THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board DANIEL L. ALKON, National Institutes of Health MEREDITH L. JONES, Smithsonian Institution and Marine Biological Laboratory GEORGE O. MACKIE, University of Victoria FREDERICK B. BANG, Johns Hopkins University JOEL L. ROSENBAUM, Yale University EDWARD M. BERGER, Dartmouth College STEVEN C. BROWN, State University of New York HoWARD A- SCHNEIDERMAN, Monsanto Company at Albany F JQHN VERNBERG, University of HARLYN O. HALVORSON, Brandeis University South Carolina J. B. JENNINGS, University of Leeds E. O. WILSON, Harvard University DR. CHARLES B. METZ, University of Miami Managing Editor VOLUME 159 AUGUST TO DECEMBER, 1980 Printed and Issued by LANCASTER PRESS, Inc. PRINCE £ LEMON STS. LANCASTER, PA. The BIOLOGICAL BULLETIN is issued six times a year at the Lancaster Press. Inc.. Prince and Lemon Streets. Lancaster, Penn- sylvania. Subscriptions and similar matter should be addressed to The Biological Bulletin, Marine Biological Laboratory, Woods : Hole Massachusetts Agent for Great Britain : Wheldon and Wesley, Limfted 2 3 and 4 Arthur Street, New Oxford Street, London, \V. C. 2. Single numbers, $10.00. Subscription per volume (thre issues), $27.00. Communications relative to manuscripts should be sent to Dr. Charles B Metz, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 between June 1 and September 1, and Charles B Metz, Institute For Molecular and Cellular Evolution University of Miami, 521 Anastasia, Coral Gables, Florida 33134 during the remainder of the year. THE BIOLOGICAL BULLETIN (ISSN 0006-3185) Second-class-postage paid at Woods Hole, Mass., and additional mailing offices. LANCASTER PRESS, INC.. LANCASTER. PA CONTENTS No. 1, AUGUST 1(>SO ANNUAL REPORT OF THE MARINE BIOLOGICAL LABORATORY 1 ARNOLD, JOHN M., AND Lois D. WILLIAMS-ARNOLD Development of the ciliature pattern on the embryo of the squid. Loligo [>calci: A scanning electron microscope study 102 BIGGER, CHARLES H. Interspecific and intraspecific acrorhagial aggressive behavior among sea anemones : A recognition of self and not-self 117 HATCH. WALTER I. The implication of carbonic anhydrase in the physiological mechanism of penetration of carbonate substrata by the marine burrowing sponge Cliona cclota (Demospongiae) 135 HILDRETH, JANE E., AND WILLIAM B. STICKLE The effects of temperature and salinity on the osmotic composition of the southern oyster drill, Thais haemastoma 148 NELSON, KEITH, DENNIS HEDGECOCK, WILL BORGESON, ERIC JOHNSON, RICHARD DAGGETT, AND DIANE ARONSTEIN Density-dependent growth inhibition in lobsters. Homarus (Decapoda, Nephropidae) 162 SEIGFRIED, CLIFFORD A. Seasonal abundance and distribution of Crangon jranciscorum and Palaemon macrodactylus (Decapoda, Caridae) in the San Francisco Bay-Delta 177 SEIGFRIED, CLIFFORD A., AND MARK E. KOPACHE Feeding of Neomysis mercedis ( Holmes) 193 STEEL, C. G. H. Mechanisms of coordination between moulting and reproduction in ter- restrial isopod Crustacea 206 SUGIMOTO, KEIJI, AND HlROSHI WATANABE Studies on reproduction in the compound ascidian, Symplcyina re plans: Relationship between neural complex and reproduction 219 WARNER, JON A., AND JAMES F. CASE The zoogeography and dietary induction of bioluminescence in the mid- shipman fish, Porichthys notatns 231 WICKHAM, DANIEL E. Aspects of the life history of Carcinonemertes errans (Nemertea: Car- cinonemertidae), an egg predator of the crab Cancer magister 247 No. 2, OCTOBER 1980 ANDERSON, ROBERT S. Hemolysins and hemagglutinins in the coelomic fluid of a polychaete annelid, Glycera dibranchiata 259 iii iv CONTENTS HADLOCK, ROBIN P. Alarm response of the intertidal snail Liitorina littorea (L.) to predation by the crab Carcinus maenas (L.) 269 HOUR, MARGARET S. AND RALPH T. HINEGARDNER The formation and early differentiation of sea urchin gonads 280 KANESHIRO. EDNA S. AND RICHARD D. KARP The ultrastructure of coelomocytes of the sea star Dermasterias imbricata 295 MARCUS, NANCY H. Photoperiodic control of diapause in the marine calanoid copepod Labi- doc era aestiva ^ MATSUMOTO, GEN AND JUNICHI SHIMADA Further improvement upon maintenance of adult squid (Doryteuthis blcckeri) in a small circular and closed-system aquarium tank 319 McCoRKLE, SUSAN AND THOMAS H. DIETZ Sodium transport in the freshwater Asiatic clam Corbicula fluminea 325 MERCANDO, NEIL A. AND CHARLES F. LYTLE Specificity in the association between H \dractinia cchinata and sympatric species of hermit crabs MILLER, CHARLES B., DAVID M. NELSON, ROBERT R. L. GUILLARD, AND BONNIE L. WOODWARD Effects of media with low silicic acid concentrations on tooth formation in Acartia tpnsa Dana (Copepoda, Calanoida) MORAN, WILLIAM M. AND RICHARD E. TULLIS Ion and water balance of the hypo- and hyperosmotically stressed chiton Mopalia miiscosa OHTSU, KOHZOH Electrical activities in the subtentacular region of the anthomedusan Spirocodon saltatrix (Tilesius) RAHAT, M. AND ORIT ADAR Effect of symbiotic zooxanthellae and temperature on budding and strobi- lation in Cassiopeia andromeda (Eschscholz) 39' SULKIN, S. D., W. VAN FlUEKELEM, P. KELLY, AND L. VAN HUEKELEM The behavioral basis of larval recruitment in the crab Callinectes sapidus Rathbun : A laboratory investigation of ontogenetic changes in geotaxis and barokinesis TURNER, KATHERINE AND TIMOTHY A. LYERLA Electrophoretic variation in sympatric mud crabs from North Inlet, South r« i- 418 Carolina YOUNG, CRAIG M. AND LEE F. BRAITHWAITE Orientation and current-induced flow in the stalked ascidian Styela 428 montereyensis ABSTRACTS OF PAPERS PRESENTED AT THE MARINE BIOLOGICAL LABORATO No. 3, DECEMBER 1980 503 ERRATUM Invited Article: ALKON, DANIEL L. tie Cellular analysis of a gastropod (Hermissenda crassicornis) associative learning CONTENTS v ALEXANDER, DAVID E., AND JOE GHIOLD The functional significance of the lunules in the sand dollar, Mellita qmn- quiesperjorata - BANG, BETSY G., AND FREDERIK B. BANG The urn cell complex of Sipitncnlns nitdus: A model for study of mucus- stimulating substances c7l BANNON, GARY A., AND GEORGE GORDON BROWN Ultrastructural characteristics of the non-expanded and expanded extra- embryonic shell of the horseshoe crab, Limulits polyphemus L 582 BOUSQUETTE, GEORGE D. The larval development of Pinnixa longipes (Lockington, 1877) (Brach- yura : Pmnotheridae), reared in the laboratory 592 BRETOS, MARTA Age determination in the keyhole limpet Fissitrclla crassa Lamarck (Archaeogastropoda : Fissurellidae), based on shell growth rings 606 CASE, JAMES F. Courting behavior in a synchronously flashing, aggregative firefly Pteroptyx tener ( , , COSTA, CHARLES J., SIDNEY K. PIERCE, AND M. KIM WARREN The intracellular mechanism of salinity tolerance in polychaetes : Volume regulation by isolated Glycera dibranchiata red coelomocytes 626 FISHER, CHARLES R., JR., AND ROBERT K. TRENCH In vitro carbon fixation by Prochloron sp. isolated from Diplosoma virens 636 GOODENOUGH, JUDITH E., AND VlCTOR G. BRUCE The effects of caffeine and theophylline on the phototactic rhythm of Chlamydomonas reinhardii HAWKINS, CLIFFORD J., PAULINE M. MEREFIELD, DAVID L. PARRY, WILTON R. BIGGS, AND JAMES H. SWINEHART Comparative study of the blood plasma of the ascidians Pyura stolonifera and Ascidia ceratodes HAWKINS, CLIFFORD J., DAVID L. PARRY, AND CRAIG PIERCE Chemistry of the blood of the ascidian Pododavella molitccensis 669 LEE, PAUL H., AND JUDITH S. WEIS Effects of magnetic fields on regeneration in fiddler crabs 681 MCNAMARA, JOHN C, GLORIA S. MOREIRA, AND PLINIO S. MOREIRA Respiratory metabolism of Macrobrachium olfcrsii (Wiegmann) zoeae during the moulting cycle from eclosion to first ecdysis 692 MlTTENTHAL, JAY EDWARD On the form and size of crayfish legs regenerated after grafting 700 MlTTENTHAL, JAY EDWARD, MARY C. OLSON, AND GLENNA D. CUM MINGS Morphology of the closer muscles in normal and homoeotic legs of crayfish 7^ MORI, TAKAO, TEIZO TSUCHIYA, AND SHONAN AMEMIYA Annual gonadal variation in sea urchins of the orders Echinothurioida and Echinoida 728 STUNKARD, HORACE W. The morphology, life history, and systematic relations of Titbit lovesicufa pinguis (Linton, 1940) Manter, 1947 (Trematoda: Hemiuridae) 737 • CONTENTS WERMUTH, TEROME F. Gamma" radiation and hydranth longevity in Campanularia flexuosa: Age-dependency of dose-response function 752 VroiN, ASHLEY I.. RICHARD A. DIENER, W. H. CLARK. JR., AND ERNEST S. CHANG Mandibular gland of the blue crab. Callinectes sapidus 760 INDEX TO VOLUME 159 773 Volume 159 Number 1 THE BIOLOGICAL BULLETIN Rjclngicaf labimiiory ? A R Y Woods Hole, Mass. PUBLISHED BY THE MARINE BIOLOGICAL Editorial DANIEL L. ALKON, National Institutes of Health MEREDITH L. JONES, Smithsonian Institution and Marine Biological Laboratory GEORGE O. MACKIE, University of Victoria FREDERICK B. BANG, Johns Hopkins University JOEL L. ROSENBAUM, Yale University EDWARD M. BERGER, Dartmouth College STEPHEN C. BROWN, State University of New York HOWARD A. SCHNEIDERMAN, Monsanto Company at Albany F. JOHN VERNBERG, University of HARLYN O. HALVORSON, Brandeis University South Carolina J. B. JENNINGS, University of Leeds E. O. WILSON, Harvard University Managing Editor: CHARLES B. METZ, University of Miami AUGUST, 1980 Printed and Issued by LANCASTER PRESS, Inc. PRINCE &. LEMON STS. LANCASTER, PA. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is published six times a year by the Marine Biological Laboratory, MBL Street, Woods Hole, Massachusetts 02543. Subscriptions and similar matter should be addressed to THE BIOLOGICAL BULLETIN, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain: Wheldon and \\Vsley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, S10.00. Subscription per volume (three issues), $27.00, (this is $54.00 per year for six issues). Communications relative to manuscripts should be sent to Dr. Charles B. Metz, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 between June 1 and September 1, and to Dr. Charles B. Metz, Institute For Molecular and Cellular Evolution, University of Miami, 521 Anastasia, Coral Gables, Florida 33134 during the remainder of the year. Assistant editor, Susan Schwartz. Copyright © 1980, by the Marine Biological Laboratory Second-class postage paid at Woods Hole, Mass., and additional mailing offices. ISSN 0006-3185 INSTRUCTIONS TO AUTHORS THE BIOLOGICAL BULLETIN accepts original research reports of intermediate length on a variety of subjects of biological interest. In general, these papers are either of particular interest to workers at the Marine Biological Laboratory, or of outstanding general significance to a large number of biologists throughout the world. Normally, review papers (except for a limited number of so- licited review papers which may be accepted after formal refereeing) , very short papers (less than five printed pages), preliminary notes, and papers which describe only a new technique or method without presenting substantial quantities of data resulting from the use of the new method can- not be accepted for publication. A paper will usually appear within four months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below; those manuscripts which do not conform will be returned to authors for correction before review. 1. Manuscripts. Manuscripts should be submitted in triplicate. The original must be typed in double spacing (including figure legends, foot-notes, bibliography, etc.) on one side of 16- or 20-lb. bond paper, 85 by 11 inches. They should be carefully proof-read and errors corrected legibly in black ink. Pages should be numbered. A left-hand margin of at least 1? inches should be allowed. Manuscripts should be divided into the following components: Title page, Intro- duction, Materials and Methods, Results, Discussion, Acknowledgments, Summary, Literature Cited, Tables (with consecutive Roman numerals) and Figure Legends (with consecutive Arabic numerals). 2. Tables, Foot-Notes, Figure Legends, etc. Tables should be typed on separate sheets and placed after the Literature Cited. Because of the high cost of setting such material in type authors are earnestly requested to limit tabular material as much as possible. Similarly, foot- notes to tables should be avoided wherever possible. If they are essential, they should be indi- cated by asterisks, daggers, etc., rather than by numbers. Foot-notes are not normally permitted in the body of the text. Such material should be incorporated into the text where appropriate. Kxplanations of figures should be typed double-spaced and placed on separate sheets at the end of the paper. 3. A condensed title or running head of no more than 35 letters and spaces should be included. 4. Literature Cited. The list of references should be headed LITERATURE CITED, should conform in punctuation and arrangement to the style of recent issues of THE BIOLOGICAL BULLETIN, and must be typed double-spaced on separate pages. Note that citations should Continued on Cover Three Vol. 159, No. 1 August 1980 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY THK MARINE BIOLOGICAL LABORATORY EIGHTY-SECOND REPORT, FOR THK YEAR 1979 — NINETY-SECOND YEAR I. TRUSTEES AND STANDING COMMITTEES 1 II. MEMBERS OF THE CORPORATION 5 III. CERTIFICATE OF ORGANIZATION 35 IV. ARTICLES OF AMENDMENT 36 V. BYLAWS 37 VI. REPORT OF THE DIRECTOR 42 VI I. REPORT OF THE TREASURER 49 VIII. REPORT OF THE LIBRARIAN 62 IX. EDUCATIONAL PROGRAMS 62 1 . SUMMER 62 2. JANUARY 70 3. SHORT COURSES 75 X. RESEARCH AND TRAINING PROGRAMS 80 1 . SUMMER 80 2. YEAR-ROUND 89 XI . HONORS 94 1 . FRIDAY EVENING LECTURES 94 2. FELLOWSHIPS AND SCHOLARSHIPS 95 XII. INSTITUTIONS REPRESENTED 95 XIII. LABORATORY SUPPORT STAFF. 99 I. TRUSTEES Including Action of the 1979 Annual Meeting OFFICERS PROSSER GIFFORD, Chairman of the Board of Trustees, \Voodro\v Wilson International Center for Scholars, Smithsonian Building, Washington, D. C. 20560 GERARD S\VOPE, JR., Honorary Chairman of the Board of Trustees, Croton-on-Hudson, New York 10520. (Deceased September 27, 1979) DENIS M. ROBINSON, Honorary Chairman of the Board of Trustees, High Voltage Kn- gineering Corporation, Burlington, Massachusetts 01803 Copyright © 1980, by the Marine Biological Laboratory Library of Congress Card No. A38-518 (ISSN 0006-3185) 2 MARINE BIOLOGICAL LABORATORY ROBERT MAIXER, Treasurer, Boston Company, One Boston Place, Boston, Massa- chusetts 02106 PAUL R. GROSS, President of the Corporation and Director of the Laboratory, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 GERALD S. FISCHBACH, Clerk of the Corporation. Harvard Medical School, Boston, Massachusetts 02215 EMERITI PHILIP B. ARMSTRONG, 51 Elliott Place, Rutherford, New Jersey 07070 ERIC G. BALL, Marine Biological Laboratory (Deceased September 4, 1979) DUGALD E. S. BROWN, 38 Whitman Road, Woods Hole, Massachusetts 02543 FRANK A. BROWN, Jr., Woods Hole, Massachusett 02543 JOHN B. BUCK, National Institutes of Health AURIN CHASE, Princeton University ANTHONY C. CLEMENT, Emory University KENNETH S. COLE, 2404 Loring Street, San Diego, California 92109 ARTHUR L. COLWIN, University of Miami LAURA H. COLWIN, University of Miami D. EUGENE COPELAND, Marine Biological Laboratory SEARS CROWELL, Indiana University HARRY GRUNDFEST, College of Physicians and Surgeons TERU HAYASHI, University of Bonn, West Germain HOPE HIBBARD, Oberlin College LEWIS KLEINHOLZ, Reed College M. E. KRAHL, Tucson, Arizona DOUGLAS MARSLAND, Cockeysville, Maryland 21030 HAROLD H. PLOUGH, Amherst, Massachusetts C. LADD PROSSER, University of Illinois JOHN S. RANKIN, Jr., Ashford, Connecticut A. C. REDFIELD, Woods Hole, Massachusetts S. MERYL ROSE, Waquoit, Massachusetts MARY SEARS, Woods Hole, Massachusetts CARL C. SPEIDEL, University of Virginia H. BURR STEINBACH, Woods Hole, Massachusetts ALBERT SZENT-GYORGYI, Marine Biological Laboratory W. RANDOLPH TAYLOR, University of Michigan GEORGE WALD, Harvard University CLASS OF 1983 XINA ALLEN, Dartmouth College HAYS CLARK, Avon Products, Incorporated DENNIS FLANAGAN, Scientific American WILLIAM T. GOLDEN, New York, New York PHILIP GRANT, University of Oregon JOEL ROSENBAUM, Yale University ANN STUART, University of North Carolina, Chapel Hill ANDREW SZENT-GYORGYI, Brandeis University KENSAL VAN HOLDE, Oregon State University CLASS OF 1982 EVERETT ANDERSON, Harvard Medical School GEORGE H. A. CLOWES, JR., The Cancer Research Institute, Boston ELLEN R. GRASS, The Grass Foundation JOHN P. KENDALL, Boston, Massachusetts TRUSTEES AND STANDING COMMITTEES EDWARD A. KRAVITX, Harvard Medical School HANS LAUFER, University of Connecticut MARJORIE R. STETTEN, National Institutes of Health WALTER S. VINCENT, University of Delaware J. RICHARD \YHITTAKKR, \Yistar Institute CLASS OF l<>,xi JOHN M. ARNOLD, University of Hawaii JANE FESSENDEN, Marine Biological Laboratory HELEN HOMANS GILBERT, Dover, Massachusetts STEPHEN \Y. KUFFLER, Harvard Medical School MAURICE LAZARUS, Federated Department Stores, Boston GEORGE PAPPAS, LJniversity of Illinois Medical Center VY. D. RUSSELL-HUNTER, Syracuse University RAYMOND E. STEPHENS, Marine Biological Laboratory CLASS OF PHILIP B.DUNHAM, Syracuse University TIMOTHY H. GOLDSMITH, Yale University BENJAMIN KAMINER, Boston University GEORGE MULLEN, Watertown, Massachusetts KEITH R. PORTER, University of Colorado LIONEL I. REBHUN, University of Virginia VV. NICHOLAS THORNDIKE, Boston, Massachusetts EDWARD O. WILSON, Harvard University STANDING COMMITTEES EXECUTIVE COMMITTEE OF THE BOARD OF TRUSTEES PROSSER GIFFORD ex officio MARJORIE R. STETTEN, 1982 PAUL R. GROSS, ex officio EDWARD A. KRAVITZ, 1981 ROBERT MAINER, ex officio KENSAL VAN HOLDE, 1981 JOEL ROSENBAUM, 1982 PHILIP GRANT, 1980 HANS LAUFER, 1980 BUDGET COMMITTEE JOHN M. ARNOLD, Chairman ROBERT MAINER, ex officio GEORGE H. A. CLOWES, JR. JEROME SCHIFF PAUL R. GROSS, ex officio HOMER P. SMITH, ex officio WILLIAM T. GOLDEN WALTER S. VINCENT BUILDINGS AND GROUNDS COMMITTEE FRANCIS HOSKIN, Chairman AUDREY E. V. HASCHEMEYER FRANK CHILD FRANK LONGO LAWRENCE B. COHEN ROBERT D. PRUSCH ALAN FEIN RONALD PRZYBYLSKI DANIEL GILBERT ALLEN SCHUETZ ROBERT GUNNING, ex officio EVELYN SPIEGEL MARINE BIOLOGICAL LABORATORY COMPUTER COMMITTEE JOHN HOBBIE, Chairman WILLIAM J. ADELMAN FRANCIS P. BOWLES ALLAH VERDI FAKMANFARMAIAN E. F. MAcXiCHOL, JR. MELVIN ROSENFELD, JR. CONSTANTINE TOLLIOS EMPLOYEE RELATIONS COMMITTEE JOAN HOWARD, Chairman CAROL EBERHARD CHARLOTTE FRANK JUDITH GRASSLE LEWIS LAWDAY DONALD LEHY RAYMOND STEPHENS HOUSING, FOOD SERVICE, AND DAY CARE COMMITTEE ANN STUART, Chairman DAN ALKON NINA ALLEN ROBERT BARLOW MONA GROSS RONALD JOYNER AIMLEE LADERMAN BRIAN SALZBERG HOMER P. SMITH, ex officio INSTRUCTION COMMITTEE BENJAMIN KAMINER, Chairman DANIEL ALKON ROBERT D. ALLEN ESTHER GOUDSMIT ARTHUR HUMES ROBERT K. JOSEPHSON MORTON MASER, ex officio MERLE MIZELL JOEL ROSENBAUM SHELDON J. SEGAL GEORGE WOODWELL INVESTMENT COMMITTEE W. NICHOLAS THORNDIKE, Chairman PROSSER GIFFORD, ex officio WILLIAM T. GOLDEN MAURICE LAZARUS JOHN ARNOLD ROBERT MAINER, ex officio LIBRARY COMMITTEE' EDWARD A. ADELBERG, Chairman FRANK M. CHILD BERNARD DAVIS JOHN DOWLING JANE FESSENDEN, ex officio NICHOLAS P. FOFONOFF J. W. HASTINGS SHINYA INOUE STEPHEN W. KUFFLER LUIGI MASTROIANNI, JR. ROBERT MORSE ARTHUR PARDEE KEITH R. PORTER SHELDON J. SEGAL DEREK W. SPENCER MARJORIE R. STETTEN STANLEY WATSON CAROLYN P. WINN, ex officio MACY SCHOLARSHIP COMMITTEE WILLIAM W. SUTTON, Chairman LOWELL DAVIS MORTON D. MASER, ex officio JAMES PERKINS EDGAR E. SMITH JAMES TOWNSEL WALTER S. VINCENT CHARLES WALKER MEMBERS OF THE CORPORATION 5 MARINK RESOURCES COMMITTEE SEARS CROWELL, Chair man ROBERT PRENDERGAST CARL J. BERG ROBERT I). PRUSCII JUNE HARRIGAN JOHN S. RANKIN TOM HUMPHREYS JOHN YALOIS, ex officio IACK LEVIN JONATHAN \\"ITTENBERG CYRUS LEVINTHAL RADIATION COMMITTEE WALTER S. VINCENT, Chairman JOHN HOBBIE EUGENE BELL E. F. MAcXicHOi., JR. FRANCIS P. BOWLES MORTON I). MASER, ex officio RICHARD L. CHAPPELL HARRIS RIPPS PAUL DE\\'EER RESEARCH SERVICES COMMITTEE NINA S. ALLEN, Chairman GEORGE PAPPAS FRANCIS P. BOWLES JEROME SCHIFF SAMUEL S. KOIDE R. BRUCE SZAMIER E. F. MAcNiCHOL, JR. JAY WELLS MORTON D. MASER, ex officio SEYMOUR ZIGMAN TOSHIO XARAHASHI RESEARCH SPACE COMMITTEE TIMOTHY GOLDSMITH, Chairman MORTON D. MASER, ex officio CLAY ARMSTRONG JOEL ROSENBAUM JOHN ARNOLD JOAN RUDERMAN ARTHUR DuBois BRIAN SALZBERG PHILIP GRANT ANN STUART FREDERICK GRASSLE GEORGE WOODWELL GEORGE LANGFORD SAFETY COMMITTEE ROBERT GUNNING, Chairman ex officio MORTON D. MASER, ex officio DANIEL ALKON RAYMOND STEPHENS LEWIS LAWDAY PAUL STEUDLER DONALD LEHY FREDERICK THRASHER E. F. MACNICHOL, JR. JAY WELLS II. MEMBERS OF THE CORPORATION Including Action of the 1979 Annual Meeting LIFE MEMBERS ADOLPH, DR. EDWARD F., University of Rochester, School of Medicine and Dentistry, Rochester, New York 14642 BEAMS, DR. HAROLD W., Department of Zoology, University of Iowa, Iowa City, Iowa 52242 BEHRE, DR. ELLINOR H., Black Mountain, North Carolina 28711 6 MARINE BIOLOGICAL LABORATORY BERTHOLF, DR. LLOYD M., Westminster Village Apt. 2114, 2025 E. Lincoln St., Bloomington, Illinois 61701 BISHOP, DR. DAVID W., Department of Physiology, Medical College of Ohio, Toledo, Ohio 43699 BOLD, DR. HAROLD C., Department of Botany, University of Texas, Austin, Texas 78712 BUIDGMAN, DR. A. JOSEPHINE, 715 Kirk Rd., Decatur, Georgia 30030 BROWN, DR. DUGALD E. S., 38 Whitman Rd., Woods Hole, Massachusetts 02543 BURDICK, DR. C. LALOR, 900 Barley Drive, Barley Mill Court, Wilmington, Dela- ware 19807 CARPENTER, DR. RUSSELL L., 60 Lake St., Winchester, Massachusetts 01890 CHASE, DR. AURIN, Professor of Biology, Emeritus, Princeton University, Princeton, New Jersey 08540 CLARKE, DR. GEORGE L., 44 Juniper Road, Belmont, Massachusetts 02178 CLEMENT, DR. ANTHONY C., Department of Biology, Emory University, Atlanta, Georgia 30322 COLE, DR. KENNETH S., 2404 Loring St., San Diego, California 92109 COLWIN, DR. ARTHUR L., 320 Woodcrest Rd., Key Biscayne, Florida 33149 COLWIN, DR. LAURA H., 320 Woodcrest Rd., Key Biscayne, Florida 33149 COPELAND, DR. D. E., 41 Fern Lane, Woods Hole, Massachusetts 02543 COSTELLO, DR. HELEN M., 507 Monroe St., Chapel Hill, North Carolina 27514 CROUSE, DR. HELEN V., Institute of Molecular Biophysics, Florida State Uni- versity, Tallahassee, Florida 32306 DILLER, DR. IRENE C., 2417 Fairhill Avenue, Glenside, Pennsylvania 19038 DILLER, DR. WILLIAM F., 2417 Fairhill Avenue, Glenside, Pennsylvania 19038 ELLIOTT, DR. ALFRED M., P. O. Box 564, Woods Hole, Massachusetts 02543 FERGUSON, DR. JAMES K. W., 56 Clarkson St., Thornhill, Ontario, L4J 2B4 Canada FISCHER, DR. ERNST, M.D., 3110 Manor Drive, Richmond, Virginia 23230 FRIES, DR. ERIK F. B., 3870 Leafy Way, Miami, Florida 33133 GRAY, DR. IRVING E., Department of Zoology, Duke University, Durham, North Carolina 27701 GRUNDFEST, DR. HARRY, Department of Neurology, Columbia University, College of Physicians and Surgeons, New York, New York 10032 GUTTMAN, DR. RITA, 75 Henry St., Brooklyn, New York 11210 HAMBURGER, DR. VIKTOR, Professor Emeritus, Washington University, St. Louis, Missouri 63130 HAMILTON, DR. HOWARD L., Department of Biology, University of Virginia, Charlottesville, Virginia 22901 HARTLINE, DR. H. KEFFER, The Rockefeller University, New York, New York 10021 HIBBARD, DR. HOPE, 143 E. College St., Apt. 309, Oberlin, Ohio 44074 HISAW, DR. F. L., 5925 S. W. Plymouth Drive, Corvallis, Oregon 97330 HOLLAENDER, DR. ALEXANDER, Associated University, Inc., 1717 Massachusetts Ave., N. W., Washington, D. C. 20036 HuGHES-ScHRADER, DR. SALLY, Department of Zoology, Duke University, Durham, North Carolina 27706 IRVING, DR. LAURENCE, University of Alaska, College, Alaska 99701 JOHNSON, DR. FRANK H., Department of Biology, Princeton University, Prince- ton, New Jersey 08540 KAAN, DR. HELEN, 62 Locust St., Apt. 244, Falmouth, Massachusetts 02540 MEMBERS OF THE CORPORATION 7 KAHLER, ROBERT, P. O. Box 423, Woods Hole, Massachusetts 02543 KILLE, DR. FRANK R., 500 Osceola Ave., Winter Park, Florida 32789 KLEINHOLZ, DR. LEWIS, Department of Biology, Reed College, Portland, Oregon 97202 LOCHHEAD, DR. JOHN H., 49 Woodlawn Rd., London, SW6 6PS, England U. K. LYNN, DR. W. GARDNER, Department of Biology, Catholic University of America, Washington, D. C. 20017 MAGRUDER, DR. SAMUEL R., 270 Cedar Lane, Paducah, Kentucky 42001 MALONE, DR. E. P., 6610 North llth Street, Philadelphia, Pennsylvania 19126 MANWELL, DR. REGINALD D., Department of Biology, Syracuse University, Syracuse, New York 13210 MARSLAND, DR. DOUGLAS, Broadmead N12, 13801 York Rd., Cockeysville, Maryland 21030 MILLER, DR. JAMES A., 307 Shorewood Dr., E. Falmouth, Massachusetts 02536 MILNE DR. LORUS J., Department of Zoology, University of New Hampshire, Durham, New Hampshire 03824 MOUL, DR. E. T., 42 F. R. Lillie Rd., Woods Hole, Massachusetts 02543 NACHMANSOHN, DR. DAVID, M.D., Department of Neurology, College of Physi- cians and Surgeons, Columbia University, 630 W. 168th St., New York, New York 10032 NICOLL, DR. PAUL A., 6636 E. Street Rd. 46, Bloomington, Indiana 47401 PAGE, DR. IRVING H., Box 516, Hyannisport, Massachusetts 02647 PLOUGH, DR. HAROLD H., 31 Middle Street, Amherst, Massachusetts 01002 POLLISTER, DR. A. W., Box 23, Dixrield, Maine 04224 POND, SAMUEL E., P. O. Box 63, E. Winthrop, Maine 04343 PORTER, DR. H. C., University of Pennsylvania, Philadelphia, Pennsylvania 19174 PRYTZ, DR. MARGARET McD., 21 McCouns Lane, Oyster Bay, New York 11771 RENN, DR. CHARLES E., Route 2, Hampstead, Maryland 21074 REZNIKOFF, DR. PAUL, M.D., 11 Brooks Rd., Woods Hole, Massachusetts 02543 RICHARDS, DR. A. GLENN, Department of Entomology, University of Minnesota, St. Paul, Minnesota 55101 RICHARDS, DR. OSCAR W., Pacific University, Forest Grove, Oregon 97116 SCHARRER, DR. BERTA, Department of Anatomy, Albert Einstein College of Medicine, 1300 Morris Pkwy., New York, New York 10461 SCHMITT, DR. F. O., 165 Allen Dale St., Jamaica Plain, Massachusetts 02130 SHEMIN, DR. DAVID, Department of Biochemistry and Molecular Biology, Northwestern University, Evanston, Illinois 60201 SICHEL, DR. ELSA K., 4 WHITMAN Rd., Woods Hole, Massachusetts 02543 SMITH, DR. DIETRICH C., 216 Oak Forest Ave., Catonsville, Maryland 21228 SONNEBORN, DR. T. M., Department of Zoology, Indiana University, Blooming- ton, Indiana 47401 SPEIDEL, DR. CARL D., 1873 Field Rd., Charlottesville, Virginia 22093 STEINBACH, DR. H. B., One Bell Tower Lane, Woods Hole, Massachusetts 02543 STRAUS, DR. W. L., JR., Department of Anatomy, The Johns Hopkins University Medical School, Baltimore, Maryland 21205 STUNKARD, DR. HORACE W., American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024 TAYLOR, DR. W. RANDOLPH, Department of Botany, University of Michigan, Ann Arbor, Michigan 48109 DR. Lois E., 4 Sanderson Ave., Northampton, Massachusetts 01060 8 MARINE BIOLOGICAL LABORATORY TRAVIS, DR. DOROTHY, 733 Sligo Ave., Apt. 503, Silver Spring, Maryland 20910 WALD, DR. GEORGE, Higgins Professor of Biology, Emeritus, Harvard University, Cambridge, Massachusetts 02138 WAI. LACK, DR. LOUISE B., 359 Lytton Ave., Palo Alto, California 94301 WARREN, DR. HERBERT S., % Leland C. Warren, 721 Conshohocken State Road, Penn Valley, Pennsylvania 19072 WKI>S, DR. PAUL, The Rockefeller University, 66th St. and York Ave., New York, New York 10016 WHITING, DR. ANNA R., WOODS HOLE, Massachusetts 02543 WICHTERMAN, DR. RALPH, 31 Buzzards Bay Ave., Woods Hole, Massachusetts 02543 Young, Dr. D. B., Main Street, North Hanover, Massachusetts 02357 ZINN, DR. DONALD J., P. O. Box 589, Falmouth, Massachusetts 02541 ZORZOLI, DR. ANITA, Department of Biology, Vassar College, Poughkeepsie, New York 12610 ZWEIFACH, DR. BENJAMIN W., % Ames, University of California, La Jolla, California 92037 REGULAR MEMBERS ABBOTT, DR. MARIE B., High St., Coventry, Connecticut 06238 ACHE, DR. BARRY W., Whitney Marine Laboratory, University of Florida, Rt. 1, Box 121, St. Augustine, Florida 32084 ACHESON, DR. GEORGE H., 25 Quisset Ave., Woods Hole, Massachusetts 02543 ADEJUWON, DR. CHRISTOPHER A., Chemical Pathology Dept., University of Ibadan, Ibadan, Nigeria ADELBERG, DR. EDWARD A., Department of Human Genetics, Yale University Medical School, New Haven, Connecticut 06511 AFZELIUS, DR. BJORN, Wenner-Gren Institute, LIniversity of Stockholm, Stock- holm, Sweden ALKON, DR. DANIEL, M. D., Head, Section on Neural Systems, Laboratory of Biophysics, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 ALLEN, DR. GARLAND E., Department of Biology, Washington LIniversity, St. Louis, Missouri 63130 ALLEN, DR. NINA S., Department of Biology, Dartmouth College, Hanover, New Hampshire 03755 ALLEN, DR. ROBERT D., Department of Biology, Dartmouth College, Hanover, New Hampshire 03755 ALSCHER, DR. RUTH, Department of Biology, Manhattanville College, Purchase, New York 10577 AMATNIEK, ERNEST, 4797 Boston Post Rd., Pelham Manor, New York 10803 ANDERSON, DR. EVERETT, Department of Anatomy and Laboratories of Human Reproductive Biology, Harvard Medical School, Boston, Massachusetts 02115 ANDERSON, DR. J. M., Division of Biological Sciences, Emerson Hall, Cornell University, Ithaca, New York 14853 ARMSTRONG, DR. CLAY M., Department of Physiology, University of Pennsyl- vania, School of Medicine, Philadelphia, Pennsylvania 19174 ARMSTRONG, DR. PP:TER B., Department of Zoology, University of California, Davis, California 95616 MEMBERS OF THE CORPORATION i) ARMSTRONG, DR. PHILLIP B., M.D., 51 Klliott Place, Rutherford, New Jersey 07070 ARNOLD, DR. JOHN MILLER, Kewalo Marine Lab., Pacific Biomedical Research Center, 41 Ahui St., Honolulu, Hawaii 96813 ARNOLD, DR. WILLIAM A., 102 Balsam Rd., Oak Ridge, Tennessee 37830 ATEMA, DR. JELLE, 9 Millfield St., Woods Hole, Massachusetts 02543 ATWOOD, DR. KIMBALL C., 100 Haven Ave., Apt. 21-E, New York, New York 10032 AUSTIN, DR. MARY L., 506| North Indiana Avenue, Bloomington, Indiana 47401 BACON, ROBERT, P. O. Box 723, Woods Hole, Massachusetts 02543 BALDWIN, DR. THOMAS O., Department of Biochemistry, University of Illinois, Urbana, Illinois 61801 BANG, MRS. BETSY, Department of Pathobiology, The Johns Hopkins University, School of Hygiene, Baltimore, Maryland 21205 BANG, DR. F. B., Department of Pathobiology, The Johns Hopkins University, School of Hygiene, Baltimore, Maryland 21205 BARKER, DR., JEFFERY L., M.D., Bldg. 36, Room 2002, National Institutes of Health, Bethesda, Maryland 20014 BARLOW, DR. ROBERT B., JR., Institute for Sensory Research, Syracuse Uni- versity, Merrill Lane, Syracuse, New York 13210 BARTELL, DR. CLELMER K., 2000 Lake Shore Drive, New Orleans, Louisiana 70122 BARTH, DR. LUCENA J., 26 Quisset Ave., Woods Hole, Massachusetts 02543 BARTLETT, DR. JAMES H., Department of Physics, University of Alabama, P. O. Box 1921, University, Alabama 35486 BAUER, DR. G. ERIC, Department of Anatomy, University of Minnesota, Minneapolis, Minnesota 55414 BEAUGE, DR. Luis ALBERTO, Department of Biophysics, University of Maryland School of Medicine, 660 W. Redwood St., Baltimore, Maryland 21201 BECK, DR. L. V., Department of Pharmacology, Indiana University, School of Experimental Medicine, Bloomington, Indiana 47401 BELL, DR. EUGENE, Department of Biology, Massachusetts Institute of Tech- nology, 77 Massachusetts Ave., Cambridge, Massachusetts 02139 BENNETT, DR. MICHAEL V. L., Department of Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Ave., New York, New York 10461 BENNETT, DR. MIRIAM F., Department of Biology, Colby College, Waterville, Maine 04901 BERG, DR. CARL J., JR., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 BERMAN, DR. MONES, National Institutes of Health, Theoretical Biology NCI, Bldg. 10, 4B56, Bethesda, Maryland 20014 BERNARD, GARY D., Department of Opthalmology and Visual Science, Yale University, 333 Cedar St., New Haven, Connecticut 06510 BERNE, DR. ROBERT W., University of Virginia School of Medicine, Charlottes- ville, Virginia 22903 BERNHEIMER, DR. ALAN W., New York University College of Medicine, New York, New York 10016 BIGGERS, DR. JOHN DENNIS, Department of Physiology, Harvard Medical School, 25 Shattuck St., Boston, Massachusetts 02115 BISHOP, DR. STEPHEN H., Department of Zoology, Iowa State University, Ames, Iowa 50010 10 MARINE BIOLOGICAL LABORATORY BLAUSTEIX, MORDECAI P., Department of Physiology and Biophysics, Washing- ton University School of Medicine, St. Louis, Missouri 03110 BLUM, DR. HAROLD F., 404 Locust Lane S., Westchester, Pennsylvania 19380 BODIAN, DR. DAVID, Department of Otolaryngology, The Johns Hopkins Uni- versity, Traylor Building, Room 424, 1721 Madison St., Baltimore, Maryland 21205 BOETTIGER, DR. EDWARD G., 5 Dunham Pond Rd., Mansfield, Connecticut 06250 BOOLOOTIAN, DR. RICHARD A., President, Science Software System, Inc., 11899 West Pico Blvd., W. Los Angeles, California 90064 BOREI, DR. HANS G., Department of Zoology, University of Pennsylvania, Philadelphia, Pennsylvania 19174 BORGESE, DR. THOMAS A., Department of Biology, Lehman College, City Uni- versity of New York, Bronx, New York 10468 BORISY, DR. GARY G., Laboratory of Molecular Biology, University of Wis- consin, Madison, Wisconsin 53715 BOSCH, DR. HERMAN F., Whipple Hill, Richmond, New Hampshire 03470 BOTKIN, DR. DANIEL B., Department of Biology, University of California, Santa Barbara, California 93106 BOWEX, DR. VAUGHN T., Woods Hole Oceanographic Institution, Redfield Bldg. 3-32, Woods Hole, Massachusetts 02543 BOWLES, DR. FRANCIS P., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 BRANDT, DR. PHILIP WILLIAMS, Department of Anatomy, Columbia University, College of Physicians and Surgeons, New York, New York 10032 BRINLEY, DR. F. J., JR., Neurological Disorders Program, NINCDS, 716 Federal Building, Bethesda, Maryland 20205 BRODWICK, DR. MALCOLM S., Department of Physiology and Biophysics, Uni- versity of Texas Medical Branch, Galveston, Texas 77550 BROOKS, DR. MATILDA M., Department of Physiology, University of California, Berkeley, California 94720 BROWN, DR. FRANK A., JR., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 BROWN, DR. JAY C., Department of Neurobiology, University of Virginia, Charlottesville, Virginia 22908 BROWN, DR. JOEL E., Department of Physiology and Biophysics, Health Sciences Center, State University of New York, Stony Brook, Long Island, New York 11794 E^ROWN, DR. STEPHEN C., Department of Biological Sciences, State University of New York, Albany, New York 12222 BUCK, DR. JOHN B., Laboratory of Physical Biology, National Institutes of Health, Bethesda, Maryland 20014 BURBANCK, DR. MADELINE PALMER, Box 15134, Atlanta, Georgia 30333 BURBANCK, DR. WILLIAM D., Box 15134, Atlanta, Georgia 30333 BURDICK, DR. CAROLYN J., Department of Biology, Brooklyn College, Brooklyn, New York 11210 BURGER, DR. MAX M., Department of Biochemistry, University of Basel, CH. 4056-Klingelbergstrasse 70, Basel, Switzerland BURKY, DR. ALBERT J., Department of Biology, University of Dayton, Dayton, Ohio 45469 BURR, DR. ARTHUR H., Department of Biological Sciences, Simon Fraser Uni- versity, Burnaby, British Columbia, Canada V5A 1S6 MEMBERS OF THE CORPORATION 11 CANDELAS, DR. GRACIELA C., 30 Christopher St., Apt. 5-A, New York, New York 10014 CARLSON, DR. FRANCIS D., Department of Biophysics, The Johns Hopkins University, Baltimore, Maryland 21218 CASE, DR. JAMES F., Department of Biological Sciences, University of California, Santa Barbara, California 93106 CASSIDY, REV. JOSEPH, O.P., Department of Biological Sciences, Northwestern University, Evanston, Illinois 60201 CEBRA, DR. JOHN J., Department of Biology, The Johns Hopkins University, Baltimore, Maryland 21218 CHAET, DR. ALFRED B., University of West Florida, Pensacola, Florida 32504 CHAMBERS, DR. EDWARD L., Department of Physiology and Biophysics, Uni- versity of Miami School of Medicine, P. (). Box 52087, Biscayne Annex, Miami, Florida 33152 CHAPPELL, DR. RICHARD L., Department of Biological Sciences, Hunter College, The City University of New York, New York, New York 10021 CHAUNCEY, DR. HOWARD H., 30 Falmouth St., Wellesley Hills, Massachusetts 02181 CHENEY, DR. RALPH H., 45 Coleridge Dr., Falmouth, Massachusetts 02540. CHILD, DR. FRANK M., Department of Biology, Trinity College, Hartford, Connecticut 06106 CITKOWITZ, DR. ELENA, 410 Livingston St., New Haven, Connecticut 06511 CLARK, DR. A. M., Department of Biological Sciences, University of Delaware, Newark, Delaware 19711 CLARK, DR. ELOISE E., National Science Foundation, 1800 G Street, Washington, D. C. 20550 CLARK, HAYS, Executive Vice-President, Avon Products, Inc., 9 West 57th Street, New York, New York 10019 CLARK, DR. WALLIS H., Aquaculture Program, Rm. 243, Department of Animal Science, University of California, Davis, California 95616 CLAUDE, DR. PHILIPPA, Primate Center, Capital Court, Madison, Wisconsin 53706 CLAYTON, DR. RODERICK K., Section of Genetics, Development and Physiology, Cornell University, Ithaca, New York 14850 CLOWES, DR. GEORGE H. A., JR., The Cancer Research Institute, 194 Pilgrim Rd., Boston, Massachusetts 02215 CLUTTER, DR. MARY, Developmental Biology Program, National Science Foundation, Washington, D. C. 20550 COBB, DR. JEWEL P., Dean, Douglass College, New Brunswick, New Jersey 08903 COHEN, DR. ADOLPH I., Department of Ophthalmology, Washington University, School of Medicine, 660 S. Euclid Ave., St. Louis, Missouri 63110 COHEN, DR. LAWRENCE B., Department of Physiology, Yale University, 333 Cedar St., New Haven, Connecticut 06510 COHEN, DR. SEYMOUR S., Department of Pharmacological Science, State Uni- versity of New York at Stony Brook, Stony Brook, New York 11790 COHEN, DR. WILLIAM D., Department of Biological Sciences, Hunter College, 695 Park Ave., New York, New York 10021 COLLIER, DR. JACK R., Department of Biology, Brooklyn College, Brooklyn, New York 11210 COOPERSTEIN, DR. SHERWIN J., University of Connecticut, School of Medicine, Farmington Ave., Farmington, Connecticut 06032 12 MARINE BIOLOGICAL LABORATORY CORLISS, DR. JOHN ()., Department of Zoology, University of Maryland, College Park, Maryland 20742 CORNELL, DR. NEAL \V., 6428 Bannockburn Drive, Bethesda, Maryland 20034 CORNMAN, DR. IVOR, 10A Orchard Street, Woods Hole, Massachusetts 02543 COUCH, DR. ERNEST F., Department of Biology, Texas Christian University, Fort Worth, Texas 76129 CRANE, JOHN O., 315 West 106th Street, New York, New York 10025 CREMER-BARTELS, DR. GERTRUD, Universitats Augenklinik, 44 Minister, West Germany CRIPPA, DR. MARCO, Department de Biologic Animale, Embryologie Moleculaire, 154 route de Malagnou, CH-1224, Chene-Bougeries, Geneve, Switzerland CROW, DR. TERRY J., Laboratory of Biophysics, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 CROWELL, DR. SEARS, Department of Zoology, Indiana University, Bloomington, Indiana 47401 DAIGNAULT, ALEXANDER T., W. R. Grace Co., 1114 Avenue of the Americas, New York, New York 10036 DAN, DR. KATSUMA, Professor Emeritus, Tokyo Metropolitan University, Meguro-ku, Tokyo, Japan DANIELLI, DR. JAMES F., Life Sciences Department, \Vorcester Polytechnic Institute, Worcester, Massachusetts 01609 DAVIS, DR. BERNARD D., M.D., Bacterial Physiology Unit, Harvard Medical School, 25 Shattuck Street, Boston, Massachusetts 02115 DAW, DR. NIGEL W., 78 Aberdeen PL, Clayton, Missouri 63105 DEGROOF, DR. ROBERT C., Department of Pharmacology, Thomas Jefferson University, 1020 Locust St., Philadelphia, Pennsylvania 19174 DEHAAN, DR. ROBERT L., Department of Anatomy, Emory University, Atlanta, GEORGIA 30322 DELANNEY, DR. Louis E., The Jackson Laboratory, Bar Harbor, Maine 04609 DEPHILLIPS, DR. HENRY A., JR., Department of Chemistry, Trinity College, Hartford, Connecticut 06106 DETTBARN, DR. WOLF-DIETRICH, Department of Pharmacology, Vanderbilt University, School of Medicine, Nashville, Tennessee 37217 DE\VEER, DR. PAUL J., Department of Physiology, Washington University, School of Medicine, St. Louis, Missouri 63110 DISCHE, DR. ZACHARIAS, Columbia University, College of Physicians and Sur- geons, 630 W. 165th Street, New York, New York 10032 DIXON, DR. KEITH E., School of Biological Sciences, Flinders University, Bedford Park, South Australia DOWDALL, DR. MICHAEL J., Department of Biochemistry, University Hospital and Medical School, Clifton Boulevard, Nottingham NG7 2UH, England, U. K. DOWLING, DR. JOHN E., Biological Laboratories, Harvard University, 16 Divinity Avenue, Cambridge, Massachusetts 02138 DRESDEN, DR. MARC H., Department of Biochemistry, Baylor College of Medi- cine, Houston, Texas 77025 DUDLEY, DR. PATRICIA L., Department of Biological Sciences, Barnard College, Columbia University, New York, New York 10027 DUNHAM, DR. PHILIP B., Department of Biology, Syracuse University, Syracuse, New York 13210 MEMBERS OF THE CORPORATION 13 EBERT, DR. JAMKS DAVID, Office of the President, Carnegie Institute of Wash- ington, 1530 P Street N.W., Washington, 1). C. 20008 ECKBERG, DR. WILLIAM R., Department of Zoology, Howard University, Wash- ington, D. C. 2005 9 ECKERT DR. ROGER ()., Department of Zoology, University of California, Los Angeles, California 00024 EDDS, DR. KENNETH T., Department ot Anatomical Sciences, State University of New York, Buffalo, New York 14214 EDDS, DR. LOUISE, College of Osteopathic Medicine, Grosvenor Hall, Ohio University, Athens, Ohio 45701 EDER, DR. HOWARD A., Albert Einstein College of Medicine, 1300 Morris Park Ave., New York, New York 10461 EDWARDS, DR. CHARLES, Department of Biological Sciences, State University of New York at Albany, Albany, New York 12222 EGYUD, DR. LASZLO G., The Sonntag Institute for Cancer Research, Boston College, Chestnut Hill, Massachusetts 02167 EHRENSTEIN, DR. GERALD, National Institutes of Health, Bethesda, Maryland 20014 EICHEL, DR. HERBERT J., Department of Biological Chemistry, Hahnemann Medical College, Philadelphia, Pennsylvania 19174 EISEN, DR. ARTHUR Z., Division of Dermatology, Washington University, School of Medicine, St. Louis, Missouri 63110 ELDER, DR. HUGH Y., Institute of Physiology, University of Glasgow, Glasgow, Scotland, U. K. ELLIOTT, DR. GERALD F., The Open University Research Unit, Foxcombe Hall, Berkeley Road, Boar Hill, Oxford, England, U. K. EPEL, DR. DAVID, Hopkins Marine Station, Pacific Grove, California 93950 EPSTEIN, DR. HERMAN T., Department of Biology, Brandeis University, Wal- tham, Massachusetts 02154 ERULKAR, DR. SOLOMON D., 318 Kent Rd., Bala Cynwyd, Pennsylvania 19004 ESSNER, DR. EDWARD S., Kresge Eye Institute, Wayne State University, School of Medicine, 540 E. Canfield Ave., Detroit, Michigan 48201 ETIENNE, DR. EARL M., Department of Anatomy, Harvard Medical School, Boston, Massachusetts 02115 FAILLA, DR. PATRICIA M., Office of the Director, Argonne National Laboratory, Argonne, Illinois 60439 FARMANFARMAIAN, DR. ALLAHVERDI, Department of Physiology and Bio- chemistry, Rutgers University, New Brunswick, New Jersey 08903 FAUST, DR. ROBERT GILBERT, Department of Physiology, University of North Carolina, Medical School, Chapel Hill, North Carolina 27514 FEIN, DR. ALAN, Laboratory of Sensory Physiology, Marine Biological Labora- tory, Woods Hole, Massachusetts 02543 FERGUSON, DR. F. P., National Institute of General Medical Science, National Institutes of Health, Bethesda, Maryland 20014 FESSENDEN, JANE, Librarian, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 FINKELSTEIN, DR. ALAN, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, New York 10461 FISCHBACH, DR. GERALD, M.D., Department of Pharmacology, Harvard Medical School, 25 Shattuck St., Boston, Massachusetts 02115 FISCHMAN, DR. DONALD A., Department of Anatomy and Cell Biology, State 14 MARINE BIOLOGICAL LABORATORY University of New York, Downstate Medical Center, 450 Clarkson Ave., Brooklyn, New York 11203 FISHER, DR. J. M., Department of Biochemistry, University of Toronto, Toronto, Ontario, Canada M5S 1A8 FISHMAN, DR. HARVEY M., Department of Physiology, University of Texas, Medical Branch, Galveston, Texas 77550 FISHMAN, DR. Louis, 143 North Grove Street, Valley Stream, New York 11580 FLANAGAN, DENNIS, Editor, Scientific American, 415 Madison Ave., New York, New York, 10017 Fox, DR. MAURICE S., Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 FRAENKEL, DR. GOTTFRIED S., Department of Entomology, University of Illinois, Urbana, Illinois 61801 FRANZINI, DR. CLARA, Department of Biology G-5, University of Pennsylvania, School of Medicine, Philadelphia, Pennsylvania 19174 FRAZIER, DR. DONALD T., Department of Physiology and Biophysics, University of Kentucky, School of Medicine, Lexington, Kentucky 40507 FREEMAN, DR. ALLAN R., Professor and Chairman, Department of Physiology, Temple University School of Medicine, 3420 N. Broad St., Philadelphia, Pennsylvania 19140 FREEMAN, DR. GARY L., Department of Zoology, University of Texas, Austin, Texas 78710 FRENCH, DR. ROBERT J., Dept. of Biophysics, University of Maryland, School of Medicine, Baltimore, Maryland 21201 FREYGANG, DR. WALTER J., JR., 6247 29th Street, N. W., Washington, D. C. 20015 FULTON, DR. CHANDLER M., Department of Biology, Brandeis University, Waltham, Massachusetts 02154 FURSHPAN, DR. EDWIN J., Department of Neurophysiology, Harvard Medical School, Boston, Massachusetts 02115 FUSELER, DR. JOHN W., Department of Cell Biology, University of Texas, Medical Branch, 5323 Harry Hines Blvd., Dallas, Texas 75235 FYE, DR. PAUL M., Woods Hole Oceanographic Institution, Woods Hole, Massa- chusetts 02543 GABRIEL, DR. MORDECAI L., Department of Biology, Brooklyn College, Brooklyn, New York 11210 GAINER, DR. HAROLD, Head, Section of Functional Neurochemistry, National Institutes of Health, Bldg. 36, Rm. 2A21, Bethesda, Maryland 20014 GALL, DR. JOSEPH G., Department of Biology, Yale University, New Haven, Connecticut 06520 GELFANT, DR. SEYMOUR, Department of Dermatology, Medical College of Georgia, Augusta, Georgia 30904 GELPERIN, DR. ALAN, Department of Biology, Princeton University, Princeton, New Jersey 08540 GERMAN, DR. JAMES L., Ill, M.D., The New York Blood Center, 310 East 67th Street, New York, New York 10021 GIBBS, DR. MARTIN, Institute for Photobiology of Cells and Organelles, Brandeis University, Waltham, Massachusetts 02154 GIBSON, DR. A. JANE, Wing Hall, Cornell University, Ithaca, New York 14850 GIFFORD, DR. PROSSER, Woodrow Wilson International Center for Scholars, Smithsonian Building, Washington, D. C. 20560 MEMBERS OF THE CORPORATION 15 GILBERT, DR. DANIEL L., Laboratory of Biophysics, NINCDS, National Insti- tutes of Health, Building 36, Room 2A29, Bethesda, Maryland 20014 OILMAN, DR. LAUREN C., Department of Biology, Box 249118, University of Miami, Coral Gables, Florida 33124 GIUDICE, DR. GIOVANNI, University of Palermo, Via Archirafi 22, Palermo, Italy GLUSMAN, DR. MURRAY, Department of Clinical Psychiatry, Columbia Uni- versity, 722 W. 168th St., New York, New York 10032 GOLDEN, WILLIAM T., 40 Wall Street, New York, New York 10005 GOLDMAN, DAVID E., 63 Loop Rd., Falmouth, Massachusetts 02540 GOLDMAN, DR. ROBERT D., Department of Biological Sciences, Carnegie-Mellon University, 4400 Fifth Ave., Pittsburgh, Pennsylvania 15213 GOLDSMITH, DR. MARY H. M., Department of Biology, Kline Biology Tower, Yale University, New Haven, Connecticut 06520 GOLDSMITH, DR. TIMOTHY H., Department of Biology, Yale University, New Haven, Connecticut 06520 GOLDSTEIN, DR. MOISE H., JR., 506 Traylor Bldg., The Johns Hopkins Uni- versity, School of Medicine, 720 Rutland Ave., Baltimore, Maryland 21205 GOOCH, DR. JAMES L., Department of Biology, Juniata College, Huntington, Pennsylvania 16652 GOODMAN, DR. LESLEY JEAN, Department of Zoology and Comparative Physi- ology, Queen Mary College, Mile End Rd., London, El 4 NS, England, U. K. GOTTSCHALL, DR. GERTRUDE Y., 315 East 68th Street, Apartment 9M, New- York, New York 10021 GOUDSMIT, DR. ESTHER M., Department of Biology, Oakland University, Rochester, Michigan 48063 GOULD, DR. STEPHEN J., Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138 GRAHAM, DR. HERBERT, 36 Wilson Road, Woods Hole, Massachusetts 02543 GRANT, DR. PHILLIP, Department of Biology, University of Oregon, Eugene, Oregon 97403 GRASS, ALBERT, The Grass Foundation, 77 Reservoir Road, Quincy, Massa- chusetts 02170 GRASS, ELLEN R., The Grass Foundation, 77 Reservoir Road, Quincy, Massa- chusetts 02170 GRASSLE, DR. JUDITH P., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 GREEN, DR. JAMES W., Department of Physiology, Rutgers University, New Brunswick, New Jersey 08903 GREEN, DR. JONATHAN P., Chairman, Department of Biology, Roosevelt Uni- versity, 430 S. Michigan Ave., Chicago, Illinois 60605 GREENBERG, DR. MICHAEL J., Department of Biological Sciences, Florida State University, Tallahassee, Florida 32306 GREGG, DR. JAMES H., Department of Zoology, University of Florida, Gainesville, Florida 32601 GREIF, DR. ROGER L., Department of Physiology, Cornell University Medical College, New York, New York 10021 GRIFFIN, DR. DONALD R., The Rockefeller University, 1230 York Avenue, New York, New York 10021 GROSCH, DR. DANIEL S., Department of Genetics, Gardner Hall, North Carolina State University, Raleigh, North Carolina 27607 16 MARINE BIOLOGICAL LABORATORY GROSS, DR. PAUL R., President and Director, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 GROSSMAN, DR. ALBERT, New York University Medical School, New York, New York 10016 GUNNING, A. ROBERT, 377 Hatchville Road, Hatchville, Massachusetts 02536 ('.WILLIAM, DR. G. P., Department of Biology, Reed College, Portland, Oregon •J7202 HALL, DR. ZACK W., Department of Physiology, University of California, San Francisco, California 94143 HALVORSON, DR. HARLYN O., Rosenstiel Basic Medical Sciences Research Center, Brandeis University, Waltham, Massachusetts 02154 HAMKALO, DR. BARBARA A., Department of Molecular Biology and Biochemistry, University of California, Irvine, California 92717 HANDLER, DR. PHILIP, President, National Academy of Sciences, 2101 Con- stitution Ave. N. W., Washington, D. C. 20418 HANNA, DR. ROBERT B., State University of New York, College of Environ- mental Science and Forestry, Syracuse, New York 13210 HARDING, DR. CLIFFORD V., JR., Professor and Director of Research, Kresge Eye Institute, Wayne State University, School of Medicine, 540 E. Canfield, Detroit, Michigan 48201 HAROSI, DR. FERENC I., Laboratory of Sensory Physiology, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 HARRIGAN, DR. JUNE F., Laboratory of Biophysics, Marine Biological Labo- ratory, Woods Hole, Massachusetts 02543 HARRINGTON, DR. GLENN W., Department of Microbiology, University of Missouri, School of Dentistry, 650 E. 25th Street, Kansas City, Missouri 64108 HASCHEMEYER, DR. AUDREY E. V., Department of Biological Sciences, Hunter College, 965 Park Avenue, New York, New York 10021 HASTINGS, DR. J. WOODLAND, The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138 HAXO, DR. FRANCIS T., Division of Marine Biology, Scripps Institution of Oceanography. University of California, La Jolla, California 92038 HAYASHI, TERU, Institute for Cytology, University of Bonn, Ulrich-Haberland Str. 61A, 5300 Bonn, West Germany HAYES, DR. RAYMOND L., JR., Department of Anatomy, School of Medicine, Morehouse College, 223 Chestnut St. S. W., Atlanta, Georgia 30314 HENLEY, DR. CATHERINE, 5225 Pooks Hill Road, Apt. 1120 North, Bethesda, Maryland 20014 HERNDON, DR. WALTER R., 506 Andy Holt Tower, University of Tennessee, Knoxville. Tennessee 37916 HERVEY, JOHN P., Box 85, Penzance Road, Woods Hole, Massachusetts 02543 Hi SSLER, DR. ANITA Y., 5795 Waverly Avenue, La Jolla, California 92037 HEUSER, DR. JOHN, M.D., Department of Physiology, School of Medicine, University of California, San Francisco, California 94143 HIATT, DR. HOWARD H., Office of the Dean, Harvard School of Public Health, 677 Huntington Ave., Boston, Massachusetts 02115 HIGHSTEIN, DR. STEPHEN M., Division of Cellular Neurobiology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 14061 HILDEBRAND, DR. JOHN G., Department of Neurobiology, Harvard Medical School, Boston, Massachusetts 02115 MEMBERS OF THE CORPORATION 17 HILL, DR. ROBERT H., Department of Zoology, University of Rhode Island, Kingston, Rhode Island 02881 HILLMAN, DR. PETER, Department of Biology, Hebrew University, Jerusalem, Israel HINEGARDNER, DR. RALPH T., Division of Natural Sciences, University of California, Santa Cruz, California 95060 HINSCH, DR. GERTRUDE W., Department of Biology, University of South Florida, Tampa, Florida 33620 HOBBIE, DR. JOHN E., The Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 HODGE, DR. ALAN J., Marine Biological Laboratory, Woods Hole, Massachusetts 02543 HODGE, DR. CHARLES, IV, P.O. Box 4095, Philadelphia, Pennsylvania 19118 HOFFMAN, DR. JOSEPH, Department of Physiology, Yale University School of Medicine, New Haven, Connecticut 06515 HOFFMANN, DR. RICHARD J., Department of Zoology, Iowa State University, Ames, Iowa 50011 HOLLYFIELD, DR. JOE G., Baylor School of Medicine, Texas Medical Center, Houston, Texas 77030 HOLTZMAN, DR. ERIC, Department of Biological Sciences, Columbia University, New York, New York 10027 HOLZ, DR. GEORGE G., JR., Department of Microbiology, State University of New York, Upstate Medical Center, Syracuse, New York 13210 HOSKIN, DR. FRANCIS C. G., Department of Biology, Illinois Institute of Tech- nology, Chicago, Illinois 60616 HOUGHTON, DR. RICHARD A., Ill, Ecosystems Center, Marine Biological Labo- ratory, Woods Hole, Massachusetts 02543 HOUSTON, HOWARD, Preston Ave., Meridan, Connecticut 06450 HOY, DR. RONALD R., Section of Neurobiology and Behavior, Cornell Uni- versity, Ithaca, New York 14850 HUBBARD, DR. RUTH, The Biological Laboratories, Harvard University, Cam- bridge, Massachusetts 02138 HUMES, DR. ARTHUR G., Boston University Marine Program, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 HUMMON, DR. WILLIAM D., Department of Biology, Ohio University, Athens, Ohio 45701 HUMPHREYS, DR. SUSIE HUNT, GRC-NIH, Baltimore City Hospital, Baltimore, Maryland 21224 HUMPHREYS, DR. TOM D., University of Hawaii, Pacific Biomedical Research Center, 41 Ahui St., Honolulu, Hawaii 96813 HUNTER, DR. BRUCE, Provost Office, Tulane University, New Orleans, Louisiana 70118 HUNTER, DR. R. D., Department of Biological Sciences, Oakland University, Rochester, Michigan 48063 HUNZIKER, H. E., Esq., P.O. Box 547, Falmouth, Massachusetts 02541 HURWITZ, DR. CHARLES, Basic Science Research Laboratory, VA Hospital, Albany, New York 12208 HURWITZ, DR. JERARD, Department of Molecular Biology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461 HUXLEY, DR. HUGH E., Medical Research Council, Laboratory of Molecular Biology, Cambridge, England, U. K. 18 MARINE BIOLOGICAL LABORATORY HYDE, DR. BEAL B., Department of Botany, University of Vermont, Burlington, Vermont 05401 ILAN, DR. JOSEPH, Department of Anatomy, Case Western Reserve University, School of Medicine, Cleveland, Ohio 44106 INOUE, DR. SADUYKI, Electron Microscopy Laboratory, McGill University Cancer Center, 3655 Drummond St., Montreal, P. A. Canada HG3 1Y6 INOUE, DR. SHINYA, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 ISENBERG, DR. IRVING, Department of Biochemistry and Biophysics, Oregon State University, Corvallis, Oregon 97331 ISSELBACHER, DR. KURT J., Massachusetts General Hospital, Boston, Massa- chusetts 02714 ISSIDORIDES, DR. MARIETTA R., Department of Psychiatry, University of Athens, Monis Petraki 8, Athens 140, Greece IZZARD, DR. COLIN S., Department of Biological Sciences, State University of New York at Albany, Albany, New York 12222 JACOBSON, DR. ANTONE G., Department of Zoology, University of Texas, Austin, Texas 78712 JAFFEE, DR. LIONEL, Department of Biology, Purdue University, Lafayette, Indiana 47907 JAHAN-PARWAR, BEHRUS, Worcester Foundation for Experimental Biology, 222 Maple Ave., Shrewsbury, Massachusetts 01545 JANNASCH, DR. HOLGER W., Woods Hole Oceanographic Institution, \Voods Hole, Massachusetts 02543 JEFFERY, DR. WILLIAM R., Department of Zoology, University of Texas, Austin, Texas 78712 JENNER, DR. CHARLES E., Department of Zoology, University of North Carolina, Chapel Hill, North Carolina 27514 JENNINGS, DR. JOSEPH B., Department of Zoology, Baines Wing, University of Leeds, Leeds LS2 9JT, England, U. K. JONES, DR. MEREDITH L., Division of Worms, Museum of Natural History, Smithsonian Institution, Washington, D. C. 20650 JONES, DR. RAYMOND F., Department of Biology, State University of New York at Stony Brook, Stony Brook, New York 11753 JOSEPHSON, DR. R. K., School of Biological Sciences, University of California, Irvine, California 92717 JOYNER, DR. RONALD W., Department of Physiology, University of Iowa, Iowa City, Iowa 52242 KABAT, DR. E. A., Department of Microbiology, Columbia University, College of Physicians and Surgeons, 630 W. 168th St., New York, New York 10032 KAFATOS, DR. FOTIS C., The Biological Laboratories, Harvard University, 16 Divinity Avenue, Cambridge, Massachusetts 02138 KALEY, DR. GABOR, Department of Physiology, Basic Sciences Building, New York Medical College, Valhalla, New York 10595 KALTENBACH, DR. JANE, Department of Biological Sciences, Mount Holyoke College, South Hadley, Massachusetts 01075 KAMINER, DR. BENJAMIN, Department of Physiology, Boston University School of Medicine, 80 E. Concord St., Boston, Massachusetts 02118 KAMMER, DR. ANN E., Division of Biology, Kansas State University, Manhattan, Kansas 66502 MEMBERS OF THE CORPORATION 19 KANE, DR. ROBERT E., Pacific Biomedical Research Center, 41 Ahui Street, University of Hawaii, Honolulu, Hawaii 96813 KANESHIRO, DR. EDNA S., Department of Biological Sciences, University of Cincinnati, Cincinnati, Ohio 45221 KAPLAN, DR. EHUD, The Rockefeller University, 1230 York Ave., New York, New York 10021 KARAKASHIAN, DR. STEPHEN J., 165 West 91st Street, Apt. 16-F, New York, New York 10024 KARUSH, DR. FRED, Department of Microbiology, University of Pennsylvania, School of Medicine, Philadelphia, Pennsylvania 19174 KATZ, DR. GEORGE M., Department of Neurology, Columbia University, College of Physicians and Surgeons, 630 West 168th Street, New York, New York 10032 KEAN, DR. EDWARD L., Departments of Biochemistry and Ophthalmology, Case Western Reserve University, Cleveland, Ohio 44101 KEMP, DR. NORMAN E., Department of Zoology, University of Michigan, Ann Arbor, Michigan 48104 KENDALL, John P., One Boston Place, Boston, Massachusetts 02108 KEOSIAN, DR. JOHN, P.O. Box 193, Woods Hole, Massachusetts 02543 KETCHUM DR. BOSTWICK H., P.O. Box 32, W^oods Hole, Massachusetts 02543 KEYNAN, DR. ALEXANDER, Vice President, Hebrew University, Jerusalem, Israel KING, DR. THOMAS J., Program Director, Division of Cancer Research, Resources and Center, National Institutes of Health, Bldg. 31, Room 10A03, Bethesda, Maryland 20014 KINGSBURY, DR. JOHN M., Department of Botany, Cornell University, Ithaca, New York 14850 KIRSCHENBAUM, DR. DONALD, Department of Biochemistry, College of Medi- cine, State University of New York, 450 Clarkson Avenue, Brooklyn, New York 11203 KLEIN, DR. MORTON, Department of Microbiology, Temple University, Phila- delphia, Pennsylvania 19122 KLOTZ, DR. I. M., Department of Chemistry, Northwestern University, Evans- ton, Illinois 60201 KOIDE, DR. SAMUEL S., Population Council, The Rockefeller University, 66th St. and York Ave., New York, New York 10021 KONINGSBERG, DR. IRWIN R., Department of Biology, Gilmer Hall, University of Virginia, Charlottesville, Virginia 22903 KOSOWER, DR. EDWARD M., Department of Chemistry, University of California, La Jolla, California 92093 KRAHL, DR. M. E., 2783 W. Casas Circle, Tucson, Arizona 85704 KRANE, DR. STEPHEN M., Massachusetts General Hospital, Boston, Massa- chusetts 02114 KRASSNER, DR. STUART MITCHELL, Department of Developmental and Cell Biology, University of California, Irvine, California 92717 KRAUSS, DR. ROBERT, FASEB, 9650 Rockville Pike, Bethesda, Maryland 20014 KRAVITZ, DR. EDWARD A., Department of Neurobiology, Harvard Medical School, 25 Shattuck St., Boston, Massachusetts 02115 KRIEBEL, DR. MAHLON E., Department of Physiology, State University of New York, Upstate Medical Center, 766 Irving Ave., Syracuse, New York 13210 KRIEG, DR. WENDELL J. S., 1236 Hinman, Evanston, Illinois 60602 20 MARINE BIOLOGICAL LABORATORY KRUPA, DR. PAUL L., Department of Biology, The City College of New York, 139th Street and Convent Avenue, New York, New York 10031 KUFFLER, DR. STEPHEN \Y., Department of Neurophysiology, Harvard Medical School, Boston, Massachusetts 02115 KUSAXO, DR. KIYOSHI, Department of Biology, Illinois Institute of Technology, 3300 South Federal Street, Chicago, Illinois 60616 LAMARCHE, DR. PAUL H., M.D., 593 Eddy St., Providence, Rhode Island 02903 LAXCEFIELD, DR. REBECCA C., The Rockefeller University, 1230 York Ave., New York, New York 10021 LAXDIS, DENNIS, M.D., Department of Neurology, Massachusetts General Hospital, Boston, Massachusetts 02114 LANDOWXE, DR. DAVID, Department of Physiology, University of Miami, Miami, Florida 33124 LAXGFORD, DR. GEORGE M., Department of Physiology, University of North Carolina, Medical Sciences Research Wing 206 H, Chapel Hill, North Carolina 27514 LASH, DR. JAMES W., Department of Anatomy, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19174 LASTER, DR. LEONARD, M.D., President, University of Oregon, Health Sciences Center, Portland, Oregon 97201 LAUFER, DR. HANS, Biological Sciences Group U-42, University of Connecticut, Storrs, Connecticut 06268 LAUFFER, DR. MAX A., Department of Biophysics, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 LAWRENCE, E. SWIFT, Pawtucket Institution for Savings, 286 Main St., Paw- tucket, Rhode Island 02860 LAZAROW, DR. JANE, 221 Woodlawn Ave., St. Paul, Minnesota 55105 LEADBETTER, DR. EDWARD R., Biological Sciences Group U-42, University of Connecticut, Storrs, Connecticut 06268 LEAK, DR. LEE VIRN, Department of Anatomy, Howard University, College of Medicine, Washington, D. C. 20059 LECAR, DR. HAROLD, Laboratory of Biophysics, National Institute of Neuro- logical Diseases and Stroke, National Institutes of Health, Bethesda, Maryland 20014 LEDERBERG, DR. JOSHUA, President, The Rockefeller University, New York, New York 10021 LEE, DR. JOHN J., Department of Biology, City College of the City University of New York, Convent Avenue and 138th Street, New York, New York 10031 LEFEVRE, DR. PAUL G., Department of Physiology, Health Sciences Center, East Campus, State University of New York at Stony Brook, Stony Brook, Long Island, New York 11794 LEIGHTON, DR. JOSEPH, M.D., Department of Pathology, Medical College of Pennsylvania, 3300 Henry Ave., Philadelphia, Pennsylvania 19129 LENHER, DR. SAMUEL, 50-C Cokesbury Village, Hockessin, Delaware 19707 LERMAN, DR. SIDXKY, Laboratory for Ophthalmic Research, Emory University, Atlanta, Georgia 30322 LERNER, DR. AARON B., Yale Medical School, New Haven, Connecticut 06510 LEVIN, DR. JACK, M.D., Hematology Division, The Johns Hopkins Hospital, Baltimore, Maryland 21205 LEVINE, DR. RACHMIEL, M.D., 2024 Canyon Road, Arcadia, California 91006 MEMBERS OF THE CORPORATION 21 LEVINTHAL, DR. CYRUS, Department of Biological Sciences, 90S Schermerhorn Hall, Columbia University, New York, New York 10027 LEVITAN, DR. HERBERT, Department of Zoology, University of Maryland, College Park, Maryland 20742 LEWIN, DR. RALPH A., Scripps Institution of Oceanography, La Jolla, California 92093 LING, DR. GILBERT, 307 Berkeley Road, Merion, Pennsylvania 19066 LINSKENS, DR. H. P., Department of Botany, University of Diiehuizerweg 2()(), Nijmegen, The Netherlands LIPICKY, DR. RAYMOND J., M.D., 886 College Parkway No. 304, Rockville, Maryland 20850 LITTLE, DR. E. P., 216 Highland Street, West Newton, Massachusetts 02158 LIUZZI, DR. ANTHONY, Department of Physics, University of Lowell, Lowell, Massachusetts 01854 LLINAS, DR. RODOLFO R., Department of Physiology and Biophysics, New York University Medical Center, 550 First Ave., New York, New York 10016 LOEWENSTEIN, DR. WERNER R., Department of Physiology and Biophysics, University of Miami, School of Medicine, P. O. Box 016430, Miami, Florida 33101 LOEWUS, DR. FRANK A., Department of Agricultural Chemistry, Washington State University, Pullman, Washington 99164 LOFTFIELD, DR. ROBERT B., Department of Biochemistry, School of Medicine, University of New Mexico, 900 Stanford N. E., Albuquerque, New Mexico 87106 LONDON, DR. IRVING M., 16-512, Massachusetts Institute of Technology, Cam- bridge, Massachusetts 02139 LONGO, DR. FRANK J., Department of Anatomy, University of Iowa, Iowa City, Iowa 52442 LORAND, DR. LASZLO, Department of Biochemistry and Molecular Biology, Northwestern University, Evanston, Illinois 60201 LURIA, DR. SALVADOR E., Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 LYNCH, DR. CLARA J., 4800 Filmore Avenue, Alexandria, Virginia 22311 MACAGNO, DR. EDUARDO R., 1003B Fairchild, Columbia University, New York, New York 10027 MAcNiCHOL, DR. EDWARD F., JR., Laboratory of Sensory Physiology, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 MAHLER, DR. HENRY R., Department of Biochemistry, Indiana University, Bloomington, Indiana 47401 MALKIEL, DR. SAUL, Sidney Farber Cancer Center, 35 Binney Street, Boston, Massachusetts 02115 MANALIS, DR. RICHARD S., Department of Physiology, University of Cincinnati, College of Medicine, Eden and Bethesda Aves., Cincinnati, Ohio 45267 MANGUM, DR. CHARLOTTE P., Department of Biology, College of William and Mary, Williamsburg, Virginia 23185 MARKS, DR. PAUL A., Health Sciences Center, Room 1802, Columbia University College of Physicians and Surgeons, 630 West 168th Street, New York, New York 10032 MARSH, DR. JULIAN B., Department of Biochemistry and Physiology, Medical College of Pennsylvania, 3300 Henry Ave., Philadelphia, Pennsylvania 19129 22 MARINE BIOLOGICAL LABORATORY MARTIN, LOWELL V., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 MARUO, DR. TAKESHI, M.D., Dept. of Obstetrics and Gynecology, Kobe Uni- versity School of Medicine, Ikuta-ku, Kobe 650, Japan MASER, DR. MORTON, Marine Biological Laboiatory, Woods Hole, Massachusetts 02543 MASTROIANNI, DR. LUIGI, JR., M.D., Chairman, Department of Obstetrics and Gynecology, Hospital of the University of Pennsylvania, 3400 Spruce St., Philadelphia, Pennsylvania 19174 MATHEWS, DR. RITA \V., Hunter College, Box 1075, 605 Park Ave., New York, New York 10021 MAUTNER, DR. HENRY G., Department of Biochemistry and Pharmacology, Tufts University School of Medicine, 136 Harrison Avenue, Boston, Massa- chusetts 02111 MAUZERALL, DR. DAVID, The Rockefeller University, 66th Street and York Avenue, New York, New York 10021 MAXWELL, DR. ARTHUR, Provost, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543 MAZIA, DR. DANIEL, Department of Zoology, University of California, Berkeley, California 94720 McCANN, DR. FRANCES, Department of Physiology, Dartmouth Medical School, Hanover, New Hampshire 03755 McCLOSKEY, DR. LAWRENCE R., Department of Biology, Walla Walla College, College Place, Washington 99324 MCLAUGHLIN, JANE A., P. O. Box 187, Woods Hole, Massachusetts 02543 McMAHON, DR. ROBERT F., Department of Biology, University of Texas, Arlington, Texas 76019 McREYNOLDS, DR. JOHN S., Department of Physiology, University of Michigan, Ann Arbor, Michigan 48104 MEINKOTH, DR. NORMAN A., Department of Biology, Swarthmore College, Swarthmore, Pennsylvania 19081 MEISS, DR. DENNIS E., Dept. of Biology, Clark University, Worcester, Massa- chusetts 01610 MELILLO, DR. JERRY M., Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 MELLON, DR. DEFOREST, JR., Department of Biology, University of Virginia, Charlottesville, Virginia 22903 METZ, DR. C. B., Institute of Molecular Evolution, University of Miami, 521 Anastasia St., Coral Gables, Florida 33134 MIDDLEBROOK, DR. ROBERT, 86 Station Road, Burley-In-Warfedale, West Yorks, England, U. K. MILKMAN, DR. ROGER D., Department of Zoology, University of Iowa, Iowa City, Iowa 52242 MILLS, DR. ERIC LEONARD, Institute of Oceanography, Dalhousie University, Halifax, Nova Scotia, Canada MILLS, ROBERT, 56 Worcester Ct., Falmouth, Massachusetts 02540 MITCHELL, DR. RALPH, Pierce Hall, Harvard University, Cambridge, Massa- chusetts 02138 MIZELL, DR. MERLE, Department of Biology, Tulane University, New Orleans, Louisiana 70118 MONROY, DR. ALBERTO, Stazione Zoologica, Villa Communale, Napoli, Italy MEMBERS OF THE CORPORATION 23 MONTROLL, DR. ELIOTT W., Institute for Fundamental Studies, Department of Physics, University of Rochester, Rochester, New York 14627 MOORK, DR. JOHN A., Department of Biology, University of California, Riverside, California 92521 MOORE, DR. JOHN W., Department of Physiology, Duke University Medical Center, Durham, North Carolina 27710 MOORE, DR. LEE E., Department of Physiology and Biophysics, University of Texas Medical Branch, Galveston, Texas 77550 MORAN, DR. JOSEPH F., JR., 23 Foxwood Drive, RR# 1, Eastham, Massachusetts 02642 MORIN, DR. JAMES G., Department of Biology, University of California, Los Angeles, California 90024 MORRELL, DR. FRANK, Department of Neurological Sciences, Rush Medical Center, 1753 W. Congress Pkwy., Chicago, Illinois 60612 MORRILL, DR. JOHN B., JR., Division of Natural Sciences, New College, Sarasota, Florida 33580 MORSE, DR. RICHARD STETSON, 193 Winding River Road, Wellesley, Massa- chusetts 02181 MORSE, ROBERT \V., Associate Director, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543 MOSCONA, DR. A. A., Department of Biology, University of Chicago, Chicago, Illinois 60627 MOTE, DR. MICHAEL I., Department of Biology, Temple University, Philadelphia, Pennsylvania 19122 MOUNTAIN, DR. ISABEL M., Vinson Hall #112, 6251 Old Dominion Drive, McLean, Virginia 22101 MULLEN, GEORGE, Belknap and McLain, 650 Pleasant St., Watertown, Massa- chusetts 02172 MUSACCHIA, DR. XAVIER J., Graduate School, University of Louisville, Louis- ville, Kentucky 40208 NABRIT, DR. S. M., 686 Beckwith Street S. W., Atlanta, Georgia 30314 NACE, DR. PAUL FOLEY, 5 Bowditch Rd., Woods Hole, Massachusetts 02543 NAKAJIMA, DR. SHIGEHIRO, Department of Biological Sciences, Purdue Uni- versity, W. Lafayette, Indiana 47907 NAKAJIMA, DR. YASUKO, Department of Biological Sciences, Purdue University, W. Lafayette, Indiana 47907 NARAHASHI, DR. TOSHIO, Department of Pharmacology, Medical Center, North- western University, 303 E. Chicago Ave., Chicago, Illinois 60611 NASATIR, DR. MAIMON, Department of Biology, University of Toledo, Toledo, Ohio 43606 NELSON, DR. LEONARD, Department of Physiology, Medical College of Ohio at Toledo, Toledo, Ohio 43699 NELSON, DR. MARGARET C., Section on Neurobiology and Behavior, Cornell University, Ithaca, New York 14850 NICHOLLS, DR. JOHN G., Department of Neurobiology, Stanford University, Stanford, California 94305 NICOSIA, DR. SANTO V., M.D., Department of Obstetrics and Gynecology, Division of Reproductive Biology, School of Medicine, University of Penn- sylvania, Philadelphia, Pennsylvania 19174 NIELSEN, DR. JENNIFER B. K., Waksman Institute for Microbiology, Rutgers University, Piscataway, New Jersey 08854 24 MARINE BIOLOGICAL LABORATORY NOE, DR. BRYAN D., Department of Anatomy, Emory University, Atlanta, Georgia 30345 OCHOA, DR. SEVERO, 530 East 72nd Street, New York, New York 10021 ODUM, DR. EUGENE, Department of Zoology, University of Georgia, Athens, Georgia 30601 OERTEL, DR. DONATA, Department of Neurophysiology, University of Wisconsin, 283 Medical Science Building, Madison, Wisconsin 53706 O'HERRON, JONATHAN, Lazard Freres and Company, 1 Rockefeller Plaza, New York, New York 10020 OLSON, DR. JOHN M., Department of Biology, Brookhaven National Laboratory, Upton, New York 11973 OSCHMAN, DR. JAMES L., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 OXFORD, DR. GERRY S., Department of Physiology, University of North Carolina, Chapel Hill, North Carolina 27514 PALMER, DR. JOHN D., Department of Zoology, University of Massachusetts, Amherst, Massachusetts 01002 PALTI, DR. YORAM, Head, Department of Biophysics, University of Maryland, Baltimore, Maryland 21201 PAPPAS, DR. GEORGE D., Department of Anatomy, College of Medicine, Uni- versity of Illinois, 808 S. Wood St., Chicago, Illinois 60612 PARDEE, DR. ARTHUR B., Department of Pharmacology, Harvard Medical School, Boston, Massachusetts 02115 PARDY, DR. ROOSEVELT L., School of Life Sciences, University of Nebraska, Lincoln, Nebraska 68588 PARMENTIER, DR. JAMES L., Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina 27710 PASSANO, DR. LEONARD M., Department of Zoology, Birge Hall, University of Wisconsin, Madison, Wisconsin 53706 PEARLMAN, DR. ALAN L., Department of Physiology, School of Medicine, Washington University, St. Louis, Missouri 63110 PEDERSON, DR. THORU, Worcester Foundation for Experimental Biology, Shrewsbury, Massachusetts 01545 PERKINS, DR. C. D., National Academy of Engineering, 2101 Constitution Ave., N. W., Washington, D. C. 20418 PERSON, DR. PHILIP, Special Dental Research Program, Veteran's Administra- tion Hospital, Brooklyn, New York 11219 PETERSON, DR. BRUCE J., Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 PETTIBONE, DR. MARIAN H., Division of Worms, W-213, Smithsonian Institu- tion, Washington, D. C. 20560 PFOHL, DR. RONALD J., Department of Zoology, Miami University, Oxford, Ohio 45056 PHILPOTT, DR. DELBERT E., Ames Research Center, Moffett Field, California 94035 PIERCE, DR. SIDNEY K., JR., Department of Zoology, University of Maryland, College Park, Maryland 20740 POLLARD, DR. HARVEY B., National Institutes of Health, F. Bldg. 10, Rm. 10B17, Bethesda, Maryland 20014 POLLARD, DR. THOMAS D., M.D., Director, Department of Cell Biology and MEMBERS OF THE CORPORATION 25 Anatomy, The Johns Hopkins University School of Medicine, 725 N. Wolfe St., Baltimore, Maryland 21205 POLLOCK, DR. LELAND \\ ., Department of Zoology, Drew University, Madison, New Jersey 07940 PORTER, DR. BEVERLY H., 14433 Taos Court, Wheaton, Maryland 20900 PORTER, DR. KEITH R. , 748 llth Street, Boulder, Colorado 80302 POTTER, DR. DAVID, Department of Neurobiology, Harvard Medical School, Boston, Massachusetts 02115 POTTER, DR. H. DAVID, Neural Sciences Program, Chemistry Building, Indiana University, Bloomington, Indiana 47401 POTTS, DR. WILLIAM T., Department of Biology, University of Lancaster, Lancaster, England, U. K. POUSSART, DR. DENIS, Department of Electrical Engineering, Universite Laval, Quebec, Canada PRENDERGAST, DR. ROBERT A., Department of Pathology and Ophthalmology, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 PRICE, DR. CARL A., Waksman Institute of Microbiology, Rutgers University, P. O. Box 759, Piscataway, New Jersey 08854 PRICE, DR. CHRISTOPHER H., Biological Science Center, 2 Cummington Street, Boston, Massachusetts 02215 PRIOR, DR. DAVID JAMES, Department of Biological Sciences, LIniversity of Kentucky, Lexington, Kentucky 40506 PROSSER, DR. C. LADD, Department of Physiology and Biophysics, Burrill Hall 524, University of Illinois, Urbana, Illinois 61801 PROVASOLI, DR. LUIGI, Haskins Laboratories, 165 Prospect Street, New Haven, Connecticut 06520 PRUSCH, DR. ROBERT D., Division of Biomedical Sciences, Brown University, Providence, Rhode Island 02908 PRZYBYLSKI, DR. RONALD J., Department of Anatomy, Case Western Reserve University, Cleveland, Ohio 44101 RABIN, DR. HARVEY, P. O. Box 239, Braddock Hts., Maryland 21714 RAMON, DR. FIDEL, Department of Physiology, Duke University Medical Center, Durham, North Carolina 27706 RANKIN, DR. JOHN S., Box 97, Ashford, Connecticut 06278 RANZI, DR. SILVIO, Department of Zoology, University of Milan, Via Celonia 10, Milan, Italy RATNER, DR. SARAH, Department of Biochemistry, The Public Health Research Institute, 455 First Avenue, New York, New York 10016 REBHUN, DR. LIONEL I., Department of Biology, Gilmer Hall, University of Virginia, Charlottesville, Virginia 22901 REDDAN, DR. JOHN R., Department of Biological Sciences, Oakland LIniversity, Rochester, Michigan 48063 REDFIELD, DR. ALFRED C., 10 Maury Lane, Woods Hole, Massachusetts 02543 REESE, DR. THOMAS S., Head, Section on Functional Neuroanatomy, National Institutes of Health, Bethesda, Maryland 20014 REINER, DR. JOHN M., Department of Biochemistry, Albany Medical College of Union University, Albany, New York 12208 REINISCH, DR. CAROL L., Department of Tumor Immunology, Sidney Farber Cancer Institute, 44 Binney St., Boston, Massachusetts 02511 26 MARINE BIOLOGICAL LABORATORY REUBEN, DR. JOHN P., Department of Neurology, Columbia University, College of Physicians and Surgeons, New York, New York 10032 REYNOLDS, DR. GEORGE THOMAS, Department of Physics, Princeton University, Princeton, New Jersey 08540 RICE, DR. ROBERT VERNON, Carnegie-Mellon Institute, 4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213 RICKLES, DR. FREDERICK R., M.D., University of Connecticut, School of Medi- cine, VA Hospital, Newington, Connecticut 06111 RIPPS, DR. HARRIS, Department of Opthalmology, New York University, School of Medicine, 550 First Ave., New York, New York 10016 ROBERTS, DR. JOHN L., Department of Zoology, University of Massachusetts, Amherst, Massachusetts 01002 ROBINSON, DR. DENIS M., 19 Orlando Avenue, Arlington, Massachusetts 02174 ROCKSTEIN, DR. MORRIS, Department of Physiology, University of Miami School of Medicine, P. O. Box 016430, Miami, Florida 33101 RONKIN, DR. RAPHAEL E., 3212 McKinley St., N. W., Washington, D. C. 20015 ROSE, DR. BIRGIT, Department of Physiology, University of Miami School of Medicine, P. O. Box 016430, Miami, Florida 33101 ROSE, DR. S. MERYL, Box 309W, Waquoit, Massachusetts 02536 ROSENBAUM, DR. JOEL L., Kline Biology Tower, Yale University, New Haven, Connecticut 06510 ROSENBERG, DR. EVELYN K., Jersey City State College, Jersey City, New Jersey 07305 ROSENBERG, DR. PHILLIP, Division of Pharmacology, University of Connecticut, School of Pharmacy, Storrs, Connecticut 06268 ROSENBLUTH, DR. JACK, Department of Physiology, New York University, School of Medicine, 550 First Avenue, New York, New York 10016 ROSENBLUTH, RAJA, 3380 West 5th Avenue, Vancouver, British Columbia, Canada V6R 1R7 ROSENKRANZ, DR. HERBERT S., Department of Microbiology, New York Medical College, Valhalla, New York 10595 ROSLANSKY, DR. JOHN, Box 208, Woods Hole, Massachusetts 02543 ROSLANSKY, DR. PRISCILLA F., Box 208, Woods Hole, Massachusetts 02543 Ross, DR. WILLIAM N., Department of Neurobiology, Harvard Medical School, Boston, Massachusetts 02115 ROTH, DR. JAY S., Division of Biological Sciences, Section of Biochemistry and Biophysics, University of Connecticut, Storrs, Connecticut 06268 ROWE, DOROTHY, 88 Chestnut St., Boston, Massachusetts 02165 ROWLAND, DR. LEWIS P., Department of Neurology, Columbia University, College of Physicians and Surgeons, 630 W. 168th St., New York, New York 10032 RUBINOW, DR. SOL I., Department of Biomathematics, Cornell University, Medical College, New York, New York 10012 RUDERMAN, DR. JOAN V., Department of Anatomy, Harvard Medical School, Boston, Massachusetts 02115 RUSHFORTH, DR. NORMAN B., Department of Biology, Case Western Reserve University, Cleveland, Ohio 44106 RUSSELL, DR. JOHN M., Department of Biophysics, University of Texas, Medical Branch, Galveston, Texas 77550 RUSSELL-HUNTER, DR. W. D., Department of Biology, Lyman Hall, Syracuse University, Syracuse, New York 13210 MEMBERS OF THE CORPORATION 27 RUSTAD, DR. RONALD C., Department of Radiology, Case Western Reserve University, Cleveland, Ohio 44106 SAGER, DR. RUTH, Sidney Farber Cancer Center, 44 Binney St., Boston, Massa- chusetts 02115 SALMON, DR. EDWARD D., Department of Zoology, University of North Carolina, Chapel Hill, North Carolina 27514 SALZBERG, DR. BRIAN M., Department of Physiology, University of Pennsylvania, 4010 Locust St., Philadelphia, Pennsylvania 19174 SANDERS, DR. HOWARD L., Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543 SATO, DR. HIDEMI, Sugashima Marine Biological Laboratory, Nagoya University, Sugashima-cho, Toba-shi, Mie-ken, 517, Japan SAUNDERS, DR. JOHN W., JR., Department of Biological Sciences, State University of New York at Albany, Albany, New York 12222 SAZ, DR. ARTHUR KENNETH, Department of Microbiology, Georgetown Uni- versity Medical and Dental Schools, 3900 Reservoir Road, N. W., Washing- ton, D. C. 20051 SCHACHMAN, DR. HOWARD K., Department of Molecular Biology, University of California, Berkeley, California 94720 SCHIFF, DR. JEROME A., Institute for Photobiology of Cells and Organelles, Brandeis University, Waltham, Massachusetts 02154 SCHLESINGER, DR. R. WALTER, Department of Microbiology, Rutgers Medical School, P. O. Box 101, Piscataway, New Jersey 08854 SCHMEER, SISTER ARLINE C., American Cancer Research Center and Hospital, 6401 W. Colfax Ave., Denver, Colorado 80214 SCHNEIDERMAN, DR. HOWARD A., Monsanto Co., 800 N. Lindberg Blvd. (D1W), St. Louis, Missouri 63166 SCHOLANDER, DR. P. F., Scripps Institution of Oceanography, La Jolla, Cali- fornia 92038 SCHOPF, DR. THOMAS J. M., Department of the Geophysical Sciences, University of Chicago, 5734 S. Ellis Avenue, Chicago, Illinois 60637 SCHOTTE, DR. OSCAR E., Department of Biology, Amherst College, Amherst, Massachusetts 01002 SHOUKIMAS, DR. JONATHAN J., Laboratory of Biophysics, NINCDS, Marine Biological Laboratory, Woods Hole, Massachusetts, 02543 SCHUEL, DR. HERBERT, Department of Anatomical Sciences, State University of New York, Buffalo, New York 14214 SCHUETZ, DR. ALLEN WALTER, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, Maryland 21205 SCHWARTZ, DR. TOBIAS L., Biological Sciences Group, University of Connecticut, Storrs, Connecticut 06268 SCOTT, DR. ALLAN C., 1 Nudd St., Waterville, Maine 04901 SCOTT, DR. GEORGE T., Department of Biology, Oberlin College, Oberlin, Ohio 44074 SEARS, DR. MARY, Box 152, Woods Hole, Massachusetts 02543 SEGAL, DR. SHELDON J., Director, Population Division, The Rockefeller Founda- tion, 1133 Avenue of the Americas, New York, New York 10036 SELIGER, DR. HOWARD H., McCollum-Pratt Institute, The Johns Hopkins University, Baltimore, Maryland 21218 SELMAN, DR. KELLY, Department of Anatomy, College of Medicine, University of Florida, Gainesville, Florida 32601 28 MARINE BIOLOGICAL LABORATORY SENFT, DR. JOSEPH P., Organic Gardening and Farming Research Center, Box 323. R.D.I, Kutztown, Pennsylvania 19530 SHANKLIX. DR. DOUGLAS R., P. O. Box 1267, Gainesville, Florida 32602 SHAPIRO, DR. HERBERT, 6025 North 13th Street, Philadelphia, Pennsylvania 10141 SHAVER, DR. JOHN R., Department of Zoology, Michigan State University, East Lansing, Michigan 48823 SHEPARD, DR. DAVID C., P. O. Box 44, Woods Hole, Massachusetts 02543 SHKPRO, DR. DAVID, Department of Biology, Boston University, 2 Cummington Street, Boston, Massachusetts 02215 SHERMAN, DR. I. W., Division of Life Sciences, University of California, Riverside, California 92502 SHILO, DR. MOSHE, Head, Department of Microbiological Chemistry, Hebrew University, Jerusalem, Israel SIEGEL, DR. IRWIN M., Department of Ophthalmology, New York University Medical Center, 550 First Avenue, New York, New York 10016 SIEGELMAN, DR. HAROLD W., Department of Biology, Brookhaven National Laboratory, Upton, New York 11973 SIMON, DR. ERIC J., New York University Medical School, 550 First Avenue, New York, New York 10016 SJODIN, DR. RAYMOND A., Department of Biophysics, University of Maryland School of Medicine, Baltimore, Maryland 21201 SKINNER, DR. DOROTHY M., Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830 SLOBODKIN, DR. LAWRENCE B., Department of Biology, State University of New York at Stony Brook, Stony Brook, New York 11790 SMITH, HOMER P., General Manager, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 SMITH, DR. MICHAEL A., Marine Biological Laboratory, Woods Hole, Massa- chusetts 02543 SMITH, PAUL FERRIS, P. O. Box 264, Woods Hole, Massachusetts 02543 SMITH, DR. RALPH I., Department of Zoology, University of California, Berkeley, California 94720 SONNENBLICK, DR. B. P., Department of Zoology and Physiology, Rutgers Uni- versity, 195 University Avenue, Newark, New Jersey 07102 SORENSON, DR. ALBERT L., Department of Physiology, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, New York 10461 SORENSON, DR. MARTHA M., Department of Neurology, Columbia University, College of Physicians and Surgeons, New York, New York 10032 SPECK, DR. WILLIAM T., M.D., Department of Pediatrics, Case Western Reserve University, Cleveland, Ohio 44106 SPECTOR, DR. A., Black Bldg., Rm. 1516, Columbia University, College of Physicians and Surgeons, New York, New York 10032 SPIEGEL, DR. EVELYN, Department of Biological Sciences, Dartmouth College, Hanover, New Hampshire 03755 SPIEGEL, DR. MELVIN, Department of Biological Sciences, Dartmouth College, Hanover, New Hampshire 03755 SPIRTES, DR. MORRIS ALBERT, M.D., Veteran's Administration Hospital, 1601 Perdido Street, New Orleans, Louisiana 70112 SPRAY, DR. DAVID C., Department of Neuroscience, Albert Einstein College of Medicine, Bronx, New York 10461 MEMBERS OF THE CORPORATION 29 STARZAK, DR. MICHAEL E., Department of Chemistry, State University of New York, Binghamton, New York 13901 STEELE, DR. JOHN HYSLOP, Director, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543 STEINBERG, DR. MALCOLM, Department of Biology, Princeton University, Princeton, New Jersey 08540 STEINACHER, DR. ANTOINETTE, Department of Biophysics, The Rockefeller University, New York, New York 10021 STEINHARDT, DR. JACINTO, 306 Reiss Bldg., Georgetown University, Washington, D. C. 20007 STEPHENS, DR. G ROVER C., School of Biological Sciences, University of Cali- fornia, Irvine, California 92717 STEPHENS, DR. RAYMOND E., Marine Biological Laboratory, Woods Hole, Massachusetts 02543 STETTEN, DR. MARJORIE R., National Institutes of Health, Bldg. 10, 9B-02, Bethesda, Maryland 20014 STETTEN, DR. DEWITT, JR., M.D., Pn.D., Senior Scientific Advisor, NIH, Build- ing 16, Room 118, Bethesda, Maryland 20014 STOKES, DR. DARRELL R., Department of Biology, Emory University, Atlanta, Georgia 30322 STRACHER, DR. ALFRED, State University of New York, Downstate Medical Center, 450 Clarkson Avenue, Brooklyn, New York 11203 STREHLER, DR. BERNARD L., 2310 N. Laguna Circle Dr., Agoura, California 91301 STRETTON, DR. ANTHONY O. W., Department of Zoology, University of Wis- consin, Madison, Wisconsin 53706 STUART, DR. ANN E., Medical Sciences Research Wing 206H, Department of Physiology, University of North Carolina, Chapel Hill, North Carolina 27514 SUMMERS, DR. WILLIAM C., Huxley College, Western Washington State College, Bellingham, Washington 98225 SUSSMAN, DR. MAURICE, Department of Life Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260 SZABO, DR. GEORGE, Harvard School of Dental Medicine, 188 Longwood Avenue, Boston, Massachusetts 02115 SZAMIER, DR. ROBERT BRUCE, Harvard Medical School, Berman-Gund Labora- tory, Eye and Ear Infirmary, 243 Charles Street, Boston, Massachusetts 02114 SZENT-GYORGYI, DR. ALBERT, Institute for Muscle Research, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 SZENT-GYORGYI, DR. ANDREW G., Department of Biology, Brandeis University, Waltham, Massachusetts 02154 TAKASHIMA, DR. SHIRO, Department of Bioengineering, University of Pennsyl- vania, Philadelphia, Pennsylvania 19174 TANZER, DR. MARVIN L., Department of Biochemistry, Box G, University of Connecticut, School of Medicine, Farmington, Connecticut 06032 TASAKI, DR. ICHIJI, Laboratory of Neurobiology, National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland 20014 TAYLOR, DR. DOUGLAS L., The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138 TAYLOR, DR. ROBERT E., Laboratory of Biophysics, National Institute of Neu- rological Diseases and Stroke, National Institutes of Health, Bethesda, Maryland 20014 30 MARINE BIOLOGICAL LABORATORY TAYLOR, DR. W. ROWLAND, 1540 Northbourne Rd., Baltimore, Maryland 21239 TELFER, DR. WILLIAM H., Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19174 DETERRA, DR. NOEL, Department of Anatomy, Hahnemann Medical College, 230 N. Broad St., Philadelphia, Pennsylvania 19102 THORNDIKE, W. NICHOLAS, Wellington Management Company, 28 State St., Boston, Massachusetts 02109 TIFFNEY, DR. WESLEY N., 226 Edge Hill Rd., Sharon, Massachusetts 02067 TRACER, DR. WILLIAM, The Rockefeller University, 66th Street and York Avenue, New York, New York 10021 TRAVIS, DR. D. M., Department of Pharmacology, University of Florida, Gainesville, Florida 32610 TRINKAUS, DR. J. PHILIP, Osborn Zoological Laboratories, Department of Zoology, Yale University, New Haven, Connecticut 06510 TROLL, DR. WALTER, Department of Environmental Medicine, New York University, College of Medicine, New York, New York 10016 TROXLER, DR. ROBERT F., Department of Biochemistry, Boston University School of Medicine, 80 E. Concord St., Boston, Massachusetts 02118 TURNER, DR. RUTH D., Mollusk Department, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138 TWEEDELL, DR. KENYON S., Department of Biology, University of Notre Dame, Notre Dame, Indiana 46656 URETZ, DR. ROBERT B., Division of Biological Sciences, University of Chicago, 950 E. 59th St., Box 417, Chicago, Illinois 60637 VALIELA, DR. IVAN, Boston University Marine Program, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 VALOIS, JOHN, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 VAN HOLDE, DR. KENSAL EDWARD, Department of Biochemistry and Biophysics, Oregon State University, Corvallis, Oregon 97331 VILLEE, DR. CLAUDE A., Department of Biological Chemistry, Harvard Medical School, Boston, Massachusetts 02115 VINCENT, DR. WALTER S., Chairman, Department of Biological Sciences, Uni- versity of Delaware, Newark, Delaware 19711 WAINIO, DR. W. W., Box 1059, Nelson Labs., Rutgers Biochemistry, Piscataway, New Jersey 08854 WAKSMAN, DR. BYRON, Department of Pathology, Yale University, New Haven, Connecticut 06510 WALKER, DR. CHARLES A., 3113 Shamrock South, Tallahasee, Florida 32303 WALL, DR. BETTY J., Marine Biological Laboratory, Woods Hole, Massachusetts 02543 WALLACE, DR. ROBIN A., P. O. Box Y, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830 WANG, DR. A., Bedford Road, Lincoln, Massachusetts 01773 WARNER, DR. ROBKRT C., Department of Molecular and Cell Biology, Uni- versity of California, Irvine, California 92717 WARREN, DR. LI-:ONARD, Department of Therapeutic Research, Anatomy- Chemistry Bldg., Rm. 337, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19174 WATERMAN, DR. T. H., 610 Kline Biology Tower, Yale University, New Haven, Connecticut 06520 MEMBERS OF THE CORPORATION 31 WATSON, DR. STANLEY WAYNE, Woods Hole Oceanogruphir Institution, Woods Hole, Massachusetts 02543 WEBB, DR. H. MARGUERITE, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 WEBER, DR. ANNEMARIE, M.D., Department of Biochemistry, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19174 WEBSTER, DR. FERRIS, 800 25th Street, N. W., Washington, D. C. 20037 WEIDNER, DR. EARL, Department of Zoology and Physiology, Louisiana State University, Baton Rouge, Louisiana 70803 WEISS, DR. LEON P., Department of Animal Biology, University of Pennsylvania, School of Veterinary Medicine, Philadelphia, Pennsylvania 19174 WEISSMANN, DR. GERALD, School of Medicine, New York University, 550 First Avenue, New York, New York 10016 WERMAN, DR. ROBERT, Neurobiology Unit, Hebrew University, Jerusalem, Israel WHITTAKER, DR. J. RICHARD, Wistar Institute for Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania 19174 WIERCINSKI, DR. FLOYD J., Department of Biology, Northeastern Illinois Uni- versity, 5500 North St. Louis Ave., Chicago, Illinois 60625 WIGLEY, DR. ROLAND L., National Marine Fisheries Service, Woods Hole, Massachusetts 02543 WILBER, DR. C. G., Chairman, Department of Zoology, Colorado State Uni- versity, Fort Collins, Colorado 80524 WILSON, DR. DARCY B., Department of Pathology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19174 WILSON, DR. EDWARD O., Department of Zoology, Harvard University, Cam- bridge, Massachusetts 02138 WILSON, DR. T. HASTINGS, Department of Physiology, Harvard Medical School, Boston, Massachusetts 02115 WILSON, DR. WALTER L., Department of Biology, Oakland University, Rochester, Michigan 48063 WITKOVSKY, DR. PAUL, Department of Ophthalmology, New York University Medical Center, 550 First Ave., New York, New York 10016 \VITTENBERG, DR. JONATHAN B., Department of Physiology and Biochemistry, Albert Einstein College of Medicine, New York, New York 10461 WTOELKERLING, DR. WILLIAM J., Department of Botany, Latrobe University, Bundoora, Victoria, Australia 3083 WOLF, DR. DON P., University of Pennsylvania, School of Medicine, 314 Medical Labs 6/3, Philadelphia, Pennsylvania 19174 WOODWELL, DR. GEORGE M., Director, The Ecosystems Center, Marine Bio- logical Laboratory, Woods Hole, Massachusetts 02543 WYTTENBACH, DR. CHARLES R., Department of Physiology and Cell Biology, University of Kansas, Lawrence, Kansas 66045 YEH, JAY Z., Department of Pharmacology, Northwestern University Medical School, 303 E. Chicago Ave., Chicago, Illinois 60611 YNTEMA, DR. C. L., Department of Anatomy, State University of New York, Upstate Medical Center, Syracuse, New York 13210 YOUNG, DR. DAVID K., Department of the Navy, NORDA. Code 334, NSTL Station, Mississippi 39529 YPHANTIS, DR. DAVID A., Department of Biochemistry and Biophysics, Uni- versity of Connecticut, Storrs, Connecticut 06268 32 MARINE BIOLOGICAL LABORATORY ZIGMAN, DR. SEYMOUR, University of Rochester School of Medicine and Den- tistry, 260 Crittenden Boulevard, Rochester, New York 14620 ZIMMERMAN, DR. A. M., Department of Zoology, University of Toronto, To- ronto, Ontario, Canada ASSOCIATE MEMBERS ACHESON, DR. AND MRS. GEORGE ACKROYD, DR. AND MRS. FREDERICK W. ADELBERG, DR. AND MRS. EDWARD A. ADELMAN, DR. AND MRS. WILLIAM J. AHEARN, MR. AND MRS. DAVID C. ALLEN, Miss CAMILLA K. ALLEN, MRS. ROBERT D. AMBERSON, MRS. WILLIAM R. ARMSTRONG, DR. PHILIP ARMSTRONG, DR. AND MRS. SAMUEL C. ARNOLD, DR. AND MRS. JOHN ATWOOD, DR. AND MRS. KIMBALL C. BACON, MR. ROBERT BALL, MRS. ERIC G. BALLANTINE, DR. AND MRS. H. T., JR. BANG, DR. AND MRS. FREDERIK B. BANKS, MRS. AND MRS. W. L. BARROWS, MRS. ALBERT W. BENNETT, DR. AND MRS. MICHAEL V. L. BERNSTEIN, MR. AND MRS. HERMAN BERNHEIMER, DR. ALAN W. BIGELOW, MRS. ROBERT BLACKBURN, DR. AND MRS. GEORGE L. BODEEN, MR. AND MRS. GEORGE H. BOETTIGER, DR. AND MRS. EDWARD G. BOLTON, MR. AND MRS. THOMAS C. BORGESE, MRS. THOMAS A. BOTKIN, DR. DANIEL B. BOWLES, MR. AND MRS. FRANCIS P. BRADLEY, DR. AND MRS. CHARLES C. BRONSON, MR. AND MRS. SAMUEL C. BROWN, DR. AND MRS. DUGALD E. S. BROWN, DR. AND MRS. F. A., JR. BROWN, DR. AND MRS. THORNTON BUCK, MRS. JOHN B. BUFFINGTON, MRS. ALICE H. BUFFINGTON, MRS. GEORGE BURGER, DR. AND MRS. MAX M. BURRELL, MRS. PETER E. BURROUGH, MRS. ARNOLD H. BURT, MR. AND MRS. CHARLES E. BUTLER, MR. AND MRS. RIIETT W. BUTLER, MRS. E. G. CALKINS, MR. AND MRS. G. N., JR. CAMPBELL, DR. AND MRS. DAVID G. CAMPBELL, MR. AND MRS. WORTHING- TON, JR. CAPOBIANCO, MR. AND MRS. PAT J. CARLSON, DR. AND MRS. FRANCIS CARLTON, MR. AND MRS. WINSLOW G. CASHMAN, MR. AND MRS. EUGENE R. CHAMBERS, DR. AND MRS. EDWARD L. CHENEY, DR. RALPH H. CLAFF, MR. AND MRS. MARK CLARK, MR. AND MRS. HAYS CLARK, MRS. JAMES McC. CLARK, DR. AND MRS. LEONARD B. CLARK, MR. AND MRS. LEROY, JR. CLARK, MR. AND MRS. W. VAN ALAN CLEMENT, DR. AND MRS. A. C. CLOWES FUND, INC. CLOWES, MR. ALLEN W. CLOWES, DR. AND MRS. G. H. A., JR. COHEN, DR. AND MRS. SEYMOUR CONNELL, MR. AND MRS. W. J. COOPER, MR. AND MRS. JOHN H., JR. COPELAND, MRS. D. EUGENE COPELAND, MR. AND MRS. PRESTON S. COSTELLO, MRS. DONALD P. CRAIN, MR. AND MRS. MELVIN C. CRAMER, MR. AND MRS. IAN D. W. CRANE, MR. JOHN CRANE, JOSEPHINE FOUNDATION CRANE, MRS. W. CAREY CROSS, MR. AND MRS. NORMAN C. CROSSLEY, MR. AND MRS. ARCHIBALD M. CROWELL, DR. AND MRS. SEARS CURTIS, DR. AND MRS. W. D. DAIGNAULT, MR. AND MRS. A. T. DANIELS, MR. AND MRS. BRUCE G. DANIELS, MRS. F. HAROLD DAY, MR. AND MRS. POMEROY DRUMMOND, MR. AND MRS. A. H., JR. DuBois, DR. AND MRS. A. B. DUNKERLEY, MR. AND MRS. H. GORDON MEMBERS OF THE CORPORATION 33 DuPoNT, MR. A. FELIX, JR. DYER, MR. AND MRS. ARNOLD VV. EASTMAN, MR. AND MRS. CHARLES E. EBERT, DR. AND MRS. JAMES D. EDWARDS, DR. AND MRS. ROBERT L. EGLOFF, DR. AND MRS. F. R. L. ELLIOTT, MRS. ALFRED M. ELSMITH, MRS. DOROTHY (). EPEL, MRS. DAVID EVANS, MR. AND MRS. DUDLEY EWING, DR. AND MRS. GIFFORD C. FENNO, MRS. EDWARD N. FERGUSON, DR. AND MRS. J. J., JR. FINE, DR. AND MRS. JACOB FIRESTONE, MR. AND MRS. EDWIN FISHER, MRS. B. C. FISHER, MR. FREDERICKS., Ill FISHER, DR. AND MRS. SAUL H. FRANCIS, MR. AND MRS. LEWIS W., JR. FRIES, DR. AND MRS. E. F. B. FYE, DR. AND MRS. PAUL M. GABRIEL, DR. AND MRS. MORDECAI L. GAISER, DR. AND MRS. DAVID W. GARFIELD, Miss ELEANOR GARREY, DR. AND MRS. WALTER GELLIS, DR. AND MRS. SYDNEY GERMAN, DR. AND MRS. JAMES L., Ill GIFFORD, MR. AND MRS. JOHN A. GIFFORD, DR. AND MRS. PROSSER GILBERT, DR. AND MRS. DANIEL L. GILBERT, MRS. HELEN H. GILDEA, DR. MARGARET C. L. GILLETTE, MR. AND MRS. ROBERT S. GLASS, DR. AND MRS. H. BENTLEY GLAZEBROOK, MRS. JAMES R. GLUSMAN, DR. AND MRS. MURRAY GOLDMAN, DR. AND MRS. ALLEN S. GOLDRING, DR. IRENE P. GOLDSTEIN, MRS. MOISE H., JR. GRANT, DR. AND MRS. PHILIP GRASSLE, MR. AND MRS. J. K. GREEN, Miss GLADYS M. GREENE, MR. AND MRS. WILLIAM C. GREER, MR. AND MRS. W. H., JR. GREIF, DR. AND MRS. ROGER L. GROSCH, DR. AND MRS. DANIEL S. GROSS, MRS. PAUL R. GRUSON, MRS. MARTHA GUNNING, MR. AND MRS. ROBERT HAIGH, MR. AND MRS. RICHARD H. HALVORSON, DR. AND MRS. HARLYN O. HANDLER, DR. AND MRS. PHILIP HARVEY, DR. AND MRS. RICHARD B. HASKINS, MRS. CARYL I'. HASSETT, DR. AND MRS. CHARLES HASTINGS, MRS. J. WOODLAND HAWKINS, MR. RICHARD H., Ill HEFFRON, DR. AND MRS. RODERICK HENLEY, DR. CATHERINE HIAM, MR. AND MRS. E. W. HIATT, DR. AND MRS. HOWARD HIBBARD, DR. HOPE HILL, MRS. SAMUEL E. HlLSINGER, MR. AND MRS. ARTHUR HlRSCHFELD, MRS. NATHAN B. HOBBIE, DR. AND MRS. JOHN HOCKER, MR. AND MRS. LON HOFFMAN, REV. AND MRS. CHARLES HORWITZ, DR. AND MRS. NORMAN H. HOUGH, MRS. GEORGE A., JR. HOUSTON, MR. AND MRS. HOWARD E. HUETTNER, DR. AND MRS. ROBERT HUNZIKER, MR. AND MRS. HERBERT E. HUTCHINSON, MR. AND MRS. JOHN HYNES, MR. AND MRS. THOMAS J. INOUE, MRS. SHINYA IRELAND, MRS. HERBERT A. ISSOKSON, MR. AND MRS. ISRAEL IVENS, DR. SUE JACKSON, Miss ELIZABETH B. JANNEY, MRS. WISTAR JEWETT, G. F., FOUNDATION JEWETT, MR. AND MRS. G. F., JR. JONES, MR. AND MRS. DEWITT, III JONES, MR. AND MRS. FREDERICK, III JORDAN, DR. AND MRS. EDWIN P. KAAN, DR. HELEN Wr. KAHLER, MR. AND MRS. GEORGE A. KAHLER, MRS. ROBERT W. KAMINER, MRS. BENJAMIN KARUSH, DR. AND MRS. FRED KEITH, MR. JEAN R. KELLEHER, MR. AND MRS. PAUL R. KEOSIAN, MRS. JESSIE KlEN, MR. AND MRS. PlETER KINGWELL, THE REV. AND MRS. WILBUR J. KINNARD, MRS. L. RICHARD KIVY, DR. AND MRS. PETER KOELSCH, MR. AND MRS. HERBERT KOHN, DR. AND MRS. HENRY I. ROLLER, DR. AND MRS. LEWIS R. 34 MARINE BIOLOGICAL LABORATORY KUFFLER, MRS. STEPHEN \Y. LADERMAN, MR. AND MRS. EZRA LASH, DR. AND MRS. JAMES LASTER, DR. AND MRS. LEONARD LAUFER, DR. AND MRS. HANS LAWRENCE, MR. FREDERICK V. LAWRENCE, MRS. WILLIAM LAZAROW, MRS. ARNOLD LEATHERBEE, MR. JOHN A. LEMANN, MRS. LUCY B. LENHER, MR. AND MRS. SAMUEL LEVINE, DR. AND MRS. RACHMIEL LEVENTHAL, Ms. MONIKA MEYER LEWIS, MR. T. HOHN LILLIE, MRS. KARL C. LILLY, MR. AND MRS. JOSIAH K., Ill LOEB, MRS. ROBERT F. LONG, MRS. G. C. LORAND, MRS. LASZLO LOVELL, MR. AND MRS. HOLLIS R. LOWENGARD, MRS. JOSEPH LOWE, DR. AND MRS. CHARLES W. LURIA, DR. AND MRS. S. E. MACKEY, MR. AND MRS. WILLIAM K. MACLEISH, MRS. MARGARET MACNARY, MR. B. GLENN MAcNiCHOL, DR. AND MRS. EDWARD F., JR. MAKER, Miss ANNE CAMILLE MARKS, DR. AND MRS. PAUL A. MARSLAND, DR. AND MRS. DOUGLAS MARTYNA, MR. AND MRS. JOSEPH MARVIN, DR. DOROTHY H. MASER, DR. AND MRS. MORTON MASTROIANNI, DR. AND MRS. L., JR. MATHER, MR. AND MRS. FRANK J., Ill MATTHIESSEN, MR. AND MRS. G. C. MAYOR, MRS. JAMES W., SR. McCuSKER, MR. AND MRS. PAUL T. MCELROY, MRS. NELLA W. McLANE, MRS. T. THORNE MEIGS, MR. AND MRS. ARTHUR MEIGS, DR. AND MRS. J. WISTER MELILLO, DR. AND MRS. JERRY THE MELLON FOUNDATION MELLON, MR. AND MRS. RICHARD P. METZ, MRS. CHARLES B. MEYERS, MR. AND MRS. RICHARD MILLER, DR. DANIEL A. MIXTER, MR. AND MRS. W. J., JR. MONTGOMERY, DR. AND MRS. CHARLES H. MONTGOMERY, MR. AND MRS. RAYMOND B. MOORE, MR. JOHN W. MORSE, MR. AND MRS. CHARLES L., JR. MORSE, MR. AND MRS. RICHARD S. MOSES, MR. AND MRS. GEORGE L. MOUL, MRS. EDWIN T. NEUBERGER, MRS. HARRY H. NEWTON, C. H., BUILDERS, INC. NICHOLS, MRS. GEORGE NlCKERSON, MR. AND MRS. FRANK L. NORMAN, MR. AND MRS. ANDREW E. NORMANDIE FOUNDATION O'HERRON, MR. AND MRS. JONATHAN O'SULLIVAN, DR. RENEE BENNETT OLMSTED, MR. AND MRS. CHRISTOPHER ORTINS, MR. ARMAND PAPPAS, DR. AND MRS. GEORGE D. PARK, MR. AND MRS. FRANKLIN A. PARK, MR. AND MRS. MALCOLM S. PARMENTER, Miss CAROLYN L. PARMENTIER, MR. GEORGE L. PATTEN, MRS. BRADLEY M. PECAN, Ms. ERENE V. PENDERGAST, MRS. CLAUDIA PENDELTON, DR. AND MRS. MURRAY E. PENNINGTON, Miss ANNE H. PERKINS, MR. AND MRS. COURTLAND D. PERSON, DR. AND MRS. PHILIP PETERSON, MR. AND MRS. E. GUNNAR PETERSON, MR. AND MRS. E. JOEL PETERSON, MR. AND MRS. RAYMOND W. PHILIPPE, MR. AND MRS. PIERRE PORTER, DR. AND MRS. KEITH R. PROSSER, MRS. C. LADD PUTNAM, MR. ALLAN RAY PUTNAM, MR. AND MRS. W. A., Ill RATCLIFFE, MR. THOMAS G., JR. RAYMOND, DR. AND MRS. SAMUEL READ, Ms. LEE REDFIELD, DR. AND MRS. ALFRED C. RENEK, MR. AND MRS. MORRIS REYNOLDS, DR. AND MRS. GEORGE REYNOLDS, MR. AND MRS. JAMES T. REZNIKOFF, DR. AND MRS. PAUL RICCA, DR. AND MRS. RENATO A. RIGGS, MR. AND MRS. LAWRASON, III CERTIFICATE OF ORGANIZATION RIINA, MR. AND MRS. JOHN R. ROBB, Ms. ALISON A. ROBERTSON, MRS. C. STUART ROBERTSON, DR. AND MRS. C. \V. ROBINSON, UR. AND MRS. DENIS M. ROGERS, MRS. JULIAN ROOT, MRS. WALTER S. Ross, MRS. JOHN ROWE, MRS. WILLIAM S. RUBIN, DR. JOSEPH RUGH, MRS. ROBERTS RUSSELL, MR. AND MRS. HENRY D. RYDER, MR. AND MRS. FRANCIS C. SAUNDERS, DR. AND MRS. JOHN W. SAUNDERS, MRS. LAWRENCE SAVERY, MR. ROGER SAWYER, MR. AND MRS. JOHN E. SCHLESINGER, MRS. R. WALTER SCOTT, MRS. GEORGE T. SCOTT, MRS. NORMAN E. SEARS, MR. AND MRS. HAROLD B. SEGAL, DR. AND MRS. SHELDON SHAPIRO, MRS. HARRIET S. SHEMIN, DR. AND MRS. DAVID SHEPROW, DR. AND MRS. DAVID SHERMAN, DR. AND MRS. IRWIN SlMKINS, MRS. WlLLARD S. SLATER, MR. DAVID SMITH, MR. AND MRS. DIETRICH C. SMITH, MRS. HOMER P. SMITH, MR. AND MRS. ROBERT I. SMITH, MR. VANDORN C. SNIDER, MR. ELIOT SONNEBEND, MR. AND MRS. PAUL SPECHT, MRS. HEINZ STEELE, MRS. M. EVELYN STEINBACH, DR. AND MRS. H. B. STETTEN, DR. AND MRS. DE\VITT, JR. STONE, DR. AND MRS. WILLIAM STRACHER, DR. AND MRS. ALFRED STUART, DR. ANN STUNKARD, DR. HORACE STURTEVANT, MRS. A. H. SWANSON, DR. AND MRS. CARL P. SWOPE, MR. AND MRS. GERARD L. SWOPE, MRS. GERARD, JR. SWOPE, Miss HENRIETTA H. TARTAKOFF, DR. HELEN TAYLOR, DR. AND MRS. W. RANDOLPH TIETJE, MR. AND MRS. EMIL D., JR. TITTLER, MRS. SYLVIA TODD, MR. AND MRS. GORDON F. TOLKAN, MR. AND MRS. NORMAN N. TOMPKINS, MRS. B. A. TRACER, MRS. WILLIAM TROLL, DR. AND MRS. WALTER TULLY, MR. AND MRS. GORDON F. VALOIS, MR. AND MRS. JOHN VAN BRUNT, MR. AND MRS. A. H., JR. VEEDER, MRS. RONALD A. VINCENT, MRS. HELEN J. WAITE, MR. AND MRS. CHARLES E. WAKSMAN, DR. AND MRS. BYRON H. WARE, MR. AND MRS. J. LINDSAY WARREN, DR. AND MRS. SHIELDS WATT, MR. AND MRS. JOHN B. WrEISBERG, MR. AND MRS. ALFRED M. WEXLER, ROBERT H., FOUNDATION WHEATLEY, DR. MARJORIE A. WHEELER, DR. AND MRS. PAUL S. WHEELER, DR. AND MRS. RALPH E. WHITNEY, MR. AND MRS. GEOFFREY G., JR. WlCHTERMAN, DR. AND MRS. RALPH WlCKERSHAM, MR. AND MRS. A. A. TlLNEY WlCKERSHAM, MRS. JAMES H., JR. WILHELM, DR. HAZEL S. WILSON, MR. AND MRS. ROBERT E., JR. WlTMER, DR. AND MRS. ENOS E. WOLFINSOHN, MR. AND MRS. WOLFE WOODWELL, MRS. GEORGE YNTEMA, DR. AND MRS. CHESTER L. ZINN, DR. AND MRS. DONALD J. ZIPF, DR. ELIZABETH ZWILLING, MRS. EDGAR III. CERTIFICATE OF ORGANIZATION (On File in the Office of the Secretary of the Commonwealth) No. 3170 We, Alpheus Hyatt, President, William Stanford Stevens, Treasurer, and William T. Sedgwick, Edward G. Gardiner, Susan Minis and Charles Sedgwick Minot being a majority of the Trustees of the Marine Biological Laboratory in compliance with the 36 MARINE BIOLOGICAL LABORATORY requirements of the fourth section of chapter one hundred and fifteen of the Public Statutes do hereby certify that the following is a true copy of the agreement of associa- tion to constitute said Corporation, with the names of the subscribers thereto :- \Ve. whose names are hereto subscribed, do, by this agreement, associate ourselves with the intention to constitute a Corporation according to the provisions of the one hundred and fifteenth chapter of the Public Statutes of the Commonwealth of Massa- chusetts, and the Acts in amendment thereof and in addition thereto. The name by which the Corporation shall be known is THE MARINE BIOLOGICAL LABORATORY. The purpose for which the Corporation is constituted is to establish and maintain a laboratory or station for scientific study and investigations, and a school for instruction in biology and natural history. The place within which the Corporation is established or located is the city of Boston within said Commonwealth. The amount of its capital stock is none. In Witness Whereof, we have hereunto set our hands, this twenty seventh day of February in the year eighteen hundred and eighty-eight, Alpheus Hyatt, Samuel Mills, William T. Sedgwick, Edward G. Gardiner, Charles Sedgwick Minot, William G. Farlow, William Stanford Stevens, Anna D. Phillips, Susan Minis, B. H. Van Yleck. That the first meeting of the subscribers to said agreement was held on the thirteenth day of March in the year eighteen hundred and eighty-eight. In Witness Whereof, we have hereunto signed our names, this thirteenth day of March in the year eighteen hundred and eighty-eight, Alpheus Hyatt, President, William Stanford Stevens, Treasurer, Edward G. Gardiner, William T. Sedgwick, Susan Mims, Charles Sedgwick Minot. (Approved on March 20, 1888 as follows: I hereby certify that it appears upon an examination of the within written certificate and the records of the corporation duly submitted to my inspection, that the require- ments of sections one, two and three of chapter one hundred and fifteen, and sections eighteen, twenty and twenty-one of chapter one hundred and six, of the Public Statutes, have been complied with and I hereby approve said certificate this twentieth day of March A.D. eighteen hundred and eighty-eight. CHARLES ENDICOTT Commissioner of Corporations) IV. ARTICLES OF AMENDMENT (On File in the Office of the Secretary of the Commonwealth) We, James D. Ebert, President, and David Shepro, Clerk of the Marine Biological Laboratory, located at Woods Hole, Massachusetts 0254,3, do hereby certify that the following amendment to the Articles of Organization of the Corporation was duly adopted at a meeting held on August 15, 1975, as adjourned to August 29, 1975, by BYLAWS 37 vote of 444 members, being at least two-thirds of its members legally qualified to vote in the meetings of the corporation : VOTED: That the Certificate of Organization of this corporation be and it hereby is amended by the addition of the following provisions: "No Officer, Trustee or Corporate Member of the corporation shall be personally liable for the payment or satisfaction of any obligation or liabilities incurred as a result of, or otherwise in connection with, any commitments, agreements, activities or affairs of the corporation. "Except as otherwise specifically provided by the Bylaws of the corpora- tion, meetings of the Corporate Members of the corporation may be held anywhere in the United States. "The Trustees of the corporation may make, amend or repeal the Bylaws of the corporation in whole or in part, except with respect to any pro- visions thereof which shall by law, this Certificate or the Bylaws of the corporation, require action by the Corporate Members." The foregoing amendment will become effective when these articles of amendment are filed in accordance with Chapter 180, Section 7 of the General Laws unless these articles specify, in accordance with the vote adopting the amendment, a later effective date not more than thirty days after such filing, in which event the amendment will become effective on such later date. In Witness whereof and Under the Penalties of Perjury, we have hereto signed our names this 2nd day of September, in the year 1975, James D. Ebert, President; David Shepro, Clerk. (Approved on October 24, 1975, as follows: I hereby approve the within articles of amendment and, the filing fee in the amount of $10 having been paid, said articles are deemed to have been filed with me this 24th day of October, 1975. PAUL GUZZI Secretary of the Commonwealth) V. BYLAWS OF THE CORPORATION OF THE MARINE BIOLOGICAL LABORATORY (Revised August 11, 1978) I. (A) The name of the Corporation shall be The Marine Biological Laboratory. The Corporation's purpose shall be to establish and maintain a laboratory or station for scientific study and investigation, and a school for instruction in biology and natural history. (B) Marine Biological Laboratory admits students without regard to race, color, sex, national and ethnic origin to all the rights, privileges, programs and activities generally accorded or made available to students in its courses. It does not discriminate on the basis of race, color, sex, national and ethnic origin in employment, administration of its educational policies, admissions policies, scholarship and other programs. II. (A) The members of the Corporation ("Members") shall consist of persons elected by the Board of Trustees, upon such terms and conditions and in accordance 38 MARINE BIOLOGICAL LABORATORY \vith such procedures, not inconsistent with law or these Bylaws, as may lie determined by said Board of Trustees. Except as provided below, any Member may vote at any meeting, either in person or by proxy executed no more than six months prior to the date of such meeting. Members shall serve until their death or resignation unless earlier removed, with or without cause, by the affirmative vote of two-thirds of the Trustees tht-n in office. Any Member who has attained the age of seventy years or has retired from his home institution shall automatically be designated a Life Member provided he signifies his wish to retain his membership. Life Members shall not have the right to vote and shall not be assessed for dues. (B) The Associates of the Marine Biological Laboratory shall be an unincorporated group of persons (including associations and corporations) interested in the Laboratory and shall be organized and operated under the general supervision and authority of the Trustees. III. The officers of the Corporation shall consist of a Chairman of the Board of Trustees, President, Director, Treasurer and Clerk, elected or appointed by the Trustees as set forth in Article IX. IV. The Annual Meeting of the Members shall be held on the Friday following the Second Tuesday in August in each year at the Laboratory in Woods Hole, Massachu- setts, at 9:30 a.m. Subject to the provisions of Article VIII (2), at such meeting the Members shall choose by ballot six Trustees to serve four years, and shall transact such other business as may properly come before the meeting. Special meetings of the Members may be called by the Chairman or Trustees to be held at such time and place as may be designated. V. Twenty five Members shall constitute a quorum at any meeting. Except as otherwise required by law or these Bylaws, the affirmative vote of a majority of the Members voting in person or by proxy at a meeting attended by a quorum (present in person or by proxy) shall constitute action on behalf of the Members. VI. (A) Inasmuch as the time and place of the Annual Meeting of Members are fixed by these Bylaws, no notice of the Annual Meeting need be given. Notice of any special meeting of Members, however, shall be given by the Clerk by mailing notice of the time and place and purpose of such meeting, at least 15 days before such meeting, to each Member at his or her address as shown on the records of the Corporation. (B) Any meeting of the Members may be adjourned to any other time and place by the vote of a majority of those Members present or represented at the meeting, whether or not such Members constitute a quorum. It shall not be necessary to notify any Member of any adjournment. VII. The Annual Meeting of the Trustees shall be held promptly after the Annual Meeting of the Corporation at the Laboratory in Woods Hole, Massachusetts. Special meetings of the Trustees shall be called by the Chairman, the President, or by any seven Trustees, to be held at such time and place as may be designated. Notice of Trustees' meetings may be given orally, by telephone, telegraph or in writing; and notice given in time <<> enable the Trustees to attend, or in any case notice sent by mail or telegraph to a Trustee's usual or last known place of residence, at least one week before the meeting shall be sufficient. Notice of a meeting need not be given to any Trustee if a written waiver of notice, executed by him before or after the meeting is filed with the records of the meeting, or if he shall attend the meeting without pro- testing prior thereto or at its commencement the lack of notice to him. BYLAWS 39 VIII. (A) There shall be four groups of Trustees: (1) Trustees (the "Corporate Trustees") elected by the Members according to such procedures, not inconsistent with these Bylaws, as the Trustees shall have deter- mined. Except as provided below, such Trustees shall be divided into four classes of six, one class to be elected each year to serve for a term of four years. Such classes shall be designated by the year of expiration of their respective in (2) Trustees ("Board Trustees") elected by the Trustees then in office according to such procedures, not inconsistent with these Bylaws, as the Trustees shall have determined. Except as provided below, such Board Trustees shall be divided into four classes of three, one class to be elected each year to serve for a term of four years. Such classes shall be designated by the year of expiration of their respective terms. It is contemplated that, unless otherwise determined by the Trustees for good reason, Board Trustees shall be individuals who have not been considered for election as Corporate Trustees. (3) Trustees ex officio, who shall be the Chairman, the President, the Director, the Treasurer, and the Clerk. (4) Trustees emeriti who shall include any Member who has attained the age of seventy years (or the age of sixty five and has retired from his home institution) and who has served a full elected term as a regular Trustee, provided he signifies his wish to serve the Laboratory in that capacity. Any Trustee who qualifies for emeritus status shall continue to serve as a regular Trustee until the next Annual Meeting whereupon his office as regular Trustee shall become vacant and be filled by election by the Members or by the Board, as the case may be. The Trustees ex officio and emeriti shall have all the rights of the Trustees, except that Trustees emeriti shall not have the right to vote. (B) The aggregate number of Corporate Trustees and Board Trustees elected in any year (excluding Trustees elected to fill vacancies which do not result from expira- tion of a term) shall not exceed nine. The number of Board Trustees so elected shall not exceed three and unless otherwise determined by vote of the Trustees, the number of Corporate Trustees so elected shall not exceed six. (C) The Trustees and Officers shall hold their respective offices until their suc- cessors are chosen in their stead. (D) Any Trustee may be removed from office at any time with or without cause, by vote of a majority of the Members entitled to vote in the election of Trustees; or for cause, by vote of two-thirds of the Trustees then in office. A Trustee may be removed for cause only if notice of such action shall have been given to all of the Trus- tees or Members entitled to vote, as the case may be, prior to the meeting at which such action is to be taken and if the Trustee so to be removed shall have been given reasonable notice and opportunity to be heard before the body proposing to remove him. (E) Any vacancy in the number of Corporate Trustees, however, arising may be filled by the Trustees then in office unless and until filled by the Members at the next Annual Meeting. Any vacancy in the number of Board Trustees may be filled by the Trustees. (F) A Corporate Trustee or a Board Trustee who has served an initial tern: of at least 2 years duration shall be eligible for re-election to a second term, but shall be ineligible for re-election to any subsequent term until two years have elapsed after he last served as a Trustee. IX. (A) The Trustees shall have the control and management of the affairs of the Corporation. They shall elect a Chairman of the Board of Trustees who shall be elected annually and shall serve until his successor is selected and qualified and who shall also preside at meetings of the Corporation. They shall elect a President of the Corpora- tion who shall also be the Vice Chairman of the Board of Trustees and Vice Chairman of meetings of the Corporation, and who shall be elected annually and shall serve until his successor is selected and qualified. They shall annually elect a Treasurer who shall serve until his successor is selected and qualified. They shall elect a Clerk (a resident 40 MARINE BIOLOGICAL LABORATORY of Massachusetts) who shall serve for a term of 4 years. Eligibility for re-election shall be in accordance with the content of Article VIII (F) as applied to Corporate or Board Trustees. They shall elect Board Trustees as described in Article VI 1 1 (B). They shall appoint a Director of the Laboratory for a term not to exceed five years, provided the term shall not exceed one year if the candidate has attained the age of 65 years prior to the date of the appointment. They may choose such other officers and agents as they may think best. They may fix the compensation and define the duties of all the officers and agents of the Corporation and may remove them at any time. They may fill vacancies occurring in any of the offices. The Board of Trustees shall have the power to choose an Executive Committee from their own number as provided in Article X, and to delegate to such Committee such of their own powers as they may deem ex- pedient in addition to those powers conferred by Article X. They shall from time to time elect Members to the Corporation upon such terms and conditions as they shall have determined, not inconsistent with law or these Bylaws. (B) The Board of Trustees shall also have the powrer, by vote of a majority of the Trustees then in Office, to elect an Investment Committee and any other committee and, by like vote, to delegate thereto some or all of their powers except those which by law, the Articles of Organization or these Bylaws they are prohibited from delegating. The members of any such committee shall have such tenure and duties as the Trustees shall determine; provided that the Investment Committee, which shall oversee the management of the Corporation's endowment funds and marketable securities, shall include the Chairman of the Board of Trustees, the Treasurer of the Corporation, and the Chairman of the Corporation's Budget Committee, as ex officio members, together with such Trustees as may be required for not less than two-thirds of the Investment Committee to consist of Trustees. Except as otherwise provided by these Bylaws or determined by the Trustees, any such committee may make rules for the conduct of its business; but, unless otherwise provided by the Trustees or in such rules, its business shall be conducted as nearly as possible in the same manner as is provided by these Bylaws for the Trustees. X. (A) The Executive Committee is hereby designated to consist of not more than ten members, including the ex officio Members (Chairman of the Board of Trus- tees, President, Director and Treasurer) ; and six additional Trustees, two of whom shall be elected by the Board of Trustees each year, to serve for a three-year term. (B) The Chairman of the Board of Trustees shall act as Chairman of the Executive Committee, and the President as V7ice Chairman. A majority of the members of the Executive Committee shall constitute a quorum and the affirmative vote of a majority of those voting at any meeting at which a quorum is present shall constitute action on behalf of the Executive Committee. The Executive Committee shall meet at such times and places and upon such notice and appoint such sub-committees as the Com- mittee shall determine. (C) The Executive Committee shall have and may exercise all the powers of the Board during the intervals between meetings of the Board of Trustees except those powers specifically withheld from time to time by vote of the Board or by law. The Executive Committee may also appoint such committees, including persons who are not Trustees, as it may from time to time approve to make recommendations with respect to matters to be acted upon by the Executive Committee or the Board of Trustees. (D) The Executive Committee shall keep appropriate minutes of its meetings and its action shall be reported to the Board of Trustees. (F.) The elected Members of the Executive Committee shall constitute as a standing "Committee for the Nomination of Officers," responsible for making nominations, at each Annual Meeting of the Corporation, and of the Board of Trustees, for candidates to fill each office as the respective terms of office expire (Chairman of the Board, Presi- dent, Director, Treasurer, and Clerk). XI. A majority of the Trustees, the Executive Committee, or any other committee elected by the Trustees shall constitute a quorum ; and a lesser number than a quorum BYLAWS 41 may adjourn any meeting from time to time without further notice. At any meeting of the Trustees, the Executive Committee, or any other committee elected by the Trustees, the vote of a majority of those present, or such different vote as may be specified by law, the Articles of Organization or these Bylaws, shall be sufficient to take any action. XII. Any action required or permitted to be taken at any meeting of the Trustees, the Executive Committee or any other committee elected by the Trustees as referred to under Article IX may be taken without a meeting if all of the Trustees or members of such committee, as the case may be, consent to the action in writing and such written consents are filed with the records of meetings. The Trustees or members of the Ex- ecutive Committee or any other committee appointed by the Trustees may also parti- cipate in meeting by means of conference telephone, or otherwise take action in such a manner as may from time to time be permitted by law. XIII. The consent of every Trustee shall be necessary to dissolution of the Marine Biological Laboratory. In case of dissolution, the property shall be disposed of in such manner and upon such terms as shall be determined by the affirmative vote of two-thirds of the Board of Trustees then in office. XIV. These Bylaws may be amended by the affirmative vote of the Members at any meeting, provided that notice of the substance of the proposed amendment is stated in the notice of such meeting. As authorized by the Articles of Organization, the Trustees, by a majority of their number then in office, may also make, amend, or repeal these Bylaws, in whole or in part, except with respect to (a) the provisions of these Bylaws governing (i) the removal of Trustees and (ii) the amendment of these Bylaws and (b) any provisions of these Bylaws which by law, the Articles of Organiza- tion or these Bylaws, requires action by the Members. No later than the time of giving notice of the meeting of Members next following the making, amending or repealing by the Trustees of any Byla\v, notice thereof stating the substance of such change shall be given to all Corporation Members entitled to vote on amending the Bylaws. Any Bylaw adopted by the Trustees may be amended or repealed by the Members entitled to vote on amending the Bylaws. XV. The account of the Treasurer shall be audited annually by a certified public accountant. XVI. The Corporation will indemnify every person who is or was a trustee, officer or employee of the Corporation or a person who provides services without compensa- tion to an Employee Benefit Plan maintained by the Corporation, for any liability (including reasonable costs of defense and settlement) arising by reason of any act or omission affecting an Employee Benefit Plan maintained by the Corporation or affect- ing the participants or beneficiaries of such Plan, including without limitation any damages, civil penalty or excise tax imposed pursuant to the Employee Retirement Income Security Act of 1974; provided, (1) that the Act or omission shall have occurred in the course of the person's service as trustee or officer of the Corporation or within the scope of the employment of an employee of the Corporation or in connection with a service provided without compensation to an Employee Benefit Plan maintained by the Corporation, (2) that the Act or omission be in good faith as determined by the Corporation (whose determination made in good faith and not arbitrarily or capriciously shall be conclusive), and (3) that the Corporation's obligation hereunder shall be offset to the extent of any otherwise applicable insurance coverage, under a policy maintained by the Corporation or any other person, or other source of indemnification. 42 MARINE BIOLOGICAL LABORATORY VI. REPORT OF THE DIRECTOR "...There is no difficulty in forming government. It is not even a matter of choice whether there shall be one or not. Like breathing, it is not permitted to de- pend on our volition. Necessity will force it on all communities in some one form or another. Very different is the case as to constitution. Instead of a matter of necessity, it is one of the most difficult tasks imposed on man to form a constitution worthy of the name, while to form a perfect one — one that would completely counteract the tendency of government to oppression and abuse and hold it strictly to the great ends for which it is ordained — has thus far exceeded human wisdom, and possibly ever will." -John C. Calhoun (1782-1850) A Disquisition on Government I have chosen this rubric, the product of an exceptional, though pessimistic mind among this nation's founding fathers, because I believe, as will emerge, that the most difficult task facing the Laboratory and those entrusted with its management is governance: its discussion, and, as may prove necessary, its modification. Not long ago, survival itself was at issue. In those circumstances, the front line is always a battle-line. Economic survival, and the compromises thought by some to be necessary in aid of it, were the issues. The weapons of battle were budget- making, belt-tightening, a cold eye cast on proposals for spending: draconic measures to insure recovery, at least in part, of the costs of doing and teaching science. In that atmosphere, governance was, quite properly, seen as a merely peacetime concern, properly to be held in abeyance until the issue of battle became clear. Remarkably, the issue has, in a short time, become clear. We have won the battle and are in the process of mopping up. The M.B.L. will survive, and it will continue into its second century as a beacon of biology. This is not to say that economic threat, to which so many sister institutions have succumbed, has vanished ; rather, we have learned how to live with it, and there is no reason to believe that we shall need, in the forseeable future, to spend principal merely to stay at work. Our children will be able to come to the M.B.L. to partake, as we did, of its unique opportunity: the opportunity for those with imagination, energy, and self-discipline, independent of background or institution of origin, to work in close contact with, and to become, leaders of biology, pure and applied. The tests we face for the immediate future are indeed tests, rather than battles. We will be tested in the responsibility to participate broadly and unselfishly, within the Corporation and the M.B.L. family generally, in the raising of what are for this place large sums of money — not for survival, but in order to insure that the full promise of this institution is realized. And we shall need to be flexible and wise about pro- cedures and governance during the half-decade of transition ahead, in a period when our aging campus is returned to a condition matched to the quality of its scientific product, and during a time in which we attempt to endow the campus and its science financially, in such a way that survival will thereafter be assured. The choice of what to investigate, and what to teach, will thus be ours, not that of transient spon- sors, nor of agencies of government, however benevolent. PROGRAMS 1. Summer Research. In the summer of 1979 the M.B.L. was fully occupied. Every square foot of usable research space was rented. Upon the basis of reports made at the General Scientific Meetings, and of results I have myself heard about in other contexts, they were occupied with an unusual productivity. By virtue of an exceptional staff job of fitting people and programs into the available laboratories, and by a considerable effort of persuasion to secure agreements for sharing, it was REPORT OF THE DIRECTOR 43 possible to accommodate all principal investigators who offered programs acceptable to our Research Space Committee, and to accommodate further the bulk of students and assistants for whom they sought places. The summer of 1979 must therefore be counted as a very special one, even by the high standards of the M.B.L. In a time of increasingly restricted funding of individual research projects, and of research costs rising with awesome speed, there was sufficient recognition among productive investigators of the value of research at the M.B.L., and sufficient recognition of that value among their study section peers, to bring about full occupancy; to enlist some thousand people in the creation of the special scientific atmosphere that is a Woods Hole summer; and to limit dis- appointments and troubles to a gratifyingly small number. The capstone of the research summer was the Friday Evening Lecture series, opened with an elegant presentation by Edward Wilson, and highlighted by a pro- vocative discussion of biological "self" offered by the man who, better than any other, has chronicled the meaning of those lectures: Lewis Thomas. The session-closing General Scientific Meetings, revivified by the efforts of Hans Laufer a year earlier, fulfilled the promise of such a summer, under the direction of Joel Rosenbaum and his able associates. Rather than make the Quixotic attempt to choose, among the nine or ten signal accomplishments of that summer, one or two for presentation in this report, I sub- stitute an anecdote: In the early spring of 1980, sitting on a panel charged with the judgment and ranking of research proposals in the fields of my own work — cellular and developmental biology — I came across no fewer than five citations of results obtained in Woods Hole during the preceding summer as critical to the proposal substance. All those proposals were approved. If ever there wras a case of success breeding headaches, this is one. The indica- tions for the summer of 1980 are more, and perhaps troublesomely more, of the same. At the time of writing the available space is not only fully claimed, but in a few cases doubly claimed. We have reached the practical limit of civilized space-sharing, and we have had an ominously small number of cancell .tions, by the standard of earlier years. Clearly, an exciting and densely-populated summer is ahead of us in 1980. The new Parasitism course will be given for the first time; we can expect a distinguished addition, thereby, to the scientific population of Woods Hole. Rather than dimin- ishing, the popularity and significance of the squid and its giant axon seem to be increasing. John Yalois and his colleagues, quietly competent as they are, will have their work cut out for them, and the fishermen will come close inshore to compete. The question is: Can we, the Corporation, the Trustees, and the Administration, deal well and in a principled way with a situation foreign to our recent history (but quite familiar to our predecessors here) : the need to turn away qualified applicants? We shall see. This resolves into one of those issues of governance for which I chose the rubric of this report. Once again, the Friday Evening Lectures bid fair to be the intellectual and even the social highlight of the coming season. The speakers scheduled for it are of a level of recognition and achievement that can be matched by few institutions. The only mitigation I can think of, in the face of such disquieting and sequential eminence, is that every one of these speakers will, at some time during his stay in Woods Hole, need to answer some searching scientific questions. 2. Year-round research. The size of our year-round research program, as measured by numbers of persons engaged or the number of dollars spent on it, has again in- creased somewhat in the interval since the last Director's Report was written. In fiscal terms, this increase is a little faster, but marginally so, than inflation. In human terms, it is a good deal faster. The indications are, however, that by the fall of 1980 a quasi-equilibrium will have been established: retirements and space-constrictions (to the Library Readership that is the Nirvana of all devout M.B.L. scientists) will match nicely the establishment of new year-round laboratories. The undertaking 44 MARINE BIOLOGICAL LABORATORY given to the Trustees in February of 1979 is to be upheld, i.e., that consolidation, improvement, and growth of the year-round science at M.B.L. will not infringe significantly, over any stretch of a few years, upon the Laboratory's capacity to host its traditional transient programs. The M.B.L. Ecosystems Center maintains its strength and a determined inde- pendence of its views. They and we look forward, mainly with happy anticipation, but in a minor way with intimations of hard work ahead, to the construction of our new Environmental Sciences Center, toward whose cost of approximately $1.2 million we received last year a grant of $0.9 million from the Fleischmann Foundation. The new facility will be a boon to all Woods Hole environmentalists. The Laboratory of Biophysics, an intramural program of the National Institutes of Health here "on location," continues to function in a productive way and in a broad range of subdisciplines in neuroscience. The first of a new series of contracts between the old H.E.W. (now Department of Health and Human Resources) and the M.B.L., for the operation of this program, has been executed, and there is every reason to believe that the next five years will be as productive as the last, and — one hopes — somewhat easier to manage. Dr. MacNichol's Sensory Physiology program continues to produce significant and technically admirable work, and to be supported for it, despite certain internecine conflicts within the cognizant advisory bodies on the question of the relevance of research on invertebrates. A small but well-deserved expansion of their space and facilities will take place shortly. Cell Biology, that quintessential M.B.L. subject, and so ably represented on the campus in prior years, has had an upward shift coincident with the arrival here of Shinya Inoue as a year round investigator, and has been aided further by the estab- lishment of Sidney Tamm as a faculty member of the B.U.M.P. program. Although Dr. Tamm has devised a number of wickedly effective methods for badgering the Director on matters of overhead, it is already clear that his work on cellular motility will flourish in Woods Hole. The only year-round program, in fact, that has not shown growth consistent with last year's expectations is Developmental Biology, and this is directly traceable to the style of the Director's life, since he is a developmentalist (in principle), and was supposed to devote some reasonable part of his time to program-building in that important subject. To be sure, the Director's own scientific work has continued, thanks to generous arrangements made by the University of Rochester, and to the quality of his Rochester colleagues and students. It is however clear that some changes will have to be made in the Director's list of responsibilities so that he can establish and help to manage a new and major year-round research program. 3. Education. The Laboratory's educational programs are in good health, due, in part, to their inherent excellence, and in part to the nonthreatening, but conscious control and overview processes that have been instituted during the two years past. The five January courses enrolled a total of 108 students, compared with 114 in 1978. This is a nonsignificant variation. The only systematic change suggested by a study of the January semester as a whole is that the quality of applicants rose detectably in 1979. Alan Fein, who had earlier served with Fred Lang as co-Instructor- in-Chief of the Neurobiology Course, took full responsibility for it less than three weeks before the start, following Fred Lang's tragic death in an automobile accident. Ted MacNichol, among others, stepped into the gap with courage and selflessness. The 1979 course was a great success, exceeded, perhaps, only by the success of the 1980 version, which had the benefit of a dedicated supervision by Daniel Alkon, one of the leaders of the M.B.L. year-round neuroscience program. Among the short courses, the design and supervision of which are ably handled by Morton Maser, there was a high esprit, the best testament to which is the body of letters received from participants. A number of new courses were mounted in 1979, including Comparative Light and Electron Microscopy in Clinical Diagnosis (Harry Carter in charge), and Mariculture — the Culture of Marine Invertebrates for REPORT OF THE DIRECTOR 45 Research Purposes (Carl Berg in charge). In 1979, nine short courses enrolled some 125 students, which approximates 87% of the planned capacity. The summer teaching session, populated in 1979 by 142 students (compared with 148 in 1978), was an exciting one. Tom Humphreys and Joan Ruderman ended their tours of duty as co-directors of the Embryology Course, which had, if possible, an even more stellar student body than the year before; and Rudolf Raff, of the Uni- versity of Indiana, took the initial steps in assuming directorship of the course for the term to come. Six important and productive years of Jerome SchifT's directorship of the Botany- course ended with this year's offering, and with an announced hiatus of one year in the offering of a marine botany course, during which a national committee of advisors will make proposals for the design of its successor. Dr. Lawrence Bogorad, the eminent Harvard botanist, will chair the committee of review, and will be in resi- dence during the summer of 1980. A core group of the 1979 botany faculty will direct a specialized research and training program, under the rules for independent investigatorship, in the summer of 1980. The Microbial Ecology course, directed by Harlyn Halvorson and Holger Jannasch, will be offered in an expanded version in 1980, residing temporarily, for that session, in the Loeb quarters normally used by Botany. A term of outstanding service to the M.B.L.'s educational venture ended in 1979 with completion of Edward Kravitz's tour as instructor-in-chief of Neurobiology. He has been succeeded by John Hildebrand and Tom Reese. The identity and scien- tific reputations of those two insure that Dr. Kravitz's achievement will have the desired continuity. Ronald Hoy accepted (generously) responsibility for the Neural Systems and Behavior course following a tragic death in the Gelperin family, the result of which was that the course's energetic founder, Alan Gelperin of Princeton, was unable to continue. Ronald Hoy's tenure will, undoubtedly, substitute for the Princetonian a certain Cornellian flavor to the enterprise, but that — however one judges Cornell as against Princeton — has already had the good effect of sharpening the substantive focus of the course, and of attracting to its important activities, at the Interface be- tween cellular neurobiology and behavioral science, a helpful amount of financial support. The Boston University Marine Program, under the directorship of Arthur Humes, solidified its important position as an element of the Woods Hole educational spectrum. The tenth anniversary of the program was celebrated this year at an occasion of a very positive character, held in November at our Swope Center. Two new faculty have joined the program: Drs. Sidney Tamm and Christopher Price. C. K. Govind, of the University of Toronto, was a visiting senior faculty member during the second half of the year. There are now 34 B.U.M.P. graduate students in residence, about two-thirds of them candidates for the Ph.D. degree, supervised by five regular faculty who have the assistance of a much larger number of colleagues. The level of B.U.M.P. activity in gestating and delivering Ph.D.'s, (in such fields as ecology, ecological genetics, physiology, and neurobiology) remains high — three the past year, and the production of quality research papers and of fruitful scientific interchanges, as, for example, in seminars, is equalled, and certainly not exceeded, by a very few other educational programs in marine biology. Planning began, in 1979, for broadening in two directions of the activities now centered about the unique late-spring course in Aquatic Veterinary Medicine ("Aquavet"), which is coordinated by Donald Abt, supported in many ways by the parent schools of veterinary medicine (Pennsylvania and Cornell), and taught by a faculty group of multi-institutional origin. One direction will be research-centered, providing systematic opportunity for veterinarians in training and for other young biologists to carry out relevant research in M.B.L. laboratories. The other, an ini- tiative that will take longer to be realized, but that will provide an important element of planning for the ultimate marine resources center and facilities, is the provision 46 MARINE BIOLOGICAL LABORATORY of diagnostic and other clinical services, under direction of qualified veterinarians, for the collection, maintenance, and use of marine animals. 4. Scientific Meetings. The skill with which Homer Smith and his staff manage the jigsaw puzzle of accommodations and facilities for scientific gatherings of various kinds received comment in the last report. Suffice it to say that in the past year— an even busier one than the preceding — there was no evidence of a decline in skill, nor of any likely reduction in the growth rate of M.B.L. as a desired location for serious scientific colloquy, especially in the quieter seasons. The whole effort can be exemplified by the efficiency with which a threatening overload of participants in the symposium held in honor of Jim Ebert (September 11-13) was handled. Some who wanted to stay here during that time could not, but there were no angry nor wounded confrontations. Almost everyone seeking to participate was accommodated in one way or another (and as far away as the Plaza end of Falmouth), and, as the audience grew with each passing day from 200 to Standing Room Only in the Lillie Auditorium, everyone was fed, and had the op- portunity to ask questions and to be heard. It is not surprising that the enthusiasm giving the event its ambience was reflected in major reports of it in the scientific press. The work of our Public Relations Office contributed importantly to the sym- posium's success. In addition to the Ebert Symposium, the following (among many more not men- tioned) were scientific gatherings held this year at the M.B.L. : The Society of General Physiologists, the Molecular Biophysics and Biochemistry Conference of Yale, a meeting on International Environmental Problems sponsored by the National Academy of Sciences and arranged by George Wood well, a Harvard Medical School conference on Immunology, the Northeast Regional meeting of the Animal Behavior Society, a meeting of the National Foundation for Cancer Research ; and a visit of Massachusetts Senator Paul Tsongas with leading Woods Hole environmental scientists. While it is not true that these influxes of our slide-and-manuscript-bearing colleagues from the big cities disturb the peace, the beauty, and the historic som- nolence of wintertime in Woods Hole, it is true that the possibility of escape, in this place, into an absence of thought, has come to be desperately small. FUND-RAISING AND COST RECOVERY Near the end of 1979, the Chairman of our Board of Trustees, Prosser Gifford, announced the Laboratory's new, very ambitious, and first fully-organized, long term fund-raising campaign, which has been named, for obvious reasons, the M.B.L. Second Century Fund. The goals and the strategy of that effort are too complex and too important to risk its misunderstanding by a mere summary in this limited space (over which the new Biological Bulletin editor, Charles B. Metz, will exert just as jealous a control as did his predecessor, "Gus" Russell-Hunter). Suffice it to say that in this year the M.B.L. mounted its first comprehensively planned capital campaign, and that as of the time of writing the totals are well ahead of the scheduled target. A full dis- cussion of the Second Century Fund will be made available to the Corporation before the end of the 1980 summer session. The cold numbers of the outcome are not without interest, however, and those I will communicate: In the year ending December, 1979, the M.B.L. actually received $1,215,144 in gifts, and those receipts were underpinned by firm pledges of an ad- ditional $1,985,500. The provenance of these gifts is as varied as the specific (or unrestricted) purposes for which they were given, but the happy implication of the totals will be evident to anyone who examines our financial statements for prior years. Details are available elsewhere in this Report. Considerable attention has been given in the past year to the much-vexed question of overhead-cost recovery for research and educational programs. Vexed it is; more so than in colleges, universities, and full-time research laboratories, because of the peculiar mixture of activities carried on at the M.B.L. Of course, few non-profit REPORT OF THE DIRECTOR 47 institutions recover anything near the actual cost of mounting externally-sponsored research programs. The allowable charges for overhead are watched penuriously and renegotiated regularly by several agencies of government, all of which are re- warded for their effectiveness in saving money (which is as it should be). The M.B.L. recovers, unfortunately, a much smaller than average share of its costs. Among the reasons for this are (1) the space-rental method employed, which has built into it a required, but really inapplicable factor of "unused capacity" and a penalty based on it, (2) the danger that even if the rental rates were to be allowed to rise realis- tically, they would be too high for individual transient investigators to pay, (3) the fact that such payments are normally made as direct costs to research grants, while the parent institution recovers full, or nearly full indirect costs nevertheless, which is a form of double jeopardy, and (4) the fact that auditors and other contract officers of the granting agencies are poorly informed about the nature, scale, and value of M.B.L. operations. They have been inclined to deal with it as though it were a hos- pital or a kind of part-time college. Among the most troublesome consequences of the system and its inherent contradictions has been unpredictable change in rates from one year to the next, causing individual investigators much difficulty in the payment of bills. Several good results have been obtained in the past year, in consequence of ini- tiatives aimed at correcting the situation. The Directors of NIH and XSF have given strong and informed support to our brief to the auditors, seeking more realistic interim arrangements. The initiatives and that support have yielded the hoped-for relief in the form of an offical agreement on rental rates, covering a three-year period and taking account of inflation, and a reduction in the penalty for so-called "unused capacity." Negotiations with the financial officers of both our major granting agencies have, furthermore, brought about a significant broadening — and a mutual one — of understanding, particularly as regards the elements of unfairness that remain features of a system that treats year-round, in-house research grants in the same way as the funds supporting transients. All parties to these negotiations have agreed that a new, simpler, fairer system of overhead recovery for the M.B.L. is to be worked out and presented for preliminary review before the end of the 1980 calendar year. The intention is to bring it into operation by the summer of 1981. Simultaneously, we are exploring, with coopera- tion of the agencies, forms of general support direct to the M.B.L., for the transient research and the educational programs of this Laboratory (and of the other LI. S. marine laboratories), with a view to minimizing the direct cost burden on the indi- vidual transient research worker. There is every reason to believe that the cost in time and effort now being paid by us for these initiatives will yield a very much larger return in the future. Our goals for that future are very simple: They are to recover the costs (and no more than the costs) of research and teaching done at the M.B.L. (assuming that the quality thereof will not decline from the present), and to make program quality, rather than the ability to pay rent, the essential determiner of participation. MANAGEMENT Biological Bulletin. Editorship of our journal has passed from the capable hands of \Y. D. Russell-Hunter to those of C. B. Metz. Russell-Hunter's was a distinguished tenure. He brought to the journal orderliness and efficiency that are rare among scientific journals in these times, and was able at the same time to impose upon it, without tyranny, something of the literacy and attention to detail that characterize his own writing. The Corporation owes him a debt of gratitude. Charles Metz, in taking up the position's challenge, brings to it a broad experience in developmental and reproductive biology and in the management of educational enterprises. He intends to broaden the substantive range of Biological Bulletin articles, making the journal more representative of scientific interests of the Laboratory and 48 MARINE BIOLOGICAL LABORATORY its current programs. This, however, he hopes to accomplish without diminution of editorial rigor, nor a decline of the high standard of style set by his predecessor. He will be aided by several new and accomplished members of the editorial board. We wish him and his associates good luck, and set down here an undertaking already made viva voce by Corporation members who were consulted in the editorial succession : to help make the Bulletin a more representative vehicle, and a primary vehicle, for important papers in fields pursued at the M.B.L., while at the same time adhering to the policies that have made it an international journal of wide distribution, rather than a mere house publication. The Board Chairman and The Treasurer. Drs. Prosser Gifford and Robert Mainer are very busy men. Each holds a demanding position and each gives time and interest to a broad range of educational and community enterprises. It was therefore with some concern that we began to seek, during 1979, a distinctly larger commitment of their time to M.B.L. management problems and to fund raising. They have responded magnanimously, and I cannot let the opportunity pass to thank them publicly, for the Trustees and for the Corporation as a whole. As must be evident, even from the skeletal indications of this Report, our management task has multiplied in difficulty and complexity. There is little of Parkinson's Effect (i.e., the expansion of management to fill all available space and budgets) in this. By the applicable standards, M.B.L.'s management team is very small for the size and complexity of the enterprise. This has been true in the past, and it is more strikingly true now. Yet the place has changed, as have the times and the requirements for getting sup- port and employing people. A large part of our management team is volunteer: the time-consuming and difficult work of the Executive Committee, for example, is done without compensation for its members. The Board Chairman and the Treasurer, who are two of the members, have a particularly heavy burden of work. \Ye have added to it by broadening the participation of Trustees in such important decisions as fund-raising initiatives, capital equipment and resources acquisitions, property management, and salary-setting. Drs. Gifford and Mainer have accepted these new responsibilities with good cheer and met them with skill. There is no way that we can pay them for the valuable time they have given, and will, we hope, continue to give, except by our honest gratitude. New Positions. The Controller and the General Manager will shortly have the assistance of a skilled personnel officer, who will relieve them of the responsibilities they now bear for personnel management, while at the same time collaborating with them and with the Director in relevant new policy-making. This is a step long in review and long needed, particularly because however decent and familial our em- ployer-employee relations have been (and they have been unusually so), the law now requires attention to record-keeping, to reporting, and to pension matters at a level of detail that cannot be managed by any one person part time. In addition, the search has begun for a senior officer, at the level of Assistant Director, who will have broad responsibility in general administration and in the Laboratory's Planning and Development activities. This is a position whose creation is indispensable for a successful outcome of long-term plans for campus rehabilitation and for enlarging the endowment. The Laboratory has received a significant grant that will underwrite the costs of this office for the first few years, during which time its performance and contribution will be observed with care. Committees. In that context the Executive Committee and the administration have begun to plan for a regular and orderly process of review, in which administrators and their positions will be subject to a study of performance, in quite the same way as staff scientists and support staff (and professors in universities) have their per- formances reviewed. It is our hope in the coming year to enlist a much closer involvement of all Trustees, and of Corporation members generally, in such processes. A deliberate and thorough review of membership rotation for the standing committees was begun early in 1979. It continues to make visible and useful progress. The work of the Library Com- REPORT OF THE TREASURER 49 mittee is particularly worthy of mention in this regard. Under the Chairmanship of Edward Adelberg, this broadly-based committee has made a study of a perennial trouble-subject: the management of our superb Library. The final report to the Corporation is due at the end of the 1980 summer session, but interim reports, com- bined with a new and heartening spirit of cooperation among the senior administrators of this Laboratory and of the Woods Hole Oceanographic Institution, have already produced important agreements. Among them is a fairer distribution of effort and budgetary support for the Library than ever before. It is now clear that the final Library proposals will entail no threatening discon- tinuities of style or services, but that they will at the same time result in a signal improvement of the Library's holdings, specialized services, and capability for at- tracting external support. CONCLUSION I return, finally, to the broader issues of constitution and government with which this report began. We have made every effort to control the changes taking place in the M.B.L.'s size, range of programs, and facilities, in such a way as to minimize (although, unfortunately, not to eliminate) dislocations and inconvenience. The directions of change have prior Corporation agreement: rehabilitation of the campus, so as to fit it for the kinds of research and teaching now done here (and to be done here in the future) ; the addition of certain new facilities, such as the much-needed marine resources center, the environmental sciences center, and a carefully planned expansion of housing; consolidation and strengthening of year-round programs within the context of the M.B.L. program as a whole, and without long-term infringement of the transient programs; reorganized management of endowment and properties. What has not, and necessarily not, been argued in nearly so much detail is the ma- chinery of governance by which these changes are to be brought about. John Calhoun's wry but perfectly accurate assessment of the central issue applies: government (or management) there must be, but the rules for it, or the constitution, must have an inherent flexibility sufficient to deal with change and to prevent tyranny. Even so the rules must be changed from time to time. This is hardly the place for a listing of the issues of governance about which we shall have to confer, since a report is, after all, concerned with past and present, rather than with the future. An indication of the nature of those issues would not, howrever, be in- appropriate. We need a reexamination of the role played by our Trustees, and particu- larly by the Board Trustees. We need some new understandings, if not more rules, about administration generally: how big it needs to be and ought to be, for this unique institution, and how its quality, whatever the agreed size, is to be maintained. We need understandings as to how the Laboratory's scientific programs are to be kept under some reasonable central control, as is demanded increasingly by granting agencies, while at the same time insisting that administration exists to serve the programs, rather than the reverse. The changes required for the M.B.L.'s survival in strength into its second century are large. Let us hope that an equally large recommitment to watchfulness, to participation, and to good will can be made this year by the Corporation member- ship and, indeed, by all friends of the Laboratory, as the countdown to our second century begins. VII. TREASURER'S REPORT The Laboratory achieved its 1979 financial objective of maintaining parity between revenues and expenses. In fact, a modest surplus of $31,000 was re- corded— a gratifying improvement over the operating deficits of 1978 and 1977. However, this result does not include provision for depreciation on the Labora- 50 MARINE BIOLOGICAL LABORATORY tory's physical plant and equipment, about which I shall have more to say in my concluding comments. Aggregate revenues of $4,766,000 in 1979 were $509,000 greater than in the prior year. Among the more important contributors to the increase are the following: 8140,000 in additional laboratory fees and overhead recoveries. A signif- icant factor here was a 13 percent increase in the approved overhead rate for federal grants and contracts ($30.50 in 1978 to $34.60 in 1979). $74,000 in additional total investment income, partly resulting from higher rates on short-term investments and partly from improved returns on endowment and quasi-endowment funds which, since February, 1979, have been under full-time professional supervision. —$443,000 in additional federal funding of direct costs of research performed under grants and contracts. Offsetting a portion of the above increases was a decline in unrestricted gifts- Si 15, 000 in 1979 versus $241,000 in 1978. However, gift income tends to be unpredictably variable. For example, 1978 saw the receipt of a gift of real estate and an unusually large foundation gift. As we build the momentum of the Laboratory's Second Century Fund campaign, we are likely to see even greater variability in unrestricted gifts, both because some longtime contributors will temporarily shift their attention to the Laboratory's capital needs and because the campaign may elicit a certain number of gifts from donors who prefer not to contribute to "bricks and mortar." Aggregate expenses increased $442,000 in 1979, reaching a total of $4,735,000. Much of this resulted from the flow-through of expenses associated with the higher level of federal contract research noted in the discussion of income. In- flation's shadow darkened virtually every category of costs in 1979 and glooms the coming months even more. I am impressed by the performance of the Laboratory's administration and staff in controlling expenses, often achieved by uncommon ingenuity in the use of limited resources. During 1979, the Woods Hole Oceanographic Institution and the Marine Biological Laboratory reached agreement on a formula for sharing joint library costs. As a result, we now have a more logical and predictable method for re- imbursement of the costs of acquiring and managing library resources for the joint benefit of WHOI and MBL. Also during 1979, the MBL negotiated a fixed three-year (1980-82) overhead rate with its government auditors. The three-year period, in contrast to the former year-to-year approvals, enables investigators to more accurately budget for multi-year research projects and permits the Laboratory to better estimate its income. The Laboratory's 1980 budget aims at an objective of $69,000 of revenues in excess of expenses, before provision for depreciation. Its achievement will be difficult. In addition to inflation's toll, the Laboratory will face the added expenses of the steps being taken to assure its future. In particular, monies must be spent now to organize and conduct the substantial fund raising effort to rehabilitate an aging physical plant, add facilities and increase endowment. This brings me to the point I must make about the Laboratory's long standing practice of not funding the depreciation of its physical assets. It is tempting to conclude that the Laboratory is doing well if its revenues cover cash outlays. Such a conclusion brings an illusory comfort. Depreciation, although a non-cash REPORT OF THE TREASURER 51 expense, reflects the loss in value of assets attributable to deterioration and obsole- scence. Moreover, in inflationary periods, stated depreciation underestimates the cost of replacing assets acquired or built during earlier times. The alternative to the funding of depreciation — assuming we are not willing to permit catabolic processes to overtake the Laboratory's viability — is to dedicate ourselves to periodic aggressive fund raising to generate the means by which the MBL's physical plant can be maintained, repaired, rehabilitated or, in some cases, replaced. It is for this reason that the recently announced Second Century campaign is vitally important and deserving of the help of everyone associated with the Laboratory. COOPERS & LY BRAND CERTIFIED PUBLIC ACCOUNTANTS A MEMBER FIRM OF COOPERS & LYBRAND (INTERNATIONAL) To the Trustees of Marine Biological Laboratory Woods Hole, Massachusetts We have examined the balance sheets of Marine Biological Laboratory as of December 31, 1979 and 1978, and the related state- ments of current funds revenues, expenditures, and other changes and changes in fund balances for the years then ended. Our examinations were made in accordance with generally accepted auditing standards and, accordingly, included confirmation from the custodians of securities owned at December 31, 1979 and 1978, and such tests of the accounting records and such other auditing procedures as we considered necessary in the circumstances. As more fully described in Note C to the financial state- ments, the Laboratory excludes certain costs of buildings and equipment from the balance sheet. In our opinion, generally accepted accounting principles require that such costs be included as invest- ment in plant in the financial statements. In our opinion, except for the effects on the financial statements of the matter discussed in the preceding paragraph, the aforementioned financial statements present fairly the financial position of Marine Biological Laboratory at December 31, 1979 and 1978, and its current funds revenues, expenditures and other changes and the changes in fund balances for the years then ended, in con- formity with generally accepted accounting principles applied on a consistent basis. Boston, Massachusetts March 21, 1980 52 MARINE BIOLOGICAL LABORATORY MARINE BIOLOGICAL LABORATORY BALANCE SHEETS December 31, 1979 and 1978 Assets 1979 1978 Current Funds: Unrestricted : Cash, including deposits at interest $ 424,763 $ 382,396 Money market securities 500,000 Accounts receivable, net of allowance for uncollectible accounts of $1,716 in 1979 and $4,500 in 1978 858,466 699,708 Other assets 11,186 15,367 Due to restricted current funds (729,275) (278,075) Due (to) from invested funds (8,151) 103,248 Total unrestricted 1,056,989 922,644 Restricted : Cash 12,285 18,973 Investments, at cost ; market value: 1979— $1,881,590; 1978— $1,297,551 (Notes B and F) 1,900,292 1,297,530 Due from unrestricted current fund 729,275 278,075 Due from invested funds 350,967 260,967 Total restricted 2,992,819 1,855,545 Total current funds $4,049,808 $2,778,189 Invested Funds: Cash 5,589 6,982 Investments, at cost; market value: 1979— $4,250,838; 1978— $3,950,184 (Notes B and F) 3,936,137 4,084,003 Due (to) from unrestricted current fund 8,151 (103,248) Due to restricted current funds (350,967) (260,967) Total invested funds $ 3,598,910 $ 3,726,770 Plant Fund: Land, buildings and equipment at cost (Note C) 12,258,651 12,421,929 Less accumulated depreciation 4,251,598 4,060,407 Total plant fund $ 8,007,053 $ 8,361,522 The accompanying notes are an integral part of the financial statements. REPORT OF THE TREASURER 53 MARINE BIOLOGICAL LABORATORY BALANCE SHEETS December M, 1979 and 1978 Liabilities and Fund Balances 1V7<> IV? '8 Current Funds: Unrestricted : Accounts payable and accrued expenses $ 268,369 $ 316,657 Deferred income 68,006 61,432 Fund balance 720,614 544,555 Total unrestricted 1,056,989 922,644 Restricted : Fund balances: Unexpended gifts and grants 2,897,679 1,777,677 Unexpended income of endowment funds 95,140 77,868 Total restricted 2,992,819 1,855,545 Total current funds $4,049,808 $2,778,189 Invested Funds: Endowment funds 1,918,170 2,174,027 Quasi-endowment funds 934,143 934,143 Retirement fund (Note D) 746,597 618,600 Total invested funds $3,598,910 $3,726,770 Plant Fund: Invested in plant 8,007,053 8,361,522 Total plant fund $8,007,053 $8,361,522 The accompanying notes are an integral part of the financial statements. 54 MARINE BIOLOGICAL LABORATORY en u ^^ ^^ ^'•O ON — -f Tf — -f ON" re~ o" CN CN 00 t^ OO ID CN O 00 ID OO NO CN C — 30 ^ , , CN ' — ' t~- i— < O >D ID £«•. CN (--* -i- f^] *— < CN --f «N o\ O cs O — t^ OO rf -f ~ OO OO ^^ 1 — r~ »o NO J^ ON i r^* | J^J" 0 00 ^ 1 »D OO OO t"^ ."^ l~-» ^^ ID o_ r*5 <*5 «— l »— 1 — rr; re CN ^" "*~ O t-- O^ O ^- o O ^« OO C-l OO oo ID" oo C ID OO o t-~" f^ oo vO 00 •<* -~, H O 2 X U X w X H O Q Z oo •o C O o ID O -H t^ CN .•^*> ^^ j^» O —i 00 >D t^ 00 ID ID re -H -H ID CN 00 O f~ 00 OO O 00 00 ID ID ^H OO oo o 00 «N ** CN" O CQ U C o J o 5 w z 5 X "W 1 u ON H 3 - z r*2 w K? OH x CD Cv W C ~ § '£ cyf OJ "^ D Q I W •o ^ OJ ON t^ W TD C 2; C/5 en Q cfl £ > t, OJ H -M ^ U W O 00 00 f*5 cs OO 00 I— >D O ID ^i O O O OO — ^f ID cs r^ c^i r^ t^ ID O CM O »— i o" o" o" f~" >D O O ON tN >D r^ 00 00 CN O NO •* OO r^ ID .-H ' ON" ON" O" <-•" t^ ID T}< ON ^H (N CN o ID _ 0 00 O ^ ON NO ^ t— ID O ,_, •* re t"** V'O ~ O *-• vH t^» V»H c~l 0- ON NO t"*» *^ ~ ID ON CN CN >D ON_ ID" re ON ON Ov O r- o" CN" vH *=J* ON" ^H NO 00 t- re 00 CN O\ t^* i — i re ** ID ro »— i ^H CN ^— 1 «— I -H O cs" as C/3 H H CO 2 8 3 C _0 •i~> u u. tt C C C cS c oj O> 03 2 '^ J Cu — 2 5 -g a; O •5 HO 03 en i- *J 0 C u 03 u S -JO 0 § ~ Cu — "B CQ o o e.s a 0° s.s^^ Q G _] CQ ^H \O ^^ PC *^" *D OO t^~ — -^ -^ PC ON PC 00" r~" oo •* -t 0 •—i — - | -]__'- i rC CM ID NO -t" * -^ PC OO •rF -^* ">! fN ON O O ~* t^ ID ID ON O O OO — ON-^^f S $ ^ o l~- C lt % O ^ ON 1-- NO 00 O t— NCt—_lD_ *~i rr"1^ e^t— _ ON . t^ t^ OO NO PC ON 00 •* O" NO PC ID O\ -^< \O PC •* NO 00 OO NO CN O O -^ PC NO ~-£ r-l ID r^ OO" 00 t— 1 00" 00 oo 00 t~- eM OO ON O "* CN oO t— ID oO O >— ' l— • PC -H ON O O PC ID_ § PN) r~- tN I—" t>^ fN ID O NO ID NO O) O\ PC t-~ PC PC 0) PC rc PC "5-^ p>f •^-H ^^ S PC Oo ON PC OO ON 0 ON -- T^ O PC 1^- ID 00 g •* o 0 ID PC ID '* -a ^ ^ 2) PC ID PC ON ID ^ PC •r-< r^ OO ID PC 00" CNl I'D PC" NO t--" PC ID PC .ts ^ OO 0^ o ON t-~ •^ PC o — < 00 o ON *— H o 0; NO 0> r*5 ON ON ID PC NO ON ID PC c>r "^ t— PC •«* ON 0) NO ID S ID" ID NO" C4 t^" NO O o" PC "•< CN '-< tS PC e$ ID NO NO_ rN PC in •O 0) a '•fi C 03 S 5 .S •^i a •ft. O c o CQ >,.a u b£ rt Q XI *O c (U 8. tu •o o a a d) « <-> O 85 G Q I < g I 0- Q. X 0) O in 0) c OJ o 4-1 'o rt &} in Is 'u c 03 C •y w C en QJ 4-> QJ CJ en 03 f^ 2 "^ rt •4-1 en en OJ •^ 4J O 4J QJ U **j f" "O C C O ^J _ri ^ "o3 t . — Ui C o. O- u- rt 03 •-+ ^ V(_( *O a -O CJ O "S -o _ i 03 •*-» — « S S -w »J S 1^ restrict en C C 3 C be QJ "c r^ p$ en c • ^- QJ C en" Qj jj 3 •a rt ri QJ 03 QJ ^= 0 0 be u •n "^ CL C T3 C 03 u CD 03 en QJ a) > *C en O tj 'S "O I QJ en c '5 g § en <" N*- •v c V4-I en c 03 rt a> ej bfi C ^ QJ -^^ 3 u C ^> £ 5 v» 4^ c o "a. c g 4-i QJ a. 03 'S, X E S "" E a. o o o o (j u u u. H 03 56 MARINE BIOLOGICAL LABORATORY a ~° s ~~. a (N « ~H 00 00 69 01 CN oc O OQ en U U £ 2 OQ Q g D oo ON i — i -a c rt ON a HI B -0 0) _N nvestments stment incomi ,o '55 C 0) a o 4-> C .0 c. ^o 01 Ol U c ii c O •-) ^ ^ c rt 6 "r5 0) aJ u > c •5 -o < ^-1 '5 H _ca 03 y CQ »f 2 % S, c OJ J= OJ a •a c 00 CJ g ^i 3 CO resear CO O) U 3 •M to 4V '55 c OJ D. O c 2- <0 a _ .2 'o LO — -f O NO" oo" t-~ r— -' CN O O lO O o 00" S S OO oo NO o CN ro" CM o -0 8 a S I OO O r~ oo" OO O t— lO CN" o" CN o •* CN o o LO CN o oo NO o" CN S OO o o \o o" CN c «u B C 0) !_ tn V ants and gifts vestment income oceeds of sale of equipment Idition to pension fund lition ses: struction, research and gene en 2 t ^ 8 £ " &? aj £ •5 •o cu "rt en C 0 en en tn _O cu C •o N "rt investments lyments to pensioners et book value of equipment epreciation fers — additions (deductions): roceeds of sale of equipment ces at December 31, 1979 p O nS £ < H 1 c t—H & Cu X O S 0- c rt ^ L. "rt cS £ CQ rt a "rt I C c rt U rt en O C b/0 .S C rt a o u u rt 58 MARINE BIOLOGICAL LABORATORY MARINE BIOLOGICAL LABORATORY NOTES TO FINANCIAL STATEMENTS A. Purpose of the Laboratory: The purpose of Marine Biological Laboratory (the "Laboratory") is to establish and maintain a laboratory or station for scientific study and investigations, and a school for instruction in biology and natural history. B. Significant Accounting Policies: Basis of Presentation — Fund Accounting In order to ensure observance of limitations and restrictions placed on the use of resources available to the Laboratory, the accounts of the Laboratory are maintained in accordance with the principles of "fund accounting." This is the procedure by which resources are classified into separate funds in accordance with activities or objectives specified. In the accompanying financial statements, funds that have similar characteristics have been combined. Externally restricted funds may only be utilized in accordance with the purposes established by the source of such funds. However, the Laboratory retains full control over the utiliza- tion of unrestricted funds. Restricted gifts, grants, and other restricted resources are accounted for in the appropriate restricted funds. Restricted current funds are reported as revenue when expended for current operating purposes. Unrestricted revenue is re- ported as revenue in the unrestricted current fund when received. Endowment funds are subject to restrictions requiring that the principal be invested and only the income utilized. Quasi-endowment funds have been established by the Laboratory for the same purposes as endowment funds, however, any portion of these funds may be expended. Certain accounts for the prior year have been reclassified to the current re- porting format. Investments Investments purchased by the Laboratory are carried at cost. Investments donated to the Laboratory are carried at fair market value at date received. For determination of gain or loss upon disposal, cost is determined based on the specific identification method. Investment Income and Distribution The Laboratory follows the accrual basis of accounting except that investment income is recorded on a cash basis. The difference between such basis and the accrual basis does not have a material effect on the determination of investment income earned on a year-to-year basis. Investment income includes income from the investments of specific funds and from the pooled investment account. Income from the pooled investment account is distributed to the participating funds on the basis of the market value at the beginning of the quarter, adjusted for the cost of any additions or disposals during the quarter. REPORT OF THE TREASURER 59 C. Land, Buildings and Equipment: Following is a summary of the plant fund assets: Classification 1979 1978 Land $ 639,693 $ 639,693 Buildings 10,190,430 10,190,430 Equipment 1,428,528 1,591,806 12,258,651 12,421,929 Less accumulated depreciation 4,251,598 4,060,407 $ 8,007,053 $ 8,361,522 The original cost of land, buildings and related initial furnishing equipment is capitalized when the assets are acquired. The cost of subsequent additions and purchases, repairs and remodeling is expensed when incurred. Equipment and remodeling expenditures amounted to approximately $76,000 and $125,000 in 1979 and 1978, respectively. Depreciation is computed using the straight-line method over estimated useful lives of 40 years for buildings and 20 years for equipment. D. Retirement Fund: The Laboratory has a noncontributory pension plan for substantially all full-time employees which complies with the requirements of the Employee Retirement Income Security Act of 1974. The actuarially determined pension expenses charged to operations in 1979 and 1978 were $81,892 and $65,673, respectively. The Laboratory's policy is to fund pension costs accrued. E. Pledges and Grants: As of December 31, 1979 and 1978, the following amounts remain to be received from pre- vious gifts and grants for specific research and instruction programs, and are expected to be received as follows : December 31, 1979 December 31, 1978 Unrestricted Restricted Unrestricted Restricted $34,666 $2,161,535 28,000 1979 1980 $ 50,500 $3,689,270 1981 49,000 215,000 1982 15,000 $114,500 $3,904,270 $34,666 $2,189,535 60 MARINE BIOLOGICAL LABORATORY 3J GO ON f^ *^f ^t" OO t^ — i E O ^> NO <^O tO -^H ON f^ ^^ i™"1 OO O CM r— oo ON o S .E § CX "^ r^i ^^ r*5 re *# i-t CT) i— i CM rsi ^ S ^O f& c ^ I 1 I ! o cO f~"> ^O ^^ O 't O C ^ ^> c$ OO ro t^ ^^ \^^ l/^ ON l~"" O ON c: -O ^ OJ Oi OO" ^ 00 *f CM CM 1—1 ^O ^^* ' — ' vO <^5 CM" ON CM t^ ^H CM CM rt (S> E ON CM" CO rt -^ ** ** & Je 1— 1 ^2 -f ON ^H \r> ON 00 •^H ^O CM ID t^» ^< ("^ tTS OO ON ON tD to 00 u, ^ c" OO ro ON I t— o" OJ J3 Ov fO -^ ON T! 1-1 -f 1 1-1 O CM r~i « OQ i— i CM Tt<" t\ *""" ff> t& -a •kJ c *2 ON OJ t^ «) ON rt ^ £ ° in *2 (U (-• > S .s g ^Q aj u. ^ rt rt X E a) •sf Ov ^H ID -^ CM CM ON NO OO NO co i— i rj< ^^ 1—1 \^^ ^^ ^^ to O ON" o" co" CM I— >-c oo 00 00 O i-i \O NO O1 NO ON SO ^H ^ ro 1-1 OO CM NO CM ID ^f 00 NO O NO ON -H r- ON O O 00 O so fn ON e f> %> I s % u E 0) O 13 E £ D C •2 c tuo o c 'z: '> '<« II •£•? H rt en 0) 3 4J ncome obligations •c 3 C _o c (U a T3 o CJ rt 4-1 0) tn c flj !_, . S. Governme orporate fixed ommon stocks referred stocks nits in combin^ rt OJ "rt 4_> "rt OJ "rt O H ess custodian f p U U OH p Di J REPORT OF THE TKKASl'KKK 61 50 5 VJ *^H 00 O OO OO r*} CN O CN" O fC I— OO •r-.' OO" 1^1 -t § O t-~ OO" •o 00 O OO OO" t— " OO -H "> OO O -f .a o\ 00 CO S k-i -»*- e o S U) 3 c 3 O -a OJ en 0) T3 C 01 3 01 C x- CO j_, .9 '•£ c j CO ^ rt 1- 1- , QJ OJ O. a o O cS •fp^ c ^ a >, -J-) _tj v. •*-a o 0 c 0 *(* FIGURE 3. Ciliated cells of the external yolk sac at Stage 20. The cells are polygonal in outline and are covered with cilia which appear to beat in an uncoordinated but unidirectional fashion. FIGURE 4. Paddle-type ciliated cells on the mantle of the Stage 25 embryo. These cells are almost invariably separated from each other by other cells of the epithelium. The indi- vidual cilia are expanded into flattened "paddles" or, more infrequently, spheroidal knobs apically or subapically. Although the beat is unidirectional, it is not synchronous. FIGURE 5. Section of one paddle-type ciliated cell. The paddles are composed of an expanded portion of the cilia membrane, which forms an electron transparent vesicle. The axoneme bends sharply within the paddle. FIGURE 6. Mantle of a Stage 20 embryo during closure of the shell sac Paddle-type ciliated cells have developed on the future ventral surface (at magnification bar) and sides of the edge of the mantle. FIGURE 7. Future dorsal surface of the Stage 20 embryo. The number of paddle-type ciliated cells has increased around the eye primordium (e) and above the mouth (mo). The arrows indicate the direction of flow of the chorionic fluid. 106 J. M. ARNOLD AND L. D. WILLIAMS-ARNOLD FIGURE 8. Lateral view of a Stage 22 embryo. The external yolk sac has elongated along with the embryo, and cells in the yolk stalk region have changed from polygonal to elongate. A few paddle-type cilia occur on the arms and distLl portion of the funnel folds (ff). The number of ciliated cells on the optic stalk has increased. FIGURE 9. Paddle-type ciliated cells have increased in number and are more or less evenly distributed on the future ventral mantle. The gill primordia (g) and proximal portions of the funnel complex remain unciliated. S(JUII) EMBRYO CIUATURE 107 FIGURE 10. The dorsal portion of the mantle is becoming ciliated by Stage 22, although the future dorsal midline is last to develop ciliated cells. The fin primordia (f) do not have any cilia on them. FIGURE 11-13. The number of paddle-type cilia on the mantle has continued to increase and the pattern seems random. Parts of the eye stalk and and some future ventral areas remain unciliated. The arrows indicate the direction of fluid flow. See text for details. FIGURES 14-16. At Stage 26, there are regions of the head which lack ciliated cells, notably above the eye and on the future dorsal head. A major change has occurred on the mantle with the appearance of "uniform-type" ciliated cells. These cells tend to form lines in which all the cilia have unidirectional and synchronous beat. (s - siphon). See text for details. On Figure 15 the arrows indicate the direction of movement of the chorionic fluid. 108 J. M. ARNOLD AND L. D. WILLIAMS-ARNOLD to have any unusual features. In SEM there is no evidence of synchrony of beat, hut in living embryos, the cilia beat toward the vegetal pole of the embryo. A second type of ciliated cell appears at about Stages 17 and 18, first on the ventral and anterior edge of the eye primordia, then on the future ventral margin of the mantle (Figs. 1 and 2). These cells are large, flattened, and separated from each other by non-ciliated neighboring cells (Fig. 4). These cilia have a unidirectional but not metachronic beat. They are larger than those on the external yolk sac and have an apical or subapical expansion three to five times the diameter of the rest of the cilium (Fig. 4). In the scanning microscope, this expanded portion frequently appears flattened into a paddle shape, although spheroidal expansions are also occasionally seen. Because of this morphology, they will be referred to as "paddle-type cilia." If the expansion is subapical, the tip of the cilium frequently emerges at an acute angle to the rest of the shaft. Trans- mission electron micrographs show the expanded portion to be a bleb of "empty" membrane at a bend in the axoneme. The axoneme has the typical "9 + 2" morphology (Fig. 5). In living cells, the paddle appears passive and follows the beat of the cilium. When seen on the return stroke with phase-contrast micros- copy, the plane of the paddles is close to and parallel with the cell surface. On the power stroke, it is perpendicular to the angle of thrust. At Stage 20, the paddle-type ciliated cells are on the sides of the mantle primor- dium, and the number of ciliated cells and area covered on the eye primordium has increased. They also appear above the mouth primordium (Figs. 6 and 7). At Stage 22 (Figs. 8-10), the paddle-type ciliated cells occur on the top of the optic stalk and on the future anterior surface of the optic vesicle, while the future posterior of the optic vesicle is unciliated. The mantle has expanded in area and turned downward to begin to cover the developing mantle cavity and the number of ciliated cells has increased accordingly. Most ciliated cells still occur singly, but occasionally two cells are in contact (Fig. 9). The fin primordia remain devoid of ciliated cells. The future dorsal region of the mantle still is temporarily devoid of ciliated cells (Fig. 10). A few paddle-type ciliated cells have appeared on the proximal portion of the arms and on the outer distal end of the funnel primordium (Fig. 8; cf. Fig. 11 ). The embryo elongates as it grows so that the external yolk sac is attached to the embryonic region proper by an ever-narrowing yolk stalk (Fig. 8). This also causes a change in shape of the individual cells, but the direction of beat apparently is unaffected. The distribution of cilia on the mantle becomes more uniform as the mantle grows over the future mantle cavity (Stage 23, Figs. 11-13). The edge of the mantle has a fairly continuous line of ciliated cells, and in some instances these cells appear larger with longer individual cilia. The fins do not have ciliated cells except at the margin between them and the base of the future dorsal surface (Fig. 12). The number of paddle-type ciliated cells on the body surface has increased but the organs or parts of the organs which will become incorporated into the mantle cavity remain unciliated (e.g., gills, anal papilla, part of the funnel folds; Fig. 13). The optic stalk has a somewhat uniform distribution of paddle-type ciliated cells except on the future posterior surface of the eye and in a band which lies between the optic stalk proper (Fig. 11 ). Later in embryonic development, this region gives rise to tissue which covers the optic vesicle and forms the definitive cornea. The tips of the arms and the developing suckers also lack ciliated cells. SQUID EMBRYO CILIATURE FIGURE 17. The uniform-type ciliated cells beat in a metachronic wave that is syn- chronized along several aligned cells. The projection in the lower left (arrow) may be a mechanoreceptor. FIGURE 18. In this developing uniform-type ciliated cell, cilia of several lengths can be seen and the beat, although unidirectional, is only partially synchronous. These cilia are of uniform diameter and end in a blunt taper. By Stage 26 (Figs. 14-16), the pattern of cilia on the head of the Stage 23 embryo has changed slightly. This includes an increase in the ciliated cells and two lines of paddle-type ciliated cells running from the base of the median pair 110 I. M. ARNOLD AND L. D. WILLIAMS-ARNOLD FiGfRK 19. The lines of uniform ciliated cells radiate from the developing Hoyle organ (hi at Stage 28. The chorionic fluid is circulated from the future posterior toward the edge of the mantle. FIGURKS 20-22. At Stage 27 the edge of the mantle has a band with relatively few- ciliated cells and the lines of uniform-type cilia are more prominent on the future dorsal and future ventral surfaces. On Figure 22, the arrows indicate the general direction of movement of the chorionic fluid. of arms back toward the optic stalk on the anterior dorsal head (Fig. 14). Ciliated cells are approximately evenly distributed on the posterior dorsal head. The ventral surface of the head is uniformly covered with ciliated cells except between SQUID EMBRYO CILIATURE HI the future ventral-most arms, in a small region over the otocysts and at the junc- tion of the optic lobe region of the optic stalk and the optic vesicle (Fig. 15). The mantle pattern of ciliature has undergone extensive changes by Stage 28 and a different type of ciliated cell has appeared amid the uniformly distributed paddle-type ciliated cells (Figs. 14-16). These cells are elongate on the anterior- posterior embryonic axis and show a strong tendency to form anterior-posterior contact with each other. The cilia are uniform in diameter and taper to a point without an expanded region (Figs. 17 and 18), and will be referred to here as "uniform-type ciliated cells." These cilia beat in a metachronic wave continuing across the cell borders. The cilia on these cells occur close to the margin of the cells. As many as seven cells occur in such an alignment, with groups of four and five common. These lines rarely branch or fork and if they do, it is with one cell. The lines of cells radiate from the future site of the Hoyle organ (hatching gland) which is situated between the fins (Fig. 19). The cilia of the cells do not grow out of the cell simultaneously : In young cells several lengths of cilia can be found and the beat pattern seems irregular (Fig. 18). At Stage 27, the mantle elongation has increased so that a band relatively sparse in ciliated cells occurs behind its anterior margin (Figs. 20—22). The ciliature pattern is unchanged except that the ciliated cell-free area of the dorsal head has increased. Aligned groups of uniform-type ciliated cells of the mantle occur in one ventral medial patch and in a "V" on the dorsal surface (Figs. 21 and 22). Between the arms of the "V" and on the sides of the mantle are exten- sive areas of paddle-type ciliated cells. As the embryo grows, the external yolk sac decreases by digestion of yolk and by transfer of its contents into the developing internal yolk sacs. The circumoral musculature constricts the hetnal space in the arm region and pulsation of the yolk sac decreases. The cilia of the yolk sac also shorten and become irregular in diameter. They still beat, however, because particles in the chorionic fluid or sea water sweep along the surface of the external yolk sac. High magnifica- tion reveals that the apical and subapical portion of these cilia have expanded to flattened discs similar to those of the "paddle-type cilia." Furthermore, these discs frequently have double ridges at their edges, each of axoneme diameter. (Fig. 23). Before hatching (Stage 29), the external yolk sac continues to decrease in size and the embryo to elongate, causing the yolk-sac cells to become rounded and have a smaller external area. The ciliature pattern on the mantle remains similar to that of the Stage 27 embryo (cf. Figs. 21 and 22 with Figs. 24 and 25). but on the dorsal surface of the head, a fourth type of ciliated cell appears (Figs. 24—26). This type occurs as two to four lines of "single-file" ciliated cells running along a central head crest ( Fig. 26). Two such rows in the central region of the head are common, but four or occasionally three lines occur. Side-to-side synchronous beat appears along the length of the "single-file" ciliated cell. Exceptionally, the tips of these cilia are apically or subapically expanded : mostly they end in a short tapered tip. After hatching (Stage 30), an abrupt change occurs in the epithelium of the mantle. Within 24-36 hr, the cells of the mantle begin to slough off and are released into the surrounding sea water as individuals or in small groups (Figs. 27 and 28). The surrounding glandular elements may also be Large gaps appear between the cells bridged by many fine branching extensions of surface material (Figs. 28 and 29). 112 J. M. ARNOLD AND L. D. WILLIAMS-ARNOLD FIGURE 23. After heart and gills assume the respiratory and circulatory functions of the external yolk sac, the sac stops pulsating and the cilia shorten and develop disc-like paddles, possibly by coiling of the axoneme. They continue to beat. FIGURES 24 and 25. At Stage 29, the ciliature pattern on the mantle has changed little from Stage 27, but "single-file" ciliated cells have appeared on the head (arrows). FIGURE 26. Higher magnification of the single-file ciliated cells show the cilia arise from a central crest of cells and beat in a lateral wave. SQUID EMBRYO CII.IATURE FIGURE 27. After hatching, the ciliated cells of the mantle are shed within 24-36 hr. The ciliated cells of the head are lost later. FIGURE 28. During shedding, cells separate from each other in longitudinal rows and are lost singly or in small groups. Mucus cells are also probably lost. FIGURE 29. As the cells of the mantle epithelium pull apart, a meshwork of branching strands which intertwine with cilia becomes apparent. Frequently, cilia can be seen sticking to the surface of non-ciliated cells. The head ciliature does not similarly change during the '. : days we have been able to keep newly hatched juveniles in good health (Fig. 27). Presumably it does so later. The external yolk sac usually is dropped and degenerates soon after hatching. Its cilia are active, however, for a short time after it is cast aside. 114 L M- ARNOLD AND L. D. WILLIAMS-ARNOLD Occasionally, non-moving bundles of cilia are observed in the mantle (Fig. 17), or more commonly on the aboral surface of the arms. Presumably these are the mechanoreceptors described by Sundermann-Meister (1978). DISCUSSION The question raised by these observations is : Why do three different types of ciliated cells exist to perform similar functions? The time of appearance and placement on the embryo may provide some insight about the function of each type ; still, all the questions cannot be answered. The function of the fourth type, the "single-file" cilia, is unknown to us. Cilia of the external yolk sac arise first. Since the external yolk sac is an embryonic digestive, circulatory, and respiratory organ, its ciliated cells are most probably associated with respiration. The hemal space of the external yolk sac is separated from the chorionic fluid by a thin layer of flexible pavement-like cells. Cilia in this position prevent localized accumulations of carbon dioxide (or excretory products) and facilitate respiratory exchange. Cilia of these cells have a directional beat, but individual cilia do not beat synchronously and unlike amphibian embryos (Kessel ct al., 1974; Billett and Courtenay, 1973) the ciliated cells are in contact with each other at their margins. Collectively, these ciliated cells must be quite effective in circulating the chorionic fluid, which in turn is able to exchange metabolic substances through the chorion. Their partial structural degeneration when the gills and branchial hearts take over the respiration and circulation of the developing embryo suggests further that they function primarily in respiratory exchange. Apparently, the development of a disk-like paddle on or near the tips of the cilia later in development is related to the coiling of the axonemes in each individual cilium. The function of the paddle-type cilia is more obscure. The development of a flattened expanded tip or subapical portion could increase efficiency of the effective stroke. By being flattened, the expanded tip would create little drag in the recovery stroke if the plane of flattening is perpendicular to the plane of bending. The sharp bending of the axoneme where it passes through the expanded portion could be structurally related to strengthening the "paddle" during the effective stroke if the curvature of the band is oriented into the plane of the effective stroke. Obser- vations with phase contrast microscopy confirm this is likely. Paddle-type ciliated cells always appear isolated from other ciliated cells and first appear on the organ primordia. They are particularly prominent and active. Their placement on the eyes, head, arms, and early mantle suggest that they generally circulate the chorionic fluid past the embryo but also prevent the embryo from directly contacting and possibly attaching to the inside of the chorion. Isolated, non-ciliated embryonic cells do stick to the chorion. Their position on the embryo would also facilitate "cutaneous" respiration. They do not beat synchronously, suggesting autonomous control. The uniform-type ciliated cells have the appearance usually associated with ciliated cells on other tissues and embryos. Their most striking feature is that they have a metachronic beat and synchrony extends from one cell to another in a continuous wave. The alignment and synchrony of the beat of these ciliated cells seems quite effective in moving large volumes of fluid. Since these ciliated cells appear about the same time (Stage 26) as mantle contraction begins and the gills and branchial hearts begin to function, they could increase chorionic fluid circulation. In fact, older embryos SQUID EMBRYO CILIATURE 115 continuously tumble inside the ever-expanding clmrinn. Dechorionatecl embryo.-, observed in a dish of sea water continuously move posteriorly along the bottom of the dish. Therefore, these rows of aligned synchronous ciliated cells possibly function in enhancing the circulation of the chorionic fluid and movement of the older embryo. This is advantageous to the embryo for respiration and excretion. The detailed cause(s) of the post-hatching sloughing of the skin of the newly hatched juvenile are unknown. Hut the phenomenon obviously is related to the breaking of cell-to-cell contacts and cells are lost either as individuals or in small groups. The filamentous material that temporarily connects the cells as they slough is probably the remnant of intercellular substances, since the cell membrane does not appear to be involved. We assume that this sloughing is followed by the development of the slimy skin tvpical of cephalopods. Hecause ol the dif- ficulties in rearing juvenile squid in the laboratory, we were unable to collect data on this. The single-file cilia do move, so it would seem unlikely that they function as mechanoreceptors ; and since they are not in an enclosed pit, it is unlikely that they are chemosensorv. They could possibly function to keep the head free ot detritus, but there is never much participate material in the chorionic fluid and since they are in a single row, it seems unlikely that they could generate much force for this purpose. However, they may move mucus produced by the gland cells of the skin on the head. How long these cilia persist and their ultimate fate is unknown. The developmental significance of morphologically different ciliated cell types is puzzling. Because they are functionally different and appear at different times during development and on different organs, there must be adequate selection pres- sure to preserve their separate phenotypes. This is striking since on other embryos, only one recognized motile ciliated cell type must provide all the surface ciliary functions for complete embryonic development. Perhaps these differences between vertebrate embryos and cephalopod embryos reflect their separate evolutionary histories. The authors wish to thank Drs. Lewis Tilney and Michael Hadfield for reading this manuscript and Mrs. Frances Okimoto and Mrs. Charlotte Daspit for preparing the typescript. This work was supported by grant #PCM 7619301 from the National Science Foundation and the Teuthis obsciira fund. Dr. Virginia Peters taught us many of the techniques used here. SUMMARY 1. Four types of beating ciliated cells, which appear at different times during embryonic development of the squid Loligo pealci, are described. 2. The ciliated cells of the external yolk sac first appear during or at onset of organogenesis, as polygonal flattened cells in a pavement-like arrangement. The cilia of these cells are uniform in diameter and end in a blunt or slightly tapered tip. Their beat is asynchronous and uncoordinated, but unidirectional toward the vegetal pole. A large hemal space develops between these cells and the yolk syncytium of the external yolk sac. In later stages of development, when primary respiratory function is assumed by the gills (and probably the skin), the 116 J- M. ARNOLD AND L. D. WILLIAMS-ARNOLD cilia on those cells develop a flattened paddle at or near their tips, which apparently is associated with coiling of the axonemes within them. 3. Paddle-type ciliated cells develop on the embryonic body proper and are always isolated from each other except at the anterior edges of the mantle. These cilia beat unidirectionally and asynchronously. The axis of the paddle is parallel to the cell surface on the return stroke, thus offering little resistance to the bending of the cilia, and is perpendicular to the direction of the power stroke, thereby increasing the effectiveness of the power stroke. Cells of this type appear on the mantle and head in patches, but not in areas to be covered by other tissues such as the mantle. 4. At or just before Stage 26, a third type of ciliated cell appears in rows on the mantle. The individual cilia are uniform in diameter and end in a tapered tip. The cilia on these cells beat in a metachronic wave synchronized among several cells aligned in rows. These lines of "uniform-type" ciliated cells are apparently very effective because the embryos begin to swim and tumble actively in the chorionic fluid and circulate it rapidly. This probably enhances respiration and prevents the embryo from sticking to the inner surface of the chorion. 5. A fourth type of beating ciliated cell, the "single-file" ciliated cell, appears on the head and ventral arms of the late Stage 28 or early Stage 29 embryo. The cilia of these cells are aligned in single file on the anterior-posterior axis and beat in a synchronized side-to-side wave. The function of these cells is unknown. 6. At hatching, the entire mantle epithelium degenerates and is sloughed off. Presumably, the epithelium of the head is also shed, but was unobserved by us. LITERATURE CITED ARNOLD, J. M., 1962. Mating behavior and social structure in Lolif/o pealei. Biol. Bull.. 123 : 53-57. ARNOLD, J. M., 1965. Normal embryonic stages of the squid I.nlif/o pealei ( Lesueur ) . Biol. Bull.. 128: 2-1-32. ARNOLD, J. M., 1971. Cephalopocls. Pp. 265-311 in G. Reverberi, Ed., in Experimental cmbr\olou\ of marine and fresh water invertebrates. North-Holland Publishing Co. ARNOLD, J. M., AND L. D. WILLIAMS-ARNOLD, 1976. The egg cortex problem as seen through the squid eye. Am. Zoo!., 16: 421-446. BERGQUIST, P. R., C. R. GREEN, M. E. SINCLAIR, AND H. S. ROBERTS, 1977. The morphology of cilia in sponge larvae. Tissue <.'•'-• Cell, 9: 179-184. BILLETT, F. S., AND T. H. CorRTENAY, 1973. A stereoscaii study of the origin of ciliated cells in the embryonic epidermis of Ambvsto/iia mcxicanwn. J. I'.mhrvol. E.rp. MorphoL, 29: 549-558'. Gut x/, H., A.-M. MuLTiER-LAjors, R. HERBST, AND G. ARKENBERG, 1975. The differentia- tion of isolated amphibian ectoderm with or without treatment with an inductor. U'illiclni Roux Arch. Entwickl.-Mech. Or,/., 178: 277-284. KESSEL, R. G., H. W. BEAMS, AND C. V. Sum, 1974. The origin, distribution, and disappearance of surface cilia during embryonic development of Ratni pipicns as revealed by scanning electron microscopy. Am. J. Anat., 141: 341-360. LANDSTKOM, LI., 1977. On the differentiation of prospective ectoderm to a ciliated cell pattern in embryos of Ambystotna mexicanum, J . Embryol. Exp. MorphoL, 41 : 23-32. l.i KT, J. H., 19M. Improvements in epoxy resin embedding methods. J. Biophys. Biocliem. Cytol, 9: 409-414. SMITH, J. L., J. C. OSBORN, AND M. STANISSTREET, 1976. Scanning electron microscopy of lithium-induced exogastrulae of Xenopus laevis. J. Emhrvol. Erp. MorphoL, 36 : 513-522. St'NDERMANN-MEisTER, G., 1978. Ein neuertyp von cilienzellen in der Haut von spatembryo- nalen und juvenilen I.o/i 19 (30' , ) 19 (30% ) 19 (30'; i 19 (3()' , ' 28 (67%) 14 (33%) 14 (33%) 7 (17%) 0 Total tested : 64 42 12S CHARLES H. BIGGER TABI.K IV Relative aggressiveness of seven A. krebsi groups. As explained in detail in the text, each block of the matrix represents the results of eight 1-hr interactions of that combination. The set of interactions for each group pairing was scored for each group: moving away — /, contracting — 2, receiving peels but remaining in place — 3, remaining in place while the other anemone contracts or moves awav — I, initiating the acrorhagial response and placing peels on the other anemone — 5, no response during any of the interaction — NR. Total score for the horizontal group was sub- tracted from the diagonal group to obtain a measure of the relative aggressiveness of the groups; i.e., 0 — both groups of a combination were equally aggressive, " " score — the horizontal group was the more aggressive of the pairing, and "-)-" score — the diagonal group was the more aggressive of the pairing. 8cl Sol 4.11 Sea 62 6e 3e 2e NR -16 -12 -6 -2 0 + 14 4dl XR -10 -9 -8 -4* 0 Sea NR NR NR NR NR 62 NR NR NR NR 6e NR NR NR 3e NR NR 2e NR * Score based on only four interactions. One group died of causes unrelated to the study toward the end of this investiga- tion, leaving one combination and its reciprocal with four observations. An anemone of each group was also removed and reintroduced to its groupmates (4 X). No groupmates responded. The red groups never responded to another red group but did have acrorhagial responses to both green groups. The green groups directed acrorhagial responses toward each other and most red groups. Some intergroup combinations consistently resulted in the same group initiating an acrorhagial response, but in others the initiating group varied. Distance between the paired groups in the field did not seem to be an indicator of the outcomes of those interactions. In 2Sc/c of the interactions involving an acrorhagial response, both anemones initiated a response. The outcome of each interaction was scored in terms of aggressiveness for each member of the anemone pairs (Table IV). For each combination of two groups, one group's score was subtracted from the other group's score and divided by the number of interactions for each combination to give a measure of the relative aggressiveness of the two groups under those conditions. Whether anemones were introduced or were residents in the interactions did not significantly (Wilcoxan signed-rank test) affect the outcome. Therefore, all interactions ( ,S ) of each group combination are combined and given in Table IV as a relative measure of the aggressiveness of the responsive groups. DISCUSSION The discovery of such a complex behavior as the acrorhagial response in the morphologically relatively simple actinians has raised a number of Questions. What SKA ANEMONE ACRORHAGIAL HEHAYlok 1 _><) is the behavioral nature of the response? What animals will elicit the response? Do factors such as prior experience, relative size, residence, and genotype of the interacting anemones affect the outcome? What is the functional significance of the response? Is the acrorhagial response related to other coelenterate self/not-self recognition systems or to an immune system ? Despite an earlier report to the contrary ( Francis, 19731) ) and a single observa- tion under unusual circumstances (Lindherg, 1976) the current study presents an observation of A. xanthogramwtica displaying an acrorhagial response under field conditions, so that all anemones with acrorhagi that have been examined are known to display a similar acrorhagial response (Abel, 1954; Bonnin, 1964; Francis, 19731) ; Bigger, 1976). As pointed out by Francis ( 19731) ) the acrorhagial response meets the general definition of an aggressive behavior. Some definitions (e.g., Hi tide. 1970) require an aggressive behavior to be directed towards the other indi- vidual, a criterion met by the application component of the acrorhagial behavior of all species examined except A. sargasscnsis. Even in that instance, one can make a case for the directed nature of the A. sargasscnsis acrorhagial response because the specificity of the peeling insures a directed nature; i.e., the response can only go to completion upon contact with the proper animal. Of special interest is the fact that anemones will respond to the same species as food or as an acrorhagial target, but at different times. Although there is a fine line between predation and aggression, in some cases, (i.e., in corals, Lang, 1971 and 1973) the same phenomenon might be considered as both. Because predation and the acrorhagial response are mutually exclusive, one need only consider the acrorhagial response in terms of aggression. Bonnin ( 1964) demonstrated that induction of an acrorhagial response in A. cqnina caused a lowering of the threshold for subsequent acrorhagial responses elicited at 10-min intervals, and that by the sixth acrorhagial response specimens of A. eqitina became unresponsive to further stimulation. In the present study, a similar threshold lowering and elevation was observed when the A. krcbsi acror- hagial response was elicited at 10-min intervals. However, not all specimens of A. krcbsi became totally refractory and at longer intervals between response elicitation (15+ min), the A. krcbsi acrorhagial response threshold remained low (Fig. 11 ). In A. krcbsi, prior experience influenced later responses over periods as long as 2 hr ; this should be considered in experimental design. Acrorhagial responses have not been elicited by non-coelenterates in the five anemone species that have been examined: A. cqnina (Bonnin, 1964), A. clcgan- tissiina (Francis, 19731) ), A. krcbsi (Bigger, 1976), A. sargasscnsis and B. cavcr- nata (this study). Therefore, these sea anemones must compete with non- coelenterates for resources in some other fashion. To date (Abel, 1954; Bonnin, 1964; Francis, 1973b; Bigger, 1976), acrorhagial responses have only been elicited by some sessile anthozoans or C. .vainachana. Past consideration of the role of the acrorhagial response has emphasized intraspecific interactions on the grounds that those were much stronger than the interspecific acrorhagial responses (Francis, 1973b). The highly predictable, full acrorhagial responses of either A. krcbsi or B. cavernata to A. sargasscnsis indicate that acrorhagial responses could be effec- tively employed against some other anthozoans, but the lack of pertinent field data limits a full assessment of the interspecific role of the acrorhagial response. This study amplifies the preliminary report (Bigger, 1976) of a difference in the response of A. krcbsi to polyps and medusae of the scyphozoan C. .raniachana. With twice the effectiveness of the medusa in soliciting acrorhagial application, the polyp appears either to have a qualitatively more effective application-eliciting 130 CHARLES H. BIGGER factor or to contain more of the eliciting factor. In the light of Bonnin's suggestion (1964) that the nematocysts of the target animal constitute the eliciting factor, it should be noted that the nematocysts of polyp tentacles and medusa mouth fronds (an area that touched the anemones) are morphologically the same and appear to be present in roughly equivalent numbers (Mariscal and Bigger, 1976; unpub- lished observations). Nematocyst toxins from Cossiopca medusae and scyphistomae have not been compared. For the most important measure of the responses, peel elicitation, 30% of the polyps, but never the medusae, elicited peels. These results point out the separation between the acrorhagial response components (acrorhagial expansion, application behavior, and ectodermal peeling; see Bigger. 1976) and suggest three possibilities: 1) Each component may require a different eliciting factor (or combination of factors). The medusae and polyps would contain the expansion factor but differ in their complement of other factors. 2) The receptors for the three components may have different thresholds for the same factor. The medusae and polyps would have enough of the eliciting factor to exceed the acrorhagial expansion threshold but would be quantitatively differentiated by the receptors of the other two components. 3) The discrimination would be based on some combination of the first two. At this time there are not enough data to suggest one possibility more than another. Francis (1973b) proposed that the acrorhagial response primarily functioned in intraspecific competition for space. Central to Francis' hypothesis is the concept that anemones distinguish clonemates from all other conspecifics, a concept derived from observations of A. clcgantissiina acrorhagial responses being elicited by all non-groupmates ("non-clonemates"). This study shows that not all A. krcbsi non-groupmates elicit a response. In fact, some specimens of A. krebsi with different color morphs, one of Francis' criteria for non-clonemates, did not respond to each other. Therefore, one cannot consider the acrorhagial response to be solely a case of an "individual" (clone) recognizing all other conspecifics as not-self and competing with them for available space. One must examine this as a case of related individuals ( individual = clone ) competing for space. These related individuals share alleles at some loci, which may include those determining surface molecules that participate in the recognition events of the acrorhagial response. The difference between the acrorhagial interaction of all A. clcgantissnna groups (Francis, 1973b) and the intergroup compatibility of some A. krcbsi groups could reflect a true species difference or, alternatively, a more homogeneous gene pool in the specimens of A. krcbsi sampled, such as the extremely limited or homo- geneous gene pool in the Maine population of the sea anemone HaUplanclla hiciac (Schick, 1976). The data of Francis (1973b) and this study demonstrating the wide number of conspecific groups recognized by A. clcgantissiina and A. krcbsi and the variability of the A. krcbsi acrorhagial responses suggest multiple alleles coding for acrorhagial recognition and perhaps many different loci, i.e., a complex polygenic phenomenon such as the mammalian histocompatibility system. There is also a major genetic influence in other coelenterate self/not-self recognition systems, e.g., overgrowth in hydroids (Ivker, 1972) and histoincom- patibility in corals (Hildemann ct a!., 1977) and gorgonians (Theodor, 1970). Thus reports of various factors, e.g., size (Brace and Pavey, 1978), controlling the initiation of an acrorhagial response must be viewed with caution unless the genetic variable is controlled. Because there are no inbred strains, this is difficult with a solely sexually reproducing anemone. Asexually reproducing anemones such as //. clcgantissiina and A. krcbsi, on the other hand, present the investigator with SKA AXKMOXE ACRORHAfil Al. HKHAYIOK a large number of genetically identical anemones which allow reproducible group combinations under a variable experimental condition. Brace and Pavey (19/S) reported a size hierarchy in the imitation of the acrorhagial response of A. equina (a solely sexually reproducing anemone), the larger anemones being first to respond. However, in approximately one third of the interactions of their study the smaller anemone was faster or as fast to initiate an attack This indicates that other factors should be considered. Very limited evidence concerning the influence of size on the initiation of acrorhagial responses in A. krehsi suggests that size plays at most a subordinate role to the particular group combination ("histoin- compatibility differences") in that species. The current study also suggests that residence in an area does not influence the outcome of A. krcbsi interactions. Because the test anemones of this study were moved while still attached to their shells, the experimental design only allowed residence to be considered in terms of the anemone's surroundings and not on the basis of pedal attachment. Ottaway (1978). in his recent field observations of Actinia tcncbrosa, noted "the successful aggressor was almost invariably the 'defender,' the anemone that had been stationary at the time of contact." Therefore, although general surroundings may not sig- nificantly influence the acrorhagial response, movement or long-term attachment may affect outcome. If the acrorhagial response functions in competition for space, as proposed by Francis ( 1973a and b. 1976). one needs to explain coexistence of competitive groups. Several models for invertebrate competitive or aggressive interactions have been used to discuss interspecific situations (e.g., Lang. 1973; Jackson and Buss, 1975; Council. 1976). A hierarchy of aggression among corals has been reported by Lang (1971 and 1973). Connell (1976) suggested that such linear hierarchies are inherently unstable and that the intervention of an outside force that selectively acted against the higher ranked members of the hierachy could explain the concurrent existence of all the groups. Jackson and Buss (1975) and Buss (1976) proposed an interaction model of "competitive networks" rather than a linear dominance, i.e., A > B > C > A, etc. Brace and Pavey (1978) reported such a "ring" situation in A. equina acrorhagial interactions and Table IV of the present study reveals one such network in A. krcbsi interactions. However, con- trary to the results of Brace and Pavey (1978) with A. equina. Table IV of this study indicates a high degree of variability in the outcomes of interactions between certain combinations of A. krcbsi. Connell (1976) found the same variable out- comes in tissue destruction and overgrowth among the corals he studied. Buss (1976) states that such competitive networks function by increasing the time required for a dominant to be established and thereby reduce the magnitude of a disturbance required to maintain diversity. Rather than the acrorhagial response being viewed in a limited sense as only the mechanism whereby a clone can capture territory from conspecific competitors, the acrorhagial response can be viewed as one of a set of ecological factors possibly maintaining a heterogeneous gene pool and. through indirect interactions with other ecological factors, causing an optimal utilization of available space. Hildemann and his associates (1975) specifically included the acrorhagial response when they categorized what they felt were four levels of "immuno- reactivity" in coelenterates. In discussing the acrorhagial response, they recog- nized that a response elicited within minutes after first contact could represent non-immunological recognition but went on to suggest that because the anemones live in a crowded habitat, prior sensitization was not ruled out. Bigger (1976) 132 CHARLES H. BIGGER reported the rapid elicitation of acrorhagial responses from A. krcbsi by several alloparric species, including Condylactis gigantca. C. .vamachana, and Ccrlanthcopsis americanus. A possible explanation, consistent with an immunological mechanism, for the rapid response to allopatric species would be that those allopatric target animals possessed a set of surface molecules so similar to those of previously encountered target animals that the two sets of molecules were perceived as the same. However, the concept of prior sensitization as a basis for the rapidity of the acrohagial response must be viewed with certain reservations. More recently (Hildemann ct a/.. 1979), it has been suggested that three mimimal criteria must be met for a phenomenon to be considered immunologic : cytotoxic or antagonistic reactions, selective or specific reactivity, and inducible memory or selectively altered reactivity on secondary contact. Thus, while self/not-self recognition is certainly the cornerstone of immunology, not all self /not-self phe- nomena are immunologic in nature. Reactions among coelenterates involving self/ not-self recognition include various cellular and behavioral phenomena in hydroids (Kato ct a!., 1967; Ivker, 1972), gorgonians ( Theodor, 1970; Bigger and Runyan, 1979), corals (Lang, 1971 ; Hildemann ct a/., 1977). and sea anemones (Abel, 1954; Purcell, 1977). While all the above coelenterate responses might be con- sidered to meet the first two criteria for an immune response, experiments examin- ing the third criterion have been performed only with corals. Although Hildemann ct al. (1977) demonstrated the characteristics of an immune system in corals, little is known about the recognition mechanisms, receptors, or molecular pathways involved, nor in some cases the effector cell types in the above mentioned coelenterate reactions. Until such information is acquired, suggestions of a common underlying recognition mechanism remain speculative. That the acrorhagial response utilizes a behavioral effector component does not preclude the use of a recognition system similar to that of other coelenterate self/not-self or immunological phenomena. However, the nature of the recognition along with many other questions about the functioning of the effector side of the acrorhagial response must await future investi- gations. Dr. R. X. Mariscal's advice during this study and his critical reading of a preliminary draft of this manuscript were greatly appreciated. Thanks are also due Dr. \Y. H. Hildemann for critically reading this manuscript and offering valuable suggestions. Dr. M. J. Greenberg for suggestions. Dr. W. Herrnkind and the FSU Psychobiology Program for the loan of equipment, Steve White for statistical assistance, and Lois Bigger for translations. Some financial support for this study was provided by XSF Grant #DEB 77-22148 to Dr. R. X. Mariscal. Portions of this paper were submitted to Florida State University in theses in partial ful- fillment of the requirements for the degrees of M.S. and Ph.D. This is contribution number 143, Tallahassee. Sopchoppy. and Gulf Coast Marine Biological Association. SUMMARY The acrorhagial responses of four sea anemones, Anthoplcnra krcbsi, Bnnodo- sonta cavernata, Ancinonia saryasscnsis, and Anthoplcnra .ranthograuiniica, are described. All four acrorhagial responses can be considered forms of aggression. The acrorhagial response is only one of several responses of sea anemones to contact with other animals ; others include several methods of avoidance and feeding. SEA AXEMONE ACRORHAG1AI. BEHAVIOR U3 Prior experience can influence the acrorhagial response. In A. krebsi, the effect of a prior encounter on the excitation threshold can he seen for at least 2 hr. Interspecific behavioral interactions were examined in ./. krebsi, B. ciwcrmihi. and A. sargassensis. With one exception, acrorhagial responses were only elicited by contact with some anthozoans. The exception is that some A. krebsi respond to the scyphistomae of the scyphozoan Cassiopea xamachana. Some C. xamachana medusae from the same clone also elicited acrorhagial expansion and application behavior but never acrorhagial peeling. Intraspecific interactions were examined in A. krebsi. Clonemates and group- mates never elicited acrorhagial responses from one another. Some non-group- mates, including different-colored groups, did not respond to one another and in some other group combinations the interactive outcome was variable. It is sug- gested the acrorhagial response involves multiple alleles and perhaps involvement of different loci coding for cell-surface recognition molecules. Several competition models were examined for these intraspecific interactions. An intergroup linear hierarchy was not found. The acrorhagial response is certainly an example of self /not-self recognition. This response has exquisite specificity and leads to cytotoxic effects. It cannot at this time be considered immunological. LITERATURE CITED ABEL, E. F., 1954. Ein Beitrag zur Giftwirkung der Aktinien und Funktion der Randsackchen. Zoologischcr Anzcit/cr. 153 : 259-268. BIGKLOW, R. P., 1900. The anatomy and development of Cassiopca xamachana. Boston Soc. Natitr. Hist. Mem.. 5: 191-236. BIGGER, C. H., 1976. The acrorhagial response in Anthopleura krchsi: intraspecific and interspecific recognition. Pages 127-136 in G. O. Mackie, Ed., Coeleuterate ecolony and behavior. Plenum Press, New York. BIGGER, C. H., AND R. RUNYAN, 1979. An in situ demonstration of self-recognition in gor- gonians. Develop. Cow par. Iwwunol.. 3 : 591-597. BOXXIN, J. P., 1964. Recherches sur la "reaction degression" et sur le fonctionnement der acrorrhages d' Actinia equina L. Bull. BioL. 98: 225-250. BRACE, R. C., AXD J. PAVEV, 1978. Size-dependent dominance hierarchy in the anemone Actinia equina. Nature. 273 : 752-753. Buss, L. W., 1976. Better living through chemistry : the relationship between allelochemical interactions and competitive networks. Pages 315-327 in F. W. Harrison and R. R. Cowden, Eds., Aspects of sponge bioloyy. Academic Press, New York. CARLGREN, O., AND J. W. HEDGPETH, 1952. Actiniaria, Zoantharia and Ceriantharia from shallow water in the Northwestern Gulf of Mexico. Tex. Univ. lust. Mar. Sci. PnbL. 2(2) : 141-172. CONNELL, J. H., 1976. Competitive interactions and the species diversity of corals. Pages 51-58 in G. O. Mackie, Ed., Coelenterate ecology and behavior. Plenum Press, New York. Dixox, G. Y., AND A. F. Dixox. 1891. Report on marine invertebrate fauna near Dublin. Proc. Royal Irish Acad.. (3)2: 19-33. DUNCAN, P. M., 1874. On the nervous system of Actinia. Part I. Proc. Royal . Land.. 22 : 263-276. EDMUNDS, M., G. W. POTTS, R. C. SWINFEN, AND V. L. WATERS, 1976. Defensive behavior of sea anemones in response to predation by the opisthobranch mollusc Acolidia papillosa ( L. ) . J. Mar. BioL Assoc. U.K., 56: 65-83. FRANCIS, L., 1973a. Clone specific segregation in the sea anemone Anthopleura elcgantissima. BioL Bull.. 144: 64-72. FRANCIS, L., 1973b. Intraspecific aggression and its effect on the distribution of Anthopleura clegantissima and some related sea anemones. BioL Bull.. 144 : 73-92. FRANCIS, L., 1976. Social organization within clones of the sea anemone Anthopleura ele- i/antissinia. BioL Bull.. 150 : 361-376. 134 CHARLES H. BIGGER GOSSE. P. H., 1860. A history of British sca-ancmoncs and corals. Van Voorst, London, 362 pp. HII.DEMANX, W. H., D. S. LINTHKUM, AXD D. C. YANK, 1975. Immunoincompatibility reactions in corals (Coelenterata). Pages 105-114 in W. H. Hildemann and A. A. Benedict. Eds., Immunologic phyloycny. Plenum Press, New York. HILDEMAXN, W. H., R. L. RAISON. G. CHEUNG, C. J. HULL, L. AKAKA, AND J. OKAMOTO, 1977. Immunlogical specificity and memory in a scleractinian coral. Nature, 270 : 219-223. HILDEMAXX, W. H., C. H. BIGGER. AND I. S. JOHNSTON, 1979. Histoincompatibility reactions and allogeneic polymorphism among invertebrates. Transplant. Proc., 11 : 1136-1141. HIXDE, R. A.. 1970. Animal behavior (2nd edition). McGraw-Hill, Xew York, 876 pp. HOLLARD, H., 1851. Monographic anatomique due genre Actinia de Linne, considerere comme type du groupe general des Polypes Zoanthaires ( d'apres les Act. scnilis et cqirina). Ann. Sci. Nat. Zool. Biol. Anim. (3)15 : 257-291. IVKER, F. B., 1972. A hierarchy of histo-incompatibility in H \dractinia cchinata. Biol. Bull.. 143: 162-174. JACKSON, J. B. C., AND L. Buss, 1975. Allelopathy and spatial competition among coral reef invertebrates. Proc. Nat. Acad. Sci.. U.S.A., 72: 5160-5163. KATO, H., E. HIRAI, AXD Y. KAKINUMA, 1967. Experiments on the coaction among hydrozoan species in the colony formation. Sci. Rep. Tohoku Unii'. Ser. IV (Biol.), 33: 359-373. KOROTXEFF, M. DE, 1876. Organes des sens des Actinies. Arch. Zool. E.rp. Gen. 5 : 203-208. LANG, J., 1971. Interspecific aggression by scleractinian corals. 1. The rediscovery of Scolyinia cubensis (Milne Edwards and Haime). Bull. Alar. Sci., 21: 952-959. LAXG, J.. 1973. Interspecific aggression by scleractinian corals. 2. Why the race is not only to the swift. Bull. Mar. Sci., 23 : 260-279. LINDBERG, W. J., 1976. Starvation behavior of the sea anemones, Anthoplenra xanthogrammica and Mclridiuin senile. The Biologist, 58: 81-88. MARISCAL, R. N., AND C. H. BIGGER, 1976. A comparison of putative sensory receptors asso- ciated with nematocysts in an anthozoan and a scyphozoan. Pages 559-568 in G. O. Mackie, Ed., Coelenterate ecology and behavior. Plenum Publishing Corporation, New York. OTTAWAY, J. R.. 1978. Population ecology of the intertidal anemone Actinia tenebrosa. I. Pedal locomotion and intraspecific aggression. Aust. J. Mar. Frcshw. Res.. 29: 787-802. PANTIN, C. F. A., 1952. The elementary nervous system. Proc. R. Soc. Loud. B. Biol. Sci., 140: 147-168. PURCKLL, J. E., 1977. Aggressive function and induced development of catch tentacles in the sea anemone Metridinni senile ( Coelenterate, Actinaria). Biol. Bull., 153: 355-368. RAIT, W., 1829. ('her die Polypen in Allgcineinen und die Aktinicn im Bcsondcrcn, Verlage des Grofsherzogl. Sachs, privileg. Landes-Industrie-Comptoirs, Weimar, 62 pp. Ross, D. M., 1974. Behavior patterns in associations and interactions with other animals. Pages 281-312 in L. Muscatine and H. M. LenhofT, Eds., Coelenterate Biology: Reviews and Ncit< Perspectives. Academic Press, New York. SciricK, J. M., 1976. Ecological physiology and genetics of the colonizing actinian Haliplanclla liiciac. Pages 137-146 in G. O. Mackie, Ed., Coelenterate ecology and hcliavior. Plenum Press, New York. THEODOR, J., 1970. Distinction between "self" and "not-self" in lower invertebrates. Nature, 227: 690-692. WALTON, C. L., 1910. On some colour variations and adaptations in Actiniae. /. Mar. Biol. Assoc. U.K., 9 : 228-235. WILLIAMS. R. B.. 1978. Some recent observations on the acrorhagi of sea anemones. J. Mar. Biol. Assoc. U.K.. 58: 787-788. Reference: Biol. Bull. 159: 135-147. (August, 1980) THE IMPLICATION OF CARBONIC ANHYDRASE I> THE PHYSIOLOGICAL MECHANISM OF PENETRATION OF CARBONATE SUBSTRATA BY THE MARINE BURROWING SPONGE C LION A CELATA (DEMOSPONGIAE) WALTER I. HATCH ncptirtincnt of Biolo, resulting in a final substrate concentration of 29.42 mM when a 2.50-ml aliquot was added to the 6-ml reaction mixture. Enzyme preparation. Alpha-, beta-, and gamma-form specimens of Cliona cclata from the Northwest Gutter of Hadley's Harbor, Naushon Island, Massa- chusetts, were maintained in running sea water. The sponges were identified by spicule examination (Old, 1941). Sponge tissue was obtained by drilling a 10-mm hole in the apex of shells in which alpha- and beta-form sponges were burrowing or by removing a plug of tissue from gamma-form sponges. This material was freed of shell fragments and sponge tissue by treatment with warm concentrated nitric acid, rinsed twice with distilled water and once with 95^ ethanol, and spread on a slide for microscopic examination. PENETRATION MECHANISM OF CLIONA 137 In preparation for enzyme extraction, living sponges were cleaned of epibiota by scraping and scrubbing briefly with a bottle brush and tap water. A sponge cell suspension was produced by manually expressing excess water, fragmenting with a hammer when necessary, and forcing the fragmented sponge tissue through a 2-mm stainless-steel sieve to remove inorganic inclusions and shell (ubris. Enzyme solutions were prepared by disrupting the cell suspension with a Waring blender. After dialysis (24 hr, distilled water at 0°C), enzyme solutions were diluted to 1-2 mg total protein/ml. In order to eliminate the effects of non-catalytic protein buffering (Addink, 1971), controls were run in the presence of 1CT3 M acetazolamide resulting in com- plete inhibition of enzyme activity. Inhibited enzyme solutions were prepared by adding 3.3 mg sodium acetazolamide to a 2.5 ml aliquot of dilute enzyme solution and titrating back to the pH of the uninhibited enzyme solution (7.25) with HC1. Assay apparatus and procedure. The enzyme assay apparatus differed from that of Carter et al. (1969) in that the reaction vessel, delivery syringes, and reservoirs for buffer and substrate storage were all housed in a water bath thermo- statically regulated to 0± 0.1 °C. In addition, provision was made for transferring substrate solutions under positive pressure to avoid degassing and the resulting fluctuations in substrate concentration. A magnetic stirring device insured rapid mixing. In use, 2.5 ml of buffer was transferred from delivery syringe to reaction vessel and 1.0 ml of either enzyme or control solution added. After 30-60 sec for electrode equilibration, 2.5 ml of substrate solution was injected and a chart recorder started simultaneously. The reaction was allowed to proceed to equilibrium before the reaction vessel was drained and flushed for the next determination. The initial rate of catalyzed and control reactions was determined by fitting a tangent to the [H+J-time curve at pH 8.70. From these tangents A pH/min could be determined and converted to mM H+/min from a buffer-enzyme calibration curve. The enzymatic rate was determined by substracting mM H+/min produced in control reactions from that produced in enzyme catalyzed reactions. The mean of triplicate runs was recorded as enzyme activity. Intraccllular localization of clionid carbonic anhydrasc Clionid cell suspensions were disrupted in 0.025 M sucrose at 0°C by alternating 30 sec of grinding with 5 min of cooling (4 X ) in a Waring blender. The resulting homogenate was centrifuged (10 min, 500 X g) to remove inorganic inclusions and unbroken cells. The supernatant was recentrifuged (10 min, 700 X g) over a layer of 0.34 M sucrose to sediment nuclei. The supernatant was again collected and recentrifuged (10 min, 21,000 X g) to sediment mitochondria-like particles Each pellet was resuspended in distilled water and, along with the supernatant, dialyzed against stirred distilled water for 24 hr at 0°C. Each retentate was then assayed for carbonic anhydrase activity. An aliquot of the supernatant ( 10 min, 700 X g) of a second sucrose homogenate was centrifuged at 21,000 X g for 15 min. A second aliquot was spun (21.000 X g) for 60 min. Third and fourth aliquots were sonicated for 15 and 30 min, respectively, prior to centrifugation (21,000 X g, 60 min). The fifth aliquot was treated with N-butanol (20%, 1 hr, 0°C) according to the standard technique for butanol treatment outlined by Morton (1955). The butanol treated aliquot was also centrifuged for 1 hr at 21,000 X g, and the aqueous and butanol phases 138 WALTER I. HATCH separated. The supernatants thus produced were all dialyzed against distilled water, diluted to a constant volume, and assayed for carbonic anhydrase activity. E.rcai'ation rate and carbonic anhydrase content of clionid sponges If clionid carbonic anhydrase is involved in the physiological mechanism of excavation, it is probable that the excavation rate is related to the concentration of the enzyme in the sponge. The difficulties involved in determining the excavation rate of clionid sponges were circumvented by developing an alternative technique (Hatch, 1974). In that as much as 90% or more of the substratum is excavated in the form of carbonate chips (Warburton, 1958), the weight of the chips expelled per unit time was used as an estimate of excavation rate. Procedure. Alpha-, beta-, and gamma-form specimens of Cliona celata were cleaned of epibiota, positively identified as previously described, and placed in petri dishes in a flowing sea-water table. Each day expelled chips were collected on Whatman #4 filter paper, rinsed with distilled water and air dried for determina- tion of calcium carbonate weight. The excavation rate was recorded as ing CaCOs expelled per day (averaged over 4 days), and the sponges were divided into actively excavating (25-50 mg CaCO3/day), slowly excavating (1-25 mg CaCO^/day), and non-excavating (no detectable carbonate chips). Four specimens from each group were extracted with butanol as previously described and assayed for carbonic anhydrase activity. In addition, gamma-form cortical and medullary tissues were extracted and assayed separately. Effects of carbonic anhydrase inhibition on excavation rate In order to establish a possible relationship between the physiological mechanism of penetration and carbonic anhydrase, the excavation rate of Cliona celata was determined in the presence of acetazolamide, a specific inhibitor of this enzyme. Procedure. Thirty actively excavating alpha-form specimens of Cliona celata within Mcrcenaria inercenaria valves were freed of epibiota, identified, and returned to a flowing sea-water table as previously described. Through visual inspection of chip production over the course of a week, the 10 most active sponges were selected and arranged convex side up in petri dishes for chip collection (Hatch, 1974). An inhibitor stock solution was prepared by dissolving sodium acetazolamide in Millipore-filtered sea water (1 X 10 4M) and adjusting the pH back to that of the sea water system (8.26) with HC1. Inhibitor solution was introduced into flowing sea water through a metered-drip intravenous-infusion apparatus at a rate sufficient to maintain the desired acetazolamide concentration. The excavation rate was determined over two 24 hr periods under the follow- ing conditions: 2.5 1/min sea water flow-through, 0.25 1/min flow-through with 2.25 1/min recirculation, and 0.25 1/min flow-through with 2.25 1/min recircula- tion in the presence of 10~5 and 10~(i M acetazolamide. The recirculation was necessitated by the sponges' requirement for a minimum current velocity and economic constraint on producing a 10~4 M inhibitor concentration with a large flow-through. At the start of each determination sponges were carefully transferred under water into clean petri dishes. Expelled chips were collected after 24 hr as previously described. PENETRATION MECHANISM OF CLIONA 139 Effect of carbonic anh\drasc inhibition on in vitro metabolic rate The possibility of toxic secondary effects of acetazolamide on the excavating ability of Cliona eclat a was investigated with in vitro respirometry utilizing a Gilson differential respirometer. Procedure. The medullary tissue of beta-form specimens of Cliona celata was dissected from between the attached closed valves of Mcrccnaria mercenaria. The tissue was pooled, rinsed in running sea water, and teased into fragments less than 1 mm on a side. Approximately 1 cm3 of this tissue was placed in each Gilson reaction flask along with 3 ml of Millipore-filtered sea water. One half of one ml of a 2Q(/( w/v KOH solution was used as the CO2 absorbant. One milliliter of the inhibitor solution (8.8 X 1O3 or 8.8 X 10 4 mg sodium acetazolamide, ml Millipore- filtered sea water, pH 8.26) was added to the side arm of each reaction flask. Fourteen replicates of each inhibitor concentration were run. The flasks were equilibrated for 30 min, after which six readings were taken at 5-min intervals to establish the control respiratory rate. The inhibitor solution was then tipped into the sea water containing the sponge tissue and the O- consumption in the presence of 10~3 and 10~'; M acetazolamide was established with six additional readings at 5-min intervals. The total protein concentration of each flask was then determined by UV absorbance (Warburg and Christian, 1942). Effects of carbonic anhydrasc inhibition on papillary contraction The ability of clonid sponges to respond to chemical sitimuli with papillary contraction has been documented (Emson, 1966). It is likely that papillary contraction and the resulting restriction of ostial and oscular openings could indirectly inhibit excavation by limiting the flow of sea water through the sponge. In vivo polarographic respirometry was utilized to reflect the degree of papillary contraction elicited by sodium acetazolamide. Procedure. Alpha-form specimens of Cliona celata contained within Mercenaria mercenaria valves were selected to provide the maximum amount of respiring tissue that could be isolated by papillary contraction. Three individual valves were cleaned of epibiota and pooled for each experimental determination. The oxygen consumption of the sponge was determined in a flow-through system in which sea water (1000 ±2 ml/min) was passed first over the sponges in a sealed Incite chamber and then over a polarographic oxygen electrode coupled to a strip-chart recorder. The control respiratory rate was recorded for min, after which an inhibitor stock solution (10~3 M sodium acetazolamide in sea water) was introduced into the inlet flow at 1 ml/min for an additional 30-min period. The inhibitor stock solution was then increased to 10"- M and a final 30-min record of the respiratory rate was obtained. Papillae were then induced to contract in response to a mechanical stimulus (tapping the Incite chamber), checking the function of the apparatus and the responsiveness of the sponges. Three replicates were run in this manner at 17.5°C. Oxygen consumption, indicative of the state of papillary contraction, was calculated from the difference in O- saturation of the sea water before and after it had passed over the sponges. 