GLOBAL ECOLOGICAL CONSEQUENCES OF THE 1982-83 EL NINO-SOUTHERN OSCILLATION
FURTHER TITLES IN THIS SERIES 1 J L MERO T...
10 downloads
762 Views
29MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
GLOBAL ECOLOGICAL CONSEQUENCES OF THE 1982-83 EL NINO-SOUTHERN OSCILLATION
FURTHER TITLES IN THIS SERIES 1 J L MERO T t E MINERAL RESOURCES OF THE SEA 2 L M FOMIN THE DYNAMIC METHOD IN OCEANOGRAPHY 3 E J F WOOD MICROBIOLOGY OF OCEANS AND ESTUARIES 4 G NEUMANN OCEAN CURRENTS 5 N G JERLOV OPTICAL OCEANOGRAPHY 6 V VACOUIER GEOMAGNETISM IN MARINE GEOLOGY 7 W J WALLACE THE DEVELOPMENTS OF THE CHLORINITY/ SALINITY CONCEPT IN OCEANOGRAPHY 8 E LlSlTZlN SEA LEVEL CHANGES 9 R H PARKER THE STUDY OF BENTHIC COMMUNITIES 10 J C J NIHOUL (Editor) MODELLING OF MARINE SYSTEMS 11 01 MAMAYEV TEMPERATURE SALINITY ANALYSIS OF WORLD OCEAN WATERS 12 E J FERGUSON WOOD and R E JOHANNES TROPICAL MARINE POLLUTION 13 E STEEMANN NIELSEN MARINE PHOTOSYNTHESIS 14 N G JERLOV MARINE OPTICS 15 G P GLASBY MARINE MANGANESE DEPOSITS 16 V M KAMENKOVICH FUNDAMENTALS OF OCEAN DYNAMICS 17 R.A.GEYER SUBMERSIBLES AND THEIR USE IN OCEANOGRAPHY AND OCEAN ENGINEERING 18 J.W. CARUTHERS FUNDAMENTALS OF MARINE ACOUSTICS 19 J.C.J. NIHOUL (Editor) BOTTOM TURBULENCE 2 0 P.H. LEBLOND and L.A. MYSAK WAVES IN THE OCEAN 2 1 C C VON DER BORCH (Editor) SYNTHESIS OF DEEP-SEA DRILLING RESULTS IN THE INDIAN OCEAN 2 2 P DEHLINGER MARINE GRAVITY 23 J C J NIHOUL (Editor) HYDRODYNAMICS OF ESTUARIES AND FJORDS 24 F T BANNER, M B COLLINS and K S MASSIE (Editors) THE NORTH-WEST EUROPEAN SHELF SEAS: THE SEA BED AND THE SEA IN MOTION 25 J.C.J. NIHOUL (Editor) MARINE FORECASTING 26 H.G. RAMMING and 2. KOWALIK NUMERICAL MODELLING MARINE HYDRODYNAMICS 27 R.A. GEYER (Editor) MARINE ENVIRONMENTALPOLLUTION 28 J.C.J. NIHOUL (Editor) MARINE TURBULENCE 29 M. M . WALDICHUK. G.B. KULLENBERG and M J ORREN (Editors) MARINE POLLUTANT TRANSFER PROCESSES 3 0 A VOlPlOIEditor) THE BALTIC SEA ’ 3 1 E.K. DUURSMA and R. DAWSON (Editors) MARINE ORGANIC CHEMISTRY 32 J.C.J. NIHOUL (Editor) ECOHYDRODYNAMICS 33 R HEKlNlAN PETROLOGY OF THE OCEAN FLOOR
3 4 J.C.J. NIHOUL (Editor) HYDRODYNAMICS OF SEMI-ENCLOSED SEAS 3 5 B. JOHNS (Editor) PHYSICAL OCEANOGRAPHY OF COASTAL AND SHELF SEAS 3 6 J.C.J. NIHOUL (Editor) HYDRODYNAMICS OF THE EQUATORIAL OCEAN 3 7 W . LANGERAAR SURVEYING AND CHARTING OF THE SEAS _3 8 -J C NlHOUL (Editor) REMOTE SENSING OF SHELF SEA HYDRODYNAMICS 3 9 -T ICHIYE (Editor) OCEAN HYDRODYNAMICS OF THE JAPAN AND EAST CHINA SEAS 40 J.C.J. NIHOUL (Editor) COUPLED OCEAN-ATMOSPHERE MODELS 4 1 H. KUNZEDORF (Editor) MARINE MINERAL EXPLORATION 4 2 J.C.J. NIHOUL (Editor) MARINE INTERFACES ECOHYDRODYNAMICS 4 3 P. LASSERRE and J.M. MARTIN (Editors) BIOGEOCHEMICAL PROCESSES AT THE LANDSEA BOUNDARY 4 4 I.P. MARTINI (Editor) CANADIAN INLAND SEAS 4 5 J.C.J. NIHOUL and B.M. JAMART (Editors) THREE-DIMENSIONALMODELS OF MARINE AND ESTUARIN DYNAMICS 4 6 J.C.J. NIHOUL and B.M. JAMART (Editors) SMALL-SCALE TURBULENCE AND MIXING IN THE OCEAN 47 M.R. LANDRY and B.M. HICKEY (Editors) COASTAL OCEANOGRAPHY OF WASHINGTON AND OREGON 4 8 S.R. MASSEL HYDRODYNAMICS OF COASTAL ZONES 49 V.C. LAKHAN and A.S. TRENHAILE (Editors) APPLICATIONS IN COASTAL MODELING 5 0 J.C.J. NIHOUL and B.M. JAMART (Editors) MESOSCALE IN GEOPHYSICALTURBULENCE SYNOPTIC COHERENT STRUCTURES 5 1 G.P. GLASBY (Editor) ANTARCTIC SECTOR OF THE PACIFIC
5
~
Elsevier Oceanography Series, 52
GLOBAL ECOLOGICAL CONSEQUENCES OF THE 1982-83 EL NINO-SOUTHERN OSCILLATI0N Edited by
P.W. GLYNN University of Miami Rosenstiel School of Marine and Atmospheric Science Miami, Florida, U.S.A.
ELSEVIER Amsterdam - Oxford - New York -Tokyo
1990
ELSEVIER SCIENCE PUBLISHERS B.V. Sara Burgerhartstraat 25 P.O. Box 2 1 1 , 1000 AE Amsterdam, The Netherlands Distributors for the United States and Canada:
ELSEVIER SCIENCE PUBLISHING COMPANY INC 655, Avenue of the Americas New York, NY 10010, U.S.A.
ISBN 0-444-88303-7
0Elsevier Science Publishers B.V., 1990 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science Publishers B.V./ Physical Sciences & Engineering Division, P.O. Box 330, 1000 AH Amsterdam, The Netherlands. Special regulationsfor readers in the USA -This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside of the USA, should be referred t o the publisher. No responsibility is assumed by the Publisher for any injury and/or damage t o persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This book is printed on acid-free paper. Printed in The Netherlands
Satellite infrared sea surface temperatures during the peak of the 1982-83 El NiiioSouthern Oscillation (above) and one year later (below). (Courtesy of R. Legeckis, National Oceanic and Atmospheric Administration.)
This Page Intentionally Left Blank
vii
PREFACE AND ACKNOWLEDGEMENTS
As originally perceived, El Niiio referred to the warm current that sets southward each year off the coasts of southern Ecuador and northern Peru. At times, the El Niiio event is particularly strong and its influence may extend over much of the tropical and subtropical eastern Pacific region. The El Niiio is a local manifestation of the El Niiio Southern Oscillation (ENSO), a largescale dynamic interaction between the worlds major low-latitude atmospheric pressure centers and basin-wide thermocline/nutricline depths across the Pacific and Indian Oceans. Some of the more obvious effects of El Niiio in the eastern Pacific are (a) anomalous sea surface warming, (b) reduced upwelling or upwelling of nutrient-poor waters, (c) a marked decline in primary productivity and fisheries stocks, (d) intensified storms with higher sea levels, and (e) high rainfall with frequent flooding. These events usually begin soon after Christmas or near Epiphany, Three King's Day (late December to early January), hence the epithets El Niiio (Christ Child), "La Comente El Niiio", and "Los Dias de El Niiio". The 1982-83 El Niiio was exceptionally severe, and was probably the strongest warming of the equatorial Pacific Ocean to occur during this century and perhaps for several centuries before that. Not only was the Ocean warming intense, but it spread over large parts of the Pacific Ocean, penetrated to greater depths than usual, and lasted longer than many previously recorded El Niiio events. Beyond the unprecedented severity of the 1982-83 ENSO, new kinds of global disturbances to marine biota were observed during this period. For example, coral bleaching (a sudden whitening of corals because of the loss of endosymbiotic algae) and widespread coral mortality occurred in numerous areas that experienced ENSO related perturbations: (a) in the equatorial eastern Pacific, including the Galapagos Islands, (b) in the central and western Pacific, (c) in Indonesia, (d) in the Persian (Arabian) Gulf, and (e) in the tropical western Atlantic. Many eastern Pacific coral reefs that had experienced nearly uninterrupted growth for several centuries (until 1983) were devastated (95% - 98% mortality) by the initial impact of this disturbance and are now undergoing extensive bioerosion and little or no recovery. The marked deepening of the thermocline and the resulting trophic impoverishment of surface waters adversely affected numerous marine species that depend upon frequent nutrient replenishment and, directly or indirectly, on algal productivity. The consequent depletion of the plant food base resulted in significant reduction in stocks of zooplankton, bait fish, and squid. This led to a mass migration and near-total reproductive failure of marine birds at Christmas Island. Numerous species faced similar food shortages in the Galapagos Islands and along the mainland coast of Ecuador, Peru and Chile: albatrosses, boobies, swallow-tailed gulls, penguins, cormorants, marine iguanas, fur seals, and sea lions. The abnormally high sea levels and rough seas that accompanied El Niiio in the Galapagos Islands made feeding difficult for certain species that graze near shore, such as marine iguanas. Heavy winter seas uprooted giant kelps along the coast of southern California and the correspondingly high El Niiio sea temperatures during the following summer interfered with the reproduction and recruitment of kelps.
viii
Terrestrial species suffered from drought in Central American rainforests, and at the same time flooding occurred along the west coast of South America. Reduced rainfall extended to western North America where drought conditions were recorded in tree rings. Severe drought conditions in Indonesia spawned forest fires that resulted in extreme damage to the rainforest habitat. In contrast, heavy rainfall along much of the Peruvian coast had a beneficial effect on the flora of the Peruvian "lomas", resulting in a brief flushing of flowering plants. Because of the numerous species affected and its global impact, it is probably fair to include severe ENSO events among the greatest natural perturbations known on our planet. Certainly the 1982-83 ENSO is the most severe on record and comes at a time when attention is being focused on the long-term impact of changes in global climate. ENSO events dramatically show the atmosphere, ocean and land links, and how small changes in sea surface temperatures can alter global climate patterns affecting a vast array of the worlds biota. This recent global disturbance underlines the intricate physical and biotic connections in the biosphere and the fragility of many tropical ecosystems to climatic disturbances. Our efforts in this volume provide detailed documentation of how a large magnitude ENSO event disrupted biotic communities. These observations may allow us to evaluate future changes that result from general warming trends if the Antarctic ice-cap continues to decrease, sea level rise continues or other broad meteorological/oceanographicphenomena occur. Some of the titles of the contributions in this volume denote the occurrence of the ENSO event through 1984 since physical effects and biological responses were often observed at some locations later than 1983. Furthermore, some of the disturbances observed from 1982 to 1984 have continued for several years following the initial disturbance period. A seven year lag in reporting these findings was common, in large part purposeful, in order to assess the effects of the disturbance, as well as post-ENS0 (secondary) disturbances and recovery processes. Emphasis in this volume is placed on disturbances to (a) near-shore populations, (b) benthic communities, especially coral reefs, (c) extratropical regions, and (d) terrestrial communities, topics not yet addressed and integrated in a comprehensive treatment of the subject. For sources emphasizing particular areas affected in 1982-83, and for information on related topics, such as meteorology, physical oceanography, zooplankton, fisheries, and El Nifio in the ancient record, the reader is referred to the references listed at the end of the preface. The contributors to this volume were selected on the basis of their research involvement with the 1982-83 ENSO and expertise in their respective fields of study. All authors were encouraged to present original observations and data to support the topics under discussion. I am grateful to J. A. Brady, J. Espinosa, and M. S. Hart, Meteorological and Hydrographic Branch, Panama Canal Commission, for providing data and insight into the nature of Central American weather systems. D. Heuer, M. Brinkley and R. Suarez, Print and Photo Service, Rosenstiel School of Marine and Atmospheric Science, assisted in numerous ways to prepare camera-ready copies of manuscripts for printing. All contributions were reviewed by at least two anonymous referees and accepted papers were revised before the final printing. In cases involving controversial issues, I have allowed the authors an opportunity to express their views, therefore the conclusions of the various
IX
contributions are not necessarily in total agreement. I am very grateful for the time and care offered by the following who helped in the review process: William M. Balch, Edward A. Boyle, Larry E. Brand, Lee. E. Branscome, K. T. Briggs, Barbara E. Brown, William Burger, F. Chavez, Anthony G. Coates, Laura E. Conkey, Paul K. Dayton, Richard E. Dodge, C. Mark Eakin, J. R. Ehleringer, N. M. Ehrhardt, David B. Enfield, Joshua S. Feingold, A. Gentry, Robert N. Ginsburg, M. P. Harris, Mark E. Hay, Paul L. Jokiel, Cadi Katzir, Gary P. Klinkhammer, Merlin P. Lawson, Egbert G. Leigh Jr., Harris A. Lessios, Ian
G. Macintyre, Harold A. Mooney, William A. Newman, Daniel K. Odell, Donald B. Olson, Richard L. Phipps, Donald C. Potts, Joseph Powers, Stanley A. Rand, Ralph Schreiber, Stephen V. Smith, Steven M. Stanley, Tod F. Stuessy, Peter K. Swart, Alina M. Szmant, Mia J. Tegner, Fritz Trillmich, Geerat J. Vemeij, Gerard M. Wellington, Dagmar I. Werner, Henk Wolda, Klaus Wyrtki, and two anonymous reviewers. Nicholas Polunin and Robin Pellew offered encouragement and helpful advice during the formative stages of this project. Robert L. Goodman and Martin Tanke of Elsevier Science Publishers kindly provided the technical guidance needed to produce the contributions in this volume. I am pleased to acknowledge the help of Symma Finn who skillfully managed editorial matters and prepared the final camera-ready versions of several contributions. Kay K. Hale and Helen D. Albertson offered expert assistance with various literature problems. Thanks are also extended to June M. Eakin, Corell L. Lundy, Lois Reid, Nora I. Rodriguez and David B. Smith for help in the preparation of several manuscripts. I am especially appreciative of the labors of Jorge Cortes who ferreted out numerous errors in the final stages of preparation, and of June Eakin and Joshua Feingold who assembled the volume indexes. Selected references with major emphasis on the 1982-83 El Niiio event: Bibliografia Sobre El Fendmeno de El Niiio Desde 1891 a 1985, 1985. J. Mariategui, A. Ch. de Vildoso and J. VClez, Bol. Inst. Mar Peni, Spec. Vol., Callao, Peni, 136 pp. Boletin ERFEN (El Estudio Regional del Fendmeno de El Niiio), 1982-to date. R. Jordan (ed.), Comisidn Permanente del Pacifico Sur, Bogoti, Colombia. Ciencia, Tecnologia y Agresidn Ambiental: El Fendmeno El Niiio, 1985. M. Vegas (ed.), CONCYTEC Press, Lima, Peni, 692 pp. El Niiio, 1984. Oceanus, 27(2), Woods Hole Oceanographic Institution, Mass. 84 pp. El Niiio: An AGU Chapman Conference, 1987. J. Geophys. Res., 92(C13), 14,187-14,479. El Niiio en Las Islas GalBpagos: El Evento de 1982-1983, 1985. G. Robinson and E. M. del Pino (eds.), Quito, Ecuador: Fundacidn Charles Darwin para las Mas Galipagos, 534 pp. El Niiio North: Niiio Effects in the Eastern Subarctic Pacific Ocean, 1985. W. S. Wooster and D. L. Fluharty, Washington Sea Grant Program, University of Washington, Seattle, 312 pp. "El Niiio", Su Impacto en la Fauna Marina, 1985. W. E. Amtz, A. Landa and J. Tarazona (eds), Bol. Inst. Mar Perh, Spec. Vol., Callao, Peni, 222 pp. Taller Nacional Fenomeno El Niiio 1982-83, 1985. lnstituto de Fomento Pesquero, Invest. Pesq., 32, 254 pp., Santiago, Chile. Taller Sobre El Fenomeno de El Niiio 1982-83, 1984. Comisidn Permanente del Pacifico Sur. Rev. Pacifico Sur, 15,423 pp., Quito, Ecuador. Tropical Ocean-Atmosphere Newsletter, 1983 and 1984. Special issues 1-111, no's. 16.21, and 28.
November 1989 Peter Glynn
This Page Intentionally Left Blank
xi CONTENTS Preface and Acknowledgements ....................................................................................................... List of Contributors .......................................................................................................................... PHYSICAL ASPECTS OF THE EL NIRO EVENT OF 1982-1983 D.V. Hansen ...................................................................................................................................
vii xix
1
Introduction ....................................................................................................................... The Global View Some Historical Perspective .............................................................................................. Development of the Event of 1982-1983 ............................. References ..........................................................................................................................
3 4 6 19
NUTRIENTS AND PRODUCTIVITY DURING THE 1982/83 EL NIRO R.T. Barber and J.E. Kogelschatz ..................................................................................................
2I
Introduction ........... ........ ........................................................................ ............................ The Enso Cycle ................................................................................................................... The Basinwide Setting Western Pacific .................................................................................................................. Eastern Pacific Normal Conditions. 5.1 The Equatorial Region .......................................................................................... 5.2 The Coastal Region ............................................................................................... Eastern Pacific Anomalous Conditions 6.1 The Equatorial Region .......................................................................................... 6.2 The Coastal Region Productivity Effects of El Niiio .......................................................................................... 7.1 The Equatorial Region .......................................................................................... 7.2 The Coastal Region .................... Conclusions ......... ......... ........... ................... ........... ............................................. .......... ...... References
21 22 26 28 30 30 32 35 35 39 43 43 43 49 50
CORAL MORTALITY AND DISTURBANCES TO CORAL REEFS IN THE TROPICAL EASTERN PACIFIC P.W. Glynn ....................................................................................................................................
55
Introduction ....................................................................................................................... Coral Bleaching, Mortality and Environmental Correlates ................ 2.1 Onset of the 1982-83 Disturbance .... .................................... 2.2 Extent ofthe 1982-83 Disturbance ...................................................................... 2.3 Condition of Bleached Corals ...... 2.4 Coral Bleaching and Sea Water Temperature Extremes ....................................... 2.5 Reliability of Sea Surface Temperature Observations .......................................... 2.6 Sea Warming and the Timing of Coral Bleaching 2.7 Further Evidence Implicating Sea Warming as th 2.8 Spatial and Temporal Occurrences of El Niiio Warming Events ......................... 2.9 Coral Bleaching and Mortality During Cooling Episodes 2.10 Non-thermal Stressors and Coral Bleaching ......................................................... Community Effects ............................................................................................................ 3. I Immediate Effects ...... 3.2 Long-term Effects ................................................................................................. 3.2. I Coral Community Changes ................................ 3.2.2 Responses and Impacts of Corallivores ............................... .............. ..... 3.2.3 Responses and Impacts of Herbivores ....................................................
55 59 59 60 63 68 70 71 75 77 81 83 87 87 94 94 96 98
1.
2. 3. 4.
1.
2. 3. 4. 5. 6. 7.
8.
I. 2.
3.
1
xii 4. 5.
6.
Interrupted Coral Growth and Reef Framework Accumulation: Indicators of Severe Event Occurrences ................................................................................ Discussion and conclusions ........................................................ 5.1 Sea Warming as the 5.2 El NiAo 1982-83 Compared with Other Disturbance 5.3 Prospects for Coral Reef Recove ry................ 5.4 Predicted Effects of Greenhouse Global Warming Summary ............................................................................. References ..........................................................................................................................
i02 107
1 16 I 17
THE EFFECTS OF THE EL NIRO/SOUTHERN OSCILLATION ON THE DISPERSAL OF CORALS AND OTHER MARINE ORGANISMS 127 R.H. Richmond .......................................................................................................................... 1..
I.
Introduction .................. 2.1 2.2
3. 4.
5.
..............................................
....
The North Equatorial Countercurrent ................................................................. The North Equatorial Current ........................
2.4 The Equatorial Undercurrent Oceanic Currents During tbe 1982-83 Transport of Marine Organisms in Oce 4.1 Transport During Non-El Niii
............................ rrents .......................................................
.........................
Conclusions ............................................................... References ......................................................,...................................................................
CORAL MORTALITY OUTSIDE OF THE EASTERN PACIFIC DURING 1982-1983: RELATIONSHIP TO EL NIRO M.A. Coffroth, H.R. Lasker and J.K. Oliver .............................................................................. I.
2.
3.
4.
127 128 129 129 130 130 130 132 132 134 137 138
...... 141
........................ Introduction ................... Causes of Coral Bleaching and Mortality ............................................................. I. I ENS0 and Coral Mortality ....................................................... ...................... 2. I Eastern Pacific .......................................... .................................. 2.2 Central Pacific .. .................................. 2.3 Western Pacific. .................................................................. 2.4 Indian Ocean/A 2.5 Caribbean Sea ........................................... ......................... Detailed Case Study 3. I Extent and Timing of Bleaching ........................................................................... 3.2 Oceanographic and Meteorological Data ............................................................. (i) Seawater Temperatures ............ ................................................... (ii) Solar Radiation .................. ................................................... (iii) Wind Speed and Direction. (iv) Rainfall ........................................ (v) Interactive Effects ................................................................................... 3.3 Relationship to El NiAo .................... Detailed Case Study - San Blas Islands, Panama ............................................................... ...................... .................................. Extent and Timing of Bleaching 4. I
(iii)
Solar Radiation ..
................................
142 142 146 146 146 148 149 151 152
153 155
155 156 158
159 159 161 161 161 162 163 164 164
...
Xlll
(iv) Wind Speed and Direction ...................................................................... (v) Rainfall and Salinity .................. Relationship to El Niiio ........................................................................................
.................................................................................
165 i67 169 171 177
EL NIRO AND THE HISTORY OF EASTERN PACIFIC REEF BUILDING M.W. Colgan ................................. .................................................................
183
4.3
Introduction .................................... ................................... Background ........................................................................................................................ 2.1 Eastern Pacific -Physical Setting and Reefs ........................................ 2.2 Past Eastern Pacific Reefs .................................................................................... The 1982-1 983 El Niiio Event and Eastern Pacific Reefs Evidence for Past El Niiio Events ....,......,......................................,.................,....... 4.1 Historical and Proxy Rec .............,......,.........................,...,..... 4.2 Holocene Record, Sedimentological Evidence ..................................................... Ocean Conditions and Past El Niiio Events ...................................................................... Uwina Bay, Galhpagos Islands ............. 6.1 El Niiio Events at Urvina Bay .............................................................................. 6.1.1 Branching Corals ......................................... 6. I .2 Massive Corals ........................ 6.2 Summary Remarks ............................................................................ Discussion ........................,.................. ................ Conclusion ............................................................................................... References ...................,....,..........................,........,.,...........................
198 20 I 202 203 21 1 214 216 218 220
REEF-BUILDING CORALS AND IDENTIFICATION OF E N S 0 WARMING EPISODES E.R.M. Druffel, R.B. Dunbar, G.M. Wellington and S.A. Minnis
233
I. 2.
3. 4.
5. 6.
7. 8.
184 187 187 189 192 196 196 196
Introduction ....................................................................................................................... I. 1 Characteristics of Skeletal Growth in Symbiotic Corals I .2 Effects of Physical Factors on Coral Growth ....................................................... I .3 Effects of Physical Factors on Skeletal Chemistry and Isotopes Study Sites .......................................................................................................................... Methods ................................... .............................................. Stable Isotope Records in Corals ....................................................................................... 4.1 6I8OResults. 4.2 6I3C Results Conclusions ........................................................................................................................ References .................
233 234 235 237 238 24 I 242 243 247 249 250
TRACE ELEMENT INDICATORS OF CLIMATE VARIABILITY IN REEF-BUILDING CORALS G.T. Shen and C.L. Sanford
255
I.
2. 3. 4. 5.
I. 2. 3.
Introduction MinorandT ........................................................... Sample Sites ....................................................................................................
5.
Oceanic Markers of El Niao ............................................................................................... 5. I Upwelling in the Eastern Equatorial Pacific ........................................................ 5.2 Precipitation in the Western Tropical Pacific River Discharge and Circulation in the Caribbean Sea........................................ 5.3
255 256 260 26 1 26 I 26 1 269 272
xiv 6.
Conclusions ........................................................................................................................ References ......,......................,............................................................................................
277 278
HISTORICAL ASPECTS OF EL NIRO/SOUTHERN OSCILLATION - INFORMATION FROM TREE RINGS J.M. Lough and H.C. Fritts ......................................................................................................... ... 285
3. 4.
.................... Site selection and Sample Collection .................. 2. I Crossdating and Measuring ............... .............................................. 2.2 ..................................... 2.3 Chronology Development .................. 2.4 Climatic Reconstruction ........................................................ Tree Rings and the Southern Oscillation: An Example Application ... 3. I Teleconnection Patterns in Dendroclimatic Reconstructions 3.2 Reconstruction of an Index of the Southern Oscillation ...................... ....................................................... Future Directions ....................... 4. I Northern Hemisphere Extra-tropics ......................... 4.2 Tropical Regions ...................... .......................................................... ,.................... .................................
285 287 288 289 289 290 29 I 29 1 298 307 307 309 310 312 314
EFFECTS OF EL NIRO 1982-83 ON BENTHOS, FISH AND FISHERIES OFF THE SOUTH AMERICAN PACIFIC COAST 323 W.E. Arntz and J. Tarazona .......................................................................................................... I. 2. 3.
4.
5.
Introduction ................................................................... Principal Abiotic Changes Induced by EN 1982-83 ......................................................... The Pelagic Subsystem ......................................... 3. I Phyto- and Zooplankton ....................................................................................... ............................................. 3.2 Pelagic Fish ........................ 3.3 Pelagic Fisheries ......................... .............. .................................... .... ............................ .................................. The Benthic Subsystem 4.1 Macrobenthos .................................................... 4.1.2 Soft Bottom 4.2 Exploited Inverte 4.3 Demersal and Co Conclusions ................................................
........................................................ ..................
324 325 329 329 330 333 336 336 336 339 342 347 350 353
EFFECTS OF THE 1982-83 EL NIRO-SOUTHERN OSCILLATION EVENT ON MARINE IGUANA (AMBLYRHYNCHUS CRZSTATUS BELL, 1825) POPULATIONS ON GALAPAGOS W.A. Laurie ................................................................................................................................... 361 2.
Study Area ..........................................................................................................................
.......,...................
...............................................................................................
3.2 3.3 3.4 3.5 3.6
Capturing and Marking Iguanas ........................................................................... Observations of Reproductive Behaviour ................ ....................... ,............... Measurements of Iguanas ..................................................................................... Growth Rates ........................................................................................................ Survival Rates.
36 I 364 364 364 364 365 365 366 366
xv
4.2
.................................... .................................. ........................... Changes in Algal Flora ..............................................
5.2 5.3 5.4
....................... t Competition for Food .................................. Growth Rates and Age at First Reproduction ................................. ........,........................ Dominant Cohorts ........................................ .................................... .......... Costs of Breeding
Results .............................
I..
.................. ..............................................
References.. .......................
THE GULF OF PANAMA AND EL NIRO EVENTS THE FATE OF TWO REFUGEE BOOBIES FROM THE 1982-83 EVENT N.G. Smith .....................................................................................................................................
...........................
2. 3. 4. 5.
.........................
t
....................
Natural History of the Boobies ....................................... ................. ......................... The Occurrence ........................ El Niiio Events and the Gulf of Panama ................................................ ........,..................................... Concluding Remarks .............................
...................................,........................................
SEABIRDS AND THE 1982-1984 EL NIRO/SOUTHERN OSCILLATION ............................................................... D.C. Duffy ......................
366 366 367 369 37 I 372 373 375 311 371 378 378 378 319
381 38 I 38 I 383 389 391 392 395
............... ........................
.............................................................................. ............................
395 396 396 396 391 398 400 40 1 402 403 403 404 405 406 408 410
EL NIRO EFFECT ON SOUTH AMERICAN PINNIPED SPECIES D. Limberger ..............................................................................................
4 I7
Introduction ......................................................................................... Galapagos Fur Seal. .............................................................................................
419
How El Niiio Affected the Galapagos Fur Seal .................. After Effects of El Niiio, During the Reproductive Season of 1983.....................
422
1.
Introduction
............................
..............................................................
....................
......................
.................
...........................
...........
References ....................................................
1.
2.
2.2 2.3
xvi 3.
4. 5.
South American Fur Seal .......................................................................... 3. I General Information ................................................................... ........ 3.2 El Niiio’s Effect on the South American Fur Seal at Punta San J How the Sea Lions in Galhpagos and Punta San Juan Survived the El Niiio Event ......... .................................... 4. I The Galapagos Sea Lion ....................... 4.2 South American Se Summary and Conclusions ..................................... ................................ References ...........................................................................
424 424 425 426 426 427 428 430
BOTTOMS BENEATH TROUBLED WATERS: BENTHIC IMPACTS OF THE 1982-1984 EL NIRO IN THE TEMPERATE ZONE P.K. Dayton and M.J. Tegner ........................................................................................................ 433
I. 2. 3.
4.
5. 6.
Introduction .............................................................................................. 1.1 Nort re Temperate El Niiios: A summary of Physical Factors ..... Biological Effects of the 1982-84 El Niiio on Temperate Pelagic Ecosystems ................... ............................. 2. I Northeastern Pacific ....... 2.2 California ............................................ .................................................................. ENSO Effects on Kelp Forests .................................................................... .......... Effects of the Storms ............................ 3. I Effects of the Warm Water ................................................................................... 3.2 .................................................. 3.3 Other California Kelp Habitats ENSO Impacts on Kelp Forest Animals ............................................. 4. I Sea Urchins ........................................................................................................... .................. 4.2 Abalones ..
.............................................. 4.4 Kelp Forest Fishes Non-kelp Benthic Systems ............................................................ 5.2 ENSO Effects on Intertidal Populations Discussion .......................................................................................................................... .......................................... 6. I Other ENSOs 6.1.2 6. I .3
7.
Japan ....................................................................................................... South Africa ...............................
Conclusions References ......................................... .... . ... . ..
........................................... ..................................
THE IMPACT OF THE “EL NIRO” DROUGHT OF 1982-83 ON A PANAMANIAN SEMIDECIDUOUS FOREST E.G. Leigh, Jr., D.M. Windsor, A. Stanley Rand and R.B. Foster ................................................ I. 2. 3.
Introduction ....................................................................................................................... The Severity of the El Niiio Drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ The Impact Upon Plants of the El Niiio Drought 3. I Signs of Stress ....................................................................................................... 3.2 Mortalit .............................................................................. ..........
..... 4. 5.
3.4 Long-term Effects on the Forest ........................................................................... The El Niiio Drought and Animal Populations ................................................................. Concluding Remarks .......................................................................................................... References ..........................................................................................................................
433 434 436 436 438 44 1 442 443 447 450 450 452 453 454 456 456 458 460 460 460 460 46 I 462 464 466
473 473 475 476 476 476 478 479 480 482 484
xvii THE BOTANICAL RESPONSE ON THE ATACAMA AND PERUVIAN DESERT FLORAS TO THE 1982-83 EL NIRO EVENT M.O. Dillon and P.W. Rundel ....................................................................................................... 487 I. 2. 3. 4. 5. 6. 7
Introduction .................................................... Lomas Form ............................................ Coastal Climate .................................................................................................................. Impact of Former Intense El Niiio Events. 1982-83 El NiRo Event ...................................................................................................... Botanical Response to 1982-83 El Niiio Event 6 .I Coastal Peru and Northern Chile ......................................................................... 6.2 Galapagos Islands ................................................................................................. Conclusions. ............................ ....................................................... References ..........................................................................................................................
487 488 49 I 493 494 495 495 498 50 I 503
AN ECOLOGICAL CRISIS IN AN EVOLUTIONARY CONTEXT: EL NIRO IN THE EASTERN PACIFIC
......................................... I. 2. 3.
505
Introduction ...... ENS0 as a Model For Extinction Events ........................................................................... Extinction in the Eastern Pacific ...........................
506
.......................................
51 5
INDICES Subject Index ................................................................................................................. Systematic Index .................................................................... Geographic Index ...............................................................................................................
554
This Page Intentionally Left Blank
XIX
LIST OF CONTRIBUTORS
W. E. ARNTZ
Alfred-Wegener-Institut fur Polar- und Meeresforschung, Columbusstrasse, D-2850 Bremerhaven, Federal Republic of Germany
R. T. BARBER
Monterey Bay Aquarium Research Institute, Pacific Grove, California 93950
M. A. COFFROTH
Division of Marine Biology and Fisheries (formerly Division of Biology and Living Resources), Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149; present address: Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York 14260
M. W. COLGAN
Earth Sciences Board, University of California, Santa Cruz, California 95064; present address: Department of Geology, College of Charleston, Charleston, South Carolina 29424
P. K. DAYTON
Scripps Institution of Oceanography, A-001, La Jolla, California 92093
M. 0.DILLON
Department of Botany, Field Museum of Natural History, Chicago, Illinois 60605-2496
E. R. M. DRUFFEL
Department of Chemistry, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543
D. C. DUFFY
Institute of Ecology, University of Georgia, Athens, Georgia 30602
R. B. DUNBAR
Department of Geology and Geophysics, and Earth Systems Institute, Rice University, Houston, Texas 77251-1892
R. B. FOSTER
Department of Botany, Field Museum of Natural History, Chicago, Illinois 60605-2496 (mailing address); and Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama.
H. C. FRITTS
Laboratory of Tree-Ring Research, University of Arizona, Tucson, Arizona 85721
P. W. GLYNN
Division of Marine Biology and Fisheries, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149-1098
D. V. HANSEN
NOAA (National Oceanic and Amiospheric Administration)/Atlantic Oceanographic and Meteorological Laboratory, 4301 Rickenbacker Causeway, Miami, Florida 33149; and Cooperative Institute for Marine and Atmospheric Studies, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33 149
J. E. KOGELSCHATZ
Monterey Bay Aquarium Research Institute, Pacific Grove, California 93950
H. R. LASKER
Department of Biological Sciences, State University of New York at Buffalo, Buffalo, New York 14260
xx
W. A. LAURIE
Max-Planck-Institut fur Verhaltensphysiologie, 8131, Seewiesen, West Germany; present address: Large Animal Research Group, Department of Zoology, University of Cambridge, Cambridge CB2 3EJ, United Kingdom
E. G. LEIGH, JR.
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama; mailing address: Smithsonian Tropical Research Institute, APO Miami 34002-0011 USA
D. LIMBERGER
13, Frayne Road, Ashton Gate, Bristol BS3 IRU, United Kingdom
J. M. LOUGH
Australian Institute of Marine Science, PMB No. 3, Townsville M. C., Queensland, Australia
S. A. MINNIS
Department of Geology and Geophysics, and Earth Systems Institute, Rice University, Houston, Texas 77251-1892
J. K. OLIVER
Sir George Fisher Centre for Tropical Marine Studies, James Cook University of North Queensland, Townsville, Queensland 48 11, Australia
A. S. RAND
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama; mailing address: Smithsonian Tropical Research Institute, APO Miami 34002-001 1 USA
R. H. RICHMOND
Marine Laboratory, University of Guam, UOG Station, Mangilao, Guam 96923 USA
P. W. RUNDEL
Laboratory of Biomedical and Environmental Sciences, and Department of Biology, University of California, Los Angeles, California 90024
C. L. SANFORD
Lamont-Doherty Geological Observatory of Columbia University, Palisades, New York 10964
G. T. SHEN
Lamont-Doherty Geological Observatory of Columbia University, Palisades, New York 10964; present address: School of Oceanography, WB-10, University of Washington, Seattle, Washington 98195
N. G. SMITH
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama; mailing address: Smithsonian Tropical Research Institute, APO Miami 34002-001 1 USA
J. TARAZONA
Grupo DePSEA, Facultad de Ciencias Biologicas, Universidad Nacional Mayor de San Marcos, Apartado 1898, Lima 1000, Peru
M. J. TEGNER
Scripps Institution of Oceanography, A-001, La Jolla, California 92093
G. J. VERMEIJ
Department of Geology, University of California, Davis, California 95616
G. M. WELLINGTON
Department of Biology, University of Houston, Houston, Texas 77204-5513
D. M. WINDSOR
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama; mailing address: Smithsonian Tropical Research Institute, APO Miami 34002-001 1 USA
1
PHYSIC?&
ASPECTS OF THE EL NIfiO EVENT OF 1982-1983
DONALD V. HANSEN NOAA/Atlantic Oceanographic and Meteorological Lab., 4301 Rickenbacker Causeway, Miami, Florida 33149 (USA), and Cooperative Institute for Marine and Atmospheric Studies, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149 (USA)
ABSTRACT Hansen, D.V., 1989. Physical aspects of the El Nifio event of 1982-1983. El Nifio events are marked by the appearance of anomalously warm ocean waters and unusual rainfall in normally arid coastal regions of Ecuador and Peru. During the past century such events have occurred at about four-year intervals on average, and nine of the events have been described as strong or very strong. In the spring of 1982 the heavy rainfall that normally characterizes the IndoPacific archipelago began to shift eastward toward the central Pacific. During the following year the region of anomalous rainfall traversed the ocean to the coast of South America, in phase with anomalous winds, currents, and sea surface temperatures. At the peak of the event in the eastern tropical Pacific, Peru and Ecuador experienced record-setting rainfall leading to flooding and avalanches, near surface ocean currents reversed from their normal direction, sea surface temperature rose t o 5 O C or more above normal, the thermocline plunged to 100 meters or more below normal, and sea level rose to nearly half a meter above normal. Upon reaching the coast, many of the oceanic perturbations propagated poleward along the continenal margins in both hemispheres, carrying the signs and effects of El Nifio to middle and high latitudes in the Pacific. The magnitude of this event made it the "event of the century" in most variables, and the event of several centuries in some. The magnitude of perturbation of the atmosphere in the tropical Pacific sector certainly carried anomalies also in distant regions of the atmosphere, and thereby secondarily in other parts of the ocean. At greater distance, however, it becomes increasingly difficult to distinguish between anomalies resulting from El Nifio and those arising from other kinds of variations of the atmospheric circulation. 1 INTRODUCTION A century or more ago, sailors in and around the port of Paita, on the northern coast of Peru, knew of a warm coastal current that flowed southward along the coast during most (Southern Hemisphere) summers. Because it usually appeared around Christmas time, Los Dias del Nifio, or the Days of the (Christ) Child, they termed it Corriente del Nifio, or El Nifio Current. This current and its warming influence are highly irregular in their occurrence. In some years they may not appear at all, but usually some of the tropical surface water that is always found north of the equator will be carried southward to the coast of Peru. At irregular intervals of a few years there is such massive intrusion of warm tropical waters into the coastal region and a change from the usual desertlike climate to moderate or heavy rainfall as to suggest that these extreme changes are something more than a particularly strong manifestation of the
2
seasonal cycle. It is these often dramatic events at several year intervals that the term El NiHo has been used to identify in recent decades. During the present century of relatively complete and reliable information, nine strong and 14 moderate events have been documented (Quinn g @., 1987). Thus, in the recent global climate regime, El Nifio events have occurred at about four-year intervals on average, and strong events have occurred about once a decade. These average intervals do not, however, imply regularity. No events occurred between 1943 and 1951, but three, including one strong event, are documented during 1939-1943. Although El Niiio is frequently thought of as an oceanic phenomenon, connection to the atmosphere is suggested by the association of unusual rainfall with El Nifio in those places where the phenomenon is most strongly manifest. Interannual variations of atmospheric patterns were documented by Sir Gilbert Walker in connection with studies of the variations of the Indian monsoons. Walker and Bliss (1932) described large-scale atmospheric pressure changes or "swings" over the southeastern Pacific and Indian Oceans that they called the Southern Oscillation. When pressure is high over the Pacific Ocean it tends to be low over the Indian Ocean, and conversely. Rainfall varies in the opposite
h
n E
W
1012-
18-
I-
La Punta/Callao (1 2"S,77"W)
1960
1965
1970 YEAR
1975
1980
Fig. 1. Historical record of atmospheric pressure variations at Tahiti and Darwin, Australia, and coastal temperature variations at Callao, Peru. Curves show 12-month running average of data.
3
direction to the air pressure. The intimate connection between these largescale pressure changes and sea surface temperature variations in the region of historic El Niao variations is illustrated in Fig. 1. The connection between the El Nifio and the Southern Oscillation as manifestations of the same large-scale, air-sea interaction process appears first to have been pointed out by Bjerknes (1969). He pointed out that the tradewind system over the tropical Pacific Ocean both sustains and is driven in part by the large-scale gradient of surface temperature, cold in the east and warm in the west. Warm water is normally driven westward by the surface winds. Warm surface waters in the west promote ascending motion and convective rainfall there, and the cold surface in the east promotes subsidence and stability in the atmosphere. It follows that if either the atmospheric or the oceanic part of this system is significantly perturbed, that perturbation is communicated to the other part and, therefore, may be sustained or amplified. 2 THE GLOBAL VIEW It is a familiar fact that tropical regions of the earth are more strongly heated by the sun than are the higher latitudes. The tropics, in fact, receive an excess of heat energy over what is locally radiated back into space, while the higher latitudes have a local deficit of radiative energy. The imbalance of radiative heating drives movements of the atmosphere and the ocean. These motions in turn carry heat from low latitudes to higher latitudes, thus maintaining the thermal regime of the earth within normal bounds, and in the process producing weather. All parts of the tropics are not the same in the process, however. Solar radiation is not particularly effective in directly heating the atmosphere. A substantial part of the solar radiation passes through the atmosphere to the surface of the earth where it is transferred to the atmosphere by several secondary processes. One of the most important of these is the evaporation of surface water and subsequent release of latent heat of evaporation when this water condenses as rainfall. Hence, those parts of the tropical atmosphere where the most convection and precipitation occurs are most strongly heated. In the normal (non-ENSO) situation the greatest tropical precipitation tends to occur over the continents, the Amazon Basin and equatorial Africa, and probably for reasons of symmetry and scaling, over the "island continent" of the IndoPacific region (Fig. 2a). These three regions of strong heating of the atmosphere may be thought of as the fire-boxes that energize the circulation of the atmosphere, and through that the ocean. Their locations establish the normal patterns of the general circulation of the atmosphere and the ocean. Probably due to the mobility and thermal properties of its ocean fraction, the island continent in the Indo-Pacific region is less effective at confining the location of precipitation than are the true tropical continents. Perturbations of the atmosphere or the ocean that lead to unusual warming in the central
4
or eastern Pacific can cause the precipitation and heating of the atmosphere to move eastward also (Fig. 2b). Furthermore, the new arrangement tends t o be self-sustaining, as pointed out by Bjerknes and described in more detail with the help of mathematical models (cf. Gill and Rasmusson, 1983). When the largescale pattern of tropical heating that drives the atmospheric circulation is changed so substantially, it is to be expected that the state and circulation of both the atmosphere and the ocean will change globally.
a.
’
Poclflc Ocean \
I
f
I
0
90E
180
9ow
b.
0
200mb
pressure
surface
-_--.
pressure
Poclflc Ocean \
I
0
90E
180
9ow
0
Fig. 2. Schematic representation of regions of strong tropical rainfall during (a) non-Nifio periods and (b) El Nifio periods. 3 SOME HISTORICAL PERSPECTIVE In-depth research on the historical El Nifio events has been published by Quinn _ et -al. (1978),Woodman (1984), and Quinn &. (1987). Information from the time prior to about 1950 is mostly inadequate to allow much more than identification of events and a general evaluation of their relative strength. Major events during the last century occurred in 1891, 1899-1900, 1911-1912, 1917, 1925-1926, 1932, 1940-1941, 1957-1958, 1972-1973, and 1982-1983. The 1891 event is relatively well described by Eguiguren (1894). Peterson (1935) gives a good account of the 1925-1926 and 1932 events, Both of the above investigators describe primarily meteorological manifestations of the events, but noted the
5
association with high ocean temperatures. Schott (1931) provided the first extensive description of the offshore character of El Nifio, based on observations of the 1925-1926 event. Woodman (1984) concluded from study of mostly rainfall and river flow data and descriptive accounts of events prior to 1925 that 1983 was by far the rainiest year in northern Peru of at least the last 200 years, and quite possibly the rainiest year of the 450 year history of the region. At the surface of the earth ( o r ocean) the changes across the tropical Pacific associated with El Nifio or ENSO warm events appear as decreased rainfall in and around Indonesia, and increased rainfall in the central and eastern Pacific Ocean and coastal regions of tropical South America. Changes in the strength and even in direction of the southeast tradewinds over the Pacific Ocean accompany the changes in the pattern of convection and rainfall. These changes can be expected to exert influences also in the more distant tropics and in higher latitudes. The atmosphere is subject to numerous other influences, however, so that the more distant changes do not appear with the same clarity as those in the tropical Pacific. In most cases they appear as statistical correlations within which counter examples occur frequently. One most investigated distant influence is the Pacific-North American pattern, or PNA, which frequently occurs in the Northern Hemisphere winter in association with ENSO warm events. The PNA is characterized by high atmosphere pressure and warm dry weather over the western half of the continent, and low pressure with unusually low temperatures over the eastern half. Yarnal and Diaz (1986) found that the PNA pattern developed in winter months during 54 percent of the ENS0 warm events during the 32 years from 1947 to 1979. However, it also occurred during 22 percent of the non-warm event winters, which comprised more than half of the occurrences of the PNA pattern, but not at all during the cold event antithesis of El Nifio. A pattern opposite in sense to the PNA was found to occur during half of the ENSO cold events, and at least once during a warm event. Many attempts have been made to explain this variety of experience in regards to the more distant aspects of the El Nifio in the tropical Pacific. Pan and Oort (1983), for instance, found evidence that details of the sea surface temperature anomaly pattern are critically important for distant influences. In particular, they found that variations of global wind and atmospheric temperature are most closely related to sea surface temperature anomalies near longitude 130 degrees west. Hamilton (1988), on the other hand, provides evidence that the PNA pattern tends to occur in association with El Niiio when sea surface temperature in the far western tropical Pacific is also anomalously warm, or at least not overly cold. Some of the distant correlates of El Nifio are surprising. Andrade and Sellers (1988), for example, found a positive correlation between El Nifio and rainfall in Arizona and western New Mexico during spring and a u t m , but not during winter or summer. The explanation seems to be that El Nifio is but one of many influences upon the global atmosphere. During any particular event the
other processes may reinforce or obliterate the distant influence of El Nifio. Climate anomalies observed in other ocean basins during El Nifio can result from anomalous surface wind in those regions. The wind anomalies may or may not be "caused" by the El Nifio. Because the ENS0 event of 1982-1983 was extraordinarily strong, its influence on distant regions may have been uncommonly strong. It should not be assumed, however, that all unusual aspects observed during this time are due to El Nifio. The PNA pattern, for example, did not develop in this winter. It has been remarked in several reports that the El Nifio event of 1982-1983 was unusual. It also has become popular to point out that no two El Nifio events are the same, but a useful description of a typical warm event was provided by Rasmusson and Carpenter (1982) on the basis of data from the seven most significant warm episodes between 1950 and 1973. They superposed data from these seven events by month of the years of onset and maximum development. One of the most familiar aspects of El Nifio is the anomalous warming of the ocean along the coast of Peru. The evolution of ocean temperature anomalies in this region during 1982-1983 is shown in Fig. 3 in comparison to that of the preceding seven events. In this measure, the event of 1982-1983 began several months earlier than the envelope of prior events, and had exceptional amplitude, but followed the pattern of previous events in the timing of its maximum development. Perhaps the major difference in timing compared to the previous events was the appearance of a secondary maximum about half a year in advance of the main peak and the rapid return to normal immediately following the main peak. Most previous events have been characterized by a rapid development of the major maximum, and several have contained a secondary maximum about half a year following the time of maximum anomaly. Some reporters (cf. Taft, 1985; Nicholls, 1987) have chosen to center the chronology of the event of 1982-1983 on 1982 rather than 1983, which brings the initial peak of this event into phase with the secondary peak in some previous events. The best overall comparison between dissimilar events appears to be as shown in Fig. 3, however. 4 DEVELOPMENT OF THE EVENT OF 1982-1983
In addition to its magnitude, the El Nifio event of 1982-1983 was exceptional in that it was the best observed event in history. Information on earlier events was mostly limited to reports from merchant ships, coastal sea level and temperature observations, as well as terrestrial rainfall. During the 1982-1983 event, there were in addition, information on oceanic precipitation, winds, and sea surface temperature from weather satellites. Also, there were in progress some scientific investigations of various aspects of the air-sea interaction, processes associated with El Nifio. In connection with these scientific investigations, observations being made from merchant vessels had been augmented by observations of subsurface temperatures.
7
5
4
3
2
0
-1
JAN
JUL
JAN
JUL
JAN
JUL
JAN
Fig. 3 . Evolution of SST anomalies off the coast of northern Peru during several prior events (thin lines) and during events of 1982-1983 (heavy line). Fig. 4 shows the development of the SST anomaly across the equatorial band of the Pacific Ocean during the same three-year period spanning the 1982-1983 event as shown in Fig. 3. In these data it appears that the first peak of SST anomalies associated with the event developed almost simultaneously everywhere east of the international dateline, and had their maximum development between longitudes llOnW and 14OOW. The principal peak of the SST anomaly shown in Fig. 3 developed near the coast and progressed westward, more or less in accord with
8
the description by Rasmusson and Carpenter (1982). Although this second peak was the strongest of the event at the coast, it is not recognizable westward of about 120"W longitude. Thus, while an observer near the coast would say that the peak of the event was in May-June 1983, an offshore observer of SST would say that the maximum anomaly had passed during December 1982. It is clear that the event is more complicated than a simple movement of anomalies on or offshore. If only the more sparse observations of prior events had been made of the 1982-1983 event, much of the early weaker but spatially more extensive SST anomaly maximum might have been missed. It is presently still not clear whether the development of this event is very unusual. Its amplitude, of course, was extraordinary.
Fig. 4. Time-longitude plot of average SST anomaly in S0N-5"S band across the tropical Pacific. Contours are in decidegrees C. (From CAC, 1984.) As mentioned earlier, the warming of surface water in the central and eastern tropical Pacific tends to draw atmospheric convection and rainfall away from its normal position in the far western Pacific. The associated cloudiness is evident in satellite cloud pictures, and is quantified and archived as outgoing longwave radiation (OLR). Cloud tops are colder than the underlying earth surface and atmosphere and, therefore, they emit less radiative energy. Anomalously cloudy areas are revealed by negative OLR anomalies. Fig. 5 shows the evolution of convection and rainfall across the tropical Pacific during the event of 1982-1983. Early 1982 was near normal, with slightly more than normal rainfall (negative OLR anomalies) in the far western Pacific, and negligible anomalies elsewhere. During the second half of 1982 the western Pacific became
9
Fig. 5. Time-longitude plot of outgoing longwave radiation anomaly (W m-*) in 5ON-5OS band across the tropical Pacific. (From CAC, 1984.) anomalously dry as the region of cloudiness and precipitation moved eastward until by January-March 1983 the entire region west of the international date1 ne was unusually dry and all to the east was rainy. The largest OLR anomalies occurred in February near 150°W. Somewhat weaker, but still large, OLR anomalies continued moving eastward, reaching the coast in April-May. During the summer the rainfall distribution rapidly returned to normal, even became drier than normal near the dateline. The global distribution of tropical and subtropical cloudiness and rainfall during the Northern Hemisphere spring (March-April-May) of 1983 is shown in Fig. 6. Regions of heavy rainfall are manifest as areas of minimum OLR over the Amazon Basin and the Congo region of Africa. The region of heavy rainfall that usually lies across the Indo-Pacific island region, however, is displaced to the central Pacific. Subsequently, the region of heavy rainfall shifted even farther eastward, heavily impacting Ecuador and northern Peru. The annual flow of the Piura River in northern Peru affords a sensitive indicator to the meteorological impact of El Nifio in this normally arid region. Fig. 7 shows the annual discharge of this river since 1952. A l l of the El Nifio event years identified by Rasmusson and Carpenter (1982) (cf. Fig. 3), except the weak to moderate events in 1963 and 1969, stand out as wet years having four to ten times the discharge of the intervening years. In 1983, in turn, the river discharge was more than five times that of 1953, the second ranking year during this thirty-two year period. These extraordinary rains had devastating terrestrial effects, and doubtless contributed to low surface salinity in the oceanic region as well.
1
90E
I
180
,
9ow
u-
I
0
I
90
I
IlOE
Fig. 6. Global pattern of cloudiness and rainfall as evidenced by average outgoing longwave radiation (W m-* ) during spring of 1983. (From CAC, 1984.)
11
12-
10-
v)
E c
8-
w
I
0
m
5
6-
LL
0 v)
Z
5d m
4-
2-
.
Of
1960 1970 YEAR
1980
Fig. 7. Annual discharges of the Piura River, Peru (from Woodman, 1984).
Fig. 8. Time-longitude plot of westerly wind anomaly ( m s - l ) in 5ON-5'S across the tropical Pacific. (From CAC, 1984.)
band
12
Associated with the displacement of the precipitation pattern was an alteration of the winds over the ocean surface. The normal wind over the equatorial Pacific Ocean is the southeast tradewind blowing generally westward. On a daily or even a monthly basis the surface winds over much of the tropical ocean are sparsely reported relative to their variability. Although not the wind at the surface which would be most valuable, useful information about variations of wind in the lower atmosphere can be obtained from movements of low level clouds in satellite pictures. In early 1982 the low level zonal winds across the equatorial Pacific were near normal. The easterlies were slightly stronger than normal in the central Pacific, and very slightly weaker than normal in the IndoPacific i;land region (Fig. 8). By July, as the region of convective rainfall moved eastward, the tradewinds in the western Pacific had weakened to the extent that west of the dateline the near equatorial winds had become westerlies. These westerly winds continued to intensify and moved into the eastern Pacific during the first quarter of 1983, while stronger than normal easterlies returned to the far western Pacific. The zonal wind had returned to near normal all across the equatorial Pacific by July 1983. Among the consequences of the rotation of the earth on its axis are strong constraints on the movements of the ocean and atmosphere. Often these movements take the form of wavelike processes on very large scale. Near the equator, wavelike movements that travel rapidly eastward are possible, and appear to be a prominent part of the way in which the ocean responds to varying surface winds, including those variations associated with El Nifio. These eastward propagating features are called equatorial Kelvin waves. A dynamically related kind of movement, the coastal Kelvin wave can propagate along the continental coastlines. Coastal Kelvin waves travel in such a direction as to have the coastline on their right in the Northern Hemisphere, and on their left in the Southern Hemisphere. In regions away from the equator the principal wavelike process is the Rossby wave. Large-scale Rossby waves travel westward. Near the equator their speed is only about one-third that of the Kelvin wave, and is even slower in higher latitudes. The sequence of events that transpires is that an equatorial Kelvin wave is generated by a change of surface wind in the tropical Pacific, and travels eastward to the coast of South America. At the coast it is partly reflected as a westward travelling Rossby wave, and partly converted to coastal Kelvin waves that travel poleward in both hemispheres. Because westward propagation of Rossby waves is so slow in higher latitudes, El Nifio effects that are carried into middle and high latitude coastal regions by the coastal Kelvin waves may persist for many months unless they are obliterated by other processes such as local winds in those regions. In the circumstance of El Nifio the Kelvin waves are of "downwelling" nature. That is, they are associated with an elevation of sea level, but a deepening of the thermocline as they propagate along the equator and the coasts.
13
M A
M J
2 J Q,
.-A S 0
N D J
F M A
2 M 0,
- J J
A S
0
Fig. 9. Zonal displacements of drifting buoys in the 4 O N - 4 O S band of the eastern tropical Pacific. The generally accepted interpretation of the observations made during 19821983 is that the early warming across much of the eastern equatorial Pacific was
a result of eastward surface current and deepening of the thermocline associated with a downwelling Kelvin wave caused by the reversal of the surface winds in the western Pacific. The wind anomaly traversed the ocean more slowly than the SST anomaly. The second peak of SST anomaly in the eastern Pacific appears to have resulted from eastward surface currents driven by the more local winds. AS will be seen, the second peak is more local in several respects. Support for these interpretations is afforded by the movements of satellitetracked drifting buoys released in the tropical Pacific to monitor the surface currents. The zonal movements of several of these buoys in the eastern Pacific is shown in Fig. 9. In the early summer of 1982 drifting buoys near the equator
14
in the eastern Pacific were moving westward at about 60 km per day, indicating moderate to strong currents in the normal direction. During the autumn the currents slowed until by December they had turned weakly eastward over most of the eastern tropical Pacific. The end of this phase coincided with the early maximum in SST anomaly. In early January 1983, and continuing into March, the surface currents again reversed, becoming powerfully westward. Moored current meter observations on the equator at 95OW and ll0W indicate that the Equatorial Undercurrent also ceased during this period (Halpern, 1987). Based on simulation using a numerical ocean circulation model, Philander and Siege1 (1985) traced the spectacular events in the eastern Pacific to the sudden return of easterly winds in the western Pacific (Fig. 8 ) . By March the patch of westerly surface winds had moved into the eastern Pacific, driving a strong jet-like eastward surface current. This eastward surface current was seen only near the equator, although probably there was eastward flow also in the North Equatorial Countercurrent. Elsewhere in the eastern Pacific the surface currents were westward. The eastward current pulse coincided with the development of the second, and at the coast largest, peak in the SST anomaly (Figs. 3, 4). In June easterly winds returned also to the eastern Pacific, the surface currents accelerated westward, and the eastern Pacific cooled rapidly. The eastward surface currents observed in association with the event must have had the effect of substantially increasing the residence time of these waters in the eastern Pacific. Waters found near the Galapagos in June 1983 would in a normal year have been displaced 5,000 to 7,000 km westward. This presumably has significant effects on ecological variables such as nutrient replacement and plankton transport as well as temperature. The disappearance of a primary feature of the tropical ocean circulation, the Equatorial Undercurrent, is extraordinary, but the ecological impact of the phenomenon is a difficult problem. The meridional pattern of evolution of the eastward surface currents is strongly reflected in that of the surface temperature anomalies. The first maximum of the SST anomaly extended into the central Pacific as did the eastward surface current anomalies associated with it. This maximum was observed far to the south along the coast of South America, with slowly attenuating magnitude (Fig. 10). The secondary peak in surface temperature anomaly that occurred only in the eastern Pacific, and was associated with strong eastward surface current anomalies only in the eastern Pacific and only near the equator, substantially exceeded the initial anomaly near the equator ( 5 O S ) in magnitude, but attenuated rapidly with poleward distance along the coast of South America. The first maximum was observed at least as far south as Talcahuano (36.7's) but not at Corral ( 4 O o S ) , Chile (Fonseca, 1985). In the Northern Hemisphere, there is evidence of both temperature anomaly peaks as far north as into the northern Gulf of Alaska (Royer, 1985), but differentiating the second peak north of Vancouver Island becomes difficult.
15
10 -
8-
60 O
4-
-
2-
-
0-2
-
I
I
1
1
I
l
1982
l
I
I
I
I
I
I
I
I
I
I
l
l
I
I
I
I
1
1983
Evolution of coastal water temperature anomalies at Paita (5OS), Puerto Chicama (7.7OS1, Chucuito (12.loS), Funta Coles (17.7OS), Arica (18.5OS), and Caldera (27's). Below the ocean surface, the thermal anomalies associated with El Niilo are most readily described as deepening of isotherms from their normal levels. In the eastern tropical Pacific, the 2OoC isotherm lies in the upper part of the main thermocline. Increases in its depth can be expected to have direct local ecological effects through change of the thermal habitat, and indirect regional effects by reducing exposure of nutrient-rich waters to surface mixing processes. In the vicinity of the Galapagos Islands and along the coast of Central America and northern South America the 20°C isotherm is normally about 30 meters deep, with temporal variations of about 15 meters (Hansen and Herman, 1988). Westward and toward the subtropical gyres of the North and South Pacific it deepens so that at 140°W it is about 120 meters deep at 20°N and 230 meters deep at 20"s. During September 1982 the 2OoC isotherm had plunged to more than 75 m below its normal depth along the equator in the eastern Pacific (Fig. 11). In subsequent months the deepening increased and propagated poleward along the coasts of Central and South America before attenuating. No intensification of the depth anomaly of the 2OoC isotherm was noted in connection with the second peak in surface temperature anomaly. The second temperature maximum was a quite superficial feature limited to the upper 50 meters or so of the ocean. Poleward of the region shown in Fig. 11, isotherm depressions were observed as far as 2OoS, and doubtless reached farther. At Iquique (20.3OS) the principal temperature m a x i m extended much deeper than 50 meters and was observed from December 1982 through March 1983. A small secondary m a x i m was observed in the upper 20 meters in June 1983 (nenzalida, 1985). In the North Pacific isotherm deepening of up to 200 meters was observed within 150 kilometers of Vancouver
16
meters
0
10
20
30
40
50
Fig. 11. Depth anomaly of the 2OoC isothermal surface in the eastern tropical Pacific during (top) September 23-October 2, 1982, (middle) November 22-December 21, 1983, and (bottom) January 21-February 21, 1983. Contours show deepening in meters, gray tone scales show uncertainty of determination in meters (from Hansen and Herman, 1989).
17
180 -
420
400 380
JARVIS
1
I
200
SANTA CRUZ 160-
140-
\
Fig. 12. Variation of sea level at Nauru Island, Jarvis Island, and Santa Cruz Island in the equatorial Pacific (from Lucas g.,1984).
e
Island during March 1983 (Tabata, 1985), and in the northern Gulf of Alaska (Royer, 1985). Water temperature continued to be above normal to depths in excess of 250 m in the northern Gulf of Alaska until the middle of 1984. The 2OoC isotherm, of course, does not occur at these high latitudes. Another important aspect of El Nifio is its effect on sea level. In normal times the surface of the Pacific ocean slopes upward from east to west due to the westward force of the surface winds. When these winds fail or reverse, as has been described for the El Nifio, the sea level also is affected (Lucas g al., 1984). Fig. 12 shows the variations of sea level at three island sites in the equatorial Pacific. At Nauru Island in the western Pacific (167"E) sea level began lowering in late 1982, falling almost 25 centimeters by March 1983, then over a period of several months returned to its normal level, but less than its level had been in 1982. At Jarvis Island, near 160°W in mid-Pacific, sea level rose to a m a x i m of almost 30 centimeters above normal during the last half of 1982, then fell to a little below normal simultaneously with the development of the initial peak'anomaly of sea surface temperature in the eastern Pacific. As sea level at Jarvis Island was lowering, that at Santa Cruz (90.3OW) in the Galapagos Islands began to rise. It reached a peak anomaly of about 44 centimeters in very early 1983. Sea level in the Galapagos reflects in detail the events that have been described in the evolution of the SST anomaly maxima and the observations of eastward surface currents. The pattern of sea level variations along the coast is shown in Fig. 13. The sea level changes observed at La Libertad, Ecuador (2OS) are similar to those at Santa Cruz, 1,000 kilometers to the west. The first peak of the sea level anomaly propagated rapidly southward along the coast with little attenuation, but the peak associated with the large SST anomaly in May-June 1983 is largest
18
0
i!2 IW
2o
,In
1
Fig. 13. Variations of monthly mean sea level at (top) La Libertad, Ecuador, (middle) Callao, Peru, and (bottom)Antofagasta, Chile. in the tropics, like the SST anomaly and the surface current variations. In the Northern Hemisphere, the first sea level anomaly m a x i m was observed far up the Pacific coast of Alaska (Cannon g g.,1985), and a secondary maximum in June was distinguishable as far as 44.6ON (Huyer and Smith, 1985). A final aspect that may be unique to the El NiTio event of 1982-1983, and one that probably has at least local ecological effects, is the extraordinary waves that assaulted exposed regions of the entire coast from California to Chile early in 1983. The unusual winds in the tropical Pacific coincided with a deepening of the Aleutian low pressure center in the North Pacific, not necessarily due to the El NiTio (Namais, 1983). This resulted in unusually strong westerly winds and a series of severe storms at unusually low latitudes in the subtropical North Pacific. Surface waves generated by these storms radiated out for thousands of kilometers. Severe coastal erosion and damage to facilities were reported from both California and Ecuador, and related problems were experienced as far south as central Chile. The fact that these unusually large waves came in the presence of higher than normal average sea level made them more destructive to coastal features than would otherwise have been the case.
19
Local ecological impacts resulted from the mechanical forces that ravaged the coast (see Dayton and Tegner, and Glynn, this volume), and increased turbidity resulting from coastal erosion, might also have an effect. 5 REFERENCES Andrade, E.R. and Sellers, W.D., 1988. El NiEo and its effect on precipitation in Arizona and western New Mexico. J. Climatology, 8: 403-410. Bjerknes, J., 1969. Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97: 163-172. CAC, 1984. Climate Diagnostic Bulletin. Published monthly by National Meteorological Center, NOAA, Washington, D.C. Cannon, G.A., Reed, R.K. and Pullen, P.E., 1985. Comparison of El Niiio events off the Pacific northwest. In: W.S. Wooster and D.L Fluharty (Editors), El Niiio North. Washington Sea Grant Program, Univ. of Washington, 75-84. Eguiguren, D.V., 1894. Las lluvias en Piura. Bol. SOC. Geograf. Lima, 4(7-9): 241-258. Fonseca, T.R., 1985. Efectos fisicos del fenomeno El Niiio 1982-83 en la costaChilena. Investigacion Pesquera, Num. Esp. Taller Nacional Fenomeno El Nino 1982-83. Santiago, 61-88. Fuenzalida, R., 1985. Aspectos oceanograficos y meteorologicos de El Niiio 1982-83 en la zona costera de Iquique. Investigacion Pesquera, N m . Esp. Taller Nacional Fenomeno El Niiio 1982-83. Santiago, 47-52. Gill, A.E. and Rasmusson, E.M., 1983. The 1982-83 climate anomaly in the equatorial Pacific. Nature, 306: 229-234. Halpern, D., 1987. Observations of annual and El Nifio thermal and flow variations at O o , llOoW and O o , 95OW during 1980-1985. J. Geophys. Res., 92(C8): 8197-8212. Hamilton, K., 1988. A detailed examination of the extratropical response to tropical El Nifio/Southern Oscillation events. J. Climatology, 8: 67-86. Hansen, D.V. and Herman, A., 1988. A seasonal isotherm depth climatology for the eastern tropical Pacific. NOAA Tech. Rept. ERL 434-AOML 33 (Rev.), 35 pp., Atl. Oceanogr. and Meteorol. Lab., Miami, FL. Hansen, D.V. and Herman, A., 1989. Evolution of isotherm depth anomalies in the eastern tropical Pacific Ocean during the El Niiio event of 1982-83. J. Geophys. Res., 94(C10): 14,461-14,473. Huyer, A. and Smith, R.L., 1985. The apparition of El Niiio off Oregon in 1982-83. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North. Washington Sea Grant Program, Univ. of Washington, 73-84. Lucas, R., Hayes, S.P. and Wyrtki, K., 1984. Equatorial sea level response during the 1982-83 El Niiio. J. Geophys. Res., 89(C6): 10,425-10,430. Namais, J., 1983. Advance signs of the western hemisphere climate observations observed in winter, spring, and summer 1983. In: Proc. Eighth Ann. Climate Diag. Workshop, NOAA, U.S. Dept. of Corn., Washington, D.C., 55-62. Nicholls, N., 1987. The El Nifio/Southern Oscillation phenomenon. Ch. 1, In: M. Glantz, R. Katz, and K. Krenz (Editors), The Societal Impacts Associated with the 1982-83 Worldwide Climate Anomalies. Report based on the workshop in Lugano, Italy, 11-13 November 1985. National Center for Atmospheric Research, Boulder, CO, 105 pp. Pan, Y.H. and Oort, A.H., 1983. Global climate variations connected with sea surface temperature anomalies in the eastern equatorial Pacific Ocean for the 1958-1973 period. Mon. Wea. Rev., 111: 1244-1258. Peterson, G., 1935. Estudios climatologicos del noroeste Peru. Bol. SOC. Geol. Peru, 7(2): 1-141. Philander, S.G.H. and Siegel, A., 1985. Simulation of El Niiio of 1982-83. In: Proceedings of the Liege Colloquium (1984), World Climate Research Program, J.C.J. Nihoul (Ed.), Coupled Ocean Atmosphere Models, pp. 517-541. Elsevier Science Fubl., Amsterdam, Holland, May 1984. Quinn, W.H., Zopf, D.O., Short, K.S. and Kuo Xang, R.T.W., 1978. Historical trends and statistics of the Southern Oscillation, El Niiio, and Indonesian droughts. Fishery Bull., 76(3): 663-678.
20
Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Nifio occurrences over the past four and a half centuries. J. Geophys. Res., 92(C13): 14,449-14,461. Rasmusson, E.M. and Carpenter, T.H., 1982. Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillationfll Niiio. Mon. Wea. Rev., llO(5): 354-384. Royer, T., 1985. Coastal temperature and salinity anomalies in the northern Gulf of Alaska. In: W.S. Wooster and D.L. Fluharty (Editors), El Nifio North. Washington Sea Grant Program, Univ. of Washington, 107-115. Schott, G., 1931. Der Peru-strom und seine nordlichen Nachbargebeite in normalen und anormaler Ausbildung. Ann. Hydr. Mar. Met., 59(5): 161-169, (6): 200-213, (7): 240-253. Tabata, S., 1985. El Nifio effects along and off the Pacific coast of Canada during 1982-83. In: W.S. Wooster and D.L. Fluharty (Editors), El Niao North. Washington Sea Grant Program, Univ. of Washington, 85-96. Taft, B.A., 1985. El Niiio of 1982-83 in the tropical Pacific. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North. Washington Sea Grant Program, Univ. of Washington, 1-8. Walker, G.T. and Bliss, E.W., 1932. world weather V. Mem. Roy. Meteor. SOC., 4: 53-84. Woodman, R.F., 1984. Recurrencia del fenomeno El NiEo con intensidad comparable a1 aiio 1982-83. In: Proc., Seminario Regional, Ciencia, Tecnologia y Agresion Ambiental, El Fenomeno El Niiio, CONCYTEC, Lima, 301-332. Yarnal, 8. and Diaz, H.F., 1986. Relationships between extremes of the Southern Oscillation and the winter climate of the Anglo-American Pacific coast. J. Climatology, 6: 197-219.
21
NUTRIENTS AND PRODUCTIVITY DURING T H E 1982/83 EL NINO R.T. BARBER and J.E. KOGELSCHATZ Monterey Bay Aquarium Research Institute, Pacific Grove, California, 93950 USA ABSTRACT Barber, R.T., and Kogelschatz, J.E. 1989. Nutrients and productivity during the 1982/83 El Niiio. The eastern Pacific Ocean, particularly in the low latitude region from 20"s to 20"N, has higher primary productivity than the equivalent region in the western Pacific. Enhanced productivity at the base of the food web supports larger populations and faster growth rates throughout higher trophic levels so that the ecological character of the eastern Pacific is distinctly different from that of the western Pacific. El Niiio is a natural, aperiodic, coupled ocean/atmosphere perturbation of the global heat budget that profoundly modifies the normal east/west asymmetry of both heat content and productivity of the Pacific basin. During the 1982/83 El Niiio a surface layer of warm, nutrient-depleted water appeared in the eastern Pacific and persisted for about 9 months. Nutrient supply, phytoplankton abundance and primary productivity were dramatically reduced by the altered physical conditions of the 1982/83 El Niiio. The decrease of new primary production available to the marine food chain caused proportional reductions in growth, reproduction and survival of marine invertebrates, fish, birds and mammals. Food deprivation together with active and passive redistribution of organisms accounts for most of the biotic changes observed in higher trophic level organisms but direct thermal and sea level effects also were observed, particularly in sessile invertebrates. 1 INTRODUCTION
El Niiio is the appearance and persistence of anomalously warm water in the low latitude eastern Pacific. El Nitio is one facet of a basinwide phenomenon called the El Niiio Southern Oscillation (ENSO) cycle. The ENSO cycle is a natural, aperiodic, coupled ocean/atmosphere cycle that determines both the climatological mean conditions and the major interannual variability of the large-scale heat flux. The ENSO cycle is arguably the most important natural process causing biotic variability in the low latitude Pacific ocean because the cycle itself is responsible for both the "normal" and anomalous ecological conditions that characterize tropical Pacific waters. In that particular context the ENSO driven interannual progression of normal and abnormal years plays a major role in determining the prevailing ecological character of the low latitude Pacific. It has been understood for many decades that the eastern boundary regions of the ocean basins have higher biological productivity than the western boundaries. This truism was perhaps first formally incorporated into a figure by Sverdrup (1955) when he used first principles to make a global estimate of the pattern of relative primary productivity. The principle Sverdrup (1955) used was that the first order process regulating ocean productivity is the transport of inorganic plant nutrients from deep water to the surface layer where there is adequate light. Where seasonal mixing, upwelling or topography enhances the vertical nutrient transport Sverdrup predicted there would be increased productivity. That Sverdrup (1955) quite accurately estimated
22 the gross pattern of primary productivity in the tropical Pacific is shown by comparison of Figure 1 with Figure 2, a map from Fleming (1957) that was based in part on observations of productivity. Later syntheses of large-scale patterns of productivity such as that by Koblentz Mishke el
QI.
(1970) confirm that the eastern Pacific is considerably more productive than the
western Pacific, i.e. there is a strong east/west asymmetry in biological productivity. To understand the biological consequences of El Niiio, it is useful to examine how the east/west asymmetry in basic productivity is created and maintained because, in simplest terms, El Niiio is a dynamic process that redistributes heat in such a manner as to greatly reduce the normal east/west asymmetry of the Pacific. Evidence of the elimination of east/west asymmetry in surface layer heat content in the Pacific basin is shown in Figure 3 where the global sea surface temperature field for the peak of the 1982/83 El Niiio (June 1983) is compared with June 1984 when the Pacific had returned to its normal condition. The global sea surface temperature
summary shows that in June 1983 no water cooler than 25°C was present anywhere in the equatorial Pacific. There is little or no east/west asymmetry in heat content. During normal conditions in June 1984, the eastern third of the equatorial Pacific is occupied by surface water cooler than 25°C and the western third is occupied by water over 29°C. The June 1984 situation shows the east/west asymmetry in heat storage that characterizes the normal Pacific condition. This contribution will examine the nutrient and productivity effects of El Niiio by describing the normal asymmetry, how it is created and maintained, and then will present observations on how the 1982/83 El NiAo modified the basinwide asymmetry causing large-scale changes in nutrient conditions which in turn caused significant decreases in primary productivity that propagated through the food web. 2 THE ENSO CYCLE The ocean/atmosphere phenomenon that determines the basinwide ecological character of the tropical Pacific is the ENSO cycle. The strong trade wind or climatological mean phase of the cycle creates the prevailing characteristics of the basinwide ecosystem (Barber, 1988), and El NiAo, or the anomalous phase of the cycle, determines the environmental extremes of temperature, sea level, and nutrient supply that set limits of abundance and distribution on marine organisms in the affected region (Barber and Chavez, 1986). The El NiAo part of the ENSO name refers to the episodic interannual redistribution of water and heat in the low latitude Pacific; the Southern Oscillation component of the name refers to a coupled oscillation of the South Pacific atmospheric high pressure system and the Indonesian atmospheric low pressure system described in this volume by D.V. Hansen. The surface pressure gradient between these two persistent pressure systems forces the easterly trade winds that dominate the equatorial Pacific. (Figure 2 A in Hansen’s contribution shows how the high and low pressure systems are vertically connected to form a Walker circulation cell and also how the Walker cell over the Pacific connects to other tropical Walker cells.) A long historical analysis of the two pressure systems has established that when pressure rises in the Indonesian low pressure system it falls in the South Pacific high; that is, the two systems show a tightly coupled oscillation that is opposite in sign. This oscillation causes the strength of the atmospheric pressure gradient across
23
Fig. I . Relative potential primary productivity. Light area is low, single hatched area is medium and the cross hatched area is estimated to have highest primary productivity. Re-drawn from Sverdrup (1955).
30
20 10 N
0 10
S 20 30
Fig. 2. Primary production in the equatorial Pacific from 20"N to 20"s in gC/m2/year. Redrawn from Fleming (1957).
24 the tropical Pacific to vary from year to year and drives the low frequency variability in the strength of the easterly trade winds. The climatological mean easterly trade winds (and mid latitude westerlies) set up and maintain the fundamental east/west asymmetry of the Pacific including a basinwide tilt of the thermocline and nutricline with these features deep in the west and shallow in the east (Figure 4). The mean large-scale winds also maintain a basinwide gradient in surface layer heat content with a heat surplus in the west and a deficit in the east. The basic east/west asymmetry is itself the product of coupled ocean/atmosphere processes. The western Pacific contains a "warm pool" of water with an annual mean temperature of > 2 9 T (Figure 3). Solar energy falling on the > 2 9 T water is transferred to the atmosphere by evaporation, convection and back radiation. The convective activity over the warm pool of the western Pacific creates the Indonesian low pressure system that draws air into the upward limb of
M
al
?5 W
c 3
3
1
1
120°€ 140'
1
1
160°€
180°
1
I
1
16OoW 140'
120'
looo
80'
60'
40'
ZOoW
0
20°E
40°N
2 00
d03
c" €0
W
c 3
200 40%
Fig. 3. Sea surface temperature fields in June 1983 and June 1984 in the Pacific and Atlantic Oceans. This analysis blends satellite and shipboard temperature measurements and was provided by Vernon Kousky of the Climate Analysis Center of NOAA. The June 1984 temperature field in the Pacific is close to the climatological mean condition with a well developed "cold tongue" reaching out from the coast of South America and a "warm pool" of water >2YC in the far western Pacific. In June 1983 the "cold tongue" is missing and there are large regions of anomalously warm water off Central and South America.
a Walker circulation cell (Figure 4). This east to west transfer of mass is the easterly trade wind system that characterizes the tropical Pacific. Along the eastern boundary of the low latitude
25 Pacific, equatorward winds force coastal upwelling and in the central and eastern equatorial region the trades force equatorial upwelling. Coastal and equatorial upwelling are circulation patterns in which there is organized vertical transport of subsurface water to replace surface water that has been horizontally displaced by wind. In the surface layer of the ocean this recently upwelled cool water extracts heat from the atmosphere, causing atmospheric subsidence and creating the South Pacific high pressure system. The subsiding air mass feeds into the easterly trades and flows to the west towards the convective center over the warm pool of the western Pacific (Figure 4). The easterly trade winds drive surface water westward in the form of the Equatorial Current System; as this water flows westward it gains heat from the sun. The net effect of the wind driven circulation is to transport heat westward into the warm pool of the western Pacific. Thus the east/west asymmetry in surface layer heat content sets up an atmospheric pressure gradient that drives winds that further accentuate the east/west asymmetry in heat content.
--
____I_, --I1
-. - -
,
,
,
,
I-
,
,
,
:
,
,
.;
,
,
,
,
,
,
,
,
,
,
,
e
Overshoot C. 50
April 1984
100
Theory
Temperature
Nitrole
Fig. 4. The El Nitio Southern Oscillation (ENSO) cycle in relation to the west vs. east temperature gradient and the basinwide thermocline and nutricline tilt. The center of the convective heating during normal conditions is usually further to the west than shown in A; Fig. 3 shows that the > 2 9 T warm pool is between 120"E and 160"E; the convective center is usually also in this region. During the mature phase of the 82/83 El Nifio, in April 1983, the center of convective activity was around 120"W as shown in Panel B. From Barber (1988). The asymmetry in heat and potential energy in the form of a higher stand of sea level in the west increases with each year of normal trade winds. Clearly this cycle of positive feedback from ocean to atmosphere and from atmosphere to ocean cannot indefinitely store more and more heat (and more water) in the western Pacific (Cane and Zebiak, 1985). El Nitio is a process that interrupts the year to year accumulation and serves to redistribute the heat and mass back across the Pacific towards the equilibrium condition; that is, El Nitio decreases the east/west asymmetry. The redistribution is set in motion by variation in the winds. The surface wind stress of the
trades establishes a hydraulic pressure head by accumulating water in the western Pacific by about perhaps 1 m above the equilibrium level. If the easterly trades slacken the wind stress decreases as the square of the change in wind speed. When wind stress is suddenly decreased the potential energy of the water accumulated in the west is released, and a pulse of water travels eastward across the Pacific, in part as a Kelvin wave whose speed of advance is 2 to 3 m/sec (O'Brien ef al., 1981). The wave takes three to four months to cross the basin and carries with it a surge of warm, western Pacific surface water (Harrison and Schopf, 1984). The arrival of the series of Kelvin waves off the coast of South America results in a depression of the thermocline by 100 to 150 m, and a warming of the surface waters. Of course, the redistribution of water causes the sea level in the western tropical Pacific to drop by 0.2 to 0.4 m and rise in the eastern Pacific by a similar amount. Figure 3 shows that in a strong El Nifio such as the 1982/83 event the sea surface temperature asymmetry was eliminated across the Pacific.
For coupled air/sea processes such as the ENSO cycle, the critical boundary is the sea surface where the transfer of heat and momentum between the ocean and the atmosphere takes place; that is, the east/west sea surface temperature asymmetry is the critical gradient driving the coupled air/sea processes. However, for determining the ecological character of an ocean region the critical property is not the sea surface temperature but the quantity of heat stored in the form of warm water in the mixed layer above the thermocline. The thickness of the mixed layer determines the ecological character of an ocean region because the major inorganic nutrient reservoir of the ocean is water below the thermocline; therefore, the thickness of the surface mixed layer is involved in regulating the rate at which nutrients can be mixed o r advected into surface waters where there is adequate light for photosynthesis. Any process that depresses the thermocline, that is, any process that increases the thickness of the mixed layer o r increases heat storage, will decrease the rate of supply of new nutrients and necessarily reduce the biological productivity of a region. Light available for photosynthesis decreases exponentially as a function of water depth even in very clear ocean water, because water molecules absorb much of the available light. The result is that in permanently stratified regions with a deep mixed layer, radiant energy and nutrients are separated spatially and primary production, particularly new primary production in the sense expressed by Dugdale and Goering (1967) and Eppley and Peterson (1979), is reduced. New primary production is that proportion of the total production supported by nutrients that are advected or mixed into the euphotic zone; in that context new production is distinguished from recycled production which is supported by nutrients that are regenerated within the euphotic zone. The ENSO cycle regulates the quantity of heat stored in the mixed layer in the eastern and western Pacific; this is the same as saying ENSO regulates the depth of the thermocline. Covarying with, but not identical to, the thermocline is the nutricline, the gradient that separates the nutrient-depleted surface mixed layer from the deep water nutrient reservoir. The critical gradient for the ecosystem that the ENSO cycle regulates is the depth of the nutricline.
3 THE BASINWIDE SETTING The waters of the eastern tropical Pacific are among the most productive of the world ocean.
27
The map of potential relative productivity (Figure 1) produced by Sverdrup (1955) and the early observations summarized by Fleming (1957) (Figure 2) emphasize the richness of the eastern equatorial and coastal Pacific. The biological richness of the region was recognized as early as the late nineteenth century by Buchanan (1886) who commented that "No waters in the ocean so teem with life as those on the west coast of South America." This region has remarkably cool ocean waters for a tropical region. The June 1984 panel of Figure 3 shows the presence of a "cool tongue" of water along the equator that has surface temperatures of <25"C. Along the coast
of South America the temperatures in June 1984 were <20"C. Despite the continual input of heat by the equatorial sun, temperatures of 25" f 2°C characterize the eastern equatorial region and temperatures of 20"
?
2°C characterize the coastal region. The association of unusually cool water
and biological richness was first made by Coker (1908 and 1918) who recognized that upwelling was the source of the cool water in this tropical region. He speculated that the greater solubility of gases in the cooler seawater might be responsible for the enhanced "growth of minute plants that form the basis of the food supply of all the marine animals." This speculation is surprisingly close to being correct: when chemical analyses became available they showed that the high concentration of nitrate, phosphate and silicate in the upwelled water was responsible for the enhanced productivity of this region (Gunther, 1936). The enhanced nutrient supply is the combined result of several ocean processes operating at different time and space scales. The largest scale at work is the thermohaline circulation of the world ocean that operates on a turnover time of 100 to 1,000 years. This large, slow circulation partitions the ocean into surface and subsurface components separated by the main thermocline. A net downward, gravity-driven flux of organic particles transports carbon, nitrogen, phosphorus and silicon through the thermocline into deep water; the organic particles are oxidized at depth to the inorganic anions of carbonate, nitrate, phosphate and silicate. The net downward flux, resulting oxidation and the long turnover time of the thermohaline circulation cause the subthermocline waters to be enriched in inorganic plant nutrients. For these nutrient elements to be taken up by plants and support primary productivity, they must re-enter the sunlit surface layer where there is enough light for photosynthesis. The process of reentering the euphotic zone involves physical processes operating at both local and basinwide scales. Upward transport by means of coastal or equatorial upwelling is a local response of the ocean to wind. In practice this means that when upwelling favorable winds occur the ocean responds on a time scale of about one day over a region of 50 to 100 km (Barber and Smith, 1981). In addition to having a relatively small horizontal domain and short timescale, the local dynamics of wind-driven upwelling appear to recruit water from a relatively shallow depth of only about 40 to 80 m. For upwelling to enhance productivity, nutrient rich water must be available in the 40 to 80 m depth range. This final condition necessary for productivity enhancement is, in fact, regulated by the basinwide east/west asymmetry set up by the large-scale trade winds. The normal basinwide slope in nutrient structure (Figure 4) makes the eastern Pacific inherently more productive because the nutricline and subsurface nutrient pool are more shallow and, therefore, closer to or overlapping the euphotic zone where radiant energy is available for photosynthesis. By regulating the depth of the nutricline, the large-scale winds control the nutrient content of
28 water that feeds into any physical process that drives advective or turbulent vertical transport. Numerous dynamic processes such as island wake mixing, shelf-break upwelling, tidal mixing, geostrophic upwelling, wind-driven mixed layer stirring or wind-driven upwelling all must reach into the nutricline to have a significant effect on productivity. The Pacific asymmetry in the depth of the thermocline and nutricline is the link between local physical dynamics that transport nutrients upward and the regional productivity response. It is the basinwide response of the ocean to large-scale winds that creates the different ecological character of the eastern and western Pacific and it is this large-scale, low-frequency response of the ocean that changes during the El Nifio phase of the ENSO cycle. 4 WESTERN PACIFIC
The mean easterly trades accumulate warm, nutrient-depleted water in the surface layer of the tropical western Pacific setting up a deep thermocline, pycnocline and nutricline as shown in Figure 4. Wind-driven mixed layer stirring, island wake mixing, and equatorial upwelling all take place throughout the warm-pool region but they have a greatly reduced ecological impact because these processes do not reach through the thick mixed layer to the nutricline; therefore, mixing and upwelling transport very few nutrients upwards. The ecological character of the western Pacific is determined by the low nutrient water that accumulates in the warm pool and by the low rates of nutrient supply to the surface layer; therefore, the ecological character of the
130
120
140
150
160
170 El8OW 170 160
40nv94
150
140
130
120
110
I00
90
80
L
30
.
Hwoii
Fig. 6 Fig 4
Fig. 5. Map of the tropical Pacific showing the location of the sections and time series stations. The numbers indicate the figures that present data from that location.
region is a direct consequence of the normal trade wind phase of the ENSO cycle. Figure 5 shows the location of the sections that are discussed in the remainder of this report. A cross equatorial profile at 165% through the warm pool of the western Pacific is shown in Figure 6. When this transect was made in February 1986, drifting buoys deployed by D.V. Hansen of the
29 Atlantic Oceanographic and Meteorological Laboratory of NOAA established that there was strong equatorial divergence; therefore, relatively strong equatorial upwelling was occurring. Figure 6 shows there was no nitrate signature of the equatorial divergence because the upwelling circulation cell was apparently confined to the mixed layer above the nutricline with the result that the upwelled water was nutrient-depleted. Figure 6 shows a subsurface chlorophyll maximum located in or just above the nutricline. A strong subsurface chlorophyll maximum poised in relation to the nutrient gradient is a common, large-scale ecological characteristic of tropical waters that are permanently stratified with a relatively deep mixed layer. In the vicinity of the equator, particularly between 0" and 2"N, the chlorophyll profile in Figure 6 shows evidence of upwelling. It appears that when this section was occupied the equatorial upwelling reached deep enough to entrain water from the top of the chlorophyll maximum layer but the absence of an upwelling temperature signal indicates that water from the thermocline was not entrained.
US/PRC #I 165"E
-S
-N
100
200
-
0
. . . . . . . . . . . . . .......
E
Y
. . .. .. ..
* 2 . .
100 f a.
0"
200
Fig. 6. Meridional profiles of temperature, nitrate and chlorophyll along 165"E from I0"N to 6"s in February 1986.
The typical nutrient depletion and low plankton biomass character of the western Pacific are further shown in Figure 7, which presents February 1986 and November 1986 meridional sections from 8"N to 18"N along 130"E. This location is just east of the Philippines. The sections show that while the temperature of the mixed layer is somewhat cooler than in the center of the warm
30 pool along 165"E, the mixed layer along 130"E is nutrient depleted and contains relatively low chlorophyll concentrations of 0.10 f 0.05 mg/m3. The top of the nutricline is never more shallow than 75 m and is deeper than 100 m over most of both sections. The mixed layer conditions shown in Figures 6 and 7 of relatively warm, nutrient-depleted water with a chlorophyll content of 0.10 f 0.05 mg/m3 and a primary productivity rate of 200 f
50 mgC/m2/day represent the basic trophic character of the western Pacific. These waters are fundamentally oligotrophic, warm and low in planktonic biomass with a relatively weak annual cycle of nutrient supply or plankton biomass. The low plankton biomass results in very clear water, and clear and continually warm waters provide the environmental conditions required by corals for reef building. The coral reef community with its complex assemblage of marine algae, symbiotic algae, invertebrates and fishes finds its maximum expression in the warm pool region of the western Pacific. In this context, the high diversity coral community is as much a result of the east/west asymmetry as is the diatom/anchovy/seabird assemblage of the coast of Peru. The normal trade wind phase of the E N S 0 cycle sets up the basic east/west asymmetry of heat storage in the surface layer of the ocean and thereby creates in the western Pacific an environment fundamentally different from that in the eastern Pacific.
130' E 0,
8.
:
9' *
10.
:
I(* 12. '
:
13' '
14' '
15'
:
16'
',:
Tronsect
17' 1 0 " ' , . I
8.
1
:
9.
10.
:
II* 12. ' :
13. 14. '
'
",
15. 16. 17. 18"
:
'
:
Fig. 7. Meridional profiles of temperature, nitrate and chlorophyll along 130"E from 18"N to 8"N in February 1986 and November 1986
5 EASTERN PACIFIC NORMAL CONDITIONS 5.1 The equatorial region The easterly trades have the same local effect in the eastern equatorial Pacific as they do in the western equatorial Pacific; that is, in a narrow band along the equator the zonal winds drive
31
equatorial upwelling. Because the nutricline topography is different in the eastern Pacific the consequence of the local upwelling is entirely different. Figures 8 and 9 show transequatorial profiles at 95"W about 300 km west of the Galapagos Islands. Considerable work done along the 95"W section (Hayes et al., 1987) has established that the hydrographic conditions and circulation at this location are characteristic of the eastern equatorial Pacific and are free of any island effect caused by the Galapagos archipelago. During the normal years of the E N S 0 cycle the trade winds show a well defined annual cycle in strength with a minimum in the northern hemisphere spring around April and a maximum in the fall around October. To illustrate the conditions during the weak trade wind season of a normal year, Figure 8 shows conditions during April 1982, well before the 1982/83 El NiAo reached 95"W. T o characterize the strong trade wind season, Figure 9 shows profiles from November 1983 after the event had ended. Beginning with strong-trade conditions (November 1983), Figure 9 shows that the surface water was between 20 and 22°C with a nitrate content of 4 to 12 yM. Equatorial upwelling is clearly evident in the temperature profile with doming between 0" and 2"s and relatively high surface layer nitrate values. The Equatorial Front with its characteristic strong temperature and nutrient gradients was present between ION and 2"N. Water north of about I"30'N was nutrient-depleted and contained a chlorophyll concentration of <0.2
Latitude on 9 5 " W
Temperature ("C)
Nitrate (PM)
Chlorophyll ( mg/m3)
Fig. 8. Cross equatorial profiles of temperature, nitrate and chlorophyll along 95"W from 2"N to 2"s. April 1982 (normal conditions) and April 1983 (mature phase of El NiAo). Nine stations sampling 12 depths were used to construct the profiles.
mg/ms. The maximum chlorophyll values were 0.8 to 1.6 mg/ms at 30 m in the strong thermocline and nutricline. During the season of weakest trade winds (April 1982). Figure 8 shows that the surface temperature was between 24 and 26°C and surface nitrate was between 4 and 8 yM. The Equatorial Front was much weaker than in the November 1983 condition and
32
Latitude on 9 5 " W 2 1"s 0 I O N 2 2
IoS
0
ION
2 (u
rn a3
s,
50 24< X< 26
B<X<4
0 . 2 < X (0.4
>
0
z
Temperature ("C)
Nitrate
(PM)
Chlorophyll ( mg/m3)
Fig. 9. Cross equatorial profiles of temperature, nitrate and chlorophyll along 95"W from 2"N to 2"s. November 1982 (onset of El Niiio) and November 1983 (normal conditions). Nine stations sampling 12 depths were used to construct the profiles. vertical temperature and nutrient stratification were also much weaker. In April 1982, the chlorophyll maximum was at the surface between 1"s and 2"s and had values from 0.4 to 0.8 mg/ms. While the degree of surface cooling was much stronger in November 1983, both periods show the high nutrient concentrations that set the eastern equatorial Pacific apart from other tropical ocean regions. The relationship between phytoplankton productivity and nutrient concentration is complex (Dugdale et al., 1981; Eppley, 1981), but for understanding the effects of El Nirio it is appropriate to define nitrate concentrations
4 pM as nutrient-rich conditions
because above this concentration the phytoplankton uptake rate versus nutrient concentration relationship is saturated for oceanic phytoplankton. Using this definition of nutrient-rich conditions, Figures 8 and 9 (April 1982 and November 1983) show that in the absence of El Niiio conditions the equatorial waters of the eastern tropical Pacific are nutrient-rich; that is, they contain saturating concentrations of nitrate in the surface layer. These concentrations support the enhanced levels of new primary production (Chavez and Barber, 1987) that characterize the eastern equatorial Pacific. 5.2 The coastal region Strong and more or less continuous coastal upwelling characterizes the eastern Pacific coastal region of Ecuador, Peru and Chile. Equatorward winds blowing parallel to the coast set a shallow surface layer in motion. The depth of the surface layer directly affected by wind stress is typically <25 m in this region (Barber and Smith, 1981). Because of the Coriolis effect the net transport of the wind-driven layer is 90" to the left of the coastal winds. This transport moves water offshore away from the coast and subsurface water is recruited from a depth of 40 to 80 m
33 to replace the divergent flow. Figures 10, 1 1 and 12 show a 400 km section that starts at the coast and extends offshore to 85"W along 5"s. This section is located close to the Peruvian fishing port of Paita and has been the site of considerable oceanographic work during the last decade (Barber and Chavez, 1983 and 1986; Chavez, 1987). This work indicates that the Paita section at 5"s is a reliable indicator of conditions that prevail in the coastal region from the equator to about 20"s (Chavez et al., 1984). Work along the Peru coast has also established that there is a considerable alongshore coherence in temperature and sea level as well as other properties (Enfield and Allen, 1980; Smith, 1983) so it is possible to use a single location such as the 5"s section as an index of temporal changes that are occurring in the region. On the basis of work in coastal upwelling areas off Peru, California and northwest Africa, Jones et al. (1983) and MacIsaac el al. (1985) have hypothesized that physical/biological coupling leads to the formation of four idealized zones of biological activity within the upwelling circulation pattern. Zone I is the zone of intense upwelling where the water is coolest, rich in nutrients, and low in phytoplankton biomass. As the upwelled water flows offshore to Zone I1 the surface layer is stabilized by solar heating, and phytoplankton adapt to the favorable light conditions and begin to increase their physiological rate processes. In Zone 111, nutrient concentrations are rapidly reduced by fast-growing phytoplankton and there is a rapid increase in biomass; physiological processes are all proceeding at maximal rates. In Zone IV, nutrient depletion occurs and phytoplankton rate processes decrease and eventually biomass decreases as well. The four zone hypothesis is useful to describe the normal upwelling structure. The July 1983 and April 1984 sections in Figure 10 and the November 1983 section in Figure 1 1 show the strong coastal upwelling that is the climatological mean state for this region. In
April 1984, water next to the coast was between 16 and 18°C and contained between 20 and 24 pM of nitrate. Offshore of the coolest and most nutrient-rich water there was a phytoplankton biomass maximum with 6 mg/m3 chlorophyll. Offshore of the chlorophyll maximum, chlorophyll levels decrease to between 1 and 2 mg/m3 but there is no Zone IV nutrientdepleted water on the entire 400 km section from the coast to 85"W. In April 1984, the entire eastern boundary current region was enriched in nutrients and phytoplankton, a condition that is frequently observed in satellite observations of ocean color (Feldman el al., 1984; Feldman, 1985) in this region. The July 1983 section in Figure 10 was occupied a month after the 1982/83 El Nifio ended and shows strong coastal upwelling next to the coast with 20 to 22°C temperatures and 12 to 16 pM nitrate. The chlorophyll maximum is located in the highest nutrient water next to the coast
so Zone I is not evident in this section, but the other zones are evident. Offshore of 84"W the water is strongly nutrient-depleted and low in phytoplankton; this area is clearly Zone IV. The November 1983 section in Figure 11 shows the coastal upwelling pattern characteristic of the strong trade wind season at 5"s. A band of water about 150 km wide next to the coast is occupied by water between 16 and 20°C and between 12 and 24 pM nitrate. Plant biomass is high close to the coast with values up to 4 mg/m5 but offshore the concentration of chlorophyll is still higher with values up to 16 mg/m3. A distinct Zone I close to the coast and Zone I1 which begins 20 km offshore are evident in the November 1983 section. Offshore of
34
Fig. 10. Cross-shelf profiles of temperature, nitrate and chlorophyll along 5"s from the coast offshore to 85"W in July 1983 (normal conditions) and April 1984 (normal conditions). The inshore chlorophyll maximum in July 1983 was 20 mg/m3. Five stations sampling 9 depths were used to construct July 1983 profiles and nine stations sampling 9 depths were used to construct the April 1984 profiles.
Fig. 11. Cross-shelf profiles of temperature, nitrate and chlorophyll along 5"s from the coast offshore to 85"W in November 1983 (normal conditions) and November 1982 (onset of El Nifio). The inshore chlorophyll maximum in November 1982 was 2 mg/m3. Ten stations sampling 10 depths were used to construct the profiles.
35 the 16 mg/m3 water a zone about 100 km wide has chlorophyll concentrations of 2 to 4 mg/m3 and nitrate of 8 to 16 pM. No nutrient depleted water was present in the 400 km band adjacent to the coast. The July 1983, November 1983 and April 1984 coastal sections and the April 1982 and November 1983 equatorial sections show the normal eastern tropical Pacific condition; that is, the local winds are favorable for upwelling and the large-scale winds have set up the basinwide topography so that the thermocline and nutricline are shallow along the equator and eastern continental margin. The combination of local and large-scale dynamics has produced the equatorial and coastal nutrient enrichment that is the normal state of the eastern tropical Pacific waters.
Fig. 12. Cross-shelf profiles of temperature, nitrate and chlorophyll along YS from the coast offshore to 85”W in March 1983 (mature phase of El Niiio) and May 1983 (mature phase of El Niiio). The inshore chlorophyll maximum in March 1983 was 6 mg/m3 and in May 1983 it was 2 mg/m3. Seven stations sampling 10 depths were used to construct the March 1983 profiles and eight stations sampling 9 depths were used to construct the May 1983 profiles.
6 EASTERN PACIFIC ANOMALOUS CONDITIONS 6.1 The Equatorial Region The temporal development of the 1982-83 event was introduced by D.V. Hansen in the preceding contribution. Additional appreciation of the causal ecological aspects of the temporal progression of El NiAo are illustrated in a Galapagos Islands time series study that began in June 1982 and ran until December 1983. This study measured ocean properties three times a week at
a station in 100 m of water offshore from Academy Bay, Santa Cruz Island (Kogelschatz et al., 1985). The progression of surface and 60 m temperatures shown in Figure 13A establish that El
36 30
I I
A.
25 h
Y E
v
2
z 20 a
:
I-
15
I 0
1 10
1
I + 2
I
V
1
3
I li
*
I
4
1
5
I
J ' J ' A ' S ' O ' N ' O
I
J ' F ' M ' A ' M ' J ' J ' A ' S ' O ' N ' C
1982
1983 1
-E
3
25 4 .
L
3
+
2
a
5 20
I-
15
15
20
25
30
Temperature (0m) 1
Fig. 13. A. Time series of tri-weekly surface and 60 m temperature observations made off Santa Cruz Island in the Galapagos Islands during 1982 and 1983. Bold line, 0 m; lighter line, 60 m. B. A plot of 60 m temperature as function of the surface temperature. The numbered arrows show the direction of temperature change. The values are represented by symbols shown in 13A for each time interval. Note that during interval 1 the 60 m temperature increases but the surface temperature is unchanged. NiAo is well defined by both the surface and 60 m temperatures; in fact, the onset of the event in the Galapagos Islands in September 1982 and recovery in June 1983 are stronger and more sharply delineated in the 60 m temperature signal. An interesting characteristic of El Niiio is that an increase or decrease in temperature at the ocean surface is always preceded by a change in temperature at 60 m; as noted previously (Enfield, 1981; Barber and Chavez, 1983), El Niiio is initially a subsurface thermocline anomaly that secondarily propagates to the ocean surface. To emphasize that subsurface changes always lead the surface changes, Figure 13B plots the surface temperature against the 60 m temperature. In Phase 1 of the event beginning in
37 September 1982, the 60 m temperature increased, but the surface temperature remained steady at 22 to 24°C. T o estimate the duration of the event, September 1982 is considered the first abnormal month. In Phase 2, from October through December, both the surface and 60 m temperature increased steadily by about 5°C. In the mature phase El Niiio condition (Phase 3 in Figure 3B), the 60 m temperature decreased while the surface temperature remained around 29°C. In June 1983, the event entered a final phase as both surface and subsurface temperatures fell rapidly towards the climatological mean. By September 1983, surface and subsurface temperatures had returned to the values that were present at the beginning of the event in the June to September 1982 period. The progression of changes in nutrient concentration in the region affected by El Niiio is shown in Figure 14 in which nitrate is used as an indicator of the nutrient conditions. In general, phosphate and silicate changes showed a pattern parallel to nitrate (Kogelschatz e t al., 19851, but in Pacific waters there is usually a slight atom ratio excess of silicate-silicon and phosphate-phosphorus relative to nitrate-nitrogen; therefore these two nutrient elements do not become as depleted in the surface layer. For describing the temporal changes associated with the 1982/83 event, it is irrelevant which inorganic nutrient is used because the El NiAo nutrient signal is strong in all three. A strong parallel with the progression of temperature changes is evident in the nitrate record.
Fig. 14. Time series of tri-weekly observations of nitrate made off Santa Cruz Island in the Galapagos Islands in 1982 and 1983. Bold line, 0 m; lighter line, 60 m. From Kogelschatz ef a ] . , (1985). Surface layer values at the beginning and end of the time series have nitrate ranges from 2 to 6 pM. As mentioned earlier, concentrations in this range saturate the nitrate uptake capability of 0
38 marine algae so these conditions are highly favorable for plant growth. What is extremely unusual about the eastern equatorial Pacific is that a nutrient-rich condition persists throughout the year during both the strong and weak trade wind seasons (Barber and Chavez, 1986). Figure
14 shows that the surface layer nitrate concentration during normal conditions (June through August 1982 and August through November 1983) is relatively variable, ranging from 2 to 6 pM, but during normal conditions the waters of the equatorial (and coastal) region are never nutrient depleted (Barber, 1988). Spectral analysis of the 0 m and 60 m nitrate time series showed that at both depths the variability had a periodicity characteristic of a lunar tide cycle such as that described by Bowman and Esaias (1981) in Long Island Sound. It appears that in the Galapagos archipelago tidal mixing is the major local process that transports nutrients to the surface given a favorable, shallow nutricline position. If one accepts that tidal mixing is one process involved in the vertical transport of nutrients, the time-series in Figure 14 shows how the basinwide nutricline topography affects overall control on nutrient supply to the surface waters. Tidal mixing must have continued unabated during the development of the 1982/83 event, so the local transport process was present in the period from November 1982 to June 1983; however, because of the deep nutricline there was no surface nitrate expression of the tidal mixing. Figure 14 also shows that strong lunar periodicity was present at the surface before and after the event. It was also present at 60 m during the event from January 1983 through May 1983. The most important information in Figure 14 is that for six and a half months, from December 1982 to mid-June 1983, the surface layer nitrate was very low, often below the detection limit of 0.02 pM. Such a concentration of nitrate clearly constitutes nutrient-poor conditions for plant growth and will necessarily decrease primary production. At 95”W in November 1982 (Figure 9), the oceanographic character of the equatorial region was changed. Thermal stratification was absent in the upper 75 m and the entire equatorial band from
2”s to
2”N was occupied by warm water (24 to 26°C) that contained from 4 to 8 pM of nitrate. The surface layer was warm, nutrient-rich, well mixed and, as shown in Figure 9, extremely low in phytoplankton-chlorophyll. A deepened mixed layer would reduce primary production by means of the Sverdrup critical-depth relationship (Sverdrup, 1953). The critical depth, as defined by Sverdrup (l953), is the depth at which the total phytoplankton photosynthesis of a water column is equivalent to the total phytoplankton respiration. The critical-depth relationship suggests that as the depth of mixed layer increases and approaches the critical-depth then net growth of the phytoplankton population will decrease. The changes observed at this time may have been partly driven by the critical-depth relationship but the decrease in productivity seemed also to be related to a qualitative change in the kind of phytoplankton. That is, there was a shift from diatoms to motile forms. The decrease in stratification apparently enhanced diffusive and sinking losses of the non-motile diatoms. In the Figure 8 profile from April 1983, the mature El NiAo conditions are expressed. The surface-layer temperature is around 29°C and there is relatively strong stratification in the upper 75 m. The stratification is caused by the increase in temperature and dramatic decrease in
salinity. Low salinity water was present in the eastern equatorial Pacific from December 1982 to
39 June 1983 (Kogelschatz el al., 1985). This is the period of anomalously heavy rainfall, so the low salinity surface water is assumed to result from local processes, not the advection of water from north of the equator (Hayes ef al., 1987). Nitrate is below detection in the surface layer where temperatures are >28"C. During this period there is a strong nutricline from 40 to 80 m where nitrate increases from 4 to 12 pM and there is some expression of a subsurface chlorophyll maximum associated with the nutricline; but surface concentrations of chlorophyll remain quite low, around 0.2 mg/ms. In general, the physical, chemical and biological conditions during the mature phase from December 1982 to mid-June 1983 resemble the typical tropical oceanic conditions that characterize a low-latitude gyre. Associated with the gyre-like oceanic conditions there are changes in the species assemblages. In Table 1, phytoplankton species identifications and counts from April 1983 are compared with April 1966. In April 1966 at 92"W and the equator, the water was 20 to 22"C, and had 10 to 12 pM nitrate and about 0.2 mg/ms chlorophyll (Barber and Ryther, 1969). If the April 1966 conditions are assumed to be "normal," the comparison in Table I demonstrates the biotic changes associated with the anomalous April 1983 condition. The major difference is that there is a 10 fold increase in the abundance of coccolithophores in April 1983 and a decrease in the abundance of diatoms. Interestingly the evenness of the diatom assemblage is much stronger during April 1983. The 1966 dominant diatom Nifzschia delicafissima decreased two orders of magnitude from lo4 to 10' cells/liter. The phytoplankton biomass expressed in chlorophyll concentration is remarkably similar (about 0.2 mg/ms) despite the difference in oceanic conditions, but while the chlorophyll concentration was similar the productivity was about an order of magnitude higher in April 1966 (Barber and Chavez, 1983). The November 1982 to April 1983 progression on the equator emphasizes that as El NiAo develops dramatic changes are taking place. The warm, nutrient-rich, well mixed, low phytoplankton biomass and low productivity condition develops into a very warm, nutrientdepleted, strongly-stratified and moderately-low biomass condition. To estimate the quantitative impact of the 1982/83 event, Chavez and Barber (1985) estimated the duration of the two phases on the basis of nutrient conditions observed in the Galapagos time series. Onset phase is defined as the period when the nutricline deepens but surface nutrient concentrations remain relatively high, about 4 to 8 pM. For the 1982/83 El Nirio, the onset phase extended from September 1982 through November 1982 (90 days). By December 1982, nitrate in the surface layer was depleted to concentrations which limit nutrient uptake (Dugdale, 1967). The low nitrate, or "mature" El Nirio, conditions extended from December 1982 through June 1983 (210 days). The return of surface nitrate concentrations to normal levels took place in mid-July 1983. Large-scale, horizontal, sea surface temperature distributions show reestablishment of the equatorial circulation (Halpern, 1987) and the cool equatorial tongue starting at this time, indicated that throughout the entire equatorial zone normal conditions were returning. 6.2 The coastal region The progression of El NiAo conditions along the west coast of South America can be seen by comparison of the November 1982 (Figure I I ) , March 1983 and May 1983 (Figure 12) sections.
40 TABLE 1 Phytoplankton species composition at 92"W and the equator during April 1966 and April 1983. The April 1966 identifications and counts were provided by Edward Hulburt of Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA. The April 1983 information was provided by Terasa Arcos of the Instituto Nacional de Pesca, Guayaquil, Ecuador.
APRIL 1966 "Normal"
APRIL 1983 El Nfio cells/liter
cells/liter COCCOLITHOPHORES Cyclococcolithusfiagilis Cyclococcolithus leptopoms Eniiliana huleyi Rliabdosphaera slylifer Total
200 50 150 50 450
DIATOMS Asteromphalus heplactis Chaetocerospew vianus Coscinodiscus sp. Heniidiscus cuneifonnis Nitzschia bicapitata Nitschia closteriuni Nitzschia delicatissima Rliizosolenia alata Rliizosolenia bergonii Thalassiotlirix sp. Unidentified pennate diatom Unidentitied diatom
100 50 600 50 100 150 10,400 100 50 100 200 5,200
Total
17,100
DINOFLAGELLATES Emviaella vagi.nula Oxytoxuni variabile Unidentified dinoflagellate
150 50
Total
250
TOTAL CELLS
50
17.800
COCCOLITHOPHORES Acanthoica sp. Cyclococcolithusfraglis Cyclococcolithus leptopoms Coccolithus pelagicus Discosphaera tubifera Emiliana huleyi Helicosphaera sp. Ophioster hidrotius Syracosphaera sp. Unidentified coccolithophore
1,146
Total
4,270
DIATOMS Chaetoceros sp. Hemialus sinensis Thalassiossira sp. Niltrchio sp. Nitzschia delicatissima Nitzschia longiksima Rliizosolenia sp. Rh izosolenia fiagidissinia Rliizosolenia stolterfothii Rhizosolenia alata Ceratulina sp.
521 938 1,536 104 104 833 104 1,875 1,979 208 104
Total
9,583
729 208 104 104 208 938 104 104 625
DINOFLAGELLATES Ceratiuni furca
104
Total
104
TOTAL CELLS
13,957
The November 1982 section was sampled during the onset phase of the event as defined by Chavez and Barber (1985). The temperature anomaly had reached 5"s in late September 1982 (Chavez et al., 1984) so that by November 1982 (Figure 11) the characteristic thermocline and nutricline depression had occurred. The 20°C isotherm, which is the middle of the thermocline,
41
was depressed about 100 m when compared with the November 1983 section. The onshore/offshore surface temperature gradient (Figure 15) is similar in magnitude (24 to 27°C in 1982 and 17 to 22°C in 1983) even though the surface temperatures are about 5°C warmer during November 1982. The existence of the onshore/offshore gradient is evidence that coastal upwelling continued during the onset phase of the event. Coastal wind observations have established that in previous El Nifio events (Enfield, 1981) and in the 1982/83 event (Smith, 1983; Huyer el al., 1987) upwelling favorable winds blowing equatorwards along the coast continue during most of the event.
In the 1982/83 event, coastal
winds and locally forced upwelling were present from the beginning of the anomaly in September 1982 until April 1983; the anomalous warming in the continued presence of upwelling is the result of large-scale thermocline changes. The thermocline was progressively depressed in the September to November 1982 period so that the water entrained into the upwelling circulation at 40 to 80 m depth was about 5°C warmer than average. In November 1982 (Figure I I ) , nutricline depression had occurred to the degree that there was a dramatic reduction in nutrients in the offshore and middle portions of the 5"s section, but an inshore band 75 km wide continued to contain nitrate of 4 to 8 pM. These inshore concentrations are still highly favorable for phytoplankton growth and in the inshore band Chavez (1987) reported that there were concentrations of chlorophyll of 2 to 10 mg/ms. These high concentrations were associated with a bloom dominated by the diatom Asterionella japonica. A comparison of November 1982 and November 1983 demonstrates one important characteristic of El Niiio: a major ecological change during El Nitio is a reduction of the size of the upwelling habitat. In November 1982, the 1982/83 event was clearly underway and many upwelling species had begun to be affected (Barber and Chavez, 1986), but in the narrow inshore band, highly productive conditions continued. The major habitat changes in November 1982 were offshore and subsurface. In strong events like 1982/83, these changes continue to move progressively inshore and towards the surface layer so that eventually the entire coastal upwelling habitat is affected. In weak or moderate El Nitio events, it is likely that the inshore band remains a refuge for species requiring the conditions provided by upwelling. In a weak event like 1975 the inshore coastal band remained very productive (Cowles
el
al., 1977) and in the
moderate event of 1976 nutrient concentrations decreased (Dugdale ef al., 1977) but phytoplankton productivity remained relatively high (see Table 4). Barber and Chavez (1986) speculate that behavior that leads fish to concentrate in the inshore upwelling refuges is adaptive for weak and moderate events, but this same behavior proves lethal in exceptional events like that of 1982/83 because fish become trapped in the inshore pockets instead of moving poleward ahead of the warm water anomaly that is progressing along the coast. The March 1983 section in Figure 12 shows that the thermocline was slightly more shallow than it was in November 1982 and that coastal upwelling continued as shown by the upward tilt of isotherms and isopleths of nitrate and the continued onshore/offshore gradient in surface temperature (Figure 15) and nitrate concentration. High chlorophyll concentrations > I .O mg/ms in March 1983 were limited to a 40 km wide band next to the coast. In contrast, the April 1984 section in Figure 10 shows that the band of high chlorophyll extended 400 km from the coast
42
Longitude on
5"s
85 84 83 82 81
Longitude on 5"s 85 84 83 82 81 I
I
l
l
Longitude on 5"s 85 84 83 82 8l0W
,
Nov 82 25
\
J u l y 83
\ A p r i l 84
20
I5
i
20
%I
Nov 83
Nov 83 10
OJ I
I
I
I
i
r
I
400 200 0 400 200 0 400 200 0 Kilometers Offshore Kilometers Offshore Kilometers Offshore Fig. 15. Surface temperature, chlorophyll and primary production along 5"s from 85"W to the coast of Peru showing the comparison between El Nixio (November 1982, March 1983 and May 1983) conditions and November 1983, April 1984 and July 1983 during normal conditions. From Barber and Chavez (1986). across the entire eastern boundary current system. The May 1983 section in Figure 12 shows the strongest development of the 1982/83 El Niiio. The surface temperatures were about 12°C above the climatological monthly mean of 16.7"C and the 20°C isotherm was depressed below 100 m. In the May 1983 section the isotherms and isopleths of nitrate slope downwards towards the coast indicating that flow was onshore and poleward in the coastal region. The chlorophyll distribution in May 1983 shows that the typically
43 rich eastern boundary region had a biomass character typical of a low-latitude ocean gyre. The highest chlorophyll values found in the May 1983 section were within 5 to 10 km of the shore and were in very low salinity, very high nitrate water. The heavy rainfall that occurred in April and May 1983 washed seabird guano off coastal rocks and formed a thin, but stable lens of high nutrient water. This high nitrate surface lens was limited in area and it varied weekly from being undetectable to having a strong local signal. PRODUCTIVITY EFFECTS O F EL NlNO 7. I The equatorial regiori Hydrographic observations taken during the 1982/83 event indicate that for about 210 days, from December 1982 through June 1983, the surface layer concentration of nitrate was very low, often below the detection limit of 0.2 pM, for an enormous expanse of the equatorial Pacific. Satellite temperature observations such as those shown in Figure 3 indicate that the entire equatorial cold tongue region was occupied by anomalously warm water. The area normally occupied by the cold tongue reaches from the coast of South America at about 80"W westward to the dateline at 180" (Levitus, 1982); therefore, the region affected reaches over about a quarter of the circumference of the earth, roughly 90" of longitude. Over this region the supply of nitrate and other nutrients to the euphotic zone was significantly reduced, necessarily decreasing new primary production. New production is defined as primary production dependent on newly available nitrogen, for example, NO3-N (Dugdale and Goering, 1967) from the deepwater nutrient reservoir.
Eppley and Peterson (1979) argue that
new production is "quantitatively equivalent to the organic matter that can be exported from the total production in the euphotic zone without the production system running down." Therefore, new primary production is newly synthesized organic matter available for export by the higher trophic levels of the food web. This single effect, the reduction of new primary production as a result of nutrient denial, is a consequence of El NiAo that affects the entire ecosystem. Estimates of equatorial productivity and the 1982/83 productivity anomaly made by Chavez and Barber (1985 and 1987), and summarized in Table 2, show that the total primary production rate during the mature phase of El Niiio was 21 to 26% of the rate during normal conditions. The empirical relationship between new production and total primary production found by Eppley and Peterson (1979) indicates that as total production falls below 300 to 400 mgC/m2/day the proportion of new production decreases sharply. As no direct measurement of new production are available for equatorial waters during this period, the global relationship was used to calculate new production. Clearly the new production rate during El NiAo was reduced more strongly than total production. Table 2 shows that the El Niiio rate of new production during the mature phase was only 5 to 6 percent of the normal rate. This 20 fold reduction in new primary production was a pervasive biotic change that affected the entire food web because the equatorial food web is adapted to continuously high levels of new primary production (Vinogradov, 1981). 7.2
The coastal region' In the low latitude coastal regions of the eastern Pacific, from the equator to about 2 0 3 ,
primary productivity is high. On an annual basis it is probably higher than in any other ocean
44 region (Barber and Smith, 1981). The difference between the eastern Pacific coastal upwelling area and other coastal and oceanic environments is a matter of quantity: the large-scale nutricline topography and more or less continuous, local, upwelling favorable winds cause the annual flux of new nutrients to the euphotic zone to exceed that of other regions. Thus, the annual new primary production is high and the quantity of organic material that can be exported from the euphotic layer as a commercial fish catch, loss to sediments and loss to adjacent intermediatedepth waters is much higher. Determining the climatological mean level of primary production is difficult because the same mechanics that make the region highly productive also make it TABLE 2 Comparison of El Nifio and normal total and new primary production rates in the equatorial Pacific. Total production was converted to new production using the model of Eppley and Peterson (1979). This table is modified from Chavez and Barber (1985) with additional data.
Normal
Rate fmnC/m2/dav)
Anomalv (96)
El Niiio
El Niiio/Normal
Onset
Mature
Onset
Mature
Total Primary Production Zonal (0"; 90"W to 180") Meridional (95"W; 2"N to 5 3 )
490 (n=132) 260 (n=8) 605 (n=8) 225 (n=7)
125 (est) 125 (n=8)
53 37
26 21
New Primary Production Zonal (0"; 90"W to 180") Meridional (95"W; 2"N to 5 % )
214 265
14 14
29 20
6 5
63 54
inherently variable (Barber el al., 1985). The ENS0 cycle is an inherent source of interannual variability because it behaves like an aperiodic oscillator (Cane and Zebiak, 1985). Table 3 presents the results of studies carried out in the coastal region during non-El Niiio years in the last two decades with a mean value of 3,840 mgC/m2/day with a standard error of f 250 mgC/m2/day.
Syntheses of existing measurements suggest that the climatological mean value is
between 2,000 and 4,000 mgC/m2/day (Barber and Smith, 1981; Walsh, 1981; Barber ef al., 1986; Chavez and Barber, 1987). Chavez et al. (1988) used a 29 year time series of wind speed and thermocline depth as inputs to a model that estimated the supply of nitrate to the surface layer and calculated the new primary production that would result if all the nitrate was taken up by phytoplankton and converted to new production. In the Peru coastal waters, Dugdale (1985) made a direct measurement of the f-ratio using nitrogen stable isotopes. With this empirical relationship new production was back converted to total production. Calculating total primary production for approximately the same periods that the studies in Table 3 examined, the model predicted a mean value of 3,180 f 70 mgC/m2/day. The agreement of the measured values (3,840 f 250) and the model estimate (3,180 f 70) to within 20% is remarkable and indicated that it would be valuable to evaluate the model predictions for mean productivity during El Niiio events. Chavez el al.
45 (1988) report that during El Nirio events the model calculates a productivity of from 1,700 mgC/m2/day in 1983 to 2,200 mgC/m2/day in 1965. The predicted values are higher than expected because the local wind increases significantly during El Niiio so that there is simply more upwelling; at the same time the nutrient content of upwelled water is so much lower that the model calculates about 50% less total primary production in the coastal region during El Niiio events. However, hydrographic observations indicate that the "El Niiio" prediction of the model may be too high for the 1982/83 El Niiio event. Measurements of currents, pressure and local winds during the 1982/83 event by Huyer el al. (1987) indicate that local upwelling favorable winds did not continue to force upwelling throughout El Nirio. At 10"s Huyer el al. (1987) found that upwelling ceased in May 1983 TABLE 3 Estimates of primary productivity from along the west coast of South America. From Chavez and Barber (1987).
Year
Month
1966 1969 1969 1974 1971 1977 1978
April April June February March November February
Productivity (mgC/m2/day) 6,260 (Barber and Smith, 1981) 5,160 (Barber and Smith, 1981) 1,240 (Walsh, 1981) 4,160 (Sorokin, 1978) 1,960 (Barber and Smith, 1981) 4,260 (Harrison and Platt, 1981) 3.830 (Barber el al., 1986) mean=3,840 f 250
despite continued strong and favorable for upwelling local winds. This cessation of upwelling occurred because the wind-induced offshore flow in the surface layer was balanced and overridden by an onshore geostrophic flow driven by the pressure gradient. The steric height along the shelf break was significantly higher close to the shelf break at 5"s than at 10% setting up a strong north to south pressure gradient. The Huyer et al. (1987) analysis indicates that geostrophic suppression of upwelling and offshore flow were present from mid-March 1983 to mid-June 1983. During this period the wind-driven primary production model of Chavez et al. (1988) would significantly overestimate productivity. This evidence suggests that during the 1982/83 event the inshore productivity supported by wind-driven upwelling may have been significantly less than the 1,700 mgC/m2/day value predicted by the model. Nevertheless, direct measurements of phytoplankton biomass and productivity (Figures 12 and 15) emphasize another important aspect of El Niiio: despite decreases in primary productivity the nearshore (<30 km) coastal region remains relatively productive, especially when compared with other tropical regions. This aspect of El Nifio accounts for population increases reported in the reviews by Arntz (1986) and Glynn (1988). After the local cool water inshore organisms were impacted by the warm conditions, other species, especially eurythermal ones, were able to invade the intertidal and inshore regions and grow rapidly. Compared with most warm water
46 environments the El Niiio conditions along the coast were relatively rich in food. With temperatures of 28°C and an estimated mean productivity of around 1.000 & 500 mgC/m2/day, the conditions in March 1983 (Figures 12 and 15) would provide eutrophic tropical conditions. Barber and Chavez (1986) believed that this nearshore food source supported the population blooms that occurred during the 1982/83 event. For example, the scallop Argopecten purpuratus underwent a rapid increase in abundance during the anomaly; in 1983 the harvest of scallops increased 40-fold to about 20,000 tons, but by January 1984, the population increase was over and the abundance had returned to its previous level (Arntz, 1986). Scallops are residents of the inshore waters of Peru; the adults are sessile and their planktonic larvae do not survive long distance transport, so it is hypothesized that the extraordinary population increase resulted from in situ increases in growth and reproduction. Another species of the genus Argopecten had
maximum growth at 28°C with a chlorophyll concentration of 2.4 mg/m3 (Kirby-Smith and Barber, 1974); the Peruvian scallop beds contained chlorophyll concentrations in this range in February 1983 when the water was 28°C. Although the phytoplankton biomass was reduced from the climatological mean, it remained high enough in the nearshore waters to support dramatic, but very short-lived, population increases in a number of species (Arntz, 1986; Glynn, 1988). These explosive increases remind us that the inshore food web was modified, but not uniformly devastated, by El NiAo. While the inshore maximum in productivity and phytoplankton biomass persisted throughout the 1982/83 event, from 30 to 200 km offshore the decrease was dramatic. Because the eastern boundary current region from the coast to about 400 km offshore has a seasonally varying productivity gradient during normal years, it is difficult to calculate the quantitative productivity decrease across this gradient for the period of the anomaly from only three cruises in November 1982, March 1983 and May 1983. Figure 15 shows that at 85"W, about 400 km off the coast, there was no difference between productivity and chlorophyll values in November 1983 vs November 1982 or in July 1983 vs May 1983, but in March 1983 vs April 1984 the 85"W surface productivity was significantly reduced. In the middle portion of the section, from 30 to 200 km offshore, there was a significant reduction in all three of the normal vs El NiAo comparisons shown in Figure 15. These waters, 30 to 200 km offshore, are the habitat of the commercially important fish species (anchoveta, sardine, mackerel, jack mackerel and hake). Reviews of the biotic impacts of the 1982/83 El NiAo by Barber and Chavez (1983 and 1986), Arntz et al. (1985) and Arntz (1986) reported that these species expressed species-specific responses to the progressive development of the anomaly. Of the important fish stocks in this region, anchoveta experienced the greatest mortality and reproductive failure during the 1982/83 El Niiio. This vulnerability can be explained in general terms if we assume anchoveta is the species best adapted to exploit the continuously high biological productivity of the region. Being best adapted to high productivity environments, it is the species most affected by the reduction in productivity that occurred in the coastal region 30 to 200 km offshore. Measurements made during the 1982/83 anomaly indicate that off Ecuador, Peru and Chile the fish were low in fat content, low in weight for a given body length and short for their age (see the review by Arntz, 1986; Santander and Zuzunaga, 1984). An indication of
47 the severity of nutritional stress as a result of low productivity is seen in the effects of the 1982/83 El NiAo on the northern anchovy (Engmulis mordax) off California where productivity decreases were similar but less than those off South America (Fiedler et al., 1986). Among the northern anchovy: (1) the body weight of spawning females was very low in 1983-84; (2) spawning occurred at smaller sizes than ever before observed; (3) growth of juveniles was 36% less than previous years; and (4) the spawning biomass was the lowest since 1962. Because the reduction in plankton biomass and productivity was proportionally much greater in the habitat of the Peruvian anchoveta, the increased reproductive failure and adult mortality off South America are not surprising. In 1983 there was a complete recruitment failure in that no larval anchoveta were found in the 1983 larval fish survey (Santander and Sandoval, 1985). One interpretation of the well established link between anchoveta variability and sea surface temperature anomaly (see, for example, Figure 9.2 in Barber, 1988) depends on the link between surface temperature, thermocline depth, nutricline depth and new primary production. (See the review edited by Pauly and Tsukayama (1987) for another interpretation.) There exists, however, an additional aspect of ENSO-driven phytoplankton changes that needs to be considered; that aspect is the shift in phytoplankton species composition from a diatom dominated community to a dinoflagellate and/or coccolithophore community. Table 1 shows the modest degree of diatom to coccolithophore shift that took place in equatorial waters in 1983 vs 1966. Along the coast of South America a particularly dramatic species shift also took place during onset of the 1976 El NiAo. This event is classed as a moderate strength El Niiio (Quinn ef al., 1987), but it had severe consequences on the anchoveta. At the onset of the 1976 El NiAo in mid March, a sudden bloom of Gymnodinium splendens took place in the coastal waters along the west coast of South America from about 2's to 1 5 3 . Unusual features in 1976 were the extent of the bloom (over 1,000 km of the coast), its duration (from March until the end of May), and the quantitative dominance of one species of dinoflagellate over the normal assemblage of diatoms (Huntsman ef al., 1981). What conditions favored the dominance of Gymnodinium splendens in 1976? The major differences in the mean conditions that distinguish 1976 from the mean of three "normal" years were increased stability and lower nitrate concentration in the surface layer (Table 4). The vertical gradient in density was stronger than normal in the upper 10 m in 1976 and the static stability, expressed as the square of the Brunt-Vaisala frequency, N2, (Phillips, 1977) was twice the average in 1976 (Table 4). Hence, the euphotic portion of the water column was unusually stable for this time of year, and this increased stability apparently gave Gymnodinium splendens a physical environment it could exploit with its motility and its positive phototaxis. The idea that decreased turbulence favors dinoflagellates over diatoms and vice versa, argued most convincingly by Margalef (1978) and Margalef et al. (1979), is supported by these results. The variance of nitrate concentration was larger in 1976 than in other years: the surface layer frequently had undetectable concentrations followed by episodes of 10 pM (Dugdale et al., 1977; MacIsaac et al., 1979). Increased stability made it possible for Gymnodinium splendens to maintain itself in the euphotic layer during the unfavorable episodes of very low nitrate. Dugdale (1979) suggests that once dominant, a surface canopy of dinoflagellates denies light to
48
the nonmotile phytoplankton that are lower in the water column insuring that the dominance will continue until disrupted by surface layer mixing. In 1976, dense Gyrntzodiniurn splendens aggregations concentrated at the sea surface during mid day with chlorophyll concentrations of over 200 mg/m3 (Dugdale ef al., 1987; Wilkerson TABLE 4 Mean values of euphotic layer properties for stations within 35 km of the shore at 15"s on the coast of Peru. Values for each station are obtained by integration from the surface to the I % light depth; the station values are averaged to obtain a mean. During March/April 1976 a moderate strength El Niiio was present; the other years were normal (Quinn ef al., 1987).
Primary Productivity (msC/m2/day) April 1966 n=15 March/April 1976 n=27 March/April 1977 n=24 February/March 1978 n=30
Stability (N2) (cycles2/hour2)
Nitrate (/AM/m2)
6,260
36
275
1,680
64
40
1,960
28
498
3,830
21
I30
ef al., 1987). In these aggregations the chlorophyll specific productivity rate was very low, not distinguishable from zero (MacIsaac ef al., 1979). However, individual dinoflagellates in the aggregations were motile and able to orient to light (Blasco, 1979), indicating that this organism can remain viable and maintain itself in the upper few meters of the sea surface while growing slowly or not at all. Diatoms cannot do this; they depend o n high growth rates (relative to dinoflagellates) and buoyancy adjustments (using photosynthesis derived energy) and local turbulence to keep them suspended in the euphotic zone (Anderson and Sweeney, 1978; Smayda, 1970 and 1980). Diatoms must grow actively or sink out of the euphotic layer; Gymnodiniurn splendens stops growing when conditions are unfavorable (episodes of zero nitrate) but can maintain itself in the surface waters ready to exploit the next nutrient input. A dramatic aspect of the Gyrnttodinium splendens bloom in 1976 was that the organism apparently had become dominant along the entire coast from
2"s to
15"s almost concurrently
(Bass and Packard, 1977; Rojas de Mendiola, 1979). The almost simultaneous dominance of this dinoflagellate along 1,000 km of coast indicates that a large-scale, rapidly propagating process was responsible for setting up the stability anomaly that favored Gymnodittiurn splendens. Model hindcasts show that a sharp trade wind decrease in early 1976 set off a train of equatorially trapped waves that reached the eastern margin (O'Brien ef al., 1981; Busalacchi el a/., 1983), then propagated poleward and produced the observed warming and stability anomalies in the coastal region (Smith, 1978 and 1983).
49
Gymnodinium splendens is excellent food for larvae of anchoveta, but the change to a dinoflagellate dominated plankton in 1976 had an extremely detrimental effect on the growth and spawning of the adult fish (Valdivia, 1978; Rojas de Mendiola, 1979). Anchoveta gain most of their adult weight and lay down lipid reserves during the March through May period when productivity is highest; by May 1976 the fish had reduced fat content, reduced weight for a given length, and reduced length at sexual maturity. The dominance of Gymnodinium splendens during March through May 1976 placed the growing fish under nutritional stress. In early 1977, recruitment of juvenile fish spawned by adults that had matured during the 1976 Gymnodinium splendens bloom was very poor, worse than recruitment during the 1972 El NiAo (Santander and
Tsukayama, 1983). but not as low as that in 1983. During the 1982/83 anomaly a large number of researchers (Avaria, 1984 and 1985; Munoz, 1985; Rojas de Mendiola et al., 1985; Ochoa el al., 1985; Chavez, 1987) reported that there were numerous diatom-to-dinoflagellate and diatom-to-coccolithophore shifts in the coastal phytoplankton species assemblage. No single, persistent, large scale dominance shift took place in 1982-83 as in 1976, but significant changes in the phytoplankton species composition developed along the coast of South America from Ecuador to Chile. Direct stomach analyses indicate that El NiAo causes both quantitative and qualitative changes in the phytoplankton food supply for higher trophic levels (Sanchez de Benites et al., 1985). Assuming that dinoflagellates and coccolithophores are a less favorable food ration than diatoms, the decrease in quality and quantity of the food supply imposed significant nutritional stress on fish and zooplankton grazers during the 1982/83 El NiAo. 8 CONCLUSIONS 1. The normal trade wind phase of the E N S 0 cycle is responsible for the usual asymmetry of
heat storage, nutrient availability and productivity of the tropical Pacific. The western Pacific in the climatological mean condition is warm and low in nutrients and productivity; the eastern Pacific is cool, rich in nutrients and highly productive. 2. El Niiio is an aperiodic redistribution of heat and water that episodically serves to decrease the normal east/west asymmetry. 3. The exceptionally strong 1982/83 El Niiio took 90 days to develop (September through November 1982) and this onset phase was characterized by warm anomalies, thermocline depression, weak stratification, moderate nutrient concentrations and low phytoplankton abundance. 4. The mature phase of the 1982/83 anomaly lasted about 210 days from December 1982 through
June 1983 and this phase was characterized by very warm surface temperatures, strong stratification, nutrient-depletion and low productivity.
5. Along the equator, from about 90"W to 180", during onset (September through December 1982) total primary production was estimated to be 53% of normal and new primary production was 29% of normal; during the mature phase total production was 26% of normal and new production 6% of normal. 6. In the coastal region, an inshore band of about 40 km remained relatively productive
50 throughout most of the 1982/83 event, but the size of the productive upwelling environment was reduced to about 10% of normal. 7. Continued coastal winds during the 1982/83 event forced coastal upwelling but the depressed nutricline reduced the quantity of nutrients upwelled, causing levels of primary production about
50 to 20% of normal in the narrow inshore band. A model forced with observed winds predicted 1,700 mgC/m2/day but a more realistic value may be 1,000 f 500 mgC/m2/day because of geostrophic suppression of upwelling. 8. Phytoplankton species composition changed from a diatom dominated community to
dinoflagellate or coccolithophore dominated communities reducing the nutritional value of the available phytoplankton for higher trophic levels. The qualitative and quantitative decreases in newly synthesized organic matter placed severe nutritional stress on the coastal food web causing reduced growth, poor physiological condition, mortality, and failed reproduction in fish and zooplankton grazers. ACKNOWLEDGEMENTS The research was supported by the National Science Foundation under grant OCE-86-13759 and by the US. TOGA program of NOAA under grant NA85AA-D-AC089.
We thank the
governments of Peru, Ecuador and the People's Republic of China for their encouragement and support of this work. The effort in the western Pacific was part of the US/PRC Cooperative Air-Sea Interaction Study. The work in the central and eastern Pacific was part of NOAA's Equatorial Pacific Ocean Climate Study (EPOCS). We express our special thanks to the crews of the U.S. and P.R.C. ships that made this work possible and enjoyable. We also wish to thank Andrea Sanico for typing the manuscript. 9 REFERENCES Anderson, L.W.J. and Sweeney, B.M., 1978. Role of inorganic ions in controlling sedimentation rate of a marine centric diatom Dirylurn brighlwelli. J . Phycol., 14: 204-214. Arntz, W.E., 1986. The two faces of El Niiio 1982-83. Meeresforschung, 31: 1-46. Arntz, W.E., Landa, A. and Tarazona, J. (eds.), 1985. "El Niiio", Su impact0 en la Fauna Marina. Bol. Inst. Mar Peru, volumen extraordinario, Callao, Peru, 222 pp. Avaria, S., 1984. Cambios en la composicion y biomasa del fitoplancton marine del norte de Chile durante el Fenomeno de El Niiio 1982-1983. Revisita de la Comision Permanente del Pacifico Sur, 15: 303-309. Avaria, S., 1985. Variaciones en la composicion y biomasa del fitoplancton marine del norte d e Chile entre Diciembre 1980 y Junio 1984. Inves. Pesq. (Chile), 32: 191-193. Barber, R.T., 1988. The ocean basin ecosystem. In: Concepts of Ecosystem Ecology. J.J. Alberts and L.R. Pomeroy (eds.). Springer Verlag, pp. 166-188. Barber, R.T., Chavez, F.P. and Kogelschatz, J.E., 1985. Biological effects of El Niiio. In: Seminario Regional Ciencia Tecnologia y Agression Ambiental: El Fenomeno "El Niiio." M. Vegas (ed.). CONCYTEC Press, Lima, Peru, pp. 399-438. Barber, R.T., Kogelschatz, J.E. and Chavez, F.P., 1986. Productividad primaria en las aguas costeras entre 10s 5"- 15"s del Pacifico suroriental. In: Bases Biologicas y Marco conceptual para el Manejo de 10s Recursos Pelagicos en el Pacifico Suroriental. Oldepesca, pp. 33-37. Barber, R.T. and Chavez, F.P., 1983. Biological consequences of El Niiio. Science, 222: 12031210. Barber, R.T. and Chavez, F.P., 1986. Ocean variability in relation to living resources during the 1982/83 El Niiio. Nature, 319: 279-285.
51 Barber, R.T. and Ryther, J.H., 1969. Organic chelators: Factors affecting primary production in the Cromwell Current upwelling. J. Exp. Mar. Bio. Eco., 3: 191-199. Barber, R.T. and Smith, R.L., 1981. Coastal upwelling ecosystems. In: Analysis of Marine Ecosystems. A.R. Longhurst (ed.). Academic Press, New York, pp. 31-68. Bass, A.E. and Packard, T.T., 1977. Physical, chemical and biological observations from Joint I1 R/V Alpha Helix, Leg 0, 5-20 March 1976. CUEA Data Report, 41: 124 pp. Blasco, D., 1979. Changes in the surface distribution of a dinoflagellate bloom off the Peru coast related to time of day. In: Toxic Dinoflagellate Blooms. D.L. Taylor and H.H. Seliger (eds.). Elsevier, North Holland, pp. 209-214. Bowman, M.J. and Esaias, W.E., 1981. Fronts, stratification, and mixing in Long Island and Block Island Sounds. J. Geophys. Res., 86: 4260-4264. Buchanan, J., 1886. On similarities in the physical geography of the great oceans. Pro. R. Geog. SOC.,8: 753-770. Busalacchi. A.J.. Takeuchi. K. and O'Brien, J.. 1983. Interannual variability of the equatorial Pacific'- revisited. J. Geophys. Res., 88 Cane, M.A. and Zebiak, S.E., 1985. A theory for El Niiio and the Southern Oscillation. Science, 228: 1085-1086. Chavez, F.P. and Barber, R.T., 1987. An estimate of new production in the equatorial Pacific. Deep-sea Res., 34: 1229-1243. Chavez, F.P., 1987. Ocean variability and phytoplankton community structure. Ph.D. Thesis, Duke University, Durham, North Carolina, 300 pp. Chavez, F.P., Barber, R.T. and Sanderson, M.P., 1988. The potential primary production of the Peruvian upwelling ecosystem: 1953- 1984. In: The Peruvian upwelling ecosystem: dynamics and interactions. D. Pauly, P. Muck, J. Mendo and I. Tsukayama (eds.). Instituto del Mar del Peru (IMARPE), Callao, Peru; Deutsche Gesellschaft fur Technische Zusammenarbeit (GTZ) GmbH, Eschborn, Federal Republic of Germany; and International Center for Living Aquatic Resources Management (ICLARM), Manila, Philippine. In press. Chavez, F.P., Barber, R.T. and Soldi, H.S.,1984. Propagated temperature changes during onset and recovery of the 1982/83 El Niiio. Nature, 309: 47-49. Chavez, F.P. and Barber, R.T., 1985. Plankton production during El Niiio. In: International Conference on the TOGA Scientific Programme. World Climate Research Publication, Geneva, pp. VI 23-32. Coker, R.E., 1908. The fisheries and guano industry of Peru. Bull. Bur. Fish., 28: 333-365. Coker, R.E., 1918. Ocean temperatures off the coast of Peru. Geogr. Rev., 5: 127-135. Cowles, T.J., Barber, R.T. and Guillen, O., 1977. Biological consequences of the 1975 El Nitlo. Science, 195: 285-287. Dugdale, R.C., 1967. Nutrient limitation in the sea: dynamics, identification, and significance. Limnol. Oceanogr., 12: 685-695. Dugdale, R.C., 1979. Primary nutrient and red tides in upwelling regions. In: Toxic Dinoflagellate Blooms. D.L. Taylor and H.H. Seliger (eds.). Elsevier, North Holland, pp. 257-278. Dugdale, R.C., 1985. The effects of varying nutrient concentrations on biological production in upwelling regions. CALCOFI Reports, 26: 93-96. Dugdale, R.C., Jones, B.H., MacIsaac, J.J. and Goering, J.J., 1981. Adaptation of nutrient assimilation. Can. Bull. Fish. Aquat. Sci., 211: 234-250. Dugdale, R.C., Wilkerson, F.P., Barber, R.T., Blasco, D. and Packard, T.T., 1987. Changes in nutrients, pH, light penetration and heat budget by migrating photosynthetic organisms. Oceanol. Acta, pp. 103-107. Dugdale, R.C., Goering, J.J., Barber, R.T., Smith, R.L. and Packard, T.T., 1977. Denitrification and hydrogen sulfide in the Peru upwelling region during 1976. Deep-sea Res., 24: 601-608. Dugdale, R.C. and Goering, J.J., 1967. Uptake of new and regenerated forms of nitrogen in primary productivity. Limnol. Oceanogr., 12: 196-206. Enfield, D.B., 1981. Thermally driven wind variability in the planetary boundary layer above Lima. J. Geophys. Res., 86: 2005. Enfield, D.B. and Allen, J.S., 1980. On the structure and dynamics of monthly mean sea level anomalies along the Pacific coast of North and South America. J. Phys. Oceanogr., 10: 557578. Eppley R.W. and Peterson, B.J., 1979. Particulate organic matter flux and planktonic new production in the deep ocean. Nature, 282: 677-680.
52 Eppley R.W., 1981. Relations between nutrient assimilation and growth in phytoplankton with a brief review of estimates of growth rate in the ocean. Can. Bull. Fish. Aquat. Sci., 210: 251263. Feldman, G., Clark, D. and Halpern, D., 1984. Satellite color observations of the phytoplankton distribution in the eastern equatorial Pacific during the 1982-1983 El Niiio. Science, 226: 1069- 1071. Feldman, G.C., 1985. Satellite observations of phytoplankton variability in the eastern equatorial Pacific. Ph.D. Thesis, State University of New York at Stony Brook, New York, New York, 217 pp. Fiedler, P.C., Methos, R.D. and Hewitt, R.P., 1986. Effects of California El Nifio 1982-1984 on the northern anchovy. J. Mar. Res., 44: 317-338. Fleming, R.H., 1957. General Features of the Oceans. In: Treatise on Marine Ecology and Paleoecology. J.W. Hedgpeth (ed.). Geological Society of America, pp. 87- 108. Glynn, P.W., 1988. El Niiio-southern oscillation 1982- 1983: nearshore population, community and ecosystem responses. Ann. Rev. Ecol. Syst., 19: 309-345. Gunther, E.R., 1936. A report on oceanographical investigations in the Peru Coastal Current. Discovery Rep., 13: 107-276. Halpern, D., 1987. Observations of annual and El Nifio thermal and flow variations at o", 110"W and o", 95"W during 1980-1985. J. Geophys. Res., 92: 8197-8212. Harrison, D.E. and Schopf, P.S., 1984. Kelvin-wave-induced anomalous advection and the onset of surface warming in El Nifio events. J. Phys. Oceanogr., 14: 923-933. Harrison, G. and Platt, T., 1981. Primary production and nutrient fluxes off the northern coast of Peru: a summary. Bol. Inst. Mar Peru, volumen extraordinario. pp. 15-21. Hayes, S.P., Mangum, L.J., Barber, R.T., Huyer, A. and Smith, R.L., 1987. Hydrographic variability west of the Galapagos Islands during the 1982/83 El Nifio. Prog. Oceanogr., 17: 137- 162. Huntsman, S.A., Brink, K.H., Barber, R.T. and Blasco, D., 1981. The role of circulation and stability in controlling the relative abundance of dinoflagellates and diatoms over the Peru shelf. In: Coastal Upwelling. F.A. Richards (ed.). Am. Geophys. Union, pp. 357-365. Huyer, A., Smith, D.L. and Paluszkiewicz, T., 1987. Coastal upwelling off Peru during normal and El Nifio times, 1981-1984. J. Geophys. Res., 9 2 14297-14308. Jones, B.H., Brink, K.H.,Dugdale, R.C., Stuart, D.W., Van Leer, J.C., Blasco, D. and Kelley, J.C., 1983. Observations of a persistent upwelling center off Point Conception, California. In: Coastal upwelling: its sediment record. E. Suess and J. Thiede (eds.). Plenum Press, New York, pp. 37-60. Kirby-Smith, W.W. and Barber, R.T., 1974. Suspension-feeding aquaculture systems: effects of phytoplankton concentration and temperature on the growth of the bay scallop. Aquaculture 3: 135- 145. Koblentz-Mishke, O.I., Volkovinsky, V.V. and Kabanova, J.G., 1970. Plankton primary production of the world ocean. In: Scientific exploration of the South Pacific. W. Wooster (ed.). Nat. Acad. Sci., pp. 183-193. Kogelschatz, J., Solorzano, L., Barber, R. and Mendoza, P., 1985. Oceanographic conditions in the Galapagos Islands during the 1982/83 El Niiio. In: El Niiio in the Galapagos Islands: The 1982/1983 Event. G. Robinson and E.M. del Pino (eds.). Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 91-123. Levitus, S., 1982. Climatological Atlas of the World Ocean. Professional Paper No. 13. National Oceanic and Atmospheric Administration. U.S. Government Printing Office, Washington, D.C., 173 pp. MacIsaac, J.J., Kogelschatz, J.E.,Jones, B.H., Paul, J.C., Breitner, N.F. and Garfield, N., 1979. Joint I1 R/V Alpha Helix productivity and hydrographic data, March-May 1976. CUEA Data Report 48: 324 pp. MacIsaac, J.J., Dugdale, R.C., Barber, R.T., Blasco, D. and Packard, T.T., 1985. Primary production cycle in an upwelling center. Deep-sea Res., 32: 503-529. Margalef, R., 1978. Life-forms of phytoplankton as survival alternatives in an unstable environment. Oceanol. Acta 1: 493-509. Margalef, R., Estrada, M. and Blasco, D., 1979. Functional morphology of organisms involved in red tides, as adapted to decaying turbulence. In: Toxic Dinoflagellate Blooms. L. Taylor and H.H. Seliger (eds.). Elsevier, Amsterdam, pp. 89-94. Munoz, S. P., 1985. Estructura y comportamiento de las comunidades fitoplanktonicas en el norte de Chile durante el fenomeno de El Nifio 1982-83, Inves. Pesq. (Chile), 32: 195-197.
53 O'Brien, J.J., Busalacchi, A. and Kindle, J., 1981. Ocean models of El Niiio. In: Resource Management and Environmental Uncertainty Lessons from Coastal Upwelling Fisheries. M.H. Glantz and J.D. Thompson (eds.). Wiley-Interscience, London, pp. 159-212. Ochoa, N., Rojas de Mendiola, B. and Gomez, O., 1985. Identification of the "El NiAo" phenomenon through phytoplankton organisms. Bol. Inst. Mar Peru, volumen extraordinario, pp. 23-32. Pauly, D. and Tsukayama, I. (eds.), 1987. The Anchoveta and its Upwelling Ecosystem: Three Decades of Changes. IMARPE, Callao, Peru; GTZ, Eschborn, Federal Republic of Germany; and ICLARM, Manila, Philippines, 351 pp. Phillips, O.M., 1977. The dynamics of the upper ocean, 2nd ed. Cambridge University Press, London and New York, 291 pp. Quinn, W.H., Neal, V.T and Antunez de Mayalo, S.E., 1987. El NiAo occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14449-14461. Rojas de Mendiola, B., 1979. Red tide along the Peruvian coast. In: Toxic Dinoflagellate Blooms. D.L. Taylor and H H. Seliger (eds.). Elsevier, Amsterdam, Holland, pp. 183-190. Rojas de Mendiola, B., Gomez, 0. and Ochoa, N., 1985. Efectos del fenomeno "El Niiio" sobre el fitoplancton. Bol. Inst. Mar Peru, volumen extraordinario, pp. 33-40. Sanchez de Benites, G., Alamo, A. and Fuentes, H., 1985. Alteraciones en la dieta alimentaria de algunos peces comerciales por efecto del fenomeno El NiAo. Bol. Inst. Mar Peru, volumen extraordinario, pp. 135- 142. Santander, H. and Tsukayama, I., 1983. The anchoveta and sardine and some events associated to the recruitment. Intergovernmental Oceanographic Commission Workshop Report 33: 1 I - 12. Santander, H. and Zuzunaga, J., 1984. Cambios en algunos componentes del ecosistema marino frente a1 Peru durante el Fenomeno El Niiio 1982-1983. Rev. Com. Perm. Pacific0 Sur, 15: 311-331. Santander, H. and Sandoval, O., 1985. Efectos del fenomeno El Niiio en el composicion, distribucion y abundancia del ictioplancton. In: Ciencia, Technologia y Agresion Ambiental: El Fenomeno El Niiio, Consejo Nacional de Ciencia y Tecnologia. Lima, Peru, pp. 355-374. Smayda, T.J., 1970. The suspension and sinking of phytoplankton in the sea. Oceanogr. Mar. Biol. Annu. Rev. 8: 353-414. Smayda, T.J., 1980. Phytoplankton species succession. In: The Physiological Ecology of Phytoplankton. I. Morris (ed.). Blackwell Scientific Publications, Boston, pp. 493-570. Smith, R.L., 1978. Poleward propagating perturbations in currents and sea level along the Peru coast. J. Geophys. Res., 83: 6083-6092. Smith, R.L., 1983. Peru coastal currents during El Niiio: 1976 and 1982. Science, 221: 1397-1399. Sorokin, Y.,1978. Description of primary production and of the heterotrophic microplankton in the Peruvian upwelling region. Oceanology 18: 62-71. Sverdrup, H.U., 1953. On conditions for the vernal blooming of phytoplankton. J. Cons. Int. Explor. Mer, 18: 287-295. Sverdrup, H.U., 1955. The place of physical oceanography in oceanographic research. J. Mar. Res., 14: 287-294. Valdivia, G.J.E., 1978. The anchovia and El Niiio. In: Marine Ecosystems and Fisheries Oceanography. T.R. Parsons, B-0. Jansson, A.R. Longhurst and G. Saetersdal (eds.). Rapp. P. -V. Reun. Cons. Int. Explor. Mer, 173 pp. Vinogradov, M.E., 1981. Ecosystems of equatorial upwelling. In: Analysis of Marine Ecosystems. A.R. Longhurst (ed.). Academic Press, New York, pp. 69-93. Walsh, J.J., 1981. A carbon budget for overfishing off Peru. Nature, 290: 300-304. Wilkerson, F.P., Dugdale, R.C. and Barber, R.T., 1987. Effects of El Niiio on new, regenerated and total production in eastern boundary upwelling systems. J. Geophys. Res., 92: 1434714353.
This Page Intentionally Left Blank
55
CORAL MORTALITY AND DISTURBANCES TO CORAL REEFS IN THE TROPICAL EASTERN PACIFIC PETER W. GLYNN Division of Marine Biology and Fisheries, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33 149-1098 ABSTRACT Glynn, P.W., 1989. Coral mortality and disturbances to coral reefs in the tropical eastern Pacific. Widespread bleaching and mortality of reef-building corals occurred in the tropical eastern Pacific region during the severe and prolonged El Nifio-Southern Oscillation (ENSO) event of 1982-83. At the height of the 10 month sea warming period, Panamanian reefs experienced 2-3 bouts of coral bleaching (loss of symbiotic zooxanthellae), which resulted in coral death 2-4 weeks later. Coral reefs in Costa Rica, Panama and Colombia suffered up to 70-90% coral mortality; in the Galapagos Islands (Ecuador), most coral reefs experienced >95% mortality. The hypothesis that sea warming caused this disturbance is supported by (a) the coincidence between the warming event and stress responses of reef organisms, (b) the correlation between the magnitude of local temperature deviations and the extent of mortality, and (c) the El Nifio simulation experiments that resulted in coral bleaching, mortality, and histopathological changes similar to those observed naturally. El Nirio warming usually does not extend north of Ecuador where eastern Pacific coral reefs are best developed. A near-decadal(l976-1988) warming trend over much of the tropical and subtropical Pacific Ocean could have affected the susceptibility of reef corals during the shortterm 1982-83 warming disturbance. Zooxanthellate corals were affected most severely in nearly all reef areas, however, other organisms -- such as benthic algae, non-zooxanthellate scleractinian corals, black corals, molluscs, barnacles, and crustacean symbiotes of corals -- often showed local negative responses associated with non-thermal, El Nibrelated conditions (e.g., nutrient depletion, low plankton abundance, high sea level, and wave assault). Secondary disturbances included (a) the elimination of coral baniers, allowing the corallivore Acanthaster access to formerly protected coral prey, (b) increased external bioerosion of reef surfaces killed in 1983 because of post-El Nirio increases in sea urchin densities, and (c) the establishment of damselfish temtories on corals that experienced partial mortality in 1983. Such disturbances are currently causing longer-term changes to their respective local communities. Estimates of the ages of massive corals that were killed or irreparably damaged, and the interruption of reef framework accumulation suggest that a disturbance comparable to that of 1982-83 probably has not occurred in the Galapagos Islands or Panama during the past 200 years. The initial damaging effects of the 1982-83 disturbance to reef coral populations, combined with persistent secondary disturbances and low coral recruitment, could prolong reef recovery for decades or possibly centuries. Periods of intense upwelling cause localized and moderate levels of cord mortality, but persistent ENSOrelated sea warming causes widespread and catastrophic coral mortality. El Niiio events of extreme severity may limit eastern Pacific reef growth and diversity as much as do distance and isolation of these reefs from the centers of reef development in the western Pacific. 1 INTRODUCTION El Niiio events signal the appearance of anomalously warm waters off northwestern South America, a region of normally cool, upwelling conditions conducive to high productivity (Wyrtki, 1975; Houvenaghel, 1978, 1984; Enfield, 1981; Philander, 1983; Rasmusson, 1984; Barber and Chavez, 1986; Cane, 1986; Longhurst and Pauly, 1987; Barber and Kogelschatz, this volume; Hansen, this volume). This sporadic movement of a tropical oceanic front to the south is accompanied by reduced nutrient replenishment to sun-lit surface waters, which depresses
56
plankton production and disrupts trophic links to a variety of consumer groups (Barber and Chavez, 1983, 1986; Glynn, 1988a). Other marine and atmospheric disturbances are typically associated with El Niiio events, e.g. increased rainfall and flooding, increased sediment loads and turbidity in coastal waters, and abnormally high sea level stands, rougher seas and coastal erosion (Caviedes, 1984; Hansen, this volume). More distant disturbances can occur during severe events, e.g. increased storm activity in the South Pacific and off western North America, droughts in southern Africa, Sri Lanka, southern India, the Philippines, Indonesia and Australia, but traditionally the focus of interest has been on disturbances to fisheries in upwelling areas off southern Ecuador and Peru during the anomalous warming periods. Disturbances that accompanied the unusually strong 1982-83 El Niiio warming event in the equatorial eastern Pacific precipitated devastating effects in some tropical marine ecosystems. The simultaneous occurrence of marked and prolonged sea warming and widespread disturbances to coral reefs in Costa Rica, Panama, Colombia and Ecuador (Galapagos Islands) suggested a causal connection linking these events. In this paper I will examine the relationships between the 1982-83 El Niiio event and disturbances to coral reefs in the tropical eastern Pacific region (Figs. 1-3). Disturbances to coral reefs and sea warming also were reported in 1983 in the Caribbean Sea, Bahama Islands and Florida Keys, and in various areas of the Indo-Pacific region (Lessios et al., 1983; Glynn, 1984a; Brown, 1987). A possible causal relationship between events outside of the eastern Pacific and the 1982-83 El Niiio-Southern Oscillation (ENSO) is examined by Coffroth et al. (this volume). Although the duration of the 1982-83 El Niiio warming lasted between 10-20 months (depending on location), persistent ecological changes are still affecting corals up to 6 years after the initial disturbance (to mid 1989). For example, (a) a sea star corallivore (Acanthaster feeds on corals that were previously unavailable to it, (b) the persistent grazing of sea urchins (Diadems m i c a n u r n and &&&vis thouarsii) on dead coral surfaces seriously erodes reef framework structures affected in 1983, and (c) territorial damselfishes (Stegastes spp.) have established and are expanding algal lawns on corals that were partially damaged by the warming event (or later by Acanthaster). The results of large-scale disturbances to coral reefs in 1982-83 and following years suggest that unusually intense El Niiio warming events may exert important controls on the development of coral reefs throughout the region. The impoverished coral fauna and meager development of reefs in the eastern Pacific could have resulted largely from infrequent but extreme warming events in concert with frequent upwelling and water bloom disturbances (Dana, 1975; Glynn, 1977; Colgan, this volume; Guzman et al., in press; Richmond and Glynn, in prep.). This focus on catastrophic warming disturbances adds another physical constraint that could interact importantly with the isolation of American coral faunas from centers of diversity (Stehli and Wells, 1971; Veron, 1985). In this paper I will examine the impact of the 1982-83 El Niiio event on eastern Pacific coral reefs, and assess the effects of conditions opposite to those prevalent during El Niiio events,
m)
57
Fig. 1. Location of eastern Pacific coral communities and coral reefs that were severely (occluded circles) or moderately (clear circles) affected during the 1982-83 El Niiio warming event. Localities and sources: COSTA RICA - 1 Playa Hermosa, 2 Conchal, 3 Samara, 4 Malpais, 5 Punta Leona, 6 Bahia Herradura and Punta Judas, 7 Punta Catedral, 8 Caiio Island, 9 San Josecito, 10 Sandalo, Golfo Dulce (1-7 after Glynn, 1984a; 8-10, J. Cortes, P.W. Glynn, H.M. Guzman, R. Richmond, pers. obs.); PANAMA - 11 Parida Island and Ladrones Islands, 12 Secas Islands, 13 Contreras Islands, 14 Coiba Island and surroundings, 15 Bona Island and Taboga Islands, 16 Pearl Islands, 17 Guayabo Grande or Ensenada Guayabo (11-16 after Glynn, 1984a; 17, J. Cubit, pers. obs.); COLOMBIA - 18 Utria Bay, 19 Gorgona Island (18 after Prahl, 1985 and H. von Prahl, pers. comm.; 19 after Glynn, 1984a); ECUADOR - 20 Ayangue and Pelado Island, 21 Santa Cruz Island, 22 Devil's Crown (Onslow Island) and Champion Island, 23 Baltra Channel, 24 Punta Espinosa, Fernandina Island, 25 Bartolome Island, 26 Espaiiola Island, 27 Marchena Island, 28 Wenman Island, 29 Culpepper Island, 30 Cocos Island, Costa Rica (20, P.W. Glynn, R. Richmond, F. Rivera, P.J.B. Scott, pers. obs.; 21 after Glynn, 1984a; 22-24, R. Espinosa, P.W. Glynn, R. Richmond, F. Rivera, G. Robinson, pers. obs.; 25, G.M. Wellington, pers. obs.; 26, David Day, pers. comm.; 27-29 after Cohen, 1985; Robinson, 1985; 30, H.M. Guzman, pers. comm.).
P A N A M A PEARL ISLAN
,
GULF OF CHlRlQUl
G U L F OF P A N A M A
c
P A C I F I C 84'
83"
I
I
82" I
8 I" I
O C E A N 80" I
79" I
78"W I
Fig. 2. Location of study sites in Costa Rica and Panama. Encircled numbers identify the following sites: 1, Cavada Island; 2, Ft. Amador Causeway; 3, Pacific entrance to the Panama Canal; 4, Taboga Island; 5, Uraba Island; 6, Bona Island; 7, Saboga Island; 8, Contadora Island.
L
59
so-called anti-El Niiio type conditions (Quinn and Neal, 1983)*. Oceanographic conditions preceding or following El Niiio events can also stress and kill corals. This information, coupled with results from a recent experiment simulating the effect of El Niiio surface warming on corals, will allow an evaluation of cause and effect relationships. Post-El Niiio disturbances on coral reefs will be examined and some speculations will be offered regarding coral recovery at selected sites. The tropical eastern Pacific is an area of marginal reef development and, perhaps for this reason, the prospects for rapid recovery are not encouraging. 2 CORAL BLEACHING, MORTALITY AND ENVIRONMENTAL CORRELATES 2.1 Onset of the 1982-83 disturbance The first sign of stress on eastern Pacific coral reefs was the bleaching or partial bleaching of corals resulting from the loss of symbiotic zooxanthellae (dinoflagellates) that normally reside within the gastrodermal cells of the coral host (Figs. 4-7). Reef-building corals exhibit a variety of colors, due largely to the photosynthetic and accessory pigments present in the zooxanthellae (Jeffrey and Haxo, 1968; Wicksten, 1989). With the expulsion of zooxanthellae during periods of stress (Steen and Muscatine, 1987), the corals bleach, i.e. display faint shades of brown, green, pink or some other color or appear stark white. The bleached appearance is due to the underlying white calcareous skeleton, which becomes visible through the translucent tissues. Corals rarely expel all of their zooxanthellae; when moderate numbers remain corals display weak brown colors, blotches or even brown streaking. Widespread coral bleaching was first observed in Panama and the Galapagos Islands in early 1983 (January-March), where it continued intermittently until July 1983 (Glynn, 1983, 1984a; Robinson, 1985). Corals showed signs of bleaching as early as January 1983 in the non-upwelling Gulf of Chiriqui, Panama (Fig. 2), according to the observations of residents of a sportfishing camp at Coiba Island (R. Griffin, pers. comm., in Glynn, 1983). By mid-February (12-13 February), extensive bleaching of most reef corals (ca. 10 species) was observed and photographed on coral reefs in the Gulf of Chiriqui (A. Averza, pers. comm., in Glynn, 1983). A similar bleaching event occurred in the Gulf of Panama (Fig. 2), an area of seasonal (January-April) upwelling, but this began in early June 1983 after a weak upwelling period. In the Galapagos Islands, reef-building corals had normal pigmentation in January 1983, but by the second half of February all the major species (ca. 5 species) were bleached at three sites (Robinson, 1985). The onset of bleaching events in Costa Rica and Colombia is unknown because the study sites there were not visited until well after the disturbance began (Prahl, 1983; Cortes et al., 1984; Guzman et al., 1987). The
* La Niiia (Spanish: the girl) was proposed recently (Philander, 1985; Ken, 1988) in place of the term anti-El Niiio to emphasize the complementary conditions prevailing between El Niiio events, i.e. strong SE trade and equatorial easterly system, cool sea water conditions, vigorous upwelling of nutrient rich water, high productivity, etc. To the Spanish speaker, La Niiia has no logical connection to El Nifio (the Christchild), which refers to an oceanographic/meteorologicevent that usually appears soon after the Christmas season. As the term anti-El Niiio has both precedence (Quinn, 1976, 1977) and a more logical basis, it will be used herein.
60
I
I
e'C"LPEPPER IS.
MARCHENA IS. GENOVESA IS. - oo P t a Espi
Bartolome Is a l t r a Channel
FER NANDINA IS.
to. Pi
- 1"s
SAN CRISTOBAL I S
50
100 K M
,Onslow Is. ( D e v i l ' s Crown) -Cormorant Bay Champion Is.
b
€i
FLoREANA Is.
1
92" W I
91"
90"
1
I
ESPANOLA IS.
89 1
Fig. 3. Location of study sites in the Galapagos Islands, Ecuador. CDRS identifies the location of the Charles Darwin Research Station, Academy Bay, Santa Cruz Island. timing of bleaching in relation to local sea temperatures during 1983 is considered at five sites in Costa Rica, Panama, Colombia and the Galapagos Islands below. 2.2 Extent of the 1982-83-d The majority of coral reef areas depicted in Figs. 1-3 were severely disturbed (occluded circles, Fig. 1 only) during the 1982-83 El Niiio event. Coral mortality ranged from about 50% to 95% of the total live cover. The Galapagos Islands were most severely affected, experiencing 97-100% mortality. The scope of damage at several sites will be examined in detail in various sections of this paper. However, some eastern Pacific areas experienced only minimal or no disturbance
61
Fig. 4. Pre-1983 appearance of the Uva Island pocilloporid reef (Uva Island, Gulf of Chiriqui, Panama, 6 m depth, 23 June 1978). Nearly all of the Pocillopora damicomis (high mound) and Pocillopora elegans (foreground) corals are alive. Reef framework mound is approximately 1 m higher than surrounding corals.
effects. The Golfo Dulce area in Costa Rica (location 10, Fig. 1, and Fig. 2), surveyed 2 years after the El Niiio disturbance (9-12 February 1985), revealed only slight or no signs of El Niiiorelated coral mortality. Coral reefs located in the inner-most reaches of Golfo Dulce (the Los Mogos and nearby Punta Islotes areas) contained mostly dead corals that appeared to have died before 1982-83. Large areas of pocilloporid corals found in growth position on the reef flat were already covered by sediments in 1978 and were likely killed in the early to mid 1970s and then buried subsequently (Glynn et al., 1983). The Pontes lobap frameworks that were alive in 1978 were largely dead in 1985, but in a highly eroded state. The extent of erosion and the relatively large regrowth increments of surviving patches indicate that most of these corals were affected before 1983 (possibly by increased sedimentation resulting from deforestation and unsound agricultural practices, J. Cortes, pers. comm.). Coral reefs located on the southwest shore of Golfo Dulce (the Sandal0 m a ) contained abundant populations of Pocillopora spp., , -P Pavona varians, and Psammocora stellata that appeared to be in equally healthy condition in 1978 and 1985. For some reason, the corals in this m a have largely escaped the El Niiio disturbance and other stressors (increased sedimentation, Corks, in prep.) that have affected corals in the inner-most gulf. During a survey 3 years after the disturbance (19 August 1986), numerous live pocilloporid colonies, some 30-40 cm in diameter, were observed further south on the Ecuadorian coast near Ayangue and at Pelado Island, a small offshore island (location 20, Fig. 1). Very few dead corals, either in growth position or otherwise, were present, whereas numerous dead corals were still present as of 1989 in severely disturbed areas of the Galapagos Islands. I could find no information on the responses of corals at higher latitudes, i.e. from Nicaragua north to Baja
62
Fig. 5. Patchy bleaching of pocilloporid corals on the upper (a) and lower (b) seaward slopes of the Uva Island reef (Uva Island, Contreras Islands, Gulf of Chiriqui, Panama, a = 4 m and b = 7 m depths, 28 April 1983). Dark patches are dead corals that were bleached in March and are now overgrown with algae. a - distance across foreground about 2 meters; b - angelfish total lengths are 20-25 cm. California, Mexico. However, in 1987, Pocillopora experienced bleaching during a period of unusually high sea temperatures in the Gulf of California (Reyes Bonilla, 1988). While the extent of refuge coral populations that survived the 1982-83 event is largely unknown, their presence as sources of reproductive propagules for recovery in distant communities could be critical.
