MBL February 2003 THE Volume 204 Number 1 BIOLOGICAL BULLETIN R D CKOSCUrtb DRAWN FROM EXPERIENCE ONLY NIKON COULD REINVENT INVERTED MICROSCOPY WITH THE NEW ECLIPSE TE2OOO SERIES. When they were launched, Nikon's Eclipse TE microscopes set new standards in optical performance, continuing a tradition of over twenty years of excellence in inverted microscopy Now, Nikon proudly introduces the next generation of Eclipse research inverted microscopes. The Eclipse TE2000 series incorporates past successes into a new design with innovations that are exactly what today's varied research techniques require. Here are just some of the special features built into this extraordinary microscope: Multi-port design permitting addition of optional illumination sources and other attachments without modifying the microscope. Unique "Noise Terminator" thoroughly eliminates stray light for greater S/N ratio in fluorescence imaging Super-nanoprecision Z-access control and increased stability for superior 3D imaging Modular motorized components for automated control Exclusive CFI60 infinity optical system for excellent resolution with high numerical apertures and longer working distances. The Eclipse TE2000 is available in three models, each configured to meet your research needs. Draw on Nikon's experience For more information, call 1 -800-52-NIKON, ext. 392 or visit www.mkonusa com In Canada, call 1 -866-99-NIKON Visit MicroscopyU at www.nikonusa.com to learn more about inverted microscopy. Nikon FEB 2 5 2003 Bathtub Theory explains light sensitivity in Salamander eyes... CURIOUS? - More than 12 million searchable journal articles - World's largest collection of free full-text articles - 4 different search tools to locate what you need - Online archives of The Biological Bulletin plus more than 330 other journals covering the sciences and medicine Find what you need at Stanford University's www.hiqhwire.Qrq frHighWire THE BIOLOGICAL BULLETIN ONLINE The Marine Biological Laboratory is pleased beginning with the October 1976 issue to announce that the full text ofllie Biological (Volume 151, Number 2), and some Tables of Bulletin is available online at Contents are online beginning with the October 1965 issue (Volume 129, Number 2). http://www.biolbull.org The Biological Bulletin publishes outstanding experimental research on the full range of biological topics and organisms, from the fields of Neurobiology, Behavior, Physiology, Ecology, Evolution, Development and Reproduction, Cell Biology, Biomechanics, Symbiosis, and Systematics. Published since 1897 by the Marine Biological Laboratory (MBL) in Woods Hole, Massachusetts, The Biological Bulletin is one of America's oldest peer-reviewed scientific journals. The journal is aimed at a general readership, and especially invites articles about those novel phenomena and contexts characteristic of intersecting fields. The Biological Bulletin Online contains the full content of each issue of the journal, including all figures and tables, beginning with the February 2001 issue (Volume 200, Number 1). The full text is searchable by keyword, and the cited references include hyperlinks to Medline. PDF files are available beginning in February 1990 (Volume 178, Number 1), some abstracts are available Each issue will be placed online approximately on the date it is mailed to subscribers; therefore the online site will be available prior to receipt of your paper copy. Online readers may want to sign up for the eTOC (electronic Table of Contents) service, which will deliver each new issue's, table of contents via e-mail. The web site also provides access to information about the journal (such as Instructions to Authors, the Editorial Board, and subscription information), as well as access to the Marine Biological Laboratory's web site and other Biological Bulletin electronic publications. The free trial period for access to The Biological Bulletin online has ended. Individuals and institutions who are subscribers to the journal in print or are members of the Marine Biological Laboratory Corporation may now activate their online subscriptions. All other access (e.g., to Abstracts, eTOCs, searching, Instructions to Authors) remains freely available. Online access is included in the print subscription price. For more information about subscribing or activating your online subscription, visit . http://www.biolbull.org ONLY THING NOT YOU. BX61 MOTOR RESEARCH MICROSCOPE. Brighter multi-color fluorescence - exquisitely balanced - achieved easier and faster than ever thought possible. D)C with the optimal balance of contrast and resolution, ck solid stability "Dazing 3D and ose imaging. button controls and Z-Drive. Automatically moves the stage up or down in increments as fine as 0.01pm. MOTORIZED reflected light illuminator. Automatic control of 6 filter cubes accommodates multi-color fluores- MOTORIZED sextuple revolving nosepiece with slider slot for Nomarski DIG. MOTORIZED filter wheels. Exchange excitation, emission and transmitted light filters with the press of a button. MOTORIZED condenser. Automatically adjusts condenser for bright- field, darkfield, phase contrast and DiC applications. It's all about imaging. And motorization lets you do things you've never been able to do before with greater ease, speed and affordability. ROCKET SCIENCE: 800 446 5967 olympusamerica.com/microscopes Cover Dinoflagellates are a diverse and ecologically im- portant group of unicellular protists. Some of them are free-living, photosynthetic or heterotrophic con- stituents of the plankton; others are symbiotic. Dinoflagellates in the genus SvnihioJiniiuu. com- monly called zooxanthellae, are intra- or intercellu- lar syrnbionts of diverse marine invertebrates, in- cluding foraminiferans, sponges, cnidarians. and molluscs. The image on the cover shows a throng of Symhio- i/iniiiiu ku\\'ci",iitii cells that were isolated from their symbiotic host, the Hawaiian stony coral Montipora cupitata ( = M. verrucosa) 1 ; the coral appears in the inset. In 1987, on the basis of cytological evidence, R. J. Blank 2 speculated that the vegetative cells of M. vermcosii are haploid; but this finding was never corroborated. The question of ploidy is important, Muraj>os, J. E. 1995. Revised checklist of extant shallow-water stony coral species from Hawaii (Cnidaria: Antho/oa: Scleructiniu). Bishop Museum Occasional Papers 42: 54-55. Blank, R. .1. 1987. Cell architecture of the dinotlagellate Symhio- diniwn sp. inhabiting the Hawaiian stony coral Montipora vcrnicosa. Mar. Biol. 94: 143-155. for it is central to our understanding of genome evolution and population genetics. Now. for the first time, the methods of molecular genetics have been applied to the problem of hap- loidy in dinoflagellates. In this issue of The Biolog- ical Bulletin (p. 10). Scott R. Santos and Mary Alice Coffroth report that vegetative cells of Symhio- iliniuin clade B symbiotic with gorgonians (sea fans and sea whips), and cultured cells from a range of hosts and locations, are haploid. Moreover, since Symbiodinium is monophyletic. S. kawagutii and other members of the genus must also be haploid. The Svnihiotlininiu cells on the cover are about 10 jam in diameter; they were photographed by Scott R. Santos (State University of New York at Buf- falo). The photograph of Montipora capitcita was taken by Frank Stanton (University of Hawaii) at a depth of about 1.5 meters; the coral is about 1 m in diameter. Materials and information for the cover and legend were provided by Fenny Cox (Univer- sity of Hawaii). The cover was designed by Beth Liles, Marine Biological Laboratory, Woods Hole, Massachusetts. THE BIOLOGICAL BULLETIN FEBRUARY 2003 Editor Associate Editors Section Editor Online Editors Editorial Board Editorial Office MICHAEL J. GREENBERG Louis E. BURNETT R. ANDREW CAMERON CHARLES D. DERBY MICHAEL LABARBERA SHINYA INOUE, Imaging and Microscopv JAMES A. BLAKE, Keys to Marine Invertebrates of the Woods Hole Region WILLIAM D. COHEN, Marine Models Electronic Record and Compendia PETER B. ARMSTRONG JOAN CERDA ERNEST S. CHANG THOMAS H. DIETZ RICHARD B. EMLET DAVID EPEL KENNETH M. HALANYCH GREGORY HINKLE NANCY KNOWLTON