140 WALTER I. HATCH 9.0 20 30 40 TIME (seconds) FIGURE 1. Clionid carbonic anhydrase reaction curves. Changes in pH are plotted against time for : A — distilled water control, B — sucrose extract prior to centrifugation, C — supernatant of sucrose extraction after centrifugation for 15 min at 21,000 X g, D — supernatant of sucrose extraction after centrifugation for 60 min at 21,000 X /min.) in the presence of acetazolamide. From this figure it can be seen that several minutes are required for the sponge to re-expand its papillae and resume normal oxygen consumption PENETRATION MECHANISM OF CLIONA 143 e Q. 30 0> §20 ON|° 1 ACETAZOLAMIDE, 10"* MOLAR ACETAZOLAMIDE, IO"5 MOLAR CONTROL 30 5 10 15 20 25 TIME (minutes) FIGURE 2. The effects of carbonic anhydrase inhibition on in vitro basal metabolism. after handling. The small peaks just before T0 may represent an increase in oxygen consumption resulting from an oxygen debt incurred during the handling period. No apparent difference in oxygen consumption can be seen from these graphs, indicating no detectable papillary contraction occurs in the presence of the concentrations of acetazolamide used to inhibit excavation. The mechanical tap, however, produces a dramatic change in oxygen consumption as the papillae contract in response to this stimulus. Visual observation during the course of the experiment confirms that no papil- Itochofilcal Stlmulut 30 60 TIME (minutes) 90 FIGURE 3. The effects of carbonic anhydrase inhibition on papillary contraction and in vitro respiratory rate. Respiratory rate of three pooled sponges is shown for three experi- mental runs. ( Baseline rate established for 30 min prior to the addition of 10~* and 10"5 M acetazolamide.) The mechanical stimulus served as an equipment check (note rapid drop in respiratory rate with papillary contraction). 144 WALTER I. HATCH lary contraction occurs in response to acetazolamide. It can then be concluded that the inhibition of clionid excavation by acetazolamide is not mediated by the ability of the sponge to detect and respond to inhibition by papillary contraction. DISCUSSION The exact manner by which carbonic anhydrase might mediate the finely controlled dissolution of the substratum by clionid etching cells is unclear. Unlike the carbonic anhydrase of Urosalpinx cinerea, which is in cytoplasmic solution in the microvillar zone of the accessory boring organ (Smarsh et al., 1969) or released in the ABO secretion (Carriker and Chauncey, 1974), clionid carbonic anhydrase appears to be associated with mitochondrial-sized particles. The presence of carbonic anhydrase in the tissues of Cliona cclata, however, indicates that the sponge has the capacity to greatly increase the speed of the reversible reaction H2O + CO2 ^± H+ + HCO3~ and accelerate the precipitation or dissolution of CaCOa by shifting the solubility equilibria. Both the tendency for rapidly excavating sponges to have higher carbonic anhydrase activities than non- excavating sponges and the inhibition of excavation rate by specific enzyme inhibitors strongly suggest the participation of this enzyme in the excavation process. The enzyme concentration of gamma-form sponges presents a problem. Although no substratum was present, the carbonic anhydrase concentration in the cortical tissues was much higher than in alpha- or beta-form sponges. The fact that the cortical tissues contained much higher concentrations of the enzyme than the medullary tissues suggests the enzyme may be localized in surface-lining cells. It is these cells that would come in contact with fresh substratum. The lower enzyme concentrations of beta-form sponges may also be attributed to a relative increase in non-excavating tissue. Rutzler and Reiger (1973) demonstrated a prominent nucleolus, abundant ribosomes, and an extremely active golgi complex within the etching cells of clionid sponges, suggesting that these cells are involved in protein and carbohydrate synthesis even in advanced stages of plasmolysis. In addition, the filopodial basket contains numerous vesicles as well as a flocculent secretory product. Although they never saw these vesicles emptying their contents to the outside (Rutzler, personal communication), they speculated that the flocculent material observed by them within the etching cells, and in the space between the cell and substratum, is responsible for the penetrating activity of these cells. It is unlikely, however, that the secretory product seen by Rutzler and Reiger represents carbonic anhydrase, as a direct catalytic breakdown of the substratum by the enzyme has been ruled out (Carriker and Chauncey, 1974). Clionid carbonic anhydrase activity can be removed from crude aqueous extracts by centrifugation, suggesting the enzyme might be associated with some subcellular structure (Morton, 1955). The enzyme could occur, therefore, in solution within a limiting semiperme- able membrane (Schneider, 1953) or intimately associated with the insoluble lipo- protein of the membrane (Morton, 1953, 1954). The fact that sonic or butanol disruption of membranes precludes loss of enzyme activity during centrifugation supports the hypothesis that the enzyme is contained within, or is associated with, the membrane of vesicles within the filopodial basket. There are several equally plausible mechanisms for the participation of carbonic anhydrase in the penetration of both organic and inorganic components of shell. First, carbonic anhydrase, within the filopodial basket, could accomplish the dis- PENETRATION MECHANISM OF CLIONA 145 solution of calcium carbonate by simply providing hydrogen ions for transport across the membrane. Thus, the mechanism may be similar to that found in parietal cells (Maren, 1967) in which the H+ ions are transported across the membrane along with Cl~ ions. It is equally valid to speculate that the mechanism of penetration involves car- bonic anhydrase in a role similar to that found in mammalian kidney (Maren, 1967). In clionid etching cells the exchange of hydrogen for bicarbonate ions would result in both the dissolution of the substratum and a lowering of the pH, with possible optimization for the enzymes responsible for the dissolution of the organic matrix. In addition, carbonic anhydrase has been implicated in the transmembrane transport of Ca-+ ions (Istin and Kirschner, 1968; Ehrenspeck et al., 1971). Thus, the penetration mechanism may involve a transmembrane flux of Ca2+ ions as well as a simple pH shift. It is also possible that the H+ ions resulting from the activity of carbonic anhydrase participate only indirectly in the dissolution of the substratum by pro- viding pH optimization for the activity of chelating agents and/or the enzyme responsible for the breakdown of the organic matrix. The experimental determina- tion of a chelating agent within the etching cell or its penetrative agent must be accomplished to confirm this hypothesis. In summary, carbonic anhydrase in Cliona celata, coupled with the demon- stration that the concentration of the enzyme in the sponge tissues is related to the excavating activity of the sponge, suggests this enzyme is involved in the physio- logical mechanism of penetration. This conclusion is supported by evidence that enzyme inhibition results in inhibition of the excavating ability of the sponge. The inability of the sponge to detect and respond to acetazolamide with papillary contraction or a depressed metabolic rate indicates that this inhibition of excava- tion rate is neither mediated by a behavioral response nor by a generalized metabolic inhibition. This work constituted part of a doctoral dissertation submitted to the Depart- ment of Biology, Boston University, and was conducted in conjunction with the Boston University Marine Program at the Marine Biological Laboratory, Woods Hole, Mass. It was supported in part by a graduate fellowship and by aid from grants through the Boston University Marine Program. SUMMARY 1. The marine burrowing sponge Cliona celata contains measurable carbonic anhydrase activity when assayed with a modified electrometric method. 2. When clionid tissues are extracted and centrifuged in isotonic sucrose, the majority of the enzyme activity is found in the mitochondrial size fraction, indicating the enzyme is membrane- or particle-bound. 3. Much of the enzyme activity can be released into solution with sonication. Treatment with isobutanol, which dissociates lipoprotein complexes, releases all of the enzyme activity into solution. 4. From a range of buffer solutions, 0.025 M trisaminomethane (pH 8.4-8.7) was chosen as the optimal buffer for the assay of clionid carbonic anhydrase. 146 WALTER I. HATCH 5. There is a positive correlation between the excavating activity of the sponge (mg of CaCO3 removed per day) and the level of carbonic anhydrase in sponge tissues. In addition, enzyme activity is concentrated in the cortical tissues of gamma-form C. eclat a. It is these tissues that are most likely to come in contact with fresh substrata in the form of shell fragments. 6. Sodium acetazolamide is capable of inhibiting the activity of clionid carbonic anhydrase in vitro with no apparent inhibition of the overall metabolism of the overall metabolism of the sponge. 7. Acetazolamide-induced inhibition of the sponge's carbonic anhydrase activ- ity markedly reduces the ability of alpha-form C. eclat a to excavate calcium car- bonate chips from the valves of Mercenaria niercenaria. 8. Similar concentrations of acetazolamide provoke no papillary contraction or reduction in oxygen consumption /';/ vivo, indicating the reduction in excavation rate is not due to the ability of the sponge to detect the acetazolamide and shut down the flow of sea water through its choanosome. It appears, therefore, that the pri- mary mechanism for the dissolution of calcium carbonate by C. cclata involves a shift in the carbonate solubility product in the microenvironment of the etching cell, mediated through the activity of the enzyme carbonic anhydrase. LITERATURE CITED ADDINK, A. D. F., 1971. Carbonic anhydrase of Sepia officinalis L. Client. Physiol.. 38 : 707-721. CARRIKER, M. R., AND H. H. CHAUNCEY, 1974. Effect of carbonic anhydrase inhibition on shell penetration by the muricid gastropod Urosalpinx cinerca. Malacologia, 12 : 247-263. CARTER, M. J., 1972. Carbonic anhydrase : isozymes, properties, distribution, and functional significance. Biol. Re?1., 47 : 465-513. CARTER, M. J., D. J. HARVARD, AND D. S. PARSONS, 1969. Electrometric assay of rate of hydration of CCX /. Physiol. , Land.. 204: 60-62. CHETAIL, M., D. BINOT, AND M. BENSALEM, 1968. Organe de performation de Purpura lapillus L. (Murcide) : histochimie et histoenzymologie. Call. Biol. Mar.. 9: 13-22. CHETAIL, M., AND J. FOURNIE, 1969. Shell-boring mechanism of the gastropod, Purpura lapillus L. : a physiological demonstration of the role of carbonic anhydrase in the dissolution of CaCO3. Am. Zoo/., 9 : 983-990. COBB, W. R., 1969. Penetration of calcium carbonate substrata by the boring sponge Cliona. Am. Zoo/., 9: 783-790. COBB, W. R., 1971. Penetration of calcium carbonate by Cliona cclata, a marine burrowing sponge. Ph.D. thesis, University of Rhode Island, 163 pages. Diss. Abstr. #72-9790. DAVIS, R. P., 1963. The Measurement of Carbonic Anhydrase Activity. Pages 307-328 in D. Click, Ed., Methods of Biochemical Analysis. (Vol. XI). Interscience, New York. EHRENSPECK, G., H. SCHRAER, AND R. SCHRAEK, 1971. Calcium transfer across isolated avian shell gland. Am. J. Physiol., 220: 967-972. EMSON, R. H., 1966. The reaction of the sponge Cliona celata to applied stimuli. Comp. Biochem. Physiol.. 18 : 825-827. GOREAT. T. F., AND W. D. HARTMAN, 1963. Boring sponges as controlling factors in the formation and maintenance of coral reefs. Pages 25-54 in R. F. Sognnaes, Ed., Mecha- nisms of Hard Tissue Destruction. Publ. No. 75, American Association for the Advancement of Science, Washington, D.C. re O1 0s 10 «r f~^ t~^ *^ O fN »f o ^ ""* ^ — -f -* vC ^1 r^ oo o t^. r f >-< r-4 (vj <*5 »—' O^' 5i <*) js^ rv) r^ r^ t^- t^ »— 1 ^S t r*"3 i i ••s» g O^ I 1 1 \ 1 1 1 1 "S ^•v >0 "*"* .R . ^ ? ^" £ * * * * S £ •? * * \s-> \r>\r> 5" IT) io' 5 £ * ^~V --^ X N X- — -, ^^v ^ — ^. ' V f N ^— X -^ ^ V_^ x__^ s^_^ %,^ , v.__^ LO W} IO ^ ^,^2. LO LO LO to IO IO ^ ->2 <^ " — ' — — "" — ' *—s ^-^ - — - - — - - — - vO 00 t^- fN -H O •&. £ ^ o • r^i ~ — rs r-* LO 10 re t~- -t — o' o O O O CN s •§ •§ c\ -f ~" ~ ^i ^— i o W} S ^ ° OO _LJ I) jj IT) I JJ JJ JJ Tl Tl Tl *- -H -H 41 r^ II II II rsj "H Tl "H : -H 41 -H 41 4^ 41 •*^ "^ IO \O tN 00 O^ ^H S-HJ rC °" S r^ S JC •«-t< rs ro O O fN <"N -f — i OO O ^H' ^4 ^ O -ci. 1 , a •<** 1 2 e s -S g'S 03 ^j *o 2 o'l 0^.0 -*-» 8 • i* rt «o ^s E'** "T? G»4 d * * ^^»^ «^ Q>) U) & * * # * * _^< «-M ^*4 f^j ^-j- ,_f« Cd 9A b) f| >* ^ IO IT) -^" to lO "^ "^ ^ '^t* •r^* *^4 ,-f4 C/3 ^J *^ ^^ ^— ' ^^ X_^ s^_^ ^_^ •^- v_^ s ^ "~^ ^^ ^^ O 00 t^ O OO fN <~n r-4 t^i 01 O ~H CO T$< ro OO f^J \O fN ^C O O 00 -H O o' ui 5 ^ "5 « ^^i ^ O Ov _U .11 JJ (^1 Tl ' 41 41 41 S 4) 41 41 0 ^ 2-ti-H-H - T) T| T) ' 4i 4i 41 t—4 O^ *^ ^ ^^v f--» fv^ r^* »/^ 10 (/>*** ^ 00 r^ «f OO O t~~ O fN m QS ^. ^ O 00 ro <*O (*O ^— ' r^ ^sj r^ r^ t^- fN T}< •— i 0s] ^^ ^^ 1 |.a *** • »^ *** 1 1 1 1 — ~£ *"• ^* **^ KB 8 O 1^ iu ^ 7r ~ >\ * 8 .S "§ ••«* * *^* !t S 5 ^ * * •e * in to to" i/^i^ir? s *- -2J •rf IO IO \O i/T^ i?T rt* LO LO i-O lO lO s~" ^ ^^^^ '1^-^' N — ^ "• — - ^ — ^ 72 ^^ ^-3 - — - - — - ' ^ — -1 •* — ^ ^ — • -— ' -^-^ ^^^~ •^— " — ^ -^_^ OOO !>. t-. QO -i* "•- "3 hi t-- 10 10 -H -H O t^- ^ CO LO O*^ ^O -f <^i ^ rsi r^a O O O wx ^^ ^ t-^. *~ OO ~" " ^^ r^i 2 f".~ t--' — ' JJ JJ JJ ^) Tl TI n 41-H-H - 41 4! 41 • -H -H 41 ^ 41 -H 41 41 41 41 a ~ •» O -f O ^O ^O C^J •^i ~^: —4 O >o oo oo r^- t— *^D ^^ ^^ CM ^ O -t O -H' rr}" ^ rs cs t-— t^- i^ I f^J '^•1 lO •— » ^-H •» s 1 | 1 •2i "5 o a 0 R t s ^"^ 'iH! e ^-k ^ ^w ^ 'C '"^ O °^ Jj< a a E u tJI UUU 5 UUU UUU UUU ,-s UUU UUU ->,ooo *J ooo ooo ^^ ooo _ ooo ooo 4-? ^OO rt 3OO ^- ^; 5O <^ OOO ^-^OOO ^>'~ * — « v £ E-^ ^, -2 ^S p^ "o w »- ^ "-^ 3 , ' — * — ^ »^ •— J ^^ «-J + ^^ ^— rt i 1 fc -SJ a Ejip^ $!^IE iJ-H "Ll^X cofl?-!lC "^^X ^ 3 k. 6 H U X S W'1" § a 2: !? -~ ° OSMOREGULATION IN THAIS HAEMASTOMA 151 New Multiple Range Test was used to find significance of differences among temperature means. A preliminary analysis of variance showed no sex effect in either the constant or fluctuating salinity conditions, so sex was not included in any further ANOVA's. Snail hemolymph osmolality and ion concentrations were com- pared with seawater values hy /-tests for each temperature-salinity combination. A significant difference (P < 0.05) would indicate regulation of hemolymph osmolality or ion concentrations as compared to seawater values. Statistical analysis: Salinity fluctuation experiments Data from the salinity fluctuation experiments were also standardized by expres- sing hemolymph osmolality and ion concentrations as hemolymph minus seawater concentrations. An ANOVA was run testing the effects of temperature, salinity regime, high versus low stage of fluctuation cycle, and time on percent tissue water, hemolymph osmolality, Cl~, Na+, Mg-+, and K+ concentrations of snails under fluctuating salinity regimes. For the ANOVAs time was broken into four classes : Days 1 and 3, 6 and 9, 12 and 15, and 18 and 21. This is a 3x2x2x4 factorial design. Dun- can's New Multiple Range Test was used to detect significant differences (P < 0.05) among means of the dependent variables over time and to compare the three temperature experiments. Maximum R-square regressions of all dependent variables against time were performed for each temperature-salinity-regime stage-of- cycle combination. Regression lines are plotted as predicted values with standard errors. All statistical analyses were done using the Statistical Analysis System, Version 76.6 (Barr ct a/., 1976). RESULTS Constant salinity Temperature, salinity, and temperature-salinity interaction all had significant effects (P < 0.05) on the difference between hemolymph and seawater osmolality. Mean values for snails held at 7.5, 10, 20, and 30#r salinity and 10°, 20°, and 30°C are given in Table I. Hemolymph osmolality increased as salinity increased. The /-test comparisons of hemolymph osmolality against seawater osmolality indi- cate a significant difference between hemolymph and seawater osmolality at 10, 20, and 30'A and 10°C (Table I). There are statistically significant differences at two salinities under both the 20° and 30° C regimes, but the differences are smaller. Hemolymph osmolality was maintained slightly above seawater values at 10°C and was isosmotic or slightly hyperionic at 20° and 30 °C. Percent tissue water was unaffected by temperature and decreased as ambient salinity and hemo- lymph osmolality increased (Table I). Hemolymph concentrations of all ions studied increased as salinity increased under steady state conditions. Temperature, salinity, and temperature-salinity interactions all had significant effect (P < 0.05) on the difference between hemo- lymph and seawater C\~, Na+, and Mg2+ concentrations (Table I). Hemolymph Cl~ concentrations were generally lower than seawater values. At 10°C hemo- lymph Na+ concentrations were greater than seawater concentrations, while hemo- lymph Na+ concentrations did not differ from ambient levels at 20° and 30°C. Hemolymph [Mg2+] was isionic or slightly hyperionic to ambient seawater at most temperature-salinity combinations examined. The /-test comparisons indi- 152 J. E. HILDRETH AND W. B. STICKLE IO°C 20°C 30°C a> 6 '5> ai h_ to 5 g = ID 0 ro 2 o 1 5 S 5 ! I - i I o I in E O K) 80 76 72 80 76 72 300 200 100 0 -too 300 200 100 0 -100 k,-H-t-r-H H "+ H H 13 69 12 15 18 21 I 3 6 9 12 15 18 21 day I 3 6 9 12 15 18 21 FIGURE 1. Regression lines for hemolymph osmolality of Tliais liacinastonia expressed as hemolymph osmolality minus seawater osmolality at the high (H) and low (L) phases of salinity cycles. Percent body water of the foot at high and low phases of salinity cycles are also given. Lines are presented for 10-30-10#f and 30-10-30^r S cycles at 10°, 20°, and 30°C. Vertical lines represent standard error at each sample point. cated significant differences between hemolymph and seawater Mg2+ concentrations at 7.5, 10, and 20#c S at both 10° and 20°C. Although neither temperature nor salinity had significant effects on K+ hemolymph-minus-seawater differences, the effect of temperature-salinity interactions was significant (P < 0.05). Hemolymph [K*| increased with increasing seawater concentration and was unaffected by temperature. The /-test comparisons for difference between seawater and hemo- lymph [K+| showed significant differences at 10 and 20% tl salinity. Fluctuating salinity experiments Hemolymph osmolality tracked that of ambient seawater at all three tempera- tures under both salinity regimes. Hemolymph tended to be isosmotic or slightly OSMOREGULATION IN THAIS HAEMASTOMA 153 hyposmotic during high salinity slackwater periods and to be hyperionic during low salinity slackwaters. This trend was most evident at 10°C. Figure 1 shows maximum R-square regression lines of hemolymph-minus-seawater osmolality for high and low salinity sample periods at 10°, 20°, and 30°C under both salinity regimes. Broken lines indicate nonsignificant, best-fit lines. Analysis of variance (ANOVA) showed temperature, salinity regime, stage of cycle (high or low), time, and all two way interactions of these variables to have significant (P < 0.05) effects on the difference of osmolality between hemo- lymph and seawater. Duncan's New Multple Range Test showed that hemolymph- minus-seawater osmolality was greatest at 10°C, intermediate at 20°C, and least at 30° C over the 21 -day salinity fluctuation experiment. Duncan's test also showed that hemolymph-minus-seawater osmolality was greater in snails under the 30-10-30/^ S fluctuating salinity regime than in snails under the 10-30-10%o S regime. A large positive difference at low salinity slackwaters, as was observed at 10°C and to a lesser extent at 20° and 30°C (Fig. 1), indicates snail hemolymph was not closely tracking ambient levels during this phase of the salinity cycle. Tissue water in the foot varied inversely with seawater and hemolymph osmo- lality, increasing with decreasing salinity (Fig. 1). ANOVA showed that tempera- ture, stage of cycle, and time had significant effects (P < 0.05) on tissue water content. Tissue water was greatest at 30° C, intermediate at 20° C, and least at 10°C according to Duncan's test. This difference in tissue water content was obvious at the time of sampling. Bleeding of the snails was more difficult at 10°C than at the other temperatures and at 30° C the foot hemolymph was evident even after blotting the severed foot. Temperature and stage of cycle, but not salinity regime, had significant (P < 0.05) effects of hemolymph Cl~ values expressed as hemolymph-minus-seawater concentrations when salinity was fluctuated. All two way interactions, including those involving salinity regime, were significant. Duncan's New Multiple Range Test showed hemolymph-minus-seawater values to be greatest at 10° C and least at 30° C. Hemolymph-minus-seawater |C1~] was greater than zero, indicating hemolymph Cl~ was maintained above seawater values at most low-salinity sample points at 10° and 20°C (Fig. 2). At 30°C hemolymph-minus-seawater Cl~ con- centration was close to zero at both high and low salinity sample points, i.e., hemo- lymph Cl" tracked ambient water concentrations closely at 30° C. Somewhat similar trends were observed for hemolymph-minus-seawater Na+ concentrations. Figure 2 shows that at 10° C hemolymph-minus-seawater Na+ concentration was consistently greater than zero during the low-salinity sample periods while there was little difference between hemolymph and seawater concen- trations at high-salinity sample points. The data were more variable at the higher temperatures but tend to show hemolymph Na' concentration tracking ambient levels more than at 10°C. ANOVA showed that temperature, salinity regime, stage of cycle, time, and all two way interactions involving these variables, except stage of cycle by time interaction, have significant effects on hemolymph-minus-seawater [Na+]. Duncan's test showed hemolymph-minus-seawater vfa* to be greatest at 10°C and least at 30° C and to be greater in snails under the 30-10-307™ salinity regime (30^ S acclimation) than in those under the 1 0-30-1 0%o > regime. Hemolymph Mg-+ concentrations showed trends similar to those observed for the other ions studied. Hemolymph [ Mg-'4 ] was hyperionic during low-salinity sample periods and usually isionic or slightly hypoionic during high-salinity sample periods (Fig. 3). Temperature, stage of cycle, time, and all two way interactions 154 J. E. HILDRETH AND W. B. STICKLE IO»C 20"C 30»C 200 e o i # § 2 • i s? e o -200 200 o 5 4 2 ^ 2 o • 2 9 >° I 2 | t I * V) O a z E -200 100 0 -100 -200 100 0 -100 -200 H H I 3 6 9 12 15 18 21 I 3 6 9 12 15 16 21 I 3 6 9 12 15 18 21 day FIGURE 2. Thais haemastoma. Regression lines for hemolymph Cl~ and Na* concentration of Thais haemastoma expressed as hemolymph-minus-seawater concentration at the high (H) and low (L) phases of salinity cycles. Lines are presented for 10-30-10#r and 30-10-30&. S cycles at 10°, 20°, and 30°C. Vertical lines represent standard error at each sample point. tested except stage of cycle by time had significant effects (P < 0.05) on hemo- lymph-minus-seawater Mg'J1. Duncan's test showed that hemolymph-minus- seawater |Mg2+] was greater at 10°C than at 20° and 30°C. Hemolymph [K+] also tended to be isionic to ambient seawater during high- salinity sample periods and to be hyperionic at low-salinity sample points (Fig. 3). Hemolymph K+ was less hyperionic at low-salinity sample points at 30 °C than at the other temperatures, indicating it tracked ambient levels to a greater extent. Temperature, salinity regime, and stage of cycle all significantly affected (P < 0.05) hemolymph Iv expressed as hemolymph-minus-seawater concentration, OSMOREGULATION IN THAIS HAEMASTOMA 155 IO°C 20°C 30°C o * S M w 8 20 10 o « 2 E o > -10 20 • j 8 ' 2 0 10 NO . "0 f 8 f 6 » 6 S •/> 4 JF 2 jf 1 L I 3 6 9 12 15 18 21 136 9 12 15 18 21 13 6 9 12 15 18 21 day FIGURE 3. Regression lines for hemolymph Mg2+ and K+ concentrations of Thais haema- stonia expressed as hemolymph-minus-seawater concentration at the high (H) and low (L) phases of salinity cycles. Lines are presented for 10-30-10^r and 30-10-30/4'r S cycles at 10°, 20°, and 30°C. Vertical lines represent standard error at each sample point. although no interactions were significant. Duncan's test showed hemolymph minus seawater [K+] was greatest at 10° C and least at 30° C. Ninh \'driu-positiz'c substances Ninhydrin-positive-substances (NFS) concentration of the foot, calculated as /iM NPS/g dry weight and as ^M NPS/g tissue water, are presented in Table II. Data are shown for 15 steady-state temperature-salinity combinations. NFS con- tent of the foot expressed as either NPS/g dry weight or \TPS/g tissue water increases with increasing salinity at all three temperatures studied. NFS content of the foot tends to be lower at 10°C than at the other temperatures. No other clear trends are evident. NPS/g dry weights were not significantly different from control (day 0) values at high- and low-salinity sample points on the first day of salinity fluctuations (Fig. 4). Snails acclimated to 30^r S had higher NFS levels in the foot even at 156 1. E. HILDRETH AND W. B. STICKLE TABLE II Ninhydrin- positive substances (NFS) in the foot of Thais haemastoma acclimated to 15 temperature- salinity is expressed as both y.M NPS/g dry weight and as \j.M NPS/g tissue ivattr. Values given are X ± 05% confidence limit (n). ND indicates no data available. 7.5 %c 10%c 20%c 30 %o 35%c 10°C /jM g dry weight A.M g tissue HiO 87.0 ± 19.4 (6) 23.6 ± 4.0 (6) 144.7 ± 3.3 (5) 40.9 ± 8.0 (4) 183.6 ± 10.5 (4) 94.3 ± 7.6 (4) 268.2 ± 31.6 (4) 94.3 ± 7.6 (4) ND ND 20°C ^M,g dry weight /jM g tissue HjO 124.3 ± 1.7 (5) 37.3 ± 2.2 (5) 129.0 ± 14.5 (5) 41.0 ± 5.2 (5) 211.0 ± 7.3 (5) 66.8 ± 9.4 (4) 312.0 ±34.5 (4) 96.5 ± 18.7 (4) 338.1 ±25.3 (1) 117.37 ± 10.4 (4) 30°C /iM/g dry weight /iM ,'g tissue H :O 115.4 ± 3.4 (4) 34.4 ± 3.2 (4) 159.9 ± 9.9 (5) 48.7 ± 4.5 (4) 217.64 ± 45.4 (3) 68.1 ± 4.2 (3) 244.6 ± 51.3 (3) 93.6 ± 5.8 (3) 241.19 ± 24.7 (3) 84.9 ± 2.0 (3) corresponding points of the salinity fluctuation cycle, that is, when both were measured at 10 or 30/^r S. NPS/g dry weight of snails acclimated to 30/£c and subjected to salinity cycles of 30— 10— 30/cf S did not change significantly over the 3-\veek experiment at 10°, 20°, or 30°C. In contrast, snails acclimated to lQ%c S and subjected to salinity cycles of 1 0-30-1 0(/cc exhibited significantly higher X PS/g dry weight values than control and day 1 values after 2 weeks and remained higher after 3 weeks of cycling salinity. There was little significant cycling of XPS levels, expressed as /*M NPS/g dry weight, with the daily salinity fluctuations. Similar trends were observed when these data were analyzed expressing NPS levels as /u.M NPS/g tissue water except that NPS levels cycled with fluctuating salinity to a greater extent, tending to be higher during high salinity slackwater sample periods (Fig. 4). DISCUSSION Although little difference existed between hemolymph and seawater osmolality or ion concentrations in Thais haemastoma at constant salinities and temperatures studied, temperature affected hemolymph values over a range of salinities. The differences between hemolymph and seawater osmolality and Cl~ and Na+ concen- trations were greatest at 10°C. Mg2+ hemolymph-minus-seawater concentration was less at 10° and 30°C than at 20°C, while hemolymph K+ levels were not affected by temperature under steady-state conditions. Pierce (1970) argued that the slight hyperosmotic concentrations of blood and pericardial fluids relative to sea water in Modiolus was due to a Gibbs-Donnan equilibrium. However, Man- gum and Johansen (1975) have shown that a Gibbs-Donnan equilibrium is not responsible for the frequently observed hyperosmoticity of body fluids relative to ambient water in osmoconforming marine invertebrates. Therefore, mechanisms responsible for T. haemastoma maintenance of hemolymph osmolality, Na+, Mg2+, K , and Cl" levels at concentrations significantly different from that of the ambient water remain to be elucidated. When salinity was cycled, hemolymph osomolality and ion concentrations were maintained above seawater values during the low phase of salinity cycles and were usually isosmotic during the high phase at 10°, 20°, and 30°C. Less cycling of hemolymph ion concentrations with salinity cycles occurred at 10°C than at 30°C, especially of hemolymph MgL>t and K+ concentrations, but there were still significant hemolymph changes during most cycles. Similar results have been reported for OSMOREGULATION IN THAIS HARMASTOMA IO°C 20°C 30°C 157 120 r i 1 CO •~) ' • »- ' 90 *** o o ~<" Y i * t •• 9 o 60 AC) 9 6 O 0» 1 Q V nb ' "•: 5 30 V w v> 120 * £ 90 2 *° a i 60 ^ / } * ^T ' ** + o o S 30 03 9 15 21 03 9 15 21 03 9 15 21 I 300 P } 1 9 i| T S V i T CO ~ a» 200 1 ? ^^Tr f t j *» 1 V 1^ 0 f s 100 • ' t t >s ' 1 300 ~K W • M g 200 2 i** f^ i cycles at 10°, 20°, and 30°C. Open circles represent high salinity sample points; closed circles, low salinity points. Lack of error bars indicates confidence limits smaller than the point on the figure. T. hacmastoma exposed to 14 days of a 2G-10-20/£C S diurnal pattern of fluctuat- ing salinity at 15°-21°C. (Stickle and Howey, 1975). The observed hemolymph-to-seawater ion differences during salinity fluctuation in Thais haemastoiua are probably passive. Hemolymph was isionic during the high phase of salinity cycles and hyperionic during the low phase, suggesting greater solute and solvent movement during periods of increasing salinity. Extracellular fluid of other osmoconforming molluscs (Stickle and Ahokas. 1975; Stickle and Howey, 1975; Hand and Stickle, 1977, Shumway, 1977) and echinoderms (Stickle and Ahokas, 1974) may also remain hypersomotic during the low phase of salinity 158 J- E. HILDRETH AND W. B. STICKLE cycles. If the hemolymph-to-seawater difference in ion concentration during the low phase of salinity cycles at 10° C were caused by metabolic-dependent activities, regulation would have been observed under steady-state conditions also. Mag- nesium was the only hemolymph ion maintained at levels significantly different from seawater values under steady-state conditions. All other hemolymph ions were not different from seawater values even after 5 weeks acclimation to 10° C, suggesting low-temperature changes in membrane permeability or changes in other passive activities during salinity fluctuations. Prusch and Hall (1978) found that Mytiliis editlis is capable of altering tissue water permeability with changes in osmolality of the ambient water. Similar permeability changes may be occurring in response to temperature and fluctuating salinity in T. haemastoma. Oxygen consumption in T. haemastoma exposed to single salinity cycles at 20° C decreases in the middle of 10-15-1 0^e, 30-10-30%f and 10-30-10%, S diurnal cycles (Findley et al, 1978), suggesting a passive mechanism is responsible for the differences between hemo- lymph and seawater during the low phase of salinity cycles. Thais haemastoma maintains higher hemolymph osmolality at lower tempera- tures under both steady-state and fluctuating salinity conditions. Hemolymph osmolality and ion concentrations are regulated above seawater values during the low phase of salinity cycles to a much greater extent at 10°C than at 30°C. Thus it seems T. haemastoma is better able to withstand dilute and widely fluctuat- ing salinities at low temperatures. Lowest field salinities in Barataria Bay, the collection site for experimental animals, occur during winter (Hewatt, 1951). T. haemastoma may be able to compensate for low environmental salinities by pas- sively regulating hemolymph composition when natural tidal salinity fluctuations occur most frequently. Since the feeding threshold of T. haemastoma is about 12.5 °C (Carton and Stickle, 1980) it cannot be concluded that T. haemastoma would be able to permanently colonize dilute waters at cold temperatures. Indeed it is generally considered to be a warm water species extending only as far north as North Carolina. Ninhydrin-positive substances (NFS) increased in foot tissue of T. haema- stoma with increasing salinity and were lower at 10°C than at other experimental temperatures under steady-state salinity conditions. Ninhydrin-positive sub- stances are principally composed of free amino acids which are important in cell volume regulation in response to salinity variation. Peterson and Duerr (1969) found the free amino acid levels of Tcgula jnncbralis maintained for 11 days at 50, 100, 120, and 160% seawater (35%/S = 100%) to increase linearly from 50 to 120% seawater and then to decline to approximately 100% seawater concentra- tion at 160% seawater when based on grams/wet-weight. Modiolus hearts (Pierce and Greenberg, 1973) and intact Mytiliis edulis (Livingstone et al., 1979) transferred directly from high to low salinity release free amino acids from the intracellular fluid compartment. Free amino acids increase in acclimation of bivalve molluscs to high salinity (Baginski and Pierce, 1977, Gainey, 1978). Ex- pressing XPS concentration in the foot of T. haemastoma on a dry weight basis indicates synthesis or degradation of NPS, principally free amino acids, whereas expressing NPS concentration on a tissue water basis provides information on the osmotic effectiveness of free amino acids. Extracellular fluid concentrations of NPS are small in comparison with entire foot NPS, suggesting that most NPS are intracellular. Stickle and Howey (1975) found hemolymph NPS concentrations to range from 0.6 to 6.4 mM in snails acclimated to 10-30/{< S. Foot tissue water NPS concentrations range from 41 to 117 mM between 10 and 30%e S. OSMOREGULATION IN THAIS IL-Ui.M.lSTOMA 159 Variation in the volume of the intracellular fluid compartment without a concomitant change in absolute XPS content will alter the osmotic importance of NFS. Staalancl ( 1970) found the intracellular fluid compartment of Biic- i'initni nndatnm to vary inversely with salinity between 10 and 33/£o whereas no change occurred in the size of the extracellular fluid compartment. If a similar relationship exists in the foot of T. haemastpma the osmotic effective- ness of NFS as intracellular osmotically active particles will be further reduced by dilution of the intracellular fluid compartment at low salinity. Under fluctuating salinity conditions, it was more difficult to obtain hemolymph at 10°C than at 20°C while hemolymph was most abundant at 30°C. Stickle and Howey (1975) found similar results with T. liaanastoina exposed to two weeks of diurnal 20-10-20'n patterns of fluctuating salinity. Total body water in the foot of stepwise acclimated T. haemastoma was not significantly different between 10° and 30° C, which suggests that the intracellular fluid compartment was probably larger at 10°C than at 20° or 30°C. Ninhydrin positive substances did not cycle in the foot of T. haemastoma in response to cycling salinities, although NFS levels increased with time in snails acclimated to IQ'/rt S and subjected to a long-term 1 0-30-1 O^c regime. In con- trast, all eight bivalve species studied by Shumway ct al. (1977) cycled adductor muscle NFS levels in response to 1-day salinity fluctuations. After 1 week of salinity fluctuation, no changes in Mytilns cdnlis tissue NFS levels were observed during a salinity cycle. Additionally, tissue NFS did not decline below the con- centration observed in 100^ seawater. On the other hand, Livingstone ct al.. (1979) found changes in Al . cdnlis NFS level to be small during short-term salinity fluctuations (30-15-30'/r cycles) but to decline markedly after exposure to 24 days of salinity fluctuations. These authors believe that the lack of NFS decline during fluctuating salinity observed in the study of Shumway ct al.. (1977) may have been due to a combination of valve closure during the low-salinity phase of each cycle and salinity effects. The mussels used by Livingstone ct al., (1979) remained open throughout salinity cycles and continued to respire. Like- wise, the T. haemastoma used in our fluctuating-salinity experiments remained attached to aquarium walls and crawled and preyed on oysters, although less activity was observed at 10° C than at other temperatures studied. Carton and Stickle (1980) have documented the feeding activity of oyster drills on oysters under identical conditions and Findley ct al., (1978) have shown aerobic respiration to occur under fluctuating salinity conditions. Thais haemastoma passively regulates hemolymph osmolality and ionic com- position slightly above environmental levels at low temperatures and constant salinity, as well as during the daily low phase of long-term salinity fluctuation cycles. Since cell fluid volume appears to be greater at low temperatures and low salinities, NFS may be less important in osmotic regulation and cell volume regulation under conditions of low environmental temperature and salinity. This research was partially supported by a PRECOL Research Fellowship to J.E.H. during the summer and was done as partial fulfillment of the require- ments for the Master of Science Degree. Tom Sabourin, Dave Carton, Preston Newman and Barbara Belisle provided technical assistance during this project. The authors wish to acknowledge with thanks the critical review of this manuscript 160 J- E. HILDRETH AND W. B. STICKLE offered by Dr. Thomas Dietz, Dr. John Fleegcr, Dr. Earl Weidner, and Mr. David Saintsing. SUMMARY 1. When Thais Iiaeiiiastoina were acclimated stepwise to constant salinity; hemolymph osmolality, Na+, Mg2+, K+, and Cl~ concentrations as well as tissue > concentration increased with salinity between 7.5 and 30/£f at 10°, 20°, and 30° C. Hemolymph osmolality was significantly greater than that of seawater at 10, 20, 3Q%c S and 10°C. Per cent tissue water was unaffected by temperature and decreased with increasing salinity and hemolymph osmolality. The hemolymph con- centration of all ions increased with increasing concentrations in seawater. Hemo- lymph [Cl~] was lower than in seawater at 20° and 30° C but hyperionic to seawater at 20 and 30'/< S and 10°C. Hemolymph [Na+J was greater than in seawater at 10°C but no difference existed at 20° and 30°C. Hemolymph [Mg-+] was significantly greater than in seawater at 10 and 20%c S at all temperatures. Neither temperature nor salinity significantly affected the hemolymph-to-seawater difference in [K+]. [NPSj in the foot of T. hacinastoina increased with salinity and was lower at 10°C than at 20° and 30 °C. 2. Hemolymph osmolality tracked seawater osmolality less well at 10°C than at 20° and 30° during 3Q-10-30#C and 1 0-30-1 0#, diurnal patterns of fluctuating salinity. Reduced tracking at 10° C was due to less decline of hemolymph osmolality during the decreasing-salinity phase of the cycles. Snail hemolymph osmolality also declined less during the decreasing-salinity phase of the 30—10— 30%c S cycle than during the 10-30-10/^ S cycles, indicating that acclimation to different constant salinities prior to the initiation of fluctuating salinity cycles has a sig- nificant effect on the response of T. hacuiastoina to fluctuating salinity. Tissue water in the foot varied inversely with seawater and hemolymph osmolality during fluctuating salinity. The difference in percent tissue water between snails acclimated to 10 and 30%0 S prior to salinity fluctuation was greatest at 30° C, intermediate at 20 °C, and least at 10°C. Hemolymph [Na+], [Mg2+], [K+], and [C1-], tracked ambient salinity least well at 10°C and most closely at 30°C. Passive factors are likely to be responsible for the differences in hemolymph osmotic and ionic fluctuation observed at 10°, 20°, and 30°C. Tissue NFS did not fluctuate with daily salinity cycles in either the 30-10-30%r or 1 0-30-1 0%o diurnal patterns of fluctuating salinity. Foot tissue NPS of snails acclimated to 30%c did not change over three weeks of exposure to a 30-1 0-30/^, S pattern of fluctuating salinity but XPS in the foot of snails acclimated to 10'/c S and subjected to a 1 0-30-1 O^r- S pattern of fluctuating salinity increased significantly in 15 days and remained significantly higher on Day 21. LITERATURE CITED BAGINSKI, R. A., AND S. K. PIERCE, 1977. The time course of intracellular free amino acid accumulation in tissues of Modiolus dcniissits during high salinity adaptaton. Coinp. Biochcm. I'hyswl.. 57 : 407-412. BARR, A. J., J. H. GOODKNIGHT, J. P. SALL, AND J. T. HELWIG, 1976. A User's Guide to SAS 76. Raleigh, North Carolina : SAS Institute Inc. DEHNEL, P. A., 19(>2 Aspects of osmoregulation in two species of intertidal crahs. Biol. Bull., 122: 208-227. DEHNEL, P. A., AND T. H. CAKEHIOT, 1965. Ion regulation in two species of intertidal crahs. Comp. Biochcm. I'hyxiol., 15: 377-397. OSMOREGULATION IN TIL-US Il.U-.MASTOMA 161 FINDI.KY, A. M., B. W. BKI.ISI.K, AND W. B. STICKLE, 1978. Effects of salinity fluctuations on the respiration rate of the Southern Oyster Drill Thais liacmastoma and tin- Blue Crab Callinectes sapidiis. Mc.r. Biol.. 49: 59-67. GAINEY, L. F., JR., 1978. The response of the Corhiculidae (Mollusca: Bivalvia) to osmotic stress: the cellular response. I'hysiol. Zool.. 51: 79-91. CARTON, D., AND W. B. STICKLE, 1980. Effects of salinity and temperature on the predation rate of Thais huemastoiini on Crassostrca Tire/mica spat. Biol. Hull., 158 : 49-57. HAND, S. C., AND W. B. STICKLE, 1977. Effects of tidal fluctuations of salinity on the peri- cardial fluid composition of the .American Oyster Crassostrca virtiitiica. Mar. Biol., 33: 309-322. HKWATT, W. G., 1951. Salinity studies in Louisiana coastal embayments west of the Mississippi River. 32 pp. Final Report of Project Nine. Texas A & M. Research Foundation College Station, Texas. KINNE, O., 1970. Temperature: Invertebrates. Pp. 407-514 in O. Kinne, Ed., Marine Ecology, Vol. 1. Environmental Factors. Part 1. London: Wiley-Interscience. KINNE, O., 1971. Salinity: Invertebrates. Pp. 821-996 in O. Kinne, Ed., Marine Ecoloyy Vol. 1. Environmental Factors. Part II. London: Wiley-Interscience. LIVINGSTONE, D. R., J. WIDDONVS, AND P. FIETH, 1979. Aspects of nitrogen metabolism of the common mussel Mytilus cdulis: adaptation to abrupt and fluctuating changes in sa- linity. Mar. Biol., 53 : 41-56. LOCKWOOD, A. P. M., 1960. Some effects of temperature and salinity of the medium on the ionic regulation of the isopod Asellns aquaticus (L.). J. Exp. Biol.. 37: 614-630. MANGUM, C. P., AND K. JOHANSEN, 1975. The colloid osmotic pressures of invertebrate body fluids. /. Exp. Biol.. 63 : 661-671. PETERSON, M. B., AND T. G. DUERR, 1969. Studies on osmotic adjustment in Teyula funebralis (Adams, 1954). Comp. Biochem. Physio!., 28: 633-644. PIERCE, S. K., JR., 1970. The water balance of Modiolus (Mollusca: Bivalvia): osmotic concentrations in changing salinities. Comp. Biochem. Pliysiol.. 36: 521-533. PIERCE, S. K., AND M. J. GREENBERG, 1973. The initiation and control of free amino acid regulation of cell volume in salinity stressed marine bivalves. /. Exp. Biol., 59 : 435-440. PRUSCH, R. D., AND C. HALL, 1978. Diffusional water permeability in selected marine bivalves. Biol. Bull., 154: 292-301. ROSEN, H., 1957. A modified ninhydrin colorimetric analysis for amino acids. Arch. Biochem. Biophys., 67: 10-15. SHUMWAY, S. E., 1977. The effect of fluctuating salinity on the tissue water content of eight species of bivalve molluscs. J. Comp. Pliysiol., 116: 269-285. SHUMWAY, S. E., P. A. GABBOT, AND A. YOUNGSON, 1977. The effect of fluctuating salinity on the concentrations of free amino acids and ninhydrin positive substances in the adductor muscles of eight species of bivalve molluscs. /. E.rp. Mar. Biol. Ecol.. 29: 131-151. STAALAND, H., 1970. Volume regulation in the common whelk, Bucciinun undatum L. Comp. Biochem. Physio!., 34 : 355-365. STICKLE, W. B., AND R. AHOKAS, 1974. The effects of tidal fluctuation of salinity on the pericardial fluid composition of several echinoderms. Comp. Biochem. Physiol., 47 : 469-476. STICKLE, W. B., AND R. AHOKAS, 1975. The effects of tidal fluctuation of salinity on the hemolymph composition of several molluscs. Comp. Biochem. Physiol., 50: 291-296. STICKLE, W. B., AND T. W. HOWEY, 1975. Effects of tidal fluctuations of salinity on hemo- lymph composition of the Southern Oyster Drill Thais liacmastoma. 309-322. TAYLOR, E. W., P. J. BUTLER, AND A. AL-WASSIA, 1977. The effect of a decrease in salinity on respiration, osmoregulation, and activity in the shore crab. Car emus maenits (L.) at different acclimation temperatures. /. Comp. Physiol., 119 TODD, M. E., 1963. Osmoregulation in Li || *• 1 1 ! .... II u Fici'KE 1. Semi-recirculnting seawater system. A, B, C, culture trays ; F, 25 and 5 /* canister filters; M, makeup water treatment box; S, S', compartments of sump; U, 50 W UV water treatment unit. Details are : a, 4 1 mixing trough with perforated distributor ; b, b', b", pairs of 4 1 mixing troughs joined by siphon manifolds; c, adjustable dam and return line ; d, d', airstones ; e, e', baffles ; f, degassing screens ; g, pump ; h, heaters ; m, untreated, and m', treated makeup water input ; o, o', overflow standpipes. Circled numbers are sampling points for water quality analyses; uncircled numbers indicate rows. Experimental subjects and protocol Two age classes of juveniles (herein called older and younger) were used. "Older" individuals were hatched from three wild-caught Homants aincncaniis females, reared in larval systems (Hand ct al., 1977) until metamorphosis, and then maintained in individual compartments at ambient temperatures (ca. 14°C) until the beginning of the experiment, at which time virtually all were nearing the 5th-to-6th-stage molt. On Julian Calendar Day 288/78, equal numbers from each of the three females (288 in all) were randomly assigned to the spaces in rows 5-10 of each tray (numbering from the influent end). "Younger" juveniles, from a single laboratory mating of a male H. aincricaniis to a female H. gaminarus, were hatched on Day 270/78 ± 3 days and reared at 19° C until Day 288/78, at which time all were in the 4th (1st post-larval) stage. (Pure H. aincricanus individuals of appropriate age and numbers were unavailable. We are confident, however, that using hybrids did not materially affect the results reported.) A total of 432 "younger" individuals were then randomly assigned to the remaining nine rows in each tray. Backup animals for each family were maintained in another semi- recirculating seawater table at 21 °C. Within the first 7 days, 37 lobsterlings died in System I, and ' > in System II. Deaths thereafter were negligible except in rows B 10-12 of System II, where warping of the plastic tray bottom allowed cannibalization of a few molted animals. All dead lobsters were replaced and data from the replacements were used in tin- analysis. Throughout most of the experiment animals were maintained on a 7 hr light : 17 hr dark photoperiod, with an additional .] hr of light approximately 6 hr into the dark period for a nightly check and molt collection. Lighting was from four standard 40 W clear white fluorescent tubes fixed 1.5 m above the culture trays and covered with two layers of green cellophane to reduce algal growth. Molted 164 NELSON ET AL. carapaces were collected twice daily and their lengths recorded when possible, before they were returned to the proper compartments. Animals were fed once daily ad libitum with adult brine shrimp and compartments were siphoned daily before feeding. Mixing troughs and sum}) compartments were siphoned as needed. Cartridge filters were replaced weekly. The experiment was terminated after 93 days, at which time weights and cara- pace lengths of all animals were recorded. The hybrid animals at this time were approximately 1 1 1 days old and the specimens of H. anicrtcanus approximately 1 month older. Only carapace length (measured from the rear edge of the orbit) and corresponding results will be reported in this paper since weight results are highly correlated with length results (Hedgecock ct a!.. 1976; Hedgecock and X el son, 1978). Chemical and physical analyses Temperature, recorded thrice daily, was maintained at 21° and 20.5°C, respec- tively, at points 1 and 4 (see Fig. 1 ) with little variation between systems or from day to day. Dissolved oxygen, pH, and ammonia were measured daily at points 1 and 4. and weekly at points 2, 3, 5, and 6. Nitrate, nitrite, and organic nitrogen were analyzed weekly at sample locations 1-6. Methods of analysis for nitrogen compounds are given in Daggett and Aronstein (1979). Non-filterable residue > 0.45 /Ain was measured weekly at points 1, 4, and 5. Bacterial counts were made according to procedures in Standard Methods (1976) on samples taken weekly from points 1 and 4. The systems were designed to create gradients of long-lived metabolite con- centration in each system, from the influent mixing trough (sampling point 1, Fig. 1) to the effluent mixing trough (sampling point 4). By using different make-up rates but the same table flow rate (2 1/min ) in the two systems, the fraction of new, non-recirculated sea water entering the table was controlled so that the gradient of long-lived metabolite concentration in System II, while no steeper, began at a threefold higher concentration. For this, the fraction of the water entering the table that was new make-up water was maintained at 50% in System I and 25f/f in System II. These fractions were used so that the concentrations of long-lived metabolites such as ammonia at points 1 and 4 in System I and points 1 and 4 in System II would be maintained in the approximate proportions 1:2:3:4, respectively, assuming equal rates of metabolite buildup in the two systems but irrespective of their actual magnitude (see compartmental analyses below). Terminal gradients from point 1 to point 4 in both systems were 50-75 /xg nitrogen/1, occasionally higher. Concentrations of ammonia in System II were always considerably higher than in System I, as anticipated ; the animals in the final rows of System II were occasionally exposed to concen- trations greater than 500 //.g nitrogen/1 ( NH;( + NH4+). These levels are never- theless very much lower than the "safe" levels suggested by Delistrary ct al. (1977) for 4th-stage lobsters. Sampling at points 5 and 6 showed a small (< 10 /xg nitrogen/1), inconsistently positive or negative effect on ammonia concen- tration by cartridge filtration, and a slight positive effect ( < 2 /Ag/1) of UV treatment. Oxygen concentration, restored to ambient values by aeration in the sump, was used as a model of the behavior of concentration of long-lived metabolites removed by filtration or aeration (see compartmental analyses below). Oxygen concentra- GROWTH INHIBITION IN HOMARUS 165 tion gradients (AOL. ) from points 1 to 4 were similar in the two systems. Only once, following pump failure, was a AOo steeper than —2 nig Oo/i observed. Terminal AO^'s averaged about -1 nig O;> 1. Weekly samples at points 1, 5, and 6 were near saturation and the gradient across points 1-4 was linear. A small negative gradient in pH was maintained in both systems (minimum —0.2 units, from 8.1 to 7.9 at points 1 and 4, respectively; generally much closer to zero). NOu. XOa, organic nitrogen, and non-filterable residue concentrations displayed no consistent gradient in either system, and very small differences from ambient values. Bacterial counts' gradients were similar in the two systems, with values of <10/ml at point 1 and highly variable values at point 4. The means of the natural logarithms of the counts at point 4 on Tables I and II, respectively, were 7.886 ± 0.416 (s.e.) and 7.979 ± 0.518. There was no evident correlation of bacterial counts with other water quality variables, except for a general increase with time. Following the experiment, logs of the absorbances of water samples taken 1, 2, 4, 8, and 16 min after methylene blue injection into 10 individual com- partments were separately regressed on time and the slopes averaged to obtain an estimate of the first-order compartment-turnover rate, k — 0.631 ± 0.033 per min. As the average compartment contained 0.25 1 and the compartment flow rate was 0.25 1 per min, the expected k is 1.0 per min, and mixing was thus not complete. The typical flow pattern within a compartment contained one or two gyres with vertical axes. Nevertheless, we will use a linear model with k — 0.631 per min in the ensuing analyses (see compartmental analyses below). In this dye test, virtually no color was observed to flow into neighboring upstream compart- ments, although this type of flow could have occurred during the experiment proper. Experimental design The block of older animals in the middle of each tray necessitates treating the design as three separate experiments (2 systems X 3 trays X k rows, with k equal to 4, 6, and 5, respectively, for the first 4 rows of younger animals, the next 6 rows of older individuals, and the last 5 rows of younger animals). While physically the experimental design is fully nested, the experimental variable is position (along the gradient from point 1 in System I to point 4 in System II) ; and we are interested in separately examining the interactions among system, tray, and row. Therefore, we treat the experiment as three three-factor, fixed-effect, fac- torial designs with eight replicates (Table I). Compartmental analyses A simple generalized catenary compartmental model (Atkins, 1969) for the concentration a-}(t) of a particular metabolite in the j-th compartment at time / will be of use in discussing both the performance of the system and the growth difference results : j/ \ i = j - 1 d((i;) ri , l'i L j, 173 Cn -~ = - + fcji — fli -- &kjflj -- feojflj, J : dt V j V; k : : j _|_ 1 where r> is the rate of production (consumption, removal) of substance in units of mass per unit time, ?'j is the volume in liters of compartment j, k^ is the turnover rate (rate constant) from the preceding compartment i, and k,tj represents the 166 NELSON ET AL. rate constant of loss (or gain, + A'joa0) of the substance from the compartment to (or from) outside the system; all A-'s are in units of reciprocal time. For our purposes all A''s will be assumed to be constants, although k0i niay be a function of the difference between flj and cr0, the concentration outside the system. In the case of oxygen utilization the r/s are negative and if we assume no addition to the compartment at the air-water interface A>J(, is zero and the right-hand term drops out. If we assume the z-'s and k's to be equal and that the water suffers no change in the mixing troughs, equation (1) has the steady-state solution a i ---- an (2) 1=1 vk with fl0 the ambient concentration of oxygen, t- = 0.25 1, and k experimentally determined as 0.631 per minute. The average weight of a lobster near the end of the experiment was somewhat less than 2 g. Assuming a 2 g weight, the oxygen concentration gradient AOL. from sampling point 1 to sampling point 4 (i.e., a4r, minus a0 ) becomes —1.5 mg Ou/1, using the O2 consumption value of 5.25 ^g per lobster per min obtained from the regression of Logan and Epifanio (1978). Our experimental values are very close to this. The gradients in O^ concentration were approximately linear ; given the flow rates and surface-volume relationships involved, it seems safe to assume that the rate constant A'(lj or kjo for loss or gain of "volatiles" at the air-water interface were not large compared with the k^s. For ammonia (NH3 + NH4+) concentration, assuming that there is loss only at the overflow o' and that makeup water concentration is negliglible, the steady state equation (2) becomes 45 „ 3 v a,= (R--l)£^+£-^ (3) ,=, vk 1=1 v k R, the recirculation multiplier, is the ratio of the flow rate through the table to the makeup flow rate into the sump, i.e., the reciprocal of the makeup fraction. (This R is not to be confused with the R in a similar model of Liao and Mayo, 1972, which they define as "% of water reused." See also Delistrary ct al., 1977.) Using an (NH3 + NH4+) production value of 0.4 mg per lobster per day for a 2-g lobster (quoted by Moffett and Fisher, 1978, from unpublished work) we obtain a gradient from point 1 to point 4 of 79 /xg/1, again consistent with our observed values. With the makeup fractions used, of course, equation (3) predicts the 1:2:3:4 ratio for a0 and a4r, in the two systems. Considering only the downstream effect of a short-lived inhibitor, the appropriate compartmental model is equation (1) with r/s greater for the older than for the younger animals (because of their larger size). We are interested only in the difference between the older and younger animals' rates of production, and therefore for simplicity we will consider the hypothetical growth-inhibiting substance as originating exclusively with the older ones and only in the last row of their block. Then the steady-state solution for the compartments downstream from the block is «•> = ri' J = 1-2,3... (4) where k\ is the rate of loss or decay of the substance, A'u is the turnover rate to the next compartment, r} is the excess production rate by the older animals and v is the compartment volume. GROWTH INHIBITION IN HOMARUS 167 For A-, " A-j = 0.631/min and r, r = 1 g per 1 per min we have plotted Un- expected steady-state concentrations ( «j ) in the rows downstream from the older animal row ("row 1") in Figure 4B. The downstream effect of pulsed production of an inhibitory substance, with the pulses far apart in time, may be simulated by equation (1 ) with the first and last terms eliminated, and with a unit impulse function as initial condition. There are repeated roots and the solutions are of the form \Yith compartment sizes and turnover rates assumed to be equal in the different compartments, the time integrals of concentration in the different compartments will be the same. One may thus equate the time integrals of the solutions (5) for each j and solve for the Aj's in terms of the initial concentration pulse At}. Equation ( 5 ) then becomes (j - 1)! which for A{} = 1 g/1 and k — 0.631 min is plotted in Figure 4A. (6) o O o c 14 t 12 c. 9) 0) u S. 0 o i 13 II B. ^4 v A \ A^ . Vv J L 4 5 10 II 15 row FIGURE 2. Average final carapace lengths of animals (ordinates) plotted against position in a tray (abscissa). Rows 5-10 contain the older Honuinis amcricanus lobsters. In A, data from corresponding rows of the two systems are pooled. Open circles, trays A ; half-filled circles, trays B ; filled circles, trays C. In B, data from corresponding rows in trays A, B, and C are pooled. Open circles, System I; closed circles. System II. 168 NELSON ET AL. RESULTS Final carapace lent/tit difference Table I gives results of ANOVAs for the three "experiments." Factor 1 ( System ) never reaches a significant F-ratio. Figure 2B displays graphically the mean carapace-length values for each row in each system, lumping over trays ; it may he seen that there is indeed no difference between Systems I and II in this respect. Factor 2 (tray position) reaches significance (P < 0.001) only in the case of the first four rows of younger animals on each tray (Table I A) ; Figure 2 A demon- strates that this is a result of a difference between the first three rows of tray A and those on trays B and C. Figure 2A illustrates also the similarity between trays for the older individuals and for the last five rows of younger animals. The F-ratio for Factor 3 (row within tray) is highly significant for the last 5 rows of younger animals (Table 1C), and significant at the P < 0.02 level for the rows of older individuals (Table IB). Figures 2A and B illustrate a consistent depression of final carapace length of greater than 15% in younger animals imme- diately downstream from the older individuals (rows 11-12), relative to the carapace lengths of those younger animals farther downsteam in rows 14-15 and also relative to those in rows 1-3 of the following tray. The corresponding depression in weight is about 40%. The second significant "row" effect is a result of a depression in older individuals' carapace lengths in the seventh, eighth, and ninth rows, relative to the fifth and tenth rows, in both systems (Fig. 2B) and at all three tray positions (Fig. 2A). Finally, when the first four rows of the last two trays only were subjected to ANOVA, the row variable just missed significance (P < 0.064). There is a suggestion in Figure 2A of depressed growth in younger animals of row 4 of tray C (just upstream from the block of older individuals). We wished to determine the average rate at which the block of older animals' negative influence upon growth of the younger ones died away with distance. Accordingly, for each of the six trays, we regressed the natural logarithm of the difference between a reference value and the average carapace length in a row (n==8) against its distance downstream from the older-animal block to obtain exponents for the corresponding geometric series model : A =d0e «, i = 1,2... (7) where d\ is the difference in mm between the carapace length for row i and the reference value, and / ( in units of rows ) is the reciprocal of the space constant of the series. We used a reference value of 0.1 mm larger than the largest row average in rows 11—15. The corresponding average value obtained for / was 0.690 ± 0.048. This value produces a "spatial half-life" for the downstream effect very close to one row. Molt interval and increment differences The 15% decrement in growth rate appears to be partitioned approximately equally between an increase in intermolt interval and a smaller molt increment. Carapaces, especially of the smaller molts, were often partially eaten before they were retrieved and measured, and many molts were missed, so that molt time information also was often lacking or uncertain. Hence, we subtracted the row GROWTH INHIBITION IN HOMARUS TAULK I Analysis of variance, final carapace lengths. Source Degrees of freedom Sum of squares Mean square F-ratio P' A. First four rows of younger animals (rows 1-4) Total 191 320.01 Replicates 7 7.24 1.03 0.69 System (1) 1 0.57 0.57 0.38 Tray (2) 2 41.04 20.52 13.73 0.000 Interaction 12 2 3.35 1.92 1.29 >0.25 Row (3) 3 9.26 3.09 2.06 >().!() Interaction 13 3 2.82 0.94 0.63 Interaction 23 6 10.06 1.67 1.12 >0.35 Interaction 123 6 10.55 1.76 1.18 >().3() Error 161 240.64 1.49 B. Six older animal rows (rows 5-- 10) Total 287 599.73 Replicates 7 16.29 2.33 1.14 >0.30 System (1) 1 4.72 4.73 2.32 >().!() Tray (2) 2 7.68 3.84 1.89 >0.15 Interaction 12 2 4.15 2.08 1.02 >0.35 Row (3) 5 29.71 5.94 2.92 0.02 > P > 0.01 Interaction 13 5 4.20 0.84 0.41 Interaction 23 10 10.40 1.04 0.51 Interaction 123 10 23.80 2.37 1.17 >0.30 Error 245 498.76 2.04 C. Last five rows of younger animals (rows 11-15) Total 239 652.32 Replicates 7 9.47 1.35 0.63 System (1) 1 1.86 1.86 0.88 Tray (2) 2 0.70 0.35 0.16 Interaction 12 2 2.35 1.17 0.55 Row (3) 4 177.75 44.44 20.89 0.000 Interaction 13 4 1.26 0.32 0.15 Interaction 23 8 8.95 1.12 0.52 Interaction 123 8 16.36 2.05 0.96 Error 203 433.61 2.14 1 Value only given for F > 1.0. average for a particular molt number from that of the preceding one to estimate the interval or increment. For this reason standard errors are given only for molt time and carapace length in Table II, which summarizes the differences between the first and fifth rows of younger animals (i.e., rows 11 and 15) following the older animal block in each tray. The "nearest-neighbor effect" is felt immediately. Animals immediately down- stream of the older individuals molt from fourth to fifth stage a day and a half later than those five rows downstream, and when their next molts are measured are found to be already 5 percent smaller. The quotient (b/a) combining the 170 NELSON ET AL. TABLE II Growth differences between first and fifth rows following the block of older animals, as reflected in intermolt interval and molt increment. Columns (a) and (b) express age and carapace length, respectively, of row 11 animals as percent of those in row 15. The last column estimates growth of row 11 animals as percent of those in row 15 . Based upon final carapace length measurements this was 83.7%. Molt to stage Row- N Age at molt (days)' x ± s.e. (a) Intermolt interval (days) (Xi — Xi-l) N Carapace length after molt (mm) y ± s.e. (b) Molt increment (mm) (yi — yi-i) Estimated growth (a X 10°) 5 It 40 28.88 ± 0.39 105.7 22 5.355 ± 0.076 95.5 0.567 90.4 15 34 27.33 ±0.31 20 5.610 ± 0.089 0.810 6 11 31 42.69 ± 0.44 107.81 13.81 21 6.256 ± 0.105 91.0 0.901 84.4 15 41 39.60 ± 0.35 12.27 20 6.875 ± 0.124 1.265 7 11 32 55.41 ± 0.60 107.25 12.70 20 7.590 ±0.193 92.6 1.334 86.3 15 40 51.66 ± 0.40 12.04 26 8.192 ± 0.122 1.317 X 11 38 71.37 ± 0.50 107.63 15.96 23 9.083 ± 0.202 91.0 1.493 84.5 15 42 66.31 ± 0.62 14.65 20 9.985 ± 0.238 1.793 9 11 26 89.00 ± 1.01 107.0 17.63 12 = 10.025 ± 0.336 85.5 0.942 79.9 15 38 83.21 ± 0.77 16.90 27 11.719 ± 0.234 1.734 1 Taking Day 270/78 as age 0. - Molt to stage 10 incomplete for this row by end of experiment. effects of molt increment decrease (b) and molt interval increase (a) shows by the second molt in the system the full \S% decrease in growth seen in the final carapace-length figures. Upon exposure to 5th- to 6th-stage juveniles, the 4th-stage lobsterling apparently switches very soon to a slower rate of growth, reflected in both intermolt interval and molt increment. DISCUSSION Results of this study permit further elucidation of the density-dependent growth inhibition observed in earlier experiments with juvenile lobsters. The experimental design allowed separation of the effects upon growth of a) long-lived metabolites not removed by filtration, air-stripping, etc., which would thus build up along the table, and be much higher in concentration in System II with its lower makeup rate (see compartmental analyses above); b) long-lived metabolites removed by air-stripping or other means, which would build up along a table, but similarly in the two systems; and c) metabolites whose effects in the system were short-lived. The gradients in ammonia concentration may serve as a paradigm of a). Were such a metabolite effective in inhibiting growth in the concentrations attained, we would expect system differences, tray differences, and row differences to be manifest in the ANOVAs of Table I. Similarly, were metabolites which were removed in the sump, filters, or UV units effective in inhibiting growth in this experiment, we would expect only tray differences and row differences to appear in the ANOVAs. In fact, there were no differences at all between the systems, a result similar to that obtained by Cobb and Tamm ( 1974). The only tray differences were between the first three rows of trays A and the first three rows in trays B and C (Fig. 2A). (Surprisingly, growth in these first three rows in the gradient was shiver. We have no ready explanations for this, although several possibilities exist. First, it is possible but not likely that supersaturation of gases in the seawater influent was not entirely removed by heating, air-stripping, and nucleation in the makeup GROWTH INHIBITION IN /IO.M.1KUS 171 0) o cr * o J I 4 5 10 row FIGURE 3. Compartment models corresponding to Figure 2. Growth rate ( ordinate, arbitrary units) is plotted against compartment number in a catenary array (abscissa). Pro- duction of inhibitory principle is higher in compartments 5-10. In A, the turnover rate con- stants A'ji and A'u (see compartmental analyses) are symmetrical; in B, the &u's are negligible; in C the A'ij's are small but not negligible. sump (Fig. 1. M). Secondly, the UV treatment could have resulted, for example, in the production of short-lived ionized molecules with an adverse effect upon growth. Third, it is possible that "natural" sea water, either because of the presence of a substance removed by lobsters or through the absence of a "conditioning" substance produced by lobsters, is not as conducive to growth of lobsters as is lobster-conditioned water (see Cobb and Tamm, 1975a). But if the effect in our systems were attributable to the makeup water, it should be more pronounced in System I with the higher makeup fraction, which a separate two-factor factorial ANOVA on just the first four rows of trays A failed to show. Therefore, if a "conditioning" substance is responsible, it must be removed from the recirculation water before the latter re-enters tray A.) Density-dependent growth inhibition appeared instead to be a short-lived effect of animals in the immediate neighborhood, i.e., an effect which dies away with distance, and was obviously asymmetrical, i.e., more pronounced downstream. Fig- ure 3A indicates schematically what we would expect from a compartmental model (see compartmental analyses above) of a 15-compartment column with symmetrical upstream and downstream turnover constants /etj and k^ for the transmission of a short-lived inhibitory influence. Figure 3B shows the pattern anticipated were the upstream flow completely negligible, and Figure 3C the pattern with dominant downstream flow but non-negligible upstream influence. In the case <.>t Figure 3A growth would be stunted symmetrically in the younger individuals on either side 172 NELSON ET AL. of the block of older animals, and also in the middle of that block. The reason for the latter effect is that the older animals at the edges of the block have only half the number of large neighbors as do the ones in the middle of the block and hence experience less inhibition. With model 3B only the older animals imme- diately downstream from the younger animals would be "released" from inhibition, and only the younger animals immediately downstream from the larger ones inhibited. If there were some slight upstream range of the inhibitory influence, the pattern of Figure 3C would result. Comparison of Figure 3 to Figure 2 suggests the model of Figure 3C, and appears to eliminate certain explanations. First, the possibility of a direct behavioral cause, for example aggressive interactions, appears to be eliminated by the trans- mission of the effect to non-adjacent compartments. It is hard to imagine an effect based upon visual or bodily contact which could produce this result. More- over, radiative influences (e.g., sound) cannot account for the upstream-downstream asymmetry. We appear to be left with some form of chemical influence, carried differentially downstream, but not accumulating. Although in the dye tests no significant back-flow of dye was observed, the possibility of small but appreciable upstream influence cannot be excluded. There appear to be only two tenable hypotheses : first, the continuous production of a chemical substance or complex whose effect is short-lived, either because of loss to the outside or because of decay or inactivation within the system ; and second, the pulsed production of a more or less long-lived inhibitory substance. Either process would be capable of producing an effect upon growth which died away with dis- tance from the source. We will examine the two hypotheses in turn. For a short-lived inhibitor with decay rate equal to the compartment turnover rate k — 0.631 per min, the steady-state concentrations in compartments downstream from the older-animal source may be calculated to be as shown in Figure 4B (see compartmental analyses above). For comparison we recall that the regressions for the downstream decay of the nearest-neighbor effect gave "spatial half-lives" of approximately one row for the downstream effect, as with these parameters. Thus we may say that if the growth inhibitory effect is due to a molecule or complex which decays or is otherwise lost from the system, the decay constant is on the order of 0.6/min or that the substance has a half-life of approximately 1 min. It appears from the oxygen results (see compartmental analyses above) that it is extremely unlikely that a substance could be lost through the air-water interface at such a rate, and surface absorption or bacterial inactivation at that rate seems similarily unlikely. If the molecule or complex is indeed lost from the system at this rate, it appears that it is unstable or easily inactivated by other factors in seawater. If the inhibitor is produced in a pulse, as for example with intermittent urinary production (Cornell, 1979), the concentration pulse will "spread out" and become lower as it travels downstream, as in Figure 4A (using the compartment turnover rate k -- 0.631/min ; see compartmental analyses). With the pulsed-production model the question arises: To which aspect of the concentration is the downstream animal sensitive? If it integrates the concentration over time, then each down- stream animal will experience the same quantitative effect upon growth. But if it is sensitive instead only to the maximum concentration experienced, or to the rate of increase in concentration in its compartment following the unit impulse up- stream (in effect differentiating the curve of rising concentration), downstream effects upon growth rate similar to those observed could be obtained. Whatever GROWTH INHIBITION IN HOMARUS 173 parameter of the pulse is presumed effective, if its duration is much greater than a minute, or its onset is less abrupt than the unit impulse, downstream smoothing will he more pronounced and the successive compartments less differentiated from one another. Pulsed release of urine containing an inhibitor is a likely candidate. However, the hypothesis of a continuously produced short-lived molecular species or complex with a half-life on the order of a minute appears at this time to be equally tenable. Without the blocks of older animals in the experiment, this rapidly-decaying effect upon growth would not have been demonstrated. However, that growth in i he middle of these blocks is inhibited relative to the lead row of the block (Fig. 2) indicates that we are not just dealing with an effect of older upon younger animals, or of H. aniericanus upon hybrids. Even with cohorts of similar sized individuals such growth inhibition may be presumed to exist, and removal of it should result in economically significant gains in growth rate. But the best way to go about this will depend on experimental answers to the questions a) Is the inhibitory substance pulsed in its release, rapidly decaying, or both? and b) Is the substance self-inhibitory ? The fast-decaying growth inhibitory effect we have observed apparently differs in several respects from the delay-of-molt phenomenon described by Cobb and 6 8 10 TIME (min) 12 14 FIGURE 4. A. The course of concentration fordinate) with time following a unit impulse of production (abscissa), in the successive compartments (rows) downstream from the produc- tion compartment (separate curves; the exponential curve represents the clearance of the production compartment). A0 = \ g per 1; k = 0.631 per min. See compartmcntal analyses. B. Steady-state concentrations (ordinate) in successive rows (abscissa) downstream from a point of continuous production of an inhibitor decaying with a rate /.'. equal to the turn- over rate to the next compartment A\., both equal to 0.631 /min. The rate of production has been set equal to 1 g per 1 min. See compartmental analyses. 174 NELSON ET AL. Tamm (Cobb, 1968, 1970; Cobb and Tamm, 1974, 1975a-c). In their experi- ments the dominant individual of a pair held together clearly delays the molt of the subordinate individual, but no mention is made of an effect upon the growth incre- ment of the subordinate. Furthermore, the molt-delay phenomenon is said to disappear in the absence of physical contact (Cobb and Tamm, 1975a). It seems instead to be an expression of dominance, possibly brought about in part by the greater activity of the subordinate animal (Cobb and Tamm, 1975c ; the effect on growth which we have found could of course also be brought about by increased activity of the downstream animals). However, the molt-delay phenomenon is not clear in groups of three or more lobsters (Cobb and Tamm, 1975b). In further contrast to our results is their suggestion that chemical communication in the absence of visual or physical contact may facilitate molting rate and synchrony, although the evidence is apparently not convincing (Cobb and Tamm, 1975a). Our results point instead to a desynchronizing influence of the fast-decaying inhibitory effect. However, the possibility of chemical facilitation of growth (need for "lobster-conditioned" water) was discussed above as an explanation of the reduced growth rates in the first three rows of tray A (Fig. 2A). The fast-decaying inhibitory effect together with dominance effects may account in full for the enhancement of growth rate variance in various crustaceans held communally (Cobb and Tamm, 1975b; Malecha, 1977; Aiken, 1977). In mass culture the dominants' behavior tends to isolate them from the more subordinate animals, which in turn often tend to crowd together in the corners, etc. (pers. obs.). Under such conditions a fast-decaying or pulsed inhibitor could exert maximum effects upon the subordinates. Whether such massing together occurs in nature is unknown. But it may under certain circumstances, e.g., spatially concentrated food or shelter. Some of the so-called "space limitation" effects in crustacean culture experiments (Ranch ct al., 1974; Shleser, 1974; Sastry ct al, 1975; Sastry and French, 1977; Van Olst and Carlberg, 1978) may also be due to a fast- decaying inhibitory effect in compartment arrays which have not been designed to differentiate between the effects of limited space per sc and the effects of nearest- neighbor inhibition (see also Cobb and Tamm. 1975c). We wish to thank C. E. Falbo for mathematical suggestions, and L. W. Botsford, J. S. Cobb, and D. E. Wickham for their helpful comments on the manuscript. SUMMARY 1. A density-dependent inhibitory effect upon the growth of juvenile lobsters was studied in two semi-recirculating seawater-table systems in which controlled metabolite- gradients were established by varying rate of makeup flow. Each system contained eight columns of 45 compartments each. Honianis anicricanus fifth-stage juveniles were placed in the compartments in rows 5-10, 20-25. and 35-40, numbering from the incurrent end of each system, and fourth-stage H. anicricanns X H. (jaiiniiants 1'", hybrids were placed in all other compartments. 2. At the end of 93 days hybrids immediately downstream from the older animals were an average of \5% shorter than those five rows downstream. No effect of metabolite accumulation could be demonstrated. GROWTH INHIBITION IN HOMARUS 175 3. Analysis by a compartmental model indicated that the growth-inhibitory effect was chemical and was due either to a rapidly decaying, continuously produced inhibitor with a half-life of about 1 niin, or to an inhibitor pulsed in production or release. 4. Decreased growth in the middle of the blocks of older animals demonstrated that the effect was neither species- nor age-specfic. The effect is substantially different from other growth-inhibitory phenomena previously described in Crustacea. LITERATURE CITED AIKEN, D. E., 1977. Molting and growth in decapod crustaceans, with particular reference to the lobster Honianis americanus. Pages 41-73 in B. F. Phillips and J. S. Cobb, Eds., U'orkslwp on Lobster and Rock Lobster Ecology and Physiolot/y. CSIRO, Division of Fisheries and Oceanography, Melbourne. ATKINS, G. L., 1969. Multicompartment Models for Biolo76 a few shallow, near shore stations (< 4 m deep ) were sampled but in 197; 5 all stations were located near midchannel and were 6-14 in deep. During the initial phase of the study (January, 1976-April, 1977) decapod shrimp were collected monthly by replicate tows with three #8 nets (mesh == 200 /*m) towed simultaneously at each 1 New York State Museum Journal Series No. 285. 177 178 CLIFFORD A. SIEGFRIED SACRAMENTO - SAN JOAOUIN RIVER SYSTEM r.«»OUINti STH/UT WM JOAOUM RIVER FIGIKK 1. Location of study area and collection sites in Suisun Bay and the Sacramento- San Joaquin River Delta. Triangles indicate sites sampled only in 1976. station: one just below the surface, one near mid-depth, and the third secured to a sled which maintained it just above the substrate. The nets consisted of a 1.3-m section of cylindrical netting, 50 cm in diameter, preceding a 2-m cone-shaped section which terminated in a removable plankton bucket. Each net was equipped with a current meter which was used to determine the volume sampled. From April, 1977, through October, 1978, single step-wise tows were made at biweekly intervals at the 32 stations indicated in Fig. 1. The nets used during this phase of the study were of 505-/xm mesh netting, with 29-cm-diameter mouths and were 1.48-m SHRIMP ABUNDANCE AND DISTRIBUTION 179 long. These tows were made during n 3-5 consecutive days between .', lir before and 1 hr after high neap tide by the California Department of Fish and (lame, Ray-Delta Fishery Project, as part of their N. incrcedis monitoring program. Conductivity and water-temperature data were collected at each sample site. All shrimp collected were placed in 10/4 buffered formalin and later transferred to 70c/f isopropyl alcohol. All shrimp collected were measured from the anterior edge of the carapace (excluding the rostrum) to the posterior tip of the telson. The sex of all speci- mens collected in 1977 and 1978 was determined by microscopic examination of the endopodite of the second pleopod. On developing and mature males this endopodite bears an appendix masculina, whereas for females the structure is similar to those of the third pleopod (Newman, 1963; Smith and Carlton, 1975). An additional distinguishing characteristic used for C. franciscorum sex determina- tion was the structure of the endopodite of the first pleopod, which is short and curved inward for males and long and straight for females. Preservation made it difficult to locate the gonopore of each shrimp so this character was not used for sex determination. For brood-size determinations the eggs of 17 ovigerous specimens of C. jran- ciscontin and 66 specimens of P. inacrodactylits were stripped and enumerated. Only females with ova in the early developmental stages were used for determination of brood sizes. Length-weight relationships were determined for 238 specimens of C. jran- cisconnn and 183 specimens of P. iiiacrodactylns. Each shrimp was measured and then dried at 60°C for 48 hr, cooled, and then weighed to the nearest 0.01 mg for shrimp > 2 mg and to the nearest 0.001 mg for shrimp < 2 mg. RESULTS The size frequency distribution of specimens of C. franciscorum collected in the study area from November, 1977, through October, 1978, is presented in Figure 2. Specimens of C. franciscorum were not caught until they were about 10 mm long. The largest influx of 10-20-mm-long C. jrancisconnn specimens occurred in May. Sexual dimorphism is typical for crangonids (Lloyd and Yonge, 1947; Meredith, 1952; Allen, 1960; Price, 1962; Durkin and Lipovsky, 1977) and was evident for specimens of C. franciscorum from the Bay-Delta. Almost all specimens of C. jrancisconnn collected from the delta that exceeded 45 mm in length were female. The largest female collected from the delta was 72 mm long while the largest male was 52 mm long. Collections of C. jrancisconnn made in San Pablo Bay in May-June and September, 1977. indicate a significantly different population struc- ture (Fig. 3). Samples from the delta in May- June, 1977, indicated a population made up of about 50r/f juvenile shrimp, median size -- 34 mm and maximum size = 50 mm. Samples from San Pablo Bay indicate a population primarily of mature shrimp, median size -- 50 mm and maximum size — 72 mm. In September the median of C. jrancisconnn collected in the delta was 38 mm (range 27-52 mm) while in San Pablo Bay the medium size was 48 mm (29-64 mm). No estimates of growth were made from the size frequency histograms. Immi- gration, emigration, and temperature-salinity differences all combine to obscure growth patterns of C. jrancisconnn in the Delta. C. jrancisconnn eggs generally hatch in the spring. Young develop into juveniles by summer and reach maturity by winter. 180 1 04- 1 o4- 20 -, 10 - o 40 - JO - 20- 10- o in 20- u. 0 10 - a 30 -i CLIFFORD A. SIEGFRIED Crangon ftancltccrufn October 1977 n= 29 NovlmMf 1977 n = 8 Januory- February l»78 11*18 March 1978 n=6 April 1978 Junt 1978 n • 258 111 July 1978 n« 222 S«pttmt>«r 1978 n = 177 Oclob«t 1978 n= 207 20 SO 4O SO TOTAL LENGTH (mm) 60 70 FIGURE 2. Size frequency distribution of specimens of Crangon jranciscorum collected in the Suisun Bay, Sacramento-San Joaquin River study area, November, 1977-October, 1978. Dark bars indicate number of males, hatched bars indicate number of females, and open bars indicate number of juveniles or undetermined individuals measured. The si/! frequency distribution of specimens of P. macro dactylus collected in the study area is presented in Figure 4. Summer is the main recruitment period for P. tnacrodai tylns in the delta. Ovigerous females were collected from May through August, with a few collected in April and September. P. macro dactylus larvae are planktonic and are very abundant in the plankton of the delta during summer. Larval specimens of P. macrodactylus are photopositive (Little, 1969) but become more photonegative as they develop (Siegfried ct a/., 1978). SHRIMP ABUNDANCE AND DISTRIBUTION 1S1 Sexual dimorphism is also apparent in P. macrodactylus. Almost all indi- viduals longer than 40 mm were female. Females reached ca. 55 mm maximum length and males ca. 44 mm. This is near previously reported sixes (Newman, 1963). Size at sexual maturity was not determined but secondary sexual char- acteristics are apparent when shrimp attain about 20 mm in length. Ovigerous females as small as 23 mm long were collected, so males would presumably mature at this or smaller sizes. Length-weight relationships for juvenile, male and female specimens of C. jranciscorum are presented in Figure 5. The regression equations describing these relationships are given below : Juveniles (n " 100) : log W = -5.41 + 2.58 log L, r ^ 0.88 Males (n == 74) : log W -6.12 + 3.27 log L, r == 0.92 Non-ovigerous females (n = 57) : log W : —6.62 + 3.57 log L, r = 0.96 (W = dry weight in grams and L = total length in mm). Analysis of covariance (Steel and Torrie, 1960) revealed significant differences in slopes between the length-weight regression of juveniles and mature shrimp. The difference is attributable, at least in part, to gonadal development. Length— weight relationships for P. macrodactylus are presented in Figure 6. Regression equations describing these relationships are given below : juveniles (n = 118) : log W = -5.44 + 2.53 log L, r == 0.95 Males (n == 45 ) : log W -5.49 + 2.95 log L, r = 0.97 Non-ovigerous females ( n = 19) : log W = -6.10 + 3.40 log L, r == 0.98 The slopes of the above regressions are all significantly different from one another (« = 0.05). P. macrodactylus is more robust than C. jranciscorum, in many cases weighing 50*/£ more than a C. jranciscorum of similar length. Ovigerous specimens of C. jranciscorum were collected in San Pablo and San Francisco Bays in the spring and fall of 1977. Most female specimens of C. jranciscorum collected in San Francisco Bay in May were ovigerous while in 40 - 30 - 20 - 10 0 40 30 20 10 10 20 May -June 1977 n = 247 September 1977 n = 312 30 40 50 TOTAL LENGTH (mm) FIGURE 3. Size frequency distribution of specimens of Cranyoii fnnicist-nnuii collected in San Pablo and San Francisco Bays, May-June and September. 197/ Dark bars indicate number of males, hatched bars indicate number of females and open bars indicate number of juveniles or undetermined individuals measured. 182 CLIFFORD A. SIEGFRIED September-October, most were non-ovigerous. Ovigerous females in our samples ranged from 48 to 67 mm long. Ovigerous specimens of P . macrodactylus were collected in the study area from April (1977) and May (1976, 197S) through August each year. More than Paloemoti mocrodoctytos 10 20 30 40 TOTAL LF.NGTH (mm) January 1978 : 27 Novtmbtr 1977 n* 14 Augutt 1978 n= 147 September 1978 n = 76 Octobtf 1978 FIGURE 4. Si/c- frequency distribution of specimens of Pulucnnni macrodactylus collected in Suisun Bay, Sacramento-San Joacjuin River study area, November, 1977. Dark bars indi- cate number of males, hatched bars indicate number of females, and open bars indicate number of juveniles or undetermined individuals measured. SHRIMP ABUNDANCE AND DISTRIBUTION 183 1000 -r; 0-100 - - I o.oio 4- 0001 Cranyon franciscorum 10 20 30 40 Total Length (mm) 50 60 70 80 90 100 FIGURE 5. Relationship between total length (mm) and dry weight (g) of juvenile (solid circles), male (open circles), female (triangle) specimens of Crane/on fraiiciscoruin collected from the study area, January, 1976-October, 1978. 50c/f of the mature females collected during these periods were ovigerous. Ovig- erous females ranged in size from 23 to 48 mm long. Female specimens of P. macrodactylus appear capable of producing more than one brood ; ovigerous females often carried a second brood in the ovaries while the first brood had not yet been released. The brood sizes of 17 specimens of C. franciscorum, ranging in size from 48 to 67 mm long, and 66 specimens of P. macrodactylus ranging in size from 23 to 48 mm long, are presented in Figure 7. The relationships between brood size and body length are given by : C. franciscorum: log N = —3.66 + 4.09 log L. r := 0.90; P. macrodactylus: log N= -1.66 + 2.96 log L, r — 0.81. The size ranges of ovigerous females of the two species have little overlap, but in the area of overlap, P. macrodactylus has a greater mean brood size than C. franciscorum. The seasonal abundance of C. franciscorum and P. macrodactylns in various parts of the study area is shown in Figures 8 and 9. There are obvious differences between 1977 and 1978 in the location of the peak population densities. In 1977. a very dry year, salinity incursion was more extensive than in 1978, and the peak- densities of both C. franciscorum and P. macrodactylus were located in the river channels. In 1978 the bulk of both populations was centered in Suisun Bay. 184 10,0.0);= 1.0, (01) -- O.I, (.01)-- .01 CLIFFORD A. SIEGFRIED Palatmon macrodoctykjs 10 2 20 3 30 4 40 5 50 6 60 7 70 8 9 10 90 90 100 Total Length (mm) FIGURE 6. Relationships between total length (mm) and dry weight (g) of juvenile (triangle), male (open circle), female (solid circle) specimens of Palacinon macrodactylus collected from study area, January, 1976-October, 1978. Inner scale = juveniles, outer scale = male and female shrimp. A significant difference between C. franciscorum and P. macrodactylus is evi- dent in their population densities in the San Joaquin River. P. macrodactylus was abundant in the San Joaquin River during the summer (particularly in 1977) while C. franciscorum was virtually absent. The San Joaquin River receives more industrial and agricultural effluents relative to its discharge than the Sacra- mento River. This may create water quality differences between the two rivers that limit C. jranciscorum distribution. Roth C. franciscorum and P. macrodactylus appeared to be limited upstream by low salinities. Few individuals of either species were collected from waters <\y,t, salinity (Fig. 10). In both 1977 and 1978, the maximum concentration of C. jranciscorum in the delta was generally in the salinity range \-7%c (Fig. 10). Palacinon appears to be more tolerant of low salinities combined with low tem- peratures than is C. franciscorum. In the spring and fall-winter it is not unusual to collect /'. macrodactylus in fresh or nearly fresh water. C. franciscoritm is generally absent from the delta channels at those times but is often abundant in Suisun Slough and adjacent sloughs from January through May where salinities are near 2%o. SHRIMP ABUNDANCE AND DISTRIBUTION 185 DISCUSSION The maximum size of C. franciscoruw collected in this study is in good agree- ment with maximum sizes reported from Yaquina Bay, Oregon (Krygier and Horton, 1975). However, collections of C. jranciscuruni off the mouth of the Columbia River indicate a population with mean size > 80 mm and maximum sizes > 110 mm (Durkin and Lipovsky, 1977). Reduction of maximum size of individuals inhabiting estuaries as compared to individuals of the same species inhabiting the sea has been reported for Crane/on crangon (Maucher, 1961). Remane and Schlieper ( 1971 ) suggest that reduction in size of marine animals, although generally slight in higher Crustacea living in brackish water, is compar- able to Bergmann's law — that is, size is related to features of the physical environ- ment. The reduction may be attributable to physiological effects of salinity, reduced food availability, or a combination of these and other factors. Studies of osmotic regulation indicate that smaller specimens of C. jranciscorum are capable of better hyper-regulation than larger individuals and that larger ones may tend to hypo-regulate better than smaller individuals (Shaner, 1978). Thus, the migra- tion of larger individuals from low salinity waters to high salinities would be energetically advantageous. 10,000 =p ipoo -- VI 0> 0> tu 100 - - 10 C fronciscoruen f> mocrodoctylvs 20 30 40 50 60 70 Total Ltngth (mm) FIGURE 7. Relationship between total length (mm) and number of eggs carried by ovig- erous specimens of Crangon jranciscorum and Palacmon macrodactylus. All ovigerous speci- mens of Crani/on jranciscorum were collected in San Pablo Bay. All ovigerous specimens of Palacmon macrodactylus were collected in the study area. 186 CLIFFORD A. SIEGFRIED 200 -i 160 - 120 - Crangon franciscorum Suisun Boy chonnel (stations 42-58) Suisun Bay north (stations 20-30, 38-40) Sacramento Rlv»r (itatlone 60-70) Son Jooquln Rlvtr (station* 72 - 82) FIGURE 8. Abundance of Cram/on franciscorum in various portions of the study area, April, 1977-October, 1978. Although C. franciscorum inhabits brackish water for much of its life cycle in the delta, it requires high salinities for reproduction. No ovigerous female speci- mens of C. franciscorum were collected in the delta at any time during the study. Ovigerous females are found year-round in San Pablo and San Francisco Bays but the peak reproductive period appears to be from December through June (Isreal, 1936). Energetic demands of osmoregulation at low salinities may pre- clude egg development and thus reproduction in the delta. Broekema (1941) has shown that low salinities, characteristic of the delta, retard egg development in crangonids. Krygier and Horton (1975) report collecting only 1 of 120 ovigerous specimens of C. franciscorum at salinities below \5%o. Salinity is important not only in relation to egg development but also in larval survival. Preliminary investigation suggest that survival of larval specimens of C. franciscorum declines at salinities below \2%o (Shaner, unpublished). The life history pattern of C. franciscorum in San Francisco Bay appears not only to minimize the energetic demands of osmoregulation but also to minimize interspecific competition with another crangonid that is abundant in San Francisco Bay, C. nigracauda. These species coexist over much of their range (Krygier and Norton, 1975), partitioning their habitat temporally and spatially. In San Fran- cisco Bay C. franciscorum spawns earlier in the year than C. nigracauda and moves 120-1 SHRIMP ABUNDANCE AND DISTRIBUTION Palaemon macrodactylus 1S7 Sulsun Bay channel (»tattOM 42-58) Suicun Bay north (•lotions 20-30,38-40) Sacramanto River (station* 60-70) San Jooqutn R1v«r (station* 72-82) T - 1 - 1 - 1 - 1 - 1 - 1 - 1 SEP NOV JAN MAR MAY JUL CO SEP NOV 1977 1978 FIGURE 9. Abundance of Palactnon macrodactylus in various portions of the study area, April, 1977-October, 1978. into the brackish waters of the delta in the summer. C. nigracauda is restricted to the higher salinity waters of the bay and ocean where it spawns during the summer months (Isreal, 1936; Siegfried, unpublished). C. nigracauda was collected only rarely, and then only at the highest salinities, from the study area. C. francisconint is also somewhat larger than C. nigracauda, suggesting the potential for trophic resource partitioning based on size differences. A third crangonid, C. nigrainaculata, also occurs in San Francisco Bay but little is known of its biology in the system. The life history of P. macrodactylus differs in several important respects from that of C. jranciscorum. Recruitment of young specimens of P. macrodactylus is separated both temporally and geographically from that of specimens of C. jranciscomui. These differences reduce niche overlap between these shrimp and may provide an alternate prey for the C. franciscoruui population. Larval and post-larval specimens of P. macrodactylus are preyed upon by specimens of C. franciscormu and may provide a "buffer" during periods of low prey availability (Siegfried, in preparation). Temporal separation of recruitment accentuates size differences between populations of C. franciscoriim and P. macrodactylus and enhances trophic resource partitioning between these shrimp (Siegfried, in prepara- tion). Differences in size and spawning season may be important factors ultimately determining the potential for coexistence of these two shrimp in the San Francisco Bay System. Reproduction by specimens of P. macrodactylus is influenced by temperature, salinity, and photoperiod. Observations of P. macrodactylus at higher salinities in the San Francisco Bay system indicate that ovigerous females are found for a longer period than in the brackish water of the delta, i.e., mid-April to late October (Little, 1969). Photoperiod appears to be an important parameter controlling spawning in P. macrodactylus. Spawning in the laboratory did not begin until 188 CLIFFORD A. SIEGFRIED 1977 1978 CRANGON FRANCISCOftUM PALAEHON UACRODACTYLUS STATION NUMBER FIGURE 10. Abundance of Crangon franciscorum, Palacinon inacrodactylus, and Neomysis mcrccdis at stations 42-70, on selected dates, May, 1977-September, 1978. Arrow pointing upward marks \f/ce salinity, arrow pointing downward marks 18%f salinity. the photoperiod was above 12 hr light : 12 hr dark, i.e. March, and occurred repeatedly at 20°C under 16 hr photoperiod (Barclay, 1978). Laboratory bioassay indicates optimum salinity for adult C. franciscorum to be around 18-20% c (Khoram and Knight, 1977). In 1977 this salinity range was present in lower Suisan Bay (stations 42^-6) yet no specimens of C. franciscorum were collected from that area in 1977 (Fig. 9). During 1978 that salinity range was located downstream of the study area throughout the year. More recent salinity tolerance investigations indicate juvenile C. franciscorum to be more tolerant of low salinity, with \00% survival at 2%c (Shaner, unpublished). Although low salinity limits the upstream distribution of specimens of C. franciscorum, other factors are important in determining their downstream distribution. The distribution of C. francisconiin in the delta is influenced by the availability of its principal prey species, N. mercedis. Analysis of gastric mill contents of C. franciscorum indicates that not only is C. franciscorum density much higher in locations were N. mercedis density is high (Fig. 9), but those individuals of C. franciscorum in areas of high prey density take more prey than those from low prey density areas (Siegfried, in preparation). The dearth of other populations of suitable prey species in the delta region may be an important factor linking the distribution of the crangonid population to that of N. mercedis. SHRIMP ABUNDANCE AND DISTRIBUTION TABLE I. Calculated indices of spatial overlap (L), and interspecific crowding between (.'. franciscorum and P. macrodactylus (Z) in the San Francisco Bay Delta, April 1 1 if each species has a clumped dis- tribution and their distributions tend to coincide. 190 CLIFFORD A. SIEGFRIED Spatial overlap (L) was calculated for each collection date in 1977 and 1978 and the results are presented in Table I. L was generally greater than 1.0, indicating a higher probability of interspecific encounter than if both C. fran- ciscorujn and P. uiacrodactylus were distributed uniformly, e.g., on July I 1978, the probability of interspecific encounter was 3.42X higher than if both species were distributed uniformly. In 1977 spatial overlap was greatest from about August through October, whereas in 1978 spatial overlap was greatest from April through July. Spatial overlap was lower in early 1977 because a large portion of the P. inacrodactyliis population was in the San Joaquin River or in the sloughs bordering Suisun Bay. In 1978 the P. niacrodactylus catch was high in the sloughs from August to October where no specimens of C. franciscoritni were collected, thus leading to reduced spatial overlap. Niche overlap is generally not reciprocal, i.e., species x generally impinges on species y to a different degree than species y impinges on species x (Hurlbert, 1978). Thus, it is of interest to measure directional overlap or mean crowding of species x by species y and vice versa (Lloyd, 1967; Hurlbert, 1978) : ZX(V) = (SxryO/x. If ZX(V) > y then species x has more experience of species y on the average than would be the case if both species were independently distributed. The mean crowding values of C. francisconun by P. uiacrodactylus, etc., are presented as part of Table I. ZXIV) is consistently greater than y. In 1977 the mean crowding of Crangon by Palacuwn, or the average number of Palaemon encountered by an individual Crangon, was generally greater than the mean crowding of Palacinon by Crangon. The reverse was true in 1978. Competition between these two species may not be occurring even in spite of considerable spatial overlap since space is not likely to be limiting. However, trophic overlap is also great (Siegfried, in preparation) and under some conditions the availability of prey, i.e., N. mercedis, could be limiting. If that is the case, competition would have been most intense in 1977 when N. mercedis densities were about 15% of normal values (Siegfried et al., 1979). It might be significant that the highest spatial overlap values occurred in the later half of 1977 when C. francisconun and P. uiacrodactylus were both abundant and N. mercedis density was very low. The association between C. franciscontin and P. uiacrodactylus is very recent in the San Francisco Bay Estuary. There is no quantitative data available to evaluate the effect, if any, on the introduction of P. uiacrodaclylus on native shrimp populations of the delta. P. macro doc tylus appears more tolerant of some environmental conditions than C. francisconun, occurring not only in the same habitats as C. franciscoruni but in additional habitats, e.g., in the San Joaquin River and in pilings near shore, not used by C. francisconun. This may provide a competitive advantage during periods in which resources become limiting, perhaps during droughts. Careful management of upstream diversions and water projects in the delta may be required to protect the native shrimp of the delta. The author wishes to thank the Biological Survey, New York State Education Department, for support during the data-analysis and manuscript-preparation phases of this work. This work was supported, in part, by a series of gifts to the author and A. W. Knight and the Regents of the University of California from Dow Chemical Company, Pittsburg, CA. I thank J. Orsi and C. Knutson of the California Depart- SHRIMP ABUNDANCE AND DISTRIBUTION mi ment of Fish and Game Delta Study Group for providing caridean shrimp samples and N. mercedis abundance data. I also thank M. Kopache and B. Louks for their assistance in the field and laboratory. K. Conway and L. Picarazzi, respec- tively, deserve special thanks for preparation of the figures and typing of several drafts of this manuscript. SUMMARY The seasonal abundance and distribution of the native caridean shrimp, Crangon franciscontm, and the introduced shrimp, Palacmon macrodactylus, in the channel areas of the San Francisco Bay Delta were studied from April, 1977, through October, 1978. C. jranciscoriim reproduces earlier in the year and grows to a larger size than P. macrodaciylns. C. jranciscoriim reproduction occurs from December to June in the high salinity waters of San Francisco Bay and the Pacific Ocean. P. macrodactylus reproduction occurs from May to September in the delta as well as in higher salinity habitats. Length-weight and length-fecundity relation- ships differ significantly between the two shrimp. Both shrimp are limited upstream by low salinities, few shrimp occurring at salinities < \%c. The downstream distri- bution of these shrimp is related to prey availability, i.e., Neomysis mercedis abundance. Indices of spatial overlap, or interspecies patchiness, indicate a high degree of overlap which varied seasonally and exhibited markedly different patterns in 1977 and 1978. Directional crowding (intraspecific patchiness) also differed between 1977 and 1978. P. macrodactylus appears more tolerant of varied environmental conditions than C. jranciscoriim, occurring in the same habitats and also in additional ones not utilized by C. jranciscoriim. This may give P. macrodactylus a competitive advantage when trophic resources become limiting. LITERATURE CITED ALLEN, J. A., 1960. On the biology of Crangon allmani Kinahan in Northumberland waters. /. Mar. Biol. Assoc. U.K.. 39 : 481-508. BARCLAY, W., 1978. Development of spawning and mass larva-rearing techniques for the euryhaline shrimp Palacmon macrodactylus. Davis, California. Pp. 137-143 in A. W. Knight, Ed., Studies on Biocncrgctics, Osmoregulation, Development, Behavior, and Survival of Several Aquacnlturc Organisms. Dep. Land, Air and Water Resources, Water Sci. and Eng. Paper No. 4507, University of California, Davis. BONNOT, P., 1932. The California shrimp industry. Calif. Dcp. Fish Game Bull. 38 : 20 pp. BROEKMA, M. M., 1941. Seasonal movements and the osmotic behavior of the shrimp Crangon crangon L. Arch. Nccrl. Zoo!., 6: 1-100. DURKIN, J. T., AND S. J. LIPOVSKY, 1977. Aquatic disposal field investigations, Columbia River disposal site, Oregon. Appendix E. Demersal fish and decapod shellfish stiidics. Tech. Rep. D-77-30, Environ. Effects Lab., U. S. Army Eng. Waterways Exp. Sta., Vicksburg, Mass. pp. 119-129. GANSSLE, D., 1966. Fishes and decapods of the San Pablo and Suisun Bays. pp. 64-94 in : D. W. Kelley, Ed., Ecological Studies of the Sacramento-San Joaquin Estuary. Calif. Dep. Fish Game Bull., 133. HULBERT, S. H., 1978. The measurement of niche overlap and some relatives. Ecologv, 59 : 67-77. ISREAL, H. R., 1936. A contribution toward the life histories of two California shrimps, Crago franciscontm (Stimpson) and Crago nigracauda (Stimpson). Calif. Dcp. Fish Game Bull., 46: 1-28. KHORRAM, S., AND A. W. KNIGHT, 1977. Combined temperature-salinity effects on grass shrimp. J. Environ. Eng. Div.. Am. Soc. Civil Eng., 103 : 381-388. 192 CLIFFORD A. SIEGFRIED KRVGIKR, E. E., AND H. F. HORTON, 1975. Distribution, reproduction, and growth of Crangon nit/ricuada and Crangon franciscorum in Yaquina Bay, Oregon. Northivest Sci.. 49: 216-240. LITTLE, G., 1969. The larval development of the shrimp Palacmon inacrodactylus Rathbun, reared in the laboratory, and the effect of eyestalk extirpation and development. Crnstaceana, 17 : 69-87. LLOYD, A. J., AND C. M. YONGE, 1947. A study of Crangon vulyaris L. in the Bristol Channel and Severn Estuary. J. Mar. Biol. Assoc. U.K. 26: 626-661. LLOYD, M., 1967. Mean crowding. J. Anim. EcoL, 36: 1-30. MATCHER, W. D., 1961. Statistische untersuchunger in den korperporotionen wischen der Nord-und Ostee form von Crangon crane/on. Kid. Mccrcsjorsch.. 17 : 219-227. MEREDITH, S. S., 1952. A study of Crangon I'lilgaris in Liverpool Bay area. Proc. Trans. Liverpool Biol. Soc.. 1950-1952 : 75-109. NELSON, S. G., M. A. SIMMONS, AND A. W. KNIGHT, 1979. Ammonia excretion and its relationship to diet in a benthic estuarine shrimp, Crane/on franciscorum Stimpson ( Crustacea : Crangonidae ). Mar. Biol. 54: 25-31. NEWMAN, V. A., 1963. On the introduction of an edible oriental shrimp ( Caridea, Palae- monidae) to San Francisco Bay. Crnstaceana, 5: 119-132. PRICE, K. S., JR., 1962. Biology of the sand shrimp, Crangon septemspinosa in the shore zone of the Delaware Bay region. Chesapeake Sci., 3 : 224-255. REMAXE, A., AND C. SCHLIEPER, 1971. Biology of brackish water. Die Binnengewasser, Vol. 25, 327 pp. SCHMITT, W. L., 1921. The marine decapod Crustacea of California. Univ. Calif. Publ. Zool., No. 23. 470 pp. SHANER, S. W., 1978. Osmotic and ionic regulation in the euryhaline shrimp Crangon francisconan (Stimpson) (Crustacea : Natantia). Pp. 106-121 in: A. W. Knight, Ed., Studies on Bioencrgetics, Osmorcgulation, Development, Behavior, and Survival of Several Aquaculturc Organisms. Dept. Land, Air and Water Resources, Water Sci. and Eng. Paper No. 4507, Univ. of Calif., Davis, California. SHARP, J. W., R. M. SITTS, AND A. W. KNIGHT, 1978. Effects of Kelthane and temperature on the respiration of Crangon franciscorum. Comp. Biochem. Physiol., 59 : 75-79. SIEGFRIED, C. A., A. W. KNIGHT, AND M. E. KOPACHE, 1978. Ecological studies on the western Sacramento-San J oaquin Delta during a dry year. Dept. Land, Air, and Water Resources, Water Sci. and Eng. Paper No. 4506. Univ. of California, Davis. 121 pp. SIEGFRIED, C. A., M. E. KOPACHE, AND A. W. KNIGHT, 1979. The distribution and abundance of Ncomvsis mcrccdis in relation to the entrapment zone in the western Sacramento- San Joaquin Delta. Trans. Am. Fish. Soc., 108: 260-268. SITTS, R. M., 1978. Ecological Aspects of the Estuarine Shrimps Neomysis mercedis, Crangon franciscorum, and Palaemon macrodactylus. Ph.D. Dissertation, University California, Davis, California, 79 pp. Diss. Abs. 7905210. SMITH, R. I., AND J. T. CARLTON, Eds., 1975. Lights Manual: Intcrtidal Invertebrates of the Central California Coast. University California Press, Berkeley, California. 716 pp. STEEL, R. G. D., AND J. H. TORRIE, 1960. Principles and Procedures of Statistics. McGraw- Hill. New York. 481 pp. Reference : B-iol. Bull., 159: 193-205. (August, 1980) FEEDING OF NEOiMVSIS MERCEDIS (HOLMES)1 CLIFFORD A. SIEGFRIED AND MARK E. KOPACHE Biological Survey. Neic York State Education Department, Albany, NY 12230; and University of ll'asliiu and preserved in Sc/c buffered formalin. Selected for gut-content analysis were specimens of N. mcrccdis of three different sizes: mature mysids 10-11 mm long, immature mysids 7 mm long, and juvenile mysids <3 mm long (January-August; few found in September- Xovember). Slides were prepared by removing the guts of 8 mature mysids, 10 immature mysids, or 20-25 juvenile mysids; teasing them open, and filtering the pooled contents onto 0.45-/zm Millipore filter pads. The entire filter pad was examined at 40 X for crustacean or rotifer remains, which were quantified by searching for rami, antennae, or loricas. All phytoplankton prey present on the filter pads were counted and identified from January through August, and per- pendicular strips of each filter pad were counted in September and November. The relative amount of each food item was determined by examining 10 microscope fields at 200 X and estimating the portion of the field occupied by each item. A Whipple ocular micrometer was used to divide the fields into smaller units for approximating percent coverage. Prey availability was determined from phytoplankton and zooplankton samples collected at the time of the mysid collections. Phytoplankton samples were obtained from depths of 1, 2, 5, 7, and 9 m and preserved in Lugol solution. Four replicate counts of each phytoplankton sample, with identifications to the generic level, were conducted by the Utermohl inverted-microscope technique (640x ) (Schworbel, 1970). A Whipple ocular micrometer was used to determine cell dimensions. Three replicate zooplankton samples were collected at the same depths as above by a diaphragm-pump method ( Herrgesell, 1975) using a #20 plankton net (mesh == 80 /mi). Counts were made of the entire sample at 45 X in all months except March, when it became necessary to sub-sample for rotifers. A mean for the entire water column was used to estimate prey availability. Laboratory feeding experiments used N. mcrccdis and Eurytcmora hinindoidcs collected in the Delta and held in laboratory culture tanks. Experiments were conducted at 17°C in sea water reconstituted to obtain test water with a salinity of 4'/ff. Four to 32 individuals of E. Itintndoides were added to 300 ml of water in each of eight 500-ml flasks. Four flasks were taped to exclude light, and four were clear. One mature unstarved mysid was added to each flask except for one light and one dark "control" flask. Feeding experiments were terminated after 12 hr and the number of copepods remaining in each flask determined. If copepocl mortality was significant in the control flask, the results were discarded. RESULTS The phytoplankton of the San Francisco Bay Estuary is diverse, with more than 120 species identified (California Department of Water Resources, unpublished). More than sixty genera, primarily diatoms and green algae, were identified in the present study, and more than half of those have been identified in the gut contents of N. uicrccdis (Table I). NEOMYSIS FEEDING 195 TAHLE I Phytoplankton genera identified from sum piles collet led in the Sacramento-San Joaquin Delta Estuary, January -November, 1V76. Genera identified in \. mercedis guts are indicated by +. CHRYSOPUVTA + Melosira Crucigenia + Achnanthes + M^ld>"1' Golenkiniopsis Amphipleura ' ^"Vlful" Gonium Am phi prom + Nttsschta Microspora + Amphora ' Pinnularia Oocystis + Aster tonella ' P^urosigma + Pediastrum Bacillaria ~*~ ^lolcosPnema Pleurotaenium Caloneis ~+~ ^^'"/)«/'"/'« Radiofilum Chaetoceros ~^~ Skeletonema Scenedesmns + Cocconeis "^ Stauroneis Selenastrutn + Coscinodiscus + Stephanodiscus Staurastrum + Cy dot el I a + •Sl"''rella Ulothrix + Cvmatableura + Synedra C™tua + Tabellana DINOFLAGELLATES Tropidoneis (PYHRRHOPHYTA) + Diploneis -\- Epit hernia -+- Eunotia (;REEN ALGAE (CHLOROPHYTA) -\- Peridinium -f- Frustulia Ankistrodesmus (CYANOPHYTA) + Gomphonema Chaetophora Anabaena -\- Gyrosigma Chodatella Merismopedia Hantzschia Closteriopsis Nodularia Mallomonas Closterium Oscillatoria The composition of the phytoplankton community changed seasonally during 1976. Various centric diatoms dominated the plankton flora from January through September, and a small blue-green alga, tentatively identified as Mersmopedia sp., was dominant in November (Fig. 1), although diatoms continued to dominate the biomass. Phytoplankton density peaked in the spring of 1976 and was lowest in August (Fig. 2). The pattern of phytoplankton population density and com- position was similar at both stations. Zooplankton dynamics within the study area in 1976 were influenced in part by low delta outflows. In January, salinities were low, and the associated fresh- water group of organisms — Cyclops, rotifers, and cladocerans — dominated the upstream area (Fig. 3). Rotifers, primarily Keratella and Notholca, dominated in March. During the remainder of 1976, the fresh-water association was gradually replaced by a brackish-water group dominated by Eurytemora hirundoides. By November, a higher-salinity association dominated by Acartia clausii and speci- mens of Synchaeta appeared in the study area. Densities reached a maximum in March, coincident with the spring diatom increase. Zooplankton densities were lowest in January. Table II lists the zooplankton collected in the study area in 1976. Although diatoms were numerically the most abundant prey in the mysid guts, unidentifiable amorphous material and animal fragments generally covered the greatest portion of the filter pads (Fig. 4). Composition of the diet varied in relation to mysid size and location in the estuary. Diatoms composed the greatest percentage of the diet at the upstream site in May. Diatoms accounted for > 50% 196 C. A. SIEGFRIED AND M. E. KOPACHE PERCENT COMPOSITION PERCENT COMPOSITION 0 20 40 60 80 100 0 20 40 60 80 OO COMMUNITY COMPOSITION | I\\\\X\VII MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS < 3mm LONG COMMUNITY COMPOSITION [ MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS S3mm LONG COMMUNITY COMPOSITION MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS * 3mm LONG COMMUNITY COMPOSITION MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIOS 7mm LONG COMMUNITY COMPOSITION MYSIDS 10mm LONG MYSIDS 7mm LONG EL B COMMUNITY COMPOSITION I IlkXVC ?j MYSIDS 10 mm LONG DOWNSTREAM UPSTREAM Steletont/na Uelotiro Cyc/ottllo CotcinodiKta OTHER DIATOMS JAN MAR MAY JUN AUG SEPT NOV OTHERS FIGURE 1. Composition of phytoplankton community and phytoplankton consumed by N. merccdis in Sacramento-San Joaquin Delta Estuary, January-November, 1976. of the gut materials on one occasion only, at the upstream site in May. The contribution of diatoms to the diet generally decreased with increasing mysid size. Crustacean fragments were not found in the guts of mysids < 3 mm long, but rotifers were present in these guts in May (Figs. 3, 4). Animal remains were relatively abundant in the guts of mysids of the larger size classes. Selective feeding is evident in every month and in all sizes of mysids (Figs. 1, 3). A strong positive selectivity for Melosira was evident among all sizes of XKOMYSIS FEEDING mysids from January through May. In March and May, although Skc was very ahundant. accounting for > 90% of the phytoplankton present, it accounted for only a small portion of the dirt. From June through Novemher, except for the upstream station in August, Coscinodiscus dominated tlie mysid diet, while all other forms were "negatively selected." UPSTREAM en LU Q 10 o * JAN MAR MAY JUL SEP NOV I976 < _l Q_ X Z> CL ^ acticoid copepods (adult and copepodites) were the crustaceans (Fig. 3) eaten most frequently. Mysids > 7 mm long caught each month except November and March had consumed an average of one to two copepods each. In November and March, the average was three to four cope- pods in each gut. The largest mysids (> 11 mm) always consumed more NEOMYSIS FEEDING TAHLK II Fauna collected in zooplankton sample* from the Sacramento- San Joaquin Estuary, Jnnmirv November, 1V76. COPEPODS E it rytemora h iru ndo ides Acartia clausii Diaptomous sp. Cyclops sp. Ectinosoma sp. Sc ot tal ana sp. undet. Harpacticoid a undet. Harpacticoid b CLADOCERA Bosmina longirostrus Daphnia laevis D. pulex D. schodleri D. gul eat a Monospilus dispar Diaphanosoma leiichtenbergianum ROTIFERA Polyarthra sp. Kellicottia sp. /''ilinia sp. Synchaeta sp. Keratella sp. Notholca sp. Brachionus sp. Platyias sp. As planch na sp. Ascomorpha sp. Tetrasiphon sp. Pleurotrocha sp. Trichotria sp. Wigrella sp. MISCELLANEOUS GROUPS Rhithropanopeus harrisii (/oea larvae) Balanus sp. (nauplii) Palaemon macrodactylus (larvae) crustaceans than mysicls 7 mm long. Juvenile mysids, i.e., < 3 mm long, did not appear to consume crustacean prey. Laboratory feeding experiments indicate that as copepod density increases, mysid predation also increases (Fig. 5). At low densities, almost no copepods were ingested, whereas at concentrations of 32/300 ml, almost half the copepods were eaten in 12 hr. Results of the field anaylsis are quite variable but indicate generally increased predation with increasing copepod density. Feeding observations suggest that N. incrcedis is not a particularly active predator. Mysids position themselves horizontally near the bottom or along the walls of the feeding chambers. Copepods were captured when drawn by feeding currents toward the first pair of thoracic appendages. Copepods were not captured until positioned ventral to the eyestalks, after which the mysid seized and held the copepod with its thoracic endopods while forming a cagelike structure with the remaining thoracic appendages. The prey was positioned below the mouthparts and consumed in its entirety, sometimes anterior first, sometimes posterior first. Copepods were often able to escape from the feeding current before becoming vul- nerable. Ingested copepods appear light to dark brown in the guts of live mysids. Ncornysis iiicrccdis appears to feed continuously, with a peak in activity for large mysids during darkness (Fig. 6). The number of copepods consumed per mysid was greater, and the percentage of unrecognizable material less, in larger mysids (> 11 mm) collected near midnight than in those collected during daylight. This pattern was not apparent in small mysids. Fine filter feeding appears to be a continuous process, since the number of diatoms per mysid did not vary sig- nificantly. 200 C. A. SIEGFRIED AND M. E. KOPACHE PERCENT COVERAGE 0 20 40 60 80 100 0 20 40 60 80 100 MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS < 3mm LONG MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS < 3mm LONG MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS < 3mm LONG MYSIDS 10mm LONG MYSIDS 7mm LONG MYSIDS <3mm LONG MYSIDS 10mm LONG MYSIDS 7mm LONG JAN MAR MAY AUG SEPT MYSIDS 10mm LONG MYSIDS 7mm LONG E NOV DOWNSTREAM UPSTREAM DIATOMS ANIMAL AMORPHOUS FIGURE 4. Percentage of gut-content filter-pad coverage attributable to phytoplankton, animal material, and amorphous materials, January-November, 197o. DISCUSSION Kost and Knight (1975) classified the amorphous material that was the most abundant material in the mysid guts as detritus. Detritus derived from vascular plants and inorganic particulate matter was rarely encountered in the present or previous study. The unidentifiable material present in the mysid guts may be the contents of fragmented algae or crustacean body fluids. Rybock (1978) found considerable "detritus" in the guts of specimens of Mysis rclicta collected from areas of Lake Tahoe where no "detritus" existed and concluded that this material NEOMYSIS FEEDING 201 was well digested and macerated prey. The role of detritus in the nutrition of Ncoinysis incrccdis requires further study. The selectivity patterns of nivsids ingesting phytoplankton represent captur- ability based on size rather than true preference. Mclosira and Coscinodi.^ its, the most frequently ingested ph\ toplankters, represent two of the largest algae available, with Coscinodiscns occurring as single cells that often exceed 50 ^m in diameter and Mclosira occurring as filaments of relatively large cells. Skclctoncma occurred Q (/) S UJ Uj 12- 8- 4- LIGHT 16- 12- e- 4- I 4- DARK 0 — o_ kj (/) d < UJ '* • 0 10 20 30 E. hirundoides DENSITY (number/ 1) i 25 50 75 100 125 E. hirundoides INITIAL DENSITY (number/ I ) FIGURE 5. Number of specimens of Eiirytciuora ingested by A", mcrccdis in relation to Eurytemora density. 202 C. A. SIEGFRIED AND M. E. KOPACHE 100 N. mercedis GUT CONTENTS MATURE IMMATURE og 50 UjQ- o o in ^ o < LiJ lOOn C3 Q 75- C/) >- 50- IE \ C/) 25- Q O 0- Q_ UJ Q. 0 O 0 41 3- 2- I- -ANIMAL -AMORPHOUS DIATOMS 0 1300 I9OO 24OO 0700 I3OO 1900 2400 0700 hrs E. hirundoides --- HARPACTICOIDS ..... --- DIATOMS FIGURE 6. Diel variation in A', mercedis gut contents, June 28, 1976. in short filaments of small cells, generally less than 10 /xm long, and in spite of its dominance in the phytoplankton was relatively unimportant in the mysid diet. Mcrismopedia, a small alga, generally occurring in a tetrad less than 5 /xm wide, was not identified in the diet of N. mercedis although it was very abundant in the fall. Smaller mysids tended to feed on the smaller diatoms as well as the larger forms. Similar results were reported for mature M. rclicta from Lake Michigan (Bowers and Grossnickle, 1978). M. relic fa was found to feed almost exclusively on the largest algal size fraction available, i.e.. filamentous diatoms. Algae that passed through a 53-/xin mesh were not consumed during M. rclicta feeding trials. Grazing pressure by mysids on large filamentous diatoms may influence phyto- plankton community composition, particularly in the spring. Peak phytoplankton concentrations were significantly lower in 1976 than in any previous year since records were initiated, i.e.. 1966 (Arthur and Ball, 1978). The low phytoplantkon concentrations, believed to be due to the upstream movement of the entrapment zone (Arthur and Ball, 1978), may have been indirectly respon- sible for the poor success of the mysid population in 1976 ( J. Orsi and C. Knutson, Calif. Dept. Fish and Game, unpublished). The negative selection for copepod nauplii evident in the present study was also evident in Rybock's (1978) study of M. relict a in Lake Tahoe. Nauplii are apparently better able to avoid the mysids' feeding currents than are later copepod life stages. Differences in feeding by various life stages of N. mercedis allow efficient par- titioning of food resources, thus reducing intraspecific competition. Larger mysids NEOMYSIS FEEDING 203 TAHI.K III Calorn value of zooplankton and diatoms in guts oj specimens of N. mercedis, from the .San Fran- cisco Bay-Delta Estuary, 22 January- 16 November, lV7f>. Upstream (Station Ml) 1 1 mm mysids 7 nun mysids Date Zooplankton Phytoplankton Zooplankton Phytoplankton cal/mysid cal/mysid cal/mysid cal/mysid x io-2 X 1()-4 x io-2 X H)-2 22 Jan 1.1 1.1 0.3 0.7 26 Mar 4.1 1.0 3.3 0.7 26 May 0.6 24.0 0.2 1.2 28 Jun 3.0 1.6 1.3 1.7 02 Aug 1.5 1.0 1.2 1.5 21 Sept 1.6 * 0.7 * 16 Nov 3.0 * 0.7 * Downstream (Station M7) 22 Jan 1.0 0.6 0.5 0.8 26 Mar 1.9 0.4 0.4 0.2 26 May 1.2 24.0 0.3 0.4 28 Jun — — — — 02 Aug 1.9 0.7 0.5 0.4 21 Sept 0.7 * 0.5 * 16 Nov 2.3 * 0.2 * * Cells fragmented, no estimates of total number of caloric equivalents made. optimize net energy per unit feeding time by "selecting" the largest prey available Juvenile mysids have a smaller range of prey available. One of the most interesting aspects of feeding by omnivores is the relative contribution of various trophic levels to nutrition. Mysids feed as herbivores when small, whereas carnivory increases in importance in larger mysids. Mean gut contents were converted to caloric equivalents by empirically determined conversion factors to assess the relative contribution of herbivory and carnivory to ingestion by N. uicrccdis (Table III). Carnivory accounted for > 90% of the energy represented by food material pres- ent in the guts of mysids > 7 mm (Table III). Mysids would be expected to pass different foods through their guts at different rates and may assimilate them with different efficiencies (Pechen-Finenko and Pavlovskaya, 1973). That could alter the above estimates (Table III) but should not change the overall patterns. The relative role of predation to the nutrition of M. rclicta in Lake Tahoe is similar (Rybock, 1978). Phytoplankton populations were very low in the Delta in 1976. That may have contributed to an increased importance of carnivory. However, conversion of gut-content information presented by Kost and Knight (1975) to caloric equivalents (assuming, conservatively, that copepods consumed -- 1 10 of crustacean fragments detected) indicates that predation accounted for about 84% of the food energy values. Although herbivory may be of direct importance during blooms (Bowers 204 C. A. SIEGFRIED AND M. E. KOPACHE and Grossnickle, 1978), predation appears to be the most important feeding mode for N. merccdis in the Sacramento-San Joaquin Delta. This work was supported, in part, by a gift to the senior author, A. W. Knight, and the Regents of the University of California from Dow Chemical Company, Pittsburg, CA. The authors thank A. \V. Knight for providing laboratory facilities, A. Hipps and K. Conway for preparation of the figures, and L. Picarazzi for typing of the manuscript. The senior author thanks the Biological Survey, New York State Education Department, for support during manuscript preparation. SUMMARY The diet of the opossum shrimp, Ncomysis merccdis, in the Sacramento River Estuary was studied in relation to food availability, i.e., plankton, from January through November, 1976. The composition of the diet of N. merccdis varied in relation to mysid size and prey availability. Mysids exhibited strong positive selec- tion for the large diatom prey species while "avoiding" small diatom prey. Although diatoms were the most abundant prey identified from the guts of specimens of N. merccdis it was determined that predation on rotifers and copepods accounted for > 80^ of the energy consumed by other-than-juvenile mysids (>7 mm in length). Juvenile mysids (< 3 mm in length) ingested rotifers when rotifers were abundant but were not found to consume copepods. Laboratory feeding experi- ments indicate a density-dependent feeding by N. mcrcedis on copepods, i.e., as copepod density increases mysid predation on copepods also increases. Feeding observations indicate that N. merccdis is not a particularly active predator, captur- ing prey drawn into its feeding current but not actively pursuing prey. N. merccdis appears to feed continuously, with a peak in activity for mature mysids during darkness, a pattern not apparent in immature mysids. Consumption of the detritus was not considered significant. Although herbivory may be of direct importance during the spring diatom increase, carnivory was the major source of energy for N. merccdis in the Sacramento River during 1976. LITERATURE CITED ARTHUR, J. F., AND M. BALL., 1978. Entrapment of suspended materials in the San Fran- cisco Bay-Delta Estuary. United States Department of the Interior, Bureau of Reclamation, Mid-Pacific Region, Sacramento, California. 106 pp. BOWERS, J. A., AND N. E. GROSSNICKLE, 1978. The herbivorous habits of Mysis r dicta in Lake Michigan. Lirnnol. Oceanoar., 23 : 767-776. Fori.Ds, J. B., AND K. H. MANN, 1978. Cellulose digestion in Mysis stcnolcpis and its eco- logical implications. LiiinwI. Oceanof/r.. 23 : 760-766. HERRGESELL, P. L., 1975. Crustacean zooplankton biomass and its relationship to larying levels of cutroph\ in Lake Bcrr\cssa, California. Ph.D. Dissertation, Univ. of Calif., Davis, California, 293 pp. Diss. Abs. 7620991. HEI-UAIII, W., R. J. Torn, AND A. J. McCREAUY, 1963. Food of the young-of-the-year striped bass (Roccus saxatilis) in the Sacramento-San Joaquin River System. Calif. Fish Game. 49 : 224-239. KOST, A. L. B., AND A. W. KNIGHT, 1975. The food of Neomysis merccdis Holmes in the Sacramento-San Joaquin Estuary. Calif. Fish Came, 61 : 35-46. LASENBY, D. C., AND R. R. LANGFORD, 1973. Feeding and assimilation of Mysis relicta. J.imnol. OccuniHir., 18: 280-285. MAUCHLINE, J., 1967. Tin- biology of Scliistomvsis sfiritns (Crustacea, Mysidacea). /. Mar. Biol. Assoc. U.K., 47 : 383-396. NEOMYSIS FEEDING 205 MAUCHLINE, J., 1971. The biology of Ncomvsis inteiier (Crustacea: Mysidacea). J. Mar. Biol. Assoc. U.K., 51 : 347-354. MORGAN, M. D., 1979. Mysis relicta Population Dynamics in Lake Talioc. Ph.D. Dissertation, Univ. of Calif., Davis, Calif. PECHEN-FINENKO, G. A., AND T. V. PATLOVSKATA, 1973. Comparative importance of detritus and algae in the food of Neomysis miralnlis. Hydrobiol. J ., 11 : 28-32. RADTKE, L. D., I960. Distribution of smelt, juvenile sturgeon, and starry ilounder in the the Sacramento-San Joaquin Delta with observations on food of sturgeon. Calif. Dcpt. Fish Game Fish Bull.. 136: 115-129. RAYMONT, J. E. C, J. AI'.STIN, AND E. LIN FORD, 1964. Biochemical studies on marine zoo- plankton. I. The biochemical composition of Ncomvsis integer . J. Cons. Cons. Intern. Explor. Mcr.. 28 : 354-363. RICHARDS, R. C., C. R. GOLDMAN, T. C. FRANTZ, AND R. WKKVVIRK, 1975. Where have all the Daphnia gone? The decline of a major cladoceran in Lake Tahoe, California- Nevada. Int. Fcr. Thcor. Angcu: Limnol. I'erh.. 19: 835-842. RYBOCK, J. T., 1978. Mysis relicta Loven in Lake Tahoe: I'crtical Distribution and Nocturnal Prcdation. Ph.D. Dissertation, University of California, Davis, California. 116 pp. Diss. Abs. 7912972. SIEGFRIED, C. A., A. W. KNIGHT, AND M. E. KOPACHE, 1978. Ecological studies on the western Sacramento-San Joaquin Delta during a dry year. University of California, Davis, Department of Water Science and Engineering Paper 4506, Davis, California. 121 pp. SITTS, R. M., 1978. Ecological aspects of the estuarinc shrimps Neomysis mercedis, Crangon franciscorum, and Palaemon macrodactylus. Ph.D. dissertation. Unversity of Cali- fornia, Davis, California. 79 pp. Diss. Abs. 7905210. SCHWORBEL, J., 1970. Methods of Hydrobiology. Pergamon Press, New York. 200 pp. STEVENS, D. E., 1966a. Food habits of striped bass, Roccus sa.ralilis, in the Sacramento-San Joaquin Delta. Calif. Dcp. Fish Game Fish Bull.. 136 : 68-96. STEVENS, D. E., 1966b. Distribution and food habits of the American shad, Alosa sapidissima, in the Sacramento-San Joaquin Delta. Calif. Dep. Fish Game Fish Bull.. 136: 97-107. TATTERSALL, W. M., AND O. S. TATTERSALL, 1951. The British Mysidacea. Royal Society, London, No. 136. 460 pp. TURNER, J. L., 1966. Distribution and food habits of Ictalurid fishes in the Sacramento-San Joaquin Delta. Calif. Dcp. Fish Game Fish Bull., 136: 130-143. WILSON, R. R., 1951. Distribution, Groivth, Feeding Habits, Abundance. Thermal and Salinity Relationship of Neomysis mercedis (Holmes) from the Nicomckl and Serpentine Rivers, British Columbia. M.A. thesis, University of British Columbia, Vancouver, British Columbia. 67 pp. Thesis call no. LE 3B1 1951A8 D5405. Reference : Biol. Bull.. 159: 206-218. (August. 1980) MECHANISMS OF COORDINATION BKT\YEEX MOULTING AND REPRODUCTION* IX TERRESTRIAL ISOPOD CRUSTACEA C. G. H. STEEL Department of Biolot/y. York University. Dcwnsriew, Ontario. M3J IPS. Canada It has long been believed that both moulting- and egg development in decapod Crustacea arc under inhibitory hormonal control by neurosecretions released from the sinus gland (C.ahe. 1%6; Adiyodi and Adiyodi. 1970; Sochasky, 1973). The observation of apparently synchronous moulting and reproduction in adult crabs (Panouse, 1947) led initially to the view that the processes were "synergistic" and that a single hormone might control both. This view subsequently gave way to the concept of "antagonism" between somatic and reproductive growth, according to which one process occurs only at the expense of the other (Passano, 1960; Charniaux-Cotton and Kleinholz. 1964; Bliss. 1966: Adiyodi and Adiyodi. 1970; Sochasky, lc>7o). Both concepts developed from consideration of the temporal relations between moulting and reproduction among decapods ; their applicability to other groups of the Crustacea is not clear. The present paper examines the mechanisms which coordinate and regulate reproduction and moulting in terrestrial isopods. In the laboratory, as in the field (Heeley. 1941). these animals moult frequently and regularly and can produce several broods of young per year. Moulting and breeding in decapods maintained in the laboratory are generally less frequent and less regular. Information on control of moulting and reproduction in isopods is based primarilv on descriptions of the effects of ablation and transplantation of the protocerebrum (Reidenbach, 1965; Besse. 1968: Besse ct al. 1969; Mocquard ct al.. 1971 : Besse and Donady. 1972: Charmantier and Trilles, 1973). The conflicting results reported may be related to lack of information concerning the relationships between these processes in normal animals and the consequent inability to perform surgical manipulations at appropriate times. The present paper examines these relationships in normal isopods and shows that moulting and reproduction are controlled by separate environmental cues and probably separate hormones. However, the relationships are more subtle than terms such as "synergism" and "antagonism" imply; rather, moulting and reproduction are coordinated by specific sensory cues which maintain precise, if complex, temporal relations between them. MATERIALS AND METHODS All of the observations and experiments described have employed Oniscns ascllns (L. ). Many of them have been repeated using Porccllio spinicornis (Say). Cyclisticus onre.vns (De Geer) and Tracheonisctts ratlikci (Brandt). The con- clusions reported are applicable to all these species ; however, quantitative differences were found between species, mainly in the total length of the intermoult cycle. The quantitative data reported are those for Oniscns ascllns. The experiments reported for Oniscns have been repeated, and have used both animals born and raised in laboratory cultures and animals collected from the field in early spring before either moulting or reproduction has commenced. All animals used in 206 ISOPOD MOULTING AND REPRODUCTION 207 experiments had a head capsule width of 2.0-2.5 mm. Animals this size in the field are 1-year-olds which have not reproduced (McQueen, 1976). Head capsule width appears to be the most reliable convenient parameter of age (Standen, 1970). Body weight was not used, as terrestrial isopods undergo substantial changes in water content (Lindqvist, 1972; Mayes and Holdich, 1975). Laboratory cultures were maintained at 21 ± 1°C in a 16L:8D daylength cycle. An individual culture consisted of 6-20 animals in a plastic Petri dish about 9 by 2.5 cm equipped with a humidity wick, crushed limestone from the vicinity of the field collections, and ground Purina rabbit chow, which Merriam (1971) found to provide an optimum diet. Females incubating broods of young were isolated in single culture dishes until the young were liberated. The adult was then removed and the young raised together in the dish. Thus the age of laboratory-born animals was known precisely. The procedure for determining the various stages of the moult cycle will be described in detail elsewhere (Steel, in prep.). In brief, the sternites of pereion segments 1— I undergo systematic changes in appearance due to the accumulation of fat cells and calcium deposits, which develop through a characteristic sequence of changes in shape and size at specific times during the cycle. Fourteen distinct configurations are recognizable on brief visual inspection of the animal's ventral surface. Stages 1 through 9 represent premoult. The same 14 stages occur with the same relative durations in all species examined, despite species variation in total cycle length. These stages are correlated with the changes associated with secretion and resorption of the exoskeleton (unpublished observations) and are readily related to the alphabetical subdivisions of the moult cycle used in decapods (Drach, 1939). Being the more familiar, the latter terminology is used in this text. The term premoult is used to refer to the period between moult initiation and ecdysis (Stage D in the terminology of Drach). Postmoult is the period of post-ecdysial secretion and calcification of the exoskeleton (Stages A, B, and Ci- C3) and intermoult is the period between the end of postmoult and initiation of the next premoult (Stage C-i). Durations given in the text are means ± S.D. RESULTS Differential effects of temperature and daylength Observations on field populations have shown that moulting occurs throughout the summer, whereas females incubating broods of eggs are found only during a part of this season (see McQueen, 1976, for details). These observations imply distinct seasons for each process. This in turn suggests separate environmental regulation of moulting and reproduction. This possibility was examined in the laboratory using both field-collected and laboratory-raised animals of the same size. Females from both sources moulted regularly at intervals of 28 days at 21 : : 1°C. Moulting in males was slightly less frequent and less regular (see Table II). Moult- ing frequency exhibited a Q10 of two between 12° and 22°C and was arrested com- pletely at 5°C. Animals maintained in intermoult at 4°C could be induced to moult by transference to 21 °C. In such animals, the first sign of of early premoult occurs 14 ± 1.9 days after transfer to 21 °C. Animals brought in from the field in either spring or autumn from outside temperatures of 0-5 °C responded to this increase in temperature in the same way. These observations suggest that premoult is initiated following a threshold number of warm days. Parallel cultures main- tained under "short" (8L: 16D) or "long" (16L:8D) daylength regimes continue 208 C. G. H. STEEL TABLE I Effect of daylength on induction of breeding in female Oniscus. In long days the 17 '"c ivhich did not breed at the first moult all did so at the second. N = 40. See text for details. Daylength regime Long days 16L:8D Short days 8L:16D Type of moult Breeding Non-breeding Breeding Non-breeding First moult Second moult 83% 73% 17% 27%. 9% 27% 91% 73% to moult throughout the year with the same frequency (except under conditions which stimulate vitellogenesis, as described below). Thus, moulting is strongly temperature dependent but is apparently independent of daylength. Adult females are capable of two different types of moults. One is similar to that of males and no change in form occurs. The other occurs only during the breeding season in the field and is morphogenetic in the sense that the oostegites of pereion segments two to five enlarge enormously to form a brood pouch in which eggs are incubated ("maternal" moult of Heeley, 1941). The influence of day- length on the morphogenetic character of the moult was examined in the follow- ing experiment. Twenty female Oniscus which had previously been held in inter- moult at 4°C were induced to commence moulting by transferrance to 21 °C. Half were placed in 8L : 16D and half in 16L : 8D. A further 20 intermoult females collected from the field in early spring were also set up under these two conditions. The proportions forming brood pouches at each of the first two moults are shown in Table I. No differences between laboratory-bred and field-collected animals were detected. Accordingly, Table I presents the combined data for all 40 animals in each regime. In long days, almost all animals formed brood pouches at the first moult, and of those that did not, all did so at the second moult. Accordingly, the 27% of 40 animals which underwent a non-breeding second moult in long days had all undergone breeding at the first moult. The majority of animals in long days underwent two breeding moults in succession. Conversely, breeding was rare in the short-day animals, even after two moults. Thus, while the occurrence of moulting appears to be independent of daylength, the morphogenetic character of the moult is strongly influenced by daylength. The interaction of temperature with daylength in Oniscits has been discussed in detail by McQueen and Steel (1980). It is sufficient here to note that tempera- ture does not directly affect whether or not reproduction will occur ; the role of temperature in regulating reproduction appears to be confined to determining the rapidity with which animals respond to those daylengths which promote reproduc- tion due to the influence of temperature on the length of time between ecdyses. Thus, temperature affects moulting but has no direct effect on reproduction. Day- length and temperature, therefore, exert differential effects on moulting and reproduction. Breeding cultures can therefore be maintained in the laboratory at 21 °C in "long" days. Although moulting continues throughout the year under these con- ditions, females do not breed throughout the year. These laboratory cultures cease breeding in September in synchrony with field populations, indicating that the cessation of breeding is brought about by different cues from those which induce ISOPOD MOULTING AND REPRODUCTION it. For the following 3 months females remain refractory to the stimulation of reproduction by long daylengths, but become responsive to them again in December, when field populations are completely dormant. Thus, there appears to be a seasonal periodicity in responsiveness to conditions promoting breeding. The absence of such periodicity influencing moulting further supports the view that moulting and reproduction are controlled by separate mechanisms which are dif- ferentiullv influenced bv environmental factors. Stages of ooycncsis Mature eggs are deposited into the brood pouch a few hours after ecdysis (Nemec. 1896). The number of eggs per brood pouch averaged 38 ± 6. In order to examine the interrelations between the development of these eggs and the moult- ing cycle, it is necessary to recognize the different stages of egg development. The paired ovaries are tubular structures extending throughout the pereion. In non-breeding animals, each ovary contains 19 ± 8 small white oocytes 100-200 /xm (mean 131 p.m) in diameter. These correspond in size and appearance to the "previtellogenic" oocytes of the amphipod Orchcstia gauiinarclla (Charniaux- Cotton, 1973). Thus, each animal contains sufficient previtellogenic oocytes for the production of exactly one brood of young without cell division. These oocytes do not change in size, appearance, or number during moult cycles which do not lead to formation of a brood pouch. It therefore seems that oocyte development is arrested in the "previtellogenic" stage until reproduction is stimulated. When reproduction is stimulated, all the previtellogenic oocytes grow synchron- ously from 200 to 600 ^m. While this is occurring they become progressively more yellow. Globules accumulate in the cytoplasm and increase in size and num- ber with growth of the oocytes. These changes in oocyte size and appearance are commonly regarded as due to the accumulation of yolk. Hence, this phase of oocyte development represents vitellogenesis. That vitellogenesis does indeed occur in Oniscus oocytes at this stage is confirmed by the finding that developing oocytes first acquire immunoreactivity to antiserum raised against purified Oniscus yolk proteins at a size of 150-200 p.m (C.G.H. Steel and R. A. Baron, unpublished observations). Mature eggs newly deposited into the brood pouch are similar in size and appearance to the largest seen in the ovaries. During the incubation of a batch of mature eggs in the brood pouch, oogonia proliferate and previtellogenic growth of oocytes occurs, thereby restoring the complement of previtellogenic oocytes by the time the brood is liberated. Brood TABLE II Modification of chronology of moult cycle by egg development and incubation. The postmoult period lasts 2-3 days. Hence intermoult is considered here to begin 2 days after anterior ecdysis and to end U'ith initiation of premoult. Thus, brood incubation time equals length of intermoult plus postmoult. Length of premoult Length of next intermoult Xo. of animals Males 16.3 ± 2.0 days 10.0 ± 5.1 days 25 Females, non-breeding 16.2 ± 1.8 7.1 ± 1.4 29 Females, first brood 33.0 ± 3.3 32.0 ±2.6 25 Females, second and sub- sequent broods 17.7 ± 3.1 33.8 ± 3.7 31 210 C. G. H. STEEL incubation lasts until the young hatch within the brood pouch and crawl out of it. At 21 °C the average duration of brood incubation is about 34 days (see Table II), longer than a complete intermoult cycle in non-breeding animals. The mechanisms by which growth of the oocytes and brood incubation are coordinated with moult- ing activities are examined below. Effects of reproduction on moulting As noted above, brood incubation takes longer than a complete intermoult cycle of non-breeding animals. This implies that the chronology of the intermoult cycle is modified during brood incubation. Since deposition of mature eggs into the brood pouch occurs shortly after ecdysis, incubation commences during the postmoult period. Females incubating eggs were inspected every 3 days for signs of premoult. No premoult was initiated until the end of brood incubation (see also below). This confirms that brood incubation occurs during the intermoult stage (Stage d of Drach, 1939). How- ever, the postmoult and intermoult stages of non-breeding females together last only about 9 days (Table II), whereas, incubation lasts about 34 days at the same temperature. Therefore, initiation of premoult must be delayed about 25 days in females incubating a brood. Thus, brood incubation influences the moult cycle by causing an extension of intermoult from 7 to 32 days (Table II). Previtellogenic growth of the oocytes occurs during brood incubation, i.e., during postmoult and intermoult. These oocytes undergo fertilization and incuba- tion following a subsequent ecdysis and formation of another brood pouch. Thus, at least one moult occurs between previtellogenic growth and fertilization. This arrangement contrasts with that in higher decapods in which both egg maturation and incubation occur within a single intermoult (Adiyodi and Adiyodi, 1970). Reproduction also modifies the chronology of premoult. This period lasts 16 days in non-breeding Oniscus females at 21 °C (Table II). However, in animals kept in intermoult at 4°C, and then transferred to 21 °C and long days to stimulate a maternal moult, the premoult period lasts 33 days (Table II). All the constituent sub-stages were found to be of their normal length except for Stage 3 (late D() in the nomenclature of Drach, 1939) which was extended from 3 to 20 days in duration. Thus, there is a lengthy pause in the early premoult of animals stimulated to undergo a first maternal moult. Dissection of the ovaries of animals which had been in this protracted late D0 for various known lengths of time revealed that the changes in size and appearance of the occytes which accompany vitellogenesis commence during this period. Moreover, the oocytes first develop immunoreactivity to anti-vitellogenin serum at this stage (C. G. H. Steel and R. A. Baron, unpublished observations). Thus, vitellogenesis commences in early premoult, at which time the moulting process undergoes suspension for 16 days, at the end of which premoult changes resume with the same chronology as in non-breeding females. However, vitello- genesis is not completed during the period of premoult suspension. Dissection of ovaries from females at various stages of premoult showed that oocyte growth is not completed until shortly before ecdysis (Table III) and most of the increase in size and accumulation of yellow globules in the oocytes occurs after the premoult has resumed. Suspension of premoult is therefore not necessary throughout vitellogenesis. ISOPOD MOULTING AND REPRODUCTION TABLE III Chronology of vitellogenesis in fi.rst and second maternal nundt cycles. The oocyte size ranges given indicate the average size increases observed between the beginning and end of the moult stage indicated. Oocyte sizes were measured for 30 oocytes (15 from each ovary) from each animal, "n" gives the number of animals dissected at each stage. Stage of moult cycle First maternal moult Second maternal moult n Oocyte size range Oocyte appearance Stage duration n Oocyte size range Oocyte appearance Stage duration Intermoult Early 17 100-200 M Clear, white 24 days 11 100-200 M Clear, white 7 clays Late Premoult 19 200-300 M Opaque, yellowish 8 days Early (D0) Late (D,-D,) 12 16 200-250 M 250-600 M Opaque, yellowish Opaque, yellow 24 days 9 days 12 8 330 M 330-600 M Opaque, yellowish Opaque, yellow 8 days 9 days A further illustration of interactions between reproduction and moulting is seen when females incubating broods of eggs are maintained in environments pro- moting reproduction, such that two or more maternal moults occur in succession, a new brood pouch being formed under the old one during the premoult following lib- eration of the brood. Under these conditions, the pause in late D0 seen in the first maternal moults is absent (Table III). The chronology of all stages of premoult is the same as in non-breeding females. Dissection of the ovaries of these animals revealed that at the beginning of premoult, the oocytes had already developed to an average size of 330 /*m (Table III) and had accumulated significant numbers of yellow globules. This indicates that the timing of vitellogenesis in the first maternal moult is different from that seen in subsequent maternal moults. During incuba- tion of the first brood, the oocytes do not cease growth at the end of the previtello- genic stage, but continue into vitellogenesis, such that by the time the brood was liberated the oocytes were about halfway through the changes in size and appearance associated with vitellogenesis (Table III). Thus, the premoult following brood liberation commences with the oocytes already undergoing vitellogenesis. How- ever, vitellogenesis is completed in late premoult, as in first maternal moults. In other words, the later part of vitellogenesis invariably occurs in late premoult, whereas the first part can occur either in early premoult (first maternal moult) or during intermoult (second and subsequent maternal moults). \Yhichever stage vitellogenesis commences in undergoes a considerable extension relative to its duration in non-breeding animals ; the extension to intermoult during incubation is 25 days and the extension to early premoult seen in first maternal moults is 16 days. Thus, the total time occupied by vitellogenesis is similar in both first and subsequent maternal moult cycles, but the position of the first part of vitel- logensis in the moult cycle is flexible. 212 C. G. H. STEEL Stimulus to prcnioidt initiation The finding that intermolt is protracted during incubation suggested the pos- sibility that the timing of the initiation of the next prenioult might be regulated by some event associated with incubation. Subsequent to fertilization of the eggs, the role of the mother in brood incubation seems to be confined to production of a watery fluid which surrounds the eggs in the brood pouch, for fertilized eggs can be grown in vitro in saline ( Sutton, 1972). The eventual liberation of the young from the pouch "seems to be simply a matter of the young crawling out when they are ready" (Sutton, 1972), over a period of up to 2 days (Heeley, 1941). The oostegites are held in a distended position by the brood, but collapse to the sternites when the young are liberated. The first visible signs of premoult are seen 3.1 ± 0.8 days after the complete collapse of the brood pouch. Premature deflation of the brood pouch was produced by rinsing away the developing eggs or embryos in a fine stream of tap water. This resulted in moult initiation, premoult becoming visible on the third day following collapse of the brood pouch as in normal animals. If one or two eggs remained stuck in the pouch, the appearance of premoult was greatly delayed, usually by about 10 days. Thus, the presence of eggs in the brood pouch is necessary to maintain the intermoult condition. In the converse experiment, the brood pouch of females which had just liberated young was stuffed with cellulose sponge moistened with the brood pouch saline of Sutton (1972). No initiation of premoult was seen in these animals. When the sponge was removed after about 10 clays, premoult was detect- able after 3 days as if a normal brood had been liberated. The initiation of premoult is therefore determined by the time of collapse of the oostegites. Presumably, proprioceptive or tactile information from these append- ages provides a nervous pathway for regulating release of the moult-controlling hormones. These observations illustrate quite strikingly that events associated with reproduction provide cues which regulate both length and timing of the moult cycle. However, in males and in non-breeding females, premoult initiation must be accomplished by a mechanism different from the above. The role of a threshold number of warm days has been mentioned earlier. However, the length of intermoult is remarkably constant, especially in females, even after months of constant temperature and day length. Even nutritional state seems not to be fundamental, for starved animals continue to moult frequently even though they become smaller at each moult. It is clear that moulting does not occur solely to permit growth. However, the mechanism of regular premoult initiation in a constant environment remains unknown. DISCUSSION Environmental regulatory factors. Both temperature and daylength influence moulting and reproduction, but have differential effects on the two processes. Moulting is strongly temperature dependent but is unaffected by daylength (McQueen and Steel, 1980). Since moulting ceases at 5°-10°C, moulting in field populations should occur when temperatures reliably exceed 5°-10°C. Local meteorological records show such field temperatures between March and October. Conversely, reproduction is induced by long days (16L:8D) and suppressed by short (8L:16D) in regularly moulting animals. This effect appears to be a ISOPOD MOULTING AND REPRODUCTION 213 conventional photoperiodic response with a critical daylength of about 11.5 hr (McQueen and Steel, 1980). At 44°N, daylength would promote reproduction in the field between April and September. These expectations are confirmed by field observations on natural populations of Porccllio in the same vicinity (McQueen, 1976) and of Oniscus in England ( Heeley, 1941). Thus, daylength confines reproduction to the central portion of the season when temperature per- mits moulting. It is concluded that moulting and reproduction are confined to their respective seasons by differential responsiveness of the two processes to the environ- mental cues of temperature and daylength. Such differential responsiveness fur- ther suggests the existence of separate physiological regulatory mechanisms for each process, as discussed below. Superimposed on these effects of temperature and daylength is a seasonal periodicity in the responsiveness of females to environments which promote breed- ing. Females become refractory to long-day stimulation in the laboratory at about the time when breeding ceases in the field, and remain so for the following 3 months. Similar events occur in cultures of Porccllio in France ( Besse and Maissiat, 1971). This refractoriness suggests the intervention of a phenomenon analogous to repro- ductive diapause. The absence of seasonal periodicity in moulting in the laboratory supports the above inference that moulting and reproduction are separately con- trolled. Stimulus to moult initiation. Brood incubation occurs during an intermoult which is prolonged to about four times its duration in non-breeding moult cycles; initiation of the next premoult is delayed until the young are liberated from the brood pouch. The collapse of the brood pouch when the brood is liberated provides a trigger which initiates premoult. Three days elapse before the appearance of the first morphological signs of premoult. Since morphological changes result from prior hormonal changes, it must be in this period between the stimulus and the appearance of morphological changes that the endocrine events associated with premoult are initiated. All other descriptions of stages of moult cycles in Crustacea rely on the appear- ance of some tissue change to define the beginning of premoult. Such descriptions of premoult in terms of effects rather than causes inevitably overlook the initial stage (s) of premoult in which key physiological changes occur. Thus, animals would be inappropriately classified as in intermoult. This may account for the high individual variation in indices of metabolic activity noted previously in osten- sibly intermoult animals (e.g., Andrieux, 1979). The proportion of the total premoult period which passes without morphological change in Oniscus is 3 out of 16 days at 21 °C, or 19%. However, most other crustacean moult staging schemes recognize premoult by the occurrence of apolysis in the integument (Drach and Tchernigovtzeff, 1967). In Oniscus, apolysis occurs in different regions of the integument between 5 and 8 days after the moult-initiating stimulus (unpub- lished observations). Thus, if premoult were recognized by apolysis in Oniscus, as it is in Sphacroma (Tchernigovtzeff and Ragage-Willigens, 1968), the pro- portion of the premoult period which would pass unnoticed would increase to as much as 50%. It is inferred that the stimulus which initiates premoult and the primary hor- monal changes occurs before tissue changes become evident, and certainly well before apolysis. Thus, premoult begins well before it is recognizable by conventional moult-stage schemes. This conclusion lends substance to that of Passano (1960), who presumed that early D0 would not be morphologically detectable, but differs 214 C. G. H. STEEL strikingly from that of Reaka (1975) and Vranckx and Durliat (1978), who propose that initiation of premoult may not occur until after apolysis. Coordination of moulting and reproduction. The mechanism discussed above, whereby intermoult is extended during incubation, is but one of several mecha- nisms which coordinate the moult cycle with reproductive events. The point in the moult cycle at which vitellogenesis commences was found to be flexible. During a first maternal moult it commences in late D0 and is accom- panied by a temporary suspension of this stage of premoult for 16 days, resulting in a premoult period of double the normal length. In animals in which a second maternal moult follows the first, this prolongation of premoult is not seen and vitellogenesis commences during the preceding intermoult. Thus, different temporal relations are found between vitellogenesis and the moult cycle under different con- ditions, which again suggests that separate mechanisms control these processes. However, regardless of the moult stage at which vitellogenesis commences, most of the growth in oocyte size occurs during late premoult. Thus, two distinct phases of vitellogenesis can be distinguished : The first phase is flexible in timing and can occur either in intermoult or during a protracted D0 ; the second phase invariably occurs during late premoult. The occurrence of vitellogenesis during both inter- moult and premoult also occurs in the amphipod Orchestia (Charniaux-Cotton, 1973 ; Meusy et a!., 1974), in which this arrangement has been ascribed to the short duration of intermoult relative to premoult (Adiyodi, 1978). However, vitellogene- sis in isopods always continues into premoult even when intermoult is prolonged fourfold. This is suggestive less of insufficient time for the completion of vitello- genesis during intermoult than of different physiological requirements for the com- pletion of the two phases. The fact that a pause in late D,, is seen only in the first maternal moult sug- gests that it is not a direct response to the occurrence of vitellogenesis but rather a response to the morphogenetic character of the moult. During a first maternal moult, the epidermis must become reorganized so that a brood pouch will be differentiated. Obviously, this reorganization must be completed before the com- mencement of cuticle deposition, i.e., prior to the end of D(>. Thus, the protraction of DO may be required to effect this reorganization of epidermal cells for differentia- tion of a brood pouch. In contrast, during a second maternal moult, the first brood pouch is replaced by a second pouch and consequently no epidermal reorganization is required. Synergism and antagonism. The relations between moulting and reproduction in crustaceans have been widely described as illustrating either "synergism" or "antagonism" between the two phenomena (references in Introduction). These two concepts are based almost entirely on the observation of either simultaneous or consecutive occurrence of moulting and reproduction in decapods. However, this terminology strongly implies that the temporal relations observed between moulting and reproduction are not merely circumstantial, but are a product of specific processes of mutual stimulation or inhibition. There is no compelling evidence of such processes, even in decapods. The present work suggests that this terminology is inappropriate, at least for isopods, and its implications potentially misleading. The finding that moulting may occur either with or without reproduction according to day-length suggests that the two processes are independent but coordinated responses to environmental cues and are not simply "synergetic" or "antagonistic." Thus, the occurrence of vitellogenesis during premoult in isopods ISOPOD MOULTING AND REPRODUCTION 215 is not interpreted as evidence of "synergism" between reproduction and moulting since there is no evidence that the occurrence of either process stimulates the other. The term "synchrony" as used to describe this phenomenon in Orchcstia (Meusy ct a!., 1977) has fewer functional implications and is considered preferable to "synergism." Adiyodi and Adiyodi ( 1970) suggest that "antagonism" characterizes the reproductive period in decapods, on the supposition that the demands of the gonad for metabolic reserves interfere temporarily with the growth of the integument. \\ bile the pause observed in early premoult of isopods during which vitellogenesis commences could be construed as illustrating this notion, most of the oocyte growth was found to occur in late premoult when cuticle changes are proceeding concur- rently. Hence, it is not the metabolic demands of the ovary which produce the pause in premoult. Similarly, the prolongation of intermoult during brood incuba- tion could also be construed as "antagonism." However, there is no need to postulate an inhibition of moulting at this time ; the initiation of premoult by sen- sory input from the brood pouch indicates coordination rather than "antagonism" between moult initiation and brood incubation. Implications for hormonal control. Substantial evidence exists for many crustacean species of inhibitory neurohormonal control by the brain of both moult- ing and reproduction (reviewed by Gabe, 1966; Adiyodi and Adiyodi, 1970; Sochasky, 1973). There is continued debate over whether separate moult- and gonad-inhibiting hormones (MIH and GIH) occur or whether both processes are regulated by a single "growth inhibiting principle" (Panouse, 1947). As in deca- pods, there are conflicting reports concerning the effects of brain lesions on moult- ing and reproduction in isopods, (reference in Introduction). The present report is the first to examine the mechanisms coordinating moulting and reproduction in normal isopods. Certain requirements of any concept of hormonal control of these processes may now be inferred. First, separate moult- and gonad-regulating hormones are necessary. All the above arguments that moulting and reproduction are separate but coordinated processes imply separate hormonal mechanisms; it is not possible to explain the interactions found between these processes in terms of a single "growth inhibitory principle." This conclusion is further supported by the finding that one of the groups of neurosecretory cells in the brain of Oniscns undergoes cytological changes correlated with the moult cycle while those of another group are correlated with vitellogenesis (unpublished observations). Second, more than one hormone seems to lie involved in vitellogenesis. Ecdysone appears to be necessary for vitellogenesis in Porccllio (Besse and Mais- siat, 1971 ). The present finding that vitellogenesis is always completed in late premoult, when whole body content of ecdysteroids peaks in other isopod species (Charmantier ct a!., 1976; Hoarau and Him, 1978) suggests that ecdysone may be necessary for the completion of vitellogenesis. However, it is highly improbable that ecdysone initiates vitellogenesis, since it may commence either in intermoult or early premoult. Rather, its initiation may be controlled by some other principle (GIH or analogous factor), perhaps acting to stimulate production of vitellogenins. Regulation of the release of this principle by daylength would explain the observed effects of daylength on reproduction. The role of ecdysone might then be in the sub- sequent uptake of vitellogenins by the oocytes, as has been suggested in Orchcstia (Meusy ct al, 1977; Blanchet ct «/., 1979). The regulation of vitellogenesis 216 C. G. H. STEEL seems to be different in isopods and decapods, for the whole of vitellogenesis usually occurs during intermoult in the latter (Adiyodi and Adiyodi, 1970, for review). Third, support is given to the proposal that isopod ovaries containing vitello- genic oocytes produce an "ovary hormone" which induces differentiation of a brood pouch (Legrand, 1955; Balesdent, 1965). During first maternal moults, the pause in late D0 coincides with the beginning of vitellogenesis and ends with apolysis. Secretion of an "ovary hormone" would therefore commence at exactly the appro- priate time to elicit the reorganization of the epidermis to form a brood pouch under the influence of ecdysone later in premoult. During second maternal moults, the ovaries contain vitellogenic oocytes at the commencement of premoult, enabling prompt secretion of hormone to ensure that the epidermis retained its ability to differentiate another brood pouch. This work was supported by a National Research Council of Canada negotiated development grant to York University. SUMMARY In terrestrial isopods, different sensory cues initiate reproduction and moulting, indicating that the two processes are controlled by different physiological mecha- nisms. A specific sensory trigger which initiates premoult is identified ; it occurs well before conventional signs of premoult become evident. Specific coordinating mechanisms adjust the chronology of moulting and vitellogenesis under conditions promoting both processes. The first phase of vitellogenesis can occur either in intermoult or early premoult according to conditions and is considered to be independent of ecdysone. The second phase invariably occurs in late premoult and may be ecdysone-dependent. The relations between moulting and repro- duction are regarded as separately controlled processes which interact via specific cues which coordinate and adjust the timing of the two processes. Implications of this concept are discussed. LITERATURE CITED ADIYODI, K. G., AND R. G. ADIYODI, 1970. Endocrine control of reproduction in decapod Crustacea. Biol. Rev., 45 : 121-165. ADIYODI, R. G., 1978. Endocrine control of ovarian function in crustaceans. Pages 25-28 in P. J. Gaillard and H. H. Boer, Eds., Comparative Endocrinology, Elsevier/North- Holland, Amsterdam. ANURIEUX, N., 1979. L'apolyse au cours du cycle d'intermue de deux Crustaces Decapodes Brachyures Carcinas macnas Linne et Carcinas mediterraneus Czerniavsky. C. R. Acad. Sci. Paris, 288 : 1595-1597. BALESDENT, M. L., 1965. Recherches sur la sexualite et le determinisme des caracteres sexuels d'Asellus aquaticus Linne (Crustace, Isopode). Bull. Acad. Soc. Lorraines Sci., 5: 1-231. BESSE, G., 1968. Contribution at 1'etude du controle neurohumoral de la maturation ovarienne et de la mue parturielle chez 1'Oniscoide Porccllio dilatatus Brandt. C.R. Acad. Sci. Paris, 266: 917-919. BESSE, G., P. JTCHAULT, J.-J. LEGRAND, AND J.-P. MOCQUARD, 1969. Contribution a 1'etude de la physiologic sexuelle femelle de Liyia occanica L. (Crustace Oniscoide). Dif- ferenciation des oostegites et controle neurohumoral de la maturation ovarienne. C.R. Acad. Sci. Paris, 269 : 733-736. BESSE, G., AND C. DONADEY, 1972. Controle neuro-endocrine du fonctionnement des caecums digestifs des Crustaces Isopodes. Etude chez deux Oniscoides : Porccllio dilatatus ("Brandt) et I.igia occanica (L.). C.R. Acad. Sci. Paris, 275: 2387-2390. ISOPOD MOULTING AND REPRODUCTION 217 BESSE, G., AND J. MAISSIAT, 1971. Action de la glande de mue sur la vitellogenese du Crustace Isopode Porccllio ciilututus (Brandt). C.R. Acad. Sci. Paris, 273: 1975-1978. BLANCHET, M. F., P. PORCHERON, AND F. DRAY, 1979. Variations du taux des ecdysteroides au cours des cycles de mue et de vitellogenese chez le Crustace Amphipode, Orchcstia gammarellus. Int. J. Invert. Reprod. 1 : 133-139. BLISS, D. E., 1966. Relation between reproduction and growth in decapod crustaceans. Am Zool, 6: 231-233. CHARMANTIER, G., M. OLLE, AND J.-P. TRILLES, 1976. Aspects du dosage de 1'ccdysterone chez Sphaeroma serration (Crustacea, Isopoda, Flabellifera) et premiers resu'.tats C.R. Acad. Sci. Paris. 283: 1329-1331. CHARMANTIER, G., AND J.-P. TRILLES, 1975. Aspects du controle ncrveux des phenomenes de la mue chez Sphaeroma serratum ( Fabricius, 1787) (Crustacea, Isopoda, Flabelli- fera). C.R. Acad. Sci. Paris, 280: 2231-2234. CHARNIAUX-COTTON, H., 1973. Description et controle de 1'ovogenese chez les Crustaces superieurs. Inst. Nat. Recli. Agron. Ann. Biol. 13 : 21-30. CHARNIAUX-COTTON, H., AND L. H. KLEINHOLZ, 1964. Hormones in invertebrates other than insects. Pages 135-198 in G. Pincus, K. V. Thimann, and E. V. Astwood, Eds., The Hormones, Vol. 4, Academic Press, N. Y. DELALEU, J. C, AND A. HOLLEY, 1976. Contribution of an electrogenic pump to the resting membrane polarization in a crustacean heart. /. Exp. Biol. 64 : 59-74. DRACH, P., 1939. Mue et cycle d'intermue chez les Crustaces Decapodes. Ann. Inst. Oceanog. (Monaco), 19: 103-391. DRACH, P., AND C. TCHERNIGOVTZEFF, 1967. Sur la methode de determination des stades d'intermue et son application generate aux Crustaces. Vie Milieu. 18 : 595-609. GABE, M., 1966. Ncurosccrction, Pergamon, Oxford, 872 pp. HEELEY, W., 1941. Observations on the life-histories of some terrestrial isopods. Proc. Zool. Soc. Land. B., Ill: 79-149. HOARAU, F., AND M. HIRN, 1978. Evolution de taux des ecdysteroides au cours du cycle de mue chez Hcllcria brcvicornis (Isopode terrestre). C.R. Acad. Sci. Paris, 286: 1443-1446. LEGRAND, J.-J., 1955. Role endocrinien de 1'ovaire dans la differenciation des oostegites chez les Crustaces Isopodes terrestres., C.R. Acad. Sci. Paris, 241 : 1083-1085. LINDQVIST, O. V., 1972. Components of water loss in terrestrial isopods. Pli\siol. Zool., 45: 316-324. MAYES, K. R., AND D. M. HOLDICH, 1975. Water exchange between woodlice and moist en- vironments, with particular reference to Oniscus asellus. Contp. Biochem. Ph\siol. A, 51 : 295-300. McQuEEN, D. J., 1976. Porcellio spinicornis Say (Isopoda) demography. II. A compari- son between field and laboratory data. Can. J. Zool., 54 : 825-842. McQuEEN, D. J., AND C. G. H. STEEL, 1980. The role of photoperiod and temperature in the initiation of reproduction in the terrestrial isopod Oniscus asclh!>';/ ma, two types of neurosecretory cells can be distinguished by electron microscopy. The character of granules surely is a reflection of function, but whether tlu:s<- cells are concerned with different functions or not must be discerned in the future. SYMPLEGMA GANGLIA AND REPRODUCTION 225 FIGURE 5. Secondary lysosomes appearing in the neurosecretory cells of 10-day-old soli- tary zooids after isolation. Lys = secondary lysosome. A neurohemal region where neurosecretory materials are released into the cir- culatory system could not be identified in the present study. Axonal endings of neurosecretory cells may exist near the ganglion, since axons containing the neurosecretory granules were found only near the ganglion. A well-defined neuro- hemal organ such as the median eminence of vertebrates, the sinus gland of crusta- ceans, and the corpus cardiacum of insects was not found in Symplcgma. Such dif- fuse and short-range distribution of the axon terminals is known in the ganglia of molluscs (Frazier ct a/., 1967; Toevs and Brackenburry, 1969; Loh ct a!., 1973). Many studies concerning the mechanism of reproduction in ascidians have been carried out since Julin (1881) postulated a homology of the neural gland with the hypophysis of vertebrates. Previous studies examining possible gonadotropic potency of the neural complex showed contradictory results: Butcher, 1930; Hogg, 1937; and Carlisle, 1951, concluded that there was a positive function but Dodd, 1955, and Sawyer, 1959, showed negative function. However, the bioassay system of earlier investigators, using tissues of vertebrates, has to be reconsidered because there is no reason to suppose that similar phenomena in such phylogenetically distinct organisms are regulated by the same materials. For example, gonadotropin and progesterone affect oocyte maturation in vertebrates, but the same phenomenon in starfish is regulated by a peptide hormone from the nervous system and 1-methyladenine (see Kanatani, 1973, for review). Therefore, in examining hor- monal activity of a species, animals of at the very least the same phylum must be used in order for bioassays to have relevance. Many studies concerning the function of neurosecretory cells on gonad develop- ment indicate that distinctly contrasting control mechanisms operate in various 226 K. SUGIMOTO AND H. WATANABE 3.2 - Days FIGURE 6. Effects of ganglion ablation on growth of zooids. Average growth curves of 35 operated zooids (solid circles) and 27 control zooids (open circles) are depicted. species. For example, control of sexual reproduction by the neurosecretory sys- tem in some species is inhibitory in character, whereas a positive gonadotropic influence exists in other species. Neurosecretory materials in Hydra act somewhat like a growth hormone and inhibit gonad development (Schaller, 1973; Schaller and Gierer, 1973). Clearly, neurosecretory cells in Hydra affect asexual reproduc- Days 10 Days FIGURE 7. Effects of ganglion ablation on oocyte growth. Average oocyte number in 35 operated zooids (solid circles) and 27 control zooids (open circles) are depicted: (A) variation in the number of oocytes in the stage of previtellogenesis ; (B) variation in the number of vitellogenic oocytes. SYMPLEGMA GANGLIA AND REPRODUCTION 227 tion, that is, zooid growth and bud formation. The same inhibitory influences of neurosecretory cells were shown in nemertines (Bierne, 1970) and in nereid polychaetes (Clark, 1965; Baskin and (folding, 1970). On the other hand, a positive function has been demonstrated in many other invertebrates, such as turbellarians (Grasso and Quaglia, 1971), molluscs (Geraerts and Algera, 1976; Wijdenes and Runham, 1976), non-nereid polychaetes (Howie, 1966; Gouedard- Couadou and Vicente, 1971), and arthropods (Tombes, 1970; Adiyodi and Adi- yodi, 1970 for reviews) . In S\niplegma, the pattern of occurrence of Gomori-positive neurosecretory cells is different from that of Hydra. Active zooid growth and bud formation were observed in both young and aged zooids isolated from the central region of a colony, but they showed only a few such neurosecretory cells in their ganglia. As shown in Figure 3, variation in the number of Gomori-positive neurosecretory cells is superficially related to oocyte development. That is, the cell number increases rapidly when the oocyte proceeds from previtellogenesis to vitellogenesis. In addition, these cells decrease in number under conditions disadvantageous for gonad development. Thus, the neurosecretory cells in Symplegma seem to be functionally insignificant in asexual reproduction but to have some significance in sexual reproduction. If this is so, surgical removal of the ganglion should affect gonad development. The most distinctive effect of the operation was the break- down of oocytes in Stage 2, while young Stage 1 oocytes were mildly affected by the operation. Similar results have been obtained in Ciona (Bouchard-Madrelle, 1967). But studies in Chelyosoma showed that neural-complex ablation has no effect on gonad development (Hisaw ct al., 1966). Additional studies to com- pare single ascidians with compound ones or to compare species with distinct 150 E 100 u o o u de la totalie de complexe neural sur le fonctionnement des gonades de Ciona intcstinnlis ( Tunicier Ascidiace) C.R. Acad. Sci. Scr. I).. 264 : 2055-2059. BUTT HER, E. O., 1930. The pituitary in the ascidians (Moh/nla manliattensis) J Exp Zool 57: 1-11. CARLISLE, D. E., 1951. On the hormonal and neural control of the release of gametes in ascidians. J. Exp. Biol., 28 : 463-472. CHAMBOST, D., 1966. Le complexe neural de dona intestinalis L. (Tunicier, Ascidiacea). fitude comparative du ganglion nerveux et de la glande asymetrique aux microscopes optique et electronique. C.R. Acad. Sci. Scr. D.. 263 : 969-971. CLARK, R. B., 1965. Endocrinology and the reproductive biology of the polychaetes. Occtinoy. Marine Biol. Ann. Rev.. 3 : 211-255. DAWSON, A. B., AND F. L. HISAW, JR., 1964. The occurrence of neurosecretory cells in the neural ganglion of tunicates. J. Morp/wl., 114 : 411-423. DODD, J. M., 1955. The hormones of sex and reproduction and their effects in fish and lower chordates. Pp. 166-187 in I. C. Jones and P. Eckstein, Eds., Comparative Pliysiolouy of Reproduction, Memoirs of the Society for Endocrinology. Cambridge Univ. Press, London. FRAZIER, W. T., E. R. KANDEL, I. KVPKERMANN, R. WA/IRI, AND R. E COGGESHALL, 1967. Morphological and functional properties of identified neurons in the abdominal ganglion of Aplysia calijornica. J. Ncurophysiol., 30: 1288-1351. GERAERTS, W. P. M., AND L. H. ALGERA, 1976. The stimulating effect of the dorsal-body hormone on cell differentiation in the female accessory sex organs of the hermaphrodite freshwater snail, Lymnaca staynalis. Gen. Com p. EndocrinoL, 29. 109-118. GOLDING, D. W., 1974. A survey of neuroendocrine phenomena in non-arthropod invertebrates. Biol. Rev., 49 : 161-224. GouEDARD-CouADOu, E., AND N. VICENTE, 1971. Sur les correlations neuroendocrines chez Spirobis (Pilcolaria) militaris (Claparede), 1870 ( Polychaeta, Serpulidae). C.R. Seances Soc. Biol. FiL. 165 : 1092-1096. GRASSO, M., AND A. QVAGLIA, 1971. Studies on neurosecrelion in planarians. III. Neuro- secretory fibers near the testis and ovaries of Polvcelis niyriu. J. Submtcrosc. CvtuL. 3: 171-180. HANCOCK, A., 1868. On the anatomy and physiology of the Tunicata. /. Linn. Soc. Lond. Zool., 9 : 309-346. HERDMAN, W. A., 1883. The hypophysis cerebri in Tunicata and Vertebrata. Nature 28 : 284-286. HIGH NAM, K. C., 1962. Neurosecretory control of ovarian development in Schistocerca gregaria. Q. J . Microsc. Sci., 103 : 57-72. HISAW, F. L., JR., C. R. BOTTICELLI, AND F. L. HISAW, 1966. A study of the relation of the neural gland-ganglion complex to gonadal development in an ascidian, Chclyosoma production Simpson. Gen. Comp. EndocrinoL, 7 : 1-9. HOGG, B. M., 1937. Subneural gland of ascidian (Polycarpa tccta) : An ovarian stimulating action in immature mice. Proc. Soc. Exptl. Biol. Mcd., 35 : 616-618. HOWIE, D. I. D., 1966. Further data relating to the maturation hormone and its site of secretion in Arcnicola marina Linnaeus. Gen. Comp. EndocrinoL, 6: 347-361. IMAI, Y., A. SUE, AND A. YAMAGUCHI, 1968. A removing method of the resin from Epoxy- embedded sections for light microscopy. /. Electron Microsc.. 17 : 84-85. JULIN, C., 1881. Recherches sur I'organization des Ascidies simples. Sur 1'hypophyse et quel- ques organes qui s'y rattachent, dans les genres Corella. Phallnsia. et Ascidia. Arch. Biol. 2 : 59-126. KANATANI, H., 1973. Maturation-inducing substance in starfishes.. Int. Re?1. C\toL, 35: 253-298. LANE, N. J., 1971. The neural gland in tunicates : Fine structure and intracellular distribution of phosphatases. Z. Zellforscli.. 120 : 80-93. LANE, N. J., 1972. Neurosecretory cells in the cerebral ganglion of adult tunicates : Fine struc- ture and distribution of phosphatases. /. Ultrastruct. Res.. 40: 480-497. 230 K. SUGIMOTO AND H. WATANABE LOH, V. P., Y. SARNE, AND H. GEINKR, 1973. Heterogeneity of proteins synthesized, stored, and released by bag cells of Aplysia calif ornica. J. Coinp. Physio!., 100: 283-295. MILLAR, R. H., 1953. Ciona. Pp. 1-123 in L. S. Colman, Ed., Liverpool Mar. Biol. Com in. Mem., Vol. 35. Univ. Press, Liverpool. PERES, J. M., 1943. Recherches sur le sang et les organes neuraux des Tuniciers. Ann. Inst. Occanogr. Monaco, 21 : 229-359. Roi'LE, L., 1884. Recherches sur les Ascidies simples des cotes de Provence ( Phallusiadees). Ann. Mus. Marseille. 2: 1-270. SAWYER, W. H., 1959. Oxytocic activity in the neural complex of two ascidians, Chelyosoma productiim and Pyura haustor. Endocrinology, 65 : 520-523. SCHALLER, H. C., 1973. Isolation and characterization of a low-molecular weight substance activating head and bud formation in Hydra. J. Embryo!. E.rp. Morphol., 29: 27-38. SCHALLER, H. C., AND A. GIERER, 1973. Distribution of head-activating substance in Hydra and its location in membranous particles in nerve cells. /. Embr\ol. Exp. Morphol., 29: 39-52. SENGEL, P., AND D. GEORGES, 1966. Effects de Peclairement et de 1'ablation du complexe neural sur la pont de Ciona intestinalis L. ( Tunicier, Ascidiace). C. R. Acad. Sci. Scr. D.. 263 : 1876-1879. TOEVS, L., AND R. BRACKENBURRY, 1969. Bag cell-specific proteins and the humoral control of egg-laying in Aplysia calijornica. Coinp. Biochcm. Physiol., 29 : 207-216. TOMBES, A. S., 1970. An Introduction to Invertebrate Endocrinology. Academic Press, New York. 203 pp. WIGGLESWORTH, V. B., 1964. The hormonal regulation of growth and reproduction in insects. Adv. Insect Physiol.. 2 : 247-336. WIJDENES, J., AND N. W. RUNHAM, 1976. Studies on the function of the dorsal bodies of Agriolimax reticulatus ( Mollusca ; Pulmonata). Gen. Comfi. EndocrinoL, 29: 545-551. Reference: Biol. Hull., 159: 231-246. (August, 1980) THE ZOOGEOGRAPHY AND DIETARY INDUCTION OF BIO- LUMINESCENCE IN THE MIDSHIPMAN FISH, PORICHTI1YS NOTATUS JON A. WARNER' AND JAMES F. CASE Department of Biological Sciences. University of California. Santa Barbara, California 93106 The light generating- systems of Porichthys notatus, the midshipman fish, cross react with those of the ostracocl I7argnla ( = Cypridind) hilgendorfii (Cormier et al., 1967), indicating that Porichthys might be able to utilize exogenous luciferin to support luminescence. This remarkable possibility is not wholly unexpected since some fishes in the two genera Apogon and Parapriacanthus are thought to acquire their luciferin from a sympatric ostracod, V. hilgendorfii (Haneda et al., 1966, 1969; Tsuji et al., 1971). The research reported here reinforces possible dependence of the midshipman on an exogenous luciferin. P. notatus occurs in coastal waters from Baja California to Southeastern Alaskan waters (Wilimovski, 1954), a range encompassing many kinds of lumi- nescent organisms which might furnish luciferin (Tsuji et al., 1971). However, luminescent ostracods, the most likely dietary source of luciferin, were not known to occur within the range of the midshipman until Kornicker and Baker (1977) described Vargula tsitjii (Mycopoda; Cyprindinae). The existence of a luminescent ostracod closely related to V . hilgendorfii but, unlike it, sympatric with the southern midshipman population, heightens interest in the possibility that Porichthys might normally utilize an exogenous source of luciferin. Midshipman luminescence is generated by an adrenergically triggered, intra- cellular, luciferase-mediated oxidation of luciferin in photocytes in hundreds of ventral and lateral photophores (Greene, 1899; Greene and Greene, 1924; Nicol, 1957; Baguet and Case, 1971; Baguet, 1975; Anctil, 1977, 1979a). Case and Strause (1979) and Anctil (1979b) proposed that the characteristic cytoplasmic vesicles and microvillous borders of the photocytes are specializations for uptake and storage of luciferin from large body stores identified by Tsuji et al., 1971. Specimens of P. notatus from Puget Sound are not bioluminescent (Strum, 1968) even though their photophores are morphologically and ultrastructurally similar to those of the southern luminescent fish (Strum, 1969a and b; Anctil and Case, 1976), except for possible reduced amounts of flocculent ground substance in cytoplasmic vesicles of nonluminescent specimens from Puget Sound (unpublished observations). Midshipman fish collected off Southern California contain luciferin in all body tissues, with larger concentrations in photophores (Tsuji et al., 1971), while fish from southern Puget Sound are luciferin deficient, both as embryos and adults (Barnes et al., 1973). Induction of luminescence capability in non-lumi- nescent fish has been achieved with intraperitoneal injection of V '. hilgendorfii luciferin (Tsuji et al., 1972) and by force feeding of whole dried V '. hilgendorfii (Barnes et al., 1973). Either technique makes luminescence possible after about 4 days. The bioluminescent response to noradrenaline continues to increase for 1 Present address : Marine Biology Research Division, Scripps Institution of Oceanography, La Jolla, California 92093. 231 232 J. A. WARNER AND J. F. CASE 5 ^^ "a u « **? ~~ n x -& • o ° 1st « I u «» »• -° SJ 8 E O C 3 j2 S^ U, »c> 0 -> a OOO 00 O OO »O 00 OOOO 0000 cs o o o o o OO O OO OO 00 O O O O rO 00 O O ^^ *o ^D VO 10 o' 10 3 0 1 ^ I O r^ oo O O O 00 O O 10 to o *— * *•— o vO OO ro OO rt a _E. j^ <0 a»»r< a. a .S*b*J*» «t*a>-r' ro co '-l^-'--i . r §r O — n BRIGHT £> MIXED LEVELS Q NOT FOUND SAN FRANCISCO CABOSAN LUCAS FIGURE 1. Collecting localities showing presence or absence of Porichthys notatus and the extent to which they exhibit luminescence and ultraviolet excited fluorescence. Shaded zone indicates distribution of Vargiila tsujii. / minutes beginning immediately after a local subcutaneous injection of 0.4 ml of 0.001 nM DL-arterenol. A conical light guide, masked so as to limit recording to a single photophore of average size, served to control for the large variation in size of fish and photophores. Feeding experiments In attempts to induce luminescence by manipulating the diet, 15 (12-21 cm) non- luminescent fish from Saanich Inlet and Imperial Eagle Channel, Vancouver Island, Canada, were fed parts or whole individuals of several luminescent organisms sympatric with the southern, luminescent P. notatus. Organisms, details of prepa- ration and effects on luminescence are shown in Tables II and III. PORICflTIIYS BIOI.UMINESCENCE 235 FIGURE 2. Arrangement used in recording luminescence from a single photophore in an intact Porichthys. Anesthetized fish were fed by inserting an appropriate size gelatine capsule of food deep within the esophagus. Close observation during recovery from anesthesia guarded against regurgitation, a rare occurrence. Each feeding greatly exceeded TABLE II Materials used in luminescence induction experiments Organism Source Method of preparation Conyaulax polyhedra Renilla kollikeri Vargida hilgendorfii Vargida tsujii Gaussia princeps Ruphausia pacifica Cennadus sp. Ophiopsila californica Stenobrachius californica Cultures supplied by Dr. B. M. Sweeney Santa Barbara Channel Vicinity of Chiba, Japan, via Dr. F. H. Johnson Collected at underwater light off dock at Catalina Marine Laboratory Midwater trawls in San Clemente Basin Same Same Naples Reef, Santa Barbara, California Midwater trawls in San Clemente Basin Concentrated by low speed cen- trifugation during night phase 5 mm wide strip from edge of rachis Probably air dried. Samples used were brilliantly lumines- cent when wetted Lyophilized after freezing alive and stored over dessicant Fed alive to test fish immedi- ately upon capture Same Same Arm sections fed fresh to test fish Frozen alive. Ventral, photo- phore-containing parts fed after brief thawing 236 J. A. WARNER AND J. F. CASE the weight of V . hilgendorfii required to induce luminescence capahility in non- luminescent fish. Studies on fluorescence induction Rate of induction of fluorescence in non-luminescent fish was estimated on one non-luminescent fish (19.5 cm total length) from Bamfield, B. C, fed 361 mg of dried V . hilgendorfii. Subsequently three photophores were removed each day and examined with a fluorescence microspectrophotometer (NanoSpec/lOS, Nanometrics Inc., Sunnyvale, Cal.) Excitation illumination was from a mercury lamp filtered with UG 1 and BG 12 filters. Fluorescence was recorded at 524 nm. Behavioral studies Luminescent responses to mechanical, visual, and electrical stimuli were studied hoth in luminescent fish from Southern California and in one luminescence-induced fish from north of Cape Flattery, Washington. Single fish were placed in a small TABLE III Results of dietary luminescence experiments. Fish number Day Food Test for lumines- cence Fish number Day Food Test for lumines- cence j 0 Gonyatilax polyedra (-) 121 (+) 2 G. polyedra 9 0 Gaussia princeps (15**) ( "~| 4 G. polyedra 9 ( ~) 9 G. polyedra 23 ( — ) 17 ( — ) 72 Euphausia pacifica (35**) 32 Beef Liver ( — ) 73 E. pacifica (15**) 47 ( — ) 74 E. pacifica (16**) 53 Vargula higendorfii ( — ) 81 (— ) (312 mg*) 95 ( — ) 71 ( +) 10 0 Euphausia pacifica ( — ) 2 0 Renilla kollikeri ( — ) 4 ( — ) 4 R. kollikeri ( — ) 19 E. pacifica ( ~"| 15 R. kolliktri ( — ) 38 ( — ) 3 0 Vargula hilgendorfii (291 mg*) (-) 11 53 0 Euphausia pacifica 1=1 8 ( "1") 7 E. pacifica 12 Beef Liver ( 4o 12 E. pacifica + Beef Liver 36 ( •}•) 30 S— ) 142 ( 4~) 41 Vargula hilgendorfii — ) 0 Vargula tsujii (3**) ( — ) (343 mg*) 7 ( — ) 51 ( ~^~) 39 V. tsujii (110 mg*) 12 0 Beef Liver ( — ) 44 -)-) 23 ( "") 75 -J-) 38 Stenobrachius californica 107 50 Euphausia pacifica ( —) 140 . 67 E. pacifica ( ~ ) 5 0 21 70 Vargula tsujii (10 mg*) V. tsujii (63**) (+) 75 82 S. californica E. pacifica (3 6 0 25 Gaussia princeps (317 mg*) 13 119 0 Gennadas sp (4***, 3 cm (I) 7 0 Gaussia princeps (28**) ( — ) animals) 1 G. princeps (24**) 9 ( — ) /'' 9 — ) 14 0 Gennadas sp (4. 3 cm (— ) f 23 0 Gaussia princeps (12**) -) animals) (V "*«>s»_ 1 G. princeps (30) — ) ~"-— • 9 ( — ) 23 ( — ) 23 ( — ) 72 Euphausia pacifica (25**) 72 73 74 81 Euphausia pacifica (28**) E. pacifica (14**) E. pacifica (16**) (-) 73 74 81 £. pacifica (13**) E. pacifica (16**) (-) 95 (-) 95 ( — ) 103 Vargula hilgendorfii \ / 15 0 Ophiopsila californica (—) (273 mg*) (-) 20 (-) * Weight of lyophilized animals. V. hilgendorfii was probably sun dried in Japan and not lyophilized. ** Numbers of living animals, other than V. tsujii. which had been lyophilized. *** Number of 3 cm animals. PORICHTHYS BIOLUMINESCENCE 237 aquarium entirely within the view of an upward-facing photomultiplier tube (EMI 9781 B). Mechanical stimulation was accomplished by a solenoid-driven rod tapping against the aquarium. Photic stimuli were of two sorts. One was a colliniated beam from a 1.5 V incandescent lamp directed across the aquarium at right angles to the photomultiplier field of view. The second was one of two models of the photophore pattern of a 14-cm midshipman. One model presented the lateral aspect and the other a ventral view of the fish. These luminous patterns were produced by templates placed over two green-emitting Sylvania "Panelescent Night- lights." Low voltage A.C. delivered between silver wire grids at each end of the aquarium invariably induced luminescence in luminescence-competent fish and was used as a control test when other methods of stimulation failed. RESULTS Evidence for a discontinuous distribution of P. notatus P. notatus has been described as ranging from Sitka, Alaska, to Baja California (Hart, 1973). The twelve sites listed in Figure 1 and Table I were sampled to establish regions within this species distribution where luminescence occurs. Figure 1 and Table I show presence or absence of fish and the presence or absence of bioluminescence and fluorescence in collected fish. We were unable to find P. notatus between Cape Flattery and Cape Medicino. The existence of this hiatus within the distribution of P. notatus was confirmed from two other sources. The first, a survey of museum collections, revealed hundreds of specimens of the midshipman from Puget Sound and Southern California waters but only four specimens collected between Cape Mendicino and Cape Flattery. These were two specimens from Winchester P>ay, Oregon (Los Angeles County Museum of Natural History and Oregon State University Ichthyological Museum), one from off the mouth of the Columbia River, and one from Coos Bay, Oregon (O.S.U. Ichthyological Museum). The second confirmation came from a series of 660 trawls along the 100-fathom line between Oxnard, California, and Cape Flattery, Washington, whose results were made available by the NOAA Northwest and Alaska Fisheries Center (Rock Fish Study, 1977, unpublished). P. notatus was common from Oxnard to approximately Cape Mendicino. No specimens were recorded between Cape Mendicino and Cape Flattery. Thus the prevalence of P. notatus in Puget Sound and in Southern California waters is in striking contrast to its paucity along the Oregon coast. / Correlation of fluorescence ^vith bioluminescence in natural populations As indicated in Table I and Figure 1, fish from the areas examined in this study were tested for the correlation of fluorescence and bioluminescence. All fish able to luminesce were fluorescent. Monterey coastal waters are evidently the northernmost limit of the entirely luminescent population, since 53 fish from the San Francisco region included non-luminescent (8%) and weakly luminescent (38%) as well as normally luminescent fish (54%). All non-luminescent fish from the San Francisco region were non-fluorescent while those with subnormal lumi- nescence had less than maximal fluorescent capacity. North of the Oregon hiatus neither fluorescence nor bioluminescence was seen. South of Monterey, all fish examined were strongly fluorescent and bioluminescent. 238 J. A. WARNER AND J. F. CASE 10 9 8 7 o ^ 4 12345 DAYS FIGURE 3. The development of fluorescence in a non-luminescent Porichthys after feeding with Vargula. Each point represents the average of three isolated photophores measured three times. Vertical bars = standard deviation. Upper curve, Vargula fed ; lower curve, unfed non-luminous fish. Fluorescence scale arbitrary with 10 approximately equal to the typical fluorescence of a southern Porichthys. Fluorescence induction Onset of photocyte fluorescence was followed in a non-luminescent 19.5-cm fish after feeding with 361 mg of V. hilgendorfii. Little or no fluorescence was record- able spectrofluorometrically from isolated photophores before the third day post- feeding. Fluorescence increased markedly from the third to fifth days, when it exceeded the calibrated range of the instrument. During this period the control fish showed no significant amount of photophore fluorescence (Fig. 3). Dietary induction of luminescence None of 13 non-luminescent midshipman fish became luminescent when fed upon specimens of eight luminescent organisms sympatric with the Southern Cali- fornia midshipman, as noted in Tables II and III. The four survivors at the end of this experiment were fed approximately 300 mg each of Vargula hilgendorfii and all became capable of luminescence upon injection of noradrenaline. In a subsequent experiment, one 19-cm nonluminescent midshipman was fed 110 mg (dry weight; about 700 specimens) of Vargula tsujii and developed bright fluo- rescence and' bioluminescence, first noted 4 days after feeding. The ability to luminesce persisted in this fish until its death 120 days later. In this experiment a second nonluminescent, 17.5-cm fish shared the same seawater as the induced fish but was fed beef liver. It developed neither fluorescence nor capacity to luminesce during the induction period for the specimen fed V '. tsujii. At the con- clusion of the experiment this control fish was fed 10 mgs (dry weight; 63 speci- mens) of V. tsujii. Between 15 and 21 days later fluorescence was apparent and the fish luminesced in response to noradrenaline injection. Luminescence capacity persisted until death of the fish 70 days later. Luminescent capability and behavior Although Barnes et al. (1973) had shown that northern midshipman fish after dietary luminescence induction could be induced to luminesce by electrical stimula- PORICIITHYS BIOLUMINESCENCE 239 TARLIC IV Luminescent responses to stimuli Stimulus type Number of tests Percent response Response characteristics Latency of response (nis) Duration (seconds) Average Minimum A. Southern Porichthys (12 fish) Light beam 111 26 1300 ± 290* 1,000 5.58 ± 3.50 Model fish 113 30 1030 ± 180 500 4.43 ± 2.36 Mechanical 102 62 660 ± 180 200 7.46 ± 5.40 Electrical 13 84 — — 11 64 ± 8.41 B. Northern Porichthys (1 fish) (First series 82 days after luminescence induction by feeding Vargula tsujii) Light beam Mechanical 5 12 33 74 1110 db 160 767 ± 98 900 500 3.41 ± 1.30 12.00 ± 7.21 (Second series on same fish, 118 days after induction) Model fish 19 0 Fish appeared to be blinded, with cornea opaque, perhaps due to fungal attack Mechanical 14 71 760 ± 210 300 5.15 ± 2.07 * ± = Standard deviation. tion, indicating the persistence at least of normal peripheral neural pathways, it is of the greatest interest to determine if the induced northern fish has behavioral control over its artificially endowed capacity to luminesce. To partially answer this question the responses of normally luminescent fish and induced fish to four stimulus types were compared. In order of increasing effectiveness in evoking bio- luminescence these were: 1) a midshipman photophore model, 2) a light beam passed through the aquarium, 3) tapping the aquarium side and 4) an A.C. current passed through the aquarium. Table IV sets forth the results of these experiments. After preliminary tests showed one induced fish to respond to mechanical stimulation, a second fish was carefully examined. Tested with light beam and mechanical stimuli, its responses appeared to fall within the range of normally bioluminescent fish with respect to response latency and rise time (Table IV, Fig. 4). First tested 82 days after an inductive feeding, this fish was left undisturbed and without feeding for an additional 36 days, when its bioluminescent responses to light and mechanical stimuli were found to be still intact. Imme- diately afterwards it expired following a forced feeding. DISCUSSION Discovery that the luminescent ostracod Vargula tsujii induces luminescence when fed to nonluminescent Porichthys further supports the possibility that a dietary deficiency causes the non-luminescent state of northern Porichthys. This 240 J. A. WARNER AND J. F. CASE LUMINESCENT RESPONSES TO STIMULI Southern Fish m (D B. O (D I*/ Light Beam Model Fish Mechanical Northern Fish D. Fish 4 (110 mg V. tsujii) Mechanical E. Fish 4 Light Beam F. Fish 8 (273 mg V. hilgendorfii) Light Beam Time 5 seconds FIGURE 4. Photomultiplier records of behavioral luminescent responses of normally luminescent southern Porichthys and Far 40 day 5 10 15 DISTANCE FROM TIP (mm) FIGURE 2. Distribution of Carcinonemertes crraus along crab egg setae for broods less than and greater than 40 days of development. N = 884 worms, < 40 days; 851 worms, > 40 days. Worm grozuth Juvenile worms remain approximately 0.5 mm in length while on the host's exoskeleton. The worms grow only during their feeding period in the host egg clutch. Average worm weight increases slowly through the first 40 days of host brooding but then begins to increase more rapidly during the second half of the ovigerous period (Fig. 3). 100- WEIGHT PER WORM Ug) 50- -o WORMS ALONE -° WORM EGG-STRINGS INCLUDED o A 20 40 6O 80 DAYS AFTER HOST OVIPOSITION FIGURE 3. Average weight of Carcinonemertes errans (±s.e.) as a function of age of their host egg clutch. Dashed line includes the weight of the worm egg strings on a per worm basis. N = 196 crabs. CARCINONEMERTES ERRANS LIFE HISTORY 251 The worms' growth appears to be sensitive to their densities within individual host egg clutches. Analysis of samples of similiar age for the relationship of average worm weight to density shows that most of the decrease in growth rate occurs as density increases to about 10 worms/ 1000 crab eggs (Fig. 4). The appearance of C. crmns changes as growth and maturity proceed. The sizes of the proboscis chamber and the stylet remain constant (stylet about 50 /zm in length) throughout life on the host. As female worms begin to increase in size their epidermises become more transparent as the white spots in the epidermis separate, eventually allowing the gut pouches to become visible along the sides. Female worms take on a segmented appearance when mature, due to the alternation of ovaries and lateral outpockets of the gut. Fggs are fertilized internally and become visible through the epidermis at maturity. ITorni reproduction Copulation by specimens of C. crrans was never observed ; however, worms are found aggregated in clusters or in pairs from about 20-40 days after host ovi- position. Copulation may occur during this period, since several female worms isolated 30 days after host oviposition with approximately 100 crab eggs for food produced fertile eggs. Worms begin laying eggs after 65-70 days in the host egg clutch. Eggs are laid in a cylindrical gelatinous matrix which is wrapped around the host egg string. Worm egg strands can be found attached all along the host egg setae but are less frequent near the tips at the periphery of the egg clutch (Fig. 5). Fecundity of C. crrans is variable and appears to be density dependent. Worms held in the laboratory exhibited high variability in both the number of egg strands produced and the number of eggs per strand (Table I). 500- WORM EGGS PER EGG STRING 250-1 501 AVERAGE WEIGHT PER WORM (jug) 25- go0 °8 o o o f>af> — I — — I — 25 50 WORM DENSITY — i — 75 IOO FIGURE 4. Number of worm eggs per egg string and average weight per worm of similarly aged samples as a function of worm density in the egg clutch (worms per 1000 crab eggs). N = 46 samples for eggs per string, N = 37 samples for average weight. 252 DANIEL E. WICKHAM 15- % WORM EGG STRINGS IC 5- 10 20 DISTANCE FROM TIP (mm) FIGURE 5. Distribution of egg strings of Carcinonemertes crrans along host egg setae. X =213 worm egg strings. Total fecundity per worm declined as the number of worms in 100-ml jars increased, for a limited number of samples (Fig. 6). In field samples the ratio of worm egg strands to worms ranges up to 7.12. The average was 1.92 egg strands per worm in 73 samples of crab eggs in their last 10 days of development (more than 80 days after host oviposition). The average number of eggs per strand, based on weight measurements of 2500 egg strands, was 265. In these samples the number of eggs per strand declined with density, with most of the decline occurring at the lower densities (Fig. 4). Infestation rates Carcinonemertes crrans differs from other congeners, being present on over 98% of all potential hosts. All non-egg-bearing crabs above 20 mm carapace width in the Bodega Bay area in Central California had worms. Similarly, only 10 samples from crab egg clutches out of more than 500 collected over a 5-year period contained no worms. Worm density (worms per 1000 crab eggs) in samples varied from 0 to 101.8 in one sample collected outside San Francisco Bay. Average worms per 1000 crab eggs for seasonal crab-egg collections from Pacific coast localities varied from a low of 0.66 in Alaska in 1979 to 25.41 just outside San Francisco Bay in 1977/78 Table II). Worm density per crab in those collections in which the size of the crabs was measured varied from 1430 in Alaska in 1979, to 20,580 in Washing- ton in 1978/79. Using the average fecundity of the measured crabs it was possible TABLE I Number of egg .strands produced by worms in the laboratory. The number of egg strands per worm and total fecundity data are based on average values per container. Individual fecundity could only be measured in jars with single worms. N = 29 worms. Kgg strands per worm Eggs per egg strand Fecundity per worm Average Range 3.1 1-8 446 79-1409 1339 847-4784 CARCINONEMERTES RRRANS LIFE HISTORY 253 5000-1 3000- TOTAL FECUNDITY . PER WORM 1000- 2468 WORM DENSITY PER 1000 CRAB EGGS FIGURE 6. Fecundity per worm for Carcinonemertcs rn'tins held in 100-ml jars at various worm densities. N = 9 samples. to estimate an average of 43,026 worms per crab outside San Francisco Bay, in 1977/78. DISCUSSION The larval life of Carcinonemertcs is poorly known. Based on the timing between maximum hatching and maximum recruitment to hosts by C. epialti on the host Hcuiigrapsus orcgonensis by Kuris (1978) and Roe (1979) the larval life TABLE II Average worm densities (worms per 1000 crab eggs) for yearly samples along the Pacific coast of North America along with coefficients of variations in density and projected average crab egg mortality from the sampled populations. Sample collection Number of samples Mean worm density Coefficient of variation Mean worms per crab (for measured crabs) Projected % crab egg mortality Kureka 1974-75 32 1.78 182.0% 10.6 1975-76 41 4.98 1177.3% 28.5 1977-78 50 4.03 221.1% 6901 30.0 1978-79 35 5.90 425.8% 12,049 43.9 San Francisco 1974-75 37 14.59 886.6% 63.3 1975-76 86 8.56 1070.9% 50.3 1976-77 44 10.02 823.9% 45.1 1977-78 48 24.96 2905.5% 62.5 1978-79 33 7.76 542.1% 44.6 Washington 1975-76 16 2.25 338.7% 4647 15.9 1978-79 19 11.14 265.4% 20,580 61.2 Alaska 1978-79 30 0.66 125.8% 1430 7.6 254 DANIEL E. WICKHAM of this worm species appears to be about 8 months. Juvenile C. crrans were not found on young-of-the-year Dungeness crabs until August or September, when the newly settled hosts exceeded 20 mm carapace width. Older crabs carried too many worms to accurately count and worms could settle on them at any time during the year. But the worms arriving on young-of-the-year crabs have to be at least 8-9 months old, having hatched the previous Decmber or January at the end of host brooding. Carcinonemertes appears to have a relatively long larval period when compared to many other planktonically dispersed marine invertebrates. Worm numbers increase with crab size during the first few months of the crab's benthic existence. This increase appears to occur throughout the host's life, since adult crabs could accrue in excess of 100,000 worms. Kuris (1978) found a similar increase of C. cpialti on //. orcgonensis of increasing size. Since H. oregoncnsis sheds its worms at molting, he proposed that worms settle continuously during the host's intermolt period and larger numbers occurred on larger crabs due to their longer intermolt periods. Recruitment to hosts by C. crrans differs from C. cpialti. C. crrans juveniles move from the exuvium to the new exoskeleton when the host molts, so hosts can acquire increasing numbers throughout their lives. P. Roe, J. Crowe, L. Crowe, and D. E. Wickham (in preparation) demonstrated that juvenile C. crrans actively absorbs dissolved primary amines from sea water and that leakage of amines across the uncalcified portion of the host's exoskeleton appears sufficient to meet dormant worms' metabolic needs. It is possible that worms arriving on a female host may remain for several years until the host broods eggs. The sites of infestation by juvenile Carcinonemertes on Pacific coast hosts differ from those described on Atlantic hosts by Coe (1902) and Humes (1942). Indi- vidual specimens of C. carcinophila on the Atlantic blue crab, Callincctcs sapidus, are found encysted between the gill lamellae on non-ovigerous hosts. Hopkins (1947) found that these worms could move back and forth from gills to egg clutches through the summer, since C. sapidus no longer molts after maturity and produces several broods during the summer. Carcinonemertes can reach a length of more than 70 mm on blue crabs, compared to a maximum of about 10 mm on Pacific coast hosts, which brood only once a year. Feeding by C. crrans begins when female Dungeness crabs oviposit their broods during October and November. This host carries up to 2,500,000 eggs for a period of approximately 90 days. C. crrans juveniles move into the egg clutch a day or two after host oviposition. During its period in the host egg clutch C. crrans differs from all other described Carcinonemertes in the lack of a sheath. Members of other species secrete a mucous sheath in which they live. Often males and females will be found together in these sheaths. C. crrans remains free-crawling while on host eggs, secreting copious quantities of mucus, apparently for adhesion. All measurable growth of C. crrans occurs during its period in the host egg clutch. Wickham (1979b) found that the average size of worms within a population from a single host was correlated with the number of crab eggs eaten per worm on that host. Female worms are larger than male worms. However, the opacity of the epidermis in C. crrans makes sex determination more difficult than in other local species which are more transparent. Worm eggs hatch near the time of host eclosion and the worm larvae become planktonic. They remain so in the laboratory for at least one month. Thus it is CARCINONEMERTES ERRANS LIFE HISTORY 255 unlikely that they reinfest the parental host as suggested for C. carcinophila by Humes' (1942). The density-dependent decline in fecundity found in laboratory-reared worms and the decrease in size of worm egg strings in field samples correlate with the decline in feeding and growth (Wickham, 1979b). This decline may reflect the action of some form of active intraspecific interference by C. errans. It occurs before food resources appear to be limiting, since in the test tube cultures crab egg mortality never exceeded 4%, even in tubes with eight worms. Carcinonemertes errans is so far the only described species of Carcinoncmcrtcs restricted to a single host. C. carcinophila has been described from 26 host species in the Atlantic, Caribbean, and Mediterranean (Humes, 1942; Kirsteuer, 1966; Vivares, 1975; Norse, 1975). C. epialti has been described on nine hosts (Humes, 1942; Kuris, 1978; Roe, 1979) and C. mitsukurii from five hosts ranging from Japan to the Indo- Pacific and Hawaii (Humes, 1942). Wickham (1978) indi- cated that morphological characteristics in the sheaths of the other locally occurring Carcinoncmcrtcs spp. suggest that host specialization in this genus is more common than previously thought. The biology of C. errans appears to adapt it specificially to its host. The fact that it was never found on alternate hosts such as Cancer gracilis, which shares the habitat of the Dungeness crab, suggests an ability to recognize its host. Its ability to transfer from the old exoskeleton during molting suggests a sophisticated behavioral repertoire, and the differential distribution on male and female exo- skeletons shows an ability to discern the sex of the host. Timing of development during the worm's trophic period appears to be an important adaptation in C. errans, especially when compared to worms on other hosts. C. errans matures and lays its eggs within about 2 weeks of the end of the host's brooding, after 65-70 days on the egg clutch. The time of development to worm hatching was not measured, but hatching was found to occur usually just prior to and during host hatching. This arrangement of timing would appear to allow worms the maximum amount of time for feeding on host eggs. Kuris (1978) and Roe (1979) found a similar synchronicity in the development of C. epialti on the host H. orcgoncnsis. Egg predators abound in nature (Orians and Janzen, 1974), but few specialize to the extent that Carcinoncmcrtcs does. The lifestyle of this genus shares features with parasitic castrators ; however, Carcinonemertes is a true predator since it eats several distinct prey individuals during its life (Kuris, 1974). The widespread occurrence of Carcinonemertes on numerous species of brachy- uran crabs, many of great ecological and economic significance, gives these ne- merteans an important role in benthic marine communities. This role is only beginning to be unraveled and the interaction of these worms with their hosts should provide stimulating material for a wide variety of studies on the adaptation of organisms to specialized environments. This work is the result of research leading to the Ph.D. degree from University of California, Berkeley. I would like to thank Cadet Hand for continuing advice and guidance throughout this study. I would also like to thank Louis Botsford, Lynn Suer, Dennis Hedgecock, and Keith Nelson for criticism of this manuscript. This work is the result of research sponsored by NOAA, Office of Sea Grant, Department of Commerce, under Grants No. NOAA 158-44121-R/A-19 and No. NOAA M01-189-R/F-52. 256 DANIEL E. WICKHAM SUMMARY 1. Specimens of Cancer inayister below 20 mm carapace width are not infested by Carcinonemertes crrans. Worms infesting young-of-the-year crabs beyond this size would have been in the plankton for 8-9 months prior to host infestation. 2. Nemertean burden on crabs increases with the crab's time on the bottom, at least through the crab's early life. Worms move from the host's exuvium to its new exoskeleton upon host molting. 3. Juvenile specimens of C. crrans were localized under the abdomen, near the copulatory appendages on male crabs. Juvenile worms were found in protected spots all over female crabs' exoskeletons. 4. Nemerteans migrate to the host egg clutch within a day or two of host ovipostion. They are peripherally distributed in the egg clutch through the early part of host brooding, but descend into the clutch to become more evenly distributed in the latter part of the brooding period. 5. All measurable growth occurs during C. errans' feeding period in the host egg clutch. Growth is inhibited as the density of worms per egg clutch increases. 6. Carcinonemertes crrans matures approximately 60-70 days after host ovi- position. Worms lay an average of 3.1 egg strings in the laboratory, each con- taining an average of 446 eggs. Fecundity per worm declines as worm density within the host egg clutch increases. 7. More than 99% of all specimens of C. inagistcr are infested by C. crrans in California waters. Numbers per host can range as high as 100,000. 8. Worms exhibit a contagious distribution among hosts, with the variance in density per host in excess of the mean density in all sample collections. The ratios of variance to mean increase as mean density increases. LITERATURE CITED BOTSFORD, L. W., AND D. E. WICKHAM, 1978. Behavior of age-specific, density-dependent models and the Northern California Dungeness crab fishery. /. Fish. Res. Board Can.. 35 : 833-843. COE, W. R., 1902. The nemertean parasites of crabs. Am. Nat.. 36: 431-450. HOPKINS, S. H., 1947. The nemertean Carcinonemertes as an indicator of the spawning history of the host, Callincctcs sapidus. J. Parasitol., 33 : 146-150. HUMES, A. G., 1942. The morphology, taxonomy and bionomics of the nemertean genus Car- cinoncincrtcs. III. Biol. Monot/r.. 18 : 1-105. KIRSTEUER, F., 1966. Uber Carcinonemertes carcinophila aus der Nordadria. Zoo/. Anz., 176 : 205-212. KURIS, A. M., 1971. Population interactions between a shore crab and ti\.'o syinbionts. Ph.D. thesis, University of California, Berkeley. 431 pp. KURIS, A., 1974. Trophic interactions : Similarity of parasitic castrators and parasitoids. Q. Rev. flio/.. 49: 129-148. KURIS, A., 1978. Life cycle, distribution and abundance of Carcinonemertes cpialti, a nemertean egg predator of shore crab, Hewiigrapsus oregonensis, in relation to host size, reproduction and molt cycle. Biol. Bull.. 154: 121-137. NORSE, E. A., 1975. The ccolo 1/16 (Fig. 1). It was unusual to detect spectrophoto- metrically any E lysis at coelomic fluid dilutions greater than 1/128. The rate of sheep E hemolysis exceeded that of rabbit E (Fig. 2). The initial rate of reaction was rapid : In 1 hr about 90% of a standard suspension of sheep E was lysed by 1/8 coelomic fluid, while under identical conditions — ' 60% of a comparable suspension of rabbit E was lysed. Total hemolysis of the sheep E was reached by 2 hr; the percentage hemolysis of rabbit E increased very slowly to about 70% by 24 hr. Hemolysis assays were routinely carried out at room temperature (~21°C). Lowering the temperature of incubation to 4°C had no effect on hemolytic activity: When the study presented in Figure 2 was repeated at 4°C, 262 ROBERT S. ANDERSON 100 80 10 "o 60 o> I 40 3* 20 Rabbit E .25 50 2 3 Time (hrs) 24 FIGURE 2. Kinetics of hemolytic activity of Glyccra coelomic fluid. Erthyrocytes (\%) \vere incubated with 1/8 diluted coelomic fluid at 21°C. Means ± standard deviations (vertical lines) (N = 5) are given for rabbit erythrocytes ; results of one representative experiment using sheep erythrocytes are included for comparative purposes. none of the mean percent hemolysis values were statistically different from those recorded at 21° C. Attempts to preserve hemolytic activity at — 80° C after quick freezing with ethanol and dry ice were unsuccessful ; apparently biological activity was lost during freezing and thawing. The hemolysin was inactivated by heating to 56° C for 30 min. If the coelomic fluid was treated with > 1 mM EDTA (ethylenediamine tetraacetic acid, disodium salt) at neutral pH, its hemo- lytic activity was markedly inhibited ( Fig. 3 ) . The hemolytic potential of coelomic fluid was considerably reduced by adsorp- tion with erythrocytes (Table I). Adsorption with rabbit E reduced subsequent lysis of rabbit E or sheep E to a comparable extent. Sheep E were more efficient in adsorbing anti-rabbit and anti-sheep hemolysins than rabbit E. Hemolysis of sheep E was more inhibited by adsorption with autologous E than was hemolysis of rabbit E, following each of the first two adsorptions. Hemolysin activity was determined daily for three days after a single injection of a 50% suspension of sheep or rabbit E (Table II). Injection with rabbit E produced no significant effect on anti-sheep hemolysin activity, but caused a reduction in the lysis of rabbit E, which lasted for the course of the experiment. Injection of sheep E caused a persistent reduction of both anti-sheep and anti-rabbit wiuu >% r^ JU T I o 0) I ^ 60 o 0 15 '£. 20 c. & 1 1 .1 .5 1 5 10 100 EDTA (mM) FIGURE 3. Inhibition of Glyccra anti-rabbit-erythrocyte lysin by EDTA. Rabbit red blood cells (1%) were incubated with EDTA-treated coelomic fluid (1/8 dilution) for 60 min. Mean % inhibition ± standard deviation (vertical lines) (N = 5) are given for each EDTA concentration. LYSINS AND AGGLUTIXINS OF GLYCRRA 263 TABU-: I Effect of adsorption idtli erythrocytes on (ilycern liemolvsin. /*,'+ for biological activity; for example, those from Lhinilits (Marchalonis and Edelman, 1968), lobsters (Hall and Rowlands, 1974), oysters (Acton et al., 1969), and tunicates (Anderson and Good, 1975). However, other invertebrate lectins do not show a dependency on divalent cations ; oyster hemagglu- tinin is active against sheep E and rabbit E in the absence of exogenous Ca-" (McDade and Tripp, 1967), and insectan hemagglutinin is active in the presence TABLE V Hemagglutinin titer in Glycera coelomic fluid 24 hr after an injection of erythrocytes (0.1 ml, 1(1' , E suspension}. Titer = Log-r^ x lowest coelomic ft u id concentration j-.iviny, visible agglutination of formaldehyde-treated erythrocytes (f E) ± SI) (N). Indicator E Uninjected Rabbit E injected Sheep E injected f Sheep E f Rabbit E 7.2 ±0.9 (10) 4.7 ±0.8 (10) 7.5 ± 1.1 (10) 4.7 ± 2.1 (10) 7.3 ± 0.9 (10) 4.4 ± 0.8 (10) 266 ROBERT S. ANDERSON of EDTA (Anderson et al, 1972). In lower vertebrates including sharks, rays, and paddlefish, and in higher vertebrates such as guinea pigs and man, agglutinins are not inhibited by EDTA; however, hemolysins are EDTA-sensitive (Gewurz et al., 1966). Both Glyccra hemolysin and hemagglutinin were heat-inactivated by holding at 56° C for 30-60 inin, as was the case for these humoral factors in oligochaetes (Cooper ct a!., 1974; Roch, 1979). Garte and Russell (1976) report that a polychaete hemagglutinin, amphitritin, is active at temperatures below 85 °C. Heat stability above — 70° C is uncommon for invertebrate lectins or hemolysins. Incu- bation of E with Glyccra hemolysin at temperatures of 4°-30° had little effect on the rate or extent of lysis. Sipunculid worm coelomic fluid also contains a hemolysin that has unaltered activity when incubated with E over a 0°-25° temperature range (Weinheimer ct al.. 1970). In this regard, it is interesting that the activity of hemolysins of poikilothermic vertebrates is increased at lower temperatures (Gushing, 1945; Gewurz ct al., 1966). Considerable activity of Glyccra hemolysin is lost after freezing and thawing, a reaction not typical of hemolysins from other invertebrates such as Mercenaria incrccnaria (Anderson, unpublished observation), the spiny lobster (Weinheimer ct al., 1969), and the earthworm (Roch, 1979). The hemagglutinin of Glyccra does not lose activity after storage at -- 80°C; this is typical of most invertebrate lectins (Pauley, 1974). Since both hemolysins and agglutinins are postulated to play a role in the immune reactions of invertebrates, their target cell specificity should be con- sidered. Based on our observations, both factors could react with E from several mammalian species : however, the intensity of the reaction against a particular type of indicator cell was different from that against another. Also, the hemolysin adsorption studies showed that, although there was considerable cross-reactivity, there was some evidence of specificity. While adsorption with either sheep E or rabbit E reduced hemolysin activity against both kinds of E, adsorption with sheep E reduced anti -sheep E lysin more than anti -rabbit E lysin. A similar directed effect on Glyccra lysins was not seen after adsorption with rabbit E. Adsorption of coelomic fluid with either E type reduced agglutination of both equally. It is not known if Glyccra hemagglutinins and hemolysins are single entities with rather broad specificities or if they exist in multiple forms with more limited specificities. Current data suggest that both of these factors in other annelids may exist in multiple forms. Garte and Russell (1976) isolated and purified a hemagglutinin from the polychaete Ainphitritc ornata. Sephadex G-100 frac- tionation yielded three active fractions with molecular weight of 30,000, 54,000, and 100,000. These fractions showed parallel specificity toward certain E, but different patterns of agglutination for other E types. The 30,000 dalton glyco- protein fraction (Amphitritin) probably was a subunit of the higher molecular weight agglutinin. Earthworm hemolysin was shown by the use of isoelectric focusing to be composed of four lipoprotein isoforms (Roch, 1979). Each iso- form had natural lytic activity against sheep erythrocytes. One isoform was found in all I*.iscnia jctida studied. The other isoforms were variably present. It was suggested that the natural hemolysin of Ilisctiia was induced and played a role during second set graft rejection (Chateaureynaud-Duprat and Izoard, 1977a). However, Roch (19/9) found no qualitative difference in the pattern of hemo- lytic proteins (isoforms) in grafted animals; the patterns remained stable regard- less of grafting, wounding, starvation, or age of the worms. LYSINS AND AGGLUTININS OF GLYCERA 267 It has been reported that hemolysins and hemagglutinins may be induced in terrestrial annelids by grafting or by the injection of erythrocytes (Cooper ct al., 1974; Chateaureynaud-Duprat and I/coard, l(^77b). We were unable to induce in Glyccra either hemolysins or hemagglutinins by either single or multiple intra- coelomic injections of erythrocytes. Cooper ct al. ( l'>74) injected Lumbricns with 0.1 ml of a 10','f E suspension; 24 lir later, lysis and agglutination of both sheep and rabbit E were enhanced. Chateaureynaud-Duprat and Izoard (1977b) reported that Lumbricns coelomic fluid had no agglutinating or lysing activity until 3—1 days after the injection of sheep E. Xone of our procedures involving single or multiple injections of 0.1 ml of 10 or 50'.; mammalian E caused any significant change in Glyccra hemagglutinin titers over a 72-96 hr period of observation. Xo induction of hemolytic activity was achieved by our protocols; in most cases, hemolysin activity was significantly decreased following E injection. This reduction was more pronounced in the coelomic fluid of those worms that received multiple E injections, suggesting that the hemolysin was simply removed from circulation by adsorption to the membranes of the injected erythrocytes. Studies of graft rejection and other manifestations of immunological memory among polychaetes have not been undertaken. However, our studies suggest that the naturally occurring hemolysins and hemagglutinins of Glyccra are not inducible. Therefore, one would not predict that these humoral factors would play a sig- nificant role in anamnestic immune reactions. This is not to exclude the possi- bility that the factors may have an important function in a more primitive immuno- logical recognition system. This work was supported in part by grant OCE-7723443 from the National Science Foundation, NCI Core Grant CA 08748, and a grant from the Whitehall Foundation. The author wishes to acknowledge the excellent technical assistance of J. D. Pearlman. SUMMARY 1. Naturally occurring hemolysins and hemagglutinins active against erythro- cytes from two mammalian species were found in the coelomic fluid of the polychaete Glyccra dibranchiata. 