63 2.3 Condition of bleached corals The condition of corals during the bleaching event (i.e., polyp expansion, mucous release, skeletal growth, the parts of colonies affected, and the water depths where bleaching occurred), observed in Panama, Colombia, and the Galapagos Islands, is considered here. Also, the known
a
b
Fig. 6. Bleached colonies of Pocillopora plegans showing uniform bleaching (a) of branch-tips and partial bleaching (b) with some centrally located branches retaining near-normal pigmentation (Cavada Island, Secas Islands, Gulf of Chiriqui, Panama, 8 m depth, 22 March 1983). Colonies are approximately 30 cm high. spatial and temporal patterns of coral bleaching are noted, as well as the examination of coral tissue microstructure in light of the possible role of an infectious agent in coral bleaching and mortality. Information on species susceptibilities to bleaching and mortality is presented below in section 3.1. Night observations of bleached field colonies of Pocillopora in the Galapagos Islands revealed that the polyps were uncharacteristically retracted (Robinson, 1985). Although no systematic observations were performed, I observed polyp expansion during the day and night in all recently bleached corals in Panama. Mucus release was greatly diminished in bleached pocilloporid corals in Colombia (Prahl, 1983), Panama (Glynn, 1985b) and the Galapagos Islands (Robinson, 1985). Bleached colonies of Pocillopora damicomis in Panama released less than half the volume of mucus that was released by pigmented colonies (Fig. 8a). No observations were reported elsewhere on the mucus production by other corals. In Panama, some 2-4 weeks elapsed between an initial bleaching event and coral death. The sloughing of dead tissues was observed in only a few pocilloporid colonies, indicating that tissue loss may have occurred rapidly in most colonies. To determine if bleached corals were capable of skeletal growth, branching (Pocillopora) and massive (Pavona and PoriteS) species were stained with Alizarin Red-S bone stain and maintained under laboratory (Glynn, 1983; Robinson, 1985) and field (Prahl, 1986) conditions. Not unexpectedly, skeletal growth was nil, indicating an absence of active calcification in bleached corals. Fully and even partially pigmented coral colonies incorporated the stain and showed skeletal growth. The loss of zooxanthellae was not uniform across the colony. Coral tissues not receiving light directly (e.g., on the sides and under surfaces of colonies or in colony fissures), or the tissues of
64
Fig. 7. A bleached colony of the massive coral Gardineroseris planulata (Contadora Island, Pearl Islands, Gulf of Panama, Panama, 4 m depth, 13 September 1983). Note the dead, algal-covered summits and the live bleached sides of the colony. Colony height approximately 0.5 m (foreground). whole shaded colonies, often retained their pigmentation longer than exposed tissues (Prahl, 1983; Glynn, 1984; Robinson, 1985). It is possible that high temperature stress increased the sensitivity of corals to damage by UV-B or visible light (Coles and Jokiel, 1978; Roth et al., 1982). In branching species, bleaching often started on branch-tips and progressed medially to the basal and inner colony branches (Fig. 6a). Although most species on Panamanian reefs that were located in cryptic habitats, e.g. Pavona varians and Pontes panamensis, did not bleach in the early months, they eventually bleached toward the end of the warming event and incurred high mortalities. The severity of coral bleaching varied with depth at the different affected sites. In the Galapagos Islands, bleaching was most pronounced in shallow water (S 10 meters) during the early phase of the El Nifio event (Robinson, 1985). However, by the end of the warming event all corals were affected to depths of about 30 meters, surpassing the maximum depth limit of most species. Still, bleaching was more intense and mortality higher for most species at shallow depths (Glynn, 1983, 1984a; Prahl, 1983; Robinson, 1985). As in the Galapagos Islands, bleaching in the Gulf of Chiriqui, Panama, was more pronounced among reef-building corals at 1-10 m than 11-20 m depths during the early El Niiio period, but as the disturbance continued more corals were bleached and killed at greater depths. And, by the end of the disturbance (October 1983) formerly large continuous tracts of live Pocillouora in the Gulf of Chiriqui at 8-10 meters were dead, but
65
P 0.02>p>0.01
I
N
PB
FB
D
Ib
pco.001
N
I
PB
FB
D
FB
D
n
N
PB
Fig. 8. Decline in mucus release (a), crustacean symbiote* density (b), and aggressive responses of crustacean symbiotes (c) in field colonies of Pocillouora damicomis during the 1983 coral bleaching event in Panama. Coral condition: N, normal-appearing with full brown pigmentation; PB, partially bleached; FB, fully bleached; D, dead. Median values, 0.95 confidence limits of medians, and number of colonies sampled (in parentheses) are shown for each condition. Kruskal-Wallis significance levels are indicated in each plot (upper right-hand comer); thick horizontal lines along abscissas join statistically equal median values (multiple comparison procedure, a = 0.15). (From Glynn, 1985b.) *Etymologically, the Greek derivation of this term requires the present suffix (Hertig et al., 1937).
66
those at 2-3 meters depth were still mostly alive (Glynn, 1984a). In contrast to other sites in the Gulf of Panama, survivorship of PocilloDora damicornis was independent of water depth over a range of 1-8 meters (Glynn, 1984a). Because of the length of the El Nifio disturbance, the irregular and sporadic responses of corals over the event interval, and delayed mortality, it was difficult to assess the effects of the warming phase on corals as a function of depth. The following quantitative surveys indicate that some coral species in Panama demonstrated a high rate of recovery at shallow depths. Approximately 14 months after the end of the 1983 coral bleaching event, before significant recruitment or secondary effects were observed, the condition of corals in reef zones dominated by pocilloporid frameworks (60-90% live coral cover before 1983) was assessed at two Panamanian reefs. Sampling was confined to reef areas whose coral populations were uniformly (100%) bleached in 1983. The results of this survey (Table 1) indicate that live Pocillopora spp. cover was highest in the upper reef slope (3-5 m depth) and lowest in the lower slope zone (7-10 m). Interestingly, within the severely disturbed m a s , two deep water coral communities that displayed conspicuous bleaching during the height of the warming event experienced only minimal or no mortality. In Panama, large tracts of Psammocora stellata at 10-18 m depth underwent partial bleaching (from brown to gray white) in April 1983, but by September-October 1983 all colonies had regained their normal pigmentation (pers. obs.). In the Galapagos Islands, a large tract of Cvcloseris mexicana at 15 m depth experienced severe bleaching (from greenish-brown to bone white), but had totally recovered shortly after the El Niiio warming period (Robinson, 1985). In Panama (Gulf of Chiriqui), bleaching was patchy on pocilloporid reefs, i.e. single colonies or continuous 1 m to 100 m patches of corals were bleached, imparting a variegated appearance to the reefscape (Fig. 5). The bleaching occurred in bouts with ca. 50430% of the original cover of pocilloporid corals affected by mid March, 80-95% affected by the end of April, and 8595% affected by late October when the disturbance ceased. Some massive coral species in Panama also bleached at different times at the same sites. Likewise in the Gulf of Panama patchy and sequential bleaching are suspected although this was not documented. The occurrence of patchy and/or sequential bleaching was not mentioned for the other eastern Pacific reef sites examined (Prahl, 1983, 1985; Cortes et al., 1984; Robinson, 1985). The repeated bleaching and patchy distribution of affected corals suggested that a possible infectious agent caused the coral bleaching and mortality. To test this hypothesis, transplantation experiments, histopathological analysis, and electron microscopic examinations were conducted on normal and affected corals in Panama during the disturbance period (Glynn et al., 1985b). The rates of coral bleaching and mortality in Pocillouora damicornis, Pocillooorp clegans, and Gardineroseris planulm were unaffected by isograft (same colony), allograft (different colonies, same species), and xenograft (different species) procedures. Colonies with full pigmentation that received bleached grafts remained in a healthy state for the duration of the study (7 months), indicating that pathogens were not responsible for the mortality. If pathogens or microorganisms were responsible for the coral bleaching, it is likely that the placement of an affected graft in direct contact with normal tissues would have resulted in transmission, as seen with other coral diseases [e.g., 'shut-down-reaction' (Antonius, 1977) or 'black band disease' (Rutzler et al., 1983)l.
TABLE 1 Mean percent live coral cover of pocilloporid-dominated frameworks at three different depths 14 months following the 1983 coral bleaching event. Reef sitesa (Sampling date)
Reef zones Reef flat RF 0.5-1 m
Upper slope
us
3-5 m
Lower slope LS 7-10 m
Number of quadrats Per zonec
statistical testsd
K-W
MC
Secasstud reef (24 Jan. IJ885)
19.3(&13.8)b
45.3(&20.8)
4.2(&7.8)
12
O.Ol>p>0.001
us RF LS
Uva Island reef (28 Jan. 1985)
16.2(&3.4)
63.4(49.8)
3.3(&2.3)
40
p
USRFLS
aThe exact locations of the reefs in the Gulf of Chiriqui, Panama, are given in Glynn and Macintyre (1977). h e 95% confidence limits of means are given in parentheses. CSampling was performed with 0.5 x 0.5 m quadrats equipped with equally-spaced cross lines giving 16 points/l/4 m2. Quadrats were dropped haphazardly near the centers of zones, then 'walked' (flipped)toward the north and sampled at every other position. dProbability levels are noted for the results of Kruskal-Wallis (K-W) testing. In the multiple comparisons (MC) testing, lines join mean values that are not significantly (a= 0.15) different.
68
The histological study of field specimens revealed that normally pigmented corals generally appeared healthy with good cellular architecture and staining characteristics. In contrast, partially or fully bleached colonies had varying degrees of tissue atrophy and necrosis (Figs. 9 and 10). For example, (a) a decrease in the numbers of mucous secretory cells in the epidermis (in all species except Pavona clavus, which showed an increase), (b) a loss of zooxanthellae from the gastrcdermal cells, and increases in the numbers of degenerating zooxanthellae (with increased vacuolation) and their pigmented debris, and (c) a lack of gonadal development in bleached corals. The disappearance of gonads in the polyps of Pocillopora damicornis during an El Niiio simulation experiment (Fig. 11) was similar to that observed in field corals in 1983. Although no microorganisms were observed in coral tissues with the light microscope, examination by transmission electron microscopy revealed circular, membrane-bound vesicles suggestive of coccoid bacteria within the gastrodermis of bleached Pavona varians (Fig. 9). Scanning electron micrographs of cryofractured gastrodermal tissues of bleached P. varians and Pocillopora glegans disclosed the presence of spherical and rod-shaped, bacteria-like objects, sometimes present in vacuoles vacated by zooxanthellae (Fig. 10). Bacteria-like objects were not found in apparently normal coral tissues. Glynn et al. (1985b) concluded that the bleaching and atrophy of coral tissues in Panama was not caused primarily by microorganisms, but that the latter probably colonized necrotic tissues secondarily. The timing of the coral bleaching event and the period of sea warming were closely correlated at several localities, suggesting a causal relationship. I will now review briefly the effects of both high and low temperatures on reef-building corals and then consider in detail the evidence linking the El Niiio warming episode with the 1983 coral bleaching and mortality event. Since coral bleaching occurs under a variety of stressful conditions, I will later consider other non-thermal stressors that could also have been responsible for this widespread disturbance (section 2.9). 2.4 Coral bleaching and sea water temDeratm extremes Following Dana's (1843) observation of the generally cool marine conditions and absence of coral reef development in the tropical eastern Pacific, several workers have substantiated the negative effects of local upwelling on reef-building corals in this region (Crossland, 1927; Glynn, 1972,1977; Glynn and Stewart, 1973; Dana, 1975; Birkeland, 1977; Glynn and Wellington, 1983; Glynn et al., 1983). Later studies have, however, c o n f i i e d the presence of coral reefs from Ecuador to Baja California, Mexico, and have concluded that they are typically small, youthful, and species poor compared with coral reefs in the central and western sectors of the tropical Pacific Ocean (Durham and Barnard, 1952; Squires, 1959; Porter, 1974; Dana, 1975; Brusca and Thomson, 1977; Glynn and Wellington, 1983; Cortes and Murillo, 1985). Generally, more importance has been assigned to low rather than high temperatures in controlling coral diversity and reef growth on a global, latitudinal scale (e.g., Wells, 1957; Yonge, 1968; Stoddart, 1969a; Newell, 1971; Stehli and Wells, 1971; Rosen, 1981; Grigg, 1982). Those studies that have demonstrated negative effects of high temperature stress were concerned with shallow reef zones that are heated intensely during diurnal tidal exposures or sea level drops when water flow
69
Fig. 9. Transmission electron micrograph of Pavona varians in a bleached condition (Uva Island, Gulf of Chiriqui, Panama, 6 m depth, 23 June 1983). Foreign, circular, membrane-bound vesicles present in the gastrodemis (upper right) are denoted by unnumbered arrows. The three large ovoid bodies are zooxanthellae in their iespective vacuoles (9,460 x). Scale bar = 5 pm. Numbered arrows: 1, pyrenoid bodies; 2, nucleus; 3, chloroplast.
Fig. 10. Scanning electron micrographs of the fractured mesenterial surfaces of normal damicomis (a) and bleached Pocillopora elegans (b). a - note zooxanthellae in separate vacuoles in gastrodemal cells (mows)and flagellated coelenteric canal, scale bar = 20 pm (Uraba reef, Taboga Islands, 4 m depth, 14 June 1983). b - note absence of zooxanthellae from the gastrodemis and presence of clusters of spherules in cells (arrows), scale bar = 5 pm (Uva Island reef, Contreras Islands, 8 m depth, 23 June 1983. Prepared by the cryofracture technique (Glynn et al., 1985b).
70
Fig. 11. Photomicrographs of pocillopora damicomit illusaating condition of polyps after week 1 (27 August 1985) and week 6 (1 October 1985) of an El Niiio simulation experiment. a - cross section of polyp near bottom of stomodaeum showing spermaries and oocytes present on mesenteries. b - cross section through gut, gonads absent and mucosecretory cells of epidermis greatly reduced in number and size. Corals from Uraba Island, Gulf of Panama, temperature treatment = 30°C, tissues stained with Movat's Pentachrome, scale bars = 100 pm. Legend ep = epidermis; gs = gastrodermis; gvc = gastrovascular cavity; mf = mesenterial filament; oc = oocyte; sp = spermary (after Glynn and D'Croz, in press). often is restricted. Numerous workers have reported coral mortality events due to excessive water warming in isolated habitats with restricted water flow in most of the worlds major reef areas, e.g. the Florida Keys (Mayer, 1918a, b; Mayor, 1924; Shinn, 1966; Jaap, 1979), the eastern Pacific (Glynn, 1976), Hawaii (Edmondson, 1928), the Marshall Islands (Fankboner and Reid, 1981), eastern Australia and Torres Straights (Mayer, 1918a), the Mariana Islands (Yamaguchi, 1975), the Ryukyus Islands (Yamazato, 1981; Kamezaki and Ui, 1984), and the Red Sea (Loya, 1976). Several laboratory studies have demonstrated that slight but sustained increases in water temperature can cause morbidity and death of reef corals (Clausen, 1971; Clausen and Roth, 1975; Coles et al., 1976; Coles and Jokiel, 1977, 1978; Houck et al., 1977; Jokiel and Coles, 1977; Yang et al., 1980). Thus, current dogma maintains that low temperatures mainly affect the latitudinal distributions of reef-building corals and high temperatures (in combination with sea level and irradiance) their upper vertical distributions (e.g. Stoddart, 1969a; Endean, 1976). The widespread coral bleaching and mortality that accompanied the 1982-83 El Niiio warming event (Glynn, 1984a, 1988a; Brown, 1987; Coffroth et al., this volume) indicates that reef corals at virtually all depths are susceptible to warm water disturbances. This hypothesis, that high temperatures can limit the vertical and horizontal distributions of corals and the extent of coral reef building, and the evidence supporting it are examined below. 2.5 w i l i t y of sea
observw
Because of the great sensitivity of reef-building corals to water temperatures (both low and high, rate of change, duration, deviations from long-term norms, etc.), it is essential that prudence be exercised in the evaluation of presumed thermal effects. The nature and reliability of sea temperature data, especially those data obtained during the 1982-83 period, are examined here in
71
relation to the timing of coral bleaching. Daily sea surface temperatures (SST) recorded from shore stations at Naos Island, Gulf of Panama (Fig. 2), and in Academy Bay (Santa Cruz Island), Galapagos Islands (Fig.3), are compared with offshore ship and satellite SST observations from the same general areas (Figs. 12 and 13). Daily extreme temperatures (maxima and minima) are noted on the curves for the permanently-based shore stations. Unlike the synoptic ship and satellite observations, the shore station records indicate the magnitude of the local ranges in SST. In Panama, monthly temperature deviations were most pronounced during the 1982 upwelling season (January - April). The extreme low temperatures during the 1983 upwelling season were considerably higher and the highest daily extremes (to 32°C) in the Gulf of Panama occurred near the beginning (June) of the 1983 warming period. Monthly extreme temperatures were more pronounced in the Gulf of Panama, with ranges as high as 6-7OC, than in the Galapagos Islands with maximum monthly ranges of about 3°C. This difference is due to the strong, intermittent episodes of upwelling in the Gulf of Panama. In general, the mean SST values (shore, ship and satellite observations) were in close agreement during the 1982-83 El Nifio period except for the 1983 record in the Gulf of Panama. The satellite SST retrievals revealed a cool bias of 1-2OC from March through December. It is possible that this negative bias was due to the El Chichon aerosol cloud (Reynolds and Gemmill, 1984), which caused an increase in the absorption of infrared light, thus resulting in an apparent lower SST. The bias was stronger in the Northern compared to the Southern Hemisphere (Strong, 1983), and is not evident in the satellite-derived SST observations from the Galapagos Islands, which were obtained from slightly below the equator (ca. 1's). In summary, this brief analysis indicates that ship and satellite derived SST data are generally in good agreement with the mean measurements obtained at shore stations during 1982-1983. Temperature extremes, however, commonly ranged from 3-4OC above and below mean values in the Gulf of Panama and from 2-3OC in the Galapagos Islands. 2.6 Sea warminp and the timine of coral bleaching Five eastern Pacific localities that experienced coral bleaching revealed a generally good correspondence between the timing of the bleaching and the period of sea warming (Fig. 14). Unfortunately, most reef areas were not under constant observation, nor were sea temperatures obtained locally on a regular basis, so that it is difficult to relate precisely the timing of coral bleaching with the onset of the warming trend. The most reliable observations are those from the Galapagos Islands and the Gulf of Panama because these areas were under nearly constant surveillance and daily hydrographic measurements were obtained there. Reasonable confidence is placed on the timing of events in the Gulf of Chiriqui because the reefs there were observed only 23 weeks after the initiation of coral bleaching. The timing of the disturbance events in Costa Rica and Colombia could have a 1-2 month margin of error because they were not observed until well after the warming event began. Reef-building corals first began losing their zooxanthellae in Panama (Glynn, 1983, 1984a) and the Galapagos Islands (Robinson, 1985). In the Gulf of Chiriqui, Panama, and the south-
72
+ N a o s Is observations
max min
Ship observations S a t e l l i t e observations
29 3 0 : 28 27
" c 26 24 25
21
:h
I".,
- 0>
I-
I
J F M A M J J A S O N D J F M A M J J A S O N D
1982
1983
Fig. 12. Mean monthly sea surface temperature curves during 1982-1983 in the vicinity of Naos Island, Gulf of Panama, Panama, derived from three different data sources. Naos Island observations were made twice daily, at about 0900 and 1500, and also show monthly maximum and minimum temperatures. Ship observations are from synoptic monthly charts prepared by F. Miller, InterAmerican Tropical Tuna Commission. Satellite observations represent 4-5 readings per month from NESS, OPC 50 km charts (read from a point 50 km offshore in the center of the Gulf of Panama), National Oceanographic and Atmospheric Administration. central sector of the Galapagos Islands, extensive coral bleaching was observed in mid February. As a rule, the coral tissues that whitened remained alive for several weeks with polyps in various stages of expansion. Thereafter, many of the affected corals died in whole or in part (partial colony mortality) depending on colony location, size, species, etc. The Galapagos area experienced 9 months of 1-2°C above-normal temperatures before the coral bleaching event whereas the Chiriqui event was preceded by 6 months of 0S"C above normal temperatures. Coral bleaching was observed somewhat later in Colombia (by April at Gorgona Island and Uma Bay, Prahl, 1985), but it is possible that it began as early as February. The pre-event warming in Colombia was considerable with temperatures consistently 1-2°C above the long-term norm over an 11 month period. In addition, one temperature reading of 3 1.5"C obtained from Gorgona Island in February 1983 (Prahl, pers. comm.), 3.2"C above the mean, suggests that Colombian coral reefs may have experienced greater warming than indicated by the synoptic temperature curve. In
73
rnax.
+
A c a d e m y Boy observations
.lll.ll..l
Ship o bse r vo t i o ns
min. 30 29 28 27
26
"c
25 24
23 22 21
20
J F M A M J J A S O N D J F M A M J J A S O N D
1982
1983
Fig. 13. Mean monthly sea surface temperature curves during 1982-1983 in the vicinity of Academy Bay, Santa Cruz Island, Galapagos Islands, derived from three different data sources. Academy Bay observations were made daily, at 0700 at the Charles Darwin Research Station, and also show monthly maximum and minimum temperatures. Ship and satellite observations are from the sources noted in Fig. 12 except that the satellite values were read from a point 30 km south of Academy Bay. Costa Rica, coral bleaching was first observed in June at Samara, Malpais and Manuel Antonio (Cortes et al., 1984), but it may have occurred earlier. According to the temperature synopsis, the duration and degree of warming were similar in Costa Rica and the Gulf of Chiriqui, Panama (Fig. 14). The timing of bleaching in the Gulf of Panama, which is a seasonally upwelling environment, provided the clue that sea warming was largely responsible for this disturbance. Coral bleaching was delayed by 4 months in the Gulf of Panama, during an upwelling period when mean SST remained below 28°C. Corals were affected after the upwelling season when mean SST increased suddenly to near 30°C in June (Glynn, 1984a). It is also evident in Fig. 14, especially from the SST curves for the Gulf of Panama, Colombia, and the Galapagos Islands, that significant warming occurred before 1983. This point is considered below with reference to the Gulf of Panama.
74
___c o r a l
COSTA RlCA 31
t
affected
i
coral a f f e c t e d - - -
PANAMA
PANAMA
25 24 23
COLOMBIA 31 t 30 -
2I
c o)-!j
o f e
d
1
7
L
L
-
1
--A
J F M A M J J A S O N D J F M A M J J A S O N D
Fig. 14. Mean monthly sea surface temperature in eastern Pacific areas with formerly (before 1983) abundant reef coral populations. Thin curves, 1982-83; thick curves, long-term means. Periods of bleaching (broken lines) and coral mortality (solid lines) are noted above for each area. The 1982 and 1983 SST curves for Costa Rica, Gulf of Chiriqui (Panama) and Colombia are based on satellite observations and for the Gulf of Panama (Panama) and Galapagos Islands on mercury thermometer readings. Twelve-year (1947-58) mean curves for Costa Rica, Panama (Gulf of Chiriqui) and Colombia from Renner (1963); 66-year (1907-1972) curve for Panama (Gulf of Panama, Balboa) from Panama Canal Commission records (Meteorological and Hydrographic Branch); 21-year (1965- 1985) curve for Galapagos Islands (Academy Bay) from Charles Darwin Research Station records. Condition of coral from: Costa Rica, Cortes et al. (1984), Cortes, pers. comm.; Gulf of Chiriqui, Gulf of Panama, Glynn (1983, 1984a), pers. obs.; Colombia, Prahl (1983, 1985), pers. comm.; Galapagos Islands, Robinson (1983, pers. comm.
75 2.7 Further ev-i
'
.
. .
sea e -w
of coral bleaching
A good correspondence in the timing of coral bleaching and sea warming has been noted. Now I will consider (a) the close relationship between the severity of coral mortality and the magnitude of temperature anomalies, and (b) the responses of corals to an El Niiio simulation experiment, which were essentially the same as observed in field populations during the 1982-83 El Niiio event. In various areas, coral mortality was related to the intensity of the warming episodes ( G l y et ~ al., in press). Mean reef-coral mortality increased significantly along a gradient from environments with high and stable temperatures to environments with low and variable temperatures: Caiio Island, Costa Rica (50% mortality), Gulf of Chiriqui, Panama (75% mortality), Gulf of Panama, Panama (85% mortality), Galapagos Islands (97% mortality) (Fig. 15). The thermal conditions along this gradient of coral mortality demonstratd (a) an increase in relative SST anomalies, (b) an increase in duration of warming, and (c) an increase in rate of warming (Fig. 16). Perhaps the strongest evidence linking sea warming with coral mortality comes from El Niiio simulation experiments in which Pocillopora damicomis was exposed to slight increases in sea temperature over a 10 week period, similar to the warming that occurred in 1983 (Fig. 17, Glynn
Fig. 15. Percent total reef coral mortality at four eastern Pacific sites. Noted for each site are mean mortality, standard error of mean, range and number of sample areas. Statistically similar mean values are joined above by horizontal lines (SNK test). (From Glynn et al., in press.)
76
and D'Croz, in press). Corals maintained at 30°C and 32OC bleached and showed a general decline over time in (a) the numbers of zooxanthellae present, (b) the concentration of chlorophyll a per zooxanthella, and (c) the protein levels of coral tissues. At 32"C, coral death occurred after 4 weeks, and at 30°C most of the corals were still alive but in a moribund state by the end of the 10 week experiment. At temperatures of 26OC and 28OC, all of the corals had retained their normal color and density of zooxanthellae to the end of the experiment*. Histological analysis of the experimental corals revealed changes with increasing time and temperature that matched the deteriorating condition observed in field corals during the 1983 warming (Glynn et al., 1985b): (a) necrosis of zooxanthellae remaining in gastrodermal tissues, (b) deterioration of epidermal mucous secretory cells, (c) erosion of epidermal and gastrodermal cell layers, (d) atrophy of longitudinal retractor muscles, (e) loss of mesogleal pleat structure, (f) appearance of necrotic nuclei in cells of most tissues, and (g) disappearance of gonads. The
2.
I
I-
0
v
J M M J S N 1980
I
I
m uwL1-v h
1981
h
k
+El 1982
I
173 I
k
Nlfio Period+ 1983
n VIV w
(Caiio Island Costa Rica
.
1984
L
1985
Fig. 16. Sea surface temperature anomalies at four eastern Pacific localities, 1980-1985. The maximum length of the El Nitio warming period (May 1982 - August 1983) is indicated on the abscissa. Least squares regression curves and slope values are indicated for each locality (from Glynn et al., in press). *Statistical testing indicated that time, temperature, and the interaction terms were all significant (p < 0.05.2-way anova), and that all linear regression lines were also significant (p < 0.05, F test). A posteriori testing showed that the low mean values of number of zooxanthellae and protein concentration at 32OC were significantly different from the means of the lower experimental temperatures.
77
(
3t\
0
0
-2
-3
3OoC
c I
\
\32OC
I
I
I
I
I
I
1
2 0 \ a l C
L
J
I
0
2
4 6 Time (weeks)
I
0
10
Fig. 17. Decline in zooxanthellae densities (a) and cord tissue protein (b) in Poci1looor;t damicornis maintained under simulated El Niiio warming conditions. Corals were collected from the Gulf of Panama and the experiment was conducted from 18 August to 21 October 1985. Twelve colonies were monitored in each treatment. Mean ambient temperature was 27.9OC (from Glynn and D'Croz, in press). All data points and detailed statistical analyses are presented in Glynn and D'Croz (in press).
condition of crustacean symbiotes associated with the experimental corals also declined markedly with time at the higher temperatures, and in a manner similar to that observed in 1983 (Glynn, 1985b; Glynn et al., 1985a). Crustacean lipid levels declined, the number of egg-carrying females decreased, emigration rates increased, defensive behavior declined, and mortality rates increased. 2.8 Spatial and temuoral occurrences of El Niiio warminr!events
To investigate a possible connection between El Niiio occurrences off northwestern South America and sea warming episodes in coral reef areas north of the equator, I have compared long-
78
term sea temperature records in the two regions. This analysis was performed employing SST observations from Puerto Chicama, Peru (1925-1987), supplied by D. B. Enfield, and from Balboa, Gulf of Panama, Panama (1915-1987), supplied by the Meteorological and Hydrographic Branch, Panama Canal Commission. The Panama records were obtained at the mine dock, Balboa (1915-May 1980) and at the pilot dock near Naos Island (May 1980-1987) (Fig 2). No significant differences in temperature were evident when the recording station was moved from Balboa to Naos Island (J. A. Brady, pers. comm.). The number of days per year with SST 2 29OC was selected as an index of El Niiio warming in Panama because temperatures this high occurred there during the 1982-83 coral bleaching event and temperatures in this range have been shown to stress corals, The condition of Pocillowra damicomis declined at accelerated rates at experimental temperatures higher than 28-29"C (Fig. 17). To verify the Panama Canal Commission's (PCC) temperature data, they were compared with temperatures measured at the Naos Island marine laboratory of the Smithsonian Tropical Research Institute (STRI). These stations are located on opposite sides of the Ft. Amador causeway, about 3 km south of the entrance to the Panama Canal (Fig. 2). The temperature curves obtained from the two stations show a similar trend over the 19 year (1969-1987) period of record (Fig. 18). The Spearman rank correlation coefficient of PCC against STRI is highly significant (1-2 = 0.774, p < 0.01), and shows a decade-long warming trend (see below). Recent strong (1972-73) and very strong (1982-83) El Nifio events in Peru were accompanied by exceptionally warm sea conditions in Panama (Fig. 19a). However, a correlation analysis (Spearman rank correlation) between the 2-yr smoothed means of the number of days/yr SST 2 29OC in Panama and the annual SST deviations at Chicama, Peru, over a 62 year period, did not reveal a significant association (r2 = 0.026, p > 0.05). Sea temperature deviations in Panama and at Puerto Chicama indicate some, but by no means total, correspondence between El Niiio occurrences and sea warming in a Central American area (Fig. 19b). A scatter plot of 2-yr smoothed temperature deviations indicates that only the 1972-73 and 1982-83 El Nifio events of strong and very strong intensity resulted in pronounced warming in both regions (Fig. 20). All earlier El Niiio events, including the very strong event of 1925-26 (Quinn et al., 1987), had only minor or no warming effect in Panama. From this analysis, it is possible to conclude that since 1925 only two of six strong to very strong El Nifio events in Peru resulted in synchronous warming episodes as far north as the Gulf of Panama. El Niiio sea level signals are usually detected in Panama (Kwiecinski and Chial, 1987), but higher SST values are not usually found. Unexpectedly, the moderate El Nifio event of 1987 (Quinn et al., 1987) was accompanied by a marked increase in SST in the Gulf of Panama (Fig. 19a). Although no coral bleaching was observed in Panama, bleaching of remnant coral populations that survived the very strong El Nifio event of 1982-83 did occur at Cocos Island, Costa Rica and in the Galapagos Islands during the warming period (Glynn, 1988a). Coral recovery occurred after 4-6 weeks with no significant mortality. The long-term temperature data for the Gulf of Panama reveal a prominent, near-decadal warming trend that began in 1976 (Fig. 19a). This is indicated by the number of days per year with high SST (2 29OC): over a 12 year period (1976-1987), 8 years experienced high SST for 100
79
1968
70
72
74
76
78
80
82
84
86
88
Year Fig. 18. A comparison of the number of days per year with high temperatures (129OC) at the Panama Canal Commission (PCC) and Smithsonian Institution (STRI) observation stations at Naos Island (1969-1987). days/year or more. In 1980, there were 160 warm days, and the 1983 El NiAo event was accompanied by 220 warm days. Only during the 1972 El NiAo did the number of warm days per year (153 days) previously exceed 100 days in the Gulf of Panama since 1915. Quinn and Neal (1983, 1984) and Quinn et al. (1987) have also identified this recent, long-term warming trend, which has affected much of the tropical and subtropical Pacific Ocean over recent years. The largely negative Easter-Darwin Southern Oscillation index anomalies, which generally co-occur with above normal SST, began in early 1976 and have persisted until early 1988 when an apparent reversal was first noted (Quinn, pers. comm.). Quinn and Neal (1983) have hypothesized that this decadal climatic change could be due to a weakening of the southeast trade system off the west coast of South America. Since the tolerance of reef corals is influenced by their thermal history (Jokiel and Coles, 1977), the decadal warming trend that encompassed the 1982-83 El Niiio event complicates our understanding of the stress responses observed during 1983. Several studies conducted on coral growth and reef accumulation rates in the eastern Pacific during this period reported vigorous reef building that rivaled the highest rates known for central and western Pacific reef areas (Chave et al., 1972; Easton and Olson, 1976; Glynn, 1977; Glynn and Macintyre, 1977; Grigg, 1982; Glynn
80
N
0
N
0
-
0
0
0
l n o l n g l n
ln
#
0 #
ln ln
ln
0
81
Fig. 19. (a) Number of days per year with daily mean sea temperature 229OC, Balboa and Naos Island, Gulf of Panama, Panama, 1915-1987. (b) Deviations about historic 2-year running average of number days/year with sea surface temperature 529OC, Panama (heavy line), and of mean annual sea surface temperature, Puerto Chicama, Peru (thin line). Strong and very strong El Niiio Occurrences from 1915-1988 are indicated above (Quinn et al., 1987). Panama temperature data from Panama Canal Commission. Peru temperature data from D. B. Enfield (unpub. data). and Wellington, 1983; Kinsey, 1983). Moreover, the histological condition of coral tissues before 1982-83 indicated a generally healthy state (Glynn et al., 1985b). Whether or not the prior, longterm sea warming period had an interactive effect with the very strong 1982-83 El Niiio event is an intriguing, but as yet unanswered question. 2.9 Coral bleachine and mortalitv dun-
1
Periods of intense sea water cooling, when temperatures drop below 18OC for several days or weeks, occur fairly commonly in the major upwelling centers of the tropical eastern Pacific region (Hubbs and Roden, 1964; Forsbergh, 1969; Dana, 1975; Glynn and Wellington, 1983; Glynn et al., 1983). Compared to non-upwelling areas with relatively rich coral faunas and well developed reefs, upwelling environments are characterized by (a) few reefs per unit area, (b) reefs of limited dimensions, (c) youthful reefs, (d) low coral growth rates, and (e) few reef associated species (Dana, 1975; Glynn and Wellington, 1983; Glynn et al., 1983). During periods of intensified upwelling, e.g. at the height of the Little Ice Age (1675-1800 A.D.), it is likely that an entire Costa Rican reef tract succumbed to environmental chilling ( G l y et ~ al., 1983). Like El Niiio, intense cooling episodes occur unpredictably, but unlike El NiRo, which shows a high inter-event variability in terms of the areas affected, cooling episodes are confined to the major upwelling centers. These thermal disturbances are also different in that extreme low temperatures are fleeting, lasting for only a few hours or days (Glynn and Stewart, 1973; Glynn and D'Croz, in press), whereas extreme high temperatures persist for weeks or months (Figs. 14 and 19). Periods of intense warming and cooling in the eastern Pacific tend to occur in succession, linked by major shifts in the Southern Oscillation. Peak anomalies in the atmospheric pressure indices, i.e. anti-El Niiio type activity with strong upwelling, are often followed by relaxation troughs and the onset of El Niiio sea surface warming (Quinn, 1976). Upwelling was unusually strong in the Pearl Islands, Gulf of Panama in 1972 (18.6T was the lowest temperature recorded), and resulted in tissue sloughing and branch tip mortality in pocilloporid corals (Glynn and Stewart, 1973). This notable upwelling season was followed by a strong El Niiio event in 1972-73 (Quinn et al., 1987) that resulted in sea warming in the Gulf of Panama (Fig. 19), but was not of sufficient intensity to cause coral bleaching or mortality (Glynn, 1977). The 1985 upwelling season was probably more intense than that in 1972, and was unusual in that it followed rather than preceded an El Niiio event (Kwiecinski et al., 1988). Anti-El Niiio type activity, with strong southeast trades and upwelling, typically precedes El Niiio events by a year or less (Quinn, 1974; Wyrtki, 1975; Quinn and Neal, 1983). Richmond (this volume) recorded minimum sea temperatures of 14.2OCin March 1985 on coral reefs in the Pearl Islands. Coral bleaching was widespread on the Saboga Island coral reef in March 1985 with 10.4%
82
PANAMA
+
15~1
-
8
I
'
083
,-.
0
40
32
0
"
" A
0
0
0
CHICAMA
0 5 8 041
0
e25
u 0
0
0
0
8';
8000
Fig. 20. Two year running mean deviations of annual mean sea surface temperature at Puerto Chicama, Peru and number of days per year with sea surface temperature 229°C at Balboa, Panama (1925-1987). Strong and very strong El Niiio years are indicated by occluded circles (Quinn et al., 1987). mortality of pocilloporid corals (Glynn and D'Croz, in press). This mortality was assessed at the end of the strong upwelling season with pulses of extreme low temperatures that began in January. The same coral population had experienced 68.5% mortality during the El NiRo warming event in 1983. The most recent coral bleaching associated with intense upwelling (with minimum temperatures of 16OC) was observed in the Pearl Islands in February 1989, but the extent of mortality due mainly to the cooling has not yet been determined (Eakin et al., 1989). Immediately preceding the 1989 bleaching event numerous reef flat corals suffered high mortality due to extreme mid-day low tidal exposures, which often accompany anti-El NiAo type conditions (see
83 below). 2.10 Non-thermal a s s ors and coral bleaching Because coral bleaching and mortality have occurred during periods of heavy fresh-water runoff and sea water dilution (Goreau, 1964; Stoddart, 1969a; Egaiia and DiSalvo, 1982), the exceptionally high rainfall that accompanied the El Niiio of 1982-83 in some areas could have conmbuted to this disturbance event. Heavy rainfall was experienced in the Galapagos Islands. For example, in Academy Bay, Santa Cruz Island, maximum 24 hour rainfall in 1982 (137.6 mm) and 1983 (122.9 mm) was about 6 times the average of maxima recorded in 1979-81 (22.1 mm) (Robalino, 1985). Total annual rainfall in 1983 (2,768.7 mm) was over 10 times the average annual amount reported in 1979-81 (265.1 mm). Heavy downpours at Floreana Island resulted in significant runoff and sea surface discoloration to slightly more than 1 km offshore (Robinson, 1985). Surface salinities that were monitored 3 times a week for 18 months at a station 7 km southeast of Academy Bay indicated slightly below-normal values during the peak El Niiio rainy period (Fig. 21). However, these offshore salinities were no lower than 32.5 o/oo (Kogelschatz et al., 1985), far above dilution levels known to stress corals (Kinsman, 1964; Endean, 1976; and see below). Also, the coral bleaching and death occurred to 30 m depth, well below the influence of surface salinities (Robinson, 1985). Rainfall on the Pacific coast of the Isthmus of Panama during the El Niiio event was close to the long-term annual average for this region. For example, at Balboa Heights the average annual rainfall for a 56 year period (1930-1985) was 1,800.6 mm and total rainfall in 1982 was 12% below and in 1983 3% above the long-term average (Meteorological and Hydrographic Branch, Panama Canal Commission). The observed salinities in the Gulf of Panama in 1982 and 1983 are in accord with the normal to low rainfall records over this period, and are similar to the seasonal range of mean values reported earlier (Fig. 21). Even the low salinity regime observed in Panama over a 2 month period in 1973 (minimum values 19-20 O/oo ) did not result in coral bleaching or death (Glynn, 1974). Thus, I conclude that El Niiio associated variations in salinity probably did not affect corals adversely in the Galapagos Islands or Panama. Heavy runoff can produce increased rates of siltation, which may cause coral bleaching and death (Johannes, 1975; Bak, 1978; Marszalek, 1981). Despite heavy rainfall and runoff in the Galapagos Islands, no evidence of coral mortality caused by heavy siltation was reported there (Robinson, 1985). Excessive runoff, which is usually accompanied by increased sedimentation and turbidity, was not reported in 1983 in Costa Rica (Cortes et al., 1984; Guzman et al., 1987), Panama (Glynn, 1984a) or Colombia (Prahl, 1985). Sudden sea level lowerings associated with ENS0 events in the west and south Pacific have caused disturbances in shallow coral reef populations. Monthly mean sea level can drop to as low as -40 to -45 cm below mean sea level (Yamaguchi, 1975; Wyrtki, 1985), with presumably additive effects on local tidal amplitudes. Reef flat exposures in Guam,during the 1972-73 El NiRo, resulted in the mass mortalities of diverse reef organisms (Yamaguchi, 1975). The disruption of skeletal growth in old reef flat corals in the New Hebrides (Vanuatu) region has also
84
a GALAPAGOS
I
ISLANDS
36
1982
34
S%o
32
1983
30 28
b/
PANAMA
1
36 34 32 30
S %o
28 26 24 22 20
18 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
J F M A M J J A S O N D J F M A M J J A S O N D
Fig. 21. Surface salinity variations in the Galapagos Islands (a) and Panama (b) in 1982-83 compared with previous, non-El Niiio periods. The 1982-83 record for the Galapagos, a few km south of Academy Bay over the 100 m depth contour, is from Kogelschatz et al. (1985); the earlier record (heavy line) is from pooled data for 1968 (Maxwell, 1974) and 1971-72 (Houvenaghel, 1978) and represents monthly means (ranging from 11-48 observations per month) from various areas throughout the archipelago. All Panama data are from Naos Island, Gulf of Panama. Monthly mean and extreme values are shown in 1982 and 1983. The earlier record (heavy line) is for 1965-1967 (United States Coast and Geodetic Survey observations), and the extreme low salinity record for the wet season in 1973 (after Glynn, 1974). been correlated with some non-seismic emergence events during ENSO occurrences (Taylor et al., 1987). And in early 1983, sea level drops in some areas in French Polynesia and the Tokelau Islands resulted in more recent ENSO disturbances to coral reefs (Glynn, 1984a; Coffroth et al., this volume). Unlike ENSO-related sea level lowerings in the west and south Pacific, eastern Pacific El Niiio events are accompanied by high sea levels (Lucas et al., 1984; Wynki, 1985; Hansen, this volume). Preliminary analysis of tidal observations in Panama (Eakin et al., 1989) indicate that exwme tidal exposures are more frequent during non-El Niiio years. No reef flat coral mortality
85 was reported in the eastern Pacific in 1982-83, nor during the strong El Niiio of 1972 and the moderate El Niiio events of 1976 and 1987. However, 4040% of the surface m a of existing reef flat corals suffered severe mortality in 1974 and 1975 (Glynn, 1976), and 97% partial to complete mortality ( > 50% tissue loss) in 1989 (Eakin et al., 1989), both disturbances occuring about 2 years after the El Niiio events. Observations by Robinson (1985) during the El Niiio disturbance have demonstrated that abnormally high sea levels [30-40 cm above long-term means (Wyrtki, 1985)], in combination with spring tides and large and contrary sea swells, caused extensive mechanical damage to shallow coral communities in the south-central Galapagos Islands. Several large colonies of massive corals (Pontes. Pavona) were dislodged and deposited above the high tide line, branching corals (PocilloDora) were tom loose and reduced to rubble, and numerous free-living coral associates, such as sea urchins (Eucidaris) and sea stars (Nidorellia), were cast ashore. A flourishing patch reef located within the Onslow Island (Devil's Crown) crater was thoroughly transformed by adverse sea conditions in early 1983 (Figs. 3 and 22). Over 90% of the pocilloporid reef frame was broken apart and this section of the reef is now a mixed coral rubblehasalt rock bottom. Most massive corals are still in their original, pre-1983 positions, but contain little (usually c 10%) live tissue. A large patch of loose Psammocora coral, formerly present at the west-central reef margin, is now covered with sand. While such physical disturbances produced dramatic effects locally in the Galapagos Islands, reef structures elsewhere in the eastern Pacific, although often mostly dead, remained largely intact immediately following the 1983 mortality event. The eastern Pacific region is geologically active and because coral reefs can be severely affected by volcanic eruptions (Wood-Jones, 1910). tectonic uplift (Stoddart, 1969a, b; Glynn and Wellington, 1983; Colgan and Malmquist, 1987; Taylor et al., 1987) and earthquakes (Stoddart, 1972), it is also necessary to consider these kinds of disturbances in relation to the 1982-83 event. The only likely link between potentially disruptive geologic events and coral mortality was the occurrence of a strong earthquake near the Panama-Costa Rica border (epicenter: 8'.80N, 83'.11W) in April 1983, during the time that corals were dying (Glynn, 1983). This was a strong earthquake (magnitude: M sub"s" = 7.2, M sub"b" = 6.3) with a deep focus (38 f 1.6 km). No surface rupture was reported and it is doubtful that any gases were released. The static displacement at the earths surface was also slight, about 10-3 microns (Bull. Int. Seismol. Centre, 1983). Moreover, since coral reefs were bleached and damaged nearly simultaneously over large parts of the eastem Pacific, at distant locations and in most instances before the Panama-Costa Rica seismic event, this disturbance can be confidently ruled out. As previously noted, coral bleaching at several localities was often most pronounced on the upper portions of colonies that received direct or nearly direct sunlight. Unfortunately, irradiance levels were not monitored near the sites where coral bleaching occurred. Moreover, no information is available on the flux of incoming solar UV radiation (especially the 290-320 nm UV-B band that can penetrate much of the euphotic zone in clear tropical waters), which can be quite damaging to many shallow marine invertebrates (Jokiel, 1980; Siebeck, 1981, 1988; Jokiel and York, 1982). The instrumentation necessary for the accurate measurement of submarine
86
4
a
-N-
Fig. 22. Planar views of the Onslow Island (Devil’s Crown) coral reef showing distributions of major coral and barnacle communities before (1976) and after (1985) the 1982-83 El NiAo disturbance. Note nearly total absence of the pocilloporid reef frame and extensive bed rock basalt exposures in 1985. Most of the massive corals present at site 1 in 1976 were absent in 1985. Similarly, the Psammocora community present at site 2 in 1976 was reduced to only a few scattered live colonies on a dominantly sand/rock bottom in 1985. The 1976 community map is from Glynn and Wellington (1983).
a7
UV radiation is costly and has been used only in a limited capacity thus far. Since there are indications of subtle interactions between high irradiance and temperature, this topic will be discussed below (also see Coffroth et al., this volume). 3 COMMUNITY EFFECTS 3.1 Immediate effects Reef-associated species responded in various ways to the 1982-83 El NiAo event. Some species were affected before or during the height of the warming period, others toward the end of the disturbance, and still others showed changes for several months to years following the disturbance. Those responses that corresponded closely in timing with the primary disturbance event will be considered first, and delayed or longer-term effects, some of which are still in progress (as of mid 1989), are considered later. The most notable community-wide change during the 1982-83 El NiAo event was the sudden appearance of large tracts of stark white coral, which contrasted with patches of brownish or greenish, normally pigmented corals. Branching scleractinian corals (Pocillopora spp.) and branching or platy hydrocorals (Milleporq spp.) were the first to bleach on reefs in Panama (Glynn, 1983), and this was most pronounced at shallow depths (< 10 m). Pocillopora damicornis also bleached before the massive coral Pontes lobata in Costa Rica (Cortes et al., 1984). At Gorgona Island, Colombia, Pocillopora evdouxi retained zooxanthellae and continued to secrete normal amounts of mucus during the early warming period (Prahl, 1983), but died in large numbers, presumably following bleaching, a month later (Prahl, 1985). No live colonies of a branching acroporid coral (Acropora valid& the only known eastern Pacific population in this family (Prahl and Mejia, 1985), could be found after 1983 (H. von Prahl, pers. comm.). Since normally pigmented colonies of this species were first observed at Gorgona Island in September 1983, two months after the main bleaching event (Fig. 14), it is not known if its disappearance was a direct result of the warming disturbance. Massive, platy and nodular corals (e.g. species in the genera Gardineroseris, Pavona, Porites, and Psammocora) generally bleached a few weeks after the branching species in Panama, but all corals were affected at about the same time in the Galapagos Islands (Robinson, 1985). Pavona gigantea and Psammocora stellata were resistant to bleaching in the Galapagos Islands. In Panama, the small, nodular coral Porites panamensis retained a normal greenish-brown coloration for 2-3 months after other corals were affected, but then suddenly died in large numbers near the end of the warming event (August-September, 1983). The high susceptibility of some branching corals to thermal stress has been observed by several workers (e.g., Mayer, 1917; Edmondson, 1928; Jokiel and Coles, 1974). Generally these corals have higher growth rates than more resistant non-branching species. It has been suggested that the sensitivity of branching corals is related to a high respiratory rate and a consequent lowering of the P:R (photosynthesishespiration) ratio at critically high temperatures (Coles and Jokiel, 1977). While most colonies of branching species and small colonies of non-branching species' that experienced severe bleaching died, most large massive colonies that experienced severe bleaching suffered only partial mortality, usually to the upper portion of the colony. Thus, massive colonies that retained live coral tissues after 1983 had the potential to regenerate, as observed in studies
88 elsewhere (Hughes and Jackson, 1980; Jackson, 1983). Some of these large colonies are undergoing regeneration, but many are continuing to lose live coral tissue due to delayed effects such as bioerosion, predation, and damselfish activities (see below). A noteworthy impact of the 1983 mortality event on coral community structure was the sudden loss of coral species and decline in species diversity. Coral diversity (H'),measured before and after the disturbance, declined markedly in Panama and the Galapagos Islands (Table 2), and to a less degree in Costa Rica (Guzman et al., 1987). Although the diversity values reported here were measured in early 1985, about 1.5 years after the disturbance, there was no indication of further significant coral mortality or recruitment in Panama at that time (Fig. 26a). Based on observations in Panama and near total mortality observed by Robinson (1985), I assume no significant recovery in coral cover in the Galapagos Islands over this same period. In the Gulf of Chiriqui, Panama mean species losses per transect along the reef base ranged from about 50% (Secas reef) to 80% (Uva reef) (see site locations in Fig. 2), and most measures of diversity declined to 0 (Table 2). Most evenness measures also declined. However, increases occurred in 3 transects. The increase in J' values was due in large part to the selective mortality of branching pocilloporid corals, which were the predominant species in these transects before 1983. In general, it appears that the loss of species had a greater effect on coral diversity indexes than changes in species evenness (relative abundances). Mean species losses in the Galapagos transects, which included all reef zones (reef flat, slope and base), amounted to 82%, and all H values declined to 0 (Table 2, Fig. 3). Mean coral species diversity (H) at Caiio Island, Costa Rica in 1980 was 0.69, declining to 0.37 in 1984 (Guzman et al., 1987). This difference was nonsignificant, which is not unexpected in light of the relatively low coral mortality observed at Caiio Island in 1983 (Guzman et al., 1987; Glynn et al., in press). Live coral cover and diversity demonstrated a declining nend along the Uva reef base during 1974 when Acanthaster was abundant and actively feeding in this habitat (Glynn, 1976). This decline in diversity during the mid 1970s was attributed to the selective predation of nonpocilloporid corals, present at low relative abundances, by Acanthaster (Glynn, 1976). Acanthaster usually avoids large, intact colonies of Pocillopora spp., the most abundant community members, because of the coral's protective crustacean guards (Glynn, 1983). A comparison of H values from 1974, median = 0.29 (converted to logarithms with base 10, n = 7 transects, Glynn, 1976), with the pre-1983 values, median = 0.24 (n = 10 transects, Table 2), indicates no significant change over the 7 year period (p > 0.05, Mann-Whitney U test). With the steady decline in Acanthaster abundance during the late 1970s to early 1980s (Glynn, 1985a), live coral cover and diversity stabilized until both suddenly declined during the 1983 El Niiio disturbance. Related to pocilloporid coral bleaching were various changes in the density and behavior of obligate crustacean symbiotes (Trapezia spp. and Alpheus lottini). Quantitative sampling in Panama revealed that crustacean densities declined significantly with the deteriorating condition of their coral hosts, from median densities of 9 inds./colony in normally pigmented corals to c 1 indkolony in dead corals (Fig. 8b). A 60% reduction from normal densities was noted for crustacean symbiotes at Gorgona Island, Colombia (Prahl, 1983). This general decline was
ro 0
E,
b s
I
L?
0000.-mmoooo
vrrl
o!o!
vim*
ec**
0000000000
**** ** ** I* ** **
*
b
I
* c*
cA*
I
*
**
wwmb *m d z22zoooooo "lo! 09
*
w
* c* m c*
m*
o
c??
vr
0000000000
Wg\
-!"!
no
0000000000
owwo
1- -! z22zoooooo
O-mbm-11mo
I t c ? l c ? c ? l l c ?
0000000000
89
(D
TABLE 2 (cont'd.)
Locations (Sampling dates)
Galapagos Islands Onslow reef (11-13 Jan. 1975 & 1-3May 1985)
0
Pre 1983
H'b post 1983
0.09 0 0 0.26 0 0.35 0.17
0 0 0 0 0 0 0
Number of Species Pre post 1983 1983
Jd 't'
testC
***
*** *** ***
Pre 1983
post 1983
0.30 0 0 0.88 0 0.74 0.55
0 0 0 0
2 1 1 2 1
0 0 0
3 2 x = 1.7
-
0 1 0 0 1 0 0 0.3
a Locations of sampling sites are indicated in Fig. 2 (Uva and Secas reefs) and Fig. 3 (Onslow reef). b Coral species cover and relative abundances were sampled with a 10-m chain transect (Glynn, 1976). H is the Shannon-Wiener index of diversity
-c
(Pielou, 1975), and was calculated from the equation, H = pi log pi. where pi is the proportion of sampling points encountered in the ith species. i= 1 Differences in H were tested employing the approximate 't'-test of Hutcheson (1970). ns, nonsignificant; *, p<0.05; **, ~ ~ 0 . 0 1 ; ***, p<0.001. dJ', an index of evenness, is based on Pielou's (1969) measure, H'rnma,.
91
probably related to a reduced food supply (mucus +attached debris and lipid depletion) available from the stressed coral hosts and to increased emigration of the crustaceans to other colonies (Glynn et al., 1985a), which involves a high risk of predation (Castro, 1978). In addition, the bleaching of pocilloporid corals increased the visibility of the orange-red crustacean symbiotes, which are normally inconspicuous against their pigmented coral hosts. This greater visibility could have led to increased predation by such visually oriented predators as wrasses, sea basses and triggefiishes. The interphyletic aggressive responses of the crustaceans (e.g., colony defense against Acanthaster) declined significantly with the deteriorating condition of their coral hosts (Fig. 8c). Day and night field observations in Panama revealed that Traoezia spp. inhabiting bleached hosts altered their usual feeding behavior (S. Gilchrist, pers. comm.). On normal hosts, the crabs stimulated the release of mucus by plucking or rubbing coral polyps with their chelipeds and pereopods (walking legs) (Knudsen, 1967; Patton, 1976). On bleached corals, the crabs executed cheliped fanning and pereopod waving, activities associated with filter feeding. These activities increased when concentrated zooplankton samples were released near the crabs. Gut content analyses of crabs treated similarly in the laboratory revealed large amounts of the same crustacean body parts that were released near the feeding crabs. Observations on facultative crustacean symbiotes (hermit crabs and the spider crab Teleophrys), which are not dependent on coralgenerated food products, showed little change in feeding behavior in relation to the condition of their coral hosts. Although not observed, it is also likely that crustacean symbiotes fed on the tissues sloughing from moribund corals during the disturbance event. Large schools of the king angelfish (Holacanthus Dasser’) were observed biting the bleached branch tips of pocilloporid corals, probably ingesting loose tissues. Dead corals were colonized quickly by microscopic and filamentous algae (Figs. 23 and 24), and after several months formerly live coral surfaces were converted into algal turf or coralline algal substrata (Cortes et al., 1984; Glynn, 1984a; Robinson, 1985). At the Uva Island reef, Gulf of Chiriqui, Panama mean live coral cover declined by 75% from pre-1983 levels and within 6-9 months following the El Niiio disturbance much of the substratum was colonized by 1-3 cm thick carpets of red algae [mainly Gelidiopsis intricata intermixed with Hypnaea pannosa and AmDhiroa beauvosii as minor components (Fig. 25)]. At the Contadora Island reef, Gulf of Panama (Fig. 2, no. 8) Wellington and Victor (1985) reported a more than four-fold increase in algal turf about 1 year after the 1982-83 El Niiio event. Dead corals in the Galapagos Islands were also overgrown by algae, and carpets of principally Giffordia and Enteromorpha were commonplace to 20 m depth by the end of 1983 (Robinson, 1985). While herbivorous reef fishes commonly fed on the new supplies of algae, there is no evidence indicating a numerical response of fishes to the generally large increases in benthic algal resources (Wellington and Victor, 1985; Glynn, unpub. data; and see below). In some communities where algae were predominant and coral cover was relatively low, both the algae and corals experienced high mortality in 1983, resulting in large areas of “free space”. Large fields of the brown, macroscopic alga Blossevillea galapagensis disappeared from the lower intertidal zone in the Galapagos Islands in 1983 (Robinson, 1985; see Laurie, this volume, for
92
Fig. 23. Bleached and dead branches of Pocillopora damicornis, Uva Island reef, Gulf of Chiriqui, Panama, 5 m depth, 22 March 1983. (a) partially bleached (upper one-third of picture) and fully bleached branches; branch diameters are ca. 8 mm. (b) diaphanous microphytes growing on dead branches of Pocillopora.
examples of intertidal algae that declined in abundance). Scattered dead pocilloporid corals in growth position were nearly all that remained of these macrobenthic assemblages. (As of April 1989, only small and scattered B. galapapensis thalli were present in Academy Bay, Santa Cruz Island, P. Wheland and F. Walsh, pers. comm.). It is likely that the greatest variety of marine organisms affected during the 1982-83 El NiAo was in the Galapagos Islands (Robinson and del Pino, 1985; Glynn, 1988a). Reef associated gorgonians, rock scallops, barnacles, and an endemic sea urchin experienced increased mortalities
Fig. 24. Bleached and dead hydrocorals (Millepora intricata) and scleractinian corals (Pocillouora spp.), Uva Island reef, Gulf of Chiriqui, Panama, 5 m depth, 22 March 1983. (a) most branches of the two colonies of Millepora (center foreground) were dead, and those of Pocillopora (surrounding the Millepora) were bleached and still alive. Diameter of the large colony of MilleDora ca. 25 cm. (b) close-up view of MilleDora showing filamentous macrophytes that had colonized the dead coral branches.
93
Fig. 25. Nearly continuous 1-2 cm thick carpets of algae overgrowing pocilloporid corals killed in 1983. Saboga Island reef, Pearl Islands, Gulf of Panama, Panama, 5 m depth, 10 July 1984.
or were encountered less frequently during the El Nifio period (Robinson, 1985). Lithophaga spp., mytilid bivalves that bore into coral skeletons, also apparently died in large numbers in the Galapagos Islands in 1983 (Scott et al., in press). Coral mortality was significantly higher in the Galapagos Islands than in Panama and Costa Rica, exceeding 95% overall (Fig. 15). This strong response was correlated with the geographic gradient of highest SST anomalies observed at these localities (Fig. 16). Thus, many of the species responses in the Galapagos could have been due directly to the water warming. However, since several of the impacted species were suspension and filter feeders it is probable that many were also adversely affected by dramatic declines in primary productivity in the Galapagos Islands (Feldman et al., 1984; Barber and Kogelschatz, this volume). Jenneria pustulata, a gastropod corallivore, was one of the few non-zooxanthellate species to die in large numbers on Panamanian reefs in 1983. It apparently suffered most from high temperature stress, but the sudden decline of its mainstay coral prey (Pocillopora spp.) also could have contributed to its demise (Glynn, 1985a). As noted earlier, significant, but isolated damage to Galapagos coral reefs was caused by increased wave assault resulting from contrary swells, spring tides and abnormally high sea level (Robinson, 1985). According to Groves (1985), some resident reef fishes in the Galapagos, largely endemic and cool water species that are also present in Peru and northern Chile, became less abundant in 1983. Several western Pacific and Panamic reef fishes, normally encountered in the warmer waters of the northern Galapagos Islands, increased in abundance throughout the archipelago in 1983 and even expanded their ranges to western island areas (Groves, 1985) where coral reef buildups are rarely present (Glynn and Wellington, 1983). Groves (1985) did not note any changes in abundance of the guineafowl puffer Arothron meleagris in 1983, but its numbers were greatly reduced from 1985 to 1988 compared with 1975 and 1976 (see below). It is impossible to test statistically for
94 changes in fish abundances in most cases because population sizes were not assessed before and during the disturbance. 3.2 Long-term effecrs The long-term monitoring of population abundances in Panama allows an assessment of the effects of the 1982-83 El Nifio warming event on certain key coral reef species. Here I present data on variations in population densities of all reef building corals in the lower reef slope and reef base zones at Uva Island, Gulf of Chiriqui, from 1975-1988. These are compared to pre- and post-El Niiio population sizes of three important corallivore species, a guild of herbivorous fishes and an echinoid bioeroder. Population changes in other reef areas will also be noted where these are known. 3.2.1 Coral communitv c h a w Mean percent live coral cover varied between about 10-20% from 1975 to early 1983 then it declined markedly to 0.6-0.7% after the El Nifio event (Fig. 26a; G l y ~ 1983, , 1984a). It is probable that this decline was more abrupt than indicated because sampling was not repeated until mid 1984. By 1988, mean coral cover had increased to over 2%,but this has not yet reached a level that is significantly different from that observed in the previous 4 years (p > 0.05, Model I1 anova, variates arcsine transformed). The decline in coral cover that occurred in late 1976 was probably due mainly to predation by Acanthaster; the moderate El Niiio event of 1976-77 (Quinn et al., 1987) was undetected on coral reefs in Panama (Fig. 19). Before 1983, a total of 7 to 8 reefbuilding coral species was usually recorded on any given sampling date at Uva Island (Fig. 26a). From most of 1983 through 1985, two species of Pocillopora were nearly the only live corals that remained in the deep reef base zone. Porites panamensis, Psammocora stellata, Millepora inmcata, and Millemra DlatvDhVlh were not found in the transects during this period. In addition to Pocilloaora damicomis only a single small colony of Pavona varians appeared in the transects. By early 1986, small colonies of all five of the above-named scleractinian species began reappearing, but the hydrocorals (Millepora spp.) were still absent from the Uva reef base mnsects. While MilleDora intricatp has been observed in low abundance on other parts of the reef since 1986, MilleDora DIatvDhvlla has not been found (as of early 1989) anywhere in the Gulf of Chiriqui after extensive searching. Significant declines in mean live coral cover (from 16.5% to 5.1%) and number of coral species (from 7 to 4)also occurred after 1983 on a study reef in the Secas Islands, Gulf of Chiriqui, and on coral reefs in the Gulf of Panama (Glynn, 1984a). By January 1984, Gardineroseris planulata and Pontes panamensis had nearly disappeared from the reefs at Caiio Island, Costa Rica (Guzman et al., 1987), and have remained scarce to February 1989 (Cortes and Guzman, pers. c o r n . ) . Pocilloporid corals virtually vanished from coral reefs in the Galapagos Islands, but recruitment to non-carbonate surfaces has occurred after five years (pers. obs.). Only a remnant (< 20 cm) of Q. glanulata remains at Pta. Estrada, Santa Cruz Island, the only known population of this species in the southem (below 0' latitude) Galapagos Islands (Glynn and Wellington, 1983; J. Feingold, pen. c o r n . ) . Frequently, crustose coralline algae have colonized dead coral surfaces and now (1988-1989)
95
Fig. 26. The 1982-83 El Niiio associated declines in live coral cover and species number (a) at Uva Island, Gulf of Chiriqui, Panama in relation to variations in corallivore (b) and herbivore (c) population densities. Mean percent coral cover (rt95% confidence limits of mean) is based on sampling 10 permanently marked, 10 meter-long chain transects (730 point counts of substratudtransect) located along the seaward reef base (see Glynn, 1976). Coral species number refers to the total number of hermatypic scleractinian corals and hydrocorals found at each sampling date. Mean corallivore densities through 1984 are from Glynn (1985a), and to 1989 from later but identical sampling procedure. Vertical lines denote ranges of densities. Diadem& densities are from Glynn (1988~)and later sampling. Herbivorous fish densities are for Scarus spp., Acanthurus spp., Ctenochaetus, and Kyphosus, based on 3 visual transects (800 m2 each) per sampling date from January 1980 - May 1986, and 9 transects per sampling date from May 1987 - February 1989. Mean values c1.0 on abscissa are denoted. occupy considerable space on most affected reefs in Costa Rica, Panama and the Galapagos Islands (pers. obs.). Unfortunately, the population changes of this important group of calcifying
96
organisms have not been studied in detail on eastern Pacific coral reefs. 3.2.2 Resuonses and impacts of corallivores The population densities of two corallivore species in the Gulf of Chiriqui, the crown-of-thorns sea star Acanthaster Dlanci and the guineafowl puffer Arothron meleagris were not noticeably affected during 1982-1983 (Glynn, 1985a). Nevertheless, reductions in coral prey availability apparently increased the impact of these corallivores. A sharp decline in the population density of Acanthaster was observed in 1981, but the specific cause is unknown. This sea star, like many invertebrates with planktonic larval stages, often undergoes pronounced fluctuations in abundance (Birkeland, 1982; Moran, 1986). Acanthaster affected the survival of certain corals after 1983, because of increased access to coral prey that were formerly situated in protected reef refuges. Before 1983, many bottom areas of the Uva reef were surrounded by continuous thickets of Pocillouora spp. coral. Acanthaster seldom crossed these thickets because of the repellent nature of the coral's nematocysts and the presence of crustacean guards that thwarted the approach of sea stars (Glynn, 1983). With the death of tracts of Pocillouora, Acanthaster could now enter these refuge areas that were inaccessible before 1983 and are still feeding (as of February 1989) on formerly protected, massive (non-branching) corals (Fig. 27). While it is too early to determine the number of colonies that will be killed, most will bear growth discontinuities resulting from feeding scars. Continued feeding by Acanthaster, the occupation of feeding scars by damselfish, and sea urchin bioerosion often exacerbate the initial feeding injury. Some of the massive corals under attack by Acanthaster are of great age. One particularly large colony (Gardineroseris planulata) on the Uva reef was cored and its skeleton studied by means of sclerochronology to determine its growth history. The core revealed a continuous growth record of 192 years, suggesting that a disturbance comparable to the 1982-83 El Niiio has not occurred in this area over the past two centuries (Glynn, 1985b; see section 4). Acanthaster predation at Cocos Island (Costa Rica), where coral mortality in 1983 exceeded 95%, has had a measureable impact on surviving corals (Guzman and Cortes, in prep.). The population size of Acanthaster has remained stable from before and after 1983, but this corallivore is now feeding on remnant patches of Porites lobata because more commonly consumed coral prey were virtually eliminated from Cocos Island during the El Niiio disturbance. This increase in mortality rate is similar to the secondary predator effects that were observed on a storm-ravaged Jamaican coral reef (Knowlton et al., 1981; and see below). There was no statistically significant change in the mean densities of Arothron meleamis (p > 0.05, t-test, 4transformation performed) from before ( 5 = 2.4 inds./1,000 m*, 24 transects, 1980-81) to after ( j I = 2.5 inds./1,000 m2, 58 transects, 1983-89) the El Nifio disturbance. Pocilloporid corals constituted about 88% of the stomach contents of puffers in 197071 (n = 14 inds., Glynn et al., 1972) and 98% (n = 10 inds., unpub. data) after the 1982-83 El Niiio. Because the standing crop of pocilloporid corals, the chief prey of Arothron on the Uva reef, was reduced by about 75% (Glynn et al., in press), it is reasonable to assume that a higher proportion of the total coral standing crop was eaten by the puffer after than before 1983. Because
97
Fig. 27. Crown-of-thorns sea star Acanthaster && feeding on a colony of Gardineroseris planulata (Uva Island, Gulf of Chiriqui, Panama, 5 m depth, 22 June 1984). Before 1983, this massive coral was surrounded and protected by continuous live thickets of the branching coral Pocillouora spp. The sea star is 26 cm in diameter (arm tip to arm tip). the puffers apparently are not food-limited (Guzman and Robertson, 1989), no changes were observed in prey selection. The feeding behavior of puffers was very different in the Galapagos Islands and Costa Rica, where a switch to alternative prey occurred. In the Galapagos Islands, where coral prey abundance was reduced by 99-100% (Glynn et al., in press), A. meleamis virtually disappeared. Before 1983, 1 to 2 inds./man-hr. could be found in areas with high pocilloporid coral cover (Glynn and Wellington, 1983). After 1983, puffer abundances declined to 0.01 inds./man-hr. (unpub. data). Coincident with this, the puffer diet shifted from dominantly pocilloporid corals to nonpocilloporid corals (including both hermatypic and ahermatypic species), molluscs, Eucidaris spines and algae (unpub. data). [Due to the low abundance of puffers in the Galapagos Islands after 1983, the diets were typically determined from the feces of live animals held for 24 hrs. and then released.] Four of 13 puffers had their stomachs packed with 1-cm long Eucidaris spine tips. Arothron also fed (two of 13 puffers) on the 1 cm-long spines of Nidorellia grmata, a sea star. These echinoderm spines have an unknown nutritive value, but are similar in size and durability to coral branch tips, suggesting prey selection on the basis of a search image. In Costa Rica, pocilloporid corals were nearly eliminated from the Cafio Island reefs by the El Nifio disturbance (1983) and later (1985) by dinoflagellate blooms (Guzmanand Cortes, 1989; Glynn et al., in press; Guzman et al., in press). Arothron then began feeding on coralline algae and Pontes coral, which were still abundant after the two disturbances. This dietary 'switching' response could aid the recovery of pocilloporid corals when they become rare (Krebs, 1978; Guzman and Robertson, 1989).
98 The corallivorous gastropod Jenneria pustulata was significantly more abundant in 1982 before the El Nifio warming period (15-28 inds./m2) than in 1984 (0-4 inds./m2) (Glynn, 1985a). Later sampling, to early 1989, indicated continued low abundances (0-3 inds.lm2) in 6 of 7 standard sampling collections at the Uva reef (Fig. 26b). Unlike the corallivorous gastropod CorallioDhila in Jamaica, which demonstrated an increase in abundance relative to its coral prey (AcroDora) following severe hurricane damage (Knowlton et al., 1981, in press), Jenneria virtually disappeared from the Uva reef after 1983. The Jamaican gastropod helped to eliminate more than 98% of the post-storm AcroDora coral survivors over a period of 5 months. This delayed mortality on a Jamaican reef species was over an order of magnitude more severe than that caused by the immediate effects of the storm. Jenneria population outbreaks have not been observed on any eastern Pacific coral reefs disturbed in the early to mid-1980s. 3.2.3 Responses and impacts of herbivores With the marked declines in coral abundance during and after 1983, many severely impacted coral reefs in Panama and the Galapagos Islands experienced dramatic increases in sea urchin populations. Before 1983, Diadema mexicanum exhibited mean population densities of 2 to 4 inds./m2 in the lower reef slope zone of the Uva reef (Fig. 26c). Non-quantitative visual observations indicated that these densities persisted in deep reef areas with high coral cover from 1978 through 1983. By mid 1984, the Diadema mean densities had approximately doubled, and from 1985 to early 1989 they have fluctuated between 60-90 inds./mz (Fig. 28). This large increase in sea urchin abundance was likely a result of high recruitment to the dead reef frame habitat that became available after 1983 (Glynn, 1988~).On the dead pocilloporid reef at Onslow Island, Galapagos Islands, the sea urchin Eucidaris thouarsii also showed a notable increase in numbers, from mean densities of 5 inds./m2 before 1983 to 30 inds./m2 after the coral mortality event (Glynn, 1988~).The increase in sea urchins on the Galapagos reef seemed more likely a result of the movement of individuals from high density, concentrated populations at the deep reef edge and in rubble substrata in deeper water. This habitat shift may have been in response to the marked changes in deep coral communities and increased availability of shelter sites and algal food on the dead pocilloporid reef frame after 1983 pig. 22). The dramatic change in the reefscape, from a dominantly live pocilloporid reef in 1982 to a sea urchin/algal dominated community in 1986, is evident in Fig. 29. It is important to note that the increased densities of sea urchins on dead reefs in Panama and the Galapagos Islands did not occur uniformly over all reef substrata, but were patchily distributed. Relatively few sea urchins were present on reef substrata that were covered by damselfish lawns. Sea urchins that attempted to enter damselfish territories were attacked by the damselfish. The fish either executed vigorous spine nipping or ejected the sea urchin by biting and grasping sea urchin spines and then lifting and swimming with them, finally dropping them beyond the temtorial boundary (Eakin, 1987; Glynn, 1988~).Eakin (1987) found that the damselfish ejection behavior in Panama was significantly correlated with lawn quality, fish size, and sea urchin density. Interestingly, Diadema feeding at night tended to avoid damselfish lawns when the damselfish were inactive and not defending their temtories (Eakin, in press). Thus, by a combination of
99
Fig. 28. Linear aggregation of Diadema mexicanum grazing on dead pocilloporid reef frame overgrown by crustose coralline algae (background) and microalgal turf (foreground). Lower forereef slope, Uva Island reef, 6 m depth, 16 October 1987, Gulf of Chiriqui, Panama.
damselfish repulsion and an avoidance of damselfish lawns, numerous reef patches are virtually free of a potentially destructive sea urchin bioeroder. Damselfish also were observed to eject Eucidaris from their territories in the Galapagos Islands, resulting in a patchwork of algal lawns without sea urchins interspersed among dead corals encrusted with coralline algae with numerous sea urchins. Compared with Diadema in Panama, Eucidaris in the Galapagos Islands did not seem to discriminate so strongly between the algae present inside and outside of damselfish territories. Several Eucidaris that were placed on algal lawns from which their host damselfish were denied access fed on the algae and, in a few cases, even excavated the underlying coral rock (Glynn, 1988~).While not immediately apparent, Eucidaris were present in large numbers on dead massive corals, such as Porites and Pavona, that accommodated damselfish territories on their summits. As expected, such colonies on the Onslow reef had low median Eucidaris densities of 1.4 inds./m2 on the summits, whereas 10.0 and 19.1 sea urchins/m2 respectively were present on the sides and undersurfaces of these colonies, outside of damselfish territories (Glynn, 1988c)*. The distribution and abundance of Diadema and Eucidaris on disturbed reefs are important because of their abilities to excavate and erode coral rock. From field studies of the actual erosion rates caused by sea urchins grazing on reef surfaces, it was found that substrata with high sea urchin densities, in both Panama and the Galapagos Islands, were undergoing erosion in excess of the net carbonate production of < 10 kg CaC03/m2/yr measured before 1982 (Fig. 30; Glynn, 1988~).In Panama, non-echinoid, largely internal bioerosion (due to clionid sponges, polychaete annelids, sipunculans, lithophagine bivalves, and cryptic crustaceans), amounted to about one-half of the total bioerosion (10-20 kg CaC03/m2/yr). In the Galapagos Islands, non-echinoid *Similar distribution patterns of Eucidaris were observed during the day and at night.
100
Fig. 29. (a) pre-1983 appearance of the Onslow Island pocilloporid reef (Onslow Island, Floreana Island, Galapagos Islands, 2 m depth, November 1982). Nearly all of the Pocillouora elegans corals were alive. Colonies in center foreground were approximately 30 cm in diameter. (b) high density of the sea urchin Eucidaris thouarsii feeding on microphytes and coralline algae encrusting dead pocilloporid framework at Onslow Island, Galapagos Islands (reef flat, 1.5 m depth, 8 February 1986). Sea urchins in foreground measured approximately 15 cm from spine tip to spine tip.
bioerosion amounted to about 114 to 1/3 of the total bioerosion (20-40 kg CaC03/m2/yr). It is evident that many eastern Pacific reef structures that have accumulated over hundreds of years are now being rapidly reduced to carbonate sediments.
101
L
A
2
E
20
PANAMA
\ 0
GALAPAGOS
-
Y
1
Gulf of Chiriqui
c
0 .-
+
-
10 -maximum
a 0 " 0
minimum
0
0 +
a,
z
10 L
x
\
N
E
\
0 Y
20
C 0 .-
I n
? a,
5
0 ._
m 0
Gams
COCO,
30
Production
Non-echinoid Eioerosion
0
Echinoid Bioerosion (mod dens 1
0
[7
0
Echinold B i o e r o s l o n (high d e n s )
40
Fig. 30. Net carbonate production pre-1982 (above) and carbonate bioerosion post-1983 (below) on coral reefs in Panama and the Galapagos Islands. Letters denote geographic areas and habitat conditions, and numbers the densities (inds./m*) of sea urchins. Production (a-d): a - Gulf of Chiriqui, Uva Island; b - Gulf of Panama, Saboga Island; c - Floreana Island, Onslow Island reef, densities of 5 and 34 70%coral cover, maximum and minimum production levels at inds./mz respectively; d - Fernandina Island, Eui&~&absent; horizontal lines mark ranges of production values (after Glynn and Macintyre, 1977; Glynn and Wellington, 1983). Bioerosion (e-j): e-g, Gulf of Chiriqui, Uva Island; e - lower seaward slope, damselfish absent; f - lower seaward slope, damselfish present; g - upper seaward slope, live pocilloporid substratum; h - Gulf of Panama, Saboga Island, lower seaward slope; i and j, Floreana Island, Onslow Island reef; i damselfish absent; j - damselfish present (from Glynn, 1988~).
Additional preliminary results from the bioerosion studies in Panama and the Galapagos Islands indicate that the rates of bioerosion varied greatly among the different reef substrata. For
102
example, in Panama mean non-echinoid bioerosion was significantly greater on dead coral blocks without damselfish (23.8 gm dry wt CaCO3/m2/day) than on dead coral blocks capped with damselfish algal lawns (9.4 gm dry wt CaCOdm2/day). Reef blocks capped with live coral revealed intermediate mean rates of bioerosion (1 1.8 gm dry wt CaCOdm2/day), which were not significantly different from either of the other values. A similar though more variable trend was found in the Galapagos Islands (Glynn, 1988~).[It was not possible to determine the rates of bioerosion of reef blocks capped with live coral in the Galapagos because this substratum type was unavailable after 1983.1 In the event that bioerosion occurs more rapidly on dead reef substrata than on live coral substrata or on reef substrata with damselfish lawns, it is possible that such differential erosion could result in the block-like morphology that is so typical of pocilloporid reefs (see section 4). It is likely that the intense grazing by sea urchins could have a destructive effect on the recruitment of corals (Schuhmacher, 1974; Sammarco, 1980, 1982). In the Galapagos Islands, numerous small pocilloporid corals were observed in 1988 and 1989 that I presume have settled and recruited since the 1983 mortality event. Surprisingly, nearly all of these corals have recruited to the tops or sides of basalt boulders in coastal areas of moderate to strong wave action, and not to dead reef surfaces. Eucidaris were rarely found on the basalt surfaces to which PocilloDora had recruited, probably because of the difficulty of maintaining a hold on such surfaces, suggesting that coral survivorship was high where Eucidaris mazing was low. Herbivorous fishes (215 cm total body length), such as parrotfishes, surgeonfishes and sea chubs, showed mean densities fluctuating around 4 to 10 inds./1,000 m2 from 1980 through 1988 (Fig. 24 c). The mean pre- and post-El Niiio densities of 7.9 and 6.8 inds./1,000 m2 respectively did not differ significantly (p > 0.05, t-test, transformation performed). Because the dead coral surfaces have been intensively grazed by herbivorous fishes (and echinoids), macroalgae have not preempted all reef substrata. Several workers have demonstrated the importance of herbivorous grazers in preventing the overgrowth of calcifying coral and coralline algae by fleshy (filamentous and frondose) algae. Coral reefs that have been subject to significant reductions in their herbivore populations, e.g. by overfishing or natural massive die-offs, have become dominated by fleshy algal communities and thereby lost much of their calcifying and reef-building capacity (e.g., Wanders, 1977; Hatcher, 1983; Hay, 1984; Lewis, 1986; Hughes et al., 1987; Morrison, 1988). While certain macroalgae that are resistant to herbivores (e.g. Caulema racemosa) may have increased in abundance in Panama, the significant increases seem to be due more to the presence of algal mats defended by damselfish that now occur on formerly live coral substrata. 4 INTERRUPTED CORAL GROWTH AND REEF FRAMEWORK ACCUMULATION: INDICATORS OF SEVERE EVENT OCCURRENCES Two independent estimates of the minimum time that has elapsed since the last El Niiio disturbance comparable to the 1982-83 event have been proposed. The initial damage in 1983 and subsequent erosion of a 192 year old massive coral in Panama has already been noted above (Long-term Effects, section 3.2, and Glynn, 1985b). The death of piiion trees on Floreana Island in 1983, planted by the first settlers sometime after 1832, led Cruz (1985) to conclude that an event
103 as severe as that of 1982-83 has not occurred for over 100 years. Here I examine the ages at mortality of massive corals and coral reef frameworks in the Galapagos Islands and Panama, and consider this evidence in the light of previous severe disturbances to eastern Pacific coral reefs. The linear growth axes of all of the larger colonies of two massive coral species were measured at several sites in the Galapagos Islands (Figs. 3 and 31). The height and maximum diameter of each colony were measured in situ, and the resulting longest growth axes (colony height or radius) are reported herein. Most colony bases were resting directly on a basaltic substratum so that underestimates of colony heights, due to burial by sediments, is not likely. Estimates of the percent live coral were also made on each colony of Pontes lobati\. Since the 1983 coral bleaching event, nearly all large Galapagos corals with a maximum linear growth axis 230 cm, have suffered progressive tissue mortality and erosion (as of early 1989). Populations of two species in Academy Bay showed significant mean size decreases in their dead colony sizes of 19.8 cm (Pontes lobata, n = 12 colonies) and 35.3 cm (Pavona clavus , n = 9) from May 1985 to April 1989 (t-tests, p < 0.01 in both cases). This erosion probably was due chiefly to large numbers of Eucidaris that were feeding on and in the massive corals at all sites (Glynn, 1988~).The hollowing of corals may lead to their collapse. The mean decreases in colony size observed at Academy Bay over the 4 year period, 21.3% in Porites and 36.8% in Pavona, were applied as correction factors to the corals measured at other sites in 1989 to compensate for erosional losses. The present analysis is confined to Galapagos corals because they are undergoing rapid erosion and most colonies will probably not regenerate. Age estimates of the modal classes and oldest corals, calculated as the product of the length of the linear growth axis and the annual growth rate (Pontes = 0.81 cm/yr, Pavonp = 1.2 cm/yr, see Glynn and Wellington, 1983) are indicated for each locality in Fig. 31. Probably most of these estimates should be regarded as minimum ages because of the relatively high incidence of partial mortality events (Hughes and Jackson, 1980; Glynn and Wellington, 1983). It is evident that numerous old corals were affected in 1983 and thereafter. Overall, five Pontes colonies that had undergone irreparable erosion since 1983 had estimated ages that exceeded 300 years (at Santa Fe and Santiago Islands) and one colony of Pavona may have been over 400 years old (Floreana Island). The estimated modal colony ages of Pontes ranged from 93 to 135 years at four islands, and several older colonies that were probably between 250 and 270 years old were present at four sites. The number of colonies more than 150 years old was highest at Santiago Island, but the reason for this is not apparent (Figs. 31 and 32). Based on the ages of the older massive corals that were killed or irreparably damaged in 1983, I conclude that an El Niiio disturbance comparable to the 1982-83 event has not occurred in the Galapagos Islands for at least 200 years and possibly for as long as 400 years. This conclusion is consistent with the results from Urvina Bay, Galapagos Islands, where colonies, exposed by a rapid tectonic uplift in 1954, attest to large Pavona clavu and P o n t a similar long, although punctuated, intervals of growth (Colgan, this volume). For example, a single large colony of Pavona c u (max. diameter, 12 m), estimated to have been 350 years old at the time of the uplift (Dunbaret al., 1987) survived at least 31 strong or very strong El Niiio events (Quinn et al., 1987) indicating that these earlier events were not as severe as the El Niiio
104
IoL
81 642-
-
-
,
S A N T A C R U Z IS. ( A c a d e m y Bay)
I PL. n = 2 4 ( o l i v e )
P
sPC.n=io
104 yrs (PL) 7 8 yrs(PC) 98 yrs (PC)
I
FLOREANA
I
I
I
I
I
I
I
l
I
1
IS.( N E shore)
el35 y r s
4
2
Y
40
20
60
8 0 100
262 yrs
423 yrs
200
400
6 0 0 800 lo3
200
400
600 800 lo3
S A N C R I S T O B A L IS. ( P t . Pitt)
I PL, n-21
(I live. ( 0 . 1 %t i s s u e )
~ 9 y3 rs
60 80 100
40
20
S A N T A F E IS. (NE anchorage)
I P L , n = 18 (12 l i v e , 1 8 %t i s s u e ) m 42c -
~ 9 y7 rs
2
1
I PL, n = 23
I
( 1 9live,
I
l
l
1111
~ 3 4 yrs 7 1 1 1 1
1
I
1
I
I
I
I
I
I*/. t i s s u e )
268 y r s
z 20
40
60
80 100
200
400
6 0 0 800
lo3
L i n e a r G r o w t h A x i s (cm) Fig. 3 1. Maximum linear growth axes (cm) in dominantly dead and eroding massive corals resulting from the 1982-83 El Niiio disturbance. Estimates of modal and oldest colony ages are noted for Porites l o b a (PL) at five sites and for Pavom (PC) at. two sites in the Galapagos Islands (April, 1989). The numbers of Pontes colonies with live coral assue and the mean percent tissue coverage of colonies that experienced partial mortality are also noted. event of 1982-83. The numbers of poritescolonies that still bore small patches of remnant live tissue are also noted in Fig. 31. Massive Porita colonies at Santa Cruz and San Cristobal Islands were totally
105
Fig. 32. Extensively eroded colonies of Pontes l u , Bartolome Island (south side), near Santiago Island, Galapagos Islands, 10 m depth, 25 April 1989 (photo by J. Feingold). Mushroom-shaped colonies in foreground, about 1.5 m high, contained only small patches of live tissue. dead except for one colony at the latter site with < 0.1% live tissue. At Santa Fe and Santiago Islands most colonies still contained some live tissue in 1989, but these were subject to intense bioerosion and fragmentation by Eucidaris and grazing fishes (Figs. 31 and 32). Large massive colonies had significantly more live tissue (% of total colony area) than small colonies (p c< 0.001, z = 0.321, Kendall rank correlation test), a size-dependent demographic correlation with survivorship that has been found among numerous clonal organisms (e.g., Hughes and Jackson, 1980, 1985; Jackson and Hughes, 1985). Long-lived pocilloporid frameworks in the Galapagos Islands and Panama also were seriously damaged in 1983. The discrete coral blocks that are often present on these reefs are now examined in terms of their size and age distributions to determine the number of years of reef growth between disturbance events. All of the larger (2 30 cm high, 250 x 50 cm square) pocilloporid blocks present at each sampling site were measured in situ. The growth of the reef framework depends not only on coral growth, but also on the growth and binding effects of other calcifying organisms, bioerosion, depositional processes, submarine lithification, etc. (Chave et al., 1972; Hubbard, 1985). Therefore, reef accumulation rates were used to estimate reef growth instead of coral growth rates. The accumulation rates (cdyr) of the reefs in Panama were determined by radiometric dating (C-14) of reef frame carbonates from different depths in the reef formation (Glynn and Macintyre, 1977). Reef accumulation rates in the Galapagos were estimated from coral net production rates, assuming a porosity of 50% and maximum coral skeletal density of 2.7 (Chave et al., 1972; Glynn and Wellington, 1983, Table 26). In the Galapagos Islands, the vertical thickness of pocilloporid frameworks in the modal class (0.6 m) indicates almost continuous reef accumulation over a period of 110-120 years (Fig. 33). A
106 GALAPAGOS I S L A N D S (7sites) n.34 122 y r s
,107-
0.2 0 4 0.6 0.8 1.0
1.2 1.4
1.6
1.8 2 0 2.2 2.4 2.6 2.8 3.0
16
18 2 0 2 2 2 4 2 6 2 8 3 0
P A N A M A (Uva Island)
8r
0 2 0 4 06 0 8
-
12
10
P A N A M A (Secas Islands) n=55
12 -
14
y189 y r s
v)
;l 0 -
E 23 m
8-
:6 -
z *-
0 L Q,
fl
E
3
z
4-
2I
I
I
I
I
I
I
I
I
I
I
,
,
I
t
I
Fig. 33. Frequency distributions of pocilloporid reef frame block thickness in the Galapagos Islands and Panama. Age estimates are calculated from pocilloporid accumulation rates of 0.56 cm/yr (low Eucidaris abundance) and 0.49 cm/yr (high Eucldaris abundance) for the Galapagos Islands (Glynn and Wellington, 1983), and from 0.74 cm/yr for Panama (Glynn and Macintyre, 1977).
107
secondary mode indicates that several reef blocks attained 1.O-1.2 m in thickness over a period of 180-200 years. The thickest accumulation of 2.7 m occurred in Cormorant Bay, Floreana Island, and suggests that the fringing reef there experienced uninterrupted growth during the last 500 years. In the Gulf of Chiriqui, Panama reef block thicknesses of 1.0 to 1.5 m were most frequently encountered, suggesting that uninterrupted reef growth occurred there commonly over a period of 135-175 years (Fig. 33). Several notably thick sections in the Secas Islands (2.0-2.5 m) suggest that some Panamanian reefs grew continuously for about 300 years before the 1983 mortality event. The great differences in coral colony and reef frame ages indicate that only a gross estimate of the time of earlier disturbances can be obtained from this sort of analysis. Thus, I conclude from the ages of corals and reef frame structures damaged in 1983 that a disturbance comparable to the 1982-83 El Niiio event has probably not occurred in Panama and the Galapagos Islands for at least 200 years, and perhaps in the Galapagos for as long as 400-500 years. The possible shorter time between disturbances on coral reefs in Panama compared with the Galapagos could be due to the continental location of the former and other types of disturbances, such as flooding, siltation, nutrient loading, Little Ice Age coolings, and dinoflagellate blooms (Glynn, 1977; Glynn and Macintyre, 1977; Glynn et al., 1983; Guzman et al., 1987, in prep.). Some of the other methods of study that can offer finer temporal resolution, e.g., dendrochronology (Lough and Fritts, this volume), sclerochronology and stable isotope analyses (Druffel, 1985; Druffel et al., this volume; Dunbar et al., 1987; Caniquiry et al., 1988) and trace element indicators (Shen and Sanford, this volume), should provide more precise information on the timing and intensity of El Niiio disturbances.
5 DISCUSSION A N D CONCLUSIONS 5.1 Sea warminp as the urimarv cause of coral bleaching and mortality Several lines of evidence, ranging from local and regional field correlations to a controlled simulation experiment, indicate that the 1983 widespread bleaching and mortality of eastern Pacific reef corals resulted from relatively small increases in sea water temperature that persisted for weeks to months. This conclusion is consistent with several earlier studies in the cenmal Pacific that have quantified coral bleaching in response to natural and anthropogenic thermal stress (e.g., Edmondson, 1928; Jokiel and Coles, 1974, 1977; Coles, 1975; Coles et al., 1976; Coles and Jokiel, 1977). In addition, coral bleaching and mortality were reported on several reefs in the tropical western Atlantic (Glynn, 1984a; Lasker et al., 1984) and the western Pacific and Indian Ocean during the sea warming events that accompanied the 1982-83 ENS0 (Glynn, 1983; Brown, 1987; Coffroth et al., this volume). More recently, coral bleaching and mortality occurred in the eastern Pacific during the 1987 El Niiio event and in the tropical western Atlantic region in 1987 (Williams et al., 1987; Glynn, 1988a,b; Hollings, 1988; Ogden and Wicklund, 1988), and in 1989 during a period of non-ENS0 activity (Williams and Bunkley-Williams, in press, in review). In most instances, these recent disturbances are associated with elevated seawater temperatures alone or in combination with high irradiance, increased sedimentation, and calm sea conditions with reduced water flow (Causey, 1988; Goenaga et al., 1988; Lang, 1988; Sandeman, 1988a).
108
Unfortunately, in most cases long-term sea temperature records (and variations in other physical conditions) are unavailable at disturbed sites so that attempts to establish correlations and causes of coral bleaching are not well founded. When particular environmental conditions are invoked to explain a given disturbance, measurements, often obtained from distant locations, may or may not be relevant. Explanations of the 1987 Caribbean-Bahamian coral bleaching event suffered from this problem. Numerous workers claimed that the bleaching occurred on coral reefs during periods of anomalously high sea water temperatures (Ogden and Wicklund, 1988; Williams and Bunkley-Williams, in press) yet a comprehensive study of the sea temperature conditions in off-reef (oceanic) settings did not find evidence of any unusual temperature increases (Atwood et al., 1988). It is difficult to reconcile this conclusion with the former reports unless detailed, in situ records become available. Furthermore, not only must the microclimatic conditions be described during the disturbance event, but it is also necessary to consider the thermal history of the study area, the duration of the warming period, and how these stresses compare with the tolerances of local populations (Brown and Howard, 1985). While numerous coral bleaching and mortality events coincident with sea warming were described before the 1980s, these involved only shallow reef flat and lagoonal populations over limited geographic areas (Stoddart, 1969a; Johannes, 1975; Endean, 1976). The warming and bleaching event of 1982-83 was remarkable in terms of its seventy, influence at all reef depths, and scale of impact over large geographic areas. The ages of massive corals and reef frame blocks that were killed or severely damaged in 1983 suggest that an ENS0 warming disturbance of this magnitude has not occurred in the tropical eastern Pacific for at least 200 years (Glynn, 1985b; and present results). The seventy, depth range, and scale of coral bleaching in the tropical western Atlantic during 1987 were also extraordinary. The widespread bleaching events that occurred on the Great Barrier Reef complex in 1982 and 1987 were probably the most severe ever reported (Williams and Bunkley-Williams, in press). The Australian event in 1982 was linked to high sea temperature (Oliver, 1985), high irradiance (Fisk and Done, 1985), and a possible interaction of both factors (Harriott, 1985; Coffroth et al., this volume). Since the majority of the organisms affected during the initial warming event were zooxanthellate species*, it is of interest to consider why this particular symbiosis is so sensitive to slight increases in temperature and/or light. Sandeman (1988a & b) has proposed a mechanism of bleaching that is consistent with the patterns observed in the Caribbean in 1987. Oxygen is produced during photosynthesis by zooxanthellae present in the coral's gastrodermal tissues. Under optimal conditions, oxygen production is balanced by reduction through diffusion and respiration. However, under conditions of accelerated photosynthesis, high concentrations of oxygen produce toxic free radicals (superoxide anions) that accumulate and exceed the levels that can be tolerated by the host cells. This toxic effect may result in the destruction of host cells and/or
*Other zooxanthellate taxa that bleached in various tropical regions in 1987 included sea anemones, soft corals, zoanthids, gorgonians and giant clams (Williams and Bunkley-Williams, in press, in review). The circumstances surrounding the bleaching of giant clams are unknown, however, some giant clams (Tridacna gigas) bleached and died during high temperature stress, but they were located in artificial ponds above the intertidal zone (Estacion and Braley, 1988).
109 the expulsion of zooxanthellae. Both high temperature and high irradiance can increase photosynthesis and thus oxygen toxicity in zooxanthellate species. The bleaching response of the delicate coral-dinoflagellate symbiosis may be a more sensitive indicator of subtle changes in environmental conditions than documented trends in sea warming or increases in solar radiation. 5.2 El Niiio 1982-83 compared with other disturbances
In light of the environmental setting of eastern Pacific coral reefs, potential sources of disturbance could involve any or all of the physical factors considered in section 2.10 (e.g., excessive fresh-water inputs, sedimentation, mechanical damage from waves and swells, volcanism, and tectonic activity). While some of these factors must have important impacts locally, and all may contribute to a "baseline" regional mortality (Coffroth et al., this volume), there is no compelling evidence that such conditions have resulted in widespread, catastrophic effects in the eastern Pacific during recent times (Porter, 1974; Dana, 1975; Glynn and Wellington, 1983; Colgan, this volume). Violent storms and associated secondary disturbances can sometimes produce devastating effects to coral reefs (Woodley et al., 1981; Knowlton et al., 1981, in press; Coffroth et al., this volume). However, the Pacific coast of Central America and northwestern South America is rarely exposed to hurricanes. Because, most hurricanes originate off the southwestern Mexican coast near 10°N latitude and move toward the northwest where they dissipate near 30°N thereby missing the reef areas (Hubbs and Roden, 1964). Since upwellings and tidal exposures are the most commonly observed physical agents of coral mortality in the tropical eastern Pacific, their effects are considered below. Coral mortalities result from strong upwelling episodes and extreme low water exposures more frequently, but with less damage, than from very strong El Niiio warming events. Since 1970 at least three upwelling and three low water disturbances have resulted in notable coral mortality in Panama (Glynn and Stewart, 1973; Glynn, 1976; Eakin et al., 1989). Occasionally coral reefs experience intense cooling and extreme tidal exposures simultaneously during anti-El Niiio conditions @akin et al., 1989). The 1985 upwelling season was remarkably strong and resulted in 10% coral mortality compared with 68% mortality in the same area during the 1982-83 El Niiio event (Glynn and D'Croz, in press). Extreme tidal exposures in 1974 resulted in 4040% coral mortality on reef flats (Glynn, 1976). These disturbance events, however, were geographically and spatially more restricted than the 1982-83 El Niiio disturbance, which resulted in 75-85% coral mortality overall on all reefs and across all zones (Glynn et al., in press). On the one hand, upwelling disturbances are more restricted because they are confined to upwelling centers, e.g., to the Gulf of Panama, and low tidal exposures affect only reef flat corals. On the other hand, El Niiio events influence marine biota over a wide geographic region, but apparently even severe El Niiio events rarely disturb coral reefs. Over the past four and a half centuries (1541-1989), only eight very strong El Niiio events probably occurred in the eastern Pacific (Quinn et al., 1987), and it is doubtful that all of these affected the tropical eastern Pacific reef tract as did the 1982-83 El Niiio event. The death of a Costa Rican coral reef tract from 150 to 300 years B.P. in the upwelling center of the Gulf of Papagayo suggests that intense cooling episodes during the culmination of the Little
110
Ice Age (from ca. 1550 to 1850 A.D., see Grove, 1988) may have had a greater impact on corals in the recent past than today (Glynn et al., 1983). The historical records of El Niiio-type activity suggest that the intensity and frequency of El Nifio events during the Little Ice Age and since, have remained similar to their activity in recent years (Quinn et al., 1987; Enfield, 1988). This raises the possibility that eastern Pacific coral reefs may have been subject to both cooling and warming disturbances during past periods of global cooling. Dinoflagellate blooms associated with anti-El Niiio conditions can cause mortality among corals and other reef organisms such as molluscs, crustaceans and fishes (Guzman et al., in press). In 1985.90-100% coral mortality was observed at Caiio Island, Costa Rica, and 13% at Uva Island, Panama (Fig. 2). both non-upwelling areas, mainly at shallow depths, during a period of intense dinoflagellate blooms. This kind of disturbance, however, apparently occurs less frequently than cooling and tidal disturbances, and probably affects much smaller areas. Corallivores, especially the sea star Acanthaster Dlanci and the gastropod DruDella cornus have caused catastrophic damage to coral reefs over large areas of the Pacific and Indian Oceans (Endean, 1973; Pearson, 1981; Birkeland, 1982; Moyer et al., 1982; Potts, 1982; Moran, 1986; Ayling and Ayling, 1988). Acanthaster and the gastropod Jenneria Dustulak commonly are present on some eastern Pacific reefs where they consume large amounts of coral. However, Acanthaster is absent from upwelling centers (Glynn, 1974), and Jenneria generally appears to have a greater impact on reefs in areas of constant high sea temperature (Glynn and Wellington, 1983, Table 25). The reported occurrence of Acanthaster in the Galapagos Islands (Moran, 1986), and a reference to outbreak populations there in 1889 (Phillips, 1987), are probably in error (Madsen, 1955; Glynn and Wellington, 1983). Although high level predation by Acanthaster and Jenneria may result in the loss of 10% (Glynn, 1973) and 26% (Glynn et al., 1972) respectively of annual coral growth, these corallivores have not been found at exceptionally high population densities on eastern Pacific coral reefs (Glynn, 1974, 1982, 1984b, 1985a; Glynn et al., 1982; Guzman, 1988a). From the evidence at hand, I conclude that no recent physical or biological disturbances can be identified in the tropical eastern Pacific that match the widespread and catastrophic coral mortality that occurred during the 1982-83 El Niiio event. By comparing the 1982-83 El Niiio coral mortalities in the eastern Pacific with other coral mortality events over the same period in other regions, it is seen that the eastern Pacific disturbance was extraordinary in its severity and scope (Brown, 1987; UNEPAJCN, 1988; Coffroth et al., this volume). Probably the only other major disturbance that was comparable with ENS0 in terms of its damage to coral reefs over large areas and at all depths were the outbreaks of Acanthaster observed in certain areas of the central and western Pacific region especially from the 1960s through the 1980s (Moran, 1986; UNEPAJCN, 1988, vol. 3). Disturbances from increased sedimentation and nutrient loading, due
w,
in large part to man's land clearing activities, are also commonly observed in all of the world's coastal waters, and are increasing in scope (Johannes, 1975; Rogers, 1985; Hallock and Schlager, 1986; Birkeland, 1987; Salvat, 1987). These changes, however, perhaps even related to increasing incidences of Acanthaster outbreaks, are often insidious and, at increasing distances, difficult to link with disturbance sources.
111
Acanthaster outbreaks in the late 1960s and early 1970s resulted in extensive damage to coral reefs in many parts of the Ryukyu Islands (Nishihira and Yamazato, 1974; Yamaguchi, 1986) and over great stretches of the 1,200 km long Great Barrier Reef complex (Moran, 1986). The severity of Acanthaster and El Niiio disturbances can result in nearly total (9599%) coral mortality over parts of their range of influence. Small cryptic corals, however, often survived Acanthaster predation (Randall, 1973; Colgan, 1982,1987; Done, 1985), but were not spared during the 1983 El Niiio event (Glynn, 1984a). Both disturbances are recurrent -- secondary outbreaks of Acanthaster predation occurred in southern Japan and on the Great Barrier Reef, and El Niiio coral bleaching and mortality recurred in the eastern Pacific and Caribbean in 1987 -- but since their study is so recent it is not possible to make further comparisons of return intervals (the average time between disturbances, see Sousa, 1984), regional frequency, and long-term effects. Paleoecological studies of the skeletal elements of Acanthaster in reef sediments (Walbran et al., 1989; Walbran and Henderson, in press) and the occurrences of El Niiio in relation to fossil reef structures (Colgan, this volume) may offer additional clues regarding the role of such disturbance events during the Holocene and perhaps earlier periods of reef development. It is noteworthy that several of the coral bleaching events of the 1980s coincided with ENSO occurrences, Some of the physical effects that accompanied the 1987 ENSO in the tropical western Atlantic, i.e. (a) elevated seawater temperatures, (b) reduced wind-driven circulation, and (c) lowered sea levels, are all potentially stressful to corals and suggest that global atmospheric disturbances may be causally related to many of the merit bleaching events (Ogden and Wicklund, 1988; Williams and Bunkley-Williams, in review). While sharing many similarities in timing and patterns of response, the 1982-83 and 1987 El Niiio events showed reversals in severity. That is, the earlier event caused a greater impact in the eastern Pacific than in the western Atlantic and the later event resulted in higher incidences of bleaching and mortality in the western Atlantic than in the eastern Pacific (Glynn, 1988a & b; Jaap, 1988; Williams and Bunkley-Williams, in press; Coffroth et al., this volume). Overall, however, coral mortality and damage to coral reefs were greatest in the eastern Pacific in 1983, where El Niiio oceanographic perturbations are most evident. Disturbance events have continued world-wide to 1989, and Williams and BunkleyWilliams (in review) have made the provocative though unfounded proposal that continuing widespread coral reef bleaching disturbances are a result of increased global temperatures, and increasing deterioration of reef-associated organisms due to a variety of local causes (e.g., sedimentation, eutrophication, industrial and/or agricultural chemicals, sewage, and disease), and ENSO events. 5.3 ProsDects for coral reef recovery Because several eastern Pacific coral reefs have been monitored over the nearly 6 year period since they were disturbed in 1983, it is possible to comment briefly on the status of some of the major processes of recovery. According to Pearson (1981), recovery involves the restoration of coral assemblages to a degree comparable to their original state before the disturbance. Although none of the monitored coral reefs has recovered by this definition, it is instructive to examine relevant restoration processes and to comment on them in relation to reef condition. I will consider
112
here the potential importance of (a) coral reproduction, (b) patch size, (c) the availability and location of source populations, (d) the impact of grazers and bioeroders, (e) competition for space, and (f)the role of secondary disturbances. Continuing studies are in progress and fuller accounts of these subjects will appear elsewhere. Recovery on many Indo-Pacific coral reefs has been greatly enhanced by sexual reproduction and the settlement of coral larvae onto damaged reef surfaces (Connell, 1973; Gngg and Maragos, 1974; Loya, 1976; Pearson, 1981; Maragos et al., 1985; Colgan, 1987). In the eastern Pacific, relatively few coral recruits have been observed on damaged reefs. The reason for this must be due in part to the low reproductive potential of Pocillouora damicornis and probably other coral species (Richmond, 1985, 1987; unpub. data). Preliminary indications from long-term (4 years) histological examination of the gonads of corals important in eastern Pacific reef building, namely, Pocillouora elegans, Pontes lobata. Pavona clavus, and Gardinerosens ulanulata, also suggest relatively low-level reproductive activity in Costa Rica, Panama, and the Galapagos Islands. A few minor reef-building corals in Costa Rica (Psammocora stellata) and Panama (Pontes panamensis, Pavona varians, and the hydrocoral Milleuora intncatd have shown moderate gonadal activity and/or recruitment, thus indicating a potential for recovery of some species by chiefly sexual means. These species are recruiting to large patches of dead reef surfaces that are 10s to 100s of meters distant from surviving source populations as predicted by Connell and Keough (1985) for sexually reproducing species. Coral recolonization also may occur by the regeneration of partially damaged colonies or by the fragmentation and regrowth of surviving colonies (Bothwell, 1981; Tunnicliffe, 1981; Highsmith, 1982). This can be an important means of recovery in disturbed patches of small size with nearby surviving corals quickly filling in the available substrata by vegetative propagation (Connell and Slatyer, 1977; Connell and Keough, 1985). Recovery by fragmentation can be important following storm disturbances (Highsmith, 1982), and recovery from the regeneration of small, cryptic remnants of tissue not killed by Acanthaster has occurred frequently on reefs subject to massive predation events (Done, 1985; Colgan, 1987). Since corals with cryptic habits and those present on open reef surfaces have suffered comparable mortalities (Glynn, 1984a), regeneration by such means has not been notable on eastern Pacific reefs. In severely impacted eastern Pacific areas, such as the Galapagos Islands, remnant tissue survival has been most frequent on large massive colonies. However, the regeneration of such remnants has been slow at best. Due to intense bioerosion and fragmentation, many coral remnants have been separated from their parent colonies. While coral fragmentation has played an important role in the recovery of branching and foliaceous coral populations elsewhere (Gilmore and Hall,1976; Bothwell, 1981; Tunnicliffe, 1981; Highsmith, 1982; Hughes and Jackson, 1985), in light of the continuing intense bioerosion it is difficult to predict the success of coral fragmentation in eastern Pacific reef recovery. To some extent, the recovery process also is being aided by biogenic fragmentation: at Catio Island, Costa Rica Porites lobata fragments have been produced and dispersed by triggerfish feeding on endolithic bivalves (Guzman and Cortes, 1989), and at Uva Island, Panama PocilloDora spp. fragments have been broken loose and dispersed by pufferfish feeding directly on the branch tips of live coral (Glynn, 1985a). Dispersal by this means typically is limited to within a few
113
meters of the parent colonies and is more successful on dead coral than sand substrata. The recovery of species richness in eastern Pacific coral communities will probably be greatly delayed because of a combination of two possible regional extinctions, and several local extinctions, and to numerous extreme reductions in population sizes. The rate of community succession is typically slowed when the area disturbed is large and the disturbance event is extreme (Connell and Slatyer, 1977). The reestablishment of AcroDora and MilleDora Dlatvphvlla in the eastern Pacific may depend upon long-distance dispersal from central Pacific populations (Richmond, this volume). Where source populations of affected coral species are nearby, such as in Costa Rica and Panama, modest coral recolonization has occurred. Little to no recolonization has occurred on the Galapagos reefs that experienced catastrophic mortality and that are far removed from surviving corals. Modest recruitment of PocilloDorq spp. to non-reefal substrata, typically basalt boulders, has been observed recently (1988 and 1989) on several islands in the Galapagos, but the source of planulae is unknown. While no obvious long-term changes in the abundances of herbivorous reef fishes have been noted following the 1983 coral mortality event (Fig. 26; unpub. obs.), dramatic increases in sea urchin abundances on dead carbonate surfaces have occurred in Panama and the Galapagos Islands (Glynn, 1988~).The high population densities and feeding activities of sea urchins, still evident after 5 years, are causing the rapid erosion of reef frame structures. This is reducing the availability of stable substrata for the settlement of coral larvae and may also be limiting coral recruitment through the direct mortality of young stages (Sammarco, 1980). Notable decreases in reef herbivore abundances also may be detrimental to reef recovery, as exemplified by recent events in the Caribbean. Following severe humcane damage on Jamaican nolth coast coral reefs in 1980 (Woodley et al., 1981), a mass mortality event of the sea urchin Diadema antillarum in 1983 (Lessios et al., 1984; Hughes et al., 1985) nearly eliminated this important grazer from reefs harboring already low herbivorous fish abundances due to overfishing (Woodley, 1979). With grazing greatly reduced, macroalgae have increased in abundance and are interfering with coral recruitment and regeneration in shallow reef areas (Hughes et al., 1987). Thus, reef recovery can be influenced greatly by unusually high or low grazer abundances (Littler and Littler, 1984). Intraphyletic competition for space among eastern Pacific corals is now practically nonexistent except where poritid and pocilloporid corals have survived as continuous stands on some Costa Rican and Panamanian reefs respectively. In such areas, coral growth is nearly vertical due to the proximity of neighboring colonies. Due to generally low abundances of macroalgae and large sessile animal populations, interphyletic competition is also virtually nonexistent. Damselfish algal lawns that have become established on dead reef surfaces since 1983 would appear to be the chief competitors of corals for space. The low coral diversity values reflect the recency and severity of the disturbance (Connell, 1978), and the low recruitment to dead reef surfaces suggests that competitive interactions among the macrobenthos are at present unimportant. Several kinds of secondary disturbances have affected eastern Pacific coral reefs after the 1983 bleaching event. The chief physical disturbances have occurred during strong upwelling and cooling (Glynn and DCroz, in press), and during mid-day reef flat exposures at extreme low tides (Eakin et al., 1989). These disturbances, coincident with anti-El Niiio type conditions, resulted in
114
limited bleaching and mortality of corals that survived the 1983 warming event. The 1987 El Nifio event of moderate strength was associated with only minor coral bleaching and mortality in the Galapagos Islands and at Cocos Island, Costa Rica (Glynn, 1988a). Reef recovery would appear to be more influenced by the varied and persistent forms of biotic disturbance. As mentioned above, sea urchins increased dramatically in abundance on many reefs after 1983 and their activities are causing severe bioerosion that could delay or prevent reef recovery. Coral regeneration and recruitment will also be adversely affected in reef areas that became accessible to Acanthaster after 1983 (Glynn, 1985b). Damselfish that have colonized dead coral surfaces have in some cases resulted in additional coral death through the enlargement of their territories (pers. obs.). The outcome of this effect is complicated, however, because damselfish also protect nearby corals from Acanthaster attack (Glynn and Colgan, in press) and from external bioeroders such as sea urchins and fishes (Williams, 1981; Wellington, 1982; Sammarco et al., 1986; Eakin, in press; Glynn, 1988~).Predator concentration, as observed in Jamaica following a primary storm disturbance (Knowlton et al., 1981, in press), was not observed on most eastern Pacific coral reefs. At Caio Island, where pocilloporid corals were severely reduced in abundance after the 1982-83 El Niiio disturbance and dinoflagellate blooms in 1985, corallivores could have
an important effect on the survivors. Guzman (1988a and b) suggested that the elevated levels of predation by specialist corallivores, especially the predatory snail Quovulq monodonta, could retard the recovery of these corals. Also, the regeneration of small patches of Pontes lobata at Cocos Island has been adversely affected by Acanthaster and Arothron (Guzman and Cortes, in prep.). In conclusion, it appears that many eastern Pacific coral reefs that were damaged in 1983, particularly those on and near mainland Costa Rica and in Panama, have the potential to recover, although the period of recovery may require several decades. Some reefscape features, such as large massive colonies and continuous fields of branching corals, will probably disappear due to changed patterns of predation and bioerosion. Whether the remnant reef smctures in the Galapagos Islands and at Cocos Island will survive long enough to provide foundations for the development of new coral communities remains to be seen. If these structures disappear, they could be replaced by non-reef building benthic assemblages. If large populations of grazing sea urchins and fishes persist, it is possible that an alternate stable community dominated by filamentous microalgae and crustose coralline algae would develop. The smucture and trophic character of this type of community would be similar to the "urchin barrens" that occasionally assume prominence in temperate waters (Leighton, 1971; Lawrence, 1975; Mann, 1977). 5.4 g&
In light of the El Nifio sea warming disturbances to eastern Pacific coral reefs, some speculations may be offered relative to the expected responses of coral reefs to predicted global greenhouse warming. Regarding temperature changes, most attention has focused on the importance of cooling in causing marine mass extinctions throughout geologic history (Newell, 1971; Stanley, 1984). It is argued that during polar cooling crises tropical marine biotas suffer most and extinctions are most evident in geographic cul-de-sacs from which escape is impossible. A few workers have adduced evidence that temperature increases could have acted in combination
115
with other stressors (e.g., regression of continental seas) causing some past reef-coral extinction events (Emiliani et al., 1981; Kauffman and Johnson, 1988). Kauffman and Johnson (1988) have noted that coral and coralline algal dominated reefs were essentially absent from a narrow equatorial belt characterized by exceptionally warm and hypersaline conditions in the Late Cretaceous seas. This putative equatorial w m zone, the Supertethys, was occupied by rudistid molluscs with reef-building corals occurring at higher, cooler latitudes. These workers further surmised that the warm-tolerant rudistids replaced reef-building corals and calcareous algae, whose thermal and chemical tolerance limits were exceeded in the Supertethys climate zone. Further study and comparison of Cretaceousnertiary boundary marine extinction events with extant coral reef ecosystem responses to global warming may well be a profitable avenue of research. Climate modeling results based on C02 doubling simulations indicate that greenhouse warming could increase tropical SST by 1.5-4.0°C (Mitchell, 1988), which is comparable to the temperature increase observed during the 1982-83 El Niiio event. A six year (1982-88) world Ocean warming trend has been observed recently, at a O.l0C/yr increase (Strong, 1989). but it is not known if this is greenhouse related. An additional important consequence of global warming is the melting of polar ice and sea level rise. Rapid sea level rise is predicted to cause reef drowning and stress to reef corals through increased sedimentation, light attenuation and nutrient loading (Hallock and Schlager, 1986; Buddemeier and Smith, 1988; Graus and Macintyre, 1988; Hopley and Kinsey, 1988). If the frequency of coral bleaching and mortality increases in actively accreting, shallow reef zones, due to slight but protracted increases in sea water temperature, then this could further diminish the potential for reefs to keep pace with rising sea level. Moreover, fast-growing, branching corals that are often responsible for rapid vertical reef growth are also among those most susceptible to warm water stress. The 1982-83 El Nifio disturbance had a differential effect on eastern Pacific reef organisms with the highest mortalities suffered by reef-building corals, while corallivores and bioeroders (except for lithophagine bivalves) were little affected or showed increases in relative or absolute abundances. If this kind of response were to occur during periods of persistent sea warming, then surviving coral populations and framework structures would be subject to increasing levels of predation and bioerosion. The capacity for coral reef recovery would be progressively diminished with each bleaching event. Moreover, if temperature increases are sudden, i.e. occurring over one or a few centuries, then it is uncertain thbt selection in long-lived species could be effective in allowing for adaptive responses to the new conditions. The strength of selection, however, probably would be high in a rapidly warming environment and could promote rapid evolution given appropriate genetic variability among reef species with short generation times (Pitelka, 1988; Bradshaw and Hardwick, 1989). A final implication that might stem from incapacitated global reef development is the shutting down of a C a sink. If estimates that 50% of the Ca flux to the oceans is precipitated on coral reefs (Smith, 1978; Ohde, in press) are correct, it follows that 0.08 GT (gigatons, 109 tons) of C/yr are likewise fixed in reef carbonate. This is a relatively small quantity, representing only
116
about 1.5% of the estimated annual global industrial C02 flux*. Therefore, the demise of global reef development would have little effect on the world-wide C02 budget. 6 SUMMARY The widespread reef-coral bleaching and mortality observed in the tropical eastern Pacific region in 1983 was caused primarily by prolonged sea warming that accompanied the very strong 1982-83 El Niiio event. A near-decadal warming trend (1976-1988) over much of the tropical and subEopica1Pacific Ocean may have exacerbated this extraordinary ENS0 disturbance. The coral bleaching response, due to the loss of symbiotic zooxanthellae from gastrodermal cells, was highly correlated temporally and spatially with the extent of the sea warming. El Niiio simulation experiments duplicated the responses to sea warming observed naturally in Panama. Considering the areal extent, severity, and secondary disturbances, the 1982-83 El Niiio wanning event had a greater impact on eastern Pacific coral reefs than other known disturbances, such as intense upwelling, extreme tidal exposures, and predation events. Zooxanthellate corals, obligate crustacean symbiotes of corals, and a corallivorous gastropod suffered immediate high mortalities. The population sizes of several corals were markedly reduced, some corals experienced local extinctions, and two species may have disappeared from the eastern Pacific. The sudden declines in coral cover and coral species diversity were followed by secondary disturbances, such as predator access to former refuges, increases in sea urchinmediated bioerosion, and damselfish occupation of partially killed coral surfaces, that caused additional mortality or interfered with recruitment and reef recovery. The death of old massive corals and accreting reef frame blocks suggest that a disturbance of this magnitude has not affected many coral reefs in the eastern Pacific for at least 200 years and perhaps for as long as 300-400 years in severely affected areas such as the Galapagos Islands. In terms of live coral cover, recovery may occur on some of the least damaged reefs in Costa Rica and Panama over a period of decades. On severely damaged Galapagos and Cocos coral reefs, with little to no coral remaining, recovery may require centuries. If global warming causes repeated and/or protracted sea temperature increases comparable to the 1982-83 El Nitio disturbance, then we could expect severe mortality of zooxanthellate reefbuilding corals, and increases in the relative abundances of predators and bioeroders that would further increase coral mortality and bioerosion, leading to a rapid destruction of framework structures, and a diminished capacity for reef recovery in present-day coral seas. Surviving species may become established at higher latitudes where they could initiate reef growth under more favorable thermal conditions. 7 ACKNOWLEDGMENTS I appreciate the information and help on various topics offered by the following persons: sea temperatures - J. A. Brady, M. S . Hart, J. Espinosa and N. H. Vasquez (Panama Canal Commission); F. R. Miller (Inter-American Tropical Tuna Commission); W. H. Quinn (Oregon *These estimates are based on a Ca dissolved river flux of 495 x 1OI2 gm/yr (Martin and Meybeck, 1979) and an anthropogenic C flux of 5.1 GT C/yr (Rotty and Masters, 1985).
117
State University); rainfall - D. M. Windsor (Smithsonian Tropical Research Institute); seismic events - J. A. Rial (University of North Carolina); coral bleaching and mortality - H. von Prahl (Universidad del Valle, Colombia); bleaching of Tridacna - W. K. Fitt (University of Georgia); statistical analyses - N. M. Ehrhardt (Rosenstiel School of Marine and Atmospheric Science); S. S. Shapiro (Florida International University). This paper benefitted substantially from the criticisms offered by M. W. Colgan, J. Cortes, C. M. Eakin, J. S . Feingold, H. M. Guzman, I. G. Macintyre, M. Moore, L. C. Peterson, W. H. Quinn, M. L. Reaka, G. T. Shen, D. B. Smith, G. M. Wellington and two anonymous reviewers. I gratefully acknowledge the support of the Scholarly Studies Program, Smithsonian Institution (grant 1234s 104) and the U.S. National Science Foundation (grants OCE-8415615 and OCE-8716726). I am also indebted to the following host institutions and respective heads for their collaboration and sustained support of this study: Centro de Investigacion en Ciencias del Mar y Limnologia (M. M. Murillo), University of Costa Rica; Centro de Ciencias del Mar y Limnologia (L. DCroz), University of Panama; Estacion Cientifica Charles Darwin (G. Reck), Galapagos Islands, Ecuador. TAME, an Ecuadorean national airline, subsidized air travel to the Galapagos Islands. Contribution from the University of Miami, Rosenstiel School of Marine and Atmospheric Science.
8 REFERENCES Antonius, A., 1977. Coral mortality in reefs: a problem for science and management. Proc. 3rd Int. Coral Reef Symp., Miami, 2: 617-623. Atwood, D.K., Sylvester, J.C., Corredor, J.E., Morell, J.M., Mendez, A., Nodal, W.J., Huss, B.E. and Foltz, C., 1988. Sea surface temperature anomalies for the Caribbean, Gulf of Mexico, Florida reef tract and the Bahamas considered in light of the 1987 regional coral bleaching event. Proc. Assoc. Is. Mar. Lab. Carib., Sarasota, 21: 47 (abstract). Ayling, A.M. and Ayling, A.L., 1988. Coral destruction resulting from grazing activities of the muricid gastropod Drupella cornus on the Ningaloo Reef Tract in NW Australia. 6th Int. Coral Reef Symp., Townsville (1988), General Predation Session (oral presentation). Bak, R.P.M., 1978. Lethal and sublethal effects of dredging on reef corals. Mar. Poll. Bull., 9: 14-16. Barber, R.T. and Chavez F.P., 1983. Biological consequences of El Niiio. Science, 222: 1,2031,210. Barber, R.T. and Chavez, F.P., 1986. Ocean variability in relation to living resources during the 1982-83 El Niiio. Nature, 319: 279-285. Birkeland, C., 1977. The importance of rate of biomass accumulation in early successional stages of benthic communities to the survival of coral recruits. Proc. 3rd Int. Coral Reef Symp., Miami, 1: 15-21. Birkeland, C., 1982. Terrestrial runoff as a cause of outbreaks of Acanthaster Dlanci (Echinodermata: Asteroidea). Mar. Biol., 69: 175-185. Birkeland, C., 1987. Nutrient availability as a major determinant of differences among coastal hard-substratum communities in different regions of the tropics. In: C. Birkeland (Editor), Comparison Between Atlantic and Pacific Tropical Marine Coastal Ecosystems: Community Structure, Ecological Processes, and Productivity. UNESCO Rpts. Mar. Sci., 46: 45-97. Bothwell, A.M., 1981. Fragmentation, a means of asexual reproduction and dispersal in the coral genus Acropora (Scleractinia: Astrocoeniida: Acroporidae)--a preliminary report. Proc. 4th Int. Coral Reef Symp., Manila, 2: 137-144. Bradshaw, A.D. and Hardwick, K., 1989. Evolution and stress -- genotypic and phenotypic components. Biol. J. Linn. SOC.,37: 137-155. Brown, B.E., 1987. Worldwide death of corals--natural cyclical events or man-made pollution? Mar. Poll. Bull., 18: 9-13.
118
Brown, B.E. and Howard, L.S., 1985. Assessing the effects of "stress" on reef corals. Adv. Mar. Biol., 22: 1-63. Brusca, R.C. and Thomson, D.A., 1975. Pulmo Reef the only "coral reef' in the Gulf of California. Ciencias Marinas, 2(2): 37-53. Buddemeier, R.W. and Smith, S.V., 1988. Coral reef growth in an era of rapidly rising sea level: predictions and suggestions for long-term research. Coral Reefs, 7: 5 1-56. Bulletin International Seismological Centre, 1983. vol. 20, no. 4 (April). Cane, M.A., 1986. El Niiio. Annu. Rev. Earth Planet. Sci., 14: 43-70. Camquiry, J.D., Risk, M.J. and Schwarcz, H.P., 1988. Timing and temperature record from stable isotopes of the 1982-1983 El Niiio warming event in eastern Pacific corals. Palaios, 3: 359-364. Castro, P., 1978. Movements between coral colonies in Trapezia fermginea (Crustacea: Brachyura), an obligate symbiont of scleractinian corals. Mar. Biol., 46: 237-245. Causey, B.D., 1988. Observations of environmental conditions preceding the coral bleaching event of 1987 -- Looe Key National Marine Sanctuary. Proc. Assoc. Is. Mar. Lab. Carib., Sarasota, 21: 48 (abstract). Caviedes, C.N., 1984. El Niiio 1982-83. Geogr. Rev., 74: 267-290. Chave, K.E., Smith, S.V. and Roy, K.J., 1972. Carbonate production by coral reefs. Mar. Geol., 12: 123-140. Clausen, C., 1971. Effects of temperature on the rate of 45calcium uptake by pocillopora damicornis. In: H.M. Lenhoff, L. Muscatine and L.V. Davis (Editors), Experimental Coelenterate Biology, University of Hawaii Press, Honolulu, pp. 246-259. Clausen, C.D. and Roth, A.A., 1975. The effects of temperature and temperature adaptation on calcification rate in the hermatypic coral Pocillopora damicomis. Mar. Biol., 33: 93- 100. Cohen, D.M., 1985. Notes on a cruise in the Galapagos Islands - May 1984. Terra, 23: 21-24. Coles, S.L., 1975, A comparison of effects of elevated temperature versus temperature fluctuations on reef corals at Kahe Point, Oahu. Pac. Sci., 29: 15-18. Coles, S.L. and Jokiel, P.L., 1977. Effects of temperature on photosynthesis and respiration in hermatypic corals. Mar. Biol., 43: 209-216. Coles, S.L. and Jokiel, P.L., 1978. Synergistic effects of temperature, salinity and light on the hermatypic coral Montipora vermcosa. Mar. Biol., 49: 187-195. Coles, S.L., Jokiel, P.L. and Lewis, C.R., 1976. Thermal tolerance in tropical versus subtropical Pacific reef corals. Pac. Sci., 30: 159-166. Colgan, M.W., 1982. Succession and recovery of a coral reef after predation by Acanthaster (L.). Proc. 4th Int. Coral Reef Symp., Manila, 2: 333-338. Colgan, M.W., 1987. Coral reef recovery on Guam (Micronesia) after catastrophic predation by Acanthaster. Ecology, 68: 1,592-1,605. Colgan, M.W. and Malmquist, D.L., 1987. The Urvina Bay uplift: a dry trek through a Galapagos coral reef. Oceanus, 30: 61-68. Connell, J.H., 1973. Population ecology of reef-building corals. In: O.A. Jones and R. Endean (Editors), Biology and Geology of Coral Reefs. Vol. 2, Biol. 1, Academic Press, New York, pp. 205-245. Connell, J.H., 1978. Diversity in tropical rain forests and coral reefs. Science, 199: 1,302-1,310. Connell, J.H. and Slatyer, R.O., 1977. Mechanisms of succession in natural communities and their role in community stability and organization. Amer. Natur., 111: 1,119-1,144. Connell, J.H. and Keough, M.J., 1985. Disturbance and patch dynamics of subtidal marine animals on hard substrata. In: S.T.A. Pickett and P.S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, New York, pp. 125-151. Cortes, J. and Murillo, M. M., 1985. Comunidades coralinas y arrecifes del Pacifico de Costa Rica. Rev. Biol. Trop., 33: 197-202. Cortes, J., Murillo, M.M., Guzman, H.M. and Acufia, J., 1984. Perdida de zooxantelas y muerte de corales y otros organismos arrecifales en el Caribe y Pacifico de Costa Rica. Rev. Biol. Trop., 32: 227-231. Crossland, C., 1927. Marine ecology and coral formations in the Panama region, the Galapagos and Marquesas Islands, and the Atoll of Napuka. The expedition to the South Pacific of the S.Y. St. George. Trans. Roy. Soc. Edinburgh, 55: 531-554. Cruz, E., 1985. Efectos de El Niiio en la Isla Floreana. In: G. Robinson and E. M. del Pino (Editors), El Niiio en las Islas Galapagos: El Evento de 1982-1983. Fundacion Charles Darwin para Las Mas Galapagos, Quito, Ecuador, pp. 497-504.
119
Dana, J.D., 1843. On the temperature limiting the distribution of corals. Am. J. Sci. & Arts, 45: 130-131. Dana, T.F., 1975. Development of contemporary eastern Pacific coral reefs. Mar. Biol., 33: 355374. Done, T.J., 1985. Effects of two Acanthaster outbreaks on coral community structure -- the meaning of devastation. Proc. 5th Int. Coral Reef Congr., Tahiti, 5: 315-320. Druffel, E.R.M., 1985. Detection of El Nifio and decade time scale variations of sea surface temperature from banded coral records: implications for the carbon dioxide cycle. The Carbon Cycle and Atmospheric C02: Natural Variations, Archean to Present. Geophys. Monogr., 32: 111-122. Dunbar, R.B., Wellington, G.M., Colgan, M. and Glynn, P.W., 1987. Galapagos massive corals: an annual record of eastern tropical Pacific climate during the past 400 years. EOS, Trans. Am. Geophys. Union, 68: 1,743 (abstract). Durham, J.W. and Barnard, J.L., 1952. Stony corals of the eastern Pacific collected by the Velero III and Velero IV. Allan Hancock Pac. Exped., 16: 1-110. Eakin, C.M., 1987. Damselfishes and their algal lawns: a case of plural mutualism. Symbiosis, 4: 275-288. Eakin, C.M., In press. Avoidance of damselfish lawns by the sea urchin Diadema mexicanum at Uva Island, Panama. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Eakin, C.M., Smith, D.B., Glynn, P.W., D'Croz, L. and Gil, J., 1989. Extreme tidal exposures, cool upwelling, and coral mortality in the eastern Pacific (Panama). Proc. Assoc. Is. Mar. Lab. Carib., Puerto Rico, 22: 29 (abstract). Easton, W. H. and Olson, E. A., 1976. Radiocarbon profile of Hanauma Reef, Oahu, Hawaii. Geol. SOC. Am. Bull., 87: 711-719. Edmondson, C.H., 1928. The ecology of an Hawaiian coral reef. Bull. Bernice P. Bishop Mus., 45: 1-64. Egatia, A.C. and DiSalvo, L.H., 1982. Mass expulsion of zooxanthellae by Easter Island corals. Pac. Sci., 36: 61-63. Emiliani, C., b u s , E.B. and Shoemaker, E.M., 1981. Sudden death at the end of the Mesozoic. Earth Planet. Sci. Lett., 55: 317-334. Endean, R., 1973. Population explosions of Acanthaster olanci and associated destruction of hermatypic corals in the Indo-West Pacific region. In: 0. A. Jones and R. Endean (Editors), Biology and Geology of Coral Reefs, Vol. 2, Biol. 1, Academic Press, pp. 389-438. Endean, R. 1976. Destruction and recovery of coral reef communities. In: O.A. Jones and R. Endean (Editors), Biology and Geology of Coral Reefs, Academic Press, London, 3: 215-254. Enfield, D.B., 1981. El Nifio: Pacific eastern boundary response to interannual forcing. In: M.H. Glantz and J.D. Thompson (Editors), Resource Management and Environmental Uncertainty: Lessons from Coastal Upwelling Fisheries, Wiley, New York, 11: 213-254. Estacion, J.S. and Braley, R.D., 1988. Growth and survival of Tridacna -juveniles in an intertidal pond. In: J.W. Copland and J.S. Lucas (Editors), Giant Clams in Asia and the Pacific. ACIAR (Aust. Cent. Int. Agric. Res.) Monogr., 9: 191-192. Fankboner, P.V. and Reid, R.G.B., 1981. Mass expulsion of zooxanthellae by heat-stressed reef corals: a source of food for giant clams? Experientia, 37: 251-252. Feldman, G., Clark, D. and Halpern, D., 1984. Satellite color observations of the phytoplankton distribution in the eastern equatorial Pacific during the 1982-1983 El Nitio. Science, 226: 1,069-1,07 1. Fisk, D.A. and Done, T.J., 1985. Taxonomic and bathymetric patterns of bleaching in corals, Myrmidon Reef (Queensland). Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 149-154. Forsbergh, E.D., 1969. On the climatology, oceanography and fisheries of the Panama Bight. Bull. Inter-Amer. Trop. Tuna Comm., 14: 49-385. Gilmore, M.D. and Hall, B.R., 1976. Life-history, growth habits, and constructional roles of Acropora Cervicornisin the patch reef environment. J. Sed. Petrology, 46: 5 19-522. Glynn, P.W., 1972. Observations on the ecology of the Caribbean and Pacific coasts of Panama. In: M.L. Jones (Editor), The Panamic Biota: Some Observations Prior to a Sea Level Canal. Bull. Biol. SOC.Wash., 2: 13-30. Glynn, P.W., 1973. AcanthasteI: effect on coral reef growth in Panama. Science, 180: 504-506. Glynn, P.W., 1974. The impact of Acanthastec on corals and coral reefs in the eastern Pacific. Environ. Conserv., 1: 295-304. Glynn, P.W., 1976. Some physical and biological determinants of coral community structure in the eastern Pacific. Ecol. Monogr., 46: 431-456.
120
Glynn, P.W., 1977. Coral growth in upwelling and nonupwelling areas off the Pacific coast of Panama. J. Mar. Res., 35: 567-585. Glynn, P.W., 1982. Acanthaster population regulation by a shrimp and a worm. Proc.4th Int. Coral Reef Symp., Manila, 2: 607-612. Glynn, P.W., 1983. Extensive 'bleaching' and death of reef corals on the Pacific coast of Panama. Environ. Conserv., 10: 149-154. Glynn, P.W., 1984a. Widespread coral mortality and the 1982/83 El Niiio warming event. Environ. Conserv., 11: 133-146. Glynn, P.W., 1984b. An amphinomid worm predator of the crown-of-thorns sea star and general predation on asteroids in eastern and western Pacific coral reefs. Bull. Mar. Sci., 35: 54-71. Glynn, P.W., 1985a. Corallivore population sizes and feeding effects following El Niiio (198283) associated coral mortality in Panama. Proc.5th Int. Coral Reef Congr., Tahiti, 4: 183-188. G l y ~P.W., , 1985b. El Niiio-associated disturbance to coral reefs and post disturbance mortality by Acanthaster planci. Mar. Ecol. Prog. Ser., 26: 295-300. Glynn, P.W., 1988a. El Niiio-Southern Oscillation 1982-1983: nearshore population, community, and ecosystem responses. Annu. Rev. Ecol. Syst., 19: 309-345. Glynn, P.W., 1988b. Coral bleaching and mortality in the tropical eastern Pacific during the 198283 El Niiio warming event. In: J. Ogden and R. Wicklund (Editors), Mass Bleaching of Coral Reefs in the Caribbean: A Research Smategy. NOAA's Undersea Res. Prog., St. Croix, U.S. Virgin Islands, Res. Rpt. 88-2: 42-45. Glynn, P.W., 1988c. El Niiio warming, coral mortality and reef framework destruction by echinoid bioerosion in the eastern Pacific. Galaxea, 7: 129-160. Glynn, P.W. and Colgan, M.W., In press. Defense of corals and enhancement of coral diversity by territorial damselfishes. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Glynn, P.W. and Macintyre, I.G., 1977. Growth rate and age of coral reefs on the Pacific coast of Panama. Proc. 3rd Int. Coral Reef Symp., Miami, 2: 251-259. Glynn, P.W. and Stewart, R.H., 1973. Distribution of coral reefs in the Pearl Islands (Gulf of Panama) in relation to thermal conditions. Limnol. Oceanogr., 18: 367-379. Glynn, P.W. and Wellington, G.M., 1983. Corals and Coral Reefs of the Galapagos Islands. University of California Press, Berkeley, 330 pp. Glynn, P.W. and D'Croz, L., In press. Experimental evidence for high temperature stress as the cause of El Niiio-coincidentcoral mortality. Coral Reefs. Glynn, P.W., Druffel, E.M. and Dunbar, R.B., 1983. A dead Central American coral reef tract: possible link with the Little Ice Age. J. Mar. Res., 41: 605-637. Glynn, P.W., Perez, M. and Gilchrist, S.L., 1985a. Lipid decline in stressed corals and their crustacean symbionts. Biol. Bull., 168: 276-284. Glynn, P.W., Peters, E.C. and Muscatine, L., 1985b. Coral tissue microstructure and necrosis: relation to catastrophiccoral mortality in Panama. Dis. Aquat. Org., 1: 29-37. Glynn, P.W., Prahl, H. von and Guhl, F., 1982. Coral reefs of Gorgona Island, Colombia, with special reference to corallivores and their influence on community structure and reef development. An. Inst. Inv. Mar. Punta de Betin, 12: 185-214. Glynn, P.W., Stewart, R.H. and McCosker, J.E., 1972. Pacific coral reefs of Panama: structure, distribution and predators. Geol. Rundschau, 61: 483-519. Glynn, P.W., Cortes, J., Guzman, H.M. and Richmond, R.H., In press. El Niiio (1982-83) associated coral mortality and relationship to sea surface temperature deviations in the tropical eastern Pacific. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Goenaga, C., Vicente, V. and Armstrong, R., 1988. Aposymbiosis in Puerto Rican zooxanthellate cnidarians. Proc. Assoc. Is. Mar. Lab. Carib., Sarasota, 21: 49 (abstract). Goreau, T.F., 1964. Mass expulsion of zooxanthellae from Jamaican reef communities after Hurricane Flora. Science, 145: 383-386. Graus, R.R. and Macintyre, I.G.,1988. Model predictions of the effects of future sea level rise on the growth of Caribbean coral reefs. Proc. 6th Int. Coral Reef Symp., Townsville (abstract). Grigg, R.W., 1982. Darwin Point: a threshold for atoll formation. Coral Reefs, 1: 29-34. Grigg, R.W. and Maragos, J.E., 1974. Recolonization of hermatypic coral on submerged lava flows in Hawaii. Ecology, 55: 387-395. Grove, J.M., 1988. The Little Ice Age. Methuen, London, 498 pp. Groves, J.S., 1985. The effects of the 1982-83El Niiio upon the ichthyofauna of the Galapagos Archipelago. In: G. Robinson and E. M. del Pino (Editors), El Niiio in the Galapagos Islands: The 1982-1983Event. The Charles Darwin Foundation for the Galapagos, Quito, Ecuador, pp. 191-198.
121
Guzman, H.M., 1988a. Distribucion y abundancia de organismos coralivoros en 10s arrecifes coralinos de la Isla del Caiio, Costa Rica. Rev. Biol. Trop., 36: 191-207. Guzman, H.M., 1988b. Feeding activity of the corallivorous gastropod aonodonta (Blainville). Rev. Biol. Trop., 36: 209-212. Guzman, H.M. and Cortes, J., 1989. Coral reef community structure at Caiio Island, Pacific Costa Rica. P.S.Z.N.1: Mar. Ecol., 10: 23-41. Guzman, H.M. and Robertson, R., 1989. Population and feeding responses of the corallivorous pufferfish melearnis to coral mortality in the eastern Pacific. Mar. Ecol. Prog. Ser., 5 5 : 121-131. Guzman, H.M., Cortes, J., Richmond, R.H. and Glynn, P.W., 1987. Efectos del fenomeno de El Niiio-Oscilacion Sureiia 1982-83 en 10s arrecifes cqralinos de la Isla del Caiio, Costa Rica. Rev. Biol. Trop., 35: 325-332. Guzman, H. M., Cortes, J., Glynn, P. W. and Richmond, R. H., In press. Coral mortality associated with dinoflagellate blooms in the eastern Pacific (Costa Rica and Panama). Mar. Ecol. Prog. Ser. Hallock, P. and Schlager, W., 1986. Nutrient excess and the demise of coral reefs and carbonate platforms. Palaios, 1: 389-398. Harriott, V.H., 1985. Mortality rates of scleractinian corals before and during a mass bleaching event. Mar. Ecol. h o g . Ser., 21: 81-88. Hatcher, B.G., 1983. Grazing in coral reef ecosystems. In: D.J. Barnes (Editor), Perspectives on Coral Reefs. Australian Institute of Marine Science, Townsville, pp. 164-179. Hay, M.E., 1984. Patterns of fish and urchin grazing on Caribbean coral reefs: are previous results typical? Ecology, 65: 446-454. Hertig, M., Tallaferro, W.H. and Schwartz, B., 1937. The terms symbiosis, symbiont and symbiotic. J. Parasitol., 23: 326-329. Highsmith, R.C., 1982. Reproduction by fragmentation in corals. Mar. Ecol. Prog. Ser., 7: 207226. Hollings, E.F. (chairman), 1988. Bleaching of coral reefs in the Caribbean. Oral and written testimony to the Commerce, Justice, State, Judiciary, and related agencies, Appropriations Subcommittee, USA Senate, 10 Nov. 1987, 142 pp. Hopley, D. and Kinsey, D.W., 1988. The effects of a rapid short-term sea-level rise on the Great Barrier Reef. In: (3.1. Pearman (Editor), Greenhouse, planning for climate change. E.J. Brill, Leiden, pp. 189-201. Houck, J.E., Buddemeier, R.W., Smith, S.V. and Jokiel, P.L., 1977. The response of coral growth rate and skeletal strontium content to light intensity and water temperature. Proc. 3rd Int. Coral Reef Symp., Miami, 2: 425-431. Houvenaghel, G.T., 1978. Oceanographic conditions in the Galapagos Archipelago and their relationships with life on the islands. In: R. Boje and M. Tomczak (Editors), Upwelling Ecosystems. Springer-Verlag, Berlin, pp. 18 1-200. Houvenaghel, G.T., 1984. Oceanographic setting of the Galapagos Islands. In: R. Perry (Editor), Key Environments, Galapagos, Pergamon, Oxford, pp. 43-54. Hubbard, D.K., 1985. What do we mean by reef growth? Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 433-438. Hubbs, C.L. and Roden, G.I., 1964. Oceanography and marine life along the Pacific coast of Middle America. In: R.C. West (Editor), Natural Environment and Early Cultures, 1, Handbook of Middle America Indians. Univ. Texas Press, Austin, pp. 143-186. Hughes, T.P. and Jackson, J.B.C., 1980. Do corals lie about their age? Some demographic consequences of partial mortality, fission, and fusion. Science, 209: 7 13-715. Hughes, T.P. and Jackson, J.B.C., 1985. Population dynamics and life histories of foliaceous corals. Ecol. Monogr., 55: 141-166. Hughes, T.P., Reed, D.C. and Boyle, M-J., 1987. Herbivory on coral reefs: community structure following mass mortalities of sea urchins. J. Exp. Mar. Biol. Ecol., 113: 39-59. Hughes, T.P., Keller, B.D., Jackson, J.B.C. and Boyle, M.J., 1985. Mass mortality of the echinoid Diadema antillarurq Philippi in Jamaica. Bull. Mar. Sci., 36: 377-384. Hutcheson, K., 1970. A test for comparing diversities based on the Shannon formula. J. Theor. Biol., 29: 151-154. Jaap, W.C., 1979. Observations on zooxanthellae expulsion at Middle Sambo Reef, Florida Keys. Bull. Mar. Sci., 29: 414-422. Jaap, W.C., 1988. The 1987 zooxanthellae expulsion event at Florida reefs. In: J. Ogden and R. Wicklund (Editors), Mass Bleaching of Coral Reefs in the Caribbean: A Research Strategy. NOAA's Undersea Res. Prog., St. Croix, U.S. Virgin Islands, Res. Rpt. 88-2: 24-29.
122
Jackson, J.B.C., 1983. Biological determinants of present and past sessile animal distributions. In: M.J.S. Tevesz and P.L. McCall (Editors), Biotic Interactions in Recent and Fossil Benthic Communities. Plenum Pub. Corp., New York, pp. 39-120. Jackson, J.B.C. and Hughes, T.P., 1985. Adaptive strategies of coral-reef invertebrates. Amer. Sci., 73: 265-274. Jeffrey, S.W. and Haxo, F.T., 1968. Photosynthetic pigments of symbiotic dinoflagellates (zooxanthellae)from corals and clams. Biol. Bull., 135: 149-165. Johannes, R.E., 1975. Pollution and degradation of coral reef communities. In: E.J. Ferguson Wood and R.E. Johannes (Editors). Tropical Marine Pollution. Elsevier, Amsterdam, pp. 1351. Jokiel, P.L., 1980. Solar ultraviolet radiation and coral reef epifauna. Science, 207: 1,069-1,071. Jokiel, P.L. and Coles, S.L., 1974. Effects of heated effluent on hermatypic corals at Kahe Point, Oahu. Pac. Sci., 28: 1-18. Jokiel, P.L. and Coles, S.L., 1977. Effects of temperature on the mortality and growth of Hawaiian reef corals. Mar. Biol., 43: 201-208. Jokiel, P.L. and Guinther, E.B., 1978. Effects of temperature on reproduction in the hermatypic coral Pocilloaora damicornis. Bull. Mar. Sci., 28: 786-789. Jokiel, P.L. and York, R.H., Jr., 1982. Solar ultraviolet photobiology of the reef coral PodloDora damicornis and symbiotic zooxanthellae. Bull. Mar. Sci., 32: 301-315. Kamezaki, N. and Ui, S., 1984. Bleaching of hermatypic corals in Yaeyama Islands. Mar. Parks I., 61: 10-13. Kauffman, E.G. and Johnson, C.C., 1988. The morphological and ecological evolution of Middle and Upper Cretaceous reef-building rudistids. Palaios, 3: 194-216. Ken, R.A., 1988. La Niiia’s big chill replaces El Niiio. Science, 241: 1,037-1,038. Kinsey, D. W., 1983. Standards of performance in coral reef primary production and carbon turnover. In: D. J. Barnes (Editor), Perspectives on Coral Reefs. Australian Institute of Marine Science, Townsville, pp. 209-220. Kinsman, D.J.J., 1964. Reef coral tolerance of high temperatures and salinities. Nature, 202: 1,280-1,282. Knowlton, N., Lang, J.C., Rooney, M.C. and Clifford, P., 1981. Evidence for delayed mortality in hurricane-damaged Jamaican staghorn corals. Nature, 294: 251-252. Knowlton, N., Lang, J.C. and Keller, B.D., In press. Case study of natural population collapse: post-humcane predation on Jamaican staghorn corals. Smith. Contrib. Mar. Sci. Knudsen, J.W., 1967. Trapezia and Tetralia (Decapoda,Brachyura, Xanthidae) as obligate ectoparasites of pocilloporid and acroporid corals. Pacific Sci., 21: 51-57. Kogelschatz, J., Solorzano, L., Barber, R. and Mendoza, P., 1985. Oceanographic conditions in the Galapagos Islands during the 1982/83 El Niiio. In: G . Robinson and E. M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983Event. Fundacion Charles Darwin para las Islas Galapagos, Quito, Ecuador, pp. 91-123. Krebs, J.R., 1978. Optimal foraging: decision rules for predators. In: J.R. Krebs and N.B. Davies (Editors), Behavioural Ecology, An Evolutionary Approach. Blackwell, Oxford, pp. 23-63. Kwiecinski, B. and Chial, B., 1987. Manifestations of El Niiio in the Gulf of Panama. Trop. Ocean-Atmos. Newsl., 42: 7-9. Kwiecinski, B., Chial, B. and Torres, A., 1988. El Niiio and post El Niiio (1982-1986) in the Gulf of Panama. Trop. Ocean. Atmos. Newsl., 44: 7-8. Lang, J.C., 1988. Apparent differences in bleaching responses by zooxanthellatecnidarians on Colombian and Bahamian reefs. In: J. Ogden and R. Wicklund (Editors), Mass Bleaching of Coral Reefs in the Caribbean: A Research Strategy. NOAA’s Undersea Res. Prog., St. Croix, U.S. Virgin Islands, Res. Rpt. 88-2: 30-32. Lasker, H.R., Peters, E.C. and Coffroth, M.A., 1984. Bleaching of reef coelenterates in the San Blas Islands, Panama. Coral Reefs, 3: 183-190. Lawrence, J.M., 1975. On the relationship between marine plants and sea urchins. Oceanogr. Mar. Biol. Annu. Rev., 13: 213-286. Leighton, D.L., 1971. Grazing activities of benthic invertebrates in southern California kelp beds. In: W.J. North (Editor), The Biology of Giant Kelp Beds (b’iacrocvsh‘s) in California. Beih. Nova Hedgwigia, 32: 1-96. Lessios, H.A., Glynn, P.W. and Robertson, D.R., 1983. Mass mortalities of coral reef organisms. Science, 222: 715. Lessios, H.A., Robertson, D.R. and Cubit, J.D., 1984. Spread of Diadems mass mortality through the Caribbean. Science, 226: 335-337.
123
Lewis, S.M., 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecol. Monogr., 56: 183-200. Littler, M.M. and Littler, D.S., 1984. Models of tropical reef biogenesis: the contribution of algae. In: F.E. Round and D.J. Chapman (Editors), Progress in Phycological Research, 3: 323-364. Longhurst, A.R. and Pauly, D., 1987. Ecology of Tropical Oceans. Academic Press, New York, 407 pp. Loya, Y., 1976. Recolonization of Red Sea corals affected by natural catastrophes and man-made perturbations. Ecology, 57: 278-289. Lucas, R., Hayes, S.P. and Wyrtki, K., 1984. Equatorial sea level response during the 1982-83 El Nifio. J. Geophys. Res., 89(C6): 10,425-10.430. Madsen, F.J., 1955. A note on the sea-star genus Acanthaster. Vidensk. Meddr. Dansk Naturh. Foren., 117: 179-192. Mann, K.H., 1977. Destruction of kelp beds by sea urchins. A cyclical phenomenon, or irreversible degradation? Helgol. Wiss. Meeresunters., 30: 455-467. Maragos, J.E., Evans, C. and Holthus, P., 1985. Reef corals in Kaneohe Bay six years before and after termination of sewage discharges (Oahu, Hawaiian Archipelago). Proc. 5th Int. Coral Reef Congr., Tahiti, 4: 189-194. Marszalek, D.S., 1981. Impact of dredging on a subtropical reef community, southeast Florida, U.S.A. Proc. 4th Int. Coral Reef Symp., Manila, 1: 147-153. Martin, J.M. and Meybeck, M., 1979. Elemental mass-balance of material carried by world rivers. Mar. Chem., 7: 173-206. Maxwell, D., 1974. Marine primary productivity of the Galapagos Archipelago. Ph.D. dissertation, Ohio State Univ., 167 pp. Mayer, A.G., 1917. Is death from high temperatures due to accumulation of acid in the tissues? Proc. Nat. Acad. Sci., Wash., 3: 626-627. Mayer, A.G., 1918a. Ecology of the Murray Island coral reef. Carnegie Inst. Wash. Pub., 213: 348. Mayer, A.G., 1918b. Toxic effects due to high temperature. Carnegie Inst. Wash. Pub., 252: 172178. Mayor, A.G., 1924. Structure and ecology of Samoan reefs. Carnegie Inst. Wash. Pub., 340: 125. Mitchell, J.F.B., 1988. Local effects of greenhouse gases. Nature, 332: 399-400. Moran, P.J., 1986. The Acanthaster phenomenon. Oceanogr. Mar. Biol. Annu. Rev., 24: 379480. Morrison, D., 1988. Comparing fish and urchin grazing in shallow and deeper coral reef algal communities. Ecol., 69: 1,367-1,382. Moyer, J.T., Emerson, W.K. and Ross, M., 1982. Massive destruction of scleractinian corals by the muricid gastropod, Drupella, in Japan and the Philippines. The Nautilus, 96: 69-82. Newell, N.D., 1971. An outline history of tropical organic reefs. Am. Mus. Novit., 2,465: 1-37. Nishihira, M. and Yamazato, K., 1974. Human interference with the coral reef community and Acanthaster infestation of Okinawa. Proc. 2nd Int. Coral Reef Symp., 1: 577-590. Ogden, J. and Wicklund, R. (Editors), 1988. Mass bleaching of coral reefs in the Caribbean: a research strategy. NOAA's Undersea Res. Prog., Res. Rpt., 88-2, 51 pp. Ohde, S., In press. Calcium carbonate production in Okinawan coral reefs. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Oliver, J., 1985. Recurrent seasonal bleaching and mortality of corals on the Great Barrier Reef. Proc. 5th Int. Coral Reef Congr., Tahiti, 4: 201-206. Patton, W.K., 1976. Animal associates of living reef corals. In: O.A. Jones and R. Endean (Editors), Biology and Geology of Coral Reefs. Vol. 3, Biol. 2, Academic Press, New York, pp. 1-36. Pearson, R.G., 1981. Recovery and recolonization of coral reefs. Mar. Ecol. Prog. Ser., 4: 105122. Philander, S.G.H., 1983. El Nifio southern oscillation phenomena. Nature, 302: 295-301. Philander, S.G.H., 1985. El Nifio and La Nifia. J. Atmos. Sci., 42: 2,652-2,662. Phillips, D.J.H., 1987. Acanthaster &&: a need to balance the books better. Mar. Poll. Bull., 18: 615-617. Pielou, E.C., 1969. An Introduction to Mathematical Ecology. Wiley-Interscience,New York, 286 pp. Pielou, E.C., 1975. Ecological Diversity. Wiley-Interscience,New York, 165 pp. Pitelka, L.F., 1988. Evolutionary responses of plants to anthropogenic pollutants. Trends Ecol. Evo~.,3: 233-236.
124 Porter, J.W., 1974. Community structure of coral reefs on opposite sides of the Isthmus of Panama. Science, 186: 543-545. Potts, D.C., 1982. Crown-of-thorns starfish -- man-induced pest or natural phenomenon? In: R.L. Kitching and R.E. Jones (Editors), The Ecology of Pests, Some Australian Case Histories. CSIRO, Canberra, Australia, pp. 55-86. Prahl, H. von, 1983. Blanqueo masivo y muerte de corales en la Isla de Gorgona, Pacifico Colombiano. Cespedesia, 12: 125-129. Prahl, H. von, 1985. Blanqueo masivo y muerte de corales hermatipicos en el Pacifico Colombiano atribuidos a1 fenomeno de El Niiio 1982-83. Bol. Erfen, 12: 22-24. Prahl, H. von, 1986. Crecimiento del coral Pocillopora damicomis durante y despues del fenomeno El Niiio 1982-1983 en la Isla de Gorgona, Colombia. Bol. Erfen, 18: 11-13. Prahl, H. von and Mejia, A., 1985. Primer informe de un coral acroporido, Acropora v a l i h (Dana, 1846) (Scleractinia: Astrocoeniida: Acroporidae) para el Pacifico americano. Rev. Biol. Trop., 33: 39-43. Quinn, W.H., 1974. Monitoring and predicting El Niiio invasions. J. Appl. Meteorol., 13: 825830. Quinn, W.H., 1976. The 1976 El Niiio and recent progress in monitoring and prediction. In: J.R. Goulet, Jr. and E.D.Haynes (Editors), Ocean Variability in the U S . Fishery Conservation Zone. NOAA Tech. Rep. NMFS Circ. 427, June 1979, pp. 93-110. Quinn, W. H., 1977. Use of Southern Oscillation index anomalies for predicting El Niiio type activity. Ciencias Marinas, 4: 40-66. Quinn, W.H. and Neal, V.T., 1983. Long-term variations in the Southern Oscillation, El Niiio, and Chilean subtropical rainfall. Fish. Bull., 81: 363-374. Quinn, W. H. and Neal, V. T., 1984. Recent long-term climatic change over the eastem tropical Pacific and its ramifications. In: Proc. 9th Annu. Clim. Diagn. Workshop. U. S. Dep. Commer., Corvallis, Oreg., pp. 101-109. (Available as NTIS-PB85-183911, from Natl. Tech. Int. Serv., Springfield, Va.) Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Niiio occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14,449-14,461. Randall, R.H., 1973. Distribution of corals after Acanthaster planci (L.) infestation at Tanguisson Point, Guam. Micronesica, 9: 213-222. Rasmusson, E.M., 1984. El Niiio: the ocean/atniosphere connection. Oceanus, 27: 5-12. Renner, J.A. 1963. Sea surface temperature monthly average and anomaly charts eastern tropical Pacific Ocean, 1947-58. U.S. Dept. Interior, Fish & Wildlife Serv., Spec. Sci. Rpt.-Fish. 442: 1-57. Reyes Bonilla, H. 1988. First report of massive loss of zooxanthellae by scleractinian corals in the Gulf of California. Proc. Assoc. Is. Mar. Lab. Carib., 21: 67 (abstract). Reynolds, R.W. and Gemmill, W.H., 1984. An objective global monthly mean sea surface temperature analysis. Trop. Ocean-Atmos. Newsl., 23: 4-5. Richmond, R.H., 1985. Variations in the population biology of Pocillopora damicornis across the Pacific. Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 101-106. Richmond, R.H., 1987. Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coral PocilloDora damicornix. Bull. Mar. Sci., 41: 594604. Robalino, M., 1985. Registros meteorologicos de la estacion cientifica Charles Darwin para 19821983. In: G. Robinson and E. M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983 Event. Fundacion Charles Darwin para las Islas Galapagos, Quito, Ecuador, pp. 83-90. Robinson, G., 1985. The influence of the 1982-83 El Niiio on Galapagos marine life. In: G. Robinson and E. M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983 Event. Fundacion Charles Darwin para las Islas Galapagos, Quito, Ecuador, pp. 153-190. Robinson, G. and del Pino, E. M. (Editors), 1985. El Niiio en Las Islas Galapagos: El Evento de 1982-1983. Fundacion Charles Darwin para Las Islas Galapagos, Quito, Ecuador, 534 pp. Rogers, C.S., 1985. Degradation of Caribbean and western Atlantic coral reefs and decline of associated fisheries. Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 491-496. Rosen, B.R., 1981. The tropical high diversity enigma -- the corals' eye view. In: P.H. Greenwood and P.L. Forey (Editors), Chance, Change and Challenge: the Evolving Biosphere. British Museum (Natural History) and Cambridge University Press, Cambridge, pp. 103-129. Roth, A.A., Clausen, C.D., Yahiku, P.Y., Clausen, V.E. and Cox, W.W., 1982. Some effects of light on coral growth. Pac. Sci., 36: 65-81.
125
Rotty, R.M. and Masters, C.D., 1985. Carbon dioxide from fossil fuel combustion: trends, resources, and technological implications. In: J.R. Trabalka (Editor), Atmospheric Carbon Dioxide and the Global Carbon Cycle. U.S. Department of Energy, DOEER-0239, Washington, D.C. pp. 63-80. Rutzler, K., Santavy, D.L., Antonius, A., 1983. The black band disease of Atlantic reef corals. 111. Distribution, ecology, and development. P.S.Z.N.1: Mar. Ecol., 4: 329-358. Salvat, B. (editor), 1987. Human impacts on coral reefs: facts and recommendations. Antenne Museum E.P.H.E., French Polynesia, 253 pp. Sammarco, P.W., 1980. Piadema and its relationship to coral spat mortality: grazing, competition, and biological disturbance. J. Exp. Mar. Biol. Ecol., 45: 245-272. Sammarco, P.W., 1982. Echinoid grazing as a structuring force in coral communities: whole reef manipulations. J. Exp. Mar. Biol. Ecol., 61: 31-55. Sammarco, P.W., Carleton, J.H. and Risk, M.J., 1986. Effects of grazing and damselfish temtoriality on bioerosion of dead corals: direct effects. J. Exp. Mar. Biol. Ecol., 98: 1-19. Sandeman, I.M., 1988a. Zooxanthellae physiology and a suggested irradiance/temperature mechanism for bleaching. Proc. Assoc. Is. Mar. Lab. Carib., Sarasota, 21: 50 (abstract). Sandeman, I.M., 1988b. Coral bleaching at Discovery Bay, Jamaica: a possible mechanism for temperature-related bleaching. In: J. Ogden and R. Wicklund (Editors), Mass Bleaching of Coral Reefs in the Caribbean: A Research Strategy. NOAA's Undersea Res. hog., St. Croix, U.S. Virgin Islands, Res. Rpt. 88-2: 46-48. Schuhmacher, H., 1974. On the conditions accompanying the first settlement of corals on artificial reefs with special reference to the influence of grazing sea urchins (Eilat, Red Sea). Roc. 2nd Int. Coral Reef Symp., Brisbane, 1: 257-267. Scott, P.J.B., Risk, M.J. and Caniquiry, J., In press. El Niiio, bioerosion and the survival of east Pacific reefs. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Shinn, E.A. 1966. Coral growth-rate, an environmental indicator. J. Palaeont., 40: 233-240. Siebeck, D., 1981. Photoreactivation and depth-dependent UV tolerance in reef coral in the Great Barrier Reef/Australia. Naturwissenschaften, 68: 426-428. Siebeck, D., 1988. Experimental investigation of UV tolerance in hermatypic corals (Scleractinia). Mar. Ecol. Prog. Ser., 43: 95-103. Smith, S.V., 1978. Coral-reef area and the contributions of reefs to processes and resources of the world's oceans. Nature, 273: 225-226. Squires, D.F., 1959. Results of the Puritan-American Museum of Natural History Expedition to western Mexico, 7. Corals and coral reefs in the Gulf of California. Bull. Am. Mus. Nat. Hist., 118: 367-432. Sousa, W.P., 1984. The role of disturbance in natural communities. Annu. Rev. Ecol. Syst., 15: 353-391. Stanley, S.M., 1984. Marine mass extinctions: a dominant role for temperatures. In: M.H. Nitecki (Editor), Extinctions. University of Chicago Press, Chicago, pp. 69-1 17. Steen, R.G. and Muscatine, L., 1987. Low temperature evokes rapid exocytosis of symbiotic algae by a sea anemone. Biol. Bull., 172: 246-263. Stehli, F.G. and Wells, J.W., 1971. Diversity and age patterns in hermatypic corals. Syst. Zool., 20: 115-126. Stoddart, D.R., 1969a. Ecology and morphology of recent coral reefs. Biol. Rev. Cambridge Phil. SOC.,44: 433-498. Stoddart, D.R., 196913. Geomorphology of the Solomon Islands coral reefs. Phil. Trans. R. SOC. B, 255: 355-382. Stoddart, D.R., 1972. Catastrophic damage to coral reef communities by earthquake. Nature, 239: _51-52. _ __
Strong, A.E., 1983. Satellite-derived sea surface temperature errors due to the El Chichon aerosol cloud. Trop. Ocean-Atmos. Newsl., 18: 14-15. Strong, A.E., 1989. Greater global warming revealed by satellite-derived sea-surface-temperature trends. Nature, 338: 642-645. Taylor, F.W., Frohlich, C., LeColle, J. and Strecker, M., 1987. Analysis of partially emerged corals and reef terraces in the central Vanuatu arc: comparison of contemporary coseismic and nonseismic with Quaternary vertical movements. J. Geophys. Res., 92: 4,905-4,933. Tunnicliffe, V.J., 1981. Breakage and propagation of the stony coral Acroporp cervicornis. Proc. Nat. Acad. Sci. (USA), 78: 2,427-2,431. UNEPWCN, 1988. Coral Reefs of the World. Volumes 1-3, S.M. Wells and M.D. Jenkins (Editors), UNEP Regional Seas Directories and Bibliographies. IUCN, Gland, Switzerland and Cambridge, U.K./UNEP, Nairobi, Kenya.
126
Veron, J.E.N., 1985. Aspects of the biogeography of hermatypic corals. Proc. 5th Int. Coral Reef Congr., Tahiti, 4: 83-88. Walbran, P.D. and Henderson, R.A., In press. Modem and ancient perspectives on the crown-ofL.) in the Great Barrier Reef province, Australia, assessed thorns starfish (Acanthaster from the sediment record of John Brewer, Green Island and Heron Island reefs. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Walbran, P.D, Henderson, R.A., Jull, A.J.T. and Head, M.J., 1989. Evidence from sediments of long-term Acanthaster predation on corals of the Great Barrier Reef. Science, 245: 847850. Wanders, J.B.W., 1977. The role of benthic algae in the shallow reef of Curacao (Netherlands Antilles) In: The significance of grazing. Aquatic Bot., 3: 357-390. Wellington, G.M., 1982. Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol. Monogr., 52: 223-241. Wellington, G.M. and Victor, B.C., 1985. El Nirio mass coral mortality: a test of resource limitation in a coral reef damselfish population. Oecologia, 68: 15-19. Wells, J.W., 1957. Coral reefs. In: J. W. Hedgpeth (Editor), Treatise on Marine Ecology and Paleoecology, Vol. 1, Ecology, Geol. SOC.Amer. Mem. 67, New York, pp. 609-631. Wicksten, M.K., 1989. Why are there bright colors in sessile marine invertebrates? Bull. Mar. Sci., 45: 519-530. Williams, A.H., 1981. An analysis of competitive interactions in a patchy back-reef environment. Ecology, 62: 1,107-1,120. Williams, E.H. and Bunkley-Williams, L., In press. Circumtropical coral reef bleaching in 19871988. Proc. 6th Int. Coral Reef Symp., Townsville (1988). Williams, E.H. and Bunkley-Williams, L., In review. The world-wide coral reef bleaching cycle. Atoll Res. Bull. Williams, E.H., Jr., Goenaga, C. and Vicente, V., 1987. Mass bleachings on Atlantic coral reefs. Science, 238: 877-878. Woodley, J.D., 1979. The effects of trap-fishing on reef communities in Jamaica. Proc. Assoc. Is. Mar. Lab. Carib., Santa Marta, 13: 27 (abstract). Woodley, J.D., Chornesky, E.A., Clifford, P.A., Jackson, J.B.C., Kaufman, L.S., Knowlton, N., Lang, J.C., Pearson, M.P., Porter, J.W., Rooney, M.C., Rylaarsdam, K.W., Tunnicliffe, V.J., Wahle, C.M., Wulff, J.L., Curtis, A.S.G., Dallmeyer, M.D., Jupp, B.P., Koehl, M.A.R., Neigel, J., and Sides, E.M., 1981. Hurricane Allen's impact on Jamaican coral reefs. Science, 214: 749-755. Wood-Jones, F., 1910. Corals and Atolls. L. Reeve, London, 392 pp. Wyrtki, K., 1975. El Niiio--the dynamic response of the equatorial Pacific Ocean to atmospheric forcing. J. Phys. Oceanogr., 5: 572-584. Wyrtki, K., 1985. Pacific-wide sea level fluctuations during the 1982-83 El N5o. In: G. Robinson and E.M. del Pino (Editors), pp. 29-48, El Niiio in the Galapagos Islands: the 19821983 Event. Fundacion Charles Darwin para las Islas Galapagos, Quito, Ecuador. Yamaguchi, M., 1975. Sea level fluctuations and mass mortalities of reef animals in Guam, Mariana Islands. Micronesica, 11: 227-243. Yamaguchi, M., 1986. Acanthaster planci infestations of reefs and coral assemblages in Japan: a retrospective analysis of control efforts. Coral Reefs, 5: 23-30. Yamazato, K., 1981. A note on the expulsion of zooxanthellae during summer, 1980 by the Okinawan reef-building corals. Sesoko Mar. Sci. Lab. Tech. Rep., 8: 9-18. Yang, R.T., Yeh, S.Z. and Sun, C.L., 1980. Effects of temperature on reef corals in the NanWan Bay, Taiwan. Institute of Oceanography, National Taiwan University, Spec. Pub. 23: 127. Yonge, C.M., 1968. Living corals. Proc. Roy. SOC.B., 169: 329-344.