MAKOTO KOBAYASHI ESTHER M. LEISE DONAL T. MANAHAN MARGARET MCFALL-NGAI MARK W. MILLER TATSUO MOTOKAWA YOSHITAKA NAGAHAMA SHERRY D. PAINTER J. HERBERT WAITE RICHARD K. ZIMMER PAMELA CLAPP HINKLE VICTORIA R. GIBSON CAROL SCHACHINGER WENDY CHILD The Whitney Laboratory, University of Florida Grice Marine Laboratory. College of Charleston California Institute of Technology Georgia State University University of Chicago Marine Biological Laboratory ENSR Marine & Coastal Center, Woods Hole Hunter College. City University of New York University of California, Davis Center of Aquaculture-IRTA, Spain Bodega Marine Lab., University of California. Davis Louisiana State University Oregon Institute of Marine Biology, Univ. of Oregon Hopkins Marine Station. Stanford University Auburn University, Alabama Millennium Pharmaceuticals. Cambridge. Massachusetts Scripps Inst. Oceanography & Smithsonian Tropical Res. Inst. Hiroshima University of Economics. Japan University of North Carolina Greensboro University of Southern California Kewalo Marine Laboratory, University of Hawaii Institute of Neurobiology, University of Puerto Rico Tokyo Institute of Technology, Japan National Institute for Basic Biology, Japan Marine Biomed. Inst., Univ. of Texas Medical Branch University of California, Santa Barbara University of California, Los Angeles Managing Editor Staff Editor Editorial Associate Subscription & Advertising Administrator Published by MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASSACHUSETTS http://www.biolbull.org CONTENTS VoLUMH 204, No. I: FEBRUARY 2003 SYMBIOSIS AND PARASITOLOGY Nixon, Julie E. J., Jessica Field, Andrew G. McArthur, Mitchell L. Sogin, Nigel Yarlett, Brendan J. Loftus. and John Samuelson Iron-dependent hydrogenases of Kntamix'bn lii^tnlytirn and (iitndin laniblia: activity of the recombinant ent- amoebic en/yme and evidence for lateral gene trans- fer . CELL BIOLOGY Santos, Scott R., and Mary Alice Coffroth Molecular genetic evidence that dinoflagellates be- longing to the genus Synibiodinium Freudenthal are haploid Coursey, Yvonne, Nina Ahmad, Barbara M. McGee, Nancy Steimel, and Mary Kimble Amebocyte production begins at stage 18 during em- bryogenesis in l.nnnlus /iiil\/>/ii'inii\. the American hoist-shoe nab . NEUROBIOLOGY AND BEHAVIOR 10 DEVELOPMENT AND REPRODUCTION Walker, Anna, Seichi Ando, and Richard F. Lee Synthesis of a high-density lipoprotein in the devel- oping blue crab (CaUinectes mpidus) 50 McBride, Richard S., and Paul E. Thurman Reproductive biology o( Hemiramphus brasiliensis and H. balao (Heniiramphidae): maturation, spawning frequency, and fecundity 57 Raskoff, Kevin A., Freya A. Sommer, William M. Ham- ner, and Katrina M. Cross ( lollection and culture techniques for gelatinous zoo- plankton 68 PHYSIOLOGY AND BIOMECHANICS Gainey, Louis F., James C. Walton, and Michael J. Greenberg Branchial musculature of a venerid clam: pharmacol- ogy, distribution, and innervation 81 ECOLOGY AND EVOLUTION Mann, Roger, and Juliana M. Harding Salinity tolerance of larval Rapnna venosa: implica- tions for dispersal and establishment of an invading predatory gastropod on the North American Atlantic coast. . 96 Buskey, Edward J., and Daniel K. Hartline High-speed video analysis of the escape responses of the copepod Acartni ttmvi to shadows 28 McGaw, I. J. Behavioral thermoregulation in Hem'tgi'tipsus mains. the amphibious purple shore crab 38 RESEARCH NOTE Aoyaina, Jun, Sam Wouthuyzen, Michael J. Miller, Ta- dashi Inagaki, and Katsumi Tsukainoto Short-distance spawning migration of tropical fresh- water eels 104 THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is published si\ limes a year by the Marine Biological Laboratory, 7 MBL Street, Woods Hole, Massachusetts 02543. Subscriptions and similar matter should be addressed to Subscription Administrator. THE BIOLOGICAL BULLETIN. Marine Biological Laboratory, 7 MBL Street, Woods Hole, Massachusetts (12543. Subscription includes both print and online journals. Subscription per year (six issues, two volumes): $280 for libraries; $105 for individuals. Subscription per volume (three issues): $140 for libraries; $52.50 for individuals. Back and single issues (subject to availability): $50 for libraries: $20 for individuals. Communications relative to manuscripts should be sent to Michael J. Greenberg. Editor-in-Chief, or Pamela Clapp Hinkle. Managing Editor, at the Marine Biological Laboratory. 7 MBL Street, Woods Hole, Massachusetts (12543. Telephone: (508) 289-714". FAX: 508-289-7422. E-mail: pclapp/ Entamoeba histolytica anil Trichomonas \aginalix Fe-hydrogencutes Sample Hydrogenase activity (nmol/min/mg of protein)'' Bacteria transformed with E. histolyticti Fe-hydrogenase Transfected E. hixiolytica with Fe- hydrogenase I ru humnmis Yu^nidlis T. viif>inalis + 0.47 mg E. histolyticti lysate" 36(2) 3.5 0.7 (3) 167 32 (41 114(2) a Averages +/ standard deviations, where possible. Number of deter- minations in parentheses. '"Calculated A", = 0.56 mg (amount of E. lii\in/yhm Usate in mg of protein to cause 50% reduction of 7". ru.i;im///.v hydrogenase activity). 1 2 3 SOObp the T. vaginalis Fe-hydrogenase was greater than that of the recombinant entamoebic GST-entamoebic Fe-hydrogenase and was inhibited by a lysate of non-transfected entamoebas (Table 1). This may explain why it was difficult to detect Fe-hydrogenase activity in lysates of nontransfected ent- amoebas, even though Fe-hydrogenase 1 mRNAs were identified from them by RT-PCR (next section). Cultured entamoehas and giurdiiifi exprexx mRNAs en- coding short Fe-hydrogenases. We isolated an fe-hydroge- nase gene of G. lanihlia. because we have frequently com- pared the fermentation enzymes of this diplomonad with those of E. hixtolytieu (Rosenthal et ul.. 1997; Field et ul., 2000; Nixon et ul., 2002). A search of the contigs predicted from the G. lamblia shogun sequences suggested that this gene, which predicts a short Fe-hydrogenase. is the only hydrogenase gene present within the giardial genome. Like the entamoebic Fe-hydrogenase 1. the predicted giardial Fe-hydrogenase lacked an N-terminal organelle-targeting sequence and had two ferredoxin-like iron-sulfur centers and a hydrogenase iron-sulfur center like those present in the short Fe-hydrogenases of T. vugimilis, Desulfovibrio sp., and Clostridiu sp. iCammark. 1992; Thompson et ai, 1994; Bui and Johnson. 1496; Homer el ul., 1996; Nicole! et ai. 1999). RT-PCR showed that cultured entamoebas and giar- dias contain mRNAs, which encode short Fe-hydrogenases (Fig. 1 A, B). Negative controls without RT showed that the RT-PCR was not amplifying DNA from the extracts of cultured entamoebas and giardias. Because the giardial con- tigs predicted only one Fe-hydrogenase, which is expressed, it is likely that the hydrogenase activity recently detected in cultures of giardias derives from this enzyme (Lloyd and Harris. 2002). In contrast, entamoebas appear to have a second long hydrogenase (see next section), so if entamoe- bic hydrogenase activity is present, it might derive from one or more enzymes. These results suggest the possibility that entamoebas and giardias use protons as electron acceptors 3 4 SOObp B Figure 1. Agarose gels of ethidium-stained RT-PCR products from entamoebic and giardial mRNAs. Images are reversed for clarity of reproduction. (A) RT-PCR of amoebic mRNAs encoding malic enzyme (lane 1). Fe-hydrogenase (lane 2). alcohol dehydrogenase E (lane 3). and ferredoxin (lane 4). Size markers are shown in lane 5. Lanes 6-9 are the negative controls for malic enzyme, hydrogenase. ADHE. and ferredoMii. respectively. (B) RT-PCR of giardial mRNAs encod- ing ADHE (lane 3). Fe-hydrogenase (lane 4). and malic enzyme (lane 5). A negative control (no RT) tor Fe-hydrogenase is shown in lane d. A positive control for Fe-hydrogenase. using Giardia lumhlia WB strain DNA. is shown in lane 2. Size markers are shown in lane I. ACTIVITY AND ORIGIN OF THE ENTAMOEBIC AND GIARDIAL Ft^-HYDROGENASES Eh2 MSTQLTPLRNKIISEWKCFKSGRFIEDIDKLPTILTDGDGWKPTSKFVHSREQEEGIYR Td IKREILVRIAKLQFEGKLQEGVHYIPREMVPRN.STPI.RCCIFHDR. .EIMR Bf VRHKLLAKLVNLWKENKLTNEIDRLPIELSPRR.SRPLGRCCIHKER. .AVYK Eh2 