2. The hemolysin required divalent cations and was inactivated by freezing and thawing ; the hemagglutinin retained activity after freezing and was active in the presence of EDTA. Both factors were thermolabile above 56° C. 3. The rate of reaction and degree of intensity of lysis or agglutination was different for each type of test erythrocyte. These differences were found in all individuals studied, and were considered to be characteristic of the species. 4. Both hemolysin and hemagglutinin were readily adsorbed from coelomic fluid by the addition of erythrocytes. Adsorption with a particular type of erythrocyte resulted not only in reduced lytic and agglutinating activity against that cell type, but also in reduced activity against red blood cells from other species. 5. The hemolysins and hemagglutinins of Glyccra were not induced by single or multiple intracoelomic injections of erythrocytes. These experimental treat- ments had no significant effect on hemagglutinin titers, and usually caused a marked reduction in hemolysin activity. The consequence of the apparent lack of inducible humoral factors on possible anamnestic immune mechanisms, such as graft recognition and destruction, has yet to be evaluated. 268 ROBERT S. ANDERSON LITERATURE CITED ACTON, R. T., J. C. BENNETT, E. E. EVANS, AND R. E. SCHROHENLOHER, 1969. Physical and chemical characterization of an oyster hemagglutinin. /. Biol. Clicm., 244: 4128-4135. ANDERSON, R. S., N. K. B. DAY, AND R. A. GOOD, 1972. Specific hemagglutinin and a modulator of complement in cockroach hemolymph. Infect. Immun., 5: 55-59. ANDERSON, R. S., AND R. A. GOOD, 1975. Xaturally-occurring hemagglutinin in a tunicate Halocynthia pyriformis. Biol. Bull., 148 : 357-369. CHATEAUREYNAUD-DUPRAT, P., AND F. IZOARD, 1973. Etude des mecanismes de defense chez Lumbricus tcrrcstris. C. R. Acad. Sci. Paris, 276: 2859-2862. CHATEAUREYNAUD-DUPRAT, P., AND F. IZOARD, 1977a. Etude comparee in vitro des reactions de defense antigreffe chez 2 genres de Lombriciens : Eiscnia et Lumbricus. C. R. Acad. Sci. Paris, Seric D, 284: 2581-2584. CHATEAUREYNAUD-DUPRAT, P., AND F. IZOARD, 1977b. Compared study of immunity between two genera of Lumbricians : Eiscnia and Lumbricus. Pp. 33-40 in J. B. Solomon and J. D. Horton, Eds., Developmental Immunobiology. Elsevier/North-Holland Bio- medical Press, Amsterdam. COOPER, E. L., 1969. Specific tissue graft rejection in earthworms. Science, 166 : 1414-1415. COOPER, E. L., C. A. E. LEMMI, AND T. C. MOORE, 1974. Agglutinins and cellular immunity in earthworms. Ann. N. Y. Acad. Sci., 234: 34-49. GUSHING, J. E., JR., 1945. A comparative study of complement. I. The specific inactivation of the components. /. Immunol.. 50: 61-74. DALES, R. P., 1970. Annelids. Hutchinson & Co., Ltd., London. 200 pp. DALES, R. P., 1978a. The basis of graft rejection in the earthworms Lumbricus tcrrcstris and Eiscnia foetida. J. Invcrtcbr. Patlwl., 32 : 264-277. DALES, R. P., 1978b. Second-set graft rejections: Do they occur in invertebrates? Pp. 203- 220 in A. S. G. Curtis, Ed., Cell-Cell Recognition. Society for Experimental Biology Symposium Xo. 32. Cambridge University Press, England. DALES, R." P., 1978c. Defense mechanisms. Pp. "479-507 in P. J. Mill, Ed., Physiology of Annelids. Academic Press, London. DUPRAT, P., 1967. Etude de la prise et du maintien d'un greffon de paroi du corps chez le lombricien Eiscnia foetida typica. Ann. Inst. Pasteur (Paris). 113: 867-881. GARTE, S. J., AND C. S. RUSSELL, 1976. Isolation and characterization of a hemagglutinin from Amphitritc ornata, a polychaetous annelid. Biochim. Biophys. Acta, 439: 368-369. GEWURZ, H., J. FINSTAD, L. H. MUSCHEL, AND R. A. GOOD, 1966. Phylogenetic inquiry into the origins of the complement system. Pp. 105-116 in R. T. Smith, P. A. Miescher, and R. A. Good, Eds., Phytogeny of Immunity. University of Florida Press, Gainesville. HALL, J. L., AND D. T. ROWLANDS, JR., 1974. Heterogeneity of lobster agglutinins. I. Puri- fication and physicochemical characterization. Biochemistry, 13 : 821-827. HOSTETTER, R. K., AND E. L. COOPER, 1972. Coelomocytes as effector cells in earthworm immunity. Immunol. Commun., 1 : 155-183. HOSTETTER, R. K., AND E. L. COOPER, 1973. Cellular anamnesis in earthworms. Cell. Immunol., 9: 384-392. McDADE, J. E., AND M. R. TRIPP, 1967. Mechanism of agglutination of red blood cells by oyster hemolymph. J. Invcrtcbr. Patlwl.. 9 : 523-530. MARCHALONIS, J. J., AND G. M. EDELMAN, 1968. Isolation and characterization of a hemagglu- tinin from Limulus polyphcmus. J. Mol. Biol., 32 : 453-465. MARKS, D. H., E. A. STEIN, AND E. L. COOPER, 1979. Chemotactic attraction of Lumbricus tcrrcstris coelomocytes to foreign tissue. Der. Com par. Immunol., 3 : 277-285. PARRY, M. J., 1978. Survival of body wall autografts, allografts, and xenografts in the earthworm Eiscnia foetida. J . Invertcbr. Patltol., 31 : 383-388. PAULEY, G. B., 1974. Physicochemical properties of natural agglutinins of some mollusks and crustaceans. Ann. N. Y. Acad. Sci., 234: 145-158. ROCH, P., 1979. Protein analysis of earthworm coelomic fluid: 1) Polymorphic system of the natural hemolysin of Eiscnia foetida andrci. Dcv. Compar. Immunol., 3 : 599-608. WEINHEIMER, P. F., R. T. ACTON, J. E. GUSHING, AND E. E. EVANS, 1970. Reactions of sipunculid coelomic fluid with erythrocytes. Life Sci., 9: 145-152. WEINHEIMER, P. F., E. E. EVANS, R. M. STROUD, R. T. ACTON, AND B. PAINTER, 1969. Comparative immunology : Natural hemolytic system of the spiny lobster, Panulirus argus. Proc. Soc. Exp. Biol. Med.. 130 : 322-326. Reference: Biol. Bull.. 159: 269-279. (October, 1980) ALARM RESPONSE OF THE INTERTIDAL SNAIL LITTORINA LITTOREA (L.) TO i'REDATIOX BY THE CRAB CARCINUS MAENAS (L.) ROBIN P. HADI.OCK ' Department of BioloMO Many aquatic organisms rely on chemical senses to detect predators. Often avoidance behavior is elicited by distance or contact chemoreception of predator "odor" or "taste" (Mackie and Grant. 1974). Some species, however, have evolved alarm or escape responses to juices from the injured tissues of crushed conspecifics; these behaviors are found in minnows (von Frisch, 1938), amphibian tadpoles (Kulzer, 1954). sea urchins ( Snyder and Snyder, 1970), sea anemones (Howe and Sheikh, 1975) and gastropod molluscs ( Kempendorff, 1942; Snyder. 1967; Snyder and Snyder, 1971; Atema and Kurd. 1975; Atema and Stenzler, 1977; Stenzler and Atema, 1977). Snyder found in laboratory studies that 19 of 30 snail species tested respond to conspecific juice. He suggested that, in general, alarm reactions are responses to predation. Predators were tested for their ability to crush snails and elicit alarm responses in the laboratory. However, until Ashkenas and Atema (1978) reported that burrowing Ilyanassa obsolcta are rarely attacked by Carcinus manias in the laboratory, no studies had tested whether responding with alarm behavior helps an individual snail avoid being eaten. Direct field observations of predation, which could support the antipredator hypothesis, have been lacking. The present study was undertaken to test this antipredator interpretation. This paper describes the alarm response of Littorina littorea and field and labora- tory observations of Carcinus predation on L. littorea and presents results of studies testing the utility of alarm behavior in preventing crab predation. MATERIALS AND METHODS Alarm behavior of Littorina littorea Field experiments on the alarm repsonse of L. littorea were performed in tide pools of the rocky intertidal mid- and high zones at Bailey Island, George- town, and Harpswell, Maine. Snails found in small pools (< 25 cm deep, < 0.75 nr area) were tested in order to present snails with a high concentration of snail juice in tide-pool water. Snails responded to crushed conspecifics by moving to sites in the pool where they were less visible to a human observer. This response was measured by placing an octagonal grid (60 cm diameter, suspended from a circular plastic frame) over the tide-pool surface. Each trial lasted 60 min and consisted of a 30 min control period (min 0-min 30) followed by the experimental 1 Present address : Department of Biology, Osborn Memorial Laboratories, Yale University, New Haven, Connecticut 06520. 269 270 ROBIN P. HADLOCK period (min 30-min 60). At the beginning of the control period (min 0), two intact snails were dropped into the center of the grid area. At min 30 (beginning of experimental period) two crushed snails were dropped into the center of the grid area (N — 6 trials). The locations of snails visible under the grid were recorded at 10-min intervals for the full 60 min of each trial (after Atema and Burd. 1975). Wet weight of snail tissue added to tide pools was determined by shell length- wet tissue weight regression. Mean wet tissue weight of intact snails added to pools was 0.68 ± 0.07 g, N = 6 trials (in this and subsequent sections, values are reported as means ± one standard error). Mean wet weight of crushed snail tissue added was 0.48 ± 0.20 g, N == 6 trials. Fifty-nine ± 24.1 (range 17-122) snails were followed per trial. A second experimental series was designed to test for responses to chemical stimulation by the snail juice alone. At min 0 (beginning of control period) 6.25 ± 0.75 ml sea water was substituted for the intact snails of the first experi- mental series and at 30 min (beginning of experimental period) 5.45 ± 0.75 ml of filtered snail juice was substituted for the crushed snails (N = 4 trials). Snail juice was prepared at poolside just prior to each trial by crushing two individuals of L. lift or ea of known shell length (distance from apex to base of aperture) in a dish, adding sea water from another pool, and filtering the mixture through Whatman #1 filter paper into a 50 ml filtration flask. Both sea-water control and snail juice were released from a pipette into the center of the grid area; again, each snail juice test followed a control trial. Separate pipettes were used to avoid contamination between control and test stimuli. The concentration of snail juice added was estimated by first determining the wet weight of crushed tissue from shell length-wet tissue weight regression. The approximate concen- tration added was 0.12 g/ml of snail tissue in sea water before filtration and release into a tide pool. In each trial 66.5 ± 33.3 snails (range 35-106) were followed. In two additional blank trials the experiment was conducted in the same way but the test stimulus at min 30 was omitted. In each trial 52 ± 36.9 snails (range 3-101) were followed. To examine the rate of crawl of individual snails, in one Bailey Island pool 11 snails were marked individually with Pla enamel (Testers Corp., Rockford, II.). Positions of six individuals were recorded at 10-min intervals during a 30- min control (sea water) and a 30-min experimental test (filtered snail juice) period. This pool was tested on April 14 and May 3, but not every snail marked was present for both tests. The movements of these individuals were also recorded during a 50-min blank trial on April 22 during which sea water, but not test stimulus, was added. Prcdation by Carcinus maenas To test the effectiveness of the L. littorea alarm response in preventing crab predation, the times required for crabs to find snails in "sheltered" and "exposed" sites were compared in the laboratory. Also, the time required for snails to hide was compared to the duration of the "consume phase" of crab feeding behavior (Fig. 2). The first comparison was simplified by using only one type of sheltered site chosen by snails in the field: a rock crevice. Two round glass bowls (each 20 cm SNAIL ALARM TO CRAB PREDATION 271 diameter, 6.5 cm deep) filled with sea water to a depth of 6.0 cm were used to simulate tide-pool habitat. The bottoms of these bowls were lined with several flat rocks. Crabs were tested with "exposed" snails by placing snails in the center of a rock surface at one end of the bowl. In trials with "sheltered" snails, snails were placed in the approximately 2.0-cm-deep crevices formed between rocks. Crabs used in these experiments were collected by commercial fishermen in Rhode Island, held in a damp refrigerated room for 3 days, and then transferred to two large (20- and 45-gal ) aquaria in a recirculating sea-water system until experiments began 4 days later. Crabs were not fed during this time. Individuals of L. littorca were collected in Narragansett, Rhode Island, on the first day of the experiment and held in a damp glass bowl thereafter. Crabs were placed in the bowls, allowed to acclimate for 10-30 min, then removed for 2-5 sec while a snail was positioned in the bowl. Both pools were used in "sheltered" and "exposed" trials. Between trials, pools were rinsed with hot water and refilled with fresh sea water. Distances between crab and snail were the same (approximately 16 cm) at the start of all 12 trials. The experiment took place in a dark room with the pools lit by a microscope illuminator. Crabs were observed for 20 min or until they had picked up a snail and moved it to the "attack" position in front of the mouthparts. In a second experiment the time required for crabs to injure and consume snails was estimated using the same glass bowls. The goal of this test was to determine how long a crab takes to consume one snail before searching for the next, since this is the period of time available to intact conspecifics to find shelter. The "consume phase" of the predation sequence begins with first injury to the snail body and release of snail juice to the surrounding water. The exact time of first injury was difficult to determine, so this moment was standardized by equating it with a behavior involving sure injury, the "pull from mouthparts" (see below). The end of the "consume phase" was marked by completion of all feeding behavior. Each crab was placed with a small (< 9.0 mm shell length), medium-sized (^ 9.0, ^ 18.0 mm), and large (> 18.0 mm) snail (Underwood, 1973) and observed until all the snails in the bowl were consumed or the crab showed no searching behavior for 10 min after consuming a snail. Field observations of Carcinus-Littorirui interactions took place in tide pools at Appledore Island, Maine. Feeding crabs were found at night by sweeping the red beam of a 9 V lantern (lens covered with a #2423 red plexiglas disc) over pool bottoms. Crabs were also observed feeding along the stony bottom of a cove at high tide during the day. RESULTS Snail alarm behavior The proportion of snails visible in the grid area decreased significantly in the 10 min period following addition of crushed snail (P ^ 0.05) or snail juice (P ^ 0.025) relative to changes in the proportion visible 10 min after introduction of intact snails or sea water (angular transformation of proportions; analysis of variance for paired data; Sokal and Rohlf, 1969). Snails tested in these trials hid by crawling into crevices, under fronds of macroalgae, or under rocks. In one pool, snails grazing at the tips of Chondrus blades moved down among the blades toward the holdfast. There was no significant change in the proportion of snails visible during the same intervals of blank trials (Fig. 1). 77? ROBIN P. HADLOCK LLJ 100 _l m 80 o h- < CL O Q_ 60 40 CO 20 INTRODUCTION OF CONTROL STIMULI INTRODUCTION OF I TEST STIMULI 0 10 20 30 40 50 60 TIME (MIN) FIGURE 1. Percent of L. littorca populations visible following introduction of control and test stimuli, relative to percent visible at min 0. Stimuli introduced : open circle = intact /-. littorca control; closed circle = crushed L. littorca (N = 6 trials). Open triangle = sea water control; closed triangle = crushed L. littorca juice (N = 4 trials). Open box indicates blank trials : intact snail or sea water was added at min 0 but no test stimulus was introduced (N = 2 trials). Symbols and bars represent means ± 1 standard error. In general, snail activity in tide pools increased after addition of crushed snail or snail juice. Individuals in one pool increased rates of locomotion significantly, from 0.32 ± 0.07 cm/min (X = 6) in the 20-min interval preceding addition of snail juice to 1.40 ± 0.31 cm/min (N -- 6) in the 20-min period following addition of juice (P ^ 0.03 ; Wilcoxon's signed-ranks test). In applying this statistical test, it was assumed that snails responded independently, although no experimental test for independent responses was conducted. There was no significant change in crawling velocity during the same periods of the hlank trial. Crab feeding behavior Feeding crabs were observed in the laboratory and in the field, description of feeding behavior was based on laboratory observations. Detailed SNAIL ALARM TO CRAB PREDATION 273 Search phase. Feeding behavior begins when the crab detects, apparently by olfaction, the presence of nearby snails. First the antennnle flicking rate increases (antennule beat, Fig. 2) and antennule position changes from primarily vertical to pointing at different angles from the carapace (antennnle point, Fig. 2). The third maxillipeds then begin to sway from side to side and one may be wiped over the other several times. This may last for several minutes until the crab begins to move forward. The advance is accompanied by chelae and walking-leg raking, in which the chelae and walking legs are extended from the carapace and Alert: Increased antennule beat Antennule point Maxilliped wipe Advance Chelae rake Walking leg rake Dactyl snap Contact snail shell SEARCH (Pounce) I Turn and test shell I Crack CRACK Pull from mouthparts CONSUME ' 1 Probe ISift --|Groom: Eye brush Antennae brush FIGURE 2. Sequence of crab (Carcinns inacnas) predatory behavior. Dashed lines indi- cate points at which crabs may return to search or cracking behavior after having begun to consume a snail. This greatly lengthens the consume phase and increases the time available for intact snails to hide. 274 ROBIX P. HADLOCK swept across the substratum with a semicircular swiping motion. While raking, the dactyl of the claw opens and closes (dactyl snap, Fig. 2). Crack phase. Crabs begin their attack after contacting the snail shell with walking leg or claw. The crab may simply pull the snail from the substratum and bring it to the attack position in front of the mouthparts. Or the crab may suddenly pounce on the snail, pinning it between the carapace and substratum and then pushing the shell forward toward the mouthparts with the walking legs. Once the shell is in front of the mouthparts, the crab turns the shell over with the chelae, pausing to insert the dactyl of the claw into the shell aperture (probe, Fig. 2). The crab then removes the dactyl from the aperture and resumes turning the shell, stopping occasionally with one claw around the shell spire and the other supporting the shell. The third maxillipeds help support the shell during this turning and testing. If the snail is small relative to crab size, the crab quickly crushes the shell with a claw or breaks off the top of the spire. If the shell is too large to crush, the crab uses alternative methods to expose the snail body ; either chipping the outer lip of the aperture until the operculum is no longer flush against the shell and then grasping the snail body behind the operculum with one claw while the other claw tugs the shell in the opposite direction ; or gradually chipping away the side of the shell. Either of these techniques requires further cracking of the shell after the first mouthful of snail tissue has been taken. Consume phase. Once the snail body is exposed, the shell is held up to the mouthparts, supported by both chelae, and the mandibles and maxillipeds tear off bits of flesh. A small cloud of fluid appears around the mouthparts. While mouthparts grip the snail body, the shell is pulled away with the claws, exposing more snail body, until it is consumed. Occasionally the shell is dropped when the snail is only partially consumed but the crab is unable to crack more of the shell. If the shell has been crushed or broken, the crab picks up the fragments again (sift, Fig. 2) after consuming the snail body. At the end of the "consume phase" the crab sits quietly, resumes searching, or grooms. A summary of the entire predation sequence appears in Figure 2. In the field, crabs were usually discovered holding periwinkles in the "attack" position, but on one occasion the entire sequence of feeding behavior (search through consume) was observed in a tide pool. Crab predation: laboratory experiments Attention was focused on two periods of this predatory behavior to test the utility of snail alarm behavior in preventing crab predation. The two periods were the "search phase" (time between a crab's becoming alerted to the presence of a snail and its attack on the shell) ; and the "consume phase" (the interval between first injury to snail body and the start of a search for the next snail victim). If responding to snail juice helps snails avoid crab attack, then crabs should require more time to locate and attack sheltered than exposed snails. Also, the response time of snails to snail juice should be less than or equal to the duration of the "consume phase" of crab feeding behavior. Crabs found and began attacking exposed snails in approximately 4 min, but required longer than 16 min to discover and begin attack on snails in crevices (P ^ 0.005, Wilcoxon two-sample test. Table I). Only three of six crabs tested SNAIL ALARM TO CRAB PREDATIOX 275 Results of tests comparing crab predatimi i»i I., littnrra in en-vices or exposed on rock snrfin cs: .Means ± one standard error. Crah size (carapace width in mtn\ exposed trials: 47.74 ± 7.75," sheltered trials: 47.44 ± 1.46. Snail size (shell length in mm) exposed: 972 ±110 P < : 0.005 range (13 465) (490-1199+) * Trial terminated at 20 min even if crab hadn't yet attacked. with sheltered snails were able to find snails and attack within the 20 min limit of a trial. Crabs became alerted (signaled by antennule pointing) to snail presence equally quickly in both cases, hut took longer or were unable to find sheltered snails. Also, once a crab's walking legs or claws had contacted snails in crevices, crabs seemed to have difficulty performing the claw movements required to extract snails from crevices. The time required to consume individuals of L. littorea depended on snail size (shell length). The regression equation relating snail size and time required to consume snails was: In consume time = -- 3.28 + 0.48X shell length, R- = 0.40, N - : 16. Crabs took longer to consume medium-sized snails than small snails (P^O.05, analysis of variance). This difference reflects different methods of attack on the two size classes of snail. All crabs consuming medium-sized snails interrupted actual feeding on snail tissue to resume attack on the shell. Small snails' shells were usually crushed immediately. Large snails were attacked (65%) but none consumed (Table II). DISCUSSION The size, shape, and structure of gastropod shells is often considered a snail's single or primary defense against shell-destroying predators such as birds, fish, TABLK II Crab predation success and the amount of time required to consume small, medium-sized, and large individuals of L. littorea. Snail si/.e Small Medium Large Number of snails presented 17 17 17 Number (proportion) attacked 13 (0.76) 14 (0.82) 11 (0.65) Number (proportion) consumed 8 (0.47) X M1.47) 0 (0.00) Consumed snails Snail shell length (mm) 7.88 ± 0.20 10.53 ± 0.29 — Crack phase duration (sec) 30 db 6 1164 ± 442 — range (2- 45) (25-3625) Consume phase duration (sec) 134 ± 52 594 ±277 P ^ 0.05 range (15-480) (163-2505) 276 ROBIX P. HADLOCK and decapod Crustacea (Heller, 1976; Vermeij, 1974. 1976, 1978; Vermeij and Covich, 1978; Hughes and Elner. 1979; Zipser and Vermeij, 1978). In this paper I have assembled evidence for an alarm response of L. liitorca and its function as a complementary antipredator device. To test the hypothesis that alarm behavior in this snail is an antipredator adaptation, answers to two questions were sought : Do crushing predators prey on L. littorca in the field? Is the snail's alarm behavior adapted to predator search and feeding behavior? Answers were derived from laboratory and field observations of crab predation, and from results of field studies of snail alarm behavior. Although further analysis of this behavior would require identification of the alarm substance, such tests were not included in this study. Both direct and circumstantial evidence suggest that crabs feed on periwinkles in the field. Carcinus was observed eating L. littorca in tide pools and a stony- bottomed cove. The abundance of broken shells found with shell injuries matching shell damage known to have been inflicted by Carcinus in the laboratory suggests that crab predation is not a rare event. Three characteristics of snail alarm behavior seem adapted to defense against the search and feeding behavior of Carcinus : the form of the alarm response, the means by which alarm is communicated, and the time taken by snails to hide. Individuals of L. littorca responded to juices of crushed conspecifics by increas- ing crawl velocities and moving toward rock crevices and under macroalgae fronds. Thus, the result of alarm behavior is movement to sites where snails are less likely to be stumbled upon by crabs. A snail in a sheltered site is more likely to avoid detection or attack than a snail exposed on a rock surface (Vermeij, 1974) or on the tide pool floor. In the present study it was found that crabs required more time to find periwinkles in crevices and were less successful in attacking once these sheltered periwinkles were found. It is likely that sites under rocks provide a similar refuge. The majority of gastropod species tested by Snyder (1967), including the mud snail Ilyanassa obsolcta, responded to conspecific juice with self-burial. In the laboratory, buried individuals of /. obsolcta were attacked by Carcinus less fre- quently than were mud snails exposed on the surface (Ashkenas and Atema, 1978). Responding to a chemical signal is adaptive in defense against activities of a noc- turnal tide pool predators such as Carcinus. It is not immediately obvious that a gastropod could avoid being consumed by simply crawling away from its predator or by moving to sheltered sites, since snails are notoriously slow creatures. The key to understanding why this strategy works is knowledge of the predator's feeding behavior and the type of refuge sought by snails. Carcinus uses different techniques to attack and devour bivalve and gastropod prey depending on prey size (Elner, 1978; Kitching ct a!., 1966; Hughes and Elner, 1979; Zipser and Vermeij, 1978). The crab employed a similar size- specific strategy for L. littorca. Small periwinkles were crushed and consumed in 3 min, while cracking and eating medium-sized snails took about 26 min longer. Large snails were never successfully consumed in the laboratory. Thus, large individuals of L. littorca appear to have a size refuge like that reported for /. obsoleta (Ashkenas and Atema, 1978), L. rndis. and L. nigrolincata (Elner and Raffaelli, 1980). Individuals of I., littorca found sheltered sites in approximately 10 min. Thus, the time required by snails to hide corresponded closely to the amount of time required by crabs to consume medium-sized snails, once first injury to snail SNAIL ALARM TO CRAB PREDATIOX 277 tissue had occurred. Although small snails are crushed and eaten too quickly to allow nearby conspecifics time to hide, with increased distance from the predator snails gain time to find shelter. If the juice of crushed conspecifics signals a rc-ul threat to intact snails, crahs must search for a second snail after consuming the first ( Snyder, 1967). All 11 crabs which consumed at least one L. littorca in the laboratory continued search- ing behavior after the first snail had been eaten. These crabs had been without food for 7-10 days when tested. In the field Cardans probably feeds more fre- quently and may never consume more than one snail per feeding period. How- ever, the single green crab observed through an entire episode of feeding in the field resumed searching as soon as the first snail was gone. Additional field observations are needed on this aspect of the snail alarm-crab predation relationship. Of course, crabs are not the only predator of Littorina able to release snail juice. Carnivorous whelks (Thais), herring gulls, ducks, fish, and lobsters have also been reported to eat L. littorca (Pettitt. 1975). The shelter-seeking behavior of the snails may also be an effective defense against visual predators, such as birds. The alarm response would be equally effective against any predator which injures snail tissue, takes more time to consume a snail than snails require to hide, and consumes more than one snail per feeding period. However, the volume and mixing of water along the shore at high tide is so much greater than the volume and mixing in a tide pool at low tide that stimulus molecules probably do not reach concentrations sufficient to affect any snails but those a few millimeters from the crushed snail. Thus, alarm responses may only occur in tide pools or other areas wrhere water is shallow and still, such as a tidal marsh at low tide. In an evolutionary race between shell-crushing predators and their gastropod prey, the evolution of elaborate shell ornamentation, short shell spires, narrow opercula, or thick shell walls may be one line of defense for a snail (Ycrmeij, 1978). However, these morphological adaptations are more often found among tropical than among temperate species. It appears that a complementary first- line strategy for the temperate L. littorca is behavioral defense: alarm behavior which helps a snail avoid detection or attack. Perhaps the most interesting chal- lenge remains : the unraveling of interactions among all selective pressures which together determine whether shell structure, alarm behavior, or a combination of the two evolves for snail defense. The section on snail alarm behavior was first prepared as an undergraduate thesis under the supervision of Beverly Greenspan and James Moulton, Depart- ment of Biology, Bowdoin College. William and Barbara Hadlock made possible the frequent field site visits. I appreciate Tom Seeley's interest and assistance with pilot field studies. The comments of Jelle Atema, J. Stanley Cobb, and Tom Seeley on different drafts of the manuscript improved its final form. This work was supported in part by the Lerner Fund for Marine Research, American Museum of Natural History ; and the Elliott Fund of Bowdoin College. SUMMARY Individuals of Littorina littorca in rocky intertidal pools crawled to pool sites where they were less visible (into rock crevices ; under rocks and macroalgal fronds) when either crushed conspecifics or juice from crushed conspecifics was added to 278 ROBIX P. HADLOCK these pools. A significant proportion of snails hid in 10 min or less; individual snails in one pool tested quadrupled their crawling velocities after snail juice was added. Field observations and laboratory experiments tested the hypothesis that this alarm behavior helps L. littorca avoid being eaten. Green crabs (Carcinus niaenas) were observed consuming individuals of L. littorca in tide pools at night and along the shore at high tide during the day. In the laboratory, crabs required more time to locate and attack periwinkles in rock crevices than periwinkles on rock surfaces. The amount of time required to consume specimens of L. littorea depended on snail size (shell length), reflecting different methods of attack by crabs. Small snails (< 9.0 mm) were crushed, then consumed in approximately 2 min 30 sec. Crabs could not consume large snails (> 18.0 mm), but destroyed medium-sized snails (^9.0, ^ 18.0 mm) by cracking the shell, tearing off bits of tissue, then resuming shell cracking to expose more snail tissue. This required a mean time of 9 min 54 sec once first injury to snail tissue had occurred, which approximately equals the 10-min response time of snails exposed to crushed snail or snail juice in the field. These findings indicate that the alarm response of L. littorca serves in defense against Carcinus inacnas. LITERATURE CITED ATEMA, J., AND G. BURD, 1975. A field study of chemotactic responses of the marine mud snail, Nassarius obsolctus. J. Chcm. Ecol., 1 : 243-251. ATEMA, J., AND D. STENZLER, 1977. Alarm response of the marine mud snail, Nassarius obsolctus: biological characterization and possible evolution. /. Chcm. Ecol., 3 : 173-187. ASHKENAS, L., AND J. ATEMA, 1978. A salt marsh predator-prey relationship : attack behavior of Carcinus inacnas (L.) and defenses of Ilyanassa obsoleta (Say). Biol. Bull., 155: 426. ELNER, R. W., 1978. The mechanics of predation by the shore crab, Carcinus macnas (L.) on the edible mussel, Mytilus cdnlis L. Occologia, 3 : 333-344. ELNER, R. W., AND D. G. RAFFAELLI, 1980. Interactions between two marine snails, Littorina rudis Maton and Littorina nigrolineata Gray, a predator, Carcinus macnas (L.), and a parasite, Microphallus similis Jagerskiold. /. Exp. Mar. Biol. Ecol., 44: 151-160. VON FRISCH, K., 1938. Zur Psychologic des Fisch-Schwarmes. Natunvisscnschaftcn, 26: 601-606. HELLER, J., 1976. The effects of exposure and predation on the shell of two British winkles. /. Zool. London. 179: 201-213. HOWE, N. R., AND Y. M. SHEIKH, 1975. Anthopleurine : a sea anemone alarm pheromone. Science. 189: 386-388. HUGHES, R. N., AND R. W. ELNER, 1979. Tactics of a predator, Carcinus macnas, and mor- phological responses of the prey, Nncclla lapilliis. J. Anim. Ecol., 48: 65-78. KEMPENDORFF, W., 1942. Uber das Fluchtphanomen und die Chemorezeption von Hclisoma (Taphius) nigricans Spix. Arch. Molluskcnk., 74: 1-27. KITCHING, J. A., L. MUNTZ, AND F. J. EsLiNG, 1966. The ecology of Lough Inc. XV. The ecological significance of shell and body forms in Nncclla. J. Anim. Ecol., 35: 113-126. KULZER, E., 1954. Untersuchung iiber die Schreckreaktion bei Erdkrotenquappen (Bujo bulo L. ) . Z. Vergl. Physiol., 36: 443-463. MACKIE, A. M., AND P. T. GRANT, 1974. Interspecies and intraspecies chemoreception by marine invertebrates. Pages 105-141 in P. T. Grant and A. M. Mackie, Eds., Chemoreception in marine organisms. Academic Press, London. PETTITT, C., 1975. A review of the predators of Littorina. especially those of L. sa.ratilis (Olivi) (Gastropoda : Prosobranchia). J. Conchol. 28: 343-357. SNYDER, N. F. R., 1967. An alarm response of aquatic gastropods to intraspecific extract. Cornell University Agricultural Experiment Station Memoir 403, 126 pp. SNYDER, N. F. R., AND H. A. SNYDER, 1970. Alarm response of Diadcma antillannn. Science, 168: 276-278. SNAIL ALARM TO CRAB PREDATION 279 SNYDER, N. F. R., AND H. A. SNYDER, 1971. Defenses of the Florida apple snail, Pomacea paludosa. Behaviour. 40: 175-215. SOKAL, R. R., AND F. J. ROHLK, 1969. Biometry: the principles and practice of statistics in biological research. W. H. Freeman, San Francisco, 77(> pp. STENZLER, D., AND J. ATEMA, 1977. Alarm response of the marine mud snail, Nassarius obsolctus: specificity and behavioral priority. J. (.'hem. lieol.. 3: 159-171. UNDERWOOD, A. J., 1973. Studies on the zonation of intertidal prosobranchs ( Gastropoda : Prosobranchia) in the region of Heybrook Bay, Plymoutli. J. Anim. Ecol.. 42: 353-372. VERMEIJ, G. J., 1974. Marine faunal dominance and molluscan shell form. Evolution. 28: 656-664. VERMEIJ, G. J., 1976. Interoceanic differences in vulnerability of shelled prey to crab preda- tion. Nature. 260: 135-136. VERMEIJ, G. J., 1978. Bioycoi/niphy and adaptation: patterns of marine life. Harvard Uni- versity Press, Cambridge, Mass., 332 pp. VERMEIJ, G. J., AND A. P. COVICH, 1978. Coevolution of freshwater gastropods and their predators. Am. Nat.. 112 : 833-843. ZIPSER, E., AND G. J. VERMEIJ, 1978. Crushing behavior of tropical and temperate crabs. /. Erp. Mar. Biol. Ecol.. 31 : 155-172. Reference: Biol. Bull., 159: 280-294. (October, 1980) THE FORMATION AND EARLY DIFFERENTIATION OF SEA URCHIN GONADS MARGARET S. HOUK AND RALPH T. HINEGARDNER Division of Natural Sciences, University of California, Santa Cruz, California 95064 Despite extensive use of sea urchin gametes in developmental biology, few modern studies deal with the origins and differentiation of the gametes from the primary germ cells (PGCs). All published studies are from the last century or the early part of this one, the major ones being those by Hamann (1887) ; Prouho (1887) ; Cuenot (1891) ; Russo (1894) ; and MacBride (1903). This and other work on gonad development has been summarized by Delavault (1966). Most authors agree that the gonadal or genital primordium, variously called the genital rudiment, genital cord, or sexual bud, is identifiable soon after metamorphosis from pluteus larva to juvenile urchin as an accumulation of large cells located in the dorsal mesentary. This mesentary, which forms from the apposition of the walls of the left and right posterior coeloms of the larva (the somatocoels of the adult), adheres to the aboral body wall in the vicinity of the madreporite and extends into the main body cavity to form the boundary between the left and right somatocoels. The dorsal mesentary also supports the stone canal, axial organ, and parts of the gut. Published studies indicate that the genital primordium later spreads circumferentially around the inside aboral surface and eventually forms a circular cord of cells, the genital rachis, which competely encircles the periproct. The gonads differentiate from five swellings on the rachis, with one gonad forming in each interradius. We initiated this study to determine the accuracy of the earlier observations and to expand upon them through electron microscopy. We have confirmed the location of the gonadal primordium in the dorsal mesentery and its subsequent development into genital rachis and gonads. We differ, however, from most pre- vious investigators in our interpretation of some features of the process. Our ultrastructural examination of the gonadal primordium sheds additional light not only on the origin of the gametes, but on the origin and differentation of the gonadal accessory cells (sometimes called the nutritive phagocytes). MATERIALS AND METHODS Selection of inaterial The sea urchins we used were out-crossed laboratory-raised individuals of the species Lytechinus pictus, grown according to the methods of Hinegardner( 1969). Juvenile urchins selected for tissue sectioning ranged from 0.3 to 6.0 mm in outer diameter and in age from 3 days post metamorphosis to 8 months post meta- morphosis. We have been able to identify the genital cells in all stages up to the time when recognizable eggs and sperm are found. Preparation for light microscopy Juvenile urchins were fixed in ethanol-acetic acid, 3:1, and dehydrated in an ethanol series. Urchins under 3 mm in diameter were embedded for serial 280 SEA URCHIN GOXAL) D! 1- FF.REXTIATK ).\ 281 sectioning first in celloiditi and then in paraffin according to the method of Akesson (1961). Larger juveniles were embedded directly in paraffin. Five p.m serial sections were used. After removal of the paraffin with xylene, sections were passed from 100 to 90';; ethanol and stained for 20 sec in Unna-Pappenheim methyl- green pyronin stain (Gurr, l'X>5). They were then washed in water for 1 min. blotted to remove excess water, and passed directly to 100 7; ethanol for 10 sec of agitation. Rapid passage through water and ethanol was necessary to prevent loss of stain. Cell counts at various growth stages were made from camera litcida drawings using semi-transparent paper so that tracings could be overlaid for comparison. For each individual in which germ cells were counted, all sections containing any germ material were traced and compared with adjacent sections to avoid counting the same cell more than once. Preparation for transmission electron microscopy Urchins were prepared for fixation by several methods depending on their size. Gonads from mature individuals ( > 10 mm in diameter) were excised from the body cavity prior to fixation. Juvenile urchins were bisected through their largest diameter. The aboral half, containing the germ cells, was then inverted to form a "bowl" and Hushed several times with fixative prior to immersion in fresh fixative. The desired tissues were then dissected from the urchin. Juveniles under 1.5 mm were fixed whole. Most material was fixed in 1.59' glutaraldehyde in sea water for 1 hr. All manipulations were carried out at pH 7.0-7.6 and at 4°C. A few samples were fixed according to the method of Bal ct al. (1968). Tissues were next washed in several changes of sea water and post-fixed for 2-2.5 hr in 1% osmium tetroxide in sea water. Dehydration was carried out in an acetone series. Tissues were embedded in Spurr's resin (Spurr, 1969). Material from early juvenile stages had to be decalcified because the minute germinal primordia could not be removed by dissection. Decalcification was carried out on material fixed overnight in 1.5% glutaraldehyde followed by osmium tetroxide as described above. Samples were either decalcified in ascorbic acid according to the method of Dietrich and Fontaine (1975) or in 5% EGTA in 0.1 M sodium phosphate buffer at pH 7 for 1 hr. Blocks of tissue were first sectioned (0.5 ju.ni) for purposes of orientation. These thick sections were stained for light microscopy with \% toluidine blue in 1% sodium borate. Thin sections were cut with a Porter-Blum MT-2 ultra- microtome using a glass knife. Sections were stained for 30 min with Millipore filtered 2% uranyl acetate and 5 min with Reynolds' lead citrate (Reynolds, 1963). The electron microscope was a JEOL GEM 100B. Preparation for scanning electron microscopy Juvenile urchins ranging from 3 to 7 mm in diameter were prepared for scanning electron microscopy by bisecting the specimen equatorially slightly on the aboral side of the greatest diameter. The aboral part was flooded with 1.5% glutaralde- hyde in sea water and any remaining pieces of gut carefully dissected out under a dissecting microscope. Specimens were dehydrated in an ethanol series, substituted with amyl acetate, then critical-point dried in carbon dioxide (Anderson, 1951). coated with carbon and gold and examined with a JEOL JSM-2 scanning electron microscope. 282 M. S. HOUR AND R. T. HINEGARDNER •^ - M lao V >»s SMf- 555s! •: 5 scO^ ':r-:--,--,;,;,;<^ igX^7^-. \\ •.•;-•:••:.•:••:;.•, -*»^ V'«-~;-r -.'- l^ : ^:> JJ-Vi .• £^^! . " : :'^v^r »» ' 7..^.\. •••••- ^ • »>. .- ^~ i^^.» —— - • • ' ki v«i - , . . ^~ -:- - 00 -»^ __»W . • w-%^— r»-«._* • ;;, ^ FIGURES 1-5. Paraffin sections through the germinal tissues of juvenile sea urchins. Figure 1 — 3-day-old juvenile 0.3 mm in diameter. Figure 2 — Month-old juvenile 1.0 mm in diameter. Figure 3 — Three-month-old juvenile, 2.0 mm in diameter, sectioned at a 45° angle to the oral-aboral axis. Figure A — Three-month-old juvenile, 2.0 mm in diameter, sectioned circumferentially. The genital primordium gr is suspended between the aboral body wall bw and stone canal sc and closely associated with the axial organ ao near where the gut gt SEA URCHIN r.OXAI) DIFFERENTIATION 283 TAHLK I Number of primordial germ cells present in the juvenile sen urchin nt different pust-metumorphosis ages and test diameters Days past Outer diameter No. urchins Average no. germ metamorphosis of test (mm ) counted cells ± SD 1-3 0.3 5 20.8 ± 5.4 29 0.68 1 21 29 0.85 3 36.0 ± 5.3 29 1.0 3 29.33 ± 3.8 29 1.2 3 (>0.3 db 12.0 93 1.5 1 60 93 2.0 2 101.0 ± 26.9 Unknown 2.7 2 1000 approx. RESULTS Differentiation prior to gonad formation In the newly metamorphosed sea urchin, the primordial germ cells (PGCs) of the genital primordium are identifiable in serially sectioned material as a group of pyrinophilic cells with large (5 ju.ni) nuclei containing one or two nucleoli. The PGCs remain suspended in the mesentery between the stone canal and the inner surface of the aboral wall throughout early juvenile development (Fig. 1). By the time the urchin reaches a diameter of 1 mm (approximately 1 month after metamorphosis), the germ cells have clustered into a distinct group of cells within the germinal primordium (Fig. 2), and are surrounded by a layer of squamous epithelium originating from the mesentery. Xo coelomic cavity has been identified in association with this group of cells. Two types of cells, as determined by nuclear morphology, can be found in the genital primordium. The first are the primary germ cells found in the mesentery at the time of metamorphosis. The other type, which we call the presumptive accessory cells, has a nucleus of about 2.5 //.m in diameter that stains more heavily than the first type with methyl-green. These cells were not seen in the dorsal mesentery of newly metamorphosed urchins, and their origin is not known. We found the size of juvenile urchins, rather than their post-metamorphic age, to be the more reliable indicator of the degree of differentiation of the genital tissues (Table I), and therefore include test diameter as an indication of the developmental meets the anus, am — ampulla of the left anterior coelom. /w/ — Junction of projecting part of body wall with the main body wall. Figure 5 shows the aboral region of a juvenile 2.5 mm in diameter. Two of the gonadal primordia yo are connected by part uf the genital rachis ?•<;. Bars = 10 /an. Figures 2-5 are at the same magnification. FIGURES 6-8. Scanning electron micrographs of the inner perianal region of juvenile sea urchins. Most of the gut has been removed. Figure 6 shows an individual 1.5 mm in diameter. The torn edges of the intestine surround the anal area at the left corner, while the axial organ and attached dorsal mesentery containing the genital primordium appear at upper right (arrow). Figure 7 depicts an urchin 3.5 mm in diameter with unbranched gonads, four of which are arranged in a circle, one each right and left, and two at the bottom of the figure. The axial organ appears just above the gut in the upper left corner (arrow). Figure 8 shows the gonads of an urchin 5.0 mm in diameter. Four gonads and axial organ (arrow) are shown, oriented as in Figure 7. All three figures are at the same approximate magnifica- tion (X 55-60). Field width for 6 is 1.04 mm, for Figure 7 is 1.65 mm, and for Figure 8, 2.02 mm. 284 M. S. HOUR AND R. T. HINEGARDNER stage of our material. In urchins as small as 1.5 mm in diameter, the genital primordium has already begun to broaden and elongate within the dorsal mesentery due to an increase in the number of PGCs (Table I). At this stage genital cells occupy most of the dorsal region of that mesentery. Subsequent gonad development is easiest to understand by visualizing the primordium as something like a letter I lying in the mesentery. As the urchin grows the PGCs migrate to the top of the 1, where the dorsal mesentery joins the dorsal body wall. The PGCs then begin to migrate laterally, converting the I to a T-shaped structure. The arms of the T gradually elongate and its stem shortens. Evenutally the stem disappears and the elongating arms form a ring around the periproct. This ring is called the genital rachis (Hamann, 1887). Its diameter is about 1/4 the diameter of the urchin. Figures 3 and 4 illustrate sections from urchins 2.0 mm in diameter, which have just begun to form a rachis. The section in Figure 3 is approximately perpen- dicular to the oral-aboral axis and shows the location of the band of genital tissue relative to the left anterior coelom (axocoel), the axial organ, and the stone canal. The rachis lies along part of the body wall which projects into the coelomic space due to the presence of the axocoel and its associated structures. Figure 4 is oriented approximately 45° to Figure 3 and shows the arms of the T-shaped genital primordium. There is a continuum of cells throughout the primordium, with no separation between those beginning to form the rachis and those still within the mesentery. The genital primordium in Figure 4 appears to be connected to the main body wall by only a narrow piece of tissue, but this appearance is clue to the angle of the section. After the rachis has formed, five thickenings appear in it, one in each interradius. These eventually contain most of the PGCs and ultimately give rise to the five gonads. Two such forming gonads and a large portion of the genital rachis, from an urchin 2.5 mm in diameter, are shown in Figure 5. Both the rachis and the developing gonads are surrounded by a layer of epithelium which seems to be derived from the epithelium for the mesentery in which the genital primordium was first found, and thus to be coelomic in origin. Both cell types found in the genital primordium are present in the rachis and in the developing gonads. In larger urchins, with differentiated gonads, the solid cord of PGCs in the rachis has disappeared. Differentiation of the gonads No gonads are visible in urchins 1.5 mm in diameter (Fig. 6). Unbranched gonads are present in a 3.5-mm-diameter urchin (Fig. 7). A raised circular area on the aboral wall connects the bases of the gonads, indicating the presence of the genital rachis. Figure 8 shows this area of an animal 5.0 mm in diameter. The gonads have branched several times. That a rachis still exists, even in an individual in which all the germ cells have undoubtedly migrated into the gonads, can be seen in the upper right portion of Figure 8, where the rachis has torn free from the underlying epithelium. We can therefore corroborate the persistence of the now empty rachis as the aboral ring canal, as reported by Cuenot (1891) and Russo (1894). Further branching continues to occur as the gonads grow, until the aboral part of the coelom contains a mass of branching tubules. Each gonad is surrounded by an epithelial layer continuous with the lining of the perivisceral coelom. Besides separating the gonad from the main body cavity, this cell layer SEA URCHIN COXA!) DIFFERENTIATION 285 forms mesenteries which attach it to the ahoral body wall. No gonoduct is present at this time. Sexual differentiation of the gonad is first apparent with the presence of recognizable oocytes and spermatids. These appear at ahout the same time or shortly after the gonads begin to branch. By the tune male urchins reach a diameter of 5 mm their gonads contain mature spermatozoa in the central cavity, but they usually cannot be induced to spawn. Females 5-6 mm in diameter have previtellogenic oocytes up to 25 /xm in diameter (mature eggs are about 110 /j,m m diameter). The structure of the female gonad at this stage does not differ substantially from that of published descriptions Of the mature gonad of other species (Holland and diese, 1965; Chatlynne, 1969; Davis, 1971; and Pearse, 1969a, b). A gonoduct leads from the central cavity of the gonad to the outer wall of the test. It consists of a single layer of cuboidal cells of unknown origin. Ultrastnicturc of the genital tissues Three stages in the development of the juvenile urchins were selected for ultrastructural studies. These stages were : 1 ) the genital primordium of newly metamorphosed urchins (0.5 mm in diameter), 2) the older genital primordium, in which the PGCs are rapidly proliferating (1.5 mm), and 3) a sexually undif- ferentiated gonad (test diameter 3.5 mm). Gonads from sexually differentiated urchins were also studied for comparative purposes and to check our results against published studies where different species and fixation procedures were used. 1. Genital primordium of )mt'l\ metamorphosed urchin. A tangential section through the genital primordium of a sea urchin with test 0.5 mm in diameter (3 weeks after metamorphosis) is shown in Figure 9. The genital primordium is roughlyr 33 /mi across and is bounded on the outside by an epithelium of ciliated squamous cells. A thick connective-tissue layer lies beneath the epithelium and contains scattered phagocytic cells. The connective-tissue layer is not strictly separated from the underlying cells, and fibrous material similar to that in the connective tissue is often found between these cells. The genital primordium is separated from the body wall by the epithelium of the mesentery and by the peritoneum of the somatocoel. A pronounced basement lamina is found under- lying the epithelial cells. The mesentery containing the genital primordium is attached to the body wall via the epithelial layers at the right of the figure. Two cell types are found within the genital primordium. The first are elongate, smooth surfaced, rounded in cross section, and have a round to oval nucleus containing a relatively electron-transparent nuclear matrix. The nucleus is 5-6 /mi in diameter. These cells correspond to the primordial germ cells seen with the light microscope. A fibrillar material, presumably chromatin, is fairly evenly distributed within the nucleus. Dense granules about 60 nm in diameter are interspersed among the chromatin material (Fig. 10). The cytoplasm contains the usual cellular organelles, most of which are concentrated in the tapered ends of the cells. Membrane-bound vesicles as large as 0.9 /mi in diameter and containing an electron-dense amorphous material are also present (Fig. 9). These are not seen in the larval material we have studied nor in later stages. Associated with the mitochondria, and sometimes apparently touching their outer membranes, are 100 nm granules containing a dense fibrillo-granular material. These are probably identical to the granules found by Longo and Anderson (1969) in sea urchin spermatogonia. The granules are sometimes clustered (Fig. 11). 286 M. S. HOUK AND R. T. HINEGARDNER FIGURE 9. Section through the genital primordium of an urchin 0.5 mm in diameter showing the mesentary epithelium me, with its underlying basement lamella (arrow), primordial germ cells pgc, presumptive accessory cells [>ac. and a phagocytic cell />/i. The epithelium of the mesentery joins that of the body wall /w at the lower right hand side of the figure. Bar = 1 fj.m. SEA URCHIX (iOXAI) DIFFERENTIATION 2S7 They are common in these cells as well as those in later stages. Since they seem to he specific to the gamete-forming cells, we will refer to them as goniosomes. Amorphous electron-dense material similar to that found in germ cells of other organisms (Eddy, 1975) is also present in the cytoplasm of PGCs. The cells that we are calling primary germ cells share a number of characteris- tics with gonial cells, of which they arc certainly the precursors. These include: nuclear morphology ; the general appearance and contents of the cytoplasm, including goniosomes ; and staining characteristics using methyl-green pyronin and toluidine blue. The second cell type is almost certainly the precursor of the accessory cells of the gonad, which it resembles in both shape and nuclear morphology. These accessory-precursor cells are smaller than the PGCs and are generally angular in cross section, with many elongate cellular processes (Figs. 9, 10 j. The nucleus is also elongate and irregular in outline. Its matrix is more electron-dense than is the nucleus of the PGC, and areas of condensed chromatin are obvious. Small dark 60 nm granules are present here as they were in the PGCs, and cytoplasmic inclusions are similar. Structures which look like the 100 nm goniosomes common in the PGC have not been found in these cells. The accessory cell primordium also contains a large electron-transparent vacuole ( Fig. 10) similar to that found in the accessory cell of the mature gonad (Holland and Giese, 1965). Flagellar cross-sections among the cells of the genital primordium indicate that PGCs and/or accessory-cell precursors are flagellated. 2. Genital primordium with rapidly proliferating PGCs. The genital pri- mordium cells of a 1.5 mm urchin are similar to those of the younger animal (Fig. 12). The PGCs have the same general appearance except for the absence of the large membrane-bound vesicles filled with electron-dense material. Mitotic cells can now be found. Otherwise, the PGCs are unchanged. The accessory-cell primordia have now become obvious accessory-cell pre- cursors. The angular nucleus is surrounded by a very thin layer of cytoplasm (Fig. 12). Extensive processes containing large numbers of membrane-bound lipid-like vesicles, similar to those found in the accessory cells of mature gonads, can be seen extending between the PGCs. Glyeogen granules, which are abundant in the accessory cells of the mature gonads, are often present in the cellular exten- sions. The vacuole has elongated. Nuclear morphology remains relatively unchanged and is similar to that of accessory cells of the mature gonad. Flagellar cross-sections are present among these cells. 3. Sexually iindiffercntiatcd gonads. A cross section of an unbranched gonad from a 3.5-mm-diameter urchin is shown in Figure 13. A many-layered gonad wall surrounds the PGCs. From outside to inside these layers consist of flagellated coelomic epithelium, a connective-tissue layer, a fluid-filled space, a musculo- epithelial layer (also flagellated), and the germinal epithelium. This order is identical to that of the layers in the wall of the mature gonad ( Kawaguti, 1965; Davis, 1971). The outer epithelium is continuous with the lining of the peri- visceral coelom. The PGCs dominate the outer germinal layer, and are arranged two or three cells thick. They are closely packed, adhering by means of junctional complexes FIGURE 10. A primordial germ cell /v/r in an urchin 0.5 mm in diameter is partially surrounded by an accessory cell pac. Both cells contain vesicles with electron dense material (arrows). The pac also has a large vacuole TO. Bar-- 1 pm. FIGURE 11. Goniosome cluster from a PGC at the same stage shown in Figures 9 and 10. Bar = 0.1 urn. 288 M. S. HOUR AND R. T. HINEGARDNER FIGURE 12. section tlirough the genital primordium of an urchin 1.5 mm in diameter, showing primordial germ ce! pf>8; Chatlynne, D72). Testes in these young urchins are organized similarly to those of mature males. All individuals which could be identified as males already had mature sperm present in their testes luminu. Presumably the spermatogenic cycle is so rapid (less than 12 days in gonads of some species of urchins; Holland and Giese, 1965) that testes containing spermatids and no mature sperm would be hard to find. Spermatogonia are clustered against the wall of the testes and resemble oogonia and PGCs. Spermatogenesis is well described in Longo and Anderson (1969). DISCUSSION While most authors agree that the gonads of echinoderms differentiate from a group of cells located in the dorsal mesentery of the newly metamorphosed juvenile, the origin of the cells is disputed. Cuenot (1891 ) and MacBride (1903) believed that the genital primordia of asteroids, ophiuroids, and echinoids have common origins with the axial organs. In fact, MacBride refers to the axial organ as the genital stolon. Prouho (1887) and Russo (1894) described the genital primordium of sea urchins as arising from the wall of the left anterior coelom independently of the axial organ. MacBride (1903) also proposed an epithelial origin for the germinal cells of sea urchins, since he believed that the genital stolon, from which both the genital primordium and the axial organ supposedly differentiated, arose from the wall of the left posterior coelom of the newly metamorphosed urchin. More recent work by Delavault (1966) on the sea star Asterina gibbosa indicates that at least in sea stars, the location of the germ cells in the dorsal mesentery may be secondary and that the genital pri- mordium may descend from mesenchyme cells which migrate to the mesentery during late larval development. Our study of newly metamorphosed juveniles of Lytcchinns pictus shows that the germ cells are already present in the dorsal mesentery at the time of metamorphosis as a cluster of cells distinct from the axial organ and not connected to the wall of the coelom. Their ultrastructure resembles that of the mesenchymal cells. If these cells do originate from the coelomic epithelium, it must happen during larval development, contrary to MacBricle's description for Echinus cscitlcntus (1903), or Cuenot's for Paracen- trotns Hindus (1891). While their location and ultrastructural appearance sug- FIGURE 15. Goniosomes closely associated with a mitochondrion in a PGC of the sexually undifferentiated urchin. Bar = 0.1 /mi. FIGURE 16. Presumptive accessory cell in unbranched gonad with lipid vesicles h, lyso- somal-like vesicles /v, large vacuole va, and glycogen granules (arrows). Bar = 0.1 /mi. FIGURE 17. Large oocyte in branched gonad of urchin 3.3 mm in diameter. The single nucleolus nu is differentiated into two regions of different opacity to electrons. Goniosomes are present in the cytoplasm (arrows). Bar = 1 /mi. 292 M. S. HOUK AND R. T. HINEGARDNER gest a mesenchymal origin for the germ cells, careful study of larval development will be necessary to decide this question. In other classes of echinoderms, particularly the asteroids, the genital rachis is described as developing from a vesicle which buds oft" the peritoneum of the left somatocoel and surrounds the germ cells. The space within the vesicle is derived from the coelom. In asteroids this space persists in the wall of the gonad and remains connected with the coelom (Delage and Herouard, 1903). Such a vesicle does not appear in Lytcchinus pictits. Rather, the germ cells begin to divide and to travel within the dorsal mesentary to the aboral body wall, extending out into a ring continuous with the mass of cells in the dorsal mesentary. This continuity indicates that the germ cells in the rachis must be between the peritoneum of the coelom and the body wall. The outer covering of both the primordium in the mesentery and the gonad wall is the peritoneal epithelium. The inner musculo- epithelial layer of the gonad wall is not present in the genital primordium stage. The origin of this secondary epithelial layer is not explained by our investigation. The layer appears to be present in the rachis, but ultrastructural examination would be necessary to determine this with certainty. The space between the two epithelia is probably a schizocoel, as suggested by Hamann (1887). The accessory cells, also called nutritive phagocytes, have been studied in the mature gonad by several authors (Holland and Giese, 1965; Verhey and Mover. 1967; Masuda and Dan, 1977). These cells are present in both sexes of the adult and change in size and internal constitution during the gametogenic cycle, giving rise to the hypothesis that they are involved in recycling nutrients from phagocytized gametes (Holland and Giese, 1965; Pearse, 1969a, b). Various origins for these cells have been suggested. Some authors have claimed that they originate from the same cells in the gonads as the gonia and therefore fit the classical definition of nurse cells (Miller and Smith, 1931 ). In our study we have seen that accessory cell primordia are present in the genital primordium soon after metamorphosis and migrate with the PGCs from the genital primordium through the rachis into the gonad. Differentiation into recognizable accessory cells consists primarily of the accumulation of cytoplasmic inclusions such as glycogen granules, lipid-like vesicles, and lysozomal-like vesicles. It begins while the cells are still in the genital primordium. This gradual accumulation of the cytoplasmic structures characteristic of accessory cells leaves little doubt that these cells differentiate from the small-nucleus cells of the genital primordium. In young juveniles 1 mm in diameter, the cytoplasm of both the PGCs and the presumptive accessory cells contains membrane-bound vesicles filled with an electron-dense material which resembles yolk. These vesicles are seldom seen in either cell type at later stages. Possibly they are diluted as the cells divide. The presence of yolk in the PGCs of sea urchins is a reasonable suggestion, since PGCs of anuran amphibians are known to contain yolk platelets (Blackler, 1958). If these vesicles prove to be generally present in PGCs of larvae as well, they may be useful as cytoplasmic markers for tracing the germ cells back through larval development. While the PGCs and the presumptive accessory cells are distinct from each other in all the stages covered by this study, the presence of yolk-like structures in both cell types suggests they may have a common embryo- logical origin. While the PGCs share some features with the accessory cells, they are much closer in appearance to gonia. They have similar nuclear morphology, general cell shape, and cytoplasmic inclusions, especially the 100-nm goniosomes which were SEA URCHIN* (iOXAl) DIFFERENTIATION 293 found at all stages examined and only in germ-line cells. Similar granules have been found in the germ lines of insects (Mahowald, 1(>71 ) and amphibia (Kalt, 1973; Kerr and Dixon, 1974). We would like to thank Janice Xowell of the Electron Microscope Facility of the University of California, Santa Cruz, for her help with the technical details of the ultrastructural work presented in this paper. \Ye also express our appreciation to Dr. John S. Pearse for his careful reading and helpful criticism of the manu- script. This work was supported by Public Health Service postdoctoral fellowship 1 FO2 HD52424-01 and 5 FO2 HD52424-02 to M. Honk and by National Science Foundation Grant PCM 76-14726 to R. Hinegardner. SUMMARY 1. The genital primordium gonads were examined at several stages during their development in the juvenile sea urchin. 2. At metamorphosis the primordial germ cells are a distinct group of cells in the dorsal mesentery, which also supports the stone canal and axial organs. They spread circumferentially along the aboral body wall to form a ring, the genital rachis, completely encircling the anal region. The gonads form as swellings in each interradius of this ring. 3. While the wall of the genitial primordium prior to rachis formation consists of a single layer of epithelial cells, that of the young gonad has two such layers separated by a fluid-filled space. 4. The primordial germs cells of each stage examined are similar in nuclear morphology, cytoplasmic inclusions, and cell shape and size. All stages have germ- cell-specific fibrillo-granular structures, which we name goniosomes. 5. A second cell type, the forerunner of the accessory cell of the mature gonad, is present in the genital tissue of all stages studied. Cellular inclusions character- istic of the accessory cell, including glycogen granules, lipid-like vesicles, and lysozymal-like vesicles, begin to accumulate in these cells even before gonad formation. LITERATURE CITED AKESSON, B., 1961. A rapid method for orienting small and brittle objects for sectioning in definite planes. Ark. Zoo!., 13: 479-482. ANDERSON, T. F., 1951. Techniques for the preparation of three dimensional structure in preparing specimens for the electron microscope. Trans. A'. Y. Acad. Sci.. 13 : 130-134. BAL, A. K., F. JUBINVILLE, AND G. H. CorsiNEAr, 1969. Nuclear activity during oogenesis in sea urchins. II. Fine structural changes and patterns of RNA synthesis during meiotic prophases of Arbacia punctidata oocytes. Z. Zcllfnrsch. Mikrosk. Auat.. 100: 180-188. BAL, A. K., F. JTBIXVILLE. G. H. Corsi\K.\r AND S. INCITE, 1968. Origin and fate of the annulate lamellae in Arhaciu punciidata eggs. /. Ultrastruct. Res.. 25: 15-28. BLACKLER, A. W., 1958. Contribution to the study of germ-cells in A num. J. Exp. Morphol.. 6: 491-503. CHATLYNNE, L., 1969. A histological study of oogenesis in the sea urchin, Strongylocentrotus piirpuratus. Biol. Bull.. 136: 167-184. CHATLYNNE, L., 1972. An ultrastructural study of oogenesis in the sea urchin Str/nic/ylo- ccntrotus furpuratus. Ph.D. Dissertation, Oregon State University. 141 pp. CUENOT, L., 1891. Etudes morphologiques sur les Echinodermes. Arch. Biol.. 11: 313-380. 294 M. S. HOUR AND R. T. HINEGARDNER DAVIS, H. S., 1971. The gonad wall of Echinodermata : a comparative study based on electron microscopy. Masters Thesis, University of California, San Diego. 99 pp. DELAVAULT, R., 1966. Determinism of sex. Pp. 615-638 in R. A. Boolootian, Ed., Physiology of Echinodermata. Interscience Publishers, New York. DELAGE, Y., AND HEROUARD, E., 1903, Traitc dc Zooloyic concrete les Echinodcnns. Schleicher Freres, Paris, 495 pp. DIETRICH, H. F., AND A. R. FONTAINE, 1975. A decalcification method for ultrastructure of echinoderm tissue. Stain Tcc/inol., 50: 351-354. EDDY, E. M., 1975. Germ plasm and the differentiation of the germ cell line. Pp. 229-280 in G. H. Bourne and J. F. Danielli, Eds., International Rcvin^' of Cytology. Vol. 43. Academic Press, New York. GURR, E., 1965. P. 211 in Rational Uses of Dyes in Biology. Williams and Wilkins Co., Baltimore. HAMANN, O., 1887. Beitrage zure Histologie das Echinodermin. Jena Z. Naturwiss., 21 : 87-262. HINEGARDNER, R. T., 1969. Growth and development of the laboratory cultured sea urchin. Biol. Bull., 137 : 465-475. HOLLAND, N. D., AND A. C. GIESE, 1965. An autoradiographic investigation of the gonads of the purple sea urchin (Strongyloccntrotus purpuratus). Biol. Bull., 128: 241-258. KALT, M. R., 1973. Ultrastructural observations on the germ line of Xenopus lacvis. Z. Zcllforsch. Mikrosk. Anat., 138: 41-62. KAWAGUTI, S., 1965. Electron microscopy on the ovarian wall of the echinoid with special reference to its muscles and nerve plexis. Biol. J. Okayama Univ., 11 : 66-74. KERR, J. B., AND K. E. DIXON, 1974. An ultrastructural study of germ plasm in spermato- genesis of Xenopus lacris. J. Embryol. Exp. Morphol., 32 : 573-592. LONGO, F. L., AND E. ANDERSON, 1969. Sperm differentiation in the sea urchins Arbacia punctulata and Strongyloccntrotus purpuratus. J. Ultrastruc. Res., 27 : 486-509. MACBRIDE, E. W., 1903. The development of Echinus esculcntus, together with some points in the development of E. iniliaris and E. acutus. Philos. Trans. R. Soc. Land. B. Biol., 195 : 285-327. MAHOWALD, A. P., 1971. Polar granules of Drosophila. II. Ultrastructural changes during early embryogenesis. /. Exp. ZooL, 167 : 237-262. MASUDA, R., AND J. C. DAN, 1977. Studies on the annual reproductive cycle of the sea urchin and the acid phosphatase activity of relict ova. Biol. Bull., 153: 577-590. MILLER, R. A., AND H. B. SMITH, 1931. Observations on the formation of the egg of Echino- dermata lucunter. Carnegie Inst. Wash. Pub'. No. 413 : 47-52. MILLONIG, G., M. Bosco, AND L. GiAMBERTONE, 1968. Fine structure analysis of oogenesis in sea urchins. /. Exp. ZooL, 169 : 293-314. PEARSE, J. S., 1969a. Reproductive periodicities of Indo-Pacific invertebrates in the Gulf of Suez. I. The echinoids Prionocidaris baculosa (Lamarck) and Lorenia elongata (Gray). Bull. Mar. Sci., 19 : 323-350. PEARSE, J. S., 1969b. Reproductive periodicities of Indo-Pacific invertebrates in the Gulf of Suez. II. The echinoid Echinoinetra niathaei (de Blainville). Bull. Mar. Sci., 19: 580-613. PROUHO, H., 1887. Recherches sur le Dorocidaris papillata et quelques autre echinoides de la Mediterranee. Arch. Zool. Exp. Gen., 15 : 213-380. REYNOLDS, E. S., 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. /. Cell. Biol., 17 : 208-213. Russo, A., 1894. Sul sistema genitale e madreporico degli Echinidi regolari. Boll. Soc. Nat. Napoli Ser. 1, 8 : 90-107. SPURR, A. R., 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. /. Ultrastruct. Res., 26: 31^3. VERHEY, C. A., AND F. H. MOVER, 1967. Fine structural changes during sea urchin oogenesis. J. Exp. ZooL, 164: 195-207. Reference: Biol. Bull.. 159: 295-310. (October, 1980) THE ULTRASTRUCTURE OF COELOMOCYTES OF THE SEA STAR DHRMASTHRIAS I M URIC ATA EDNA S. KANESHIRO AXU RICHARD I). KARP Department of Biological Sciences, University of Cincinnati. Cincinnati, Ohio 15221 It has been previously reported that the sea star. Dcnnastcrias iinbricata, possesses a true adaptative cell-mediated immune response, as reflected by the specific chronic rejection of integumentary allografts (Karp and Hildemann, 1976). The reaction appears to be mediated by infiltration by phagocytic cells. Since the predominant phagocyte in the sea star is the coelomic amoebocyte (Boolootian and Giese, 1958; Endean, 1966), this cell may mediate the graft reaction. This hypothesis is supported by the recent report of Bertheussen (1979) that phagocytic amoebocytes of the sea urchin Strongylocentrotus drocbachicnsis demonstrate in vitro cytotoxic reactions against both allogeneic and xenogeneic echinoid cells. Although several excellent studies characterize the coelomocytes of holothurians and echinoids (Hetzel, 1963; Johnson, 1969a, b, c; Chien ct al., 1970; Fontaine and Lambert, 1973, 1977; Bertheussen and Seljelid, 1978; Bertheussen, 1979), only sparse information deals with asteroids. To further understand the immunologic capacities of the sea star, Dcnnastcrias, and to compare them with those of other animals, we have characterized the ultrastructure of the coeolomic amoebocyte and have reaffirmed its phagocytic role. Ultrastructural observations on Tiede- mann's bodies were also made to determine whether or not the cells of this organ participate in phagocytic activities. We experimented on animals stressed by inter- mittent extraction of coelomic fluid to follow the recovery of coelomocytes, and to determine whether or not this recovery is due to cell division in the circulating coelomic fluid. MATERIALS AND METHODS Animals Specimens of Dermasterias iinbricata collected off the California coast were purchased from Pacific Biomarine (Venice, CA) and maintained at 15 °C in aquaria containing artificial sea water adjusted to a specific gravity of 1.03-1.04 (Instant Ocean, Aquarium Systems, Wickliffe, OH). The animals were fed lake smelt three times per week. Animals exhibited active locomotion and extension of tube feet, indicating good health. Isolation of coelomocytes Coelomic fluid (containing coelomocytes) was obtained by inserting a tuberculin syringe with a #27 gauge needle into the interradial area through the aboral surface of the central disc. Before fluid withdrawal, individual animals were weighed on an Ohaus Harvard trip balance after they were drained of excess sea water for 30 sec. 295 296 E. S. KANESHIRO AND R. D. KARP Coelomic fluid was rapidly added to equal volumes of a fixative solution containing 6% glutaraldehyde and 10 mM ethylene bis (oxyethylene-nitrilo) tetra- acetic acid (EGTA) in CV"-free artificial sea water (M.B.L. formula; Cavanaugh, 1964) adjusted to pH 7.4. The addition of the Ca-+ chelator and use of Ca-+-free sea water minimized aggregation of coelomocytes ( Edds. 1977). Coelomocytes were viewed at X40 magnification on a Zeiss light microscope and their numbers determined with a hemocvtometer. Light microscopy Coelomic fluid was placed on glass slides and coelomocytes were observed by phase contrast microscopy as smears or hanging drop preparations. Living coelo- mocytes were stained for lysosomes with 0.1% neutral red. Cells were allowed to adhere to glass slides coated with bovine serum albumin, then dried and treated with Giemsa stain to aid observations on nuclear configurations. Electron microscopy Coelomocytes were prepared for scanning electron microscopy (SEM) and transmission electron microscopy (TEM) by fixation in Z% glutaraldehyde in M.B.L. -formula sea water, post-fixation in \r/c OsO4, and dehydration in ethanol. Coelomocytes for SEM were transferred into acetone and placed on glass cover slips which were coated with poly-L-lysine (Sanders ct a!., 1975). The prepara- tions were subjected to critical point drying (Denton DCP-1, Denton Vacuum Inc., Cherry Hill, NJ), coated with gold ( PS-2 sputter coater. International Scientific Instruments, Glen Ellyn, IL), and viewed on a Stereoscan 600 SEM (Cambridge Scientific Instruments, Ossining, NY). Cells used for TEM were transferred into propylene oxide, infiltrated, and embedded in plastic (Spurr, 1969). Sections were cut with a diamond knife on a Sorvall MT-1 ultratome and stained with uranyl acetate and lead citrate. Thin sections were viewed on an AEI-6B electron microscope operated at 60 kV. Tiedemann's bodies were removed from animals which had been injected with 1 ml of the fixative solution. Excised organs were then prepared for TEM as described above. Clearing of bacteria from coelomic fluid Specimens of a marine, gram-negative short rod with a single polar flagellum, Acromonas sp., UF-B (isolated and characterized by G. G. Holz, Jr., S.U.N.Y., Upstate Medical Center), were grown in an enriched sea water medium (Soldo and Merlin, 1972) at 37°C for 1 day. Bacteria were concentrated by centrifugation at 2500 < g and washed with M.B.L.-formula sea water. Approximately 3.5 X 107 bacteria in 1 ml of sea water were injected into the coelomic cavity of each sea star tested. Coelomic fluid was withdrawn at various time intervals, and a 0.1 ml portion (undiluted or diluted 1 : 100) was spread on agar plates prepared with the bacterial growth medium. Bacteria numbers were obtained by colony counts following incubation of agar plates at 37°C for 2 days. Another 2 ml portion was prepared for TEM in the glutaraldehyde fixative described above. SEA STAR COELOMOCYTES 2()7 RESULTS Surface of coelomocytes The phagocytic amoebocyte is tin- doniinant cell type in the coelomic fluid of Dermasterias iinbricata. Smaller mononuclear. nonflagellated cells have also been observed. The dynamic nature of the living coelomocytes was apparent by phase microscopy of the cells. Petaloid (bladder), lamellipodial, and filopodial forms were observed. The filopodial form was predominant when cells were allowed to adhere to glass slides. Coelomocytes were in petaloid or lamellipodial configura- tions when cells were observed immediately after coelomic fluid was sampled, or when cells were fixed immediately upon withdrawal of coelomic fluid from the sea stars. Surfaces of living coelomocytes rapidly and constantly changed shapes, forming pseudopodial extensions that quickly collapsed as still others formed. Cells prepared for SEM and TEM were fixed within seconds of withdrawal of coelomic fluid. Observations with SEM (Fig. 1) confirmed light microscopic studies which indicated that the lamellipodial or petaloid configurations were prob- ably characteristic of these cells in livo. Cells from one animal not only had extensive pseudopods, but also had microvilli protruding over the cell surfaces (Fig. 2). In thin sections viewed by TEM the surface membranes of these cells were greatly folded. Extensions that formed bladders were present in numerous cell profiles (Fig. 3). The extensive folding of the surface made it difficult to dis- tinguish surface membranes from what appeared to be "clear" vesicles. Lysosoincs The cytoplasm contained many lysosomal vesicles. Cells stained with neutral red (Fig. 4) and viewed by light microscopy showed large numbers of granules, positive for the stain, concentrated in the central region of the cells. These granules were not observed in pseudopodial extensions. Cells in Figure 4 were photographed after they adhered to the microscope glass slide and hence displayed the filopodial configuration. In TEM the type of vesicles located centrally in coelomocytes, and seen in large numbers within a cell, contained electron-dense material sur- rounded by less dense material (Fig. 5). These vesicles were of various densities, perhaps reflecting their stage of development. The denser areas within these vesicles .may be regions of condensation or "crystallization" of material packed within the vesicles (Fig. 6). Another structure, which may be related to a stage in the maturation of these vesicles (or may be an artifact of sample preparation), is shown in Figure 7. These very dense structures often contained a clear area which in turn contained material of even greater density than the main structure. These very dense struc- tures are probably also lysosomes, since such structures are associated with rod- shaped material (Fig. 8) that resembles the material within lysosomal vesicles (cf. Fig. 6). Hence, vesicles numerous in the cell and containing dense material surrounded by less dense matrix material are probably the lysosomal neutral-red granules observed by light microscope cytochemical methods. Furthermore, struc- tures resembling the dense rod-shaped material of these vesicles occur in digestive vacuoles (see below) and support our interpretation of the identification and function of these organelles in the cell. 298 E. S. KANESHIRO AND R. D. KARP 1.0pm FIGURE 1. Scanning electron micrograph of coelomocytes of Dertnastcrias imbricata show- ing lamellipodial extensions of the cell surface. X 2000. SKA STAR COELOMOCYTES 299 Other cell organdies \\here surface extensions formed lamellipods, the protrusions were filled with microfilaments (Fig. 9). Thick linear bundles of filaments in the cytoplasm radiated to the surface of the pseudopod. Organelles such as Ivsosomes and the nucleus were excluded from these protrusion zones. The nucleus of coelomocytes was usually bean-shaped or notched, although various profiles were seen because of varying section planes (Fig. 9). Coelomo- cytes from normal and experimental sea stars were treated with Giemsa stain and observed by light microscopy at various times following removal of coelomic fluid. No nuclear division or mitotic activity was observed. A Golgi complex was often seen in the notched region of the nucleus as well as in other parts of the cytoplasm. ( iolgi complexes were generally well developed and had several evenly and closely spaced stacks of lamellae (Fig. 10). Mito- chondria were elongate (Fig. 11). Coelomic fluid samples contained a few cells which had centrioles (basal bodies) associated with striated rootlets (kinetodesmal fibers) (Fig. 12). The combination of these structures suggested that those cells were ciliated or flagellated. These may have been cells normally at other sites but moved during coelomic fluid sampling by a hypodermic needle and syringe. On the other hand these cells may be analogs to the "vibratile" cells described as unique to sea urchins (Johnson, 1969b). Strongyloccntrotus "vibratile" cells move by a single flagellum in the coelomic fluid. Also uncommon within coelomic fluid samples was a cell with a large vesicle filled with filamentous material (Figs. 13, 14). Again this may not be an in situ cell of the coelomic fluid. It is also possible that this cell normally can come from, or go to another tissue such as the Tiedemann's bodies (see below). Phagocytosis \Yashed suspensions of a pure culture of a marine Acroinonas were injected into sea star coeloms to characterize the phagocytic activity of coelomocytes. Coelomocytes were prepared for TEM at various periods from 5 min to 3 days after injection. By 5 min after the injection, coelomocyte morphology had changed. Large numbers of vacuoles contained structures undergoing digestion (Fig. 15). Bacteria engulfed by pseudopodial activity (Fig. 16) were observed within phagosomes in the cell (Fig. 17). Some bacteria in phagocytic vacuoles were lysed, indicating that digestion had begun (Fig. 18). Vacuoles containing heterogeneous structures were present (Fig. 19). The vacuoles in Figure 19 also contained dense material typical of the lysosomal matrix (see above). Myelin- like figures, indicative of cellular digestive processes, were abundant. These myelin-like figures were present in vesicles (Fig. 20), large vacuoles (Fig. 21), FIGURE 2. Dcnnastcrias imbricata coelomocytes with microvilli protruding from their surfaces. Most cells in these samples had microvilli. X 2780. FIGURE 3. Thin section of a coelomocyte, showing tiie hladder or petaloid surface mor- phology. Note the highly folded surface membrane and numerous cytoplasmic vesicles. X 6750. FIGURE 4. Light micrograph of a coelomocyte stained for Ivsosomes witli neutral red. The cell had adhered to the microscope glass slide and thus exhibits filopodial surface exten- sions. Neutral red-stained granules are restricted to central regions of the cell and do not extend into filopods. X 920. FIGURE 5. Section through a coelomocyte showing high concentrations of lysosomal vesicles. X 13,390. 300 E. S. KANESHIRO AND R. D. KARP _ 8 1.0pm 10 1.0pm FIGURE 6. Lysosomal vesicles containing more condensed or electron-dense material surrounded by less dense matrices. Denser material often appears in rod-like profiles. This cell was taken from an animal 60 inin after bacteria were injected into the coelom. Lysosomal vesicles are similar in cells of untreated animals. X 24,700. FIGURE 7. Dense vesicles containing clear regions in the cytoplasm. The clear regions contain very dense material (cj. more typical lysosomal vesicles next to them and in Fig. 6). SEA STAR COELOMOCYTES 301 and cell cytoplasm. In many cell profiles it was not possible to determine whether structures that appeared to he related to digestion were distinct phagosomes (secondary lysosomes) or autophagous structures. \Ye did not determine whether coelomocytes self-destruct after they perform their phagocytic function, recover and remain in the coelomic fluid, or are removed from the coelomic fluid. Acronionas introduced into the coelomic fluid of sea stars by direct injection of a bacterial suspension were effectively cleared from the coelomic fluid within 2-3 days. No bacterial colonies formed when 0.1 nil of coelomic fluid, withdrawn before injection of bacteria, was plated. This indicates that the coelomic fluid is normally aseptic. Figure 22 shows the rate of disappearance of bacteria from the coelomic fluid of two animals. Also shown is the concentration of coelomocytes in these animals over this time period. Relationship with the Tiedemann's bodies Tiedemann's bodies may serve as sites from which coelomocytes arise, through which coelomocytes pass, or where coelomocytes can associate. Our studies do not establish coelomocyte genesis but do suggest a possible relationship and some similarities between cells of the organ and circulating coelomocytes. Tiedemann's bodies are located on the inner side of the peristomial ring and are outpockets of the ring canal of the water vascular system. The paired bodies are approximately 2 mm in diameter and are hollow and highly folded. The folds are lined with columnar or squamous epithelial cells (Hyman, 1955). Figure 23 is a section through an infold showing the lining of epithelial cells and a cell (coelomocyte) in the lumen. The animal was not injected with bacteria but clearly contained numerous bacteria that were probably present in the natural environment and/or the laboratory aquarium. Bacteria were identified in phagocytic vacuoles (Fig. 23) and also within vesicles that resembled lysosomes, i.e. appeared to be filled with matrix material (Fig. 24). Some profiles of epithelial cells show notched nuclei (Fig. 23) suggesting that these cells have the same nuclear configuration as do nuclei of coelomocytes. A cell (presumably a coelomocyte) within the lumen, which is contiguous with the water vascular system, contained a large vacuole with a structure having filamentous material (Figs. 25, 26) identical to that observed in a cell in the coelomic fluid (see above and Fig. 13). The nature of the filamentous material is not known, but it resembles smooth muscles seen in Tiedemann's bodies. These observations suggest that coelomocytes can migrate from coelomic fluid, which bathes the exterior of Tiedemann's bodies, to the interior of the organ, which is part of the water vascular system. This cell was taken from an animal 5 min after bacteria were injected into the coelomic cavity. These structures are also present in cells of untreated animals. X 39,1 10. FIGURE 8. A vesicle with structures resembling those shown in Figures 6 and 7. The rod- shaped element within the very dense matrix suggests these are lysosomes in different stages of maturity. This cell was taken from an animal 5 min after bacteria were injected into the coelom. X 37,660. FIGURE 9. A cell with filamentous material in pseudopodial protrusions. Thick bundles (arrows) extend from the central region of the cell to the cell surface. Organelles such as lysosomes and the nucleus are excluded from protrusion zones. This cell was taken from an animal 5 min after bacteria were injected into the coelomic fluid. X 8410. FIGURE 10. A Golgi apparatus in the cytoplasm of a coelomocyte. Lamellae are even and numerous vesicles are present on the convex mature face of the complex. This cell was taken from an animal 5 min after it was injected with bacteria. Golgi complexes are the same in cells of untreated animals. X 32,880. FIGURE 11. A mitochondrion sectioned along its length. The outer membrane and cristae are visible. Electron-dense deposits (also see mitochondria in Fig. 8), probably divalent cation deposits, are present. This cell was taken from an animal 60 min after it was injected with bacteria. Mitochondria have the same morphology in cells of untreated animals. X 37,180. 302 E. S. KANESHIRO AND R. D. KARP • 'Si* 1"-'A ' 4 — . ^ I kfB&Sfe* FIGURE 12. A cell with basal bodies and striated rootlets, from a coelotnic fluid sample. The rootlet appears to associate with the nucleus. X 16,320. FIGURE 13. A cell with a large vacuole filled with filamentous material from a coelotnic fluid sample (cf. Figs. 25, 26). The nature of the filamentous material is not known. This cell was taken from a sea star 60 min after the animal was injected with a bacterial suspension. X 6300. SEA STAR COELOMOCYTES 303 Some cells within the Tiedemann's bodies were flagellated (Fig. 27) and these organelles were associated with elaborate striated rootlets. As described above, we have observed a cell in the coeloniic fluid with a basal body and rootlet. Concentration of coelomocytes in coeloniic fluid Concentration of coelomocytes varied widely from animal to animal. Cell counts ranged from 1.5 X lO"'/ml to 1.2 X 10: ml in 57 untreated sea stars. There was no obvious correlation of cell number with the animal's size. Similar hetero- geneity probably occurs in the natural population. A wide range of percent change in coelomocyte concentration was observed after withdrawal of 7.5 ml (experi- mental) or 0.5 ml (control) coeloniic fluid. Increased concentrations of coelomo- cytes were observed in 3 out of 6 animals after 1 day ; 9 out of 1 1 after 2 days ; 12 out of 15 after 3 days; and 5 out of 10 after 4 days. Some animals did not show any change and a few showed lower coelomocyte concentrations. Weights of animals taken before withdrawal of 7.5 ml fluid and weights taken at 3 or 4 days after coeloniic fluid removal indicated that the animals regained or increased their total body weights. At day 3 there was a mean increase of 5.3 g (±8.6 SD, N = 18) and at day 4 a mean increase of 29.5 g (±11.5 SD, N = 8). DISCUSSION Cell surface The coelomocytes of the sea star, Dcnnasterias iinbricata, have the same general morphology as those reported in echinoids and holothurians (Johnson, 1969a, b, c; Fontaine and Lambert, 1973, 1977). Transformation from lamellipodial or petaloid forms to the filopodial form was observed when cells adhered to glass slides. This transformation has recently been carefully documented in echinoid coelomocytes by Edds (1977) who detected reorganization of microfilament bundles when cells were allowed to settle on glass. The cell shape changes are probably controlled by an actin-myosin-based system, since microfilaments are numerous in pseudopods. Actin has been isolated from echinoid coelomocytes in milligram quantities and its migration in gels characterized (Edds, 1977, 1979; Otto ct al., 1979). Echinoid coelomocytes change from petaloid to the lamellipodial forms upon settling on a glass substratum and then to the filopodial form upon hypotonic shock (Otto ct al., 1979). The transformation of Dcnnasterias coelomocytes to the filopodial form apparently did not require hypotonic shock, since the change occurred in cells in undiluted coelomic fluid. Microvilli of various cell types are also known to contain microfilaments and actin-like molecules (Mooseker and Tilney, 1975). Our observations that "villous" structures can appear on surfaces of sea star coelomocytes indicate that these cells have the same capacities as vertebrate macrophages and lymphocytes to produce microvilli (Cohn, 1968). Whether or not smooth vs. "villous" surface mor- phologies were the results of fixation procedures was not examined in this study. The usual "villous" surface morphology of human T- and B-lymphocytes appear FIGURE 14. A higher magnification of the filamentous material shown in Figure 13. X 31,790. FIGURE 15. A coelomocyte 60 min after bacteria were injected into the coelomic cavity of the sea water. Bacteria are in vacuoles and digestive processes are indicated by the presence of myelin-like figures, and by numerous vacuoles of mixed contents. X9100. FIGURE 16 A bacterium being "lassoed" by pseudopodial activity of the coelomocyte surface. This sample of coelomic fluid was taken 5 min after bacteria were introduced into the coelom. X 35,200. 304 E. S. KANESHIRO AND R. D. KARP FIGURE 17. Several bacteria (arrows) within phagocytic vesicles 5 min after bacteria were introduced into the coelomic cavity of sea stars. X 12,530. SKA STAR COELOMOCYTES 305 2 - o < CD 1 - o o m |— O ^ O o m C/> x o CD TIME (days) FIGURE 22. The rate of clearing of bacteria from the coelomic fluid of two sea stars (open and closed squares) and concentration of coelomocytes in the t\vo animals (open and closed circles). Animals with bacteria-free coelomic fluid were injected with 3.5 X 107 cells of Acromonas sp. smooth after various fixation procedures (Alexander ct al., 1976). Our observa- tions establish that unlike many cell types, these coelomocytes can exhibit micro- villi-covered surfaces when prepared for and viewed by SEM. Phagocytosis Numerous studies indicate that echinoclerm coelomocytes can actively phago- cytize and clear foreign material within a few days after its injection ( Bang and Lemma, 1962; Endean, 1966; Johnson, 1969c; Johnson ct al., 1970; Reinisch and Bang, 1971; Unkles and Wardlaw, 1976; Wardlaw and Unkles. 1978; Coffaro, 1978). Dermasterias, whose normal coelomic fluid was found to be aseptic, also rapidly disposed of injected gram-negative bacteria. Johnson ct al. (1970) provided ultrastructural evidence that coelomocytes from Strongylocentrotus phagocytize bacteria in vitro in hanging drops. The cells, however, did not have recognizable primary lysosomes and bacterial cells taken up did not appear FIGURE 18. A bacterium within a phagosome shows signs of being digested. The arrow points to a broken region of the bacterial surface. This cell was taken from the animal 60 min after the animal was injected with bacteria. >< 33,860. FIGURE 19. Large digestive vacuoles are common in the cytoplasm of cells taken 60 min after sea stars were injected with bacteria. The vacuoles contain rod-shaped elements (arrow) similar in density and character to those seen in lysosomes (cf. Fig. 6). X 53,210. FIGURE 20. Vesicles with homogeneous matrix material similar to that seen in lysosomes contained myelin-like figures. This vesicle was in a cell taken 5 min after the animal was injected with bacteria. X 22,160. FIGURE 21. Vacuoles containing myelin-like figures in a cell taken 5 min after the animal was injected with bacteria. X 21,030. 306 E. S. KANESHIRO AND R. D. KARP 23 FIGURE 23. A section through a fold within a Tiedemann's body removed from an untreated sea star. Epithelial cells form a circle around a cell in the lumen. The arrow points to an epithelial cell sectioned at a notched region of the nucleus. Crosses indicate bacteria within vesicles in epithelial cells, and within the cell (presumably a coelomocyte) in the lumen. X 6380. degraded. Several other studies (Bang and Lemma, 1962; Ghiradella, 1965; Enclean, 1966; Reinisch and Bang, 1971) have reported that coelomocytes phago- cytize foreign matter and then clump within dermal papulae (branchiae). These SEA STAR COELOMOCYTES 307 ••* '-. V/ I 26 FIGURE 24. Lysosome-like vesicles within a Tiedemann's-body cell. These vesicles with homogeneous matrices contain bacteria. ~X 10,900. FIGURE 25. Low magnification of a section through a fold in a Tiedemann's body. The rectangle indicates a cell in the lumen. X 1570. FIGURE 26. A higher magnification of the cell in the lumen (marked in Fig. 24). This cell has a large vacuole filled with filamentous material (cf. Fig. 13). X 8340. FIGURE 27. A flagellated cell in the Tiedemann's body. The pair of basal bodies and associated striated rootlet are similar to that seen in a cell in a sample of coelomic fluid. X 14,370. 308 E. S. KANESHIRO AND R. D. KARP structures then rupture, releasing the coelomocytes to the exterior. Our ultra- structural observations suggest that coelomocytes undergoing extensive phagocytic activity not only are filled with digestive vacuoles but may also participate in autophagous activity. Since the initial bacterial concentration was less than the coelomocyte concentraton, not all digestive vacuoles may have been involved with bacterial elimination. Autophagy in these cells suggests that spent phagocytic coelomocytes of Dcnnastcrias self-destruct in the coelomic fluid. Another pos- sibility is that they are expelled from the animal, as observed in other echinoderms (Bang and Lemma, 1962; Endean, 1966; Reinisch and Bang, 1971). It is interest- ing that the coelomocytes described here show a great deal of similarity to verte- brate macrophages in that they: 1) can produce microvilli on their surfaces, 2) have numerous lysosomes, well developed Golgi complexes (usually located in the nuclear "hof"), notched nuclei, and elongated mitochondria and 3) phagocytize foreign particles forming heterophagic as well as apparent autophagic vacuoles (cf. Cohn, 1968). Tiedemann's bodies Tiedemann's bodies are organs that may function as primitive lymphoid tissue in sea stars. The organs are outpockets of the water vascular system and are bathed on their outer surfaces by coelomic fluid. Hence, all major fluid systems of the organism circulate around the cells of this organ. The cells of Tiedemann's bodies are capable of endocytosing bacteria. The lumina of the deep folds on the water vascular system side of the organ are filled with cells that resemble coelomo- cytes (cf. Hyman, 1955). The epithelial cells of the Tiedemann's bodies are flagellated, as are those of the radial and ring canals (Hyman, 1955), and possibly serve to move fluid in the water vascular system. Vanclen Bossche and Jangoux (1976) have reported that asteroid coelomocytes originate from the coelomic epithelium including the epithelium of Tiedemann's bodies. Since dividing cells were not observed in the coelomic fluid, it appears that circulating coelomocytes of Dcnnastcrias might also come from other tissue sources. Animals recovering from extensive withdrawal of coelomic fluid regained total body weights. If that recovery (or increase) in weight reflects coelomic fluid replacement, then the recovery (or increase) in coelomocyte concentrations must result from coelomocyte recruitment. It is possible that coelomocytes originate from epithelial cells, e.g. of Tiedemann's bodies. Vanden Bossche and Jangoux (1976) reported that epithelial cells of sea stars became detached and rapidly lost their flagella. The rare cells in coelomic fluid with basal bodies and striated rootlets may represent incompletely transformed recent recruits. The findings reported here reveal indirect evidence that this organ may contribute to the coelomocyte population as well as play an important role in the animal's defense system. 1'hagocytic cells participate in rejection of integumentary allografts in Dcnna- stcrias. The rejection site is heavily infiltrated by large phagocytic cells and smaller mononuclear cells. Autografts do not contain these cell types (Karp and Hilde- mann, 1976). A number of recent findings implicate the phagocytic amoebocyte as the effector cell. Karp and Johns (1978) reported that coelomic amoebocytes can be stimulated in vitro by such mitogenic substances as bacterial lipopolysac- charide and concanavalin A. Bertheussen (1979) reported that the echinoid coelomic phagocyte was able to recognize and react to allogeneic and xenogeneic cells in culture. Other echinoid cell types did not show this activity. Various studies have reported that filopodial coelomocytes participate in "clot" formation SEA STAR COELOMOCYTES 309 and in the elimination of foreign material ( Karp and Coffaro, 1(JSO). In addi- tion, preliminary studies indicate that the coelomocyte of Ih-niiastcrias has surface proteins with molecular weights similar to vertehrate (human and murine) trans- plantation antigens (L. A. Rheins, J. D. Stinnett, and K. 1). Karp, University of Cincinnati, unpublished results). Thus, the phagocytic coelomocyte may mediate specific immune defense reactions in the sea star, and possibly in i-chinoderms in general. This cell may provide insight into how the sophisticated immune mecha- nisms of vertebrates evolved. This work was supported in part by a University of Cincinnati Research Council and Biomedical Research Support awards to E.S.K. and a University of Cincinnati Research Council award and a N.I.A.I.D. research grant (AI 15601) to R.D.K. The authors thank Dr. C. A. Huether and Ms. V. K. Beckham for their assistance in the project and Dr. K. T. Edds for suggestions and comments concerning the manuscript. SUMMARY 1. The structure of coelomocytes from Dermasterias iinbricata was characterized by light microscopy and scanning and transmission electron microscopy. The petaloid or lamellipodial configurations were the prevalent forms in freshly drawn coelomic fluid. Cells produced filopodia when they adhered to glass slides. 2. Lysosomes were stained with neutral red and observed by light microscopy, and variations in density and structures of their matrices described by TEM. 3. Animals injected with a bacterial suspension cleared the bacteria from their coelomic fluid within 2-3 days. Ultrastructure of coelomocytes in these animals indicated that phagocytosis and digestion of bacteria were rapid and involved lysosomes. 4. With TEM numerous bacteria were observed within cells of Tiedemann's bodies from untreated animals, indicating that this organ may play an important role in clearing the coelomic fluid and water vascular system of foreign particles. 5. Giemsa-stained coelomocytes of untreated animals and animals recovering from coelomic fluid removal indicated that coelomocytes did not undergo mitosis within the coelomic fluid. Removal of coelomic fluid resulted in recovery or increase in total body weights and coelomocyte concentrations. Hence, circulating coelomocytes must be recruited from other tissue sources in the animal. LITERATURE CITED ALEXANDER, E., S. SANDERS, AND R. BRAYLAN, 1976. Purported difference between human T- and B-cell surface morphology is an artifact. Nature, 261 : 239-241. BANG, F. B., AND A. LEMMA, 1962. Bacterial infection and reaction to injury in some echino- derms. /. Invertcbr. Pathol, 4: 401-414. BERTHEUSSEN, K., 1979. The cytotoxic reaction in allogcneic mixtures of echinoid phagocytes. Ex p. Cell Res., 120: 373-381. BERTHEUSSEN, K., AND R. SELJELID, 1978. Echinoid phagocytes in vitro. E.rf>. Cell Res., Ill: 401-412. BOOLOOTIAN, R. A., AND A. C. GiESE, 1958. Coelomic corpuscles of echinoderms. Biol. Hull., 115: 53-63. CAVANAUGH, G. M., Ed., 1964. Marine Diolot/ical Laboratory Formulae and Methods V. Mar. Biol. Lab., Woods Hole, Massachusetts. P. 52. CHIEN, P. K., P. T. JOHNSON, N. K. HOLLAND, AND F. A. CHAPMAN, 1970. The coelomic elements of sea urchins ( Stromjylocentrotus) . IV. Ultrastructure of the coelomo- cytes. Protoplasnw, 71 : 419-442. 310 E. S. KANESHIRO AND R. D. KARP COFFARO, K. A., 1978. Clearance of bacteriophage Tt in the sea urchin, Lvtechimus Rictus, J. Invert ebr. Pat ho/.. 32 : 384-385. COHN, Z. A., 1968. The structure and function of monocytes and macrophages. Adv. Immunol.. 9: 163-214. EDDS, K. T., 1977. Dynamic aspects of filopodia formation hy reorganization of microfilatnents. /. Cell Biol.. 73: 479-491. EDDS, K. T., 1979. Isolation and characterization of two forms of a cytoskeleton. /. Cell Biol., 83: 109-115. ENDEAX, R., 1966. The coelomocytes and coleomic fluids. Pp. 301-328 in R. A. Boolootian, Ed., Physiology of Echinodcnnata, Interscience, New York. FONTAINE, A. R., AND P. LAMBERT, 1973. The fine structure of the haemocyte of the holo- thurian, Cucumaria niiniata (Brandt). Can. J. Zoo/., 51: 323-332. FONTAINE, A. R., AND P. LAMBERT, 1977. The fine structure of the leucocytes of the holo- thurian, Cucumaria niiniata. Can. J. Zoo/., 55 : 1530-1544. GHIRADELLA, H. T., 1965. The reaction of two starfishes, Patiria niiniata and Asterias forbes-i, to foreign tissue in the coelom. Biol. Bull., 128 : 77-89. HETZEL, H. R., 1963. Studies on holothurian coelomocytes. I. A survey of coelomocyte types. Biol. Bull., 125 : 289-301. HYMAN, L. H., 1955. The Invertebrates: Echinodcnuata. the Coclomatc Bilateria, Vol. IV. McGraw-Hill, New York. Pp. 274-283. JOHNSON, P. T., 1969a. The coelomic elements of sea urchins ( Strongyloccntrotus}. I. The normal coelomocytes, their morphology and dynamics in hanging drops. /. Invertebr. Pathol., 13: 25-41. JOHNSON, P. T., 1969b. The coelomic elements of sea urchins (Strongylocentrotus). II. Cytochemistry of the coelomocytes. Histocheinie, 17 : 213-231. JOHNSON, P. T., 1969c. The coelomic elements of sea urchins (Strongyloccntrotus).- III. In litro reaction to bacteria. /. Invertebr. Pathol., 13 : 42-62. JOHNSON, P. T., P. K. CHIEN, AND F. A. CHAPMAN, 1970. The coelomic elements of sea urchins (Strongyloccntrotus). V. Ultrastructurc of leucocytes exposed to bacteria. /. Invertebr. Pathol.. 16: 466-469. KARP, R. D., AND W. H. HILDEMANN, 1976. Specific allograft reactivity in the sea star, Dcrmastcrias inihricata. Transplantation, 22 : 434-439. KARP, R. D., AND J. D. JOHNS, 1978. Evolution of immune reactivity: Mitogenic responsive- ness in the sea star, Dcrmasterias iinbricata. 7St!i Annual Meeting, Am. Soc. Micro- bio}., p. 48. KARP, R. D., AND K. A. COFFARO, 1980. Cellular defense systems in the Echinodermata. In N. Cohen and M. M. Sigel, Eds., The Reticuloena'othelial System: A Comprehensive Treatise, Vol. V. Plenum Press, New York. In press. MOOSEKER, M. S., AND L. G. TiLNEY, 1975. The organization of an actin filament-membrane complex : Filament polarity and membrane attachment in the microvilli of intestinal epithelial cells. J. Cell Biol., 67 : 725-743. OTTO, J. J., R. E. KANE, AND J. BRYAN, 1979. Formation of filopodia in coelomocytes: Localization of fascin, a 58,000 dalton actin cross-linking protein. Cell, 17 : 285-293. REINISCH, C. L., AND F. B. BANG, 1971. Cell recognition reactions of the sea star (Asterias vulgaris) to the injection of amebocytes of the sea urchin (Arbacia pune/ulata) . Cell. Immunol., 2: 496-503. SANDERS, S. K., E. L. ALEXANDER, AND R. C. BRAYLAN, 1975. A high yield technique for preparing cells fixed in suspension for scanning electron microscopy. /. C ell Bio/., 67: 476-480. SOLDO, A. T., AND E. J. MERLIN, 1972. The cultivation of symbiote-free marine ciliates in axenic medium. J. Protozool., 19: 519-524. SPURR, A. R., 1969. A low viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res., 26: 31-43. UNKLES, S. E., AND A. C. WARDLAW, 1976. Antibacterial activity in the sea urchin, Echinus esculentus. Soc. Gen. Microbiol. Proc., 3 : 182. VANDEN, BOSSCHE, J. P., AND M. JANGOUX, 1976. Epithelial origin of starfish coelomocytes Nature, 261 : 227-228. WARDLAW, A. C., AND S. E. UNKLKS, 1978. Bactericidal activity of coelomic fluid from the sea urchin, Echinus esculentus. J. Invertebr. Pathol., 32 : 25-34. Reference: Biol. Bull., 159: 311-318. (October, 1980) PHOTOPERIODIC CONTROL OF DIAPAUSE IN THE MARINE CALAXOID COPEPOD LA1UDOCKRA AI'STll/A ' NANCY II. MARCUS ll'oods Hole Occanoyniphic Institution, ll'ootls Hole, Massachusetts <>25-f3 It has long been recognized that many marine copepods are seasonally present or absent in the plankton. Fish and Johnson (1937) postulated that such species probably survived as resting eggs on the sea bottom during periods unfavorable for their existence in the plankton. Resting eggs in marine bottom sediments have now been documented for several temperate neritic calanoid copepods and cladocerans (Grice and Gibson, 1975, 1977; Johnson, 1980; Kasahara ct a!., 1974, 1975; Onbe, 1973, 1978), and it has been shown for some of these species that the maximum abundance of resting eggs in the sediments alternates seasonally with the maximum abundance of individuals in the plankton (Kasahara et al., 1975). However, it has been demonstrated experimentally for only a few species that these resting eggs are in diapause rather than quiescence (Grice and Gibson, 1977; Johnson, 1980; Kasahara and Uye, 1979; Marcus, 1979; Zillioux and Gonzalez, 1972). Diapause typically necessitates complex changes at the biochemical and physio- logical levels (see Clutter, 1978). In an environment that fluctuates in a predict- able way, the ecological and evolutionary success of species must depend on individuals being able to accurately forecast the changes so that sufficient time is allowed for the requisite adaptive responses prior to the onset of unfavorable con- ditions. Marcus (1979) demonstrated that the onset of egg diapause in the marine calanoid copepod Labidocera acstiva precedes the decline in surface water tem- peratures by 2 weeks, and suggested that some factor other than temperature alone was important in triggering the onset of dormancy in this species. Data presented in this paper demonstrate that photoperiod significantly affects the induction of diapause in L. acstiva. MATERIALS AND METHODS Rearing of nauplii, copepodites, and adults Labidocera acstiva were reared from the first naupliar stage through to repro- ductive maturity in 19 1 glass carboys containing 5-//.m filtered sea water. The carboys were mounted at an angle on a frame which rotated at 2 rpm. This rotation served to keep the algal food (see below) in suspension. The entire apparatus was in a walk-in incubator equipped with temperature controls and a 24-hr light : dark cycle, with fluorescent lights providing illumination of 200-300 ft-candles. All developmental stages were fed a mixture of Gymnodinium nclsoni. Gonyaula.r polycdra, rroroccntnim inicans, and I'cndininui troclioidcmn. Dino- flagellates were obtained from Robert Guillard, Woods Hole Oceanographic Insti- tution, Woods Hole, MA, and were subsequently cultured in Fernbach flasks con- taining Guillard's f/2 media (Guillard, 1972) at 17°C and 12L:12D cycle. 1 Contribution Number 4554 of the Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543. 311 312 NANCY H. MARCUS Before the copepods reached reproductive maturity, the contents of each carboy were siphoned weekly through a 63 /an Nitex screen to collect surviving copepods. The vessels were washed, refilled with fresh filtered seawater, and supplied with food. The dinoflagellates were added in equal numerical concentrations to give a final density in the carboy of 7.0-10.0 >: 10L> cells/ml. Surviving copepods were returned to each carboy. When egg production began, the number of males and females in each carboy was equalized to a density of 25-30 individuals of each sex by removing adults at random with a wide-mouth pipette from the filtrate. Subse- quently the carboys were siphoned every 2-3 days so that eggs could be collected before they hatched. Eggs were removed from the filtrate (retained by the screen) with a micropipette and placed in 5 -/tin filtered sea water. Adults were returned to the carboy. The weekly schedule for changing the sea water and food remained the same. Experimental procedures Carboy populations were initiated with 200 nauplii derived from subitaneous eggs (carboys 7, 8, 9. 12, 14, 15) or chilled resting eggs (carboys 4 and 6) produced either in the laboratory (first generation) or by freshly collected field females. Copepods were reared at 13.5°-15.5°C in five carboys exposed to a short-day photoperiod (8L: 16D), and three carboys exposed to a long-day photo- period (18L:6D). Eggs were collected from each carboy every 2-3 clays for 8—10 days. At each collection the total numbers of eggs and surviving males and females were determined, and a sample of 100-120 eggs (from each collection day) was placed in glass-fiber-filtered sea water and incubated at 25°C and 12L: 12D to hasten the time to hatching. After 4—5 days the percent hatch was determined, and unhatched eggs were placed in two jars (50 eggs/jar) at each of 5° and 25 °C. The eggs were kept at these temperatures a minimum of 40 days. At the end of this interval the 5°C eggs were warmed at 25 °C and the percent hatch ascertained. At the same time, the hatch of eggs incubated at 25 °C was also determined. In addition, a minimum of 20 adult males and females from each carboy were preserved in 5% buffered formalin and subsequently measured with a stereomicroscope to determine cephalothorax and total body length. The body sizes of these copepods and adults collected in the field under similar temperature conditions were compared, to assess the suitability of growth conditions in the laboratory. RESULTS The age at reproductive maturity of Labidoccra a estiva reared in the laboratory at 13.5°-15.5°C ranged from 272 ± 1.3 days at 8L : 16D to 22.3 ± 1.5 days at 18L : 6D. The survival of individuals to reproductive adulthood was good (65-75%) for each experimental carboy. Copepods reared under the long-day regime of 18L:6D produced more eggs during the experimental period (800-1000 eggs in each carboy per day) than the individuals reared under the short-day regime of 8L: 16D (227-580 eggs in each carboy per day). Table I shows the percent hatch data for one set of eggs for each collection day for carboy 6, and is representative of results for the other egg set as well as the other carboys. For carboy 6, final hatch at 25 °C was high (> 90%) after chilling at 5°C for incubation periods ranging from 49 to 123 days. The hatch of eggs kept at 25°C appears to increase (26-43%) as the incubation period DIAPAUSE IN LABIDOCI-.R.I AESTIVA 313 TAHLK I Percent hatch of eggs produced by females reared at 13.5°-15.5°C and KL:16D in Carbav f>. Eggs that did not hatch ivithin 4 -5 days at 25°C (initial) u'ere incubated in jars at 5° or 25°C. The final hatch of these eggs at 25° C is indicated, as icell as the period of incubation. Final hatch Date- Initial hatch Incubation collected at 25°C (days) 5°C 25°C 3 August 1979 4'; 49 95 '•; 2(.' , 6 August 1979 r, 62 97'; 36^ 8 August 1979 7% 81 ')()', 399; 10 August 1979 2% 99 933 37% 13 August 1979 7% 123 91', 43% increases. The percent hatch of eggs and adult body lengths for each carboy is shown in Table II. The values for percent initial hatch for each carboy represent an average of values determined each time the eggs were collected. Similarly, the values for percent final hatch for each carboy after incubation at 5° and 25° C represent an average of values derived from the unhatched eggs subsequently incubated from each collection day. Subitaneous (quickly hatching) and diapause egg production were strikingly different between the two photoperiocl regimes. At 8L : 16D, 13.2 ± 4.3% of the eggs produced hatched within 4-5 days (classified as subitaneous), whereas at 18L: 6D, 89.9 ± 1.4% of the eggs were subitaneous. The similarity of subitaneous egg production among replicates was slightly less at 8L:16D than at 1SL:6D, ranging from 2.9 ± 0.8%- to 23.8% ± 1.9, and 87.3 ± 2.5% to 92.0 ± 1.7%, respec- tively. Eggs produced at 18L:6D that did not hatch within 4-5 days appeared brown and granular. They were dead eggs (Marcus, unpublished), and were not incubated at 5° or 25°C. Most of the unhatched eggs produced at 8L : 16D were green and had a clear perimeter. These eggs were incubated at 5° and 25 °C and. as shown in Table II, an average of 85.9 ± 3.1% hatched after chilling and were classified as diapause eggs. The remainder, which did not hatch, were dead. Some of the eggs (38.1 ± 5.3%) kept at 25°C did hatch during the incubation interval. The difference in body lengths of adult males and females reared in replicate carboys for a given photoperiod regime was small. The widest range of values resulted at 8L: 16D for male total length (2.10 ± 0.11 mm to 2.24 ±0.13 mm). For each experimental regime females were longer than males (both total length and cephalothorax). Moreover, individuals reared under the short-day regime (8L:16D) were larger than individuals reared at 18L:6D. DISCUSSION The results presented in this paper provide the first evidence of the importance of photoperiod as a trigger for the induction of diapause in a marine copeopd, although this effect has been shown for several fresh- water copepods and cladocerans (Stress and Hill. 1965; Stress, 1969a, b; Watson and Smallman, 1971; Bunner and Halcrow, 1977). Copepods which developed at 13.5°-15.5°C under 18L:6D produced 89.9 ± 1.4% subitaneous eggs, whereas individuals reared at 8L:16D produced only 13.2 ± 4.3% subitaneous eggs. Moreover, 85.(>±3.1% of the non-subitaneous eggs produced at 8L:16D hatched synchronously at 25 °C after 314 NANCY H. MARCUS o 15 3 ;. a j O B , O S i °° o ~. "a 2 a -a: ~e i: ~ ^ = a a <-, *> -o >^ ii » A *> is ^ ft, •- 5 1 ^ OO O 00 ^ S ^ ~ c o J -H -H -H -H -H ^ r*; £i -r o -H -H -H -H 01 72 'i 25 -H i i i i 111 "' r- 10 os vo t- in oo 10 o rv] ^ h U o — ~ — c; -H -H -H -H -H -H -H -H -H -H ^ r- r. r, . <: : 0 O LO LO uo T. a u '•?. O C: C- oo oo 0 — ; 0 0 ~ \O ^i C^ t^- o H 41 HH -H -H -H t-* O r*- O OC -H -H -H -H i o r- t^. oo vrj o -f GO ^O O cs "^ '. ^ ""^ "~^. — . ~~. ^ "I ^ — . U -U -H -H -H -H -f C: oo O 1^1 -H -H -H -H LO O O O r^. ^O ^O f^** o ^1 -f — -f t- M* U o LO r-l -H -H -H -H -H O t^. — -f -t< -H - o ^ -t -t -t- -t OC c/> c OO ^^ OO OO — QJ U o -H 4i -H -H -H -H 0^ C 41 O OO O^ 00 OO 1-^ 00 0. TJ c 0 ^1 •J-J rt r<^ C^l ^< OO C- r<0 ^- < TJ< O ^— i -H -H -H -H -H — ^i O Ov 00 ^ LO oo r— -t r*i _ 1 ^1 -H -H -fl 4J ^l ^ -t O rC r^* O^ -r OS 00 o u rt U -f \O r^ OO Os CN ^ IO Photoperiod C 0 00 c *— < 00 H DIAPAUSE IN LABIDOCERA AESTIVA 315 chilling at 5°C. However, chilling was not an ahsolute requirement to induce hatching, since some eggs kept at 25°C for a prolonged period did hatch (Tahles I. II). The hatching response of diapause eggs produced in the lahoratory under a short-day regime is similar to the pattern demonstrated for diapause eggs pro- duced by field-collected females (Marcus, 1979). Copepods reared in the laboratory under a long-day regime did not produce diapause eggs. The genetic composition of a female ultimately restricts the potential range of egg types she can produce, but the egg type actually realized is determined by the environmental conditions she experiences during development. This is evident from the production of diapause eggs at 8L : 16D by females which developed from both diapause eggs (carboys 4 and 6) and subitaneous eggs (carboys 7, 8, and 9). Similarly, subitaneous eggs were produced at 18L: 6D by females which developed from subitaneous eggs (carboys 12, 14, and 15). The data on body sizes of L. acstiva reported herein provide a valid estimate (see Paffenhoffer, 1970) of the suitability of the laboratory conditions tested in this study. The total lengths and cephalothorax lengths attained by adult males and females reared at 13.5°-15.5°C under both long- and short-day regimes (Table II) are comparable to the sizes of field adults collected at a similar water tempera- ture in Vineyard Sound, MA (Marcus. 1979). The fact that individuals reared at 18L:6D were smaller than individuals reared at 8L : 16D was probably due to temperature. The incubator in which the copepods were reared undergoes a slight temperature change as a result of the light cycle. When the lights were on. the air temperature in the incubator was maintained at 15.5°C. but with the lights off the temperature dropped to 13.5°C. Presumably the water in the carboys also experienced this temperature shift, although to a lesser extent. Thus, the copepods reared under the 18L : 6D regime experience slightly warmer temperatures overall (average 15.0°C/L: D cycle), than individuals reared at 8L : 16D (average 14.2°C/L:D cycle). Based upon the inverse relationship between body size and temperature which has been documented for Labidoccra acstiva collected in the field (Marcus, 1979), the differences observed for the laboratory-reared copepods can perhaps be accounted for by the slightly different thermal regimes. An alternative explanation, based on feeding patterns, could also account for the observed differences. If L. acstiva feeds primarily at night, it is possible that under the short-day regime (8L:16D) individuals were able to consume more food and therefore attained a larger size. If this were the case, however, it might be expected that greater food consumption would result in an increase in the number of eggs produced. These results were not obtained. In fact, copepods reared at 8L: 16D produced fewer eggs. It may be that a diapause egg is able to overwinter because it contains more storage nutrients than a subitaneous egg. This could result in the production of fewer diapause eggs to balance the increased energy requirement. Without more information on L. acstiva feeding pattern and egg biochemistry, this problem cannot be clarified. Nevertheless, sizes attained by labora- tory animals suggest that conditions used to rear L. acstiva in the laboratory ade- quately simulated the conditions for good development, growth, and reproduction ex- perienced by this species in the field. Therefore, it is assumed that the factors which were shown to stimulate the production of diapause and subitaneous eggs in the laboratory reflect a similar function in the field. The influence of photoperiod on the life history of L. acstiva, as suggested by this study, correlates well with the annual life cycle of this species in Vineyard Sound, MA. (Fig. 1). L. acstiva adults first appear in the plankton in early 316 NANCY H. MARCUS Hours of Daylength 10 12 14 15 14 12 10 9.5 SUBITANEOUS DIAPAUSE ADULTS J FMAMJ J ASQN D Month FIGURE 1. Schematic diagram of seasonal occurrence of adults and production of sub- itaneous and diapause eggs of Lahuloccni ucstii'a in Vineyard Sound, MA, with respect to day length. summer when surface water temperatures have reached 18°-20°C (Marcus, 1979) and the photoperiod is 15L:9D (U. S. Dept. of Commerce, 1979). In August, daylength is somewhat shorter (14L), and water temperature is maximal (22°- 23°C). By mid-September, surface water temperature has declined to 19°C and photoperiod is 12L:12D. Temperatures drop to 15°C by mid-November, and daylength (9 3/4L) is further reduced. In mid-December when nauplii, cope- podites, and adults disappear from the plankton, surface water temperature has dropped to 6°C, and daylength is minimal at 9 1/2L:14 1/2D. Labidoccra acstiva produces both subitaneous and diapause eggs in November in Vineyard Sound (Marcus, 1979). Similarly, L. acstira reared in the laboratory under the regime (13.5°— 15.5° C, 8L:16D) that approximates Vineyard Sound in November produce both subitaneous and diapause eggs. Moreover, the majority of eggs produced by this combination of photoperiod and temperature in both the field and laboratory are diapause eggs. The same combination of factors again prevail in Vineyard Sound in mid-March, but at this time L. acstira adults are not present. When they do reappear in early summer the photoperiod is approxi- mately 15L:9D, and only subitaneous eggs are produced. The results for the alter- nate regime (13.5°-15.5°C, 18L:6D) tested in this study, while not exactly identical to the field situation experienced by L. acstiva, nevertheless suggest that under long-day photoperiods, this species produces subitaneous eggs. The results do not imply that photoperiod is the only factor influencing the production of diapause eggs by L. acstiva, nor that it is of primary importance. Other factors (e.g., temperature, density, food) may have a similar function and/or may modify the effects of photoperiod, as has been shown for insects (Saunders, 1976), fresh- water copepods (Watson and Smallman, 1971), and cladocerans (Runner and Halcrow, 1977; Stress, 1969b ; Stross and Hill, 1965). Thus, it is possible that changes in the quantity or quality of the food supply during the weekly feeding interval may have influenced the type of eggs produced by L. acstira in this study. Most recently, it has been suggested that temperatures below 15°C induce the production of dormant eggs by the estuarine calanoid copepod, .-Icartia cali- forniensis (Johnson, 1980). Gaining insight into the factors which control diapause of marine copepods is fundamental to elucidating the mechanism that regulates their seasonal occurrence in the plankton. DIAPAUSE IN LABIDOCERA AESTIVA 317 I thank George Grice and Timothy Cowles for their helpful comments on the manuscript, and Carol Riley for IUT assistance with tin- re-search. This work was supported by NSF Grant OCE78-08857. SUMMARY The calanoid copepod Labidoccnt acstiva was reared in the laboratory at 15°C. Individuals that developed under a photoperiod regime of 18L:6D produced subitaneous eggs, whereas copepods exposed to a short-day regime of 8L:16D produced mostly diapause eggs. The results indicate that photoperiod is an important factor controlling the life cycle of L. acstiva. It is suggested that in Vineyard Sound, MA. this species produces subitaneous eggs during the summer in response to long daylengths, and in the fall produces mostly diapause eggs in response to short daylengths. LITERATURE CITED BUNNER, H. C, AND K. HAi.cRow, 1977. Experimental induction of the production of ephippia by Daphnia tmujna Straus (Cladocera). Crustaceana, 32: 77-86. CLUTTER, M., 1978 (ed. ) Dormancy and Developmental Arrest. Academic Press, New York, 316 pp. FISH, C., AND M. JOHNSON, 1937. The hiology of zooplankton populations in the Bay of Fundy and Gulf of Maine with special reference to production and distribution. J. Fish. Res. Bd. Can., 3 : 189-233. GRICE, G., AND V. GIBSON, 1975. Occurrence, variability, and significance of resting eggs of the calanoid copepod, Lalndocera aestiva. Mar. Biol.. 31 : 335-357. GRICE, G., AND V. GIBSON, 1977. Resting eggs in Pontella mcadi (Copepoda : Calanoida). /. Fish. Res. Bd. Can.. 34: 410-412. GRICE, G., AND T. LAWSON, 1976. Resting eggs in the marine calanoid copepod, Lalndocera aestiva Wheeler. Crustaccana. 30: 9-12. GUILLARD, R., 1972. Culture of phytoplankton for feeding marine invertebrates. Pp. 29-60 in W. Smith and M. Chanley, Eds., Culture of Marine Invertebrate Animals, Plenum Press, New York. JOHNSON, J., 1980. Effects of temperature and salinity on production and hatching of dor- mant eggs of Acartia californiensis (Copepoda) in an Oregon estuary. Fishery Bulletin. Canada, 77 : 567-584. KASAHARA, S., AND S. UYE, 1979. Calanoid copepod eggs in sea-bottom muds. V. Seasonal changes in hatching of subitaneous and diapause eggs of Tortanus jorcipatus. Mar. Biol., 55 : 63-68. KASAHARA, S., S. UYE, AND T. ONBE, 1974. Calanoid copepod eggs in sea-bottom muds. Mar. Biol., 26: 167-171. KASAHARA, S., S. UYE, AND T. ONBE, 1975. Calanoid copepod eggs in sea-bottom muds. II. Seasonal cycles of abundance in the populations of several species of copepods and their eggs in the Inland Sea of Japan. Mar. Biol., 31 : 25-29. MARCUS, N., 1979. On the population biology and nature of diapause of Labidocera acstiva '(Copepoda : Calanoida) . Biol. Bull, 157 : 297-305. ONBE, T., 1973. Preliminary notes on the biology of the resting eggs of marine cladocerans. Bull. Plankton Soe. J pn., 20: 74-77. ONBE, T., 1978. Distribution of the resting eggs of marine cladocerans in the bottom sedi- ment of Ise Bay and Uragannii Inlet, Central Japan. Bull. J pn. Soc. Sci. Fish.. 44: 1053. PAFFENHOFFER, G., 1970. Cultivation of Calauus hcli/olandicus under controlled conditions. Hclgo'l. Wiss. Mccresunters.. 20: 346-359. SAUNDERS, D. S., 1976. Insect Clocks. Pergamon Press, Oxford. 279 pp. STROSS, R., 1969a. Photoperiod control of diapause in Daphnij. II. Induction of winter diapause in the Arctic. Biol. Bull.. 136 : 264-273. STROSS, R., 1969b. Photoperiod control of diapause in Daphnia. III. Two-stimulus control of long-day, short-day induction. Biul. Bull.. 137 : 359-374. 318 NANCY H. MARCUS STROSS, R., AND J. C. HILL, 1965. Diapause induction in Daphnia requires two stimuli. Science, 150: 1462-1464. UNITED STATES DEPARTMENT OF COMMERCE, 1979. Tide Tables, East Coast of North and South America. U. S. Government Printing Office, Washington, D. C. 293 pp. WATSON, N., AND B. SMALLMAN, 1971. The role of photoperiod and temperature in the induc- tion and termination of an arrested development in two species of freshwater cyclopoid copepods. Can. J. Zoo/.. 49: 855-862. ZILLIOUX, E., AND J. GONZALEZ, 1972. Egg dormancy in the neretic calanoid copepod and its implications to overwintering in boreal waters. Arch. Limnol. European Marine Biology Symposium, 5 : 217-230. Reference: Biol. Bull.. 159: 319-324. (October, 1980! FURTHER IMPROVEMENT UPON MAINTENANCE OF ADULT SQUID (DORYTEUTHIS HLEEKERI] IN A SMALL CIRCULAR AND CLOSED-SYSTEM AQUARIUM TANK GEN MATSUMOTO AND JUNICHI SHIMADA Electrotechnical Laboratory, Optoelectronics Section, Tsukuba Science L ity. Ihaniki 3<>5. Japan The squid Loligo, Ih>r\tcuthis and Sepioteuthis have large nerve fibers that make them particularly amenable to experimentation (Arnold ct al., 1974). How- ever, difficulties in squid experiments result from regional and seasonal limitations on the availability of natural squid stocks. These limitations can be partially elimi- nated if the squid are maintained in an aquarium in the laboratory (Matsumoto, 1976). We tried maintenance of adult squid at our laboratory located at Tanashi, a 3-5 hr drive from the fishing site near Jogashima Island, Kanagawa Prefecture (Matsumoto, 1976). We adopted a closed system with a circular tank where sea water was circulated peripherally along the tank wall to mitigate head-on collisions of the squid with the wall. We adopted this approach mainly because of the con- clusions of Summers and McMahon (1974), and Summers ct al.. (1974), that skin damage resulting from collisions with tank walls was a major factor limiting the survival of the squid in a small (e.g., 1.68 nr) tank. Similar closed seawater systems with tank capacities of 1000 and 10,000 1 were used by Hanlon ct al. (1978) to maintain Loligo plci and Loligo pcalci. with mean survival times of 13-25 days. O'Dor ct al. ( 1977) obtained 32-82 day survival of specimens of Illc.r illeccbrosns by mitigating collisions using a 15-m-diameter circular tank in an open system. We maintained squid (Dorytcuthis blcckcri) in our closed system tank for as long as 2 weeks. We concluded that filtering ability was essential to squid survival in our closed system, and that skin damage was minor among causes of mortality in a 2-week maintenance period (Matsumoto, 1976). The question became : Can survival be improved if filtering ability is increased ? MATERIALS AND METHODS The source of squid (Dorytcittliis blcckcri), method of transporting them, and the aquarium system used in these experiments was similar to the one described in Matsumoto (1976) : It consisted of an inner and an outer circular tank, three filter- ing stages, a charcoal filter, a dirt trap, a circulation pump, a temperature controller, and an air bubbler (Fig. 1). The outlet was directed so that filtered sea water flowed peripherally along the circular walls. Net patterns were drawn with black vinyl paint on the inner and the outer tank walls (Fig. 2). The peripherally circulating flow and these net patterns were particularly useful in reducing squid collisions with the tank walls (Matsumoto, 1976). Inner and outer tank walls were light brown polyethylene. Their diameters were 1.5 and 0.5 m, respectively, and their depth 1.2 m. Filtering stage 3, the charcoal filter, dirt trap, circulation pump, and temperature controller were the same as in Matsumoto ( 1976). Filtering stage 1 was composed of layers of commercially available zeolite and sand. The average diameter of zeolite grains was about 3 mm, and the thickness of the layer 15 cm. The 319 320 G. MATSUMOTO AND J. SHIMADA Inner E CM Outer tank -0.5m- ^.^-..-••'.c^--.-..-,-;,/ •:£\t'r-'£''^--'*. 1.5 m •.'->! -:»-./.:.-ir»V3r £-*-•';.:'> :-:'. '•-•"•'.:.'.-?. •><>>:• ^» Dirt trap CharcoaT filter Circulation pump and Temperature controller FIGURE 1. Configuration of the aquarium system. The arrows show the sequence of recirculating sea water through the system. average diameter of the sand grains was 2 mm and the layer's thickness in most runs was 20 cm. The zeolite in the filter weighed 20 kg. In our previous system (Matsu- moto, 1976) filter 1 contained a 5-cm layer of zeolite. Filtering stage 2 was the same as described in our previous report, except that 10 kg of crushed oyster shell was used to cover the filter surface. No direct sunlight entered the room when our laboratory was located at Tanashi, Tokyo. After our move to Tsukuba, Ibaraki prefecture, a 6-12 hr drive from the fishing site near Jogashima Island, the aquarium system was roofed to protect it from the rain but direct sunlight reached the system, resulting in the growth of greenish algae on the tank walls. At Tanashi and Tsukuba, a 40 W incandescent light about 1 m above the aquarium was kept lit day and night to help the squid see the tank walls. Maintenance conditions were the same as those in Matsumoto (1976). Sea water flowed at the rate of 20 1/min through the system, and sea water temperature was kept between 15° and 18°C by a temperature controller. When squid were trans- ferred to the aquarium from a transportation tank, aquarium temperature was adjusted to that in the tank. After transfer the temperature of sea water in the aquarium was lowered at less than 0.5°C hr to between 15° and 18°C. RESULTS Maintenance of unfed squid In November and December, 1978, we tried three runs of 2-week maintenance of squid (Ihiryteuthis blcekcri) by putting 15, 18, and 20 squid, respectively, into the aquarium without feeding them. The three runs resulted in no natural deaths, but eight squid were killed by cannibalism. No cannibalism was observed during first 3 days after the squid were transported from the fishing site. Survival was appreciably improved as compared with that in our previous experiment (see the dashed line in Fig. 3). The system and maintenence conditions had not been changed, except for the replacement of the sand by 20 kg zeolite in filter 1 and the addition of oyster shell. The shell alone did not appreciably SQUID IN A CLOSED-SYSTEM TANK 321 FIGURE 2. Tank walls' painted net patterns helped squid avoid collisions with the walls. improve survival. When the tickness of zeolite layer in filter 1 was reduced to the original 5 cm, squid survival became similar to that of our previous experiment. Thus, improved survival was a function of zeolite. To further test the effect of zeolite upon survival, we tried three runs of 15, 12, and 5 squid, respectively, with filter 1 containing 80 kg zeolite. Average survival was 4.5 days (thin solid line in Fig. 3). Thus, 80 kg zeolite may be harmful to the squid. Probably, there exists an optimum amount of zeolite for survival. In our system, 20-40 kg of zeolite appeared optimum among amounts tested : 5, 20, 40, and 80 kg. Maintenance of fed squid Longer maintenance of squid (Dorytcitthis blcekcri) by feeding them was tried February through April, 1979. Ten squid (one female, the rest male) were placed 322 G. MATSUMOTO AND J. SHIMADA in the aquarium at the end of February. Four to five red goldfish (Carassins anratus) 4—5 cm in length per squid were put into the tank for food every day after the third day from the beginning of the experiment. Feeding was conducted twice a day, at 9 a.m. and 5 p.m. At first the fish were not eaten, but after about a week, squid ate the fish within 30 sec after the fish were put into the aquarium. Figure 3 shows survival of 10 squid over a 2-month span. The survival times were 43-60 days. The first squid (female) died after spawning. Five squid that survived for over a month died shortly after their skins were scraped at the mantle tips. Once this area of skin was scraped, the squid rubbed their mantle tips more frequently on the tank walls. Another experience indicated that even a slight injury at the tip was fatal: When we stapled a small mark sheet on the squid to discriminate old from new squids, those stapled at their mantle tips died sooner. DISCUSSION Our previous report (Matsumoto, 1976) suggested that the cause of death of squid in a closed-system aquarium, even with enough oxygen dissolved in the sea wrater, was primarily the squids' lack of oxygen-adsorbing ability. Our present experiment further suggests that an obstacle to oxygen adsorption is organic sub- stances generated in the aquarium or a substance secondarily produced from these organic substances, which can be overcome by using zeolite in a filter. In main- taining squid (Illc.v illcccbrosns} for more than 30 days in an open-system tank of 15 m in diameter, O'Dor ct al. (1977) reported that sea water was pumped from the Northwest Arm of Halifax Harbor through intakes located at a depth of 15 m, 0.7 m off the bottom, and that the water quality was relatively high. 10 3 CT O> > c I - CT " **— - 0 o> \ \ \ \ \ *~^~ < — - Present result V i \0ur previous result \ \ \ \ — 3 0 0) 1 D n . n Survival percentc "n \ Result for \ the system \ with 80 kg \ zeolite in \ weight \ Maintenance days 10 FIGURE 3. Survival of 10 squid over 2 months. The dashed line shows an old record of squid survival over 8 days (Matsumoto, 1976). The fine solid line shows results obtained for the system in Figure 1 when filter 1 contained 80 kg of zeolite. Calibration of squid survival record (discontinuous solid line) is in numbers of live squid, while calibrations for both fine solid and dashed lines (continuous lines) are in percentage of squid surviving. SQUID IN A CLOSED-SYSTEM TANK 323 With a closed or open system for maintenance of squid in the aquarium, high quality sea water seems to be crucial for squid survival. In squid maintained for more than a month, skin damage was found to he fatal. In our maintenance system, the peripherally circulating flow of sea water, net patterns drawn on the tank walls, and continuous light enabled the squid to avoid hitting the walls. However, skin damage was caused by faint but repeated con- tacts with the tank walls, rather than by a harsh collision with the wall as suggested by Summers ct al. (1974). Flores ct al. reported that mortality after a 10-day survival test of squid (Todarodcs) in an open system was high with the autumn- captured squid, possibly because of spread of a skin infection. O'Dor ct al. (1977) concluded that deaths during the first week after specimens of Illc.v illecebrosus were put into the tank were associated with skin damage during capture. The reason why skin abrasion is fatal is not yet well understood. One explanation may be that skin abrasion provides an invasion site for bacteria, resulting in a disease leading to a quick death (Leibovitz ct al., 1977 ; Hulet ct a!., 1979). The condition of the squid when they were put into the tank obviously affected survival, as concluded by Summers ct al. (1974) and O'Dor ct al. (1977). In this respect, our squid-capturing method using Japanese squid jiggers (Matsumoto. 1976) is satisfactory. Initial condition also depends upon time between capture and arrival at a laboratory. In our case, shipboard transportation required 3-10 hr. and trucking from the seashore to our laboratories at Tanashi and Tsukuba took 3-5 and 6-12 hr, respectively. When truck transport took more than 8 hr, some squid were found to lie down on the bottom of the transportation tank. They usually recovered in the aquarium, but died within a week. Two-week survival tests showed the importance of feeding squid. We tried live red goldfish and frozen sardines as food, and found the goldfish more suitable. Small frozen sardines, 9-12 cm long, were purchased from a fish seller. One sardine per squid a day was defrosted, tied at the tail by white thread and hung in the aquarium until captured by squid. The fatty sardines made the sea water oily, so that we had to make the sea water overflow by adding reserved sea water to the aquarium. Goldfish (Carassius auratus) usually live in fresh water, and do not survive in sea water longer than 10 min. However, squid were able to catch the live goldfish within 30 sec after the fish were put in the aquarium. The live goldfish were cheap (about 4 cents each), and squid can swallow them, so that sea water is kept clean. Efforts to provide a steady supply of squid for neurosciences should now be directed to rearing from eggs in the aquarium (LaRoe, 1971 ; von Boltzky, 1971). SUMMARY Ten adult squid (Dorytcuthis blcckeri) were put into a small (1.37 nr in area) circular tank in a closed system. Half of the squid survived 43-60 days, the other half longer. Filtering ability was essential to squid maintenance, and filtering ability was satisfactorily supported by zeolite in appropriate amounts. In long-term (over a month) maintenance, skin damage caused by faint but repeated contacts with the tank walls became a major cause of death. Feeding squid was found to be important for squid maintenance, and live goldfish (Carassiits anratns) were found to be satisfactory as food for long-term maintenance of squid. 324 G. MATSUMOTO AND J. SHIMADA LITERATURE CITED ARNOLD, J. M., W. C. SUMMERS, D. L. GILBERT, R. S. MAXALIS, X. W. DA\V, AND R. J. LASEK, 1974. .-I t/uide to the laboratory use of the squid Loligo pealei. Marine Biological Laboratory, Woods Hole, Massachusetts. Pp. 9-17. FLORES, E. E. C., S. IGARASHI. AND T. MIKAMI, 1977. Studies on squid behavior in relation to fishing. II. On the survival of squid, Torarodcs pacificus Stecnstrup. in experi- mental aquarium. Bull. Fac. Fish. Hokkaido Univ., 28 : 137-142. HANLON, R. T., R. F. Hixox, AND W. H. HULET, 1978. Laboratory maintenance of wild- caught loliginid squids. Fisheries and Marine Service Technical Report, 833 : 20. 1-20.14. HULET, W. H., M. R. VILLOCH. R. F. HIXON, AND R. T. HANLON, 1979. Fin damage in captured and reared squids. Lab. Anim. Sci. 29 : 528-533. LEIBOVITZ, L., T. R. MEYERS, R. ELSTON, AND P. CHANEY, 1977. Xecrotic exfoliative dermatitis of captive squid (Lolitjo pealei). J. Invertebr. Pathol., 30: 369-376. LARGE, E. T., 1971. The culture and maintenance of the loliginid squids Sepioteuthis sepioidea and Dorytcuthis plei. Mar. Bio!.. 9: 9-25. MATSUMOTO, G. 1976. Transportation and maintenance of adult squid (Dor\tcuthis hleekeri) for physiological studies. Biol. Bull.. 150: 279-285. O'DoR, R. K., R. D. DURWARD, AND N. BALCH, 1977. Maintenance and maturation of squid. (Ilh-x illccehrosns) in a 15 meter circular pool. Biol. Bull.. 153: 322-335. SUMMERS, W. C., AND J. J. McMAHON, 1974. Studies on the maintenance of adult squid (Loligo (>ealei). I. Factorial survey. Biol. Bull.. 146: 279-290. SUMMERS, W. C., J. J. McMAHON, AND G. X. P. A. RUPPERT, 1974. Studies on the main- tenance of adult squid (Loliuo pealei). II. Empirical extensions. Biol. Bull., 146: 291-301. VON BOLETZKY, S., M. V. VON BOLETZKY, D. FROSCH, AND V. GOTZI. 1971. Laboratory rear- ing of Sepiolinae (Mollusca: Cephalopoda). Mar. Biol., 8: 82-87. Reference : Biol. Bull.. 159: 325-336 . (October, 1980) SODIUM TRANSPORT IN THE FRESHWATER ASIATIC CLAM CORBICULA FLU MI NBA SUSAN McCORKLE AND THOMAS H. D1KTX Department of Zoology and Physiology, Louisiana Stale University, Baton Rouge, LA 70803 Freshwater bivalves have a low blood-solute concentration that reduces the concentration gradients between their body fluids and the external medium and minimizes passive ion movements (Dietz and Branton, 1975; Potts, 1954; Prosser, 1973). Ion loss is still a problem due to the hyperosmotic blood and must be balanced by active ion uptake. Active Na and Cl transpoit systems have been reported in freshwater bivalves (Dietz, 1978, 1979; Krogh, 1939), the possible sites of Na transport being the mantle, kidney, and gill epithelia (Saintsing and Towle, 1978). Members of the molluscan superfamily Sphaerioidea have a higher rate of Na transport than do members of the superfamily Unionoidea. The sphaeriid species Corbicula fluminea ( = C. manilensis) and Musculium transversum ( = Sphaerium transversum] transport Na at rates up to twice those of the Unionoidea, approach- ing the rates of brackish water species (Dietz, 1979). Corbicula fluminea is considered to be a freshwater animal although it has been reported to inhabit brackish water up to 5%0 salinity (Filice, 1958; Hayashi, 1956). Geologically, the Corbiculidae are recent invaders of freshwater (Keen and Casey, 1969). Few physiological studies have been reported for C. fluminea (Dietz, 1979; Gainey, 1978a, b; Gainey and Greenberg, 1977). Because this species has been reported in both estuarine and freshwater habitats, the present study was under- taken to elucidate its mechanism of Na regulation. Sodium balance in C. fluminea was examined by partitioning Na flux into four processes: 1) active transport, in which ions are moved against electrochemical gradients, 2) passive diffusion, 3) exchange diffusion, an obligatory exchange of an ion with an identical ion located on the opposite side of a membrane, and 4) excretion. MATERIALS AND METHODS Specimens of Corbicula fluminea were collected from local bayous and main- tained unfed in an aerated aquarium containing artificial pondwater (0.5 mM NaCl, 0.4 mM CaCl2, 0.2 mM NaHCO3, 0.05 mM KC1). The clams were placed in a 12L:12D photoperiod at room temperature (20°-25°C) and were allowed to acclimate to laboratory conditions for at least 7 days before use. Animals under- going salt depletion were placed in deionized water that was changed frequently. Salt depletion for more than 3 months resulted in less than 5% mortality. Measurement of sodium flux Sodium influx was determined from calculations based on the slope of a line representing a decrease in 22Na radioactivity of the bathing medium as a function of time (Dietz, 1978, 1979; Dietz and Branton, 1975). All experiments were begun at approximately 1 1 a.m. to minimize the influence of physiological rhythms (McCorkle et al., 1979). Clams were placed in a de- 325 326 S. McCORKLE AND T. H. DIETZ ionized water hath for 1 hr prior to each study to remove adsorbed ions and waste products from the valves and mantle cavity. Clams were then moved to individual 50-ml beakers containing 30-40 ml of a bathing solution. The volume of the bath depended upon the type of experiment. Except where noted, the bathing solution was Na2SO4 to eliminate any Na/Cl transport interactions, as the SO4 ion is non-permeating (Dietz, 1978; Ehrenfeld, 1974). Within an hour the valves opened and the animals were siphoning the bath. Turbulence at the air-bath interface verified that the animals were in contact with and circulating the bathing medium. An initial bath sample was taken at 0 hr with a second sample taken 1-3 hr after the onset of the flux study. Sample volumes were not replaced. The time interval between samples allowed a 10-20% decrease in bath radioactivity. Bath volume and time interval were varied to prevent more than a 25rr decrease in the radioactivity of the bathing medium. After the second sample was taken, the animals were removed from the bath, drained of mantle-cavity water, blotted dry, and weighed for determination of total wet weight. The valves were opened and the soft tissues removed and oven- dried at 95°C for 24 hr for measurement of dry tissue weight, shell excluded. Sodium concentrations were measured using a flame photometer. A Triton X-100, toluene, p-terphenyl scintillation "cocktail" was added to an aliquot of each bath sample, and radioactivity was assayed by a liquid scintillation counter. Net Na flux (JnNa) was calculated from changes in the Na concentration of the bathing medium. The net flux is positive when there is a net absorption of the ion by the animal, and negative when there is a net loss of the ion to the external medium. Unidirectional Na influx (JiNa), ions taken up from the bath by the animal, was calculated from the rate of —Na disappearance from the bath. Sodium efflux (J,,Na), ions lost by the animal to the bath, was determined from the relationship J,, = Ji -- Jn- All flux rates are given as ^M Na/(g dry tissue -hr). Blood ion composition Blood samples were obtained from the clams by cardiac puncture (Fyhn and Costlow, 1975). The blood was centrifuged for 2 min at 8000 X g to remove the cells, and the supernatant used for ion analyses. Chloride concentrations were measured by electrometric titration. Total solutes were determined from undiluted blood using a freezing-point depression osmometer. Identification of exchange ions The pH of the bathing medium was measured at 0 and 3 hr during Na flux study to estimate the net H efflux (Jn11). The bath was buffered initially to pH 7.3 with 1 mM tris (hydroxy methyl) aminomethane to stabilize the pH. Bath samples were sonicated for 2 min to remove dissolved respiratory CO-j. A pH decrease was noted in each experiment. The 3-hr sample was titrated back to the original pH at 0 hr with 4.8 mN NaOH to determine JnH- The release of NH4 by the clams was assayed using the phenolhypochlorite method of ammonia determination (Solorzano, 1969). Five-mi bath samples were taken at 0 and 3 hr during an Na flux experiment. Due to tris interference in NH4 determinations, an unbuffered 0.5 mM Na2SO4 medium was used. Net NH4 flux (JnN"4) was calculated from the difference between the 3- and 0-hr NH4 concentrations. NA TRANSPORT IX A FRESHWATER MUSSEL Amiloride (0.5 and 1.0 mM 1) and ouabain (0.5 niM/1), both potential inhibitors of Na transport, were tested in the Na^SO, bath as aids to identification of the possible exchange ion(s) for Na. Amiloride is thought to act on the out- side surface of the epithelium, revcrsibly blocking the entry of Na (Ussing et a!., 1974). Inhibition of Na influx coupled with inhibition of H efflux has been demonstrated in crayfish, rainbow trout, frogs, and the unionid Carnnridina texasensis (Ehrenfeld, 1974; Kiischner et SO,, bathing medium. Data appear as mean ± SEM (** P < 0.01). Treatment N AiM Na/(g dry tissue -hr; Jn Ji Jo Pondwater Salt-depleted 32 35 -2.67 ± 0.86 9.79 ± 1.23** 7.90 ± 0.79 18.53 db 2.10** 10.57 ± 0.95 8.74 ± 1.27 Salt depletion caused a marked reduction in Na and Cl ions in the blood of C. fluminea (Table I). The total blood-solute level was reduced by 23% in salt-depleted clams, and blood Na and Cl concentrations dropped significantly. The combined decrease of Na and Cl in the blood accounted for 91%. of the de- crease in total blood solutes of salt-depleted clams. Sodium flux Salt depletion enhanced the rate of unidirectional Na uptake from an Na2S()4 bathing medium (Table II). The JiNa increased 234%, in salt-depleted clams compared to pondwater-acclimated animals (P < 0.01). The J0Na was not significantly different in pondwater and salt-depleted clams. Due to increased JiNa in the salt-depleted animals, JnNa was positive. Kinetics of sodium transport The rate of Na influx was dependent on the external Na concentration in pondwater-acclimated and salt-depleted C. fluminea. A non-linear relationship was found between J;Na and external Na concentration (Fig. 1). The transport system was saturable. The maximum rate of J;Na, Vmax, for pondwater clams was 12.90 ± 3.01 MM Na/(g dry tissue -hr) in a 0.87 mM Na/I bathing solution. The affinity of the transport system, Km, is the external Na concentration at which half of the maximum rate of Na influx occurs. The affinity of pondwater clams was 0.05 mM Na/1. Sodium influx of salt-depleted clams was variable, but the influx rates were always higher than the rates of pondwater animals in 0) "> 3O i/> -J w 20, • PW o SD O.I 0.5 1.0 1.5 2.0 No* concentration (mM/l) 3.0 FIGURE 1. Effect of Xa2SO4 concentration of the bathing medium on unidirectional sodium influx in Corbicula fluminea. Each point represents mean ± SEM for five animals either ac- climated to pondwater (PW) or salt-depleted (SD). 330 S. McCORKLE AND T. H. DIETZ TABLE III The effects of amiloride and ouabain on sodium transport in pondwater-acclimated specimens of Corbicula fluminea. Amiloride and ouabain were dissolved in 0.5 mM Na^SOt. Data are pre- sented as mean ± SEM (* P < 0.05). pM Na/(g dry tissue-hr) T M Jn J, Jo Control 16 -0.04 ± 1.49 7.26 ± 1.01 7.29 ± 1.40 Amiloride, 0.5 mM and 1.0 mM 11 4.71 ± 1.66* 9.26 ± 1.83 4.55 ± 1.23 Ouabain, 0.5 mM 11 -3.28 ± 2.62 7.09 ± 1.13 10.37 ± 2.49 all bath concentrations tested. The maximum transport rate for salt-depleted From the V Km was animals was 28.66 ± 2.17 //M Na/(g dry tissue-hr). estimated for salt-depleted clams as 0.04 mM Na/1. Effects of amiloride and ouabain on sodium transport Amiloride at concentrations of 0.5 and 1.0 mM/1 resulted in no significant changes in JiNa or J0Na and the data were pooled (Table III). Although there appeared to be a significant difference between control and amiloride-treated net fluxes, the difference may be an artifact. The fluxes from individual experi- ments using amiloride were not consistent. The treatment appeared to be stimulatory, inhibitory, or showed no effect on rates of Na flux at both concentra- tions. Ouabain (0.5 mM/1) added to the bathing solution had no effect on Na fluxes. Partitioning of the sodium flux The outward movement of Na to deionized water (Jt,Dif fusion _j_ JQ Excretion) was 2.87 ± 0.76/uM Na/(g dry tissue-hr) for pondwater-acclimated clams (Table IV). Rates of diffusive and excretory Na loss measured in Na2SO4 were assumed to be unchanged from the rates measured in deionized water. The J0Total for pondwater clams was 8.94 ± 0.76 ^M Na/(g dry tissue-hr). Therefore, 5.91 ± 0.80 ;uM Na/(g dry tissue-hr) was the estimated value for exchange diffusion. The outward diffusive/excretory component of Na movement decreased signif- icantly (P < 0.05) from 2.87 ± 0.76 MM Na/(g dry tissue-hr) to 0.87 ± 0.38 MM Na/(g dry tissue-hr) in salt-depleted animals. Total inward movement of Na, JiTotal, was 8.82±0.60^M Na/(gdry tissue-hr) for pondwater clams (Table V). The maximum passive diffusive influx was estimated with Ussing's flux ratio equation and was 0.50 /xM Na/(g dry tissue •hr) for pondwater-acclimated clams (Table V). The inward diffusion of Na doubled in salt-depleted clams. The measured potential difference across the epithelia was --7.0 ± 1.2 mV (N = 10), blood negative to the bathing medium. There was no significant difference between the TEPs of pondwater-acclimated and salt-depleted specimens of C. fluminea and the data were pooled. The active transport component of Na movement was estimated to be 2.41 ^M Na/(g dry tissue-hr) for pondwater-acclimated clams and increased fivefold in salt-depleted animals. Salt-depleted clams exhibited highly significant increases in rates of exchange diffusion, total Na influx, and total Na efflux (Tables IV and V). NA TRANSPORT IN A FRESHWATER MUSSEL 331 TABLE IV A partitioning of unidirectional sodium efflux in Corbicula fluminea. Total sodium efflux is the unidirectional efflux measured in a 0.5 mM Na^SOt bath. The diffusive /excretory component was measured in a deionized water bath. Exchange diffusion was calculated as /0Totul - (/oniifusion _^_ yoExcretion) Data appear as mean ± $EM (* P < 0.05, ** P < 0.01). f Means were calculated from individual animals and this accounts for the lack of agreement with the average Treatment jtM Na/(g dry tissue-hr) i Total Na T Exchange , / j Diffusion j r Excretion^ Jo JoDlffuslon VJo i Jo Pondwater 8.94 ± 0.76 (N = 6) 5.91 ± O.SOf (N = 6) 2.87 ±0.76 (N == 11) Salt-depleted 16.92 ± 0.84** (N = 7) 16.05 ± 0.67** (N = 7) 0.87 ± 0.38* (N == 14) Exchange ions Hydrogen and ammonium ions may be exchanged for Na in C. fluminea. Net ammonium flux was measured in a 0.5 mM Na2SO4 bath and a deionized water bath to determine the NH4 loss through active transport and the diffusive/ excretory NH4 loss. The value of JnNH4 in deionized water was not significantly different from the JnN"4 in Na2SO4, indicating that NH4 efflux in pondwater- acclimated clams was mostly passive and not involved in exchange for Na ions (Table VI). The JnNH< of salt-depleted animals in deionized water, 1.28 ± 0.18 ^M NH4/(g dry tissue-hr), was not significantly different from the JnNH4 of pond- water animals. However, the JnN"4 when salt-depleted animals were in a Na2SO4 bathing medium increased to 4.15 ± 0.18 pM NH4/(g dry tissue-hr), a highly significant rise in NH4 efflux. Net hydrogen ion flux, JnH, in pondwater animals was 5.12 ± 0.53 ^M H/(g dry tissue-hr) (Table VI). The JnH of salt-depleted animals, 7.62 ± 2.00 MM H/(g dry tissue-hr), was not significantly different from the J,,H of pondwater animals. The sum of JnH and JnN"4 in Na2SO4, 11.77 AiM/(g dry tissue-hr), balanced 83% of the Na actively transported into the salt-depleted animals (Fig. 2). A linear regression of J;Na and JnH+NH4 for pondwater and salt- TABLF, V A partitioning of the unidirectional influx of sodium in Corbicula fluminea. Total sodium influx is the unidirectional influx measured in a 0.5 mM Na-iSO* bath. Inward diffusion of sodium was calculated from the flux ratio equation. Active transport was calculated as JiTotal — (7; /.Diffusion) Data appear as mean ± SEM (** P < 0.01). Na/(g dry tissue-hr) i reaimem T Total Xa T.^£ha,nge 4- T Diffusion _i_ I Active Ji JiDlffuslon Ji J iTransport Pondwater 8.82 ± 0.60 5.91 ± 0.80 0.50 2.41 (N = 6) (N = 6) Salt-depleted 31.42 ± 1.39** (N = 7) 16.05 ± 0.67** (N = 7) 1.17 14.20 332 S. McCORKLE AND T. H. DIETZ TABLE VI A comparison of the net NH* flux in 0.5 mM Na^SOt and deionized water baths and net H flux in 0.5 mM Na-iSOt of pondwater-acclimated and salt-depleted Corbicula fluminea. Data appear as mean =t SEM (** P < 0.01). Means that are not significantly different are indicated by N.S. Treatment dry tissue-hr) in Na2SO4 JnNH« in deionized H2() T H Jn in Na2SO< Pondwater Salt-depleted 1.17 ± 0.22 (N = 6) 1.77 ± 0.18 (N = 14) N.S. 4.15 ± 0.18** (N = 7) 1.28 ± O.li (N = 14) 5.12 ± 0.53 (N = 10) 7.62 ± 2.00 (N = 7) depleted animals resulted in the significant regression coefficient value r = 0.66 (P < 0.02). DISCUSSION Corbicula fluminea exhibits an Na transport mechanism different from that of unionid mussels. The high J;Na and J0Na for C. fluminea agree with those pre- the unionid X z + X 25 -, 20 - I 5 - 10 - 5 - • PW O SD 5 10 , No 15 20 25 ( pM / g dry tissue • hr ) FIGURE 2. Relationship between unidirectional sodium influx and the sum of the net loss of H and NH4 of pondwater-acclimated (PW) and salt-depleted (SD) Corbicula fluminea in 0.5 mM Na2S(>4. The broken line represents the baseline excretory H + NH4 efflux for pondwater- acclimated clams. The equation for the regression line is JnH+NH4 = (1.48 ± 0.80) + (0.65 ± 0.60) JiNa. The regression coefficient, r, is equal to 0.66 (P < 0.02). NA TRANSPORT IN A FRESHWATER MUSSEL 333 viously reported by Dietz (1979). Sodium influx of C. fluminea is approximately six times the rate of the unionids Carunculina texnsensis and Ligumia subrostrata and twice that of Margaritifera hembeli (Dietz, 1978, 1979). Sodium efflux is an order of magnitude higher than that of any of the unionid mussels studied by Dietz (1979). Corbie ula Jinminea also has a higher Na uptake rate relative t<> other freshwater invertebrates examined: 1.28 juM Na/(g wet tissue-hr) as com- pared to 0.29 p.M Na/(g wet tissue-hr) in (he gastropod Limnaea stagnalis, 0.65 ^M Na/(g wet tissue-hr) in the crayfish Astacus pallipes and 0.70 /iM Na/(g wet tissue-hr) (at 18°C) for the earthworm Lnmbricus terrestris ((ireenaway, 1970; Shaw, 1959; Dietz and Alvarado, 1970). The major difference between the fluxes of C. Jinminea and other freshwater animals is the presence of a substantial exchange diffusion component. Sixty- seven percent of the total inward Na movement is due to exchange diffusion. Only 22% of the JiNn in the crayfish and minimal amounts of the JjNa in Carun- culina texasensis are attributable to exchange diffusion (Shaw, 1959; Dietz, 1978). In addition to the large percentage of exchange diffusion, the rate of active transport of Na into pondwater-acclimated specimens of C. fluminea, 2.41 juM Na/(g dry tissue-hr), is almost double the approximate value of 1.3 /iM Na/ (g dry tissue-hr) recorded for the unionids (Dietz, 1979). Absence of the active transport component could place the animals in a negative ion balance, depending on the rate of Na efflux relative to the inward Na movement. The magnitude of active Na transport depends upon the degree of salt depletion. Table II indicates transport rates for animals salt-depleted for an average of 40 days. Although the fluxes were not partitioned for these studies, the active transport component is at least equal to the net Na uptake. The clams used for the partitioning studies were salt-depleted for a longer period (about 90 days) and display a further elevation of active Na transport (Table V). For all studies, Na-Na exchange diffusion is about half of the total Na influx. Exchange diffusion is characteristic of corbiculid Na transport. Non-equilibrium conditions may be inferred by comparing observed TEP with TEP estimated for animals in a steady-state condition. In a steady state Ussing's flux ratio equation simplifies to the Nernst equation : E = (RT/F)ln (Ci/Co), giving the TEP (E) necessary to maintain Na in electrochemical equilibrium. If there is no active transport, observed TEP would equal the TEP estimated by the Nernst equation. The estimated TEP of -74 mV does not equal the ob- served TEP of -7 mV, suggesting active Na transport in C. fluminea. Previous studies on other freshwater clams have indicated that TEP is a calcium-diffusion potential independent of Na, Cl, or SO4 (Dietz and Branton, 1975). Transport of Na in C. fluminea is efficient. The affinity of the transport system for Na ions in these animals is greater than the affinity in the unionid mussels. Even with twice the active transport rate the affinity, Km, in C. fluminea is 0.05 mM Na/1 as compared to the Km of 0.15 mM Na/1 in Carunculina texasensis (Dietz, 1978). The Km in C. fluminea is low relative to literature K,,, values of 0.2-0.7 mM Na/1 in other freshwater animals (Dietz and Alvarado, 1970, 1974;Greenaway, 1970 ; Maetz, 1973; Prosser, 1973; Shaw, 1959). Sodium influx increases two to three times in salt-depleted animals without a significant change in Kni, suggesting that more epithelial transport sites are activated during the salt-depletion treatment. Passive diffusion and/or excretion of Na out of the salt-depleted animal is reduced (Table IV). 334 S. McCORKLE AND T. H. DIETZ The active transport systems in Carunculina texasensis and several other freshwater animals are primarily Na/H exchanges with minimal involvement of NH4 as an exchange ion (Dietz, 1978, 1979; Ehrenfeld, 1974; Kerstetter et al., 1970; Maetz, 1973; Maetz et al., 1976). Corbicula fluminea acclimated to pond- water does not differ from other freshwater animals in this respect. The JnNH4 of pondwater-acclimated clams appears to be primarily excretory since no change in NH4 output was noted between Na2SO4 and deionized-water bathing solutions. Although the JnH was not measured in a deionized water bath, the excretory J0H may be estimated. Assuming that Na/H exchange occurs as a 1:1 ratio, active Na transport may be subtracted from total JnH, leaving 2.71 /zM H/(g dry tissue -hr) as the excretory output. Na/H exchange has been reported to occur on a 1:1 basis in two species of amphibia and in crayfish (Garcia Romeu et a/., 1969; Garcia Romeu and Ehrenfeld, 1975; Ehrenfeld, 1974). The baseline ex- cretory H -f- NH4 efflux for pondwater animals may then be approximated as 4.12MM/(gdry tissue -hr) (Fig. 2). Salt-depleted and pondwater-acclimated specimens of C. fluminea show essentially the same JnN"4 in deionized water; however, the JnNH4 for salt- depleted clams in Na2SO4 quadrupled. Although Na/H exchange appears to be the primary exchange mechanism under normal conditions (Maetz and Garcia Romeu, 1964; Maetz et al., 1976), C. fluminea may be activating a Na/NH4 exchange as an auxiliary mechanism under stress of salt depletion. The active transport exchange ratio of Na : (H + NH4) was 2.2:1 in salt-depleted clams. The lack of stoichiometry suggests there may be a change in epithelial Na perme- ability when animals are in Na2SO4 solutions. The calculated exchange diffusion in Na2SO4 may therefore by underestimated and active Na transport over- estimated. Salt depletion is often used to stimulate active Na uptake in freshwater animals, but important changes in the animals should be recognized. Salt depletion caused a significant reduction in dry tissue weight of C. fluminea. Concurrently, the percent body water increased. The total blood-solute level was reduced, mostly by highly significant reductions in Na and Cl concentrations. These animals have undergone salt-depletion in the laboratory for long periods of time without mortality, but part of the elevated Na influx may be due to the 20% loss in dry tissue weight. The small tissue weight of this particular clam species makes the changes in weight and body water even more important. Significant loss of dry tissue weight is correlated with elevated NH4 efflux in salt- depleted clams. Tissue protein hydrolysis generates free amino acids, which may become major intracellular solutes contributing to water gain. Free amino acids are probably a major energy source under these laboratory conditions. Catabo- lism of amino acids would generate the additional H and NH4 necessary for Na exchange in the salt-depleted clams. Specimens of Corbicula fluminea show little sensitivity to amiloride and ouabain as inhibitors of Na transport. The apparent lack of inhibition in pondwater-acclimated animals may be a result of the small fraction of total Na turnover attributable to active Na transport. These data suggest a fundamental difference in the mechanism of Na-Na exchange diffusion and active Na trans- port, rather than exchange diffusion being an artifact of active transport. The auxiliary Na/NH4 exchange in salt-depleted animals may have been an important exchange mechanism when C. fluminea inhabited brackish water. Ammonia efflux against a gradient has been reported in the marine invertebrates Rangia cuneata, Nereis succinea, and Callinectes sapidus (Mangum et al., 1976, NA TRANSPORT IN A FRESHWATER MUSSEL 335 1978). The large exchange-diffusion component of C. fluminea might also be a vestige of its brackish water habitation. Exchange diffusion in estuarine blue crabs living in freshwater is reported to comprise 50% of Na influx (Cameron, 1978). Exchange diffusion in the unionids, long term inhabitant^ of freshwater (Haas, 1969), is negligible. Many of the differences in the Na balance mechanism of Corbicula fluminea as compared to other freshwater animals may be attribut- able to its recent brackish water heritage. We thank Deborah M. McMullan for typing the manuscript. This paper was submitted by Susan McCorkle to L.S.U. in partial fulfillment of the require- ments for the M.S. degree. This research was supported by NSF grants PCM75- 05483 AO1 and PCM 7 7-088 18. SUMMARY The Na transport mechanism was examined in pondwater-acclimated (PVV) and salt-depleted (SD) specimens of Corbicula fluminea. The Na influx in 0.5 mM Na2SO4of 7.90 ± 0.79/zM Na/(gdry tissue -hr), higher than most freshwater animals, increased to 18.53 ± 2.10 //M Na/(g dry tissue -hr) in SD animals. Saturation of the transport system is typical of Michaelis-Menten enzyme kinetics. Maximum influx of PW clams was 12.90 ± 3.01 /uM Na/(g dry tissue •hr), with a Km of 0.05 mM Na/1. The maximum rate in SD clams was 28.66 ±2.17 nM Na/(g dry tissue -hr), with little change in Km. Sodium movement in C. fluminea may be partitioned into passive diffusion, excretion, exchange diffusion and active transport. Exchange diffusion com- prises a substantial portion of Na movement: 5.91 ± 0.80 pM Na/(g dry tissue •hr) in PW animals and 16.05 ± 0.67 ^M Na/(g dry tissue -hr) in SD clams. Passive inward diffusion of Na was 0.50 /zM Na/(g dry tissue -hr) for P\Y clams and 1.17 ^M Na/(g dry tissue -hr) for SD clams. The primary exchange ion for Na is H, although a Na/NH4 exchange is functional in SD animals. In PW clams, 2.41 /uM H/(g dry tissue -hr) is trans- ported in a 1 : 1 exchange with Na. In SD clams, the net NH4 flux quadrupled contributing to a Na: (H + NH,) exchange. LITERATURE CITED CAMERON, J. N., 1978. NaCl balance in blue crabs, Callinectes sapidus, in fresh water. /. Comp. Physiol, 123: 127-135. DIETZ, T. H., 1978. Sodium transport in the freshwater mussel, Carunculina texasensis (Lea). Am. J. Physiol., 235: R35-R40. DIETZ, T. H., 1979. Uptake of sodium and chloride by freshwater mussels. Can. J. Zool. 57: 156-160. DIETZ, T. H., and R. H. Alvarado, 1970. Osmotic and ionic regulation in Lumbricus terrestris, L. Biol. Bull., 138: 247-261. DIETZ, T. H., and R. H. ALVARADO, 1974. Na+ and Cl~ transport across gill chamber epithelium of Rana catesbeiana tadpoles. Am. J. Physiol., 226: 764-770. DIETZ, T. H., and W. D. BRANTON, 1975. Ionic regulation in the freshwater mussel, Ligumia subrostrata (Say). J. Comp. Physiol., 104: 19-26. DIETZ, T. H., and W. D. BRANTON, 1979. Chloride transport in freshwater mussels. Physiol. Zool., 52: 520-528. EHRENFELD, J., 1974. Aspects of ionic transport mechanisms in the crayfish Astacus leptodactylus. J. Exp. Biol, 61 : 57-70. FILICE, F. P., 1958. Invertebrates from the estuarine portion of San Francisco Bay and some factors influencing their distribution. Wasmann J. Biol., 16(2): 159-211. FYHN, H. J., AND J. D. COSTLOW, 1975. Anaerobic sampling of body fluids in bivalve molluscs. Comp. Biochem. Physiol., 52A : 265-268. 336 S. McCORKLE AND T. H. DIETZ GAINEY, L. F., 1978a. The response of the Corbiculidae (Mollusca: Bivalvia) to osmotic stress: the organismal response. Physiol. Zoo!., 51 : 68-78. GAINEY, L. F., 1978b. The response of the Corbiculidae (Mollusca: Bivalvia) to osmotic stress: the cellular response. Physiol. Zoo/., 51: 79-91. GAINEY, L. F., AND M. J. GREENBERG, 1977. The physiological basis of the molluscan species abundance-salinity curve: a speculation. Marine Biol., 40: 41-49. GARCIA ROMEU, F., AND J. EHRENFELD, 1975. In vivo Na+ and Cl~ independent transport across the skin of Rana esculenta. Am. J. Physiol., 228: 839-843. GARCIA ROMEU, F., A. SALIBIAN, AND S. PEZZANI-HERNANDEZ, 1969. The nature of the in vivo sodium and chloride uptake mechanisms through the epithelium of the Chilean frog, Calyptocephalella gayi (Dum et Bibr., 1841). J. Gen. Physiol., 53: 816-835. GREENAWAY, P., 1970. Sodium regulation in the freshwater mollusc Limnaea stagnalis (L.) (Gastropoda: Pulmonata). /. Exp. Biol., 53: 147-163. HAAS, F., 1969. Family Unionidae. Pp. 411-463 in R. C. Moore, Ed., Treatise on Invertebrate Paleontology, Pt. N, Geological Society of America, Boulder, Colorado. HAYASHI, V., 1956. On the variation of Corhicida due to environmental factors. Venus, 19: 54-60. KEEN, M., AND R. CASEY, 1969. Family Corbiculidae. Pp. 665-669 in R. C. Moore, Ed., Treatise on Invertebrate Paleontology, Pt. X, Geological Society of America, Boulder, Colorado. KERSTETTER, T. H., L. B. KIRSCHNER, AND I). I). RAFUSE, 1970. On the mechanisms of sodium ion transport by the irrigated gills of rainbow trout (Sal mo gairdneri). J. Gen. Physiol., 56 : 342-359. KIRSCHNER, L. B., L. GREEN WALD, AND T. H. KERSTETTER, 1973. Effect of amiloride on sodium transport across body surfaces of freshwater animals. Am. J. Physiol., 224: 832-837. KROGH, A., 1939. Osmotic Regulation in Aquatic Animals. Cambridge University Press, London. MAETZ, J., 1973. Na+/NH4+, Na+/H+ exchanges and NH3 movement across the gill of Carassins auratus. J. Exp. Biol., 58: 255-275. MAETZ, J., AND F. GARCIA ROMEU, 1964. The mechanism of sodium and chloride uptake by the gills of a freshwater fish, Carassius auratus II. Evidence for NH4+/Na+ and HCOs~/Cl~ exchanges. J. Gen. Physiol., 47: 1209-1227. MAETZ, J., P. PAYAN, AND G. DERENZIS, 1976. Controversial aspects of ionic uptake in fresh- water animals. Pp. 77-92 in P. S. Davies, Ed., Perspectives in Experimental Biology, Vol. I, Pergamon Press, New York. MANGUM, C. P., J. A. DYKENS, R. P. HENRY, AND G. POLITES, 1978. The excretion of XH4+ and its ouabain sensitivity in aquatic annelids and molluscs. J. Exp. Zool., 203 : 151-157. MANGUM, C. P., S. U. SILVERTHORN, J. L. HARRIS, D. W. TOWLE, AND A. R. KRALL, 1976. The relationship between blood pH, ammonia excretion and adaptation to low salinity in the blue crab Callinectes sapidus. J. Exp. Zool., 195: 129-136. MCCORKLE, S., T. C. SHIRLEY, AND T. H. DIETZ, 1979. Rhythms of activity and oxygen con- sumption in the common pond clam, Ligumia subrostrata (Say). Can. J. Zool., 57: 1960-1964. POTTS, W. T. W., 1954. The inorganic composition of the blood of Afytilus edulis and Anodonta cygnea. J. Exp. Biol., 31 : 376-385. PROSSER, C. L., 1973. Comparative Animal Physiology. W. B. Saunders, Philadelphia. SAINTSING, D. G., AND D. W. TOWLE, 1978. Na+ -- K+ ATPase in the osmoregulating clam, Rangia cuneata. J. Exp. Zool., 206: 435-442. SHAW, J., 1959. The absorption of sodium ions by the crayfish, Astacus pallipes I ereboullet I. The effect of external and internal sodium concentrations. J. Exp. Biol., 36: 126-144. SOLORZANO, L., 1969. Determination of ammonia in natural waters by the phenolhypochlorite method. Limnol. Oceanogr., 14: 799-801. USSING, H. H., 1949. The distinction by means of tracers between active transport and diffusion. Ada Physiol. Scand., 19: 43-56. USSING, H. H., D. ERLIJ, AND U. LASSEN, 1974. Transport pathways in biological membranes. Ann. Rev. Physiol., 36: 17-49. Reference : Biol. Bull., 159: 337-348. (October, 1980) SPECIFICITY IN THE ASSOCIATION BETWEEN HYDRACTINIA ECHINATA AND SYMPATRIC SPECIES OF HERMIT CRABS1 NEIL A. MERCANDO AND CHARLES F. LYTLE Department of Biolouy. The Pennsylvania State University, Oyontz Campus, Ahiiiyton, Pennsylvania 19001, and Department of Zoology, North Carolina State University, Raleigh . North Carolina 27650 Symbiotic associations between hermit crabs and cnidarians, particularly sea anemones, have been widely investigated. From his extensive behavioral and physiological studies of many hermit crab-sea anemone associations, Ross (1960, 1974) identified varying degrees of specificity, ranging from intimate obligatory relationships, e.g., that between Pagnrus pridcau.vi and its "cloak anemone," Adamsia palliata ; to non-obligatory, possibly chance associations. Since the work of Schijfsma (1935, 1939), the association between hermit crabs and the colonial hydrozoan. Hydractinia cchinata, has received only limited attention. Specimens of Hydractinia cchinata form an encrusting, spinous, mat-like covering on their usual substrate, namely, gastropod shells. The occurrence of hydrozoans (e.g., Hydractinia} on hermit-crab-occupied shells influences the ecology and behavior of the host hermit crabs (Jensen, 1970; and Grant and Ulmer, 1974). In their studies of hermit crab populations from the Gulf of Mexico, Wright (1973), Conover (1976), Fotheringham (1976), and Mills (1976a) all discuss the differential occurrence of hydractiniids on shells occupied by P. longicarpus, P. pollicaris, and Clibanarius vittatns. Mills (1976a) states that smaller, common, sympatric hermit crab species (e.g., P. annulipes) "un- doubtedly utilize some of the same shells as young individuals of the other three [larger] species," and that "the extent of their participation in the asso- ciation has not been studied." Although Grant and Ulmer (1974) and Fothering- ham (1976) have suggested that the association plays an important role in partition- ing the gastropod shell resource, additional work is needed to substantiate and explain this. Hermit crabs exhibit distinct behavior patterns in selecting shells (Reese, 1962, 1963). Most studies involving selection of shells have focused on the importance of such shell characteristics as species, morphology, weight, volume, aperture size, or a combination of these factors (Reese, 1962, 1963; Markham, 1968; Childress, 1972; Kuris and Brody, 1976; and Conover, 1978). Various aspects of shell availability and utilization as well as intra- and interspecific competition for suit- able shells have been described by Orians and King (1964), Vance (1972a, b), Hazlett (1974), Kellogg (1976), Spight (1977) and Scully (1979). As recog- nized by Reese (1969), the shell functions as a microhabitat of the crab, and consequently, shell selection by hermit crabs should be regarded as a special form of habitat selection. The purposes of this study were 1) to investigate the degree of specificity be- tween species of hermit crabs and Hydractinia; 2) to determine the seasonality of 1 Paper No. 6574 of the Journal Series of the North Carolina Agricultural Research Service, Raleigh, N. C. 337 338 N. A. MERCANDO AND C. F. LYTLE the association; 3) to determine any pattern of specificity in shell-selection be- havior of the sympatric hermit crabs Pagurus annulipes. P. brevidact\lus, P. longi- carpn-s, and P. pollicaris; and 4) to consider the occurrence of Hydractinia as a factor in the crabs' partitioning of the gastropod shell resource. MATERIALS AND METHODS Four to six samples were obtained each month from November, 1971, to February, 1973, from an area of Bogue Sound southwest of the Morehead City, North Carolina, port turning basin. Using a 1-m scallop dredge covered with 1/4-in mesh netting, collections from the oyster shell substrate were made at depths varying from 3 to 5 m after dredging for 15 min at a constant speed. From the contents of each dredge haul, all gastropod shells were collected and placed in buckets of sea water for later identification and analysis. For each sample, the numbers of unoccupied, snail-occupied, and hermit-crab-occupied shells were recorded. The species of shell and hermit crab also were identified. All shells were examined with a dissecting microscope for the presence of Hydractinia polyps. Behavioral studies of three hermit crab species (Pagurus annulipcs, P. longi- carpus, and P. polUcarls) were conducted in the laboratory. All animals and shells were maintained in filtered sea water at 22°-24°C, 34— 3S(/f( salinity, and under a 16-hr light : 8-hr dark photoperiod. Before testing, all crabs were removed from their shells and held separately in plastic divider-boxes for at least 1 week. Crabs were evicted by removing the shell apex with wire cutters and inserting a thin piece of wire that was wiggled until the crab vacated the shell. Only crabs possessing all appendages and free of obvious external parasites and injuries were used. Crabs were fed shrimp pellets for 1-2 hr on alternate days. The compartments were then cleaned and refilled with sea water. Illynassa obsoleta shells, the most common shell species for all crab species, were maintained separately in sea water. The interior of each shell used was cleaned by repeated brushing with wire pipe cleaners and rinsing. The shell exteriors without Hydractinia colonies were cleaned by light brushing under flowing tap water. Only shells completely encrusted with a living mat of Hydrac- tinia were used as Hydractinia shells (HS), and only those without the hydroid and free of other fouling organisms were used as plain shells (PS). The Hydrac- tinia colonies were fed freshly hatched Artcuiia larvae for 1-2 hr on alternate days. Their compartments were then cleaned and refilled with sea water. Due to a shortage of appropriate Hydractinia-covered shells and the scarcity of the hermit crab P. brevidactylns, this species was not included in the laboratory studies. (While this report was in press we learned of a paper by McLaughlin, 1975, which indicates that the crab identified in our report as Pagurus brevidactylns has been described as a new species, Pagurus carolincnsis. ) Considerable evidence (Reese, 1962, 1963; Vance, 1972a; Kellogg, 1976) indicates that hermit crabs exhibit shell-size specificity. From his extensive studies on shell utilization by hermit crabs, including the species in this study, Kellogg (1971) has shown that the length of the anterior shield (ASL) of the hermit crab carapace is a reliable measure of crab size. The anterior-shield length (the dis- tance from the top of the rostrum to the midpoint of the cervical suture) was measured (± 0.01 mm) with a dissecting microscope fitted with an ocular microm- eter. Kellogg (1971) also indicated that shell width and shell weight were reliable measures of shell size. We froze specimens of P. annulipcs, P. longicarpus. HERMIT CRAE-HYDR.-ICTI.\f.l SPECIFICITY FIGURE 1. Shell width ( SW ) measurement of gastropod shell size. and P. pollicaris occupying /. obsoleta shells that were not fouled, damaged, or covered with Hydractinia. Upon thawing, the crabs were easily extracted intact, and ASL measurements were made. Shell widths and shell weights (wet) were obtained for each vacated shell. Between 14 and 30 observations were made for each of the three crab species. The relationships between ASL-shell width and ASL-shell weight were established by standard regression methods. For all three crab species, shell width accounted for more of the variation in crab size than did shell weight. Therefore, shell width (SW) was selected as the primary variate of interest and used in matching suitably sized shells to crabs to be tested. The equations describing the SW-ASL relationship for each crab species are : for P. annnlipcs, SW = 4.815 + 1.602 (ASL), for P. longicarpus, SW = 1.486 + 2.836 (ASL) ; and for P. pollicaris, SW == 5.812 + 1.545 (ASL). Shell width (SW) was defined as the distance between the juncture of the outer lip and first suture on one side of the shell and the point first contacted by calipers on the body whorl directly opposite (Fig. 1). This measurement was always made perpendicular to the longitudinal axis of the shell, i.e., from the shell apex to the aperture of the anterior siphonal canal. Shell width measure- ments were with vernier calipers graduated in 0.1 mm and read with the aid of a dissecting microscope. The shell-selection experiments were conducted in an illuminated test chamber where thirty 150-ml culture dishes each containing 100 ml of filtered sea water, were arranged in five rows of six dishes each. With the remainder of the room in darkness, the crabs could be observed through viewing ports with little risk of disturbing them. Into each of these dishes containing 100 ml of filtered sea water a naked (without a shell) hermit crab of known size was placed. Two crabs of each species were randomly selected and assigned within each row. Empty gastro- pod shells were added after 24 hrs acclimation. In Experiment I (HS z's. PS) 30 individuals of each crab species were tested for preference for Hydractinia shells (HS) or plain shells (PS). For each crab, a pair of /. obsoleta shells, one covered with a Hydractinia colony and one without, was selected according to the shell width (SW) appropriate for the ASL of the crab. Attempts were made to match the shells as closely as possible regarding SW (±0.05 mm), color, and condition, so that the only noticeable difference between the matched shells would be the presence or absence of a colony of Hydractina. Shells were matched to within ± 0.3 mm of the expected acceptable shell size for the crab to be tested. The paired shells were added simultaneously to each dish in a row within the test chamber. The shell occupied after 24 hr was considered the "preferred" shell. 340 N. A. MERCANDO AND C. F. LYTLE TABLE I Hermit crab abundance and the utilization of Hydrncimia-shells: PIS = shells with Hydractinia, and PS = shells u'ithout Hydractinia. * Based on the overall occurrence of Hydractinia-covered shells in the total hermit crab population. Hermit crab species ', of PS HS <% HS 9 X' P* total Pagurus longicarpus 36.2 700 288 29.1 48.8 <0.001 P. annul ipes 28.7 780 3 0.4 191.0 <0.001 P. pollicaris 18.6 257 251 49.4 268.4 <0.001 P. brevidactvlns 14.6 397 3 0.8 94.0 < 0.001 Paguristes humnii 1.2 30 21 Clibanarius vittatus Petrochirus diogenes 0.4 0.3 8 5 2 3f 13.7 1.4 >0.100 Paguristes tortugae <0.1 1 <>J Total 2178 552 20.2 In Experiment II (DS vs. PS), the same criteria for shell pairing and matching to crabs were employed as in Experiment I. However, shells covered with the perisarcal crust of a dead Hydractinia colony (DS) were substituted for shells covered with a living hydroid colony. No crabs used in Experiment I were retested in Experiment II. The observations on each species of crab were made in the same manner as described for Experiment I. To indicate the degree of "rejection" of shells not chosen in Experiments I and II, a second pair of experiments was conducted. Experiment III was con- ducted to determine if crabs that rejected HS shells in Experiment I would accept a Hydractinia shell when given no other shell to enter. Experiment IV tested whether crabs that rejected a perisarc shell (DS) in Experiment II would accept such a shell when given no other to enter. In these experiments crabs that chose a PS shell in Experiments I or II were evicted and returned naked to culture dishes containing new sea water. After a 24-hr acclimation period, the crabs from Experi- ment I were retested given only the original HS shell to enter. Similarly, in Experiment IV crabs that rejected a DS shell were given only the original DS shell to enter. RESULTS Utilization of gastropod shells by hermit crabs Eight species of hermit crabs occupied 46c/c of all gastropod shells collected from Bogue Sound. Of the remaining shells, 32% were unoccupied (i.e., did not contain a hermit crab or snail), and 22% housed a living snail. The four most common species of hermit crabs were Pagurus annulipcs, P. brcvidactylits, P. longicarpus, and P. pollicaris, which together comprised over 98% of the 2730 individuals obtained (Table I). Twenty-three species of gastropod shells were recorded from the total of 5885 shells collected in 78 dredge hauls, with 10 species comprising over 96% of the shells (Table II). Only six shell species, however, were inhabited by all four common crab species. Each of these shell species (Euplenra caitdata, Fascw- laria hunterii, I. obsolcta, Nassariits ribc.r, Terebra dislocata, and Urosalpinx HERMIT CRAE-HYDRACTINIA SPECIFICITY 341 TAHI.I-: II .Shell utilization and the occurrence of 1 lydractinia on common shell species occupied by her/nil crnl>. PS = shells without Hydractinia. IIS — shells with Hydractinia. /'. annulipes 1'. brevida tylux /'. longicarpus /'. [>olliraris PS IIS % HS PS IIS % HS PS HS •; us PS IIS '••; us Anachis spp. 36 (i 0.0 7 0 0.0 — — /tusvcon spp. — — 1 0 0.0 3 i) ii. i) 58 49 45.8 Eupleura caudata 42 0 0.0 16 0 0.0 9 1 10.0 3 0 0.0 1'ascialaria hunlerii 3 0 0.0 2 0 0.0 10 6 37.5 34 46 57.5 llvanassa obsoleta 337 1 0.3 241 1 0.4 483 223 31.6 87 71 44.9 Littorina irrorala — — — — — 18 10 35.7 3 6 66.7 Nassarius t'ibex 242 2 0.8 55 9 3.6 106 27 20.3 7 1 12.5 Polinices duplicatiis — — — 14 6 30.0 36 66 64.7 Terebra dislocata 31 0 ().() 6 0 0.0 6 1 14.3 0 1 100.0 Urosalpinx cinera 77 0 0.0 65 0 0.0 24 12 33.3 17 4 19.1 Other 12 0 0.0 4 0 0.0 21 2 8.7 12 7 36.8 Total 780 3 0.4 397 3 0.8 700 288 29.1 257 251 49.4 cinerea) also supported colonies of Hydractinia and, in total, constituted 86% of the shells occupied by hermit crabs. Even though many shell species were available, one species, / obsoleta, was most frequently utilized by all common hermit crab species in Bogue Sound (Table II). P. annulipes primarily inhabited /. obsoleta (43.2%) and Nassarius vibex (31.2%) shells, whereas P. brcz'idactylus mostly inhabited /. obsoleta (60.5%) and Urosalpinx cincrca (16.2%) shells. Although P. longicarpus in- habited the greatest variety of shell species, it was also primarily found in /. obsoleta shells (71.5%). The greatest number of P. pollicaris (3l.\% ) were collected in /. obsoleta shells. Larger members of this species, however, were frequently collected in two species of Busycon, Fasciolaria huntcrii, and Polinices duplicatiis shells. Occurrence of Hydractinia on hermit crab-occupied shells Nearly all of the Hydractinia-colonized shells (98%) were occupied by hermit crabs. The remaining colonized shells were found unoccupied. Chi-square con- tingency table analysis indicated significant differences in the incidence of Hydrac- tinia on shells occupied by different crab species. Since Hydractinia colonized 20% of all hermit-crab-occupied shells, it would be expected that approximately 20% of each crab species would occupy a Hydractinia shell (Table I) if the asso- ciation were independent of species. The percentages of Hydractinia shells occupied by P. annulipes (0.4%) and P. breridactylns (0.8%) were less than expected (P < 0.001) while the percentages of Hydractinia shells occupied by P. longicarpus (29%) and P. pollicaris (49%) were greater than expected (P < 0.001). In addition, the occurrence of the hydroid with P. annulipes was not significantly different (P < 0.50) from that with P. brcz'idactylus, while its occurrence with P. longicarpus and P. pollicaris was significantly different between these species as well as between these and the other two species (P < 0.001). If the occurrence of Hydractinia were related to shell species and not to hermit crab species, then the hydroid would be present only on certain shell species without regard to the species of crab inhabiting the shell. Therefore, we compared the occurrence of Hydractinia on the six shell species occupied by all four hermit crab species (Tables II, III). The hydroid was found on all six shell species; but more 342 N. A. MERCANDO AND C. F. LYTLE TABLE III Occurrence of Hydractinia on shells occupied by four species of hermit crabs from Bogne Sound: PS = shells u'ithout Hydractinia, and HS = shells with Hydractinia. * Based on the occurrence of HS shells occupied by the total hermit crab population as indicated in each section of the table. Hermit crabs PS HS 7c HS •7 X" P* Pooled results of the six shell species inhabited by all four common hermit crab species P. annulipes 732 3 0.4 146.9 <0.001 P. brevidactylus 385 3 0.8 74.3 <0.001 P. longicarpus 638 270 29.7 97.4 < 0.001 P. pollicaris 148 123 45.4 148.7 <0.001 Total 1903 399 17.3 467.4 <0.001 Same as above but excluding Ilyanassa obsoleta shells P. annulipes 395 2 0.5 49.8 <0.001 P. brevidactylus 144 2 1.4 15.6 <0.001 P. longicarpus 155 47 23.3 24.4 < 0.001 P. pollicaris 61 52 46.0 123.2 <0.001 Total 755 103 12.0 213.0 <0.0()1 Occurrence of Hydractinia on /. obsoleta shells only P. annulipes 387 1 0.3 82.6 <0.001 P. brevidactvliis 241 1 0.4 58.5 <0.001 P. longicarpus 483 223 31.6 57.7 <0.001 P. pollicaris 87 71 44.9 60.5 <0.001 Total 1148 296 20.5 264.9 < 0.001 frequently than expected (P < 0.001 ) on shells occupied by P. longicarpus and P. pollicaris and less frequently than expected (P < 0.001) on shells occupied by P. annulipes and P. brevidactylus (Table III). Also, Hydractinia occurred much more often on shells housing P. pollicaris and P. longicarpus (P < 0.001 ) than on shells occupied by other species. Its frequency on shells housing P. annulipes did not differ from its frequency on P. brevidactylus shells (P < 0.70). Since one shell species, /. obsoleta, comprised over 54% of all hermit-crab- occupied shells, one might suspect that the pattern of association between hermit crab species and Hydractinia could result directly from the abundance of I. obsoleta shells. To resolve this question, we analyzed the frequency of Hydractinia on those shell species commonly occupied by all four crab species, excluding /. obsoleta from the pooled totals. Here again, P. annulipes and P. brevidactylus inhabited Hydrac- tinia shells significantly less often (P < 0.001 ) than did P. longicarpus or P. pollicaris (Table III). Using contingency tables and Fisher's Exact Probability Test, an analysis of the occurrence of Hydractinia on /. obsoleta shells alone also showed the same pattern of hermit crab specificity (Table III). The frequency with which each hermit crab species inhabited Hydractinia shells was calculated for each monthly collection to determine seasonal differences. In all monthly collections, Hydractinia was rarely found on shells occupied by P. HERMIT CRAE-HYDR ACTINIA SPECIFICITY 343 10- 30- 10- o o 10- 10 - • ALL SHELLS OCCUPIED BY SPECIES n HYDRACTINIA-SHELLS OCCUPIED P, POLL. P. LONG P, BREV DJFMANJJASONDJF MONTHS FIGURE 2. Seasonal abundance of Hydractinia and species of hermit crabs from Bogue Sound, P. annul. = Pagurus annulipcs. P. brer. — P. brcz'idactylus, P. lony. = P. longicarpus, P. poll. = P. pollicaris. annnllpcs or P. brcz'idactylus but was frequently found on shells inhabited by P. longicarpus and P. pollicaris. Hermit crabs and Hydractinia shells were more abundant from November through February than from March through October (Fig. 2). From November, 1972, to February, 1973, numbers of hermit crabs increased markedly with P. longicarpus showing the greatest increase. Concurrent increases in Hydractinia shells occupied by P. longicarpus and P. pollicaris occurred during these months ; however, no appreciable increase was evident in Hydractinia shells occupied by P. annulipcs or P. brevidactylus. Shell-selection experiments Two types of experiments were conducted to determine whether each hermit crab species exhibited shell preference behavior relating to the presence of Hydrac- tinia or the perisarcal remains of a dead Hydractinia colony. Experiments I and II ("two-choice" experiments) showed that shell preferences exist, while experi- ments III and IV ("single-choice" experiments) indicated the degree of rejection of a hydroid-covered or perisarc-covered shell. Experiment I: Hydractinia shell (HS) vs. plain shell (PS). In this experi- ment P. annulipcs and P. longicarpus rejected HS shells more often than expected on the basis of chance and P. pollicaris accepted HS shells more often than expected 344 N. A. MERCANDO AND C. F. LYTLE TABLK IV Shell selection experiments. Two-choice experiments (Random shell selection expected) Single-choice experiments (100% shell selection expected) Hermit Crabs HS or DS PS X2 P HS or DS NS x- P I. Hydractinia shell (HS) vs. plain shell (PS) III. Hvdraftinia shell (HS) vs. no shell (XS) P. annulipes P. longicarpus P. pollicaris 2 8 23 28 22 7 22.5 6.5 8.5 <0.001 <0.05 <0.005 14 22 7 13 0 0 12.5 0 0 < 0.001 II. Perisarc shell (DS) vs. plain shell (PS) IV. Perisarc shell (DS) vs. no shell (NS) P. annulipes P. longicarpus P. pollicaris 8 9 13 12 1 1 6 0.8 0.2 2.6 >0.05 >0.05 >0.05 11 1 1 3 0 0 0 — — (Table IV). Only 2 of 30 P. annulipes (7%} chose Hydractinia-covered shells (HS) after 24 hr. According to the chi-square test, this result is significantly different from random (P < 0.001). Twenty-three of 30 (77%) P. pollicaris. on the other hand, chose HS shells (P < 0.005). Eight of 30 P. longicarpus chose HS shells. This result was significantly different from the expected random selection (P < 0.05). Experiment II: Perisarc shell (DS) Z'S. plain shell (PS}. Experiment II paired shells with only the spinous perisarcal crust remaining from a dead Hy- dractina colony (DS) with plain shells (PS). Random shell-selection behavior was exhibited by all three crab species tested in this experiment. Thus, the selection of DS shells as opposed to PS shells by P. annulipes, P. longicarpus, and P. pollicaris was not significant (chi-square, P > 0.05). Although only 40% of the P. annulipes selected DS shells as compared to 68% of the P. pollicaris, this difference was not significant (adjusted chi-square, P > 0.05). DS shell selection by P. longicarpus also did not differ from that by P. annulipes or P. pollicaris (P > 0.05). Experiment III: Hydractinia shell (HS) vs. no shell (NS). Those individuals rejecting HS shells in Experiment I were retested and exposed only to HS shells. After 24 hr in the test dishes, 13 of the 27 P. annulipes remained naked (i.e., without a shell), while lOO'/f of the P. longicarpus and P. pollicaris entered the Hydractinia shell (Table IV). Thus, 48% of the P. annulipes did not enter a HS shell, even though it was the only shell available. According to Fisher's Exact Probability Test, HS selection by P. annulipes differed significantly from P. longicarpus and P. pollicaris (P < 0.02). Experiment IV: Perisarc shells (DS) vs. no shell (NS). All crabs tested in this "single-choice" experiment entered the DS shells rather than remain without a shell (Table IV). Therefore, the DS shells did not elicit shell rejection behavior by P. annulipes as did the HS shells. Ten crabs used in shell selection experiments I and II were treated for 24 hr prior to testing in a sea water solution containing 25 ppb HgClo, to determine if sublethal levels of HgCl-j would alter shell selection behavior. The results indicated no significant difference in the selection of HS or DS shells by HERMIT CRAB-in'DKACTINI.-l SPECIFICI TV 345 crabs dosed with HgClL. and untreated controls. Fisher's Exact Probability Test was used to determine homogeneity of this data. Since the results for all three hermit crab species showed no significant difference between treated and control animals, these data were pooled. DISCUSSION The presence of Hydractinia on hermit-crab-occupied shells is clearly not random in Bogue Sound, N. C. The frequency of Hydractinia-colonized shells occupied by P. anindipcs and P. brcridactyliis was less than \c/c, while the fre- quencies of such shells housing /'. lonf/icarpns and /'. pollicaris were 29c/< and 4()r/c , respectively. This pattern of specificity was significantly different from the ex- pected 20%, the frequency of Hydractinia for the total hermit crab population. Field studies from different areas support the concept of non-random occur- rence of hydractiniid hydrozoans on hermit-crab-occupied gastropod shells. Grant and Ulmer (1974) showed that P. acadianus inhabits Hydractinia shells more frequently than P. pitbesccns in the Frenchman Bay area of Maine. Wright (1973) found that P. longicarpus and P. pol/icaris in Galveston Bay, Texas, often occur in shells supporting either Hydractinia or Podocoryne cornea, while Clibanarius rittatits was rarely observed in shells with hydroids. In other studies of Gulf Coast hermit crabs. Mills ( 1976a, b ) reported 7\c/t of specimens of P. policaris and 30(/( of P. longicarpus with either Hydractinia or Podocorvnc selcna, and Fotheringham (1976) observed the same two crab species utilizing hydroid-covered shells with a frequency of 62% and 30f/r respectively. Both of the latter authors found hydractiniid colonies absent or rare on shells occupied by C. vittatns. Neither reports any data on P. annulipes. We have rarely seen C. vittatus in a Hydractinia-covered shell along the North Carolina coast. It is well established that certain species of hermit crabs exhibit clear pref- erences for the shells of certain gastropod species (Reese. 1962, 1963; Orians and King, 1964; Markham, 1968; and Hazlett, 1971). It has also been suggested that Hydractinia frequently colonizes the same species of shells most often occupied by certain crab species (Crowell, 1945). Thus, we must ask whether hermit crabs are selecting for Hydractinia or for shell species. If the species of shell were the major factor influencing the association, then certain species of shells commonly occupied by each crab species, especially P. longicarpus and /'. pollicaris, should not be found supporting Hydractinia colonies. However, our data shows that those shell species commonly occupied by all four hermit crab species (except E. caitdata) frequently supported colonies of the hydroid when occupied by P. longicarpus and P. pollicaris. Conversely, the same shell species rarely sup- ported Hydractinia when occupied by P. annitlipcs or P. brevidactylus. Similar results by Grant and Ulmer (1974) indicate that the species of shell colonized had little apparent effect on the preference for Hydractinia shells exhibited by P. acadianus. Their work also suggests that the presence of Hydractinia must be an important factor in the selection of shells by certain crab species. Seasonal factors influence various aspects of the biology of hermit crabs (Reese, 1968; Samuelsen, 1970; and Fotheringham, 1975). The seasonal occurrence of hermit crabs in association with Hydractinia, however, has not previously been reported. In each of the monthly collections of this study, Hydractinia occurred most frequently on shells occupied by P. longicarpus or P. po/licaris and rarely on those occupied by P. annitlipcs or P. brci'idactylus. The former two crab species 346 N. A. MERCANDO AND C. F. LYTLE were most abundant from November through February, when the highest fre- quencies of Hydractinia shells were also found. Thus, the population densities of both P. longicarpus and P. poUicaris appear to have an important relationship with the abundance of the hydroid. Various investigators (Jensen, 1970; Grant and Ulmer, 1974; and Conover, 1976) suggest that shell-selection behavior is influenced by the presence of Hydractinia. This suggestion is supported by the results of our shell preference studies where P. annnlipcs frequently rejected Hydractinia shells even when hydroid-covered shells were the only ones available. On the other hand, P. longi- carpus and P. poUicaris readily accepted Hydractinia shells. The rejection of hydroid-covered shells even when no other shells were available was also reported by Wright (1973) for C. vittatus. A naked hermit crab that rejected a non- preferred shell in its natural environment would be more vulnerable to predation. The encrusting mat of perisarc produced by Hydractinia gives the shell a rough-textured surface. It might be hypothesized that shell selection or rejection might be due to the crust rather than the living hydroid. Our laboratory experi- ment showed, however, that each species of hermit crab selected shells without regard to the presence of the perisarcal crust. The living Hydractinia colony, therefore, and not the perisarcal crust, appears to provide the stimulus for the differential selection of Hydractinia shells by various species of hermit crabs. This conclusion is in accord with the findings of Jensen (1970), who showed that P. bcrnliardns exhibited a clear preference for shells covered with living colonies of Hydractinia. We suggest that shell selection behavior of the crabs plays the primary role in determining the pattern of association between symbiotic hydractiniid hydroids and various hermit crab species. Coexistence of sympatric species depends upon many factors. Vance (1972a) demonstrated that both resource and habitat partitioning are important mechanisms facilitating the coexistence of hermit crab species. The occurrence of Hydractinia on available gastropod shells may act as a resource partitioning mechanism, as suggested by Grant and Ulmer (1974) and Fotheringham (1976). Such a mechanism could reduce competition for shells among the sympatric species of hermit crabs from Bogue Sound, N. C. From his studies of seven sympatric populations of hermit crabs from nearby Reaufort Harbor, N. C., Kellogg (1977) concludes that co-existence of these populations depends upon habitat differences, shell-size partitioning, and shell- species partitioning, in that order. His earlier report (Kellogg, 1976) suggests that the restricted abundance of shells of adequate size and condition may be a factor limiting the size of hermit crab populations, particularly those of the largest species (P. poUicaris}. The three most abundant species in this area partition the shell re- source according to size, with P. annnlipcs commonly occupying the smallest shells, P. longicarpus the intermediate, and P. poUicaris the largest sizes (Kellogg, 1977). This, however, does not preclude size overlap. In our studies, individuals of each species ranging in ASL size from 1.9 to 2.9 mm were commonly collected. Although smaller shells are more abundant, so are the numbers of species and individuals com- peting for these shells. Our findings indicate that one species of shell (/. obsoleta} is most frequently occupied by each crab species. Shells of this species were most likely not produced in the subtidal habitat of Bogue Sound, since few live gastropods were collected in this area (Mercando, 1975). These factors could only increase competition for shells of small or intermediate size. Since P. poUicaris indicates strong preferences for Hydractinia shells while P. annnlipcs and P. brcvidactylus HERMIT CRAB-IIYDRACriNIA SPECIFICITY 347 do not, younger members of the former species effectively compete only with P. longicarpus for hydroid-covered shells. Therefore, in addition to shell-size selection, Hydractinia-she\] selection behavior may act to further partition tin- gastropod shell resource from Rogue Sound. The authors wish to express their gratitude to: 1) Drs. J. Costlow and \V. Kirby-Smith and the staff of the Duke University Marine Laboratory, Beaufort, North Carolina, for advice, assistance and the use of their facilities for most of the field studies; 2) Drs. G. \V. Thayer and M. Kjelson, and Mr. M. W. LaCroix and Mr. Nelson Johnson of the National Marine Fisheries Service, Atlantic Estuarine Fisheries Center, National Oceanic and Atmospheric Administration, Beaufort, North Carolina, for their assistance with some of the field collections ; and 3) Drs. R. L. Dixon and B. A. Fowler of the Pathologic Physiology Branch, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina, for the funds, space, and equipment utilized in various aspects of the laboratory phase of this study. Dr. T. Clemmer also provided assistance with statistical problems. We also thank Dr. Thomas G. Wolcott, North Carolina State University, for reviewing the manuscript. SUMMARY 1. A distinct pattern in the occurrence of Hydractinia on shells occupied by four sympatric species of hermit crabs was identified. Hydractinia occurred most frequently on shells occupied by Pagitms longicarpus and P. pollicaris and rarely on shells occupied by I', annulipes and P. brevidactylus. The abundance of Hydractinia fluctuated seasonally, concurrent with changes in the abundance of P. longicarpus and P. pollicaris. 2. Species of shell appeared to be a less significant factor than species of crab in determining the pattern of association between Hydractinia and hermit crabs. 3. In laboratory studies, naked P. annulipes rejected Hydractinia-covered shells even when offered only such shells to enter, while P. longicarpus and P. pollicaris readily accepted Hydra-ctinia-covered shells. None of the three species showed any preference for or against shells supporting only the perisarcal crust of dead Hydractinia colonies. 4. Our results indicate that occurrence of Hydractinia on gastropod shells acts to partition this resource and thereby tends to reduce competition among the four sympatric species of hermit crab in Bogue Sound, N. C. LITERATURE CITED CHILDRESS, J. R., 1972. Behavioral ecology and fitness theory in a tropical hermit crab. Ecology. 53: 960-964. CONOVER, M. R., 1976. The influence of some symbionts on the shell-selection behavior of the hermit crabs, Pagurus pollicaris and Pagurus longicarpus. Anim. Bchav.. 24: 191-194. CONOVER, M. R., 1978. The importance of various shell characteristics to the shell-selection behavior of hermit crabs. J. E.rp. Mar. Biol. Ecol.. 32 : 131-142. CROWELL, S., 1945. A comparison of shells utilized by Hydractinia and Podocorvnc. Ecologv, 26: 207. FOTHERINGHAM, N., 1975. Structure of seasonal migrations of the littoral hermit crab Clibanarius vittatus ( Bosc. ) . /. Exp. Mar. Biol. Ecol.. 18: 47-53. FOTHERINGHAM, N., 1976. Population consequences of shell utilization by hermit crabs. Ecology. 57: 570-578. GRANT, W. C., AND K. M. ULMER, 1974. Shell selection and aggressive behavior in two sympatric species of hermit crabs. Biol. Bull.. 146: 32-43. 348 N. A. MERCANDO AND C. F. LYTLE HAZLETT, B. A., 1971. Influence of rearing conditions on initial shell entering behavior of a hermit crab (Dccapoda. Paguridac). Crustaccana. 20: 168-170. HAZLETT, B. A., 1974. Field observations on interspecific agnostic behavior in laboratory reared hermit crabs. Bull. Mar. Set.. 15 : 616-633. JENSEN, K., 1970. The interaction between Payurus bernhardus ( L. ) and Hydractinia cchinata (Fleming). Ophelia. 8: 135-144. KELLOGG, C. W., 1971. The role of gastropod shells in determining the patterns of distribu- tion and abundance in hermit crabs. Ph.D. Thesis, Duke University, Durham, North Carolina. 210 pp. Diss. Abst. Order No. 72-11.101. KELLOGG, C. W., 1976. Gastropod shells: a potentially limiting resource for hermit crabs. /. Exp. Mar. Biol. Ecol.. 22: 101-111. KELLOGG, C. W., 1977. Coexistence in a hermit crab species ensemble. Biol. Bull., 153 : 133-144. KURIS, A. M., AND M. S. BKODY, 1976. Use of principal components analysis to describe the snail shell resource for hermit crabs. J '. Exp. Mar. Biol. Ecol., 22 : 69-77. MARKHAM, J. C., 1968. Notes on growth-patterns and shell-utilization of the hermit crab Payurus bernhardus (L. ). Ophelia. 5: 189-205. McLAL'GHLiN, PATSY A., 1975. On the identity of Put/urns brci'iductylus (Stimpson) ( Decapoda : Paguridae), with the description of a new species of Payurus from the Western Atlantic. Bull. Mar. Sci.. 23: 359-376. MERCANDO, N. A., 1975. Specificity in the symbiotic association between Hydractinia cchinata and four species of hermit crabs. Ph.D. Thesis, North Carolina State University at Raleigh, Raleigh, North Carolina. 79 pp. Diss. Ahst. Order No. 76-14,332. MILLS, C. E., 1976a. The association of hydractiniid hydroids and hermit crabs, with new- observations from North Florida. Pages 467-476 in G. O. Mackie, Ed., Coclcntcratc Ecoloyy and Behavior. Plenum Press, New York. MILLS, C. E., 1976b. Podocoryne sclcna. a new species of hydroid from the Gulf of Mexico, and a comparison with Hydractinia cclunata. Biol. Bull.. 151 : 214-222. ORIANS, G. H., AND C. E. KING, 1964. Shell selection and invasion rates of some Pacific hermit crabs. Pac. Sci.. 18 : 297-306. REESE, E. S., 1969. Behavioral adaptations of intertidal hermit crabs. Atn. Zoo/., 9: 343-355. REESE, E. S., 1963. The behavioral mechanisms underlying shell selection by hermit crabs. Behaviour, 21 : 78-126. REESE, E. S., 1968. Annual breeding seasons of three sympatric species of tropical hermit crabs, with a discussion of factors controlling breeding. /. Exp. Mar. Biol. Ecol., 2: 308-318. REESE. E. S., 1969. Behavioral adaptations of intertidal hermit crabs. Am. Zool., 9: 343-355. Ross, D. M., 1960. The association between the hermit crab Eupatiurus bernhardus (L.) and the sea anemone Calliactis parasitica (Couch). Proc. Zool. Soc. London, 134: 43-57. Ross, D. M., 1974. Evolutionary aspects of associations between crabs and sea anemones. Pages 111-125 in W. B. Vernberg, Ed., Symbiosis In The Sea. University of South Carolina Press, Columbia, S. C. SAMUELSEN, T. J., 1970. The biology of six species of Anomura ( Crustacea, Decapoda) from Raunefjorden, Western Norway. Sarsia. 45 : 25-52. SCHIJFSMA, K., 1935. Observations on Hydractinia cchinata ( Flem. ) and Eupayums bern- hardus (L.). Arch. Necrlandaiscs Zool., 1: 261-313. SCHIJFSMA, K., 1939. Preliminary notes on early stages in the growth of colonies of Hydrac- tinia cchinata (Flem.). Arch. Ncerlandaiscs Zool., 4: 93-102. SCULLY, E. P., 1979. The effects of gastropod shell availability and habitat characteristics on shell utilization by the intertidal hermit crab Payurus lonyicarpus Say. /. Exp. Mar. Biol. Ecol.. 37 : 129-152. SPIGHT, T. M., 1977. Availability and use of shells by intertidal hermit crabs. Biol. Bull., 152: 120-133. VANCE, R. R., 1972a. Competition and mechanism of coexistence in three sympatric species of intertidal hermit crabs. Ecoloyy, 53 : 1062-1074. VANCE, R. R., 1972b. The role of shell adequacy in behavioral interactions involving hermit crabs. Ecology. 53 : 1075-1083. WKIGHT, H. O., 1973. Effect of commensal hydroids on hermit crab competition in the littoral zone of Texas. Nature, 241 : 139-140. Reference : Biol. Bull., 159: 349-363. (October, 1980) EFFECTS OF MEDIA WITH LOW SILICIC ACID CONCENTRATIONS ON TOOTH FORMATION IN ACARTIA TONSA DANA (COPEPODA, CALANOIDA) CHARLES B. MILLER.' DAVID M. NELSON 1, ROBERT R. L. GUILLARD 2 AND BONNIE L. WOODWARD - 1 School of Oceanography, Oregon State University, Corrallis, Oregon 97331, and - Woods Hole 0 ceanographic Institution, U'oods Hole, Massachusetts 025-13 Many calanoid copepods, the dominant herbivores in marine pelagic habitats, have elaborate siliceous teeth set in sockets on the mandibnlar gnathobase (Bek- lemishev. 1954, 1959; Sullivan ct al., 1975; Vyshkvartseva, 1972). The existence of these siliceous teeth raises the possibility that silicon availability is important for copepod survival, growth, or at least tooth formation. Silicic acid is present in concentrations of 150-175 /iM at depth in the Pacific and Indian Oceans and of 40-50 fj.M in most of the Atlantic Ocean (Armstrong, 1965). However, surface layers of such huge oligotrophic regions as the Sargasso Sea (Menzel and Ryther, 1960) and Central Pacific gyre (SIO Ref. 67-5, 1967) become depleted seasonally to levels as low as 0.3 /xM. These levels inhibit diatom growth even in the presence of an excess of other nutrients (Paasche, 1973a; Guillard et al., 1973; Harrison ct al., 1976). Therefore, it is possible that other organisms with siliceous parts are inhibited by low silicic acid concentrations as well. If low silicic acid con- centration at levels that occur in nature directly inhibits copepod development, then its role in the interaction of phytoplankton and herbivorous zooplankton is complex and requires evaluation. We have conducted a first study of this possibility, employing the neritic cope- pod species Acartm tonsa Dana. This form is abundant in the waters around Woods Hole from early summer to early fall. It has three siliceous teeth : a curving spine on the ventral end of the tooth list (V), a heavy crown with four rounded points in the center (Ci), and a small bifurcate tooth just to right of center (Cs). The normal form of these teeth is shown in Figure 1. Acartia tonsa has been reared by Zillioux and Wilson (1966), Heinle (1969), and others. It tolerates a wide range of food mixtures, container sizes, salinities, and contaminants. We reared A. tonsa in media of low silicic acid concentration, trying to demon- strate the level at which development or tooth formation is inhibited. MATERIALS AND METHODS Our experimental sequence is shown in Table I. Low silicate media (LSM) were prepared from 1) Sargasso Sea surface water (1.56 /*M reactive silicate), and 2) various lots of Vineyard Sound, MA, surface water (1.26-1.56 //.M reactive silicate). Water was "stripped" of silicate by growing either Thalassiosira pseudonana (Hustedt) Hasle and Heimdal (clone 3H, method of Guillard ct al., 1973) or Phaeodactylum tricornutum Bohlin (clone Pet Pd, method of D'Elia et al., 1979) in 15- to 18-1 lots enriched with nitrate, phosphate, vitamins, and trace metals to f/2 level (fluillard and Ryther. 1962). Incubation was in poly- ethylene bags ('Cubitainers') under four Sylvania cool white fluorescent lamps on 349 350 MILLER, NELSON, GUILLARD, AND WOODWARD FIGURE 1. Mandibular gnathobase from adult female Acartia tonsa reared at 12.6 silicic acid (experiment L). The black bar represents 10 /j.m. Symbols: V = ventral tooth, Ci = large central tooth, Q = smaller central tooth, L = limit of siliceous caps, N = nonsiliceous dorsal spine. Scanning electron micrograph by Ann Cornell-Bell (A.C.-B.) using JEOL T20 SEM at 12,500 V accelerating potential. a 16 hr on: 8 hr off cycle. Temperature was held at 18°C. Increase in cell density of T. pscudonana was monitored daily by determining the in rivo fluo- rescence of the suspension with a Turner Designs fluorometer. At the end of exponential growth the cells were filtered from the water with 0.45 /*,m Millipore COPEPOD TOOTH FORMATION Experimental sequence for producing Imc silicate medium (A.SM/.i mid testing its effects on tooth formation in Acartia tonsa. Collect surface seawater of lowest possible silicate concentration Knrich to f/2 less Si(Oll), I (.row diatoms to stationary phase 1 Hlter out diatoms to obtain LSM ('.row flagellate food ( '.ro\v Acartia tonsa cultures cultures in LSM in LSM Evaluate survival, development rate and tooth formation HA filters using plastic filter holders and flasks. Vacuum was limited to 30 mm Hg. Silicic acid concentration in the LSM was determined as "reactive silicate" by the acid molybdate method (Strickland and Parsons, 1972). P. tricornittiini cells were removed when reactive silicate was substantially depleted. This occurred at a high cell density, though less than that at stationary phase. Stripping with P. trlcornntutn produced LSM with 0.08 /xM silicic acid. However, this medium was immediately toxic to all three of the food plants used in the copepod rearing. It produced cell lysis. Its pH was found to be 9.5, presumably because of a basic extracellular product of the algae. Vigorous bubbling for several hours with CO-, reducing the pH to 6.2, followed by bubbling with cotton filtered air raised the pH back to 7.5, removed this toxicity, and increased the silicic acid concentration only slightly. All the food plants and A. tonsa grew normally in the LSM treated in this manner. The toxicity is presumably due simply to the high pH, not to the specific base involved. No such effects were noted for LSM produced with T. pscudonana, which did not reach such high cell densities and presumably produced less extracellular material. Three species of flagellated phytoplankton were grown in LSM to serve as food for A. tonsa cultures. These were Isoclirysis galbana Parke (clone ISO), Diinaliclla tcrtiolccta Butcher (clone DUN), and Proroccntruni sp. (clone EXUV). Each 3-4 days new cultures of each species were started in 450 ml of LSM. Acartia tonsa was netted from Vineyard Sound on several occasions and placed in cultures by pipetting ten females and four males into 350 ml of water. A mix- ture of food was added to produce a final total of about 150,000 cells/ml in a volume of 500 ml. Experimental rearing in LSM was done with groups of eggs or very early nauplii produced over 2-3 days in these stock cultures. Some second generation young were used. Each culture used as a source of animals was poured through a 53 /zm mesh strainer, rinsed with LSM. and transferred to LSM in a plastic petri dish. Xauplii or eggs wetv counted under a dissecting microscope 352 MILLER, NELSON, GUILLARD, AND WOODWARD « "3 ^ -• a C: ^i tri *-" ^ I— o ;s _>, ~C M V V V 3? rt ~s 4— • 4<-) 4J rt n3 c^ ^ .S 'm _^ SEE c 5 § $ 3 O »- u u 000 - _, C C ^ -tx 2 ^ Cri -c O _c j= ° c ^ o .t; — .t: .t: K-> i! ^- is c S is if 1,^ in c in in b> is is rt nj rt **? '" •h ^ tn O tn tn s js ^ tf) "^ I— O -D O O i.' - -3. ' 2 2 c S 3 ^) !SJ t-l o -1° 1 ^ o <0 ^ '> ° ^ S IO IT5 \O ^J ' ^ *"*"""• 13 '> ^ o \O o ;_, ^i ^ 10 CX. 3 r-* ^ • o C/) "•s,"! Q. ,_, -t ON r^ 5 2 Cd "^ •-, ~ '-C C •2J _2 t '>£> & "a "| •*-» o 'c LH en *•*"" ^ C* ~ a rt rt ^ ~e S 3 (__^ fc d C^ M tl M C-l "I ~ o C Jvt^ C 00— JcXaX cWrWrW CJ^T • - "^ "0 Ov "* Cy ! - ^'S OO^o OOO OOCN •Is. •t-J "^ r^ T-JH ^H rr; i/-) rr^ T-^ -^- — f- ^> Vj *•"* W 2 2 % ^ 11 rt t^i (U C/) "O S « S a. td Ss -o 2 5 I? _« « in c * ~ -• — -- ^^ Q) .-.— < v>^ £ "^ a! •- £ ~ ••Co •• Oi C *-< -H o -^ fN CX nl *J% C^ 4? t/5 ^ c/5 ~ _ -^ rt _ -^ >, u '^ o •— c Oi-jou 3. U *O K uS £ -^ ,_, -f Ort4^ O3"^5 .§ g W 6 U U ^1 O £ 4J nj ^"^ a •s. W = 6 a e 0 <<< £££ uuu ^) Q I-M d e < COPEPOD TOOTH FORMATION J u! U) X> c c _0 .0 3 -C 'w '-^J be ^ O CJ 3 ® 3 3 < CO ;S ^' J= 0 ii U • *J l- "^ iZ S "^ u — *^ ^) """j ^ CO — ' CO .i| 2 CO to jjT Ul o 4— > 'J^ -*— ' w u « «Q Q 3 - C 3 § tn P4 J=U -C r- 1 2 •z$ ~l -z ~<3 O «| » £ 2^ •f-> '5 f '£ o ""o o o •| = s s s -2 " "2 "" - 1 ,c — 1 — 1 CO ^ 00 C-A~I oV 'r ^" •o "1 f T ^*^ *"^ •^ c/5 cj to i i 3 3 1 3 i — i f _ °° !§ > s ^ ^ OO ^-1 OO 3 . ^0 ~" t/5 i i u *» •» C 8 A 8 O fc U ^H r-i ^^ r^j rvj o i '£ ;; 3 O *3 be > r^; 10 OO ^-3 O r^- ^-— ' * 3 > ,— , ,— « ^H ^H jj ^^ ^^ u j *O 3 •^ Q *4 — ' 21 00 S H — 1 s ^. ,0 „ ^, _,. ^ _ 'c _a en "cO -C CO N 9, e, « U3 -^ E "^ ^ '^ ^ 6§ ^ 'c Ul ; a co — -J E '^ o — o X -+ O.e co C^ S CU 4. . ov O o3 ID 4-J ^H 'E _c •D ID -a < O •>» CO -a "23 Sll C 41 . x w u 4V "B _o •ti 3 r- o -C -C _cj; T3 -r-! C C ^ *-" i •"•*-" +•• X ^ n 4> O ^ r ^ CU 4> Q. u > j. ^ rt ^ "^^Q4_; cn co cnQwVcn tn -g'^-Cu-0 03 03 o3uU-,o3 03 o3 "JL v^^ oGjo3 tn cn EQ cu ,S£ 0) in _.S,o_5uu ^u S 1 __ 1 1 03 ^ ^^^!JS ^ g ^ ^ 5^^ CM S CO I I u -OOiDOO o 0 0 in 10 osOx^DOv 00 ^ -D ~ rC ^,r-Ov Ifl o > **" u 3 ^^rvj^H ^^OCr^JiD^in^Hlo^lfD ID OJ i^l ^^ OJ c CO o c Animals O* C*» Oi CM >^ "7 "7 ^7 ^7 ^7 b/D>Trbfi ^bCM '^blObO f-* f-< £-\ f-* £-* £-<£-* f~t f-+ £-* ^ o d Q.CO * o c -o CO * ta« ''Cl> '*'CL-^H ''C4* •'CXu'— i ^» ^ .2 "^ ^4^03. ^* ^.S-^^.S-?3- ~j o -ti g "ocajoj ^ "o •"— g "S +3 cu cu § g^2 g3^? § g^2gll§ d U U d U U 4> tJ ^ 6 Q. = — • (N W T iO ^ ^jj CO 1" irt jj-jjj P!^^s^ zzzzz 660 cCcCoT 6?aa COPEPOD TOOTH FORMATION 355 as they entered a small pipette and were placed in polymethyl pentene beakers of LSM. These young were fed LSM food. Total cell density was, again, about 150.000 cells/ml. Experiments were done in 500 and 130 ml volumes. One experiment was run in polycarbonate bottles. In some cases the LSM was changed after 1 week. In others development was completed in the initial medium. One set of experimental control animals was reared in LSM with addition of dissolved sodium metasilicate to approximately f/4 levels to test the direct effect of resupply- ing silicic acid. Another set was reared in unstripped Sargasso Sea or Vineyard Sound water to test the possibility of adverse effects of the stripping process on water quality. Reactive silicate in the water from all containers was measured at the end of the development period. RESULTS Table II lists the completed experiments and their results. Initial silicic acid concentrations in LSM were as low as 0.06 ^M. In all cases, concentrations increased substantially during the experimental period. The lowest level after rearing was complete was 0.18 /*M. There are two likely sources for this silicate: dissolution of bits of diatom silica which passed the filters used in producing the LSM, and dust. Since reactive silicate did not increase in stock containers of LSM, dust is the more probable source. Specimens of Acartia tonsa survived as well in LSM as in the controls, and developed at normal rates. Several control groups lagged behind the LSM groups in development. This is probably not attributable to treatment differences, but to use of different stocks in establishing the groups. There were small variations in age at the start, and since there were different parents, the groups presumably had somewhat different inherent development rates. This is particularly evident in comparison of the I replicates, which started the experimental period as eggs, with the J replicates, which were early nauplii at the start. Copepodites and adults of Acartia tonsa formed teeth at all of the silicic acid levels tested. However, typical specimens from low silicic-acid levels had teeth much lower in relief than those from high silicic-acid levels. This reduction was most evident in specimens from the 0.18 to 0.3 ju.M concentrations. Most teeth of specimens reared at levels of 0.8 /xM and above closely resembled the teeth shown in Figure 1, which were formed at 12.6 /xM. Reduction of teeth in specimens reared at low levels was variable both among individuals and among teeth on the same mandible. Various examples are shown in Figures 2, 3, and 4. There were exceptions (about 10-207^) in hoth directions: high teeth from specimens reared in the media lowest in silicic acid, and reduced teeth from specimens reared in intermediate levels. In a number of cases the ventral tooth was so reduced that in the scanning electron microscope there appeared to be a hole completely through the siliceous structure to the chitin or tissue underneath. The jaw shown in Figure 2 has a hole of this kind in the low hummock at the position of the ventral tooth. Higher mag- nification, to 15,000 diameters, did not resolve the character of these holes any- better (Fig. 2B). In fact, virtually no well-defined structure was apparent at any magnification anywhere on the surfaces of reduced teeth. The slight grooves apparent on the sides of all teeth were all that could be resolved better at high magnification (see Fig. 4B). 356 MILLER, NELSON, GUILLARD, AND WOODWARD FIGURE 2. A. (Left) Mandibular gnathohase from adult female A. tonsa reared at 0.22 silicic acid (experiment H:<). The black bar represents 10 /um. The reduced ventral tooth (3 on the reduction rating scale) is on the lower right. B. (Right) Detail of same jaw shows the character of the apparent hole in the ventral tooth. The black bar represents 2.5 /urn. SEM by A.C.-B. By scanning electron microscopy, it was difficult to determine the statistical frequency of reduction. Therefore, a rating scheme was devised using light microscopy of a large sample of the jaws by four observers who did not know the origin of the jaws. The rating scale was 1 for teeth like those of Figure 1 (essentially no reduction), 2 for ventral teeth like that in Figure 3 and central teeth like that in Figure 2 (substantial reduction), and 3 for ventral teeth like those of Figure 2 and central teeth like that in Figure 3 (tooth nearly missing). The sample series was 30 right jaws from female specimens, 10 reared at each of three silicic acid levels. The results are given in Table III. Substantial reduction was found to be primarily and usually present in the specimens from the lowest silicic acid levels tested, about 0.2 /xM. While different observers gave different nu- merical ratings, their agreement on the direction of the differences between speci- mens was excellent. Using mean ratings for each cell in the tables, combining the mean ratings of 2 and 3 (that is, combining all reduced teeth), and combining the mean ratings for the 0.8 and 10 //.M silicic acid levels produced 2x2 contingency tables with sufficiently large expectations for statistical testing. Com- bining categories after the fact in this way is not strictly legitimate; however, the resulting measure of contingency of tooth reduction on low silicic acid was significant for both teeth (XT > 10, P<0.01). Thus the differences between the specimen in Figure 1 and those in Figures 2, 3, and 4 are typical. Mandibles of stage V copepodites reared in medium of 0.2 /iM silicic acid concentration are shown in Figure 5. There is substantial reduction in one speci- men, but not in the other. Both the reduction and its variability occur in the younger stages as well as in adults. COPEPOD TOOTH FORMATION 357 FIGURE 3. Mandibular gnathobase from adult female Acartia tnnsa reared at 0.27 silicic acid (experiment Ki). The central tooth shows severe reduction (3 on the reduction rating scale). The black bar represents 5 /urn. The ventral tooth is at left. SEM by Alfred H. Soeldner (A.H.S.) using ISI Mini-SEM. DISCUSSION Acartia tonsa successfully extracted silicic acid from concentrations in seawater as low as 0.2 /xM and formed it into teeth. On the other hand those teeth were not normal. Formation of normal teeth required concentration of 0.8 /xM or above. Thus we must consider both 1 ) the ability of copepods to draw silicon for tooth formation from lower concentrations and 2 ) the possible ecologic significance of low silicic acid availability. Before we conclude that Acartia tonsa can draw enough silicon for tooth forma- tion from media with low concentrations of silicic acid, several other possibilities must be considered. It could be that moderate amounts of silicon are present in seawater and LSM which do not react in the acid-molybdate analysis. Silicic acid can form polymers that do not produce the molybdate complex (Alexander, 1953), but which could possibly be used by copepods for tooth formation. How- ever, Burton ct al. (1970) have done a careful study of this problem. They found "no detectable amount of unreactive silicon in any of the natural water samples analyzed. Polymeric silicon added to the [ sea j water samples was unstable and depolymerized completely within a few days." It seems unlikely, therefore, that copepods have any source of silicon in our experiments beside the low levels of reactive silicate which we measure. The result itself argues against another source. Normality of teeth improved with increasing reactive silicate. 