127
THE EFFECTS OF THE EL NInO/SOUTHERN OSCILLATION ON THE DISPERSAL OF CORALS AND OTHER MARINE ORGANISMS
R.H. RICHMOND Marine Laboratory, University of Guam, UOG Station, Mangilao, Guam 96923 USA ABSTRACT Richmond, R.H., 1989. The effects of the El Niiio/Southern Oscillation on the dispersal of corals and other marine organisms. The El Nido/Southern Oscillation is a global event that results in anomalous circulation patterns in the Pacific Ocean. Since the dispersal of marine organisms is directly related to the timing, speed, and direction of oceanic currents, El Niiio events can have a major effect on the distribution of species. Changes in the direction of flow open new routes for dispersal, while the increased velocity of currents enables organisms and larvae to disperse greater distances in shorter periods of time. As such, El Niiio events may be responsible for species introductions,range extensions, establishment of expatriate populations, as well as setting the stage for evolution and speciation in populations of marine organisms.
1 INTRODUCTION The study of the distribution of marine organisms (marine biogeography) relies on two types of theories to explain observed patterns, namely, vicariance and dispersal. Vicariance theories explain the present distribution of organisms as resulting from modification of pre-existing faunas through tectonic events, extinctions, and speciation (Croizat et al., 1974; Heck and McCoy, 1978; Nelson and Rosen, 1981; Springer, 1982).
These tectonic events include movement of
plates and their corresponding land masses, as well as uplift and subsidence of islands. According to the vicariance theories, once faunas become separated, extinction and speciation act to alter the species composition. The result is a distribution pattern in which geologically separated areas have shared elements, but noticeable differences. Dispersal theories include both rafting of individuals and dispersal of planktonic larvae. Rafting of organisms attached to floating debris has been hypothesized to explain the dispersal of numerous species, including barnacles, crabs, corals, and other benthic invertebrates (Jokiel, 1984). Rafting theory is supported by numerous accounts of animals found attached to driftwood, pumice, fishing floats, and miscellaneous man-made materials encountered on
beaches as well as in oceanic currents far from land masses. Long-distance dispersal of larvae via oceanic currents is also invoked to explain extant distribution patterns of marine organisms.
This theory is
based on the ability of larvae to remain competent for periods long enough to allow immigration from a source area some distance away, where competency is defined as the ability of a larva to settle successfully and metamorphose. There may be a pre-competency period during which larvae must undergo a degree of development before settlement is possible, and a post-competency period during which larvae are viable, but are no longer capable of settling and undergoing metamorphosis (Jackson and Strathmann, 1981).
For such dispersal
hypotheses, longer pre-competency and competency periods translate into greater dispersal potential. It is important to note that although the aforementioned theories of vicariance, rafting, and long-distance dispersal of larvae differ in their explanation of how the present distribution of organisms arose, these theories are not mutually exclusive, and the observed distribution patterns can best be explained by a combination of all three. Since oceanic currents are called upon as the mechanism for dispersal in two of the three hypotheses central to marine biogeography, the study of such currents is of extreme importance to our understanding of distribution patterns. We need to know parameters including current direction, velocity, depth, and duration. Since ocean currents are dynamic, varying their direction and speed seasonally, mean or average values are often used for ecological and biogeographic studies, Of greater importance, however, are the extreme values, since we are concerned with events on the geologic time scale. A strong El Nifio, such as that experienced in 1982-83, is an example of an extreme with respect to ocean currents. During this period, currents in the tropical and subtropical Pacific differed from "normal" conditions. These differences, and their effect on the distribution of marine organisms, are the subject of this study. 2 OCEANIC CURRENTS IN THE TROPICAL AND SUBTROPICAL PACIFIC
The El Nifio is responsible for numerous anomalies in weather patterns, sea surface temperatures, and oceanic currents (see Hansen, this volume).
In order
to understand the biogeographic implications of the event, it is necessary to review ocean current patterns during normal conditions. For the purposes set forth in this paper, the three components of interest are time, velocity and direction; each of which will be discussed for the North Equatorial Countercurrent, the Equatorial surface currents, and their corresponding undercurrents. The discussion of these currents will be brief, concentrating on information of interest to biologists. Here, currents will be considered as
129 conduits for the transport and dispersal of marine organisms. For more detailed (and accurate) accounts of currents, see Wyrtki (1965, 1966, 1973) and Magnier et al. (1973). 2.1 The North Eauatorial Countercurrent The North Equatorial Countercurrent (Fig. 1) has been the proposed route for the transport of benthic invertebrate larvae from the Central and Indo-West Pacific to the eastern Pacific (Dana, 1975; Richmond, 1987a; Scheltema, 1986). The position, width, direction, and velocity of this current vary seasonally
(Wyrtki, 1965, 1966). The Countercurrent is strongly affected by the position of the intertropical convergence. It flows eastward from May through December, with a mean velocity of 0.39 m/sec (0.75 knots).
During this period, the southern boundary
occurs between 4" and 6' N, and the northern boundary reaches as far as 11' N. The Countercurrent begins to dissipate in January, breaking up into disjunct cells by February and March, and is totally absent in April.
In May, as the
intertropical convergence moves northward away from the equator, the North Equatorial Countercurrent becomes re-established.
+ + Tihiti
I
yo"
200-
lfOO
1000 I
Fig. 1. Current patterns in the tropical Pacific Ocean during non-El Nifio years. 2.2 The North Eauatorial Current The North Equatorial Current is found to the north of the Countercurrent (Fig. 1).
For the purposes of this paper, discussion will be limited to the
130 band from 10" N to 20" N, and the boundary with the Countercurrent, with particular emphasis on the eastern region of the Pacific Ocean. As
with the Countercurrent, the North Equatorial Current varies seasonally.
From March through June, most water enters from the California Current. From July through December, the North Equatorial Current is fed by the Countercurrent, predominantly along the eastern Pacific boundary.
Flow i s
always westward (with a slight northward component in the region of interest) at a mean velocity of 0.15 m/sec. 2.3 The South Eauatorial Current The South Equatorial Current actually extends north of the equator, until it comes in contact with the North Equatorial Countercurrent at approximately
4" N. While flow is predominantly to the west, there are periods when eastward transport occurs, which may be the result of the Equatorial Undercurrent reaching the surface (Wyrtki, 1965).
While the South Equatorial Current
usually serves as a barrier to south-north transport of marine organisms, of notable interest to biologists, water from the Southern Hemisphere can enter the North Equatorial Countercurrent in May, when that eastward flow once again commences. Transport from the south to the north, and eventually to the east, can occur as these various currents converge. 2.4 The Eauatorial Undercurrent The Equatorial Undercurrent is a narrow, subsurface current, centered on the equator. Its core depth varies from approximately 50 m to 200 m, and it may actually surface on occasion. It is characterized as thin (200 m), flowing eastward with peak velocities of up to 1.5 m/sec, and i s cooler than the overlying surface currents (Taft and Jones, 1973; Knauss, 1978). This current is particularly noteworthy as being the only eastward corridor along the equator during non-El Nifio years. Under normal conditions, only the North Equatorial Countercurrent and the Equatorial Undercurrent could transport organisms from the areas of highest diversity, in the Tropical Central and Indo-West Pacific, to the more depauperate eastern Pacific. The cooler temperatures of the Undercurrent could potentially increase the pre-competency and competency period of larvae, by reducing the rate of metabolic expenditure. 3 OCEANIC CURRENTS DURING THE 1982-83 EL NIB0 The El Nifio/Southern Oscillation is a global phenomenon. The oceanic currents associated with it are the result of meteorological events that occur during the "pre-El Niilo period", which may last for several months. Most notably, south easterly winds push water from the eastern Pacific region of the Southern Hemisphere, across the equator, into the Northern Indo-West Pacific
131
(Fig. 2).
The force and duration of these winds will determine the extent of
changes in current patterns from normal conditions.
*.vawaii
Fig. 2. Pre-El Nifio conditions, during which south easterlies force water north and west.
200-
TLhiti
140"
100"
Fig. 3 . El Nifio related current patterns in the tropical Pacific Ocean with a strong eastward surface flow along the equator and subsequent intrusions of warm water up and down the coast of the Americas.
132
As a result of the sustained winds, seawater accumulates to the west until the south easterlies cease. Without the wind to maintain this unbalanced state, the stockpiled seawater is released, responding to the difference in sea level. The resulting anomaly, with higher sea levels at the equator and lower sea levels to the north and south, indicates a directional re-adjustment of the sea surface, called the Kelvin Wave.
This signals the actual onset of the El
Nitio. Associated with this is a strong anomalous surface flow to the east (Fig. 3 ) .
With the exception of occasional surfacing of the Equatorial
Undercurrent, this is the only time when there is an eastward surface flow along the equator. During the 1982-83 El Nifio, mean current velocity along the equator from mid-September through mid-Decemberwas 0.74 m/sec, with a peak flow of 1.40 m/sec (Wyrtki, 1985).
The result of this current is a 5,000km
displacement eastward of water parcels, and all organisms contained within them. Such an event could easily transport live material from the Line Islands to the eastern Pacific. 4 TRANSPORT OF MARINE ORGANISMS IN OCEANIC CURRENTS
Oceanic currents have been invoked in the dispersal of fish and marine invertebrates as well as for the dispersal of plants, seeds, and even terrestrial organisms (MacArthur and Wilson, 1967; Carlquist, 1974; Newman, 1986; Scheltema, 1986).
The dispersal and distribution of tropical, shallow-
water, benthic invertebrates are of particular interest, as this group provides for a conservative estimation of what may be possible. Many of these organisms are stenotypic, exhibiting a relatively narrow range of tolerances to environmental factors, and are quite specific in their settlement requirements. Their distribution is directly related to the pre-competency and competency periods of their larvae, the speed of the currents in which they are transported, and the distance between suitable habitats (Thorson, 1961; Scheltema, 1971; Jackson and Strathmann, 1981; Richmond, 1987a). The occurrence of larvae of benthic invertebrates in oceanic currents has been documented from plankton tows performed on numerous oceanographic cruises (Scheltema, 1986).
During the Helios Expedition (R.V. Melville, Sept.
-
Oct.
1987, San Diego, California to Tahiti), larvae representing five different invertebrate phyla, including Annelida, Cnidaria, Echinodermata, Mollusca, and Sipuncula, were collected in a single plankton tow from a station in the North Equatorial Countercurrent (6" N, 135' W; September, 1987; Scheltema, Richmond, in prep.). 4.1 TranSDOrt Durinrr Non-El Nido Years For the tropical Pacific, the longest jump between shallow-waterbenthic habitats is between the Central Pacific and the eastern Pacific, a distance of
133 approximately 5,200 km.
This distance has been referred to as Ekman's Barrier,
since there is a noticeable decline in species diversity from west to east, and the expanse of open ocean, without any island stepping-stones,appears to form a barrier to dispersal for most tropical invertebrate and fish species (Vermeij, 1987).
McKenna (1973), described the distance as a "filter", which
is a more accurate depiction, because numerous species have apparently crossed at some time, establishing populations in the eastern Pacific. For corals, the filtering effect of distance on diversity is striking, with a distinct gradient from west to east. There are approximately 350 hermatypic scleractinians recorded from the Philippines, 275 from the Mariana Islands, 240 from the Marshall Islands, 70 from the Line Islands, 42 from Hawaii, and only 11 species from the eastern Pacific reefs of Panama (Wells, 1954; Stehli and Wells, 1971; Maragos, 1974, 1977; Glynn and Wellington, 1983). The only surface route from the Central Pacific to the eastern Pacific during periods of normal circulation is the North Equatorial Countercurrent. At the mean current velocity of 0.39 m/sec, it would take 155 days to traverse the distance from the Line Islands to the eastern Pacific. Satellite-tracked drifter buoys actually covered the distance in 100 days, averaging 0.54 m/sec (Wyrtki et al., 1981).
Other buoys released simultaneously, exhibited speeds
as slow as 0.35 m/sec, and as fast as 0.90 m/sec.
In this example, the
observed buoy took less time to traverse the given distance than was predicted by the mean velocities measured, illustrating the importance of considering extremes rather than means in biogeographic studies. Actual larval competency data are scarce: however, based on the above values for transport time, certain benthic invertebrate species could make the jump under normal conditions (Thorson, 1950; Scheltema, 1974; Hadfield, 1978; Richmond, 1987a).
Larvae of two pan-Pacific scleractinians, Pocillouora
damicornis (Fig. 4) and Tubastrea (Richmond, 1987a; in prep.).
remain competent for over 100 days
Birkeland et al. (1971) reported a competency
period of 14 months for larvae of the temperate asteroid Mediaster aeaualis, while Kempf (1981) found that larvae of the tropical gastropod Aulvsia juliana remained competent for over 300 days. Additionally, subsurface larval transport from west to east could occur in the Equatorial Undercurrent. A major argument against the Undercurrent as a conduit for tropical organisms and larvae has been the cooler temperatures encountered at the interface between the surface current and the undercurrent (ca. 16°C
-
20°C).
While it is true that many tropical organisms are
relatively stenothermal, data indicate that the survival of some tropical invertebrate larvae is possible at these temperatures (Scheltema, 1968), and cooler temperatures,within limits, may slow the metabolism of lecithotrophic larvae, thus enhancing competency.
134
-
Fig. 4 . Planula larvae of the coral PocilloDora damicornis, from Enewetak populations (scales 0.5 mm). (a) 42 day old larva with extruded mesenterial filaments, which enable feeding while in the planktonic state. (b) 100 day old larva prior to successful settlement and metamorphosis. Pigmentation in both larvae is due to the presence of symbiotic zooxanthellae (arrows). In addition to the east - west "barrier" or "filter", there exists a similar obstruction to the north
-
south transportation of larvae across the
equator. With the exception of limited mixing at the boundaries between the South Equatorial Current and North Equatorial Countercurrent, and the North Equatorial Countercurrent and the North Equatorial Current, larvae would have to travel all the way to the east or most of the way to the west to move between the north Central Pacific and the south Central Pacific. Newman (1986) and others have commented on the faunal affinities between Hawaii and Polynesia, but no satisfactory dispersal mechanism has been proposed to explain this cross-equatorial distribution. 4 . 2 TransDort Durinp. Pre-El Niiio and El Niiio Conditions
During pre-El Niiio conditions, the south easterlies transport water from the eastern and central portions of the South Pacific to the western portion of the North Pacific (Fig. 2 ) .
At these times, larvae and organisms can be swept
from Polynesia, across the equator, into the North Equatorial Countercurrent, as well as past other island systems. As such, pre-El Nitio conditions may provide an explanation for the faunal affinities between Hawaii and Polynesia, where Hawaii possesses a subset of the Polynesian biota. Once the south easterlies cease, the water that has accumulated in the Western Pacific is released. As the El Niiio conditions commence (signaled by the passing of the Kelvin wave), eastward surface flow occurs along the equator, making it possible for organisms from the Line Islands to be swept into the eastern Pacific (Fig. 3 ) .
The Line Islands possess a rich marine
fauna that could be the source of eastern Pacific populations. During the
135 1982-83 El Niiio, eastward surface flow averaged 0.74 m/sec, with a maximum of 1.40 m/sec (Wyrtki, 1985).
This translates to a 5,200 km trip in 81 days at
the mean velocity, and in less than 50 days at peak flow. Conceivably, such a halving of the time necessary to make the longest jump across the Pacific would more than double the number of species that could qualify by the competency criterion. Scleractinian corals provide support for the theory of long distance dispersal of larvae via oceanic currents. PocilloDora damicornis is the predominant coral on the Pacific reefs of Panama (Glynn, 1984; Richmond, 1985). With a larval competency period exceeding 100 days, its wide distribution pattern and presence in the eastern Pacific can be explained under normal circulation conditions. Corals of the genus AcroDora have a more restricted distribution pattern in the Pacific, exhibiting a greater degree of endemism than many other coral genera (Veron, 1986).
Recent experiments performed on
AcrODOra larvae indicate that they have a much shorter competency period than PocilloDora larvae (< 25 days; Richmond, 1988).
Von Prahl and Mejia (1985)
reported the first record of live Acrouora from the eastern Pacific, finding several colonies of AcroDora valida around Gorgona Island, Colombia. While they attribute this distribution to vicariance mechanisms, the fact that all of the colonies were of similar size and were found in the same area suggests that the population resulted from a rare dispersal event associated with anomalous oceanic conditions, like those of El Nifio. This premise is further supported by the absence of AcroDora from the eastern Pacific fossil record (Heck and McCoy, 1978). Conditions in the tropical eastern Pacific can vary drastically during the year. Due to coastal upwelling, temperatures on eastern Pacific coral reefs can dip to daily lows below 15°C for periods of several days, and to daily lows below 17°C for weeks (Glynn and Stewart, 1973; Richmond, pers. obs.)*.
The
warm waters of the El Niiio may allow tropical larvae to immigrate into upwelling-affectedeastern Pacific habitats and, subsequently, for populations to become established before conditions return to normal. Organisms living at their limits of tolerance to environmental conditions, such as temperature, may be reproductively constrained, forming pseudo- or expatriate populations that are viable, but non-reproductive and dependent upon immigration or vegetative propagation for recruitment (Mileikovsky, 1971).
This may be the case for the
pan-Pacific coral Pocillouora damicornis that releases planula larvae monthly on the reefs of the Central and Indo-West Pacific, but has never been observed to planulate in the eastern Pacific, where its reproduction appears to be
*On two occasions in March 1985, I recorded minimum temperatures of 14.2-C on a coral reef in the Gulf of Panama.
136 solely vegetative (Richmond, 1987a,b). Grigg et al. (1981) observed that the same may be true for Acrouora spp. in the Northwestern Hawaiian Islands. Another effect of El Niao on the distribution of marine organisms is the poleward transport of individuals from indigenous and introduced populations in the eastern Pacific. The extraordinary eastward flow along the equator forces water masses north and south upon reaching the American coastline (Fig. 3). During the 1982-83 event, warmer surface water temperatures were recorded both north and south of the equator (Glynn, 1988).
These warmer water masses
contained both adult organisms and larvae and were responsible for range extensions as well as introductions (Karinen et al., 1985).
Brodeur (1986)
reported that the euphausiid NvctiDhanes simDlex, that normally occurs off the California coast, was found off the coasts of Oregon and Washington through the fall of 1983 and in reduced abundances in 1984. Discontinuous coastal distributions of subtropical organisms, with northern and southern, but no equatorial populations, can also be explained by El Niiio-related circulation patterns. Finally, El Niiio events may be important in evolution and speciation in marine organisms ( s e e also Vermeij, this volume). The introduction of larvae into new areas with different physical and biological conditions than those experienced in the habitat of origin can lead to allopatric speciation. Populations originating from the immigration of a few individuals can display different gene frequencies than the population of origin due to the founder effect and genetic drift (Mayr, 1971).
The amount of gene flow between
geographically isolated populations can be affected by events including the E l Niiio/Southern Oscillation. If such events occur relatively often, they can act to maintain genetic homogeneity among populations. If, on the other hand, extreme events are rare, they can set the stage for isolation, evolution and speciation. Some eastern Pacific reef populations of the coral identified as Pocillopora damicornis are morphologically similar to, but biologically distinct from, those of the Central and Indo-West Pacific. The geographically separated populations differ in life history characteristics including fecundity, growth rate, reproductive allocation, age specific mortality, and interspecific competitive ability (Richmond, 1985; 1987b).
Most notably,
eastern Pacific populations have not been observed to produce larvae. As such, some contemporary eastern Pacific reef populations may represent pseudo- or expatriate populations of Central/Indo-West Pacific origin (Mileikovsky, 1971), resulting from the introduction of larvae into a new habitat via long-distance chance dispersal events like the El Nitlo. The unique physical and biological characteristics of the eastern Pacific reef environments may have provided strong selection pressure in favor of genotypes that do not planulate, but
137 fragment and grow rapidly (Richmond, 1985; 1987b). The occurrence of non-reef populations of p. damicornis, i.e. those relatively small, individual colonies attached to nearly vertical or steeply sloping basaltic walls, may be the result of recent dispersal and immigration events from the Central Pacific.
Alternatively, these colonies of apparent
larval origin may be the result of "polyp bail-out,''as sloughed tissue and polyps from stressed adult p. damicornis colonies are capable of dispersal, settlement and development into new colonies (Sammarco, 1982; Richmond, 1985) 5 CONCLUSIONS Pre-El NiAo and El Niiio conditions result in anomalous circulation patterns, extremes in current velocities, and elevations in sea-surface temperatures. As such, these periodic events can result in species introductions, range extensions, and the establishment of expatriate populations beyond those that could be accounted for under normal circulation conditions. El Nifio events can virtually break east-west and north-south dispersal barriers, or at least increase the pore size of the "filters", allowing larvae with shorter competency periods to cross long distances. The increased current velocity associated with the El Nifio can reduce the competency period requirements for making connections by as much as 50%. The anomalous circulation patterns open new dispersal routes, allowing for transport of organisms and effecting gene flow between geographically separated populations. Specifically, pre-El NiAo circulation patterns may provide a mechanism to explain the faunal affinities between Polynesia and the Hawaiian Archipelago. One major El NiAo-related dispersal event every 200 years is equivalent to 5,000 such events per million years. It is the author's opinion that El Nitio events, over geological time, have played an important role in establishing observed biogeographic distribution patterns in the Pacific Ocean. ACKNOWLEDGEMENTS The author gratefully acknowledges Drs. M.G. Hadfield, W.A. Newman, R.S. Scheltema, P.L. Jokiel, K. Wyrtki, and B.R. Rosen for helpful suggestions and discussions, T. Rock for help with figures and illustrations, and Dr. P.W. Glynn, for sparking my interest in the effects of the El NiAo on coral and larval biology. Data on the occurrence of benthic invertebrate larvae in oceanic currents were collected in collaboration with Dr. R.S. Scheltema, aboard the R.V. Thomas Washington (Papatua Expedition), and the R.V. Melville (Helios Expedition), with support from NSF grant 8614579 (to R. Scheltema), and Research Opportunity Awards from NSF and the Smithsonian Institution. Field work in the eastern Pacific was supported by the Office of Fellowships and
138 Grants, Smithsonian Institution, and the National Science Foundation (grant OCE-8415615 to P.W. Glynn).
This is contribution number 247 of the University
of Guam Marine Laboratory. 6 REFERENCES Birkeland, C., Chia, F. and Strathmann, R.R., 1971. Development, substratum selection, delay of metamorphosis and growth in the seastar Mediaster aeaualis Stimpson. Bio. Bull., 141: 99-108. Brodeur, R.D., 1986. Northward displacement of the euphausiid Nvctiuhanes simDlex Hansen to Oregon and Washington waters following the El Niiio event of 1982-83. J. Crust. Biol., 6(4): 686-692. Carlquist, S . , 1974. Island Biology. Columbia University Press, New York, 660 pp. Croizat, C.B., Nelson, G.J. and Rosen, D.E., 1974. Centers of origin and related concepts. Syst. Zool., 23: 265-287. Dana, T.F., 1975. Development of contemporary eastern Pacific coral reefs. Mar. Biol., 33: 355-374. Glynn, P.W., 1984. Widespread coral mortality and the 1982-83 El Niiio warming event. Environ. Conserv., 11: 133-146. Glynn, P.W., 1988. El Niiio-SouthernOscillation 1982-83:nearshore population, community and ecosystem responses. Annu. Rev. Ecol. Syst. 19: 309-345. Glynn, P.W. and Stewart, R.H., 1973. Distribution of coral reefs in the Pearl Islands (Gulf of Panama) in relation to thermal conditions. Limnol. Oceanogr., 18: 367-379. Glynn, P.W. and Wellington, G.M., 1983. Corals and Coral Reefs of the Galapagos Islands. University of California Press, Berkeley/Los Angeles, 330 pp. Grigg, R.W., Wells, J.W. and Wallace, C., 1981. Acrouora in Hawaii. Part 1. History of the scientific record, systematics, and ecology. Pac. Sci., 35: 1-13. Hadfield, M.G., 1978. Growth and metamorphosis of planktonic larvae of Ptychodera flava (Hemichordata: Enteropneusta). In: M.E. Rice and F. Chia (Editors), Settlement and Metamorphosis of Marine Invertebrate Larvae. Elsevier, Amsterdam/Oxford/New York/Tokyo, pp. 247-254. Heck, K.L. and McCoy, E.D., 1978. Long distance dispersal and the reefbuilding corals of the eastern Pacific. Mar. Biol., 48: 349-356. Jackson, G.A. and Strathmann, R., 1981. Larval mortality from offshore mixing as a link between precompetent and competent periods of development. Am. Nat., 118: 16-26. Jokiel, P.L., 1984. Long distance dispersal of reef corals by rafting. Coral Reefs, 3: 113-116. Karinen, J.F., Wing, B.L. and Straty, R.R., 1985. Records and sightings of fish and invertebrates in the eastern Gulf of Alaska and oceanic phenomena related to the 1983 El Niiio event. In: W.S. Wooster and D.L. Fluharty (Editors), El NiSio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Washington Sea Grant Publication WSG-WO 85-3, Seattle, pp. 253-267. Kempf, S.C., 1981. Long-lived larvae of the gastropod Aulvsia iuliana: do they disperse and metamorphose or just slowly fade away? Mar. Ecol. Prog. Ser., 6: 61-65. Knauss, J A., 1978. Introduction to Physical Oceanography. Prentice-Hall, Inc., New Jersey, 338 pp. MacArthur, R.H. and Wilson, E.O., 1967. The Theory of Island Biogeography. Monographs in Population Biology 1. Princeton University Press, New Jersey, 203 pp. Magnier, Y., Rotschi, H., Rual, P., and Colin, C., 1983. Equatorial circulation in the Western Pacific. Progress in Oceanography, 6 , B.A. Warren (Editor). Pergamon Press, Oxford, pp. 29-46. Maragos, J.E., 1974. Reef corals of Fanning Atoll. Pac. Sci., 28: 247-255.
139 Maragos, J.E., 1977. Scleractinia. In: D.M. Devaney and L.G. Eldredge (Editors), Reef and Shore Fauna of Hawaii, Section 1: Protozoa through Ctenophora. Bishop Museum Press, Honolulu, pp. 158-242. Mayr, E., 1971. Populations, Species and Evolution, Harvard University Press, Cambridge, Massachusetts, 453 pp. Mileikovsky, S.A., 1971. Types of larval development in marine bottom invertebrates, their distribution and ecological significance: a reevaluation. Mar. Biol., 10: 193-213. McKenna, M.C., 1973. Sweepstakes, filters, corridors, Noah's Arks, and beached Viking funeral ships in palaeogeography. Implications of Continental Drift to the Earth Sciences, Vol. 1. D.H. Tarling and S . K. Runcorn (Editors), Academic Press, New York, pp. 295-308. Nelson, G. and Rosen, D.E. (Editors), 1981. Vicariance Biogeography: A Critique, Columbia University Press, New York, 616 pp. Newman, W.A., 1986. Origin of the Hawaiian marine fauna: dispersal and vicariance as indicated by barnacles and other organisms. Crustacean Biogeography, In: R. H. Gore and K. L. Heck (Editors). A.A. Balkema Press, Rotterdam, pp. 21-49. Richmond, R.H., 1985. Variations in the population biology of Pocillouora damicornis across the Pacific Ocean. Proc. 5th Intl. Coral Reef Congr., Tahiti, 6: 101-106. Richmond, R.H., 1987a. Energetics, competency, and long-distance dispersal of planula larvae of the coral Pocillouora damicornis. Mar. Biol., 93: 527533. Richmond, R.H., 1987b. Energetic relationships and biogeographical differences among fecundity, growth and reproduction in the reef coral Pocillouora damicornis. Bull. Mar. Sci., 41: 594-604. Richmond, R.H., 1988. Competency and dispersal potential of planula larvae of a spawning versus a brooding coral. Proc. 6th Intl. Coral Reef Symp., Townsville, in press. Sammarco, P.W., 1982. Polyp bail-out: an escape response to environmental stress and a new means of reproduction in corals. Mar. Ecol. Prog. Ser., 10: 57-65. Scheltema, R.S., 1968. Dispersal of larvae by equatorial ocean currents and its importance to the zoogeography of shoalwater tropical species. Nature, 217: 1159-1162. Scheltema, R.S., 1971. The dispersal of larvae of shoal-water benthic invertebrate species over long distances by ocean currents. 4th Eur. Mar. Biol. Symp., D. Crisp (Editor). Cambridge University Press, London, pp. 7-28. Scheltema, R.S., 1974. Biological interactions determining larval settlement of marine invertebrates. Thal. Jugosl., 10: 263-296. Scheltema, R.S., 1986. Long-distance dispersal by planktonic larvae of shoalwater benthic invertebrates among Central Pacific islands. Bull. Mar. Sci., 39: 241-256. Springer, V.G., 1982. Pacific plate biogeography, with special reference to shorefishes. Smith. Contr. Zool., 367: 1-182. Stehli, F.G. and Wells, J.W., 1971. Diversity and age patterns in hermatypic corals. Syst. Z o o l . , 20: 115-126. Taft, B . A . and Jones, J.H., 1973. Measurements of the equatorial undercurrent in the eastern Pacific. Progress in Oceanography, 6, B. A. Warren (Editor). Pergamon Press, Oxford, pp. 47-110. Thorson, G., 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev., 25: 1-45. Thorson, G . , 1961. Length of pelagic life in marine invertebrates as related to larval transport by ocean currents. Oceanography, M. Sears (Editor). Amer. Assoc. Adv. Sci. Washington, pp. 455-474. Vermeij, G.J., 1987. The dispersal barrier in the tropical Pacific: implications for molluscan speciation and extinction. Evolution 41: 10461058.
140
Veron, J.E.N., 1986. Corals of Australia and the Indo-Pacific. Angus and Robertson Publishers, Australia, 644 pp. Von Prahl, H. and Mejia, A., 1985. Primer informe de un coral acroporido, Acropora valida (Dana, 1846) (Scleractinia: Astrocoeniida: Acroporidae) para el Pacific0 americano. Rev. Biol. Trop., 33: 39-43. Wells, J.W., 1954. Recent corals of the Marshall Islands. Prof. Pap. U.S. Geol. Surv., 260: 385-486. Wyrtki, K., 1965. Surface currents of the eastern tropical Pacific Ocean. Bull. Inter-Am. Trop. Tuna. Corn., 9: 271-304. Wyrtki, K., 1966. Oceanography of the eastern equatorial Pacific Ocean. Oceanogr. Mar. Biol. Ann. Rev., 4: 33-68. Wyrtki, K., 1973. Teleconnections in the equatorial Pacific Ocean. Science, 180: 66-68. Wyrtki, K., 1985. Sea level fluctuations in the Pacific during the 1982-83 El Niiio. Geophy. Res. Lett., 12: 125-128. Wyrtki, K., Firing, E., Halpern, D., Knox, R., McNally, G.J., Patzert, W.C., Stroup, E.D., Taft, R., and Williams, R., 1981. The Hawaii to Tahiti shuttle experiment. Science, 211: 22-28.
141
CORAL MPRTALITY OUTSIDE OF THE EASTERN PACIFIC DURING 1 9 8 2 - 1 9 8 3 : TO EL NINO
RELATIONSHIP
M.A. COFFROTH Division of Biology and Living Resources, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4 6 0 0 Rickenbacker Causeway, Miami, FL 3 3 1 4 9 [USA1 and Department of Biological Sciences, State University of New York at Buffalo, Buffalo, NY 1 4 2 6 0 (USA) H.R. LASKER Department of Biological Sciences, State University of New York at Buffalo, Buffalo, NY 1 4 2 6 0 IUSAI J.K. OLIVER Sir George Fisher Centre for Tropical Marine Studies, James Cook University of North Queensland, Townsville, Queensland 4 8 1 1 (Australia)
ABSTRACT Coffroth, M.A., Lasker, H.R., and Oliver, J.K. 1 9 8 9 . Coral mortality outside of the eastern Pacific during 1 9 8 2 - 1 9 8 3 : relationship to El Niiio. 1 9 8 2 - 8 3 witnessed the most widespread coral bleaching and mortality in recorded history. In early 1 9 8 2 coral bleaching and subsequent mortality was reported from the Great Barrier Reef (GBRJ. These reports were followed in early 1 9 8 3 with observations of coral bleaching and mortality in the Java Sea, Tokelau Islands, French Polynesia, Galapagos Islands, and on the Pacific coast of Panama and in m i d - 1 9 8 3 with bleaching in the southwestern Indian Ocean, southern Japan and the Caribbean. This world-wide coral bleaching and mortality coincided with one of the strongest El Nizos on record. Many ecological anomalies that occurred during 1 9 8 2 - 8 3 have been related to this strong El Nifio/Southern Oscillation IENSOJ event and we have sought to ascertain if the observed coral bleaching and morta1it.y outside of the eastern Pacific can also be attributed to ENSO-related effects, such as shifts in wind patterns, water mass movements and solar insolation. In the central Pacific the relationship is clear. ENSO-related anomalies included an increase in the frequency of cyclones and a lowering of sea level that were directly responsible for the extensive coral mortality observed there. In other regions, where the signal of ENSO is less well defined, a causal relationship between ENSO and coral death is more difficult to establish. On the Great Barrier Reef, high temperatures and high light may have played a role in the bleaching. However, the proximal cause of bleaching is not known, and ENSOrelated anomalies that were observed in the area during 1 9 8 2 - 8 3 did not coincide with the bleaching event. Therefore, a causal relation cannot be established. On the Caribbean coast of Panama (San Blas Islands), elevated water temperatures are thought to have led to the observed bleaching. Analyses of available meteorological and oceanographic data failed to reveal a mechanism for the onshore warming nor is there a clear link between the bleaching and the ENSO. The lack of knowledge on the mechanisms and global interactions of ENSO, as well as the lack of appropriate meteorological and oceanographic data at the sites of bleaching/mortality, have hindered our ability to establish a connection between the 1 9 8 2 - 8 3 ENSO event and coral mortality.
142 1 INTRODUCTION
In early 1983, during the height of the 1982-83 El NiEo, Glynn and co-workers observed a widespread mortality of corals throughout the eastern Pacific (Glynn, 1983; 19841. Those mortality events have been convincingly related to increases in water temperature associated with El NiEo IGlynn, 1984; Glynn, et al., 1988; see Glynn, this volumel. However, a large percentage of the damage to corals during the 1982-83 El NiEo occurred outside of the eastern Pacific, the area most commonly associated with El NiEo occurrences (Table 1 1 . Starting in early 1982 and continuing through mid-1983 there were observations of coral bleaching and mortality from most of the world's tropical oceans. Bleached corals were observed on the Great Barrier Reef (early 19821, in the Java Sea, Tokelau Islands, French Polynesia, Galapagos Islands and on the Pacific coast of Panama (early 19831 and in the southwestern Indian Ocean, southern Japan and the Caribbean (mid-19831. El Nifio is part of an atmospheric-oceanic interaction, the El NiEoSouthern Oscillation (ENSO, see Hansen, this volume). The 1982-83 ENSO generated dramatic ecological effects on a global scale. Because of the extensive nature of the ecological anomalies, there has been a tendency to relate all anomalies during 1982-83 to the ENSO phenomenon. However, some, or perhaps many, of these events may have been spuriously correlated with the ENSO episode. In this study we examine whether the coral mortality observed during 1982-83 can be attributed to environmental perturbations caused by ENSO-related effects, such as shifts in wind patterns, water mass movements and solar insolation. We then look in detail at the relationship between ENSO-related changes in the physical environment and coral mortality at two sites, the Great Barrier Reef IGBRJ in the western Pacific and the San Blas Islands on the Caribbean coast of Panama. 1.1 Causes of Coral Bleaching and Mortality Much of the coral mortality observed during 1982-83 was preceded by coral bleaching. It is therefore important to identify the proximal causes of coral bleaching in order to relate the bleaching event to the ENSO. Bleaching refers to the whitening of coral tissue due to the loss of zooxanthellae, intracellularly located photosynthetic symbionts. Almost any physiological stress can cause a coral to expel its zooxanthellae including thermal stress, altered light levels, siltation, and reduced salinity. The ability of corals to survive and regain their zooxanthellae varies among species and with the extent of bleaching.
143 Of the factors listed above, temperature and light are the most frequently cited agents of coral bleaching. These factors act separately and
TABLE 1 Coral bleaching and mortalities observed during 1 9 8 2 - 8 3 Location
Date
Observation Probable Cause
Eastern Pacific/ Galapagos
2/83
Bleaching/ mortality
Central Pacific: French Polynesia
Tokelau Is. Nukunonu Atoll Western Pacific: Great Barrier Reef, Australia
elevated water temp.
exposure; early mortality/ cyclonic 1983 physical destruction early mortality 1983
early bleaching/ 1982 mortality
Reference Glynn, 1 9 8 3 ; 1984
Laboute, 1 9 8 5 ; Harmelin-Vivien & Laboute, 1 9 8 6
aerial exposure
Glynn, 1 9 8 4
unknown high temp? high light?
Fisk & Done, 1 9 8 5 ; Harriott 1 9 8 5 ; Oliver, 1985
elevated water temp.
Kamezaki & Ui, 1 9 8 4
Indian Ocean/Arabian Sea: 3/83 mortality Arabian Sea
oil spill?
Burchard & McCain, 1 9 8 4
Java Sea
1 /83 3-5/83
bleaching/ mortality
elevated water temp.
Suharsono & Kiswara, 1 9 8 4 Brown, 1 9 8 7
Reunion Island
1983
bleaching/ morta1ity
elevated water temp. low tides
Guillaume et al., 1 9 8 3
Mayotte Island
5-6/83
bleaching/ mortality
elevated water temp. sedimentation
Faure et al.,
Southern Japan
8/83
bleaching/ mortality
1984
Caribbean: Bahamas
9/83
bleaching
unknown
Glynn, 1 9 8 4
Florida Keys
9/83
bleaching/ mortality
elevated water temp.
Jaap, 1 9 8 5
Costa Rica
6/83
bleaching/ mortality
elevated water temp.
Cortes et al.,
bleaching/ mortality
elevated water temp.
Glynn, 1 9 8 4 Lasker et al., 1 9 8 4
San Blas Is., Panama
6/83
1984
144
synergistically to stress corals lColes and Jokiel, 1 9 7 8 1 . For example, under natural conditions increased solar irradiance elevates water temperatures. Heating due to decreased rainfall, calm weather and midday low tides can further accentuate the temperature increases. Shinn 1 1 9 6 6 ) and Jaap ( 1 9 7 9 ) have reported periods of increased water temperature and concurrent coral bleaching from the Florida Keys. Fankboner and Reid ( 1 9 8 1 ) observed massive expulsions of zooxanthellae at Enewetak Atoll, which they attributed to thermal stress. Yamazato 1 1 9 8 1 ) observed bleaching in both scleractinians (Seriatopora sp., Stylophora pistillata, Pocillopora damicornis, and Montipora sp.) and alcyonarians (Sinularia sp., Lobophytum sp., and Sarcophyton sp.). He attributed the extensive bleaching to elevated water temperature resulting from a combination of increased solar irradiance, reduced precipitation and midday spring tidal exposures IYamazato, 1 9 8 1 ) . Mass mortalities of reef flat corals (Puerto Rico-Glynn, 1 9 6 8 ; GuamYamaguchi, 1 9 7 5 ; and the Red Sea-Loya, 1 9 7 6 ) have also been attributed to solar heating of waters with circulation restricted by low tides and calm weather. Thermal stress, desiccation from exposure at low tides and direct damage from ultraviolet ( U V ) radiation contributed to the mortality of reef flat corals. Low temperatures as well as high temperatures can induce bleaching in corals. Cold water incursions from the shallow waters of Florida Bay have killed corals in the Florida Keys IShinn, 1 9 6 6 ; Roberts, et al., 1 9 8 2 ) and irregularities in coral density bands show that these events have occurred frequently [Hudson, et al., 1 9 7 6 ) . Coral death due to low water temperature also has been observed in the Dry Tortugas (Porter, et al., 1 9 8 2 ) and the Persian Gulf (Shinn, 1 9 7 6 ) . Low light levels also result in coral bleaching. Coral colonies maintained in the dark for long periods whitened and eventually died (Yonge and Nicholls, 1 9 3 1 ; Franzisket, 1 9 7 0 1 . Tropical storms are important agents of reef mortality. Mechanical damage from tropical storms has caused extensive coral mortality in numerous instances (Moorhouse, 1 9 3 6 ; Stephenson, et al., 1 9 5 8 , Stoddart, 1 9 6 3 ; Glynn, et al., 1 9 6 4 , Ball, et al., 1 9 6 7 ; Baines, et al. 1 9 7 4 ; Woodley et al., 1 9 8 1 1 . Hurricane Hattie resulted in near total destruction of some reefs off Belize in 1 9 6 1 (Stoddart, 1 9 6 3 ) . Hurricane Allen had a similar impact on Jamaican reefs in 1 9 8 0 with mortality as high as 9 4 % in some areas (Woodley et al., Furthermore, heavy rains and flooding associated with tropical storms can lower water salinity, which induces bleaching. E g a k and DiSalvo ( 1 9 8 2 ) suggested that heavy rains ( 9 7 . 6 mm in 2 4 h ) and the resultant reduced salinity led to the bleaching of several Pocillopora spp. in the vicinity of Easter Island in 1 9 8 0 . Goreau ( 1 9 6 4 ) observed bleaching of over 2 0 species of 1981).
Jamaican reef anthozoans following heavy rains accompanying Hurricane Flora,
145
and reported mortalities of over 5 0 % on many reefs. Similar coral mortality due to the heavy rains and flooding associated with tropical cyclones have occurred on the Queensland coast of Australia and the Great Barrier Reef (Stephenson et al., 1958; Slack-Smith, 1959; Stoddart, 1969). Finally, high levels of siltation, as a result of runoff from heavy rains and from human activities, may induce corals to bleach. Bak 119781 observed loss of zooxanthellae and subsequent mortality in Porites astreoides due to high sedimentation on the reefs at Curasao. Marszalek (19811, in a study of the impact of dredging on a Florida reef community, observed that extended exposure to siltation and high turbidity caused bleaching in some corals. Acevedo and Goenaga 11986) attributed bleaching of several coral species on a Puerto Rican reef to high turbidity from mainland erosion and sediment-loaded river discharge after heavy rains. Many of the bleaching and mortality events reported during 1982-83 were attributed to factors such as elevated water temperatures, reduced salinities, sedimentation, aerial exposures, and tropical storms. Thus, in many respects the coral mortality events of 1982-83 are not unusual. The uniqueness of 1982-83 is the global extent of the disturbance--coral mortality was reported
€rom all the world's tropical oceans (Fig. 1 and Table 1 ) .
I'
Fig. 1 . Location of coral bleaching and mortality observed during 1982-1983 1 1 ) Eastern Pacific/Galapagos; 1 2 1 French Polynesia (Moored, Tuamotu Archipelago); 13) Tokelau Islands, Nukunonu Atoll; 1 4 ) Great Barrier Reef, Australia; ( 5 ) Java Sea; 1 6 1 Southern Japan; ( 7 ) Reunion Island/Mayotte, western Indian Ocean; 1 8 ) Arabian Sea; I91 Bahamas; ( 1 0 ) Florida Keys; I 1 1 1 San Blas Islands, Panama/Costa Rica.
146
2 ENSO
AND CORAL MORTALITY
The 1 9 8 2 - 8 3 ENSO was one of the most intense on record IRasmusson and Wallace, 1 9 8 3 ; Rasmusson et al., 1 9 8 3 a ; Kousky et al., 1 9 8 4 ; Hansen, this volumel. Although the "mechanism of origin" of the ENSO event is described in terms of the Pacific basin, meteorological and oceanographic studies have shown ENSO to be a global event IHalpern, 1 9 8 3 ; Philander, 1 9 8 3 ; Pan and Oort, 1983;
Gill and Rasmusson, 1 9 8 3 ; Rasmusson and Wallace, 1 9 8 3 ; Kousky et al.,
1984;
Hamilton, 1 9 8 5 ; Michelchen, 1 9 8 5 ) .
Events correlated with ENSO include
severe droughts in northeastern Brazil, northern Australia, Indonesia, and southern Africa (Hastenrath, 1 9 7 6 , 1 9 7 8 ; Covey and Hastenrath, 1 9 7 8 ; Philander, 1 9 8 3 ; Rasmusson and Wallace, 1 9 8 3 ; Kousky et al., 1 9 8 4 1 , changes in the timing of onset of the Indonesian monsoon (Barnett, 1 9 8 4 1 , anomalous sea surface temperatures off the northwest coast of Africa IMichelchen, 1 9 8 5 1 , reduced hurricane activity in the Atlantic [Gray, 1 9 8 4 , 1 9 8 5 1 , and altered temperatures and precipitation over North America IHorel and Wallace, 1 9 8 1 ; Chen, 1 9 8 3 ; Pan and Oort, 1 9 8 3 ; Philander, 1 9 8 3 ; Hamilton, 1 9 8 5 ; Ropelewski and Halpert, 1 9 8 6 ; Lough and Fritts, this volume). To investigate the relationship of this global event to reports of coral mortality, we will summarize briefly what is known of the effect of ENSO on the physical parameters of the world's oceans and compare these effects with reports of coral bleaching and mortality. These comparisons are summarized in Table 2 . 2.1
Eastern Pacific The presence of anomalous high sea surface temperatures ISSTsJ in the
equatorial eastern Pacific is a key signal of El NiEo IHansen, this volumel. The massive coral mortality that occurred along the coastal regions of the eastern Pacific and in the Galapagos Islands was attributed to these high temperatures and, possibly to high light intensities IGlynn, 1 9 8 4 , 1 9 8 8 , this volume; Robinson, 1 9 8 5 1 . 2.2
Central Pacific In the central Pacific the signature of the ENSO episode is relatively well
defined and three effects have been observed that could affect corals: la1 changes in rainfall, l b l changes in the incidence of tropical storms and lcl sea level fluctuations. Heavy rainfall eastward of the dateline in the south central Pacific resulted from a shift eastward during September and October 1 9 8 2 of the Indonesian-New Guinea convergence region and an intensification and southward displacement of the Intertropical Convergence Zone IRasmusson et al., 1 9 8 3 a , 1 9 8 3 b ; Gill and Rasmusson, 1 9 8 3 , Philander, 1 9 8 3 ) .
147
TABLE 2 A summary of ENSO-correlated anomalies observed in the world's oceans that could potentially affect corals, the probable cause of the observed bleachingimortality in these regions and whether a causal relationship between the 1 9 8 2 - 8 3 ENSO and coral bleaching/mortality has been established. See text for details and references.
Selected Anomalies correlated with ENSO Central Pacific: - Increased rainfall - Increased storms - Drop in sea level Western Pacific: Increased SST prior to ENSO [late 1 9 8 2 1 Decreased SST (June 1982 I Droughts over Australasia (late 1 9 8 2 ) Anomalous westerly winds (June, 1 9 8 2 )
Probable Cause
- Cyclones - Prolonged exposures
-
Indian Ocean/ Arabian Sea: - Early onset of westerly monsoon-prior to ENSO - Anomalous easterly winds during ENSO - Drought in southern India, southern Africa
Causal Relationship?
Yes
Unknown Not established Elevated water temperature? High light levels?
Elevated water temperature Low tides Sedimentation
Caribbean: - Anomalous westerly winds - Elevated water temperature - Increased surface trade winds - Droughts in NE Brazil
Not established
Not established
The origins of tropical cyclones also shifted eastward due partly to an eastward shift of the maximum SST (Eldin and Donguy, 1 9 8 3 ; Rasmusson and Wallace, 1 9 8 3 ; Gill and Rasrnusson, 1 9 8 3 1 . East of the dateline this led to a cyclone season unsurpassed in intensity in recent history (Chen, 1 9 8 3 ; Eldin and Donguy, 1 9 8 3 ; Sadler and Kilonsky, 1 9 8 3 ; Rasmusson et al., 1 9 8 4 ; Revel1 and Goulter, 1 9 8 6 ) . Sea levels were depressed across the Central Pacific from the Solomon Islands to Tahiti (Wyrtki, 1 9 8 5 ) . Sea levels 3 0 - 4 0 cm below normal were recorded in the western and south-central Pacific at Funafuti (9" S 1 7 9 O El
148
and Penrhyn 19's
158' WJ during the spring and summer of 1983 with above
normal sea levels observed at Noumea 122O S 166' EJ and in the eastern Pacific at the Galapagos IWyrtki, 1982; 1984; Cane, 1983). These sea level fluctuations during the ENSO were generated by the movement of warm water across the western Pacific to the eastern Pacific by surface currents generated by a Kelvin wave (Wyrtki, 1975, 1982; Hansen, this volume). The ENSO effects in the central Pacific had direct ramifications for reefs in the region. At Nukunonu Atoll in the Tokelau Islands, the large areas of coral mortality observed on reef flats in early 1983 were attributed to a drop
in sea level of approximately 60 cm and the subsequent exposure and reduced circulation (Glynn, 1984). Salvat (in Glynn, 19841 attributed extensive coral mortality at Moorea Island, French Polynesia, to subaerial exposures due to a drop in sea level, as well as physical destruction from several cyclones in early 1983. On the outer slopes of the reefs of Tikehau and Takapoto atolls, Tuamotu Archipelago, cyclone damage accounted for mortalities of 80% ITikehauJ and 50-100% (TakapotoJ (Laboute, 1985; Harmelin-Vivien and Laboute, 1986). The observed sea level anomalies and the shift in the zone of tropical cyclogenesis to east of the dateline were clearly the principal causes of coral mortality in the Central Pacific. Since these anomalies have been unambiguously related to the 1982-83 ENSO, the causal relation between ENSO and coral mortality in this region is clear. 2.3 Western Pacific Meteorological and oceanographic anomalies associated with the ENSO event in the western Pacific are generally of smaller amplitude than those recorded
in the central and eastern Pacific IHastenrath and Wu, 1983; Nicholls, 1984a; 1984331. Thus, a causal relationship between ENSO and the observed coral mortalities is more difficult to establish. Sea surface temperature, winds and rainfall are among the most relevant and best documented parameters that are significantly correlated with the Southern Oscillation ( S O J . There is a strong negative correlation between SSTs off northern Australia and SST off Peru INicholls, 1984a; 1984b; Wright, 1984). SSTs off northern Australia are generally high during the year prior to an El Niso, but then become anomalously low. Winds in the western Pacific also show a distinct and anomalous pattern during an ENSO event. Anomalous westerlies first appeared in the western equatorial Pacific in June 1982, and propagated eastward, generating in turn warm water and high sea levels in the eastern Pacific ISadler and Kilonsky, 1983; Rasmusson et al., 1983a, 198333; Holland and Nicholls, 1985). El NiEo years are also characterized by widespread drought in the Australasian region with the shift of the Indonesian-New Guinea convergence
149
zone to the east during the initial phase of ENSO (Pittock, 1 9 7 5 ; Quinn et al., 1 9 7 8 ; Philander, 1 9 8 3 ; Rasmusson and Wallace, 1 9 8 3 ; Rasmusson et al., 1 9 8 3 a , 1983331. Coughlan I 1 9 8 3 1 found that up to 4 0 % of the yearly rainfall variation in southeast Australia could be accounted for by the SO. In the western Pacific, bleaching and subsequent mortality on the Great Barrier Reef IGBRJ occurred on nearshore, mid-shelf and shelf-edge reefs in January-March 1 9 8 2 , a year before the sea surface warming off Peru. This incident affected at least 25 species of scleractinians and 2 species of alcyonaceans IFisk and Done, 1 9 8 5 ; Harriott, 1 9 8 5 , Oliver, 1 9 8 5 1 , and occurred primarily in shallow waters. UV radiation, subaerial exposure, and warmer water temperatures have been proposed as the potential causes. Harriott ( 1 9 8 5 ) reported slightly warmer GBR water temperatures in summer 1 9 8 1 - 1 9 8 2 than in the summers of either 1980-81 or 1 9 8 2 - 8 3 . Coral mortality was observed 1 8 months later in August 1 9 8 3 in southern Japan (Kamezaki and Ui, 1 9 8 4 1 . Although the shallow reef areas were most affected, bleaching of actinians, hydrocorals, scleractinians and alcyonaceans occurred to 1 0 m depth on the islands of Okinawa, Yaeyama, and Iriomote (Kamezaki and Ui, 1 9 8 4 ) . Sea surface temperatures in excess of 3 l o C were reported and high temperature was the probable cause of mortality. ENSO related anomalies were apparent in the western Pacific during the 1 9 8 2 - 8 3 ENSO, but the relationship to the observed bleaching events is not clear. The principal effect of ENSO in Australia, the extensive droughts that occurred in late 1 9 8 2 , cannot explain the GBR mortalities, as the latter
events occurred early in 1 9 8 2 . The warm SST anomalies observed north of Australia in late 1 9 8 1 started just prior to the GBR coral mortalities. Those anomalies may have produced elevated water temperatures further south along the GBR. However, at this level of analysis, we are unable to substantiate a causal relationship between ENSO and the bleaching in the western Pacific. We examine the possible causes of the GBR bleaching in greater detail below. 2 . 4 Indian Ocean/Arabian Sea
In the Indian Ocean, as in other areas outside of the eastern Pacific, the signals of ENSO are weak and poorly defined (Pan and Oort, 1 9 7 8 1 . Manifestations of changes in the SO index are generally reported in terms of the monsoons, which dominate the climate in this region. Barnett ( 1 9 8 4 ) has proposed that wind anomalies, which lead to the ENSO episode, originate in the Indian Ocean. For example, a strong burst of westerly winds in October and November 1 9 8 1 triggered an early onset of the westerly monsoon, a year prior to the 1 9 8 2 - 8 3 El Niso peak warming in the eastern Pacific. This early onset of the summer monsoon has been proposed as a predictor of the ENSO event and is correlated with positive SST anomalies north of Australia. Positive SST
150
anomalies in the Indian Ocean are also correlated with the SO phenomenon [Weare, 1 9 7 9 ; Cadet and Diehl, 1 9 8 4 ) .
During ENSO, anomalous easterly winds
were observed west of Australia [Hastenrath and Wu, 1 9 8 3 ) and droughts occurred in southern India, Sri Lanka and southern Africa IRasmusson and Wallace, 1 9 8 3 . There were several reports of coral bleaching and mortality in the Arabian Sea/Indian Ocean during 1 9 8 3 . In January 1 9 8 3 bleaching and subsequent mortality were observed among corals on reefs throughout the Java Sea (Cook,
in Glynn,
1984;
Suharsono, 1 9 8 4 ; Suharsono and Kiswara, 1 9 8 4 ; Brown, 1 9 8 7 ) .
Seventy-two species were affected and mortalities of 8 0 - 9 0 % were observed by the end of May ISuharsono and Kiswara, 1 9 8 4 ) . Bleaching and mortality occurred to a depth of 1 5 m, but was highest on the reef flats ISuharsono, where in some cases coral cover was reduced from 5 0 % to less than 1 0 % in four months [Brown, 1 9 8 7 1 . Brown ( 1 9 8 7 1 and Suharsono and Kiswara 1 1 9 8 4 1 reported water temperatures greater than 29OC in Indonesian waters and greater than 3OoC around Pulau Seribu in the Java Sea. Burchard and McCain 1 1 9 8 4 ) reported extensive mortality of a variety of marine life, including several species of reef corals, in the western Arabian Gulf during March and April, 1 9 8 3 . Initial mortalities were observed in mid-March, but had ceased by late April. No abnormal environmental conditions were reported, and the mortalities have been attributed to an oil spill in the region. In the Indian Ocean, off the east African coast, coral bleaching and death occurred at Reunion Island in March, 1 9 8 3 [Guillaume et al., 1 9 8 3 ) and at Mayotte Island [Mozambique Channel, West Indian Ocean) in May and June, 1 9 8 3 IFaure et al., 1 9 8 4 ) . At Reunion Island, bleaching occurred to a depth of 20 m, affecting 20-50% of the hard and soft corals and killing up to 80% of the shallow water Acropora colonies IGuillaume et al., 1 9 8 3 ) . Death of reef flat corals was attributed to exceptionally low austral summer tides IGuillaume et al., 1 9 8 3 ) . High levels of sedimentation and elevated lagoonal water temperatures were given as the probable causes for the bleaching and death of over 2 0 species of hydrocorals and scleractinians observed at Mayotte Island [Faure et al., 1 9 8 4 ) . Sedimentation was attributed to land clearing and land development activities. Due to a lack of data on the oceanic response in the region, the linking of observed coral bleaching and mortality in the Indian Ocean/Arabian Sea to ENSO-generated environmental changes is difficult. Two of the observed coral mortalities [Mayotte Island and the Arabian Gulf) appear unrelated to the ENSO episode. However, the droughts that occurred throughout the area during 1 9 8 2 - 8 3 have been associated with increased solar insolation. This could have produced the elevated water temperatures observed in both the lagoon at 19841,
151
Mayotte and the Java Sea. Contrary to this, however, is the observation of anomalously cool water to the north of Australia during the El Nifio event [Holland and Nicholls, 1 9 8 5 ; Wright, 1 9 8 5 ) . Thus, at this time, the data to substantiate a connection with ENSO are ambiguous. Caribbean Sea In the tropical Atlantic, as in other areas outside of the eastern and central Pacific, a number oE perturbations were observed in 1 9 8 3 . Anomalous westerlies extended from the eastern Pacific, over the Caribbean and 2.5
equatorial Atlantic, surface trade winds were greater than normal, the Intertropical Convergence Zone was weakened and displaced northward and droughts occurred in NE Brazil [Hovel et al., 1 9 8 6 ) . During the 1 9 8 2 - 8 3 ENSO, anomalous upper-level westerlies in the eastern Pacific extended over the Caribbean and surface trades over the equatorial Atlantic intensified. These anomalous winds are attributed to the intense convection and ascending flow centered over the warm eastern equatorial Pacific INamias, 1 9 8 3 ; Gray, 1 9 8 5 ; Philander, 1 9 8 6 ; Horel et al., 1 9 8 6 ; Shapiro, 1 9 8 7 ) . These enhanced winds may have reduced hurricane activity in the Atlantic in 1 9 8 3 (Gray, 1 9 8 5 1 . A northward displacement of the north Atlantic storm track occurred during the 1 9 8 2 - 8 3 ENSO, which allowed for greater-than-normal insolation and lighter-than-normal winds to the south IRopelewski, 1 9 8 4 ) . This was associated with positive SST anomalies in the north Atlantic IRopelewski, In the southeast Atlantic, warm water and a deepening of the thermocline off South Africa also coincided with ENSO IShannon, 1 9 8 3 ; Tourne and Rasmusson, 1 9 8 4 ) . Horel et al. ( 1 9 8 6 1 proposed that these perturbations were part of a "remote readjustment of the tropical circulation to the unusually intense convection and rising motion over the eastern equatorial 19841.
Pacific." The occurrence of the severe droughts in northeastern Brazil ["secas"),as well as the intensity of the rainy season in the Caribbean, have been correlated with SST anomalies off the coast of Peru (Hastenrath, 1 9 7 6 , Covey and Hastenrath, 1 9 7 8 ; Kousky et al., 1 9 8 4 ) . However, the manifestation of these global scale connections remains unclear INobre et al., 1978;
1985).
During the summer of 1 9 8 3 , extensive bleaching was reported at a number of Caribbean localities. Glynn 1 1 9 8 4 ) noted bleaching at sites in the Florida Keys, Bahamas, Costa Rica and Panama. More detailed reports include that of Cortes et al. 1 1 9 8 4 1 , who reported bleaching and mortality among 1 2 species of coral on the Caribbean coast of Costa Rica. Lasker et al. 1 1 9 8 4 ) reported bleaching among 25 species of zooxanthellate anthozoans in the San Blas Islands, Panama during June and July of 1 9 8 3 with total coral mortality
152 reaching 35% of living cover on some shallow reefs. Jaap 11985) observed extensive coral bleaching in the Florida Keys in September, 1983, which included 15 species. In this area mortality was low 15-15%), and most colonies had recovered after 7 months IJaap, 1985). Increased SST has been cited as the cause for the observed bleaching and mortality in the Caribbean. Most of the events occurred during the northern hemisphere summer when water temperatures are at a maximum. In Costa Rica, sea surface temperatures as high as 35OC were reported ICortes et al., 19841. Bleaching and mortality in the San Blas Islands, Panama, was closely correlated with an elevation of subsurface sea temperatures I4 m depth) to 32°C (Lasker et a1.,1984). Subsurface temperatures in San Blas remained at or above 31°C between May 25 and June 30, 1983. Satellite SST data also show 1983 Caribbean water temperatures warmer by about one degree than in 1982. However, those observations do not always correspond to the timing of the bleaching events. Teleconnections between the eastern Pacific and Atlantic exist and ENSO events can be correlated with anomalies in the Atlantic, but at this level of analysis, there is not a specific mechanism to couple ENSO with the observed bleaching throughout the Caribbean. Because of the global nature of the warm episode, there were manifestations of El Nifio and the Southern Oscillation in each of the world's oceans where coral mortality was observed. El Nifio and its related phenomena had a direct and obvious effect on many reefs in the eastern and central Pacific. Outside of this region the ENSO signature is not well-defined. Consequently, it becomes more difficult to attribute cases of coral mortality to the El Nifio phenomenon. Similarly, it becomes more difficult to disprove an El Nifio coral mortality hypothesis. In most of the cases of coral mortality outside of the central and eastern Pacific, we can note possible El Nifio effects that may have indirectly affected the reported mortality. However, the connections between the putative El Nifio effects and the proximal causes of the mortality events have been unclear.
- GREAT BARRIER REEF In this and the following section 14) we attempt a more detailed analysis of coral mortality at two sites--the Great Barrier Reef in the western Pacific 3 DETAILED CASE STUDY
and the San Blas Islands in the western Caribbean. In the analyses we attempt to identify the immediate cause of the coral mortality. We examine environmental parameters, such as solar insolation and water temperature, that could directly cause bleaching in corals as well as wind direction that could act indirectly. Analyses are restricted to the periods for which there are
153
meteorological data and reliable data on bleaching. Finally, we discuss the extent to which these physical changes could result from the ENS0 episode. 3.1 Extent and Timing of Bleachinq
In Australia, large scale bleaching of corals and other reef coelenterates occurred in the central and northern GBR in early 1 9 8 2 . Conspicuous bleaching of corals was observed at 14 sites extending over 500 km, and encompassed a wide range of habitats, from protected inshore reefs in turbid water to exposed, wave-swept, shelf-edge reefs in clear oceanic conditions (Fig. 2 ) . The amount of bleaching varied considerably between sites. Magnetic Island and Myrmidon Reef, which lie at opposite extremes of the inshore-offshore environmental gradient, were affected the most severely. At other reefs, such as Davies and Britomart, bleaching was conspicuous but usually confined to pocilloporid corals, which constitute a small proportion of the living cover on these reefs. Bleaching was generally limited to shallow, well illuminated areas above 9 m depth (Harriott, 1 9 8 5 ; Oliver, 1 9 8 5 ) .
144O
152'
Fig. 2 . Northern Queensland, Australia and the northern Great Barrier Reef where coral bleaching and mortality were observed in early 1 9 8 2 .
154
In total, at least 25 species of scleractinian corals were either totally or partially bleached [Oliver, 1985). Of these, 8 species exhibited extensive bleaching at more than one locality lat least 40% of the colonies of each species were affected). Pocilloporid corals, i.e., species of Pocillopora, Stylophora and Seriatopora, were most commonly bleached. They were affected on the majority of reefs where bleaching occurred, were among the first to become bleached, and suffered the highest subsequent mortality IFisk and Done, 1985; Harriott, 19851. Within the family Acroporidae, plate-forming species of Montipora, and digitate and arborescent forms of Acropora were extensively affected and suffered high mortality. Other taxa exhibited more sporadic bleaching, but occasionally formed extensive areas of bleached coral on some reefs 1e.g. Goniopora, family Poritidael. The most complete chronology of events during the Australian bleaching episode was obtained for the fringing reefs around Magnetic Island, which lies approximately 3 km off Townsville, North Queensland (Oliver, 19851. Bleaching was first noticed in early January, 1982, and reached a maximum in terms of area and number of species affected during mid-February. Bleaching persisted relatively undiminished until late March, when the first signs of mortality or recovery were detected. By September, 1982 most colonies had either died or recovered. Visual estimates in these areas indicated that mortality was very high among bleached colonies in general, and that in zones dominated by Acropora and Montipora living coral cover was reduced by up to 50%. Harriott I19851 and Fisk and Done 11985) compared population mortality rates during the bleaching event to prior non-bleaching periods. At Lizard Island, Pocillopora damicornis suffered a 4 month mortality rate of 26% during the bleaching event compared to rates of 17, 18 and 0 5 during the previous four month periods IHarriott, 1985). At Myrmidon Reef, the bleaching was much more severe. Fisk and Done 119851 recorded annual mortality rates as high as 47% and 7 5 % at two shallow sites 10-1 ml compared with 11% or less in years
preceding and following 1982.
At two shallow sites on Myrmidon Reef, the
percentage cover was reduced by 54% and 82%, and in the latter case community structure shifted from a coral/coralline algae-dominated to a turf/coralline algae-dominated community. Significant and widespread bleaching of coral during the summer months has also occurred on at least one other occasion in the last 10-15 years (Oliver 1985). In 1980 bleaching was recorded at 4 different locations between Lizard Island and Townsville. Although less detailed information is available for this period, species in the family Pocilloporidae were the most commonly affected. Bleaching at Magnetic Island was not so severe as during the 1982
event; however, the large number of reports indicate that bleaching in 1980 was as extensive in geographic range as in 1982.
155 3.2
Oceanoqraphic and Meteoroloqical Data
There were no striking meteorological anomalies (cyclones, heavy rainstorms, temperature extremes) or any other conspicuous, potentially stress-inducing events during the 1982 bleaching period IFisk and Done, 1985; Oliver, 1985). Harriott 119851 showed that seawater temperature at 10 m near Lizard Island was slightly higher during the summer of 1981-82 than for the same period during two previous years. She presented meteorological data on rainfall and number of sun-hours at Townsville that indicated departures of -47% and +17% respectively from long term (34 yrl averages. Although these departures are suggestive, further analysis of the data shows that these departures were not markedly greater than those that occur in non-ENS0 years (see below]. Data on sea temperature, wind speed and direction, and solar radiation were examined for anomalies that corresponded to the known periods of coral bleaching 11979-1980 and 1981-19821. We define a significant anomaly as any case in which the values during one or both bleaching years were significantly different from all non-bleaching years. The 1981-82 bleaching event was the most severe, and this should be reflected in any analysis of environmental variables. Therefore, significant differences found for 1979-80 but not for 1981-82 are considered to be ambiguous. (il Seawater Temperatures. Seawater temperatures were taken at irregular intervals at Magnetic Island over 5 years at depths between 3 and 5 m on the fringing reef slopes at Nelly Bay (Fig. 3). In all years, summer values (December through March) reached or exceeded 31OC. More detailed historical records from shallow (1-4 ml water near Townsville Harbour indicate that the mean temperatures for January and February are 31.2O and 30.5OC respectively (Kenny, 19741. These mean temperatures were slightly higher than the 1982 temperatures at Nelly Bay (30.7O and 29.7OC respectively). These data indicate that temperatures attain potentially stressful levels every summer, and that there was no obvious anomaly during the bleaching periods. The single exceptionally low temperature in January 1981 occurred during a prolonged period I22 days1 of heavy rainfall, which caused anomalous values for several parameters (Fig. 3 1 . However, this period was not associated with any significant bleaching. Water temperatures in the midshelf area off Townsville are generally cooler than in the shallow embayed waters near Townsville and Magnetic Island (Pickard, 1977; Wolanski et al., 1981; Andrews, 1983; Andrews et al., 1984). Continuous temperature data from June 1981 through June 1982 near John Brewer reef (20 m depth1 showed that temperatures rose to 28.OoC by December 28, 1981, and then fluctuated between 28.0 and 29.3OC until January 21, 1982 (J.C. Andrews, pers. comm.1. These temperatures fall within the range of summer
156
...
32-0
x
. O O ~
30-x h
0
2 2 Q)
O
O&
28-
X
X
88' 0
O
003
0
0 0
4-
e
8
0 a30
26-.
0 0 0
Q08
a0
0
0 00 0
m o00
0
E
a . OO
0
24-
22 -
0
s"
0
0
0
0
0 0
8300 0
. 0 h 0 8
0
Fig. 3. Seawater temperatures at Magnetic Island 1979-1980 bleaching; 1981-1982 bleaching).
( 0non-bleaching
periods; X
temperatures recorded during other summers from this area IPickard, 1977; Wolanski et al., 1981; Andrews et al., 1 9 8 4 ) . Although no data on bleaching are available for midshelf reefs, the temperature records indicate that broad-scale temperature anomalies did not occur at this time. Monthly records of SSTs from ship-based and satellite observations €or the north Australian region (5-1!i0S, 120-16O0E) have been cross-checked and combined to provide mean monthly anomalies during the period 1978-1982 (data supplied by N. Nicholls, Australian Bureau of Meteorology; Fig. 41. During late 1981 there was a moderate positive anomaly, succeeded by a striking fall in temperature during all of 1982. This anomaly appears to be associated with the 1982-83 ENS0 event, but the anomaly occurred in the late winter and spring (July to September) when temperatures were low, and had virtually disappeared by December 1981. The negative anomaly did not develop until June 1982, when bleaching had already reached a maximum. lii) Solar Radiation. The frequent observations of partial bleaching on only the upper surfaces of colonies suggest that light played an important role in coral bleaching IHarriott, 1985; Oliver, 1985). Harriott ( 1 9 8 5 ) noted that the number of sun-hours during the months of the 1981-82 bleaching period was above the 34 year average. However, at Townsville similar or more extreme
157
1978
1979
1980
1981
1982
1983
Year Fig. 4. Mean monthly sea surface temperature ISST) anomalies for northern Australia 15-15OS, 120-160°El from ship based and satellite observations.
levels occurred in two non-bleaching years. The most extreme year, in terms of sun-hours, was during 1982-83 when no bleaching was observed.
A
posteriori comparisons of monthly sun-hours from different years do not reveal any consistent differences between bleaching and non-bleaching years IMann-Whitney
U-tests, p)0.051. Solar irradiation was measured daily with a pyroheliometer at Townsville airport [Fig. 5).
Although solar radiation was somewhat higher during the two
bleaching periods compared with the same months during non-bleaching years,
a
posteriori comparisons based on the Mann-Whitney U statistic [Sokal and Rohlf, 1969) indicate that bleaching years were not consistently different from nonbleaching years. Ambient levels of UV radiation can have a detrimental effect on reef biota IJokiel, 1980) and can reduce growth rates in corals IJokiel and York, 1982). Two factors that affect the amount of UV radiation reaching corals are water depth and the concentration of particulate matter IJerlov, 1950; Smith and Baker, 19791. Low tides and clear water could contribute to stressful levels of both UV and visible light. Continuous measurements of water depth at
several locations during the bleaching period do not show anomalously low tides during the bleaching period [J.C. Andrews, pers. corn.). Furthermore, during the bleaching event the lowest daily spring tides occurred at night.
158
-
W
1
78-79
79-80
80-81
81-82
82-83
December
83-84
84-85
Year
Fig. 5. Monthly means for total solar irradiation measured at Townsville, Australia.
Fisk and Done 11985) measured turbidity levels on the outer shelf at monthly intervals for a year and found the levels were lowest in January 1982. However, these data do not indicate that abnormally high water transparency was the causal factor of the bleaching because bleaching was already extensive on most reefs during January (Oliver, 1985). liii) Wind speed and direction. Anomalies in the wind may serve as an indicator of extremes in other more directly stressful factors. Fluctuations in the wind speed and direction can cause displacement of water across the shelf, and alter temperature patterns within the GBR lagoon (Andrews, 19831.
Similarly, turbidity at near shore stations is strongly correlated with wind speed (Walker, 1980). Thus, consistently light winds may indicate periods of low turbidity and potentially high subsurface light levels. Daily 9 AM wind speed and direction at Cape Cleveland (1978-1985) were converted into NE (long-shore) and SE (cross-shelf) component vectors. The monthly means did not show any major anomalies during the bleaching periods IFig. 6 1 .
Although
there was significant between year variation in the S.E. wind vector in each of the three months and in the N.E. wind vector during January (ANOVA and SNK tests; all values p(0.05), there were years that showed greater deviations from the mean than did 1981-82. A similar analysis of the total daily windrun lthe sum of velocities in each octant over 24h) at Cape Cleveland from 1979 to
159
S.E. Wind Vector
February
-
61
N.E. Wind Vector
f
E i 5
U
4
P) P)
P 0 U
s2 0 78-79
79-80 80-81
81-82 82-83 83-84
84-85
85-86
Year
Fig. 6. Monthly mean SE (cross-shelf) and NE (long-shelf)component vector (speed and direction) for winds measured at Cape Cleveland, Australia, 1 9 7 8 - 1 985.
1986
also failed to indicate any significant anomaly during the bleaching
period. livl Rainfall. Increased rainfall can cause stress in reef areas by lowering salinity and by increasing turbidity and sedimentation through river discharge in near-shore areas, while decreased rainfall can lead to higher near-shore salinities and water clarity. Monthly rainfall records suggest that rainfall during 1 9 7 9 - 8 0 and 1 9 8 1 - 8 2 was lower than in most other years (Fig. 7 ) . However, it is difficult to relate low onshore rainfall with coral stress on offshore reefs such as Myrmidon, where salinity is buffered by adjacent oceanic water, and turbidity is low at all times. (v) Interactive Effects. Although no major anomalies could be found in any of the major environmental parameters considered singly, it is possible that near stressful levels of two or more parameters acting in concert, with interactive effects, may have caused the observed bleaching and mortality.
160
800 h
v
December January February
600
0 78-79 79-80 80-81 81-82 82-83 83-84 84-85 85-86
Year Fig. 7 . Monthly rainfall for the central Great Barrier Reef region.
Temperature and light are known to act synergistically in corals to cause stress and to impede subsequent recovery (Coles and Jokiel, 1 9 7 8 ) . A simple index of the possible synergistic effects of temperature, light and windrun (which can affect both of the former) was calculated by multiplying these three parameters together. Values were slightly higher than in most non-bleaching years, but there were no obvious anomalies during the bleaching periods. The bleaching years were never significantly different from all lor most) other years (p)0.05, SNK test), although there were significant dif€erences between years (p<0.05,ANOVA). To summarize, there is no clear evidence from the limited oceanographic, and more detailed meteorological data, that any major anomalies occurred in the Townsville area during either the 1 9 7 9 - 8 0 or 1 9 8 1 - 8 2 bleaching events. Oliver ( 1 9 8 5 ) presented data on coral branch extension rates at Magnetic Island that show a reduction in growth during each summer period ( 1 9 7 9 - 1 9 8 1 ) . If reduction in growth rate is an indicator of stress, then his results suggest that corals are sublethally stressed every summer. In this condition, it may take only a slight, perhaps statistically undetectable, increase in the level of one or more parameters to cause bleaching and death. Light and temperature both reached high levels during the bleaching period. In addition, the pattern of bleaching on the upper surfaces, the restriction of large-scale bleaching to shallow waters, and the similarity of this phenomenon with the laboratory observations of Jokiel and Coles ( 1 9 7 7 1 , all point to
161
light and temperature as the likely stress leading to bleaching. As pointed out by Harriott ( 1 9 8 5 1 , and Fisk and Done 1 1 9 8 5 1 , the 1 9 8 1 - 8 2 bleaching period was characterized by high temperatures, high light levels, low rainfall, and light winds, but in almost all cases, more extreme levels of these parameters also occurred during other years when bleaching was not observed. Alternatively, some other variable such as disease, or pollution may have provided the additional stress that caused the bleaching and mortality. Unfortunately, there is not sufficient information available on such factors to evaluate their possible involvement. 3.3
Relationship to El NiEo
As noted above, meteorological and oceanographic anomalies do occur in the Australian region up to one year before the major ENSO events manifest themselves, but the timing of these anomalies (rainfall, SST and wind) do not coincide with coral bleaching. The decreased rainfall, characteristic of El NiEo years in Australia, occurred one year after the 1 9 8 2 bleaching and the development of anomalous westerly winds occurred 6 months after the bleaching. The positive SST anomalies that occurred off northern Australia prior to El
Nice had already reversed by the time of bleaching on the Great Barrier Reef [Fig. 4 ) . Thus, despite the likelihood that both oceanographic and meteorological anomalies will occur in the GBR region during an ENSO event, the lack of proximate cause of the Australian bleaching event, the inadequacy of local records, and the lack of synchrony between bleaching and the onset of the major environmental perturbations associated with the 1 9 8 2 - 8 3 ENSO make it difficult to establish the existence or mechanism of a causal link between ENSO and coral bleaching on the GBR. However, it is noteworthy that both bleaching events in Australia occurred prior to an ENSO-related phenomenon. The 1 9 7 9 - 8 0 bleaching event occurred prior to a minor "abortive" ENSO in the western Pacific, which failed to develop in the east [Donguy et al., 1 9 8 2 ) . This interesting coincidence in the two phenomena should be examined in more detail during future episodes. DETAILED CASE STUDY - SAN BLAS ISLANDS, PANAMA 4 . 1 Extent and Timinq of Bleachin3 The first cases of bleaching in the San Blas Islands were observed in early 4
June 1 9 8 3 , and by June 1 5 , 1 9 8 3 many colonies of Millepora spp. and Aqaricia spp. had bleached. Colonies of an additional 2 3 species showed signs of bleaching in the following month and included scleractinians, anemones, gorgonians, and hydrocorals. The most severe bleaching occurred on shallow reef flats, but bleaching was observed down to 2 0 m as well. As on the Great
162
Barrier Reef, the timing and pattern of bleaching varied among taxa. Millepora and Aqaricia species were among the first to exhibit bleaching, with bleaching first appearing at the growing tip of the colony and then progressing toward the base. Among other scleractinian corals (i.e., Montastrea) bleaching first appeared at the base of the colony and then spread toward the top of the colony, while in others only scattered small patches were bleached li.e., Siderastrea). In many cases the affected coelenterates recovered from the bleaching, but in some, bleached sections of the colony died. The timing of recovery or mortality was again species specific. Millepora spp. showed signs of recovery by early July, whereas bleaching of Montastrea annularis colonies continued to worsen through mid-August. The extent of the initial bleaching was not an indicator of the extent of mortality. Millepora spp., which were among the most severely bleached species, recovered during subsequent months. However, Aqaricia aqaricites and A . tenuifolia, which were common members of the shallow fore- and back reefs of the San Blas, suffered 53% mortality between June 1983 and January 1984. The species-specific nature of the mortality also led to a wide range of effects on different reefs. Reefs with exceptionally high Aqaricia cover exhibited up to 100% bleaching and mortality, whereas on more "typical" shallow reefs, living coral cover decreased by 20-35%.
Reefs with few Aqaricia colonies suffered relatively
little mortality. A more complete account of the bleaching and mortality is presented in Lasker et a l . (19841. 4.2 Oceanoqraphic and Meteoroloqical Data
The 1983 bleaching event in Panama corresponded closely with increased subsurface water temperatures (Lasker et al., 19841. Subsurface sea temperatures ( 4 m depth) were measured on alternate days in the San Blas Islands during May 1 1 to August 26, 1983 and again during July 4 to August 8, 1984 and June 4 to July 31, 1986 (Fig. 8 ) . Temperatures at 4 m depth rose above 3OoC approximately 3 weeks prior to the first observations of bleaching,
and rose further to 32OC one week prior to the event. The most severe bleaching was observed on shallow reef flats where temperatures during some Elevated temperatures ()30°C) were not observed in low tides reached 34'C. 1984 or 1986 nor were there other reports of coral bleaching during 1981-1986 (excluding 19831. As noted above, thermal stress is a well documented cause of coral bleaching. Thus, we conclude that elevated temperatures caused the observed bleaching and predicate subsequent analysis on this assumption. In attempting to identify the environmental conditions that led to this warming, we will consider two general explanations for the unusual temperatures observed in 1983: large scale warming of waters throughout the western
163
1983 1984 X 1986
+
.. .... . a
. .. ..
0
30 --
0
. .
. . . ... . . 0
.
0 .
b
xx
251
'
10
I
I
20
30
MaY
+ tx+xx* t* + *+f xx
x x x x x x x x x x x x x x x xX+ ~ ~ * ; + t
I
I
I
10
I
20
30
10
June
4-
1
20
July
1
30
I
10 August
Fig. 8. Subsurface ( 4 m depth1 temperatures measured on a coral reef in the San Blas Islands, Panama.
Caribbean and/or unusually great warming of coastal waters in the San Blas region. (il Large-scale warming of offshore waters. Sea surface temperatures for the San Blas region for 1982 were obtained from the monthly maps published by Glynn 119841, and for 1983 and 1984 from biweekly maps of satellite-measured sea surface temperature (MCSST charts produced by NOAA, NESDISI. During May-August, 1982, SST remained below 29OC. In contrast, during May-August, 1983, SST rose above 29OC during seven weeks and averaged 28.3'C ls.d.=0.8). During May-August, 1984, SSTs fluctuated between 27O and 29OC with only a single observation of 29OC and a mean of 27.7OC (s.d. = 0.6). The SST data suggest an overall pattern of warmer waters in the summer of 1983 than in 1984 [Fig. 9). However, the onset of the warmer SSTs in 1983 did not occur until late June, well after the San Blas bleaching event had started. This lack of correlation may be in part an artifact of the low l0C resolution of the SST charts, but the 1-2O offshore warming alone is not sufficient to account for the 32OC temperatures observed in the San Blas area. Nearshore temperatures in 1984 were approximately l o higher than the observed offshore values, presumably due to local warming. To raise the 1983 offshore temperatures to the values observed nearshore in the San Blas, onshore warming must have elevated temperatures by as much as 5 O . This suggests that while there may have been a large scale warming of Caribbean waters in 1983, additional
30r L
28 -
........
26 24
22
t-
-1984
...... 1983
-
Fig. 9. Offshore SST for Panama region from satellite observations.
warming occurred nearshore in the San Blas during 1983. This warming may have been due to local processes leading to a warming of coastal waters. (ii) Coastal Warming. Unusually great coastal warming could occur through any combination of the following mechanisms: decreased wind speeds or changes in wind direction that caused increases in residence times of shallow waters subjected to solar warming; decreased cloud cover that resulted in increased solar insolation, and altered rainfall that affected salinity and thereby water column stability. To determine whether the unusual temperatures observed in 1983 could be attributed to unusual onshore warming in the San Blas, we examined wind direction and speed, rainfall, and salinity to detect differences between 1983, and non-El Nifio years. Since data on all of the variables were not collected at any single site, it was necessary to utilize meteorological data from a number of different sites within and near the San Blas Islands. Firstly, we concentrate on a comparison between 1983 and 1984, the summers for which we have temperature data from the San Blas. Secondly, we compare 1983 to other years to determine if 1983-84 differences can be correlated with El Nifio. (iii) Solar Radiation. Solar radiation was measured at Barro Colorado Island between 1975 and 1986 (D. Windsor, Smithsonian Environmental Monitoring Program, unpubl. data). This inland site (Lake Gatun, Panama Canal) is the
165
furthest from the San Blas of our reporting sites (approximately 125 km) and the values may differ in magnitude from San Blas radiation levels. However, they should be representative of yearly variation. June 1983 had the highest levels of insolation recorded during June of the 8 year data set, but May 1983, the month in which seawater temperatures first rose, did not have especially high levels of insolation (Fig. 10). Thus, in as far as Barro
Colorado Island insolation is representative of the San Blas Islands, solar insolation alone did not account for the observed seawater temperatures. (ivl Wind speed and direction. Hourly wind speed and direction at Galeta Island, a reef flat 100 km west of the San Blas Islands, were obtained from the Smithsonian Institution Environmental Monitoring Program [Cubit et al., 1988).
Wind data during 1983 were reported to the nearest 45O arc whereas
those from 1984 were reported to the nearest 0.lo.
In subsequent analyses,
the 1984 data were first rounded to the nearest octant in order to apply uniform rounding error to all of the data sets. Analysis of wind data for the summers of 1983 and 1984 suggests several differences between the years in both the strength of the winds and their primary direction. Windrun, calculated by summing the velocities observed in each octant over 24 hours and averaging the results over 7 day periods, shows
.
h
E
6000 C
.-0 4-
.-m
-
1
i
! X
k I
U
2
Y
4000-
i
;;I 0 cn 2000.m n -
0'
X
'
J
I
F
I
M
I
A
I
M
I
J
X
!
I
J
8 I
I
X
I
8
X
.
f
I
I
I
I
I
A
S
O
N
D
Month Fig. 10. Average daily solar radiation at Barro Colorado Island, Panama, 1975-1986. Each point is a monthly average X, 1983; 0 , 1975-1986, excluding 1983 [data for January and February 1983 were not available).
166
that during May-August 1 9 8 3 , the El NiEo year, the dominant wind direction was from the north during 1 3 of the 1 8 weeks and was out of the northeast in only 1 of 1 8 weeks [Fig. 1 1 ) . In 1 9 8 4 , the dominant wind direction was from the northeast during 1 2 of the 1 8 weeks. In 1 9 8 3 , 8 of 1 8 weeks had average total windruns of 2 4 0 km/day or less [i.e., a mean wind speed of 1 0 km/h). However, winds were even calmer in 1 9 8 4 when there were 1 5 weeks with average daily windruns of less than 2 4 0 km/day. Similarly whereas there were 1 5 days with net wind speeds of less than 5 km/h in 1 9 8 3 , there were 75 such days in 1 9 8 4 . Average wind speeds were significantly different between the two years in both July and August (Table 3 ) .
In both those months, 1 9 8 4 wind speeds were lower
than those observed in 1 9 8 3 . Although there were differences in wind direction and speed between 1 9 8 3 and 1 9 8 4 , there was no correlation between the timing of the differences and the rise in sea temperature. Elevated temperatures were observed in 1 9 8 3
600
1
. .. ..... .. North
-Northeast
Months Fig. 1 1 . Windrun from N and NE octants at Galeta Island, Panama for 1 9 8 3 and 1 9 8 4 . Area under N curve represents windrun, area between N and NE curves represents NE windrun, and area between total and NE curve represents wind-run from the remaining octants.
167
between May 19 and July 25. However, the pattern of northerly winds observed in 1983 developed during the beginning of the week of March 14, 1983, well in advance of the elevation in sea surface temperatures and the observed bleaching. As can be seen in Figure 1 1 , the period of elevated temperatures was characterized by relatively low wind speeds. The week of May 16, 1983 marked the start of the warming phase and was a particularly calm period. After a week of slightly stronger winds, 7 weeks of low windruns followed, which closely coincided with the period of warm water. While reduced winds and the resultant reduction in water transport could have produced local warming, the pattern of wind speeds in 1984 was remarkably similar with even calmer winds, but no warming resulted. (v1 Rainfall and Salinity. The reduced wind speeds alone do not explain the disparate temperatures observed in May-June of 1983 and 1984. However, reduced water movement in concert with changes in runoff could elevate sea temperatures. Rainfall was measured in the San Blas between May 9 and August 27, 1983 and monitored throughout 1974-1985 at Galeta Island.
TABLE 3 Averages (and standard errors1 of net wind direction and speed' at Galeta Island, Panama; North=Oo. DIRECTION
MONTH
1983
APRIL * * * MAY ** JUNE * JULY AUGUST
354.5 331.0 251.8 327.5 305.1
1984 (4.0) 19.71 I171 (111 1101
15.5 10.1 310.1 313.5 329.1
(3.51 (111 (141
(151 (121
SPEED (KM/HI MONTH
983
APRIL
6.6 2.0 4.0 9.1 9.5
MAY JUNE JULY * * * AUGUST * * *
984 1.01 1.71 0.41 1.21 1.01
4.0 8.2 3.7 3.0 4.0
(0.91 (1.31 (0.81 10.71
(1.01
* 1983 vs 1984, t-test, P(O.05 * * 1983 vs 1984, t-test, P(O.01
* * * 1983 vs 1984, t-test, P(O.001 'Daily net wind direction and speed was determined by vector addition of hourly winds.
168
The San Blas rainfall data from 1983 illustrate the effect of freshwater runoff on local sea temperature. May, June and most of July, 1983 were characterized by moderate rainfall (Fig. 121. In late July, rainfall increased with accumulations of 55.0 and 141.5 mm measured on July 28 and 30 respectively. Those rains were associated with the appearance of fresh water lenses on the sea surface, a drop in sea temperature from 31.5' to 29.OoC, and a drop in salinity from 34.5 O/oo to 32.0 O / o o . Several days after the rains, salinity rose to 33.4 O/oo and temperatures again rose to 29.5'C. Thereafter salinity remained below 34.0 O/oo and sea temperature remained at or below 29.5OC. Thus, the heavy July rains and associated runoff coincided with the return of San Blas water temperatures to below 30'C. Although we do not have rainfall data for the San Blas from other years, the data from Galeta Island can be used as an index of the relative rainfall for different months and years. The Galeta data do not suggest any strong
150r
36 rr
I
I
l
I
I
I
I
I
I
I
1
10 20 30 10 20 30 10 20 30 10 20 30 May June July AUQ Fig. 12. Rainfall, salinity, and sea surface temperature measured in the San Blas Islands, Panama for May-August, 1983.
169 difference in rainfall between 1983 and 1984 (Cubit et al,, 19881. Salinities followed the opposite trend from that expected by unchanged rainfall. Salinities measured in the San Blas between May 1 1 and August 2 7 , 1983, were lower than those observed in July 1984 IMann-Whitney U Test, p(O.0011. This same trend is present in measurements made daily from Galeta (Cubit et al., 1988). The lower surface salinities suggest greater runoff that could have resulted in lowered temperatures, but reduced surface salinity could also have created greater stability in the water column that could enhance the effects of solar heating. In summary, there is no evidence from the 1983-1984 solar radiation, wind, rainfall or salinity data to suggest that local meteorological conditions could have caused coastal warming in 1983 that would have been greater than in 1984. However, the lack of detailed meteorological and oceanographic data from the San Blas, the actual site of the bleaching, as well as our ignorance of local currents and basin dynamics, prevent a definitive conclusion. At this time we can only speculate that the warming may have been generated by the combined effects of the greater water column stability created by reduced salinities and perhaps slight changes in circulation caused by the wind shift. 4.3 Relationship to El Niao The available data from the San Blas Islands and nearby sites show the presence of three differences between the sununers of 1983 and 1984, with 1983 having: ( 1 1 more northerly winds, I 2 1 reduced salinities, I31 slightly warmer offshore waters. We have speculated that these events resulted in the raised seawater temperatures and the subsequent bleaching, although the exact mechanism is still uncertain. Our lack of long-term sea temperature data makes it unclear whether the warmer waters of the Caribbean in 1983 can be interpreted as an El Niao effect. However, analysis of wind and salinity data from Galeta suggest that the 1983 "anomalies" fall well within the year to year variability observed in non-El NiEo years. Multivariate analysis of variance [simultaneous analysis of all 8 octants1 of 1975-1985 windruns revealed significant month by year (FZ97; df=840,3146; p(O.001 1, month lFg14.13; df=88,3146; ~(0.0011 and year effects (Fg14.55; df=80,3146; p(O.0011. Yearly values differed from each other but the pattern of between year variation was dependent on which month was being analysed. Univariate analysis of windruns from each of the octants for June 1975-1985 indicated significant variation in winds from all directions except the northeast (all other cases p(0.05 or lower). However, the 1983 data did not distinguish themselves by being significantly different from the mean and in most cases the 1983 data were very close to the 1 0 year mean (Fig. 131. Although 1983 was notably different from 1982 and 1984, the years preceding and following El
170
NiKo, the 1983 windrun data are not statistically distinguishable from many of the non-El Nifio years. This same trend is present in 1977-1985 salinity data from Galeta [analysis of monthly averages). There was significant variation between months (F=13.68;df=11,88; p(O.001) and years lF=2.23; df=8,88; p(O.OOl), but 1983 was close to an average year. Analysis of the San Blas bleaching event illustrates the difficulties of retrospective analyses. Relevant data are difficult to find or unavailable and often one must use data from sites more distant than desirable. Although we remain convinced that the 1983 bleaching was caused by unusually warm
North
75
76
75
76 77
77
78
79
80
81
82
83 84 85
600 h
m
9 E 400 Y K 3 L -0 K
5
200
0
78 79 80 a1 82 a3 84 85 Year
Fig. 1 3 . Ten year 11975-19851 windruns from N and NE octants measured at Galeta Island, Panama.
171
waters, we cannot fully explain the events that led to those elevated water temperatures. Offshore warming may have played some role, and we can look for an El Nizo causation in that event. However, we have been unable to relate any explanation of the unusually great onshore warming in 1 9 8 3 to El Nizo. 5
CONCLUSIONS The relationship between the coral mortality observed throughout the
world's oceans and the 1 9 8 2 - 8 3 ENSO is problematic. Assuming for the moment that these events were ENSO-related, we may ask whether they will have longterm effects on the world's coral reefs. Since mortality on reefs is not restricted to ENSO years, the importance of ENSO events is dependent on the presence of mortality that is qualitatively or quantitatively different from mortality during non-ENS0 years. In any year there is a "baseline" mortality attributable to continuous, non-catastrophic processes such as competitive overgrowth, grazing, disease and bioerosion/physical disturbance. Non-catastrophic mortality has been measured at several locations. Connell 1 1 9 7 3 ) and Hughes and Connell 1 1 9 8 7 ) report annual mortality rates that range from 0% to 3 0 % for an Australian reef flat and Fisk and Done 1 1 9 8 5 ) report Great Barrier Reef slope mortality rates of 1-17%. In the Caribbean, mortality rates on reef slopes averaged 26% at Curacao (Bak and Luckhurst, 1 9 8 0 1 and as high as 60% among Agaricia spp. at deeper sites in Jamaica (Hughes and Jackson, 1 9 8 0 1 . Excluding the eastern Pacific, coral mortality during the 1 9 8 2 - 8 3 ENSO ranged from an overall mortality of 5 - 1 5 % on Florida reefs (Jaap, 1 9 8 5 ) to 5 0 - 1 0 0 % at Takapoto (Laboute, 1 9 8 5 ; Harmelin-Vivien and Laboute, 1 9 8 6 1 . At many sites the level of mortality during the 1 9 8 2 - 8 3 ENSO was well above baseline, noncatastrophic mortality rates. Catastrophic storms and low tides also occur during non-ENS0 years and have caused high mortality that is comparable to that observed in 1 9 8 2 - 8 3 . Hurricanes are regular events that dramatically affect reefs on a scale of tens to hundreds of years (Moorhouse, 1 9 3 6 ; Stephenson et al., 1 9 5 8 ; Stoddart, 1 9 6 3 ; Glynn et al., 1 9 6 4 ; Ball et al., 1 9 6 7 ; Baines et al., 1 9 7 4 ; Woodley et al., 1 9 8 1 ) . For instance, in 1 9 8 0 Hurricane Allen effected mortality in Jamaican corals that ranged from 2 % in Aqaricia spp. at 1 2 m to 9 9 % in Acropora spp. at 6 m (Woodley et al., 1 9 8 1 1 . The wave damage and sediment scour created by hurricanes affect corals most severely in shallow waters and disproportionately affect branching species (Woodley et al., 1 9 8 1 1 , but at some sites rubble "avalanches" cause nearly 1 0 0 % mortality at depths as great as 40 m (Harmelin-Vivien and Laboute, 1 9 8 6 ) . Mass mortalities associated with temperature stress occur more regularly than hurricanes and are among the most common of the catastrophic events
172
affecting reef corals. Mass mortalities of this type most commonly occur when extreme low tides and high solar insolation raise temperatures of reef flat waters. Mortality in these habitats can reach 8 0 - 9 0 % (Glynn, 1 9 6 8 ; Loya, 1 9 7 6 1 . Thermal stress is most likely to occur in shallow waters and can be species specific. For example, a distinct taxonomic trend in likelihood to
bleach occurred on Okinawan reefs in 1 9 8 0 in a bleaching event attributed to elevated water temperatures (Yamazato, 1 9 8 1 ) .
On those reefs, Seriatopora
sp. followed by Stylophora pistillata, Pocillopora damicornis and Montipora foliosa were the most commonly observed bleached corals. As in the case of storm-related damage, the patterns of mortality in these non-ENS0 year events are indistinguishable from the 1 9 8 2 - 8 3 events. In comparing the "ENSO-related" mortalities to other catastrophic mortality events, it is obvious that in most cases the 1 9 8 2 - 8 3 mortalities did not involve unique causes of mortality. The physical disturbance associated with hurricanes, temperature stresses created by extreme low tides, and low salinities associated with heavy rains observed in non-ENS0 years are similar to those that occurred during the 1 9 8 2 - 8 3 ENSO episode. Therefore, the importance of the ENSO event is dependent on the magnitude, distribution and frequency of the "ENSO-induced'' effects relative to other catastrophic events. As we have noted, hurricanes are regular sources of mortality on many
reefs. However, cyclones rarely occur in French Polynesia, the last major cyclones having occurred in 1 9 0 3 - 1 9 0 6 (Harmelin-Vivien and Laboute, 1 9 8 6 1 . The extraordinary frequency of cyclones in 1 9 8 2 - 8 3 has been attributed to the ENSO. Thus, in French Polynesia the ENSO generated large scale mortality unlike that which had occurred for over 7 0 years. Given the degree of mortality associated with hurricanes and recovery times of approximately 50 years (Pearson, 1 9 8 1 ; Harmelin-Vivien and Laboute, 1 9 8 6 1 , we must conclude that the 1 9 8 2 - 8 3 ENSO had a major long-term eEfect on some French Polynesian reefs. If ENSO-related storm track changes are responsible for many of the severe hurricanes affecting French Polynesia we might further speculate that ENSO-related events are a major force in controlling the structure of these reef communities. The bleaching observed in the San Blas Islands, Panama, resulted in mortality similar to that observed elsewhere in bleaching events during non-ENS0 years. This suggests that the 1 9 8 2 - 8 3 "ENSO-related" mortalities were equivalent to other causes of mortality on Caribbean reefs such as "normal" bleaching events and storms. This is probably also true for the effects of the ENSO in much of the Caribbean. However, in reviewing its significance on reefs in the San Blas we must consider that the 1 9 8 2 - 8 3 mortality occurred in
a
region not frequently affected by large scale physical
disturbances. Our limited historical knowledge of the area indicates that
173 bleaching events are rare. More importantly, Panama lies well to the west and south of the storm track for Caribbean hurricanes IGlynn, 1 9 7 3 ) . Thus, an event such as the 1983 bleaching may play an important role in the dynamics of shallow Panamanian reefs. As at other bleaching localities, the bleaching event on the Great Barrier Reef was highly visible and at some sites led to mortality similar to that caused by cyclones and Acanthaster planci outbreaks. As reported from other Indo-Pacific bleaching events, pocilloporid corals, along with certain acroporids (Acropora spp. and Montipora spp.), were most affected. These are among the many species affected by
A.
planci outbreaks, an important
corallivore on the GBR lMoran, 1986). Thus the effects of the 1982 bleaching were not unique, but on some reefs bleaching and subsequent mortality may have been an important mechanism for opening space on the reef. The frequency of ENSO-related bleaching relative to storms and A. planci outbreaks will therefore determine the significance of bleaching events for GBR reef dynamics. Perhaps the most unique of the 1982-83 mortality events was generated by sea level lowering in the central Pacific. In some cases, mortality was attributed to thermal stress that occurred on reef flats where water circulation was restricted. At other sites, reef flats became completely exposed. These events generated nearly 100% mortality on some reef flats. In "normal" years, reef flats are areas of relatively high turnover (Connell, 19731, and recovery rates from similarly high mortality events such as
Acanthaster planci outbreaks range from 10 to 15 years (Pearson, 1981). The high turnover and relatively rapid recovery of reef flat communities suggest that the 1982-83 ENSO, though highly visible, probably did not cause changes very different in intensity than those that occur on reef flats on the order of tens of years. When viewed on a world-wide basis, it must be acknowledged that at no other time in recorded history has coral bleaching and mortality occurred on such a scale. The extent of these mortalities marks 1982-83 as a significant period for the world's coral reefs. However, the 1982-83 coral mortality varied in magnitude and importance depending on location. In areas such as the central Pacific, mortalities caused unusual and significant changes in reef communities, while in areas such as the GBR, the magnitude and frequency of the mortality was comparable to mortality events that occur more often than severe ENSO episodes. Can the 1982-83 mortality event be attributed to the 1982-83 ENSO? The answer to this question is limited by our ability to identify the proximal cause of the mortality and then relate it to larger scale ENSO phenomena. An alternative to the detailed analysis of events in 1982-83 is to search for
174
correlations between previous coral mortalities and previous El Ni%o events. There have been numerous reports of mass mortalities among reef corals. These are generally attributable to tropical storms (both mechanical damage and physiological stress), low tide exposures or unseasonably low water temperatures (Table 4 ) . Of 29 mass mortalities reported from 1876-1980, 10 occurred during moderate to strong El NiEo events, as identified by Quinn et al. 11978) and Quinn et al. (1987). In only one case was a causal relationship with ENS0 discussed (1972--Yamaguchi, 1975).
TABLE 4 Historical occurrences of mass coral mortalities Year Disturbance 1876
Location
"dark water" Cocos Keeling
1878* "dark water" Dry Tortugas 1906 cyclone
Observed Mortalities
Reference
lagoon corals destroyed Stoddart, 1969 corals killed
Tikehau, F.P. not recorded, believed to be extensive
Stoddart, 1969 Harmelin-Vivien & Laboute, 1986
1918* heavy rains/ Queensland low tides coast, Australia
coral death
Stoddart, 1969
1920 heavy rains/ Samoa low tides
"1000's of corals killed"
Mayor, 1924
1920 heavy rains/ Tutuila low tides
coral mortality
Mayor, 1924
1934 cyclone
Low Isles, GBR
branching corals destroyed
Moorhouse, 1936; Stoddart, 1969
1945 heavy rains
Low Isles, GBR
coral mortality
Stephenson et al., 1958
1950 cyclone
Low Isles GBR
physical destruction
Stephenson et al., 1958
1951* cyclone/ heavy rains
Coral Point, Sarina, Austra1ia
coral mortality
Stephenson et al., 1958
* Years of El NiEo-type events (moderate-strong) as classified by Quinn et al., 1978; Quinn et al., 1987
175
TABLE 4 (continued) Historical occurrences of mass coral mortalities Year Disturbance
Location
Observed Mortalities
Reference
1956 heavy rains
Peel Island, Australia
10-70% coral mortality
SlackSmith, 1959
1956 heavy rains
Jamaica
mass mortality-shallow marine organisms
Goodbody, 1961
1958* heavy rains
Jamaica
mass mortality-shallow marine organisms
Goodbody, Ball et al., 1967
1961
1960 hurricane
Florida Keys
branching corals destroyed
1961
Belize
overall reef damage: Stoddart, 75-805 1963 A. cervicornis-100% -A. palmata - 80% -M. annularis - 50% mortality
1963 hurricane
Puerto Rico
windward reefs-)50-100% inner reefs-10-50% mortality
Glynn et al., 1964
1963 hurricane/ heavy rains
Jamaica
bleaching
Goreau, 1964
1964 decreased water temp.
Persian Gulf
mortality of Acropora
Shinn, 1976
1965* hurricane
Florida Keys
branching corals destroyed
Perkins & Enos, 1968
1965* hurricane
Fiji
mortality to shallow water corals
Cooper, 1965
1965* heavy rains
Kaneohe Bay, Hawaii
mortality of all corals Banner, in 1-2 meters depth 1968
1967 cyclone
Heron Island, extensive damage to Australia some areas
Pearson, 1981
1969 decreased temperature
Florida Keys
80-90% mortality
Hudson et al., 1976
80-90s; mortality of reef flat corals
Loya, 1972; Fishelson, 1973
hurricane
Red Sea 1970 low water/ aerial exposure 1971 red tide
Gulf of Mexico coral mortality
Pearson, 1981
* Years of El Nifio-type events (moderate-strong) as classified by Quinn et al., 1978; Quinn et al., 1987
176
TABLE 4 (continued) Historical occurrences of mass coral mortalities Year Disturbance
Location
Observed Mortalities
Reference
Heron Island, extensive in some Australia areas
Pearson,
aerial exposure
Guam
high mortality of reef flat organisms
Yamaguchi,
1976*
decreased temperature
Dry Tortugas
9 6 % mortality of shallow water corals
Porter et al., 1 9 8 2
1980
hurricane
Jamaica
cover reduced to 4 0 % of 1 9 7 7 value, near
Woodley et al.,
1972*
1972*
cyclone
total destruction of
1981
1975
1981
Acropora spp,
* Years of El Nifio-type events [moderate-strong) as classified by Quinn et al., 1 9 7 8 ; Quinn et al., 1 9 8 7
1982-83
witnessed the most widespread coral bleaching and mortality in
recorded history. At this time we can link ENSO to coral mortality only in the central Pacific (excluding the close connection in the eastern Pacific, see Glynn, this volume), and we are unable to conclude that the 1 9 8 2 - 8 3 E l Niso event had a world-wide effect on coral reefs. Several factors contribute to our inability to link conclusively the 1 9 8 2 - 8 3 coral bleaching to the ENSO event. Firstly, the actual cause of coral mortality has been identified at only a few sites. Because corals bleach in response to many sub-lethal stresses, the perturbation (or combination of perturbations) may not have been 'anomalous" enough in either magnitude and/or duration to be detected over the noise of year to year variability. Secondly, even when the cause of mortality has been identified, the factors that lead to the environmental perturbation have not been determined. This is often due to a paucity of the appropriate data. Since extreme events, such as the 1 9 8 2 - 8 3 coral mortality, are seldom predictable, the collection of data prior to or concurrent with the event often is fortuitous. Finally, while it is accepted that ENSO is an interaction between the ocean and atmosphere, the mechanism and nature of these interactions on a global scale, especially far removed from the strong signals of the eastern Pacific, are still largely theoretical. As the mechanisms and teleconnections of the ENSO phenomenon become better established, causal relationships to such coral mortalities may become apparent. In the interim, we can only weakly correlate the 1 9 8 2 - 8 3 ecological
177
anomalies of the world's coral reefs with meteorological/oceanographic anomalies that are themselves correlated to the 1 9 8 2 - 8 3 ENSO. As a final note, in 1 9 8 7 , bleaching was again recorded on the GBR and throughout the Caribbean. Although detailed observations have not been collated, it appeared that the bleaching on the GBR was not so severe as the 1 9 8 2 event, but was at least as extensive (J. Oliver, pers. obs.). Bleaching on the GBR was first observed in April, 1 9 8 7 , and again the corals most severely and frequently affected were species in the family Pocilloporidae IJ. Oliver, pers. obs). The 1 9 8 7 bleaching in the Caribbean was much more extensive than that observed there in 1 9 8 3 .
Bleaching was first reported from
the northern coast of Colombia and the Florida Keys during mid-July, 1 9 8 7 , with subsequent bleaching observed in Jamaica, southern Florida, Puerto Rico, the Bahamas, and St. Croix and St. John, USVI (Williams et al., 1 9 8 7 ; Roberts, 1 9 8 7 ; Ogden and Wicklund, 1 9 8 8 ) . Again scleractinians, gorgonians, zoanthids, actinians and hydrocorals, as well as sponges with zooxanthellae were
affected. Early speculation is that elevated water temperatures is a likely cause of the extensive bleaching (Roberts, 1 9 8 7 ; Ogden and Wicklund, 1 9 8 8 ) . This bleaching coincides with a weak/moderate ENSO event in 1 9 8 6 - 8 7 (Kerr, 1 9 8 7 ; Quinn et al., 1 9 8 7 ) . As in many past events, the cause of the bleaching is unknown and it is unclear whether these two events are causally related to ENSO. 6
ACKNOWLEDGEMENTS
We thank J. Cubit, H. Caffey, and D. Windsor for providing Galeta and BCI data and J.C. Andrews and N. Nicholls for Australian oceanographic data. We are grateful to D. Olson for discussions on the physical oceanographic aspects of ENSO and to I. Downs for help with French translations. We also thank T. Done, M. McGowan, D. Olson, G. Podesta, L. Shapiro, and A. Szmant for comments on the manuscript and J. Stamos for preparing the figures. 7
REFERENCES
Acevedo, R. and Goenaga, C., 1 9 8 6 . Note on coral bleaching after a chronic flooding in southwestern Puerto Rico. Carib. J. Sci., 2 2 : 2 2 5 . Andrews, J.C., 1 9 8 3 . Water masses, nutrient levels and seasonal drift on the outer central Queensland shelf (Great Barrier Reef). Aust. J. Mar. Freshw. Res., 3 4 : 8 2 1 - 8 3 4 . Andrews, J.C., Dunlap, W.C. and Bellamy, N.F., 1 9 8 4 . Stratification in a small lagoon in the Great Barrier Reef. Aust. J. Mar. Freshw. Res., 3 5 : 2 7 3 - 2 8 4 . Baines, G.B.K., Beveridge, P.J. and Maragos, J.E., 1 9 7 4 . Storms and island building at Funafuti Atoll, Ellice Islands. In: A.M. Cameron et al. (Editors),Proc. 2nd Int. Symp. Coral Reefs, Brisbane, 2 : 2 8 5 - 4 9 6 . Bak, R.P.M., 1 9 7 8 . Lethal and sublethal effects of dredging on reef corals. Mar. Poll. Bull., 9 : 1 4 - 1 6 . Bak, R.P.M. and Luckhurst, B.E., 1 9 8 0 . Constancy and change in coral reef habitats along depth gradients at Curaqao. Oecologia, 4 7 : 1 4 5 - 1 5 5 .
178
Ball, M.B., Shinn, E.A. and Stockman, K.W., 1967. The geologic effects of hurricane Donna in South Florida. J. Geol., 75: 583-597. Banner, A.H., 1968. A freshwater "kill" on the coral reefs of Hawaii. Hawaii Inst. Mar. Biol. Tech. Rep., 15: 1-29. Barnett, T.P., 1984. Interaction of the monsoon and Pacific trade wind system at interannual time scales. Part 111: A partial anatomy of the Southern Oscillation., Mon. Wed. Rev., 112: 2388-2400. Brown, B.E., 1987. Worldwide death of corals-natural cyclical events or man-made pollution? Mar. Poll. Bull., 18: 9-13. Burchard, J.E. and McCain, J.C., 1984. Marine life kill in the Arabian Gulf, March/April 1983. Proc. Symp. Coral Reef Environ. Red Sea. Jeddah, Jan. Abstr., pp.392-393. Cadet, D.L. and Diehl, B.C., 1984. Interannual variability of surface fields over the Indian Ocean during recent decades. Mon. Wed. Rev., 112: 1921-1935.
Cane, M.A., 1983. Oceanographic events during El NiKo. Science, 222: 1189-1 195.
Chen, W.Y., 1983. The climate of spring 1983 - a season with persistent global anomalies associated with El Nizo. Mon. Wea. Rev., 111: 2371-2384. Coles, S.L. and Jokiel, P.L., 1978. Synergistic effects of temperature, salinity and light on the hermatypic coral Montipora verrucosa. Mar. Biol., 49: 187-195.
Connell, J.H., 1973. Population ecology of reef building corals. In: O.A. Jones and R. Endean IEditorsl, Biology and Geology of Coral Reefs, Vo1.2 Biol.1: 205-245. Cooper, M.J., 1965. Destruction of marine flora and fauna in Fiji caused by the hurricane of February, 1965. Pac. Sci., 20: 137-141. Cortes, J., Murillo, M.M., Guzmdn, H.M., and AcuEa, J., 1984. Pkrdida de zooxantelas y muerte de corales y otros organismos arrecifales en el Caribe y Pacific0 de Costa Rica. Rev. Biol. Trop., 32: 227-231. Coughlan, M., 1983. Australian drought and the Southern Oscillation. Trop. Ocean-Atmos. Newsl., 21: 2- 3. Covey, D.L. and Hastenrath, S., 1978. The Pacific El Nifio phenomenon and the Atlantic circulation. Mon. Wea. Rev., 106: 1280-1287. Cubit, J.D., Thompson, R.C., Caffey, H.M., Windsor, D.M., 1988. Hydrographic and meteorologic studies of a Caribbean fringing reef at Punta Galeta, Panama: Hourly and daily variations for 1977-1985. Smithson. Contrib. Mar. Sci., 32: 1-220. Donguy, J.R., Henin, C., Mortiere, A., and Rebert, J.P., 1982. Appearances in the western Pacific of phenomena induced by El NiEo in 1976-1980. Trop. Ocean-Atmos. Newsl., 10: 1-2. Egafia, A.C. and DiSalvo, L.H., 1982. Mass expulsion of zooxanthellae by Easter Island corals. Pac. Sci., 36: 61-63. Eldin, G. and Donguy, J.R., 1983. Heat content and cyclogenesis in the south Pacific. Trop. Ocean-Atmos. Newsl., 21: 5-6. Fankboner, P.V. and Reid, R.G.B., 1981. Mass expulsion of zooxanthellae by heat-stressed reef corals: a source of food for giant clams? Experientia, 37: 251-252.
Faure, G., Guillaume, M., Payri, C., Thomassin, B.A., van Praet, M. and Vasseur, P., 1984. Sur un phknomene remarquable de blanchiment et de mortalitk massive des madrkporaires dans le complexe retifal de l'ile Mayotte ISW Ocean Indien). C. R. Acad. Sci. Ser., 111,299: 637-642. Fishelson, L., 1973. Ecological and biological phenomena influencing coral species composition on the reef tables at Eilat (Gulf of Aqaba, Red Sea). Mar. Biol., 19: 183-196. Fisk, D.A. and Done, T.J., 1985. Taxonomic and bathymetric patterns of bleaching in corals, Myrmidon Reef (Queenslandl. In: C. Gabrie et al. [Editors), Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 149-154. Franzisket, L., 1970. The atrophy of hermatypic reef corals maintained in darkness and their subsequent regeneration in light. Int. Rev. Gesamten Hydrobiol., 55 : 1 - 1 2.
179
Gill, A.E. and Rasmusson, E.M., 1 9 8 3 . The 1 9 8 2 - 8 3 climate anomaly in the equatorial Pacific. Nature, 3 0 6 : 2 2 9 - 2 3 4 . Glynn, P.W., 1 9 6 8 . Mass mortalities of echinoids and other reef-flat organisms coincident with midday, low-water exposures in Puerto Rico. Mar. Biol., 1 : 2 2 6 - 2 4 3 . Glynn, P.W., 1 9 7 3 . Aspects of the ecology of corals reefs in the western Atlantic region. In: O.A. Jones and E. Endean [Editors). Biology and Geology of Coral Reefs. Vol. I, Biol. 1 Academic Press, London and New York. pp. 2 7 1 - 3 2 4 . Glynn, P.W., 1 9 8 3 . Extensive "bleaching" and death of reef corals on the Pacific coast of Panama. Env. Conserv., 1 0 : 1 4 9 - 1 5 4 . Glynn, P.W., 1 9 8 4 . Widespread coral mortality and the 1 9 8 2 - 8 3 El Niso warming event. Env. Conserv., 1 1 : 1 3 3 - 1 4 6 . Glynn, P.W., 1 9 8 8 . El Niso-Southern Oscillation 1 9 8 2 - 8 3 : Nearshore population, community and ecosystem responses. Ann. Rev. Ecol. Syst., 1 9 : 309-345.
Glynn, P.W., Almodovar, L.R. and Gonzalez, J.G., 1 9 6 4 . Effects of Hurricane Edith on marine life in La Parguera, Puerto Rico. Car. J. Sci., 4 : 3 3 5 - 3 4 5 . Glynn, P.W., Cortes, J., Guzman, H.M., and Richmond, R.H. 1 9 8 8 . El NiEo ( 1 9 8 2 8 3 1 associated coral mortality and relationship to sea surface temperature deviations in the tropical eastern Pacific. Proc. 6th Int. Coral Reef Symp., Townsville. in press. Goreau, T.J., 1 9 6 4 . Mass expulsion of zooxanthellae from Jamaican reef communities after Hurricane Flora. Science, 1 4 5 : 3 8 3 - 3 8 6 . Gray, W.M., 1 9 8 4 . Atlantic seasonal hurricane frequency. Part I: El NiEo and 3 0 mb Quasi-Biennial oscillation influences. Mon. Wed. Rev., 1 1 2 : 1649-1 668.
Gray, W.M., 1 9 8 5 . Association of Atlantic seasonal hurricane activity with El Niso and stratospheric QBO. Trop. Ocean-Atmos. Newsl., 2 9 : 1 . Goodbody, I.M., 1 9 6 1 . Mass mortality of a marine fauna following tropical rains. Ecology, 4 2 : 1 5 0 - 1 5 5 . Guillaume, M., Payri, C., Faure, G., 1 9 8 3 . Blatent degradation of coral reefs at La Reunion Island [West Indian Ocean). Int. SOC. Reef Stud., Annu Meet. Nice (Abstr.). Halpern, D., 1 9 8 3 . Editor's Note. Trop. Ocean-Atmos. Newsl., 2 1 : 1 . Hamilton, K., 1 9 8 5 . Evidence for tropical-midlatitude teleconnections n the eighteenth and nineteenth centuries. Trop. Ocean-Atmos. Newsl., 3 0 : - 2 . Harmelin-Vivien, M.L. and Laboute, P., 1 9 8 6 . Catastrouhic imuact of hurricanes on atoll outer reef slopes in the Tuamotu (French Polynesia). Coral Reefs, 5: 5 5 - 6 2 . Harriott, V.J., 1 9 8 5 . Mortality rates of scleractinian corals before and during a mass bleaching event. Mar. Ecol. Prog. Ser., 2 1 : 8 1 - 8 8 . Hastenrath, S., 1 9 7 6 . Variations in low latitude circulation and extreme climatic events in the tropical Americas. J. Atmos. Sci., 3 3 : 2 0 2 - 2 1 5 . Hastenrath, S., 1 9 7 8 . On modes of tropical circulation and climate anomalies. J. Atmos. Sci., 3 5 : 2 2 2 2 - 2 2 3 1 . Hastenrath, S. and Wu, M-C., 1 9 8 3 . On upper-air mechanisms of the Southern Oscillation. Trop. Ocean-Atmos. Newsl., 1 7 : 2 - 4 . Holland, G.J. and Nicholls, N. 1 9 8 5 . A simple predictor of El NiEo. Trop. Ocean-Atmos. Newsl., 3 5 : 8 - 9 . Horel, J.D. and Wallace, J.M., 1 9 8 1 . Planetary-scale atmospheric phenomena associated with the Southern Oscillation. Mon. Wea. Rev., 1 0 9 : 8 1 3 - 8 2 9 . Horel, J.D., Kousky, V.E. and Kagano, M.T., 1 9 8 6 . Atmospheric conditions in the Atlantic sector during 1 9 8 3 and 1 9 8 4 . Nature, 3 2 2 : 2 4 8 - 2 5 1 . Hudson, J.H., Shinn, E.A., Halley, R.B. and Lidz, B., 1 9 7 6 . Sclerochronology: a tool for interpreting past environments. Geology, 4 : 3 6 1 - 3 6 4 . Hughes, T.P. and Jackson, J.B.C., 1 9 8 0 . Do corals lie about their age? Some demographic consequences of partial mortality, fission and fusion. Science, 209: 713-715.
Hughes, T.P. and Connell, J.H., 1 9 8 7 . Population dynamics based on size or age? A reef-coral analysis. Am. Nat., 1 2 9 : 8 1 8 - 8 2 9 .
180
Jaap, W.C., 1 9 7 9 . Observations of zooxanthellae expulsion at Middle Sambo Reef, Florida Keys. Bull. Mar. Sci., 2 9 : 4 1 4 - 4 2 2 . Jaap, W.C., 1 9 8 5 . An epidemic zooxanthellae expulsion during 1 9 8 3 in the lower Florida Keys coral reefs: hyperthermic etiology. In: C. Gabrie et al. (Editors),Proc. 5th Int. Coral Reef Congr., Tahiti, 6: 1 4 3 - 1 4 8 . Jerlov, N.G., 1 9 5 0 . Ultraviolet radiation in the sea. Nature, 1 6 6 : 1 1 1 . Jokiel, P.L., 1 9 8 0 . Solar ultraviolet radiation and coral reef epifauna. Science, 2 0 7 : 1 0 6 9 - 1 0 7 1 . Jokiel, P.L. and Coles, S.L., 1 9 7 7 . Effects of temperature on the mortality and growth of Hawaiian reef corals. Mar. Biol., 4 3 : 2 0 1 - 2 0 8 . Jokiel, P.L. and York, R.H. Jr., 1 9 8 2 . Solar ultraviolet photobiology of the reef coral Pocillopora damicornis and symbiotic zooxanthellae. Bull. Mar. Sci., 3 2 : 3 0 1 - 3 1 5 . Kamezaki, N. and Ui, S., 1 9 8 4 . Bleaching of hermatypic corals in Yaeyama Islands. (in Japanese) Mar. Parks J., 6 1 : 1 0 - 1 3 . Kenny, R.P., 1 9 7 4 . Inshore sea surface temperatures at Townsville. Aust. J. Mar. Freshw. Res., 2 5 : 1 - 5 . Kerr, R.A., 1 9 8 7 . Another El NiEo surprise in the Pacific, but was it predicted? Science, 2 3 5 : 7 4 4 - 7 4 5 . Kousky, V.E., Kagano, M.T. and Cavalcanti, I.F.A., 1 9 8 4 . A review of the Southern Oscillation: oceanic-atmospheric circulation changes and related rainfall anomalies. Tellus, 36A: 4 9 0 - 5 0 4 . Laboute, P., 1 9 8 5 . Evaluation of damage done by the cyclones of 1 9 8 2 - 1 9 8 3 to the outer slopes of the Tikehau and Takapoto Atolls (Tuamotu Archipelago). In: C. Gabrie et al. (Editors),Proc. 5th Int. Coral Reef Congr., Tahiti, 3: 323-329.
Lasker, H.R., Peters, E.C. and Coffroth, M.A., 1 9 8 4 . Bleaching of reef coelenterates in the San Blas Islands, Panama. Coral Reefs, 3 : 1 8 3 - 1 9 0 . Loya, Y., 1 9 7 2 . Community structure and species diversity of hermatypic corals at Eilat, Red Sea. Mar. Biol., 1 3 : 1 0 0 - 1 2 3 . Loya, Y., 1 9 7 6 . Recolonization of Red Sea corals affected by natural catastrophes and man-made perturbations. Ecology, 5 7 : 2 7 8 - 2 8 9 . Marszalek, D . S . , 1 9 8 1 . Impact of dredging on a subtropical reef community, southeast Florida, U.S.A. In: E.D. Gomez, C.E. Birkeland, R.W. Buddemier, R.E. Johannes, J.A. Marsh, Jr., R.T. Tsuda (Editors), Proc. 4th Int. Symp. Coral Reefs, Manila, 1 : 1 4 7 - 1 5 3 . Mayor, A.G., 1 9 2 4 . Structure and ecology of Samoan reefs. Pap. Dept. Mar. Biol. Carnegie Inst. Wash., 1 9 : 1 - 2 5 . Michelchen, N., 1 9 8 5 . The Southern Oscillation and the interannual upwelling variations off northwest Africa and in the Gulf of Guinea. Trop. Ocean-Atmos. Newsl , 30 : 1 7-1 8 . Moran, P.J., 1 9 8 6 . The Acanthaster phenomenon. Ocean. Mar. Biol. Ann. Rev.,
.
24: 379-480.
Moorhouse, F.W., 1 9 3 6 . The cyclone of 1 9 3 4 and its effects on Low Isles with special observations on Porites. Rep. Great Barrier Reef Corn., 4: 3 7 - 4 4 . Namais, J., 1 9 8 3 . Advance signs of the strong subtropical westerlies associated with the 1 9 8 3 El NiEo. Trop. Ocean-Atmos. Newsl., 2 1 : 2 1 . Nicholls, N., 1 9 8 4 a . El NiEo-Southern Oscillation and North Australian sea surface temperature. Trop. Ocean-Atmos. Newsl., 2 4 : 1 1 - 1 2 . Nicholls, N., 1 9 8 4 b . The Southern Oscillation and Indonesian sea surface temperature. Mon. Wed. Rev., 1 1 2 : 4 2 4 - 4 3 2 . Nobre, P., Moura, A.D. and Nobre, C.A., 1 9 8 5 . Planetary-scale circulation anomalies associated with droughts over northeast Brazil. Trop. Ocean-Atmos. Newsl., 3 0 : 1 1 - 1 3 . Ogden, J. and Wicklund, R. 1 9 8 8 . (Editors) Mass bleaching of coral reefs in the Caribbean: A research strategy. NOAA Research Report 8 8 - 2 . Oliver, J., 1 9 8 5 . Recurrent seasonal bleaching and mortality of corals on the Great Barrier Reef. In: C. Gabrie, et al. (Editors), Proc. 5th Int. Coral Reef Congr., Tahiti, 4 : 2 0 1 - 2 0 6 . Pan, Y.H. and Oort, A.H., 1 9 8 3 . Global climate variations with sea surface
181
temperature anomalies in the eastern equatorial Pacific Ocean for the 1 9 5 8 - 7 3 period. Mon. Wed. Rev., 1 1 1 : 1 2 4 4 - 1 2 5 8 . Pearson, R.G., 1 9 8 1 . Recovery and recolonization of coral reefs. Mar. Ecol. Prog. Ser., 4 : 1 0 5 - 1 2 2 . Perkins, R.D. and Enos, P., 1 9 6 8 . Hurricane Betsy in the Florida-Bahamas area - geologic effects and comparison with Hurricane Donna. J. Geol., 7 : 7 1 0-71 7 .
Philander, S.G.H., 1 9 8 3 . El NiEo-Southern Oscillation phenomena. Nature, 3 0 2 : 295-301,
Philander, S.G.H., 1 9 8 6 . Unusual conditions in the tropical Atlantic Ocean in 1 9 8 4 . Nature, 3 2 2 : 2 3 6 - 2 3 8 . Pickard, G.L., 1 9 7 7 . A review of the oceanography of the Great Barrier Reef. Australian Inst. Mar. Sci. Monogr. Ser. Vol. 2 . Pittock, A.B., 1 9 7 5 . Climatic change and patterns of variations in Australian rainfall. Search, 6 : 4 9 8 . Porter, J.W., Battey, J.F. and Smith, G.J., 1 9 8 2 . Perturbation and change in coral reef communities. Proc. Natl. Acad. Sci., 7 9 : 1 6 7 8 - 1 6 8 1 . Quinn, W.H., Zopf, D.O., Short, K . S . , and Kuo Yang, R.T.W., 1 9 7 8 . Historical trend and statistics of the SO, El NiEo and Indonesian droughts. Fish. Bull., 7 6 : 6 6 3 - 6 7 8 . Quinn, W.H., Neal, V.T., and Antunez de Mayolo, S.E., 1 9 8 7 . El NiEo occurrences over the past four and a half centuries. J. Geophys. Res., 9 2 : 14449-1 4461,
Rasmusson, E.M. and Wallace, J.M., 1 9 8 3 . Meteorological aspects of the El Nii?o/Southern Oscillation. Science, 2 2 2 : 1 1 9 5 - 1 2 0 2 . Rasmusson, E.M., Reynolds, R.W., Arkin, P.A., 1 9 8 4 . 1 9 8 2 - 8 3 ENSO episode revisited. Trop. Ocean-Atmos. Newsl., 2 8 : 5 - 6 . Rasmusson, E.M., Arkin, P.A., Krueger, A.F., Quiroz, R.S. and Reynolds, R.W., 1 9 8 3 a . The equatorial Pacific atmospheric climate during 1 9 8 2 - 8 3 . Trop. Ocean-Atmos. Newsl., 2 1 : 2-3. Rasmusson, E.M., Arkin, P.A., Carpenter, T.H., Koopman, J., Krueger, A.F. and Reynolds, R.W., 1 9 8 3 b . A warm episode in the eastern equatorial Pacific Ocean. Trop. Ocean-Atmos. Newsl., 1 6 : 1 - 3 . Revell, C.R. and Goulter, S.W., 1 9 8 6 . South Pacific tropical cyclones and the Southern Oscillation. Mon. Wed. Rev., 1 1 4 : 1 1 3 8 - 1 1 4 5 . Roberts, L., 1 9 8 7 . Coral bleaching threatens Atlantic reefs. Science, 2 3 8 : 1228-1 2 2 9 .
Roberts, H.H., Rouse Jr. L.J., Wallace, N.D. and Hudson, J.H., 1 9 8 2 . Cold-water stress in Florida Bay and northern Bahamas: a product of winter cold-air out breaks. J. Sed. Pet., 5 2 : 1 4 5 - 1 5 5 . Robinson, G., 1 9 8 5 . The influence of the 1 9 8 2 - 8 3 El NiEo on Galapagos marine life. pp. 1 5 3 - 1 9 0 . In: G. Robinson and E.M. del Pino (Editors). E l NiEo in the Galapagos Islands: The 1 9 8 2 - 1 9 8 3 event. Charles Darwin Foundation for the Galapagos Islands. Quito, Ecuador. Ropelewski, C.E., 1 9 8 4 . The climate of 1 9 8 3 - a season of contrasts and extremes. Mon. Wed. Rev., 1 1 2 : 5 9 1 - 6 0 9 . Ropelewski, C.E. and Halpert, M.S., 1 9 8 6 . North American precipitation and temperature patterns associated with the El Niiio/Southern Oscillation (ENSO). Mon. Wea. Rev., 1 1 4 : 2 3 5 2 - 2 3 6 2 . Sadler, J.C. and Kilonsky, B.J., 1 9 8 3 . Meteorological events in the central Pacific during 1 9 8 3 associated with the 1 9 8 2 - 8 3 El NiEo. Trop. Ocean-Atmos. Newsl., 2 1 : 3 - 5 . Shannon, L.V.,1 9 8 3 . Preliminary report of the southern Benguela warm event, 1 9 8 2 - 8 3 . Trop. Ocean-Atmos. Newsl., 2 2 : 8-9. Shapiro, L., 1 9 8 7 . Month-to-month variability of the Atlantic tropical circulation and its relationship to tropical storm formation. Mon. Wed. Rev., 1 1 5 : 2 5 9 8 - 2 6 1 4 . Shinn, E.A., 1 9 6 6 . Coral growth rate, an environmental indicator. J. Paleontol., 4 0 : 2 3 3 - 4 0 . Shinn, E.A., 1 9 7 6 . Coral reef recovery in Florida and the Persian Gulf. Environ. Geol., 1 : 2 4 1 - 2 5 4 .
182
Slack-Smith, R.J., 1 9 5 9 . An investigation of coral deaths at Peel Island, Moreton Bay, in early 1 9 5 6 . Pap. Dept., Zool. Univ. Qld., 1 : 2 1 1 - 2 2 2 . Smith, R.C. and Baker, K.S., 1 9 7 9 . Penetration of UV-B and biologically effective dose-rates in natural waters. Photochem. Photobiol., 2 9 : 3 1 1 - 3 2 3 . Sokal, R.R., and Rohlf, F.J., 1 9 6 9 . Biometry. W.H. Freeman and Company, San Francisco, 7 7 6 pp. Stephenson, W., Endean, R., Bennett, I., 1 9 5 8 . An ecological survey of the marine fauna of Low Isles, Queensland. Aust. J. Mar. Freshw. Res., 9 : 261-31 6.
Stoddart, D.R. 1 9 6 3 . Effects of Hurricane Hattie on the British Honduras reefs and cays, October 3 0 - 3 1 , 1 9 6 1 . Atoll Res. Bull., 9 5 : 1 - 1 4 2 . Stoddart, D.R., 1 9 6 9 . Ecology and morphology of recent coral reefs. Biol. Rev., 4 4 : 4 3 3 - 4 9 8 . Suharsono, 1 9 8 4 . Coral death in the Java Sea, Indonesia. Reef Encounter Newsletter of the ISRS, 2 : 6 . Suharsono and Kiswara, W., 1 9 8 4 . Natural death of corals in the Java Sea (Indonesia). Oseana IX, 1 : 3 1 - 4 0 . Tourne, Y. and Rasmusson, E., 1 9 8 4 . The tropical Atlantic region during the 1 9 8 2 - 8 3 equatorial Pacific warm event. Trop. Ocean-Atmos. Newsl., 2 5 : 1 - 2 . Walker, T.A., 1 9 8 0 . Biological oceanography of Cleveland Bay. PhD manuscript, James Cook University, Queensland. Weare, B.C., 1 9 7 9 . A statistical study of the relationship between ocean surface temperatures and the Indian monsoon. J. Atmos. Sci., 3 6 : 2 2 7 9 - 2 2 9 1 . Williams, E.H., Goenaga, C., and Vicente, V. 1 9 8 7 . Mass bleachings on Atlantic coral reefs. Science, 2 3 8 : 8 7 7 - 8 7 8 . Wolanski, E., Jones, M., and Williams, W.T., 1 9 8 1 . Physical properties of Great Barrier Reef lagoon waters near Townsville. 11. Seasonal variations. Aust. J. Mar. Freshw. Res,, 3 2 : 3 2 1 - 3 3 4 . Woodley, J.D., Chornesky, E.A., Clifford, P.A., Jackson, J.B.C., Kaufman, L.S., Knowlton, N., Lang, J.C., Pearson, M.P., Porter, J.W., Rooney, M.C., Rylaarsdam, K.W., Tunnicliffe, V.J., Wahle, C.M., Wulff, J.L., Curtis, A.S.G., Dallmeyer, M.D., Jupp, B.P., Koehl, M.A.R., Neigel, J., Sides, E.M., 1 9 8 1 . Hurricane Allen's impact on Jamaican coral reefs. Science, 2 1 4 : 749-754.
Wright, P.B., 1 9 8 5 . Relationship of SST near northern Australia to the Southern Oscillation. Trop. Ocean-Atmos. Newsl., 2 7 : 1 7 - 1 8 . Wyrtki, K., 1 9 7 5 . El Nifio - the dynamic response of the equatorial Pacific Ocean to atmospheric forcing. J. Phys. Oceangr., 5 : 5 7 2 - 5 8 4 . Wyrtki, K., 1 9 8 2 . The Southern Oscillation, ocean-atmosphere interaction and El NiEo. Mar. Technol. SOC. J., 1 6 : 3 - 1 0 . Wyrtki, K., 1 9 8 4 . A southward displacement of the subtropical gyre in the South Pacific during the 1 9 8 2 - 1 9 8 3 , El Nino. Trop. Ocean-Atmos. Newsl., 2 3 : 14-15.
Wyrtki, K., 1 9 8 5 . Pacific-wide sea level fluctuations during the 1 9 8 2 - 1 9 8 3 El Nifio. pp. 2 9 - 4 8 . In: G. Robinson and E.M. del Pino (Editors). El Nifio in the Galapagos Islands: The 1 9 8 2 - 1 9 8 3 event. Charles Darwin Foundation for the Galapagos Islands. Quito, Ecuador. Yamaguchi, M., 1 9 7 5 . Sea level fluctuations and mass mortalities of reef animals in Guam, Mariana Islands. Micronesica, 1 1 : 2 2 7 - 2 4 3 . Yamazato, K., 1 9 8 1 . A note on the expulsion of zooxanthellae during summer 1 9 8 0 by Okinawa reef building corals. Sesoko Mar. Sci. Lab. Tech. Rep., 8: 9-1 8.
Yonge, C.M. and Nicholls, A.G., 1 9 3 1 . Study on the physiology of corals. IV. The structure, distribution and physiology of the zooxanthellae. Sci. Rep. Gt. Barrier Reef Exped. 1 9 2 8 - 1 9 2 9 , 1 : 1 3 5 - 1 6 7 .
183
EL
m0 AND THE HISTORY OF EASTERN PACIFIC REEF BUILDING
MITCHELL W. COLGAN Department of Geology, College of Charleston, Charleston, South Carolina 29424 (USA)
ABSTRACT Colgan, M. W., 1989. El Niiio and the history of eastern Pacific reef building. The intense 1982-83 El Niiio-Southern Oscillation (ENSO) event significantly raised sea surface temperature devastating eastern Pacific coral reefs. In the aftermath of this disturbance, slow recovery has led to extensive reef erosion. The death of corals and the subsequent reef erosion have stimulated a reevaluation of the history and the causes for the small, low diversity reefs of the eastern Pacific. Today's eastern Pacific reefs differ markedly in size and species composition and richness from this region's past reefs and other present-day Pacific reefs. The closure of the Panamanian isthmus in the Plio-Pleistocene marked a deterioration in the reef building environment in the eastern Pacific. With this seaway closed, the modern Pacific surface circulation developed, and the components necessary for the ENSO events came together. The termination of the Pacific-Atlantic exchange coincided with the beginning of glacial-interglacial cycles. Since the closure of the Panamanian seaway and the onset of the glacial cycles, the eastern Pacific has faced two different climatic states that restricted reef growth and development: one during glacial periods with cool waters and lowered sea-levels and the other during interglacial periods with higher sea-levels, warmer waters, and ENSO events. During this latest high sea-level stand, between 18 to 65 ENSO events of the 1982-1983 magnitude may have disturbed the eastern Pacific. An uplifted reef at Urvina Bay, Galapagos Islands provides an opportunity to determine how eastern Pacific reefs develop during sea-level high stands. Here, recurrent intense ENSO events start the coral communities on a cycle of death, erosion, and recolonization stunting long-term reef growth. ENSO events, acting in concert with other physical and biological forces, have prevented the buildup of a substantial reef-framework. During the tranSition to inter-glacial times, melting ice raised sea-level and improved the environmental conditions for coral-reef growth enabling coral recruits to colonize the shelf. In the western and central Pacific, where reefs survived subaerial exposure, corals colonized and built new reefs on these antecedent structures enabling carbonate accumulation to continue. Whereas in the eastern Pacific, where small reefs eroded during low sea-level stands, recruits had to settle on
184
basaltic or other consolidated outcrops rather than on previous carbonate build-ups. New recruits face not only harsh reef building conditions (e.g., upwelling and intense grazing) but recurrent intense and lethal ENS0 events. After coral mortality, bioerosion removed much of the coral build-up. This repeated process prevents the coral community from increasing in diversity or developing to a resistant structure that can withstand erosion after death. Thus, one generation's growth is not transferred to the next, and large, persistent reef frameworks are not constructed. 1 INTRODUCTION The intense 1982-83 El Niiio event raised sea-surface temperatures killing vast stretches of coral reefs in the eastern Pacific (Glynn, 1983a, 1984, 1988a, this volume; Lessios et al., 1983; Glynn et al., in press). Before this severe event, the detrimental effects of El Niiio events on eastern Pacific reef building were unknown, and thus prior reef studies paid little attention to them (e.g., Durham, 1947; Stoddart, 1969; Stehli and Wells, 1971; Porter, 1974; Dana, 1975; Glynn et al., 1983; Glynn and Wellington, 1983). Although uncertainties exist concerning the evolutionary consequences of El Niiio events (Vermeij, this volume), this newly recognized environmental factor nevertheless has forced a reevaluation of reef development and history. In this paper, I examine whether ecologically rare but recurring intense El Nifio disturbances significantly influence the long-term structure and history of eastern Pacific reef building. On the continental shelf of the Pacific basin, reef growth initiated 8,000 years ago, and sea-level stabilized 6,200 years ago (Thom and Chappell, 1975; Chappell, 1983). The average growth rates of some eastern Pacific reefs (e.g., 3.1-3.9 m/1,000 yr, Gulf of Chiriqui, Panama, [Glynn and Macintyre, 19771) are similar to many reefs elsewhere in the Pacific (e.g., 3.2 m/1,000 yr, Great Barrier Reef, One Tree Reef [Davies and Marshall, 19801). Despite the similar starting times and growth rates, eastern Pacific reef development has been meager compared to that of the western and central Pacific (Glynn et al., 1972; Glynn, 1976). Eastern Pacific reefs typically develop as small, isolated, monospecific thickets of Pocillopora or massive colonies of Porites and Pavona growing usually no more than 3 m thick and 150 m wide, however, some reefs obtain thicknesses as great as 13 m with accumulation rates approaching 10 m/1,000 years (Glynn, 1976; Glynn and Macintyre, 1977; Glynn and Wellington, 1983; Wellington, 1984). Previous explanations for the enigma of these small, low diversity eastern Pacific reefs invoked only physical and biological causes that are daily or seasonally present. Such physical causes include wave action, low water exposure, high turbidity, shallow thermocline, cool waters, and
185
upwelling (e.g., Glynn and Stewart, 1973; Dana, 1975; Guzman and Cortes, 1989; Birkeland, in press); biological factors include predation (Glynn, 1976), distance from the source of high coral diversity (e.g.. Stehli and Wells, 1971) and bioerosion (e.g., Glynn et al., 1972; Glynn et al., 1979; Glynn and Wellington, 1983; Glynn 1988b; Scott et al., 1988) all of which act to retard coral growth, reduce sexual reproduction, and slow carbonate accumulation. However, if reef building - a long-term, historically contingent process depends on past as well as present conditions (Wilson, 1975; Davies, 1983; Grigg and Epp, 1989), then explanations that rely primarily on present-day biological and physical causes will provide only a partial answer to why eastern Pacific reefs are small. Coral reefs tolerate a limited range of environmental conditions (Wells, 1954; Stoddart, 1969; Buddemeier and Kinzie, 1976; James, 1983; Fagerstrom, 1987). making them particularly vulnerable to environmental vicissitudes, such as lowering of sea-level (Davies, 1983; Potts, 1984; Grigg, 1988), productivity changes (Birkeland, 1982, in press), and siliciclastic sedimentation (Hubbard and Pocock, 1972; Dodge et al., 1974; Dodge and Vaisnys, 1977; Cortes and Risk, 1985), which halt or reduce reef growth. Because reefs are also affected by climatological shifts, large, modern reefs have grown episodically rather than continuously. Today's reefs, only a few thousand years old, veneer older karstic and eroded reefs that grew during previous high sea-level stands (Thurber et al., 1965; Goreau, 1969). This growth developed upon inherited topography retaining still earlier development [e.g., The Great Barrier Reef (Hopley, 1982) and Enewetok Atoll (Ladd, 1973)l. Indeed, much of a recent western and central Pacific reefs topographic relief comes from this underlying structure (Hoffmeister and Ladd, 1944; Stoddart, 1973; Purdy, 1974). The formation of these large reefs is a long process that requires: 1) vigorous growth during high sealevel stands, 2) partial preservation of some of this build-up despite low sea-level subaerial erosion, and 3) initiation of new growth on an antecedent platform. For example, in the Northern Great Barrier Reef over a kilometer of reef carbonate has accumulated over Miocene sediments but recent growth accounts for only 20 m (Davies et al., in press). In contrast, in the eastern Pacific, no Miocene to Pleistocene age reef foundations exist, and the oldest coral found at the bottom of a reef structure is only 5,600 years old (Glynn, 1976). When sea-level stabilized, approximately 6,200 years ago (Chappell, 1983), corals colonized the rocky substrate, starting reef building de novo. This raises the question of why eastern Pacific reef formations had not accumulated enough carbonate during earlier periods of reef growth to provide a surface for subsequent
186
h
0
A.S-1 U
0 0
2co
B.
10
160I
,
20
I
I
I
I
I
I
I
I
I
I
40 60 80 100 120 Years (X 1,000) Before Present
I
1 0
Fig. 1. Sea - level changes. A. Changes in 618 0 values correspond to variation in the sea-level with lower readings during high sea-level stands (modified from Berger, 1982). B. Fluctuation in sea-level during the last 140,000 years (modified from Grigg and Epp, 1989).
growth (seven high sea-level periods over the last 700,000 years [Fig. l])? The partial answer to this question, as well as to why there are only small reefs, may result from the recurrence of intense El Niiio disturbances that cause widespread coral mortality. However, it would be an oversimplification to assume that El Niiio events alone caused the small, impoverished reefs of the eastern Pacific. Rather, El Niiio warming events acting in concert with other biological and physical factors have limited the size and the distribution of reefs. Since the closure of the Panamanian isthmus in the mid-Pliocene, sea-level changes and El Niiio warmings have greatly influenced the course of eastern Pacific reef development. Even without El Niiio events disrupting reef development during optimal reef growing periods (high sea-level stands), the eastern Pacific would still be a marginal reef-growing environment because of cool upwelling, intense
187
bioerosion, predators, siliciclastic sedimentation, tectonic changes, and periodic short-term and long-term cooling episodes*. Recurrent intense El Niiio events further aggravate these conditions exacerbating an already harsh environment for hermatypic corals (Glynn, 1988a). Before examining El Niiio events and their consequences, I review the physical environment of the eastern Pacific and examine past and present reef building. 1 \qicaragua
900
I
8'0 O
1
CARIBBEAN N SEA
Gulf o costa Papagayu Rica
Panam Gulf .of
.
Chiriqui
COCOS ' Islands
I 5 O
Gulf of panma]
Malpelo Island
-
Island
00
A;-
,Galapagos Islands
Urvina Bay
0.
p
Ecuador
Fig. 2. Eastern Pacific location map. 2 BACKGROUND 2.1 Eastern Pacific -- phvsical setting and reefs Although there is strong affinity at the species level with the central and western Pacific, which has over 510 species (Veron and Kelly, 1988), the eastern Pacific reefs are much less diverse with only 18 species of scleractinian and 3 species of hydrozoan corals forming reefs (Glynn and Wellington, 1983). The eastern Pacific reefs discontinuously occur from the tip of Baja California, in the north, to the coast of Ecuador, in the south, and include offshore islands (UNEPDUCN, 1988; Fig. 2). Reef growth is restricted
*
The Little Ice Age being the most recent of these events (Grove, 1988).
188
by turbidity, siliciclastic sedimentation, the relatively narrow continental shelf (Stehli and Wells, 1971), and the region's shallow (20 m) and fluctuating thermocline (Allison, 1959; Dana, 1975). Sea surface temperatures generally range between 24" to 29°C (Dana, 1975), and fall within the range for vigorous reef growth (i.e., 25" to 28"C, Wells, 1956).
Uva Reef Sea Level
A
Non-pocilloporid c o r a l s
Secas Reef
A
Non-pocilloporid c o r a l s
Fig. 3. Zonation of two reefs (Uva Reef = patch reef, Secas Reef = fringing reef) in the Gulf of Chiriqui, Panama (modified from Glynn, 1976). Coral reefs of the eastern Pacific, furthermore, contend with the seasonal upwelling of cool nutrient-rich waters that negatively affect reef growth, consequently reefs are best developed in the nonupwelling areas (Glynn and Stewart, 1973; Dana, 1975; Glynn, 1977; Glynn and Macintyre, 1977; Glynn and Wellington, 1983). For example, in the Gulf of Chiriqui, where there is no upwelling, small (1 to 2 hectares in lateral extent) reefs occur that are among the finest reefs in the eastern Pacific. The weakly zoned reef framework, 8 m thick, confined to shallow ( c 10 m) depth (Fig. 3), consists of poorly cemented frameworks of interlocking branches of P o c i l l o p o r a weakly bound by calcareous algae (chiefly Porolithon [Glynn et al., 1972; Glynn and Macintyre, 19771). P a v o n a , Porites, and Millepora are commonly found on
189
the reef along with the frame-building Pocillopora (Glynn et al., 1972). In the Gulf of Panama, where upwelling occurs, reefs are more poorly developed (Wellington, 1982) with a median reef thickness ranging between 3.4 - 4.9 m (Glynn and Wellington, 1983). The cooler water in the Gulf of Panama slows the mean rate of Pocillopora growth to 3.1 cm/yr compared to 3.9 cm/yr in the Gulf of Chiriqui (Glynn, 1977).
60"
0"
60" 180"
120"
60"
O0
6 Oo
120"
180"
Fig. 4. Circumtropical surface circulation 25 million years ago (modified from Romine and Lombari, 1985). 2.2 Past eastern Pacific reefs Today's small, low diversity, eastern Pacific reefs differ markedly in size and species composition from past reefs of the region. After the global collapse of reef building at the end of the Cretaceous, 65 million years ago (Newell, 1972), the eastern Pacific experienced sporadic bouts of reef building and, until the closing of the Isthmus of Panama (about 3.5 million years ago; Keigwin, 1978), reef building was more extensive and species diversity was greater (Durham, 1966; Frost, 1977). When the Isthmus of Panama was open, the eastern Pacific was the westernmost extension of the Teyths sea (Fig. 4). and consequently shared many species with the Atlantic (Durham and Allison, 1960; Durham, 1966 [Appendix 11). During the mid-Eocene as many as 27 genera of shallowwater hermatypic corals grew throughout the region (Appendix 1). In
190
particular, small biostromal reefs were formed in Nicaragua (Brito Formation; Vaughan, 1919), Mexico (San Juan Formation; Frost and Langenheim, 1974), and Panama (Gatuncillo Formation; Woodring, 1957). the end of the Eocene, reef building ceased. 30
I
EaJ
g 20 0 b
4
0
P
Foundinthe western Atlantic and eastern Pacific G l Found in the Indo-Pacific and eastern Pacific El Foundinthe
10
E
Indo-Pacific, western Atlantic, and eastern Pacific
2 =I
I
At
0
Pal Eoc Olig Mi0 Pli Pleis Holo Time
Fig. 5 . The graph shows whether the coral genera found in the eastern Pacific were also present in either the western Atlantic or Indo-Pacific region (see Appendix 1). Pal = Paleocene, Eoc = Eocene, Olig = Oligocene, Mio = Miocene, Pli = Pliocene, Pleis = Pleistocene, Holo = Holocene During the Oligocene, reefs grew vigorously, comparing more favorably in size, diversity, and species composition to present Indo-Pacific reefs (Frost and Langenheim, 1974), and reached their maximum development in the eastern Pacific and the adjacent Caribbean (Frost, 1977). For example, one well zoned barrier reef stood more than 46 m thick and contained 31 coral species (e.g., species belonging to the genera Stylophora, Acropora, Porites, Porites (Synaraea), Favites, Goniastrea, and Diploastrea [Frost and Langenheim, 19741). Throughout the eastern Pacific, 25 genera of corals were discovered, and they had a strong affinity to the Caribbean region (Fig. 5; Appendix 1). This vigorous period of reef building came to an abrupt end in the late Oligocene (approximately 30 million years ago) with the extinction of many genera of shallow water corals (Frost, 1977; Veron and Kelly, 1988). During the Miocene and early Pliocene, the movement of the Caribbean Plate and the uplift of southern Central America constricted flow through the Panamanian Seaway (Sykes et al., 1982; Fig. 6) marking drastic changes in global oceanic and climatic patterns (Berggren, 1982; Romine, 1985). Subsequently, coral-algal mounds rather than true reefs developed in the
191
eastern Pacific (e.g., Panama Formation [Woodring, 19571 and Imperial Formation [Vaughan, 19171), and shallow water reef building corals declined from 25 to 12 genera (Vaughan, 1919; Fig. 5 ; Appendix 1).
&
North American P l a t e
11
I
East-Pacific-
I
North American
Late Miocene 7 Ma
Fig. 6. Plate movement and the closing of the Panamanian Seaway in the Tertiary. Teeth marks are found on the upper plate in the overthrust zones. Dash lines indicate where the land was probably submerged (see Fig. 10). Newly created sea floor is represented by hatched lines (modified from Sykes et al., 1982). Around 3.5 to 3.2 million years ago (Keigwin, 1978; Keller et al., 1989). the Isthmus of Panama closed completely, and the source of eastern Pacific
192
coral recruits switched from the Caribbean to the Indo-Pacific (Fig. 5). The closing of the central American passage, critical in the reorganization of oceanic paleocirculation (Romine, 1985; Gartner et al., 1987). produced the oceanographic and atmospheric conditions necessary for intense El Niiio events (see following sections). Since the Pliocene, the eastern Pacific reef record is very poor (Dana, 1975). The absence of extensive fossil reefs is consistent with the establishment of the present marginal reef growing conditions with its seasonal upwelling, El Niiio warming events, and bioerosion, which inhibits the preservation of coral reefs. The PlioPleistocene interval that started with the full closure of the Panamanian isthmus thus marked a deterioration in the reef building environment in the eastern Pacific. 3 THE 1982-1983 EL NIfiO EVENT AND EASTERN PACIFIC REEFS The El Niiio warming event, a complex meteorological and oceanographic phenomenon (Caviedes, 1984; Vallis, 1986; Deser and Wallace, 1987; Hansen, this volume), often coincides with global climatic changes (Bjerknes, 1969, 1972; Cane, 1983, 1986; Rasmusson, 1985). The thermal gradient between the colder waters of the eastern Pacific and the warmer waters of the western Pacific drives intense El Niiio events (Cane, 1983, 1986; Hansen, this volume). Normally, the thermal difference between the cooler eastern Pacific (high pressure) and the warmer western Pacific (low pressure) causes tradewinds to move from east to west. About one year before an El Nifio event, the winds shift and increase in strength, and a deeper layer of warm water accumulates in the western Pacific (Cane, 1983, 1986; Enfield, 1987). The low pressure system migrates eastward (southern oscillation), the tradewinds weaken and the north equatorial current intensifies (Fig. 7). The warm water that had “piled-up” in the western Pacific now moves in Kelvin waves (Wyrtki, 1975, 1979; Miller et al., 1988) toward the eastern Pacific causing the El Niiio-Southern Oscillation (ENSO) event (Bjerknes, 1969, 1972). As a result, in the western Pacific, sea-level lowers (e.g., Guam [Yamaguchi, 19751 and Truk Lagoon), and rain fall decreases (Wyrtki, 1985). Whereas, in the eastern Pacific, sea-levels rise, sea surface temperatures increase, and equatorial and coastal upwelling of cool, nutrient-rich waters relaxes (Cane, 1986; Hansen, this volume). Only recently has this warming event been recognized as a negative force on reef building (Glynn, 1984; Robinson, 1985). Before 1982-1983, ENSO events were thought beneficial to coral reefs because elevated sea-water temperatures would enhance coral growth (Glynn and Wellington, 1983) and increased current rates would speed planulae transport thereby promoting
193
coral recruitment (Dana, 1975; Glynn and Wellington, 1983; Richmond, this volume). Although less intense El Niiio events might potentially be favorable to coral communities, the more intense 1982-1983 occurrence raised mean sea surface temperatures 3"-4"C (Glynn et al., in press; Glynn and D'Croz, in press) and dramatically demonstrated the detrimental outcome of significant and prolonged sea-water warming on coral reefs and other marine communities (Barber and Chaves, 1983, 1986; Glynn, 1988a; Barber and Kogelschatz, this volume; Dayton and Tegner, this volume).
Fig. 7. Ocean surface circulation and the Equatorial undercurrent in the tropical eastern Pacific (modified from Grove, 1984). Elevated sea surface temperatures of the 1982-83 El Niiio event catastrophically affected eastern Pacific coastal reefs from northwestern Costa Rica to southern Ecuador as well as the offshore reefs of COCOSand Galapagos Islands (Fig. 8). Early in 1983, the El NiRo warming stressed reef building corals, so that they began expelling their endosymbiotic algae (zooxanthellae), thereby turning vast stretches of reef from golden-brown to white (Glynn, 1983a, 1984). This bleaching extended to coral colonies living as deep as 20 m, and within 2-4 weeks, the coral tissue atrophied and the colonies died (Glynn, 1983a; Glynn et al., 1985b). Coral mortality ranged
194
form 51 % on Caiio Island, Costa Rica to 97 % in the Galapagos Islands (Glynn et al., in press). Although all reefs were devastated, in Panama coral mortality was greater in upwelling areas (e.g., 84 %, Gulf of Panama) than in non-upwelling localities (e.g., 76 %, Gulf of Chiriqui [Glynn et al., in press]). Death extended to all coral species, but most heavily impacted the fastgrowing Pocillopora and Millepora colonies, and some massive colonies of Gardineroseris (Glynn, 1983a, 1984; Guzman et al., 1987).
9oow
80*W
Fig. 8. Sea surface temperatures before (May 1982) and during (May 1983) the 1982-1983 El Niiio event (from Glynn et a]., in press). Along with coral colony deaths, El Niiio effected the local and regional extinction of some coral species. Possible new recruits from the central Pacific, Acropora valida in Colombia (Prahl and Mejia, 1985) and Porites (Synaraea) rus in Costa Rica (Cortes and Murillo, 1985) may not have survived the El Niiio warming (Prahl, 1985). Along with these new recruits, other coral species became locally extinct or exceedingly rare; two of three species of Millepora have not been found alive, Pocillopora colonies were drastically reduced or absent from some reefs in the Galapagos Islands and Panama (Glynn, 1983a, 1984), and Gardineroseris planulata and Porites panamensis were nearly eliminated from Caiio Island, Costa Rica (Guzman et al., 1987). Coral mortality was not limited to the initial warming events. El Niiio bleaching of Pocillopora colonies caused a decline in mucus production
195
thereby depriving of their food the commensal crustaceans that protect the coral from Acanthaster planci predation (Glynn, 1980, 1983b; Glynn et al., 1985a). Soon after the Pocillopora colonies died the crustaceans either fled or died (Glynn et al., in press). The widespread death of Pocillopora colonies removed a biotic barrier that once protected massive colonies that lived within these coral thickets exposing them to A . planci predation. For example, Pocillopora colonies that encircled 22 massive colonies of Gardineroseris planulata suffered 95 % mortality, and, within a year, Acanthaster preyed upon 95 % of those encircled massive corals (Glynn, 1985). Perhaps the most devastating factors affecting long-term build-up of coral reefs is the acceleration of bioerosion following coral mortality (Glynn, this volume). Even after death, an intact corallum can support future recruitment so that reef accretion can continue. However, echinoderm grazers erode carbonate and may determine whether reefs accrete or not (Stearn and Scoffin, 1974; Hutchings, 1986; Birkeland, in press). In Panama, D i a d e m a mexicanum recruited to the dead Pocillopora framework increasing their mean densities from 3 inds./m2 to 80 inds./m2. In the Galapagos Islands, Eucidaris thouarsii moved to dead coral colonies covered with benthic algae, swelling their numbers from 5 inds./m2 to 50 inds./m2, thus increasing the rate of bioerosion by 3 to 5 times over that previously reported (Glynn, 1988b, this volume). However, in Panama and Galapagos, damselfishes established new territories and excluded urchins thereby reducing the intensity of bioerosion (Glynn, 1988b). although internal eroding sponges persist (Pang, 1973). Nevertheless, in both areas, bioerosion by external grazers (e.g., molluscs and sea urchins) and internal borers (e.g., Lithophaga mussels [Scott et al., 19881 and sponges) exceed the rate of carbonate production and thus removed the carbonate foundation for future reef growth (Glynn, this volume). In contrast to the benefits once associated with El Nifio warmings, the exceedingly strong 1982-1983 occurrence (Gill and Rasmusson, 1983; Hansen, this volume) demonstrated its harmful consequences in three ways (Glynn, 1988a). First, elevated sea-water temperatures caused widespread coral mortality without a concomitant decrease in coral predators and bioeroders (Glynn, 1983a, 1984, 1988a; Robinson, 1985). Secondly, with some coral sanctuaries breached, Acanthaster planci entered and killed many of the surviving corals (Glynn, 1985). Thirdly, with coral reef growth halted or at a near stand-still, internal borers (e.g., lithophagine bivalves and sponges) and external bioeroders (e.g., sea urchins, D i a d e m a and Eucidaris) reduced coral build-ups to rubble (Scott et al., 1988; Glynn, 1988b, this volume).
196
4 EVIDENCE FOR PAST EL N f i O EVENTS For El Niiio events to play an important role in the long-term development of reefs, they must have an equally long history. In this and the following section, I intend to demonstrate that El Niiio events may have a history stretching back at least to the closing of the Isthmus of Panama some 3.2 - 3.5 million years ago, a span of time coinciding with the latest interval of eastern Pacific reef building. The relative short duration of El Niiio events and their ambiguous sedimentological signals produce an equivocal record (De Vries, 1987). However, three lines of evidence successively extend the age of El Niiio events: historical and proxy data (as old as 450 years), sedimentological and paleontological records (as old as 7,500 years), and paleoceanographic and paleoclimatologic reconstructions (as old as 3.2 million years). 4.1 Historical and proxv records
Historical analysis of reports and ships’ logs dating back to the early 1500s indicate the sort of climatic changes that are consistent with El Nifio events along the usually dry northwestern coast of South America (Quinn et a]., 1987). Quinn and coworkers divide these events into five categories, with the intense events like the one in 1982-1983 classified as “very strong”. Between 1803-1987, El Niiio events (“moderate”, “strong”, or “very strong”) recurred every 3.8 years (Fig. 9). From the early 1500s, 32 strong to very strong El Niiio events have affected the eastern Pacific. These intense El Niiio events have recurred on the average of every 9.9 years, but with as much as 20 years between the appearances. Recent strong events occurred in 1940-41, 1957-58, 1972-73, and eight very strong events occurred between 1500 and 1982-1983 (Quinn et al., 1987). Proxy records using tree-rings (Lough and Fritts, this volume), ice core records (Thompson et al., 1984), and corals (Druffel, 1985; Shen et al., 1988; Druffel et al., this volume; Shen and Sanford, this volume) have been used to trace the appearance and the recurrence interval of El Niiio events. Although of limited temporal span, these data establish that El Niiio events are a normal consequence of climatic fluctuation rather than of human environmental impact.
4 . 2 Holocene record. sedimentological evidence Cool ocean waters normally bathe the shores of Peru’s arid, narrow coastal plain. When El Niiio events warm these coastal waters, precipitation increases; during severe sea-surface warming events, torrential rains flood
197
vs
c 1500
I
1600
1700
1800
15
Fig. 9. Frequency of “strong” (S) to “very strong” (VS) El Nifio events over the last four and half centuries from Quinn et al. (1987). Thin lines less than one year, thick lines greater than a year’s duration. U.B. = time of the Urvina Bay uplift (modified after Glynn, 1988a).
the dry South American coastal plains (Philander, 1983; Hansen, this volume: Dillon and Rundel, this volume). Several lines of geological evidence that rely on flood-related sedimentary features can be used to date the occurrence of El Niiio events in the long term geologic past. Along the arid coast of Santa, Peru, a series of eight subparallel beach ridges is interpreted by some as having formed after intense El Niiio events (Sandweiss, 1986). However, for others, the association between these beach ridges and El Nifio events is ambiguous (De Vries, 1987). Less ambiguous sedimentary evidence is provided by alluvial, overbank, flooding depositional sequences consisting of sheets of sandy gravel. The flood conditions associated with the 1982-1983 El Niiio event produced a high energy, high sediment-load alluvial deposit, useful in determining the occurrence of past events (Wells, 1987). Using these sedimentologic criteria and aided by archaeological relics, Wells (1987) was able to date four El Nifio layers with radiocarbon dates of 1720 f 60 A.D., 1460 t 20 A.D., 1380 t 140 A.D., and 1230 t 60 B.C. A minimum of 15 “very strong” El Niiio events, severe enough to produce enough sediment to leave a long-term depositional record, have affected the Peruvian coastal plains in the last 7,500 years (Wells, 1987). For example, around 900 years ago, a flood estimated to be two to four times as strong as the 1925 flood* devastated the Moche Valley, Peru (Nials et al., 1979). Another three catastrophic floods ravaged the same valley around 2,500, 1,900, and 1,500 years ago (Moseley et al., 1981). Still
*
A very strong El Nifio event occurred in 1925 (Quinn et al., 1987).
198
older floodplain, overbank deposits, now exposed in river cuts, reveal a minimum of 21 late Pleistocene flood events possibly indicative of earlier El Nifio events (Wells, 1987).
5 OCEAN CONDITIONS AND PAST EL NIfiO EVENTS Sedimentary records discussed above suggest that the 1982-1983 El Niiio event was not a unique occurrence, and further that some previous events were even more intense. For at least the last 7,500 years, during the time of optimal reef-growth, El Niiio events produced elevated sea surface temperatures, brought torrential rains, and devastated the coral reef tract i n the eastern Pacific. How often have such El Niiio events recurred during the long history of eastern Pacific reef building? Without unequivocal data pointing to the exact timing of past El Niiio events, one must rely on circumstantial evidence to reconstruct the events leading to the start and the periodicity of El Niiio events (De Vries, 1987). El Niiio events are not isolated unusual events or the outcome of human environmental tampering, and they appear to have started when present day climatic and oceanographic conditions came about (Rasmusson and Wallace, 1983; Rasmusson, 1985; Enfield, 1989). In particular, the oceanographic El Niiio event and the atmospheric southern oscillation form as a consequence of the present configuration of the Pacific Basin, which was developed about 3.5 million years ago (Kennett et al., 1985). Throughout the Mesozoic and early Tertiary, a sluggish Tethyan circumequatorial circulation dominated the Pacific (Kauffman and Johnson, 1988; Fig. 4). Closure of the Indonesian and Panamanian ocean gateways fragmented the global circumequatorial circulation, and altered current flow and climate patterns (Berggren, 1982; Fig. 10). The movement of the Caribbean plate constricted flow in the case of the Panamanian seaway (Sykes et al., 1982; Fig. 6) and reorganized oceanic paleocirculation (Gartner et al., 1987), with the final closing occurring about 3.5 to 3.2 million years ago* (Keigwin, 1978; Keller et al., 1989). In the eastern Pacific, the isthmus barrier favored colder, upwelling waters and enhanced the influence of the equatorial undercurrent (Romine, 1985). About 2.8 million years ago wind stress increased (Hays et al., 1989) steepening the thermocline and increasing the east-west thermal gradient (Kennett et al., 1985; Fig. 10) bringing together the elements that influence modern eastern Pacific circulation (Romine, 1982).
*
Deep water connection between the Pacific and the Caribbean ceased about 3.8 million years ago, but a shallow sea way may have existed between 3.0 to 2.2 million years ago (Berggren and Hollister, 1974).
199
r
I
-b
ECC C
t Thermocline
zone -Photx KT?3
/;:;:;:;:;:\ ...... . .. ._.,.... . . . .