EKVLSVLGF.VDGEYDDITPLHVYAQKALERT.SLHEPVFGISQKGCNKCHFNGYFVTQA Td HRVIARLGCSLENYDEEKT . LAQFAKEALERE . KPTWPMLTVLDEACNACVKSKYMITNA Bf YKLFPLLGFDMTDETDELTSLSEYARQALERKNKQKENILCVIDEACSSCVQVNYEVTNL Eh2 CEGCTSRPCSVNCPKKCISFGEDGRAVINQNNCIKCGRCYKFCPYGAIISKSVPCVKACP Td CQACVARPCMMNCPKTAIAIS . GGRARIDEEKCINCGICLKNCPYHAVIKIPVPCEEACP Bf CRGCVARSCYMNCPKDAIRFRKNGQAKIDHDACISCGKCHQSCPYHAIVFIPVPCEEACP Eh2 CGAMLDSPEGVKTIDFEKCINCGGCMRACPFGAILPRSNLIDVLK. ILPTKKVVACPAPS Td VGAISKDENGKERIDYHKCIFCGNCMRECPFGAMMDKGQIVDVIKHLMSGKKVSALYAPA Bf VKAISKDENGIEHIDESKCIYCGKCLNACPFGAIFEISQAFDVLEGIRSGEKMIAIPAPS X x Eh2 IAAHFGKYDLALVSGGLIQVGFTSVEDVSYGADLCALNEAKEFEERIVKNKKDFMTTSCC Td VAAQF . KAVPGQLESALKKAGFNKVWEVAIGADITADREASEFEERMEHGHI . LMTTSCC Bf ILGQF. NTSIEAVYGALRQMGFADWEVAQGAMDTVSHEAAELKEKLEEGQP . FMTTSCC o Eh2 PAYINAINKHMPELKENVSHTPTPMHFATQAVKDRDQETVTVFIGPCNAKRWETLQDSTT Td PAYVRAVKKHVPALVPCISDTRSPMHYTAELAKKEDPDCVTVFIGPCLAKRREGLEDEFV Bf PSYIELVNKHIPGMKPYVSSTGSPMYYAARIAKERHPDAKIVFIGPCVAKRKEARRDECV o Eh2 DYCLTFDEIFGLFEGSGIDLSKVQPYTFVDKAHKEGKIFAVSGGVASAVASLLPKEVPDG Td DYVLSIEELGALLTAKEIDISKEEALPGKITPTSSGRGFAASGGVAEAVRVRL.KKPEN. Bf DYILTFEEMASIFEGLDIQLEQTQPFSVLYTSVREAHGFAQAGGVMGAIKAYLGEEAKK. Eh2 VIKPTIIDGFSQENFKRLKNFKKNI TGNLVEVMVCEGGCAYGPGCPGLNTP Td . LRPVLINGLNKEGMKQLASYGKIQSGELPHDSSTPNLVEVMSCEGGCIGGP BF . FSAIQVSDLNKKNIGLLRAAAKTG KAQGQFIEVMACEGGCISGP o o Eh2 ATSAKIKIAVDKMEAHPEGRWVGLPNSQIKPIKVEN 504 Figure 2. Alignment of the predicted Entumnebu liixralvticu long Fe-hydrogenase 2 (Eh2) with predicted long Fe-hydrogenases ot Treponema ilcniici>lu (Td) and Bacteroides fi'ii^His (Bf). Conserved Cys residues, which are shaded, include those that coordinate putative |4Fe-4S] iron-sulfur centers (marked with x's) and those that coordinate putative hydrogenase iron-sulfur centers (marked with o's). Other conserved Cys residues, which may be involved in coordinating iron-sulfur centers, are marked with asterisks. Amino acids at the beginning and end of the conserved Fe-hydrogenase domain are underlined. when the organisms are growing under strictly anaerobic conditions in the bowel lumen (Brown et ai. 1998; Huston and Petri. 2001; Lloyd and Harris, 2002). E. histolytica has a hydrogenase 2 gene encoding a long Fe-hydrogenase. The assemblies of the shotgun sequences of the E. histolytica genome predicted a long Fe-hydroge- nase 2 (Fig. 2) in addition to the short Fe-hydrogenase 1. The entamoebic Fe-hydrogenase 2 was 504 amino acids long and had an N-terminal sequence, which included pos- itively charged Lys and Arg that are often present at the N-termini of organellar proteins (Claros and Vincens, 1996). In addition, the N-terminus of Fe-hydrogenase 2 contained Ser and Leu residues, which are present at the N-termini of crypton and hydrogenosomal proteins (Bui et til.. 1996; Mai et /., 1999). However, in the absence of experimental evidence, we cannot be sure that the entamoe- bic long Fe-hydrogenase is targeted to the crypton. The entamoebic Fe-hydrogenase 2 was much more sim- ilar (>38% amino acid identities) to predicted long Fe- hydrogenases of Bucteroides frag His and Treponema den- ticola than to short Fe-hydrogenases of entamoebas. giar- dias. trichomonads. and other anaerobic bacteria (<28% amino acid identities; Fig. 2). The entamoebic Fe-hydroge- nase 2 and the predicted long Fe-hydrogenases of B. fragilis NIXON ET AL. 96/95 |C. reinhardtii 1 [C. reinhardtii 2\ 96/1 OO T. maritima 1 |P/romycessp.| | N. frontalis\ 99/100 */100 80/65 T. gallinae] L \T. vaginalis 2\ T. vaginalis 3 \T. vaginalis 4\ M. elsdenii T. maritima 2 100/100 |S. barkhanus] 99/691 IG. lamblia \ |E histolytica 1\ 95169 96/99 */99 T. tengcongensis 1 T. maritima 3 T. denticola B. fragilis R. albus C. perfringens 1 C. acetobutylicum 98/100 c~ C. perfringens 2 D. vulgaris 1 r V. tengcongensis 2 "jr C. thermocellum * E. acidaminophilum 98/100 64/* C. saccharobutylicum C. pasteurianum C. perfringens 3 I/V. ovalisl D fructosovorans D. desulfuricans 1 D. vulgaris 2 D. desulfuricans 2 0.1 ACTIVITY AND ORIGIN OF THE ENTAMOEBIC AND GIARDIAL FE-HYDROGENASES and T. dcnticola each contained Cys residues that likely coordinate two ferredoxin-like [4Fe-4S] iron-sulfur centers (marked with x's in Fig. 2) and hydrogenase iron-sulfur centers (marked with o's). which have previously been identified in structures of short and long Fe-hydrogenases (Peters et ai. 1998; Nicolet et ai, 1999). In addition, the predicted entamoebic Fe-hydrogenase 2 had eight other N-terminal Cys residues, which aligned with those of the bacteroides and treponema long Fe-hydrogenases (marked with asterisks). Although these Cys residues probably co- ordinate other iron-sulfur centers, they remain unidentified, because they do not align with the N-terminal iron-sulfur centers of the long Fe-hydrogenase of C. pastenriannni. which has been crystallized (Peters et a/.. 1998). The entamoebic and giardial short fe-hydrogenase I genes appear to share a common ancestn: while the ent- tiinoehic lout; fe-hydrogenase 2 gene appears to have been laterally transferred from a prokaryote. Phylogenetic trees of Fe-hydrogenases from eubacteria and eukaryotes are star-shaped and contain few basal nodes that are strongly supported (Fig. 3). This result suggests that the Fe-hydro- genases are widely divergent and that little phylogenetic signal remains. For example, Fe-hydrogenases of closely related eukaryotes either trichomonads, green algae (Chlamydomonas reinhanltii, Scenedesinus obliqiiiis. and Chlorella fused), or chytrid fungi (Piromyces sp. and Neo- callimastix frontalis) each grouped together, but Fe-hydrog- enases of unrelated eukaryotes did not group together. In particular, our analysis does not support recent conclusions that hydrogenases of trichomonads are monophyletic with those of chytrid fungi ( Voncken et at., 2002) or with those of E. histolytica and 5. barkhanus (Horner et cii, 2000). The short fe-hydrogenase genes of G. Iambi ia, S. barkha- nus, and E. histolvtica appear to share a most recent com- mon ancestry, although a particular bacterial donor was not identified. Remarkably, the short Fe-hydrogenase of G. Iain- Miii was more similar to that off. histol\ticu than to that of S. harkhaims. Because G. lumblia and S. barkhanus are diplomonads, which share a recent common ancestor in phylogenetic trees of rRNA and proteins (Sogin and Silber- man, 1998), a possible explanation of these results is that the E . histolytica fe-hydrogenase gene was laterally trans- ferred from a diplomonad (Rosenthal et ai. 1997; Doolittle, 1998, 1999; Miiller, 1998; de Koning ft til.. 2000; Field et al.. 2000; Nixon et ai. 2002). This lateral gene transfer would not have occurred recently, because the Fe-hydroge- nases of entamoebas and giardias showed only a 40% amino acid identity with each other, and each fe-hydrogenase gene has the codon usage of its host. Alternatively, the diplomonad-E. histolytica sub-clade could be incorrectly rooted by the long branch connecting it to the remainder of the tree. The common ancestry of genes encoding the E. histo- lytica long Fe-hydrogenase 2 and those of B. fragilis and T. denticola is strongly supported. This appears then to be an example of lateral gene transfer, as Entamoeba is not a close relative of either of these eubacteria (Rosenthal et ill., 1997; Doolittle. 1998, 1999; Muller. 1998; de Koning et ai. 2000; Field et ai, 2000; Nixon et ai, 2002). There was weak support for the pairing of Fe-hydrogenases of the ciliate N. avails and Desulfavibrio sp., as has been previously noted (Horner et ai. 2000; Voncken et ui, 2002). This suggests that the ciliate hydrogenase was derived by lateral gene transfer, but does not prove it. Conclusions This is the first time that an Fe-hydrogenase from a protist has been expressed as a GST-fusion protein in bacteria. This is also the first time that an fe-hydrogenase gene (encoding the long hydrogenase of entamoebas) has been inferred to have been laterally transferred from a bacterium, although numerous genes encoding fermentation enzymes (e.g.. al- cohol dehydrogenases, malic enzyme, and acetyl-CoA syn- Figure 3. Phylogenetic relationships of Fe-hydrogenases, inferred using a distance matrix generated by the Dayhoff+I + r model and the Fitch-Margoliash algorithm. Bootstrap values obtained using PUZZLEBOOT and PROTPARS, respectively, are shown at the relevant nodes. Bootstrap values below 50% are marked with an asterisk if the other bootstrap value is >50%. If both bootstrap values are below 50%, neither is marked. The scale bar indicates estimated sequence divergence per unit branch length. Sequences from eukaryotes. which are boxed, were from Chlamydomonas reinhunltii 1 and 2 [accession # 16945126 and 18026272]: Chlorella fusca [21732235]; Enuinioebti histolytic.i 1 and 2 [9963974 and AY172963]; Giardia Uimhlin [13506793]: Neocal- limastix fruntalis [19547863]: Nyctnlherus avulis [4034791]; Piromyces sp. [19548180]: Scenedesinus obliquus 1 and 2 [12581498 and 13311187]; Spinmitcleiix harkhamis [11127703]; Trichoimmas gallinae [19548182]; Trk-hiinimiiix vuximilix 1.2, 3, and 4 [19547859. 