358 MILLER, NELSON, GUILLARD, AND WOODWARD FIGURE 4. A. (Above) Mandibular gnathobase from adult female A. tonsa reared at 0.21 silicic acid (experiment Oi). Both teeth show substantial, but not severe reduction (2 on the reduction rating scale). The black bar represents 5 /nm. The ventral tooth is on the right. B. (Below) Detail of central tooth shows the proximal-to-distal grooves usually on the surfaces of the teeth. The black bar represents 2.5 MITI. SEM by A.H.S. Another possibility is that Acartia tonsa eggs might be supplied with sufficient silicon to support tooth formation throughout the life cycle. In that case full depletion of the supply would require several generations. Some calculations show this to be unlikely. Eggs of A. tonsa have a diameter of 70 /mi and a volume of 1.8 X 105 /mi3 = 1.8 X 10~10 liters. If silicic acid were stored in the egg at a concentration equivalent to seawater saturation (ca. 1600 ^M), then the content of an egg would be about 2.9 X 10~7 /mioles. Plasticene models of the teeth of A. tonsa based on SEM pictures like Figure 1 show the volume of the teeth to be approximately V = 7.2 X 10- /mi', Q ^ 1.7 X 103 /mi3, and C2 = 1.7 X 102 />im3. The total tooth volume on both jaws is 5.1 XlO3 /mi3. Applying the density of opal (2.1-2.3 gm cm"3) this is equivalent to 1.1 X 10~* gm. The molecular weight of solid, hydrated silica depends upon the amount of water included. For the formula SiO2 + 2H2O the molecular weight is 96 gm mole"1, and the amount of silica in the teeth of one adult would be about 1.1 X 10~4 COPEPOD TOOTH FORMATION TABU-: III Results of rating by each of four observers of right ja-cs from female Ararlia tonsa according to tin1 state of reduction of the ventral ( V) and central (Ci) teeth. Ruling was on a scale from 1 for no reduc- tion (Fig. 1) to 3 for severe reduction. A. Raw data. The four numbers in each cell are for the four observers. B. Mean numbers of ratings. C. Means combined into 2X2 table. A. Ventral tooth, Y Central tooth, Ct Rating Silicic acid levels — /iM Rating Silicic acid levels — /u.\I Low 0.2 Interni. 0.8 High 10-12 Low 0.2 Interni. 0.8 High 10-12 1 2 3 2|2 8|7 10|10 1 2 3 3|2 9|8 9|8 2 2 8 8 8|8 2|2 9|9 0|0 2|2 5 4 7|8 1|2 10|7 1 2 1 7 0|0 0 1 o|i 1 1 o|o 8 2 2|4 2 | 2 o|o 0 2 0|0 7 1 2|1 0 0 0|6 1|0 0|1 B. Ventral tooth, V Central tooth, Ci Low Interm. High Low Interm. High 1 2 3 Mean rating 2.00 6.00 2.00 7.75 1.25 1.00 9.50 0.50 0.00 1 2 3 2.25 4.75 3.00 8.00 1.75 0.25 8.50 1.25 0.25 2.00 1.33 1.05 2.08 1.23 1.18 C. Ventral tooth, V Central tooth, Ci Low Int. + High Low Int. + High 1 2 & 3 2.00 8.00 17.25 2.75 xr = 12.7 1 2 &3 2.25 7.75 16.50 3.50 xr == 1LO ^.moles. This is more than the maximum probable content of an egg by a factor of about 380. Additional smaller quantities would have to be present for each copepodite stage, since teeth are lost at each molt. Solid phase silica would have to be present within the egg for the entire life cycle's supply to be contained there. The weak teeth formed in our lowest reactive silicate levels also argue against this possibility. We are certain that Acartia tonsa can draw sufficient silicon for tooth formation from media with only slightly more than 0.2 //.M silicic acid. While we have applied a density appropriate for opal in this calculation, we have some evidence that for Caloniis the siliceous teeth are crystalline. Fragments of tooth have irregular arrays of very sharp spots as their electron diffraction pat- tern. This is typical of disrupted or somewhat irregular crystals, but not of amorphous opal. Further work on the mineral character of copepod teeth is in progress. 360 MILLER, NELSON, GUILLARD, AND WOODWARD FIGURE 5. A. (Above) Mandibular gnathobase from stage V copepodite of A. tonsa reared at 0.21 juM silicic acid (experiment L). Both V and Ci show extreme reduction. The black bar represents 5 fj.m. Ventral tooth is on left. B. (Below) Mandibular gnathobase from another stage V copepodite reared at 0.21 ^M silicic acid (also experiment L). Neither V nor C. shows much wear. The black bar represents 5 /j.m. SEM by A.H.S. A probable explanation for the reduced form of the teeth from LSM is that they have crumbled during use because of insufficient mineralization. The individuals from which all of the jaws in all of the figures came were 1-3 days past their terminal molt when they were preserved for examination. Their teeth had had some time to wear. Figure 5 shows a similar syndrome in stage V copep- odites. This stage lasts approximately 1 day at 18°C, and that is apparently sufficient time for substantial wear to occur (if wear is the mechanism of reduction). Alternately, animals reared in LSM may have failed to deposit teeth of normal shape. Silicic acid concentrations as low as 0.2 //.M occur very rarely in the oceans. Thus it seems unlikely that low silicic-acid availability would ever directly inhibit .•Icartia tonsa growth, development, or tooth formation in the field. The approxi- mately 1.5 /uM levels encountered in the waters over the continental shelf of the eastern United States that are the habitat of this animal are fully sufficient. A COPEPOD TOOTH FORMATION 361 simple observation extends this conclusion to other copepods of pelagic habitats comparably low in silicic acid. We examined a variety of calanoid copepods from surface water at a station in the central Pacific at 30° N, 143° W (samples collected from R/V Alpha Helix in November, 1971). The ambient concentra- tion of reactive silicate was 1.04 /*M. Siliceous teeth were found in representatives of all genera studied, except Candacia, which has a very reduced mandible. These genera were Calanus, Clausocalanus, Paracalanus, Scolecithrix, Euchaeta, Acartia, Centropagcs, Pachyptilus, and Pontella. It seems very likely that the capability for making siliceous teeth at silicate levels less than 2 /nM is quite general in cope- pods of oligotrophic habitats. We performed our experiments with a species obviously accustomed to low silicic-acid levels. Therefore, it may be that we chose a form least likely to demonstrate an inhibition. Perhaps Acartia clausi Giesbrecht, which is present in Vineyard Sound from fall to spring, and which does not experience severe silicic- acid depletion in its habitat, will prove more susceptible. Thus, it is still possible that silicic acid availability plays a role in the ecologic succession of copepod species. The flagellated forms used in our cultures surely present no severe challenge for mastication, even for individuals with the sorts of reduced teeth that develop at low silicic-acid concentration. However, extremely low silicic-acid levels are often found in the field just as a diatom bloom reaches a silicate-limited stationary phase (Dugdale and Goering, 1970; Schelske and Stoermer, 1971 ; Hafferty et al., 1978). It is possible that copepods undergoing their last maturation molts in such conditions would find themselves unfit for sustained feeding on the abundant diatom food all about them. This could reduce their ability to produce eggs, since most copepods must eat to reproduce, and thus cause a substantial delay in reduc- tion of the bloom. Documenting such an event in a real pelagic ecosystem would require that the investigator be on hand exactly when it occurred, something notably difficult for rare or transient phenomena of all kinds. It is also possible that sufficient silicic acid could be drawn from diatoms in the gut of maturing copeopds to form the teeth for the next phase. Silicic acid concentrations as low as 0.2 />iM, which impair tooth formation in Acartia tonsa, are also strongly limiting to silicic acid uptake, silica deposition, and cell division in most planktonic diatoms (Paasche, 1973a, 1973b ; Guillard ct al., 1973; Azam, 1974; Nelson et a!., 1970). With the single known exception of Phaeodactylum tricornutmn (Lewin ct al., 1958), the diatoms have an absolute silicon requirement for growth (Lewin, 1902). Many diatoms can take up silicic acid from concentrations of the order of 1 /iM or less. This is in contrast to the silicic-acid uptake capabilities of other algae. It has recently been discovered that many planktonic algae, including P. tricornutum and representatives of several major taxa other than diatoms, take up and deposit substantial amounts of silicic acid when it is available ( Fuhnnan ct al., 1978; Bankston ct al., 1979). Kinetic experiments with P. tricornutmn and Platynionas sp. (D. Nelson, unpublished data) show their uptake rates to be very low at silicic acid concentrations less than 30 juM. Their uptake systems thus have far less affinity for silicic acid than those of the silicon-requiring diatoms. We think that this difference in affinity for silicon implies a substantially greater expense in metabolic energy or molecular complexity for systems with high affinity. Only those organisms that strictly require silicon for life meet this expense. The comparability of the silicic acid affinity of Acartia tonsa to that of 362 MILLER, NELSON, GUILLARD, AND WOODWARD diatoms, in which it is an absolute necessity, thus argues that sound teeth are vitally important for survival in this copepod. The experiments described were performed by C. B. Miller during the course of a guest investigatorship at Woods Hole Oceanographic Institution in the labora- tories of R. R. L. Guillard and Peter H. Wiebe. The work was supported by a grant from the National Science Foundation (OCE-7825762) to Oregon State University and by X.S.F. grant ( )CE-78-08858 to R.R.L.G. Scanning electron microscopy was done by Ann Cornell-Bell at Marine Biological Laboratory, Woods Hole, and Alfred Soeldner at Oregon State University. We would like to thank Nathanial Corwin, Larry Brand, Steven Boyd. Nancy Marcus, Timothy Cowles, Everett Hogue, William Peterson, and Waldo Wakefield for help. This is con- tribution No. 4651 from the Woods Hole Oceanographic Institution. SUMMARY Acartia tonsa Dana can extract sufficient silicon for formation of siliceous teeth from media with concentrations of silicic acid as low as 0.2 /xM. Low silicate media were produced by growing diatoms in seawater collected from oligotrophic habitats and enriched with nutrients other than silicic acid, then removing them by filtra- tion when they reached a silicic acid-limited stationary phase. Most teeth formed by A. tonsa in 0.2 /tM medium were greatly reduced in profile, probably by in- creased effects of wear on insufficiently mineralized teeth. Teeth formed at silicic acid levels comparable to those in the field habitat of the copepod (1.5 /xM ) were normal. The extent of this ability to remove silicon from dilute media is comparable to that of diatoms. It is unlikely that low silicate levels in the field are an impor- tant limiting factor for A. tonsa or other copepods normally found in oligotrophic, pelagic habitats in the sea. LITERATURE CITED ALEXANDER, G. B., 1953. The reaction of low molecular weight silicic acid. J. Atncr. Chon. Soc., 75 : 5655-5657. ARMSTRONG, F. A. J., 1965. Silicon. Pages 409-432 in J. P. Riley and G. Skirrow, Eds., Chemical Oceanography. Vol. 1. Academic Press, New York. AZAM, F., 1974. Silicic acid uptake in diatoms studied with [^Ge] germanic acid as tracer. Planta (Berlin), 121: 205-212. BANKSTON, D. C., N. S. FISHER, R. R. L. GUILLARD, AND V. T. BOWEN, 1979. Application of dc plasma optical emission spectrometry to studies of element incorporation by marine phytoplankton. U.S.D.O.E. Environ. Q. Report, EML 356: 509-531. BEKLEMISHEV, K. V., 1954. The discovery of siliceous formations in the epidermis of lower Crustacea. Dokl Akad. Nauk SSSR, 97: 543-545. BEKLEMSHEV, K. V., 1959. On the anatomy of masticatory organs of copepoda. Report 2: the masticatory edge in mandibles of certain species of Calanidae and Eucalanidae. Tr. Inst. Okcanol., 30: 148-155. BURTON, J. D., T. M. LEATHERLAND, AND P. S. Liss, 1970. The reactivity of dissolved silicon in some natural waters. Limnol. Occanogr., 14 : 473-476. D'EuA, C. G., R. R. L. GUILLARD, AND D. M. NELSON, 1979. Growth and competition of the marine diatoms Phaeodactylum tricornutum and Thalassiosira pscudonana. I. Nutrient effects. Marine Biology, 50: 305-312. DUGDALE, R. C. AND J. J. GoERiNG, 1970. Nutrient limitation and the path of nitrogen in Peru Current production. Anton Bruun Report No. 4, Texas A & M Press, pp. 5.3-5.8. FUHRMAN, J. A., S. W. CHIOLM, AND R. R. L. GUILLARD, 1978. Marine alga Platymonas sp. accumulates silicon without apparent requirement. Nature, 272 : 244—246. COPEPOD TOOTH FORMATION 363 GUILLARD, R. R. L., AND J. H. RvTHER, 1962. Studies of marine planktonic diatoms. I. C\clotella nana Hustedt, and Dctonula confcrracca (Cleve) Gran. Can. J. Microhiol 8: 229-239. GUILLARD, R. R. L., P. KILHAM, AND T. A. JACKSON, 1973. Kinetics of silicon-limited growth in the marine diatom Thalassiosira pscndonana Hasle and Heimdal (Cvclotella nana Hustedt ). y. PhvcoL. 9 : 233-237. HAKFERTY, A. J., L. A. CODISPOTI. AND A. HUYER. 1978. JOINT-II R/V Melville legs I, II, and IV, R/V Iselin leg II bottle data March 1977-May 1977. WOE Coastal Upwcllinij Ecosystem Analysis Data Report 45. Univ. Wash., Seattle. 779 pp. HARRISON, P. J., H. L. CON\VAY, AND R. C. DUGDALE, 1976. Marine diatoms grown in chemostat under silicate or ammonium limitation. I. Cellular chemical composition and steady-state growth kinetics of Skeletonema costatmn. Marine Bioloq\, 35 : 177-186. HEINLE, D. R., 1969. Culture of calanoid copepods in synthetic seawater. J. Fish. Res. Bd. Can., 26: 150-153. LEWIN, J. D., 1962. Silification. Pp. 445-455 in R. A. Lewin, Ed., Physiology and Bio- chemistry of Aluae. Academic Press, New York. LEWIN, J. D., R. A. LEWIN, AND D. E. PHILPOTT, 1958. Observations on Phaeodactylum tn- cornutuin. J. Gen. Microbiol., 18: 418-426. MENZEL, D. W., AND J. H. RYTHER, 1960. The annual cycle of primary production in the Sargasso Sea off Bermuda. Dccp-Sca Res., 6: 351-367. NELSON, D. M., J. J. GOERING, S. S. KILHAM, AND R. R. L. GUILLARD, 1976. Kinetic of silicic acid uptake and rates of silica dissolution in the marine diatom Thalassiosira pscudo- nana. J. Phycol.. 12 : 246-252. PAASCHE, E., 1973a. Silicon and the ecology of marine plankton diatoms. I. Thalassiosira pscudonana (Cyclotclla nana) grown in a chemostat with silicate as the limiting nutrient. Marine Biology, 19: 117-126. PAASCHE, E., 1973b. Silicon and the ecology of marine planktonic diatoms. II. Silicate- uptake kinetics in five diatom species. Marine Biology, 19 : 262-269. SCHELSKE, C. L., AND E. F. STOERMER, 1971. Eutrophication, silica depletion and predicted changes in algal quality in Lake Michigan. Science, 173 : 423-424. SCRIPPS INST. OCEANOGR., 1967. Data report: physical, chemical and biological data Ursa Major Expedition. SIO Ref. No. 67-5. 43 pp. STRICKLAND, J. D. H., AND T. R. PARSONS, 1972. A practical handbook of seawater analysis, rev. edn. Bull. Fish. Res. Bd. Can., 167: 1-311. SULLIVAN, B. K., C. B. MILLER, W. T. PETERSON, AND A. H. SOELDNER, 1975. A scanning electron microscope study of the mandibular morphology of boreal copepods. Marine Biology, 30: 175-182. VYSHKVARTSEVA, N. V., 1972. Structure of the mandibles in the genus Calanus s.l. in relation to latitudinal zonality. Pp. 186-199 in B. E. Bykhovskii and Zh. A. Zvereva, Eds., Geographical and Seasonal Variability of Marine Plankton. [Translation ISBN-O- 7065-1494-7.] ZILLIOUX, E. J., AND D. F. WILSON, 1966. Culture of a planktonic calanoid copepod through multiple generations. Science, 151 : 996-998. Reference: Biol. Bull.. 159: 364-375. (October, 1980) ION AND WATER BALANCE OF THE HYPO- AND HYPEROSMOTICALLY STRESSED CHITON MOPALIA MUSCOSA WILLIAM M. MORAN ' AND RICHARD E. TULLIS Moss Landing Marine Laboratories, California State Universities, Moss Landing, California; and Dcpt. of Biological Sciences. California State University Hayivard, Hayzvard, California 94542 Molluscs inhabiting rocky intertidal zones and estuaries often experience salinity stress, as freshwater run-off after heavy rain (Boyle, 1969) or tidal fluctuations of salinity (Stickle and Ahokas, 1975). A rapid reduction of external salinity causes these organisms to gain water osmotically, with the result that physiological functions such as respiration, feeding, and locomotion may be severely impaired (Oglesby, 1975). The water taken up can be considered to enter both the extra- cellular space (in the case of animals with open circulatory systems, the blood) and the intracellular space. Volume regulation of both is required for survival. Cell volume regulation has been extensively investigated in marine and estuarine invertebrates (Pierce and Greenberg, 1973). Control of cell volume occurs through changes in the concentration of intracellular free amino acids (FAA). For example, during hyposmotic stress these acids leave the cell intact followed by osmotically obligated water, and cell volume is restored (reviewed by Watts and Pierce, 1978). Much less is known about whole-animal volume regulation of molluscs. This aspect of volume regulation is complex. It may simultaneously involve salt and water movements between the animal's body fluids and the medium, shifts of free amino acids between cells and extracellular fluids (cell volume regulation), change of osmotic permeability, and change in excretion rates (Fletcher, 1974b). In addi- tion, these mechanisms may contribute to volume control at different times after exposure of the organism to hyposmotic sea water. Most work on volume regulation of molluscs has been performed with bivalves and gastropods. Both groups have an encompassing shell which makes accurate wet-weight determinations difficult. The polyplacophoran molluscs, the chitons, do not have an encompassing shell. Instead, they have eight serially arranged plates embedded in the girdle on their dorsal surface. In addition, chitons readily attach themselves to petri dishes, facilitating weighing of the animal. The chiton's wet weight is determined by subtracting the weight of the petri dish. The chiton Mopalia muscosa inhabits the rocky intertidal zone of the Pacific coasts of the United States and Canada. The purpose of this study was to characterize whole-animal volume regulation in M. muscosa. We also report the blood-ion concentrations and muscle-tissue water content of chitons acclimated to different salinities. 1 Present address : Department of Zoology, University of Maryland, College Park, MD 20742. 364 CHITON ION AND WATER BALANCE 355 MATERIALS AND METHODS Collection of animals and maintenance conditions Specimens of Mopalia miiscosa were collected on the south jetty of Elkhorn Slough, Moss Landing, California. All chitons were taken within the intertidal zone, and between —0.3 and +0.6 in of mean lower low water. Collections were restricted to this intertidal range to limit intraspecific differences resulting from different habitats. At the laboratory, the chitons were placed in a maintenance tank which held aerated normal sea water (33.3-34.5%^ salinity; pH = 7.7^8.1) at 13° C ±2°. All chitons were allowed 6-13 days to adjust to these conditions before experiments were performed. The chitons ranged in weight from 5 to 22 g. Source and preparation of experimental SW Sea water was obtained from the mouth of Elkhorn Slough at high tide, filtered with a Glass Fiber Type A filter, and approximately 11 g of an artificial sea-salt mixture (Instant Ocean) was added to a liter of filtered sea water (SW) in order to raise the salinity to 42.\'/lf (125% SW) ; lower salinities were made by diluting the 125% SW with deionized water. The salinity of all solutions was determined before use with an induction salinometer (± 0.01/£0). Acclimation of chitons to experimental SW Most experiments were conducted at 13.3°C; the pH varied between 8.0 and 8.3. The effect of salinity on blood-ion concentrations was examined by acclimating batches of three chitons to experimental media of 125, 100, 75, and 60% SW for 48 hr. Then a sample of hemolymph was removed from each damp-dried chiton by puncturing the body wall medially just beneath the eighth valve with a needle and 1 cc syringe. The pericardium is near the region where hemolymph was obtained. However, pericardial fluid would be expected to be colorless when with- drawn, due to ultra-filtration of hemolymph through the heart wall. In each case blue fluid was obtained, indicating the presence of hemocyanin, and assuring that the fluid was blood and not pericardial fluid. The hemolymph collected from each animal was centrifuged to remove cells. The supernatants were used for ion-concentration determinations. The effect of hyposmotic SW on the kinetics of change of chiton blood Na+ and Cl" concentrations was investigated by placing chitons in 60% SW and then removing three chitons at 2, 4, 6, 8, 12, and 24 hr for hemolymph-ion analysis. Three chitons from the maintenance tank served as time 0 samples. Measurement of SW and blood-ion concentrations Sodium, K+, Caa+, and Mg2+ concentrations of hemolymph and SW were deter- mined by atomic-absorption spectrophotometry (Perkin-Elmer 305B). Sodium, Ca'->+, and Mg2+ concentrations were measured by diluting hemolymph and SW appropriately with 1000 ppm K/, in order to prevent ionization effects, in \% HNOs; the diluting solution for the Ca-+ and Mg-' determinations also contained \% La3+ to prevent phosphate interference with Ca-+ and Mg1!l determinations. The diluting solution for determination of K+ concentrations contained 1000 ppm Na+, again to prevent ionization effects, in \% HNO3. Standards were prepared with the appropriate diluting solution. Chloride concentrations were determined 366 W. M. MORAX AXD R. E. TULLIS with an Oxford Titrator by the mercuric titration method of Schales and Schales (1941). Measurement oj wet -weight changes as an indicator oj volume regulation Chitons were placed in the maintenance SW on preweighed plastic petri dishes and allowed 2-3 days to attach to them. The petri dish plus the chiton was weighed and then placed into 2 1 of aerated experimental SW and weighed at 1, 2, 4, 6, 8, 12, and 24 hr. Wet weights were determined by removing the petri dish with the attached chiton from the S\V, blotting dry. and immediately weighing on an analytical balance to the nearest 0.01 g. Nine replicate weighings of a single chiton in 100% SW gave a standard deviation of ±0.8% body wet weight. The small volume of water trapped in the mantle cavity of the chitons was assumed to remain unchanged during the experiment. The resulting small overestimates of the chiton's true wet weight were too small to seriously affect conclusions about volume regulation. Lange and Mostad (1967) have criticized volume-regulation experiments using whole animals : They state that weight variations could be caused by excretion and loss of fecal pellets. Few fecal pellets were seen in the containers during volume- regulation experiments with M. nntscosa; defecation probably did not contribute to the variability in wet-weight determinations. However, excretion remains a possible source of variation. The effect of temperature on volume regulation was studied by holding the chitons for 2-3 hr in 100% SW at either 7° or 19°C and then placing them in 60% SW of the same temperature and monitoring wet weights at 0, 2, 4, 6, 8, 12, and 24 hr. Calculations of osmotic permeability (I',,*) and oj chitons as osmoiiictcrs Calculations of osmotic permeability, Pos, and the theoretical rate of weight change of chitons responding as if they were perfect osmometers were based on the wet weight of the soft parts, determined as follows. Sixteen chitons were blotted dry and their wet weight determined by weighing on an analytical balance. Next, the chitons' plates were removed by scraping all girdle tissue from the plates, and all eight plates were weighed. Then, a regression line of plate wet weight vs. total-body wet weight was calculated. The equation was Y = 0.34 — 0.6X, and r, the coefficient of correlation, was 0.998. Thus plate wet weight could be obtained from a measurement of total-body weight. The wet weight of the soft parts could then be found by subtracting the plate wet weight from the total- body wet weight. In order to base all Pos values and theoretical values for chitons responding as perfect osmometers on the soft-part water content, the total body water content had to be determined. This was done by drying four chitons to constant weight at 67 °C. Using the above regression equation, the soft-part water content of the chitons could be determined. The osmotic water permeability, Pos, was calculated according to Fletcher (1974a) : r 7 r ' P = t(Ci + C0' - C0) '-- (Ecluatlon CHITON ION AND WATER BALANCE 367 where P,,s is in kg H^O/kg animal X hr X unit osmolal concentration difference, C0 is the osmolalily of the acclimatization medium ( 100% SW), C,/ the osmolality of chiton hemolymph in the acclimating medium, Z,, the water content of the animal (based on soft parts) (kg !!•_•( )/kg animal when acclimated to ('„), d is the osmolality of the experimental medium (60/4 SW ) and f is the fractional increase in weight t hrs after transfer to the experimental medium. Whenever the term C0'--C0 occurred in the equation it was cancelled, since chitons apparently are osmoconformers (McGill, 1975; Simonsen, 1975; Boyle, 1969). At various times after the chitons were placed in 60% SW, f was measured. P,,s was calculated for each time, and a graph of P,,s against t plotted. The theoretical increase in the weight of soft parts at a given time t of chitons behaving as if they were perfect osmometers was calculated according to Fletcher (1974b) : t = 1 7 r ' ^••O^ 1) p //"• _i_ r* i v 1, C0) r r, - r ' -- r v 1 1 v- (i V'O 7 (T - (\ ) ^•ov*— o ^-^ I/ ] - f (EquatIon Initial values of Pos must he used to calculate the theoretical osmometer curves ; later values would be reduced by the mechanisms regulating chiton volume. The initial values were obtained from the plot of Pos against time. Determination of muscle-tissue water content After 9-10 chitons had acclimated for 48 hr to the various experimental media, the muscle-tissue water content was determined. Small portions of foot muscle (60-110 nig) were oven dried at 96°C for 20 hr, and the percent water content was obtained from the difference between the wet and dry weights. These data were arcsine transformed for statistical purposes (Sokal and Rohlf, 1969). How- ever, the untransformed data are presented graphically. Statistical methods Means of two treatment groups were compared by Student's t test after deter- mining homogeneity of variance with an 77 test (Sokal and Rohlf, 1969). In a few cases the non-parametric Mann Whitney U test was used to test for statistical differences between two treatment groups. Regression analysis was performed by the method of least squares, and the significance of the regression coefficient was established by a t test (Sokal and Rohlf, 1969). Statistical significance is con- sidered to be the 5% level of probability. RESULTS Effect of salinity on hemolymph ion concentrations Following acclimation, the hemolymph Na+ and Cl~ concentrations of Mopalia muscosa were equivalent to the SW concentrations at all salinities tested. Hemo- lymph K+ and Mg-+ concentrations were also isoionic to the SW, except at 60% SW where the hemolymph K+ and Mg-* concentrations were greater than the SW values (P < 0.05, K+ ; P < 0.05, Mg2t, Mann Whitney U test). The blood Ca2+ 368 W. M. MORAX AND R. E. TULLIS TABLE I Blood ion concentrations of M. muscosa (mEq/l), at different salinities, compared to the SW concen- tration. N = 3 for each ion at each salinity. Values are expressed as the mean ±SE. % SW Blood Sea Water ci- 125 100 75 60 673.2 ± 10.00 531.2 ±- 9.10 407.5 ± 1.70 343.7 ± 3.80 686.2 ± 3.70 543.5 ± 0.60 402.4 ± 2.10 348.1 ± 1.10 Na+ 125 100 75 60 590.8 ± 3.50 483.2 ± 13.00 374.9 ± 5.10 303.4 ± 4.90 604.5 ± 1.00 479.2 ± 6.00 377.0 ± 8.10 318.8 ± 6.00 K+ 125 100 75 60 11.6 ± 0.60 10.2 ± 0.50 7.3 ± 0.12 5.8 ± 0.07* 11.0 ± 0.25 8.6 =b 0.20 6.9 ± 0.05 5.5 ± 0.05 Ca2+ 125 100 75 60 27.3 ± 0.50 21.8 ± 0.29 17.5 ± 0.13** 15.4 ± 0.18** 26.7 ± 0.10 20.8 ± 0.01 14.8 ± 0.00 12.7 ± 0.01 Mg2+ 125 100 75 60 119.5 ± 2.75 96.9 ± 1.18 75.2 ± 0.13 61.2 ± 0.21*** 122.9 ± 0.20 97.2 ± 1.60 75.1 ± 0.20 59.9 ± 0.00 * P < 0.05, hemolymph vs. SW (/ test) ** P < 0.001, hemolymph vs. SW (/ test) *** P < 0.05, hemolymph vs. SW (Mann Whitney U test) concentration of chitons acclimated to 125% and 100% SW were equal to that of SW ; but the hemolymph CaLH concentration was hyperionic in chitons acclimated to 75 and 60% SW (Table I). Hemolymph Cl~ concentration of chitons transferred to 60% SW from 100% SW reached equilibrium with the 60% SW Cl~ concentration 8 hrs after transfer. Blood Na+ concentration came into equilibrium between 8 and 12 hr after transfer of chitons from 100 to 60% SW (Fig. 1). Volume regulation In the fall, M. muscosa regulates volume in salinities as low as 60% SW. Chitons exposed to 60% SW for 6 hr showed a maximum increase of approxi- mately 20%- in total-body wet weight, whereas chitons exposed to 75% SW for 2 hr showed a maximum increase of 10%. In both 60 and 75% SW there is a decline of 6-7%, total-body wet weight after the maximum weight gain. In addi- tion, the rates of weight loss were equivalent in both hyposmotic salinities (Fig. 2). A regression of maximum weight gain of fall chitons exposed to 60% SW on total-body wet weight showed no correlation between chiton size and maximum weight gain. Chitons returned to 100%) SW following exposure to 60 or 75% SW showed a 3—4% undershoot of the original wet weight during the first hour of exposure CHITON ION AND WATER BALANCE 369 600 500 o § 400 300 480 440 - 400 o 360 12 16 20 24 o o 320 280 12 HOURS 16 20 24 FIGURE 1. Solid circles indicate the effect of time and salinity on blood Na+ and Cl~ concentration of M. muscosa. Chitons were transferred from 100 to 60% SW at time zero. Each point represents the mean, and vertical bars ± 1 SE, of three chitons. Open circles = 60% SW Na+ and Cl" concentration. (Fig. 2). Two to four hr after transfer to 100% SW the undershoot peaked at approximately 6% loss of body weight. Over the next 20 hr a gradual but sig- nificant (P < 0.01, 60%, SW; P < 0.05, 75%) SW) increase in weight occurred. Chitons exposed to 125% SW lost 8-9% of their original total-body weight and showed limited volume regulation after 12 hr exposure to the hyperosmotic medium. 4 8 12 16 20 24 28 32 36 40 44 48 HOURS FIGURE 2. Volume regulation of M. muscosa in fall at different salinities. Chitons in 60% SW, N=6 (solid circles); in 75% SW, N = 6 (solid squares); in 100%. SW, N = 4 (solid hexagon); in 125% SW, N = 3 (solid triangle). Vertical bars indicate ± 1 SE. At the arrows the chitons exposed to 60 and 75% SW were returned to 100% SW. 370 W. M. MORAN AND R. E. TULLIS 72 FIGURE 3. Volume regulation of M. muscosa. Spring chitons in 100% SW, N = 5 (solid triangles) ; in 75% SW, N = 7 (solid circles) ; in 60% SW, N = 5 (solid squares). Fall chitons in 100% SW (open triangles), in 75% SW (open circles) and in 60% SW (open squares). Sample size of fall chitons is shown in Fig. 2. Error bars are omitted for clarity. Another set of volume regulation experiments were performed over 72 hr to determine if M. niiiscosa was capable of further reducing its weight if specimens spent more than 24 hr in 60% SW. These experiments were done in the spring. An additional weight reduction of only 2% occurred over the next 48 hr (Fig. 3). The volume regulation response of fall chitons is also shown in Figure 3 to illustrate the difference in volume regulation of fall and spring chitons. Fall chitons in both 60 and 75% SW gain significantly (P < 0.01, both salinities) more weight than spring animals. In addition, spring chitons regulate volume more quickly than do fall chitons in both 60 and 75% SW. High temperature, 19° C, did not affect volume regulation of M. muscosa in 60% SW. However, low temperature, 7°C, reduced the chitons' ability to regulate volume (Fig. 4). Significant differences between the 7° and 13 °C curves occurred at 2 (P<0.02), 8 (P<0.05). and 12 hr (P < 0.05). The 6 hr values were not significantly different because of the large amount of variation associated with the 6 hr value of the control curve, 13°C. 12 HOURS 16 20 24 FIGURE 4. Volume regulation of M. muscosa in 60% SW at three temperatures in the spring. Chitons at 13°C, N = 5 (solid squares); 7°C, N = 6 (solid triangles); 19°C, N=6 (solid circles). Data points are offset on the ordinate for clarity of diagram. CHITON ION AND WATER BALANCE 371 Osmotic permeability and chitons responding theoretically as osmometers The osmotic permeability, Pos, (calculated according to equation 1) was 0.71 ± 0.045 kg H^.O/kg animal X hr ' X unit osmolal concentration + is involved in release of neurotransmitters from presynaptic neurons of the squid Loligo vulgaris (Miledi, 1973), and the action potential in the heart of the bivalve Modiolus dcuiissus is dependent primarily on Ca-' (\Yilkens, 1972). Furthermore, ciliary movement is dependent on extracellular Ca'-'* concentration (Eckert, 1972; Murakami and Eckert, 1972). The blood Xa+ and Cl~ concentrations are reduced most rapidly during the first 2 hr of exposure to 6Qc/f SW (Eig. 1). The reduction of blood salt concentration could be due to salt loss (either diffusive across the body wall or through the urine) and to dilution of the blood by the water influx. The relative importance of salt loss and water influx in reducing blood Na+ and Cl" concentrations remains uncertain but cannot be ignored, since the diffusive salt loss would contribute to volume control by reducing the driving force for osmotic water flux existing across the body wall of hyposmotically stressed chitons. Diffusive salt loss in chitons has been demonstrated indirectly. Specimens of Sypharochiton pclliscrpentis exposed to a seawater/sucrose mixture isosmotic to SW lost weight. This sug- gests that an isosmotic solution of salt and water was lost from the chitons, and that diffusive salt loss may contribute to volume control (Boyle, 1969). Although blood osmotic pressure was not measured, the blood ion data, especially that of blood Na+ and Cl~ concentrations, of salinity stressed chitons suggest that M. mnscosa is an osmoconformer. Osmoconformity has been demon- strated in members of the genus Mopalla (Boyle, 1969) ; and the chitons Cyanoplax hartwegii (McGill, 1975), Nuttalina californica (Simonsen, 1975), and Sypharo- chiton pelliscrpcntis (Boyle. 1969) are osmoconformers in hypo- and hyperosmotic SWr. Since chitons are probably osmoconformers, intracellular volume regulation must occur during salinity stress. In most marine and estuarine invertebrates intracellular free amino acids are the source of solute for this regulation ( Pierce 84 8 so 76 z o o 72 ' *• * 68 60 80 100 % SW 120 140 FIGURE 6. Water content of the foot muscle of M. inuscosa as a function of salinity. Observed water content of muscle (closed circles) ; expected water content of muscle responding as if it were an osmometer (open circles). For chitons in 60, 75, and 100% SW, X — 10 ; 125% SW, N = 9. Vertical bars = t 1 SE. CHITON ION AND WATER BALANCE 373 and Greenberg, 1973). Thus, osmotically stressed muscle tissue which relies on intracellular volume regulation would not be expected to alter its water content as would a Boyle- van't Hoff osmometer. This was found for the foot muscle- tissue of M. Jiinscosa (Fig. 6). Volume regulation has been studied in other chitons and in bivalves. For example, the chiton Cyanopla.v hartu'cgii can regulate volume in both ', i and 125% SW although the volume response is faster and more complete in / than in the hyperosmotic medium ( McGill, 1975). McGill considers the hyp- osmotic volume-control response adaptive in chitons because it will limit swelling of the foot and allow continued attachment to the substrate. Volume regulation has also been studied in vertically separated populations of the chiton Nuttalina califomica. Low-intertidal Nuttalin-a gain more weight in 50% SW than do high- intertidal specimens. Volume regulation was not demonstrated in either group ; however, the maximum weight increase of both populations was less than that of chitons responding as if they were perfect osmometers (Simonsen, 1975). Like- wise, the chiton Sypharochiton pelliserpentis does not regulate volume in hyp- osmotic SW. Instead, adaptation to hyposomotic media is accomplished by diffusive salt loss and by considerable tolerance to dilution of body fluids (Boyle, 1969). Volume regulation in hyposmotic SW has also been demonstrated in the bivalves Modiolus demissits granosissimus and M. sqitainosus. Volume control was not shown in either species when they were exposed to hyperosmotic SW (Pierce, 1971). After a 24 hr exposure to either 60 or 75% SW (fall experiments), the chitons were returned to 100% SW, where they exhibited an undershoot of their original weight; that is, on return to 100% SW the chitons lost weight until they weighed less than their original weight at time zero. This may be due to loss of solute from the organism during exposure to hyposmotic SW (Gross, 1954). However, if this were the only cause the extent of undershoot should be correlated with salinity. This correlation was not found, suggesting that the chitons exposed to 60 and 75% SW lost the same amount of solute. A somewhat surprising result of returning the chitons to 100% SW after exposure to hyposmotic SW was their steady weight gain after peak weight loss (Fig. 2). These chitons appear to begin volume regulation within 6 hr on return to 100% SW, whereas chitons exposed to 125% SW shrink passively and show limited ability to regulate volume. Thus, previous exposure to hyposmotic SW appears to activate volume regulation of M. muscosa re-exposed to 100% SW. These results differ from those of Pierce (1971 ) in a study of volume regula- tion in various species of Modiolns. Re-exposure of Modiolus demissus granosis- simum, M. modiolus, M. demissus, and M. squauiosus to full-strength SW after exposure to hyposmotic SW resulted in an undershoot of their original weight, and none of the species examined was able to regain its original volume. The physio- logical basis for this difference between chitons and bivalves is not known. Volume control in an organism exposed to hyposmotic media may begin before the usually observed decline in weight. For example, M. muscosa begins to limit its rate of weight increase after I hr of exposure to hyposmotic SW (Fig. 5). Therefore, volume control mechanisms in Mopalia begin to function several hours before the observed decline in weight. Likewise, Machin (1975), using a graphic analysis of weight and water-content regulation, demonstrated that volume com in the polychaete Glyccra dibranchlata begins before osmotic equilibrium is re The volume control mechanisms responsible for the chitons' early deviation : 374 W. M. MORAX AXD R. E. TULLIS perfect osmometer behavior are not known. However, this effect could result from a change in epithelial permeability to water, an increase in water excretion or dif- fusive salt loss. Low temperature (7°C) reduced rate of weight gain of M. muscosa exposed to 60% SW. Likewise, the weight-loss rate was reduced during volume regulation. Therefore, temperature seems to affect processes controlling both osmotic influx and volume regulation. Volume regulation in Mopalia may vary with season (Fig. 3). Chitons on the U. S. -Canadian \Yest Coast reproduce in the spring, and this difference could be related to the animals' reproductive state. Furthermore, all chitons in this study were taken within a restricted vertical range in the intertidal zone: Animals sepa- rated vertically in the intertidal zone have different periods of exposure, and, as mentioned above, Simonsen (1975) has demonstrated that low-intertidal specimens of Nnttalina califomica gain more weight in hyposmotic SYV than high-intertidal animals. Thus, season and vertical position in the intertidal zone, should be con- sidered in future studies concerning water balance of intertidal organisms. In conclusion, M. muscosa responds to short-term salinity stress by quickly activating mechanisms which control volume. This is of adaptive significance to an organism inhabiting the intertidal zone, where salinity may change rapidly. The results reported in this paper were part of a Master's thesis submitted to the Graduate School of the California State University at Hayward. The authors wish to thank Dr. John Martin of the Moss Landing Marine Laboratories for allowing us to use the atomic absorption spectrophotometer. We also thank Dr. Robin Burnett of Hopkins Marine Station of Stanford University for check- ing some of the calculations in this paper. We thank Dr. Ralph Smith of the Uni- versity of California at Berkeley, and Dr. Sidney Pierce, Jr., and Chuck Costa, both of the University of Maryland at College Park, for critically reviewing the manuscript. SUMMARY 1. The effects of external salinity changes on whole-animal volume, blood ions, and muscle-tissue water content of the chiton Mopalia muscosa were investigated. The data indicated that short-term low-salinity adaptation in the chiton M. muscosa is achieved by volume control mechanisms that are quickly activated. 2. Blood Na+, Cl~, K+, and Mg-+ were isoionic to the SW concentration in salinities ranging between 60 and 125% SW. Blood CaL>+ concentration was hyper- regulated in hyposmotic SW. 3. Regulation of cell volume occurred in salinity-stressed chitons, as water content of foot-muscle tissue was regulated in both hypo- and hyperosmotic media. 4. When exposed to hyposmotic SW the chitons at first (0—4 hr) gained weight ; this was followed by a period in which the rate of weight gain approached zero (4-6 hr) and finally by a period of weight loss (6-24 hr) (volume regulation). Exposure to hyperosmotic SW resulted in weight loss and little volume regulation. 5. Volume control in hyposmotic media is accomplished, in part, by a loss of solute. 6. Low temperature, 7°C, decreased both the water influx and volume reg- ulation. CHITON ION AND WATER BALANCE 375 7. Volume regulatory mechanisms are activated within 1 hr after exposure to SW. LITERATURE CITED BOYLE, P. R., 1969. The survival of osmotic stress by Sypharochiton pclliscrpc, Mollusca1 Polyplacophora). Biol. Bull.. 136 : 154-166. ECKERT, R., 1972. Bioelectric control of ciliary activity. Science, 176: 473-481. FLETCHER, C. R., 1974a. Volume regulation in Nereis diversicolor — I. The steady Comp. Biochem. Physio!.. 47A : 1199-1214. FLETCHER, C. R., 1974b. Volume regulation in Nereis diversicolor — III. Adaptation to a reduced salinity. Comp. Biocltcm. Physiol., 47A : 1221-1234. GROSS, W. J., 1954. Osmotic responses in the sipunculid Dendrostomum zostcricolum. J. Exp. Biol., 31 : 402-423. LANGE, R., AND A. MOSTAD, 1967. Cell volume regulation in osmotically adjusting marine animals. /. Exp. Mar. Biol. Ecol., 1 : 209-219. MACHIN, J., 1975. Osmotic responses of the blood worm Glyccni dibranchiata Ehlers : A graphical approach to the analysis of weight regulation. Comp. Biochem. Ph\siol., 52A: 49-54. McGiLL, V. L., 1975. Responses to osmotic stress in the chiton, Cyanoplax hartwegii (Mol- luscs: Polyplacophora). Veligcr, 18 (Suppl.) : 109-112. MILEDI, R., 1973. Transmitter release induced by injection of calcium ions into nerve terminals. Proc. R. Soc. Land. B. Biol., 183 : 421-425. MURAKAMI, A., AND R. ECKERT, 1972. Cilia : activation coupled to mechanical stimulation by calcium influx. Science, 24 : 1375-1377. OGLESBY", L. C., 1975. An analysis of water content regulation in selected worms. Pp. 181- 204 in F. J. Vernberg, Ed., Physiological Ecology of Estuarine Organisms. University of South Carolina Press, Columbia, South Carolina. PIERCE, S. K., JR., 1971. Volume regulation and valve movements by marine mussels. Comp. Biochem. Physiol., 39A : 103-117. PIERCE, S. K. JR., AND M. J. GREENBERG, 1973. The initiation and control of free amino acid regulation of cell volume in salinity-stressed marine bivalves. /. E.rp. Biol., 59 : 435- 440. PIPER, S. C., 1975. The effects of air exposure and external salinity change on the blood ionic composition of Nuttalina californica. Vcliger, 18 (Suppl.) : 103-108. SCHALES, O., AND S. S. SCHALES, 1941. A simple and accurate method for the determination of chloride in biological fluids. /. Biol. Chem., 140: 879-884. SIMONSEN, M., 1975. Response to osmotic stress in vertically separated populations of an intertidal chiton, Nuttalina californica. Veligcr, 18 (Suppl.): 113-116. SOKAL, R. R., AND F. J. ROHLF, 1969. Introduction to Biostatistics. W. H. Freeman and Co., San Francisco, 368 pp. STICKLE, W. B., AND R. AHOKAS, 1975. The effects of tidal fluctuation of salinity on the hemolymph composition of several molluscs. Comp. Bwchcm. Physiol., 50A : 291-296. WATTS, J. A., AND S. K. PIERCE, JR., 1978. A correlation between the activity of divalent cation activated adenosine triphosphatase in the cell membrane and low salinity tolerance of the ribbed mussel, Modiolus demissus dcmissus. J. E.rf>. Zoo/., 204 : 49-56. WEBBER, H. H., AND P. A. DEHNEL, 1968. Ion balance in the prosobranch gastropod, Acmaca scutum. Comp. Biochem. Physiol., 25 : 49-64. WILKENS, L. A., 1972. Electrophysiological studies on the heart of the bivalve mollusc, Modiolus demissus. II. Ionic basis of the action potential. /. Exp. Biol., 56: 293-310. Reference : Biol. Bull.. 159: 376-393. (October, 1980) ELECTRICAL ACTIVITIES IN THE SUBTENTACULAR REGION OF THE ANTHOMEDUSAN.SYJ/tfOrO/)OA< SALTATRIX (TILESIUS) KOHZOH OHTSU Ushniittdo Marine Laboratory, Okayama University, Kas/iino, Usliiuiado, Oku. Okayuiiui. 701-43 Japan Electrophysiological studies of hydrozoans have revealed various kinds of elec- trical pulses of epithelial, muscular, or nervous origin. One striking feature of hydrozoans is that the conducting systems of those pulses often lie parallel to one another. It is interesting to determine how such parallel systems function and therefore, much effort has been made to show the origin, nature, and function of each conducting system, e.g., in Hydra ( Passano and McCullough, 1962, 1964, 1965; Josephson and Macklin, 1967; Josephson, 1967; Macklin and Josephson, 1971 ; Rushforth and Burke, 1971 ; Rushforth, 1971 ), in Tubularia (Josephson and Mackie, 1965; Josephson, 1965, 1974; Josephson and Rushforth, 1973; Kruijf, 1977), in hydromedusans (Spencer, 1975, 1978; Mackie, 1975) and in siphono- phores (Mackie, 1976a, 1978). Spirocodon, the hydrozoan medusa used in this study, also has such parallel systems in some tissues (Ohtsu and Yoshida, 1973). Only those in the area between the ocelli and the nerve ring called the "sub- tentacular region" will be considered here. Spirocodon saltatri.r has many morphological charactristics in common with other species of Anthomedusae (Horridge, 1955; Mackie and Passano, 1968; Spencer, 1979; King and Spencer, 1979) ; but it shows some striking differences from them in the marginal region of the umbrella. Figure 1 shows a part of the subtentacular region. According to Uchida (1927), the subtentacular region is formed in the following manner: In early developmental stages, single ocelli are present at the base of the four perradial and the four interradial tentacles, close to the nerve rings. As tentacles grow, these ocelli move away from the inner margin of the umbrella, forming an inter- vening expanse of tentacular tissues, the radial streaks, between the ocelli and the nerve ring while new ocelli and tentacles are formed at both sides of and slightly below those already formed. The newer the ocelli, the shorter the radial streaks. Thus a crescent-like area called the subtentacular region is produced between the ocelli (tentacular bases) and the nerve rings (velar base). A fully grown medusa has eight subtentacular regions at the lower margin of the umbrella. In large specimens, each subtentacular region contains more than 60 radial streaks. Sketches of the subtentacular region and the whole body are shown in Ohtsu and Yoshida (1973). The subtentacular region should provide pathways for excitations traveling from the periphery (tentacles and ocelli) to the center (nerve ring). Histological as well as electrophysiological investigations were carried out to identify cell types within the subtentacular region in order to provide a morphological basis for inter- preting data on pulse generating sites characterizing each conducting system. 376 ELECTRICAL ACTIVITIES OF SPIROCODON -ExuEp SubuEL- Subu M SphM 377 FIGURE 1. The subtentacular region. Four and a half radial streaks are shown. Ect, ectoderm; End, endoderm ; Exu EP, exumbrellar epithelium; IXR, inner nerve ring; TM, transparent mesogloea ; Oc, ocellus; ONR, outer nerve-ring; Rd S, radial streak; Rn C, ring canal ; Sph M, marginal sphincter muscle ; Subt. R, subtentacular region ; Subu EL, subumbrellar endodermal lamella; Subu M, subumbrellar muscle sheet; Te, tentacle; Te C, tentacular canal; OM, opaque mesogloea ; Ve, velum. For explanation, see text. MATERIALS AND METHODS Specimens of Spirocodon saltatri.i- were collected from the Seto Inland Sea, Japan, and kept in running sea water. Half of the subtentacular region, with a small fraction of mesogloea lying under it, was dissected from the margin of the umbrella and used for histological observations and electrical recordings. For electron- and light-microscopical observations, the ocelli and radial streaks were dissected from the subtentacular region. They were fixed in chilled 2% glutaraldehyde for 2 hr and post-fixed in 2% osmium tetroxide for 2.5 hr. To regulate osmolarity, 0.9 M sucrose or sea water was used. In some cases, fixation was done at 15°C for the first hour to retain microtuble structure. Fixatives were buffered with 0.1 M Na-cacodylate, pH 8.0. Post-fixation was omitted in the dye-marking experiment described later. The specimens were dehydrated through an ethanol series and embedded in TAAB embedding resin (TAAB Laboratories). Sections were cut with a Porter-Blum MT-2B ultramicrotome. Silver to gray sections were stained with uranyl acetate and then lead citrate and observed with a Hitachi electron microscope (HS-8). Thicker sections of 1-2 ^m were stained with methylene blue for light microscopy. Experiments were performed at room temperature ( 18-21 °C). The arrange- ment of the recording chamber was similar to that described in Ohtsu and Yoshida (1973). During recording, the surface of the preparation was exposed to air to reduce short-circuit currents. During intracellular recording, preparations were almost totally immersed in sea water. Recording electrodes were advanced with a hydraulic micromanipulator in conjunction with a mechanical advance. For extracellular recordings, two types of glass micropipette electrodes were used : rigid or floating. The latter consisted of a coiled piece of tungsten wire (30 /j.m diameter) joined with wax to the proximal opening of a micropipe electrode tip a few millimeters long. This type of electrode allowed vigorous move 378 KOHZOH OHTSU merits of the preparation without electrical artifacts. Both types of electrodes were filled with sea water. External tip diameters ranged from 1 to 5 /xm. For intracellular recordings, micropipettes were filled with 2 M KC1 and had DC-resistances of 20-100 MQ. Rigid electrodes were used. Preparations were treated with a 0.3-0. 5 r/f pronase/sea-water solution for 4-8 min to facilitate penetration, aided by a jolting device ( Tomita ct al., 1967). A 1-2 hr immersion in pronase was used to dissociate the endodermal cells for examination with Normarski optics (Leitz, Orthoplan). To measure depth of penetration of the recording electrode tips, a pair of elec- trodes was used, unless stated otherwise. Initially, both electrodes were placed on the surface of the radial streak and then one electrode was advanced through the tissues, leaving the other on the surface. Because smooth penetration was impos- sible, depths were estimated during withdrawal of the electrode, using a micro- manipulator scale. A micropipette filled with M/2 K-ferrocyanide instead of sea water was used to mark the recording sites. The pipettes with K-ferrocyanide were not used as recording electrodes because leakage of the electrolyte from the large pipette tips might have deleterious effects and because tissue penetration with such pipettes often resulted in failure of electrophoretic ejection of the fluid, probably because the pipette tips became plugged with tissue debris during penetration. After recording, the pipette with K-ferrocyanide was attached to the recording electrode tip and the electrolyte was expelled into tissues by inonophoretic current. One or two drops of ferric chloride solution were delivered to make a Berlin blue precipi- tate. Cross-sectional dimensions of tissues were estimated from light micrographs compared with readings from the micromanipulator scale. The most serious difficulty with the rigid electrode was immobilizing the preparations. Since tissues can contract locally, mechanical restraint was impos- sible. Instead, various anaesthetics, such as urethane, menthol, MS-222, and magnesium chloride, were tried. Urethane was most successful. The urethane concentration was varied within the range of 1.0-3.0 X 10"1 M, since the effect was not the same on every preparation. Pulses can be evoked by termination of adapting light (Ohtsu and Yoshida, 1973). Light from a tungsten lamp (35 W) was delivered through a heat-absorb- ing filter and a light guide, the illuminance of which was 3500 lux in the plane of the preparation. Electrical stimulation was by means of paired silver wires of 200 /xm diameter through an isolation unit (Nihon Kohden, MSE-JM). In some cases, however, stimuli were delivered through the recording electrode. Electrical responses were fed to a two-channel preamplifier formed by two FET-coupled operational amplifiers (Teldyne Filbrick, 142502) arranged in parallel. One was equipped with a Wheatstone bridge circuit, permitting measure- ment of membrane resistance changes. Two DC/AC amplifiers (Nihon Kohden, AVM-9S, AVH-9) were employed; the former, equipped with a beam chopper device, was most frequently used. The latter was used as a differential amplifier or to monitor photosignals. All responses were displayed on a dual-beam cathode ray oscilloscope (Nihon Kohden, VC-9A), photographed by a continuous recording camera (Nihon Kohden, PC-2B) and visually monitored on a second oscilloscope (Iwatsu, SS-5100). ELECTRICAL ACTIVITIES OF SPIROCODON Ca FIGURE 2. Photomicrographs of cross-sections through the radial streaks. Ca, canal ; EC, endodermal cell; Ect, ectoderm; Ne, nematocyst ; Oc, ocellus; OM, opaque mesogloea ; Subt NB, subtentacular nerve bundle; TM, transparent mesogloea. Arrows in D show the sub- tentacular nerve bundles. Horizontal bars, 50 /um. RESULTS Histological observations The ectoderm of the radial streaks is similar to that of the tentacles of other hydromedusans. Figure 2A shows a cross-section of two radial streaks. Figures 2B and C show enlargements of parts of A. Endodermal epithelial cells, sur- rounding the tentacular canal situated at the center, project their fine cytoplasmic processes towards a surrounding thin mesogloeal layer, resulting in many large endodermal spaces separated by the processes. The endodermal spaces may look larger in the photomicrographs than they actually are (Fig. 2) because the fine endodermal processes are often smaller than 0.1 /im, and thus beyond the limit of resolution of the light microscope. The mesogloeal layer is far more opaque than the jelly-like mesogloea found elsewhere. Therefore it will be called the opaque mesogloea and the latter, the transparent mesogloea. A layer of ecto- dermal cells surrounds the opaque mesogloea. The transparent mesogloea cannot be seen because it has the same refractive index as Jie embedding resin. The subtentacular region is folded between adjacent radial streaks, forming a groove (Fig. 2A) where a single nerve bundle runs (Fig. 2C). This bundle will be called a subtentacular nerve bundle. The grooves become flatter near the nerve-rings. In contrast, going towards the ocelli away from the nerve-rings, the grooves become deeper, i.e., the ectodermal cell layer extends progressive!; around the endoderm, while the subtentacular nerve bundle divides into tv% branches, each running in a corner of the groove (Fig. 2D). When the ecto completely surrounds the endoderm, the ocelli appear in the ectoderm on the brellar side (Fig. 2E). Beyond the ocelli, the radial streaks become the KOHZOH OHTSU FIGURE 3. Electron micrographs of cross-sections through the radial streaks. A, the ectoderm and a part of the endoderm ; B, high magnification of the endoderm ; C, the sub- tentacular nerve bundle and high magnification of a part of it (asterisk) ; D, endodermal cells. Ca, canal ; End, endoderm ; End CP, fine endodermal cell process ; EP, the epithelial process of the myoepithelial cell; MBV, membrane bound vesicle; MBDV, membrane bound digestive ELECTRICAL ACTIVITIES OF SPIROCODON 381 The construction of the tentacles, therefore, necessarily resembles that of the radial streaks. In the vicinity of the ocelli, the nerve bundles are missing, probably because the nerves have dispersed. It seems that at least some axons, if not all of them, receive input from the ocelli and function in the photon- The major functions of the subtentacular nerve bundle, however, are unknov Figures 3A-C show the ultrastructure of the ectoderm in cross sections The thickness of the layer usually lies in the range 10-50 /u.m though it different preparations and in different parts of the sections. The muscular proa of the myoepithelial cells are smooth muscle fibers running along the axis of the tentacle. They lie in a layer on the innermost side of the ectoderm, anchored to the mesogloea (A). They are connected with each other by structures resembling the gap junctions reported in other hydrozoan cells (Hand and Gobel, 1972). The muscle layer becomes thinner towards the nerve-ring and thicker towards the tentacle. The epithelial proccesses lying at the outermost surface and forming almost the whole surface of the ectoderm contain the nuclei and extremely large vacuoles lying in relatively undifferentiated cytoplasm (A). The opaque meso- gloea is perforated in places by a tissue similar to that of the endodermal cell processes, allowing direct contact between the ectoderm and the endoderm, as reported in other anthomedusans (Hoelsterli, 1977). A number of nerve axons with variable diameters (Figs. 3A, B) and some- times neuronal somata of about 5-10 //.m in diameter are present. Both contain a number of microtubules and often membrane-bound vesicles. They are almost always adjacent to the myoepithelial cells. Some cells with a number of membrane- bound vesicles (recalling neurosecretory granules), are also observed (Fig. 3B). Synapse-like structures, however, have not been found so far between the cells identified, though they are frequently observed around the ocelli (Toh et al, 1979). The subtentacular nerve bundles consist of a large number of fine axons which contain many microtubules and membrane-bound vesicles (Fig. 3C). The number of axons differs from preparation to preparation but gradually decreases going towards the ocelli. No neuronal somata have so far been found along the bundle. They probably lie elsewhere, e.g., adjacent to the ocelli or outside the nerve bundle. It must be noted that muscle processes are weakly developed near the sub- tentacular nerve bundle. Cnidocytes are also present among the myoepithelial cells in the subtentacular region but they are far less abundant than in the tentacles. The fine structure of the endodermal epithelial cells is shown in Figure 3D. A single type of endodermal cell can be identified in the subtentacular region except near the ring canal, where endodermal nerves can be seen as reported in Stoniotoca (Mackie and Singla, 1975) and Polyorchis (Singla, 1978; Spencer, 1979). The endodermal cell cytoplasm expands on the side facing the tentacular canal, and the endoplasmic reticulum, Golgi apparatus, and mitochondria are in this region. A number of microvilli and cilia are present on the surface adjacent to the canal. Septate desmosomes are also observed connecting cells on their lumenal sides (Fig. 3D, inset). Membrane-bound digestive vacuoles are often observed, sug- gesting active phagocytosis. The nuclei lie towards the periphery. More peripherally, the cell processes become thin but often form enlargements includ- ing mitochondria and even nuclei on occasion, and extending in sheets of two vacuole ; MP, the myonal process of the myoepithelial cell; N, nucleus; NA, nervi OM, opaque mesogloea; SJ, septate junction; Vc, vacuole. Horizontal bars, 0.2 /j.m in ( D insets and 2 fj.m in the others. 382 KOHZOH OHTSU or more layers as far as the opaque mesogloea, creating many large spaces between the cell processes (Figs. 2, 3A, 3D inset). Where they reach the opaque mesogloea, the epithelial cell processes line its inner side (Fig. 3A). Although the large endodermal spaces seen among the epithelial cell processes appear to he extracellular, this may not be the case. The epithelial process almost always consists of at least two layers of endoderm (Fig. 3D, inset). The narrow gap between the two layers is extracellular space, often observed to be connected with the central canal via a septate junction. If the large endodermal spaces were extracellular spaces, then they, too, should often be connected to the canal. But such cases have not been observed in electron micrographs. It is likely, therefore, that the large endodermal spaces represent extraordinarily large vacuoles in endodermal epithelial cells. Supporting evidence was obtained from experiments involving dissociation of the tissues with pronase. As the endodermal cells dissociate, the endodermal spaces become spherical and some of the cells break free as single spherical cells. Vacuoles occupy the greater part of the cytoplasm and the nuclei and major portion of the cytoplasm lie near the periphery of the cell. Cells isolated from the endoderm vary greatly in size. For instance, in one treatment the cells measured 15-20 /tin and sometimes 30 fj.ni, but in another treatment some cells measured 100 Conduction velocities of pulses recorded When the recording electrode is placed at the surface of the subtentacular region, three types of pulses are recorded: slow monophasic pulses (SMPs; Figs. 4A, C), quick and slow compound pulses (QSCPs; Figs. 4A-D), and tentacle contraction pulses (TCPs; Fig. 4D). The first two were reported by Ohtsu and Yoshida (1973). The QSCP consists of a very short phase preceding a slow 0.5 mV 100 msec FIGURE 4. Pulses recorded in the subtentacular region. A and B show conduction of QSCPs. In A, a single electrical stimulus was delivered to a part of a tentacle distal to the recording sites. In B, a stimulus was delivered to an adjacent tentacle. C shows TCPs (t) following long after QSCPs and an SMP. D shows a TCP (t) following soon after two successive QSCPs, suggesting triggering of the TCP by the QSCPs. The distance between the recording electrodes and the stimulus site is much larger in D than C, as can be seen from the longer delay of the QSCPs after the electrical stimulus. E shows multiple firing of the quick phase of QSCPs over the course of the prolonged slow phase induced by urethane treatment. Arrows, electrical stimuli, "q" and "s" indicate a QSCP (or QSCPs) and an SMP. respectively. ELECTRICAL ACTIVITIES OF SPIROCODON 383 phase which bears a remarkable resemblance to the SMP in terms of its time course and waveform (Fig. 4A). A preliminary experiment showed that the conduction velocity ' QSCPs in the subtentacular region is approximately 60 cm/sec (Ohtsu and 1973). This value represented only a single measurement, however, so furt letailed bidirectional measurements were performed on the tentacles, taking ; : of their greater length. Electrical stimuli were employed to evoke QS proximal parts of moderately relaxed tentacles, conductive velocities were calcula from the time difference between the quick components of the QSCPs recorde two recording electrodes. For distal to proximal conduction, stimuli were livered at distal parts of the tentacles with recording electrodes placed more proximally (Fig. 4A), while for proximal to distal conduction a neighboring tentacle was stimulated (Fig. 4B). In the latter case, QSCPs would first travel proximally up the stimulated tentacle and would reach the recording sites by travelling distally down the second tentacle after crossing the subtentacular region. The longer post-stimulus delays of the pulses in Figure 4B, compared with those in Figure 4A, reflect the longer distance travelled. The mean value obtained from 25 different tentacles from eight animals was 75 cm/sec (range, 50-104 cm/sec) in distal to proximal direction and 73 cm/sec (range, 57-95 cm/sec) in the opposite direction. In most cases. QSCPs appeared coupled with each other on a one-to- one basis within the subtentacular region and the tentacles associated with it. SMPs can be conducted for only a short distance within the same radial streak (Ohtsu and Yoshida, 1973). The generating sites of spontaneously occurring SMPs frequently changed, as shown in Figure 7A, where with the first SMP pair, a pulse first appears on the lower trace and then on the upper trace, while in the second pair, pulses appear in the reverse order. Moreover, SMPs some- times disappeared between the two electrodes and were not conducted (Fig. 4C). Although the pulses can be induced by termination of adapting light with latencies of 0.2-1.1 sec, probably mediated by the ocelli (Ohtsu and Yoshida, 1973), it was difficult to restrict pulse generation to a single fixed site ; one should be cautious in estimating the conduction velocity because the pulses were often generated between two recording electrodes. Waveforms of SMPs became variable as they moved distally along the tentacles so that it was often difficult to identify the corresponding phases of pulses recorded at two points. Estimation of conduc- tion velocity, therefore, was performed in the subtentacular region, using spontane- ous pulses or pulses triggered by lights-off and measuring the peak-to-peak time delay, since the initial rising phase was not steep enough to allow initiation time to be accurately determined. Cases with extremely short time delays were excluded, as they were probably due to initiation of pulses between the two recording electrodes. The mean value from 32 different radial streaks from eight animals was 16 cm/sec (range, 9-38 cm/sec). It must be noted that SMPs with conduction velocities less than 9 cm/sec were seen, especially during sequences where later pulses tended to be conducted far more slowly than earlier ones. However, these data were omitted from the above measurements, since they were obtained in experi- ments using urethane. There may be more than one conducting pathway for SMPs. When electrical recordings were performed on the subtentacular region or the tentacles, extremely slow deflections with monophasic and sometimes more com- plicated waveforms were observed, accompanied by muscle contractions 4C) ; the waveforms may include electrical artifacts due to tentacular movi These pulses are referred to hereafter as tentacle contraction pulses 384 KOHZOH OHTSU Measurement of their conduction velocity was done as for SMPs but using pulses elicited by electrical stimulation on the tentacles and recorded in the subtentacular region. Recordings were not made on the tentacles because the pulses were larger and of simpler waveform on the subtentacular region. The mean value from 29 different radial streaks from nine animals (range, 2.2-8.6 cm/sec), which is very slow compared with the conduction velocity of other pulses, suggests that TCPs are conducted without involvement of the nervous system. Figure 4C illustrates the great difference in the conduction velocities of QSCPs and TCPs, whereas the TCP in Figure 4D occured shortly after the QSCPs. A possible explanation for this is that TCPs can be induced directly by QSCPs in the course of conduction. Such examples, indicating triggering of TCPs, were often encountered when QSCPs appeared in rapid succession. Waveform changes of pulses associated zvith electrode insertion depth To determine where QSCPs originate, one of a pair of recording electrodes was inserted into a radial streak, while the other was placed on the surface. Urethane at 2 X 10'1 M concentration was used to immobilize the preparation although this concentration of urethane did not always completely abolish movement. This treat- ment, however, often prolonged the duration of the slow phase of QSCPs con- siderably and sometimes resulted in multiple firing of the quick phase over the course of the slow phase (Fig. 4E). This was never seen in untreated preparations. Immobile preparations with normal pulse waveforms were selected for these experiments. Treatment with urethane further abolished the lights-off evocation of QSCPs and SMPs so that the pulses had to be evoked by electrical stimulation of distal parts of the tentacles, resulting in the disappearance of locally conducted SMPs. During insertions of recording electrodes into the radial streak, smooth pene- tration was always interrupted in two places, once near the surface and again at considerable depth. The depths at which these interruptions occurred seem to correspond to the interfaces between the ectoderm and opaque mesogloea near the surface, and between the opaque and transparent mesogloea on the opposite side of the radial streak (see Fig. 2). Results of one penetration experiment are shown in Figure 5 : When both recording electrode tips were placed side by side on the surface of a radial streak, QSCPs with the same waveform appeared in the upper and the middle traces, with no deflection in the differential, lower trace (A). At a depth of 635 /mi, a very small deflection with a waveform similar to that of the upper trace was observed (B). It must be noted, however, that in B and also in the following records, the pulses recorded on the surface were smaller than the one in A, presumably due to damage to cells near the surface by the electrode as it advanced and as the wirier part of tapered electrode tip entered the surface tissues. With- drawal of the electrode in 70 /mi steps resulted in reversal of the polarity of the deflection in the middle trace (C). As the electrode was withdrawn further, positivity of the middle trace was gradually increased (D, E) and reached a maximum in or near the canal (F-H). With further withdrawal the slow deflection gradually became less positive (I-K) and eventually reversed its polarity again (I.). Approaching still close to the surface, the negative slow deflection grew larger (M, N) and reached its maximum amplitude on the ecto- dermal surface (O). These facts clearly indicate that the surface of the radial ELECTRICAL ACTIVITIES OF SPIROCODON 385 B D -B ImV 100 ms FIGURE 5. Waveform changes of QSCPs associated with the depth of the recording electrode tip. The recording electrode was inserted into the radial streak to a depth of 700 /j.m and then withdrawn stepwise. The upper traces were recorded from the surface electrode, the middle traces from the inserted electrode, and the lower traces show differential recordings (upper records minus middle records). The right inset diagram shows a cross-section of the radial streak used here. A, surface; B, 635 fum; C, 565 /mi; D, 515 pm ; E, 465 /j.m ; F, 415 Aim; G, 365 /mi ; H, 315 /um ; I, 265 /j.m; J, 215 jum ; K, 115 urn; L, 65 /xm ; M, 45 pm; N, 25 /urn O, surface. QSCPs were evoked by electrical stimuli before initiation of the recordings. streaks becomes negative (about 1.3 mV in H) compared with the central region of the endodermal canal during the slow phase of QSCPs. Similar results were obtained in four other preparations. Difference in DC-level measured between the ectoderm and the endoderm was variable, but in most cases was endodermal- side positive by a few millivolts. The polarity of the quick phase of QSCPs, on the other hand, remained unchanged throughout the experiment. The amplitude was very small deep in the radial streak (Fig. 5B-J) and became larger when the electrode was moved to 100 /mi or less from the surface (K-O). These changes are shown more clearly in Figure 6, where the quick phases were much larger than the slow phases, indicat- ing that the recording site was nearer the nerve ring than the ocelli. The quick phases of QSCPs usually dominated the slow phase near the nerve ring, and vice versa near the ocelli ; more nervous elements could be found near the former. Both the quick and slow phases of QSCPs abruptly increased when the electrode was withdrawn to within 40 /um of the surface of the ectoderm, strongly sug- gesting that both phases of QSCPs originate in the ectoderm. The waveforms of the pulses in Figure 6E closely resemble those at the surface (Fig. 6A) and such waveforms were often encountered when electrode tips obviously passed from the endoderm into the transparent mesogloea. This suggests that they are due to electrotonic spread from the surface. In the preparation, almost no deflec was observed in the endoderm, implying that the endodermal positive phase anc slow negative phase of QSCPs have different origins. 386 KOHZOH OHTSU 0.5 mV 100 msec B-V r~ FIGURE 6. Amplitude changes of QSCPs associated with insertion depth of the recording electrode tip. Depth measurement \vas done with a single electrode. A, surface B, 40 jum ; C, 90 urn ; D, 190 Mm ; E, 290 Mm. Waveform changes of SMPs associated ivith electrode insertion depth To permit SMPs to be triggered by ligbts-off stimulation, preparations were moderately anesthetized with 1O1 M urethane. Since this concentration did not immobilize animals completely, the experiment had to be done quickly. Initially, two electrode tips were placed a few mm apart from each other (not as in Figure 5, where they were placed directly side by side). The results are shown in Figure 7. When both recording electrodes were placed on the surface of a radial streak, a pair of QSCPs and two pairs of SMPs appeared in both records (A). With the insertion of the electrode, very small negative deflections (B, upper trace) were seen. As the electrode was withdrawn the potentials reversed their polarity, in- creased in amplitude, and then became negative again, in a manner closely resem- bling Figure 5 (C-J). Similar results were obtained in 13 other experiments from seven animals. Positive pulses in the endodermal canal corresponded with QSCPs and SMPs and will be called PEPs. The maximum PEP was usually much larger than the maximum slow phase of QSCPs and the maximum SMP; in extreme cases without urethane treatment, the PEP reached about 4 mV, being four times larger than the corresponding SMP. The PEP sometimes showed a biphasic wave- form with a negative phase, suggesting that it is a conducted event. It is note- worthy that QSCPs and SMPs together show coupling with PEPs, except for the last PEPs in D and E and the last SMP in the upper trace of J. These might have been lost in the course of conduction. Pulses recorded form the ectoderm zvith high-resistance electrodes When recording electrodes with resistances of 20-50 MO were placed on the surface of the radial streaks and brief vibrations were delivered by means of a jolt- ing device (Tomita ct al, 1967). SMPs and the slow phase of QSCPs abruptly changed polarity immediately after a negative shift in the DC-level of 20-50 mV (mean, 37 mV ; measurements were done in 12 cases). Large, positive, long duration pulses (ectodermal large pulses), amplitudes of which usually ranged ELECTRICAL ACTIVITIES OF SPIROCODON 387 between 5 and 10 mV, were common in the 26 cases studied. They always cor- responded either to SMPs or to the slow phase of QSCPs (Fi It is quite clear that they originate in certain ectodermal cells of the radial streaks because the fine tips of electrodes used could not penetrate the opaque mes >ea without being damaged. To estimate membrane resistance changes during the pulses, current pulses were delivered through a balanced bridge circuit but unexpectedly, no changes were observed in four cells measured ( Fig. 8B). However, inje; ion of positive current of about 10 s A through the internal electrode often induced slow positive deflection similar to the ectodermal large pulses (Fig. 8D), resembling intracellularly recorded subthreshold pulses in the epithelio-muscular cell of Nanomia (Spencer, 1971). Pulses could not be evoked by delivering current extracellularly through such fine electrode tips. Even using electrodes with much larger tips of 20-30 /^m, current of 10~5 A was needed. Furthermore, resistances were measured before and after withdrawal of the inserted electrode and a mean difference of 9.3 Mfi (variation, 3-4 Mfi) was obtained in five measurements. Therefore, it appears likely that the ectodermal large pulses are recorded within cells having an excitable membrane. Similarly, the experimental results indicate that SMPs and the slow phase of QSCPs originate in the same kind of cells and can therefore be regarded as extracellular forms of the ectodermal large pulses. The s pmV 100msec B V q FIGURE 7. Waveform changes of SMPs (s) and QSCPs (q) associated with the depth of the recording electrode tip (upper trace). Lower trace is reference electrode on surface. A, surface; B, 460 /tm ; C, 360 /mi ; D, 310 /urn ; E, 210 /mi; F, 110 /mi; G, 60 /mi ; H, 40 /mi; I, 20 /im J, surface. Originating sites of SMPs often changed; in the first pair of . appeared first in the lower trace but in the second pair, the order was reversed, line shows possible pairs of pulses. All the pulses were evoked by terminating adapting ligh slightly before initiation of each recording. 388 KOHZOH OHTSU B FIGURE 8. Intracellular recordings from ectoderm and endoderm. A : recording from the ectoderm of a radial streak with a high-resistance electrode (lower trace) and simultaneous extracellular recording from the ectodermal surface (upper trace), respectively. B: result of the resistance-change measurement. Current pulses with duration of 7 msec were delivered at 33 Hz. C : recordings from the endoderm of a radial streak with a high-resistance electrode (lower trace) and simultaneous extracellular recording from the ectodermal surface (upper trace). D: pulse evoked by electric current through a high-resistance electrode penetrating intracellularly in the ectodermal layer. Note the summation of ectodermal large pulses in A and B. Arrow: artifact due to current injection. Horizontal bar, 100 msec; vertical bar, 1 mV in the upper traces of A and C, 5 mV in B and the lower traces of A and C, 12.5 mV in D and 3.4 Mfi in B. q = QSCP, s = SMP. two ectodermal large pulses occurring closely in time showed summation (Fig. 8A, B). Pulses recorded from the endoderm with high-resistance electrodes Since the fine tips of the high-resistance electrodes could not penetrate the opaque mesogloea, they were inserted into the endoderm of the radial streaks through cuts made at the bases of the tentacles. The insertion of electrodes with resistances of 80-100 Mfi sometimes resulted in positive pulses which were small, but somewhat larger than PEPs, showing amplitudes of 2-A mV (Fig. SC). Such pulses were obtained immediately after an abrupt negative shift of DC-level of 36-44 mV (mean, 40 mV in eight measurements). These pulses will be called endodermal large pulses. In Figure 8C, the SMP did not appear coupled with the endodermal large pulses because the inserted electrode was separated from the surface electrode by a distance greater than SMPs could travel, whereas in other records with shorter distances, SMPs and the endodermal large pulses often showed correspondence. Membrane resistance changes during the course of the endodermal large pulses could not be measured reliably because the high resistance of the electrodes resulted in unstable tip resistance. In one case where tip resistance was relatively stable, the resistance was measured before and after withdrawal of the inserted electrode and a difference of 13 MQ was obtained, suggesting that the endodermal large pulses are also of intracellular origin. DISCUSSION QSCPs and SMPs rapidly increase in amplitude as the electrode is withdrawn to within 100 /xm of the tissue surface and reach their maximum amplitudes at the ectodermal surface, suggesting that they are generated by ectodermal cells. A depth of 100 /xm is greater than the thickness of the ectodermal layer (10-50 /xtri). The effects may be attributed to a current-guide effect of the hole made in the ELECTRICAL ACTIVITIES OF SPIROCODON 389 tissues (especially in the opaque mesogloen) by the shank of the electrode during deep penetration. The quick phase and the slow phase of QSCPs appear to !u • their origins in different cells, because the quick phase can be isolated using urethai (Fig. 4E). On the other hand, the slow phase of QSCPs and SMPs becomes <• - towards the nerve ring while the quick phase of OSCPs gets larger. This ; the histological observation that in this region, the muscle process of the myo Dithelial cells become weaker and more nervous elements are observed. The slow phase of QSCPs and SMPs appear coupled with PEPs, and SMPs cannot usually cross over the groove between two radial streaks, where few myonal processes can be observed. This suggests that the muscle processes may generate the slow phase of QSCPs and SMPs; and the nervous elements, the quick phase of QSCPs. The latter seems more likely given the fact that the quick components have similar conduction velocities to pulses in the outer nerve-ring (nMPs. Ohtsu and Yoshida, 1973) and are usually coupled to them on a one-to-one basis. Indeed, a number of axons are observed among the myoepithelial cells. It is not clear if SMPs are triggered events or not. Speculating from the fact that SMPs occur at lights-off, they seem to be triggered by something that could be described as a light-information carrier system. SMPs can also originate spon- taneously without light stimuli. Non-nervous elements do not generally show re- peated spontaneous firings, whereas SMPs do. Furthermore, extremely small deflections with very short duration were sometimes observed in the subtentacular region (Fig. 9A). It is likely, therefore, that those pulses trigger SMPs. How- ever, recordings are not sufficient to permit definite conclusions. Another pos- sibility is that conduction of SMPs is linked with the conducting system of PEPs. 1 FIGURE 9. A : extremely small pulses (dots) recorded from the surface of a radial streak. B: simultaneous recordings from the exumbrellar epithelium near the ocelli (middle trace) and surface of a radial streak (lower trace). The first QSCP (q) and SMP (s) are spon- taneous. Note: epithelial pulses (EPs) can invade into the radial streak, where they have longer durations and larger amplitudes than in the exumbrella, but QSCPs and SMPs are not conducted to the exumbrella. C : simultaneous recordings from the endoderm (middle trace) and the ectoderm (lower trace) of a radial streak. The PEPs (middle trace) appear coupled with the QSCPs and the SMPs. The EPs were evoked by a single shock of « stimuli (arrows) delivered to the exumbrella and the QSCPs and the SMPs were indue lights-off (downward shifts of the upper traces in B and C). Horizontal bar, vertical bar, 0.1 mV in A, 0.4 mV in B and 1 mV in C. 390 KOHZOH OHTSU Although intracellular dye-marking was not feasible, it appears from Figure 8 that the ectodermal large pulses are recorded intracellularly from the cells which also generate SMPs and the slow phase of QSCPs extracellularly. Histological investigations have revealed at least three types of cells — myoepithelial cells, nerve cells, and nematocytes — in the ectodermal layer of the subtentacular region. The latter two seem not to be responsible for generating the ectodermal large pulses because the region used for experiment includes only a small number of nemato- cytes. and neuronal somata large enough to be penetrated intracellularly are rare. Penetrations regularly showing the ectodermal large pulses would not be expected with so few cells. The epithelial processes of the myoepithelial cells, which cover almost all the subtentacular region, are therefore the most probable sites of origin of the ectodermal large pulses. If the myoepithelial cell is penetrated by a micropipette electrode, then it is likely that the tip of the electrode usually lies in a large vacuole. This may explain why no resistance change is observed ; the membrane of the vacuole, which might have a high resistance, impedes the resistance change occurring in the cytoplasmic membrane from reaching the inside of the vacuole. Another possible reason for the lack of resistance changes is that some sensitive cells are actively excited while other, less sensitive ones receive passively conducted depolarizations spreading electronically from the former. The cell generating the ectodermal large pulses in Figure 8B might belong in the latter class, the loss of excitability being caused by the electrode impalement or some other factor. Electronic spread from other cells is possible if low-resistance intercellular pathways are present between the myoepithelial cells, as reported in other hydrozoan cells (Hand and Gobel, 1972; Mackie, 1976b; King and Spencer, 1979). Indeed, gap junctions are observed between muscle processes of Spirocodon. When two pulses, whether QSCPs or SMPs, appear close together in time, summation is often observed in the ectodermal large pulses (Fig. 8A). This is reminiscent of the observation that closely spaced large QSCPs can evoke a TCP (Fig. 4D). The ectodermal large pulses, therefore, may develop into much larger action potentials inducing tentacular contraction when the membrane potential of the cell, probably the myoepithelial cell, is raised to a certain level (a critical depolarization potential) due to summation. This can be seen in the dissociated ectodermal epithelial cells of Hydra, in which spontaneous oscillations develop into large spikes (Kass-Simon and Diesl, 1977), and in the epithelio- muscular cell of the siphonophore Nanoinia. where subthreshold pulses are thought to develop electrical responses with larger amplitude, causing twitch responses (Mackie, 1976a). The mechanism generating PEPs is puzzling. The reversed polarity of PEPs and ectodermal slow pulses (Figs. 5, 7) suggests that the two are in a sink-and- source relationship in an electrogenerative system. Arguing against this idea is the long distance between the ectoderm, where SMPs (and the slow phase of QSCPs) are observed, and the central region of the endoderm, where the maximum PEPs are observed. Moreover, the opaque mesogloea between the ectoderm and the endoderm may function as a current barrier. The opaque mesogloea, however, is perforated in places with a tissue resembling an endodermal cell process, per- mitting endodermal cells direct contact with ectodermal cells. It is possible, there- fore, that depolarizations of ectodermal cells reach electronically peripheral processes of endodermal cells and then are conducted electrogenerativly to the central region, generating PEPs. ELECTRICAL ACTIVITIES OF SPIROCODON 391 The endodermal large pulses appear from Figure 8C to be recorded intra- cellularly from endodermal cells which also generate PEPs extracellularly. The small amplitude of the endodermal large pulses, however, argues against regarding them as intracellular recordings. The small amplitude is explicable if it is assumed (a) that the pulses recorded are clue not to active excitation of the endodermal cells penetrated, but to electronic spread from other actively :cited endodermal cells or (b) that recordings are from vacuoles. PEPs sometimes show diphasic waveforms followed by negative deflections, suggesting that endodermal cells are excitable and conductive. If so, PEPs may be conducted back to ectodermal cells and cause small deflections. A similar case of endodermal pulses evoking pulses in ectodermal cells electrotonically has been reported in the siphonophore, Nanoinia (Mackie, 1976a). The pulses of Spirocodon are comparable to electrical potentials described in several hydrozoan medusae (e.g.. Mackie and Passano, 1968; Spencer, 1975; Mackie, 1975, 1976a ; Spencer, 1978). As described above, the quick phase of QSCPs is best interpreted as a nervous event associated directly or indirectly with the outer nerve-ring, and the slow phase as an electrical concomitant of myo- epithelial cell activity, triggered by the quick phase. In Spiricodon QSCPs (and SMPs) occurring singly are not usually accompanied by tentacular contractions, as far as can be observed under a dissecting microscope. This is different from Proboscidactyla (Spencer, 1975) but similar to Sarsia (Passano ct al, 1967). Tentacular contraction can be induced, however, when two pulses are close together in time, probably by summation or facilitation of two slow phases. The slow phase of QSCPs decreases or disappears in excess Mg2+ and becomes larger stepwise by successive electrical stimuli, especially in a relatively depressed condition. In this respect, QSCPs obviously resemble the tentacle pulse (TP) in Stomotoca described by Mackie (1975) ; the quick phase appears to correspond to the pretentacle pulse (PTP) of Mackie (1975). Spencer (1978) has reported the TP without initial quick phase in the tentacles of Polyorchis. Also in Spirocodon, the quick phase was often extremely small, especially in tentacles, but QSCPs were readily distinguishable from SMPs and TCPs because the former was conducted only locally and the latter showed extremely slow conduction associated with muscle contraction. Large specimens of Spirocodon with umbrellar diameter of 4-7 cm, as used in this study, show almost no crumpling in response to pricking the exumbrellar ecto- derm with fine forceps. But such stimulation can evoke epithelial pulses (Ohtsu and Yoshida, 1973) similar to the CrPs of Spencer (1975, 1978). Small specimens of about 1 cm clearly show the crumpling response. Epithelial pulses distinct from SMPs and QSCPs (Fig. 9B, C) can also invade the subtentacular region, as in Sarsia. The pulses are larger in the endoderm than the ectoderm (Fig. 9C) but their generating origin remains to be investigated. I should like to thank Professor G. O. Mackie of University of Victoria, Canada, and Dr. A. N. Spencer of University of Alberta, Canada, for their help in preparing this manuscript. Work supported in part by grants-in-aid to M. Yoshida from the Ministrv of Education. j SUMMARY 1. Three types of cells — myoepithelial cells, nerve cells, and nematocytes found in the ectoderm of the subtentacular region except near the nerve ring. 392 KOHZOH OHTSU single type of endodermal cell is identified in the endoderni of the subtentacular region except near the ring canal. 2. Three types of conductive pulses — QSCPs, SMPs, and TCPs — are recorded in the ectoderm. When recording electrodes are inserted into the endoderni, SMPs and the slow phase of QSCPs (negative pulses) are replaced by PEPs (positive pulses). 3. Using high-resistance electrodes, two types of pulses (ectodermal and endo- dermal large pulses) are recorded (presumably intracellularly) from the ectoderm and the endoderm, respectively. They probably originate in the myoepithelial cells and in the endodermal cells, respectively. 4. SMPs and the slow phase of QSCPs occur coupled with the ectodermal large pulses and appear to be generated in the same cells, as the extracellular forms of the latter. The quick phase of QSCPs is probably an event generated by nerves in the ectoderm. 5. PEPs occur coupled with the endodermal large pulses and may be generated in endodermal cells as the extracellular forms of the endodermal large pulses. 6. Pulses with extremely small amplitudes suggesting the presence of another conducting system can be recorded. 7. SMPs and QSCPs are different from the crumpling pulses of Spencer (1975). LITERATURE CITED BOELSTERLI, U., 1977. An electron microscopic study of early developmental stages, myo- genesis, oogenesis and cnidogenesis in the anthomedusa, Podocorync carnca M. Sars. /. Morphol., 154 : 259-290. HAND, A. R., AND S. GOBEL, 1972. The structural organization of the septate and gap junc- tions of Hydra. J. Cell Biol.. 52: 397-408. HORRIDGE, G. A., 1955. The nerves and muscles of Medusae. II. Gcryonia proboscidalis Eschscholtz. J. Exp. Biol.. 32 : 555-568. JOSEPH SON, R. K., 1965. Three parallel conducting systems in the stalk of a hydroid. /. E.vp. Biol., 42: 139-152. JOSEPH SON, R. K., 1967. Conduction and contraction in the column of Hydra. J. Exp. Biol., 47: 179-190. JOSEPHSON, R. K., 1974 Factors affecting muscle activation in the hydroid Tubularia. Biol. Bull., 147 : 594-607. JOSEPHSON, R. K., AND M. MACKLIX, 1967. Transepithelial potentials in hydra. Science, 156: 1629-1631. JOSEPHSON, R. K., AND G. O. MACKIE, 1965. Multiple pacemakers and the behaviour of the hydroid Tubulana. J. Exp. Biol., 43: 293-332. JOSEPHSON, R. K., AND N. B. RUSHFORTH, 1973. The time course of pacemaker inhibition in the hydroid Tubulana. J. Exp. Biol., 59: 305-314. KASS-SIMON, G., AND V. K. DIESL, 1977. Spontaneous and evoked potentials from dissociated epithelial cells of Hydra. Nature, 265 : 75-77. KING, M. G., AND A. N. SPENCER, 1979. Gap and septate junctions in the excitable endoderm of Polyorchis pcnicillatus (Hydrozoa, Anthomedusae). /. Cell Sci., 36: 391-400. KKUIJF, H. A. M., 1977. Bursting pacemaker activity in the solitary hydroid Tubularia soli- taria. J. Exp. Biol.. 68 : 19-34. MACKIE, G. O., 1975. Neurobiology of Stomotoca. II. Pacemakers and conduction pathways. /. NcurobioL, 6: 357-378. MACKIE, G. O., 1976a. The control of fast and slow muscle contractions in the siphonophore stem. Pp. 647-659 in G. O. Mackie, Ed., Coelcntcratc ecology and behavior. Plenum Press, New York. MACKIE, G. O., 1976b. Propagated spikes and secretion in a coelenterate glandular epithelium. J. Gen. Physiol., 68 : 313-325. MACKIE, G. O., 1978. Coordination in physonectid siphonophores. Mar. Bchai'. Pltvsiol., 5 : 325-346. ELECTRICAL ACTIVITIES OF SPIROCODON 393 MACKIE, G. O., AND L. M. PASSANO, 1968. Epithelial conduction in Hydromeclusae J Gen Physio!., 52: 600-621. MACKIE, G. O., AND C. L. SINGI.A, 1975. Neurobiology of Stoiuotoca. I. Action systems /. Ncurobiol.. 6: 339-356. MACKLIN, M., AND R. K. JOSEPHSON, 1971. The ionic requirements of ir, Hal poten- tials in Hydra. Biol. Bull., 141 : 299-318. OHTSU, K., AND M. YOSHIDA, 1973. Electrical activities of the anthomedusan, Spin 'ta- trix (Tilesius). Biol. Bull., 145 : 532-547. PASSANO, L. M., AND C. B. McCuLLOUGH, 1962. The light response and the rhythmic IM/TIT of hydra. Proc. Nat. Acad. Sci. U. S. A., 48 : 1376-1382. PASSANO, L. M., AND C. B. McCuLLOUGH, 1964. Co-ordinating systems and behaviour in H\dra. I. Pacemaker system of the periodic contractions. J '. Exp. Biol., 41 : 643-664. PASSANO, L. M., AND C. B. McCuLLOUGH, 1965. Co-ordinating systems and behaviour in Hydra. II. The rhythmic potential system. /. Exp. Biol., 42: 205-231. PASSANO, L. M., G. O. MACKIE, AND M. PAVANS DE CECCATTY, 1967. Physiologic du com- portement de Fhydromeduse Sarsia tuhulosa Sars. Les systemes des activites spon- tanees. C. R. Acad. Sci. (Paris), 264 : 614-617. RUSHFORTH, N. B., 1971. Behavioral and electrophysiological studies of hydra. I. Analysis of contraction pulse patterns. Biol. Bull., 140 : 255-273. RUSHFORTH, N. B., AND D. S. BURKE, 1971. Behavioral and electrophysiological studies of hydra. II. Pacemaker activity of isolated tentacles. Biol. Bull., 140: 502-519. SINGLA, C. L., 1978. Fine structure of the neuromuscular system of Polyorchis pcnicillatus ( Hydomedusae, Cnidaria). Cell Tissue Res.. 193: 163-174. SPENCER, A. N., 1971. Myoid conduction in the siphonophore Nanotnia bijuga. Nature, 233 : 490-491. SPENCER, A. N., 1975. Behavior and electrical activity in the hydrozoan Proboscidactyla flavicirrata (Brandt). II. The medusa. Biol. Bull., 149: 236-250. SPENCER, A. N., 1978. Neurobiology of Polyorchis. I. Function of effector systems. J. Ncurobiol., 9 : 143-157. SPENCER, A. N., 1979. Neurobiology of Pol\orchis. II. Structure of effector systems. J. Ncurobiol., 10: 95-117. Ton, Y., M. YOSHIDA, AND H. TATEDA, 1979. Fine structure of the ocellus of the hydro- medusan, Spirocodon saltatrix. I. Receptor cells. /. Ultrastruct. Res., 68 : 341-352. TOMITA, T., A. KANEKO, M. MURAKAMI, AND E. L. PAUTLER, 1967. Spectral response curves of single cones in the carp. I'ision Res.. 7 : 519-531. UCHIDA, T., 1927. Studies on Japanese Hydromedusae. I. Anthomedusae. /. Fac. Sci. Imp. ' Univ. Tokvo, Section IV, 1: 145-241. Reference : Biol. Bull.. 159: 394-401. (October, 1980) EFFECT OF SYMBIOTIC ZOOXANTHELLAE AND TEMPERATURE ON BUDDING AND STROBILATION IN CASSIOPEIA ANDROMEDA (ESCHSCHOLZ) M. RAHAT AND ORIT ADAR * Department of Zooloyy, The Hebrew' University, Jerusalem, Israel Many algal-invertebrate symbioses are based on mutual biotrophic benefit. The heterotrophic host is supplied with photosynthates from its phototrophic symbiont, and the latter obtains both nutrients and an optimal habitat (Smith et a/., 1969; Cook, 1972; Richmond and Smith, 1979). Symbiotic associations were probably initiated by a chance meeting of the cosymbionts. A loose faculative symbiosis could evolve in time into a symbiosis obligatory for at least one of the partners. Such symbioses can affect both form and function of the cosymbionts, and through coevolution new species may have evolved (Margulis, 1976). The symbiogenic theory of the origin of mitochondria and chloroplasts is based on the above assumptions. To illustrate such evolutionary trends, the following examples are usually cited: The platyhelminth Amphiscolops langcrhansi is unable to achieve sexual maturity without its symbiotic algae (Taylor, 1971), and the closely related flat- worm Convoluta roscoffcnsis is completely dependent on its symbiotic algae "for its bulk nutrients and growth stimuli" (Provasoli ct a!., 1968). It has also been claimed that the scyphozoan medusa Mastigias papna will strobilate ephyrae only if infected with zooxanthellae, and "if a scyphistoma is deprived of its zooxanthellae, it can never strobilate" (Sugiura, 1964). Ludwig (1969) stated similarily that aposymbiotic scyphistomae of Cassiopeia andromcda obtained in his laboratory "have only a limited vitality, and cease their strobilation completely". The above statements allege an obligate dependence of the hosts on their sym- biotic algae. It is thus of interest that all the above invetebrates form aposymbiotic eggs and larvae, and the symbiotic algae are acquired anew at each generation. In view of this, we reexamined the life cycle of C. andromcda. In this study we report strobilation and formation of aposymbiotic ephyrae by aposymbiotic scyphistomae, and the effect of temperature on budding and strobilation. MATERIALS AND METHODS Mature Cassiopeia medusae were collected from the Gulf of Eilath, brought to the laboratory within 8 hr, and maintained in aquaria at 18° ± 2°C. During transport and in the aquaria, the medusae released many aposymbiotic planulae. Planulae left in the aquaria, or maintained in water from the latter, developed into polyps hosting algae (Fig. la). In addition to the polyps obtained from medusae collected at Eilath, polyps were also obtained from a culture maintained in our department for several years. * This study is part of a Master's thesis submitted by the junior author to the Hebrew University. 394 STROBILATION IX CASSIOPEIA 395 FIGURE 1. Polyps of C. androincda. a. Symbiotic polyp. The dark dots in the calyx are endocellular zooxanthellae. b. Aposymbiotic polyp with three buds. Scale bar : 1 mm. To obtain aposymbiotic polyps (Fig. Ib), planulae were collected from the aquaria, and washed and maintained in sterile sea water, in crystallizing dishes or finger bowls. Aposymbiotic polyps were also obtained by incubation of symbiotic polyps in 10-('M DCMU (3,3,4-dichlorophenyl. 1.1-dimethyl urea), for 35-40 days. Polyps were fed 3x a week to repletion, with 2-4-day-old Arteniia salina larvae. The cultures were washed and media changed within 24 hr after feeding. Sea water (from the Mediterranean, salinity 37-39/{c) was sterilized in an autoclave. Aposymbiosis of the various forms of Cassiopeia was verified by fluorescence microscopy. Stock cultures of polyps were maintained at 20° ± 2°C under con- tinuous illumination of 3-5 W/nr from white fluorescent lamps. Crude homogenates of Cassiopeia were prepared from the oral lobes and sub- umbrellae of adult medusae, using a glass homogenizer. To reinfect aposymbiotic polyps, about 20 p\ of fresh homogenate was injected into the polyp's mouth using a drawn-out Pasteur pipette. For "feeding" experiment controls, the same homogenates were heated for 30 min in a water bath at 45 °C and stored — 18°C until used. Some homogenates were deep-frozen directly to without prior heating. 396 150 100 3 CO 50 M. RAHAT AND O. ADAR 25r g O OO OO O O O ODOOOO AAAA A A A lit *M B 20 15 O X 10 oo A oo OO A OOO AAA A A AAA A o A A A A O A o A A O OO 20 40 60 80 0 Days 20 40 60 80 250 200 • 17°-18°C o 20°- 22° A 25°- 26° A 28°- 30° 150 o A A " 7 0 ' r A •o 3 CD 100 A m A 6 o 0 A A ° • >4 A a A 50 : •* 2 AA 1 A A o ' O i i i ~ i » 20 40 Da y s 20 40 80 FIGURE 2. Effect of temperature on budding and strobilation of symbiotic and apo- symbiotic C. andromcda. N = 7. A and B, symbiotic polyps. C and D, aposymbiotic polyps. RESULTS Budding and strobilation of symbiotic polyps Groups of seven symbiotic polyps were placed in crystallizing dishes at four different temperature ranges between 17° and 30°C. The numbers of buds and STROBILATION IN CASSIOPEIA 397 TABLK I Budding and strobilation of C. andromeda under various conditions. Polyps °C Buds/polyp Ephyrae /polyp Symbiotic* 17°-18° 19.6 (N = 7) 20°-22° 3.1 3.4 25°-26° 1.0 2.8 28°-30° 0.3 3.1 Aposymbioticf 17°-18° 25.7 0 (N = 7) 20°-22° 34.3 0.1 25°-26° 29.1 1.0 28°-30° 23.1 0.4 Reinfected 20°-22° 2.0 2.0 (N = 10) Injected 20°-22° 62.5 1.0 (N = 10) * Some symbiotic polyps formed up to five ephyrae in succession. t Aposymbiotic polyps never formed more than one ephyrae. For time-rate of budding and strobilation, see Figures 2 and 4. ephyrae formed at each temperature were recorded, as shown in Figure 2 (A and B) and Table I. At higher temperatures budding was fully arrested after 20-25 days, and only ephyrae were formed. Strobilation was about equal at all temperatures between 20° and 30° C. Sometimes, budding and strobilation occurred simul- taneously, as was noted by Hoffmann et a/., (1978). Budding and strobilation oj aposymbiotic polyps Experimental conditions for aposymbiotic polyps were identical to those described above for symbiotic polyps. Budding was equally frequent at all temperatures tested, and aposymbiotic ephyrae were formed at the higher tempera- tures (Fig. 2B, C; Table I). In this and similar experiments, a total of 26 aposymbiotic ephyrae (Fig. 3) were obtained from 130 aposymbiotic polyps at temperatures above 25° C. Only two ephyrae were obtained from several hundred aposymbiotic polyps below this temperature. During strobilation, a yellow-green color appeared in the upper part of the calyx of the aposymbiotic scyphistomae. This color was also observed in the free-swimming young ephyrae, but it disappeared 7-10 clays after their release. Effect of reinfection and homogenate injection Ten aposymbiotic polyps previously maintained at 18° C were reinfected with zooxanthellae by injection of a fresh homogenate into their coelenterons. The polyps were then maintained at 20°-22°C (Fig. 4A). Within 5 days, endosym- biotic algae could be seen with the naked eye. Strobilation could be seen on about the 10th day, and on the 18th day two ephyrae were released. All buds and ephyrae formed by the reinfected polyps contained zooxanthellae. Ten similar aposymbiotic polyps maintained at 20°-22°C were injected every 2 days with a homogenate containing heat-killed algae (Fig. 4B). Strobilatic was first observed on the 34th day, and the first ephyrae released on the day. The above experiments are summarized in Table I. In these and in : "feeding" experiments, a total of 17 ephyrae and 2025 buds were obtaine< 398 M. RAHAT AND O. ADAR FIGURE 3. Symbiotic and aposymbiotic ephyrae of C. andromeda. a. Algae-hosting ephyrae : the dark central area is heavily infected with zooxanthellae. b. Enlarged margin of same : zooxanthellae are seen as bright dots through the transparent umbrella, on dark back- ground, c. Aposymbiotic ephyrae : note transparent central area. The dark oral lobes contain a violet-blue or brown pigment, d. Enlarged margin of same : note transparent umbrella on dark background, and white clusters of nematocytes. In a and c, milimetric scale. 44 injected aposymbiotic polyps. Strobilations were also obtained in experi- ments in which homogenates with freeze-killed algae were used. There was no reinfection in any of the polyps. DISCUSSION This study shows that elevated temperatures and endosymbiotic algae act synergistically to induce strobilation in C. andronicda. The direct effect of temperature on morphogenesis has been observed in both symbiotic and aposymbiotic polyps. In the former, there was a marked preference for bud formation at temperatures of 17°-18°C. At higher temperatures budding ceased completely and only ephyrae were formed. In the aposymbiotic polyps, budding rate was similar at all tested temperatures between 17° and 30°C. Strobi- lation, however, was definitely temperature dependent, and occurred almost only at above 25 °C (Fig. 2, Table I). Such a requirement for elevated temperatures to induce strobilation has also been reported for the symbiotic scyphozoan medusa Mastigias papua (Sugiura, STROBILATION IN CASSIOPEIA 399 1965), and for the nonsymbiotic Chrysaura quinquecirrha (Loeb, 1972). In both cases, precooling to 20° C for several weeks was required for later strobila- tion at 22° and 26° C, respectively. The stock cultures of pc -ed in our experiments were maintained at 20° ± 2°C, but we did not thoroi%;': >vestigate whether such precooling is required also for C. andronicda. The morphogenic effect of endosymbiotic algae was shown lower temperatures at which strobilation occurred in the symbiotic polyps, by the relative number of ephyrae and buds formed at each temperature (Table I). The lowest temperature at which aposymbiotic strobilation occurred was 20°-22°C. Symbiotic polyps formed some ephyrae even at 17°-18°C. We therefore looked for a strobilation-enhancing factor (s) in the endosymbiotic algae. Reinfection of aposymbiotic polyps enhanced strobilation at 20°-22°C. Twenty ephyrae were formed in 60 days by 10 reinfected polyps (Fig. 4A), vs. only one ephyrae formed at the same temperature from seven aposymbiotic polyps after 73 days (Fig. 2D). A similar though less marked effect on strobilation was obtained by injecting preheated or frozen homogenates into aposymbiotic polyps. In one experiment (Fig. 4B), 10 ephyrae were formed by 10 aposymbiotic polyps in 60 days. At the same time, however, 600 buds were formed. This relatively high rate of budding correlates with the high budding rate shown in Figure 2C for aposymbiotic polyps. In all these experiments (Figs. 2, 4), initiation of strobila- tion was accompanied by a synchronous decrease in budding rate. The lag time to the release of the first ephyrae was much shorter in the symbiotic polyps than in the aposymbionts (Fig. 2). A similar relation was found between the reinfected polyps and the aposymbionts injected with homogenates (Fig. 4). From the results obtained in our study, we suggest that morphogenic processes in C. andromeda shift from budding to strobilation at relatively higher metabolic rates. 30r 20 TJ 3 1O 2O o 10 • 600 500 400 300 200 100 B 20 o oo n 10 o o 4O 60 20 40 60 Days FIGURE 4. Formation of buds and ephyrae by aposymbiotic polyps. N - 22°C. A. Polyps reinfected with zooxanthellae on day O. B. Polyps injected with hea homogenate of symbiotic mature medusae. 400 M. RAHAT AND O. ADAR A higher metabolic rate can be obtained either by a rise in ambient temperature, or by a factor(s) that enhances metabolism. Such a factor is formed by Cassiopeia polyps, and it is probable that its accumulation enables strobilation of aposymbionts even at low temperatures. The formation of this factor by Cassiopeia is facilitated by a metabolite formed by the zooxanthellae. This algal metabolite is non- specific, as it is present also in non-strobilating symbiotic adult medusae. Further experiments are required to elucidate quantitatively and qualitatively the characteristics and effects of the proposed factor (s) and its relation to effects of temperature and the metabolites of the algal symbionts. The ecological implications of the above described combined effects of tem- perature and endosymbiotic algae can only be speculated upon. In its natural habitat in the Gulf of Eilath, C. andromeda can be found along the open coast, as well as in closed lagoons and mangroves. In the open coast the yearly tem- perature fluctuation is 17°-27°C, but in lagoons and mangroves a range of 13°-33°C has been recorded (For ct a/., 1977). Budding would thus be enhanced in winter, while at the higher summer temperatures ephyrae would be formed. In the sea, no aposymbiotic C. andromeda have yet been reported. By the ease at which algal infections occur in these polyps, it is doubtful whether aposymbiotic specimens ever occur in their natural habitat. We do not know yet if aposymbiotic ephyrae can grow into sexually mature medusae, and thus complete an aposymbiotic life cycle in this species. In pre- liminary experiments we obtained a twofold increase in diameter of aposymbiotic ephyrae (from 5.1 to 11.5 mm), accompanied by normal morphological development. Our observations on aposymbiotic strobilation in C. andromeda raise some questions with regard to the evolution of obligatory symbioses. Loss of capacity in a host to complete its life cycle in absence of its symbionts would indicate an evolutionary genetic integration in the sense described by Margulis (1976). It would thus be possible to claim that we have at our hands a new species, made up respectively of animal and algal tissues. Such an integration, however, should express itself also in the capability of the host to directly inherit its endosymbionts through the germ cells. In C. andomeda, as well as in Mastigias papua, Amphi- scolops langcrhansi, and Convoluta roscoffcnsis, this is not the case. Aposymbiotic larvae are formed by these species, and the symbiotic algae are acquired anew at each generation. On the basis of our results that strobilation in C. andromeda is not dependent upon a symbiont, vs. the hypothesized obligatory dependence of the above species on their respective endosymbiotic algae, it would be of interest to reexamine the algae for the extent of host-symbiont integration. Such a study would add to our understanding of the effect of symbiosis on evolution. SUMMARY The role of the zooxanthellae in the life cycle of Cassiopeia andromeda was reexamined. Symbiotic and aposymbiotic polyps were maintained at various temperatures between 17° and 30°C, and numbers of buds and ephyrae formed at each temperature were recorded. The number of buds formed by the symbiotic polyps was inversely related to temperature. After 20 days at above 20° C, only ephyrae were formed. In aposymbiotic polyps, buds were formed at all temperatures tested. Strobila- tion occurred above 25 °C, and aposymbiotic ephyrae were obtained. Aposymbiotic STROBILATIOX IX CASSIOPEIA 4()1 polyps reinfected with zooxanthellae formed ephyrae at 22°C. A similar effect was obtained by repeatedly injecting aposymbiotic polyps with heat- or cold-inacti- vated homogenate derived from adult medusae. No reinfection occurred in the latter. It is concluded: 1. At higher metabolic rates, strobilation supers' , budding in C. androineda. 2. A higher metabolic rate is obtained at elevated tenr >eratures, and by a biotrophic effect of the symbiotic zooxanthellae. 3. Symbiotic zooxanthel- lae are not essential for strobilation in C. androineda. LITERATURE CITED COOK, C. B., 1972. Benefit to symbiotic zoochlorellae from feeding by green hydra. Biol. Bull., 142: 236-242. HOFFMANN, D. K., R. NEUMANN, AND K. HENNE, 1978. Strobilation, budding and initiation of scyphistoma morphogenesis in the rhizostome Cassiopeia androineda (Cnidaria: Scyphosoa). Mar. Biol. 47: 161-176. LOEB, M., 1972. Strobilation in the Chesapeake Bay sea nettle Chrysaura quinquecirrha. I. The effects of environmental temperature changes on strobilation and growth. /. Exp. Zool. 180 : 279-292. LUDWIG, F. D., 1969. Die zooxanthellen bei Cassiopeia andromcda Eschscholz 1829 (polyp- stadium) und ihre bedeutung fiir die strobilation. Zool. Jb. Abt. Anat. Ontog. Tiere 86: 238-277. MARGULIS, L., 1976. Genetic and evolutionary consequences of symbiosis. Exp. Parasitol. 39: 277-349. POR, F. D., I. DOR, AND A. AMIR, 1977. The mangal of Sinai : Limits of an ecosystem. Hclgol. Wiss. Mcercsuntcrs., 30: 295-314. PROVASOLI, L., T. YAMASU, AND I. MANTON, 1968. Experiments on the resynthesis of sym- biosis in Convoluta roscoffcnsis with different flagellate cultures. J. Mar. Biol. Assoc. U. K. 48 : 465^79. RICHMOND, M. H., AND D. C. SMITH, Eds., 1979. The cell as a habitat. The Royal Society, London. SMITH, D. C., L. MUSCATINE, AND D. H. LEWIS, 1969. Carbohydrate movement from auto- trophs to heterotrophs in parasitic and mutulistic symbiosis. Biol. Rev. 44: 17-90. SUGIURA, Y., 1964. On the life history of rhizostome medusae. II. Indispensability of zooxanthellae for strobilation in Mastigias papua. Embryologia, 8 : 223-233. SUGIURA, Y., 1965. On the life history of rhizostome medusae. III. On the effects of tem- perature on the strobilation of Mastigias papua. Biol. Bull., 128 : 493-496. TAYLOR, D. L., 1971. On the symbiosis between Amphidinium klcbsri (Dinophyccae) and Ainphiscolops langcrlmnsi (Turbcllaria: Acocla). J. Mar. Biol. Assoc. U. K. 51: 301-313. Reference : Bio!. Bull.. 159: 402-417. (October, 1980) THE BEHAVIORAL BASIS OF LARVAL RECRUITMENT IN THE CRAB CALLINECTES SAPIDUS RATHBUN: A LABORATORY INVESTIGATION OF ONTOGENETIC CHANGES IN GEOTAXIS AND BAROKINESIS l S. D. SULKIN, W. VAN HEUKELEM, P. KELLY, AND L. VAN HEUKELEM Horn Point Environmental Laboratories, Center for Environmental and Estuarine Studies, University of Maryland, Cambridge, Maryland 21613 Estuarine invertebrates that have planktonic larvae must maintain their popula- tions in the face of net seaward flow of water. One mechanism for this is active retention within the estuary of sufficient off spring to at least replace the population ( Sandif er, 1 975 ; Scheltema, 1 975 ) . Retention mechanisms that combine behavioral adaptations with characteristic estuarine circulation have been described for larvae of several estuarine inverte- brates, including barnacles and mud crabs (Bousfield, 1955), oysters (Carriker, 1951; Wood and Hargis, 1971) and decapods (Sandifer, 1975). However, it is unlikely that all planktonic larvae will be retained in the face of net downstream flow of water. Indeed, larvae of many estuarine crabs are present in the waters of the continental shelf off the east coast of North America (Nichols and Keney, 1963 ; Sandifer, 1973; Dudley and Judy, 1971). Among the most common are species of the genus Callincctcs, including the blue crab CaUincctes sapidus Rathbun (Sandifer, 1975; Goy, 1976). C. sapidus inhabits entire estuaries as an adult, but spawns only in high salinity regions near estuary mouths (Hopkins, 1943; Van Engel, 1958). These cir- cumstances increase the probability of loss of larvae from the estuary. It is critical to an understanding of the recruitment process, and hence population dynamics, to determine whether larvae exported from estuarine spawning grounds to ocean waters represent a significant loss from the adult habitat and, if so, whether they may be recruited back to it in significant numbers. Mechanisms for larval retention within an estuary or exchange between an estuary and its adjacent coastal region depend upon characteristic circulation (Pritchard, 1952; Bousfield, 1955; Scheltema, 1975). Because currents vary in direction and rate with depth, vertical distribution of larvae is significant in de- termining their ultimate destinations. For some crab species, there is evidence that vertical distribution of larvae varies predictably with age (Bousfield, 1955; Sandifer, 1975; Goy, 1976). There is further evidence that these vertical dis- tribution patterns result from oriented movement controlled by behavioral responses to exogenous stimuli (Sulkin, 1973, 1975; Forward and Costlow, 1974; Ott and Forward, 1976; Latz and Forward, 1977). Any C. sapidus adaptation for retention or exchange would be likely to be mani- fest in behavioral patterns that regulate vertical distribution through ontogeny. We have examined the behavioral responses of C. sapidus at various larval stages to stimuli likely to influence vertical movement. We report here the results of experi- 1 Contribution No. 1094 from the Center for Environmental and Estuarine Studies, University of Maryland. 402 BEHAVIORAL ECOLOGY OF CRAB LARVAE 403 merits on orientation and swimming rate in response to gravity, hydrostatic pressure, temperature, and salinity. The results are evaluated in term of their effects on vertical distribution of C. sapidus larvae throughout their >ntogenv and consequent patterns of dispersal of this species in the estuarine am ital marine environments. MATERIALS AND METHODS Larval culture Ovigerous blue crabs were obtained from lower Delaware Bay and were taken to the laboratory with egg masses intact. Eggs were stripped from the females when they reached the black eyespot stage, or immediately if they were collected at that stage. Approximately 500 eggs from a single female were then placed in each of several 125 ml Erlenmeyer flasks containing 50 ml of filtered seawater at 30^o salinity (S) and 20°-30°C. Culture water contained either penicillin (60 mg/1) and streptomycin (50 mg/1) or chloramphenicol at 5 mg/1. Unpublished data show that use of antibiotics improves egg and larval survival without altering larval behavior. Eggs were transferred to fresh medium every second day. Upon hatching, larvae from a given female were placed in several glass culture dishes containing 100 ml of 30/^f- S seawater. The standard culture temperature was 25 °C, although experiments occasionally dictated acclimation of larvae to 15°C. Each dish initially contained approximately 200 zoeae. Zoeae were transferred to fresh medium daily and fed either the rotifer Brachionus plicatilis Muller (during the first 14 days of development) or freshly- hatched nauplii of the brine shrimp Artcmia salina L. (after clay 15 ; Sulkin, 1978). Records of mortality and molting were kept. Samples of larvae were removed for testing as experiments dictated. In order to avoid complicating effects of rhythmicity in locomotory activity (Sulkin et al., 1979), all behavior experiments were run during afternoon hours. Experiments were conducted on zoeal stages I, IV, and VII. Sinking rates Larvae of various stages were narcotized using 3c/r ethyl carbamate. Individuals were timed as they sank vertically through a 10-cm-long space marked off midway along a vertical cylinder filled with seawater (30%c S; 25 C). At least 50 indi- viduals from three different broods were tested. Data are presented as mean sinking rates (cm/sec.). Differences among stages were tested by analysis of variance. Geotaxis Typically the presence and sign of geotaxis in crab larvae has been determined by noting changes over time in net distribution of organisms placed in a vertically oriented test chamber in total darkness (Sulkin, 1973; Ott and Forward, 1976; Latz and Forward, 1977). Oriented movement upward has been termed negative geotaxis ; downward, positive geotaxis. To account more effectively for non-oriented or random movement that occur simultaneously with oriented movement or geotaxis, the following experiment design was used. Thirty to fifty larvae were drawn at random from the and divided between two test chambers, one positioned vertically ; the 404 SULKIN, VAN HEUKELEM, KELLY, AND VAN HEUKELEM zontally. Each chamber was constructed of 3-mm-thick transparent Incite, mea- sured 25 X 5 X 5 cm, and was marked off into four sections of equal length. Larvae were placed at the bottom of the vertical experimental chamber and at one end of the horizontal control chamber. At 10-min intervals, the larvae in each quadrant were counted. A dim deep-red light was used to silhouette the larvae when counts were made. This procedure did not appear to disrupt the positions of larvae. Change in distribution over time in both horizontal and vertical chambers was analyzed by construction of a "kite" diagram, which shows the proportion of total sample in each quadrant. Movement of larvae away from the end of the horizontal chamber was con- sidered random or non-oriented and thus served as a control against which to compare the distributional change which occurred in the vertical chamber. Ad- mittedly, the control was imperfect in that it did not account for sinking, which occurred intermittently in the vertical chamber. Nevertheless, movement along the axis from the initial point source in the vertical chamber exceeding that for the horizontal chamber can be attributed to oriented response — in this case, negative geotaxis. On the other hand, movement in the horizontal control exceeding that in the vertical tank can be attributed to positive geotaxis. If there is no dif- ference between the two, the presumption is that no oriented response can be attributed to geotaxis. Distributions after 30 min were analyzed. This period proved long enough to separate oriented from random movement while providing for stable distribution in the vertical chamber and random dispersal in the horizontal chamber. Larvae used in these experiments were acclimated to darkness for at least 2 hr prior to testing and not fed during the experiment. For each larval stage, experiments were conducted at 25, 30, and 35/^ S in the following fashion. A group of larvae was first tested at 3Q'/co S. Then they were placed in test chambers which contained either 25(/co S seawater or 35^( S seawater, and dis- tribution shift experiment was repeated. A minimum of three replicates with off- spring from three different crabs was conducted for each salinity. The experimental design permitted statistical analysis as well as the qualita- tive analysis typically used in behavioral experiments. For each experimental and control replicate, an average distribution at 30 min ("mean position value") was calculated by assigning weights from 0-3 to the quadrants (the end to which the larvae were initially added == 0), multiplying the weights by the number of larvae in the quadrants, and dividing the product by the total number of larvae. The mean position value is a quantitative descriptor of the distribution and provides a commonly derived set of data for all experimental and control replicates. For each salinity, the mean position values for all experimental replicates were com- pared against those for all control replicates using the non-parametric Mann- Whitney U test. Swimming rate Figure 1 shows the apparatus used to measure larval swimming rates. A sample of larvae from a particular brood at a specified stage of development was placed in the behavior chamber. Larvae were attracted to one end of the chamber by means of a broad-spectrum light (75 W/irr). The larvae were then induced to swim along the axis of the tank by reversing the direction of the light. Approxi- mately 20 individuals were timed during each experiment as they swam through a 5-cm-long section of the chamber. BEHAVIORAL ECOLOGY OF CRAB LARVAE 405 D MERCURY D SEAWATER A LIGHT SOURCES B BEHAVIOR CHAMBER C PROXIMAL RESERVOIR D DISTAL RESERVOIR E PULLEY SYSTEM FIGURE 1. Diagrammatic representation of apparatus used to measure swimming rate as a function of hydrostatic pressure. The effect of pressure on swimming rate was measured by repeating the pro- cedure described above at various pressure increments. Pressure was increased by raising the distal end of the mercury manometer shown in Figure 1 (Sulkin, 1973). In nature, pressure increases with depth at a rate of 1 atm per 10 m of water. In one set of experiments the pressure increments used were 0, 20, 40, and 60 cm Hg (0, 0.26, 0.52, 0.78 atm, respectively). In a second set of experi- ments, the increments were 0, 1, 2, and 3 atm. The pressures were tested sequentially. Experiments were repeated with larvae from several broods, with total sample sizes ranging from 70 to 140 individuals. The effects of temperature and salinity on swimming rate were measured by repeating the procedure described above. Larvae were acclimated to the specified temperature or salinity for 24 hr before testing. RESULTS Sinking rates Mean passive sinking rates are shown for zoeal stages I, IV, and VII in Figure 2. Analysis of variance confirms a significant increase in sinking rate from the hatching stage to zoeal stage VII (P < 0.001). G eo taxis Figure 3 shows the distributions at 30 min for all replicates conducted on stage I larvae at the three salinities. The mean distribution for each salinity is shown at the bottom of Figure 3. It is apparent that negative geotaxis is strong and pervasive in stage I larvae. It also appears that a $'/« salinity change between 25 and 35#0 S has little, if any, effect upon the presence or sign of geotaxis in zoeal stage I. Statistical analysis supports these conclusions. The results in Table I indicate that the experimental mean position values were higher than the control in ever case and that the differences were significant in all three salinities. Figure 4 shows the distributions at 30 min for all replicates conduct stage IV larvae at three salinities. The mean distribution for each sali 406 SULKIN, VAX HEUKELEM, KELLY, AND VAN HEUKELEM 1.2 r 1.0 01 0.8 2 ° 6 o 2 0.4 0.2 I IV VII STAGE OF DEVELOPMENT FIGURE 2. Mean passive sinking rates for zoeal stages I, IV, and VII of Callinectes sapidus. Vertical bars represent ± 1 standard error. TABLE I Analysis of relative distributions between vertical experimental and horizontal control tests for stage I larvae at three salinities. Calculation of mean position value described in text. Salinity Mean position value Mann-Whitney t/test Experimental Control probability 25 2.50 1.17 2.50 0 <0.05 2.95 0.33 30 2.85 1.31 2.11 0.45 3.00 0.45 2.32 1.28 2.40 0.76 <0.001 2.76 0.76 1.61 1.05 2.10 0.55 2.80 0.80 35 2.35 0.85 2.45 0.85 <0.05 3.00 0.68 BEHAVIORAL ECOLOGY OF CRAB LARVAE 407 shown at the bottom of Figure 4. In the tests at 30%c S, there was considerable variability in responses shown in experimental replicates. h it appears that negative geotaxis is exhibited by a portion of the sanipl u:h replicate, there is also an indication in all replicates thai a portion of ti. may be responding with a positive geotaxis. A reduction in salinity from , 25%o S decreases ihe proportion of the sample responding with a negative- while an increase from 30 to 35#c S increases the proportion exhibiting negative The fourth stage may be in a transition period during which the sign of geotaxis is variable and is sensitive to salinity change. The data in Table II show thai al 25'/ S, for six of eight replicates at 30/£t S, and for two of three replicates at 35(/f(, S. The Mann-Whitney U test TABLE II Analysis of relative distributions bet-ween vertical experimental and horizontal control tests for stage IV larvae at three salinities. Calculation of mean position value described in text. Salinity Mean position value Mann-Whitney U test probability Experimental Control 25 0.55 0.61 0.48 1.04 1.33 1.36 = 0.10 0.32 0.91 30 1.30 0.72 2.00 1.03 1.82 0.84 0.37 1.64 1.44 0.64 0.92 0.90 >0. 0.40 0.25 1.87 1.26 1.04 0.84 1.95 1.70 35 1.26 0.90 2.65 2.48 0.78 0.72 = 0.05 2.20 0.95 BEHAVIORAL ECOLOGY OF CRAB LARVAE 409 TABLE III Analysis of relative distributions between vertical experimental and horizontal v.s/.s for stage VII larvae at three salinities. Calculation of mean position value described in text. Salinity (%) Mean position value Mann- Whitney U test probability Experimental Control 25 0.19 1.25 0.25 0.55 <0.05 0.06 2.38 30 0.33 1.05 0.42 0.10 0.67 0.94 0 0.08 0.56 0.48 <0.05 0.39 0.75 0.94 0.80 0.25 0.86 35 1.00 0.72 0.30 0.86 >0.05 0.05 1.24 indicates significant differences between distributions in vertical and horizontal tanks at 25 and 30%c S, but not at 35'A S. Barokinesis Because early larval stages of estuarine crabs are sensitive to small increases in hydrostatic pressure (Sulkin, 1973; Bentley and Sulkin, 1977; Wheeler and Epifanio, 1978), initial experiments were designed to determine response of blue crab larvae to small pressure increments. The data in Figure 6 show mean 1.0 - O 2 1.0 0.1 - x N^ r T lX'VXT T ----- I " I 4— — f-, ~ ~ STAGE VII STAGE IV J- • STAGE I 0 10 40 60 PRESSURE INCREMENT (cm H9J FIGURE 6. Mean swimming rates (± 1 standard error) of Callincctes sapidus zoea 1, IV, and VII in pressure increments up to 60 cm Hg (0.78 atm), in seawater and 30& S. 410 SULKIN, VAX HEUKELEM, KELLY, AND VAN HEUKELEM O 2 vo I w IU S oc. O 8 0123 PRESSURE INCREMENT (atm) 0123 PRESSURE INCREMENT (atm) FIGURE 7. (Left.) Mean swimming rates (±1 standard error) of Callincctcs sapidus zoeal stage I (open circle), stage IV (closed circle), and stage VII (half-closed circle) as a function of hydrostatic pressure in filtered seawater at 25°C and 25r/rf S. FIGURE 8. ( Right. ) Mean swimming rates ( ± 1 standard error ) of Callincctcs sapidus zoeal stage I (open circle), stage IV (closed circle), and stage VII (half-closed circle) as a function of hydrostatic pressure in filtered seawater at 25°C and SO'/,, S. swimming rates of zoeal stages I, IV, and VII in small pressure increments up to 60 cm Hg (0.78 atm). The experiments were conducted at 25° C, 30f/V S. In contrast to results for other estuarine species, stage I blue crab larvae do not respond to small increases in hydrostatic pressure. Two-way analysis of variance (development stage X pressure) indicates significant difference in swimming rate due to development stage (P < 0.005), but no differences due to pressure (P > 0.05). The interaction term, however, was significant (P < 0.005), probably due to the trend in zoeal stage IV for swimming rate to decrease with increased pressure. Mean swimming rates for the same three stages at pressure increments up to 3 atm are shown in Figures 7, 8, and 9 at salinities of 25, 30, and 35/^f S, re- spectively. Larvae were acclimated to the test salinity. Temperature was kept at 25 °C. Three-way analysis of variance showed significant differences due to develop- ment stage (P < 0.001), pressure increment (P < 0.005), and salinity (P < BEHAVIORAL ECOLOGY OF CRAB LARVAE 411 0.001). The only significant two-way interaction was stage (pressure (P < 0.025). Higher salinities have little effect on stage I larvae, but depress swimming rate in zoeal stages IV and VII. As the significant interaction 1 suggests, the effect of increased pressure is stage dependent. There is an increase swimming rate as pressure increment exceeds one atm in zoeal stage I (hig okinesis) and a reduction in swimming rate with pressure increase in zoeal stages I VII (low barokinesis). To determine the effect of reduced temperature on barokinesis, swimming rates were measured at 15°C at ambient pressure and at 3 atm pressure increment (Fig. 10). A two-way analysis of variance was conducted for each stage, testing temperature against pressure. Appropriate data collected at 0 and 3 atm increments at 25 °C, 30/^f S, were used in the analyses. For zoeal stage I, no significant dif- 2.0 1.8 1.6 E w 1.2 o 2 1.0 So. 0.6 0.4 0.2 0123 PRESSURE INCREMENT (atm) FIGURE 9. Mean swimming rates (±1 standard error) of Callincctcs safrdiis zoeal I (open circle), stage IV (closed circle), and stage VII (half-closed circle) as a funct hydrostatic pressure in filtered seavvater at 25°C and 35f/e( S. 412 SULKIN, VAN HEUKELEM, KELLY, AND VAN HEUKELEM ferences were attributable to temperature (P > 0.05). Reduced temperature sig- nificantly reduces swimming rate only in zoeal stages IV (P < 0.001) and VII (P < 0.001), with significant interaction witb pressure only in zoeal stage IV (P < 0.01 ) . However, the reduction in temperature appears to moderate both high and low barokinesis to some degree in all developmental stages. Data reported in Figures 8 and 9 indicated that acclimation to higher salinity reduces swimming rate in zoeal stages IV and VII. Shown in Figure 11 are the results of an experiment designed to determine the effect of a 10/£c S increase on larvae acclimated to the lower salinity. A two-way analysis of variance indicated significant interaction between development stage and salinity (P < 0.005) with stages significant (P < 0.001) and salinity not significant (P > 0.05). Because of the significant interaction, a series of one-way analysis of variance tests were con- ducted for each stage. The results indicated a significant difference in zoeal stage I (P < 0.05), but no significant differences in either zoeal stage IV or VII. Thus a Wytc S increase from that of acclimation increases swimming rate in zoeal stage I, but has little effect on zoeal stages IV and VII. DISCUSSION Thorson (1950) and others have suggested there is adaptive value in a changing pattern of vertical distribution in developing planktonic larvae of benthic species. For example, migration towards the surface in early larval stages may enhance dispersal, provide for a reliable source of food and an optimum of temperature and light, and reduce immediate competition with adult members of the population (Thorson, 1950; Mileikovsky, 1972). However, the late stages of benthic species must move to the bottom for metamorphosis or post-metamorphic development. In order to understand the regulation of vertical distribution through ontogeny, it is necessary to investigate the component factors which control vertical move- ment and how these factors change as development proceeds. The vertical posi- tion of a negatively buoyant planktonic animal is the net result of a complex combination of factors, including its sinking rate, the direction of active orienta- tion, swimming rate, and the proportion of time spent sinking and swimming. Changes in vertical position among various planktonic larval stages can be due to changes in one or more of these factors as development proceeds. Orientation and swimming are highly responsive to external stimuli. Because in nature many of these stimuli are variable and unpredictable, it is difficult to generalize about the presence of adaptive behavioral patterns. However, behavioral adaptations which produce a characteristic pattern of vertical distribution through ontogeny probably would evolve in response to ubiquitous and predictable selec- tive pressures. If this is true, it may be necessary to look only at behavioral responses to such conservative stimuli to detect the presence of a characteristic adaptive pattern of vertical distribution. The results reported here for response to gravity and hydrostatic pressure suggest that this is the case for C. sapidus. The finding of negative geotaxis in the first larval stage is consistent with earlier reports for crab larvae (Sulkin, 1973; Latz and Forward, 1977). Negative geo- taxis tends to orient the larvae so that active swimming will result in upward movement. Moreover, as larvae move deeper, increased pressure will result in increased swimming rate. The combination of negative geotaxis and high baro- kinesis thus produces a depth regulatory mechanism (Sulkin, 1973). Sensitivity to pressure change is common in early larval instars of crabs (Hardy and Bain- BEHAVIORAL ECOLOGY OF CRAB LARVAE 413 I6r 3 E — i.o iu OC o oo 0.8 0.6 0.4 1O i 2.8 ,- 2.4 J ^u LU 1.6 O Z " 00 0.4 11 0 3 PRESSURE INCREMENT (atm) 25 35 TEST SAL/N/TV(ppt) FIGURE 10. (Left) Mean swimming rate (±1 SE) of Callinectcs sapidus zoeal stage I (open circle), stage IV (closed circle), and stage VII (half-closed circle) at ambient pres- sure and at a pressure increment of 3 atm in filtered seawater at 30#r S and 15°C. FIGURE 11. (Right) Mean swimming rate (± 1 SE) of Callinectcs sapidus zoeal stage I (open circle), stage IV (closed circle), and stage VII (half-closed circle) in test salinities of 25'j, and 35f/f( (25°C). All larvae acclimated to 25'/,< S and 25°C. bridge, 1951 ; Rice, 1964, 1966; Sulkin, 1973; Bentley and Sulkin, 1977; Wheeler and Epifanio, 1978). However, the high threshold for pressure sensitivity in C. sapidus zoeae suggests that the pressure mechanism is not activated in shallow systems. On the other hand, the increase in swimming rate which occurs with a salinity increase (Fig. 11) would complement the pressure mechanism in a stratified, deep system or substitute for it in a more shallow one. If stage I larvae descend from surface waters, they will encounter increased pressure, increased salinity, and decreased temperature. Zoeal stage I responds to gravity, pressure, salinity, and temperature in ways which will produce upward movement. Stage I larvae thus exhibit behavioral adaptations likely to promote maintenance of vertical position high in the water column. By the fourth larval instar, changes in behavior have occurred which are likely to produce deeper vertical position. Fourth stage zoeae have a sinking rate greater than that of zoeal stage I and are in a period of transition between nega- tive and positive geotaxis. In any sample of stage IV larvae, some individual will exhibit negative geotaxis and others positive geotaxis. Furthermore, sign of geotaxis in individual larvae is sensitive to salinity change, tion in salinity induces increased incidence of positive geotaxis, whereas increase induces increased incidence of negative response. This is SULKIN, VAN HEUKELEM, KELLY, AND VAN HEUKELEM responses reported by Latz and Forward ( 1977) for Rhithropanopeus harrisii larvae. In a stratified estuary or the near-shore marine environment, these complex re- sponses help regulate depth. More larvae become positively geotactic as their development proceeds. As these larvae swim downward, however, they will en- counter increased salinity, which tends to reverse the sign of geotaxis. In this way, intermediate stages will tend to migrate down from surface waters, but main- tain their position up in the water column. Unlike stage I larvae, the fourth stage does not respond to a salinity increase by increasing swimming rate, a response which, in combination with reversal of geotaxis, would carry them upward. Indeed, as fourth stage larvae become accli- mated to higher salinities, their swimming rate drops. It also drops with reduction in temperature or with increase in pressure. These adaptations help insure that once fourth stage larvae move downward from surface waters, they will be retained at depth. By the seventh, or final, instar the transition to positive geotaxis is complete. Positive geotaxis is pervasive and insensitive to salinity change. Our results cannot be attributed either to inactivity or to increased sinking rate in zoeal stage VII. Random movement of stage VII larvae, as measured in the horizontal controls, exceeds that for zoeal stage I and is equal to that for zoeal stage IV. Although sinking rate increases 3.2-fold between zoeal stages I and VII, swimming rate increases 4.4-fold (measured at 30/£t S, ambient pressure). The positive geotaxis reported here is in contrast to the persistent negative geotaxis reported for zoeae of other estuarine species (Sulkin, 1973). However, the last zoeal stage of the mud crab Rhithropanopeus harrisii shows positive geotaxis (Latz and Forward, 1977) as does the megalopa of the crab Lcptodins floridanus (Sulkin, 1973). The same factors which reduce swimming rate in stage IV larvae as they move deeper are operative in zoeal stage VII. Swimming rate does not change with an increase in salinity, but does decrease as larvae become acclimated to higher salinity, or are subjected to increased pressure or reduced temperature. These responses are likely to produce a deep vertical position in zoeal stage VII. Thus behavioral response to pressure and gravity, as modified by temperature and salinity, provide the basis for differential vertical distribution during larval development of C. sapidus. Early stage larvae that enter surface waters, either by hatching in shallow areas or by migrating upward, are likely to stay near the sur- face. Changes in behavioral response will stimulate larvae in later zoeal stages to move deeper. Behavioral evidence that suggests a prevalence of early stages in surface waters and late stages in deep water is consistent with field evidence (Sandifer, 1975; Goy, 1976). This adaptive pattern of vertical distribution has predictable conseqences to larval dispersal in an estuarine and near-shore marine environment. The proximity of spawning grounds to the estuary mouth and the complementary behavioral adaptations which insure the presence of early stages in seaward-flowing surface waters will result in tremendous losses of larvae from the estuarine habitat. How- ever, a mechanism for recruitment back to the estuary exists. Deep waters of the continental shelf along the Atlantic coast of North America drift landward (Bumpus, 1965; Harrison ct al., 1967; Beardsley ct al, 1976), with particularly noticeable landward components at the mouths of major estuaries (Harrison ct al., 1967). Scheltema (1975) suggests that larvae positioned in this deep, landward- flowing drift would be carried shoreward and perhaps into an estuary. Offshore BEHAVIORAL ECOLOGY OF CRAB LARVAE 415 recruitment thus may be particularly significant in highly stratified estuaries with significant upstream non-tidal drift along the bottom. The adaptive vertical distribution pattern reported here wil of course, be modified by the vagaries of the environment. Furthermore, i galopa stage undoubtedly is also important in recruitment and may exhibit di.'u :. behavioral responses from those reported here for late zoeal stages (Naylor and ] , 1973). Nevertheless, a mechanism for larval exhange between the estuary and th open sea exists. This mechanism is fundamentally the same as that for retention, described for other estuarine species (Bousfield, 1955; Wood and Hargis, 1971), and results from a combination of hydrographic features common in estuarine and coastal marine environments and behavioral responses that produce an adaptive- pattern of differential vertical distribution during the period of larval development. C. sapidus differs from other estuarine species in that its offspring originate close to the estuary mouth. As a result, the probability of retention is reduced and the potential significance of offshore recruitment is increased. In stratified estuaries, population dynamics of C. sapidns may be influenced profoundly by such offshore recruitment. This work is part of a broader program of research on crab larval behavior supported by the Maryland Sea Grant program (Project No. R/F-8). We wish to thank Mr. Robert Miller and Mr. Robert McConnaughey for their technical assistance. SUMMARY 1. The first zoeal stage of Callincctcs sapidus shows negative geotaxis unaffected by salinity changes of 5(/(( ; high barokinesis at pressure increments above 1 atm ; an increase in swimming rate with a salinity increase ; and maintenance of swimming rate as temperature drops. 2. Stage IV larvae show both positive and negative geotaxis. As salinity drops, positive geotaxis prevails ; as it increases negative geotaxis prevails. Stage IV larvae show a tendency to reduce swimming rate as pressure increases, as tempera- ture drops, and as they become acclimated to higher salinities. 3. Stage VII larvae show positive geotaxis and reduced swimming rate in response to increased pressure, reduced temperature, and as they are acclimated to increased salinity. 4. Between hatching and the seventh (terminal) zoeal stage, passive sinking rate increases 3.2-fold, while swimming rate increases 4.4-fold. 5. These responses to environmental stimuli produce a pattern of early stages moving to surface waters and later stages to deeper waters. 6. Because of characteristic circulation in lower estuarine and coastal marine systems, this pattern of vertical distribution could provide a mechanism for exchange of larvae between the estuary and the coastal marine environment. 7. In stratified estuaries, offshore recruitment may significantly influence popula- tion dynamics in C. sapidus. LITERATURE CITED BEARDSLEY, R. C., W. C. BOICOURT, AND D. V. HANSEN, 1976. Physical oceanograpl the Mid-Atlantic Bight. American Society of Limnology and Oceanography Symposium, 2 : 20-34. 416 SULKIN, VAN HEUKELEM, KELLY, AND VAN HEUKELEM BENTLEY, E., AND S. D. SULKIN, 1977. The ontogeny of barokinesis during the zoeal develop- ment of the xanthid crab Rhithropanopcus harrisii (Gould). Mar. Bchai1. Physiol. 4 : 275-282. BOUSFIELD, E. L., 1955. Ecological control of the occurrence of barnacles in the Miramichi Estuary. Nat. Mus. Can. Bull., 137 : 1-69. BUMPUS, D. F., 1965. Residual drift along the bottom on the continental shelf in the middle Atlantic Bight area. Linmol. Occanogr., 10 (Supplement): R50-53. CARRIKER, M. R., 1951. Ecological observations on the distribution of oyster larvae in New Jersey estuaries. Ecol. Monoyr., 21 : 19-38. DUDLEY, D. L., AND M. H. JUDY, 1971. Occurrence of larval, juvenile, and mature crabs in the vicinity of Beaufort Inlet, North Carolina. NOAA Tech. Report, NMFS, Special Scientific Report, Fisheries. 637 : 1-10. FORWARD, R. B., JR., AND J. D. COSTLOW, JR., 1974. The ontogeny of phototaxis by larvae of the crab Rhithropanopeus harrisii. Mar. Bio!., 25 : 27-33. GOY, J. W., 1976. Seasonal distribution and retention of sonic decapod crustacean larvae within the Chesapeake Bay, Virginia. Masters thesis. Old Dominion University. 334 pp. HARDY, A. C, AND R. BAINBRIDGE, 1951. Effects of pressure on the behavior of decapod larvae. Nature, 167: 354-355. HARRISON, W., J. J. NORCROSS, N. A. PORE, AND E. M. STANLEY, 1967. Circulation of shelf waters off Chesapeake Bight — surface and bottom drift of continental shelf waters between Cape Henlopen, Delaware and Cape Hatteras, North Carolina. June 1963- December 1964. United States ESS A Professional Papers. 3: 1-82. HOPKINS, S. H., 1943. The external morphology of the first and second zoeal stages of the blue crab, Callincctcs sapidus Rathbun. Biol. Bull., 87: 145-152. LATZ, M. I., AND R. B. FORWARD, JR., 1977. The effect of salinity upon phototaxis and geotaxis in a larval crustacean. Biol. Bull., 153 : 163-179. MILEIKOVSKY, S. A., 1972. The "pelagic larvaton" and its role in the biology of the World Ocean, with special reference to pelagic larvae of marine bottom invertebrates. Mar. Biol.. 16: 13-21. NAYLOR, E., AND M. J. ISAAC, 1973. Behavioral significance of pressure responses in megalopa larvae of Callincctcs sapidus and Macro pipits sp. Mar. Bchav. Physiol., 1 : 341-350. NICHOLS, P., AND P. M. KENEY, 1963. Crab larvae (Callincctcs) in plankton collections from cruises of M/V Theodore N. Gill. South Atlantic coast of the United States, 1953-1954. U. S. Fish and Wild. Serv. Spec. Sci. Rep. Fish., 448: 1-14. OTT, F. S., AND R. B. FORWARD, JR., 1976. The effect of temperature on phototaxis and geo- taxis by larvae of the crab Rhithropanopeus harrisii (Gould). /. Ex p. Mar. Biol. Ecol., 23 : 97-107. PRITCHARD, D. W., 1952. Salinity distribution and circulation in the Chesapeake Bay estuarine system. J. Mar. Res., 11 : 106-123. RICE, A. L., 1964. Observations on the effects of changes in hydrostatic pressure on the be- havior of some marine animals. /. Mar. Biol. Assn. U. K., 44 : 163-175. RICE, A. L., 1966. The orientation of pressure responses of some marine Crustacea. Proceedings of Symposium on Crustacea, Part III, Mar. Biol. Assn., India, 1966: 1124-1131. SANDIFER, P. A., 1973. Distribution and abundance of decapod crustacean larvae in the York River Estuary and adjacent lower Chesapeake Bay, Virginia, 1968-1969. Chesapeake Sci., 14 : 235-257. SANDIFER, P. A., 1975. The role of pelagic larvae in recruitment to populations of decapod crustaceans in the York River Estuary and adjacent lower Chesapeake Bay, Virginia. Estuarine Coastal Mar. Sci., 3 : 269-279. SCHELTEMA, R. S., 1975. Relationship of larval dispersal, gene-flow, and natural selection to geographic variations of benthic invertebrates in estuaries and along coastal regions. Estuarine Research, 1 : 372-391. SULKIN, S. D., 1973. Depth regulation of crab larvae in the absence of light. J. Exp. Mar. Biol. Ecol., 13: 73-82. SULKIN, S. D., 1975. The influence of light in the depth regulation of crab larvae. Biol. Bull., 148: 333-343. SULKIN, S. D., 1978. Nutritional requirements during larval development of the portunid crab, Callincctcs sapidus Rathbun. /. Exp. Mar. Biol. Ecol., 34: 29-41. SULKIN, S. D., I. PHILLIPS, AND W. VAN HEUKELEM, 1979. On the locomotory rhythm of hrachyuran crab larvae and its significance in vertical migration. Marine Ecology — Progress Scries, 1 : 331-335. BEHAVIORAL ECOLOGY OF CRAB LARVAE 417 THORSON, G., 1950. Reproductive and larval ecology of marine bottom invertebrates. Kiol. Rev. of Camb. Philos. Sac., 25 : 1-45. VAN ENGEL, W. A., 1958. The blue crab and its fishery in Chesapeake Bay. I. Reproduction, early development, growth, and migration. Count. Fish. Rev., 20 : WHEELER, D., AND C. E. EPIFANIO, 1978. Behavioral response to hydrostatic :n-fssure in larvae of two species of xanthid crabs. Mar. Biol., 46: 167-174. WOOD, L., AND W. J. HARGIS, 1971. Transport of bivalve larvae in a tidal estuary 29-44 in D. J. Crisp, Ed., Fourth European Marine Biology Symposium, Cambridj • Univ. Press, Cambridge, England. Reference : Biol. Bull.. 159 : 418-427. ( October, 1980 ) ELECTROPHORETIC VARIATION IN SYMPATRIC MUD CRABS FROM NORTH INLET. SOUTH CAROLINA KATHERINE TURNER AND TIMOTHY A. LYERLA Department of Biolotjy, Clark University, Worcester, Massachusetts 01610 In the decapod crustaceans, electrophoretic screening of enzymes and proteins has shown that this group, in general, possesses low levels of detectable genetic variation (Hedgecock ct a/., 1976; Gooch, 1977; Nemeth and Tracey, 1979). Gooch (1977) has hypothesized that this may be a characteristic of the decapods, and Nemeth and Tracey (1979) suggest it may be due to low rates of mutation or of intracistronic recombination. Since both evolutionary (Lewontin, 1974) and ecological (Nevo, 1978) inferences are made from genetic variation in organisms measured with this technique, it is important that the generalization of low vari- ability in decapod crustaceans be thoroughly tested by sampling several different members of this group under diverse ecological conditions. To this end, we have assessed genetic variation in xanthid mud crab populations at North Inlet, South Carolina, using gel electrophoresis. These studies provide three additional populations to add to the data concerning electrophoretic variability in the decapods, and deal with genera that have been studied previously as to their ecological relationships (McDonald, 1977) and a species in which taxonomic "forms" have been described (Rathbun, 1930). At North Inlet, both Panopcns hcrbstii and Eurypanopeus dcprcssus, which are frequently associated with oyster bar communities along the eastern Atlantic seaboard of the United States (Lunz, 1937; Ryan, 1956; Williams, 1965), coexist in the mud and crevices between and under shells on intertidal oyster bars. P. hcrbstii is the larger of the two in adult stages but overlaps sizes of E. deprcssus during juvenile stages. Trophic differences between juveniles appear to be a major mechanism of niche partitioning that permits their successful coexistence (McDonald, 1977). In addition, morphological varieties or "forms" of P. hcrbstii occur at this site, including the forms "simpsoni" and "obesa," among the four types described within the geographic range of this species (Rathbun, 1930). The form "typica" also occurs at North Inlet but was not found in sufficient numbers for this study. MATERIALS AND METHODS Collection Adult specimens of E. deprcssus and the two forms of P. hcrbstii were collected from October, 1977, through October, 1978, from Town Creek at North Inlet, South Carolina. Data for this study were derived from summer (July) and fall (October) collections of 1978. Animals collected during the fall (October) and winter (December) of 1977 and spring (March) of 1978 were used primarily to assess techniques. Data from these collections indicated no overt seasonal dif- ferences. Although E. dcpressus was found throughout the intertidal region, the two forms of P. hcrbstii occurred in generally discrete but partly overlapping areas : 418 GENETIC VARIATION IN MUD CRABS 419 "obesa" in the upper interticlal region where the marsh grass, Sparlina alteniiflora, was abundant, and "siinpsoni" in the lower areas where oyster rubble and clumps of the oyster Crassostrca rirginica covered the surface. In the mid-intertidal, which was characterized by a relatively flat mud surface, both forms could be found, although "obesa" appeared to be more abundant. Crabs were identified by morphological characteristics outlined by Rathbun (1930) and McDonald (1977). Electrophoresis As these crabs were so small, II. depressus homogenates were prepared from whole animals. The appendages, dorsal carapace and hepatopancreas were removed from P. hcrbstii before samples were homogenized. Otherwise, the homogenization procedure was the same for both genera. The homogenizing buffer was comprised of 10 mM Tris and 10 mM maleic acid, pH 7.4, with 10 mM disodium ethylene diaminetetraacetic acid (EDTA), 10 mM MgCl^, 0.4 mM nicotinamide adenine dinucleotide phosphate (NADP+), 0.1 % (v/v) 2-mercaptoethanol and 10% (v/v) 1° octanol added. Homogenates (4Qf/o w/v) were prepared in an ice bath with Polytron PCU-2 and then centrifuged for 20 min at 900 X g in a Beckman Microfuge B in a cold room at 4°C. Soluble and participate fractions were separated in some samples in order to evaluate the banding patterns of enzymes. For these separations, homogenates were prepared according to the methods of Lyerla ct al. (1979), using 0.3 M sucrose in 10 mM Tris/citrate buffer, pH 7.4, with lO^o (v/v) 1° octanol added to reduce foaming. Supernatants were saved and final pellets dispersed in the regular homogenizing buffer to be analyzed electrophoretically as the soluble and participate fractions, respectively. Samples were screened on 13% (w/v) Connaught starch gels using standard horizontal electrophoresis techniques for all systems except amylase. Electro- phoretic buffer systems included the following : A) 0.1 M Tris/maleate (Shaw and Prasad, 1970), B) 0.41 M Sodium citrate/ citrate (Brewer, 1970), C) 0.155 M Tris/citrate (Shaw and Prasad, 1970), D) 0.5 M Tris/disodium EDTA/borate (Shaw and Prasad, 1970), and E) 0.05 M K2HPO4/KH2PO4, pH 7.0, with gel buffer a 1 : 15 dilution of electrode buffer. Electrophoresis was 70-95 V for 16 hr using buffer systems A, C, D, and E, and for 4 hr with buffer system B. All runs were standardized with bromophenol blue tracking dye, which migrated 8 cm from the origin. Staining recipes are given below and are modifications of standard methods (Shaw and Prasad, 1970). Dyes, substrates, co-factors and auxiliary enzymes were purchased from Sigma Chemical Co. Unless otherwise noted, gels were incubated at 37° C until well defined bands were present (usually within 1-2 hr). Staining recipes were: Aldolasc (ALD) : 275 mg fructose- 1, 6-diphosphate (trisodium salt), 100 U glyceraldehyde-3-phosphate dehydrogenase, 75 mg sodium arsenate, 25 mg nicotinamide adenine dinucleotide (NAD*), 15 mg nitroblue tetrazolium (NBT), 2 mg phenazine methosulfate (PMS) in 50 ml of 0.4 M Tris/HCl buffer, pH 8.0. Alkaline phosphatasc (AKPH) : 50 mg sodium a- naphthyl acid phosphate, 50 mg fast blue BB. 30 mg MgCU, 30 mg MnCl 1 g NaCl in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Bcta-glucuronida. (b-GLU) : 20 mg 4-methylumbelliferyl-/2-D-glucuronide in 5 ml of O.Of Na2HPO4/citric acid buffer, pH 4.0. Gel surface flooded and incubated at 420 K. TURNER AND T. A. LYERLA for 45 min. Rinsed once with HL-O and 7.4 N NH(OH added dropwise to cover surface of the gel. Enzyme activity marked by bands of fluorescence when gel viewed immediately under UV illumination. Catalase (CAT) : 0.0015% HoO-. (v/v) in 0.01 M XaH,PO,/XaOH buffer, pH 7.0. Gel soaked for 15 min at room temperature, rinsed once with H2O and stained with 0.09 M KI. White bands against dark blue background appear after 20 min at room temperature, indicating sites of enzyme activity. Est erase (EST) : Gel soaked for 30 min in 0.05 M boric acid at 4°C before staining. 30 mg a-naphthyl acetate, 30 mg a-naphthyl butyrate, 15 mg a-naphthyl proprionate, 100 mg fast blue RR in 50 ml of 0.05 M Xa^HPO.t/XaH^PC^ buffer, pH 7.0. Glutamate-oxaloacetate transaniinasc (GOT) : 1 mg pyridoxal-5'-phosphate, 200 mg L-aspartic acid, 100 mg a-keto- glutaric acid, 150'mg fast blue BB in 50 ml of 0.2 M Tris/HCl buffer, pH 8.0. Glucose-6-phosphate dchydrogcnasc (G6PD) : 200 mg glucose-6-phosphate (disodium salt), 50 mg MgC^, 15 mg NADP+, 15 mg NET, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Glutamatc-pyrnvatc transaminase (GPT) : 30 mg NADH, 50 mg L-alanine, 20 mg a-ketoglutaric acid, 300 U lactate dehydro- genase (Type III) in 10 ml of 0.02 M Tris/HCl buffer, pH 8.0. Gel surface flooded wth 5 ml stain and incubated at 37°C until non-fluorescent sites (enzyme activity) in fluorescent gel surface appeared under UV illumination. Isocitratc dehydro- dehydrogenase (IDH) : 200 mg isocitrate (trisodium salt), 50 mg MgCU, 15 mg NADP+, 15 mg NBT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Malatc dehydrogenase (MDH): 600 mg L-malate (monosodium salt), 25 mg XAD\ 15 mg XBT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Malic enzyme (ME) : 600 mg L-malate, 50 mg MgCU, 15 mg XADP+, 15 mg NBT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. 6-phosphogluconate dehydrogenase (6-PGD) : 100 mg 6-phosphogluconic acid, 50 mg MgClo, 15 mg NADP+, 15 mg XBT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Phosphoglucose isomerasc (PGI): 80 mg fructose-6-phosphate (disodium salt), 50 mg MgClo, 80 U glucose-6-phosphate dehydrogenase (Type XI) , 7.5 mg NADP+, 10 mg MTT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Phospho- glucomutase (PGM) : 300 mg glucose-1-phosphate (disodium salt), 80 U glucose- 6-phosphate dehydrogenase, 50 mg MgCl2, 15 mg NADP+, 20 mg MTT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 7.0. Protein (PRO): 200 mg nigrosin in 25 ml 95% methanol, 25 ml H^O, 5 ml glacial acetic acid. Gel stained for 10 min at 37 °C and destained with several washes of methanol/acetate acid fixative without nigrosin over 1-2 days at room temperature. Tetrasolium o.vidase (TO) : 25 mg XAD+, 15 mg XBT, 2 mg PMS in 50 ml of 0.04 M Tris/HCl buffer, pH 8.0. Amylase activity was visualized by electrophoresis on horizontal polyacrylamide gel slabs \6% acrylamide (w/v)] followed by incubation of the acrylamide gel with an underlying gel of 2 parts agar (Difco Bactoagar) and 1 part potato starch (Sigma Chemical Co.) with 3.4 mM XaCl in gel buffer for 90 min at 37° C (modified from Evans ct al., 1977). Buffer system C was used for both gels. After incubation, the acrylamide gel was discarded, and the agar gel stained with a 0.3% (w/v) KI, 0.15% (w/v) Ij solution with amylase activity appearing as translucent bands against a dark blue background. Putative allelic variants of polymorphic loci were identified by the distance, in mm, of their migration from the origin to the center of the band of activity. Only those bands differing by 1.5 mm or greater were considered as allelic variants resolved in these systems. GENETIC VARIATION IN MUD CRABS 421 TABLE I Summary of electrophoretically scorable loci for three populations of xanthid mud crabs (P. herbstii form "simpsoni," P. herbstii forw "obesa" and K. cleprcssus). Enzyme/protein Abbreviation/ no. loci Buffer system f Variationft :ons 1 . Amylase AMY/1 C P All tl., 2. Catalase CAT/1 B M All throe 3. Beta-glucuronidase b-GLU/1 B M All three 4. Glutamate-oxaloacetate transaminase GOT/1 C, E M All three 5. Glutamate-pyruvate transaminase GPT/1 A M All three 6. Glucose-6-phosphate dehydrogenase G6PD/1 E M All three 7. 6-phosphogluconate dehydrogenase 6PGD/1 A M All three 8. Tetrazolium oxidase TO/1 D M All three 9. Alkaline phosphatase AKPH-1 A, E M E. depressus AKPH-2 M E. depressus 10. Aldolase ALD-1 B M All three ALD-2 M All three 11. Isocitrate dehydrogenase IDH-1 B M P. herbstii IDH-2 M All three 12. Malate dehydrogenase MDH-1 A, C M/P* All three MDH-2 M/P** All three 13. Malic enzyme ME-1 D M All three ME-2 M All three 14. Phosphoglucose isomerase PGI-1 D M All three PGI-2 M All three 15. Phosphoglucomutase PGM-1 C M All three PGM-2 M All three 16. Esterase EST-1 D M form "simpsoni" EST-2 M E. depressus EST-3 M E. depressus EST-4 M All three 17. Protein PRO-1 C M All three PRO-2 M All three PRO-3 M All three * In form "simpsoni" only. ** In form "simpsoni" and E. depressus. t See Materials and Methods, tt P = polymorphic; M = monomorphic. RESULTS Sixteen different enzyme systems, as well as soluble proteins, were screened in both genera (Table I). These gave a total of 27 scorable loci in E. depressus, 25 in P. herbstii "simpsoni" and 24 in P. herbstii "obesa." Eight enzymes exhibited a single band of activity, and were considered as products of a single structural gene in each population. Seven enzyme systems appeared as double bands from whole, unfractionated homogenates. From fractionation studies, it was determined that the double banded condition was due to the presence of both soluble and participate forms, and therefore the two bands were considered products fror two structural genes. The anodal-most band was designated "-1" and the oti "-2". Esterase (EST) and soluble protein (PRO) exhibited multiple ba 422 K. TURNER AND T. A. LYERLA TABLE II Allele frequencies at polymorphic loci in E. deprcssus (a), P. herbstii form "obesa" (b) and P. herbstii form "simpsoni" (c). Sample size* Allele frequency \ ori Allele** L^\j\^i a b c a b c Amylase 54 20 79 125 — 0.300 — 116 — 0.100 — 113 — 0.200 0.158 109 — 0.150 0.392 105 — 0.125 0.210 103 — 0.125 0.240 100 0.676 — — 95 0.324 — — Hobs 0.278*** 0.600*** 0.468*** Hexp 0.442 0.826 0.724 MDH-1 45 20 89 120 — 1.000 0.950 111 — — 0.050 100 1.000 — — Hobs — — 0.101 Hexp — — 0.096 i\IDH-2 51 16 109 100 0.931 1.000 0.985 94 — — 0.015 92 0.069 — — Hobs 0.096 — 0.029 Hexp 0.126 — 0.029 * Sample size represents the number of individuals screened at each locus. ** Alleles were numbered by assigning "100" to the most common allele at each locus in E. depressus and adding or substracting the difference in mobility (in mm) for other alleles. *** Observed and expected heterozygosities were significantly different from the expectations of Hardy-Weinberg equilibrium (P < 0.05). P. herbstii "simpsoni" had the greatest number of esterase bands, but only two were consistently resolved: EST-1, which was unique to "simpsoni," and EST-4, which co-migrated with an esterase zone found in both P. herbstii "obesa" and E. depressus. Esterase zymogram patterns in P. herbstii "obesa" were clearly distinct from those of P. herbstii "simpsoni" or E. depressus, but only one zone of activity, the common EST-4 site, was consistently resolved. The esterase pat- tern from E. depressus was also readily distinguished from those of both forms of P. herbstii and had two zones unique to this genus, EST-2 and EST-3, as well as the third, common site, EST-4. Thus, four electrophoretically separable zones of esterase activity found among the three populations could be resolved in all samples and are presumably coded for by separate structural genes. Electrophoretic variations that could be ascribed to presumptive allelic variants for these sites were not found, and these genes are scored as monomorphic (Table I). Three bands of soluble protein were consistently resolved in all samples. The arrays from both forms of P. herbstii were identical. In E. depressus, the two anodal-most bands were faster migrating than those in P. herbstii, whereas the third site co-migrated with that found in this genus. As in the esterase zones of activity, no putative allelic variants of the three sites of soluble protein were observed. These were scored as the products of three different structural genes, PRO-1. -2 and -3, which were monomorphic in each population (Table I). GENETIC VARIATION IN MUD CRABS 423 Nineteen monomorphic loci were scorable in all three populations (Table I). E. depressus was also monomorphic at AKPH-1 and -2, EST-2 and -3, and MDH-1, while P. hcrbstii "simpsoni" was monomorphic at KST-1 .;iul -4 and IDH-1 and /'. hcrbstii "obesa" at KST-4, 11)11-1. and MDH-1 and - Although E. depresses and both forms of P. hcrbstii were monomorphic at the same 19 loci, at only three of these, EST-4, GOT and PRO-3, did the three populations share electromorphs. The remaining 16 loci were fixed for electrophoretically different sites in the two genera. The "100" allele at MDH-2 was also common to both genera. Allelic frequencies at polymorphic loci are presented in Table II. A locus was classified as polymorphic if the frequency of the most common allele was less than 0.99 (Nei, 1975). The expected heterozygosity at each locus was calculated using Levene's (1949) formula for finite samples and was compared to the observed heterozygosity to determine if there was a significant deviation (P < 0.05) from the expectations of Hardy-Weinberg equilibrium (Table II). Disequilibrium was observed for AMY, but not MDH, in all three populations. Values for the parameters P and H for each population are given in Table III. Using a Mest for comparisons (Crow and Kimura, 1970), the observed and ex- pected H values within each population are not significantly different, and neither the observed H values nor the expected H values between populations differ at the 0.05 level of significance. A difference between the two forms of P. hcrbstii was observed at the AMY locus, with six allozymes present in the "obesa" form and four in "simpsoni" (Table II; Fig. 1). The four allozymes in the "simpsoni" population were shared by "obesa," but at lower frequencies in "obesa." Other possible genetic differences between the two forms are at the two MDH loci. Both were polymorphic in "simpsoni" and appeared to be monomorphic in "obesa." However, due to the small sample size of "obesa" ( N = 20 at MDH-1 ; N = 16 at MDH-2) and the low frequency of heterozygotes at these loci in "simpsoni" (H = 0.101 at MDH-1 ; H = 0.029 at MDH-2), it is possible that "obesa" is polymorphic at these loci. That is, the probability of not having sampled a heterozygote in "obesa" if the fre- quencies of homozygotes at these loci are the same in each population is greater than 0.05 [ (0.899)-" == 0.12 at MDH-1 ; (0.971 )li; = 0.62 at MDH-2J. DISCUSSION In this study, the classification of a particular P. hcrbstii crab as a "simpsoni" or an "obesa" form variant by morphological characters was supported by amylase TABLE III Proportion of polymorphic loci (P) and mean heterozygosities (H) in E. depressus, P. herbstii form "obesa" and P. herbstii form "simpsoni." Hobs = observed and Hexp = expected heterozygosities. SE = standard error. Population Loci P Hobs ± SE Hexp ± SE E. depressus 27 0.074 0.015 ± 0.0004 0.021 =t 0.0012 P. herbstii form "obesa" 24 0.042 0.025 ± 0.0025 0.034 ± 0.0049 P. herbstii form "simpsoni" 25 0.120 0.024 ± 0.0006 0.034 ± 0.0017 424 K. TURNER AND T. A. LYERLA 123 4 5678 9 10 11 12 13 14 15 16 FIGURE 1. Polyacrylamide gel zymogram of amylase activity in Panopcus licrbstii. Chan- nels 1-10 are duplicate samples from five different P. licrbstii form "simpsoni" individuals, and channels 11-16 are duplicates from three different P. licrbstii form "obesa" individuals. Their inferred genotypes are as follows: channels 1-2 and 3-4, 109/105; channels 5-6, 113/113; channels 7-8, 103/103; channels 9-10, 105/105; channels 11-12, 125/105; channels 13-14, 125/103; channels 15-16, 105/105. At most, only two bands of amylase activity were resolved from a given individual, indicating a monomeric protein with electrophuretically separable allelic variants. and/or esterase zymogram patterns. Using Nei's index (Nei, 1975) to estimate genetic similarity, however, the two forms are genetically identical (I = 0.997 ± 0.011 SE, where I ranges from 1, or total identity, to 0, with no shared alleles in two populations). This is due to the large number of monomorphic loci at which the two forms shared electromorphs (20 out of 23). On an individual gene basis, those contributing to an identity of less than 1 were polymorphic: MDH-1 and AMY. Yet the two forms' dissimilar esterase zymogram patterns and the presence at the AMY locus of two alleles in "obesa" not found in "simpsoni" may indicate a greater genetic difference between these two populations than that implied by their designation as form variants ( Rathbun, 1930). Differences between two populations of loci such as those coding for non-specific esterases may mean these could be considered as genetically divergent populations (Sarich, 1977). Although the two genera, P. hcrbstii and E. dcprcssus, were readily dis- tinguished at most loci, they shared electromorphs at EST-4, GOT, PRO-3 and MDH-2. Other similarities were also found. The scorable polymorphisms occurred only at the AMY and MDH loci, and the frequency of heterozygotes was low for MDH and high for AMY in both genera. Also, allcle frequencies at the MDH loci in both genera conformed to the expectation of Hardy-Weinburg equi- librium but not at AMY in either genera. These results provide some insights into current problems of observed genetic variation in decapod crustaceans. The electromorphic differences of some 19 out of 23 scorable loci between P. hcrbstii and E. depressus indicate there has been sufficient evolutionary time (and a high enough mutation rate) to allow a genetic distinction, even at the single gene level, between these two sympatric genera. Their ecological niches may overlap considerably (McDonald, 1977), but this obviously does not require overlap of electromorphic forms of structural genes. Also, GENETIC VARIATION IN MUD CRABS 425 genetic variation within each genus, as measured by mean heterozygosities, is low and similar to estimates for other decapod crustaceans (Gooch .- id Schopf, 1972; Hedgecock ct al.. 1976; Gooch, 1977; Cole and Morgan, 1978: ,,! I'.jsol, 1978). This estimate is made without respect to the types of gene: xmtributing to it. Yet P. hcrbstii and /:. depressns are polymorphic at the same two loci, AMY and MDH, and monomorphic at more than 19 other scorable genes. MDH has usually been included in studies of genetic variability in other decapod species, and there does not appear to be a tendency for these genes to be more frequently polymorphic than others (Gooch and Schopf. 1972; Tracey et al., 1975; Gooch. 1977; Cole and Morgan, 1978; Costa and Bisol. 1978; Nemeth and Tracey, 1979). In addition to this study, electrophoretic variation of AMY has apparently been reported in other decapod crustaceans only by Nemeth and Tracey (1979) in their survey of crayfish populations (Orconcctcs spp. and Catuharus spp. ). The locus was monomorphic in all but one- population. It seems reasonable to assume, then, that the specificity of the observed polymorphisms in P. hcrbstii and E. depressus is of biological importance to both genera. The disequilibrium at the AMY locus and the low likelihood of having the same two genes and only these two. showing allelic variation in two sympatric genera suggest that selective in- fluences rather than random changes account for the nature of this variation. Finally, the low values of H in all three populations support the suggestion that low genetic variability is a phylogenetic character of the decapod crustaceans (Gooch, 1977). Gooch (1977) examined three separate populations and three different stages of development for electromorphic variation of some 15 loci in the xanthid crab, Rithropaiiopcits harrisii and found no life-cycle or geographic vari- ability, with the possible exception of a polymorphic peptidase locus in a Maryland population. As Gooch (1977) points out, these estimates of low genetic variability in decapods, expressed as the low incidence of electrophoretic variability at protein loci, are likely to remain valid even if a larger number of loci were screened. In this study, where the number of monomorphic loci is biased upward by including only those esterase bands that exhibited no genetic variation, refinement of techniques or parent-offspring data that might provide possible allelic variants in other, poorly resolved zones would not appreciably alter the overall heterozygos- ity estimates for these genera. Nemeth and Tracey (1979) have suggested that low mutation or recombination rates may explain the low variability in this taxonomic group. However, the genetic distinctions between these two sympatric genera, P. hcrbstii and E. deprcssus, are compatible with average mutation rates expected to account for the accumulation of structural gene differences between two genera at selectively neutral genes (Sarich, 1977). The specificity of their polymorphic loci imply a selective advantage for some of the allelic variants of AMY and MDH. and it is here that further work with genetic variation in these two genera would seem most appropriate. We are grateful to Dr. Jack McDonald, Smithsonian Institution, Fort Pierce Bureau, Fort Pierce, Florida, for providing much valuable information and to Dr. Alan M. Young, Nasson College, Springvale, Maine, for his assistance in collecting animals; for the use of the facilities at the Belle \Y. Baruch Institute fo Marine Biology and Coastal Research, University of South Carolina; and to Denni Allen and the staff at the Baruch Institute for their assistance and hospitality This research was supported in part by funds from the Lerner Fund for 426 K. TURNER AND T. A. LYERLA Biology of the American Museum of Natural History and by a Grant-in-Aid for Research from Sigma Xi to Katharine Turner. It was completed in partial fulfillment of the requirements for the Master of Arts degree in Biology from Clark University to Katherine Turner. SUMMARY The genetic variation of 27 scorable genes in Eurypanopeus dcpressus, 25 in Panopeus herbstii form "simpsoni" and 24 in P. herbstii form "obesa" at Town Creek, North Inlet, South Carolina, was studied by gel electrophoresis. The three populations had low levels of genetic variability, comparable to those found in other decapod crustaceans, with polymorphisms observed only at the amylase and malate dehydrogenase loci. The two forms of P. herbstii could be dis- tinguished both by previously described morphological differences and by genetic differences found in this study. The amounts of genetic variation in E. dcpressus and P. herbstii were similar, as measured by H, but shared electromorphs at only 4 of 23 scorable loci. The common, but specific, polymorphisms at amylase and malate dehydrogenase were taken to imply selective influences as a factor in their maintenance within these two sympatric genera. LITERATURE CITED BREWER, G. J., 1970. An Introduction to Isoz\mc Techniques. Academic Press, New York, 186 pp. COLE, M. A., AND R. P. MORGAN, 1978. Genetic variation in two populations of blue crab, Callincctcs sapidus. Estuaries, 1 : 202-204. COSTA, R., AND P. M. BISOL, 1978. Genetic variability in deep sea organisms. Biol. Bull., 155: 125-133. CROW, J. F., AND M. KIMURA, 1970. An Introduction to Population Genetics Theory. Harper and Row, New York, 591 pp. EVANS, M. J., L. L. HUANG, AND W. D. DAWSON, 1977. Genetic variation of amylases in deer mice. /. Hcred., 68 : 313-316. GOOCH, J. L., 1977. Allozyme genetics of life cycle stages of brachyurans. Chesapeake Sci., 18: 284-289. GOOCH, J. L., AND T. J. M. SCHOPK, 1972. Genetic variability in the deep sea: relation to environmental variability. Evolution, 26 : 545-552. HEDGECOCK, D., R. A. SHLESER, AND K. NELSON, 1976. Applications of biochemical genetics to aquaculture. /. Fish. Res. Bd.. Can.. 33: 1108-1119. LEVENE, H., 1949. On a matching problem arising in genetics. Ann. Math. Stat., 20 : 91-94. LEWONTIN, R. C, 1974. The Genetic Basis of Evolutionary Change. Columbia University Press, New York, 346 pp. LUNZ, G., 1937. Xanthidae (mud crabs) of the Carolinas. Charleston Museum Leaflet No. 9: 9-27. LYERLA, J. H., T. A. LYERLA, AND S. E. JOHNSON, 1979. Malate dehydrogenase isozyme patterns in species of terrestrial isopods. Cotnp. Biochem. Physiol., 63 : 19-24. McDoNALD, H. J., 1977. The comparative intertidal ecolot/y and niche relations of the sympatric mud crabs Panopeus herbstii Milne-Edwards and Eurypanopeus depressus (Smith) at North Inlet, South Carolina, U. S. A. ( Dccapoda: Brachyura: Xanthidae). Ph.D. Thesis, University of South Carolina, 233 pp., DAI 38/098, p. 4109. NEI, M., 1975. Molecular Population Genetics and Evolution. American Elsevier Publ. Co., Inc., New York, 288 pp. NEMETH, S. T., AND M. L. TKACKY, 1979. Allozyme variability and relatedness in six crayfish species. /. Hcred., 70 : 37-43. NEVO, E.. 1978. Genetic variation in natural populations: patterns and theory. The or. Popul. Biol., 13 : 121-177. GENETIC VARIATION IN MUD CRABS 427 RATHBTN, M., 1930. The cancroid crahs of North America of the families Euryalidae, Portunidae, Atelecyclidae, Cancridae, and Xanthidae. U. S. Nut. Mus. Bull. 152 : 1-609. RYAN, E., 1956. Observations on the life histories and the distribution of th< Xanthidae of Chesapeake Bay. Am. Nat.. 56: 138-162. SARICH, V. M., 1977. Rates, sample sizes, and the neutrality hypothesis in evolutionary studies. Nature, 256: 24-28. SHAW, C. R., AND R. PRASAD, 1970. Starch gel electrophoresis — a compilation of recipes. Biochem. Genet., 5 : 297-320. TRACEY, M. L., J. NELSON, D. HEDGECOCK, R. A. SHLESER, AND M. L. PRESSICK, 1975. Bio- chemical genetics of lobsters : genetic variation and the structure of the American lob- ster (Homarus amcricanus) populations. J. Fish. Res. Bd., Can.. 32: 2091-2101. WILLIAMS, A., 1965. Marine decapod crustaceans of the Carolinas. Fishery Bulletin [U. S. Fish and Wildlife Service], 65 : 1-298. Reference : Biol. Bull., 159: 428-440. (October, 1980) ORIENTATION AND CURRENT-INDUCED FLOW IN THE STALKED ASCIDIAN STYELA MONTEREYENSIS CRAIG M. YOUNG AND LEE F. BRAITHWAITE Department of Zoology. Uiiiirrsity of Alberta, Edmonton, Alberta T6G 2E<), Canada; and Department of Zoology, Brit/lnuii Y tinny Uuii'crsity, Provo, Utah 84602 Virtually all sessile suspension feeders rely on ambient water movement to renew their food supplies and to carry away waste products and depleted water. In addition, species in many taxa have developed hydrodynamic mechanisms for exploiting the energy in the velocity gradient between the substratum and the moving water. For instance, exogenous currents induce flow through tubes, burrows, or internal chambers, and across external food-collecting surfaces (Vogel, 1978). Where currents are predictably unidirectional or bidirectional, colonial organisms frequently exhibit a permanent orientation which maximizes exposure to current (Wainwright and Dillon, 1969; Grigg. 1972; Meyer, 1973). In areas with turbulent water or currents of unpredictable direction, orientation-independent mechanisms are common (Warner, 1977). These are exemplified by animals with irregular or radial forms, such as sponges (Vogel, 1974) and some crinoids (Meyer, 1973). While barnacles (Crisp and Stubbings, 1957) and brachiopods (La Bar- bara, 1977) reorient actively when the current direction changes, certain sea anemones (Koehl, 1976), conical stalked hydroids, and erect bryozoans allow currents to orient them passively (Warner, 1977). Monniot (1967) has reported that the subtidal ascidian Microcosmos viilgaris normally orients its siphons up-current, but is capable of actively reorienting when suspended material becomes too dense. Surprisingly, evidence for induced flow has been presented for only one ascidian species, Stycla hlicata (Hretz, 1972). Stycla montereyensis is a common stolidobranch ascidian in the Northeast Pacific. Johnson and Abbott (1972) have redescribed the species clearly, calling attention to morphological variation within and among local populations. The animal is anchored by an irregular tunic holdfast and held more or less erect in the water by a long stalk. Although both siphons are inserted anteriorly as in most ascidians, the larger incurrent siphon is recurved to point either posteriorly or ven- trally. The entire body appears longitudinally plicated due to alternating thick and thin tunic areas. Preliminary diving observations indicated that the flexible stalk of Styela allows ambient currents to reorient the animal passively, facilitating induced flow. The mechanism by which this occurs operates regardless of the direction from which the current comes. This feature is appropriate since Stycla characteristically occupies shallow water where the surge oscillates, sometimes in unpredictable direc- tions, every few seconds. In this paper, we document current-induced flow and describe aspects of orientation, morphological variation, habitat utilization, and larval habitat selection related to the use of currents by Stycla. 428 ASCIDIAN CURRENT UTILIZATION FIGURE 1. A) Known geographical range of Stycla inontcrcycnsis on the Pacific Coast of North America, showing main study areas. Subtidal sites are capitalized. Unlabelled arrows indicate sites where Stycla was not found. B) Mills Peninsula in Barkley Sound, British Columbia, showing 2 protected outer coast sites and a nearby protected site. C) Study sites on the Monterey Peninsula, California. MATERIALS AND METHODS We collected or observed Styela montereyensis at 11 sites from southern Cali- fornia to Vancouver Island (Fig. 1), thereby covering the species' known geo- graphical range (Van Name, 1945). Although Styela characteristically occurs in areas of vigorous water movement (Fay and Vallee, 1979), its occasional appear- ance in quiet bays allowed us to compare morphology of specimens collected from a variety of current regimens. We made no quantitative measurements of current velocities, so our designations of sites as "open coast," "protected outer coast," or "protected bay" are based on subjective evaluation of local geography, surf, and fauna, following the criteria of Ricketts and Calvin (1962). The nature of the substratum, approximate incline of the surface of attachment, orientation of the attachment surface relative to the direction of the prevailing surf or surge, depth or tide level, and approximate dimensions of the substratum were recorded for each individual we encountered while diving or collecting in; tidally. Relative stalk length of each animal was determined in the laboratory measuring the compression-resistant portion of the stalk and expressing thi as a fraction of the total body length exclusive of expanded siphons. 430 C. M. YOUNG AND L. F. BRAITHWAITE Living animals were transported in cold seawater (on ice, where necessary) to aquaria at Hopkins Marine Station, California; FYiclay Harbor Laboratories, Washington State ; Bamfield Marine Station, British Columbia ; or Brigham Young University, Utah. Larval cultures and live adults were maintained at 8°-14°C, depending on the season and the seawater system used. At Neah Bay, Washington, we recorded the permanent orientations of ascidians relative to the horizontal. This was done at low tide by loosely holding each animal perpendicular to the piling, viewing it from the anterior end, and drawing an arrow on a slate to represent the direction the incurrent siphon pointed. The angles of the arrows were later measured using a protactor and placed in a "rose diagram" frequency distribution which was compared with a circular-normal distribution using chi-square (Batschelet, 1965). Internal flow rates were estimated in a running seawater flume in which current velocity could be varied by regulating incoming water volume and the size of the outlet channel. Current speed in the tank was measured by repeatedly timing the passage of fluorescein dye along a 1 m course. Each animal was tied by its holdfast to a small weight in the flume, with the animal's incurrent siphon directed upstream. When the animal acclimated, about 1 ml of red food coloring (F.D.C. red #9,40) in seawater solution was introduced into the incurrent siphon with a pipette. A hand-held stop watch was used to measure the time elapsed before reappearance of the dye at the excurrent siphon. Twenty trials were made with each animal at each of several current speeds. The ascidians showed no adverse reaction to the dye, though they closed up in response to low concentrations of fluorescein or rhodamine. To assess the possible influence of stalk flexibility on feeding, the rates of ingestion by animals free to sway with the currents were compared with those of animals restrained upright. Specimens of intermediate size (12 ±2 cm) were collected from the Cannery Row site and held in clean seawater aquaria for 48 hr to clear food from their guts. Animals were paired by size ( ± 0.1 g, dry weight) and one of each pair randomly designated as the experimental animal. Each control animal was wired by its holdfast to the surface of a commercial structural brick, so that it could move freely. Each experimental animal was either anchored within a hole in the brick or caged above the brick in a narrow tube of 0.5 cm plastic mesh (Fig. 2). By restraining the experimental animals at two different vertical levels in this fashion, we hoped to control for boundary-layer and turbulence effects ; some animals would feed at the vertical level of a flexed control animal while others would feed at the level of an erect control animal. The bricks with tunicates attached were placed in Monterey Bay at 4 m depth and left for 24 hr. They were then retrieved and brought into the laboratory, where the ascidians were removed and isolated in separate bowls of seawater. All feces expelled over the subsequent 24 hr were collected on pre-tared filter papers and washed with distilled water. Filter papers and animals were oven dried at 50°C and weighed on an analytical balance sensitive to 0.001 g. Tadpole larvae were reared from spawned gametes or gametes removed by dissection. Responses of the settling larvae to light and gravity were assessed by examining the settling distributions of larvae in 16-ml-capacity molded-polystyrene petri dishes completely filled with water and covered on one side with black plastic tape (Young and Braithwaite, 1980). The dishes were cooled in a shallow seawater table under incandescent light at an intensity of 1500 Ix. ASCIDIAN CURRENT UTILIZATION n = 5 n=11 n = 6 FIGURE 2. Schematic drawing of field feeding experiment, showing positioning of restrained and control animals in calm water (top) and current (bottom). In a second experiment, larvae were placed in a circular channel, 5.75 cm wide and 3 cm deep, created by positioning a glass stacking dish in the center of a large shallow battery jar, half darkened with opaque black plastic. Twin air jets at opposite sides of the dish moved water and larvae around the channel at an average speed of 3 cm/sec. The tadpoles, slowly moving between the light and dark sides of the dish, were thus given the opportunity of settling anywhere along two continuous light-dark gradients, simulating conditions they might encounter in the field, where currents move them over illuminated and shaded portions of the substratum. After all animals settled, the distribution of zooids was plotted to the same scale on a large piece of paper. A protractor was used to divide the circular plot into 10° sectors. The number of animals in each sector was recorded, and each sector was paired with each opposite sector for a two-way analysis of variance in which the within-treatment variance tested the position of the sector pair relative to the air jets and the between-treatment variance tested light versus dark. In a final series of experiments, groups of larvae were maintained in bowls of filtered seawater in continuous light or continuous dark. The dishes were moni- tored every few hours to assess the difference in settling times for larvae rear* under these radically different lighting conditions. Yamaguchi (1970) and and Ghobashy (1971), working with other ascidian species, established prec for this experimental design. 432 C. M. YOUNG AND L. F. BRAITHWAITE FIGURE 3. (Top.) Typical siphon inclinations of Styela montereyensis collected from the open coast (A) and protected bays (B, C). Arrows indicate incurrent siphon aperatures. FIGURE 4. (Bottom.) Current-facilitated orientation in Styela montereyensis. RESULTS Phenotypic plasticity Two major morphological characteristics, incurrent siphon curvature and rela- tive stalk length, varied among animals inhabiting different current regimes. By observing numerous animals in laboratory aquaria over periods of up to 2 months, we concluded that each consistently points its expanded incurrent siphon in a specific direction relative to its own longitudinal axis. The curvature which deter- mines this direction apparently results from differential growth in the walls of the siphon. Specimens collected from all sites on the open coast or protected outer coast, either intertidally or subtidally, characteristically had siphons curved nearly 180° from the anterior point of insertion. The aperature was thus directed posteriorly. With few exceptions, animals collected from the two protected sites displayed ventrally-directed incurrent siphons (Fig. 3), as did some animals collected from deeper than 12 m in Monterey Bay. ASCIDIAN CURRENT UTILIZATION 433 TAHLK I Relative stalk lengths of Styela montereyensis at several sites (ranked subjectively by exposure). Variation among group means significant at P < 0.001 (Cattle Point site nut included in analysis because of small sample size). OC: open coast; PC: protected outer coast; I'h: :>; ducted bay. Site Exposure N Mean relative stalk length Standard deviation Still water Cove OC 15 0.659 0.054 Barkley Sound PC 14 0.655 0.066 Crescent Beach PC 5 0.541 0.045 Breakwater PC 50 0.572 0.066 Dillon Beach PC 13 0.545 0.054 Bamfield Inlet PB 9 0.470 0.031 Neah Bay PB 80 0.388 0.061 Cattle Point PB 2 0.461 0.033 Table I gives relative stalk lengths of animals from each site. Stalks are significantly longer in more exposed habitats, with the between-group variance component being significant at P < 0.001 by one-way analysis of variance. Orientation and induced flow Different orientation mechanisms are employed by Styela in habitats with dif- ferent current regimes. In subtidal habitats with strong oscillating surge, Styela orients as follows : Each wave that passes onshore bends the flexible stalk far enough to align the longitudinal axis of the animal with the current. The subse- quent backsurge instantly flips the animal 180° aligning it once again (Fig. 4). In either position, the recurved incurrent siphon points directly upcurrent. If the H12 FIGURE 5. (Top.) Orientations of 95 Styela adults on vertical pilings in Neah Bay, as percentage of animals with incurrent siphon orientation falling within each 20° sector. Orienta- tion is non-random ( P < 0.001) by chi-square. FIGURE 6. (Bottom.) Orientations of 95 juvenile Stylca (<3 cm long) on pilings at } Bay, shown as percentages as in Figure 5. Orientation is random (P>0.05). 434 C. M. YOUNG AND L. F. BRAITHWAITE 8 6- 4- 2- 10 r=-0.86 p<005 '-K-,... "I 8 6 p<0.05 10 10- ~ B - ^ 1 J 8- 0) o> " r=-0.79 g_ p>0.05 tV""% t r=-0.87 ""-^ 1 p>0.05 S 6H in ° ""- I 6~ ^ ^ >S II ^x,. 4- "-^ T3 T i ^ 1 o ~* -»T ^ ^ * 2- vx 2- 0> E - x — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — ' 04 8 12 16 20 24 28 32 0 4 8 12 16 20 24 28 32 C F 8 r= -0.96 8_ • r=-0.75 - p<0.01 p< 0.001 6- 6^ 4- T ~ - - 4- ' ^'>. * 2- " "r - - _ i 2- -i 1 1 1 1 1 1 — i 1 1 1 1 1 1 1 1 r 0 4 8 12 16 20 24 28 32 0 4 8 12 16 20 24 28 32 Ambient Current Speed (cm/sec) FIGURE 7. Relationship between current speed and internal filtering rate (measured by the time elapsed while dye passed from the incurrent to the excurrent siphon) in Stycla montcrcycnsis. Each point represents mean of 20 trials, shown with 95% confidence interval. A-D) 11 to 12 cm long animals. E) 15 cm long animal. F) Pooled data from A-D (shown without confidence intervals for clarity). animal is anchored without nearby obstructions, orientation is effective regardless of the direction from which the current comes. In quiet bays, the surge is seldom enough to sway the short-stalked animals. Individuals in Neah Bay attached to wooden pilings and floats display a permanent orientation. A significant majority of animals longer than 3 cm have their mid- saggital planes aligned parallel with the surface of the water (Fig. 5). Thus, the ventrally-directed incurrent siphons of most animals are pointed upcurrent (into either the surge or the backsurge) 509^ of the time. Very small individuals are pointed randomly (Fig. 6), so the adult orientation apparently results from rotational growth rather than larval behavior or reorientation shortly after settling. Further indication of this is that the longitudinal furrows of Neah Bay specimens often spiral in the stalk region, while those of subtidal outer coast speci- mens most frequently run straight from the holdfast to the anterior end. ASCIDIAN CURRENT UTILIZATION 435 TABLE II Fecal production by restrained and unrestrained Styela in subtidal feeding experiment. Restrictive Animal dry weight (g) : Fecal weight (g)/ ody weight (g) : treatment Unrestrained Restrained Unrestrained Restrained hole 0.192 0.186 0.052 0.037 hole 0.408 0.286 0.046 0.017 hole 0.345 0.258 0.029 0.012 hole 0.332 0.256 0.057 0.004 hole 0.194 0.199 0.072 0.015 cage 0.175 0.167 0.068 0.036 cage 0.150 0.151 0.186 0.033 cage 0.214 0.229 0.168 0.030 cage 0.901 0.855 0.026 0.019 cage 0.767 0.741 0.027 0.013 cage 0.755 0.689 0.039 0.013 In laboratory flume experiments utilizing animals from Monterey Bay, flow through the branchial basket and atrium increased at higher current speeds. Furthermore, in most cases the relationship between external current speed and internal flow was significantly linear (Fig. 7). We augmented the number of points available for regression analysis by pooling data from all animals 11-12 cm long. Despite obvious behavioral differences between individuals, the points conformed to a linear model at the 0.001 level of significance. The maximum current speed tested with each animal was determined by the animal itself. Beyond a certain point, ranging from 14 to 30 cm/sec, each animal contracted its siphons and stopped filtering. In ascidians, correlation between internal and external flow is not necessarily indicative of enhanced feeding, for two reasons. First, ascidians may be capable of increasing the activity of their lateral cilia when detecting current. Second, they are capable of filtering without actually feeding, by interrupting the secretion of endostyle mucus (MacGinitie, 1939). Our field experiment, in which ingestion (measured by fecal production) was compared in restrained and unrestrained animals, was designed to meet both of these potential objections. All animals in the field experiments survived their respective treatments without apparent abrasion or other damage, and all animals fed while in the field. \Yithin each pair of animals, the restrained animal always consumed less food (Table II). There was no consistant difference between the two restrictive treatments, so data from all pairs were pooled. Paired one-way analysis of variance revealed a significant difference (F<0.01) between means of the restricted and control groups. Assuming all animals move food through their guts at a constant rate and that this rate is more or less independent of the volume of food ingested, we conclude that induced flow actually enhances feeding in the field. C. K. Goddard (University of New South Wales, personal communication, 1978), in using markers to trace the passage of food through the guts of other stolidobranch species, has shown both of the above assumptions to be true. Habitat utilization Styela montercyensis is found on hard substrata from the mid-intertidal to least 15 m depth. Although the overwhelming majority of animals are attaci 436 C. M. YOUXG AND L. F. BRAITHWAITE TABLE III Percentage of animals at each site free to swuv horizontally in all directions, obstructed on one or more sides, or in surge channels. Intertidal (I) No. man-dives N % Unobstructed % Obstructed % In surge or subtidal (S) or low tides within radius within radius channel Breakwater S 9 89 98.9 1.1 0.0 Cannery Row S 4 29 100.0 0.0 0.0 Carmel River Beach S 4 11 81.8 0.0 18.8 Stillwater Cove S 3 21 66.7 0.0 33.5 Crescent Beach S 3 9 0.0 loo.o 0.0 Dillion Beach I 3 14 0.0 21.4 78.6 Barkley Sound I 3 13 23.1 23.1 53.8 large subtidal boulders or rock reefs, specimens of S'ycla are also found on a variety of other substrata including wooden pilings and floats, styrofoam floats, rubber tires, steel pipe, and cement. They occur as epizooites on the barnacles Balanus nubilus and Balaniis cariosus, the mussel Myiilns californianus, the ascidians Stye/a inontcrcycnsis and Pyiira haustor, and the algae Cystoccira osmun- dacea and Gigartina ca/ifornica. In addition, some specimens at the Monterey Breakwater site are found anchored in sand. These latter animals have unusually large and ramified holdfasts for securing themselves in soft sediment. Since surge currents are generally horizontal (Bascom, 1964), we hypothesized that animals on vertical surfaces might be in danger of battering against their substratum when the surge approaches head-on. Field observations support this hypothesis. Table III compares the number of animals at each site that are free to sway in any horizontal direction within their own radius with animals not free to sway. A significant majority of animals occupy the unobstructed positions, which are nearly always on the upward-facing surfaces of rocks or reefs. Animals on vertical surfaces are usually in surge channels where there are physical constaints on the direction of water movement. At Crescent Beach, an extensive survey failed to reveal any specimens except on a single face, angled about 45° to the horizontal. Although the absence of an opposing face in close proximity precludes our classifying this site as a surge channel, we note that the attachment surface is aligned perpendicular to the beach in a small bay where the surf approaches from a single direction. In the rocky intertidal, most animals occur either in surge channels or on the back sides of boulders, where water passes around the rocks on the backwash, creating effective surge channels where Stycla orients by swaying in the usual fashion. Larval settling behavior We obtained abundant spawned or dissected gametes during every month of the year except March, July, and October, when we made no attempts. Embryo- genesis (fertilization to hatching) takes 2 days (at 8°C) and the larvae may delay metamorphosis up to 6 days. Planktonic life may therefore be as long as 8 days. Styela tadpoles are not strongly photonegative at settling. In eight replicate polystyrene dishes witli an average of 100 tadpoles per dish, there was no significant preference for light or dark (P = 0.844; Wilcoxon signed-rank test). We ran two light-dark choice experiments in which larvae drifted passively around the dish with air-driven currents, one with 2441 tadpoles from Neah Bay and one with 183 ASCIDIAN CURRENT UTILIZATION 100- 80, o> 20 40 60 80 Hours from Hatching 100 FIGURE 8. Settlement of Stycla montcrcycnsis larvae in complete darkness (dotted lines) and continuous light (solid lines). Each line represents a replicate run with 50 tadpoles. A) three experiments using tadpoles from Monterey Bay animals. B ) three experiments using tadpoles from Neah Bay animals. tadpoles from Monterey. The difference in settlement between light and dark was non-significant (P < 0.5, P<0.1) by two-way analysis of variance, while the variance attributed to current effects was non-significant in the Neah Bay experiment (P > 0.75) and significant in the Monterey experiment (P < 0.05). This last difference is probably not due to a behavioral preference for a particular current regime ; non-laminar flow in the dish apparently caused pooling in some regions, resulting in uneven settlement. Stycla larvae tend to be relatively lethargic, especially at the end of larval life when they generally rest on the bottom or drift passively rather than swim up in the water column. In the petri dish experiments, significantly more larvae settled on the bottoms of the dishes than on the undersides of the lids (P = 0.008; Wilcoxon signed-rank test). Ascidian species with more active tadpoles often show the opposite response in identical experiments (Young, unpublished data). Both dona iutcstinalis (Yamaguchi, 1970) and Diplosouia listerianum (Crisp and Ghobashy, 1971) delay metamorphosis in continuous light. By contrast, Styela nwntercyensis tadpoles delay metamorphosis in continuous darkness (Fig. 8). This response is stronger in animals of the open coast than those of protected bays, but the reason for this difference is not known. DISCUSSION Among animals that utilize ambient currents for feeding, Styela monta represents an extreme in adaptiveness. Its feeding mechanism is unusual 438 C. M. YOUNG AND L. F. BRAITHWAITE efficient induced flow depends on precise orientation in habitats where current direction oscillates on a scale of seconds. Stycla uses the force of the surge not only to induce flow, but also to effect the orientation that makes this current utilization possible. Thus, foraging may require little energy beyond initial growth and normal body maintenance. Vogel (1978) has outlined three transducing mechanisms for inducing flow in biological systems .'dynamic force, the Bernoulli effect, and viscous entrainment. At least two of these apparently play some role in the feeding of Styela. When Stycla is aligned with the current, water is pushed through the feeding apparatus primarily by dynamic force against the incurrent siphon, but this effect is probably reinforced by viscous entrainment in which some water is drawn out of the excurrent siphon. In Xeah Bay animals, dynamic force and viscous entrainment aid filtering when water comes from one direction only. On the backsurge, viscous entrainment should theoretically draw some water out of the incurrent siphon. We hypothesize that the effect of this entrainment is minimal and does not significantly offset the advantage obtained by part-time orientation. Current utilization is enhanced in different habitats by the phenotypic plasticity of the species. Both open-water and protected bay forms are sometimes present in closely adjacent areas separated by a sharp exposure gradient. For example, the exposed beaches of Barkley Sound, British Columbia (Nudibranch Point and Brady's Beach), are only about 2 km from Bamfield Inlet and currents flow freely between the sites. Yet the animals at each site are distinctly different in form, There is no reason to suspect a barrier to larval dispersal, so differences are prob- ably due to ecophenotypic (physiological) adaptation rather than genetic isolation. Styela uiontcrcyensis rarely occurs in quiet water near the southern end of its range (Fay and Vallee, 1979). For example, in Newport Bay, California, where the stalked Styela clava has been introduced (Johnson and Abbott, 1972), Stycla montereycnsis is not found, despite the proximity of open water populations from which a bay population could be recruited. Styela clava has a ventrally- or anteriorly-directed incurrent siphon and a short stalk ; it is almost identical in form to the Neah Bay Stycla montereycnsis. It is not known whether induced flow aids Stycla clava in feeding or whether intrageneric competition excludes Stycla montercyensis from Newport Bay. Previous studies have shown that tadpoles of many ascidian species are photo- positive at hatching or upon release from the parent colony, and then develop a strong photonegative response at the time of settling. This results in a preference for cryptic sites. By contrast, Stycla montercyensis tadpoles in the dark delay metamorphosis for a short time. Beyond this, they are quite non-discriminating with regard to light or substratum. It is reasonable to suppose that photonegative behavior has been selected out of the behavioral repertoire of Styela as a maladaptive trait, since cryptic sites are probably the least suitable places for taking advantage of surge currents. Despite the larvae's apparent non-discrimination, a large percentage of adults are found in unobstructed positions where they can orient freely. When animals die, their stalks often remain intact for a time. We have frequently found more stalks than animals on intertidal vertical surfaces, lending credence to the idea that the observed pattern of distribution results primarily from differential mortality following a nearly-random settlement. Jorgensen (1955) has pointed out that most filter-feeders must expend a large proportion of their assimilated energy on food collection because of the relatively ASCIDIAN CURRENT UTILIZATION 439 low concentration of usable particulate matter suspended in the sea. Because of its induced flow foraging method, Stycla may be an exception to thi> generalization. At Neah Bay, animals recruited in the spring of 1979 grew to an ; :rage length of 13 cm by November. In Monterey Bay. we have found specimens as icng as 32 cm, though Van Name (1945) reported the upper size limit of the .spt ^ he 20 cm. Stycla has a lower basal metabolic rate than congeneric species found in quiet bays (C. Lambert, California State University, Fullerton. personal communication, 1978), and it is capable of spawning viable gametes in every season of the year. We suggest that a low energy demand for food gathering permits low metabolism and also frees energy for large size, indeterminate growth, and continuous reproduc- tion. Numerous stalked ascidians from the deep sea have recurved incurrent siphons (Van Name, 1945; Kott. 1969; Mnnniot and Monniot, 1978). Of special interest are species in several families with completely unciliated branchial baskets. The incurrent siphons of these forms are often hypertrophied. Although their modes of particle capture remain a mystery, Kott (1969) and Monniot and Monniot (1978) have independently speculated that bottom currents could greatly enhance their feeding by inducing flow. While Stycla inontcrcycnsis, living in strong surge, has a strong branchial basket with small stigmata, the stigmata of deep sea forms are large, seeming to provide less resistance to the dynamic pressure of even slight currents. The fact that many of these species are found in shallower water only where there are strong upwelling currents (Monniot and Monniot, 1978) supports the hypothesis that they collect food in much the same way as Stycla. We wish to thank Dr. D. P. Abbott, D. Alexander, L. Cameron, Dr. C. K. Goddard, and G. Lambert for helpful discussions. Drs. J. R. Barnes, P. V. Fank- boner, J. Farmer, R. L. Fernald, C. D. Jorgensen, and E. N. Kozloff critically reviewed various drafts of the manuscript. Laboratory space at Hopkins Marine Station and Friday Harbor Laboratories was provided by Drs. D. P. Abbott and A. O. D. Willows, respectively. We thank L. Cameron, R. Otto, A. Murray, and A. Rowe for serving as diving buddies and Dr. E. N. Kozloff for directing our attention to the Neah Bay population of Stycla. This project was supported in part by a graduate internship award from Brigham Young University. SUMMARY 1. Ambient water currents enhance internal flow and feeding in Stycla montereyensis by forcing water through the branchial basket. Induced flow depends on upstream orientation of the incurrent siphon. 2. In open-coast and protected-outer-coast habitats, most animals occur in surge channels or on upward-facing horizontal surfaces where they sway freely with the surge. Orientation takes place as the flexible stalk bends with the passing of each wave, keeping the posteriorly-recurved incurrent siphon directed upcurrent. 3. In calmer water, animals' growth form is characterized by relatively shorter stalks and ventrally-inclined incurrent siphons. Adults in these habitats are mosth oriented with the mid-saggital plane parallel to the surface of the water, orientation allows the ascidians to utilize the dynamic force in water surges from a single direction. 440 C. M. YOUNG AND L. F. BRAITHWAITE 4. Observed patterns of microdistribution probably result from differential mortality ratber than habitat selection. Larvae delay metamorphosis for a short time in complete darkness and show no strong preferences for shaded substrata. Otherwise, they seem almost non-discriminatory with regard to attachment sites. LITERATURE CITED BASCOM, W., 1964. H'ai-es and Beaches. The dynamics of the ocean surface. Doubleday and Co., New York, 267 pp. BATSCHELET, E., 1965. Statistical methods for the analysis of problems in animal orientation and certain biological rhythms. Amer. Inst. Biol. Sci. Monogr., 57 pp. BRETZ, W. L., 1972. Effects of water current speed on the filtration rate of Stycla plicata, a tunicate and epibenthic suspension feeder. Am. Zoo!., 12(4) : 720. CRISP, D. J., AND A. F. A. A. GHOBASHY, 1971. Responses of the larvae of Diplosoma listeri- annin to light and gravity. Pp. 443-465 in D. J. Crisp, Ed., Fourth European Marine Biology Symposium. Cambridge Press, Cambridge, England. CRISP, D. J., AND H. G. STUBBINGS, 1957. The orientation of barnacles to water currents. J. Anim. EcoL, 26: 179-196. FAY, R. C., AND J. A. VALLEE, 1979. A survey of the littoral and sublittoral ascidians of Southern California, including the Channel Islands. Bull. South. Calif. Acad. Sci. 78(2) : 122-135. GRIGG, R. W., 1972. Orientation and growth form of sea fans. Linniol. Occanogr., 17 : 185-192. JOHNSON, J. V., AND D. P. ABBOTT, 1972. The ascidians Stycla barnharti, S. plicata, S. clava, and S. montcrc\cnsis in California!! waters. Bull. South. Calif. Acad. Sci., 71(2): 95-105. JORGENSEN, C. B., 1955. Quantitative aspects of filter feeding in invertebrates. Biol. Rev. Camb. Philos. Soc., 30 : 391-454. KOEHL, M. A. R., 1976. Mechanical design in sea anemones. Pp. 23-31 in G. O. Mackie, Ed., Coelcntcratc Ecology and Behavior. Plenum Press, New York. KOTT, P., 1969. Antarctic Ascidiacea. American Geophysical Union Antarctic Research Series, 13 : 239 pp. LA BARBARA, M., 1977. Brachiopod orientation to water movement I. Theory, laboratory be- havior, and field orientation. Palcobiology, 3(3): 270-287. MACGINITIE, G. E., 1939. The method of feeding of tunicates. Biol. Bull., 77(3) : 443-447. MEYER, D. L., 1973. Feeding behavior and ecology of shallow-water unstalked crinoids (Echinodermata) in the Caribbean Sea. Mar. Biol., 22(2) : 105-129. MONNIOT, C., 1967. Problemes ecologiques poses par 1'observation des ascidies dans la zone infralittorale. Helgol. li'iss. Mecrcsuntcrs., 15: 371-375. MONNIOT, C., AND F. MONNIOT, 1978. Recent work on the deep-sea tunicates. Occanogr. Mar. Biol. Ann. Rev., 16: 181-228. RICKETTS, E. F., AND J. CALVIN, 1962. Between Pacific Tides. Third Edition, revised by J. W. Hedgpeth. Stanford Press, Stanford, Calif., 461 pp. VAN NAME, W. G., 1945. The North and South American ascidians. Bull. Am. Mus. Nat. Hist., 84: 1-476. VOGEL, S., 1974. Current induced flow through the sponge, Halichondria. Biol. Bull., 147(2) : 443-456. VOGEL, S., 1978. Organisms that capture currents. Sci. Am. 239(2) : 128-139. WAINWRIGHT, S. A., AND J. R. DILLON, 1969. On the orientation of sea fans (genus Gorgonia). Biol. Bull., 136(1) : 130-139. WARNER, G. F., 1977. On the shapes of passive suspension feeders. Pp. 567-576 in B. F. Keegan, P. O. Ceidigh, and J. J. S. Boaden, Eds., Biology of Bcnthic Organisms. Pergamon Press, New York. YAMAGUCHI, M., 1970. Spawning periodicity and settling time in ascidians, dona intestinalis and Stycla plicata. Rcc. Occanogr. Works //>«.. 10(2) : 147-155. YOUNG, C. M., AND L. F. BRAITHWAITE, 1980. Larval behavior and post-settling morphology in the ascidian, Chelvosoma production Stimpson. J. Exp. Mar. Biol. EcoL, 42(2) : 157-169. ABSTRACTS OF PAPERS PRESKX'i AT THE GENERAL SCIENTIFIC MEETINGS OF THE MARINE BIOLOGICAL LABORATORY AUGUST 21-23. 1980 Abstracts arc arranged alphabetically by first author ivithin the folloiving categories: Cell motility and cytoskclcton, comparative physiology, developmental genetics, ecology, fertilization and reproduction, microanatom\ and microtechniques, molec- ular biology and biochemistry, ncurobiologv, parasitology and immunology, and plant pigments and photosynthesis. Author and subject references will be found in the regular volume index in the December issue. CELL MOTILITY AND CYTOSKELETON Labile components of the dogfish erythrocyte cytoskclcton. DIANA C. BARTELT AND WILLIAM D. COHEN. We used the dogfish (Mustcla canis) erythrocyte to study the relationship between cell morphology and cytoskeletal structure and composition. Dogfish erythrocytes are flattened ellipitical cells with a nuclear bulge. Triton lysis under microtubule stabilizing conditions produces "cytoskeletons" consisting primarily of a nucleus, a continuous circumferential marginal band (MB) of microtubules, and a surrounding fibrous network (FN). The MB has been shown to be cold labile in vivo while the FN is cold stable. When lysates of cells at 0° and 25°C are examined by polyacrylamide gel electrotrophoresis (PAGE), proteins comigrating with a tubulin standard are augmented in the 0°C lysate but no material is found in either lysate comigrating with a spectrin standard. This is consistent with disappearance of the MB and stability of the FN. Dogfish erythrocytes undergo a Ca2*-dependent change in morphology (from flattened ellipse to sphere) in Ringer's solutions containing ionophore A23187 and varying Ca2* concentrations. Examination of the spherical cell "cytoskeleton" (phase con- trast) indicates that both MB and FN are Ca2+-labile in vivo. PAGE analysis of lysates from cells incubated in A23187 + Ca2+ versus A23187 + EGTA shows proteins in both tubulin and spectrin regions of the gel to be more prominent in the Ca2+ lysate, consistent with in rivo disassembly of both a tubulin-containing MB and a spectrin-containing FN. MB and FN of cytoskeletons prepared under our usual microtubule stabilizing conditions are stable at 0°C or in 5 mM Ca2* (room temp.) over long periods. A major difference between the in vivo state and the in vitro conditions is the presence of KG in the former. Inclusion of 150 mM KC1 in lysing and washing media greatly enhances the in vitro lability of cytoskeletal com- ponents to both low temperature and Ca*\ The MBs of cytoskeletons prepared and washed in the presence of KC1 and 5 mM EGTA are labile to both 0°C and 5 mM Ca-+. The FN is partially labile with long term exposure to 5 mM Ca'+ but stable at 0°C. Tubulin is present in both 0°C and Ca2+ extracts of these cytoskeletons, with spectrin appearing only in the long term Ca2* extract. Supported by NIH grant HL 20902 and CUNY grants 13313 and 13051. Release of cytoskeletal proteins into the pcrfusatc of squid giant axons. BAUMGOLD AND PAUL GALLANT. Squid giant axons were used to study the effect of various physiological manipulate the cytoskeletal proteins underlying the axolemma. The axons were intracellularK 441 442 ABSTRACTS FROM M.B.L. GENERAL MEETINGS and the bulk of the axoplasm removed using brief perfusion with a trypsin-containing solution followed by prolonged perfusion with a trypsin-free solution. The proteins dissolved in the perfusate were then analyzed by polyacrylamide gel electrophoresis after having been labeled with 125I-Bolton-Hunter reagent. After prolonged perfusion, tubulin was the major protein eroded into the perfusate. The amount of protein released into the perfusate increased sig- nificantly following each of the following manipulations : potassium depolarization, electrical stimulation, and the application of 20-40 nM tetrodotoxin (TTX) to the external surface of the axon. The proteins released by potassium deplorization included tubulins, actin, 69K protein, and a number of minor ones. No neurofilament protein was detected. Application of TTX released the following proteins: tubulin, actin, 69K protein, 65 K protein, a protein migrating between a and /3 tubulin, and several other minor ones. Since the influx of calcium is dramatically increased when a nerve is depolarized or electrically stimulated, the mechanism underlying this release of cytoskeletal proteins during repetitive stimulation and potassium depolarization could involve this increased calcium influx. However, this mechanism cannot explain the release of protein in response to the application of TTX, since TTX does not increase the calcium influx. These experiments thus demonstrate that by manipulating the excitability of the giant axon, we can affect changes in cytoskeletal proteins. We therefore conclude that an intimate relationship must exist between the intrinsic proteins in the axolemma and the underlying cytoskeleton. Light-scattering studies of sheared and unshearcd actin polymerization. W. B. BUSA, K. E. VAN HOLDE, AND M. S. MOOSEKER. Light-scattering by polymerizing actin solutions was observed using a Perkin Elmer Model MPF-44 fluorescence spectrophotometer (excitation and emission wavelengths — 520 nm). Dogfish muscle actin (0.4 mg/ml) in 2mM Tris (pH 8), 0.2 mM ATP, 0.5 mM dithiothreitol, and 0.2 mM CaCL ws polymerized in the sample cuvette by addition of 50 mM KC1 and 2 mM MgSOi. Unsheared polymerization in the absence of ligands demonstrated a slow increase in the intensity of light scattered, reaching a plateau within 200 min. Polymerization in the presence of 18 ^M phalloidin revealed an increase in rate of assembly but no change in the plateau. Cytochalasin B (5 /*M.) slightly increased assembly rate and significantly lowered the plateau. The effect of shearing on assembly was observed by Pasteur pipetting five times at each of three 5-min intervals during polymerization. Shearing early during assembly caused a dramatic increase in assembly rate. Further shearing during assembly caused less, if any, increase. The plateau was achieved within 15 min, and equalled the plateau for controls. Shearing fully polymerized F-actin caused a very rapid increase in scattering which decayed with time to the original plateau. This effect was not inhibited by phalloidin. Theoretical considerations indicate shearing per sc should not alter scattering. The increase and decay observed may, however, reflect an effect of shearing and reannealing. This work was performed under NIH Training Grant GM 00265. The marginal band system o\ the dogfish crytlirocytc. WILLIAM D. COHEN, DENEENE WHITEHEAD, AND RICHARD JAEGER. The erythrocyte marginal band (MB) is a relatively simple and potentially useful micro- tubule system with which to approach problems in cellular morphogenesis and cytoskeletal function. We have been studying dogfish (Mustclus cams) erythrocyte MBs with respect to structure, function, isolation, composition, and biogenesis. Dogfish erythrocytes are relatively large, obtainable in quantity, and have the flattened, elliptical, nucleated morphology typical of non-mammalian vertebrate erythrocytes. When cells at room temperature (18°-22°C) are lysed with Triton X-100 under microtubule-stabilizing conditions, "cytoskeletons" consisting of MB, nucleus, and a sub-surface network enclosing the MB are obtained. Cells incubated at 0°C for 1 hr or more retain normal morphology. Lysis and thin sectioning show that MBs are absent but the network present, suggesting that the network maintains cell shape in the induced absence of the MB. MBs reassemble in living cells re-warmed about 1 hr. Colchicine does not induce MB disassembly, but prevents reassembly after 0°C-MB disassembly. MB influence on cell shape is evident in atypical erythrocytes with singly or doubly-pointed ends, which appear in cell suspensions stored at room temperature for long periods. Lysis reveals correspondingly pointed MBs, some with straight crossing extensions accounting for straight polar projections sometimes seen on living pointed cells. Points are lost and CELL MOTILITY AND CYTOSKELETON 443 ellipticity regained in living cells at 0°C. This is consistent with the idea that cell mor- phology corresponds to that of a sub-surface network "set" for flattened, elliptical mor- phology, but subject to reversible deformation by the MB. Appearan-.-i of free ("isolated) MBs in some cytoskeleton preparations has led to development of a mass Ivi . . lion procedure. The isolated MBs are ribbon-like (TEM whole mounts) with inter-microtubule cross-bridges (thin sections). Recent observations on probable centriole participation in reassembly of an invertebrate (blood clam) erythrocyte MB prompted close examination of MBs in rewarming dogfish erythrocytes. Initial observations suggest that centrioles may function in vertebrate MB biogenesis as well, and raise the possibility that the plane of MB formation (plane of cell flattening) might correspond to the plane denned by a fixed right-angle centriole pair. Supported by NIH grant HL 20902 and CUNY grants 13313 and 13051. The theory of chemotaxis and the ability of a cell to sense its position and orienta- tion within a tissue. R. P. FUTRELLE. In embryogenesis, cells can move in coordinated ways while patterns of cytodifferentiation develop concurrently. Cells do this by producing and responding to chemical signals. It is possible to understand such chemical communication in fundamental terms (Bers, H. C. & E. M. Purcell, 1977, B-iophys. J. 20: 193-219). I'll give an example of chemotaxis, a critical part of cellular slime mold (CSM) development. Consider the orientation and chemotaxis occurring in response to N molecules emitted by a source cell and diffusing to a receiver cell. A concentration gradient develops at the receiver. The receiver counts the number of molecules hitting its two halves and finds a difference An, its orientation "signal." Due to thermal motion of the signal molecules, this count has an error, or "noise," given by the square-root-of-n law. The orientation will be reliable when the signal-to-noise ratio is > 1, or N > (2r/d)3 with r the source-receiver separation and d the receiver cell diameter. This is a new kind of result which relates a molecular quantity N to the macroscopic tissue geometry. For example, if d = 1/2 mm, a typical embryonic field dimension, and d = 10 /um, N > 10". Experiments show that CSM, leukocytes, and bacteria come close to the theoretical limit. N = 10" molecules corresponds to 4 /*M in a 10 /im cell — not metabolically demanding. A cell can thus have many signal and receptor types and carry on many "pattern conversions" simultaneously. Complex pattern formation is described by reaction-diffusion models. It will now be possible to estimate how many signal molecules a tissue needs to reliably generate a pattern. I have been able to reduce such models to the signal-cell paradigm by the mathematical technique of the functional derivative. This gives the perturbation in the pattern due to local cell function changes. Experimentally it corresponds to observing the effects of localized lesions or stimuli, an approach which has been successful in studies of CSM aggregation. This research supported by NSF, under PCM79-04242. Mitotic spindle behavior in unequal cleavage of Spisula solidissima. SHINYA INOUE AND KAT§UMA DAN. In sea urchin eggs the first three cleavages are nearly equal. At the fourth division the animal cells cleave equally, but the vegetal four cells cleave unequally. Prior to division in the vegetal cells, the resting nuclei migrate to the vegetal pole where they attach themselves by a centrosome. Thus at nuclear envelope breakdown, the spindle is excentrically situated with two differently sized asters — a spherical proximal one and a distal flat or truncate one. In the zygote of Spisula solidissima, the two pronuclei migrate to the middle of the cell and immediately give rise to the definitive mitotic apparatus (MA) ; a resting nucleus is hardly found. In the zygote and CD cell of Spisula, the completed MA, first situated at the center of the cell, moves in toto to the egg periphery slightly higher than the equator. The distal end of the MA oscillates as though seeking the correct attachment site. The oscillation ceases and the spindle backs off from the cell cortex and proceeds through anaphase. Unequal cleavage ensues, as determined by the position and orientation of the early anaphase spindle. In the sea urchin egg, the nuclei themselves migrate to the vegetal pole ; the nuclei and the centrosomes are already positioned, anticipating the micromere-forming division ph before the spindle is formed. In Spisula. the spindles are formed in the middle of but are positioned and oriented by the onset of anaphase presaging the ensuing cleavage factor common to sea urchin and Spisula unequal cleavages seems to be neither the 444 ABSTRACTS FROM M.B.L. GENERAL MEETINGS nor the spindle, but the centrosome. We shall turn to their behavior in relation to the cell surface or other determinants to look for the factors that initiate cell differentiation. Grant support : NIH GM 23475-15, NSF PCM 81351. Effects of cytochalasin B and phalluidin on F-actin and G-actin. MICHAEL P. MCCARTHY. The effects of cytochalasin B (CB), and phalloidin on actin polymerization and F-actin steady-state viscosity were studied using Ostwald viscometry. Conventionally prepared scup and rabbit skeletal muscle actin was polymerized in 1 mM Tris-Cl, pH 8.0, 0.1 mM ATP, 0.25 mM DTT, 0.1 mM CaCl,, 50 mM KC1, and 1 mM MgSCX As has been previously shown, actin polymerized in greater than stoichiometric amounts of phalloidin polymerizes more rapidly and reaches a higher steady-state viscosity than actin alone (Wieland, 1977, Natur- wissenschaften 64, 3037, whereas actin polymerized in substoichiometric amounts of CB reaches a lower steady-state viscosity (MacLean-Fletcher and Pollard, 1980, Cell 20, 329). These changes can be correlated to altering the critical actin concentration below which actin won't associate; phalloidin lowers the critical concentration while CB raises it. The lower, CB- induced steady-state viscosity is the same regardless of whether CB is added before, during, or at steady-state polymerization. In competition experiments, the presence of phalloidin (25 /uM), and CB (2 yuM) during polymerization yields a steady-state viscosity intermediate be- tween that of actin with phalloidin, and actin alone (7.5 /*M). The intermediate steady-state viscosity induced by phalloidin and CB is also achieved by their addition to F-actin. Supported by NIH Training Grant GM 00265. In vivo disassembly /reassembly of the marginal band in an invertebrate erythrocyte. IRIS NEMHAUSER AND WILLIAM D. COHEN. Erythrocytes of the Arcidae, like those of nonmammalian vertebrates and some inverte- brates, are flattened and ellipitical and contain nuclei and marginal bands (MBs) of micro- tubules. The "blood-clam" erythrocyte MBs are notable for the association of each with a pair of centrioles. Furthermore, these MBs are cold labile (at 0-4°) in vivo, and reassemble with re-warming to 25°C. This system thus seemed ideal for investigating the possible role of centrioles in MB formation. Blood of Noctia pondcrosa was diluted 1 : 10 in "Instant Ocean" artificial sea water and incubated at 0°C. After 1.5 hr, cell lysis with microtubule-stabilizing medium containing 0.4% Triton X-100 revealed the absence of MBs (phase contrast). Pres- ent were nuclei, centrioles, and remnants of oiher cellular organelles. Upon rewarming, the reassembled MBs once again were associated with the centrioles. Observation of the cells at intervals during rewarming reveals a developmental sequence : after 2 min centrioles are seen near the cell periphery, with fibers having free ends apparently emanating from them. After 5 min thin MBs have formed; they are pointed at one end with the centrioles located at the apex. After 15 min, the somewhat less pointed MBs appear to have thickened. By 120 min, MBs are observed with elliptical shape similar to that in controls. Thin sectioning (TEM) confirms the above results for untreated controls, and 0°C, 15-min-revvarmed, and 120-min- rewarmed cells. In addition, it shows that the centrioles are frequently in right-angle pairs and that the cell contents are enclosed by a fibrous network. After 15 min the reassembled MBs contain in cross section about 35 somewhat loosely packed microtubules (compared to about 50 tightly packed microtubules in controls). However, as in controls, the MBs are located in protrusions of the network, giving the impression that they could be producing tension in the latter. The 120 min rewarmed preparation is similar in appearance to the one obtained after 15 min, and reappearance of the MB in the former has been documented by electron microscopic examination of negatively-stained whole mounts. Experi- ments are underway to investigate further the 2 min time point, using immunofluorescence and electron microscopy. Supported NIH grant HL 20902 and CUNY 13313 and 13051. Vanadate inhibits ciliary beating in intact snail salivary glands. G. SALAMA, D. M. SENSEMAN, I. S. HORWITZ, AND B. M. SALZBERG. In the salivary gland of the freshwater snail, Ilclisonni trivoh-is, secretion may be regulated by the patterning of electrical activity generated by the effector neurons. To study CELL MOTILITY AND CYTOSKELETON 445 the mechanism ( s) coupling excitation to secretion, voltage dependent optical signals from dif- ferent regions of the gland were detected with a 25 element photodiode array. Intact glands were mounted on the stage of a microscope, s an vith a solution con- taining 25-200 fj-g/m\ Merocyanine-oxazolone dye (NK. 23(>7), and then washed. Each element of the array recorded optical "spikes" at 680 ± 5 nm from 0.01 mm' of tissue. We observed higher than expected levels of optical noise, which we attributed to oscillatory beating of the cilia that line the acini of the gland. In contrast to experiment urchin sperm flagella, we found that sodium me.avanadate, applied externally, blocked ciliary beating. In the presence of 10 mM vanadate, ciliary motion was reversibly blocked in 5 min. Occa- sionally, the presence of the vanadate enhanced the level of spontaneous electrical activity of the gland, and hyperpolarized the acinar cells by 5.0 ± 0.4 mV. The latter effects could be reversed in 5 min by returning the gland to vanadate-free Ringer's, in which the cilia remained immobile for more than 30 min. Pharmacological effects resulting from the higher dye concentration could be prevented by a five-fold increase in Ca~+ concentration, to 20 mM, during staining. Consequently, staining the salivary glands in a 200 /ug/ml solution of NK 2367 plus 20 mM CaL'+, followed by vanadate treatment for 5 min, resulted in a 20-fold increase in the signal-to-noise ratio of the optical recording of membrane potential in this preparation. The large optical signals and the spatial resolution provided by the photodiode array can be used to study conduction pathways in this and other elecirical syncitia. The authors are grateful to Joel Rosenbaurn and Dave Spray for suggesting the use of vanadate. This work was supported by NSF grant BNS 77-05025, NIDR grants De 05271 and DE 05536 and an M.B.L. Steps Towards Independence Fellowship to G. Salama. The interaction of phalloidin with G- and F-actin. FRANCINE R. SMITH, K. E. VAN HOLDE, AND MARK S. MOOSEKER. The interaction of phalloidin with actin from rabbit skeletal muscle was studied using sedimentation velocity and electron microscopy. Conventionally prepared G-actin contains a small population (3-5%) of oligomeric actin which may be removed by gel filtration, yielding "pure" actin (MacLean-Fletcher and Pollard, 1980, Cell 20, 329). At low ionic strength (2 mM Tris-Cl, 0.2 mM ATP, 0.5 mM DTT, 0.2 mM CaCU, pH 8.0 at 25°C), saturating amounts of phalloidin do not alter the sedimentation behavior of "pure" G-actin. Under identical buffer conditions, the addition of phalloidin to "conventional" actin slowly promotes the formation of two actin species (Sai, »• = 35S, S2°o, W = 74S) in addition to monomeric actin ( S^.. «• = 3S). The rapidly sedimenting species correspond to filaments of approximately 7 /t in length and to large aggregates, respectively. Similar species are observed in sedimentation velocity studies of actin polymerized by the addition of 50 mM KC1 and ImM MgSOi (San, w = 31S, Sso. w = 41S). Electron microscopic examination of filaments induced by phalloidin at low ionic strength indicates that they are morphologically identical to F-actin filaments polymerized by KC1 and MgSO4 addition. Supported by NIH Training Grant GM 00265. Observations on the isolated mitotic apparatus ghost. GEORGE W. SMITH. A remnant of the isolated mitotic apparatus (MA) of eggs was partially characterized by electron microscopy and SDS gel electrophoresis. These mitotic apparatus ghosts (MAGS) were formed whenever the mitotic appara.us was depleted of its tubulin. MAGS were formed by subjecting MAs (isolated by a modified microtubule polymerizing medium) to either low temperature (0°-4°C) or 1-100 t*.M of CaL'+. They were isolated from four species of marine organisms and were morphologically indistinct from the regular MA when viewed under phase optics. Polarized light revealed a dramatic decrease in birefringence to about 15-20% of original in sea urchin and to almost negligible readings in the surf clam. Thin sectioning of MAGS revealed the absence of microtubules in the spindle or astral fibers. Microtubules were present in the centrioles. Oriented fibers, consisting of ribosomes and an amorphous material, were aligned where microtubules had been. It is thought that the residual birefringence of MAGS is due to this oriented material. SDS polyacrylamide gel electrophoresis revealed a normal MA pattern with about an 80% reduction in the tubulin band. High molec ular weight proteins were also present but in amounts less than in a normal MA pattern. No differences were noted between MAGS produced by cold or Ca"+ depolymerization the exception that millimolar concentrations of Ca2+ frequently dissolved the entire Part of this study was supported by a Macy Foundation Grant to Marine Laboratory. 