12- 16 ma
Figure 10. Gateway closing and the formation of tropical Pacific surface circulation (modified from Kennett et al., 1985). Between 20 and 10 million years ago, northward movement of the Australian plate constricted and finally closed the Indonesian Seaway preventing the surface circulation between the Indian and Pacific Oceans. Encountering this barrier, warm westward flowing equatorial waters amassed on the western side of the Pacific basin, and redirected current flow creating the Equatorial Under Current (EUC) and strengthening the Equatorial Counter Current (ECC). The movement of the Caribbean plate closed the flow through the Panamanian seaway and strengthened the westward flowing South Equatorial Current (SEC), the eastward streaming ECC, and the eastward moving subsurface EUC. The depth of the EUC parallels the thermocline depth and flows at 150-250 m depth in the west and shoals to less than 50 m in the east. Therefore, starting some 2.8 million years ago, the continental configuration and thermal gradients produced the modern Pacific surface circulation systems. And, thus for the first time, all the components necessary for the El Niiio events came together, and may have coincided with the starting point for intense ENS0 events. The termination of the Pacific-Atlantic exchange through the Panamanian seaway coincided with the beginning of glacial-interglacial cycles (Berggren and Hollister, 1974; Imbrie and Imbrie, 1979; Berger, 1982; Romine, 1985;
200
Gartner et al., 1987). How did ENSO events respond to the oscillation between the glacial and inter-glacial extremes? Over the last 700,000 years, there have been seven interglacial periods, and at these times sea-level and climate were similar to today, and thus should have experienced ENSO (Fig. 1). High sea-level stands are also the time of maximum reef growth (Davies, 1983), thus appearance of ENSO events coincides with the times of optimal reef growth. During glacial times, when reef building diminishes, oceanographic and climatic conditions are different, and Quinn (1971) and Salinger (1981) have suggested that during glacial periods, ENSO events were reduced in frequency and intensity. Four lines of evidence for climatic changes suggest that ENSO events were absent during the lowered sea-levels. 1) The widespread equatorial dry zone (Hutchinson, 1952; CLIMAP, 1976, 1981; Stoddart and Scoffin, 1983; Aharon and Veeh, 1984; Rea et al., 1986) may indicate a reduction in southern oscillations. 2) The southern oscillation is linked to the periodic monsoonal storms (Rasmusson, 1984). During the last glacial phase, Australia and southern Asia were drier (Bowler et al., 1976; Kershaw, 1976, 1978) and monsoonal rains and upwelling were reduced (COHMAP, 1988). 3) In the tropical eastern Pacific, where El Niiio associated rains account for much of the precipitation (Colinvaux, 1984), sedimentologic and paleontologic data indicate that climatic conditions were drier than at present. The stratigraphy of El Junco Lake on San Cristobal Island, Galapagos Islands, suggests that rainfall was greatly reduced between 34 - 10 x 103 years ago (Colinvaux, 1972, 1984). 4) Lacustrine paleoenvironment records from Mexico (Bradbury, 1989) and Peru (Markgraf, 1989) show lower lake levels with arid and semiarid conditions dominating large portions of equatorial South America (Damuth and Fairbridge, 1971). In the eastern Pacific, between 10,000 to 11,000 years ago a rising sealevel and the establishment of the present atmospheric and oceanic circulation patterns (Romine and Moore, 1981) enabled ENSO events to once again occur. Sea-level stabilized around 6,500 years ago (Chappell, 1983), and corals could once again inhabit the eastern Pacific’s continental shelf. During this latest high sea-level stand (the last 6,500 years), how many events have disturbed the eastern Pacific reef building? A wide range of estimates surround the recurrence interval of intense El Niiio events like the 1982-1983 incident. Quinn et al. (1987) find that “very strong” events occur every 50 years (e.g., 1891, 1925-1926, and 1982-1983), however, the 1891 event most closely resembled the 1982-1983 El Niiio episode (Enfield, 1987) giving nearly a 100 year interval. The existence of large, massive corals from Panama (Glynn, 1985) and the Galapagos Islands (Dunbar et al.. 1988) that did not survive the 1982-1983 El Nifio event may indicate that such a
20 1
strong event has not occurred for at least 100 to 350 years (Glynn, this volume). Thus, over the last 6,500 years, between 18 to 65 El Niiio events of the 1982-1983 magnitude could have disturbed the eastern Pacific reef building region. Also, ENS0 events of a significantly greater magnitude than the one in 1982-1983 may recur about every 500 years (Wells, 1987). Thus, far from being rare during the long history of reef building, El Niiio events of the 1982-1983 magnitude frequently disturbed eastern Pacific reefs during the latest, and no doubt during other sea-level high stands.
6 URVINA BAY, GALAPAGOS ISLANDS An uplifted reef at Urvina Bay, Galapagos Islands provides an opportunity to examine the interplay between normal environmental conditions and periodic El Niiio warming events to determine how eastern Pacific reefs develop during repeated sea-level high stands. The Galapagos Islands, located on the equator some 1,000 km off the west coast of Ecuador, experienced the greatest El Niiio related coral mortality during the 1982-1983 event (Glynn et al., in press, this volume). Here, recurrent El Niiio events that significantly raise sea surface temperature start coral communities on a cycle of death, bioerosion, and recolonization, stunting long-term reef growth. Urvina Bay’s uplifted corals show the consequences of such past El Niiio events (Fig. 11). At Urvina Bay, El Niiio events, acting in concert with other physical and biological forces, have prevented the build-up of a substantial reef-framework. Instead, corals formed small, isolated reefs. In 1954, along the west-central coast of Isabela Island, the upward movement of magma suddenly raised a portion of Urvina Bay* more than 7 m above sea-level and drove the shoreline 1.2 km seaward (Malmquist et al., 1986; Fig. 12), and exposed several km2 of a marine shelf including invertebrate and vertebrate remains as they were in 1954 (Couffer, 1956; Colgan and Malmquist, 1987). A geologically young “aa” lava flow (around 1,000 yr. old according to relative lava flow stratigraphy; K. Howard, pers. comm.) underlies Urvina Bay. Before the uplift, the lava’s irregular topography caused variable water depth and wave exposure, which produced seven microhabitats where corals and calcareous algae settled and grew (Fig. 13). In addition to attached coral colonies, two locations contain unattached coralliths, mobile, rounded, branching and massive coral colonies (Glynn, 1974). Along with corals and
* Local
uplifts, often by the intrusion of magma, commonly accompany the growth of islands in the Galapagos (McBirney and Williams, 1969).
202
91'20'
9 1" 15'
Fig. 11. Location map of Urvina Bay, Isabela Island, Galapagos Islands. calcareous algae, a rich invertebrate community contains abundant sponges, ahermatypic corals, annelids, bryozoans, brachiopods, molluscs, arthropods, echinoderms, and tunicates. Some of the most abundant invertebrates were bioeroding sea urchins (e.g., Eucidaris thouarsii and Echinometra vanbrunti), and boring bivalves (e.g., Lithophaga). During three field seasons, the uplift deposit was mapped using stadia rod and alidade. With the freedom of walking, rather than diving, we were able to measure and map most of the coral colonies and sediment deposits on the uplift.
6.1 El Nifio events at Urvina Bay The following sections show how the appearance of catastrophic El Nifio events altered the size structure, growth and community development of the Urvina Bay corals. The most recent intense El Nifio event to affect Urvina Bay before the 1954 uplift occurred in 1941 (Fig. 9), and many of the corals, branching and massive, apparently died or were partially killed as a result
203
of that warming event. The coral skeletons of surviving corals recorded that sea surface temperature was elevated and productivity was reduced (Druffel, 1985; Linn, 1988). Seven hermatypic coral species lived in the shallow waters of Urvina Bay (Table 1) with three of these species, Pavona clavus, Pocillopora damicornis, and Porites lobata producing 17 small, isolated, nearly monospecific reefs. Among these small reefs, some individual colonies attained an enormous size (i.e., P . clavus, 12 m diameter [Fig. 141). TABLE 1 Hermatypic corals of the Urvina Bay Uplift, Isabela Island, Galapagos Islands. Higher taxonomic classifications rely on the interpretations of Veron (1 986). Sclerac tinia Astrocoeniina Pocilloporidae Pocillopora damicornis Pocillopora elegans Fungiina Siderastreidae Psammocora (Stephanaria) stellata Agariciidae Pavona clavus Pavona gigantea Pavona varians Poritidae Porites lobata
6.1.1 Branching corals Before the uplift, three branching corals grew at Urvina Bay, Psammocora ( S . ) stellata, Pocillopora damicornis, and Pocillopora elegans, and as at many locations in the eastern Pacific, P . damicornis formed the most extensive, interlocking coral framework. During the 1982-1983 ENS0 warming event, Pocillopora colonies died in large numbers (Glynn, 1983a, 1984). At Urvina Bay, although producing the most extensive reef structures, probably none of the P . damicornis colonies were alive at the time of the uplift (broken, urchin-eroded, over-turned, and algal encrusted upper and outer branches testify to death prior to the uplift). The widespread death of the P . damicornis colonies points to the 1941 El Niiio event as the cause. If, following the death of the Pocillopora colonies, new corals had been recruited to and grew on the dead framework then the structure would have
204
Fig. 12. Top aerial photograph of Urvina Bay taken in 1946 before the uplift. The arrow points to the white carbonate sand beach. Lower aerial photograph of Urvina Bay taken after the 1954 uplift, the white carbonate sand beach is now nearly a kilometer from shore. remained intact and framework development would have continued. However, either coral recruits failed to settle or did not survive after settlement on the dead framework. Without this new growth, bioeroders (echinoderms, sponges, and mollilscs) and currents rapidly tore-down the P. damicornis framework and previous growth was turned to rubble. Once the coral had died, crustose and fleshy algae covered the coralla, and large
Channel Subtidal Beach
[3 Bay
Fig. 13. Urvina Bay facies map.
206
Fig. 14. Large colony of Pavona clavus (max. diameter, 12 m). Bioerosion caused a portion of the colony to collapse and fragment before the 1954 uplift. numbers of herbivores grazed the algae and eroded the coral skeletons (Glynn, 1988b). On the uplift, Pocillopora damicornis formed the most continuous reef structure, and these colonies grew-in and filled-up a central channel (Fig. 15). In the area dominated by P . damicornis, three echinoid species, E u c ida r is tho u a rs i i , E c h i n o m e t r a vanb run t i , and L y t e c h in us s e m i t u b e r c u 1 a t u s thrived in and around the Pocillopora colonies. A single P . damicornis colony covering 34 m2 and standing at its maximum, 0.4 m high (average height 0.2 m), housed 487 urchins ( E . thouarsii - 4.2/m2, E . vanbrunti - 3.9Im2, and L . semituberculatus - 6.21m2). This concentration of sea urchins (14.3 urchinslm2) on the P . damicornis colony suggests a removal rate of 230 gm of corallday, a rate of erosion that in itself cannot quickly reduce the 22,390 kg of coral framework to rubble (Malmquist, in prep.). But the degree of destruction is not measured only by the amount of coral removed. Sea urchins do not remove the corallum like a carpenter planes a wooden plank, but rather urchins act somewhat like beavers undercutting the framework and toppling portions of the coral colonies (Figs. 16, 17). Thus, although relatively small amounts of coral are removed, large portions of the colony are destroyed. Before the uplift, damselfishes (Stegastes spp.) probably colonized the dead P . damicornis colonies, and protected their algal lawns by excluding sea urchins (Williams, 1979; Sammarco and Williams, 1982; Glynn and Wellington, 1983; Eakin, 1987). This protection of a portion of the coral
Fig. 15. Urvina Bay facies map with location of some of the coral communities. 207
208
Fig. 16. Sea urchins grazing and undercutting a small build-up of P o c i f f o p o r a damicornis. The scattered tests of Eucidaris thouarsii are found at the base of the colony. Left-hand scale bar = 10 cm. colony resulted in a local decrease in bioerosion creating an irregular corallum topography with towers and pits (Eakin, in press; Glynn, this volume). However, sea urchins can potentially attack these damselfish towers from below (Fig. 17), eventually toppling them; leveling even these protected areas. In one survey area at Urvina Bay (200 m2), all 22 colonies examined were either urchin-pitted, undercut, or toppled. Toppled colonies were apparently further broken up by currents and wave action to form unconsolidated rubble, which fills an adjacent channel with gravel-sized (mean size > 4 mm diameter) poorly sorted (og=1.03) P o c i l f o p o r a branches and carbonate sands. At the time of the uplift, in contrast to the P . damicornis, most of the Pocillopora elegans colonies were alive and small. Clusters of discrete colonies of P . elegans, spread across the raised surface (Fig. 18), are uniformly small with mean widths ranging from 18.9 cm (s=9.9, n=60) to 22.5 cm (s=12.1, n=131). In five P. elegans clusters, there were no significant differences between the size distribution of the colonies found at each of the sites (one-way anova F[4, 566]=2.3ns; p>0.05 [Fig. 191). Given an average growth rate of 2.4 mm/mo (Glynn and Wellington, 1983) all likely settled after the 1941 El Niiio event. ~
209
Fig. 17. A remnant of a possible damselfish tower on a Pocillopora damicornis colony with sea urchins undercutting the tower. The tower rose over 23 cm above the rest of the coral colony. Damselfish protect their algal lawns from sea urchin grazing producing this type of irregular topography (Eakin, in press; Glynn, this volume). Sea urchin tests are found at the base. None of the P . elegans colonies grew on top of dead coral colonies, because the intense grazing pressure of sea urchins may have excluded coral recruits from previous build-ups (Glynn, this volume). Instead, the recruits survived on top of basalt outcrops, where urchin concentrations were less. The exclusion of recruits from previous build-ups forced reef building to start in new locations. The population structure of P . elegans thus can be explained as a consequence of the 1941 El Niiio event. First, the sea-surface warming killed most of the P . elegans colonies. Second, sea urchins and other bioeroders either removed or toppled most of the pre-1941 colonies, leaving few standing at the time of the uplift. Third, P . elegans recruited on the uplift, but, removed from areas of intense urchin grazing, they survived only on the upper surface of lava outcrops, and fourth, nearby settlement of additional recruits (Lewis, 1974) produced discrete clusters of P . e l e g a n s colonies. If such events are repeated, coral skeletons do not accumulate to form a
210
Fig. 18. Colonies of Pocillopora elegans that settled on the uppermost surface of a basalt outcrop. Left-hand scale bar = 10 cm.
I
Mean Width (Cm)
Fig. 19. Size distribution of 404 Pocillopora elegans colonies from five different locations on the uplift.
21 1
permanent framework. Coral reef size is small because growth occurs only during a short time span between El Niiio events: growth after each event is forced to a new location while the reef at the old location is destroyed. 6.1.2 Massive corals During the 1982-1983 El Niiio event, the upper surfaces of massive corals Porites lobata, Pavona clavus and Pavona gigantea were killed whereas the shaded sides and undersurface of some of the colonies commonly remained alive (Glynn, 1984, this volume, and pers. obs.). A similar response to the El Niiio event in the early 1940s is suggested by the massive corals present on the uplifted surface at Urvina Bay. If El Niiio warming has not killed the entire colony, regrowth from the surviving patches can cover the dead surface and heal the colony (Hughes and Jackson, 1980). Because the dead surface first becomes encrusted with algae and epifauna, subsequent growth encloses this within the colony where it forms an easily recognized scar (Fig. 20). Most of the massive colonies at Urvina Bay show this scar with an average of 6.0 cm (s=2.18, n=108, range of 2 cm to 12 cm) of growth above the hiatus. For Porites lobata, this amount of growth represents approximately 7.4 years of growth (growth rate of 0.81 cm/yr [Glynn and Wellington, 1983]), indicating that growth occurred after the 1941 El Niiio event ended. In the aftermath of El Niiio events, corals that suffered partial mortality may be colonized by sessile epifaunal filter feeders such as vermetid molluscs and barnacles and/or invaded by bioeroders like Lithophaga mussels, and sponges (Fig. 20). Bioeroding organisms remove calcium carbonate, which weakens the coral’s structural integrity (Scott and Risk, 1988). Lithophaga grew ubiquitously within the massive coral skeletons. For example, in one sample area, 97 % of 119 P . lobata colonies contained lithophagine boreholes with an average of 6.6 holes1100 cm2. El NiRo events that kill portions of a massive colony would favor the infestation of some boring bivalves that preferably bore dead corals (Warme, 1977; Kleemann, 1982: Jones and Pemberton, 1988). In massive Pavona clavus colonies, the density of Lithophaga holes was greatest in the dead portions of the colony (25/100 cm2, s=19.8, n=40: F[1,1581=92.86, p
212
Fig. 20. Regrowth from the sides of a colony of Porites lobata overtopping sessile organisms and the dead upper portion of the colony. Left-hand scale bar = 10 cm. colonies, with 12.4 urchinslm2, compared to 7.8 urchinslm2 on live colonies. Sea urchins erode the skeletons of massive corals as they do branching corals, and remove large quantities of carbonate (Glynn and Wellington, 1983). This prevents coral recruits from surviving on the dead corals (Glynn, this volume). On colonies where the outer surface is alive, sea urchins may attack the colony from the underside. The scraping of sea urchins hollows-out the colony leaving a thin outer shell of living coral that is then easily toppled (Fig. 22). In one sample area at Urvina Bay, urchins hollowed-out between 50 to 80% of the total skeletal volume of Pavona clavus, and between 6 to 12 sea urchins were found within each coral colony (Table 2). Bioerosion also leads to extensive fragmentation of coral colonies (Highsmith, 1980, 1981). For example, at Urvina Bay a single, isolated large colony of Pavona clavus, whose base covered 3.25 m2, had collapsed and fragmented into 121 smaller remnants (mean width=28.8 cm, s=l9.7) scattered over 11.05 m2. At the time of the uplift, 32 % of the fragments were alive, and therefore represent successful asexual propagation. Fragmentation that spawns many small fragments may involve toppling
213
Fig. 21. Sea urchin burrows pit a massive colony of Porites lobata (mean colony diameter = 5.5 m and mean height = 2 m). The widths of the burrows range from 3 cm to 30 cm. TABLE 2 The amount of coral missing from the interior of five massive Pavona c l a v u s colonies (see Fig. 22). Presented are the mean dimensions in cm of the coral colonies, the percent eroded from the interior, and the number of E. thouarsii tests that were present in each colony. Colony Number
Length (cm)
Width ( cm )
Height ( cm )
%Coral Missing
Number of E. thouarsii
1 2 3 4
240 146 117 110 156
200 135 94 140 173
57.6 56.8 41.2 53.1 49.3
50.7 71.3 70.2 59.9 46.8
7 6 6 12 8
5
large portions of colonies (Fig. 14). Without recruitment, large coral colonies, as consequences of intense bioerosion, will be destroyed rendering them unavailable as a platform for
214
Fig. 22. Massive colonies of Pavona clavus that have been hollowed out by the grazing of sea urchins (see Table 2). continued carbonate accumulation (Fig. 23). At Urvina Bay, long-term erosion reduced 39 large (74 cm to 1,249 cm wide), massive colonies (i.e.. Porites lobata, Pavona clavus, and Pavona gigantea) to thin traces rising on average between 9 to 17 cm above the sediment (Table 3). 6.2 Summary remarks Urvina Bay's small reefs, like most in the eastern Pacific, apparently owe their size to several factors: seasonal upwelling, low reproductive success and recruitment, intense bioerosion, and periodic ENS0 warming events. Upwelling of nutrient-rich waters promoted the growth of algae and phytoplankton leading to an abundance of filter feeders, suspension feeders, and grazing organisms. Important among filter feeders were boring bivalvcs
215
Fig. 23. A portion of a Porites lobara colony (mean diameter = 8.5 m) most probably levelled by bioerosion (see Table 3). Arrows point to a section of the perimeter (approximately 1.5 m wide). TABLE 3 Widths and heights in crn of levelled massive corals. Expected heights calculated by a simple regression of mean height versus mean diameter of 51 massive corals (height = (0.476 * width) + 10.54). Coral Species and Locations
Mean Width
Mean Height
( cm )
Expected Height
n
( cm )
580
(164 -1,249)
9
(5 - 14)
287
7
302
(128 -
579)
11
(5 - 15)
154
4
160 450 444
( 52 ( 74 -
293) 955) (187 -1,174)
9 17 9
(5 - 12) (5 - 45) (1 - 18)
86 224 222
10 6 12
Range
Range
Pavona clavus Lagoon Pavona gigantea Lagoon Porites lobata Dead donkey Coral line Lagoon
and clionid sponges, that can weaken a coral's skeleton, and among grazing organisms were sea urchins and parrot fish that erode coral colonies. At Urvina Bay, as in other eastern Pacific localities, the coral communities
216
apparently survived near their physiologic limits. Stressed by the cool, nutrient-rich, upwelling waters that slow coral growth (Glynn and D'Croz, in press), and isolated from a large reproductive stock, these coral communities were small and species poor and were unable to form substantial frameworks. An El Nifio warming event may then have caused additional stresses killing many coral colonies. In addition, those organisms that survived El Niiio warming, such as sea urchins, destroyed existing coral build-ups and excluded new coral recruits, thus denying a foundation for renewed framework growth and reef recovery. The constant cycle of El Nifio perturbations and bioerosion scattered coral recruits and prevented corals from growing and maintaining a reef framework. 7 DISCUSSION In concert with El Nifio events, short-term, unseasonably cool upwelling events and long-duration, glacial episodes (e.g., the Little Ice Age) have disturbed eastern Pacific reef building. The most recent short-term, cooling episode occurred in 1988 when unseasonably intense upwelling lowered sea surface temperatures by 10'-ll°C and produced a 10% coral mortality in the Gulf of Panama (Glynn and D'Croz, in press). However, to underscore the lethal effects of warming events, the 1982-1983 El Niiio event raised temperatures 3"-4OC and caused between 5 to 9 times more coral mortality than occurred during the 1988 cooling event (Glynn and D'Croz, in press). Six longer duration cooling events, lasting several hundred years, have occurred in the last 7.000 years (Grove, 1988). The latest of these glacial episodes, the Little Ice Age, lowered global temperatures 1°-2"C (Grove, 1988). In the Gulf of Papagayo, Costa Rica, during the Little Ice Age (150 to 300 years ago), cool waters killed many corals (Glynn et al., 1983). However, large coral colonies in the Galapagos Islands (Dunbar et al., 1988) and in Panama (Glynn, 1985), that survived this cooling episode, attest to the limited impact of the Little Ice Age coolings in the tropical eastern Pacific. Another region where there are low diversity reef coral communities occurs along the west coast of Africa in the eastern tropical Atlantic (Stehli and Wells, 1971). As in the eastern Pacific, El Niiio-type warming events occur along the western African coast (Glynn, 1988a). and likewise may also negatively affect reef growth. These warming events, known as the Benguela Nifio, cause similar ecologic disturbances, although they are less intense and occur less frequently than those in the eastern Pacific (Duffy et al., 1984; Shannon et al., 1984, 1986; Duffy, this volume). In most reef building regions following extensive, catastrophic coral mortality, rapid recovery occurs through the growth of surviving corals, recruitment of new corals, and survival of the reef framework (Pearson,
21 7
1981; Colgan, 1987). In the eastern Pacific, however, the return of intensified upwelling of nutrient-rich waters (anti-El Niiio type conditions) and the survival of coral predators (Glynn, this volume) hampers recovery of surviving coral reefs by reducing recruitment success and eroding the reef framework. Two important coral predators influence eastern Pacific reef development, Acanthaster planci in Panama (Glynn, 1976), and Eucidaris thouarsii in the Galapagos Islands (Glynn et al., 1979). In the aftermath of the 1982-1983 El Niiio event, A . planci and E . thouarsii survived and continued preying on the few surviving corals causing secondary El Niiio related mortality (Glynn, 1984; 1985; 1988b). After the primary and secondary mortalities resulting from the 19821983 El Nifio disturbance, upwelling of cool, nutrient-rich waters resumed (Hansen, this volume). Seasonally cool waters retard coral growth (Stoddart, 1969; Jokiel and Coles, 1977; Glynn and D’Croz, in press) as do increased nutrients possibly by suppressing calcification (Kinsey and Davies, 1979), or indirectly by stimulating the growth of benthic algae and other organisms that compete for space with surviving coral colonies and new recruits (Birkeland, in press). After coral death, algae cover the skeleton, and are cropped by grazing herbivores, which also erode the coral (Glynn et al., 1972; Birkeland, in press). Increased primary productivity, also, favors suspension and filter feeding bioeroders (Birkeland, 1977; Highsmith, 1980, 1981; Hallock and Schlager, 1986). Unlike some areas in the western Pacific, where grazing enhances coral recruitment, the intensity of eastern Pacific sea urchin predation excludes coral recruits (Glynn, this volume). The near elimination of coral recruits from dead colonies, impedes the rapid accumulation of carbonate and abandons the existing build-ups to the forces of biological and mechanical erosion. While some build-ups are being torn down, coral recruits may settle at new locations, where grazing pressures are less, thereby moving the site where corals grow and reefs develop. Thus, eastern Pacific reefs are short-lived structures growing between major perturbations without a continuity of space and time necessary for the formation of large reefs and a good fossil reef record. The fossil record has long been recognized as imperfect (Darwin, 1859), but the lack of data, at times, is revealing (Eldredge and Gould, 1972). Since the Pliocene, the fossil reef record of the eastern Pacific is very poor (Dana, 1975), and the absence of a reef fossil record might reflect the cycle of death and bioerosion confronting reef communities during all high, sea-level intervals. When sea-level lowers, the small dispersed build-ups easily erode
218
leaving neither a foundation for future growth nor an extensive paleontologic record of their existence. The low species diversity of eastern Pacific coral reefs has been explained as a result of the distance from central and western Pacific coral source areas (Stehli and Wells, 1971). However, for over 7,000 years sealevel and currents have favored surface transport of long-lived, coral larvae across the Pacific basin. Even corals with shorter larval life-spans should have access to the eastern Pacific by transport on pieces of flotsam (Jokiel, 1984; Richmond, this volume). In the last ten years, two new coral species (Acropora valida and Porites (Synaraea) rus) from the western Pacific were discovered on the shores of Central and South America (Cortes and Murillo, 1985; Prahl and Mejia, 1985), but, during the 1982-1983 El Niiio event, these latest presumed colonists may have died. Other ephemeral western Pacific coral colonists include a Montipora colony, collected probably from La Paz (Baja California), and an Acropora colony, collected from the Galapagos Islands, both in the nineteenth century; neither have been found again in the eastern Pacific (Durham, 1947; Wells, 1983; respectively). Repeated lethal El Niiio events may thus favor the continued existence of corals that survived previous events (Vermeij, this volume). As a result, species richness in the eastern Pacific may not increase rapidly, rather the number of species may decline or stabilize.
8 CONCLUSION “It may be said that natural selection is daily and hourly scrutinising” (Darwin, l859), or is it? The structure and composition of long-lived ecosystems are historical consequences of past and present biological and physical forces. Often, observable and testable daily processes that “scrutinize” the population are, by default, determined as the primary factors that ultimately structure a community. However, rare but intense physical and biological disturbances are also potent determinants of a community’s ultimate composition and structure (e.g., Connell, 1978; Paine and Levin, 1981; Connell and Keough, 1985; Moran, 1986). The intense 1982-83 ENSO event was one such uncommon event that perturbed the long-lived eastern Pacific coral reef ecosystem. The recurrence of intense ENSO events brings new perspective to the history of eastern Pacific reef building. Since the closure of the Panamanian seaway and the onset of glacial cycles, the eastern Pacific has faced two different climatic states that restricted reef growth and development: one during glacial periods with cool waters and lowered sea-levels, and the other during interglacial periods with higher sea-levels, warmer waters, and ENSO events. During glacial periods in the eastern Pacific, lowered sea-level
219
restricts all potential reef growth to the edge of the continent. However, intense upwelling doubled productivity (De Vries and Schrader, 1981; Reimers and Suess, 1983; Schramn, 1985; Lyle et al., 1988) further restricting or preventing reef development (Kinsey and Davies, 1979; Hallock and Schlager, 1986; Hallock, 1988). Because of these harsh environmental conditions, possibly few if any hermatypic eastern Pacific coral species survived the glacial sea-level lowerings. On the exposed shelf, the small carbonate build-ups from the previous high sea-level stands eroded in the subaerial environment removing the previous carbonate accumulation (Harris and Matthews, 1968; Matthews, 1968; Bathurst, 1971; Longman, 1980). During the transition to inter-glacial times, melting ice raised sea-level and improved the environmental conditions for coral-reef growth thus enabling coral recruits to colonize the shelf. In the western and central Pacific, where previous high stand reefs survived subaerial exposure, corals colonized and built new reefs on these antecedent structures enabling carbonate accumulation to continue (Goreau, 1969; Stoddart, 1973; Purdy, 1974; Davies, 1983). Whereas in the eastern Pacific, where small reefs eroded during low sea-level stands, recruits had to settle on basaltic or other consolidated outcrops and not upon previous carbonate build-ups (Glynn, 1976). New recruits faced not only harsh reef building conditions (e.g., upwelling and intense grazing) but recurrent intense El Niiio events, which caused repeated coral death. After coral mortality, bioerosion removed much of the coral build-up. This repeated process prevents the coral community from increasing in diversity or developing to a size that can withstand El Niiio related erosion. Thus, one generation's growth is not transferred to the next, and large, persistent reef frameworks are not constructed. 9 ACKNOWLEDGEMENTS For assistance in the field, I thank, Linda Anderson, William Anderson, Sai'n Chai Colgan, Rene Espinosa, David Hollander, E. Allison Kay, Melissa Keppel, LCo F. Laporte, Margaret Laporte, David Malmquist, and Tom Smalley. I thank the following for their assistance in the Galapagos Islands, Sylvia Harcourt, Hank Kasteleijin, and Gunther Reck at the Charles Darwin Research Station, Juan Black at the Charles Darwin Foundation, and Miguel Cifuentes and Humberto Ochoa at Parque Nacional Galapagos. Research support was also given by Rob Dunbar, Gene Gonzales, Sarah Griscom, Keith Howard, Edge Lopez, R. Larry Phillips, and Gerard Wellington. TAME, an Ecuadorian national airline, subsidized air travel to the Galapagos Islands.
220
Discussions with Margaret Dalaney, Peter Glynn, Sarah Gray, Leo Laporte, David Malmquist, and Don Potts and the comments of three anonymous reviewers greatly improved the clarity of this manuscript. This research was funded by NSF award EAR 8508966 and supported by NSF award OCE 8415615 to Peter W. Glynn. Charles Darwin Foundation contribution Number 449.
10 REFERENCES Aharon, P. and Veeh, H. H., 1984. Isotope studies of insular phosphates explain atoll phosphorization. Nature, 309: 614-617. Allison, E. C., 1959. Distribution of Conus on Clipperton Island. Veliger, 1 : 32-34. Barber, R. T. and Chaves, F. P., 1983. Biological consequences of El Nifio. Science, 222: 1.203-1,210. Barber, R. T. and Chaves, F. P., 1986. Ocean variability in relation to living resources during the 1982-1983 El Niiio. Nature, 319: 279-285. Bathurst, R. G. C., 1971. Carbonate Sediments and Their Diagenesis, 2nd Edition. Elsevier, Amsterdam, xix + 658 pp. Berger, W. H., 1982. Climate steps in ocean history - lessons from the Pleistocene. In: W. H. Berger and J. C. Crowell (Editors), Climate in Earth History. National Academy Press, Washington D. C., pp. 26-42. Berggren, W. A., 1982. Role of ocean gateways in climatic changes. In: W. H. Berger and J. C. Crowell (Editors), Climate in Earth History. National Academy Press, Washington D. C., pp. 118-126. Berggren, W. A. and Hollister, C. D., 1974. Paleogeography, paleobiogeography and the history of circulation in the Atlantic. In: W. W. Hay (Editor), Studies in Paleo-oceanography. SOC.Econ. Paleo. Mineral. Spec. Pub., 20: 126-186. Birkeland, C., 1977. The importance of biomass accumulation in the early successional stages of benthic communities to the survival of coral recruits. Proc. 3 d Int. Coral Reef Symp., 1 : 15-21. Birkeland, C., 1982. Terrestrial runoff as a cause of outbreaks of Acanthaster planci (Echinodermata: Asteroidea). Mar. Biol., 69: 175-185. Birkeland, C., In press. Geographic comparisons of coral-reef community processes. Proc. 6l.h Int. Coral Reef Symp. Bjerknes, J., 1969. Atmospheric teleconnections from the equatorial Pacific. Mon. Weath. Rev., 97: 163-172. Bjerknes, J., 1972. Large-scale atmospheric response to the 1964-65 Pacific equatorial warming. J. Phys. Oceanogr., 15: 1,255-1.273. Bowler, J. M., Hope, G. S., Jennings, J. N., Singh, G. and Walker, D., 1976. Late Quaternary climates in Australia and New Guinea. Quatern. Res., 6: 359394. Bradbury, J. P., 1989. Late Quaternary lacustrine paleoenvironments in the Cuenca de Mexico. Quatern. Sci. Rev., 8: 75-100.
22 1
Buddemeier, R.W. and Kinzie, R., 1976. Coral growth. Oceanogr. Mar. Biol. Annu. Rev., 14: 183-225. Cane, M. A., 1983. Oceanographic events during El Niiio. Science, 222: 1,189-1,194. Cane, M. A., 1986. El Niiio. Annu. Rev. Earth Planet. Sci., 14: 43-70. Caviedes, C. N., 1984. El Niiio 1982-1983. Geogr. Rev., 74: 267-290. Chappell, J., 1983. Sea level change and coral reef growth. In: D. J. Barnes (Editor), Perspectives on Coral Reefs. Aust. Inst. Mar. Sci., Manuka, pp. 69-94. Clark, B. L. and Durham, J. W., 1946. Eocene faunas from the department of Bolivar, Colombia. Geo. Soc. Am. Mem., 16: 1-126. CLIMAP Members, 1976. The surface of the ice-age earth. Science, 191: 1,131-1,137. CLIMAP Members, 1981. Seasonal reconstruction of the earth's surface at the last glacial maximum. Geol. SOC.Am. Map & Chart Ser., MC-36. COHMAP Members, 1988. Climatic changes of the last 18,000 years: observations and model simulations. Science, 241 : 1,043-1,052. Colgan, M. W., 1987. Coral reef recovery on Guam (Micronesia) after catastrophic predation by Acanthaster. Ecology, 68: 1,592-1,605. Colgan, M. W. and Malmquist, D., 1987. The Urvina Bay Uplift: a dry trek through a coral community. Oceanus, 30: 61-66. Colinvaux, P. A., 1972. Climate and Galapagos Islands. Nature, 240: 17-20. Colinvaux, P. A., 1984. The Galapagos climate: past and present. In: R. Perry (Editor), Key Environments, Galapagos. Pergamon Press, Oxford, pp. 5 5 70. Connell, J. H., 1978. Diversity in tropical rain forests and coral reefs. Science, 199: 1,302-1,310. Connell, J. H. and Keough, M. J., 1985. Disturbance and patch dynamics of subtidal marine animals on hard substrata. In: S. T. A. Pickett and P. S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, pp. 125-151. Cortes, J. and Murillo, M. M., 1985. Comunidades coralinas y arrecifes del Pacific0 de Costa Rica. Rev. Biol. Trop., 33: 197-202. Cortes, J. and Risk, M. J., 1985. A reef under siltation stress: Cahuita, Costa Rica. Bull. Mar. Sci., 36: 339-356. Couffer, J. C., 1956. The disappearance of Urvina Bay. Nat. Hist., 65: 78-83. Damuth, J. E. and Fairbridge, R. W., 1971. Equatorial Atlantic deep-sea arkosic sands and Ice-Age aridity in tropical South America. Geol. SOC. Am. Bull., 81: 189-206. Dana, T. F., 1975. Development of contemporary eastern Pacific coral reefs. Mar. Biol., 33: 355-374. Darwin, C., 1859. On the Origin of Species. John Murray, London, x + 490 pp. Davies, P. J., 1983. Reef growth. In: D. J. Barnes (Editor), Perspectives on Coral Reefs. Aust. Inst. Mar. Sci., Manuka, pp. 69-94. Davies, P. J. and Marshall, J. F., 1980. A model of epicontinental reef growth. Nature, 287: 37-38. Davies, P. J., Symonds, P. A., Freary, D. A. and Pigram, C. J., In Press. The evolution of the carbonate platforms of Northeast Australia. SOC. Econ. Paleo. Mineral. Spec Pub.
222
Deser, C. and Wallace, J. M.. 1987. El Niiio events and their relation to Southern Oscillation: 1925-1986. J. Geophys. Res., 92: 14,471-14,479. De Vries, T. J., 1987. A review of geological evidence for ancient El Niiio activity in Peru. J. Geophys. Res., 92: 471-479. De Vries, T. J. and Schrader, H., 1981. Variation of upwelling/oceanic conditions during the latest Pleistocene through Holocene off the Central Peruvian coast: a diatom record. Mar. Micropaleo., 6: 157-167. Dodge, R.E., Aller, R.C. and Thomson, J., 1974. Coral growth related to resuspension of bottom sediments. Nature, 247: 574-577. Dodge, R. E. and Vaisnys, J. R., 1977. Coral populations and growth patterns: responses to sedimentation and turbidity associated with dredging. J. Mar. Res., 35: 715-730. Druffel, E. M., 1985. Detection of El Niiio and decade time scale variations of sea surface temperature from banded coral records: implications for the carbon dioxide cycle. Geophys. Monogr., 32: 111- 121. Duffy, D. C., Berruti, A., Randall, R. M. and Cooper, J., 1984. Effects of the 1982-1983 warm water event on the breeding of South African seabirds. S. Afr. J. Sci., 80: 65-68. Dunbar, R. B., Wellington, G., Colgan, M. and Glynn, P. W., 1988. Galapagos massive corals: an annual record of eastern tropical Pacific climate during the past 400 years. EOS, 68: 1,743. Durham, J. W., 1947. Corals from the Gulf of California and the north Pacific coast of America. Geol. SOC.Am. Mem., 20: 1-68. Durham, J. W., 1966. Coelenterates, especially stony corals, from the Galapagos Islands. In: R. I. Bowman (Editor), The Galapagos, proceedings of the symposia of the Galapagos International Scientific Project. Univ. Calif. Press, Berkeley, pp. 123-135. Durham, J. W. and Allison, E. C., 1960. The geologic history of Baja California and its marine faunas. Syst. Zool., 9: 47-91. Eakin, C. M., 1987. Damselfish and their algal lawns: a case of plural mutualism. Symbiosis, 4: 275-288. Eakin, C. M., In press. Avoidance of damselfish lawns by the sea urchin Diadema mexicanum at Uva Island, Panama. Proc. 6th Int. Coral Reef SYmP* Eldredge, N. and Gould, S. J., 1972. Punctuated equilibria: an alternative to phyletic gradualism. In: T. J. Schopf (Editor), Models in Paleobiology. Freeman, Cooper, and Co., San Francisco, pp. 82-115. Enfield, D. B., 1987. Progress in understanding El Niiio. Endeavour, 11: 197204. Enfield, D. B., 1989. El Niiio, past and present. Rev. Geophys., 27: 159-187. Fagerstrom, J. A., 1987. The Evolution of Reef Communities. John Wiley & Sons, New York, 628 pp. Frost, S. H., 1977. Cenozoic reef systems of Caribbean- prospects for paleoecologic synthesis. In: S. H. Frost, M. P. Weiss and J. B. Saunders (Editors), Reefs and Related Carbonates--Ecology and Sedimentology. Am. Assoc. Petrol. Geol. Stud. Geol., 4: 93-110. Frost, S. H. and Langenheim, R. L., 1974. Cenozoic Reef Biofacies. North. Illinois Univ. Press, De Kalb, xii + 388 pp.
223
Gartner, S., Chow, J. and Stanton, R. J., 1987. Late Neogene paleoceanography of the eastern Caribbean, the Gulf of Mexico, and the eastern equatorial Pacific. Mar. Micropaleo., 12: 255-306. Gill, A. E. and Rasmusson, E. M., 1983. The 1982-1983 climate anomaly in the equatorial Pacific. Nature, 306: 229-234. Glynn, P. W., 1974. Rolling stones among the Scleractinia: Mobile coralliths in the Gulf of Panama. Proc. Int. Coral Reef Symp., 2: 183-197. Glynn, P. W., 1976. Some physical and biological determinants of coral community structure in the eastern Pacific. Ecol. Monogr., 46: 43 1-456. Glynn, P. W., 1977. Coral growth in upwelling and nonupwelling areas off the Pacific coast of Panama. J. Mar. Res., 35: 567-587. Glynn, P. W., 1980. Defense by symbiotic crustacea of host corals elicited by chemical cues from predator. Oecologia, 47: 287-290. Glynn, P. W., 1983a. Extensive 'bleaching' and death of coral reefs on the Pacific coast of Panama. Environ. Conserv., 10: 149-154. Glynn, P. W., 1983b. Crustacean symbionts and the defense of corals: coevolution on the reef? In: M. H. Nitecki (Editor), Coevolution. Univ. Chicago Press, Chicago, pp. 11 1-178. Glynn, P. W., 1984. Widespread coral mortality and the 1982-83 El Niiio warming event. Environ. Conserv., 11: 133-146. Glynn, P. W., 1985. El Niiio-associated disturbance to coral reefs and post disturbance mortality by Acanthaster p l a n c i . Mar. Ecol. Prog. Ser., 26: 295-300. Glynn, P. W., 1988a. El Nifio-Southern Oscillation 1982-1983: nearshore population, community and ecosystem responses. Annu. Rev. Ecol. Syst., 19: 309-345. Glynn, P. W., 1988b. El Niiio warming and reef framework destruction by echinoid bioerosion in the eastern Pacific. Galaxea, 7: 129-160. Glynn, P. W. and D'Croz, L., In press. Experimental evidence for high temperature stress as the cause of El Niiio-coincident coral mortality. Coral Reefs. Glynn, P. W. and Macintyre, I. G., 1977. Growth rate and age of coral reefs on the Pacific coast of Panama. Proc. 31h Int. Coral Reef Symp., 2: 251259. Glynn, P. W. and Stewart, R. H., 1973. Distribution of coral reefs in the Pearl Islands (Gulf of Panama) in relationship to thermal conditions. Limnol. Oceangr., 18: 367-379. Glynn, P. W. and Wellington, G. M., 1983. Corals and coral reefs of the Galapagos Islands. Univ. Calif. Press, Berkeley, xvi + 330 pp. Glynn, P. W., Druffel, E. M. and Dunbar, R. B., 1983. A dead Central American coral reef tract: possible link with the Little Ice Age. J. Mar. Res., 41: 605637. Glynn, P. W., Perez, M. and Gilchrist, S. L., 1985a. Lipid decline in stressed corals and their crustacean symbionts. Biol. Bull., 168: 276-284. Glynn, P. W., Peters, E. C. and Muscatine, L., 1985b. Coral tissue microstructure and necrosis: relation to catastrophic coral mortality in Panama. Dis. Aquat. Org., 1: 29-37.
a
224
Glynn, P. W., Stewart, R. H. and McCosker, J. E., 1972. Pacific coral reefs of Panama: structure, distribution and predators. Geol. Rundschau, 61 : 483519. Glynn, P. W., Wellington, G. M. and Birkeland, C., 1979. Coral reef growth in the Galapagos: limitation by sea urchins. Science, 203: 47-49. Glynn, P. W., Cortes, J., Guzman, H. M. and Richmond, R. H., In press. El Niiio (1982-1983) associated coral mortality and relationship to sea surface temperature deviations in the tropical eastern Pacific. Proc. 6 h Int. Coral Reef Symp. Goreau, T. F., 1969. Post Pleistocene urban renewal in coral reefs. Micronesica, 5: 323-326. Grove, J., 1984. At the heart of El Niiio. Oceans, 17(3): 3-8. Grove, J. M., 1988. The Little Ice Age. Methuen, London, xxii + 498 pp. Grigg. R. W., 1988. Paleoceanography of coral reefs in the Hawaiian-Emperor chain. Science, 240: 1,737-1.743. Grigg, R. W. and Epp, D., 1989. Critical depth for the survival of coral islands: effects on the Hawaiian Archipelago. Science, 243: 638-640. Guzman, H. M., and Cortes, J., 1989. Coral reef community structure at Caiio Island, Pacific Costa Rica. P. S. Z. N. I.: Mar. Ecol., 10: 23-41. Guzman, H. M., Cortes, J., Richmond, R. H. and Glynn, P. W., 1987. Efectos del fenomeno de “El Niiio-Oscilacion Sureiia” 1982/83 en 10s arrecifes coralinos de la Isla del Caiio, Costa Rica. Rev. Biol. Trop., 35: 325-332. Hallock, P., 1988. The role of nutrient availability in bioerosion: consequences to carbonate buildups. Palaeogeogr., Palaeoclimat., Palaeoecol., 63: 275-291. Hallock, P. and Schlager W., 1986. Nutrient excess and the demise of coral reefs and carbonate platform. Palaios, 1: 389-398. Harris, W. H. and Matthews, R. K., 1968. Subaerial diagenesis of carbonate sediments: efficiency of solution-reprecipitation process. Science, 160: 77-79. Hays, P. E., Pisias, N. G. and Roelofs, A. K., 1989. Paleoceanography of the eastern equatorial Pacific during the Pliocene: a high-resolution radiolarian study. Paleoceanogr., 4: 57-73. Highsmith, R. C., 1980. Geographic patterns of coral bioerosion: a productivity hypothesis. J. Exp. Mar. Biol. Ecol., 117: 193-198. Highsmith, R. C., 1981. Coral bioerosion: damage relative to skeletal density. Am. Nat., 117: 193-198. Hoffmeister, J. E. and Ladd, H. S., 1944. The antecedent platform theory. J. Geol., 53: 388-402. Hopley, D., 1982. The Geomorphology of the Great Barrier Reef: Quaternary Development of Coral Reefs. John Wiley & Sons Inc., New York, xiv + 453 PP. Hubbard, J. A. E. B. and Pocock, Y. P., 1972. Sediment rejection by recent scleractinian corals: a key to palaeoenvironmental reconstruction. Geol. Rundschau, 61 : 598-626. Hughes, T. P. and Jackson, J. B. C., 1980. Do corals lie about their age? Some demographic consequences of partial mortality, fission and fusion. Science, 209: 713-715.
225
Hutchings, P. A., 1986. Biological destruction of coral reefs. Coral Reefs, 4: 239-252. Hutchinson, G. E., 1952. The biogeochemistry of phosphorus. In: L. F. Wolterink (Editor), The Biology of Phosphorus. Michigan State College Press, pp. 1-35. Imbrie, J. and Imbrie, K. P., 1979. Ice Ages Solving the Mystery. Enslow Pub., Short Hills, N. J., 224 pp. James, N. P., 1983. Reef environments. In: P. D. Scholle, D. G. Bebout and C. H. Moore (Editors), Carbonate Depositional Environments. Am. Assoc. Petrol. Geol. Mem., 33: 345-440. Jokiel, P. L., 1984. Long distance dispersal of reef corals by rafting. Coral Reefs, 3: 113-116. Jokiel, P. L. and Coles, S. L., 1977, Effects of temperature on the mortality and growth of Hawaiian reef corals. Mar. Biol., 43: 201-208. Jones, B. and Pemberton, S. G., 1988. Lithophaga boring and their influence on the diagenesis of corals in the Pleistocene Ironshore Formation of Grand Cayman Island, British West Indies. Palaios, 3: 3-21. Kauffman, E. G. and Johnson, C. C., 1988. The morphology and ecological evolution of middle and upper Cretaceous reef-building rudistids. Palaios, 3: 194-217. Keigwin, L. D., 1978. Pliocene closing of the Isthmus of Panama based on biostratigraphic evidence from nearby Pacific Ocean and Caribbean Sea cores. Geology, 6: 630-634. Keller, G., Zenker, C. E. and Stone, S. M., 1989. Late Neogene history of the Pacific-Caribbean gateway. J. South Amer. Earth Sci., 2: 73-108. Kennett, J. P., Keller, G. and Srinivasan, M. S., 1985. Miocene planktonic foraminifera1 biogeography and paleoceanographic development of the Indo-Pacific region. Geol. SOC. Am. Mem., 163: 197-236. Kershaw, A. P., 1976. A late Pleistocene and Holocene pollen diagram from Lynch’s Crater, northeastern Queensland, Australia. New Phytol., 77: 469498. Kershaw, A. P., 1978. Record of last interglacial-glacial cycle from northeastern Queensland. Nature, 272: 159- 16 1 . Kinsey, D. W. and Davies, P. J., 1979. Effects of elevated nitrogen and phosphorus on coral reef growth. Limnol. Oceanogr., 24: 935-940. Kleemann, K. H., 1980. Boring bivalves and their host corals from the Great Barrier Reef. J. Moll. Stud., 46: 13-54. Ladd, H. S., 1973. Bikini and Eniwetok atolls, Marshall Islands. In: 0. A. Jones and R. Endean (Editors), Biology and Geology of Coral Reefs. Vol. 1 , Geol. 1 , Academic Press, New York, pp. 93-112. Lessios, H. A., Glynn, P. W. and Robertson, D. R., 1983. Mass mortalities of coral reef organisms. Science, 222: 715. Lewis, J. B., 1974. Settlement and growth factors influencing the contagious distribution of some Atlantic reef corals. Proc. 2lld Int. Coral Reef Symp., 2 0 1 -206. Linn, L. J., 1988. Trace metals in Galapagos corals: quarter-annual and annual cycles of copper, manganese, cadmium, and lead. Masters Thesis, Univ. Calif., Santa Cruz, vi + 86 pp.
226
Longman, M., W., 1980. Carbonate diagenetic textures from near surface diagenetic environments. Am. Assoc. Petrol. Geol. Bull., 64: 461 -487. Lyle, M., Murray, D. W., Finney, B. P., Dymond, J., Robbins, J. M. and Brooksforce, K., 1988. The record of Late Pleistocene biogenic sedimentation in the eastern tropical Pacific Ocean. Paleoceanogr., 3: 3959. Malmquist, D. L., Colgan, M. W. and Espinosa, R. E., 1986. Evidence for catastrophic vs. attritional mortality in the fossil record: Urvina Bay, Isabela Island, Galapagos Archipelago, Ecuador. Geol. SOC. Am. Bull. Prog. Abstr., 18: 680. Markgraf, V., 1989. Palaeoclimates in Central and South America since 18,000 BP based on pollen and lake-level records. Quatern. Sci. Rev., 8: I 24. Matthews, R. K., 1968. Carbonate diagenesis: equilibrium of sedimentary mineralogy to the subaerial environment; coral cap of Barbados, West Indies. J. Sed. Pet., 38: 1,110-1,119. McBirney, A. R. and Williams, H., 1969. Geology and petrology of the Galapagos Islands. Geol. SOC.Am. Mem., 118: 1-197. Miller, L., Cheney, R. E. and Douglas, B. C., 1988. GEOSAT altimeter observations of Kelvin Waves and the 1986-87 El Niiio. Science, 239: 5254. Moran, P. J., 1986. The Acanthaster phenomenon. Oceanogr. Mar. Biol. Annu. Rev., 24: 379-480. Moseley, M. E., Feldman, R. A. and Ortloff, C. R., 1981. Living with crises: human perception of process and time. In: M. Nitecki (Editor), Biotic Crises in Ecological and Evolutionary Time. Academic Press, New York, pp. 231-267. Newell, N. D., 1972. The evolution of reefs. Sci. Am., 226: 54-65. Nials, F. L., Deeds, E., Moseley, M. E., Pozorski, S. G., Pozorski, T. G. and Feldman, R. A., 1979. El Niiio: the catastrophic flooding of coastal Peru. (Part 11). Bull. Field Mus. Nat. Hist., 50: 4-10. Paine, R. T. and Levin, S. A., 1981. Intertidal landscapes: disturbances and the dynamics of pattern. Ecol. Monogr., 51: 145-179. Pang, R. K., 1973. The ecology of some Jamaican excavating sponges. Bull. Mar. Sci., 23: 227-243. Pearson, R. G., 1981. Recovery and recolonization of coral reefs. Mar. Ecol. Prog. Ser., 4: 105-122. Philander, S. G. H., 1983. El Niiio Southern Oscillation phenomena. Nature, 302: 295-301. Porter, J. W., 1974. Community structure of coral reefs on opposite sides of the Isthmus of Panama. Science, 186: 543-545. Potts, D. C., 1984. Generation times and the Quaternary evolution of reefbuilding corals. Paleobiol., 10: 48-58. Prahl, H., von, 1985. Blanqueo masivo y muerte de corales hermatipicos en el Pacifico colombiano, atribuido a1 fenomeno de El Niiio 1982-1983. Bol. Erfen, 12: 22-24. Prahl, H., von and Mejia, A., 1985. Primer informe de un coral acroporido, Acropora valida (Dana, 1846) (Scleractinia: Astrocoeniida: Acroporidae) para el Pacifico americano. Rev. Biol. Trop., 33: 39-43.
227
Purdy, E. G., 1974. Reef configurations: cause and effect. In: L. F. Laporte (Editor), Reefs in Time and Space. SOC.Econ. Paleo. Mineral. Spec. Pub., 18: 9-76. Quinn, W. H., 1971. Late Quaternary meteorological and oceanographic developments in the equatorial Pacific. Nature, 229: 330-33 1. Quinn, W. H., Neal, V. T. and Antunez de Mayolo, S. E., 1987. El Niiio occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14,449-14,461. Rasmusson, E. M., 1984. El Niiio: the ocean/atmosphere connection. Oceanus, 27: 5-12. Rasmusson, E. M., 1985. El Niiio and variation in climate. Am. Sci., 73: 168177. Rasmusson, E. M. and Wallace, J. M., 1983. Meteorological aspects of the El Niiio/Southern Oscillation. Science, 222: 1,195-1,202. Rea, D. K.,Chambers, L. W., Chuey, J. M., Jenecek, T. R., Leinen, M. and Pisias, N. G., 1986. A 420,000-year record of cyclicity in oceanic and atmospheric processes from the eastern equatorial Pacific. Paleoceangr., 1: 577-586. Reimers, C. E. and Suess, E., 1983. Late Quaternary fluctuations in the cycling of organic matter off central Peru: a proto-kerogen record. In: J. Thiede and E. Suess (Editors), Coastal Upwelling: Its Sedimentary Record, Part B: Sedimentary Record of Ancient Coastal Upwelling. Plenum Press, New York, pp 497-526. Robinson, G., 1985. The influence of the 1982-1983 El Niiio on Galapagos marine life. In: G. Robinson and E. M. del Pino (Editors), El Nifio in the Galapagos Islands: the 1982- 1983 Event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 153-190. Romine, K., 1982. Late Quaternary history of atmospheric and oceanic circulation in the late eastern equatorial Pacific. Mar. Micropal., 7: 163187. Romine, K., 1985. Radiolarian biogeography and paleoceanography of the North Pacific at 8 ma. Geol. SOC.Am. Mem., 163: 237-272. Romine, K. and Lombari, G., 1985. Evolution of Pacific circulation in the Miocene: radiolarian evidence from DSDP site 289. Geol. SOC.Am. Mem., 163: 273-290. Romine, K. and Moore, T. C., 1981. Radiolarian assemblage distribution and paleoceanography of the eastern equatorial Pacific Ocean during the last 127,000 years. Palaeogeogr., Palaeoclimat., Palaeoecol., 35: 281-314. Salinger, M. J., 1981. Palaeoclimates north and south. Nature, 291: 106-107. Sammarco, P. W. and Williams, A. H., 1982. Damselfish territoriality: influence on D iadema distribution and implication for coral community structure. Mar. Ecol. Prog. Ser., 8: 53-59. Sandweiss, D. H., 1986. The beach ridges at Santa, Peru: El Niiio, uplift, and prehistory. Geoarch., 1: 17-28. Schramm, C. T., 1985. Implication of radiolarian assemblages for the late Quaternary paleoceanography of the eastern equatorial Pacific. Quatern. Res., 24: 204-218. Scott, P. J. B. and Risk, M. J., 1988. The effect of Lithophaga (Bivalvia: Mytilidae) boreholes on the strength of the coral Porites lobata. Coral Reefs, 7: 145-151.
228
Scott, P. J. B., Risk, M. J. and Carriquiry, J., 1988. El Niiio and the survival of east Pacific reefs. Proc. 6 h Int. Coral Reef Symp., abstr., p. 92. Shannon, L. V., Crawford, R. J. M. and Duffy. D. C., 1984. Pelagic fisheries and warm events - a comparative study. S. Afr. J. Sci., 80: 51-60. Shannon, L. V., Boyd, A. J., Brundrit, G. B. and Taunton-Clark, J., 1986. On the existence of an El Nino-type phenomenon in the Benguela system. J. Mar. Res., 44: 495-520. Shen, G. T., Dunbar, R. B., Colgan, M. and Glynn, P. W., 1988. Equatorial upwelling at Galapagos: Cd record in corals since the 17th century. EOS, 68: 1,744. Stearn, C. W. and Scoffin, T. P., 1974. Carbonate budgets of a fringing reef, Barbados. Proc. 2L?h Int. Coral Reef Symp., 2: 471-476. Stehli, F. G. and Wells, J. W., 1971. Diversity and age patterns in hermatypic corals. Syst. Zool., 20: 115-126. Stoddart, D. R., 1969. Ecology and morphology of recent coral reefs. Biol. Rev., 44: 433-498. Stoddart, D. R., 1973. Coral reefs: the last two million years. Geography, 58: 3 13-323. Stoddart, D. R. and Scoffin, T. P., 1983. Phosphate rock and coral reef islands. In: A. S. Goudie and K. Pye (Editors), Chemical Sediments and Geomorphology: Precipitates and Residua in the Near-Surface , Environment. Academic Press, London, pp. 369-401. Sykes, L. R., McCann, W. R. and Kafka, A.L., 1982. Motion of the Caribbean plate during last 7 million years and implications for earlier Cenozoic movements. J. Geophys. Res., 87: 10,656-10,676. Thom, B. G. and Chappell, J., 1975. Holocene sea levels relative to Australia. Search, 6: 90-93. Thompson, L. G., Mosley-Thompson, E. and Morales, B., 1984. El NiiioSouthern Oscillation events recorded in the stratigraphy of the tropical Quelccaya ice cap, Peru. Science, 226: 50-53. Thurber, D. I., Broecker, W. S., Blanchard, R. L. and Potratz, H. A., 1965. Uranium series ages of Pacific atoll corals. Science, 149: 55-58. UNEPflUCN, 1988. Coral Reefs of the World, Volume 1 : Atlantic and eastern Pacific. UNEP Regional Seas Directories and Bibliographies. IUCN, GlandKambridge, xvii + 373 pp. Vallis, G. K., 1986. El Niiio: a chaotic dynamical system. Science, 232: 243245. Vaughan, T. W.. 1917. The reef-coral fauna of Carrizo Creek, Imperial County, Cal., and its significance. U. S. Geol. SUIT. Prof. Pap., 98T: 355-386. Vaughan, T. W., 1919. Fossil coral from Central America, Cuba, and Porto Rico, with an account of the American Tertiary, Pleistocene. and Recent coral reefs. U. S. Nat. Mus. Bull., 103: 189-524. Veron, J. E. N., 1986. Coral of Australia and the Indo-Pacific. Angus & Robertson Publ., London, xii + 644 pp. Veron, J. E. N. and Kelly, R., 1988. Species stability in reef corals of Papua New Guinea and Indo-Pacific. Mem. Assoc. Australas. Palaeontols., 6: 1 69.
229
Warme, J. E., 1977. Carbonate borers--their role in reef ecology and preservation. In: S. H. Frost, M. P. Weiss and J. B. Saunders (Editors), Reefs and Related Carbonates--Ecology and Sedimentology. Am. Assoc. Petrol. Geol. Stud. Geol., 4: 261-279. Wellington, G. M., 1982. Depth zonation of corals in the Gulf of Panama: control and facilitation by resident reef fishes. Ecol. Monogr., 52: 223-241. Wellington, G. M., 1984. Marine environment and protection. In: R. Perry (Editor), Key Environments, Galapagos. Pergamon Press, Oxford, pp. 247265. Wells, J. W., 1954. Recent corals of the Marshall Islands: Bikini and nearby atolls, Part 2, Oceanography (Biologic). U . S. Geol. Surv. Prof. Pap., 260-1: 385 -486. Wells, J. W., 1956. Scleractinia. In: R. C. Moore (Editor), Treatise on Invertebrate Paleontology. Part F. Coelenterata. Geol. SOC. Am., Univ., Kansas Press, Lawrence, pp. F328-F444. Wells, J. W., 1983. Annotated list of the scleractinian corals of the Galapagos Islands. In: P. W. Glynn and G. W. Wellington, Corals and Coral reefs of the Galapagos Islands. Univ. Calif. Press, Berkeley, pp. 212-295. Wells, L. E., 1987. An alluvial record of El Niiio events from northern coastal Peru. J. Geophys. Res., 92: 14,463-14,470. Williams, A. H., 1979. Interference behavior and ecology of threespot damselfish (Eupomacentrus planiforns). Oecologia, 38: 223-230. Wilson, J. L., 1975. Carbonate Facies in Geologic History. Springer Verlag, New York, xiv + 471 pp. Woodring, W. P., 1957. Geology and paleontology of Canal Zone and adjoining parts of Panama. U. S. Geol. Surv. Prof. Pap., 306-A: 1-145. Wyrtki, K., 1975. El Niiio, the dynamic response of the equatorial Pacific Ocean to atmospheric forcing. J. Phys. Oceangr., 5: 572-584. Wyrtki, K., 1979. The response of sea surface topography to the 1976 El Niiio cycles. J. Phys. Oceangr., 9: 1,223-1,231. Wyrtki, K., 1985. Pacific-wide sea level fluctuations during the 1982-83 El Niiio. In: G. Robinson and E. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983 event. Charles Darwin Foundation for the Galapagos Islands. Quito, Ecuador, pp. 29-49. Yamaguchi, M., 1975. Sea level fluctuation and mass mortalities of reef animals in Guam, Manana Islands. Micronesica, 11: 227-243.
APPENDIX 1 Suborders, families, and genera of hermatypic corals (Scleractinia) and the families and genera of the orders Coenothecalia and Milleporina that have inhabited the eastern Pacific since the Paleocene. IP = Indo-Pacific, EP = Eastern Pacific, WA = Western Atlantic, A = Ahermatypic coral, and 't = extinct. Sources: Vaughan, 1919; Clark and Durham, 1946; Durham, 1947; Wells, 1956; Durham and Allison, 1960; Frost and Langenheim, 1974; Glynn and Wellington, 1983; Veron, 1986; Veron and Kelly, 1988.
N
W 0
APPENDIX 1 (cont’d.) P a l e o c e n e IEocene IPIEPIWAI IP~EP
23 1
N W A
APPENDIX 1 (cont’d.)
Total Genera E. Pacific & W. Atlantic E. Pacific & Indo-Pacific Common to all areas
*
Ahermatypic coral genera not counted.
h,
w
N
233
REEF-BUILDING CORALS AND IDENTIFICATION OF ENSO WARMING EPISODES E.R.M. DRUFFEL Department of Chemistry, Woods Hole Oceanographic Institution, Woods Hole MA 02543 R. B. DUNBAR Department of Geology and Geophysics & Earth Systems Institute, Rice University, Houston TX 77251-1892 G.M. WELLINGTON Department of Biology, University of Houston, Houston TX 77204-5513 S . A. MINNIS Department of Geology and Geophysics & Earth Systems Institute, Rice University, Houston TX 77251-1892
ABSTRACT Druffel, E.R.M., Dunbar, R.B.,Wellington, G.M., and Minnis, S.A. 1989. Reefbuilding corals and identification of ENSO warming episodes. Many chemical and physical changes that occur in the atmosphere and surface ocean are recorded in coral skeletons. In this chapter, we discuss some of the physical, chemical and isotopic changes observed in coral skeletons and their relationships with various climate-related forcing functions. Among the specific factors affecting coral growth and the chemical and isotopic composition of the accreted aragonite are those encountered during ENSO events: abnormal temperature and salinity, reduced nutrient and zooplankton supply, and variation in light intensity. We also determine the usefulness of coral skeletons for reconstructing ENSO records in the eastern tropical Pacific. We present records of stable oxygen and carbon isotopes in corals from the Galapagos Islands and the coast of Panama. We conclude that stable isotopic studies are adequate for tracking most minor and major ENSO events, but are insufficient for unequivocally determining the occurrences of all events or of catastrophic events, such as that which occurred during 1982-83. The catastrophic environmental stresses imposed on corals during such events result in cessation of skeletal accretion, and thus an absence of a significant portion of the record. 1 INTRODUCTION Since the discovery that skeletal density banding patterns in reef-building corals are accreted on an annual basis, analogous to tree rings (Knutson et al., 1972), researchers have been able to use this information to: i) assess variation in colony growth rates (intra- and interspecific: Buddemeier et al., 1974; Baker and Weber, 1975; Highsmith, 1979; Wellington and Glynn, 1983; Huston, 1985); ii) estimate environmental effects on growth (Dodge and Vaisnys, 1975; Weber et al., 1975; Hudson et al., 1976; Schneider and Smith, 1982; Dodge and Lang, 1983; Glynn and Wellington, 1983; Hubbard and Scaturo, 1985); and iii) evaluate the actual and potential affects of human-related
234 perturbations on the environment (Dodge, 1982; Dodge et al., 1984; Dodge and Gilbert, 1984).
In addition, the stable isotopic composition of the coral
skeleton has proven useful in monitoring surface ocean climate, in particular changes in sea surface temperature (Weber and Woodhead, 1972; Fairbanks and Dodge, 1979; Dunbar and Wellington, 1981; Pltzold, 1984; Druffel, 1985).
By
combining isotopically-derived temperature profiles with chronologies available from annual banding, we can derive valuable climatic information on recent El Niiio/Southern Oscillation (ENSO) events and potentially identify the extent and magnitude of similar events over the past several hundred years. In this chapter we show that chronologies of stable isotopes in coral skeletons can be used to construct a long term history for ENSO events in the eastern Pacific. An accurate proxy record is contingent on having both a reliable chronology and an accurate interpretation of the associated environmental conditions. In general, corals have been shown to meet both of these criteria but ambiguities can arise.
For example, while most minor and
major El Niiio events alter coral growth, exceptional ENSO events such as occurred during 1982-83 can impose a severe stress on corals leading to a signature similar to those recordod from other types of stress. To understand the potential confounds involved in analyzing coral banding/isotope patterns, we first review our current understanding of the banding phenomenon and discuss the environmental factors that can influence this process. Later, we present isotopic data from eastern Pacific corals documenting several recent and past ENSO events. 1.1 Characteristics of skeletal growth in symbiotic corals Although morphology of reef-building symbiotic corals is quite variable, ranging from plate-like to highly branched, all corals have apical growth (for review: Buddemeier and Kinzie, 1976).
In corals with a plating morphology,
growth is highest at the edge of the colony while in branching forms it occurs at the tips.
In contrast to other morphologies, growth in massive corals is
more evenly distributed over the entire surface except for the underside, and distinct annual growth bands are accreted along the axis of growth. An annual band is comprised of a high and low density portion representing changes in both the rate of linear skeletal extension and calcification. For massive corals in the eastern Pacific (genus Pavona), the high density band occurs as a result of both reduced skeletal extension and calcification, relative to the low density band (Wellington and Glynn, 1983; see Dodge and Brass, 1984 for a description of banding i n the Caribbean coral, Montastrea). Banding i n symbiotic corals is thought to result from differential rates of protein matrix production (an intracellular process) and calcification (an extracellular process).
Goreau and Hayes (1977) have predicted that periods
235 favoring matrix production, relative to calcification, should lead to the formation of a low density band.
Several attempts have been made to correlate
banding with the physical factors that could influence these processes but as yet no clear cause and effect relationship has been established (Weber et al., 1975; Highsmith, 1979; Wellington and Glynn, 1983).
Past research has focused
primarily on the importance of annual variation in light and temperature since these factors are known to influence the productivity of the endosymbiotic algae (zooxanthellae), which mediate calcification and contribute to tissue growth in corals (Goreau and Goreau, 1959; Clausen, 1971; Chalker and Taylor, 1975; Clausen and Roth, 1975; Coles and Jokiel, 1977; Jokiel and Coles, 1977). Wellington and Glynn (1983) found in Panama that seasonal variation in light intensity provides the most consistent correlation to band formation: low and high density bands are associated with high and low light levels, respectively. However, this relationship is often confounded by other exogenous (e.g., temperature: see Highsmith, 1979) and endogenous (e.g., reproductive periodicity) correlates. Despite the absence of a verified model to account for this banding phenomenon, any seasonal variations (or nonseasonal perturbations) in physical parameters that affect rates of calcification and linear growth (coral tissue and skeleton) in corals should also influence band formation. We briefly review the major physical and biological factors that are known to affect coral growth
--
some of which govern band formation and dictate skeletal isotopic
composition. 1.2 Effects of uhvsical factors on coral erowth Perhaps the best studied physical factor is temperature. In general, reefbuilding corals thrive between 18" and 26°C (Wells, 1957; Stoddart, 1969). Laboratory studies have shown that in branching corals, maximum growth (calcification and metabolic efficiency) coincides with average seasonal high temperatures, and is slowest during cooler water periods (Clausen, 1971; Jokiel and Coles, 1977).
These results have been verified in the field (Glynn, 1977).
For massive corals, however, the opposite has been observed in field studies showing that accelerated growth (calcification and linear skeletal extension) occurs during seasonally cooler periods (Highsmith, 1979; Wellington and Glynn, 1983).
Why this difference exists is not clear, but may be related to a
differential response to increased productivity (i.e., some species may be more reliant on heterotrophic food sources), or reflective of seasonal patterns in spawning that result from the reallocation of energies from growth to reproduction (Wellington and Glynn, 1983).
In any case, under conditions of
normal seasonal variation in temperature, coral skeletons are very useful in both documenting annual variation in growth and as an isotopic thermometer to
236 monitor surface ocean climate (see Fairbanks and Dodge, 1979; Dunbar and Wellington, 1981).
However, under extreme thermal conditions interpreting this
record can be more difficult. When seasonal temperature variations exceed seasonal averages, growth in both massive and branching morphologies can show negative (stress) effects. For example, when sea surface temperatures rose to 30-31°C (3" to 4°C above normal seasonal highs) in the eastern Pacific during the 1982-83 ENS0 event and persisted for several weeks, nearly all the extant reef-building corals in this region experienced partial or complete loss of their endosymbiotic algae (Glynn, 1983, 1984; see Glynn, this volume for details).
Many corals
eventually died, those that survived incurred a hiatus in skeletal growth (see below).
While the physiological basis for this response is not entirely
understood, it is well known that zooxanthellae are critical to the calcification process (Kawaguti and Sakamoto, 1948; Goreau, 1961; Pearse and Muscatine, 1971) and are involved in translocation of photosyntheticallyderived carbon to the coral host (Smith et al., 1969; Muscatine and Cernichiari, 1969).
Occurrences of high water temperatures leading to loss of
zooxanthellae have been cited many times (Yonge and Nicholls, 1931; Jaap, 1979), most recently in the Caribbean (Roberts, 1987, 1988).
A
similar
response has been described for corals exposed to exceptionally cool water events (Hudson et al., 1976; Jaap, 1979; Porter et al., 1982; Glynn et al., 1983), and its banding (Hudson, 1981) and isotopic (Emiliani et al., 1978; Glynn et al., 1983) signatures have been characterized in both massive and branching species. Light is also an important physical parameter that influences skeletal banding and isotopic composition. Lower light levels decrease photosynthetic rates, resulting in an overall decrease in algal productivity. Hence, the amount of carbon translocated to the coral host decreases while the respiratory demands of the algae increase proportionally. Even though algae are capable of efficiently adapting to low light conditions (e.g.,Wethey and Porter, 1976), circumstantial evidence suggests that reduced insolation is at least partially responsible for the lower linear growth rates observed in corals at depth. For example, colonies of massive species collected along a depth gradient show an overall decrease in band width with the high density portion representing a greater proportion of the annual band (Highsmith, 1979).
The absence of
significant temperature changes along this depth gradient indicates that light alone is the probable cause of reduced growth. In addition to observational data, experiments conducted in the field show that shading alone substantially reduces growth (Wellington, 1982).
Expulsion of zooxanthellae and subsequent
death will ensue if shading is severe (Yonge and Nicholls, 1931; Franzisket 1970; Rogers, 1979).
237 A s with temperature, variation in light levels can influence band formation
and subsequent interpretation of environmental conditions. When these parameters are correlated in time, as is usually the case, it can be difficult to evaluate their relative effects on growth. Also, care must be taken in interpreting interannual differences in growth rates within and among geographically separated populations without detailed knowledge of habitat conditions. In addition to temperature and light, changes in salinity can indirectly influence band formation, especially when accompanied by increased siltation from runoff (Dodge and Vaisnys, 1975; Dodge et al., 1977; Dodge and Lang, 1983).
When lowered salinity is associated with detrital runoff, it is
possible to detect the presence of terrestrial, plant-derived fluorescent compounds deposited in the coral skeleton (Isdale, 1984; Boto and Isdale, 1985).
This technique can be used in conjunction with sclerochronology and
isotope analyses to assess past climatic changes. Finally, it should be noted that several important nonclimatic factors can also affect growth patterns, some of which can complicate climatic interpretations. These include expulsion of zooxanthellae following physical damage to the coral (Stoddart, 1962; Woodley et al., 1981), heavy sedimentation (Dodge and Vaisnys, 1975), and grazing by fish and invertebrates that can interrupt a continuous growth record leaving an undefined hiatus. Despite the fact that interactions between various physical parameters are not well understood, and that there exist potential pitfalls in interpreting the skeletal record, we believe that reef-building corals can provide valuable proxy records for low latitude environmental changes. In the following section we combine sclerochronology with stable isotopic analyses to reveal the signature of recent ENS0 events at several sites in the tropical eastern Pacific. 1.3 Effects of physical factors on skeletal chemistry and isotopes There is increasing interest among oceanographers and geochemists in the exploitation of chemical and isotopic records from coral skeletons. Some of the recent studies are discussed briefly. Buddemeier et al. (1981) studied the distribution of alkaline earth (M) elements in corals and concluded that skeletal M/Ca ratios are in near equilibrium with respect to their sea water ratios (thus KD =I), except for Mg which is discriminated against in the skeleton presumably due to its small ionic radius. Smith et al. (1979) demonstrated that Sr/Ca ratios are proportional to sea surface temperature (SST) at the time of accretion in Hawaiian corals. Trace element concentrations in coral skeletons have also been studied (see
238 Shen and Sanford, this volume).
Shen and Boyle (1987) have reported changes in
lattice-bound Cd/Ca ratios in Galapagos corals that correlate with ENS0 occurrences during the past 20 years. As cadmium distributions exhibit nutrient-like behavior in sea water, Cd/Ca in corals provides a recorder of surface nutrient concentrations in the past.
Input of anthropogenic lead to
the surface ocean has also been quantified using Pb/Ca measured in Atlantic and Pacific corals (Shen et al., 1987).
Manganese measured in Galapagos corals has
been shown to be controlled by changes in ocean circulation, e.g., the movement of the Galapagos Front (Linn et al., 1987). Several studies have investigated the distributions of natural and bombproduced isotopes in corals. Natural and bomb radiocarbon have been measured in corals from the Atlantic (Nozaki et al., 1978; Druffel & Linick, 1978; Druffel and Suess, 1983), Pacific (Druffel, 1981; Toggweiler, 1983; Konishi et al., 1981; Druffel, 1987) and Indian Oceans (Toggweiler, 1983) to determine A14C values and to reconstruct time histories of the invasion of bomb radiocarbon into the surface ocean. The bomb fallout radioisotope 90Sr has been measured in corals from several locations in the temperate and tropical oceans (Toggweiler and Trumbore, 1985; Madzsar et al., 1987; Purdy et al., 1987). Stable oxygen isotope ratios (6180) in zooxanthellate corals are significantly shifted toward lighter values from that expected for equilibrium precipitation of aragonite, however, changes in 6180 are directly proportional to changes in SST, salinity and isotopic composition of the water (Weber and Woodhead, 1972; Fairbanks and Dodge, 1979; Dunbar and Wellington, 1981; Weil et al., 1981; Chivas et al., 1983; Swart, 1983; Druffel, 1985). Stable carbon isotope ratios (613C) in zooxanthellate corals are controlled by ambient light, growth rate and sea water 613C composition. Detailed discussion of both oxygen and carbon isotopes in corals appears below. 2 STUDY SITES We discuss stable isotope records from corals located in two major areas of the eastern tropical Pacific, the Galapagos Islands and the coastal Panama region. Corals from four sites in the Galapagos Islands (Urvina Bay [Isabela Island], Academy Bay [Santa Cruz Island], Champion Island and Hood Island) and two sites off Panama (Uraba and Uva Islands) were studied (Fig. la and b). Oxygen isotope records from Urvina Bay and Hood Island were published previously (Druffel, 1985) and are included here for comparison. Information regarding depth, coral species, time span of coral growth studied, and approximate range in seasonal SST for each site are summarized in Table 1. The Galapagos Islands are located in the South Equatorial Current, which is fed by the equatorward flowing Peru Current off the coast of South America.
239 Surface waters are nutrient-rich due to the influence of upwelling along the South American coast and from Cromwell Current upwelling and divergence at the equator. As the southeast trade winds decrease in intensity toward the end of the year, generic El Nifio conditions prevail. This period, which lasts from about December-March,is marked by higher SST (24-25°C versus 2O-2l0C), reduced salinity (34O/oo versus 35O/oo) and low nutrient levels [e.g. nitrate values are often
m compared to an average of 5-7 pm in November (Kogelschatz, et
al., 1985)l. During normal (non-ENSO) years, the SE trade winds increase by April-May and upwelling waters once again reach the surface. Seasonal SST ranges from about 20" to 24°C along the coasts of most of the Galapagos Islands. A typical ENSO event, however, would be marked by continued warm, low salinity waters, and much higher rainfall rates continuing for 4-18 additional months. The catastrophic ENSO of 1982-83was atypical, as the onset of the warming began in mid-1982 and continued through June of 1983 (see Hansen, this volume, for more details). Surface waters along the Pacific coast of Panama are derived from the eastward flowing North Equatorial Counter Current and an arm of the Peru Current that flows northward along the coast of Ecuador and Colombia, setting up a general counterclockwise circulation pattern within the Panama Bight. The hydrography of the area is discussed in Forsbergh (1969). Our Central American study sites are best contrasted by the strong seasonal upwelling that occurs in the Gulf of Panama (Uraba Island), but is generally absent in the Gulf of Chiriqui (Uva Island).
The Inter-TropicalConvergence
Zone (ITCZ) shifts seasonally across the Panamanian Isthmus; at its southernmost excursion during January through April, the NE Trade Wind system is positioned over Panama. During this period, the tropical dry season, the NE trades blow unimpeded across central Panama and induce coastal upwelling.
Sea
surface temperatures in the Gulf of Panama typically drop 4" to 8°C from a wet season mean of 28" to 29°C (Fig. 2).
Lower temperatures (15" to 20'C)
have
been recorded but the duration of such events is typically less than one week (Glynn, 1972; Glynn and Stewart, 1973).
Salinities in the Gulf of Panama vary
from about 3ho/oo during the dry season (January through April) to as low as 28O/oo during the wet season (May through December).
The coasts of western
Panama and southern Costa Rica lie within the warmest and seasonally most stable waters of the eastern tropical Pacific (Wyrtki, 1964).
Strong seasonal
upwelling is absent in the Gulf of Chiriqui because the higher elevations of the western Panamanian highlands intercept the NE Trades (Fig. lb).
January
through April are typically the most cloud-free months of the year in Panama and in the absence of upwelling, sea surface temperatures may reach a maximum of 29" to 30°C in the Gulf of Chiriqui at this time (Fig. 2). seasonally by about 3°C in the Gulf of Chiriqui (Dana, 1975).
SST values vary
240
t
a
oMarchena p Genovesa
-
0"
-
1"s
p
Cristobal
5
SantaMarla
.. (Hood)
Galapagos Islands
830 w
8 20
81"
8 00
79"
780
770
1O"N
9"
8"
24 1
Fig. 1. Sampling locations of corals from (a) the Galapagos Islands and (b) the Gulfs of Panama and Chiriqui. Hatched areas indicate altitudes >1000m. *s represent collection localities. TABLE 1 Locations, periods of growth, and average seasonal sea surface temperature (SST) information for corals used in this study. Location
Depth
Species
Time span
Sampling
SST Range
annual seasonal seasonal annual
19-23°C 21-25'12 20-24.5'C 20-24.5"C
1933-1980
annual
27- 30°C
1973-1983 1961-1983
seasonal seasonal
27-30°C 19-28°C
(m)
GALAPAGOS ISLANDS Urvina Bay1 Academy Bay Champion Island Hood Island1
3
pavonq Flavus P. clavus p . clavus P. clavus
1929-1954 1970-1983 1972-1982 1962-1976
PANAMA COAST Uva -A Uva-1,-56 Uraba
5 5
<8
Gardineroseris planulata P. clavus E gieanteq
6 l80 data reported by Druffel (1985).
3 METHODS Coral colonies were collected using SCUBA, rinsed with fresh water and dried. Slabs were cut along the growth axes and x-rayed to reveal annual variations in the density of the accreted aragonite (Table 1). The annual nature of the bands in the Galapagos and the Uva Island corals were verified using bomb radiocarbon measurements published elsewhere (Druffel, 1981, 1987). Subannual samples were drilled (0.5 or 0.7 nun diameter) from the coral slabs; 7 to 16 samples per year were obtained. Where sample size was limiting, due to
inadequate carbonate deposition (Urvina Bay and Uva-A), analyses were performed on annual bands that were cut with a bandsaw (1 nun width). The Galapagos and Uva-A samples were prepared and the stable isotopes measured according to previously described methods (Druffel, 1985). The Uraba, Uva-1, and Uva-56 samples were prepared and analyzed at Rice University with a V.G. Isogas 602E mass spectrometer following procedures outlined by Dunbar and Wellington (1981) and Minnis (1986). One standard deviation for multiple replicates of samples of Uraba, Uva-1, and Uva-56 measured at Rice University is 0.13°/oo for both 6180 and 613C. Similar results were obtained for multiple replicates of Galapagos samples (Urvina and Hood) at the Woods Hole Oceanographic Institution (WHOI).
242
--Y W
a I)
t-
a
c a
W
2
OF PANAMA
22
2
181 I
I
I
I
I
I
'4
1
I
I
I
I
I
I
I
*FIX+-
Low
High
, . .Cloud Cloud Cover . . . . . . .Cover . . . . . . . ,....................................... .........
........................... ............................
....................................
,
Low Density Band
Non - reproductive
High Density Band
, Reproductive 1
, I
Fig. 2 . Seasonal water temperature variations in the Gulf of Panama (weekly averages, 1978-1979) and the Gulf of Chiriqui (monthly averages, 1979-1980) and average cloud cover (from Wellington and Glynn, 1983). As indicated above, reproductive periodicity of massive corals in this region is coincident with the onset of high density band formation and may be an additional factor regulating linear growth and calcification rates (Wellington and Glynn, 1983). 4
STABLE ISOTOPE RECORDS IN CORALS The data presented here are selected sets from larger compilations for the
243 eastern tropical Pacific region from both the Rice and WHO1 laboratories. The compilations are reported in their entirety elsewhere with a broader emphasis on stable isotopes as potential paleothermometers and paleoproductivity indicators (Druffel, 1985 and in prep.; Minnis, 1986; Dunbar et al., 1987; Wellington, et al., in prep.). 4.1 6 l80 Results A composite of the annual and annually-averaged seasonal 6180 results appears in Figure 3. In general, the Galapagos corals are enriched in l80 relative to the Panama corals. The major reason for the offset is the cooler (by 4-5"C), more saline (by 2-3O/oo) waters that are present in the Galapagos. Both lower SST and higher salinity have the effect of enriching the accreted aragonite in l8O. We would expect the 6180 offset to be approximately equal to the observed value of 1.5O/00, considering the magnitude of the SST and salinity differences and the relationships d6180/dS
- 0.12°/00
per
O/oo
(Dunbar
and Wellington, 1981); not taking into account a change in 6180 of sea water and d6180/dT-0.220/oo per "C (Fairbanks and Dodge, 1979).
However, different
species were examined at the Galapagos (Pavona clavus) than at Panama (Gardineroseris and Pavona eigantea), possibly contributing to a non-linear shift of the 6180 values. McConnaughey (1986) concluded that oxygen isotope thermometry in Galapagos corals is valid only in high growth rate sections of corals (>5 mm/yr), where the displacement from isotopic equilibrium is approximately constant. All of the corals included in this study displayed linear extension rates of >5 mm/yr, thus conforming to the criteria of McConnaughey (1986). One data set (Urvina Bay) from the Galapagos displayed slightly higher 6180 values than the rest. This is from a site of enhanced upwelling on the western coast of Isabela Island, due to the surfacing of the equatorial undercurrent (Cromwell Current).
Surface waters are about 2°C cooler in this region than
they are at the other coral study sites in the Galapagos, hence the 0.4°/00 shift toward heavier 6180 in the Urvina Bay coral (assuming gradients in salinity are small in comparison). Overall, there are lower 6180 values in the Galapagos records during ENSO events than during non-ENS0 periods (for a detailed discussion see Druffel, 1985).
The average annual decrease of 6180 appears more prominent in corals
from Hood, Academy Bay and Champion Islands, presumably due to the absence of the influence of the Cromwell Current. This approach of using annuallyaveraged 6180 is sufficient for detecting ENSO events of significant duration (>9 months) and/or intensity. 6180 appears to be somewhat proportional to the
severity of the pre-1983 ENSO events, that is 6180 is lowest during the major ENSO's of 1940 and 1972-73,noting that the isotope excursions in the Urvina
244
-3
J
'
i
1930
1
1940
- 4
- --- -
1950
1960
ENSO
I
I
1970
1980
YEAR Fig. 3. Annually-averaged 6180 results for corals from Panama [OUraba (L gigantea), Gulf of Panama; W Uva-A (L p l a n u b ) , A Uva-1 (Eclam%) , A Uva56 (Eclams), Gulf of Chiriqui] and the Galapagos Islands [OUrvina,OHood,O Champion and oAcademy (all p . clavus)]. Annual averages for specimens Uraba, Uva-1, Uva-56, Academy and Champion represent the mean isotopic composition of 5 to 16 individual analyses collected between successive February-March markers in the corals (as determined by density banding and isotope maxima/minima). Bay specimen are only slightly larger than the analytical precision.
The
excursion during 1974 does not correspond to an ENSO year, indicating that it is not possible to detect all ENSO's unambiguously and that there are some light isotopic excursions that do not reflect ENSO's. We also need to point out that because density banding is used as the basis for "annual" sampling, some bias may be introduced when banding becomes difficult to interpret or when the time of band formation shifts relative to the temperature signal. The annually averaged Panama data sets show excursions towards lower 6180 values during 1969 and 1982; there is no clear correlation between the isotopic signal and other ENSO periods. Reasons for this are presented below in the description of the seasonal 6180 record. The average isotopic composition of the three Uva Island samples varies by about 0.5°/00. We attribute this range to differences in location on the Uva Island reef and to species specific offsets from equilibrium (e.g., Gardineroseris Dlanulata and Pavona clavus, see Table 1). In two Galapagos corals (Champion Island and Academy Bay), seasonal variations of 6180 ranged from 0.6-0.8°/oo during the ENSO events of 1972-73, 1976 and 1982-83 (Fig. 4a).
The reduction of the seasonal signals demonstrates
that the coral is recording the warm, low salinity waters that persist at the Galapagos during most of an ENSO year.
245
-6
-
-5
--
-4
--
-3
-
a
P O
I ENS0 1
I
I
1970
I
1972
I
I
1974
I
-
I
I
1976
,
1978
,I, 1 - 3
I
I
-6
1982
1980
1984
YEAR
b
-8
-7
6% -6
-5
-7
-6
-5 1
1962
1
1
1964
1
1
1966
1
1
1968
1
1
1970
1
1
1972
I
I
1974
I
I
1976
I
I
1978
I
I I I I I 1980 1982 1984
Year (Feb.-March) Fig. 4 . Seasonal 6180 results for (a) Champion Island ( 0 ,P. clams) and Academy Bay ( 0 ,p . in the Galapagos; (b) Uraba, Gulf of Panama ( 0 , E gieantea), Uva-1 ( 0 , E slams) and Uva-56 ( 0 , E clams) Gulf of Chiriqui, Panama. The time stratigraphy is assigned using both the coral density banding as observed through x-radiography and the seasonal oxygen isotope cycles (after Dunbar et al., 1988). February-March are typically the months of lowest temperatures in the Gulf of Panama and highest temperatures in the Gulf of Chiri ui and at the Galapagos. The year assi ments were therefore made at 180,120 maxima for the Uraba data set and at k80/160minima for the other data sets and confirmed by the position of one major growth band between isotope peaks. The precision on age assignments is estimated at 3 months.
a)
246 However, the relative severity of the ENSO events is not manifest by these data. On the scale of Quinn et al. (1987), the 1982-83 event was longer and more severe (higher SST) than the 1972-73 event, which in turn was longer than the 1976 event. The 6180 data, however, displayed similar seasonal records for the 1976 and 1982-83 events. The intermediate ENSO of 1972-73 was marked by the lowest 6180 values in the Academy Bay coral and values equal to the 1982-83 event in the Champion Island coral. We suspect that the reason the corals did not adequately record the catastrophic 1982-83 event was due to the unprecedented SST of 3O-3l0C,which persisted for at least 3 months in the Galapagos. Bleaching of the corals occurred and accretion of calcium carbonate ceased under these conditions. Much milder conditions prevailed during the 1972-73 and 1976 events, which did not cause bleaching of corals in the Galapagos. Seasonal variation of 6180 ratios was greater in corals from the coast of Panama than in the Galapagos (Fig. 4b), and ranged from an average of 1.g0/oo in the Gulf of Panama (Uraba Island) to l.lO/oo in the Gulf of Chiriqui (Uva Island).
The greater range in the Gulf of Panama reflects the 5" to 8°C
decrease in water temperature during seasonal upwelling as well as the dry season/wet season contrast in salinity. In the Gulf of Panama, Dunbar and Wellington (1981) have estimated that about 30% of the seasonal isotopic variation in branching corals results from salinity changes, and the remainder from temperature. The peaks of isotopic enrichment or depletion are offset by about 4 to 6 months between the Gulf of Chiriqui and Gulf of Panama, as predicted from consideration of the thermal regimes (Fig. 2). As observed in the Galapagos data, the seasonal isotopic range for the Uraba Island specimen is generally reduced during ENSO intervals. This is particularly true for the 1965, 1969, and 1982 events, and observed to a lesser extent in 1972. The relatively minor 1976 ENSO is not evident in either data set and is in fact characterized by some of the greatest enrichments observed in 6180 during the upwelling or cool months of the year. The reduced seasonal isotopic variation results from both lesser l80 enrichment during the upwelling period and lesser l80 depletion during the warmer months of the tropical wet season. The physical process that produces this effect is not yet clear, but it is evident that this type of attenuation of the isotopic signal will obscure the ENSO signal in annually averaged samples. In the Uraba specimen between late 1969 through early 1973, annual growth rates were reduced to about 4 mm/year from a twenty-year average of about 7 mm/year.
This reduction in
growth rate may reflect the deleterious influence of two strong ENSO events on the corals. The reduction in growth rate furthermore reduces the likelihood that the full range of environmental perturbations may be resolved from a coralline isotopic record. The Uva-1 data set reveals a shift to lower
247 180/160ratios during 1982 through early 1983. The cool season maxima and warm
season minima are shifted about 0.5O/oo to lower values relative to the 19731982 averages, corresponding to a seawater warming of about 2°C (Fig. 4b). 4.2 6I3C Results Annually-averaged 6I3C results appear in Figure 5 and seasonal 6I3C results appear in Figure 6. In general, Pavona clavus samples from both the Galapagos gieantea values are about
and Panama have about the same 613C values; the
l0/oo depleted in 6I3C and the Gardineroseris ulanulata samples are about ~ . ~ O / Oenriched O
in 6I3C relative to the P. clavus results.
It appears that
species specific offsets from equilibrium are the dominant control on this large range of values rather than differences in the 6I3C of total dissolved C02 in sea water. There are several other factors controlling the 6I3C value in coralline aragonite. The overall 6I3C signature of the surrounding sea water is considerably lower in waters whose origins are from subsurface water masses, e.g., equatorial surface waters (Kroopnick, 1985).
This could presumably
account for 0.2-0.3°/oo of the offset. Other factors that affect the 613C in skeletal material are ambient light as a function of coral depth in the water column (Weber and Woodhead, 1972; McConnaughey, 1986), cloud cover (Fairbanks and Dodge, 1979), position on the coral colony (McConnaughey, 1986; Wefer and
tI
I
1930
I
- I
1940
I
- - - I
1950
1960
4
1
1970
- ENS0 I
I980
YEAR Fig. 5. Annually-averaged 613C results for corals from Panama [OUraba (E pigantea), Gulf of Panama; mUva-A (G- planulata), rUva-1 (P. clavus), AUva56 (P. clavus ) , Gulf of Chiriqui] and the Galapagos Islands [OUrvina,@Hood,O Champion and OAcademy (all P. clams). Annual averages for specimens Uraba, Uva-1, Uva-56,Academy and Champion represent the mean isotopic composition of 5 to 16 individual analyses collected between successive February-Marchmarkers in the corals (as determined by density banding and isotope maxima/minima).
248 -4
-
-3
--
-2
--
-I
-
a Champion
P
6 "c
-4
Academy Bay
,
1
I
1970
I
I
1972
1
1974
I
I
I
1976
1
1
1978
-3
I
J - I
1
1
I
1982
1980
1984
YEAR
b -4
6'3C -3
-2
-1 I
1962
I
I
1964
1
l
1966
I
I
- -
1968
l
l
1970
I
1
1972
I
I
I
1974
I
1
1976
I
)
1978
I
1
1980
I
I
1982
I
I
1984
Year (Feb.-March)
Fig. 6. Seasonal 613C results for (a) Galapagos Islands: Champion Island ( O E clavus) and Academy Bay ( O E clavua) and (b) Gulf of Panama: Uraba Island ( 0 E gieantea) and Gulf of Chiriqui: Uva Island ( O U v a 56 and Uva 1, both clavus) ,
249
Piitzold, 1985), and shading from adjacent coral colonies. Excess C02 from fossil fuel consumption and biotic sources, admitted to the oceans over the past 100 or so years, has lowered the 613C of the surface sea water by about 0.So/oo in the North Atlantic (Druffel and Benavides, 1986).
This effect may
be present in the annual Uva-A record (Fig. 5 ) , though it is difficult to separate it from other effects that may also have contributed to the range in values noted. The subannual carbon isotope results from Uraba and Uva Islands (Fig. 6b) all show somewhat well-defined seasonal variations with a general tendency towards enrichment in 13C in aragonite accreted during January through April. As water temperatures are at a maximum in the Gulf of Chiriqui and a minimum in the Gulf of Panama during the tropical dry season, we attribute this response primarily to the low cloud cover prevalent over both Gulfs at this time. Small phase offsets between carbon and oxygen isotope minima and maxima are common and variable with the result that l80 - 13C scatterplots often reveal negligible correlation between the two isotope ratios. There is no unique carbon isotope anomaly evident in the seasonal data during ENSO events. 5 CONCLUSIONS 1. We have identified isotope anomalies in a number of coral specimens that show environmental perturbations during the major ENSOs of the past 25 years. 2. Stable oxygen and carbon isotopes do not correctly identify all ENSO events in an unequivocal fashion. 3.
Galapagos corals record higher average water temperature/lower average
salinity during ENSO years. 4.
Seasonal 6180 in Panama corals reveals most ENSO anomalies, but the signal
is complex and may not be readily apparent in annually averaged samples. 5.
An ENSO signature common to both Galapagos and Panama study sites is a
reduction in the seasonal range of 6180 values. 6.
Major ENSOs (as well as any major stress event) may result in cessation of
"normal" growth, thereby reducing the efficacy of the coral as an environmental recorder. 6 ACKNOWLEDGEMENTS We thank Mike Dehn, Sheila Griffin, C. Eben Franks and Danuta Kaminski for assistance with the isotope analyses. We thank Peter Swart and an anonymous reviewer for comments on the manuscript. We are grateful to Molly Lumping and Wanda Blackman for typing the manuscript and to Amy Witter for drafting the figures. This work was supported by a grant from the Petroleum Research Fund of the American Chemical Society (RBD), American Philosophical Society (GMW, RBD), and by NSF through Grant Nos. OCE-8315260 and OCE-8608263 (ERMD), and
2 50
OCE-8415615 (to P.W. Glynn).
7 REFERENCES Baker, P.A. and Weber, T.N., 1975. Coral growth rate: variation with depth. Earth Planet. Sci. Lett., 27: 57-66. Boto, K. and Isdale, P., 1985. Fluorescent bands in massive corals result from terrestrial fulvic acid inputs to nearshore zone. Nature, 315: 396-397. Buddemeier, R.W. and Kinzie 111, R.A., 1976. Coral Growth. Oceanogr. Mar. Biol. Annu. Rev., 14: 183-225. Buddemeier, R.W., Maragos, J.E. and Knutson, D.W., 1974. Radiographic studies of reef coral exoskeletons: rates and patterns of coral growth. J. Exp. Mar. Biol. Ecol., 14: 179-200. Buddemeier, R.W., Schneider, R.C. and Smith, S.V., 1981. The alkaline earth chemistry of corals. Proc. 4th Int. Coral Reef. Symp., Manila, 2: 81-85. Chalker, B.E. and Taylor, D.L., 1975. Light-enhanced calcification of the coral, AcroDora cervicornis. Proc. Roy. SOC. London (B)., 201: 179-189. Chivas, A.R., Aharon, P., Chappell, J., Viastuin, C. and Kiss, E., 1983. Trace elements and stable-isotoperatios of annual growth bands as environmental indicators. In: J.T. Baker, R.M. Carter, P.W. Sammarco and K.P. Stark (Editors), Proceedings of the Inaugural Great Barrier Reef Conference, Townsville, JCU Press, pp. 83-88. Clausen, C.D., 1971. Effects of temperature on the rate of 45Ca uptake by PocilloDora damicornis. In: H.M. Lenhoff, L. Muscatine and L.V. Davis (Editors), Experimental Coelenterate Biology. University o f Hawaii Press, Honolulu, pp. 246-259. Clausen, C.D. and Roth, A.A., 1975. Effects of temperature and temperature adaptation on calcification rate in the hermatypic coral, Pocillopora damicornis. Mar. Biol., 33: 93-100. Coles, S.L. and Jokiel, P.L., 1977. Effects of temperature on photosynthesis and respiration in hermatypic corals. Mar. Biol., 43: 209-216. Dana, T.F., 1975. Development of contemporary eastern Pacific coral reefs. Mar. Biol., 33: 355-374. Dodge, R.E., 1982. Effects of drilling muds on the reef-building coral Montastrea annularis. Mar. Biol., 71: 141-147. Dodge, R.E. and Vaisnys, J.R., 1975. Hermatypic coral growth banding as environmental recorder. Nature, 258: 706-708. Dodge, R.E. and Lang, J.C., 1983. Environmental correlates of hermatypic coral (Montastrea annularis) growth on the East Flower Gardens Bank, northwest Gulf of Mexico. Limnol. Oceanogr., 28: 228-240. Dodge, R.E. and Brass, G.W., 1984. Skeletal extension, density and calcification of the reef coral, Montastrea annularis: St. Croix, U.S. Virgin Islands. Bull. Mar. Sci., 34: 288-307. Dodge, R.E. and Gilbert, T.R., 1984. Chronology of lead contained in banded coral skeletons. Mar. Biol., 82: 9-13. Dodge, R.E., Aller, R.G. and Thompson, J., 1974. Coral growth related to resuspension of bottom sediments. Nature, 247: 574-577. Dodge, R.E., Turekian, K.K. and Vaisnys, J.R., 1977. Climatic implication of Barbados coral growth. Proc. 3rd Int. Coral Reef Symp., Miami, 2: 361-366. Dodge, R.E., Wyers, S.C, Firth, H.R., Knapp, A.H., Smith, S.R. and Sleeter, T.D., 1984. The effect of oil and oil dispersants on the skeletal growth of the hermatypic coral DiDloria strigosa. Coral Reefs, 3: 191-198. Druffel, E.M., 1981. Radiocarbon in annual coral rings from the eastern tropical Pacific. Geophys. Res. Lett., 8: 59-62. Druffel, E.R.M., 1985. Detection of El Nifio and decade time scale variations of sea surface temperature from banded coral records: implications for the carbon dioxide cycle. In: The Carbon Cycle and Atmospheric C02: Natural Variations Archean to Present. Geophysical Monograph 32, AGU. Druffel, E.R.M., 1987. Bomb radiocarbon in the Pacific: annual and seasonal timescale variations. J . Mar. Res., 45: 667-698.
25 1
Druffel, E.M. and Linick, T.W., 1978. Radiocarbon in annual coral rings of Florida. Geophys. Res. Lett., 5: 913-916. Druffel, E.M. and Suess, H.E., 1983. On the radiocarbon record in banded corals: exchange parameters and net transport of 14C02 between atmospheric and surface ocean. J. Geophys. Res., 88: 1271-1280. Druffel, E.R.M. and Benavides, L.M., 1986. Input of excess C02 to the surface ocean based on 13C/12C ratios in a banded Jamaican sclerosponge. Nature, 321: 58-61. Dunbar, R.B. and Wellington, G.M., 1981. Stable isotopes in a branching coral monitor seasonal temperature variation. Nature, 293: 453-455. Dunbar, R.B., Wellington, G.M., Colgan, M. and Glynn, P.W., 1987. Galapagos massive corals: an annual record of eastern tropical Pacific climate during the past 400 years. EOS, 68: 1743. Emiliani, C,, Hudson, J.H., Shinn, E. and George, R.Y., 1978. Oxygen and carbon isotopic growth record in a reef from the Florida Keys and a deep-sea coral from Blake Plateau. Science, 202: 627-627. Fairbanks R.G. and Dodge, R.E., 1979. Annual periodicity of the l80/l6O and 13C/1fC ratios in the coral piontastrea annularis. Geochim. Cosmochim. Acta, 43: 1009-1020. Forsbergh, E.D., 1969. On the climatology, oceanography and fisheries of the Panama Bight. Bull. Inter-Am.Trop. Tuna Comm., 14: 49-385. Franzisket, L., 1970. The atrophy of hermatypic reef corals maintained in darkness and their subsequent regeneration in light. Int. Rev. Gesamten. Hydrobiol., 55: 1-12. Glynn, P.W., 1972. Observations on the ecology of the Caribbean and Pacific coasts of Panama. In: M.L. Jones (Editor), The Panamic Biota: Some Observations Prior to a Sea-level Canal. Bull. Biol. SOC. Wash., 2: 13-30. Glynn, P.W., 1977. Coral growth in upwelling and nonupwelling areas off the Pacific coast of Panama. Environ. Conserv., 11: 133-146. Glynn, P.W., 1983. Extensive bleaching and death of reef corals on the Pacific coast of Panama. Environ. Conserv., 11: 133-146. Glynn, P.W., 1984. Widespread coral mortality and the 1982/83 El Nido warming event. Environ. Conserv., 11: 133-146. Glynn, P.W. and Stewart, R.H., 1973. Distribution of coral reefs in the Pearl Islands (Gulf of Panama) in relation to thermal conditions. Limnol. Oceanogr., 18: 367-379. Glynn, P.W. and Wellington, G.M., 1983. Corals and Coral Reefs of the Galapagos Islands. University of California Press, Berkeley, 330 pp. Glynn, P.W., Druffel, E.M. and Dunbar, R.B., 1983. A dead Central American coral reef tract: possible link with the Little Ice Age. J. Mar. Res., 41: 605-637. Goreau, N.I. and Hayes, R.L., 1977. Nucleation catalysis in coral skeletogenesis. Proc. 3rd Int. Coral Reef Symp., Miami, 2: 439-455. Goreau, T.F., 1961. Problems of growth and calcium deposition in reef corals. Endeavor, 20: 20-39. Goreau, T.F. and Goreau, N.I., 1959. The physiology of skeleton formation in corals. 111. Calcium deposition by hermatypic corals under various conditions in the reef. Biol. Bull., 117: 239-250. Highsmith, R.C., 1979. Coral growth rates and environmental control of density banding. J. Exp. Mar. Biol. Ecol., 37: 105-125. Hubbard, D.K. and Scaturo, D., 1985. Growth rates of seven species of scleractinian corals from Cane Bay and Salt River, St. Croix, USVI. Bull. Mar. Sci., 36: 325-338. Hudson, J.H., 1981. Growth rate in Montastrea annularis: a record of environmental change in Key Largo coral reef marine sanctuary, Florida. Bull. Mar. Sci., 31: 444-459. Hudson, J.H., Shinn, E.A., Halley, R.B. and Lidz, B., 1976. Sclerochronology: a tool for interpreting past environments. Geology, 4: 361-364. Huston, M., 1985. Changes in coral growth rates with depth at Discovery Bay, Jamaica. Coral Reefs, 4: 19-25.
252 Isdale, P., 1984. Fluorescent bands in massive corals record centuries of coastal rainfall. Nature, 310: 578-579. Jaap, W.C., 1979. Observation on zooxanthellae expulsion at Middle Sambo Reef, Florida Keys. Bull. Mar. Sci., 29: 414-422. Jokiel, P.L. and Coles, S.L., 1977. Effects of temperature on the mortality and growth of Hawaiian reef corals. Mar. Biol., 43: 201-208. Kawaguti, S . and Sakamoto, D., 1948. The effect of light on the calcium deposition of corals. Bull. Oceanogr. Inst. Taiwan.,4: 65-70. Knutson, D.W., Buddemeier, R.W. and Smith, S.V., 1972. Coral chronometers: seasonal growth bands in reef coral. Science, 177: 270-272. Kogelschatz, J . , Solorzano, L., Barber, R., Mendoza, P., 1985. Oceanographic conditions in the Galapagos Islands during the 1982/1983 El Niiio. In: G. Robinson and E.M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983 Event. Publication of the Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 91-123. Konishi, K., Tanaka, T. and Sakanoue, M., 1981. Secular variation of radiocarbon concentration in seawater: sclerochronological approach. Proc. 4th Int. Coral Reef Symp., Manila, 1: 181-185. Kroopnick, P.M., 1985. The distribution of 13C of sigma-C2 in the world oceans. Deep-sea Res., 32: 57-84. Linn, L.J., Druffel, E.R.M. and Delaney, M.L., 1987. Trace metal concentrations in a Galapagos coral: an ENS0 indicator. EOS, 68: 1743. Madzsar E.M Benninger, L.K. and Freeman, J . H . , 1987. Combined fallout 90Sr and '239*2d0Puin the annual bands of Montastrea annularis, Broward County, Florida. EOS, 68: 1743. McConnaughey, T.A., 1986. Oxygen and carbon isotope disequilibria in Galapagos corals: isotopic thermometry and calcification physiology. Ph.D. Dissertation, University of Washington, 340 pp. Minnis, S.A., 1986. Stable isotope profiles of hermatypic corals: indicators of changing environmental conditions in upwelling and non-upwelling regions o f the Eastern Tropical Pacific. (unpubl.) M.S. Thesis, Rice Unviersity, Houston, Texas, 122 pp. Muscatine, L. and Cernichiari, E., 1969. Assimilation of photosynthetic products of zooxanthellae by a reef coral. Biol. Bull., 137: 506-523 Nozaki, Y., Rye, D.M., Turekian, K.K., and Dodge, R.E., 1978. 13C and I4C variations in a Bermuda coral. Geophys. Res. Lett., 5: 825-828. PBtzold, J . , 1984. Growth rhythms recorded in stable isotopes and density bands in the reef coral Porites lobata (Cebu, Philippines). Coral Reefs, 3: 8790. Pearse, V.B., and Muscatine, L., 1971. Role of symbiotic algae (zooxanthellae) in coral calcification. Biol. Bull., 141: 350-363. Porter, J.W., Battey, J.F. and Smith, G . J . , 1982. Perturbations and change in coral reef communities. Proc. Nat. Acad. Sci., 79: 1678-1681. Purdy, C., Druffel, E.R.M. and Livingston, H., 1987. Strontium-90 activities in banded corals from the North Atlantic. EOS, 68: 1743. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Nitio occurrences over the past four and a half centuries, J . Geophys. Res., 92: 14,449-14, 461.
Roberts, L., 1987. Coral bleaching threatens Atlantic reefs. Science, 238: 1228-1229. Roberts, L., 1988. Corals remain baffling. Science, 239: 256. Rogers, C.S., 1979. The effect of shading on coral reef structure and function. J . Exp. Mar. Biol. Ecol., 41: 269-288. Schneider, R.C. and Smith, S.V., 1982. Skeletal Sr content and density in Porites spp. in relation to environmental factors. Mar. Biol., 66: 121-131. Shen, G.T. and Boyle, E.A., 1987. Lead in corals: reconstruction of historic industrial fluxes to the surface ocean. Earth Planet. Sci. Lett., 82: 289304. Shen, G.T., Boyle, E.A. and Lea, D.W., 1987. Cadmium in corals as a tracer of historical upwelling and industrial fallout. Nature, 328: 794-796.
253 Smith, D.C., Muscatine, L. and Lewis, D.H., 1969. Carbohydrate movement from autotrophs to heterotrophs in parasitic and mutualistic symbiosis. Biol. Rev., 44: 17-90. Smith, S.V., Buddemeier, R.W., Radalje, R.C. and Houck, J.E., 1979. Strontiumcalcium thermometry in coral skeletons. Science, 204: 404-407. Stoddart, D.R., 1962. Catastrophic storm effects on the British reefs and cays. Nature, 196: 512-515. Stoddart, D.R., 1969. Ecology and geology of recent coral reefs. Biol. Rev., 44: 433-498. Swart, P.K., 1983. Carbon and oxygen isotope fractionation in scleractinian corals: a review. Earth-Sci. Rev., 19: 51-80. Toggweiler, J . R . , 1983. A six zone regionalized model for bomb-radiotracers and Cog in the upper kilometer of the Pacific Ocean. Ph.D. Dissertation. Columbia University. Toggweiler, J . R . and Trumbore, S . , 1985. Bomb-test Sr-90 in Pacific and Indian Ocean surface water as recorded by banded corals. Earth Planet. Sci. Lett., 74: 306-314. Weber, J.N. and Woodhead, P.M.J., 1972. Temperature dependence of oxygen-18 concentration in reef coral carbonates. J . Geophys. Res., 77: 463-473. Weber, J.N., White, E.W. and Weber, P.H., 1975. Correlation of density banding in reef coral skeletons with environmental parameters: the basis for interpretation of chronological records preserved in the coralla of corals. Paleobiology, 1: 137-149. Wefer, G. and Patzold, J . , 1985. 13C/12C record of atmospheric C02 increase in a coral head from the Philippines (Cebu). Scripps Institution of Oceanography Reference 85-31: 133-135. Weil, S.H., Buddemeier, R.W., Smith, S.V. and Kroopnick, P.M., 1981. The stable isotopic composition of coral skeletons: control by environmental variables. Geochim. Cosmochim. Acta, 45: 1147-1153. Wellington, G.M., 1982. An experimental analysis of the effects of light and zooplankton on coral zonation. Oecologia, 52: 311-320. Wellington, G.M. and Glynn, P.W., 1983. Environmental influences on skeletal banding in eastern Pacific (Panama) corals. Coral Reefs, 1: 215-222. Wells, J.W.,1957. Coral reefs. Mem. Geol. SOC. A m . , 67: 609-631. Wethey, D.S. and Porter, J.W., 1976. Sun and shade differences in productivity of reef corals. Nature, 262: 281-282. Woodley, J.D. and 17 co-authors,1981. Hurricane Allen's impact on Jamaican coral reefs. Science, 214: 749-755. Wyrtki, K., 1964. The thermal structure of the eastern Pacific Ocean. Dtsch. Hydrogr. Z . , 8: 1-84. Yonge, C.M. and Nicholls, A.G., 1931. Studies on the physiology of corals. VI. The structure, distribution and physiology of the zooxanthellae. Sci. Rep. Great Barrier Reef Exped., 1: 177-211.
This Page Intentionally Left Blank
255
TRACE ELEMENT INDICATORS OF CLIMATE VARIABILITY IN REEF-BUILDING
cow
G. T. SHEN and C. L. SANFORD Lamont-Doherty Geological Observatory of Columbia University, Palisades, New York 10964 (U.S.A.)
ABSTRACT Shen, G.T. and Sanford, C.L., 1989. Trace element indicators of climate variability in reef-building corals.
An active couple exists between the atmosphere and surface ocean to produce regional-to-global changes in earth's climate. While numerous terrestrial recording systems have revealed meteorological perturbations over the continents, however, historical changes in the surface ocean are largely unknown. This contribution describes recent efforts to develop new paleochemical indicators trapped in the aragonite lattice of reef-building corals. The methodology requires chance combinations of circumstances (crystal lattice compatibility and mechanisms that promote variable surface ocean composition) and great analytical care, yet several applications have surfaced in relation to El Niiio. Cadmium and manganese comprise paleo-upwelling indicators that display great sensitivity in the eastern tropical Pacific Ocean. Barium levels in Caribbean corals reflect seasonal river discharge from South America that may in turn be modulated by the El Niiio Southern Oscillation. Manganese and barium may also track lateral movement of water masses as in continental shelf water advection or river plume migration. Finally, coralline cadmium is seen to respond to ENSO-driven rainfall anomalies in the western equatorial Pacific, though the mechanism behind this response is not clear.
1 INTRODUCTION Clues as to previous climates in Earth's history have come in various forms from the land and sea. Among the recording substrates that exist in the oceans, calcareous plankton have proven to be most useful. The shells of fossil foraminifera in deep-sea sediments have, for example, been used to reconstruct changes in glacial-interglacial ice volumes and ocean temperature and circulation. Periodicities unmasked by such studies have implicated orbital forcing as an important trigger in resetting climate on time scales of tens of thousands of years. In the nearer term past, skeletons of reef-building corals offer the time resolution necessary for detailed reconstructions of climate effects on the surface ocean. Druffel and co-
workers describe aspects of coral growth and use of stable oxygen and carbon
256
isotopes as El Niiio indicators in a separate contribution in this volume. This study introduces the potential of several skeletally-bound trace elements as complementary chemical indicators of El Niiio-related perturbations to the surface ocean. Elements of particular interest include cadmium, manganese, and barium. Depending on the locale, these and possibly other trace seawater constituents may be used to reconstruct historical changes in such diverse phenomena as winds, upwelling, rainfall, river discharge, and advection of surface ocean waters. 2 MINOR AND TRACE ELEMENT GEOCHEMISTRY OF CORALS Chemical studies of fossil marine carbonates to reconstruct paleo-environments began with the alkaline earth metals Mg, Sr, and Ba (Odum, 1951; Chave, 1954; Turekian, 1955; Thompson and Chow, 1955; and many others). The rationale behind investigation of these elements was their chemical similarity to Ca, and expected uptake in CaC03 minerals. Eventually, wider coverage of elements in aragonitic corals included the alkali metals (Amiel et al., 1973; Swart, 1981), various transition metals (Harriss and Almy, 1964; Veeh and Turekian, 1968; Livingston and Thompson, 1971; St. John, 1974), uranium (Broecker, 1963; Gvirtzman et al., 1973; Flor and Moore, 1977; Cross and Cross, 1983), and the rare earth elements (Scherer and Seitz, 1980; Shaw and Wasserburg, 1985). Results of many of these pioneering studies, however, must be interpreted carefully insofar as inferred host phases and incorporation mechanisms. Particularly in the case of the transition and rare earth metals where lattice abundances in corals are in the ppmppb range, tissue or contaminant phase inventories can rival or exceed structurallybound metal levels. A great variety of coralline elements have been identified where because of size and/or charge imbalances, incorporation mechanisms are very uncertain (e.g. Na+, K+, Nd+3). Chemical complexation with seawater anions also raises questions as to deposition mechanisms, especially in the case of metals that exist as anionic complexes (e.g. U, V - see Swart and Hubbard, 1982; Shen and Boyle, 1987). In considering skeletal components as paleoenvironmental indicators, it is imperative that the constituents reflect conditions at the time of carbonate deposition and that they be reproducibly analyzed. This largely rules out organically-bound markers because the organic content of corals is expected to vary with species, location, and age. Studies of metal adsorption (e.g. Cd2+, Mn2+, Zn2+, Co2+) on inorganic calcium carbonate have established that surface coatings may be important, particularly when chemisorption and precipitation co-occur as endpoints of a continuum process (Kitano et al., 1976; McBride, 1979; Morse, 1986; Comans and Middelburg, 1987; Davis et al., 1987, and many others). Still other studies suggest that substitution of metal ions may occur heterogeneously via trace secondary phases (e.g. calcite) (Amiel et al., 1973; Houck et al., 1975; Angus et al., 1979; Swart and Hubbard, 1982). Fortunately, the extent of diffusion, precipitation,
257
and/or recrystallization in biogenic marine carbonates often appears limited, as attested to by paleochemical studies of foraminifera and corals (Druffel and Linick, 1978; Dodge and Gilbert, 1984; Boyle, 1986; Shen and Boyle, 1987, 1988) and successful radiometric dating applications (21oPb, 228Ra, 14C, U-Th), which rely upon closed system behavior (Dodge and Thomson, 1974; Chappell and Polach, 1972; Edwards et al., 1987). Even were surface chemisorbed/precipitated metals to comprise persistent phases (Amiel et al., 1973), though, they would prove difficult to isolate in the presence of other heterogeneous surface phases, and thus are not likely to offer much hope as paleochemical indicators. Conversely, latticesubstituted components cannot be added or removed unless mineral dissolution or alteration occurs, so they have potential as permanent markers. Demonstration of true lattice substitution by elements at ppb levels is not possible by conventional analytical means. Instead one must appeal to physicochemical constraints, observed structures of related minerals, and empirical measures of trace element abundances and partitioning. For example, the ionic radius of Cd2+ is nearly identical to that of Ca2+, as is cadmium's outer electron configuration (4d105d2 versus 4 ~ 2 ) . Furthermore, rhombohedral CdC03 is a known mineral phase (otavite) that exhibits solid solution behavior with calcite (Chang and Brice, 1971). Paleochemical studies of benthic foraminifera by Boyle (1988) have revealed a consistent partitioning of Cd between ocean bottom waters and shells of 4 different species ( K ~ = 2 . 9 ) . Taken together, this information suggests that it is reasonable to assume Boyle's measured Cd levels to be lattice-bound. Furthermore, it is reasonable to envision that parallel Cd substitution in aragonite is possible, though little is actually known of such solution behavior. Structural uptake is argued by Shen et al. (1987) and Shen and Boyle (1988) in their determination of baseline Cd levels in corals from the Florida Keys, Bermuda, and Galapagos Islands. Reported distribution coefficients for divalent cations in aragonitic corals, however, vary widely in the literature and have thus caused controversy as to inferred controls over precipitation. Howard and Brown (1984) interpret this observed variability as evidence of extraskeletal uptake and conclude that structural inclusion by corals does not occur for most metals. Actually, some of the historical discrepancies can be trimmed by correcting earlier estimates of KD using more recently determined concentrations of metals in seawater. Specifically, reliable measurements of Mn, Pb, Cd, Fe, Cu, Zn, Co, and rare earth elements have largely been attained only within the last decade. Elements whose concentrations in seawater are sensitive to contamination (all of the above) may also be prone to contamination as solid aragonitic components because of their trace levels. Retrospective corrections of this nature are not possible. In spite of these complications, observations have indicated that the larger alkaline earth cations (Sr2+, Ba2+, Ra2+) are indiscriminately precipitated from seawater by corals (e.g. Veeh and Turekian, 1968; Buddemeier et al., 1981). In comparing the metal
258
content of skeletons and tissues of both hermatypic and ahermatypic corals, Buddemeier and co-workers (198 1) further concluded that the calcification process is insensitive to bulk tissue concentrations of metals. Extension of this precipitation behavior to several transition metals and possibly rare earth elements has recently been proposed by Shen and Boyle (1988). Such a uniform mineralization process is very convenient for reconstructing seawater paleoenvironments, but what is the physicochemical basis for such distributions? Let us consider thermodynamic controls on uptake of a few of the closest aragonite-compatible elements to see whether C a C 0 3 precipitation by corals is an equilibrium process. If we express metal substitution as:
the resulting equilibrium constant follows
where y s are solid and solution phase activity coefficients, fs represent the fraction of dissolved metal available in uncomplexed divalent form, and Kapp is an apparent equilibrium constant. We can re-express K as the ratio of solubility products for C a C 0 3 (aragonite) and MCO3 which yields the following expression for the apparent constant (or apparent distribution coefficient):
259
Equation (3) can be used to predict the relative equilibrium distributions of various metals in aragonite. Table 1 lists the computed magnitudes of Kapp f o r eight metals that bracket Ca2+ in ionic radius, given our best estimates of yand f. The predicted thermodynamic Ks span 3 orders of magnitude, primarily as a result of large differences in solubilities of the metal carbonates (e.g. PbC03 is sparingly soluble in comparison to SrC03). In contrast, observed distribution coefficients in corals lie conspicuously close to unity. There exist large uncertainties in the activity and species complexing coefficients used in these simple calculations (particularly for Y ~ c o -3 see Stumm and Morgan, 1981, pp. 287-291), however, the results suggest that kinetic control via rapid coprecipitation may supersede thermodynamic equilibrium.
TABLE 1 Predicted thermodynamic and observed distribution coefficients for trace elements in aragonitic corals
Sr2+
1.42 1.29 1.26
ca2+
1.12
Ba2+ Pb2+
1.10 0.96 0.92 0.90 0.90
1.20 8.1 104 6.5
0.86 0.03 0.71
1.3 3 103 5.6
0.81 1.2 io3rd] 12 2.9 x lo2 58 60
0.03 0.58 0.69 0.58 0.46
46 8.6 2.5 x lo2 4.1 34
=1 0.1-1 1-30 =1 <10
[h] [i] ti] [k] [I]
* radii are relative to a coordination number of VIII (aragonitic Ca2+ CN = IX, h o w e v e r estimates are not available for some elements) # assumes Yca2+ = yM2+ = 0.22 and ' f c a c 0 3 = Y M C O ~ = 1 [see equation (3)] [a] Shannon, 1976 [b] Smith and Martell, 1976 [c] Turner et al., 1981 [d] Davis et al., 1987 [el Ohde et al., 1978: Buddemeier et al., 1981; Shen and Boyle. 1988 [fl Shen and Boyle, 1987 [g] Smith et al., 1979; Cross and Cross, 1983 [h] Shen et al., 1987 [i and j ] this paper: Shen and Boyle, 1988: Shen, unpublished results (uncertainties due to lack of dissolved phase measurements in immediate vicinity of reefs. Estimates are based on Klinkhammer and Bender, 1980 and Landing and Bruland, 1987.) [k] Veeh and Turekian, 1968 [I] Shen and Boyle, 1988
260
3 SAMPLE SITES A variety of reef environments have been sampled to test for oceanic chemical responses to climatological phenomena (Fig. 1). In the eastern tropical Pacific Ocean, the influence of upwelling is investigated at two sites within the Galapagos Islands (Punta Pitt, San Cristobal Island and Urvina Bay, Isabela Island) and one location in the Gulf of Panama (Contadora Island). Upwelling at the Galapagos Islands occurs via vertical mixing by the Equatorial Undercurrent and equatorial divergence set up by southeasterly trade winds that have their strongest easterly component from June - November (Pak and Zaneveld, 1974; Donguy and Henin, 1980). In contrast, the Gulf of Panama experiences coastal upwelling that is induced by the meridional component of the northeast trade wind system during January - April (Wellington and Glynn, 1983). Although controls on upwelling at these locations are fundamentally different, both regions are influenced by El Niiio. Farther west, preliminary measurements of a coral from Tarawa (Republic of Kiribati - lo24'N, 17300'E) have been performed to investigate possible influences of local upwelling or precipitation. Tarawa lies in the core of the Pacific rainfall anomaly region, which is subject to interannual El Niiio-related variability of order 300-400% (Ropelewski and Halpert, 1987). Finally, a terrestrial connection to the El Niilo-Southern Oscillation is sought in Caribbean corals from Barbados and Tobago. Both of these locations are influenced by seasonal Amazon and Orinoco River discharge (Steven and Brooks, 1972; Froelich et al., 1978; Borstad, 1982), which may in turn be modulated by altered precipitation patterns wrought by ENSO.
Figure 1 Locations of coral sampling sites, (A) = Barbados; (B) = Tobago; (C) = Gulf of Panama; (D) = Galapagos Islands; (E) = Tarawa
26 1
4 METHODS Corals were either drilled or collected intact by a number of investigators (see Table 2). All specimens were dated by annual band counting of x-radiographs. Low and high-density aragonite accretion periods have been judged from published studies (Table 2) and subannual time assignments estimated accordingly. Sample preparation and analysis for trace elements were performed according to previously described methods (Shen and Boyle, 1987). Analysis by furnace atomic absorption spectrophotometry, however, has since been modified to extend graphite tube life and improve measurement precision. Cadmium and manganese are both atomized from graphite platforms (normal ungrooved graphite tubes were used) at 1,600OC and 2,4OOOC, respectively. (Char temperatures have also been adjusted to 720OC and 1,37OOC, respectively.) Replicability of Cd and Mn analysis by GFAAS is +1-2%, however, overall measurement reproducibility ( l a = 25%)is affected by pH sensitive recovery of transition metals by dithiocarbamate chelate (APDC) coprecipitation. Larger errors for replicate Ba analysis (in the range +5-10%) result from this element's refractory nature, which necessitates very high furnace temperatures to achieve atomization.
5 OCEANIC MARKERS OF EL N S O Within the oceans, climate effects as shortlived as ENS0 can only be detected in the upper water column. Here, chemical distributions can be altered by a variety of causes; for example: rainfall, winds (upwelling/aeolian transport), continental runoff, etc. Though sedimentary records of such changes typically only resolve shifts on millenium time scales, a number of investigators have exploited varved sediments in nearshore settings to enhance temporal resolution. In the context of El Niiio, Baumgartner et al. (1985) and Lange et al. (1987a, b), for example, have identified numerous microfossil assemblage and geochemical changes in cores from the Gulf of California and the Santa Barbara Basin. The unique aspect of corals as paleoenvironmental recorders is that surface ocean signals are preserved without having to transit the water column and sediments. In this study, we specifically seek subtle, temporal changes in the chemical make-up of the surface ocean. With the exception of a few of the platinum group metals (Hodge et al., 1985), all of the naturally occurring elements of the periodic table have been mapped to varying degrees in the ocean. Controls over certain types of elemental distributions have been inferred; these suggest several possible paleochemical applications. 5.1 Uawelline: in the eastern eauatorial Pacific Cadmium is perhaps the archtypal nutrient analogue among all the elements. Boyle and co-workers (1976) and Martin et al., (1976) first brought attention to the
TABLE 2 Coral collection data Location
Genus/species
Water depth (m)
Sample age
Dense Band formation*
Collector
Galapagos Islands Punta Pitt, San Cristobal Island
Pavom clavus
15
1965-1979
Feb-April [a]
Mc Connaughey
Urvina Bay, Isabela Island
Pavona clavus
3
1969-1978
Feb-July [a]
Colgan/Glynn
Urvina Bay, Isabela Island
Pavona clavus
=4-5
1583-1954
Feb-July [a]
Colgan/Glynn
Pavona gigantea
-5-6
1971-1980
July-Dec [b]
Dunbar/Wellington
Hydnophora microconos
6
1959-1978
irregular banding
Fairbanks
Barbados Bellairs Research Institute
Montastrea
3
1965-1977
Aug-Oct [c]
FairbankdDodge
Tobago Bucco Reef
Diploria strigosa
4
1975-1983
Sept-Nov [d]
Brown
G ~ l fof Panama Contadora Island
Tarawa
*
annularis
assuming linear growth rate
[a] G l y ~and Wellington, 1983 [b] Wellington and Glynn, 1983 [c] Fairbanks and Dodge, 1979; Weber et al., 1975 [d] assumed similar to Barbados
263
similarity in distributions of Cd and the micronutrients PO4 and NO3. Since then, this relationship has been confirmed globally (Boyle et al., 1981; Bruland and Franks, 1983; Danielsson and Westerlund, 1983; and others). Apparently, similar recycling efficiencies give rise to strikingly parallel Cd and nutrient distributions as seen in Fig. 2. In the absence of a demonstrated biochemical pathway, however, the mechanism of Cd uptake by phytoplankton remains unknown. The critical aspect of the oceanic Cd distribution with regard to El NiRo is the steep gradient in the upper several hundred meters. In addition to anomalous warming of surface waters in the eastern equatorial Pacific, a hallmark of El NiRo is a decline in surface nutrient levels, which spurs dramatic shifts in regional ecology. These temperature and nutrient responses are brought about by a deepening of the thermocline in the eastern basin to the extent that wind-driven mixing cannot penetrate the newly emplaced warm surface layer (Wyrtki, 1975; Rasmusson and
PI IOSPHATE (pM) or MANGANESE (nM)
CADMIUM (pM)
(a 1
1'1 IOSl'l IhTE (pM)
(b)
Figure 2 (a) Similarity of phosphate and cadmium biogeochemical cycles as exemplified by a station in the eastern North Pacific (33'N; 145OW .Bruland, 1980). Also shown is the distribution of total dissolvable Mn in the eastern tropical Pacific (6'33"; 92'48'W - Klinkhammer and Bender, 1980). Shelf advected Mn in the upper 200 m causes a surface distribution opposite to that of Cd and PO4 in this area. (b) Cadmium-phosphate correlation based on a global data set (adapted from Boyle, 1988).
264
Carpenter, 1982). Unfortunately, nutrients are not expected to leave a permanent record in corals. Dodge et al. (1984) have detected phosphorus in untreated skeletons of two Atlantic/Caribbean coral species; however, the incorporation mechanism is unknown. At least part of the skeletal phosphorus measured by these investigators was not liberated by HC1 dissolution and must therefore have been organically bound. Evidence of sewage outfall effects described by Dodge and co-workers, however, suggest that phosphorus uptake by corals should perhaps be studied more closely. Alternatively, if a strong Cd-nutrient relationship persists in upwelling areas, a proxy-nutrient record is possible, provided corals bind Cd in fixed proportion to ambient dissolved levels. The success of the Cd paleofertility tracer is evident in a 15-year seasonallyreconstructed Cd/Ca time series shown in Fig. 3a from the Galapagos Islands. Increases in Cd/Ca mole ratios in Pavona clavus from San Cristobal Island generally accompany seasonal decreases in water temperature (due to enhanced upwelling) as recorded in Academy Bay, Santa Cruz Island, 65 km to the west (note temperature axis is inverted). Conversely, skeletal Cd levels fall during warm periods when upwelling is less vigorous (February-July). A small amount of stretching and compressing of the Cd/Ca time scale (+ several months) seems warranted in a few instances (e.g. 1966-67; 1971-72) and is defensible in view of sectioning difficulties. Quarterly anomalies in temperature and Cd/Ca ratio from the 1965-1980 means (Fig. 3b) highlight individual annual cycles that were characterized by unusually weak (El Niiio) or strong upwelling. For the most part, the temperature and Cd indices concur, though the above mentioned time scale adjustments would improve the correspondence. The 1965 El Niiio appears only weakly in the Cd/Ca determination, and 1977 and 1978 Cd levels deviate from the temperature pattern for reasons that are presently unknown. In spite of these inconsistencies, however, the overall resemblance between two very different upwelling tracers is obvious. A second brief Cd/Ca time series based on a young coral (Pavona clavus) collected from Urvina Bay, Isabela Island (Galapagos Islands) exhibits similar Cd levels, but correspondence to regional temperature anomalies is in this case poorer than at
Figure 3 (a) Cd/Ca mole ratios in P . clavus (Punta Pitt, San Cristobal Island, Galapagos Islands) compared with sea surface temperature at Academy Bay (Santa Cruz Island, Galapagos Islands) for period 1965-1979. C d C a measurements were determined on approximate 3-month growth increments; temperatures are 3-month averages. (Temperature data courtesy of Charles Darwin Research Station, Galapagos Islands) (b) Quarterly anomalies in sea surface temperature and C d C a ratio in P . clavus from 1965-1980 means. Note that Cd/Ca negative anomaly from 1969-1970 should be compressed and positive anomaly during late 1970 expanded due to a band sectioning error.
265
1-9 6 5
1970
1980
1975
Year
(a 1
1865 2.5-r
1970 I
t I
1975
1
1980
1
I
r3
266
Punta Pitt (Fig. 4a). Nonetheless, the 1972 and 1976 El Niiio events are evident in the record, which covers eight annual cycles. Curiously, the 1976 event is reflected by a more severe Cd decline than recorded in 1972, though the 1972 event was stronger by most instrumental indices. More significant though, is the behavior of Mn at Urvina Bay as recorded by this coral. Pronounced seasonal cycling appears to occur in phase with temperature changes and six months out of phase with respect to Cd. Such a mirror-image response is expected on the part of Mn because of its reverse gradient in eastern tropical Pacific surface waters relative to Cd and nutrients (Fig, 2a). A strong surface Mn pulse observed north of Galapagos by Klinkhammer and Bender (1980) is an advective feature originating from reducing sediments of the CentraWouth American continental shelf. Similar lateral injection of Mn has been observed at VERTEX stations west of California by Martin et al. (1985). A deeper Mn maximum (450 m) associated with the intense oxygen minimum found in this part of the Pacific Ocean lies beyond the reach of windinduced Ekman pumping. The greater consistency and amplitude of the coral Mn signal in relation to Cd may derive from differences in the vertical gradients in dissolved Mn and Cd in the upwelled waters off Urvina Bay. Interestingly, Mn cycling at Punta Pitt (Fig. 4b) appears not so steady, though traces clearly persist (shaded cycles). A lower average Mn/Ca ratio at Punta Pitt (80 versus 100 nmol/mol at Urvina Bay) suggests that a supplemental Mn flux at Urvina Bay augments that originating from the continental shelf one thousand kilometers to the east. This source is likely to be the Urvina Bay shelf ( 4 0 0 m depth), which extends for some 30 km southwest of the collection site. A second significant observation in the case of the Punta Pitt coral is that Mn cycles appear to flatten precisely during El Niiio episodes (i.e. 1965, 1969, 1972, 1976). In locations similar to Punta Pitt, this behavior may prove instrumental in identifying and corroborating past occurrences of ENS0 activity. Linn et al. (1987) have interpreted Urvina Bay coral Mn levels from the 1950s in this context. Interpretation of Cd and Mn as El Niiio markers in the coastal upwelling environment of the Gulf of Panama is more complex. Reefs dwelling within the coastal zone of Central America are directly influenced by metal fluxes released from reducing sediments along the continental shelf. This source would be expected to be particularly strong for Mn since advective pulses of this metal are visible for thousands of kilometers to the west (Martin et al., 1985). Skeletal Mn/Ca levels of 300-800 nmol Mn/mol Ca in Pavona gigantea from Contadora Island (Fig. 5) (compared with 65-150 nmol Mn/mol Ca at Galapagos) confirm the presence of substantial diagenetic Mn. Cadmium levels, too, are enhanced on average by a factor of 3-4 over those observed in Galapagos corals. Nearshore Cd excesses are likely to result from a combination of remobilization from sediments, fluvial discharge, and particulate desorption (Boyle et al., 1984).
267
6-
I*
12 -
Urvina Bay
- 160
(Pavona clavus)
m
h
2
0
10-120
0 5
!?& '0
3
8-
-E
6-
Mn/Ca
. -80
0 Cd/Ca
4-
eaf ==sz =
- 40 0
2
Punta Pitt (Pavona clavus)
1965
1970
Year
1975
25 1980
Figure 4 (a) CdICa and MnKa mole ratios in P. clavus from Urvina Bay (Isabela Island, Galapagos Islands). Sampling period is 1969- 1978; sampling frequency is approximately tri-monthly. Note that Cd and Mn annual cycles are roughly 6-months out-of-phase due to upwelling of mirrorimage distributions in the upper ocean. (b) Cd/Ca and MnICa distributions in P. clavus from Punta Pitt (San Cristobal Island, Galapagos Islands). Sampling period spans 1965-1979; sampling frequency is approximately trimonthly. Stippled areas in CdICa series represent El Nifio periods. Note how Mn cyclicity appears to weaken during El Nifio at this location.
i3
-E
268
In spite of such grossly enriched background levels of Cd and Mn, a few features in the Contadora Island record are recognizable. For example, upwelling seasonality is plainly evident in the Mn/Ca time series even though sampling intervals were relatively coarse (every 6 months). In contrast to what was observed at the Galapagos Islands, however, strong upwelling (January-June) is reflected by elevated Mn/Ca ratios, indicating that a more nutrient-like gradient exists in slope and shelf waters. This distribution may be controlled by sediment injection of Mn into the shallow water column near Panama. The vertical gradient here must also be very steep as the amplitude of seasonal Mn oscillations in the coral sometimes reaches several hundred nanomoles of Mn/mole Ca (a difference of 100 nmol Mn/mol Ca would be equivalent to a dissolved Mn concentration shift of 10 nM if K D = 0.1). The range of observed interannual Cd/Ca ratios is also quite wide (9-20 nmol Cd/mol Ca), suggesting that the background level of dissolved Cd is variable. An obvious 6-month cycle in the Cd record is not evident. El Niiio events are difficult to specify unambiguously in the Gulf of Panama. Cadmium levels during the strong 1972 event are certainly low, but similar levels are found the year before and after, as well as during 1976-1979. The 1969 El Niiio is not accompanied by unusual activity in either trace metal index. It should be pointed out, however, that the temperature/chemical history of this region is not known in any detail. Local effects such as observed in the Gulf of California, can substantially alter the expected timing and sequence of events associated with El Niiio (Baumgartner et al., 1985).
, ,””
“V
Contadora Island
-
- 900
(Pavona gigantea)
40 -
m
’
- 700
0
30-
9 5 -00 0
=
20-
- 300
CdICa
I
10
o
‘
luz
MnlCa -500
E
-:
3
-
9 c5 I-
-E
- 100 ~
’
’
’
’
r
”
Figure 5 Cd/Ca and Mn/Ca variability in Panama. Measurements over 1967- 1979 approximate 6-month growth increments. persists despite an enhanced background remobilization.
’
.
~
’
.
.
.
,-100
P. gigantea from Contadora Island, Gulf of were performed on Mn upwelling seasonality attributable to diagenetic shelf
269
5.2 PreciDitation in the west ern tropical Pacific Historical changes in precipitation patterns have been inferred using annual ringwidth variations in trees (see Fritts, 1987; Lough and Fritts, this volume), shoreline features associated with changes in lake levels (e.g. Maley, 1981; Street and Grove, 1976), and reconstructed changes in terrestrial vegetation cover (e.g. Heusser, 1978). In the oceans, the most sensitive indicator of rainfall/salinity change is the 1 8 0 / 1 6 0 isotopic ratio, which is also recorded in corals. Temperature effects on oxygen isotope fractionation, however, can compete with and often override subtle changes in salinity due to rainfall or river discharge. Additionally, the isotopic content of tropical precipitation is not always readily predictable (Dansgaard, 1964), though dramatic progress has been made in mapping of global isotopic precipitation fields (Jouzel et al., 1987). Fairbanks and Dodge (1979) and Cole and Fairbanks (in prep.) have proposed a second potential oceanic tracer of precipitation in the carbon isotopic record of corals. The essence of this tracer system is solar radiational control over photosynthetic fractionation of 12C and 1 3 C by algal symbionts that reside within corals. If confounding influences of photoinhibition (McConnaughey, 1986) and shading (Patzold, 1986) are minimized, the carbon isotopic content of skeletal material is observed to become lighter during the cloudy (rainy) season. Simply put, reduced radiation causes reduced photosynthesis, which in turn causes relatively less 13C enrichment of the local dissolved inorganic carbon pool (due to reduced carbon fixation in organic tissues and attendant fractionation). Figure 6 depicts the results of a pilot study of a coral from Tarawa Atoll (Republic of Kiribati) to investigate Cd variability in relation to precipitation. Tarawa is expected to be a favorable test site since absolute rainfall levels are high (approximately 150 cm/yr) and interannual variations are extreme (300-400%) and appear to be driven by ENS0 activity (Taylor, 1973; Cole and Fairbanks, in prep.). Figures 6a and b show C d K a ratios determined on 4-12 month coral increments that grew with anomalously enriched or depleted 1 8 0 and 13C. The results indicate that Cd increases both with decreasing 61 8 0 (increasing rainfall) and decreasing 6 l 3 C (increasing cloudiness/rainfall). The Cd-613C relationship is also consistent with an upwelling mechanism (S13C seawater distribution is opposite to nutrients; see Kroopnick et al., 1970), however, chemical and thermal gradients in central equatorial Pacific surface waters are very mild (e.g. 1-20C annual surface temperature range), so pronounced variations due to upwelling are not expected. Furthermore, the rainy season (August-January) and SST cold phase (FebruaryJuly) occur at opposite times during the year, so over an annual cycle, these effects should tend to cancel rather than be additive. The Cd-6180 relationship is exactly the reverse of that expected by upwelling-induced temperature change. Thus, the distribution of Cd in the surface ocean near Tarawa appears to be responding in some way to rainfall. We can test the plausibility of direct wet deposition of Cd by estimating the required Cd content of tropical rains at Tarawa and comparing this
270
value to remote measurements reported in the literature. The average salinity change from June to December in the upper 75 m near Tarawa is about +0.4 O/oo (based on data compiled by the Geophysical Fluid Dynamics Laboratory). This salinity increase can be accounted for by the addition of 0.86 m of rainwater to the surface mixed layer each winter. If the dissolved seawater Cd concentrations implied by coral measurements (4 - 25 pM assuming KD = 1.0; see Shen et al., 1987) represent the equivalent seasonal range, we would require an effective rainwater Cd concentration of 985 pM. This result appears high in comparison with wet deposition measurements at Bermuda (535 pM -- Jickells et al., 1984), and even less likely in light of SEAREX measurements of only 19 pM at Enewetak (1 lo2O'N, 162020E) (Arimoto et al., 1985). One possible explanation might be that shallow reef waters near Tarawa amplify chemical changes associated with rainfall through salinity stratification. The observed 6180 range of nearly 10/00 suggests that this effect may in fact be important. Nevertheless, if the Cd value of 19 pM recorded by Arimoto and co-workers is representative of western tropical Pacific precipitation, then rainfall cannot be a significant source of Cd to the surface ocean at Tarawa in any circumstance. The only other viable explanation is island runoff and/or enhanced resuspension and release of particulate Cd in the reef environment of Tarawa during the rainy season. It is difficult to make quantitative estimates as to the influence of such a process as coastal observations are few and the atoll environment unique. Figure 7 depicts one example of the potential influence of an island land mass on dissolved and total Cd levels in surface waters. The samples are from James Bay, Santiago Island in the Galapagos archipelago. While the transect has little to do with rainfall (average precipitation at Galapagos is a scant 30 cm/yr), it does show that dissolved Cd levels are enhanced in the coastal zone (0-2 km) about this small island. This is most likely a result of diagenetic release from nearshore sediments, but it seems plausible that runoff resulting from heavy rainfall (up to 0.6 m/month during ENSO) might produce comparable effects. Despite the improbability of a surface ocean Cd anomaly directly attributable to
Figure 6 (a) Cd/Ca versus 6 1 8 0 in H. microconos from Tarawa Atoll. Increasing Cd/Ca follows depletion of 1 8 0 caused by heavy tropical precipitation. (b) Cd/Ca versus 613C in H. microconos. Cd/Ca increases as 13C becomes depleted; the carbon shift most likely results from decreased photosynthesis during seasons of high rainfall and cloud cover. ( 6 1 8 0 and 613C values were estimated from measurements made along a separate sampling transect on the same coral slab. Stable isotope data were kindly provided by Julia Cole of Lamont-Doherty Geological Observatory.)
27 1
Estimated 6 1 8 0 (a 1
Estimated 613C (b)
272
wet deposition, it is interesting to consider the influence of large precipitation volumes on other trace element distributions in the ocean. Such elements as Zn, Se, and Fe appear to be appreciably enriched in rain (Arimoto et al., 1985) relative to seawater and might therefore show related temporal changes. The problems remain of locating suitable sample sites and establishing the presence of such metals in a permanent host phase within the skeletal structure of corals.
5.3 River discharge and circulation in the Caribbean Sea Pursuing a related strategy but changing settings to the Caribbean Sea, it may be possible to gain perspective on terrestrial responses to ENSO on the South American continent. A number of climate adjustments in South America have been recognized as ENSO teleconnections. Most dramatic are the heavy rains that accompany El Niiio in normally arid northern Peru. Historical accounts of flooding of this coastal desert have provided a means of extending the chronology of El Niiio
60
Cadmium in Seawater James Bay, Santiago Island, Galapagos Islands
50
( 10 / 1 9 / 8 6 )
40 0.
v
3
30 20
0
1
2
3
Distance from shore Figure 7 Total and dissolved (<0.45 Fm) Cd the Galapagos Islands in October 1986. High levels within 2 km of Santiago Island indicate resuspended Cd. Beyond 3 km, the dissolved stabilize at the open ocean background level.
5
4
6
(km)
concentrations in nearshore waters of dissolved and particulate the presence of Cd concentration appears to
273
events to as far back as 1541 (Hamilton and Garcia, 1986; Quinn et al., 1987). Wells (1987) and Rollins et al. (1986) have examined overbank flood deposits and marine faunal distributions along the west coast of South America to explore the incidence of El Nirio over past centuries to millenia. Elsewhere on the South American continent, a correlation exists between rainfall over northeast Brazil and ENSO (Hastenrath and Heller, 1977; Rao et al., 1986; Ropelewski and Halpert, 1987) whereby El Nirio episodes tend to be accompanied by drought conditions. Much the opposite is observed in a narrow band along southeastern South America (Ropelewski and Halpert, 1987). Evidence that such precipitation anomalies affect river flow discharge (Amazon - Richey et al., submitted; Trombetas, Jy Parana, Parana - Molion and Moraes, 1986) suggests a possible route to ENSO reconstruction via historical river flow records in corals. Among the possible chemical tracers delivered to the oceans by rivers, barium and radium offer great potential for two reasons. The first is that the river flux of both these elements to the ocean is large. In the case of Ba, for example, a concentration of 125 nM in the Amazon River (Boyle, 1976) is about three times that of the surface ocean. Barium desorption from riverine particulate matter boosts the effective freshwater endmember concentration to about 265 nM (Boyle, 1976). Similar behavior has been observed in the Mississippi and Zaire River estuaries (Hanor and Chan, 1977; Edmond et al., 1978). Amazon endmember concentrations of 226Ra and 228Ra are actually not very different from those of seawater, however, radium addition via particulate desorption and diffusive flux from estuarine sediments results in a large effective river source (Moore and Edmond, 1984; Key et al., 1985). The 228Ra/226Ra activity ratio is a particularly useful tracer of river penetration in surface ocean waters as 228Ra is preferentially enriched during estuarine mixing to yield a 228Ra/226Ra ratio that is nearly an order of magnitude greater than that found in seawater (Moore et al., 1986). The second reason Ba and Ra are especially attractive candidate tracers is that both alkaline earth metals are known to form orthorhombic carbonates (see Speer, 1983). Thus, they are likely to comprise permanent markers in reef-building corals. Terrestrial markers such as these might be used very effectively in concert with previously identified flourescent marker compounds in corals (fulvic acids) that also derive from continental runoff (Isdale, 1984; Boto and Isdale, 1985). We chose to test Ba as a paleo-river flow indicator because of its expected wide dynamic range [effective endmember compositions of 265 nM (fluvial - Boyle, 1976) and 43 nM (seawater - Chan et al., 1977)] and relative measurement ease in comparison to radiochemical assay. Benninger and Dodge (1986) have detected 228Ra levels of 20-35 dpm/kg in Montastrea annularis from St. Croix that are high in comparison to open ocean and river endmember concentrations of 1 and 45 dpm/100 kg (Moore et al., 1986) (KD = 1 implies endmember coral 2 2 8 R a concentrations of 1 and 45 dpm/kg CaC03). Apparently, a significant local source of
274
22813, exists at Tague Bay, St. Croix, that must be subtracted in order to evaluate Amazon River influence at this location using this tracer. Ultimately, use of 2 2 8 R a is also limited by its brief half-life of 5.7 years. Application of 226Ra (t1/2 = 1,600 yrs) is not limited in the modern time domain, but the z26Ra signal is expected to be smaller than that of Ba since maximum estuarine concentrations are only about three times higher than surface seawater (Key et al., 1985). Figure 8 shows a tenyear Ba/Ca timeseries (1967-1977) extracted from a coral collected from Barbados. Though these measurements were simpler to carry out than paired 226Ra-228R a determinations, Ba analysis by furnace atomic absorption is problematic. Since Ba is atomized only at very high temperatures (> 2,600OC) and tends to form refractory carbides, graphite tube life is generally short and peak shape is difficult to optimize. Additionally, broad band emission from the furnace can increase noise. For these reasons, precision is often poor (+lo% or worse), particularly between runs. We believe our within run precision is closer to 5% ( l a ) , though there is uncertainty as to the relative placement of the data series from 1971-1973, which was run independently. Despite these analytical difficulties, it is very clear that a seasonal cycle exists in the Ba record.
275
Barbados BalCa
150
Amazon Streamflow @I Manacapuru 25 1966
1968
1970
1972
1974
1976
1978
Year
Figure 8 Top: Quarter-annual Ba/Ca ratios in M. annularis from Barbados, 19661977. Vertical grid lines mark the boundary between low- and highdensity growth bands estimated to occur near August of each year. Bottom: Monthly average Amazon discharge at Manacapuru for period 1966-1977 (calculated from daily stage of Manaus vs. Manacapuru - data courtesy of Jeff Richey). Amazon discharge cycle (peak generally in MayJuly) anticipates Ba/Ca cycle by about 2-4 months (see text).
275
It has been known for many years that low salinity waters intrude upon Barbados annually during the summer (minimum salinity usually occurs in August) (Parr, 1937; Steven and Brooks, 1972). Though the source of these low salinity plumes has been ascribed to Amazon and Orinoco discharge and tropical Atlantic precipitation, the relative importance of these inputs has been argued. More recently, Froelich et al. (1978) and Borstad (1982) appear to have converged on a value near 60% as the contribution of Amazon waters to measured silicate and salinity anomalies in the area. The remainder is likely supplied from rainfall within the westward flowing North and South equatorial currents. Local precipitation effects have been shown to be minimal (Steven and Brooks, 1972). If in fact the seasonal Ba oscillation at Barbados is driven primarily by variations in Amazon discharge, then the relative timing of these cycles must be consistent. Vertical grid lines in Fig. 8 denote the base of the high-density coral bands, which we estimate are accreted from August to October. Band widths ranged from 6-11 mm/yr leaving in some cases only about one millimeter of usable material after "trimonthly" sectioning with a jeweler's saw. As a result, time assignment errors of several months are easily possible due to sectioning errors, but for the most part we appear to have captured consistently the same phase of the annual cycle. The observed phasing indicates that maximum Ba levels occur from August-October, which is consistent with the August salinity minimum at Barbados. Therefore, the two signals appear to share a common source. This conclusion can be reinforced by a calculation as to the expected amplitude of the Ba cycle. The average annual salinity change at Barbados appears to be of the order 20/00 (only 4 complete annual cycles have been documented - Steven and Brooks, 1972; Borstad, 1982). Assuming that this shift is entirely due to river discharge, then at the height of the salinity minimum, about 6% of the surface water at Barbados would have to be of fluvial origin. Given freshwater and seawater endmember Ba concentrations of 265 nM and 43 nM, respectively, the maximum Ba anomaly should be +13 nM. This is almost exactly the peak-to-trough amplitude of the coral signal when it is converted to a dissolved Ba scale (since K D = ~ ,multiply Ba/Ca axis by 10 and change units to nM). In converting the coral measurements to seawater Ba concentrations, a final detail emerges: the time series baseline falls at 48 nM Ba versus the cited value of 43 nM measured some distance away at GEOSECS station 33 (210N, 54OW). This minor discrepancy points out either that KD
for Ba in corals is actually closer to 1.2 than 1.0, or that a background fluvial Ba component (7 nM) is always present at Barbados (maximum salinity 360/00), which does not exist at GEOSECS station 33 (surface salinity 37.l0/0o). The actual timing of Amazon peak discharge (May-June), however, occurs 2-4 months prior to the minimum salinity/maximum Ba effect observed at Barbados (Fig. 8). This time lag is reasonable in light of the horizontal distance involved (2,000 km), but influences on the travel path are numerous and varied. Using
276
satellite images and drifting buoys, Muller-Karger et al. (1988) determined that throughout much of the year, Amazon discharge tends to be carried offshore from Brazil and then eastward toward Africa by the North Equatorial Countercurrent. Eventually, much of this flow is entrained by the North Equatorial Current to arrive in the Caribbean from several months to a year later. From February through May, however, Amazon discharge is advected more rapidly into the Caribbean along a narrow boundary current flowing to the northwest. Within this boundary current, speeds of 0.9 m/sec suggest a travel time of one month to Barbados, however, additional time would realistically be needed as the boundary current diverges from the continent beyond Guiana and becomes more diffuse. Thus, it appears that timing of the salinity minimum (and Ba maximum) may be influenced as much by regional circulation as by Amazon river stage. To compound the difficulty of separating these effects, dispersal of the freshwater plume varies from year to year (Muller-Karger et al., 1988) and exhibits patchiness (Deuser et al., 1988). We believe it is premature to attempt an analysis of the differences between the Ba/Ca and Amazon River stage timeseries shown in Fig. 8 because of analytical uncertainties and the above mentioned complexities. Ultimately, Barbados may prove too distant from the Amazon estuary to enable an unambiguous reconstruction of historical discharge. A longer, carefully analyzed timeseries is needed to evaluate definitively the potential of this site. In Fig. 9, we show additional results from what might constitute a better location, the island of Tobago. Located only 180 km north of the mouth of the Orinoco River, corals here might record chemical anomalies associated more directly with South American river flow. Unfortunately, several corals collected from Bucco Reef for this study exhibited slow growth (2-6 mm/yr) and/or very faint banding. The study coral selected in Fig. 9 was analyzed at annual increments because of the narrowness of growth layers. In averaging out an anticipated strong seasonal cycle (Orinoco annual discharge is only 15% that of the Amazon, but seasonal flow varies by a factor of 6 ) , we observe no interannual variability beyond measurement errors in the nine-year period spanning 1975-1983. Overall levels are similar to those found in Montastrea annularis at Barbados. The fact that a major anomaly is not perceptible during the 1982-83 El Nifio may indicate that precipitation over the Orinoco drainage basin in Venezuela is not sensitive to ENS0 perturbations. Apparently, a similar conclusion has been inferred in a preliminary statistical examination of Orinoco streamflow data (T. Herbert, pers. comm.). These Caribbean trace element records represent exploratory studies that have established the existence of a useful historical river discharge indicator in corals. It remains to identify other climatologically-sensitive areas where continuous coral sections, can be used to reconstruct land-sea exchanges in the past.
277
-
Tobago (Diploria strigosa)
6-
0
?+I-+
mo
% I a m
5-
-5
4-
m -
-
1974
1976
1978
1980
1982
1984
Year
Figure 9
Annual Ba/Ca mole ratio in D. strigosa from Tobago for period 19751983. Within lo errors (based on triplicate analysis), there is no apparent inter-annual variability. An expected strong seasonal signal associated with Orinoco streamflow has been averaged out by annual sampling; this coral grew at only 3-4 mm/yr.
6 CONCLUSIONS The surface ocean is influenced by a wide variety of physical phenomena that are associated with climate changes. A few of these phenomena may actually help drive such changes; for example, equatorial Pacific tradewinds and sea surface temperature anomalies are thought to excite internal waves that bring about El Nifio. Others may be viewed more appropriately as responses to altered climate states such as enhanced or diminished river flow in South America. Still other phenomena may emerge from an initial disequilibrium state to then themselves regulate subsequent climatic development -- examples here might include altered rainfall and surface ocean circulation patterns. Long term records of each type of physical change carry information that can be interpreted from diverse perspectives. Our findings constitute in large measure, groundwork for future studies of historical climate using new paleochemical indicators. We have established that trace but reproducibly analyzed quantities of Cd, Mn, and Ba exist in reef-building corals that passively record historical perturbations to the surface ocean environment. Cadmium appears sensitive to changes in upwelling by virtue of its nutrient-like distribution and also exhibits an apparent rainfall response in the western equatorial Pacific that is not completely understood. Manganese can be exploited to deduce changes in long range transport of continental shelf waters, as well as upwelling variability in the reverse sense of Cd. Barium (and in all likelihood, radium) is potentially an effective tracer of continental precipitation and
278
runoff due to its abundance in rivers and estuaries in relation to surface seawater. Undoubtedly, the list will go on as additional elements are explored and analytical capabilities improve. A suite of trace element indicators when combined with existing stable isotopic, radiocarbon, and band width/density indices, will greatly enhance the interpretability of the coral record. 7 ACKNOWLEDGMENTS We wish to thank Robert Dunbar, Gerard Wellington, Mitchell Colgan, Peter Glynn, Richard Fairbanks, and Eric Brown for recovering and sharing the coral samples used throughout this study. Julia Cole is thanked for her help in tracking down appropriate coral sections and for furnishing the stable isotopic data from the Tarawa coral. We are grateful to Jeffrey Richey for supplying Amazon River stage data. Thoughtful comments by Edward Boyle, Gary Klinkhammer, Steven Smith, Julia Cole, and Peter Glynn contributed much to improve this manuscript. This work was supported by NSF grant ATM-86-18064 (G.T.S.) and an NSF undergraduate research grant to LDGO (C.L.S.). We also acknowledge NSF grant OCE8415615 (P.W. Glynn) through which the Panamanian coral was collected.
8 REFERENCES
Amiel, A.J., Friedman, G.M. and Miller, D.S., 1973. Distribution and nature of incorporation of trace elements in modern aragonitic corals. Sedimentology, 20: 47-64. Angus, J.G., Raynor, J.B. and Robson, M., 1979. Reliability of experimental partition coefficients in carbonate systems: Evidence for inhomogeneous distribution of impurity cations. Chem. Geol., 27: 181-205. Arimoto, R., Duce, R.A., Ray, B.J. and Unni, C.K., 1985. Atmospheric trace elements at Enewetak Atoll: 2. Transport to the ocean by wet and dry deposition. J. Geophys. Res., 90: 2391-2408. Baumgartner, T., Ferreira-Bartrina, V., Schrader, H. and Soutar, A., 1985. A 20-year varve record of siliceous phytoplankton variability in the central Gulf of California. Mar. Geol., 64: 113-129. Benninger, L.K. and Dodge, R.E., 1986. Fallout plutonium and natural radionuclides in annual bands of the coral Montastrea annularis, St. Croix, U.S. Virgin Islands. Geochim. Cosmochim. Acta, 50: 2785-2797. Borstad, G.A., 1982. The influence of the meandering Guiana Current and Amazon River discharge on surface salinity near Barbados. J. Mar. Res., 40: 421-434. Boto, K. and Isdale, P., 1985. Fluorescent bands in massive corals result from terrestrial fulvic acid inputs to nearshore zone. Nature, 315: 396-397. Boyle, E.A., 1976. The marine geochemistry of trace metals. Ph.D. thesis, Massachusetts Inst. Technol./Woods Hole Oceanogr. Inst., Mass., 156 pp. Boyle, E.A., 1986. Paired carbon isotope and cadmium data from benthic foraminifera: implications for changes in oceanic phosphorus, oceanic circulation, and atmospheric carbon dioxide. Geochim. Cosmochim. Acta, 50: 265-276. Boyle, E.A., 1988. Cadmium: Chemical tracer of deepwater paleoceanography. Paleoceanogr., 3: 47 1-489.
279
Boyle, E.A., Sclater, F. and Edmond, J.M., 1976. On the marine geochemistry of cadmium. Nature, 263: 42-44. Boyle, E.A., Huested, S.S. and Jones, S.P., 1981. On the distribution of copper, nickel and cadmium in the surface waters of the North Atlantic and North Pacific Ocean. J. Geophys. Res., 86: 8048-8066. Boyle, E.A., Reid, D.F., Huested, S.S. and Hering, J., 1984. Trace metals and radium in the Gulf of Mexico: An evaluation of river and continental shelf sources. Earth Planet. Sci. Lett., 69: 69-87. Broecker, W.S., 1963. A preliminary evaluation of uranium series inequilibrium as a tool for absolute age measurement on marine carbonates. J. Geophys. Res., 68: 2817-2834. Bruland, K.W., 1980. Oceanographic distributions of cadmium, zinc, nickel, and copper in the North Pacific. Earth Planet. Sci. Lett., 47: 176-198. Bruland, K.W. and Franks, R.P., 1983. Mn, Ni, Cu, Zn and Cd in the western North Atlantic. In: C.S. Wong, E. Boyle, K.W. Bruland, J.D. Burton and E.D. Goldberg (Editors), Trace Metals in Seawater. Plenum, New York, pp. 395-414. Buddemeier, R.W., Schneider, R.C. and Smith, S.V., 1981. The alkaline earth chemistry of corals. Proc. 4th Int. Coral Reef Symposium, Manila, 2: 81-85. Chan, L.H., Drummond, D., Edmond, J.M. and Grant, B., 1977. On the barium data from the Atlantic GEOSECS Expedition. Deep-sea Res., 24: 613-649. Chang, L.L.Y. and Brice, W.R., 1971. Subsolidus phase relations in the system calcium carbonate - cadmium carbonate. Am. Mineral., 56: 338-341. Chappell, J. and Polach, H.A., 1972. Some effects of partial recrystallization on 1% dating late Pleistocene corals and molluscs. Quat. Res., 2: 244-252. Chave, K.E., 1954. Aspects of the biochemistry of magnesium. 1. Calcareous marine organisms. J. Geol., 62: 266-283. Cole, J. and Fairbanks, R.G, in prep. Indonesian low migration during El Nino/Southern Oscillation events recorded by the stable isotopes of a coral. Comans, R.N.J. and Middelburg, J.J., 1987. Sorption of trace metals on calcite: Applicability of the surface precipitation model. Geochim. Cosmochim. Acta, 5 1: 2587-2591. Cross, T.S. and Cross, B.W., 1983. U, Sr and Mg in Holocene and Pleistocene corals. J. Sediment. Petrol., 53: 587-594. Danielsson, L.G. and Westerlund, S . , 1983. Trace metals in the Arctic Ocean. In: C.S. Wong, E. Boyle, K.W. Bruland, J.D. Burton and E.D. Goldberg (Editors), Trace Metals in Seawater. Plenum, New York, pp. 85-96. Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus, 16: 436-468. Davis, J.A., Fuller, C.C. and Cook, A.D., 1987. A model for trace metal sorption processes at the calcite surface: Adsorption of Cd2+ and subsequent solid solution formation. Geochim. Cosmochim. Acta, 51: 1477-1490. Deuser, W., Muller-Karger, F.E. and Hemleben, C., 1988. Temporal variations of particle fluxes in the deep subtropical and tropical North Atlantic: Eulerian versus LaGrangian effects. J. Geophys. Res., 93: 6857-6862. Dodge, R.E. and Thomson, J., 1974. The natural radiochemical and growth records in contemporary hermatypic corals from the Atlantic and Caribbean. Earth Planet. Sci. Lett., 23: 313-322. Dodge, R.E. and Gilbert, T.R., 1984. Chronology of lead pollution contained in banded coral skeletons. Mar. Biol., 82: 9-13. Dodge, R.E., Jickells, T.D., Knap., A.H., Boyd, S. and Bak, R.P.M., 1984. Reef-building coral skeletons as chemical pollution (phosphorus) indicators. Mar. Poll. Bull., 15~178-187.
280
Donguy, J.R. and Henin, C., 1980. Surface conditions in the eastern equatorial Pacific related to the intertropical convergence zone of the winds. Deep-sea Res., 27A: 693-714. Druffel, E.M. and Linick, T.W., 1978. Radiocarbon in annual coral rings of Florida. Geophys. Res. Lett., 5: 913-916. Edmond, J.M., Boyle, E.A., Drummond, D., Grant, B. and Mislick, T., 1978. Desorption of barium in the plume of the Zaire (Congo) River. Neth. J. Sea Res., 12: 324-328. Edwards, R.L., Chen, J.H. and Wasserburg, G.J., 1987. 238U-234U-23oTh-232Th systematics and the precise measurement of time over the past 500,000 years. Earth Planet. Sci. Lett., 81: 175-192. Fairbanks, R.G. and Dodge, R.E., 1979. Annual periodicity of the 1 8 0 / 1 6 0 and 13C/12C ratios in the coral Montastrea annufaris. Geochim. Cosmochim. Acta, 43: 1009-1020. Flor, T.H. and Moore, W.S., 1977. Radium/calcium and uranium/calcium determinations for Western Atlantic reef corals. Proc. 3rd Int. Coral Reef Symp., Miami, FL, 2: 555-561. Fritts, H.C., 1987. Tree-ring analysis (dendroclimatology). In: J.E. Oliver and R.W. Fairbridge (Editors), The Encyclopedia of Climatology. Van Nostrand Reinhold, New York, pp. 858-875. Froelich, P.N., Atwood, D.K. and Giese, G.S.,1978. Influence of Amazon River discharge on surface salinity and dissolved silicate concentration in the Caribbean Sea. Deep-sea Res., 25: 735-744. Glynn, P.W. and Wellington, G.M., 1983. Corals and Coral Reefs of the Galapagos Islands. University of California Press, Berkeley, CA, 330 pp. Gvirtzman, G.,Friedman, G.M. and Miller, D.S., 1973. Control and distributions of uranium in coral reefs during diagenesis. J. Sediment. Petrol., 43: 985-997. Hamilton, K. and Garcia, R.R., 1986. El Niiio/Southern Oscillation events and their associated mid-latitude teleconnections, 1531-1841. Bull. Am. Meteor. SOC., 67: 1354-1 361. Hanor, J.S. and Chan, L.H., 1977. Non-conservative behavior of barium during mixing of Mississippi River and Gulf of Mexico waters. Earth Planet. Sci. Lett., 37: 242250. Harriss, R.C. and Almy, C.C., 1964. A preliminary investigation into the incorporation and distribution of minor elements in the skeletal material of scleractinian corals. Bull. Mar. Sci. Gulf Caribb., 14: 418-423. Hastenrath, S. and Heller, L., 1977. Dynamics of climatic hazards in northeast Brazil. Quart. J. R. Met. SOC., 103: 77-92 Heusser, L., 1978. Pollen in Santa Barbara Basin, California: A 12,000-yr record. Geol. SOC. Am. Bull., 89: 673-678. Hodge, V.F., Stallard, M., Koide, M. and Goldberg, E.D., 1985. Platinum and the platinum anomaly in the marine environment. Earth Planet. Sci. Lett., 72: 158172. Houck, J.E., Buddemeier, R.W., and Chave, K.E.. 1975. Skeletal low-magnesium calcite in living scleractinian corals. Science, 189: 997-999. Howard, L.S. and Brown, B.E., 1984. Heavy metals and reef corals. Oceanogr. Mar. Biol. Ann. Rev., 22: 195-210. Isdale, P., 1984. Fluorescent bands in massive corals record centuries of coastal rainfall. Nature, 310: 578-579. Jickells, T.D., Knap, A.H. and Church, T.M., 1984. Trace metals in Bermuda rainwater. J. Geophys. Res., 89: 1423-1428.