1 1 127701. I 171 1 17. and 1345094]. Eubacterial sequences were from Bacteroides fragilis [unfinished microbial database]; Closlridium acetobutylicum [15896476]; Clostridiitm fiiiMcnrUmiim [557064]; Chsiridiiini pcrfrin^ns I. 2. and 3 [18311557, 18309258. and 1831 1328]; Closlridium saccharobutylicum [488597]; Clo.ilridiiim thermncellnm [4927278]; Desulfavibrio fructosovorans [1914864]; Dfsiilfovibrio destilfiiricans 1 and 2 [4930044 and 13022069]; Desulfovihrio vulgaris 1 and 2 [97381 and 66319, respectively]; Eiihucieriiini acidominophilum [14250935); Megasphaera elsdenii [66509851; Ruminococcui albus [unfinished microbial database); Theriiuiaiwerubiicter teiif-coiitn'iM* 1 and 2 [20807184 and 20515894]; Thermotoga marilimu 1. 2. and 3 [15644177, 7433127. and 4981985]; and Treponema Jenticnlii [unfinished microbial database]. 8 NIXON ET AL. thases) appear to have been laterally transferred from pro- karyotes to amoebas and giardias (Rosenthal et oi. 1997; Field et til.. 2000; Nixon et til.. 2002). Although the evi- dence is weak, this may also be the first time that a gene (encoding the short hydrogenase of entamoebas) has been inferred to have been laterally transferred from another protist. Because the hypothesized lateral gene transfer would probably have occurred after the acquisition of the fe-hydrogenase gene by the diplomonad lineage, this par- ticular result does not disprove the hydrogen hypothesis (Martin and Miiller, 1998). However, the failure to demon- strate that the eukaryotic Fe-hydrogenases share a common ancestry, or to identify an a-proteobacterial donor for these eukaryotic fe-hydrogenase genes (Horner et ai. 2000), dampens our enthusiasm for the hydrogen hypothesis. These results suggest that the mitochondria! endosymbiont was selected for a property other than hydrogen production (e.g.. its ability to consume oxygen) (Andersson and Kur- land, 1999) and that the presence of Fe-hydrogenases and other fermentation enzymes of microaerophilic eukaryotes may reflect a secondary adaptation to their anaerobic envi- ronment (Rosenthal et til.. 1997; Doolittle, 1998, 1999; de Koning et til., 2000; Field et til.. 2000; Lloyd and Harris, 2002; Nixon et til.. 2002). Acknowledgments This work was supported by NIH grants ( AI33492 to J.S., AI43273 to M.L.S.. and AI46516 to B.J.L.). Literature Cited Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. ,). Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389-34(12. Andersson, S. G., and C. G. Kurland. 1999. Origins of mitochondria and hydrogenosomes. (.'HIT. Opin. Microhiol. 2: 535-541. Bapteste, E., H. Brinkmann. J. A. Lee, D. V. Moore, C. \V. Sensen, P. Gordon, I,. Duruflc, T. Gaasterland, P. Lopez, M. Miiller. and H. Philippe. 20(12. The analysis of 100 genes supports the grouping of three highly divergent amoebae: DictytKtcliiini. Eniinnnchii. and A/II.V- tif>innoehti. Piot. Nail. Acini. Sci. USA 99: 1414-141'). Bradley, P. J., C. J. Lahti, E. Plumper, and P. J. Johnson. 1997. Targeting and trunslocation of proteins into the hydrogenosome of the prolix! Tnehoiiionii\: similarities with mitochondria! protein import. EMBO J.\f>: 3484-3443. Brown, D. M., J. A. Upcroft, M. R. Edwards, and P. Upcroft. 1998. Anaerobic bacterial metabolism in the ancient eukaryote Gil\phei>nis embryos consist of relatively few cell types. In the ventral region and the growing appendages, the epider- mal cells are typically columnar in shape, while the under- lying mesodermal cells are flattened in appearance. Coelo- mic cavities have formed within some of the developing appendages, and within these we occasionally observe cells that have a fibroblast-like appearance. Also within the ven- tral plate are occasional cells having dark-staining cytoplas- mic inclusions. These cells are most likely muscle precur- sors. In later stage embryos and larvae, cells with similarly stained inclusions are often seen adjacent to developing muscles. As one moves dorsally away from the ventral plate, there is a gradual transition in the epidermal cells from columnar through cuboidal to a flattened appearance. Similarly, the mesodermal cell layers in the dorsal region are very thin flat sheets that are often difficult to discern. By stage 18, the extension of the mesoderm over the central yolk mass appears to be complete or nearly so. The central region of the embryo is filled with yolk, within which are distributed numerous yolk nuclei. No evidence of internal organs is seen at this stage, although hemocoel cavities have begun to form. The cavities are located between the meso- dermal cell layers, and within the cavities granular pro- amebocytes and elements of connective tissue are fre- quently observed. Also located in the dorsal regions of the embryo are cells that appear to be producing chitin-like material (based on the staining properties of the material). Finally, as mentioned before, the lateral organs, composed of distinct goblet-shaped cells, have begun to develop. Thus we are able to identify at most seven to eight distinct cell types in the stage 18 L polyphemus embryos. How this compares with T. triilenUitns embryos at the same stage of development remains to be determined. Do the yolk nuclei represent u pool of multipotent cell precursors? In contrast to many arthropods, horseshoe crabs retain significant numbers of yolk nuclei after cellular blastoderm formation (Kishinouye. 1893; Kingsley, 1892, 1893; Kimble et al., 2002). The yolk nuclei persist throughout embryonic development. During the mid- to late stages of embryogenesis. some yolk nuclei probably function as vitel- lophages. After hatching, the residual yolk is incorporated into the developing midgut and digestive diverticulum. a network of blind-end caeca that extends throughout the prosoma. We have previously shown that some of the re- sidual yolk nuclei cellularize to form the columnar epider- mal lining of the digestive caeca, while others form a layer of flattened cells that surround the individual caeca (Kimble et til.. 2002). In most arthropod species, the yolk nuclei or yolk cells function only as vitellophages, degenerating before the end of embryonic development (Anderson, 1973; Campos-Or- tega and Hartenstein, 1997). In the terrestrial chelicerates, spiders and scorpions, most of the cleavage nuclei partici- pate in blastoderm formation. Subsequently some cells re- populate the yolk mass, where they function as vitello- phages. Eventually the vitellophages migrate to the surface of the yolk mass and form the endoderm epithelium (Ander- son. 1973). Thus, a role for the yolk nuclei or vitellophages in formation of the gut endoderm appears to be common to most if not all chelicerates. However, participation in for- mation of the mesodermal components of the gut is appar- ently unique to the Xiphosura. If. as suggested here, some yolk nuclei cellulari/.e and differentiate as amebocytes dur- ing late embryogenesis, it would suggest that retention ot large numbers of yolk nuclei in horseshoe crab embryos provides the embryos with a pool of undetermined nuclei thai can be utili/ed in a variety of distinct tissues during development. Acknowledgments We thank undergraduate students Evelyn Wurth, Carrie Ottoson, Nicole Tremblay, Kimberly Demon. Patrick Mella, AMEBOCYTES IN L1MVLUS EMBRYOS 27 and Victoria Davis for assistance with sectioning and stain- ing of the embryos. These studies were supported in part by grants to MK from the USF Research Council, and from NOAA, Office of Sea Grant, Department of Commerce. Grant # NA76RG-0120. The U.S. government is authorized to produce and distribute reprints for governmental pur- poses not withstanding any copyright that may appear hereon. YC was supported by grants/scholarships from Sigma Xi. Sigma Delta Epsilon Graduate Women in Sci- ence. Aylesworth/Old Salt, Sea Space, The American As- sociation of University Women, and the Florida and Tampa Garden Clubs. NA was supported in part by the McNair Post-Baccalaureate Achievement Program. Literature Cited Agarwala, K. L., S. Kawahata, Y. Miura, Y. Kuroki. and S. Iwanaga. 