446 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Mitochondria! movements in curly development of Ciona. DAVID STOPAK AND ROSARIA DESANTIS. We have traced the movement of mitochondria during the development of the ascidian Ciona intcstinalis by use of the laser dye Rhodamine 123. Previous studies have shown that mitochondria are concentrated after fertilization in the yellow crescent (myoplasm), which is segregated into the muscle lineage. Eggs were preincubated with 10 ng/ml Rhodamine 123, which specifically binds to mitochondria, and development was followed in the living embryo by fluorescence microscopy. In the mature egg mitochondria are already restricted to the vegetal 2/3 of the egg. The first shift from uniform distribution occurs at the time of germinal vescicle breakdown. The second dramatic shift occurs after fertilization, but only after meiosis is complete and the second polar body extruded. At this time the mitochondria form a posterior crescent just above the vegetal pole. By third cleavage the mitochondrial crescent is partitioned into the posterior vegetal pair of blastomers. The next two cleavages are unequal, occurring at the midpoint of the crescent. As early as the four-cell stage, however, a small clear protuberance forms at this location, suggesting that the cortex here may be different from other areas. By this method the muscle lineage can be followed into late gastrula. Other cells contain mitochondria at later stages, but the muscle lineage continues to emit the strongest fluorescent signal. To test whether cytoskeletal elements are responsible for crescent formation we have treated eggs with Cytochalasin B (2 /ug/ml) and Colchicine (5X10~4M), concentrations sufficient to block cleavage. We find that both drugs inhibit crescent formation. However, neither is completely effective. In many cases mitochondria appear concentrated near the vegetal pole but do not form a normal crescent. We conclude that the movement of mitochondria is linked to both the microtubule and microfilament systems, but is not solely dependent on either one. The mechanism of infra-plate ciliary synchrony in ctenophores. SIDNEY L. TAMM. A ctenophore comb plate consists of hundreds of thousands of long cilia which beat together as a unit. In lobates (i.e., Mncmiopsis) the plates are triggered to beat by the interplate ciliated groove (ICG), which runs to the center of the base of each plate. A signal must therefore be transmitted outward from the ciliated groove junction to activate the beating of cilia on either side of the plate. The nature of the synchronizing signal was investigated by microsurgical experiments on single comb plates of Mnemiopsis. A physical gap between two parts of a plate was created by holding back a sliver of the plate with a needle. Only the part of the plate connected to the ICG beat during passage of a wave; the other part did not beat. This effect was reversible: upon releasing the sliver, the entire plate heat as a unit. A mechanical block between two parts of a plate, without preventing the movement of an intermediate piece, was made by slitting a plate (but not the underlying tissue) with a razor blade. Again, only the side of the plate connected to the ICG was stimulated to beat, but both parts beat synchronously after the razor blade was removed. Finally, the tissue at the base of a plate was completely cut across, producing a narrow separation in the plate as well. Nevertheless, the two severed parts beat together. However, if the smaller part of the plate was moved away from the main part, the smaller part stopped beating. Upon allowing the two parts to come together, the entire plate resumed synchronous beating. In ctenophores without an ICG (i.e., cydippids, beroids, cestids), the beating of adjacent plates is triggered by hydrodynamic interaction between them, obviating the need for an intra- plate synchronizing mechanism. However, the first comb plate in each row is stimulated by the ciliated groove running from the aboral statocyst. Microsurgical experiments on the first plate in Plcurobrachia give results identical to those on lobates. Thus, in cases where intra-plate coordination occurs, the cilia within a plate are syn- chronized by hydrodynamic coupling between them, not by cell-to-cell electrical transmission via the gap junctions between comb plate cells. The flange-like compartmenting lamellae of these cilia undoubtedly contribute to their mechanical coupling. This research was supported by NIH grant GM 27903. Binding of dyncin to isolated meiotic spindles of the surf clam, Spisula solidissima. BRUCE R. TELZER AND LEAH T. HAIMO. Recently, Haimo, Telzer, and Rosenbaum (1979, Proc. Nat. Acad. Sci. U.S.A. 76: 5795- 5763) demonstrated that flagellar dynein bound to and crossbridged rnicrotubules assembled COMPARATIVE PHYSIOLOGY 447 in vitro, and revealed intrinsic microtubule polarity. Experiments were undertaken to deter- mine if dynein could also hind to native microtubules present within the mitotic a])])aratus. Dynein was isolated from axonemes of Tctrahymcna, and meiotic spindles were obtained from activated eggs of Spisula. Isolated spindles were incubated in the presence of dynein for up to 60 min at 22°C and were collected by centrifugation. Analysis by gel electorphoresis demonstrated that dynein cosedimented with these spindles. The specific ATPase activity [/trnol Pi/(min-mg)] of the dynein preparation alone equaled 0.20 while that of the spindles was 0.04. The activity of spindles incubated in dynein for 45 min was 0.24, indicating that dynein had bound. That the specific ATPase activity of the spindles containing dynein was higher than that of the soluble dynein indicated stimulation of the dynein's ATPase activity. In addition, spindles incubated in dynein exhibited greater stability and birefringence than those incubated in its absence. The retardation of spindles immediately after isolation was 1.44 nm. After 45 min spindles incubated in the absence of dynein exhibited little if any birefringence while those incubated with dynein exhibited a retardation of 1.00 nm, further supporting interaction of the dynein with the mitotic microtubules. Ultrastructral analysis is being undertaken to visualize the binding of dynein to the mitotic microtubules and, thus, to determine their polarity. We are grateful to Joel L. Rosenbaum for generous use of laboratory equipment and for helpful discussions. We thank Ted Salmon and David Begg for suggestions and assistance. B.R.T. was supported by a Steps Toward Independence Fellowship. L.T.H. was a post- doctoral fellow of the American Cancer Society. COMPARATIVE PHYSIOLOGY Effects of temperature acclimation on nitrogen metabolism in two littorinid snails. DAVID W. ALDRIDGE, ROBERT F. McMAHON, AND W. D. RUSSELL-HUNTER. Temperature acclimation of nitrogenous excretion and oxygen uptake was investigated in a high littoral snail, Littorina rudis ( = sa.vatilis), and the low intertidal species, L. obtusata, collected at Nobska Point, Massachusetts, and acclimated for 15-19 days at 4° or 21 °C. Using an Orion ammonia probe, weight specific ammonia and urea (after urease treatment) excretion rates were determined based on tissue dry weight (TDW) as ng NH;i/(mg TDW- hr). Concurrent oxygen uptake rates (see Russell-Hunter, McMahon, and Aldridge, 1980, Biol. Bull., 159: 452) were determined with polarographic oxygen electrodes. For L. obtusata (at 5 mg TDW) acclimated at 4°C total ammonia (NH:,T urea as NHO excretion rates were 180 at 4°C; 159 at 11 °C; and 363 at 21 °C, yielding Qtl, values of 0.84 (4°-H°C), 2.28 (H0-21°C), and 1.51 (4°-21°C). Rates for 21°C acclimated snails were 62, 182, and 312, yielding Qu, values of 4.66, 1.71, and 2.95, respectively. The mean urea NH:i to ammonia NHa ratios for 4°C acclimated L. obtusata were 1.40, 1.29, and 1.01, respectively. These ratios for 21°C acclimated snails were 1.35, 0.73, and 0.46. Moles Oe : moles NH, ratios for 4°C acclimated L. obtusata were 2.61, 5.97, and 4.52. These mole-ratios for 21°C acclimated snails were 6.41, 3.64, and 5.28. For L rudis (at 3 mg TDW) acclimated at 4°C, rates in ng NH./mg-hr were 151 at 4°C; 175 at H°C; and 1325 at 21°C, yielding Q1(, values of 1.23 (4°-ll°C), 7.57 (11°-210C), and 3.59 (4°-21°C). Rates for 21°C acclimated snails were 117, 179, and 897, yielding Qm values of 1.84, 5.01, and 3.31. The mean urea NH., to ammonia NH, ratios for 4°C acclimated L. rudis were 1.82, 1.86, and 2.15, respectively. The ratios for 21°C acclimated snails were 1.39, 0.72, and 1.07. The O2 : NH, ratios for 4°C acclimated L. rudis were 1.26, 3.74, and 1.51. These mole-ratios for 21°C acclimated snails were 2.19, 4.20, and 1.79. There is no evidence for compensatory adjustments in excretion rates except for 4°C acclimated L. obtusata at low temperatures. The more aerial L. rudis excretes proportionally more urea than the more aquatic L. obtusata. Greater relative protein catabolism in L. rudis may reflect a dietary difference. Supported by National Science Foundation research grant DEB-7810190 to Dr. W. Russell-Hunter, and funds from the Biology Department of The University of Texas at Arlington to Dr. Robert F. McMahon. Janus green B: A vital stain during the process of mucus secretion. B. G. AND S. J. COOPERSTEIN. The 80 /im urn cell complex of Sifiunculus nudus responds to mucus-stimulating sub (MSS) in vitro by secreting streams of mucus from 4-6 loci of synthesis. 448 ABSTRACTS FROM M.B.L. GENERAL MEETINGS was a 1 : 10 seawater dilution of human serum heated at 85° for 4i min. Added to urns, it induced tails of mucus 2-5 X the length of the urn itself. Janus green B (JGB) is a blue dye comprising diethylsafranin linked to dimethylaniline by an azo bond. When added to urn cells the dye was bound by the synthetic loci of the cells and reduced, producing red-violet loci. When serum MSS was added, the emerging stream of mucus was blue. When secretion was stimulated before dye was added, the original secretion remained colorless but newly emer- gent mucus was blue. When JGB was reduced in a test tube and then added to urns, the loci did not bind the dye, and when MSS was added the emergent mucus was pink, indicating that the dye was bound to extracellular mucus. Janus green G differs from JGB only by having methyl rather than ethyl groups on the diethylsafranin moiety ; it cannot be reduced. Added to urns, JGG stained the loci pure blue, and emergent MSS-stimulated mucus was colorless. Anaerobically, JGB killed most stimulated urns, but when glutathione was added to the MSS, urn cell death was prevented and long refractile blue mucous tails emerged. Together, results suggest that the synthetic apparatus of the urn cell complex binds the oxidized (not the reduced) form of JGB, which is then metabolically reduced; the reduced (not the oxidized) form is incorporated into the mucus granules during synthesis, and then is re-oxidized at exocytosis, probably by a component in the heated serum. Supported by NIH 5 P50 HL 19157. Thermostability of fish hemoglobins. THOMAS A. BORGESE, JOAN M. BORGESE, JOHN P. HARRINGTON, AND RONALD L. NAGEL. The structural basis of protein heat stability is poorly understood, although electrostatic interactions (Perutz), hydrophobic interactions (Bigelow) or both (Kauzman) have been suggested. We have studied a set of homologus proteins (fish hemoglobins) and found that, as a group, their thermostability is lower than mammalian hemoglobins as measured by the time required for 50% precipitation of a 0.2 g/dl buffered hemoglobin solution at 50°C (tj)- By the same criterion, however, skate hemoglobin (type I) is more stable than human hemoglobin. When the pH dependency of thermostability was studied we found dogfish ( Mustclus canis) hemoglobin (type II) to be stable at acid pH but unstable at alkaline pH. Conversely, toadfish hemoglobin (type III) is stable at alkaline but unstable at acid pH. Type IV hemoglobins (goosefish, tautog, ocean pout, scup, and sea robin) are unstable over the pH range (6.0-8.0) studied. Thermostability of type I hemoglobin is unaffected by KC1 concentration up to 1.0 M. Type II is stabilized and types III and IV destabilized by salt. All hemoglobins were progressively destabilized by alkylureas in direct relationship to the length of their hydro- carbon side chains (methyl < ethyl < propyl < butyl ). All hemoglobins were stabilized when converted to the carbonmonoxy or cyanmethemoglobin form. The state of the heme group, in order of increasing effectiveness for thermostability is methemoglobin < oxyhemoglobin < carbonmonoxy and cyanmethemoglobin. We conclude that ( 1 ) a few sequence changes in homologous proteins are sufficient to produce large differences in thermostability (2) the effect of salt, and consequently of electro- static interactions, appears to be complex and variable for the four different hemoglobin types described (3) all fish hemoglobins were systematically destabilized when hydrophobic inter- actions were interfered with and (4) the ligand state of the home and the strength of the heme to globin attachment are important determinants of hemoglobin thermostability. Supported by PSC-BHE grants 12212 and 13321. Mosaic development in the polyclad turbellarian Hoploplana inquilina and its evolutionary implications. BARBARA BOYER. Spiral cleavage and mosaic development occur together in annelids and molluscs. My earlier work, however, showed that the acoel turbellarian embryo, with duet spiral cleavage, is regulative, suggesting that spiral cleavage and mosaicism originated independently in evolutionary history. In this study cell deletions were done on the polyclad Hoploplana inquilina to determine if the embryos of this closely related turbellarian order with typical quartet spiral cleavage are regulative or mosaic. Single blastomeres were deleted at the two-cell stage by puncture through the egg mem- branes with tungsten needles. Of 70 complete deletions, 33 survived to form larvae which were compared with the Mullers-larva controls for such characteristics as general morphology, lobe development, number and position of eyes, and swimming behavior. In all cases the COMPARATIVE PHYSIOLOGY 449 experimental larvae were abnormal, exhibiting a shape suggestive of the Mullers larva but with rudimentary lobes, only one or no eyes, and aberrant swimming behavkr. Most significantly 33% of the experimental larvae were eyeless and 51% had only one ey< suggesting a dif- ference in the morphologenetic capacities of the blastomeres at the two-( stage and possible interaction between these blastomeres in normal development to form twi The results demonstrate that the polyclad Hoplophma is mosaic witi, embryos always forming deficient larvae and suggests that mosaicism became associated with spiral cleavage in the quartet form during the evolutionary history of the Turbellaria. This association has evidently remained a permanent feature of the turbellarian-annelid-mollusc lineages. This work was supported by an MBL Steps Toward Independence Fellowship and NSF grant PCM 77 16269 to Dr. John Arnold. Substitution of calcium by polycations in sponge aggregation factor interaction. WERNER BURKART AND MAX M. BERGER. Aggregation factor (AF) from the marine sponge Microciona prolifcra promotes species-specific reaggregation of Microciona cells. Although highly specific binding of AF to cells is calcium independent, AF mediated cell aggregation probably based on AF-AF inter- actions needs high calcium concentrations of 10-20 mM. Using AF-coated Sephadex Superfine beads and iodine-125 labeled AF to assay AF-AF interaction, it was shown that minute amounts of polycations like polylysine, histones, or P-120 can substitute for calcium. As a control, low background binding of AF to bovine serum albumin derivatized beads was shown not to be affected by these polycations. The concentra- tions needed to get binding as high as with seawater concentration of calcium (450 fj-g/ml = 10 mM) were about 1 Mg/ml for P-120 and histones and 10 /xg/ml for polylysine. Based on a comparison of charges, the polycations exert their effect at a charge concentration which is 10.000X lower than that needed for calcium. Spermidine, with only three charges per molecule at concentrations of up to 30 Mg/ml, had no effect on AF-AF interaction. Neither did poly- anions like hyaluronic acid or chondroitin sulfate interfere with the binding of iodine-125- labeled AF to AF-carriers in the concentration range tested. In order to measure the interaction of single AF molecules with polycations and cal- cium, the distribution of iodine-125-labeled AF in a two-polymer aqueous phase system com- posed of dextran T500 and polyethylene glycol 6000 was measured. By charging the system, the huge AF molecule (20-10* dalton), having polyanion characteristics at neutral pH, is driven into the upper phase. Addition of 1 fig/ml polylysine or histones inverts the effect, sug- gesting the formation of complexes having a net positive charge. The most potent polycation, P-120, having more widely spaced charges, brings the partition coefficient near 1 over a wide concentration range by just neutralizing the AF-molecule. Based on these results, we propose that large areas with a multitude of charged binding sites are involved in AF-AF interaction. Although the specificity of the single binding site would be very low, cooperative effects could still provide enough selectivity to ensure species- specific AF-AF interaction and, together with the high binding specificity of AF to the cell surface, cell sorting. This work was supported by Swiss National Science Foundation grant 3.513-079. Gas secretion in the sunmbladdcr of shallow ivater fishes compared to a deep ocean fish (3000 meters}. EUGENE COPELAND, RICHARD HAEDRICH, AND JONATHAN WITTENBERG. By scanning electron microscopy, the gas-secreting epithelia of the swimbladder of five shallow water fish (cunner, toadfish, Fundulus, spadefish, and bluefish) are compared to that in the deep sea (3000 m) rat-tail fish, Clialinura (Coryphaenoides) . The shallow- water fish show evidence of a bubble release from a relatively smooth surfaced epithelium, as postulated earlier at the transmission electron microscope level (Copeland, 1969, Zeit. Zcllforscl 93, 305-331). The appearance of the surface of the secretory epithelium in Chalinura is com- pletely different. The capillaries of the rete mirable enlarge and loop onto the surf.' expanded tubules whose surfaces are bound by circular ridges (like galvanized drain pipe There is no sign of bubble release under conditions when free gas is secreted against 300 atm pressure. Conclusion : either the interpretation of the mechanism of gas rel the shallow environment is at fault or the very deep-living fish have a different mcd 450 ABSTRACTS FROM M.B.L. GENERAL MEETINGS L-Glutamate-specialist chemoreceptors on the leys of the lobster Homarus americanus. CHARLES DERBY AND JELLE ATEMA. Lobsters taste with chemoreceptors on their legs. These chemoreceptors function in food location and consumption. Multi-unit neurophysiological recordings from leg chemoreceptors demonstrated that, among the 48 compounds tested, L-glutamate, hydroxy-L-proline, and ammonium chloride, in that order, are the most stimulatory. L-Glutamate-sensitive cells were studied in more detail by using single-unit extracellular techniques, in order to determine their threshold and specificity. L-Glutamate-sensitive cells have thresholds for L-glutamate near 5 X 10~s M. This cell type responds with less than 8% of its L-glutamate response to 19 other equimolar compounds. Any change in the L-glutamate molecule causes a drastic reduction in the degree of excitation of these cells: Alteration of the carbon chain length ( DL-a-aminoadipic acid, L-asparatate), alteration of either of the carboxyl groups ( L-glutamine, -y-amino-n-butyric acid, L-glutamate- •y-methyl ester, norvaline, 4-amino-n-valeric acid), substitution on the a-carbon ( DL-a-methyl- glutamate), addition of a second or third amino acid (L-glutamyl -L-glutamate, glutathione), isomerization (D-glutamate), deamination (glutarate), and cyclization ( L-pyroglutamate). These cells also do not respond to other compounds which, in multi-unit recordings, are very stimulatory ; these compounds include hydroxy-L-proline, ammonium chloride, L-arginine glycine, betaine, and taurine. Thus, there are L-glutamate-specialist chemoreceptors in the taste organs of lobsters. Since L-glutamate is one of the common free amino acids in the tissues of many prey species of lobsters, these L-glutamate specialists are probably involved in detection and identification of food items. Thrust and drag of bluefish (Pomatomus saltatrix) at different buoyancies, speeds, and swimming angles. ARTHUR B. DuBois AND CHRISTOPHER S. OGILVY. Factors which affect the thrust and drag of swimming fish were studied using a water circuit in which bluefish averaging 52 cm and 1.53 kg were placed. Thrust and drag (Fu) in Newtons were calculated from Newton's law using the mass of each fish and the acceleration and deceleration recorded from a miniature accelerometer implanted near the center of gravity. Speeds were varied between 0 and 1 m per sec. Buoyancy was adjusted to neutral (about 75 cc), then between 45 cc negative and 15 cc positive buoyancy, by changing the volume of air in a balloon implanted in the swimbladder. The angle of inclination of the tunnel with respect to horizontal was varied between 23° head down and 15° head up. The regression line for mean drag on speed during neutral buoyancy, which was independent of angle, in nine bluefish was: Ft, = (0.51 X speed) +0.15, r - 0.73, S.E. of estimate (E) 0.090, in kg m/S2, over the range 0.15-0.95 m/S. At 30 cc negative buoyancy, horizontal, Fi. = (0.26 X speed) + 0.31, S.E. of E 0.008, over the range 0.31-0.70 m/S, where 0.31 is stalling speed. With buoyancy negative 30 cc, head down 17°, Fi, - (0.72 X speed) -0.011, S.E. of E 0.023, over a range 0.31-0.90 m/S. Negative buoyancy 30 cc, head up 15°, F,, = (0.31 X speed) +0.29, S.E. of E 0.005, at a range 0.31-1.3 m/S. With 15 cc positive buoyancy, 9° head down, Fi, = (0.68 X speed) + 0.039, S.E. of E 0.072 in the range 0.23-0.82 m/S. The bluefish were able to use their pectoral fins, whose mean projected area was 34.5 cnr, to sustain their weight in water at speeds above stalling speed, but at the cost of extra energy. They would not actually glide downward, since the force of drag, Fn, exceeded the force of their weight in water multiplied by the sine of the glide angle. Relationship of body colors to environmental light conditions in "poster colored" fishes. JOSEPH S. LEVINE AND EDWARD F. MAcNicnoL, JR. Marine and freshwater fishes exhibit great variety in their phototopic visual pigment sys- tems. Some of the variations in wavelengths of peak absorption may be related to water color in the habitat. Additional differences are related to the unique visual tasks facing each species, including visually mediated intra-specific communication, where color cues are often important. In an analysis aimed at describing visual communication quantitatively, working hypotheses were: 1) Contrast between adjacent black and white areas should be highly conspicuous under most conditions ; 2) The range of hues useful in generating color contrast should narrow with depth as the spectral-band width of the available illumination decreases; 3) The specific colors COMPARATIVE PHYSIOLOGY 451 useful as conspicuous markers should be predictably different in water with different transmission characteristics. Colors in high-quality photographs of fishes were quantified using chips from the Munsell Book of Color. All colors — and the achromatics white, black, grey, and silver were sampled along six body transects. Total and relative frequencies of occurrence and areas covered were combined to obtain an importance value for each color. One study group included all Hawaiian members of five coral reef families, divided into shallow (^20 m) and deep water (> 20 m) species. The other was South American (reddish-brown water) and African i blue water) cichlids. In both groups, black and white were the most important colors. Among actual hues, blue and yellow were most common among blue water species, while green and red predominated in the red-brown water forms. Contrary to predictions, deep water reef fishes showed minimal decrease in the importance of red coloration compared to shallow-water con- familials. However, red and orange are visually equivalent to black in the blue deep-water environment, and may be relatively easy to concentrate from ingested algae and invertebrates, whereas melanin must be synthesized from tryosine, possible with a greater energy requirement. Respiration in the polychacte worm Magelona : rcponscs to temperature, hypoxia and tentacle ablation. ROBERT F. McMAHON AND W. D. RUSSELL-HUNTER. Magelona sp., a small polychaete worm (family Magelonidae) with the unusual blood pigment, haemerythrin, burrows in sands of low oxygen content in the shallow sublittoral. The pair of feeding tentacles, extended from the burrow, apparently function in gas exchange in the well-oxygenated waters just above the substratum. Living specimens of Magelona were collected from Stony Beach, Woods Hole, Massachusetts, during the summer of 1980. Polaro- graphic oxygen electrodes were utilized to record oxygen consumption rate continuously as VO- [fj.1 O2/(mg dry tissue weight -hour) ]. Uptake rates were recorded with decreasing oxygen concentrations at 20°C from near air saturation for oxygen ( PO- = 159 torr) until uptake ceased at 1-10 torr, and at near air saturation from 10°-45°C in 5°C intervals both for whole indi- viduals, and for those with tentacles removed. Magelona appears adapted to reducing sands. It is a good regulator of VO2 with a critical PO3 of 56-60 torr. Tentacle ablation has no significant effect (P > 0.1) on VOE below 140 torr although intact specimens maintain slightly elevated VO2's over 140-160 torr. Mean VO2 was 0.460 /tl O,/(mg-hr) at 10°C and increased to 2.026 Ml O2/(mg-hr) at 30°C. Qio values for each 5°C increase from 10° to 30°C ranged from 1.88 to 2.43. VO= and Qi,, values in specimens without tentacles were very similar to those of intact specimens over 10°-30°C. VO3 increase markedly to 7.648 /xl O3/(mg-hr) at 40°C, with Q10 values for 30°- 35°C being 5.04 and for 35-40°C being 2.83; and declines at 45°C, which is lethal to Magelona. The VO2 of specimens without tentacles was significantly (P <; 0.1) inhibited at higher tempera- tures, being only 56-70% that of intact specimens at 35° and 40°C, respectively (CVs : 30°-35°C ; 2.57 and 35°-40°C; 3.57). This reduction of VO- in specimens without tentacles occurs under the elevated oxygen demand induced by temperature stress. Therefore the tentacles of Magelona, which are only 2-3% of the dry weight, appear to function as respiratory surfaces. Supported by National Science Foundation research grant DEB-7810190 to Dr. W. D. Russell-Hunter, and funds from the Biology Department of The University of Texas at Arling- ton to Dr. Robert F. McMahon. Interstitial fluid pressures of smooth dogfish (Mustelus canis) and bhtefish (Pomatomus saltatrix) titled in air. CHRISTOPHER S. OGILVY AND ARTHUR B. DuBois. Interstitial fluid pressure (IFP) can be measured using a cotton wick and polyethylene tubing (PE160) inserted subcutaneously. One was inserted in the head region and another in the caudal region of live, unanesthetized dogfish and bluefish. The fish were placed hori- zontally on a V-board in air while the gills were perfused with sea water. The five dogfish tested had an average pre-tilt head IFP of 2.0 cm H-O (S.E. ±2.3) and a caudal IFP of 2.9 cm H»O (S.E. ± 1.9). The fish were tilted head up to 30° and IFP in the caudal region rose to 15.8 cm H2O (S.E. ±7.4). The head IFP only rose 0.3 cm H2O. After 30 min tilt the fish were lowered. At the end of a 30 min follow-up period the head IFI H2O (S.E. ± 1.8) while the caudal IFP was 2.6 cm H^O (S.E. ± 1.2). In four bluefi average pre-tilt head IFP was 0.4 (S.E. ±0.2). During the tilt the head IFP rose to 2.1 cm 452 ABSTRACTS FROM M.B.L. GENERAL MEETINGS H2O (S.E. ±0.9) while the caudal IFF rose slightly to 0.8 cm H2O (S.E. ±1.2). After the tilt the IFF in the head was -0.6 cm ELO (S.E. ± 1.5) and the caudal IFF was -0.1 cm H2O (S.E. ±0.9). These results show that when a dogfish is tilted head up in air, fluid enters the interstitial space in the caudal region. This is reflected by a large positive IFF. Because only a slight increase in caudal IFF is observed in bluefish, we conclude they can com- pensate for the tilt. The dogfish tested typically showed blood oozing from the tail and died a few hours after the tilt, whereas the bluefish survived well. Perhaps the hydrodynamic forces acting on the fast swimming bluefish pre-adapted this animal to tolerate gravity while the slower swimming dogfish has much less adaptation. Lack of respiratory response to temperature acclimation in two littorinid snails. W. D. RUSSELL-HUNTER, ROBERT F. McMAHON, AND DAVID W. ALDRIDGE. Rate functions of physiological processes, such as oxygen uptake, may adapt to temperature changes over three distinct time-scales: (a) directly, within minutes or hours, (b) by com- pensatory acclimation over days or weeks, and (c) by natural selection over many generations. The commonest snail of the midlittoral, Littorina littorca, shows some thermoregulation of respiration (McMahon and Russell-Hunter, 1977, Biol. Bull., 152: 182-198; Newell and Roy, 1973, Physiol. Zoo!., 46: 253-275) but little acclimation. The high littoral snail, L. rudis (— saxatilis) , and the low intertidal species, L. obtusata, differ from L. littorea in their temperature responses. Thus possible acclimation was investigated in conjunction with simul- taneous studies on excretion. Both species were collected at Nobska Point, Massachusetts, and acclimated (at 4° or 21° C) for 15-19 days. For each test temperature (4°, 11°, 21°), oxygen uptake rates were based on 10-15 determinations using polarographic oxygen electrodes, and computed from log-log regressions against tissue dry weight (TDW) to give VO2 in /JL\ C>2/(mg'hr). Such VO^ values for L. obtusata (at 5 mg TDW) acclimated to 4°C were: 0.620 at 4°C, 1.256 at 1TC, and 2.157 at 21°C, yielding Q1(1 values of 2.74 (4°-ll°C), 1.72 (H°-21°C) and 2.08 (4°-21°C). For 21°C acclimated L. obtusata, corresponding VO2 values were: 0.530, 0.872, and 2.163, yielding Qio's of 2.04, 2.48, and 2.29. Corresponding VOs values for L. rudis (at 3 mg TDW) acclimated to 4°C were 0.250, 0.863, and 2.626, yielding Q10's of 5.87, 3.04, and 3.98. For 21 °C acclimated L. rudis, corresponding VO2 values were 0.339, 0.988, and 2.116, yielding Qi,,'s of 4.61, 2.14, and 2.94. Irrespective of acclimation history, L. rudis shows unusually high Qio values over the 4°-ll°C range. Neither species shows any significant difference in VO2 (P > 0.05) between 4° and 21 °C acclimated individuals at any test temperature. It seems possible that this lack of an acclimatory response is an adaptation to the wide short-term temperature variations experienced on rocky seashores. These two littorinid species are exposed to more marked diurnal and semilunar environmental changes than are nonmarine and deeper marine forms whose capacity for acclimatory compensation has evolved to fit sustained longer-term tempera- ture changes. Supported by National Science Foundation research grant DEB-7810190 to Dr. W. D. Russell-Hunter, and funds from the Biology Department of The University of Texas at Arling- ton to Dr. Robert F. McMahon. Stability of dogfish lens fiber cell membranes. SEYMOUR ZIGMAN, TERESA PAXHIA, AND TERESA YULO. Lens fiber cells may be the largest and most stable cells present in soft vertebrate tissues. While the hardness and elasticity of the lenses of different vertebrates varies remarkably, the basic units comprising most of the substance of the lens are quite similar. In the nucleus of the lens, these are organized (by many strong gap junctional connections) into closely contigu- ous membranes that contain excesses of extrinsic lens proteins. Nearly all cytoplasm and cell particles are absent from these highly differentiated cells. After sucrose -gradient (5-20%) purified fibers were extracted with tris-buffer (pH 7.2) 8M urea, and \% SDS, scanning EM revealed that the basic ribbonlike appearance of these fibers in Mustclus canis remained intact. Changes observed were fiber thinning and the removal of the characteristic interdigitating knobs. While the peptides dervied from the proteins extracted from the lens membranes were common to the total soluble proteins, as shown by SDS-PAGE, two bands were present only in the membrane extract (55,000 and 45,000 daltons). These were lost when DTT was used DEVELOPMENTAL GENETICS 453 in the extraction solvent, indicating that -SS- bonds are involved in crosslinking. Membrane lipid analyses showed the following: protein : lipid of 5:1; PL : C + CE of 2.5. fluidity 28% ; PE — Sph > other PL. The combination of the physical stability of the lens fiber cell mem- branes and the continued binding of formerly soluble lens proteins to them may explain the conservative process of lens nucleus growth and hardening during aging. Grant support: NIH (EY00459); Pledger Fund; Mullie Fund; RPB, Inc DEVELOPMENTAL GENETICS Healing of Xenopus eye fragments prevnarked b\ chiniaeric pigment grafts. KEVIN CONWAY AND R. KEVIN HUNT. Albino (albp) Xenopus laevis embryonic (stage 24-28) eyebuds were marked with small orthotopic grafts from normally pigmented donors. About a day later (stage 33-37) half the eyebud was deleted, leaving a locally marked half-bud fragment. These fragments were observed and photographed as they rounded up to form whole small eyes in the post-operative days, and then normally sized eyes in the following weeks of tadpole life. Marked dorsal, ventral, anterior, and posterior fragments were prepared, each type ( e.g. dorsal) bearing one of three types of marker grafts (e.g. posterior, dorsal, or anterior). In the first post-operative day, marked tissue near the cut edge migrated along the cut edge, except for ventral tissue, which did not move, presumably due to the ventral fissure. Marked tissue away from the wound did not move much in the first day. However, in the succeeding days, anterior marks in anterior fragments moved dorsally, and increased their angular extent. Similarly, posterior marks in posterior fragments moved dorsally, and expanded. Dorsal marks in dorsal fragments and ventral marks in ventral fragments expanded without moving. The clonal configuration did not change substantially a week after the operation, and was elaborated radially as the eye grew, in a manner similar to the normal growth of eyebuds with orthotopic pigment-marked clones. Thus the healing of eyebud fragments is asymmetric, in that ventral tissue is not free to contribute to the early (Day 1) melting of tissue into the wound. Also, the healing is not strictly an epimorphic regulation accomplished only by cells at the wound edge. As cells far from the cut move and expand, some morpholactic mechanism must instruct them to change their fate. This work was supported by an NIH Training grant (GM 07231) to K. Conway, and NSF (PCM-77-26987) to R. K. Hunt. Pigmentation mosaicism in the choroid of the eye after embryonic grafting. R. KEVIN HUNT, STEVE ECKMAN. AND KEVIN CONWAY. An interesting assymetry exists when half eye-buds are exchanged between pigmented and albino Xenopus embryos at stage 32. When a posterior half-bud is grafted orthotopically into an albino host, the resulting adult mosaic shows a distinct boundary between the black and white halves. When the reciprocal graft is done, the adult eye shows no boundary on external view. Histologic sections show a sharply bounded mosaic of pigment retina, but in the white- into-black chimera, the choroid layer outside the white graft was pigmented. Whole eye exchanges at stage 32 pigmented autonomously, showing that new choroidal pigment was not coming from extraocular parts of the host. Observations on orthotopic half-bud grafts at stage 32 were continued over several post-operative weeks. Choroid pigment from the black host fragment appeared to "migrate" in a dorsal-to-ventral direction. At one week, the boundary had moved significantly and by two weeks the white-into-black half-eye mosaic appeared com- pletely pigmented. Grafting anterior half-buds, white-into-black, produced a similar blackening and disappearance of the boundary. "Migration" was observed in 17 of 18 white-into-black half-eye mosaics, while 80-90% of black-into-white cases retained a sharp boundary on external view. It appears that the choroid follows different rules of development from the more orderly "sector" patterns of the pigment retinal epithelium. More work is needed to discover how the pigment "migrates": Is it actually motile? Does it divide and migrate? Does the moving pigment boundary accurately reflect the shifting positions of wild-type cells? theless, the unusual pattern of choroidal development helps us understand the complex pn> of organogenesis in the vertebrate eye. We thank NSF (PCM-77-26987) and the Alfred Sloan Foundation. 454 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Retinotectal patterns and patterns of genetic mosaicism in ploidy chimerae oj Xenopus eye. R. KEVIN Hrxr, BEN G. S/ARO. ROBERT TOMPKINS, AND DANA REIN SCHMIDT. Frog retina is a model system for study of organ morphogensis and development of nerve patterns. We have prepared genetic mosaics of Xenopus retina by microsurgically grafting small wedges of eye-bud tissue at stages 31-36 from pigmented tetraploid donors into diploid host embryos homozygous for the albino (alb1') mutation. These orthotopic grafts ranged in size from small (30°) sectors of eye-bud to half-buds. Two batches of such chimerae (N = 71, N = 76) were observed closely during healing: the 4n tissue (Tompkins ct al., 1980, Biol. Bull., 159: 455) acquired its black phenotype during the late 30s stages, and the two fragments healed together leaving a contiguous but variable "arc" of marked cells on the ciliary marginal growth ring, from which annuli of new cells are added to the periphery of the eye throughout larval growth. Pigment retinal polyclones, in the 25 individuals reared through metamorphosis, showed "radial line" boundaries ; the pigmented 4n cells were contiguous and occupied a "sector" radiating out from the optic disc to the ciliary margin. The detailed structure of polyclone boundaries, and the nuances of form seen in polyclones growing out from different angular positions, conformed closely to patterns seen in mosaics prepared with other markers. Extra- cellular electrophysiologic recording was used to analyze the visual field projection from the chimeric eye to the contralateral optic tectum. Ten frogs, whose retinal mosaic patterns ranged from small sectors (at 5, 6, 7, 3, 9, or 6 o'clock) to half-and-half patterns, all showed visual projections of normal continuous topography and normal metrics. We conclude that the tetraploid strain offers a powerful marker which is minimally intrusive into retinal growth and retinotectal patterning. We thank NSF (PCM-79-03827; PCM-77-26987) and the Alfred Sloan Foundation. A developmental analysis of an unusual homocotic mutation, proboscipedia, in Drosophila melanogaster. ANNA W. SEITZ, PETER B. MONK, AND THOMAS C. KAUFMAN. Current theories of pattern regulation in limb development propose that all limb structures proximal to the most distal structure present be represented. The homoeotic mutation probo- scipedia of Drosophila mclanoi/astcr effects labial structures in the adult fly. One allele of this locus, pbr>, causes the transformation of labial structures to prothoracic leg structures in which femoral and tarsal, but not tibial segments are present. The above theory leads us to expect that presumptive tibial cells existed at one point during limb formation, but that these cells then died either after the limb tissue was competent to replace them or as fast as they were formed. To test this hypothesis the labial discs from pb5 homozygous larvae were examined for cell death. No cell death was observed in discs of early third instar larvae. Some cell death was seen in late third instar larvae, but this did not correlate with the expected location of presumptive tibial tissue. When the morphology of the mutated labial discs was carefully examined, the left and right discs of a pair were found to differ and the morphology of neither disc resembled that of a leg disc. Mapping of the mutated disc is in progress and will clarify the observed morphology. The absence of cell death in the expected areas could be explained by the folding of the mutated disc, such that the observed cell death is tibial cell death. Alternatively, pattern regulation which operates at a very early stage in disc development could give rise to presumptive tibial cells that fail to proliferate. The asymmetry observed in pairs of discs is also manifested in adult structures. This asymmetry could reflect the unusual nature of labial discs, which are known to duplicate themselves in conditions where other discs do not. This work was supported by Public Health Service grants 5 T32 H DO 7067-04 and T32 HD07098 (A.W.S.), GM-21558 (P.B.M.), SO5 RR7031, and RO1 GM24299-01 (T.C.K.). We would like to thank Brooke Kirby for her interest and encouragement in this project. Reaction-diffusion models of morphogenesis: an application to pattern formation in Xenopus retina. SARAH A. SHOAF, KEVIN CONVVAY, AND R. KEVIN HUNT. We present reaction-diffusion model calculations for pattern formation on a disk (repre- senting a wide variety of embryonic anlagen including frog eyes), in order to examine the sensitivity of patterns to changes in initial conditions and perturbations in the geometry of the ECOLOGY 455 morphogen-producing space. The models of Kauffnian, Shymko, and Trabert and of Gierer and Meinhardt were tested : Analysis of the linearized equations produced appropriate parameters and disk sizes for pattern growth. A computer-implemented finite element method was used to solve the non-linear model equations reiteratively. For the Gierer-Meinhardt model, initial activation (varying in size over two orders of magnitude) of one point on the disk's edge was sufficient to generate the primary gradient. Various parts of the disk were removed (remaining only as diffusible space) from the morphogen-producing cycle to investigate the effects of cells dropping out of the cycle due to cell death and malfunction (single point removed) or differentiation (center removed), as occur in Xcnopus eye-bud. The resulting patterns had the same general shape and amplitude as normal gradients. Nor did a 2-fold increase in disk size affect the pattern-generating ability of the model. Disk fragments bearing their primary gradient patterns were fused (with gradients in opposite directions, but each parallel to the fusion line). The resulting patterns generated by the model showed many similarities to results of compound eye experiments in Xcnopns, including both regulative and mosaic patterns. We conclude that the Gierer-Meinhardt model is remarkably stable subject to a wide range of perturbations in the diffusible space, thus allowing it to cope with normal biological variability, and offering an exciting range of possibilities for reaction-diffusion models as mechanisms underlying the spatial patterns of tissue structures. We thank NSF (PCM-77-26987) and Dean's Fund of Johns Hopkins University. Application of a polyploid marker to clonal analysis in Xenopus eye. ROBERT TOMPKINS, DANA REINSCHMIDT, KEVIN CONWAY, AND R. KEVIN HUNT. Clonal analysis of Xcnof>iis eye development has been undertaken by producing chimeric eyes, at embryonic stages, of normal tissues and tissues marked with the periodic albino (albp) or anucleolate ( 1-nu) mutant. Analysis of the pattern of marked cells in the resulting larval and adult eye facilitates understanding the growth of the eye and can be correlated with the generation of positional information in the chimeric retina, as revealed by its retinotectal map in a normal tectum. Limitations in the above markers led us to develop a new marker, tetra- ploidy (4 n). Tetraploid frogs, X. lacz'is, were produced by suppressing first cleavage in normal embryos. The low yield of 4n embryos and the high incidence of abnormalities do not allow direct use of pressed embryos ; however, selection of tetraploids by karyotypic and cell-size analysis permitted isolation of fertile tetraploid animals. The yield was one in 250,000 eggs pressed. The selected animals differ from tetraploid amphibia produced by other methods in that they produce eggs larger than the expected doubling of diploid volume, and the cell size of 4n offspring remain unexpectedly large (in retina and brain as well as many other tissues), at least through early larval stages. These large cells — as well as nuclear size, DNA con- tent, and nucleolar number — facilitate the identification of the orgins of individual cells in chimeric eyes as well as other experimental procedures (Hunt ct al., 1980, Biol. Bull., 159: 454). About 150 half-eye replacements and orthotopic wedge-grafts were performed at stages 31-36. Preliminary analysis of albino (diploid)/4n (pigmented) chimeric eyes showed that 4n tissues neither overgrow nor undergrow relative to diploid tissues. Our results confirm findings using other markers ; and the large cell size of this easily analyzed marker should make the 4n strain attractive to vertebrate neurophysiologists. Support was provided by NSF (PCM-79-03827; PCM-77-26987) and the Alfred Sloan Foundation. ECOLOGY Food preferences and population dynamics of the amphipod Talorchestia longicornis. GLORIA ALLENDE AND NANCY DISE. Two distinct populations of the amphipod Talorchestia longicornis were found grazing at night on a Massachusetts salt marsh, one along the shoreline in piles of Zostcra marina Ascophyllum nodosum, and Fucus vcsiculosis, and another on a blue-green algal mat on the lee side of the dunes. Feeding experiments in divided petri dishes revealed the algal mat population preferred blue-green algae and Enter omorpha sp. over two species of sulfur bact. also found on the algal mat. Both populations preferred Ascophyllum over blue-green algae and Zostcra, although the mat population showed a more pronounced difference in preferences. 456 ABSTRACTS FROM M.B.L. GENERAL MEETINGS The amphipods on the algal mat were smaller than the amphipods found along the shore ; mixing of the two populations may be prevented by ecological or physical barriers. A mark- recapture study of the algal mat population confirmed that the amphipods are highly mobile within their range, and allowed a late summer population estimate of approximately 10,000 adults. Microbial ecology of algal extracellular products: the specificity of alga-bacterial interactions. WAYNE H. BELL. Kinetics analyses of utilization of 14C-labeled algal extracellular products (EP) were performed upon a steady-state bacterial population maintained in continuous culture with the diatom, Skeletonema costatuin. The culture system selected for development of bacteria adapted well to this alga. Adaptation was evidenced by evolution of a kinetics pattern indicat- ing that both uptake and metabolism of 14C-labeled S. costatum EP were limited by the same substrate or substrates. Labeled EP from two other algal species, Thalassiosira pseudonana and Dunaliclla tertiolecta, produced kinetics patterns inconsistent with bacterial adaptation to these algae; such patterns included diffusion-limited transport and lines for uptake and respiration that failed to intersect at a common X -intercept. Nevertheless, the absolute rates of EP utilization were similar no matter what the source of the products. Competitive inhibi- tion analyses with unlabeled algal culture filtrates indicated that the major components of the EP pools of S. costatum and T. pseudonana are chemically similar and differ considerably from those from D. tertiolecta. Thus, although rapid uptake of T. pseudonana EP by bacteria adapted to 5\ costatwn can be explained by the similarities of EP pools of these two algae, rapid uptake of D. tertiolecta EP cannot be so explained. The results indicate that bacteria adapted to a given alga-mediated environment retain sufficient metabolic diversity to rapidly assimilate products from other algal species. Stimulation of bacterial activity by an algal bloom does not restrict bacterial metabolism to the principle components of the available EP pool, but enhances the microbial population's ability to rapidly utilize soluble organic compounds from other sources. The author is grateful to the Steps Toward Independence Program of the Marine Bio- logical Laboratory and to Hamilton College for support of this research. Studies oj methanogenic bacteria from intestinal tracts of marine fishes. HELMUT BRANDL, J. R. PATEREK, FRANCESCA MOLLURA, C. D. TAYLOR, AND E. P. GREENBERG. The intestinal contents of five scup (Stenototnus chrysops] and one smooth dogfish (Mustclus canis) were examined with respect to populations of methanogenic bacteria. Oxygen levels in the intestines of scup were less than 1 fj.M , the limit of detection of implanted oxygen- microelectrodes. This suggested that scup intestines may serve as anaerobic habitats for methanogenic bacteria. Samples of intestinal contents were serially diluted in an enrich- ment medium and incubated under an atmosphere of H2 and CO2 (80:20). The Hungate anaerobic tube technique was employed in order to avoid contact of samples with oxygen After 3-5 days negative pressure had developed within the enrichment tubes, indicating con- sumption of H2 and/or CO2. However, at this time methane could not be detected by gas chromatography. Since sulfate-reducing bacteria were found at densities of at least 10"/ml in intestinal contents from both S. clirysops and M. canis, these bacteria may have consumed H2 in the enrichments for methanogenic bacteria. After an incubation period of 20 days, methane was detected in the atmosphere above cultures from each of the fish, indicating the presence of methanogenic bacteria. Examination of material from enrichment-tube cultures by epifluores- cence microscopy revealed three different morphological cell types that exhibited the char- acteristic fluorescence of methanogenic bacteria : a Methanococcus-type and a large and a small Mcthanobacteriu»i-type. Apparently, methanogenic and sulfate-reducing bacteria coexist within intestines of certain fishes. This research was supported in part by grants from NASA (NAGW-72) and the Founda- tion of Microbiology. Distribution and migratory behavior of Ilyanassa obsoleta in Barnstable Harbor. G. A. BRENCHLEY. The distribution and migratory behavior of the mudsnail, Ilyanassa (Nassarius) obsoleta, was monitored at four sites in order to evaluate the snail's role in creating distributional hetero- ECOLOGY 457 geneity of infauna in Barnstable Harbor, Mass. Millions of adult snails (16-22 mm) immigrat- ing from the subtidal in the spring onto muddy, Zostcra-iree areas at Huckins Island and Calves Pasture Point became widely distributed in moderately high density (500 per m8). This pattern continued throughout the summer. In contrast, '/.osicra beds near low water con- taining the introduced periwinkle, Litlorina littorca, acted as migration barriers to the mud- snails, concentrating them into dense (1000+ per nr) patches which maintained their integrity as migrating swarms throughout the summer. By late June, about one million snails had entered between two extensive Zostcra beds on the sandflat off Indian Trail and were proceeding toward the marsh. Most reproductive individuals (16-22 mm) then moved east at 8-10 m per day toward Bone Hill Road, alternating between two /.. littorca habitats: Spartina in the high, and Zostcra m the low intertidal zone. Migration continued after the reproductive period, possibly related to the abundance of diatoms, sulfur bacteria, and currents. In the laboratory, /. obsolcta would not lay eggs in the presence of L. littorca; the latter dislodged, but did not ingest, mudsnail egg masses. Smaller individuals (8-14 mm) remained at the marsh edge near the initial point of contact. They expanded into the Spartitia only when L. littorca were rare or manually removed. Another population of small snails (8-18 mm) off Indian Trail remained over the study period within 4 m of the marsh edge. Patterns of infaunal dis- tribution and abundance correlate with abundances of I. obsolcta: patchiness was most evident on the sandflat containing migratory snails. Thus the structure of the benthic community is influenced by the presence of /. ohsolcta and L. littorca. Supported by a Steps Towards Independence Fellowship. Symbiosis of chemoautotrophic bacteria and marine invertebrates. COLLEEN M. CAVANAUGH. Transmission electron microscopic (TEM) examination and analysis of lipopolysaccharide indicate that procaryotic cells make up the bulk of the trophosome tissue in a newly discovered species of Vcstimentijcra found near deep sea hydrothermal vents. High activities of enzymes characteristic of sulfide oxidation and of CO^ fixation have been measured in the trophosome tissue of this species (H. Felbeck, Scripps Institution of Oceanography, pers. comm.). Thus trophosome procaryotic cells may provide an internal chemoautotrophic source of nutrition to these large tubeworms, which lack a mouth and a gut. Based on these studies, the possibility that chemoautotrophic bacteria are symbionts in other marine invertebrates was examined. Tests for ribulose-l,5-bisphosphate carboxylase activity were positive in Solonya velum, an Atlantic coast bivalve known to have a very small gut and to live in sulfide-rich sediments. The enzyme activity appears to be located in the gills and/or mantle cavity. Initial examination of Solcmya gill tissue with TEM indicates the presence of membrane-bound procaryotic cells which appear to be intracellular. Studies are in progress to determine the potential chemo- autotrophy of the procaryotes found in Solcmya and their relationship to the animal. Similar procaryotic inclusions can be seen with TEM in the gill tissue of a new species of large white clam found at the deep sea vents. The possible contribution by chemoautotrophic sym- bionts to the nutrition of these animals remains to be assessed. This new type of symbiosis may be important to some marine invertebrates inhabiting sulfide-rich environments. This research was supported in part by the National Science Foundation grant PCM 79- 06638 and the M.B.L. Microbial Ecology Course. The predatory habits of two lycosid spiders in Great Sippewissett Marsh. MICHELLE CORK AND DAVID HUGHES. Prey preferences for two lycosid spiders were investigated. Gcolycosa pikci, found in sand dunes near the algal mat, was observed in laboratory experiments to prey upon Talorchcstia longicornis, an amphipod found in high densities around the algal mats. Gcolycosa, 13 mm in body length, showed size preference toward small Talorchcstia, approximately 5 mm in length. Lycosa hcllue, found in the high marsh near the strand line, was placed in cages in the field with known numbers of three types of prey: Orchcstia (jrillns. an amphipod; Melampus bidcntatus, a snail; and Philoscia •uittata, an isopod. Of the three, isopod mortality was highesl but differences in mortality between experimental and control cages with prey only not statistically significant. Amphipods and Melampus may have been too large for the Lye to prey upon. More replications, with smaller amphipods and snails, are necessary to : substantial conclusions about this predator (Lycosa) -prey interaction. 458 ABSTRACTS FROM M.B.L. GENERAL MEETINGS The relationship between diet and growth rate of the grass shrimp Palaemonetes pugio. RANDY CHAMBERS AND ANTHONY PIRES. The effect of diet on the growth of juvenile grass shrimp, Palaemonetes pugio, was investigated. The laboratory experiment was a Latin square design, with five shrimp monitored for each of five diets (mussel mantle, control and nitrogen-enriched sediments, and two algal species, Entcromorplia intcstinalis and Gracilaria I'crrucosa). Growth on each diet after 20 days was recorded as percent increase in total body length. Only the mussel diet yielded growth comparable to field population growth, averaging a 57% length increase (the largest on any experimental diet) over the course of the study. No significant differences (P = 0.05) in growth were shown between sediment diets (control, 31%; nitrogen-enriched, 24%), although growth on an Entcromorpha diet (25%) was greater than growth on Gracilaria (11%). Length increases were correlated with nitrogen content and palatability of the foods. These results suggest that in the field, Palaemonetes pugio must find concentrated protein sources (for instance, meiofauna or food items scavenged from other predators) that can be translated into fast growth. Mechanism of heavy metal inhibition of amino acid transport in the intestine of marine fishes. A. FARMANFARMAIAN, ROBIN Socci, AND THOMAS POLIDORE. Heavy metals, such as cadmium and mercury, are released from various pollution sources into coastal waters and sediments. Marine fish accumulate heavy metals via food. Small invertebrates, such as annelid worms and clams, which are important sources of food for demersal fishes, have been shown to contain high levels (10-30 ppm) of heavy metals. The effect of such toxicants, released from food in the gastrointestinal canals of fishes, are not known. We have started a project examining the effects of heavy metals (CdCla, CHaHgCl, and HgCla) on digestive-absorptive functions in the intestines of several marine fishes. The absorption of 14C-labelled L-leucine from buffered fish Ringer at 20°C in vivo and in vitro was measured in the presence and absence of heavy metals at three concentrations (2- 30 ppm). In toadfish in vivo experiments CdCL- and CH.iHgCl had no significant effect but HgCIj inhibited the absorption rate by 57 and 80% at approximately 10 and 20 ppm of mercury respectively, after 10 min of incubation. In sea robin in vivo experiments intestinal tissues were incubated for 10 min. The uptake of L-leucine was not affected by CtLHgCl but in the presence of HgCL> 21 and 44% inhibition was observed at the respective concentrations mentioned. Similar levels of inhibition were recorded for winter flounder when tissues were exposed to HgCL. The oxygen uptake of the tissue, although inhibited by high levels of HgCU, is not affected as drastically as amino acid uptake during a 10 min incubation. This indicates that Hg"* inhibition involves a direct binding to the intestinal membrane transport mechanism and does not lower amino acid uptake secondarily via reduction in cellular energy production. This view is further supported by the observation that CHnHgCl, in which mercury has a single positive charge, does not cause appreciable inhibition of leucine transport but reduces oxygen uptake significantly. Mapping vegetation and topography of Crcat Sippewissett Salt Marsh, Mass. AMY FRIEDLANDER, FRANCIS BOWLES, JUDITH GALE, AND BRUCE PETERSON. Great Sippewissett Salt Marsh on Cape Cod is a small pocket marsh of approximately 50 hectares. Mapping field work was conducted in the field seasons of 1979 and 1980. The resulting map is to serve three functions : to provide a detailed vegetation map that can be used as a baseline for comparison in the future, to provide a finely contoured relief map that will allow estimates of water volume in the marsh at different stages in the tidal cycle, and to determine the area covered by each major grass species, and the extent to which these areas are exposed or flooded at different tide heights. Approximately 4000 data points of elevation and vegetation were taken at borders of vegeta- tion types and/or creeks. Elevations were determined by differential leveling, with a level and ECOLOGY rod and metric tape. The elevations are referenced to U.S. Geological Survey benchmarks, which provides a baseline datum of mean low water. The data have been used to determine location and elevation of sites used in peat v i eolation, litter decomposition, and fertilized plot studies. They have also been used to determine the elevation of a tide gauge, and in studies of slope of the water's surface, which have shown that in ebbing and flooding tide there is a maximum slope of as much as 30 cm from the upper reaches of the marsh to Buzzards Bay. Anticipated uses include analyzing present vegetation patterns and possible future changes, looking at the structure of the barrier beach and subsequent changes in its shape by dune migration or erosion, and calculating volumes of water held in the marsh at various tide heights. We thank Ian Bowles, Benedicte Misner, Carol Xilson, and Susan Pilling for their assistance in the field. This work was supported by the National Science Foundation under grant DEB 78-03557. The fea-sibility of seaweed aqiiaciilfure in the Great Sippeicissctt salt marsh. RODNEY M. FUJITA. Salt marsh tidal creeks provide certain advantages for the cultivation of marine organisms. High natural concentrations of inorganic nutrients and water exchange due to tidal currents obviate the need for commercial fertilizers and artificial water exchange in the cultivation of marine macroalgae. Gracilaria rcrrucosa and Chondrus crispus. red algae valuable for their phycocolloid content and potential energy yield through byconversion to methane, were grown in floating cage culture at two sites in the Great Sippewissett salt marsh : in a creek draining an area which had been fertilized for 1 year, and in an unfertilized control creek. During July, Gracilaria in the control creek grew at a rate (10% -day"1) lower than but comparable to those achieved in intensive tank culture, but did not grow well in the fertilized creek. Chondrus failed to grow at either site. During August, Gracilaria in the control creek grew at a rate of 18% -day"1 before becoming infested with tunicates. Since the cages in the fertilized creek quickly became covered with Entcromorpha, an epiphytic green alga, it was thought that water exchange and light intensity limited growth. These results suggest that fouling will be a major problem for aquaculture in naturally eutrophic marsh environments. Two possible methods for controlling epiphytes were investigated. Gracilaria was grown in the control creek and exposed to the fertilized creek every 10 days for 40 hr. This treat- ment eliminated Entcromorpha, while growth rate remained comparable to that of untreated controls. The ability of various marsh animals to selectively graze Entcromorpha was investi- gated in the laboratory. It was found that all animals tested preferred Entcromorpha to Gracilaria, and thus have potential as biological control organisms. Statistical mechanics of geomagnetic orientation in sediment bacteria. MICHAEL K. GILSON AND AD. J. KALMIJN. Last year we reported on time-of-transit experiments in which magnetically orienting bacteria crossed a 1-mm stretch in the direction of a uniform magnetic field. The bacteria were found to behave as tiny self-propelled compass needles subject both to magnetic field alignment and to the randomizing effect of thermal agitation. In strong fields, magnetic bacteria are held in tight alignment ; in weaker fields, their swimming paths meander more and transit times are greater. Paul Langevin derived an expression for the distribution of orientation in an ensemble of free-moving dipole particles as a function of ambient field strength. His theory becomes applicable to our experiments when bacterial migration is analyzed as a sequence of short steps during each of which the cell swims in a direction randomly selected from the Langevin distribution. The duration of each step, At, is actually a time constant of the cell's loss of directionality due to thermal agitation. By thus treating the migration as a process of random walk with drift, we are able to predict the mean and variance of the time of transit across a 1-mm stretch. The behavior of the model depends on three parameters: randomization time At, the cell's intrinsic dipole moment m, and the speed of propulsion We use nonlinear regression analysis to estimate these parameters and to fit the beha.\ of the model to that of the bacteria. We also determine the goodness of fit of the model in entirety, and the approximate confidence limits of the parameter estimates. The estimated ran- domization times are in accord with preliminary calculations of rotational diffusion rates. 460 ABSTRACTS FROM M.B.L. GENERAL MEETINGS dipole strengths agree well with those expected on the basis of the number and size range of the bacteria's intracellular magnetite crystals. Our values are slightly lower due to the inevitable impurities and imperfections in alignment of the crystals, and to additional agitation resulting from swimming movements. In short, the dipole moments direct the bacteria magnetically despite thermal agitation and swimming noise. As statistical mechanics suffice to explain the orientation of magnetic bacteria, there is no need to invoke an active orientation mechanism. (Kalmijn's project on electric and magnetic detection operates under the auspices of the Office of Naval Research, Oceanic Biology Program, X00014-79-C-0071.) Larval settlement on inicrobial films: A model system. STEPHEN GRAHAM, DAVID KlRCHMAN, AND RALPH MlTCHELL. We have found a tube-forming polychaete, Janna (Dcxiospira) hrasilicnsis ( Sedentaria : Spirorbidae) useful in studies of larval settlement on microbial films. Spirorbid and serpulid worms are important marine fouling organisms. Janua is a good model animal for several reasons. It is small (2-3 mm) and hermaphroditic. It is abundant on a variety of surfaces, especially on Zostcra (eelgrass) from Woods Hole to Buzzards Bay, MA. At least 50% of a typical population is brooding eggs at any one time. Larvae are readily obtained from external brood sacs located in the operculum, and most (60%) settle and metamorphose within 2 hr. One limitation we encountered in Woods Hole was that fertility rates declined over the summer, possibly associated with seasonal temperature. By comparison, Janna sampled off Long Beach, CA, produced ample numbers of larvae (65% of adults brooded eggs). We demonstrated that larvae prefer to settle on surfaces coated with microbial films. Settlement and metamorphosis does not occur on films of the diatom Nitzchia. In the absence of microbial films, Janua larvae failed to settle. Gama-aminobutyric acid did not induce settling. Uni-bacterial cultures of Pscudomonas marina and a few other specific bacteria induced settle- ment and metamorphosis, but most of the marine bacteria tested lacked this ability. We attempted to identify the metamorphic trigger produced by the bacteria. Chloramphen- icol did not block the triggering action of the bacteria, providing evidence that protein synthesis is not essential for settlement. Formalin-treated films did not inhibit metamorphosis, indicating that viable cells are not required. No metamorphosis was detected on surface films prepared from either cell wall components or intracellular material produced by lysis of the bacteria. The metamorphic trigger may be associated with either extracellular polysaccharides or other loosely-bound bacterial polymers. This work was supported in part by NOAA grant NA79AA-D-00091 and ONR contract N00014-76-C-0042 to Harvard University. The relationship bctivecn organic and nitrogen content of marsh sediments and feed- ing rates in Uca pugnax. ERICH F. H ORGAN AND MARGARET S. RACE. The ability of the deposit-feeding crab Uca pugna.r to alter its feeding rate in response to sediments varying in organic and nitrogen content ( % total dry weight ) was investigated in the laboratory. Crabs starved for one day were observed feeding on six different sediment types: the first three sediments contained 45% organic matter with nitrogen values ranging from 1.26 to 1.35%, while the second three sediments contained 20% organic matter with nitrogen values ranging from 0.38 to 0.63 %. Under both organic content regimes, feeding rates were inversely correlated with the nitrogen content of the sediments : as nitrogen content increased, feeding rates significantly decreased. These results suggest a relationship between feeding behavior and nutrient turnover in benthic systems. We extend thanks to Ivan Valiela and John Teal for their helpful ideas, and appreciation to the Founders of the Marine Biological Laboratory Scholarship Fund in assisting in making this research possible. Preferred food sources and the limitation of local distributions of the isopod crusta- cean Philoscia vittata. DAVID HUGHES AND MICHAEL USEM. Food preference of the isopod Philoscia vittata, testing Spartina plant parts and detrital material, was determined in laboratory experiments. Preference, measured by the observed production of fecal pellets in petri plates containing agar suspensions of the foods, was marked ECOLOGY 461 for detrital material from two sources. No significant difference between preference for S. patens or S. alterniflora detritus was found, suggesting that food choice does not determine distribution of the isopod. In the field, isopods were caged in areas of the marsh above and below areas of known distribution to test the effects of water immersion and dessication. Sur- vivorship after 8 hr under water was 100% ; survivorship in wooded areas bordering the marsh was variable, increasing with the amount of grass cover. Dessication may limit the upper dis- tribution of the isopod in the marsh, but no lower limit factors were identified. Tidal ivater exchanges between Great Sippewissett Salt Marsh and Buzzards Bay. DAVID W. JUERS, FRANCIS P. BOWLES, AND BRUCE J. PETERSON. The construction of a sulfur budget for Great Sippewissett Marsh requires accurate esti- mates of water volume exchanges between the marsh and Buzzards Bay. These estimates are derived from 24-hr continuous records of tide height and current velocity made during monthly samplings. To examine the patterns of water transport implied by these data, ebb and flood volumes were measured on four tides which had amplitudes ranging from 0.85 to 1.25 m. Throughout each tidal cycle, water velocity measurements were made with an EMF-type current meter, at 20 cm depth intervals every 4 m across the channel behind the inlet to the marsh. The area at each location was computed from depth measurements taken during the tidal cycle. A second current meter, connected to a data logging system, continuously recorded water velocity 30 cm from the surface, at the first transect location. Flow rates at each location were calculated and summed to yield total water flow. Integrated over time, these measured rates yielded tidal exchange volumes which varied from 85,000 to 220,000 m3 for flood tides, and 110,000 to 160,000 m' for ebb tides. Predictions for flood tides were obtained by fitting a modified logistic equation to the cumulative volume curves. Further conditioning the equation with factors relating peak water velocity to tidal amplitude, and tidal amplitude to the length of the flood tide phase, yielded predictions of total volume that differed from those measured by only 3-8%. For any time during the ebb, the amount of water leaving the marsh was found to be a linear function of the water volume input on the flood tide. This work was supported by NSF DEB 78-03557. Bacterial epiphytes on Zostera marina surfaces. D. L. KIRCHMAN, L. MAZZELLA, R. MITCHELL, AND R. S. ALBERTE. A complex microbial community, comprised of bacteria, diatoms, and other microorganisms, is present on Zostera marina L. (eelgrass) leaf surfaces. To increase our understanding of the relationships between the epiphytic community and Zostera, we tested whether bacteria on the leaf surface obtain organic carbon from the leaves to support their production. Individual Zostera leaves were placed with their tips in dark chambers and their bases in illuminated chambers. Sodium [14C] bicarbonate was injected into the illuminated chambers, allowing radio- activity to appear in the dark chambers only by translocation through the Zostera. Epiphytes were removed from the Zostera leaf surface with a razor blade with no detectable damage to leaves as determined by microscopic examination. Nearly 20% of the total translocated radiolabel was recovered in the dark chamber epiphytes. The absolute amount of radioactivity was approximately three times background levels. Radioactivity in the dark chamber epiphytes can only be due to epiphytic bacterial uptake of organic carbon from the Zostera, since dark "CO* fixation is extremely low. In order to eliminate possible accumulation of radiolabel independent of epiphytes on the leaf surface in the dark, epiphytes were removed from leaf surfaces, and then the leaves were incubated as described above. The amount of radiolabel associated with the epiphytes in the dark chamber dropped to only 2% compared with 20% in the previous experiments in which epiphytes were present. This amount was barely above background levels. An estimate was made of epiphytic bacterial production specifically supported by the Zostera tissue. Bacterial production per cell was estimated to be (1.3-13) X 10~7 /igC-hr"1, based on literature values. Bacterial numbers on Zostera surfaces were approximately 107 cm"2. Total bacterial production was then calculated to be 1.3 to 13 MgC-hr^-arr2. Since measured carbon fixation rate of Zostera was 3.5 /ugC-hr^-cnr2, and since recovery of from Zostera photosynthesis in the epiphytes was 20% of that translocated, it was calcula that 0.7 MgC-hr^-cm"2 was available to the epiphytes from the Zostera. Therefore, ', of total bacterial production could be supported by carbon fixed by Zostera alone. These 462 ABSTRACTS FROM M.B.L. GENERAL MEETINGS calculations and the experiments described above suggest that a significant amount of epiphytic bacterial production is supported directly by Zostcra photosynthesis. This work was supported in part by NOAA grant NA79AA-D-0091, ONR contract X00014-76-C-0042, and NSF grants PCM 79-06638 and PCM 78-10535. Contextual relationships in food wchs inrolviin/ nieiofanna. JOHN J. LEE AND MONICA J. LEE. Conceptual models of energy flow in shallow water benthic marine food webs involving small animals (microfauna and meiofauna) and microflora have a common weakness. They fail to account for food-quality-related aspects of energy flow. Potential energy in these com- munities as detritus, or as the substance of bacteria, microphytes, and fungi is always in excess. Gnotobiotic nutritional experiments suggest most successful protozoa and micrometazoa have the potential to optimize their energetic needs by selectively consuming microfloral species. The properties of quality related (or informational) energy flow are not only intrinsic to the molecular constitution of food, but also to molecular processing after consumption and nutritional requirements of consumers. Enigmatically, most meiofaunal species raised gnotobiotically are fecund on diets of algae with low abundances. It could be argued that nutritional experiments involving one or only a few potential food species are misleading. They present fewer choices than encountered in the "real" world. In the absence of competition the variety of microorganisms available as potential food in any benthic community should satisfy nutritional requirements of small grazing herbivores. We have been testing this hypothesis. Selected species of meiofauna were inoculated into natural assemblages of aufwuchs microflora and detritus ( from which all animals had been mechanically removed) and incubated either in the nearby Greater Sippewissett salt marsh or in the lab in tissue culture flasks with nylon windows (3 nm pore size) to permit free passage of sea water. Of the four animals tested, only Allogromia laticollaris, a foraminiferan, steadily increased in all natural assemblages into which it was introduced. Populations of Chromadorina i;cr- manica (nematode) and Nitocra typica ( harpacticoid copepod) reproduced vigorously in some natural assemblage contexts but were unable to do so in others. Rhabditis marina (nema- tode) failed to reproduce in any of the nutritional contexts tested. Diatom assemblages are being studied to see if the animals modify the communities in which they grazed. Supported by NSF grant OCE 7929485. The effect of chromium on microbial activity in salt marsh sediments. BRUCE LEIGHTY. A basis of marsh productivity is the decomposition performed by the microbes of the salt marsh sediments. As the health and performance of these microbes are functions of the environment, the effect of pollution is of great importance. This study was conducted to test the effect of a heavy metal pollutant, chromium, on microbial activity in salt marsh sediments. Two experimental plots with controls were set up in a stand of uniform Spartina altcrni- flora (short form). The experimental plots were subjected to thrice weekly applications of 10 ppm Cr solutions. Scrape samples from the aerobic surface of the sediment were collected and homogenized in the laboratory. Inoculums were placed in blackened biological-oxygen- demand bottles with 1 cc of sterile substrate and filled with 50% sterile seawater. Oxygen levels in the bottles were monitored every 12 hr for 48 hr. Exposure to Cr was found to reduce oxygen uptake by 80% after 12 hr. However, when samples were incubated without Cr, oxygen uptake immediately returned to normal, suggesting that exposure to Cr for the length of this study had no long term inhibition on microbial activity. In tests for metal tolerance, both experimental and control plot microbes were incubated with Cr. It was found that following Cr applications, the microbes from the experimental plots were more tolerant of the presence of Cr and exhibited greater oxygen uptake than the con- trol plot microbes, implying that Cr resistance had developed. To test if this tolerance was a specific response to Cr, or a general response to metal stress, microbes were also incubated with copper. Experimental plot microbes were also found to be more tolerant of Cu stress than controls. ECOLOGY 4ft.? Interactions between two species of littorines — Littorina littorea and L. saxatilis— along Nen.' England shores. RANDA A. MANSOUK. Aproximately 110 years ago the snail I.ittorina littorea was introduced to Nova Scotia from Western Europe. /.. littorea has since spread as far south as Virginia. In New England it has become the most abundant periwinkle on both exposed and sheltered shores. /,. littorea prefers moist areas and is found at high densities from the low- to mid-intertidal. The native periwinkle, L. saxatilis. inhabits the mid- to supra-intertidal, increasing in abundance in the higher tidal zones. The changes that occurred with the introduction of L. littorea have not been documented. The purpose of this study was to determine the nature of the interaction between L. littorea and /.. saxatilis. On 24 June 1980, 18 cages measuring 8x8x3 cm were attached to the pilings of a private dock near Nobska Point, Cape Cod, Mass. Cages were placed in the supra- and mid-intertidal zones. Ten L. littorea, 10 /.. sa.ratilis, or 5 animals of each species were assigned to a cage. One cage at each level was empty. On 12 July 1980 snails were removed, growth was measured as lip increments, and the relative food abundance in each cage was determined. Both species grew significantly more in the mid- than in the supra-intertidal. Intraspecific competition for food limited the growth of L. littorea in the supra-intertidal. At the mid- intertidal L. littorea depressed the growth of L. saxatilis. These results indicate that in the absence of L. littorea. L. sa.ratilis would have a broader vertical range. Cadmium resistance in bacteria from the Great Sippewissctt Marsh. F. MOLLURA. Plots of Spartina in the Great Sippewissett Marsh have been fertilized with commercially- prepared sewage sludge containing substantial amounts of heavy metals. The study addresses the question of whether cadmium (Cd) resistant strains of bacteria have developed, and begins preliminary study regarding mechanism of detoxification. Isolates of bacteria from the upper centimeter of marsh sediments were obtained from sludge-fertilized and unfertilized plots of Spartina. These were tested for ability to grow on Cd~+ supplemented and unsupplemented agar media. Two types of populations were grown, one on nutrient-rich Zobell's (2216E) agar and the other on diluted medium (2216E/ 100). Clones that appeared to be Cd resistant by their ability to grow on 10" ' M Cd2* supplemented 2216E were found on the fertilized plots but not on the unfertilized one. More colonies of resistant bacteria were found on the plot fertilized for 6 years than the plot fertilized for < 1 year, suggesting that adaptation and ability to grow in the presence of Cd2t requires time. Bacteria grown on the diluted medium were unable to grow with 10~:1 M Cd2+, and grew very poorly on 2216E/100 with 10"4 M Cd2*. These same bacteria grew very well on 2216E with 10~4 M Cd2*. Isolates from 2216E with 10"'1 M Cd2* were unable to tolerate 10"3 M Cd2* concentration on 2216E/100. This suggests that the presence of greater amounts of organic material may provide a more favorable environment for detoxification. This work was supported in part by grants from NASA (NAGW-72), The Foundation of Microbiology, and Le Moyne College, Syracuse, NY. Gratitude is expressed to Dr. Jeanne S. Poindexter for her assistance with this project. Control of drilling fluid discharge from petroleum development on Georges Bank. ELIZABETH MULLIN. Drilling muds, which are pumped down the bore hole to cool the drill bit, control pressure, and bring cuttings to the surface, may be discharged during drilling for oil and gas on Georges Bank. Studies to predict their fate and effects on marine ecosystems have yielded conflicting and ambiguous results. However, studies do suggest that drilling fluids could be widely dis- persed throughout the water column and concentrated in certain depositional areas. There is also evidence that drilling fluid components, which include heavy metals and bactericides, may cause lethal or sublethal effects on marine hiota. Several strategies are available to mitigate possible impacts. Transportation of used muds to shore for land-based recycling or disposal, or reinjection of muds into the rock forma- tion would minimize the potential for contact with marine organisms. Piping or barging of effluent off the continental shelf could reduce risk to the fisheries of Georges Bank, volume of muds entering the ocean would not change. 464 ABSTRACTS FROM M.B.L. GENERAL MEETINGS The oil companies favor shunting drilling mud through a pipe into the water column, but the efficacy of this technique is questionable Shunting is unlikely to localize impacts because of circulation patterns in the region. Controls over discharge rates and dilution of muds before discharge may reduce initial concentrations, but, like shunting, do not reduce the amount of mud discharged on Georges Bank. Physical or chemical treatment of wastes before dis- charge could provide some additional protection. Altering the composition of muds could reduce the amount of toxic components. However, chronic, low level contamination of Georges Bank may cumulatively or synergistically affect the fisheries. Food choice and palatability in a salt marsh dctritivorc, Melampus bidentatus. CAROL S. RIETSMA. Detritivores play an important role in decomposition of higher plants by comminuting and assimilating detritus. A number of properties of detritus affect its palatability to detritivores. This research investigated effects of plant species, age, nitrogen content, amino acid content, ferulic acid content, salinity, and pH of the detritus on feeding by the common salt marsh detritivore, Melampus bidentatus. Relative palatability was measured by counting the number of feeding marks left by snails on the surface of agar suspensions of detritus in petri dishes with four compartments. The snails were offered the choice of feeding on either two or four different foods. Melampus bidentatus preferred to feed on detritus from Juncus gcrardi over that from Spartina alterniflora and S. patens, both of which were preferred over Distichlis spicata. Natural aging of detritus from S. alterniflora in the field reduced its palatability. Irrespective of age, a higher nitrogen content enhanced its palatability, whereas phenylalanine and leucine reduced it. Aspartic acid, glutamic acid, glycine, and betaine had no effect. Ferulic acid, a feeding inhibitor, interacted with salinity and pH to affect the palatability of old detritus. Properties of detritus such as its plant species, age, and nitrogen content, etc., determine whether or not it is consumed by detritivores and thus affect the rate of decomposition of detritus. This research was supported by an NSF grant DEB 7905127. Denitrification potentials in a snccessional sequence oj northern hardwood forest stands. ]. P. SCHIMEL, P. A. STEUDLER, J. M. MELILLO, J. G. WARREN, C. M. ZACKS, AND J. D. ABER. Denitrification potentials were measured using an acetylene technique in four north-central New Hampshire hardwood forest stands — 2, 3, 30, and 50 years since clear cutting. The pro- cedure used involved the following steps: (1) short term (less than 2.5 hr) in situ incubation of freshly taken, unamended soil in a closed vessel with an atmosphere containing 10% acetylene in nitrogen; (2) periodic sampling of gas in the incubation vessel and storage of the gas samples in Vacutainer tubes; and (3) analysis of gas samples for nitrous oxide concen- trations using "''Ni electron-capture gas chromatography. Soil samples were analyzed in the lab for moisture, organic matter, pH, and nutrient concentrations. The results of our study show that denitrification potential was highest in the forest floor of the 2-year-old stand (560 ng N2O-N-g organic matter"1 -hour"1), intermediate in the forest floors of the 50- and 3-year-old stands (150 and 47 ng N2O-N-g organic matter"1 -hr"1, respectively) and lowest in the 30-year-old stand (5.8 ng N2O-N-g organic matter"' -hr"1). Po- tentials in the mineral horizons were comparable to the forest floor except in the 2-year-old stand which gave lower values (180 and 360 ng N^O-N-g organic matter'1 -hr"1 for the A and B horizons). Within stand variability among denitrification potential measurements was high (S.D./ mean — 1). Correlations of soil nitrate, ammonia, and total inorganic N to denitrification potentials were high (r>0.9) across the successional sequence, but low within each stand (r<0.5). These findings indicate that northern hardwood forest ecosystems, especially recently cut systems, have the potential for nitrogen loss by denitrification in spite of their high acidity (forest floor pH range 3.5-3.9), and their generally well-drained condition. ECOLOGY 46- Effcct of yrasiny by Uca pugnax on the microbial population in salt marsh sedi- ment. ANNETTE SPIES AND FREDRIC LII-SCH n/i/. The fiddler crab, Uca f>u.>issett Salt Marsli Falmouth, Mass. The effects of Ucu grazing on sediment microbial populations was deter- mined by caging Uca on 0.2 nr/plots, and then counting bacterial numbers by acridine-orangr epifluorescence. Uca were either enclosed in caged plots at 28 animals -cage'1 or excluded entirely. Caged and control plots were established on silty sediments of creek bottoms and mudflats where Uca primarily grazes. Small sediment cores were removed at 1, 2, 7, 14, and 21 days and the top 1 cm and subsurface 2 cm sectioned and preserved in 10% formalin. Bacterial counts were done on 10 random fields at 100X magnification. Bacterial numbers doubled from approximately 2 X 101" cc'1 to 4 X 10"' cc~' within 48 hr in the mudflat exclusion plots compared to controls. This was statistically significant (P < 0.05). Xumbers remained high for a week but declined to control values by Day 21. Popu- lation densities at the creek site also increased rapidly and then declined. Meiofauna were enumerated from the plot sediments after 21 days : As with bacterial numbers, populations of worms, protozoans, polychaetes, and copepods were identical in control and crab inclusion plots. Mudflat exclusion plots, however, had four times as many worms and six times as many protozoans as the controls. This was statistically significant (/J<0.05). Removal of a major predator, Uca puyna.v, from the sediment surface resulted in rapid increase of bacterial numbers from steady state levels, followed by greatly increased meiofaunal numbers, possibly in response to higher bacterial prey densities and release from crab predation. Effects oj closed-culture competitive interactions on t/rou'th of Teredo navalis larvae. GREGORY A. TRACEY, CARL }. BERG, AND KTTH I). TURNER. Larvae of the shipworm Teredo navalis were reared in closed culture under experimental conditions designed to distinguish between genetic variation and possible size-dependent com- petitive interactions. No differences in growth rates, survivorship, or maximum size of the larvae after 14 days of growth were observed for culture densities of 250-1000/1 and volumes of 1-3 1. Differential growth rates were observed among larvae obtained from populations divided by sieving into large and small fractions at 90 and 120 /u. mean height. In both cases, the growth rates of the smaller fraction were greater than the larger fraction or controls until larvae had achieved a size of 140 n in height, at which point their size-frequency distributions overlapped. This phenomenon cannot be totally explained by genetic variation, as the size- frequency distributions of large and small larvae, seen as a bimodal peak in the control groups, remained distinct throughout this period. It remains unclear what mechanism of competition might be occurring, or if size-dependent competitive interactions are possible in the natural habitat. The effect oj sewage fertilization on benthic macroinvertebrates in salt marsh creeks. TIMOTHY UYEKI AND WENDY WILTSE. Species composition and population densities of benthic invertebrates were studied in salt marsh tidal creeks treated with three dosage levels of sewage sludge fertilizer (0.8, 2.5, and 7.5 g N • irT2 • week"1 ) , urea fertilizer (2.5 g N-m"1'- week'1), and untreated controls. Oligochaetes and polychaetes predominated, with the greatest number of species occurring in creeks treated with the highest dosage of sewage fertilizer. Densities increased during the spring, reached a maximum in early summer, and decreased dramatically by August under all treatments. The highest density (340,000 m~') occurred in June at the highest level of sewage fertilization, suggesting that food availability influenced recruitment or survival of benthic invertebrates. Densities below 100 irT" in all creeks in August indicated that intense fish and crab predation depleted populations of all species during summer. Sewage fertilization may increase produc- tion of benthic invertebrates in salt marsh creeks, but increased densities were evident only in early summer when predation was not intense. Rifampin as a selective agent for the isolation and enumeration oj Spirochaeta f salt marsh habitats. FREDERICK H. WEBER AND F. P. GREEN BERG. Members of the genus Spirochaeta from various salt marsh habitats were enumerate isolated directly using a prereduced complex agar medium containing rifampin as the 466 ABSTRACTS FROM M.B.L. GENERAL MEETINGS agent. In the absence of rifatnpin, Spiroclwcta-type colonies constituted 7% of the total colony- forming units (CFU) from samples of the top cm of sediment from a pan in the Great Sip- pewisset salt marsh. At a rifampin concentration of 10 /tg/ml, Spirochacta-type colonies con- stituted 56% of the total CFU. At higher rifampin concentrations (30 ^g/ml) formation of Spirochaeta colonies was inhibited. Thus the medium used in subsequent experiments con- tained rifampin at 10 ng/m\. Spirochetes were enumerated and a variety of morphological types isolated from the top cm of sediment from several habitats in Great Sippevvisset salt marsh including Spartina stands, cyanobacterial mats, sandy and muddy creek bottoms, and pan areas. Spirochacta-type CFU generally ranged between 104 and 4 X 105/g wet weight of sediment but exceeded 10"/g wet weight in two samples from pan areas in the Barnstable marsh. Studies of vertical distributions in cores of the top 5 cm of sediment from pan areas and Spartina stands revealed that Spirochaeta-type CFU were most prevalent in the upper 1-2 cm — generally at least 10-fold more frequent than in deeper sediments. Apparently, the habitats studied provided favorable environments for certain members of the genus Spirochaeta. This research was supported in part by grants from NASA (NAGW-72) and the Founda- tion of Microbiology. The effectiveness oj National Park Service policies in protecting barrier island eco- systems within the Cape Cod National Seashore. JOHN P. WARGO. This paper analyzes the effectiveness of National Park Service policies in protecting barrier island ecosystems within the Cape Cod National Seashore, established in 1961. The types and intensities of uses permitted by Park Service policies were analyzed. The impacts of such uses on beach, dune, salt marsh, and mud flat ecosystems were evaluated. Recent literature indicates that structural development and use of oversand recreational vehicles has had substantial adverse effects on the stability of these dynamic ecosystems. Legislation permits continued private ownership of improved property within the park boundary, if local governments adopt zoning ordinances meeting Interior Department stan- dards. Development not meeting such standards is subject to federal condemnation. Zoning ordinances adopted by towns have been effective in preventing development of new residential, commercial, and industrial uses on private lands within the Seashore. They have not been effective in preventing the substantial expansion of existing residential uses. Present Park Service policy permits the reconstruction of residential structures on barriar island ecosystems if the homes are destroyed by storms. Park Service policy permits the use of recreational vehicles on established routes. In 1979, 30,000 vehicle trips occurred on federal lands within the Seashore and 5000 vehicle trips on town land. Legislation permits continued town ownership of land within the Seashore. The Park Service has no control over structural development or the use of recreational vehicles on more than 2200 acres of such land. The Park Service has no method but condemnation to enforce land-use regulations on private land. Enforcement is therefore dependent upon the availability of funds to buy land. FERTILIZATION AND REPRODUCTION Change in cvclic nucleotide content during germinal vesicle breakdown in Spisula oocvtes. ( )YEVVOLE ADEYEMO, HIROKO SIIIRAI, AND SAMUEL S. KOIDE. Spisula solidissinni prophase-ar rested oocytes can be activated by insemination or applica- tion of chemicals. Reports suggest a relationship between germinal vesicle breakdown (GVBD) and cyclic nucleotide content, r.//. in Xcnopus lucrts, injection of a regulatory sub- unit of cAMP-dependent protein kinase induces GVBD while injection of a catalytic subunit inhibits GVBD induced by progesterone. In Kaua pipicns a reduction in cAMP and cGMP occurs shortly after progesterone application. In the present experiment fluctuation in cyclic nucleotide content during GVBD in Spisula was investigated. Oocytes from 5-7 animals were suspended in scawater (0.5 nil packed oocytes/5 nil). GVBD was induced by adding 0.5 ml of 0.5 M KC1 or 1/30 dilution dry sperm to 5 nil of oocyte suspension. At appropriate time intervals ( 0, 1.2, 4, 6, 8, 10, 20 min) oocytes were collected by centrifugation. To the pellet 5 nil of 0.5 M perchloric acid in 25% ethanol (v/v) was added. The mixture was homogenized and centrifuged. The supernatant was neutralized with solid KHCO:i, centri- FERTILIZATION AND REPRODUCTION 467 fuged, and dried (55C.N,). Assay buffer (0.5 nil of 0.05 M Tris-HCl, 4 mM EDTA, pH 7.5) was added and the pH adjusted to 7.5. We estimated cGMP and cAMP (four experiments) by radioimmunoassay and the values expressed as pmol/10" oocytes. The ratio of cAMP to cGMP content in unstitnulated oocytes was 1.7. There was an initial consistent fall in cGMP level (2.8) 1-4 min after stimulation of GVBD, followed by a return to the zero-time level (5.6). A second decrease occurred at 8-10 min or thereafter. No significant change in cAMP level (9.0) throughout GVBD was observed. These findings sugt robable in- volvement of cGMP-dependent processes in GVBD in Spisulu oocytes and further indicate that the two nucleotides have different metabolic pathways. This investigation was supported in part by the Rockefeller Foundation, Steps to Inde- pendence fellowship from M.B.L. (H.S.), a travel grant from Yamada Science Foundation (H.S.). and NICHHD grant R01-HD131S4. rroperties of isolated fertilization envelopes from Arbacia ])tinctulata. Lucio CARIELLO. JAY C. BROWN, AND BRIAN STORRIE. Arbacia eggs were fertilized in the presence of 1 mM amino-triazole (ATA), a perox- idase inhibitor, to prevent hardening of the fertilization envelope. Eggs were then passed through a 51 /tm Nitex filter to strip raised fertilization envelopes off the eggs. Fertilization envelopes and eggs were separated by repeated settling. Envelopes prepared by this procedure were found to be intact and substantially free from cellular contamination. Sodium dodecyl sulfate polyacrylamide gel electrophoresis of solubilized envelopes revealed that they contain several protein species with molecular weights between 20,000 and 200,000. Phospholipid could not be detected in purified envelope preparations. Lactoperoxidase-catalyzed ml labeling of intact Arbacia eggs resulted in the labeling of a single high molecular weight protein. After fertilization several lower molecular weight components are also labeled. Any or all of these '"I-labeled proteins could be components of the fertilization envelope. Isolated envelopes were stable to treatment with either \% Triton X-100 or 0.01 M dithiothreitol, but they were com- pletely disrupted by exposure to 2 mM EDTA or 2 mM EGTA for 30 min at room tem- perature. This suggests that Ca2+ plays a key role in stabilizing the fertilization envelope before it is cross-linked by the cortical granule peroxidase. Envelopes could be stabilized by pretreatment with H2O2 (0.003%) for 30 min at room temperature. ATA completely blocked this effect. These results suggest that fertilization envelopes contain endogenous peroxidase activity which may be derived from the cortical granules. Pretreatment of isolated fertiliza- tion envelopes with Triton X-100 extracts of unfertilized eggs (in the presence of H2O-) also stabilized envelopes against EDTA or EGTA. This should provide an assay for the purifica- tion of the peroxidase. Gossypol inhibits motility of Arbacia sperm. MARIO H. BURGOS, CHANG CHIH-YI, LEONARD NELSON, AND SHELDON J. SEGAL. Gossypol, the principal pigment of cotton seed, has been given to men to suppress fertility. Clinical investigations have not established the basis for the action of this orally administered compound although direct effect of gossypol on fructose utilization by ejaculated human sperm has been observed. Here, we report that gossypol concentrations from 10-100 ^M stop motility of sperm of Arbacia punctulata within 12 min. Sperm motility was measured by centrifugal-field quantitative evaluation as well as direct microscopic observation. There is a close correlation between the two methods when they are used to evaluate the effect of lower concentrations of gossypol, but with higher concentrations (50 or 100 /*M) agglutination of sperm to form clusters interferes with the optical density readings involved in the centrifugal- field method. Trypan blue staining can be used to evaluate alterations in cell membrane permeability. One hr exposure to 50 /uM gossypol results in trypan blue absorption by 50% of treated spermatozoa. Trypan blue absorption is not manifested by unfertilized eggs or develop- ing embryos of Arbacia exposed to the same concentration of gossypol. The compound does not inhibit ciliary motion of free swimming blastulae or plutei. SEM observations of gos- sypol-treated sperm reveal a high incidence of alterations in the shape of head, midpiece, and tail. Blebbing of the membrane is seen along the entire length of the spermatozoon, membranes appear to become extremely adhesive, so that in sperm clusters attachment points are randomly established. Efforts to wash gossypol-treated sperm to reinitiate motility not been successful. Gossypol-treated eggs, subsequently washed, are successfully fertili: 468 ABSTRACTS FROM M.B.L. GENERAL MEETINGS normal sperm. The action of gossypol on gametes appears to be selective. It inhibits the motility of sperm, probably by altering cell membrane permeability, but does not damage eggs or developing embryos. Oxygen free radical generation by gossypol: a possible mechanism of antifertilitv action in sea urchin sperm. MICHAEL COBURN, PETER SINSHEIMER, SHELDON SEGAL, MARIO BURGOS AND WALTER TROLL. Gossypol, a constituent of cottonseed oil, is a potential male antifertility agent. The poly- phenolic dialdehyde structure resembles that of pyrogallol ( 1,2,3-benzenetriol ) in containing multiple oxidizable functional groups. Pyrogallol reacts with oxygen to form superoxide and hydrogen peroxide during autoxidation. We hypothesized that gossypol, which also takes up oxygen and is autoxidizecl, may also generate free radicals. Gossypol is a highly reactive compound that inhibits the fertilizing capacity of sperm of the sea urchin Arbacia punctulata. We tested the possibility that gossypol toxicity to sperm is due to the production of toxic oxygen metabolites by using the enzymes catalase and super- oxide dismutase (SOD) to degrade hydrogen peroxide and superoxide, respectively. Gossy- pol at 20 £tM concentrations completely destroyed the capacity of sperm to fertilize eggs. In the presence of catalase or SOD, however, sperm treated with gossypol fertilized eggs nor- mally. This suggests that the action of gossypol is dependent on the generation of hydrogen peroxide and superoxide. Pyrogallol, at concentrations similar to those of gossypol, also damages sperm fertilizing capacity, and this action of pyrogallol is also inhibited by catalase and SOD. When used together at lower concentrations, the two enzymes protect sperm against damage by gossypol or pyrogallol more completely than either enzyme alone. Boiled enzyme was without effect in preventing such damage. It thus appears that the spermicidal action of gossypol results from the production of toxic oxygen metabolites. This finding is consistent with the known sensitivity of sea urchin sperm to small amounts of hydrogen peroxide. We acknowledge grant support from NIH-CA-16060 to W. T. and from the New York University School of Medicine Honors Program to M. C. Mechanism of nicotinamide action on germinal reside breakdown in oocytes of Spisula solidissima. AKIRA MOMII. F. MOMII AND S. S. KOIDE. Nicotinamide (Nm) at 5 mM inhibits germinal vesicle breakdown (GVBD) of Spisula oocytes. By studying Nm action, the molecular mechanism of GVBD can be clarified, as the intracellular NAD* level falls drastically after insemination, and Nm raises the intracellular NAD* level. We have shown that Nm is rapidly converted to nicotinic acid in the Spisulu oocyte, and that trace amounts of Nm are detected after incubation of oocytes with [14C]Nm for 3 min. This metabolic conversion is due to an active nicotinamide deamidase system in the Spisnla oocyte (839 ±73 nmol-mg protein"1 -hr"1). This finding suggests that the agent responsible for the inhibition of GVBD may not be Nm, but rather a metabolite of the deamidase action, namely nicotinic acid or ammonia. Ammonia is not the inhibitor since it activates Spisula oocytes. Moreover, nicotinic acid does not affect GVBD, although it is incorporated into oocytes at the same rate as Nm. In starfish (Asterias jorbcsi) oocyte GVBD is induced by 1-methyladenine, and nicotina- mide deamidase activity is low (2 nmol-mg protein"1 -hr'1). Nm incorporated into starfish oocytes remains unchanged after 3 min incubation, and the intracellular concentration of Nm was higher in starfish than in Spisula oocytes. This finding suggests that intracellular Nm may not be responsible for the inhibition, and that Nm might affect GVBD indirectly via some mechanism involving interaction with the cell membrane. The present findings support the hypothesis that Nm blocks an NAD*-splitting enzyme, thereby increasing intracellular NAD* levels. This proposition is based on the reports that incubation of oocytes of Spisula and sea urchin with Nm results in a significant increase in NAD* level. Our result indicates that the concentration of NAD* synthesized from adminis- tered [14C]Nm after 3 hr of incubation can account for less than 0.5% of the reported level. Thus, the rise in intracellular level is probably the result of an inhibition of the NAD*-splitting enzyme rather than synthesis of NAD* from administered Nm. Supported in part by the Rockefeller Foundation and NICHHD grant RO1-HD13184. FERTILIZATION AND REPRODUCTION 469 Receptors, drug interactions and calcium in regulation oj Arbacia spenn cell junc- tion. LEONARD NELSON. Neurochemical control of motility of sperm cells containing both recvptors for and enzymes engaged in synthesis and hydrolysis of acetylcholine, presumably operates through modulation of calcium uptake and intracellular distribution. In earlier studies, epinephrine and nor- epinephrine showed no evidence of affecting cell movement. Now propranolol, a beta-adrener- gic blocker, was found to inhibit cell progression at 1 nM/1 while causing a moderate increase at 10 /uM/1 of filtered sea water (FSW). Quinidine, a beta-adrenergic agonist, had only minimal effects by itself. A pronounced interaction between these drugs appeared : in cells pre-incubated in 0.1 mM or 1.0 ,uM propranolol/1 FSW, 1 mM of quinidine depressed motility while 10 y^M sharply increased their swimming speed. Propranolol reportedly also counteracts the effect of the cardiac glycoside, ouabain, which is a specific inhibitor of Mg-activated, Na,K-ATPase. Micromolar propranolol increases the motility of cells pretreated with one micromolar ouabain, but with millimolar ouabain, propranolol slowed them down over a range of concentration from 0.1 nM to 1.0 mM. The local anesthetic procaine, which competes for cell surface calcium-binding sites, depressed movement in this study between 10 mM and 10 /uM/1 FSW. However, in calcium-free artificial sea water (Ca-f-ASW), the same amounts of procaine, after initially depressing, subsequently double the speed over that of the controls. Moreover, sperm spawned in and diluted with Ca-f-ASW showed increases in both speed and duration of swimming. It thus appears that sperm cells are endowed with a number of receptors which may be involved in motility regulation by modulation of calcium transport. Supported by N.S.F. grant PCM 80-02358. Tension generation in ovarian wall hv l-methyladenine during starfish spawning. HIROKO SHIRAI AND YASUAKI YOSHIMOTO. Starfish spawning is induced by l-methyladenine (1 Ma) produced by gonads under the influence of a peptide hormone. Since ovarian components such as oocytes, follicle cells, and ovarian walls adhere to each other before spawning and since ovarian walls shrink during spawning, release from adhesion within the ovary and contraction of the wall are essential events involved with spawning. To investigate mechanisms of ovarian contraction we measured tension in ovaries of Astcrina fcctinijcra with an isometric tension meter having sensitivity of 0.1 mg. "Isolated walls" without oocytes were prepared by incising ovarian fragments. "Jelly" was prepared from spawned eggs by acidification of egg suspension and concentration (1/100 of the intact jelly layer). Intact ovarian fragments (5 mm in length) generated tension after 1 Ma application (10~5 M), with a lag time of 15 min (22°C). The value reached a high plateau (mean 45 mg/fragment) 1 hr after 1 Ma application and remained at this level at least 2 hr (designated as "large tension," LT). Discharge of eggs began 20 min after 1 Ma application. Incised ovarian fragments (with oocytes) generated LT (mean 101 mg) with a lag time on 1 Ma application. On the other hand "isolated walls" (without oocytes) did not generate LT on simple application of 1 Ma. However, LT was generated (mean 26 mg) with no lag time on application of "jelly." Moreover "isolated walls" prepared from 1 Ma- treated ovarian fragments retained large tension (mean 27 mg) only under the presence of 1 Ma. The present findings indicate that 1 Ma indirectly induces contraction of ovarian walls, that "jelly" substance can be considered as a direct inducer of ovarian contraction, and that 1 Ma enables ovarian walls to maintain the generated contraction. This investigation was supported in part by a Steps to Independence fellowship from M.B.L., a research grant from Rockefeller Foundation, and a travel grant from Yamada Science Foundation to H.S. The toxic effects oj vitamin A on sea urchin gametes. PETER SINSHEIMER, MICHAEL COBURN, AND WALTER TROLL. The block to polyspermy in the sea urchin Arbacia puntulata requires the release of HsO- from the egg during fertilization. The release of H-O^ from the egg is triggered by entry of the first sperm. The toxic effects of HaO2 incapacitate nearby sperm capabl fertilization, thus preventing polyspermy. Two agents which interfere with the eliminat and production of H2O2 after fertilization, permitting polyspermy, are soy bean tripson : hibitor (SBTI) and catalase. SBTI prevents the formation of H2O2 at the egg's through its membrane interaction, while the enzyme catalase degrades H2O2. We have shown 470 ABSTRACTS FROM M.B.L. GENERAL MEETINGS that retinol or vitamin A also cause polyspermy in Arbacia eggs. Like SBTI, vitamin A pre- vents the production of H,.Oj hy the eggs upon fertilization. Both of these compounds also prevent the production of HaO2 and superoxides in human leukocytes upon phagocytosis, sug- gesting a general action on biological membranes. Vitamin A at 0.2 nig/ml was incubated with Arbacia eggs, washed, and resuspended in sea water. After fertilization, these eggs were nearly 100% polyspermic. HjOi> release, measured colorimetrically, was diminished 80% in eggs treated with vitamin A. Vitamin A is also toxic to sea urchin sperm at concentrations which caused high levels of polyspermy (0.1 mg/ml). Vitamin A inhibited the ability of sperm to effectively fertilize eggs. This action was inhibited by catalase and superoxide oismutase, which suggests that sperm damage occurred because of the toxic effects of oxygen free radicals. Thus vitamin A induces opposite effects on eggs and on sperm at the time of fertilization. In eggs, vitamin A inhibits the H-O- generating system necessary for an effective block to polyspermy, while in sperm, it causes loss of fertilization capacity through the generation of superoxides and H-jOs. We acknowledge grant support from NIH-CA-16060. Sperm-oocytc interaction in the surf clam Spisula solidissima. ALOYS G. TUMBOH- OERI AND SAMUKL S. KOIDK. Fertilization is a complex sequence of events that involves species-specific fusion of plasma membranes of the sperm and egg. It is likely that the initial interactions between sperm and egg are mediated by surface receptors that facilitate recognition and binding. To test this hypothesis, a study was undertaken to determine whether there are sperm receptors on the surface of Spisula solidissima oocytes. Germinal vesicle breakdown (GVBD) in the oocyte after insemination in vitro was used to assay occurrence of fertilization. When normal oocytes were treated with 0.0001% Triton X-100 for 30 min, washed, and inseminated, GVBD did not occur. This suggested that Triton X-100 had removed oocyte surface components responsible for sperm recognition. A supernatant was obtained from detergent-treated oocytes and Triton X-100 removed by treatment with several batches of SM-2 beads. The extract was concen- trated and incubated with washed detergent-treated oocytes for 60 min. Approximately 20% of reconstituted oocytes underwent GVBD. In addition, when normal Spisula spermatozoa were pre-incubated with the concentrated extract and subsequently used to inseminate normal oocytes, GVBD did not occur. To determine some of the properties of the oocyte receptor extract, the concentrated extract was boiled for 2 min then used to reconstitute detergent- treated oocytes. These oocytes failed to undergo GVBD when inseminated, indicating that the receptor activity is destroyed by heat. Further, reconstitution of detergent-treated oocytes with sea urchin egg surface-membrane extract failed to induce GVBD, showing that the receptors were species-specific. Finally, treatment of oocytes with 100 Mg/ml of the lectins ConA, PNA, DBA, RCA-I and SBA prior to insemination did not affect GVBD, whereas the lectin VEA-I reduced GVBD by 50%. This finding suggests that a-L-Fucose is involved in sperm-oocyte interaction in Spisula. It can be concluded that receptors present on the surface of Spisula oocytes are involved in recognition and binding of sperms. These receptors are heat labile and appear to be species-specific. Supported in part by The Rockefeller Foundation and NICHHD grant RO1-HD13184. MICROANATOMY AND MICROTECHNIQUES Relationships between dendrite and spine neck diameters in jrcczc-jractured rat hippocampal jorjitation. KRISTKN M. HARRIS. Theoretical mathematical models have described how dendritic spine necks might provide synaptic input attenuation and isolation, as well as impedence matching to the dendrite. These models require that as the dendrite narrows, so do spine necks. Granule cells from the dentate gyrus of rat hippocampal formation were chosen to study this relationship empirically because their spiny dendritcs taper with distance from the cell body. In Golgi -impregnated granule cells it is possible to see dendritic taper, count many spines, and observe thin spine necks ; but difficult to resolve spine neck diameters. Conversely, exami- nation of serial thin sections through spines yields accurate estimates of their neck diameters, but it is difficult to obtain many observations. The freeze-fracture complimentary-replica MICROANATOMY AND MICROTECHNIQUES 471 technique seemed promising because replicas can be viewed at high magnifications with good resolution of many spine neck diameters. The diameters of cross-fractured spine necks were measured in six adjacent 50 yuni regions from the granule cell bodies to the hippocampal fissure. Cross fractures were measured because they appeared more frequently and were similar to profile diameter^, v'en in the same field. The distribution of spine neck diameters was skewed in each of the : regions, with smaller diameters always occurring more frequently. The dendritic diameter was measured at the location of each spine neck. Although this measure is subject to variability in the freeze fracture plane, we have no reason to believe that narrow and wide dendrites will be selectively fractured at different relative depths. Measurements from 162 spines showed that spine necks narrow as dendrites taper so that the ratio of spine neck diameter to dendrite diameter does not change significantly. Thus, our preliminary results demonstrate a way to analyze many spine necks, and provide an em- pirical anatomical basis for models relating spine necks to impedance matching. The author wishes to thank Drs. Thomas Reese and Dennis Landis for their help and encouragement. This project was done during the MBL Neurobiology course, summer of 1980, and supported by IRP-NINCDS. Characterization of periodic order in the neuroplasmic lattice oj Hufo, Loligo and Hermissenda a.vons b\ a I'ourier analytical technique in scanning transmission electron jnicroscopy (STEM ). ALAN J. HODGE AND WILLIAM J. ADELMAN, JR. We have previously demonstrated the presence of an ordered neuroplasmic network or lattice in Loligo and Hermissenda axons using stereoscopy and optical autocorrelation on trans- mission electron micrographs of relatively thick (0.1-0.5 turn) sections. The lattice consists primarily of neurofilaments and their periodically arrayed side projections (40-45 nm apart) which act as cross-bridges. Neurotubules and other fibrous elements are often included in the lattice. The same type of lattice structure has now been observed in Bufo peripheral axons, with the order best preserved in constricted zones associated with Ranvier nodes and Schmidt-Lanterman clefts. The cross-bridge spacing found is also 40-45 nm. This type of structure accounts for axoplasm's gel-like character and anomalously weak optical anisotropy. The neuroplasmic lattice is also seen in stereomicrographs taken by STEM, a technique suited to analytical methods for objective characterization of video line signals comprising the picture raster. One method involves orientation of one axis (usually longitudinal) of the structure along the line-scanning direction and synchronized recording of the single-line-mode video signal. Inherent noise in single line traces was reduced by averaging numerous (typi- cally 256) single-picture line repeats. Data was analyzed by computer using Fourier methods. Another method, applicable to highly ordered structures such as myelin, involves signal averaging 16-64 successive lines in high resolution picture mode (1000 lines/frame). This is equivalent to integrating the image along the y-axis before Fourier analysis. Myelin and tropomyosin crystals were used as standards. These techniques were used on longitudinal sections of axons in Bufo peripheral nerves and Hermissenda connectives. In all cases, the neuroplasmic lattice exhibited an apparent longitudinal spacing of 40-45 nm with indications of a larger unit cell having a n-fold screw axis. It seems clear that this axoplasmic structural order is primarily a property of neuro- filaments and associated cross-bridges in the species examined. Oriented particle assemblies in the plasma membrane of Tetrahymenu : their deploy- ment relative to cell surface topography, cellular morphogenesis, and sensitivity to stimuli. LINDA HUFNAGEL. A freeze-fracture ultrastructural analysis of membrane structure in the ciliate, Tetra- hymcna, revealed three different types of particle assemblies: 1) plate-like arrays, 2) paired linear arrays, and 3) hexagonal arrays. The distribution of these arrays relative to cell surface topography and positions of cellular organelles has now been extensively analyzed from freeze-fracture replicas and thin sections. In addition, the surface of Tetrahymena has been examined following deep-etching of cells fixed in glutaraldehyde, washed thoroughly distilled water, and rapidly frozen in Freon 22. Paired linear arrays appear to be restric to regions where the cell surface undergoes sharp contour changes, such as the edge of oral cavity and the crests of interkinetal ridges. These arrays thus appear to be important 472 ABSTRACTS FROM M.B.L. GENERAL MEETINGS in cytoskeletal -membrane associations required to establish and maintain surface topography. Hexagonal arrays are limited to the floors of the kinetal valleys, near the anterior tip of the cell ; they, too, may function in forming and maintaining the specialized architecture of the deeply grooved anterior end of the cell. The morphology and distribution of the plate-like arrays are most consistent with a role in reception of stimuli. For example, surface analysis of deep-etched cells reveals that the plate-like arrays are manifested externally. Furthermore, thin sections demonstrate that the arrays are locations where the cell membrane, underlying inner and outer alveolar membranes, and internal cytoplasm of the cell have a structural con- tinuity not found in other regions of the cell. Liposome intraocular drn, rhombohedral calcite crystals large enough to determine the directions of a-axes were grown epitaxially on the etched surface. In calcite crystals observed along the c-axis, three ridges surrounding the c-axis form 120° angles with one another and a-axes are 30° to the three ridges. By observing etched and decorated samples under SEM, the following conclusions were drawn: (1) the spine is a single crystal with its c-axis parallel to its morphological long axis, (2) the interambulacral plate is also a single crystal with its c-axis perpendicular to the surface of the plate, (3) the marnmelon of the tubercle is a polycrystalline aggregate in both interambulacral and ambulacral plates, (4) the ambulacral plate is an assembly of seven single crystals, (5) the directions of the three a-axes in the plate are nearly parallel to the three edges of the plate, and (6) the direction of the a-axes of the plate have no fixed relationship to those of the attached spine. Light micrographs of gold-palladium coated, decorated tests showed reflection patterns indicative of the crystalline axes of individual plates due to the functional front-surface mirrors formed. This work was partially supported by a Hargitt Fellowship to K. O. and NSF Grant PCM 78-22242. MOLECULAR BIOLOGY AND BIOCHEMISTRY 473 Evaluation oj low light Ici'd intensifier I'idicon detectors for microscop\. GKO. T. REYNOLDS. Increased interest in applying image intensifier-TV systems to low light level microscopy lias resulted in availability of several commercial systems. Evaluation for application to specific problems requires information on dynamic range, resolution, and system noise. Each of these terms is limited in that separately they provide an incomplete description of the response of the system and they are not independent. Evaluation criteria developed for systems used in astronomy and X-ray diffraction studies are applicable in biological applications. The detective quantum efficiency (DQE) compares the signal-to-noise ratio of the output of the device to that of the input signal and thus measures the noise introduced by the system in a measurement. It can be readily measured and used to characterize detector response over a full range of intensity and spatial resolution. The DQE for film, intensifier-silicon vidicon (SIV), and silicon intensified target (SIT) vidicons has been measured. In addition a method of determining the rms video noise, which limits the sensitivity of TV detectors, has been developed using a calibrated light emitting diode (LED). In this way the rms noise equivalent input has been determined for a Plumbicon, Chalnicon, and three different commercial SIT's. Values ranged from 3.5 X 10~4 to 5.0 X 10~7 foot candles (fc). The equivalent input for an intensifier (gain 10r') -Plumbicon system was found to be 1.0 X 10~7 fc (the maximum of the LED emission was at 560 nm). Since image tubes with gains above 10" are available, limiting sensitivities of 10~s fc, or approximately 10" photons per cm2sec, can be realized. Using slow scan TV readout and time integration, higher sensitivity can be achieved. Supported by DOE Contract EY-76-S-02-3120. MOLECULAR BIOLOGY AND BIOCHEMISTV Subcellular localization and characterization oj islet hormone-degrading enzymes in anglerfish islet tissue. G. ERIC BAUER, BRYAN D. NOE, MICHAEL N. MORAN, STEPHEN KAHLER, AND JEFFREY B. NEUSTADT. Evidence was obtained for the presence of insulin- and glucagon-degrading enzymes in subcellular fractions of islet tissue. Islets were homogenized in 0.25 M sucrose ; unbroken cells and nuclei were removed by centrifugation. Remaining particulate and cytosolic components then were separated by centrifugation at 142,000 X g for 1 hr. Particulates were reconstituted in 0.25 M sucrose and freeze-thawed before assay. Hormone degradation was assayed with tracer concentrations of 125I-insulin (porcine) or '"'I-glucagon (bovine-porcine) in sodium acetate-citric acid buffer for 30 min at 37°C. Intact hormone was separated from fragments by adding trichloroacetic acid (TCA) to 5%, and centrifugation. TCA-soluble and precipi- table radioactivities were determined by gamma spectrophotometry. When pH optima were studied, insulin- and glucagon-degrading enzyme activities of cytosol were identical (pH 7.3), suggesting that one enzyme is involved, corresponding to "insulin-specific protease" of other tissues. Also, there were acidic enzyme activities (pH 3.4-4.3) in the cytosol as well as in the particulate subcellular fraction. In order to more precisely localize hormone-degrading enzymes, islet tissue homogenates were fractionated into nuclear, lysosome-rich, Golgi-rich, secretion granule, microsomal, and cytosolic subfractions by published procedures. Acidic hormone-degrading enzymes had the highest specific activity in the lysosome subfraction. Also by this procedure, the major "insulin-specific protease" activity (pH 7.3) was found in the cytosol. "Insulin-specific protease" was sensitive to dithiothreitol-reversible thiol-protease inhibition. Acidic enzymatic activity was sensitive to thiol-protease inhibitors, as well as leupeptin, antipain, and especially pepstatin. Hormone-degrading enzymes were relatively insensitive to metallo-protease and serine-protease inhibitors. Further enzyme characterization and purification studies are now in progress. Support: NIH AM-19223. Further observations on the phosphor ylation oj ribosomal proteins after fertilization of Arbacia punctulata eggs. D. BALLINGER AND T. HUNT. We have previously shown that increased labeling of a 31,000 Mr protein of the ribosomal subunit occurs 4-5 min after fertilization of Arbacia punctulata eggs pre-1 474 ABSTRACTS FROM M.B.L. GENERAL MEETINGS We tested the hypothesis that this phosphorylation increases the probability that the modified subunit will become involved in protein synthesis by analyzing post-mitochondrial supernatants from ^PGVlabeled 30-, 70-, 120-, 180-, and 240-min embryos on sucrose gradients, and assaying for the presence of the phosphorylated protein. The specific activity of the phosphorylated 31,000 Mr protein is approximately the same in free ribosomes and polysomes, although accurate quantification is difficult because of the small proportion of polysomes. Thus, it appears that the phosphorylation may be necessary, but not sufficient, for ribosomal activity. To elucidate the mechanism of this increased labeling we searched for kinase and phos- phatase activities in cell-free extracts of eggs and embryos. We could not detect specific kinase activity, but did detect a phosphatase specific for the 31,000 Mr protein in these extracts. We prepared a substrate for the phosphatase assays by incubating embryos for 4 hr in the presence of :BPO<, and removing unincorporated label. This substrate was added as tracer to extracts of eggs and embryos. The label in the 31,000 Mr protein is specifically removed by extracts of eggs or 5 min embryos, but not by extracts of older embryos. We believe that the label is removed by a phosphatase activity because it is extremely specific for the phos- phorylated protein, and it is inhibited by various phosphatase inhibitors such as NaF, EDTA, and 10-100 mM PGY There is apparently a phosphatase specific for the 31,000 Mr protein in eggs and early embryos that is inactivated shortly after fertilization, leading — perhaps in con- junction with increased kinase activity — to the increased labeling of the ribosomal protein. This work was supported by XIH Training Grant GM 00265. ADP ribosylation in nuclei isolated from different embryonic stages of Arbacia. REBECCA P. ELLISON. Adenosine diphosphate ribosylation is a ubiquitous enzyme-dependent reaction in which the ADPR moiety of NAD+ is covalently attached to a protein acceptor. The effect of the re- action depends on the cellular function of the acceptor protein. In eucaryotes, the enzyme reaction takes place on chrornatin, and various chromosomal proteins, including histones, are modified. Nuclei were isolated from five Arbacia developmental stages and incubated with :'H-NAD+ under various conditions. Incubation of 3 mg/ml protein in 0.1 mM NAD+ and 5 mM MgCU at pH 6.8 and 15°C resulted in linear incorporation of ADPR for 30 min, which was totally inhibited by 4 M nicotinamide and 70% inhibited by 5 mM ADPR. PEI-cellulose chroma- tography of the :'H -NADMncubated nuclei after snake venom phosphodiesterase hydrolysis showed half the counts migrating with ADPR and half with AMP, indicating the reaction product is poly (ADPR) with an average chain length of 2 ADPR monomers. Blastula and pluteus nuclei incubated with :'H-NAD+ were fractionated into nuclear sap (0.15 M NaCl soluble), histone/HMG (0.4 N H,SC>4 or 0.25 M HC1 soluble) and residual proteins. Nuclear sap contained no 20% TCA precipitable material, histone/HMG contained 1.5% (blastula) and 0.35% (pluteus) of the total incorporated ADPR, and the remainder was found in the residual fraction. The specific activity was 0.54 (±0.01) nmoles ADPR/mg protein for blastula histone/HMG, 0.06 nmoles ADPR/mg protein for pluteus histone/HMG, and 35.9 (±2.5) nmoles ADPR/mg protein for residual proteins of both stages. Supported by the Research Foundation of the State University of New York. Electron spin resonance studies of protein-methylglyoxal complexes. PETER GASCOYNE, JANE MCLAUGHLIN, AND ALBERT SZENT-GYORGYI. Pethig, McLaughlin, and Szent-Gyorgyi have shown that the brown color that arises when bovine serum albumin (BSA) is complexed with methylglyoxal ( 1,2-oxo-propanal) results from the formation of a Schiff base linkage between methylglyoxal and the e-amino groups of the protein lysine residues. Electron spin resonance (ESR) and spectrophotometric mea- surements were made on the BSA-methylglyoxal complex in order to investigate the link between the brown color and the unpaired spin concentration that also accompanies the treat- ment of protein with methylglyoxal. The intensities of both the brown color (measured at 318 nm) and the ESR signal (at g — 2.005 ) were directly proportional to the amount of methylglyoxal complexed with BSA. Whereas the intensity of the ESR signal increased rapidly in the presence of molecular oxygen, the optical absorhance remained constant within experimental error. These results show that the ESR signal and the brown color of the BSA- methylglyoxal complex arise from different molecular species. It is proposed that methyl- MOLECULAR BIOLOGY AND BIOCHEMISTRY 475 glyoxal initially reacts with the e-amino groups of the lysine residues of the protein to form a hemiacetal. The hemiacetal can undergo rearrangement to form either a Schiff base (ac- counting for the brown color) or else a diol in the ene form. The diol can undergo stepwise oxidation in the presence of oxygen first to form a semidione (accounting for the HSR signal at y — 2.005) and then a dionc. Such studies are considered to he of direct relevance to the radical signal observed in living systems where methylglyoxal has been observed complexed to proteins. Part of the ESR signal of living systems may arise from complexed methyl- glyoxal semidione radicals rather than from quinones or flavins as is currently believed. This work is supported by the National Foundation for Cancer Research. Partial purification ami characterization oj a cytoto.ric protein from squid (Loligo pealei ) posterior salivary glands. WILLIAM R. KKM AND JAMES D. SCOTT. Homogenization of freeze-dried posterior salivary glands in 1 M ammonium acetate (pH 6) yielded a cytotoxic supernatant fraction that was further purified by molecular sieve chromatography and isoelectric focusing. Cytotoxicity was determined using a \% suspension of washed rat erythrocytes in 145 mM XaCl, 2 mM Cad-, and 10 mM Tris, pH 7.4. After incubation of the red cells with toxic samples for 1 hr at 37° C, the suspension was centrifuged, and hemoglobin release measured at 540 nm. All of the hemolytic activity eluted from an AcA-44 LTltrogel column as a sharp peak at 1.8 V,, (void volume) ; thus the apparent molecu- lar size of the cytotoxin was abovit 30,000-50,000 daltons. The cytotoxicity apparently is due to one or more proteins, since it was readily inactivated by boiling or by 30 min exposure to low concentrations of trypsin, pronase, or 1 mM dithiothreitol. The serine protease inhibitor phenylmethylsulfonylfluoride did not affect hemolytic activity. Flat-bed isoelectric focusing demonstrated that the major cytotoxin possessed a pi of 7.2 at 5°C, whereas a minor hemolytic component possessed a pi of 7.8. The hemolytic action of the Lolii/o toxins occurred in the absence of calcium although elevated concentrations ( 10-40 mM) of calcium progressively prevented hemolysis. Reducing pH to 6 enhanced hemolysis by about 70%, whereas raising the pH to 8.5 had no significant effect. The steady-state hemolysis evoked by Loligo toxins was greatly dependent upon temperature, being much greater at 37°C than at 23°, 13°, or 5°C. This research was supported by NIH grant GM 25849. Studies oj protein synthesis in isolated toadfish iiepatocytes. RITA \Y. MATHEWS AND AUDREY E. V. HASCHEMEYER. Active hepatocyte preparations have been obtained from toadfish (Opsanus tan) by col- lagenase perfusion based on the method of Seglen. Anesthetized fish maintained with cir- culating seawater were cannulated through the hepatic portal vein and perfused with 150 ml Hepes-buffered balanced salts (pH 7.8), followed by medium containing 2 mM MgSO<, \% bovine serum albumin (BSA), and 100 units/ml collagenase Type IV (Sigma). Cells were combed into balanced salts medium and washed by low-speed centrifugation and resuspension. Incorporation of I4C-leucine into protein was assayed in freshly isolated cells in suspension at 0.04 g/ml in a medium containing 0.16 M NaCl, 0.005 M KC1, 0.001 M MgSO,, 0.1 % glucose, 25 mM Hepes (pH 7.8), \% BSA, 0.29 mM each of 19 amino acids, and 0.6 ^Ci/ml 14C- leucine at 0.014—0.13 mM. At low external leucine concentration, incorporation was linear with time at all temperatures, and reached levels of 30,000 cpnvhr'MOO M' sample'1 at 30°C. The temperature coefficient (Qm) was 5 for the 4°-21°C range and 1.3-1.7 from 21° to 30°C. A similar temperature dependency was obtained with cells in stationary culture 2 days after plating to plastic dishes in serum-free Waymouth's medium. The behavior of the hepatocytes up to 21°C closely resembles that observed for protein synthesis in liver in vivo. At high external leucine concentrations, incorporation fell to low levels and Qio increased to 5 (20- 30°C) and 20 (15-20°C), suggesting that uptake of the radioactive precursor became rate-limit- ing under these conditions at low temperatures. Further study of this system will require combined analysis of transport and protein synthetic rates. Supported by National Science Foundation grant PCM 79-21091. Histochemical identification oj N-acctyl-(3-(/lncosaininidase activity in early embryo, oj Lytechinus pictus. KATHLEEN MORGAN. A histochemical technique specific for the localization of N-acetyl-/3-glucosaminidase activity was applied to unfertilized eggs and early embryos of Lytechinus pictus to determine 476 ABSTRACTS FROM M.B.L. GENERAL MEETINGS if the enzyme activity is present in early embryos and if the activity is temporally and/or spatially localized. N-acetyl-p-glucosaminidase (EC 3.2.1.30) is a lysosomal exoglycosidase specific for non- reducing terminal hexosamines of oligo- and polysaccharides, in invertebrates and vertebrates. The enzymatic activity has recently been found associated with ligatin, a low-molecular-weight, surface-binding protein isolated from vertebrate cell types and also from sea urchin eggs and embryos. Such a lytic activity, if present at the cell surface, could reflect or underlie cellular differentiation. Lytcchimts pictus embryos were fixed briefly in cold 80% ethanol at regular intervals from unfertilized egg through early gastrulatiun. In the first step embryos were incubated for 2 lir at 37°C in a reaction mixture containing Xaphthol-AS-BI-N-acetyl-/3-glucosaminide (9 X 10"4 M), or, in the case of controls, the reaction mixture minus the substrate. After washing, the embryos were incubated in a post-coupling reaction mixture containing 1 mg/ml Fast Blue RR salt, a diazo compound which reacts with the released naphthol to produce a blue/violet precipitate at the site of enzymatic activity. Enzymatic activity is present in the unfertilized egg throughout the cytoplasm, but ex- cluded from the nucleus and not associated with the plasma membrane. From fertilization to early gastrulation the activity remains cytoplasmic and is not selectively segregated to par- ticular cells. This research was supported by U.S. Public Health Service Grants 2T32GM07231-06 and T32H007098. The status of inKXA in eggs and early embryos. ANDREW MURRAY AND RONALD SOSNOWSKI. We measured the translatibility of the mRNA in extracts of Arbacia pnnctulata eggs by adding them to mRNA-dependent reticulocyte lysate (MDL). The eggs or embryos were homogenized in 45.5 mM KC1, 2.3 mM MgCU, 69.6 mM Kgluconate, 50 mM HEPES, 10 mM EGTA, 618 mM glycerol titrated to pH 6.9 with KOH, and centrifuged at 12,000 X g to yield a post-mitochondrial supernatant (PMS). Protein synthesis was assayed as the incorporation of :i5S-methionine into TCA insoluble material. Protein synthesis in this system was linear for 45 min and was 90% inhibitable by edeine, showing that re-initiation occurred on PMS mRNA. By 20 min post-fertilization the PMS- directed incorporation was twice that of unfertilized eggs. A similar increase in polysomal mRNA was detected as increased edeine insensitive incorporation. It remains unclear whether non-polysomal mRNPs contribute to translation in the MDL. To demonstrate masked mRNA we translated extracted RNA and PMS at equivalent concentrations, assuming complete re- covery of the RNA. We precipitated RNA by adding one volume of 4 M LiCl, 8 M urea, 0.5 mM EDTA, and 20 mM Tris pH 7.5 at 0°C, and phenol -extracted the pellet. Purified RNA directed 2-4 times more incorporation than the PMS from eggs or early embryos. Since the PMS is non-inhibitory to translation of purified mRNA we conclude that the PMS contains functionally masked mRNA. The clam Spisula solidissima undergoes a change in the pattern of protein synthesis at fertilization. To investigate the protein components of mRNPs Spisula oocyte PMS was passed over oligo-dT cellulose columns to which poly-A, oocyte mRNA, or no added com- ponents had been bound. Proteins were eluted with increasing salt concentrations and analyzed on SDS-polyacrylamide gels. The 0.1-2 M eluates from all three columns were identical, while the 4 M NaCl eluate from the oligo-dT-mRNA column showed seven additional stained bands. Supported by NIH training grant GM 00265. Inhibition of islet prohormone to hormone conversion by incorporation of arginine and lysine analogs. HRYAN D. NOE, G. ERIC BAUER, MICHAEL N. MORAN, STEPHAN KAHLER, AND JEFFREY B. NEUSTADT. Previous work using isolated anglerfish (AF) islets has established the existence of bio- synthetic precursors for glucagon and somatostatin as well as for insulin. Most polypeptide prohormones have pairs of basic amino acid residues at the prohormone cleavage sites. AF islets were incubated with 3 mM canavanine (CAN), thialysine (TL), and/or norleucine (NL), and 3H-trp and "C-ile to determine whether: a) the analogs of arginine (CAN) and lysine (TL, NL) can be incorporated into islet prohormones, and b) prohormone conversion MOLECULAR BIOLOGY AND BIOCHEMISTRY 477 is inhibited by analog incorporation. After incubation of islets, 2 M acetic acid extracts were analyzed for prohormoncs and hormones by gel filtration. Prohormone-hormone ratios from extracts of tissue incubated with and without analogs were compared. Incubation with CAN alone resulted in 81, 72, and 10% reduction of proinsulin (Pro-I), proglucagon (Pro-G), and prosomatostatin (Pro-SS) conversion, respectively. Incubation with TL alone reduced Pro-I conversion 81%, Pro-G 53%, and Pro-SS 23%. Incubation with NL reduced Pro-I conversion 3%, Pro-G 17%, and Pro-SS 22%. Incubation with CAN + TL reduced conversion of Pro-I 94%, Pro-G 83%,, and Pro-SS 46%. CAN + NL reduced Pro-I conversion 77%, Pro-G 73%, and Pro-SS 32%. All results are means of 3 determinations. Arginine, lysine, or their analogs (3 mM), added to lysed secretory granule preparations, did not inhibit conversion of non-analog containing prohormones, indicating that the reduction of prohormone conversion in intact tissue was not due to inhibition of the converting enzyme (s) by the analogs. More- over, conversion of CAN + TL containing precursors by lysed granules was approximately 90% less than conversion of non-analog containing precursors. Analog-containing prohormones also were significantly less degraded than normal precursors after limited trypsinization. These results indicate that CAN, TL, and NL are incorporated into islet prohormones, result- ing in inhibition of prohormone to hormone conversion. Supported by NIH grants AM-16921, AM-26378, and AM-00142. Inhibition of L-lcucinc transport in toad fish liver b\< various natural atnino acids. ROGER PERSELL AND AUDREY E. V. HASCHEMEYER. Liver uptake of amino acids has been studied in rii'o by monitoring the distribution of "C-amino acids and of :'H-inulin between liver and blood at short times after pulse injection into the hepatic portal vein. Previous studies have shown that transport of L-leucine at 20°C is a saturable carrier-mediated process operating through a combination of active transport and facilitated exchange. The major competing amino acids are isoleucine and phenylalanine with lesser effects by valine, methionine, and aspartic acid. The present study explores the effect of reduced temperatures on competition characteristics of this system. Toadfish ac- climated to ambient summer seawater temperature (20° ± 1°C) were cooled to 10° ± 1°C 1 hr prior to experiments. Under benzocaine anesthesia fish were injected with 2 //Ci ["C(U)]- leucine and 4 juCi [:'H(G) jinulin in 0.1 ml balanced salts medium, pH 7.8. Leucine concen- tration was 0.1 mM ; competing amino acids were tested at 15 mM. The liver was excised after 5 min, and blood collected from the cut hepatic vein. Ratios of 14C/:|H and total re- coveries of the two isotopes in liver and plasma were determined. The results showed a high level of inhibition of leucine uptake by isoleucine (63%) and phenylalanine (54%), as at 20°C. Significant increases in inhibitory activity were found with valine (66%), methionine (63%), and aspartic acid (49%), compared with 20°C data. Histidine also was active (36%) at 10°C. The results suggest qualitative changes in the leucine uptake system at reduced temperatures. Supported by National Science Foundation grant PCM 79-21091. Influence of u Ca > Sr » Mg. Membrane lysis did not con- tribute to the perceived translocation as the liposomes remained impermeable to EDTA, EGTA, arsenazo III, or Mg. Liposomes with phosphatidic acid or the trienoic acids pre- incorporated at 1-5 mole percent of total lipids also permitted translocation of Ca but not Mg. Release of Ca from liposomes which had entrapped arsenazo III-Ca complexes into medium rich in EGTA permitted calculation of efflux induced by ionophores, whether these were added to the outside of liposomes or preincorporated. Data suggest that phosphatidic and trienoic acids (formed in natural membranes after phospholipase activation) act as calcium ionophores in model bilayers, and could play a similar role in cells as "endogenous ionophores." Dark-adaptation effects on photobchavior of Hermissencla crassicornis (Gastro- poda: nudibranchia). EDWARD BARNES AND IZJA LEDERHENDLER. The diurnally-dependent photopositive behavior of the sea snail Hermissenda crassicornis may be described in terms of three component phases : initiation of movement, approach to light, and selection of an intensity level within an illumination gradient. We found previously that variations in the photobehavior depend on an individual's past experience with light. We report here the effect of one aspect of this : amount of time in the dark before testing (dark adaptation). We measured the behavioral components as follows: time to leave a start area (start latency) ; the time, after starting, to cross within 8 cm from the center of illumination (latency to enter the light) ; and the time spent in the light within the 8 cm boundary. Animals were maintained on a 12 hr light: 12 hr dark cycle and tested during the light porti of the cycle. Individuals were dark-adapted for 0, 5, or 15 min, and then placed in the of a shallow aquarium illuminated from above at one end. The resulting gradient de at least three orders of magnitude from a maximum of approximately 100 Joules m~* sec 480 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Increased dark-adaptation significantly increased the start latency but did not alter latency to enter the light. After entering the light, however, individuals receiving more dark-adapta- tion remained in the light for significantly less time. Thus dark-adaptation seems to contribute to reduced photopositivity. This conclusion is also supported by our finding that among ani- mals that entered the light, dark-adapted individuals spent more time in relatively less illumi- nated areas. While the start latencies were not correlated with the latencies to enter the light, other aspects of the approach behavior, including velocity, turning, and heading, may be important. The photopositive behavior of Hermissenda crassicornis is an important feature of a cellu- lar model of learning that is being developed for this animal. Further detailed study of its mucus trails will help specify the exact nature of the change in behavior. Supported by IRP, NINCDS, NIH. Guanosinc phosphates and the control of membrane potential in Limulus ventral photoreccptors. S. R. BOLSOVER AND J. E. BROWN. Solutions of guanosine phosphates were injected into Limulus ventral photoreceptor cells. Where possible, the nucleotides were mixed with the dye Fast Green FCF, which is without effect when injected alone. Analogues of guanosine triphosphate (GTP) had a striking effect on the photoreceptors. Injection of GMP-PNP (guanylyl imidodiphosphate), GMP- PCP (guanylyl ( p, y methylene) diphosphate) and GTP7S (guanosine 7-thio triphosphate) substantially increased the frequency of "quantum bumps" observed in darkness. Injections were monitored visually in a relatively bright light. After this adapting illumination, the amplitudes of both light-induced and drug-induced bumps increased. In some cells, a second bright illumination was necessary before the elevation of dark bump frequency was observed. Sensitivity to light (peak light-induced current per stimulus intensity) was unchanged by GMP-PNP injection. Injections of imidodiphosphate, 5'-guanosine monophosphate and 3',5'- cyclic guanosine monophosphate were without effect on dark bump frequency. When injected in the light GTP had no effect on the subsequent membrane voltage. However, injection of GTP in the dark produced a slow, smooth depolarization, reaching a maximum about 10 sec after injection and decaying over 1-2 min. The response to a light flash was much attenuated during the GTP-induced depolarization and recovered over a few minutes. Although injection of too large a volume of any solution can cause a depolarization due to membrane damage, we observed the smooth, transient depolarization only after GTP injection. In voltage clamped cells, the steady-state current-voltage relation measured in constant light intersects with that measured in the dark at the reversal voltage (< + 30 mV). When measured similarly, reversal voltage for the change in membrane current induced by GTP injection (—"+55 mV) was substantially more positive than that of the light-induced current. Supported by NIH EY-01914 and EY-01915. Preliminary evidence for taste-az'ersion learning in the nudibranch mollusc Aeolidia papillosa. M. B. BOYLE AND L. B. COHEN. We have undertaken to develop taste-aversion learning in an opistobranch mollusc using a paradigm similar to that used by Sahley, Rudy, and Gelperin on the pulmonate Limax ina.rinuts. In our experiment, each slug was exposed once per day to two anemones, Tcalia lofotcnsis and Epiactis prolijcru, with 3 hr between presentations. As an aversive stimulus, quinine (5-10 mM) in seawater was administered following 1 min of feeding on one of the two foods. Four of the slugs received quinine paired with one species, the other four with the second species. We predicted the development of a selective aversion to the paired food. Two response measures were used. First, tines before biting into both foods were measured during each trial and the difference of the two tines (paired minus unpaired) calculated. We expected that the mean differences would become significantly positive after training. Second, the slugs were tested in an olfactometer in which the odors of the two anemones were pre- sented simultaneously. A choice score was calculated for each slug based on its position in the box at 10, 15, 20, and 25 min after the start. We asked whether the preferences of the two groups became different. For both measures considered on a day-by-day basis, the data were infrequently significant at the 0.05 level or below (one-tailed t test). However, if for each measure the data points from the last six of the nine trials were combined into two groups of three consecutive days, the data was significant at the 0.01-0.04 level. We believe NEUROBIOLOGY 481 that we have conditioned a selective aversion in these animals, and plan in the near future to work on modifying the paradigm to achieve a more mlnist effect. Supported by XIH grant XS0837. Presynaptic injection oj calcium facilitates transmitter re/ease in the squid giant synapse. MILTON I'. (.'IIARLTON. KOHF.RT S. /IVKKR. AND STKIMIK> J SMITH. Facilitation of transmitter release involves secretion of increasing amounts of transmitter evoked by successive action potentials in a presynaptic terminal. This facilitation could be due to "residual calcium" which lingers in the presynaptic terminal following an action potential. To test this hypothesis we mimicked residual calcium by direct injection of calcium into presynaptic terminals of the squid giant synapse. Current was passed between the barrels of a thick-septum theta-tube microelectrode, one tube filled with 0.5 M CaCU and 1.5 M KC1, and the other with 3 M KC1. Electrodes were tested for calcium ejection in a solution of Arsenazo III. When 5-50 nA was passed through the calcium barrel to inject calcium into the pre- synaptic terminal the membrane potential of the postsynaptic cell became noisy and was de- polarized by as much as 1 mV. This result was routine and indicative of a large increase in spontaneous transmitter release caused by the additional calcium. Synapses placed in low cal- cium saline gave subthreshold synaptic potentials which sometimes increased 5-10% during calcium injection but were severely depressed for several minutes after the injection. More consistent results were obtained when transmission was eliminated in 1 mM CaCl^, 7 mM MnClj saline and was restored at a restricted area by focal application of calcium from an extracellular pipette located adjacent to the intracellular calcium pipette. In all such ex- periments intracellular calcium injection caused increased spontaneous release, increased (5-30%) evoked release and long lasting depression of synaptic responses. Twin-pulse facili- tation was little affected by the additional intracellular calcium. There was no effect on pre- synaptic resting or action potentials. Injection of potassium had no effect. The results indicate that additional intracellular calcium can cause facilitation of transmitter release and lend strong support to the residual calcium hypothesis. M.P.C. and S.J.S. are Steps Toward Independence Follows. Supported by XIH grant NS15114 and a Sloan Foundation Fellowship to R.S.Z. A GTP binding component regulates discrete i^are production in Limulus ventral photoreceptors: pharmacological evidence. D. WESLEY CORSON AND ALAN- FEIN. The receptor potential of Limnlus ventral photoreceptors is composed of discrete waves at low light intensities. Recently we reported that fluoride induced production of discrete waves in the dark. We now report that vanadate ions (VO:f) and GTP-7-S ( guanosine-5'-O- (3 thiotriphosphate) ) induce the production of discrete waves in the dark and prolong the response to a flash of fixed intensity. The time course of discrete waves induced by vanadate and GTP-7-S (a hydrolysis resistant analog of GTP) appears to be similar to the time course of spontaneous and light-induced discrete waves. The vanadate and GTP-7-S induced discrete waves also undergo adaptation in response to a bright adapting light. We tentatively con- clude that vanadate and GTP-7-S induced discrete waves arise from a process similar or per- haps identical to visual excitation. Microelectrodes with 100 mM solutions of vanadate, GTP-7-S, or the control substances GTP, ATP, or ATP-7-S were used for iontophoretic injection. Control substances were without comparable effects on discrete wave production. The effects of vanadate injection were largely reversible within 6 hr. GMP-PNP (guanylyl-imidodiphosphate), a second hydrolysis resistant analog of GTP, did not consistently produce discrete waves in the dark. GTP-7-S, vanadate, and fluoride are members of the class of compounds known to activate hormone regulated enzymes through direct activation of GTP-binding regulatory proteins. From the observed effects of these compounds on ventral photoreceptors, we suggest that GTP-binding regulatory protein is very likely to be involved in the regulation of discreate wave production. Supported by grants from the NIH and the Rowland Foundation. 482 ABSTRACTS FROM M.B.L. GENERAL MEETINGS An X-ray study oj the retinal f>hotureccf>tor structure of squid. A. R. CZETO, A. R. \YORTHINGTON, AND C. R. WoRTHINGTON. We have investigated the molecular structure of the invertebrate rhabdomeric visual cells of squid by low-angle X-ray diffraction. It is known from previous X-ray studies by Worth- ington et al. (1976, Nature 262, 626) that X-ray patterns can be recorded from squid retina after fixation with glutaraldehyde. This previous study demonstrated that the two-dimensional array of microvilli was hexagonal with a = 620 A and also that the light direction was coinci- dent with the (11) direction of the two-dimensional array. All later attempts to obtain X-ray patterns from live squid retina after transportation from M.B.L. to our X-ray labora- tory at Pittsburgh have failed. In the present study the X-ray experiments were run at M.B.L. The squids were first dark adapted for 45 min. In the X-ray experiments a 1 mm strip was cut from the freshly dissected eye and mounted in a specimen chamber with excess Ringer's solution. The X-ray exposure was started about 1.5 hr after the arrival of the squid in the M.B.L. holding tank. Slit collimation with exposure times of 1 or 2 hr was used to explore whether low-angle X-ray diffraction could be recorded from the intact, untreated strips of squid retina. Initially, only poor patterns were obtained, indicating that the micro- villi array was disordered. It was soon found that the ordering of the array was extremely sensitive to the chemical composition and the osmolarity of the Ringer's solution. Poor X-ray patterns were routinely obtained using the various artificial sea waters even after adjusting the osmolarity of the solution to the osmolarity of the vitreous humor of the squid eye, namely about 900 milliosmoles. A Ringer's solution based upon a study by Robertson (1953, J. Exp. Biol. 30, 277) was designed to match the osmolarity of the squid eye. We report here that excellent X-ray patterns showing diffraction orders 2-8 of d «=* 640 A were obtained from intact untreated squid retina when using this Ringer's solution. The diffraction extends out to a minimum spacing of about 35 A, which is typical of membrane diffraction. This work was supported by NIH grant NS 09329. Field experiments on electrically evoked feeding responses in the dogfish shark, Mustelus canis. BENJAMIN G. DAWSON, GAIL W. HEYER, RENE E. EPPI, AND AD. J. KALMIJN. From previous laboratory experiments, we learned that sharks, skates, and rays have an electric sense that enables them to detect voltage gradients as low as 0.01 jciV/cm within the frequency range from DC up to 8 Hz. The animals use their electric sense in predation, cuing in on the bioelectric fields commonly produced by fish and aquatic invertebrates. To quantify the response, we analyzed the feeding behavior of the shark Mustelus canis in Vine- yard Sound off Cape Cod, Mass. An electrode panel was embedded in the ocean substrate in a water depth of 2-3 m. Two salt-bridge electrodes, simulating a small prey fish, were placed 2 cm apart at a distance of 15 cm from a centrally located odor source. Another pair of salt-bridge electrodes, simulating a larger fish, were placed 5 cm apart at a distance of 30 cm on the other side of the odor source. DC current of 8 /iA was applied to either one or both pairs of electrodes. Observations were made at night from a Boston Whaler with a glass- bottomed observation well. Liquified herring chum attracted and motivated sharks. Markings on the electrode panel enabled observers to measure the distances from the electrodes at which the sharks initiated attacks. Data were recorded on the size of the sharks (30-40 cm pups or 90-120 cm adults), the vigor of response, and the pattern of attack. Out of 136 responses, first-year sharks attacked the 2-cm electrodes 49 times from a distance of 15 cm or more measured along the dipole axis, which corresponds to a sensitivity of 0.04 I(4) »I(3), where I(h) refers to the integrated intensity of the diffraction order h. The X-ray patterns of nerve myelin from fresh water fish are however anomalous in that 1(2) > 1(4) ==* 1(3), that is, there is a strong third order reflection. In the present study, X-ray patterns were recorded from marine teleosts, toad fish and sculpin, and from the elastmobranch skate. The Ringer's solutions were a calcium-free marine teleost solution and a skate Ringer's solution which contained 333 mM urea. The optic nerve and a branch nerve (a fin nerve) of toad fish and the optic nerve and spinal cord nerve of sculpin were examined. These marine teleost nerves had a repeat distance d — 153-155 A and had the same intensity variation as shown by fresh water fish nerves. The X-ray patterns of the two marine teleosts did not change over 7 days. The optic nerve and a branch nerve (a lateral line nerve) of skate were examined. The skate nerves had larger repeat periods : d = 167 A for optic nerve and for spinal cord and d = 177 A for the branch nerve. Moreover, the skate nerves had a different intensity variation from the other fish nerves : the intensity variation was identical to that of the central nervous system myelins from the other vertebrates, namely, 1(2) > 1(4) » 1(3). In an earlier X-ray study on fish nerves, Blaurock and Worthington (1969, Biochem. Biophys. Acta 173, 419) identified a second phase with a larger spacing of d « 180 A which was superimposed on the original pattern. The toad fish and sculpin nerves did not show this second phase, but the skate nerves readily show this pattern. In the skate the original pattern at first dominant but the larger period pattern with d = 200 A replaces the original pattern within 1 day. This work was supported by NIH grant NS 09329. An ion-permeable channel produced by venom oj the janged bloodivorm Glycera dibranchia. BRUCE L. KAGAN, HARVEY B. POLLARD, AND ROBERT B. HANNA. Venom from the poison glands of the polychaete annelid Glycera convoluta has been shown to dramatically increase the frequency of miniature endplate potentials at the frog and cray- fish neuromuscular junctions without causing detectable ultrastructural changes. We report that addition to Glycera dibranchia venom to one side of a lipid bilayer results in formation of ion-permeable channels in the membrane. Glycera dibranchia from Maine were immobilized on ice and the poison glands dissected free. Glands were then homogenized in 10 mM sodium phosphate buffer, pH 7.2, and centrifuged at 48,000 X g for 30 min. The supernatant was passed over Sephadex G-25, and 1-10 fj.1 of the void volume fraction (MW>5000) was added to one side of a planar phospholipid bilayer membrane. The final concentration of venom was about 5 milliglands/ml. The conductance of a single channel is about 330 pmho in 0.1 J and is ohmic. The channels exhibit moderate (but not ideal) cation selectivity in ] KC1 gradients. Other selectivity measurements suggest that Ca~* and Mg2* are also permeable The channels show a slight voltage sensitivity. The steady state conductance at (side opposite venom) is about 6 times the conductance at +60 mV. We suggest that these 486 ABSTRACTS FROM M.B.L. GENERAL MEETINGS channels in the venom may evoke transmitter release at neuromuscular junctions either by (1) depolarizing the pre-synaptic terminal and thus opening voltage-dependent Ca'~'+ channels, or (2) directly allowing Ca2+ to enter the terminal. The Ca"* requirement for the venom's effect at neuromuscular junctions is consistent with either mechanism. Black widow spider (Latro- dectus) venom is known to produce similar effects on neuromuscular junctions and lipid bilayers. Although worms bear little apparent resemblance to spiders, the single channel con- ductances and ionic selectivities of the channels found in the venoms of Glyccra and Latro- dcctus are strikingly similar. (Supported by NIH grant 5T32 GM 7288 to B.L.K. and SUNY Research Foundation award 7142 to R.B.H.) Coupling between horizontal cells in the carp retina examined b\ diffusion of Lucifer yellow. AKIMICHI KANEKO AND ANN E. STUART. The carp retina has four morphological types of horizontal cells (HCs) : HI, H2, and H3, all connecting to cones; and a rod HC. Because each type connects with a different set of photoreceptors, it can be identified from its spectral responses. HCs have large receptive fields thought to be due to their electrical coupling. To test this hypothesis, and also to de- termine whether the coupling occurs only among cells of the same type, we examined diffusion of Lucifer yellow, a fluorescent marker dye, after it had been injected into each type of HC. The penetrated cell was identified by its responses to narrow-band monochromatic lights from green (565 nm), yellow (585 nm), and red (635 nm) light-emitting diodes. As previously reported, HI cells gave luminosity (L-) type responses with highest sensitivity to red light, rod HCs gave L-type responses with highest sensitivity to blue-green, H2 cells gave biphasic chromaticity (C-) type responses, and H3 cells gave triphasic C-type responses. In 39 of 56 injections into the cell body, Lucifer yellow spread to immediately surrounding cells and, in some cases, even to secondary neighbors. In all preparations the diffusion was limited to the cells of the same morphological type as the injected one. When dye was in- jected into the expanded axon endings of the HCs, it diffused to certain other axon expansions, so that the appearance in flat-mounted retinas was of a coarse meshwork. This result indi- cates that neighboring axonal endings are also coupled and that the coupling is limited, prob- ably to the same cell type. Axons with expanded endings were seen in almost all rod HCs, a finding not previously reported. Supported by a Rand Fellowship to A.K. and by NIH grant EY03347 to A.E.S. Recording jrom the Limulus ventral eye in situ : is there a circadian rhythm? EHUD KAPLAN, RANJAN BATRA, AND ROBERT B. BARLOW, JR. The sensitivity of the lateral eyes and median ocelli of the horseshoe crab, Limulus poly- phcmus, increases at night and decreases during the day. The brain imposes this circadian rhythm on the lateral eye via efferent optic nerve fibers which affect both the physiology and the morphology of the photoreceptors (Science 197, 86-89, 1977). In the ventral eye nerve, small fibers terminate on the photoreceptors in what are believed to be neurosecretory synapses, which presumably mediate efferent activity. We investigated the possibility that the ventral eye undergoes circadian changes in sensi- tivity. We recorded responses to brief flashes from the ventral eye "wart" in situ under con- stant darkness during the day and at night. No surgery or anesthesia was used. Thus far we have failed to detect any cyclic changes in visual sensitivity. In other experiments we excised the ventral eye and recorded light responses from the unsheathed nerve, using a suction elec- trode. Shocking the proximal stump of the nerve to activate the efferent fibers did not affect the recorded responses. Likewise, intracellular responses from impaled photoreceptors in the excised ventral eye were unaffected by electrical shocks to the nerve, except for a 10-20 mV depolarization that subsided in about 10 min. Finally, we attempted to influence the light responses of the ventral eye by using octopamine, which B. Battelle (National Eye Institute, N.I.H.) recently found in the efferent fibers of the ventral eye. Although we found octopamine to increase the ERG from the lateral eye, neither the intracellular nor the extracellular re- sponses from the ventral eye were affected by a wide range of octopamine concentrations. We conclude that under the conditions of our experiments the neurosecretory fibers of the ventral eye nerve do not affect the visual sensitivity of the ventral photoreceptors. Supported by NIH grants EY188, EY108, EY667, and NSF grant BNS77- 19436. NEUROBIOLOGY ^7 Octopamine increases the ERG oj the Linmlus lateral eve. LEONARD KASS AND ROBERT B. BARLOW, JR. When the horseshoe crab is kept in constant darkness, the lateral eye produces larger electroretinographic responses (ERG) during the night than during the day (Barlow et al., 1977, Science 197: 86-89). The elevated retinal sensitivity at night is mediated by efferent activity in the optic nerve trunk. During the day, the efferent activity stops and the sensi- tivity returns to a low, constant level. Pulses of current delivered to the distal end of the cut optic nerve during the day simulate the efferent activity and elevate the ERG to the nighttime level. B. Battelle (National Eye Institute, NIH) found high endogenous levels and the syn- thesis of octopamine in the lateral and ventral eyes of Limulus. She also found that in the ventral eye octopamine is localized within what appear to be small-diameter efferent fibers and terminals. We report here that injecting octopamine during the day beneath the cornea of the lateral eye in situ increases the amplitude of the ERG. Octopamine (1-10 n^l) injected at 1 /xl/min for 15 min increased the amplitude of the ERG 2-4X the daytime level, equivalent to about half the elevated nighttime level. Clozapine (25 /xM), a demonstrated antagonist of octopa- mine, reversibly decreased the ERG amplitude of the lateral eye at night. During the day, clozapine (25 /xM) blocked the increase in the ERG by octopamine (1 (J.M.). Elevation of the ERG by octopamine injection is reversible, is graded with concentration, and exhibits a time course similar to that caused by the endogenous efferent activity. Supported by the Grass Foundation, NIH grant EY 00667, and NSF grant BNS 77-19436. Are there membrane surface charges in the I'icinitv of the sodium pump.' GEORGE R. KRACKE AND PAUL DE WEER. According to the diffuse double-layer theory of Gouy and Chapman, the magnitude of the surface potential resulting from fixed charges on the exterior of the cell membrane will vary with the ionic composition of the bathing solution, becoming smaller as ionic strength is in- creased. Consequently, the attraction toward, or repulsion from, the membrane that ions undergo as a result of this surface potential will also decrease as ionic strength is increased. We have exploited this phenomenon in an attempt to determine whether the membrane of the squid giant axon bears external net charges in the vicinity of the sodium pump. The rate of inhibition of the sodium pump by cardiotonic steroids, which follows pseudo-first-order kinetics, was taken as a measure of the effective concentration of these compounds at the membrane/solution interface. Three artificial seawater solutions of normal, low, and high ionic strength were prepared. In the normal ionic strength (NIS) solution, 90% of the usual NaCl (425 mM) and MgCL- (50 mM) was replaced with N-methylglucamine-HCl. In the low ionic strength (LIS) solution these ions were replaced with sucrose, and in the high ionic strength (HIS) solution with MgSCX The three cardiotonic steroids used were un- charged ouabian, anionic actodigin hemisuccinate (CS~), and cationic digitoxigenin amino- glycoside (CS+). We found the rate of pump inhibition by ouabain, relative to that at NIS, to be 0.67 at LIS and 2.8 at HIS. This variation must be ascribed to chemical effects of the bathing solution changes. However, the corresponding relative rates for CS" were 0.42 at LIS and 6.6 at HIS, suggesting that the relative concentration of anions at the interface was reduced to 0.63 in LIS seawater and enhanced 2.4-fold in HIS. This is to be expected if the membrane in the vicinity of the sodium pump carries negative charges at a density of one electronic charge per 160 to 480 (average 280) A2. Preliminary data with CS+ suggest that, at least at HIS, a pattern opposite to that described for CS~ is followed, as expected. Further studies with a variety of charged steroids are in progress. Supported by NIH and the Grass Foundation. Inside-out voltage clamp in the squid giant a.von. J. LOPEZ-BARNEO, R. D. MAT- TESON, AND C. M. ARMSTRONG. In the study of membrane channels by noise analysis, reduction of background noisi essential. We have developed a new voltage clamp that is potentially quieter than the con- ventional one and is well suited to noise measurements, which require isolation of a small patch of membrane. This clamp has been tested by recording Na and K voltage lamp currents. 488 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Segments of axons were internally perfused and three electrodes were inserted: 1) a pipette to record the internal voltage (Vi) 2) a platinized current wire and 3) a glass pipette to record currents through patches of membrane. This pipette was bent at 90° in the last 150-200 fi.m, and had a tip diameter of 20-30 fim. The Vi pipette was wired to a clamp amplifier which maintained the axon interior at virtual ground potential. The external voltage (V0) was measured by an external electrode, and the membrane voltage (Vo-Vi) was then clamped by means of a second clamp amplifier (CA), whose output was wired to large external current electrodes by way of a 200 ohm resistor. Total membrane current (It) was measured as the IR drop across the resistor. Current from a small membrane patch (IP) was gathered by the bent pipette, and fed to a high gain current-to-voltage converter. The major advantages of this design are: a) stray capacitance of the Vi pipette is essen- tially eliminated, thus improving the frequency response; b) a low resistance path between the external current recording electrode and ground, present in the conventional clamp, is eliminated, thereby reducing noise; c) potential inside the axon is the same as in the bent pipette, eliminating stray capacitance effects in this electrode. Na and K currents recorded as I( were of similar magnitude and time characteristics to the ones recorded with conventional technique. Na currents recorded through the patch are quite similar to It. Thus far we have not achieved good isolation of a membrane patch, but perfusion with low conductivity solutions improves isolation and increases IP (up to 35 nA). Seventy successive IP records at the same potential were taken to calculate the ensemble variance of current fluctuations. Our analysis suggests that an important source of the fluctuations recorded thus far has been a small variation in the holding potential. Such sources of error must be carefully eliminated before making inferences about the contribution of the stochastic opening of the Na channels to the current fluctuations. Temperature effects on peak and steady state sodium currents in squid giant axons. RICK MATTESON AND CLAY M. ARMSTRONG. We have studied the effects of temperature on Na currents and Na-channel gating cur- rents with the intention of estimating the energy barrier encountered either by a mobile charged gating particle or an Na ion as it traverses the membrane field. Some preliminary results indicate that QKI measurements of many Na-channel properties are difficult to interpret since these measurements do not reflect the activity of a single process. A decrease in temperature in the range of 15°-1°C decreases the peak Na current generated by voltage clamp steps to +20 mV. The Qio of this effect range between 1.83 and 2.44. Temperature changes had little or no effect on the steady state sodium current at +20 mV in inactivation intact axons. In the absence of inactivation (i.e. after pronase treatment) steady state Na currents had a large Qio (about 2.0). Peak Na currents could be increased by delivering large positive prepulses (to + 100 mV) 20-30 msec before the test pulse. Steady state Na currents were essentially unchanged by this procedure. This potentiation effect was more pronounced at low temperatures: at 1°C peak current could be increased by up to 40% whereas at 12° C the increase was about 10%. In the absence of a prepulse, we often observed Na currents that had both a fast and slow activation phase, indicating that some channels may turn on very slowly. Following a prepulse the slow activation phase is eliminated. One consistent explanation of these results involves the hypothesis that there is a second type of Na channel that normally activates very slowly and does not inactivate. The activa- tion kinetics of these putative slow channels are presumably increased by positive prepulses so that they may contribute to peak Na current. Decreases in temperature seem to transform normal channels into slow channels, accounting for some of the decrease in peak Na current. As a result of this phenomenon, Qw measurements of peak sodium currents are difficult to interpret, since they do not reflect a single process such as temperature effects on single- channel conductance. Oscillation-free compensation jor the series resistance in voltage clamped squid axons. JOHN MOORK, DAVID TAUCK, AND MICHAEL HINES. In their classic voltage clamp experiments on squid axons, Hodgkin, Huxley, and Katz observed a small resistance, R», in series with the membrane and showed that it causes the voltage across the membrane to deviate from the desired command potential by an amount equal to the product of RH and the membrane current density, Im. By including a signal pro- NEUROBIOLOGY portional to I,,, in the control circuit, in principle they could compensate for the unwanted voltage drop across R«. In practice full compensation was rarely used because of the danger of destroying the axon by oscillations of the voltage clamp. Most investigators still employ this method of Rs compensation and use only partial (~-j) compensation. For accurate measurement of the kinetics and amplitude of membrane conductances, full compensation is necessary. We developed a method to allow complete compensation during ionic current flow without oscillations or ringing. An electronic bridge circuit subtracts from Im a current equal to the capacitive current, !<•; we apply a fraction of the bridge output, the ionic component of Im, to the summing point of the control amplifier through a poten- tiometer set to a value proportional to the measured value of Rs. Such a circuit is so stable that a 2- to 3-fold over-compensation for Rs is obtainable without ringing. The only compro- mise is that the Rslc error from the capacitive component of Im still causes a lag of a few /usec in the membrane potential rise following a command step. With this method, we obtain very accurate records of ionic membrane currents. Compared to the uncorrected records, the sodium current is markedly decreased and slowed in the nega- tive conductance region ; at all other potentials the currents are increased in magnitude. Electrical membrane properties of single and two-cell preparations from Chironomus salivary (/land. ANA LIA OBAID AND BIRGIT ROSE. Single and two-cell preparations were obtained from salivary glands of Chironomns larvae. These preparations' long-term stability in terms of electrical parameters such as resting po- tential (—30 to —50 mV) and input resistance (4-8 Mn) enabled us to determine nonjunc- tional (rm) and junctional (rj) membrane resistances under various experimental conditions, in particular those known to affect cell-to-cell coupling. In cell pairs with coupling coeffi- cients close to 1.0, rj ranges between 50 and 400 kfi (129 ±23 S.E., n—23), with rm usually 20-30 X rj. These values are of the order predicted by cable analysis of measurements on the intact gland. In the intact gland, exposure to Propionate medium (pH 6.5) lowers intra- cellular pH (pHi) — as measured with pH-glass microelectrodes — and uncouples cells. In the isolated cell pairs, rj under these conditions increases by at least 3 orders of magnitude, while rm decreases several fold. In the organ, increasing pH of the Propionate medium to 7.4 raises pHi from 6.5 to 7.3 and recouples cells, at times only transiently. Similarly, in the isolated cell pairs rj transiently decreases from 100 Mfl to 400 kft, with a subsequent renewed rise up to 11 M$2. pHi during these times of renewed uncoupling or renewed increase of rj is found to be unaltered at 7.3, indicating that pHi does not control junctional resistance in this case. Treatments which raise intracellular CaL'+, such as XaCX or alkalinization by washout of Propionate medium, uncouple the cells in the gland and raise the value of rj in the cell pairs up to 50 Mft. The cell pairs obtained from Chironomus salivary gland are a viable preparation, well suited for electrophysiological studies. The data obtained so far from cell pairs are in con- cordance with and confirm the results obtained with the organ. Mechanics and energetics of contraction in striated muscle of the sea scallop, Placopecton magellanicus. JACK A. RALL. The striated adductor muscle is responsible for the characteristic propulsive swimming behavior of scallops. This muscle is also interesting because contractile force is thought to be under Ca"'* control via thick filaments, in contrast to the exclusive thin-filament regulation of vertebrate striated muscle. The purpose of the present study was a general mechanical and energetic characterization of this muscle. Animals were kept at 12°C in continuously running sea water. Experiments were conducted at 10°C in aerated sea water on isolated muscle bundles (n = 7) with average dimensions of 3.1 X 0.25 XO.ll cm weighing 87 ± 10 mg, from scallops with an average shell length of 9.3 cm (about 4 years of age). Energy utilization was derived from measurements of the change in muscle temperature during contraction and relaxation, excluding recovery, using thermopiles. Maximum isometric tetanus force averaged 132 ± 5 mN/mnr with a twitch to tetanus ratio of 0.58 ± 0.02. In a twitch, time to peak force development was ms and time to half relax from peak force was 104 ± 2 ins. Muscles stimulated repetitiv< for 1.5 sec displayed a pronounced fatigue, i.e., force at last stimulus was 51 - tetanus force. A twitch elicited after a tetanus was transiently potentiated with peak force (132 ±8% of control twitch) occurring 5-15 sec post-tetanus and returning to control values 490 ABSTRACTS FROM M.B.L. GENERAL MEETINGS by 3 min post-tetanus. Twitch energy liberation averaged 7.6 ± 0.5 mj/g with a force to heat ratio of 10.6 ± 1.0 (dimensionless). This ratio decreased by 25 ± 2% when measured in twitches 3 min post-tetanus. Thus the same twitch force after a tetanus utilized 25% more energy than before a tetanus. In general the striated adductor muscle displayed properties consistent with its function of brief, rapid force development rather than slow, maintained force generation. This research was supported by the Ohio State University College of Medicine. Orthogonal polarization sensitivities of squid photoreceptors: implications for a retinal design. WILLIAM M. SAIDEL. Single- and multi-unit responses were recorded with suction electrodes from squid photo- receptor axons in an isolated eye preparation. Individual units responded to graded flashes and light steps with graded trains of spikes. The number of spikes and duration of the train depended upon state of retinal adaptation and relative angle of plane-polarized light. For example, a dark-adapted photoreceptor generated an 11.0 sec train of 186 spikes to a flash. When light-adapted, the response to a similar stimulus was a 0.5 sec train of 25 spikes. Re- sponses to steps of light displayed a changing character with adaptation. Dark-adapted photo- receptors responded to a series of graded steps with a higher frequency of spikes. As the cell adapted to light, the response was transformed into a phasic on/off response. With bright background illumination, only a phasic on response was elicited by a step. Individual photoreceptors also responded as analyzers of polarization. Spike trains from the same receptor in response to orthogonally plane-polarized light steps varied in frequency from 25-40%. Maximal variation in frequencies occurred with orthogonally polarized stimuli. As predicted from the orientations of microvilli in photoreceptor outer segments, two classes of receptors were observed whose maximal sensitivities in the polarization domain varied by 90°. Polari- zation sensitivity of a single photoreceptor may be considered as a broad band filter. Thus, the squid retina contains a two-channel mechanism for analyzing objects in a watery world illuminated by naturally polarized submarine light. Preferential sensitivity of photoreceptors to polarized light along angles parallel and perpendicular to the water surface would minimize stimulation by scattered or reflected light having random planes of polarization. This con- struction would enhance perception of contrast. This work was supported by the Grass Foundation and an NIMH Training Grant to the Department of Psychiatry at the University of California, San Diego, Medical School. Circadian rh\thin of photoreceptor cells in the Limulus lateral eve: further studies. TAKEHIKO SAITO, EHUD KAPLAN, AND ROBERT B. BARLOW, JR. The sensitivity of the Limulus lateral eye undergoes a circadian rhythm. At night a clock in the brain transmits nerve impulses via efferent fibers to the ommatidia of the com- pound eyes. The efferent input increases the amplitude of the photoreceptor response to light and decreases the frequency of discrete potential fluctuations in the dark (Nature, 286, 393- 395, 1980). We report here that cutting the optic nerve at night reverses the action of the circadian clock on the photoreceptors. After nerve section the frequency of the spontaneously occurring discrete waves increases and the amplitude of the receptor potential decreases. The signal- to-noise ratio of the retinular cells is thereby reduced. No consistent changes in cell resting potential were detected after nerve section. Spontaneously occurring nerve impulses recorded from eccentric cells J'H situ were triggered by discrete fluctuations in membrane potential ; impulses caused by injury (nerve section) were not. The potential fluctuations of eccentric cells arise from discrete waves of retinular cells via electrical coupling of the retinular cells to eccentric-cell dendrite. The origin of spontaneous optic nerve activity in situ can thus be traced to the photoreceptors. Under conditions of normal clock input, the decline of photoreceptor noise precedes the increase of retinal sensitivity. In the early evening hours when the endogenous efferent activity begins, the frequency of spontaneously occurring discrete waves of the photoreceptors decreases to a low level before the amplitude of the ERG increases significantly. This result is consistent with our previous finding, which indicates that separate cellular mechanisms underlie the clock-induced changes in the signal and noise characteristics of the photoreceptors. Supported by NIH grants EY-00667, EY-00108, EY-00188, and NSF grant BNS 77-19436. NEUROBIOLOGY An optical determination oj the resistance in series with the axolemma of Loligo pealei. B. M. SALXHKRG, F. BEZANII.LA, AND II. V. DAVILA. The membrane capacitance in squid giant axons has, in series with it, a small resistance arising primarily in the narrow Schwann cell clefts. The true transmembrane potential there- fore differs from that recorded between voltage electrodes in a voltage clarnp by an amount proportional to the membrane current. If this resistance is not properly compensated, serious errors are introduced into membrane conductance and AC impedance measurements. We have used an optical method to measure the series resistance by exploiting the potential -dependent changes in light absorption exhibited by an axon stained with a mcrocya- nine-oxazolone dye, NK 2367. This dye behaves as a linear potentiometric probe with a microsecond time constant. The transmission at 720 ± 15 nm of a perfused voltage clamped giant axon, mounted on the stage of a compound microscope, was recorded by a silicon photodiode positioned in the objective image plane during hyperpolarizing and depolarizing potential steps (+70 mV from —70 mV). The optical record during the hyperpolarization closely resembled the voltage clamp step, since no ionic current crosses the series resistance. During the depolarizing step, however, the transmitted intensity revealed a component propor- tional to the membrane current and series resistance. The feedback compensation (Hodgkin, Huxley, and Katz, 1952) was then varied until the optical signal assumed a square shape during both the hyperpolarizing and depolarizing potential steps. The value of the series resistance could then be read from a calibrated potentiometer in the feedback circuit. In natural seawater, the value of the series resistance obtained in this manner was 2.8 ± 0.6 ohm cnr. The method described here is not essentially electrical, but employs a 15 A molecular voltage probe to afford an independent determination of the series resistance. Because the probe is located between the axolemma and the external series resistance, it is capable of distinguishing between the effects of the series resistance itself, and the anomalous dispersion of the membrane dielectric. This work is dedicated to Kenneth S. Cole on the occasion of his 80th birthday, and was supported by NSF grant BNS 77-05025 and PHS grants AM 25201 and NS 12253. The birefringence response of voltage-clamped internally perfused a.rons. VIRGINIA SCRUGGS AND DAVID LANDOWNE. The birefringence of squid axons was measured by placing the axon between crossed polarizing crystals and measuring the transmitted light. When the membrane potential was changed from —10 to —140 mV the birefringence of the axon increased by about 10"4. The birefringence response to a pulse from —70 to 0 mV was smaller and rose more slowly when compared with the response to a hyperpolarizing pulse. Pairs of pulses were used to compare the birefringence response to a depolarizing pulse with and without sodium inactivation. When the records with a conditioning pulse to —140 or to 0 mV were made to coincide at the beginning of the second pulse they diverge after an initial fast phase. After the first 100 /usec the response corresponding to the inactivated sodium current has a smaller slope. Both of these assymmetries in the birefringence response were also seen with symmetrical pulses with the depolarizing pulse beyond the sodium equilibrium potential, and also in the presence of tetrodotoxin, demonstrating they are not simple artifacts produced by current flow. Suported by NIH research grant NS 13789. Arsenazo III reveals long-lasting intraccllular calcium transient following the action potential in squid giant prcsynaptic terminal. STEPHEN J SMITH, MILTON P. CHARLTON, AND ROBERT S. ZUCKER. The calcium indicator dye arsenazo III was used to measure cytoplasmic calcium concen- tration transients in presynaptic terminals of the stellate-ganglion giant synapse of Loligo pealei. Terminals were filled with arsenazo by passing a steady current from a dye-filled micropipette inserted near the tip of the terminal digit. Light absorption by the dye was measured by focusing filtered light from a tungsten-halogen source on the presynaptic terminal, and conducting transmitted light to a photodiode using a light pipe. Individual presynaptic action potentials elicited rapid changes in light transmittance, with the spectral properties expected from an increase in calcium binding to the dye: A maximal 492 ABSTRACTS FROM M.B.L. GENERAL MEETINGS transmittance decrease was observed at 660 nni, a smaller decrease at 610 nm, and no change at 578 or 690 nm. The optical signal reached a peak in about 1 msec beginning coincident with the peak of the presynaptic action potential. Recovery of baseline optical transmittance was very slow, requiring approximately 20 sec at 15°C. Action potentials elicited repetitively over a wide range of intervals always gave incremental signals of similar amplitude. At high dye concentrations, a marked suppression of synaptic transmission was usually observed but good optical signals were obtained by signal averaging at low dye concentrations, where transmission appeared essentially normal. Detecting calcium ions in cytoplasm at this short latency supports the currently accepted view that transmitter release is triggered by such a calcium transient. The very slow time course of dye signal recovery compared to transmitter release termination probably reflects spatial redistribution of calcium ions in the cytoplasm following entry at the surface mem- brane. The detection of residual free calcium long after the action potential is consistent with earlier suggestions that residual calcium may underlie certain aftereffects of transmission such as presynaptic facilitation. SJ.S. and M.P.C. are Steps Toward Independence Fellows. R.S.Z. supported by a Sloan Foundation Fellowship and NIH grant NS15114. Calcium and potassium activities in the hemolymph of the squid, Loligo pealei. RODERIC E. STEELE AND DANIEL L. GILBERT. Although the Ca and K concentrations in the hemolymph of the squid, Loligo pealei, have been previously measured, the activities of these ions have not. Physiological functions pre- sumably depend upon the chemical activities, rather than the concentrations. Activity ratios were measured using microelectrodes (2-5 nm) filled with CaL>+ exchanger (WP Instruments, Xew Haven, CT) or K* exchanger (Orion Research, Cambridge, MA) and saturated KC1 reference electrodes. For K+, a double well wax sample holder was used to prevent K+ leakage from the reference electrode into the sample being measured. The activity ratio of the Ca'+ in the hemolymph to the Ca"+ in artificial sea water (9.27 mM CaCL, 9.0 mM KG, 423 mM NaCl, 22.9 mM MgCU, 25.5 mM MgSOO was 1.05 ± 0.02 (SEM)(n = 9 squid). For K+, this ratio was 1.21 ± 0.04 (SEM) (n^ 11 squid). Using activity coefficients (Smith, CRC Handbook of Marine Science, Vol. I, CRC Press, Cleveland, 1974), the calculated activities in millimoles/kg H2O are 2.0 for Ca2+ and 6.8 for K+. When compared to the natural sea water in which the squid were swimming, the activity ratios become 0.99 ± 0.02 (SEM) for Ca2+ and 1.26 ± 0.08 (SEM) for K+. However, Requena ct al. (1977, J. Gen. Physiol. 70: 329) found that the Ca"+ concentration in the external media had to be lowered to 3 mM (corresponding to a concentration ratio of about 0.3) in order for the excised squid nerve to be in a steady state for Ca. Perhaps this difference is due to a change in the active transport or permeability for Ca in the excised nerve. The K+ activity ratio of 1.26 can be compared with molal concentration ratios of hemolymph/sea water of 2.22 (Robertson, 1965, /. Exp. Rial. 42: 153), 1.94 (Hodgkin, 1958, Proc. Roy. Soc. B. Biol. 148: 1), 1.51 (Manery, 1939, J. Cell. Comp. Physiol. 14: 365), and 1.34 ( Shoukimas ct al., 1977, Biophys. J. 18: 231). K* activities were elevated by 20-50% in 3 samples, which were taken 2 min after the initial sample, presumably due to K+ leakage from cells. Such a leakage could explain the higher K* values found by some other investigators. We wish to thank Kenneth S. Cole for stimulating this investigation. Elimination of synapses from identified lobster motor neurons during dei'elop/uent. PHILIP J. STEPHENS AND C. K. GOVIND. During development, inherent limitations are imposed on the peripheral fields of motor neurons. However, the mechanisms that impose such peripheral-field limitations are poorly understood. We have categorized the peripheral fields of two identifiable lobster motor neurons during normal development. Pairs of deep extensor muscles are located in each abdominal segment, and in each hemi- segment the muscle is divided into medial (DEAM) and lateral (DEAL) bundles. The second root nerve carries six motor axons to the deep extensor muscle in each hemisegment. The common excitor (CE) and common inhibitor (CI) motor axons in each root nerve can be identified at the periphery since they are the only axons that innervate both the DEAL and the DEAM. NEUROBIOLOGY In abdominal segments 2-5 of embryonic lobsters the CE and CI axons make functional connections with deep extensor muscle fibers in their o\\n segment and in adjacent anterior and posterior segments. Junctional activity was not produced via central pathways and there- was no detectable electrical coupling between muscle fibers in adjacent segments. Therefore the CE and CI motor axons must branch and make synaptic connections with deep extensor muscle fibers in three abdominal segments. This innervation pattern was found for the CE and CI axons in embryonic, larval, and early juvenile lobsters. However at juvenile stage 6 the CE and CI connections were' con- fined to the deep extensor fibers in their own segment. This is the typical distribution of CE and CI synapses in the adult. Our results suggest that one way for a motor neuron to define its peripheral field involves an initial over-production of synapses followed by selective elimination. Supported by a Grass Fellowship (P.J.S.), NIH-NIXCDS and the Muscular Dystrophy Association of Canada (C.K.G. and H. L. Atwood). Gap junctions: quantitative comparison of reduction in conductance />v // and b\ Ca ions in an internally perfused preparation. }. H. STKRN, D. C. SPRAY, A. L. HARRIS, AND M. V. L. BF.NXKTT. Conductance of gap junctions between Anibystotia and Fundulus blastomeres is decreased at low intracellular pH. The relation is well fit by a Hill plot with KH — 50 mM (pK.n~7.3) and Hill coefficient of 4.5 (Spray, Harris, and Bennett, Science, in press). To assess the relative effect of calcium ions we perfused one cytoplasmic face of the junctions with defined and rapidly changeable solutions. One cell of a coupled pair of Fundulus blasto- meres (-"100 /j. diameter) was sucked into an internally perfused pipette ( — 20 n internal diameter) until the second cell sealed against the pipette. The first cell was broken and its cytoplasm washed away by perfusate, leaving a membrane patch which included gap junctions. With current applied between bath and pipette interior, the ratio of voltages across the outer membrane of the intact cell and the patch gave the ratio of conductances of the patch (gp) and outer non Junctional membrane (gn). Changes in gp/gn were presumably largely in junc- tional membrane since perfusion of a patch of a single cell had little effect. Perfusion solu- tions contained 100 mM KC1, 5 mM NaCl, 10 mM HEPES, and 10 mM EGTA or NTA. CaCU and MgCl2 were added to obtain 1 mM free Mgf+ and desired levels of buffered Ca"+, confirmed with Ca~*-sensitive macroelectrodes. Changes in g,,/gn with perfusates at pCa 7.0 and pH 6.8, 7.2, and 7.8 were rapid, reversible, and consistent with our previous measurement on intact cells. gP/gn was essentially unchanged by perfusates at pH 7.8 and pCa 7.0-4.0 but was progressively reduced by perfusates at pCa 3.9-3.0, with an apparent Kca ~~/ 0.4 mM (pKca — 3.4). Thus Junctional conductance is in- sensitive to cytoplasmic free Ca1'* concentrations below 0.1 mM (pCa 4) whereas Junctional conductance is profoundly depressed by 0.1 fiM hydrogen ion (pH 7). External K^ and Rb^ retarded closing of potassium channels. R. P. SWENSON, JR.. AND C. M. ARMSTRONG. Gating properties of ionic channels have previously been considered to be insensitive to changes in the rate or species of ions moving within them. For example, Hodgkin and Huxley (1952) reported no change in Na channel kinetics when external Na concentrations were lowered, and Chandler and Meves (1965) found identical Na channel kinetics with either Xa* or K+ as the current carrier. In contrast to the immunity of channel gating to ions moving through the pore ; blocking cations, e.g. local anesthetics, which bind within the Na channel, can prevent the Na channel from closing. In contrast to previous results, we have found the closing kinetics of K channels to be slowed following equimolar substitution of extracellular Na+ by either K* or Rb*. Using a double pulse protocol to avoid complications arising from changes in K* concentrations in the confined extracellular space near the membrane, we were able to accurately measure K chan- nel closing kinetics. At —70 mV the time constants were 3.0, 4.5, and 6.5 msec in artificial sea water (ASW), 200 K*-ASW, and 200 Rb*-ASW respectively at a temperature of 8°C. stantaneous I-V relations in the same solutions suggest that external K* and Rb* enter the channel equally well, but Rb* remains in the channel about five times longer at inside negative potentials. The longer dwelling time of Rb+ in the K channel than the K* itself correlates well with the slower rate of closing when Rb+ is present externally. 494 ABSTRACTS FROM M.B.L. GENERAL MEETINGS When blocking BaLV ion binds near the external end of the potassium channel, K channels close with normal kinetics, but binding of TEA* near the inner end of the channel slows closing. In light of these observations, we suggest that a cation, permeant or not, at a site near the inner end of the channel is responsible for the retarded closing of potassium channels. This work was supported by an XIH post-doctoral fellowship NS 06201-01 to R P S and XIH XS 12547 to CM. A. Swelling of squid giant a.von during action potentials. I. TASAKI AND K. IWASA. Rapid pressure changes and surface displacements in the squid giant axon were demon- strated during action potentials. To measure pressure changes, a Gulton piezo-ceramic bender was used in conjunction with a voltage follower. For detecting pressure changes, a Fotonic sensor placed about 0.1 mm above a small piece of gold leaf on a giant axon was employed. The output of the Gulton bender or of the Fotonic sensor was amplified with a capacity- coupled amplifier with a gain of 1000 and was led to a signal averager. The time course of the mechanical response of the axon observed by these two methods was diphasic. The first phase was found to coincide with the depolarizing phase of the action potential and to represent swelling of the axon. The magnitude of the outward displacement of the axon surface at the peak of swelling was about 1 nm and the corresponding pressure rise was, on the average, 10 fj.g per 0.3 mnr. The second, shrinkage phase of the mechanical response started roughly at the onset of the hyperpolarizing after-potential : at its peak the magnitude of the surface movement observed was 1-4 nm and that of the pressure change was 10-50 ^g per 0.3 mm2. The mechanical responses associated with action potentials induced by lowering the external Ca-ion concentration were very similar to those evoked by electrical stimulation. External application of ouabain (up to 1 mM) or trypsin (1 mg/ml) had no clear effect on the mechanical responses for more than 1 hr after application ; this finding suggests that Schwann cells outside the axon play little or no role in producing mechanical responses. The present findings indicate that action potential production is accompanied by drastic changes in the macromolecules in or near the axolemma. Many aspects of the experimental results are con- sistent with the old theories of excitation proposed by Loeb, Hober, Teorell, and others. Presynaptic calcium currents and facilitated transmitter release in the giant synapse of Loligo pealei. ROBERT S. ZUCKER, MILTON P. CHARLTON, AND STEPHEN J SMITH. At moderate levels of transmission, the giant synapse in the stellate ganglion of the squid exhibits facilitation, such that the second of two EPSPs (excitatory postsynaptic potentials) is larger than the first. We tested the possibility that synaptic facilitation is caused by a larger calcium influx during the second presynaptic impulse. We measured the current carried by calcium ions using a three-electrode voltage clamp of the presynaptic terminal. Sodium currents were eliminated by external tetrodotoxin (2 X 10~7 g/ml) and potassium currents were blocked by 2 mM 3,4-diaminopyridine and intra- cellular injection of tetraethylammonium. Leak and capacitative currents were reduced and calcium currents smoothed by averaging the currents accompanying several equal and opposite pulses. When two spike-like depolarizing pulses (50 mV amplitude, 1 ms duration), giving a moderate level of transmitter release (5 mV EPSPs), were separated by 3 ms, the second EPSP was about 50% larger than the first. Both pulses were accompanied by a cadmium- sensitive inward calcium current of about 40 nA in the end terminal region. In most experiments the currents during the two pulses were identical. In a few prepara- tions, the peak inward current during the second pulse appeared about 5% larger than that accompanying the first. When a single depolarizing pulse was increased in amplitude or duration to give an EPSP equal to the facilitated EPSP, the calcium current was increased by 12-25%. We conclude that calcium channels facilitate very little if at all, and that facilita- tion of calcium influx cannot be primarily responsible for facilitation of transmitter release at this synapse. Supported by NIH grant NS15114. R.S.Z. is an Alfred P. Sloan Research Fellow and M.P.C. and S.J.S. are Steps Toward Independence Fellows. PARASITOLOGY AND IMMUNOLOGY 4<;5 PAKAS1TOLOCY AND IMMUXOLOGY Solid-phase radioimmune assay to detect antibodies to Entamoeba histolvtica. SKRGIO ARIAS-XEGRETE, DIANNK McMAHON-PRATT, WILLY F. I'IKSSKNS, AND IAN ROSENBERG. We describe a solid-phase radioimmune assay for detecting antibodies to Entamoeba his- tolytica strain HM-1 cultivated axenically in TYI-S-33 (trypticase, yeast extract, iron and serum) media. To prepare the E. histnlytica antigen, the amebas were pooled and centrifuged for 3 min at 500 X g, and washed three times with phosphate buffered saline (PBS) at pH 7.2. The trophozoites were disrupted ultrasonically and the insoluble material removed by centrifugation for 5 min at 12,000 X g. Phenylmethylsulfonylfluoride and Aprotinin were added to a final concentration of 2 mM and 0.5 units, respectively. The antigen was stored at —70°. Five female mice BALB/cJ were injected intraperitoneally with 75,000 trophozoites. On Day 14 post-inoculation, the animals were bled from the retro-orbital sinus. The serum was heated 30 min at 56° C. Polyvinylchloride plates were coated with 4-5 /j.g of E. histolytica antigen in 0.05 ml of PBS and incubated overnight at 4°C. The plates were then washed five times with PBS containing 4% fetal calf serum. Serial dilutions of 0.05 ml samples were made in the plates and incubated 6-7 hr at 4°C. To detect mouse anti-ameba serum bound to the plates, 0.04 ml of ^I-rabbit anti-mouse IgG were added and incubated for 1 hr in an ice bath. The plates were washed and the samples counted in a gamma counter. The results showed that mouse anti-E. histolytica serum specifically bound the antigen as compared to the normal mouse serum. This antigen-antibody reaction is time- and temperature-dependent. This assay is more sensitive than enzyme-linked immunosorbent assay in detecting anti-ameba antibodies produced during the primary immune response to E. histolytica. Inflammation in the sea star, Asterias forbesi. F. B. BANG. Inflammation is a response of living tissue to local injury, leading to a local accumulation of blood cells and fluid. Following the injection of foreign pigmented cells of the sea urchin Arbacia into small (6-10 gm) sea stars, plugs of phagocytosed Arbacia cells accumulated within the transparent papulae of the skin of the sea stars. During the early phase (8-24 hr) these cells migrated through the papulae, often producing holes in the papular wall. A change of permeability of the papulae was detected after 4 hr by immersing the star in an 0.1% seawater solution of Evans blue dye for 20 min. This change in permeability disappeared by 48 hr with beginning resolution of the lesion. Xot all of the urchin cells were carried out of the otherwise normal sea stars in this way ; many were digested, and brownish pigment persisted in the papulae for a week or more. Inert material such as carmine powder is also known classically to be disposed of by penetration, but large amounts of carmine are found circulating within the amebocytes of the coelom weeks after injection. The inflammatory process induced by urchin cells spread to the separate water vascular system, especially after several injections of urchin cells into the coelom, as shown by marked clumping of amebocytes within the antenna and feet of the star. Edema, consisting of a ballooning of the epithelium, especially around the spines and paxillae of the skin, occurred usually after several injections of urchin cells, often in the injected limb, and lasted for several days. About 50 individual stars have been followed through this sequence of changes, and study of the influence of external variables on the response has been initiated. Supported by NIH 5 P50 HL 19157. Bacterial kidney disease oj rainboi^ trout in sea water. JAMES C. CARLISLE AND JAN SPITSBERGEN. To investigate the influence of sea water acclimation on the pathogenesis of kidney disease (KD) in rainbow trout, Salino gairdneri, 16 juveniles were infected on day 0 by intra- peritoneal injection of 109 Rcnibactcrium sulmoninamm. the causative agent, and 16 were sharn injected. Each group was subdivided into a fresh water group and a group to be expose gradually increasing salinity to reach full strength hy day 31. Each of the four groups was kept in a 20-gal aerated and filtered aquarium at 15°C with 20% of the water replaced daily. Dissolved Oa remained above 9 mg/1, NH4 below 0.1 mg/1, and NO-- below 0.3 mg/1. 496 ABSTRACTS FROM M.B.L. GENERAL MEETINGS A fish from each group was skin tested weekly by injecting 5 X 107 R. salmoninarum into the adipose fin, then bled and killed 1 week later. Humoral and cellular immunity were evaluated by bacterial agglutination, indirect fluorescent antibody test, and leukocyte migration inhibition. Spleen cells were tested for their ability to bind antigen. Total mortality in the sea water infected group was 55% with a mean of 11.6 days between infection and death. In fresh water, mortality was 57.2% with a mean of 24 days until death. Among the uninfected fish in sea water, 22.2% died after an average of 32 days in increasing salinity. No fresh water control fish died. Skin tests did not demonstrate grossly visible delayed type hypersensitivity. Infected fish showed lymphocyte-mediated migration inhibition on days 35 and 42. Humoral antibodies as measured by bacterial agglutination were present in infected fish on days 14, 21, and 28, and in one control fish on day 28. Antibody activity measured by indirect immunofluorescence was present in infected trout on days 28 and 35, and in control trout on day 28. On the 28th day all fish had been skin tested 1 week previously, while on day 35 the controls had not been skin tested. Antigen binding cells were present in the spleens of infected trout on days 21 and 28. SchistosomiasJs immune erosion. FRANCIS \Y. KLOTZ, SARA ANDKRSON, SERGIO ARIAS-NEGRETE, DAVID KOECH, BARBARA SHERRY, AND ALAN SHER. Lung forms of Schistosojiia inansoni acquire host antigens and are not recognized by the im- mune system as foreign. We investigated this immune evasion by lung schistosomula of S. inansoni. To determine the sites of immune effector mechanisms and the efficacy of immunization against schistosomula, we infected inbred mice with 5". inansoni cercariae skin-derived schisto- somula, and lung-derived schistosomula. Cercariae collected from 5". iiiansoni-'miecied Biompha- laria glabrata were irradiated with a 20 Krad dose of X rays from a ""Co gamma source. Thirty C57BL/6J female mice were each exposed percutaneously to 500 X-irradiated cercariae. Six weeks later these mice and 30 age-matched controls were challenged with 200 normal cer- cariae percutaneously, 200 skin-derived schistosomula intravenously or 5-day-old lung-derived schistosomula intravenously. The lung schistosomula were derived from syngeneic, age-matched donors. Four weeks post-challenge, the mice were killed and the adult worms recovered by per- fusion from the hepatic portal system. We recovered 5 ± 6 adult worms from the immunized, cercaria-challenged animals compared to<71 ± 27 from the controls for 93% protection. We recov- ered 10 ± 7 adults from the immunized, skin-schistosomula-challenged animals compared to 83 ± 10 from the controls, indicating 88% protection. We recovered 6 ± 7 adults from immunized, lung-schistosomula-challenged mice compared to 67 ± 18 from the controls, indicating 91% protection. The data is significant at the 99% confidence limit. Immune evasion by lung schistosomula did not occur in this preliminary experiment. A natural protozoan-derived ionophore: a possible mechanism of cytoto.vicity by Entamoeba histolytica. EILEEN LYNCH, ANDY HARRIS, IAN ROSENBERG, AND CARLOS GITLER. Using the technique of planar lipid bilayers, we have reconstituted a natural ionophore channel from extracts of the pathogen E. Histolytica. Both the conditioned media in which amoebas had been grown and cell homogenates of the organism exhibited spontaneous channel- forming activity, whereas control media did not. Addition of microgram quantities of extract to the bilaycr results in stepwise increases in membrane conductance equal to 1600 pmhos in 1 M KC1. Such increments occur at a linear rate to increase membrane conductance up to 5 orders of magnitude. Ion channels so induced display a moderate cation selectivity with the permeability ratio of potassium to chloride = 4.6 and that of sodium to chloride = 2.5. The amoeba pore exhibits a time variant, voltage dependent behavior with both negative and posi- tive values of transmembrane field decreasing membrane conductance. More than 75% of the induced conductance is inactivated by protease addition to the system. Additionally, under recording conditions, we show that microgram quantities of amoeba extract irreversibly decrease the input membrane resistance of an impaled Fundiilus blastomere 1.12 mega -ohms. Such a resistance change corresponds to a conductance increase of 630 nmhos ; such an increase would be produced if 4000 amoeba channels were inserted across the blastomere membrane. These data demonstrate that extracts of E. Histolytica can impose striking changes upon the membrane permeability of both living and artificial systems. Furthermore, these data PARASITOLOGY AND IMMUNOLOGY 4<;7 suggest that the in I'iro meclianisni of cytotoxicity and patliology induced by the amoeba may operate through the insertion of parasite-derived channels into the host cell membrane. This work was supported in part by XIH Training Grant T-32-GM-7288, and the M.B.L. Labeling of inicrofilariac of Krugia malayi. WILLY F. PJESSE> , IAN KOSKNBERG, CARLOS GITLKR, AND STUDENTS OK THE BIOLOGY OK PARASITISM COURSE. The antigens that play a role in the biologic interaction between microfilariae of B. malayi and the infected host remain to be identified. To identify the antigens, we developed methods to label parasite materials with lectins and by iodination procedures. Of seven lectins tested, only wheat germ agglutinin bound to the sheath of microfilariae, in a pattern that appeared linear and homogeneous by direct immunofluorescence. Microfilariae could also be labeled with ''"I by the lactoperoxidase, Bolton and Hunter, iodogen, and iodonaphtylazide techniques. Each labeling method produced a different banding pattern on linear SDS-polyacrylamide gradient gels of deoxycholate extracts of labeled microfilariae. A single band (mol. wt. 67,000 daltons) was labeled by all four methods. Whether any of the labeled materials can be adsorbed onto wheat germ agglutinin-sepharose columns or precipitated by immune serum remains to be determined. Monoclonal antibodies to Leishmania enriettii ami tubnlin. D. McMAHON PRATT AND STUDENTS OK THE PARASITOLOGY COURSE. Kohler and Milstein demonstrated in 1975 that by fusion of myeloma cells with spleen cells from immunized animals, cloned hybrid cell lines could be produced which secrete anti- bodies specific for the immunizing antigen. Such monoclonal antibodies can potentially be of great use in the diagnosis and treatment of parasitic diseases. In the Parasitology Course at Marine Biological Laboratories this summer, as a class experiment, monoclonal antibodies to the protozoan parasite Leishmania enriettii were produced. The experimental procedure was as follows : Balb/c mice, immunized with membranes from L. enriettii, were boosted with antigen. Three days later the spleens from these animals were removed and fused with cells of the myeloma cell line NS-1 using polyethylene glycol 1000. The fused cells were cultured in 96 well microtiter plates. Hybrids were selected by growth in HAT (hypox- anthine-aminopterin-thymidine) medium. Of the total 2612 wells cultured, hybrid cells grew in 904 wells. After 14 days of culture, antibody production by the hybrid cells was evaluated using a solid phase radioimmune assay. Two hundred and sixteen wells were determined to be secreting antibody. Selected antibody producing wells were cloned by limiting dilution on a feeder layer of x-irradiated (2500 rads) mouse spleen cells. Two weeks later the cloned cells were evaluated for antibody production. Of the initial 42 antibody producing wells, 19 resulted in cloned antibody producing hybrids. Several of these monoclonal antibodies to L. enriettii are being further characterized with respect to both the subclass of immunoglobulin and the Leishmania antigen recognized. Of these, one monoclonal antibody has been demon- strated specific for tubulin, a major membrane protein constituent of Leishmania. The identi- fication of the tubulin specific monoclonal antibody was done in collaboration with Marcel Hommel and Mary Porter. Stages in the life cycle of the digenetic treniatode, Lasiotocus minutus (Manter, 1931) Thomas, 1959. HORACE W. STUNKARD. Lasiotocus minutus, a parasite in the intestine of Mcnidia menidia, was described and named by Manter (1931) from specimens taken at Beaufort, North Carolina. It is a member of the Monorchiidae, a large family with 12 subfamilies, about 30 genera, and 100 species, which infect marine fishes in all parts of the world. Observations on different stages in the life history of certain species have been recorded and the complete life cycle of a single species, Monorchcidcs cumingiac, was worked out by Martin (1940) at the M.B.L. The paper by Stunkard and Uzmann (Biol. Bull. 116: 184-193) on the life cycle of Proctoeces maculatus included the description of a microcercous cercaria, designated as Cercaria adrano- ccrca, found in Gemma gemma collected at Boothbay Harbor, Maine. During the summer of 1980, examination of specimens of G. acmma taken in the Woods Hole area yielded infec- tions with C. adranocerca and study has elucidated the life cycle of the species. The cer- cariae, which proved to be the larval stage of L. minutus, are microcercous and cannot swim. 498 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Soon after their emergence from sporocysts, they become encysted in the haemocoele of the clam. They may be extruded singly or in strands embedded in a jelly-like matrix. The metacercariae were fed to small fishes and a series of developmental stages was recovered from the silversides, Mcnidia mcnidia. Xo infections were established in Fundnlus hetcro- clitus or 0-year Pseudopleuronectes amcricanus. Revisions of classification of monorchiid trematodes have resulted in many changes and the species long known as Genalopa minuta is now recognized as Lasiotocus minutns ( Manter, 1931 ) Thomas, 1959. Investigation supported by NSF-DEB 80-06150. I'urifieation of parasite niKX.-l. I). \\'IRTII, R. CARTER, L. MILLER, AND M. HO MM EL. To identify stage specific mRXA in two parasite systems analyzed, total mRNA was purified from each parasite and translated in a cell-free system. The products were analyzed on protein gels. In Plasmodium (/allinaccum, the gamete and asexual stage RNAs were isolated. There appear to be several parasite-specific proteins in the translation products of RXA isolated from gametes and these proteins will be analyzed using monoclonal antibodies raised against gametes. The RNA isolated from L. cnricttii amastigote stage directs the syn- thesis of many proteins and these need to be identified using specific antisera. Tubulin genes in parasites. D. WIRTH, C. GITLER, AND I. ROSENBERG, (IN COL- LABORATION WITH C. FRENCH, D. MISCHKE, AND M. PARDUE). Two parasites examined, Trypanosome brucci and L. cnricttii were known to contain tubulin ; no tubulin had been detected in the third parasite, Entamocba histolytica. To detect tubulin genes in parasites, DNA was purified from each parasite, cut with a restriction enzyme (Eco RI ) and fractionated on an agarose gel. The fractionated DNA was trans- ferred to nitrocellulose and hybridized to a radiolabeled tubulin gene isolated from Drosophila. The Drosophila tubulin gene clearly hybridized to a single Ecol fragment of L. cnricttii DNA and to two Ecol fragments of T. brucci DNA. Xo hybridization to Entamocba histolytica DXA was seen in the first experiment and weak, possibly non-specific hybridization was found in a second experiment. Cloning genes of Leishmania enriettii in K. coli. I). WIRTH, DIANNE McMAHON PRATT, JOHN DAVID AND STUDENTS IN THE BIOLOGY OF PARASITISM COURSE. The goal of this experiment was to make a genomic library of L. cnricttii DNA and to detect expression of parasite-specific antigens in E. coli. If such expression did occur, antigens suitable for vaccination might be produced in bacteria. Total DNA was purified from L. cnricttii promastigotes, cut with the restriction enzyme PST I and ligated to the plasmid pBR322, which had also been cut with PST I. E. coli bacteria were transformed with these recombinant plasmids and the transformed bacterial clones selected for growth in the presence of tetracycline, a drug resistance to which was encoded by the plasmid. Approximately 1000 transformed E. coli clones arose per microgram of recombinant plasmids. Of the transformed clones 82% contained a segment of L. cnricttii, as indicated by the sensitivity of the clones to ampicillin. The transformed E. coli clones were screened for the expression of parasite antigens using a solid phase radioimmunoassay. The antiserum used in this assay was raised against crude L. cnricttii membranes and contained antibodies which react with several proteins. Five of 10,000 clones screened appeared to react with the antiserum. These clones have been isolated and will be retested. PLANT PIGMENTS AND PHOTOSYNTHESIS Colorless proteins in phycobilisomes oj rhodophycean and eyanobacterial species. R. S. ALBERTE, T. A. KURSAR, AND R. F. TROXLER. Phycobilisomes are multimeric aggregates composed of phycobiliproteins which occur on the outer surfaces of thylakoid membranes. They function as light-harvesting antennae for photosynthesis in rhodophycean and eyanobacterial cells. Phycobilisomes contain allophyco- ^^^•^B PLANT PIGMENTS AND PHOTOSYNTHESIS 4<;<; cyanin (A PC; X max. = 650 nm), phycocyanin ( PC ; X max. = 620 mn) and in some species, pliycoerythrin ( PE ; X max. = 500, 545, 565 nm). Recently, it lias been shown that phycobili- somes contain 8-10 colorless proteins in addition to the phycobiliproteins and it has been sug- gested that these colorless proteins function as structural components by linking the various phycobiliproteins to one another. In the present work, phycobilisomes from the unicellular rhodophyte, Cyanidium caldariittn, the unicellular cyanobacterium, Aiwcystis nidulans (both producing APC and PC), were compared to those of the macrophytic ml algae, Ncogardluclla sp. and Ccramium sp. (both producing APC, PC, and PE). C. caldarium phycobilisomes, not previously described, displayed a single absorption maxi- mum at 624 nm (due to PC). Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of C. caldarium phycobilisomes revealed approximately nine colorless proteins in the 24-75 kd range after staining with Coomassie Brilliant Blue. A. nidulans phyco- bilisomes displayed a broad absorption band (X max. = 620 nm) due to PC, and SDS-PAGE revealed approximately nine colorless proteins in stained gels. A new procedure was developed to enable isolation of phycobilisomes from macrophytic red algae. Phycobilisomes from Neogardhiella sp. and Ccramium sp. displayed absorption maxima at 498, 544, and 568 nm (due to PE), at 615 nm (due to PC) and at 650 nm (due to APC). Analysis of phycobilisomes from these species by SDS-PAGE revealed the 7 sub- unit of PE (mol. wt. s; 31 kd) in addition to approximately nine hands ascribed to colorless proteins (after staining) in the 24-105 kd range. The pattern of colorless proteins on gels displayed minor differences among species and differed significantly with respect to the amount of protein (Coomassie Brilliant Blue staining) in a given Mr range. The present study is the first to demonstrate the presence of colorless proteins in phyco- bilisomes of C. caldarium, Neogardhiella sp., and Ccramium sp. and provides a point of de- parture for demonstrating the function of the respective colorless proteins in attaching the various phycobiliproteins to one another. Supported by NATO grant 1721, NSF PCM79 06638, and NIH GM 22822 and GM 23944. Phycocyanobilin synthesis jroin exogenous lieine. S. B. BROWN, R. F. TROXLER, AND R. S. ALBERTE. In animals, biliverdin evidently is derived from protoporphyrin IX via the henie of hemo- proteins. In plants, it is likely that, as with biliverdin, the carbon skeleton of phycobilins is derived from protoporphyrin IX, but it is uncertain whether ring cleavage occurs via a mag- nesium porphyrin derivative or via heme. Recently, we showed that the unicellular rhodo- phyte, Cyanidium caldarium, can take up heme from the suspending medium. This organism synthesizes chlorophyll-a, allophycocyanin, and phycocyanin when grown autotrophically but does not produce these pigments when grown heterotrophically. Photosynthetic pigments are synthesized when heterotrophic cells are placed in the light. When dark-grown (heterotrophic) cells were incubated with 14C-heme and exposed to light, the phycocyanobilin chromophore was found to be 14C-labeled, indicating conversion of heme to bile pigment by this organism. We have repeated these experiments using the cyanobacterium, Anacystis nidulans, which unlike C. caldarium, cannot be grown heterotrophically but produces the same photosynthetic pigments (chlorophyll a, allophycocyanin, and phycocyanin). In the first experiment, 18 mg of 14C-heme (11.7 dpm/mole) was added to 4500 ml of growth medium containing 225 ml inoculum of fully grown A. nidulans cells. After 6 days, growth was complete and the phyco- cyanobilin, cleaved from apoprotein and purified as the dimethylester, revealed little or no significant radioactivity. In a second experiment, addition of radiolabeled heme was delayed until cultures were grown to half their final cell density. The remaining growth occurred within 24 hr, after which phycocyanobilin dimethylester, obtained as before, again contained no significant radioactivity. In a third experiment, cells of A. nidulans in exponential growth were resuspended in nitrogen depleted medium, resulting in loss of phycocyanin and allo- phycocyanin. After nitrogen bleaching (36 hr), nitrogen and radiolabeled heme were added to the medium. When resynthesis of phycobiliproteins was completed, phycocyanobilin was again isolated as the dimethylester but it contained no radioactivity. The failure of A. nidulans cells to convert UC heme to phycocyanobilin is probably due to the inability of these cells to take up heme from the suspending medium, although the possi- bility that heme might not be a precursor of phycocyanobilin in this organism cannot be dis- counted from the experimental results. 