.
28 1
Jouzel, J., Russell, G.L., Suozzo, R.J., Koster, R.D., White, J.W.C. and Broecker, W.S., 1987. Simulations of the HDO and H2180 atmospheric cycles using the NASA GISS general circulation model: The seasonal cycle for present-day conditions. J. Geophys. Res., 92: 14739-14760. Key, R.M., Stallard, R.F., Moore, W.S. and Sarmiento, J.L., 1985. Distribution and flux of 226Ra and 228Ra in the Amazon River estuary. J. Geophys. Res.. 90: 6995-7004. Kitano, Y., Kanamori. N. and Yoshioka, S., 1976. Adsorption of zinc and copper ions on calcite and aragonite and its influence on the transformation of aragonite to calcite. Geochem. J., 10: 175-179. Klinkhammer, G. P. and Bender, M.L., 1980. The distribution of manganese in the Pacific Ocean. Earth Planet. Sci. Lett., 46: 361-384. Kroopnick, P., Deuser, W.G. and Craig, H., 1970. Carbon 13 measurements on dissolved inorganic carbon at the North Pacific (1969) Geosecs station. J. Geophys. Res., 75: 7668-7671. Landing, W.M. and Bruland, K.W., 1987. The contrasting biogeochemistry of iron and manganese in the Pacific Ocean, Geochim. Cosmochim. Acta, 5 1: 29-43. Lange, C.B., Berger, W.H., Burke, S.K., Casey, R.E., Schimmelmann, A., Soutar, A. and Weinheimer, A.L., 1987a. El NiAo in Santa Barbara Basin: Diatom, radiolarian and foraminiferan responses to the "1983 El Niiio" event. Mar. Geol., 78: 153-160. Lange, C.B., Schimmelmann, A., Burke, S.K., Tont, S.A., Berger, W.H., Soutar, A., Casey, R.E. and Weinheimer, A.L., 1987b. Microfossil and geochemical responses to the 1983 El NiAo in the Santa Barbara Basin, California. EOS, Trans. Am. Geophys. Union, 68: 1778. Linn, L.J., Druffel, E.R.M. and Delaney, M.L., 1987. Trace metal concentrations in a Galapagos coral: An ENS0 indicator. EOS Trans. Am. Geophys. Union, 68: 17431744. Livingston, H.D. and Thompson, G., 1971. Trace element concentrations in some modern corals. Limnol. Oceanogr., 16: 786-796. Maley, J., 1981. Etudes palynologiques dans le bassin du Tehad et paleoclimatologie de 1'Afrique nord-tropicale de 30 000 ans a l'epoque actuelle. Paris: ORSTROM. Martin, J.H., Bruland, K.W. and Broenkow, W.W., 1976. Cadmium transfer in the California Current. In: H. Windom and R. Duce (Editors), Marine Pollutant Transfer. Lexington Books, Lexington, MA, pp. 159-184. Martin, J.H., Knauer, G.A. and Broenkow, W.W., 1985. VERTEX: the lateral transport of manganese in the northeast Pacific. Deep-sea Res., 32: 1404- 1427. McBride, M.B., 1979. Chemisorption and precipitation of Mn2+ at CaC03 surfaces. J. Soil Sci. SOC. Am., 43: 693-698. McConnaughey, T.A., 1986. Oxygen and carbon isotope disequilibria in Galapagos corals: Isotopic thermometry and calcification physiology. Ph.D. thesis, University of Washington, Seattle, WA, 341 pp. Molion, L.C.B. and Moraes, J.C., 1986. Southern Oscillation and river discharge in tropical South America. In: Chapman Conference on El Niiio (abstracts), Guayaquil, Ecuador, Oct. 27-31, 1986, p. 21. Moore, W.S. and Edmond, J.M., 1984. Radium and barium in the Amazon River system. J. Geophys. Res., 89: 2061-2065. Moore, W.S., Sarmiento, J.L and Key, R.M., 1986. Tracing the Amazon component of surface Atlantic water using 228Ra, salinity and silica. J. Geophys. Res., 91: 25742580. Morse, J.W., 1986. The surface chemistry of calcium carbonate minerals in natural waters: An overview. Mar. Chem., 20: 91-112. Muller-Karger, F.E., McClain, C.R. and Richardson, P.L., 1988. The dispersal of the Amazon's water. Nature, 333: 56-59.
282
Odum. H.T., 1951. Note on the strontium content of sea water, celestite radiolaria and strontianite snail shells. Science, 114: 21 1-213. Ohde, S., Ohta, N. and Tomura, K., 1978. Determination of trace elements in carbonates by instrumental neutron activation analysis. J. Radioanal. Chem., 42: 15 9- 167. Pak, H., and Zaneveld, J.R.V., 1974. Equatorial Front in the Eastern Pacific Ocean. J. Phys. Oceanogr., 4: 570-578. Parr, A.E., 1937. A contribution to the hydrography of the Caribbean and Cayman Seas based upon observations made by the research ship "Atlantis", 1933-1934. Bull. Bingham Oceanogr., Coll. 5: 1-110. Patzold, J., 1986. Temperature and C 0 2 changes in tropical surface waters of the Philippines during the past 120 years: Record in the stable isotopes of hermatypic corals, Berichte - Reports, Geo1.-Palaont. Inst. Univ. Kiel, Nr. 12, 82 PP. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Niiio occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14449-14461. Rao, V.B., Satyamurty, P., De Brito, J.I.B., 1986. On the 1983 drought in north-east Brazil. J. Climatol., 6: 43-51. Rasmusson, E.M. and Carpenter, T.H., 1982. Variations in tropical sea surface temperature and surface wind fields associated with the Southern-Oscillation/El Niiio. Mon. Weather Rev., 110: 354-383. Richey, J., Nobre, C. and Deser, C., submitted. Amazon River discharge and climate variability: 1903-1985. Rollins, H.B., Richardson, J.B. 111, and Sandweiss, D.H., 1986. The birth of El Niiio: Geoarcheologic evidence and implications. Geoarcheology, 1 : 3- 15. Ropelewski, C. F. and Halpert, M.S., 1987. Global and regional scale precipitation patterns associated with the El Niiio/Southern Oscillation. Mon. Weather Rev., 115: 1606-1626. St. John, B.E., 1974. Heavy metals in the skeletal carbonate of scleractinian corals. Proc. 2nd Int. Coral Reef Symp., Great Barrier Reef Comm., Brisbane, Qld., pp. 461469. Scherer, M. and Seitz, H., 1980. Rare earth element distributions in Holocene and Pleistocene corals and their redistribution during diagenesis. Chem. Geol., 28: 279-289. Shannon, R.D., 1976. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr., A32: 751767. Shaw, H.F. and Wasserburg, G.J., 1985. Sm-Nd in marine carbonates and phosphates: Implications for Nd isotopes in seawater and crustal ages. Geochim. Cosmochim. Acta, 49: 503-518. Shen, G.T. and Boyle, E.A., 1987. Lead in corals: reconstruction of historical industrial fluxes to the surface ocean. Earth Planet. Sci. Lett., 82: 289-304. Shen, G.T. and Boyle, E..A., 1988. Determination of lead, cadmium and other trace metals in annually-banded corals. Chem. Geol., 67: 47-62. Shen, G.T., Boyle, E.A. and Lea, D.W., 1987. Cadmium in corals as a tracer of historical upwelling and industrial fallout. Nature, 328: 794-796. Smith, R.M. and Martell. A.E., 1976. Critical stability constants, Vol. 4. Plenum, New York, N.Y., 257 pp. Smith, S.V., Buddemeier, R.W., Redalje, R.C. and Houck, J.E., 1979. Strontium-calcium thermometry in coral skeletons. Science, 204: 404-407.
283
Speer, J.A., 1983. Crystal chemistry and phase relations of orthorhombic carbonates. In: R.J. Reeder (Editor), Carbonates: Mineralogy and Chemistry. Min. SOC. Am., 1 1 : 145-190. Steven, D.M. and Brooks, A.L., 1972. Identification of Amazon River water at Barbados, W. Indies, by salinity and silicate measurements. Mar. Biol., 14: 345348. Street, F.A. and Grove, A.T., 1976. Environmental and climatic implications of late Quaternary lake-level fluctuations in Africa. Nature, 261: 385-390. Stumm, W. and Morgan, J.J., 1981. Aquatic Chemistry. John Wiley and Sons, New York, 780 pp. Swart, P.K., 1981. The strontium, magnesium and sodium composition of recent scleractinian coral skeletons as standards for paleoenvironmental analysis. Paleogeogr, Paleoclimatol., Paleoecol., 34: 115-136. Swart, P.K. and Hubbard J.A.E.B., 1982. Uranium in scleractinian coral skeletons. Coral Reefs, 1: 13-19. Taylor, R.C., 1973. An Atlas of Pacific Islands Rainfall. Office of Naval Research, HIG-73-9, Data report No. 25, 167 pp.. Thompson, T.G. and Chow, T.J., 1955. The strontium-calcium ratio in carbonatesecreting marine organisms. Deep-sea Res., 3 (suppl.): 20-39. Turekian, K.K., 1955. Paleoecological significance of the strontium-calcium ratio in fossils and sediments. Bull. Geol. SOC. Am., 66: 155-158. Turner, P.R., Whitfield, M. and Dickson, A.G., 1981. The equilibrium speciation of dissolved components in freshwater and seawater at 25OC and 1 atm pressure. Geochim. Cosmochim. Acta, 45: 855-88 1. Veeh, H.H. and Turekian, K.K., 1968. Cobalt, silver and uranium concentrations of reef-building corals in the Pacific Ocean. Limnol. Oceanogr., 13: 304-308. Weber, J.N., White, E.W. and Weber, P.H., 1975. Correlation of density banding in reef coral skeletons with environmental parameters: the basis for interpretation of chronological records preserved in the coralla of corals. Paleobiol., 1: 137-149. Wellington, G.M. and Glynn, P.W., 1983. Environmental influences on skeletal banding in eastern Pacific (Panama) corals. Coral Reefs, 1: 215-222. Wells, L.E., 1987. An alluvial record of El Nifio events from northern coastal Peru. J. Geophys. Res., 92: 14463-14470. Wyrtki, K., 1975. El Nifio - The dynamic response of the equatorial Pacific Ocean to atmospheric forcing. J. Phys. Oceanogr., 5: 572-584.
This Page Intentionally Left Blank
285
HISTORICAL ASPECTS OF EL NIRO/SOUTHERN OSCILLATION - INFORMATION FROM TREE RINGS
J.M. LOUGH Australian Institute of Marine Science, PMB No.3, Townsville M.C., Queensland, Australia. H .C.FRITTS Laboratory of Tree-Ring Research, University of Arizona, Tucson, Arizona 85721, U.S.A.
ABSTRACT J.M. Lough and H.C. Fritts. 1989. Historical aspects of El NiiiolSouthern Oscillation information from tree rings. Characteristics of the annual growth rings of trees can provide accurately dated proxy climatic information for periods prior to the start of instrumental climate records. To derive information about El NiiiolSouthem Oscillation (ENSO) events from tree rings depends upon the trees being located at sites where climate is significantly and consistently influenced by ENSO events and also the trees' ability to faithfully record this influence. Such a tree ring/ENSO linkage is illustrated with tree-ring data from sites in western North America that have been used to develop large-scale climate reconstructions back to 1602. The instrumental surface temperature and precipitation fields in pal-is of North America show significant teleconnections with ENSO events and these linkages are demonslrated to be preserved in dendroclimatic reconstructions of the same climate variables. A significantly calibrated and verified reconstruction of a Southein Oscillation (SO) index has been developed back to 1601 from the westein North American tree-ring data base. The practical significance of this reconstruction is assessed by comparison with an independent record of historical El Niiio events. Dendroclimatic techniques are limited in their application to tree species from temperate and sub-polar latitudes clue to problems in dating species from tropical regions. Thus. no tree-ring data are presently available from the land areas in the tropics where climate is most strongly influenced by ENSO events. Prospects for deriving more information about past ENSO events from tree-ring data in temperate regions outside of North America are discussed. Ultimately. the most reliable history of climatic phenomena associated with ENSO events for periods prior to instrumental records will be obtained by combining a variety of independent sources of proxy climate infoimation. I INTRODUCTION
The Southem Oscillation is now recognized to be a major source of interannual global climate variability. The strong signal acting over large areas with inherent lag relationships gives this phenomenon great impoitance to the prediction of short-term climate varialions (e.g. Gill and Rasmusson. 1983: k i t h . 1984: Rasmusson and Arkin. 1985: Wright. 1985). Although the most pronounced oceanic and atmospheric anomalies of the low SO index phase (ENSO events) occur in the equatorial Pacific region (Rasniusson and Carpenter. 1982). associated atmospheric circulation and suiface climate anomalies (known as
286 teleconnections) extend throughout tropical regions and into the extra-tropics of both hemispheres (e.g. Walker and Bliss, 1932: Bjerknes. 1966. 1969: Wright. 1977: Horel and Wallace. 1981; van Loon and Madden. 198 I: Ropelewski and Halpert. 1987a.b; among others). ENSO events typically occur at intervals of between two and ten years (Trenberth. 1976: Quinn et al.. 1978). The 1982-83 ENSO event had significant climatic. biological and economic impacts in many regions of the world (Rasmusson and Hall. 1983: Hansen, this volume). As measured by the Tahiti minus Darwin sea-level pressure difference (the most commonly used index of the SO). the 1982-83 event was the most extreme within the 12 I years. 1866 to 1986. of continuous instrumental records (Ropelewski and Jones, 1987). The well-studied 1982-83 event was unusual in both its intensity and its atypical evolution over time compared to previous events (Rasmusson and Wallace. 1983: Rasmusson and Arkin. 1985). How frequent is an ENSO event like that of 1982-83? Has the frequency of recurrence of ENSO events and/or their intensity varied significantly with time? Are ENSO events likely to be more or less frequent in the warmer world climate anticipated to occur as a result of enhanced COJtrace gas concentrations'? These questions. among others. about the historical occurrences of ENSO events. cannot be adequately addressed with the instrumental records of climate. These are only widely available from the end of the 19th century and are generally too short in length to provide a realistic perspective of the range of past and. therefore. possible future climate variations. such as those associated with the SO. To establish how the SO has varied prior to this time we must examine documentary and proxy records of past climate that are capable of resolving climate variations on at least annual time scales and at periods within the high frequency (two to ten year range) of the climatic spectrum. Detailed descriptions of climate variations within the short term range of 1 to lo3 years may also enable study of the short-term development of climatic patterns that can contribute to an understanding of the physical processes involved (Kutzbach. 1975). Documentary or historical references to unusual climatic phenomena are one type of means of extending knowledge of past climate of the past several hundred years. Quinn et al. ( I 978) have. for example. dated occurrences of El Niiio events. associated with the lowSO index phase. back to 1763. To do this they made use of early instrumental records and information contained in written references to unusual events. such as disruption of the anchovy harvest. off the coast of Pew. The work of Quinn et al. (1978) was extended hy Hamilton and Garcia ( 1986) who examined the historical sources relating to Peruvian precipitation to produce a probable chronology of strong El Niiio occurrences between 153 I and 1841. Quinn et al. (1987) recently revised their chronology of past El Niiio occurrences and also extended it to cover the past 450 years. This new chronology contains some differences from the earlier series and also that of Hamilton and Garcia (1986). Indirect or proxy records of past climate can also be developed from natural hiological
287
or geological systems that are in some way affected hy climate and that permanently incorporate into their structure some measure of this dependency. Extraction of this climatic information or signal usually requires comparisons with instrumental climate records and is based on the principle of uniformitarianism. i.e. currently observed relationships operated in the same manner in the past. Proxy indicators of past climate vary according to spatial coverage. period. dating accuracy and the environmental variable involved. Accurate dating is of fundamental impoitance to all paleoclimatic studies. Proxy records that are capable of resolving the climate variahle to at least a specific year are of most importance to developing climate histories relating to the SO. Tree rings. annually layered ice sheets. coral skeletons and valved lake sediments are all potential so~ircesof such high resolution proxy climate data. In this chapter we review how one type of terrestrial biological system, the annual growth layers of long-lived trees. can provide information about climate variations related to the SO. We first of all review the role of tree rings as climate proxies. We then present an example application of tree rings to the understanding of past ENSO events and a reconstruction of an SO index developed from tree rings. Finally. we discuss the potential of tree-ring series from various areas of the world to add to our knowledge of past ENSO variations. 2 TREE RINGS AS A SOURCE OF INFORMATION ON PAST CLIMATE Well defined growth rings are formed in the wood of many tree species throughout subpolar and temperate regions of the world. The widths. wood density. and cell characteristics of the rings vary to a greater or lesser extent from one growing season to the next. Measurements of these characteristics may be correlated with variations in macroclimatic factors if these factors are linked strongly enough to conditions that directly or indirectly influence tree growth. In such cases. the rings from old trees can be used to reconstruct seasonal to millennia-long variations in past climate. Dendroclimatology is the study and application of tree rings to reconstruct and analyze variations in past climate. Descriptions of the principles and practices of dendroclimatology can be found in Fritts (1976). Hughes et al. (1982). Brubaker and Cook (1983) antl Stockton et al. (1985). Here. we briefly review some of the important tree-ring characteristics and dentlroclimatic procedures that make tree rings a unique paleoclimatic data so~irce. Gymnospeims have most commonly been used for tlenclrocliniatic study but some angiosperm genera. particularly Quercus. have also been used extensively. The inner cell layer of the annual ring begins to form hy cell division under the bark in spring. I n gymnosperm species the first cells that are produced are largc antl thin-walled. forming low density wood, referred to as earlywood. As the season progresses through mid antl late summer, the newly-formed cells do not enlarge as much and the cell walls beconic thicker. forming wood of higher density. the latewood. In temperate forest gymnosperms. the wood density variations are relatively gradual within each ring. but there is usually an ahrupt density change marking the boundaries between the annual rings.
Growth is more variable in most subtropical and tropical trees. There may be several growing periods within a year: ring boundaries may not be apparent: or there is no way to determine which boundaries represent the annual growth increments. Thus few species of tropical trees have been used in dendroclimatic work. While most trees growing in temperate and sub-polar regions foim well defined rings. few of them are of suitable age or are sufficiently affected by climatic variations to be useful in dendroclimatic work (FI-itts. 1976: Brubaker and Cook. 1983: Stockton et al.. 1985)
2. I Site Selection and Sample Collection To obtain a tree-ring record that faithfully reflects past climatic variations. dendroclimatologists search for the following: i) Trees must have sound stems and be old enough to span the inteival over which climate is to be reconstructed. ii) Trees must produce visible annual layers during most years and must grow on sites where annual variations in climatic conditions have been growth limiting. The rings from trees on arid sites are likely to reflect variations in precipitation while those from trees at high altitudes and latitudes are likely to reflect vaiiations in temperature. Trees respond to the total environment. Thus they may contain identifiable responses to sunlight. wind. evaporation, drifting snow and carbon dioxide as well as temperature and precipitation (LaMarche et al.. 1984: Graumlich and Brubaker. 1986). Climatic variables such as atmospheric pressure may also be correlated with tree-ring characteristics because the larger-scale atmospheric circulation determines the precipitation. sunlight. temperature and winds that most directly influence tree growth. Sites are selected so that a particular growth limiting factor predominates. e.g. water on arid sites or temperatui-e near the upper treeline. iii) One or more ring features must vary in a perceptible manner from one year to the next and the same pattern of ring variation must be visible in large numbers of trees from a given area. This similarity of ring character variation is evidence that large-scale climatic variations have been limiting to ring growth and is used in the procedure of "crossdating". Crossdating is important to distinguish between the features resulting from climatic variation and those resulting from non-climatic influences such as fire. forest disturbance and pollution. Trees influenced by the latter are generally avoided because the presence of non-climatic influences dilutes the common patterns corresponding to the "signal" of climatic infoimation. iv) There must be adequate numbers of acceptable trees to provide statistical replication. Generally 10 to 30 sample trees are selected from habitats that are as similar as possible. When the climatic "signal" appears weak. it may be necessary to collect a larger number of samples. At the Laboratory of Tree-Ring Research two cores are usually extracted from each tree. the cores are examined for evidence of non-climatic influences. and then processed in the laboratory for crossdating analysis and measurement,
289 2.2 Crossdating and Measuring Crossdating is a procedure that uses the similaritics of ring characteristics (usually associated with climate) through time to identify the exact year of ring formation (Stokes and Smiley. 1968). Thus the date in which each ring was foimed is identified by crossdating rather than simply counting the rings from the outside to the stem center. Crossdating is tedious and time-consuming. but it is basic to dendroclimatic analysis. Sometimes a ring may be absent from a particular sample because growing conditions were highly limiting, but it is present as a narrow ring in other samples. In other circumstances a tree may produce more than one ring in a given year. There are other circumstances that may lead to improper ring age determination in particular specimens. In all of these situations the dates using simple ring counting would be incorrect for all rings inside the first encountered problem growth layer. Such dating problems may he encountered several times in a sample. Without tree-ring dating. any measurements of the ring characteristics would be hopelessly scrambled when the data are averaged among the samples to obtain a chronological sequence of variations in yearly growth. Computer programs can be used to assist the dating procedure (Hughes et al.. 1982: Holmes. 1983). but the application of these programs usually requires a basic understanding of the crossdating operation and some technical training. including hands-on experience with known materials. I t is relatively easy to optically measure ring width. X-ray technology can be used to measure wood density variations within the rings (e.g. Schweingruber. 1982). These data can be converted to measurements of maximum density. minimum density. earlywood width and latewood width that may reflect different climatic information from that contained in ring widths. Computer systems are now used to facilitate all types of measurements (e.g. Lenz et al.. 1976: Robinson and Evans. 1980). Most of the present day tree-ring chronologies are based upon variations in ring width. While clensitometric measurements can provide more climatic information than those from ring widths. the technology is costly and time-consuming. I t is only comparatively recently that dendroclimatic reconstructions have been developed that incorporate wood density variations (e.g. Hughes et al.. 1984: Conkey. 1986: Briffaet al.. 1988). 2.3 Chronology Development The time series of many tree ring characteristics may have substantial trends and nonstationary variance that are associated with the changing geometry of the ring. increasing tree age. and slowly changing features of the tree's immediate environment. An exponential or polynomial curve. cubic spline. or some other growth function is fitted by least-squares techniques. and the ring measurements for each year are divided by the values of the fitted curve for that year to obtain a standardized growth index (Fritts. 1976: Graybill. 1982). The indices from all dated cores are then averaged by year to obtain a standardized chronology for each site. Both the mean and variance of the standardized indices for most ring-width time series are relatively stationary over time. normally distributed. and the
290 autocorrelation is usually higher than those for annually averaged instrumental climatic measurements. Since differences associated with the growth function itself can contribute to a large percentage of the ring-width variance. the removal of this function by stantlardization enhances the climatic "signal" remaining in the indexed measurements. In addition. the differences in measurements between trees that represent non-climatic "noise" are niore-orless random, Their variance is reduced by the averaging processes. which further enhances the ratio of the "signal" to "noise". A variety of chronology statistics and analyses can be used to evaluate the climatic signal content (Fritts. 1976). 2.4 Climatic Reconstruction To reconstruct climate in an objective manner. i t is necessary to obtain a transfer function. A transfer function is some sort of statistical model that translates the related climatic infoimation in the tree-ling chronologies (predictors) to estimates of the climatic (predictand) data set. Time series of both climatic and tree-ring data that overlap through both time and space are assembled. The period of overlap is usually clividetl into two intervals. one for calibration and the other for verification tests. i ) Calibration : Both the climatic and tree-ring data may be averaged antl calibrated using linear regression techniques. However. spatial or seasonal relationships can he considered and the climatic information enhanced for large data sets by entering the station data and tree-ring chronologies into separate matrices and using multivariate methods to calibrate the two sets (Fiitts et al.. 1971. 1979: Fritts. in press). In addition. principal component analysis may be used to convert the data to orthogonal variables and to reduce the size of the original data sets. Thus a reasonable strategy is to enter the individual standardized chronologies representing different localities. species. and site types into the columns of the predictor data set with the rows corresponding to the averaged growth index (Fritts antl Shatz. 1975: Cropper and Fritts. 1981). These data may be A R M A modeled if significant persistence is present (Rose. 1983: Monserud. 1986). Seasonal or annual climatic data for different stations are entered in the columns of the second matrix. If there are more columns than rows. the numbers can be reduced by extracting the most important principal components from each data set. The principal components of climate are regressed on the principal components of tree growth. This can be accomplished for each climatic element of the grid using principal component regression (Briffa et al.. 1988) or for all climatic principal components using canonical regression techniques (Glahn. 1968: Blasing. 1978: Fritts. in press). ii) Verification : It is always possible that a transfer function may have statistically significant calibration statistics but fail when it is applied to intlepentleni climatic data (data from the verification interval that were not used for calihration). While the reliahility of regression estimates almost always cleclines for such claia. i t is iinportant to assess 17s
291
statistical means the magnitude of the decline and to Lest whether the independent reconstructions are unlikely to have arisen by chance (Gordon. 1982). Atlditional verification of the statistically significant reconstructions can be obtained by comparing them to historical data and other proxy climatic records with a similar climatic response. 3 TREE RINGS AND THE SOUTHERN OSCILLATION : A N EXAMPLE
APPLICATION The tropical antl sub-tropical land masses represent a large gap in the global dendroclimatological data base because of the difficulties (discussed in the preceding section and see also various papers in Bormann and Berlyn. I 9 8 I ) of applying dendrochronological techniques to tropical species. The application of tree-ring chronologies to the study of past ENSO events. therefore. depends upon the anomalous climate distributions (teleconnections) that occur in the extra-tropics as a result of tropical forcing. The nature and piior history of the atmospheric and oceanic circulation at the time of tropical forcing can influence the magnitude and characteristics of the subsequent extratropical teleconnection pattern (Rasmusson and Wallace. 1983: Hamilton. 1988). Considerable variability has been observed in the extra-tropical responses of individual years when compared to the "average" signal (e.g. Emery and Hamilton, 1985). In addition. a particular climate teleconnection pattern may not be a unique product of tropical forcing by ENSO events. Numerical modelling (Lau. 198 I : Simmons et al.. 1983) and empirical studies (Douglas et al.. 1982: Namias and Cayan. 1984) have demonstrated the occurrence of some extra-tropical teleconnections in the ahsence of non-seasonal tropical forcing. There will. therefore. he an upper limit to the inferences that can he made about the tropical climatic vaiiations associated with ENSO events using data from extratropical latitudes. whether instrumental or proxy. The latter data also contain error terms unrelated to climate (Fritts. 1976). With these words of caution we describe some of the results of a study relating the SO to tree-ring chronologies from North America (Lough and Fritts. 1985). First. we demonstrate that teleconnection patterns associated with the SO can be identified in spatially detailed dentlroclimatic reconstnictions and second. that an index of the SO can be reconstructed from tree-ring chronologies.
3. I Teleconnection Patterns in Dendroclimatic Reconstructions The extra-tropical teleconnections of the SO appear to be especially well developed during Northern Hemisphere winter (December to Fehi-uary) over North America and the Noith Pacific (Wright. 1977: Douglas and Englehart. I98 I : Horel and Wallace. I98 I : Trenberth and Paolino. I98 I : van Loon and Madden. I 9 8 I : van Loon and Rogers. I98 I : Chen. 1982: Rasmusson and Wallace. 1983: Quiroz. 1983: Ropelewski and Halperl. 1986. 1987a.b). Above average temperatures in the Pacific NorthwestiAIaslta region antl helow average temperatures and increased precipitation in the south/soulheast United States are
292
found in the winter following an ENS0 event. These surface climate anomalies are associated with a deepened and southeastward displaced Aleutian Low and appear to he reasonably consistent and strong responses in winter to the tropical forcing. Spatially detailed dendroclimatic reconstructions have heen developed back to I602 for these same regions and climate variables. These spatial reconstructions of climate and the reconstruction of an SO index described in the following section were developed from a network of 65 arid-site tree-ring chronologies from western North America (Fig. 1 ). These chronologies were selected on the basis of greatest number of trees sampled. the statistical characteristics of the data. longest records and spatial distribution of the sites (Fritts and Shatz. 1975). All data span the period from 1600 to 1963 but they are best replicated after 1700. Tree growth at the 65 sites is related to precipitation as it affects water storage and temperature as it affects evapotranspiration (Fritts. 1974. 1976). The patterns of temperature and precipitation variations are in turn related to the overlying atmospheric circulation. Features of the latter tend to move from west to east. from the North Pacific to the eastern United States. over the tree-ring sites (Bryson and Hare. 1974). This array of tree-ring chronologies can. therefore. be used to examine climate variations in regions well beyond the area of the tree-ring sites (see also, Kutzbach and Guetter. 1980). Sea-level pressure was reconstructed at 96 grid points between IOOOE and 80OW. 20° and 70°N. and surface temperature at 77 stations and precipitation at 96 stations in the United States and Southwestern Canada. The climate reconstructions were developed using the methods outlined by Fritts et al. (1979) and Fritts (in press). The predictors were the major principal components of the tree-ring chronologies and the predictands the major principal components of the respective climate variables. Seasonal values of the climate variables were calibrated by means of canonical regression with the tree growth over the periods 1901 to 1963 for temperature and precipitation and 1899 to I963 for sea-level pressure. A transfer function was obtained and then applied to the tree growth data to estimate the particular seasonal climate variable back to 1602. The statistical nature of the transfer function model requires that its validity be verified over data independent of that used to calibrate the model. The temperature and precipitation models were verified with independent instrumental data prior to 1901. The general features of the final sea-level pressure niodels were verified using sub-sample replication within the calibration period as there were insufficient data available prior to I899 for statistical analysis. Full details of the verification techniques are given by Gordon ( 1982). Several models with significant calibration and verification statistics were developed for each season and variable. The reconstnictions from the best (based on ranking of the calibration and verification statistics) models were then averaged and these averaged estimates were found to generally have better statistics of calihration and verification than the individual component models (see Fritts and Lough. 1985: Lough and Fritts. 1985). Verified seasonal estimates of temperature. precipitation and sea-level pressure thus were
293
120"
100"
50"
-50"
40"
-40"
-30"
30
500
MILES
20 "
Fig. I . Locations of the 65 arid-site tree-ring chronologies froni western North America.
294
developed for each station or grid point and year from I602 to I96 I . Analyses of the characteristics of these reconstructions suggest that the large-scale regional patterns of climate are calibrated at the expense of precision at individual stations of grid points. Despite these drawbacks. the reconstructed data have been demonstrated to contain information in common with other independent sources of proxy climate data and to be applicable to the study of particular climatic phenomena (Fritts and Lough. 1985: Gordon et al.. 1985: Lough and Fritts. 1985. 1987: Lough et al.. 1987: Fritts. in press). TABLE 1
Years of extreme low antl high SO index. within the dependent ( I90 I - I96 1 ) and independent (I852 - 1900) time periods. used to calculate the average difference niapz. 1901 - 1961 High
LOW
I903 I905 I906 1912 1914 1915 1919 1926 1941 I958
1917 1918 1921 I934 I938 I939 I945 I947 19.50 I956
18.52 - I900 Low High I856 186.5 I867 I868 I869 I874 I877 I878 1881 1889
1852
I862 I863 I870 I873 I879 I880 I890 I893 I899
The instrumental records of sea-level pressure. temperature and precipitation (the data used to calibrate the reconstruction models) were first tested for their ability to I-eprotluce the SO teleconnection patterns described in the literature. Extreme low index values represent E N S 0 events and the lowiwet years as defined by van Loon and Madden ( I98 I ). Conversely extreme high index values represent the high/dry years of van Loon antl Madden ( I98 I ) and what have been termed La Niiia events by Philander ( 1985). Winter data were averaged for the ten most extreme low-index and the ten most extreme higliindex years and the differences between these two extremes (average for low-index minus average for high-index years) were calculated. tested for statistical significance (using the ttest for difference between two means (Mitchell et al.. 1966))and mapped. This analysis was then repeated using reconstructed data for the same set of years. These two sets of maps would be expected to be very similar as the calihration procedure forces the estimates to be most like the instrumental record of the 20th century. Finally. for the real test of the reconstructions. the analysis W B Sperformed using the ten most extreme low-index and the ten most extreme high-index years and the reconstructed data prior to the 20th century when the data are independent of that used for calihration. The extreme SO years (see Table I ) were identified using Wright's (1975) index of the SO. The averaging antl
295 differencing procedures should help to maximize the signal associated with the SO and minimize the noise terms. which in the dendrocliniatic estimates can include a non-climatic component (Fritts. 1976). The collective or field significance of the maps of the differences between low- and high-index years (Figs. 2 anti 3) was determined following the procedures presented hy Livezey and Chen ( I 983). The critical thresholds for significance of the maps (representing the higher percentage value determined from the binomial distribution antl Monte Carlo simulations) were as follows for the instrumental and reconstructed data (in parentheses) : 17.5% (10. I % ) for sea-level pressure: 15.5% (9.4%)for temperature and 27.5% (10.1 % ) for precipitation. Correlations between the map patterns of the differences were also computed. The effects of spatial correlation on the degrees of freedom and hence the statistical significance of the correlation coefficients was approximated from the number of principal components required to explain at least 90 percent of the variance in each data set. The critical 5 percent significance levels for the correlation coefficients between maps were detetmined to be 0.58 for sea-level pressure. 0.67 for temperature and 0.8 I for precipitation. I t was considered that this technique would provide a reasonable estimate of the true values of spatial correlation for the different fields. The difference maps (low- minus high-index averages) are presented in Fig. 2 for sealevel pressure and Fig. 3 for temperature antl precipitation. The instrumental records (top map in each figure) highlight the patterns of suiface climate anomalies identified in previous studies to follow ENS0 events. Forty three percent. 34 percent and 25 percent of the differences were significant at the 5 percent significance level for sea-level pressure. temperature and precipitation. respectively. The sea-level pressure and temperature maps thus have collective significance and that for precipitation falls just below the threshold val Lie. The reconstructed data averaged for the same sets of years (midclle set of maps) showed. as expected. many similarities with thc patterns identified in the instrumental clata. For each variable the differences wcre smaller in magnitude and fewer are statistically significant than found in the instrumental record. A certain amount of decrease is expected partly because only a portion of the insti-umental cliniale variance was reconstructed. Seven percent (sea-level pressure). 8 percent (temperature) and 14 percent (precipitation) of the differences were significant. so only the precipitation map has collective significance. The general similarity between the instrumental and reconstructed maps was supported. however, hy pattern correlations of 0.88 (sea-level pressure). 0.93 (temperature) antl 0.79 (precipitation). The correlations for sea-level pressure antl temperature are signilicant at the 5 percent significance level when allowing for the effects of spatial correlation on the degrees of freedom. The main difference between the sea-level pressure patterns from the instrumental and reconstructed data was found over Asia. This is the region fuithest removed from the treering predictor sites where the reconstructions are known to he least reliable. The
296
WINTER PRESSURE (LOW lOOE
120
140
lOOE
120
140
160
180
"MENTAL 160
180
- HIGH S.O. INDEX)
160
140
120
100
8OW
120
100
BOW
MTA 1901-1981 160
140
70 N
0 60
50
40
30
20
RECONSTRUCTION 1952 - 1900
Fig. 2. Winter sea-level pressure (mb) difference between low and high SO index year averages for instrumental data I90 I - I96 I (top), reconstructed data I90 1 - I96 I (middle) and reconstructed data 1852-1900 (bottom). Black dots represent grid points where the difference is significant at the 5 percent level.
297
WINTER TEMPERATURE (LOW
- HIGH S O
WINTER PRECIPITATION (LOW
INDEX)
-
HIGH S.O. INDEX)
INSTRUMENTAL DATA 1901-1961
INSTRUMENTAL DATA 1901.1961
0
RECONSTRWT~ON 1901-1961
RECONSTRVCTION 1852-1900
RECONSTRUCTION 1901-1961
RECONSTRUCTION 1852 -1900
Fi . 3 . Winter temperature (degrees Centigrade) and winter precipitation (percent) dikerences between low and high SO index year averages for instrumental data 1901- I96 I (top), reconstructed data 1901-1961 (middle) and reconstructed data 1852- 1900 (bottom). Black dots represent stations where the difference is significant at the 5 percent level.
298 precipitation maps showed the greatest difference hetween the instrumental and reconstructed data over the eastern United States where the enhanced precipitation i n Florida following ENSO events is not well estimated. Again this is a region where the reconstructions are known to be weak and where Atlantic influences on the climate patterns are not well reproduced from trees at sites in western North America. The patterns for the independent period (lower set of maps) showed general similarities with the maps from the calibration period though the magnitude and significance of the differences diminished further. This was particularly apparent for sea-level pressure with no significant differences and only the deepening of the Aleutian Low weakly defined. The correlation of this pattern of differences with that of the instrumental data of 0.62 was. however. significant. Five percent of the temperature stations showed significant differences in the independent period (not collectively significant) though the pattern was significantly correlated (r=0.87) with the instrumental pattern. Twelve percent of the precipitation stations show significant differences (collectively significant) and the map has a correlation of 0.77 (below the 0.8 I threshold for significance) with the instrumental map. These comparisons suggested that of the three reconstructed variables. winter temperatures most clearly retained the features of the teleconnection pattern identified in the instrumental record. The winter precipitation reconstructions contained the principle features of the instrumental pattern in the western United States but were unable to reproduce features to the east, the region furthest removed from the tree sites. The scalevel pressure reconstructions only identified a very weak deepening of the Aleutian Low in the independent period. Analysis of the spectral characteristics of the reconstructions has shown that those for temperature and precipitation contain a higher proportion of the high frequency climatic infomiation in the instrumental recorcls than the sea-level pressure estimates. Such a bias of the sea-level pressure reconstructions towards lower frequencies would reduce the amount of variance that could he associated with the mid- to highfrequency variations of the SO. These results suggest that variations in North American climate associated with the SO make some impact on tree-ring width variations. This informalion is modelled with some degree of success in the transfer function used to develop the I-econstructions of climate. This is particularly true for temperature and precipitation in the westein United States. In addition. the teleconnections from the SO to extra-tropical latitudes appear to have had a similar influence on North American climate in the past as found for the 20th century. 3.2 Reconstruction of an Index of the Southein Oscillation As the reconstructions from tree rings appeared to contain some information associated with the SO. it seemed reasonable to test whether past variations of the SO could he estimated directly from the tree-ring chronologies. The hasis for this I-econstniction is that significant teleconnections from tropical to extra-tropical regions. asociated with ENSO.
299
influence the suiface climate conditions that help to regulate tree growth. Climate information. whether instrumental or proxy. can only he expected to estimate a part of the variance of the SO. Bainett (1981: see also Wright. 1985) found. for example. that less than half of the variance in North American suiface temperatures could he explained hy tropical forcing. The ability to reliably estimate variations of the SO will also be limited by the strength of the climate signal in the tree-ring chronologies. the ability ofthe linear regression model to translate this climate signal and the presence of non-SO factors affecting climate in the vicinity of the trees. Lough and Fritts (1985) developed a reconstruction of the SO index of Wright ( 1975). This index. covering the period 185 I to 1974. was the longest ancl most homogeneous index available at that time. The estimates calibrated about SO percent of the SO variance with about 20 percent of the variance verified over an independent period. We have recently developed a new reconstruction using the same predictor data set and techniques as Lough and Fritts ( 1985) but estimating the recently extended Tahiti-Daiwin sea-level pressure SO index (Ropelewski and Jones. 1987). This index. starting in 1866. has two advantages. First. it represents the most commonly used measure of the strength of the Southern Oscillation. Second. the index is based on the same two stations throughout the length of the record. Wright's ( 1975) index used different conihinations of stations through time. which is likely to cause inhomogeneities. The reconstruction hased on the TahitiDaiwin index. presented here, is. therefore. considered to he the more reliahle and more relevant to studies of past E N S 0 events. though the calibration and verification statistics are similar in the two reconstructions. The 65 tree-ring chronologies (Fig. I ) were first prewhitenetl using all significant autoregressive-moving-average terms (Rose. 1983: see also Monserucl. 1986). The first fifteen eigenvectors of the correlation matrix. which explain 68 percent of the total variance. antl their principal components formed the predictor data set . The pretlictantls were the twelve seasonal values of the SO index from December-Febriiary one year prior to the year of tree growth (DJF,~,) to September-November in the year following tree growth (SON,,,). (This is slightly different from Lough and Fritts (1985) who used SON, to JJA,, . ) This combination of predictands would allow lag cffects in the relationship hetween the tree rings ancl the SO to be included (see Fritts antl Gordon. 1982). The multivariate transfer function was based on stepwise canonical regression analysis (Blasing. 1978: Fritts et al.. 1971. 1979) and the model peiformance was assessed o w r both the calibration and verification periods using live statistical tests discussed by Gordon (1982) and listed at the end of Table 2. The reduction or error statistic (RE) was also used
,
to assess the model's peiforniance over the verification period (Lorenz. 1956: Frilts. 1976: Gordon. 1982). This statistic compares the estimated values over the intlepentlent period to those values based solely on the mean of the calibration perintl. Although i t cannot he formally tested for significance. a positive value clemonstratcs that some useful information is contained in the regression estimates (Gordon antl LeD~ic.1981) . The transfer lunction was tirst calibrated over the pcriotl 1916 to 1963 antl the resulting estimates veril'iccl over
300 the independent period. I867 to 1915. The calibration antl verification periods were then reversed to test the stability of the regression model. TABLE 2 Summary of calibration (CAL) and verification (VER) statistics for the SO estimates derived from the prewhitened western North American tree-ring predictor set for the 1916-1963 and 1867-1915 calibration periods. I916 - 1963 CAL -
VER -
Predictand variables
R’
Np
R?
DJFt - I MAMt - 1 JJAt - 1 SON1 - I DJFt MAMt JJAt SONt DJFt 1 MAMt + 1 JJAt 1 SONt + I
0.41 0.24 0.52 0.44 0.57 0.25 0.29 0.36 0.37 0.36 0.15 0.20
4 4 5 5 5 4 4 5
0.04 0.04 0.04 0.10 0.33 0.30 0.08 0.00 0.01 0.09 0.08 0.02
+ +
Note:
5 5
3 2
1867 - 1915
RE
CAL
VER -
~
Np
-0.74 2 -0.43 0 -0.23 I 0.052 032 5 5 -m 2 -0.72 0 -0.55 0 -0.10 3 0.05 3 2
-m
R’
Np
0.47 0.27 0.36 0.18 0.59 0.45 0.34 0.22 0.24 0.15 0.34 0.20
5 5 5
R?
RE
0.00 0.02 0.1 I 0.14 0.29 0.16 0.0 I 0.00 0.00 0.05 0.00 0.00
4 5 5 5 5 4 4 5 4
Np
-0.52 I -0.60 0
0.10 5 0 x 4 0x74 -m3 -0.36 0 -0.30 0 -0.29 0 -0.34 0 -0.59 0 -0.42 0
R2 is conelation coefficient squared: RE is reduction of error statistic with positive values underlined: and N p is the numher of statistical tests passed at the 95 percent confidence level out of a total of 5 . The tests are the correlation coefficient. the correlation coefficient of the lint differences. the sign test. the sign test of the first differences antl the cross-product mean5 tesl.
As found by Lough and Fritts ( I 985) most variance (as measured hy R?) is explained. in both the earlier and later calibrations. for the four seasons : JJA, . SON, . DJF, antl MAM, (see Table 2). Between 18 antl 59 percent of the variance in the SO was calilmted with between 4 and 33 percent of the variance accounted for over the verification periods.
,
,
The majority of the statistical tests were passed antl. with two exceptions. the R E statistics were positive. The negative RE statistics for J J A , ~(1916 , to I963 calihration) antl M A M , (1867 to 1915 calibration) may be clue to a small number of erroneously large eslimates. to which the RE statistic can be sensitive (see Lough antl Fritts. 1985). Direct comparisons between the old and new reconstructions are not possible hecause of the different SO indices used and also the different calihralion antl verification periods (the earlier reconstruction was based on the 1908 to 1963 calibration period and verilid from 1853 to 1907). Comparison of the two instrumental indices of the SO over the conininn period. 1866 to 1974. gives correlations (all significant at the 5 percent level) of0.72 (DJF). 0.55 (MAM). 0.66 (JJA) antl 0.73 (SON). The homogeneity o f Wright‘s index i s probably most suspect in the earlier part of the series. In the old reconstructions. [he
301 residuals of the estimates tended to show greater variability in the 19th than the 20th centuries. The correlations between the old and new reconstructions over the period I60 I to 1962 (all significant at the 5 percent level) are 0.80 (DJF). 0.77(MAM). 0.80 (JJA) and 0.70 (SON). The agreement between the old and new estimates is lowest. for all four seasons. in the 30-year period. 163 I to 1660 when the correlations drop to 0.5 to 0.6. The average explained variance is similar in the two reconstructions. For all twelve variables over the two calibration periods. the explained variance averages 30 percent for the old and 34 percent for the new series. For the four "best" seasons (JJA, to MAM,). the explained variance averages 48 percent for the old and 43 percent for the new series. Because of the limitations of this particular reconstructions technique (outlined earlier) this is probably as good as can be expected. The new reconstruction is. however. likely to be more homogeneous through time and may have slightly more practical value as it is based on the more commonly used index of SO strength. Of the four seasons (Table 2). the reconstruction for DJF appears to be most reliable with 57 to 59 percent variance calihrated and 29 to 33 percent variance verified. Positive RE statistics (0.32 and 0.30) in both calibrations also indicate stability of this seasonal reconstruction. The other three seasons show more variability in the calibration and verification statistics of the different periods. Significant coherence between the instrumental and reconstructed seasonal values (for the period I867 to 1962 with 25 lags. at the 5 percent significance level) is found in DJF at 5.6 years and MAM at 3.1 years. The contributions of the individual chronologies to the transfer function models for the two time periods and the four best seasonal estimates are shown in Fig. 4. The regression coefficients (beta weights) of the principal components of the predictors were multiplied by the respective eigenvectors to derive the beta weights in terms of the original tree-ring chronologies (see equation (15) in Blasing. 1978). Briffa et al. ( 1986) have used a similar procedure, though using principal component multiple regression. in their work reconstructing sea-level pressure patterns over the British Isles. General similarities in the patterns of the beta weights are apparent both between seasons and between the two calibration periods. The correlations between the 65 beta weights from the two periods are 0.65 for JJA,.,, 0.68 for SON, 0.41 for DJF, and 0.27 for MAM, (all significant at the 5 percent level, though no allowance has been made for the effects of spatial correlation o n the degrees of freedom). The major. consistent features of the maps are a large negative centre, which extends into Mexico. in the southern portion of the grid. a negative centre of lesser magnitude in the Pacific Northwest (though this feature is not evident in JJA, ant1 SON,., in the 1867 to 1915 calibration) and a positive centre in the central area of the grid. Little contribution to the final regression estimates is madc by Iree-ring chronologies from California. As low values of the SO index are associated with E N S 0 events. these maps can be interpreted as follows : increased tree growth (i.e. wider annual rings) in the southern part
,
,.
,
302
w
0
N
Fig. 4. Regression coefficients for the 05 free-ring chronologies for the SO index reconstruction models based on the I9 16- I963 and 1867- I9 IS calibration periods.
303 of the grid is the pi-imary factor contrihuting to estimates of low SO index values (i.e. ENSO events). Increased tree growth in the Pacific Northwest and reduced tree growth in the central region make lesser contributions to the final regression model. The tree-ring chronologies in the south lie within the region of enhanced winter precipitation following ENSO events (see Fig. 3. top right). Tree growth at these sites is enhanced by increased winter precipitation (see Fritts. 1976). so a direct link can he established : ENSO forcing of the extra-tropical atmospheric circulation deepens and displaces the Aleutian Low causing the winter stoims to cross North America at a more southerly latitude than usual. This leads to above average winter precipitation in the south which encourages tree growth and a wider ring than normal is foimed. This relationship appears to be the primary factor influencing the SO-tree ring calibrations over both time periods. Reduced winter precipitation may contribute to narrower growth rings in the central region and the wider growth rings in the Pacific Northwest may he a response to increased winter temperatures. The time series of the four best seasonal estimates of the SO based on the 1916 to 1963 calibration is presented in Fig. 5 . Significant spectral peaks are found in each of the four seasons at about 10. 5 and 3 years (over the period 1601 to 1962 with 90 lags). Over three 120-year subperiods ( I60 I to 1720. 172 I to I840 and I84 I to 1960). all four seasons showed significant peaks at about 10 years in the earlier two periods and at the higher frequency of 3 to 4 years in the I84 I to I960 period. This suggests that ENSO events may have been more frequent in the past 120 years though see also Table 4. TABLE 3 Years of most extreme low and high SO index values reconstructed from the prewhitened western North American tree-ring predictor set dui-ing the period I60 I to 1962. Rank
LOW INDEX
HIGH INDEX
I 2 3 4 5 6 7 8 9
1855- I 856 1815-1816 1940-I94 I I83 1 - 1 832 1854- 1855 1914- 19 15 1792-1793 1680-I68 I 172 1 - I722 1876-I877
1924-I925 1747- I748 1622- I623 1789-I790 1846- I847 1695-I696 1632-I633 1715-1716 1920-I92 I 1675- I676
10
The most extreme reconstructed low-index and high-index years between I601 and 1962 are preseited in Table 3. These were identified as years in which all tour seasons had negative (or positive for high-index) values with at least two of these seasons ha\ing estimated values at least I standard deviation from the long-term mean. During the 20th century. the low index event of 1940 10 1941 is estimated to he the most extreme.
304
N
0
In 0
?
0
z
Lo 0
m
0
m
r? I >
305 Rasmusson and Wallace (1983) have suggested that of recent ENSO events that of 1940 to 1941 was most like the intense event of 1982 to 1983. The latter year could not be included in our calibration period as the majority of the tree-ring chronologies ended in I963 or soon after. From the estimated series we suggest that events of comparahle magnitude may also have occurred in 1792 to 1793. I8 I5 to I8 16. I83 I to 1832. I854 to 1855. I855 to 1856. and 1914 to 1915 (the estimated seasonal index values average at least two standard deviations below the mean for each of these events). Thus. although uncommon (seven large events estimated in 362 years). the 1982 to 1983 ENSO may not be unprecedented. Four high-index (or La NiAa events) are estimated to be of comparahle magnitude to the 1924 to 1925 event. The estimates for 1622 to 1623. 1747 to 1748. 1789 to I790 and 1846 to I847 all average at least two standard deviations above the mean. TABLE 4
Estimated number of low and high SO index events for each SO-year period between 1601 and 1950.
Period
Number of Events
LOW
I60 I - I650 I65 I - 1700 I70 1 - 1750 I75 I - I800 I801 - I850 I85 I - 1900 I90 1-1950
HIGH
9 12 10
8 12 6 9
The estimates were also examined for evidence of changes in the frequency of low- and high-index events such as those identified in the 20th century (Berlage. 1957: Quinn and Neal. 1983). A count was made of the number of estimated low- and high-index events for each 50-year period from 1601 to 1950 (see Table 4). On this basis we see that eight to twelve such low-index events characterized six of the seven 50-year periods. In the second half of the 19th century. however. only six events of similar magnitude were estimated to have occurred. The number of low-index events was estimated hy the old reconstruction to be only five for the whole of the 19th century! The number of high-index events was also estimated to have been lowest in the second half 0 1 the I91h century. only four compared to between eight and twelve for the other 50-year subperiods.
306 TABLE 5 Estimated values of the SO index for twenty years of strong or very strong El NiHo events off the coast of Peru itlentifietl by Quinn et ill. (1987). R4AM. JJA antl SON iire for the year of El Niiio and DJF and M A M lor the year following.
Year of El Nino
MAM
JJA
SON
DJF
MAM
-0.59 I .oo 0.63 I .48 -0.4 I -0.78 -0.24 -0.s1 -0.64 -0.30 0.54 -0.23 -0.29 0.52 -0.2 I I .05 -0.97 -0.20 -1.39 -0.20
-0.39 0.75 0.4 I I .07 -0. I 2 -0.45
0.00 0.67
0.07 0.43
I607 1614 1728 I747 1761 1791 I803 I828 I844 I864 1871 1877 I884 1891 I899 1917 I925 I932 I940 1957
-0.25 0.28 0.68 -0.42 0.2 I -0.34 0.4 I 0. I6 -0.35 0.4 I 0.24 -0.50 -0.2 1 -0.24 0.2 I 0.71 1.19 0.02 0.10 0.40
-1.14 0.99 -0.04 I .67 -0.98 -0.93 0.02 -1.43 - I .75 0.10 -0.38 -0.8 I 0.86 -0.27 I .oo -0.72 -0.28 -2.17 0.10
-0.89 0.72 0.09 I .O? -0.84 -0.54 -0.60 -1.01 -0.94 -0. I7 -0. I I -0.47 -0.66 0.2 I -0.69 0.98 -0.62 -0.13 - I .06 -0.25
Mean S.D.
0.02 0.37
-0.23 0.97
-0.32 0.62
0.65
-0.13
-0.10 -0.28 -0.10 0.3x
-0.2 I -0.12 0.35 -0.04 0.68
-0.47 0.03 -0.82 -0.04
The SO estimates were compared to the historical recorcls of El NiHo events at the Peruvian coast. which have heen dated hack to the 16th centur)) (Quinn el al.. 1987). 'Table 5 shows the estimated SO values for the five seasons from MAh4 of the El Niiio year to M A M of the year following (the seasons for which the instrumental values of the SO index show the largest negative departures associated wilh El NiRo). The analysis is conlinetl to dates of strong or very strong El Niiios with the highest confidence rating of live (see Quinn et al.. 1987 for further details). Although the estimates averaged over the 20 evcnts are not significantly below the long-term mean. they do average negative for JJA antl SON. More than half of the estimated values are negative for JJA through M A M . The events o f 1607. I79 I . 1844. I877 antl I884 arc all estimated to have heen helow average index values in all five seasons antl in I761 . 1x28. 1899. I925 and I940 in lour out of the live seasons. There thus appears to he sonic agreement hetween the reconstructctl SO index
values and this independent measure of El Niiio occurrences. The agreement is better than that found between the old reconstruction and the El Niiio dates 0 1 Quinn el al. ( 1976) and
307 Hamilton and Garcia (1986). The El Niiio events are identified and their strength ascertained from conditions along the south American coast (Quinn et al.. 1987). Deser and Wallace (1987) suggest that the coupling of El Niiio and the SO is not as strong as previous studies imply. They note (within the period 1925-1986) that i ) the timing of El Niiio events varies with respect to episodes of above normal Darwin pressure (their measure of the SO), ii) El Niiio events occurred in the absence of marked positive pressure anomalies at Darwin and iii) major negative swings of Darwin pressure occurred without accompanying El NiAo events. The level of agreement found between our SO estimate and Quinn et al.’s (1987) El Niiio dates is. therefore. encouraging. Concluding. these analyses demonstrate that. allowing for the errors present. dendroclimatic reconstructions of North American climate contain information similar to that found in the instrumental record about the teleconnections between climate and the SO. This is important as it suggests that proxy climate data of this sort can be used in more than a simple descriptive capacity to study past occurrences of mid- to high-frequency climatic factors such as ENSO events, i.e. the short-teim evolution of climatic patterns. 4 FUTURE DIRECTIONS Here we consider what additional information the global tree-ring network may provide about the SO. Any useful relationship between tree rings and ENSO events will depend upon ENSO forcing of a consistent and strong climate signal in the vicinity of the tree-ling sites. Associations between tree rings and the SO will also be constrained by the locations of climatically sensitive trees and the season of the tree-ring response. The main climatic anomalies associated with ENSO events have been identified in numerous regional studies and on a global scale by Wright (1977). Stoeckinius (1981). Wright et al. (1985) and Ropelewski and Halpei-t (1987a.b). Here we look at the regional extent of these patterns and their overlap with tree-ring chronologies suitable for climatic reconstruction work. 4. I Northern Hemisphere Extra-tropics We were fortunate in the location of the western North American tree-ring grid. Ring width variations in this region are associated with climate variations over an area that appears to experience the most significant teleconnections of those between ENSO events and climate of the extra-tropical landmasses of the Northern Hemisphere (e.g. Wright. 1977: Horel and Wallace. I98 1 ; van Loon and Madden. I98 I : van Loon and Rogers. 1981 : Chen. 1982: Yarnal. 1985). Although detailed tlendroclimatic reconstructions have been developed for Europe (e.g. Briffa et al.. 1983. 1986). there is no clear ENSO signal in this region. A very large positive temperature anomaly was obseivetl over Eurasia during the 1982-1983 event (see Quiroz. 1983). hut this feature doe5 not appear to be consistently associated with ENSO (Wright, 1977). Similarly. the ENSO climate signal over easi Asia appears to be weak. There is some indication of warmer winters in Japan (van Loon and Madden. I98 I ). which Hamilton and Garcia ( 1986) find some evidence for in a prox) record of winter severity dating from the 16th century. Summer temperatures in northeast
308 China tend to be cooler in association with ENSO events (e.g. Wang. 1984: Zhang and Zeng. 1984). Although dendroclimatic work is developing in these regions (e.g. Ko,io. 1987) no reliable reconstructions of climate have. as yet. been produced (Stockton et al.. 1985). Over North America. however, a characteristic ENSO ”signal” has been identified with warming in the northwest. cooling in the southeast and wetter conditions in the southern states and Great Basin (Ropelewski and Halperl. 1986). The 65 western North American tree-ling chronology grid appears able to capture major aspects of this pattern as i t affects western North America (see section 3). Improvements in this reconstruction could probably be made by including additional moisture-sensitive trees from the southeastern United States and temperature-sensitive trees from the northwestern United Slates and southwestein Canada. The western United States is probably the most extensively sampled region in the world for dendrochronological purposes. both for arid sites (Fritts and Shatz. 1975) and temperature-sensitive upper treeline species (LaMarche and Stockton. 1974). Tree-ring studies of North America now extend into the mesic forest sites of the eastern United States where reliable reconstiuctions of precipitation related variables have been developed (e.g. Blasing and Duvick, 1984: Cook and Jacohy. 1979. 1983: Cook. 1987. in press). Temperature information has also been extracted for eastern North America using both ring-width and density measurements (Conkey. 1986). Such reconstructions may conlain historical information about climate anomalies associated with extreme ENSO events similar to that of 1982-1983 (Quiroz. 1983). The climate responses in these regions are not. however. a consistent feature of ENSO events. Thus the tree-ring information is unlikely to contribute to identifying past ENSO events. though the reconstructions may help to identify the characteristics of the climate response to independenlly dated ENSO events of the past (see. for example. Hamilton and Garcia. 1986). Tree-ring studies also extend into the more humid. subtropical climate of the southeast United States (see Stahle in Stockton et al.. 1985). A reliable reconstruction of droughl in Arkansas has. for example. been developed from swamp-grown baldcyprcss trees for 450 years (Stahle et al.. 1985). This area starts to overlap with the region of enhaiicecl winter rainfall associated with ENSO (see Fig. 3: Ropelewski and Halpert. 1986). The ENSO signal is. however. confined to winter. whereas the tree-ring variations responded primarily to summer moisture conditions in this region and are thus unlikely to contain information related to ENSO. Out of 2 1 strong or very strong El Niiio events since I600 (Quinn ct al,. 1987). this drought index reconstruction shows a change to drier conditions in 62 percent of the years following El Nifio. Drier conditions in the winter are indicated (see Fig. 2 . top right). The considerable potential of trees in the southern states to provide long. statistically reliable climate reconstructions is further illustrated h y the recently puhlished precipitation related reconstructions for various southern Unilcd States regions (Blasing et al.. 1988: Stahle and Cleaveland. 1988: Stahle et 81.. I98X).
309 We again meet a conflict between the season of ENSO response and that of the trees when we seek to expand the western North American tree-ring grid northwards into western Canada antl Alaska where a significant increase in winter temperatures follows ENSO events. Chronology series now exist for several locations in northern North America (see Cropper and Fritts. 1981; Jacoby in Stockton et al.. 1985) and these chronologies have been used to develop a number of reliable tree-ring reconstructions of predominantly summer temperature related variables (Garfinkel and Brubaker. 1980; Cropper. 1982: Jacoby and Cook. 1981: Jacoby et al.. 1985. 1988). Tree-ring width variations from upper treeline of the Cascade Range in Washington have been linked to summer temperature and spring snow depth and used to reconstruct annual temperatures at Longmire back lo IS90 (Graumlich and Brubaker. 1986). Such a reconstruction may contain some information about the winter temperature response associated with ENSO but examination of the reconstructed temperature change associated with 2 I El Niiio years (Quinn et al.. 1987) shows 48 percent of years warm and 52 percent of years cool in the year following El Niiio. Tree-ring chronologies from the western North American grid have also been used to reconstruct sea surface temperatures off the coast of southern California and Ba,ja California from 1671 (Douglas. 1980). Sea surface temperatures in this region tend to vary with ENSO events. in phase with the eastern equatorial Pacific warming (Wright et al.. 1985: Lough. 1986a). Improved historical information about ENSO from tree-ring chronologies in North America is. therefore. probably limited to the western region and southeast United States. 4.2 Tropical Regions The major ENSO surface climate signals occur in the tropical regions. Unfortunately. tree-ring studies of tropical species have proved difficult to develop (see section 2) clue primarily to the lack of clearly defined rings in most species and intra-annual growth bands in others which present major problems for dating (Ogden. 1978. 1981: Norton, 1988). Precipitation tends to decrease and surface temperature increase with ENSO in southern
Central and northern South America (Stoeckinius. I98 1 ; Ropelewski and Halpert. 1987a.b; Rogers. 1988). The major impact of the SO on South American climate appears to be during Southern Hemisphere summer in the tropical part of the continent (Aceituno. 1988). Jacoby (in Stockton et al.. 1985) has found no suitable tree-ring data for climatic reconstiuction purposes in this area. though Villalba et al. (1985) have recently been able to develop four chronologies from tropical species of the low-latitude forests of northem Argentina and Bolivia. about 24OS. with good free-ring characteristics. Central India has decreased summer monsoon rainfall antl warmer temperatures associated with ENSO events (Angell. I98 I : Rasmusson and Carpenter. I98 I : Stoeckinius. I98 I : Nicholls. 198.3: Shukla and Paolino. 1983: Ropelewski and Halpert. 1987a.b). No suilablc tree-ring chronologies have been developed to date for this country. Nicholson antl Eiitekhahi ( 1986) have found the ENSO signal in African rainfall to he geographically limited to the south/southeast
310 where drier antl warmer conditions are associated with ENSO (see also Rasmusson and Wallace. 1983: Ropelewski and Halpert. I9X7a.h) though Ogallo ( I98X) has presented evidence of some significant teleconnections between the SO and precipitation in near equatorial east Africa during the Northern hemisphere aiituriin a n d summer. The relationship in south and southeast Africa seems best defined in the late summer season (January to March) (Lindesay. 1988). Dunwiddie and LaMarche (1980) puhlishccl a 4 13-year chronology from the Cape Province in South Africa. which is significantly related to spring and early summer precipitation. hut this is well to the west of the area most strongly related to ENSO. Climatically sensitive tree ring width chronologies have also been developed for Morocco. These chronologies can provide infoimation about precipitation (Guiot et al.. 1982: Till. 1988) in the Rif and Atlas mountains. though there does not appear to he an ENSO response there. A large area of reduced rainfall with ENSO occurs in the land areas of the tropical west Pacific and tropical Australia. the area with prohably the greatest response in the Southern Hemisphere (van Loon and Shea. 1987). Linkages between Australian precipitation antl ENSO are well documented (Troup. 1965: Trenherth. 1976: Coughlan. 1979: Sroecltinius. I98 I : Pittock. 1975. 1984: McBride and Nicholls. 1983: Allan. 1985. 1988; Nicholls. 1985: Holland. 1986). Tropical antl suh-tropical regions of eastern antl northern Australia show relatively stable associations with the SO with u p to alwut a third of the variance in the rainfall in given locations and seasons accountable by variations in the SO. I n 1982- 1983. drought in eastern Australia reached "historic proportions" (Quiroz. 19x3). Growth of the predominant tree species in these regions. Eucalyptus and Callitris. is primarily confined to the wet season antl although the growth hands are largely annual they are difficult to crossdate and the species are not long-lived (see Ogden. 198 I ). Tropical species in Australia should not. however. he discounted totally. The tropical rain forests of eastern Australia. for example. contain long-lived species and Ogden ( I98 I ) suggests that improved understanding of the phenology and ecology of tropical trees may help identify their complex growth response and may. cventually. lead to application of denclrochronological techniques. 4.3 Southern Hemisphere Extra-lropics From the point of view of dendroclimatological studics. the most obvious limitation in the Southern Hemisphere is the paucity of land in temperate latitudes. Collections of treering data for this heniisphere were initiated much h e r than i n the Northern Hemisphere antl it is only within the past ten to fifteen years that large numbers of chronologies have heen developed. Norton ( 1988) provides a comprehensive review of the tlevelopnient of dendrochronology in the Southern Hcniisphere. He lists a total of I I9 published chronologies for the Soiithern Hemisphere. which compares with over 700 chronologies held by the Tree-Ring Laboratory. Tucson. Arizona. for the western United States alone (see Dean in Stockton et d..1985).
31 1 A program of extensive sampling o f trees from teniperate latitudes of the Southern Hemisphere was undertaken between I973 and I979 and the resulting chronologies published by LaMarche et al. (1979.a.b.c.d.e). Some 200 sites were sampled in South America. southein Africa. Australia and New Zealantl. Lough and Fritts (1985) attempted a
reconstruction of the SO using 33 of these chronologies. The calibration and verification statistics of this reconstruction indicated that it was much less reliable than that obtained from western North American chronologies (see section 3) with only 17 to 38 percent of the SO variance calibrated and less than I8 percent of the variance verified. This might he due in part to the poor replication and low correlations hetween trees of these chronologies (see Norton, 1988). In addition, as stressed by Norton. although the potential of Southern Hemisphere dendrochronology is great, considerable work needs to be (lone hefore this potential is realized. He emphasizes the current lack of knowledge of the ecology and life history of many of the species that is necessary to understand how climate conditions may influence tree growth. Such understanding is essential for site and tree selection ( a niqjor component of dendrochronology. see section 2) to ohtain the most climatically sensitive trees. Hughes et al. (1982, p.78) have suggesletl that dendrochronology in the Southern Hemisphere may be most effective in reconstructing major features of the general circulation (such as the SO) than climate at the local level. Pittoclc ( 1984). for example. found that a variable describing about I2 percent of mid-latitude annual precipitation variations in the Southern Hemisphere (as itlentifkl b y principal component analysis. I93 I to 1960) has a significant correlation (r=0.59) with an index of the SO. Pittock ( 1980a.b) has also demonstrated that precipitation. though not temperature. in parts of Argentina and Chile is associated with ENSO. Increased precipitation with ENSO is found along the westein side of the Andes (between 20" and 38"s)and in eastern Argentina (the latter region is also identified by Ropelewski and Halpert. 1987). Snow cover in mountainous regions and south of20"S in South America has also been linked to ENSO (Ceiwny et al.. 1987). Denclrochronologically useful tree-ring data are available for these countries (LaMarche et al.. 1979a.b) but development of dendi-oclimatic reconstructions has heen limited. Holmes et al.( 1979) have reconstructed stream flow for Argentina and Boninsegna and Holnies (1985) have developed the longest Soulhern Hemisphere tree-ring chronology (back to 44 I A.D.) from Argentina. 'The locations of trees analyzed L I to ~ the present do
not. however. coincide with the area of strong ENSO-related precipilation signal in southeastern South America. In New Zealantl. ENSO is associated with increased precipitalion in the South lslancl and decreased precipitation in the noilheasl of the North lslancl and north of Ihr South lslancl (Pittock and Salinger. 1982: Salinger. 1982). Low SO index is d s c ) associated wilh cooler temperatures over most of the country (Salinger. 1980a.h: Wright. 19X.S). Temperatures appeal- to be more closely related to ENSO events than precipitation
312
(Gordon, 1986). Noiton and Ogden (1987) review the current status of dendrochronology in New Zealand, which. with its wide range of habitats. is probably the potentially most productive area for tree-ring studies in the Southern Hemisphere. Associations have heen found between tree growth and both precipitation and temperature and reconstructions have been developed for river flow and precipitation (Norton. 1987) and summer temperatures (Norton. pers. comm.. 1987). The maximum correlation between New Zealand summer (DJF) temperatures (from Salinger. 1 9 8 0 ~and ) an index of the SO (Ropelewski and Jones. 1987) over the period 1866 to 1975 is only 0.30 (significant at the 5 percent level). This compares with correlations of up to -0.56 with precipitation in Florida and -0.47 with temperature in the Pacific Northwest. U.S.A. Southein Australia and Tasmania tend to have dryer conditions from May to October associated with ENSO events, though this response is weaker than that found in the eastern tropical and sub-tropical regions of Australia (Ropelewski and Halpert. I987a: Whetton. 1988). Preliminary reconstructions have been made of river flow and temperature in Tasmania (Campbell. 1982; LaMarche and Pittock. 1982). Only the reconstructions of liver flow for Tasmania are for a region where climate is linked to ENSO and. even then. the season of the reconstruction does not match that of the ENSO signal. It is thus evident that understanding of the tree ring-climate response and the size and location of the Southern Hemisphere tree-ring data base is too limited at present to make reliable inferences about past ENSO occurrences. This situation can only improve as the research effort in this hemisphere continues to expand and reliable climate reconstnictions are developed.
5 SUMMARY Characteristics of the annual growth rings of trees from certain regions of the world can be applied to develop accurately dated proxy climate series. Such clendroclimatic reconstructions are presently limited in their spatial distribution to temperate antl sub-polar areas. Problems with dating of tropical tree species have. to date. curtailed the application of dendrochronological techniques in the land areas where climate is most strongly affected by ENSO events. Information from tree rings about past ENSO events must. therefore. rely on the correlations at a distance (i.e. teleconnections) between the tropical forcing and surface climate at extra-tropical latitudes. Stable and consistent teleconnections tend to be limited in space. climate variable most affected. and season of year. For tree-ring chronologies to be useful for understanding the past history of ENSO events. thcy must be located in these same regions and respond to the same climate variable antl in the same season as that most closely linked to ENSO. Temperatures and precipitation in paits of North America during Northern Hemisphere winter show consistent teleconnections with ENSO. Spatially detailed dendroclimatic reconstructions have been developed for these variables. season and regions from tree-ring chronologies at sites in western North America. These reconstructions. particularly temperature, are demonstrated to contain similar teleconnection patterns as those ohseived
313 in the instrumental climate record. though. as expected. with reduced magnitude and significance of the climate departures. A reconstruction of past SO index values was also developed from westein North American tree-ring chronologies. This reconstruction calibrates a significant amount of the SO variance. is significantly verified over an independent time period. and shows spectral characteristics similar to those of instrumental indices of the SO. The reconstruction shows some agreement with an independently derived series of historical dates of El Nifio. though the practical significance of the reconstruction is still probably limited. Other extra-tropical regions that show significant teleconnections between surface climate and ENSO events are identified antl compared to currently available high-quality tree-ring material. The most important matching of areas is probably between ENSO and temperatures in New Zealand. Reliable reconstructions of the latter climate variahle have recently been developed. Although high-quality tree-ring series already exist or are k i n g developed for other areas of the extra-tropics. they are not generally suitahle for identifying past ENSO events. This is due either to the tree sites not lying within areas o f climate associated with ENSO andlor the tree-ring response not being related to the same season and climate variable most closely linked to ENSO. Thus. prospects for improving estimates of SO using tree-ring chronologies from regions outside of North America are prohably limited. Dendroclimatic infoimation from tree rings can still contribute to the understanding of the global influences of past ENSO events dated from other sources. As more high quality dendroclimatic reconstructions are developed for the past few hundred years i t will be possible to map in detail the estimated surface climate conditions following the ENSO events. This will help to understand whether the extensive glohal climate anomalies that were associated with the recent extreme 1982- 1983 ENSO event occurred in the past and. therefore. may occur again in the future. Independent sources of infomiation about ENSO include historical tlocuments describing conditions along the coast of Peru. which have been ~isetlto date strong El Niiios back to the 16th century (Quinn et al.. 1978. 1987: Hamilton antl Garcia. 1986). Environmental records from cei-tain corals growing in tropical marine waters may iilso help to substantiate past ENSO dates and the nature of associated marine climate anomalies. Glynn and Wellington (1983) have. for example. linked variations in linear growth rates of corals from the Galapagos Islands with warmer sea suiface temperatures and El Nifio. Associations between stable oxygen isotope ratios in corals antl sea suiface tempcratureslEl Niiios in the eastern tropical Pacific have also been suggested lo exist (Driift'el. 19x5 antl Druffel et al.. this volume). as well as the cadmium content of the corals (Shen et al.. 1987: and Shen and Sanford. this volume). The snow accumulation rates of tropical ice caps are linked to regional precipitation which may. in turn. be linkctl to ENSO occurrences (Thompson et al.. 1984. 1985). The ice core records can. at least. pro\.ide detailed proxy climate histories for tropical South America ('rhompson a n d Thompson.
314
1987) and similar high resolution data from China (Thompson et al.. in press) will complement the extensive historical records of past climate from that region (Central Meteorological Bureau, 1981). All such high resolution proxy climate data, even if not directly related to ENSO, will help to provide a background description of climate in which to place historical ENSO occurrences. Any study of past climatic phenomena. pai-ticularly one as globally extensive in its effects as ENSO, must draw on all available sources of proxy climate information. To study past ENSO events, such series must be capable of resolving climatic infoimation with at least annual resolution. Such proxy records can only. however. provide a partial history of past climate. Each will contain bias and random error terms which are unrelated to climate (National Academy of Sciences, 1975). The most reliable description of past ENSO events and their climatic impact will only be obtained by comparing and combining the information contained in independent proxy climate records. Such a study has not been undertaken to date. The authors have, however. been compiling high quality, annuallyresolved, proxy climate records (from tree rings. documentary and other sources) for the past four centuries (see Lough. I986b,c) and examining the similarities and diffei-ences between the independent records (Lough. 1986d). Using the probable dates of El NiAo (Quinn et al.. 1987) it is now possible to examine in a fairly detailed manner the proxy evidence for consistent spatial patterns following ENSO events for the period prior to the start of instrumental climate records. A fir51 approximation to the extratropical teleconnections by this means could be complemented by the information about the local variations that are most closely coupled to ENSO events that may be obtained from, for example. coral skeletons. Such a study would emphasize the use of paleoclimatic data in an applied rather than just a simple descriptive role. ACKNOWLEDGEMENTS This work was supported by the National Science Foundation under Grant Nos. ATM 81 15754 and ATM 8319848. T.J. Blasing. G.A. Gordon and G.R. Lofgren made substantial contributions to the development of the spatial dendrocliniatic reconstructions.
6 REFERENCES
Aceituno. P., 1988. On the functioning of the Southern Oscillation in the South American sector. Part 1: surface climate. Mon. Wea. Rev.. I 16: 505-524. Allan. R.J., 1985. The Australian summer monsoon. teleconnections and flooding in the Lake Eyre Basin. South Australian Geographical Papers. No. 2. Royal Geographical Society of Australia. University of Adelaide Printing Scction. 47pp. Allan. R.J.. 1988. El Nitio Southein Oscillation influences in the Australasian region. Prog. Phys. Geogr.. 12: 313-348. Angell. J.K.. 198 I , Comparison of variations in atmospheric quantities with sea surface temperature variations in the eqiiatorial eastern Pacific. Mon. Wea. Rev.. 109: 230243.
315
Barnett. T.P.. I98 I. Statistical prediction of North American air temperalures from Pacific predictors. Mon. Wea. Rev.. 109: 1021-1041. Berlage. H.P.. 1957. Fluctuations of the general atmospheric circulation of more than one year. their nature and prognostic values. K . Ned. Meteorol. Inst., Mecled. Verh., 69, I52pp. Bjerknes. J., 1966. A possible response of the atmospheric Hadley circulation to equatorial anomalies of ocean temperature. Tellus, 18: 820-829. Bjerltnes. J.. 1969. Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev.. 97: 163- 172. Blasing, T.J.. 1978. Time series and multivariate analysis in paleoclimatology. In: H.H. Shugart (Editor). Time Series and Ecological Processes. SIAM-SIMS Conference Series. No. 5 . Society for lndustrial and Applied Mathematics. Philadelphia. PA, pp. 2 1 1-226. Blasing. T.J. and Duvick. D., 1984. Reconstruction of precipitation history in North American corn belt using tree rings. Nature. 307: 143-145. Blasing. T.J., Stahle. D.W. and Duvick. D.N.. 1988. Tree ring-based reconstruction of annual precipitation in the south-central United States from 1750 to 1980. Water Resour. Res.. 24: 163- I7 I. Boninsegna, J.A. and Holmes. R.H.. 1985. Fitzroya cupressoides yields 1534-year long South American chronology. Tree-Ring Bull.. 45: 37-4 I. Bormann, F.H. and Berlyn, G. (Editors) 1981. Age and Growth Rate of Tropical Trees: New Directions for Research. Bulletin No. 94. School of Forestry and Environmental Studies. Yale University. Briffa. K.R., Jones. P.D. and Schweingruher. F.H.. 1988. Summer temperature patterns over Europe : a reconstruction to 1750 A.D. based on maximum latewood density indices of conifers. Quat. Res., 30: 36-52. Briffa. K.R., Jones. P.D.. Wigley, T. M.L.. Pilcher, J.R. and Baillie. M.G.L.. 1983. Climate reconstruction from tree rings: Part I . basic methodology and preliminary results for England. J. Climatol.. 3: 233-242. Briffa. K.R., Jones. P.D.. Wigley. T.M.L., Pilcher. J.R. and Baillie, M.G.L., 1986. Climate reconstruction from tree rings: Part 2, spatial reconstruction of summer mean sea-level pressure patterns over Great Britain. J. Climatol.. 6: I - 15. Brubaker. L.B. and Cook. E.R., 1983. Tree-ring studies of Holocene environments. In : H.E. Wright Jr (Editor), Late Quaternary Environments of the United States, V01.2. University of Minnesota Press. pp 222-235. Bryson, R.A. and Hare, F.K., 1974. World Survey of Climatology. Climates of North America. H.E. Landsberg (Editor). Vol. 1 1 , Elsevier. New York. Campbell. D.A., 1982. Preliminary estimates of summer streamflow for Tasmania. In: M.K. Hughes, P.M. Kelly, J.R. Pilcher and V.C. LaMarche Jr. (Editors). Climate from Tree Rings. Cambridge University Pre5s. pp. 170-176. Central Meteorological Bureau, I98 I . Yearly Charts of DiynesslWetness in China for the Last 500-year Period. Cartographic Publishing House. Beijing. 332pp. Ceiveny. R.S., Skeeter, B.R. and Dewey. K.F.. 1987. A preliminary investigation of a relationship between South American snow cover and the Southern Oscillation. Mon. Wea. Rev., 115: 620-623. Chen. W.Y., 1982. Fluctuations in Noithem Hemisphere 700mb height field associated with the Southern Oscillation. Mon. Wea. Rev.. 110: 808-823. Conkey. L.E., 1986. Red spruce tree-ring widths and densities in eastern North America as indicators of past climate. Quat. Res.. 26: 232-243. Cook. E.R.. 1987. Past drought from tree rings in the eastern United States : statistical methods and reconstruction fidelity. J. Paleoclimatol. (in press). Cook, E.R. and Jacoby, G.C.. 1979. Evidence for quasi-periodic July drought in the Hudson Valley. New York. Nature. 282: 390-392. Cook, E.R. and Jacoby. G.C.. 1983. Potomac River streamflow since 1730 as reconstructed by tree rings. J. Clini. Appl. Meteorol.. 22: 1659-1672. Coughlan. M.J.. 1979. Recent variations in annual-mean maximum temperatures over Australia. Q. J. R. Meteorol. S O C . . 105: 707-719. Cropper. J.P.. 1982. Climate reconstructions (1801 to 1938) inferred from tree-ring width chronologies of the North American Arctic. Arct. Alpine Res.. 14: 223-241.
316 Cropper. J.P. and Fritts. H.C.. 1981. Tree-ring width chronologies from the Noith American Arctic. Arct , Alpine Res.. 13: 245-260. Deser. C. and Wallace, J.M.. 1987: El Nifio events and their relation to the Southein Oscillation: 1925-1986. J. Geophys. Res.. 92(C13): 14.189-14.196. Douglas. A.V.. 1980. Geophysical estimates of sea-surface temperatures off western North Americasince 1671. CALCOFI Rep. Vol. XXI: 102-1 12. Douglas, A.V. and Englehart. P.J.. I98 I . On a statistical relationship between autumn precipitation in the central equatorial Pacific and subsequent winter precipitation in Florida. Mon. Wea. Rev.. 109: 2377-2382. Douglas. A.V.. Cayan. D.R. and Namias. J., 1982. Large-scale changes in North Pacific and North American weather patterns in recent decades. Mon. Wea. Rev.. I 10: 185 I 1862. Druffel. E.R.M.. 1985. Detection of El Nifio and decade time scale variations in sea surface temperature from banded coral records: implications for the carbon dioxide cycle. In: The Carbon Cycle and Atmospheric CO,: Natural Variations Archean to Present. Geophysical Monograph 32. American Geophysical Union, pp. I I I - 122. Dunwiddie. P.W. and LaMarche, V.C.. 1980. A climatically responsive tree-ring record from Widdringtonia cedarbergensis. Cape Province. South Africa. Nature, 286: 796797. Emery. W.J. and Hamilton. K.. 1985: Atmospheric forcing of interannual variability in the northeast Pacific Ocean : connections with El Nifio. J. Geophys. Res.. 90(CI): 857868. Fritts. H.C.. 1974. Relationships of ring widths in arid-site conifers to variations in monthly temperature and precipitation. Ecol. Monogr., 44: 4 11-440. Fritts. H.C., 1976. Tree Rings and Climate. Academic Press. 567pp. Fritts. H.C. Reconstructing Large-Scale Climatic Patterns from Tree-Ring Data : A Diagnostic Study. University of Arizona Press (in press). Fritts. H.C. and Shatz. D.J.. 1975. Selecting and characterizing tree-ring chronologies for dendroclimatic analysis. Tree-Ring Bull.. 35: 3 1-40. Fiitts, H.C. and Gordon. G.A.. 1982. Reconstructed annual precipitation for California. In: M.K. Hughes. P.M. Kelly. J.R. Pilcher and V.C. LaMarche Jr. (Editors), Climate from Tree Rings. Cambridge University Press, pp. 185-191. Fritts, H.C. and Lough. J.M., 1985. An estimate of average annual temperature variations for North America. 1602-1961. Clim. Change. 7: 203-224. Fritts. H.C.. Lofgren. G.R. and Gordon. G.A.. 1979. Variations in climate since 1602 as reconstructed from tree rings. Quat. Res.. 12: 18-46. Fritts. H.C.. Blasing. T.J.. Hayden. B.P. and Kutzbach. J.E.. 1971. Multivariate techniques for specifying tree-growth and climate relationships and for reconstructing anomalies in paleoclimate. J. Appl. Meteorol.. 10: 845-864. Garfinkel. H.L. and Bruhaker. L.B.. 1980. Modern climate-tree-growth relationships and climate reconstruction in sub-Arctic Alaska. Nature. 286: 872-874. Gill. A.E. and Rasmusson. E.M.. 1983. The 1982-83 climate anomaly in the equatorial Pacific. Nature. 306: 229-234. Glahn, H.R., 1968. Canonical correlation and its relationship to discriminant analysis and multiple regression. J. Atmos. Sci.. 25: 23-31. Glynn. P.W. and Wellin ton. G.M.. 1983. Corals and Coral Reefs of the Galapagos Islands. University o f California Press. 330pp Gordon. G.A.. 1982. Verification of dendroclimak reconstructions. In: M.K. Hughes. P.M. Kelly, J.R. Pilcher and V.C. LaMarche Jr. (Editors). Climate from Tree Rings. Cambridge University Press. pp. 58-6 I . Gordon, G.A. and LeDuc. S.K.. 198 I . Verification statistics for regression models. In: Proceedings Seventh Conference on Probability and Statistics in Atmospheric Science, Monterey. American Meteorological Society. pp. 129-133. Gordon. G.A., Lough. J.M., Fritts, H.C. and Kelly. P.M.. 1985. Comparison of sea level pressure reconstructions from western North American tree rings with a proxy record of winter severity in Japan. J. Clini. Appl. Meteorol.. 24: 1219-1224. Gordon. N.D.. 1986: The Soulhein Oscillation and New Zealand weather. Mon. Wea. Rev.. 114: 371-387.
317 Graumlich. L. and Biubaker. L.B.. 1986. Reconstruction of annual temperature (15901979) for Longmire. Washington derived from tree rings. Quat. Res.. 25: 223-234. Graybill. D.A., 1982. Chronology development and analysis. In : M.K. Hughes, P.M. Kelly. J.R. Pilcher and V.C. LaMarche Jr (Editors). Climate from Tree Rings. Cambridge University Press. pp. 2 1-30. Guiot. J.. Berger, A.L.. Munaut. A.V. and Till. C.. 1982: Some new mathematical procedures in dendrochronology with examples from Switzerland and Morocco. TreeRing Bull., 42: 33-48. Hamilton. K.. 1988. A detailed examination of the extratropical response to tropical El NifiolSouthern Oscillation events. J . Climatol.. 8: 67-86. Hamilton. K. and Garcia. R.R., 1986. El Nifio/Southem Oscillation events and their associated midlatitude teleconnections I53 I - I84 1 . Bull. Am. Meteorol. SOC..67: 13541361. Holland. G.J.. 1986. Interannual variability of the Australian summer monsoon at Darwin: 1952-82. Mon. Wea. Rev.. 114: 594-604. Holmes. R.L.. 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bull.. 43: 69-78. Holmes, R.L., Stockton. C.W. and LaMarche. V.C.. 1979. Extension of river flow records in Argentina from long tree-ring chronologies. Water Resour. Bull.. 15: 108 1 - 1085. Horel, J.D. and Wallace, J.M.. 198 I . Planetary-scale atmospheric phenomena associated with the Southern Oscillation. Mon. Wea. Rcv.. 109: 813-829. Hughes. M.K., Kelly. P.M.. Pilcher. J.R. and LaMarche, V.C. .Jr. (Editors) 1982. Climate from Tree Rings. Cambridge University Press. 223pp. Hughes. M.L.. Schweingruber. F.H., Cartwright. D. and Kelly, P.M.. 1984. July-August temperature at Edinburgh between 172 I and 1975 from tree-ring density and width data. Nature. 308: 341-344. Jacoby. G.C. and Cook. E.R.. 198 I . Past temperature variations inferred from a 400-year tree-ring chronology from Yukon Territory. Canada. Arci. Alpine Res., 13: 409-4 18. Jacoby. G.C.. Cook, E.R. and Ulan. L.D.. 1985. Summer degree days in Alaska since 1524. Quat. Res.. 23: 18-26. Jacoby, G.C., Ivanciu. I.S. and Ulan. L.D.. 1988. A 263-year record of summer temperature for northern Quebec reconstructed from tree-ring data and evidence of a major climatic shift in the early 1800s. Palaeogeogr., Palaeoclimatol., Palaeoecol., 64: 69-78. Kojo. Y . , 1987: A dendrochronological study of Cryptomeriajaponicain Japan. Tree-Ring Bull. 47: 1-2 I . Kutzbach. J.E.. 1975. Diagnostic studies of past climates. In: The Physical Basis of Climate and Climate Modelling. GARP Publication Series Number 16. ICSU/WMO. WMO. pp. 119-126. Kutzbach. J.E. and Guetter, P.J.. 1980. On the design of paleoenvironmental networks for estimating large-scale patterns of climate. Quat. Res. 14: 169- 187. LaMarche. V.C. and Stockion. C. W.. 1974. Chronologies from temperature-sensitive bristlecone pines at upper treeline in western United Stales. Tree-Ring Bull.. 34: 2145. LaMarche. V.C. and Pittock. A.B.. 1982. Preliminary temperature reconstructions for Tasmania. In: M.K. Hughes, P.M. Kelly. J.R. Pilcher and V.C. LaMarche Jr. (Editors), Climate from Tree Rings. Cambridge University Press, pp. 177- 185. LaMarche. V.C.. Holmes. R.L., Dunwiddie. P. W. and Drew, L.G.. 1979a. Tree-ring chronologies of the Southern Hemisphere, Vol. 1 : Argentina. Chronology Series V. Laboratory of Tree-Ring Research, University of Arizona, Tucson. AZ. 69pp. LaMarche. V.C.. Holmes. R.L.. Dunwiddie. P.W. and Drew. L.G.. 1979b. Tree-ring clironologies of the Southern Hemisphere. Vol. 2: Chile. Chronology Series V. Laboratory of Tree-Ring Research. University of Arizona. Tucson, AZ, 43pp. LaMarche, V.C.. Holmes. R.L., Dunwiddie. P. W. and Drew, L.G., 1 9 7 9 ~Tree-ring . chronologies of the Southern Hemisphere, Vol. 3: New Zedland. Chronology Series V. Laboratory of Tree-Ring Research. University of Arizona. Tucson. AZ. 77pp.
318 LaMarche. V.C.. Holmes, R.L., Dunwiddie. P. W. and Drew. L.G.. I979d. Tree-ring chronologies of the Southern Hemisphere. Vol. 4: Australia. Chronology Series V. Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ. 89pp. LaMarche. V.C.. Holmes. R.L.. Dunwicldie. P.W. and Drew. L.G.. 1979e. Tree-ring chronologies of the Southern Hemisphere. Vol. 5 : South Africa. Chronology Series V. Laboratory of Tree-Ring Research. University of Arizona, Tucson, AZ, 27pp. LaMarche, V.C., Graybill. D.A.. Fritts. H.C. and Rose. M.R.. 1984. Increasing atmospheric carbon dioxide: tree ring evidence for growth enhancement in natural vegetation. Science, 225: 1019-1021. Lau. N.-C.. 1981. A diagnostic study of recurrent meteorological anomalies appearing in a 15-year simulation with a GFDL general circulation model. Mon. Wea. Rev.. 109: 2287-23 1 I . k i t h . C.E.. 1984. Global climate research. In: J.T. Houghton (Editor), The Global Climate. Cambridge University Press. pp. 13-24. Lenz. 0.. Shar. E. and Schweingruber. F.H.. 1976. Methodische Probleme bei der radiographisch-densitometrischen Bestimmung der Dichte und der Jahrringbreiten von Holz. Holzforschung. 30: 114-123. Lindesay. J.A., 1988. South African rainfall. the Southern Oscillation and a Southern Hemisphere semi-annual cycle. J. Climatol.. 8: 17-30. Livezey, R.E. and Chen. W.Y.. 1983. Statistical field significance and its determination by Monte Carlo techniques. Mon. Wea. Rev.. 1 I I : 46-59. Lorenz, E. N . , 1956. Empirical orthogonal functions and statistical weather prediction. M.I.T. Statistical Forecasting Prqject. Scientific Report I , Contract AF(604)-1566, 49PP. Lough. J.M.. 1986a. Reconstructed sea surface temperature of the eastern North Pacific: comparisons with other dendroclimatic reconstructions. Technical Note Number 44. Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ. 40pp. Lough. J.M.. 198613. Description of climatic data sets assembled for Climatic Regimes project. Technical Note Number 4 I . Laboratory of Tree-Ring Research. University of Arizona. Tucson. AZ. 27pp. Lough, J.M.. I986c. Climatic data and climatic events indexes and bibliographies. Climate Regimes pro'ect. Technical Note Number 42. Laboratory of Tree-Ring Research. University o! Arizona. Tucson. AZ. 95pp. Lough. J.M., 1986d. Climatic regimes of the western hemisphere since 1600 A.D. : an examination of the evidence from climate/proxy climate series. Technical Note Number SO. Laboratory of Tree-Ring Research. University of Arizona, Tucson. AZ. 90pp. Lough. J.M. and Fritts. H.C.. 1985. The Southern Oscillation and tree rings: 1600-1961. J. Clim. Appl. Meteorol.. 24: 952-966. Lough. J.M. and Fritts. H.C.. 1987. An assessment of the possible effects of volcanic eruptions on North American climate using tree-ring data. I602 to I900 A.D. Clim. Change. 10: 2 19-239. Lough. J.M.. Fritts. H.C. and Wu Xiangding. 1987. Relationships between the climates of China and North America over the past four centuries: a comparison of proxy records. In: Ye Duzheng. Fu Congbin, Chao Jiping and M. Yoshino (Editors). The Climate of China and Global Climale. China Ocean Press. Springer-Verlag, pp. 89- 105. McBride. J.L. and Nicholls. N.. 1983. Seasonal relationships between Australian rainfall and the Southern Oscillation. Mon. Wea. Rev.. I I I : 1998-2004. Mitchell. J.M.. Dzerdzeevsltii. B.. Flohn. H.. Hofmeyer. W.L.. Lamb. H.H., Rao, K.N. and Wallen. C.C.. 1966. Climatic Change. World Meteorological Organization Technical Note Number 79. WMO. Geneva. Switzerland. Monserud. R.A., 1986. Time-series analysis of Wee-ring chronologies. For. Sci.. 32: 349372. Namias. J. and Cayan. D.R.. 1984. El Niiio : the implications for forecasting. Oceanus. 27: 41-47. National Academy of Sciences. 1975. Untlel-standing Climatic Change. Nalional Academy Press. Washington. D.C. Nicholls. N., 1983. Predicting Indian monsoon rainfall from sea-surface temperature in the Indonesia-north Australia region. Nature. 306: 576-577.
319
Nicholls. N.. 1985. Impact of the Southern Oscillation on Australian crops. J. Climatol.. 5 : 553-560. Nicholson, S.E. and Entekhabi. D., 1986. ‘rile quasi-periodic behaviour of rainfall variability in Africa and its relationship to the Southern Oscillation. Arch. Meteorol. Geophys. Biokliniatol. Ser. A. 34: 31 1-348. Norton, D.A., 1987. Reconstruction of past river flow and precipitation in Canterbuiy, New Zealand from an analysis of tree rings. N . Z. J. Hydrol.. 26: 161-174. Norton. D.A., 1988. Historical development of Southern Hemisphere dendrochi-onology. In: E. R. Cook and L. Kairiukstis (Editors). Methods of Tree-Rings Analysis: Applications in the Environmental Sciences. IIASA. Laxenberg. Reidel Press. Dordrecht. Norton. D.A. and Ogden, J.. 1987. Dendrochronology: A review with emphasis on New Zealand applications. N. Z. J. Ecol., 10: 77-95. Ogallo, L.. 1988. Relationships between seasonal rainfall in east Africa and the Southern Oscillation. J. Climatol.. 8: 31-43. Ogden. J . , 1978. On the dendrochronological potential of Australian trees. Aust. J. Ecol.. 3: 339-356. Ogden. J.. I98 1 . Dendrochronological studies and the deteimination of tree ages in the Australian tropics. J. Biogeogr.. 8: 405-420. Philander. S.G.H.. 1985. El Niiio and LaNiiia. J. Atmos. Sci.. 42: 2652-2662. Pittock. A.B., 1975. Climatic change and the patterns of variation in Australian rainfall. Search. 6: 498-504. Pittock. A.B.. 1980a. Patterns of climatic variation in Argentina and Chile - I. Precipitation. 1931-60. Mon. Wea. Rev.. 108: 1347.1361. Pittock. A.B.. 1980b. Patterns of climatic variation in Argentina and Chile - 11. Temperature. 1931-60. Mon. Wea. Rev.. 108: 1362-1369. Pittock. A.B.. 1984. On the reality. stability. and usefulness of southern hemisphere teleconnections. Aust. Meteoi-ol. Mag., 32: 75-82. Pittock. A.B. and Salinger. M.J.. 1982. Towards regional scenarios for a warmed earth. Clim. Change, 4: 23-40. Quinn. W.H. and Neal, V.T.. 1983. Long-tern1 variations in the Southern Oscillation. El Niiio and Chilean subtropical rainfall. Fish. Bull.. 8 I : 363-376. Quinn. W.H., Zopf, D.O., Short. K.S. and Kuo Yang. R.T.W.. 1978. Historical trends and statistics of the Southern Oscillation. El Niiio and Indonesian droughts. Fish. Bull.. 76: 663-678. Quinn. W.H., Neal, W.T. and Antunez de Mayolo. S.E.. 1987. El Niiio occui-rences over the past four and a half centuries. J. Geophys. Res.. 92(C13): 14.449-14.461. Quiroz. R.S.. 1983. The climate of the El Niiio winter of 1982-83 - a season of extraordinary climatic anomalies. Mon. Wea. Rev.. 1 I I : 1685-1706. Rasmusson. E.M. and Carpenter. T.. 1982. Variations in tropical sea suiface temperature and surface wind fields associated with the Southern Oscillation/El Niiio. Mon. Wea. Rev.. 110: 354-384. Rasmusson. E.M. and Hall. J . M . . 1983. The great equatorial Pacific Ocean wainiing event of 1982-1983. Weatheiwise. 36: 166-175. Rasmusson. E.M. and Wallace, J.M.. 1983. Meteorological aspects of the El Niiio/Southem Oscillation. Science. 222: I 195-1202. Rasmusson. E.M. and Arkin. P A.. 1985. Interannual cliniate variability associated with the El Niiio/Southem Oscillation. In: J.C.J. Nihoul (Editor). Coupled OceanAtmosphere Models. Elsevier. pp. 697-725. Robinson. W.J. and Evans. R.. 1980. A microcomputer-based trec-ring measuring system. Tree-Ring Bull., 40: 59-64. Rogers. J.C.. 1988. Precipitation variability over the Caribhean antl tropical Americas associated with the Soulhern Oscillation. J. Climate. I : 172- 182. Ropelewski. C.F. and Halpcrt. M.S.. 1986. North American precipitation antl temperature patterns associated with the El Nifio/Southern Oscillation (ENSO). Mnn. Wea. Rev.. 114: 2352-2362. Ropelewslti. C.F. arid Halpert. M.S.. I987a. Global antl regional scale precipitation patterns associatecl with the El NiiioiSnuthcin Oscillation. Mon. Wea. Rev.. I 15: 16061626 .
320 Ropelewski, C.F. and Halpert. M.S.. 198713. Global and regional scale temperature patterns associated with the El Niiio/Southern Oscillation. Proceedings of the Eleventh Annual Climate Diagnostics Workshop, U . S . Department of Commerce/NOAA. Champaign. Illinois, October 14- 17. 1986. pp. 69-72. Ropelewski, C.F. and Jones, P.D., 1987. An extension of the Tahiti-Daiwin Southern Oscillation index. Mon. Wea. Rev., 115: 2161-2165. Rose, M., 1983. Time domain characteristics of tree-ring chronologies and eigenvector amplitude series from western North America. Technical Note Number 25, Laboratory of Tree-Ring Research, University of Arizona, Tucson. AZ. 42pp. Salinger, M.J.. 1980a. New Zealand climate: I . Precipitation patterns. Mon. Wea. Rev., 108: 1892-1904. Salinger, M.J., 1980b. New Zealand climate: 11. Temperature patterns. Mon. Wea. Rev., 108: 1905-1912. . New Zealand temperature series. Climate Monitor, 9: 112-1 18. Salinger, M.J., 1 9 8 0 ~The Salinger, M.J., 1982. New Zealand climate scenarios for a warn high-C02 world. Weather and Climate, 2: 9-15. Schweingruber, F.H., 1982. Measurement of densitometric properties of wood. In: M.K. Hughes, P.M. Kelly. J.R. Pilcher and V.C. LaMarche Jr. (Editors). Climate from Tree Rings. Cambridge University Press, pp. 8- 12. Shen, G.T., Boyle, E.A. and Lea, D.W.. 1987. Cadmium in corals: chronicles of historical upwelling and industrial fallout. Nature, 328: 794-796. Shukla, J. and Paolino, D.A., 1983. The Southern Oscillation and long-range forecasting of the summer monsoon rainfall over India. Mon. Wea. Rev.. 1 I I : 1830- 1837. Simmons, A.J.. Wallace, J.M. and Branstator. G.W., 1983. Baratropic wave propagation and instability, and atmospheric teleconnections. J. Atmos. Sci., 40: 1363-1392. Stahle, D.W. and Cleaveland, M.K.. 1988. Texas drought history reconstructed and analysed from 1698 to 1980. J. Climate, 1: 59-74. Stahle, D.W., Cleaveland, M.K. and Hehr. J.G.. 1985. A 450-year drought reconstruction for Arkansas, United States. Nature. 3 16: 530-532. Stahle. D.W., Cleaveland, M.K. and Hehr. J.G.. 1988. North Carolina climate changes reconstructed from tree rings: AD 372 to 1985. Science, 240: 1517-1519. Stockton. C.W.. Boggess. W.R. and Meko. D.M., 1985. Climate and tree rings. In: A.D. Hecht (Editor). Paleoclimate Analysis and Modeling. John Wiley and Sons, pp. 7 I - 15 I . Stoeckinius. T . , 1981, Interannual variations of tropical precipitation patterns. Mon. Wea. Rev., 109: 1233-1247. Stokes, M.A. and Smiley, T.L.. 1968. An Introduction to Tree-Ring Dating. University of Chicago Press, Chicago. Thompson. L.G. and Mosley-Thompson, E., 1987. Evidence of abrupt climatic change during the last 1,500 years recorded in ice cores from the tropical Quelccaya ice cap, Pew. In: Abrupt Climatic Change, W.H. Berger and L.D. Labeyrie (Editors). D. Reidel. pp. 99-1 10. Thompson, L.G., Mosley-Thompson, E., and Morales Amao. B., 1984. El Niiio- Southern Oscillation events recorded in the stratigraphy of the tropical Quelccaya ice cap. Peru. Science, 226: 50-53. Thompson, L.G., Xiaoling. Wu, Mosley-Thompson. E. and Zichu. Xie. 1988. Climatic records from the Dunde ice cap, China. Ann. of Glaciol.. 10: (in press) Thompson, L.G., Mosley-Thompson, E.. Bolzan, J.F. and Koci, B.R. 1985. A 1500-year record of tropical precipitation in ice cores from the Quelccaya ice cap. Peru. Science, 229: 971-973. Till. C., 1987. The summary response function of Cedrus atlantica (Endl .) Carriere in Morocco. Tree-Ring Bull.. 47: 23-36. Trenherth, K.E.. 1976. Spatial and temporal variations of the Southern Oscillation. Q. J. R. Meteorol. Soc., 102: 639-653. Trenberth. K.E. and Paolino. D.A.. 1981. Characteristic patterns of variability of sea level pressure in the Northern Hemisphere. Mon. Wea. Rev., 109: 1 169- I 189. Troup, A.J.. 1965. The Southern Oscillation. Q. J. R. Meteorol. Soc.. 91: 490-506. van Loon. H. and Madden. R.A.. 1981. Tlie Southern Oscillation. Part I : Global associations with pressure and tempernlure i n northem winter. Mon. Wea. Rev., 109: 1150-1 162.
321 van Loon, H. and Rogers, J.C.. 1981. The Southern Oscillation. Part 11: Associations with changes in the middle troposphere in the northern winter. Mon. Wea. Rev.. 109: 1 1631168. van Loon. H. and Shea. D.J., 1987. The Southern Oscillation. Part VI: Anomalies of sea level pressure of the Southern Hemisphere and of Pacific sea surface temperatures during the development of a warm event. Mon. Wea. Rev., I 15: 370-379. Villalba. R., Boninsegna. J.A. and Holnics. R.L.. 1985. Cedrela angustifolia and Juglans australis: two new tropical species useful in dendrochronology. Tree-Ring Bull.. 45: 2535. Walker. G.T. and Bliss, E.W.. 1932. World Weather V. Meni. Roy. Meteorol. SOC..4: 53-84. Wang. Shao-Wu, 1984. El Niiio and summer temperatures in northeast China, 1860-1980. Trop. Ocean-Atmos. Newsl.. 25: 4 . Whetton, P.H., 1988. A synoptic climatological analysis of rainfall variability in southeastem Australia. J. Climatol.. 8: 155-177. Wright, P.B.. 1975. An Index of the Southem Oscillation. CRU RP4. Climatic Research Unit. Norwich. England, 22pp. Wright, P.B.. 1977. The Southern Oscillation - patterns and mechanisms of the teleconnections and the persistence. Hawaii Institute of Geophysics. MIG-77- 13. University of Hawaii. 107pp. Wright. P.B., 1985. The Southern Oscillation: an ocean-atmosphere feedback system? Bull. Am. Meteorol. SOC..66: 398-412. Wright. P.B., Mitchell. T.P. and Wallace, J.M.. 1985. Relationships between surface observations over the global oceans and the Southern Oscillation. NOAA Data Report ERL PMEL- 12. Pacific Marine Environmental Laboratory. Seattle. Washington. 6 I pp. Yarnal. B.. 1985. Extratropical teleconnections with El Niiio/Southem Oscillation (ENSO) events. Prog. Phys. Geogr.. 9: 315-352. Zhang, Ming-Lie and Zeng. Zhau-Mei. 1984. Teleconnections between sea surface temperature in the tropical eastern Pacific and air temperature in northeast China. Trop. Ocean-Atmos. Newsl., 23: 5-7.
This Page Intentionally Left Blank
323
EFFECTS OF EL NIRO 1982-83 ON BENTHOS, FISH AND FISHERIES OFF THE SOUTH AMERICAN PACIFIC COAST WOLF E. ARNTZ Alfred-Wegener-Institut fur Polar- und Meeresforschung, Columbusstrasse, D-2850 Bremerhaven (Federal Republic of Germany) JUAN TARAZONA Grupo DePSEA, Facultad de Ciencias Biol6gicas, Universidad Nacional Mayor de San Marcos, Apartado 1898, Lima 100 (Per$ ABSTRACT Arntz, W.E., and Tarazona, J., 1989. Effects of El Niiio 1982-83 on benthos, fish and fisheries off the South American Pacific coast. Biological effects of El Niiio (EN) were first detected on the Peruvian coast and have been reported for some of the major events during this century, especially those of 1925-26, 194041 and 1972-73. EN 1982-83 exceeded all these earlier events in strength and duration, and was the first to be investigated in more detail, both on a latitudinal gradient and including, for the first time, the benthic subsystem of the Humboldt Current upwelling area off the west coast of South America. Despite the fact that the coverage of the area of major impact (over 4,000 km of coastline) was by no means complete, and that most of the causes and mechanisms behind the observed changes remained unclear due to the lack of experimental (e.g. tolerance) studies, the observational evidence and the fluctuations observed in the landing statistics have contributed a great deal to our understanding of the way in which a major EN event affects the upwelling system as a whole. In the pelagic subsystem, a surface temperature increase of up to 1l0C, the effects of which could be traced down to at least 1,000 m depth, led to a tropicalization of the ecosystem, the disruption of the normal food web, and induced changes in species composition and migrations of a large number of fish and invertebrate species populations. This resulted in a general impoverishment of this system. The consequences for some commercially important species, which had already been damaged by former EN events and the impact of overfishing, were severe off Peni and Ecuador although there were certain positive effects due to the inshore migration of various oceanic mackerels and the southward migration of (sub)tropical fish species such as dolphinfish and skipjack tuna. These beneficial effects were, however, only of minor importance for the fisheries, which were not equipped to deal with the new target species. In the benthic subsystem, a factor of major importance, besides the increase in temperature, was the marked increase of dissolved oxygen at the seafloor during and after the event. Despite mass mortalities of many of the cold-water adapted invertebrate species on sandy beaches, shallow soft bottoms and in the rocky intertidal zone, the overall effect on fisheries was rather positive. In the case of commercially exploited shellfish ("mariscos"), this was due to an increased immigration of subtropical species, mostly crustaceans, as well as increases in abundance of some local species more tolerant of warm water such as the Peruvian scallop. This led to an unprecedented boom of some fisheries on the northem and central Peruvian coast. For the demersal fisheries and the artisanal finfish fishery, the immediate effects of EN were detrimental due to a general dispersal and downslope migration of many species, but the long-term effects appeared to have been beneficial: improved feeding conditions at the seafloor and decreased fishing pressure caused a major recruitment success of some species such as Pacific hake, and this may provide increased catches several years after EN. The primary EN effects described in this review apparently induced many secondary effects. The disruption of the pelagic food web caused mortalities and reproductive failures of guano birds and seals; mortalities (including marine algae) and increases in population abundance in shallow water led to temporary changes in species composition and successional processes that remain to be studied in more detail. On the whole, however, recovery of most species occurred rapidly, and relatively few long-term effects have remained, demonstrating the remarkable resilience of the upwelling ecosystem.
324 1 INTRODUCTION Only recently has it become evident that the El Niiio cunent (EN), although mainly associated with the Southern Oscillation (Rasmusson and Wallace, 1983), is a phenomenon with global effects. Data collected during the particularly strong event of 1982-83 revealed its teleconnections and worldwide impact to their full extent (Dayton and Tegner, 1984; Emery and Hamilton, 1985; Gunnill, 1985; Wooster and Fluharty, 1985; La Cock, 1986; Mysak, 1986; Paine, 1986; Glantz et al., 1987; Pearcy and Schoener, 1987; Tegner and Dayton, 1987). Despite this, the most direct oceanographic and biological effects, with enormous impacts on fisheries, agriculture and other human activities, occur in the southeast Pacific. Here warming in the summer was "detected' by local fishermen around Christmas time and this event was first described by marine scientists as a "current" that can develop to a marked extent, affecting coastal ecosystems, every couple of years (see Chirinos de Vildoso, 1976; Enfield, 1989a). Only during the last two decades did the true nature of EN become evident due in large measure to the work of Wyrtki (1975, 1979, 1982). Although information on the occurrence of EN is available for the whole timespan since the Spanish conquest about four and a half centuries ago (Quinn et al., 1987), retrospective proxy evidence of EN also exists for the period prior to the European discovery of South America (Thompson et al., 1984; Rollins et al., 1986; DeVries, 1987; Lough and Fritts, this volume). First accounts of changes in the spatial distribution of fish species, such as anchovy, were published early this century (Lavalle y Garcia, 1917; Murphy, 1926). Before EN 1982-83, almost all information on this event was limited to the immigration of oceanic fish and to the emigration or mortality of local fish species, which resulted in high mortality of guano birds (Vogt, 1940, 1957; Del Solar, 1942; Schweigger, 1953, 1960, 1964; Jordh, 1964; JordCln and Fuentes, 1966). Since the 1950s, the industrial anchovy fishery developed rapidly, resulting in very high rates of exploitation, which required management actions to preserve this resource. After EN 1972-73, a somewhat more systematic although still incomplete evaluation of changes in the pelagic subsystem was initiated by the Instituto del Mar del P e d (Chirinos de Vildoso, 1976; Valdivia, 1978; Samamt et al., 1978) and others (Jimtnez, 1982; Del Solar, 1983). It was only at the beginning of the eighties, and especially during and after EN 198283, that more integrated research on the entire marine ecosystem developed, which included macrobenthos as well as demersal fish and their fisheries (Arntz and J. Valdivia, 1985; F O P , 1985; Robinson and del Pino, 1985; Wooster and Fluharty, 1985; Arntz, 1986; Glynn, 1988). In recent years it has been shown that EN effects range from the cellular to the ecosystem level, implying physiological, biochemical and ecological changes, many of which have been investigated in a very fragmentary way. This, together with the enormous latitudinal extension of the direct impact of EN, complicates the study of its biological effects. This is especially true for benthos and nekton, which include species with short and long lifespans that suffer immediate and long-term effects from event-related disturbances. In this review the principal effects of EN are summarized, particularly with reference to the 1982-83 EN impact on macrobenthos, finfish and shellfish ("mariscos") of the South American Pacific coast (Fig. 1). The bulk of the information centers on Perd and northern Chile, but
325
.la
Q:G
ua y a q u i I
(.:... P .. .-....... .._ ...
;a pagos
v ..;i t l
5"s
K
\ 10"
k 15"
HlLE ... ... : 20"
90"W I
85" I
80" I
75" I
Fig. 1. Map of area of major impact of EN on the South American west coast. also includes some data on Ecuador and Colombia from which less information is available. The paper further points out some of the most likely mechanisms and processes underlying the observed changes, and compares the impact of events of varying intensity on the fisheries. In this context, and for reasons of simplicity, the term "normal years" is used to denote non-"EN years" although we are aware of the fact that EN is an integral part of the dynamics of the southeast Pacific ecosystem (Amtz and J. Valdivia, 1985). 2 PRINCIPAL ABIOTIC CHANGES INDUCED BY EN 1982-83 Atmospheric and oceanographic changes during EN, especially those during 1982-83, have been referred to in a large number of publications, among others by Cane (1983, 1986),
326
Rasmusson and Wallace (1983), Barber and Chilvez (1983, 1986), McCreary and Anderson (1984), Leetma et al. (1987), Enfield (1989b) and Fahrbach et al. (in press). Many of these changes and specific EN manifestations are suspected to be of ecological significance for the Humboldt Current* upwelling system during an EN event: Kelvin, Rossby and Yanai waves, trapped waves; changes in currents and sea level, air and water temperature and the depth of the thermocline, salinity and the depth of the pycnocline; increased rainfall, inundations and increased sedimentation; wind force, swell and agitated seas ("marejadas"); changes in upwelling and their effects on nutrient availability; dissolved oxygen increase and H2S decrease at the seafloor. In most cases, however, we are only beginning to understand the full impact of these conditions on species populations and communities. The mechanisms behind the observed changes are mostly unclear. Off mainland Ecuador the sea level, which was occasionally above the long-term average between May and September 1982, rose sharply starting in October with the arrival of the Kelvin wave front and reached a first peak of about 40 cm in January 1983. After a decline in the following months, a second peak of about the same height was observed in May 1983 before a steady decline that reached pre-EN levels in August [Cucalh, 1987; INOCAR (Instituto Oceanogrilfico de la Armada, pers. corn.)]. The sea level increase off Per& of up to 40 cm, was about the same height as observed further north (Enfield and Lukas, 1984). Maxima of around 30 cm were reported as far south as Antofagasta (23"s) in Chile (Fonseca, 1985). The sea level along about 25 degrees of latitude peaked in January 1983 and again between April and June of the same year. The exact southern extent of these increases was never fully documented for EN 1982-83. However, Enfield and Allen (1980) showed that former EN sea level changes could be detected from Alaska to southern Chile. Sea level rise during EN is often accompanied by strong swells and agitated seas ("marejadas") that can be harmful to intertidal organisms. Coastal erosion often occurs at such times as well as the destruction of buildings, boats, etc. in Ecuador and northern Peni (Moreano, 1984; Enfield, 1989b). The sea level elevation due to the arrival of the Kelvin wave front also led to an enhancement of poleward flow by coastal Kelvin waves (Hansen, this volume), and thus an advection of warm water into the traditionally cool upwelling system of the Humboldt Current area. The flow of the Humboldt Current was strongly reduced and even reversed its direction at times in the fall of 1983 (see review by Barber and Kogelschatz, this volume). At lo's, the Peni Undercurrent increased its poleward speed from about 4 cm sec-1 to over 25 cm sec-1 with peak velocities of about 36 cm sec-1,returning to its former slower speeds after mid-July 1983 (Smith, 1983, 1984). Further north, along the equator west and east of the Galilpagos Islands, the usual westward equatorial flow stopped, and the surface currents to the west of the islands were reversed by the end of 1982 carrying warmer water from the west into the Galilpagos area. The Equatorial Undercurrent apparently disappeared or even partly reversed
*The Humboldt Current and Perti Current are identical and refer to the system of relatively shallow currents flowing generally equatorward along the west coast of South America.
327 its direction during this period (Firing et al., 1983; JimCnez and Intriago, 1984; Taft, 1985; Leetma et al., 1987). Sea surface temperatures (SSTs) off Ecuador in February 1983 were 4°C higher than in normal years, but temperatures at 50 m depth exceeded the normal values by 9°C revealing that the subsurface signal of EN in this area was much stronger than the surface signal (Cucalbn, 1987). Measurements off central Peni, however, indicated a steady decline of the temperature anomalies from the surface towards deeper waters (Arntz et al., 1985). In November 1982, SSTs were 23 - 26'C along the Peruvian coast, that is 4 - 6°C warmer than normal. The warm waters at that time extended as far south as Arica (18'30) in Chile, with a few cold water tongues of 19 - 20'C off San Juan and Atico in southern Perli (Amtz, 1986). Already during December these cooler upwelling areas shrank drastically and were restricted to areas close to the coast. In most cases, however, upwelling no longer reached below the thermocline and thus brought only warm, nutrient poor waters to the surface. Only during EN 1982-83 did it become obvious that the reason for the observed reduced primary production was not to be sought in a weakening of coastal winds, and thus a general interruption of the upwelling process, but rather in the deepening of the thermocline and nutricline (Barber and ChBvez, 1983; Huyer et al., 1987; Leetma et al., 1987). In shallow nearshore areas, the decline in nutrients did not seem to be so serious, although even the levels of primary production decreased to 20-50% of normal values (Barber and Kogelschatz, this volume). Relatively high levels of productivity in nearshore areas were measured not only in northem Peni (Barber and Chrivez, 1983), but also along the central Peruvian coast (Tarazona et al., 1985b) even during the peak period of the event. SSTs reached their first peak off Peni in January 1983, with temperatures between 26 30'C north of Callao (12's). In southern Peni this maximum was delayed until March, attaining 23 - 27°C. In northern Chile SSTs increased to 28"C, exceeding the long-term mean by 4.5'C. Southward the warming extended at least to Concepcibn, Chile (37's; Silva and Rojas, 1984; Kelly, 1985). While in southem Peni and northern Chile surface temperatures gradually started to decline from March 1983 (Hydrographic Institute of Peruvian Navy, pers. comm.; Fuenzalida, 1985), warming north of Callao continued until June, with a second peak during April and May 1983 (Fig. 2), and in some areas rose to positive SST anomalies of up to 11'C. Only by mid-July did temperatures in the north decline drastically and attain mostly normal levels by October 1983 (Arntz, 1986). All along the South American Pacific coast the anomalous SSTs were accompanied by large scale warming at depth. Off central Peni, the 15°C isotherm, which is normally found at about 50 m depth nearshore, descended to > 200 m in December 1983 (GuillCn et al., 1985). In northern Chile it dropped to 100 - 150 m depth (Blanco and Dfaz, 1985). At the peak of the event, oceanographers traced the warming down to 800 m (GuillCn et al., 1985), and in some cases even beyond 1,000 m (Leetma et al., 1987). A secondary effect of the deepening of the thermocline, and thus the increase of the mixed layer, may have been the dispersal and deepening of phytoplankton populations, which in some cases may have been transported out of the euphotic zone. However, sunlight penetrates the
328
4
3 2
1 0 26 24 22 20 18 16 14 12 10 Aug-81
Sep-82
Oct-83
Nov-84
Dec-85
Jan-87
Feb-88
Fig. 2. Changes in near-bottom oxygen concentration and sea temperature at 15 m depth in Anc6n Bay, Peni, during the period 1981 - 1987. Shaded areas: EN 1982-83 and EN 198687. clear EN waters to a greater depth than the turbid upwelling waters found during normal periods. Nearshore in northern Peni and off Ecuador, where torrential rainfall increased the river discharge, turbidity and sedimentation of coastal waters were strongly elevated during EN, and locally the salinity of surface waters was reduced by about 1 O/OO (Leetma et al., 1987). Generally, however, salinity changes in either direction -- clearly detectable as inflow of equatorial or oceanic waters -- occurred at a lower range, and did not seem to have as important an effect on the upwelling flora and fauna (see below) as did changes in, for example, temperature, sea level and dissolved oxygen (Arntz, 1986; Arntz and Tarazona, 1988). Dissolved oxygen was reduced somewhat in the surface waters off Peni, from normal values > 6 ml l-1 to values around 5 ml l-1, whereas at depths of about 200 m the intruding waters induced a strong 0 2 increase (Fig. 2) in areas that are normally hypoxic or even anoxic (Rosenberg et al., 1983; Arntz et al., 1985; Tarazona et al., 1985a). A comparison of R.V. "Humboldt" data from the cruises 8103/04 (1981, normal conditions) and 8212/8301 (1982-1983, EN) indicated a 3 to 7 fold increase in dissolved oxygen at the seafloor below 50 m off northern and central Peni during EN. At depths c 100 m, 0 2 values often exceeded 3 ml l-1, and between 100 and 200 m they reached 2.5 ml 1-1 (Arntz et al., 1985). Increased
329 oxygenation during EN has been observed all along the Peruvian continental shelf (Fig. 3); there are, however, marked differences between strong and weak events (GuillCn et al., 1985). In contrast to these conditions, surface waters off northem Chile during EN 1982-83 revealed 0 2 values < 1 ml l-1 (Alvial, 1985; Fonseca, 1985; Fuenzalida, 1985).
3 THE PELAGIC SUBSYSTEM 3.1 Phvto- and zooulankton The deepening of the thermo- and numclines, which hampered the transport of numents into the euphotic layer, the strong warming of the surface waters and possibly also the slight reduction of dissolved oxygen in the uppermost offshore layers, induced important changes in the pelagic system. The most obvious consequences were a drastic reduction of biomass and production of phytoplankton, equally strong changes in species composition, and a "tropicalization" due to the transport of equatorial and oceanic species poleward and towards the coast (Barber and ChBvez, 1983; Avaria, 1985; Muiioz, 1985; Rojas de Mendiola et al., 1985).
1
'
1
'
I 3-
1
'
-
1
'
1
'
1
'
1
'
2-
1 10
t-
I
3
1981
I
5
\ : ; \ I
I
7
I
1
I
I
9
LATITUDE
11
( O
I
I
13
1985 I I
I
15
S 1
Fig. 3. Latitudinal changes of temperature and dissolved oxygen close to the seafloor at 91 140 m depth during EN (1982-83, 1987) and non-EN years. Peruvian coast, summer values only with the exception of 1982 when spring (November) data were included also.
17
330 According to Peruvian and Chilean net plankton samples during EN 1982-83, the small diatoms characteristic of the upwelling region gradually disappeared and were replaced by (sub)tropical dinoflagellates, coccolithophorids and large diatoms. The extent of these changes depends on the strength of the EN event; during EN 1982-83 it occurred over > 15 degrees of latitude and in a coastal fringe comprising several hundred kilometers. The retreat of the autochthonous diatoms, associated with a reduction of phytoplankton biomass, occurred stepwise (Avaria, 1985; Avaria and Muiioz, 1987). Plankton recovery in most areas off Peni and northern Chile began in September 1983 (Avaria et al., 1988). However, plankton net catches failed to record the minute (picoplankton to nanoplankton) single-celled species that have recently been found responsible for a very large part of the carbon production in tropical seas (Joint, 1986; Stockner, 1988). Although we have not found any literature referring specifically to EN 1982-83,these small organisms must have been a major component of the (sub)tropical plankton community that replaced the traditional upwelling community during this event. The changes at the base of the food web -- reduced food availability and exchange of the autochthonous phytoplankters for both minute and larger species equally unsuitable as food -influenced the higher levels of the food chains, i.e. zooplankton and pelagic fish feeding on plankton, in an unfavorable way. At the same time, the zooplankton and pelagic fish species of the upwelling system were affected themselves by the increased water temperatures. Within these groups, too, the local species were displaced or suffered from mortality, and were replaced by tropical species. In particular there was a change from herbivorous copepods, the biomass of which was reduced to about one-sixth the value before EN, to chaetognaths and other large, predatory organisms such as salps, jellyfish and siphonophores (Tsukayama and Santander, 1986; Carrasco and Santander, 1987). In many cases, even these large forms could not make up for the biomass loss caused by the disappearance of the autochthonous small copepods; since most of the invading plankters were voracious predators, they may even have contributed considerably to the general shortage of food. The reduction of the local holoplankters and the increase of tropical holoplanktonic organisms were accompanied by a reduction of the pelagic larvae of benthic organisms (meroplankton), which spend only part of their lives in the water column. At the same time tropical meroplankton, such as the larvae of shrimps, extended their areas of distribution over several degrees of latitude towards the south (Tarazona et al., 1985b; Carrasco and Santander, 1987). In Anc6n Bay, Peni (ca. 12"N) meroplankton increased before and during the initial phase of EN (Fig. 4) due to the intrusion of large numbers of gastropod larvae. During the first peak of the thermal anomaly, the abundances of gastropod larvae collapsed, and meroplankton densities were low. After EN, bivalve larvae became prominent for some time, and the rapid recovery of intertidal mussel populations followed (see below).
3.2 Pelagic fish Previous EN events to some extent favored the sardine (Surdimps sagax), which contrary to the anchovy (Engruulis ringens) tends to avoid the cold water plumes of the upwelling
33 1
m
5.0
LD
cu-
4.0
0
3.0
E
o g m -
> L
m
2.0
f-l
U .?I
K
0 CI
Y K
m
1.0 12.0
10.0 8.0
6.0 4.0
2.0 0.0 Apr-81
Nov-81
May-82
Dec-82
Jul-83
Jan-84
Aug-84
Fig. 4. Changes in numbers of total meroplankton (above) and of planktonic gastropod and bivalve larvae (below) at a shallow water station in Anc6n Bay between 1981 and 1984. The values correspond to 10 vertical hauls between 15 and 0 m depth taken by standard net and pooled, mesh width 200 pm. centres (Zuzunaga, 1985). However, the 1982-83 event had negative biological effects on both pelagic fish species of greatest commercial importance in the Humboldt Current area. Off Ecuador, relative values of sardine abundance based on monthly standard catches dropped sharply between October 1982 and August 1983 (Menz, 1986) whereas the sardine biomass estimated by echoacoustics off Peni increased between March and April (Santander and Zuzunaga, 1984), leading the latter authors to the assumption that part of the sardine stock must have migrated southward. Off Peni, towards the end of 1982, both sardines and anchovies reacted to the warming and the impoverishment of the epipelagic zone with migrations of different kinds (Zuzunaga, 1985; Amtz, 1986). Part of the shoals withdrew to the remnants of upwelling centres, which at that time still contained some cool, nutrient-rich water close to the shore (cf. section 2). Continued deepening of the thermocline, however, converted these areas into traps, and the fishery took large catches from the densely aggregated shoals as it had done in 1972, when on a single day 170,000 t of anchovies were caught by the Peruvian purse seiner fleet (J. Valdivia, 1978). Another part of the anchovy and sardine populations withdrew to deeper water, often > 100 m depth, where there was little food and a lower oxygen content than there is normally at the surface, but where the temperatures were also lower than in the epipelagic zone (Santander and Zuzunaga, 1984; Amtz, 1986). A third group, especially
332 the sardines, avoided the unfavorable conditions off Peni by migrating even further southward where the impact of EN was less dramatic (Caiibn, 1985). Off northern Chile, where the temperature increase was lower, the sardines remained close to the surface and to the shore (Martinez et al., 1989). There, the feeding conditions were more favorable than at greater depth and further offshore, but they could be taken by the purse seiners, which fish only in the uppermost 50 m (cf. section 3.3). The biological effects on the anchovy off Chile seem to have been similarly as negative as those off Peni, judging from the extremely low catches during 1983 (FAO, 1989). Tagging experiments have been carried out on a relatively small scale; the few data from sardines that were tagged off Peni and recovered later off Chile confirm that southward migration does indeed occur (Torres et al., 1985). The disruption of the pelagic food web and the withdrawal of the pelagic shoaling fish into areas that were almost devoid of food, warmer and less oxygenated than their normal habitat were reflected in changes in stomach contents, poor condition, reduced growth, reduced spawning activity and extremely poor spawning success of both anchovy and sardine. Instead of feeding on their normal food, i.e. small herbivorous copepods and diatoms, sardines off Peni in 1982 mainly fed on subtropical camivorous copepods and dinoflagellates, which occurred in much lower densities, and even fed on fish (Alamo et al., 1988). The average stomach contents were greatly reduced, which led to weight losses of 10 - 20 %, and in extreme cases up to > 30 %, in the sardines in Peni (Dioses, 1985) and weight losses of around 20 % in adult sardines in northern Chile (Martinez et al., 1984). Poor sardine condition off northern Chile was recorded throughout 1983 (Mujica et al., 1985; Alcocer and Kelly, 1987). The consequence was stagnation of growth (Aguayo et al., 1985; CCdenas and Chipollini, 1988), reduction of lipids in the body tissues (Caiibn, 1985; Romo, 1985), and reproductive failure (Pastor, 1984; Santander and Zuzunaga, 1984; Retamales and Gonzlilez, 1985) both off Peni and Chile during 1983. However, the sardines demonstrated partial recovery towards the end of that year and exhibited normal spawning in 1984. Anchovies off Peni lost 30 % or more of their weight (Santander and Zuzunaga, 1984). A complete breakdown in reproduction of the Peruvian part of the anchovy stock occurred throughout 1983 (Santander and Tsukayama, 1984). No published anchovy data seem to be available over this period in Chile. The other two important shoaling fish species of the Humboldt upwelling area, horse mackerel (Truchurusmurphyi) and Spanish mackerel (Scomberjuponicus peruanus), live further offshore than anchovy and sardine under normal conditions. The occurrence of mackerel larvae is positively correlated with SSTs (Muck et al., 1987). During EN the two species approached the shore all the way from Ecuador to Chile. Nearshore migrants of S. juponicus were reported as far south as Puerto Chacabuco (45'30's) in Chile where in midJanuary 1982 a shoal of "caballa" (Pacific mackerel) suffered massive mortality close to the shore presumably from a combination of low 0 2 concentration, low salinity and high temperature (Zamaet al., 1984). Mackerels, like anchovies, often stayed beyond the reach of seines as they remained at greater depths (Vflchez et al., 1988). Neither species suffered from a scarcity of food because they inhabit oceanic waters that are much poorer in food than up-
333 welling areas, and they feed mainly on fish (Konchina, 1982; Muck and Sinchez, 1987). Fish (weak anchovies and sardines) were not scarce during EN in deeper water. Accordingly, horse and Spanish mackerels lost at most 10 % of their body weights (Dioses, 1985), spawning was normal, and the density of larvae even increased (Santander and Zuzunaga, 1984). A large number of predatory tropical fish invaded the pelagic zone of the Peruvian-Chilean upwelling area in 1983 (Hoyos et al., 1985; Kong et al., 1985; VClez and Zeballos, 1985), among them sierra (Scomberomorus sierra), skipjack (Karsuwonuspelamis), yellowfin tuna (Thunnus albacares), dolphinfish (Coryphaena hippurus) and different species of oceanic sharks. Bonito (Sardu chiliensis chiliensis and, possibly also, S. orientalis , which may have extended its range to the south during EN; see Pauly et al., 1987), which had virtually disappeared after the decline of its principal anchovy food, in the first half of the seventies, once again returned to Peruvian coastal waters. All these species may have contributed substantially to an increase in predation on shoaling fish. Most of the invaders disappeared from the upwelling area when the water temperatures returned to normal. Close to shore another pelagic fish species, the silverside (Odontesthes regia regia), which is important in the artisanal driftnet fishery, virtually disappeared from Peruvian waters in January 1983 and did not return until 1985 (cf. Fig. 14, section 4.3). It is not known if these fish migrated southward or withdrew to deep water. 3.3 Pelagic fisheries The effects of EN on the pelagic (i.e. industrial) fisheries were predominantly negative, as can be seen from the reduction of Latin American fish catches by about 3 million t (metric tons) as compared with 1982 (when they were 11.6 million t), a decrease in fish meal production of 64 % in the first 10 months of 1983, and the collapse of fish oil production in that year. However, there were important regional differences. Off Ecuador, the purse seine fishery on Etrumeus teres (Round herring) and other small fish collapsed in 1983; however, 25,000 t of horse mackerel were caught, which in normal years conmbutes minimally to the catches. Off Peni, landings of pelagic fish by the purse seine fleet declined from 3.3 million t in 1982 (of which 1.7 million t were anchovies and 1.5 million t were sardines) to 1.4 million t i n 1983 (1.1 million t sardines and 0.1 million t anchovies; the latter from the first quarter of the year and close to the Chilean border). In 1984 the Peruvian anchovy catches were practically nil, whereas the sardine catches doubled to 2.8 million t. In 1986, the anchovy landings were 3.5 million t and similar to the 1974-1976 level. The sardine catches dropped to only half the landings of the previous year. The horse and Spanish mackerel landings increased during and after EN 1982-83, reaching 270,000 t in the post-EN year 1984, but they never approached the 1977-78 record catches of nearly 0.5 million t (Fig. 5 ) . The Peruvian anchovy landings indicate that each EN after 1960 has had a marked effect on the availability of the stock to the purse seine fishery; even warmer years like 1969 and 1979, which are not considered EN years according to strict thermal criteria, caused a decline in the catches (Fig. 5). However, the landing statistics also reveal that the Peruvian anchovy catches were at a very low level before the appearance of EN 1982-83. An estimate of the Peruvian
334 .. .. .. .. .. .. ., .. .. ... ... ... ... ~
... ... ,.. .
1:::
W 0
a
0
H
0.5
z
j;j;i
. .. .. .. .... .. .. .. ..
... ... ... ...
X
c,
0
... ... ... ...
.... .. ... ,..
d
Z
... ... ... ..
... ... ...
-
m
... ... ... ... ... ... ... ... ... ... ... ... . . .. .. .. .. ... ... ... ... ... ...
a A
0.4
_I
a
3 Z
z a
0.3
0.2 0.1
n 1951 1955 1959 1963 1967 1971 1975
1979 1983 1987
Fig. 5. Annual landings of the principal pelagic shoaling fish off Per6, 1951-1987. Shaded areas: EN and other years with positive temperature anomaly > 2'C. spawning population of anchovy in 1981, based on the "egg production method", revealed an extremely low value of < 2 million t (Santander et al., 1984) indicating that an excessive share of the stock was being taken at that time. Northern Chile's anchovy catches started in the early sixties and were around 1 million t, revealing a certain impact of larger EN as occurred in Peni, until they dropped to a very low level during EN 1972-73 (Jordh, 1983). During the past five years they contributed < 5 % to total Chilean pelagic catches except in 1986, when they reached a record low value (33 %; Martinez et al., 1989). However, in 1987 catches declined again during the occurrence of another EN. Low anchovy catches per unit of effort during EN years (Serra, 1986) indicate reduced accessibility of the stock to the fisheries during these times, as occurred in Peni (see section 3.2). Sardine (S. s a g a ) catches in Ecuador started to decline in October 1982 and remained at a very low level until August 1983 (Jimknez and Herdson, 1984). After EN 1982-83, landings increased sharply despite the fact that the size of the purse seine fleet remained the same (Maridueiia 1986 fide S e r a and Tsukayama, 1988). In Peni, sardine catches also declined in October 1982, a consequence of reduced availability to the fisheries. They reached their lowest values in June 1983. Catches per unit of effort are not available, but the size of the industrial fleet (= 350 purse seiners) remained about the same during 1981 - 1985 with a reduction of
335 one-third in 1984. Interestingly, the landings in 1983 and 1984 increased steeply (Fig. 6) showing that large quantities of sardines had again become available to the fishery almost immediately after the return to normal conditions. Off Chile, the southward migration of the sardines, their nearshore concentration and the presence of shoals in more superficial waters favored purse seining contrary to the situation in Ecuador and Perli. As a result, Chile became one of Latin America's most important fishing nations. From the 3.9 million t catch in 1983,2.8 million t were sardines (1.0 million t more than in the preceding year). From this total, 3.2 million t were processed into fish meal and oil.
3.5
-
3
W
0 4
X
2.5
c,
v
Ln
a
2
Z
H
0
Z
a
1.5
_I _I
a
3
1
Z
Z
a 0.5 0 1964
1968
1972
1976
1980
1984
1988
Fig. 6. Annual landings of "sardina", S. sugar, in the countries of major EN impact, 1964 1987. Shaded areas: EN and other years with positive temperature anomaly > 2'C.
Despite these immediate positive effects on the sardine fishery, the medium-term effects were unfavorable. The catches declined by 33 % in 1987 due to continued overfishing, and because the 1983 yearclass, which should have entered the fishery that year, was absent (Martinez et al., 1989). The 1983 year class recruitment failure was confirmed by a population analysis carried out by Martinez and co-workers. Overfishing was caused by a tremendous increase (850 %) in the number of vessels and the modernization of the fishing fleet. Accordingly, catches per unit of effort of sardines that had increased until the first half of 1983 declined by 58 % until 1988 (Martinez et al., 1989). The horse mackerel catches off Chile dropped from 1.5 million t in 1982 to 0.9 million t in 1983 because the fleet concentrated on the easily accessible sardines, and because part of the
336
horse mackerel stock migrated northward. Tropical immigrants such as dolphinfish and skipjack became locally important items of the artisanal fishery during EN, especially off Peni, but almost immediately disappeared from the area when the sea temperatures returned to normal. In 1983, dolphinfish catches reached over 30 t in Chimbote during several summer months, and > 100 t in Callao in May; skipjack landings in the same port were 113 t in April (Vtlez and Zeballos, 1985). Silverside catches, with > 20 t still landed in the port of Callao in December 1982, collapsed in January 1983 for two complete years. Apparently similar shifts occurred during former EN events although dolphinfish never were as abundant off Peni as in 1983 (Del Solar, 1983). 4 THE BENTHIC SUBSYSTEM 4.1 Macrobenthos 4.1.1 Hard bottoms and rockv shores Most of the information available on the impact of EN 1982-83 and 1986-87 on rocky shores is based on qualitative observations; only for hard bottom communities at Anc6n (1 1'46's) and Independencia Bay (14O15'S) are quantitative time series data available (Tarazona et al., 1985b, 19884 Romero et al., 1988). Unfortunately, quantitative sampling has not been continued after these EN events. In the following, reference to an unqualified recent EN refers to the severe event of 1982-83. Under normal, non-EN conditions the intertidal and shallow subtidal zones of the rocky shores of Peni and northern Chile are dominated by rich populations of algae, mytilids and balanids, which compete for the available space. The floral and faunal community smcture seems to be controlled by grazers (sea urchins, chitons, limpets and other gastropods) and predators (sea stars, brachyuran crabs and fish), respectively, resulting in a rather stable balance and the absence of space monopolization by just one group of organisms (Castilla, 1981; Hoyos et al., 1985; Tokeshi et al., 1988, 1989a). Both the mytilid associations and large brown algae -- particularly LRssonia nigrescens and in deeper water "forests" of Macrocystis pyrifera -- provide many niches and refuges for numerous associated species. About
150 animal species live associated with the rhizoids of Lessonia spp.; they can reach densities of up to 8,000 and about 200 g wet weight per rhizoid (Romero et al., 1988). and over 90 species are associated with two, densely packed intertidal mussels Semimytilus algosus and
Perumytilus purpurutus (Paredes and Tarazona, 1980; Tarazona et al., 1988d; Tokeshi et al., 1989b). During EN 1982-83 the balance between the different components of the hard bottom ecosystem was upset. The principal reason for these changes, unlike the changes observed in many soft bottom communities below the intertidal zone, was not due to an increased 0 2 concentration during EN because 0 2 is not limiting in these kinds of communities under normal conditions. Instead, most changes seem to have been caused by a combination of high temperatures, changes in sea level and increased swell, and from biological interactions that resulted from the impact of these physical perturbations. The changes along rocky shores during the first phase of the event involved the mass mortalities of key species, leading to a
337
general impoverishment of the communities in terms of density, biomass and species numbers (Soto, 1985; Tarazona et al., 1985b; Tomicic, 1985). This disturbance was accompanied by predators immigrating from tropical areas (mainly swimming crabs). At a later stage, when ample space was available due to the mass mortalities of species that had formerly occupied the rock substratum, algae increased dramatically in abundance in the intertidal zone and the algae were able to monopolize the available space because of the absence of grazers. Apparently, grazer populations also needed more time to recover than their food: while the algae, especially Ulva cosrara, developed prominent growths in Laynillas (south of Pisco) from about May 1983, the first juveniles of limpets and other snails became visible only as late as December 1983 (Arntz, 1986). Commercial sea urchin stocks had not even recovered by 1988 (Wosnitza-Mendo et al., 1988). In Anc6n Bay, increases in abundance of U . costafa started in March 1983, when sea level increased, and extensive bleaching of the green alga occurred in October, after sea level had returned to normal. In Anc6n Bay, U . cosfata increased in abundance starting in March 1983 when sea level was anomalously high. The green alga also experienced extensive bleaching in October, after sea level had returned to normal. Recolonization by Sernirnyrilus on rocks left bare by the impact of EN in this area (Fig. 7) started only in October 1984 (Tarazona et al., 1988b). In the deeper intertidal and shallow subtidal zones, kelp suffered almost complete mortality during the first months of 1983, and most of the species associated with the rhizoids and foliage of the laminarians were also killed. The post-EN period was characterized by a multitude of biological interactions among the various components of the intertidal hard bottom community (Tarazona et al., 1988d), finally resulting in the reestablishment of the usual set of species, the gradual disappearance of the invaders and the reduction of algal cover and biomass to normal levels. This process took about two years in the intertidal zone. In the subtidal zone, however, recovery occurred much later since the kelps started to recolonize their former habitat only after about three years, and they may still need another couple of years to grow to their former size. At an intertidal location at Ancbn, the normally dominant mytilids were reduced to about 5 % of their former density during the first phase of the 1982-83 EN event. Polychaetes and brachiopods, which normally live in the deeper sublittoral areas, became predominant together with green algae (Ulva costata , which replaced U . lacruca) and tropical species such as the stalked barnacle Pollicipes elegans, which recruited by larval settlement (see section 4.2). Towards the end of EN, several benthic invertebrate species and algae coexisted without any one species monopolizing the available space, and evenness reached an unusually high value for this community (Tarazona et al., 1985b). Apparently, the monopolization of the substrate by a few species, which is typical for these kinds of communities, can be replaced by species assemblages with more even space distributions during EN conditions. However, at other sites where even higher temperatures prevailed, all zoobenthos suffered high mortalities, with the result that algae monopolized the substrate. Alternatively, in some exposed areas populations of stalked barnacles that had settled during the event became community dominants only after EN 1982-83 (Arntz, 1986; Kameya and Zeballos, 1988). Another study site in the Anc6n area, at 5 m depth, was dominated by populations of the
338
4-
I I
E
In
m s o + m
\z L
3' -
2-
g: E d 3
Z
-
l-
06N
E
d
0 -
. d
5-
4-
o +
\z m
3
L
rnm
2:
2
3
Z
Ex
0
May-82
Ju 1l-83
Jan'-84
'
AUCJ-84
Fig. 7. Density changes of planktonic bivalve larvae (above) and the mussel, Semimytilus nlgosus (below), in the rocky intertidal of Ancdn Bay during and after EN 1982-83. On the left side, the lower graph presents data from a field experiment (Tarazona et al., 1985b) where colonization was monitored on rock surfaces cleaned of their mussel cover. On the right side, natural recolonization is indicated on rock surfaces freed by EN 1982-83 (Tarazona et al., 198Xd).
mytilid Aulacomya ater and the laminarian Macrocystis pyrifera before EN. Both species died during the event. Experimental substrates were almost totally colonized by the polychaetes Hydroides norvegica and Pomatoceros sp., which occupied nearly all of the space released by the former two species. The reason for the success of the polychaete worms may have been that they were more tolerant of high water temperatures and could not be eaten by the immigrant tropical predators, principally swimming crabs and shrimps. Interestingly, the two tubicolous polychaetes, which normally live on offshore islands, appeared in large numbers in layers 5 cm thick on the hulls of fishing vessels as early as December 1982 whereas they colonized nearshore rocks several months later (Tarazona et al., 1985b). Possibly these vessels acted as dispersion agents for the two polychaete species, which have a short planktonic stage (Thorson, 1946; Wisely, 1958). Generally, the impact of EN on the rocky shores of the northern Humboldt Current area seems to have been predominantly negative, leading to profound changes in the biotic
339 composition and structure of the communities. However, in northern Chile the impact of EN 1982-83 on the rock fauna seems to have been less negative, in some cases even beneficial to certain populations. An example is the mytilid, Semimytilus algosus, which developed favorably in the Iquique-Antofagastd area where the changes brought about by this EN were not so drastic as in central Peni (Soto, 1985; Tomicic, 1985). The implications for commercial invertebrate species are dealt with in section 4.2. 4.1.2 Soft bottoms Soft bottom habitats in the Humboldt Current area include intertidal sandy beaches, shallow subtidal zones, where the sand contains an increasing amount of silt, and the mud bottoms of the oxygen minimum zone, which are hypoxic or even anoxic under normal conditions. EN effects were different in these areas, and the three zones will be dealt with individually for this reason. For more oxygenated habitats below the 0 2 minimum zone, extending from about 700 m to the deep sea (Rosenberg et al., 1983), we have no data. Sandy beach communities in the intertidal zone of Peni and in northern Chile are composed of only a limited number of macrobenthic species (about 30), but they often have high population densities and high biomasses due to the relatively large size of their dominant species, e.g. the surf clams Mesodesma donacium and Donarperuvianus, and the mole crab
Emerita analoga. Four years prior to EN, in Asia, Peni (south of Lima) the average density of the macrobenthos was 3,720 individuals m-2 and the mean biomass was 15.1 kg m-2 wet weight (Tarazona et al., 1986). Both the density and biomass of M . donacium can be even higher at times (see section 4.2). During EN 1982-83, the three numerically dominant species of sandy beaches suffered high mortalities (Arntz et al., 1987). While D. peruvianus and E . analoga survived at low densities, M . donacium, which had exhibited changes in reproductive behaviour in the months before EN, did not survive the high temperatures in February 1983 and became absent in central Peni (Fig. 8). When conditions normalized, small polychaetes of the genera Dispio and Scolelepis became pre-eminent. D . peruvianus and E. analoga increased in density but never reached the dominant role of M . donacium before the event. At this writing (1989), the intertidal sand beach community has not fully recovered and is far below the high biomass values observed before EN 1982-83. Seafloor communities at moderate depths along the South American coast, which are often affected by hypoxic conditions (Gallardo, 1963; Ramorino and Muiiiz, 1970; Gallardo et al., 1972; Rosenberg et al., 1983; Tarazona, 1984; Tarazona et al., 1985a, 1988c), responded to EN in quite a different way. Our observations are based principally on data from two stations at 15 and 34 m depths in Anc6n Bay (Tarazona et al., 1988b). Two months before the front of the Kelvin waves arrived in the first week of October 1982, dissolved oxygen levels increased close to the bottom and maintained higher than normal values until May 1984, nearly one year after the return to normal temperatures (cf. Fig. 2). Several community attributes more or less followed these changes. Between June and July 1982 the number of species increased; between May and September 1984 the number of species decreased and finally returned to the
340
.
,I
B
I
n
a
I I
8,000
u)
I
U
I I
-
Mean and standard error
EL NIRO 1982-83
I
.
I
j
No M. donaclum found
Mean only (no replicate available)
Fig. 8. Density changes and mortality during EN 1982-83 of the surf clam, Mesodesma donacium, in the intertidal zone of the Santa Maria del Mar sandy beach, 1981 - 1983. Vertical lines denote standard error of the mean. Modified after Arntz et al. (1987). low level found before EN. Density and biomass of the community started to increase in October 1982, these attributes decreased to some extent during the periods of maximal thermal anomalies in January and May 1983, and they returned to normal low values between May and August 1984. At 15 m, the species number doubled during the event, the biomass rose from < 1 g dry weight to > 18 g m-2, and total individual density increased from < 4,000 to > 40,000 m-2. At the 34 m station, species number increased by a factor of 5 (Fig. 9), biomass--normally at 0 g or negligibly above--rose to 7 g dry weight, and densities climbed from about 400 to > 13,000 m-2 (Fig. 10). At both stations, species diversity (H) was about 2 - 3 times higher during than before the event. Trophically, there was a change from a dominance of deposit to suspension feeders at the shallower station, and a species replacement among the deposit feeders at the deeper station (Tarazona et al., 1988b). The favorable development of the benthic communities in Anc6n Bay seems to have been caused mainly by the increased 0 2 concentration at the seafloor during and after the event, possibly in connection with higher temperatures (Tarazona, 1984). Under these conditions, colonization on the otherwise oxygen deficient bottom is improved both for autochthonous species and immigrants from tropical or oceanic waters, and growth and production of all faunal components are accelerated. The flexibility of this subsystem, i.e. its capacity for immediate response to a change of environmental conditions, is surprising. Information on the deeper parts of the oxygen minimum zone on the shelf and the upper continental slope is more fragmentary, but the marked increase in dissolved oxygen referred to in section 2 apparently had beneficial effects on macrobenthos whereas it negatively affected the
341
24 N
E
20
N d
0
\
16
m
a, .rl
U
a,
n cn
12
Y-
O
8
L
a,
n
E 3
Z
4 0 Aug-81
Feb-82
Sep-82
Mar-83
Oct-83
Apr-84
Nov-84
Fig. 9. Changes in total number of macrobenthic species and appearance of "new" species (n) before, during and after EN 1982-83 at a 34 m deep station in Anc6n Bay .
1981
1982
1983
1984
Fig. 10. Density changes of macrobenthos at 15 and 34 m deep stations in Anc6n Bay before, during and after EN 1982-83. Vertical lines denote standard errors of means. From Tarazona et al., 1988b.
342 normally dominant spaghetti bacteria (Prokaryota, genus Thioploca) (Amtz et al., 1985; Salzwedel et al., 1988). There were, however, geographical differences: in northern Perh, where 0 2 values at the seafloor are high during normal periods, hardly any changes were observed, but between 10°30 and 12"spolychaetes and nemerteans increased in biomass and population density. We do not know why such positive effects were not evident among some mollusc species, but predation by tropical immigrants may have concealed it. Immediately after EN, increased densities of molluscs were observed in some areas (Amtz et al., in prep.). Species number and diversity of macrobenthos increased all along the coast, even north of 1Oo30'S. The favorable development of the macro-benthos, together with the improved 0 2 conditions, also affected the behavior of demersal fish (see section 4.3). In normal years, swimming crabs do not occur in the Humboldt Current area south of Paita (5"s) but are restricted to the extreme north of Peni, Ecuador and Colombia. In December 1982, large numbers of juveniles of the species Euphylax robustuy and Portunus acuminatus were taken by RV "Humboldt" off Huacho (1 1"s). Within a short time, these two species and three others (Portunus asper, Callinectes arcuatus and Arenaeus mexicanus) appeared all along the Peruvian coast, with C. arcuatus even present in northern Chile. A sixth species (E. dovii), which had been common during EN 1972-73, appeared on the Peruvian coast in 1984 (Arntz and E. Valdivia, 1985; Amtz, 1986). Many swimming crabs spawned during EN in their new habitat and thus enlarged their populations. Bycatches of several hundred kg per night were taken in the shrimp trawl fishery off Chimbote towards the end of EN, and in the shallow waters of Paracas Bay densities of 100 individuals m-2were recorded (Amtz, 1986). When the water temperatures returned to normal, most populations in Peruvian waters were killed but some survived through 1984. In Anc6n Bay A. mexicanus, the only species for which we have a time series, managed to reproduce once more under post-EN conditions (Fig. l l ) , but recruitment was not successful, and the population suffered total mortality soon afterwards. Swimming crabs, besides being a nuisance to the fishermen, are voracious predators and may have been largely responsible for the devastation of benthic fauna during EN both in shallow and deep waters. In Ancdn Bay the food of A. mexicanus during EN consisted mainly of mole crabs (E. analoga), which were taken by all predators both large and small. Some mussels, e.g. S. ulgosu.7, were eaten also, revealing that the swimming crabs took their food both from soft and hard bottoms. After EN, juvenile swimming crabs nearly exclusively fed on small S. algosus, which had become available again in large quantities. 4.2
Exdoited invertebrates ("mariscos") The profound changes in the benthic subsystem described in section 4.1 also affected the commercially exploited invertebrate species. Negative or positive EN effects often become more clearly visible in shellfish populations, even from catch statistics that for some of these species are the only available information (Figs. 12, 13). Certain factors, however, (e.g. changes in the preference for certain shellfish by fishermen and the associated shifts in the distribution of the artisanal fishing fleets, or closed seasons) sometimes conceal population changes. Since catches per unit of effort are seldom available, the only possible approach is to
343
1.5 1.2 0.9 0.6 0.3
0.0 400
300
Arenaeus mexicanus
200
100 0 Apr-81
Nov-81
May-82
Dec-82
Jul-83
Jan-84
Aug-84
Fig. 11. Changes in numbers of brachyuran zoeas (above) and the swimming crab, Arenaeiis mexicanus (below), a tropical invader, in Anc6n Bay, 1981 - 1984. The upper graph pools 10 standard net hauls, 0 - 15 m, mesh size 200 pm; the lower curve presents values from standard beach seine catches taken alongshore in the shallow subtidal zone (see Hoyos et al., 1985). use a combination of observations along the shore with landing statistics. EN effects on "mariscos" off Peni have been discussed extensively by Arntz and E. Valdivia (1985) and Arntz et al. (1987, 1988). The sorts of effects observed during EN 1982-83 and former events during this century (cf. Chirinos de Vildoso, 1976, 1984 and Del Solar, 1983) included (a) mass mortalities, (b) immigration from (sub)tropical areas further north combined with area extensions and southward shifts of certain populations, (c) emigration from shallow areas towards greater depths, (d) increases in abundance of local populations more tolerant of high temperatures and capable of using alternate sources of food, and (e) after EN the temporary survival of exotic species that managed to recruit locally during the event. Mass mortality occurred on sandy beaches, in the rocky intertidal zone and in shallow subtidal areas. This disturbance mostly affected mollusc and crustacean populations, but also sea urchins and ascidians (the latter are used as fishing baits in P e h and for direct human consumption in Chile), and it seems to have been connected mainly to the unusual increase in sea temperature. Other factors that may have contributed to these mortalities were changes in sea level, rough seas, the destruction of kelp stocks that provide shelter for many invertebrates
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
Fig. 12. Annual landings of three shellfish of major importance (scallop, A . purpuratus; mussel, A. ater; and shrimps of the penaeid family) off the Peruvian coast, 1970 - 1987. Shaded areas: EN and other years with positive temperature anomaly > 2°C. Modified after Arntz et al. (1988). (see Dayton and Tegner, this volume), and the decline in other algae that serve as food for grazers. In northern Peni and Ecuador, where torrential rainfall increased river discharge (and thus sedimentation in coastal waters), decreases in salinity and increases in sedimentation rates also occurred. All mussel beds of Aulacomya ater, which before EN 1982-83 was the most important shellfish species in Perli above 15 m depth from Huacho to Pisco (1 1 - 14OS), died and detached from the substrate in March 1983 although this species demonstrated accelerated growth until February. Only in deeper water (> 30 m) did some populations survive (Soenens, 1985). The destruction of mussel beds induced important population shifts among many species that normally live in this habitat. The shallow water clams Semele spp. and Tagelus dombeii, together with other bivalves of minor commercial importance, became virtually extinct in some areas whereas another clam (Cari solida) living at a greater depth survived and contributed to fishermen's catches in the later phase of EN. Similar depth differences in survival were observed in crab populations. In the case of Cancer spp., C . setosus suffered almost complete mortality similar to another commercially important species (Platyxanthus orbignyi) (Fig. 13) whereas C .porteri and C. coronatus withdrew to deeper water and reappeared in catches half a year after EN (Amtz and Arancibia, in press). On Peruvian rocky shores, thick layers of shells of many grazers, including limpets (Fissurella
345
.............. .............. .............. .............. .............. ..............
40
i:: A
100 80
=
6o 40
e 1
h
25
3 2o
5 15
a 10 5
1981
D
1982
D
1983
D
1984
D
1985
D
Fig. 13. Monthly landings of brachyuran crabs (Cancer setosus and Platyxanthus orbignyi) and clams of the "almeja" type (Semele spp., Gari solida and others) at the port of Pisco, Peni, 1981 - 1985. Modified after Arntz et al. (1988). spp.), other prosobranch gastropods, chitons and sea urchins, were found in March 1983, while the substrate, normally covered with a rich invertebrate fauna, was practically bare. The disappearance of grazers subsequently led to the proliferation and succession of algae, which, unfortunately, were never studied in detail (Amtz, 1986). However, on Chilean rocky shores mortality of invertebrates was much less intense, and some species even benefited from the EN conditions, e.g. Semimytilus algosus (Soto, 1985) and Acanthopleura echinata (Pefia et al., 1987). On sandy beaches, the surf clam Mesodesma donacium, an easily accessible species for local exploitation, which before EN in some areas produced > 35 kg m-2 (3 kg shell-free dry weight) (Amtz et al., 1987), did not survive along 7 degrees of latitude on the central Peruvian coast at its northern most distribution. At the same time, however, it increased in abundance near its southern distribution limit in the area of Valdivia, Chile (40's). M. donacium is one of the few shellfish species for which population data were available before, during and after EN (cf. Fig. 8). Immigration from tropical and subtropical areas near the equator included both active invasion by juveniles and adults, and an inflow of eggs and larvae of invertebrate species that later became established in the upwelling area off Pert5 and northern Chile. In many cases the immigrant species even reproduced. For most species, it is not possible to distinguish between adult, larval or post-colonization cohorts on the basis of the available data. Only among the crustaceans did immigrant species attain commercial significance although among the molluscs there appeared some spectacular large species such as pearl oysters (Pinctada sp., Pteria sterna), pen shells (Arrina maura) or large gastropods such as Maleu ringens. By far the most important invaders by commercial standards were the penaeid shrimps (cf. Fig. 12) and among these, the species Xiphopenaeus riveti, which in normal periods abounds off Colombia and Ecuador. The first specimens of penaeids appeared in the bottom trawl catches of the Peruvian research vessel "Humboldt" off Huacho as early as December 1982. During 1983 X . riveri extended its range by 13 degrees latitude down to the Peruvian south coast and gave rise to a
P
346
new fishery both with beach seines along the shore and with trawls and trammel nets in deeper water. Even after the pronounced temperature decline in July 1983, catches per boat per night south of Chimbote were between 0.5 and 3 t. During 1984 the shrimp boom off Peni gradually ceased, and towards the end of the year the fishery was once again resmcted to the Paita (5's) area and further north. Some of the larger penaeid species maintained small stocks, e.g. south of Pisco (14OS), where they were caught incidentally in purse seines as late as 1985. Off Colombia and Ecuador the abundances of shallow water shrimps, which were mainly responsible for the invasion in Peni, decreased (as did those of the swimming crabs that also invaded Peruvian waters but were not exploited by the fishery), giving the impression that a major part of the tropical shallow water ecosystem had shifted southward. The deep water shrimp catches, however, especially Penaeus brevirostris and P . californiensis, increased (Mora et al., 1984; Martinez, 1989). While the Ecuadorian shrimp trawl fishery apparently fully exploits this resource, with a relatively constant number of fishing vessels in operation since the beginning of the seventies, the very strong recent increase in shrimp production is a result of shrimp farming that also has expanded recently due to an extension of the area under cultivation (McPadden et al., 1988). After EN, spiny lobsters (Panulirusgracilis) and stalked barnacles (Pollicipes elegans) increased in abundance along the Peruvian central coast; however, the larvae of both species already had been transported into the area during EN. P . elegans occupied spaces in crevices and on rocks that had been vacated during the disappearance of local species during EN. In some areas the stalked barnacles remained alive until 1986 (Kameya and Zeballos, 1988). Emigration from shallow areas, apart from that from the tropical zones just mentioned, occurred on a large scale in upwelling waters. Thermally intolerant species withdrew to greater depths where the temperature increase was not so pronounced. Most of the species in question temporarily disappeared from the landings. Among the brachyurans, two species already mentioned, Cancer porteri and C . coronatus, apparently managed to survive at greater depths during the period of temperature increase (see above). A similar case among the gastropods involved the "false abalone" Concholepas concholepas, a carnivore that was largely deprived of its kelp shelter in shallow water but survived in part at greater depths (Amtz and E. Valdivia, 1985). Similar effects off northern Chile can be seen from the Cruz Grande data presented by Geaghan and Castilla (1987). Increases in the abundance of local populations more tolerant of high temperatures, mostly by species derived from tropical waters, induced the second positive development during and after EN 1982-83, which in economic terms was even more important than the shrimp invasion from the north. The principal species concerned were the Peruvian scallop Argopecten purpuratus (Fig. 12; Wolff, 1984, 1985; lllanes et al., 1985), and to a lesser degree the purple snail Thais chocolata and the octopus Octopusfontaneanus, species that are present at relatively low densities off central and southern Peni and northern Chile and that only modestly contribute to the landings in normal years. Under EN conditions these species are favored possibly not only by the increase in temperature but also because of their ability to exploit alternative food sources. Reduced competition with coexisting species that suffered high
347 mortality during EN may also have been an important factor. Favorable conditions for scallop larvae seem to be more important than favorable conditions during gonadal development and spawning of the adults, as can be seen from the weak correlation between parent stock size and recruitment success (Illanes et al., 1985; Wolff, 1987). The scallop population boom in Paracas Bay, south of Pisco, started with a heavy recruitment event in 1983 and with an extension of the area populated by this species into shallow water. The scallops grew to market size in only 6 months, and with > 100 individuals per square meter (5 - 8 kg m-2) the stock size was about 60 times greater than in normal years. An important processing industry developed at Pisco, and large shell middens covered many square kilometers in the surrounding desert. In 1984 there was a short period of diminished yield due to overfishing of the Paracas Bay stock, but in 1985 another boom occurred in Independencia Bay, south of the Paracas peninsula, which attracted even more fishermen to the area. Despite certain protective measures and closed seasons, the scallop fishery suddenly declined by the end of 1986 (cf. Fig. 12) mainly due to overfishing of the stock and the dumping of shells and waste onto the fishing grounds (Mendo et al., 1988). It is not clear at this time whether a more cautious exploitation would have resulted in a temporal extension of this fishery; in fact, a sustained yield of this species over a long period may be an illusion (Wolff, 1987). Interestingly, however, large ancient shell middens in the Peruvian coastal desert reflect former high abundances of this species that may well have been associated with similarly strong EN events in the past (Amtz et al., 1987). The post-EN development of shellfish was characterized by (a) withdrawal of the allochthonous species, especially the shrimps, although some penaeids managed to survive in the upwelling area for up to two years, (b) mortality of many invaders (especially most of the swimming crabs), (c) persistence of a few economically important invader stocks such as stalked barnacles and rock lobsters, and (d) recovery or reappearance of most autochthonous species within a period of about two years. Mainly as a result of the invasion of shrimp and the increase in abundance of scallops, the overall effect on the shellfish fisheries was rather beneficial. Total landings strongly increased, but their composition changed drastically. Moreover, the changes brought about by EN favored only a small part of the artisanal fishermen, those who owned boats and diving equipment for scallop exploitation, or those who were able to adapt their gear to shrimp fishing. However, the small-scale, shore-based subsistence fishery largely failed during the event and in the years following it.
4.3 Demersal and coastal fish Under normal conditions, the fishery for demersal fish in the Peruvian/Chilean upwelling area is restricted to the well oxygenated waters nearshore, and most of the catch is taken by small-scale artisanal fishermen using trammel and drift nets as well as hook and line. A demersal trawl fishery for finfish is carried out only in a limited area off northern Peni, between Paita (5"s) and Chimbote (9OS), whereas further south the seafloor is practically devoid of demersal fish below 30 m due to the extremely low oxygen values and the
348
narrowness of the shelf, which only north of Chimbote reaches a width of > 100 km. In this fishery, Peruvian hake (Merluccius gayi peruanus) provides about 70 % of the landings (Fig. 14). Part of these catches, however, are taken by purse seiners since this species often moves to shallow depths in the pelagic zone. Other species of commercial importance include a number of sciaenids (especially ayanque, Cynoscion analis, coco, Paralonchurus peruanus, cabinza, Isacia conceptionis, corvina, Sciaena gilberri, and lorna, S. deliciosa), sea bass (cabrilla, Paralabra humeralis), flatfish (Paralichthys adpersus and others), dogfish (tollo, Mustelus spp.), rays (Myliobatis spp.) and mullets (liza, Mugil cephalus and others). Coastal pelagic fish include cojinoba (Seriolella violacea),machete (Opisthonemalibertate) and silverside or pejerrey (Odontesthes regia regia). The increase of temperatures and dissolved oxygen at the seafloor during EN, and the resulting positive development of small benthic species that serve as food for demersal fish, affected the finfish populations just as much as the shelfish although in a somewhat different way.
320 280 240 200 160 120 80 40 0 28 24 20 16 12 8 4 0 1 Fig. 14. Annual landings of hake (M.gayi peruanus), dogfish (Mustelus spp.) and silverside (0.regia regia) off Peni, 1951 - 1987. Shaded areas: EN and other years with positive temperature anomaly > 2OC.
349 Soviet scientists working aboard the research vessel "Prof. Mesyatsev" were the first to detect higher 0 2 values at the bottom and latitudinal shifts in demersal fish populations during EN (Romanova, 1972). In 1972, hake and associated species were detected on the deeper shelf off Pisco (14"S), 5 degrees latitude south of their normal area. In 1982-83, the entire Peruvian and Ecuadorian demersal fish community in the Humboldt Current area apparently shifted to the south and, at the same time, extended its distribution beyond the edge of the continental shelf to the upper slope, thus covering a much wider area than during normal years. A large number of independent observations elucidated this general pattern. Demersal standard hauls obtained off Ecuador for research purposes yielded 30 - 50 % lower catches in April May 1983 than in normal years (Martinez, 1989). Gurnards (Prionotus stephanophrys), which at the onset of EN virtually disappeared from Ecuadorian waters, revealed a reduction in their share of the demersal fish catches from 21 - 36 % to 3 % (Herdson, 1984), and became abundant off Pisco where they are not normally caught. Hake were not recorded from their normal distribution area south of Paita between December 1982 and February 1984 (Castillo, 1985), but were concentrated on the upper slope where they encountered favourable 0 2 and food conditions, as well as temperatures similar to those in their usual habitat during non-EN years (Espino et al., 1985). South of Chimbote R.V. "Humboldt" obtained large catches of dogfish, rays, loma and sea bass in December and January 1982-83, in areas where demersal fish are not normally caught. Loma, which are usually restricted to within 20 km of the shore, were found up to 150 km offshore (Arntz, 1986). In April 1983, most of the local finfish species were absent from the Gulf of Guayaquil whereas tropical species (Fam. Lutjanidae and others) had immigrated into the gulf from more northerly areas (Herdson, 1984). During the much weaker EN 1986-87 most demersal fish again migrated towards the south (to about 10's) and to the margin of the continental shelf, but on a reduced scale. Fish were more dispersed on the seafloor, shoals of pelagic coastal fish could no longer be detected, and the size and age structure of the catches taken by R.V. "Humboldt" in January 1987 was also different from normal years (ComitC Cientifico CPPS, 1987). Thus it appears that demersal fish can benefit from the improved 0 2 and food conditions on the seafloor during EN, avoiding at the same time the high water temperatures found in the northern sector of the Humboldt Current upwelling area and in nearshore shallow waters. According to Espino and Wosnitza-Mendo (1986) hake cannibalism, which is an important source of natural mortality in that species, decreases substantially when hake extend their distribution area, as occurs during EN conditions. If at the same time predation by pelagic shoaling fish on hake larvae becomes less intense, as assumed by Wosnitza-Mendo and Espino (1986a), improved recruitment of Peruvian hake must be expected during strong EN events. During EN 1982-83 the demersal fish also changed their feeding habits. Hake, with a diet normally consisting almost exclusively of fish and euphausiids from the water column, consumed predominantly benthic crustaceans in 1983. Some hake and gurnard stomachs even contained bathypelagic fish that had formerly been recorded only off Central America. Lorna and sharks fed heavily on anchovies that had taken refuge at greater depths. Generally the food of demersal and coastal fish, in contrast to pelagic species, was found to be more diverse
350 during EN conditions (Hoyos et al., 1985; Sinchez de Benites et al., 1985; Tarazona et al., 1988a). Improved feeding conditions during EN 1982-83 may have been the reason for premature spawning in many demersal fish such as seabass (SamamC et al., 1985). Even among species spawning normally, spawning success and recruitment should have been improved because of favorable conditions on the seafloor and a closed fishing season that was enforced due to the dispersal of fish concentrations. During EN years the demersal fish catches off Peni are lower than in normal years (Fig. 14) because of the dispersal of fish and their movement into deeper waters, which decreases their availability and increases the difficulty of detection by means of echoacoustics. In Peni, demersal fish catches decreased from 94,000 t in 1981 (of which 69,000 t were hake) to 51,000 t (hake = 26,000 t) in 1982 and 26,000 t (hake = 6,000 t) in 1983. In 1984 landings rose to 64,000 t but hake landings (12,000 t) remained much below the pre-EN mean. Off Chile, the catches nearly doubled to 71,000 t in 1982 and persisted at a high level in 1983 with 56,000 t, probably due to the southward migration of part of the demersal stocks (IMARPE, unpub.).