1996. Limulus intracellular coagulation inhibitor type 3. Purification, characterization. cDNA cloning, and tissue localization. J. Biol. Chem. 271: 23,768-23.774. Anderson, D. T. 1973. Embryology und Phytogeny in Annelids and Arthropods. Pergamon Press. New York. 495 pp. Armstrong, P. B. 1985. Adhesion and motility of the blood cells of Limulus. Pp. 77-124 in Blood Cells of Marine Invertebrates: Experi- mental Systems in Cell Biology and Comparative Physiology, W. D. Cohen, ed. Alan R. Liss. New York. Armstrong, P. B., and F. R. Rickles. 1982. Endotoxin-induced degran- ulation of the Limulus amebocyte. Exp. Cell Res. 140: 1 5-24. Bang, F. B. 1956. A bacterial disease of Limulus po/vphemus. Bull. Johns Hopkins H,,sp. 98: 325-351. Bang, F. B. 1979. Ontogeny and phylogeny of response to gram-negative endotoxins among the marine invertebrates. Pp. 109-123 in Biomedi- cal Applications of the Horseshoe Crab (Limulidae), W. D. Cohen, ed. Alan R. Liss. New York. Campos-Ortega, J. A., and V. Hartenstein. 1997. The Embrvonu Development of Drosophila melanogaster. 2nd ed. Springer-Verlag. Berlin. 405 pp. Dumont, J. N., E. Anderson, and G. Winner. 1966. Some cytologic characteristics of the hemocytes of Limulus during clotting. J. Morphol. 119: 1X1-208. Hilly, J. B., and D. G. Gibson. 1989. Culture of amebocytes on opened gill lamellae of the horseshoe crab. Limulus po/yphemu.s. (abstract). Am. Zoo/. 29: 1 1 2A. Iwanaga, S. 20(12. The molecular basis of innate immunity in the horse- shoe crab. Curr. Opin. Immunol. 14: 87-95. Kimble, M., Y. Course), N. Ahmad, and G. W. Hinsch. 2002. Behav- ior of the yolk nuclei during embryogenesis, and development of the midgut diverticulum in the horseshoe crab. Limulus polyphcmiis. In- vcrtcbr. Biol. 121: 365-377. Kingsley, J. S. 1892. The embryology of Limulus. ./. Morphol. 1: 35-68. Kingsley. J. S. 1893. The embryology of Limn/us. Part II. ./. Morphol. 8: 195-268. Kishinouye, K. 1893. On the development of Limulus longispina. ./. Coll. Sci. Imp. Univ. Japan 5: 53-100. Levin, J. 1985a. The history of the development of the Limulus amebo- cyte lysate test. Pp. 3-28 in Bacterial Endotoxins: Structure. Biomed- ical Significance, and Detection with the Limulus Amebocyte Lysate Test. Prog. Clin. Biol. Res. 189. Levin, J. 1985b. The role of amebocytes in the blood coagulation mech- anism of the horseshoe crab Limulus polvphcnn/s. Pp. 145-163 in Blood Cells of Marine Invertebrates: Experimental Svstems in Cell Biology and Comparative Physiology, W. D. Cohen, ed. Alan R. Liss. New York. Liang, P., T.-K. Cheng, Y.-Q. Wu, and W.-H. Wu. 1990. Ultrastruc tural observations on hemocytopoiesis in embryos of the horseshoe crab. Tachvpleus tridenlatus. Proceedings of the XII International Congress of Electron Microscopy. August 12-18, 1990, Seattle, WA. 3: 506-507. Miura, Y., S.-I. Kawabata, Y. Wakamiya, T. Nakamura, and S. Iwanaga. 1995. A Limulus intracellular coagulation inhibitor type 2. Purification, characterization. cDNA cloning, and tissue localization. J. Biol. Chen,. 270: 558-565. Miyata, T., M. Hiranaga, M. Umezu, and S. Iwanaga. 1984. Amino acid sequence of the coagulogen from Limulus po/yphemnx hemocytes. J. Biol. Chem. 259: 8924-8933. Ornberg, R. L. 1985. Exocytosis in Limulus amebocytes. Pp. 127-142 in Blood Cells of Marine Invertebrates: Experimental Systems in Cell Biology and Comparative Physiology. W. D. Cohen, ed. Alan R. Liss, New York. Sawada, T., and S. Tomonaga. 1996. The immunocytes of protostomes and deuterostomes as revealed by LM. EM and other methods. Adv. Comp. Environ. Physiol. 23: 9-40. Sekiguchi, K. 1973. A normal plate of the development of the Japanese horseshoe crab, Tachvpleus tridentatits. Sci. Rep. Tokvo kvoikti Daigaku Sect. 15: 153-162. Sekiguchi. K. 1988. History of the study. Pp. 1-9 in Biology of Horse- shoe Crabs, K. Sekiguchi. ed. Science House. Tokyo. Sekiguchi, K., Y. Yamamichi, and J. D. Costlow. 1982. Horseshoe crab developmental studies. 1. Normal embryonic development of Limulus polyphemu.s compared with Tachvp/ens tndentatiis. Pp. 53-73 in Physiology and Biology of Horseshoe Crabs: Studies on Normal and Environmentally Stressed Animals, J. Bonaventura. C. Bonaventura. and S. Tesh, eds. Alan R. Liss, New York. Sekiguchi, K., Y. Yamamichi, H. Seshimo, and H. Sugita. 1988. Nor- mal development. Pp. 133-181 in Bio/ogv of Horseshoe Crabs, K. Sekiguchi, ed. Science House. Tokyo. Yeager, J. F., and O. E. Tauber. 1935. On the hemolymph cell counts of some marine invertebrates. Biol. Bull. 69: 66-70. Reference: Bio/. Bull. 204: 2X-37. (February 2003) 2003 Marine Biological Laboratory High-Speed Video Analysis of the Escape Responses of the Copepod Acartia tonsa to Shadows EDWARD J. BUSKEY 1 * AND DANIEL K. HARTLINE 2 1 Marine Science Institute. 750 Channel View Drive, Port Aransas, Texas 78373: and ' Bekesv Laboratory of Neurobiology, Pacific Biomcdical Research Center. Universitv of Hawaii at Manoa. 1W3 East-West Road. Honolulu. Hawaii Vf>S22 Abstract. The copepod Acartia tonsa exhibits a vigorous escape jump in response to rapid decreases in light intensity, such as those produced by the shadow of an object passing above it. In the laboratory, decreases in light intensity were produced using a fiber optic lamp and an electronic shutter to abruptly either nearly eliminate visible light or reduce light intensity to a constant proportion of its original inten- sity. The escape responses of A. tonsa to these rapid de- creases in visible light were recorded on high-speed video using infrared illumination. The speed, acceleration, and direction of movement of the escape response were quan- tified from videotape by using automated motion analysis techniques. A. tonsa typically responds to decreases in light intensity with an escape jump comprising an initial reori- entation followed by multiple power strokes of the swim- ming legs. These escape jumps can result in maximum speeds of over 800 mm s~' and maximum accelerations of over 200 m s~ 2 . In .4. tonsa. photically stimulated escape responses differ from hydrodynamically stimulated re- sponses mainly in the longer latencies of photically stimu- lated responses and in the increased number of power strokes, even when the stimulus is near threshold; these factors result in longer escape jumps covering greater dis- tances. The latency of responses of A. tonsa to this photic stimulus ranged from a minimum of about 30 ms to a maximum of more than 150 ms, compared to about 4 ms for hydrodynamically stimulated escape jumps. Average re- sponse latency decreased with increasing light intensity or increasing proportion of light eliminated. Little change was Received 7 June 2(102: accepted 26 November 2002. To whom correspondence should he addressed. H-m;iil: buskeyfs'utmsi. utexas.edu observed in the vigor of the escape response to rapid de- creases in visible light over a wide range of adaptation intensities. Introduction Planktonic copepods are an important link in marine food webs between microplankton and higher trophic levels. Copepods are well known for their vigorous escape re- sponses (e.g., Singarajah. 1969; Fields and Yen, 1997). which play an important role in predator avoidance (e.g.. Drenner et a/.. 1978; Viitasalo et al.. 1998). These escape responses can be elicited by both hydrodynamic (Hartline et al.. 1999; Kiorboe et al.. 1999; Lenz and Hartline, 1999) and photic stimuli (Buskey et al., 1986. 