500 ABSTRACTS FROM M.B.L. GENERAL MEETINGS Supported by NATO grant 1721, XSF PCM79 06638, PCM78 20535, and NIH GM 22822. Relationships between photosystem II and phycobilisomes in red algae and cvano- bacteria. T. A. KURSAR, I). MATZERALL. AND R. S. ALBERTE. The phycobiliproteins of the cyanobacteria and red algae are aggregated into complexes termed phycobilisomes (PBS). The PBS are localized on photosynthetic membranes and efficiently transfer excitation energy to reaction center II (RC-II). The ratio of chlorophyll (Chi) to RC-II was measured by the method of Emerson and Arnold, and the Chi -O"1- flash number'1 was divided by four, the number of electrons necessary to make O2. The values obtained were 330 (±25) Chl/RC-II for Anacystis nidulans, a cyanobacterium, and 320 (±25) Chl/RC-II for Neoagardhiella bailcyi, a macrophytic red alga collected at Woods Hole. Fil- tration of algal homogenates through Celite removed essentially all of the Chi. The bili- protein concentrations in the filtrates were determined from specific extinction coefficients. The PBS of Anacystis and Neoagardhiella were shown to contain about 90% of the cellular biliprotein. If we assume that 10-15% of the PBS is composed of "uncolored proteins", the PBS-associated protein per RC-II was 3.0 and 3.7 X 10" daltons for Anacystis and Neoagardhi- ella, respectively. Based on published chromophore compositions of biliproteins, the total bilin chromophores per RC-II was 190 (±4) for Anacystis and 520 (±50) for Neoagardhiella. Molecular weights of isolated PBS were determined from S values obtained from sedimenta- tion velocity measurements and diffusion coefficients measured by quasi-elastic light scattering (Paul Missel, MIT). Anacystis PBS were determined to be 3.6x10" daltons and Neo- agardhiella PBS to be 12 X 10" daltons. Another PBS molecular weight was determined assuming a six-rod model for PBS structure, using literature data on tripartite particles (800 kd PBS building blocks from red algae), and using spectroscopically determined bili- protein compositions of Anacystis and Neoagardhiella PBS. The results yielded molecular weights of 5.1 X 10" and 14.1 (±2.0) daltons for Anacystis and Neoagardhiella, respectively. Consequently, the ratio of PBS per RC-II would be 1.45 (±0.25) for Anacystis and 3.4 (±0.5) for Neoagardhiella. It is possible that the RC-IIs in red algae are clustered into groups of three or four such that several RC-IIs share a single PBS. Research supported by NSF grants PCM 77-09102, PCM 78-10535, and PCM 79-06638, and NIH grant GM -23944. Photosynthetic light adaptation features of Zostera marina L. (eelgrass}. L. MAZ- ZELLA, D. MAUZERALL, AND R. S. ALBERTE. The light adaptation features of photosynthesis in Zostera marina L. were studied in plants collected in a Woods Hole tidal bed. Our previous studies on Zostera showed a gradient in leaf pigmentation and photosynthetic activity along the leaf axes. Therefore, we sought to characterize in the laboratory the photosynthetic characteristics of Zostera leaves developed in situ under a range of light environments. Photon flux densities measured at the collection site (July, 1980) showed a gradient from the level of the tips of submerged plants to the bottom (2 m) (830-250 and 400-70 /tE-m~2-sec"1 on cloudless and cloudy days, respec- tively). The maximal photosynthetic rates per unit area, measured as COa-dependent Oa evolution, increased almost two-fold (3.0-5.6 /uMol Oa'dnr'-min"1) from the leaf bases to the tips. However, when rates were expressed on a chlorophyll (Chi) basis, leaf bases had higher rates than tips (4.0 /xMol O^ compared with 3.0 /j.Mo\ O^-mg Chl^-min1). The dif- ferences in Chi and area rates can be attributed to the fact that there is a 40% increase in Chi per dm" in leaf tips compared with bases. Light saturation of photosynthesis occurred 100 and 150 yuE-nr'-sec'1 for all leaf types examined. Initial slopes of the photosynthetic light saturation curves, defined as photosynthetic efficiency, increased from leaf bases to tips when rates were expressed per unit area. Photosynthetic activities of young, old epiphytized, old non-epiphytized, and old leaves with epiphytes removed were compared along the leaf axes. In almost all cases photosynthetic activity increased from the bases to the tips. It is clear that epiphytes contribute a significant portion of the total leaf photosynthesis per unit area. Photosynthesis measured as 14CO--fixation rates paralleled the O^ evolution rates described above. Such studies provide a molecular and biochemical framework for further investigations aimed at defining the adaptive physiology of a highly productive coastal marine species. Research supported by NSF grants PCM 79-06638 and PCM 78-10535. PLANT PIGMENTS AND PHOTOSYNTHESIS 501 Photosynthetic activity oj jtint/al injected and noninjcctcd Laminaria saccharina Lam. SCOTT SCIIATZ AND THOMAS A. KURSAR. The sporophytic phase of the brown alga, Laminaria saccharina Lam., becomes susceptible to infection by the ascomycete, Phycomelaina latninariac (Rostr.) Kohlrn. after 1 year of growth. Coincident with reduced rates of growth, diseased plants display a more limited photosynthetic capacity under saturating light conditions when compared with healthy plants. The base of the healthy blade tissue has a CO--dependent oxygen evolution rate of 1.1 /*moles O2-mg Chr'-min"1 while the base of the blade of infected plants have rates of 0.44 /imoles Oa-mg Chr'-miir1. The rate of O- evolution in stipes, the site of infection, was 0.08 ^moles O2-mg Chl-'-min'1 in healthy stipes and 0.01 yurnoles O^-mg Chl^-min"1 in infected stipes. The rate of respiration was found to exceed the net photosynthetic rate of infected stipes and may be attributable to increased rates of respiration due to the presence of the fungal pathogen. The functional organization of the photosynthetic apparatus, the photosynthetic unit (PSU), can be defined as the minimum number of chlorophyll molecules necessary to drive oxygen evolution. Interestingly, the PSU size in the base of blade tissue is similar in both healthy and infected plants (2100 ± 200 Chi a + c). The PSU size of the tip of the healthy blade was much greater (4500 Chi « + <•)• These data, taken together with previous PSU data on mid-blade tissues of healthy plants, indicate that as the blade of healthy plants ages there is a loss in the number of PSUs. It is possible that the decrease in photosynthetic rate in infected plants may be related to the numbers of functional PSUs or due to decreased activity of enzymes involved in the dark reactions of photosynthesis. It was not possible to determine whether the process of "whole plant senescence" led to increased susceptibility to infection, or whether fungal infection was the causal factor in decreased photosynthetic capacity. Research supported by NSF grant PCM-79-06638 and PCM-780535. Chemical conversion oj a chlorophyll-^, derivative to a bile pigment. K. M. SMITH, S. B. BROWN, AND R. F. TROXLKR. Little is known about the mechanisms of chlorophyll degradation in spite of the fact that millions of tons are removed from the biosphere annually during leaf senescence and grain ripening. In contrast, heme degradation to bile pigments is reasonably well understood and the origin of the bile pigment lactam oxygen atoms has been studied using lbO labeling of molecular oxygen in a variety of systems. Bile pigments have never been observed during photochemical chlorophyll breakdown in vivo or in vitro. Smith and coworkers have recently developed a chemical model reaction in which chlorophyll-a, after chemical derivatization, reacts with molecular oxygen to produce a bile pigment. In the present work, we have developed conditions whereby the reaction can be carried out in a closed system appropriate for 1SO labeling techniques used in heme de- grading systems. Zn( II )methyl-5-mesotrifluoroacetoxypyropheophorbide-(/ was prepared by treating methyl- pyropheophorbide-a (derived from chlorophyll-a) with zinc acetate and thallium trifluoroacetate. We dissolved 5 mg of Zn(II)methyl-5-mesotrifluoroacetoxypyropheophorbide-a in chloroform (5 ml) and added basic alumina (2-3 g). After shaking for 15 min in air, the reaction mix- ture was passed through a sintered glass funnel, the nitrate Was evaporated to dryness, and the residue was applied to a silica gel G plate. The green mixture rapidly separated into two green bands, the upper (minor) corresponding to starting material and the lower (major) corresponding to ring-opened product. This quickly turned blue on the plate, presumably due to demetallation of the zinc complex of the bile pigment. The product was eluted with chloroform and evaporated to dryness yielding product as a residue. An identical result was obtained when the reaction was run in the absence of light indicating that it is not a photo- oxidation. Using this method, it should be possible to carry out the reaction in a closed system under lh'lhO-. Mass spectrometry of the ring-opened product will reveal whether the lactam oxygen atoms are derived from the same oxygen molecule or from different oxygen molecules as in heme catabolism. Supported by NATO grant 1721, NSF PCM79 06638, NIH GM 22822, NIH AM 25714, and a grant from the Nuffield Foundation. 502 ABSTRACTS FROM M.B.I.. GENERAL MEETINGS Synthesis of chlorophyll-^ jroin 8-aminoleviilinate in ilark-yrown Cyanidium cal- darium cells. R. F. TKOXLKK AND S. B. BRO\\ x. Autotrophic wild-type C. caldarintn cells synthesize chlorophyll-o. allophycocyanin and phycocyanin but lack these photosynthetic pigments when grown heterotrophically in the dark. It was shown previously that dark-grown cells administered 5-aminolevulinate (ALA) in the dark excrete porphyrins and phycocyanohilin (the chromophore of allophycocyanin and phyco- cyanin) into the suspending medium. The cell pellet displayed a greenish-brown color, in contrast to the yellow color of non-ALA treated controls. The present work examined whether cells given ALA in the dark synthesixe chlorophyll <;. Cells incubated with ALA were suspended in 50 mM Tris-HCl, pH 8, disrupted in a French pressure cell (20,000 p.s.i.) and centrifuged at 30,000 X g for 20 min. The greenish- brown pellet was suspended in 6 volumes of petroleum ether : acetone (5:1, v/v) to which saturated NaCl (1/4 volume), water (1/4 volume) and diethyl ether (1 volume) was added. After shaking, the green ether phase was dried with Na-SOi, reduced to dryness, and the residue was applied as a band to silica gel G plates which were developed with petroleum ether : acetone (7:3, v/v). A green band was observed at Rf = 0.66 which chromatographed with authentic chlorophyll a prepared as above from autotrophic cells. The green band was eluted with 80% acetone and gave a spectrum identical to chlorophyll a with absorption maxima at 663, 618, 582, 536, and 433 nm. The pellets from disrupted, ALA-treated cells were extracted with 0.1% sodium dodecyl sulfate and examined electrophoretically on 7.5% polyacrylamide gels. Complex I (P7oo chlorophyll a protein) extracted from thylakoid membranes of autotrophic cells was absent in membrane fragments of cells incubated with ALA. Measurements on autotrophic cells indi- cated a photosynthetic rate of approx. 1 ^1 O-'min'^mg chlorophyll'1 whereas O2 evolution from cells incubated with ALA could not be detected. The data indicated the C. caldarium cells synthesize chlorophyll mf>. Physiol.) C. All abbreviated components must be followed by a period, whole word components must not (i.e. J. Cancer Res.) D. Space between all components (e.g. J. Cell. Comp. Physiol., not J.Cell.Comp. Physiol.) E. Unusual words in journal titles should be spelled out in full, rather than employing new abbreviations invented by the author. For example, use Rit Visindajjelags Islendinga without abbreviation. F. All single word journal titles in full (e.g. Veliger, Ecology, Brain). G. The order of abbreviated components should be the same as the word order of the com- plete title (i.e. Proc. and Trans, placed where they appear, not transposed as in some BIOLOGICAL ABSTRACTS listings). H. A few well-known international journals in their preferred forms rather than WORLD LIST or USASl usage (e.g. Nature, Science, Evolution NOT Nature, Land.; Science, N. Y.; Evolution, Lancaster, Pa.) 6. Reprints, charges. The Biological Bulletin has no page charges. However, authors will be requested to pay printing charges in excess of $70 for reproducing tabular material, mathemati- cal formulae, and figures. Reprints may be ordered at time of publication and normally will be delivered about two to three months after the issue date. CONTENTS ANDERSON, ROBERT S. Hemolysins and hemagglutinins in the coelomic fluid of a poly- chaete annelid, Glycera dibranchiata .V 259 HADLOCK, ROBIN P. Alarm response of the intertidal snail Littorina littorea (L.) to predation by the crab Carcinus maenas (L.) 269 HOUK, MARGARET S. AND RALPH T. HINEGARDNER The formation and early differentiation of sea urchin gonads 280 KANESHIRO, EDNA S. AND RICHARD D. KARP The ultrastructure of coelomocytes of the sea star Dermasterias imbricata 295 MARCUS, NANCY H. Photoperiodic control of diapause in the marine calanoid copepod Labidocera aestiva 311 MATSUMOTO, GEN AND JUNICHI SHIMADA Further improvement upon maintenance of adult squid (Doryteuthis bleekeri) in a small circular and closed-system aquarium tank 319 MCCORKLE, SUSAN AND THOMAS H. DlETZ Sodium transport in the freshwater Asiatic clam Corbicula fluminea . 325 MERCANDO, NEIL A. AND CHARLES F. LYTLE Specificity in the association between Hydractinia echinata and sympatric species of hermit crabs 337 MILLER, CHARLES B., DAVID M. NELSON, ROBERT R. L. GUILLARD, AND BONNIE L. WOODWARD Effects of media with low silicic acid concentrations on tooth forma- tion in Acartia tonsa Dana (Copepoda, Calanoida) 349 MORAN, WILLIAM M. AND RICHARD E. TULLIS Ion and water balance of the hypo- and hyperosmotically stressed chiton Mopalia muscosa 364 OHTSU, KOHZOH Electrical activities in the subtentacular region of the anthomedusan Spirocodon saltatrix (Tilesius) 376 RAHAT, M. AND ORIT ADAR Effect of symbiotic zooxanthellae and temperature on budding and strobilation in Cassiopeia andromeda (Eschscholz) 394 SULKIN, S. D., W. VAN HUEKELEM, P. KELLY, AND L. VAN HUEKELEM The behavioral basis of larval recruitment in the crab Callinectes sapidus Rathbun : A laboratory investigation of ontogenetic changes in geotaxis and barokinesis 402 TURNER, KATHERINE AND TIMOTHY A. LYERLA Electrophoretic variation in sympatric mud crabs from North Inlet, South' Carolina 418 YOUNG, CRAIG M. AND LEE F. BRAITHWAITE Orientation and current-induced flow in the stalked ascidian Styela montereyensis 428 ABSTRACTS OF PAPERS PRESENTED AT THE MARINE BIOLOGICAL LABORATORY. 441 Volume 159 Number 3 i } IB 6 1931 ; I L~ THF BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board DANIEL L. ALKON, National Institutes of Health MEREDITH L. JONES, Smithsonian Institution and Marine Biological Laboratory GEORGE O. MACKIE, University of Victoria FREDERICK B. BANG, Johns Hopkins University IT™ r. n/r T>^ T-* J°EL L- ROSENBAUM, Yale University EDWARD M. BERGER, Dartmouth College STEPHEN C. BROWN, State University of New York HOWARD A. SCHNEIDERMAN, Monsanto Company at Albany F. JOHN VERNBERG, University of HARLYN O. HALVORSON, Brandeis University South Carolina J. B. JENNINGS, University of Leeds E. O. WILSON, Harvard University Managing Editor: CHARLES B. METZ, University of Miami DECEMBER, 1980 Printed and Issued by LANCASTER PRESS, Inc. PRINCE & LEMON STS. LANCASTER, PA. THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is published six times a year by the Marine Biological Laboratory, MBL Street, Woods Hole, Massachusetts 02543. Subscriptions and similar matter should be addressed to THE BIOLOGICAL BULLETIN, Marine Biological Laboratory, Woods Hole, Massachusetts. Single numbers, $10.00. Subscription per volume (three issues), $27.00 (this is $54.00 per year for six issues). Communications relative to manuscripts should be sent to Dr. Charles B. Metz, Editor; or Susan Schwartz, assistant editor, at the Marine Biological Laboratory, Woods Hole, Massachusetts 02543 between June 1 and September 1, and at the Institute For Molecular and Cellular Evolu- tion, University of Miami, 521 Anastasia, Coral Gables, Florida 33134 during the remainder of the rear. Copyright © 1980, by the Marine Biological Laboratory Second-class postage paid at Woods Hole, Mass., and additional mailing offices. ISSN 0006-3185 INSTRUCTIONS TO AUTHORS THE BIOLOGICAL BULLETIN accepts original research reports of intermediate length on a variety of subjects of biological interest. In general, these papers are either of particular interest to workers at the Marine Biological Laboratory, or of outstanding general significance to a large number of biologists throughout the world. Normally, review papers (except for a limited number of so- licited review papers which may be accepted after formal refereeing) , very short papers (less than five printed pages), preliminary notes, and papers which describe only a new technique or method without presenting substantial quantities of data resulting from the use of the new method can- not be accepted for publication. A paper will usually appear within four months of the date of its acceptance. The Editorial Board requests that manuscripts conform to the requirements set below; those manuscripts which do not conform will be returned to authors for correction before review. 1. Manuscripts. Manuscripts, including figures, should be submitted in triplicate. (Xerox copies of photographs are not acceptable for review purposes.) The original manuscript must be typed in double spacing (including figure legends, footnotes, bibliography, etc.) on one side of 16- or 20-lb. bond paper, 8^ by 11 inches. Manuscripts should be proofread carefully and errors corrected legibly in black ink. Pages should be numbered consecutively. Margins on all sides should be at least 1 inch (2.5 cm). Manuscripts should conform to the Council of Biology Editors Style Manual, 4th Edition (Council of Biology Editors, 1978). Unusual abbreviations should be kept to a minimum and should be spelled out on first reference as well as defined in a footnote on the title page. Manuscripts should be divided into the following components: Title page, Ab- stract (of no more than 200 words), Introduction, Materials and Methods, Results, Discussion, Acknowledgments, Literature Cited, Tables, and Figure Legends. In addition, authors should supply a list of words and phrases under which the article should be indexed. 2. Figures. Figures should be no larger than 83 by 1 1 inches. The dimensions of the printed page, 5 by 7f inches, should be kept in mind in preparing figures for publication. We recommend that figures be about lj times the linear dimensions of the final printing desired, and that the ratio of the largest to the smallest letter or number and of the thickest to the thinnest line not exceed 1:1.5. Explanatory matter generally should be included in legends, although axes should always be identified on the illustration itself. Figures should be prepared for reproduction as either line cuts or halftones. Figures to be reproduced as line cuts should be unmounted glossy photographic reproductions or drawn in black ink on white paper, good-quality tracing cloth or plastic, or blue- lined coordinate paper. Those to be reproduced as halftones should be mounted on board, with both designating numbers or letters and scale bars affixed directly to the figures. All figures should be numbered in consecutive order, with no distinction between text and plate figures. The author's name and an arrow indicating orientation should appear on the reverse side of all figures. 3. Tables, footnotes, figure legends, etc. Authors should follow the style in a recent issue of The Biological Bulletin in preparing table headings, figure legends, and the like. Because of the Continued on Cover Three ERRATA THE BIOLOGICAL BULLETIN, Volume 159, Number 1, Pages 166, 167, 168. Equations were printed incorrectly in the article by Keith Nelson, Dennis Hedgecock, Will Borgeson, Eric Johnson, Richard Daggett, and Diane Aronstein, "Density-dependent growth inhibition in lobsters, Hoiiiarns (Decapoda, Nephro- pidae" (1980, Biol. Bull., 159: 162-176). The equations should read as follows: Page 166: 1=1 - 1 Vk Page 167: aj(/) »= A-^-lerkt (5) Page 168: di = d0e~li, i = 1,2... 503 Vol. 159, No. 3 December 1980 THE BIOLOGICAL BULLETIN PUBLISHED BY THK MARINE BIOLOGICAL LABORATORY Reference: Biol. Bull., 159: 505-560. (December, 1980) CELLULAR ANALYSIS OF A GASTROPOD (HERMISSENDA CRASSICORNIS} MODEL OF ASSOCIATIVE LEARNING DANIEL L. ALKON Section on Neural Systems, Laboratory of Biophysics, IRP, NINCDS, National Institutes of Health, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 Abbreviations : Excitatory post-synaptic potential, EPSP ; inhibitory post-synaptic po- tential, IPSP; long-lasting depolarization, LLD ; tetramethyl ammonium, TMA ; ethylene- glycol tetraacetic acid, EGTA ; artificial sea water, ASW; cerebropleural ganglion, CPG ; resting potential, R.P. ABSTRACT Intuitive formulations concerning the nature of cellular substrates for associa- tive learning are considered. Invertebrate models of vertebrate learning are dis- cussed. A cellular analysis of a long-term associative behavioral change is reviewed for the nudibranch mollusc Hermissenda crassicornis. Convergence of neuronal pathways which mediate natural sensory stimuli used for associative training have been determined. Primary biophysical changes within single identified neurons are analyzed for their causal role in learning behavior. A sequence of neural processes is proposed to underlie acquisition and retention of an associative behavioral change of Hermissenda. Paired sensory stimuli enhance voltage-dependent currents by means of converging sensory pathways. Repeated pairings cause progressive mem- brane depolarization, which in turn causes long-lasting conductance changes. Bio- chemical correlates of these phenomena are presented. INTRODUCTION The defining features of classical conditioning, which Pavlov produced in his experiments on dogs (Pavlov, 1927), have also been observed in other examples of associative learning. These features (Miller, 1967; Gormezano and Moore, 1969) include (1) the necessity for temporal association of distinct sensory stimuli, (2) improvement of the learned behavior with practice, also known as acquisition. (3) long duration, and (4) stimulus specificity, i.e. responses to the conditioned stimulus, but not to other stimuli, are modified. 505 Copyright © 1980, by the Marine Biological Laboratory Library of Congress Card No. A38-518 (ISSN 0006-3185) 506 DANIEL L. ALKON Although we can readily conceive of human emotional responses being condi- tioned by associated trauma, how do we conceptualize the nature of other human learning, far more extensive, but less emotionally charged, in our experience? As humans we learn to remember and recognize another person as an infinitely rich perceptual experience evoked by a name : a visual space or matrix filled with specific ordering of colored points, an auditory space filled with ordering of sound waves by amplitude and frequency, and an emotional space composed of a subtle balance of feelings. What meaning does the simple conditioning of a dog have for these memories of which human consciousness is constructed? Even less obvious is the relevance of behavioral changes of gastropod molluscs. What can a snail be for other snails except a source of a few specific stimuli which elicit stereotypic ag- gressive, escape, or copulatory behavioral patterns ? According to a reductionist's approach to human learning, implicit in Pavlov's experiments, complex memories might be considered as generated from a host of elemental memories or associations. A view contrary to the reductionist's would hold that human learning is so unique that hosts of elementary associations are not sufficient explanation. Neither view is clearly favored by our current knowledge of animal nervous systems. Although there are many characteristic differences be- tween neural elements of humans as compared to those of other species, none of these differences themselves present obvious advantages for long-term information storage which could be important for learning. This lack of apparent functional uniqueness of human neural cells for long-term change allows at least for the pos- sibility that elemental associative learning processes can be accomplished by ele- mentary neural systems, and that multiplication of these neural systems in higher species provides for multiplication, and thereby increased complexity, of associations rather than for an entirely new or unique learning process. With the advent of electrophysiologic recording techniques, vertebrate studies have allowed us to conclude that neuronal impulse activity in certain cortical and subcortical structures changes specifically as a result of training procedures. Most of these studies were conducted with electrodes outside of neurons (John and Killam, 1959; Adey et al, 1960; Jasper ct al., 1960; Doty, 1961 ; Buchwald et al, 1966; Livingston, 1966; John, 1967; Cohen, 1969; Olds and Hiramo, 1969; Olds et al., 1972; Evarts, 1973; John et al., 1973; Cohen and Macdonald, 1976; Norman et al., 1977; Berger and Thompson, 1978a, b ; Toledo-Morrell ct al., 1979). These studies monitored for the most part impulse frequency changes as a function of training. More recently. Woody and his colleagues have utilized intracellular electrodes to measure changes of excitability as a function of a conditioning proce- dure (Woody and Engel, 1972; Woody and Yaronsky, 1972; Woody and Black- Cleworth, 1973; Woody et al., 1976). What these studies have not revealed are the primary cellular and subcellular changes which cause learned behavior. For example, do the primary changes occur at synapses, axonal branch points, or neural cell bodies? Are there long-lasting biochemical modifications which control mem- brane conductances? Are there morphological changes? Do neurons grow new processes? To determine primary neural mechanisms for learned behavior we are limited by existing technology to using a behavioral model in an animal where precise determination of networks is possible. Among so-called "simple-system" prepara- tions which have been investigated as potential models of learning are the crayfish (Bullock and Quarton, 1966; Krasne and Woodsmall, 1969; Zucker et al., 1971 ; Zucker, 1972a, b), the locust (Horridge, 1959; 1962a, b; Hoyle, 1964, 1965; GASTROPOD MODEL OF LEARNING 507 FIGURE 1. A three-dimensional representation (upper sketch) of the metathoracic gang- lion of a young adult male cockroach. Note that all the cells except those associated with nerve 2 send their axons out from the ipsilateral nerve trunk. The lower outline indicates some of the cell bodies and connectives investigated (Cohen and Jacklet, 1967). Note that all the cells except those associated with nerve 2 send their axons out from the ipsilateral nerve trunk. Horridge, 1965; Runion and Usherwood, 1968; Burrows, 1973; Burrows and Hoyle, 1973 ; Hoyle and Burrows, 1973a, 1>; Burrows and Horridge, 1974; Pearson and Goodman, 1979), and the cockroach (Eisenstein and Cohen, 1965; Wilson, 1965, 1966; Cohen and Jacklet, 1967; Aranda and Luco, 1969; Disterhoft et al, 1968; Disterhoft et al., 1971; Pearson, 1972). Thus far the nervous systems of these preparations (Fig. 1) have proven rather complex for a comprehensive neural network analysis. To achieve such an analysis, other investigators turned to gastropod molluscs for learning models. A non-associative behavioral change, habituation, was produced with gastropods many years ago (Humphrey, 1930; Harris, 1943; Thorpe, 1963). Hughes and Tauc (1963) produced a decreased amplitude of compound excitatory post-synaptic potentials (EPSPs) in an Aplysia neuron by repeated mechanical stimulation applied to the animal's skin. Bruner and Tauc produced a similar decrease in the giant cell of the left pleural ganglion by repeated electrical stimulation of afferent fibers (1964). Subsequently, Kandel. Peretz, Tauc, and their colleagues established more clearly the relationship of de- creases in unitary EPSPs to habituation of Aplysia behavior (Bruner and Tauc, 1964; Kandel and Tauc, 1964; Bruner and Tauc, 1965, 1966a; Kupferman ct a/.. 1970; Peretz, 1970; Carew ct al., 1972; Carew and Kandel, 1973; Eisenstein and Peretz, 1973; Peretz and Howieson, 1973; Peretz and iMoller, 1974). These and other workers also analyzed behavioral and cellular aspects of dishabituation, a non-associative stimulus-evoked restoration of a previously habituated response (Kandel and Tauc, 1964; 1965a, b; Tauc and Epstein, 1967; von Baumgarten and Djahanparwar, 1967; von Baumgarten and Hukuhara, 1969; von Baumgarten, 508 DANIEL L. ALKON 1970; Epstein and Tauc, 1970; Peretz, 1970; Castellucci et al, 1970; Carew et al, 1971). Strumwasser, Jacklet, and others have used Aplysia to study neuronal mechanisms of circadian rhythms of behavior (Strumwasser, 1965; Strumwasser ctal., 1969; Jacklet, 1977). ' During the last 10 years other investigators have turned to a number of gastro- pod molluscs for examples of associative learning behavior mediated by analyzable neural pathways. With the exception of the nudibranch Hermissenda crassicornis (Fig. 2; Alkon, 1974a, b) these molluscs experience chemical (and/or touch) stim- ulation of the tentacle or oral veil during the training paradigm. In considering possible preparations it seemed reasonable to me that to under- stand quantitatively how a nervous system causes associative learning behavior, it would be advantageous to use natural stimuli transduced by sensory receptors during an organism's normal life cycle. The effects of these stimuli might then be followed step by step from the receptors to interneurons and ultimately to motor neurons and muscular activity. Gastropods were attractive because of their relatively small number of neurons yet fairly elaborate behavioral capability. It also seemed reasonable that with such a small number of neurons, the animal was more likely to be capable of intermodal than of intramodal sensory associations. Natural stimuli for the training regimen seemed important to me for another reason : the patterns of natural stimuli encountered by the animal in its environment should determine the range of possible learned behaviors adaptive within a given biological FIGURE 2. The nudibranch mollusc Hermissenda crassicornis. Note the small black spot (the right eye) at the base of the lower rhinophore. Length of the animal is ~ 4 cm. GASTROPOD MODEL OF LEARNING 5o<) niche. Brief dissection of a transparent dorsal integument reveals the circum- esophageal nervous system of Hermissenda. including the receptor organs (Fig. 3A) of three distinct sensory pathways: visual, statocyst, and chemosensory. The easy accessibility of the eyes, statocysts, and central ganglia to electrophysiologic re- cording techniques, and the close proximity of the eyes to the statocysts, encouraged consideration of Hermissenda as a model system. During the light period of a 6.5 hr light : dark cycle, light elicits a positive be- havioral response from Hcnnisscnda (Fig. 4A). The animal increases its general level of motor activity and moves toward a light source ( Alkon, 1974 ; Lederhendler et al., 1980). In response to rotation the animal moves toward the center of rota- tion, thereby reducing the centrifugal force it experiences. Similarly, the animal tends to move up a vertical gradient, i.e., away from the direction of the earth's gravitational force (Alkon, 1973, 1976). In response to vigorous rotation the animal also contracts the musculature of its body wall and "clings" to surfaces (Fig. 4B). This "clinging" minimizes the turbulence the animal would experience were it free of any surfaces. In all then, rotation can be regarded as a negative or aversive stimulus with effects the animal reduces by its responses. When light that elicits a positive response is repeatedly paired with rotation, an aversive stimulus, the animal's subsequent reaction to light is markedly less positive (Alkon, 1974), i.e., the animal no longer approaches a light source (Fig. 5). Crow and Alkon (1978) showed that this behavioral change is truly associative (that is, randomized light and rotation did not produce the same effect), persists for at least several days after training, and increases with practice (Fig. 6A, B). Specificity of this training effect was suggested by the fact that trained animals did not show changes in response to food. Specificity was further indicated by the observation that animal's response to light but not darkness is modified by stimulus pairing (Crow and Offenbach, 1979; Farley and Alkon, in preparation). Stimulus pairing also did not modify the animals' negative geotactic responses as demonstrated with movement on a vertical gradient (Farley and Alkon, in preparation). Thus, this behavioral change of Hermissenda shows defining features of, and can serve as a model for, associative learning as described for vertebrate species. What is the possible physiologic significance of this associative behavioral change? Hermissenda, like other intertidal molluscs, can be seen to approach the water's sur- face during daylight hours and is found at lower depths at night (Harrigan and Lederhendler, in preparation). This behavior can be explained at least in part by the animal's positive phototaxis during the day, as demonstrated by Lederhendler et al. (1980), and negative phototaxis at night. This cyclic movement has adaptive value in that hydroids, microorganisms on which Hcnnisscnda feeds, are concen- trated in illuminated areas. During periods of increased oceanic turbulence, such as might occur in a storm, any Hermissenda approaching the water's surface during daylight would be shaken vigorously — i.e., they would be shaken in association with light. This shaking was approximated by the rotation stimulus conditions originally paired (Alkon, 1974) with light. Hermissenda, then, as it approached the water's surface would experience more light in association with turbulence. It is possible that repetition of this association in the field, like the association of light and rotation in the labora- tory, would result in reduced movement of the animals toward the water's surface, thereby minimizing the turbulence subsequently experienced. Persistence of re- 510 DANIEL L. ALKOX CON ST 50 (jm FIGURE 3A. (Top) Hertnissenda circumesophageal nervous system. The photomicro- graph is of a living preparation. The statocysts lie between the two central cerebroplural GASTROPOD MODEL OF LEARNING 511 1 cm FIGURE 4A. (Left) Response of animals in experimental chamber to test light spot. A, B, C animals move toward and into light spot within 20 min. Spot (indicated by dashed lines) is obscured by stroboscopic flash in B, C. Calibration: 1 cm (Alkon, 1974). FIGURE 4B. (Right) Response of Hermisscnda to rotation. A, immediately before and B, immediately after 20 sec of rotation at 200 rpm. Immediately after and during rota- tion the entire animal is contracted (Alkon, 1974). ganglia and the two lateral pedal ganglia. Immediately above the statocysts are the optic ganglia (transparent) and above them are the eyes (darkly pigmented). Fed. 1 indicates a putative motor neuron. FIGURE 3B. (Bottom) Combined (isolated) sensory structures. E, eye; OG, optic ganglion; ST, statocyst ; and CON, capsule of connective tissue ( Heldman ct a!.. 1979). 512 DANIEL L. ALKON duced positive phototactic behavior over several days of a storm, then, could on bal- ance have survival value. For Pleurobranchaca (Fig. 7A). another gastropod used as a learning model, a food substance touched the tentacle, paired (or unpaired) with electric shock to the same tentacle and/or the head region (Mpitsos and Collins, 1975). A behav- ioral change was measured when the animals withdrew on subsequent touching with the food substance instead of the pre-training "bite and strike" feeding be- havior. This behavioral change was assessed by repeated touching of the tentacle until withdrawal or feeding was elicited. More animals withdrew from the food following pairing with electric shock than following an unpaired training regimen. Gelperin found that the terrestrial slug Lima* maxim us (Fig. 7B) will avoid food substances after paired presentations with COo poisoning (Gelperin, 1975). Within the last year, Walters ct al. (1979) paired chemical stimulus presentations with elec- tric shocks to the head region of Aplysia californica. Aplysia responses to test shocks across the tail were modified by paired but not by unpaired stimulus presentations. These unpaired stimulus presentations control only to a limited extent for non- associative behavioral changes of Pleurobranchaea and Aplysia. There is a fixed tem- poral relationship or "association" of stimuli even with these unpaired paradigms. A more unequivocal control (Rescorla, 1967) , in which food presentation has no fixed temporal relationship to shock, would clearly demonstrate the associative nature of this behavioral change. This is not entirely possible for the Pleurobranchaea, Limax, or Aplysia models, since only a few stimuli are presented over several hours during a training session. Range of variation in the temporal relationships among stimulus presentations adequate to approximate their randomization is therefore not realistic. In a more recent study, stimulus specificity for the be- 10 20 30 40 50 60 70 80 90 100 110 120 TIME (mm) 0 10 20 30 40 50 60 70 80 90 100 110 120 TIME (mm! FIGURE 5A. (Left) Kaplan-Meier (1958) curves for percent of animals which have reached a test light spot as a function of time. Squares: Animals tested after 10 min of dark- ness in experimental chamber. Triangles : Animals tested after 3 hr of 70-sec light intervals (see Methods). Circles: Animals tested after 3 hr of complete darkness. These data were obtained from photographic records. Animals were not removed when they reached the test spot nor was interaction between animals prevented. By the Gehan-Breslow test procedure the three groups are not significantly different (Alkon, 1974). FIGURE SB. (Right) Kaplan-Meier (1958) curves for percent of animals which have reached a test light spot as a function of time. These data were obtained by direct observation of animals entering the test light spot. Animals were removed when they reached the test spot and interaction between animals was prevented. Squares : Animals tested after 3 hr of light intervals. Circles : Animals tested after 3 hr of rotation intervals associated with light. The two groups are significantly different (P < 0.001) by the Gehan-Breslow test procedure (Alkon, 1974). GASTROPOD MODEL OF LEARNING 513 ACQUISITION .60 r RETENTION REACQUISITION g < .50 - % .40 - o O. -ir\ CO •ou UJ cr I '20 m .10 8 9 II FIGURE 6. (Top) Training and testing apparatus. The response latencies to enter a light spot projected onto the center of the turntable by an overhead illuminator were recorded automatically when the Hertnissenda moved toward the light source (direction of arrows) and interrupted the light between illuminator and photocells (arrowhead). (Inset) Hertnissenda was subjected to different behavioral treatments, consisting of light and rotation, while con- fined to the end of glass tubes filled with sea water. (Bottom) Median response ratios for acquisition, retention, and reacquisition of a long-term behavioral change in response to a light stimulus in Hennissenda. Solid circle, random rotation; open square, random light; open triangle, unpaired light and rotation; solid triangle, random light and rotation; solid square, nothing; and open circle, paired light and rotation. The response ratio in the form of 1-A/(A + B) compared the latency during the test (A) with the baseline response latency (B). Group data consist of two independent replications for all control groups and three independent replications for the experimental group (Crow and Alkon, 1978). havioral change of Plcurobranchaca was assessed following pairing of food sub- stances with electric shock; Davis ct al., 1980). A clear demonstration of stimulus specificity would add support to the hypothesis that a truly associative behavioral change was involved. Significant differences, however, between testing an avoid- ance response with squid extract and testing with Corynactis extract were not dem- onstrated following paired presentations of squid and electric shock. In fact, on 514 DANIEL L. ALKOX FIGURE 7A. (Top) Pleurobranchaca californica: a large notaspidian opisthobranch common to the subtidal waters of the California coast. It has lost its shell completely and bears a gill on the right side of the body under the large mantle cover, which is continuous with the expanded frontal veil. The veil also bears the tentacles (animal size = 15 cm long) (courtesy of Dr. Alan M. Kuzirian). FIGURE 7B. (Bottom) Limax maximus: common terrestrial slug (Pulmonata: Stylom- matophora) having secondary reduction and concealment of the shell and a pair of eyes borne on the tips of the tentacles (animal size = 4.5 cm) (courtesy of Dr. Alan M. Kuzirian). the third day following training the Corynactis avoidance response was significantly greater than controls, whereas the squid response was not. Also inconclusive were attempts to demonstrate differential avoidance responses after paired presentation of shock and one food substance together with unpaired presentation of a second food substance. Differential avoidance response behavior was observed on the second but not the first or third days. Some significant differences occurred for a "bite-strike" response. Most apparent, however, for the differential training pro- cedures, was the finding that the animals in general showed more withdrawal be- havior than for the single stimulus pairing procedure. This suggests that animals receiving more food stimulus presentations, regardless of their relation to shock, withdraw more. Since the animals were unfed for the entire 2 weeks of training and testing, the food used and ingested during training might significantly influence the animals' relative states of starvation. In fact, a general difficulty with those gastropod models using chemical (food substance) stimuli is the control of satiation effects. In none of these studies (using food paired with other stimuli) were the animals fed daily or were the GASTROPOD MODEL OF LEARNING 515 effects of inanition monitored. The electric shocks used for the Plcurobranchaca and Aplysia models and the COo poisoning for Liuia.v, when paired with food, might influence the animal's ability to feed and thus its state of satiation. The ani- mal's subsequent response to food might reflect how much it ingested specific food substances or extracts during and/or between stimulus presentations, rather than the learning of an association between a chemical stimulus and shock (Plcuro- branchaca and Aplysia) or food and COo poisoning (Limax}. States of inanition have recently been found to have important effects on gastropod responses to re- productive (Davis ct of., 1974), visual (Alkon ct al.. 1978), and touch (Lukowiak, 1979) stimuli, and to influence vertebrate neuronal responses to sensory stimuli (Nienhuis and Olds, 1978). A final difficulty with these models just mentioned concerns any possible meaningfulness for the animals in their natural environments. The use of electric shocks of sufficient magnitude to affect a substantial portion of the animal's body, or CO2 poisoning, which has diffuse toxic effects, obviates com- parisons to behavior in response to natural stimuli. Cellular analyses of associative learning Associative learning of any species should be explainable by the functions of cells. Theoretically, behavioral changes must result from neural changes and neural changes must depend on biophysical and biochemical processes capable of long- term transformation. Developmental changes of neural systems, such as those ob- served in cats by Hubel and Wiesel, Hirsh and Spinelli, Blakemore, and others (Wiesel and Hubel, 1965; Pribram et al., 1967; Blakemore et al., 1970; Hubel and Wiesel, 1970; Hirsh and Spinelli, 1971) are too slow to explain rapid, sometimes one-trial, learning. Growth of neural branches over any significant distance (> 1 /xm) requires hours if not days. Rapid learning, therefore, seems most reasonably accomplished via pre-existing neural circuits or systems. Again, this is an intuitive formulation which motivates a working hypothesis : genetically specified neural cir- cuits hold the potential of environmental specification by learning from sensory experience. Investigation of neural changes during learning must begin with the question, what are the genetically specified neural circuits ; what are the invariant aspects of gastropod neural systems ? The second question is : Given genetically determined limits of variance, what is the variance specific to and causal of learning behavior ? We must return, therefore, to what is known of the neural systems in these gas- tropod preparations: in general, very little. Although there are relatively few neurons, only some are identifiable from preparation to preparation, and many are not large and not superficially located on the ganglia, thereby making intracellular recordings less practical. In addition, considerable variability of neural structure has been demonstrated for gastropod and arthropod preparations (Lubbock, 1858; Burrows, 1973; Macagno ct al., 1973; Pitman ct al., 1973; Stretton and Kravitz. 1973; Altman and Tyrer, 1977; Goodman, 1978; Pearson and Goodman, 1979; Calabrese, in preparation). Because most synaptic interactions of gastropod neurons occur on axons at considerable distances from the cell bodies impaled by the intracellular electrodes, distortion of synaptic signals is likely and identification of direct or monosynaptic interactions by the necessary combination of electro- physiologic and morphologic criteria has only rarely been achieved. In fact, in many respects, vertebrate sensory pathways are better understood than those of gastropods. We know, for example beginning with the sensory receptors, successive neuronal groups which constitute the pathways for touch, 516 DANIEL L. ALKON FIGURE 8. Tentacle of Hermissenda stained for neural cells (Bodian technique). A. Longi- tudinal section of tentacle tip. Darkly stained areas represent loci of presumed chemoreceptors (200 X). B. Higher magnification (1000X) of presumed neural network immediately below receptors. Arrows indicate putative individual neurons, ~ 3 M'" in diameter. C. Tentacular response of Hermissenda to touch of food substance (squid). Animal is ~ 3 cm long. vibration, visual, auditory, and to some extent olfactory sensation (Shepherd, 1979). For neural systems such as the retina (Dowling and Boycott, 1965; Lasansky, 1971) and cerebellum (Llinas et al., 1972; Llinas and Hess, 1976) syn- GASTROPOD MODEL OF LEARNING 517 aptic interactions occur on or near cell bodies, and readily recognizable ultra- structural features can be used to demonstrate synaptic loci. Finally, because of the considerable redundance within these vertebrate neural systems, knowledge of neural circuitry for a unitary neural group (e.g. receptor-horizontal-bipolar- amacrine-ganglion cell aggregate) can be extended to some degree to a host of other comparable neural groups. Where, however, does chemosensation begin for Pleurobranchaea, Helix, or Aplysia (Stinnakre and Tauc, 1969; Fredman and Jahan-Parwar, 1980) ? Our experience with Hermissenda (Agersborg, 1922, 1925 ; Alkon et al., 1978) suggests the presence of receptors and neural plexi (Fig. 8) on the tentacular structures. Even after these chemoreceptor loci are unequivocally identified, intracellular recording, if past experience with chemoreceptor recording (Getchell, 1977) is any guide, may be difficult to accomplish. Higher order chemosensory neurons and their synaptic interactions have also yet to be comprehensively studied in gas- tropods, although this work has begun (Fig. 9; Gillette and Davis, 1977; Spray et al., 1980; Gillette et al., 1980). The neural pathways which mediate the effects of electric shocks used for training gastropods are even less well understood. Considering the preliminary nature of neural network analysis of chemoreception and sensation of electric shocks, it is not surprising that no primary neural correlate specific to training with these stimuli in a paired paradigm has been clearly demonstrated. Determining the primary neural changes specific to pairing might be especially difficult if, as has been suggested, neural networks with a large number of small neurons are situated on the tentacles themselves. It is for this primary neural change specific to pairing, however, that gastropods are particularly needed. As already mentioned, a number of workers have recorded neural changes following and accompanying associative learning by vertebrates (Fig. 10; Berger and Thomp- son, 1978a-c). The question that these workers may not be able to address with present technology is : What is the causal sequence, beginning with the training stimuli, that ultimately results in the behavioral change? Recording, from gastro- pod command or motor neurons, neural changes specific to pairing of chemosensory and shock stimuli, will not substantially enhance our knowledge of mechanisms of associative learning if these changes are simply a consequence of changes in synaptic input from other neurons that are themselves the site of primary or causal neural modification responsible for the learning behavior. L r3 withdr mn Lvwc JLJJUUJUIM FIGURE 9. Intracellular stimulation of the left ventral white cell (Lvwc) drives the feed- ing rhythm, recorded intracellularly from a withdrawal motor neuron (withdr mn) and extra- cellularly from roots 2 and 3 of the buccal ganglion (Rr2, Lr3) (Gillette ct al., 1980). 518 DANIEL L. ALKON vV NM • CAT A * NM CA1 FIGURE 10. Hippocampal multiple unit response recorded during NM classical conditioning. Upper trace: average NM response for one block of eight paired trials. Lower trace: hippo- campal multiple unit poststimulus histogram (15-msec time bins) for same block of eight paired trials. First cursor indicates tone onset ; second cursor indicates air puff onset. Total trace length equals 750 msec here and in all subsequent histograms. (A) Second block of con- ditioning training, Day 1. (B) Tenth block of conditioning training, Day 2. Vertical calibra- tion in A and B equals 25 unit counts per 15-msec time bin (Berger and Thompson, 1978). NEURAL ORGANIZATION IN HERMISSENDA To find primary neuronal changes responsible for learning, a comprehensive study of Hermissenda pathways was undertaken, beginning with intracellular record- ing from and histologic identification of photoreceptors in the eye and the hair cells in the statocyst, a primitive vestibular organ. Investigations then focused on second order neurons, other internetirons and, more recently, putative motorneurons. Sites of convergence between the visual and statocyst pathways and later the chemosensory pathway were also determined (Alkon and Fuortes, 1972; Alkon and Bak, 1973; Alkon, 1973a, b; Detwiler and Alkon, 1973; Alkon, 1974, 1975, 1976a, b ; Alkon ct «/., 1978 ; Alkon and Grossman, 1978 ; Akaike and Alkon, 1980). In general the interactions encountered between neural elements within these sensory pathways included both excitatory and inhibitory chemical synapses and electrical synapses. One example of an apparent non-synaptic interaction (Alkon and Grossman, 1978) was also found. Many of the chemical synaptic interactions appeared to be monosynaptic. In these cases discrete synaptic potentials followed with a brief latency, in a one-for-one fashion, impulses of the presynaptic cell (Figs. 11, 12). However, it is not unequivocally established that these are mono- synaptic interactions. GASTROPOD MODEL OF LEARNING 519 L [ I I I I I I I u I 10 — I I I I I 0 02 04 mv 60 50 10 30 20 10 0 1.0 SECOND 1 LL V FIGURE 11. (Above and middle) Responses recorded from soma and axon of same cell. The distance between the two electrodes was 80-100 fj,m. A) Responses to a step of depolarizing current through axonal electrode. B) Response to a step of current through the soma elec- trode. In either case the spike is larger and earlier in the axon. C) Two electrodes were inserted, respectively, in the soma and axon of one receptor and a third electrode \vas inserted in the soma of a second receptor. A step of depolarizing current through this third electrode evokes a spike in one receptor and a hyperpolarizing synaptic potential in the other. The synaptic potential is larger and has shorter time to peak in the axon. D) Responses to a flash of light. The generator potential is larger at the soma (loivcr trace) while the spikes are larger at the axon. Down-going bars in (A) and (B) indicate timing of current steps (Alkon and Fuortes, 1972). FIGURE 12. (Below) Simultaneously evoked EPSPs in a second order (cerebropleural ganglion) visual neuron and IPSPs recorded from an ipsilateral Type A photoreceptor. The second order neuron was hyperpolarized by injecting-1.0 nA D.C. Current. Voltage calibration: 20 mV. Arrows indicate IPSPs in the Type A cell simultaneous with small EPSPs in the second order visual neuron (Akaike and Alkon, 1980). In our investigation of Hermissenda's neural systems we used mainly electro- physiologic criteria (obtained by simultaneous intracellular recording from pre- and post-synaptic elements) to characterize neural interactions. Ultimately, how- ever, morphologic criteria (Figs. 13. 14) must be relied on for more complete de- scription of these interactions. Our observations, although incomplete, directed our efforts within these neural systems to a causal sequence of changes likely to be 520 DANIEL L. ALKOX FIGURE 13. Type B photoreceptor stained by iontophoretic injection of Procion yellow. Dark field photomicrographs demonstrate the fluorescent dye within the cell and its processes. A) cross section of cell soma and axon. B-D) details of axon and terminal branches. 480 X (Alkon, 1976). E) Nomarski optical section of HRP-labeled terminals of a photoreceptor in Hermissenda crassicornis. X 2400. Scale bar = 10 microns ; thick arrow = connecting process ; thin arrow = terminal swelling (courtesy of S. Senft). responsible for behavioral changes acquired during the associative learning paradigm. This was possible only after we had substantial confidence in the reproducibility from animal to animal of the observed interactions. For this it was necessary to GASTROPOD MODEL OF LEARNING 521 study repeatedly, in many animals, the same interactions, as will now be briefly described. Photoreceptors Each of the two Hcrmisscnda eyes contains five photoreceptors. Two of these, the Type A photoreceptors, are located in the ventral anterior part of the eye and the other three in the dorsal posterior region. Type A photoreceptors have spikes of 45 mV (based on intracellular recordings from > 150 cells), are silent in dark- ness, are less sensitive to light than Type B cells, and have little or no interaction with each other. (Of Type A cell pairs within the same eye from 12 different ani- mals, only one type A cell showed any response, a slight hyperpolarization, to the other Type A's firing at 20 Hz. elicited by a positive current pulse). In darkness. Type B cells, with spikes of 15 mV average amplitude (based on intracellular re- cordings from > 400 cells) develop depolarizing waves (3-8 mV) at infrequent and irregular intervals (0.5-30 sec). At least some of these waves are excitatory post-synaptic potentials (EPSPs) which arise from the ipsilateral optic ganglion (Tabata and Alkon, 1979). Type B photoreceptors are all mutually inhibitory (> 100 pairs in different preparations). In an early study (Alkon and Fuortes. 1972) of synaptic interactions between Type A and Type B photoreceptors (50 pairs), the following relations were found: reciprocal inhibition in 22 pairs; inhibi- tion of A upon B in 10 pairs ; inhibition of B upon A in 12 pairs, and no interactions in 6 pairs. Subsequent recording experience from A-B pairs has indicated that it is the lateral Type A photoreceptor which has reciprocal inhibitory interactions with the Type B cells in the same eye. Simultaneous intracellular recordings from three or more photoreceptors in the same eye indicated that there are, in each eye, two Type A and three Type B photoreceptors. This is consistent with an earlier histologic study (Fig. 15; Stensaas ct a!., 1969), which showed five axons exiting from the eye. The location of the microelectrode impaling class A cells suggests that they correspond to cells I and II, classified by histologic criteria (Stensaas ct a/., 1969) and that the class B cells correspond to cells III through V (Fig. 15). All five photoreceptors depolarize in response to flashes of light. With very dim flashes at least one and probably two Type B photoreceptors de- polarize and the two Type A photoreceptors hyperpolarize. A number of experiments indicated that the locus of the synaptic interactions between photoreceptors (as well as between photoreceptors and other neural ele- ments described below) is at their terminal branchings (Fig. 16) within the pleural ganglia. Iontophoresis of the fluorescent dye Procion Yellow into the photoreceptor somata within the eye reveals an axon which passes through the optic ganglion (Fig. 13) without branching and ends in a spray of fine branches immediately after entering the pleural ganglion. A high resolution Nomarski technique for visualizing these endings (Senft ct a/., in preparation) reveals detailed structures of these branches and their endings (Fig. 13E). Severing the optic nerve immediately before it enters the pleural ganglia abolishes the inhibitory inter- actions while preserving the response to light and sometimes the spikes. Finally, recording simultaneously in the axon (at its point of entry into the pleural ganglion) and soma of a photoreceptor shows a decrement of the inhibitory potential (elicited by impulses in another impaled photoreceptor) which is never less than the decre- ment of hyperpolarizing potentials produced by currents in the axon electrode (Fig. 11). This last experiment indicated that the inhibitory post-synaptic poten- tials (IPSPs) cannot arise proximal to the axonal recording site. A more re- cent investigation involved injection of horseradish peroxidase into two Type B 522 DANIEL L. ALKON Y ' 'IN yfefr < • * •'•::$- l^^'.-x:>-^^L, r.& 10 "J f *N . T. GASTROPOD MODEL OF LEAKXIXC 523 CELL IV CELL V FIGURE 15. Schematic representation of Hcnitisscnda eye showing the general configura- tion of the five photoreceptors (dark stipple), the rhabdomere (black) beneath the lens, and the course of axons which enter the optic nerve (black dot). Cell IV is situated above cell V, and their rhabdomeres are apposed (Stensaas ct al., 1969). photoreceptors within the same eye. Electron microscopic sections revealed a faint and a darkly stained cell. The axons of these two cells could then be traced by thick and thin sections to sites of apposition between faint and darkly stained endings —' 30 p.m from the optic nerves' entry point into the pleural ganglion (Fig. 14). These experiments, taken together, provide strong evidence that the photoreceptors synapse on each other at their terminal branchs within the pleural ganglion. Optic ganglion cells There are 14 cells in each optic ganglion (Stensaas ct al., 1969). The axons of these cells join those of the photoreceptors to form the optic nerve (10-20 /xm), which enters the lateral aspect of the pleural ganglion (Fig. 16). The axons of optic ganglion cells continue to course in a well-defined tract approximately 300 fj.m into the pleural ganglion. Optic ganglion cells (except for one, the largest) FIGURE 14A. (Numbered 7) Electron micrograph showing HRP-labeled photoreceptor soma (s) and cross-section of optic nerve. The pigment cells contain numerous pigment gran- ules (pg). The optic nerve consists of five axons (1-5) surrounded by a glial sheath (gs). A darkly labeled axon (a) corresponds to the labeled soma. A partially filled area indicated by arrow (bottom center) may be the result of some leakage from a previous unstable electrode penetration (Crow et al., 1979). A faintly stained axon, apparent in the optic nerve, was stained during this previous penetration. FIGURE 14B. (Numbered 8) Electron micrograph of an axon hillock of an HRP-labeled photoreceptor. The area contains numerous vesicles 60-80 nm in diameter, and granular ma- terial thought to be glycogen (gly). Outside the stained axon are collagen fibrils (col). A small process on the right of the micrograph invaginates the axon (Crow ct al., 1979). FIGURE 14C. (Numbered 9) Electron micrograph of a terminal process following in- jection of a single Type B photoreceptor. The terminal process (gray area, upper center, with asterisk) contains numerous clear round vesicles and abuts several clear processes. Scale bar: 0.5 Mm (Crow ct al., 1979). FIGURE 14D. (Numbered 10, 11) A moderately labeled terminal process (asterisk, upper center) surrounded by three darkly labeled processes following injection of HRP into two Type B photoreceptors. Scale bar: 0.5 /j.m (Crow et 100 pairs) but not Type A photoreceptors (demonstrated with 30 pairs) inhibit ipsilateral optic ganglion cells. This synaptic inhibition produced the hyperpolarizing response of optic ganglion cells in response to light (Fig. 17). Present evidence suggests that the Type B photoreceptor in- hibits the optic ganglion cell monosynaptically (Alkon, 1973). A subsequent analysis of photoreceptor-optic ganglion cell interactions has revealed many addi- tional details of Type B photoreceptor interactions with optic ganglion cells (Tabata and Alkon, 1979; Tabata and Alkon, in preparation). Of particular interest is a synaptic excitation of the Type B cells by at least one ipsilateral optic ganglion cell. The site of these EPSPs probably occurs within or near the optic ganglion itself (i.e., nearer than the site of IPSPs between photoreceptors) since they can often be transiently observed after the optic nerve has been severed near its point of entry into the pleural ganglion (Tabata and Alkon, in preparation). GASTROPOD MODEL OF LEARNING 525 B 0 1.0 0 10 20 30 40 FIGURE 17A. Depolarizing currents to Type B photoreceptor produce impulses associated with IPSPs in simultaneously impaled optic ganglion cell (Alkon, 1973). B. Responses of "C" cell to light flashes of increasing intensity. Note increased frequency of firing after the hyper- polarization for dim to moderate flashes. The duration of hyperpolarization with cessation of firing increases only for the brighter flashes (Alkon, 1973). Values beneath intracellular voltage recordings refer to attenuation with neutral density filters of quartz-iodide source with intensity on preparation of 105 ergs -cm"2 -sec'1 designated as "1". At least one type of optic ganglion cell (the "C" cell), recognized by its char- acteristic double spike, is inhibited by both ipsilateral and contralateral Type B photoreceptors (demonstrated by triple recordings in 12 different animals). Optic ganglion cells within the same optic ganglion do not have synaptic interactions (demonstrated with 25 pairs). Optic ganglion cells do interact, however, with other optic ganglion cells in the contralateral optic ganglion. Neurochemical features of these photoreceptors and optic ganglion cells were also helpful in reproducibly identifying cells within the Hermissenda visual system. A number of observations, for instance, indicated a cholinergic basis for inhibitory 526 DANIEL L. ALKON 1000 1 Bet ACh Ch FIGURE 18. Radiochemical scan of electrophoretogram. A : combined sensory structures (eye, optic ganglion, statocyst ) incubated with labeled choline. B: connective tissue incubated with labeled choline. C : eyes incubated with labeled choline. Ch, choline ; ACh, acetylcholine ; and Bet, betaine ; labeled phospholipids (not shown) migrated very slightly ( Heldman et al., 1979). interactions between the photoreceptors (Heldman ct al., 1979). The eyes, after isolation from the rest of the nervous system, synthesized and accumulated acetylcho- line (as determined hy electrophoretic separation of products that incorporated radioactive label) but not other putative neurotransmitter substances (Fig. 18). Electron micrographic (Crow et al., 1979) sections showed clear round vesicles (consistent with synthesis and storage of acetylcholine) within the photoreceptors' somata, axon hillocks, and terminal branches. Pharmacologic experiments indi- cated cholinergic receptors on the terminal branches of the photoreceptors (Fig. 19). Neuronal endings within the optic tract, near the photoreceptor's terminal branches, stained for acetylcholinesterase. Another example of genetic specification of neuro- chemical characteristics was demonstrated within the optic ganglion. Using a histofluorescence technique to detect and localize small quantities of catecholamines or serotonin, we found that only one cell body within the optic ganglion fluoresced (Fig. 20). For 15/15 preparations this was clearly green long-lasting fluores- cence (indicative of catecholamines) which appeared only after treatment of the freeze-dried ganglia with paraformaldehyde. GASTROPOD MODEL OF LEARNING 527 Hair cells The statocysts of Hermissenda consist of 12-13 disk-shaped hair cells which are 40-50 /mi in diameter and 5-10 /mi in thickness. Scanning electron micro- graphs of normal specimens show the statocyst to be a sphere, 80-110 /mi in diam- eter, surrounded by strands of a connective tissue bundle which also envelops each of the two Hermissenda eyes and optic ganglia. Fach of the hair cells has approx- imately 120—150 modified cilia (motile hairs) (—8-10 /mi long) which project into the lumen of the cyst and move a cluster of 150-200 crystalline structures known as statoconia (Figs. 21, 22). Gravitational sense — that is, information about the direction of gravity relative to body position — is provided by the otolith organ in higher animals and by the statocysts of invertebrates. Both the statocyst and otolith organ consist of a cavity lined with mechano-sensitive hair cells ; the cavity contains fluid and heavy (with respect to the density of the fluid) particles. A change in the direction of gravity causes the particles to move, in turn mechanically exciting the hair cells. Gravity effects on the Hermissenda statocyst can be simulated by rotating the isolated cir- cumesophageal nervous system with intact statocysts (Fig. 22) while recording in- tracellular potentials from the statocyst hair cells. Hair cells located on the stato- cyst so as to be in front of the centrifugal force vector (Fig. 22) show an increased number and amplitude of small depolarizing waves (here termed "voltage noise"), depolarize, and increase firing in response to rotation. The hairs of these cells in A MEMBRANE POTENTIAL (mV) \ --55 CURRENT (nA) 0-5 IPSP REVERSAL L-80 MEMBRANE POTENTIAL (mV) V50 06 0.8 1.0 CURRENT (nA) IPSP REVERSAL FIGURE 19. Photoreceptor current-voltage relations before and after perfusion with car- bachol. I-V curves of Type A (A) and Type B (B) photoreceptors were plotted in control seawater (filled circles) and in presence of 10~4 M carbachol (open squares) for at least 10 min. The intersection of these two curves indicates the reversal potential of the drug-induced hyperpolarization. Recovery from the carbachol effect was seen 15 min after wash with sea- water (open triangles). Arrows indicate the reversal potentials of the natural IPSPs in these same cells (Heldman et al, 1979). 528 DANIEL L. ALKON FIGURE 20. Formaldehyde-induced fluorescence in the optic ganglion (OG). A: phase contrast optics showing the ganglion and individual axons within the optic nerve (ON). B: fluorescence microscopy of the above. One cell within the optic ganglion has intense greenish fluorescence indicative of catecholamines. Note that the fluorescent material is concentrated in the cytoplasm and in the axon, but not in the nucleus (Heldman ct al., 1979). front of the force vector experience the weight of the statoconia (as accelerated by rotation) and move with a fundamental frequency of 7 Hz. Hair cells located on the statocyst so as to be behind the force vector hyperpolarize, show a decreased voltage noise, and decrease firing in response to rotation. The hairs of these cells behind the force vector do not interact with the statoconia, which have been moved (as accelerated by rotation) to the opposite half of the statocyst (Fig. 22). These hairs move with a fundamental frequency of 10 Hz. These observations and a number of additional experiments involving alteration of hair movement (with such treatments as cooling, hypertonicity, chloral hydrate, vanadate, and 4,4'-dithiobisphenyl azide) indicate that the hair cell generator po- tentials arise from summation of increased voltage noise (DeFelice and Alkon, 1977; Grossman et al., 1978; Stommel ct al., in press) which in turn arises from force exerted by the statoconia on the hairs but not from hair bending per sc. This force then causes sustained membrane distortion and consequent conductance changes at the base of the cilia (Fig. 23) rather than on the ciliary shafts. The hair cells in Hermissenda, unlike vertebrate hair cells, have axons which join together to form the static nerve, which extends for 30-40 //.m before entering the pleural ganglion. Cutting the nerve proximal to this point of entry can elimi- nate all spontaneous impulses and synaptic potentials (as was the case for cutting GASTROPOD MODEL OF LEARNING 529 the photoreceptor axons) or those elicited by depolarizing current pulses. Thus, in an intact hair cell a depolarizing generator potential spreads passively from the hair cell's luminal surface and sonia down its axon to an excitable zone where im- pulses are triggered. The impulses spread (probably in a regenerative fashion) into the hair cell's terminal branches, where they elicit synaptic potentials on neigh- boring hair cells and other post-synaptic neurons. The synaptic organization of the statocyst enhances the differences in responses from hair cells located 180° with respect to each other on a statocyst equatorial locus. Thus, a kind of lateral inhibition provides increased contrast between gravitational responses of the hair cells for different orientations of the animal. Of 75 pairs of hair cells recorded within the same statocyst (Detwiler and Alkon, 1974) 43 pairs showed unidirectional inhibition, i.e., stimulation of one cell of a pair caused inhibi- tion of the other cell, but not vice versa. Unlike the IPSPs between photoreceptors or from Type B photoreceptors onto optic ganglion cells, this inhibition of a hair cell occurred only in response to a train of spikes in the presynaptic hair cell. Of these 43 pairs of hair cells, 36 were located at ~ 90° with respect to each other on statocyst equatorial loci. Of the 75 intracyst hair cell pairs, 16 showed reciprocal or mutual inhibition. The hair cells of all 16 pairs were located at — 180° with respect to each other on statocyst equatorial loci. Of the 75 intracyst pairs, 7 showed very weak electrical interaction. In 61 pairs of intercyst hair cells (i.e., from each of the two statocysts), 15 showed substantial electrical coupling, 20 showed unidirectional or reciprocal inhibition, and 22 did not show any interaction. Intersensory interactions The same approach to investigating neural organization as described for photo- receptor, optic ganglion, and hair cell interactions, was also taken for interactions between the statocyst and visual pathways. In an initial study (Alkon, 1973a), I found that some hair cells inhibit ipsilateral and contralateral photoreceptors (11/59 photoreceptor-hair cell pairs). Type B photoreceptors inhibit ipsilateral and contralateral hair cells (4/36 pairs). Type A photoreceptors excite ipsilateral hair cells (3/17 pairs). Hair cells inhibit and are inhibited by ipsilateral optic ganglion cells (2/14 and 3/14 pairs). In subsequent studies (Tabata and Alkon. 1979, and in preparation) the nature of these intersensory interactions was corre- lated with the position of the hair cell on the statocyst. We found that only caudal hair cells (i.e., on a dorso-ventral statocyst equator) inhibit the ipsilateral photo- receptors and Type B photoreceptor impulses can eliminate (Fig. 24) a source of tonic inhibition onto caudal hair cells from the ipsilateral optic ganglion. The pre- synaptic source of these IPSPs within the ipsilateral optic ganglion is also the source of EPSPs onto the ipsilateral Type B photoreceptor (Fig. 25). This pre- synaptic source of EPSPs and IPSPs is also inhibited by caudal hair cell impulses (Fig. 26). Chemoreceptors Agersborg (1922, 1925) originally presented histologic evidence for the loca- tion of chemoreceptors within the Hermissenda tentacles. Consistent with this evidence were our own preliminary histologic investigations (Fig. S) and the finding that stimulation of the tentacle with food substances (squid extract, egg yolk) caused synaptic excitation of small neurons on the ventral side of the cerebro- pleural ganglion (CPG) (Alkon ct a!., 1978). These EPSPs summated with 530 DANIEL L. ALKOX UJ I- i- < Z H- LJ O <-> IT o: o o GASTROPOD MODEL OF LEA KX ING 5JU higher concentrations of stimulus solutions to cause steady membrane depolarization and increased impulse activity of these neurons (11/26 cells). Glycine (> 10"8 M) and DL-methionine (> 10 In M) solutions also caused KI'SPs and depolarization of ventral CPG neurons. These ventral neurons were affected similarly by elec- trical stimulation of the nerve existing from the ipsilateral tentacle (44/55 cells). Other experiments showed that the photoreceptors, optic ganglion cells, and hair cells received synaptic inhibition from the tentacular chemosensory pathway. Mnltiscnsory interneurons and nwtorneurons In the pleural ganglion two classes of central neurons (Akaike and Alkon, 1980) were identified, which received input from the three sensory pathways just discussed : the visual, statocyst, and chemosensory. Of 74 neurons on the dorsal surface of the pleural ganglion (in a restricted region surround the optic tract's point of entry) 26 cells depolarized, 16 hyperpolarized, 4 depolarized and then hyperpolarized, and 28 cells showed no potential change in response to light. Dis- crete EPSPs were recorded from 14/26 neurons which depolarized and discrete IPSPs from 10/16 neurons which hyperpolarized in response to light. All of the discrete synaptic potentials followed impulses of ipsilateral Type B photoreceptors in a one-for-one fashion. Statocyst and chemosensory pathway stimulation pro- duced on these central visual neurons synaptic effects which paralleled the effect of light. Thus, central neurons which received EPSPs (IPSPs) during light also, for most cases, received EPSPs (IPSPs) during statocyst and chemosensory pathway stimulation. Our knowledge of the interneurons in these sensory pathways is still, of course, in- complete (Fig. 27). This is also true of putative motorneurons. Recently we have explored the largest cell (and two adjacent neighbors) in the Hcrurisscnda nervous system. This cell, designated Pedal 1 for its location in the left pedal ganglion, receives synaptic input from the visual pathway and. upon electrical stimulation, controls movement of the cerata along the entire length of the animals (Jerussi and Alkon, 1980). In addition to mapping central ganglia motor neurons with intra- cellular electrodes we are also making extracellular records of impulse activity FIGURE 21 (Sideways on page). Scanning electron micrograph of Hcnnisscnda statocyst. (Upper) On the luminal surface of the statocyst, hairs make contact with statoconia. (Lower) Higher magnification of the above (Grossman ct a/., 1979). FIGURE 22 (Sideways). Upper left: Light micrograph of vital preparation of Hcrmisscnda statocyst. Dorsal view (upper cyst) of equatorial cross-section shows statoconia maintained in a spherical cluster by hair movements. Same preparation after tilting microsocope 90° shows (lower cyst) statoconia moved by gravity to the lower half of the statocyst. Statocysts are ~ 100 /*m in diameter. Upper center: (A) Vital preparation. Motile hairs in Hcrmissenda statocyst visualized with Xomarski optics (1,250 X) in an intact statocyst placed on a microscope tilted 90°. Apical half of the motile hairs could not be photographed but could be visualized. (B) Preparation fixed with 0.5% glutaraldehyde. The full lengths of the hairs, now immobilized, are apparent in the photograph. A single statoconium (S) is seen among the hairs (Grossman ct al., 1979). Upper right: Photograph of modified Garrard turn- table and recording apparatus (f) Faraday cage; (p) pedestal for preparation; (a) amplified; (s) slip rings. Lower left: Diagram illustrating the loci of hair cells with respect to a centrifugal force vector. Lower right : Response to rotation of hair cell with luminal surface opposite to the centrifugal force vector. The axon of the hair cell was cut to eliminate all synaptic but not impulse activity. Maximal centrifugal force is indicated at the left of each lower trace monitoring the rotation. The amplitude of monitor signal is proportional to the angular velocity of the turntable. The interval between monitor signals equals the period of the turntable's rotation. Impulse peaks are not included in the bottom record (Grossman ct a!., 1979). 532 DANIEL L. ALKON R P 20 mV I sec FIGURE 23. Conceptual model for hair-cell voltage-noise and generator-potential produc- tion. The hair is inherently motile (range of motion indicated by dash lines). (A) When the hair cell (oriented in the same direction as the centrifugal force vector) experiences a nega- tive centrifugal force, the hair moves freely through a maximum range of motion producing little voltage noise (see noise record on right). (B) When the hair cell on that statocyst equator visualized when looking dorsoventrally is exposed to zero centrifugal force, the hair has some interaction with the statoconium. The statoconium exerts little net force on the hair although collisions of the statoconium with the rigid hair increase the voltage noise vari- ance and frequency (voltage noise record on right). (C) When the hair cell (with lutninal surface opposite the force vector) is exposed to a substantial centrifugal force (e.g. 1.0 g), the statoconium exerts a considerable net force on the hair, limiting its range of movement and reducing its movement frequency as well as dramatically increasing the voltage noise variance (see noise record on right). The increase of voltage noise variance and frequency results, by summation of individual events, in a depolarizing generator potential. Dash line indicates a positive centrifugal force exerted by rotation. All noise records were taken from actual hair cells under the conditions described (Grossman et a/., 1979). within connectives exiting the circumesophageal nervous system and coursing peripherally to innervate the animal's musculature. We have been able to relate this output activity to input (light) responses of the photoreceptors (Richards, Lederhendler, and Alkon, in preparation). This approach should enable us to better understand the behavioral meaningfulness of changes in Type B firing frequency elicited by natural stimuli as well as consequent to training paradigms. XEURAI. CORRELATES OF HERMISSENDA LEARNING BEHAVIOR With an extensive, though by no means complete, knowledge of neural organiza- tion within and between H ermissenda sensory pathways, it was possible to ask what changes of this organization can be correlated specifically with the associative learning of the animal. More important, if such changes could be detected, it might also be possible to construct a causal sequence for the behavioral change beginning GASTROPOD MODEL OF LEARNING 533 wU — l - sees FIGURE 24. Simultaneous intracellular recordings of EPSPs from Type B photoreceptor (lower records in A, B, C) and IPSPs from ipsilateral caudal hair cell. Type B impulse (in response to 30.0 sec step at + 1.0 nA) eliminates IPSPs (B). EPSPs and IPSPs have in- creased frequently following current (C) (Tabata and Alkon, in prep.). with the first neural site where the two distinct stimuli associated during training are associated by the animal's nervous system. As a working hypothesis, a change at a well-defined convergence point between the statocyst and visual pathway seemed to be a reasonable primary neural locus where light paired with rotation could be associated by the Hcriuissenda neural systems. In fact, the first neural correlate I found of the short-term loss of Hennisscndds positive phototaxis following associative training was a marked reduction in excitation of ipsilateral hair cells by the type A photoreceptor (Fig. 28A, B). A subsequent study (Alkon, 1976) revealed that training with light paired with rotation (but not light alternating with rotation, i.e. unpaired presenta- tion, nor light alone or rotation alone) caused the Type A cell to respond with far fewer impulses than during its depolarizing response to light alone (Fig. 29). This and other related observations suggested that the Type A cell was somewhat hyperpolarized following associative training because of increased inhibitory input from the Type B photoreceptors in the same eye (Alkon, 1976). Supporting this hypothesis was the finding (Alkon and Grossman, 1978) that even a single paired presentation of light and rotation was followed by a prolonged increase of Type B firing and an increased number of IPSPs onto the Type A cell as compared to un- paired or individual stimulus presentations (Fig. 30). These observations and their interpretation then suggested that the Type A photoreceptors would respond 10 mV sees FIGURE 25. Simultaneous EPSPs (lower record) recorded from Type B photoreceptor and IPSPs (upper record) recorded from ipsilateral caudal hair cell. The presynaptic source of both types of synaptic potentials is within the ipsilateral optic ganglion (same cell as in Fig. 24). 534 DAXIEL L. ALKOX I 5 mV •L, ,',- vj J.JJ jjjjj J J JJ i j <~ f*S - /- . «v- _u_^__f ^ FIGURE 26. Depolarization of Type B photoreceptor (lower record) follows impulse train of ipsilateral caudal hair cell (upper record). A positive current pulse (+0.5 nA) injected into hair cell elicits increased firing of hair cell and slight inhibition of Type B cell. After high fre- quency firing hair cell remains slightly hyperpolarized and without impulse activity while Type B cells show increased EPSPs and prolonged depolarization ( Tabala and Alkon, in preparation). with fewer impulses and the Type B with more impulses in response to light during retention of long-term associative behavioral change specific to training with paired (but not random) light and rotation (Crow and Alkon, 1978). Intracellular record- ings from nervous systems isolated from animals trained in this way showed that Type B photoreceptors' impulse activity following a light step was particularly en- hanced in animals which had associative training (X = 2.84), compared (t = 2.47; P < 0.025) to random (X = 2.09) and other control training paradigms. These and other changes of Type B photoreceptor responses, specific to training with paired stimuli, also persisted during long-term retention of the associative behavioral change (Crow and Alkon, 1980). As another example of a change specific to associative training. Type B input resistance for the paired group (X — 88.5 Mn) was significantly greater (Mann-Whitney U test, U - 4; P — 0.026) than that for the random control group (X = 62.7 Mn). Dr. Crow and I also found that there was a significant positive correlation (Spearman P = 0.63 ; P < 0.01) between the Type B impulse frequency after training and the behavioral response latencies to INTERNEURONS STATOCYST FIGURE 27. Schematic summary (partial) of ipsilateral neural interactions within and between the visual, statocyst, and chemosensory pathways. Branches perpendicular to lines (representing axons) indicate inhibitory synapses. Double-lined branches indicate excitatory synapses. Dash lines indicate synaptic interactions presently under study. No contralateral ipsilateral interactions (which have been investigated) are included. GASTROPOD MODEL OF LEARNING 535 enter the light field for the paired light and rotation group (Crow and Alkon, 1978; Crow and Alkon, 1980). This last observation suggested a causal relationship be- tween Type B impulse activity and the animal's associative behavioral change. Dr. Toshiaki Takeda and I have, in addition, observed similar changes following associ- ative training in Type B photoreceptors' firing frequencies after a light step, which were correlated with and very possibly causal of changes in the firing frequency of the Pedal 1 cell (or putative motor neuron) after a light step (Takeda and Alkon, in preparation). Cellular causality of associative learning Long-term enhancement of Type B photoreceptor responses can account for the other observed neural changes (decreased hair cell excitation by Type A cells, de- creased firing of Type A cells in response to light, decreased firing of Pedal 1 after a light step) produced by associative training. Since Type B firing is also known to control output activity in nerves exiting from the circumesophageal nervous sys- tem, these cells are very likely the site of a primary (if not the primary) neural change responsible for the associative learning of Hcrmisscnda. If increased firing frequency of Type B cells has a primary causal role in the acquisition of the associative behavioral change retained by Hermisscnda, it should not be due to a long-lasting increase in excitatory synaptic potential frequency. In fact, hyperpolar- izing photoreceptors to the reversal potential of the IPSPs (between photorecep- tors) did not reveal any difference in EPSP frequency between cells from the trained and random control animals (Crow and Alkon, 1980). Type B photoreceptors from the paired group also required a greater level of hyperpolarization from their rest- ing level (X = 13.9 mV) to block spike activity as compared with random controls (X = 7.6 mV; t = 4.66; P < 0.005). These results together with other observa- tions on Type B cells without impulses or synapses (see below) indicated that asso- ciative training at first results in a persistent non-synaptic depolarization of the Type B photoreceptors. Such non-synaptic depolarization should also be observable when the same sen- sory stimuli, light steps paired with rotation, are presented repetitively to the iso- lated Hcrmissenda nervous system. Intracellular recordings during and after light and rotation stimulus regimens demonstrated that depolarization of type B photoreceptors following stimulus pairing (Fig. 31) accumulated with repetition. This accumulation was specific to pairing of light and rotation (vs. light alone, or explicitly unpaired stimuli) and to the orientation of the nervous system with re- spect to the center of rotation (Farley and Alkon, 1980; Alkon, in press). This cumulative depolarization following stimulus pairing did not occur when the cir- cumesophageal nervous system was oriented with its cephalic pole way from the center of rotation (cephalic orientation). Indeed, for the cephalic orientation, paired light and rotation (but not explicitly unpaired stimuli) produced a slight hyperpolarization. This cumulative Type B depolarization, specific to stimulus pairing and orienta- tion, most likely represents the beginning of the long-term neural change which was observed after acquisition and during retention of the associative behavioral modi- fication described above. If cumulative depolarization of the Type B cell is causally related to acquisition of the learned behavior, we would predict, on the basis of the intracellular recordings during training of the isolated nervous system (cf. Fig. 31 ), that restricting the animal to the caudal orientation should not prevent 536 DANIEL L. ALKON mV 20 0L • ! mV 60 40 20 I I I I 1 3.5 3.0 2.0 .0 0.5 1.0 20 30 4.0 5.0 0 1.0 2.0 30 4.0 50 s s CONTROL = 5 % (C) (D)! o Q. O) Q TEST (A) ¥ (B) o Q. 01 Q. I i..,:...! . «/ N = 20 N- 9 T N = 19 NHO N=39 N'-22 N 12 tf 6 FIGURE 28A. Hair cell responses to light. Upper records are of hair cells in untrained animals. Left marker indicates duration of light flash to ipsilateral eye. Right marker indi- cates flash to contralateral eye. (Lower records) The hair cell of associatively trained ani- mal responds with an isolated hyperpolarization to illumination of the ipsilateral eye, and shows no response with illumination of the contralateral eye. Intensity of flashes: 6 X 103 ergs-cm-2- sec"1. Base-line activity of lower hair cell : 210 impulses per min ; of upper hair cell : 140 impulses per min (Alkon, 1975). FIGURE 28B. Hair cell responses to illumination (6 X 103 ergs -cm-2- sec"1) of ipsilateral eye measured by the ratio of hair cell firing frequency for 10 sec immediately after to the fre- quency for 10 sec immediately before a 1-sec test flash. Each measurement was the average of two to four responses. Test group A, B : Animals exposed to 3 hr of rotation associated with light ; A : maintained with 6.3 hr of daily light. B : maintained with 18 hr of daily light. Control group C : Animals exposed to 3 hr of rotation in darkness, maintained as Group A. Control group D : Animals taken directly from aquarium, maintained as Group A. Control GASTROPOD MODEL OF LEARNING 537 ,1 ' ' f 10 mV ^*WiKJx**/ I I ' 1 I v^|^*vv^^vys^k FIGURE 29. Effect of rotation and light stimulus regimen on responses of type A photo- receptor to light steps. A. Steady-state response before stimulus regimen. B. Response im- mediately after stimulus regimen. C. Response 20 min after stimulus regimen. Bottom trace in each record indicates duration of the light stimulus. Note that the primary effect of stimulus regimen is a decrease of firing frequency during the type A response. This effect is reduced 20 min after stimulus regimen. Spike amplitudes in B are slightly smaller in comparison to spike amplitudes in A and C for approximately equal firing frequencies. First two frames in A, B, C interrupted by 1 sec. Second and third frames interrupted by 20 sec (Alkon, 1976). the training effects and might even exaggerate them. We would also predict that restricting the animal to the cephalic orientation will prevent training effects and may even reverse them. These predictions were confirmed by behavioral experi- ments (Farley and Alkon, 1980, and in preparation). Although incomplete, the evidence for the Type B photoreceptor's primary re- sponsibility for Hermissenda's associative learning is quite substantial. The evi- dence for a change located within this cell's soma was complete and unequivocal. Dr. Terry Crow and I measured increased input resistance (Fig. 32), membrane depolarization, and facilitated light responses which were specific to associative training (Crow and Alkon, 1978; Crow and Alkon, 1980) even after the cell's axon group E: Animals, maintained as Group A, exposed to 3 hr of light intervals (identical to those used for the test groups but without associated rotation). Horizontal extent of bars is proportional to the percentage of animals within a given ratio interval. N = Number of hair cells. N' = Number of animals (Alkon, 1975). 538 DANIEL L. ALKON TABLE I Comparison (Mann-Whitney U test) of membrane characteristics for Type B cells (cut nerve) Paired vs. Random Characteristic (U) (P) Input resistance 10 0.025 Membrane potential 5 0.024 Hyperpolarizing phase of light response 5 0.024 Long-lasting depolarization 7 0.053 and terminal branches had been removed by a razor cut (Table I). After such a cut the Type B cell is electrically isolated from all other cells in the Hermissenda nervous system (Fig. 35 ; Alkon, 1979). This is so because the photoreceptors are not electrically coupled and because no chemical synapses occur on the somata. Thus, these changes could not be explained as secondary to changes of synaptic FIGURE 30. Effect of paired light and rotation stimuli on Type A and B photoreceptor responses. A. Responses of intact lateral type A photoreceptor to light alone (lower record) and light paired with rotation (upper record). The end of a 10-sec light step (intensity in —log units) is indicated by bottom trace. The end of a rotation stimulus (onset 5 sec before light), with a maximum of 1.3 g, is indicated by top trace. The number of IPSPs during the LLH (long-lasting hyperpolarization) is greatly increased by paired stimulation. Maximum hyper- polarization is also greater after light step but before rotation has ceased entirely. B. Responses of intact type B photoreceptor to light alone (lower record) and light paired with rotation (upper record). The end of a 20-sec light step (intensity in — log units) is indicated by bottom trace. Rotation, indicated by top trace, as above. The number of impulses during the LLD is greatly increased by paired stimulation (Alkon and Grossman, 1978). These recordings were made with a brush pen recorder (frequency response to 40 Hz). Note the Type A recording in A saturates the pen's maximum range following stimulus pairing. GASTROPOD MODEL OF LEARNING 539 A EYE OPTIC GANGLION STATOCYST INHIBITION OF INTERSENSORY 2nd ORDER INHIBITION VISUAL CELLS /„ POSITIVE 2nd ORDER FEEDBACK NEGATIVE 2nd ORDER FEEDBACK MUTUAL INHIBITION B ROTATION HAIR CELL V TYPE B PHOTORECEPTOR V sees LIGHT FIGURE 31. Neural responses to stimulus pairing. (A) Neural system (schematic and partial diagram) responsive to light and rotation. Each eye has two type A and three type B photoreceptors ; each optic ganglion has 13 second-order visual neurons ; each statocyst has 12 hair cells. The neural interactions (intersection of vertical and horizontal processes) identi- fied to be reproducible from preparation to preparation are based on intracellular recordings from hundreds of pre- and post-synaptic neuron pairs. Abbreviations : In HC, hair cell ~ 45° lateral to caudal north-south equatorial pole of statocyst; S, silent optic ganglion cell, elec- trically coupled to E cell ; E, optic ganglion cell, presynaptic source of EPSPs in type B photo- receptors. The E second-order visual neuron causes EPSPs in type B photoreceptors and cephalad hair cells and simultaneous inhibitory postsynaptic potentials (IPSPs) in caudal hair cells. (B) Intracellular recordings (simultaneous) from caudal hair cell and type B photo- receptor shows increase of EPSP's (type B cell, lower record) and simultaneous IPSPs (caudal hair cell, upper record) after light paired with rotation. The LLD after stimulus pairing is greater than that after light alone (line of long dashes). The line of short dashes indicates level of resting membrane potential. The lowest trace indicates light duration ; top trace, angular velocity of turntable (effecting 1.2 g) (Alkon, 1979). 540 DANIEL L. ALKON B -40 -30 -20 -10 »— • PAIRED > — o RANDOM Em<(mV) 0.1 0.2 0.3 0.4 I (nA) PAIRED RANDOM Id t j 1 sec 10 mV 0.2 nA PAIRED R.P. Light FIGURE 32. Cellular changes in cut nerve preparations from experimental and random control animals. (A) Receptor potentials of dark-adapted Type B photoreceptor from an ex- perimental animal (light paired with rotation). Responses evoked by brief light flashes of in- creasing intensity (4), —log 3.0; (3), —log 2.0; (2), —log 1.0; —log 0.5. Dash line indi- cates resting membrane potential and lower trace indicates duration of light flash. The ab- sence of spikes and synaptic potentials indicate that the photoreceptor soma was successfully isolated from the area of spike initiation and synaptic input. (B) Representative linear cur- rent-voltage relationship of dark-adapted isolated (cut nerve) Type B photoreceptors from experimental (paired) and random control groups. (C) Examples of changes in membrane potential of dark-adapted isolated B photoreceptors from experimental and random control animals. Electrotonic potentials evoked by hyperpolarizing square current pulses (bottom traces) through a balanced bridge. Resistance measurements taken with the single electrode- bridge circuit are consistent with data from experiments in which the photoreceptors were im- paled simultaneously with two microelectrodes for current injection and voltage recording (Crow and Alkon, 1980). (D) Voltage representation of membrane changes during associa- tive learning. Theoretical voltage recordings of Type B (cut nerve) responses to light follow- ing three days of paired and randomly associated light and rotation. These responses are based on data referred to in Table I and actual responses of Fig. 32A, B, C. input, for instance, from the optic ganglion (see above). Furthermore, since the Type B cell is at the input stage of the Hermissenda nervous system, it can be ex- pected to affect post-synaptic neurons at various stages within the visual pathway. GASTROPOD MODEL OF LEARNING 541 A quantitative assessment of the degree to which a change in Type B cell responses to light causes a change in output impulse activity and motor responses certainly will require further extensive investigation. Additional evidence will be provided by recordings from cells of intact (Jerussi and Alkon, 1980) and semi-intact (Fig. 33) behaving animals. Elimination of Type A cell activity by negative current injection might reduce or eliminate the behavioral response to light (as occurs with increased impulse activity of Type B cells and the consequent increased inhibition of the Type A cells). Hyperpolarization of the Type B cell after associative training should help restore the normal behavioral response to light. Similar ex- periments at each station of the visual pathway of behaving animals will further define the causal role of known neural elements such as the Type B photoreceptors in the visual behavior and its modification by training. The difficulties of precisely defining processes which cause changes of animal be- havior is illustrated by work on habituation of the gill withdrawal reflex of Aplysia. Although it has been possible to recognize some large identified neurons in the Aplysia abdominal ganglion, a comprehensive description of neural circuits which entirely account for even this relatively simple reflex and its habituation has never been completed (Jacklet and Rine, 1977). Thus, although the excitatory post- synaptic potentials (EPSPs) produced by sensory cells on motor neurons are known to be responsible for much of this behavior (Castellucci et a!., 1970), peripheral CNS Tent *~- Rhm Esoph SIPHON \ S.N. / / / / ,• "m ^ ////// ; / / 24 SN Groups MN LOG, LDG2 L7 etc \J / / / / / / / / i 0 / /Br>Ct.N. \ :: / / / / / • < //////////*///////// Peripheral system GILL MUSCLES FIGURE 33. (Left) Semi-intact preparation of Hcrmisscnda circumesophagael nervous system. The visual, statocyst, and chemosensory pathways are intact although most of the animal's foot has been removed. FIGURE 34. (Right) Diagram of the synaptic sites for habituation of the gill withdrawal reflex activated by tactile stimulation of the siphon. Sensory information is conveyed to the central nervous system (CNS) via the siphon nerve (S.N.) and other nerves (Ct.N and Br.N) to central sensory neurons (SN). The synapses (site 1) of sensory neurons with motor neurons (MN) are subject to synaptic depression and the synapses (site 2) of motor neurons with gill muscles are subject to frequency-dependent homosynaptic facilitation. Dis- habituation causes heterosynaptic facilitation of the sensory neuron terminals (site 3) a-nd the firing of motor neurons, which facilitates neuromuscular junctions (site 2). The facilitation and its subsequent decay is frequency-dependent and may carry over from trial to trial. Addi- tionally, the gill is activated (most effectively in the absence of the PVG) by a peripheral path- way (15) that shows habituation and dishabituation, presumably at chemical synapses (site 4). Other synaptic influences of excitatory or inhibitory sign and central or peripheral origin (site 5) may also influence habituation. For example, inhibitory junction potentials in gill musculature (14) and the suppressive influence of the PVG on the peripheral pathway (15) may contribute. Also, the branchial (gill) ganglion, which is important in gill habituation to direct gill stimulation (21, 22) may contribute. Responses of the isolated siphon to direct siphon stimulation also readily habituate (23-25) (Jacklet and Rine, 1977). 542 DANIEL L. ALKON neuronal interactions have also been implicated (Peretz, 1970; Peretz and Estes, 1974). Interneurons which have interactions with both sensory and motor neurons are thought to exist (Fig. 34), but their influence has not been fully assessed. In- teractions between sensory cells and between motor neurons also have not been rigorously analyzed. The close correlation of EPSPs (recorded from the motor neurons) and behavioral decrement during habituation of the gill withdrawal reflex has been observed by Kandel and his colleagues (Carew et al., 1972; Carew and Kandel, 1973). Other studies revealed, however, that central ganglionic neuronal changes were not alone sufficient to explain habituation and that peripheral neuronal changes had to be included (Peretz, 1970; Bruner and Kennedy, 1970; Lukowiak and Jacklet, 1972; Peretz and Howieson, 1973; Lukowiak and Peretz, 1974; Peretz et al., 1976; Jacklet and Rine, 1977). VOLTAGE-DEPENDENT CONDUCTANCES AND ASSOCIATIVE LEARNING To better understand the ionic basis of the observed changes within the soma membrane of the Type B photoreceptor during associative learning of Hermissenda and to examine how stimulus pairing is specifically encoded by the effects of synap- tic interactions on this cell's ionic conductances, voltage-clamp analysis in our labora- tory was undertaken. The Type B photoreceptor, located in the anterodorsal por- tion of the eye, is 35 //.m in diameter. Its axon, —* 1 ^m in diameter, leaves the base of the eye, enters the optic nerve, passes ensheathed through the optic ganglion (for — 50 /u.m), enters the optic tract, and terminates in a spray of fine endings — 20 /xm from its entry into the cerebropleural ganglion (Fig. 13). The synaptic interactions and impulses described only occur along the distal axon and terminal endings. Thus, cutting through the nerve near the impulse-generating zone leaves a cell body that contains the phototransduction apparatus without impulses or synaptic interactions (Fig. 35). Since, following lesion, input resistance tends to increase slightly, significant shunting of currents injected or measured across the soma membrane is unlikely. Voltage- and current-clamp recordings can then be made with two electrodes in the cell body under favorable space-clamp conditions. Type B photoreceptors in cut-nerve preparations depolarize for 1-2 min after a 30-sec light step of moderate intensity of 510 nm (103-104 ergs -cm'2 -sec"1). This long-lasting depolarization (LLD) is always associated with a membrane re- sistance 1.5-3 times higher than the resting or dark value. Positive current in- jection, causing depolarization comparable to that produced by light, is not followed by an LLD or increased membrane resistance (Fig. 36). With sufficient hyper- polarization during the light response, however, the LLD and its associated con- ductance decrease could be eliminated. These observations suggested that the LLD arises at least in part from a voltage-dependent decrease of K+ conductance within the Type B soma membrane. Furthermore, since the LLD increased when the light step was paired with rotation, this increase with pairing could be due to the LLD's voltage-dependence. Voltage-clamp experiments provided additional insight into the nature of voltage-dependent conductances of the Type B soma membrane. In darkness a voltage-dependent increase of total membrane conductance (Fig. 37) was apparent with steady-state changes of holding potential from resting ( — • — 55 mV) to more positive levels (Alkon, 1979). Step clamp commands (Shoukimas and Alkon, 1980) from a holding potential of — 60 mV elicit an early, transient outward cur- rent which appears to be carried predominantly by K+ ions. Both steady-state ac- GASTROPOD MODEL OF LEARNING Cut i Cut n 543 Rotation FIGURE 35. Excitation and inhibition of type A photoreceptor. Diagram of type A photoreceptor soma with rhabdome, axon, excitable focus, and terminal branches. Light de- polarized photoreceptors at rhabdomes (Rh). Type B photoreceptors (of which there are three) inhibit, at synaptic endings, lateral type A, directly (1) and indirectly by inhibiting (2) optic ganglion (O.G.) cells (of which there are 13). Ipsilateral hair cells (H.C.) receive less inhibition from optic ganglion cells (3) when the type B fires and thus increases its inhibi- tion of the lateral type A (4). Hair cells also inhibit type A when they depolarize in re- sponse to rotation. In response to light, the type A cell depolarizes without impulses or after- hyperpolarization with cut I lesion. It depolarizes with impulses but without after-hyper- polarization with cut II lesion. The response of an intact type A cell is represented at upper right (Alkon and Grossman, 1978). tivation and inactivation of the conductance are voltage-dependent. Unlike some other gastropod neurons in which similar currents have been reported, such as Anisodoris (Connor and Stevens, 1971), the kinetics of inactivation, like the steady- state value, show voltage-dependence. Voltage-dependent conductances were also indicated during and following light steps (as well as in darkness as described above). In brief, the three currents induced during a light step were most likely carried by : 1. Na+, a rapidly decreasing inward current, greatly reduced by replacing ex- ternal Na+ with equimolar tetramethyl ammonium ions (TMA) and reversing at ~ + 60mV (Fig. 37A). 2. K+, a rapidly decreasing outward current, greatly reduced by 4 mM external 4-aminopyridine, intracellular injection of tetraethyl ammonium ions (TEA) and ethylene-glycol tetraacetic acid (EGTA) and 1 mM external Ni2*, and reversing at — — 80 mV. The reversal potential became more positive when external K+ was raised (Fig. 37, 38). 3. Ca2+, a sustained inward current that decreased slowly following the ces- sation of the light step, was present in Na+-free artificial sea water (ASW), re- duced by lowering external Ca2*, increasing external Mg2+, and 1 mM external Ni21 (Figs. 37; 38B, C; 39). This current was reduced by intracellular injection of 544 DANIEL L. ALKON 10 mV I 1.0 nA FIGURE 36. Effect of light and positive-current steps on intact type B photoreceptor. Cell was impaled simultaneously with two microelectrodes : one to measure voltage only, the other to measure voltage as well as to inject current (lower voltage traces in each pair). A: LLD associated with decreased membrane conductance following a 30-sec light step (indi- cated by lowest trace and intensity expressed in —log units). Membrane conductance was measured by injection of negative-current pulses (indicated by lowest tract). Dash line indicates level of resting membrane potential. B : Response to injection of positive-current step, 1.5 nA. Membrane conductance, measured by injection of negative current pulses (indicated by lowest trace), increases during positive current (delayed rectification). Note the brief hyperpolarizing wave following current step. Pulses are — 0.5 nA. Note changes of paper speed in A, B (Alkon and Grossman, 1978). cyclic- AMP, but not by TEA, and was increased by injection of EGTA. This Ca2+ current was markedly voltage-dependent, not appearing at holding potentials < 35 mV and reversing at holding potentials > + 100 mV. With elevated external K+ (120 mM) the Ca2+ current was still observed at the approximate re- versal potential for the light-induced K+ current. After a light step there was a prolonged outward current. This current in- creased with holding potentials more positive than the resting potential (—55 mV) and decreased with more negative holding potential in ASW and Na+-free ASW. This late outward current was eliminated by 4 mM 4-aminopyridine (Fig. 39), intracellular TEA and EGTA injections, and 1 mM Ni2+ added to ASW. These data indicate that the outward current at holding potentials > R.P. is a K+ current which depends on a light-induced Ca2+ current. Following a light step a substantial prolonged Ca2+ current (most apparent in 4-aminopyridine, which blocks the late outward current) will only occur at holding potentials > R.P. The light-induced K* current (s), therefore, can be enhanced by a light-induced increase of intra- cellular Ca2* only at holding potentials > R.P. How can these voltage-dependent conductances together with the known synaptic interactions encode stimulus pairing and ultimately result in a long-lasting non- synaptic depolarization of the Type B photoreceptor ? Because of the synaptic inter- actions within and between the visual and statocyst pathways (Fig. 31), rotation paired with light is followed by disinhibition of the Type B cell and an increased GASTROPOD MODEL OF LEARNING 545 number of EPSPs in this cell. Because membrane resistence (and thus the mem- brane time constant) is increased following a light step, the synaptic depolarization (due to disinhibition and the EPSPs) is enhanced. The synaptic depolarization following stimulus pairing, however, also facilitates the light-induced depolarization that was shown to arise from voltage-dependent Ca2t and K+ conductances. ASW A — A Dark Current • — • Light Current (peak) -95 -75 I (nA) 120 mM K ' No' - free ASW A— A Dark Current • --• Light Current ( 2 minimum) KnA) 8 6 4 B No - free ASW £> — A Dark Current O— • O Light Current ( peak at ~ 5" ) • — • Light Current (minimum at —2 -155 -135 -115 -95 mV) 5 -75 -55 -35 L^^ 5 25 ^4^' -2 I i - Em(mV) ^^^ •'' -« AA • '' -14 - No - free ASW • --• K ' current ( lote) O""O Co current KnA) 1.0 08 06 .O' -120 -80 .O' C--4 ?"9 5 25 4! 1 1 1 i -2 - 4 E^lmV -02 :-0.4 Vo.6 20 \ 60 -i — nin Em(mV) -1.88 -20^-. FIGURE 37. Current-voltage plots from voltage-clamp recordings in darkness and during and after light steps (unfiltered) ; negative current = inward current. (A) Dark current and maximum initial inward current (Na+) in ASW in response to 1.8 X 106 erg -cm"2 -sec"1 at dif- ferent holding potentials. Arrow indicates reversal potential extrapolated from linear portion of current-voltage of I-V plot of Na* current. (B) Dark current and isochronal (2 and 5 sec after light onset) values of initial outward current (K*) and sustained inward current (Ca2+) in Na+-free ASW in response to 1.8 X 109 and 1.1 X 10" erg -cm"2 -sec"1, respectively, at different holding potentials. There is no measurable sustained inward current at holding potentials < RP. (C) Dark current and isochronal (2 sec after light onset) value of initial outward current (K+) in 120 mM K+ in Na+-free ASW in response to 1.8 X 10" erg -cm""2 -sec"1 at different holding potentials. Reversal potential (arrow) of outward current has shifted toward more positive holding potential. (D) Absolute values of sustained light-induced inward current (Ca2*) and late outward current (K*) as a function of holding potential. Ca2* value was measured 3 sec after light step cessation ; K+ value, 35 sec after light step cessation. Note the parallel relation of the' two currents to holding potential. Light intensity, 1.1 X 10" m^-sec--1 (Alkon, 1979). 546 DANIEL L. ALKON ASW No - free ASW No - free ASW sec sec I5mV I 5mV I 5 mV + 45 mV + 20 mV + 45mV , — y -5 mV - 25 mV N -5mV -30 mV f\ -20mV -55 mV -80mV -105mV K-80 mV . -85mV -75 mV' > " '55mV -90 mV -130 mV r -155 mV FIGURE 38. Voltage-clamp recordings during and after light steps ; up = inward current. Values at left are absolute holding potentials. Dash lines indicate steady-state current in dark- ness for each holding potential. Voltage traces at top were recorded simultaneously with current (flowing from reference element in the bath) recordings. Lowest trace in each set indicates duration of light step. A sustained inward current (Ca2*) is induced by light at holding potentials > RP in Na*-free ASW. The early outward current (K+) is marked at higher light intensity (middle column). Light intensity: in middle and left column: 1.8 X 10" erg -cm"2 -sec"1 (unfiltered) ; in right column: 1.1 X 10" erg -cm"2 -sec"1 (unfiltered). Paper re- cording amplifier saturates for some records in ASW (Alkon, 1979). Thus, a kind of regenerative or positive feedback mechanism is suggested as a neurophysiologic basis for the associative learning of Hermissenda. Light-induced depolarization enhances synaptic depolarization, which in turn enhances the light- induced depolarization, and so forth. With each successive trial, residual depolari- zation could potentiate and add to depolarization following the next stimulus pair. How does cumulative depolarization (Farley and Alkon, 1980; Alkon, in press) of the Type B membrane, in turn, cause a very long-lasting change which could account for the animals' decreased positive phototactic behavior ? A causal sequence which is consistent with the currently available data is as follows. Cumulative depolariza- tion, specific to stimulus pairing, causes inactivation of dark voltage-dependent K+ conductances (cf. Shoukimas and Alkon, 1980). Inactivation of voltage-dependent K+ conductance (s) causes increased total membrane resistance of the Type B mem- brane when it is depolarized in response to a light step. An increased membrane resistance would be associated with greater depolarization (of the Type B cell) re- sulting from light-induced ionic currents. Greater depolarization of the Type B cell in response to light causes facilitation of the voltage-dependent light-induced Ca2* currents. In brief, cumulative depolarization (following associative training) causes inac- tivation of a dark K4^ conductance and facilitates the light-induced responses of the GASTROPOD MODEL OF LEARNING 547 10 mV _r - 15 mV J2 nt> -55 mV -85 mV FIGURE 39. Effects of blocking agents on voltage and voltage -clamp recordings during and after light steps (A) Voltage recordings from type B cell in ASW containing 1 mM NiCk Two microelectrodes recorded voltage simultaneously (upper two recordings). Current pulses injected through one microelectrode via a bridge circuit caused smaller voltage change during light (compared to before light) and transiently a larger voltage change after the light (com- pared to voltage change before light) ; this indicates increased membrane resistance after light (compared to membrane resistance before light). The LLD after light persisted in NiClz when membrane resistance was not elevated. This was not so without NiCU. The LLD was reduced at more negative potentials (produced by injecting steady negative current through one micro- electrode) ; light intensity (indicated by bottom trace), 1.1 X 10" erg -cm'2 -sec"1 (unfiltered). (B) Voltage-clamp recordings from same cell as in (A) under same conditions at comparable holding potentials ; up = inward current. Note the long-lasting inward current after light step. Voltage traces at top were recorded simultaneously with current recordings. (C) Voltage-clamp recording during and after light steps in presence of blocking agents. Voltage traces at top were recorded simultaneously with current recordings. Upper recordings in NaMree ASW, 1 mM 4-aminopyridine (4 A-P), and 1 mM NiCU were all at — 5 mV absolute holding potential. Cur- rent records at lower holding potentials were for same cell in 4-aminopyridine. Note absence of measurable early or late outward currents during and after light (intensity, 1.8 X 109 erg -cm"4- sec"1, unfiltered), and presence of slowly decreasing sustained inward current (Ca2+). Dash line indi- cates steady-state current level in darkness (Alkon, 1979). Type B cell. The Type B cell with its enhanced response to light now, via synaptic inhibition, more effectively reduces the Type A cell impulse activity in response to light and this finally is expressed as a decreased positive phototaxis. Conductance changes in non-gastropod learning models Woody and his colleagues observed increased excitability of cells in the coronal pericruciate cortex to current injection in cats conditioned to eye blink by a Pav- lovian training procedure (Woody et al., 1974). Woody and Black-Cleworth (1974) proposed that increased neuronal input resistance could underlie these 548 DANIEL L. ALKON changes in cortical excitability. Increased input resistance due to a decrease of conductance (presumably to potassium) would cause membrane depolarization and thus increased excitability. A similar explanation has been offered regarding changes of impulse frequency produced during training of the isolated cockroach leg preparation (Woollacott and Hoyle, 1977). Berger and Thompson's finding (Fig. 10) that hippocampal units increase their firing during and following classical conditioning of the rabbit nictitating membrane also might be due to a similar depolarization and decrease of membrane conductance. These measurements of neural changes with non-gastropod learning models, however, are preliminary. Other causes for these neural changes have not been ruled out. Primary changes in neurons which synaptically affect the cells studied in cockroach, pericruciate cortex, and hippocampal neurons could account for the neural correlates thus far observed witli these preparations. As these measurements are extended, however, it will be interesting to compare, across species, their bio- physical basis. Even if a precise causal sequence for associative learning cannot be constructed for the more complex nervous systems, the generality of what we learn with Hermissenda and other gastropods will be increased by the identification of common neuronal changes. The neuronal functions of Hermissenda are not unique. The voltage-dependent conductances within the soma membrane of the Type B photoreceptor are not unique. Voltage-dependent K+ and/or Ca2+ conductances have been inferred or demonstrated in spinal motor neurons (Blankenship, 1968), frog motor neurons (Barrett and Barrett, 1976), Helix neurons (Eckert and Lux, 1976), in vertebrate rods (Fain and Quandt, 1977), barnacle muscle fibers (Hagiwara et al., 1974), molluscan neurons (Connor, 1979; Byrne, 1980), starfish egg (Hagiwara et al., 1974) and hippocampal neurons (Hotson and Prince, 1980). However, the long- term modification of these conductances, as it arises from paired stimulus effects on converging sensory pathways, could be unique to the process of associative learning in a variety of species, all of which have neurons with similar biophysical characteristics. Biochemical correlates of associative learning Long-term modification of conductances means long-term membrane changes which must have a biochemical basis. The presence of a slowly decaying light- induced Ca2+ current and Ca2+-dependent K+ current during the long-lasting de- polarization of the Type B cell after a light step suggested the possible involvement of cyclic nucleotide metabolism. Recent experiments further suggested some in- volvement of cyclic nucleotide metabolism in the generation of Type B membrane currents. Intracellular injection of the catalytic subunit of protein kinase (Alkon, Olds, Kuzma, and Neary, in preparation) enhanced the LLD following a light step. Perfusion with IB MX (a phosphodiesterase inhibitor which elevates intracellular cyclic nucleotide levels) also caused prolonged enhancement of the LLD after a brief initial reduction of the LLD. This initial reduction might be due to smaller intra- cellular cyclic-AMP increments, which were previously found to reduce the light- induced Ca++-current (Alkon, 1979). Interaction of Ca2+ and cyclic nucleotide levels is now thought to occur in a number of tissues (Rasmussen, 1970; Rasmussen et al., 1975; Berridge, 1976; Rasmussen and Goodman, 1975; Malaisse, 1973; Borle, 1974; Lipton et al., 1977; Boron et al., 1978). Intracellular Ca2+ could control enzymes which catalyze the GASTROPOD MODEL OF LEARNING 549 synthesis (Rasmussen and Goodman, 1975) or the- degradation of cyclic nucleotides. There is also evidence of cyclic nucleotide-dependent phosphorylation of proteins in retinal rods (Farber et al, 1979), Ca-+-dependent protein phosphorylation in neural membranes (Schulman and Greengard, 1978), depolarization-induced protein phos- phorylation mediated by Ca2+ influx in synaptosomes (Krueger et al., 1977) and light-dependent phosphorylation of rhodopsin (Shichi and Somers, 1978). More recently phosphorylation studies with Hcnnisscnda eyes have been conducted in our laboratory. We detected seven phosphoprotein bands in all of the eyes isolated from animals of four groups: trained, unpaired random, and normal controls (Neary et al., 1980). There were significant overall differences between paired and con- trol groups (P<0.01) for bands with approximate molecular weights of 23,000 and 20,000 daltons. The paired group was significantly different from both the random and unpaired control groups (P < 0.01 ) while the control groups were not significantly different from each other. The changes of the two phosphoprotein bands' density (increased 49 and 56% for paired vs. random animals) might be related to the depolarization and increased input resistance of the Type B cell fol- lowing associative training. It will of course be necessary to localize the phos- phoprotein changes to photoreceptors within the eye and not neighboring pigment and epithelial cells. We will also attempt to characterize these phosphoproteins and localize them within cellular compartments. The causal sequence for this biochem- ical change will also be of great interest. How, for instance, might Type B de- polarization affect protein kinases and/or phosphatases? Are there direct effects on cyclic-AMP levels, phosphodiesterase activity, etc. ? Is protein assembly in- fluenced? Like the biophysical changes, the biochemical changes need not be unique to serve within the context of the neural systems' response to training stim- uli, in a manner unique to the process of associative learning. Again, if the bio- chemical process can be characterized for Hcnnissenda and other gastropods, its generality might be questioned for other species. Further questions Photoreception and photoreceptor adaptation are now known to be associated with shifts of intracellular cyclic-GMP (Woodruff and Bownds, 1979) changes of phosphodiesterase activity (Rasmussen, 1970; Rasmussen et al., 1975; Lipton et al., 1977), and protein phosphorylation, as well as changes of intracellular Ca2+ concentration (Lisman and Brown, 1972). It could be argued that the biochemical as well as the biophysical changes specific to associatively trained Hermissenda involve mechanisms important to photoreception and are therefore less likely to have general significance. As already emphasized, the voltage-dependent con- ductances which were observed with voltage-clamp of the Type B somata are present in a host of sensory and central neurons in a number of different vertebrate and invertebrate species. Cyclic-nucleotide-Ca2+-protein phosphorylation interaction has also been described for a host of preparations, from synaptosomes to brain slices to retinal rods. These biophysical and biochemical phenomena obviously are avail- able to nervous systems for other functions which are not at all restricted to photo- reception. In addition, the behavioral and neural changes specific to associatively trained Hermissenda cannot be explained or simulated by differences in light adaptation of the Type B photoreceptors. Light stimuli alone (i.e., in the absence of rotation) produced no behavioral or neural changes (Alkon, 1974; 1975; 1976; Crow and Alkon, 1978; 1980). When animals were exposed to twice the duration 550 DANIEL L. ALKON of light each day (Lederhendler and Barnes, unpublished observations) their latency to move toward a light source was not reduced (as observed following paired light and rotation). Furthermore, when the photoreceptors are light-adapted during the light phase of the diurnal cycle, the animals approach a light source but do not dur- ing the dark phase (Lederhendler ct a!., 1980). Light adaptation of the photo- receptors, then, is associated with the behavioral effect opposite that with associative training. Increasing intensity of light stimuli during training (i.e., increasing the level of light adaptation) cannot approximate the effects of stimulus pairing. Insertion of additional steps of light during training with paired light and rotation does not enhance the behavioral change (Farley and Alkon, in preparation). Stimulus pairing in essence shifts the Type B membrane potential (by the retro- grade spread of the synaptic effects following the stimuli) to more positive levels without bleaching more visual pigment molecules (in this case, rhodopsin) of either Type B or Type A photoreceptors. Increased light intensity will bleach more pigment molecules of both photoreceptor types and thus decrease their response amplitudes to subsequent repeated presentation of the light stimuli during a training period. Cumulative depolarization of the Type B cell produced by light alone would then approach a limit defined by pigment molecule bleaching. This limit of cumulative depolarization could then be exceeded only by an external effect such as the synaptic depolarization which follows stimulus pairing. Light steps of in- creasing duration (> 1 min) therefore, can be expected and in fact were observed, followed by long-lasting depolarization of the same or decreasing duration (Alkon, unpublished observations). This is also explained by the fact that light-induced Na+, K+, and Ca2+ conductances have different light intensity-response relationships and adaptation properties (Alkon, 1979). The light-induced K+ current, for in- stance, requires brighter lights but shows much less adaptation than do the Na+ and Ca2+ currents (Alkon, 1979). Increasing light duration beyond levels which produce a maximum depolarization during and after a light step can be expected to cause light-induced K+ currents, larger with respect to the Na+ and Ca2+ currents, that cause more hyperpolarization. In brief, for the Type B cell, a light-induced Ca2+ conductance is enhanced and a voltage-dependent K+ conductance is reduced by the retrograde spread of synaptic depolarization following stimulus pairing. Stimulus pairing produced this en- hanced conductance in a manner which cannot result from light stimuli alone because : 1. The Ca2+, K+, and Na+ conductances have different voltage-dependence, and 2. The relative voltage-dependence of these conductances is quite different from their relative light intensity dependence. A stimulus pair is therefore encoded by the synaptic effect on the Type B cell depolarization after light. Similarly, stimulus pair repetition is encoded by the cumulative depolarization which follows associative training. It is not difficult to imagine that a voltage-dependent synaptic (Magelby and Stevens, 1972) or impulse-generating conductance mediating one sensory stimulus could be enhanced by the synaptic effects caused by pairing with a second stimulus. Stimulus pairing would more likely be encoded by the enhancement of voltage- dependent soma, dendritic, synaptic, and/or axonal conductances in integrating neurons of vertebrate species whose central nervous systems allow a myriad of pos- sible convergence points for associated distinct stimuli within and between sensory modalities. Training with these associated stimuli might then cause a cumulative GASTROPOD MODEL OF LEARNING 551 membrane potential and/or conductance change(s) which, like the Type B depolar- ization, arise from repeated enhancement of voltage-dependent conductances. CONCLUSION The usefulness of gastropod models of associative learning can only be fully assessed in retrospect. Study of these models was undertaken after a number of assumptions, intuitive formulations, and hypotheses were made. It makes intuitive sense that new connections between neurons are not formed during associative learn- ing (Purpura, 1967), because axons and their endings don't grow as quickly as many associations can be learned. It makes intuitive sense that the stimuli associ- ated during learning activate neural pathways which at some point must converge. Presumably, paired stimulation by means of his convergence of pathways produces specific neural signals and neural changes. These in turn cause associative learn- ing behavior. By this line of intuitive reasoning, the convergence point must al- ready exist in order for two stimuli to be associated during learning. If this is true, the number of convergence points will determine the capacity for learning. The number of convergence points and learning capacity are obviously vastly different for gastropods and humans. Yet if we are true to the intuitive formula- tions just mentioned there should be a chance, and it is only a chance, for some common underlying physiology. If the gastropod associative learning behavior closely resembles examples of elementary learning (e.g. conditioning) of humans, and if primary neural changes mediated by convergence points are identified, the chance of some general significance is there. The more basic our understanding of these primary neural changes, the greater the chance that we can test generality in other species whose neural systems may never be described in their totality. There are, however, always the caveats that intuition may misguide investiga- tions and models may cloud issues. It may be possible, for instance with a gastro- pod, to produce a behavioral change which is long-lasting, dependent on paired vs. unpaired training and specific to the training stimuli. Yet the mechanism for one animal's "association" of the two training stimuli may be peculiar to this animal and its training experience. As an example, strong shock to a gastropod's head and/or tentacles may stimulate a host of neuronal pathways, none of which con- verge with the "food" or chemosensory pathway. Instead, strong shock may stimu- late the "food" pathway directly, and, when paired with food, produce neural and behavioral changes without the convergence points which we intuitively expect are necessary in a variety of species for associative learning. The mechanism for a neural change produced by the paired effects of shock and food on the same sensory pathway, i.e., without convergence of inputs, seems less likely to have significance for other species in which neuronal convergence is necessary for associative learning. From our hypotheses about how learning is mediated and limited by convergence points within the nervous system, we would predict that a knowledge of evolutionary constraints on species' learning capacity will help reveal neural mechanisms which may transcend the unique adaptive function of those species. Our gastropod model, then, should be constructed from an understanding of how the animal be- haves in and interacts with its natural environment — how it can learn from naturally occurring stimulus patterns. Gastropod models will then less likely be imposed on and more likely be generated and suggested to us by the learning phenomena of interest. 552 DANIEL L. ALKON The difficulty of designing the best experiment and choosing the most appro- priate model with a minimum of knowledge and helpful intuitions has always con- fronted scientists. It is a source of reassurance to recall Pavlov's analogous predica- ment. His productivity as an experimentalist is not in small part, I believe, due to the insight of his intuitive formulations : "Thus, we may reach the limit of analysis of which a given animal may be capable, i.e. most minute natural phenomena may become special stimuli for a definite activity of the organism. We may think that by the same process by which connexions are formed between cortical cells and subcortical centres connexions are also formed between the cortical cells themselves. The excitations produced by phenomena taking place simultaneously in the outside world are thus complex. These complex excitations may become, under corresponding conditions, condi- tioned stimuli, and be differentiated by means of the just-indicated process of in- hibition from other closely related complex stimuli" (Pavlov, 1927). ACKNOWLEDGMENTS I wish to acknowledge the invaluable assistance of Mrs. Jeanne Kuzirian in all aspects of this article's preparation. I am also grateful to Ms. Linda and Mr. Bob Colder for their gifted and dedicated help in preparing the illustrations. LITERATURE CITED ADEY, W. R., C. W. DUNLOP, AND C. E. HENDRIK, 1960. Hippocampal slow waves: distribu- tion and phase relationships in the course of approach learning. Arch. Neurol., 3 : 74-90. AGERSBORG, H. P. K., 1922. Some observations on qualitative chemical and physical stimula- tions in nudibranchiate mollusks with special reference to the role of the rhinophore. /. Exp, Zool, 36 : 423-444. AGERSBORG, H. P. K., 1925. The sensory receptors and the structure of the oral tentacles of of the nudibranchiate mollusk Hcrmisscnda crassicornis syn. Herrnissenda opalescens. Acta Zool. (Stockholm), 6: 167-182. AKAIKE, T., AND D. L. ALKON, 1980. Sensory convergence on central visual neurons in Hermissenda. J. Ncurophysiol., 44: 501-513. ALKON, D. L., 1973a. Intersensory interactions in Hcrmisscnda. J. Gen. Physiol., 62: 185-202. ALKON, D. L., 1973b. Neural organization of a molluscan visual system. /. Gen. Physiol., 61 : 444-461. ALKON, D. L., 1974a. Associative training of Hcrmisscnda. J. Gen. Physiol., 64: 70-84. ALKON, D. L., 1974b. Sensory interactions in the nudibranch mollusc Hcrmisscnda crassi- cornis. Fed. Proc., 33 : 1083-1090. ALKON, D. L., 1975. Neural correlates of associative training in Hcrmisscnda. J. Gen. Phys- iol., 65 : 46-56. ALKON, D. L., 1976a. The economy of photoreceptor function in a primitive nervous system. Pp. 410-426 in F. Zettler and R. Weiler, Eds., Neural Principles in Vision, Springer Verlag, New York. ALKON, D. L., 1976b. Neural modification by paired sensory stimuli. /. Gen. Phvsiol., 68: 341-358. ALKON, D. L., 1979. Voltage-dependent calcium and potassium ion conductances : A contingency mechanism for an associative learning model. Science, 205 : 810-816. ALKON, D. L., 1980. Membrane depolarization accumulates during acquisition of an associ- ative behavioral change. Science, 209 : in press. ALKON, D. L., T. AKAIKE, AND J. F. HARRIGAN, 1978. Interaction of chemosensory, visual and statocyst pathways in Hcrmisscnda. J. Gen. Physiol., 71 : 177-194. ALKON, D. L., AND A. BAK, 1973. Hair cell generator potentials. /. Gen. Physiol., 61 : 619-637. ALKON, D. L., AND M. G. F. FUORTES, 1972. Responses of photoreceptors in Hermissenda. J. Gen. Physiol., 60: 631-649. GASTROPOD MODEL OF LEARNING 553 ALKON, D. L., AND Y. GROSSMAN, 1978a. Evidence for non-synaptic neuronal interaction. J. Neurophysiol, 41 : 640-653. ALKON, D. L., AND Y. GROSSMAN, 1978b. Long-lasting depolarization and hyperpolarization in eye of Hcnnisscnda. J. Neurophysiol., 41 : 1328-1342. ALTMAN, J. A., AND N. M. TYRER, 1977. The locust wing hinge stretch receptors. II. Varia- tion, alternative pathways and "mistakes" in the central arborizations. /. Comp. Ncur.. 172: 431-440. ARANDA, L. C, AND J. V. Luco, 1969. Further studies on an electrical correlate to learning: experiments in an isolated insect ganglion. Physiol. Bchav., 4: 133-137. BARRETT, E., AND J. BARRETT, 1976. Separation of two voltage-sensitive potassium currents, and demonstration of a tetrodotoxin-resistant calcium current in frog motorneurones. /. Physiol.. (Land.), 255: 737-774. BAUDRY, M., AND G. LYNCH, 1980. Regulation of hippocampal glutamate receptors: Evidence for the involvement of a calcium-activated protease. Proc. Nat I. A cad. Set. U.S.A., 77 : 2298-2302. BAUDRY, M., AND G. LYNCH, 1980. Hypothesis regarding the cellular mechanisms responsible for long-term synaptic potentiation in the hippocampus. Exp. N enrol., 68 : 202-204. BERGER, T. W., AND R. F. THOMPSON, 1978a. Identification of pyramidal cells as the critical elements in hippocampal neuronal plasticity during learning. Proc. Nat. Acad. Sci. U.S.A., 75: 1572-1576. BERGER, T. W., AND R. F. THOMSON, 1978b. Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. I. The Hippocampi. Brain Res., 145 : 323-346. BERGER, T. W., AND R. F. THOMPSON, 1978c. Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. II. Septum and mammillary bodies. Brain Res., 156: 293-314. BERRIDGE, M. J., 1976. Interaction of cyclic nucleotides and calcium in the control of cellular activity. Adv. Cyclic Nucleotide Res., 6: 1-98. BLAKEMORE, C., J. NACHIMIAS, AND P. SUTTOX, 1970. The perceived spatial frequency shift: evidence for frequency-selective neurons in the human brain. /. Physiol. (Loud.), 210: 727-750. BLANKENSHIP, J. E., 1968. Action of tetrodotoxin on spinal motoneurons of the cat. /. Neurophysiol, 31 : 186-194. BODIAN, D., 1936. A new method for staining nerve fibers and nerve endings in mounted paraffin sections. Anat. Rec., 65: 89-97. BORLE, A. B., 1974. Cyclic AMP stimulation of calcium efflux from kidney, liver and heart mitochondria. /. Membr. Biol., 16: 221-236. BORON, W., J. RUSSELL, M. BRODWICK, D. KEIFER, AND A. Roos, 1978. Influence of cyclic AMP on intracellular pH regulation and chloride fluxes in barnacle muscle fibers. Nature, 276: 511-513. BRUNER, J., AND D. KENNEDY, 1970. Habituation : occurrence at a neuromuscular junction. Science, 169: 92-94. BRUNER, J., AND L. TAUC, 1964. Les modifications de 1'activite synaptique au cours de 1'habitua- tion chez I'Aplysie. J. Physio!. (Paris), 56: 306-307. BRUNER, J., AND L. TAUC, 1965. La plasticite synaptique impliquee dans le processus d'habitua- tion chez I'Aplysie. J. Physiol. (Paris), 57: 230-231. BRUNER, J., AND L. TAUC, 1966a. Habituation at the synaptic level in Apl\sia. Nature, 210: 37-39. BRUNER, J., AND L. TAUC, 1966b. Long-lasting phenomena in the molluskan nervous system. Symp. Soc. for Exp. Biol., 20 : 457-475. BUCHWALD, J. S., E. S. HALAS, AND S. SCHRAMM, 1966. Changes in cortical and subcortical unit activity during conditioning in chronic cats. Physiol. Bchav., I: 11-22. BULLOCK, T. H., AND G. C. QUARTON, 1966. Simple systems for the study of learning mecha- nisms. Neurosci. Res. Prog. Bull.. 4: 203-327. BURES, J., AND O. BURESOVA, 1967. Plastic changes of unit activity based on reinforcing prop- erties of extracellular stimulation of single neurons. J. Neurophysiol., 30: 98-113. BURROWS, M., 1973. Physiological and morphological properties of the metathoracic common inhibitory neuron of the locust. /. Coinp. Physiol., 82 : 59-78. BURROWS, M., AND G. A. HORRIDGE, 1974. The organization of inputs to motoneurons of the locust metathoracic leg. Philos. Trans. R. Soc. Land. B. Biol., 269: 49-94. 554 DANIEL L. ALKON BURROWS, M., AND G. HOYLE, 1973. Neural mechanisms underlying behavior in the locust Schistoccra gregaria. III. Topography of limb motoneurons in the metathoracic gang- lion. /. Neurobiol., 4: 167-186. BYRNE, J. H., 1980a. Analysis of ionic conductance mechanisms in motor cells mediating inking behavior in Aplysia calif ornic a. J. Neurophysiol, 43 : 630-650. BYRNE, J. H., 1980b. Quantitative aspects of ionic conductance mechanisms contributing to firing pattern of motor cells mediating inking behavior in Aplysia californica. J. Neurophysiol., 43: 651-668. CAREW, T. J., AND E. R. KANDEL, 1973. Acquisition and retention of long-term habituation in Aplysia: correlation of behavioral and cellular process. Science, 182: 1158-1160. CAREW, T. J., H. M. PINSKER, AND E. R. KANDEL, 1972. Long-term habituation of a de- fensive withdrawal reflex in Aplysia. Science, 175: 451-454. CAREW, T. J., V. F. CASTELLUCCI, AND E. R. KANDEL, 1971. An analysis of dishabituation and sensitization of the gill-withdrawal reflex in Aplysia. Int. J. Neurosci., 2: 79-98. CASTELLUCCI, V. F., H. M. PINSKER, I. KUPFERMAN, AND E. R. KANDEL, 1970. Neuronal mechanisms of habituation and dishabituation of the gill-withdrawal reflex in Aplysia. Science, 167: 1745-1748. COHEN, D. H., 1969. Development of a vertebrate experimental model for cellular neuro- physiologic studies of learning. Conditional Reflex, 4 : 61-80. COHEN, D. H., AND R. L. MACDONALD, 1976. Involvement of the avian hypothalamus in defen- sively conditioned heart rate change. /. Comp. Neurol., 167: 465-480. COHEN, M. J., AND J. W. JACKLET, 1967. The functional organization of motor neurons in an insect ganglion. Philos. Trans. R. Soc. Land. B. Biol., 252: 563-571. CONNOR, J. A., 1979. Calcium current in molluscan neurones: measurement under conditions which maximize its visibility. /. Physiol. (Lond.), 286: 41-60. CONNOR, J. A., AND C. F. STEVENS, 1971. Voltage-clamp studies of a transient outward mem- brane current in gastropod neural somata. /. Physiol. (Lond.}, 213: 21-30. CROW, T. J., AND D. L. ALKON, 1978a. Neural correlates of an associative behavioral change in Hennissenda crassicornis. Soc. Neurosci., 4: 191. CROW, T. J., AND D. L. ALKON, 1978b. Retention of an associative behavioral change in Hennissenda. Science, 201 : 1239-1241. CROW, T. J., AND D. L. ALKON, 1980a. Associative behavioral modification in Hennissenda : cellular correlates. Science, 209 : 412-414. CROW, T., AND D. L. ALKON, 1980b. Neural correlates of a long-term associative behavioral modification in Hennissenda crassicornis. Fed. Proc. 39 : 596. CROW, T., E. HELDMAN, V. HACOPIAN, R. ENOS, AND D. L. ALKON, 1979. Ultrastructure of photoreceptors in the eye of Hennissenda labeled with intracellular injections of horseradish peroxidase. /. Neurocytol. 8: 181-195. CROW, T., AND N. OFFENBACH, 1979. Response specificity following behavioral training in the nudibranch mollusk Hennissenda crassicornis. Biol. Bull., 157: 364. DAVIS, W. J., G. J. MPITSOS, AND J. M. PINNEO, 1974. The behavioral hierarchy of the mol- lusk Pleurobranchaca. I. The dominant position of feeding behavior. /. Comp. Physiol. 90 : 207-224. DAVIS, W. J., J. VILLET, D. LEE, M. RIGLER, R. GILLETTE, AND E. PRINCE, 1980. Selective and differential avoidance learning in the feeding and withdrawal behavior of Pleuro- branchaea californica. J. Comp. Physiol., 138: 157-165. DEFELICE, L. J., AND D. L. ALKON, 1977. Voltage noise from hair cells during mechanical stimulation. Nature, 269: 613-615. DETWILER, P. B., AND D. L. ALKON, 1973. Hair cell interactions in the statocyst of Hennis- senda. J. Gen. Physiol., 62: 618-642. DISTERHOFT, J. F., J. NuRNBERGER, AND W. C. CORNING, 1968. "P-R" differences in intact cockroaches as a function of testing interval. Psychonomic Science, Section on Animal and Physiological Psychology, 12 : 205-206. DISTERHOFT, J. F., R. HAGGERTY, AND W. C. CORNING, 1971. An analysis of leg position learning in the cockroach yoked control. Physiol. Bchav., 7 : 359-362. DOTY, R. W., 1961. The role of subcortical structures in conditioned reflexes. Ann. N.Y. Acad. Sci., 92 : 939-945. DOWING, J. E., AND B. B. BOYCOTT, 1965. Neural connections of the retina: fine structure of the inner plexiform layer. Cold Spr. Harb. Symp. Quant. Biol., 30: 393-402. GASTROPOD MODEL OF LEARNING 555 DUNWIDDIE, T. V., AND G. LYNCH, 1979. The relationship between extracellular calcium con- centration and the induction of hippocampal long-term potentiation. Brain Res., 169 : 103-110. ECKERT, R., AND H. D. Lux, 1976. A voltage-sensitive persistent calcium conductance in neuronal somata of Helix. J. Physiol. (Land.) 254: 129-151. EISENSTEIN, E. G., 1968. Assessing the influence of pharmacological agents on shock avoidance learning in simpler systems. Brain Res. 11: 471-480. EISENSTEIN, E. M., 1970a. A comparison of activity and position response measures of avoid- ance learning in the cockroach, P. aincricana. Brain Res., 21 : 143-147. EISENSTEIN, E. M., 1970b. The retention of shock avoidance learning in the cockroach, P. atncricana. Brain Res. 21 : 148-150. EISENSTEIN, E. M., AND M. J. COHEN, 1965. Learning in an isolated prothoracic insect gang- lion. Anim. Bchav. 13: 104-108. EISENSTEIN, E. M., G. KEMENY, I. D. SKCRNICK, AND J. T. HARRIS, 1972. Transfer of informa- tion within the cockroach central nervous system during shock avoidance learning : is there a nerve impulse code(s) for learned information? Int. J. Netirosci., 3: 93-98. EISENSTEIN, E. M., AND G. H. KRASILOVSKY, 1967. Studies of learning in isolated insect ganglia. Pp. 329-332 in C.A.G. Wiersma, Ed., Invertebrate Nervous Systems, Chicago. The University of Chicago Press. EISENSTEIN, E. M., AND B. PERETZ, 1973. Comparative aspects of habituation in invertebrates. Pp. 1-34 in H. Peeke and M. J. Herz, Eds., Habituation, Vol. 2, Electrophysiological Substrates. EPSTEIN, R., AND L. TAUC, 1970. Heterosynaptic facilitation and posttetanic potentiation in Aplysia nervous system. /. Physiol., 209 : 1-23. EVARTS, E. V., 1973. Motor cortex reflexes associated with learned movement. Science, 179: 501-503. FAIN, G. L., AND F. N. QUANDT, 1977. Calcium-dependent regenerative responses in rods. Nature, 269 : 707-709. FARBER, D. B., M. B. BROWN, AND R. N. LOLLEY, 1979. Cyclic nucleotide dependent protein kinase and the phosphorylation of endogenous proteins of retinal rod outer segments. Biochemistry, 18 : 370-378. FARLEY, J., AND D. L. ALKON, 1980. Neural organization predicts stimulus specificity for a retained associative behavioral change. Soc. N curosci., 6 : 786. FETZ, E. E., AND D. B. FINOCCHIO, 1971. Operant conditioning of specific patterns of neural and muscular activity. Science, 174: 431-435. FREDMAN, S. M., AND B. JAHAN-PARWAR, 1980. Processing of chemosensory and mechanosensory information in identifiable Aplysia neurons. Comp. Biochem. Physiol., 66A : 25-34. FUSTER, J. M., AND G. E. ALEXANDER, 1971. Neuron activity related to short-term memory. Science, 173: 652-654. GELPERIN, A., 1975. Rapid food-aversion learning by a terrestrial mollusk. Science. 189: 567-570. GETCHELL, T. V., 1977. Analysis of intracellular recordings from salamander olfactory epi- thelium. Brain Res., 123: 275-286. GILLETTE, R., AND W. J. DAVIS, 1977. The role of the metacerebral giant neuron in the feeding behavior of Plcurobranchaea. J. C<»np. Physiol., 116: 129-159. GILLETTE, R., M. V. GILLETTE, AND W. J. DAVIS, 1980. Action-potential broadening and en- dogenously sustained bursting are substrates of command ability in a feeding neuron of Pleurobranchaea. J. Neurophysiol., 43 : 669-685. GOODMAN, C. S., 1978. Isogenic grasshoppers: genetic variability in the morphology of identi- fied neurons. /. Comp. N enrol, 182 : 681-706. GORMEZANO, L, AND J. W. MOORE, 1969. History and method. Pp. 121-203 in M. M. Marx, Ed., Learning Processes, Macmillan, Toronto. GORMEZANO, L, N. SCHNEIDERMAN, E. DEAUX, AND I. FUENTES, 1962. Nictitating membrane: classical conditioning and extinction in the albino rabbit. Science 138 : 33-34. GROSSMAN, Y., D. L. ALKON, AND E. HELDMAN, 1979. A common origin of voltage noise and generator potentials in statocyst hair cells. /. Gen. Physiol., 73: 23-48. HAGIWARA, S., J. FUKUDA, AND D. C. EATON, 1974. Membrane currents carried by Ca, Sr, and Ba in barnacle muscle fiber during voltage-clamp. J. Gen. Physiol., 63: 564-578. 556 DANIEL L. ALKON HARRIS, J. D., 1943. Habituatory response decrement in the intact organism. Psvchol. Bull., 40 : 385-422. HELDMAN, E. Y. GROSSMAN, T. P. JERUSSI, AND D. L. ALKON, 1979. Cholinergic features of photoreceptor synapses in Hermissenda. J. NcurophysioL, 42: 153-165. HIRSCH, H. V. B., AND D. X. SPINEI.LI, 1970. Visual experience modifies distribution of hori- zontally and vertically oriented receptive fields in cats. Science. 168: 769-871. HIRSCH, H. V. B., AND D. N. SPINELLI, 1971. Modification of the distribution of receptive field orientation during development. Exp. Brain Res., 13: 509-527. HORRIDGE, G. A., 1959. Analysis of the rapid responses of Nereis and Haruwthoe (Annelida). Proc. R. Soc. Land. B Bint., 150 : 245-262. HORRIDGE, G. A., 1962a. Learning of leg position by the ventral nerve cord in headless insects. Proc. R. Soc. Land. B Biol., 157: 33-52. HORRIDGE, G. A., 1962b. Learning of leg position by headless insects. Nature, 193: 697-698. HORRIDGE, G. A., 1965. The electrophysiological approach to learning in isolated ganglia. Anim. Behav., 12, 163-182. HOTSON, J. R., AND D. A. PRINCE, 1980. A calcium-activated hyperpolarization follows repeti- tive firing in hippocampal neurons. /. NcurophysioL, 43: 409^119. HOYLE, G., 1964. Exploration of neuronal mechanisms underlying behavior in insects. Pp. 346-376 in R. F. Reiss, Ed., Neural Theory and Modeling, Stanford University Press, Stanford. HOYLE, G., 1965. Neurophysiological studies of "learning" in headless insects. Pp. 203-232 in J. E. Treherne and J. W. L. Beament, Eds., The Physiology of the Insect Central Nervous System, Academic Press, New York. HOYLE, G., 1970. Cellular mechanisms underlying behavior-neuroethology. Advances in In- sect Physiology 7 : 349-444. HOYLE, G., AND M. BURROWS, 1973a. Neural mechanisms underlying behavior in the locust Shistocerca gregaria. I. Physiology of identified motoneurons in the metathoracic ganglion. /. Neurobiol. 4: 3-41. HOYLE, G., AND M. BURROWS, 1973b. Neural mechanisms underlying behavior in the locust Shistocerca gregaria. II. Integrative activity in metathoracic neurons. /. Neurobiol., 4 : 43-67. HUBEL, D. H., AND T. N. WIESEL, 1970. The period of susceptibility to the physiological ef- fects of unilateral eye closure in kittens. /. Physiol., 206 : 419-436. HUGHES, G. M., AND L. TAUC, 1963. An electrophysiological study of the anatomical relations of two giant nerve cells in Aplysia dcpilans. J. Exp. Biol., 40 : 469-486. HUMPHREY, G., 1930. Le Chatelier's role, and the problem of habituation and dehabituation in Helix albolabris. Psychol. Forsch., 13: 113-127. JACKLET, J. W., 1977. Neuronal circadian rhythm: phase shifting by a protein synthesis in- hibitor. Science, 198: 69-71. JACKLET, J. W., AND J. RINE, 1977. Facilitation at neuromuscular junctions: contribution to habituation and dishabituation of the Aplvsia gill withdrawal reflex. Proc. Natl. Acad. Sci. U.S.A., 74: 1267-1271. JASPER, H. H., G. RICCI, AND B. DOANE, 1960. Microelectrode analysis of cortical cell dis- charge during avoidance conditioning in the monkey. EEC Clin. Neurophysiol. (SuppL), 13: 137-155. JENNINGS, K. R., L. K. KACZMAREK, AND F. STRUMWASSEK, 1979. Protein phosphorylation during afterdischarge of the neuroendocrine bag cells in Aplysia. Soc. Neurosci., 5 : 249. JERUSSI, T. P., AND D. L. ALKON, 1980. Light responses of motor units in a marine mollusc. Fed. Proc. 39 : 596. JOHN, E. R., 1967. Electrophysiological studies of conditioning. Pp. 690-704, in G. C. Quarton, T. Melnechuk, and F. O. Schmitt, Eds., The Neuroscicnccs: A Study Pro- gram, Rockefeller University Press, New York. JOHN, E. R., F. BARTLETT, M. SHIMOKOCHI, AND D. KLEINMAN, 1973. Neural readout from memory. /. Neurophysiol., 36 : 893-924. JOHN, E. R., AND K. F. KILLAM, 1959. Electrophysiological correlates of avoidance condition- ing in the cat. /. Pharmacol. Exper. Ther., 125 : 252-274. KACZMAREK, L. K., K. R. JENNINGS, AND F. STRUMWASSER, 1978. Neurotransmitter modu- GASTROPOD MODEL OF LEARNING 557 lation of phosphodiesterase inhibitor effects and cyclic AMP correlates of afterdischarge in peptidergic neurites. Proc. Nail. .lead. Sci. U.S.A.. 75: 5200-5204. KACZMAREK, L. K., K. R. JENNINGS, AND F. STRUMWASSER, 1979. Cyclic AMP analog gen- erates afterdischarge in Aplysia bag cell neurons. Soc. Neurosci., 5 : 249. KANDEL, E. R., AND L. TAUC, 1964. Mechanisms of prolonged heterosynaptic facilitation. Nature 202 : 145-147. KANDEL, E. R., AND I. TAUC, 1965a. Heterosynaptic facilitation in neurons of the abdominal ganglion of Aplysia dcpilans. J. Physio!., 181 : 1-27. KANDEL, E. R., AND L. TAUC, 1965b. Mechanisms of heterosynaptic facilitation in the giant cell of the abdominal ganglion of Aplysia dcpilans. J. Physio!., 181 : 28-47. KATER, S. B., 1974. Feeding in Hclisoma trivolvis: the morphological and physiological basis of a fixed action pattern. Am. Zoo!., 14: 1017-1036. KEENE, J. J., 1973. Reward-associated inhibition and pain-associated excitation lasting seconds in single intralaminar thalamic units. Brain Res., 64: 211-224. KRASNE, F. B., AND K. S. WOODSMALL, 1969. Waning of the crayfish escape response as a re- sult of repeated stimulation. Anim. Bchav., 17: 416-424. KRUEGER, B. K., J. FORN, AND P. GREENGARD, 1977. Depolarization-induced phosphorylation of specific proteins, mediated by calcium ion influx, in rat brain synaptosomes. /. Biol. Chem.,252: 2764-2773. KUPFERMAN, L, V. CASTELLUCCI, H. PINSKER, AND E. R. KANDEL, 1970. Neuronal correlates of habituation and dishabituation of the gill-withdrawal reflex of Apl\sia. Science, 167: 1743-1745. LASANSKY, A., 1971. Synaptic organization of cone cells in the turtle retina. Phil. Trans. R. Soc. Land. B Biol., 252: 265-281. LEDERHENDLER, I. I., E. S. BARNES, AND D. L. ALKON, 1980. Complex responses to light of the nudibranch Hcrurissenda crassicornis. Bchav. Neural Biol., 28: 218-230. LEE, R. M., AND R. J. LIEGEOIS, 1974. Motor and sensory mechanisms of feeding in Pleuro- branchaca. J. Ncurobiol., 5 : 545-564. LIPTON, S. A., H. RASMUSSEN, AND J. D. DOWLING, 1977. Electrical and adaptive properties of rod photoreceptors in Bujo viarimis. J. Gen. Physiol., 70: 771-791. LISMAN, J. E., AND J. E. BROWN, 1972. The effects of intracellular iontophoretic injection of calcium and sodium ions on the light responses of Limulus ventral photoreceptors. /. Gen. Physiol., 59 : 701-719. LIVINGSTON, R. B., 1966. Brain mechanisms in conditioning and learning. Neurosci. Res. Prog. Bull., 4 : 349-354. LLINAS, R., R. BAKER, AND W. PRECT, 1972. Blockage of inhibition by ammonium acetate action on chloride pump in cat trochlear motorneurons. /. Neurophysiol. 37 : 522-532. LLINAS, R., AND T. HESS, 1976. Tetrodotoxin-resistant dendridic spikes in avian Purkinje cells. Proc. Natl. Acad. Sci. U.S.A. 73: 2520-2523. LUBBOCK, J., 1958. On the digestive and nervous systems of Coccus hespcridum. Proc. R. Soc., 9 : 480-486. LUKOWIAK, K., 1979. Food increases CNS suppressive control over gill reflex behaviors in Aplysia. Soc. Neurosci., S : 252. LUKOWIAK, K., AND J. W. JACKLET, 1972. Habituation and dishabituation : interactions be- tween peripheral and central nervous systems in Aplysia. Science, 178: 1306-1308. LUKOWIAK, K., AND J. JACKLET, 1975. Habituation and dishabituation mediated by the periph- eral and central neural circuits of the siphon of Aplysia. J. Ncurobiol., 6: 103-200. MACAGNO, E. R., V. LOPRESTI, AND C. LEVINTHAL, 1973. Structure and development of neuronal connections in isogenic organisms : variations and similarities in the optic system of Daphnia magna. Proc. Nat. Acad. Sci. U.S.A., 70: 57-61. MAGELBY, K. L., AND C. F. STEVENS, 1972. A quantitative description of end-plate currents. /. Physiol. 223: 173-197. MALAISSE, W. J., 1973. The stimulus secretion coupling of glucose induced insulin release. Part 12. Effects of diazoxide and gliclazide upon calcium-45 efflux from perfused islets. Diabetologica, 9 : 1-5. MILLER, N. E., 1967. Certain facts of learning relevant for its physical basis. Pp. 643-652 in G. C. Quarton, T. Melnechuk, F. O. Schmitt, Eds., The Ncurosciences: A Study Pro- grain, Vol. 2. New York, Rockefeller University Press. 558 DANIEL L. ALKON MPITSOS, G. J., AND S. D. COLLINS, 1975. Learning: rapid aversive conditioning in the gas- tropod mollusk Plcurobranchaca. Science, 188 : 954-956. NEARY, J. T., T. J. CROW, AND D. L. ALKON, 1980. Biochemical correlates of a long-term associative behavioral modification in Hermissenda: changes in the levels of specific phosphoprotein bands following a conditioning procedure. Soc. Neurosci., 6: 591. NEHER, E., AND C. F. STEVENS, 1977. Conductance fluctuations and ionic pores in membranes. Ann. Rev. Biophys. Bioeng., 6: 345-381. NIENHUIS, R., AND J. OLDS, 1978. Changes in unit responses to tones after food reinforcement in the auditory pathway of the rat: intertrial arousal. Exp. Neural., 59: 229-242. NORMAN, R. J., J. S. BUCHWALD, AND J. R. VILLABLANCA, 1977. Classical conditioning with auditory discrimination of the eye blink in decerebrate cats. Science, 196: 551-553. OLDS, J., J. F. DISTERHOFT, M. SEGAL, C. L. KORNBLITH, AND R. HIRSCH, 1972. Learning cen- ters of rat brain mapped by measuring latencies of conditional unit responses. /. Ncuro- physiol, 35 : 202-219. OLDS, J., AND T. HIRAMO, 1969. Conditional responses of hippocampal and other neurons. EEC Clin. Neurophysiol. 26 : 159-166. PAVLOV, I. P. S. P., 1927. Conditioned Reflexes. Translated by G. V. Anrep. Oxford Uni- versity Press, London. PEARSON, K. G., 1972. Central programming and reflex control of walking in the cockroach. /. Exp. Biol., 56: 173-193. PEARSON, K. G., AND C. S. GOODMAN, 1979. Correlation of variability in structure with vari- ability in synaptic connections of an identified interneuron in locusts. J. Coinp. Ncurol., 184 : 141-166. PERETZ, B., 1969. Central neuron initiation of periodic gill movements. Science, 116: 1167-1172. PERETZ, B., 1970. Habituation and dishabituation in the absence of a central nervous system. Science, 169: 373-381. PERETZ, B., AND J. ESTES, 1974. Histology and histochemistry of the peripheral neural plexus in the Aplysia gill. /. Neurobiol. 5: 3-19. PERETZ, B., AND D. B. HOWIESON, 1973. Central influence on peripherally mediated habituation of an Aplysia gill withdrawal response. /. Cotnp. Physiol., 84: 1-18. PERETZ, B., J. W. JACKLET, AND K. LUKOWIAK, 1976. Habituation of reflexes in Aplysia : Contribution of the peripheral and central nervous systems. Science, 191 : 396-399. PERETZ, B., AND R. MOLLER, 1974. Control of habituation of the withdrawal reflex by the gill ganglion in Aplysia. J. Neurobiol., 5: 191-212. PITMAN, R. M., C. D. TWEEDLE, AND M. J. COHEN, 1973. The form of nerve cells: determi- nation by cobalt impregnation. Pp. 83-97 in S. B. Kater and C. Nicholson, Eds., Intraccllular Staining in Neurobiology. Springer-Verlag, New York. PRIBRAM, K. H., D. N. SPINELLI, AND M. C. KORNBACK, 1967. Electrocortical correlates of stimulus response and reinforcement. Science, 157 : 94-96. PRITCHATT, D., 1968. Avoidance of electric shock by the cockroach Periplancta aincricana. Anim. Behav., 16: 178-185. PURPURA, D. P., 1967. Comparative physiology of dendrites. Pp. 372-393 in G. C. Quarton, T. Melnechuk, and F. O. Schmitt, Eds., The Neuroscicnces: A Study Program. Rockefeller University Press, New York. RASMUSSEN, H., 1970. Cell communication, calcium ion, and cyclic adenosine monophosphate. Science, 170: 404-412. RASMUSSEN, H., AND D. B. P. GOODMAN, 1975. Calcium and cAMP as interrelated intra- cellular messengers. Ann. N.Y. Acad. Sci., 253: 789-796. RASMUSSEN, H., P. JENSEN, W. LAKE, N. FRIEDMAN, AND D. B. P. GOODMAN, 1975. Cyclic nucleotides and cellular calcium metabolism. Adv. Cyclic Nucleotide Res. 5 : 375-394. REINGOLD, S. C., AND A. GELPERIN, 1979. Feeding modulation in Limax: behavior and physi- ology. Soc. Neurosci., 5 : 259. RESCORLA, R. A., 1967. Pavlovian conditioning and its proper control procedures. Psvchol. Rev., 74 : 71-80. RUNION, H. I., AND P. N. R. USHERWOOD, 1968. Tarsal receptors and leg reflexes in the locust and grasshopper. /. Exp. Bio!., 49 : 421-436. SCHULMAN, H., AND P. GREENGARD, 1978. Ca2*-dependent protein phosphorylation system in membranes from various tissues, and its activation by "calcium-dependent regulator." Proc. Nat. Acad. Sci. U.S.A., 75: 5432-5436. GASTROPOD MODEL OF LEARNING 559 SHEPHERD, G., 1979. The Synoptic Organization of the Brain. Oxford University Press, New York. 436p. SHICHI, H., AND R. L. SOMERS, 1978. Light-dependent phosphorylation of rhodopsin. /. Biol. Chan.. 253 : 7040. SHOUKIMAS, J. J., AND D. L. AI.KON, 1980. Voltage-dependent, early outward current in :i photoreceptor of Hermissenda craxsicornis. Sue. Neurosci., 6: 17. SPRAY, D. C, M. E. SPIRA, AND M. V. L. BENNETT, 1980. Peripheral fields and branching pat- terns of buccal mechanosensory neurons in the opisthobranch mollusc, Xaranax inennis. Brain Res., 182 : 253-270. STENSAAS, L. J., S. S. STENSAAS, AND O. TRUJILLO-CENOZ, 1969. Some morphological aspects of the visual system of Hermissenda crassicornis (Mollusca: Nudibranchia). /. Ultra- struct. Res., 27: 510. STINNAKRE, J., AND L. TAUC, 1969. Central neuronal response to the activation of osmorecep- tors in the osphradium of Aplysia. J. Exp. Biol. 51 : 347-361. STOMMEL, E. W., R. E. STEPHENS, AND D. L. ALKON, 1980. Motile statocyst cilia transmit rather than transduce mechanical stimuli. /. Cell. Biol., 84: in press. STRETTON, A. O. W., AND E. KRAVITZ, 1973. Intracellular dye injection: the selection of Procion yellow and its application in preliminary studies of neuronal geometry of the lobster nervous system. Pp. 21-40 in S. B. Kater and C. Nicholson, Eds., Intracellular Staining in Ncurobiology. Springer-Verlag, New York. STRUMWASSER, F., 1965. The demonstration and manipulation of a circadian rhythm in a single neuron. Pp. 442-462 in J. A. Schoff, Ed., Circadian Clocks. North Holland Publishing Co., Amsterdam. STRUMWASSER, F., J. W. JACKLET, AND R. B. ALVAREZ, 1969. A seasonal rhythm in the neural extract induction of behavioral egg-laying in Aplysia. Comp. Biochem. Physiol. 29 : 197-206. TABATA, M., AND D. L. ALKON, 1979. Control of synaptic feedback in a molluscan visual system. Invest. Ophthahnol. Vis. Sci., Suppl.: 77. TAUC, L., AND R. EPSTEIN, 1967. Heterosynaptic facilitation as a distinct mechanism in Aplysia. Nature, 214 : 724-725. THOMPSON, E. B., J. H. SCHWARTZ, AND E. R. KANDEL, 1976. A radioautographic analysis in the light and electron microscope of identified Aplysia neurons and their processes after intrasomatic injection of L{3H} fucose. Brain Res., 112: 251-281. THOMPSON, R. F., AND W. A. SPENCER, 1966. Habituation : A model phenomenon for the study of neuronal substrates of behavior. Psycho!. Rev., 73 : 16-43. THOMPSON, W. F., AND G. S. STENT, 1976. Neuronal control of heartbeat in the medicinal leech. II. Intersegmental coordination of heart motor neuron activity by heart inter- neurons. /. Comp. Physiol., Ill : 281-307. THORPE, W. H., 1963. Learning and Instinct in Animals. Harvard University Press, Cam- bridge, MA. TOLEDO- MORRELL, L., T. J. HOEPPNER, AND F. MORRELL, 1979. Conditioned inhibition: Selective response of single units. Science, 204 : 528-530. VORONIN, L. L., 1971. Microelectrode investigations of cellular analogs of learning. Sov. Neural. Psych., 4 : 99-125. VON BAUMGARTEN, R., 1970. Plasticity in the nervous system at the unitary level. Pp. 260-271 in F. O. Schmitt, Ed., The Neuroscicncc: Second Study Program. Rockefeller Uni- versity Press, New York. VON BAUMGARTEN, R., AND B. DJAHANPARWAR, 1967. Time course of repetitive heterosynap- tic facilitation in Aplysia californica. Brain Res., 4: 295-297. VON BAUMGARTEN, R., AND T. HUKUHARA, 1969. The role of interstirnulus interval in heterosynaptic facilitation in Aplysia calijornica. Brain Res., 16: 369-381. WALTERS, E. T., T. J. CAREW, AND E. R. KANDEL, 1979. Classical conditionng in Aplysia cali- fornica. Proc. Nat. Acad. Sci. U.S.A., 76: 6675-6679. WEIGHT, F. F., 1974. Physiological mechanisms of synaptic modulation. Pp. 929-941 in F. O. Schmitt and F. G. Worden, Eds., The Neuroscicnces: Third Study Program. MIT Press, Cambridge, Mass. WIESEL, T. N., AND D. H. HUBEL, 1965. Comparison of the effects of unilateral and bilateral eye closure on cortical unit responses in kittens. /. Neurophysiol., 28: 1029-1040. 560 DANIEL L. ALKON WILSON, D. M., 1965. Proprioceptive leg reflexes in cockroaches. /. Exp. BioL, 43: 397-409. WILSON, D. M., 1966. Insect walk-ing. Ann. Rev. Entomol., 11 : 103-123. WOODRUFF, M. L., P. BOWNDS, S. H. GREEN, J. L. MORRISEY, AND A. SHEDLOVSKY, 1977. cGMP and the in vitro physiology of frog photoreceptor membranes. /. Gen. Physiol., 69 : 667-679. WOODRUFF, M. L., AND S. M. D. BOWNDS, 1979. Amplitude, kinetics and reversibility of a light-induced decrease in guanosine 3',5'-cyclic monophosphate in frog photoreceptor membranes. /. Gen. Physiol.. 73 : 629-653. WOODY, C. D., AND P. BLACK-CLEWORTH, 1973. Differences in excitability of cortical neurons as a function of motor projection in conditioned cats. /. Neurophysiol., 36: 1104-1116. WOODY, C. D., D. CARPENTER, J. D. KNISPEL, T. CROW, AND P. BLACK-CLEWORTH, 1974. Pro- longed increases in resistance of neurons in cat motor cortex following extracellular iontophoretic application of acetylcholine (ACh) and intracellular current injection. Fed. Proc., 33 : 399. WOODY, C. D., AND J. ENGEL, JR., 1972. Changes in unit activity and thresholds to electrical microstimulation at coronal-pericruciate cortex of cat with classical conditioning of different facial movements. /. Neurophysiol., 35: 230-241. WOODY, C. D., J. D. KNISPEL, T. J. CROW, AND P. A. BLACK-CLEWORTH, 1976. Activity and excitability to electrical current of cortical auditory receptive neurons of awake cats as affected by stimulus association. /. Neurophysiol., 39: 1045-1061. WOODY, C. D., AND P. J. YARONSKY, 1972. Conditioned eye blink using electrical stimulation of coronal-precruciate cortex as conditional stimulus. /. Neurophysiol.. 35 : 242-252. WOOLLACOTT, M. H., AND G. HOYLE, 1977. Neural events underlying learning in insects : changes in pacemaker. Proc. R. Soc. Land. B. BioL, 195: 395-415. ZUCKER, R. S., 1972a. Crayfish escape behavior and central synapses. I. Neural circuit ex- citing lateral giant fiber. /. Neurophysiol., 35 : 599-620. ZUCKER, R. S., 1972b. Crayfish escape behavior and central synapses. II. Physiological mechanisms underlying behavioral habituation. /. Neurophysiol., 35: 621-637. ZUCKER, R. S., D. KENNEDY, AND A. I. SELVERSTON, 1971. Neuronal circuit mediating escape responses in crayfish. Science, 173: 645-650. Reference: Biol. Bull., 159: 561-570. (December, 1980) THE FUNCTIONAL SIGNIFICANCE OF THE LUNULES IN THE SAND DOLLAR, MELLITA QU1NQU1ESPERFORATA DAVID E. ALEXANDER AND JOE GHIOLD Department of Zoology, Duke University, Durham, NC 27706; and Itistitut und Museum fur Geologic und Palaontologie, Universitdt Tubingen, Sigwartstrassc 10, 7400 Tubingen 1, West Germany ABSTRACT Experiments indicate that the lunules in Mellita quinquiesperforata function as food gathering devices. Allometric analysis of dead specimens shows that the lunules grow faster than does the body as a whole. This is analogous with the situation in certain fossil brachiopods, in which the lophophore (feeding structure) grows faster than the body. Field experiments involving sand dollars with plugged lunules show that righting and burrowing are not affected. However, animals with plugged lunules cannot feed as well as normal individuals, based on analysis of stomach contents. Thus, the lunules play an important role in feeding, although the exact mechanism has not been conclusively demonstrated. Our study, as well as previous work, suggest that the lunules act as a site for the passage of food particles from the aboral (upper) surface to the oral (lower) surface. Flow tank observations suggest that the lunules may act as devices to bring water, and thus food particles, up from the sand below the animal. In scutellid sand dollars, the fossil record shows an increase in lunule number over time. In Western Atlantic species, there is also an increase in lunule number with decreasing latitude. Per- haps these correlations reflect progressive adaptation by sand dollars to less pro- ductive waters. INTRODUCTION The slots and notches termed "lunules" are structures common to many scutellid sand dollars. Many authors (Berrill and Berrill, 1957; Hyman, 1958; Bell and Frey, 1969; Barnes, 1974) have speculated upon the function of the lunules, but there has been little experimental work. Ikeda (1941) found that specimens of Astriclypeus if their lunules are plugged cannot turn upright when inverted, and burrowing speed is impaired. However, Hyman (1958) and Ghiold (1979) observed that Mellita quinquiesperforata does not use its lunules in bur- rowing, and Weihe and Gray (1968) found that M. quinquiesperforata does not use its lunules in righting when inverted. The suggestion of Berrill and Berrill (1957) that the lunules strengthen the connection between the aboral and oral skele- tal surfaces seems unlikely, as the internal calcareous columns already pro- vide such support (Hyman, 1958). Furthermore, some scutellid sand dollars lack lunules entirely. Many workers have suggested other functions for the lunules in M. quinquiesperforata. These suggestions include removal of feces via upward Received June 8, 1979; accepted September 25, 1980. 561 562 D. E. ALEXANDER AND J. GHIOLD movement of sand through the lunule (Bell and Frey, 1969) ; maintenance of "direct communication with the substrate surface" by buried sand dollars (Bell and Frey, 1969, p. 557) ; and maintenance of a slightly inclined posture by passing sand downward through the anal lunule (Barnes, 1974). Podia and miliary spines, noted to function in food gathering, are concentrated along the inner walls of the lunules of M. quinquiesperjorata. This characteristic led Ghiold (1979) to con- clude that the lunules in this species are involved with feeding. \Ve studied the function of lunules in M. quinquiesperjorata by analysis of lunule growth and development, and by observation and experimentation on living and dead specimens. MATERIALS AND METHODS Lunule growth analysis A growth analysis was conducted to determine possible allometric relationships for the morphology of the lateral, posterior, and anal lunules. Seventy-four dead tests were collected for study along the intertidal margin of Bird Shoal, a small tidal flat in the Beaufort Inlet area of North Carolina. The maximum diameter of the echinoid test, measured along the anterior-posterior axis, was used as the prime index for sand dollar size. The maximum diameters for the sample popula- tion ranged from 98 to 32 mm. Lunule length measurements were plotted against this index to isolate those features displaying allometric growth. (This relation- ship does not account for test diameters smaller than 32 mm and is valid only for those specimens measured.) Measurements were recorded in millimeters with the aid of vernier calipers. A computer program was designed to handle calculations from which a stan- dard logarithmic plot was generated using the power function, y = bxa. The pro- gram plotted the log of lunule size against the log of test size and drew a least- squares regression line through the data points for each graph. The joint mean, "D" (67 mm test diameter) was selected as an objective if somewhat arbitrary estimation of the beginning of the adult population. A dashed line which has a slope, "a" in the equation above, of 1.0 was drawn through "D" and represents a line of isometry (Gould, 1977). The resulting graphs are shown in Figures 1 and 2. The slopes of all regression lines were tested to see whether lunule size changes disproportionately with changing body size. A multivariate regression technique was used to test this; Bonferroni simultaneous confidence intervals (Morrison, 1976) were calculated for the slope of each line with log (test diameter) as the independent variable and log (lunule length) as the dependent variable. This sta- tistical treatment is not unbiased, because least-squares regression assumes that there is no error in the independent variable. Both our variables probably have about the same error, but any error in the x-direction artificially lowers the regres- sion coefficient, making the test of whether the slope is different from 1.0 conserva- tive if the calculated slope is greater than 1.0. Flow tank experiment For this experiment, dead tests of M. quinquiesperjorata were collected by hand near Bird Shoal and preserved specimens were obtained from Carolina Biological Supply, Inc. Specimens were tested in a small Plexiglas flow tank as described LUNULE FUNCTION IN MELLITA 563 8=1.40 ±0.306 (P<0005) N--68 8=1.39*0.300 (P« 0.005) 0.00 1.60 1.80 200 2.20 2.40 2.60 LOG TEST DIAMETER (MM) 000 160 1.80 E.OO ?20 2.40 2.60 LOG TEST DIAMETER (MM) B=1.51± 0.290 (P<0.005) "(JO 160 1.80 2.00 2.20 2.40 2.60 LOG TEST DIAMETER (MM) 8 = 1.53* 0439 (P<0.05) N=56 D.OO 160 180 2.00 2.20 2.40 LOG TEST DIAMETER (MM) 260 FIGURE 1. Length of left (a) and right (b) lateral lunules, and left (c) and right (d) posterior lunules measured against test length. Solid line is least-squares linear regression line; dashed line is a line of isometry (slope = 1.0) through the point D, estimating the be- ginning of the adult population (log (67 mm) or 1.83 on the x-axis). B: slope of regression line ±. Bonferroni confidence interval at the indicated level of significance. N : sample size, which includes some specimens for which one or more lunule could not be measured ; only 43 specimens with all lunules intact were used to calculate the multivariate confidence intervals. by Vogel and LaBarbera (1978). Sand dollars were placed on a bed of sand and covered with a thin layer of sand. After injecting dye (Rotamine B) under the test, observations of flow patterns were made at current speeds ranging from 5-15 cm/sec. Field experiment In situ experiments were carried out on sand dollars along Middle Marsh, near Beaufort, North Carolina. Sand dollars were collected from an abundant popula- tion on a sand bar near Middle Marsh. Only large, healthy-looking sand dollars 564 D. E. ALEXANDER AND J. GHIOLD CXJ _ <=> _ B=1.10±0.139 (DIFFERENCE FROM 1.0 N.S.) N=74 000 160 1.80 2.00 2.20 LOG TEST DIAMETER (MM) 2.40 FIGURE 2. Length of the anal lunule measured against test length. Solid line is least- squares regression line, dashed line is line of isometry (see Fig. 1 for further explanation). Note slope confidence limits include 1.0. were used. In long-term experiments individuals were marked by tying colored threads through the lunules and were positioned near the anchor of a float. By carefully raking the sand around the float anchor where the animals were placed, over 90% of the organisms were recaptured routinely. The burrowing speed of each of five sand dollars was measured two times before and two times after the lunules were plugged. We found that the time it takes an uncovered sand dollar to become completely covered with sand is an excellent indicator of burrowing speed ; thus, the burrowing time of several sand dollars was measured before and after plugging the lunules with soft paraffin. To determine whether the lunules are used in righting inverted sand dollars, 12 sand dollars were collected and half of them had their lunules plugged. The sand dollars were left inverted near the float anchor and observed at the next low tide (for unknown reasons, Mellita qtiinquicsperjorata will right itself in the field but not in the laboratory.) To test the importance of lunules in feeding, 24 sand dollars were collected from the area near the float anchor and tagged. Twelve had their lunules plugged with soft paraffin, and the remaining animals were used as controls. Each control individual was exposed to air for the length of time that it took to plug the lunules of an individual in the other group. All the animals were returned to the sand near the float anchor within a few meters of where they were collected. Two days later, the sand dollars were recaptured. Animals were killed in the field by im- mersion in fresh water so as to minimize movement of the stomach contents. The organisms were then returned to the laboratory and placed in 95% ethanol for 4—6 days to harden their tissues. Stomachs were dissected out, placed on glass slides, and dried overnight at 100 °C. The stomachs were then weighed to the nearest 0.0001 g on a Mettler balance. The laboratory work in this experiment was car- ried out at Duke University Marine Laboratory, which also provided field equipment. LUNULE FUNCTION IN MELLITA 565 CURRENT INDUCED FLOW FIGURE 3. Illustration of flow tank experiment results. Schematic cross-section through the sand dollar shows water moving up through the lunule in response to a current. RESULTS Lunule growth analysis The population measured for lunule growth exhibited an allometric relation for the length of the two lateral and two posterior lunules (Fig. 1). The anal lunule showed a non-significant allometric trend in the same direction (Fig. 2). The slopes of the regression lines are shown with the Bonferroni confidence intervals on the figures. The allometric relationships suggest that the lunules grow more rapidly than does the body. We feel confident that this is a real developmental pattern and not an artifact of adult size distribution, because sand dollar growth appears to be highly indeterminate. There is at least a two- or three-fold range in adult size (personal observation) and there is no reason to believe that large adults are the same age as small adults. There may be no functional significance in the allometric trend alone, but the allometry raises the possibility that the functional importance of the lunules differs as body size changes. Flow tank experiment The flow tank study illustrates an unusual property of the lunule : When an appreciable current (8 cm/sec or more) was passed over the sand dollar, water from the sand beneath the lunule was drawn up through the lunule (Fig. 3). This was probably due to a combination of Bernoulli effect and viscous entrainment, what Vogel (1977) has called "induced flow." Although the process occurred to a small extent at very low current speeds, at higher speeds — above about 8 cm/sec — a significant stream of dye appeared in a few seconds. Field experiments We found strikingly little difference in the burrowing speed of five sand dollars before and after plugging their lunules: The mean time for animals with open lunules was 230 sec (s.e. ± 10.6) ; for the animals with plugged lunules, it was 231 sec (s.e. ± 18.0). Also, the righting ability of sand dollars with plugged lunules did not seem to surfer : of the six sand dollars with plugged lunules and six with open lunules that were inverted in the field, all 12 were right-side-up by the next low tide. Data from the feeding experiment show that for a given body size, sand dollars with plugged lunules had lighter stomachs than sand dollars with normal lunules (Fig. 4). Sand dollars have flat, translucent stomachs, and visual inspection ex- plained the weight difference: 11 of the 12 control animals had stomachs approxi- mately 1/4 to 3/4 full. In contrast, eight animals with plugged lunules had 566 D. E. ALEXANDER AND J. GHIOLD completely empty stomachs, and the two others recaptured had far less in their stomachs than any of the controls. Unexpectedly, paraffin from one lunule of a plugged specimen was missing upon recapture. When the stomach centents of this specimen were weighed, they clearly fell in the range of sand dollars with open lunules (Fig. 4). This strongly suggests that the paraffin itself has no damaging effects on sand dollar feeding behavior, other than to stopper the lunules. DISCUSSION Our burrowing data show that Mellita quinqnicspcrforata does not use its lunules in burrowing. Unlike Ikeda's (1941) observations, our righting experiment clearly shows that open lunules are not necessary for righting. Unfortunately, we could not watch the sand dollars over a whole tidal cycle in the field, so we do not know if plugging the lunules slows righting. Our evidence, although indirect, strongly suggests that the lunules of M. quinquiespcrjorata are involved in the food gathering process and apparently ex- pedite the transfer of food particles to the mouth. Our observations on the lunules are at variance with the observations of Bell and Frey (1969) as well as Ikeda. Bell and Frey suggested that the lunules some- how and for some unknown reason function to keep the echinoid in contact with the surface of the sand, based on small pits in the sand over the lunules. We ob- served these pits commonly over animals beached by the falling tide, but very rarely in normal burrowing animals. Also, Bell and Frey's description of sand dollars pushing sand up through the large anal lunule and leaving a well-defined cylinder of sand behind the organism is unlike any behavior Hyman (1958), Weihe and Gray (1968), or we observed. From observations of specimens in the field and in the laboratory, we noticed that sand dollars occasionally pump with the anal papilla just enough to clear a small portion of the anal lunule of sand and eject a cloud of flocculent fecal matter ; at no time did we observe a significant amount of sand pass- ing up through any of the lunules. In fact, as the sand dollars burrow, they char- acteristically raise the spines around the lunules into a fence-like structure so that sand must pile up until it is high enough for some grains to fall over into the lunule. Field experiments in our study strongly suggest that the lunules are involved in feeding. There are several ways the lunules may function as feeding devices. They 0.0 70-- "5-060 1.050 O ^040 0-030 |.020 ^010 .000 > = plugged lunules 00 1.0 7.0 8.0 9.0 100 11.0 MAXIMUM BODY WIDTH (cm) FIGURE 4. Results of the feeding experiment show that for a given body size, animals with plugged lunules (open circles) have lighter stomachs than control animals (closed circles). A sand dollar with only four of its five lunules plugged is indicated by the open square. Lines were obtained by linear regression. LUNULE FUNCTION IN MELLITA 567 may be sorting and collecting areas for food particles trapped on the aboral (upper) surface (Ghiold, 1979) and may function as an additional "edge" to pass particles from the aboral to the oral side as observed in M. scxicspcrjorata by Goodbody (1960). Another possibility is suggested by our flow tank study : The sand dollars may be using induced flow (Vogel, 1977) to draw water and small particles up from the underlying sand. Presumably, the water then passes through the lunule, where spines and podia trap potential food particles. As it is known that water flows down into and up out of ripples in a sand or gravel bed under a current (Webb and Theodor, 1972), even a sand dollar buried too deeply to take advantage of induced flow could passively exploit the movement of water through the sub- strate in the manner described for induced flow. The allometric growth of lunules also suggests that lunules are involved in feeding. Our data show that lunules grow faster in length than does the animal's body. Ghiold (1979) has demonstrated that the spines of the body and lunule margins are well suited for food collecting. Seilacher (1979) mentioned that lunules may result in an allometric increase in the spine-fringed margin or edge, and he commented that the margin must have an important food gathering role. Seilacher also noted that some sand dollars seem to show a negative allometry of the lunule width. If the lunule is approximated by a long, narrow rectangle, ele- mentary geometry shows that with negative allometry of the width, positive al- lometry of the length can still produce an allometric increase in perimeter. As the perimeter of the lunule seems to be the lunule's important food gathering char- acteristic, any negative allometry of the width is probably unimportant relative to the positive allometry of the length. The allometric increase in perimeter is im- portant because the body increases considerably in weight for a given increase in length, although sand dollars are so nearly two-dimensional that the square : cube relationship is not a good approximation. But even considering the extreme case of weight being proportional to area, the area still increases -n- times the increase in length. Thus, assuming the weight-specific metabolic rate stays constant, a sand dollar's appetite will always grow faster than its linear dimensions. A similar relationship is found in some fossil brachiopods : the lophophore, an internal food gathering structure, increases in relative size and complexity with increasing body size for Aemula inusitata, a species from the Cretaceous of Northwestern Europe (Gould, 1977). Thecidelliniform brachiopods from the Jurassic display similar feature changes (Gould, 1966). The geographic distribution of recent scutellid sand dollars along the North American east coast suggests a possible relationship between climate and lunule number. Echinorachnins parma is a cold-temperate species found in New England waters and has no lunules. M. quinquicsperjorata, a mild-temperate to tropical species found south of Cape Hatteras, North Carolina, has five lunules (very rarely, six). M. sexiesperjorata, which has six lunules, and Encope michlini, which has five notches and one very large anal lunule, are found in tropical regions and are extremely rare in temperate areas (Kier and Grant, 1965). One environ- mental factor that varies between cold and tropical waters is the nutrient level : cold waters tend to be richer in nutrients than tropical waters (Russell-Hunter, 1970). Perhaps the presence of lunules is a response to the reduced food supply. Lunules and notches (incomplete lunules) may represent an evolutionary modification to enhance food gathering in sand dollars. As Seilacher (1979) pointed out, the change in plate growth pattern from the regular urchins to that 568 D. E. ALEXANDER AND J. GHIOLD FIGURE 5. Evolution of lunules in scutellid sand dollars. Vertical spacing proportional to time except for Pleistocene. This diagram is modified from Durham (1966, p. 453) notably in the family Mellitidae where Mellita aclincnsis, M. Sexiespcrjorata, M. quinquiespcrjorata and the genus Encopc are depicted. The genus Echinodisctts is also shown in the family Astriclypeidae. of the irregular urchins allowed new constructional possibilities such as lunules. The important role of the lunules is emphasized by the fact that lunules appear to have arisen independently in six different sand dollar lineages (Seilacher, 1979). According to the fossil record the first lunules appeared during the Oligocene in LUNULE FUNCTION IN MELLITA 569 the suborder Scutellina, and over time, lunules increased in number and develop- ment, especially within the family Mellitidae (Fig. 5). One exception to this trend is the descent of the modern Mcllitu quinquiesperforata from M. aclinensis. M. aclinensis was a six-lunuled fossil species, which lived during the late Miocene and Pliocene in warm, subtropical waters (Kier, 1972). Kier believes that the mild-temperate M. quinquiesperforata evolved directly from the six-lunuled tropical species M. aclinensis to its modern form with commonly five, but rarely six, lunules. Thus, the fossil record also shows a correlation of lunules with warm climates, and perhaps with food supply. Lunules and notches of M. sexicsperforata and Encope tnichelini were not observed, but the distribution of these external perforations and indentations is so similar to that of M. quinquiesperforata that they undoubtedly function in the same manner. We emphasize, however, that further examination of other sand dollars is necessary before the ideas presented here can be applied to lunules in other families. ACKNOWLEDGMENTS D. E. A. would like to thank Alexander Motten, Helen Miller, and Garth Ware for much helpful discussion, Diane Campbell for important statistical help, and Steven Vogel for his plentiful encouragement and support. The section on lunule growth draws heavily upon parts of a senior thesis com- pleted by J. G. while at Harpur College, SUNY-Binghamton. Instrumental in completion of that work and a source of stimulating discussion on growth analysis were Don L. Kissling and James R. Beebower. Steven R. Dickman provided valuable assistance with the computer program. Both authors thank Robert D. Barnes, Gettysburg College; David L. Pawson, U. S. National Museum, and Patricia Timko Withers, Duke University, for critically reviewing the manuscript, and Adolf Seilacher, University of Tubingen, for commenting on revisions. Steven J. Gould, Harvard University, kindly identified himself as a reviewer and offered many pertinent suggestions for prepara- tion of the final manuscript. LITERATURE CITED BARNES, R. D., 1974. Invertebrate zoology, 3rd edtn. W. B. Saunders Company, Philadelphia. 870 pp. BELL, B. M., AND R. W. FREY, 1969. Observations on ecology and the feeding and burrowing mechanisms of Mcllita quinquiespcrjorata (Leske). /. Palcontol. 43: 553-560. BERRILL, N. J., AND J. BERRILL, 1957. 1001 Questions Answered About the Seashore. Dodd, Mead and Co., New York. 305 pp. DURHAM, J. W., 1966. Clypeasteroids. Pp. 450^491 in R. C. Moore, Ed., Treatise on In- vertebrate Paleontology, Part U, Echinodermata 3. Geol. Soc. Am. and Univ. Kans. Press, Lawrence, Kansas. GHIOLD, J., 1979. Spine morphology and its significance in feeding and burrowing in the sand dollar Mcllita quinquiespcrjorata (Echinodermata: Echinoidea). Bull. Mar. Sci., 29 : 481-490. GOODBODY, I., 1960. The feeding mechanism in the sand dollar Mellita sexiesperjorata (Leske). Biol. Bull., 119: 80-86. GOULD, S. J., 1966. Allometry and size in ontogeny and phylogeny. Biol. Rev., 41 : 587-640. GOULD, S. J., 1977. Ontogeny and Phylogeny. Belknap Press of Harvard Univ. Press, Cam- bridge, Mass. 501 pp. HYMAN, L. H., 1958. Notes on the biology of the five-lunuled sand dollar. Biol. Bull., 114: 54-56. 570 D. E. ALEXANDER AND J. GHIOLD IKEDA, H., 1941. Function of the lunules of Astriclypcus as observed in the righting move- ment (Echinoidea). Annot. Zool. Jpn., 20: 79-82. KIER, P. M., 1972. Upper Miocene echinoids from the Yorktown formation of Virginia and their environmental significance. Smithson. Contrib. Palcobiol., 13 : 1-41. KIER, P. M., AND R. E. GRANT, 1965. Echinoid distribution and habits, Key Largo Coral Reef Preserve, Florida. Smithsonian Misc. Coll., 149(6): 1-68. MORRISON, D. F., 1976. Multivariate Statistical Methods. McGraw-Hill Book Co., New York. 420 pp. RUSSELL-HUNTER, W. D., 1970. Aquatic Productivity. Macmillan Publishing Co., Inc. New York. 306 pp. SEILACHER, A., 1979. Constructional morphology of sand dollars. Paleobiologv, 5(3) : 191- 221. VOGEL, S., 1977. Flows in organisms induced by movements of the external medium. Pp. 285-297 in T. J. Pedley, Ed., Scale Effects in Animal Locomotion. Academic Press, Inc., New York. VOGEL, S., AND M. LABARBERA 1978. Simple flow tanks for research and teaching. BioScience, 28 : 639-643. WEBB, J. E., AND J. THEODOR, 1972. Wave induced circulation in submerged sands. /. Mar. Biol. Assn. U. K., 52 : 903-914. WEIHE, S. C, AND I. E. GRAY, 1968. Observations on the biology of the sand dollar Mellita quinquiespcrforata (Leske). /. Elisha Mitchell Sci. Soc., 84: 315-327. Reference: Biol. Bull.. 159: 571-581. (December, 1980) THE URN CELL. COMPLEX OF SIPUNCULUS NUDUS: A MODEL FOR STUDY OF MUCUS-STIMULATING SUBSTANCES1 BETSY G. BANG AND FREDERICK B. BANG Department of Patlwhiology, The Johns Hopkins University, School of Hygiene and Public Health, Baltimore, Maryland 21205, and Station Biologique, 29211 Roscoff, France ABSTRACT The secretory systems of the living urn cell complex of Sipunculus nudus were studied by light and fluorescence microscopy, using vital stains. The complex comprises two morphologically distinct secretory systems which respond to different stimuli, secrete at different rates, and respond differently to vital dyes. One system apparently resides in the ciliated base cell of the complex, the other in a cluster of small secretory "R" cells (presumptive regulatory cells) attached to the central membrane of the base cell. R cells slowly secrete mucus that is selectively sticky for cell debris and foreign particulates. When stimulated by pathogenic bacteria or other defined substances, the other system rapidly synthesizes streams of mucus from four to five synthetic loci. The loci are at rest until stimulated. In vivo the two secretory systems keep S. nudus coelomic fluid sterile and free of debris. Both systems respond in vitro. In in vitro suspensions, R cells are depleted over time; concomitantly, streams of mucus from stimulated loci become longer, suggesting a degree of regulation by R cells. R cells contain large granules of neutral red- staining material, but do not stain with Janus green B. The synthetic loci stain intensely with the purple form of Janus green B but do not stain with neutral red. INTRODUCTION The mucus-producing urn cell complexes of the marine invertebrates Sipunculus nudus respond in vitro to mucus-stimulating substances in the body fluids of humans and rabbits. There is a quantitative change in responses to the same fluids during the course of clinical disease (Bang and Bang, 1979). Light-microscopic studies of vitally stained living urn-cell complexes indicate that each of two secretory systems of the complex produces one or more morphologically distinctive types of mucus, each of which has a particular function. The present report reviews the functional dynamics of each secretory system as seen in vitally stained living urn-cell complexes. It also considers some of the variables inherent in the use of urn cell complexes as a biomedical assay system. MATERIALS AND METHODS Methods of collection and maintenance of S. nudus, of obtaining suspensions of urn cell complexes, and of observing the behavior of urn cell complexes in vitro Abbreviations: BFSW — boiled filtered seawater ; JGB — Janus green B; R cells — presumptive regulatory cells; DASPMI — diaminostyrylpyridinium-iodine. Received June 26, 1980; accepted September 25, 1980. 1 Supported by NIH grant #5 P50 HL-19157. 571 572 B. G. BANG AND F. B. BANG have been described (Bang and Bang, 1962). In the original studies (Bang and Bang, 1976), a series of vital dyes was used for the purpose of evaluating the specificity of a given dye for a given organelle. On the basis of those results, neutral red and Janus green were used to further define the separate characteristics of two apparently distinct secretory systems within the complex. For the present studies, a 1:500 solution of neutral red (Matheson, Coleman, and Bell) was prepared in distilled water, then diluted in boiled filtered seawater (BFSW) for a final dilution of 1:5000. Janus green B (JOB) (British Drug House) was prepared in the same way for a final 1 : 1000 solution, which was then twice filtered through a Whatman #1 filter. The seawater used to dilute for dyes and sera in all experiments was natural seawater boiled for 5 min, twice filtered through #1 Whatman filters, and stored at4°C. The stock stimulus for eliciting secretion of the free-flowing mucus from the loci was human serum from the same donor, diluted 1:10 in BFSW and heated to 85 °C for 6 min to produce a strong stimulus. Serial dilutions of this stock provided moderate to mild stimuli as required. For fluorescence microscopy, a seawater buffered 0.01% solution of acridine orange was used to identify cell nuclei. For vital staining of mitochondria, a 2 ju,M solution of the specific vital mitochondrial probe diaminostyrylpyridinium-iodine (DASPMI) (Bereiter-Hahn, 1976) was prepared in BFSW. On a glass slide, 5 /A of fluorescent stain were added to 5 p.1 of supernatant urn cell complexes at room temperature, and the appearance of fluorescence was monitored at 5-min intervals. Both acridine orange and DASPMI showed some fluorescence at 20 min, and reached sustained brightness by 40 min. To stimulate cells, 5 p\ of a 1:2 dilution of stock stimulus was added simultaneously with the dye. All fluorescent preparations were stained and photographed at the Station Biologique by Dr. Christian Sardet, Groupe cle Biologic Marine, Station Zoologique, Ville- franche-sur-Mer. RESULTS The urn cell complex. Free urn cell complexes are normal components of the coelomic fluid (blood) of S. nudits. They originate as fixed epithelial cell complexes in the coelomic-lining epithelium, where they function to trap and remove participate debris and to hyper- secrete mucus in response to bacterial pathogens. Thousands of fixed urn cell complexes periodically detach from the body wall and become free-swimming in the blood, where they perform the same functions. By producing two types of mucus, the secretory cells of the complex participate in three known major functions in the living animal: removal of foreign and cellular debris, cell clotting, and secretion of mucus that immobilizes and isolates bacteria. Non-stimulated, healthy, autologous cells are never entrapped in urn-cell- complex mucus. All of these responses are reproducible in vitro in suspensions of urn cell complexes maintained in their own coelomic fluid. The complete urn cell complex (Fig. 1) ranges in diameter from about 65-75 ^.m. It was first understood to consist of two cells, a "vesicle cell" and a mucociliated base cell (Bang and Bang, 1965). But preliminary electron microscope obser- vations by Reissig (Reissig et «/., 1979) have shown that the complex comprises three cell types : a "vesicle cell," a mucociliated base cell, and a cluster of small URN CELL MUCUS 573 ' R FIGURE 1. Profile of unstained urn cell complex: vesicle cell (v), base cell (b), and R cells (R) covered by adherent amebocytes. Cilia do not show. (about 2 /AMI) secretory cells attached to the external membrane of the base cell (Fig. 2). The vesicle cell is very thinly stretched to form a transparent hollow dome (Fig. 3, inset), the base of which is attached to the upper rim of a quadrifoil saucer-shaped mucociliated base cell by junctional complexes. The two cells thus enclose a transparent matrix. Attached to the center of the outer membrane of the base cell, also by junctional complexes, is the cluster of small secretory R cells (so designated in honor of Reissig). Between the R cells and the mucociliated rim of the base cell are several (usually four) loci of synthesis of free-flowing mucus secreted in response to defined stimuli. The static and active relationships are diagramed in Figures 2 and 3. There is an apparent division of labor between the two secretory systems : the R cells continually and slowly produce brief tails of granular mucus that is selectively sticky for cell debris and foreign participates; the other system is at rest until the loci on the nonciliated portion of the base cell membrane are stimulated to secrete streamers of free-flowing mucus. Evidence for tivo separate secretory systems The R cells. If a drop of fresh whole blood from S. nudns is diluted about 1 :20 in seawater and put on a glass slide, it can be seen that the rapidly swimming urn- cell complexes have few, if any, autologous cells adherent to them. However, in fresh preparations of supernatant urn cell complexes, the R cells nearly always have moderate loads of adherent amebocytes (Fig. 1), presumably activated as a 574 B. G. BANG AND F. B. BANG primary loci FIGURE 2. (Left) Diagram of base cell of urn cell complex, viewed from exterior surface and in profile: A. R cells in center of non-ciliated membrane; muco-ciliated rim (mcr), cilia not drawn ; B, profile, secretory granules indicated by black dots ; C, position of hypersecretory loci in central membrane; D, profile. Inset: relationships of vesicle (v) and base cell (b) with attached R cells. FIGURE 3. (Right) Diagram of interrelations of R cells and actively secreting loci, urn cell complex swimming forward (arrow on vesicle cell) while secreting tail emerges (arrows) pull- ing R cells with it for varying distances. Inset: relationship of vesicle cell (v), base cell (b), and R cells, shown as though junctional complexes were removed. result of bleeding. When these adherent cells are removed by centrifugation at 12,000 rpm for 1 min, the cluster of R cells and tags of the secretions are revealed (Fig. 4). From 80 to 100% of active urn cell complexes from freshly bled healthy S. nndus have clusters of 4—7 adherent R cells. Neutral red staining of urn cell complexes in individual suspensions over time has shown that beginning late in the first week and continuing for the subsequent 3-4 weeks, R cells gradually detach from the base cells of individual urn cell com- plexes (Table I). During this gradual in zntro loss of R cells, the tails of mucus induced in the free-flowing locus system change visibly, generally becoming longer and slimmer with time. In about 5% of suspensions, the R cells persist, and fore- shortened tails predominate for the life of the suspension. Thus the presence of the full complement of R cells evidently has some control over the quality of free- flowing secretions in vitro. Whether R cells may be replaced in vivo is not known. The large packages of granules in R cells stain brilliantly with neutral red, so that it is easy to follow the behavior of R cells as their membranes stretch when, during response to a strong stimulus, they are pulled distalward by traction of the emerging locus secretions. The nuclei are in the distal portion. The membrane often stretches until it breaks, and the nuclear portion is cleaved from the base cell. We had originally interpreted the R cells as independent sacs of the base cell mem- brane, possibly mucus reservoirs (Bang and Bang, 1965), but Dr. Reissig's electron-microscopic studies revealed that they are separate cells (Reissig et al., 1979). The fact that R cell nuclei remain in the distally stretched portion of R cells during hypersecretion of the free-flowing system was confirmed by acridine orange vital staining (Fig. 5). URN CELL MUCUS 575 TABLE I Loss of R cells from four preparations of supernatant urn cell complexes, three monitored over time. The numbers of visible R cells on 100-110 urn cell complexes were counted at random, using a cell counter. Ages of individual S. nudus unknown. Preparation # Days in culture >3 R cells to; \ \ /c) <2 R cells (%) Lacking R cells (%) 1 80 20 0 3 60 37 2 1 5 56 35.5 8.5 7 32 61 8 10 14 71 15 2 11 36 54 10 13 14 49 38 3 17 0 24 77 19 0 2.5 97.5 21 2 54 44 4 25 0 52 48 28 0 46 53 31 0 21 83 Cleavage of R cells from the base cell. Not only were R cells gradually detached from base cells over time in culture, but they were observed to be cleaved from the base cell in z>itro by bacterial action. Further, Dr. Moon Shin, Associate Professor of Pathology, University of Maryland School of Medicine, was able to induce subtotal cleavage of R cells from the base cells of stimulated urn cell complexes by incubating them with neuraminidase prior to stimulation. 6". nudus spends much of its life in partially anaerobic conditions in burrows under the sand of the ocean floor. We investigated the capability of urn cell complexes to respond in vitro, over time, to a mild heated-serum stimulus after I 17.5/1 FIGURE 4. (Left) Lower portion of urn cell complex killed by formalin gas, showing immobilized cells (c), R cells with secretory granules stained with neutral red (R), and granular secretions revealed after removing attached amebocytes by centrifugation. FIGURE 5. (Right) Stretched R cells of stimulated living urn cell complex, stained with acridine orange, showing R cell nuclei (n) and stretched membranes (m) with enclosed secre- tory granules. Fluorescence microphotograph by Dr. Christian Sardet. 576 B. G. BANG AND F. B. BANG they were placed in a sealed flowing-nitrogen gas chamber. A small vial contain- ing a 1 -day-old suspension of urn cell complexes with full complements of R cells, and another vial containing the stimulus, were placed in the chamber for 2 hr ; then at hourly intervals for the next 6 hr 10 /u.1 of stimulus were added to 10 /A! of the complex-containing suspension in a depression slide under flowing nitrogen. A coverglass was sealed in place by silicone, and responses were immediately observed in the light microscope. After 4 hr under nitrogen, many R cells appeared swollen, and R cells were seen to be pushed distally and detached from the base cell. By the 6th hour, rampant bacterial infection was present in both the suspension of urn cell complexes and the stimulus preparation ; there were no R cells on the urn cell complexes, and the base cell loci were hypersecreting thin mucus at very rapid rates. In Dr. Shin's unpublished experiments, newly bled urn cell complexes were incubated for 3 hr in Vibrio cholerae neuraminidase (2 units/ml in BFSW). Then a moderately strong human serum stimulus was added. Cells were killed after they had secreted for 10 min, and those which showed stretching of R cell mem- branes and cleavage of the nuclear portion of the cell were counted. Only \% of control complexes showed a lack of R cells, while 64% of the enzyme-treated ones showed separation of the nuclear portion of R cells from the base cell by breakage of the over-stretched cell membrane. As in urn cell complexes, which lose R cells over time, the stimulus-induced tails subsequently produced by com- plexes from which R cells had been cleaved were two to three times longer than those of complexes which had retained the R cells. There was no stretching of R cells in unstimulated complexes. R cell pathology. The only condition thus far identified as one in which R cell hypersecretion occurs is a transmissible pox-like skin disease of S. nudus which sometimes appears in laboratory-maintained animals. R cell abnormalities were found in coelomic fluid samples from eight S. nndus experimentally infected with the skin disease. In five cases there was R cell hypertrophy, in one case marked excess R cell secretion, and in the other two cases R cells were absent and possibly cleaved. During these events, there was no evidence of any activity from the hyper- secretory loci. The serum of animals infected with this disease was consistently found to contain a previously identified heat-labile lysin against the ciliate Anophrys (Bang, 1966). The hypersecretory loci. \Yhile R cell intracellular granules stain intensely with neutral red, neither the granules nor the extracellular mucus stains with Janus green. The loci which produce free-flowing mucus in response to specific stimuli do not stain with neutral red, nor does the extracellular mucus which emerges from these loci. The loci do, however, stain within a few minutes of exposure to the vital basic azo dye Janus green, concentrating the purple (reduced) form of the dye (Bang and Bang, 1976). The dye kills the urn cell complexes unless the loci are stimulated to produce mucus. The dye is evidently metabolized in some way during the secretory process and is re-excreted with the mucus, expressing the color which indicates the reduced or oxidized form of the dye (Lazarow and Cooper- stein, 1953) depending on the stimulus. This dye is known to be preferentially adsorbed on the insoluble protein fraction of serum albumin (Lazarow and Cooper- stein, 1953). R cells and loci: separate secretory systems. Three sets of light microscope evidence indicate that the R cells and loci are separate secretory systems. URN CELL MUCUS 577 FIGURE 6. ( Left) Mucociliated rim of base cell of 1-day-old living urn cell complex 30 min after staining with the mitochondria! fluorescent stain DASPMI, showing the lacework of mitochondria. Fluorescence microphotograph by Dr. Christian Sardet. FIGURE 7. (Right) Same area of base cell of living 20-day-old urn cell complex, showing small rounded mitochondria. Pale oval is nucleus (n), blank areas are presumably loci. Fluorescence microphotography by Dr. Christian Sardet. DASPMI stain. First, when new suspensions of urn cell complexes were incubated in neutral red in direct sunlight for 90 min, by which time the stained R cells were presumably partly inactivated, the loci immediately concentrated Janus green and hypersecreted vigorously in response to the stock stimulus. Second, in 20-25 day suspensions in which many complexes visibly lacked R cells, the loci also quickly concentrated Janus green and subsequently secreted when stimulated. Finally, when complexes that lacked R cells were stimulated with either serous nasal fluid or centrifuged lacrimal fluid, each of which induced a separate refractile stream of secretion, the point of attachment of each stream to its locus of origin on the base cell was clearly visible at 400 X magnification. The ultrastructure of both of the secretory systems awaits electron microscopic analysis. Mitochondria. The base cell of the urn cell complex is very active metabolically, since the cilia and interciliary secretory granules are in the outer rim of this cell. The specific fluorescent mitochondria! stain DASPMI revealed the extremely rich lacework of mitochondria in this area in freshly cultured urn cells (Fig. 6). The mitochondria in the central membrane of the cell shown were obscured by nonspecifically fluorescent R cells and adherent amebocytes. Figure 7 shows the same area in a 20-day-old urn cell complex that lacked both R cells and adherent amebocytes : the separate, rounded mitochondria of the central membrane occupy all areas except those taken up by the large nucleus and by the presumptive loci, which appear as blank spaces. During observation of this moribund old cell, the vesicle became separated from the base cell and revealed a remarkable concentration of mitochondria in the rim of the vesicle at its site of attachment to the base cell. Interactions of R cell and locus secretions. We have not yet been able to develop a stimulus which clearly induces concomitant R cell and locus secretions. When 578 B. G. BANG AND F. B. BANG autologous heat-killed erythrocytes were added to a new suspension of urn cell complexes during active secretion from the loci, the red cells tended to adhere predominantly to the proximal part of the secreted tail at the level of the R cells (Fig. 8), suggesting that polymerization of some R cell product may interact with the freely flowing secretion at that level. Effect of S-H compounds. Because of their capacity to break S-H bonds in respiratory, intestinal, and cervical mucus, the cysteine compounds are frequently used to solubilize mucus for therapeutic and experimental purposes. We are interested in the effects of these amino acid compounds on the synthesis of mucus by urn cell complexes and in their effects on previously secreted or final mucus. Incubation of a suspension of urn cell complexes in equal volumes of 0.01 M concentrations of any of the eight free-thiol compounds listed below for 5 min before adding a strong stock stimulus induced urn cell complexes to produce very long, periodic compact tails within 2-5 min (Fig. 9 inset). This was in contrast to the cloudy, precipitate-covered tails induced in the same sample of urn cell com- plexes by the same stmulus without pre-incubation in the free-thiol compounds (Fig. 9). At no time was there solubilization of the secreted mucus. Nine S-H compounds that altered and enhanced secretion produced by urn cell complexes following stimulus by heated serum were L-cysteine, D-cysteine, N-acetyl-L- cysteine, 1,4-dithio-L-threitol (L-DTT), dithioerythretol (DTE), S-carboxymethyl- L-cysteine, 2-mercapto-ethanol, thioglycolic acid, and glutathione. Also tested were 6 amino acids without free S-H terminals on the molecule. Incubation of urn cell complexes in each of these before addition of the stock stimulus produced no change in the morphology of the tail, nor was there any effect \ I 100/J FIGURE 8. (Left) Living urn cell complexes were stimulated with mucus-stimulating human serum, then heat-killed S. nudus erythrocytes were added. These were treated as non-self and were scavenged by urn cell complexes. Red cells accumulated most avidly at R cell level. FIGURE 9. (Right) Living urn cell complexes photographed 10 min after exposure to mucus- stimulating human serum : copious secretory tails covered with background precipitate. Inset : Living urn cell complex from same culture incubated 5 min in glutathione, then exposed to same stimulus and photographed in 10 min : long, thin, periodic, compact secretory tail. URN CELL MUCUS 579 on the secreted mucus. The following !_' amino acids failed to enhance or alter secretion : L-cystine, L-cysteic acid, L-alanine, X-acetyl-L-alanine, L-phenylala- nine, L-asparagine. L-arginine. L-glutamic acid, L-proline, L-serine, and L- methionine. The junction of the secretory systems in nature. There is no way of knowing the ages of individual S. nitdits that are sources of supernatant urn cell complexes. Nor can one judge the spectrum of "ages" (i.e.. the time after release from the epithelial lining) of individual complexes within a supernatant population. Also unknown is the recent history of exposure of each S. niidns to various types of stress in its habitat. For these reasons, laboratory-maintained 5". nitdus cannot be expected to provide urn cell complexes which give statistically standard responses, a variable which must be taken into account in any comparative study. It is advisable to use urn cell complexes from the same supernatant preparation for any comparative study. At least 500 tests can be carried out with equal portions of supernatant from 2 to 3 ml of whole 5\ nitdus blood. \Ye tried to gain some insight into the amount of stimulus to which the secretory systems might be routinely exposed in nature. Thirty animals were bled in the the field under sterile conditions. Then, after a 45-min drive back to the laboratory, the supernatant urn cell complexes of each of the 30 test-tube preparations were examined microscopically for "spontaneous" hypersecretion. Urn cell complexes in 13, or 43c/c, of the preparations were producing small tails of varying lengths and shapes, usually a clear granular secretion. Agar plate cultures of a drop of blood from each of these same animals had been prepared in the field, and after 3 days' culture at room temperature, there were no visible bacterial colonies. Thus, the spectrum of infectious agents or toxins which induce urn cell complexes to produce secretory tails of differing qualities in nature remains to be determined. DISCUSSION Our previous papers on responses of urn cell complexes to stimuli have empha- sized effects on the locus system, since it is this system which responds differently to body fluids of humans and rabbits in normal and dseased states. It is now clear, however, that the effects of R cells and their products on the nature of the final (combined) urn cell complex secretion may be crucial in 5". nitdus in vivo. There are two points to be made about the biological activity of R cells. One is that in vivo trapping of activated and dead cells by R cells is the first stage in progressive phagocytosis of this cell debris by autologous leukocytes. We suggest therefore that R cells may themselves be phagocytic : like vertebrate macrophages, they rapidly and avidly pinocytose neutral red, and vertebrate physiologists now recognize that the classical macrophage is a secretory cell (Unanue ct al.. 1976). Dybas (1976) described the phagocytic properties of the free urn cell complex of another Sipun- culid. Phascolosojna agassizii, a complex comprising three constituent cell types, two of which are selectively phagocytic. The second point is that understanding the relationship between R cell secretory activity and the nature of locus secretion may be very useful in understanding the nature and control of analogous secretory responses in human diseases that affect mucous secretory systems. In human respiratory mucous membranes, for example, there is continuous interaction between serous and mucous gland cells. The surface mucus derives from mixed mucous-serous acini and from goblet cells, ontogenically different cell systems (Tos, 1976; Bang, 1964). Qualitative changes in the surface 580 B. G. BANG AND F. B. BANG mucus may result from airborne stimuli (allergens or irritants), from infectious agents, or from systemic defects. Cystic fibrosis exemplifies the latter : the respiratory mucous membranes of cystic fibrosis patients function normally until bacterial infection sets in ; normal individuals recover mucociliary function with no adverse sequelae, while cystic fibrosis patients are unable to produce mucus of normal consistency after such an infection ; tenacious mucus is anchored in the gland cells and produces varying degrees of mucociliary stasis. In living specimens of 6\ Niidns, during acute bacterial infections all urn cell complexes are engaged in producing mucus which immobilizes, traps, and probably lyses the bacteria. This race between bacterial replication and urn cell complex activity ends either with death of the 6". audits or elimination of the bacteria. During low-grade infections, there is a clear advantage to having the hypersecretory (locus) system controlled by the R cells: enough mucus is produced to trap the bacteria without energy-expensive hypersecretion of large volumes of mucus. During acute infection, however, bacterial products themselves may cleave the R cells from the base cell and allow uncontrolled hypersecretion to proceed to eliminate the bacteria. The data suggest that R cells and loci each have receptors for stimuli that have particular molecular conformations. The nature of the stimuli and of the receptors, the composition of R cell secretion and locus secretions, the ultrastructure of the respective secretory systems, and the role of the fluid space enclosed by the vesicle and the base cell all remain to be investigated. Radio isotope labeling of metabolic precursors and of the active components of defined stimuli may begin to resolve the biochemical questions ; electron microscopy should identify the secretory apparatuses. The urn cell com- plex provides a system in which the regulation and production of mucus can be systematically explored in vitro, and a biological system in which mucus stimulating substances in invertebrate and vertebrate body fluids can be comparatively assayed in health and disease. ACKNOWLEDGMENTS The flowing-nitrogen chamber was kindly loaned to us at Roscoff by Dr. Manfred Grieshaber of the Zoologisches Institut der Westfalischen Wilhelms-Universitat, Minister. West Germany. We thank Dr. Moon L. Shin, Associate Professor of Pathology, University of Maryland School of Medicine, for her innovative contributions to the model, and the administrative and field staffs of the Station Biologique, Roscoff, for their continued cooperation. LITERATURE CITED BANG, B. G., 1964. The mucous glands of the developing human nose. Ada Anat., 59: 297-314. BANG, B. G., AND F. B. BANG, 1965. Mucus hypersecretion in a normally isolated non-inner- vated cell. Cah. Biol. Mar., 6: 257-264. BANG, B. G., AND F. B. BANG, 1976. The mucous secretory apparatus of the free urn cell of Sipunculus uudus. Call. Biol. Mar., 17 : 423-432. BANG, B. G., AND I<". B. BANG, 1979. Mucus-stimulating substances in human body fluids assayed in an invertebrate mucous cell system. Johns Hopkins Mcd. J ., 145 : 209-216. BANG, F. B., 1966. Serologic response in a marine worm, Sipidiculus nudus. J. Inununol., 96: 960-972. BANG, F. B., AND B. G. BANG, 1962. Studies on sipunculid blood: immunologic properties of coelomic fluid and morphology of "urn cells." Cah. Biol. Mar., 3 : 363-374. URN CELL MUCUS 581 BEREITER-HAHN, J., 1976. Dimethylaminostyrylmethylpyridiniuin- iodine (DASPMI) as a fluorescent probe for mitochondria in situ. Bioeliiin. IHcptiys. Acta. 423 : 1-14. DYBAS, L., 1976. A light and electron microscopic study of the ciliated urn of Phascolosoma ac/assizii ( Sipunculida ) . Cell Tissue Res., 169: 67-75. LAZAKOW, A., AND S. J. COOI-KKSTKIN, 1953. Studies on the mechanism of Janus green B stain- ing of mitochondria. Ex p. Cell Res., 5 : 56-97. REISSIG, M., B. G. BANG, AND F. B. BANT,, 1979. Mucus secretion in the urn complex of SipHncHliis nudus. J. Cell Biol.. 83: SP2512 (abstract). Tos, M., 1976. Mucous elements in the nose. Rhiiiolouy. 14: 155-162. UNANUE, E. R., D. I. BELLER, J. CALDERON, J. M. KIKI.Y, AND M. J. STA DECKER, 1976. Reg- ulation of immunity and inflammation by mediators from macrophages. Am. J. Patlml.. 85: 466-478. Reference: Biol. Bull., 159: 582-591. (December, 1980) ULTRASTRUCTURAL CHARACTERISTICS OF THE NON-EXPANDED AND EXPANDED EXTRA-EMBRYONIC SHELL OF THE HORSESHOE CRAB, LIMULUS POLYPHEMUS L. GARY A. BANNON 1 AND GEORGE GORDON BROWN Department of Zoology, Iowa State University, Ames, lozva 50011 and Hopkins Marine Station oj Stanford University, Pacific Grove, California 93950 ABSTRACT During embryonic development of the horseshoe crab, Limulus polyphemus, a structural layer, the extra-embryonic shell (EES), develops in the periembryonic space. Late in development the egg envelope breaks away and the EES expands slowly (over a period of several days) to approximately twice its original diameter. The EES is composed of non-cellular material organized into three distinct layers. The outermost (layer 1) is electron translucent and exhibits numerous hairlike projections. The middle layer (layer 2) is differentiated from layer 1 by its greater density under the electron beam. The innermost layer (layer 3) comprises 99% of the mass of the EES and is intermediate in electron density between layers 1 and 2. The inside surface of the non-expanded EES is smooth, and easily dis- tinguished from the outside surface, which is rough and accentuated by polygonal structures. The deep indentations separating these structures are probable sources of preformed surface area utilized during EES expansion. The expansion of the EES occurs concomitantly with cracking of the egg envelope when the embryo has reached stage 20. As expansion continues, the distance between the once tightly packed polygonal structures is greatly increased. However, the thickness of the expanded shell is less than expected if expansion of the EES is only the result of utilization of surface area stored in the indentations. Therefore, EES expansion appears to be made possible by at least two mechanisms: (1) Utiliza- tion of preformed surface area stored in the indentation, and (2) Stretching of the shell — implying that the EES exhibits limited elasticity during development. Expansion of the EES is necessary to provide the space required for later larval development. INTRODUCTION The envelopes of developing embryos are important for protection from me- chanical injury, maintenance of a constant environmental medium, and passage of respiratory gases (Wigglesworth, 1946). These envelopes, depending on their origin, have been classified as primary, secondary, or tertiary (Quattropani and Anderson, 1969; Ludwig, 1874). Primary envelopes are produced by the oocyte, secondary by ovarian follicle cells, and tertiary by an extra-ovarian source (e.g., the developing embryo) (Quattropani and Anderson, 1969). The formation, structure, and function of primary and secondary envelopes have been described Abbreviations : EES, extra-embryonic shell ; ASW, artificial seawater. Received August 10, 1980 ; accepted September 25, 1980. 1 All communication should be with this author. 582 EXTRA-EMBRYONIC SHELL OF LIMULUS 583 for a number of invertebrates (Regier ct al., 1980; Mazur et al., 1980; Turner and Mahowald, 1976; Barbier and Chauvin, 1974; Salkeld, 1973; Quattropani and Anderson, 1969; Cheung, 1966). The tertiary envelope, however, has been less thoroughly examined, probably because many organisms do not form a true envelope of this origin (e.g., the fertilization membrane of sea urchins can be de- scribed as one of primary and tertiary origins because it is formed from a combina- tion of the oocyte-produced vitelline layer and materials excreted by the embryo during the cortical reaction ; Chandler and Heuser, 1980 ; Boyan, 1970a, b). In the American horseshoe crab, Limulus polyphemus, all three types of en- velopes are found. Eggs possess an egg envelope consisting of an inner vitelline envelope of primary origin and an outer basement lamina of secondary origin. The formation and structure of these layers have been previously described (Shoger and Brown, 1970; Dumont and Anderson, 1967). However, in Limulus and the closely related Japanese horseshoe crab, Tachypleus tridentatus, the formation and structure of the extra-embryonic shell (a true tertiary envelope) has been examined only superficially (Kingsley, 1892; Sekiguchi, 1970). The extra-embryonic shell (EES) appears in the periembryonic space during early development of the Limulus embryo and is completed prior to stage 17 (Brown and Clapper, 1980). During stage 20 (approximately 21 days after insemination at 20°-21°C) the egg envelope cracks off and the EES slowly expands to ap- proximately twice its previous diameter. In this expanded sphere, the larva is active, and several days after the cracking of the egg envelope moults into a trilobite larva (stage 21). Eventually, the trilobite larva will hatch from the EES and become free-swimming. Similar observations have also been recorded by Seki- guchi (1973) and Sugita and Sekiguchi (1979) on Tachypleus tridentatus. This report describes the organization of the extra-embryonic shell (EES) before and after expansion. Possible mechanisms by which the EES expands are also discussed. MATERIALS AND METHODS Source of Animals Specimens of Limulus polyphemus L. (obtained from the Florida Marine Bio- logical Specimen Co., Inc., Panama City, Florida) were maintained at 15°C in Instant Ocean Aquaria in artificial sea water (ASW) (Aquarium Systems, Inc., Eastlake, Ohio) on a 14 :10 L :D photoperiod. Gamete collection and insemination Gametes were collected by brief electrical stimulation (1-1.5 V, 0.5-1.0 mA, ac) of the gonoducts proximal to the gonopores (Brown and Clapper, 1980). Semen was diluted with 100% ASW (960 m-osmoles) to produce a W% sperm suspen- sion (109 spermatozoa/ml). Eggs (40-50) were placed in a plastic petri dish containing 25 ml of ASW. Two drops of the sperm suspension were added and the mixture was gently swirled. All cultures were maintained at room tem- perature (20°-21°C). Procurement of embryos before and after expansion of EES Specimens representing different steps of expansion of the EES were collected for electron microscopic preparation before, during, and after cracking of the egg 584 G. A. BANNON AND G. G. BROWN FIGURE 1. Expansion of the extra-embryonic shell : (a) The embryo is enclosed by the extra-embryonic shell and the egg envelope. Bar = 1.0 mm. (b) The embryo has moulted EXTRA-EMBRYONIC SHELL OF LIMULUS 585 envelope and subsequent expansion of the EES. The first step was obtained when embryos were in stage 1(> (approximately 19 days after fertilization). The second step was obtained after stage 20 appeared and cracking of the egg envelope was in process. Specimens for the third step were usually collected several days after cracking had occurred. The expansion process was measured daily until hatching and was recorded photographically using a Wild M-5 dissecting micro- scope. Electron microscopy Properly staged specimens were fixed in a 2. 5 % glutaraldehyde solution (cacodylate buffer; 4°C) for 12-24 hr, post-fixed in 2% osmium tetroxide and stored in double distilled H2O for one or more days. For critical point drying, utilizing liquid carbon dioxide, fixed specimens were dehydrated in ethanol and cleared into atnylacetate. Some dried specimens were fractured with a sharp razor blade. For freeze-drying, fixed specimens were snap-frozen, one at a time, in liquid nitrogen, placed on pre-cooled brass blocks, and evacuated (5 X 10~:' Torr) overnight. All specimens were mounted on stubs with silver paint, carbon and gold coated, and examined on a Cambridge Mark 2A "Stereoscan" scanning electron microscope. Similarly staged specimens, to be examined by transmission electron microscopy, were fixed in a trialdehyde solution (Kalt and Tandler, 1971 ; acrolein, 0.178 M ; DMSO, 0.32 M ; formaldehyde, 0.66 M ; glutaraldehyde, 0.33 M : pH, 7.0) for 3 hr, post-fixed in 1% osmium tetroxide and dehydrated. All fixation procedures were carried out at 4°C. The sample was infiltrated by puncturing the embryo with a sharp probe before embedding it in Spurrs resin (Standard Medium A; Poly- sciences Data Sheet No. 127). All tissue was sectioned on a LKB Ultrotome, stained with aqueous uranyl acetate and lead citrate and examined on a Hitachi HU-11E-1 electron microscope. RESULTS General observations Approximately 14 days elapse from the first step of the expansion process (Figs, la-c) to hatching (19th-33rd day after fertilization). The diameter of the enclosing envelopes during this process increases from an initial 3.8 to 6.4 mm at hatching. This increase in diameter of the EES is very slow during expan- sion, even immediately after sloughing of the egg envelope (Fig. Ib). The EES forms in the periembryonic space between the egg envelope and embryo proper and appears to be morphologically complete by stage 14 (limb-bud stage, 12 days after fertilization). Obvious polygonal structures can be observed easily after the egg envelope cracks (Fig. Id). into stage 20 and the egg envelope is cracking away. The extra-embryonic shell is exposed and is slowly expanding. Full expansion takes several days. Same scale as (a), (c) The fully expanded extra-embryonic shell allows an increased space in which the stage 20 embryo can be active and undergo the final embryonic moult into the trilobite larva (stage 21). Same scale as (a), (d) Scanning electron micrograph (SEM) of the cracking of the egg envelope (EN). The extra-embryonic shell (EES) is characterized by the presence of numerous packed polygonal structures which gives the surface a rough, irregular appearance. Bar = 0.5 mm. 586 G. A. BANNON AND G. G. BROWN • • - - . • r. EXTRA-EMBRYONIC SHELL OF LIMULUS 587 Extra-embryonic shell before expansion The EES before expansion is evident in a section through a stage 19 embryo (Fig. 2a). Also apparent are the egg envelope and the developing embryo. The egg envelope consists of two morphologically distinct layers, a 5-/xm-thick base- ment lamina and a 40-//,m-thick vitelline envelope (Shoger and Brown, 1970; Dumont and Anderson, 1967). The EKS is approximately 33.16 ^m thick and consists of swirls of non-cellular materials with the outer third containing regularly spaced indentations. Closely apposed to the inside surface of the EES is the chitinous exoskeleton of the embryo. The EES is composed of three distinct layers (Figs. 2b, d). The outermost (layer 1) is approximately 0.02 /u,m thick, electron translucent (Fig. 2d), and the source of numerous hairlike projections. The middle layer (layer 2) is ap- proximately 0.14 yum thick and electron dense. Layer 3, the innermost, comprises the majority of mass of the EES and is approximately 33 /u.m thick and intermediate in electron density between layers 1 and 2. Fibrous bundles are observed through- out layer 3 (Figs. 2b, e). Associated with the indentations observed in the EES in Figures 2a and b are electron-dense bodies (Fig. 2c). Extra-embryonic shell after expansion After expansion the exterior surface of the EES consists of numerous variably- shaped polygonal structures separated by a large interpolygonal region with a very irregular surface (Fig. 3a). As the EES expands this region is formed by elevation and flattening of the previously described indentations between the polygonal structures. On the surface, the interpolygonal region appears to be composed of many ridges or "fibers" radiating from each polygonal structure, and interdigitating with "fibers" from other polygonal structures (Fig. 3a). This arrangement can also be observed on an unfixed expanded EES with light microscopy. Between these "fibers" hairlike projections give the appearance of cross-linking (Fig. 3b). Cross sections reveal the surface is very irregular and these "fibers" are ridges on the outer surface of the EES (Fig. 3c, d). The thickness of the fully expanded shell (measured at widest portion) is reduced to 12 /im, a 64 % reduction from the non- expanded shell. The three layers which comprise the EES retain the characteristics seen in the non-expanded form, except for the absence of fibrous bundles and electron-dense bodies in layer 3 (Figs. 3c, d). DISCUSSION For xiphosurid embryology, the use of the term extra-embryonic shell (EES) is recommended since this structure is formed outside the embryo, is composed of noncellular material, and is formed by embryonic secretions which may occur over FIGURE 2. Electron microscopy of non-expanded shell : (a) SEM of fractured stage-19 embryo and surrounding coats. The EES develops between the egg envelope (EN) and the chitinous exoskeleton (*) of the embryo (EM) in the periembryonic space. Indentations in the EES are clearly visible (arrows). Bar = 5 fj.m. (b) TEM of non-expanded EES showing large indentations and fibrous bundles (FB). Bar = 3 /im. (c) Longitudinal section of an indentation in the non-expanded shell. Dense bodies are associated with the indentations. Bar = 1 fim. (d) Cross-section of the non-expanded EES detailing the three layers which compose this structure. Bar = 0.5 /urn. (e) Fibrous bundles of layer 3 in non-expanded EES. Bar = 0.5 jum. 588 G. A. BANNON AND G. G. BROWN EXTRA-EMBRYONIC SHELL OK LIMULUS 589 several developmental stages. Other terms have been used : blastoderm cuticle (Roomval. 