5 CONCLUSIONS EN owes its negative reputation as a merely catastrophic event mainly to the dramatic collapse of the anchovy fishery that began in the early 1970s (which, however, was a consequence of continuous overfishing, although in some way it was also related to reproductive failures caused by EN; Jordin, 1983) as well as to certain impacts on organisms that catch the observers’ eye, such as the mortality of guano birds and seals. Before EN 198283, the literature concentrated on these negative aspects (e.g. Jordin, 1964; Boerema et al., 1965; Valdivia, 1978), and only during this exceptionally strong event did it become obvious that there were positive effects as well (Amtz, 1984), and that these beneficial effects occurred predominantly in benthic and nearshore subsystems that had been previously neglected (Amtz and J. Valdivia, 1985). We now know that it is necessary to consider the different subsystems separately. Even within the Humboldt Current upwelling area off South America there were strong gradients in effects from the equator to the south, from oceanic regions towards the shore in pelagic biotas, and from shallow to deep water in benthic environments (Amtz and Tarazona, 1988). The major ecological factors acting in these subsystems were different, ranging from an overall impact of increased temperatures to more local conditions such as increased sedimentation close to some river mouths. In many cases, although the effects of EN were clearly visible, we are not yet able to refer specific causes to them. Much work is still necessary to elucidate the mechanisms that ultimately cause the observed primary effects and the secondary biotic interactions that resulted from them. In the pelagic system, most effects of EN 1982-83 were indeed catastrophic although they were hardly visible in the Peruvian finfish landing figures because continued overfishing before the event has maintained low biomasses of the two principal species (anchovy and sardine).
35 1 Off Chile, however, the effects on the fishery were rather positive due to increased recruitment of sardines prior to the 1982-83 EN event and to the southward migration and near-surface concentration of sardines during the event, which resulted in increased vulnerability and much greater temporary landings of this species. Overall, the general biological impact on anchovies and sardines was negative; the changes in food composition and the almost complete breakdown of their underlying food base led to reproductive failure, lack of recruitment and considerable weight losses in these two shoaling species as well as to reduced stocks after EN. The different fate of the Peruvian and Chilean sardine fishery during the strong 1982-83 EN and later (i.e. during the moderate 1987 event), elucidated particularly well the variable impact of the disturbance on the behavior of pelagic shoaling fish along a latitudinal gradient. The mackerels, which are usually more oceanic, withstood the changes brought about by EN much better, but their presence in nearshore waters did not contribute a great deal to the fisheries. Many immigrant finfish species from oceanic and tropical waters entered the upwelling system, but were only of local importance for a short period in 1983. The total extent of this immigration must have been considerable taking into account the more than 50 fish species recorded off Perli and Chile that are not normal inhabitants of this area (Kong et al., 1985; VClez and Zeballos, 1985). The majority of demersal and nearshore fish such as hake, sciaenids and flatfish withdrew from shallow waters and moved to the edge of the continental shelf, and also undertook southward migrations at the same time, as did many immigrants from (sub)tropical waters (gurnards, snappers, dogfish, and rays, among others). The causes of these movements have to be sought in the increase of dissolved oxygen at the seafloor, the resulting improvement of benthic food resources at the bottom, and most likely in the temperature changes as well: while hake and accompanying species remained in waters of cooler temperatures at greater depth (Wosnitza-Mendo and Espino, 1986b), the tropical immigrants followed the warm waters further south. The effects on the fisheries were unfavorable during and immediately after EN, but the fish themselves could recover, and recruitment was enhanced, which may result in increased catches a few years after the event like those observed in 1978 as a consequence of the 1972-73 EN (Wosnitza-Mendo and Espino, 1986a). The effects on macrobenthos in general and on shellfish in particular were quite varied. Where normally hypoxic conditions prevail, the benthos benefitted from improved 0 2 conditions and developed unusually rich populations during EN and one year after the event. Then, with the final decrease in oxygen concentration, the communities returned to the impoverished state they had shown before EN. On sandy beaches and in the rocky intertidal zone, most invertebrate species (including shellfish) suffered mass mortalities caused presumably by a variety of impacts, with high temperatures, sea level changes, strong swell and increased predation as the most likely sources of disturbance. However, some local species such as scallops, purple snails and octopus, proliferated to unprecedented population levels. At the same time, some tropical immigrants, most of them crustaceans, became enormously abundant. Landings from either group -- especially those of scallops and shrimps - increased the shellfish catches much above the pre-EN level, both in amounts taken and in
352 value. The post-EN effects on the artisand fishery were also favorable at first since the scallop boom continued until 1986. At the present time, most shellfish that suffered during the event seem to have recovered to their pre-EN levels. This review is restricted to the impact of EN on benthos, fish and fisheries. There are many other effects of EN on the communities and species populations of the upwelling system (Amtz, 1986), and some of them are consequences of the effects described in this paper. For example, the disruption of the pelagic food web caused mass mortalities and reproductive failures of guano buds (see Smith, this volume, and Duffy, this volume) and seals (see Limberger, this volume); mortality of grazers in shallow water led to increases in algal biomass, changes in species composition and successional processes that remain to be studied in more detail. On the whole, however, recovery of most species occurred rather quickly, resembling recovery events in temperate areas where most species return to pre-disturbance abundances within 2 - 3 years (Arntz and Rumohr, 1982, 1986; Amtz and Arancibia, in press), and relatively few long-term effects have remained, thus demonstrating the remarkable resiliency of the upwelling ecosystem. One of the long-term effects is the continuing absence of the surf clam M.donucium from central Peruvian waters. Fluctuations of this species have been observed in former years as well, however, and are evident in shell middens along the Peruvian coast. Variations in abundance of surf clams and scallops (A. purpurutus) in shell middens may reflect former cold, anti-EN and warm, EN periods (Arntz et al., 1988). Finally, we would like to stress that EN 1982-83 was an exceptionally strong event (Quinn et al., 1987). During a minor EN event, such as 1976, only some effects were apparent, and the impact was probably resnicted to a limited section of the South American west coast off Ecuador and northem Perii. Only a long-term monitoring program including all major research institutions between Colombia and Chile will enable us to answer the many questions that have remained even after EN 1982-83, and in particular, to clarify the causal mechanisms underlying the ecosystem, community, and population responses described in this paper and, in a wider context, by Amtz (1986) and Glynn (1988). 6 ACKNOWLEDGMENTS Our thanks are offered to the members of the DePSEA group at San Marcos University, Lima, for their cooperation in the field work. Claudia Willeweit has been very helpful with the preparation of the manuscript. Alodia Holierhcek kindly removed some of the worst errors in the English language. Peter Clynn and two anonymous reviewers undertook considerable efforts to improve this paper both from a scientific and language point of view, and called attention to major inconsistencies in the first draft of the text. This work has been completed under an Alexander von Humboldt Foundation fellowship to Juan Tarazona in Germany, which is gratefully acknowledged. Contribution No. 260 from the Alfred-Wegener-Institut fur Polar- und Meeresforschung.
353 7 REFERENCES Aguayo, M., Erbs, V. and Estay, E., 1985. Influencia del fen6meno El Niiio 1982-83 en el crecimiento de sardina espaiiola (Surdimps sugar) y jurel (Truchurus murphyi) en el norte de Chile. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 175-184. Alamo, A., Bouchon, M. and Navarro, I., 1988. Variaciones en el factor de condicibn, coeficiente alimentario y alimentaci6n de la sardina peruana (Surdimps sugar sugar) durante el period0 1975-1986. In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinrimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Perh-Callao (special issue): 273-277. Alcocer, C. and Kelly, R., 1987. Variabilidad ambiental y su incidencia en aspectos fisiol6gicos de sardina (Surdinops sugar). Stock norte de Chile. Invest. Pesq. (Chile), 34: 5'7-63. Alvial, A., 1985. Programa de vigilancia del fendmeno El Niiio en la zona norte de Chile. Resultados y perspectivas. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 3 69-77. Amtz, W.E., 1984. El Niiio and Perfi: Positive aspects. ceanus 27(2): 36-39. Amtz, W.E., 1986. The two faces of El Nifio 1982-83. Meeresforsch. 31: 1-46. Arntz, W.E. and Arancibia, H., in press. El ecosistema bentodemersal del Pacific0 Suroriental y el del norte de Europa: una comparaci6n. Proc. CPPS Conference on the Living Resources and Fisheries of the SE Pacific, Viiia del MadChile, May 1988. Arntz, W.E. and Rumohr, H., 1982. An experimental study of macrobenthic colonization and succession, and the importance of seasonal variation in temperate latitudes. J. Exp. Mar. Biol. Ecol., 64: 17-45. Arntz, W.E. and Rumohr, H., 1986. Fluctuations of benthic macrofauna during succession and in an established community. Meeresforsch., 31: 97-1 14. Arntz, W.E. and Tarazona, J., 1988. Una retrospectiva a El Niiio 1982-83: QuC hemos aprendido? In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinrimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Peni-Callao (special issue): 353-364. Arntz, W.E. and Valdivia, E., 1985. Incidencia del fen6meno "El Niiio" sobre 10s mariscos en el litoral peruano. In: Amtz, W.E., Landa, A, and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Peni-Callao (special issue): 91101. Arntz, W.E. and Valdivia, J., 1985. Visi6n integral del problema: Introduccibn. In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Perli-Callao (special issue): 5-10. Arntz, W.E., Valdivia, E. and Zeballos, J., 1988. Impact of El Niiio 1982-83 on the commercially exploited invertebrates (mariscos) of the Peruvian shore. Meeresforsch., 32: 3-22. Amtz, W.E., Brey, T., Tarazona, J. and Robles, A., 1987. Changes in the structure of a shallow sandy-beach community in Perli during an El Niiio event. In: Payne, A.I.L., Gulland, J.A. and Brink, K.H. (Editors), The Benguela and comparable ecosystems, S. Afr. J. Mar. Sci., 5: 645-658. Amtz, W.E., Flores, L.A., Maldonado, M. and Carbajal, G., 1985. Cambios de 10s factores ambientales, macrobentos y bacterias filamentosas en la zona de m'nimo de oxigeno frente a1 Peni. In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Peni-Callao (special issue): 65-77. Avaria, S., 1985. Variaciones en la composici6n y biomasa del fitoplancton marino del norte de Chile entre diciembre 1980 y junio 1984. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 191-193. Avaria, S. and Muiioz, P., 1987. Effects of the 1982-1983 El Niiio on the marine phytoplankton off Northern Chile. J. Geophys. Res., 92: 14,369-14,382. Avaria, S., Muiioz, P. and Braun, M., 1988. El fitoplankton frente a Peninsula Los Molles, Valparafso, Chile (32'45's) y su relaci6n con "El Niiio" 1982-83. Rev. Biol. Mar., Valparaiso, 24: 1-35. Barber, R.T. and Chivez, E.P., 1983. Biological consequences of El Niiio. Science, 22: 1,203-1,210. Barber, R.T. and Chrivez, E.P., 1986. Ocean variability in relation to living resources during the 1982-83 El Niiio. Nature, 319: 279-285.
8
354 Blanco, J.L. and Diaz, M., 1985. Caracten'sticas oceanogrificas y desarrollo de El Niiio 1982-83en la zona costera de Chile. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fendmeno El Niiio 1982-83,Invest. Pesq. (Chile), 32:53-60. Boerema, L.K., Saetersdal, G.,Tsukayama, I., Valdivia, J. and Alegre, B., 1965. Report on the effects of fishing on the Peruvian stock of anchovy. F A 0 Fish. Tech. Paper, 55: 41 pp. Cane, M.A., 1983. Oceanographic events during El Niiio. Science, 22: 1,189-1,195. Cane, M.A., 1986. El Niiio. Ann. Rev. Earth Planet. Sci., 14:43-70. Caiidn, J.R., 1985. La variabilidad ambiental en la zona norte de Chile y su influencia en la pesquexfa peligica durante El Niiio 1982-83.In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fendmeno El Niiio 1982-83,Invest. Pesq. (Chile), 32: 119-128. Cirdenas, G.and Chipollini, A., 1988. Crecimiento de la sardina peruana (Surdimps sugm s a g a ) y composicidn por edades de sus desembarques durante 1981-85.In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Peni-Callao (special issue): 279-284. Carrasco, S . and Santander, H., 1987. The El Niiio event and its influence on the zooplankton off Perk J. Geophys. Res., 92: 14,405-14,410. Castilla, J.C., 1981. Perspectivas de investigacidn en estructura y dinimica de comunidades intermareales rocosas de Chile central. II.Depredadores de alto nivel trdfico. Medio Ambiente, 5: 190-215. Castillo, R., 1985. Breve comentario sobre la situacidn de 10s recursos "sardina" y "merluza" en el Puerto de Paita a fines de 1984.Bol. ERFEN, 12:25-26. Cucaldn, E., 1987. Oceanographic variability off Ecuador associated with an El Niiio event in 1982-1983.J. Geophys. Res., 92: 14,309-14,322. Chirinos de Vildoso, A., 1976.Aspectos del fendmeno El Niiio 1972-73.Parte 1: Distribucidn de la fauna. F A 0 Inf. Pesca, 185:62-79. Chirinos de Vildoso, A., 1984. El Niiio: Eventos desde 1891 hasta 1958,con tnfasis en la bibliografia peruana. Rev. Com. Perm. Pacifico Sur, 15: 363-392. CPPS (Permanent South Pacific Commission), 1987 . Caracterizacidn de El Niiio en el Pacifico sudeste. Bol. ERFEN, 22:3-30. Dayton, P.K. and Tegner, M.J., 1984. Catastrophic storms, El Niiio, and patch stability in a southern California kelp community. Science, 224:283-285. Del Solar, E., 1942. Ensayo sobre la ecologfa de la anchoveta. Bol. Cia Adm. Guano, 18(1):
1-23. Del Solar, E.M., 1983. Variacidn poblacional de algunos de 10s recursos vivos del mar por causa del fendmeno de "El Niiio" y el hombre. Bol. Lima, 27:59-66. DeVries, T.J., 1987. A review of geological evidence for ancient El Niiio activity in Perk J. Geophys. Res., 90:857-868. Dioses, T., 1985. Influencia del fendmeno "El Niiio" 1982-83en el peso total individual de 10s peces peligicos sardina, jurel y caballa. In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna manna, Bol. Inst. Mar Per& Callao (special issue): 129-134. Emery, W.J. and Hamilton, K., 1985. Atmospheric forcing of interannual variability in the northeast Pacific Ocean, connections with El Niiio. J. Geophys. Res., 90:857-868. Enfield, D.B., 1989a. El Niiio, past and present. Rev. Geophys., 27: 159-187. Enfield, D.B., 1989b. Is El Niiio becoming more common? Oceanography, l(2): 23-27,59. Enfield, D.B. and Allen, J.S., 1980. On the structure and dynamics of monthly mean sea level anomalies along the Pacific coast of North and South America. J. Phys. Oceanogr., 10:
557-578. Enfield, D.B. and Lukas, R., 1984. Low-frequency sea level variability along the South American Coast in 1981-1983,Trop. Ocean-Atmos. Newsl., 28:2-4. Espino, M. and Wosnitza-Mendo, C., 1986. Efecto de la concentracidn en la mortalidad natural y su aplicacidn en el anilisis de cohortes. Int. Symp. Long Term Changes Mar. Fish Pop., Vigo (Spain): 235-252. Espino, M., Benites, C. and Maldonado, M., 1985. Situacidn de la poblacidn de merluza durante "El Niiio". In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna manna, Bol. Inst. Mar Ped-Callao (special issue): 159-
162. F A 0 (Food and Agricultural Organization of the United Nations), 1989. Fishery statistics, catches and landings 1987.F A 0 Statistical Yearbook, 64:428-430.
355
Fahrbach, E., Trillmich, F. and Amtz, W.E., in press. The time sequence and magnitude of physical effects of El Niiio in the eastern Pacific. In: Trillmich, F. and Ono, K.A. (Editors),The Ecological Effects of El Niiio on Otariids and Phocids: Responses of Marine Mammals to Environmental Stress. Univ. Calif. Press, Berkeley. Firing, E., Lukas, R., Sadler, J. and Wyrtki, K., 1983. Equatorial undercurrent disappears during 1982-1983 El Niiio. Science, 222: 1,121-1,123. Fonseca, T.R., 1985. Efectos ffsicos del fen6meno El Niiio 1982-83en la costa chilena. In: Inst. Fomento Pesquero-Chile(Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 61-68. Fuenzalida, R., 1985. Aspectos oceanogrificos y meteorol6gicos de El Niiio 1982-83en la zona costera de Iquique. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 47-52. Gallardo, V.A., 1963. Notas sobre la densidad de la fauna bentonica en el sublitoral del norte de Chile. Gayana, Zoologia, 10: 1-15. Gallardo, V.A., Castillo, J.G. and Yaiiez, L.A., 1972. Algunas consideraciones sobre la ecologia bent6nica de 10s fondos sublitorales blandos en la Bahia de Concepci6n. Bol. SOC. Biol. de Concepci6n, 44: 169-190. Geaghan, J.P. and Castilla, J.C., 1987. Population dynamics of the loco (Concholepas concholepas) fishery in Central Chile. Invest. Pesq. (Chile), 34: 21-31. Glantz, M., Katz, R. and Krenz, M., 1987. The societal impacts associated with the 1982-83 worldwide climate anomalies. Nat. Cent. Atmos. Res., Colorado, 105 pp. Glynn, P.W., 1988. El Niiio-Southern Oscillation 1982-83: Nearshore population, community, and ecosystem responses. Ann. Rev. Ecol. Syst., 19: 309-345. GuillCn, O., Lostaunau, N. and Jacinto, M., 1985. Caracteristicas del fen6meno El Niiio 1982-83. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Per6-Callao (special issue): 11-21. Gunnill, F.C., 1985. Population fluctuations of seven macroalgae in Southern California during 1981-83 including effects of severe storms and of El Niiio. J. Exp. Mar. Biol. Ecol., 85: 149-164. Herdson, D., 1984. Changes in the demersal fish stocks and other marine life in Ecuadorian coastal waters during the 1982-83 El Niiio. Trop. Ocean-Atmos. Newsl., 28: 14-16. Hoyos, L., Tarazona, J., Shiga, B. and Chiong, V., 1985. Algunos cambios en la ictiofauna y sus relaciones tr6ficas durante el fen6meno "El Nifio" en la Bahia de Anc6n. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Per6-Callao (special issue): 163-171. Huyer, A., Smith, R.L. and Paluszkiewicz, T., 1987. Coastal upwelling off Per6 during normal and El Niiio times, 1981-1984. J. Geophys. Res., 92: 14,297-14,308. IFOP (Instituto de Fomento Pesquero), 1985. Taller nacional fen6meno El Niiio 1982-83. Invest. Pesq. (Chile), 32: 253 pp. Illanes, J.E., Akaboshi, S. and Uribe, E., 1985. Efectos de la temperatura en la reproducci6n del ostidn del norte Chlamys (Argopecten)purpuratus en la Bahia Tongoy durante el fendmeno El Niiio 1982-83. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 167-173. JimCnez, R., 1982. The 1976 El Niiio: Biological response off Ecuador. Paper presented at the Joint Oceanographic Assembly. August 1982. Halifax, Canada. JimCnez, R. and Herdson, D., 1984. Efectos de "El Niiio" 1982-83 sobre 10s recursos pesqueros en Ecuador. Rev. Com. Perm. Pacifico Sur, 15: 269-291. JimCnez, R. and Intriago, P., 1984. Equatorial upwelling processes in the Galfipagos Islands after the 1982-83 "El Niiio". Chapman EN Conference, Guayaquilficuador, Book of Abstracts: p. 39 Joint, I.R. 1986. Physiological ecology of picoplankton in various oceanographicprovinces. In: Platt, T. and Li, W.K.W. (Editors), Photosynthetic picoplankton. Can. Bull. Fish. Aquat. Sci., 214: 287-309. Jordfin, R., 1964. Las emigraciones y mortalidad de las aves en el invierno y otoiio de 1963. Inf. Inst. Invest. Recurs. Mar Per6-Callao. 27: 1-31. Jordin, R., 1983. Variabilidad de 10s recursos peligicos en el Pacifico Sudeste. Inf. Pesca FAO, 291: 113-130. Jordin, R. and Fuentes, H., 1966. Las poblaciones de aves guaneras y su situaci6n actual. Inf. Inst. Mar Per&Callao, 10: 1-31.
356 Kameya, A. and Zeballos, J., 1988. Distribuciones y densidad de percebes Pollicipes eleguns (Crustacea: CirriDedia) en el mediolitoral ueruano (Yasila, Paita; Chilca, Lima). Bol. Inst. Mar Peni-Calla6, 12: 1-22. Kellv. R.. 1985. Amectos eenerales de El Niiio 1982-83. In: Inst. Fomento Pesauero-Chile (Editor), Taller iacionarfen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 5-7. Konchina, Y.V., 1982. Feeding of the Pacific mackerel, Scomber juponicus (Scombridae), near the Peruvian coast. J. Ichthyol., 22: 102-111. Kong, I., Tomicic, J. and J. Zegers, 1985. Ictiofauna asociada a1 fendmeno El Niiio 1982-83 en la zona norte de Chile. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fendmeno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 215-223. La Cock, G.D., 1986. The southem oscillation,environmental anomalies and mortality of two southern African seabirds. Clim. Change, 8: 173-184. Lavalle y Garcia, J.A., 1917. Informe preliminar sobre la causa de mortalidad anormal de las aves ocumda en el mes de marzo del presente aiio. Mem. Comp. Adm. Guano, 8: 61-83. Leetma, A., Behringer, D.W., Huyer, A., Smith, R.L. and Toole, J., 1987. Hydrographic conditions in the eastern Pacific before. during and after the 1982/83 El Niiio. Proe. Oceanog., 19: 1-47. Martinez, C., Bohm, G. and Caballero, L., 1989. La pesquen'a peligica de la zona norte de Chile v su relaci6n con las condiciones ambientales. Bol. ERFEN. 29: 31-36. Martinez,-C., Salazar, C. and Bohm, G.,1984. La pesqueria cerquera ejercida en la zona norte de Chile y su relaci6n con 10s cambios biolbgico-pesquerosasociados al fen6meno de "El Niiio 1982-83". Rev. Com. Perm. Pacifico Sur, 15: 203-221. Martinez, J., 1989. Desarrollo, problemitica y perspectivas de 10s recursos demersales en el Ecuador. Inf. Subsecr. Rec. Pesq./Inst. Nac. Pesca, Ecuador (mimeo): 16 pp. McCreary, J.P. Jr. and Anderson, D.L.T., 1984. A simple model of El Niiio and the Southem Oscillation. Mon. Weather Rev., 112: 934-946. McPadden, C., Barragdn, J. and Rodrfguez, C., 1988. A study of the Ecuadorian shrimp fishery. Bol. Cient. Ttcn., Inst. Nac. Pesca, Ecuador, 9(4): 46 pp. Mendo, J., Valdivieso, V. and Yamashiro, C., 1988. Cambios en densidad, n ~ m e r oy biomasa de la poblaci6n de concha de abanico (Argopecten purpurutur) en la Bahia Independencia (Pisco, Peni). In: Salzwedel, H. and Landa, A. (Editors), Recursos y dindmica del ecosistema de afloramiento peruano. Bol. Inst. Mar Peni-Callao (special issue): 154-162. Menz, A,, 1986. Patrones de captura y tendencias. Ecuador. Oldepesca Doc. Pesca, 1: 57-59. Mora, O., Barreto, C. and Cuestas, S . , 1984. Cambios en la abundancia de camarones en la costa del Pacifico colombiano durante el fen6meno de "El Niiio" 1982-83. Rev. Com. Perm. Pacifico Sur, 15: 247-257. Moreano, H.R., 1984. The weakening of the 1982-83 El Niiio in Ecuadoran waters. Trop. Ocean-Atmos. News]., 28: 13-14. Muck, P. and Sinchez, G., 1987. The importance of mackerel and horse mackerel predation for the Peruvian anchoveta stock (a population and feeding model). In: Pauly, D. and Tsukayama, I. (Editors),The Peruvian anchoveta and its upwelling ecosystem: Three decades of change. ICLARM Studies and Reviews, 15: 276-293. Muck, P., Sandoval de Castillo, 0. and Carrasco, S., 1987. Abundance of sardine, mackerel and horse mackerel eggs and larvae and their relationship to temperature, turbulence and anchoveta biomass off Peni. In: Pauly, D. and Tsukayama, I. (Editors), The Peruvian anchoveta and its upwelling ecosystem: Three decades of change. ICLARM Studies and Reviews, 15: 268-275. Mujica, A., Rojas, 0. and Aranis, A., 1985. Aspectos reproductivos de la sardina espaiiola (Surdinops sugar Jenyns) en la zona norte de Chile y su variaci6n durante el fen6meno El Niiio 1982-83. In: Inst. Fomento Pesquero-Chile(Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 153-160. Muiioz, P., 1985. Estructura y comportamiento de las communidades fitoplanct6nicasen el norte de Chile durante el fen6meno de El Niiio 1982-83. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 195-197. Murphy, R.C., 1926. Oceanic and climatic phenomena along the west coast of South America during 192.5. Geogr. Rev., 16: 26-54. Mysak, L.A., 1986. El Niiio, interannual variability and fisheries in the northeast Pacific Ocean. Can. J. Fish. Aquat. Sci., 43: 464-497. Y
Y
357
Paine, R.T., 1986. Benthic community - water column coupling during the 1982-1983 El Niiio. Are community changes at high latitudes atmbutable to cause or coincidence? Limnol. Oceanogr., 31: 35 1-360. Paredes, C. and Tarazona, J., 1980. Las comunidades de mitilidos del mediolitoral rOcoso del Departamento de Lima. Rev. Per. Biol., 2: 59-71. Pastor, A., 1984. El Niiio 1982-83 y su incidencia en algunos aspectos biol6gicos y pesqueros en la costa sur del litoral peruano. Bol. ERFEN, 10: 14-22. Pauly, D., Chirinos de Vildoso, A., Mejia, J., SamamC, M. and Palomares, M.L., 1987. Population dynamics and estimated anchoveta consumption of bonito (Sarda chiliensis) off Per& 1953 to 1982. In: Pauly, D. and Tsukayama, I. (Editors), The Peruvian anchoveta and its upwelling ecosystem: Three decades of change. ICLARM Stud. Rev., 15: 248-267. Pearcy, W.G. and Schoener, A., 1987. Changes in the marine biota coincident with the 19821983 El Niiio in the northeastern subarctic Pacific Ocean. J. Geophys. Res., 92: 14,41714,428. Peiia, R., Zdiiiga, 0. and Rodriguez, L., 1987. Variaci6n estacional del indice gonadosomitico en Acanthpleuru echinaru (Barnes, 1823) (Mollusca: Polyplacophora). Estud. Oceanol., 6: 59-65. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S., 1987. El Niiio Occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14,449-14,461. Ramorino, L. and Muiiiz, L., 1970. Estudio cuantitativo general sobre la fauna de fondo de la Bahia de Mejillones. Rev. Biol. Mar, Valparaiso, 14: 79-93. Rasmusson, E.M. and Wallace, J.M., 1983. Meteorological aspects of the El Niiio/Southern Oscillation. Science, 222: 1,195-1,202. Retamales, R. and Gonzilez, L., 1985. Incidencia del fen6meno El Niiio 1982-83 en el desove de sardina espaiiola (Sardinops sagau). In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Nifio 1982-83, Invest. Pesq. (Chile), 32: 161-165. Robinson, G.and del Pino, E.M. (Editors), 1985. El Nifio en las Islas Galipagos, el evento de 1982-1983.Fundacidn Charles Darwin para las Islas Galipagos, Quito, Ecuador, 534 PP. Rojas de Mendiola, B., G6mez, 0. and Ochoa, N., 1985. Efectos del fendmeno "El Niiio" sobre el fitoplancton. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Peni-Callao (special issue): 3340. Rollins, H.B., Sandweiss, D.H., and Rollins, J.C., 1986. Documentation of large-magnitude El Niiio events using molluscs from coastal archaeological sites (Abstract). In: Chapman conference on El Niiio. Guayaquil, Ecuador, 29. Romanova, N.N., 1972. Investigaciones cientifico-pesqueras en las aguas del OcCano Pacific0 adyacentes a la costa del Peni durante el invierno de 1972.Distribucidn de bentos en la plataforma y en el talud continental de la costa peruana. Ser. Inf. Esp. Inst. Mar Per& Callao, 128: 127-132. Romero, L., Paredes, C. and Chivez, R., 1988. Estructura de la macrofauna asociada a 10s rizoides de Lessonia sp. (LaminarialesPhaeophyta). In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Peni-Callao (special issue): 133-139. Romo, D., 1985. Composici6n quimica de la harina de pescado chilena durante el fen6meno El Niiio 1982-83. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fendmeno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 141-151. Rosenberg, R., Arntz, W.E., Chumin de Flores, E., Flores, L.A., Carbajal, G., Finger, I. and Tarazona, J., 1983. Benthos biomass and oxygen deficiency in the upwelling system off Peni. J. Mar. Res., 41: 263-279. Salzwedel, H., Flores, L.A., Ch. de Flores, E., Zafra, A. and Carbajal, G.,1988. Macrozoobentos del sublitoral peruano antes, durante y despuCs de El Niiio 1982-83. In: Salzwedel,H. and Landa, A. (Editors), Recursos y dinimica del ecosistema afloramiento peruano. Bol. Inst. Mar Perii-Callao (special issue): 17-98. Samamt, M., Castillo, J., Flores, L.A. and Vilchez, R., 1978. Estructura, distribuci6n y abundancia de peces demersales. Inf. Inst. Mar Peni-Callao, 47: 1-28. SamamC, M., Castillo, J. and Mendieta, A., 1985. Situaci6n de las pesquerias demersales y 10s cambios durante la presencia del fen6meno "El Niiio". In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Perd-Callao (special issue): 153-158.
358
Sinchez de Benites, G., Alamo, A. and Fuentes, H., 1985. Alteraciones de la dieta alimentaria de algunos peces comerciales por efecto del fen6meno "El Niiio". In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Per6-Callao (special issue): 135-142. Santander, H. and Tsukayama, I., 1984. The anchoveta and sardine and some events associated to their recruitment. IOC Worksh. Rep., 33: 11-14. Santander, H. and Zuzunaga, J., 1984. Impact of the 1982-83 El Niiio on the pelagic resources off Peni. Trop. Ocean-Atmos. Newsl., 16: 10-12. Santander, H., Alheit, J. and Smith, P.E., 1984. Estimaci6n de la biomasa de la poblaci6n desovante de anchoveta peruana, Engraulis ringens, en 1981 por aplicacidn del "MCtodo de producci6n de huevos". Bol. Inst. Mar Peni-Callao, 8: 209-250. Schweigger,E., 1953. Situaci6n veraniega en el litoral peruano en 1953. Bol. Cient. Cia Adm. Guano, 1: 1-18. Schweigger,E., 1960. Fen6menos hidrogrificos y biol6gicos en el sur del Peni y en el norte de Chile. Rev. Biol. Mar. Valparaiso, 10: 51-68. Schweigger, E., 1964. El litoral peruano. Segunda edicidn auspiciada por la Univ. Nac. Federico Villareal, Lima, 414 pp. Serra, R., 1986. Desarrollo de la pesqueria de anchoveta (Engraulis ringens) y 10s cambios de su abundancia. Invest. Pesq., 33: 13-24. Serra, R. and Tsukayama, I., 1988. Sinopsis de datos biol6gicos y pesqueros de la sardina Sardinops sagax (Jenyns, 1842) en el Pacifico Suroriental.F A 0 Sinopsis sobre la Pesca, 13 (Rev. 1): 60 pp. Silva, N. and Rojas, R., 1984. Presencia del fendmeno El Niiio 1982/1983 frente a la costa norte de Chile. Rev. Com. Perm. Pacifico Sur, 15: 121-140. Smith, R.L., 1983. Peni coastal currents during El Nifio: 1976 and 1982. Science, 221: 1,397-1,399. Smith, R.L., 1984. Coastal currents off Perli during El Niiio of 1982-83. Prenatal to senescent conditions. Trop. Ocean-Atmos. Newsl., 28: 7-8. Soenens, P., 1985. Estudios preliminares sobre el efecto del fen6meno "El Niiio" 1982-1983 en comunidades de Aulacomya ater. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna manna, Bol. Inst. Mar PeniCallao (special issue): 51-53. Soto, R., 1985. Efectos del fen6meno El Niiio 1982-83 en ecosistemas de la I Regi6n. In: Inst. Foment0 Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 199-206. Stockner, J.G., 1988. Phototrophic picoplankton: An overview from marine and freshwater ecosystems. Limnol. Oceanogr., 33: 765-775. Taft, B.A., 1985. El Niiio of 1982-83 in the tropical Pacific. In: Wooster, W.S. and Fluharty, D.L. (Editors), El Niiio North, Niiio effects in the eastern subarctic Pacific Ocean. Wash. Sea Grant Program, Seattle: 1-8. Tarazona, J., 1984. Modificacionesde la infauna bent6nica de una bahia con deficiencia de oxigeno durante El Niiio 1982-83. Rev. Com. Perm. Pacifico Sur., 15: 223-238. Tarazona, J., Amtz, W.E. and Hoyos, L., 1988a. Repartici6n de 10s recursos alimenticios entre tres peces bent6fagos frente a1 Peni antes, durante y desputs de El Niiio 1982-83.In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Perli-Callao (special issue): 107-114. Tarazona, J., Paredes, C. and Igreda, M., 1986. Estructura del macrobentos en las playas arenosas de la zona de Lima, Perli. Rev. Ciencias U.N.M.S.M., 74: 103-116. Tarazona, J., Salzwedel, H. and Arntz, W.E., 1988b. Oscillations of macrobenthos in shallow waters of the Peruvian central coast induced by El Niiio 1982-83. J. Mar. Res., 46: 593-611. Tarazona, J., Salzwedel, H. and Arntz, W.E., 1988~.Positive effects of "El Niiio" on macrozoobenthos inhabiting hypoxic areas of the Peruvian upwelling system. Oecologia, 76: 184-190. Tarazona, J., Paredes, C., Romero, L. and Guzmin, S., 1988d. La recolonizaci6n de las comunidades de mitilidos en la costa central del Peni desputs de El Niiio 1982-83. In: Salzwedel, H. and Landa, A. (Editors),Recursos y dinimica del ecosistema afloramiento peruano. Bol. Inst. Mar Peni-Callao (special issue): 41-49.
359 Tarazona, J., Amtz, W.E., Canahuire, E., Ayala, Z. and Robles, A., 1985a. Modificaciones producidas durante "El Niiio" en la infauna bentdnica de Leas someras del ecosistema de afloramiento peruano. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Penl-Callao (specid issue): 5563. Tarazona, J., Paredes, C., Romero, L., Blaskovich, V., GuzmBn, S. and SBnchez, S., 1985b. Caracteristicas de la vida planctdnica y colonizacidn de 10s organismos bentonicos epiliticos durante el fen6meno "El Niiio". In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fendmeno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Peni-Callao (special issue): 41-49. Tegner, M.J. and Dayton, P.K., 1987. El Niiio effects on Southern California kelp forest communities. Adv. Ecol. Res., 17: 243-279. Thompson, L.G., Moseley-Thompson, E. and Morales-Arnao, B., 1984. El Niiio-Southern Oscillation events recorded in the stratigraphy of the tropical Quelccaya Ice Cap, Peni. Science, 226: 50-53. Thorson, G., 1946. Reproduction and larvel development of Danish marine bottom invertebrates, with special reference to the planktonic larvae in the Sound (0resund). Medd. Komm. Danm. Fisk.-og Havunders., Ser. Plankton, 4: 523 pp. Tokeshi, M., Estrella, C. and Tarazona, J., 1988. Estudio preliminar de las relaciones predator-presa en Heliasrer helianthus (Asteroidea: Echinodermata). In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinimica del ecosistema afloramiento peruano. Bol. Inst. Mar Pen-Callao (special issue): 141-145. Tokeshi, M., Estrella, C. and Paredes, C., 1989a. Feeding ecology of a size-structured predator population, the South American sun-star Heliasrer helianrhur. Mar. Biol., 100: 493-505. Tokeshi, M., Romero, L. and Tarazona, J., 1989b. Spatial coexistence of mussel-associated free-ranging polychaetes in a subtropical intertidal habitat. J. Anim. Ecol., 58: 681-692. Tomicic, J.J., 1985. Efectos del fendmeno El Niiio 1982-83 en las comunidades litorales de la Peninsula de Mejillones. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fendmeno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 209-213. Torres, A., Martinez, C. and Oliva, J., 1985. Migraciones de la sardina espaiiola en el Pacifico Suroriental durante el fendmeno El Niiio 1982-83 y en 1984. In: Inst. Fomento Pesquero-Chile (Editor), Taller nacional fen6meno El Niiio 1982-83, Invest. Pesq. (Chile), 32: 95-100. Tsukayama, I. and Santander, H., 1986. Cambios bidticos y efectos sobre 10s recursos pesqueros y las pesquerias en el Peni. Rev. Com. Perm. Pacifico Sur, 16: 97-166. Valdivia, J., 1978. The anchoveta and "El Niiio". Rapp. P.-v. RCun. CIEM, 173: 196-202. VClez, J.J. and Zeballos, J., 1985. Ampliaci6n de la dismbuci6n de algunos peces e invertebrados durante el fendmeno "El Niiio" 1982-83. In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fenbmeno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Peru-Callao (special issue): 173-180. Vilchez, R., Muck, P. and Gonzales, A., 1988. Variaciones en la biomasa y en la distribucidn de 10s principales recursos peligicos del Peru e n m 1983 y 1987. In: Salzwedel, H. and Landa, A. (Editors), Recursos y dinimica del ecosistema de afloramiento peruano. Bol. Inst. Mar Peru-Callao (special issue): 255-264. Vogt, W., 1940. Una depresidn ecoldgica en la costa peruana. Bol. Cia Adm. Guano, 16(10): 307-329. Vogt, W., 1957. Informe sobre las aves guaneras. Bol. Cia Adm. Guano, 33(3): 1-132. Wisely, B., 1958. The development and settling of a serpulid worm, Hydroides norvegica, Gunnerus (Polychaeta). Aust. J. Mar. Freshw. Res., 9: 351-361. Wolff, M., 1984. Impact of the 1982-83 El Niiio on the Peruvian scallop Argopecren purpuratus. Trop. Ocean-Atmos. News]., 28: 8-9. Wolff, M., 1985. Abundancia masiva y crecimiento de preadultos de la concha de abanico peruana (Argopecrenpurpurarus) en la zona de Pisco bajo condiciones de "El Niiio" 1983. In: Arntz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impacto en la fauna marina, Bol. Inst. Mar Ped-Callao (special issue): 87-89. Wolff, M., 1987. Population dynamics of the Peruvian scallop Argopecten purpuratus during the El Niiio phenomenon of 1983. Can. J. Fish. Aquat. Sci., 44: 1,684-1,691. Wooster, W.S. and Fluharty, D.L. (Editors), 1985. El Niiio North, Niiio effects in the eastern subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. of Wash., Seattle, 312 PP.
360
Wosnitza-Mendo, C. and Espino, M., 1986a. The impact of "El Niiio" on recruitment of the Peruvian hake (Merluccius gayi peruanus). Meeresforsch, 31: 47-5 1. Wosnitza-Mendo, C. and Espino, M., 1986b. La construcci6n de curvas de rendimiento sostenible (Shepherd, 1982) para la merluza peruana. Bol. Inst. Mar Peni-Callao, lO(7): 165-183. Wosnitza-Mendo, C., Espino, M. and VCliz, M., 1988. La pesqueria artesanal en el Peni durante junio de 1986 a junio de 1988. Inf. Inst. Mar Peni-Callao, 93: 144 pp. Wyrtki, K., 1975. El Niiio - the dynamic response of the equatorial Pacific Ocean to atmospheric forcing. J. Phys. Oceanogr., 5: 572-584. Wyrtki, K., 1979. El Niiio. La Recherche, 106: 1,212-1,220. Wyrtki, K., 1982. The Southern oscillation, ocean-atmosphere interaction and El Niiio. Mar. Technol. SOC.J., 16(1): 3-10. Zama, A., Rueda, T. and Cirdenas, E., 1984. Unusual arrival of chub mackerel Scornber juponicus at Puerto Chacabuco, southern Chile (Pisces, Scombridae). Rev. Biol. Mar., Valparaiso, 20: 61-76. Zuzunaga, J., 1985. Cambios del equilibrio poblacional entre la anchoveta (Engruulis ringens) y la sardina (Surdinops sugm) en el sistema de afloramiento frente al Peni. In: Amtz, W.E., Landa, A. and Tarazona, J. (Editors), El fen6meno El Niiio y su impact0 en la fauna marina, Bol. Inst. Mar Peni-Callao (special issue): 107-117.
361
EFFECTS OF THE 1982-83 EL NIRO-SOUTHERN OSCILLATION EVENT ON MARINE IGUANA (AMBLYRHYNCHUS CRISTATUS BELL, 1825) POPULATIONS ON GALAPAGOS
W. ANDREW LAURIE Max-Planck-Institut fur Verhaltensphysiologie, 8 131, Seewiesen, West Germany, and Large Animal Research Group, Department of Zoology, University of Cambridge, Cambridge CB2 3EJ, United Kingdom. ABSTRACT Laurie. W. A.. 1989. Effects of the 1982-83 El Niiio-Southern Oscillation event on marine iguana (Amblhhynchus Gristatu Bell, 1825) populations on Galapagos. Y
The effects of the 1982-83 El Niiio-Southern Oscillation event (ENSO) on marine iguanas (Amblvrhvnchus cristatus) were observed during a long term study of marine iguana population dynamics on Santa Fe, Galapagos, begun in 1981. The 1982-83 ENSO was the most severe ever recorded: there were record sea-surface temperatures, sea-levels and rainfall, and a major change in marine algal flora resulted in disappearance of most of the iguanas' preferred food species and colonization of the intertidal zone by the brown alga Giffordia mitchelliae, a species not previously recorded in Galapagos. This led to widespread starvation, with about 60% of the Santa Fe population dying between March and August 1983, and similar mortality on other islands of the archipelago. Adult males and juveniles suffered the highest mortality, with 1982 hatchlings being almost completely exterminated. Body condition and growth rates were greatly depressed during ENSO, with adult growth ceasing almost entirely, but both increased again rapidly after the population crash and reached levels higher than before ENSO. There was almost no breeding in the post ENSO season (1983-84) but since then frequency of breeding, age at first breeding, and clutch size have all increased above pre-ENS0 levels. It is suggested that the increases in rates of growth and reproduction are due to a reduction in competition for food after the return of normal feeding conditions at greatly reduced population density. 1 INTRODUCTION A long term study of marine iguana (AmblYrhvnchusGristatus) population dynamics in Galapagos, begun in 1981, provided an opportunity to study the effects on the iguanas of the 1982-83 El Niiio-Southem Oscillation (ENSO) event (Philander, 1983), the most severe on record (Quinn et al., 1978, 1987; Glynn, 1988). ENSO events are characterized by a massive advection of warm, low salinity, nutrient poor surface water to the south in the eastern tropical Pacific, mainly along the coasts of Ecuador and Peru (Houvenaghel, 1978, 1984; Rasmusson, 1984; Hansen, this volume). The biological productivity of the euphotic zone declines rapidly leading to reduced survival and reproduction of animals at higher trophic levels (Barber and Chavez, 1983, 1986; Trillmich and Limberger, 1985; Arntz, 1986; Barber and Kogelschatz, this volume), although the increased rainfall leads to increased reproduction in land-based ecosystems, for example, in Darwin's finches on Galapagos (Gibbs and Grant, 1987; Grant and Grant, 1987) and in the floras of the "lomas" formations in the Atacama and Peruvian deserts (see Dillon and Rundel, this volume).
362
The marine iguana is endemic to Galapagos and is widely distributed over the archipelago with highest concentrations on the western islands (Laurie, 1983a). The iguanas feed on fleshy macrophytic marine algae, either diving for them or grazing on exposed intertidal rocks at low tide (see Trillmich and Trillmich, 1986). They are sexually dimorphic, with adult males typically weighing 70% more than adult females. Adult male body weight varies from a maximum of 12.3 kg on southern Isabela to about 1.2 kg on Genovesa (Laurie, in prep.). Males defend mating territories during the breeding season, and females lay one to six eggs about one month after copulation. The eggs take three months to incubate in nests dug 30-80 cm deep in sand or volcanic ash. The time of the breeding season varies between islands (Fig. l), being earliest (nesting in January) on Santa Fe and latest (nesting in March-April) on southern Isabela and Espaiiola (Laurie, in prep.). During 1983 unusually high mortality of marine iguanas was observed in populations on all the islands of the archipelago, except Wenman and Culpepper, which were not visited (Laurie, 1983b). A major change in marine algal flora was observed during the same period and abnomially high rainfall, sea-surface temperatures (SST) and sea-levels associated with the
Fig. 1. Map of the Galapagos Islands, showing main study site, Miedo. 1982-83 El Niiio-Southern Oscillation event (ENSO) were recorded from November 1982 until July 1983. The mean monthly SST anomaly reached +4.3OC in June 1983 (Fig. 2), and the pattern of SST fluctuations was similar to those on the coast of Peru (Chavez, 1987). The
363
tradewinds failed almost completely, and on Santa Cruz, where the mean annual rainfall between 1965 and 1981 was 374 mm, 3,264 mm of rain fell between November 1982 and July 1983 (Robalino, 1985). There was an increase in sea level over the same period that varied between 20 and 45 cm (Wyrtki, 1985). ENS0 events occur frequently but are poorly predictable and vary
I
I
I
I
1965
1970
1975
1980
1985
1970
1975
1980
1985
30
V
0
W
n 3
$ 25 a: W a W I
cu W
a
k
20
ul 3
a ul W
15 1965
YEAR
Fig. 2. Monthly mean sea-surface temperature (SST) anomalies (above) and monthly mean seasurface temperature (below) observed on the shore in Academy Bay, 1965-8 (courtesy Charles Darwin Research Station). Broken curve denotes long-term (22 years) annual mean SST. greatly in intensity, extent of influence and duration. Quinn et al. (1987) have documented 24 events of near moderate to very strong intensity since 1900 and 50 between 1800 and 1987, a mean frequency of one event in 3.8 years. The average interval between strong or very strong
364 ENSO events, with mean monthly sea surface temperature anomalies of 3.0 to 5.0OC, is 12.3 years (Quinn et al., 1978,1987) but the 1982-83 ENSO was exceptional, and there is evidence that the last event of comparable magnitude occurred in 1877-78 (Kiladis and Diaz, 1984). 2 STUDY AREA The main study area was at Miedo, on the south coast of Santa Fe (Fig. 1) and consisted of 2 km of low, rocky coastline with extensive intertidal flats and abundant marine algae. There is an old, uplifted beach deposit 300 m inland at the base of an escarpment, and most of the marine iguanas in the study area nested there. A number of other sites were chosen on other islands for comparative observations during regular visits over the study period. The climate of Galapagos is biseasonal: January to May is the hot season, with the only substantial rainfall, and June to December is cool, and frequently overcast, with persistent, very light drizzle (Colinvaux, 1984). 3 METHODS 3.1 Census technicues In order to collect comparative data on population densities and composition on Santa Fe and other islands binoculars were used to count animals and classify them according to sex and size. Iguanas were divided into 11 different size classes, based on snout-vent length and, with practice, animals could be classified accurately to size class without being captured or otherwise disturbed. Differences between the sexes in body size, head width and nuchal crest development were sufficient to determine the sex of most adults without capture. Prominent hemipenes were visible in some younger males when held in the hand but even after capture sex determination was generally possible only for the older animals. Counts of iguanas along the same stretch of coastline produced different results in terms of both numbers and population composition according to the time of day and state of the tide. Experiments showed that the censuses in the late afternoon gave the highest counts, and that consistent estimates of population size and composition can be made by a simple mark, release and count method (Laurie, 1982). An annual census was made each April on Santa Fe using this method: the animals released after weighing and measuring constituted the marked population. 3.2 CaDturine and marking ieuanas A total of 3,833 iguanas was marked over the study period: 3,482 on Santa Fe and 351 on Caamaiio, a small islet off Santa Cruz (Fig. 1). A further 1,440 captures were made on other islands but the animals were weighed, measured and released without marking and may not all be different individuals. Fifty-two recaptures were made of animals marked earlier by C. Rohrbach and N. Rauch on Caamaiio and Punta Nuiiez. Adult iguanas were caught by hand on the shore in the early morning when still torpid, or later with the help of a running noose on a bamboo pole. Hatchlings were caught on emergence at the nesting area in an enclosure fenced by 45 cm high plastic sheeting. An intensive effort was made to recapture all marked animals each year between February and May for measuring and weighing,
365
and many adult males and females were recaptured each October or November too, at the start of the breeding season. Permanent marking was achieved by hot branding with wire brands heated in a portable gas burner. Coloured glass beads, threaded on nylon line through the nuchal crest, provided a second method of marking that was less permanent but allowed identification of individuals at greater distances. For very rapid identification of adults during limited periods of detailed observation, particularly over the breeding season, animals were painted with white, orange or yellow paint on both flanks with numbers 5 cm high. A small patch of orange or yellow paint on the neck or base of the tail was used to distinguish animals caught at different sites. There was no evidence that the numbers or the small patches of colour affected the iguanas' behaviour. 3.3 Observations of reproductive behaviour Intensive observations were made during each breeding season from 1981-82 to 1985-86. Observations were made from suitable vantage points above colonies on the coast and the nesting area, during continuous (0700-1730) daytime watches that spanned and were maintained throughout each breeding season. These involved two or three observers who worked in shifts every day for eight weeks each season. Checks were also made during the night, with some prolonged nocturnal observations at the nesting area. The proportion of females that nested each year was estimated from a combination of direct observations and changes in weight of females caught at the beginning and the end of each nesting season (Laurie, in prep.). 3.4. Measurements of ivuana Standard linear measurements were made to the nearest 1 mm on each occasion that animals were captured. They included snout to vent length (SVL), tail length (TL),maximum head width (HW) (at the point on the living animal with maximum width across the quadrates, just in front of the tympanum) and length of longest nuchal crest spine (SL). A 600 mm rigid stainless steel rule and Vernier calipers (Mitutoyo) were used for these measurements. Spring balances (Pesola) were used to determine body weight (WT) to the nearest 1 g up to 100 g, to the nearest 10 g up to 1,000 g, to the nearest 50 g up to 2,500 g and to the nearest 100 g above 2,500 g. Two people are needed to measure an iguana accurately. There appeared to be more potential for error in the way the iguanas were held for measuring, than in the actual measuring, particularly for SVL. So, as I was present throughout the study, I always held the iguanas and a number of different people did the measuring. Accuracy checks were carried out within each season using repeat measurements of 100 animals within an hour of fiist measuring. The standard deviations of measurements were: SVL, 2.2%, TL, 0.8%, HW, 2.5%, SL, 5.0% and WT, 0.5%; with only small differences between size classes. The length measurements of adult males were slightly less consistent (2.25% sd for SVL) than for smaller animals (hatchlings: 2.13%), due to small variations in the extent to which the animals were stretched for measurement. The major error appeared to be in SVL measurement so I tested for consistency between seasons and assistants by comparing the ratio of SVL to total length (SVL + TL) for each size class in each year, first discarding all measurements of animals that had lost part of their tails. If SVL and TL are assumed
366
to have independent errors and the variances are small, the standard deviation of 3.04% in S W ( S V L + TL) corresponds to a standard deviation of 2.15% in SVL measurement between seasons; i.e. the same as within seasons. 3.5 Growth Growth rates were calculated as annual rates, and corrected for differences between years in the actual date of capture, usually not more than one month. Differences in growth rates between years and between size, sex and age classes were tested by an&jsis of variance and the t-test for the difference between means, making full use of the different types of data involved: paired increments (for the same individuals in two years), unpaired increments (for the means of different individuals of the same size class at the beginning of the years in question) and data for hatchlings that can be separated into each year's cohort. 3.6 Survival r m Three sources of data were available for estimating survival rates in each sex and age class: annual recapture of marked individuals, recovery of corpses (marked and unmarked) and the annual censuses. The best data are those on recapture of marked individuals. Estimation of survival was based on the method of Pollock (1981) and reported by Laurie and Brown (in press a). 4 RESULTS 4.1 Morta litv. a w e d with E m Only ten corpses were recorded from the coastline of the study area between April 1981 and October 1982 during eight months on Santa Fe, but between November 1982 and July 1983 more than 800 corpses were recovered during six months on the island. Most corpses were washed away by the sea so the figures indicate an enonnous mortality in a population estimated to number less than 8,000 individuals in June 1982. The data for recaptures of marked individuals were used to estimate annual mortality rates (April to April) for each sex and size class and cohort (Laurie and Brown, in press a). During the 1982-83 ENSO the relative rates for each group were checked using the data from recovery of corpses and the annual census, which, although less accurate, gave very consistent results. Animals began to die as a result of ENSO as early as November 1982, so the annual mortality rates shown in Table 1 do not accurately indicate the size of the effect of ENSO. Further analysis, using November recaptures, showed that mortality over the period November 1982 to November 1983 rose from a pre-ENS0 level of 8-15% in adult males, 2-4% in adult females and 46% in hatchlings to 58% in adult males, 47% in adult females and 84% in hatchlings. These rates have since returned towards pre-ENS0 levels (Table 1) but adults of both sexes have lower survival than before ENSO and the 1985 hatchlings had a survival rate in the first year similar to hatchling survival rate in 1982 and 1983 and considerably lower than in 1981 (Laurie and Brown, in press a). Fig. 3 shows the percentage of animals that survived to the end of each year according to sex (in adults) and cohort (year of emergence).
367 TABLE 1 Estimated percentage annual mortality (April to April) for each adult sex class and cohort. COHORTS Adultmales N = 464 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87
14.7 7.5 57.7 5.3 27.7 (7.2)
Adultfemales 453 3.8 1.5 47.1 13.6 16.9
1980 140
1981 643
1982 404
22.9 11.1 77.7
35.7 47.3 72.9 17.5 13.1 (14.6)
60.6 83.7 7.1 21.5
-
-
1983 422
55.6 13.1 3.1 (14.6)
1985 722
60.4 (10.9)
1986-87 figures are over-estimates (see Laurie and Brown, in press a).
The abnormal mortality started in December 1982 and continued until August 1983, with the highest mortality occurring between April and July 1983 (Laurie, 1983b). Adult iguanas weighed a mean of 54.2% (s.e. 0.9%. n = 42) of their normal weight at death, and were extremely emaciated, with no fat reserves and considerable reduction of musculature, particularly at the base of the tail (Cooper and Laurie, 1987). They were very weak and in the last few days before death could hardly move. Their stomachs generally contained very little: the mean weight of the contents of 89 adults' stomachs was 17 g (s.e. 3 g, range 0-220 g) compared with a mean of 196 g (s.e. 22 g, range 95-228 g) for 6 adults' stomachs collected during normal conditions. Apart from algae, stomach contents included small stones, pieces of crab IGraDsus iguana skin, iguana and sea-lion (7alophus d o r n i a n u ) faeces, sea-lion hairs, and earth. These other items were also observed being picked up on land by animals apparently too weak to venture into the water or the intertidal zone. The algae present in the stomach and the intestines consisted mainly of Giffordia mitchelliap and were largely undigested. In marked contrast to the normal semi-liquid state of algae in marine iguana intestines, Giffordia was relatively dry and fibrous and remained so in the faeces, which are normally liquid and amorphous containing few recognizable parts of algae. Gross and histopathological examination of iguanas that died during ENSO and comparison with others that died under normal conditions indicated that the former died of starvation (Cooper and Laurie, 1987).
m,
4.2 Changes in aleal flora
The normal red algal turf, consisting of Polvsiphonium, aelidium, C e n t r w and Spermothamnium spp. had almost entirely disappeared by March 1983 and was replaced by the brown alga Giffordia mitchelliac, which dominated the intertidal and splash zones (Laurie, 1983b). Giffordia mitchelliae has not been recorded before in Galapagos but may have been present in small quantities. It tolerates a wider range of temperatures than the red algae and was thus able to
a
I
1981
1
1982
1
1983
1984
I
7
1985
1986
1985
1986
YEAR
'"1
O
1981
1982
i 1983
1984
YEAR
Fig. 3. The percentage of animals that survived to the end of each year, adults according to sex ( 3 4 and juveniles according to cohort, or year of emergence (3b).
369 colonize the sites that the red algae had occupied previously (J. Price, pers. comm.). In vitro digestibilities of algae with sheep rumen fluid (Tillev and Terry, 1963) showed that the organic matter digestibility of the Gjffordia was 21% compared with a mean of 78% for the Chaetomorpha and Enteromorpha preferred red algal species and 64% for the green algae spp.) (Laurie and Uryu, in prep.). Brown algae generally contain more cellulose than red and green algae (Black, 1955; Paterson, 1984) and thus would be expected to be more difficult to digest. The sea level and sea-surface temperatures in Galapagos had returned to the normal range for the time of year by September 1983, and the dense mat of Giffordia algae had begun to disappear by early November and was almost completely gone by December, being slowly replaced by red algal turf, and Chaetomorpha spp. The response of the marine iguanas was almost immediate: there was no more than normal mortality after August 1983 and the surviving adults had recovered to near their 1981 condition by November 1983 (Laurie, 1987). 4.3 Growth r a m Mean annual relative increase in snout-vent length decreased in adults from 6.8% in 1981-82 to 0.5% in 1982-83, increased slightly to 1.2% in 1983-84, most of the growth being after August 1983, and then returned to pre-ENS0 levels of 6.3% in 1984-85. Juvenile growth rates recovered much faster after the ENSO, 1983 hatchlings grew faster than 1981 hatchlings in their first year. Figure 4 shows the growth rates of males and females separately; the mean annual increase in SVL (April to April) is plotted against SVL at the beginning of the year for 29 one cm size classes. The data include all records of individuals captured at the beginning and the end of the year under consideration; younger animals that did not reach 225 mm SVL during the study were not sexed, these juveniles of unknown sex are included in both curves. For some individuals there are data for only one year; for others up to six years. Points on the graphs are the means of between 8 and 178 individuals' growth rates. There was a significant decrease in growth rates from 1981-82 to 1982-83 in all size classes (1 1.2 > t > 3.4; p c 0.01), followed in 1983-84 by increased growth rates in hatchlings (16.2 > t > 3.6; p < 0.001), but continued low growth rates in adults (t = 4.73; p c 0.001 for animals of more than 22 cm SVL). Adults of both sexes hardly grew at all between April 1982 and April 1983 (mean of 0.5% increase in SVL). Growth in the year April 1983 to April 1984 was concentrated in the last part of the year after the recovery of the algal flora. The 1984-85 growth rates exceeded the 1982-83 and 1983-84 rates in all size classes (3.0 < t < 13.2; p < 0.01). For juveniles the 1984-85 growth rates were considerably higher, more than double in some size classes, than the pre-ENS0 growth rates of 1981-82 (t = 18.7; p c 0.001 for animals of less than 22 cm SVL). However, adult growth rates in 1985-86 were similar to those of 1981-82, although juvenile growth rates were still considerably higher than in 1981-82. Juvenile growth rates in 1986-87 were lower than in 1985-86 (4.3 c t < 6.2; p c 0.01) and appear to be decreasing towards pre-ENS0 levels. Figure 5 shows the growth curves for each hatchling cohort from 1980-1986 (there were extremely few hatchlings in 1984 and none was measured on emergence). The highest first year growth was recorded for 1985 hatchlings and had decreased again for the 1986 cohort. The rapid
Fig. 4. Mean annual increment in SVL plotted against SVL at beginning of the year for males (4a) and females (4b), 1981-82 to 1986-87.
9
W
F.
P
R
0
.z
Q
4
D nh) v-Q v
2
-4
DO
=h)
n
0
A
0
A
h)
W
P
cn
n
rn v,
I-
b
I
rn
0)
U
MEAN GROWTH RATE CM SVL PER YEAR
-1
0 A
h)
W
I
6
P
-I - -
VI
'
MEAN GROWTH RATE CM S V L PER YEAR a,
0
4
w
370
37 1
0 C v)
1981
82
83
84
85
86
87
Year Fig. 5. Growth curves of hatchlings of 1980, 1981,1982, 1983, 1985 and 1986 cohorts. growth rates of the 1983 hatchlings resulted in that cohort having a greater mean size at 2 years of age than the 1981 hatchlings had at 3 years of age (Tables 2 and 3). There was also a considerable overlap in size in 1985 between 1982 and 1983 hatchlings; growth rates increased again after the slowed growth during ENSO, showing a plasticity of response to environmental conditions. Figure 6 shows the mean predicted growth in SVL for both sexes, based on the 1981-82 and the 1985-86 data on annual increments in SVL for each one cm size class (Fig. 4). It shows a clear shift to the left, so that both males and females could be expected to reach the mean size of 1981 breeding animals about two years earlier than in 1981.
4.4 Condition The relationship between log SVL and log WT was examined in order to find a suitable index of condition applicable across sex and size classes. Simple regression and principal component analyses both showed that WT/SvL3 varied little over sex and size classes at the time of measurement (Wong, 1985). Figure 7 shows the changes in mean condition of adult iguanas (SVL > 225 mm) on Santa Fe over the seven years 1981-1987. Results for other islands, with the exception of the small islet CaamaAo, are similar (Table 4). The clear trough in 1983, when many animals lost almost 50% of their body weight before either dying or recovering, was followed by a sharp rise in condition to well above the pre-ENS0 level, and then a return to that level. There was a marked decrease in condition of animals well before the 1982-83 ENSO and before the first negative sea surface temperature anomaly of the event (-1.OOC) was recorded in Galapagos by A.
372 TABLE 2 Age specific snout-vent lengths of hatchlings of different cohorts, 1981-1987. Cohort
Emergence
1 yr
2 P
3 P
4 P
5 P
6 P
1981 SVL s.e. n
a 106.7 (0.2) 643
a 136.1 (0.4) 282
a 170.4 (0.8) 181
a 197.7 (1.6) 41
a 247.2 (2.0) 37
a288.5 (3.7) 22
297.0 (5.4) 22
1982 SVL s.e. n
a 107.0 (0.2) 404
a 137.4 a 167.9 (3.1) (0.6) 140 14
b 217.9
(2.5) 19
b 261.2 (3.0) 12
a 286.7 (5.2) 7
1983 SVL s.e. n
ab 107.5 (0.2) 422
b 146.0 b 204.1 (0.8) (1.2) 137 118
c 252.4 (1.6) 94
c 273.1 (2.3) 75
Almost no reproduction took place in 1983-84
1984-1985 SVL s.e. n
a 107.2 (0.1) 743
c 162.0 (0.7) 222
1986 SVL s.e. n
b 108.2 (0.4) 109
d 149.0 (0.7) 75
c 199.4 (1.1) 156
Within years, cohorts with different letters (a,b,c, etc.) have significantly different snout-vent lengths (p < 0.001).
Matson (pers. comm., 1983) in May 1982. There is an inverse correlation (rs = -0.6, p < 0.05) between the condition index (Fig. 7) and the mean monthly sea surface temperature anomaly between 1981 and 1987 (Fig. 8).
4.5 Effects on reproduction The most obvious effect of ENS0 on reproduction was the almost complete failure of breeding in the 1983-84 season. On Santa Fe, territorial defence was less intense than normal, with only 25% of the normal number of territorial males and fewer extended fights (Laurie, 1984), but the main difference from previous years was in the reactions of the females, who consistently avoided the males' approaches. Not a single copulation was seen during one month of observation, although during the same period of observation in each of the other four years of the study between 55 and 70 copulations were recorded. In the 1981-82 and 1982-83 seasons the males finished mating by early January, but in the 1983-84 season temtorial defence continued until early March, with the difference that the territory holders fed more frequently and lost significantly less weight than in a normal year.
373 TABLE 3
Age specific weights of hatchlings of different cohorts, 1981-1987. Cohort
Emergence
1 yr
2yr
3yr
4yr
5yr
6yr
1981 WT(g) s.e. n
a 58.8 (0.3) 643
a 121.1 a203.4 (3.9) (1.3) 282 181
a436.8 (12.7) 41
a 851.9 a 1,288.6 1,413.6 (21.3) (44.7) (81.6) 37 22 22
1982 WT s.e. n
a 59.4 (0.4) 404
b 105.5 b 251.4 (2.1) (12.8) 140 14
b 584.7 (19.9) 19
a 922.5 a 1,271.4 (37.0) (102.9) 12 7
1983 WT s.e. n
b 56.5 (0.4) 422
c 166.8 c468.4 (8.1) (2.8) 137 118
c 826.4 b 1,101.2 (26.0) (15.7) 94 75
Almost no reproduction took place in 1983-84 1984-1985 WT s.e. n
c 62.9 (0.3) 743
d 212.7 (2.8) 222
1986 WT s.e. n
c 63.5 (0.7) 109
e 183.7 (4.4) 75
d 423.5 (6.8) 156
Within years, cohorts with different letters (a,b,c, etc.) have significantly different weights @
374 n
E
s
Males
1985-86 data
Fern--a I es __--------------------
1
1
1
1
1
1 2
1
3
4
5
6
1
l
7 8
1
1
1
1
1
1
9 1011121314
Age / Years Fig. 6. Mean predicted growth in SVL for males and females based on 1981-82 and 1985-86 data.
TABLE 4 Mean index of condition, measured as WT/SVL3 (kg m-3) for each island at each visit. Island
n
Santa Cruz 235 Caam~o 233 Plaza Sur 48 Santiago 35 Genovesa 78 Marchena 60 Pinta 80 Isabela 251 Femandina 415 Floreana 31 EspaA o1a 78 Santa Fe 4,168 SeymourNorte 21
1981
1982
1983
52.5 52.9 55.5 55.7 56.1 55.3 54.9 58.8 57.7 54.9 54.2 55.0 57.5
52.0 51.5 48.8 48.9
57.1 49.6
43.8 50.1 41.7 42.5 45.4 41.3 41.7 43.1 39.5
47.1
39.9 37.2
YEAR 1984 55.2 53.9 54.5 56.9
54.0 53.6 55.8
1985
1986
1987
54.3 55.0 54.7
51.9 44.3
62.1 64.7 60.5 58.8 58.7 55.0 53.5 57.0 60.3
54.2 55.3 54.1
54.2
Each figure is the mean of the condition index of three classes of iguanas: adult males, adult females and juveniles. sd in each case is between 1.5 and 13.5 with overall sd = 10.8. classes being almost completely eliminated and hardly any breeding taking place during the 198384 post-ENS0 breeding season. Since the population crash there have been large increases in rates of growth, survival and reproduction, with growth rates and fecundity exceeding pre-ENS0 levels, but survival rates still lower. Male mortality was greater than female mortality under the
375
70-
60X
a
3
c
I
.-.w 50.-
'C
m
c
0
0 40-
Fig. 7. Condition measured as WT/SVL3 (WT in kg, SVL in m) for adult iguanas on Santa Fe between April 1981 and February 1987. conditions of low food availability, as has been found in several species of mammals and birds, for example, in pinnipeds (Limberger, this volume) and deer (Cervus elaDhus)(Flook, 1970), and this is analysed and discussed by Laurie and Brown (in press a, b). The evidence from post-mortem examination suggests that iguanas died slowly of starvation after their preferred food species of red algae were replaced by the brown alga Giffordia ~ t c h e l l i a s (Cooper and Laurie, 1987; Laurie and Uryu, in prep.). The iguanas showed typical signs of starvation and gut impaction, and death by ingesting toxins was considered unlikely. W. Fenical (pers. comm., 1983) has found toxins in Bifurcana. ' Laurencia and O c h t a spp., all of which are avoided by marine iguanas, but not in Giffordia spp. 5.1 Indirect comaetition for food Marine iguanas are limited by the tides in the time they can spend feeding. On islands where iguanas are larger, more sub-tidal feeding takes place (Trillmich and Trillmich, 1986), but juveniles and many adult females are restricted to intertidal feeding. Whenever the sea is rough, animals go without food, sometimes for several consecutive days. Food type, abundance and ease of grazing
81
1982 1983
1984
1985
1986 1907
YEAR Fig. 8. Mean monthly sea-surface temperature anomaly at Academy Bay, Santa Cruz between 1981 and 1987.
TABLE 5 Estimated proportion of females that bred in each year. Year
N
% Females
nested 1981-82 1982-83 1983-84 1984-85 1985-86
99 157 202 206 228
41.1 40.4 1.0 87.9 86.1
90% confidence limits 34.2-52.2 30.9-58.0 78.2- 100 75.0-97.2
varies considerably between islands and between sites on the same island, and probably accounts for a large amount of the size differences between populations (Laurie, 1983a). On Santa Fe in the ENS0 year, growth rates and survival were higher on some parts of the coastline, where grazing was easier and the lower intertidal zone approachable, than on steeply shelving unsheltered
377 neighbouring areas (Laurie, 1987; Laurie and Brown, in press b). The most likely explanation for the increase in rates of juvenile survival, growth and reproduction after the 1982-83 ENSO is reduction in competition for food after the return of the normal food species and the disappearance of the colonizing Giffordia sp. Marine iguanas do not fight at feeding sites and there is generally abundant food in the grazing areas. However, individuals are limited in their feeding times by the state of the tide, the strength of the swell and the temperature of the water (Carpenter, 1966; White, 1973) and there is an advantage in occupying feeding sites easy of access, nearest to the shore or particularly well sheltered from the waves, which can knock iguanas off exposed sites. So there is indirect competition for food in that individuals have to feed on inferior feeding sites if the best ones are already occupied, and they may have to spend relatively more of their foraging time in reaching the feeding site and clinging on as waves pass over them, losing more heat in the process and further reducing the time available for feeding. The inshore and sheltered feeding sites are often grazed down very short like a lawn, whereas the lower, more exposed sites have a lusher growth of algae. There appears to be an optimum level at which to feed for each state of the sea: low enough to avoid the heavily cropped sites yet high enough to avoid spending too much time and energy searching and holding on in heavy surf. On Santa Fe there is relatively little sub-tidal feeding (Trillmich and Trillmich, 1986), but similar effects could result in indirect feeding competition in sub-tidal feeders also. By 1985 the algae had returned to the normal species composition, and the fewer iguanas grazing could take more food in a shorter time than before the 1982-83 ENSO. Population size has returned to near the pre-ENS0 level by 1989, although compared to 1981 the population composition is heavily skewed towards juveniles (T. Dellinger, pers. comm., 1989). 5.2 Growth rates and aee at first reproduction The growth curves of iguanas of both sexes were shifted about two years forward immediately after the 1982-83 ENSO, leading to a two year decrease in the age of first breeding. This flexibility in growth rates, typical of reptiles (Dunham, 1978; Vogel, 1984) and the year of first breeding being related to size rather than age, means that there is relatively fast recovery of population size and structure after periods of high mortality. As the population density increased again, particularly in juveniles, growth rates have begun to decrease. 5.3 Dominant cohorts The almost complete loss of the 1982 cohort, and heavy losses to the 1980 and 1981 cohorts, and the failure to breed in 1984 have resulted in the 1983 cohort dominating the recruitment from the five years 1980-1984. This has important implications for investigation of whether individual reproductive strategies have adaptive significance, or whether there is so much uncertainty in the environment, both in terms of feeding conditions (Laurie, 1987), predation (Laurie, in prep.) and distribution of females between mating territories (Laurie, in prep.), that any variation is a direct response to environmental conditions (Howe, 1978; Price, 1985).
378 5.4 costs 0f breeding It appears that marine iguanas are limited in growth rates and frequency of reproduction by food availability. Adult females did not recover breeding condition in time to breed in consecutive years, and adult males that mated with high numbers of females in one year often spent the next year as a non-territorial male and returned to their territories after two years (Laurie, 1984). Skipping opportunities for reproduction has been recorded for a large number of reptiles and amphibians (Ballinger, 1977; Bull and Shine, 1979). 5.5 Other effects o f the 1982-83 ENSO in Galapapm Land animals thrived during the ENSO (Gibbs et al, 1984; Gibbs and Grant, 1987; Grant and Grant, 1987), but seabirds and sea mammals all suffered high mortality, failure of breeding or loss of whole cohorts (Trillmich and Limberger, 1985; Gibbs et al., 1987; also see contributions by Duffy, Limberger, and Smith, this volume.). The 1982-83 event was exceptional (Cane, 1983; Glynn, 1988) but nevertheless important in that repeated Occurrences within a relatively short time would be expected to lower the average population size of affected species, and could be critical for species with small, isolated populations with a slow recovery rate such as some fishes and corals (Glynn, 1988). The marine iguanas have shown a rapid recovery, but the fur seals are considerably slower, and more vulnerable to repeated occurrences of strong ENSO events (F. Trillmich, pers. comm.; Limberger, this volume).
5.6 Pole of E P The 1982-83 ENSO event has had profound effects on the population size and composition, but other, weaker events have not been shown to affect the iguanas in the same way. No widespread mortality of marine iguanas was recorded by scientists or others during previous ENSO events, but there is evidence that some events led to a decline in condition and possibly, therefore, reduced reproduction (G.K. Trillmich, pers. comm.). Certainly ENSO events must be considered as potential regulators of population density in marine iguanas. There was another, weaker ENSO event in early 1987 that led to a reduction in the standing crop of red algae and the ' in February, the first reappearance, at a few places high on the shoreline, of Giffo& record since 1983. However, no increase in mortality has been recorded, so the event was presumably tea brief to affect the marine iguana population. 6 ACKNOWLEDGMENTS I am very grateful to the Galapagos National Park Service (GNPS) for permission to work in Galapagos and to the Charles Darwin Research Station (CDRS) for logistic support. M. Cifuentes, F. Cepeda and H. Ochoa of the GNPS and F. Koster and G.Reck of the CDRS and their staff were always ready to help. Don Ramos and the crews of many boats were invaluable in supplying me with water and food on Santa Fe. I thank J. Marshall, H. Uryu, R. and M. Zavala, D. Watling, the late D. Villalba, M. Eckstein, C. Fairhurst, T. Woollard, M. Wells, D. Hams, B. Iglesias, A. Dik, E. Cruz, G. Molina, C. Cevallos, A. Balmford, F. Trillmich and T. Dellinger for help in the field, and D. Brown for help with analysis. The research was funded by The
379 Leverhulme Trust, The Royal Society and the Max-Planck Gesellschaft, and the data analysis was carried out at the Large Animal Research Group of the Department of Zoology, University of Cambridge. This is contribution number 504 of the Charles Darwin Foundation. 7 REFERENCES Arntz, W. E., 1986. The two faces of El Niiio. Meeresforschung, 31: 1-46. Ballinger, R. E., 1977. Reproductive strategies: food availability as a source of proximal variation in a lizard. Ecology, 59: 628-635. Barber, R. T. and Chavez, F. P., 1983. Biological consequences of El Niiio. Science, 222: 1,2031,210. Barber, R. T. and Chavez, F. P., 1986. Ocean variability in relation to living resources during the 1982-83El Niiio. Nature, 319: 279-285. Bell, T., 1825. A new genus of Iguanidae. Zool. J., 2: 204-208. Black, W. A. P., 1955. Seaweed and their constituents in foods for man and animal. J. SOC. Chem. & Ind., 74: 1,640-1,645. Bull, J. J. and Shine, R., 1979. Iteroparous animals that skip opportunities for reproduction. Am. Nat., 114: 296-303. Cane, M. A., 1983. Oceanographic events during El Niiio. Science, 222: 1,189-1.194. Carpenter, C. C., 1966. The marine iguana of the Galapagos Islands, its behaviour and ecology. Proc. Calif. Acad. Sci., 4th Ser., 34(6): 329-376. Chavez, F. P., 1987. The annual cycle of SST along the coast of Peru. Trop. Ocean-Atmos. Newslet., 37: 4-6. Colinvaux, P. A., 1984. The Galapagos climate: present and past. In: R. Peny (Editor), Key Environments - Galapagos, Pergamon Press, Oxford, pp. 55-69. Cooper, J. E. and Laurie, W. A., 1987. Investigation of deaths in marine iguanas on Galapagos. J. Comp. Pathol., 978: 129-136. Dunham, A. E., 1978. Food availability as a proximate factor influencing individual growth rates in the iguanid lizard Scelopom merriami. Ecology, 59: 770-778. Flook, D. R., 1970. A study of sex differential in the survival of wapiti. Can. Wildl. Serv. Rep., Ser. No. 11, Ottawa, 71 pp. Gibbs, H. L. and Grant, P.R., 1987. Ecological consequences of an exceptionally strong El Niiio event on Darwin's finches. Ecology, 68: 1,735-1,746. Gibbs, H. L., Grant, D. R. and Weiland, J., 1984. Breeding of Darwin's finches at an unusually early age in an El Niiio year. Auk, 101: 872-874. Gibbs, H. L., Latta, S. C. and Gibbs, J. P., 1987. Effects of the 1982-83 El Niiio event on blue footed and masked booby populations on Isla Daphne Major, Galapagos. The Condor, 89: 440-442. Glynn, P. W., 1988. El Niiio-Southern Oscillation 1982-83: nearshore population, community and ecosystem responses. Annu. Rev. Ecol. Syst., 19: 309-345. Grant, P. R. and Grant B. R., 1987. The extraordinary El Niiio event of 1982-83: effects on Darwin's finches on Isla Genovesa, Galapagos. Oikos, 49: 55-66. Houvenaghel, G. T., 1978. Oceanographic conditions in the Galapagos Archipelago and their relationships with life on the islands. In: R. Boje and M. Tomczak (Editors), Upwelling Ecosystems. Springer Verlag, Berlin, pp. 181-200. Houvenaghel, G . T., 1984. Oceanographic setting of the Galapagos Islands. In: R. Perry (Editor), Key Environments - Galapagos. Pergamon Press, Oxford, pp. 43-54. Howe, H. F., 1978. Initial investment, clutch size and brood reduction in the Common Grackle (Ouiscalus-L.). Ecology, 59: 1,109-1,122. Kiladis, G. and Diaz, H. F., 1984. A comparison of the 1982-83 and 1877-78 ENS0 events. Trop. Ocean-Atmos. Newslet., 25: 7-8. Laurie, A., 1982. Marine iguanas - where have all their babies gone? Noticias de Galapagos, 35: 17-19. Laurie, W. A., 1983a. Marine iguanas in Galapagos. Oryx, 17: 18-25. Laurie, W. A., 1983b. An ill wind for iguanas. New Sci., 100: 108. Laurie, W. A., 1984. Marine iguanas: the aftermath of El Niiio. Noticias de Galapagos, 40:9- 11. Laurie, W. A., 1987. Marine iguanas and the Galapagos marine reserve. Oceanus, 30: 54-60.
380 Laurie, W. A. and Brown, D., In press a. Changes in annual survival rates of marine iguanas (AmblvrhynchyS)on Galapagos, and the effects of size, sex, age and fecundity in a population crash. J. Anim. Ecol. Laurie, W. A. and Brown, D., In press b. The factors affecting survival of marine iguanas (Amblvrhvnchus on Galapagos. J. Anim. Ecol. Paterson, I. W., 1984. The foraging strategy of the seaweed eating sheep of North Ronaldsay, Orkney. Ph. D. Dissertation, University of Cambridge. Philander, S. G. H., 1983. El Niiio Southern Oscillation phenomena. Nature, 302: 295-301. Pollock, K. H., 1981. Capture-dependent models allowing for age-dependent survival and capture rates. Biomemcs, 37: 521-530. Price, T., 1985. Reproductive responses to varying food supply in a population of Darwin's finches: clutch size, growth rates and hatching synchrony. Oecologia, 66: 41 1-416. Quinn, W. H., Neal, V. T. and Antunez de Mayolo, S. E., 1987. El Nifio Occurrences over the past four and a half centuries. J. Geophys. Res., 92: 14,449-14,461. Quinn, W. H., Zopf, D. O., Short, K. S. and Kuo Yang, R. T. W., 1978. Historical trends and statistics of the Southern Oscillation El Niiio, and Indonesihn droughts. Fish. Bull., 7(3): 663678. Rasmusson, E. M., 1984. El Niiio: the ocean/atmosphere connection. Oceanus, 27: 5-12. Robalino, M., 1985. Registros meteorologicos de la Estacion Cientifica Charles Darwin para 1982-83. In: G. Robinson and E. M. del Pino (Editors), El Niiio en las Islas Galapagos: El Evento de 1982-83. Charles Darwin Foundation for the Galapagos Islands, Quito, pp. 83-90. Tilley, J. M. A. and Terry, R. A., 1963. A two stage technique for the in vitro digestion of forage crops. J. Brit. Grassl. SOC.,18: 104-111. Trillmich, F. and Limberger, D., 1985. Drastic effects of El Niiio on Galapagos pinnipeds. Oecologia, 67: 19-22. Trillmich, G. K. and Trillmich, F., 1986. Foraging strategies of the marine iguana, Amblvrhvnchusgristatus. Behav. Ecol. Sociobiol., 18: 259-266. Vogel, P., 1984. Seasonal hatchling recruitment and juvenile growth of the lizard Anolis lineatopus. Copeia, 1984: 747-757. White, F. N., 1973. Temperature and the Galapagos marine iguana - insights into reptilian thermoregulation.Comp. Biochem. Physiol., 45A: 503-513. Wong, K. T., 1985. An investigation of growth and survival rates of marine iguanas on the Galapagos. Diploma in Statistics Dissertation, University of Cambridge. Wyrtki, K., 1985. Pacific-wide sea level fluctuations during the 1982-83 El Niiio. In: G. Robinson, and E. M. del Pino (Editors), El Niiio en las Islas Galapagos: El Evento de 198283. Charles Darwin Foundation for the Galapagos Islands, Quito, pp. 29-48.
38 1
THE GULF OF PANAMA AND EL NIB0 EVENTS: THE FATE OF TWO REFUGEE BOOBIES FROM THE 1982-83 EVENT
N.G. SMITH, Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic of Panama
ABSTRACT Smith, N. G . , 1989. The Gulf of Panama and El NiAo events: the fate of two refugee boobies from the 1982-83 event. Thousands of Peruvian and Blue-footed Boobies, birds that normally breed in highly productive cool water regions, fled the most severe El Nifio of this century and arrived at the Gulf of Panama during February 1983 in surprisingly good physical condition. Their behavior and physical condition were studied from February until April 1985, when almost none remained. Their occurrence in Panama suggested that the Gulf might be a refugium from the effects of El Niiio. The physical conditions of the Gulf of Panama at the time of historical El Nifios were considered. However, qualitative comparisons of three indicators o f upwelling and high productivity: sea surface temperature, total wind, and wind direction, for 71 years of data in Panama, revealed no relationship between these factors and the occurrence of an El Nifio. 1 INTRODUCTION Invasions of animals into areas outside their normal ranges are of interest to evolutionary biologists for they may provide information about the factors that limit the range of a species (Mayr, 1965).
This paper deals with
observations of an invasion of large numbers of two refugee boobies (Aves: Sulidae) that occurred in the Gulf of Panama during the 1982-1983El Niiio.
It
contrasts the differential survival of these boobies and speculates on their potential causes. This required a survey of the possible oceanographic and/or meteorological connection between the Gulf of Panama, a seasonally cool water upwelling zone, and the areas where these birds normally live. Is the dry season productivity of the Gulf of Panama influenced by these erratic El Nifio events and thus does it regularly act as a temporary refugium for animals normally associated with cool water, high productivity environments?
2 NATURAL HISTORY OF THE BOOBIES The two species considered herein are the Peruvian Booby, (Fig. 1) and the Blue-footed Booby,
pebouxii (Fig. 2).
refugee species such as the Inca Tern (Larosterna
varieFata
Other El Nifio
m)and Guanay Cormorant
382 (Phalacrocorax bouzainvillli) also occurred in the Gulf of Panama, but in much lower numbers and are not discussed here. The Peruvian Booby nests, often in very large colonies (105 adults), on islands and coastal cliffs in Peru and northern Chile (Nelson, 1978).
During
periods when the productivity of the Peruvian coast is reduced by the invasion of warm water and/or the upwelling of nutrient-poorwater (Hansen; Barber and Kogelschatz; this volume), the numbers of Peruvian Boobies crash from 4 million
Fig. 1
Fig. 2
Fig. 1. The Peruvian Booby, Fig. 2 .
varietxata.
The Blue-footed Booby, Sula nebouxii.
to well under 1 million birds (Nelson, 1978; Barber and Chavez, 1983; Tovar and Cabrera, 1985).
At the northern end of its breeding range, it overlaps with
the Blue-footed Booby on Isla Lobos de Tierra, Peru (Figs. 3 , 4 ) .
During the
non-breeding season it normally strays only as far north as the Gulf of Guayaquil, Ecuador (Fuentes, 1965).
Adults and juveniles differ in color.
Males are slightly smaller than females. It lays the largest clutches of any booby ( 3 - 4 egg clutches are not uncommon).
In Peru, most eggs are laid between
November and December, but eggs have been found throughout the year (Nelson, 1978). The degree of sexual size dimorphism is more extreme in the Blue-footed than the Peruvian Booby. The high-pitched calls of the smaller males differ markedly from the hoots of the females. The Blue-footed Booby has a relatively wide breeding range
--
from the Gulf of California, the Gulf of Panama
383 (population
100 pairs, pers. obs.), the Galapagos, and southward to its
overlap with the Peruvian Booby on Isla Lobos de Tierra, Peru. It is more tolerant of water temperatures higher than 22'C
than the Peruvian Booby, but
never nests far from upwelling zones where at least 23-25°C water is regular. Nowhere does it occur in colonies as large as those that characterize the Peruvian Booby.
90
10
85
80
75
70
0
5
5
0
3 Galapagos Islands 5
Lobos de Tierra
10
FLED Fig. 3. Presumed dispersion routes taken by Blue-footed Boobies during the 1982-83 El Nifio.
On Isla Lobos de Tierra, Peru, the two species forage in different zones:
Peruvian Booby to the south where water temperatures below 2 2 ° C are encountered, and Blue-footed Booby to the north into 23°C water (Duffy, pers. corn.).
The Peruvian Booby appears to feed primarily on anchoveta (Enaraulis
rin.eens), while the Blue-footed Booby is more catholic in its diet. 3 THE OCCURRENCE Approximately 7,000 Peruvian Boobies and 9,000Blue-footed Boobies arrived
384
90
85
80
75
70
10
5
3 0
.
Golopogos Islands
5 Lobos do Tierro
10
Sula variegata -
Fig. 4 . Presumed fates and dispersion routes of Peruvian Boobies during the 1982-83 El Nirlo. en masse in the Gulf of Panama during the second week of February, 1983 (local fishermen, pers,.comm.), approximately 4 months after the first sea surface warming in northern Peru (Tovar and Cabrera, 1985; Hughes, 1985; Ainley et al., 1988).
According to Hughes (1985), refugees of both species from the northern
colonies fled south to the Mollendo district (17"s) of Peru in October 1982. The northern Peruvian colonies of Blue-footed Boobies often have more than
10,000 individuals (Duffy, pers. comm.) so the Blue-footed Boobies that entered the Gulf of Panama may have included individuals from the Galapagos (Fig. 3 ) , and perhaps even individuals from Baja California. Figure 4 represents the apparent dispersal paths of Peruvian Boobies between October 1982 and March 1983 (Hughes, 1985; Ainley et al., 1988; Paul Sharf, pers. comm.). Ornithologists became aware of the invasion about one month later (Aid et al., 1985).
In April 1983, Montgomery (pers. comm.) collected 4 Peruvian and 3
Blue footed Boobies; all were fat and the three female Peruvian Boobies had large atretic follicles (Fig. 5 ) .
I presumed that these were degrading
follicles because of their "cheesy" texture and color, which was unlike the shiny orange follicles of developing eggs. The Blue-footed Boobies were all
385 males with recognizable testes, as was the case with the single male Peruvian Booby. About every two months, from April 1983 through June 1984, I visited all of the islands within 90 kilometers of Panama City to census boobies and to collect the dead birds.
I made three trips with the anchovy fleet to a
location 40 to 60 km east of Panama City where the anchovetas (Cetennraulis
Fig. 5. Physical condition of the female Peruvian Booby, April, 1983 in the Bay of Panama. Note the large amount of visceral fat (arrow 1) and recently atretic ovarian follicles (arrow 2 ) .
mvsticetus) spawned. There I observed the catch and the behavior of the birds. Roca San Jose, a major booby roost at the mouth of the Panama Canal, was visited on a biweekly basis to catch boobies at night. These birds were weighed to ca 10 g precision, and palpated for the presence of visceral fat pads. Dead birds were usually dissected but not weighed. The weights and general physical condition of birds varied greatly (Table 1).
Female Blue-footed Boobies had a mean weight during April-May of 1,860 g, with a range of 1,411 to 2,075 g. Males were lighter, with a mean weight of 1,488 g, but again the spread was wide.
Similarly female Peruvian Boobies had
a mean weight in April of 1,570 g and males 1,353 g.
I dissected visceral fat
pads from two birds, and these were ca 41 g for a female Peruvian (weighing 1,419 g) and 37 g for a male (1,353 g).
Blue-footed and Peruvian Boobies that
appeared to have recently died at their roost, typically had lost between 31% and 37%, respectively, of their mean weights in April. None of the six female Blue-footed Boobies that were examined in early 1983 had atretic follicles as were found in the Peruvian Boobies.
Possibly, the
Blue-footed Boobies had deserted their breeding areas earlier than did the
386 TABLE 1 A comparison of two El Nifio refugee populations of Sula spp. in the Gulf of Panama 1983-1985. Date
Peruvian Booby (S, varieeata)
Apr. 83
6,000-7,000inds. very fat (N-30) ovaries large, atretic (N-30) testes smali (N-22) no breeding
9,000-11,000inds. some fat (N-11) weight - normal gonads vary (N-11) some breeding (residents?)
same
weight down (N-7) gonads small (N-7)
less, 7,000 inds. lean (concave breast muscles) (N-13) weights vary (N-13) a few dead (N-10)
Aug. 83
1,300-2,000inds. weights down (N-10) many dead (N-173)
ca 2,000 inds. Eights down (N-13) a few dead (N-10)
Dec. 83
600-1,200inds. starvation more dead (N-52) gonads small (N-22)
less than 1,000 inds. weights normal (N-20) some breeding (N-51)
Apr. 84
< 300 inds. weights stable (N-15) more dead (N-15) lean gonads small (N-15)
650-900 inds. some breeding
Jun. 84
none observed, remains of dead birds in roosts
Jun. 83
no visible fat (N-7)
I
ca 75 inds.
Nov. 84
22 pairs breeding
Apr. 85
report 60 inds. breeding?
Peruvian Boobies. Adults and juveniles differ in color, and unlike the Peruvian Boobies, a large proportion of the Blue-footed Boobies were immature. The Peruvian Boobies apparently reabsorbed their yolks for no female after 20 June 1983 was found with developed ovaries. Most of those seen in Panama in 1983 were adults. During November to February, there is no legal anchovy fishing in Panama, but 13 vessels regularly and illegally fished the shoreline 40-60 km east of Panama City from February to May 1983. Figures 6a, b, and c show a typical scene. Both boobies, and the local Brown Pelicans (Pelicanus occidentalis)
387
Fig. 6a
Fig. 6b Figs. 6a, b, c. This purse seine anchoveta boat is concentrating the fish and driving them to the surface where they became available to the birds present during February-April 1983, Bay of Panama. followed the boats in swarms. The anchoveta spawning grounds are in shallow water, with a mud bottom and high turbidity. Once the fish were concentrated by the
100 m net, boobies
and pelicans fed in a frenzy, which lasted until the nets were drawn in. The fishing vessels switched their activities to the Gulf of San Miguel, Darien, and to deep water, open bay fishing. In the absence of this concentrated inshore fishing, the Peruvian Boobies at Roca San Jose remained and foraged only
388
Fig. 6c
Fig. 7. Peruvian Boobies on Roca San Jose, Bay of Panama, July, 1984. The birds oriented their dark backs to the sun. At this stage, all were starving and never left the rock.
sporadically during the period from August to November 1983 (Fig. 7 ) . Although I had seen them feeding independently in April and May, their dominant feeding behavior was with the fishing fleet. Because of the Peruvian Booby’s habit of diving in flocks, this bird appeared better adapted to exploiting food at artificially high concentrations than was the Blue-footed Booby. Although there were more Blue-footed Boobies than Peruvian Boobies in April 1983, by December their numbers were about equal and a tenth of the April numbers (Table 1). I found many more dead Peruvian Boobies than Blue-footed
389
Fig. 8. Physical condition of female Peruvian Booby, June, 1983 in the Bay of Panama. Note the absence of fat and the degeneration of the breast muscle (arrow).
Boobies (Fig, 8).
By December, Blue-footed Boobies were breeding on various
islands in the Gulf of Panama in normal numbers (M 60 pairs).
Importantly,
Blue-footed Boobies were seen again at Galapagos colonies, and by December 1983 were breeding at about 70% of their pre-El Nitio population levels (Valle, 1985).
Duffy and Merlen (1986) made two passages through the Galapagos in
January 1985 and counted more birds than had been recorded in similar transects in 1982. Fragmentary information (Gonzalo Castro, pers. corn.) indicated a very different story for Peru, for while the cool water had returned by late 1983, the Peruvian Boobies were breeding at about 10% of their former numbers (November 1983).
This suggests that most of the Peruvian Boobies probably
died. The refugium of the Gulf of Panama was better suited for the Blue-footed Boobies because, although they undoubtedly profited by the artificial prey concentrations produced by fish nets, they were more adept at solo foraging when fish were more dispersed, as was the case after May.
4 EL NIB0 EVENTS AND THE GULF OF PANAMA Recent reviews have suggested that many unusual biological phenomena occurring in 1982-83were related directly to that El Nitio (La Cock, 1986; Ainley et al., 1988; Duffy, this volume).
However, I could not detect any
significant correlations with the biology of the Gulf of Panama aside from drought on the isthmian mainland (Rand and Rand, 1982; Leigh et al., this volume). Data on physical conditions that might demonstrate a relationship with the occurrence of El NiRo were compiled. The wind data from 1915 to 1954 were digitized from the plots in Shaefer et al. (1958), Fig. 9. The wind data from
390
n
0 0 0
s
7
45
I-
30
Y
v
n
15
Z
3
0 1920
1940
1960
1980
Fig. 9. Relationship between trade wind intensity during January to April at Balboa, Panama, and the relative strengths of El Nifio events. Strong ( 4 ) events are in black. Data from 1915 to 1954 are digitized from Schaefer et al. (1958).
1955 to 1985, and all sea surface temperatures (SST), were provided by the Panama Canal Commission. Figure 10 plots Balboa SST and with largely subjective rankings of relative strengths of El Nifio events (Quinn et al., 1978) during 1915-1985. One of the earlier theories suggested that El Nifio development is related to reduced trade wind intensity that allows warm water to move southward from the Panama Bight into Peruvian waters (Bjerknes, 1961, 1966).
Figure 9 plots the northern component of the trade winds in Panama and
El Niiio strengths during 1915-1985. Note the decreasing trend in the wind data over time. This probably resulted from data quality problems (e.g., deterioration and/or relocation or changes of the anemometer) rather than an actual change in wind speed. Panama Canal Commission records indicate that a change from four to three cup anemometers was made in 1927 (M.S. Hart, pers. comm.).
I found no clear trends linking either wind speed or SST with historical El Nifio events. Kwiecinski and Chial (1987) also found little correlation between northerly wind passage and SST anomalies during El Nifio years, but correlations between Balboa sea level and Chicama, Peru SST anomalies were high during El Niiio years. However, Kwiecinski et al. (1988) did find a strong negative correlation between northerly wind passage and Balboa SST anomalies during the 1982-83 El Nifio event. The accepted theory is that it is the N.E. trade winds which, not impeded by any mountain barrier at the Isthmus of Panama, result in upwelling of the underlying colder, nutrient enriched water (Fleming, 1940; Schaefer et al.,
39 1
30 A
0 0
W
n' 25
2 w
+ 20 I-
15 1920
1940
1960
1980
Fig. 10. Relationship of average maximum and average minimum water temperatures during the January-Aprildry season in the Bay of Panama, and the relative strengths of El Nitlo events. Strong ( 4 ) events are in black.
1958; Smayda, 1966; Forsbergh, 1969; Glynn and Stewart, 1973).
I found no
correspondence between the strengths of El Niiio events and any aspect or vector of the northerly trade winds (Fig. 9).
The year 1985 had the coldest water
temperature values since 1915, yet the winds, from any vector, were approximately the same or less than during years with high water temperatures. Glynn and Stewart (1973) pointed out that the complex bottom profile of the Gulf of Panama, coupled with a number of mountain-like islands, could in part be responsible for the very complex pattern of sea water temperatures. They also noted that the sea water temperatures are taken at a dock near Balboa that is influenced by fresh water drainage. Yet, aside from the data imperfections, some other factor(s) besides local trade wind stress is having an effect on the water temperature of the Gulf of Panama. 5 CONCLUDING REMARKS
In February, 1983 about 7,000 Peruvian Boobies and 9,000 Blue-footed Boobies entered the Gulf of Panama, apparently fleeing the disastrous effects of the 1982-83 Pacific warm water event. Both species were in good condition, and the Peruvian Boobies had almost developed into full reproductive condition, but had begun to regress. They encountered a concentration of fish that was in part normal for the upwelling Gulf of Panama, and partly peculiar because fishing had concentrated the prey. Shortly after the fishing effort ceased, the boobies began to lose weight and die. Many more Peruvian Boobies died,
392 apparently because they are specialized at feeding on dense shoals of fish, while the Blue-footed Boobies are less so. Evidence from their normal breeding areas in the following years suggests that most of the Blue-footed Boobies survived and returned after December 1983. The Peruvian Boobies did not. Most of the population that entered the Gulf of Panama died. An examination of meteorological and water temperature data for the Gulf of Panama from 1915 to 1985 revealed neither a correspondence with El Nitlo events nor revealed any correspondence between El Nifio events and the upwelling in the Gulf of Panama. Additionally, the local influence of trade winds (recorded at Balboa) apparently is not directly responsible for cool water upwelling and high productivity, which affects a broad spectrum of organisms in the Panama area. Montgomery and Martinez (1984) presented information suggesting that Brown Pelican reproductive activity was proximately influenced by water temperature. Plankton, corals, and boobies are in one way or another affected by these physical parameters. I believe that we have not yet attained an understanding of this local oceanographic event, let alone an understanding of a phenomenon as vast as the El Nirlo/Southern Oscillation. 6 ACKNOWLEDGMENTS My colleague, Donald M. Windsor, kindly aided me in data interpretation and plotted all statistical matter. J. Brady of the Panama Canal Commission provided the wind and water temperature information. Gustavo Justines provided information and data on the anchovy fishery. A. Velarde, L. Cruz, J. Bryant, A. Rodaniche were captains of various vessels. N.E. Smith and M.L. Jimenez typed the manuscript. I thank J. Karr, D. Duffy, A.S. Rand and two anonymous reviewers for providing comments. 7 REFERENCES Aid, C.S., Montgomery, G.G. and Mock, D.W., 1985. Range extension of the Peruvian Booby to Panama during the 1983 El Nirlo. Colonial Water Birds, 8: 67-68. Ainley, D.G., Carter, H.R., Anderson, D.W., Briggs, K.T., Coulter, M.C., Cruz, F., Cruz, J.B., Valle, C.A., Fefer, S.I., Hatch, S.A., Schreiber, E.A., Schreiber, R.W. and Smith, N.G., 1988. ENS0 effects on Pacific Ocean marine bird populations. In: H. Ouellet (Editor), Proceedings of the XIX International Ornithological Congress. National Museum of Natural History, Ottawa, Canada, pp. 1,747-1,758. Barber, R.T. and Chavez, F.P., 1983. Biological consequences of El Niiio. Science, 222: 1203-1210. Bjerknes, J., 1961. El Nitlo study based on analysis of ocean surface temperatures 1935-57. Bull. Inter-her. Trop. Tuna Comm., 5: 219-303. Bjerknes, J., 1966. Survey of El Nitlo 1957-58 in its relation to tropical Pacific meteorology. Bull. Inter-her. Trop. Tuna Comm., 12: 25-86. Duffy, D.C. and Merlen, G. 1986. Seabird densities and aggregations during the 1983 El Niiio in the Galapagos Islands. Wilson Bull., 98: 588-591.
393 Fleming, R.H., 1940. A contribution to the oceanography of the Central American region. Proc. Sixth Pac. Sci. Cong., 3: 167-175. Forsbergh, E.D., 1969. On the climatology, oceanography and fisheries of the Panama Bight. Bull. Inter-her. Trop. Tuna Comm., 14: 49-259. Fuentes, H., 1965. Informe sobre el viaje efectuado a Guayaquil con el proposito de realizar observaciones de aves guaneras. Inf. Esp. Instit. Mar, Peru-Callao,4: 1-11. Glynn, P.W. and Stewart, R.H., 1973. Distribution of coral reefs in the Pearl Islands (Gulf of Panama) in relation to thermal conditions. Limnol. Oceanogr., 18: 367-379. Hughes, R.A., 1985. Notes on the effects of El Nifio on the seabirds of the Mollends district, southwest Peru, in 1983. Ibis, 127: 385-390. Kwiecinski, B. and Chial, B., 1987. Manifestations of El Nifio in the Gulf of Panama. Trop. Ocean-Atmos. Newsl., 42: 7-9. Kwiecinski, B., Chial, B. and Torres, A . , 1988. El Nifio and post El Nido (19821986) in the Gulf of Panama. Trop. Ocean-Atmos. Newsl., 44: 7-8. La Cock, G.D., 1986. The Southern Oscillation, environmental anomalies and mortality of two southern African seabirds. Climatic Change, 8: 173-184. Mayr, E., 1965. The nature of colonizations in birds. In: H.G. Baker and G.L. Stebbins (Editors), The Genetics of Colonizing Species. Academic Press, New York, pp. 29-47. Montgomery, G.G. and Martinez, M.L., 1984. Timing of Brown Pelican nesting on Taboga Island in relation to upwelling in the Bay of Panama. Colon. Waterbirds, 7: 10-21. Nelson, J.B., 1978. The Sulidae. Oxford University Press, Oxford, 1,012 pp. Quinn, W.H., Zopf, D.O., Short, K.S., and Yang, R.T., 1978. Historical trends and statistics of the Southern Oscillation, El Niiio, and Indonesian droughts. Fishery Bull., 76: 663-678. Rand, A.S., and Rand, W.M., 1982. Variation in rainfall on Barro Colorado Island. In: E.G. Leigh, A.S. Rand, and D.M. Windsor (Editors), The Ecology of a Tropical Forest, Seasonal Rhythms and Long-Term Changes. Smithsonian Institution Press, Washington, D.C. pp. 47-59. Schaefer, M.B., Bishop, Y.M. and Howard, G.V., 1958. Some aspects of upwelling in the Gulf of Panama. Bull. Inter-Amer.Trop. Tuna Comm., 3: 79-131. Smayda, T.J., 1966. A quantitative analysis of the phyto-plankton of the Gulf of Panama. Bull. Inter-Amer.Trop. Tuna Comm., 11: 355-612. Tovar, H. and Cabrera, D., 1985. Las aves guaneras y el fenomeno "El Nifio". In: W.E. Arntz, A. Landa and J. Tarazona (Editors), El Fenomeno El Nido y su Impact0 en la Fauna Marina. Bol. Inst. Mar, Peru-Calla0 (volumen extraordinario), 181-186. Valle, C.A., 1985. Alteracion en las poblaciones del cormoran no volador, el pinguino y otras aves marinas en Galapagos por efecto de El Nifio 1982-83 y su subsecuente recuperacion. In: G. Robinson and E.M. del Pino (Editors), El Nifio en las Islas Galapagos: El Evento de 1982-83, Fundacion Charles Darwin para las Islas Galapagos, Quito, Ecuador, pp. 245-258.
This Page Intentionally Left Blank
395
SEABIRDS AND THE 1982-1984 EL NIRO-SOUTHERN OSCILLATION
DAVID CAMERON DUFFY Institute of Ecology, University of Georgia, Athens, GA 30602 ABSTRACT Duffy, D.C., 1989. Seabirds and the 1982-1984 El Niiio-Southern Oscillation The 1982-1984El Niiio and associated events affected seabirds in the Pacific and Atlantic oceans. Effects ranged from extralimital dispersal to nest desertion and adult mortality, and appeared most severe in the eastern Pacific upwellings off Peru and California, and in the central Pacific. Up to 85% of seabirds died in Peru. In contrast, effects were mild in the Caribbean Sea. Other areas such as the North Atlantic and Indian Oceans were apparently unaffected. El Niiio appears to be an important force shaping life-histories only for seabirds in the Peruvian upwelling, where El Niiio events are both frequent and severe. In other areas, with milder or less frequent events, responses to normal environmental variability, such as deferred breeding, readiness to abandon nesting efforts, and low reproductive effort, may be sufficient to buffer seabirds against all but the most severe El Nifio events. On an evolutionary scale, El Nifio, by generating massive extralimital dispersals or by exterminating colonies, may have caused disjunct ranges or even speciation in Pacific seabirds. Our ability to study the effects of El Niiio on seabirds is limited by the lack of long-term,intensive studies of seabird populations. Without such studies, our knowledge is likely to remain anecdotal and superficial. We also lack information on the environmental features to which seabirds respond during El Niiio. These may include sea and air temperature, rainfall, food abundance or parasite loads.
1 INTRODUCTION Much of the early work concerning El Niiio-SouthernOscillation (ENSO) resulted from efforts to protect Peruvian guano production from periodic decreases in yields caused by the emigration and mortality of guano-producing seabirds during such events (e.g. Murphy, 1925; Vogt, 1942; Schweigger, 1964; Jordan and Fuentes, 1966). ENSO was seen as an isolated, climatic aberration, unique to Peru (but see Vogt, 1940 for an early discussion of effects in other areas). More recently, after El Niiio events in 1972-1973 and 1982-1984,a global perspective has linked ENSO with a mosaic of devastating droughts, floods, and changes in storm-tracksand commercial fish stocks far beyond the borders of Peru (Quinn et al., 1978; Glantz and Thompson, 1981; Rasmusson 1985).
Such events have received increased attention with the aim of
understanding their global consequences and forecasting future occurrences (cf. Glantz, 1981). Although seabirds are no longer the reason for most ENSO research, they remain important indicators of the effects of such events at upper trophic
396 levels in marine systems (e.g. Jordan and Fuentes, 1966; Duffy et al., 1984; Schreiber and Schreiber, 1984). This paper summarizes the world-wide effects on seabirds of the events of 1982-1984and examines the implications of these effects for the evolution of seabird life histories. Two sets of events are documented: 1) those in the Pacific and Atlantic directly linked to changes in wind circulation and internal, marine waves (El Nifio), and 2 ) 'teleconnections', shifts in global atmospheric circulation responding to the main event in the Pacific (Cane, 1983; Rasmusson and Wallace, 1983; Hansen, this volume). The overall event is referred to as El NifioSouthern Oscillation (ENSO hereafter). Since our understanding of ENSO is only at its initial stages, some of the apparent effects on seabirds may prove to be coincidences. But, by casting a wider net, I hope to stimulate ENSO research, much as investigation of guano bird mortality in the past provoked research on El Nifio in Peru. 2. RESULTS 2.1 The Pacific Ocean Within the Pacific Basin, oceanographic events during ENSO are triggered by three main forces that result from changes in atmospheric circulation: surfacewater moving eastward along, or polewards from, the Equator; subsurface waves that displace cold water downwards, leading to the upwelling of warm water, poor in usable nutrients; and teleconnections. (i) Central Pacific During the 1983 ENSO, the usually dry islands of the equatorial central Pacific experienced heavy rainfall, storms, and high winds as warm surface-water moved eastward past them (Rasmusson and Wallace, 1983). Schreiber and Schreiber (1984, 1986) reported nearly total reproductive failure and a mass exodus of adult seabirds from Christmas Island
(01°52'N,
157O2O'W). Beginning in November 1982, burrowing and ground-nesting species such as shearwaters and petrels suffered flooding, either from heavy rains or from unusually high sea-levels.Arboreal nests of Black Noddies Anous minutus and Greater Frigatebirds Fregata minor were destroyed by rain. Nestling Lesser Frigatebirds F. ariel and Masked Boobies
dactvlatra weighed less than in
normal years. By October 1983, many species had resumed nesting but in greatly reduced numbers, suggesting that adult mortality or emigration had occurred, although at least one species, Crested Tern Sterna bereii, was unaffected (Ainley et al., 1988), perhaps because its inshore foraging environment suffered less change than offshore areas. Other species, such as Sooty Terns Sterna fuscata, apparently increased their subsequent breeding numbers. This may have resulted from greater synchronization of breeding efforts in the subsequent breeding season (Ainley et al., 1988). Such synchronization could
397 have arisen if most of the birds, having been prevented from breeding during ENSO, simultaneously found conditions suitable for reproduction. At French Frigate Shoals (23'45'N,
166'10'W)
and Jarvis Island (OO023'S,
16Oo02'W), Fefer (in Ainley et al., 1988) found a variable response to ENSO within the seabird community. Grey-backed Terns Sterna lunata and Masked
-
Boobies abandoned nests but returned in pre-ENS0 numbers the following year. Sooty Tern breeding numbers dropped by half. Black Noddy numbers did not decrease, but they fledged fewer and lighter young during ENSO. Nestling Redfooted Boobies Sula sula were also lighter. Breeding populations of Blackfooted Diomedea
and Laysan
immutabilis albatrosses, Red-footed
Booby, Brown Anous stolidus and Black noddies and White Terns &is
alba were
not affected during or after the 1983 ENSO, perhaps because ENSO conditions are not very different from those in their normal foraging habitats. (ii) Galauaeos Islands Earlier studies during ENSO events in Galapagos suggested that nesting failure occurred, but that adult mortality did not (Maridueila, 1977; Boersma, 1978; Harris, 1979). Coverage of the event in 1982 was sparse; however, observations during 1983 suggested a far more devastating event than any previously documented. Heavy rainfall, rapid vegetation growth, and flooding from higher sea-levels physically damaged nesting efforts (e.g. Rechten, 1985; Gibbs et al., 1987), while food apparently became less available, as suggested by adult mortality (Sosa, 1985) and decreased sizes of foraging aggregations (Duffy and Merlen, 1986). Numbers of two endemic species, Galapagos Penguin Suheniscus mendiculus and Flightless Cormorant Nannouterum
w,decreased to
23% and 51%, respectively, of 1980 levels (Valle, 1984).
By 1985, cormorant numbers had recovered completely, and penguin numbers, although still only 38% of 1980 counts, were increasing (Valle, 1986; Valle and Coulter, 1987). Other species experienced breeding failure: Waved Albatross Diomedea irrorata (Rechten, 1985), Greater Frigatebird (Hernandez and de Vries, 1985), Brown Pelican Pelecanus occidentalis, Blue-footed S . nebouxi and Masked S, d
m boobies (Gibbs et al., 1987), and Swallow-tailed Creagrus furcatus
and Lava Larus fuleinosus gulls (Valle, 1985). Many individuals of these species apparently emigrated or died, since they were not seen elsewhere in the islands (Valle, 1985) or in adjacent waters (Duffy and Merlen, 1986). Adult mortality was reported (Sosa, 1985; Valle, 1985), but its extent was unknown. The rapid increase of birds in sea-transectsfollowing the 1983 ENSO (Duffy and Merlen, 1986) suggests that, rather than dying, many of the birds dispersed and returned. In contrast to the negative effects, Magnificent Frigatebirds maenificens were relatively unaffected, with approximately half their
398 population breeding through the event (cf. Valle, 1985). Red-footed Boobies also bred without interruption (Merlen, 1985, but see Valle, 1985). These species normally feed in low productivity warm water areas,
so
that ENSO
conditions might represent an extension of normal foraging habitat. Breeding success of Dark-rumped Petrels Pterodroma Dhaeonveia increased in 1983 compared to 1981, but growth was slower, and birds took longer to fledge (Cruz and Cruz, 1985). Nest and burrow destruction, primarily by flooding or collapse, rose from 20% to 60%, but this was counterbalanced by the success of a rat control program that reduced predation from 24% to zero (Cruz and Cruz, 1985). The apparent increase in overall success may have had little to do with ENSO. High nesting success in 1984 (Cruz et al., 1984) supports this interpretation. Many species resumed nesting in October 1983 (Valle, 1985), returning to their colonies as early as June 1983 (Merlen, 1985). The thermocline was reestablished at its normal depths in June 1983 (Halpern, 1984). and the heavy rainfall characteristic of ENSO ended in July 1983 (Merlen, 1985). Increased vegetation caused by the 1982-83 ENSO reduced available nesting space for Blue-footed Boobies on Isla Daphne Major until at least January 1986
and apparently limited Masked Booby nesting space as well (Gibbs et al., 1987). The growth of lantana (Lantana camara) during ENSO also reduced nesting space for Dark-rumped Petrels on Isla Floreana (Cruz et al., 1986). (iii) Southeastern Pacific The effects of ENSO on the Peruvian and Chilean upwelling ecosystems are well known (e.g. Glantz and Thompson, 1981; Bernal et al., 1982; Arntz, 1986; Hansen, this volume). The onset of ENSO is marked by the arrival of eastward-moving subsurface waters on the west coast of South America, depressing the thermocline. Upwelling continues, but the cool nutrient-rich water that normally serves as the water-source is overlain by warm, nutrient-poorwater (Smith and Huyer, 1983). Cold patches of water may remain along the coast, concentrating the marine fauna typical of upwelling. Successive cold-water patches of surviving prey are overwhelmed by southwardmoving internal waves (Arntz, 1986). Prey such as “anchoveta” Eneraulis rineens and mackerel Scomber IaDonicus move southward, offshore to deeper water, or die (Vogt, 1942; Zama et al., 1984; Arntz, 1986). Anchovy and the herring CluDea betincki typically experience reduced reproductive success during ENSO, but the reverse is true of the “sardina“ Sardinous (Tsukayama and Alvarez, 1981; Ware and Tsukayama, 1981; Bernal et al., 1982; Veloso and Arrizaga, 1985). Occurring at approximately 3.8-year intervals (Quinn et al., 1978, 1987), ENSO causes adult mortalities and nesting failure of the resident seabirds, especially the Guanay Cormorant Phalacrocorax boupainvillii, Peruvian Booby variepata and Peruvian Brown Pelican Felecanus occidentalis thaaus (Vogt,
399 1942; Jordan and Fuentes, 1966; Duffy, 1983). Effects occur sequentially down the coast (Vogt, 1940; Jordan and Fuentes, 1966), apparently as cold-water refuges are swamped by southward-movingwarm water (Duffy, 1983). During the most recent ENSO event, Peruvian sea-surface temperatures began to rise in early October 1982, indicating the arrival of the subsurface wave (Smith, 1983). Emigration and mortality of guano birds started as early as October 1982 in northern Peru, but not until March 1983 in southern Peru (Tovar and Cabrera, 1985). Large numbers of Guanay Cormorants and Peruvian and Bluefooted boobies occurred in Ecuador in April-May 1983, presumably having emigrated from northern Peru or Galapagos (Herdson, 1984). Both booby species were present in Panama in April 1983 (Aid et el., 1985). Mortality of these species subsequently occurred in both areas (Herdson, 1984; Ainley et al., 1988; Smith, this volume). Blue-footed Boobies were also reported from northern Chile (22-23O S ) in February 1983, along with an influx of warm-water fish (Guerra, 1983). In southern Peru, southward movement of guano birds peaked during November 1982
-
February 1983 (Fig. 1; after Hughes, 1985). Increased numbers, probably
coinciding with a northward return of pelicans and cormorants, occurred from October onward in 1983. In Chile, an unquantified but large die-off of Guanay Cormorants, Peruvian Brown Pelicans and Peruvian Diving Petrels Pelecanoides garnoti took place from December 1982 to February 1983 in the vicinity of Arica, ca. 18OS (M. Pinto, via B. Araya, pers. comm.). At 2loS, Guerra and FitzPatrick (ms.) found that Grey Gulls
Larus modestus
shifted their molt
patterns and did not breed during 1982-1984. Guano birds experienced a major influx from the north, then an 84% mortality. Local breeding of pelicans was disrupted by the influx of guano birds and of sea lions Otaria bvronia. Pelicans were able to exploit larger prey than did the other guano birds, feeding on sardines SardinoDs
moving in from the north, and did not
experience food shortages as did the Guanay Cormorant and Peruvian Booby that require smaller prey. Duffy et al. (1988) calculated a mean mortality of 0.22 along the Peruvian coast at the height of the 1982
-
-
0.25 birds/m
1983 ENSO, based on data
from Tovar and Cabrera (1985). The initial Peruvian guano bird population in March 1982, at the end of the breeding season, was six million adults and 2.9 million juveniles (Tovar and Cabrera, 1985). Toward the end of the event, in May 1983, only 330,000 birds remained in Peru (Fuentes, 1984) and numbers declined further to reach 110,000 by June (Tovar and Cabrera, 1985). By March 1984, ten months after the end of the ENSO, about 1.2 million birds were present in the area 12O
-
18OS
(Fuentes, 1984) (birds north of 12OS were not
counted), suggesting a maximum mortality of approximately 87%, one of the highest ENSO mortalities recorded (cf. Jordan and Fuentes, 1966; Duffy, 1983).
400
-4-
Peruvian Booby
+ Guanay Peruvian Pelican
A
M
J
J
A S 1982
0
N
O l J
F
M
A
M
J 1983
J
A
O
S
N
D l J
F
1984
Time
Fig 1. Numbers of birds passing north or south offshore from Mollendo, Peru, 17"S, April 1982 - February 1984 (after Hughes, 1985).
Humboldt Penguin SDheniscus humboldti numbers were reduced severely, with only 2,100 - 3,000 individuals surviving a 65% die-off of the initial population (Hays, 1986). Juveniles constituted 76% of 21 penguins found dead on beaches (Hays, 1986) and mortality was heavier (80%) in the northern part of Peru ( S o
-
12O11'S) than in the center (64%: 12O11'S
-
15O3O'S). Similarly, in
Chile, Araya (pers. comm.) found that Humboldt Penguins in the northern part of the country (Tarapaca, Antofagasta, and northern Atacama) experienced a greater population decrease than did penguins farther south. Scavengers such as Turkey Vultures Cathartes aura, and Kelp
Larus
dominicanus and Band-tailed L. belcheri gulls increased, apparently attracted to the dead birds (Arntz, 1986).
In Paracas Bay, Black Skimmers Rvnchous
m,inshore foragers in sheltered bays, appeared unaffected, whereas Chilean Flamingoes Phoenicouterus chilensis increased, presumably as a result of emigrants escaping the concurrent drought in the Andean altiplano (Arntz, 1986). (iv) Central Eastern Pacific Central America experiences drought during ENSO events (Rand and Rand, 1982; Vega, 1987), linked to similar conditions in the Caribbean (Rasmusson, 1985). Effects of ENSO events are best documented for
40 1
landbirds (e.g. Foster, 1977; Wheelwright, 1986). In 1982, the rainy season was drier and shorter than usual and drought conditions persisted into June 1983, caused by strong easterly winds that prevented the build-up of the cumulus clouds characteristic of the rainy season (Vega, 1987). A colony of 400-500 pairs of Brown Pelicans at Isla Guayabo, Gulf of Nicoya, Costa Rica deserted in March 1982 and did not nest at all in 1983 (M. McCoy, pers. comm.). Pelicans breed during the dry season (Stiles, 1984), extending from December to April, when the Gulf of Nicoya is wind-mixed as a result of strong northerly winds (Peterson, 1960) spawned by polar cold-fronts (Hasenrath, 1966; Coen, 1983). In Panama, Peruvian guano birds, including Peruvian Booby (see above), Guanay Cormorant, and Inca Tern Larosterna (Smith, this volume).
a, occurred during the
ENSO event
(v) Northeastern Pacific Although 'warm events' in the northeastern Pacific have occurred in the past and are frequently synchronous with Peruvian events (Sette and Isaacs, 1960; McLain et al., 1985; Dayton and Tegner, this volume), this is not always the case (cf. Chelton et al., 1982). The California Countercurrent strengthens during ENSO conditions (McLain and Thomas, 1983), transporting warmer water and planktonic warm-water organisms northward (cf. Hubbs, 1948; Radovich, 1961). Pacific mackerel experience increased reproductive success during ENSO events (Sinclair et al., 1985), and pelagic red crabs Pleuroncodes ulanipes are swept north by the countercurrent (McLain and Thomas, 1983). Anderson (1973), Ainley and Lewis (1974) and Ainley (1976) report on some effects of the 1972 ENSO on seabirds. In 1983, large-scale warming along the North American west coast began in January; temperatures were above normal from Mexico to Alaska; and sea-levels rose to values higher than those reported in the previous severe ENSO of 19571958 (Lynn, 1983). The countercurrent reversed inshore in February - March 1983 and off northern California by April (Lynn, 1983), suggesting a return to more typical oceanographic conditions. Warm waters (anomalies in excess of 2'C) were present in the Gulf of Alaska from April through October 1983 (Royer and Xiong, 1984). Among seabirds, changes in their distribution began in the boreal autumn of 1982, with fewer cold-water and more warm-water species in northern California waters (Ainley et al., 1988). Unusual warm-water species such as Black-vented Shearwater Puffinus ouisthomelas, Black Storm-Petrel Oceanodroma melania, Craveri's Murrelet endomvchurp SraverL, and Brown Booby Sula leucoeaster began to occur near the Farallon Islands, off northern California, in early to mid 1983; the murrelet and storm-petrelhaving disappeared from their normal ranges
402 in the Gulf of California, Mexico (Ainley et al., 1988). Adult mortality included cormorants in California, murres and guillemots in California and Oregon, and Short-tailed Shearwaters puffinus tenuirostris and Black-legged Kittiwakes
-
Larus tridactvla in Alaska and Kamchatka (Nysewander and Trapp,
1984; Graybill and Hodder, 1985; Hodder and Graybill, 1985; Hatch, 1987; Ainley et al., 1988). Many breeding sites were deserted during the 1983 ENSO. Postevent populations were 50% lower for Tufted Puffins
cirrhata on the
Farallons (Ainley et al., 1988) and 19% lower for Pigeon Guillemots CeDDhUS columba in Oregon (Hodder and Graybill, 1985). Although some populations showed no decrease after the 1983 ENSO, Ainley et al. (1988) suggest that this apparent stability could have resulted from an influx of non-breeders. Reduced nesting success occurred in a wide variety of birds. Pelagic E pelapicus and Brandt’s E pencillatus cormorants and Common Murre Uria aalge produced fewer young in Orgeon in 1983 compared to 1982, but Pigeon Guillemot and Western Gull
Larus occidentalis success was
not affected (Hodder and
Graybill, 1985; Bayer, 1986). At the Farallon Islands, six of seven species (Brandt’s and Pelagic cormorants, Western Gull, Common Murre, Pigeon Guillemot, and Cassin’s Auklet PtvchoramDhus aleuticus, but not Ashy Storm-petrel Oceanodroma homochroa), had their lowest nesting success in 1983, compared to data from 1981-1985 (Ainley et al., 1988). In Alaska, Black-legged Kittiwakes experienced nesting failure in 1983 (Ainley et al., 1988), but desertions have also been common in recent, normal years (Hatch, 1987). Common and Thick-billed
p. lomvia murres and Horned Puffins had normal or improved breeding success in 1983, suggesting that diving species were less affected by ENSO, compared to surface-feedingkittiwakes (Hatch, 1987). (vi) Southwestern Pacific
During ENSO events, New Zealand and Australia
are protected from eastward-moving cold-fronts by a ‘blocking high‘ pressure area between eastern Australia and New Zealand, resulting in drought in Australia and unusually cold water off New Zealand (van Dijk et al., 1983). Beached-bird surveys (Veitch, 1975, et seq.; Powlesland, 1983, et seq.) showed few dead birds during the 1972, 1977, and 1983 ENSO events, but counts were higher in years following such events (Fig. 2). There was a major invasion and beach-wreck of southern ocean species in New Zealand, Australia, and Tasmania in 1984 (Carter, 1984; Powlesland, 1986). The absence of fronts, blocked by the high pressure area during ENSO events, would have resulted in calmer conditions, less likely to cause mortality of seabirds. The subsequent increase in counts of dead birds may reflect especially vigorous circulation in the Southern Ocean following ENSO events.
403 New Zeolond
South Africa
-
Fig 2. Annual number of beached birds/km coast for South Africa (1977 1983) and New Zealand (1970 - 1983), after Veitch (1975; et seq.); Powlesland (1983, 1984, 1985); and Avery (1985; pers. comm). 2.2 Atlantic Ocean (i) South-west Atlantic
The northward-moving shelf-waters off Argentina
are sub-antarctic in origin, with an influx of water from the Straits of Magellan (Carreto et al., 1986). Local winds are strong from the west and southwest, especially in summer (Charpy and Charpy, 1977), resulting in coastal upwelling and productive inshore communities. Offshore, conditions are determined by the relative strengths of the warm, southward-flowing Brazil Current and the cold, northward-flowing Falkland Current; increased penetration by the Brazil Current and warmer conditions occur in summer (de Ciechomski and Sanchez, 1983).
In the South Georgia area, krill biomass in the upper water column during the austral winter of 1983 was considerably reduced compared to summer 19811982 (Heywood et al., 1985). Breeding failure of Gentoo Penguins PvPoscelis ~ a ~ and u a Black-browed Albatrosses Diomedea melanoDhris occurred during the
1983-1984 summer, as it had following the 1976-1977 ENS0 (Croxall and Prince, 1987). Breeding success of Grey-headed Albatrosses and Macaroni Penguins Eudvptes chrvsoloDhus was also minimal, even though the former had been unaffected by the earlier event (Croxall and Prince, 1987). I n Chubut Province, King Cormorants -P
atriceps did not breed at all during
404
their normal austral-summer breeding season in 1982-1983,while Rock Cormorants maeellani had a limited and late nesting season (Boersma, in Duffy et al., 1988). In 1983-1984,both species had poor seasons. Magellanic Penguins Suheniscus maeellanicw had a relatively successful season in 1982-1983,but, in subsequent years, productivity and mean body size of young were reduced, Both these indices of breeding success are significantly and positively linked to January sea-surface temperatures. On the one hand, this suggests that warmer conditions favor penguin reproduction, while the reverse may be true for cormorants. On the other hand, Scolaro and Bodano (1986) reported an 86% mortality of fledgling Magellanic Penguins caused by heavy rains in 1984, as well as during 1976. Both these were El Nifio years, suggesting that both terrestrial and marine conditions during ENSO events determine nesting success. In a limited series of counts of birds found dead during 1969-1985 along beaches at Costa Bonita, Buenos Aires, Argentina, 3ao45’S,58O5O’W,Narosky and Fiameni (1986) found greatest numbers of dead Sooty Shearwaters, Slender-billed Prions PachvDtila belcheri and Magellanic and Rockhopper penguins EudvDtes chrvsocome in the austral winter of 1984, coinciding with increased seabird mortalities in Australia. Stronger than usual southward penetration by the Brazilian Current may occur during Pacific ENSO events when west winds increase because of a northward shift of the South Atlantic atmospheric high-pressure cell (Gillooly and Walker, 1984). At present, we can only speculate that this leads to increased sea temperatures and improved availability of anchovy eneraulis anchoita, the major food of Magellanic Penguins (Scolaro and Bodano, 1986), and thus presumably to their increased nesting success. (ii) Southeastern Atlantic The southwestern coast of Africa is a productive, upwelling system analogous to that off Peru. In the far south, upwelling occurs primarily during summer, whereas off Namibia upwelling occurs throughout the year, with some seasonal variation (Shannon, 1985). The southern upwelling is controlled by the South Atlantic high-pressure cell that moves south in summer, blocking eastward-moving cold fronts and producing southeast trade winds that result in coastal upwelling. In winter, the cell moves north, allowing fronts to penetrate, leading to onshore winds and cessation of upwelling (Andrew and Hutchings, 1980). Two separate ENSO mechanisms affected the marine environment during 19821984 in southern Africa. In 1982-1983,the South Atlantic high-pressure cell remained north of its usual position during much of the summer, allowing fronts to move through and leading to reduced upwelling and warmer sea-surface temperatures (Nelson and Walker, 1984). Fishery landings yielded small catches of anchovy (Shannon et al., 1984). These conditions were local, mostly
405
affecting the southwestern Cape Province of South Africa. In 1984, the El Nifiolike event in the tropical Atlantic (Horel et al., 1986; Shannon et al., 1986) led to an intrusion of warm water, heavy rains, and subsequent reproductive failure of the anchovy Engraulis iauonicus cauensis off Namibia (Boyd et al., 1985). During 1982-1983,a mass breeding failure (75-95%) of Cape Cormorants Phalacrocorax cauensis occurred in Saldanha and Lambert's bays of the western Cape, but not farther south or east. The failure seemed to be caused by local food shortages that forced breeding adults to undertake very long foraging trips and to neglect their nests (Duffy et al., 1984). Dead birds, especially Cape Cormorants, on beaches were more abundant during the 1982 and 1977 ENSO events than in normal years (Avery, 1985). This may represent increased mortality or simply reflect diminished offshore transport of dead birds during reduced upwelling. During 1982-1983,breeding success of Cape Gannets Morus capensis was normal, but post-fledging mortality was higher than in previous years. African Penguins Suheniscus demersus deserted nests at Marcus and St. Croix islands (Duffy et al., 1984). Post-molting adults may have experienced higher than normal mortality (Avery, 1985). Unfortunately, almost no information exists on the effects of the 1984 Atlantic EN in Namibia. No ornithologist worked on Namibian seabirds during this period. In South Africa, Cape Cormorants again experienced mass reproductive failure in Saldanha Bay (approximately 10,000nests) during November 1983 (pers. observ.), but this may have been unrelated to events farther north. Survival of post-breeding adult African Penguins from Marcus Island dropped to a six-year low of 33.3% during 1984-1985,from a mean of 61.7% (La Cock et al., 1987). (iii) Caribbean Sea
Events within the Caribbean and their relation with
ENSO appear exceptionally complex. The 1983 ENSO coincided with a major drought (Norton, 1983a,b,1985) that was periodically broken as fluctuations of the jet-stream brought brief periods of heavy rain (Rasmusson and Wallace, 1983; Norton, 1983b). Marine productivity would have been affected by two factors: short-term variations in inshore productivity from local freshwater run-off, and longer-term,delayed variations in regional productivity influenced by runoff from the Orinoco and other rivers on the surrounding continents (Norton, in Duffy et al., 1988). Local effects of the 1983 ENSO were mild, with a slight die-off among young Roseate Terns Sterna douFalliC during an anomalous heat wave, and a reduction in nesting success of Sooty Terns (Norton, in Duffy et al., 1988). The 1984 Atlantic EN may also have affected the Caribbean, as egg-volumes of four
406 inshore foragers were smaller in 1984 than in 1985, while the reverse was true for three offshore foraging species; and commercial fisheries in the Virgin Islands were greatly reduced during November 1983-March 1984 (Norton, in Duffy et al., 1988). 3 DISCUSSION OF EFFECTS AND EVOLUTIONARY CONSEQUENCES OF ENSO The effects of ENSO on seabirds appear to fall into three classes of increasing biological importance: 1) extralimital dispersals and range expansions; 2 ) reproductive failure; and 3) adult mortality. Extra-limital dispersal seems characteristic of ENSO events. Warm-water seabird species invaded the California upwelling; Peruvian guano birds moved south into Chile and north to Panama; and northern migrants were unusually abundant in the Galapagos. Many of the vagrants died, although enough may have survived in past events so that species occasionally established themselves in new areas. Elegant Terns Sterna eleeans are believed to have extended their breeding range 500 km south along the California coast to San Diego following the 1957-1958event (Schaffner, 1986). The presence of Guanay Cormorants in Argentinian Patagonia may have resulted from dispersal of this species during the 1965 or earlier ENSO events (Erize, 1972). The disjunct ranges of other species such as Blue-footed Boobies could have been caused either by dispersing birds reaching and colonizing new areas during El Niiio events, or by extinctions of colonies in intervening areas. El Niiio might even be a cause of speciation under such circumstances. Unfortunately, post hoc zoogeographic explanations, invoking El Nifio, can never be satisfactorily tested, but changes in range and new colonizations should be watched for following future events. Nesting failure is a characteristic of many seabirds, especially those in tropical waters (cf. Nelson, 1978). Nesting failure is also one of the commonest effects of ENSO. As one more cause of such failures, ENSO probably exerts little selective force on seabird biology, because long-lived adults can survive to breed another year. However, if ENSO conditions persist over a number of years or if ENSOs increase in frequency
so
that reproduction is
depressed for a long period, seabird populations would decrease, as may have happened a century ago to Cassin’s Auklets in California (Ainley and Lewis, 1974). ENSO may also trigger shifts between warm and cold phases in marine ecosystems, leading to longer term changes in environmental suitability for seabirds (Shannon et al., 1984). Adult mortality may be a much more serious result of ENSO, because populations of long-lived, slowly-reproducingseabirds can require decades to recover from a single major mortality (Southwood, 1981). Unfortunately, we lack data on the extent of mortality in most seabird populations during ENSO events. Counts of unmarked seabirds at colonies are often not especially useful, as
407
unchanged populations after an ENSO event might reflect greater nesting synchrony or increased recruitment of non-breeders, even if extensive mortality has occurred. Two main factors determine the importance of ENSO in shaping the breeding and population biology of seabirds: the frequency and severity of ENSO events in relation to the generation time of the birds, and the degree of densitydependent mortality during an event. Investigating frequencies of ENSO is difficult because of the absence of long time-series of data. Even when data are available, differing intervals between ENSO events may result simply from time-series of variable lengths being recorded in different areas. Nevertheless, such comparisons serve as useful starting points for further work and may indicate the time-scales necessary for future studies. In Peru, moderate ENSO events (approximately 17% adult mortality:
Duffy,
1983) occur at approximately five-year intervals (Quinn et al., 1978) and severe ENSOs (47% adult mortality: Duffy, 1983) occur at 12.3-year intervals (Quinn et al., 1978). Off southern Africa, Shannon et al. (1986) report only two ENSO-like events since the early 1950s: 1963 and 1984, an interval of 21 years. Along the west coast of North America, major El NiAo events occurred in 1940-41 (Pearcy and Schoener, 1987) and in 1957-58 (Sette and Isaacs, 1960), but little research effort was expended on seabirds. Mortality occurred in the 1972 and 1983 events (Ainley et al., 1988), suggesting the minimum time between ENSO events with adult mortality in California is likely to be on the order of 10-15 years. Milder events may have passed unnoticed because of normal, interannual variation (cf. Hatch, 1987). In other areas, the frequency of biologically ‘significant’ENSO events appears to be similar. In Galapagos, major events occurred in 1972 and 1983. In the central Pacific, 1983 appears to have been the first ENSO studied (e.g. Schreiber and Schreiber, 1984; Ainley et al., 1988). Earlier severe events would have coincided with major events in Peru, as the ENSO mechanism involves both areas (Cane, 1983). Southwood (1981) has argued that, when a species’ average generation time (the years between first reproduction of an animal and of its young) is approximately equal in length to periods of favorable environmental conditions, the species will tend to lead a ‘boomfiust’ existence, with highly unstable populations, short generation times, high intrinsic rates of population increase, and mortality occurring irrespective of population size. In contrast, a species with a generation time shorter than typical periods of favorable environment will be long-lived,with a stable population, a low intrinsic rate of population increase, and birth rates sensitive to population densities. Conditions in Peru would appear to favor seabird species typical of the first
408
class of organisms, while conditions in the central Pacific islands, despite the events of 1983, favor species of the second. In neither case do birds have adaptations to survive ENSO events; they have adaptations to the environments that occur between events. A seabird species adapted to survive Peruvian ENSOs would exhibit reduced reproductive effort to avoid depleted energy reserves that might jeopardize its survival at the onset of anENSO. It would thus leave fewer offspring than other individuals during normal years. This might be advantageous if such ‘bethedging’ adults did indeed have a greater probability of surviving the next ENSO. If survival is mostly a matter of luck, however, the more offspring produced in good years, the greater the chance of some surviving the next ENSO. In tropical waters, adaptations to survive events that occur less than twice a century are less likely to be selected for than are adaptations to survive in the environment that occurs more than 95% of the time. Adaptations for this ‘normal’ environment include delayed maturity, small clutch size and frequent desertion (Goodman, 1974; Nelson, 1978). These serve to reduce the risk to adults caused by environmental variability. Such ’bet-hedging‘features may also be sufficient to allow birds to survive all but the strongest ENSO events (cf. Curio, 1983). Other marine environments may fall between the Peruvian and Pacific Island situations, with moderate mortality. Unfortunately, we usually lack the basic demographic data necessary to test this. 4 CONCLUSIONS AND F’UTURE WORK
The effects of ENSO on seabirds can be divided into two major themes: 1) proximate effects on breeding success and adult survival, and 2 ) ultimate effects on populations and the evolution of seabird species. Despite all the data presented above, we know little of either aspect. Seabirds frequently desert their nests during an ENSO event, leaving their eggs and young to die. Some seabirds appear to anticipate the onset of an event (Vogt, 1940; La Cock, 1986; Schreiber and Schreiber, 1986). We need to investigate the forces that trigger desertions. Are desertions physically induced, by rain, high tides, anomalously high sea or air temperatures, or are food shortages and outbreaks of parasites responsible? Do desertions precede or coincide with the onset of an ENSO event? What kills adults: lack of food, parasites, diseases, or are most apparent mortalities in reality just emigrations from study areas? The 1983 ENSO was physically one of the most severe ever recorded and seems to have affected areas usually uninfluenced by ENSO or where environmental
‘noise’hides the effects of milder events (cf. Hatch, 1987, for a discussion of ‘noise‘ and ENSO in Alaska). Nevertheless, ENSO is only one of a suite of
409
‘cycles’ that may effect seabirds and marine ecosystems (cf. Cushing, 1982; Freon, 1983). Some of the effects described here may not have been caused by ENSO alone, or even by ENSO at all. We can’t be sure that the 1983 event wasn’t a 40, 100, or even 1,000 year exception to normal patterns of environmental
variability. Untangling ENSO from other sources of variability will be difficult. We can start by examining the importance of different temporal scales of variability that affect seabirds (see Hunt and Schneider, 1987 for a brief discussion), as biologists have done for other marine organisms (e.g. Stommel, 1963; Lasker, 1978; Smith, 1978; Walsh, 1978). We need to work more closely with researchers studying ‘teleconnections’between different parts of the globe. Our lingua franca must be time-series of quantitative data. Unfortunately, most studies of seabird populations last only a year or two. Some exceptions of relevance to ENSO studies include work on the seabirds of the Farallon Islands (Ainley and Lewis, 1974; Ainley et al., 1988) and on guano harvests and management reports from Peru and southern Africa (e.g. Jordan and Fuentes, 1966; Duffy, 1983; La Cock, 1986; Duffy and Siegfried, 1987; Schneider and Duffy, 1988). If we can’t be bothered to study seabird populations at scales appropriate to their biology, we must consider whether we should study them at all. Long-term studies ideally should include assessments of interannual variability in reproductive and life-history parameters, such as clutch size, young fledged, adult survival, age of first breeding, breeding frequency, and population size. These data will be most useful if diet and environmental parameters are sampled concurrently, at appropriate scales. We must also model populations (e.g. Leigh, 1981; MacCall, 1984). Modeling may allow us to determine whether our insights into how seabird populations work are reasonable, long before we have the data to test them directly. Major ENSO events occur at approximately 12-year intervals (Quinn et al., 1978). The probability is high (p 0.8 - 0.9) of a strong event occurring by
-
2002 - 2003 (Glynn, 1988). Milder events, such as the one occurring in 1987, may also yield valuable data; however, if we are not to amass yet another collection of anecdotes in future reviews, we must have studies ‘up and running‘ in time to capture events, rather than waiting until one is underway. 5 ACKNOWLEDGEMENTS
This article is dedicated to the memories of William Vogt whose pioneering work on E l Nifio (Vogt, 1940, 1942) was among the least of his many contributions, and to Ralph Schreiber whose contributions to ENSO studies had only just begun. I thank D. G. Ainley, B. Araya, P. A. Arkin, W. E. Arntz, G. Avery, A . Berruti, P. D. Boersma, A. J. Boyd, D. Cabrera, J . Cooper, R. Crawford, R.
410 Fraga, H. L. Gibbs, C. Guerra, J. C. Haney, S . Harding, J. H. Hugel, M. L. Jimenez, M. McCoy, D. R. McLain, G. Merlen, R. L. Norton, P. G. Ryan, D. C. Schneider, R. W. Schreiber, L. V. Shannon, N. Smith, and H. Tovar S. for comments, data, discussion, or preprints. Writing of this review was supported by the Benguela Ecology Programme of the South African National Committee for Oceanography, by the University of Cape Town, and by the Graduate Program in Wildlife Biology, Escuela de Ciencias Ambientales, Universidad Nacional de Costa Rica. 6 REFERENCES Aid, C.S., Montgomery, G.G. and Mock, D.W., 1985. Range extension of the Peruvian Booby to Panama during the 1983 El Nifio. Col. Waterbirds, 8:67-68. Ainley, D.G. 1976. The occurrence of seabirds in the coastal region of California. Western Birds, 7:33-68. Ainley, D.G. and Lewis, T.J., 1974. The history of Farallon Island marine bird populations, 1854-1972. Condor, 76:432-446. Ainley, D.G., Carter, H.R., Anderson, D.W., Briggs, K.T., Coulter, M.C., Cruz, F., Cruz, J.B., Valle, C.A., Fefer, S.I., Hatch, S.A., Schreiber, E.A., Schreiber, R.W. and Smith, N.G., 1988. ENS0 effects on Pacific Ocean marine bird populations. In: H. Ouellet (Editor), Proceedings of the XIX International Ornithological Congress. National Museum of Natural History, Ottawa, Canada, pp. 1,747-1,758. Anderson, D.W. 1973. Gulf of California seabird breeding failure. Event Notification Report, Smithson. Instit., 1653. Andrew, W.R.H. and Hutchings, L., 1980. Upwelling in the southern Benguela Current. Prog. Oceanogr., 9:l-81. Arntz, W.E. 1986. The two faces of El NiRo 1982-83. Meeresforch., 31:l-46. Avery, G. 1985. Results of patrols for beached seabirds conducted in southern Africa in 1983. Cormorant, 13:3-15. Bayer, R.D. 1986. Breeding success of seabirds along the mid-Oregon coast concurrent with the 1983 El Niiio. Murrelet, 67: 23-26. Bernal. P.A., Robles, F.L. and Rojas, O., 1982. Variabilidad fisica y biologica en la region meridional del sistema de corrientes Chile-Peru. In. J.C. Castilla (Editor), Segundo Seminario Taller: Bases Biologicas para el Us0 y Manejo de Recursos Naturales Renovables: Recursos Biologicos Marinos. Santiago de Chile, pp. 75-102. Boersma, P.D. 1978. Breeding patterns of Galapagos Penguins as an indicator of oceanographic conditions. Science, 200:1481-1483. Boyd, A.J., Hewitson, J.D., Kruger, I. and le Clus, F., 1985. Temperature and salinity trends off Namibia from August 1982 to August 1984, and their relation to plankton abundance and the reproductive success of pelagic fish. Collect. Sci. Pap. Int. Comm. Southeast Atl. Fish., 12:53-58. Cane, M.A. 1983. Oceanographic events during El Nifio. Science, 222:1189-1194. Carreto, J.I., Benavides, H.R., Negri, R.M. and Glorioso, P.A., 1986. Toxic red-tide in the Argentine Sea: phytoplankton distribution and survival of the toxic dinoflagellate Gonvaulax excavata in a frontal area. J. Plankton Res., 8:15-28. Carter, M. 1984. A petrel strike. R.A.O.U. Newsletter, 61:l. Charpy, C.J. and Charpy, L.J., 1977. Biomass phytoplanctonique, production primaire et facteurs limitante la fertilite des eaux de golfe San Jose (Peninsula Valdes, Argentina). These Doct. Spec. Oceanologie, Univ. AixMarseilles 11, 185 pp. Chelton, D.B., Bernal, P.A. and McGowan, J.A., 1982. Large-scale physical and biological interactions in the California Current. J. Mar. Res., 40:10951125.
41 1
Coen, E. 1983. Climate. In. D.H. Janzen (Editor), Costa Rican Natural History. University of Chicago Press, Chicago, pp. 35-46. Croxall, J.P. and Prince, P.A., 1987. Seabirds as predators on marine resources, especially krill, at South Georgia. In. J.P. Croxall (Editor), Seabirds: Feeding Ecology and Role in Marine Ecosystems, Cambridge University Press, Cambridge, pp. 347-368. Cruz, F. and Cruz, J . , 1985. The effect of El Nido on the breeding of the Darkrumped Petrel on Cerro Pajas, Floreana. In, G. Robinson and E.M. del Pino (Editors), El Nido en las Islas Galapagos: el Evento de 1982-1983. Fundacion Darwin, Quito, pp. 259-272. Cruz, F., Cruz, J., Astudillo, J. and Sanchez, H. 1984. Protecting the Hawaiian Petrel. Not. Galap., 40:8. Cruz, F., Cruz, J. and Lawesson, J . E., 1986. Lantana camara L.: a threat to native plants and animals. Not. Galap., 43:lO-11. Curio, E. 1983. Why do young birds reproduce less well? Ibis, 125:400-404. Cushing, D.H. 1982. Climate and Fisheries. Academic Press, London, 373 pp. de Chiechomski, J.D. and Sanchez, R.P., 1983. Relationship between ichthyoplankton abundance and assorted zooplankton biomass in the shelf waters off Argentina. Biol. Oceanogr., 3:77-101. Duffy, D.C. 1983. environmental uncertainty and commercial fishing: effects on Peruvian guano birds. Biol. Conserv., 26:227-238. Duffy, D.C. and Merlen, G., 1986. Seabird densities and aggregations during the 1983 El Niiio in the Galapagos Islands. Wilson Bull., 98:588-591. Duffy, D.C. and Siegfried, W.R., 1987. Historical variations in food consumption by breeding seabirds of the Humboldt and Benguela upwelling regions. In. J.P. Croxall (Editor), Seabirds: Feeding Ecology and Role in Marine Ecosystems, Cambridge University Press, Cambridge, pp. 327-346. Duffy, D.C., Berruti, A., Randall, R.M. and Cooper, J. 1984. Effects of the 1982-83 warm water event on the breeding of South African seabirds. S . Afr. J. Sci., 80:65-69. Duffy, D.C., Arntz, W.E., Tovar S.H., Boersma, P.D. and Norton, R.L., 1988. A comparison of the effects of El Niiio and the Southern Oscillation in Peru and the Atlantic Ocean. In: H. Ouellet (Editor), Proceedings of the XIX International Ornithological Congress. National Museum of Natural History, Ottawa, Canada, pp. 1,740-1,746. Erize, F. 1972. The Guanay Cormorant Phalacrocorax boueainvillii nesting on the Atlantic coast of South America. Bull. Brit. Ornith. Club, 92:117-118. Foster, M.S. 1977. Ecological and nutritional effects of food scarcity on a tropical frugivorous bird and its fruit source. Ecology, 58:74-85. Freon, P. 1983. Production models as applied to sub-stocks depending on upwelling fluctuations. Proc. FAO-IOC Expert Consultation to Examine Changes in Abundance and Species Composition of Neritic Fish Stocks. San Jose, Costa Rica, 18-29 April 1983, pp. 1047-1064. Fuentes, H. 1984. Las aves guaneras en la zona sur del Peru, en marzo de 1984. Bol. ERFEN, 11:10-15. Gibbs, H.L., Latta, S.C. and Gibbs, J.P., 1987. Effects of the 1982-83 El Niiio event on Blue-footed and Masked Booby populations on Isla Daphne Major, Galapagos. Condor, 89:440-442. Gillooly, J.F. and Walker, N.D., 1984. Spatial and temporal behaviour of seasurface temperatures in the South Atlantic. s . Afr. J. Sci., 80:97-100. Glantz, M.H. 1981. Considerations on the societal value of an El Niiio forecast and the 1972-1973 El Niiio. In M.H. Glantz and J.D. Thompson (Editors), Resource Management and environmental Uncertainty: Lessons from Coastal Upwelling Fisheries. John Wiley and Sons, New York, pp. 449-476. Glantz, M.H. and Thompson, J.D., (Editors), 1981. Resource Management and environmental Uncertainty: Lessons from Coastal Upwelling Fisheries. John Wiley and Sons, New York, 491 pp. Glynn, P.W. 1988. El Niiio-SouthernOscillation 1982-83: nearshore population, community and ecosystem responses. Ann. Rev. Ecol. Syst., 19: 309-345. Goodman, D. 1974. Natural selection and a cost ceiling on reproductive effort. Amer. Nat., 108:247-268.
412 Graybill, M.R. and Hodder, J . , 1985. Effects of the 1982-83 El Niiio on reproduction of six species of seabirds in Oregon. In W.S. Wooster and D. L. Fluharty (Editors), El Niiio Norte. Washington Sea Grant Program, Univ. Washington, Seattle, pp. 205-210. Guerra, C. 1983. El piquero de patas azules, nebouxi Milne-Edwards, 1882; nuevos registros para Chile. Estud. Oceanol., 3:63-67. Halpern, D. 1984. Upper ocean heat content in the eastern equatorial Pacific during the 1982-83 ENS0 event. Trop. Ocean-Atmos. Newsl., 24:14-16. Harris, M.P. 1979. Population dynamics of the Flightless Cormorant NannoDterum harrisi. Ibis, 112:135-146. Hasenrath, S . 1966. On general circulation and energy budgets in the area of the Central American seas. J. Atmos. Sci., 23:694-711. Hatch, S.A. 1987. Did the 1982-1983 El Niiio Southern Oscillation affect seabirds in Alaska? Wilson Bull., 99:468-474. Hays, C. 1986. Effects of the 1982-83 El Nifio on Humboldt Penguin colonies in Peru. Biol. Conserv., 36:169-180. Herdson, D. 1984. Changes in the demersal fish stocks and other marine life in Ecuadorian coastal waters during the 1982-83 El Niiio. Trop. Ocean.-Atmos. Newsl., 28:14. Hernandez, C. and de Vries, T., 1985. Fluctuaciones en la poblacion de fragatas Freeata minor en la Bahia Darwin, Genovesa durante 1975-1983. In. G. Robinson and E.M. del Pino (Editors), El Niiio en las Islas Galapagos: el Evento de 1982-1983. Fundacion Darwin, Quito, pp. 239-244. Heywood, R.B., Everson, I. and Priddle, J . , 1985. The absence of krill from the South Georgia zone, winter 1983. Deep-sea Res., 32:369-378. Hodder, J . and Graybill, M.R., 1985. Reproduction and survival of seabirds in Oregon during the 1982-1983 El Nifio. Condor, 87:535-541. Horel, J.D., Kousky, V.E. and Kagano, M.T., 1986. Atmospheric conditions in the Atlantic sector during 1983 and 1984. Nature, 322:248-251. Hubbs, C.L. 1948. Changes in the fish fauna of western North America correlated with changes in ocean temperatures. J. Mar. Res., 7:459-482. Hughes, R.A. 1985. Notes on the effects of El Niiio on the seabirds of the Mollendo district, southwest Peru, in 1983. Ibis, 127:385-388. Hunt, Jr., G.L. and Schneider, D.C., 1987. Scale-dependentprocesses in the physical and biological environment of marine birds. In. J.P. Croxall (Editor), Seabirds: Feeding Ecology and Role in Marine Ecosystems. Cambridge University Press, Cambridge, pp. 7-41. Jordan, R. and Fuentes, H., 1966. Las poblaciones de aves guaneras y su situacion actual. Inf. Instit. Mar Peru Callao, 1O:l-31. La Cock, G.D. 1986. The Southern Oscillation, environmental anomalies, and mortality of two southern African seabirds. Climatic Change, 8:173-184. La Cock, G.D., Duffy, D.C. and Cooper, J . , 1987. Population dynamics of the African Penguin SDheniscus demersus at Marcus Island in the Benguela Upwelling Ecosystem: 1979-85. Biol. Conserv., 40:117-126. Lasker, R. 1978. Ocean variability and its biological effects--regional review--northeastPacific. Rapp. P.-V. Reun. Cons. Int. Explor. Mer, 173~168-181. Leigh Jr., E.G. 1981. The average life-time of a population in a varying environment. J . Theor. Biol., 90:213-239. Lynn, R.J. 1983. The 1982-83 warm episode in the California Current. Geophys. Res. Lett., 10:1093-1095. MacCall, A. 1984. Seabird-fishery trophic interactions in eastern Pacific boundary currents: California and Peru. In. D.N. Nettleship, Sanger, C . A. and P.F. Springer (Editors), Marine Birds: Their Feeding Ecology and Commercial Fisheries Relationships. Canadian Widlife Service, Ottawa, pp. 136-149. Mariduefia, L.M. 1977. Influencia del fenomeno de "El Nifio" en la nidacion de nebouxi excisa (piquero de patas azules). Mem. I1 Simp. Latinoamericano sobre Ocean. Biol., pp. 101-116.
413
McLain, D.R. and Thomas, D.H., 1983. Year-to-yearfluctuations of the California Countercurrent and effects on marine orsanisms. Calif. COOP. Oceanic Fish. Invest. Rep., 24:165-181. McLain, D.R., Brainard, R.E. and Norton, J.G., 1985. Anomalous warm events in eastern boundary current systems. Calif. Coop. Oceanic Fish. Invest. Rep., 26:51-64. Merlen, G . 1985. The 1982-83 El Nifio: some of its consequences for Galapagos wildlife. Not. Galap., 41:8-15. Murphy, R.C. 1925. Bird Islands of Peru. Putnarns, New York, 362 pp. Narosky, S . and Fiameni, M. A., 1986. Aves pelagicas en Costa Bonita, Buenos Aires, Argentina. El Hornero, 12:281-285. Nelson, J.B. 1978. The Sulidae: Gannets and Boobies. Oxford University Press, Oxford, 1,012 pp. Nelson, G. and Walker, N., 1984. Comparison of summer winds on the west coast of South Africa between 1979 and 1983 and the response of coastal upwelling. S. Afr. J. Sci., 80:90-93. Norton, R.L. 1983a. West Indies Region. Am. Birds, 37:342-344. Norton, R.L. 1983b. West Indies Region. Am. Birds, 37:916-917. Norton, R.L. 1985. West Indies Region. Am. Birds, 39:965-966. Nysewander, D.R. and Trapp, J.L., 1984. Widespread mortality of adult seabirds in Alaska: August September 1983. U.S.F.W.S. Wildl. Assist. Admin. Rep., Anchorage, 23 pp. Pearcy, W.G. and Schoener, A., 1987. Changes in marine biota coincident with the 1982-1983 El Niiio in the northeastern subarctic Pacific Ocean. J. Geophys. Res., 92:14417-14428. Peterson, C.L. 1960. The physical oceanography of the Gulf of Nicoya, Costa Rica, a tropical estuary. Bull. Inter-Amer.Trop. Tuna Comm.. 4:138-316. Powlesland, R.G. 1983. Seabirds found dead on New Zealand beaches in 1981. Notornis, 30:125-135. Powlesland, R.G. 1984. Seabirds found dead on N. 2. beaches in 1982 and a review of penguin recoveries since 1960. Notornis, 31:155-171. Powlesland, R.G. 1985. Seabirds found dead on New Zealand beaches in 1983 and a review of albatross recoveries since 1960. Notornis, 32:23-41. Powlesland, R.G. 1986. Seabirds found dead on New Zealand beaches in 1984 and a review of Fulmar recoveries since 1960. Notornis, 33:171-184. Quinn, W.H., Zopf, D.O., Short, K.S. and Kuo Yang, R.T.W., 1978. Historical trends and statistics of the Southern Oscillation, El Niiio, and Indonesian droughts. Fish. Bull. U. S . , 76:663-678. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Niiio occurrences over the past four and a half centuries. J. Geophys. Res., 92:14449-14461. Radovich, J. 1961. Relationships of some marine organisms of the northeast Pacific to water temperatures, particularly during 1957 through 1959. Fish Bull. Calif., 112:l-62. Rand, A.S. and Rand, W.M., 1982. Variation in rainfall on Barro Colorado Island. In. E.G. Leigh, Jr., A.S. Rand, and D.M. Winsor (Editors), The Ecology of a Tropical Forest. Smithsonian, Washington, D. C. pp. 47-59. Rasmusson, E.M. 1985. El Niiio and variations in climate. Amer. Sci., 73:168177. Rasmusson, E.M. and Wallace, J.M. 1983. Meteorological aspects of the El Niiio/Southern Oscillation. Science, 222:1195-1202. Rechten, C. 1985. The Waved Albatross in 1983--E1Nitlo leads to complete breeding failure. In. G. Robinson and E.M. del Pino (Editors), El Niiio en las Islas Galapagos: el Evento de 1982-1983. Fundacion Darwin, Quito, pp. 227-237, Royer, T.C. and Xiong, Q., 1984. A possible warming in the Gulf of Alaska due to the 1982-83 El Nifio Southern Oscillation. Trop. Ocean.-Atmos. Newsl., 24~4-5. Schaffner, F.C. 1986. Trends in Elegant Tern and Northern Anchovy populations in California. Condor, 88:347-354. I
-
414
Schneider, D. and Duffy, D.C., 1988. Historical variation in guano production from the Peruvian and Benguela upwelling ecosystems. Climatic Change, 13~309-316. Schreiber, R.W. and Schreiber, E.A., 1984. Central Pacific seabirds and the El Nifio Southern Oscillation: 1982 to 1983 perspectives. Science, 225:713-716. Schreiber, R.W. and Schreiber, E.A., 1986. Christmas Island (Pacific Ocean) seabirds and the El Niiio Southern Oscillation (ENSO): 1984 perspectives. In MEDMARAVIS and X. Monbailliu (Editors), Mediterranean Marine Avifauna. NATO AS1 Series G 12. Springer-Verlag,Berlin., pp. 397-408. Schweigger, E. 1964. El Litoral Peruano. Universidad Nacional "Federico Villarreal", Lima. Scolaro, J . A., La conservacion del pinguino de Magellanes: un problema de conflict0 e intereses que requieren de argumentos cientificos. An. Mus. Hist. Nat. Valparaiso, 17:113-119. Scolaro, J.A. and Bodano, L.A., 1986. Diet of the Magellanic Penguin Suheniscus maeellanicus during the chick-rearingperiod at Punta Clara, Argentina. Cormorant, 13:91-97. Sette, O.E. and Isaacs, J.D. (Editors), 1960. The changing Pacific Ocean in 1957 and 1958. Calif. Coop. Oceanic Fish. Invest. Rep., 7:13-217. Shannon, L.V. 1985. The Benguela ecosystem. Part I. Evolution of the Benguela, physical features and processes. Oceanogr. Mar. Biol. Ann. Rev., 23:105182. Shannon, L.V., Crawford, R.J.M. and Duffy, D.C., 1984. Pelagic fisheries and warm events: a comparative study. S . Afr. J. Sci., 80:51-60. Shannon, L.V., Boyd, A.J., Brundrit and G.B., Taunton-Clark,J . 1986. On the existence of an El Nifio type phenomenon in the Benguela system. J. Mar. Res., 44: 495-520. Sinclair, M., Tremblay, M.J. and Bernal, P., 1985. El Niiio events and variability in a Pacific Mackerel (Scomber iauonicus) survival index: support for Hjort's second hypothesis. Can J . Fish. Aquat. Sci., 42:602608.
Smith, P.E. 1978. Biological effects of ocean variability: time and space scales of biological response. Rapp. P.-V. Reun. Cons. Int. Explor. Mer, 173:117-127. Smith, R.L. 1983. Peru coastal currents during El Nifio: 1976 and 1982. Science, 221~1397-1399. Smith, R.L. and Huyer, A , , 1983. The Peru coastal upwelling region during a normal (1977) and El Nifio (1983) austral summer. Trop. Ocean.-Atmos. Newsl., 21:28-29. Southwood, T.R.E. 1981. Bionomic strategies and population parameters. In R.M. May (Editor), Theoretical Ecology: Principles and Applications. Blackwells, Oxford, pp. 30-52. Sosa, A.C. 1985. Fluctuaciones en las poblaciones de aves en Quinta Playa, Isla Isabela, Galapagos. In. G. Robinson and E.M. del Pino (Editors), El Nifio en las Islas Galapagos: el Evento de 1982-1983. Fundacion Darwin, Quito, pp. 245- 258. Stiles, F.G. 1984. Status and conservation of seabirds in Costa Rican waters. In. J.P. Croxall, P.G.H. Evans and R.W. Schreiber (Editors), Status and Conservation of the World's Seabirds. Int. Coun. Bird Preserv. Tech. Rept., 2:223-229. Stommel, H. 1963. Varieties of oceanographic experience. Science, 139:572-576. Tovar, H. and Cabrera, D., 1985. Las aves guaneras y el fenomeno "El Nido". In. W. Arntz, A. Landa and J. Tarazona (Editors), El fenomeno El Nifio y su impacto en la fauna marina. Bol. Instit. Mar Peru Callao, (volumen extraordinario): 181-186. Tsukayama K., I. and Alvarez V., M.A., 1981. Fluctuaciones en el stock de anchovetas desovantes durante las temporadas reproductivas de primavera 1964-1978. Bol. Inst. Mar Peru (Callao), (volumen extraordinario): 50-54. Valle, C.A. 1984. Breve nota sobre el impacto de El Nifio en las poblaciones de aves marinas en Las Galapagos. Bol. ERFEN, 11:18-19.
41 5
Valle, C.A. 1985. Alteracion de las poblaciones del cormoran no volador, el pinguino y otras aves marinas en Galapagos por efecto de El Niao 1982-83y su subsequente recuperacion. In. G. Robinson and E.M. del Pino (Editors), El NiAo en las Islas Galapagos: el Evento de 1982-1983. Fundacion Darwin, Quito, pp. 245-258. Valle, C.A. 1986. Status of the Galapagos Penguin and Flightless Cormorant populations in 1985. Not. Galap., 43:16-17. Valle, C.A. and Coulter, M.C., 1987. Present status of the Flightless Cormorant, Galapagos Penguin and Greater Flamingo populations in the Galapagos Islands, Ecuador, after the 1982-83 El Nido. Condor, 89:276-281. van Dijk, M., Mercer, D. and Peterson, J. 1983. Australia's drought and the southern climate. New Scientist, 7 April 1983: 30-32. Vega G., N. 1987. The effects of El Niiio in Costa Rica, 1982-83. Trop. Ocean.Atmos. Newsl., 39:l-4. Veitch, C.R. 1975. Seabirds found dead in New Zealand in 1973. Notornis, 22~231-240. Veitch, C.R. 1976. Seabirds found dead in New Zealand in 1974. Notornis, 23~168-178. Veitch, C.R. 1977. Seabirds found dead in New Zealand in 1975. Notornis, 24:4149. Veitch, C.R. 1978. Seabirds found dead in New Zealand in 1976. Notornis, 25~144-148. Veitch, C.R. 1979. Seabirds found dead in New Zealand in 1977. Notornis, 26~361-368. Veitch, C.R. 1980. Seabirds found dead in New Zealand in 1978. Notornis, 27~115-124. Veitch, C.R. 1980. Seabirds found dead in New Zealand in 1970, 1971 and 1972. Notornis, 27:369-385. Veitch, C.R. 1981. Seabirds found dead in New Zealand in 1979. Notornis, 28:4147. Veitch, C.R. 1982. Seabirds found dead in New Zealand in 1980. Notornis, 29:4147. Veloso F.C. and Arrizaga M.A., 1985. El Niao y la pesqueria de la sardina comun CluDea (Strannomera) bentincki de Talcahuano. Invest. Pesq. (Chile), 32:185-190. Vogt, W. 1940. Una depresion ecologica en la costa peruana. Bol. Cia. Admora. Guano, 16:307-329. Vogt, W. 1942. Aves guaneras. Bol. Cia. Admora. Guano, 18:l-132. Walsh, J.J. 1978. The biological consequences of interaction of the climatic, El Nifio, and event scales of variability in the eastern tropical Pacific. Rapp. P.-V. Reun. Cons. Int. Explor. Mer, 173:182-192. Ware, D.M. and Tsukayama, I., 1981. A possible recruitment model for the Peruvian anchovy. Bol. Instit. Mar. Peru Callao, (volumen extraordinario): 55-61. Wheelwright, N.T. 1986. A seven-year study of individual variation in fruit production in tropical bird-dispersed tree species in the family Lauraceae. In. A. Estrada and T.H. Fleming (Editors), Frugivores and Seed Dispersal. Dr. W. Junk, Dordrecht, pp. 19-35. Zama, A., Rueda L.T. and Cardenas G.E., 1984. Unusual arrival of Chub mackerel Scomber iauonicus at Puerto Chacabuco, south Chile (Pisces, Scombridae). Rev. Biol. Mar., Valparaiso, 20:61-76.
This Page Intentionally Left Blank
41 7
EL NIfiOO'SEFFECT ON SOUTH AMERICAN PINNIPED SPECIES D. Limberger 13, Frayne Road, Ashton Gate, Bristol BS3 1RU (United Kingdom)
ABSTRACT Limberger, D. 1989. El NiAo's effect on South American pinniped species. Marine mammals living in the tropical upwelling ecosystem of Galapagos and Peru were severely affected by the influx of tropical warm water during the 1982/83 El NiAo. The Galapagos fur seal (ArctoceDhalus ealauapoensis) population suffered considerable mortality; the 4 youngest age classes were almost completely lost, along with 30% of the adult population. All larger males, which were territorial during El Nitlo, disappeared and presumably died. The Galapagos sea lion (ZaloDhus californianus wollebaeki) population lost most of the 1982 year class and had a greatly reduced birth rate in the breeding season following El Niiio. High mortality was also reported for the South American fur seal (ArctoceDhalus australis) and the South American sea lion (Otaria bvronia) on the southern coast of Peru. Periods with El Nifio conditions cause a severe disruption of the food web, thus creating acute food shortage for both fur seals and sea lions.