1987). Despite the scarcity of direct evidence that chemosensory stimuli, by themselves, can produce vigorous escape responses in cal- anoid copepods, there is evidence that such stimuli can cause copepods to exhibit changes in swimming activity (e.g.. Katona, 1973; Buskey, 1984). Predator-specific chem- icals have also been shown to alter vertical migration be- havior in freshwater zooplankton (e.g., Tjossem, 1990; Rin- gelberg, 1991) and marine crab larvae (Forward and Rittschof, 2000). but similar effects are yet to be demon- strated in marine copepods (e.g.. Bollens et al.. 1994). Light has long been known to have an important effect on the behavior of planktonic organisms, with much research emphasizing the effects of light on vertical migration (re- viewed in Forward, 1988). Photophobic responses of cope- pods are thought to play a role in planktonic predator-prey interactions both in terms of a predator-deterrent role of bioluminescence in dark-adapted copepods (Buskey and Swift. 1983, 1985) and in terms of a predator-avoidance role for copepods exposed to shadows in light-adapted copepods (Buskey et a/.. 1M86). Alterations in behavior of planktonic COPEPOD ESCAPE RESPONSES TO SHADOWS organisms in response to decreases in light intensity have been demonstrated in neritic calanoid copepods (Buskey et al.. 1987) and crab larvae (Forward. 1977). Recent studies have used strain gauges and high-speed video to provide the high temporal resolution necessary to describe, in detail, the kinetics of the escape responses of both tethered (Len/ and Hartline. 1999; Hartline H.M' parameters t<"' tululi A'mii/o ami malc\ <>/ Acartia lonsa cv/xwJ Jo plume Miniiili Parameter Females Males P Latency (ins) 68.2(1.8:29-159) 62.2(1.3:35-138) / Inverichrutes. W.H. Freeman. San Francisco. 1714 pp. (Ch. 19). Burkenroad, M. D. 1943. A possible function of bioluminescence. J. Mar. Res. 5: 161-164. Buskey, E. J. 1984. Swimming pattern as an indicator ul the roles of copepod sensory systems in the recognition ol fond. Mar. Biol. 79: 165-175. Buskey, K. .1. 1993. Annual pattern of micro- and mesozooplankton abundance and biomass in a subtropical estuary. J. Plankton Res. 15: 9(17-924. Buskey, E. J., and E. Swift. 1983. Behavioral responses of the coastal copepod Acurtiu Inidsonica to simulated dinoflugellale hiolumines- cence. / F..\p. Mar. Hi, !. '>/. 72: 43-5S. COPEPOD ESCAPE RESPONSES TO SHADOWS 37 Buskey, E. ,)., and E. Swift. 1985. Behavioral responses of oceanic zooplankum lo simulated bioluminescence. Biol. Bull, 168: 263-275. Buskey. E. J., L. Mills, and E. Swift. 1983. The effects of dinoflagellate bioluminescence on the swimming behavior of a marine copepod. Limnol. Oceanogr. 28: 575-579. Buskey, E. J., C. G. Mann, and E. Swift. 1986. The shadow response of (he estuarine copepod Acania tonsa. J. .v/>. Mar. Biol. Ecol. 103: 65-75. Buskey, E. J., C. G. Mann, and E. Swift. 1987. Photophobic responses of calanoid copepods: possible adaptive value. ./. Plankton Res. 9: 857-870. Buskey, E. J., P. H. Lenz, and D. K. Hartline. 2002. Escape behavior of planktonic copepods to hydrodynamic disturbances: high speed video analysis. Mar. Ecu/. Prog. Ser. 235: 135-146. Camhi. J. M.. and T. G. Nolen. 1981. Properties of the escape system of cockroaches during walking. J. Coinp. Phvsiol. 142: 339-346. Davis, A. D., T. M. Weatherby, D. K. Hartline, and P. H. Lenz. 1999. Myelin-like sheaths in copepod axons. Nature 398: 571. Drenner, R. W., J. R. Strickler, and W. J. O'Brien. 1978. Capture probability: the role of zooplankton escape in the selective feeding of planktivorous fish. J. Fish. Res. Board Can. 35: 1370-1373. Eaton, R. C., R. A. Bombardier!, and D. L. Meyer. 1977. The Mauthner-initiated startle response in teleost fish. J. Ev/>. Biol. 66: 65- M. Eaton, R. C.. R.K.K. Lee, and M. B. Foreman. 2001. The Mauthner cell and other identified neurons of the brainstem escape network of fish. Prog. Neumbinl. 63: 467-485. Elofsson, R. 1966. The nauplius eye and frontal organs of the non- Malacostraca (Crustacea). Sarsia 25: 1-128. Fields, D. M., and J. Yen. 1997. The escape behavior of marine cope- pods in response to a quantifiable fluid mechanical disturbance. / Plankton Res. 19: 1289-1304. Forward, R. B., Jr. 1977. Occurrence of a shadow response among brachyuran larvae. Mar. Biol. 39: 331-341. Forward, R. B., Jr. 1988. Diel vertical migration: zooplankton photo- biology and behavior. Oceanogr. Mar. Biol. Annu. Rev. 26: 361-393. Forward, R. B., Jr., and D. Rittschof. 2000. Alteration of photore- sponses involved in diel vertical migration of a crab larva by fish mucus and degradation products of mucopolysaccharides. J. E.\p. Mar. Biol. Ecol. 245: 277-292. French, A. S. 1992. Mechanotransduction. Annu. Rev. Phvsiol. 54: 135- 152. Goldsmith, T. H. 1991. Photoreception and vision. Pp. 171-245 in Neural and Integrative Animal Physiology. C.L. Prosser, ed. Wiley- Liss, New York. Hartline, D. K., E. J. Buskey. and P. H. Lenz. 1999. Rapid jumps and bioluminescence elicited by controlled hydrodynamic stimuli in a me- sopelagic copepod. Pleuromamma xiphias. Biol. Bull. 197: 132-143. Hess, K., and F. White. 1974. A numerical tidal model of Narragansett Bay. Sea Grant Marine Technical Rep. 20. University of Rhode Island. Kingston. RI. Katona, S. K. 1973. Evidence for sex pheromones in planktonic cope- pods. Limnol. Oceanogr. 18: 574-583. Kiarboe, T., and A. Visser. 1999. Predator and prey perception in copepods due to hydromechanical signals. Mar. Ecol. Prog. Ser. 179: 81-95. Kiorboe, T., E. Saiz, and A. Visser. 1999. Hydrodynamic signal per- ception in the copepod Acania tonsa. Mar. Ecol. Prog. Ser. 179: 97-111 Laverack, M. S. 1969. Mechanoreceptors. photoreeeptors. and rapid conduction pathways in the leech. IliruJo nicilii -inulis. J. E.V/J. Hiol. 50: 129-140. Lenz, P. H., and I). K. Hartline. 1999. Reaction times and force production during escape behavior of a calanoid copepod Uiulinula vulgaris. Mar. Biol. 133: 249-258. Lenz, P. H., D. K. Hartline, and A. D. Davis. 2000. The need for speed I. Fast reactions and myelinated axons in copepods. J. Coinp. Phvsiol. A 186: 337-345. Lowe, E. 1935. On the anatomy of a marine copepod. Ca/anus fininar- chicits (Gunnerus). Trans. R. Soc. Edinb. 63: 560-603. Mensinger, A. F., and J. F. Case. 1992. Dinoflagellate luminescence increases susceptibility of zooplankton to teleost predation. Mar. Hiol. 112: 207-210. Morin, J. G. 1983. Coastal bioluminescence: patterns and functions. Bull. Mar. Sci. 33: 7S7-8I7. Neumeister, H., B. Ripley. T. Preuss, and \V. F. Gilly. 2000. Effects of temperature on escape jetting in the squid Loligo opalescent. J. Exp. Biol. 203: 547-557. Park, T. S. 1966. The biology of a calanoid copepod Epilabuloccra ainphiirite.s McMurrich. Cellule 66: 129-251. Porter, K. G., and J. W. Porter. 1979. Bioluminescence in marine plankton: a co-evolved antipredator system. Am. Nat. 114: 458-461. Ringelberg, J. 1991. Enhancement of the phototactic reaction in Daph- iiin Inalina by a chemical mediated by juvenile perch ( Perca fluviati- lis}. ./. Plankton Res. 13: 17-25. Singarajah. K. V. 1969. Escape reactions of zooplankton: avoidance of a pursuing siphon tube. J. Exp. Mar. Biol. Ecol. 3: 171-178. Stearns, D. E., and R. B. Forward, Jr. 1984a. Photosensitivity of the calanoid copepod Acartia tonsa. Mar. Biol. 82: 85-89. Stearns, D. E., and R. B. Forward, Jr. 1984b. Copepod photobehavior in a simulated natural light environment and its relation to nocturnal vertical migration. Mar. Biol. 82: 91-100. Stubblefield, C. L., C. M. Lascara, and M. Vecchione. 1984. Vertical distribution of zooplankton in a shallow turbid estuary. Contr. Mar. Sci. 27: 93-104. Tett, P. B., and M. G. Kelly. 1973. Marine bioluminescence. Oceanogr. Mar. Biol. Annu. Rev. 11: 89-173. Thurm, II. 1965. An insect receptor potential. Cold Spring Harbor Symp. Quant. Biol. 30: 83-94. Tjossem, S. F. 1990. Effect of fish chemical cues on vertical migration behavior of Chaohorus. Limnol. Oceanogr. 35: 1456-1468. Tomita, T. 1970. Electrical activity of vertebrate photoreeeptors. Quart. Rev. Biopliy. 3: 179-222. Viitasalo, M, T. Kinrboe, J. Flinkman, L. Pedersen, and W. W. Visser. 