1944), blastodermhaut ( Kingslt-y. 1S<>2), deutovum (Iwanoff, 1933), inner egg membrane (Sekiguchi, 1970), and egg membrane (French, 1979). These are inappropriate because of the above characteristics. However, the EES of Lhnuliis has some similarities to a typical insect cuticle. An insect cuticle is constructed asymmetrically with the outermost layer represent- ing the first layer produced by the epidermis and subsequent layers distinguished by their different electron densities as observed with electron microscopy (cf. Neville, 1975). The insect cuticle may be composed of numerous layers (Caveny, 1971 ). The EES is also asymmetrical but consists of only three layers, which could correspond, at least spatially, to the epicuticle, exocuticle, and endocuticle of insect cuticle. The EES differs from a typical insect cuticle in its relative simplicity and its separation from any cellular layer by a periembryonic space. The expansion of the EES involves utilization of preformed surface area stored in the large indentations observed in the non-expanded shell (Fig. 2b). However, the size increase of the totally expanded shell requires more surface area than found in these indentations. For example, before expansion, the thickness of the EES from the bottom of the indentations to the inside surface of the EES is 22 /mi. But when the expanded shell reaches maximum size, its width is only 12 /xm at the widest point, indicating that the noncellular components of the EES must be rearranged (stretched) to compensate for the decreased thickness of the expanded shell. As measured in this study and others (Sekiguchi, 1970; Roon- wal, 1944), the EES continues a slow expansion until hatching of the larva. The mechanism for expansion has been considered by various investigators (Kingsley, 1892; Iwanoff, 1933; Roomval. 1944; Sugita and Sekiguchi, 1979). Several theories have been presented. A favorite one involves osmotic effectors produced by the embryo to allow an increase in fluid content in the periembryonic space. However, little supporting evidence has been provided. Recently, using Tachyplens, Sugita and Sekiguchi (1979) have examined the complement of proteins present in the periembryonic ("perivitelline") space and described four classes of proteins, H, B-l, B-2, and "rest." Changes in concentration of these proteins were followed during development. The "rest" proteins increased dra- matically before expansion of the EES, indicating their possible involvement in this process. According to Yamamichi and Sekiguchi (1974) and Sugita and Sekiguchi (1979) the EES is formed from material secreted into the periembryonic space by blastoderm cells. The timetable for secretion of these components into the perivitelline space and the mechanisms of transport are not known. However, it is possible that some of the components reach the perivitelline space via the egg cortical reaction (Bannon and Brown, 1980). The question now being ad- dressed in our laboratory is whether these components are synthesized during FIGURE 3. Electron microscopy of expanded shell : (a) SEM of exterior surface of ex- panded EES showing polygonal structures (PS) and the large interpolygonal region with interdigitating "fibers." Bar = 5 nm. (b) SEM of interpolygonal region. This region is very irregular and numerous hairlike projections can be seen originating from the "fibers." Bar = 1 /urn. (c) TEM of expanded shell. The large indentations, fibrous bundles, and electron dense bodies are no longer present. Bar = 2 t*m. (d) Cross section of expanded EES showing that the basic structure of the shell does not change after expansion (1, layer 1; 2, layer 2; 3, layer 3). (Cf. Fig. 2d.) The irregular surface topography represents the "fibers" ob- served in 3a. Bar = 0.5 590 G. A. BANNON AND G. G. BROWN oogenesis and stored in the early embryo until needed, or synthesized by the embryo during development. ACKNOWLEDGMENTS The authors acknowledge the use of Dr. Walt Humphreys' SEM facilities at the Department of Biology, University of Georgia, and give special thanks to Dr. George Dearlove, Department of Biology, Hofstra University, Hempstead, New York, for his critical reading of the manuscript. This research was sup- ported in part by the Iowa State University Graduate College, an Iowa State University Research Grant, an Iowa State University Faculty Improvement leave, and a gift from Houston Endowment, Inc. LITERATURE CITED BANNON, G. A., AND G. G. BROWN, 1980. Vesicle involvement in the egg cortical reaction of the horseshoe crab, Limulus polyphemus L. Dcv. Biol., 76: 418^27. BARRIER, R., AND G. CHAUVIN, 1974. The aquatic egg of Nymphula nymphaeata (Lepi- doptera : Pyralidae). On the fine structure of the egg shell. Cell Tiss. Res., 149: 473-479. BOYAN, J., 1970a. The isolation of a major structural element of the sea urchin fertilization membrane. /. Cell Biol., 44 : 635-644. BOYAN, J., 1970b. On the reconstitution of the crystalline components of the sea urchin fertilization membrane. /. Cell Biol., 45 : 606-614. BROWN, G. G., AND D. L. CLAPPER, in press. Procedures for collecting gametes and culturing embryos of the horseshoe crab, Limulus polyphemus. In Ralph Hinegardner ct a/., Eds., Marine Invertebrates, Laboratory Animal Maintenance. National Academy of Science, Washington, D. C. CAVENY, S., 1971. Control of Cuticle Architecture in Insects. Ph.D. thesis, Oxford Univ., England. CHANDLER, D. E., AND J. HEUSER, 1980. The vitelline layer of the sea urchin egg and its modification during fertilization. /. Cell Biol., 84 : 618-632. CHEUNG, T. S., 1966. The development of egg membranes and egg attachment in the shore crab, Carcinus maemts and some related decapods. J. Mar. Biol. Assn. U. K., 46 : 373^00. DUMONT, J. N., AND E. ANDERSON, 1967. Vitellogenesis in the horseshoe crab, Limulus polyphemus. J. Microsc., 6 : 791-806. FRENCH, K. A., 1979. Laboratory culture of embryonic and juvenile Limulus. Pp. 61-71 in Biomedical Applications of the Horseshoe Crab (Limulida). Alan R. Liss, Inc., New York. IWANOFF, P. P., 1933. Die embryonale Entwicklung von Limulus moluccanis. Zool. Jahrb., 56: 164-346. KALT, M. R., AND B. TANDLER, 1971. A study of fixation of early amphibian embryos for electron microscopy. J ' . Ultrastruct. Res., 36 : 633-645. KINGSLEY, J. S., 1892. The Embryology of Limulus. J. Morphol., 7 : 35-68. LUDWIG, H., 1874. Uber die Eibildung in Thierreiche. ll'uzb Zool. hist. Arb., 1: 287-510. MAZUR, G. D., J. C. REGIER, AND F. C. KAFATOS, 1980. The silkmoth chorion: Morpho- genesis of surface structures and its relation to synthesis of specific proteins. Dev. Biol., 76: 305-321. NEVILLE, A. C., 1975. Biology of the Arthropod Cuticle. Springer Verlag, Berlin. 448 pages. QUATTROPANI, S. L., AND E. ANDERSON, 1969. The origin and structure of the secondary coat of the egg of Drosophila mclanogaster. Z. Zcllforsch., 85 : 495-510. REGIER, J. C., G. D. MAZUR, AND F. C. KAFATOS, 1980. The silkmoth chorion: Morphological and biochemical characterization of four surface regions. Dcv. Biol., 76 : 286-304. ROONWAL, M. L., 1944. Some observations on the breeding biology and on the swelling, weight, water content and embryonic movement in the developing eggs of the Moluccan king crab, Tachyplcus yigas (Muller). Proc. Indian Acad. Sci. Sec. B, 20: 115-129. SALKELD, E. H., 1973. The chorionic architecture and shell structure of Amathes C-Nigrum (Lepidoptera : NoctuidaeJ. Can. Entomol., 105: 1-10. EXTRA-EMBRYONIC SHELL OF LIMULUS 591 SEKIGUCHI, K., 1970. On the inner egg membrane of the horseshoe crab. Zool. Mag. (Tokyo), 79: 115-118. SEKIGUCHI, K., 1973. A normal plate of the development of the Japanese horseshoe crab, Tachyplcus tridcntatus. Sci. Rep. Tokyo Kyoiku Daigaku Sect. B, 15 : 153-162. SHOGER, R. L., AND G. G. BROWN, 1970. Ultrastructural study of sperm-egg inteactions of the horseshoe crab, Liiimlus Polyphemus L. ( Merostomata : Xiphosura). J. Suh- microsc. Cytoi, 2: 167-179 SUGITA, H., AND K. SEKIGUCHI, 1979. Protein components in the perivitelline fluid of the embryo of the horseshoe crab, Tachyplcus tridcntatus. Dcv. Bio!., 73 : 183-192. TURNER, F. R., AND A. P. MAHOWALD, 1976. Scanning electron microscopy of Drosophila embryogenesis : 1. The structure of the egg envelopes and the formation of the cel- lular blastoderm. Dcv. Biol. 50 : 95-108. WIGGLESWORTH, V. B., 1946. The Principles of Insect Phvsioloyv. Chapman and Hall, London. Pp. 1-26. YAMAMICHI, Y., AND K. SEKIGUCHI, 1974. Embryo and organ cultures of the horseshoe crab, Tachypleus tridcntatus. Dev. Growth Differ., 16: 295-304. Reference: Bio/. Bull., 159: 592-605. (December, 1980) THE LARVAL DEVELOPMENT OF PINNIXA LONGIPES (LOCKINGTON, 1877) (BRACHYURA: PINNOTHERIDAE), REARED IN THE LABORATORY GEORGE D. BOUSQUETTE1 Pacific Marine Station, Dillon Beach, California, and University of the Pacific, Stockton, California ABSTRACT Larvae of Pinnixa longipes, a commensal with the polychaete Axiothella rubro- cincta, were reared to the megalopa stage in the laboratory. The zoeae were cultured at 34#f, 20° C, fed a combination of Dunaliella tertiolecta and Phaeodac- tyluiii tricornntuni daily, and provided with a photoperiod of 12 h each of light and dark. The five zoeal and one megalopal stages are described and figured in detail, including the arrangement of chromatophores and various types of setae. Zoeae of P. longipes can be distinguished from other Pinnotherid zoeae, for which descriptions are available, on the basis of morphological characteristics. INTRODUCTION The pea crab, Pinnixa longipes (Lockington, 1877), is commensal with the tube building polychaete, Axiothella rubrocincta (Johnson, 1901). This worm con- structs U-shaped tubes in sand-mud substrata of bays and estuaries along the Pacific coast (Kudenov, 1977). The pronounced lateral elongation and diminutive size of P. longipes enables it to move easily in the narrow tube with the worm. The species has occasionally been found with other polychaetes, including Pectinaria and F^ista (Carlton and Kuris, 1975). According to Schmitt et al. (1973), Pinnixa longipes ranges from Tomales Bay to Laguna Beach, California. There are no published accounts regarding the biology of this commensal crab. The present paper describes in detail the larval development of P. longipes reared under laboratory conditions. These results are compared with previous descriptions of pinnotherid larvae. MATERIALS AND METHODS Ovigerous specimens of Pinnixa longipes were collected on 13 and 18 November 1978 from Axiothclla rubrocincta tubes at White Gulch in Tomales Bay, California. In the laboratory, the ovigerous females were placed in finger bowls (10-cm diameter) with filtered seawater at 34'^ salinity. This salinity was maintained by mixing Instant Ocean or glass-distilled water with filtered seawater. A constant temperature of 20° C was maintained in a refrigerator supplied with fluorescent lights to provide 12 h each of light and dark. Adult females were not fed during the experiment. The filtered seawater was changed daily and the bowls inspected for larvae. Received September 9, 1979; accepted August 20, 1980. 1 Current address : 2480 I St., Petaluma, CA 94952. 592 PINNIXA LONGIPES LARVAL DKVELOPMENT After hatching, the larvae were placed in finger howls ( 10-cin diameter) and maintained under the same salinity, temperature, and light parameters as the adults. From 10 to 20 zoeae were put in each howl . Each day. after changing the medium, approximately 2 nil each of concentrated Ditnaliclla tcrtiolccta and Phaeodactylum tricornutuw were added to the howls. The algae were cultured in a half-strength enriched seawater medium "f" ((luillard and Ryther. 1962). During the experi- ment, algal material was clearly evident within the larval digestive tracts. (Pre- liminary culturing observations in the early summer of 1978 demonstrated that Artcinia salina is an inappropriate food for /'. lanyipcs larvae.) Larvae of each stage were preserved in ethvlene glycol along with the dead zoeae and exuviae. Pigmentation was observed on both living and preserved specimens. Drawings were made from preserved larvae and exuviae using a camera lucida, calibrated with a stage micrometer. Appendages were dissected with fine needles. Descriptions and illustrations were checked against at least five specimens from each stage of development. RESULTS Pigmentation The location of chromatophores is fairly consistent throughout the zoeal develop- ment of Pinni.va lonyipes. The arrangement of chromatophores is diagramed in Figure 1. The eyes are dark reddish brown in color. A pair of yellow chromato- phores is located on abdominal segment 5 near its attachment to the telson, and a similar pair is found on the posterior-lateral portion of abdominal segment 4 (Fig. 1 A and C, 1 and 2 respectively). A single yellow chromatophore is present on abdominal segments 2 and 3 in a posterior-medial position (Fig. 1A and C, 3 and 4). On the first abdominal segment a yellow chromatophore is situated anteriorly and medially (Figs. 1A and C, 5). A yellow chromatophore is located on the distal portion of the basipodite of the first maxilliped (Fig. 1A and B, 6), and also on the mandible (Fig. 1A, 7). One FIGURE 1. Diagrammatic sketches of arrangement of chromatophores on Zoea II of Pinnixa longipcs : A, side view ; B, frontal view of cephalothorax ; C, ventral view of abdomen. 594 GEORGE D. BOUSQUETTE FIGURE 2. Lateral view of zoeae of Pintiixa longifcs: A, Zoea I; B, Zoea II; C, Zoea III; D, Zoea IV; E, Zoea V. Scale is 0.5 mm. dark red chromatophore is evident on the labrum (Fig. 1A, 8). Various chro- matophores are found on the carapace itself. A large red one appears at the base of each lateral spine (Fig. 1A, B, 9), between the eyes dorsally and ventrally (Fig. 1A and B, 10 and 11 ), and at the base of the dorsal spine in a posterior position (Fig. 1A and B, 12). Also, single large yellow chromatophores occur along the posterior-ventral margin of the carapace below each lateral spine (Fig. 1A, 13), dorsal and anterior to each lateral spine (Fig. 1A, 14), and near the base of each antennule-antenna set (Fig. 1A and B, 15). A large yellow chromatophore is found above the heart and stomach (Fig. 1A and B, 16), while a red chromatophore occurs near the base of each mandible (Fig. 1A, 17). Description uj larval stages Five zoeal stages and one megalopa were observed. The duration of each inter- molt period was as follows: Zoea I, 6-7 days, Zoea II, 4—5 days, Zoea III, 4—6 days, Zoea IV, 6-8 days, Zoea V, 6-9 Days. After 8 days the megalopa had exhibited no further molting. The average time required to reach the megalopal stage was 26.75 days (n = 4, s.d. = 0.9574). The major morphological character- PINNIXA LONGIPES LARVAL DEVELOPMENT 595 FIGURE 3. Frontal view of the carapace of Pj;jHi.ra longipcs zoeae : A, Zoea I ; B, Zoea II ; C, Zoea III; D, Zoea IV; E, Zoea V. Scale is 0.5 mm. istics of each developmental stage are detailed below : Zoea I (Fig. 2A, 3A). Dorsal, rostral, and lateral spines are well developed. Minute plumose setae arise between the dorsal and lateral spines (Fig. 2A, 3 A). The eyes are sessile. The abdomen (Fig. 4A) is composed of a telson and five segments. Segments 2 and 3 have rounded lobes projecting laterally. The fifth segment has lateral extensions which overlap the telson. Each furcal ramus of the telson bears three serrate setae on its inner margin. The medial and lateral margins of each ramus possesses a row of very fine acuminate setae. A thickening overlies the anal opening. The antennule is conical in outline with two long aesthetascs and one short acuminate seta located terminally (Fig. 5A). The antenna has an elongated protopodite with two opposing rows of minute serrations on its distal half. From the proximal portion of the protopodite an acuminate seta projects laterally (Fig. 5F). The mandibles are symmetrical with two incisors located ventrally and a molar process located dorsally (Fig. 6A). Very fine acuminate setae are found on the proximal margin of the coxal and basal endites of the maxillule (Fig. 5K) and the maxilla (Fig. 5P). On the maxilla, these setae are also found on each side of the endopod, and entirely fringing the scaphognathite except where plumose setae are located. The first and second maxilliped are pictured in Figures 7A and 8A respectively, while the third maxilliped has not developed. 596 GEORGE D. BOUSQUETTE FIGURE 4. Dorsal view of the abdomen of Pinnixa longipcs zoeae : A, Zoea I ; B, Zoea II ; C, Zoea III; D, Zoea IV; E, Zoea V. Scale is 0.2 mm. Zoea II (Fig. 2B, 3B). The characteristics of the cephalothorax and abdomen are similar to Zoea I except that the eyes are now stalked. Figure 5 illustrates the changes in form and setation which have occurred on the antennule, antenna, maxillule, and the maxilla. The setal arrangements found on the first and second maxillipeds are shown in Figures 7 and 8, respectively. The third maxilliped has not yet developed. Zoea III (Fig. 2C, 3C). The cephalothorax is similar to the preceding stage except that there are now four plumose setae on the posterior-ventral margin of the carapace. A long, acuminate seta arises dorsally from the distal portion of the first abdominal segment (Fig. 4C). A bulge occurring at the base of the serrated portion of the protopodite of the antenna represents the endopodite bud (Fig. 5H). The third maxillipeds are still undeveloped. Zoea IV (Fig. 2D, 3D). Seven plumose setae are now evident on the posterior- ventral margin of the carapace. Pleopod buds are visible for the first time, with a pair found on each of segments 2-5. Another incisor has been added to the two already present ventrally and a fourth is now evident on the same plane, but is more dorsal in position (Fig. 6D). A medial extension of the coxa is evident on the second maxilliped of this stage (Fig. 81) ). Rounded buds of the third maxilliped and pereiopods are beneath the carapace (Fig. 2D). The claw of the first pereiopod is clearly bifurcated, as is the third maxilliped (Fig. 9A). PINNIXA LONGIPES LARVAL DEVELOPMENT 597 Zoca r (Fig. 2E. 3E). The cephalothorax is similar to that in the preceding stage except that 12 plumose setae are now present on the posterior-ventral margin of the carapace. The pleopod buds have increased in si/.e ( Fig. 9B). The mandible now lias five teeth present on one plane, three ventral and two dorsal, while the more dorsally located molar plane has acquired four teeth at its margins (Fig. 6E ) . The third maxilliped and pereiopod buds have greatly increased in size (Fig. 9C). Indications of segmentation are visible in the third maxilliped and the pereiopods. FIGURE 5. Antennules, antennae, maxillules, and maxillae of Pinnixa loni/ipcs zoeae : A-E, antennules of Zoea I-V; F-J, antennae of Zoea I-V ; K-O, maxillules of Zoea I-V ; P-T, maxillae of Zoea I-V. Scale is 0.2 mm. 598 GEORGE D. BOUSQUETTE FIGURE 6. Mandibles of Pinnixa longipcs zoeae : A, Zoea I; B, Zoea II; C, Zoea III; D, Zoea IV; E, Zoea V. Scale is 0.2 mm. Megalopa (Fig. 10A-C, 11). The carapace of P. longipes is \l/2 times wider than it is long, with a squared rostrum separating the stalked eyes. The dorsal surface of the carapace is smooth except for the presence of 42 minute, acuminate setae and one small plumose seta just posterior and lateral to each eye (Fig. 10A). FIGURE 7. (Left) First maxillipeds of Pinnixa longipcs zoeae: A. Zoea I; B, Zoea II; C, Zoea III ; D, Zoea IV ; E, Zoea V. Scale is 0.2 mm. PINNIXA LONG1PES LARVAL DEVELOPMENT FIGURE 8. (Right) Second maxillipeds of Pinnixa longipcs zoeae : A, Zoea I; B, Zoea II ; C, Zoea III ; D, Zoea IV; E, Zoea V. Scale is 0.2 mm. Twenty-five plumose setae are in evidence along the length of the ventral border of the carapace starting near the cheliped and extending to the fourth pereiopod (not pictured). The mandible consists of a cutting edge with five teeth and a basal palp which bears 14 small, plumose setae terminally and a longer one subterminally (Fig. 10D). The very fine acuminate setae found on the maxillule and maxilla of the zoeal stages are no longer evident. The dactylus of the first pereiopod has a medial lobe facing the longer fixed finger of the propodus, which has a subterminal notch into which the tip of the dactylus fits (Fig. 10A). Proceeding from the second pereiopod to the fifth they become more paddle-like. The fourth pereiopod bears two small, rounded lumps on the posterior margin of the ischium, and seven more each on the posterior and anterior margins of the merus. Two similar protuberances are located on tin- posterior margin of the ischium of the fifth pereiopod along with a single one on its FIGURE 9. Third maxillipeds, pereiopods, and pleopod buds found on Pittnixa lon(/if>cs zoeae: A, buds of third maxilliped and pereiopods on Zoea IV; B, pleopod buds on Zoea V; C, third maxilliped and pereiopod buds on Zoea V. Scale is 0.2 mm. 600 GEORGE D. BOUSQUETTE FIGURE 10. Megalopa of Pinnixa lonc/ifcs: A, dorsal view; B, lateral view of abdomen; C, thoracic sternum; D, mandibles; E, antennule ; F, antenna; G, maxillule ; H, maxilla; I, second maxilliped ; J, first maxilliped. Scale is 0.5 mm for A-C, and 0.2 mm for D-J. anterior margin. Two more are found on the posterior margin of the merus of the fifth pereiopod. The pleopods are found on abdominal segments 2-5 (Fig. 12A-D). Two hook setae are located terminally on the endopod of each pleopod. DISCUSSION Pinnotherid crabs have attracted considerable attention through the years, primarily due to their highly modified and varied commensal relationships with PIN NIX A LONGIPES LARVAL DEVELOPMENT 601 FIGURE 11. Third maxilliped of Pinnixa lonyipcs megalopa. Scale is 0.2 mm. other invertebrates. In recent years, the need for accurate identification of plank- tonic larval forms has been recognized by ecologists undertaking population and community research. Unfortunately, except for a few species infesting hosts of commercial importance, there have been few investigations of the larval develop- ment of pinnotherids. Larval descriptions are presently available for 24 species in nine genera within the Pinnotheridae (Table I). Of these species, the complete larval development is known for only 11 species, 2 of which belong to Pinnixa. Based upon morphological characters little difficulty is encountered in separating the zoeae of Pinnixa from the other Pinnotherid genera. P. longipes exhibits the unique fifth abdominal segment which Hyman (1924) reports to be characteristic for the genus. The lateral and posterior extension of this segment was noted for /'. chactopterana and P. sayana by Faxon (1879), and for P. rathbuni by Sekiguchi (1978). It is evident from the drawings and description supplied by Irvine and Coffin (1960) that Fabia subqnadrata also possesses this special abdominal modifi- cation, and is dissimilar in only a few characteristics from P. longipes. The zoeae of P. longipes have two setae on the first two segments of the endopodite of the first maxilliped, and a small acuminate seta on the antennule in addition to the aesthetascs. On the other hand, Fabia snbqiiadrata, which has been reported from Alaska to San Diego, California (Schmitt ct ai., 1973), has only one seta on the first two segments of the endopodite of the first maxilliped, and no acuminate seta on the antennule. In addition, Irvine and Coffin (1960) report a humped condition in zoeae of F. subqitadrata. They pictured this as lateral ridges on the carapace extending between the eyes and dorsal spine. Faxon (1879) and Hyman (1924) described the first zoeal stage of P. chaetoptcrana, whose range extends from Massachusetts to Brazil (Schmitt ct a/., 1973), making its occurrence with P. longipes unlikely. One particular character, FIGURE 12. Pleopods of Pinnixa longipes megalopa: A, first pleopod ; B, second pleopod ; C, third pleopod ; D, fourth pleopod. Scale is 0.2 mm. 602 GEORGE D. BOUSQUETTE TABLE I Sunismn (Decapoda, Brachyura). Crtistaceana Suppl. (I.ieden), 0 (Suppl. 5) : 22-30. RICK, A. L., 1975. The first zocal stages of Cancer pn-18 months, and which presented well defined growth rings on their shells. Mean diameter and standard deviation of growth rings formed during the ob- servation period are shown in Table II. Disturbance rings, assumed to be due to markings, were disregarded. Growth and age were calculated following the methods described in Kicker (1975). Walford's graphic (\Yalford. 1946) was constructed using mean ring sizes to calculate maximum or asymtotic size. The least square regression of Walford plot has the general linear form: Y = a + bX. This expression becomes: Lt+i = L«(l • - k) + kLt. Here X -- Lt, the size at time t; b -- k, the slope of the Walford regression line; a -- L (1 --k), the Y-axis intercept from which L, the asymptotic size, can be calculated in the following derived expression : L* = ( - - L oc (1 - k))/(k - 1). The von Bertalanffy curve for Fissurclla crassa was fitted using the parameters obtained in the Walford regression. This curve is represented by the equation (von Bertalanffy, 1938) : Lt = _ e-K(t+to)) Sexual dimorphism in ring formation or growth in the experimental group could not be studied because sex could not be determined from external morpho- logical features. Control groups Group A: A sample of 99 animals collected in the intertidal rocky shore of Los Lobitos (20° 28' S) (Fig. 2) ranged from 19.8 to 59.8 mm in shell length, and 88% of them measured less than 47 mm. Shell lengths and ring sizes were TABLE II Growth ring diameters (and standard deviation) detected in Fissurella crassa from Huayquique, Los Lobitos, and Los Verdes. Sizes underlined are assumed to correspond to annual growth rings. Los Lobitos Huayquique Los Verdes Growth Mean Mean Mean Estimated rings n diameter s.d. n diameter s.d. n diameter s.d. age — years (mm) (mm) (mm) I _ _ _ . — — — — — II 65 26.0 ± 2.7 • — — — — — — 1 III 28 35.1 ± 1.9 5 35.8 ± 1.5 47 36.3 ± 2.0 U IV 8 43.3 ± 1.7 17 44.5 ± 2.5 104 43.5 ± 2.0 2 V 5 51.4 ± 2.0 30 51.2 ± 2.4 109 50.<; ± 2.5 2k VI 1 56.6 20 58.1 ± 1.8 112 57. S ± 2.1 3 VII — — 6 63.5 db 1.6 70 63.5 ± 1.3 3i VIII — — — 6 68.5 ± 1.3 87 6S.4 ± 1.3 4 IX 3 71.9 ± 0.5 44 7J.4 ± 1.1 4* X — — — 2 75.7 ± 1.3 28 75.4 ± 0.9 5 XI — • — — 1 78.0 — 20 7S.2 ± 1.0 5* 610 MARTA BRETOS 100 80 E E 60 20 L, = 94 5U8 G-. - 0 1590 ( t + 1 9999 ) JTTI. 234 6 AGE (YEARS 10 FIGURE 3. Growth curve of shell length for Fissurella crassa at Huayquique. measured as in the experimental group. Mean ring sizes and standard deviation were calculated (Table II). Group B: It was impossible to find large living specimens of F. crassa due to man's intensive predation on this species in northern Chile. For this reason, a second comparative group was analyzed for growth ring sizes in middle- and large-sized specimens. Fossil shells of F. crassa were collected in an archaeo- logical site at Los Verdes (20° 26' S) (Fig. 2). This sample of 189 shells ranged from 44.7 to 90.3 mm in shell length. Shell lengths and growth ring sizes were measured as indicated above. Mean ring sizes and standard deviation were calculated (Table II). RESULTS A disturbance ring formed on shells marked with notches. These rings were not considered in age and growth estimates. Two periodic rings were formed per year in Fissurella crassa at Huayquique, except in one case where only one ring was formed during a year (asterisk in Table I). Ring formation occurred usually in winter and summer (45 cases), rarely in spring or/and autumn (4 cases) (Table I). Formation of rings III-XI was observed during the study period in the ex- perimental group at Huayquique. Sizes of these rings were used to determine age and growth. The Walford regression obtained using these data was: Lm = 13.9007 + 0.8529 Lt, with correlation coefficient = 0.999 significant at P = 0.005. Growth and age of Fissurella crassa calculated according to the von Bertalanffy equation are shown in Figure 3. Asymptotic shell length of this species at Huayquique was estimated at 94.5 mm. According to these results, the animals in the experimental group were 1.5- 5. 5 years old (Table II). Shells of animals of the group from Los Lobitos showed rings II and III, and occasionally rings TV, V, or VI (Table II). Rings III-XI were distinct on shells from Los Verdes (Table II). Other rings were also present in this group, but deciphering them was confusing. In middle- and large-sized specimens, the AGE DETERMINATION IN A LIMPET 611 first ring is often obscure or disappears due to wearing away of the shell or to enlarging of the apical hole. Rings beyond the eleventh are not always clearly defined on account of slow growth at this age. Mean ring sizes in the three groups were similar. DISCUSSION Bretos (1978) experimentally determined the growth rate in Fissurella crassa in its natural habitat at Huayquique, Northern Chile. This species shows a shell length increment of 19.8 mm in the second year of life. 14.5 mm in the third, 9.7 mm in the fourth, and 6.8 mm during the fifth year (Fig. 3). These values are similar to the size differences (13.7 mm, 10.3 mm, and 7.2 mm) observed between odd rings formed in the experimental group in the present study (Table II). According to the mean ring sizes obtained in the present investigation, a 1 -year- old F. crassa has a shell length of about 26 mm (Table II). By adding the annual mean shell increment (Bretos, 1978) to the shell sizes at ages 1-5 (Table II), one obtains shell lengths very close to those calculated using the von Bertalanffy equation (Table III). These results agree with those of Bretos (1978), demonstrating that Fissurella crassa exhibits a seasonal growth pattern, with diminishing growth rates during winter and early spring and in summer or autumn, bringing about formation of growth rings. Other molluscs can form two rings per year on their shells, for example, the bivalves Chlamys varia (Dalmon, 1935), Drcissena polyniorpha (Morton, 1969), Corbicida fluminea (Morton, 1977), and Tcllina jabula (Salzwedel, 1979). Picken (1980) reports the formation of two types of bands on the shell of the patellid limpet Nacella (Patinigcra) concinna, dark bands denoting winter shell growth and light bands, summer shell growth ; so that 1 year's growth is repre- sented by a pair of bands. In molluscs with two phases of shell growth, slowing of shell growth in winter can be attributed to a decrease in water temperature (a time when soft tissues can grow faster) (Salzwedel, 1979), whereas the reduced shell growth rate observed in summer could be related to metabolic changes or may coincide with spawning (Dalmon, 1935; Morton, 1969; Salzwedel, 1979). No information is available on reproductive, physiological, or biochemical pro- cesses in the keyhole limpet Fissurella crassa to explain its observed growth pat- tern. It can be concluded that F. crassa at Huayquique, northern Chile, has two growth phases a year, resulting in formation of two periodic growth rings, at TABLE III Comparative shell lengths according to annual growth increments (Bretos, 1978) and mean annual ring sizes. Growth ring numbers are in parenthesis. Age (years) Annual increment (mm) Shell length (mm) Annual ring size (mm) 1 (25.8) (25.8) 25.8 (II) 2 19.8 45.6 44.5 (IV) 3 14.5 60.1 58.1 (VI) 4 9.7 69.8 68.5 (VIII) 5 6.8 76.6 75.7 (X) 612 MARTA BRETOS least in animals from 1-6 years old. Unfortunately, the shell growth ring method of determining age could be applied only over a limited size range, because in older specimens shell growth is slow and rings are hard to identify. The largest specimen analyzed in this study had a shell length of 90.3 mm and a cal- culated age of about 10 years. Longevity of F. crassa in other regions of Chilean coast could be different since most of the Fissurella species living in Northern Chile attain greater sizes in central Chile. ACKNOWLEDGMENTS I am indebted to my husband, Mr. Jose Ignacio Moraga, for encouragement throughout the investigation and for making the figures. Thanks are also due to members of the Developmental Biology and Malacology Group for their help in the laboratory and in the field. This work was supported in part by SERPLAC I Region ( Regional Development Funds ) . LITERATURE CITED BERTALANFFY, L. VON, 1938. A quantitative theory of organic growth (Inquiries on growth laws. II.) Human Bio!.. 10: 181-213. BRETOS, M., 1978. Growth in the keyhole limpet Fissurella crassa Lamarck (Mollusca: Archaeogastropoda) in Northern Chile. The Veligcr, 21 : 268-273. BROUSSEAU, D. J., 1979. Analysis of growth rate in My a arenaria using the von Bertalanffy equation. Mar. Biol., 51 : 221-227. COE, W. R., 1947. Nutrition, growth and sexuality in the pismo clam (Tivcla stultorum). J. Exp. Zool. 104: 1-24. DALMON, J., 1935. Note sur la biologic du petoncle. Rev. Trav. Off. Pechcs maril., 8: 268-281. GORDON, J., AND M. R. CARRIKER, 1978. Growth lines in a bivalve mollusk : Subdaily pat- tern and dissolution of the shell. Science, 202 : 519-521. JONES, D. S., I. THOMPSON, AND W. AMBROSE, 1978. Age and growth rate determinations for the Atlantic surf clam Spisula solidissima ( Bivalvia : Mactracea), based on in- ternal growth lines in shell cross-sections. Mar. Biol., 47 : 63-70. MORTON, B. S., 1969. Studies on the biology of Dreissena polymorpha Pall. 3. Population dynamics. Proc. Malacol. Soc. Land., 38 : 471—482. MORTON, B. S., 1977. The population dynamics of Corbicula fluminca (Miiller, 1974) (Bivalvia: Corbiculacea) in Plover Cove reservoir, Hong Kong. J. Zool. Land., 181: 21-42. NEWMAN, G. G., 1968. Growth of the South African abalone llaliotis inidae. Invest. Rep. Div. Sea Fish. S. Ajr., 67 : 1-24. PICKEN, G. B., 1980. The distribution, growth and reproduction of the Antarctic limpet Nacclla (Patinigera) concinna (Strebel, 1908). /. Exp. Mar. Biol. Ecol., 42: 71-85. RICKER, W. E., 1975. Computation and interpretation of biological statistics of fish popula- tions. Bull. Fish. Res. Bd. Can., 191 : 1-382. SAKAI, S., 1960. The formation of the annual ring in the shell of the abalone Haliotis discus var. hannai Ino. Tohoku J . Ayric. Res., 11 : 239-244. SALZWEDEL, H., 1979. Reproduction, growth, mortality, and variations in abundance and biomass of Tcllina falntla (Bivalvia) in the German Bight in 1975/76. VeroQ. Inst. Mceresjorsch. Brcmcrh., 18: 111-202. WALFORD, L. A., 1946. A new graphic method of describing the growth of animals. Biol. Bull., 90: 141-147. WILBUR, K. M., AND G. OWEN, 1964. Growth. Pp. 211-242 in K. M. Wilbur and C. M. Younge, Eds., Physiology of Mollusca. Volume 1. Academic Press, New York. Reference: Biol. Bull., 159: 613-625. (December, 1980) COURTING BEHAVIOR IN A SYNCHRONOUSLY FLASHING, AGGREGATIVE FIREFLY, PTEROPTYX TENER JAMES F. CASE Department of Bioloi/ical Sciences, University of California, Santa Barbara, CA 93106 ABSTRACT Display, courting, and mating behavior of the aggregative, synchronously flash- ing firefly Pteroptyx tencr were analyzed with the aid of high-gain video recording. Males flash synchronously with other males under a variety of conditions, including flight and many phases of courtship, competition with other males, and mating. The courting male perches on the back of the female, rocks backwards and forwards, flashes his lantern directly into her eyes, and strikes her abdomen with his hind pair of legs. Characteristic flash exchanges occur before copulation, which is also accompanied by flashing by both sexes. The behavior of interloping males and the counter-behavior of the primary male are described. The import of these behaviors is discussed in relation to theories of the evolution and adaptive significance of synchronous flashing. It is suggested that the brilliant illumination of the female's eyes by the male may prevent her from seeing signals from other males or they from recognizing her emission. INTRODUCTION Synchronous flashing by large groups of male fireflies of several species of the Asian genus Pteroptyx is associated with aggregation of males and females together in trees and is believed to facilitate mating. The flashing has been studied visually and recorded from individuals and groups of fireflies in field and laboratory (Bassot and Polunin, 1967; Buck and Buck, 1968; Hanson, 1978; Hanson et al., 1971; Lloyd, 1973b) and several speculations concerned with how flash synchronization might function during courtship and mating have been advanced (Buck and Buck, 1966, 1978; Lloyd, 1973a, 1979). However, aside from occasional sightings of antennation, mounting, or copulation, no observations have been made of actual courtship, here defined as interactions between male and female after physical contact and before intromission. The dearth of crucial data about reproductive behavior in habitually synchro- nizing firefly species is easily understandable. In the first place the fireflies are minute (most species are 5-7 mm long, with light organ areas of only a few mm- ; Figs, la, b). Second, flash duration is usually very brief (40-100 msec). Third, the light, though of low absolute intensity, is dazzling when it impinges suddenly on the dark -adapted human eye at close range. Hence, even if the observer is not distracted by the simultaneous flashes of the many nearby fireflies in the swarm, the problem of focusing on a tight pair of small insects at a somewhat uncertain point in space during an instant of illumination in otherwise total darkness is considerable. To record sexual behavior instrumentally has also proven impossible, prior to the present investigation, because of inadequate spatial resolution. These Received July 25, 1980; accepted September 25, 1980. 613 614 JAMES F. CASE k c FIGURE 1. Individuals of Ptcroptyx tcner photographed in darkness with an electronic flash (X5). (a) Male (5.2 mm body length), ventral view, showing single luminous organ covering surface of sixth abdominal segment and two smaller, roughly triangular lateral organs in the seventh segment. (The mostly white fifth segment is not luminous.) (b) Female, ventral view showing luminous organ occupying sixth abdominal segment, (c) Two males, ventral view. Male at bottom of picture is displaying to the other male by twisting his abdomen so as to direct the light at the other, (d) Courting pair, ventral view. The male is standing on the back of the female in head-to-head alignment, and has flexed and twisted his abdomen almost 180° to bring his light organs close to, and almost directly over, the female's eyes, (e) Courting pair, dorsal view, illustrating how the male's abdomen (flexed to the readers left) is placed close to the female's eyes and showing also that neither of his hindmost pair of legs is used to hold on to the female, (f) Courting pair, rear end view. Male's body axis is turned about 15° to the left with respect to the female's and his abdomen is flexed forward (reader's left) and twisted to aim the light organ toward the female's head. Again the metathoracic legs are seen not to be used in grasping. technical difficulties have now been overcome by the twin expedients of visual observation and video targeting via red light, to which fireflies are essentially blind, and of recording luminescence via a portable, high sensitivity, close focus, video camera. These techniques have revealed several close range male courting and competitive behaviors that are not only remarkably complex but appear to provide one or more plausible rationales for the function of flash synchronization in courtship and mating. FIREFLY COURTING BEHAVIOR 615 MATERIALS AND METHODS Fireflies, rtcropty.v tcncr (Olivier) (Figs. la. 1>), were observed, and collected alive for darkroom study, from aggregations occurring along the lower (tidal) 10 km of the Kuala Selangor River, Selangor, Malaysia, in March, 1980. The fire- flies were viewed from a boat and from a platform built in the dense riverside vegetation where large nightly displays of /'. tencr occur. Light emission and other behavior was recorded with a Panasonic WV-1350A video camera (Matsushita Communications Industrial Co. Ltd.) equipped with a Panasonic "Newvicon" video tube having a light sensitivity of about 0.03 fc in the visible spectrum and a markedly greater, but unmeasured, sensitivity in the far red. Fitted with a 16 mm lens of f/1.6 aperture, this equipment recorded flashing behavior up to a distance of about 4 m. By moving the video tube relative to the lens, focus could be varied from about 5 cm to infinity. Illumination of the scene with a 1.5 V flashlight provided with a deep red filter (Wratten No. 89B, 670 nm cut-off ) allowed excellent viewing of behavioral details with little or no disturbance of the insects, as might be expected from the observations of Lall ct al., 1980, who found that the sensitivity of the compound eye of the male of Pho thins pyralis at 650 nm is only about 5% of its maximum value at 570 nm. Video data and oral comments were recorded on a Panasonic VHS recorder and monitored with a Sony AVF-3250A electronic viewfinder. The entire system was powered by an AC converter and a 12 V car battery. For viewing luminescence at farther than 4 in the video camera was coupled with an image intensifier ( Varo, Inc. ) providing a light gain of about 40,000. It proved very difficult to document detailed and complete behavioral episodes in the field, owing to wind movement of vegetation and to the incessant movement of other fireflies near perched, displaying males. However, it was found that after fireflies had been in light for a few hours, exposure to darkness at any time of day triggered normal flashing activity. Consequently, detailed recordings were made in a darkroom and confirmed by visual observation and video recording of parts of behavioral sequences in the field. Fireflies used in the darkroom studies were netted just before we left the river each evening and transported in large plastic bags. Within a few hours males and females were separated and left overnight in light. The next day in the darkroom appropriate numbers of males and females were released into a terrarium supplied with fresh foliage. Synchronous flashing began within 0.5 hr of darkening the room; courting and mating soon followed. Darkroom temperature fluctuated around 21 °C while evening temperatures on the river were in the range of 27°-28°C. Still photographs of insects were made with a hand-held 35 mm camera and 105 mm macro lens by triggering a strobe while focusing on flashes in darkness. The strobe was never used near a firefly swarm under investigation and fireflies illum- inated by the strobe were never used again. Temporal relationships of flashing by courting and mating pairs were analysed by the following procedures. Appropriate video-taped flashing sequences were displayed on a video monitor at approximately 5x magnification. Hand-held 0.5- mm-diameter light guides (Edmund Scientific) were positioned to record images of the light organs of two fireflies. The light guides conducted light to a pair of photomultiplier tubes whose outputs led to a Grass polygraph through current to voltage converters. Adjustment of the video brightness and contrast allowed recording of remarkably clean signals. Time resolution was limited to 16 msec by 616 JAMES F. CASE the video frame refresh time. This was a small price to pay for the tremendously increased spatial resolution of the method, which made it possible to record sep- arately the emissions of two light organs a few mm apart from a distance of 20 cm. Both in 1979 and 1980 Pteroptyx tencr males and females were also studied in a darkroom in Santa Barbara, where they lived and behaved normally for up to 6 weeks. The observations described below are based on 4 hr of video recording in the rield and in the Kuala Lumpur darkroom. Four complete matings were observed, three in the darkroom and one in the field. Many incomplete courtships, male to male display encounters, and interactions between interloper males and courting pairs were recorded in both field and darkroom. RESULTS General aspects of behaznor Synchronous flashing becomes established rapidly at dusk as males and females, with little flashing, fly directly out from river bank vegetation and assemble in trees along the water's edge, most frequently the mangrove Sonneratia cascolaris. The first flashes were seen at about 7:15 p.m. local time, when light on the open river under a cloudless sky was about 0.4 ,uW/cm- as measured with an upwards-directed photometer ( United Detector Technology ) . These flashes were from males, usually perched at the tips of leaves. The frequency of flashing was about 3.7 flashes/sec (period 0.27 sec) at the typical sunset environmental temperature of 28°C. This is at or near the frequency of the full mass display by thousands of insects up and down the river, which is established within the next 0.5 hr and continues for much of the night. The synchronized flashing is produced wholly by males and is highly precise, typically being visible only in the first, 22nd, and 23rd frames of each successive 23 video frames. The synchronized males in the trees engage in two types of flashing, differing in flash intensity and duration, and in body attitude while flashing, but not in frequency. Much time is spent sitting or walking slowly, making rhythmic single flashes with the abdomen held parallel to the leaf surface. The emission in these flashes is relatively dim, may be as short as 70 msec, and may involve only the most anterior parts of the lantern (Fig. la) or in any event be initiated in the anterior organ. At irregular intervals the male changes to display flashing, which occurs usually with the male standing prominently on the edge or tip of the leaf, and in which the abdomen is twisted longitudinally and usually somewhat flexed so that his light is directed not down toward the leaf surface but laterally. During display flashing the intensity of light is much greater than in non-display flashing. The flashes may be up to 300 msec in duration and be multiple-peaked, so that the result is a virtually continuous, pulsating light emission (Fig. 2). Video records show that the multipeaked display flash is initiated by a low intensity flash from the most posterior parts of the lantern followed by a brilliant flicker of the whole organ. Displaying to nearby males is a major male activity. During such display the light organ is aimed so that its surface is roughly perpendicular to a line from the displaying male to the other. The aiming is usually lateral (Fig. Ic) but a different direction was seen in a video sequence showing a male displaying while running to a pair in courtship. In this instance the approaching male directed his light forward FIREFLY COURTING BEHAVIOR 617 FIGURE 2. Photomultiplier record from video record of a male displaying in the field. Each comiXHincl flash begins as the the preceeding one dies away. Avrrage Hash duration : 0.28 sec; period = 0.28 sec; time marks =- 1 sec. towards the pair while the male of the pair displayed towards him. The trigger for male-male display can he purely photic, as a captive male will display at his own reflection from a mirror. When not responding to a nearby male, a displaying male aims his light away from the mass of synchronizing fireflies. Individual males do not display for more than a few minutes at a time but whether displaying or not displaying, and some- times even when flying (Fig. 3), they maintain flash synchrony with the other males of the swarm. The only exceptions are rapid flickers made when they are disturbed, and certain brief episodes of flashing during courtship (see below). Females never flash in synchrony with males. They produce single-peaked flashes lasting up to nearly 1 sec, usually at a much slower rate than the male's display frequency and nearly always in an irregular pattern (Fig. 4). Such be- havior is markedly obvious in the mass of displaying insects. Females may emit continuous glows when mechanically disturbed. Although the female light organ spans the width of the abdomen (Fig. Ib) the entire organ is virtually never active. Usually only the outer edges flash, giving the impression of a pair of luminous dots (Fig. 7g). Further details of the biology of P. tcner will be presented elsewhere. FIGURE 3. Flash synchrony in flight. Frame at upper left is a composite video record showing five successive flash positions (largest images) as closer male flew from top to bottom of field, flashing in synchrony with several perched males. The other three photographs are of single video frames from the sequence, showing that the flying male flashed each time in coincidence with the perched individuals. None of the video frames separating the frames showing flashes (there being about 22 frames between episodes) showed any flashing. 618 JAMES F. CASE FIELD RECORD MALE AT A DISTANCE 1 --. I ^ ' •*•—/ ^~^_/ ^--~__/ ~^~~~^J X J X-^_ FEMALE KL LABORATORY MALE INCOMPLETE COPULATION ATTEMPT FEMALE SECONDS FIGURE 4. Luminescence of female in relation to different male behaviors. Top two traces show flashes of a non-displaying hiale and concurrent flashes of a female 10 cm distant from him. Bottom two traces show display flashes of mounted male, including those during a copulation attempt, and the concurrent luminescence of the female. Flat flash peaks in second trace are artifacts and the slow baseline shifts in all records are caused by variation in video background illumination. Courtship and mating Courtship begins when the male, flashing synchronously with the swarm, or not flashing at all, rapidly approaches a female, who may or may not be flashing. Video records show that walking males tend to make their first contact with females by touching their heads to the end of the female's abdomen. He climbs upon her back with his head facing in the same direction as hers and maintains his perch with only his first two pairs of legs. His abdomen is acutely flexed and twisted so that his light organ is held just above the head of the female (Figs. Id, e). He continues or begins flashing, along with frequent bouts of waving his light organ over the female's head, while simultaneously striking her abdomen with his last pair of legs. The male's abdomen is held a short distance from the head of the female (Fig. If) but never seems to touch her until copulation is attempted. These acts are accompanied by rapid fore and aft swaying of the male's entire body parallel to the female's body axis. In the darkroom, courtship sometimes lasted as long as half an hour before ending in copulation or in separation of the pair. During all this male activity the female may stand quietly or may walk about. She may luminesce or not. When luminescing she sometimes increases her flash frequency even above that of the male, and at times her flashes may roughly alternate with the male's emissions (Fig. 5). In the field this joint luminescence is often what first brings to notice the presence of a courting pair. The female seems to make few other visible responses to attentions from the male, and within 30 sec after his departure from an aborted courtship her flashing has slowed to well below the typical male frequency. In one video sequence the female used a leg to rub her eye and then to push at the male's brilliantly flashing light organ. Occasionally the abdominal striking seemed to cause the female to lift her abdomen. This did not seem to be strongly correlated with attempted copulation, however. Courtship is interspersed with copulation attempts. Initiation of an undisturbed copulation attempt is signalled by 2 or 3 sec of darkness followed by a short burst FIREFLY COURTING BEHAVIOR ALTERNATE FLASHING DURING COURTSHIP 619 FEMALE MALE SECONDS FIGURE 5. Flashing of a courting pair, showing occasional flash coincidence and more frequent alternate flashing of male and female. of especially brilliant and more slowly delivered flashes by the male, whose lantern still is held over the head of the female (Fig. 6, third trace). The female commonly begins to flash, if she is not already doing so, just before the male's brief dark period (Fig. 6, fourth trace). The male, continuing to flash intensely, then rapidly moves his abdomen to cover the female's terminalia from above and attempts to copulate (Fig. 7). Unsuccessful copulation attempts last only a few seconds before the male instantly returns to brilliant flashing and vigorous striking at the female's abdomen for a brief interval before trying again. No consistent cue for an undisturbed male to switch from flashing at the female's head to attempting to copulate was detected. The one definite trigger to an attempt is approach of another male, which invariably causes the courting male to increase the intensity of his flashes (Fig. 8) and to rapidly alternate the abdomen between the head flashing posture and the copulation initiation position, in which the male terminalia cover the female's from the dorsal aspect. LUMINESCENCE ASSOCIATED WITH MATING MALE FEMALE SECONDS COPULATION FEMALE FIGURE 6. Luminescence during SO sec preceding mating. Upper pair of records con- tinuous with lower pair. Male maintained continuous low intensity flashing after an initial episode of bright flashes accompanied by abdominal striking (visible on video). Just before copulation his flashing ceased for about 3 sec immediately before four maximal flashes occurring simultaneously with his abdominal movements to effect copulation. Subsequent to these flashes both male and female flashed brightly but their contributions could not be resolved separately by the light guides. The four bright flashes by the female just before copulation were unique to this mating. From video recording in Kuala Lumpur laboratory. 620 JAMES F. CASE FIGURE 7. (a-f) Single video frames of the copulation recorded in Figure 6. (a) male perched on female with lantern held over her head, neither flashing; (b) male flashes, with his light appearing paired since it is seen through and around his abdomen; (c) female flashes; (d) both flash; (e) the instant of copulation with both members of pair flashing as interloper arrives from right; (f) pair rotates still flashing, male at lower left, female center, interloper at right, flashing and about to initiate courtship, (g-i) Three video frames of another copula- tion viewed through glass from below; (g) courtship position, male's lantern at top; (h) male moves tail towards female genitalia to start copulation; (i) copulation in progress. Once copulation is accomplished, both male and female continue to flash brilliantly, often in alternation. Within a minute the pair rotates to a tail-to-tail position, remaining in copulation for at least several hours. Flashing rapidly subsides and the pair commonly takes up a more sheltered position. They are then difficult to induce to flash, even by an interloping male (Fig. 9). How the male distinguishes between male and female when initiating courtship is not clear. The discriminatory powers of the normal male may not be especially acute, as the field video records show one instance in which a male courted and attempted copulation with the male of a courting pair before disengaging and exiting, still flashing. Similarly, another video record, made in the darkroom on the day after collecting, shows a successful copulation occurring simultaneously with arrival of an interloper who undertook active courtship with the female immediately after the copulators rotated (Fig. 7). This interloper remained with the pair for at least 3 hr. Under laboratory conditions in Santa Barbara, courtships between males were common even when many females were present. The ridden member of such male- COURTING MALE INTERLOPER APPROACHES SECONDS FIGURE 8. Flashing of courting male before (top trace) and after (bottom trace) inter- loper approaches, showing brightening of flashes. FIREFLY COURTING BEHAVIOR 621 INTERLOPER ATTEMPTS COPULATION WITH MATED PAIR MATED PAIR INTERLOPING MALE COPULATION ATTEMPT — . 1 • > . . 1 . , , . I , , , , I , , , , I , , . . I . _ SECONDS FIGURE 9. Second (interloper) male attempts copulation with rotated mating pair. Neither member of mating pair emits any light ( bumps in trace are artifacts caused by ]>en from lower trace). Note that this attempt to copulate with a female in copula has the characteristic form of a normal mating attempt. to-male pairings continued to behave normally, even making typical male displays to passing males. This behavior was not observed in the field, and so may not be normal. While normal synchronizing behavior and low mortality characterized groups of insects maintained in Santa Barbara for up to 45 days, no complete copulations were seen among those insects. Events subsequent to mating remain obscure. Pairs remain in copula for hours, perhaps all night. Both males and females leave the display sites in increasing numbers towards dawn and fly into nearby vegetation just above the high tide mark. Early in the evening females are seen in small numbers along the river bank near the display sites. They are invariably flying slowly and hovering a few inches off the ground while making long, single-peaked flashes. They are perhaps getting ready to oviposit. Males are never seen during the night in such places. DISCUSSION The present study, in showing that Pteroptyx tencr males engage in at least three modes of in-tree flashing, in addition to at least three close range interactions with females, has enlarged the male's known repertory while at the same time showing how much remains to be discovered. The male display flashes aimed outward from the tree are very probably long-range recruiting or convocational signals, as assumed in other synchronizing species, but whether they attract males and females indiscriminately or specifically remains unknown. No instance of a female coming to a male, or indeed engaging in any type of long range or mid- range signaling with a male was recognized. Neither was any clue found as to the significance of the dim, un-aimed flashing which occupies a major part of the male's flashing time. The bright display flashing directed at nearby males, however, is clearly suggestive of a competitive or comparison display that often ends with what the observer takes to be defeat and departure of one individual. It is thus quite similar to the aggressive flashing between pairs of walking P. malaccae males reported by Buck and Buck (1978), but without the actual physical combat seen in that species. In neither Pteroptyx, however, is there any actual evidence that females select males on the basis of such a competition. Rather, its visible con- sequence would seem to be the regular spacing of displaying males through the vegetation. Competition among males displaying to a female on the basis of comparative flash intensity, suggested for P. malaccae by Buck and Buck (1978), seems 622 JAMES F. CASE excluded in P. tcncr, since displaying males were not grouped in relation to a female. Xo behavior suggesting one male preempting another's place in a dialogue (Lloyd's 1973a "interloping") was seen. In fact, no male-female inter- action prior to actual contact was detected in P. tcncr except the possibility that pheromonal attraction draws the male to the end of the female's abdomen just prior to mounting, in the sort of "change in communicative channel" postulated by Lloyd (1973a). However, none of the other possible pre-mounting female behaviors or sexual interactions considered by Lloyd seems to occur. The female of P. tcncr seems singularly inactive once within the tree, in striking contrast with the female's precise photic role in dialogue in many roving fireflies, her active flight in others, and her participation in complex aerial sexual chases in still other species (e.g., Ludola obsoleta; Lloyd, 1972). It is in close range male-female interactions that the present work has yielded the most new information. Courtship in P. tcncr is the most complex yet described and contrasts strongly with behavior in many non-synchronizing species in which the male, after locating the female by photic signaling, may copulate seemingly without further preliminaries and with little or no luminescence. Though there is no evidence of whether synchronizing species derive from non-synchronizers or vice versa, certain aspects of P. tcncr behavior may possibly have analogs among non- synchronizers. Lloyd (1964), for example, observed that the mounted male of Pyractonicna dispersa "tapped the abdominal tergites of an apparently unresponsive female with his parameres. . ." before attempting copulation. If such tactile stimulation is a general releaser of receptivity the P. tcncr male's use of his hind legs for this purpose, rather than his genitalia, could be a compromise forced by the overriding need to hold his abdomen near the female's head. Another parallel may be the long periods of luminescent interactions between male and female in the New Guinea species Ludola obsoleta. Here Lloyd (1972) noted a series of very bright flashes from the male before or while mounting. Although copulation was not seen to occur, mated pairs were found in the tail-to-tail position with both members glowing or flashing. L. obsoleta might be considered intermediate in courting behavior between roving fireflies and the aggregative synchronizer P. tcncr. The genus Ludola contains at least one synchronizer, L. pupttla, and L. obsoleta is somewhat aggregative, though not synchronic. The coupling of abdominal movement with flashing is widespread in the Lampyridae, and lantern aiming by the displaying P. tcncr male may have its roots in the almost universal practice of the females of dialogue species of aiming their lanterns towards the male while answering. Further, Case and Buck (unpublished) have found that the male of Photinns grccni points his lantern away from the substrate while flashing prior to taking flight. However, the holding of the male's lantern over the female's head while flashing is a major element of P. tener court- ship that seems totally without parallel in firefly behavior and invites inquiry into its possible physiological consequences. First, flooding the female's head with light may reduce her dark adaptation and thereby limit her ability to see luminescences of other males. To be sure, the mounted male continues to synchronize with other males, showing that he is not self-blinded, but in Pteroptyx cribcllata it is known that the time of the male's flashing is the time of his minimum sensitivity to visual input (Buck ct al., unpublished). In any case, if the male's eyes are twice as far from his lantern as the female's (Figs, le, f ) his chances of being dazzled would be only one quarter of hers. Second, point-blank discharge, by the mounted male, of long-duration display FIREFLY COURTING BEHAVIOR 623 flashes directly into the female's eyes is likely to be perceived as a sinusoidally vary- ing illumination that would mask dimmer, more discrete flashes of other, more distant, males that might possibly have communicative significance. A third pos- sibility is that the male's flashing might, through recurrent pacemaker resetting (Hanson ct al., 1971; Hanson, 1978) or photic inhibition (Case and Buck, 1963; Case and Trinkle, 1968), alter any possibly attractive flashing by the female to a type less attractive or less obvious to searching males. Thus present records show that under courtship conditions female flashing may speed up and blur together with the male's flashes (Fig. 5). While male and female emissions are then still separable spatially by the human eye, or instrumentally (Fig. 9), a firefly might see this joint luminescence as a single strong male display. Hence the observation of a displaying male running up to a pair engaged in alternated flashing might have had territorial rather than sexual significance. Fourth, strong photic stimulation might be a necessary releaser for female copulatory behavior. This is suggested by the brief set of unusually brilliant and long flashes which immediately precede virtually every copulation attempt (Figs. 6, 9). Though the long period of courtship, and occasionally observed unsuccessful terminations, suggest that testing is going on during the mounted phase of courtship it is not obvious how sexual selection is exercised by the female. Her flashing is erratic but there is a clear tendency for her to increase flash intensity and frequency, often associated with a crescendo of mechanical and photic stimulation by her partner. The only movement, aside from slow walking, that she seems to make is occasionally raising the end of her abdomen. The observation that the female often begins to flash immediately before the brief interruption in the male's flashing that just precedes his attempt to copulate suggests a receptivity cue, but this female behavior is not invariant nor are subsequent male attempts always successful. Furthermore, the increased frequency and brightness of her flashes might just as well be signals to attract other suitors, since her luminescence also increases in the presence of an interloper. Any of the first three suggested effects of head illumination by the mounted male (above) appears to obviate photic selection of males by females in that phase of courtship. Yet as already mentioned, there is no evidence of signaling or selection before the male comes to close range, and it is hard to understand why the female often seems to remain passive when a male, without flashing, arrives and initiates courtship. The fact that the male is constrained to supply such an over- whelming photic stimulus to the female might mean that the ultimate stage of mate selection in Pteroptyx tener has shifted away from one in which the female's evaluation of male photic performance was paramount. Perhaps as a result of increasing population density, such as occurs in Malaysian Pteroptyx swarms, the female would tend to be in a continual state of photic confusion from the barrage of male display signals and thus sexual selection on the grounds of inspection of the luminescences of the male from a distance would break down. This seems a pos- sibility because at least the female of Plwtinns greeni, while engaged in timing photic signals, integrates all those seen without making any spatial discrimination (Case and Buck, 1979). Consequently, the female's photic release threshold might rise markedly, necessitating point-blank stimulation, or the female might shift to non-photic selection criteria which the male must meet in addition to photically masking the female from the light of other males or disguising her flashes. It is hoped that the discovery of the wholly unexpected and unpredicted apparent photic coercion of the P. tener female by the mounted male will stimulate 624 JAMES F. CASE new studies and re-evaluation of observations on other species. Particularly necessary are further efforts to understand the adaptive significance of the synchro- nized flashing of male fireflies. In P. tcncr and certain other species synchrony is all-pervasive in short range, long range, display, and non-display flashing, indicating that it has been genetically selected in a variety of interlocking contexts. However, known differences in the behaviors of different habitual synchronizers (Buck and Buck, 1978) and in their physiological control mechanisms (Hanson, 1978) caution against assuming that all synchronizing species have the same photic reproductive adaptations and behaviors. This investigation has not touched on the cause of aggregation, but since the rhythm of flashing in habitually synchronizing species is controlled by a specialized pacemaker which becomes obligatorily entrained to rhythmic external photic signals (Hanson ct al., 1971 ), mass synchronous flashing depends on mass congre- gation, and the functional significance of males and females coming together in huge numbers is an integral part of the puzzle of synchronous flashing fireflies. ACKNOWLEDGMENTS I am most particularly indebted to John and Elisabeth Buck and to Frank E. Hanson for much help and advice. I gratefully acknowledge my introduction to P. tencr by Dr. Ivan Polunin. In Malaysia I have been ably assisted over the past several years by Thian Heng Lee and, in 1980, by Devinder Kumar. Profs. J. Furtado, Y. C. Siew, and others of the Department of Zoology, University of Malaysia, have been most helpful. Jallaludin, of Kampong Kuantan, Malaysia, orga- nized construction of the observation platform and provided transport on the river. Research support was provided by NSF Grant BNS76-80246, University of California Faculty Research Funds, the FBN Fund, and, in its early stages, by the Office of Naval Research. LITERATURE CITED BASSOT, J. M., AND I. V. POLUNIN, 1967. Synchronously flashing fireflies in the Malay Peninsula. Sci. Rep. Yokosuka City Mus., 13 : 18-22. BUCK, J., AND E. BUCK, 1966. Biology of synchronous flashing of fireflies. Nature, 211 : 562-564. BUCK, J., AND E. BUCK, 1968. Mechanism of rhythmic synchronous flashing of fireflies. Science, 159: 1319-1327. BUCK, J., AND E. BUCK, 1978. Toward a functional interpretation of synchronous flashing by fireflies. Am. Nat., 112: 471-492. CASK, J. F., AND J. BUCK, 1963. Control of flashing in fireflies. II. Role of central nervous system. Biol. Bull.. 125 : 234-250. CASE, J. F., AND J. B. BUCK, 1979. Limitations on firefly mate selection by luminescent signalling. Western Soc. Naturalists, Abstr, 60th Ann. Meeting, 28. CASK, J. F., AND M. S. TRINKLK, 1968. Light-inhibition of flashing in the firefly Photuris missouriensis. Biol. Bull., 135 : 476-485. HANSON, F. E., 1978. Comparative studies of firefly pacemakers. Fed. Proc., 37: 2158-2163. HANSON, F. E., J. F. CASE, E. BUCK, AND J. BUCK, 1971. Synchrony and flash entrainment in a Ne\v Guinea firefly. Science, 174: 161-164. LALL, A. B., R. M. CHAPMAN, C. O. TROUTH, AND J. A. HOLLOWAY, 1980. Spectral mechanisms of the compound eye in the firefly Photinus f>\ralis (Coleoptera: Lampyridae). /. Coinp. Physio!., 135 : 21-27. LLOYD, J. E., 1964. Notes on flash communication in the firefly Pyractomena dispcrsa (Coleo- ptera: Lampyridae). Ann. Entomol. Soc. Am., 57: 260-261. FIREFLY COURTING BEHAVIOR 625 LLOYD, J. E., 1972. Mating behavior of a New Guinea Luciola firefly ; a new communicative protocol (Coleoptera: Lampyridae). Coleopt. Bull., 26: 155-164. LLOYD, J. E., 1973a. Model for the mating protocol of synchronously flashing fireflies. Nature, 245 : 268-270. LLOYD, J. E., 1973b. Fireflies of Melanesia : bioluminescence, mating behavior, and synchronous flashing (Coleoptera: Lampyridae). Environ. Entoniol. 2: 991-1002. LLOYD, J. E., 1979. Sexual selection in luminescent beetles. Pp. 293-341 in Blum, M. S., and N. A. Blum, Eds. Sexual selection and reproductive competition in insects. Academic Press, N. Y. Reference : Biol. Bull., 159 : 626-638. (December, 1980) THE INTRACELLULAR MECHANISM OF SALINITY TOLERANCE IN POLYCHAETES: VOLUME REGULATION BY ISOLATED GLYCERA DIBRANCHIATA RED COELOMOCYTES * CHARLES J. COSTA, SIDNEY K. PIERCE, AND M. KIM WARREN Department of Zoology, University of Maryland, College Park, Maryland 20742 ABSTRACT Glycera dibranchiata is an osmoconforming polychaete that lives in seawater osmotic concentrations ranging from 1366 to 374 mosm/kg. G. dibranchiata uses free amino acids (FAA) to reduce intracellular solute concentrations during hypoosmotic stress : Both body wall tissue and isolated red coelomocytes taken from worms adapted to dilute seawater for 14 days showed substantial decreases in several amino acids when compared to full-strength seawater controls. Measure- ments of cell volume made on isolated red coelomocyte suspensions during acute exposure to dilute media indicate that volume regulation is rapid but incom- plete. The volume of red coelomocytes does not return to control values during 120 min. The volume recovery of the isolated coelomocytes during hypoosmotic stress is accompanied by salinity-dependent efflux of FAA proline and taurine. A comparison of this efflux with the FAA pool in full-strength-seawater acclimated cells indicates that the efflux is accomplished by a selective change in membrane permeability to proline and taurine. INTRODUCTION Euryhaline marine invertebrates which survive large changes in extracellular osmotic pressure have provided excellent models for studying the mechanisms underlying cellular volume regulation (Gilles and Schoffeniels, 1969; Lang and Gainer, 1969a, b; Gerard and Gilles, 1972; Pierce and Greenberg, 1972, 1973; Watts and Pierce, 1978; Amende and Pierce, 1980). Most of these studies used isolated solid tissues as experimental preparations. While the solid tissue prepara- tions have yielded a great deal of information, they have the disadvantages that cell volume cannot be directly measured, and that access of the medium to the cells may vary with tissue thickness. A homogeneous suspension of large num- bers of single cells (coelomocytes or blood cells for example), would solve both the volume-measurement and medium-access problems. However, suitable cells are relatively rare in invertebrates with sufficient salinity tolerance to be of use in cell-volume regulation experiments. One such homogeneous cell suspension can be prepared from the coelomocytes of the glycerid polychaete Glycera dibranchiata. This worm is also very much a euryhaline osmoconformer, tolerating salinities from 12.4 to 46.5%e (Machin, Received July 8, 1980 ; accepted September 7, 1980. Abbreviations used: FAA, free amino acid; ASW, artificial seawater; mosm, milliosmoles per kilogram water; MCV, mean cell volume. 1 Contribution No. 157 from the Tallahassee, Sopchoppy, & Gulf Coast Marine Biological Association, Inc. Supported by the N1H Grant GM-23731. 626 GLYCERA RED CELL VOLUME REGULATION 627 1975). The coelomic fluid of G. dibranchiata contains three readily identifiable cell types. Most numerous, except in gravid individuals, is the nucleated, hemo- globin-containing coelomocyte (hereafter, red coelomocyte). These cells have been described both morphologically (Seamonds and Shumacher, 1972) and physiologically as oxygen carriers (Hoffmann and Mangum, 1970). Second, gametes, either eggs or sperm, are often present in numbers dependent on the breeding condition of the worm. Finally, always present, though in relatively small numbers, are non-pigmented amoebocytes. Machin and O'Donnell (1977) provided a preliminary description of the volume regulatory capacity of Glycera coelomocytes and found these cells able to partially recover cell volume following osmotic swelling during exposure to hypoosmotic media. The mechanism ef- fecting the volume decrease was not examined. As a first approach to identify the volume control mechanism of the red coelomocytes, we have characterized the water balance of both the animal and the isolated cell. We have investigated the effect of acclimation to hypoosmotic seawater on the FAA pool of G. dibranchiata body-wall muscle and of isolated red coelomocytes. We have also examined the volume regulatory responses of isolated red coelomocytes to both acute and long term hypoosmotic stresses. Finally, we have begun to identify the volume recovery mechanism used by G. dibranchiata coelomocytes. The results show that the FAA pool of both body wall muscle and red coelomocytes in acclimated worms declines in a salinity- dependent manner. Further, the response of the red coelomocyte to acute hypo- osmotic stress is a transitory and rapid swelling followed by a substantial volume recovery accompanied by an efflux of the intracellular amino acids taurine and proline. MATERIALS AND METHODS Animals and solutions Specimens of Glycera dibranchiata were purchased from the Maine Bait Co., Newcastle, Maine. The worms were shipped to College Park via air mail special delivery and routinely arrived in excellent condition. Upon arrival, single worms were placed into individual lengths of glass tubing, both ends of which were capped with plastic screen cloth. The screen allowed free flow of water through the tube, while preventing the worms from escaping. The worms in their tubes were placed into full-strength artificial seawater (ASW) (Instant Ocean, salinity = 3S%0 = 1001 mosm/Kg H2O ; mosm hereafter) at 15°C. All experimental ac- climation salinities were made by dilution of ASW with distilled water. The osmotic concentration of all solutions was determined before use witli a freezing- point depression osmometer (Osmette, Precision Systems). Coelomic fluid osmotic concentration Worms were kept in ASW for 48 hr. Then a large group was transferred into 697 mosm ASW and allowed to acclimate to that solution for 48 hr. Care was taken to completely empty each tube and to refill each with the lower salinity water. Next, a portion of this group was similarly moved to 463, 374, and finally to 230 mosm ASW, all at 48-hr intervals. All worms in this last dilution died after 48 hr so no further dilutions were attempted, and the use of 230 mosm ASW as an experimental medium was abandoned. The worms were allowed to ac- 628 COSTA, PIERCE, AND WARREN climate to the other salinities for at least 14 days, during which time no mortality occurred. At the end of the 2-week acclimation period, five animals were removed from each test salinity together with a seawater sample, and the osmotic concentrations of the coelomic fluid from each worm and of the seawater were determined. The coelomic fluid was prepared as follows. The worm was removed from the glass tube, placed on a piece of filter paper and gently blotted to remove external fluid. The worm was then held over a beaker, and the body wall was cut in several places with scissors. Coelomic fluid was allowed to drain into the beaker, and the carcass was saved for subsequent determination of water and amino acid content of body wall tissue (see below). The coelomic fluid was then centrifuged at 5000 X g (4°C) to remove coelomocytes and debris. Occasionally, at this point, an individual sample would show evidence of coelomocyte lysis ; such samples were discarded and replaced. The supernatant was then removed, and the osmotic concentration of that solution determined with the osmometer. Care was taken throughout the procedure to keep samples covered and cold to minimize evapora- tion. The data were analyzed by standard regression techniques, resulting in the least-squares regression line and the 99% confidence belt for the regression line between 374 and 1001 mosm (Steel and Torrie, 1960). The slope of the regres- sion line was compared to the slope of the isosmotic line (1.0) using Student's t test for slopes (Snedecor and Cochran, 1967) . Water and intracellular free amino acids in body wall tissue Portions of the body wall were dissected from the carcass remaining from the above experiment and weighed on an analytical balance. The pieces of body wall were then frozen on dry ice, lyophilized overnight, and re-weighed. Tissue water as a percentage of the wet weight was calculated from the following expres- sion : Tissue water = (wet wt— dry wt) • wet wt'1 • 100. The dry tissue was homogenized in 40% ethanol in a glass homogenizer. The homogenate was brought to a boil to precipitate protein and then centrifuged at 20,000 X g for 20 min. The supernatant was lyophilized, the residue dissolved in an appropriate amount of lithium citrate buffer (pH 2.2) and the amino acid content of this last solution determined with an amino acid analyzer (JEOL, model JLC-6AH). The effect of acclimation salinity on the body wall intracellular FAA pool was tested by analysis of variance and the significant treatment means identified by the Student-Newman-Keuls multiple range test. Free amino acid content of red coelomocytes Another group of blood worms was acclimated to 1001, 695, 463, and 374 mosm AS W in the manner described above. After the 2-week acclimation period the coelomic fluid was collected from five worms in each test salinity, as already described. Hemoglobin-containing coelomocytes were isolated from other coelomo- cytes by centrifugation at 5000 X g (4°C) for 10 min. The red coelomocytes, now free of gametes and amoebocytes, were then resuspended in 2 ml of ASW at the osmotic pressure of acclimation. The cells were recentrifuged (1000 X g, 5 min, 4°C), washed three times and finally resuspended in 0.5 ml of ASW at the salinity of acclimation. At this point considerable lysis of the coelomocytes taken from worms acclimated to the lowest salinity (374 mosm) had occurred. GLYCERA RED CELL VOLUME REGULATION 629 Thus, we were unable to measure accurately the aminn acid concentrations of those cells. Lysis did not occur in coelomocytes from any other salinity. The hematocrit of this final suspension was usually 5-10^. Cell number in a 0.1 -ml sample of each suspension was determined with an electronic particle counter (Electrozone Celloscope, Particle Data, Inc.). The remaining 0.4 ml of each suspension was recentrifuged (5000 X y, 10 min, 4°C) and the supernatant solu- tion replaced with 1 ml of distilled water followed by 1 ml of S0% ethanol. This solution was brought to a boil and then centrifuged at 20,000 X g (4°C) for 20 min. The supernatant solution was lyophilized and the residue dissolved in an appropriate volume of lithium citrate buffer (pH 2.2). The amino acid content of this last solution was determined with the amino acid analyzer and concentra- tions expressed as nmol/10'! cells. We have only reported concentrations of the acidic and neutral amino acids for body wall and coelomocytes, since preliminary analyses indicated that the basic amino acids, excepting arginine, were present in very low concentrations (<0.5 /xmol/gm dry weight or 0.5 nmol/10G cells). Arginine concentrations were higher, but they did not vary systematically with salinity. Data from this experiment were statistically analyzed as in the previous section. Identification of hypotaurine Our preliminary amino acid analyses of both body wall and coelomocyte ex- tracts included a substantial peak which exactly coeluted with the urea peak of our standards. Urea has a low color index with ninhydrin and, if the peak was in fact urea, its contribution to the total osmotically active solute pool in Glycera would be enormous. Amende and Pierce (1978) showed that a similar peak on chromatograms of extracts of the bivalve Noctia ponderosa was not urea but hypotaurine. Therefore, we performed experiments designed to identify this substance in the G. dibranchiata extracts. Using identical methods to those described by Amende and Pierce (1978), we determined that the Glycera "urea" peak was also hypotaurine. Red coelomocyte volume regulation: the acute response Red coelomocytes, collected from G. dibranchiata acclimated to 1000 mosm ASW, were isolated from contaminating cell types as described above. Follow- ing isolation, the red coelomocytes were resuspended in 996 mosm ASW (as formulated by Wilkins, 1972) and their volume regulatory ability determined as follows. Suspensions of coelomocytes with a final density of 10,000-20,000 cells/ml were made in approximately 20 ml of ASW of the appropriate osmotic pressure. During the experimental period, cell volume was measured repeatedly to within 75 //.m3 with a Coulter Counter (model ZB, Coulter Electronics Inc., Hialeah, FL) equipped with a Coulter Channelyser (model C-1000) using a 100 /xm aperture, and calibrated against 18.18-/mi-diameter polystyrene beads. The performance characteristics of the Coulter Counter were unaffected by the temperature-salinity combinations used in this study. Mean cell volume (MCV) at each sampling time was calculated from the Channelyser 100-channel volume-frequency distribu- tion consisting of volume measurements on 15,000-25,000 coelomocytes by the following method : MCV = 2Vifj/2fi, where Vs is the volume of the cells in each channel and f; is the number of cells in the same channel. 630 COSTA, PIERCE, AND WARREN Using this procedure, we determined the volume-regulating ability of G. dibranchiata coelomocytes in two experiments. Each experiment was repeated three times with three different coelomocyte suspensions collected from two or three worms per suspension. In the first experiment, cell volume was measured in coelomocytes taken from worms adapted to 996 mosm ASW and exposed to 996 mosm (control) or 505 mosm AS\Y, at either 5° or 20°C. The MCV was determined 0, 2, 4, 12, and 20 min after mixing with the experimental medium. The data were plotted as the mean change in MCV from the 996 mosm control group at zero time (cell volume changet = MCV,, — MCVt). In the second experiment, cell volume was measured in coelomocytes taken from worms adapted to 996 mosm ASW and exposed to 996 mosm (control) or 592 mosm ASW maintained at 2° or 15°C. The cell volume was monitored for 2 hr, with determinations of MCV at 0, 2, 4, 10, 20, 60, and 120 min after mixing with the experimental medium. After the 120 min measurement was completed, the 592 mosm-2°C cells were warmed to 15°C. These cells were then incubated 20 min longer at 15°C and the MCV determined as above. The data were plotted as the mean change in MCV as above. Statistical significance was determined by analysis of variance, and significant treatment means were identified using the Student-Newman-Keuls multiple range test. Red coelomocyte volume regulation: cell volume in acclimated cells Glycera were acclimated to 1000 or 755 mosm ASW for 14 days. Following the acclimation period, the coelomocytes were collected separately from 14—15 worms in each acclimation salinity. The red coelomocytes were isolated from other coelomic cell types as described above, with all wash and resuspension media at the osmotic concentration of the acclimation medium. The MCV of the red coelomocytes from each worm was then determined by suspending the cells in 20 ml of ASW at the osmotic concentration of acclimation and measuring the cell volume of 15,000-25,000 cells with the Coulter Counter and Coulter Channelyser. The volume regulatory capacity of these acclimated coelomocytes was estimated by suspending cells from each worm adapted to 1000 mosm ASW in 20 ml of 755 mosm ASW. After 3 min, which allows the coelomocytes to attain a near-maxi- mum volume, the MCV was determined as above. Red coelomocyte amino-acid efflux during acute hypoosmotic stress Red coelomocytes were isolated from worms acclimated to 1000 mosm ASW as described above. Using the methods described above for measuring cell volume regulation during acute stress, a small sample from each batch of coelomocytes (3.0 X 107 to 7.0 X 107 cells/batch) was tested to establish a positive volume regulatory response to 592 mosm ASW. Cell batches that did not display a positive response were not used in this experiment. Following this preliminary test, half of the cells from each batch were suspended in 2.0 ml of either 592 mosm or 996 mosm ASW. The cells were incubated for 40 min at 15°C with agitation every 5 min. During the incubation period, quadruplicate 10 /*! samples were removed from each tube for determination of number by the Coulter Counter. At the end of the incubation period, the cells were centrifuged at 1000 X g for 5 min, and the supernatant analyzed for amino acids on the amino acid analyzer. Incidental hemolysis was determined by converting the hemoglobin in the su- GLYCERA RED CELL VOLUME REGULATION 631 pernatant to cyanmethemoglobin (Eilers, 1967) and measuring absorbance spec- trophotometrically at 540 nm. The red coelomocyte pellet was then hemolysed in distilled water, and samples analyzed for amino acid and hemoglobin concentra- tions. All amino acid values presented are corrected for the small amount of hemolysis which occurred in both high and low osmotic concentration media. Statistically significant salinity-dependent changes in amino acid efflux were identified by a paired / test (Steel and Torrie, 1960). RESULTS Coelomic fluid osmotic concentration and body wall tissue water The osmotic concentration of G. dibranchiata coelomic fluid varies directly with the osmotic concentration of the acclimation media over the range of salinities tested (Fig. 1). Although the slope of the regression line describing this rela- tionship (TTCF — 1-005 (TTMED) + 4.8 mosm where CF = coelomic fluid and MED = medium) is not significantly different from 1.0 (P>0.3), the coelomic fluid is not isosmotic with the environment. Rather, the osmotic concentration is sig- nificantly higher (P<0.01) than the environment at every point, averaging 8 mosm hyperosmotic to the acclimation medium. Body-wall-tissue water, as a percent of wet weight, varied by 12.6% over the range of acclimation salinities tested (Table I ) . Body wall and coelomocyte intraccllular free amino acids The intracellular FAA pool of both body wall and coelomocytes decreased significantly (P < 0.005) with lowered external salinity (Figs. 2, 3). Both the 1000 900 800 ~- 700 O 600 O E 500 400- 300- 200 300 4OO 500 600 7OO 800 (mOsm/ Kq H20)Q FIGURE 1. Osmotic concentration of Glyccra dibranchiata coelomic fluid from worms acclimated to 1001, 697, 463, 374, and 230 mosm/Kg H,O ASW. The points represent the mean osmotic concentration measured from five animals at each salinity. The lower limit of the 99% confidence belt (not shown) about the regression line is hyperosmotic to isosmotic over the entire range of salinities tested. 632 COSTA, PIERCE, AND WARREN TABLE I Body-wall-tissue water from specimens of Glycera dibranchiata acclimated to reduced salinity. N = 5 in each case. Acclimation osmotic concentration (mosm/Kg H2O) Body-wall-tissue water (wet wt-dry wt \ - X 100 (±SE ) wet wt / 1001 697 463 374 70.6 (±0.34) 75.5 (±0.09) 80.4 (±0.61) 83.2 (±0.70) composition of the FA A pool, and the amino acids which changed most drastically as a function of the salinity, were different in the two tissues. In the body wall of worms acclimated to 1001 mosm ASW, asparagine, alanine, serine, threonine, and hypotaurine account for 91% of the FAA pool; taurine was less than 2% of the total and proline undetectable. As the external salinity was decreased, the con- centrations of asparagine, alanine, serine, and threonine decreased sharply (Fig. 2) while hypotaurine concentration (mean = 78.9 /xinol/g dry wt) was independent of salinity change. In contrast, the coelomocyte FAA pool in 1001 -mosm acclimated worms con- sisted mainly of taurine, proline, asparagine, alanine, glutamine, and serine (Fig. 3). Taurine and proline together accounted for 54% of the 1001-mosm coelomo- cyte FAA pool. Among the six amino acids composing the bulk of the 1001-mosm FAA pool, proline, taurine, alanine, and glutamine decreased significantly (P < 0.05) with decreased acclimation salinity. Asparagine and serine did not change significantly in a salinity-dependent manner. Hypotaurine content, <20 nmol/10f> cells, was too low to be important as osmotically active solute. 300 100 Glycera body woH 71 m 71 T Tl ffl 1001 697 463 mOsm/ Kg 71 [fin 395 FIGURE 2. Body wall intracellular free amino acid pool of Glycera dibranchiata acclimated to 1001, 697, 463, and 395 mosm/Kg H,-O ASW. The values presented are the mean amino acid content measured from five animals at each salinity. Error bars indicate 1 SE. GLYCERA RED CELL VOLUME REGULATION 633 300r = 200 100 0L t— r n Aspartote— » I •- Glulomme FIGURE 3. a. Two-dimensional radiochromatogram of the medium in which Prochloron was incubated in NaH14CO3 in the light. PW., phenol : water ; B.P.A.W.. butanol, propionic acid: water, b. Two-dimensional radiochromatogram of the ethanol extract of the same cells. 646 C. R. FISHER, JR. AND R. K. TRENCH lysis. Similarly, the difference between the light and dark fixation patterns indi- cates that the 14C fixed in the dark controls was a result of heterotrophic dark fixa- tion and not a small amount of photosynthesis due to light leaks in the apparatus. In other associations involving algae and invertebrate hosts, the in vitro release of photosynthate corroborated evidence of in vivo translocation of photosynthetically fixed carbon (Muscatine, 1967; Trench, 1971a, b, 1974). Trench (1971c) and (1971c) and Muscatine et at. (1972) found that adding homogenates of the tissues of the host animals to suspensions of isolated "zooxanthellae" resulted in an in- creased release of photosynthetically fixed carbon. In the case of Prochloron, the quantity of released carbon never exceeded 0.178 //.gC- (/ig Chi a^-hr'1 (or 5A% of the carbon fixed). By extrapolation, it can be concluded that if the cyanobac- teria function in vivo as they do in vitro, then Prochloron probably contributes mini- mally to the carbon metabolism of its hosts. Several phytoplankters have been shown to release (excrete) from 1 to 17% of the carbon they fix photosynthetically (Hellebust, 1965). Glycolic acid was found to be the major product excreted by Chlorella pyrenoidosa (Fogg, 1962) and by several species of Chlamydomonas (Lewin, 1957). Therefore, as far as the release of photosynthetic products is concerned, Prochloron resembles more closely the free-living phytoplankton than symbiotic algae such as "zooxanthellae" and "zoochlorellae" (see Trench, 1979). Because of the presence of chlorophyll b in Prochloron, it has become somewhat popular to speculate that an organism such as Prochloron may have been the evo- lutionary precursor of the chloroplasts of green plants. In fact the presence of chlorophyll b prompted Lewin (1976) to propose a new Division for these or- ganisms. The search for other characteristics relating these cells to the Chloro- Giutamate Asparlale Glutamme P.W FIGURE 4. Two-dimensional radiochromatogram of ethanol extract of Prochloron incubated in NaH14CO3 in the dark. CARBON FIXATION BY PROCHLORON 647 phyta continues. As far as the patterns of photosynthetic and heterotrophic car- bon fixation is concerned, Prochloron demonstrates characteristics typical of cyano- bacteria (Kremer et al, 1979). The absence of sucrose as a photosynthetic end product also distinguishes Prochloron from chlorophytes (Meeuse, 1962). ACKNOWLEDGMENTS We wish to thank the Director, Dr. John Caperon, and the faculty and staff of Hawaii Institute of Marine Biology for the use of their facilities. We are grateful to Professor L. Muscatine for his advice and suggestions. Partial support for this work was provided by NSF, PCM78-15209 to R.K.T. LITERATURE CITED AKAZAWA, T., E. H., NEWCOMB, AND C. B. OSMOND, 1978. Pathway and products of CO2- fixation by green prokaryotic algae in the cloacal cavity of Diplosoma z'irctis. Mar. Biol., 47 : 325-330. BASSHAM, J. A., AND M. CALVIN, 1957. The path of carbon in photosynthesis. Prentice-Hall, Englewood Cliffs, N. J. 104 pp. FOGG, G. E., 1962. Extracellular products. Pp. 475-489 in R. A. Lewin, Ed., Physiology and biochemistry of algae. Academic Press, New York. GIDDINGS, T. H. JR., N. W. WITHERS, AND L. A. STAEHELIN, 1980. Supramolecular structure of stacked and unstacked regions of the photosynthetic membranes of Prochloron sp., a prokaryote. Proc. Nat. Acad, Sci. U.S.A., 77 : 352-356. HELLEBUST, J. A., 1965. Excretion of some organic compounds by marine phytoplankton. Linmol. Oceanog., 10: 192-206. JEFFREY, S. W., AND G. F. HUMPHREY, 1975. New spectrophotometric equations for deter- mining chlorophylls a, b, ci, and c2 in higher plants, algae and natural phytoplankton. Biochem. Physiol. Pflanzen., 167 : 191-194. KREMER, P. P., L. KIES, AND A. ROSTAMI-RAKET, 1979. Photosynthetic performance of the cyanelles in the endocyanomes Cyanophora, Glaucosphaera, Glocochaetc, and Glauco- cystis. Z. Pflansenp. 92 : 303-317. LEWIN, R. A., 1957. Excretion of glycolic acid by Chlamydomonas. Bull. Jpn. Soc. Phvcol., 5 : 74-75. LEWIN, R. A., AND N. W. WITHERS, 1975. Extraordinary pigment composition of a prokaryotic alga. Nature, 256 : 735-737. LEWIN, R. A., 1976. Prochlorophyta as a proposed new Division of algae. Nature, 261 : 697-698. LEWIN, R. A., 1977. Prochloron, type genus of the Prochlorophyta. Phycologia, 16: 217. MEEUSE, B. J. D., 1962. Storage products. Pp. 289-313 in R. A. Lewin, Ed., Physiology and Biochemistry of Algae. Academic Press, New York. 120: 191-193. MORIARTY, D. J. W., 1979. Muramic acid in the cell walls of Prochloron. Arch. Microbiol., MUSCATINE, L., 1967. Glycerol excretion by symbiotic algae from corals and Tridacna and its control by the host. Science, 156: 516^519. MUSCATINE, L., E. CERNICHIARI, AND R. R. POOL, 1972. Some factors affecting selective re- lease of soluble organic material by zooxanthellae from reef corals. Mar. BwL, 13: 298-308. THINH, L. V., AND D. J. GRIFFITHS, 1977. Studies of the relationship between the ascidian Diplosoma virens and its associated microscopic algae. I. Photosynthetic character- istics of the algae. Aust. J. Mar. Freshiv. Res., 28: 673-681. THINH, L. V., 1978. Photosynthetic lamellae of Prochloron (Prochlorophyta) associated with ascidian Diplosoma virens (Hartmeyer) in the vicinity of Townsville. Aust. J. Bot., 26 : 617-620. THORNE, S. W., E. H. NEWCOMB, AND C. B. OSMOND, 1977. Identification of chlorophyll b in extracts of prokaryotic algae by fluorescence spectroscopy. Proc. Nat. Acad. Sci. US. A., 14: 575-578. " TOENNIES, G., AND J. J. KOLB, 1951. Techniques and reagents for paper chromatography. Anal. Chem., 23 : 823-826. 648 C. R. FISHER, JR. AND R. K. TRENCH TRENCH, R. K., 1971a. The physiology and biochemistry of zooxanthellae symbiotic with marine coelenterates. I. The assimilation of photosynthetic products of zooxanthellae by two marine coelenterates. Proc. R. Soc. Land. B. Biol., 177 : 225-235. TRENCH, R. K. 1971b. The physiology and biochemistry of zooxanthellae symbiotic with marine coelenterates. II. Liberation of fixed a'C by zooxanthellae in vitro. Proc. R. Soc. Land. B. Biol., 177: 237-250. TRENCH, R. K., 1971c. The physiology and biochemistry of zooxanthellae symbiotic with marine coelenterates. III. The effect of homogenates of host tissues on the excre- tion of photosynthetic products in ritro by zooxanthellae from two marine coelenterates. Proc. R. Soc. Land. B. Biol.. 177: 251-264. TRENCH, R. K., 1974. Nutritional potentials in Zoantlnis sociatus. Hclgol. Wiss.Meeresunters,, 26: 174-216. TRENCH, R. K., 1979. The cell biology of plant-animal symbioses. Ann. Rci'. Plant Ph\siol.. 30: 485-531. TREVELYAN, W. F., D. P. PROCTOR, AND J. S. HARRISON, 1950. Detection of sugars on paper chromatograms. Nature. 166 : 444-445. WHATLEY, J. M., 1977. The fine structures of Prochloron. Nczc Phytol.. 79: 309-313. WHATLEY, J. M., P. JOHN, AND F. R. WHATLEY, 1979. From extracellular to intracellular : The establishment of mitochondria and chloroplasts. Proc. R. Soc. Lond. B. Biol. 204 : 165-187. WITHERS, N., W. VIDAVER, AND R. A. LEWIN, 1978. Pigment composition, photosynthesis, and fine structure of a nonblue-green prokaryotic algal symbiont (Prochloron sp. ) in a didernnid ascidian from Hawaiian waters. Phycologia, 17: 167-171. Reference : Biol. Bull.. 159 : 649-655. (December, 1980) THE EFFECTS OF CAFFEINE AND THEOI'H YLLINE ON THE PHOTOTACTIC RHYTHM OF CHLAMYDOMONAS REINHARDII JUDITH E. GOODENOUGHi AND VICTOR G. BRUCE 2 1 Department of Zoology, University of Massachusetts, Amherst, Mass. 01003, and - Department of Biology, Princeton University, Princeton, N. J. 08540 ABSTRACT Caffeine and theophylline were found to increase the period length of the photo- tactic rhythm of Chlaiuydoinonas rcinliardii in a dose-related manner. Treatment with caffeine resulted in average period increase of 11.2% (3 mM), 16.4% (5 mM), and 29% (10 mM). Theophylline caused smaller period increases than did caffeine: an average period lengthening of 4.2% (3 mM) and 6.1% (5 mM). INTRODUCTION Many of the important physiological processes in plants and animals are rhythmic. Therefore, it is of great importance to learn the nature of the clock- driving these rhythms. One method of attempting to decipher the mechanism of the clock has been to subject the organism to chemicals with known effects in the hope that the clockworks will be altered in a way that will reveal some aspect of their machinery. If a drug reaches the clock and interferes with some process im- portant to its timekeeping, sustained administration of the drug would be expected to result in a change in the period length of the observed rhythm. Although we are still far from knowing the molecular basis for circadian rhythmicity, we are beginning to identify some processes that are directly or indirectly involved in the timing mechanism. Translation of proteins on the SOS ribosomes appears to be important to the functioning of the clock. This gen- eralization is based on the observation that inhibitors of protein synthesis on the 70S ribosomes, such as streptomycin and chloramphenicol, usually have no effects on the period or phase of most biological rhythms (Karakashian and Hastings, 1962, 1963; Sweeney et al, 1967; Enright, 1971; Mergenhagen and Schweiger, 1975a) while inhibitors of protein synthesis on the SOS ribosomes, such as cycloheximide, puromycin, and anisomycin, often affect the timing process (Brink- mann, 1971, 1973; Feldman, 1967; Mergenhagen and Schweiger, 1975b; Kara- kashian and Schweiger, 1976a, b; Rothman and Strumwasser, 1977; Jacklet, 1977 ; Walz and Sweeney, 1979). Some primarily circumstantial evidence indicates that membranes are involved in generating circadian oscillations. Substances such as ethanol (Keller, 1960; Running and Baltes, 1962; Running and Moser, 1973; Brinkmann, 1974, 1976; Sweeney, 1974), K+ and valinomycin, an ionophore of K+ (Eskin, 1972; Hiinning and Moser, 1973) and Li+ (Engelmann 1972, 1973; Engelmann et al., 1974; Eskin. 1977) modify rhythms and are also known to alter membrane function. Received May 22, 1980; accepted September 25, 1980. Abbreviations : + mating type, mt+ ; - - mating type, mt" ; high salt-concentration medium, HSM. 649 650 J. E. GOODENOUGH AND V. G. BRUCE A role for cyclic AMP (cAMP) in the functioning of the biological clock has also been postulated (Felclman, 1975; Sweeney, 1976). A reason for this specula- tion that led to testing the hypothesis chemically is that the enzymes which control the levels of cAMP, adenyl cyclase and cAMP phosphodiesterase, are often membrane-bound. In addition, Cummings (1975) has suggested a more intimate involvement of cAMP in the timing process. He proposes a feedback system in which the levels of cAMP control and in turn are controlled by the activity of adenyl cyclase and phosphodiesterase. Chlamydomonas reinhardii has a well denned rhythm in phototaxis (Bruce, 1970, 1972). In an attempt to test the speculations that cAMP may be involved in the functioning of the biological clock, we subjected Chlamydomonas to in- hibitors of cyclic AMP phosphodiesterase and observed the effect on the organisms' rhythm in phototaxis. MATERIALS AND METHODS Culture techniques Both the + mating type (mt+) and the — mating type (mt~) of strain 137 F of Chlamydomonas reinhardii were used in these experiments. The strains were originally obtained from Lutz Wiese in 1960 and have been maintained at Prince- ton University since then. Experimental cultures were grown on a shaker table under continuous illumination from cool white flourescent lamps (1500-3000 lux) at 22° ± 1°C. The cultures were grown in 100 ml of 0.3 HSM medium (high salt-concentration medium) (Bruce, 1972) in 200 ml Erlemneyer flasks to a density of approximately 1-2 X 10° cells/ml. Phototactic assay Phototaxis was assayed on 1.5 ml samples of the culture placed in plastic tissue culture trays (Linbro Plastic) to which the drugs were added. The cells were placed in the apparatus in darkness that was interrupted at 2 hr intervals with 0.5 hr of a vertical light beam. While the light was on, organisms in a positive phototactic state were attracted to the beam and accumulated there, dimin- ishing the light reaching a photocell positioned above each beam. The photocell measured the intensity of transmitted light during the last 10 min the light was on. Thus the light beam not only elicited the phototactic response, but also provided a means of measuring its strength. The apparatus is described in more detail in Bruce, 1970. Chemicals Caffeine and theophylline, obtained from Sigma Chemical Co., were each added to final concentrations of 3, 5, or 10 mM to two replicate samples of each mating type. RESULTS The period lengthening of the phototactic rhythm of mt~ Chlamydomonas by caffeine treatment is shown in Figure 1. The effect is already apparent in the second peak of the rhythm, even at the lowest concentration tested (3 mM). When treated with 10 mM caffeine, the cultures became arrhythmic after 2-3 days. CHLAMYDOMONAS CLOCK 651 24 48 TIME 72 96 (HOURS) 120 FIGURE 1. The rhythmic phototactic response of the — mating type of strain 137 F of Chlamydomonas rcinhardii and its modification by treatment with caffeine. Each point repre- sents the mean of two samples. Figure 2 shows the period lengthening of the phototactic rhythm of mt~ caused by theophylline treatment. Although the period increases are not as great as those caused by equal concentrations of caffeine, a slight lengthening is evident by the second day of treatment. The data suggest that the phototactic response becomes arrhythmic after about 5 days of exposure to 5 mM theophylline. But the rhythm was only followed for 6 days, so it is uncertain whether the 6th peak- would have been present. A comparison of the period lengthening effects of caffeine and theophylline on both mt+ and mt~ Chlamydomonas is presented in Figure 3. The caffeine in- creased the period length more than theophylline did and the responses of mt+ and mt" were very similar, although the period increase of mt~ was slightly greater at almost all drug concentrations. The period lengthening with both drugs was dose-related. A 3 mM caffeine treatment caused a 9.6% increase in the rhythm's period length in mt+ and a 12.7% increase in that in mt". With 5 mM caffeine, the period lengths increased 12.9% in nit* and 16.9% in mt". The increases in the period lengths were substantially greater with a 10 mM caffeine treatment: 27 and 31% in rnt+ and mt", respectively. The theophylline treatments generated 652 J. E. GOODENOUGH AND V. G. BRUCE 24 48 72 96 TIME (HOURS) 120 FIGURE 2. The rhythmic phototactic response of the — mating type of strain 137 F of Chlamydomonas rcinhardii and its modification by treatment with theophylline. Each point represents the mean of two samples. smaller changes in period length. At 3 mM there was a 5.5% increase in mt+ and 2.8% increase in mt~. Five mM theophylline resulted in 5.7 and 6.4% in- creases in the period lengths of mt+ and mt~, respectively. DISCUSSION Theophylline and caffeine both increased the period length of the phototactic rhythm of Chlamydomonas almost immediately and in a dose-related manner. These results are consistent with previous tests of the effects of these substances, but extend the findings to a new circadian system. In previous studies it was shown that sustained administration of theophylline produced an increase in the period length of the sleep movement rhythm of the bean plant, Phaseolus (Keller, 1960), of the conidiation rhythm of the bread mold Neurospora (Feldman, 1975) and of the leaf movement rhythm of Trifolium (Bollig ct al., 1978). Pulses of theophylline produced phase changes in the sleep movement rhythm of Phaseolus (Mayer ct al., 1975) and in the body temperature rhythm of Rattus rattus (Ehret et al., 1975). Caffeine has been less widely tested. When continuously ad- ministered, it caused period lengthening in Neurospora's conidiation rhythm CHLAMYDOMONAS CLOCK 653 CAFFEINE THEOPHYLLINE 035 10 DRUG CONCENTRATION (mM) FIGURE 3. Period lengthening effects of caffeine and tlu-ophylline on both the + and - mating types of strain 137 F of Chlamydomonas rcinhardii. Each point represents the mean of two replicate samples. For 9 of the 12 points the standard deviations ranged from 0 to 0.35. For the + mating type 5 mM and the 10 mM caffeine-treated samples, the standard deviations were 0.71 and 1.4, respectively. For the -- mating type the standard deviation for treatment with 10 mM caffeine was 0.71. (Feldman, 1975), and when pulsed, it produced phase changes in the Phaseolus sleep movement rhythm (Mayer and Scherer, 1975). In our experiments caffeine was found to have a much greater effect on the biological timing process than did theophylline at equal concentrations. This is consistent with the results of Feldman (1975) on the conidiation rhythm of Neurospora. The period lengthening effect of these drugs suggests that they are acting at or close to the biological timing process. The best known effect of these drugs is the inhibition of cAMP phosphodiesterase, which results in an increase in cellular cAMP levels. Amrhein and Filner (1973) demonstrated that theophylline inhibits cyclic AMP phosphodiesterase in Chlamydomonas. However, other experiments suggest that the drugs may not be influencing the clock through their alteration of cAMP levels. Scott and Solomon (1973) found that in Neurospora, theophylline (5 mM) inhibited phosphodiesterase more effectively than caffeine; 83% of the phosphodiesterase activity remained in the caffeine-treated samples and only S2% of the activity remained in the theophylline- treated samples. If the action of the drugs on the timing mechanism of the clock is via their effects on cAMP phosphodiesterase, it is surprising that in Feldman's (1975) study on the conidiation rhythm of Xcitrospora and in this experiment, caffeine was found to increase the period length considerably more than theo- phylline. In addition, Feldman ct a!. (1979) have shown that mutants of Neuro- spora that have less than \% of the normal levels of cyclic AMP still have a normal clock, indicating that the major pools of cAMP in the cell are not essential for normal clock operation. Furthermore, if theophylline affected the clock by altering cAMP levels through its effect on phosphodiesterase, one might expect that pulses of cAMP and of theophylline would have similar effects on the clock and that these would 654 J- E. GOODENOUGH AND V. G. BRUCE be observed as similar phase-response curves. However, although both cAMP and theophylline increase the period length of the Trijolimn leaf movement rhythm, when the drugs were administered as pulses the phase-response curve for cAMP had only delays, while the one for theophylline had both advances and delays (Bollig ct al, 1978). Another way to alter cAMP metabolism would be to inhibit adenyl cyclase activity with a substance such as quinidine. It has been demonstrated that quinidine (26-58 /xg/ml) has no effect on the phototactic rhythm of Chlamy- donwnas (Carter, 1975), although higher concentrations (259 ftg/ml) inhibited phototaxis completely. If caffeine and theophylline were slowing biological tim- ing processes through their effects on cAMP levels, one would expect that quinidine would also affect the system. Caffeine and theophylline are known to have effects on other cellular processes, such as ion transport, macromolecular synthesis and the phosphorylation of pro- teins. Perhaps it is better to consider the period lengthening effects of these drugs as consistent with hypotheses that involve these cellular processes. LITERATURE CITED AMRHEIN, X., AND P. FILNER, 1973. Adenosine 3' : 5' cyclic monophosphate in Chlamy- domonas rcinhardtii: isolation and characterization. Proc. Nat. Acad. Sci. U.S.A., 70: 1099-1113. BOLLIG, I., K. MAYER, W-E. MAYER, AND W. ENGELMANN, 1978. Effects of cAMP, theo- phylline, imidazole, and 4-(3,4-dimethoxybenzyl)-2-irnidazolidone on the leaf move- ment rhythm of Trifoliuin re pens — a test of the cAMP hypothesis of circadian rhythms. Planta, 141 : 225-230. BRINKMANN, K., 1971. Metabolic control of temperature compensation in the circadian rhythm of Euglcna gracilis. Pp. 567-593 in M. Menaker, Ed., Biochronomctry. Na- tional Academy of Sciences, Washington, D. C. BRINKMANN, K., 1973. The role of actidion in the temperature jump response of the circadian rhythm in Euglcna gracilis. Pp. 523-529 in B. Chance, Ed., Biological and Bio- chemical Oscillators. Academic Press, New York. BRINKMANN, K., 1974. The effect of ethanol on the metabolism and circadian rhythms of Euglena gracilis. J. Intcrdiscip. Cycle Res., 5 : 186. BRINKMANN, K., 1976. The influence of alcohols on the circadian rhythm and metabolism of Euglcna gracilis. J . Intcrdiscip. Cycle Res., 7 : 149-170. BRUCE, V. G., 1970. The biological clock in Chlainvdomonas rcinhardi. J. Protozool., 17: 328-334. BRUCE, V. G., 1972. Mutants of the biological clock in Chlamydomonas rcinhardi. Genetics, 70 : 536-548. BUNNING, E., AND J. BALTEs, 1962. Wirkung von Athylalkohol auf die physiologische Uhr. Natunvisscnschaften, 49: 19-20. BUNNING, E., AND I. MOSER, 1973. Light-induced phase shifts of the circadian leaf move- ments of Phasco Ins: comparison with the effects of potassium and ethyl alcohol. Proc. Nat. Acad. Sci. U.S.A., 70 : 3387-3389. CARTER, A., 1975. Antibiotic effects on the phototactic response of Chlamydomonas rcinhardi. Sr. Honors thesis, Princeton University. CUMMINGS, F. W., 1975. A biochemical model of the circadian clock J. Theoret. Biol., 55: 355^170. EHRET, C. F., N. R. POTTKR, AND K. W. DOBRA, 1975. Chronotypic action of theophylline and of pentobarbital as circadian zeitgeibers in the rat. Science, 198: 1212-1215. ENGELMANN, W., 1972. Lithium slows down the Kalanchoe clock. Z. Naturforsch., 27b: 477. ENGELMANN, W., 1973. A slowing down of circadian rhythms by lithium ions. Z. Natur- forsch., 28c: 733-736. ENGELMANN, W., A. MAUER, M. MUHLBACH, AND M. JOHNSON, 1974. Action of lithium ions and heavy water in slowing circadian rhythms in petal movement in Kalanchoe. J. Interdiscip. Cycle Res., 5 : 199-205. CHLAMYDOMONAS CLOCK 655 ENRIGHT, J. T., 1971. The internal clock of drunken isopods, Z. I'gl. Physiol., 72: 1-16. ESKIN, A., 1972. Phase shifting a circadian rhythm in the eye of Aplysia by high potassium pulses. J. Comp. Physiol., 80 : 353-376. ESKIN, A., 1977. Neurophysiological mechanisms involved in the photoentrainment of the circadian rhythm from the Aplysia eye. J. Ncurobiol., 8: 273-299. FELDMAN, J., 1967. Lengthening the period of a biological clock in Euglcna by cycloheximide an inhibitor of protein synthesis. Proc. Nat. Acad. Sci. U.S.A., 57 : 1080-1087. FELDMAN, J., 1975. Circadian periodicity in N euros f>ora: alteration by inhibitors of cyclic AMP phosphodiesterase. Science. 190: 789-790. FELDMAN, J. F., G. GARDNER, AND R. DENISON, 1979. Genetic analysis of the circadian clock of Ncurospora. Pp. 57-66 in M. Suda, O. Hayaishi and H. Nagagawa, Eds., Bio- logical rhythms and their central mechanism, Elsevier/North Holland Biomedical Press. HASTINGS, J. W., 1960. Biochemical aspects of rhythms : phase shifting by chemicals. Cold Spring Harbor Symp. Quant. Biol., 25 : 131-143. JACKLET, J. W., 1977. Neuronal circadian rhythm : phase shifting by a protein synthesis in- hibitor. Science, 198 : 69-71. KARAKASHIAN, M. W., AND J. W. HASTINGS, 1962. The inhibition of a biological clock by actinomycin D. Proc. Nat. Acad. Sci. US. A., 48: 2130-2137. KARAKASHIAN, M. W., AND J. W. HASTINGS, 1963. The effects of inhibitors of macromolecu- lar biosynthesis upon the persistent rhythm in luminescence in Gonyaulax. J. Gen. Physiol., 47 : 1-12. KARAKASHIAN, M. W., AND H-G. SCHWEIGER, 1976a. Evidence for a cycloheximide-sensitive component in the biological clock of Acetabularia. Exp. Cell Res., 98: 303-312. KARAKASHIAN, M. W., AND H-G. SCHWEIGER, 1976b. Temperature dependence of the cyclo- heximide-sensitive phase of the circadian cycle in Acetabularia mediterranea. Proc. Nat. Acad. Sci. U.S.A., 73: 3216-3219. KELLER, S., 1960. Uber die Wirkung chemischer Faktoren auf die tagesperiodischen Blatt- bewegungen von Phascolus multifloris. Z. Bot., 48 : 32-57. MAYER, W., R. GRUNER, AND H. STRUBEL, 1975. Periodenverlangerund der circadianen Rhythmik von Phaseolus coccincus durch theophyllin. Planta, 125 : 141-148. MAYER, W., AND I. SCHERER, 1975. Phase shifting effect of caffeine in the circadian rhythm of Phascolus coccineus. Z. Naturjorsch, 30C : 855-856. MERGENHAGEN, D., AND H-G. SCHWEIGER, 1975a. Circadian rhythm of oxygen evolution in cell fragments of Acetabularia mediterranea. Exp. Cell Res., 92 : 127-130. MERGENHAGEN, D., AND H-G. SCHWEIGER, 1975b. The effect of different inhibitors of transcrip- tion and translation on the expression and control of circadian rhythms in individual cells of Acetabularia. Exp. Cell. Res., 94: 321-326. ROTHMAN, B. S., AND F. STRUMWASSER, 1977. Manipulations of a neuronal oscillator with inhibitors of macromolecular synthesis. Fed. Proc. 36 : 2050-2066. SCOTT, W. A., AND B. SOLOMON, 1973. Cyclic 3',5'-AMP phosphodiesterase of Neurospora crassa. Biochcm. Biophys. Res. Commun., 53 : 1024-1030. SWEENEY, B. M., 1974. The potassium content of Gonyaulax polycdra and changes in the circadian rhythm of stimulated bioluminescence by short exposure to ethanol and valinomycin. Plant Physiol., 53 : 337-342. SWEENEY, B. M., 1976. Evidence that membranes are components of circadian oscillators. Pp. 267-282 in J. W. Hastings and H-G. Schwieger, Eds. The Molecular Basis of Circadian Rhythms, Abakon Verlagsgestellschaft, Berlin. SWEENEY, B. M., C. F. TUFFLI, AND R. H. RUBIN, 1967. The circadian rhythm in photo- synthesis in Acetabularia in the presence of actinomycin D, puromycin and chlor- amphenicol. /. Gen. Physiol., 50 : 647-659. WALZ, B., AND B. M. SWEENEY, 1979. Kinetics of the Cycloheximide-induced phase changes in the biological clock in Gonyaulax. Proc. Nat. Acad. Sci. U.S.A., 76: 6443-6447. Reference : Biol. Bull., 159 : 656-668. (December, 1980) COMPARATIVE STUDY OF THE BLOOD PLASMA OF THE ASCIDIANS PYURA STOLONIFERA AND ASCJDIA CERATODES CLIFFORD J. HAWKINS.i PAULINE M. MEREFIELD,' DAVID L. PARRY,! WILTON R. BIGGS,2 AND JAMES H. SWINEHART-' 1 Department of Chemistry. University of Queensland, Brisbane, Australia, 4067; and -Department of Chemistry, University of California, Davis, California 95616 ABSTRACT The components of the blood plasma of the ascidians, Pyura stolonifera and Ascidia ceratodcs, have been separated by Sephadex G75 chromatography and gel electrophoresis. The absorption spectra of the plasma and the fractions for both species and the corresponding circular dichroism spectra for Pyura stolonifera have been recorded. The components for each species consisted of a number of proteins and two N-acetylaminosugar compounds, probably aminopolysaccharides. Although there were similarities in the properties of these components between the two species, a number of important differences were observed, especially with regard to the uptake of metals as studied by 59Fe, 54Mn, 48V, and 51Cr radiotracers. Electron spin resonance measurements detected manganese (II) in a N-acetylamino- sugar component of the plasma of Pyura stolonijcra. INTRODUCTION The blood of each suborder of the Ascidiacea (Aplousobranchia, Phlebo- branchia, and Stolidobranchia) seems to offer its own unusual characteristics. For example, members of the Phlebobranchia extract vanadium from sea water and concentrate it in morula blood cells called vanadocytes. Vanadium is also observed in some species within the Aplousobranchia, but where it is found within the animal is unknown (Biggs and Swinehart, 1976). Ascidians containing only iron as the dominant biological metal are found in the suborders Aplousobranchia and Stolidobranchia. For instance, dry cells of the stolidobranch Pyura stolonifera contain up to 5.1 mg/g iron, localized in blood cells termed ferrocytes (Endean, 1955a). Iron is also observed in the blood cells of vanadium-containing ascidians, but at an order of magnitude less than the vanadium. For example, Ascidia ceratodes blood cells contain 0.5 mg/g dry weight iron and 6-8 mg/g vanadium (Swinehart ct ai, 1974). Besides vanadium and iron, a number of other metals have been detected in ascidians. Manganese has been found in dona intcstinalis (Noddack and Noddack, 1940; Carlisle, 1968), Didcmnum albidiinn (Noddack and Noddack, 1940), Didemnum candidum (Carlisle, 1968), and Pyura stolonifera (D. A. Buckingham and J. M. Webb, Australian National University, personal communication). Chro- mium has been detected in the body tissue of Ciona intcstinalis (Bielig et al., 1961), and in whole zooids (Levine, 1962) and whole colonies (Swinehart ct al., 1974) of Eudistoma ritteri. It can be argued that these observations represent localized Received July 20, 1979; accepted August 20, 1980. Abbreviations: 2,4,6-tripyridyl-s-triazine (TPTZ), periodic acid Shiff (PAS), circular dichroism (CD), counts per minute (CPM). 656 BLOOD PLASMA OF ASCIDIANS 657 metal concentrations in sea water reflected in analyses of total animals. It is im- portant to learn whether ascidians take up metals such as manganese and chromium and incorporate them in their plasma and blood cells. In any process in which the final repository of the metal is blood cells, it would seem logical that the metal would make a transient or permanent appearance in plasma. To this end, the properties of the plasma of two phylogenetically diverse solitary ascidians, Pyura stolonifera and Ascidia ccratodcs, have been examined to ascertain the similarities and differences between them. P. stolonifera is an iron- containing species in the suborder Stolidobranchia, and A. ceratodcs is a vanadium- containing species in the suborder Phlebobranchia. The interaction of the plasma with metal ions has been examined by studying the uptake of 59Fe (as Fe(III) ), 54Mn (as Mn(II) ). r>1Cr (as Cr(III) ], and 48V (as V(Y) ) by animals and plasma. MATERIALS AND METHODS Specimens of the solitary ascidians A. ceratodcs and P. stolonifera were collected at Bodega Bay (California), U.S.A.; and Hastings Point (New South Wales), Stradbroke Island, and Noosa (Queensland), Australia, respectively. Both species were maintained in aquaria in aerated local sea water. Specimens of A. ccratodcs showed changes in properties of the plasma after short-term residence in an aquarium, so they were kept a minimum time under such conditions. The specimens of P. stolonifera studied could be kept in aquaria for long periods, even in excess of 6 months, without significant change in blood composition provided water temperature was maintained between 15° and 19°C and the aquaria had a balanced ecology. If the water temperature exceeded 19° C, for example in summer with no cooling, the blood cell count sank to a very low level in healthy animals and mortality increased markedly. Blood samples were removed from A. ceratodes by cardiac puncture, and from P. stolonifera by cutting the base and creating a small depression into which blood flowed. Blood cells were removed from the plasma by centrifugation. The plasma of both species becomes cloudy if left standing for a short time, so samples were used soon after blood was removed from the animal and centrifuged. The nature of the white cloudy material is unknown. The plasma samples were tested for N-acetylaminosugars (Reissig et a!., 1955), protein (biuret test), and after digestion with HC1O4-HNO3 mixture (Allan, 1959), for iron (colorimetrically using 2,4,6-tripyridyl-s-triazine [TPTZ] [Collins et al, 1959], and by atomic absorption). Gel electrophoretic experiments were carried out using analytical (6 or 8% polyacrylamide ) and, in some cases, stacking gels prepared by standard techniques. Bromophenol blue was used as the tracking dye, and the current was 2-3 ma/gel. The gels were fixed with \5C/C trichloroacetic acid, stained with \% Coomassie Brilliant Blue, and destained with 5% glacial acetic acid for protein analysis; for carbohydrate analysis the gels were treated with aqueous 40c/c ethanol and 5% glacial acetic acid, then oxidized with periodic acid, and finally stained with Schiff base stain (PAS) (Pearse, 1960). In the in situ radioactivity experiments about 4-15 liters of sea water was inoculated with an appropriate isotope, 59Fe, "Mn, 48Y, or 51Cr, and allowed to equilibrate for several hours. Animals were added and air was bubbled through the system. Animals were removed at various times and their blood removed and analyzed as described above. Plasma was either separated using liquid chro- 658 HAWKINS ET AL. u GD cc o en CD 05 - 600 500 400 300 WAVELENGTH, nm - -25 200 FIGURE 1. Ultraviolet visible and circular dichroism (CD) spectra of the fresh plasma from P. stolonijera (solid line) and A. ceratodcs (dashed line). Cell pathlength 1 cm. Dilution shown where applicable. matography (Sephadex) or gel electrophoresis and counted, or counted directly. In plasma inoculation experiments, plasma was maintained at 4°C for P. stolonijera and at 12°C for A. ceratodes. For both types of experiments with P. stolonijera, the plasma was dialyzed against distilled water for 1 hr before counting or chromatography, and the individual chromatography fractions were dialyzed for a further 2 hr before counting. In some cases solid samples {e.g., test, hydroids, algae) were counted, and where possible the counts were reduced to a per minute per gram basis. The counting instruments used were a Beckman Gamma 310 Radiation Counter (A. ceratodes} and a Nuclear Chicago counting well system (P. stolonijera}. Spectroscopic studies employed Cary Models 17 and 14 spectrophotometers for the ultraviolet (UV) visible spectra, a Jobin Yvon Mark III Dichrograph for the circular dichroism (CD) spectra, a Varian V 4502 15X band spectrometer operating at 9.104 GHz for the ESR spectra, a Varian Techtron Model 1200 spectrometer for the atomic absorption studies, and a Lab Test Multielement Analyser for the inductively coupled plasma studies. For ESR measurements, fractions of the plasma were separated by chromatography, dialyzed against distilled water for 3 hr at 10°C. freeze dried, and studied as powders at 77°K. RESULTS Properties of the plasma of the iron-containing ascidian P. stolonijera and the vanadium-containing ascidian A. ceratodes are presented in two parts: spectral and separative properties of the plasma and its constituents, and association of metals with the plasma or its constituents. Association was measured by direct analysis, electron spin resonance (ESR), or radiotracer incorporation. Spectral and separative properties Plasma from P. stolonijera is pink; that of A. ceratodes is yellow. Figure 1 contains the UV visible spectra of both species. Common to both spectra were bands in the 260-275 and 300-330 nm regions. In addition, P. stolonijera had an BLOOD PLASMA OF ASCIDIANS 659 absorbance at about 500 nm which imparted the pink color to its plasma. The CD spectrum of the plasma of P. stolonifera is recorded in Figure 1 for comparison with chromatographed constituents of the plasma. The plasmas of both species were chromatographed through Sephadex G75 with 0.05M NaCl as eluent. The elution patterns detected at 275 nm for both species and also at 325 mn for P. stolonifera are given in Figure 2. For both species two major bands, one each, in fractions 1-2 and 8-13 were observed. The second band from P. stolonifera could contain several components. In addition, a pink compound in fractions 3-5 was observed between the two major bands. The first band eluted (fractions 1-2) was in the void volume of the column and should have contained compounds with molecular weights greater than the exclusion of the G75 column (30,000). It gave a positive test for protein, and negative tests for N- acetylaminosugar and iron for both species. It had a maximum at 275 nm in the absorption spectrum for P. stolonifera and at 282 for A. ccratodes. Fractions 3-5 differed for the two species. Both give positive tests for protein and negative tests for N-acetylaminosugar and iron. However, the absorption spectrum for P. stolonifera (Fig. 3) contained maxima at 497 and 275 nm, whereas the 497 nm band was missing for A. ccratodes. This pink band from /'. stolonifera was eluted with water in fractions 2-4. The CD spectrum of the pink band had a positive Cotton effect at 516 nm, a negative at 387 nm, and a positive at 293 nm. Below 260 nm the CD was negative. Fractions 8-13 for A. ceratodes had absorption maxima at 260 and 314 nm with the ratio of the absorbance at the two wavelengths increasing as the fraction number increases, suggesting that more than one compound was present. For P. stolonifera, on some occasions only a 327-nm maximum was observed in fractions 8-13, and on these occasions the spectrum did not change significantly across the band. On other occasions a 275-nm maximum was obtained, and the absorbance ratio for 275 and 325 nm increased with increasing fraction number, again showing the presence of more than one compound. When fractions from P. stolonifera were dialyzed, a compound having the UV visible and CD spectra shown in Figure 3 was isolated in the external solution. The same dialysis properties were observed for comparable fractions from A. ceratodes, and resulted in the separation of a 1234 5 6 78 9 10 II 12 13 14 15 16 10 ml FRACTIONS FIGURE 2. Elution pattern of plasma of P. stolonifera from Sephadex G75 at 275 (solid line) and 325 (dashed line) nm, and A. ceratodes at 275 nm (dotted line). Fractions 3-5 from P. stolonifera were pink. 660 HAWKINS ET AL. 0.2 0 I Ul o m tr. o °5 04 0 3 02 0 I 0 J_ I 0 0.5 O z < m or O in CD -05 o x 50 — LLJ O Z < DQ K O CO CD -50 400 350 300 250 WAVELENGTH, nm 200 FIGURE 3. Absorption and CD spectra of (A) the "pink fraction" (fractions 3-5) from the plasma of P. stolonifcra, and (B) the "327 nm" compound isolated from the plasma of P. stolonifcra (solid line) and the absorption spectrum of the "314 nm" compound from A. ceratodcs (fractions 8-13) (dashed line). compound having a 314 nm maximum (Fig. 3). The separation in this case was not as clean. An alternate method for separating the P. stolonifera compounds with 327 and 275 nm absorption was to freeze dry the fractions when water was used as eluent. The aqueous extract of the residue gave the same spectrum as in Figure 3A. A white insoluble residue became purple with time. Fractions 8-13 and isolated compounds from both species tested positive for N-acetylaminosugar, and negative for protein. With 0.05 M NaCl eluent, the P. stolonifera fractions tested positively for iron. But this appeared to be due to a contaminant in the sodium chloride, because when distilled water was used as eluent, a negative test for iron was obtained. A. ceratodes tested negative for iron. Both species gave different elution patterns if subjected to trauma during trans- portation or in aquaria. When species of A. ceratodcs were transported long dis- tances and/or remained in an aquarium for as little as 2 days, the pattern had increased absorption for fractions 3-7, resulting from increasing absorption below 300 nm, showing no maximum down to 200 nm. A. ceratodes seemed more sensi- tive to these influences in winter. P. stolonifcra appeared to be more robust, but if subjected to trauma due to temperature increase, for example in summer, all fractions except 1 and 2 showed markedly decreased absorption. BLOOD PLASMA OF ASCIDIANS 661 The gel electrophoretic patterns (Fig. 4) for plasma showed a saccharide hand (stained by PAS) at RfO for hoth species and a numher of protein hands (stained by Coomassie Blue). For P. stolonijera, six protein hands were observed with the major component at Rf 0.45. Two patterns (A and \\ in Fig. 4) were ohserved for the protein hands in thr A. ccratodcs plasma from animals collected from two sites in Bodega Bay, California. The pattern was not related to the collection site. Gel electrophoresis of fraction 2 from /'. stolonijera showed a major and three minor protein bands, whereas for A. ccratodcs two protein bands were observed. The gel electrophoresis pattern of fraction 5 ( /'. stolonijcra ) usually contained two major bands that were stained by Coomassie Blue, but for some animals four separate protein bands were observed. For A. ccratodcs at least five proteins were present in this fraction. Fractions 8-13 for both species yielded no separation of the different compounds on the gel electrophoretogram, with only a band (positive for saccharide ) observed at RfO. Metal association studies One reason for interest in the elution and gel electrophoretic patterns is to learn which plasma constituents, if any, participate in the uptake of metals and their subsequent transport to the blood cells. Therefore, uptake studies involving Fe, Mn, Cr, and V using the radiotracers 5!'Fe, °*Mn, :ilCr, and 48V were carried out. In addition, cells and animal surface (test) were analyzed. Animals of both species were incubated with 5!'Fe for up to 150 hr. Plasma activity levels of both species increased with time, whereas the sea water activity decreased rapidly to approximately a fifth of its original value at 150 hr. The Fe level increased in the blood cells as well as in the plasma. On a counts per minute (CPM) per ml whole-blood basis, the 59Fe content of the blood cells exceeded that A. ceratodes P. sto/on/fera 0 Rf 0.5 i n P A asr na B /////// _ Plasma 2B 5B 9B M FIGURE 4. Gel electrophoretograms : 8 and 6% polyacrylamide gels, respectively, were used for plasma and Sephadex separated fractions of A. ccratodcs and P. stolonijcra. Bands stained by Coomassie Blue are shown in solid lines and those stained by PAS in broken lines. 662 HAWKINS ET AL. 2000 1000 E Q. 0 >-" H 800 > O < 600 400 200 0 ' / \.\ A V» B 600 -c 400 200 0 800 600 400 200 2 4 6 8 10 12 14 16 " ; FRACTION 4. D ^1044 1 1 I I I I II 1 1 I I I I I I I I I I I I I I I I I I 1 I I I I \\ \ I I \ \ \ I \ A / \ I \ / \ 4 6 8 10 12 14 16 FIGURE 5. Elution patterns for 59Fe activity : In plasma of P. stolonifera for two sets of whole animal experiments (A) 48 hr (closed circles), 118 hr (open circles), and (B) 24 hr (closed circles), 72 hr (closed squares) and 120 hr (open triangles) ; (C) in plasma incubation of P. stolonifera: 0.5 hr (open circles), 24 hr (open squares), 60 hr (closed triangles; (D) in whole animal incubation of A. ccratodcs: 48 hr (closed triangles), 144 hr (open triangles). of the plasma at any time for either species. Upon lysing the blood cells, 59Fe activity appeared in both the lysate and residue. Iron 59 activity versus plasma fraction eluted from Sephadex G75 for both species is summarized in Figure 5. In the P. stolonifera whole-animal incubation experiments (Figs. 5A, B), 59Fe activity correlated well with the three elution bands discussed earlier: fractions 1-2, 3-5, and 8-13. Even when the 275^497 nm compound normally associated with fractions 3-5 was shifted to fractions 5-7 in the 24 hr sample (Fig. 5B), the 59Fe activity also shifted. When P. stolonifera plasma was incubated, a different pattern (C in Fig. 5) was obtained. The activity was not maximized where the 275^497 nm compound was eluted (normally fractions 3—5). Indeed at fraction 5, the activity was at a minimum. For P. stolonifera, the pattern of 59Fe uptake was independent of iron concen- tration. This was not the case for A. ceratodes. At low activity levels (0.1 yu,Ci in 4 1 sea water) activity was restricted to fractions 8—13, whereas at higher Fe levels some activity was found in all peaks eluted from Sephadex G75, with similar patterns obtained from whole animal and plasma incubations (D in Fig. 5). BLOOD PLASMA OF ASCIDIANS 663 TABLE I 69 Fe radiotracer results for organs and sections of test of Pyura stolonifera. Test sections had the dimensions 0.5 X 1.5 X 1.0 cm and were consecutive slices taken from a cross-section of the base with TI adjacent to the exterior. Organ or Radioactivity (CPM/g) test section 24 hr incubation 72 hr incubation 120 hr incubation Gut 14,730 32,950 2,300 Sac 2,250 7,530 2,550 Gonad 160 180 82 Liver 2,547 19,700 11,580 T, 780 690 330 T, 560 2,600 762 T3 2,970 630 630 When sea water inoculated with 59Fe was chromatographed through Sephadex G75 using 0.05 M NaCl as eluent, the free metal was eluted in fractions 8-13, the same fractions as the "314 nm" and "327 nm" compounds from A. ceratodes and P. stolonifera, respectively. To test whether the 59Fe detected in fractions 8-13 in the above experiments was free or bound, the plasma fractions were dialyzed for 3 hr before counting. The activity remained, showing the 59Fe was bound to the organic compound (s). Although the various fractions took up 59Fe in the whole animal and plasma experiments with both species, no natural iron was detected by the TPTZ method or by atomic absorption for these fractions. However, an inductively coupled plasma study of the isolated "327 nm" compound and the insoluble residue from fraction 8-15 of P. stolonifera found that the "327 nm" compound had 0.05% iron whereas the insoluble fraction had 0.74% iron. We concluded that the natural concentrations of iron in these fractions were too low to be detected by the TPTZ or atomic absorption methods, given the small amounts of materials used. When 5iiFe was administered to A. ceratodes and P. stolonifera, significant activity appeared on the surface (test) of the animals. This means that much of the 59Fe was unavailable for uptake by the internal organs and raises the question of the reason for surface absorption. In a typical experiment using A. ceratodes. comparable 1 cm- areas of the outer surface of the test had 28,300 CPM (48 hr) and 25,000 CPM (144 hr) while the same area 1 mm below the test had 50 CPM (48 hr) and 800 CPM (144 hr). Since blood vessels run through the test, some of the activity observed below the surface of the test could be due to incorporation of 5l)Fe into the plasma. Table I shows the incorporation of 59Fe into various P. stolonifera organs and layers of test. Again, the surface of the test accumulated 59Fe and part of the activity below the surface must have arisen from accumula- tion of Fe in the blood. For both A. ceratodes and P. stolonifera the chemical com- position of the test guaranteed that it would be a good complexing agent for metals, and Fe accumulation on the surface would be expected. Enclean examined the test of P. stolonifera and found it to be composed of polysaccharides and aminopolysaccharides (Endean, 1955b). An interesting sidelight on 59Fe incorporation is its uptake by algae, hydroids, and possibly bacteria attached to the surface of P. stolonifera and A. ceratodes. Table II shows CPM/g dry weight for algae and hydroids found on the surface of 664 HAWKINS ET AL. TABLE II 69 Fe uptake by algae and hydroid on surface of Pyura stolonifera Organism Radioactivity (CPM/g) 24 hr incubation 72 hr incubation 120 hr incubation Gracilaria edulis (Rhodophyte) Halimeda discoidea (Chlorophyte) Caularpa nexicana (Chlorophyte) Hydroid sp. 31,960 123,000 36,550 39,450 81,330 26,860 11,710 9,450 25,120 127,000 P. stolonifera. Comparable observations were made for algae associated with A. ceratodes. Manganese 54 uptake experiments with A. ccratodes showed con- siderable uptake by a hydroid attached to the surface. Clearly some of the 59Fe and 54Mn on the test of the two species resulted from uptake by organisms on the surface. Whole-animal 54Mn studies for both species showed uptake in the plasma and blood cells. In both animal and plasma incubation experiments the activity of r'4Mn in the plasma was maximized at the same position in the elution pattern from Sephadex G75 as the N-acetylaminosugar (fractions 8-13). When sea water was inoculated with 54Mn as Mn(OH^)6J+ and chromatographed, it too gave a peak at this position. However, 54Mn in the plasma fractions was not removed by short term dialysis. During whole-animal incubation experiments the sea water activity decreased relatively rapidly with time, reaching a tenth of its original activity in less than 100 hr. For P. stolonifera, the "327 nm" compound and its associated insoluble material from fractions 8—13 were analyzed for manganese by inductively-coupled plasma optical emission. Values of 0.04 and 0.12%, respectively, were obtained. ESR spectra of the freeze-dried chromatographic fractions (distilled water as eluent) of P. stolonifera gave no signal for fractions 1-7 but gave an intense typical six-line Mn(II) resonance (see, for example, Chan et al., 1967; Meirovitch and Lanir, 1978) with g = 2.001 ± 0.001 and A = 92 G for the "330" band. The quantities TABLE III Absorbance ratio for "275" and "330" bands, and the positions of the maxima (in nm, in parenthesis) for fractions 9-12 from chromatography of blood plasma of Pyura stolonifera. Incubation experiment Fraction 9 Fraction 10 Fraction 11 Fraction 12 Natural 0.5 (275, 327) 0.8 (275, 327) 1.3 (273, 325) 1.7 (271, 325) 24 hr 69Fe 0.7 (272, 325) 0.7 (272, 323) 0.9 (272, 323) 1.3 (268, 322) 54 hr 6»Fe 0.6 (273, 327) 0.5 (273, 327) 0.6 (273, 326) 0.8 (273, 325) 120 hr 69Fe 0.5 (275, 329) 0.5 (272, 327) 0.7 (273, 327) 1.3 (273, 325) 24 hr" Mn 0.8 (278, 324) 0.6 (285, 324) 0.6 (282, 323) 0.9 (280, 322) 98 hr 64Mn 0.4 (285, 327) 0.4 (275, 327) 0.6 (274, 326) 0.8 (275, 326) (283) (282) (282) 64Mn in plasma 0.3 (273, 323) 0.6 (273, 323) 0.5 (273, 323) BLOOD PLASMA OF ASCIDIAXS 665 g and A are the gyromagnetic ratio and hyperfine splitting constant, respectively. When this band was separated into the "327 nm" compounds and the "275 nm" insoluble fraction, a relatively weak Mn(II) signal was obtained for the former, but no signal was observed for the latter. As mentioned above, the isolated insoluble fraction turned purple on standing, and it was this sample that had been studied. The loss of most of the Mn(II) signal when the "330" band was separated into its components could be due to the Mn(II) being oxidized during the change in color of the insoluble fraction. In 59Fe and r>4Mn experiments with P. stolonifcra. the metals affected the spectral characteristics of the various fractions throughout the N-acetylaminosugar band. The data for the fractions immediately on elution from the column are presented in Table III. The results show that with r'"Fe the largest effect on the positions of the bands was after 24-hr incubation, where both bands shifted to about 3 nm lower wavelength. The 24-hr :'4Mn whole-animal incubation also shifted the "327 nm" band to lower wavelength by about 3 nm, but the "275" band moved up to 10 nm higher wavelength, and with 98-hr incubation a new shoulder appeared at about 282 nm. In the 54Mn plasma-incubation experiment, little variation in the ratio of the absorption bands was observed for fractions 9, 10, and 11, and no vari- ation in the absorption maxima that were in similar positions to the bands in the 24-hr 59Fe experiment. For A. ceratodes the absorbance maximum of fraction 9, normally at 314 nm, was transformed to 325-330 nm by incubation of the animal with 59Fe, and to a wavelength lower than 314 nm by incubation with r>4Mn. This effect was not reproduced by direct interaction of these metals with the plasma. When specimens of P. stolonifcra were exposed for 24 and 72 hr to sea water containing an aquahydroxo complex of 51Cr(III), the chromium was taken into the plasma, but this activity was lost upon dialysis. The failure of 51Cr to bind to any of the plasma fractions could be due to the inertness of Cr(III) to substitute reactions, in contrast to the other metal ions. The uptake of 48V by whole animals and plasma alone was studied. Both species' plasma had 48V activity. However, all activity was dialyzed from P. stolonifera plasma, showing that there is no strong binding between vanadium and the plasma components of this species. Dialysis experiments witli A. ceratodes were not conducted. The plasma was chromatographed and had activity in fractions 1-2 and 8-13 corresponding to the two main chromatography bands detected by UV-visible spectroscopy. Sea water innoculated with 48V (V) and chromatographed gave activity in fractions 8-13. Therefore, it is possible that the 48V in the plasma is not bound to the aminosugar component. In a gel electro- phoresis experiment, the majority of the activity traveled with the front, showing that most of the vanadium in the plasma is not bound to the components, but probably is present as a low molecular weight, negatively charged compound such as H2VO4~, present in water at this pH. DISCUSSION Plasma spectra (Fig. 1) are similar for P. stolonijcra and A. ceratodes, iron- containing and vanadium-containing ascidians respectively. Both spectra have shoulders or bands in the 260-275 and 300-330 nm regions. The main difference between the two species is the presence of pink compound (s) having maxima at 275 and 497 nm in the P. stolonifcra plasma. The spectrum of plasma of A. nigra, a species in the same genus as A. ceratodes, has shoulders in the 260-275 and 300- 330 nm regions, and at about 380 nm (Kustin et al., 1976). Thus there appears HAWKINS ET AL. to be some correlation between the plasma spectra for A. ceratodes and A. nigra. We have observed that lysing of blood cells can cause a transitory absorbance in the 380 nm region in the plasma. Elution of the plasma of P. stolonifera and A. ceratodes from Sephadex G75 results in similar patterns for absorbance vs. fraction collected (Fig. 2). Tests of the fractions for proteins and N-acytylaminosugars show 1-2 and 3-5 to be proteins, and 8-13 to be N-acytylaminosugar for both species. Gel electrophoretic patterns of the plasma of A. ceratodes and P. stolonifera (Fig. 4) confirm the above mentioned similarities between the plasma of the two species. The proteinaceous materials are contained in earlier elution fractions 1-2 and 3-5 and show dissimilar electrophoretic patterns between species (Fig. 4). The last compounds eluted from Sephadex G75 for both species (fractions 8-13, Fig. 2) test as N-acetylaminosugars and show no protein. The fact that there is a single band in the gel electrophoretic patterns of fractions 9 is not inconsistent with the possible presence of two closely related compounds or two compounds that interconvert, as is suggested by the variations of the absorption ratio (275/325 nm) with increasing fraction (8-13). In any case the two compounds referred to above are saccharides and the compound(s) having an absorption in the 300-330 nm region test as N-acetylaminosugars. The gel electrophoretic pattern for A. ceratodes (A, B, Fig. 4) suggests equilibria between some proteinaceous compounds. This may be the basis of the changes in the plasma composition noted in this species. The difference between the two patterns is the disappearance of the two bands between 0.50-0.55 and the appearance of single bands at 0.35 and 0.91, suggesting as one explanation that two proteins of intermediate molecular weight can yield proteins of higher and lower molecular weight or z'icc versa. The appearance of a 0.50-0.55 band in fraction 5B from a plasma sample containing only a 0.35 band indicates that the conversion can take place outside the animal. The CD spectra of the pink compound contained in Sephadex fractions 3-5 from P. stolonifera dominates the CD spectrum of the plasma of this species. This pink compound is not present in A. ceratodes, but compounds with similar spectra occur in the plasma of other species in the genus Pyura (unpublished results, J. H. Swinehart). The role of these compounds in the plasma is not yet understood. The N-acetylaminosugars found in fractions 8-13 of the elution patterns from Sephadex G75 (Fig. 2), and represented by the spectra in Figure 3B, appear to be important in association with metal ions. The spectral properties of these com- pounds also reflect the metal association. Figure 3 shows that the compounds isolated from these fractions for P. stolonifera and A. ceratodes have maxima at 327 and 314 nm, respectively. If A. ceratodes is incubated in 59Fe (or Fe) the 314 nm maxima shifts to 325-330 nm, while incubation with a4Mn (Mn) shifts the maxima below 314 nm. Incubation of P. stolonifera with 54Mn results in a shift of the 327 nm band to shorter wavelengths, and 59Fe incubations a shift to slightly longer wavelengths. Thus, on the basis of the metal-induced spectral changes, metal analyses, and electron-spin-resonance studies carried out on these compounds it seems reasonable to focus on them as possible metal carriers in the plasma. Accordingly, animals were incubated in various metal radiotracers, and plasma eluted from Sephadex G75. Whole-animal 59Fe uptake into the plasma of P. stolonifera (Fig. 5) shows associations with all classes of compounds eluted from Sephadex G75. This distri- BLOOD PLASMA OF ASCIDIANS 667 bution occurs at both high and low MIFe levels. Very little can be done to correlate the absolute activity with the sequence of Fe transfer in the plasma because each elution experiment represents an animal, and as the experiment proceeds the animal tends to degenerate so that uptake requirements and plasma chemistry may vary with time for a particular animal. However, comparison of whole animal data with that for uptake of r';'Fe by the plasma itself shows that only in whole animal incuba- tion is M'Fe incorporated in the "pink" compound in fractions 3-5. If A. ccratodes is incubated at low 59Fe levels (0.1 ju.Ci in 4 1 sea water) incorporation occurs only into the band contained in fractions 8-13. At higher levels of r'!'Fe, A. ccratodes incorporates the metal into all classes of bands eluted from Sephadex G75. Thus it appears that for A. ceratodes, in contrast to /'. stolonifcra, there is a preferred uptake of Fe in fractions 8-13 and, if sufficient Fe is present, it is indis- criminately removed by several components of the plasma. This difference in the uptake of ™Fe by the two species could be pertinent to the difference in their ability to concentrate iron in the blood cells. The r>4Mn uptake experiments show that there is not a rapid equilibration of manganese between sea water and plasma. If such an equilibration were present, a rapid increase in the activity of the plasma followed by a close correspondence to the sea-water activity curve should occur. The maximization and subsequent decrease in activity of the plasma can result from slow re-equilibration with the sea water, uptake by other organs, and uptake by the blood cells. For both species the 54Mn activity of the blood cells increases over the entire course of the experi- ments, and in the case of P. stolonifcra, an iron-containing fraction from the blood cells, so named "ferriascid," has been isolated and shows increased r>4Mn activity, as do the cell residues. Manganese-54 uptake by both species results in the incorporation of the isotope in fractions 8-13 of the elution pattern from Sephadex G75. This incorporation was substantiated by the observation of the Mn(II) ESR signal in samples of plasma from P. stolonifera eluted from Sephadex G75. For both 59Fe and S4Mn, dialysis of these plasma fractions does not result in a loss of the metal. It is concluded that 59Fe and 54Mn are bound to the N-acetylaminosugar(s) in these fractions. The result of 48V uptake for A. ceratodes can be interpreted in terms of 48V binding to the proteins in fraction 1-2, but it is unclear whether the activity in fractions 8-13 is completely due to free vanadium species or partly to 48V bound to the N-acetylaminosugar compound. A major difference between the two species is that the proteins in fractions 1-2 bind vanadium even at low concentrations in A. ccratodes, a species that concen- trates vanadium in its blood cells, but not in P. stolonifcra. The reverse was found to be true for 59Fe. In P. stolonifcra, a species that concentrates iron in its blood cells, the proteins in fractions 1-2 were found to bind iron at both high and low concentrations, but in A. ceratodes at low concentrations no iron was detected in these fractions. In summary, studies with 59Fe, r'4Mn, 51Cr and 48V show that the chemistry of the sea water-plasma interface does not discriminate between these elements, but that selectivity occurs during the binding of the plasma components to the metals. The N-acetylaminosugar compounds bind 54Mn and M'Fe in both species, and possibly 48V in A. ceratodes . The proteins in fractions 1 and 2 bind 59Fe in both species at high concentrations but only in P. stolonifcra at low concentration, and HAWKINS ET AL. SV only in A. ceratodcs at either high or low concentrations. The proteins in fractions 3-5 bind 59Fe in P. stolonijcra. ACKNOWLEDGMENTS The authors wish to thank Dr. R. Bramley of the Research School of Chemistry of the Australian National University for the ESR measurements, Dr. Patricia Mather of the Queensland Museum for useful discussions, and the University of Queensland for financial support for Professor J. H. Swinehart during his stay at the University on sabbatical leave from the University of California, Davis. Research support is acknowledged from the Australian Research Grants Committee (C. J. Hawkins) and the Research Corporation (J. H. Swinehart). LITERATURE CITED ALLAN, J. E., 1959. The determination of iron and manganese by atomic absorption. Sfectrochemica Acta, 15 : 800-806. BIELIG, H. J., K. PFLEGER, W. RUMMEL, AND E. SEIFEN, 1961. Aufnahme und Verteilung von Vanadin bei der Tunicate Ciona intcstiiialis L. Hoppe-Seyler's Z. Plivsiol. Chan., 326: 249-258. BIGGS, W. R., AND J. H. SWINEHAUT, 1976. Vanadium in selected biological systems. Pp. 141- 195 in H. Siegel, Ed., Metals in Biological Systems. M. Dekker, Inc., New York, New York. CARLISLE, D. B., 1968. Vanadium and other metals in ascidians. Proc. R. Soc. Land. Scr. B., 171: 31-42. CHAN, S. I., B. M. FUNG, AND H. LUTJE, 1967. Electron paramagnetic resonance of Mn(II) complexes in acetonitrile. /. Chcm. Phys., 47 : 2121-2130. COLLINS, P. F., H. DIEHL, AND G. F. SMITH, 1959. 2,4,6-Tripyridyl-s-triazine as a reagent for iron. Anal. Chcm., 31 : 1862-1867. ENDEAN, R., 1955a. Studies of the blood and tests of some Australian ascidians. I. The blood of Pyura stolonifera (Heller). Aust. J. Mar. Freshzv. Res., 6: 35-39. ENDEAN, R., 1955b. Studies of the blood and tests of some Australian ascidians. III. The formation of the test of P\ura stolonijcra (Heller). Aust. J. Mar. Frcshiv. Res., 6: 157-164. KUSTIN, K., D. S. LEVINE, G. C. McLEOD, AND W. A. CURBY, 1976. The blood of Ascidia nigra: Blood cell frequency distribution, morphology, and the distribution and valence of vanadium in living blood cells. Biol. Bull., 150: 426-441. LEVINE, E. P., 1962. Studies on the structure, reproduction, development and accumulation of metals in the colonial ascidian Endistoma rittcri Van Name, 1945. /. Morphol., Ill : 105-137. MEIROVITCH, E., AND A. LANIR, 1978. Electron spin resonance of manganese (II) complexes with proteins. Chcm. Phys. Letters, 53: 530-535. XODDACK, I., AND W. NODDACK, 1940. Die Haiifigkeiten der schwermetalle in Meerestieren. Ark. Zoo/., 32 : 1-35. PEARSE, A. G. E., 1960. Histochcmistry: Theoretical and Applied, 2nd Ed., J. & A. Churchill Ltd., Lond. 998 pp. REISSIG, J. L., J. L. STROMINGER, AND L. F. LELOIR, 1955. A modified colorimetric method for the estimation of N-acetylamino sugars. /. Biol. Chcm. 217 : 959-966. SWINEHART, J. H., W. R. BIGGS, J. D. HALKO, AND N. C. SCHROEDER, 1974. The vanadium and selected metal contents of some ascidians. Biol. Bull., 146: 302-312. Reference : Biol. Bull., 159 : 669-680. (December, 1980) CHEMISTRY OF THE BLOOD OF THE ASCIDIAN PODOCLAVELLA MOLUCCENSIS CLIFFORD J. HAWKINS, DAVID L. PARRY, AND CRAIG PIERCE Department of Chemistry, University of Queensland, Brisbane, Australia 4067 ABSTRACT The blood plasma of the aplousobranch Podoclavella moluccensis was found to have features similar to the plasma of ascidians from different suborders, including the presence of a yellow saccharide. The plasma concentrates a number of transi- tion metals, especially vanadium, manganese, zinc, and chromium. The dominant blood cells were yellow morula cells and blue granular cells. A yellow ethanolic extract of the cells had a visible spectrum similar to a /3-carotene with an additional intense absorption at about 270 nm. Magnesium and a trace of iron (III) were the only metals detected in that fraction. A subsequent purple acidified-ethanol extract, also possibly a carotenoid, and the cell residue contained a number of metals. Rela- tive to the plasma, these two fractions concentrated calcium, and also iron, vanadium, and chromium, to a much greater extent than magnesium. This is the first re- port of chromium in the blood cells of ascidians or of significant concentrations of zinc in the plasma. The vanadium was not present as V(III) in the morula cells, as it is in some phlebobranchs, but appears to be in the granular cells, and the cell residue gave an ESR signal characteristic of VO2+. The UV-visible and circular dichroism spectra of the various blood fractions are reported and compared to similar fractions from the aplousobranch Sigillina cyanea and the stolidobranch Polycarpa pedunculata. INTRODUCTION A companion paper from this laboratory compares the chemistry of the blood plasma of two ascidians : Pyura stolonifera from the suborder Stolidobranchia, and Ascidia ceratodes from the suborder Phlebobranchia (Hawkins et al., 1980) . Plasma components were separated by gel chromatography and gel electrophoresis, and uptake of metals by these components was studied using radiotracers with whole animals and plasma. Besides proteins, each species has in its plasma an N-acetyl- aminosugar compound that has different spectroscopic characteristics for the two species and that takes up iron and manganese in both, and possibly vanadium in Ascidia ceratodes. Aminosugar compounds have also been isolated from the blood cells of these two species (Biggs, 1977; Dorsett, Hawkins, Merefield, Parry, and Stern, in preparation). Except for metal concentrations, little has been published concerning the chem- istry of blood from the third suborder, Aplousobranchia. Relatively large vanadium and iron concentrations have been found in certain species (Biggs and Swinehart, 1976). In contrast to the phlebobranchs, in which the vanadium is in the (III) Received July 20, 1979; accepted August 20, 1980. Abbreviations: Circular dichroism (CD), periodic acid Shiff (PAS), inductively-coupled plasma optical emission (ICP), gyromagnetic ratio (g). 669 HAWKINS, PARRY, AND PIERCE oxidation state localized in morula "green cells" called vanadocytes, the aplouso- branchs have vanadium in the (IV) oxidation state and its location is unknown. In this study, concentrations of metals in various fractions of the blood were de- termined for the aplousobranch Podoclavclla moluccensis (Sluiter, 1904), which belongs to the family Clavelinidae. The UV-visible absorption and circular di- chroism (CD) spectra of the fractions were measured and compared with spectra of similar fractions from other ascidians. Electron spin resonance studies were car- ried out to determine the oxidation states of the metals. MATERIALS AND METHODS Ascidians were collected in the Capricorn Reefs in the vicinity of Heron Island, Queensland, Australia, in late spring, 1978, and were identified by Dr. Patricia Mather (Kott) of the Queensland Museum. Blood collection and frac- tionation were carried out at the Heron Island Research Station. Two species of the suborder Aplousobranchia were studied : Podoclavclla mohtccensis, collected from a rubble substrate at about 15 m, and SigiJlina cyanca, (Colella cyanea, Herd- man, 1899; Eudistoma cyanea, Kott, 1957) collected from a sandy bottom at 30 m. Two distinct forms of the stolidobranch Polycarpa pednnculata (Heller, 1878) were collected. One, found both at 30 m on a sandy bottom and at about 8 m attached to the reef rock on the drop-off, had a bright orange test and blood. The other, collected from reef rock at 8 m, had a whitish test with yellow blood. Following collection, the ascidians were kept in aquaria in running sea water. The blue blood of Podoclavclla moluccensis was extracted within 24 hr of col- lection from individual stolons of the zooids, using a syringe. The blood was cen- trifuged and the cells mechanically lysed in 90% ethanol, giving a bright yellow solu- tion. Extraction of the cells with ethanol was repeated until a clear solution was obtained. The residue was then exhaustively extracted with 0.3% HC1 in ethanol, giving a purple solution. The blue cell residue and both extracts were stored at 4°C, and the pale yellow plasma was stored frozen. Sigillina cyanca zooids have a common stolon. Blood was removed from this species and from the solitary ascidian Polycarpa pcduncnlata by cutting the stolons near the base, creating a well when the animals were inverted, and by sucking the blood from this well with a syringe. A sample of plasma from a colony of Podoclavclla moluccensis was dialyzed against distilled water for 1 hr and chromatographed through Sephadex G75 with distilled water as eluent. The fractions were monitored by measuring absorbance at 275 and 325 nm. Individual fractions were tested for proteins by the biuret test, and for saccharides by the PAS method (Pearse, 1968). Metal ion concentrations in the various blood fractions were determined by inductively-coupled plasma optical emission (ICP) using a Lab Test Multielement Analyser. The ethanolic solution samples were evaporated to dryness under vac- uum at room temperature and the aqueous solutions including the plasma were freeze-dried prior to digestion with HNO3-HC1O4. Spectroscopic studies were carried out with a Cary Model 17 spectrophotometer for the UV-visible absorption spectra, a Jobin Yvon Mark III Dichrograph for CD spectra, and a Varian V-4502-11X band spectrometer equipped with a multipurpose cavity operating at 9.104 GHz for electron-spin-resonance measurements. Cells were stained both in whole blood and after fixing with 5% formalin in sea water. The stains were as follows: oil red, neutral red, methyl red, osmium tetroxide, Turnbull's iron(II), and Perl's iron(III). BLOOD OF PODOCLAVELLA MOLUCCENSIS 671 B FIGURE 1. Blood cells, showing relative diameters. Approximate relative populations are given below in parentheses. (A) Mature morula cell: Podoclavella iiwlucccnsis, yellow (50%) ; Sigillina cyanea, yellow (60%) ; Poly car pa pcdunculata (white test), yellow (80- 90%) ; Polycarpa pcdunculata (orange test), orange (80-90%). (B) Non-acidic granular cell: Podoclavella moluccensis, blue with Brownian granules (30-40%); purple (2%). (C) Purple-black globular cell: Sigillina cyanea with Brownian granules (30-35%). (D) Macro- phage with ingested cells: each species (2%). (E) Lymphocyte: each species (10%). RESULTS Blood cells Morula cells (Fig. 1) were the predominant blood cells in each of the species studied. In Podoclavella moluccensis, Sigillina cyanea, and the specimens of Polycarpa pcdunculata with white test, these cells were yellow and made up 50, 60, and 80-90%, respectively, of the cell populations. The orange form of Polycarpa peduncidata had bright orange morula cells (80-90% of total population) which readily agglutinated, and the blood in the test of this form went black on relatively brief exposure to air. Globules within the yellow morula cells from specimens of Podoclavella moluccensis and Sigillina cyanea gave positive acid tests with oil red, neutral red, and methyl red, although the whole blood was neutral. No staining tests were conducted with blood from specimens of Polycarpa pcdunculata. The amount of acid in the blood cells from the species Podoclavella molluccensis must be small because lysis in water (freezing was necessary for lysis) and in alcohol gave a neutral solution. A non-acidic blue granular cell was the second most predomi- nant cell type for Podoclavella moluccensis (30-40%). Non-acidic purple granular cells (about 2%) were also present. For Sigillina cyanea the second major cell type (30-35%) contained large purple to black globules. Lymphocytes (< 10%) and some large macrophages were also present in each species. The cells are diagramatically represented in Figure 1. Cells from specimens of Podoclavella moluccensis were stained for vanadium and iron. Osmium tetroxide, known to stain cells containing V(III) black (Henze, 1913), and cell membranes containing fats grey (Pearse, 1968), stained the morula cells grey, suggesting that the vanadium found in the blood is not in the V(IIT) form. Turnbull's solution turned the morula cells a bright turquoise blue, indicating the presence of Fe(II). A negative test was obtained with Perl's stain for Fe(III) although the blue granular cells took on a more purplish tint. 672 HAWKINS, PARRY, AND PIERCE 8 0 700 600 500 400 300 WAVELENGTH , nm 200 500 400 300 WAVELENGTH, nm FIGURE 2. Absorption and CD spectra of blood fractions. (A) from blood cells of Podoclavella nwluccensis: ethanol extract (dotted line); hexane solution of solid from ethanol extract (dashed line) ; and acidified ethanol extract (solid line). (B) from blood cells of Sigillina cyanca: ethanol extract (dotted line); hexane (dashed line) and 85% aqueous methanol (solid line) solutions of the blue solid isolated from the ethanol extract. (C) from blood cells of Polycarpa pedunculata (orange form) : ethanol (dotted line) and acidified ethanol (solid line) extracts. (D) plasma of Podoclavella nwluccensis, (dotted line), and orange (dashed line) and white (solid line) forms of Polycarpa pcdunculata. Cell pathlength 1 cm ; dilution shown where applicable. The animals were collected late in spring, when specimens of Podoclavella moluccensis were very active in producing larvae. It is possible that this could have affected the relative cell populations and perhaps the blood chemistry of this species. Fractions isolated from blood cells Blood cells of Podoclavella moluccensis did not lyse in water unless the water was frozen, nor did they lyse in dilute acid. Ethanol, on the other hand, caused rapid lysis of cells, giving a bright yellow extract with the absorption and CD spectra shown in Figure 2. After evaporation to dryness, the yellow fraction was partitioned between hexane and 85% aqueous methanol. It dissolved completely in the hexane layer giving the spectrum in Figure 2, and did not change color on acidification. It gave an ESR spectrum at 77 K with resonances at gyromagnetic ratio (g} values of 2.00 and 3.78 (Fig. 3). After exhaustive extraction with ethanol, the residue was extracted with acidified ethanol, giving a deep purple extract with the absorption spectrum shown in Figure 2 but no observable CD nor ESR spectra. The blue cell residue had a strong ESR resonance of g 2.01 with a fine structure characteristic of VO2+ (Fig. 3). BLOOD OF PODOCLAVELLA MOLUCCENSIS 673 Sigillina cyanca blood cells lysed in ethanol gave a bright blue lysate with the spec- trum shown in Figure 2. After exhaustive extraction with ethanol, no material was extracted into acidified ethanol. The ethanol extract gave a blue solid on evaporation to dryness. Upon partitioning between hexane and 85% methanol, 75% of this solid went into the hexane and the remainder into the methanol. The spectra of these two solutions are given in Figure 2. The blue extract had no detectable CD. Cells of the orange form of Polycarpa pedunculata were also lysed in ethanol and exhaustively extracted with that solvent, giving a bright yellow, solution. Sub- sequent extraction with acidified ethanol also gave a yellow solution. Figure 2 shows the absorption and CD spectra of both fractions. Not enough blood was isolated from specimens of Polycarpa pednnculata with white tests for these types of studies. Blood plasma Absorption and CD spectra of the plasma of Podoclavella tnolucccnsis and the two forms of Polycarpa pednnculata are shown in Figure 2. Unfortunately the Sigillina cyanca plasma was destroyed before its spectrum was measured. The elution pattern of Podoclavella ntoluccensis plasma chromatographed through Sephadex G75 is given in Figure 4. A colorless protein band that gave a positive biuret test but a negative PAS test for saccharide was eluted in the first 10 ml frac- tion following the exclusion volume of the column. Fractions 2 and 3, part of the same band as fraction 1, were yellow and tested positive for protein and negative for saccharide. Fractions 9—12 came under the second major elution band. They were yellow and tested negative for protein but positive for saccharide. The absorption spectra for the individual fractions under this chromatographic band varied across the band (Fig. 5). The CD spectrum for fraction 10 is included in Figure 5. 4,9-3-78 100 G H FIGURE 3. ESR spectra of the ethanol extract of the blood cells (solid line) and the cell residue (dashed line) of Podoclavella moluccensis. HAWKINS, PARRY, AND PIERCE GO oe o CO 00 < 0-8 0-6 0-4 0-2 u — 275nm- 325nm i — i — i i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 10ml FRACTIONS FIGURE 4. Elution pattern of plasma of Podoclavella nwliiccciisis through Sephadex G75 with water as eluent. Metal concentrations in blood fractions The various fractions of the blood of Podoclavella moluccensis were analyzed by TCP for Ca, Mg, V, Cr, Mn, Fe, Cu, and Zn. In Table I, the results are compared o z < DO a. o CO 00 --0-1 400 350 300 WAVELENGTH, nm 250 FIGURE 5. Absorption spectra of fractions 10 (dashed line) and 11 (solid line) and the CD spectrum of fraction 10, from the chromatographic separation of the plasma of PodoclavcUa moluccensis. Cell pathlength 1 cm ; dilution shown where applicable. BLOOD OF PODOCLAVELLA MOLUCCENSIS 675 with the mean of the limits of the range of concentrations found in sea water (Parker, 1972). DISCUSSION The visible absorption spectrum of the yellow fraction extracted from Podoclavclla moluccensis blood cells possessed a broad band centered at 450 nm with shoulders at 470 and 428 nm, very similar to the corresponding band in the /2-carotene spectrum (Vetter et al., 1971). However, the yellow fraction had more intense absorption in the UV than /^-carotene, and had a band at 270 nm, as is found in spectra of carotenoproteins such as ovorubin (Cheesman, 1958), However, carotenoproteins are usually not soluble in ethanol, as the protein moiety is denatured by the sol- vent. The result of partitioning between hexane and 85% methanol was charac- teristic of an epiphasic carotene, and not of a carotenoprotein. Carotenoids have been isolated previously from a number of ascidians in the three suborders (Good- win, 1952). The acidified-ethanol extract's spectrum had an absorption maximum at 584 nm with shoulders on the low wavelength side of the band. Further shoulders are observed at about 360 and 280 nm with a large peak at 250 nm. On evaporation, this compound changed to a reddish-orange solid, which dissolved in acidified ethanol to give a spectrum with a maximum at 475 nm, shoulders on the high wavelength side of the band and at 355 and 280 nm, and a peak at 248 nm. The purple compound is possibly also a carotenoid, as dodecapreno-/3-carotenes, for ex- ample, are known to absorb at 589 ± 14 nm (Vetter et al., 1971). The spectrum also had a shape similar to those of a number of carotenoproteins (Lee, 1977) but its solubility in acidified ethanol was not consistent with its being a protein. The UV-visible spectrum of the blue ethanol extract of Sigillina cyanca blood cells had an intense peak at 593 nm with a shoulder at 555 nm, a broad band at 375 nm, and peaks at 327 and 277 nm (Fig. 2). Absorptions in the visible spectrum were similar to those of the purple acidified-ethanol extract from specimens of Podoclavella moluccensis. Both extracts had no detectable CD spectrum. The absorption spectrum of the extract from specimens of Sigillina cyanea again sug- gested a carotenoid, and the partitioning between hexane and 85% methanol was consistent with a xanthophyll (Liaaen-Jensen, 1971). The absorption and CD spectra of the ethanol and acidified-ethanol extracts of the cells from the orange form of the stolidobranch Polycarpa pedunculata (Fig. 2) were similar to one another, but completely different from the spectra of the frac- tions from the two aplousobranchs. The ethanol extract had a peak at 380 nm and a shoulder at 250 nm in its absorption spectrum, and a negative Cotton effect at 380 nm with a negative shoulder at 280 nm in the CD spectrum. The acidified-ethanol TABLE I Metal concentrations in blood fractions of Podoclavella moluccensis expressed in ng/g dry -weight of fraction. Sea water concentrations (ng/l) are the mean of the concentration-range limits (Parker, 1972). Fraction Ca Mg V Cr Mn Fe Cu Zn Plasma — whole 1,115 1,720 4.17 0.35 0.45 1.31 0.66 4.37 Plasma— Fr. 9-12 19,980 46,310 96 940 45 4,530 130 370 EtOH extract 0 6,050 0 0 0 0 0 0 EtOH/HCI extract 2 X 105 460 480 58 3 624 18 66 Cell residue 7.000 950 681 22 7 964 16 49 Sea water 0.4 X 106 1.3 X 106 2.4 1.3 0.7 30 13 14 HAWKINS, PARRY, AND PIERCE extract's spectrum had a peak at 396 nm and shoulders at 340, 305, and 250 nm. Its CD spectrum had a negative Cotton effect at 390 nm. Visible absorption bands of both fractions lacked the fine structure characteristic of the carotenoids' main visible absorption band as found for extracts from the aplousobranchs. The visible CD band was negative for both extracts from Polycarpa pcdunculata, whereas the ethanol extract from Podoclavclla rnolucccnsis had a small negative Cotton effect at 439 nm and a large positive at 371 nm, and the blue-purple fractions from the two aplousobranchs had no detectable CD. As has been found previously for the plasma of Ascidia nigra (Kustin et al., 1976), Ascidia ceratodes, and Pyitra stolonifera (Hawkins et al., 1980), the absorption spectra of the plasma of Podoclavella molucccnsis and the two forms of Polycarpa pedunculata (Fig. 2) had bands in the 300-330 and 260-275 nm regions. The spectrum for the form of Polycarpa pcdunculata with white tests tailed further into the visible than the spectrum for the orange form. This absorption gave the plasma a pink color like that observed for another stolidobranch, P\ura stolonifera, with a definite absorption band at 500 nm in the spectrum of its plasma (Hawkins et al., 1980). The CD spectra of the plasma of Podoclavella rnolucccnsis and the two forms of Polycarpa pcdunculata had negative Cotton effects with maxima be- tween 360 and 380 nm ( Fig. 2 ) , as has also been found for Pyitra stolonifera. How- ever, for the first three of these species, the plasma had a negative Cotton effect with a maximum in the region 270-290 nm, whereas the plasma for Pyura stoloni- fera has a positive in this region. The elution pattern for the plasma of Podoclavella molucccnsis, chromatographed through Sephadex G75 (Fig. 2), was very similar to that for the Pyura stolonifera plasma with water as eluent (Hawkins et al., 1980). A major protein band was eluted following the exclusion volume of the column. A colored compound was eluted on the side of this band. This was yellow for both Podoclavella moluccensis and Ascidia ceratodes, with an absorption peak at 275 nm and a tail in the visible. The corresponding fractions for Pyura stolonifera were pink, with absorption bands at 497 and 275 nm. The elution of a saccharide in the second major chromato- graphic band was also consistent with the findings for P\ura stolonifera and Ascidia ceratodes. For these latter two species the saccharide was identified as an N-acetylamino- sugar, but insufficient material for this test was available from Podoclavella moluc- censis. The absorption spectra for the individual fractions for each of the three species varied across the chromatographic band, showing that at least two com- pounds were present (Fig. 5). The ratio of the absorption peaks' intensities (270: 327 nm) increased across the chromatographic band. For Pyura stolonifera and Ascidia ceratodes the two compounds were separated. The yellow compound had an absorption maximum at 314 nm for Ascidia ceratodes and at 327 nm for Pyura stolonifera. From the spectra in Figure 5 it would appear that the Podoclavella molucccnsis yellow compound also absorbed at 327 nm. Its CD spectrum was also very similar to the corresponding compound from Pyura stolonifera, with negative Cotton effects at 350 and 300 nm, and a positive at 268 nm (Fig. 5 ) . The spectrum went negative below 250 nm. Many published papers have been concerned with metal ions in ascidians, but most have dealt with the concentrations of metals in whole animals (see, for ex- ample, Biggs and Swinehart, 1976). It is difficult to evaluate from these studies whether ascidians have used the metals, or whether they are present through con- tamination via the ascidians' filter feeding. If the metals accumulate in the blood BLOOD OF PODOCLAVELLA MOLUCCENSIS 677 plasma or, more importantly, in the blood cells, there is a much greater possibility of the metals fulfilling a biological function. For this reason, the various fractions of the blood of Podoclai'dla mohtcccnsis were analysed by TCP for a range of metals. The concentrations of the metals in the plasma were in the order : Mg > Ca > Zn — V > Fe > Cu > Mn > Cr. Using Mg as the reference, the transition metal ion concentrations can be seen to have increased markedly in the plasma relative to the normal levels of dissolved metals in sea water, especially vanadium, and to a lesser extent, manganese, zinc, and chromium. This is true whether the mean value or the concentration range in sea water is used as the basis for comparison. Fur- ther, the relative concentrations in the plasma did not mirror those normally found in oceanic sea water. The apparent very marked accumulation of the above four metals in the plasma, relative to the others studied, could be a direct result of higher than normal concentrations of these metals, both dissolved and in participate form, in the water pumped by these animals in their coral reef habitats. The present study is the first report of high zinc levels in the blood plasma of ascidians. The skeletal carbonate of the corals in the reefs of the Capricorn Group, where the col- onies of Podoclavella molnccensis were collected, has been found to have relatively high levels of zinc (St. John, 1974). For example, in samples of Acroporidac 1400 ng/g Zn was found compared to 740 ng/g Fe and 210 ng/g Cu. St. John claimed that the concentrations of trace elements in the corals reflected concentra- tions in the water of the environment, and hence these results would indicate larger concentrations of zinc in the vicinity of the reefs compared to oceanic sea water. The high levels of zinc in specimens of Podoclavella uiolucccnsis could also be a result of the ingestion of particles of coral containing high levels of zinc rather than the presence of larger than normal concentrations of dissolved zinc in the sea water. The values taken for the concentrations of manganese and vanadium in sea water might also be misleading because they refer to dissolved metals. Concentrations of manganese in participate matter (Parker, 1972), and of vanadium in marine sediments and plankton (Biggs and Swinehart, 1976), which are known to be high, were not included. The apparent accumulation of Cr in the plasma might have a similar explanation, but in the case of Cr and also V relatively high levels of these metals were found in blood-cell fractions, and hence the levels in the plasma might not be simply a consequence of the concentrations in the animals' food intake. The high Ca/Mg concentration ratio in the plasma (0.65) compared to sea water (0.31) might also be a consequence of the relative abundance of these ele- ments in the food supply of species collected from coral reefs, which, of course, are composed largely of calcium carbonate. In contrast, specimens of Phallusia mam- mill at a (Henze, 1912; Robertson, 1954) and Pyura stolonifcra (Enclean, 1955), collected from a rock substrate, have been found to have Ca/Mg ratios in the plasma almost identical to that in sea water. The variations found in the Ca/Mg ratios for the various blood fractions are also interesting. In the saccharide fraction from the plasma, the Ca/Mg value of 0.42 is lower than that in the whole plasma (0.65). This is somewhat surprising, as some sugars have been found to form much more stable complexes with calcium than with magnesium (Angyal, 1973). In the blood-cell fractions, no calcium was detected in the yellow ethanol extract, whereas 6050 fjig/g Mg was found. In direct contrast, in the purple acidified-ethanol extract about 2 X 105 /Ag/g Ca and 460 /xg/g Mg were found — a Ca/Mg ratio of 422. The cell residue had 7000 ^.g/g Ca and 950 /ig/g Mg, with Ca/Mg equal to 7.4. Since the chloride and sulfate salts of calcium were soluble enough in ethanol to be ex- HAWKINS, PARRY, AND PIERCE tracted into ethanol, the fact that no detectable levels of Ca were found in the ethanol extract indicates that Ca was either precipitated out as an insoluble salt other than the chloride or sulfate, or was coordinated in a complex insoluble in ethanol. At least one calcium compound was soluble, however, in acidified ethanol. Perhaps Ca was bound to the purple pigment. The magnesium in the yellow ethanol extract could be free Mg-+ or bound to the yellow pigment. The saccharide fraction from the plasma had high concentrations of Fe and Cr, and to a lesser extent Zn. However, a comparison of these concentrations to those in the plasma shows that only Fe and Cr were concentrated by the saccharide com- pound to any marked extent. No detectable levels of the transition metals were found for the yellow ethanol ex- tract of the blood cells by ICP, but relatively high concentrations of iron and vanadium were found in the purple acidified-ethanol extract (624 and 480 p-g/g, respectively) and also in the cell residue (964 and 681 /*g/g, respectively). Fe/V concentration ratios in the two fractions were similar. However, a trace of iron was present in the yellow ethanol extract, as an ESR signal typical of non-heme iron (Peisach and Blumberg, 1969) was detected. The spectrum was identical to that obtained for an iron-containing N-acetylaminosugar compound isolated from the morula blood cells of Pyura stolonijcra (unpublished results, C. J. Hawkins and D. L. Parry). No iron or vanadium ESR signal was observed at 77 K for the purple extract shown to contain both elements. This could result from the rapid relaxation of the electron spins for the two elements. An ESR signal characteristic of VO2+, de- tected for the blue cell residue, is similar to the spectrum obtained for V(IV) in the zooids of the aplousobranch Eudistoma diaphanes (Swinehart et al., 1974). Superimposed on this spectrum is a relatively weak signal at about g 2.06 that could result from the iron. No attempt was made to exclude oxygen during the isola- tion of these fractions, and therefore it is possible that the vanadium had a lower oxidation state in whole cells than in the isolated fractions. However, a negative stain for V(III) was obtained for the whole cells. In the purple fraction, besides Fe and V, significant concentrations of Zn and Cr were found with traces of Cu and Mn. Comparing the metal concentrations in this fraction to those in the plasma obtained the following ratios: V, 115; Cr, 166; Mn, 6; Fe, 476; Cu, 27; and Zn, 15. The same comparison for the cell residue, yields the following ratios: V, 163; Cr, 63; Mn, 16; Fe, 736; Cu, 24; and Zn, 11. From these results it can be concluded that Fe, V, and Cr accumulate in the blood cells. Chromium has been found in whole zooids (Levine, 1962) and in whole colonies (Swinehart et al., 1974) of the aplousobranch Eudistoma rittcri and in the body tissues of dona intcstinalis (Bielig ct al., 1961), but the present paper is the first report of Cr in ascidian blood cells. However, in terms of the actual concentrations of transition metals in the blood cells, Fe and V are the predomi- nant metals, with an Fe/V ratio of about 1.3. Iron and vanadium are also the predominant transition metals in the blood cells of Ciona intcstinalis, with an Fe/V ratio of 0.53 (Bielig et al., 1961 ). The vanadium in that species, traditionally classi- fied as a phlebobranch, also occurs in the (IV) oxidation state (Swinehart ct al., 1974). In the blood cells of species of the family Ascidiidae belonging to the sub- order Phlebobranchia, V(III) is the predominant transition metal ion (Biggs and Swinehart, 1976), although significant concentrations of Fe have been reported for the blood cells of Phallusia mamniiUata (Bielig ct al., 1961) and in the blood of Ascidia ccratodcs (Swinehart ct al., 1974). In the blood cells of the stolidobranch BLOOD OF PODOCLAVELLA MOLUCCENSIS 679 Pyura stolonifera, no V has been detected, but relatively large concentrations of Fe have been found (Endean, 1955). In the Ascidiidae the V(III) is localized in the morula "green" cells called va- nadocytes (Biggs and Swinehart, 1976). In Ciona intestinalis, the vanadium is present in cells "that look somewhat different" from the vanadocytes (Rummel et a!., 1966). In Podoclavella molucccnsis, vanadium is found in the purple fraction and in the blue cell residue, and not in the yellow fraction. Therefore, it is tempting to conclude that the vanadium is in the purple and blue granular cells and not in the yellow morula cells. ACKNOWLEDGMENTS The authors wish to thank especially Dr. Patricia Mather of the Queensland Museum for identifying the ascidians collected in the Capricorn Group and for useful discussions. The authors are also grateful to Mr. R. W. Garrett of this department for the ESR measurements, to Dr. A. J. Bruce, director of the Heron Island Research Station, for use of the facilities at the Station, and to staff and student members of the College of Idaho for assistance in the collection of the ascidi- ans at Heron Island during their study program in Australia. Research support is acknowledged from the Australian Research Grants Committee. LITERATURE CITED ANGYAL, S. J., 1973. Complex formation between sugars and metal ions. Pure Appl. Chew., 35: 131-146. BIELIG, H. J., E. JOST, K. PFLEGER, W. RUMMEL, AND E. SEIFEN, 1961. Aufnahme und Verteilung von Vanadin bei der Tunicate Pliallusia mamillata Cuvier. Hoppc-Sevler's Z. Physiol. Chem., 325: 122-131. BIELIG, H. J., K. PFLEGER, W. RUMMEL, AND E. SEIFEN, 1961. Aufnahme und Verteilung von Vanadin, bei der Tunicate Ciona intestinalis L. Hoppe-Sevler's Z. Physiol. Chem., 326 : 249-258. BIGGS, W. R., 1977. Transition metals in ascidians : The blood of Ascidia ccratodes. Ph.D. thesis, University of California, Davis, Calif., 151 pp. Diss. Abstr. 38: 4799(B). BIGGS, W. R., AND J. H. SWINEHART, 1976. Vanadium in selected biological systems. Pp. 141-195 in H. Seigel, Ed., Metals in Biological Systems. M. Dekker, Inc., New York. CHEESMAN, D. F., 1958. Ovorubin, a chromoprotein from the eggs of the gastropod mollusc Pomacea canaliculata. Proc. R. Soc. Lond. Ser. B, 149: 571-587. ENDEAN, R., 1955. Studies of the blood and tests of some Australian ascidians. I. The blood Pyura stolonifera (Heller). Aust. J. Mar. Freshzv. Res.. 6: 35-39. GOODWIN, T. W., 1952. The Comparative Biochemistry of the Carotenoids. Chapman & Hall Ltd., London, 365 pp. HAWKINS, C. J., P. M. MEREFIELD, D. L. PARRY, W. R. BIGGS, AND J. H. SWINEHART, 1980. Comparative study of the blood plasma of the ascidians, Pyura stolonifera and Ascidia ccratodes. Biol. Bull., 159: 656-668. HELLER, C., 1878. Beitrage zur nahern Kenntniss der Tunicaten. Sitzbcr Akad. Wiss. Wein, 77: 83-110. HENZE, M., 1912. Blood of ascidians. Hoppe-Seyler's Z. Physiol. Chem.. 79: 215-228. HENZE, M., 1913. Blood of ascidia. Hoppe-Seyler's Z. Physiol. Chem.. 86: 340-344. HERDMAN, W. A., 1899. Descriptive Catalogue of the Tunicata in the Australian Museum. T. Dobb, Liverpool, 139 pp. KOTT, P., 1957. The ascidians of Australia. II. Aplousobranchiata Lahille : Clavelinidae Forbes and Hanly and Polyclinidae Verrill. Aust. J. Mar. Freshw. Res., 8: 64-110. KUSTIN, K., D. S. LEVINE, G. C. McLEOD, AND W. A. CURBY, 1976. The blood of Ascidia nigra : Blood cell frequency distribution, morphology, and the distribution and valence of vanadium in living blood cells. Biol. Bull., 150: 426-441. LEE, W. L., Ed., 1977. Carotene proteins in Animal Coloration. Dowden, Hutchinson & Ross Inc., New York, 395 pp. HAWKINS, PARRY, AND PIERCE LEVIXE, E. P., 1962. Studies on the structure, reproduction, development and accumulation of metals in the colonial ascidian Eudistonia ritteri Van Name, 1945. /. Morpho/., Ill: 105-137. LIAAEN-JENSEN, S., 1971. III. Isolation, Reactions in Carotenoids, Pp. 63-177 in O. Isler, Ed., Carotcnoids, Birkhauser Verlag, Basel. PARKER, C. R., 1972. Water Analysis b\ Atomic Absorption. Varian Techtron, Australia, 78pp. PEARSE, A. G. E., 1968. Histochemistry : Theoretical and Applied. 3rd Ed. J. & A. Churchill Ltd., London, 759 pp. PEISACH, J., AND W. E. BLUMBERG, 1969. Metalloproteins as studied by electron paramagnetic resonance. Pp. 71-96 in Teh Fu Yen. Ed., Electron Spin Resonance of Metal Complexes, Plenum Press, New York. ROBERTSON, J. D., 1954. The chemical composition of the blood of some aquatic chordates, in- cluding members of the Tunicata, Cyclostomata and Osteichthyes. /. E.vp. Biol.. 31 : 424-442. RUMMEL, W., H. J. BlELIG, W. FORTH, K. PpLEGER, W. RlJDIGER, AND E. SEIFEN, 1966. Ab- sorption and accumulation of vanadium by tunicates. Protides Biol. Fluids Proc. Colloq., 14 : 205-210. SLUITER, C. P., 1904. Die Tunicaten der Sibogo-Expedition. I. Die Socialen urid holosomen Ascidien. Sibogo-Exped., 56a : 1-126. ST. JOHN, B. E., 1974. Heavy metals in the skeletal carbonate of Scleractinian corals. Pp. 461-469 in A. M. Cameron et a!., Eds., Proceedings of the 2nd International Coral Reef Symposium 2. GBRC, Brisbane. SWINEHART, J. H., W. R. BIGGS, D. J. HALKO, AND N. C. SCHROEDER, 1974. The vanadium and selected metal contents of some ascidians. Biol. Bull.. 146: 302-312. VETTER, W., G. ENGLERT, N. RIGASSI, AND U. SCHWIETER, 1971. IV. Spectroscopic Methods. Pp. 189-265 in O. Isler, Ed., Carotenoids, Birkhauser Verlag, Basel. Reference: Biol. Bid!., 159: 681-691. (December, 1980) EFFECTS OF MAGNETIC FIELDS ON REGENERATION IN FIDDLER CRABS PAUL H. LEE i AND JUDITH S. WEIS - Department of Zoology and Physiology, Rutgers, the State University, Ncivark, N. J. 07102 ABSTRACT After autotomy, fiddler crabs (Uca fiitgilator and U. pugnax) were placed in magnetic fields (in the 100 Oersted range) produced by ceramic plate magnets in such a way that some crabs were in proximity to the N pole, and others to the S pole. Control crabs were several feet away from the magnets. Regardless of the number of limbs autotomized, direction of the field, species, or salinity of the water, crabs in proximity to the S pole regenerated and molted sooner than con- trols, and those by the N pole were delayed in relation to controls. Factors such as magnetic effects on enzymatic reactions and ionic flux, orientation of animals within the magnetic field, and production of electric currents in the limb stumps may be involved in these effects. The opposite responses to N and S poles may be due to the opposite direction of the gradient vector. Regeneration of juvenile crabs, and of adult crabs during the summer months, was not significantly affected by the magnetic fields. INTRODUCTION Many studies have been done on the effects of magnetic fields on biological systems. Barnothy et al. (1956) found a decrease in leukocyte formation in mice subjected to a strong magnetic field. This was interpreted as retardation of the mitotic rate. Mulay and Mulay (1961) found that sarcoma 37 ascites tumor cells degenerated in response to a magnetic field of 4000 Oersteds, and Malinin et al. (1976) found inhibition of growth in various mammalian cell lines. Barnothy (1963) found that the growth rate of mice was retarded in a static magnetic field, though Eiselein et al. (1961) found no such effect. Reno and Nutini (1963) and Cook et al. (1969) found effects on metabolic rate: Rapidly growing tissues (tumors and embryonic kidney) exhibited a decrease in oxygen consumption. Some experiments have revealed different effects of north and south poles. Boe and Salunkhe (1963) found that tomato ripening was accelerated in a mag- netic field, and that tomatoes placed by the S pole of a magnet ripened faster than those placed by the N pole. Taneeva (1978) found differences in effects of N and S poles on Artemia growth and oxygen uptake. Limb regeneration in crabs involves a series of processes, starting with autot- omy, and followed by growth and differentiation of a limb bud. The regenerating limb bud grows in a folded position within a layer of cuticle and unfolds when the animal molts. Peripheral nerves are necessary for normal regeneration. Since regeneration ends with ecdysis, factors which influence the molt cycle and its con- Received October 8, 1979 ; accepted September 25, 1980. Abbreviations used: R, (limb bud length -H carapace width) X 100; N, north; S, south. 1 Present address : College of Medicine and Dentistry, Newark, N. J. - Author to whom correspondence should be addressed. 681 P. H. LEE AND J. S. WEIS trolling hormones can alter the rate of regeneration. Removal of eyestalks, a source of molt-inhibiting hormone, is a standard way of producing precocious molting and accelerated regeneration (Bliss, 1960). Skinner and Graham (1972) found that multiple autotomy, producing many regenerating limb buds, can also ac- celerate regeneration and ecdysis. Growth of regenerating limb buds in Crustacea is generally expressed in terms of "R value" which is ( limb bud length -=- carapace width) X 100 (Bliss, 1956). This is useful for comparing crabs of different sizes. Environmental factors including temperature, salinity, light, and water quality have been shown to affect regeneration and ecdysis (Bliss and Boyer, 1964; Rao, 1965 ; Weis, 1976a, b, 1977). The experiments to be reported here were designed to study the effects of magnetic fields on regeneration and ecdysis in the fiddler crabs Uca pugilator and Uca pugnax. MATERIALS AND METHODS Crabs were collected from N. J. or Long Island, X. V., or purchased from Gulf Specimen Co., Panacea, Fla. After autotomy (by pinching the merus) crabs were maintained in shallow artificial sea water (Instant Ocean) at 30/£f salinity and 25 °C in individual plastic cups 5.0 cm in diameter at the bottom. These cups were placed by either the north ( X ) or south ( S ) pole of the magnets. The close proximity of N and S crabs assured that other environmental factors would be the same for all. Cups of control crabs were beyond the magnetic field. A piece of non-magnetized metal was placed with some controls. All animals were fed Purina Fly Chow twice a week before measurement and changing of the water. Limb buds of the first walking leg were measured under a stereo- microscope with a calibrated ocular micrometer, and values converted to R values. R values for this limb bud approach 20 prior to ecdysis. We used ceramic chromate plate magnets 16 X 5 X 1.2 cm with the north pole on one flat surface and the south pole on the other. The measured strength of these magnets was non-uniform, but approximately 400-500 Gauss at the magnet, with a gradient of about 100 Gauss/cm close to the magnet, as measured by a Hall- effect gaussmeter. The gradient lessened further away from the magnet. The cups containing the crabs were placed next to the plate magnets, which stood on edge inside a plastic container (Fig. la) that was covered with a transparent cover. The inhomogenious field was about 250-50 Oersteds across the cups. Attempts to restrain crabs so they would have a permanent orientation with re- spect to the magnets were unsuccessful, as such animals died. Experiment 1 was performed in the late fall 1978 on specimens of U. pugnax from N. J. (14 mm mean carapace width) with four limbs autotomized. Six crabs were exposed to N, six to S, and six served as controls. Experiment 2, performed in December, used specimens of U. pugilator from Long Island (18 mm carapace width) with two limbs autotomized. Six were exposed to N, 6 to S, and 12 served as controls. Experiment 3, performed in February, used specimens of U. pugilator from Florida (15 mm mean carapace width) with six limbs re- moved. Twelve were exposed to N, 12 to S, and 12 served as controls, all in water of W/«, salinity. Experiment 4, performed in March, used specimens of U. pugilator from Florida (15 mm mean carapace width) with five limbs removed. Twelve were exposed to X, 12 to S, and 12 were controls. Experiments 5 and 6, MAGNETS AND CRAB REGENERATION 683 SN SN NS IQ Ib FIGL-RE 1. a: Arrangement of six cups by magnet for single field exposures, as in experi- ments 1-4. b : Arrangement of twelve cups by magnets for double field exposures, as in ex- periments 7-9. done in May and June, placed specimens of Florida LJ. pugilator (IS mm mean carapace width ) with seven limbs removed, in a vertical magnetic field, above and below the magnets, which were oriented horizontally. One magnet was oriented S pole upward, and another N pole up. Each was placed on top of a 8.5 cm diameter, 3.5 cm high fingerbowl containing three or four crabs. Another finger- bowl with three or four crabs was placed on top of each magnet. These crabs were therefore exposed to N or S from above or below. In experiments 7-9 crabs were placed between pairs of magnets (Fig. Ib) so that six were exposed to N on both sides (X-X), six were exposed to S on both sides (S-S), and six were exposed to X on one side and S on the other (X-S). (This last group is comparable to most of the previous experments cited in the introduction, in which organisms were placed between the poles of horseshoe magnets.) Additional crabs were placed on the outsides of the pairs of magnets to be exposed to a single N or S field (12 crabs each), and 12 crabs served as controls. Experiment 7 was done on specimens of Florida U. pugilator (17 mm mean carapace width) in June, and experiments 8 and 9 on specimens of Long Island U. pugilator during July and August respectively. Crabs in all these ex- periments had six limbs autotomized. Experiments 10-12 were done on juvenile specimens of U. pugilator (<13 mm carapace width). Experiment 10, done in April, exposed them to single horizontal fields, 12 crabs to X, 12 to S, and 12 controls. Experiment 11 (July) and 12 (August) exposed them to vertical fields as described in experiments 5 and 6. Sixteen crabs were exposed to N and 16 to S. Crabs in these experiments all had six limbs autotomized. Experiment 13 was done the following August (1980) on L.I. crabs (18 mm) with three limbs autotomized. Orientation : Orientation was investigated by recording the position of re- generating crabs on either side of the magnets, or on either side of a similar sized piece of cardboard acting as a "dummy magnet." The position of each animal (three crabs in front of a N pole, three behind a X pole, three in front of a S pole, three behind a S pole, six in front of the "dummy magnet," and six behind it) was recorded every 15 min for 6-hr periods on four consecutive days. Crabs were scored as facing toward, away from, or parallel to the magnet. "Toward" and "away" were defined as within the <179° roughly facing towards or away from the magnet. Only when animals were exactly parallel to the magnet were they recorded as parallel. Care was taken not to disturb the crabs during the observa- tion periods. P. H. LEE AND J. S. WEIS 16 14 12 10 R 8 18 16 14 12 R 10 10 20 30 DAYS 40 i 50 10 20 DAYS 30 40 FIGURE 2. Left: Mean R values of limb buds of crabs in Experiment 1. Right: Mean R values of limb buds of crabs in Experiment 4. Open circles = controls, closed circles = adjacent to N pole, squares = adjacent to S pole. Vertical lines — standard deviations. RESULTS The results of experiment 1 are shown in Figure 2, left. The crabs exposed to S pole regenerated limbs at an accelerated rate compared to controls, and those exposed to N pole regenerated more slowly than controls. All animals by the S pole molted by day 29, whereas 66% of the controls molted by day 31, and 0% of crabs by N pole molted by day 31. In experiment 2, crabs with only two limbs autotomized regenerated them much more slowly. Among the six crabs exposed to S pole, the R value at day 39 was 8.0 ± 0.8 (s.e.) and that for controls was 2.4 ± 0.5. Among those crabs by the N pole, five had not begun to regenerate, and one had reached an R value of 0.6. In experiment 3, in water of 8'/« salinity, the S pole again stimulated, and N pole retarded, growth compared to controls. For example, on day 11, the crabs at the S pole had R values of 5.2 ± 0.4 (s.e.), controls had R values of 2.4 ± 0.3, and those by the N pole had R values of 0.8 ± 0.2. In experiment 4, MAGNETS AM) CRAB REGENERATION 685 similar results were seen (Figure 2, right). By day 35, all of the crabs by S pole, 30% of controls, and one of the crabs by the N pole had molted. This one crab molted before its limbs were fully grown. The vertical fields in experiments 5 and 6 produced similar effects. On day 10 in experiment 5, crabs by the S pole had K values of 8.4 ± 0.9 (s.e.) whereas those by X pole had 5.2 ±1.1. In experiment 6, at 12 days after autotomy, the S-exposed crabs had R values of 8.3 ± 1.0, X-exposed had 4.1 ± 0.9, and controls had 6.0 ±0.9 (s.e.). In experiment 7, crabs exposed to X or to X-X were retarded, those exposed to S or to S-S were accelerated, and those exposed to X-S were retarded in com- parison to controls. However, differences were not as great as in previous ex- periments, and were often not statistically significant, whereas differences in previous experiments were highly significant. For example, on day 14, R values of controls were 9.6 ± 1.4 (s.e.), those for crabs at X were 6.3 ± 1.7 (n.s.), those for X-X were 4.7 ± 1.5 (significant to 0.05), those at S were 13.6 ± 1.4 (significant to 0.05), those at S-S were 11.1 ± 1.6 (n.s.) and those at X-S were 6.8 ± 1.3 (n.s.). By day 24, 66% of S, 50% of S-S, 30% of controls. \2% of X-S, and 0% of X and X-X had molted. When this was repeated in July (experiment 8), effects were generally reduced to non-significance, except for the retardation by X. For example, on day 10 controls had R values of 9.9 ± 0.7 (s.e.), S crabs had 9.7 ± 0.6, S-S had 10.1 ± 0.4. X crabs had 6.4 ± 0.9 (significant to 0.05). X-X crabs had 7.4 ± 1.3, and X-S crabs had 11.3 ± 1.5. In experiment 9 (August) the same phenomenon occurred. On day 13, con- trols had R values of 16.8± 1.0 (s.e.), S crabs had 16.6 ± 1.2, S-S crabs had 17.0 ± 1.0, X crabs had 13.2 ± 1.3. X-X crabs had 12.5 ± 1.4. and X-S crabs had 15.6 ± 1.5. Only growth of X-X crabs was significantly different from controls. By day 17, 50% of the controls, 50% of S-S crabs, 36% of S crabs, 20% of X-S crabs, 8% of X crabs, and 0% of X-X crabs had molted. Similarly, in experiment 13, no significant differences occurred. The experiments on the juveniles (experiments 10-12) revealed that they were unaffected by the horizontal or vertical magnetic fields. For example, in experiment 10, on day 9, R values of N-exposed crabs were 4.1 ± 0.6 (s.e.) and those of S-exposed were 4.5 ± 0.7. In experiment 11 (vertical field ) on day 7, those exposed to S had R values of 7.0 ± 1.0, and those exposed to N had 5.6 ± 0.9 (s.e.). On day 6 in experiment 12 (a repeat of 11 ), R values for those ex- posed to S was 4.9 ±0.5, and for those exposed to X were 5.1 ± 0.8. Xone of these are significantly different from each other. Results of all experiments are summarized in Table I. Orientation : In the orientation study, crabs in front of the X pole were re- corded as facing toward the magnet 72 times and away from the magnet 18 times (80% toward). Crabs in front of the S pole were facing toward the magnet 79 times, and away 11 times (88% towards). Crabs in front of the cardboard "dummy magnet" were facing towards it 154 times and away 28 times (85% towards). Of the crabs behind the magnets, those behind the X pole were facing towards it 64 times, and away 24 times (73% towards), those behind the S pole were facing towards it 58 times and away 30 times (66% towards), and those behind the cardboard were facing towards it 94 times, and away 51 times (65% towards). The differences in total numbers of readings are due to varying numbers of readings of orientation parallel to the magnets, and the presence of P. H. LEE AND J. S. WEIS TABLE I Growth of regenerating limb buds of fiddler crabs exposed to magnetic fields. N = north pole, S = south pole, C = control. In "significance" column, inner value is for difference from controls, or difference of S from N if there was no control (experiments 5, 10-12). Outer numbers indicate difference of S from N, or SS from NN, as indicated by the brackets. Expt. Month Carapace width #Legs off Days of growth Pole (n) R value Signif. by / test 2 Dec. 18 mm 2 39 N (6) 0 + C (12) 2.4 ± 0.5 (s.e.) S (6) 8.0 ± 0.8 0.001 3 Feb. 15 mm 6 11 N (12) 0.8 ±0.2 (s.e.) 0.001 C (12) 2.4 ± 0.3 S (12) 5.2 ± 0.4 0.001 5 May 18 mm 7 10 N (6) 5.2 ± 1.1 (s.e.) S (6) 8.4 ± 0.9 0.05 6 June 18 mm 7 12 N (8) 4.1 ± 0.9 (s.e.) n.s C (11) 6.0 ± 0.9 >0.01 S (8) 8.3 ± 1.0 0.05 7 June 17mm 6 14 NN (6) 4.7 ± 1.5 (s.e.) 0.05 N (12) 6.3 ± 1.7 n.s. ~~-~-^ C S (10) (12) 9.6 ± 1.4 13.6 ± 1.4 ^X>.oT>0.05 SS (6) 11. 1 ± 1.6 n.s. NS (6) 6.8 ± 1.3 n.s. 8 July 17 mm 6 10 NN (6) 7.7 ± 1.3 (s.e.) n.s. N (12) 6.4 ± 0.9 0.05 C (12) 9.9 ± 0.7 /0.01 S (12) 9.7 ± 0.6 n.s. SS (6) 10.1 ± 0.4 n.s. NS (6) 11.3 ± 1.5 n.s. 9 Aug. 15 mm 6 13 NN (6) 12.5 ± 1.4 (s.e.) 0.05 N (12) 13.2 ± 1.3 n.s. C (12) 16.8 ± 1.0 S (12) 16.6 ± 1.2 n.s. SS (6) 17.0 ± 1.0 n.s. NS (6) 15.6 ± 1.5 n.s. 10 Apr. 1 1 mm 6 9 N (12) 4.1 ± 0.6 (s.e.) S (12) 4.5 ± 0.7 n.s. 11 July 1 1 mm 6 7 N (16) 5.6 ± 0.9 (s.e.) S (16) 7.0 ± 1.0 n.s. 12 Aug. 11 mm 6 6 N (16) 5.1 ± 0.8 (s.e.) S (16) 4.9 ± 0.5 n.s. 13 July '80 18 mm 3 10 N (10) 5.2 ± 0.5 (s.e.) n.s. C (12) 7.2 ± 1.1 S (11) 6.6 ± 0.6 n.s. twice as many crabs by the cardboard. ( )f the parallel readings, no differences were seen in numbers of times spent facing left or right in any group. In all groups, crabs spent most of the time facing towards the magnet. DISCUSSION In experiments 1-7 proximity to the S pole was correlated with accelerated growth of regenerating limbs, and proximity to the X pole was correlated with retarded growth. A number of possible mechanisms may be involved in effects of magnetic fields on living systems. Haberditzl (1967) studied enzyme activities and found that carboxydismutase, glutamate dehydrogenase, and trypsin in solution produced gradients in response to a non-uniform magnetic field. Similarly, Barnothy (1964a) has stated that external magnetic fields can change the motion of elec- trolytes in cells, leading to aggregations of chemical substances at variance with MAGNETS AND CRAB REGENERATION 687 their normal distribution. Xeurath (1969) found that development of frog eggs was impaired if they were placed in a strong inhomogeneous field with their animal/vegetal gradient parallel to the field and gradient direction. Presumably, this disrupted the polarity of the eggs, which is necessary for proper development. Libofif (1965) has stated that magnetic fields interfere with diffusion of ions, and Labes (1966) has shown that a magnetic field of 1000 Oersteds influenced charge transport, both ionic and electronic, as well as reaction rates in biological systems. Gross (1964b) felt that the inhibition of biochemical reactions caused by magnetic fields was due to interference with rotation of paramagnetic molecules, such as free radicals, which can be intermediates in biological processes, and to alterations in bond angle orientation, which would impair closeness of fit of enzymes and substrates. Gualtierotti and Capraro (1964) found that magnetic fields changed the potential of frog skin by decreasing the influx of sodium ions. Since cell growth is related to diffusion of substances across the plasma membrane, these chemical fluxes in response to magnetic fields may be associated with the growth effects. One can imagine that as a crab moves through a magnetic field, voltage is in- duced across the limb stump and currents are produced. However, the animal need not move to generate a current. Barnothy and Barnothy (1974) have found that conduction currents are produced in inhomogeneous fields in which different parts of the experimental specimen are exposed to different field strengths. These currents can effect the nervous system. Becker (1961) found that trans- verse D.C. voltages were obtained when a steady state magnetic field of 2500 Oersteds was applied vertically to a non-moving salamander limb, but only when nerves were intact. When the magnetic field was removed, the original baseline voltage was resumed. Oberg (1973) has found magnetic effects on nerve activity in frogs, and Russell (1969) found that magnetic fields affected nerve impulses in cockroaches. Bioelectricity has been found to play an important role in regeneration. Borgens ct al. (1977a) have reported that currents normally leave the tip of re- generating salamander limb buds. Furthermore, the same investigators (1977b) found that the application of a 0.2 /u,A current to the limb stumps of adults frogs (Rana pipiens) could stimulate partial regeneration if the current was cathodal (distally negative). The application of anodal current (distally positive) caused extensive destruction of the limb stump. Control frogs (sham treated and un- treated) merely healed. These effects of oppositely directed current may be related to the effects of N and S poles observed here. Borgens ct al. (1979) found that causing a steady small current to leave limb stumps of Xcnopus en- hanced limb regeneration in this anuran. In both Rana and Xcnopns the cath- odally stimulated regenerates had increased the innervation to the regenerating tissue, and the authors felt that the current's effects were mediated by early en- hancement of nerve growth. That electric currents can affect crab limb regenera- tion was demonstrated by Mantel and Levin ( 1973). The regenerating limb buds of Uca are pointed in a generally upward angle, which is the orientation of the merus. If somehow the crabs by the S pole had currents induced upward (and therefore out of the limb stump) and those by the N pole had current induced downward, this could help explain the effects. A problem with these possible mechanisms is that the crabs were not im- mobilized to be subjected to a magnetic field of constant direction, but were free to face in any direction. However, it has been observed that various organisms P. H. LEE AND J. S. WEIS orient in magnetic fields (Brown et al., 1960a, Brown et al., 1960b, Barn well and Brown, 1964). Experiments of Gross (1964a) and Gross and Smith (1964) produced results (in retarding tumor growth and wound healing) in unrestrained mice in a horizontal magnetic field. Barnothy (1964b) says that unrestrained animals in a horizontal magnetic field will show attenuation rather than anullment of magnetic effects, since animals will not have opposite orientations for the same length of time, and therefore will not totally cancel out cumulative reversible ef- fects. The preferred orientation of crabs, facing towards the magnets, supports this idea. Gerencser ct al. (1964) found that inhomogeneous fields reduced growth of bacteria due to a paramagnetic phenomenon. Rod-shaped bacteria were more affected than spherical ones. They suggested that the effects did not cancel out because, although the bacteria were free to change their position, they could have been oriented by the magnetic field and therefore tended to keep their position relative to the field and the gradient. Recently, Frankel et al. (1979) found that some bacteria do, in fact, orient in a magnetic field due to the presence of iron as magnetite within them. Magnetite has also been found in the abdomens of bees (Gould et al., 1978), and in the heads of pigeons (Walcott et al., 1979), which also can orient in magnetic fields. Visalberghi and Alleva (1979) found that magnets placed N-pole-up on the heads of homing pigeons caused disorienta- tion. This was not observed in birds with magnets placed S pole up. These studies show that magnetic fields can be perceived by these animals, and can play a role in their normal activities. It is therefore possible that the earth's magnetic field might have adaptive effects on fiddler crabs' daily activities. Otoliths can be highly paramagnetic and can be influenced by magnetic fields (Barnothy, 1964b). However, the data on orientation of crabs showed that they oriented toward the piece of cardboard as well as toward the magnets. This indi- cates that the presence of the object, rather than its magnetic field, was responsible for the orientation in these experiments. It is possible, however, that a more ex- tensive and rigorous investigation of orientation would have revealed differences between orientation to the cardboard and to the magnets. Animals placed in a vertical magnetic field can be unrestrained and still remain in a constant position relative to the field and gradient vectors, since they do not become supine. In our experiments with vertical fields on adult crabs (5 and 6), clear effects were seen. Furthermore, these crabs were extremely active, much more so than any crabs in horizontal fields or in no applied magnetic field. Their hyperactivity may indicate perception of the field and an attempt to orient them- selves differently (vertically ?) within it, which they were unable to do. An asymmetry in the experiments as they were set up is as follows : crabs at both sides of the magnet are exposed to a magnetic field of the same absolute spatial direction (i.e. N— >S). However, the crabs at the N pole of the magnet are exposed to an inhomogeneous field whose gradient is in the opposite direction, since the gradient vector points in the direction in which increase in field strength occurs. These crabs had reduced growth. Conversely, the crabs at the S pole are experiencing an inhomogeneous field whose gradient vector is in the same direction as the field vector, due to their placement immediately north of the magnet's S pole. Some phenomena have been shown to be gradient sensitive rather than field sensitive (Barnothy, 1964a). These include arrest of tumor growth and lethal effects in Drosophila and mice. Barnothy (1964b) has said that an inhomogeneous field exerts an acceleratory force in the direction of the gradient vector on molecules that have an intrinsic magnetic moment, or whose MAGNETS AND CRAB REGENERATION 689 susceptibility differs from that of the environment. This force is independent of the field vector, but depends only on the gradient vector. Our observed effects may be dependent on this property of the inhomogeneous field. The lack of effects on juveniles and on adults in summer is puzzling. It may be related to the fact that regeneration is normally very rapid during this time, as the crabs approach their normal time of ecdysis, and that they tend to be more refractory to environmental influences. Crabs also have been found to have reduced sensitivity to heavy metal toxicants during this time of year (Weis, 1976b). Thorp and Lake (1974) found a similar seasonal change in the sensi- tivity of shrimp (Paratya tasmaniensis) to cadmium. Juvenile crabs normally go through their molt cycle more rapidly, and their growth has been found to be more refractory to other environmental influences as well (Rao, 1965). It would appear that individuals with more rapid growth rates are less sensitive to environ- mental effects. However, Fingerman and Fingerman (1979) found that indi- viduals with faster growing limb buds were more sensitive to Aroclor 1242 than those with slowly growing limb buds. Therefore, the reasons for the lack of sensitivity of juveniles and of adults in the summer are not clear. It is possible that juveniles would respond to magnetic fields in the winter months, or if fewer limbs were removed. It is also possible that since the volume of limb buds is less and less tissue is being affected, no significant differences could be observed. We do not expect that the preceding discussion has thoroughly explained the phenomenon we have observed. Whether or not we can explain it completely, the effects noted are of interest and should be investigated further. ACKNOWLEDGMENTS We wish to thank Drs. V. Santarelli, B. Sonnenblick, S. McDowell, and P. Weis for reviewing the manuscript and offering their comments. Ms. Jennifer Weis is thanked for her technical assistance. This research was supported by Biomedical Research Support Grant RR7059. P.H.L. was a student in the Honors Program during the course of this research. LITERATURE CITED BARNOTHY, J. M., 1963. Growth rate in mice in static magnetic fields. Nature, 200: 86-87. BARNOTHY, J. M., 1964a. Introduction. Pp. 3-24 in M. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. I. Plenum Press, New York. BARNOTHY, J. M., 1964b. The vector character of field and gradient and its possible im- plications for biomagnetic experiments and space travel. Pp. 56-62 in M. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. I. Plenum Press, New York. BARNOTHY, J. M., M. F. BARNOTHY, AND I. BOSZORMENYI-NAGY, 1956. Influence of a mag- netic field upon the leukocytes of the mouse. Nature, 177 : 577-578. BARNOTHY, M. F., AND J. M. BARNOTHY, 1974. Magnetobiology. Pp. 313-349 in N. B. Slonim, Ed. Environmental Physiology. C. B. Mosby Co., St. Louis. BARNWELL, F. H., AND F. A. BROWN, JR., 1964. Responses of planarians and snails. Pages 263-278 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. I. Plenum Press, New York. BECKER, R. O., 1961. Search for evidence of axial current flow in peripheral nerves of the salamander. Science, 134 : 101-102. BLISS, D. E., 1956. Neurosecretion and the control of growth in a decapod crustacean. Pp. 56-75 in K. G. Wingstrand, Ed. Bcrtil Hanstrom, Zoological Papers in Honor of his Sixty-fifth Birthday, Nov. 20, 1956. Zool. Inst. Lund, Sweden. BLISS, D. E.( 1960. Autotomy and regeneration. Pp. 561-589 in T. Waterman, Ed. Physiol- ogy of Crustacea. Academic Press, New York. P. H. LEE AND J. S. WEIS JLISS, D. E., AND J. R. BOVER, 1964. Environmental regulation of growth in the decapod crustacean Gccarcinus latcralis. Gen. Comp. Endocrinol. 4: 15-41. BOE. A. A., AND D. K. SALUNKHE, 1963. Effects of magnetic fields on tomato ripening Nature, 199: 91-92. BORGENS, R., J. VANABLE, AND L. JAFFE, 1977a. Bioelectricity and regeneration: Large cur- rents leave the stumps of regenerating newt limbs. Proc. Nat. Acad. Sci. U. S. A 74 : 4528-4532. BORGENS, R., J. VANABLE, AND L. JAFFE, 1977b. Bioelectricity and regeneration I. Initiation of frcg limb regeneration by minute currents. /. Exp. Zoo/., 200 : 403^16. BORGENS, R., J. VANABLE, AND L. JAFFE, 1979. Small artificial currents enhance Xcnopus limb regeneration. /. Exp. Zoo/., 207 : 217-226. BROWN, F. A., JR., M. F. BENNETT, AND H. M. WEBB, 1960a. A magnetic compass response of an organism. Bill. Bull., 119: 65-74. BROWN, F. A., JR., W. J. BRETT, M. F. BENNETT, AND F. H. BARNWELL, 1960b. Magnetic response of an organism and its solar relationships. Biol. Bull., 118: 367-381. COOK, E. S., J. C. FARDON, AND L. G. NUTINI, 1969. Effect of magnetic fields on cellular respiration. Pp. 67-78 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. 2. Plenum Press, New York. EISELEIN, B. S., H. M. BOUTEI.L, AND M. W. BIGGS, 1961. Biological effects of magnetic fields— negative results. Aerosp. Mcd., 32 : 383-386. FINGERMAN, S., AND M. FIN-GERMAN, 1979. Effects of two polychlorinated biphenyls (Aroclors 1242 and 1254) on limb regeneration in the fiddler crab, Uca pugilator, at different times of the year. Vie Milieu, 28-29 : 69-75. FRANKEL, R. B., R. P. BLAKEMORE, AND R. S. WOLFE, 1979. Magnetite in freshwater mag- netotactic bacteria. Science, 203 : 1355-1356. GERENCSER, V. F., M. F. BARNOTHY, AND J. M. BARNOTHY, 1964. Inhibition of bacterial growth in fields of high paramagnetic strength. Pp. 229-239 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. 1. Plenum Press, New York. GOULD, J. L., J. L. KIRSCHVINK, AND K. S. DEFFEYES, 1978. Bees have magnetic remanence. Science, 201 : 1026-1028. GROSS, L., 1964a. Lifespan increase of tumor-bearing mice through pretreatment. Pp. 132-139 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. 1. Plenum Press, New York. GROSS, L., 1964b. Distortion of the bend angle in a magnetic field and its possible magneto- biological implications. Pp. 74-79 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. 1. Plenum Press, New York. GROSS, L., AND L. W. SMITH, 1964. Wound healing and tissue regeneration. Pp. 140-145 in M. F. Barnothy, Ed. Biological Effects of Magnetic Fields. Vol. 1. Plenum Press, New York. GUALTIEROTTI, T., AND V. CAPRARO, 1964. The action of magnetic field on sodium transport across cell membrane. Pp. 311-316 in M. Florkin and A. Dollfus, Eds. Life Sciences and Space Research. A Symposium. John Wiley & Sons, New York. HABERDITZL, W., 1967. Enzyme activity in high magnetic fields. Nature, 213: 72-73. LABES, M. M., 1966. A possible explanation for the effect of magnetic fields on biological systems. Nature, 211 : 968. LIBOFF, R. L., 1965. A biomagnetic hypothesis. Biophys. J., S: 845-853. MALININ, G. L, W. D. GREGORY, L. MORELLI, V. K. SHARMA, AND J. C. HOUCK, 1976. Evi- dence of morphological and physiological transformation of mammalian cells by strong magnetic fields. Science, 194 : 844-846. MANTEL, L. H., AND R. S. LEVIN, 1973. Effects of multiple limb removal and stimulation on regeneration and molting in Uca pugilator. Am. Zoo/., 13: 1339. MULAY, I. L., AND L. W. MULAY, 1961. Effect of a magnetic field on Sarcoma 37 ascites tumor cells. Nature, 190 : 1019. NEURATH, P. W., 1969. The effect of high gradient, high strength magnetic fields on the early embryonic development of frogs. Pp. 177-187 in M. F. Barnothy, Ed., Bio- logical Effects of Magnetic Fields. Vol. 2. Plenum Press, New York. OBERG, P. A., 1973. Magnetic stimulation of nerve tissue. Mcd. Biol. Eng., 11: 55-64. RAO, K. R., 1965. Studies on the influence of environmental factors on growth in the crab Ocvpode macrocera H. Milne Edwards. Crustaceana, 11 : 257-276. MAGNETS AND CRAB REGENERATION 691 RUSSELL, D. R., 1969. Effect of a constant magnetic field on invertebrate neurons. Pp. 227-232 in M. F. Barnothy, Ed., Biological Effects of Magnetic Fields. Vol. 2. Plenum Press, New York. RENO, V. R., AND L. G. NUTINI, 1963. Effect of magnetic fields on tissue respiration. Nature. 198: 204-205. SKINNER, D. M., AND D. E. GRAHAM, 1972. Loss of limbs as a stimulus to ecdysis in Brachyura (true crabs). Biol. Bull.. 143 : 222-233. TANEEVA, A. I., 1978. Changes in physiological processes in hydrobionts under the effect of a constant magnetic field. Gidrobiol. Zh., 114 : 56-63. THORP, V. J., AND P. S. LAKE, 1974. Toxicity bioassays of cadmium on selected freshwater invertebrates and the interaction of cadmium and zinc on the freshwater shrimp, Paratya tasmamcnsis (Riek). Austr. J. Mar. Frcshu*. Res.. 25 : 97-104. VISALBERGHI, E., AND E. A M.EVA, 1979. Magnetic influences on pigeon homing. Biol. Bull., 156: 246-256. WALCOTT, C., J. L. GOULD, AND J. L. KIRSCHVINK, 1979. Pigeons have magnets. Science, 205 : 1027-1028. WEIS, J. S., 1976a. Effects of environmental factors on regeneration and molting in fiddler crabs. Biol. Bull.. ISO: 52-62. WEISS, J. S., 1976b. Effects of mercury, cadmium, and lead salts on regeneration and ecdysis in the fiddler crab, Uca pugilator. Fish. Bull.. 74 : 464-467. WEIS, J. S., 1977. Limb regeneration in fiddler crabs: species differences and effects of methylmercury. Biol. Bull., 152 : 236-274. Reference : Biol, Bull., 159 : 692-699. (December, 1980) RESPIRATORY METABOLISM OF MACROBRACHIUM OLFERSI1 (WIEGMANN) ZOEAE DURING THE MOULTING CYCLE FROM ECLOSION TO FIRST ECDYSIS JOHN C. McNAMARA,1 GLORIA S. MOREIRA, AND PLINIO S. MOREIRA Institute Occanografico and Institute dc Biocicncias, Univcrsidadc dc Sao Paulo, Brasil ABSTRACT The respiratory metabolism of first stage Macrobrachiuui olfersii (Wiegmann) zoeae was measured at 12 hr intervals with Cartesian diver microrespirometers throughout the 120 hr period from eclosion to first ecdysis. Weight-specific oxygen consumption rates, although slightly higher immediately after hatching, remained constant and did not differ significantly during the first moulting cycle. Larval dry weight decreased from 21.8 to 18.9 /xg during this period. There was no evidence of a diurnal metabolic pattern. The moulting cycle was subdivided according to morphological characteristics, based on the degree of cytoplastnic homogeneity, epidermal retraction, and stage of setal development in the telson. Larval metab- olism is suggested to be uniform throughout the eclosion-first ecdysis period, in spite of a distinct and divisible moulting cycle. INTRODUCTION Although some aspects of adult crustaceans' physiology have been well studied in relation to the moulting cycle (see Passano, 1960; Charniaux-Cotton and Klein- holz, 1964; Drach and Tchernigovtzeff, 1967; Novales et al., 1973; and Kleinholz, 1976 for references), the relationships between respiratory rate and moult- inhibiting hormone, and the putative control of respiration by a separate eyestalk hormone, still are not well understood (Silverthorn, 1975a, b; Kleinholz, 1976). Several authors have studied the respiratory metabolism of intact adult crustaceans as a function of the moulting cycle (Scudamore, 1947; Roberts, 1957; Costlow and Bookhout, 1958; Halcrow and Boyd, 1967; Bulnheim, 1974; Hagerman, 1976). However, such studies of the relationship between larval metabolism and moulting are lacking. The moulting cycle of some larval decapod crustaceans has been subdivided and morphologically characterized (Rao et al., 1973; Van Herp and Bellon-Humbert, 1978; Huner and Colvin, 1979; Freeman and Costlow, 1980), although palaemonid larvae have received little attention. The present study evaluates the possible influence of moulting-cycle-related events on the respiratory metabolism of the first zoea of a palaemonid shrimp, Macrobrachinrn olfersii (Wiegmann), and offers a tentative morphological basis for the subdivision of the moulting cycle of early zoeae of this species. Received March 25, 1980; accepted September 25, 1980. 1 Address for reprint requests : Departamento de Fisiologia, Institute de Biociencias, Caixa Postal 11.461, Universidade de Sao Paulo, 05421 Sao Paulo, Brasil. 692 LARVAL RESPIRATION AND MOULTING 693 MATERIALS AND METHODS Ovigerous females of the freshwater palaemonid shrimp Macrobrachinin oljcrsii were maintained in the laboratory in aerated aquaria filled with river water (5'/cc salinity). Specimens were collected from the lower reaches of the Guaeca River (approx. 23° 49' 18" S ; 45° 27' 18" W) in the State of Sao Paulo, Brazil. Larvae, usually released in the early evening between 2000 and 2200 hr, were collected by pipette, transferred to small glass bowls containing 100 nil of river water mixed with seawater to a final salinity of \4'/,(, and kept in a constant temperature chamber at 20°C under 12 hr light: 12 hr dark. The larvae were not fed throughout the 5-day experimental period, since our previous results showed larval yolk reserves to be sufficient nourishment during the first zoeal stage (Moreira ct al., 1979). Oxygen consumption of individual Stage I zoeae was measured using Cartesian diver microrespirometers (Holter, 1943), each having a total volume of between 8 and 13 p.