1 INTRODUCTION
Detailed studies of the biological consequences of El NiAo are extremely difficult to conduct because of its unpredictable nature, especially in its timing and severity. Although El Niiio is a recurring event with an average interval of about 4 years it was not until recently that a detailed analysis of its biological impact was suggested to be attempted (Walsh, 1972). The biological importance of El Niiio became apparent indirectly, through its economic effects. In the course of the 1957/58 El Nido event, high seabird mortality threatened the collapse of the then main export industry of Peru, the agricultural fertilizer guano, harvested from islands containing huge seabird breeding colonies (see Duffy, this volume). Similarly, the Peruvian fisheries industry, based on anchovy, collapsed when the extensively fished population was further decreased during the 1972 El Niiio (Wyrtki, 1979). The meteorological and oceanographic changes characterizing El Niiio (like heavy rainfall, the decrease or disappearence of the southeast trade winds, increased sea surface temperature,high sea levels, depressed thermocline, and changes in salinity) disrupt the high productivity of the normally rich waters of the eastern equatorial Pacific (Barber and Chavez, 1983; Cane, 1983; Rasmusson and Wallace, 1983; Wyrtki, 1985). It is not surprising, therefore,
41 8
that all organisms which depend on the continual upwelling of nutrient rich waters suffer during El Nido conditions, the marine mammals restricted to such areas being no exception. Our knowledge of the impact on marine mammals, however, is very limited and is based on the coincidence of long term studies carried out on two species of eared seals (otariids), the Galapagos fur seal (Arctoceuhalus ealanapoensis) (Fig.1) by Trillmich and co-workers (Trillmich and Mohren, 1981; Trillmich, 1984; Arnold and Trillmich, 1985; Kooyman and Trillmich, 1986a; Limberger et al., 1986; Trillmich, 1986a) and on the South American fur seal (9, australis) by Majluf and co-workers (Majluf and Trillmich, 1981; Trillmich and Majluf, 1981; Trillmich et al., 1986; Majluf, 1987).
Fig. 1. Galapagos fur seals: adult male (left) and female nursing pup (right). The species of fur seals studied on Galapagos and in Peru are very closely related (Repenning et al., 1979) and also quite similar in behavior. They are shallow divers, feeding on fish and squid that come to the surface during the night on their vertical migration. In addition, occasional observations were noted on the local sea lion Zalouhus californianus wollebaeki (Fig. 2 ) on Galapagos and Otaria bvronia in Peru. All tropical species of eared seals depend on cool upwelling zones for feeding, which on Galapagos are surrounded by warmer tropical waters with much lower primary production. It is assumed that this prevents the seals from
41 9
migrating away. Only during extreme conditions like the 1982/83 El Niiio did the absolute dependence of the seal population on the local upwelling become fully apparent,
Fig. 2. Gal.apagos sea lions: young individual with pox sores (left, July 1983, courtesy G. Robinson) and adolescent (right, February 1988, courtesy J.S. Feingold) . 2 GALAPAGOS FUR SEAL
2.1 General biolopy The Galapagos fur seal, endemic to the Galapagos archipelago, is found year round on many of the islands' coasts, with the most dense breeding colonies around the northwest, close to areas of strong upwelling (Houvenaghel, 1984). Total population numbers were between 30,000 and 40,000before the
1982/83 El
Nido, estimated by a population census in 1977/78 (Trillmich, 1984). Most of the information available on the Galapagos fur seal is based on an ongoing study started in 1977 at Cab0 Hammond, Fernandina (glow, O"28'S) by Trillmich and co-workers. A large proportion of the studied breeding population has been individually tagged, and observations have been carried out at least during each breeding season and annually since 1977. Fernandina is the most westerly island of the archipelago; it is uninhabited and free of introduced animals, which might interfere with the fur seal population. Its climate is strongly
420 affected by the upwelling of the cool Cromwell current, with dry desert conditions at the island's lowest elevations (Colinveaux, 1984). The Galapagos fur seal is the smallest of all fur seal species. Adult males weigh 64 kg on average, with a body length of about 152 cm; an adult female weighs 27 kg, with a body length of 120 cm (Trillmich, 1984). They spend about 30% of adult life ashore in an extremely untypical habitat for fur seals, at equatorial latitudes and on lava cliffs where they experience air temperatures frequently up to and above 30°C (Limberger et al., 1986). From August to December, falling within the Galapagos cool season (June - December), the largest males defend breeding territories along a narrow coastal zone about 20 m wide and parallel to the waterline. Thus almost all territories have direct access to the water, at least during high tide. Territory tenure may last a maximum of 50 days, but averages only 25 days, during which the territory holder never leaves for feeding and depends solely on his stored fat reserves. A territorial male may lose up to 25% of his initial body weight during tenure
and is under constant threat of being replaced by an intruding male in better physical condition. During the period of territorial defense, females enter a male territory to give birth. The peak of pupping on Fernandina is September. The females remain with their usually single pup for about 1 week, after which they mate, usually with the territory holder. Females then resume their 2 - 4 day foraging trips to sea alternating with rest periods of about 1 day ashore during which they meet their pup for nursing. Attendance patterns may vary with the age of the pup and the lunar cycle (Trillmich, 1986a). Young fur seals are not weaned until 1-2 years old and frequently when even older. First year mortality is about 20%. Only a small proportion of all females, however, give birth in successive years, and it appears that these are the largest and oldest individuals. It is very unlikely that pups with an older, dependent sibling will survive. Studies on the diving behavior of the Galapagos fur seal have shown that they feed predominantly at night, when 95% of all dives are performed. The majority of dives recorded were shallower than 20 m, with occasional dives as deep as 100 m although these deep dives are far more energy consuming (Kooyman and Trillmich, 1986a). Their main diet consists of fish and squid, which come close to the surface during vertical migration at night. This, the general pattern in "normal years", was seriously disrupted when feeding conditions deteriorated during the 1982/83 El Niilo. The oceanographic conditions characteristic of El Niilo had arrived at Fernandina by July 1982 and worsened during the reproductive period of the fur seals, a time of high energy demand.
421 2.2 How El Nifio affected the GalaDaeos fur seal At the onset of the reproductive season, water temperature at Cab0 Hammond
is about 18"C, but in 1982 temperatures were 5°C higher. By the end of the reproductive season they were 26'C and continued increasing until March 1983, when they reached 30"C, more than 1O'C higher than in normal years (Limberger et al., 1983). However, El Nifio is primarily a subsurface anomaly and only
secondarily affects the surface (Kogelschatz et al., 1985), so it may have already caused a reduction in food availability well before the onset of the pupping season. During El Nifto, pupping rate was nearly as high as in previous years. In a subsection of the studied colony 18 pups were born in 1982, whereas the average for the 3 previous years in the same area was 24. However, stillbirths, which are extremely rare in normal years, were recorded on several occasions in 1982. The pups were born 9% lighter than in the 3 previous years (Trillmich and Limberger, 1985). Only 67% survived their first month of life, in contrast to 95% in the years 1979 - 81 (Trillmich, 1987). Within 5 months of birth all but 1 of 90 pups counted in the entire colony had died (Fig. 3). The pups weighed less than the average birthweight at death; this clearly indicated that they had starved, which was also manifested by their poor body condition. Although some mothers came ashore regularly to nurse them, their pups still starved. In general, however, the females stayed offshore for increasingly longer periods (Table 1). Median duration o f feeding trips shifted from 1.3 days for 1979/80 to 2 days by October 1982 and was even 3.5 days in late 1982. A few females stayed away for 10 days or longer before returning (Trillmich and Limberger, 1985).
[%1
mortality of pups
. .
. . OJ
, , 10
,
r
,
50
,
, , , , , , , , , 100
150
days after birth
Fig. 3. Pup mortality in 1982 as cumulative distribution. Under normal conditions as few as about 20% die during their first year of life.
422
Although it did not become apparent until early 1983, these extended feeding trips also affected the 1 and 2 year olds, since they still partially depended on their mothers' milk. Many of these immatures were very skinny and disappeared or were found dead during February and March 1983 (Limberger et al., 1983). Similarly, many of the adults were emaciated and disappeared. During February and March the number of fur seals encountered ashore was reduced to about 30% of those counted in November/December 1982, taking counts around full moon when the maximum number of fur seals come ashore (Trillmich and Mohren, 1981). In previous years, the numbers of fur seals ashore at this time remained constant throughout the year. It is very likely that many fur seals, including adults, had died although part of the reduction may be explained by prolonged feeding trips at sea. It seems improbable that the fur seals had migrated to other areas, although some may have occasionally gone ashore temporarily on other parts of the coast when extremely high surf prevented them from reaching the steep cliff face. However, there are no records of tagged seals found on any other island in the archipelago, nor did any tagged seal which had disappeared during El NiAo reappear after conditions had normalized. TABLE 1 Duration of foraging trips (days) of females with pups. Period
Range
Median
n
1979-1980 (pre-EN)
1 - 5
1.3
131
Sep.-Oct. 1982 (during EN)
1 - 6
2.0
32
Nov.-Dec. 1982 (during EN)
1
-
27
3.5
54
1983 (Post-EN)
1 - 4
0.6
39
2.3 After effects of El Nifio. durine the reuroductive season of 1983 El Nifio had ended by June/July 1983 (Chavez et al., 1984a), several weeks before the next reproductive season. By September sea surface temperature was below 20°C (Limberger, 1985) and the meteorological conditions associated with El Niiio had ended. It was only then that the changes in the studied population at Cab0 Hammond became fully apparent (Fig.4). The most drastically affected were the young, who had been dependent on their mothers. All but 1 of the age class 1 - 3 years old had died. Of the 3 year olds weaned at the onset of El NiAo only 33% were resighted in 1983. Survival was higher for older fur seals classified as adults. Ninety-six
423 females (73%) were counted as compared to an average of 132 for the 3 years preceding El Niiio. Sixty-eight percent of small males (11-17) counted in previous years were found in 1983. Almost all the large males, which usually defend territories, had disappeared. None were seen in our study area, and in the entire colony only 2 maximally sized males were seen. The main study area, which was in previous years divided up between 15 territorial males, was now patrolled by only 5 males, that were clearly
0 =J
.?
200
U T e r r . Male mother Males =Adult Females Gi 1-3 Years 0 Nevborn
2 -0
c \e
100
0
0
=
o 1979
1980
1981
1982
1983
YEAR Fig. 4. Maximal number of fur seals ashore during peak reproductive season before and after EN. All age/sex classes counted in the main study area (180 m shoreline), except for pups, counted over entire colony (530 m); large number of pups present in 1981, but no counts were made. smaller than average territorial males. These 5 tried to defend areas about 4 times as large as the usual 200 m2 territory. This resulted in a temporary dominance hierarchy between them and smaller males that attempted to mate and quite frequently succeeded. It seemed that all females were reproductively active. Pupping rate was, however, sharply reduced to 10%. Newborn pups in 1983 were underweight by about lo%, as in 1982. The duration of foraging trips by nursing females was back to under 1 day and pup mortality was less than 10% in their first month of life. The effects of El Niiio on the fur seals, as it progressed together with the changes in sea surface temperature, are summarized in Table 2. Trillmich, who returned to Cab0 Hammond in 1984, gave an account of the fur seals there (Trillmich, 1985).
The fur seals on Fernandina did very well,
probably due to increased productivity of the surrounding waters and also to reduced competition because of the high losses. The few pups born in 1983, now
1 year old, were almost as heavy as 2 year olds; the females were about 20% heavier than average and the small males had grown rapidly. A territorial system had been re-established and almost all females gave birth, whereas in previous years only about half the female population pupped. This was possible
424
because all the El NiAo survivors had fully recovered and only a few were nursing young from the previous year. TABLE 2
El Niiio's effects on fur seals and sea surface temperature. Period
Fur seal responses
Mean SST
Sep.-Dec. 1982
still births 24.0-C low birth weights extended feeding trips of females
Dec.1982-Jan. 1983
deaths of young
26.0"C
Feb.-Mar. 1983
deaths of 1 and 2 year olds
30.0"C
Sep.-Nov. 1983
absence of 1-3 year olds adults reduced by 30% few territories hardly any pups more female than male pups low birth weights feeding trip duration normal normal survival of pups
l8.5'C
3 SOUTH AMERICAN FUR SEAL 3.1 General information The South American fur seal is widely distributed along the western and eastern coasts of southern South America (Fig. 5), including the Falkland Islands (Vaz-Ferreira,1982). The world population is estimated to be about 325,000 individuals (Gentry and Kooyman, 1986). It is reported to be a non-
migratory species. The females revisit the breeding grounds regularly round the year to nurse their young. The locations of the fur seal colonies are close to areas of upwelling cool waters with temperatures between 15'C and 18°C and high primary productivity (Zuta et al., 1978). The South American fur seal is larger than the closely related Galapagos fur seal. Its sexual dimorphism is also more pronounced: males weigh 150 kg and more, females about 45 kg. In contrast, they defend smaller territories, only about 50 m2, and live in much higher densities. Their peak reproductive season is around November/December, when females give birth to a single young. Females copulate after a postnatal attendance period of about 8 days and then resume their foraging trips to sea. Foraging excursions last about 2 3 days and
-
alternate with periods ashore, during which the females join their pups and nurse them, on average for a one day period. Pup mortality is usually quite low, about 10% during the first few months of life (Majluf and Goebel, 1986).
425 The fur seals' main diet consists of fish, predominantly anchovy (Eneraulis rinpens), cephalopods and crustaceans (Majluf and Goebel, 1985). Dive records obtained under non-El Niito conditions (see Gentry and Kooyman, 1986, for the techniques employed) reveal that the South American fur seal forages mainly during the night, with most of the dives around dawn and dusk. Its mean diving depth is around 25 m, rarely exceeding 40 m (Majluf and Goebel, 1985). This general pattern was quite altered during the 1982/83 El Nino.
Fig. 5. Occurrences of South American fur seals (///) and South American sea lions (\\\) from southern Brazilflruguay (Bonner, 1981; Vaz-Ferreira, 1981) to Peru (Majluf and Trillmich, 1981). Arrow indicates approximate location of the studied colonies at Punta San Juan, Peru.
3.2 1 sea at Punt San Ju n Most of the information referred to in this section is based on studies carried out in Peru at Punta San Juan, 75"12'W, 15"22'S (Fig. 5). This is the northernmost breeding colony of the South American fur seal and is estimated to contain about 60% of the Peruvian population, about 20,000 animals, based on a population census in 1979 (Majluf and Trillmich, 1981). Punta San Juan is a peninsula at the edge of the desert, surrounded by an area of cool upwelling water rich in demersal fish stocks. Punta San Juan is a protected guano bird colony, where the fur seals frequent the steep and rocky slopes, with a rocky shelf that provides tidal pools during low tides. As environmental conditions change during the day, the seals use different parts of the beach (Trillmich and Majluf, 1981). The situation reported from Punta San Juan during January and February 1983 very much confirms the findings on Galapagos fur seals. Sea surface temperatures had risen by an average of 7°C and the thermocline had dropped from about 40 m to 100 m. During this time pups were about 2 months old and
426 still totally dependent on their mother's milk. However, the time females were offshore had increased from the norm to an average of 4.7 days, and at the same time the duration of the females' visits ashore had decreased (Trillmich et al., 1986). By late February, 2
-
3 month old pups weighed about as much as
a newborn pup. Only as few as 8% of all pups examined by visual inspection were in normal physical condition. As many as 40% died during a 2 2 day period (Limberger et al., 1983). Many immatures and adults also appeared emaciated and many more carcasses were washed up onto the beaches than in corresponding periods of normal years (Majluf, 1985; Tovar et al., 1985). The dive records obtained during El Nido show marked differences in the females' behavior from those obtained under normal conditions. Not only did the fur seals dive throughout the night, they also made deeper dives, 60% being deeper than 40 m. A series of reports demonstrate that feeding conditions changed dramatically during El Nido. At an early stage prey organisms usually making vertical migrations stayed below the thermocline to avoid the warm surface water (Santander and Zuzunaga, 1984). Although a number of species from tropical warm waters replaced the typical ichthyofauna of the upwelling waters (Velez et al., 1984), they may not have been suitable prey for the seals. Majluf (1985), however, found that the seals' diet did include some of these unusual tropical fish species. This would suggest that the seals' diet was at least temporarily supplemented while their main prey was reduced or absent. It appears that some seals tried to escape these unfavorable feeding conditions by migrating. Small rookeries in the south were reported to have dramatically increased in number, while the northern ones decreased. There are even some reports of subadult males migrating to the north where the South American fur seal is usually never encountered (Majluf, pers. comm.).
4 HOW THE SEA LIONS IN GALAPAGOS AND PUNTA SAN JUAN SURVIVED THE EL NIRO EVENT 4.1 The GalaDanos sea lion The Galapagos sea lion is widely distributed over the archipelago. It prefers gently sloping rocky shores and sandy beaches. Its breeding season is less defined than that of the Galapagos fur seal and varies from island to island, starting earliest in the west of the archipelago (Trillmich, 1986b). Young are usually weaned before becoming one year old. The females usually spend the day foraging at sea and come back to suckle their young at night. Median dive depth is about 45 m, which is almost twice as deep as Galapagos fur seal dives. Their maximum dive depth recorded was 186 m. Dives to depths greater than 80 m were rare and made up only 3% of the total (Kooyman and Trillmich, 1986b). It seems that the sea lions were affected by El Nido later than the fur seals, since in February 1983 pups still appeared to be healthy with no
427
indication of starvation at a colony on Fernandina (Limberger et al., 1983). However, during a census in late 1983 only a few young born in the previous year were encountered. Pup production in 1983 was also greatly reduced, to about 30% of that of normal years (Trillmich and Limberger, 1985). There are no data on the impact of El Niito on adult sea lion numbers. On some parts of the Galapagos more sea lions than usual seemed to have suffered from the so called sea lion pox (see Figure 2 ) , a skin disease (Robinson, 1985). 4.2 South American sea lion The South American or Southern sea lion has a very similar distribution along the southern coast of South America to that of the South American fur seal (Fig. 5), although some breeding colonies are found even further north, on the Peruvian islands Lob0 de Tierra and Lobos de Afuera ( 6 " s ) . It is also more evenly distributed than the fur seal (Majluf and Trillmich, 1981). Male sea lions weigh up to 300 kg and females about 140 kg. They too have a polygynous mating system, with the pupping season from December to February. As in the Galapagos sea lion, mating occurs shortly after giving birth. A very similar attendance pattern with alternating foraging at sea and nursing ashore has been described, although details are not known (Majluf and Trillmich, 1981). In January 1983, all foraging trips recorded were longer than 5 days and presumably longer than in normal years (Majluf, pers. corn.). At Punta San Juan, sea lion pup mortality was very high during a visit in January to early February 1983. Females appeared to be in poor condition and many more dead animals than usual were seen on the beaches (Limberger et al., 1983; Majluf, pers. corn.). Population censuses were conducted during the period of June 1982 through July 1983 at Punta San Juan, covering the critical period of El Nifio. Up to October 1982, the numbers of sea lions counted were more or less constant around 4,500 and then increased to about 6,500 individuals during the reproductive season. This was probably due to congregations of animals on the breeding grounds and to inclusion of newborn pups. By July 1983, the number was reduced to 4 , 2 0 0 . Separation into sex and age groups suggests that this decrease was mainly due to the disappearance of nearly half the pups and juveniles, decreasing from 1,300 to 800 individuals. The numbers of adult males and females remained fairly constant throughout El Niito (Tovar et al., 1985). Approximately 920 carcasses, 520 of which were juveniles, were found between January 1 and February 28, with the highest mortality reported for all categories in the first 2 weeks of February. Juveniles, however, were affected most severely in late January (Tovar et al., 1985).
5 SUMMARY AND CONCLUSIONS The El Nido event of 1982/83 was probably the most intense for a century. Its effects were geographically far ranging, severely disrupting nutrient regeneration in surface waters along the western shores of the South American continent and near islands such as the Galapagos. El Nido 1982/83 was also accompanied by very widespread oceanographic effects, eventually causing dramatic changes in the depth of the thermocline both near the shore and across the open ocean (see Hansen, this volume). These disruptive changes in the physical environment had profound effects on the entire food web (Chavez et al., 1984b; Barber and Kogelschatz, this volume). Primary production, which is very high in nutrient-richupwelling waters, decreased in intensity to become nearly equivalent to that found in the nutrient poor waters of most tropical with this a classical cascade effect occurred throughout the water masses food chain.
-
In the past, the occurrence of an El Nido event was usually detected by decreased fish catches and the exodus or death of sea birds from breeding colonies. As would be expected with massive changes in ocean surface waters, the normal species composition changed dramatically (Grove, 1984; Herdson, 1984). Numerous organisms from all trophic levels suffered. Primary production was poor (Feldman et al., 1984; Barber and Kogelschatz, this volume), zooplankton density decreased and so did the number of squid and planktonfeeding fish (Herdson, 1984). These fish species not only became scarce but were in poor condition (Dioses, 1985). Finally, at the top of the food chain, sea birds (Schreiber and Schreiber, 1984; Rechten, 1985; Tovar and Cabrera, 1985; also see studies by Duffy and by Smith, this volume) and marine mammals experienced prolonged periods of extreme scarcity of prey. The most dire consequences of El Nido on the marine mammals occurred during their breeding season. Breeding seasons are temporally phased to meet the most favourable environmental conditions for the successful rearing of the young. Ideally, breeding should occur during times of prey surplus since energy requirements are high. Territorial defence by males is costly; accordingly, males that had exhausted themselves during territory tenure were not able to survive further food shortage, whereas the category of smaller, nonterritorial males were better able to cope with El Nido conditions. The nursing period adds to the female's energy requirements, as she needs food not only for herself but also to convert into considerable quantities of milk for the growing young. There is also a demand for obtaining this energy in a temporally efficient manner, within a restrictive temporal framework. Only a finite period of time can be spent foraging, because the female must return to the pup within narrow time limits. The importance of this was made quite obvious by observations of the breeding colony at Fernandina Island during El Nido. Over a period of weeks
429 the time spent by the female for each foraging trip continually increased, with briefer and briefer periods spent ashore suckling the young. Judging by the reactions of the young the female usually left well before the pup was sated. It is probable that even with an extended period of time ashore the female was not able to supply the required volume of milk. A clear conflict developed with the female's attempts to forage at sea, not only to replenish her own reserves but also to acquire enough food to produce sufficient milk for the pup. While the mother was at sea, pups on shore continually searched and called for their mothers. In desperation, some attempted to secure meals by suckling on another female. Almost without exception the females reacted aggressively to such behavior. The condition of the pup would further deteriorate if it obtained a physical injury from such an encounter. The biological consequences of El Nifio on the fur seal and sea lion populations clearly reveal that the latter species suffered less than the former. Fur seals are shallow divers compared to the sea lions. Only the sea lions dived near or below the thermocline where prey items were concentrated and the typical diet species found. This, together with a more varied diet of the sea lions ensured that they suffered less mortality than the fur seals. The two sections of the populations that suffered highest mortality were undoubtedly the newborn and the territorial males. The newborn died because their mothers' emaciated condition prevented production of sufficient milk in the restricted time period. The juveniles suffered increased mortality probably because they were less skilled foragers than adults. The territorial males probably died because defending the territory leads to a poor nutritional state by the end of the breeding season even in the best of years. These males, like the juveniles, probably died before the next breeding season because they were unable to recover nutritionally and either starved, were killed by predators or succumbed to disease from which they would ordinarily recover. Whatever the case they certainly were not seen in the following year. Events such as El Nifio represent the most severe selection pressure. Not only do entire age classes perish but a considerable number of adults as well. The population structure at the end of El Nifio varies markedly from that at the beginning. Such dramatic intensity of selection pressure does have parallels in other environments. Avian aerial plankton feeders such as swifts and martins can suffer nearly complete age class losses when there are prolonged periods of precipitation during the nestling rearing periods. Other animals suffer in periods of drought when annual rains are expected. And yet, sufficient individuals survive such unpredictable yet recurring dramatic changes, like the fur seals and sea lions occasionally exposed to El Nifio events. Although such events dramatically alter the population structure temporarily, the fur seal and sea lion populations may recover rapidly and regain their previous sizes,
430 provided that similarly severe El Nido events do not occur too frequently. After all, these environmental instabilities are part of the evolutionary framework that shaped the contemporary fur seal and sea lion populations along the South American coast. 6 REFERENCES Arnold, W. and Trillmich, F., 1985. Time budget in Galapagos fur seal pups: the influence of the mother's presence and absence on pup activity and play. Behaviour, 92: 302-321. Barber, R.T. and Chavez, F.P., 1983. Biological consequences of El Niiio. Science, 222: 1203-1210. Bonner, W.N., 1981. Southern fur seals - ArctoceDhalus (Geoffroy Saint-Hilaire and Cuvier, 1826). In: S.H. Ridgway and R . J . Harrison (Editors), Handbook of marine mammals. Academic Press, London, pp. 161-208. Cane, M.A., 1983. Oceanographic events during El Niiio. Science, 222: 1189-1195. Chavez, F.P., Barber, R.T. and Soldi, H., 1984a. Propagated temperature changes during onset and recovery of the 1982-83 El Niiio. Nature, 309: 47-49. Chavez, F.P., Barber, R.T., Kogelschatz, J.E. and Thayer, V.G., 1984b. El Niiio and primary productivity: potential effects on atmospheric carbon dioxide and fish production. Trop. Ocean-Atmos. Newsl., 28: 1-2. Colinvaux, P.A., 1984. The Galapagos climate: present and past. In: R. Perry (Editor), Key Environments - Galapagos. Pergamon Press, Oxford, pp. 55-69. Dioses, T., 1985. Influencia del fen6meno El Niiio 1982-83 en el peso total individual de 10s pesces pellgicos: Sardina, Jurel y Caballa. In: W. Arntz, A. Landa and J. Tarazona (Editors), El Niiio - su impact0 en la fauna marina. Bol. Inst. Mar Peni, Spec. Vol., Callao, Peru. pp. 129-134. Feldman, G., Clark, D. and Halpern, H., 1984. Satellite color observations of the phytoplankton distribution in the eastern equatorial Pacific during the 1982-83 El Niiio. Science, 226: 1069-1071. Gentry, R.L. and Kooyman, G.L., 1986. Methods of dive analysis. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 28-40. Grove, J . S . , 1984. Influence of the 1982-83 El Nifio on the ichthyofauna of the Galapagos archipelago. Trop. Ocean-Atmos. Newsl., 28: 18-19. Herdson, D., 1984. Changes in the demersal fish stocks and other marine life in Ecuadorian coastal waters during the 1982-83 El Niiio. Trop. Ocean-Atmos. Newsl., 28: 14-16. Houvenaghel, G. T., 1984. Oceanographic setting of the Galapagos Islands. In: R. Perry (Editor), Key Environments - Galapagos. Pergamon Press, Oxford, pp. 43-54. Kogelschatz, J . , Solorzano, L., Barber, R.T. and Mendoza, P., 1985. Oceanographic conditions in the Galapagos Islands during the 1982/83 El Niiio. In: G. Robinson and E.M. del Pino (Editors), El Nifio in the Galapagos Islands: the 1982-1983 event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 91-123. Kooyman, G.L. and Trillmich, F., 1286a. Diving behavior of the Galapagos fur seals. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 186-195. Kooyman, G.L. and Trillmich, F., 1986b. Diving behavior of the Galapagos sea lions. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 209-219. Limberger, D., 1985. El Nido on Fernandina. In: G. Robinson and E.M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-83 event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 211-225. Limberger, D., Trillmich, F., Kooyman, G.L. and Majluf, P., 1983. Reproductive failure of fur seals in Galapagos and Peru in 1982-83. Trop. Ocean-Atmo's. Newsl.. 21: 16-17.
43 1 Limberger, D., Trillmich. F., Biebach, H. and Stevenson, R.D., 1986. Temperature regulation and microhabitat choice by free ranging Galapagos fur seal pups (Arctoceuhalus ealaDaeoensis). Oecologia, 69: 53-59. Majluf, P., 1985. Comportamiento del Lob0 fino de Sudamdrica (ArctoceDhalus gustralis) en Punta San Juan, Peni, durante El Niiio 1982-1983. In: W. Arntz, A. Landa and J . Tarazona, (Editors), El Niiio su impacto en la fauna marina. Bol. Inst. Mar Peni, Spec. Vol., Callao, Peru, pp. 187-193. Majluf, P., 1987. Reproductive ecology of female South American fur seals in Peru. Ph. D. Thesis, Cambridge. Majluf, P. and Goebel, M., 1986. Fur seal foraging behavior during El Niiio. Talk at Chapman Conference on El Niiio, Guayaquil, Ecuador (unpublished). Majluf, P. and Trillmich, F., 1981. Distribution and abundance of sea lions (Otaria bvronia) and fur seals (&ctoceuhalus australis) in Peru. 2. Saeugetierkd., 46: 384-393. Rasmusson, E.M. and Wallace, J.M., 1983. Meteorological aspects of the El Niiio/Southern Oscillation. Science, 222: 1195-1202. Rechten, C., 1985. The waved albatross in 1983 - El Niiio leads t o complete breeding failure. In: G. Robinson and E.M. del Pino (Editors), El Niiio in the Galapagos Islands: the 1982-1983 event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 227-237. Repenning, C.A., Peterson, R.S. and Hubbs, C.I., 1979. Contributions to the systematics of'the southern fur seals, with particular references to the Juan Fernandez and Guadelupe species. In: W.H. Burt (Editor), Antarctic Research Series, Vol. 18. Antarctic Pinnipedia. Am. Geophys. Union, Washington, D.C., pp. 1-34. Robinson, G.. 1985. The influence of the 1982-83 El Niiio on Galapagos marine life. In: G. Robinson and E.M. del Pino (Editors), El Nifio in the Galapagos Islands: the 1982-1983 event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 153-190. Santander, H. and Zuzunaga, J . , 1984. Impact of the 1982-83 El Niiio on the pelagic resources off Peru. Trop. Ocean-Atmos. Newsl., 28: 9-10. Schreiber, R.W. and Schreiber, E.A., 1984. Pacific seabirds and the El Niiio Southern Oscillation. Science, 222: 1195-1202. Tovar, H. and Cabrera, D., 1985. Las aves guaneras y el fen6meno El Niiio. In: W. Arntz, A. Landa and J. Tarazona, (Editors), El Niilo - su impacto en la fauna marina. Bol. Inst. Mar Peni, Spec. Vol., Callao, Peni, pp. 181-186. Tovar, H., Cabrera, D. and del Pino, M.F., 1985. Impact0 del fen6meno El Nitio en la poblacion de lobos marinos en Punta San Juan. In: W. Arntz, A. Landa and J. Tarazona, (Editors), El Niiio su impacto en la fauna marina. Bol. Inst. Mar Peru, Spec. Vol., Callao, Peru, pp. 196-200. Trillmich, F., 1984. The natural history of the Galapagos fur seal (ArctoceDhalus ralauaeoensis). In: R. Perry (Editor), Key Environments Galapagos. Pergamon Press, Oxford, pp. 215-223. Trillmich, F., 1985. Effects of the 1982/83 El Niiio on Galapagos fur seals and sea lions. Noticias de Galapagos, 42: 215-223. Trillmich, F., 1986a. Attendance behavior of Galapagos fur seal females. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 168-185. Trillmich, F., 1986b. Attendance behavior of Galapagos sea lion females. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 196-208. Trillmich, F., 1987. The Galapagos fur seal Arctoceuhalus Palauaroensis. In: J.P. Croxall and R.L. Gentry (Editors), Status, biology, and ecology of fur seals. NOAA Tech. Rep. NMFS 51: 23-27. Trillmich, F. and Limberger, D., 1985. Drastic effects of El Niiio on Galapagos pinnipeds. Oecologia, 67: 19-22. Trillmich, F. and Majluf, P., 1981. First observations on colony structure, behavior and vocal repertoire of the South American fur seal (Arctoceuhalus australis) in Peru. 2. Saeugetierkd.,46: 310-322. Trillmich, F. and Mohren, W., 1981. Effects of the lunar cycle on the Galapagos fur seal (ArctoceDhalus WalaDaPoensis). Oecologia, 48: 85-92.
-
-
432 Trillmich, F., Kooyman, G.L., Majluf, P. and Sanchez-Griiian,M., 1986. Attendance and diving behavior of South American fur seals during El NiAo in 1983. In: R.L. Gentry and G.L. Kooyman (Editors), Fur seals, maternal strategies on land and at sea. Princeton University Press, pp. 153-167. Vaz-Ferreira, R., 1981. South American sea lion - gtaria flavescens (Shaw, 1800). In: S.H. Ridgway and R.J. Harrison (Editors), Handbook of marine mammals. Academic Press, London, pp. 39-65. Vaz-Ferreira,R., 1982. 4rctoceDhalus australis Zimmermann, South American fur seal. In: Mammals in the sea, Vol. 4: Small Cetaceans, Seals, Sirenians, and Otters. FA0 Fish. Ser. 5 , pp. 497-508. Velez, J . , Zaballos, J . and Mendez, M., 1984. Effects of the 1982-83 El NiAo on fishes and crustaceans off Peru. Trop. Ocean-Atmos. Newsl., 28: 10-12. Walsh, J . J . , 1972. Implications of a systems approach to oceanography. Science, 176: 969-975. Wyrtki, K., 1979. El Nitlo. La Recherche, 10: 1210-1220. Wyrtki, K., 1985. Pacific-wide sea level fluctuations during the 1982-83 El NiAo. In: G . Robinson and E.M. del Pino (Editors), El NiAo in the Galapagos Islands: the 1982-1983 event. Charles Darwin Foundation for the Galapagos Islands, Quito, Ecuador, pp. 29-48. Zuta, S . , Rivera, T. and Bustamante, A., 1978. Hydrologic aspects of the main upwelling areas off Peru. In: R. Boje and M. Tomczak (Editors), Upwelling Ecosystems. Springer Verlag, Berlin, Heidelberg, New York, pp. 235-57.
433
BOITQMS BENEATH TROUBLED WATERS: BENTHIC IMPACTS OF THE 1982-1984 EL NINO IN THE TEMPERATE ZONE PAUL K. DAYTON and MIA J. TEGNER Scripps Institution of Oceanography, A-001, La Jolla, California 92093 (USA) ABSTRACT Dayton, P.K. and Tegner, M.J., 1989. Bottoms beneath troubled waters: benthic impacts of the 1982-1984 El Niiio in the temperate zone. The massive ENSO that profoundly affected the pelagic community of the northeast Pacific from 1982-84 offered an opportunity to evaluate and scale mechanisms coupling benthic communities to the water column. The California Current system off Central and Southern California had massive changes associated with a deepened thermocline and warmed mixed layer; nutrients, chlorophyll a, and macrozooplankton biomass were extremely low, and transport patterns were altered. These responses were felt throughout the pelagic food chain from the water column to seabirds and pinnipeds. The most pronounced benthic responses were observed in kelp forests. Disturbances of these productive communities were the largest ever recorded and appeared to result from both ENSO-associated storms and low nutrients. In most cases, the main impact was on adults of the giant kelp ' pyrifera that were either tom out by the storms or starved by the low nutrient conditions. However, at one site the storms killed sea urchins and released the kelps from intense cropping by these grazers. Following disturbance, most kelp forests quickly recovered because of unusually good light penetration in the clear water apparently ' ' beds that lost their associated with reduced nutrients. Exceptions were the Macrocysris angusbfolla biotic substrata in the storms. Several kelp associated animals were affected, usually via recruitment anomalies, but they also generally recovered quickly. The recruitment of some motile epibenthic species such as crustacea and some fishes apparently was affected by the ENSO and this may result in modest secondary effects in the future. Few if any ENSO effects were observed in the intertidal zone or among infaunal species. The benthic system did not respond as strongly as the pelagic system with the exception of ' grows through the water column and was thus exposed to the full -st?GyE%w nutrient conditions; it represents a special case for benthic species. Most of the benthic responses that were observed resulted from recruitment anomalies associated with water column processes such as advection and possibly larval starvation. The fundamental differences between benthic and pelagic systems such as relative mobility, longevity, and population turnover times probably contribute to the contrasting results of the ENSO event. 1 INTRODUCTION An evolutionary justification of community ecology can be expressed with the following statement: Populations evolve in the milieu of an integrated community; our challenge is to understand the integrating factors. Of ultimate importance is the identification of the relative importance of various specific environmental factors integrating the community and influencing selection of its populations. Much has been written of the futile effort to separate biological mechanisms such as competition and predation or biotic "versus" (rarely "and") abiotic factors. And only recently has there been much focus on the proper scale of investigation because practical logistical problems force most ecologists to work in local areas over relatively brief temporal periods. Alternative options of working at large scales are rarely available and questions about the
434 most appropriate scale for particular projects are almost never asked. Benthic marine systems are among the most difficult in this regard because they are intimately tied to the pelagic bath overhead. The pelagic system influences availability of nutrients and light and often has a strong influence on the propagules entering a benthic community. Furthermore, the pelagic realm is an important source of risk from predation and sedimentation as well as several types of physiological stress (e.g., extremes of temperature, salinity, water motion, etc.). The pelagic system is an essential component of the "integrated milieu" yet its large scale importance to the benthos is difficult to study. How tightly are benthic marine populations coupled to the water column above? What factors are particularly important; that is, what are the ties that bind? What is the relative importance of nutrients, propagules, and physical stress from waves or sediment? Do populations in different communities have parallel patterns of responses to water column perturbations or do they respond individualistically? Clearly this line of thought does not lead to clean replicated experiments, nor does it lead to many conclusions that can be considered universally true. How can hypotheses of this scale be tested? A water column perturbation as enormous as the 1982-84 El Niiio-Southern Oscillation (ENSO) event should result in marked and general benthic changes if the water column and benthos were tightly coupled. Here we review the literature of benthic effects with an emphasis on Southern California kelp communities that we know were impacted by the 1982-84 ENSO event (Tegner and Dayton, 1987). Unfortunately, the benthic communities in the northeastern Pacific (Fig. 1) are characterized by very different levels of pre-existing understanding and, furthermore, there were very different levels of research on them during and after the ENSO. In many areas the ENSO event had run its course by the end of 1983, but in California there was a year lag and very strong El Niiio conditions persisted through 1984,
._
1.1 Northern hemisphere mDerate El NInos: asummarvof phvsical f a c m Only very recently have major interannual variations in the California Current k e n tied to ENSO events. Large scale, low frequency shifts in the California Current result in interannual variability in physical and biological parameters (Bernal, 1981). Chelton et al. (1982) demonstrated that the strength of the California Current is closely correlated with El Niiio events of 1957-1958, 1964,1969, and 1972 plus some minor events at other times. The most important changes in the California Current during these events are higher water temperatures and increases in sea level that correspond to anomalous levels of poleward flow. Negative (or anti) El Niiio conditions or, as suggested by Kerr (1988), "La Niiia" conditions are characterized by anomalously low temperatures, low sea levels, and strong equatorward flow of the California Current (Bernal, 1981; Chelton et al., 1982). While the correlation between tropical ENSO events and changes in the California Current is very strong, the relationship occasionally breaks down (Chelton et al., 1982; Mysak, 1986). Furthermore, temperate El Niiio events vary qualitatively in such things as the sign of the salinity anomaly (McCowan, 1985) and the occurrence of large wave events generated by mid-latitude storms (Namias and Cayan, 1984; Seymour et al., 1984). Energy transfer from the tropics to midlatitudes during ENSO events occurs by both oceanic and atmospheric processes (Chelton et al.,
435 130"
140"
120"
50"
50'
40°
40'
Son Benito Is. Cabo Thurloe
30"
140'
30"
130"
120"
Fig. 1. Map of the place names mentioned in the text. 1982; Norton et al., 1985); the differences between events may reflect shifts in the relative importance of the two pathways. While the differences between events are not well understood, there may be biological implications; Ainley and Boekelheide (in press) report that different feeding guilds of sea buds at the Farallon Islands responded to planktonic conditions during each of the last three El Niiio events. The 1982-1984 event was the largest temperate El Niiio ever measured. Anomalously high sea level, corresponding to strong poleward flow along the coast, became apparent in Southern California during late spring of 1982 and was about three standard deviations above normal by fall. This condition persisted through most of 1983, finally returning to normal by December (Cayan and Flick, 1985). The 1982-1983 winter storm season was the most severe until that time (Namias
436
and Cayan, 1984; Seymour et al., 1984), and the extratropical atmospheric circulation was clearly linked to the warming of the east equatorial Pacific Ocean surface (Quiroz, 1983). The unusually deep Aleutian Low and the intensification of the westerlies resulted in an extraordinary number of severe storms with large waves making landfall along the southwest coast of the United States (Seymour et al., 1984). Mesoscale changes in sea surface temperature in the Southern Bight were first evident in December of 1982 and the temperature was about 20C above normal for the first three months of 1983 (Fiedler, 1984; Tegner and Dayton, 1987). Surface manifestations were not apparent during April and May along the coast; upwelling apparently increased to near normal levels. But during that same period offshore, sea surface temperatures were up to three standard deviations above normal (Simpson, 1983). By July of 1983, sea surface temperatures were some 4OC above normal in coastal areas (Fiedler, 1984). Temperature, salinity, and dissolved oxygen measurements of coastal waters were consistent with enhanced onshore transport of subarctic water induced by large-scale atmospheric forcing. This onshore transport leads to increased sea level and large-scale depression of the thermocline; geostrophic readjustment to these changes causes enhanced poleward flow (Simpson, 1983,1984). In Southern California, the depth of the nutrient-depleted waters appeared to be defined by the 14OC isotherm (Fiedler et al., 1986); its enormous depression rendered coastal upwelling ineffective and eliminated internal waves, a critical source of nutrients for kelps during the summer (Zimmerman and Robertson, 1985), from kelp forest depths. While the tropical ENSO event waned by the fall of 1983, the anomalous oceanographic conditions in much of the California Current persisted through 1984 (Simpson, unpub. ms). The large-scale anomalies of subsurface positive temperature and negative salinity anomalies (Simpson, 1984) were very stable. Without any major storms to overturn the water column, these conditions persisted through the mild winter of 1983-1984 and became even warmer in 1984 with normal seasonal heating. As a result, the temperature anomaly in 1984 was larger than 1983 (Tegner and Dayton, 1987). Slow erosion and storm mixing in the fall of 1984 led to near normal conditions by the end of the year. The 1940-1941 and 1957-1959 El Nifio events also persisted longer off California than in the tropics (Simpson, unpub. ms). It is important to note what appears to be a gradient of temperate El Niiio anomalies that decrease with latitude; some of the larger tropical events in the last 100 years were not observed north of Central California (Cannon et al., 1985). Only the events of 1940-1941, 1957-1958, and 1982-1983 produced major changes in sea level off the Pacific Northwest. Cannon et al. (1985) report that the temperature anomalies were smaller and the offshore extent of the anomalies in the Pacific Northwest was only about half that observed off California for these three events.
m0
2 BIOLOGICAL EFFECTS OF THE 1982-1984 EL ON TEMPERATE PELAGIC ECOSYSTEMS 2.1 Northeastern Pacific The papers in Wooster and Fluharty (1985) represent the most complete early summary of the 1982-1983 ENS0 event in the eastern subarctic Pacific Ocean. This event was far-reaching; the
437
ocean and the atmosphere responded at least as far north as the Bering Sea (Niebauer, 1985). Maximum sea surface temperature anomalies were about 2OC off southeastern Alaska (Karinen et al., 1985), British Columbia (although there was a subsurface positive anomaly of 5OC at middepths over the shelf, Tabata, 1985), and Oregon (Huyer and Smith, 1985). Germann et al. (1987) reported decreased availability of nitrate, an important kelp nutrient, in Barkley Sound, British Columbia in 1983 relative to 1982. Zooplankton abundance in 1983 off Oregon was about 30% of a non-El N 5 o year and there was a prolonged persistence and recurring dominance of southern species (Miller et al., 1985). Apparently as a result, there was greatly increased adult mortality and decreased average size for Oregon's coho and chinook salmon (Johnson, 1988). The 1983 upwelling index was the lowest since measurements began in 1946, salmon growth was slow, most of the coho landed in the summer had empty stomachs, and average fecundity per female declined. Coho jack returns in the fall of 1983 indicated that smolts entering the Ocean in the spring of 1983 survived poorly. Interestingly, chinook stocks that were reared in Ocean areas south of Vancouver Island showed poor survival, while Oregon chinooks that migrated to more northerly Ocean areas were affected moderately, if at all (Johnson, 1988). However, salmonid problems were widespread; 1983 was a disastrous year for salmon fisheries off Oregon, California, and Washington. Such effects on the salmon were probably general as Pearcy et al. (1985) document major shifts in the rank order of abundance of nekton caught in purse seines off Oregon during 1983 and 1984, as well as the appearance of rare and unusual species from southern waters. Schoener and Fluharty (1985) summarized biological anomalies observed off the state of Washington associated with the 1982-83 ENSO period. They considered four types of changes: 1) range extensions, 2) range anomalies, 3) habitat anomalies (deeper or shallower water), and 4) abundance changes. Changes were reported for 4 invertebrates, 14 fishes, the leatherback turtle and 3 species of birds. Fulton and Le Brassuer (1985) present a table of unusual sighting of marine species off the coast of British Columbia during the same period. Their list expands the Schoener and Fluharty list by 6 species of fishes, 13 plankton species (not covered by Schoener and Fluharty), and 6 species of birds. They discuss a northward shift of the subarctic zoogeographic boundary that may account for the biotic changes. Finally, 18 other species were observed to have unusual northward range extensions off the coast of Alaska (Karinen et al., 1985). The most spectacular range extension was that of a triggerfish, which was found in Alaska, some 2,800 km north of its prior northern record (Pearcy and Schoener, 1987). Schoener and Fluharty (1985) compare biotic anomalies from the recent El Niiio with those from two very strong ENSO events in the past, 1940-1941 and 1957-1959. These data suggest that, at least for the coast of Washington, an increase in the commercial landings of squid and the range anomaly of sand crabs are the only examples of species affected by all three of these massive ENSO events. In all other cases, their data suggest that a species may have been affected by one or two but never all three events, perhaps an indication of differences between ENSOs. However, some of these observations were based on single specimens and it is unlikely that an El Niiio during the war years drew as much attention as the recently highly publicized event; i.e., lack of reports from previous El Niiio events may simply reflect lack of search rather than absence. In
w,
438 summary, more than 10 species of fish normally occumng off California were observed off Washington. For plankton and other drifting organisms, transport by altered current patterns presumably produced the observed redistribution of southern species into more northern latitudes, whereas strong swimmers probably migrated in pursuit of favored physical conditions or preferred food (Schoener and Fluharty, 1985). In an exceptionally thorough and scholarly paper, Meams (1988) evaluated over 1,OOO records of unusual biological occurrences along the west coast of North America. Meams systematically evaluated published records starting in 1915, but unfortunately was not able to incorporate data from the 1982-84 ENSO. Not only does this report offer a remarkable compendium of unusual biological records, it includes an evaluation of oceanographic data with a special focus on El Niiio and anti-El Niiio events. The main conclusion is that unusual occurrences of marine organisms are very common -- only 13 years had low numbers of unusual occurrences. But when considered collectively, unusual occurrences formed signals that often correlated with either cool, oceanic, or warm water episodes. Perhaps the most important message of Meams'review is that there are many unusual records unrelated to El Niiio or anti-El Nifio events and they may herald other, as yet undefined, physical processes along this coast. 2.2 California The 1982-1984 ENSO event produced a very strong biological signal in California, especially in the Southern California Bight, which extends from Point Conception to the Mexican border and includes the Channel Islands. McGowan (1985), using long term California Cooperative Oceanic Fisheries Investigations' (CalCOFI) data, documented massive changes in the California Current System including a large intrusion of anomalously warm, low salinity water into the Southern California Bight. There was a pronounced deepening of the thermocline and a warming of the mixed layer. Nutrient concentrations were extremely low in the euphotic zone, algal biomass was redistributed to deeper water, and macrozooplankton biomass reached record lows for the 30 year history of measurements (Fig. 2, McGowan, 1985). Chlorophyll a concentration dropped markedly in mid summer of 1983 in the waters around the Scripps Pier in La Jolla and the fall phytoplankton assemblage consisted of oceanic warm water species (Reid et al., 1985). Putt and Prezelin (1985) found that over 80% of the phytoplankton biomass in the Santa Barbara Channel in 1983 was picoplankton (4 pm) dominated by chroococcalean cyanobacteria typical of oligotrophic warm waters. Commercial fisheries showed drastic declines in 1983: anchovy 90%, squid 89%, salmon 86%, ocean shrimp 74%, Dungeness crab 70%,rockfish 35%, and jack mackerel 30% (Klingbeil, 1984). Some of these declines can be atnibuted to northward displacement of mobile stocks, e.g., the squid fishery increased off Washington in 1983 (Schoener and Fluharty, 1985), and jack mackerel became relatively more abundant off Oregon (Pearcy et al., 1985). To some degree these losses were offset by significant increases in albacore, yellowfin tuna, skipjack, and swordfish. The sportfishing industry prospered because of the migration of more southerly "big game" into California waters (Klingbeil, 1984) including the above three species, marlin, and dorado as well as increased numbers of yellowtail and white seabass. Lea and Vojkovich (1985) report an
439
a
A
1,000
100
200
0
DISTANCE OFFSHORE ( k m ) I
90.65
.60
I
1
.55 .53
I
.50
I
A5
1
,442
1
.37
I
l
l
.32 .30 .28
LINE 90 STATIONS 90.28 THRU 90.65 Fig. 2. Zooplankton biomass vs. distance offshore along line 90 of the CalCOFI grid. All samples taken in years 1949-1969 (not including the 1957-1959 El Nifio) in the months July, August, and September are shown. Data from 1983 are shown as X and their range stippled. The dashed line connects the station medians. From McGowan (1985).
440
extremely rare occurrence of the tropical Pacific burrfish,
a, in Long Beach
Harbor, and Lea et al. (1988) add an extension from Baja California to Newport Bay of the white mullet, &giJm; both were clearly related to the ENS0 event. Pelagic red crabs DlaniDes), normally found off central Baja California, moved north during this El Niiio to a record high latitude, Fort Bragg, by early 1985 (Pearcy et al., 1985). Radovitch (1961) reviews similar fishery changes and the many biological anomalies of the 1957-1959 warming event. Fiedler et al. (1986) summarized the physical and biological effects on the habitat of the northern anchovy; the biological effects include the reduction of phytoplankton and zooplankton and an extreme deepening of the chlorophyll maximum layer. The anchovy population responded with an expanded spawning range, reduced fecundity, abnormally high yolk-sac mortality, low recruitment, and reduced growth of juveniles and adults. Spawning effort in 1984 was high in spite of the reduced size and abundance of spawning females. The length-at-age and stock size returned to pre-El Niiio levels by early 1985 because of the strong 1984 year class. In contrast, rockfishes have limited behavioral mobility to pursue better conditions; as adults, these animals tend to have home ranges or to be territorial. In 1983, Lenarz and Echeverria (1986) reported reduced levels of visceral and gonadal fat in yellowtail rockfish (Sebastes flavidus')from Central California, which appeared to result from decreased availability of planktonic prey. It was not clear whether this would result in delayed parturition and/or lower reproductive effort. Ven Tresca (pers. comm.) found extremely low gonadal production in blue rockfish (Sebastes mvstinus') and poor larval recruitment in 1982 and 1983 in Monterey Bay. Seabirds and pinnipeds are top level predators in the California Current system and thus their survival and reproduction integrate changes in lower trophic levels. Ainley et al. (1988) describe changes in seabird distributions with shifts in their preferred water types, e.g., species that vacated the Gulf of California during 1983 reappeared at the Farallon Islands in Central California. Some subsurface-feeding seabird species that require fairly reliable food resources experienced high mortality, and all but one of the species that nest at the Farallons showed reduced reproductive success in 1983. El Niiio effects on seabird reproduction were apparent in 1982, before the physical environment had changed enough to be noted by oceanographers, and continued through 1984 (Ainley et al., 1988). Brown pelican ( P e l e c a n u s m ' califomicus)reproductive success in the Southern California Bight is closely related to the anchovy spawning biomass (MacCall, 1986). The number of young fledged per pair of adults was not affected by the El Niiio, but there was a substantial decline in the number of pairs nesting, especially in 1984 (MacCall, pers. comm.). While some fish-eating species suffered, the many large swanns of red crabs that washed up on San Nicolas and San Miguel Islands during 1983 provided an easily exploited food source for sea gulls, which Stewart et al. (1984) suggest benefited from the El Niiio. Sea anemones and even black abalones likewise profited (VanBlaricom and Stewart, 1986). Further north, the 1983 breeding season was an exceptionally poor one for three of Oregon's nesting seabird species where nest abandonment by cormorants was widespread (Graybill and Hodder, 1985). In Alaska, there was also a broadscale breeding failure of some surface feeding seabirds and a large die off of several species of adult birds in 1983, apparently due to starvation (Hatch,
44 1
1987). Pearcy and Schoener (1987) review the poor nesting success and high mortality of other seabirds in the northeast Pacific in 1983. Stewart et al. (in press) report that some of the Channel Island pinniped populations exhibited strong declines in births and perhaps survival during the recent ENSO event, evidently because of reductions in prey availability. Lactating California sea lions (ZaloDhusmlifornianu ) on San Nicolas Island spent much more time at sea and offered less suckling time with less milk; not unexpectedly, their pups suffered higher mortality (Ono, 1987; see Lmberger, this volume). The effects on the local population may go beyond the poor recruitment of this cohort because preceding cohorts may have experienced lower survival and slower physical growth during the El NiAo. While details are lacking, Hubbs (1948), Radovitch (1961), and Kuhn and Shepherd (1984) report many types of historical evidence, including long term air and sea temperature data, indicating the occurrence of large El Niiio events in the past. Hubbs (1948) reported on good biological collections as far back as the middle of the 19th century in which the warm periods were associated with remarkable northward displacements of fish faunas. In many cases, there were other El Niiio indicators such as northward occurrences of pelagic red crabs (Radovitch, 1961). Estimated sea "surface" temperatures (from 5 m below the surface) spanning 300 years imply that both El Niiio and "anti El Niiio" conditions are part of a long term recurrent pattern (Dayton and Tegner, 1984b; Kuhn and Shepherd, 1984).
3 ENSO EFFECTS ON KELP FORESTS Kelp forests are among the richest and structurally most diverse of temperate marine communities. They include several algal guilds, suspension feeding species, and an abundance of herbivore and carnivore species occupying an assortment of microhabitats and composing complicated food webs. Like most natural communities, kelp forests are composed of distinct patches and are influenced by various physical and biological processes. All kelp forests are different from each other to some degree; nevertheless, all have similar substratum needs and are to some extent structured by similar processes such as the availability of light, nutrients, and the effects of herbivory, storms, etc. To the degree that kelp forests do share these same processes, any particular kelp forest can be used as a model to evaluate the mechanisms; we have used the large kelp forest off Point Loma, San Diego, as our model. Where available we also discuss data from other sites. These communities generally are characterized by the giant kelp, M a c r m Dvrifera, a species noted for large standing stocks and high rates of growth, production, and turnover (North, 1971; Mann, 1982). A great deal of kelp forest structure is determined by substrata heterogeneity and by various types of physical and biological interactions. Many kelp habitats have distinct patches of seaweeds that can be grouped by general morphology or structure of the vegetation layers they produce. In Southern California there is (1) a canopy composed of Macrocvstis, m p h v c u , and Emepia, which support their fronds at or near the surface by floats of various types, (2) a canopy supported h and Eisenia), (3) a canopy in which the fronds lie on above the substratum by stipes or immediately above the substratum &aminaria and &vum), (4) a mulaspecific turf composed
442
primarily of species of red algae, and (5) a relatively clean pavement of encrusting coralline algae. While subject to many types of disturbances, the patches in the Point Loma kelp forest exhibited considerable stability between 1970 through 1981 (Dayton et al., 1984); for most species this persistence covered several generations. In contrast, patch smcture may not be so clearly maintained in more heavily disturbed communities or habitats with broken, heterogeneous substrata. Like most coastal communities, kelp forests naturally are exposed to a wide array of disturbances. The most important disturbances to this community are from storms and grazing invertebrates (see summaries in Dayton et al., 1984; Ebeling et al., 1985; Tegner and Dayton, 1987). The effects of storms vary widely between kelp forests as a function of the frequency and magnitude of the storms and the direction of exposure; all modified by local topography and the presence of protecting islands or points of land. The stability of the substratum is another factor especially important to the persistence of kelp plants. Finally, the physiological or demographic status of the existing populations are difficult to categorize but can make the timing of the disturbance critical. Spatially-small disturbance patches are common and usually are colonized by members of the existing patch type. Larger disturbance patches might be colonized by other species depending upon their relative dispersal abilities, various physiological thresholds, and reproductive seasonality (Dayton, 1985). An important second order result of large-scale disturbances is the effect on the availability of detached algal material that is the major food of sea urchins. When the supply of drift algae does not meet urchin needs, the urchins starve and tend to forage out from refuges and can destroy large areas of kelp forest (Harrold and Reed, 1985; Ebeling et al., 1985; see references in Dayton, 1985 and VanBlaricom and Estes, 1988). These points are important to understand some of the events subsequent to the 1982-84 ENSO. For example, the kelp forest on Naples Reef in Santa Barbara had earlier been devastated by a 1980 storm that came from an unusual direction and tore out much of the algal standing stock. This resulted in intense grazing pressure on extant plants and the reef soon was taken over by sea urchins. Thus Naples Reef differed from many other kelp forests in that it was an urchin barren immediately before the 1982-84 ENSO event (Ebeling et al., 1985). 3.1 Effects of the stormS The extraordinary storm season of 1982-1983 set many records, e.g. of 18 storms with wave heights exceeding six meters since the turn of the century, 6 of these storms occurred during this winter. There were eight storms between January and March with wave periods between 17-22 seconds compared with only one storm with a period approaching 17 seconds during the previous three years (Seymour et al., 1984; Seymour and Sonu, 1985). Thus there were more storms, with ' canopies. The Point Loma larger waves and longer periods, and they devastated canopy, which covered 600 ha in the fall of 1982, was eliminated. In addition to removing the canopy, storms also ripped holdfasts and their attached stipes from the substratum, and they became entangled with attached plants either ripping them free or causing extensive stipe loss. + plants at our five study sites in the Point Loma kelp forest From 16 to 66%of the were killed by the storms; the damage was greatest in shallow water, decreasing with depth and at
443
the long-shore edges of the forest (Dayton and Tegner, 1984a). The consequences of the storms for subsequent changes in the community included (1) high light levels on the bottom due to loss of the Macrocvstis canopy, (2) the provision of much open space, and (3) a much wider dispersal of algal reproductive material than we saw between 1971 and 1980 (Dayton et al., 1984). Upwelling approached near normal levels during the spring of 1983 and bottom temperatures dropped to 13OC (Tegner and Dayton, 1987). These physical factors created a "recruitment window" (Deysher and Dean, 1986) that led to wide spread, extremely heavy recruitment of MacrocysUs ' that appeared to be general throughout the Southern California Bight. At Point Loma this recruitment, which also included m uh a and Laminaria, was much more apparent in the more heavily disturbed, shallow sites where we observed hundreds of juvenile kelps per square meter. In addition, several opportunistic species . . such as D e s m a r e m DlctvoDtens ' and' ,normally rare at Point Loma, became abundant and occurred in high density patches. Desmaresaa ' was especially abundant in the more shallow sites, Dictvou- ' in the deeper sites, while Acroson'm was most abundant in the north and south sites (Tegner and Dayton, 1987). This burst of kelp recruitment also occurred in Central California as VanBlaricom (pers. comm.) reported m u h o r a densities at Point Piedras Blancas commonly over 6OO/m2 in 1983. 3.2 Effects o f the warm warn For the coastal areas of Southern California, the extraordinarily warm water associated with the El Niiio occurred during the late summer and fall of 1983 and through most of 1984. The normal sources of nutrients to the kelp forest were essentially eliminated for this period, upwelling was ineffective and the thermocline dropped below kelp forest depths as indicated by anomalously high bottom temperatures (normally below 16OC but above that for three months of 1983, summarized ' which, with in Tegner and Dayton, 1987). The warm water per se does not h a m abundant nutrients, grows well even at 25OC (reviewed by North and Zimmerman, 1984). The main relevance of this high temperature is that it is correlated with low nutrients; there a~ negligible amounts of nitrate, the nutrient most likely to limit growth (North et al., 1982) above 15OC (Jackson, 1977; Gerard, 1982; Zimmerman and Kremer, 1984). Due to good growth conditions during the spring, many of the Macrocv s h plants that survived the storms had fronds on the surface by early summer; however, 20 to 67% of these survivors died of apparent nutrient smss during the late summer and fall of 1983 (Dayton and Tegner, 1984a; Tegner and Dayton, 1987). Interestingly, in this case the northern and southem ends of the kelp forest had much lower mortality (22 and 20%) than the central part (58 to 67%) where there was no offshore-onshore gradient. The hypothesis that essentially all this mortality results from nitrogen deficiencies is strongly supported by a considerable body of research by Gerard (1982, 1984), North et al. (1982) and North and Zimmerman (1984). Gerard has shown that 1% tissue nitrogen (on a dry weight basis) is a critical level below which internal nitrogen reserves are depleted and nitrogen starvation will lead to rapid deterioration, Her study of the physiological effects of the El Niiio on a Macrocv& population off Laguna Beach (Fig. 3) illustrates the relationship between canopy cover, temperature, tissue nitrogen, and growth rates (Gerard, 1984).
A
I
I
I
I
I
B
A
M
J
J
A
Fig. 3. Data for the Macrocvstis pyrifera population at Laguna Beach, California, during AprilAugust, 1983. A. Canopy density. B. Daily minimum and maximum temperatures at 2 m depth. C. Nitrogen content of mature blades on canopy fronds (-x-), subcanopy fronds (-0-), and small fronds (+); si k 1 SE, n = 5-6. Data are not shown for older blades on canopy and subcanopy fronds that had slightly higher N-content than mature blades. 0. Standard growth rate of fronds as in C; jT & 1 SE, n = 3-10. From Gerard (1984).
445
She and others (Gerard, 1986; Gerard and Du Bois, 1988) have more recently argued that both nitrogen starvation and high temperature can result in reduced carbon assimilation. Our Point Loma data from 1983 and 1984 corroborated Gerard's conclusions regarding nitrogen content and canopy deterioration (Tegner andDayton, 1987). The fact that the longshore edges of the kelp forest had greater survivorship also supports the nutrient starvation hypothesis, because with no thermocline present, there was no cross shore internal wave pumping and the nutrient flux would have to come via longshore currents that would offer the ends of the forest first chance at whatever nutrients were available. In any case, the 1983 winter and summer mortality at Point Loma was much higher than comparable mortality patterns during 1970-1981 (Dayton and Tegner, 1984a). Of the heavy Macrocvshs ' recruitment at Point Loma in the spring of 1983, very few of these recruits survived long. Their mortality appeared toresult from several factors. Most understory populations survived the storms, although they may have lost considerable blade tissue (Dayton and Tegner, 1984a); these are known to inhibit the recruitment of kelps (reviewed in Dayton, 1985) and in some areas this understory competition may have had an effect on the young Macrocvs&. As understory plants recovered from storm damage, they precluded further growth of young Macrocysas ' that recruited into those patches. However, very high densities of also recruited in cleared areas in early 1983 and were growing very rapidly until the warm, presumably nutrient-poor water appeared in midsummer. By September 1983, most of these plants had died back to within 2-3 meters of the bottom, and even these were often discolored and diseased. Desmaresha ' ,the opportunist that bloomed along with the kelps at the shallowest site, also appeared to interfere with young Macrocvstis. While the growth of juvenile Macrocvst is in coastal kelp forests of Southern California is usually limited by irradiance, Dean and Jacobsen (1986) experimentally demonstrated in the San Onofre forest that the poor growth during 1983 and 1984 was a result of nutrient limitation rather than temperature stress or low light. Indeed, they showed that irradiance levels were much higher from 1982 through 1984 than those observed in the previous few years. Interestingly, cooler water reappeared in November 1983, surviving Macrocvsn's adults began to form a canopy, and surviving spring recruits grew into the adult classification (an arbitrary definition of 4 stipes, which is usually the point at which sporophyll growth begins); these data are presented in Fig. 4. There was also another heavy recruitment of ' in some areas that winter. However, the surface warmed much faster in 1984 than 1983, and summer temperatures were higher than the year before. Deterioration of the Point Loma canopy was underway by June of 1984, when nitrogen values of surface blades were well below Gerard's critical level. The canopy was largely gone by August (Tegner and Dayton, 1987). The extent of the disturbances to the Macro' canopy can be seen in the harvest data; the top 1.2 m is cut periodically for the production of alginates. In 1982, Kelco, the local kelp harvesting company, collected 75% of its long term average at Point Loma, but nothing was harvested in 1983, and only 9% in 1984, all at the beginning of the year (R. McPeak, quoted in Tegner and Dayton, 1987). The upwelling that occurred in the middle of the 1982-84 ENS0 apparently did not occur in the 1957-59 event from which the kelp forests recovered very slowly (North, 1971). Thus the presence or absence of such upwelling events may have very long term consequences.
u
.
u
1983
I
.
.
.
I 1984
.
.
.
I 1985
.
.
.
I 1986
.
. 1987
Year Fig. 4. Density of adult (four or more stipes) Macr' at 12 m site at Point Loma. Plants that recruited in the spring of 1983 entered the adult classification by December. The number of adults declined during the warm summer of 1984 and increased again as a new cohort of recruits attained this size. Oceanographic conditions returned to "normal" by the end of 1984 and were very conducive to ' kelp growth; in 1985 flourished throughout most of Cennal and Southern California. At Point Loma ' recovered its competitive dominance and the canopy reformed rapidly during the winter. Then, during a period with essentially no storms and cool temperatures during the season when kelps grow the fastest, the canopy started to disappear in large areas, especially in the southern part of the forest. We found that there was an infestation of gammarid amphipods, especially the kelp curler, Amphithoe humeralis. Macrocvstis growth tips deteriorated, growth rates were anomalously low, stipe counts declined, and there was a burst in the availability of algal drift. The early stages of the infestation resembled nutrient-stressed plants despite relatively cold water. Nitrate determinations indicated that the relationship between temperature and nutrients had not broken down. Eventually the increasingly apparent grazing damage made it obvious that the amphipods were responsible for all the damage. Grazing damage occurred throughout the 10 km long forest but large areas of the southern end, including one of our major study sites and two control areas, were completely denuded of all macroalgae except the encrusting corallines (Tegner and Dayton, 1987). We documented patterns of grazing, amphipod dispersal, and plant loss. form nests (or curls) in Macrocystis and understory kelps by laminating or folding the edges together
447
while distal portions of the blade are eaten or severed. This accounted for the increases in algal drift, which included many amphipod nests. We observed nocturnal swanning that suggested infestation through the water column as well. Soon grazing damage and curls were spread throughout the water column but were most common on the bottom half; in severe cases there were no Macrblades below 6m. Planimetric analysis of aerial photographs indicated that the Macro' canopy declined from 632 ha in January to 275 ha in July (R. McPeak, pers. comm.), almost a 60% reduction during the season when the kelps normally grow the fastest (Tegner and Dayton, 1987). The best explanation for this infestation appears to be a population reduction of the guild of fishes that specialize on invertebrates associated with w o c v. stis fronds, especially the kelp surf perch, ' ' frenatus,studied by Coyer (1979). is normally closely associated with kelp canopies and its populations decline or disappear entirely when the canopy is reduced (Coyer, 1979; Ebeling et al., 1980). These fishes live for two to three years (Hubbs & Hubbs, 1954) and are viviparous (Baltz, 1984). The canopy at Point Loma was lost three times in only . . . There are no pre-ENS0 two years, and this must have had a severe effect on the . . quantitative data at Point Loma, but swarms of Brachvlstlus were common (pers. obs.; E. . . DeMartini, pers. comm.). B r a c h v i m were rare during the period when the amphipods were destroying the Macrocvs&canopy. By late 1985, the fish populations were beginning to recover and respond to the amphipod infestation. While they are known primarily as daytime feeders (Hobson and Chess, 1976), large animals from south Point Loma switched to night time feeding, presumably to take advantage of the nocturnal amphipod swarming behavior. We found that 100% . . of the B r a c h v i w sampled at Point Loma had gammarid amphipods in their guts compared with 43% from uninfested locations. Amphipod outbreaks were also observed in coastal forests around Santa Barbara and near Point Dume. It is interesting that a similar outbreak was observed after the 1957-59 ENSO, when Jones (1965) described a peracaxid crustacean infestation of Macrocystis at Point Loma. In this instance the canopy was completely eliminated between stations 2.4 km apart during the winter of 1965 when Jones reported large numbers of Amphithor and the isopod Iodothearesecataconsuming the kelp fronds. The 1965 phenomenon was associated with canopy deterioration, but at that time the plants appeared normal below 3m. The infestation of 1985 was much more severe because the grazing affected &lacrocw throughout the water column, including the understory as well in some areas (Tegner and Dayton, 1987). 3.3 Other Californiakelp habit& The storms had dramatic effects on the surface canopies of Macrocvs throughout Central and Southern California, whereas W o c v s n's mortality induced by the storms and the warm water summer-fall seasons was variable (Tegner and Dayton, 1987). With the exception of the lee side of Santa Catalina Island (Zimmeman and Robertson, 1985). the storms caused massive damage to forests from San Diego (Dayton and Tegner, 1984a), to the San Onofre Area (Elliott and Foster, 1987), to Palos Verdes (Wilson and Togstad, 1983), Naples Reef (Ebeling et al., 1985) and San Nicolas Island (Fig. 5,Estes and Harrold, 1988) in Southern California, to Diablo Canyon (Kimura, 1985) and the Monterey Bay Region (M. Foster, J. Roughgarden, G.R. VanBlaricom,
448
2000
1,500
cn
w
U
g
1,000
2r 500
0
JAN 1981
JAN 1982
JAN 1983
JAN 1984
JAN 1985
JAN 1986
'
Fig. 5. Temporal variation in the areal extent of the surface canopy at San Nicolas ' gyrifera Island, 1981-1985. From Estes and Harrold (1988). The total area of Macrocvm canopy at San Nicolas Island is approximately 2,000 ha. pers. c o r n . ) in Central California (Fig, 6). The Point Loma kelp canopy was reduced from well over 600 ha in the fall of 1982 to essentially zero by March 1983 (Dayton and Tegner, 1984a), and the densities of adult kelp plants at San Onofre sites were reduced from about 50 to 75% (Elliott and Foster, 1987, fig. 4.7). During only three weeks in January, the canopy at Palos Verdes Peninsula in Los Angeles County declined from 196 to 18 ha (Wilson and Togstad, 1983). The large kelp forests in the Santa Barbara Channel were essentially eliminated (R. McPeak, pers. c o r n . ) . The loss of the Santa Barbara kelp forests may have long term consequences because these forests are composed of anrmstifolla ' ,a species (or perhaps a variety of fi pyriferil) that forms a very large holdfast system that is loosely anchored in the sand and produces a great deal of asexual growth; these expanded holdfast systems act as spore settlement sites in an otherwise unacceptable sand habitat. When the storms eliminated these kelps they also eliminated recruitment sites, which means that natural recovery could be very slow. While not well documented, similar devastation occurred to kelp forests from Isla Asuncion, Baja California almost to the mouth of San Francisco Bay, the entire range of ' Dvrifera forests in North America. This damage resulted in the lowest kelp standing stock in history of the local kelp harvesting company (R. McPeak, Kelco Company, pers. c o r n . ) . Again, an exception to this pattern was at Naples Reef, Santa Barbara County, where the kelp forest had been transformed into a sea urchin barren (sensu Lawrence, 1975) after the 1980 storm. The storms of 1983 washed the exposed urchins off the reef and released the kelps from these grazers. The kelps responded with strong recruitment (Ebeling et al., 1985), in effect reversing the more typical Southern California pattern of storm devastation of kelp forests. Another interesting
449
Fig. 6. Aerial photographs from Soberanes Point, Monterey County, Central California, courtesy of G. VanBlaricom. Shown is the variation in &lacrow& prifera canopy cover over spring (first column), late summer (middle column) and winter conditions (third column) from 1983 (A-C) through 1986 (J-L). consequence of the 1983 storms at Naples Reef was that the reef itself was heavily damaged, leading to the exposure of considerable freshly exposed rock. These new surfaces supported ten times higher ’ recruit density than did the old rock, where most of the developing flora were red algae regenerating from old basal stems; the new rock surfaces were covered with dense growths of various species of filamentous brown algae that apparently facilitated the survival of -by protecting it from grazing fishes (Harris et al., 1984). Published negative impacts of the warm water events as compared with storm damage are more limited (Tegner and Dayton, 1987). Apparently the only effects were in the Southern California Bight, with major effects reported at Santa Catalina Island (Zimmerman and Robertson, 1985), and along the mainland coast from Laguna Beach (Gerard, 1984) to San Diego (Dayton and Tegner, 1984a). Most of the biomass of a healthy population is in the upper one meter of water (North et al., 1982), and this is the part of the water column with the most serious nutrient
450 depletion (Jackson, 1977). Surface water temperatures offshore of the Palos Verdes Peninsula in Los Angeles County were anomalously warm for an extended period of 1983 but the canopy flourished (Tegner, unpub. data). B. Jones (pers. comm. reported in Tegner and Dayton, 1987) has recently found high concentrations of presumably from the nearby Los Angeles County sewage outfall in kelp forest depths during the summer. It is interesting to note that the kelp forests in the lee at Santa Catalina Island were protected from the 1982-83 storms but were ravaged by the warm, nutrient-poor water associated with the ENSO event. These kelps at Santa Catalina depend upon vertical excursions of the thermocline for their summer nutrients but the ' above thermocline was depressed well below 50 m, twice as deep as usual, and all the 10 m had disappeared by November 1983 (Zimmerman and Robertson, 1985). Further south in Baja California, where the waters were even warmer and more nutrientdepleted, giant kelp beds disappeared completely from some areas. This is nicely documented by Hernandez (in press) who measured the surface biomass, growth, survival, and recovery of M a c r o c a in relation to surface and bottom water temperatures. He did this at six kelp forests throughout their distribution along the west coast of Baja California and Baja California Sur (B.C.S.). In this area it is important to record the water temperatures because some of these sites have strong upwelling and cold temperatures, which strongly contrast with the much warmer temperatures along the rest of the coast pawson, 1951). Hernandez documented the deterioration ' of in temperatures often over 26OC. He tagged plants and watched them die in a matter of days. The forests at the northern sites recovered, but interestingly, the best recruitment was in the relatively southern Bahia Tortugas, B.C.S. where they had suffered 100% mortality. In contrast, the far southern Bahia Asuncion and Punta Prieta, B.C.S. sites were invaded by the understory kelp, which apparently prevented recovery even after the water temperatures had returned to normal.
m+,
m,
4 ENSO IMPACTS ON KELP FOREST ANIMALS 4.1 S e a urchins ' Sea urchins are one of the most important groups of invertebrates in temperate coastal systems. This is especially true in the North Pacific where they often are conspicuous members of the community, and when they occur in sufficient density, sea urchins are consistently the most effective and potentially devastating grazers in temperate algal systems (reviewed in Dayton, 1985). Investigations of the Point Loma kelp bed in the late 1950s and early 1960s revealed large populations of sea urchins in areas where the kelp forest had disappeared (IMR,1963; North, 1971). While the kelp decline was associated with dense populations of w l o c entrotys franciscanusand S,DUTt)UTatUS, the role of the 1957-59 ENSO is not clear. The ENSO could have resulted in 1) a loss of drift algal forage followed by destructive urchin grazing, 2) a reduction in the populations of predators of urchins leading to a release in the urchin populations, 3) increased settlement of urchins which then overgrazed the kelps, 4)any combination of the above, or 5) none of the above. While the cause and effect relationships are not clear, there was evidence that the recovery of the kelp forest at Point Loma was enhanced by urchin control measures (North and
45 1 Pearse, 1970; Leighton, 1971). These urchin species normally feed on algae that drift to their microhabitats in reefs and rock piles. Hungry urchins will leave their habitats in search of food (Dean et al., 1984; Harrold and Reed, 1985). These urchins often aggregate and move as fronts denuding large areas of kelps, leaving only a pavement of encrusting coralline algae. Enough foraging urchins remain to consume all kelp recruitment so that these barren patches may persist for years (reviewed in Dayton, 1985). Since this shift could be caused by the ENSO event as a result of a reduction or complete loss of drift algal stocks, we quantified the amount of drift at Point Loma and compared it to data from other sites and times (Tegner and Dayton, 1987). We found that while our drift abundance was lower than that found by Gerard (1976) in Monterey Bay, it was comparable to data from pre-El Niiio years at San Nicolas Island (Harrold and Reed, 1985). The kelp canopy completely disappeared during the summers of 1983 and 1984, but during this period there was considerable recruitment and death of subsurface kelps producing a continuous if small amount of drift algae. With the exception of one transect line in the southern part of the kelp forest, sea urchin barrens did not form at any of our five sites during the ENSO years (Tegner and Dayton, 1987). Our sea urchin recruitment data for non-El Niiio periods (Tegner and Dayton, 1981, 1987, and unpub.; Dayton and Tegner, 1984b; Tegner and Barry, 1987) suggest significant annual recruitment in the San Diego area and the southeastern Channel Islands, apparently as a result of the Southern California Eddy, and low but regular recruitment in the northwestern Channel Islands. We had previously found that the recruitment of S, franciscanuswas highest on the outside edge of the Point Loma kelp bed, or in the center only after localized disturbance of the canopy by storms (Tegner and Dayton, 1981). When the 1982-83 storms devastated the canopy, we assumed that there would be strong recruitment of urchins throughout the kelp forest; the initial and summer losses of the canopies followed by the partial winter recoveries minimized the edge effect for almost two years compared to pre-ENS0 canopies. However, urchin recruitment during the 1982-84 ENSO event was much lower than we had seen before and it was zero in some areas (Tegner and Dayton, 1987). One explanation for the reduced recruitment could be a reduction in the availability of the larvae, which are planktonic for about 50 days (Cameron and Schroeter, 1980). This could result from a decline in egg production, starvation of larvae, or a change in the currents. We did not monitor sea urchin gonadal development during this period, but the roe yield of the commercial fishery was lower than normal, often to the point of closing the fishery (Fig. 7). Thus there could have been a general reduction in larval availability. Certainly the California Current was very much changed during this period and the surface waters had extremely low nutrient concentrations and primary and secondary productivity (McGowan, 1985). As a result, larval survival may have been poor and those that survived may have been transported to unusual areas; in fact, there was a burst in urchin recruitment in the northern Channel Islands (Tegner and Barry, 1987) and Davis (1985) has followed a burst of the sea urchin , apparently related to the El Niiio, that has subsequently resulted in major algal losses. Thus all three mechanisms were probably involved in the poor recruitment of red urchins observed at Point Loma during the ENSO event vegner and
12000 10000
8000
6000 4000 2000 0
70
72
74
76
78 80 YEAR
82
84
86
Fig. 7. Landings from the Southern California red sea urchin rStrongvlocentrom b c i s c a u ) fishery, 1970-1986. Landings, almost all of which are exported to Japan, are sensitive to fluctuations in the monetary exchange rate as well as environmental problems. The storms of November 1982 through April 1983 reduced access during the major fishing season. Reduced landings through the rest of 1983 and 1984 probably reflect reduced kelp stocks and poor nitrogen content of the remaining kelp in many areas. Data from the California Department of Fish and Game. Dayton, 1987). 4.2 Abalones There are no comparable recruitment data for abalones, but Tegner (Tegner and Dayton, 1987) has studied recruitment and gonadal development of green abalones, in a shallow (4-6 m) kelp forest off the Palos Verdes Peninsula. Here the storms removed all the ,and the abalones, which also rely on drift algae, had virtually nothing to eat from January to May, 1983. This is usually a period with ample food, and the time of gonadal growth before the late spring spawning. In the spring of 1983, however, gonadal growth was severely depressed have a very short planktonic period, 6 (Tegner and Dayton, 1987). Since the larvae of to 8 days, (Leighton et al., 1981), it is likely that there was a severe local reduction in recruitment. Back calculations from size frequency data collected from somewhat deeper (6-12 m) water at also had suppressed Palos Verdes in 1986 suggest that red abalones recruitment during the ENSO event, but almost all the pink abalones (Haliotis)appeared to have recruited during this time (Tegner and Dayton, 1987). Summarizing, it appears that the recruitment of all three species of abalones was affected by the ENSO event, which supports the conclusion that coastal larvae were strongly influenced by the altered oceanographic conditions. That the larvae of the pink abalones apparently responded in diametrically opposite directions from
m,
453
the other larvae emphasizes how much more we need to learn about the biology of even relatively simple, short-lived larvae. 4.3 Asteroids and lobsterS The ENSO may have had an effect on the kelp forest asteroids with cascading ramifications (Tegner and Dayton, 1987). A dramatic effect of warm water was observed in shallow-water populations of starfish, especially the batstar, miniata. has suffered several mass mortalities in the last decade, which Dixon and Schroeter (unpub. ms) have shown to be associated with warm water. In the Channel Islands National Park kelp forest survey, the numbers dropped from over 1.43/m2 in 1982 to 0.86 in 1983, and 0.14 in 1984 (Davis, 1985). Our data from Pt. Loma showed similar declines on top of another order of magnitude drop in the late 1970s for and several other species of asteroids (Tegner and Dayton, 1987). Where the white sea urchin occurs, predation appears to be able to restrict its distribution to areas outside of kelp forests (Schroeter et al., 1983), and in the Channel Islands, Davis (1985) reported that densities increased from 0.97/m2 in 1984 to 8.27/m2 in 1985. By 1985, destructive grazing by Lvtechinus was reported from the Channel Islands, Palos Verdes, and San Onofre kelp forests of the Southern California Bight (reviewed in Tegner and Dayton, 1987). Before the ENSO event, the effect on the San Onofre kelp forest was restricted to grazing on the small stages of kelps and temporarily reducing kelp recruitment (Dean et al., 1984). But -densities fell from 1-2/m2 in 1981 to almost zero in 1983, and the ; by creating a refuge for -populations exploded and began grazing on adult understory species, they may have a long lasting effect on the kelp forest (Dixon and Schroeter, unpub. ms). The California spiny lobster, Panulirusjntermpw ,is a more important predator of strongylocentrotids than are the asteroids in southern California, because the urchins have much less of a growth escape from lobsters, which are capable of killing much larger urchins than most asteroids (Tegner and Dayton, 1981; Tegner and Levin, 1983). Spiny lobsters recruit into shallow waters and as they grow they become migratory, moving into and out of kelp forests (Engle, 1979). Cowen (1985) reported that spiny lobsters recruited heavily at San Nicolas Island in 1983, whereas he had not seen any recruitment during the previous four years. We have anecdotal evidence from sporadic pre-ENS0 observations of a few, very large and presumably old lobsters at the more northerly San Nicolas Island where we never saw any sign of recruitment or even juvenile lobsters. The northernmost San Miguel Island has very few but very large lobsters. These observations support Cowen's argument that the recruitment at the far north of their range may depend upon unusual dispersal events that could occur during El Niiio conditions. Pringle (1986) reviewed CalCOFI collections of lobster larvae and came to the same conclusion. Most of the lobster range is far to the south and the effect of ENSO events seems to be that of moving competent larvae north of the normal settling range. Thus El Niiio events are especially important in the more northern area of a species' range.
a
a
454 4.4 Kelp forest fishes Cowen (1985) and Cowen et al. (1988) consider the recruitment of coastal fishes to be functions of larval source areas, the duration of competency and, most importantly, patterns and variation of current flow. For Southern California populations with larval sources in both the north and the south, there is relatively consistent year-to-year recruitment. For populations in which there are no northern larval sources (upstream in the California Current), recruitment is
variable and depends upon anomalous events that they separate into low and high level anomalies. Low level events occur in situations in which the flow is mostly from the north but some southerly water becomes mixed along the axis of the Southern California Eddy or via small meandering intrusions. These events have short time periods of days and weeks. High level anomalies are typically associated with El Niiio events that bring large amounts of warm southern water northwar& these events have a much longer duration of months and years. Upstream, northerly dispersal into areas where reproductive adults are rare or lacking depends upon these anomalous oceanographic events. The sheephead, w o s s y p h u a southern species with a relatively long-lived (about two months) larva, showed a dramatic increase in recruitment at the northern end of its range in 1983 (Fig. 8). Cowen (1985) suggests that fish like the sheephead are more likely to have occasional dispersal via the ephemeral low level events than species with short lived larvae such as gobies; such low level events probably account for the low levels of sheephead recruitment seen in
e,
non-El Nifio years at San Nicolas (Fig. 8). Cowen summarizes evidence substantiating such patterns for several other species of coastal fishes and invertebrates. In addition, he demonstrates that there is a strong relationship between density of such species and the frequency of recruitment events. In effect, for southern species, the further upstream the more evidence for episodic and increasingly rare recruitment events. This can be seen for long-lived species such as the sheephead (which live 20-25 years) in the strongly skewed age/size distributions, but short-lived species such as the gobies simply disappear after a high level event. Clearly such high level oceanographic events integrated with the duration of larval competency and adult life span are critical factors in the biogeography of coastal species. These events are also important to kelp forests because spiny lobsters and sheephead were the two major predators of the strongylocentrotid urchins in Southern California (Tegner and Dayton, 1981; Cowen, 1983; Tegner and Levin, 1983) before the recent transplant of sea otters to San Nicolas Island Populations of both are related to El Niiio events, but their effects on urchin populations will be delayed several years after ENS0 events until they are large enough to handle the adult urchin prey. Patton (1985), in surveys of many mainland sites spread throughout Southern California, found that the mean abundances of most southern fish species increased and most northern species decreased during 1983 and 1984 relative to 1979-1982. He also documented that fishes of each group tend to occupy a broader range of habitats when temperatures are optimal. Ebeling and Laur (1988) report examples of similar shifts of southern and northern species at Naples Reef even though the results were confounded by the fishes' responses to storm-induced changes in structural habitat. Recruitment of juveniles provided clearer evidence; e.g., there was a large
455
I
SAN
m
s1
z 0 L
I
CAB0 THURLOE
20
10
75 76 77 7a 79 a0 a1 a2 a 3
YEAR
Fig. 8. Relative annual recruitment success of sheephead ( S e m i c o s s v m pulcher) near the northern end of its range (San Nicolas Island) and at three sites in Baja California towards the southern end of its range. ND - no data; 0 - site visited but no recruits observed. Data prior to 1980 based on the age structure of the population. Data from 1980-1983 based on field counts and given as 1 S.E. From Cowen (1985).
*
456 increase in the young of the tropical family Labridae -- sheephead, seiioritas (oxviulism, and rock wrasses (J3alichoeres semicinctua). The density of adult sheephead on Naples Reef more than quadrupled between 1981 and 1986 resulting from the record recruitment during the El Niiio and subsequent growth (Ebeling and Laur, 1988). Finally, Cowen and Bodkin (in press) report many other species of southern fishes with apparent ENSO-induced recruitment at San Nicolas Island. Fish communities in Central California kelp forests were affected by both the storms and the northward transport of w m , southern water north. Bodkin et al. (1987) reported three incidents of mass mortality of nearshore fishes at Piedras Blancas corresponding with three large wave events during the winter of 1983. Beachcast fishes occurred about in proportion to patterns of abundance in the fish assemblage in the adjacent kelp forest. Later in the year, Bodkin (1986) noted an increase in seiiorita populations and a decrease in blue rockfkh, the usual dominant, which he related to the ENSO event. Sheephead recruited as far north as Monterey in 1983, nearly 250 km north of their normal range (Cowen, 1985). While the southern species had a banner year in Central California in 1983, there was nearly a complete recruitment failure of rockfishes associated with the kelp forests of Monterey (Gaines and Roughgarden, 1987). Summarizing, the 1982-84 ENSO event is associated with anomalous northern recruitment of many southern coastal species but what about northern-affinity fishes in Southern California? Cowen and Bodkin (in press) found no significant changes in northern species at San Nicolas Island, which is heavily influenced by the cooler waters of the California Current. However, further to the southeast in the warmer area of the Southern California Bight, the pattern was different. Stephens et al. (1986) have studied the larval and adult fish assemblages at King Harbor near Los Angeles from 1974 through 1985, a period which includes the minor El Niiio of 1978-79 as well as the recent strong event. As observed by others, they found that the warming trends of recent years have increased the proportion of warm temperate fishes and reduced the abundances of cool species. In particular, five species of surf perches have declined in abundance and have shown reduced juvenile numbers since the cold anomaly of 1975-76; a similar pattern was shown by the rock fishes. But because Stephens et al. (1986) sample predominantly young larval stages before their offshore dilution with the coastal larval pool, their results reflect the reproductive condition of the adult population. As many of these species are long lived, the effects will be apparent for many years.
5 NON-KELP BENTHIC SYSTEMS 5.1 ENS0 effects on subtidal soft bottom UODDespite considerable searching, we have found little evidence of ENSO effects on benthic species not associated with kelp habitats. Effects could be either short term--i.e., a reduction in commercial harvest during the El Niiio, or long term--effects on reproduction or recruitment that may take many years to become apparent. Smith (1985) compiied short term data for shelf and slope fisheries from California waters (Table 1); all but one group showed some reduction in harvest in 1983 relative to 1982. While many ground fishes were minimally affected, there was a 75% decline in the Pacific Ocean shrimp (Pandalus-) fishery between 1982 and 1983;
457 landings increased in 1984 and 1985 (Klingbeil, 1986). The Dungeness crab (Cancer magista), which supports a large epibenthic fishery in northern California, undergoes cycles in catch with a 9-10 yr period. Recent analysis (Johnson et al., 1986) suggests that the cycles are related to windTABLE 1 Commercial landings of shelf and slope species from California waters in 1982 and 1983 (from Smith, 1985). 1982 Species Chinook salmon Dover sole Dungeness crab English sole Lingcod Pacific herring Petrale sole Rex sole Rock crab Rockfish Rockfish (b-C*) Rockfish (t**) Sablefish Sea urchin Shrimp (PO***) Totals
Rank 11 6 12 18 19
5 22 25 unranked 3 14 16 7 8 17
Weight
1983 Value Rank
3.3 10.0 3.0 1.4 1.4 10.3 0.8 0.7
18.8 5.1 7.2 1.0 0.7 9.7 1.0 0.5
21.8 2.3 2.0 9.5 8.3 2.0
10.6 1.0 1.0 5.2 3.1 2.2
76.8
67.1
-
25 5 18 15 19 6 23 21 22 4 11 13 8 7 24
Weight ratio
Weight
Value
0.4 8.4 0.9 1.2 0.8 8.0 0.6 0.6 0.6 14.0 3.5 1.7 6.1 7.2 0.5
1.8 4.1 3.1 0.8 0.5 12.5 0.8 0.5 1.1 8.6 1.8 0.8 3.2 3.3 0.9
0.12 0.84 0.30 0.86 0.57 0.78 0.75 0.86
54.5
43.8
0.71
+
0.64 1.52 0.85 0.64 0.87 0.25
Weight is in thousands of metric tons. Value is in millions of dollars to the fishermen. *b-c boccacio and chilipepper rockfish. **t thomyhead rockfish ***PO Pacific Ocean shrimp
driven, southward onshore transport of late stage larvae. The stock cycles appear to be a function of cycles in southward wind stress that are independent of ENSO effects. Pearcy and Schoener (1987) report that razor clams (Siliaua @decreased in abundance along Washington beaches in 1983. Red crabs, pelagic as juveniles and benthic as adults, are normally found off the coast of Baja California. Juveniles were canied north during the El Nifio (Smith, 1985), and we observed adults settling in large numbers in nearshore habitats in southern California. These were a bonanza for varied carnivores (from anemones to fishes to night herons) along the shoreline. In terms of non-commercial species, again, so far as we have found, no monitoring program has produced any reports of important infaunal responses to the ENSO despite the greatly reduced productivity overhead. Longer term effects on reproduction and recruitment may not be apparent for several years, given the variable times to recruitment to fishable stocks and the lag in reporting landing data, yet
458
may be the most important results of El Niiio events. The Pismo clam (Tivela mltorum), a southern species that until recently was the basis of a major recreational fishery in Central California, may represent a soft bottom species strongly affected by ENSO events. Radovitch (1961) noted that survival of Pismo clams was poor at Pismo Beach during the 10 cool years preceding 1957, but that there were sizeable sets during 1957,1958, and 1959, the period of a strong El Niiio. Norton (pers. comm.), using the data of Wendell et al. (1986) from the same location, points out that most of the subsequent sets (through 1981) relate to moderate ENSO events. However, Cce (1956) found virtually no settlement at Pismo Beach during the strong El Niiio of 1940-1941. Summarizing, there were few documented effects of the El Niiio of 1982-1984 on subtidal benthic species outside of kelp forests, but more subtle effects such as changes in growth rates and productivity or long term effects on recruitment and reproduction may not have been studied or are not yet apparent. 5.2 ENSO effects on intertidal pagu lation5 Intertidal communities appear to have been less severely affected than nearby kelp forests by the ENSO event. Barry (1989) observed dramatic storm induced effects on the tube building worm, PhramnatoDoma californica, which responded with a tremendous population explosion that will probably persist for many years. This polychaete has an unusual reproductive response in which severe physical disturbances stimulate the release of larvae (Pawlik, pers. comm.), and Barry observed massive recruitment pulses. As a result, bare patches produced by the storms were colonized by,which has resulted in a new community configuration that will probably persist until another unusually severe storm season. VanBlaricom (in press) found no cormpata)in dense effect attributable to the ENSO event on the numbers of black abalones intertidal populations on San Nicolas Island despite extreme wave events and heavy mortality of kelp populations that provide the drift on which these animals feed. Not surprisingly, mussel growth rates reflected the depressed phytoplankton populations in nearshore waters (e.g., Fiedler, 1984; McCowan, 1985). Searcy et al. (1983) found that the growth rates of Mvtilus edulis and J!J., UlifomianyS in 1983 were 38% and 61% of their growth rates in 1982 at an exposed location glandula at Monterey in northern Baja California. There was a strong recruitment pulse of that Gaines and Roughgarden (1987) attribute to both a reduction in offshore transport of larvae associated with upwelling and to less larval predation by juvenile rockfishes when the larvae returned to shore. The purple sea urchin (Stronwlocentrou purpuraw, an important inter- and subtidal grazer from Baja California to British Columbia, may have latitudinal differences in recruitment success (Ebert and Russell, 1988). While recruitment appears to be frequent in Southern and Baja California, it is highly episodic in the north. Ebert (1983) observed a strong 1963 cohort in Oregon, the only significant recruitment event in 20 years of observation. Paine (1986) summarizes long-term data from the intertidal zone in northwest Washington that include Ebert's 1963 Oregon recruitment followed by a recruitment event in Washington in 1969, and finally a massive recruitment in 1982-83. These years were El Niiio yeais, but only the latter two ENSO
459
events were recognized north of California, and at 480N, the 1982 recruitment event would be very early for and possibly even precede the ENSO event. However, there is reason to believe that the 1982 urchin recruitment may have been linked to the beginning of the 1982-84 ENSO because sea level was anomalously high in Washington (Norton et al., 1985) and planktonic populations had changed at the Farallon Islands (Ainley et al., 1988) by mid-1982. Paine (1986) points out that recruitment events were associated with positive sea surface while all three S. temperature anomalies, there were greater temperature anomalies in non-ENS0 years (1967,1978) with no urchin recruitment. However, Mysak (1986) shows that local warmings and anomalous currents cannot be determined from a catalog of ENSO years, and temperature per se is not as good a predictor of ENSOs as the sea level anomaly. And this does correlate with the S, purp~raubs recruitment. The sand crab, Emerita -, which normally extends into southern Oregon, has been collected off the Washington coast or even further north during or after the El Niiio events of 1940-1941,1957-1958, and 1982-1983 (Schoener and Fluharty, 1985). As such anomalous populations do not persist (Efford, 1970), these observations strongly support the hypothesized larval transport to the north during ENSO events.
The effects of the ENSO on intertidal plant communities were evaluated in both Southern and Central California, and at Tatoosh Island, Washington. Gunnill(l985) had long term data on the populations of seven species of macroalgae at La Jolla when the ENSO occurred, and he found no consistent effect on the algae. However, individuals of all species, especially those on loose rocks, were damaged or killed by the waves. There were further large losses of two laminarians associated with the prolonged warm water period but net recruitment by the other species was high. In the San Diego area, the highest mortality for these intertidal algae normally results from desiccation during Santa Ana winds in the fall (Gunnill, 1980a, b). During the ENSO event, however, the fall season was marked by cloud and fog cover, which, along with the higher sea level, probably minimized desiccation stress, and the plants had high survivorship (Gunnill, 1985). Murray and Horn (1987) began baseline observations in 1978 near Piedras Blancas in Central California; despite much quantitative data and an elaborate analysis, they found significant variation in algal cover in 1983 only in limited instances where crustose species had increased their surface cover. This suggests a release from competition with the overstory algae (sensu Dayton, 1975), but the only overstoty with a significantly decreased cover was Mastocmuis which is not likely to have such an effect. VanBlaricom (pers. comm.) reported one large patch of PostelsiaDalmaeforrms ' that was lost apparently as a result of the El Niiio. Paine (1986) published several sets of long-term data fromthe coast of northern Washington, which he interpreted as showing no convincing ENSO effect; e.g., no effect was identified on the recruitment, mortality, or growth of a brown alga, Dalmaeformis. While there was some response, probably at least in part from the storm related disturbance, the long-term data showed similar anomalies during non-ENS0 years. In all three cases there would appear to be remarkably limited response to such a major disturbance, but all sites exhibit so much natural seasonal and interannual variation that it is difficult to evaluate the effects of the 1982-1983 El Niiio. Summarizing, while there was some response in the intertidal community, especially in recruitment and growth rates, the 1982-1983 ENSO appears to have had little effect on the most
w,
460 commonly measured parameter, adult abundance. This is probably a consequence of several factors. First, the ability of a large number of intertidal algae to perennate, to lose much of the frond biomass annually and then regenerate from the holdfast, would aid in recovery from wave damage. Second, of course, the entire evolutionary history of intertidal species has depended upon adaptations to cope with physical stress during low tide periods, especially desiccation and temperature shifts. Perhaps equally important, the higher sea levels associated with the ENSO must have had the effect of buffering the intertidal species from desiccatory or air temperature extremes. Third, intertidal species are adapted to a wider range of water temperatures than subtidal organisms. The only non-storm related ENSO factor that may have had an effect, then, might have been the presumed lower nutrients in the water, and most of the species seemed buffered against this type of stress. As Paine (1986) concludes, the problem may be that without sufficiently long term data, the effects of an episodic disturbance such as El Niiio cannot easily be separated within such an ecologically complex assemblage as the rocky intertidal community from the normal background of interannual variation. 6 DISCUSSION 6.1 Other EN= 6.1.1 While the 1982-1984 El Niiio exacted biological changes along much of the west coast of America, by far the most significant effects occurred off Ecuador and Peru where temperature anomalies were up to 110C (Arntz, 1986) versus 4 to 5OC off California. Amtz (1985a,b; 1986), and Amtz and Tarazona (this volume) review the massive changes to pelagic and benthic ecosystems. See also Glynn (1988) and Glynn (this volume) for a review of El Niiio effects on eastern Pacific coral reef communities. Benthos development in deeper waters of the Peruvian shelf was affected by changes in oxygen concentrations, temperature increases, and predation by invading tropical carnivores. Demersal fisheries crashed as the animals retreated to cooler conditions at the edge of the continental slope. In the nearshore, there was mass mortality of lntegnfolla and the animals associated with kelps in Peru and northern Chile. In stark contrast with the northeastern Pacific, the almost total mortality of intertidal animals including grazers, on the other hand, led to nearly unlimited growth of algae in this habitat. Grazers began to recruit by the end of 1983 but since they only affect juvenile algae, the once invertebrate-dominated shore may have shifted into a more simple but more resilient algal type community. Subtidal (5- 15 m) mussel banks, commercially a very important resource, were similarly devastated leaving free space that was soon colonized by invading species. Mangroves and their associated fauna were damaged by the high sea level (up to 2 m above normal versus a maximum anomaly of about 30 cm in California, Flick, pers. comm.) with its sediment load, high temperatures, and low salinities caused by continuous rainfall. There was an almost complete shift of the species caught by coastal fishermen as many traditional fisheries crashed but were replaced somewhat by invading tropical species (Amtz, 1985a & b, 1986; Arntz and Tarazona, this volume). The various ENSO induced changes had several types of consequence to different populations. For example, there were mass mortalities of crabs, several species of molluscs, and sea urchins;
46 1 other species such as scallops, octopus, and purple snails experienced large population increases; and several others including shrimps, rock lobsters, swimming crabs, and staked barnacles enlarged distribution patterns (Amtz, 1986). Many of the tropical invaders demonstrated a competitive advantage in space utilization and significant impact as predators, suggesting that ENSO events play a critical role in nearshore community shucture off the coast of South America (reviewed by Glynn, 1988). Of considerable general interest is the degree and persistence of the changes in the shallow sandy beach communities of Peru (Amtz et al., 1987). These communities had been dominated by coexisting populations of the surf clams, Mesodesma hnacium and Donax gemvianu, and the anomuran mole crab Emerita a-. This was an apparently very persistent association, but Mesodesma became locally extinct and had not recolonized the area by July 1986, three years after the retum of normal temperatures. The environment was changed from a highly productive community with easily accessible protein to a completely unproductive resource for local fisheries. Interesting and paradoxically, some of the changes began several months before the warm water occurred in October, 1982 (Amtz et al., 1987); this observation parallels the strong 1982 recruitment of the sea urchin, SgQngvlocentrotyS gumuram, in the northeastem Pacific, which preceded the warm water but was correlated with elevated sea level. While some shallow water communities off Peru and Chile show a high resilience with a fast recovery after each ENSO event, Amtz et al. (1987) and Arntz and Tarazona (this volume) discuss other apparently long-term changes in communities exacted by this and earlier El Nifio events. It is certainly clear that the South American ENSO devastated the benthic community whereas the effects in the North Pacific seemed restricted to kelps, and even in those cases seem relatively transitory compared to the South American experience. 6.1.2 There also is evidence that ENSO events affect the northwestern Pacific. Kawabe (1985) documents interannual variability in the Kuroshio and Oyashio Currents near Japan. The Oyashio, which originates in the Arctic, extends further to the south and water temperatures along the northeastern coast of Honshu reach minima during most El Niiio events. Sakai (1962) has shown that abalone production in this area correlates with the harvest of Undaria pinnatifh, the most preferred food of abalones, and that fluctuations in production of Undaria and other algae are closely related to variations of the Oyashio Current. 6.1.3 South Africa Like the normal communities off Peru-Ecuador and the west coast of North America, the South African and Namibian faunas are dependent on upwelling for nutrient input (Walker et al., 1984). Glynn (1988) reviews evidence for an El Niiio-type event off the SW African coast with a significant temporal (concurrent or slightly preceding) relationship to ENSO events in the Pacific Ocean. Although less intense and less frequent, the Benguela Niiio also has been implicated with declines in anchovies and sardines and seabird reproductive failures (see Duffy, this volume). Birkett and Cook (1987) show that the 1982-1983 temperature anomaly caused irregular spawning of bivalve (Donax serra) populations off South Africa. Anomalously warm temperatures led to premature spawning and a drastic recruitment failure in 1983. Walker et al. (1984) conclude that
462 the 1982-1983 Benguela warm event resulted from abnormalities in the summer wind regime, not advection from equatorial regions, and suggest that the interoceanic link may be provided by the westerly wind belt in the southern hemisphere. Interestingly, 1984 was a year of high sea level and warm temperatures along the entire eastern South Atlantic Ocean that Brundit et al. (1987) regard as evidence for poleward propagation and Glynn (1988, Fig. 1) illustrates as an intrusion of warm Angolan water. 6.2 1 Dvrifera forests are found from mid Baja California to a location in Central California between Santa Cruz and San Francisco, a range of more than 100 latitude and considerable variation in average temperature (Foster and Schiel, 1985). Because of the warmer temperatures, presumably associated with lowered nutrients, stands of Macrocvst is along the
coasts of Baja and Southern California develop into large forests at sites where upwelling of cool, nutrient-rich water is regular (e.g., Dawson, 1951). Such forests occur west of Santa Barbara, off Palos Verdes, and La Jolla-Point Loma. (In contrast, the northwestern Channel Islands, consistently the coolest area of the Southern California Bight [Fig. 91, support large forests because of the strong influence of the California Current.) The combination of the elimination of the normal sources of summer nutrient input to these forests, upwelling and thermocline motion, with the apparent transport of warm, nutrient-poor water into the area, pushed the southeasternmost coastal habitats out of the range of suitability for Macmvs&. Here a benthic community was critically impacted by abiotic aspects of the pelagic environment. Further to the north, given the decrease in temperature with latitude, the addition of the El Niiio temperature anomaly to normal summer-fall temperatures apparently maintained the environment within the range of ’ suitability (i.e., nutrients did not become limiting) for and may thus have only acted to increase growth rates. Conversely in the normally warmest, most nutrient-depleted parts of the Macrocvstis habitat, such as Santa Catalina Island in Southern California and much of Baja California, the effects of the warm summer-fall temperatures were most severe (Zimmerman and Robertson, 1985; Hernandez, in press; Estes and Harrold, 1988). While southern Macrocvs forests were clearly impacted by the ENSO, there is little evidence that other benthic North American habitats were much disturbed, especially in comparison with events off Ecuador, Peru and northern Chile. In part this reflects the magnitude of the physical anomalies, which were more severe off South America (e.g., Amtz, 1986; Hansen, this volume; Barber and Kogelschatz, this volume). While presumably all communities were affected by the major decline in nutrients and primary productivity (e.g., McGowan, 1985), it may have been apparent only in Macrocv- ‘ whose scales of size and growth rate are so much larger than those of ’ in the northern part of its range, ENSO environmental other organisms. Similar to perturbations may have been within the range of suitability for most intertidal organisms whose habitat is naturally much more variable than nearby kelp forests. Like Macrocv. stis confined to upwelling locations, some pelagic populations were behaviorally not free to move to better conditions and so were strongly affected by the El Niiio. These included rockfishes that tend to have home ranges or be territorial as adults (Lenarz and Echevema, 1986;
463 Ven Tresca, pers. comm.), salmon tied to spawning streams (Pearcy and Schoener, 1987; Johnson, 1988) and seabird reproduction linked to nesting sites (reviewed by Pearcy and Schoener, 1987; Ainley et al., 1988). Recruitment, one of the most important processes mediated by the pelagic environment, may be losing some of its unexplained variability to a better understanding of ENS0 events and ocean climate generally. Cowen's (1985) and Pringle's (1986) work clearly demonstrated the importance of El Niiio transport to recruitment of species at the northern end of their ranges in southern California, especially for two species important to the structure of kelp forests in this region, spiny lobsters and sheephead. Gaines (1987), Gaines and Roughgarden (1987) and Roughgarden (pers. comm.) relate increased barnacle settlement at a Central California site in part to a reduction in offshore transport associated with the reduction in upwelling in 1983. Recruitment of pelagic fish stocks appears to be similarly affected by variations in transport. Sinclair et al. (1985) found essentially a one-to-one relationship between the interannual variability in the survival of Pacific mackerel to age one and sea level between 1928 and 1965, and associated the good year classes with El Niiio-related reductions in southward transport. In this case there was a striking inverse relationship between plankton biomass and early survivorship suggesting a more prominent role for hydrographic processes than for biological interactions. Norton (1987) studied recruitment of two rockfishes for the years 1965-1980 and found their most successful recruitment years to occur under different environmental conditions. Widow rockfish recruitment pulses were associated with warm years, strong Aleutian lows and El Niiio events, increased northward transport, and downwelling; chilipepper rockfish did well in cool years with weak Aleutian lows, strong southward transport, and reduced downwelling. There was no apparent relationship to zooplankton abundance in either case. The common thread of these examples is that interannual anomalies in transport processes cause fluctuations in recruitment success as predicted by Panish et al. (1981). The kelp forests of Southern California, despite being subjected to several major disturbances within a short time, recovered very rapidly relative to other communities. The same can be said of the kelp forests of Baja California except in the far southerly end of their range where, according to Hernandez (in press), they may have been replaced by m.Glynn (1988) reports El Niiio-induced mortality of reef building corals ranging from 50 to 98% in the eastern equatorial Pacific region, including the possible regional extinction of two species of hydrocorals. The death of the coral framework, which provides most of the structure in these communities, led to the elimination of refuges for numerous reef-associated species, changes in predator distributions, and major bioerosion. Recovery of these communities may take decades. In contrast, the heteromorphic life history of kelps provides considerable resilience to disturbance. The life cycle of all kelps consists of an alteration of generations between a microscopic haploid gametophyte and a large diploid sporophyte. The gametophytes appear to be able to survive for months or longer providing a stabilizing storage tactic similar to that of seed banks in some higher plants (Dayton, 1985). Thus once temperatures returned to normal, the kelp community structure recovered rapidly in most areas, despite the delay caused by the amphipod infestation (Tegner and Dayton, 1987). Repeated or extraordinarily severe disturbances may eliminate the seed bank; e.g.,
464
Macrocvstis remains locally extinct in some habitats in Southern California (Tegner, pers. obs.) and central Peru (Amtz, 1986) well after the ENSO. Assuming that competitors do not usurp the plants. space, recolonization of these habitats will depend on drifting reproductive In summary, there were many qualitative similarities between ENSO effects off the coasts of Ecuador and Peru and off the west coast of North America, as well as substantial differences that appear to reflect the relative magnitude of environmental changes. There were species displacements, changes in relative abundance, and changes in rates of growth, survival, and recruitment. Habitats were altered and shifted by water movement, and those species that could not move suffered, or in some cases benefited depending on the properties of the new environment (Arntz, 1986; Pearcy and Schoener, 1987; Sharp, 1987; Arntz and Tarazona, this volume). In both cases there were north-south gradients of severity with greater survival in higher latitudes. Arntz (1986) attributes some of the resilience of the South American system to rapid transport of eggs, larvae, and adults from areas unaffected by the ENSO via the Humbolt Current; the California Current will act in the same way. Furthermore, while the time period between very strong ENSOs such as 1957-1959 and 1982-1984 is long compared to the life span of most organisms, El Niiio events are a regular feature of both environments and must be an important evolutionary factor.
7 CONCLUSIONS The 1982-84 El Niiio had a massive effect on planktonic communities from the equator poleward to surprisingly high latitudes in the eastern Pacific (reviewed in Wooster and Fluharty, 1985; Arntz, 1986, Glynn, 1988). In the northeastern Pacific, planktonic effects were variable with regard to the particular species, and the latitude and duration of various ENSO effects. Despite the massive effect on the pelagic community, the benthic populations of the Northern Hemisphere temperate zone were but marginally affected by the ENSO phenomenon. In both habitats there were diametrically opposed results in which some species increased and others decreased. And of course, for most species there are so few time series data that the trends are statistically ambiguous. Certainly all the communities are complex and characterized by inadequate time series data such that it is always risky to discuss episodes such as the El Niiio in relation to their normal interannual variations. Perhaps the classic work of Hjort (1914) represents a means of summarizing some of the apparently contradictory observations in recruitment patterns. Hjort suggested that two basic processes influence recruitment: 1) the availability of appropriate food for the larvae and 2) drift of the larvae away from the proper recruitment habitat. The former is the classical explanation for declines in most plankton populations during ENSOs (McGowan, 1985), but there are also many cases of local populations being enhanced, apparently by the northerly transport (Sette and Isaacs, 1960; Sinclair et al., 1984; Cowen, 1985; Pringle, 1986; Pearcy and Schoener, 1987). These two processes are sufficiently general that they may in principle explain how the two communities responded differently. For example the planktonic populations that people have studied tend to be composed of small individuals with minimal food reserve. While Sinclair et al. (1984) considered larger fish, most of the populations that collapsed in the plankton (McGowan, 1985) are composed
465 of small individuals that need nutrients on a regular basis. Be they larvae or simply small adults on the one hand, Hjort's first hypothesis is probably applicable to many plankters. On the other hand, many of the important benthic populations that people have studied are composed of larger organisms that might be expected to have stored reserves sufficient to allow the individuals to survive the temporary effects of low nutrients and reduced primary production, which in any case is probably several steps removed from many of the benthic individuals. With the exception of Gallardo (1985),apparently no one has looked for ENSO effects on the very small benthic organisms per se, but for the most part, these consume detrital food and would not be expected to respond very quickly to the assumed reduction in production. Where non-kelp benthic effects have been seen, they tend to be consistent with Hjort's second hypothesis regarding a change in larval availability in the system. Again, note that this discussion does not relate to the South American situation where many benthic organisms were temporarily replaced by tropical species advected into the area. There are probably other reasons for the different responses between the benthos and the plankton. There are many fundamental differences between the two systems (Dayton, 1984; Branch et al., 1987). Certainly variability in time and space has always been the dominant theme in coastal plankton communities; the individuals are motile and characterized by high turnover rates with relatively short lifespans and shortlived developmental stages. While motile, the individuals drift with the water masses and except for large nekton, are locked into the physical phenomena driving the system. The benthic community seems characterized by relatively longer term stability and the structure of the communities is often more influenced by biological relationships; the relative longevity of the organisms, their interactions with each other and, importantly, their frequent ability to modify their physical environments thus homogenizes some of the seasonally induced shifts and contributes to a resistance to the major disturbances such as those associated with the ENSO phenomenon.
8 ACKNOWLEDGEMENTS We thank R. Butler, T. Klinger, C. Lennert, P. Pamell, K. Plummer, M. Pruett-Jones, and L. Walcheff for their many hours underwater and in the laboratory. For allowing us to use their unpublished and published data, we are very grateful to D. Ainley, J. Barry, J. Estes, R. Flick, M. Foster, P. Glynn, A. MacCall, R. McPeak, J. Pawlik, J. Roughgarden, J. Simpson, B. Stewart, G. VanBlaricom, and D. Ven Tresca. We appreciate manuscript reviews by C. Barilotti, J. Barry, R. Cowen, C. D'Elia, P. Fairweather, V. Gerard, G. Jackson, L. Deysher, A. Ebeling, P. Glynn, J. McGowan, W. North, R. McPeak, G. VanBlaricom and R. Zimmerman. We thank J. Pineda and D. Rice for translating the Hernandez manuscript. We are very grateful to L.S. Jackson for typing too many drafts and E. Crissman for proofreading. This work is the result of research sponsored by the National Science Foundation, the University of California San Diego Academic Senate, and the National Oceanic and Atmospheric Administration, National Sea Grant College Program, Department of Commerce, under grant number NA80AA-D-00120, project number R/NP-l-l2F, through the California Sea Grant Program. The U.S. Government is authorized to reproduce and distribute this document for governmental purposes.
466 9 REFERENCES Ainley, D.G. and Boekelheide, R.J., In press. Farallon Island Seabirds: Ecology, Dynamics, and Structure in an Upwelling-Based Community. Stanford University Press, Stanford, Calif. Ainley, D.G., Carter, H.R., Anderson, D.W., Briggs, K.T., Coulter, M.C., Cruz, F., Cruz, J.B., Valle, C.A., Fefer, S.I., Hatch, S.A., Schreiber, E.A., Schreiber, R.W. and Smith, N.G., 1988. ENS0 effects on Pacific Ocean marine bird populations. In: H. Ouellet (Editor), Proceedings of the XIX International Ornithological Congress. National Museum of Natural Sciences, Ottawa, Canada, pp. 1,747-1,758. Arntz, W.E., 1985a. El Niiio and Peru: positive aspects. Oceanus, 27: 36-39. Arntz, W.E., 1985b. Zur Entstehung von Organisrnenansammlungen: "El Niiio" 1982- 1983 vor Peru. Natur und Museum, 115(5): 134-151. Arntz, W.E., 1986. The two faces of El Niiio 1982-1983. Meeresforsch., 31: 1-46. Arntz, W.E., Brey, T., Tarazona, J., and Robles, A., 1987. Changes in the structure of a sandybeach community in Peru during an El Niiio event. S. Afr. J. Mar. Sci., 5: 645-658. Baltz, D.W., 1984. Life history variation among female surf perches (Perciformes: Embiotocidae). Environ. Biol. Fishes, 10: 159-171. Barry, J.P., 1989. Reproductive responses of a marine annelid to winter storms: an analogue to fire adaptation in plants? Mar. Ecol. Prog. Ser. (in press). Bernal, P.A., 1981. A review of the low frequency response of a pelagic ecosystem in the California Current. Cal. Coop. Ocean. Fish. Inv. Rep., 22: 49-64. Birkett, D.A. and Cook, P., 1987. Effect of the Benguela temperature anomaly, 1982-1983,on the breeding cycle o f m m R o d i n g . S . Afr. J. Mar. Sci., 5: 191-196. Bodkin, J.L., 1986. Fish assemblages in Macrocvstisand' kelp forests off Central California. Fish. Bull. U.S.,84: 799-808. Bodkin, J.L., VanBlaricom, G.R.,and Jameson, R.J., 1987. Mortalities of kelp-forest fishes associated with large oceanic waves off Central California, 1982-1987.Environ. Biol. Fishes, 18: 73-76. Branch, G.M., Barkai, A., Hockey, P.A.R., and Hutchings, L., 1987. Biological interactions: causes or effects of variability in the Benguela ecosystem. S. Afr. J. Mar. Sci., 5: 425-445. Brundit, G.B., Cuevas, B.A., and Shipley, A.M., 1987. Long term sea level variability in the eastern South Atlantic and a comparison with that in the Eastern Pacific. S. Afr. J. Mar. Sci., 5 : 73-78. Cameron, R.A. and Schroeter, S.C., 1980. Sea urchin recruitment: effect of substrate selection on juvenile distribution. Mar. Ecol. Prog. Ser., 2: 243-247. Cannon, G.A., Reed, R.K., and Pullen, P.E., 1985. Comparison of El Niiio events off the Pacific Northwest. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 75-84. Cayan, D.R. and Flick, R.E., 1985. Extreme sea levels in San Diego, California winter 19821983. Scripps Inst. Oceanogr. Ref. No. 85-3: 1-58. Chelton, D.B., Bernal, P.A. and McGowan, J.A., 1982. Large-scale interannual physical and biological interactions in the California Current. J. Mar. Res., 40: 1,0951,125. Coe, W.R., 1956. Fluctuations in populations of littoral marine invertebrates. J. Mar. Res., 15: 212-232. Cowen, R.K., 1983. The effect of sheephead (SemicossvDhus predation on red sea urchin m v l o c e n t r o t u ffZl0ciscanu) populations: an experimental analysis. Oecologia, 58: 249-255. Cowen, R.K., 1985. Large scale patterns of recruitment by the labrid, Semicossvohus Dulcher: causes and implications. J. Mar. Res., 43: 719-742. Cowen, R.K. and Bodkin, J.L., In press. Annual and spatial variation of the kelp forest fish assemblage at San Nicolas Island, California In: Proc. Third Channel Islands Symposium, Santa Barbara, Calif., 2-6 March 1987. Cowen, R.K., Bowman, M.J. and Fouke, S.S., 1988. Time scales, space scales and fish scales: viscous vortices as voluted vivaria? Abstract 84, So. Cal. Acad. Sci. Ann. Meet., May 6-7, 1988. Coyer, J.A., 1979. The invertebrate assemblage associated with Macrocvst is pyrifera and its utilization as a food resource by kelp forest fishes. Ph.D. Dissertation. 364 pp. Univ. So. Cal., Los Angeles.
e)
467 Davis, G.E., 1985. Kelp forest dynamics in Channel Islands National Park, California, 19821985. Channel Islands Nat. Park, Nat. Mar. Sanctuary, Nat. Sci. Stud. Rep. CH15-86-001: 1-11. Dawson, E.Y., 1951. A further study of upwelling and associated vegetation along Pacific Baja California, Mexico. J. Mar. Res., 10: 39-58. Dayton, P.K., 1975. Experimental evaluation of ecological dominance in a rocky intertidal algal community. Ecol. Monogr., 45: 137-159. Dayton, P.K., 1984. Processes structuring some marine communities: are they general? In: D.R. Strong, D. Simberloff, L.G. Abele, and A.B. Thistle (Editors), Ecological Communities: Conceptual Issues and the Evidence. Princeton University Press, Princeton, New Jersey, pp. 181- 197. Dayton, P.K., 1985. Ecology of kelp communities. Annu. Rev. Ecol. Syst., 1 6 215-245. Dayton, P.K. and Tegner, M.J., 1984a. Catastrophic storms, El Niiio, and patch stability in a Southern California kelp community. Science, 224: 283-285. Dayton, P.K. and Tegner, M.J., 1984b. The importance of scale in community ecology: a kelp forest example with terrestrial analogs. In: P.W. Price, C.N. Slobodchikoff and W.S. Gaud (Editors), Novel Approaches to Interactive Systems. J. Wiley and Sons, New York, pp. 457481. Dayton, P.K., Cume, V., Gerrodette, T., Keller, B., Rosenthal, R., and Ven Tresca, D., 1984. Patch dynamics and stability of some Southern California kelp communities. Ecol. Monogr., 54: 253-289. Dean, T.A. and Jacobsen, F.R., 1986. Nutrient limited growth of juvenile kelp, &facrocvstiS pyrifera, during the 1982-1984 "El Niiio" in Southern California. Mar.Biol., 90: 597-601. Dean, T.A., Schroeter, S.C., and Dixon, J., 1984. Effects of grazing by two species of sea urchins @trongvlocentrotys franciscanusand LvtechinySmamesus) on recruitment and survival of two species of kelp (Macrocvstis pyifem and m g o phora californica). Mar. Biol., 78: 301-313. Deysher, L.E. and Dean, T.A., 1986. Interactive effects of light and temperature on sporophyte production in the giant kelp Macrocvs pvrifeq. Mar.Biol., 93: 17-20. Ebeling, A.W. and Laur, D.R., 1988. Fish populations in kelp forests without sea otters: effects of severe storm damage and destructive sea urchin grazing. In: G.R. VanBlaricom and J.A. Estes (Editors), The Community Ecology of Sea Otters. Springer-Verlag, Berlin, pp. 169-191. Ebeling, A.W., Larson, R.J. and Alevizon, W.S., 1980. Habitat groups and island-mainland distribution of kelp-bed fishes off Santa Barbara, California. In: D.M. Power (Editor), Multidisciplinary Symposium on the California Islands. Santa Barbara Mus. Nat. Hist., Santa Barbara, Calif., pp. 403-431. Ebeling, A.W., Laur, D.R. and Rowley, R.J., 1985. Severe storm disturbances and reversal of community structure in a Southern California kelp forest. Mar. Biol., 84: 287-294. Ebert, T.A., 1983. Recruitment in echinoderms. In: M. Jangoux and J.M. Lawrence (Editors), Echinoderm Studies. A.A. Balkema, Rotterdam, pp. 169-203. Ebert, T.A. and Russell, M.P., 1988. Latitudinal variation in size structure of the west coast purple sea urchin: a correlation with headlands. Limnol. Oceanogr., 33: 286-294. Efford, LA., 1970. Recruitment to sedentary marine populations as exemplified by the sand crab, Emerita analoea (Decapoda, Hippidae). Crustaceana, 18: 293-308. Elliott, J.L. and Foster, M.S., 1987. Comparison of change in population structure and abundance of large benthic algae and invertebrates at two Southern California kelp forests. In: Southern California Edison Company, Report on 1986 Data, Marine Environmental Analysis and Interpretation. San Onofre Generating Station, chapter 4, 34 pp. Engle, J.M., 1979. Ecology and growth of juvenile California spiny lobster, Panulirus interruptus (Randall). Ph.D. Dissertation. 298 pp. Univ. So. Cal., Los Angeles. Estes, J.A. and Harrold, C., 1988. Sea otters, sea urchins, and kelp beds: some questions of scale. In: G.R. VanBlaricom and J.A. Estes (Editors). The Community Ecology of Sea Otters. Springer-Verlag, Berlin, pp. 116-150. Fiedler, P.C., 1984. Satellite observations of the 1982-83 El Niiio along the U.S. Pacific coast. Science. 224: 1.25 1- 1,254. Fiedler, P.C., Methot, R.D. and Hewitt, R.P., 1986. Effects of the California El Niiio on the northern anchovy. J. Mar. Res., 44: 317-338. Foster, M.S. and Schiel, D.R., 1985. The ecology of giant kelp forests in California: a community profile. U.S. Fish Wildl. Serv. Biol. Rep., 85(7.2): 1-152. 1-
468 Fulton, J.D. and LeBrasseur, R.J., 1985. Interannual shifting of the subarctic boundary and some of the biotic effects on juvenile salmonids. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 237-252. Gaines, S.D., 1987. Effects of upwelling in the California current system on recruitment to intertidal barnacle population. Abstr. West. Soc. Nat. 68th Annu. Meet., Long Beach, Calif., pp. 19-20. Gaines, S.D. and Roughgarden, J., 1987. Fish in offshore kelp forests affect recruitment to intertidal barnacle populations. Science, 235: 479-481. Gallardo, V.A., 1985. Efectos del fenomeno de "El Niiio" sobre el bentos sublitoral frente a Concepcion, Chile. In: W. Arntz, A. Landa and J. Tarazona (Editors), Instituto del Mar del, Peru Bol. (special issue), Callao, Peru, pp. 79-85. Gerard, V.A., 1976. Some aspects of material dynamics and energy flow in a kelp forest in Monterey Bay, California. Ph.D. Dissertation. 172 pp. Univ. California , Santa Cruz. Gerard, V.A., 1982. Growth and utilization of internal nitrogen reserves by the giant kelp -Dvrifera in a low-nitrogen environment. Mar. Biol., 66: 27-35. Gerard, V.A., 1984. Physiological effects of El Niiio on giant kelp in Southern California. Mar. Biol. Lett., 5: 317-322. Gerard, V.A., 1986. Photosynthetic characteristics of giant kelp (Macrocv .stis Dvrifera) determined in situ. Mar. Biol., 90: 473-482. Gerard, V.A. and Du Bois, K.R., 1988. Temperature ecotypes near the southern boundary of the kelp Laminaria scchan'na. Mar. Biol., 97: 575-580. Gemann, I., Druehl, L.D. and Hoeger, U., 1987. Seasonal variation in total and soluble tissue nitrogen of gardnei (Phaeophyceae:Laminariales) in relation to environmental nitrate. Mar. Biol., 96: 413-423. Glynn, P.W., 1988. El Niiio-Southern Oscillation 1982-83: nearshore population, community and ecosystem responses. Ann Rev. Ecol. Syst., 19: 309-345. Graybill, M.R. and Hodder, J., 1985. Effects of the 1982-1983 El Niiio on reproduction of six species of seabirds in Oregon. In: W.S. Wooster and D.L. eluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 205-210. Gunnill, F.C., 1980a. Demography of the intertidal brown alga Pelvetia fastigim in Southern California USA. Mar. Biol., 59: 169-179. Gunnill, F.C., 1980b. Recruitment and standing stocks in populations of one green alga and five brown algae in the intertidal zone near La Jolla, California during 1973-1977.Mar. Ecol. Prog. Ser., 3: 231-243. Gunnill, F.C., 1985. Population fluctuationsof seven macroalgae in Southern California during 1981-1983including effects of severe storms and El Niiio. J. Exp. Mar. Biol. Ecol., 85: 149164. Hams, L.G., Ebeling, A.W., Laur, D.R. and Rowley, R.J., 1984. Community recovery after storm damage: a case of facilitation in primary succession. Science, 224: 1,336-1,338. Harrold, C. and Reed, D.C., 1985. Food availability, sea urchin grazing, and kelp forest community structure. Ecology, 66: 1,160-1,169. Hatch, S.A., 1987. Did the 1982-1983El Niiio-Southern Oscillation affect seabirds in Alaska? Wilson Bull., 99: 468-474. Hernandez, C.G., In press. Recuperacion de 10s mantos de sargaso gigantes @lacrocvstiS pyriferd) en Baja California, despues del fenomeno de "El Niiio" de 1982-83. "Simp. Invest. Biol. Oceanogr. Pesq. Mexico, 28-30 de Abril de 1987" CICMAR. Hjort, J., 1914. Fluctuations in the great fisheries of northern Europe viewed in the light of biological research. Rapp. P.-V. Reun. Cons. Int. Explor. Mer, 20: 1-228. Hobson, E.S. and Chess, J.R., 1976. Trophic interactions among fishes and zooplankters near shore at Santa Catalina Island, California Fish. Bull. U.S., 74: 567-598. Hubbs, C.L., 1948. Changes in the fish fauna of Western North America correlated with changes in Ocean temperature. J. Mar. Res., 7: 459-482. Hubbs, C.L. and Hubbs, L.C., 1954. Data on the life history, variation, and ecology of the kelp surf perch, B r a c h m frenatus,an embiotocid fish of the Californias. Calif. Fish Game, 40: 183-198. Huyer, A. and S'mith, R.L., 1985. The signature of El Niiio off Oregon, 1982-1983. J. Geophys. Res., 90: 7,133-7,142. IMR., 1963. An investigation of the effects of discharged wastes on kelp. Final Report. Univ. Cal. Inst. Mar. Resour. IMR Ref. 63-6, 305 p.
469 Jackson, G.A., 1977. Numents and production of the giant kelp, Macrocym ' pyrifera, off Southern California. Limnol. Oceanogr., 22: 979-995. Johnson, D.F., Botsford, L.W., Methot, R.D. and Wainwright, T.C., 1986. Wind stress and cycles in Dungeness crab (Cancer maeister') catch off California, Oregon, and Washington. Can. J. Fish. Aquat. Sci., 43: 838-845. Johnson, S.L., 1988. The effects of the 1983 El Niiio on Oregon's coho (Qncorhynchus kisutch) and chinook (QBhawvtsch3 salmon. Fish. Res., 6: 105-123. Jones, L.G., 1965. Canopy grazing at southern Point Loma. In: Kelp Habitat Improvement Project Annual Report 1 Feb. 1964 - 31 Mar. 1965. W.M. Keck Lab. Environ. Health Eng. Cal. Inst. Tech., Pasadena, pp. 62-63. Karinen, J.F., Wing, B.L. and Straty, R.R., 1985. Records and sightings of fish and invertebrates in the eastern Gulf of Alaska and oceanic phenomena related to the 1983 El Nifio event. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 253267. Kawabe, M., 1985. El Niiio effects in the Kuroshio and western North Pacific. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 31-43. Kerr, R.A., 1988. La Niiia's big chill replaces El Niiio. Science, 241: 1,037-1,039. Kimura, S.,1985. Differences in long-term patterns of algal change in Diablo Cove, Central California subtidal habitats. Abstr. West. Soc. of Nat. 66th Annu. Meet., Monterey, Calif., pp. 45-46. Klingbeil, R. (comp.), 1984. Review of some California fisheries for 1983. Calif. Coop. Oceanic Fish. Invest. Rep., 25: 7-15. Klingbeil, R. (comp.), 1986. Review of some California fisheries for 1985. Calif. Coop. Oceanic Fish. Invest. Rep., 27: 7-15. Kuhn, G.G.and Shepherd, F.P., 1984. Sea Cliffs, Beaches, and Coastal Valleys of San Diego County: Some Amazing Stories and Homfying Implications. Univ. Cal. Press, Berkeley, 193 PP. Lawrence, J.M., 1975. On the relationships between marine plants and sea urchins. Oceanogr. Mar. Biol. Annu. Rev., 13: 213-286. Lea, R.N. and Vojkovich, M., 1985. Record of the Pacific burrfish from southern California. Bull. South. Calif. Acad. Sci., 84: 46-47. Lea, R.N., Swift, C.C. and Lavenberg, R.L., 1988. Records of Wil curema Valenciennes, the white mullet, from southern California Bull. South. Calif. Acad. Sci., 87: 31-34. Leighton, D.L., 1971. Grazing activities of benthic invertebrates in kelp beds. Nova Hedwigia, 32: 421-453. Leighton, D.L., Byhower, M.J., Kelly, J.C., Hooker, G.N. and Morse D.E., 1981. Acceleration of development and growth in young green abalone @Ialio& using warmed effluent seawater. J. World Maricult. Soc., 12: 170-180. Lenarz, W.H. and Echeverria, T.W.,1986. Comparison of visceral fat and gonadal fat volumes of yellowtail rockfish, Sebastes flavidus, during a normal year and a year of El Niiio conditions. Fish. Bull. U.S., 84: 743-745. Mann, K.H., 1982. Ecology of Coastal Waters, A Systems Approach. Univ. Calif. Press, Berkeley, Chap. 3, pp. 53-82. MacCall, A.D., 1986. Changes in the biomass of the California Current ecosystem. In: K. Sherman and L.M. Alexander (Editors), Variability and Management of Large Marine Ecosystems. Am. Assoc. Adv. Sci., Wash., D.C., pp. 33-54. McGowan, J.A., 1985. El Niiio 1983 in the Southern Calif. Bight. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 166-184. Mearns, A.J., 1988. The "odd fish": unusual Occurrences of marine life as indicators of changing Ocean conditions. Chapter 7, In: D.F. Soule and G.S. Kleppel (Editors), "Marine Organisms as Indicators". Springer-Verlag, New York, 342 p. Miller, C.B., Batchelder, H.P., Brodeur, R.D. and Pearcy, W.G., 1985. Response of the zooplankton and ichthyoplankton off Oregon to the El Niiio event of 1983. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 185-187. Murray, S.N. and Horn, M.H., 1987. Variations in standing stocks of intertidal macrophytes at a Central California site before and after the 1982-83 El Niiio. Abstr. West. Soc. Nat. 68th Annu. Meet., Long Beach, Calif., pp. 34-35.
w)
470 Mysak, L.A., 1986.El Niiio, interannual variability and fisheries in the northeast Pacific Ocean. Can. J. Fish. Aquat. Sci., 43:464-497. Namias, J. and Cayan, D.R., 1984.El Niiio: implications for forecasting. Oceanus, 27:41-47. Niebauer, H.J., 1985.Southern Oscillation/El Niiio effects in the Eastern Bering Sea. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 116-120. North, W.J. (ed.)., 1971.The Biology of Giant Kelp Beds @lacrocv&U' ) in California. Nova Hedwigia, 32: 1-600. North, W.J. and Pearse, J.S., 1970.Sea urchin population explosion in Southern California waters. Science, 167:209. North, W.J. and Zimmerman, R.C., 1984.Influences of macronutrients and water temperatures ' canopies. Hydrobiologia, 116/117:419-424. on summertime survival of North, W.J., Gerard, V. and Kuwabara, J., 1982.Farming Macrocvstis at coastal and oceanic sites. In: L.M. Srivastava (Editor), Synthetic and Degradative Processes in Marine Macrophytes. Walter de Gruyter, Berlin, pp. 247-262. Norton, J., 1987.Ocean climate influences groundfish recruitment in the California Current. Proc. Int. Rockfish Symp., Anchorage, Alaska. Univ. Alaska Sea Grant Rep. No. 87-2,pp. 73-98. Norton, J., McLain, D., Brainhard, R. and Husby, D., 1985.The 1982-83El Niiio event off Baja and Alta California and its ocean climate context. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 44-72. Ono, K.A., 1987.The effects of El Niiio on mother-pup biology in the California sea lion. Abstr. Third Channel Islands Symp., 2-6March 1987,Santa Barbara, Calif. Paine, R.T., 1986.Benthic community-water column coupling during the 1982-1983El Niiio. Are community changes at high latitudes attributable to cause or coincidence? Limnol. Oceanogr.,
31: 351-360. Panish, R., Nelson, C. and Bakun, A., 1981.Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceanog., l(2): 175-203. Patton, M.L., 1985.Changes in fish abundance in the Southern California bight during a warm water episode. In: Southern California Edison Company, Report on 1984 Data, Mar. Environ. Anal. Interpretation, San Onofre Generating Station, Chap. 8,34pp. Pearcy, W.G. and Schoener, A., 1987.Changes in the marine biota coincident with the 19821983 El Niiio in the northeastern subarctic Pacific Ocean. J. Geophys. Res., 92: 14,417-
14,428. Pearcy, W.G., Fisher, J., Brodeur R. and Johnson, S . , 1985. Effects of the 1983 El Niiio on coastal nekton off Oregon and Washington. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 188-204. Pringle, J.D., 1986.California spiny lobster (Panulirus interrupt& larval retention and recruitment: a review and synthesis. Can. J. Fish. Aquat. Sci., 43:2,142-2.152. Putt, M. and Prezelin, B.B., 1985.Observations of die1 patterns of photosynthesis in cyanobacteria and nanoplankton in the Santa Barbara Channel during "El Niiio". J. Plankton Res., 7:779-790. Quiroz, R.S., 1983.The climate of the "El Niiio" winter of 1982-83--a season of extraordinary climatic anomalies. Mon. Wea. Rev., 1 1 1: 1,685-1,706. Radovitch, J., 1961.Relationships of some marine organisms of the northeast Pacific to water temperatures particularly during 1957 through 1959.Calif. Dept. Fish and Game Fish Bull.,
112: 1-62. Reid, F.M.H., Lange, C.B. and White, M.M., 1985.Microplankton species assemblages at the Scnpps Pier from March to November 1983 during the 1982-1984El Niiio event. Bot. Mar.,
28:443-452. b n n a i Ino. III. Study on the Sakai, S., 1962.Ecological studies on the abalone mechanism of production of the abalone in the region of Onagawa Bay. Bull. Jap. Soc.Sci. Fish., 28: 891-898(English abstract). Schoener, A. and Fluharty, D.L., 1985.Biological anomalies off Washington in 1982-1983and other major Niiio periods. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 211-225. Schroeter, S.C., Dixon, J. and Kastendiek, J., 1983.Effects of the starfish w m i n i a u on the distribution of the sea urchin Lvtechinu in a Southern California kelp forest. Oecologia (Berlin), 56: 141-147.
47 1
Searcy-Bernal, R., Garcia-Pamanes, L., Salas-Garza, A., Garcia-Pamanes, F., LizarragaVizcarra, S. and Cancino-Franklin, L., 1983. Comparacion del crecimiento de l% ddk y h!L califmianus cultivados en una costa expuesta y una semiprotegida de Baja California. In: Informe Anual 1983 del Project0 "Estudios para el Desarrollo del Cultivo Comercial de Mejillones en Baja California". Inst. Invest. Oceanol. Univ. Aut6noma de Baja California, Ensenada, B.C. Mexico. Sette, O.E. and Isaacs, J.D. (eds.), 1960. Symposium on "The Changing Pacific Ocean in 1957 and 1958". Calif. Coop. Oceanic Fish. Invest. Rep., 7: 13-217. Seymour, R.J. and Sonu, C.J., 1985. Storm waves and shoreline damage on the California coast, winter of 1983. Japanese J. Coastal Engineering, 70: 59-64. Seymour, R.J., Strange, R.R. In, Cayan, D.R. and Nathan, R.A., 1984. Influence of El Niiios on California's wave climate. In: B.L. Edge (Editor), Nineteenth Coastal Engineering Conference, Proc. Int. Conf., Sept. 3-7, 1984, Houston, Texas. Am. Soc. Civil Engineers, New York, I: 577-592. Sharp, G.D., 1987. Climate and fisheries: cause and effect or msnaging the long and short of it all. S. Afr. J. Mar. Sci., 5 : 811-838. Simpson, J.J., 1983. Large-scale thermal anomalies in the California Current during the 19821983 El Niiio. Geophys. Res. Lett., 10: 937-940. Simpson, J.J., 1984. El Niiio-induced onshore transport in the California Current during 19821983. Geophys. Res. Lett., 11: 241-242. Sinclair, M., Tremblay, M.J. and Bernal, P., 1985. El Niiio events and variability in a Pacific mackerel ( S c o r n b e r m ) survival index: support for Hjort's second hypothesis. Can. J. Fish. Aquat. Sci., 42: 602-608. Smith, P.E., 1985. A case history of an anti-El Niiio to an El Niiio transition on plankton and nekton distribution and abundances. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North. Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Promam. . Univ. Wash:, Seattle, pp. 121-142. Stewart, B.S., Yochem, P.K. and Schreiber, R.W., 1984. Pelagic red crabs as food for gulls: a uossible benefit of El Niiio. The Condor, 86: 341-342. Stewart, B.S., Yochem, P.K., DeLong, R.L: and Antonelis, G.A., In press. Trends in abundance and status of pinnipeds on the Southern California Channel Islands. In: Proc. Third Channel Islands Symp., Santa Barbara, Calif., 2-6 March, 1987. Stephens, J.S., Jr., Jordan, G.A., Moms, P.A., Singer, M.M. and McGowen, G.E., 1986. Can we relate larval fish abundance to recruitment or population stability? Calif. Coop. Oceanic Fish. Invest. Rep., 27: 65-83. Tabata, S., 1985. El Niiio effects along and off the Pacific coast of Canada. In: W.S. Wooster and D.L. Fluharty (Editors), El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, pp. 85-96. Tegner, M.J. and Dayton, P.K., 1981. Population structure, recruitment and mortality of two sea in a kelp forest. Mar. Ecol. Prog. urchins (Stronmlocenmtu h c i s c a n m and S, Ser., 5: 225-268. Tegner, M.J. and Barry, J.P., 1987. Physical mechanisms of larval dispersion affect recruitment and survival of red sea urchins (Saanpvlocentrotus franciscanus) populations in the southern California Bight. EOS, 68: 1,751. Tegner, M.J. and Dayton, P.K., 1987. El Niiio effects on Southern California kelp forest communities. Adv. Ecol. Res., 17: 243-279. Tegner, M.J. and Levin, L.A., 1983. Spiny lobsters and sea urchins: analysis of a predator-prey interaction. J. Exp. Mar. Biol. Ecol., 73: 125-150. VanBlaricom, G.R., In press. Dynamics and distribution of black abalone populations at San Nicolas Island. In: Proc. Third Channel Islands Symp., Santa Barbara, Calif., 2-6 March, 1987. VanBlaricom, G.R. and Estes, J.A. (eds.), 1988. The Community Ecology of Sea Otters. Springer-Verlag, Berlin, 247 pp. VanBlaricom, G.R. and Stewart, B.S., 1986. Consumption of pelagic red crabs by black abalone at San Nicholas and San Miguel Islands, California. The Veliger, 28: 328-329. Walker, N., Taunton-Clark, J., and Pugh, J., 1984. Sea temperatures off the South Africa West Coast as indicators of Benguela warm events. S. Afr. J. Sci., 80: 72-77. Wendell, F.E., Hardy, R.A., Ames, J.A., and Burge, R.T., 1986. Temporal and spatial patterns in sea otter, Enhvdra M,range expansion and in the loss of Pismo clam fisheries. Calif. Fish Game, 72: 197-212.
472
Wilson, K.C. and Togstad, H., 1983.Storm caused changes in the Palos Verdes kelp forests. In: W. Bascom (Editor), The Effects of Waste Disposal on Kelp Communities. Inst. Mar. Res., Univ. Calif., pp. 301-307. Wooster, W.S. and Fluharty, D.L. (eds.), 1985.El Niiio North, Niiio Effects in the Eastern Subarctic Pacific Ocean. Wash. Sea Grant Program, Univ. Wash., Seattle, 312 pp. Zimmerman, R.C. and Kremer, J.N., 1984.Episodic nutrient supply to a kelp forest ecosystem in Southern California. J. Mar. Res., 42:591-604. Zimmerman, R.C. and Robertson, D.L., 1985.Effects of the 1983 El Niiio on growth of giant kelp, Macrocvstls ’ Dvrifera, at Santa Catalina Island. Limnol. Oceanogr., 30: 1,298-1,302.
473
THE IMPACT OF THE "EL NIROO" DROUGHT OF 1982-83 ON A PANAMANIAN SEMIDECIDUOUS FOREST
EGBERT G. LEIGH, JR., D.M. WINDSOR, A. STANLEY RAND, and ROBIN B. FOSTER, Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Panama (mailing address: Smithsonian Tropical Research Institute, APO Miami 34002-0011,USA)
ABSTRACT Leigh, E.G. Jr., Windsor, D.M., Rand, A . S . and Foster, R.B., 1989. The impact of the "El Niiio" drought of 1982-83 on a Panamanian semideciduous forest. The 1982-83 El Niiio event afflicted Barro Colorado Island, in central Panama, with a dry season that began a month early. Although the drought ended "on schedule", it was especially severe during the last 13 of its 24 weeks. It is considered the "worst" dry season since the start of Barro Colorado's rainfall records in 1929, and quite possibly the worst in 190 years. Tree mortality seems to have been 5 times higher than usual in 1983, and Barro Colorado's forest was visibly stressed, but its animal populations were little affected. The forest seemed to have recovered fully by 1984's rainy season. Damage to the forest was much less severe or lasting than that caused by a contemporaneous drought in east Borneo. Moreover, the El Niiio drought did not initiate a progressive deterioration of Barro Colorado's forest community comparable to the progressing disaster unleashed by Hurrican Allen on the reefs of the north coast of Jamaica. The sharp dry season that strikes Barro Colorado every year seems to have "preadapted" its forest for surviving El Niiio droughts without excessive damage, so long as they do not occur "out of season."
1 INTRODUCTION Thanks to events associated with an unusually strong El Niiio, the dry season of 1982-83 in central Panama began on 14 November 1982, a full month early. This dry season ended "on schedule" on 27 April 1983. Although this Panamanian drought was far less devastating than the contemporaneous El Niiio drought that afflicted east Borneo (the eastern portions of Sabah and Kalimantan), where, over a ten-month period, rainfall was one third the average and harsh dry seasons occurred where they are not the rule (Leighton and Wirawan, 1986), the 1982-83 dry season impressed observers in central Panama as far surpassing the ordinary. This paper will address three questions. How severe was Panama's El Niilo drought? How much, and in what ways, was it more severe than a normal dry season? How did this unusual drought affect the plants and animals of local forests? We will answer these questions with data and observations from Barro Colorado Island, a 1,500-hectareisland isolated from the surrounding mainland
474 in 1914 by the rising waters of Gatun Lake, a lake destined to form the central portion of the Panama Canal. This island was declared a reserve in 1923, and is now administered by the Smithsonian Tropical Research Institute (Leigh et al., 1982). Perhaps the best way to introduce Barro Colorado to the reader is by way of comparisons with forest communities of east Borneo. On Barro Colorado Island, the mature forest of the central plateau has an average of 417 trees over 10 cm in diameter at breast height (dbh), representing an average of 93 species, per hectare. This forest is roughly 35 m tall, and has a basal area of 28.5 m2/ha (Foster and Hubbell, in press and unpublished; Foster and Brokaw, 1982).
The forest of Sebulu, east Borneo, is 60 m tall (Kira, 1978);
judging from data on Research Plot 17 in Fox (1973), the forest of Sepilok, in Sabah, has basal area exceeding 42 m2/ha, and 627 trees over 10 cm dbh, representing 154 species, per hectare: 417 such trees would contain roughly 130 species. We have calculated these species counts from the numbers given by Fox (1973) for RP 17, assuming that Fisher's alpha (Fisher et al., 1943; also see Note 1) is relatively independent of plot size, as is true on Barro Colorado (Foster and Hubbell, in press). Barro Colorado has no dominant family of trees analogous to the dipterocarps of Borneo. Average annual rainfall for Barro Colorado is 2,600 mm (Windsor, in press), which is roughly comparable to the annual rainfall of sites in east Borneo, where Sepilok, Sabah, receives 3,150 mm per year (Fox, 1973), and Balikpapan in east Kalimantan receives 2,232 mm per year (Kira, 1978). East Borneo, however, lacks a severe dry season: averaPe rainfall exceeds 100 mm for every month of the year at nearly all sites there. Barro Colorado, by contrast, has a severe dry season, which normally begins some time in December and normally ends within two weeks of the first of May. Over the past 57 years, rainfall on Barro Colorado has averaged 12 mm per week for the first 16 weeks of the year, compared to 69 mm per week for the next 32 (Windsor, in press). This predictable seasonal drought has influenced every aspect of the biology of Barro Colorado (Leigh et al., 1982).
Its forest has many deciduous,
or facultatively deciduous, canopy trees. The production of fruit and new leaves - - vital food for many animal populations - - has two seasonal peaks, one centered about the beginning of the rainy season, in May, and the other in September. Many evergreens also season. The regularly recurring crucial role in regulating Barro b; Leigh and Windsor, 1982). In
flush new leaves at the beginning of the dry shortages of fruit and new leaves play a Colorado's animal populations (Foster, 1982a, the dipterocarp forests of Malesia (Malaya and
the Sunda Shelf), fruit supply is vastly more irregular and less predictable in both space and time (Fogden, 1972). There, much of the forest's fruit is produced during the famous episodes of gregarious fruiting of dipterocarps and associated canopy trees, which occur once every several years (Janzen, 1974).
475 2 THE SEVERITY OF THE EL NINO DROUGHT How severe, and how unusual, was the 1983 El Nifio drought? We base our assessments on data compiled by Windsor (in press). This drought was indeed truly remarkable. The total rainfall for the last 8 weeks of 1982 and the first 16 of 1983 was 121 mm, compared to an average of 787 mm for this period. Let us compare, period by period, the El Nifio dry season of 1982/83 with other notable dry seasons of the last 60 years. This will involve us in a thicket of minima, quartiles and medians, but may provide some perspective on what was unusual about the 1982/83 dry season. The 1976/77 dry season, the longest of the last 60 years, was longer than that of 1982/83, starting on 15 (rather than 14) November, and ending on 17 May (rather than 27 April), but the 1977 dry season was considerably less intense (Table 1). During the last 6 weeks of 1982 and the first 16 of 1983, 88 mm of rain fell, compared to 184 mm for the next driest year (1976-77) of our 57-year record, a lower quartile of 315 mm, and a median of 567 mm (very close to the average for this period of 573 nun). lowest on m m ) . The median of
The rainfall during the last six weeks of 1982 was 61 mm, the second record (the lowest was 52 mm in 1976, and the lower quartile, 207 rainfall during the first four weeks of 1983 was 23 mm, compared to a 42 mm, a lower quartile of 12 mm, and a minimum of 2 mm (in 1950 and
1975). The rainfall during the next 11 weeks of 1983 was 3 mm, the lowest on record: the next lowest was 13 mm (in 1948): the lower quartile, median and average for this period were 38, 69 and 90 mm respectively. Over the first 16 weeks of 1983, total rainfall was 26 mm, a tie with 1959 for the record low: the lower quartile, median and average for this period were 84, 138 and 172 mm respectively. Comparing rainfall figures for selected periods can be tricky. If we look at the total rainfall for the last six weeks of one year and the first 19 of the next, then we find that the minimum rainfall was 184 mm, in 1976/77; the next lowest was 292 mm, in 1948/49, while the rainfall in 1982-83 was 328 nun, and the lower quartile and median were 442 and 664 mm respectively. Yet all agree that the El Niito dry season of 1982/83 was "much worse", largely because it seemed "off the scale" compared to their collective experience of the previous 15 or 20 years. Temperatures were extraordinarily high in March and early April of 1983 (Table 1). Soil moisture contents did not achieve record lows in 1982/83 (Table 2), thanks perhaps to an earlier change in sampling technique. On the other hand, the moisture content (by weight) of 15 0.25 m2 samples of leaf litter taken on 22 April 1983 averaged 5%, compared to 18% for similar series of samples taken in the driest of the last four weeks of the dry seasons of 1976, 1977, and 1978 (Wheeler and Levings, 1988). In these other years, the moisture content of the driest sinele sample of the 15 from the year's driest week was roughly 5% (Wheeler and Levings. 1988). The 1983 dry
476 season apparently eliminated from the leaf litter the damp patches that are normally available as refuges for moisture-loving litter invertebrates. This was indeed a severe drought.
3 THE IMPACT UPON PLANTS OF THE EL NInO DROUGHT 3.1 Siens of Stress Robin Foster observed that, in 1983, wilting was already noticeable among forest plants by the first week of March, and signs of stress multiplied as the dry season progressed. Whole branches of Gustavia died, their leaves wilted, as if sacrificed to preserve the water balance of the rest of the tree (Tyree and Sperry, 1988). Robin Foster also found that the El Nido drought affected second-growth forest more severely than mature forest, forest on shallow soils worse than forest on deeper ones, plants along the seasonal streamlets and swamps of Barro Colorado worse than those on slopes and ridges, plants on those shores and ridges of the island that were exposed to trade winds worse than plants in more sheltered areas, and plants in forest clearings worse than plants in tall, shady forest. In 1983 Gatun Lake dropped to its lowest level since 1977, and plants along its edge seemed particularly hard hit, especially palms and Gustavia. Damage was worst where the soil was shallow and the water was reaching only to bedrock, not even touching the mineral soil. When the rains came on 27 April 1983, even severely wilted plants recovered turgor. Nevertheless, on 30 April, Robin Foster noted that below the thick layer of leaf litter, the soil was still fairly dry and cracked, throughout the forest. The day after a rain on 3 May, he observed that the litter layer was wet and matted down, but that cracks were still visible in the soil: these cracks may have remained until after three days of rain starting 13 May.
The
standing crop of litter, the deepest Foster had ever seen, thus appears to have absorbed much of the first rains. The ground was
so
dried out that two major
rains did not suffice for the soil to recover its normal structure. 3.2 Mortality The El Nido drought imposed an unusually severe mortality on the forest of Barro Colorado. Our best data are from the 50-hectare "forest dynamics plot" on the island's central plateau (Hubbell and Foster, unpublished).
Of the
trees 20 cm dbh and over, 11.4% died between 1982 and 1985, a death rate o f 3.8% per year. By contrast, the death rate of such trees on three 1-hectare tracts within this plot averaged 1.4% per year between 1975 and 1980 (Putz and Milton, 1982). The drought appears to have imposed an extra mortality of roughly 7% on trees 20 cm dbh and above, which afflicted all size classes
477
TABLE 1 Weekly rainfall (mm) and weekly average maximum temperature dry seasons.
Month
Nov .
Dec .
43 44 45 46 47 48 49 50 51 52
76 74 38 198 58 20 196 91 127 25
89 41 104 74 18 0
1 2
3
3
3
4
Mar.
5 6 7 8 9 10 11 12 13
Apr .
May
during selected
Rainfall Average maximum daily temperature Week in successive weeks of: in successive weeks of: number 1975 1976 1977 1981 1982 1975 1976 1977 1981 1982
Jan.
Feb .
("C)
14 15 16 17 18 19 20 21
Dry seasons ending in:
3
0 28 3
94 61 155 91 86 48 3 13 41 0
5 3 8 10 8 10 0 38 0 3 0 1 3 0 5 0 0 0 0 1 3
13
0 5 3
0 0 8 3
8 0 0 3
3 1 0 0 0 0 3 6 0 0 1 0 48 3 15 0 5 8 4 13 8 76 33 5 30 3 15 18 23 94 89 46 94 46
69 84 102 119 279 79 9 7 48 28
94 46 25 10 13 15 0 0 20 13
30.4 30.0 30.0 30.2 30.0 30.1 29.4 28.8 30.2 30.7
102 3 5 5 13 3 3 0 0 0 0 5 0 0 94 0 0 13 71 25 0
5 8 0 10 0 0 0 0 0 0 3 0 0 0 0 0 91 81 61 66 13
30.4 31.4 32.1 30.7 31.4 30.6 3 0 . 3 30.3 32.2 30.4 30.5 31.5 3 0 . 6 3 1 . 1 30.9 30.3 31.0 30.4 31.0 30.6 31.6 31.2 31.0 31.8 30.7 3 1 . 1 31.9 31.0 31.0 32.1 3 1 . 1 30.5 30.5 30.7 3 1 . 3 31.2 31.3 31.5 31.8 32.6 31.2 31.8 n/d 30.4 31.6 3 1 . 3 32.3 31.1 31.1 31.1 3 1 . 6 31.1 32.1 31.2 32.4 30.7 32.2 31.7 31.0 31.1 31.7 31.8 29.8
330
1976 1977 1978 1982 1983
30.1 31.6 31.2 30.5 29.8 31.8 31.9 31.3 30.1 30.4 30.0 3 1 . 3 3 0 . 3 29.5 29.3 n/d 30.9 31.0 29.7 3 0 . 6 31.5 30.7 30.4 31.2 31.4 30.0 30.5 31.9 31.7 30.5 28.4 31.8 30.9 30.7 30.8 31.4 30.2 32.2 31.6 31.4 30.1 30.9 31.4 31.2 32.0 31.2 32.1 31.8 32.0 31.8 32.6 32.5 32.6 32.7 32.1 31.9 31.8 33.6 31.4 33.4 32.8 34.2 32.5 34.2 32.9 33.4 32.7 34.5 31.1 34.1 32.5 34.7 32.8 3 1 . 3 3 3 . 0 31.9 31.8 32.0 3 3 . 0 31.0 31.9 32.9
1976 1977 1978 1982 1983
Data from Tables A2 and B1 of Windsor, in press. Temperatures are from a Stevenson screen in the laboratory clearing. n/d, no data. nearly equally. On the 50-hectare plot, 7.9% of the trees 10-19 cm dbh, and 8.9% of the shrubs and treelets 1-9 cm dbh, died between 1982 and 1985. The trees 10 cm dbh and over that died during this period represented 10.8% (3.11 m2/ha) of the basal area (28.5 m2/ha) of trees on the plot in 1982 10 cm dbh and over. During the week of 26 April 1983, Foster checked the 10,794 stems 1 cm dbh and over on 2.5 hectares of the plot, and found that 69 were freshly dead, with
478 TABLE 2 Changes in soil moisture content late in the dry season, and early in the rainy season, of selected years. Per cent soil moisture by weight (average of ten samples, from top 10 cm of soil in 1982 and 1983, and from top 5 cm in earlier years, taken on the same day of successive weeks in a ravine near the laboratory clearing) in years indicated at tops of columns. Month
Week number
1976
1977
1978
1982
1983
March
10 11 12 13 14 15 16 17 18 19 20 21
29.2 26.7 27.8 25.6 24.9 26.4 33.9 33.1 34.7 40.2 34.1 38.5
28.5 29.5 30.0 32.7 27.0 30.7 28.1 28.1 31.3 30.5 37.5 39.8
32.1 28.9 30.3 29.6 34.0 35.7 31.8 37.5 40.9 38.1 40.0 41.8
32.1 30.8 31.8 29.5 29.0 37.2 33.9 31.3 32.0 37.6 35.3 33.6
n/d 27.3 n/d 27.4 28.1 28.2 26.5 34.3 35.6 36.5 n/d 37.6
April
Data from Table F1 of Windsor, in press. n/d, no data. dried leaves still attached - - 0.63% of the stems checked. The mortality imposed by the El Niilo drought was by no means instantaneous. Different species responded very differently to the drought. The total numbers of stems over 1 cm dbh on the plot of the streamside shrub Acalvuha diversifolia (Euphorbiaceae), originally represented there by 1,582 stems, and the moisture-loving trees Poulsenia armata (Moraceae), Pectandra whitei (Lauraceae, listed as Qcotea skutchii by Hubbell and Foster, in press) and Virola surinamensis (Myristicaceae), initially represented by 3,437, 1,149 and 300 stems, declined by 23, 22, 16 and 14% respectively between 1982 and 1985 (Hubbell and Foster, in press and ms.). Of the Trichilia tuberculata (Meliaceae, the most common big tree on the plot), 17% over 32 cm dbh, but only 9% of its saplings and smaller trees, died during this period. Of the Quararibea asteroleuis (Bombacaceae) 8% over 32 cm dbh, but only 5% of its smaller stems, died during this period, while mortality over these three years of Prioria couaifera (Leguminosae), a common species in local swamps, lay between 1 and 3% for stems of all sizes (Fig. 4 of Hubbell and Foster, in press). 3.3 Phenoloev During the 1983 dry season, more trees lost their leaves than usual, but leaf flush and fruit production generally proceeded on schedule (D.M. Windsor, unpublished data). One exception was palms, numbers of stems of all eight
479 genera of which declined between 1982 and 1985 (Hubbell and Foster, ms.).
The
tall, wet forest palm Socratea exorrhiza, the blowdown colonist Oenocaruus maDora, the swamp oil-palm Elaeis oleifera, and the understory palms Chamaedorea wendlandiana, Bactris maior and Geonoma cuneata, normally put out new leaves all year round, some of these at a nearly constant rate. In 1983, however, leaf production during the dry season was nil, or nearly s o , in these six species (De Steven et al., 1987).
Perhaps because of the drought, which
visibly affected this species in other ways, fruit crops of Virola surinamensis were the lowest in five years of record, although still half the average level (H. F. Howe, personal communication). In 1983, fruit crops of Ouararibea asteroleuis were even lower than the poor 1982 crops (D. M. Windsor, unpublished data): two such poor crops in succession is rather rare. 3 . 4 Lone-term Effects on the Forest
Interested biologists had the impression that the forest on Barro Colorado recovered very quickly after the rains came in 1983. Many of the surviving palms, for example, made up later in the year for their low dry season leaf production (De Steven et al., 1987). Forest-wide,however, recovery was not quite immediate. In the rainy season of 1983, the median, upper quartile, and lower quartile of the proportion of solar radiation (direct and diffuse) reaching to within 60 cm of the forest floor along a line of 400 points spaced 2.5 m apart in the forest dynamics plot, were 1.7%, 3.1% and 0.87% respectively, compared to 0.3%, 0.79% and 0.1% a year later (A.P. Smith and P.F. Becker, personal communication).
The 1984 figures seem normal (cf. Kira,
1978), indicating that, one year after the end of the El Niao drought, the forest's capacity for leaf production had fully recovered. The unusual amount of light penetrating to the forest understory presumably explains the 7% increase in the number of stems 1 - 1.9 cm in diameter between 1982 and 1985 on the forest dynamics plot. This increase took place despite the high mortality that branches falling from drought-killed trees must have imposed, and despite the outgrowth of 17.8% of the original stems into the next larger size classes (Hubbell and Foster, in press). To judge by Table 1 of Hubbell and Foster (in press), surviving stems of all size classes were growing vigorously between 1982 and 1985. During this period, 93% of the basal area lost in dying trees was replaced by diameter growth in surviving stems (Hubbell and Foster, unpublished data). Spot checks on selected hectares of the 50-hectare forest dynamics plot indicate that, after the 1985 census, tree death rates as low as they were between 1975 and 1980, one to two per cent per year ( S . P. Hubbell, pers. comm.). Death rates between one and two per cent per year appear to be the norm for tropical forests the world around (Swaine et al., 1987).
480 In sum, the El Niilo drought did not disrupt the "fabric" of the forest on Barro Colorado, as did its deadly counterpart in east Borneo, where drought killed 25% of canopy trees, and understory fires, set by humans on the borders of the forest and allowed to spread by the unusually dry conditions, killed 90% of the forest's shrubs and treelets (Leighton and Wirawan, 1986).
The El Nirio
drought did kill an unusual number of Barro Colorado's trees.
The El Niiio drought may have drained Barro Colorado's forest of some of its reserves. Phyllis Coley (personal communication) reports that the rate of consumption by insects of mature leaves on saplings of the second-growth "pioneer" species Annona snraauei and Cecropia insianis was 4 to 12 times higher in the rainy season of 1983 than in rainy seasons following earlier,
less severe, dry seasons. Do these high rates of herbivory imply that the plants could not afford to stock their leaves with defensive compounds as fully as usual? It may be very fortunate for central Panama that drought years have no tendency to follow each other in succession, as they were said to do in biblical Palestine (I Kings, 17, 18).
Two or three such years in a row might
well have reduced Barro Colorado to the condition of east Borneo.
4 THE EL NIRO DROUGHT AND ANIMAL POPULATIONS It is harder to assess the effects of the El Niiio drought on animal than on plant populations. Animals move about, and some like to hide: their populations are accordingly more difficult to sample, and one cannot always tell whether a change in number of animals observed along a census route represents a change in numbers, a change in their habits, or a change in their tolerance of observers (due perhaps to the timing of the last poaching incident). Animals, moreover, are far shorter-lived than trees; a population fluctuation that would signal a major event for a tree species may represent nothing unusual for most kinds of animals. Nevertheless, it appears that, although the trees of Barro Colorado suffered markedly during the El Niiio drought, the forest's animals, with a few striking exceptions, were surprisingly little affected. This circumstance contrasts sharply with 1970, when a wet dry season was followed by a failure of the September fruiting peak, causing obvious distress and a most unusual mortality among mammals, and unusual migrations of frugivorous birds, while plants appeared to suffer nothing worse than disruptions of their schedules of flowering and fruiting (Foster, 1982b). The numbers, diversity, and timing of appearance of the (herbivorous) homopterans caught in light traps on Barro Colorado during and after the dry season of 1983 were in no way unusual (H. Wolda, personal communication). Populations of most of the common species of terrestrial mammals showed no unusual changes in response to this dry season (Table 3), perhaps because
48 1 Diutervx, a critical food species for many animals (De Steven and Putz, 1984), bore fruit in extraordinary abundance early in 1983, and Gustavia fruit was abundant later on (W.E. Glanz, personal communication to A.S. Rand). Recaptures late in 1984 of frugivorous bats netted two or more years earlier was, however, markedly less than one would have inferred from recapture rates between 1978 and 1982 of bats first marked two years before the recapture (Charles 0. Handley, Jr., unpublished data).
Unfortunately, no one netted bats during the
two years preceding late 1984, so we do not know what role, if any, the El Niiio drought played in this mortality. TABLE 3 Sighting rates of mammal species on Barro Colorado (Number of animals, or *number of groups, sighted per kilometer of trail walked by census-takers) Based on Table 4 of Glanz (in press).
Census year:
1977-78
1982
Dry Season 1983-84
Diurnal species 0.56 0.10 2.51 3.26 0.22 0.09
0.81 0.21 1.20 1.35 0.29 0.13
0.84 0.33 1.05 2.26 0.18 0.36
DidelDhis (common opossums) DasvDus (armadillos) Acouti (pacas) Proechimvs (spiny rats) Potos (kinkajous) Mazama (deer)
0.67 0.55 0.70 0.50 0.39 0.28
0.31 0.63 0.61 0.17 0.23 0.49
0.33 0.04 0.98 0.33 0.25 0.66
Total km walked per census yr: daytime nighttime
314 62
168 57
101
Alouatta* (howling monkeys) (white-facemonkeys) Sciurus (squirrels) Daswrocta (agoutis -* (coatis) Tavassu* (peccaries)
-*
Nocturnal species
24
Birds appeared weaker, and more easily stressed by capture, during the El Nifio dry season than in other dry seasons (Rachel Levin, James Karr, personal communications to A.S. Rand).
The El Nifio drought appeared, however, to have
no lasting effect on populations of understory insectivorous antbirds (Greenberg and Gradwohl, 1986). Iguanas and crocodiles near Barro Colorado normally nest during the dry
482
season. They nested at the normal date in 1983, showing no response to that dry season's early start. Hatching success appeared normal for both species. The El Nirlo dry season affected forest floor animals in various ways. Armadillo numbers dropped sharply, but coati numbers dropped much less markedly (Table 3).
The small forest floor lizard Anolis limifrons was reasonably
abundant, and its individuals in good condition, in early April, 1983. By August, 1983, however, populations of this lizard had crashed to lows never observed in the previous 12 years, an event that had no obvious connection with the El Nitio dry season. These populations took years to recover (Robin Andrews, personal communication to A.S. Rand). A.S. Rand observed that amphibian populations generally, and forest floor species especially, appeared to be low in the rainy season of 1983.
Bufo
tvuhonius, a common forest floor toad that breeds in temporary pools during the dry season, was particularly hard hit: its numbers, like those of Anolis limifrons, took years to recover.
Bufo twhonius, however, has suffered
similar population crashes in the past that were not clearly related to droughts. Wheeler and Levings (1988) compared the impact of the 1983 dry season on litter arthropods to those of previous dry seasons, as reported by Levings and Windsor (1982, 1985).