1998. Predation vulnerability of planktonic copepods: consequences of predator foraging strategies and prey sensory abilities. Mar. Ecol. Prog. Ser. 175: 129-142. Weeks, F. L, and G. Duncan. 1974. Photoreception by a cephalopod retina: response dynamics. Exp. Eve Res. 19: 493-509. Wine, J. J., and F. B. Krasne. 1982. The cellular organization of crayfish escape behavior. Pp. 241-292 in The Bio/ogv of Crustacea Vol. 4. Neural Integration and Behavior. D.C. Sandeman and H.L. Atwood, eds. Academic Press. New York. Wyman, R. J., J. B. Thomas, L. Salkoff, and D. G. King. 1984. The Dmsopliilii giant fiber system. Pp. 133-161 in Neural Mechanisms oj Startle Behavior. R.C. Eaton, ed. Plenum. New York. Yen, J., P. H. Lenz, D. V. Gassie, and D. K. Hartline. 1992. Mech- anoreception in marine copepods: eleetrophysiological studies on the first antennae. J. Plankton Res. 14: 495-512. Reference: Biol. Bull. 204: 3S-44. (February 2003) O 2003 Marine Biolo-jic.il Lahoraton Behavioral Thermoregulation in Hemigrapsus nudus, the Amphibious Purple Shore Crab I. J. McGAW Department of Biological Sciences, University of Nevada Las Vegas, 4505 Maryland Parkway, Las Vegus, Nevada S9 1 54-4004; and Bainfie/d Marine Sciences Centre. 100 Pachena Road. Bamfield, British Cohtnihiu VOR 1BO. Canada Abstract. The thermoregulatory behavior of Hanigrap- sns niuliis. the amphibious purple shore crab, was examined in both aquatic and aerial environments. Crabs warmed and cooled more rapidly in water than in air. Acclimation in water of 16 C (summer temperatures) raised the critical thermal maximum temperature (CTMax): acclimation in water of 10 C (winter temperatures) lowered the critical thermal minimum temperature (CTMin). The changes oc- curred in both water and air. However, these survival re- gimes did not reflect the thermal preferences of the animals. In water, the thermal preference of crabs acclimated to 16 C was 14.6 C, and they avoided water warmer than 25.? C. These values were significantly lower than those of the crabs acclimated to 10 C; these animals demonstrated temperature preferences for water that was 17 C, and they avoided water that was warmer than 26.9 C. This temper- ature preference was also exhibited in air. where 10 C acclimated crabs exited from under rocks at a temperature that was 3.2 C higher than that at which the 16 C accli- mated animals responded. This behavioral pattern was pos- sibly due to a decreased thermal tolerance of 16 C accli- mated crabs, related with the molting process. H. niidiis was better able to survive prolonged exposure to cold tempera- tures than to warm temperatures, and there was a trend towards lower exit temperatures with the lower acclimation (10 C) temperature. Using a complex series of behaviors, the crabs were able to precisely control body temperature independent of the medium, by shuttling between air and water. The time spent in either air or water was influenced more strongly by the temperature than by the medium. In the field, this species may experience ranges in temperatures Received 10 May 2002: accepted 17 October 2002. F.-mail: IIIIL>:,I\\ '"ccmail nevada.edu of up to 20 C; however, it is able to utilize thermal microhabitats underneath rocks to maintain its body tem- perature within fairly narrow limits. Introduction Intertidal organisms experience abrupt, frequently large, changes in temperature as a result of alternating episodes of exposure to air and water (Vernberg and Vernberg, 1972). These changes in temperature may pose an additional bur- den to amphibious organisms that are already challenged by the switch between ventilatory media (Greenaway et ai, 1996). Hemigrapsus audits, the purple shore crab, is a common species in the mid- to high-intertidal zone of rocky shores along the northeastern Pacific (Schmitt, 1921; Dehnel, 1960; Low, 1970; Daly, 1981). These crabs are involun- tarily exposed as the tide recedes, but they are active in air (Burnett and McMahon, 1987). The species can tolerate temperatures up to 33.6 C for short periods of time (Todd and Dehnel, 1960); however, exposure to suboptimal tem- perature regimes is associated with compensatory physio- logical responses in decapod crustaceans. The aerobic metabolism of crustaceans, like that of most other aquatic organisms, is temperature dependent. Oxygen uptake increases in Carcinus inaenas, the green shore crab. as the temperature of the water is raised (Taylor and Wheatly, 1979); likewise, oxygen consumption in Homanis f>tiininunts. the European lobster, decreases (for a short time) as temperature is lowered (Whiteley et ai. 1995). Increases in temperature also influence oxygen delivery to the tissues by causing a reduction in the carrying capacity ot the hemolymph for oxygen and the binding of oxygen to the hcmocyanin (Taylor. 1981 ; Truchot. 1983). Heart rate is directly related to temperature in a number 38 THERMOREGULATION IN HEMIGRAPSUS NUDUS 39 of crustacean species (deFur and Mangum, 1979; Taylor and Wheatly. 1979: DeWachter and McMahon, 1996; Still- man and Somero. 1996; Pirro et a/.. 1999: Jury and Watson, 2000: Fredrich etui., 2000). Heating increases heart rate and cardiac output but decreases stroke volume in Cancer nui- gister, the Dungeness crab (DeWachter and McMahon, 1996). This is associated with an increase in hemolymph perfusion of the carapace, gonads. and musculature of the pereiopods (DeWachter and McMahon. 1996). Cooling causes a decrease in cardiac parameters: heart rate and cardiac output drop sharply in low temperature, and hemo- lymph flow is directed away from anterior structures to more ventral structures (Fredrich et al., 2000). Since exposure to high or low temperatures can be met- abolically costly, the ability of crabs to sense temperature and orient to a "thermal niche" should be advantageous in minimizing physiological stress. In addition, many pro- cesses such as molting, growth, reproduction, and matura- tion of eggs are temperature dependent (Sastry, 1983a, b); therefore, selection of optimal temperatures should also maximize growth and reproductive potential (Hutchison and Maness. 1979). A number of crustacean species are known to exhibit behavioral thermoregulation. Homarus america- iiiis, the American lobster, can thermoregulate precisely for up to 6 days, preferring temperatures in the 15-21 "C range (Reynolds and Casterlin, 1979a; Crossin et nL 1998). Lob- sters are able to detect water temperature differences of as little as 1 C and exhibit directional taxis (Jury and Watson, 2000). Carcinus maenas, the green shore crab, avoids ad- verse temperatures, showing emersion responses at 28 "C in the laboratory (Taylor and Wheatly, 1979). Procambarus clarkii. the red swamp crayfish, has a broad temperature tolerance (Payette and McGaw. 2001) and prefers water with a mean temperature of 23-24 C (Espina et ai, 1993; Ramirez et ai. 1994). An animal's thermal preference can be also be influenced by the acclimation temperature. Ac- climation to warm temperatures results in a higher temper- ature preference in Homarus americanus, the American lobster (Crossin et ai. 1998), and in Astacus astacits, a crayfish (Kivivuori, 1994). Temperature acclimation has an opposite effect on the crayfish Orconectes immunis, with animals acclimated to warm water selecting cooler temper- atures than those acclimated to cold water (Crawshaw, 1974). Most of the articles on behavioral thermoregulation in decapod crustaceans have concentrated on fully aquatic species (see Crossin et ai, 1998). Much less information exists on amphibious species that are emersed in the inter- tidal zone twice daily (Thurman, 1998). H. nuthis is exposed to a wide range of temperatures on both a tidal and diurnal basis (Todd and Dehnel. 