\. Measurements were made at 12 hr intervals beginning 2 hr after hatching, for 120 hr, which in this species effectively represents the first larval moulting cycle. The respiratory rates of a minimum of seven larvae were deter- mined at each 12 hr interval. Zoeae were micropipetted into 0.8 /A of dilute seawater, \4(/cc salinity. Following an equilibration period of 30 min, diver read- ings were taken at 30 min intervals over 1 hr, during which values were constant. Such measurements represent "routine metabolism" (Prosser, 1973), since zoeae could move slightly, but not swim, inside the divers. All experiments were per- formed at 20°C (controlled by a thermostat regulating with 0.01°C). In order to render the equilibrium pressure independent of barometric changes, one end of the manometer was connected by a thick-walled capillary tube to a 4-1 barostatic bottle completely submerged in a water bath. Each zoea was used only once and was examined after removal from the respirometer to verify condition and degree of activity. Results are expressed both as jul O;> consumed • mg dry wt'Mir"1 and /A O^ consumed- larva"1 -hr"1. Significant differences of means were tested according to Parl (1967). To determine successive larval dry weights, 15 larvae from selected batches (at 24, 72, and 96 hr) were briefly rinsed with distilled water, dried overnight at 80°C, placed in a dessicator for 2 hr and weighed on a Calm Gram electrobalance (0.1 /xg sensitivity). Groups of larvae, after being used for respirometric determinations, were pre- served separately in 4% formalin. The tails of these larvae were severed at the junction of the telson and last abdominal segment, mounted in glycerine, and covered with a coverslip. Photomicrographs of the inner 4th and 5th telson setae were taken with Kodak Panatomic 32 ASA film using an Olympus photomicro- scope. RESULTS Figure 1 shows the respiratory rates of newly hatched Stage I Macrobrachium oljersii zoeae, measured at 12 hr intervals over a 120 hr period. Larval dry weights were calculated as 21.8, 19.9, and 18.9 /mg at 24, 72, and 96 hr, respectively, after hatching. The small weight loss was presumably due to the larvae's use of yolk- reserves. Weight-specific oxygen-consumption rates were slightly higher immedi- ately after hatching (2 hr) but quickly reached a stable level after 26 hr. No significant difference in respiratory rate was recorded during the subsequent experi- mental period, including possible alterations associated with measurements made 694 McNAMARA, MOREIRA, AND MOREIRA 12 4 - 14 26 38 50 62 74 86 98 no (22 HOURS AFTER HATCHING FIGURE 1. Rate of oxygen consumption by State I Macrobrachiiun olfcrsii zoeae during 120 hrs from eclosion to first ecdysis. Solid line, ^10;. consumed • larva"1 • hour"1; broken line, |il02 consumed • milligram dry weight"1 • hour"1. Vertical bars = SE of the mean. during daylight and night hours. Weight-specific oxygen-consumption rates appeared to be somewhat lower just before moulting, but differences were not statistically significant. One zoea actually moulted inside the diver with no modifi- cation in respiratory rate recorded. Oxygen consumption per animal steadily diminished over the experimental period, although no significant differences were recorded after 26 hr post-hatching, in spite of the concomitant decrease in larval dry weight. During the 120 hr experimental period most larvae underwent a complete moult- ing cycle (Figs. 2-7). During the first 26 hr after hatching, larvae pass from Stage A to Stage B. In Stage A (Fig. 2) the epidermal and setal cytoplasm is dense, with cellular elements clearly evident. The epidermis completely fills the spaces under the setae and their articulations. In Stage B (Fig. 3) the epidermal cytoplasm becomes more homogenous, still filling the space under the cuticle. Cone-like structures begin to form at the setal bases. After 50 hr, the larvae have reached substage D0 (Fig. 4) in which the epidermis has begun to retract from the cuticle between the setae. The setal cytoplasm, however, remains clearly connected to the telson epidermis. By 74 hr, most larva have attained at least substage D/ (Fig. 5), characterized by a well defined epidermal retraction between the setae and at their bases. The partially formed new seta is surrounded by a bulb-like invagination, the proximal end of which is not well defined in the epidermal matrix. After 96 hours, most larvae have reached substage D/ ' ' (Fig. 6). Setal invagination is complete, although the open ends of the new setae are not well defined. Barbules can be seen on the new setal walls. New setules are evident in the space between the retracted epidermis and the cuticle. Immediately after the moult to the second zoea (betwen 96 and 120 hr), setal and epidermal components (Fig. 7) appear identical to those seen in Stage A of the newly hatched zoea I (Fig. 2). DISCUSSION Scudamore (1947), Scheer and Scheer (1954), Roberts (1957), Halcrow and Boyd (1967), Paranjape (1967), Bulnheim (1972, 1974), and Hagerman (1976) all reported increased oxygen consumption associated with the immediate pre- and/or post-moult periods in a variety of intact adult crustaceans, including amphi- pods, decapods, euphausiids, and isopods. Skinner (1962) has also shown that LARVAL RESPIRATION AND MOULTING 695 oxygen consumption by isolated integumentary tissues increases during exocuticle synthesis, immediately prior to ecdysis in Gecarcinus latcralis. However, some researchers have reported results similar to those of the present study, e.g., Costlow and Bookhout (1958) and Barnes and Barnes (1963), studying the cirripedes Balanus amphitrite and Balanus balanoidcs, respectively, demonstrated there was either no significant increase (former) or only a small increase (latter) in respira- tory rate at ecdysis. Vernberg and Costlow (1966) noted a lack of significant difference in respiratory rate for Uca puynax first zoeae during 3 days. Crustacean metabolism has been postulated to be either directly regulated by specific eyestalk hormones (Silverthorn, 1975a, b) or indirectly modified by moult- related hormones also released from eyestalk neurosecretory centers (Passano, 1960). On this basis, a functional analysis of the neurosecretory structures in the larval crustacean eyestalk is essential to understanding of larval respiratory responses. Hubschman (1963) showed that the sinus gland is not recognizable until the fifth zoeal stage in several species of Palaemonetes and that the ganglionic X-organs are not functional during larval development. Significantly, eyestalk removal had no effect on the larval moulting cycle nor on metamorphosis. Bellon- Humbert et al. (1978) also showed that the sinus gland in Palaemon serratus is not discernable until the fifth zoeal stage, becoming active only after metamorphosis. The ganglionic X-organs (medullae externa and terminalis) are not recognizable before metamorphosis in this species, appearing only during the postlarval phase. Our findings that moulting-cycle-relatecl events do not affect the respiratory metabolism of Macrobrachium olfersii first zoeae, combined with the ontogenetic data of Hubschman (1963) and Bellon-Humbert ct al. (1978) above, suggest that if a moult-inhibiting hormone is present in early palaemonid zoeae in the eyestalks or elsewhere, it does not appear to affect respiration. Such evidence also indirectly suggests that the larval moulting cycle may be uninhibited, i.e., the larval Y-organ, if it exists and is functional, may produce moult-promoting substances continuously. The rapid moulting cycles (3-4 days) of Macrobrachium olfersii and M. holthusi early zoeae (Moreira et al., 1979) certainly suggest this. It is unfortunate that so little is known of larval neurosecretory processes in relation to moulting control. Clearly, as suggested by Hubschman (1963), such processes are very different from those known to operate in adults. Few authors have studied the morphological changes in epidermal and cuticular structure accompanying larval moulting cycles. Broad and Hubschman (1963), observing the larval development of Palaemonetes kadiakensis, noted gross morpho- logical details associated with post-moult stages and suggested that the proecdysial stages may be of relatively long duration. Rao et al. (1973) briefly characterized Stages C and Do-D/' ' of fourth and fifth stage Homarus americanus larvae. Like the first zoea of M. olfersii, Stage D0 in these larvae was characterized by retraction of the epidermis, and Stages D/-D/ by formation of new setae. However, contrary to the situation in M. olfersii first zoeae, the new setae are completely separated from the old cuticle and setae. Van Herp and Bellon-Humbert (1978) have produced the most complete subdivision of the larval moulting cycle to date, using second-fourth stage Astacus leptodactylus larvae. The characteristics of each stage of the moulting cycle resemble those of M. olfersii first zoeae, although the nerve fiber described from A. leptodactylus setae was not observed in M. olfersii setae. Similarly, the conelike structures described at the setal bases in Stage B and the bulblike structures (setal organ?) described in Stage D/ of M. olfersii larvae were not noted in A. leptodactylus larval uropods. Huner and Colvin 696 McNAMARA, MOREIRA, AND MOREIRA 4 'A LARVAL RESPIRATION AND MOULTING FIGURE 2. Telson and setae of Macrobrachium olfcrsii zoea I in Stage A. Cellular elements (arrows) prominent in both setal and epidermal cytoplasm. 300X. FIGURE 3. Stage B. Both setal and epidermal cytoplasm is homogeneous. Cone-like struc- tures visible (arrows) at setal bases. 300X. FIGURE 4. Substage D,,. Epidermis has begun to retract from cuticle leaving a well denned space s, narrower in the intersetal areas (arrow). 300X. FIGURE 5. Substage D/. Epidermal retraction s closely follows curves of cuticle with uniform spacing. Bulb-like invagination b surrounds the base of each forming seta. 300X. FIGURE 6. Substage Di' ' '. Setal invagination i is complete. Barbules, b, project from internal walls of new seta. Setules, s, appear in the space between old cuticle and retracted epidemis. 300 X. FIGURE 7. Telson and setae of Macrobrachium olfcrsii zoea. II in Stage A. Cellular elements (arrows) are evident in heterogeneous setal and epidermal cytoplasm. 300 X. (1979) described moulting cycle subdivisions for juvenile Pcnaeiis calijomicnsis and P. stylirostris that approximate those of M. olfersii first zoeae, differing only in some details, e.g., the absence of inner setal cones in Stage C of Penaeus juve- niles. Freeman and Costlow (1980) characterized the zoeal and megalopal moult- ing cycles of Rhithropanopcus harrisii using the morphological changes occurring in the integument, particularly in the spines and antennae. In R. harrisii larvae, the general events correlate with those taking place in M. oljcrsii first zoeae. However, a direct comparison was not possible as these authors did not describe phases of setal development. Although conforming to the general scheme outlined by Drach and Tcherni- govtzeff (1967) for crustacean moulting cycle subdivision, our preliminary data for M. olfersii first zoeae have revealed more morphological details than previously described for larval crustaceans. As previously suggested by McNamara (1979), detailed studies at the cellular level are necessary to elucidate the complex modifica- tions occurring in the crustacean integument during the moulting cycle. The first McNAMARA, MOREIRA, AND MOREIRA zoea of M. olfersii has proved to be excellent material for such studies owing to its fine transparent cuticle and simple epidermis. The present study has emphasized that although modifications in metabolic rate are usually associated with the moulting cycle of crustaceans, such modifications do not necessarily occur in early developmental stages, suggesting that studies throughout the developmental sequence are necessary to a better understanding of the relationship between metabolism and the crustacean moulting cycle. ACKNOWLEDGMENTS The authors would like to express their gratitude to Dr. L. Ludovico George of the Biomedical Sciences Institute, University of Sao Paulo, for the use of photo- micrographic equipment. This study was supported in part by the Organization of American States (OAS). LITERATURE CITED BARNES, H., AND M. BARNES, 1963. The relation of water uptake and oxygen consumption of the body tissues to the moulting cycle in Balanus balanoidcs (L. ). Crustaccana, 6: 62-68. BELLON-HUMBERT, C., M. J. TnijssEN, AND F. VAN KERF, 1978. Development, location and relocation of sensory and neurosecretory sites in the eyestalks during the larval and post-larval life of Palacmon scrratus (Pennant). /. Mar. Biol. Assoc. U. K., 58: 851-868. BROAD, A. C., AND J. H. HUBSCHMAN, 1963. The larval development of Palacmonctcs kadiakensis. M. J. Rathbun in the laboratory. Trans. Am. Microsc. Soc., 82 : 185-197. BULNHEIM, H.-P., 1972. Vergleichende Untersuchungen zur Atmungsphysiologie euryhaliner Gammariden unter besonderer Berkiicksichtigung der Salzgehaltsanpassung. Hclgo- lander Wiss. Meeresunters., 23 : 485-534. BULNHEIM, H.-P., 1974. Respiratory metabolism of Idotca balthica (Crustacea, Isopoda) in relation to environmental variables, acclimation processes and moulting. Hclgolandcr Wiss. Meeresunters., 26: 464-480. CHARNIAVX-COTTON, H., AND L. H. KLEIN HOLZ, 1964. Hormones in invertebrates other than insects. Pp. 135-198 in G. Pincus, K. V. Thimann, and E. B. Astwood, Eds., The Hormones. Academic Press, New York. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1958. Molting and respiration in Balanus amphitritc var. dcnticulata Broch. Physiol. Zool., 31 : 271-280. DRACH, P., AND C. TCHERNIGOVTZEFF, 1967. Sur la methode de determination des stades d'intermue et son application generale aux crustaces. Vic ct Milieu, Ser. A., 18 : 595-607. FREEMAN, J. A., AND J. D. COSTLOW, 1980. The molt cycle and its hormonal control in Rhithropanopcus Iwrrisii larvae. Develop. Biol., 74: 479-485. HAGERMAN, L., 1976. The respiration during the moulting cycle of Crangon vulgaries (Fabr.) (Crustacea, Natantia). Ophelia, 15: 15-21. HALCROW, K.. AND C. M. BOYD, 1967. The oxygen consumption and swimming activity of the amphipod Gammarus occanicus at different temperatures. Comp. Biochcm. Physiol., 23 : 233-242. HOLTER, H., 1943. Technique of the Cartesian diver. C. R. Trav. Lab. Carlsberg. Scr. Chiin., 24 : 399-478. HUBSCHMANN, J. H., 1963. Development and function of neurosecretory sites in the eyestalks of larval Palacinonctes (Decapoda: Natantia). Biol. Bull., 125: 96-113. HUNER, J. V., AND L. B. COLVIN, 1979. Observations on the molt cycle of two species of juvenile shrimp, Pcnacus califoruiensis and Penacus stylirostris (Decapoda: Crustacea). Proc. Nat. Shellfish. Assn., 69 : 77-84. KLEIN HOLZ, L., 1976. Crustacean neurosecretory hormones and physiological specificity. Amer. Zool., 16: 151-166. McNAMARA, J. C., 1979. Ultrastructural observations on the endoplasmic reticulum of epidermal cells in the branchiostegite of Macrobrachiwn holthuisi (Crustacea: Decapoda). Biol. Fisiol. Animal, Univ. S. Paulo, 3: 73-80. LARVAL RESPIRATION AND MOULTING 699 MOREIRA, G. S., J. C. MCNAMARA AND P. S. MORKIRA. 1979. The combined effects of temper- ature and salinity on the survival and moulting of early zoeae of Macrobrachium holthuisi (Decapoda: Palaemonidae). Biol. Fisiol. Animal, Univ. Sao Paulo, 3: 81-93. MOREIRA, G. S., J. C. MCNAMARA, P. S. MOREIRA, AND M. WEINRICH, 1980. Temperature and salinity effects on the respiratory metabolism of the first zoeal stage of Macrobrachium holthuisi Genofre & Lobao (Decapoda: Palaemonidae). /. Exp. Mar. Biol. Ecol., 47 : in press. NOVALES, R. R., L. I. GILBERT, AND F. A. BROWN, JR., 1973. Endocrine mechanisms. Pp. 857- 908 in C. L. Prosser, Ed., Comparative Animal Physiology. W. B. Saunders Co., Philadelphia. PARANJAPE, M. A., 1967. Molting and respiration of euphausiids. ./. Fish. Res. Bd. Canada, 24: 1229-1240. PARL, B., 1967. Basic Statistics. Doubleday and Co., New York, 364 pp. PASSANO, L. M., 1960. Molting and its control. Pp. 473-536 in T. H. Waterman, Ed.. The Physiology of Crustacea. Vol. 1, Academic Press, New York. PROSSER, C. L., 1973. Oxygen: respiration and metabolism. Pp. 163-211 in C. L. Prosser, Ed., Comparative Animal Physiology. 3rd Edition. W. B. Saunders Co., Philadelphia. RAO, K. R., S. W. FINGERMAN, AND M. FiNGERMAN, 1973. Effects of exogenous ecdysones on the molt cycles of fourth and fifth stage American lobsters, Homarus amcricainis. Com p. Biochcm. Physiol., 44: 1105-1120. ROBERTS, J. L., 1957. Thermal acclimation of metabolism in the crab Pachyyrapsus crassipcs Randall. I. The influence of body size, starvation and moulting. Physiol. Zool., 30: 232-242. SCHEER, B. T.. AND M. A. SCHEER, 1954. The hormonal control of metabolism in crustaceans. VIII. Oxygen consumption in Lca.ndcr scrratus. Pubbl. Staz. Zool. Napoli, 25 : 419^26. SCUDAMORE, H. H., 1947. The influence of the sinus glands upon moulting and associated changes in the crayfish. Physiol. Zool. 20: 187-208. SILVERTHORN, S. S., 1975a. Hormonal involvement in thermal acclimation in the fiddler crab Uca pugilator (Bosc). I. Effect of eyestalk extracts on whole animal respiration. Comp. Biochem. Physiol., 50: 281-283. SILVERTHORN, S. S., 1975b. Hormonal involvement in thermal acclimation in the fiddler crab Uca pugilator (Bosc). II. Effects of extracts on tissue respiration. Comp. Biochem. Physiol., 50 : 285-290. SKINNER, D. M., 1962. The structure and metabolism of a crustacean integumentary tissue during a molt cycle. Biol. Bull., 123 : 635-647. VAN HERP, F., AND C. BELLON-HUMBERT, 1978. Setal development and molt prediction in the larvae and adults of the crayfish. Astacus leptodactylus (Nordman, 1842). Aqua- culture, 14: 289-301. VERNBERG, F. J., AND J. D. COSTLOW, JR., 1966. Studies on the variation between tropical and temperature-zone fiddler crabs of the genus Uca. IV. Oxygen consumption of larvae and young crabs reared in the laboratory. Physiol. Zool., 39 : 36-52. VERNBERG, W. B., G. S. MOREIRA, AND J. C. MCNAMARA, 1980. The effect of temperature on the respiratory metabolism of the developmental stages of Pagurus criniticornis (Dana) (Anomura: Paguridae). Mar. Biol. Letters, in press. Reference: Biol. Bull., 159: 700-713. (December, 1980) ON THE FORM AND SIZE OF CRAYFISH LEGS REGENERATED AFTER GRAFTING JAY EDWARD MITTEXTHAL Department of Biological Sciences, Purdue University, II'. Lafayette, Indiana 47907 ABSTRACT The control of form and size in regenerated legs of crayfish was investigated by interchanging basipodites of the cheliped and first walking leg on one side. Ho- motypic, homoeotic, and mosaic legs regenerated. A C( cheliped) -like or W( walk- ing-leg) -like leg usually regenerated with normal proportions, independent of size and sex of the animal and of host site. These factors affected the type and size of a leg. The dimensions of the propo- dite increased allometrically with body length and were larger for chelipeds of males than of females. The C site promoted more rapid growth than the W site. Legs adjacent to a regenerating leg apparently influenced its type and size : C-type tissue growing at either host site tended to inhibit the regeneration of C-type tissue, and to decrease the growth rate of either type of leg, at the adjacent site. In Crustacea a limb bud may be partitioned into exopodite and endopodite com- partments. Atypical development in an exopodite compartment may underlie the observed regeneration of atypical motile exopodites. The majority of mosaic walking legs have the anterior face W-like and the posterior face C-like ; the characteristic boundary between these faces may separate anterior and posterior compartments in the endopodite. INTRODUCTION Serially homologous structures in a segmented animal can be used to analyze pattern formation during development. The five pairs of legs (pereiopods) of decapod crustaceans show deviations from strict serial homology and, in some spe- cies, from bilateral symmetry. This variety of similar but distinctive forms, the ca- pacity for regeneration of legs, and the relatively large size of the animals invite the study of pattern formation through surgical perturbation (reviewed by Needham, 1965). Although the relationships between the time course of regeneration and the molt cycle in crustaceans have been intensively investigated (e.g.. Bliss, 1960 ; Durand, 1960), the available information about development or regeneration in crustacean legs is largely descriptive (e.g., Emmel, 1910). In this study, regeneration after interchange of a proximal segment between stumps of two crayfish legs showed that the host site can influence the size and pos- Received June 15, 1979; accepted September 25, 1980. Abbreviations: C: cheliped (first pereiopod) ; W: walking leg (second pereiopod) ; C-like C: cheliped-like leg regenerated at site normal for C; W-like W: W-like leg regenerated at site normal for W; W-like C: W-like leg regenerated at site normal for C; C-like W: C-like leg regenerated at site normal for W; homotypic leg: leg of type normally at the host site (e.g., C-like C, W-like W) ; homoeotic leg: leg of type normally at a different host site (e.g., W-like C, C-like W). 1 Present address : Department of Biology, University of Oregon, Eugene, Oregon 97403. 700 CRAYFISH LEG REGENERATION: CONTROL 701 sibly the type of regenerated leg. Host sites and adjacent legs seemed to interact in determining the size and type of each regenerated leg. Each type of leg usually had proportions independent of its size and host site. MATERIALS AND METHODS Animals and maintenance Juvenile specimens of the crayfish Procambarus clarkii were obtained from Monterey Bay Hydroculture Farms, Santa Cruz, California, or raised from eggs. Animals were maintained individually after surgery in circulating filtered well water at 18°-22°C. They were fed various combinations of carrots, beef or chicken liver, Tetramin fish food, and a pelletized crayfish food obtained from the supplier. The light : dark cycle was approximately 10 hr light : 14 hr darkness. Grafting Basal segments of the anterior two pairs of legs, the cheliped (C) and the first walking leg (W; Fig. 1), were interchanged on one side of each animal. Opera- tions were performed at all phases of the molt cycle (Stevenson, 1968) later than 2 days after a molt. Animals often died when chilled soon after molting. Grafts transplanted 1 or 2 days before a molt were often lost in molting. The operation was as follows: A crayfish of 20—30 mm body length (tip of ros- trum to caudal margin of telson) was anesthetized by gradual chilling to 0°-5°C in well water. The animal was confined ventral side up with tungsten staples in a groove cut into a slab of Sylgard (Dow-Corning) in a petri dish. The animal was immersed at 0°-5°C in the saline of van Harreveld (1936), pH 7.0, with 10 mM Tris replacing bicarbonate buffer. A C or W was broken off at the basipodite- ischiopodite joint by crushing the meropodite or twisting the ischiopodite ventral- ward (cf. Wales et a/., 1970). It is unclear whether this discarding of distal seg- ments is a reflex (autotomy) or a fracture at a region of weakness (autospasy. Wood and Wood, 1932), particularly in the walking legs. On the C and W the coxopodite-basipodite joint was then transected with iridectomy scissors and the tip of a fine syringe needle (26—28 gauge). Each basipodite was pressed into the host coxopodite in normal orientation. After the operation the animal was re- turned to well water at 0°C and warmed slowly to room temperature. Early in the study the above procedure was not performed in a single operation. Rather, the C and W were broken off, and then the animal was maintained until after the next molt occurred. Each basipodite was then transplanted with its re- generating limb bud. For later animals fracture and interchange were combined in one operation. Both procedures yielded similar legs, which are described here. As a "remove-and-replace" control series, basipodites of the C and W on one side were ablated and then pressed into the coxopodites of the donor legs. A second control series is described in "Results." Muscle in the coxopodite and transplanted basipodite was often opaque a few days after transplantation. Such opacity is presumably a sign of cell death in and around the graft, since healthy muscle is translucent. A transplanted limb bud projecting from the coxopodite was usually opaque. It wras often unclear whether a grafted basipodite had been retained after a scab and opacity obscured the interior of the coxopodite. 702 JAY EDWARD MITTENTHAL A2 A3 B, M D I II Pr FIGURE 1. Cheliped and first walking leg of Procambarus, Leg segments: ischiopodite (I), meropodite (M), carpopodite (Ca), propodite (Pr), dactylopodite (D). The terms anterior, posterior, dorsal, and ventral refer to the position of structures when the leg is fully extended perpendicular to the antero-posterior axis of the body. A : Normal walking leg, posterior (Ai, A4), ventral (A2, A5), and anterior (A:)) views. No tubercles are present; arcs on the pro- podite represent small pits. The rows of large setae are dorsal (left in Ai) and antero-ventral (right in A2, left in An). A second row of smaller setae occurs on the ventral meropodite pos- terior to the first; it is sometimes more prominent than in this animal. A4, A5: Parameters used to characterize size and shape of propodites. LP : propodite length ; L : nianus length ; W : manus width ; T : manus thickness. The measurement of Lp from the ventral view was used only when the index curved markedly in this view. B : Normal cheliped. Because of torsion at the Ca-Pr joint the views corresponding to those for the walking leg are, meropodite : Bi : dorsal, B2 : ventral; propodite: Bi : posterior, B2: anterior. Note the two somewhat irregular rows of large tubercles on the ventral meropodite (B2) and the rows of tubercles on both faces of the propodite. Form and size of legs Legs were observed on animals of body length 55-80 mm. These animals had molted at least 3-9 times after the interchange operation, 3-13 months earlier. Di- mensions of propodites were measured on drawings of the legs made at 3X-12X with a Wild M5 microscope with drawing attachment. RESULTS T\pes of regenerates At the site of the cheliped and walking leg regenerated legs were cheliped-like, walking leg-like, or mosaic. Discrete markers allowed recognition of C-like and W-like regions in mosaic legs (Fig. 1). CRAYFISH LEG REGENERATION: CONTROL 703 Mosaic legs The majority (15/27) of the mosaic legs at the \Y site had the anterior half W-like and the posterior half C-like (Fig. 2). The division between W-like and C-like halves was especially clear on the ventral surface of the meropodite, where an anterior row of W-like setae paralleled a posterior row of C-like tubercles of variable prominence. Occasionally (2/27 animals), one or two tubercles appeared in the anterior row. A few mosaic walking legs appeared YV-like except that tubercles replaced setae in the posterior row. In mosaic W with differing anterior and posterior portions (antero-posterior), the posterior surfaces of the meropodite and carpopodite, which are smooth in nor- mal W, often had low tubercles. The carpopodite was abnormally broad but had W-like anterior rows of setae. Tubercles occur on a normal C, and setae on a normal W, on the propodite and dactylopodite. The propodites of antero-posterior mosaic W had setae on the anterior face and tubercles on the posterior face. The tubercles differed in promi- FIGURE 2. Antero-posterior mosaic walking legs. All four legs shown are from right side of animal. Ai, A», and C are backlighted to show setae, so tubercles appear as dark spots in these pictures. Ai, A2: Typical antero-posterior mosaic W, ventral view. Rows of W-like setae occur on the anterior side of the meropodite and propodite (left side of picture), and rows of C-like tubercles on the posterior side. Curvature of the index toward the anterior side is greater than in a walking leg. B, C : Variant mosaic W. B : Anterior view of proximal segments, ventral view of propodite. The leg resembles those shown in Ai and A« except for a few tubercles on the anterior side of the meropodite and carpopodite (arrows). C: poster ior view. Postero-dorsal aspect is C-like, ventral and anterior aspects W-like. JAY EDWARD MITTENTHAL nence, orderliness of rows, and dorsoventral distribution among animals, but were always restricted to the posterior half of the leg. In the antero-posterior mosaic W the index and dactylopodite often curved more than normally toward the anterior. Posterior C-type cells may proliferate faster than anterior W-type cells to increase distal curvature. The curvature of index and dactylopodite sometimes differed, with larger tubercles proximal to the member with greater curvature. In six mosaic W not of antero-posterior type, numerous tubercles appeared in the anterior half of one or more leg segments. Five of the six legs had two ven- tral rows of tubercles, and the sixth had anterodorsal tubercles, on the meropodite. On other segments tubercles were restricted to a part of the circumference, differing among legs. Patches containing tubercles occurred only in one segment in some legs, but extended across joints in others. No type of mosaic C predominated. The seven mosaic C examined spanned the spectrum for mosaic W. Two of the four mosaic C with differing anterior and posterior halves were C-like anteriorly and W-like posteriorly. The index and dacty- lopodite of these mosaic C curved toward the posterior face, presumably because the anterior C-type surface grew more rapidly. The other two antero-posterior mosaic C approximated large antero-posterior mosaic W, but the boundary between C-like and W-like regions was more variable. Two control series of operations were performed to investigate why mosaic legs regenerate. In the "remove-and-replace" series, 3 of 23 animals regenerated antero-posterior mosaic W ; the remaining regenerates appeared normal for the site. Thus the operation used in the experimental series to interchange basipodites can induce mosaic regeneration ; the interchange of basipodites between C and W sites is not essential. Damage to a basipodite during surgery might make it more vulnerable to in- fluence from the host site, increasing the probability that a mosaic or host-like leg would regenerate. Specifically, inadvertent damage to the anterior proximal margin of the cheliped basipodite might favor regeneration of an antero-posterior mosaic W. To test this possibility, in a second control series the proximal 30% of homolateral C and W basipodites was ablated with a cut parallel to the distal margin of the basipodite. This cut removed both anterior and posterior parts of the proximal margin. The residues of the basipodites were then interchanged. The W site regenerated a mosaic leg in 15 of 18 crayfish; 10 of these legs were antero-posterior mosaic W. Thus damage to the C basipodite often allowed the W site to influence the regenerated leg. However, regeneration of an antero- posterior mosaic W does not require preferential damage to the anterior proximal margin of the C basipodite. Exopodites A small fraction of the regenerated legs were biramous, unlike typical uniramous C and W. These atypical legs had an exopodite attached to the dorsal part of the basipodite at the joint with the ischiopodite (Fig. 3). Each exopodite consisted of a larger proximal segment and a series of smaller distal segments bearing setae. Like the exopodites of the maxillipeds, the atypical exopodites beat intermittently. Exopodites of pereiopocls and maxillipeds usually did not beat in synchrony. How- ever, both seemed to beat at roughly the same frequency — about 1-2 beats/sec at 3°C in a 70 mm male. One animal regenerated exopodites on a mosaic C and on CRAYFISH LEG REGENERATION": CONTROL 705 FIGURE 3. Exopodites of regenerated periopods. A: cheliped; B: mosaic cheliped; C, D: walking leg-like cheliped. A, B are anterior views ; C, D are posterior views. In D the tip of the expedite is blurred because the exopodite was beating. the adjacent C-like W ; these exopodites could beat independently of each other and of the exopodites of the maxillipeds. Exopodites were also observed on two W-like C's. Regeneration of an ex- opodite does not occur only on homeotic or mosaic legs, however. In the "remove- and-replace" control series, 2 of 23 animals regenerated exopodites on C-like C. Surgery is not necessary, and regeneration of a leg may not be necessary, to pro- duce a post-embryonic exopodite : A cheliped having a motile exopodite occurred in one unoperated 29 mm male. Examination of the most immature animals avail- able, four preserved normal specimens of Procambarus in the second stage after hatching, showed no exopodites on the pereiopods. Motility of regenerated legs Most regenerated legs were used in the manner typical for an unoperated leg at the host site. A crayfish walking on a level substrate held a leg at the C site off the substrate, while a leg at the W site touched the substrate periodically in the stepping rhythm. (One exceptional animal held a C-like W off the substrate dur- ing walking. This leg was lost at the next molt, and neither C site nor W site sub- sequently regenerated a leg.) Typically the regenerated legs showed normal clos- ing and defense reflexes involving the dactylopodite. These observations suggest JAY EDWARD MITTENTHAL TABLE I Ratios of dimensions of propodites; mean -\- s.d. (number of legs). "Control" legs are from un- operated animals measured as obtained from the supplier. "Normal" legs are unoperated legs from operated animals. Manus length ranged from 2.2 to 15.2 mm in chelipeds, and from 2.7 to 6.7 mm in walking legs, used to estimate these ratios. Propodite length Manus width Manus thickness Manus length Manus length Manus length Control C Normal C C-like C C-like W 2.63 ± 0.21 (13) 2.54 ± 0.18 (21) 2.61 ± 0.22 (9) 2. 73 ±0.10 (3) 0.79 ± 0.11 (13) 0.81 ± 0.10 (21) 0.76 ± 0.09 (9) 0.76 ± 0.03 (3) 0.56 ± 0.07 (12) 0.55 ± 0.06 (20) 0.51 ± 0.08 (9) 0.47 ± 0.01 (3) Control W Normal W W-like W W-like C 2.04 ± 0.08 (11) 2.13 ± 0.11 (24) 2. 14 ±0.18 (6) 2.00 ± 0.21 (10) 0.48 ± 0.05 (11) 0.53 ± 0.05 (25) 0.53 ± 0.05 (6) 0.48 ± 0.06 (10) 0.33 ± 0.03 (10) 0.34 ± 0.03 (23) 0.34 ± 0.02 (6) 0.35 ± 0.04 (9) that the regenerated leg usually made functionally adequate neural connections with the ganglion at the host site. Size of the legs Propodite length, manus length, manus width, and manus thickness (Fig. 1) were used to compare normal and regenerated legs. Sixteen unoperated animals and 21 animals bearing legs regenerated after grafting were measured. C-like legs or W-like legs maintained constant proportions, regardless of site or time of growth (Table I ; but note thickness/length for manus of C-like legs). C-like and W-like legs showed distinct patterns of proportional growth ; the propodite length, manus width, and manus thickness increased more rapidly, relative to manus length, in C-like than in W-like legs. Although walking legs of unoperated males and fe- males grew at the same rate, sexual dimorphism occurred in chelipeds : In unoper- ated animals the propodite of the cheliped was about 1.25 times as long in males as in females (Fig. 4). C-like, W-like, and mosaic regenerates were larger at the C site than at the W site (Table II). The regenerated legs were intermediate in size between the con- tralateral unoperated C and W. Frequency distribution of types of regenerates At the C site and at the W site three types of legs regenerated : C-like, W-like, and mosaic. Thus nine pairings of types might have regenerated at the two sites. Table III summarizes the prevalence of the types, singly and in pairs, in 47 ani- mals. At both sites the majority of regenerated legs were not of graft type, but were host-like or mosaic. Also, at both sites regeneration of heteromorphic legs was biased toward smaller, less specialized types — at the C site, W-like C rather than mosaic C ; at the W site, mosaic W rather than C-like W. Three of the nine pairings of leg types occurred in about two-thirds of the ani- mals. In the most common pairing (15/47), the two more common types of heteromorphic legs, W-like C and mosaic W, occurred together. In 16/47 animals a C-like C accompanied a mosaic W or W-like W. Of the remaining six pairings CRAYFISH LEG REGENERATION': CONTROL 707 B J I L 40 5O 60 70 80 100 Body Length, mm FIGURE 4. Propodite length vs. body length. Circle, chelipid ; triangle, walking leg ; blackened symbol, male ; open symbol, female. A. Unoperated animals, measured as obtained from the supplier. Lines are drawn by eye. B. Normal C and W of operated animals. Lines are from part A, for unoperated animals. Arrows indicate legs for which mass of closer muscle was measured ; see Mittenthal et al., this volume. of leg types, the most common was W-like C with C-like W. This pairing regen- erated in less than 10% of the animals. Table III includes estimates of the number of animals expected to bear each of the nine pairings if legs regenerated independently at the C site and W site. Of the six pairings in which the two regenerated legs were of different types, four oc- curred more often than expected from independent regeneration. All three of the pairings having both legs of the same type occurred less often than expected. The probability that the two legs regenerated independently is less than 0.16. As might be expected from the prevalence of mosaic W, legs at the C site were paired with mosaic W at least as often with W-like W or C-like W. It was there- fore interesting to ask whether each of the three types of regenerates at the C site was preferentially paired with antero-posterior mosaic W or with other mosaic W. Table IV shows that the larger was the type of leg regenerated at the C site, the greater the probability of its being paired with an antero-posterior mosaic W. DISCUSSION In these experiments homolateral basipodites of the cheliped and first walking leg of crayfish were interchanged. If each graft had retained its initial state of JAY EDWARD MITTENTHAL w m ^J 13 t ^_^ d d* * # _; A A V V V to f/1 t/) 0 oj — O -f C ON O O 0 0 O O 0 d to c-5 •* -H -H -H-H-H-H NO -LI 5 --1 ON -H 10 t^ -H 00 '— ' ,^5 •^ C 01 ro ^O C1 to OO ^ d oi d d '•O t*** OO '•'O 00 d — •^ O 01 O 0 s c/) 3 •£ o o o o o o OJ ° •M C rt IH ij _M -U -11 -LI -LI Tl Tl Tl Tl <» -H o s ^ OO rf> *-* t~~ NO t^ o -+ u 0 -H r^ <*O to -^ OO -o 1— l' rt —' d — i d d 00, _ O -H — 1 O | •£ O O -H -H 0 0 0 O 4-1 4J JJ -LJ Tl Tl Tl Tl od o a) to \o O ON OO OO d ^ (i ~~ 0 — < <~^ ^ *& to ON •~ o — i O d 'o 15 O _£ vO 00 -H -H ro ro *-H t^ ^ 'E 03 ^^ J> ^ ^ U ^ U uu §1 o := U CT3 03 E E u. t-, 0 O ctf rt cd rt E E E E U U Id IH O O O O rt rt E E o o _E C C c c c c c c o 0) a TJ £ ^ ^ u ^ ^ U U > 2 _D o y y u P* LJ II .-^ r^ cti rt cd rt CO CO (0 CO ^ J< bfl 1 1 ^O 0 O 0 1 1 OJ SSSS U U CRAYFISH LEG REGENERATION' : CONTROL TABLE III Number of animals with various types of regenerated legs in cheliped and walking leg coxopodites. For each entry, the number on the left is observed. The number on the right, in parentheses, is calcu- lated on the hypothesis that the types of legs regenerated in cheliped and walking leg coxopodites are independent. The probability that the number of mosaic C and W-like C is ^6 is 0.154 on this hypothesis, by Fisher's exact test (Blalock, I960). For this test the entries of Table I were pooled by using leg types C-like C, W-like W, other C, other W. In ^^\^ In cheliped walking ^\roxopodite leg coxopodite^\. C-like C Mosaic C \V-like C Number of animals with given type leg in walking leg coxopodite W-like W 7 (4.98) 3 (1.94) 3 (6.09) 13 Mosaic W 9 (10.34) 3 (4.02) 15 (12.64) 27 C-like W 2 (2.68) 1 (1.04) 4 (3.28) 7 Number of animals with given type leg in cheliped coxopodite 18 7 22 47 determination, and if no grafts were lost, each operated animal would have re- generated two homoeotic legs, a walking leg-like cheliped and a cheliped-like walk- ing leg. However, in addition to homoeotic legs, host-type and mosaic legs re- generated. Similarly, Edwards and Sahota (1967) found homoeotic, mosaic, and host-type regenerates in crickets after they grafted the bud of a regenerating cercus to the stump of a mesothoracic leg. Mosaic appendages may have originated from the grafts through a combination of cell death, invasion of the graft by host tissue, and transdetermination of graft tissue. These processes or loss of grafts could have produced host-type regenerates. The results of this experiment show that the proportions of the leg are inde- pendent of its size and are determined with its type (rf. Prange, 1977). However, other distinctive correlates of leg type, the surface features of the cuticle, do change with size in decapod Crustacea. As a crayfish grows, the number and prominence of tubercles on the cheliped increase. Factor (1978) has shown that the distribution of tubercles and setae on the perioral appendages of the lobster Homarus changes during the molts between larval and postlarval stages. The host site for regeneration of a leg affects its growth rate ; the cheliped site promotes more rapid growth than the walking leg cite. The use of a leg may also affect its growth rate. The normal cheliped of operated animals tended to be TABLE IV Number of animals having two types of mosaic W paired with various types of limbs at the C site. A-p = antero- posterior. For this table, mosaic W with any tubercles on the anterior face were in- cluded in "not a- p." Type of \Type of leg C-like C Mosaic C W-like C mosaic W\at C site a-p 6 2 6 not a-p 1 1 5 JAY EDWARD MITTENTHAL larger than a cheliped of an unoperated animal of the same size. The normal cheliped may be used more, and may hypertrophy, after the operation. In decapod crustaceans a cheliped can influence the determination of the con- tralateral cheliped. The Alphcidae (snapping shrimp) have asymmetric chelipeds, a large specialized snapper and a smaller, less specialized pincer. After autotomy of the snapper, the pincer is transformed to a snapper, and a pincer replaces the lost snapper (Przibram, 1931; Darby, 1934). To explain reversal of asymmetry, it has been postulated that the snapper suppresses the transformation of the pincer to a snapper (Wilson, 1903; Darby, 1934; Mellon and Stephens, 1978). The present results suggest that in crayfish the cheliped and first walking leg are analogous to the snapper and the pincer, respectively, in interactions between legs. Cheliped-type (C-type) epidermis in either leg apparently tends to prevent regeneration of C-type tissue in the other leg. Pairing of a walking leg-like (W-like) regenerate at one host site and a C-type or mosaic regenerate at the other site was prevalent ; pairing of two C-like legs was rare. C-like tissue may also decrease the growth rate of legs at adjacent sites. (See Huxley, 1972, for an alternative hypothesis on growth modulation.) A regenerated C-like C tends to be smaller than the contralateral normal C. However, a W-like W tends to be larger than the contralateral unoperated W, perhaps because a small C-site regenerate exerts subnormal inhibition on growth of the W-site regenerate. This hypothesis predicts a greater enlargement of the W-site regenerate, the less C-like is the regenerated leg at the C site, as was observed. If C-type tissue reduces growth of adjacent legs, then contralateral unoperated legs, growing under reduced inhibition from smaller regenerating legs, should tend to be larger than homologous legs of unoperated animals. This is so : A propodite of an unoperated C or W tends to be longer in an operated animal than in a comparable unoperated animal (Figure 4B), especially for male chelipeds. Thus in Procainbarus chelipeds may inhibit the growth of adjacent legs. C-type tissue in a cheliped or mosaic leg may suppress formation of C-type tissue in adja- cent legs. It remains to be seen whether suppression of growth and of C-typc determination are causally related. The majority of mosaic walking legs were W-like on the anterior half, adjacent to the C site, and C-like on the posterior half. The prevalence of this pattern sug- gests that the proposed suppression of C-type development by adjacent C-type tis- sue is spatially graded, decreasing with distance from the C-type tissue. Polariza- tion of a mosaic W toward this pattern should then be more frequent, the more C-like is the regenerate at the C site. The results in Table IV agree with this prediction. In the antero-posterior mosaic W the boundary between W-type and C-type epidermis is particularly clear on the ventral surface of the meropodite. There an anterior row of W-like setae parallels a posterior row of C-like tubercles. It is difficult to believe that graded inhibition from the C site falls below a threshold for developing C-like tissue exactly between the rows of protuberances in most mosaic W. The phenomenon of compartmentalization in Drosopliila offers a model for interpreting a characteristic W-C boundary in an antero-posterior mosaic W. In Drosophila determination of an imaginal disc proceeds through a sequence of com- partmentalization events (Crick and Lawrence, 1975; Garcia-Bellido, 1975). In each event boundaries are established at characteristic positions in the disc, de- limiting new compartments. The progeny of cells within a compartment do not CRAYFISH LEG REGENERATION: CONTROL 711 cross its boundaries, although cells may cross boundary lines not yet established. Mutants are known that change the state of determination in particular com- partments, apparently leaving the rest of the disc unaltered. During regeneration of a disc, boundaries previously set can be crossed by clones, and compartmentaliza- tion proceeds anew as the disc develops (Szabad ct al., 1979). The Minute (M) mutant has been used to demonstrate compartmentalization. X-irradiation of a Drosophila larva heterozygous for Minute may induce formation of an M+/M+ cell by somatic crossing-over. This cell proliferates faster than sur- rounding M/M+ cells. The M+/M+ patch, revealed by genetic markers, often grows fast enough to fill most of a compartment, so that the boundaries of patch and com- partment coincide extensively (Morata and Ripoll, 1975). Partition of a developing leg into anterior and posterior compartments may oc- cur in crayfish, as in Drosophila (Tokunaga, 1962; Steiner, 1976). The compart- ment boundary might be visible in antero-posterior mosaic W because a patch of C-type cells in the posterior compartment grew faster than surrounding W-type cells and so occupied most of the compartment. Anterior and posterior compart- ments may also occur in appendages of cockroaches (French, 1976) and crickets (Edwards and Sahota, 1967). The segments of a crayfish leg distal to the basipodite represent one ramus, the endopodite, of an appendage that was biramous during ontogeny. Normally asta- curans shed the other ramus, the exopodite, of each pereiopod at the molt from the last mysis stage to the first post-larval stage (Anderson, 1973). In the present study most regenerated legs lacked an exopodite ; some, however, had one. The observation that an appendage of Procambarus, or of Asellus (Needham, 1950), may regenerate either in uniramous or biramous form suggests that the two rami develop from separate compartments. Presence or absence of a ramus could then correspond to two interpretations of the positional information in a compartment. Atavistic regeneration, in which a phylogenetically earlier type of ramus replaces a lost ramus (e.g. Schultz, 1905; reviewed by Needham, 1965), could represent yet another interpretation of the same positional information. It may seem remarkable that a motile appendage can regenerate after the normal appendage of the same type was lost several molts previously. However, Gurney (1942, pp. 106-107) mentions several decapods in which an appendage is lost at one stage of development but later reappears in functional form. In some Crustacea (Bittner, 1973) axotomy of a mature motoneuron often does not kill the neuron, but initiates changes that culminate in regeneration of its axon. Davis and Davis (1973) have found that in Houiarns the motoneurons innervating pleopod muscles appear to develop normally, at least to the sixth post-hatching stage, after their target muscles are destroyed at hatching. Perhaps the motoneurons innervating exopodites of pereiopods also survive the normal loss of their target muscles for several stages and can reinnervate a regenerated exopodite. ACKNOWLEDGMENTS I thank Dr. Arthur Winfree and Ms. Vickery Trinkaus-Randall for a critical reading of an early draft of the manuscript, and Ms. Glenna D. Cummings for aid with photography. This research was supported in part by NSF grant 77-24149. JAY EDWARD MITTENTHAL LITERATURE CITED \XDERSON, D. T., 1973. Embryology and Phytogeny in Annelids and Arthropods. Pergamon Press, New York, xiv, 495 pp. BITTNER, G. D., 1973. Degeneration and regeneration in crustacean neuromuscular systems. Am. Zoo/., 13 : 379-408. BLALOCK, H. M., JR., 1960. Social Statistics. McGraw-Hill Book Co., Inc., New York, xiv, 465 pp. BLISS, D. E., 1960. Autotomy and regeneration. Pp. 561-589 in T. H. Waterman, Ed., Physi- ology of Crustacea. Academic Press, New York, Vol. I. CRICK, F. H. C., AND P. A. LAWRENCE, 1975. Compartments and polyclones in insect develop- ment. Science, 189 : 340-347. DARBY, H. H., 1934. The mechanism of asymmetry in the Alphcidac. Carnegie I nst. U'ash. Publ, 435 : 357-361. DAVIS, W. J., AND K. B. DAVIS, 1973. Ontogeny of a simple locomotor system : Role of the periphery in the development of central nervous circuitry. Am. Zoo!., 13: 409-425. DURAND, J. B., 1960. Limb regeneration and endocrine activity in the crayfish. Biol. Bull.. 118: 250-261. EDWARDS, J. S., AND T. S. SAHOTA, 1967. Regeneration of a sensory system : The formation of central connections by normal and transplanted cerci of the house cricket Achcta domesticus. J. Exp. Zoo/., 166: 387-396. EMMEL, V. E., 1910. A study of the differentiation of tissues in the regenerating crustacean limb. Am. J. Anat. 10: 109-159. FACTOR, J. R., 1978. Morphology of the mouthparts of larval lobsters, Homarus amcricanus (Decapoda: Nephropidae), with special emphasis on their setae. Biol. Bull., 154: 383-408. FRENCH, V., 1976. Leg regeneration in cockroach, Blatclla germanica. II. Regeneration from a noncongruent tibial graft-host junction. /. Embryo!. E.\-p. Morphol., 35: 267-301. GARciA-BELLioo, A., 1975. Genetic control of wing disc development in Drosophila. Pp. 161- 182 in Cell Patterning (CIBA Foundation Symposium 29, New series). Associated Scientific Publishers, New York. GURNEY, R., 1942. Larvae of decapod Crustacea. Ray Society, London, vi, 306 pp. HUXLEY, J., 1972. Problems of Relative Grozvth. Second Edition, Dover Publications, Inc., New York, xxvii, 312 pp. MELLON, D., JR., AND P. J. STEPHENS, 1978. Limb morphology and function are transformed by contralateral nerve section in snapping shrimps. Nature, 272 : 246-248. MORATA, G., AND P. RipOLL, 1975. Minutes : Mutants of Drosophila autonomously affecting cell division rate. Dcv. Biol., 42 : 211-221. MORRISON, D. F., 1967. Multivariatc Statistical Methods. McGraw-Hill Book Co., New York, xiii, 338 pp. NEEDHAM, A. E., 1950. Determination of the form of regenerating limbs in Ascllus aquaticus. Q. J. Microsc. Sci., 91 : 401-418. NEEDHAM, A. E., 1965. Regeneration in the arthropods and its endocrine control. Pp. 283- 323 in V. Kiortsis and H. A. L. Trampusch, Eds., Regeneration of Animals and Re- lated Problems. North Holland Publishing Co., Amsterdam. PRANCE, H. D., 1977. The scaling and mechanics of arthropod exoskeletons. Pp. 169-181 in T. J. Pedley, Ed., Scale Effects in Animal Locomotion. Academic Press, New York, xx, 545 pp. PRZIBRAM, H., 1931. Connecting Laivs in Animal Morphology. University of London Press, Ltd., London, 62 pp. SCHULTZ, E., 1905. Ueber atavistische regeneration bei Flusskrebsen. l-Vilhclm Roux' Arch. Entwichlungsmech. Org., 20 : 38-47. STEINER, E., 1976. Establishment of compartments in developing leg imaginal discs of Droso- phila mclanogastcr. H'ilhclm Roux' Arch. Entu'ichlungsiuccli. Org. 180: 9-30. STEVENSON, J. R., 1968. Metecdysial molt staging and changes in the cuticle in the crayfish Orconectes sanborni (Faxon). Crnstaccana, 14: 169-177. SZABAD, J., P. SIMPSON, AND R. NOTHIGER, 1979. Regeneration and compartments in Drosophila. J. Embryol. Exp. Morph., 49: 229-241. TOKUNAGA, C., 1962. Cell lineage and differentiation on the male foreleg of Drosophila melanogaster. Dei: Biol. 4: 489-516. VAN HARREVELD, A., 1936. A physiological solution for fresh-water crustaceans. Proc. Soc. Exp. Biol. and Mcd., 34 : 428-432. CRAYFISH LEG REGENERATION: CONTROL ~U WALES, W., F. CI.ARAC, M. DANDO, AND M. LAVKKACK, 1970. Innervation of the receptor- present at the various joints of the pereiopods and third maxilliped of Hniimnts //<;;;/- nuinis (L.) and other niacruran decapods (Crustacea). 7.. l'i'rt/1. f'hysinl.. 68: 345 384. WILSON, E. B., 1903. Xotes on the reversal of asymmetry in the regeneration of the chelae in Alphcus hctcrochcHs. Biol. Bull.. 4: 197-210. WOOD, F. D., AND H. F. WOOD. II., 1932. Autotomy in decapod crustacca. ./. /:.r/>. /.mil. 62: 1-55. Reference: Biol. Bull., 159: 714-727. (December, 1980) .MORPHOLOGY OF THE CLOSER MUSCLES IN NORMAL AND HOMOEOTIC LEGS OF CRAYFISH1 JAY EDWARD MITTENTHAL,- MARY C OLSON, AND GLENNA D. CUMMINGS Department of Biological Sciences. Purdue University, West Lafayette, Indiana 47907 ABSTRACT Regeneration of homoeotic legs in crayfish was evoked by grafting a proximal leg segment to an ectopic host site, in order to investigate the relative roles of the host site and the donor leg type in determining morphological characteristics of a leg muscle, the dactylopodite closer. To provide a frame of reference for results from closers of homoeotic chelipeds and first walking legs, closers from the corresponding legs of unoperated animals were also examined. The mass, fiber length, and fiber diameter of normal and regenerated closers increase in proportion to appropriate external dimensions of a leg : Mass is pro- portional to a measure of the volume of the manus, where the muscle occurs. Mean fiber length for the region of the closer sampled is proportional to the maximum thickness of the manus. The square of mean fiber diameter for this region is proportional to a measure of the area over which fibers attach. The latter result suggests that the number of fibers in a closer muscle is nearly constant during growth. The spectrum of sarcomere lengths in the sampled region is bimodal in most muscles, though fibers with short sarcomeres were absent from some samples. The length and fraction of long sarcomeres increase more slowly, but over a longer period, in chelipeds than in walking legs. In chelipeds fibers elongated by addition of sarcomeres as well as by elongation of sarcomeres. The closer muscles of homoeotic legs resemble the closer muscles of normal donor-type legs of the same size. A walking leg regenerated from the coxopodite of a cheliped grows larger in external dimensions and in closer muscle mass, fiber length, and fiber diameter than a normal walking leg. A cheliped regenerated from the coxopodite of a walking leg is smaller than a normal cheliped in these param- eters. Thus the host site modulates the growth rate of the homoeotic leg, externally and internally, toward the growth rate of the leg normally present there. INTRODUCTION Muscle fibers develop through the interaction of intrinsic properties of myoblasts with extrinsic agents, including motoneurons and attachment surfaces (Finlayson, 1975). In a homoeotic appendage muscle fibers attach to surfaces in the abnormal appendage but are subject to the remote controls (motoneurons, blood supply) Received June 15, 1979; accepted September 25, 1980. Abbreviations: C, cheliped; W, walking leg; K, mean ratio of fiber thickness to fiber diameter; s \,,nK, mean length of long sarcomeres. 1 This paper is dedicated to the memory of Fred Lang, whose untimely death impoverishes the science and the lives of those who were his friends and colleagues. 2 Present address : Department of Biology, University of Oregon, Eugene, Oregon 97403. 714 CRAYFISH I.K<; REGENERATION: MTSCLK 715 normal for a different appendage. Therefore the traits of muscles in a homoeotic appendage should suggest what contributions attachment surfaces and remote controls make to muscle properties. In arthropods muscle fibers attach to the epidermis, which also controls external form. Thus homoeosis of external form must modify the influence of attachment surfaces on muscle fibers. Motoneurons may influence muscle properties in Crustacea (Atwood, 1973), as in vertebrates (Close, 1972). Therefore homoeotic regenerated legs of crayfish (Mittenthal, 1980) allow analysis of extrinsic modula- tion of muscle development. Wiersma (1955) found that the closer muscles of the cheliped and first walking leg in the lobster Howarits rulgaris differed in the number of impulses required to elicit noticeable closing of the dactylopodite. The present work shows that the homologous muscles of crayfish differ in the distributions of fiber length, fiber diameter, and sarcomere length. For a given type and size of leg these distributions are similar, whether the leg develops at the cheliped site or the walking leg site. Thus in the homoeotic legs the determinants of leg type carried with the graft dictate the morphology of muscle fibers that regenerate. However, the host site alters the size of the leg and of its muscle fibers, keeping the proportions of muscles to leg nearly normal. MATERIALS AND METHODS Transplantation Crayfish were cultured and operated as described by Mittenthal (1980). For legs from which closer muscles were assayed, a basipodite bearing a small limb bud (0-2 mm) was grafted. Therefore the donor site may have influenced the regenerated muscles. However, the limb bud typically became necrotic after the interchange operation (Mittenthal, 1980). Extensive de- and re-differentiation of structures in the limb bud, including muscles, probably occurred after operation. Selection of legs for analysis Two criteria were used to select regenerated legs for analysis of closer muscles. The leg resembled a normal cheliped or walking leg in its qualitative proportions and discrete surface markers ; and the leg was used in a manner typical for the host site (spontaneous behavior, tactile closing reflex). In four animals the ventral nerve cord and its roots were exposed when legs were removed, to investigate innervation of the regenerated legs. In all these cases visual inspection showed nerves joining the ganglion of the host segment, and not of the donor segment, to the regenerate. Transection of the connectives on either side of the host segment ganglion did not abolish the closing reflex, but cutting the nerves from this ganglion to the leg abolished the motility of the leg, in the two animals thus tested. Histology Closer muscles were fixed according to a procedure modified from that of Lang et al. (1977). After autospasy the leg was immersed at 0°C in crayfish saline (van Harreveld, 1936) with 10 mM Tris replacing bicarbonate, pH 7.0 at 25°C. Ca2+ was reduced to 0.25 of normal (to 2.4 mM) ; this greatly weakens or abolishes contraction of crayfish muscle in response to nerve stimulation (Mittenthal, unpublished observation). MITTENTHAL, OLSON, AND CUMMINGS The carpopodite-propodite joint was transected. The propodite and dactylo- podite were firmly wired to a stainless steel screen with the dactylopodite in the maximally open position. The propodite was perfused through a syringe needle inserted near the index-manus junction; the perfusate filled the up-tipped propodite and flowed from its proximal end. After perfusion for 10 min with 1/4 Ca saline at room temperature, the perfusate was changed continuously over 10-15 min to formaldehyde in 1/4 Ca saline, pH 7.2. A smoothly increasing concentration of formaldehyde was mixed in an enlarged version of a linear gradient maker (Blattner and Abelson, 1966). The propodite was immersed in. and perfused with, the 109^ formaldehyde saline for at least 2 hr. The closer muscle was then exposed by scraping away exoskeleton on the anterior face of the walking leg or the homologous face of the cheliped. Only fibers from this side of the pinnate closer muscle were subsequently analyzed. The muscle was resubmerged in lO^c formaldehyde saline at 5°C for 48 hr, then removed from the propodite in 1/4 Ca saline, superficially dried by light blotting, weighed in 1/4 Ca saline, and stored in 70% ethanol. Fibers were sampled from the distal dorsal region of the closer muscle (distal 1/2, dorsal 1/3). This region and its subdivisions are defined on the apodeme, not at the superficial attachment surface of the muscle. With electrolytically sharpened tungsten needles an equal number of fibers (from 4 in some muscles to 10 in others) were teased from each of the six subdivisions shown in Table III. The length and diameter of each fiber were measured with an ocular micrometer in a binocular microscope. The fibers are roughly rectangular in cross section ; fiber diameter was estimated as the average of the maximum and minimum widths of a fiber lying on its broader side. Sarcomere lengths were measured from phase contrast photomicrographs of fibers (walking leg) or of bundles of myofibrils freed by teasing (cheliped). Three regions of each fiber were photographed ; in each region a sarcomere length was estimated from the length of 5-15 (typically 10) sarcomeres in series in a myo- fibril. If the three estimates of sarcomere length thus obtained did not agree within 25% of the minimum estimate, data for that fiber were not used. Franzini- Armstrong ( 1970) found a maximum variation of 25% in sarcomere lengths within single fibers of a limb muscle in the crab Port units. Only one fiber of more than 750 fibers measured failed to meet this criterion. This uniformity of sarcomere length suggests that localized contraction in parts of fibers during fixation was rare. RESULTS Cheliped-like (C-like) or walking-leg-like (W-like) legs which regenerated at the site of a normal cheliped (C-site) will be called C-like C and W-like C, respectively. Similarly, C-like W and W-like W signify regenerated legs at the site m.Tmal for a walking leg. Normal C and W are unoperated. Closer muscles Fibers of the closer muscle attached to a similar zone of epidermis in all legs of a given type (C-like or W-like). A closer muscle grows approximately in proportion to its pmpodite (Fig. 1A) ; the closer in C-like legs occupies about 40%-, and in \Y-like legs about 20%, of the volume of a rectangular parallelepiped CRAYFISH LEG REGENERATION : MUSCLE 717 1000 5 10 20 50 100 200 500 1000 Manus Length x Manus Width x Manus Thickness (mm3) 40 60 100 Body Length (mm) FIGURE 1 : A. Mass of closer muscle vs. volume of a rectangular parallelepiped enclosing the manus. Filled circle: normal C and C-like C; ring around filled circle: C-like W; x: normal W and W-like W ; ring around x : W-like C. T over a symbol designates normal limb of an operated animal. No difference between males and females was evident. Lines were fitted to data for unoperated animals with a nonlinear regression program (Dixon, 1975) for y = axb. For 11 chelipeds, a == 2.9±6.1, b = 0.69±0.31 ; for 5 walking legs, a = 0.21±0.21, b = 0.91 ±0.27. If muscle mass is proportional to the volume of the parallelepiped, b = 1. The dashed line represents y = axb with b = 1. B. Mass of closer muscle vs. body length for chelipeds. Filled circle : male ; open circle : female. Three parallel lines have been drawn by eye through points for males, females, and normal chelipeds of operated males. that contains the manus. The closer of a normal C is 40-100 times more massive than the closer of the normal W in the same animal. The host site alters the size of a homoeotic leg: W-like C are slightly larger than normal W, and C-like W are smaller than normal C (Mittenthal, 1980). The relations between closer mass and manus volume for normal legs roughly predict the closer mass for homoeotic legs. However, the ectopic host site affects closer mass slightly more than manus volume ; two W-like C have larger closer mass, and three C-like W have smaller mass, than expected for normal propodites of the same size. Because chelipeds of different sizes are geometrically similar, cheliped closer mass should be proportional to the cube of propodite length. The propodite of a normal cheliped is about 1.25 times longer in unoperated males, and as much as 1.7 times longer in operated males, than in unoperated females of the same body length (Mittenthal, 1980). Muscle mass might therefore be 1.95 times larger in unoperated males, and as much as 4.9 times larger in operated males, than in unoperated females. A plot of muscle mass vs. body length (Fig. IB) showed measured factors of increase about 1.75 and 7.3. Thus, after the grafting operation 718 MITTENTHAL, OLSON, AND CUMMINGS 10 9 I8 ^ 7 c* 6 o> 1 0 01 23456789 Manus Thickness (mm) B ll -J— i 10 40 50 60 70 80 100 Body Length, mm FIGURE 2 : Mean±s.d. of length of fibers from distal dorsal region. A. Fiber length vs. manus thickness. Symbols as in Figure 1A. The line has slope 1 and passes through the origin. B. Fiber length vs. body length for chelipeds. Symbols as is Figure IB. Straight line is drawn by eye through data for unoperated animals. the unoperated cheliped grew more rapidly in closer mass than in external size, relative to the cheliped of an unoperated animal. Dimensions of fibers The length and diameter of a muscle fiber allow estimation of its volume and cross-sectional area. These parameters of fiber size are correlated with the size of the leg. In the closer, a pinnate muscle, the fibers extend from a central apodeme to opposite surfaces of the leg. Fiber length should increase with the distance between these surfaces — that is, with manus thickness. The mean length of fibers from the distal dorsal region is roughly equal to the manus thickness, in C-like and W-like legs at either site (Fig. 2A). If the cross-sectional area of each fiber increases in proportion to the area of the apodeme, and if the number of fibers remains constant during growth, then fiber cross-sectional area should be proportional to the product of manus width and manus thickness. If a cross-section of each fiber retains constant proportions as the fiber grows, then its area is proportional to the square of fiber diameter. As these assumptions predict, fiber diameter is proportional to (manus length X manus width) - in C-like and W-like legs at both sites (Fig. 3A). CRAYFISH LEG REGENERATION : MUSCLE 719 These results allow estimation of the ratio of the number of closer muscle fibers in a cheliped and in a walking leg. Approximately. muscle mass / number \ \of fibers \ /mean fiherX /mean fiberY' / \ length / \ diameter / manus envelope volume manus (length -width) -manus thickness K is the mean ratio of fiber thickness to fiber diameter. In the distal dorsal region of the closer, two ratios are the same in C and \V : the ratio of mean fiber length to manus thickness, and the ratio of the square of mean fiber diameter to the product of manus length and width. If these ratios and K were known for an entire muscle, the number of fibers in the muscle could be estimated. To estimate the ratio of the fiber numbers in C and W, we assume that the above two ratios and K are the same in C as in W if fibers from the entire closer muscle are sampled. Then the ratio of the number of closer muscle fibers in C and in W is the same as the ratio, for C— W, of muscle mass to manus envelope volume. Since muscle mass divided by manus envelope volume equals approximately 0.4 in C and 0.2 in W, a cheliped closer has about twice as many muscle fibers as a walking-leg closer. Bittner and Traut (1978) showed by counting fibers that a cheliped opener muscle also has about twice as many muscle fibers as a walking-leg opener (265/135 = 1.96). The opener and closer might both have twice as many fibers in a C as in a W if the serially homologous apodemes were twice as large in a C as in a W at the time muscle fibers formed, and if newly formed fibers had the same diameter spectra in homologous muscles. The relations between fiber size and manus size are slightly abnormal in operated animals. In C-like W the mean fiber diameter is disproportionately small (Fig. 3A). Although the mean fiber lengths in C-like W are normal relative to 600 500 ^ 400 E 300 a. ~ 200 E 100 80 60 40 20 A 10 20 50 100 200 Manus Length x Manus Width (mm2) \ i i 60 100 Body Length (mm) FIGURE 3 : A. Mean±s.d. of diameter of fibers from distal dorsal region vs. manus length X manus width. Symbols as in Figure 1A. The line, fitted by nonlinear regression (cf. Fig. 1) has a = 22.3 ±1.6, b = 0.50±0.01, for 48 fibers. B. Fiber diameter vs. body length for chelipeds. Symbols as in Figure IB. MITTENTHAL, OLSON, AND CUMMINGS TABLE I Fiber diameter (urn) in several legs of an animal; mean ± s.d. (number of fibers). Values separated by asterisks correspond to histograms which differ significantly according to the Kolmogorov-Smirnoff test (Blalock, 1960) with P = 0.01; width of a histogram column (bin) = 20 pm. Leg type 62 mm Female 62 mm Male 76 mm Female 70 mm Female Normal C 123 ± 30 227 ± 60 281 ± 58 * C-like C 105 ± 23 * * * C-like W 71 ± 19 W-like C 90 ±35 77 ± 23 \V-like W 33 ± 12 Normal W 60 ± 22 67 ± 24 58 ± 17 41 ± 11 manus thickness, the manus thickness is abnormally low (Mittenthal, 1980). In \\-like C the closer mass is disproportionately large ; scatter in the data may conceal a corresponding enlargement of fibers. The mean fiber length and diameter, as well as the closer mass, are disproportionately large in the unoperated cheliped of an operated animal (Figs. 1B-3B). The ectopic host site modulates the growth of a homoeotic leg. In a given crayfish C-like and W-like regenerated legs tend to have mean fiber diameters and lengths intermediate between those of the contralateral unoperated C and W (Tables I and II ). However, a homoeotic leg deviates further from the unoperated donor- type leg than does a donor-type leg regeneraed at the donor site. In summary, the foreign host site for a homoeotic leg biases the diameter and length of closer fibers away from donor-like values and toward host-like values. Surgery and regeneration alone do not suffice to produce this bias. The host site biases muscle morphology and external dimensions together, in such a way that muscle parameters remain related to measures of external size in homoeotic legs as in unoperated legs. The changes in muscle parameters slightly exceed those in external size. TABLE II Fiber length (mm) in several legs of an animal; mean ± s.d. (number of fibers) ; * : as in Table I; bin width = 50 \j.m. Leg type 62 mm Female 62 mm Male 76 mm Female 70 mm Female Normal C 3.62 ± 0.36 4.36 ± 0.34 4.47 ± 0.55 C-like C 3.53 ± 0.33 * if C-like W 3.08 ± 0.34 W-like C 1.75 ± 0.19 2.18 ±0.17 * * * W-like W 1.41 ± 0.33 Normal W 1.44 ± 0.11 1.29 ± 0.32 1.99 ± 0.18 1.37 ± 0.22 CRAYFISH LEG REGENERATION : MUSCLE 721 Sarconicrc length The distribution of sarcomere lengths in the distal dorsal region of the closer muscle is usually bimoclal. In animals of 50-110 nun body length most of the shorter sarcomeres are 2—1- /xin long (short), but some are 4-6 //.m (intermediate; Fig. 4). As an animal grows the fraction of sampled cheliped fibers having shorter sarcomeres tends to decrease toward zero (Fig. 5). In walking legs from the same animals the fraction of short sarcomeres tended to be intermediate between the fractions for chelipeds of small and large animals. It is unclear whether short sarcomeres were absent from the distal dorsal region or were missed in sampling from some legs. Most fibers with shorter sarcomeres were in proximal subdivisions of the distal dorsal region (Table III). The non-uniformity of distribution appears especially striking in the cheliped. However, the apparent difference in distribution between chelipeds and walking legs might result from the different choice of subdivisions in the two types of legs. As Figure 4 shows, the mean length of long sarcomeres. Si,,,,,,, tends to increase with manus length for normal chelipeds. The increase in sarcomere length is insufficient to account for the elongation of muscle fibers ; mean fiber length increases more than threefold (Fig. 2A ) while s\ 7 — i E1 4 l A D Q) c i_ t> - \ 0) ^ E 5 _ B 0 J o o 4 1 x ( B ^ * . * * 3 *C< •) a> * * 2 1 - n I I I 1 111 l l i 1 1 1 1 1 1 1 34 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Manus Length (mm) FIGURE 4 : Sarcomere length of fibers in the distal dorsal part of the closer muscle, vs. manus length. Mean±s.d. calculated separately for long and for short sarcomeres. Letters A-U designate particular animals ; body length increases with position of letter in alphabet. Filled circle: normal C and C-like C; ring around filled circle: C-like W; x: normal W and W-like W; ring around x : W-like C. 722 MITTENTHAL, OLSON, AND CUMMINGS A .6 .4 Fraction .2 B Fraction .4 .2 0 Cheliped Walking Leg -•-• -L-Xl FIGURE 5 : Figure 4. 50 60 70 80 90 100 110 Body Length (mm) Fraction of fibers with short sarcomeres vs. body length. Symbols as in regions of a muscle from plots of fiber length and sarcomere length vs. external dimensions (Figs. 2, 4). Dispersion in sarcomere length may obscure constancy of the number of sarcomeres per fiber as the fibers grow. We have therefore estimated the number of sarcomeres per fiber as the mean of the ratio of fiber length to sarcomere length, for all long-sarcomere fibers sampled from each of the six sub- divisions of the distal dorsal region, in chelipeds of six normal animals. These estimates, in Table IV, show that the number of sarcomeres per fiber tends to increase with size of the crayfish in all six subdivisions and in the distal dorsal region as a whole. Thus cheliped closer fibers lengthen by addition as well as by elongation of sarcomeres. The relative contributions to elongation of walking-leg fibers by elongation and by addition of sarcomeres are unclear because of scatter in the data. The s\ong values for walking legs cluster near the maximum slonK in large chelipeds. The ratio of Si,mg values for normal C and W of one animal shifts from less than 1 for small crayfish (<65 mm; animals D, E) to greater than 1 in large crayfish (>90 mm long; animals J, P, U). In homoeotic legs (two W-like C. three C-like W), ^iong has roughly the value expected for a normal propodite of donor type with the observed manus length. Unlike mean fiber length and diameter, Si,mK for a homoeotic leg was not clearly displaced toward the value for the contralateral host-type leg. TABLE III Distribution in normal chelipeds and walking legs of sampled fibers having short sarcomeres, among the subdivisions of the distal dorsal region. Data for unoperated and operated animals are pooled, as no difference was noted. A. In normal chelipeds, fraction of 60 fibers from nine animals in the subdivisions. B. In normal walking legs, fraction of 15 fibers from four animals. A. Dorsal B. 0.20 0.067 0 0.333 0 r>irfr il 0 V*} n 0.60 0.117 0.017 0.2 0.133 Ventral CRAYFISH LEG REGENERATION : MUSCLE 723 TABLE IV Number of sarcomeres per fiber. Mean ± s.d. (number of fibers') of fiber length/mean sarcomere length for the six subdivisions of the distal dorsal region, and for the entire region, in chelipeds of six unoperated male animals. Body length (mm) 51 59 72 87 88 91 Manus length (mm) 4.54 6.76 10.8 11.9 11.2 14.3 Subdivision 1 204±19 (3) 376 ± 64 (5) 529±34 (5) 544 ± 32 (5) 709 ± 69 (5) 673 ± 19 (6) 2 223 ±24 (5) 339 ±106 (5) 44.1 ± 6 (5) 560 ± 25 (5) 734 ± 80 (5) 710± 34 (6) 3 311 ±63 (3) 480±114 (5) 587 ±44 (5) 730±112 (5) 737 ± 42 (5) 713± 80 (6) 4 280 ±26 (5) 592 ± 48 (3) 564 ±69 (5) 642 ± 39 (5) 817 ± 19 (5) 773 ± 70 (6) 5 371 (1) 387±119 (5) 517 ±63 (5) 815 ± 47 (5) 696 ±126 (5) 819±132 (6) 6 - (0) — (0) 604 (1) 790 ± 73 (3) 899 (1) 1006± 97 (6) Distal dorsal region 261 ±57 (17) 421 ±122 (23) 530 ±68 (26) 672 ±121 (28) 745 ± 86 (26) 783 ±135 (36) DISCUSSION Transplantation has been used extensively to study the development of motor systems in vertebrates (e.g., Detwiler, 1964; Hollyday et a/., 1977). However, in arthropods similar methods have been restricted to insects (Young, 1972; Westin and Camhi, 1975a, b). Recently a method for transplanting regenerating crustacean limb buds has been applied to motor systems in fiddler crabs (Trinkaus-Randall and Mittenthal, 1978) and in crayfish, as reported here. The effects of an ectopic host site on muscles in homoeotic legs will be discussed after a summary of the post- embryonic growth of crustacean muscle in normal legs. A muscle may grow by an increase in the number or in the dimensions of its fibers. In vertebrates, fibers usually grow in length and cross-sectional area but remain constant in number as the animal grows (reviewed by Goldspink, 1972). However, in hypertrophy of muscles induced by exercise the number of fibers in the muscle has been observed to increase significantly (Gonyea et al., 1977). The number of fibers in the opener muscle of the crayfish cheliped is constant at about 300, in normal and regenerated chelipeds having a 10-fold range of propodite lengths (Bittner, 1968). Bittner found that fiber length and fiber diameter increase linearly with propodite length. The dry mass of the opener muscle increased with propodite length in the way expected if the number of fibers remains constant and if the propodite maintains constant proportions as it grows. The measurements of propodite dimensions reported by Mittenthal (1980) confirm Bittner's assumption that chelipeds of different sizes are geometrically similar. Our findings on the closer muscles of the cheliped and first walking leg are consistent with Bittner's observations and assumptions for the cheliped opener: The wet mass of the fixed closer muscle is proportional to the volume of a box enclosing the manus, which contains the closer and opener. The mean length and diameter of closer fibers are proportional to linear dimensions of the manus. The latter proportionality means that the number of fibers remains nearly constant during growth. However, Davison (1956) found that abdominal flexor muscle of Procambarus alleni grows by addition of new fibers as well as by enlargement of extant fibers. MITTENTHAL, OLSON, AND CUMMINGS istological examination of the muscles showed aggregation and fusion of myo- blasts. Over a 4000-fold range of body weight, the number of fibers increased about 5-fold for a 1000- fold increase in weight. If the number of fibers in the closer muscle increased at the rate found by Davison, the data of Fig. 3A should fit a straight line of slope 0.33 rather than the observed slope of 0.5. Evidently some crustacean muscles, such as the opener in a crab (Lang, Sutterlin, and Prosser, 1970) grow only by enlargement of fibers. However, others, such as abdominal flexors of crayfish and the stretcher muscle of lobster pereiopods (Jahromi and Atwood, 1971a) also grow by increasing the number of fibers. Differences in the structure or function of muscles correlated with these two modes of growth are unknown. As a crustacean muscle fiber elongates, the number and/or length of its sarcomeres increases (Bittner, 1968; Goudey and Lang, 1974; Govind ct a/., 1974; Jahromi and Charlton, 1979). In some crustacean muscles fibers grow mainly or solely by elongation of sarcomeres (Bittner and Traut, 1978). However, in other muscles sarcomeres elongate and new sarcomeres are added (Govind ct al., 1977). Fibers of the cheliped closer muscles in P. clarkii grow in both ways. Mean sarcomere length increases slightly with manus length in chelipeds, but much of the increase in fiber length is contributed by addition of sarcomeres. Growth of sarcomeres and an increase of their number during elongation of fibers have also been observed in other arthropods (Aronson, 1961; Shafiq, 1963; Auber, 1965) and in vertebrates (Goldspink, 1968; Williams and Goldspink, 1971). Muscles in several species of decapod crustaceans have bimodal distributions of sarcomere length (Procainbarus: present work; lobster, Homarns: Jahromi and Atwood, 1971b; snapping shrimp, Alpheus: Stephens and Mellon, 1979a). In crayfish the spectrum of sarcomere lengths in the walking leg attains its asymptotic character early in development, while slow changes in the spectrum continue in the cheliped. A parallel situation occurs in small juvenile lobsters: The closer muscle of the cutter has nearly attained the adult fraction of short sarcomeres by stage 11, whereas the crusher is not yet adult-like at stage 16 (Govind and Lang, 1978). Evidently in the crayfish and lobster, the closer muscles of the larger, more specialized types of leg (crayfish cheliped, lobster crusher) have evolved in part through a prolongation of growth processes that proceed more rapidly and stop earlier in the smaller, less specialized legs. A slowing and prolongation of growth often accompanies the evolution of larger, more specialized structures (Gould, 1977). In the present work measurements of propodite dimensions, muscle mass, and fiber morphology in normal legs provide a background for evaluating the contri- bution of the host site to the properties of homoeotic legs. The homoeotic legs resembled donor legs in muscle morphology as well as external form. The host site produced a slightly greater alteration of size in muscle fibers than in external form. Similar grafting experiments in fiddler crabs also yielded regenerated legs with donor-like external form and muscles. In a large cheliped regenerated on a male specimen of Uca pugnax from the basipodite of a male specimen of Uca pugilator, sarcomere lengths of fibers from the major carpopodite extensor muscle were those expected for a donor cheliped (ca. 10 /xm ) rather than a host cheliped (13-14 /xm) (Trinkaus-Randall and Mittenthal, 1978). It is noteworthy that in some circumstances the host site can modulate the external form and muscle properties of a leg. In Alpheus loss of the snapper CRAYFISH LEG REGENERATION: MUSCLE 725 cheliped causes reversal of cheliped asymmetry : The pincer transforms to a snapper, and a pincer regenerates in place of the former snapper. Recent studies of cheliped muscles (Stephens and Mellon, 1979a, b; Mellon and Stephens, 1978, 1979) show that the closer muscles of snapper and pincer have homologous innervation but differ in spectra of fiber length, fiber diameter, and sarcomere length, and in the size and facilitatory characteristics of excitatory junctional potentials. When a pincer is transformed to a snapper, these muscle properties are transformed corre- spondingly. Evidently the site where a leg develops, and interactions between adjacent legs, can modulate the type of leg that develops, evoking a co-ordinated transition of external form and muscle properties. Recent work on the development of lobster chelipeds shows that the normal co-ordination between external form and muscle properties can be disrupted. In Hoinanis the closer muscles of crusher and cutter chelipeds have different distribution of sarcomere lengths. Normally the cutter closer has about 70% short sarcomere fibers (2-4 /xm sarcomere length) ; 30% of the fibers have sar- comeres greater than 6 p.m long (Jahromi and Atwood, 1971b). Lang et al. (1978) found that lobsters raised in smooth-bottomed tanks retained two cutter-like chelipeds. One of these chelipeds had 90% short-sarcomere fibers; the other, a "false" cutter, had 50-60%. Govind and Lang (1979) examined lobsters having paired crusher-like chelipeds. In one of these chelipeds the closer resembled a normal crusher closer, with no short sarcomeres. However, the contralateral cheliped, though crusher-like in external form, had a closer with about 40% short sarcomere fibers. These observations show that the closer muscles of lobster chelipeds tend to be asymmetric, even when the chelipeds are externally symmetric. The preceding discussion suggests a model for the control of leg development in decapod Crustacea. During ontogeny separate control systems may generate the external form and muscle properties characteristic of each leg. Normally the hypodermis and muscle generators are co-ordinated, in a way not understood, to generate typical legs. Errors in coupling the two generators may occur, so that the hypodermis generator typical of one leg and the muscle generator typical of another leg must co-operate to produce a leg. Apparently this happens in lobsters with symmetric chelipeds. Although the hypodermis and muscle generators play a major role in determining leg characteristics, the site on the body where a leg develops and the control systems generating adjacent legs can modulate the develop- ment of a leg. Such modulation can occur even if the coupling between hypodermis type and muscle type is normal, as in the homoeotic crayfish legs studied in the present work. This model has testable implications. For example, lobsters might be found in which the chelipeds appear normal externally, but the muscle asymmetry is the reverse of the external asymmetry. To screen for lobsters with reversed closer muscle asymmetry one could test the speed of the dactylopodite closing reflex, looking for animals with unusually fast crusher closing and unusually slow cutter closing. ACKNOWLEDGMENTS We thank Dr. Arthur Winfree and Ms. Vickery Trinkaus-Randall for a critical reading of the manuscript. We are indebted to Ms. Judith Franklin for the non- linear regressions. Dr. Fred Lang provided invaluable advice on histology, and helpful criticism in discussions. Dr. Lang and Dr. George Bittner supplied MITTENTHAL, OLSON, AND CUMMINGS manuscripts in press. This research was supported in part by NSF grant no Bl S77-24149. LITERATURE CITED ARONSON, J., 1961. Sarcomere size in developing muscles of a tarsonemid mite. /. Biophys. Biochcm. Cytol., 11: 147-156. ATWOOD, H. L., 1973. An attempt to account for the diversity of crustacean muscles. Am. Zoo/., 13 : 357-378. AUBER, J., 1965. L'accroissement en longeur des myofibrilles et la formation de nouveaux sarcomeres on cours du developpement des muscles chez Calliphora erythrocephala. C.R. A cad. Sci. (Paris), 261: 4845. BITTNER, G. D., 1968. The differentiation of crayfish muscle fibers during development. /. Exp. Zoo/., 167: 439-456. BITTNER, G. D., AND D. L. TRAUT, 1978. Growth of crustacean muscles and muscle fibers. ./. Comp. Physio!., 124: 277-285. BLATTNER, F. R., AND J. N. ABELSON, 1966. A precision method for generation of solvent gradients. Anal. Chctn., 38: 1279-1281. BLALOCK, H. M., 1960. Soda! Statistics. McGraw-Hill Book Co., Inc., New York, xiv, 465 pp. CLOSE, R. I., 1972. Dynamic properties of mammalian skeletal muscles. Plivsiol. Rev., 52: 129-197. DAVISON, J., 1956. An analysis of cell growth and metabolism in the crayfish (Procambarus allcni). Biol. Bull., 110: 264-273. DETWILER, S. R., 1964. Ncurocmbryology. Hafner Publ. Co., New York, x, 218 pp. DIXON, W. J., 1975. BMDP: Biomedical Computer Programs. University of California Press. Berkeley, Los Angeles, xi, 792 pp. FINLAYSON, L. H., 1975. Development and degeneration. Pp. 75-149 in P. N. R. Usherwood, Ed., Insect Muscle. Academic Press, New York. FRANZINI-ARMSTRONG, C., 1970. Natural variability in the length of thin and thick filaments in single fibers from a crab, Par hums dcpurator. J. Cell Sci., 6 : 559-592. GOLDSPINK, G., 1968. Sarcomere length during the post-natal growth of mammalian muscle fibers. /. Cell Sci., 3 : 539-548. GOLDSPINK, G., 1972. Postembryonic growth and differentiation of striated muscle. Pp. 181- 236 in G. H. Bourne, Ed., The Structure and Function of Muscle. Second edition. Vol. 1. GONYEA, W., C. G. ERICSON, AND F. BONDE-PETERSON, 1977. Skeletal muscle fiber splitting induced by weight-lifting exercise in cats. Ada Physiol. Scand., 99 : 105-109. GOUDEY, L. R., AND F. LANG, 1974. Growth of crustacean muscle : Asymmetric development of the claw closer muscles in the lobster, Homarus americanus. J. Exp. Zoo/., 189 : 421-427. GOULD, S. J., 1977. Ontogeny and Phylogeny. Harvard University (Belknap) Press, Cam- bridge, Mass, ix, 501 pp. GOVIND, C. K., H. L. ATWOOD, AND F. LANG, 1974. Sarcomere length increases in developing crustacean muscle. /. Exp. Zoo/., 189: 395-400. GOVIND, C. K., AND F. LANG, 1978. Development of the dimorphic claw closer muscles of the lobster Homarus americanus. III. Transformation to dimorphic muscles in juveniles. Biol. Bull., 154 : 55-67. GOVIND, C. K., AND F. LANG, 1979. Physiological asymmetry in the bilateral crusher claws of a lobster. /. Exp. Zoo/., 207 : 27-32. GOVIND, C. K.. J. SHE, AND F. LANG, 1977. Lengthening of lobster muscle fibres by 2 age- dependent mechanisms. Expcricntia, 33 : 35-36. HOLLYDAY, M., V. HAMBURGER, AND J. M. G. FARRIS, 1977. Localization of motor neuron pools supplying identified muscles in normal and supernumerary legs of chick embryo. Proc. Nat. Acad. Sci. U.S.A. 74: 3582-3586. JAHROMI, S. S., AND H. L. ATWOOD, 1971a. Electrical coupling and growth in lobster muscle fibers. Can. J. Zoo/., 49 : 1029-1034. JAHROMI, S. S., AND H. L. ATWOOD, 1971b. Structural and contractile properties of lobster leg- muscle fibers. /. Exp. Zoo/., 176 : 475-486. JAHROMI, S. S., AND M. P. CHARLTON, 1979. Transverse sarcomere splitting: A possible means of longitudinal growth in crab muscles. /. Cell Biol., 80 : 736-742. CRAYFISH LEG REGENERATION: MUSCLE 727 LANG, F., W. J. COSTELI.O. AND C. K. GOVIND, 1977. Development of the dimorphic claw closer muscles of the lobster Honnirus amcricanus : I. Regional distribution of muscle fiber types. Biol. Bit!!., 152 : 75-83. LANG, F., C. K. GOVIND, AND W. J. COSTELLO, 1978. Experimental transformation of muscle fiber properties in lobster. Science, 201 : 1037-1039. LANG. F., A. SUTTERLIN, AND C. L. PROSSKR, 1970. Electrical and mechanical properties of the closer muscle of the Alaskan king crab Paralithoides cuintsclnitica. Coin p. Biocliein. Physiol., 32: 615-628. MELLON, D., JR., AND P. J. STEPHENS, 1978. Limb morphology and function are transformed by contralateral nerve section in snapping shrimps. Nature, 272: 246-248. MELLON, D., JR., AND P. J. STEPHENS. 1979. The motor organization of claw-closer muscles in snapping shrimp. /. Coinp. Physiol., 132: 109-115. MITTENTHAL, J. E., 1980. On the form and size of crayfish legs regenerated after grafting. Biol. Bull, (in press). SHAFIQ, S. A., 1963. Electron microscopic studies on the indirect flight muscles of Drosophila melanogaster. J . Cell. Biol., 17 : 363-373. STEPHENS, P. J., AND DfiF. MELLON, JR., 1979a. Modification of structure and synaptic physiology in transformed shrimp muscle. J. Coinp. Physiol., 132: 97-108. STEPHENS, P. J., AND DEF. MELLON, JR., 1979b. Morphology and synaptic physiology of the main closer muscle in regenerating claws of Alplicus. J. Conip. Physiol., 134: 17-27. TRINKAUS-RANDALL, V., AND J. E. MITTENTHAL, 1978. Intra- and interspecific transplantation of limb buds in Crustacea : A new method for studying central and peripheral inter- actions in crustacean limbs. /. Ex p. Zool., 204: 275-281. VAN HARREVELD, A., 1936. A physiological solution for fresh-water crustaceans. Proc. Soc. Exp. Biol. and Mcd., 34: 428-432. WIERSMA, C. A. G., 1955. An analysis of the functional differences between the contractions of the adductor muscles in the thoracic legs of the lobster Homarns amcricanus. Arch. Neerl. Zool., 11: 1-13. WESTIN, J., AND J. M. CAMHI, 1975a. Regeneration of peripheral nerves to supernumerary legs in cockroaches. /. Exp. Biol., 63 : 483-496. WESTIN, J., AND J. M. CAMHI, 1975b. Motor innervation within supernumerary legs in cock- roaches. /. Exp. Biol., 63 : 497-503. WILLIAMS, P. E., AND G. GOLDSPINK, 1971. Longitudinal growth of striated muscle fibers. /. Cell Set., 9: 751-767. YOUNG, D., 1972. Specific re-innervation of limbs transplanted between segments in the cock- roach, Periplancta amcricana. J. Exp. Biol., 57: 305-316. Reference: Biol. Bull., 159: 728-736. (December, 1980) ANNUAL GONADAL VARIATION IN SEA URCHINS OF THE ORDERS ECHIXOTHURIOIDA AND ECHINOIDA TAKAO MORI,1 TEIZO TSUCHIYA,2 AND SHONAN AMEMIYA 3 1 Zoological Institute, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113; 2 Department of Physiology, School of Medicine, Teikyo University, Itabashi-ku, Tokyo 173; and 3 Misaki Marine Biological Station, Faculty of Science, University of Tokyo, Minra-shi. Kanaya?\.'a 238-02 ABSTRACT The echinothurioid sea urchins, Asthenosoma ijiinai and Araeosoma ou'stoni, were collected throughout the year. Their gonads were studied histologically by light microscopy and compared with those of some species of the order Echinoida. The ovaries contained degenerating and large developing oocytes throughout the year, showing that, unlike species of the order Echinoida so far studied, the Echinothuri- oid sea urchins lack the so-called resting state in which the ovaries contain only oogonial clusters. In the testes of this species, differentiation of spermatocytes into more advanced stages of male germinal cells appears to be blocked until at least 5 months before spawning. Spermatids appeared 3 months before spawning and spermatozoa at least 1 month before. Factors involved in differences in annual gonadal variation between the sea urchins of the orders Echinoida and Echinothuri- oida are discussed. INTRODUCTION Many species of sea urchins exhibit an annual reproductive cycle, with oocytes and spermatocytes becoming mature germinal cells during a few months before the breeding season (Boolootian, 1966). Although reproductive cycles have been studied in detail in some species of the orders Echinoida, Arbacioida, Clypeasteroida, etc. (Boolootian, 1966), little information is available on the gonadal cycles in the sea urchins of Echinothurioida. Recently, Amemiya and Tsuchiya (1979) suc- ceeded in fertilizing eggs of the Japanese echinothurioid, Asthenosoma ijiinai, in mid-September and followed subsequent development for 2 months. They pointed out important differences between development in this species and in the order Echinoida — differences probably ascribable to taxonomic and phylogenetic dif- ferences between the two groups. This is the first report comparing the annual reproductive cycle in this sea urchin with the cycles in some species of the order Echinoida. MATERIALS AND METHODS Reproductive cycles of two echinothurioids, Asthenosoma ijimai Yoshiwara and Araeosoma oivstoni Mortensen, were compared with those of three species of the order Echinoida, Hemiccntrotns pnlchen-imus (A. Agassiz), Anthocidaris cras- sispina (A. Agassiz), and Pseitdocentrotus depressus (A. Agassiz). Received June 5, 1980; accepted August 20, 1980. 728 SEA URCHIN GONADS 729 Adult specimens were collected at varying intervals between April, 1977, and October, 1979, along tbe coast around Misaki Marine Biological Station, Kanagawa, Japan. For histological observations, the middle portion of one of tbe five gonads in individual urcbins was removed and fixed in sea water and Bouin's rluid, de- hydrated in ethanol, cleared in xylene, and embedded in paraplast. Serial sections cut at 7 /mi were stained with Delafield's hematoxylin and eosin. Five stages of the gonadal cycle were distinguished according to the criteria described by Fuji (1960), i.e. stage I (recovery), stage II (growing), Stage III (premature), stage IV (mature), and stage Y (spent). The general morphology and staining characteristics of the ovary were used to define oocyte stages (Conor, 1973). For quantitative studies, oocytes and mature ova were counted in five ovaries randomly taken from sea urchins of each of the three species Asthenosoma ijimai, Hemicentrotus pulcherrimus, and Anthocidaris crassispina. At 70 X magni- fication, all normal-appearing oocytes with germinal vesicles and mature ova were counted in every 200th section of the ovaries, with oogonia and oocytes measuring less than 15 ju.ni in their largest diameter excluded. At least 2000 oocytes, divisible into four or five size groupings, were counted per specimen, at several different times throughout the year. Frequency distributions of oocytes and different sizes were determined from their percentages in the total oocytes counted. RESULTS There were no marked differences of the gonadal cycles and structures among the three species Hemicentrotus pulchcrrimus, Anthocidaris crassispina, and Pseudo- ccntrotus dcprcssus. Around Misaki Marine Biological Station, the breeding season usually continued from January through March in H. pulchcrrimus, from June through September in A. crassispina, and from October through December in P. dcprcssus. Each of five separate ovaries located under the peritoneum consists of branched saccules (ovarioles) with final blind acini. Young oocytes were on the ovarian wall, which was made up of connective tissue and smooth muscle fibers. The central portion of the acinous lumen was occupied by germinal cells, which ma- tured to ova, and accessory cells, here called nutritive phagocytes. In H. pulcherrimus and A. crassispina, oocytes were counted in late December, 1978; late March, 1979, and late June, 1979. Oocytes were arbitrarily divided into four size classes: diameter 15-30 /tin (young oocytes), diameter 30-65 /iiii (develop- ing oocytes), largest diameter more than 65 /*m (large primary oocytes and sec- ondary oocytes), and mature ova. The size distributions of oocytes clearly demon- strate the annual changes in ovarian activity (Figs. 1 A, B). The ovaries in stage I were filled with accessory cells containing eosinophilic globules and large dark granules in the cytoplasm. Some germinal cells were at- tached to the wall in small clusters. In stage II (growing ovaries) a large num- ber of oocytes less than 65 /xm in diameter were arranged in a single uniform layer lining the acinous wall. The acinous lumen was totally occupied by a large number of globular accessory cells. In stage III (premature ovaries), numerous large oocytes projected markedly into the acinous lumen from the ovarian wall. These oocytes eventuality left the wall of the acini. A few mature ova about 90 p.m in diameter, with small female pronuclei, were in the lumen. Ovarian acini in this stage also contained developing oocytes of varying sizes, including small young oocytes on the wall. In stage IV (mature ovaries), the acinous lumen was largely packed with mature ova, although oocytes of varying sizes and a few ac- cessory cells were on the wall. In stage V ovaries, the acinous lumen was largely 730 MORT, TSUCHIYA, AND AMEMIYA rh Y D I M OIC 0 L M MAI D L M JUN B tfl V 0 L M DEC Y D L M JUN 100 o < so rti ±U S Y SO LO L DEC Y SD LD L APR £ f hi. r ft , 1 * S Y SO LD L S Y SD LD L S Y SD LD L JUN E. SEP L SEP FIGURE 1. Frequency distribution of different-sized oocytes in the ovaries of Hcuiiccntm- tus pulcherrimus (A) and Antliocidaris crassispina (B), collected in December, 1978 (DEC), March, 1979 (MAR) and June, 1979 (JUN). Y = young oocytes 15-30 /mi in largest diameter, D = developing oocytes 30-65 /mi, L — large primary and secondary oocytes larger than 65 /mi diameter, M = mature ova. (C) Frequency distribution of different-sized oocytes in the ovaries of Asthenosoma ijimai. collected in December, 1978 (DEC), April, 1979 (APR), June, 1979 (JUN), early September, 1979 (E. SEP) and late September, 1979 (L. SEP), S = small oocytes 15-100 /mi in largest diameter, Y = young oocytes 100-400 /mi diameter, SD — small de- veloping oocytes 400-800 /mi diameter, LD = large developing oocytes 800-1100 /mi diameter, L = large primary oocytes over 1100 /mi in diameter. Vertical lines represent standard errors. occupied by accessory cells with vacuolated cytoplasm, but a few small oocytes measuring less than 20 /xm in diameter and clusters of small germinal cells were still on the wall. The wall of the acinous cavity of the sea urchin testis is composed of an outer visceral peritoneum, a middle fibromuscular layer composed of connective tissue and muscle fibers, and an inner layer consisting of germinal as well as non-germinal cells. In the non-breeding seasons, the acinous lumen was obliterated by prolifera- tion of globular accessory cells. Some clusters of germinal cells were present on the acinous wall. In the premature stage spermatogenesis was very active. In almost all acini, sperm clusters had already formed in the central parts of the lumens. The acinous wall was lined by a sheet of accessory cells. In the mature stage, the testes were distended with mature spermatozoa, and some accessory cells were ob- servable around sperm aggregations. The echinothurioid sea urchin, A. ijimai, has five separate ovaries, each branching into saccules (ovarioles) ending in blind acini. However, the ovaries differ in histological structure from those of echinoid species. The echinothurioid ovaries studied were under the peritoneal epithelium, and had a wall composed of col- lagenous connective tissue and smooth muscle fibers. From the inner wall, thin SEA URCHIN1 GOXADS connective tissue partitions extended interiorly, dividing the acini into several compartments containing germinal and non-germinal cells (Fig. 2A). The par- titions ramify and grow toward the central part of the acinous lumen. Oocytes of varying sizes attaching to the ramified partitions arc distributed without apparent order in the ovarian acini. The ovary is readily distinguishable from the testis FIGURE 2. (A) Sections of ovarian lobes of Asthcnosoma ijiinai. Peripheral part of the ovarian acinus. Arrows indicate connective tissue partitions. (B) Part of ovarian acinus from a specimen collected on December 21, 1978. (C) Cluster of germinal cells (arn>\\ i attached to ovarian wall. (D) Part of ovarian acinus from a specimen collected on September 14, 1979. (E) Cortical portion of an oocyte, undergoing dissolution. A, C, and E X 400. Scale line in A represents 20 /urn. B and D X 40. Scale line in B represents 200 /j.m. 732 MORI, TSUCHIYA, AND AMEMIYA throughout the year (cf. Fig. 3), since the oocytes in the peripheral region of the ovarioles can be seen with the naked eye. Oocytes counted in late December, 1978; early April, 1979; late June, 1979; early September, 1979; and late September, 1979, were divided into five size group- ings by largest diameter : those from 15-100 /mi (small oocytes), 100-400 /mi (young oocytes), 400-800 /mi (small developing oocytes: oocytes over 400 /mi had eosinophilic granular cytoplasm, suggesting vitellogenesis), 800-1100 /mi (large developing oocytes), and over 1100 /mi (large primary oocytes whose dense, yolk- filled cytoplasm stained pink). Primary oocytes were not distinguishable from secondary oocytes, as reliable criteria for the species are not known. Unlike spe- cies of the order Echinoida, in A. ijiniai the size distribution diagrams of oocytes did not show any clear annual changes (Fig. 1C), demonstrating that many small primary oocytes continue to grow throughout the year, and that oocytes of almost all sizes occur at any itme. In December, ovaries contained oocytes of varying sizes. The largest ones, about 100 /Jim in largest diameter, were surrounded by vacuolated accessory cells and stained homogeneously with eosin. The large oocytes had germinal vesicles about 200 /mi in diameter. The acini contained many irregular- or polygonal- shaped degenerating oocytes with large vacuoles in the cytoplasm. Oocytes over 400 /mi in largest diameter were eosinophilic, while smaller ones were basophilic. Germinal vesicles of all oocytes contained a single, large, spherical, and strongly basophilic nucleolus (Fig. 2B). Clusters of small germinal cells, composed of closely packed cells of varying sizes with indistinct boundaries, were distributed over the inner connective tissue partitions (Fig. 2C). The cells sometimes ap- peared as aggregates of small, strongly staining nuclei, occasionally in the process of mitosis or meiosis. It was not known when the primary oocytes were produced from oogonia during the annual reproductive cycle. Globulated accessory cells were sparsely scattered in the ovarioles. In April, the largest oocytes reached about 1100 /mi in largest diameter. Many irregular-shaped degenerating oocytes were still present. In June, the ovarioles contained more oocytes and fewer accessory cells than in April. The largest oocytes measured about 1300 /mi in diameter. However, ovarioles still contained degen- erating oocytes. From late August through early September, ovarian sructure was almost the same as in June. Some oocytes were degenerating. Accessory cells decreased in number and almost all of them became vacuolated. Clusters of small germinal cells were arranged over the connective tissue septa. In mid-September, the ovaries contained many large oocytes 1100-1300 /mi in largest diameter. Crowded together within the narrow ovarioles, these oocytes be- came irregularly polygonal in shape (Fig. 2D). Careful examination of serial sec- tions revealed that the oocytes had no nuclei. A few accesseory cells were visible on the acinous wall. Oocytes in the ovaries, which seemed to have already spawned, were invariably less than 800 /mi in diameter. A decrease in number of oocytes was followed by a rapid increase in number of empty-appearing accessory cells. In late September, after the spawning season, almost all remaining oocytes were degenerating but had increased in size up to 800 /mi. The degenerating oocytes had disintegrating cytoplasm and were irregular in shape. In other seasons, the egg cortex of degenerating oocytes remained relatively unchanged for a while. By constrast, in post-spawning ovaries, degeneration of oocytes began with the dis- solution of the cortical layer, and the disintegration of the inner cytoplasm followed (Fig. 2F). Empty-appearing accessory cells in post-spawning ovaries increased in SEA URCHIX GO NADS 733 C 1 i*?&3 ~ F V*H ' •. F FIGURE 3. Sections of testicular lobes of Astlicnosoma ijiinai. Parts of testicular acini from specimens collected December 21, 1978 (A), April 7, 1979 (B), June 25, 1979 (C). August 14, 1979 (D) and September 12, 1979 (E). All figures except DxlOO; D X 400. Scale line in A represents 100 /urn and applies to A, B, C, and E; scale in D represents 20 /j.m. number continuously and exhibited many dark granular inclusions. By October, oocytes measuring 400-600 /xm in diameter had increased in number. The largest oocytes attained 800 /tin, with germinal vesicles of 200-/im diameter. A few de- generating oocytes were also found. Accessory cells, still empty looking, in- creased in number. The testicular acini of A. ijiinai are composed of an outer visceral peritoneum, a middle fibromuscular layer, and an inner layer consisting of germinal and non- germinal cells. The inner surface of the inner layer is covered with a thin con- nective tissue membrane, as in the ovarioles. MORI, TSUCHIYA, AND AMEMIYA FIGURE 4. Ovaries of Araeosoma owstoni (left) and of Heniicentrotus pulcherrimus (right). 1.1 X. Scale line indicates 1 cm. In late December, the testes were filled with globulated accessory cells and some clusters of germinal cells located over the inner connective tissue partitions. In some individuals, the acinous lumen contained small masses of relict sperm (Fig. 3A). In early April, the testicular acini had numerous folds on the wall ex- tending into the acinous lumen. Vacuolated empty-looking accessory cells were arranged on this wall (Fig. 3B). Clusters of germinal cells less than 10 jam in diameter were on the inner connective tissue partitions. In late June, the lumen of the testicular acini widened (Fig. 3C). Spermatogenesis, taking place on the con- nective tissue partitions, resulted in production of a few spermatids in the acini, but spermatozoa were not yet present. The primary and secondary spermatocytes were not distinguishable. In August, spermatogenesis was going on even in the central part of the acini (Fig. 3D). In early September, the acinous lumen was totally occupied by mature spermatozoa and the testicular acini were greatly en- larged (Fig. 3E). In late September, the testes were approximately the same in structure as in early September, except for a decrease in numbers of spermatozoa. Ovaries of Araeosoma ozvstoni are similar in structure to ovaries of A. ijhnai (Fig. 4). In mid-March, the largest A. o^vston^ oocytes measured about 1000 ju.m in diameter and the germinal vesicles about 200 /xm. Oocytes of varying sizes were distributed without apparent order in the ovarian acini. In early April, before the spawning season, the largest oocytes reached 1200 pm in diameter, with germinal vesicles measuring 340 //,m. In late April, after the spawning season, the maximum diameter of the oocytes dropped to 910 /tin and remained there until early October. Degenerating large oocytes were present throughout the year. The testes of A. owstoni showed active spermatogenesis from March through April. In late May, spermatogenesis was still going on in the acini but empty- appearing accessory cells increased in number. From early June through early October, spermatogenesis was no longer visible, the main components of the testes being vacuolated accessory cells. SEA URCHIN GONADS 7.S5 DISCUSSION Ovaries of the species of the order Echinoida exhibit a definite annual cycle, demonstrated by the frequency distributions of oocytes of different sizes at differ- ent times of the year (Pearse and Giese, 1966; Gonor, 1973). Mature ova and developing oocytes were totally absent during non-breeding seasons, until just be- fore onset of the breeding season (Boolootian, 1966; Masuda and Dan, 1977; the present study). By contrast, in the echinothurioid sea urchins A. ijiinai and A. ou'stoni no such distinct cyclic changes occurred in oocyte growth and differen- tiation. Small young and large developing oocytes are found in the ovaries almost throughout the year, although spawing is restricted to a short definite season. However, testes show a distinct annual activity cycle in these echinothurioid species. Lunar periods, food, salinity, light, and temperature, among other factors, are known to influence reproductive cycles. It is particularly well established that fluctuations in quantity and quality of food are involved in changes in gonadal ac- tivity: Adequate food permits maturation of a large number of oocytes (Boolootian, 1966), while starved sea urchins (Stronglyoccntrotts purpuratus) failed to repro- duce (Giese, 1959). These findings suggest that energy reserves in the gonads are important in promoting reproductive activity in the sea urchin. The echinothurioid sea urchins dealt with in this paper live on the rocky bottom 20-30 m deep (A. ijimai] or on sandy bottom 80-100 m deep (A. ou'stoni), where the effects of seasonal fluctuations may be much mitigated. The intestines of the sea urchins examined contained mainly nereids, barnacles, and sponges, showing that the urchins were carnivorous. Thus they may be able to get abundant food almost year-round, in contrast to herbivorous echinoids living on seaweed that grows seasonally. This may largely account for the lack of marked rhythmic changes in ovarian morphology in the echinothurioids. However, further detailed studies are needed. Factors other than food may be largely responsible for the annual cycle of spermatogenesis in the same species. In species of the order Echinoida, both oogenesis and spermatogenesis occur from separated germinal cells or in small groups of germinal cells within the ovarian wall of connective tissue and muscle fibers (Holland and Giese, 1965 ; Chat- lynne, 1969; Masuda and Dan, 1977). Holland (1967) reported similar findings in the cidarid, Stylocidaris affinis. In species of the order Echinothurioida, ex- tentions from the inner connective tissue wall on which young germinal cells are located grow into the central part of gonadal acini, so that young germinal cells are in the central as well as the peripheral parts of the acini and give rise to large oocytes or spermatozoa. Thus, oocytes of varying sizes are distributed throughout the lumen of the ovarian acini. ACKNOWLEDGMENTS This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan. We wish to express our thanks to Emeritus Professor K. Takewaki of the University of Tokyo for his critical reading of the manuscript and to Dr. M. Shigei for his valuable advice. This paper is dedicated to Emeritus Professor J. Ishida of the University of Tokyo on his 70th birthday. LITERATURE CITED AMEMIYA, S., AND T. TSUCHIYA, 1979. Development of the echinothurid sea urchin Astheno- soma ijiinai. Mar. Biol. 52 : 93-96. MORI, TSUCHIYA, AND AMEMIYA BOOLOOTIAN, R. A., 1966. Reproductive physiology. Pp. 561-613. in R. A. Boolootian, Ed.. Physiology of Echinodermata. Interscience Publishers, New York. CHATLYNNE, L. G., 1969. A histochemical study of oogenesis in the sea urchin Strongylocen- trotus purpuratus. Biol. Bull, 136: 167-184. FUJI, A., 1960. Studies on the biology of the sea urchin. I. Superficial and histological gonadal changes in gametogenic process of two sea urchins, Strongylocentrotus nudus and 5. intermedium. Bull. Fac. Fish. Hokkaido Univ.. 11 : 1-14. GIESE, A. C., 1959. Reproductive cycles of some west coast invertebrates. Pp. 625-638. in R. Withrow, Ed., Conference on Photopcriodisin and Related Phenomena in Plants and Animals. Am. Assoc. Adv. Sci., Washington, D. C. GONOR, J. J., 1973. Reproductive cycles in oregon populations of the echinoid, Strongylocentro- tus purpuratus (Stimpson). I. Annual gonad growth and ovarian gametogenic cycles. /. Ejcp. Mar. Biol. Ecol. 12 : 45-64. HOLLAND, N. D., 1967. Gametogenesis during the annual reproductive cycle in a cidaroid sea urchin (Stylocidaris affinis). Biol. Bull., 133: 578-590. HOLLAND, N. D., AND A. C. GIESE. 1965. An autoradiographic investigation of the gonad s of the purple sea urchin (Strongylocentrotus purpuratus}. Biol. Bull.. 128: 241-258. MASUDA, R., AND J. C. DAN, 1977. Studies on the annual reproductive cycle of the sea urchin and the acid phosphatase activity of relict ova. Biol. Bull., 153 : 577-590. PEARSE, J. S., AND A. C. GIESE, 1966. Food, reproduction and organic constitution of the common antarctic echinoid Stercchinus neuma\eri (Meissner). Biol. Bull.. 130: 387-401. Reference: Biol. Bull., 159: 737-751. (December, 1980) THE MORPHOLOGY, LIFE HISTORY, AND SYSTEMATIC RELATIONS OF TVIWLOl'ESICULA ri\C,('IS (LINTOX, 1940) MAXTER, 1947 (TREMATODA: HEMIURIDAE)1 HORACE W. STUNKARD American Museum of Natural History, Central Park U'est at 79th Street, Nn<< York. New York; and Marine Biological Laboratory, Woods Hale, Massachusetts ABSTRACT Linton (1910) described Diniirus riibcus n. sp., a digenetic trematode from species of Lycodontis at Tortugas, Florida, and Manter (1931) descril)ed a second species, Dinurns niagnus, from Synodns foctcns at Beaufort, Xorth Carolina. He (1947) transferred both species to the genus Stomachicola Yamaguti, 1934, as Stomachicola rnbca and Stomachicola magna. Meanwhile, Linton (1940) de- scribed Dinurns pinguis from Menidia incnidia at Woods Hole, Massachusetts, and Manter (1947) transferred this species, pinguis, to Tubulovesicula Yamaguti, 1934. Sinclair ct al. (1972) in a 2-year study of Stomachicola rubca noted that the worms attain maximum size only in the stomach of "true" definitive hosts, large fishes, and that small fishes, which ingest planktonic invertebrates, serve as "transfer hosts." In these hosts, the parasites may wander about in the tissues, or may be encysted. These authors predicated that T. pinguis is, in such a transfer host, a stage in the life cycle of ^. rubca. Stunkard (1973) described the develop- ment of T. pinguis in M. incnidia. The worms matured without acquiring any of the essential characters of the genus Stomachicola, and, although the life cycles of both species were unknown, the proposal of Sinclair ct al. was rejected. With the discovery of the life cycle and asexual generations of T. pinguis, described in this report, the morphological differences between Stomachicola and Tubulovesicula are supplemented with bionomic characteristics that clearly establish the validity of T. pinguis as a valid and independent species. HISTORICAL BACKGROUND Linton (1940) described Dinurus pinguis n. sp. from Menidia incnidia notata taken at Woods Hole, Massachusetts. The text was supplemented by figures (Plate 9, Fig. 96 and Plate 10, Figs. 97-100). Also, specimens taken in former years from other fishes were included in the new species. Records of collection, measurements of worms, and numbers of specimens deposited in the Helmintho- logical Collection of the U. S. National Museum were noted. The list of hosts included: (1), Anguilla rostrata; (2), Roccus lincatus; (3), Fundulus hctero- clitus; (4), Menidia rnenidia ; (5), Synodus joetcns; (6), Tylosurits marinus; (7), Menticirrhus americanus ; (8), Merluccius bilincaris; (9), Menticirrhus saxatilis; (10), Paralichthys dentatus; (11). Opsanns tan; (12), Mcrulinus caro- linus; and (13), Sphyraena borealis. Received June 16, 1980; accepted September 25, 1980. 1 Investigation supported by NSF-DEB 80-06150. 737 HORACE W. STUNKARD The identification of worms from the stomachs of marine fishes is a precarious task. The problem becomes almost impossible when a specimen is juvenile, dis- torted, or partially digested. The likelihood of predation is great, since fishes eat other smaller fishes. It follows, therefore, that the report by Linton of Dinurus pinguis in many host species is largely irrelevant. The species, pinguis, was transferred from the genus Dinurus Looss, 1907 to Tubnlovesiciila Yamaguti, 1934 by Manter (1947). It is a member of the sub- family Dinurinae Looss, 1907; one of 25 subfamilies listed by Yamaguti (1971) in the family Hemiuridae Lube, 1901, an enormous taxonomic group comprising hundreds of genera. Certain authors recognize the assemblage as a superfamily, Hemiuroidea Faust, 1929 or order, Hemiurata (Markevich, 1951) Skrjabin and Guschanskaja, 1954, with the subfamilies of Yamaguti elevated to family status. Members of the group are chiefly stomach parasites of marine fishes, with a few species in freshwater fishes. The subfamily Halipeginae Esjmont, 1931 ( = Hali- pegidae Poche, 1925) contains species parasitic in freshwater fishes, amphibians and reptiles. Indeed, Parukhin (1969) described specimens from the sea-snake, Laticuada sp., taken in the Gulf of Tonkin, off the coast of Vietnam, as Tubo- lovcsicnla laticnadai n. sp. It is the only species of Dinurinae reported from reptiles. Yamaguti (1934) reported on parasites of fishes from the Inland Sea of Japan. In the subfamily Dinurinae he erected three new genera: Stomachicola, with S. mnraencsocis n. sp., from the stomach of Muraeneso.v cinercns as type species; Tubulovesicnla, with T. spari n. sp., based on a single specimen from Sparus uiacroceplialus as type; and Magnacetabulum, based on M. trachuri n. sp., from Trachurns japonica. Referring to the genus Tubulovesicnla, he predicated, p. 70, "Lecithastcr lindbcrgi Layman, 1930, undoubtedly belongs in this genus." In the genus he included also Tubnlovesiciila anguillac n. sp., from Anguilla japonica and Tubulovesicnla muracne n. sp. from M. cincreits. He noted, p. 474, "This species is closely related to Tubulovesicnla anguistacauda (Nicoll, 1915) from M. cinerens, but differs markedly in the size of eggs and in the posterior extent of the vesicula seminalis and pars prostatica. The uterus is confined to the body proper in young individuals, so that the differences in its posterior extent cannot be utilized for specific diagnosis." Although Yamaguti (1934) did not specifically transfer Ectenurus angusticauda Nicoll, 1915 to Tubulovesicula, the statement just quoted has that effect. Manter (1940) described digenetic trematodes of fishes from the Galapagos Islands. In the subfamily Dinurinae, he erected two new genera : Elytrophallns and Mecodenis. The first genus was compared with the related genera, Stcrrhnrns Looss, 1907, Tubulovesicula Yamaguti, 1934, and Clupenurus Srivastava, 1935. The last genus, Clupenurus, he noted, should be considered a synonym of Tubulovesicula. The digenetic trematodes of marine fishes at Tortugas, Florida, were studied by Manter (1947). In the subfamily Dinurinae he listed the diagnostic features and recognized species of the following genera : Lccithocladium Lube, 1901 ; Dinurus Looss, 1907; Ectenurus Looss, 1907, Stomachicola Yamaguti, 1934; Tubulovesicula. Yamaguti, 1934; Erileptus Woolcock, 1935; Clupenurus Srivastava, 1935; Elytrophallus Manter, 1940; and Mecodcrus Manter, 1940, Clupcnurus Srivastava, listed as a synonym in Manter (1940), was accepted as a valid genus. In the genus Tubulovesicula he recognized T. spari Yamaguti, 1934; T. anguillae Yamaguti, 1934; T. uiuraenesocis Yamaguti, 1934; T. cali- jornica Park, 1936; T. pseudorhonibi Yamaguti, 1938; T. lindbergi (Layman, LIFE CYCLE OF TUBULOVESICU LA PINGUIS 739 1930) Yamaguti, 1934; T. nanaimoensis (McFarlane, 1955) n. comb, (synonym Dinurus nanaimoensis McFarlane, 1935); T. pinguis (Linton, 1940) n. comb, (synonym Dinurus pingiiis Linton, 1940; T. angusticauda (Nicoll, 1915) Yama- guti, 1934. He gave a brief diagnosis of Stomachicola; in addition to the type species, S. muracnesocis Yamaguti, 1934, he included S. secunda Srivastava, 1959 ; S. uiagna (Manter, 1931) n. comb, (synonym Dinurus ntagniis Manter, 1928); S. rube a (Linton, 1910) n. comb, (synonym Dinurus rubcus Linton, 1910). Manter (1954) revised the genus Tubulovcsicula: T. nanaimoensis (Mc- Farlane, 1935) Manter, 1947 and T. madnrcnsis Nigrelli, 1940 were considered synonyms of T. lindbcrgi. Tubulovesicula calijornica Park, 1936 was described from a single specimen found in Enophrys bison at Dillon Beach, California. Manter declared, p. 546, "I cannot find significant differences between this species and T. spari Yamaguti, 1934, T. pseudorhombi Yamaguti, 1934, and therefore con- sider them all to be T. spari, the type of the genus. T. anguillac differs only in that the ecsoma is as long as the body." This species was suppressed as a synonym of T. spari by Manter (1954) and as a synonym of T. lindbergi by McCauley (1960). Tubulovesicula spari was regarded as a synonym of T. lindbergi by Sogandares-Bernal (1959) and by Zhukov (1960). After reviewing the synonymy in Tubulovesicula, Sogandares-Bernal (1959, p. 104) stated, "Since T. spari ap- pears to be a synonym of T. lindbergi, the latter becomes the type species of the genus. T. serrani Nagaty, 1956, which was described from one specimen, also appears to be a synonym of T. lindbcrgi. I believe that only four species of Tubulovesicula are valid : T. lindbergi, T. anguisticaiida, T. pingiiis, and T. magnacetabulum." Linton (1910) described Dinurus rubcus n. sp., from Lycodontis moringa and Lycodontis junebris, taken at the laboratory of the Carnegie Institution at Tortugas, Florida. The description was supplemented by his Figures 149-154. A second species, Dinurus magnus, was described by Manter (1931), based on two specimens from the stomach of the lizardfish, Synodus foetens, and several immature specimens from the stomach and air-bladder of the spotted trout, Cynoscion nebulosus, taken at Beaufort, North Carolina. Both species were transferred to Stomachicola Yamaguti, 1934 by Manter (1947). Corkum (1959) described gravid specimens from the stomach of S. foe tens taken on the Mississippi Gulf Coast as 5\ magnus (Manter, 1931). He (1966) found one mature and many immature specimens of 5. magnus in the stomach and intestines of the flounder, Paralichtys lethostigma. The paucity of mature and presence of a great many immature specimens suggested that the flounder is not a true definitive host, but rather one in which a few worms reach maturity while the majority remain immature. In the tonguefish, Symplurus palgiusa, 61 of 96 fishes had hemiurid metacercariae encysted in the body musculature. Every region of the body was involved and the encysted larvae were indicated by large, dark brown spots. The encysted worms were identified as immature specimens of S. magnus. Corkum concluded, p. 49, "It seems doubtful that these fishes are true intermediate hosts since Crustacea are the usual hosts for larval hemiurids." In essence, Corkum suggested that fishes serve as second intermediate hosts of S. magnus. In a paper dealing with T. pinguis, Sinclair ct al. (1972) predicated that this species is a stage in the life cycle of Stomachicola rubeus (Linton, 1910) Manter, 1947. They reported 5". rubeus from 28 species of marine fishes taken near Sapelo Island, Georgia, between 1 October 1969 and the autumn of 1971, noting the incidence of infection and seasonal development of the parasites. They re- HORACE W. STUNKARD called that the species was described by Linton (1910) as Dinurus rubeus from species of Lycodontis at Tortugas and that Planter (1931) had described a second species, Dinurus inagnits, from Synodus foctcns. Both species were transferred to the genus Stomachicola by Manter (1947); the specific names, rubeus and niagnus, were incorrectly emended to rubea and magna respectively, apparently on the mistaken belief that Stomachicola is a feminine name. Sinclair et al. (1972) suppressed 5\ inagnns as a synonym of 5". rubeus and predicated, p. 253, "an additional synonymy with S. rubeus is the designation Distomuin tornatum (in part) of Linton (1940) and Dawes (1946 and as Tubulovesicula pinguis by Manter (1947, 1954), Skrjabin and Guschanskaja (1954), and Yamaguti (1958), Sogandares (1959), and Overstreet (1968)." According to Sinclair ct al., S. rubeus uses a large number of small fishes as "transfer-hosts." These individuals may belong to species which mature at a small size or are the young of larger species which presumably have similar food sources. In these hosts the parasites wander freely in the body cavity or in the tissues of various organs, often invading the liver, spleen, heart, kidney, swim- bladder, and body musculature. The worms may mature at a length of 3-3.5 mm and often leave trails of eggs as they move about. It was postulated that the transfer-hosts became infected in the spring and summer and that the worms con- tinue to grow and mature until they are trapped in the tissues by defense reactions of the host, walled off in melanated cysts, and then "mummified" in winter and spring. Or they may be eaten by "true" definitive hosts and remain in the stomach and mature into 5. rubeus. These authors confirmed Corkum's findings of en- cysted .5*. inagnus in Paralichthys Icthostignia, with heavy infections in one-third of the fishes examined. They identified the specimens as 5. rubeus and suggested that Corkum's report of metacercariae in the tonguefish should not be interpreted literally, as it is doubtful that the organisms reported by him were metacercariae. They regarded encystment in muscular tissue as one of the variable sites selected by younger stages of the postmetacercarial specimens of this species. They noted that Corkum's account concerned only encysted worms observed during summer, thus nearly eliminating any chance of observing mature specimens in similar situa- tions. In the tonguefish, S. plagiusa, Sinclair ct al. ( 1972) found the worms more commonly encysted in the stomach-wall. Sinclair et al. (1972) reported that 5. rubeus develops maximum size only in "true" definitive hosts; the American eel (Angnilla rostrata), tarpon (Mega- lops atlantiea), lizardfish (Synodus foetcns), and kingfish (Mcnticirrhits ameri- canus). They noted, p. 257, "The growth pattern of Stomachicola rubea in transfer hosts is quite diverse and depends on microhabitat. Worms encysted in the stomach wall of hosts enlarge considerably in body dimensions with little growth of the ecsoma and no apparent genital differentiation. Those free in the coelom or in body tissues elongate with comparable length of soma and ecsoma and eventually become ovigerous. Thus the diversity of forms in definitive hosts seemingly reflects the developmental stage attained in the many transfer hosts before they are ingested by a definitive host." Sinclair et al. traced the develop- ment of the parasite in the lizardfish (a definitive host) from June to October, with a relative increase in the length of the ecsoma as the worms grew. They reported, p. 257, "This same pattern is portrayed in coelomic inhabitants of transfer hosts, with only quite large and mature worms being found during the late fall and winter months. The larger species of transfer host appear to pro- vide a more favorable milieu; in the kingfish and Cynoscion spp., the worms attain LIFE CYCLE OF TUBULOVESICULA PINGUIS 741 a length of 9.44-14.40 (avg. 11.48) mm in midwinter. From the pattern ob- served it would appear that the very largest worms (22-25 mm) of this species could well be more than 1 year of age." Sinclair et al. (1972) observed that the lizanlfish, Synodits foetcns. and the kingfish, Mentidrrhus amcricanus, are "true" definitive hosts of S. rnbcits, but that young, small individuals of these species may serve as transfer hosts. They noted, p. 256, "This dual site for infection in the same host species depending on size of the host would suggest that residence in a transfer host is a necessary part of the life-cycle of 5\ rnbea, as it is unlikely that the definitive hosts acquire it from the metacercarial host (most likely a small crustacean)." This statement suggests or implies a four host life cycle in 5. rubcits. If development in a transfer host is essential before the metacercaria can mature in a definitive host, the life cycle is affected. Overstreet (1968, p. 451) reported S. magna (Manter, 1931) Manter, 1947 from 5. foetcns in South Florida. He noted specimens over 2 cm in length and other immature worms which appear to be S. magna. Whether or not worms from Mcnidia menidia, identified as Tubitlovesicula pinguis (Linton, 1910) Manter 1947 can continue to grow and transform into S. rubcits remains undetermined. The present report is a contribution toward the solution of the problem. The development of T. pinguis in Menidia menidia was described by Stunkard (1973). The morphology of the worms, taken from the body cavity and tissues of the fishes, was traced from juveniles, 0.50 mm long and 0.10 mm wide to gravid specimens 6.5 mm long and 1.50 mm wide. The ecsoma in the immediate postmetacercariae stage is merely a retractable protuberance, which elongates as the worm grows until it becomes as long as the soma in sexually mature specimens. The systematic relations of T. pinguis were considered in a discussion of the life cycles and classification of the hemiurid trematodes. For more than 50 years, the mollusks of the Woods Hole region have been examined for infection by digenetic trematodes and only one hemiurid species has previously been discovered. It occurs in Odostomia trifida (Totten, 1834) ; the discovery and description of the stages in the snail were made by Hunninen and Cable (1943) and their experimental demonstration that the parasite is the asexual generation of Lccithastcr conjusus Odhner, 1905, was the first life cycle of a marine hemiurid trematode completed under controlled conditions. On 20 June 1976 a second hemiurid infection was found. It occurred in Nassarius trivitattits and is described in the present report as the asexual generation of Tubulovesicula pinguis. MATERIALS AND METHODS The experimental work detailed in this report was done at the Marine Bio- logical Laboratory, Woods Hole, Massachusetts, in the summer months, 1972- 1979. Since 1973, special efforts have been made to find the larval stages and complete the life cycle of T. pinguis. Thousands of specimens of common and even rare mollusks have been collected and examined for infection. To find the asexual generations of digenetic trematodes, snails and bivalves were isolated in bowls, 10-30, depending on size, in each bowl. On alternate days, the snails from each bowl were transferred to a bowl of fresh seawater and the water in the original bowl was examined for cercariae. If cercariae were found, the snails were isolated singly, to identify the infected one. After an isolation period 742 HORACE W. STUNKARD of some 2 weeks, snails that had not liberated cercariae were crushed to look for immature infections. In a collection of 110 specimens of Nassariits trivitattatus, made 20 June 1976, in the outer harbor at Ouissett, Massachusetts, one snail was shedding cyto- cercous cercariae. This snail has been studied for many years ; hundreds of specimens have been collected along the coast of Xew England, and no infection by a trematode had previously been reported. Over 500 additional specimens were taken during the summer of 1976, and no other infection was found. The infected snail continued to liberate 20-200 cercariae daily, chiefly at night, until it was crushed on 16 August to obtain the redial generations of the parasite. The tissues of the snail were intact and showed little injury as a result of the infection. There were six large rediae in the digestive gland. One (Fig. 1) was ap- parently almost exhausted ; it contained few developing cercariae. The others were filled with developing cercariae. No small daughter rediae were found. There were no encysted cercariae in the rediae or in the haemocoele of the snail. The cercariae did not accumulate in the haemocoele, but emerged from the snail soon after leaving the rediae. They did not encyst in the snail, but when they appeared in the seawater, the body was encased in a thin-walled cyst, to which the tail was attached. Encystment must have been rapid, almost instantaneous, and presumably occurred when the cercariae emerged from the gills of the snail into seawater. The cercariae swam vigorously at all levels in the water. There was no apparent response to light, but lashing of the tail caused the larvae to rise in the water. The cercariae were hardy ; they lived for 4 days at room tempera- ture and for 10 days at 12°C. When exhausted by swimming, the cercariae died, without excystment, on the bottom of the container. In an attempt to complete the life cycle, the cercariae were fed to copepods, Acartia tonsa, which ingested them avidly. The copepods were maintained in bowls set in running seawater in the labora- tory and in a cold room at 12 °C. Cultures of diatoms were provided as food. The copepods did not live well in the laboratory, although the temperature of the seawater was lower than the ambient air and some bowls were supplied with compressed air. At the reduced temperature of the cold room, the copepods lived well for 2-3 weeks. Copepods that had been fed were dissected ; cercariae still in their cysts were taken from the stomachs and excysted metacercariae were found free in the body cavities of the copepods (Fig. 10). Growth was rapid in the copepods and a series of developmental stages were obtained, from recently acquired experimental to naturally acquired mature metacercariae (Fig. 11). The largest metacercariae closely resembled the smallest juveniles of T. pingnis and had conspicuous ecsomas (Fig. 17). The probability of specific identity was postulated, but convincing evidence was not obtained. The rediae, cercariae, and metacercariae were studied alive, after fixation and staining with various techniques. A brief preliminary account of the asexual stages of the parasite was made (Stunkard, 1976) . In the summer of 1977, more than 600 N. trivitattits were collected and isolated, but no cercariae were obtained. One snail, taken on 9 June was crushed on 20 June, contained an immature infection. In the hope of obtaining cercariae later, no more snails were crushed until mid-August, but no other infection was found. In the summer of 1978, the collection, isolation, and examination of more than 600 N. trivitattus yielded no infection by hemiurid trematodes. (A specimen of LIFE CYCLE OF TUBULOVESICULA PINGUIS 743 N. trivitattus, taken 20 July, was shedding a large, ophtlulmotrichocercous cer- caria; this was the second trematode species found in this snail. The identity and life history of the parasite were determined later ; it proved to be the larva of Lepocrcadium arcolatitiu ( Linton, 1900) described by Stunkard, 1980.) At- tempts to infect N. trivitattus specimens by feeding them eggs of T. pinguis were futile. Mcnidia incnidia did not appear until the middle of July and then they were small, 2-3 cm long, and their digestive tracts were filled with algae. Later, larger specimens, 3-4 cm long, had veligers and copepods in their stomachs. The first gravid worms were taken in August. Eggs of T. pinguis were added to bowls containing N. trivitattus. Examination of the snails later gave only negative results. The eggs of this species measure 0.018 by 0.012 mm, are almost transparent, and hardly visible with a dissecting microscope. Collection, isolation, and examination of more than 600 N . trivitattus continued the investigation in 1979. An immature infection was disclosed in a snail crushed on 25 June (Fig. 12) and another in a snail crushed on 6 August, but no swimming cercariae were obtained. During a period of eight summers, only one snail was found shedding cystocercous cercariae, the specimen of N. trivitattus taken on 20 June 1976. Accordingly, another method was needed to disclose the life cycle of T. pinguis, and efforts were made to infect snails. In June 1979 bowls were prepared, with three snails of different sizes in each bowl. To ensure the health and vigor of these carnivorous snails, bits of such food as squid, clam, or crab tissue were provided on alternate days, before they were transferred to bowls of fresh sea water. When gravid T. pinguis became available, late in July, worms were placed in half tap : half seawater in watchglasses, and eggs in the terminal part of the uterus were extruded. Fresh eggs and eggs that had been embryonated in seawater at room temperature for 10 days were fed to snails that had been isolated since early June. In previous years, the eggs had been allowed to settle on algae which was added to the bowls with the snails. In the 1979 experiments, the eggs were added to minced squid, clam, or crab tissue, to increase the likeli- hood that they would be ingested. Eggs were fed beginning 27 July. Examina- tion of snails began 25 August. In three snails, small spherical to oval deeply staining cellular masses (Fig. 15) were found. They are recognized as sporocysts. Larger oval sporocysts (Fig. 16) were embedded in snail tissue. Motile sporo- cysts were not obtained. The brief periods of summer research and the late appearance of M. incnidia precluded consummation of the asexual generation, but experimental infection of the snail apparently completed the life cycle. DESCRIPTIONS Redia (Figs. 1,2,12} There are successive generations of rediae. Daughter rediae produce daughters, (Fig. 2), and more than 100 rediae were counted in a recent, immature infection. One of these rediae (Fig. 12) 4.00 mm long, contained thousands of germ-balls and developing cercariae. The rediae are essentially cylindrical, about 10 times as long as wide, and up to 5 mm in length. The body wall is composed of three layers. An external, syncytial tegument rests on a delicate layer of circular and longitudinal muscle fibers. Below the muscle layer is a stratum of parenchyma! cells, and the lumen may be extensive in old, exhausted rediae ( Fig. 1 ) , or filled completely by developing cercariae in young rediae (Fig. 12). Small rediae are active but as they grow and the body fills with thousands of developing cercariae, HORACE W. STUNKARD 2 6 FIGURE 1. Tubulovcsicula pinguis redia from snail collected 20 June 1976, dissected 16 August ; almost exhausted ; the cercaria marked with the arrow is shown in Figure 6. FIGURE 2. Daughter redia, 1.14 mm long, containing six daughter rediae. FIGURE 3. Young cercaria, 0.09 mm long : beginning of tripartite form. FIGURE 4. Larger cercaria, 0.12 mm long. FIGURE 5. Larger ccrcaria; body 0.077 mm, cyst 0.077 mm, tail 148 mm long. FIGURE 6. Cercaria in redia (Fig. 1) ; body 0.045 mm, cyst 0.045 mm, tail 0.081 mm long. motility is restricted and large rediae slowly change shape and are unable to move. The pharynx is seldom visible in large gravid specimens, but in smaller individuals it is sometimes discernible, measuring 0.06-0.08 mm in diameter. There is a LIFE CYCLE OF TUBVLOVESICULA PINGUIS 745 birth pore near the pharynx and a cercaria was seen to emerge tail foremost. Germ-balls first appear as small clusters of deeply staining cells, embedded in or attached to the parenchyma! layer of the body wall, or free in the body cavity. Cercariae (Figs. 3-7) The germ-balls in the cavity of the redia are spherical to oval. As they grow, they become ovate. The narrower end elongates and two constrictions pro- duce a tripartite form (Figs. 3-6). The larger part becomes the body of the cercaria, the medial section becomes the cyst, and the distal part becomes the tail of the cercaria. The body does not increase greatly in size; greater growth occurs in the medial and distal sections. At this stage (Fig. 6), the cercariae become motile in the redia. The cells in the body stain deeply ; as development proceeds the cells in the medial section enlarge, no longer stain, and collapse, producing the cavity of the cyst and forming the delivery tube. As the cavity forms in the cyst, a strand of tissue extends from the caudal tip of the body to the tail. In encystment, it appears that this strand draws the body into the cyst, which closes over it. In the cyst, the body is bent in an inverted U-shape (Fig. 7), with the ventral side internal. The cyst is oval to pyriform, 0.05-0.055 mm in diameter with a thin-walled bulge on one side near the attachment of the tail. The tail is about 0.20 mm long and 0.015 to 0.020 mm wide at the base. A small area in the cyst between the body and tail contains what appears to be a coiled filament, but is actually the collapsed delivery tube. Fixed, stained, and mounted speci- mens are slightly smaller. Cercariae killed in hot water and fixed with AFAG or Duboscq-Bresil solutions have the following average measurements : diameter of the cyst, almost spherical, 0.045 mm with a conspicuous posterolateral bulge; tail 0.15 mm long and 0.012 mm wide at the base. Living specimens under cover- glass pressure protruded the thin-walled posterolateral bulge of the cyst, adjacent to the attachment of the tail. A double-walled sac-like structure was everted and the body slowly passed into it (Fig. 8). The wall of the delivery tube ap- peared soft and flexible as the body of the cercaria emerged from the cyst and passed through the tube. In the tube wall, about 0.05-0.06 mm from the cyst, there was a ring of what appeared to be large vesicles, whose function is unclear. The delivery tube (Figs. 8, 13, 14) was about 0.20 mm long and 0.01-0.015 mm wide. Its distal tip was somewhat narrower than the rest. The tube was narrower and more constricted after the cercaria had emerged. The body was long and slender while emerging, but when free it became short and broad (Fig. 13). There is a strong tendency for the anterior end to turn ventrad (Fig. 9). Mean- while, the tail remained attached to the empty cyst wall. Under pressure, the cyst wall sometimes ruptured, freeing the cercaria that thus did not pass through the delivery tube (Fig. 14). Fixed, stained, and mounted, emerged cercariae averaged 0.08 by 0.02 mm. The oral sucker was about 0.019 and the acetabulum, 0.022 mm in diameter. Metaccrcariae When cercariae are ingested by copepods, excystment begins in the stomach ; in the intestine the delivery tube pierces the wall and ejects the cercaria into the haemocoele. In the body cavity of copepods, the metacercariae grow rapidly and specimens removed 2 weeks after ingestion (Fig. 10) measured 0.10 by 0.023 mm. During the period from July to September, 1976, dissection of copepods yielded a continuous series of developing stages from 0.10 to 0.30 or 0.40 mm in length. 746 HORACE W. STUNKARD 9 FIGURE 7. Cercaria measured alive; cyst 0.050 by O.OSS mm, tail 0.20 mm long, 0.008 wide at base. Heat killed, in alcohol ; cyst 0.045 mm, tail, 16 mm long. FIGURE 8. Cercaria, under slight pressure, beginning to emerge through the extruded delivery tube. FIGURE 9. Cercaria after emergence: anterior end bent ventral, flattened, 0.11 by 0.035 mm. LIFE CYCLE OF TUBULOVESICULA PINGUIS 747 The reproductive organs may remain in a very immature condition or the gonads may increase substantially in size. Presumably the degree of development is correlated with the time the specimen has been in the copepod. The largest metacercariae from copepods are about half as large and morphologically very similar to juveniles encysted in the stomach wall (Fig. 18) or free in the body cavity of Menidia mcnidia, as reported by Stunkard (1973). If the metacercariae are immature when ingested by M. mcnidia. they emerge from the intestine and invade the tissues where they may be encysted. If encysted, the body grows, the appendix becomes distinct, but the reproductive organs do not develop. A specimen removed from the wall of the stomach of M. mcnidia is shown in Figure 18. If they are not encysted, the metacercariae move about, invade the tissues, and eventually become mature, leaving trails of eggs. If the metacercariae are large and mature when ingested, they bore their way out of the intestine, attach themselves to the liver or one of the large veins on the wall of the intestine, begin to suck blood, and become gravid. It appears that ingestion of blood is correlated with the development of sexual maturity. DISCUSSION The systematic relations of the genera Stomachicola and Tubulovesicula were called in question by the report of Sinclair et al., (1972). If T. pinguis is merely a stage in the life cycle of S. rubeus, the validity and integrity of the genus Tubulovesicula is compromised if not undermined. It is important to recognize that Sinclair et al. studied infections in many species of fishes for 2 years. They noted that the infection is seasonal, that 28 species of fish serve as transfer hosts, that they become infected in spring or early summer, and that growth is continuous. In these small fishes the worms may be encysted or may wander freely in the tissues or in the body cavity. They observed that these postmetacercarial juveniles may be killed by immune reactions of the hosts and become "mummified" in melanoid cysts. Larger piscivorous fishes apparently become infected by ingestion of "transfer hosts." In definitive hosts the worms remain in the stomach and average 11.5 mm in length by midwinter. Specimens 22-25 mm in length may be more than 1 year of age. The only "true" definitive hosts were the American eel, Anguilla rostrata ; tarpon, Megalops atlantica ; lizard- fish, Synodus foetens; and kingfish, Menticirrhus americanus. Supporting their conception of the life cycle, Sinclair et al. quoted an observation of Overstreet (1968), "A low incidence of Stomachicola in small fish (Synodus foetens, lizard- fish) accompanied by a higher incidence in the larger Synodus foetens collected at the same time, suggests that the parasite is acquired from an intermediate host not normally fed on by smaller individuals of S. foetens." Sinclair et al. added, "As well, the southern kingfish, which when small, is one of the more common transfer hosts, carries S. rubea in the stomach when it reaches a larger size (above 30 cm). This dual site for infection in the same host species depending on size of the host would suggest that residence in a transfer host is a necessary part of the life-cycle of S. rubea, as it is unlikely that the piscivorous definitive hosts acquire it from the metacercarial host (most likely a small crustacean)." FIGURE 10. Metacercaria, 2 weeks in haemocoele of Acartia tonsa; 0.10 by 0.028 mm. FIGURE 11. Metacercaria from A. tonsa, natural infection; 0.32 mm long, acetabulum 0.10 mm, oral sucker 0.032 mm, testes 0.07 mm in diameter. 748 I HORACE W. STUNKARD « • M ;&*%*• ~~ £f* ^ •** * ^ '*+ v 15 18 ;\ 13 16 14 17 FIGURE 12. Redia taken 25 June 1979; body wall ruptured with emission of cercariae. FIGURE 13. Empty cyst and delivery tube, and cercaria that had emerged through the tube. LIFE CYCLE OF TUBULOVESICULA PINGUIS 740 There is much evidence to support the concept of Sinclair ct al. that typically. S. rubeus is acquired by small transfer fishes that feed on plankton and in which it develops to a condition in which it may remain in the stomach of a larger predator. But the postulate that T. pinyuis is a stage in the life cycle of S. rubcus has not been demonstrated. The problem concerning the relationship of Stomachicola and Tnbitloz'csicnla may be resolved by a study of the comparative morphology of the two genera. Linton (1905) described specimens collected in July and August. 1901 and 1902. at Beaufort, North Carolina, as Distomum tornatitm Rudolphi. Mature worms 8-22 mm in length were found in the stomach of Coryphacna cqnisctis, Coryphaena hip- purus, Menticirrhus americanus, and Synod -its foetens, and juvenile specimens. 6 mm long, were removed from cysts in the intestine of Tylosurus marimis. Linton (1910) described Dinurus rubcus from the stomach of Lycodontis funebris and Lycodontis moringa at Tortugas. Florida. The worms (his figs. 149-154) mea- sured 5-25 mm in length. Alanter (1931) described Dinurus uiagnus from two mature specimens (Figs. 24, 25) from the stomach of the lizardfish, -5". joetens, and immature worms from the stomach and air bladder of spotted trout, Cynoscion nubitlosus, taken at Beaufort, North Carolina. The adult worms were 11 and 22 mm in length and one of the juveniles was 6.5 mm long. He noted the resemblance to D. rubcus and included the specimens allocated by Linton (1905) to D. tornatuiu in D. uiagnus. Alanter (1947) transferred D. rubens and D. uiagnus to Stomach- icola. Sinclair ct al. (1972) suppressed 6". uiagnus as a synonym of S. rubeus. Stomachicola is known by only two species, 6". muraenesocis Yamaguti, 1934 from the Sea of Japan and ^. rubeus (Linton, 1901) Manter, 1947 from the south Atlantic coast of North America. The morphology of the two species is well described and illustrated by adequate figures. Tiibiilovcsicula is represented by several species, although the validity of certain of them is equivocal. The morphology of the adult stage is represented by adequate descriptions of different species, including the account by Stunkard (1973) of development of T. pingnis in Menidia menidia. The discovery of the asexual generations and life cycle, described in the present report, presents additional infor- mation to support the integrity of the generic concept. The internal organization of Stomachicola and Tubulovesicula is similar. The testes are situated immedi- ately posterior to the acetabulum; the seminal vesicle is anterior to the testes; the pars prostatica is followed by an ejaculatory duct that joins the end of the metraterm and both are included in a muscular sac that opens into the genital atrium, situated ventral to the pharynx. The ovary is post-testicular ; there is a seminal receptacle but no Laurer's canal ; typically there are seven (3 + 4 ) cylindrical vitelline lobes ; the uterus passes posteriad for a short distance and then forward to join the metraterm. The major differences between the genera are in overall morphology. In Stomachicola, the body is slender, 8-25 mm long ; the acetabulum is about its diam- eter from the oral sucker; there is a constriction of the body in the postovarian FIGURE 14. Cyst ruptured by pressure, with extrusion of delivery tube and release of cer- caria, but not through the delivery tube. FIGURE 15. Two small sporocysts, 0.059 mm in diameter, from snail fed eggs of T. pingnis. FIGURE 16. Sporocyst 0.16 mm long in snail tissue, experimental infection. FIGURE 17. Metacercarcia, natural infection from A. tonsa, 0.33 mm long. FIGURE 18. Metacercaria removed from cyst in stomach wall of Menidia menidia, 0.50 mm long. HORACE W. STUNKARD region and an enormous contractile tail, not a retractile ecsoma. The gonads are in the anterior one-sixth to one-fourth of the body length, and the uterine coils do not extend into the posterior half of the tail. In Tubitlovesicula, the body is fusiform, 3—4.5 mm long ; the ovary is near midbody, immediately posterior to the testes ; and the uterus extends into the ecsoma which is one-fourth to one-half the body length. The most conspicuous differences between the genera are in the location of the gonads and in the elongate body and enormous tail of Stomachicola. The morphological and bionomic differences between 5". rubeus and T. pinguis clearly demonstrate generic distinctions and controvert the proposal that T. pinguis is a paratenic stage in the life cycle of S. rubeus. ACKNOWLEDGMENT Grateful acknowledgment is accorded Mr. Michael DiSpezio, a student in the Boston University Marine Program, for assistance in the collection and care of copepods. LITERATURE CITED CORKUM, K. C., 1959. Some trematode parasites of fishes from the Mississippi Gulf coast. Proc. La. Acad. Sd., 22: 17-29. CORK CM, K. C., 1966. The digenetic trematodes of some flatfishes from Barataria Bay, Louisiana. Proc. La. Acad. Sci., 29: 45-51. HCNNINEN, A. V., AND R. M. CABLE, 1943. The life-history of Lccithaster confusus Odhner (Trematoda: Hemiuridae) . J. Parasitol., 29: 71-79. LINTON, E., 1905. Parasites of fishes of Beaufort, North Carolina. Bull. Bur. Fish., 24: 323- 428. (1904). LINTON, E., 1910. Helminth fauna of the Dry Tortugas. II. Trematodes. Carnegie Inst. Wash., Pnbl. No. 133. Papers Tortugas Laboratory, 4: 11-98. LINTON, E., 1940. Trematodes from fishes mainly from the Woods Hole region. Proc. U. S. Nat. Museum, 88: 1-172 McCACLEY, J. E., 1960. Some hemiurid trematodes of Oregon marine fishes. /. Parasitol., 46 : 84-89. MANTF.R, H. W., 1931. Some digenetic trematodes of marine fishes of Beaufort, North Carolina. Parasitology, 23: 396-411. MANTER, H. W., 1940a. Digenetic trematodes of fishes from the Galapagos Islands and the neighboring Pacific. Rep. Allan Hancock Pacij. Expcd., 2(14) : 325-497. MANTER, H. W., 19405. The geographic distribution of digenetic trematodes of marine fishes of the tropical American Pacific. Rep. Allan Hancock Pacij. Expcd., 2(16) : 531-547. MANTER, H. W., 1947. The digenetic trematodes of marine fishes of Tortugas, Florida. Am. Midi. Nat., 38 : 257-416. MANTER, H. W., 1954. Some digenetic trematodes from fishes of New Zealand. Trans. R. Soc., N. Z., 82(2) : 475-568. PARCKHIN, A. M., 1969. Trematode family Dinuridae Skrjabin and Guschanskaia, 1954, found in sea-snake from Tonking Bay of North Vietnam. Uchen. Zapiski, No. 99, Biol. Sciences Ser., pp. 26-28. OVERSTREET, R. M., 1968. Parasites of the inshore lizardfish, Synodus foetcns, from south Florida, including a description of a new genus of Cestoda. Bull. Marine Sci., 18 : 444-470. SINCLAIR, N. R., F. G. SMITH, AND J. J. SULLIVAN, 1972. The Stomachicola rubea: Tubulovcsicola pinguis enigma. Proc. Helm. Soc. Wash., 39: 253-258. SKRJABIN, K. I., AND L. K. GUSCHANSKAJA, Eds., 1954. Trematodes of Animals and Man, Vol. 9. Moscow. 344 pp. (In Russian) SOGANDARES-BERNAL, F. 1959. Digenetic trematodes of marine fishes from the Gulf of Panama and Rimini, British West Indies. Tulanc Studies Zool., 7: 69-117. STUNKARD, H. W., 1973. Observations on Tubulovcsicula pinguis (Linton, 1910) Manter, 1947 and on the systematics of the hemiurid trematodes. Biol. Bull., 145 : 607-626. STUNKARD, H. W., 1976. A new cystophorous cercaria from Nassarius trivitattus. Biol. Bull., 151: 433 (Abs.) LIFE CYCLE OF TUBULOVESICU LA PINGUIS 751 STUNKAKU, H. W.. 1°KO. The morphology, life-history, and taxonomic relations of Lepocreadutm areolatutn (Linton, 1()00) Stunkanl, 1%<) ( Trematoda : Digenea ) . Biol. Bull.. 158: 154-163. YAMAGUTI, S., 1934. Studies on the helniinth fauna of Japan. I'art 2. Trematodes of fishes. Jpn. J. Zuol., 5: 249-541. YAMAGTTI, S., 1971. Svn«[>sis «/ (lif/cuctic trctmitmics nj I'crtchrntcs. 2 Vols. Vol. I. 1074 pp., Vol. II. 349 plates. 1794 figs. ZHTKOV, E. V., 1969. Endoparasitic worms of fishes from the sea of Japan and South Kevule Shoal. 7>. /?,»,./. hist. Akad. Nauk SSR.. 28: 1-146. Reference: Biol. Bull., 159: 752-759. (December, 1980) GAMMA RADIATION AND HYDRANTH LONGEVITY IN CAMPANULARIA FLEXUOSA : AGE-DEPENDENCY OF DOSE-RESPONSE FUNCTION JEROME F. WERMUTH Department of Biology, Purdue Utm'crsity Calumet, Hammond. Indiana 46323 ABSTRACT Colonies of Campanularia flexuosa were subjected to doses of gamma radiation of 0, 16, 24, 40, and 80 Krads. Observations were made on the longevity of hydranths relative to their age on the day of irradiation, and to the dose of radiation applied. The average life span of hydranths increased with increasing radiation dose. The increase was greater for older hydranths than for younger ones. It is suggested that the delay in hydranth regression does not approximate the aging process in higher animals. Irradiated hydranth positions do not regenerate; irradiated whole colonies eventually die. The observations on hydranth longevity are discussed in terms of radiation damage to the DNA of developing hydranths of Campanularia flexuosa. INTRODUCTION Study of the effects of ionizing radiation on growth and development of cnidarians seems to have begun with the observation by Puckett (1935) that regen- eration of hydranths in colonies of Pennaria tiarella was suppressed by an x-ray dose of 10,000 R. Daniel and Park (1951, 1953) produced damage to the tentacles of the common brown hydra at an x-ray dose of 9600 R. Wermuth and Barnes (1967, 1968) showed an increase in stolon growth of the colonial hydroid Campanularia fle.vuosa when starting material of new colonies received an x-ray dose of 81,000 R on the fifth day after colony initiation. Brock and Strehler (1963) and Strehler (1964) demonstrated what they con- sidered an increase in average life span of hydranths of Campanularia flcxuosa with x-ray doses of 500-210,000 R applied to 10-day-old colonies. This phenomenon, more properly considered as radiation-induced delay or disruption of the normal regression-replacement cycle of hydranths of this species, was confirmed by Wermuth (1968) for x-irradiation, and by Wermuth and Barnes (1973, 1975) for gamma-irradiated Campanularia flcxuosa hydranths. Wermuth and Barnes (1975) monitored several other post-irradiation growth functions in this species : average amount of new stolon material added per colony, average number of new hydranth positions added to colony starting material, average number of new uprights added to new stolons, and average number of hydranth positions added to uprights of new stolons. All decreased with increasing gamma-radiation doses. Implicit but not stated in any of the above papers is the deleterious effect of ionizing radiation on entire colonies of Campanularia flexuosa. The author of this paper found that doses greater than 16 Krad resulted in colony death, even though hydranth regression was delayed. Received August 20, 1979; accepted August 19, 1980. 752 HYDRANTH LONGEVITY AND 7 RADIATION 753 Campanularia fle.vuosa hydranths normally undergo a regression-replacement cycle (Crowell, 1953) whose duration is a function (among other things) of the temperature at which the colonies are raised. Generally, the warmer the temper- ature, the shorter the length of time between the initiation of a hydranth and its regression, followed by regeneration of a new hydranth at that position. Studies by Brock and Strehler (1963) and Strehler (1964) represent dose- response studies on a single growth function in Campanularia flexuosa, but the data were not presented as dose-response curves. These studies did not consider the effect that the age of the hydranth may have on its susceptibility to delayed regres- sion with increasing dosage of x-irradiation. The present paper shows that age at the time of irradiation does affect the increase in delay of regression. The data also suggest that the increase in delay of hydranth regression may well be the result of interference with several biological operations. MATERIALS AND METHODS Experimental colonies of Campanularia flexuosa were initiated as described in Wermuth and Barnes ( 1969) . Stock colonies from which experimental colonies were derived were obtained from Dr. Sears Crowell, Department of Biology, Indiana University, Bloomington, Indiana. The stock colonies (strains E and Q) used by the author were originally collected by Dr. Crowell and Dr. Charles Wyttenbach near Woods Hole, Massachusetts. Each experimental group consisted of 15 colonies. Three colonies were cultured on single microscope slides, with five slides per experimental group. Each group of five slides was kept in a 10-slide tissue-staining rack. Groups of colonies were maintained at constant temperature in Instant Ocean medium either in a 20 gal Instant Ocean aquarium or in 200 ml tissue-staining dishes in a Hot Pack environmental chamber on a shaker table. The medium of colonies maintained in tissue-staining dishes was replaced every other day. Colonies were fed twice daily by placing the staining racks in a beaker containing freshly hatched brine shrimp (Artemia) nauplii. The nauplii were filtered and washed once in Instant Ocean medium before resuspension in fresh Instant Ocean medium for feeding purposes. Hydranth developmental stages were observed daily. The ability to capture nauplii was used to determine whether a polyp was a mature hydranth, or had not regressed. This ability was determined in doubtful cases by gently squirting a stream of brine shrimp nauplii past the polyps' tentacles. Capture of at least one brine shrimp by a tentacle was considered sufficient to determine a polyp as a mature non-regressed hydranth. No attempt was made to select equal sized groups of hydranths for statistical purposes. All irradiated hydranths were included in the study. Data presented in Figure 1, and Tables I, II, and III were collected only from polyps produced by new stolon growth from starting material. Data from starter material polyps are not included. To equalize ages, a hydranth was considered to have an age of +1 days if it acquired the ability to feed on the day of irradiation. A polyp of age —2 days did not become a feeding hydranth until 2 days after the day of irradiation. A polyp of +3 days began feeding 3 days before the day of irradiation. Thus, the data contained in Figure 1, and Tables I, II, and III are reported as age in days, as mature polyp, on the day of irradiation. Gamma irradiation of colonies was accomplished by placing the colonies and tissue staining racks in tissue staining dishes containing 200 ml of fresh Instant 754 JEROME F. WERMUTH u s o • *•» •"s -a a S S 13 1 ' So •J8 a r*i CN •* •*' o ro OO CN r*5 O CN ^J1 ** O "3 ' ~ -H - -H - -u CN n CN n cd u u "8 q ^^ sO r^ ^> t^ CN O 00 CN rt •a ~ 2 ~^ 4i -f ON ^ ^ -H ^*^ ^^ ^^ :< -H ** CN O c o en a so 10 W) lO t^- CN r^ ro 0 J*1 *"* u") ro o r-i fN O -i< so' CN -f C -H' •—I t-~ o a + ~ -H ^ -H ~ -H CN II ^^ u rt CN <- r^ 10 ror- q i-~ E en 7 1 - -H H -H "" -H CN n be rt •4-1 C r^ sO ^0 r^ O q O rt <~O CN O •^< CN O f} ^ O CN so' O fO O O -o ~ -H ^ -H rt -H ^ -H CN II I ^^ ^2 OO O t- ^ CN ^H O O O CN O CN ^H O •rf fO O "". — (• — r-O f-' ~H -~ -H -~-H -•H ~ -H H -H „, t-» *-H <^ SO CN O 1 -. OO r*i «-< *-O ^O C*>1 1 ~ -H ^ -H ^"^ - -n " -H q ir) r^. t^ "0 ^O o fO CN O CN >-< C5 0 f»3 «/5 ~H 1 - -H ' -H - -H 8 en en en S*» en en ^ en en en ^> o -a c rt c".o ^-8 2 c'-- S^g rt c".2 III rt c" .2 _o rt tg"f t|| til r* f/J K I— I —i j. "° a .2 •O rt rt •3 •^ o> o> •" 4) 4) •^ OJ 4J •*•" O) CU "^ ijj (U rt ^0 it: "^ 'o S "Q ••• ^o S O 1 ""1 M T, "° •O "O TJ "U St llr Lri ^ & UN .a 0. 3 _^ DO rt jgjgj -0 -0 rt -o 5 E S c v 3 > 3 ilifl 1|§1 lljfl O * /^ <• ur> i^ RADIATION 755 Ocean medium, and placing the tissue staining dishes at set distance from a Co60 source. The colony-containing slides were moved to the five slide positions in the tissue staining rack closest to the radiation source for the period of irradiation, and were moved to the other rack positions post-irradiation. The radiation source consisted of a series of Co60 rods which could be raised or lowered from a well located in the center of the floor of the radiation room, which was approximately 4x4 m. Colonies to he irradiated were placed on the floor adjoining the source rods; control colonies were kept in the adjoining control room, from which the experimenter could raise or lower Co60 rods by remote control. Radiation was delivered in a single 80 min dose. The temperature of the Instant Ocean medium surrounding non-irradiated control colonies rose no more than 1°-2°C during the course of the irradiation ; the temperature of the medium surrounding irradiated colonies rose no more than 2°-3°C during that same time. Immediately post- irradiation, all colonies were returned to the aquarium or to the environmental chamber. Colonies returned to the chamber were placed in fresh Instant Ocean medium. Gamma-radiation dosage was monitored by the HE-6 dosimetry method ( McLaughlin et al., 1971). Vials of the dosimetry reagent were placed alongside the staining racks containing the colony-bearing slides at the level of the slides. RESULTS The following data were recorded from a cloned strain (designated E) of Campanularia flexnosa. All experimental colonies were derived from stock ( Ii .IIN, I' , , / ,11 , \ir im mlu , mr lipnl'i i .iK mm lonophoi, . \ i i imp. H i .on, i-m ploying nrwnd •• ill m iipo-.omi -i, in i \\ . . n known loiiophon • > ^ ' ' ' ' - i « n o m \ 1 1 n i .mil pin i-.piii iiiph I-., i. it i \ .n ii !•.. proHi i Hi M. i . .mil irl moii I-., I /'' \M«li;.i>\. Knmiii ',. I Ii mnl \ .nr. .iml llcm. i^-lnl m i ir. m ilu .••.!, mm Mm, I ol .1 p. .1 \ i h.irl r . i n iii 1 1 1 1 i ./>..• ,i i//, 'if , i id Ii 1,1 Iii . ' ,' ' \ NIH i; .1 IN, ', MM, -ice I i .im i-. \\ K lot .-. l''d \nKlc.li-.ll, l*lH, 17 I. 176 \nim.il I'on.nl.il \.iii.ilniii 111 ,..i mi Inn-, ol I In- i H ilri -. T i Imml Inn miil.i .in. I I i In m M, I i / 'S \min.il 1 1 pi M I ol I In M .n im I in i|. H- n .1 1 I .1 1 mi .1 I in \ , I \ ii I 1 1 01 m 1 1 ir.,111, ili i I i n .1 1 .n I i \ 1 1 1 1 . in .nl . I nil .n ill. n 1 1 -y ion, ' / d \ p| .In .1 1 ion o I .1 pol \ pi i M. I m 1 1 I, i i I n i Ii m. 1 1 .ni.il\ -.r. in \ i-iii'/-ii •> i \ i I . ' \. |ii.n mm, m. mil i n. im i ol -.i|iiiil m, U '' \il;i, in. Id/, ld,H, Id'), I/ t, I/ I, l/d. I')', III i iii \\ii I .im.n i l> I \i i li.in i^.r. \l i- me m I n .1 nc Iipi. I . i .1 Ii I mn n moplioi i . \ Am li. I I > . -..-<• I I' S< him, -I. Inl \li-.li.nl-. ol p.ipci-. pii'.intcil .il I he I'liiii.il ••> icill ilu inrrlim;-. ol ilu M.iimc (tin 1 0^1 1 .il I .il HM .il m \ . \m;n-.l 'I ' *. I ''SO. Ill I.. :•//,; t,'i,\,i, I 01 m.i I ion ol llllCCOUM I., ill. >!'' \i I ol ll.it; I.I I i r-.p. Mi-.r in -.c.i . incnii nn ••.. I I / \, IIN. II' \n Mi. ( hi i i . -.ir M N.ili.ii , i /i. \|l| I M VN. \\ II I MM | . | H ('/ ill , S(|||ii I .1 xon siiilmm ili.umcl-. .lie blocked ii\ri-.il.l\ li\ .iph.inl ox m iliiivi.l limn .1 piok.iixoli, .ilv.. ir. I /S •„-,- M.III | lloili;,-, I /I \l>INil\|(l. < >\ I \\OII t'l ,ll , ( II. 111^.1- III (\illi ii IK Irol 11 Ii- i ( ml cnl cluriHU e.ci mm. 1 1 \c-.iili In c.iKiInu n m SfHSHld nn<\li-'.. loo \|)|' i il MI'. \ l.il ion in mn In r.nl.ilcd hoin 1 1 ill ci cnl cm I n \ i mn -.1 .1 i'i -. ol I i /'iii iii , I / I lii'liiliii f'ii f'lllinii I'.ln^.), IH(I i.i i n m m i In- keyhole liopoil.i I r,-. mclli. Lie) I. .r.c. I on -.lull ^•1 nu I Ii i m^.'., nOd Xl'.l'.1 ' '•'•! \ ' licll.l\lol .iinon^; '.c.i .incmonc'.. I I/ \r.ni(.;. m ( ii in f-ii n nl,i i hi tlf\ iii>\,i , i '\ ' \i m ii 1 1 , K S , sec S. It Hinwii. I1'1' ,- 1 ,ll , ( olol 1 1". -. pi ol c lie. III pi IS ( ill n 1 1'. I .mil |i.| < .IM'.I I., I In .1 r | I n|ir/ ll.il IK n, IS'/ , . I',, I \l ni. ion, IHK .1 . i-1 r '.\\ i n .mi, i'j t I mn I i nn. 1 1 -.ij;nih( .UK i i .1 I In Innnlc-. Alir.i >| n. |i mr. M , \ '.n I . M . I > Will I \ M • m llic '..mil ilnll.il, Mrllitti ,/llllli/ltli"~f'i-l An:. m n, I »i \ i Ii i| nnciil i il ilu iili.ilnn pill, in mi lln inililV" "'I lln iii|iinl, I .'hi'" f'<;ilrl A iK.ininnr iliilnm nnin. MOpQ Study, I"' Inliltil. '>dl Air.. ic, -l.Sd. .17 K, I''M, MM), Mil .iml Cassiopeia androtntda, (''i A I Ki i'., I ) \'J|| I , I ..nl liipnil I Id'; ol . i •.•...( i.i A HUNS I I r,. I >l \:n , .. i I', cil h ' l<-\-.,,n. Id', Mil I I\T li .n >•. SO.S '.i ' ii,i/ii I .' '.i, III i c\ i-.il1; I m i j' I. c.i i n^ ml i .n ( Mill il i ,il. Him ii.ni'.H nl lolloWmB IMP " Imn poll nli.il in .ipinl , i M,I pi «•'. \ n. i pi n Icimm.il, I'M \ ,i lil iii i i'i iilni/i . , I . I. M n I pIllNIHH, d'.d \ .. i. hin, .i-c './\'(7(; inniil, 1 1 yi n :i .. I' niln, 1,1 .; ll,i \iiiNnr, (,iiiin\, ANI» NANCY I >i .1 rnml preference! nnd popul&tion > Is n.imn . .1 i In- impliip..,! I'llliili llr;tlil loill'l, HI III:, 1 , - .1 inl'V'li'iiii lil.r.l (inn H •., I1' t \\ll Ml -, \. SlIONAN. HIT I . ili. in Mm I. 7 ' \mmii ,i(|(|-., m (il\'ii-ni .il V.llinil'i ';.ilmili. 626 Ilinlll, < rir.l\, I'vilin \lnli, in /. i i i-ntlotli-\, '•\ 773 774 INDEX TO VOLUME 159 Ascidian blood chemistry, 656, 669 Aspects of the life history of Carcinonemertes errans (Nemertea: Carcinonemertidae), an egg predator of the crab Cancer magister, 247 Associative learning, in gastropods, 505 Asterias forbesi, see starfish ATEMA, JELLE, see Charles Derby, 450 Axons, 471, 478, 481, 483, 484, 485, 487, 488, 491, 494 B Bacteria, 456, 457, 459-463, 465 Bacterial epiphytes on Zostera marina surfaces, 461 Bacterial kidney disease of rainbow trout in sea water, 495 BALLINGER, D., AND T. HUNT, Further ob- servations on the phosphorylation of ribosomal proteins after fertilization of Arbacia punctulata eggs, 473 BANG, BETSY G., AND FREDERIK B. BANG, The urn cell complex of Sipunculus nudus: A model for study of mucus- stimulating substances, 571 AND S. J. COOPERSTEIN. Janus green B: A vital stain during the process of mucus secretion, 447 BANG, FREDERIK B., see Betsy G. Bang, 571 Inflammation in the sea star, Asterias forbesi, 495 BANNON, GARY A., AND GEORGE GORDON BROWN, Ultrastructural characteristics of the non-expanded and expanded extra- embryonic shell of the horseshoe crab, Limulns polyphemus L., 582 BARLOW, ROBERT B., JR., see Ehud Kaplan, 486 see Leonard Kass, 487 see Takehiko Saito, 490 BARNES, EDWARD, AND IZJA LEDERHENDLER, Dark-adaptation effects on photobehavior of Hermissenda crassicornis (Gastropoda : nudibranchia), 479 Barokinesis, in crab Callinectes sapidus larvae, 402 BARTELT, DIANA C, AND WILLIAM D. COHEN, Labile components of the dogfish eryth- rocyte cytoskeleton, 441 Bathynomus giganteus (sea roach), 478 BATRA, RANJAN, see Ehud Kaplan, 486 BAUER, G. ERIC, et al., Subcellular localiza- tion and characterization of islet hor- mone-degrading enzymes in anglerfish islet tissue, 473 see Bryan D. Noe, 476 BAUMGOLD, JESSE, AND PAUL GALLANT, Re- lease of cytoskeletal proteins into the perfusate of squid giant axons, 441 Behavioral basis of larval recruitment in the crab Callinectes sapidus Rathbun : A labo- ratory investigation of ontogenetic changes in geotaxis and barokinesis, 402 BELL, WAYNE H., Microbial ecology of algal extra-cellular products: the specificity of alga-bacterial interactions, 456 BENNETT, M. V. L., see J. H. Stern, 493 BERG, CARL J., see Gregory A. Tracey, 465 Better fluorescent probes for optical measure- ment of changes in membrane potential, 484 BEZANILLA, F., see B. M. Salzberg, 491 BIGGER, CHARLES H., Interspecific and intra- specific acrorhagial aggressive behavior among sea anemones: A recognition of self- and not-self, 117 BIGGS, WILTON R., see Clifford ]. Hawkins, 656 Bile pigments, 499, 501 Binding of dynein to isolated meiotic spindles of the surf clam, Spisula solidissima, 446 Biochemistry, abstracts, 473 Bioluminescence, see luminescence Birefringence response of voltage-clamped in- ternally perfused axons, 491 Blood chemistry of ascidians, 656, 669 Blue crab, see Callinectes sapidus BOLSOVER, S. R., AND J. E. BROWN, Guanosine phosphates and the control of membrane potential in Limulus ventral photorecep- tors, 480 BORGESE, JOAN M., see Thomas A. Borgese BORGESE, THOMAS A., et al., Thermostability of fish hemoglobins, 448 BORGESON, WILL, see Keith Nelson, 162, 503 BOUSOUETTE, GEORGE D., The larval develop- ment of Pinnixa longipes (Lockington, 1877) (Brachyura: Pinnotheridae) reared in the laboratory, 592 BOWLES, FRANCIS, see Amy Friedlander, 458 see David W. Juers, 461 BOYER, BARBARA, Mosaic development in the polyclad turbellarian Haploplana inquilina and its evolutionary implications, 448 BOYLE, M. B., AND L. B. COHEN, Preliminary evidence for taste-aversion learning in the nudibranch mollusc Aeolidia papillosa, 480 BRAITHWAITE, LEE F., see Criag M. Young, 428 BRANDL, HELMUT, et al., Studies of methano- genic bacteria from intestinal tracts of marine fishes, 456 BRENCHLEY, G. A., Distribution and migra- tory behavior of Ilyanassa obsoleta in Barnstable Harbor, 457 BRENOWITZ, M., see K. E. Van Holde, 478 BRETOS, MARTA, Age determinations in the keyhole limpet Fissurella crassa, 606 BROWN, GEORGE GORDON, see Gary A. Bannon, 582 BROWN, JAY C., see Lucio Cariello, 467 BROWN, J. E., see S. R. Bolsover, 480 INDEX TO VOLUME 159 BROWN, S. B., el al., Phycocyanobilin syn- thesis from exogenous heme, 499 see K. M. Smith, 501 see R. F. Troxler, 502 BRUCE, VICTOR G., see Judith E. Goodenough, 649 Budding, of Cassiopeia andromeda, effect of symbionts, 394 Bufo, neuroplasmic lattice, 471 BURGER, MAX M., see Werner Burkart, 449 BURGOS, MARIO H., et al., Gossypol inhibits motility of Arbacia sperm, 467 see Michael Coburn, 468 BURKART, WERNER, AND MAX M. BURGER, Substitution of calcium by polycations in sponge aggregation factor interaction, 449 Burrowing sponge, carbonic anhydrase im- plicated in mechanism of penetration, 135 BUSA, W. B., et al.. Light-scattering studies of sheared and unsheared actin polymeriza- tion, 442 Cadmium resistance in bacteria from the Great Sippewissett Marsh, 463 Caffeine, effects on Chlamydomonas reinhardii, 649 Calcium, 449, 470, 480, 481, 485, 491, 494 and potassium activities in the hemolymph of the squid, Loligo pealei, 492 Callincctes sapidus, behavioral basis of larval recruitment, 402 mandibular gland, 760 Calmodulin from the axoplasm of the squid, 485 Campanularia flexuosa, 752 Cancer magister, predation on eggs by Car- cinonemertes errans, 247 Cape Cod National Seashore, 466 Carbonate substrata, penetration of by bur- rowing sponge, 135 Carbon fixation, by Prochloron isolated from Diplosoma virens, 636 Carbonic anhydrase, in Cliona celata burrowing, 135 Carcinonemertes errans, crab-egg predator, life history, 247 CARIELLO, Lucio, et al., Properties of isolated fertilization envelopes from Arbacia punc- tulata, 467 CARLISLE, JAMES C., AND JAN SPITSBERGEN, Bacterial kidney diseases of rainbow trout in sea water, 495 CARTER, R., see D. Wirth, 498 Cartesian diver, 692 CASE, JAMES F., Courting behavior in a syn- chronously flashing aggregative firefly, Pteroptyx tener, 613 see John A. Warner, 231 Cassiopeia andromeda, budding and strobila- tion, 394 CAVANAUGH, COI.LKKN M., Symbiosis of chemo- autotrophic bacteria and marine in- vertebrates, 456 Cell membrane, 471 Cell motility and cytoskeleton, abstracts, 441 Central projections of the lateral eyes of Balunus nubilis (Pacific giant barnacle) differ from those of the median eye, 483 CHAMBERS, RANDY, AND ANTHONY PIKES, The relationship between diet and growth rate of the grass shrimp Palaemonetes pugio, 458 Change in cyclic nucleotide content during germinal vesicle breakdown in Spisula oocytes, 466 Change in protein synthetic pattern and niRNA population during Spisula em- bryogenesis, 478 CHANG, ERNEST S., see Ashley I. Vudin, 760 Characterization of periodic order in the neuroplasmic lattice of Bufo, Loligo, and Hermissenda axons by a Fourier analytical technique in scanning transmission elec- tron microscopy (STEM), 471 CHARLTON, MILTON P., el al., Presynaptic injection of calcium facilitates transmitter release in the squid giant synapse, 481 see Stephen J Smith, 491 see Robert S. Zucker, 494 Chemical communication in Homarus, 162 Chemical conversion of a chlorophyll a de- rivative to a bile pigment, 501 Chemistry of the blood of the ascidian Podo- clavella moluccensis, 669 Chemoreceptors, on legs of Homarus ameri- cannus, 450 Chemotaxis, 443 CniH-Vi, CHANG, see Mario H. Burgos, 467 Chironomus, 489 Chiton water balance and salinity stress, 364 Chlamydomonas reinhardii, effects of caffeine and theophylline on phototactic response of, 649 Chlorophyll a, 501, 502 Chromium, 462 in ascidian blood, 669 Cilia, 444, 446 pattern, on squid embryo, 102 Ciona, 446 Circadian rhythm of photoreceptor cells in the Limulus lateral eye: further studies, 490 Clams, freshwater, sodium regulation in, 325 CLARK, \V. H., JR., see Ashley I. Vudin, 760 Cliona celata, carbonic anhydrase implicated in mechanism of burrowing, 135 Cloning genes in Leishmania enriettii in E. coh, 498 COBURN, MICHAEL, et al., Oxygen free radical generation by gossypol : a possible mecha- nism of antifertility action in sea urchin sperm, 468 see Peter Sinsheimer, 469 776 INDEX TO VOLUME 159 Coelomic fluid, hemolysins and hemagglutinins in Glycera, 259 Coelomocytes, red, salinity tolerance and volume regulation in Glycera, 626 ultrastructure, of Dermasterias imbricata, 295 COHEN, L. B., see M. B. Boyle, 480 see A. Grinvald, 484 COHEN, WILLIAM D., et a!., The marginal band system of the dogfish erythrocyte, 442 see Diana C. Bartelt, 441 see Iris Nemhauser, 444 Colorless proteins in phycobilisomes of rhodo- phycean and cyanobacterial species, 499 Color, of bodies of fishes in relation to envi- ronmental light, 450 Comparative study of the blood plasma of the ascidians Pyura stolonifera and Ascidia ceratodes, 656 Competition, in Homarus, 162 in sea anemones, 117 in Teredo navalis, 465 resource partitioning in sympatric hermit crabs, 337 Contextual relationships in food webs in- volving meiofauna, 462 Control of drilling fluid discharge from pe- troleum development on Georges Bank, 463 CONWAY, KEVIN, AND R. KEVIN HUNT, Heal- ing of Xenopus eye fragments premarked by chimaeric pigment grafts, 453 see R. Kevin Hunt, 453 see Sarah A. Shoaf, 454 see Robert Tompkins, 455 COOPERSTEIN, S. J., see B. G. Bang, 447 COPELAND, EUGENE, et al., Gas secretion in the swimbladder of shallow water fishes compared to a deep ocean fish (3000 meters), 449 Copepod, formation of siliceous teeth by Acartia tonsa, 349 Labidocera aestiva, control of diapause, 311 Corbicula fluminea, sodium balance, 325 CORR, MICHELLE, AND DAVID HUGHES, The predatory habits of two lycosid spiders in Great Sippewissett Marsh, 456 CORSON, D. WESLEY, AND ALAN FEIN, A GTP binding component regulates discrete wave production in Limulus ventral photorecep- tors: pharmacological evidence, 481 COSTA, CHARLES J., et al., The intracellular mechanism of salinity tolerance in poly- chaetes: Volume regulation by isolated Glycera dibranchiata red coelomocytes, 626 Coupling between horizontal cells in the carp retina examined by diffusion of Lucifer yellow, 486 Courting behavior in a synchronously flashing, aggregative firefly, Pteroptyx tener, 613 Crab, see Callinectes sapidus, Cancer magister, hermit crabs, Panopeus herbstii, Pinnixa, Uca Crangon franciscorum, seasonal abundance and distribution, 177 Crayfish, see Procamberus clarkii Crowding, effects in Homarus, 162 Crystalline axes of the test and spine of the sea urchin Strongylocentrotus purpuratus— a new method of analysis, 472 Ctenophores, 446 CUMMINGS, GLENNA D., see Jay E. Mittenthal, 714 Current, effects on ascidian Styela montereyen- sis, 428 Cyanobacteria, 498-500, 502, 636 Cytoskeleton, 441, 471 CZETO, A. R., et al., An X-ray study of the retinal photoreceptor structure of squid, 482 D DAGGETT, RICHARD, see Keith Nelson, 162 DAN, KATSUMA, see Shinya Inoue, 443 Dark-adaptation effects on photobehavior of Hermissenda crassicornis (Gastropoda : nu- dibranchia), 479 DAVID, JOHN, see D. Wirth, 498 DAVILA, H. V., see B. M. Salzberg, 491 DAWSON, BENJAMIN G., et al., Field experi- ments on electrically evoked feeding re- sponse in the dogfish shark, Mustelus cants, 482 Dendritic spine necks, 470 Denitrification potentials in a successional sequence of northern hardwood forest stands, 464 Density-dependent growth inhibition in lob- sters, Homarus (Decapoda, Nephropidae), 162, 503 DERBY, CHARLES, AND JELLE ATEMA, L-Glu- tamate-specialist chemoreceptors on the legs of the lobster Homarus americanus, 450 DERIEMER, SUSAN A., AND EDUARDO R. MACAGNO, Positional correlation of syn- aptic boutons in pairs of mechanosensory cells in the leech, 482 Dermasterias imbricata, ultrastructure of coelo- mocytes, 295 DESANTIS, ROSARIA, see David Stopak, 446 Developmental analysis of an unusual ho- moeotic mutation, proboscipedia, in Dro- sophila melanogaster, 454 Development of the ciliature pattern on the embryo of the squid, Loligo pealei: A scan- ning electron microscope study, 102 DE WEER, PAUL, see George R. Kracke, 487 Diapause, photoperiod control in Labidocera aestiva, 3 1 1 DIET/, THOMAS H., see Susan McCorkle. 325 INDEX TO VOLUME DIENER, RICHARD A., sec Ashley I. Vudin, 760 DILLAMAK, RICHARD, see Kayo Okazaki. 472 Dinurus, 737 Diplosoma virens, photosynthesis by Prochloron isolated from, 636 DISE, NANCY, see (iloria Allende, 455 Distribution and migratory behavior of Ilya- nassa obsoleta in Barnstable Harbor, 457 Dogfish, 441, 442, 451, 452, 472, 482 Dormancy, of copepod Labidocera aesliva, 311 Dorvteuthis bleekeri, maintenance in aquaria, '319 Drosophila melanogaster, proboscipedia, in, 454 Drug delivery to dogfish lens, 472 DuBois, ARTHUR B., AND CHRISTOPHER S. OGILVY, Thrust and drag of bluefish (Pomatomns saltatrix) at different buoy- ancies, speeds, and swimming angles, 450 see Christopher S. Ogilvy, 451 DUNHAM, P., see P. Anderson, 479 E Ecdysone, and terrestrial isopod Crustacea, 337 Echinoderms, see starfish Echinoida, sea urchins, reproductive cycle, 728 Echinothurioida, sea urchins, reproductive cycle, 728 ECKMAN, STEVE, see R. Kevin Hunt, 453 Ecology, abstracts, 455 Effectiveness of National Park Service policies in protecting barrier island ecosystems within the Cape Cod National Seashore, 466 Effect, of chromium on microbial activity in salt marsh sediments, 462 of cytochalasin B and phalloidin on F-actin and G-actin, 449 of grazing by Uca pugnax on the microbial population in salt marsh sediment, 465 of sewage fertilization on benthic macro- invertebrates in salt marsh creeks, 465 of symbiotic zooxanthellae and temperature on budding and strobilation in Cassiopeia andromeda (Eschscholz), 394 of temperature acclimation on nitrogen metabolism in two littorinid snails, 447 Effects, of caffeine and theophylline on the phototactic thythm of CMamydomonas reinhardii, 649 of closed-culture competitive interactions on growth of Teredo Navalis larvae, 465 of magnetic fields on regeneration in fiddler crabs, 681 of media with low silicic acid concentra- tions on tooth formation in Acartia tonsa Dana (Copepoda, calanoida), 349 of temperature and salinity on the osmotic composition of the southern oyster drill, Thais haemastoma, 148 Electrical, activities in the subtentacular region of the anthomedusan Spirocodon saltatrix (Tilesius), 37<> membrane properties of single and two-cell preparations from Chinniomus salivary gland, 489 Electron spin resonance studies of protein- methylglyoxal complexes, 474 Electrophoretic variation in sympatric mud crabs from North Inlet, South Carolina, 418 Electrophysiology, of Spiricodon saltatrix, 376 Abstracts, of MBL General Meetings, 441 ff. Elimination of synapses from identified lob- ster motor neurons during development, 492 ELLISON, REBECCA P., ADP ribosylation in nuclei isolated from different embryonic stages of Arbacia, 474 Embryogenesis, of squid, 259 Embryonic development, of Limit/us, 582 Entamoeba histolytica, 495, 496 Enzyme activity in early embryos, 475 EPPI, RENE E., see Benjamin G. Dawson, 482 Erythrocytes, 441 Europanoepus depressus, electrophoretic vari- ation in, 418 Evaluation of low light level intensifier vidicon detectors for microscopy, 473 External K+ and Rb+ retarded closing of potassium channels, 493 Extra-embryonic shell of Limulus, 582 Eye, see photoreceptors, lens FARMANFARMAIAN, A., el al., Mechanism of heavy metal inhibition of amino acid transport in the intestine of marine fishes, 458 Feasibility of seaweed aquaculture in the Great Sippewiseett Salt Marsh, 459 Feeding, and induction of bioluminescence in Porychthys notatus, 231 of Neomysis mercedis (Holmes), 193 of Styela montereyensis, 428 FEIN, ALAN, see D. Wesley Corson, 481 Fertilization, abstracts on, 466 ff., 473, 474, 478 Fiddler crabs, see Uca Field experiments on electrically evoked feed- ing response in the dogfish shark, Mustelus canis, 482 Firefly, flashing and courting, 613 FISHER, CHARLES R., JR., AND ROBERT K. TRENCH, In vitro carbon fixtion by Pro- chloron sp. isolated from Diplosoma virens, 636 Fishes, 449-451, 457, 458, 485, 486, 495 see also dogfish, toadfish, Porychthys notatus Fissurella crassa, shell growth rings and age determination, 606 778 INDEX TO VOLUME 159 FOHLMEISTER, JuRGEN F., see William J. Adelman, Jr., 478 Food choice and palatability in a salt marsh detritivore, Melampus bidentatus, 464 Food preferences and population dynamics of the amphipod Talorchestia longicornis, 455 Formation and early differentiation of sea urchin gonads, 280 Fourier analytical technique, characterizing order in neuroplasmic lattice, 471 Freeze-fracturing, 470, 471 FRENCH, C., see D. Wirth, 498 FRENCH, KATHLEEN A., AND ANN E. STUART, The central projections of the lateral eyes of Balanus nubilis (Pacific giant barnacle) differ from those of the median eye, 483 FRIEDLANDER, AMY, et al., Mapping vegeta- tion and topography of Great Sippewissett Salt Marsh, Mass., 458 Frog, see Xenopiis FUJITA, RODNEY M., The feasibility of seaweed aquaculture in the Great Sippewissett Salt Marsh, 459 Functional significance of the lunules in the sand dollar, Mellita quinquiesperforata, 561 Fundulus, 493 Further improvement upon maintenance of adult squid (Doryteuthis bleekeri) in a small circular and closed-system aquarium tank, 319 Further observations on the phosphorylation of ribosomal proteins after fertilization of Arbacia punctulata eggs, 473 FUTRELLE, R. P., The theory of chemotaxis and the ability of a cell to sense its posi- tion and orientation within a tissue, 443 GALE, JUDITH, see Amy Friedlander, 458 GALLANT, PAUL, see Jesse Baumgold, 441 Gamma radiation and hydranth longevity in Campanularia flexuosa: Age-dependency of dose-response function, 752 Ganglion ablation and gonad development in ascidian Symplegma reptans, 219 Gap junctions: quantitative comparison of reduction in conductance by H and by Ca ions in an internally perfused prepara- tion, 493 GARDNER, KAREN, see Judith Megaw, 472 GASCOYNE, PETER, et al., Electron spin reso- nance studies of protein-methylglyoxal complexes, 474 see Ronald Pethig, 477 Gas secretion in the swimbladder of shallow water fishes compared to a deep ocean fish (3000 meters), 449 Gastropod models of associative learning, 505 Genetic differences, in sympatric mud crabs, 418 Genetics, abstracts, 453 Geotaxis, in crab Callinectes sapidus larvae, 402 Germinal vesicle breakdown, 466, 468 GHIOLD, JOE, see David E. Alexander, 561 GILBERT, DANIEL L., see Roderic E. Steele, 492 GILLY, WM. F., AND CLAY M. ARMSTRONG, Interaction of external transition metal ions and the mobile gating charges of Na and K channels in squid axon, 483 GILSON, MICHAEL K., AND AD. J. KALMIJN, Statistical mechanics of geomagnetic orien- tation in sediment bacteria, 459 GITLER, CARLOS, see Eileen Lynch, 496 see Willy F. Piessens, 497 see D. Wirth, 498 L-Glutamate-specialist chemoreceptors on the legs of the lobster Homarus americanus, 450 Glycera, humoral immunity, 259 salinity tolerance, 626 venom and axons, 485 Gonad development and ganglion ablation in ascidian Symplegma reptans, 219 Gonads, annual cycle in Echinoida and Echinothurioida, 728 sea urchin, formation and early differentia- tion, 280 GOODENOUGH, JUDITH E., AND VICTOR G. BRUCE, The effects of caffeine and theo- phylline on the phototactic rhythm of Chlamydomonas reinhardii, 649 GOODMAN, E., see P. Anderson, 479 Gossypol, 467, 468 GOULD, ROBERT M., Phospholipid synthesis in axoplasm from the squid giant fiber, 484 GOVIND, C. K., see Philip J. Stephens, 492 GRAHAM, STEPHEN, et al., Larval settlement on microbial films: A model system, 460 GREENBERG, E. P., see Frederick H. Weber, 465 see Helmut Brandl, 457 GRINVALD, A., et al., Better fluorescent probes for optical measurement of changes in membrane potential, 136 Growth inhibition in lobsters, 162, 503 Growth rings, in Fissurella crassa shell, 606 GTP binding component regulates discrete wave production in Limulus ventral photo- receptors: pharmacological evidence, 481 Guanosine phosphates and the control of membrane potential in Limulus ventral photoreceptors, 480 GUILLARD, ROBERT, R. L., see Charles B. Miller, 349 GUPTA, R., see A. Grinvald, 484 H Habitat, Styela montereyensis, 428 see also ecology HADLOCK, ROBIN P., Alarm response of the intertidal snail Littorina littorea (L.) to INDEX TO VOLUME 159 779 predation by the crab Curchnis maenas (L.), 269 HAEDRICH, RICHARD, see Eugene Copeland, 449 HAIMO, LEAH T., see Bruce R. Telzer, 446 HANNA, ROBERT B., see Bruce L. Kagan, 485 HARRINGTON, JOHN P., see Thomas A. Borgese, 448 HARRIS, A. L., see J. H. Stern, 49,} HARRIS, ANDY, see Eileen Lynch, 496 HARRIS, KRISTEN, Relationships between den- drite and spine neck diameters in freeze- fractured rat hippocampal formation, 470 HASCHEMEYER, AUDREY E. Y., see Rita \V. Mathews, 475 HATCH, WALTER L, The implication of car- bonic anhydrase in the physiological mechanism of penetration of carbonate substrata by the marine burrowing sponge Cliona celata (Demospongiae), 135 HAWKINS, CLIFFORD J., et al., Chemistry of the blood of the ascidian Pododavella rnoluccensis, 669 HAWKINS, CLIFFORD J., et al., Comparative study of the blood plasma of the asci- dians Pyura stolonifera, and Ascidia cera- todes, 656 HEAD, JAMES F., AND BENJAMIN KAMINER, Calmodulin from the axoplasm of the squid, 485 Healing of Xenopus eye fragments premarked by chimaeric pigment grafts, 453 HEDGECOCK, DENNIS, see Keith Nelson, 162 Hemagglutinins in the coelomic fluid of Clycera dibranchiata, 259 Hemocyanin, of sea roach, 478 Hemoglobins, fish thermostability, 448 Hemolysins and hemmagglutinins in the coe- lomic fluid of a polychaete annelid, Glycera dibranchiata, 259 Hermissenda, 471, 479 associative learning in, 505 Hermit crabs, specificity of association with Hydractinia, 337 HEYER, GAIL W., see Benjamin G. Dawson, 482 HILDESHEIM, R., see A. Grinvald, 484 HILDRETH, JANE E., AND WILLIAM B. STICKLE, The effects of temperature and salinity on the osmotic composition of the southern oyster drill, Thais haemastoma, 148 HINEGARDNER, RALPH T., see Margaret S. Houk, 280 HINES, MICHAEL, see John Moore, 488 Histochemical identification of N-acetyl-j3- glucosaminidase activity in early embryos of Lytechinus pictus, 475 HODGE, ALAN J., AND WILLIAM J. ADELMAN, JR., Characterization of periodic order in the neuroplasmic lattice of Bufo, Loligo, and Hermissenda axons by a Fourier analytical technique in scanning trans- mission electron microscopy (STEM), 471 Homarus, see lobster HOMMEL, M., see I). \Virth, 498 Hoploplana inuuilina, mosaic development, 448 HORGAN, ERICH F., AND MARGARET S. RACK, The relationship between organic and nitrogen content of marsh sediments and feeding rates in L'ca pugnux, 460 Hormone-degrading enzymes in anglerfish islet tissues, 473 Horseshoe crab, see Limulus HORWITZ, I. S., see G. Salama, 449 HOUK, MARGARET S., AND RALPH T. H INK- GARDNER, The formation and early dif- ferentiation of sea urchin gonads, 280 HUFNAGEL, LINDA, Oriented particle assem- blies in the plasma membrane of Tetra- hymena: their deployment relative to cell surface topography, cellular morphogene- sis and sensitivity to stimuli, 471 HUGHES, DAVID, AND MICHAEL USEM, Pre- ferred food sources and the limitation of local distributions of the isopod crusta- cean Philoscia vittata, 460 see Michelle Corr, 456 HUNT, R. KEVIN, see Kevin Conway, 453 et al., Pigmentation mosaicism in the choroid of the eye after embryonic grafting, 453 et al., Retinotectal patterns and patterns of genetic mosaicism of ploidy chimerae of Xenopus eye, 454 see Sarah A. Shoaf, 454 see Robert Tompkins, 455 HUNT, T., see D. Ballinger, 473 Hydractinia, symbiosis with hermit crabs, 337 Hydranth longevity and gamma radiation in Campanularia flexuosa, 752 Hydrozoan, see Hydractinia, Spirocodon sal- tatrix Ilyanassa obsoleta, distribution and migration, 457 Immune response, of Dermasterias imbncata, 295 Immunology, abstracts, 495 annelid, 259 Implication of carbonic anhydrase in the physiological mechanism of penetration of carbonate substrata by the marine burrowing sponge Cliona celata (Demo- spongiae), 135 Inflammation in the sea star, Asterias forbesi, 495 Inhibition, of islet prohormone to hormone conversion by incorporation or arginine and lysine analogs, 476 of L-leucine transport in toadfish liver by various natural amino acids, 477 780 INDEX TO VOLUME 159 INDUE, SHINYA, AND KATSUMA DAN, Mitotic spindle behavior in unequal cleavage of Spisula solidissima, 466 INOUYE, H., et al., An X-ray study of fish nerves, 485 Inside-out voltage clamp in the squid giant axon, 487 Interaction, of external transition metal ions and the mobile gating charges of Na and K channels in squid axon, 483 of phalloidin with G- and F-actin, 445 Interactions between two species of littorines — Littorina littorea and L. Saxatilis — along New England shores, 463 Interspecific and intraspecific acrorhagial ag- gressive behavior among sea anemones : A recognition of self- and not-self, 117 Interstitial fluid pressures of smooth dogfish (Mustelus canis) and bluefish (Pomatomus saltatrix) tilted in air, 451 Intracellular mechanism of salinity tolerance in polychaetes: Volume regulation by isolated Glycera dibranchiata red coelamo- cytes, 626 In vitro carbon fixation by Prochloron sp. isolated from Diplosoma virens, 636 Ion and water balance of the hypo- and hyper-osmotically stressed chiton Mopalia muscosa, 364 permeable channel produced by venom of the fanged bloodworm Glvcera dibranchia, 485 regulation in freshwater clams, 325 regulation in Thais haemastoma, 148 Iron, in ascidian blood plasma, 656 Islet tissue, 473, 476 Isopods, terrestrial, coordination of moulting and reproduction, 206 see Bathynomus giganteus IWASA, K., see I. Tasaki, 494 KANEKO, AKIMICHI, AND ANN E. STUART, Coupling between horizontal cells in the carp retina examined by diffusion of Lucifer yellow, 486 KANESHIRO, EDNA S., AND RICHARD D. KARP, The ultrastructure of coelomocytes of the sea star Dermasterias imbricata, 295 KAPLAN, EHUD, et al., Recording from the Limulus ventral eye in situ: is there a circadian rhythm? 486 see Takehiko Saito, 490 KARP, RICHARD D., see Edna S. Kaneshiro, 295 KASS, LEONARD, AND ROBERT B. BARLOW, JR., Octopamine increases the ERG of the Limulus lateral eye, 487 KAUFMAN, THOMAS C., see Anna W. Seitz, 454 KELLY, P., see S. D. Sulkin, 402 KEM, WILLIAM R., AND JAMES D. SCOTT, Partial purification and characterization of a cytotoxic protein from squid (Loligo pealei) posterior salivary glands, 475 KIRCHMAN, DAVID, see Stephen Graham, 460 KIRCHMAN, D. L., et al., Bacterial epiphytes on Zoster a marina surfaces, 461 KLOTZ, FRANCIS W., et al., Schistosomiasis immune evasion, 496 KOECH, DAVID, see Francis W. Klotz, 496 KOIDE, SAMUEL S., see Oyewole Adeyomo, 466 see Akira Momii, 468 see Aloys G. Tumboh-Oeri, 470 KOPACHE, MARK E., Feeding of Neomysis mercedis (Holmes), 193 KRACKE, GEORGE R., AND PAUL DE WEER, Are there membrane surface charges in the vicinity of the sodium pump? 487 KURSAR, T. A., see R. S. Alberte, 498 et al., Relationships between photosystem II and phycobilisomes in red algae and cyanobacteria, 500 see Scott Schatz, 501 JAEGER, RICHARD, see William D. Cohen, 442 Janus green B : A vital stain during the process of mucus secretion, 447 and urn cell mucus, 571 JOHNSON, ERIC, see Keith Nelson, 162, 503 JUERS, DAVID W., et al., Tidal water ex- changes between Great Sippewissett Salt Marsh and Buzzards Bay, 461 K KAGAN, BRUCE L., et al., An ion-permeable channel produced by venom of the fanged bloodworm Glycera dibranchia, 485 KAHLER, STEPHEN, see G. Eric Bauer, 473 see Bryan D. Noe, 476 KALMIJN, AD. J., see Michael K. Gilson, 459 see Benjamin G. Dawson, 482 KAMINER, BENJAMIN, see James F. Head, 485 Labeling of microfilariae of Brugia malayi, 497 Labile components of the dogfish erythrocyte cytoskeleton, 441 Lack of respiratory response to temperature acclimation in two littorinid snails, 452 LANDOWNE, DAVID, see Virginia Scruggs, 491 Larval behavior, Styela montereyensis, 428 Larval development, geotaxis and barokinesis, and recruitment in crab Callinectes sapidus, 402 of Pinnixa longipes (Lockington, 1877) (Brachyura: Pinnotheridae), reared in the laboratory, 592 Larval respiration, and metabolism, of Macro- brachium olfersii, 692 Larval settlement, on microbial films: A model system, 460 Lasiotocus minutus, 497 1XDEX TO VOLUME 159 7S1 Learning, 480 associative, in gastropods, 505 LEDERHENDLER, IZJA, see Edward Barnes, 479 LEE, JOHN J., AND MONICA J. LEK, Con- textual relationships in food webs in- volving meiofauna, 462 LEE, MONICA J., see John J. Lee, 462 LEE, PAUL H., AND JUDITH S. WEIS, Effects of magnetic fields on regeneration in fiddler crabs, 681 Leech, mechanosensory cells, 483 LEIGHTY, BRUCE, Effect of chromium on microbial activity in salt marsh sedi- ments, 482 Leishmania enrietii, 497, 498 Lens (dogfish), 452, 472 LERMAN, SIDNEY, see Judith Megaw, 472 LEVINE, JOSEPH S., AND EDWARD F. MAC- NICHOL, JR., Relationship of body colors to environmental light conditions in "poster colored" fishes, 450 Life history, of Carcinonemertes errans, 247 of copepod Labidocera aestiva, 311 of Lasiotocus minutus, 497 of shrimp Crangon franciscorum and Pa- laemon macrodactylus, 177 of Tubulovesicula pinguis, 737 Light-scattering studies of sheared and ,un- sheared actin polymerization, 442 Limpet, Fissurella crassa, age determination, 606 Limulus, 480, 481, 486, 487, 490, 582 Liposome intraocular drug delivery, 472 LIPSCHULTZ, FREDRIC, see Annette Spies, 465 Littorina littorea and saxatilis, interactions, 463 Lobster, 450, 492 (Homarus), density-dependent growth in- hibition, 162, 503 Loligo, see squid LOPEZ-BARNEO, J., et al., Inside-out voltage clamp in the squid giant axon, 487 Luminescence, diet and induction of in Po- rychthys notatus, 364 firefly courting, 613 Lunules, function in Mellita, 561 LYERLA, TIMOTHY, see Katherine Turner, 418 LYNCH, EILEEN, et al., A natural protozoan- derived ionophore : a possible mechanism of cytotoxicity by Entamoeba histolytica, 496 Lytechinns pictus, enzyme activity in early embryo, 475 formation and early differentiation of go- nads, 280 LYTLE, CHARLES F., see Neil A. Mercando, 337 M MACAGNO, EDUARDO R., see Susan A. De- Riemer, 483 MACNICHOL, EDWARD F., JR., see Joseph S. Levine, 450 Macrobrachium olfersii (shrimp), larval res- piration, 692 Ma gel on a, 451 Magnetic, orientation in bacteria, 45V fields, and fiddler-crab regeneration, 6S1 Mandibular gland of the blue crab, Callinectes sapidus, 700 MANSOUR, RANDA A., Interactions between the two special of littorines — Littorina littorea and L. saxatilis — along New England shores, 463 Mapping vegetation and topography of Great Sippewissett Salt Marsh, Mass., 458 MARCUS, NANCY 1L, Photoperiodic control of diapause in the marine calanoid copepod Labidocera aestiva, 311 Marine Biological Laboratory, annual report, 1 MATHEWS, RITA \Y., AND AUDREY \V. V. HASCHEMEYER, Studies of protein syn- thesis in isolated toadfish hepatocytes, 475 Mating behavior in Pteroptyx tener (firefly), 613 MATSUMOTO, GEN, AND JUNICHI SHIMADA, Further improvements upon maintenance of adult squid (Doryteuthis bleekeri) in a small circular and closed-system aquarium tank, 319 MATTESON, R. I)., see J. Lopez-Barneo, 487 and Clay M. Armstrong, Temperature ef- fects on peak and steady state sodium currents in squid giant axons, 488 MAUZERALL, D., see L. Mazzella, 500 see T. A. Kursar, 500 MAZZELLA, L., et al., Photosynthetic light adaptation features of Zostera marina L. (eelgrass), 500 see D. L. Kirchman, 461 MCCARTHY, MICHAEL P., Effects of cyto- chalasin B and phalloidin on F-actin and G-actin, 449 MCCORKLE, SUSAN, AND THOMAS H. DIETZ, Sodium transport in the freshwater Asiatic clam, Corbicula fluminea, 325 MCLAUGHLIN, JANE, see Peter Gascoyne, 474 see Ronald Pethig, 477 McMAHON, ROBERT F., et al., Respiration in the polychaete worm Magelona: responses to temperature, hypoxia and tentacle ablation, 451 see David W. Aldridge, 447 see W. D. Russell-Hunter, 452 McMAHON-PRATT, DiANNE, see Sergio Arias- Negrete, 495 and students of parasitology course, Mono- clonal antibodies to Leishmania enriettit and tubulin, 497 see D. \Virth, 498 McN'AMARA, JOHN C., et al., Respiratory metabolism of Macrobrachium olfersii (Wiegmann) zoeae during the moulting cycle from eclosion to first ecdysis, 692 Mechanics and energetics of contraction in striated muscle of the sea scallop, Placo- pecton magellanicus, 489 INDEX TO VOLUME 159 Mechanism, of heavy metal inhibition of amino acid transport in intestine of marine fishes, 458 of intra-plate ciliary synchrony in cteno- phores, 446 of nicotinamide action in germinal vesicle breakdown in oocytes of Spisula soli- dissima, 468 Mechanisms of coordination between moulting and reproduction in terrestrial isopod Crustacea, 206 MEGAW, JUDITH, et al., Liposome intraocular drug delivery, 472 Meiofauna, food webs, 462 Melampus bidentatus, food choice, 464 MELILLO, J. M., see J. P. Schimel, 464 Mellita quinquiesperforata, 561 MERCANDO, NEIL A., AND CHARLES F. LYTLE, Specificity in the association between Hydractinia echinata and sympatric species of hermit crabs, 337 MEREFIELD, PAULINE M., see Clifford J. Hawkins, 656 Metals, in ascidian blood, 656, 669 Microanatomy and microtechniques, abstracts, 470 Microbial ecology of algal extracellular prod- ucts: the specificity of alga-bacterial in- teractions, 456 Midshipman fish, see Porichthys notatus, 231 MILLER, CHARLES B., DAVID M. NELSON, ROBERT R. L. GUILLARD, AND BONNIE L. WOODWARD, Effects of media with low silicic acid concentrations on tooth forma- tion in Acartia tonsa Dana (Copepoda, Calanoida), 349 MILLER, L., see D. Wirth, 498 MISCHKE, D., see D. Wirth, 498 MITCHELL, R., see D. L. Kirchman, 461 MITCHELL, RALPH, see Stephen Graham, 460 Mitochondria! movements in early develop- ment of Ciona, 446 MITTENTHAL, JAY E., et al., Morphology of the closer muscles in normal and homoe- otic legs of crayfish, 714 MITTENTHAL, JAY E., On the form and size of crayfish legs regenerated after grafting, 700 Molecular biology, abstracts, 473 MOLLURA, F., Cadmium resistance in bacteria from the Great Sippewissett Marsh, 463 MOLLURA, FRANCESCA, see Helmut Brandl, 457 MOMII, AKIRA, et al., Mechanism of nicotin- amide action on germinal vesicle break- down in oocytes of Spisula solidissima, 468 MOMII, F., see Akira Momii, 468 MONK, PETER B., see Anna W. Seitz, 454 Monoclonal antibodies to Leishmania enriettii and tubulin, 497 MOORE, JOHN, et al., Oscillation-free com- pensation for the series resistance in voltage clamped squid axons, 488 MOOSEKER, M. S., see W. B. Busa, 441 see Francine R. Smith, 445 Mopalia muscosa, salinity stress, 364 MORAN, MICHAEL N., see G. Eric Bauer, 473 see Bryan D. Noe, 476 MORAN, WILLIAM M., AND RICHARD E. TULLIS, Ion and water balance of the hypo- and hyperosmotically stressed chiton Mopalia muscosa, 364 MOREIRA, GLORIA S., see John C. McNamara, 692 MOREIRA, PLINIO, see John C. McNamara, 692 MORGAN, KATHLEEN, Histochemical identifi- cation of N-acetyl-j3-glucosaminidase ac- tivity in early embryos of Lytechinus pictus, 475 MORI, TAKAO, et al., Annual gonadal varia- tion in sea urchins of the orders Echino- thurioida and Echinoida, 728 Morphology, life history, and systematic rela- tions of Tubulovesicula pinguis (Linton, 1940) Manter, 1947 (Trematoda: Hemi- uridae), 737 of crayfish legs, 700, 714 of mandibular gland of Callinectes sapidus, 760 of Styela montereyensis, 428 of the closer muscles in normal and ho- nioeotic legs of crayfish, 714 Mosaic development in the polyclad turbella- rian Hoploplana inquilina and its evolu- tionary implications, 448 Mosaicism of pigmentation in frog eye, 453, 454 Motility, and squid embryo, 102 Abstracts, 441 Moulting, coordination with reproduction in terrestrial isopods, 206 cycle and metabolism of Macrobrachium olfersii larvae, 692 influence of mandibular gland of blue crab, 760 Mucus secretion, and Janus green, 447 in urn cell complex of Sipunculus nudus, 571 Mud crabs, elect rophoretic variation in, 418 MULLIN, ELIZABETH, Control of drilling fluid discharge from petroleum development on Georges Bank, 463 MURRAY, ANDREW, AND RONALD SOSNOWSKI, The status of mRNA in eggs and early embryos, 476 Muscle, in crayfish legs, 700, 714 Mutation, in Drosophila, 454 N NAGEL, RONALD L., see Thomas A. Borgese, 448 INDEX TO VOLUME 159 7 S3 Natural protozoan-derived ionophore: a pos- sible mechanism of cytotoxicity by Enta- moeba histolytica, 496 NELSON, DAVID M., see Charles B. Miller, 34;) NELSON, KEITH, DENNIS HEDGECOCK, WILL BORGESON, ERIC JOHNSON, RICHARD DAG- GETT, AND DIANE ARONSTEIN, Density- dependent growth inhibition in lobsters, Homarus (Decapoda: Nephropidae), 162, 503 NELSON, LEONARD, see Mario H. Burgos, 467 Receptors, drug interactions, and calcium in regulation of Arbacia sperm cell func- tion, 469 NEMHAUSER, IRIS, AND WILLIAM D. COHEN, In vivo disassembly/reassembly of the marginal band in an invertebrate erythro- cyte, 449 Neomysis mercedis feeding, 193 Neurobiology, abstracts, 478 ff. of Hermissenda, 505 Neuroplasmic lattice, 471 Neurosecretory cells and neural complex re- lated to gonad development in ascidian Symplegma reptans, 219 NEUSTADT, JEFFREY B., see G. Eric Bauer, 473 see Bryan D. Noe, 476 Nicotinamide action in germinal vesicle break- down, 468 Ninhydrin positive substances, and Thais haemastoma, 148 Nitrogen metabolism, in littorinid snails, 447 NOE, BRYAN D., see G. Eric Bauer, 473 et al., Inhibition of islet prohormone to hormone conversion by incorporation of arginine and lysine analogs, 476 OBAID, ANA LIA, AND BIRGIT ROSE, Electrical membrane properties of single and two- cell preparations from Chironomus salivary gland, 489 Observations on the isolated mitotic apparatus ghost, 445 Octopamine increases the ERG of the Limulus lateral eye, 487 OGILVY, CHRISTOPHER S., AND ARTHUR B. DuBois, Interstitial fluid pressures of smooth dogfish (Mustelus canis) and bluefish (Pomatomus saltatrix) tilted in air, 451 see Arthur B. DuBois, 450 OHTSU, KOHZOH, Electrical activities in the subtentacular region of the anthomedusan Spirocodon saltatrix (Tilesius), 376 OKAZAKI, KAYO, AND RICHARD DILLAMAN, Crystalline axes of the test and spine of the sea urchin Strongylocentrotus purpura- tus — A new method of analysis, 472 OLSEN, MARY C, see Jay E. Mittenthal, 714 On the form and size of crayfish legs regen- erated after grafting, 700 Opsanus tau, see toadfish Optical determination of the resistance in series with the axolemma of Loligo pealei, 491 Orientation and current-induced How in the stalked ascidian Styela montereyensis, 428 Oriented particle assemblies in the plasma membrane of Tetrahymena: their deploy- ment relative to cell surface topography, cellular morphogenesis, and sensitivity to stimuli, 471 Orthogonal polarization sensitivities of squid photoreceptors : implication for a retinal design, 490 Oscillation-free compensation for the series resistance in voltage-clamped squid axons, 488 Osmoregulation, and freshwater clam, 325 and chiton, 364 in Thais haemastoma, 148 in Glycera, 626 Oyster drill, Thais haemastoma, 148 Oxygen free radical generation by gossypol : a possible mechanism of antifertility action in sea urchin sperm, 468 Paguridae, symbiosis with Hydractinia, and shell selection, 337 Palaemonetes pugio, relationship between diet and growth, 458 Palaemon macrodactylus, seasonal abundance and distribution, 177 Panopens herbstii, elect rophoretic variation in, 418 Parasites, Lasiotocus minutus, 497 Tubulovesicula pingnis, 737 Parasitology, abstracts, 495 PARDUE, M., see D. Wirth, 498 PARRY, DAVID L., see Clifford J. Hawkins, 656, 669 Partial purification and characterization of a cytotoxic protein from squid (Loligo pealei) posterior salivary glands, 479 PATEREK, J. R., see Helmut Brandl, 457 PAXHIA, TERESA, see Seymour Zigman, 452 PERSELL, ROGER, AND AUDREY E. V. HA- SCHEMEYER, Inhibition of L-leucine trans- port in toadfish liver by various natural amino acids, 477 PETERSON, BRUCE, see Amy Friedlander, 458 see David W. Juera, 461 PETHIG, RONALD, et al., Influence of water structure on proton diffusion in proteins, 477 Petroleum drilling effects, 463 Phagocytosis, of Dermasterias imbricata, 295 Phalloidin, 444, 445 Philoscia vittata, 460 Phospholipid synthesis in axoplasm from the squid giant fiber, 484 784 INDEX TO VOLUME 159 Photoperiod, and terrestrial isopod Crustacea, 206 Photoperiodic control of diapause in the marine calanoid copepod Labidocera aestiva, 311 Photoreceptors, 480, 481, 483, 486, 487, 490 in Hermissenda, 505 Photosynthesis, abstracts, 498 ff. by Prochloron isolated from Diplosoma virens, 636 Photosynthetic activity of fungal infected and noninfected Laminaria saccharina Lam., 501 Photosynthetic light adaptation features of Zostera marina L. (eelgrass), 500 Phototactic response, effects of caffeine and theophylline on Chlamydomonas reinhardii, 649 Phycocyanobilin synthesis from exogenous heme, 499 Physiology, abstracts, 447 PIERCE, CRAIG, see Clifford J. Hawkins, 669 PIERCE, SIDNEY K., see Charles J. Costa, 626 PIESSEKS, WILLY F., Labeling of microfilariae of Brugia malayi, 497 see Sergio Arias- Negrete, 495 Pigmentation mosaicism in the choroid of the eye after embryonic grafting, 453 Pinnixa longipes, larval development, 582 PIRES, ANTHONY, see Randy Chambers, 458 Placopecton magellanicus (sea scallop), 489 Plant pigments and photosynthesis, abstracts, 498 Plasma, of Pyura stolonifera and Ascidia ceraiod.es blood, compared, 656 Podoclavella moluccensis, 669 POLIDORE, THOMAS, see A. Farmanfarmaian, 458 POLLARD, HARVEY B., see Bruce L. Kagan, 140 Polychaete, see Glycera Pomatomus saltatrix, see bluefish Positional correlation of synaptic boutons in pairs of mechanosensory cells in the leech, 483 Potassium channels, 483, 485, 492, 493 Predation, by Carcinonemertes errans on Cancer magister, 247 see also food, feeding Predatory habits of two lycosid spiders in Great Sippewissett Marsh, 456 Preferred food sources and the limitation of local distributions of the isopod crustacean Philoscia vittata, 460 Preliminary, characterization of the hemo- cyanin of the giant sea roach, Bathynomus giganteus, 478 evidence for taste-aversion learning in the nudibranch mollusc Aeolidia papillosa, 480 Presynaptic, calcium currents and facilitated transmitter release in the giant synapse of Loligo pealei, 494 injection of calcium facilitates transmitter release in the squid giant synapse, 481 Primordial germ cells, of sea urchins, 280 Procambarus clarkii, see crayfish Prochloron, photosynthetic capacity isolated from Diplosoma virens, 636 Proline, in Glycera, 626 Properties of isolated fertilization envelopes from Arbacia punctitlata, 467 Protein, complexes, 474 cytotoxic, 475 plant, 498, 500 proton diffusion, 477 synthesis, 475, 478 calmodulin, 485 Pteroptyx tener, (firefly) courting and flashing, 582 Purification of parasite mRNA, 498 Pyura stolonifera blood plasma, 656 R RACE, MARGARET S., see Erich F. Horgan, 460 RAHAT, M., AND ORIT ADAR, Effect of sym- biotic zooxanthellae and temperature on budding and strobilation in Cassiopeia andromeda (Eschscholz), 394 RALL, JACK A., Mechanics and energetics of contraction in striated muscle of the sea scallop, Placopecton magellanicus, 489 Reaction-diffusion models of morphogenesis: an application to pattern formation in Xenopus retina, 454 Recording from the Limulus ventral eye in situ: is there a circadian rhythm? 486 Recruitment, of larvae of Callinectes sapidus, 402 Regeneration, of crayfish legs, 700, 714 of fiddler crab legs in magnetic fields, 681 in gamma-irradiated Campanularia flexuosa, 752 REINSCHMIDT, DANA, see R. Kevin Hunt, 454 see Robert Tompkins, 455 REITSMA, CAROL S., Food choice and pala- tability in a salt marsh detritivore, Melampus bidentatus, 464 Relationship, between diet and growth rate of the grass shrimp Palaemonetes pugio, 458 between dendrite and spine neck diameters in freeze-fractured rat hippocampal for- mation, 470 between organic and nitrogen content of marsh sediments and feeding rates in Uca pugnax, 460 between photosystem II and phycobilisomes in red algae and cyanobacteria, 501 of body colors to environmental light con- ditions in "poster colored" fishes, 450 Release of cytoskeletal proteins into the perfu- sate of squid giant axons, 441 Reproduction, coordinated with moulting in terrestrial isotops, 206 IXDEX TO VOLUME 159 785 in ascidian Symplegmo, reptuns, 219 see also Fertilization Reproductive cycles, of the sea urchins of Echinothurioida and Echinoida, 72S Resource partitioning, see competition Respiration in the polychaete worm Magelona: responses to temperature, hypoxia and tentacle ablation, 451 Respiratory metabolism of Macrobrachium olfersii (Wiegmann) zoeae during the moulting cycle from eclosion to first ecdysis, 692 Retinotectal patterns and patterns of genetic mosaicism of ploidy chimerae of Xenopus eye, 454 RKYNOLDS, GEO. T., Evaluation of low light level intensifier vidicon detectors for mi- croscopy, 473 Rhythm, circadian, 486, 490 in Chlamvdomonas reinhardii, affected by caffeine and theophylline, 649 Rifampin as a selective agent for the isolation and enumeration of Spirochaeta from salt marsh habitats, 465 mRXA, 476, 478, 498 ROSE, BIRGIT, see Ana Lia Obaid, 489 ROSENBERG, IAN, see Sergio Arias-Negrete, 495 see Eileen Lynch, 496 see Willy F. Piesens, 497 see D. Wirth, 498 RUDERMAN, JOAN, see Teresa Tansey, 478 RUSSELL-HUNTER, VV. D., et al., Lack of respiratory response to temperature ac- climation in two littorinid snails, 452 see David W. Aldridge, 447 see Robert F. McMahon, 451 SAIDEL, WILLIAM M., Orthogonal polarization sensitivities of squid photoreceptors : im- plication for a retinal design, 489 SAITO, TAKEHIKO, et al., Circadian rhythm of photoreceptor cells in the Limulns lateral eye: further studies, 490 SALAMA, G., et al., Vanadate inhibits ciliary beating in intact snail salivary glands, 449 Salinity stress, and chitons, 364 and freshwater clams, 325 and Thais haemastoma osmoregulation, 148 in Glycera, 626 SALZBERG, B. M., et al., An optical determina- tion of the resistance in series with the axolemma of Loligo pealei, 491 see G. Salama, 449 Sand dollar, see Mellita SASNER, JOHN J., JR., see William J. Adelman, Jr., 478 Scanning electron microscopy, of cilia on squid embryo, 102 SCHATZ, SCOTT, AND THOMAS A. KURSAR, Photosynthetic activity of fungal infected and noninfected Laminaria saccharina Lam., 501 SCHIMEL, J. P., <•/ al., Denitrification poten- tials in a successional sequence of northern hardwood forest stands, 4(>4 Schistosomiasis immune evasion, 496 SCOTT, JAMES I)., see William R. Kern, 475 SCRUGGS, VIRGINIA, AND DAVID LANDONVM , The birefringence response of voltage- clamped internally perfused axons, 491 Sea anemone aggressive behavior, 1 17 Seasonal abundance and distribution of Cran- gon franciscorum and Palaemon macro- dactylus (Decapoda, Caridael in the San Francisco Bay-Delta, 117 Sea star, see starfish Sea urchin, see Arbacia, Echinoida, Echino- thurioida, Lytechinus pictus, Strongylocen- trotus Seaweed, aquaculture, 459 see also algae SEGAL, SHELDON, see Mario H. Burgos, 467 see Michael Coburn, 468 SEITZ, ANNA W., et al., A developmental analysis of an unusual homoeotic muta- tion, proboscipedia, in Drosophila melano- gaster, 454 Selective feeding: by Neomysis mercedis, 193 see also food, feeding Self /not-self recognition in sea anemones, 117 SENSEMAN, D. M., see G. Salama, 449 SERHAN, C., see P. Anderson, 479 Sewage, 465 Shell selection, by hermit crabs, 337 SHER, ALAN, see Francis W. Klotz, 496 SHERRY, BARBARA, see Sara Anderson, 496 SHIMADA, JUNICHI, see Gen Matsumoto, 319 SHIRAI, HIROKO, see Oyewole Adeyomo, 466 and Yasuaki Yoshimoto, Tension generation in ovarian wall by 1-methyladenine during starfish spawning, 469 SHOAF, SARAH A., et al., Reaction-diffusion models of morphogenesis: an application to pattern formation in Xenopus retina, 454 Shrimp, see Crangon, franciscorum, Macro- brachium, olfersii, Neomysis mercedis, Pa- laemon macrodactylus, Paluemonctes pugio SIEGFRIED, CLIFFORD A., AND MARK E. KOPACHE, Feeding of Neomysis mercedis (Holmes), 177 Seasonal abundance and distribution of Crangon franciscorum and Palaemon macro- dactylus (Decapoda, Caridae) in the San Francisco Bay-Delta, 177 Siliceous teeth, production by Acartia tonsa, 349 Silicic acid, deficiency, and effects on tooth formation in Acartia tonsa (copepod), 349 production of media low in, 349 SINSHKIMKR, PETER, see Michael Coburn, 468 et al., Toxic effects of vitamin A on sea urchin gametes, 469 786 IXDEX TO VOLUME 159 Sippewissett Marsh, 456, 458, 459, 461, 463 Sipunculus nudus, mucus secretion, 571 SMITH, FRANCINE R., et a!., The interaction of phalloidin with G- and F-actin, 445 SMITH, GEORGE W., Observations on the isolated mitotic apparatus ghost, 445 SMITH, K. M., et a/., Chemical conversion of a chlorophyll a derivative to a bile pigment, 501 SMITH, STEPHEN J, et al., Arsenazo III reveals long-lasting intracellular calcium transient following the action potential in squid giant presynaptic terminal, 491 see Milton P. Charlton, 481 see Robert S. Zucker, 494 Snails, 447, 463, 452, 457 see Ilyanassa obsoleta, Littorina littorea Socci, ROBIN, see A. Farmanfarmaian, 458 Sodium channels, 479, 483, 485, 487, 488 Sodium transport in the freshwater Asiatic clam Corbicula fluminea, 325 Solid-phase radioimmune assay to detect anti- bodies in Entamoeba histolytica, 495 SOSNOWSKI, RONALD, see Andrew Murray, 476 Specificity in the association between Hy- dractima echinata and sympatric species of hermit crabs, 337 Sperm, 467, 468, 470, 728 Sperm-oocyte interaction in the surf clam Spisula solid/is sima, 470 Spiders, in Sippewissett Marsh, 456 SPIES, ANNETTE, AND FREDRIC LIPSCHULTZ, Effect of grazing by Uca pugnax on the microbial population in salt marsh sedi- ment, 465 Spindles, 443, 445, 446 Spirochaeta, 476 Spirocodon saltatrix, electrical activities, 376 Spisula, 443, 446, 466, 468, 470, 476, 478 SPITSBERGEN, JAN, see James C. Carlisle, 495 Sponge, aggregation factor, 449 Cliona celata, burrowing and carbonic an- hydrase, 135 SPRAY, D. C., see J. H. Stern, 493 Squid, 471, 475, 481-485, 487-492, 494 axon sodium channels are blocked reversibly by aphantoxin derived from a prokaryotic alga, 478 development of ciliature pattern on the embryo, 102 Doryteuthis bleekeri, maintenance in aquaria, 319 Stability of dogfish lens fiber cell membranes, 452 Stages in the life cycle of the digenetic trema- tode, Lasiotocus minutus (Manter, 1931) Thomas, 1959, 497 Starfish, 469, 495 ultrastructure of coelomocytes, 295 Statistical mechanics of geomagnetic orienta- tion in sediment bacteria, 459 Status of mRNA in eggs and early embryos, 476 STEEL, C. G. H., Mechanisms of coordination between moulting and reproduction in ter- restrial isopod Crustacea, 206 STEELE, RODERIC E., AND DANIEL L. GILBERT, Calcium and potassium activities in the hemolymph of the squid, Loligo pealei, 492 STEPHENS, PHILIP J. AND C. K. GOVIND, Elim- ination of synapses from identified lobster motor neurons during development, 492 STERN, J. H., etaL, Gap junctions: quantitative comparison of reduction in conductance by H and by Ca ions in an internally perfused preparation, 493 STEUDLER, P. A., see J. P. Schimel, 464 STICKLE, WILLIAM B., see Jane E. Hildreth, 148 Stomachicola, D17 STOPAK, DAVID, AND ROSARIA DE SANTIS, Mitochondrian movements in early de- velopment of Ciona, 446 STORRIE, BRIAN, see Lucio Cariello, 467 Strobilation, of Cassiopeia andromeda, effect of symbionts, 394 Strongylocentrotus skeletal-cell structure, 472 STUART, ANN E., see Kathleen A. French, 483 see Akimichi Kaneko, 486 Studies, of methanogenic bacteria from intes- tinal tracts of marine fishes, 457 of protein synthesis in isolated toadfish hepatocytes, 475 on reproduction in the compound ascidian, Symplegma reptans: Relationship between neural complex and reproduction, 219 STUNKARD, HORACE W. Morphology, life his- tory, and systematic relations of Tubn- lovesicula pinguis (Linton, 1940) Manter, 1947 (Trematoda, Hemiuridae), 737 Stages in the life cycle of the digenetic tre- matode, Lasiotocus minutus (Manter, 1931) Thomas, 1959, 497 Styela montereyensis, ascidian; orientation, larval behavior, morphology, ecology, 428 Subcellular localization and characterization of islet hormone-degrading enzymes in angler- fish islet tissue, 473 Substitution of calcium by polycations in sponge aggregation factor interaction, 449 SUGIMOTO, KEIJI, AND HlROSHI WATANABE, Studies on reproduction in the compound ascidian Symplegma reptans: Relationship between neural complex and reproduction, 219 SULKIN, S. D., W. VAN HUEKELEM, P. KELLY, AND L. VAN HUEKELEM, The behavioral basis of larval recruitment in the crab Callinectes sapidus Rathbun : A laboratory investigation of ontogenetic changes in geotaxis and barokinesis, 402 Swelling of squid giant axon during action po- tentials, 494 INDEX TO VOLUME 159 787 SWENSON, R. P., JR., AND C. M. ARMSTRONG, External k+ and Rb+ retarded closing of potassium channels, 493 Swimbladder, 448 SWINEHART, JAMES H., see Clifford J. Hawkins, 656 Symbiosis, of Cassiopeia andromeda and algae, 3)4 of chemoautotrophic bacteria and marine invertebrates, 456 hermit crabs and Hydractinia, 337 of ProMoron and Diplosoma virens, 636 Sympatric, hermit crabs, specificity of associa- tion with Hydractinia echinata, 337 mud crabs, electrophoretic variation in, 418 shrimp Crangon franciscorum and Palaemon macrodactylus, 177 Symplegma reptans, neural complex and re- production, 219 Synapse, 481, 491, 492, 494 Synthesis of chlorophyll a from 5-aminolevulin- ate in dark-grown Cyanidium caldarium cells, 502 SZARO, BEN G., see R. Kevin Hunt, 454 SZENT-GYORGYI, ALBERT, see Peter Gascoyne, 474 see Ronald Pethig, 477 Talorchestia longicornis, 455 TAMM, SIDNEY L. The mechanism of intra- plate ciliary synchrony in ctenophores, 446 TANSEY, TERESA, AND JOAN RUDERMAN, Change in the protein synthetic pattern and mRNA population during Spisula embryogenesis, 478 TASAKI, I., -AND K. IWASA, Swelling of squid giant axon during action potentials, 494 TAUCK, DAVID, see John Moore, 488 Taurine, in Clycera, 626 Taxonomy, of Tubulovesicula pingnis, 737 TAYLOR, C. D., see Helmut Brandl, 456 TELZER, BRUCE R., AND LEAH T. HAIMO. Bind- ing of dynein to isolated meiotic spindles of the surf clam, Spisula solidissima, 446 Temperature acclimation and nitrogen metab- olism in littorinid snails, 447 Temperature, effect on budding and strobilation in Cassiopeia andromeda, 394 Temperature effects on peak and steady state sodium currents in squid giant axons, 488 Temperature-salinity interaction, and Thais haemastoma, 148 Tension generation in ovarian wall by 1-methyl- adenine during starfish spawning, 469 Teredo navalis (shipworm), effect of competition on larvae in closed culture, 465 Tetrahymena, 471 Thais haemastoma, osmoregulation, 148 Theophylline, effects on Chlamydomonas rein- hardii, 649 Theory of chemotaxis and the ability of a cell to sense its position and orientation within a tissue, 443 Thermost ability of fish hemoglobins, 448 Thrust and drag of bluefish (Pomatomus salt- atrix) at different buoyancies, speeds, and swimming angles, 450 Tidal water exchanges between Great Sippe- wissett Salt Marsh and Buzzards Bay, 461 Toadfish (Opsanus tan), 475, 477 TOMPKINS, ROBERT, et «/., Application of a polyploid marker to clonal analysis in Xenopus eye, 455 see R. Kevin Hunt, 454 Toxic effects of vitamin A on sea urchin gametes, 469 TRACEY, GREGORY A., et al., Effects of closed- culture competitive interactions on growth of Teredo navalis larvae, 465 Trematodes, 737, 497 TRENCH, ROBERT K., see Charles R. Fisher, Jr., 636 TROLL, WALTER, see Michael Coburn, 468 see Peter Sinsheimer, 469 TROXLER, R. F., see R. S. Alberte, 498 see S. B. Brown, 499 and S. B. BROWN, Synthesis of chlorophyll a from 5-aminolevulinate in dark-grown Cyanidium caldarium cells, 502 see K. M. Smith, 501 TSUCHIYA, TEIZO, see Takao Mori, 728 Tubulin genes in parasites, 498 TULLIS, RICHARD E., see William M. Moran, 364 TUMBOH-OERI, ALOYS G., AND SAMUEL S. KOIDE, Sperm-oocyte interaction in the surf clam Spisula solidissima, 470 TURNER, KATHERINE, AND TIMOTHY A. LYERLA, Electrophoretic variation in sympatric mud crabs from North Inlet, South Carolina, 518 TURNER, RUTH D., see Gregory A. Tracey, 465 TV, see video u Uca (fiddler crab: feeding rates on sediment, 460 grazing and microbes, 465 regeneration and magnetic fields, 681 Ultrastructural characteristics of the non- expanded and expanded extra-embryonic shell of the horseshoe crab, Limulus poly- phemus, 582 infrastructure, abstracts, 44 Iff, 470 of coelomocytes of the sea star Dermasterias inbricata, 295 Urn cell complcs of Sipunculus nudus: A model for study of mucus-stimulating sub- stances, 571 USEM, MICHAEL, see David Hughes, 460 788 INDEX TO VOLUME 159 I"\i KI, TIMOTHY, AND WENDY WILTSE, The effects of sewage fertilization on benthic macroinvertebrates in salt marsh creeks, 465 Vanadate inhibits ciliary beating in intact snail salivary glands, 444 Vanadium, in ascidians, 656, 669 VAN HOLDE, K. E., see W. B. Busa, 442 see Francine R. Smith, 445 and M. Brenowitz, A preliminary character- ization of the hemocyanin of the giant sea roach, Bathynomus giganteus, 478 VAN HUEKELEM, L., see S. D. Sulkin, 402 VAN HUEKELEM, W., see S. D. Sulkin, 402 Video, and firefly mating, 613 and low light level microscopy, 473 Vision, see photoreceptors, lens Vitamin A, 469 Vitellogenesis, in terrestrial isopod Crustacea, 206 Volume regulation, sec osmoregulation, salinity WICKHAM, DANIEL E., Aspects of the life his- tory of Carcinonemertes errans (Nemertea: Carcinonemertidae), an egg predator of the crab Cancer magister, 247 WILLIAMS-ARNOLD, Lois, see John M. Arnold, 102 WILTSE, WENDY, see Timothy Uyeki, 465 WIRTH, D., Cloning genes of Leishmania en- riettii in E. coli, 498 Purification of parasite mRXA, 498 Tubulin genes in parasites, 498 WITTENBERG, JONATHAN, see Eugene Copeland, 449 WOODWARD, BONNIE L., See Charles B. Miller, 349 WORTHINGTON, C. R., see A. R. Czeto, 482 see H. Inouye, 485 X Xenopus (frog) : eye, 453, 454 X-ray, study of fish nerves, 485 study of retinal photoreceptor structure of squid, 482 W WARGO, JOHN P., The effectiveness of National Park Service policies in protecting barrier island ecosystems within the Cape Cod Natural Seashore, 466 WARNER, JOHN A., AND JAMES F. CASE, The zoogeography and dietary induction of bioluminescence in the midshipman fish, Porychthys notatus, 231 WARREN, J. G., see J. P. Schimel, 464 WARREN, M. KIM, see Charles J. Costa, 626 WATANABE, HIROSHI, see Keiji Sugimoto, 219 Water balance, see osmoregulation WEBER, FREDERICK H. AND E. P. GREENBERG, Rifampin as a selective agent for the iso- lation and enumeration of Spirochaeta from salt marsh habitats, 465 WEIS, JUDITH S., see Paul H. Lee, 681 WEISSMANN, G., see P. Anderson, 479 WERMUTH, JEROME F., Gamma radiation and hydranth longevity in Campamdaria flex- uosa: Age-dependency of dose-response function, 752 WHITEHEAD, DENEENE, see William D.Cohen, 442 YOSHIMOTO, YASUAKI, see Hiroko Shirai, 469 YOUNG, CRAIG M., AND LEE F. BRAITHWAITE, Orientation and current-induced flow in the stalked ascidian Stvela montereyensis, 428 YUDIN, ASHLEY L, RICHARD A. DIENER, W. H. CLARK, JR., AND ERNEST S. CHANG, Mandibular gland of the blue crab, Calli- nectes sapidus, 760 YULO, TERESA, see Seymour Zigman, 452 ZACKS, C. M., see J. P. Schimel, 464 ZIGMAN, SEYMOUR, et a!., Stability of dogfish lens fiber cell membranes, 452 Zinc in ascidian plasma, 669 Zoogeography and dietary induction of biolu- minescence in the midshipman fish, Porichthys notatus, 231 Zostera maria (eel grass), 461, 500 ZUCKER, ROBERT S., et al., Presynaptic calcium currents and facilitated transmitter release in the giant synapse of Loligo pealei, 494 see Milton P. Charlton, 481 see Stephen J Smith, 491 Continued from Cover Two high cost of setting tabular material in type, authors art- asked to limit such material as much As possible. Tables, with their headings and footnote*, should be typed on separate sheets, numbered with consecutive Roman numerals, and placed after the Literature Cited. Figure legends should contain enough information to make the figure intelligible' separate from the text. Legends should be typed double spared, with consecutive Arabic numbers, on a separate sheet at the end of the paper. Footnotes should be limited to authors' current addresses, acknowledgments or contribu- tion numbers, and explanation of unusual abbreviations. All such footnotes should appear on the title page. Footnotes are not normally permitted in the body of the text. 4. A condensed title or running head of no more than 35 letters and spaces should appear at the top of the title page. 5. Literature cited. In the text, literature should be cited by the Harvard system, with papers by more than two authors cited as Jones et al., 1980. Persona) communications and material in preparation or in press should be cited in the text only, with author's initials and institutions, un- less the material has been formally accepted and a volume number can be supplied. The list of references following the text sould be headed LITERATURE CITED, and must be typed double spaced on separate pages, conforming in punctuation and arrangement to the style of recent issues of The Biological Bulletin. Citations should include complete titles and inclusive pagination. Journal abbreviations should normally follow those of the U. S. A. Standards Institute (USASI), as adopted by BIOLOGICAL ABSTRACTS and CHEMICAL ABSTRACTS, with the minor differences set out below. The most generally useful list of biological journal titles is that published each year by BIOLOGICAL ABSTRACTS (BIOSIS List of Serials; the most recent issue). Foreign authors, and others who are accustomed to use THE WORLD LIST OF SCIENTIFIC PERIODICALS, may find a booklet published by the Biological Council of the U.K. (obtainable from the Institute of Biology, 41 Queen's Gate, London, S.W.7, England, U.K. useful, since it sets out the WORLD LIST abbreviations for most biological journals with notes of the USASI abbreviations where these differ. CHEMICAL ABSTRACTS publishes quarterly supplements of additional abbreviations. The following points of reference style for THE BIOLOGICAL BULLETIN differ from USASI (or modified WORLD LIST) usage : A. Journal abbreviations, and book titles, all underlined (for italics) B. All components of abbreviations with initial capitals (not as European usage in WORLD LIST e.g. J. Cell. Comp. Physiol. NOT J. cell. comp. Physiol.) C. All abbreviated components must be followed by a period, whole word components must not (i.e. J. Cancer Res.) D. Space between all components (e.g. J. Cell. Comp. Physiol., not J.Cell.Comp. Physiol.) E. Unusual words in journal titles should be spelled out in full, rather than employing new abbreviations invented by the author. For example, use Rit Visindafjelags Islendinga without abbreviation. F. All single word journal titles in full (e.g. Veliger, Ecology, Brain). G. The order of abbreviated components should be the same as the word order of the com- plete title (i.e. Proc. and Trans, placed where they appear, not transposed as in some BIOLOGICAL ABSTRACTS listings). H. A few well-known international journals in their preferred forms rather than WORLD LIST or USASI usage (e.g. Nature, Science, Evolution NOT Nature, Lond.; Science, N. Y.; Evolution, Lancaster, Pa.) 6. Reprints, charges. The Biological Bulletin has no page charges. However, authors will be requested to help pay printing charges of manuscripts that are unusually costly due to length or numbers of tables, figures, or formulae. Reprints may be ordered at time of publication and normally will be delivered about two to three months after the issue date. Authors (or delegates or foreign authors) will receive page proofs of articles shortly before publication. They will be charged the current cost of printers' time for corrections to these (other than correction of printers' or editors' errors). CONTENTS ERRATUM 503 Invited Article: ALKON, DANIEL L. Cellular analysis of a gastropod (Hermissenda crassicornis) model of associative learning 505 ALEXANDER, DAVID E., AND JOE GHIOLD The functional significance of the lunules in the sand dollar, Mellita quinquiesperforata 561 BANG, BETSY G., AND FREDERIK B. BANG The urn cell complex of Sipunculus nudus: A model for study of mucus-stimulating substances 571 BANNON, GARY A., AND GEORGE GORDON BROWN Ultrastructural characteristics of the non-expanded and expanded extra-embryonic shell of the horseshoe crab, Limulus polyphemus L 582 BOUSQUETTE, GEORGE D. The larval development of Pinnixa longipes (Lockington, 1877) (Brachyura: Pinnotheridae), reared in the laboratory 592 BRETOS, MARTA Age determination in the keyhole limpet Fissurella crassa Lamarck (Archaeogastro- poda : Fissurellidae), based on shell growth rings 606 CASE, JAMES F. Courting behavior in a synchronously flashing, aggregative firefly, Pteroptyx tener 613 COSTA, CHARLES J., SIDNEY K. PIERCE, AND M. KIM WARREN The intracellular mechanism of salinity tolerance in polychaetes : Volume regulation by isolated Glycera dibranchiata red coelomocytes 626 FISHER, CHARLES R., JR., AND ROBERT K. TRENCH In vitro carbon fixation by Prochloron sp. isolated from Diplosoma virens 636 GOODENOUGH, JUDITH E., AND VICTOR G. BRUCE The effects of caffeine and theophylline on the phototactic rhythm of Chlamydomonas reinhardii 649 HAWKINS, CLIFFORD J., PAULINE M. MEREFIELD, DAVID L. PARRY, WILTON R. BIGGS, AND JAMES H. SWINEHART Comparative study of the blood plasma of the ascidians Pyura stolonifera and Ascidia ceratodes 656 HAWKINS, CLIFFORD, J., DAVID L. PARRY, AND CRAIG PIERCE Chemistry of the blood of the ascidian Podoclavella moluccensis 669 LEE, PAUL H., AND JUDITH S. WEIS Effects of magnetic fields on regeneration in fiddler crabs 681 MCNAMARA, JOHN C., GLORIA S. MOREIRA, AND PLINIO S. MOREIRA Respiratory metabolism of Macrobrachium olfersii (Wiegmann) zoeae during the moulting cycle from eclosion to first ecdysis 692 MlTTENTHAL, JAY EDWARD On the form and size of crayfish legs regenerated after grafting 7CO MlTTENTHAL, JAY EDWARD, MARY C. OLSON, AND GLENNA D. CUMMINGS Morphology of the closer muscles in normal and homoeotic legs of crayfish 714 MORI, TAKAO, TEIZO TSUCHIYA, AND SHONAN AMEMIYA Annual gonadal variation in sea urchins of the orders Echinothurioida and Echinoida 728 STUNKARD, HORACE W. The morphology, life history, and systematic relations of Tubulovesicula pinguis (Linton, 1940) Manter, 1947 (Trematoda : Hemiuridae) 737 WERMUTH, JEROME F. Gamma radiation and hydranth longevity in Campanulariaflexuosa: Age-dependency of dose-response function 752 YUDIN, ASHLEY I., RICHARD A. DIENER, W. H. CLARK, JR., AND ERNEST S. CHANG Mandibular gland of the blue crab, Callinectes sapidus 760 INDEX TO VOLUME 159 .. 773