Those groups of litter arthropods that are normally most
abundant in the rainy season, such as Araneae, opilionids, Diplopoda, Hemiptera, beetles, pseudoscorpions, and holometabolous larvae, were roughly as uncommon at the end of 1983's dry season as at the end of the shorter, but severe, dry season of 1976, and much less abundant than at the ends of the long but less severe dry season of 1977 and the mild dry season of 1978. The same applies to ants, but fewer species of ants were active i n April 1983 than in April 1976. Psocoptera and Thysanoptera are normally most abundant during the dry season. Psocoptera were enormously more abundant in April 1983 than at the end of any of the three previously studied dry seasons, while Thysanoptera were about as abundant as they were at the end of the 1977 dry season, and more abundant than at the ends of the 1976 and 1978 dry seasons. The moistureloving groups of litter arthropods, however, sprang back to normal levels of abundance and diversity with unexpected speed, once the rains came in 1983. It is as if the El Nitio dry season affected the "phenology" of these arthropods without damaging their populations. 5 CONCLUDING REMARKS
Taking into account both length and severity, the El Nitio dry season of 1983 appears to be the worst to hit Barro Colorado in the last 60 years. If intense droughts only occur on Barro Colorado in connection with intense El Nitio events, then, judging by the time since the last such spell before 1983 of
El Nifio-relatedcoral mortality on the Pacific coast of Panama, the 1983 dry season was quite possibly the worst on Barro Colorado in the past 190 years (Glynn, 1985).
In view of this circumstance, Barro Colorado's forest suffered
surprisingly little damage
---
much less than the nearby coral reefs of the
tropical eastern Pacific (Glynn, 1985).
The El Nido drought most decidedly
failed to destabilize Barro Colorado's forest community, in contrast to the way Hurricane Allen destabilized the coral reef communities of Jamaica's north coast. There, the immediate destruction inflicted by the hurricane (Woodley et al., 1981) was merely the prelude to a progressive deterioration of the surviving remnants of the reef community (Knowlton et al., 1981, in press). Yet Jamaica's north coast may well suffer such destructive hurricanes once or twice per century (Woodley et al., 1981). Severe droughts in both Barro Colorado and east Borneo appear to be responses to the same nearly world-wide signal, severe El Nifio events (Rasmusson and Wallace, 1983; Windsor, in press; Leighton and Wirawan, 1986). Insofar as the severity of an El Nifio event governs the severity of the droughts associated with it, the "return time" on Barro Colorado for a drought as severe as that of 1982/83 should be roughly the same as the return time in east Borneo for droughts of the severity observed there in 1982/83. Yet east Borneo suffered far more than Barro Colorado from the El Nifio droughts of 1982/83: the forests of east Borneo lost 25% of their canopy trees to this drought (Leighton and Wirawan, 1986). Perhaps more surprisingly, there is no evidence that the forest of the 50hectare forest dynamics plot on Barro Colorado has ever suffered wildfires during the last 2,000 years, although there was some human activity, and even some settlement, on the plot during the first 1,500 years A.D. (Piperno, in press).
By contrast, wildfires are known from rain forests that do not
normally suffer severe dry seasons, both in Amazonia (Sanford et al., 1985) and east Borneo (Leighton and Wirawan, 1986).
Are we missing something, or are
severe droughts less devastating for seasonal forests such as those of Barro Colorado than for less seasonal rain forests? Does Barro Colorado benefit by suffering a relatively severe drought every year, rather than the much less regular and predictable droughts that afflict dipterocarp forests? Was the El Nido merely a relatively severe version of what Barro Colorado's forest is "primed" to experience at this time of year? After all, the El Nirlo drought did not impair the basic seasonal rhythm of Barro Colorado's forest community, as it very clearly did do in east Borneo. Therein may lie the essential difference. 6 ACKNOWLEDGEMENTS We are most grateful to the Environmental Sciences Program of the
484
Smithsonian Institution, and the many individuals who have labored on its behalf, for the variety of "background" data that allow us to compare one year with another; to S.P. Hubbell for the supply of unpublished data on tree growth and mortality; and to Una Smith for preparing the tree data in usable form. We are also deeply indebted to the many members of the scientific community of Barro Colorado Island, who contributed unpublished data and observations to this endeavor. Finally, we thank the plants and animals of this island that are such a j o y to try to understand. Note 1: Fisher's alpha is a measure of diversity. It is most appropriate for communities where the distribution of species abundances in a sample follows the "log series", whereby the number f(N) of species with N sampled individuals is f(N)
- axN/N,
where the parameters a (Fisher's alpha) and x are given by the equations
Here,
S
and NT are the total numbers of species and individuals in the samples
(Fisher et al.,1943;May, 1975).
When the distribution of species abundances
in a sample follows the log series (as is true for various plots in Malaysian forests; Foster and Hubbell, in press), the value of a is independent of sample size. Moreover, if the distribution of species abundances in a sample follows
the log series, then D, Simpson's (1949) measure of diversity, given by
may be expressed as
which is nearly a. 7 REFERENCES De Steven, D. and Putz, F.E., 1984. Impact of mammals on recruitment of a tropical canopy tree, DiDtervx panamensis, in Panama. Oikos, 43: 207-216. De Steven, D., Windsor, D.M., Putz, F.E., and de Leon, B., 1987. Vegetative and reproductive phenologies of a palm assemblage in Panama. Biotroiica, 19: 342-356.
485 Fisher, R.A., Corbet, A.S. and Williams, C.B., 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology, 12: 42-58. Fogden, M.P.L., 1972. The seasonality and populations of equatorial forest birds in Sarawak. Ibis, 114: 307-342. Foster, R.B., 1982a. The seasonal rhythm of fruitfall on Barro Colorado Island. In: Leigh et al. (1982), p p . 151-172. Foster, R.B., 1982b. Famine on Barro Colorado Island. In: Leigh et al. (1982), pp. 201-212. Foster, R.B. and Brokaw, N.V.L., 1982. Structure and history of the vegetation on Barro Colorado Island. In: Leigh et al. (1982), pp. 67-81. Foster, R.B. and Hubbell, S.P., in press. Estructura de la vegetacion y composicion de especies en un lote de cincuenta hectareas de la isla de Barro Colorado. In: Leigh et al. (in press). Fox, J.E.D., 1973. A Handbook to Kabili-Sepilok Forest Reserve. Sabah Forest Record no. 9. Borneo Literature Bureau, Kuching, Sarawak, Malaysia. Glanz, W.E., in press. Neotropical mammal densities: how unusual is the community on Barro Colorado Island, Panama? In: A.H. Gentry (Editor), Four Neotropical Forests. Yale University Press, New Haven, CT. Glynn, P.W., 1985. El Nifio-associateddisturbance to coral reefs and postdisturbance mortality by Acanthaster planci. Mar. Eco1.-Prog. Ser., 26: 295- 300. Greenberg, R. and Gradwohl, J . , 1986. Constant density and stable territoriality in some tropical insectivorous birds. Oecologia, 69: 618625. Hubbell, S.P. and Foster, R.B., in press. Structure, dynamics, and equilibrium status of old-growth forest on Barro Colorado Island. In: A.H. Gentry (Editor), Four Neotropical Forests. Yale University Press, New Haven, CT. Hubbell, S.P. and Foster, R.B., ms. Short-term population dynamics of trees and shrubs in a neotropical forest: El Nifio effects and successional change. Submitted for publication. Janzen, D.H., 1974. Tropical blackwater rivers, animals, and mast fruiting by the Dipterocarpaceae. Biotropica, 4: 69-103. Kira, T., 1978. Community architecture and organic matter dynamics in tropical lowland rain forests of Southeast Asia, with special reference to Pasoh Forest, West Malaysia. In: P.B. Tomlinson and M.H. Zimmermann (Editors), Tropical Trees as Living Systems. Cambridge University Press, Cambridge, England. Knowlton, N., Lang, J.C and Keller, B.D., in press. Case study of natural population collapse: post-hurricane predation on Jamaican staghorn coral. Smithsonian Contributions to Marine Science. Knowlton, N., Lang, J.C., Rooney, M.C. and Clifford, P., 1981. Evidence for delayed mortality in hurricane-damaged Jamaican staghorn corals. Nature, 294: 251-252. Leigh, E.G. Jr., Rand, A.S. and Windsor, D.M. (Editors), 1982. The Ecology of a Tropical Forest. Smithsonian Institution Press, Washington, D.C. Leigh, E.G. Jr., Rand, A.S. and Windsor, D.M. (Editors), in press. Ecologia de un bosque tropical. Leigh, E.G. Jr., and Windsor, D.M., 1982. Forest production and the regulation of primary consumers on Barro Colorado Island. In: Leigh et al. (1982), pp. 111-122. Leighton, M. and Wirawan, N., 1986. Catastrophic drought and fire in Borneo tropical rain forest associated with the 1982-1983 El Nifio Southern Oscillation event. In: G.T. Prance (Editor), Tropical Rain Forests and the World Atmosphere. Westview Press, Boulder, Colorado, pp. 75-102. Levings, S.C. and Windsor, D.M., 1982. Seasonal and annual variations in litter arthropod populations. In: Leigh et al. (1982), pp. 355-387. Levings, S.C. and Windsor, D.M., 1985. Litter arthropods in a tropical deciduous forest: relationships among years and arthropod groups. Journal of Animal Ecology, 54: 61-69.
486 May, R.M., 1975. Patterns of species abundance and diversity. In: M.L. Cody and J.M. Diamond (Editors), Ecology and Evolution of Communities. Harvard University Press, Cambridge, MA, pp. 81-120. Piperno, D.R., in press. Fitolitos, arqueologia y cambios prehistoricos de la vegetacion en un lote de cincuenta hectareas de la isla de Barro Colorado. In Leigh et al. (in press). Putz, F.E. and Milton, K., 1982. Tree mortality rates on Barro Colorado Island. In Leigh et al. (1982), pp. 95-100. Rasmusson, E.M. and Wallace, J.M., 1983. Meteorological aspects of the El Nido/Southern Oscillation. Science, 222: 1195-1202. Sanford, R.L. Jr., Saldarriaga, J., Clark, K.E., Uhl, C. and Herrera, R., 1985. Amazon rain-forest fires. Science, 227: 53-55. Simpson, E.H., 1949. Measurement of diversity. Nature, 163: 688. Swaine, M.D., Lieberman, D. and Putz, F.E., 1987. The dynamics of tree populations in tropical forest: a review. J. Trop. Ecol., 3 : 359-366. Tyree, M.T. and Sperry, J.S., 1988. Do woody plants operate near the point of catastrophic xylem dysfunction caused by dynamic water stress? Plant Physiology, 88: 574-580. Wheeler, D.E. and Levings, S.C., 1988. Impact of El Niiio on litter arthropods. In: J.C. Trager (Editor), Advances in Myrmecology. Flora and Fauna, Gainesville, FL, pp. 309-326. Windsor, D.M., in press. Climate and moisture availability in a tropical forest: long-term records from Barro Colorado Island, Panama. Smithsonian Contributions to the Earth Sciences. Woodley, J.D. et al., 1981. Hurricane Allen’s impact on Jamaican coral reefs. Science, 214: 749-755.
487
THE BOTANICAL RESPONSE O F THE ATACAMA AND PERUVIAN DESERT FLORAS TO THE 1982-83 EL NINO EVENT M. 0. DILLON Department of Botany, Field Museum of Natural History, Chicago, IL 60605-2496 (USA) P. W. RUNDEL Laboratory of Biomedical and Environmental Sciences and Department of Biology, University of California, Los Angeles, CA 90024 (USA)
ABSTRACT Dillon, M. 0. and Rundel, P. W. 1989. The botanical response of the Atacama and Peruvian Desert floras to the 1982-83 El Niiio event. The Atacama and Peruvian Deserts form a continuous belt along the western escarpment of the Andes for more than 3,500 km from the Peru/Ecuador border (5°00,S) to northern Chile (29O55'S). This region owes its severe aridity to a constant temperature inversion generated, in part, by the cool, north-flowing Humboldt (Peruvian) Current. The atmospheric influences associated with a positionally stable, subtropical anticyclone result in a mild, uniform coastal climate and regular formation of thick stratus clouds below 1,000 m during the winter months. Where coastal topography is low and flat, this stratus layer dissipates inward over broad areas with little biological impact. However, where small isolated mountains or steep coastal slopes are present, this stratus layer forms a fog zone concentrated against the hillsides. These moist fogs allow for the development of rich fog-zone vegetation termed lornus formations. These floristic assemblages essentially function as terrestrial islands separated by hyperarid habitat where plant life is completely absent or virtually so, with the exception of Cactaceae or terrestrial TiZlunhiu species. Significant desert rains occur only in association with rare, but recurrent, El Niiio perturbations. What was estimated as the most severe El Niiio of the century began in June of 1982. By early 1983, rains had moved down the length of coastal Peru and into northern Chile, and showers continued sporadically through June 1983. Botanical fieldwork throughout the Peruvian lomas in 1983 indicated the presence of a rich bloom. Species number and density had not been at such heights since a major El Niiio event in 1925. The dynamics of lomas endemism and of how plants came to colonize the fog-island archipelago are potentially related to rare episodes of El Niiio rainfall. 1 INTRODUCTION
The effects of the 1982-83 El Niiio Southern Oscillation (ENSO) event upon marine biota are well documented (Barber and Chhvez, 1983; Arntz et al. 1985; Robinson and del Pino, 1985; Wooster and Fluharty, 1985; Glynn, 1988; cf. present volume); however, the short and long term effects of these disturbances upon terrestrial systems are not as
488
well documented (Hamann, 1985; Luong and Torro, 1985; Leighton and Wirawan, 1986; Leigh et al., this volume). One geographic area that was profoundly influenced by the 1982-83 ENSO is the western coast of South America. The following discussion combines observations made by the authors during fieldwork before, during, and after the 1982-83 ENSO event and observations drawn from published accounts documenting strong ENSO events. Topics presented here include, (1) a brief description of the lomas formations of the western South American deserts, including structure and composition, (2) a description of the prevailing climate of this area and how it is altered by ENSO events, (3) a few historical observations made during the intense ENSO event of 1925, (4) an account of the observed effects upon vegetation, both on the coastal lomas formations of Peru and northern Chile and the nearby GalApagos Islands, and lastly (5) a few speculative remarks about the possible effect of El Niiio events upon the distribution and evolutionary history of the vegetation of the lomas formations. 2 LOMAS FORMATIONS
The Peruvian and Atacama Deserts form a continuous belt from the Peru/Ecuador border (S'OO'S) to La Serena (29'55's) in northern Chile, a distance of more than 3,500 km. These deserts owe their severe aridity to a climatic regime dominated by a constant temperature inversion generated, in large part, by the cool, north-flowing Humboldt (Peruvian) Current. Also important is the influence of strong atmospheric subsidence associated with a positionally stable, subtropical anticyclone (Trewartha, 1961). This results in a mild, uniform coastal climate with the regular formation of thick stratus cloud banks below 1,000 m during the winter months. Where coastal topography is low and flat, the stratus layer dissipates inward over broad areas with little biological impact. However, where isolated small mountains or steep coastal slopes are present, a fog zone forms, concentrated against the hillsides. These moist fogs, termed the curnunchucu in Chile and the gudu in Peru, allow the development of fog zone vegetation known as lomas formations. The fogs, especially abundant during ENSO events, are the key to the extent and diversity of vegetation in the Atacama and Peruvian Deserts (Rundel et al., 1990). The physical and ecological parameters of the lomas formations have recently been discussed in detail (Rundel et al., 1990), but a brief review is advisable for discussion. The term lomas literally refers to small hills and is used to denote the coastal, fogdependent vegetation of Peru and northern Chile. The distribution of plant communities along the coast is dependent upon the presence of sufficient moisture to support growth. Since available moisture is at scattered localities, the vegetation does
not occur as a solid strip along the coast, but rather exists as isolated communities or
489
"terrestrial islands" with expanses of hyperarid desert in between (Fig. 1). A diagrammatic representation of the lomas in Peru and Chile (Fig. 2) shows them to be similar in their utilization of fog for available moisture; however, the species composition of each formation is unique, reflecting diverse floristic origins and differences in substratum, topography and climatic history. The distribution of plant species within each formation is also variable and controlled by habitat structure, available moisture, and competition. The most common sources of lomas vegetation include montane Andean disjuncts, amphitropic disjuncts, semi-arid Ecuadorian and central Chilean taxa, pantropical weeds, and lomas endemics. Current tabulation of the entire lomas flora shows it to contain ca. 1,000 species of angiosperms and ferns (Chile: 73 families, 196 genera, 375 species; Peru: 72 families, 284 genera, 560 species), of which over 40% are narrow endemics (Dillon, unpublished data). Unlike many desert areas in the world, the hyperaridity of the Atacama and southern Peruvian Deserts appears to be geologically very old (Alpers and Brimhall, 1988). Climates similar to those that prevail today have existed since the Middle Miocene. These conditions arose with the establishment of the ancestral Humboldt Current as an increase in the upwelling intensity of the Antarctic ice cap began 13-15 million years ago, and to an uplift of the central Andean Cordillera to at least half of its present height at this same time (Alpers and Brimhall, 1988). This long history of extreme aridity combined with changing climates has had a profound impact on plant evolution and phytogeographic patterns. A 250 km segment of desert near the Peruvian/Chilean border, between Tacna, Peru and Iquique, Chile, marks a significant biological discontinuity between the floras of each region. While the two countries share many families and genera the congruence at the species level is quite low with only 68 species common to communities on both sides of this area. This is reflected, for instance, in the distribution of the Cactaceae within the lomas, where only 4 of the 14 genera represented have species distributed in both Peru and Chile. Furthermore, the cacti floras of the two regions are nearly 100% endemic. It would appear, therefore, that this area continues to act as an effective barrier to north-south dispersal. Conditions to which lomas communities are adapted have been most persistent and intense in the south between about 25" and 19" S latitude and perhaps to a somewhat lesser degree between 19" and 16" S latitude. The presence of salt accumulations in soils, but not nitrate deposits, between 16" and 13" S latitude should reflect a decline in the persistence or intensity of extremely arid conditions. This may correlate with north-south vacillation in the extent of upwelling during the Pleistocene. North of about 13" S latitude, appropriate conditions to support lomas vegetation have been relatively
490
/'
1 ,1"
1
I
>
\ CERRO CHIMBOTE-
J-
1
I
Fig. 1. Geographic features, including lomas localities referred to in text, within the Atacama and Peruvian Deserts. ephemeral. Although lomas communities may have temporarily colonized the region during interstadials, their present northernmost range was only established about 5,000 years ago (Rollins et al., 1986). It has been suggested that the El Niiio phenomenon did not occur during Pleistocene glacial advances because the low pressure center over
49 1
herbaceous perennials
~
_
Fig. 2. Diagrammatic view of vegetation zonation in the coastal fog zone of southern Peru (redrawn from Ellenberg, 1959) and northern Chile at Paposo (Rundel and Mahu, 1976). Indonesia would not have weakened to allow a warm-water displacement across the Pacific and down the coast of Peru (Simpson, 197Sa, 1975b; Richardson, 1981). The composition and distribution of the present-day lomas vegetation reflect not only the past climatic and geologic events, but also the present-day climate. 3 COASTAL CLIMATE
Values of mean annual precipitation reported for coastal cities of northern Chile are the lowest for any long term records in the world. Iquique and Arica with more than 50 years of precipitation records average 2.1 mm and 0.6 mm respectively. Even these low figures are misleading, however, because of the extreme rarity of days with precipitation. A single storm that dropped 10 mm of precipitation on Arica in January of 1918
_
492
accounted for nearly a third of all the precipitation received by the city over the past half century. In fact, Miller (1976) reports that there have been only about half a dozen days in the past 30 years when more than 1 mm of precipitation fell at any of the coastal cities from Antofagasta to Arica. Only about two out of five years on the average have any measurable precipitation, and drought periods of four to eight years are relatively common. Precipitation levels are higher as one moves southward in the Atacama. La Serena, at the southern margin of the desert, has a mean of 127 mm per year, but with high variability. Although Peruvian coastal cities occasionally receive torrential rains associated with El Nixio conditions, heavy storms are virtually unknown along the north coast of Chile. The maximum recorded precipitation for a 24 hour period for Antofagasta, Iquique and Arica was 28 mm, 13 mm, and 10 mm respectively (Miller, 1976). There may be some connection between these storms and El Nifio conditions. For example, in 1925, a strong El Nixio year, the entire coast from Ecuador to Chile received unusually high levels of precipitation. However, other El Nixio years have failed to produce increased precipitation in northern Chile. The relatively strong El Niiio of 1965 failed to increase precipitation levels in central Peru and areas to the south. In contrast, the El Nixio event of 1983 produced conditions leading to heavy rains from northern to southern Peru, and a relatively high 7.3 mm of precipitation in Iquique, but no notable rains further south (Romero and Garrido, 1985; Rutllant, 1985). Nevertheless, the fogs appear to have been unusually dense that year, producing excellent flowering conditions in the coastal "fertile" zone (Prenafeta, 1984). Perhaps the most prominent feature of the coastal climate of Peru, much as in Chile, is its extraordinary stability. Dense stratocumulus clouds form along the coast almost continuously from May to October, particularly in central and southern Peru. This fog formation occurs when long-wave radiation loss to the atmosphere cools the upper surface of the stratus bank. Prohaska (1973) describes the winter of 1967 in Lima in which 90% of all days were overcast, and continuous periods of cloud cover remained as long as 44 days. Clear skies occurred only 1.6% of the time over this period. Such conditions can be considered typical. The summer months, in contrast, are characterized by clear and sunny weather, although occasional clouds do form. Fog is a steady winter feature of virtually the entire coastal region of Peru from Chiclayo south. Below 100 m the fog is primarily a phenomenon of the night and early morning hours but it may be a continuous condition in the zone from 100-800 m. Studies by Prohaska (1973) at Jorge Ch6vez International Airport, on the coast near Lima, found that drizzle occurred 27% of the time through the winter. Despite this nearly 900 hours of fog precipitation, only 6.1 mm of moisture collected in rain gauges.
493
Although little moisture collects in this manner in rain gauges, the lomas vegetation and rock outcrops in the upland fog zone intercept moving fog droplets and act to condense considerable quantities of water. It is fog drip, both from plant structures and from these outcrops, which allows the existence of plant growth. An artificial moisture capture system in the shape of a tetrahedron, four meters in height and 25 m2 at its base, was reported to trap as much as 100 liters of fog moisture per day at the Lomas de Lachay near Lima (Oka, 1986). Precipitation levels are extremely low all along the coast of Peru, but the extreme variability of annual levels gives little significance to mean values. There is no clear latitudinal pattern of change along the coast. Many cities in southern and central Peru have drought periods similar to those already discussed for northern Chile. Pisco (13"42'S, 10 m elevation), for example, sits below the fog zone and averages only 2 mm yearly precipitation. Tacna in southern Peru sits inland within the fog zone at 558 m elevation and receives 44 mm yearly, much of this fog drizzle in August and September (Johnson, 1976). The major problem in interpreting climatic records for most other Peruvian coastal cities, however, is the rare episodes of rain which may obscure the long-term pattern of virtually zero precipitation. Some inland cities receive occasional showers spilling over from stratus in the Andes but these are less a heavy problem than those of the El Niiio years. 4 IMPACT OF FORMER INTENSE EL NINO EVENTS
Stimulation of precipitation levels by ENSO events has occurred several times within the last 100 years. Strong El Nifio conditions occurred in 1891, 1925, 1953 and 1965, with weaker effects felt commonly but not invariably at about seven year frequencies (Johnson, 1976). The periodicity of El Niiio events has been analyzed (Quinn et al., 1978, 1987) and a discussion is provided in the present volume (see Hansen). Strong perturbations producing the rains of the 1982-83 or 1925 magnitude probably have a long-term statistical periodicity of only once or twice per century (Nials et al., 1979; Thompson et al., 1979; Moseley et al., 1981; Quinn and Neal, 1983). The 1925 ENSO was the last event prior to 1982-83 to have had a profound effect upon the terrestrial biota of western South America. A number of observations on the climatic and biological effects of the 1925 El Niiio event are available from descriptions by American biologists who worked in the area at that time, Robert Cushman Murphy
(1926) in Peru, William Beebe (1926) aboard the Arcturus between Panama and the GalApagos Islands (Wooster, 1980), and Ivan M. Johnston (1929) collecting plants in southern Peru and northern Chile, October 1925 through January 1926. Oceanographic observations of both the 1891 and 1925 El Niiio events were made by German merchant
494
ships along the west coast of South America, and these data have been summarized by Schott (1931). Murphy (1926) made fairly detailed measurements of changes in sea surface temperatures through time along the coast of Peru. At Talara (S"45'S) in Peru the temperatures began rising rapidly in midJanuary 1925, increasing 6.7"C to 26.6"C over a 10-day period. At this latitude, sea surface temperatures equalled or exceeded mean atmospheric temperatures. By mid-March these high sea surface temperatures extended much further south, with 26.7"C at Callao (lZ"05'S) and 26.1"C at Mollendo. Trujillo (8'07's) had sea temperatures 8.3"C higher than normal. Antofagasta (23'40's) in northern Chile had only slightly elevated temperatures at 21.1"C. Torrential rains began in northern Peru on January 27, 1925, and continued well into April. Murphy (1926) provided detailed logs of heavy and continuous precipitation in 1925 for a number of cities. In Trujillo, for example, 395 mm of rain fell in March, more than 20 times the normal annual total, causing widespread destruction (Knoch, 1930). Chicama (7"51'S) with a mean annual precipitation of 4 mm and Lima (12"OO'S) with a mean of 46 nun were deluged by 394 mm and 1,524 mm of rain respectively (Goudie and Wilkinson, 1977). Unlike other El Niiio years where environmental anomalies did not extend beyond central or southern Peru, enhanced precipitation was seen as far south as Antofagasta. While rainfall was certainly not torrential, Johnston (1929) reported 17 mm of precipitation falling in a single day there, one of the highest levels experienced in that city at any time this century. No quantitative data on plant diversity and density was gathered, however, Murphy (1926, p. 50) mentioned that the vegetation along the Peruvian coast was abundant and marked the first major bloom recorded for 40 years. Between 28 October 1925 and 18 January 1926, Johnston made 930 collections from the coastal vegetation of northern Chile, and these collections contained over 80% of all plant species recorded for this area and yielded 55 taxa new to science (Johnston, 1929). 5 1982-83 EL NINO EVENT What is being estimated as the most severe El Niiio perturbation of the century began in June 1982. By early 1983, rains had moved down the length of coastal Ecuador and Peru, and showers continued through June 1983. The climatic effects upon Ecuador in 1982-83 have been summarized by Naranjo (1985). Increases in atmospheric pressure and ocean surface water temperatures were first seen in June 1982, four months before the onset of heavy rains. Sea surface temperatures eventually reached 10°C above normal, the highest level on record. Air temperatures were also higher than normal, with Quito, Ecuador averaging 2°C above normal in August 1983. Heavy rains began to
495
fall in November 1982. Guayaquil, Ecuador recorded 150 mm in this normally dry month. Precipitation continued to be heavy in December, also a dry month in normal years, and turned to torrential rains (and associated catastrophic floods) with the beginning of the wet season in January 1983. All but one of the six months from January through June received more than 600 mm, with the highest level reaching 830 mm in March. Moderately heavy rains continued through July, well into the normal dry season, before dropping off to normal levels. The total precipitation at Guayaquil in 1983 was 3,949 mm, nearly four times the mean level of 1,120 mm. Inland stations at higher elevations in Ecuador also received greater than normal amounts of precipitation during the El Niiio period but not to the same extent as coastal stations. Esmeraldas and Quito recorded 1,766 mm and 1,983 mm of precipitation respectively in 1983, compared to mean annual totals of 760 mm and 1,553 mm. The amount of rain falling in 1982-83 in northern and central Peru was remarkable as well. Piura, with a mean annual precipitation of 50 mm, received more than 1,200 mm between December 1982 and April 1983 and over 2,000 mm for all of 1983 (Mujica, 1984; Hansen, this volume). Likewise, Tumbes with 1,537 mm, Chiclayo with 211 mm, Trujillo with 8.8 mm and Chimbote with 32.4 mm, all received levels of precipitation several times greater between December 1982 and April 1983 than 20-30 year averages. To the south, actual recorded rainfall did not reflect the increased duration and density of the fogs.
6 BOTANICAL RESPONSE TO 1982-83 EL NINO EVENT 6.1 Coastal Peru and Northern Chile
The effects of the increased available moisture from the 1982-83 El Nirio event were quite noticeable along coastal Peru and to a lesser extent, northern Chile. While no quantitative data are available, the increase in available moisture did produce luxuriant carpets of vegetation not present in "normal" years. Conversations with botanists and local residents indicated that the effects of the 1982-83 event were extraordinary, and most compared it to the conditions witnessed in 1925. The vegetation along the entire Peruvian coast was surveyed by the senior author during January-February and again October-December of 1983. Population sizes for all annual species in the lomas communities were extremely high with ground cover nearly complete in many areas. Perennial species also flowered abundantly, especially later in the season. At Cerro Campana (Fig. 3), for example, of the 168 species known from previous collections, over 90% were in flowering and/or fruiting condition between January and October 1983 (Sagbtegui et al., 1988). Similar levels of diversity and abundance were found in the other lomas formations during this period, including the
496
Fig. 3. Lomas de Cerro Campana, photograph from 600 m elevation, facing the Pacific Ocean (4 January 1983). Upper slopes support dense vegetation, while the lower slopes near sea level are nearly devoid of vegetation. Lomas de Lachay (120 spp.), Lomas de Atiquipa (146 spp.), Lomas de Mollendo (122 spp.), Lomas de 110 (60 spp.), and Lomas de Sama y Tacna (144 spp.). In most communities, ground cover was solid with mixed populations from a wide variety of families, but with a predominance of representatives from the Asteraceae, Boraginaceae, Gramineae, Leguminosae, Nolanaceae, Malvaceae, and Solanaceae. Community composition changed over time, especially from January to October of 1983. Some formations were dominated sequentially by various annual species. For example, vast portions of the Lomas de Lachay were initially covered with dense stands of Loasa wens Jacq. and later Nicotianapaniculata L. Large areas of the Lomas de Carnank were covered by grasses, predominately the endemic, Eragrostis peruviana (Jacq.) Trin. At the Lomas de 110, the ground was nearly covered with annual herbs, including Palaua weberbauen Ulbr., Nolana conjinis Johnst., and N. spathulata R. & P. At other sites, such as Lomas de Cachendo and Tacna, other annual herbs predominated, notably Tiquilia litoralis (Phil.) Richardson, Solanurn rnultiwurn Lam. and
497
Argyliu rudiutu (L.) D. Don.
The lomas formations were utilized in 1983 as a grazing resource for livestock brought in from the Sierra. Cattle, sheep and goats were driven on foot or by truck to nearly all formations, and their grazing was often quite destructive. At the Lomas de Camanfi, the grazing pattern of many of the goats and sheep was obvious on the hillsides (Fig. 4). Other formations, including the Lomas de Chfippara, were nearly denuded by November of 1983 and only unpalatable Grindeliu glutinosa (Cav.) Dunal and/or Nicotiunu puniculatu L. were left untouched. In one portion of the Lomas de Tacna, both corn and wheat were cultivated successfully, indicating the presence of enough available moisture for agriculture in these areas. It should be noted that at these times, the opportunities for the introduction of seeds from Andean sources and weedy species associated with agriculture are greatest. During field studies throughout the Peruvian coast during the subsequent years of 1984-86, the lomas formations were found to be either quite depauperate or in some
Fig. 4. Lomas de Caman& photograph illustrating the damage created by grazing sheep and goats (5 November 1983).
498
cases completely devoid of vegetation. The dried remains of annuals remained relatively unchanged for three years following the bloom. For example, in October of 1983, the Lomas de Mejia contained over 40 flowering species (Fig. 5, A), including Palaua velutina Ulbr. & Hill, Eragrostis peruviana (Jacq.) Trin., Portulaca pilosissima Hook., Weberbauerella brongniartioides Ulbr., and CriStaria multiwa Cav. However in subsequent years, species diversity and density were considerably reduced. In November 1986, only four species were in bloom and many dried remains of the 1983 bloom were still evident (Fig. 5, B). Only in northern and central Peruvian localities were perennial
herbs and woody vegetation found in flower during 1984, 1985 and 1986. The authors did not make direct observations in northern Chile during 1983, however, Prenafeta (1984) indicated that plant communities were effected by the increased cloudiness and available moisture. An indication of how these communities respond to ENS0 events was witnessed after a rare rainstorm struck northern Chile in July 1987, as a low pressure system moved north from Central Chile. While 1987 was characterized as a moderately strong El Nifio year, it is not clear if this storm was connected to that event. By late September 1987, a spectacular bloom was underway throughout the area that received rains, principally between Caldera and El Cobre (cf. Fig. 1). Annual herb communities were dominated by Nolana elegans (Gaud.) Johnst.,
N. aplocatyoides (Gaud.) Johnst., Calandnnia cymosa Phil., and Tetragonia ovata Phil., among others. Many perennials also responded with increased blooming of such endemics as Oxalis gigantea Barn., Euphorbia lactifIua Phil., Balbisia pedunculata (Lindl.) D. Don, and various Heliotropium species. In particular, the areas immediately north of Paposo were quite well developed with vegetation (Fig. 6). When these localities were visited again in September-December 1988, the diversity of annuals was much reduced, but perennials were nearly all in blooming condition. By July of 1989, the same areas were nearly devoid of plants in flower. 6.2 Galhpagos Islands
It is relevant to mention here the effects of the heavy rains on the GalBpagos Islands during the 1982-83 event. These islands are separated from the mainland by some 500 miles, have an arid climate, and share floristic affinities with western South America. The unusual quantity of precipitation in 1983 caused dramatic, and potentially long-term, changes in vegetation structure. These changes, described in some detail by Hamann (1985) and Luong and Torro (1985), were particularly spectacular in the arid coastal communities that are normally restricted by water availability. A diverse assemblage of herbaceous species became conspicuous in these communities, most notably an assemblage of sedges, herbs and vines such as Cypencr elegans subsp. rubiginosus (Hook.
499
Fig. 5. Lomas de Mejia, (A) photograph of vegetation taken on 25 October 1983, with nearly all plant species either blooming or in fruit. (B) Photograph taken at the same locality on 17 November 1986, with few living individuals and dried remains dating from the 1983 bloom.
500
Fig. 6. Photograph of coastal lomas vegetation north of Paposo, taken from lower slope at ca. 50 m; clouds are covering upper slopes at ca. 900 meters (14 October 1987).
f.) Eliass., Desmodium procumbens (Mill.) Hitchc., Chlork virgata Sw., Merremia aegyptica (L.) Urban, and Ludwigia erecfa (L.) Hara. Seedlings of woody species such as Croton
scoulen Hook. f., Lantana peduncularis Anderss. and Scalesia hellen Robins. became established in great numbers in these same communities. The abundance of these seedlings demonstrated that soil seed reserves had been plentiful prior to the rains and these were subsequently restored by heavy fruit set from the reproducing individuals. Not all arid zone perennials responded favorably to the torrential rains. Waterlogged soils and saturated water storage tissues caused the death of large numbers of two arborescent cacti, Opuntia echios Howell and Jusminocereus thouursii (Weber) Backbg., and to a lesser degree Bursera grmeolens (HBK.) Trian. & Planch. Seedlings of these species were also killed. Despite this dramatic decline in populations of these community dominants, Hamann (1985) suggested that increased shrub establishment may provide favorable microsites for successful establishment of these species in the future. Trees of
Elythrina velufina Willd. and Crofon scoulen' Hook. f. fell on several islands (Luong and Toro, 1985). Waterlogged soils appear to have been a major factor in some more mesic
50 1
communities as well. Many trees of Sculesiu pedunculutu Hook. f. were killed by these conditions, while other tree species such as Psidium gulupegium Hook. f., Zunthoxylum
fugwu (L.) Sarg. and Phoniu floribunda Hook. f. did not appear to be affected (Hamann, 1985). While the decline in numbers of S. pedunculufa Hook. f. was significant, there is evidence to suggest this species regenerates very well after extended drought periods that open up the forest understory (Kastdalen, 1982), and thus there may be a cyclical dynamic to stand structure. Plant distribution in the Galhpagos Islands was also strongly influenced by the 1982-
83 El Nifio event. Many species normally restricted to more mesic upland communities were abundant in the coastal arid zone (Luong and Toro, 1985). A number of weedy species also expanded their range significantly in this same arid zone, particularly near the littoral where high tides deposited drifting seeds in favorable microsites (Luong and Toro, 1985). Lastly, in addition to the impact on vascular plants, the high tides and torrential rains associated with the El Nifio conditions had damaging effects upon many lichens and bryophyte populations (Weber and Beck, 1985). None of the destructive effects mentioned here were witnessed in mainland communities.
7 CONCLUSIONS The lomas formations owe their existence to a complex set of physical and meteorological conditions found only along the western coast of South America. They occur as spatially discrete plant communities, separated by tracts of otherwise completely barren desert. Each lomas formation is unique in its species composition and typically contains an array of annuals, herbaceous perennials and woody plants. The overall plant diversity and density from year-to-year are variable, depending upon available moisture. El Niiio events, which stimulate higher moisture levels, have important influences upon reproductive cycles, dispersal events, and potential for colonizations.
As the number of reproducing individuals becomes greater during El Niiio events, opportunities exist for populations to temporarily expand and merge in response to favorable conditions. Subsequently, in the absence of sufficient moisture, they become re-isolated by unfavorable habitats. Increased moisture availability also leads to high seed productivity allowing the replenishment of seed banks, a necessity for annuals within unpredictable environments. It is also during these periods that seedling establishment can occur in herbaceous and woody perennials. Whenever a majority of the population is brought into flowering condition simultaneously, it allows the maximum potential re-shuffling of genetic material and subsequent selection of 20-50 year periods of drought. As mentioned previously, the possibility for the introduction of seeds from montane
502
plant formations is especially great during El Niiio events. The movement of livestock from the Sierra at these times allows for transfer of seeds and propagules. These plants can become established under optimal conditions and persist in suitable micro-habitats. How plants colonized these formations is undoubtedly related to periods of increased moisture availability. The fact that the lomas formations contain such high levels of endemic species suggests that isolation has been effective. The dynamics of endemism within the lomas archipelago is potentially related to rare episodes of El Niiio rainfall providing possible "bridging conditions" for populations to expand their normally restricted distributions and then become isolated during intervening dry periods. Should subsequent selection fix differences between populations, the isolation afforded by the surrounding desert can effectively maintain these differences. Only when communities merge is this isolation disturbed, either during optimal periods of El Niiio rainfall or longer-term merging in response to eustatic change during glacial cycles.
A greater understanding of the processes that shape the lomas formations is only possible when they are viewed in the context of models for the geological and climatic dynamics. The 1982-83 El Niiio event provided an opportunity to witness the climatic effects likely responsible for establishing and shaping these unique communities. 8 ACKNOWLEDGMENTS
We wish to thank M. Moseley and M. Ono for valuable discussions and exchange of ideas and J. Beach and T. Duncan for help with floristic data analysis. P. Glynn and anonymous reviewers provided useful comments and these are greatfully acknowledged. Field studies by the senior author were supported, in part, by grants from the National Geographic Society (2706-83) and National Science Foundation (BSR-8513205). The junior author was supported by grants from the National Science Foundation and the Office of Health and Environmental Research of the U.S. Department of Energy. Many people extended hospitality or otherwise helped with field work in South America, including V. Ascencio, F. Cabiesas, E. Carrillo, E. Cerrate, M. Chanco, C. Dim, R. Ferreyra, L. Gonzhles, S. Leiva, A. Mpez, C. Marticorena, U. Molau, G. Montenegro, J. Mostacero, M. Muiioz, B. Palma, R. Perucci, V. Poblete, M. Quezada, J. Santisteban, A. Saghtegui, N. Tay, S. Teillier, J. de la Torre, M. Villarroel, and L. Watanabe. C. Simpson-Richardson is greatfully acknowledged for her help with the illustrations. Diane Dillon is given special thanks by the senior author for valuable assistance in plant and insect collecting during the 1983 and 1988 field seasons and for help in editing and proofing the final manuscript.
503
9 REFERENCES Alpers, C.N. and Brimhall, G.H., 1988. Middle Miocene climatic change in the Atacama Desert, northern Chile: evidence from supergene mineralization at La Escondida. Geol. SOC.h e r . Bull. 100(10): 1,640-1,656. Arm, W. E., Landa, A, and Tarazona, J., (eds.), 1985. "El Niiio", Su Impact0 en la Fauna Marina. Bol. Inst. Mar Perk 222 pp. Barber, R. T. and ChAvez, F. P., 1983. Biological consequences of El Niiio. Science, 222: 1,203-1,210. Beebe, W., 1926. The Arctuns Adventure. G. P. Putnam, New York, 439 pp. Ellenberg, H., 1959. Uber den Wasserhaushalt tropischer Nebeloasen an der Kiistenwiiste Perus. Ber. Geobot. Forsch. Inst. Riibel 1958: 47-74. Glynn, P.W., 1988. El Niiio-Southern Oscillation 1982-83: Nearshore population, community and ecosystem responses. Annu. Rev. Ecol. Syst., 19: 309-345. Goudie, A. and Wilkinson, J., 1977. The Warm Desert Environment. Cambridge Univ. Press, Cambridge, 88 pp. Hamann, O., 1985. El Niiio influence on the GalApagos vegetation. In: G. Robinson and E. M. del Pino (Editors), El Nifio en las Islas GalApagos: El Evento de 1982-83. Charles Darwin Foundation for the GalApagos Islands, Quito, pp. 299-330. Johnson, A. M., 1976. The climate of Peru, Bolivia and Ecuador. In: W. Schwerdtfeger (Editor), Climates of Central and South America. World Survey of Climatology, 12: 147-218. Johnston, I.M., 1929. Papers on the flora of northern Chile. Contrib. Gray Herb., 4: 1-172. Kastdalen, A., 1982. Changes in the biology of Santa Cruz Island between 1935 and & 1965. Noticias de GalApagos, 35: 7-12. Knoch, K., 1930. Klimakunde von Sudamerika. In: W. Koppen and R. Geiger (Editors), Handbuch der Klimatologie, Vol. 2, Part G. Borntraeger, Berlin. Leighton, M. and Wirawan, N., 1986. Catastrophic drought and fire in Borneo tropical rain forest associated with the 1982-83 El Niiio Southern Oscillation event. In: G. T. Prance (Editor), Tropical Rain Forest and the World Atmosphere. Westview Press. Boulder, CO, pp. 75-102. Luong, T.T. and Torro, B., 1985. Cambios en la vegetaci6n de las Mas Galapagos durante "El Niiio" 1982-83. In: G. Robinson and E. M. del Pino (Editors), El Niiio en las Islas GalApagos: El Evento de 1982-83. Charles Darwin Foundation for the GalApagos Islands, Quito, pp. 331-342. Miller, A., 1976. The climate of Chile. In: W. Schwerdtfeger (Editor), Climates of Central and South America. World Survey of Climatology, 12: 113-145. Moseley, M.E., Feldman, R.A. and Ortloff, C.R., 1981. Living with crisis: human perception of process and time. In: M. Nitecki (Editor), Biotic Crises in Ecological and Evolutionary Time. Academic Press, New York, pp. 231-267. Mujica, R., 1984. Departamento de Piura rainfall in 1983. Trop. Ocean-Atmos. Newsl. 28. Murphy, R.C., 1926. Oceanic and climatic phenomena along the west coast of South America during 1925. Geog. Rev., 16: 26-54. Naranjo, P., 1985. El fenomeno El Niiio y sus efectos en el clima del Ecuador. In: G. Robinson and E. M. del Pino (Editors), El Niiio en las Islas GalApagos: El Evento de 1982-83. Charles Darwin Foundation for the Galapagos Islands, Quito, pp. 3-27. Nials, F.L, Deeds, E.E., Moseley, M.E., Pozorski, S.G., Pozorski, T.G. and Feldman, R.A., 1979. El Niiio: the catastrophic flooding of Peru. Field Mus. Nat. Hist. Bull., 50(7): 4-14. (Pt. I) and 50(8): 4-10 (Pt. 11). Oka, S., 1986. On a trial measurement of the moisture in fog on Lama Ancon--in relation to an investigation into conditions required for development of lomas communities. In: M. Ono (Editor), Taxonomic and Ecological Studies on the Lomas Vegetation in the Pacific Coast of Peru. Makino Herbarium, Tokyo
504
Metropolitan University, pp. 41-51. Prenafeta, S., 1984. El paraiso se mudo de casa. Creces, 5(1/2): 12-14. Prohaska, F., 1973. New evidence on the climatic controls along the Peruvian coast. In: D.H.K. Amiran and A.W. Wilson (Editors), Coastal Deserts, their Natural and Human Environments. Univ. Arizona Press, Tucson, pp. 91-107. Quinn, W.H. and Neal, V.T., 1983. Long-term variations in the southern oscillation, El Niiio, and Chilean subtropical rainfall. Fish. Bull., 81: 363-374. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E., 1987. El Niiio occurrences over the past four and a half centuries. J. Geophys. Res., 92(C13): 14,449-14,461. Quinn, W. H., Zopf, O., Short, K. S. and Kuo Yang, R. T. W., 1978. Historical trends and statistics of the Southern Oscillation, El Nifio, and Indonesian droughts. Fish. Bull., 76: 663-678. Richardson, J.B., 1981. Modeling the development of sedentary maritime economies on the coast of Peru: A preliminary statement. Ann. Carnegie Mus., 5 0 139-150. Robinson, G. and del Pino, E.M., (eds.), 1985. El Niiio en las Islas GalApagos: El Evento de 1982-83. Charles Darwin Foundation for the Galtipagos Islands, Quito, 534 pp. Rollins, H.B., Richardson, J.B. and Sandweiss, D.H., 1986. The birth of El Niiio: Geoarchaeological evidence and implications. Geoarchaeology, 1: 3-15. Romero, H. and Garrido, A.M., 1985. Influencias geneticas del fenomeno El Niiio sobre 10s patrones climaticas de Chile. Invest. Pesq. (Chile), 32: 19-35. Rundel, P. W. and Mahu, M., 1976. Community structure and diversity in a coastal fog desert in northern Chile. Flora, 165: 493-505. Rundel, P.W., Dillon, M.O., Palma, B., Mooney, H.A., Gulmon, S.L. and Ehleringer, J.R., 1990. The phytogeography and ecology of the coastal Atacama and Peruvian deserts. Aliso, in press. Rutllant, J., 1985. Algunos aspectos de la influencia climatica, a nivel mundial y regional, del fenomeno El Nifio. Invest. Pesq. (Chile), 3 2 9-17. Saghtegui, A., Mostacero, J. and Upez, S., 1988. Fitoecologfa del Cerro Campana. Bol. SOC.Bot. La Libertad, 14: 1-47. Schott, G., 1931. Der Peru-Strom und seine nordlichen Nachbargebiete in normaler und anormaler Asbildung. Ann. Hydrogr. Mar.Meterol., 59: 161-169. Simpson, B.B., 1975a. Glacial climates in the eastern tropical South Pacific. Nature, 253: 34-36. Simpson, B.B., 1975b. Pleistocene changes in the flora of the high tropical Andes. Paleobiology, 1: 273-294. Thompson, L. G., Hastenrath, S. and Amao, B.M., 1979. Climatic ice core records from the tropical Quelccaya ice cap. Science, 203: 1,240-1,243. Trewartha, G.T., 1961. The Earth’s Problem Climates. Univ. Wisc. Press, Madison, 334
PP.
Weber, W.A. and Beck, H.T., 1985. Effects on cryptogamic vegetation (lichens, mosses, and liverworts). In: G. Robinson and E. M. del Pino (Editors), El Niiio en las Islas Galtipagos: El Evento de 1982-83. Charles Darwin Foundation for the Galtipagos Islands, Quito, pp. 343-363. Wooster, W.S., 1980. Early observations and investigations of El Niiio: the event of 1925. In: M. Sears and D. Merriman (Editors), Oceanography: The Past. Springer-Verlag, New York, pp. 629-641. Wooster, W.S. and Fluharty, D.L., (eds.), 1985. El Niiio North, Nifio Effects in the Eastern Subarctic Pacific Ocean. Seattle: Wash. Sea Grant Program. 312 pp.
505
AN ECOLOGICAL CRISIS IN AN EVOLUTIONARY CONTEXT: EL NIfiO IN THE EASTERN PACIFIC
GEERAT J. VERMEIJ, Department of Geology, University of California, Davis, CA 95616 ABSTRACT Vermeij, G.J. 1989. An ecological crisis in an evolutionary context: El Nifio in the eastern Pacific. In this essay I ask whether and how El Nifio-SouthernOscillation (ENSO) events have influenced extinction and speciation among marine organisms in the tropical eastern Pacific during the Pliocene and Pleistocene, and whether events of this type could serve as small-scalemodels for the greater biotic crises that have marked the history of life from time to time. The warming and nutrient depletion that characterize the eastern Pacific during ENSO events are both too frequent and on too small a scale to result in species extinctions. Although a reduction in primary productivity has been implicated in many of the biotic crises of the past, warming has apparently not played a significant role. These arguments, together with empirical evidence of low rates of extinction among eastern Pacific invertebrates during the past three million years, suggest that ENSO events are not good models for mass extinctions. Similarly, the changes in oceanic circulation associated with ENSO events are too frequent to have resulted in speciation. This conclusion is supported by the general lack of divergence of eastern Pacific populations that have dispersed from the central and western Pacific since the Pliocene.
1 INTRODUCTION Any circumstance that kills large numbers of organisms has the potential for bringing about evolutionary consequences. What these consequences are depends on at least four attributes of the circumstance in question: (1) the way in which it departs from normal tonditions; ( 2 ) its frequency of occurrence; ( 3 ) its severity; and ( 4 ) its geographical extent. There are three possible consequences: (1) a change in the relative importance of agencies effecting natural selection on surviving individuals; ( 2 ) extinction of populations; and ( 3 ) the splintering or founding of populations which, given appropriate subsequent conditions, may form new species. As summarized in other contributions to this volume and by Glynn (1988), the principal physical changes effected by El Nifio-SouthernOscillation (ENSO) events relative to normal conditions in the central and eastern Pacific are as follows: (1) higher temperatures on the eastern margin of the temperate and tropical Pacific; ( 2 ) a drastic (ten-fold) reduction in primary planktonic productivity associated with the absence of usual upwelling in much of the eastern Pacific; ( 3 ) increased incidence of severe storms along temperate coasts of the eastern Pacific; ( 4 ) a reduction or reversal in flow of the westward-flowing Peru Current; and (5) increased flow of the eastward-flowing
506 North Equatorial Countercurrent across the Pacific. These changes brought about widespread mortality at all trophic levels in the eastern and central Pacific, and may have enabled propagules of some western Pacific species to colonize the eastern Pacific. Have events of the ENSO type been significant in evolution? This question may be approached at several levels. At the ecological time-scale (years to decades), we may ask how crises such as ENSO affect patterns of natural selection, gene flow, extinction of local populations, and recolonization of affected habitats. Interesting as these questions are, I can find no data bearing on their answers in the literature on the effects of ENSO on marine organisms; we simply know too little about the genetic structure of eastern Pacific populations to make intelligent guesses about how gene flow and regimes of selection might be altered by the great climatic variations associated with ENSO events. Consequently, I am forced to circumvent this topic, much as I should have liked to have done otherwise. At the longer time-scale of speciation and extinction, we may ask whether ENSO events provide opportunities for large-scale evolutionary change, and how the fossil record of the appearance and disappearance of species in the eastern Pacific, where the effects of El Nifio are especially strong, differs from that in the adjacent western Atlantic, where physical effects of ENSO are apparently much less severe. Here, too, there is no overabundance of empirical evidence, but it is at the level of extinction and speciation to which I shall devote myself in this essay. Circumstances that bring about extinction, speciation, and immigration clearly are not everyday occurrences. Because they are beyond the experience of most present-day observers, it is important to ask whether hugely scaled-up versions of ENSO-like disturbances can serve as models for the circumstances that brought about the biological upheavals observable in the fossil record. 2 ENSO AS A MODEL FOR EXTINCTION EVENTS
The catastrophic mortality of marine organisms and of terrestrial animals that are ecologically tied to the marine food web along the coasts of Peru and Ecuador suggests either or both of two principal causes during El Nifio years, namely, high temperatures and depletion of normally abundant nutrients. High temperatures have also been implicated in the widespread mortality of corals in the tropical eastern Pacific (Glynn, 1985a,b, 1988, and this volume), and nutrient loss is believed to be responsible for reproductive failure of seabirds in the central Pacific and Galapagos (see Duffy and Smith, this volume). Would these circumstances, if widespread enough, bring about extinction? Consider first the case of high temperatures. To the best of my knowledge, no
507 one has suggested warming as a cause of mass extinction in the fossil record. Several models suggest that warming may have taken place at the end of the Cretaceous, but proponents of these scenarios (Mchan, 1978; Rampino and Volk, 1988) believe that the rise in temperature came after (and therefore did not cause) the mass extinction of phytoplankters that characterizes the endCretaceous crisis. On the one hand, Mchan's (1978) hypothesis is that the carbon dioxide that would have been taken up during normal times in photosynthesis by phytoplankton instead entered the atmosphere after the extinction. In the atmosphere, the carbon dioxide contributed to a greenhouse effect with the consequence that temperatures increased. Rampino and Volk (1988), on the other hand, point to the connection between the release of dimethylsulphides by oceanic phytoplankton (Charlson et al., 1987) and the creation of condensation nuclei in clouds, which in turn cause precipitation and keep temperatures down. With the decline of the phytoplankton at the end
of the Cretaceous (and probably during other biotic crises as well), this biological control on temperature was temporarily disrupted, allowing temperatures to rise. In his analysis of crinoid extinctions at the end of the Middle Ordovician, Eckert (1988) suggested that a rise in sea level coupled with the resulting depletion of oxygen in quiet bottom waters of shallow seas covering large expanses of tropical continental shelves was responsible for the impoverishment of the rich crinoid fauna of North America. Anoxia has frequently also been invoked by Hallam (1976, 1977, 1982, 1986, 1987) to explain regional extinctions of pelecypods in Europe, especially during the Early Jurassic. Rising sea levels are often associated with climatic warming, reduced latitudinal thermal contrasts, more sluggish oceanic circulation, and therefore with depletion of oxygen in deep waters (Fischer and Arthur, 1977; Hallock, 1987). In these interpretations, it is the anoxia rather than high temperature that is believed to cause extinction. Even then, only those species that are confined to waters susceptible to deoxygenation are at risk; those whose ranges extend to waters agitated by waves and currents would remain unaffected. Others have concluded that many of the great extinction crises were brought on by lower rather than by higher temperatures. Strong evidence for this conclusion, ably summarized by Stanley (1984, 1988), is that the magnitude of extinction among marine organisms was generally much higher at tropical latitudes than in the biotas of colder oceans at high latitudes. The sharp reductions in the reef biotas of the Late Ordovician and Late Devonian, for example, have been linked to oceanic refrigeration at low latitudes (Copper, 1977, 1986; Fagerstrom, 1983; Stanley, 1984, 1988). This cooling is believed to be a consequence of continental suturing, which interrupted east-to-west circulation in the tropical ocean and forced a greater exchange of oceanic
508
waters between the polar regions and the tropics. Cooling during the Neogene (especially during the last 3.5 million years) has also been associated with extinctions and with changes in circulation. As the Central American isthmus closed and the Bering Strait opened about 3.5 million years ago, east-west circulation at low latitudes was interrupted, and the currents bringing warm water toward higher latitudes and cool water toward lower latitudes intensified (Weyl, 1968; Kaneps, 1979; Herman and Hopkins, 1980; Kennett, 1983; Keigwin, 1986; Stanley, 1986). Evidence from biogeography, paleoceanography, and the biology of survivors suggests that cooling was also a significant cause of extinction during the Late Eocene (Corliss, 1979; Keigwin 1980; Corliss et al., 1984; Hansen, 1987).
The death of reefs off the Pacific coast of northwestern
Costa Rica during the seventeenth century was attributed by Glynn et al. (1983) to intense upwelling of cool water during what has come to be known as the Little Ice Age. Besides cold, a reduction in primary productivity (and therefore of food supply for animals at higher trophic levels) is thought to have played a significant role in several of the major Phanerozoic extinction events. Data from carbon-isotope ratios in marine sediments seem to imply a sharp reduction in phytoplankton productivity at the end of the Cretaceous (Hsu et al., 1982; Zachos and Arthur, 1986; Arthur et al., 1987).
The fact that many sediments
marking the stratigraphical boundary between the Maastrichtian stage of the Late Cretaceous and the Paleocene stage of the Paleogene consist of calciumpoor clay rather than calcium-rich oozes in deep-sea settings is consistent with this interpretation,because the presence of large amounts of calcium carbonate before and after the crisis indicates the presence of numerous photosynthesizing skeletonized phytoplankters and other pelagic organisms (Percival and Fischer, 1977).
A reduction in primary productivity is
consistent with most of the proposed underlying physical causes of the endCretaceous crisis
--
large-scale volcanism, the impact of a large asteroid,
darkening of the sky following gigantic fires - - and would have affected terrestrial as well as marine communities. Hsu (1986) has made a plausible if speculative case for the assertion that events marking the ends of geological eras (especially the Paleozoic and Mesozoic, but also the period immediately preceding the Cambrian explosion) were characterized by profound reductions in primary productivity. In our analysis of patterns of extinction and survival among tropical molluscs following the uplift of the Central American isthmus during the Pliocene, we suggested that a reduction in phytoplankton productivity in the western Atlantic (but not the eastern Pacific) accounts in part for the observation that the magnitude of extinction in the Atlantic was about twice that in the eastern Pacific (Vermeij and Petuch, 1986).
This inference was
509 supported by the observation that many molluscs which during the Pliocene had extensive geographical ranges encompassing parts of both the Atlantic and Pacific in tropical America have since become restricted to the shores of the eastern Pacific, northern South America, and Brazil, areas all characterized by intense nutrient-rich upwelling or by extensive terrestrial runoff of nutrients from the land. Moreover, Keigwin (1982) found that bottom waters in the western Atlantic changed isotopically in a manner suggesting lower primary productivity since the Central American uplift, whereas eastern Pacific bottom waters did not (see also Gartner et al., 1987, for a similar conclusion based on analyses of radiolarian assemblages). Even in the tropical western Pacific, where the magnitude of extinction since the Pliocene has been very small, many previously more widespread molluscs have contracted their ranges to the coasts of high islands and continents, where phytoplankton productivity is high (Vermeij, 1986, 1987b). Just how nutrient depletion brings about extinction is not fully understood. One suggestion, based on the effects of El Niiio events in Peru, is that the death and decay of vast numbers of fish and other animals following a spell of unusually high temperature causes the bacteria in the water to deplete oxygen levels, making the waters anoxic and therefore inimical to aerobic life. Moreover, toxins may be produced which could kill organisms that would otherwise be able to tolerate periods of anoxia. If anoxic waters well up after the El Niiio event is over, organisms that escaped the direct effects might still be killed. This scenario is believed by Wilde and Berry (1984) to explain at least some of the great extinction crises, which these authors see as ENS0 events on a gargantuan scale. Another possibility is that a reduction or interruption of primary productivity will quickly result in the death of animals whose high metabolic rates require a high food intake (Vermeij, 1987a). Single-celled organisms with high rates of photosynthesis, such as diatoms, would also be severely affected unless they had inert resting stages (Hallock, 1987). Only species with low metabolic requirements or those that depend on foods whose supply is not directly tied to primary production regimes (such as some deposit-feeders) would be able to tolerate a reduction or temporary cessation in primary productivity (Sheehan and Hansen, 1986). Cooling, darkening, and any number of other causes could be responsible for a reduction in photosynthesis. Whichever cause is responsible, organisms most likely to be affected are those with the greatest potential for competitive and defensive superiority (Vermeij, 1987a). The opposite interpretation - - extinction episodes are correlated with
--
higher primary productivity has been invoked to explain the demise of coral reefs, and especially of the dominant photosynthesizing reef-building animals
510
and coralline algae (Hallock and Schlager, 1986; Hallock, 1987, 1988).
Reef-
builders with photosynthesizing microalgae in their tissues prevail competitively over nonphotosynthesizing invertebrates such as barnacles, hydroids, and pelecypods in clear, warm, nutrient-poor waters, but they do not do well when nutrient concentrations are high (Birkeland, 1977).
Moreover,
bioeroding organisms that destroy the skeletal framework of reefs achieve much higher densities and have a greater effect in communities bathed by nutrientrich waters than in those living in areas of nutrient-poor water (see also Highsmith, 1980; Glynn et al., 1983).
A possible resolution to the conflict
between these interpretations may lie in the difference in scale between the various causes of death. In the modern tropical ocean, eutrophic conditions (that is, abundant nutrient supplies) are localized to a few areas. It is striking, for example, that reef-building corals and coralline algae persist in the generally productive eastern Pacific. In this region, areas of intense upwelling sufficient to kill reef-builders on a large scale or to prevent their successful settlement are confined to the Gulf of Panama, the Gulf of Tehuantepec, and northwestern Costa Rica. Adjacent waters, though still relatively rich in nutrients, nevertheless support reef-builders and sometimes even reefs (Glynn and Wellington, 1983). Similarly in the Indo-West-Pacific region, coral reefs of extraordinarily high diversity and great complexity thrive in the waters of the Indo-Malayanarchipelago, where nutrient inputs may be locally very high. In contrast to the localized nature of high nutrient concentrations, cold spells of the kind that killed corals in Florida during the late 1970s (Porter et al., 1982) and the sudden nutrient depletion brought on by ENSO events occur on a much larger scale, with the result that there are few local refuges for the species that are adversely affected by these changes. In short, although these disturbances still are not on a scale sufficient to cause species extinctions in the present day, their effects are more likely to bring about extinctions than are the apparently much smaller-scale episodes of nutrient enrichment. Can ENSO events serve as models for the major extinction crises? I think the data at hand are insufficient to provide a clear answer. At the very least, however, we should resist the temptation to regard the great crises as merely large-scale versions of ENSO events. In particular, the warming that apparently killed large numbers of eastern Pacific corals during the 1982-83 ENSO event did
so
without anoxia, and warming by itself has not been invoked to
explain the demise of ancient reefs. Nutrient depletion and anoxia of the kind observed during ENSO events in Peru may have been important causes of ancient extinctions, but external stimuli other than ENSO events could have brought on such conditions. A still unanswered question is whether ENSO events could occur on a geological scale large enough to encompass the ranges of entire
51 1
species, even within the confines of the eastern Pacific. If even the most dramatic ENSO events are confined to the central and eastern Pacific, most of the world's marine biota would remain unaffected, with the result that the ensuing extinction would, like Hallam's (1986) Jurassic events, be a regional rather than a global crisis. 3 EXTINCTION IN THE EASTERN PACIFIC
Despite the likelihood that ENSO events are not good models for ancient biotic crises, they could still have resulted in widespread extinctions of species in the eastern Pacific, and they may continue to pose a threat in the future. It is difficult to assess the impact of ENSO events on regional extinctions in the eastern Pacific because we do not know how long ENSO events have affected this part of the world.
If the Central American uplift
interrupted east-west circulation in the tropical ocean, it may have intensified ENSO events by exacerbating the tendency for waters to move meridionally rather than parallel to the equator. Whatever the time of origin of ENSO events, the distribution of oceans and continents has been sufficiently uniform since the mid-Pliocene that we may assume ENSO events to have struck the eastern Pacific with regularity from that time to the present (see Colgan, this volume). Generally, the Pliocene and Pleistocene history of the tropical eastern Pacific following the mid-Pliocene Central American uplift suggests that the magnitude of extinction was relatively small, and therefore indirectly that ENSO events have been small enough to decimate populations but not to eliminate entire species. Some coral species with restricted distributions went extinct locally, e.g. Milleuora ulatvuhvlla in Panama and Acrouora valida in Colombia (see Glynn, 1988). In our study of patterns of extinction among tropical American molluscan subgenera, we found that the magnitude of extinction in the eastern Pacific
(15%) was less than half that in the adjacent western Atlantic (32%) (Vermeij and Petuch, 1986).
In fact, the eastern Pacific has served as an important
biogeographical refuge for a large number of taxa whose distribution during the Pliocene encompassed both the western Atlantic and eastern Pacific (Woodring, 1966; Vermeij, 1978, 1986; Vermeij and Petuch, 1986).
Nothing is known about
the causes or times of extinction of the eastern Pacific subgenera that did perish during the Late Pliocene and Pleistocene. A role for ENSO events obviously cannot be excluded, but whatever this role might have been, its effects were small compared to the devastations visited upon the marine biotas of the tropical and warm-temperate Atlantic (see also Stanley, 1986) where ENSO effects today are small.
512
Additional evidence for the relatively low magnitude of extinction of Pleistocene eastern Pacific invertebrates comes from recent studies of fossils from the Galapagos Islands. Zullo (1986) recovered five rock-dwelling Pleistocene balanomorph barnacles in the Galapagos; two of these, though locally extinct in the islands, still survive on the mainland coast of tropical west America. Of twenty Late Pliocene or Pleistocene molluscs from tuff cones on Isla Santa Cruz in the Galapagos, two (not identified to species) probably represent extinct taxa known only from the Galapagos, and six are found today only on the mainland of tropical western America (Pitt et al., 1986).
In
summary, although there have apparently been several cases of geographical range restriction (and therefore local extinction), relatively few species have become extinct in the Galapagos during the last three million years despite the likelihood that ENSO events have ravaged the islands frequently. 4 SPECIATION IN THE EASTERN PACIFIC
By changing the pattern and intensity of oceanic circulation, ENSO events could provide good opportunities for the establishment of founder populations which then could diverge and achieve the status of fully fledged species. However, just as the eastern Pacific seems to have been relatively free of extinctions of marine species since the mid-Pliocene, it has also been evolutionarily quiescent with respect to speciation. One indication of this comes from Cronin’s (1985, 1987) study of the tropical American ostracod genus puriana. Cronin showed that, whereas the Early Pliocene (4 to 3 million years before present) was a time of active speciation in the genus, especially in the western Atlantic, only one eastern Pacific species arose after this time. He suggests that the climatic transition from a relatively warm to a much cooler pattern during the Pliocene provided opportunities for speciation, whereas the relatively much more rapid and more frequent shifts in climate during the Late Pliocene and Pleistocene were too frequent (occurring approximately once every one hundred thousand years) to effect speciation. If this interpretation is supported by data from other groups, then even the most severe ENSO events might be too frequent to have lasting evolutionary consequences. If one of the oceanographical characteristics of ENSO events is to speed up eastward flow of the North Equatorial Countercurrent, it follows that propagules of Indo-West-Pacificspecies have a better chance of being established in the eastern Pacific during ENSO events than at other times (see Richmond, this volume). If conditions in the eastern Pacific differ from those further west, and if the propagules remain isolated for a sufficiently long period, the invading populations could undergo differentiation from their ancestors and attain the status of full-fledgedendemic species. The evidence
513 at hand suggests that this has not happened on a large scale. Among eastern
Pacific corals, most of which are believed to have arrived from the west (Porter, 1972; Glynn and Wellington, 1983), at most one (pocillouora capitata) can be regarded as an endemic derivative of a widespread Indo-West-Pacific species. The others are apparently also known from both the eastern Pacific and at least some island groups in the central Pacific (Wells in Glynn and Wellington, 1983). It may turn out that some of these corals are descended from previously more widespread species native to tropical America (Heck and McCoy, 1978) and that they therefore did not colonize the eastern Pacific from the west. If s o , the endemism of E caDitat8 and perhaps other species cannot be interpreted as the result of divergence from Indo-West-Pacificinvading stocks.
Among the more than sixty molluscs that have invaded the eastern Pacific from the west (Emerson, 1978; Vermeij, 1987b, and references therein), most are morphologically identical to widespread Indo-West-Pacificspecies. Only four eastern Pacific species (Polinices caurae. Conus dalli, C. nux, and Terebra maculata roosevelti) are interpretable as endemic offshoots from Indo-WestPacific immigrant stocks. Reid (1986) has shown that Littoraria schmitti and Littoraria pullata, described from Clipperton Island and Baja California respectively, are merely dark forms of the widespread primitive Indo-WestPacific Littoraria uintado, and not distinct species or subspecies endemic to west America as had been thought. In the same vein, electrophoretic comparisons between eastern Pacific and Indo-West-Pacificpopulations of fish species with a trans-Pacific distribution show little or no divergence of the eastern Pacific populations (Rosenblatt and Waples, 1986).
Trans-Pacific dispersal is apparently frequent enough to
prevent isolation and divergence of propagules in the eastern Pacific (Thresher and Brothers, 1985). In short, although a few stocks of Indo-West-Pacificorigin have apparently diverged in the tropical eastern Pacific, there have been very few clear-cut speciation events there. Again, the oceanographical shifts in circulation due to ENSO events may have been both too frequent and too ephemeral for significant evolution to have taken place. Indeed, it could be argued that, because climatic conditions in the easten Pacific vary greatly on a small timescale between "normal" regimes and extreme ENSO or extreme upwelling events, the species now living in the eastern Pacific are able to cope with considerable short-term changes, and that oceanographical shifts would have to be greater in extent and severity than elsewhere in the world in order to result in evolution of the kind observable in the fossil record. In fact, the extreme differences between El Nido and "normal" years may have brought about a certain evolutionary stability in the eastern Pacific.
514
Species unable to cope with the great variations were presumably exterminated when ENSO events first affected the eastern Pacific. Those that remained would, for the most part, have been able to persist during most ENSO events, with some populations (especially recently established propagules) perhaps becoming extinct. Because of the profound short-term variations, predictable and consistent regional differences in selective regimes would be swamped. By effecting great climatic swings, ENSO events thus may prevent regional differentiation of populations and thereby choke off opportunities for speciation. 5 CONCLUDING REMARKS There are far too many uncertainties about the evolutionary consequences of
ENSO events to draw firm conclusions. Several points are worth emphasizing, however. The first is that, despite widespread disruptions, ENSO events may be so
frequent that most species in the tropical eastern Pacific seem to persist
and to be able to recoup their losses. It may be that the devastations were far worse during the early history of ENSO events, when many species were presumably unable to cope with the periodic oceanographical disturbances. A similar argument has been made for the effects of glaciations. When the climate cooled precipitously for the first time during the mid-Pliocene in the northern hemisphere, large numbers of species became extinct in the Atlantic; but during subsequent cooling events, especially during the Pleistocene glaciations, most of the species or lineages that had survived through the earlier crises were evolutionarily "prepared" to cope with the new disturbances, and weathered these events as well (Stanley, 1984, 1986, 1988). The apparent immunity of species to extinction during ENSO events also puts into perspective the severity of the conditions that did result in widespread extinctions in the geological past. In the eastern Pacific at least, events of a scale and magnitude much greater than historical ENSO events or cooling events like the Little Ice Age would be required to bring about species extinctions. Similarly, the magnitude of events bringing about speciation must in general be greater than any of the changes that have been historically observed in the eastern Pacific, however profound these changes appear to be to us. Several interesting questions emerge from these considerations. What was the time course of Pliocene and Pleistocene extinction (including local extinction) in the eastern Pacific? When did divergences of species that formed during the Late Pliocene and Pleistocene in the eastern Pacific in fact take place? Did the events of speciation occur as several distinct clumps, or were individual divergences temporarily scattered and independent? What has been the recent history of extinction and speciation on the temperate west
515 coast of South America, where the effects of ENSO events are most dramatically displayed today? Research on fossils and on the phylogenetic relationships and inferred times of divergence of living species will go a long way toward answering these questions and toward placing our understanding of the evolutionary dimension of ENSO events on a firmer footing. 6 REFERENCES Arthur, M.A., Zachos, J.C. and Jones, D.S., 1987. Primary productivity and the Cretaceous/Tertiary boundary events in the oceans. Cretaceous Res., 8: 4354. Birkeland, C . , 1977. The importance of rate of biomass accumulation in early successional stages of benthic communities to the survival of coral recruits. Proc. Third. Int. Coral Reef Symp., Miami, 1: 15-21. Charlson, R.J., Lovelock, J.E., Andre, M.O. and Warren, S . G . , 1987. Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature, 326: 655-661. Copper, P., 1977. Paleolatitudes in the Devonian of Brazil and the FrasnianFamennian mass extinction. Palaeogeogr., Palaeoclimatol.,Palaeoecol., 21: 165-207. Copper, P., 1986. Frasnian/Famennian mass extinction and cold-water oceans. Geology, 14: 835-839. Corliss, B.H., 1979. Response of benthic deep-sea Foraminifera to development of the psychrosphere near the Eocene-Oligoceneboundary. Nature, 282: 6365. Corliss, B.H., Aubry, M.-P.,Berggren, W.A., Fenner, J.M., Keigwin, L.D. Jr. and Keller, G., 1984. The Eocene/Oligocene boundary event in the deep sea. Science, 226: 806-810. Cronin, T.M., 1985. Speciation and stasis in marine Ostracoda: climatic modulation of evolution. Science, 227: 60-63. Cronin, T.M., 1987. Evolution, biogeography, and systematics of Puriana: evolution and speciation in Ostracoda, 111. Paleont. SOC. Mem. 21, J. Paleont. 61 (suppl. to no. 3): 1-71. Eckert, J.D., 1988. Late Ordovician extinction of North American and British crinoids. Lethaia, 21: 147-167. Emerson, W.K., 1978. Mollusks with Indo-Pacific faunal affinities in the Eastern Pacific Ocean. Nautilus, 92: 91-96. Fagerstrom, J.A., 1983. Diversity, speciation, endemism and extinction in Devonian reef and level-bottom communities, eastern North America. Coral Reefs, 2: 65-70. Fischer, A.G. and Arthur, M.A., 1977. Secular variations in pelagic realm. SOC. Econ. Petrol. Mineral. Spec. Publ., 25: 19-50. Gartner, S . , Chow, J. and Stanton, R.J., 1987. Late Neogene paleoceanography of the eastern Caribbean, the Gulf of Mexico, and the eastern equatorial Pacific. Mar. Micropaleont., 12: 255-304. Glynn, P.W., 1985a. Corallivore population sizes and feeding effects following El Nitio (1982-1983) associated coral mortality in Panama. Proc. Fifth Int. Coral Reef Congr., Tahiti, 4: 183-188. Glvnn. P.W.. 1985b. El Nitio-associated disturbance to coral reefs and Dost disturbance mortality by Acanthaster ulanci. Mar. Ecol. Prog. Ser., 26: 295-300. Glynn, P.W., 1988. El Nitio-SouthernOscillation 1982-83: nearshore population, community and ecosystem responses. Ann. Rev. Ecol. Syst., 19: 309-345. Glynn, P.W. and Wellington, G.M., 1983. Corals and coral reefs of the Galapagos Islands, with an annotated list of the scleractinian corals of the Galapagos by John W. Wells. University of California Press, Berkeley, 330 pp. .
,
I
516 Glynn, P.W., Druffel, E.M. and Dunbar, R.B., 1983. A dead Central American coral reef tract: possible link with the Little Ice Age, J . Mar. Res., 4 1 : 605-637. Hallam, A., 1976. Stratigraphic distribution and ecology of European Jurassic bivalves. Lethaia, 9: 245-259. Hallam, A., 1977. Jurassic bivalve biogeography. Paleobiology, 3 : 58-73. Hallam, A., 1982. Patterns of speciation in Jurassic Grvuhaea. Paleobiology, 8: 354-366. Hallam, A., 1986. The Pliensbachian and Tithonian extinction events. Nature, 319: 765-768. Hallam, A., 1987. Radiations and extinctions in relation to environmental change in the marine Lower Jurassic of northwest Europe. Paleobiology, 13: 152-168. Hallock, P., 1987. Fluctuations in the trophic resource continuum: a factor in global diversity cycles? Paleoceanography, 2: 457-471. Hallock, P., 1988. The role of nutrient availability in bioerosion: consequences to carbonate buildups. Palaeogeogr., Palaeoclimatol., Palaeoecol., 63: 275-291. Hallock, P. and Schlager, W., 1986. Nutrient excess and the demise of coral reefs and carbonate platforms. Palaios, 1: 389-398. Hansen, T.A., 1987. Extinction of Late Eocene to Oligocene molluscs: relationship to shelf area, temperature changes, and impact events. Palaios, 2: 69-75. Heck, K.L. Jr. and McCoy, E.D., 1978. Long-distance dispersal and the reefbuilding corals of the eastern Pacific. Mar. Biol., 48: 349-356. Herman, Y. and Hopkins, D.M., 1980. Arctic oceanic climate in Late Cenozoic time. Science, 209: 557-562. Highsmith, R.C., 1980. Geographic patterns of coral bioerosion: a productivity hypothesis. J. Exp. Mar. Biol. Ecol., 46: 177-196. Hsu, K.J., 1986. Environmental changes in times of biotic crisis. In: D.M. Raup and D. Jablonski (Editors), Patterns and Processes in the History of Life. Springer, Berlin, pp. 297-312. Hsu, K.J., He, Q., McKenzie, J.A., Weissert, H., Perch-Nielsen,K., Oberhiinsli, H., Kelts, K., LaBrecque, J., Tauxe, L., KrBhenbuhl, U., Percival, S.F. Jr., Wright, R., Karpoff, A.M., Petersen, N., Tucker, P., Poore, R.Z., Gombos, A.M., Pisciotto, K., Carman, M.F. Jr. and Schreiber, E., 1982. Mass mortality and its environmental and evolutionary consequences. Science, 216: 249-256. Kaneps, A.G., 1979. Gulf Stream: velocity fluctuation during the Late Cenozoic. Science, 204: 297-301. Keigwin, L.D. Jr., 1980. Paleoceanographic change in the Pacific at the Eocene-Oligocene boundary. Nature, 287: 722-725. Keigwin, L.D. Jr., 1982. Isotopic paleoceanography of the Caribbean and East Pacific: role of Panama uplift in Late Neogene time. Science, 217: 350353. Keigwin, L.D. Jr., 1986. North Atlantic record of Late Neogene climatic change. South Afr. J . Sci., 82: 499. Kennett, J.P., 1983. Paleo-oceanography:global ocean evolution. Revs. Geophys. Space Phys., 21: 1,258-1,274. McLean, D.M., 1978. A terminal Mesozoic "greenhouse": lessons from the past. Science, 201: 401-406. Percival, S.F. Jr. and Fischer, A.G., 1977. Changes in calcareous nanoplankton in the Cretaceous-Tertiarybiotic crisis at Zumaya, Spain. Evol. Theory, 2: 1-35. Pitt, W.D., James, M.J., Hickman, C.S., Lipps, J.H. and Pitt, L.J., 1986. Late Cenozoic marine mollusks from tuff cones in the Galapagos Islands. Proc. Calif. Acad. Sci., 44: 269-282. Porter, J . W . , 1972. Ecology and species diversity of coral reefs on opposite sides of the isthmus of Panama. Bull. Biol. SOC. Washington, 2: 89-116. Porter, J.W., Battey, J . F . and Smith, G.J., 1982. Perturbation and change in coral reef communities. Proc. Nat. Acad. Sci. U.S.A., 79: 1,678-1,681.
517 Rampino, M.R. and Volk, T., 1988. Mass extinctions, atmospheric sulphur and climatic warming at the K/T boundary. Nature, 332: 63-65. Reid, D.G., 1986. The littorinid molluscs of mangrove forests in the IndoPacific region: the genus Littoraria. British Museum of Natural History, London, 227 pp. Rosenblatt, R.H. and Waples, R.S., 1986. A genetic comparison of allopatric populations of shore fish species from the eastern and central Pacific Ocean: dispersal or vicariance? Copeia, 1986: 275-284. Sheehan, P.M. and Hansen, T.A., 1986. Detritus feeding as a buffer to extinction at the end of the Cretaceous. Geology, 14: 868-870. Stanley, S.M., 1984. Temperature and biotic crises in the marine realm. Geology, 12: 205-208. Stanley, S.M., 1986. Anatomy of a regional mass extinction: Plio-Pleistocene decimation of the Western Atlantic bivalve fauna. Palaios, 1: 17-36. Stanley, S.M., 1988. Paleozoic mass extinctions: shared patterns suggest global cooling as a common cause. Amer. J. Sci., 288: 334-352. Thresher, R.E. and Brothers, E.B., 1985. Reproductive ecology and biogeography of Indo-West Pacific angelfishes (Pisces: Pomacanthidae). Evolution, 39: 878 - 887. Vermeij, G.J., 1978. Biogeography and Adaptation: Patterns of Marine Life. Harvard University Press, Cambridge, 332 pp. Vermeij, G.J., 1986. Survival during biotic crises: the properties and evolutionary significance of refuges. In: D.K. Elliott (Editor), Dynamics of Extinction. Wiley, New York, pp. 231-246. Vermeij, G.J., 1987a. Evolution and Escalation: An Ecological History of Life. Princeton University Press, Princeton, 527 pp. Vermeij, G.J., 1987b. The dispersal barrier in the tropical Pacific: implications for molluscan speciation and extinction. Evolution, 41: 1,046-1,058. Vermeij, G.J. and Petuch, E.J., 1986. Differential extinction in tropical American molluscs: endemism, architecture, and the Panama land bridge. Malacologia, 27: 29-41. Weyl, P.K., 1968. The role of the oceans in climatic change: a theory of the ice ages. Meteorol. Monogr., 8: 37-62. Wilde, P. and Berry, B.N., 1984. Destabilization of the oceanic density structure and its significance to marine "extinction" events. Palaeogeogr., Palaeoclimatol., Palaeoecol., 48: 143-162. Woodring, W.P., 1966. The Panama land bridge as a sea barrier. Amer. Phil. SOC. Proc., 110: 425-433. Zachos, J.C. and Arthur, M.A., 1986. Paleoceanography of the Cretaceous/Tertiary boundary event: inferences from stable isotopic and other data. Paleoceanography, 1: 5-26. Zullo, V.A., 1986. Quaternary barnacles from the Galapagos Islands. Proc. Calif. Acad. Sci., 44: 45-66.
This Page Intentionally Left Blank
519
INDICES SUBJECT INDEX Abalone black 472 green 452 pink 452 red 452 Abalone prcduc tion 46 1 Abiotic 433,462 Absorption spectrophotometry 261 Acunthmter (see Crown-of-Thorns starfkh) Aeolian aansport 261 Aerosol cloud El Chichon Volcano 7 1 Agoutis 481 Ahermatypic coral 97,202,258 Air-sea heat transfer 26,49, 151 Air-sea interaction 3, 5, 6, 12, 17, 25, 158, 190, 381 Air-sea momentum transfer 26 Aleutian Low 298,303 4 3 brown 367,375 green 337 recoverv 369 Algae 55,&2, 91-94, 97, 99, 100, 102, 114, 115, 154, 188, 193, 195, 201, 202, 204, 206, 211, 214,217, 336, 362,364,367, 369, 375, 377, 378,442,449, 451,452,459-461, 510 benthic 217 brown 91, 336,449 calcareous 115, 188,201,202 communities 102 community change 362 competition 445 crustose 204 filamentous 91, 102, 114 fleshy 102,204 frondose 102 green 337 intertidal 92,459,460 kelp (see Kelp) macroalgae 102, 113,336-338,446,459 microscopic 91 sea urchin food 442,45 1 planktonic (see Phytoplankton) red 91, 367, 369, 375,378,442,449 symbiotic (see Zooxanthellae) Algal biomass 438 Algal drift 446,447 Algallawns 206 Algal turf 91,367, 369 Alizarin Red-S 63 Alkaline earth elements 237 Allograft 66 Alternate stable community 114 Amphibian populations 482 Amphipod infestation 447 nests 446 nocturnal swarming behavior 447
520
Amphipod (continued) outbreaks 447,463 Analysis of Variance (ANOVA) 366 Anemones 108, 161,440,457 Angiosperms 287 Anions seawater 256 superoxide. 108 Annelids 202 Anoxia 510 Anoxic 339,509 Anti-El Niiio 59, 81, 82, 109, 110, 113, 294, 305,434, 438, 441 Ants 482 Aragonite 238, 241, 243, 247, 257, 259 Archaeological relics 197 Armadillos 481 Arthropods 202 leaflitter 482 Ascidians 343 Asteroid 508 Atmospheric forcing 436 Atmospheric temperature 288, 291,292,294, 295, 298, 299, 303, 309, 311, 312, 475, 488 anomalies 2,5 changes 307-309,401,507 high 408,420,475,477,494 Bacteria 509 coccoid 68 spaghetti 342 Balanids 336 Barium 256,261,273-276,277 Barnacle 55, 86,92, 127, 211, 461, 463, 510,512 recruitment 458 settlement 463 stalked 337 Bats frugivorous 481 Beavers 206 Beetles 482 Benguela Niiio 216,461,462 Benthic marine systems 339-342,350,351,434,465 Bioeroders 94,96, 99, 116, 204, 211, 510 Bioerosion 55, 56, 88, 96.99-102, 105, 112, 114-116, 171, 184, 185, 187, 192, 195, 201, 206, 208,212-217, 219,463 calcium carbonate 101, 185, 187, 192, 195, 201, 206, 208 damselfish absent, present 101 non-echinoid 100, 102 Biogeography 127, 128,133,508 boundary shift in subarctic 437 corals 187 cross-equatorial affinities 134 marine faunal distributions 273 EGological anomalies 437,440 Biological interactions 336 Biotic 114, 195, 433, 437, 505, 507, 511 Birds (also see Seabirds) 375, 381, 382, 385-389, 397-408,428,435437,440, 480, 481 albatross 97,403 antbirds 481 auklet 402,406 booby 381-392, 395-401,406 cormorant 107, 381,397-399,401-405
52 1
Birds (continued) darwin finches 361 flamingo 400 guano birds 324 guillemot 402 gull 397,400,402,440 kittiwake 402 murre 402 murrelet 401 noddies 396 pelican 392, 397, 398,401,440 penguin 397,400,403,404 petrel 395,398,401 prion, slender-billed 409 puffin 402 seabird 30, 428, 435, 506 condition 440 reproduction 463 reproductive failures 461 shearwaters 396,401,404 skimmer 400 tern 381,396,397,405 vulture 400 Bivalves endolithic 112 lithophage 99, 115, 195 mytilid 93 Bleaching (see Coral bleaching) Bomb radiocarbon 241 Boundary current 33,42,46 Brachiopods 202,337 Bryozoans 202 Cadmium 238,256,257,261,264,266,268-270,277,313 Calcareous algae 115, 188,201,202 Calcification 63,217,234,235 Calcium 237,238,256 Calibration statistics 292 Cannibalism hake 349 Carbon dioxide 249 doubling simulations 115 flux, global industrial 116 global budget 116 sink 115 Carbonate erosion 219 net production 99, 101 Caribbean Plate 190 Cephalopod (also see Squid or Octopus) 425 Chaetognaths 330 Chemical tracers 273 Chemicals indusmal and agricultural 111 Los Angeles County sewage outfall 450 Chemisorption 256 Chitons 345 Chlorophyll a 76,433,438 Chlorophyll concentrations 29, 33, 39,41,43,46,48 Chlorophyll maximum layer 440 Clam 344
522 Clam, giant 108 Climate anomalies 2, 5, 8, 11-14.237, 242, 285, 292, 295, 473, 475, 494, 495 Climatic patterns 3, 10, 190,237,474,491-493 Climatological shifts 185 Cloud cover 8,9, 164,239,247,249 Coastal climate stability 492 Coastal erosion 326 Coati 482 Cobalt 257 Coccolithophorids 39,40,47, 49,50,330 Community ecology, evolutionary perspective 433 Competition space 113,445 understory 445 Continental runoff 261 Continental seas regression 115 Continental shelf 340 Continental slope 340 Cool marine conditions 68, 81, 82, 216, 445,446,462 Copepods 332 Copper 257 Coral acroporid 87, 113, 135, 136, 175, 194,218 age estimates 103 ages 107 ahermatypic (see Ahermatypic coral) bleached and dead 92 calcification 217,234-236 carbonate accumulation 185,214 colonization 185,209 competition 171 condition 66 cryptic 111 diversity 88, 89, 133, 185, 187, 218, 219 endosymbiotic algae (see Zooxanthellae) environmental records 313 erosional losses 103 extinction 94, 190, 194 framework 105,203, 204, 206 gastrodermis 68-70 geochemistry 256-259 hermatypic (see Hermatypic coral) histological condition 81 histopathology 66 interspecifc competition 136 isotopic composition 236,238,243-249,269 life-histories 136 linear extension rates 243 massive 99 maximum linear growth axes 104 mesenterial filament 70 mesenteries 70 morphology 234 mucus 63,65, 87,91, 194 necrosis 69 net production rates 105 physiologic limit 216 pocilloporid 61-63, 66, 81, 82, 85, 86, 88, 91-94, 96-102, 105, 106, 113, 114, 153, 154, 173
523 Coral (continued) pocilloporid blocks 105, 106 pocilloporid branch tip mortality 81 pocilloporid tissue sloughing 81 predation 88,96,97, 110, 111, 185, 187, 195,217 recolonization 201 recovery 59,78, 154, 160,216 regeneration 112, 114 regrowth 61,211 remnant live tissue 104 river discharge indicator 237,274-276 scleractinian 55, 87, 92,95, 133, 135, 144, 154, 162, 187 skeletal banding 233-235,244,275,287 skeletal precipitation 258 species loss 88 spermaries 70 tissue atrophy 68 Coral-algal mounds 190 Coral bleaching 55, 59, 60, 62-66,68,70-75, 78, 81-83, 85, 87, 88, 91, 103, 107-109, 111, 113-116 141-147, 149-163, 167, 169, 170, 172, 173, 175, 176, 193, 194,236, 246 Caribbean-Bahamian bleaching event (1982-83) 108 cool periods 81 frequency 108, 173 geographic analysis 108, 171-173 Gulf of Panama 73 response 116 sea warming 71,75, 107, 155, 163-167 seventy 108, 153, 154, 162 thermal stress 107, 162 timing 68,71, 154, 161, 162 Coral community changes 87, 88,94-96,204,212 Coral cover 66, 67, 88,91,94,95,97,98, 101, 116, 150, 154, 162 Coral disease 66, 161, 171 pathogens 66 Coral diversity 68,88, 113, 185 Coral growth 79, 81, 102, 103, 105, 110, 113, 185, 192, 216, 217, 235, 236 discontinuities 96 interruption 102 rates 103, 160, 184, 192, 238, 246 Coral mortality 55, 60, 61,70,74,75, 83-85, 88,93, 96, 98, 103, 109-111, 113, 116, 141, 142, 144-146, 148, 149, 151, 152, 160, 162, 171-176, 184, 186, 193-195, 201, 209, 21 1,216,219,483,506 ElNiiio 463 gradient 75 mass mortalities 83, 154, 171, 172, 174 partial mortality 55, 87, 103, 104.21 1 Coral pigments 59 Coral recruitment 102, 112, 114, 135, 193, 194, 204, 209, 212, 214, 216, 217, 219 pocilloporid coral 102, 113 Coral reef 150, 152, 154, 171,173,177 accumulation rates 79, 105 age 107, 185 antecedent platform 185,219 building 192 communities 30-88,91-93,96-102 cover 154 demise 204,509 development 56,68, 184 framework 188 fringing reef growth 107
524
Coral reef (continued) fossil 217 foundation 185 growth 184, 185, 187 low diversity 216 midshelf 156 monospecific 203 pocilloporid 115, 192 recovery 59, 111, 112, 116 shelf edge 153 size 211 tectonic uplift 201,203, 206, 211 thickness 107, 189 zonation 86, 188, 190 Coral reefs barrier reef 190 Cocos Island 116 CostaRica 109 eastern Pacific 110, 114, 135, 142, 184-220 Galapagos Islands 93 Indo-Pacific 112 Jamaica 96, 113 Onslow Island 86, 100 Panama 93 pocilloporid 61,85, 86, 98-100, 106,203-211 Coral reef herbivores 95,98,206,208-212 Coral refuge populations 62 Coral remnants 112 Coral reproduction 2 14 fragmentation 105, 112,212 gonadal development 68 polyp bail-out 137 sexual 135, 136, 185 vegetative 135, 136 Coral stress cyclones 148 desiccation 144 dredging 145 fresh water 83, 144, 145, 167, 168 illumination 64, 144, 146, 160, 161 insolation 152 mainland erosion 145 non-thermal 83, 85 pollution 161 precipitation 148, 159 radiation 155 river discharge 145 sea level fluctuation 148, 149, 185, 186 sedimentation 145, 150, 185, 186, 188, 237 siltation 83,237 storms 171, 172, 174 synergistic interactions 160 temperature 59, 62, 64, 68, 71, 73, 87, 93, 108, 135, 144, 149, 150, 152, 155, 160, 161, 163, 171, 172, 184, 186, 187, 193, 195,203,209 thermocline shallowing 184, 188 tidal 85, 150, 171, 172, 174, 185 tropical storms 144 turbidity 145, 184, 188 ultraviolet irradiation 144, 149 upwelling 185, 186, 188, 194, 214, 216
525 Coral stress (continued) wave action 184 Coralline algae 94, 97, 99, 100, 102, 114, 154, 442, 451, 510 substrata 91 Coralliths 201 Corallivore 95, 96, 110, 114, 115, 173 coral prey availability 96 gastropod 93,98, 116 responses 96 snail 114 Coriolis Effect 32 Crab 91,367,438,457,459,461 anomuran 461 brachyuran 336 coral 88,91 dungeness 438,457 mass mortalities 460 mole 342 pelagic 401,440,441 rock 457 sand crab range anomaly 437 spider 91 swimming 337,342,346 Cretaceous 189,507,508 CretaceousD’ertiary boundary 115 Crinoid extinctions 507 Crocodiles 481 hatching success 482 Cromwell Current (see Equatorial Undercurrent) Crossdating 288, 289 Crown-of-Thorns starfish 88,94,96,97 access to coral prey 96 feeding on remnant patches of Porites lobata 96 outbreaks 110, 111 skeletal elements 111 Crustacean 55, 65, 77, 88, 91, 96, 99, 110, 116, 195, 425, 447 cryptic 99 Crustacean symbiotes 55, 65,77, 88,91,96, 116, 195 aggressive responses 65,91 behavior 88 coral guards 88,96 defensive behavior 77 density 65, 88 egg-canying females 77 emigration rates 77,91 greater visibility 91 lipid levels 77 mortality rates 77 mucus feeding 91 predators 91 reduced food supply 9 1 Crustose coralline algae (CCA) 99, 114,442,459 Currents Brazil current 403,404 California Countercurrent 401,412 California Current 130,433,434,436,438,440,451,454,456,462,464,470 Equatorial Current system 25 Equatorial Undercurrent (Cromwell Current) 14, 128, 130, 132, 133, 198, 236, 239, 241, 243,258,260, 326,420 velocity 130
526
Currents (continued) Equatorial Undercurrent cooling 508 nutrient depletion 509 primary productivity 508 Falkland Current 403 Humboldt (Peruvian) Current 326, 331, 332, 338, 339, 342, 349, 350, 464, 487, 488 Kuroshio Current 461 North Equatorial Countercurrent 128-130, 132-134,237,239,274,505,506,512 velocity 129, 133, 135 North Equatorial Current 129,130,134,192,274 velocity 130 Oyashio Current 461 Peru Current 236,237,239,505 Peru Undercurrent 326 South Equatorial Current 130,134,236 velocity 130 Southern California Eddy 451,454 Cyanobacteria 438 Cyclones 145, 147, 148,155, 172, 173 Darwin (Australia) atmospheric pressure 307 Deer 375,481 Deforestation 61 Dendrochronology 269,308,310-312 Dendroclimatic analysis 289, 308 reconstructions 291,292, 307, 31 1, 312 Dendroclimatology 287,288 Dendroclimatological data base 291 Deposit feeders 340 Desert areas Atacama 489 floral discontinuity 489 southern Peru 489 Desorption 266,273 Devonian 507 Diatom 30, 38, 39, 41, 47-50, 330, 332, 509 Dinoflagellate 47-50, 59,97, 107, 109, 110, 114, 330, 332 blooms 97, 107, 110, 114 Dipterocarps 474,483 Dispersal 135-137, 506, 512, 513 barriers 130, 133, 134 birds 406 corals 135-137 currents 132 long-distance 113, 128, 136,443 marine invertebrates 132,343 Dissolved oxygen 328,436 Distribution Blue-footed booby 382,383 Galapagos fur seal 419 Peruvian booby 382 rock lobsters 461 shrimps 461 South American fur seal 424 stalked barnacles 461 swimming crabs 461 Disturbances 26, 36-39, 56, 59, 70, 81, 83-85,97, 103, 107, 109-114, 116, 172, 184, 186, 216,218,442,445,463,465,487, 506,510,514 biotic 43-45, 50, 114
527
Disturbances (continued) global atmospheric 3-6, 12,111 long-term effects 94-102, 111,218, 445 recurrence 111,245,248,264,265 return intervals 111 regional frequency 49, 111,246, 247,267, 286, 463 secondary 55, 109, 112, 113, 116 storm 112, 114, 144 world wide 111,483 Diversity centers 56 Diversity index Fisher’s alpha 484 Shannon-Wiener (H) 89,90,340 Downwelling 12, 13 Dredging 145 Drought 56, 146, 147, 148-151, 310, 389, 395, 400,402, 405,473, 482, 483 seasonal 474 Earlywood 287,289 Easter-Darwin Southern Oscillation index 79 Echinoderms 202,204 Echoacoustics 331,349 Ecological changes 56 Eigenvector 299,301 Ekman pumping 266 Ekman’s Barrier 133 El Niiio dry season comparisons 475,482 effects on temperate pelagic ecosystems 436-441 evolutionary stability 5 13 frequency 49, 111, 197,246,247,267,286,463,493 heat storage 49 heat aansfer (latitudinal) 3,25 heat aansfer (oceWatmosphere) 26 history 4, 5 intensity 2, 111,363 nutricline regulation 26,39 oceanic current alteration 128 propagation 2, 12, 14,36, 39 sea level changes 17, 18,78, 83-85, 111, 132, 147, 148, 173, 390, 434-436,459 simulation experiments 68,70, 75, 116 thermocline regulation 26 timing 6, 8 water mass changes 39 El Nitio 1982-83 comparison with other disturbances 109, 116 differential mortalities 115, 162, 171, 174-176, 193,216, 366-369, 374, 404, 405,427,443, 445,476 drought (also see Drought) 475,476,480,482 duration 37,56 onset 36, 59 severity 81,108, 110,286,475 upwelling index 437 weather conditions 494 El Niiio 1987 107 El Nitio events comparisons 7, 111,407,435,483 historical evidence 4-6,441 moderate 78,407 strong 78, 81,82,493 very strong 6,7,49,78,81,82
528
El Niiio persistence 436 El Niiio type event off SW Africa 461 Electrophoretic comparisons 513 Emmigration 343,397 Endemic species 5 12 Endemism 135,502,513 Energy tmnsfer tropics to mid-latitudes 434 ENS0 effects abalones 450 asteroids and lobsters 450 biological impact 417 comparisons between northern and southern hemispheres 14, 17,441-450 corals (see Coral bleaching, Coral mortality, Coral stress) desert floras 487 eastern Pacific coral reef communities 56-81 evolution 506 extinction 113, 116, 194, 406,505, 506, 510, 511, 514 intertidal populations 458-461 Peru and northern Chile 335-339 Washington intertidal zones 459 kelp forest fishes 450 kelp forests 441-450 Macrocystis off Mexico and California 462-464 natural selection 506 nutrient regeneration 428 pelagic and benthic communities 464 preadaptation of intertidal species 460 recruitment 98, 116, 135,216,217,407,433,440-443, 4 5 4 8,450-459,461-464 reproduction 361, 372-374, 378,404,457 resilience to kelp disturbance 463 selection pressure 429 soft bottom populations 456-459 South American coast 460,461 speciation 406,464,482,512 Eocene 189,508 Epipelagic zone 331 Equatorial Current (see Current, Equatorial) Equatorial divergence 28,33 Equatorial dry zone 200 Eauatorial front 31 Eq'uatorial surface water 247 Erosion 18, 56, 61,76, 99, 102, 103, 113, 145, 183-185, 206, 214, 217, 219, 436 coastal 18,56 Euphausids 349 Euphotic layer 329 Euphotic zone 26,27,43,44,47,327 Ewythermal 45 Eustatic change 502 Eutrophic 46 Eutrophication 111 Evaporation 3 Evapotranspiration 292 Evenness (J') 88-90,337 Extinction 113-116, 190, 194, 406,463, 505, 506, 508, 512, 514 Feeding behavior Trapezia 91 Filter feeders 214 Fires 288 forest understory 480
529 Fish albacore tuna 438 anchovy 30946, 47,49, 286, 324, 330-334, 351, 385, 386, 398, 404, 405, 417, 425, 438, 440 growth and life history 440 spawning biomass 440 blue rockfish 440,456 boccacio 457 bonito 333 burrfish 440 cabinza 348 cabrilla 348 chilipepper rockfish 457,463 coco 348 cojinoba 348 corvina 348 damselfish 55, 88, 96,98,99, 101, 102, 113, 114, 116, 195, 206, 208, 209 disturbances 116 ejection behavior 98 territories 55,98,99, 114, 206 dogfish 348 dolphinfish 333, 336 dorado 438 dover sole 457 English sole 457 flatfish 348 gobies 454 guineafowl puffer 93,96 diet shift to non-pocilloporid corals 97 searchimage 97 stomach contents 96 switch to alternative prey 97 hake 46,348,349 hemng Pacific 457 round 333 kelp surf perch 447 king angelfish 91 lingcod 457 liza 348 lorna 348 machete 348 mackerel 46,351,398 horse 332,333, 335 jack 46,438,457 Pacific 332 Spanish 332,333 marfin 438 mullet 348 parrotfish 215 pufferfish 112, 121 rays 348 rexsole 457 rockfish 438,440,456,457,463 sablefish 457 salmon 437,438,457,463 change in size 347 Chinook 437,457 fecundity per female 347 mortality 347
530 Fish (continued) salmon smolt survival 347 sardine 46, 330-335 biomass 331 growth 332 reproduction 332 sciaenids 348 sea bass 348 seiioritas 456 sharks 333 sheephead 454-456.463 sierra 333 silverside 333, 348 skipjack 333, 336,438 surfperch 447 swordfish 438 thomyhead rockfish 457 triggerfish 112,437 white mullet 440 white seabass 438 widow rockfish 463 yellowfin tuna 333,438 yellowtail 438,440 yellowtail rockfkh 440 Fish meal 333,335 Fisheries 333-336 anchovy 350,386-389,404,417 artisanal 342 commercial 44,347,406,438,457 Peruvian 417 red sea urchin 452 salmon 347 shellfish 347,437 squid 347,438 traditional 460 Fishes 91,93,94, 102, 105, 110, 113, 114, 378,433,437,447,449,454,456,457 abundances on reef 93-95 behavior (also see Fish, Damselfish) 41 coastal 454,456 demersal 324, 342, 347, 349, 351, 425, 460 dispersal 132, 133,454, 513 feeding patterns 349 herbivorous 91, 95, 105, 113 increased abundances 456 numerical response 91 Panamic reef 93 pelagic 463 recruitment 47,49,454-456 reef herbivores 91,93-95, 113 reproduction 46,50 spawning 47,48 Flooding 1,56, 107, 144, 145, 197, 198, 396-398 Fluvial discharge (see River discharge) Fogzone 488 Food web 22,43,46,50,330 marine 506 pelagic 351 Foraminifera 255,257
53 1 Forest dipterocarp 474 dynamics plot 479 floor animals 482 long term effects 479 mature 474,476 second growth 476 soil depth variation 476 temperate 287 wet forest palm 479 Forest communities East Borneo 474 Forest disturbance 288 Fossil fuel consumption 249 Fossil record 506,507 Fossils 515 Fresh water inputs 109,405 Fruit Gustavia 481 Fulvic acid 273 Fur seal Galapagos 418-424 Southern 418,424-426 Gammarid amphipods grazing damage 446 infestation 446 kelp curler 446 Gannet 405 Garua 488 Gastropod 93,98, 110, 116, 133 Gene flow 506 Geochemistry of corals 256-259 Geologic disturbances 85 earthquakes 85 tectonic uplift 85, 103, 109, 127, 202,208, 211, 364 volcanic eruptions 85, 109 Geostrophic flow 45 Geostrophic readjustment 436 Glacial cycles 199, 200,216, 218,502 periods 218 Glaciation 514 Global temperatures 111,216 Gonads 68,70,76, 112, 386 Gorgonians 92, 108, 161, 177 Grazers abundance 113 echinoderm (see Sea urchin, bicerosion and grazing) macroalgae 102,450,460 mortality 352 Grazing (also see Sea urchin grazing) 171, 217, 237 Greenhouse effect 507 Greenhouse global wanning 114, 115 Guano 43,324,395,409,417 Gut 70, 91, 375 Gymnosperms 287 Gyre 39,43 Habitat anomalies 437 Hard bottom community 337 Herbivores (also see Fishes, herbivorous and Sea urchins) 206
532 Herbivory high rates by insects 480 Hermatypic coral 95,97, 133, 187, 189, 203, 219, 258 Holocene 111, 190, 196 Holoplankters 330 Howler monkeys 481 Human activity and settlement 483 Hurricanes 98, 113, 146, 151, 172 Allen 144,171,483 Flora 144 Hattie 144 Hydrocorals 87,92, 94, 95, 149, 150, 161, 177, 187, 463 bleaching and mortality 92 regional extinction 463 Hydroids 509 Hypoxic 339,351 Icecores 196 Ice sheets 287 Ice volumes 255 Iguana (see Marine iguana) Immigration 128, 135,343,345,506 invading tropical species 460 Imperial Formation 191 Infrared light absorption 71 Infectious agent 63,66 Insolation 150-152 Interglacial cycles 199,200 periods 200,218 Internal waves 398,436 Intertidal sand beach community 339 Intertidal zone 45.337 Intertropical convergence zone 129,146, 151,239 Iron 257.272 Irradiance 70, 85, 87, 107-109, 142, 144, 146, 165, 172,445 high 87, 107-109 solar (see Solar radiation) ultraviolet (see Ultraviolet radiation) visible 64,157 Island stepping stones 133 Isograft 66 Isopleth 41 Isopod 447 Isotherm 41 Isotope analyses 107 anomalies 249 carbon, stable 238, 247, 249, 255, 269 depletion 246 enrichment 246 nitrogen, stable 44 oxygen, stable 238,243,244, 246, 249,255,269 stable 234, 237, 243, 278 Isotopic precipitation 269 Jellyfish 330 Jetstream 405 Jurassic 507, 5 1 1 Kelp 343,400, 433, 434, 436, 437, 441-443,445-448, 450-454,456, 458, 460-463,465 decline in aerial coverage 447
533 Kelp (continued) giant 433,441,450 growth 450 harvestdata 445 lifecycle 463 local extinctions 464 Macrocystis canopy 443,444 Macrocystis forest 462 Macrocystis mortality 442 mortality 445 Point Loma density 446 recovery 450 recruitment 443,445,451,453 recruitment density, Naples Reef 449 surface biomass 450 survival 450 Kelp communities Southern California 434.458 Kelp forest asteroids 453 carnivores 441 disturbances 442 fish recruitment 454-457 herbivores 441 PointLorna 442 SanOnofre 445 SantaBarbara 448 structure 441 warm water effects 443-447 Kelvin Wave 12, 13, 26, 132, 1 148, 19 326,339 Kinkajous 481 Krill 403 La Niiia (see Anti El-Niiio) Lake sediments 287 Laminarians 337,338,459 Land clearing activities 110 Lantana 398 Larvae 128 Acropora 135 bivalve 330 crab 459 fish 47, 49, 454 gastropod 330 holometabolous 482 invertebrate 129, 132, 345 lecithotrophic 133 mackerel 332 plankton 127, 128, 132,451 planulae 134, 192 Pocillopora 135, 136 scallop 347 Sea urchin 451 shrimp 330 sub-surface transport 133 thermal tolerance 133 transport 132-134, 136,453,457,459 Larval competency 128, 130, 132, 133, 135,433,454 Latewood 289 Latitudinal distributions 70 Lava 201.209
534
Lead 238,257 Lead carbonate 259
Leafconsumption insects 480 Leafflush 478 Leaflitter arthropods 482 invertebrates 476 moisture content 475 Leafproduction 479 Leatherback turtle 437 Light (see Irradiance or Solar radiation) Limpet 336,337,344 Linear regression analysis 290 Lipid 49 Little Ice Age 81,107, 110, 187,216,508,514 Lomas formation 488 grazing destruction 497 Lomas vegetation 489 community composition 496,498 Low tide exposures 82, 109 reef flat exposures 113 Lower reef slope 94,98 Lower slope zone 66 Maasmchtian stage 508 Macrobenthos 324,336-339 Macroclimatic factors 287 Macrozooplankton biomass 433,438 Magnesium 256 Malaysian forests 484 Mammal sighting rates 481 Manganese 238,256,257,261,266,268,277 Mangroves and associated fauna 460 Marine iguana 361,362,364,365,367,369,371,374,375,377-380,481 growth 369-371 hatching success 482 measurement 365 mortality 362,366-369 optimal foraging 377 reproduction 372, 373,376 stomach contents 367 Mariscos (shellfish seafood) 342,343 Mass mortalities (also see Coral mortality) 83, 113, 144, 171, 174, 336, 337, 343, 351,453,460 fish 456 Meroplankton 330 Mesic community 500,501 Mesic forest 308 Metal adsorption 256 Microalgal turf 99 Microclimatic conditions 108 Microfossil 261 Micronutrients 261 Microphytes 92, 100 Midshelf 155 Migrations fmgivorous birds 480 Miocene 185, 190 Modeling 409 Molluscs 202,204,509, 511-513
535
Monsoon Indian Ocean 149,309 Indonesian 146 Monsoonal storms 200 Mortality 347 birds 350, 395400,402, 404, 405,407,408, 417 coral (see Coral mortality) delayed 66,98,478 fur seal 420, 421,423,424 mammal 480 marine iguana 362,366-369 sea lion 424,427,429 seals 350 Multivariate analysis 290 Mussel 336 growth rates 344,458 subtidal banks 460 Mytilid 336, 338,339 Namibian faunas 461 Nanoplankton 330 Natural selection 218 Necrosis 68,76 Neogene 508 Nekton 324 Nematocysts 96 Nemerteans 342 Nesting failure 406 Night herons 457 Nitrate 29, 31, 33, 38, 39,43, 239 decreased availability 437 Non-seismic emergence 84 Non-upwelling 59, 81, 110, 194 Nutricline 28, 29, 31, 38, 39 depression 28, 39, 40, 327, 329 depth 26, 28, 30, 39, 47 shallowing 35 topography 3 1,44 Nutrient 14, 15, 22, 31,41,43,44, 55, 59, 107, 110, 115, 188, 192, 214, 216, 217, 361, 382, 390, 398, 418,428,433,436-438,443,445,446,449-451,461, 462, 505, 506, 509, 510 availability 434 concentration 32, 37-39,41,47,238,239,348,438,510 depletion 29, 31, 33, 35, 38, 39, 43, 41, 239, 443, 450,451, 506, 509, 510 enrichment 35,43 loading 107, 110, 115 micronutrients 261 new 26 nitrate 29, 31, 41, 43,44,437, 446 nitrogen deficiencies 443 saturation 32,37 stratification 15, 32,436 supply 30,420,445 transport 29 uptake 32 upwelling 33, 214, 216, 217, 390, 398, 461 Nutrient depleted waters depth 436 Nutrient starvation hypothesis 445 Oceanic circulation 3, 128,255 equatorial mixing 134
536
Oceanic circulation (continued) geostrophic mixing 28 island-wake mixing 28,31 shelf break mixing 28 surface 277 surface mixed-layer 15,26,28, 29,32,38,48, 199 tidal mixing 28, 38 turbulence 47,48 wind-driven mixing 261 Oceanic doming 3 1 Oceanicfront 55 Oceanic warming 115,438 Octopus 346,461 Oilspill 150 Oligocene 190 Oligotrophic 30 Oocytes 70 Opossums 481 Opportunistic species 443 Orbital forcing 255 Ordovician 507 Organic material, export 44 Outgoing longwave radiation (OLR) 8,9 Oxygen toxicity 109 Pacas 481 Paleo-river flow indicator 273 Paleoceanographic reconstructions 196 Paleoceanography 508 Paleocene stage 508 Paleochemical indicators 257,261,277 Paleocirculation 192, 198 Paleoclimate 287, 314 Paleoclimatologic reconstruction 196,200 Paleoenvironments 256,258 lacustrine records 200 Paleoenvironmental indicators 256 Paleoenvironmental recorders 261 Paleoproductivity indicators 243 Paleothermometers 243 Palms 476,478,479 Panama Formation 191 Panamanian Seaway 190 Parasites 408 Peccaries 481 Pelagic system/community 434,436-441,460 Pelagic zone 348 Pelecypod 510 Pen shells 345 Penaeid shrimp 345 Peruvian continental shelf 329 Phanerozoic 508 Phenology 310,478,482 Phosphorous 264 Photoinhibition 269 Photosynthesis/respiration (P:R ratio) 87 Physical perturbations 336 Phytoplankton 214,438,440,458, 507-509 abundance 49,327,440,458,507 assemblage 438 autochthonous 330
537 Phytoplankton (continued) biomass 33,46,329, 438 bloom 47,48 cadmium uptake 261 critical-depth 38 extinction 507 motility 47 near upwelling 33,41,214,328-330 phototaxis 47 productivity 32,41, 329, 505,508 species composition 49,330 uptake of nutrients 32 upwelling community 330 Picoplankton 330,438 Pinniped 440,441 Piiion trees 102 Pismo clam settlement and survival 458 Plankton 14, 55, 56, 132, 392,428,429,431, 438, 463-465 biomass 29, 30, 47 calcareous 255 dispersal 132 nonoplankton 330 picoplankton 330,438 Plant mortality 476 wilting 476 Plant production fruit 474,478,479 new leaves 474,478,479 Plate tectonics 127 Pleistocene 185, 192, 198 Pliocene 186, 190, 192,217 Poleward flow 434-436 Pollution 161,288 Polychaete 99, 337, 342,458 tubicolous 338,458 Population increases molluscs 460 octopus 461 purplesnail 461 scallops 461 sea urchins 98, 110, 460 Precipitation 3, 5, 8-10,43, 144, 146, 148, 149, 155, 159, 161, 164, 167-169, 196, 239, 260, 261,269,270,273, 275, 277, 286,294,295,298, 303, 308, 326,498, 507 African 309,310,474 annual values 83,363,474,477,491,492-494 Australia 159-161,200, 310 Central America 167-169,309,474,475 Chile 491,492 decrease 5, 159, 161, 303, 309, 310,473-475 increase 5, 83, 168, 196,291, 303, 361, 362,396, 397,405,494, 495 New Zealand 311,312 Peru 5,492,493 regional 146,297, 313 South America 309 torrential rain 198,328, 344,494,495 variation 2, 4, 146, 149,200, 311 west Pacific 3 10 Precipitation patterns, reconstruction 312
538 Predation 88, 91, 94, 96, 110-112, 114-116, 185, 195, 217, 377, 398, 433, 434, 453, 458, 460 Panulirus interruptus on strongylocentrotids 453 secondary coral predator effects 96 Predator concentration 114 Principle Component Analysis 290, 371 Productivity 27,38, 46, 55, 59,93, 185, 203, 217, 219, 361, 381, 382, 398,404,405,417, 423,424,451,457,458,462, 501,505,508, 509 coastal 43 critical-depth 38 high 55, 59, 381, 392, 417, 424 models 44,45 new primary 26,32,43,44,47,49 primary 22,26,38,44,45, 50,93,217,424,462,505,508,509 primary decrease 93, 327,428,462,508,509 primary rate 30,44 total primary 43,44,49 Propagules 62,434, 502,506, 512-514 Proxy climate records 291,299, 307,312,314 Proxy indicators E N S 0 196,234,324 nutrients 264 paleoclimate 255, 286,287 Pseudoscorpions 482 Purple snails 461 Pycnocline depth 28,326 Pyroheliometer 157 Radiocarbon 278 bomb 241 Radiochemical assay 273 Radiolarian 509 Radiometric dating 104,257 Radium 273,274,277 Rafting 127, 128 Rainfall (also seePrecipitation) 1-6, 8-10, 12, 56, 83, 144, 146-149, 155, 159-161, 164, 167169, 200, 308-310, 361-364, 395-398,417,460,473-475,477,487, 494, 495, 502 Rainfall anomaly region 259 Range extensions (species) 127, 136, 137,406,437,439,453,454,456, 463 Rare earth elements 256 Razor clams 457 Recruitment 347 birds 407 crustaceans 433 fishes 433 Hjort's hypotheses 465 larval settlement 337 lobster 453 marineiguana 377 rareevents 454 sardines 351 sheephead 454-456 Strongylocentroturpurpuratus 458,459 Recruitment patterns 464 Reduction of error statistic (RE) 299,300,301 Reefdevelopment 55,59,68, 111, 115, 116, 184,217,219 Reef drowning 115 Reef flat 61, 67, 82-85, 88, 100, 108, 109, 113, 144, 150, 161, 165, 171-173, 175 Reef framework 55,56,61, 102, 105, 188,216 Reef refuges 96
539
Refuges predator access to 116 Regression analysis 37 1 Reproduction coral (see Coral reproduction) increase 361,377 penguin 404 River discharge 256,260,266,273,275,277, 328,344 River flow 312 Riveroutput 9 Rocky intertidal zone 35 1 Rossby waves 12,326 Rudistid molluscs 115 Runoff 83, 167-169,237 Salinity 9, 83, 84, 142, 144, 145, 159, 164, 167-170, 172, 237, 361, 417, 434, 436, 438, 460 anomaly 434,436 decrease 38, 39, 144, 168, 169, 239, 244, 344 increase 243 values 168,239 variations 84 Salps 330 Sandy beach community Peru 461 Sandy beaches 351 Santa Ana winds 459 Saplings 478,480 Satellite cloud movement 12 imagery 33 temperature sensing 43,7 1 Scallop 46, 346, 347,461 rock 92 Scanning electron micrographs 68,69 Scleractinian coral (see Coral, scleractinian) Sclerochronology 96, 107, 192, 261,276 flourescent bands 237,273 growthrecord 96 Sea anemones 108,440 Sea conditions 56,107 Sea floor communities 339 Sea level 1, 6, 12, 17, 18,25, 26, 33,55, 56, 68,70,78, 83, 84,93, 115, 132, 141, 146-149, 173, 185,200,201,217,326,336,337,363,369,390,434-436,459-463,496,507 French Polynesia 84 high 56, 84, 85, 93, 148, 396, 397,435, 436,459, 462 low 111, 147, 173,434 pressure (atmospheric) 294,295,298 rise 17, 115,507 south Pacific 83,84 Tokelau Islands 84 values 17, 18, 83, 147, 148, 186, 296 west Pacific 83, 84 Sea level pressure index 286,298 Sea lion 399,418,426,427,441 Sea lion pox 427 Seaotters 454 Sea stars (also see Starfish) 85,96 Sea subsurface temperature increase 152, 162, 163 values 152, 163, 327,437,444 Sea surface discoloration 83
540
Sea surface temperature, SST (also see Sea water temperature) 6, 35,47,234,294,295, 309, 313,425 amplitude 6, 8 anomaly 5-8, 13-15, 17, 18, 22, 75, 76, 128, 146, 149, 151, 156, 161, 193, 264, 277, 327, 362, 363, 371,372, 390,459,473 cool bias 71 dailyextremes 71 decrease 36, 37, 151 deviations 71,81,82 distributions 39 extreme low temperatures 71,81,82,216,422 geographic gradient 93 heat storage 22,49 heatsurplus 24 increases 5, 36-38, 144-147, 152, 155, 167, 170, 177, 184, 192, 193, 197, 236, 336, 404, 425 longitudinal heat storage 30 mean montly curves (1882-1983) 72-74 mesoscale changes 436 monthly extremes 71, 156,362 reliability 7 1 satellite 71, 152, 163, 164 ship stations 71, 156 shore stations 7 1 stratification 32,38 values 2,7, 8, 14,22, 24, 29-32, 34-39,42,46,72-74,76,78,80-82, 149, 150, 152, 155157, 164, 168, 188, 194,236,239,242,246,265, 327-329,363, 391,422,424,494 variation 3,148,163 Sea urchin 55,92, 94,96, 98-100, 102, 113, 116, 202, 206, 208, 209, 211-213, 216, 217, 336, 337, 448,450-453,457,458,461 abundance 98, 113, 114,206,209,448,450 aggregation 99,211,212, 451 bioerosion 94,96,99-102, 116, 195,204, 206,208, 209, 211-214, 216 grazing 56,99, 102,442,450, 451 larvae 451 purple 458 recruitment 102,451,458,459,461 red 452 spine nipping 98 white 453 Sea urchidalgal dominated community 98 Sea urchin barren 114,442,448,451 Sea warming (also see Sea surface temperature) 81 adaptive responses 115 primary cause of coral bleaching 107, 116 prolonged 116 Sea water temperature (also see Sea surface temperature) elevation 107, 108, 111,401,434-459 extremes 68 mixed layer 26,29,38 upwelling 424 Seas late Cretaceous 115 Seawater density gradient 47 Sediment 56, 145, 171, 197,202,214,434,460 Sediment load 460 Sedimentary records 261 Sedimentation 61, 83, 107, 109-111, 115, 143, 145, 147, 150, 159, 185, 187, 188, 326, 434 increase 344
54 1 Sediments deposition 197 reduction 266 Selection pressure 429 Selection regimes 506 Selenium 272 Sewage 111,264,450 Shannon-Wiener diversity index (H') 90 Shrimp 438,456,457 Ocean 438,456,457 Shrubs and treelets 478,500 deathrate 477 Silicate 275 Siltation 83, 107, 142, 145 Siphonophores 330 Sipunculans 99 Snow cover 311,313 Soft corals 108, 150 Soil moisture 475,478 Solar heating 3,4,24, 33, 169 Solar radiation 155, 157, 164, 165,479 values 157 Southern oscillation index 79, 149, 285-287, 291,294, 295,298, 299, 301, 302, 305, 306, 311, 313 Spawning Donarserra 461 Speciation 506,5 14 Species diversity Fisher's alpha 484 Shannon-Wiener (H') 89,90,340 Species richness 113 Spiny lobster 346,453 Spinyrats 481 Sponges 195,204,215 Sporophyll growth 445 Sportfishing industry 438 Spring tides 85,93, 157 Squid 418, 420, 428,437, 438 Squirrels 481 SST (see Sea surface temperature) Stable isotope (see Isotope, stable) Stable isotope analyses 107 Stalked barnacle 337 Starfish (also see Sea star) 453 batstar 453 Static stability 47 Stenothermal 133 Steric height 45 Storm effects delayed 483 Macrocystis canopies 442 Phragmatopoma catifornica 458 Storm induced changes dispersal of algae 443 light levels 443 provision of open space 443 Storms 18, 109, 144-147, 171~174,200,303,396,433,434,436,441-443,445-452,456, 458,492,505 activity 56 mid-latitude 434
542 Storms (continued) mixing 436 severe 18, 109,436, 505 Stress 55, 59, 64,68, 78, 79, 83, 87, 93, 107, 108, 115, 142, 144, 155, 159, 160-162, 171174, 198,391,434,443,445,457, 459,460,476 Strontium 237,238 Strontium carbonate 259 Subaerial erosion 185 Subsidence 127 Substratum stability 442 Subsurface water temperature 15,35-37 Subtidal zone 337 Supertethys 115 Surfclams 461 Suspension feeders 93,214,340,441 Swamp oil-palm 479 Swell 85, 93, 326, 336 Synergistic 160 t-test 294, 366 Tahiti-Darwin sea-level pressure index 286,298 Tectonic uplift (see Uplift) Teleconnections 152, 176,272,286,294,298,307, 310, 312, 313, 324, 409 Extra-tropical 291,298,314 Temperate El-Niiio, northern hemisphere 434 Temperate forest 287 Tethyan circulation 198 Tethys Sea 189 Supertethys 115 Thermal history 79, 108 Thermal stratification 38 Thermocline 1, 12, 14, 26, 29, 31, 151, 184, 188, 198, 199, 327, 398, 417, 425,426, 428, 429,433,436,438,443,445,450,462 anomaly 36 depression 12, 13, 26,40, 41, 49, 151, 261, 327, 329, 331, 348, 417, 425 depth 28,44,47 shallowing 35,41 steepening 198 vemcal excursions in summer 450 Thermohaline circulation 27 Tidal exposures 82,84, 109, 116, 144 Tides high 408 low 82, 109, 157, 172 lunar 38 Trace element coral skeletons 234,256,261,264-278 indicators 107,264-278 seawater 256 Tracegas 286 Trade winds 12, 22, 24-28, 30, 31, 33, 38, 48,49, 151, 277 northeast 239, 390 southeast 79, 81, 239,404, 417 Treegrowth 479 evapotranspiration 292 increase 301 precipitation 288,292 regulation 299 standardized growth index 289,290 sunlight 288 temperature 288,292
543 Tree growth (continued) water storage 292 wind 288 Tree ring 196,233,285-314 chronology 290,291,301, 303,307,309 density 308 grid 309 width variation 298,308, 309 Trees 102,285, 287,288, 290,292,298,299, 307-312,474,476-480,483, 500, 501 death rate 476,479 deciduous canopy 474 sub-polar 287, 288 sub-tropical 288 temperate 287,288 tropical 288 Tropical rain forest 310 Tunicates 202 Turbidity 18,56, 83, 145, 157-159, 184, 188, 328, 387 Ultra-violet radiation 85, 87, 149, 157 UV-B 64,85 Uplift Central American 508,5 11 tectonic 85, 103, 109, 127,202, 208, 211, 364 Upper reef slope 66 Upper slope 67 Upwelling 25, 27-29, 55, 56, 59, 68, 71, 73, 81, 82, 109, 110, 113, 116, 135, 184-186, 188, 189, 192, 194, 198, 200, 214, 216, 217, 219, 243, 246, 256, 260, 266, 327, 331, 346, 347, 381-383, 390-392, 395, 396, 398, 403-406, 417-420, 424-426, 428, 436, 437, 443, 445, 450, 458, 461-463, 489, 505, 508-510, 513 California 406,436,443,463 coastal 32, 33, 35, 41,44, 45, 50, 403,404 Cromwell Current 239,420 equatorial 28, 31, 35,239,260 geostrophic 28 geostrophic suppression 45 Humboldt Current 326,350,489 Macrocystis 445,462 mixing 28,260 nutrient-poor water 382,396,398 PeruviadChileanarea 347 physical/biological coupling 33 reef development 81, 109, 185, 186, 189,216,510 seasonality 268,404 shelf-break 28 strong 71,81, 82, 109,419,450,508 tracer 264 variability 277, 489 wind-driven 27,28,41,44,45,239,260,261 Upwelling centers 81, 109, 110 Upwelling ecosystem 351,398 Uranium 256 Varved lake sediments 287 Vegetative propagation 112 Verification statistics 292 Vicariance 127, 128, 135 Volcanism 508 Walker circulation cell 22.24 Warming duration 75, 108 rate 75
544 Warming trend 78,79,116 Washington (state) beaches 457 Water clarity (see Turbidity) Water column perturbations 434 Water flow poleward 435 reduction 107 restriction 70 Water motion 434 Wave assault 55,93 Waves 18,48,396 Westerlies anomalies 148, 149, 151 intensification 12,436 White-face monkeys 481 Wildfires 483 Wind-driven circulation 32,41, 111 Windforce 326 Wind pattern shifts 142 Wind speed 44,165 anomalies 151, 155, 158, 161 decreases 164, 166 increases 166 values 158,166 Windrun 158, 160, 165, 167, 169, 170 Winter storm season 435 Winter storms 303 Wright's index 300 Xenograft 66 X-ray 289 Yanaiwaves 326 Zinc 257,272 Zoanthids 108. 177 Zoobenthos 337 Zoogeography 406,437 Zooplankton 91,428,437,439,440,463 abundance 437,463 biomass 439 density 428 grazers 49,50 reproduction 50 Zooxanthellae 30,55, 59, 63,68, 69,71,76,77, 87, 108, 109, 116, 134, 142, 144, 145, 193, 235-237 degeneration 67 densities 77 expulsion 109 necrosis 76 photosynthesis 235 Zooxanthellate species 108
545 SYSTEMATIC INDEX
Acalypha diversifolia 478 Acanthasterplanci 55,56,88,91,94-97, 110-112, 114, 117-120, 123-126, 173, 195,217 Acanthoica SD. 40 Acanthopleia echinata 345 Acanthurus spp. 95 Acropora 98, 135-136, 150, 154, 171, 173, 175, 176, 190,218,230 A. cervicornis 175 A.palmata 175 A. valida 87, 113, 135, 194, 218, 511 Acroporidae 230 Acrosorium 443 Actinacididae 23 1 Actinacis 231 Agaricia 161, 162, 171 A. agaricites 162 A. tenuifolia 162 Agariciidae 203,230 Agarum 441 Agathiphyllia 23 1 Agouti 481 Alouatta 481 Alpheus lottini 88 Alveopora 23 1 Amblyrhynchus cristatus 361-378 Amphiroa beauvosii 9 1 Amphithoe 446,447 A. humeralis 446 Annelida 132 Anolis limifrons 482 Annona spraguei 480 Anous minutus 396 Anous stolidus 397 Antiguastrea 231 Antilloseris 230 Aplysia juliana 133 Araneae 482 Arctocephalus australis 418,424-426 Arctocephalus galapagoensis 418-424 Arenaeus mexicanus 342,343 Argopecten 46 A . purpuratus 46,344, 346, 352 Argylia radiata 497 Arothron meleagris 93,95-97, 114 Asteraceae 496 Asterina miniata 453 Asterionellajaponica 41 Asteromphalus heptactis 40 Astreopora 230 Astrocoeniina 203,230 Atrinamaura 345 Aulacomya ater 338,344 Bactris major 479 Balanus glandda 458 Balbisiapedunculata 498 Bifurcaria 375 Blossevillea galapagensis 91,92 Bombacaceae 478 Boraginaceae 496
546
Brachyistius frenatus 447 Brachyura 343 Bufo typhonius 482 Bursera graveolens 500 Cactaceae 481,489 Calamophylliidae 230 Calandrinia cymosa 498 Callinectes arcuam 342 Cancer spp. 344 C. coronatus 344,346 C. magister 451 C. porteri 344,346 C. setosus 344, 345 Cathartesaura 400 Caulerpa racemosa 102 Cebus 481 Cecropia insignis 480 Centroceras 361 Cepphus columba 402 Ceratiumfurca 40 Ceratulina sp. 40 Cervus elaphus 315 Cetengraulis mysticetus 385 Chaetocerosperuvianus 40 Chaetoceros sp. 40 Chaetomorpha 369 Chawdorea wendlandiana 419 Chilomycterus affinus 440 Chloris virgata 500 Clahcora 231 Clupea betincki 398 Cnidaria 132 Coccolithuspelagicus 40 Coenothecalia 230,232 Colpophyllia 231 Concholepas concholepas 346 Conus ahlli 5 13 Conus nux 513 Capiapoa 491 Coralliophila 98 Coryphaena hippurus 333 Coscinaraea 230 Coscinodiscus sp. 40 Creagrusfurcatus 391 Cristaria multijiah 498 Croton scouleri 500 Ctenochaetus 95 Cyathoseris 230 Cyclococcolithfragilis 40 Cyclococcolithus leptoporus 40 Cycloseris 23 1 Cycloseris mexicana 66 Cynoscion analis 348 Cyperus elegans subsp. rubigirwsus 498 Dasyprocta 481 Dasypus 481 Dendrophylliidae 232 Dendrophylliina 232 Desmarestia 443,445 Desmodim procumbens 500
547
Diadema 56,95,98,99, 113, 195 D. antillarum 113 D. mexicanum 56,98,99, 195 Diaseris 231 Dictyopteris 443 Didelphis 481 Diomedea immutabilis 397 Diomedea irrorara 397 Diomedea melanophris 403 Diomedea nigripes 397 Diploastrea 190,231 Diplopoda 482 Diploria strigosa 262,277 Dipteryx 48 1,484 Discosphaera tubifera 40 Dispio 339 Donax peruvianus 339,461 Donaxserra 461 Drupella cornus 110, 117 Echinodermata 132 Echinometra vanbrunti 202,206 Egregia 441 Eisenia 441 E. arborea 450,463 Elaeis oleifera 479 Emerita analoga 339,342,459,461 Emiliana huleyi 40 Endomychura craveri 401 Engraulis anchoita 404 Engraulisjaponicus capensis 405 Engraulis mordax 47 Engraulis ringens 330, 334,398,383,425 Enteromorpha 91,369 Eragrostis peruviana 496,498 Erythrina velutina 500 Etrumeus teres 333 Eucidaris thouarsii 56, 85, 97-103, 105, 106, 195, 202, 206,208,211, 213, 217 Eudyptes chrysocome 404 Eudyptes chrysolophus 403 Eulychnia-Copiapoa 49 1 Eulychnia-Puya 491 Euphorbia lactijlua 498 Euphorbiaceae 478 Euphorbia-Eulychnia 491 Euphylax dovii 342 Euphylax robustus 342 Exuviaella vaginula 40 Favia 231 Faviidae 231 Faviina 231 Favites 190,231 Fissurella spp. 344, 345 Fregara ariel 396 Fregata magnij7cens 397 Fregata minor 396 Fungiidae 231 Fungiina 203,230 Gardineroseris 230 G. planulata 64, 66, 87, 94, 96, 97, 112, 194, 195, 241, 243, 244, 247 Gari solida 344,345
548
Gelidiopsis intricata 91 Gelidium 367 Geonoma cuneata 479 Giffordia 91, 367, 369, 375, 377 G. mitchelliae 361, 367, 369,375, 378 Goniastrea 190,231 Goniopora, family Poritidae 154,231 Gramineae 496 Grapsus grapsus 367 Grindelia glutinosa 497 Gustavia 476,481 Gygis alba 397 Gymnodinium splendens 47-49 Haimesastraea 230 Halichoeres semicinctus 456 Haliotis corrugata 452,458 Haliotisfulgens 452 Haliotis rufescens 452 Helicosphaera sp. 40 Heliopora 232 Helioporidae 232 Heliotropium 498 Hemialus sinensis 40 Hemidiscus cuneiformis 40 Hemiptera 482 Holacanthuspasser 91 Hydnophora 23 1 H . microconos 262,270,271 Hydroides norvegica 338 Hypnaea pannosa 91 Iodothea resecata 447 Isacia conceptionis 348 Jarminocereusthouarsii 500 Jenneria pustulata 93,9598, 110 Katsuwonus pelamis 333 Kyphosus 95 Laminaria 44 1,443 Lantanacamara 398 Lantana peduncularis 500 Larosterna inca 38 1,401 Larus belcheri 400 L a m dominicanus 400 Larus fulginosus 397 Larus modestus 399 Larus occidentalis 402 Larus tridactyla 402 Lauraceae 478 Laurencia 375 Leguminosae 478,496 Leptastrea 231 Leptoseris 230 Lessonia nigrescens 336 Lessonia spp. 336 Lithophaga 93, 195,202,211 Linoraria pintado 513 Littoraria pullata 513 Littoraria schmitti 5 13 Loasaurens 496 Lobophytum 144 Ludwigia erecta 500
549
Lunda cirrhata 402 Lutjanidae 349 Lytechinus 206, 451,453 L. anamesus 453 L. semituberculatus 206 Macrocystis 433,441-450,452, 453, 460,462, 464 M. angustifolia 433,448 M. integrifolia 460 M . pyrifera 336,338,433,441,444,448,449,462 Madracis 230 Malea ringens 345 Malvaceae 496 Mastocarpuspapillatus 459 Mazama 481 Meandrina 232 Meandrinidae 232 Mediaster aequalis 133 Megabalanur 86 Meliaceae 478 Melichthys niger 437 Merluccius gayi peruanus 348 Merremia aegyptica 500 Mesodesma donacium 339, 340, 345, 352,461 Millepora 87,92, 94, 112, 113, 161, 162, 188, 194, 232, 511 M . intricata 92, 94, 112 M. platyphylla 94, 113, 51 1 Milleporidae 232 Milleporina 230,232 Mollusca 132 Montastrea 162,231,234 M . annularis 162, 175,262,273,274,276 Montipora 118, 144, 154, 173, 218,230 M . foliosa 172 Montlivaltiidae 23 1 Moraceae 478 Morus capensis 405 Mugil cephalus 348 Mugil cerema 440 Mussidae 232 Mustelus spp. 348 Mycetophyllia 232 Myliobatis spp. 348 Myristicaceae 478 Mytilus californianus 458 Mytilus edulis 458 Nannopterum harrisi 397 Nasua 481 Nectandra whitei 478 Nicotiana paniculata 496,497 Nidorellia 85,97 N.armata 97 Nitzschia bicapitata 40 Nitzschia closterium 40 Nitzschia alelicatissima 39, 40 Nitzschia longissima 40 Nitzschia sp. 40 Nolana aplocaryoides 498 Nolana confnis 496 Nolana elegans 498 Nolana spathulata 496
550 Nolanaceae 496 Nyctiphanes simplex 136 Oceanodroma hmochroa 402 Oceanodroma melania 401 Ochtodes 375 Ocotea skutchii 478 Octopusfontaneanus 346 Odontesthes regia regia 333,348 Oenocarpus mapora 479 Ophiaster hidron’us 40 Opilionids 482 Opisthonema libertate 348 Opuntia echios 500 Otaria byronia 399,418,427 Oxalis gigantea 498 Oxyjulis California 456 Oxytoxum variabile 40 Pachyptila belcheri 404 Palaua velutina 498 Palaua weberbaueri 496 Pandalusjordani 456 Panulirus gracilis 346 Panulirus interruptus 453 Paralabrax humeralis 348 Paralichthys aakpersus 348 Paralonchurus peruanus 348 Pavona 61, 63, 64,68,69, 85-87,94,99, 103, 104, 112, 184, 188,203, 206, 211-215, 230, 234 P. clavus 68, 103, 104, 112, 203, 206, 211-215, 241, 243-245, 247, 248, 262, 264, 265, 267 P. gigantea 87,203,211,214,215,241,243-245,247,248,262,266,268 P. varians 61, 64,68, 69,94, 112, 203 Placosmilia 231 Pelagophycus 441 Pelecanoides garnoti 399 Pelecanus occidentalis 386,397 Pelecanus occidentalis thagus 398 Pelecanus occidentalis californicus 440 Penaeidae 344 Penaeus brevirostris 346 Penueus californiensis 346 Perumytilus purpuratus 336 Phalacrocorax atriceps 403 Phalacrocorax bougainvillii 398,382 Phalacrocorax capensis 405 Phalacrocorax magellani 404 Phalacrocoraxpelagicus 402 Phalacrocoraxpencillams 402 Phoenicopterus chilensis 400 Phragmatopoma californica 458 Pinctada sp. 345 Pironastrea 230 Pisonia Joribunda 501 Plantyxanthus orbignyi 344,345 Platygyra 231 Platyxanthus orbignyi 344,345 Plesiastrea 231 Pleuroncodes planipes 401,440 Pocillopora 61-66, 68-70.75, 77, 78, 85-88, 92-94, 96, 97, 100, 102, 112, 113, 118, 133-136, 144, 154, 184, 188, 189, 194, 195, 203, 206, 208-210, 230, 513
55 1
P. capitata 513 P. damicornis 61, 63,65,66, 68-70,75-78, 87,92, 94, 112, 133-137, 144, 154, 172, 203, 204, 206-209 P. elegans 61, 63, 66, 68, 69, 100, 112, 203, 207-210 P. eydouxi 87 Pocilloporidae 203,230 Polinices caprae 5 13 Pollicipes elegans 337, 346 Polysiphonium 367 Pomatoceros sp. 338 Porites 61, 6 j , 64, 85-87, 94,96, 97, 99, 103-105, 112, 114, 145, 184, 188, 190, 194, 203.211-215.218.231 P. astreoides 145 P. lobata 61, 87,96, 103-105, 112, 114, 203, 211-215 P. panamensis 64,87,94, 112, 194 P. (Synaraea) 190,231 P. (Synaraea) rus 194,218 Poritidae 203,231 Porolithon 188 Portulaca pilosissima 498 Portunus acuminatus 342 Portunus asper 342 Postelsia palmaeformis 459 Potos 481 Poulsenia armata 478 Prionotus stephanophrys 349 Prioria copaifera 478 Proechimys 48 1 Prokqota 342 Psammocora 61,66,85-87,94, 112,230 P. stellata 61, 66, 87,94, 112 P. (Stephanaria)stellata 203 Psidium galapegium 501 Psocoptera 482 Pteria sterna 345 Pterodroma phaeopygia 398 Pterygophora 441,443 Ptychoramphus aleuticus 402 PufJinusopisthomelas 401 Puffinustenuirostris 402 Puriana 512 Pygoscelis papua 403 Quararibea asterolepis 478,479 Quoyula monodonta 114 Rhabdosphaera stylifer 40 Rhizosolenia alata 40 Rhizosolenia bergonii 40 Rhizosoleniafragilissima 40 Rhizosolenia sp. 40 Rhizosolenia stolterfothii 40 Rynchops niger 400 Sarcophyton 144 Sarda chiliensis chiliensis 333,334 Sarda orientalis 333 Sardinops sagax 330,334,335,398,399 Scalesia helleri 500 Scalesiapedunculata 501 Scarus spp. 95 Sciaena deliciosa 348 Sciaena gilbet-ti 348 '
552
Sciurus 481 Scolelepis 339 Scomber japonicus 332,334,398 Scomber japonicus peruanus 332 Scomberomorus sierra 333 Sebastes flavidus 440 Sebastes mystinus 440 Semele spp. 344, 345 Semicossyphus pulcher 454,455 Semimytilus algosus 336-339, 342,345 Seriatopora 144, 154, 172 Seriolella violacea 348 Siderastrea 162,230 Siderastreidae 203,230 Sideroseris 230 Siliqua patula 457 Sinularia 144 Sipuncula 132 Socratea exorrhiza 479 Solanaceae 496 Solanum multifdum 496 Spermothamnium 367 Spheniscus demersus 405 Spheniscus humboldti 400 Spheniscus magellanicus 404 Spheniscus mendiculus 397 Stegastes spp. 56,206 Sterna bergii 396 Sterna dougallii 405 Sterna elegans 406 Sternafuscata 396 Sterna lunata 397 Strongylocentrotusfranciscanus 450-452 Strongylocentrotus purpuratus 450,458,459,461 Stylinidae 230 Stylophora 154, 190,230 S.pistillata 144, 172 Sula dactylatra 396,397 Sula leucogaster 401 Sula nebouxii 381-383,386,397 Sulasula 397 Sula variegata 381,382,384, 386, 398 Sulidae 381 Syracosphaera sp. 40 Tagelus dombeii 344 Tayassu 481 Teleophrys 91 Terebra maculata roosevelti 5 13 Tetragonia ovata 498 Thais chocolata 346 Thalassiossira sp. 40 Thulassiothrin sp. 40 Thioploca 342 Thunnus albacares 333 Thysanoptera 482 Tillandsia 49 1 Tiquilia litoralis 496 Tivela stultorum 458 Trachurus murphyi 332,334 Trapezia 88,91, 118, 122
553 Trichilia tuberculata 478 Tridacna gigm 108 Trochoseris 230 Tubastrea a r e a 133 Turbinuria 232 Ulva 369 U. costata 337 U.lactuca 337 Undaria 461 U. pinnatifida 461 Uriaaalge 402 Uria lomvia 402 Virola surinamensis 478, 479 Weberbauerellabrongniam'oides 498 Xiphopenaeur riveti 345 Zalophus calijornianus 367,441 Zalophus calijornianus wollebaeki 418,426,421 Z a n t b q l m fagara 501
554 GEOGRAPHIC INDEX Africa 3, 9, 33, 151, 216, 276, 309-312, 395, 403-405, 461 Congo 9 Marcus Island 405 southern Africa 56, 146, 147, 150, 311,404,407,409 St. Croix Island 405 Amazon Basin 3,9 Amazon River 260,273-276 Amazonia 483 Andean Altiplano 400 Andean Cordillera 489 Andes 311,487,493 Antarctic ice cap 489 Arabian Gulf 150 Arabian Sea 143,145, 147, 149, 150 Arctic 461 Argentina 309,311,403,425,490 Argentinian Patagonia 406 Buenos Aires 404 Costa Bonita 404 Asia 200,295,307 Asia (Peru) 339,490 Atlantic Ocean 4, 24, 145, 146, 152, 183, 189, 190, 199,216, 238, 275, 298, 396, 403405, 514 eastern tropical Atlantic 216 North Atlantic 151,249,395 South Atlantic 404,462 South-west Atlantic 403 Southeastern Atlantic 404 tropical western Atlantic 107, 108, 111 western Atlantic 107,108, 111,506,508,509,511,512 Atlantic Oceanographic and Meteorological Laboratory (NOAA, Miami) 29 Atlas Mountains 308,310 Australasia 147, 148 Australia 2, 56,70, 143, 145, 146, 148-151, 153, 157-159, 161, 174, 175, 176, 200, 310-312,402,404 Cairns 153 Cape Cleveland 158,159 Coral Point 174 Darwin 2 Great Barrier Reef 108, 111, 141-143, 145, 149, 152, 153, 158, 160, 161, 171, 173, 174, 184, 185 Heron Island 175, 176 Lizard Island 153-155 Magnetic Island 153-156, 160 Myrmidon Reef 153,154,159 Nelly Bay 155 One Tree Island Reef 184 Peel Island 175 Queensland 145, 153, 154, 174 Sarina 174 Torres Straights 70 Townsville 28, 153-158, 160 Australian Bureau of Meteorology 156 Australian Plate 199 Bahama Islands 56, 143, 145, 151 Barbados 260,262,274-276 Belize 144, 175 Beringsea 437
555
Beringstrait 508 Bermuda 257,270 Bolivia 309,490 Borneo (Kalimantan) 473,474,480,483 Balikpapan 474 Sabah 473,474 Sebulu 474 Sepilok 474 Brazil 146, 147, 151, 273, 276, 403,425, 509 Manacauuru 274 Manaui 274 British Columbia 437,458 Barkley Sound 437 Vancouver Island 14, 15,435.437 British Isles 301 California Cooperative Oceanic Fisheries Investigations 438,439,453 Canada 292, 308,309 Caribbean Plate 191, 199 Caribbean Sea 56,57,108, 111,142,143,151, 187,198,236,272,276,400,405,406 Cenaal America 24, 109,239,266,309,349,400,508 isthmus 508 Pacific coast 109 Charles Darwin Research Station (Galapagos Islands) 60,73,74,219,264,363,378 Chile 14, 18, 32, 46, 49, 311, 324, 326, 327, 329, 330, 333, 335, 343, 351, 352, 382, 398-400,406,425,460-462,487-489,491-498 Antofagasta 18, 325, 326,400, 490,492,494 Arica (Azapa) 15,327,399,490-492 Atacama Desert 361,400,487-490,492 Caldera 15,490,498 CerroMoreno 490 Chanaral490 Cobija 490 Concepcion 327 Copiapo 490 Corral 14 El Cobre 490,498 Huasco 490 Iquique 15,489-492 Iquique-Antofagasta 339 Juan Fernandez Islands 425 LaChimba 490 La Serena 488,490,492 MiguelDiaz 490 northern Chile 93,332,334,336,339,345,346,462 Pampa de Tamarugal 490 PandeAzucar 490 Paposo 491,498,490,500 Pisagua 490 Puerto Chacabuco 332 Rio Copiapo 490 Rio Huasco 490 RioLoa 490 southern Chile 326 Talcahuano 14 Taltal490 Tarapaca 400 Tocopilla 490 Valdivia 345 Vallenar 490 China 308. 314
556 Christmas Island 396 Clipperton Island 5 13 Colombia 55-57, 59, 60, 63, 71-74, 83, 87, 88, 135, 177, 194, 239, 325, 342, 345,346,352,511 Buenaventura 325 Gorgona Island 57,72, 87, 88, 135, 187 Malpelo Island 57, 187 Utria Bay 57,72 Costa Rica 55-61, 71, 73-75,78, 81, 83, 85, 87, 88, 93-97, 110, 112-114, 116, 143, 145, 151, 152, 187, 193, 194, 216,239, 395,401,508,510 Bahia Herradura 57 Cario Island 57, 58, 75, 76, 88, 94, 97, 110, 112, 114, 194 Cocos Island 57,78,96, 114, 116, 174, 187, 193 Cocos Plate 191 Conchal 57 Golfo Dulce 57,58,61 Gulf of Nicoya 401 Gulf ofPapagayo 109,187,216 Isla Guayabo 401 Los Mogos 58,61 Malpais 57,73 Manuel Antonio 73 Playa Hermosa 57 Punta Catedral 57 Punta Islotes 61 Punta Judas 57 PuntaLeona 57 Samara 57,73 San Josecito 57 Sandal0 57,58, 61 Curaqao 145, 171 Easter Island 144 Ecuador 1, 9, 17, 18, 32, 46, 49, 55-57, 60,68, 187, 193, 201, 237, 239, 325-328, 331-335, 342, 344-346, 349, 352,361,382,399,425,460-462,464,487,488, 490,492,494,495,506 Ayangue 57,61 Esmeraldas 495 Guayaquil 325,382,495 LaLibertad 17, 18 Paita 349 Pelado Island 57,61 Quito 494,495 Enewetak Atoll 134,144,185.268 Equator 1, 12-15, 25, 29,7I,77, 129-132, 134, 136, 201, 396, 464, 490, 51 1 Eurasia 307 Europe 307,507 Falkhd Islands 424,425 Fiii 175 Frknch Frigate Shoals 397 French Polynesia 84, 141-143, 145, 148, 172 Moorea Island 145,148 Tahiti 2,129 Takapoto Atoll 148, 171 Tikehau Atoll 148, 174 Tuamotu Archipelago 145,148 Funafuti 147 Galapagos Islands 14, 15, 17,28, 31, 35-38, 55-57, 59-61, 63, 64, 66,71-75, 78, 8385, 87-90, 91-95, 97-107, 110, 112-114, 116, 141-143, 145, 146, 148, 187, 193195, 200, 201, 203,216-219, 231, 238, 239, 241, 243-249, 257, 260, 262, 264, 266,267,270,272,313, 325,326, 361, 362, 364, 367,369, 371,373, 378,
557
Galapagos Islands (continued) 383, 384, 389, 397, 399,406,407,417-421,424-428,488,493,498, 500, 501, 506. 512 Academy Bay 35, 60,71,73,74, 83, 84,92, 103, 104,238,241,243-249, 264, 363, 376 Baltra Channel 57,60 Baltra Island 362 Bartolome Island 57,60, 105 Bolivar Channel 202 Caamaiio Island 364,37 1,374 Cab0 Hammond 419,421-423 Champion Island 57,60, 238,241, 243-248 Cormorant Bay 60,107 Culpepper Island 57,60, 362 Devil's Crown 57, 60, 85, 86 El Junco Lake 200 Espaiiola Island 57,60,362, 374 Femandina Island 57,60, 101,202, 362,374,419,420,423,427,428 Floreana Island 60,83, 100-104, 107, 362, 374 Genovesa Island 60,362,374 Hood Island 238,241,243,244,247 James Bay 270,272 Isabela Island 60, 201-203, 238, 243, 260, 262, 264, 267, 362, 373, 374 Isla Daphne Major 398 Isla Floreana 398 Marchena Island 57, 60, 362, 374 h4iedo 362,364 Onslow Island 57, 60, 85, 86, 90,98, 100, 101 Pinta Island 60, 362, 374 Pinzon Island 362 Plaza Sur Island 374 Punta Espinosa 57,60 Punta Estrada 60,94 Punta Nuiiez 364 Punta Pitt 60, 104, 260,262, 264-267 Rabida Island 362 San Cristobal Island 60, 104,200,260,262,264,267,362 Santa Cruz Island 17, 35-37, 57,60,71,73,76,83,92,94, 104, 238,264, 362364,374,376,512 Santa Fe Island 60, 103-105, 361, 362, 364, 366, 371, 372, 374-378 Santiago Island 60, 103-105,270,272, 362,374 Sevmour Norte Island 374 ~ i g u s c o v e202 Urvina Bay 103, 183, 187, 196, 197, 201-205, 207, 208,211,212, 214, 215, 238, 241-243. 245.260.262.264.266.267 , , Wenman Island 57,66,362 Galapagos National Park Service 378 Gorgona Island 135 Guiana 276 Gulf of Alaska 14, 15, 17, 401, 402,440 Gulf of California 62, 261,268, 382,402,440 Gulf of Chiriqui 57-59, 61-64, 66, 67, 69, 71-76, 88, 91, 92, 94-97, 99, 101, 107, 184, 187-189, 194, 239, 241, 242-249 Gulf of Guayaquil 349,383,384 Gulf of Mexico 175 Gulf of Panama 57-59, 64, 66,70-76, 77-79, 81, 83, 84, 91, 93, 94, 101, 109, 135, 187, 189, 194,216, 239, 241, 242,244-249,260, 262,266,268, 381, 382, 384, 386,389, 391,392,510 Gulf of San Miguel 387 Gulf of Tehuantepec 5 10
558 India 309 southern India 56,147,150 Indian Ocean 2,4,107, 110,141-143,145,147,149, 150,238 Cocos Keeling Islands 174 Mayotte Island 143,150 Reunion Island 143,145,150 Indonesia 5, 56, 146, 148,150, 198, 199,491 Indonesian Ocean gateway 198 Indonesian Seaway 199 Sunda Shelf 474 Indo-Malayan archipelago 521 Instituto del Mar del Peru 324 Instituto Nacional de Pesca (Guayaquil, Ecuador) 40 InterAmerican Tropical Tuna Commission (La Jolla, California) 72 Jamaica 96,98,113, 114, 144, 171, 175-177,473,483 Japan 111, 141-143,145,149,305,307,452,461 Honshu 461 Iriomote 149 Okinawa 149,172 Ryukyus Islands 70,1 1 1 Yaeyama 149 Jarvis Island 17,397 Java Sea 141-143,145,150,151 Pulau Seribu 150 Jy Parana River 273 Kamchatka 402 Laboratory of Tree-Ring Research (Tucson, Arizona) 288,310 Large Animal Research Group (University of Cambridge) 379 Latin America 333,335 Line Islands 132-134 Long Island Sound 38 Malaya 474 Malaysia 484 Malesia 474 Mariana Islands 70, 133 Marshall Islands 70,133 Meteorological and Hydrographic Branch (Panama Canal Commission) 74,78,83 Mexico 5,62,68,109, 175, 190,200,299,301,401,402,435 Bahia Asuncion 450 Bahia Tormgas 450 Baja California 61,62,68,187,218,307,309,384,435,440,448,450,455, 457,
458,462,463,513 Baja California Sur 450 Cab0 Thurloe 435,455 El Chichon Volcano 71 Guadalupe Island 435,455 Isla Asuncion 435-448 LaPaz 218 Punta Eugenia 435 PuntaPrieta 450 San Benitos Island 435,455 southwest coast 109 Mississippi River 273 Morocco 310 Namibia 404,405 Marcus Island 405 Saldanha Bay 405 National Oceanographic and Atmospheric Administration (NOAA, Miami) 72 NauruIsland 17 New Guinea 146,148, 199
559
New Hebrides (Vanuatu) 83 New Zealand 311-313,402,403 North Island 31 1 South Island 31 1 Nicaragua 57,61, 187, 190 Brito Formation 190 North America 56, 146,291-293,298-300,303, 307-309, 311-313,407,438, 448,461,464,507 Great Basin 308 North American Plate 191 Northern Hemisphere 5, 9, 12, 14, 17, 152, 291, 307, 310, 312,434,464,514 Noumea 148 Orinoco River 260,275,276 Pacific Ocean 2-7, 9, 17, 22,24, 55, 58,68, 79, 116, 127, 129-131, 137, 238, 266, 396,436,456,457,461,490,496 central eastern Pacific 400 central equatorial Pacific 269 central Pacific 1, 9, 12, 14, 79, 107, 110, 113, 129, 132-137, 141, 143, 146-148, 151, 152, 173, 176, 183-185, 194, 219, 395, 396, 407, 408, 506, 513 dateline (equatorial) 43 eastern North Pacific 263 eastern Pacific 3, 12-15, 17, 22, 30-43, 55-57, 59, 60, 66,68,70, 71, 74-76, 79, 81, 84, 85, 87, 96, 98, 100, 103, 107-116, 129, 130, 132-137, 141-143, 145, 148, 149, 151, 152, 171, 176, 183-193, 196, 198, 200,201,203, 215-219, 234, 236, 395,400,436,460,464,483, 505, 506, 508-514 east-Pacific-Caribbean Plate 191 equatorial eastern Pacific 30-32, 35-39.56, 146,263, 309,417,463 equatorial Pacific Ocean 12, 17, 22, 23,43, 148, 151, 277, 285, 309,417, 436, 463 Indo-Pacific 1, 3, 9, 12, 56, 112, 173, 190, 192 Indo-West-Pacific 129, 130, 135, 136,510, 512, 513 North Pacific 15, 18, 134,145,261,291,292,450,461 northeastern Pacific 401,434,436-438,460,461,464 Pacific Basin 146, 184, 198, 199,218, 396 Pacific Northwest 289,291,299,301,303,310,312,436 South Pacific 15, 22, 25, 56, 83, 84, 134-145 Southeast Pacific 324, 325,398 tropical eastern Pacific 55,56,59,68,81, 108-110, 116, 135, 193,216,238, 243, 260,263,266, 313,483, 505, 506, 511,513,514 tropical Pacific 1, 3, 5, 8, 9, 11, 13, 14, 18, 22, 24, 68, 116, 128, 129, 131, 132, 313, 361,440,505 West tropical Pacific 269,270, 310 western Pacific 8, 12, 14, 17, 22, 24-26,28, 30, 55, 79, 93, 107, 110, 134, 142, 143, 147-149, 152, 161, 187, 192,217,218,505,506,509 Pacific Northwest 291,301,303,312,436 Palestine 480 Panama 28,57-67,74,80,88,235,238,239,243,245-247,249,268, 399,473,493 Balboa 58,74,78,81-83, 390-392 Balboa Heights 83 Barro Colorado Island 164, 165,473, 474, 476,479-484 Bona Island 57, 58 Cavada Island 58,63 Coiba Island 57-59 Contadora Island 58, 64, 91,260, 266, 268 Contreras Islands 57,62,69 Darien 387 Ensenada Guayabo 57 Ft. Amador Causeway 58,78 Galeta Island 165-170 Gatun Lake 164,474,476 Guayabo Grande 57
560 Panama (continued) Isthmus of Panama 83, 186, 189, 191, 192, 196,389,390 Ladrones Islands 57,58 Naos Island 58, 71, 72.78, 79, 81, 84 Panama Bay 383-385,387-389,391 Panama Bight 239,390 Panama Canal 58,74,78,79,81, 83, 116, 164,385,390,474 Panamacity 385 Panama Formation 191 Parida Island 57,58 Pearl Islands 57,58, 64, 81, 82,93 Roca San Jose 385,387,388 Saboga Island 58,81,93, 101 San Blas Islands 141-143, 145, 151, 152, 161-165, 167-170, 172 Secas Islands 57,63, 89, 94, 107 Secas reef 58,67,88, 188 Taboga Islands 57,58,69 Uraba Island 58, 69, 70, 238, 239, 244-249 Uva Island 58, 61, 62, 67, 69, 88, 89,91,92, 94, 95, 97, 99, 101, 110, 112, 238, 239,241,244-249 Uva reef 88,94,96,98, 188 Panama Bight 239 Panama Canal Commission 78,79, 81, 390 Panama-Costa Rica border 85 Panamanian Seaway 190,191,198,199 Parana River 273 Penrhyn Atoll 148 Persian Gulf 144, 175 Peru 1, 2, 5-7, 9, 11, 18, 30, 32, 33,42, 44,46, 48, 56, 57, 78, 81, 82, 93, 148, 149, 151, 196, 197, 200,272,286,306, 313, 324, 326, 328-336, 339, 342-346, 349351, 361, 362, 382-384,389,390, 395,396,399,400,404,407, 409,417,418, 425,460-462,464,487-498, 505,506,509,510 Amara 490 Ancon Bay 328,331,336-341,343 Arequipa 490 Atico 327,490 Atiquipa 490 Atocongo 490 Barranca 490 Bayovar 490 Cachendo 490 Calla0 2, 18, 28, 325, 327, 336,494 Camana 490 Canete 490 Casma 490 central Peru 327,339 Cerro Cabezon 490 Cerro Campana 490,495,496 Cerro Chimbote 490 Cerro Chiputur 490 Cerro Reque 490 CerroViru 490 Chala 490 Chaparra 490 Chicama 15,78,80-82,325,390,494 Chiclayo 490,492,495 Chimbote 325,336,342,346-349,495 Chucuito 15 Desierto de Sechura 490 Huacho 342,344,345
56 1
Peru (continued) Huarmey 325 Ica 490 Iguanil 490 110 325,490 Independencia Bay 336,347 Isla San Gallan 490 Isla San Lorenzo 490 Ite 490 Jahuay 490 Jorge Chavez International Airport 492 Lachay 490 Lagunillas 337 Lima 339,490,492-494 Lobo(s) de Tierra 382-384,427 L o b s de Afuera 427 Lomas 490 Lomas de Atiquipa 496 Lomas de Caman 496,497 Lomas de Cachendo 496 Lomas de Camana 496,497 Lomas de Chappara 497 Lomas de 110 496 Lomas de Lachay 493,496 Lomas de Mejia 498,499 Lomas de Mollendo 496 Lomas de Sama y Tacna 496 Lomas de Tacna 496,497 Lupin 490 Lurin 490 Mejia 490 Moche Valley 197 Mollendo 384,400,490,494,496 northern Peru 5.7, 344,352 Ocona 490 Paita 1, 15, 33, 325, 342, 346, 347, 349 Paracas Bay 342,347,400 Pasamayo 490 Pativilca 490 Peruvian Desert 487 Pisco 325, 337, 344-347, 349, 490, 493 Piura 9, 11,495 Puerto Chicama 15,78,81,82 Punta Coles 15 Punta San Juan 425-427 Sama Grande 490 San Juan 325,327,490 San Nicolas 490 Santa Maria del Mar 340 Santa 197 Tacna 489,490,493,496,497 Talara 494 Trujillo 490,494,495 Tumbes 495 Philippines 29,56, 133 Piura River 9, 11 Polynesia 134, 137 Red Sea 70, 144, 175,452 Republic of Kiribati 260,269 Rice University (Houston, Texas) 241,243
562 Rif mountains 310 St. Croix 273,274 TagueBay 274 Samoa 174 Smithsonian Tropical Research Institute (Panama) 78,79,474 Solomon Islands 147 South Africa 151,310,395,403,405,461 Cape Province 3 10,405 LambertBay 405 SaldanhaBay 405 South America 1, 5, 12, 14, 15, 24, 26, 27, 39, 43, 45, 47, 49, 79, 196, 200, 218, 238, 239, 266, 272, 273, 276, 277, 307, 309, 311, 313, 324, 325, 326, 339, 350, 352, 383, 384, 398,424, 425, 421, 460-462,488, 493, 494,498, 501, 502, 509, 515 northwestern South America 5577,109 Pacific coast 327 South American Plate 191 South Georgia 403 Chubut Province 403 Southern California Bight 436,438,440,443,449,453,456,462 Southern Hemisphere 1,71, 130, 309-312,462 Southern Ocean 402 Sri Lanka 56, 150 Straits of Magellan 403 Tarawa Atoll 260,262,269-271 Tasmania 3 12,402 Tobago 260,276,277 Bucco Reef 262,276 Tokelau Islands 84, 141-143, 145, 148 Nukunonu Atoll 143, 145, 148 Trombetas River 273 Tropic of Capricorn 490 TrukLagoon 192 Tutuila Island 174 U.S. Virgin Islands St. Croix 177,405 St. John 177 United States 84, 291, 292,298, 308-310, 312, 436 Alaska 14, 15, 17, 291, 309, 326, 401,402,408, 437, 440 southeastern Alaska 437 Arizona 5,310 Tucson 310 Arkansas 308 California 18, 33, 47, 136, 266, 301, 309, 401, 402,406, 407,434-459, 462-465 Central California 435, 436,440,443, 446-448,449,456,458,459, 462, 463 Channel Islands 438,441,451,453,462 Channel Islands National Park 453 Diablo Canyon 435,447 Farallon Islands 401-403.409, 435,440,459 Fort Bragg 435,440 Imperial Formation 191 King Harbor 456 La Jolla 435,438,459,462 Laguna Beach 435,443,444,449 Long Beach Harbor 440 Los Angeles 435,456 Los Angeles County 448,450 Monterey 435, 440,447, 449,451, 456, 458 Monterey Bay 440,441,447,451 Monterey County 449
563 United States (continued) California Naples 442, 447-449,454,456 NaDles Reef 442.. 447.448. . . 449,454.456 . . NebportBay 440 Palos Verdes 447,448, 450,452, 453,462 Palos Verdes Peninsula 435,448,450,452 . . Piedras Blancas 435,443,456,459 Pismo Beach 458 Point Conception 435,438 Point Dume 435,447 Point Loma 441-443, 445-448, 450,451,453,462 San Diego 132,406,435,441,447,449,451,459,465 San Francisco 435,448,462 San Miguel Islands 435,440,453 San Nicolas Island 435, 440,441,447, 448,451, 453-456,458 San Onofre 435,445,447,448,453 Santa Barbara 435,438,442,447,448,462 Santa Barbara Basin 261 Santa Barbara Channel 435,438,448 Santa Catalina Island 435,447,449,450,462 Santa Cruz 363,364,435,462 Soberanes Point 449 Southern California 309,433-436,438,440,441,443,445-449,451-454,456, 457,459,462-464 Southern California Bight 438,440,449,456,462 Florida 70. 143-145, 151, 152, 171, 175,296,298, 312, 510 Dry Tortugas 144,174,176 FloridaBay 144 Florida Keys 56,70, 143-145, 151, 152, 175, 177,257 Guam 83,144, 176, 192 Hawaii 70, 129, 133, 134, 136, 137, 175, 237 Kaneohe Bay 175 NewMexico 5 Oregon 116, 136,402,435,437,438,440,458,459 PuertoRico 145 Washington 136, 309, 435,437,457-459 Cascade Range 309 Long& 307 Tatoosh Island 435,459 Woods Hole OceanographicInstitution (WHOI) 40,241,243 Uruguay 425 Vanuatu (New Hebrides) 83 Venezuela 276 Virgin Islands 406 ZaireRiver 273
This Page Intentionally Left Blank
-
On the cover of the book, and on the two title pages, El Nino should read El Niiio.
-
page 324, 1st line: instead of "the El Niiio current" it should read El Niiio.
-
page 325, Fig. 1: All arrows placed west of "Humboldt Current (Peru Current)" should be omitted; the current by no means reaches that far out into the sea!
-
page 327, 2nd paragraph, 16th line: should read "although even there (i.e. opposed to offshore waters) the levels o f . . . ' I .
-
page 330, 4th paragraph, 6th line: 12's (not N).
-
page 335, 2nd paragraph, 3rd and 4th lines: Chile became Latin America's most important fishing nation (not "one of . . . ' I ) .
-
page 337, 11th line: delete entire sentence beginning, "In Ancon Bay,.. ' I .
-
page 341, Fig. 9, legend: "n").
-
page 342, 2nd paragraph, 11th line: instead of 100 individuals m-2 it 2 should read "I, individual per 100 m .
...
appearance of "new" species (m) (not
. ' I .
The publisher regrets the errors and emphasizes that this was neither a fault of the author or the editor.
This Page Intentionally Left Blank