1960; Greenaway et al., 1996). Therefore, the aim of this study was to investigate the thermal ecology of this amphibious species and to assess the role of behavioral reactions, in both water and air. in min- imizing the effects of thermal stress. Materials and Methods Adult male and female purple shore crabs, Hemigrapsm nudiis, of 25-40-mm carapace width, were collected inter- tidally in Barkley Sound, British Columbia, during the months of May to August in 2000 and 2001. They were transferred to 40-liter aquaria at the Bamfield Marine Sci- ences Centre and maintained in aerated seawater at a salin- ity of 33 ppt 0.5 ppt on a natural light-dark cycle. The crabs were held at water temperatures of either 16 C 0.5 C or 10 C 0.5 C for at least 2 weeks. These temper- atures approximated those measured in the field during summer and winter respectively (Gosselin and Chia, 1995). More extreme temperatures were not used because the an- imals tended to molt at higher temperatures and become lethargic at lower temperatures. The crabs were fed sea lettuce, Ulva lactuca, ad libitum. Approximately equal numbers of each sex were used, and individual crabs were not re-used in any experiment. Rate of change of hod\ temperature Changes in the body temperature of H. nudiis (n = 10) were studied in water and air. To measure blood tempera- ture, a catheter-mounted (PE90) thermocouple (Physitemp IT18) was inserted through a small hole drilled in the first abdominal segment and guided to lie against the sternal artery. The crabs were returned to the holding tank and allowed to settle for 15 min. Animals (n = 10) were then transferred to water or air of 5 C or 20 C. The amount of time required for the body temperature to equilibrate with the surrounding medium was recorded at 30-s intervals using a BAT 12 digital thermometer (Physitemp Instru- ments). Critical thermal maximum and minimum temperatures The critical thermal maximum (CTMax) and critical ther- mal minimum (CTMin) temperatures of H. niiclim were assessed in air and in water (n = 30). Crabs were accli- mated to 10 C or 16 C and were studied separately, with the starting temperature being 10 or 16 C, respectively. The temperature of the air was raised (or cooled) at 0.5 C/min in an incubator (Percival Instruments [Boone. Iowa]; model 135LL), and the temperature was monitored at 1-min intervals with a Physitemp BAT 12 digital thermom- eter. A volume of 5 liters of water was used, and the temperature was raised (or cooled) at 0.5 C/min by way of a recirculating water bath (VWR Scientific Instruments). The water was aerated, and the temperature was monitored with the Physitemp thermometer. At random intervals, the crabs were turned on their backs until the first animal 40 I. J. McGAW reached its CTMax or CTMin; that is. until the animal could no longer right itself within 1 min (Cuculescu ct al., 1998). Thereafter, all remaining crabs were inverted together every minute and the CTMax or CTMin was recorded for each individual. Temperature preference hehavior The temperature preference range of H. inulus was de- termined using an elongated (length, 300 cm) cylindrical (diameter. 12 cm) chamber that was orientated horizontally. Heating and cooling recirculating water baths at either end of the chamber maintained the temperature gradient be- tween 7 C and 30 C. The placement of the heating and cooling water baths was alternated between each trial, to eliminate any bias for either end of the chamber. Airstones minimized any vertical thermal stratification in the gradient and ensured that the water did not become hypoxic. Shelters (broken glass beakers) were placed along the length of the chamber to reduce stress, H. niultis is highly thigmotactic and will remain active, attempting to escape, unless there is a place to shelter (McGaw, 2001). This atypical behavior could obscure thermoregulatory responses. Crabs a max- imum of five at any one time (8 repetitions; total // = 40) were introduced into the gradient at random locations; using this number of H. inulits in experiments does not affect the thermotolerance of an individual (Todd and Dehnel. I960). After 3 h. a temperature reading was taken at the position of each individual crab. Those crabs accli- mated to either 10 C or 16 C were studied separately. In control experiments, the temperature was maintained at either a constant 10 'C or a constant 16 C; crabs were then introduced randomly into the apparatus, and their position was recorded after 3 h. Temperature avoidance The following two experiments were designed to test the responses of the crabs after they had sensed a change in temperature. Behavioral responses consisted of migration from underneath a shelter as the temperature changed. Ex- periments were performed in both aquatic and aerial envi- ronments. The first experiment (aquatic) was carried out in a mod- ified two-choice chamber (Fig. 1 ), which contained seawater (32 ppt), in one side, as well as pieces of broken glass beakers for shelter. The chamber was held in an incubator (Percival Instruments Model 135LL). which allowed inde- pendent control of air temperature. Five animals per trial (5 repetitions; /; :: 25) were placed in the seawater and allowed to settle for 30 min. Any animals that exited the water within this period were not used in the experiments. The starting temperature of experiments was either 10 'C or 16 C for each group of acclimated crabs. The temperature of the seawater was raised at 0.5 C/min usinu a recirculat- Ramp Aerial environment Mesh screen To water bath Glass shelters Figure 1. Modified two-choice chamber used to measure exit temper- atures from water into air. shuttling behavior between air and water, and behavioral control of body temperature. ing water bath (VWR Scientific Instruments). The temper- ature at which the crabs made a voluntary migration into air was recorded; this behavior was defined as "emigration" (Taylor and Wheatly. 1979). The experiment was repeated with air temperatures of 5 C, 20 C, and 35 C, each at 50%-70% relative humidity. Experiments were then carried out to assess the lower preference range. The water was cooled at 0.5 C/min, and emigration temperature from the seawater was recorded at the three air temperatures. The water side of the chamber was alternated between trials to avoid any preference associated with either side of the chamber. For the second experiment, temperature avoidance was tested in air using a chamber measuring 45 cm X 45 cm X S cm deep, with a gauze bottom to allow air to circulate. Flat tiles were placed in the chamber. Five animals were then introduced into the chamber and allowed to settle under the tiles. Any animals that migrated from under the tiles within 30 min were not used in the experiments. The chamber was held in an incubator (Percival Instruments), with the starting temperature for the two acclimated groups being either 10 C or 16 C. The air temperature was then raised by 0.5 C/min, and the temperature (measured under the tiles) at which the crabs exited from under the tile shelters was recorded (n = 25). The experiment was repeated by low- ering the temperature, by 0.5 C/min, and observing the temperature at which the crabs exited from under the shel- ters. All recordings were made in constant dim red light. Shuttling behavior A time-lapse video recorder and camera (Panasonic AG- RT600AS VCR and Panasonic WV-BP120 camera) was used to monitor the shuttling behavior of individual crabs THERMOREGULATION IN HEMIGRAPSUS NUDUS 41 25 -i air/water temperature \ air/water temperature 10 20 30 Time (min) 40 50 60 Figure 2. Changes in body temperature (mean SEM) of 10 //<;/ i,'ni/i.v/.\ inulm. after transfer from 10 "C water to 20 C air (A), from 10 C water to 20 C water (A), from 16 C water to 5 C air (O), and from 16 C water to 5 C water (). In some cases error bars are smaller than the symbols. between air and water at temperatures of 10 C. 20 C, and 30 C. The choice chamber was set up in an incubator with glass shelters in both air and water. Four crabs (acclimated to 16 C) were placed in the water (2 repetitions, total n = 8 for each treatment). The number of shuttles, duration of shuttles, and total time spent in air and water were recorded over a 24-h period in constant dim red light. Behavioral control of body temperature The body temperature of eight crabs (acclimated to 16 C) was recorded with a thermocouple (Physitemp IT 18) introduced through the first abdominal segment. The ther- mocouple was connected to a BAT 12 digital thermometer (Physitemp Instruments); data were recorded on an ADIn- struments Powerlab data acquisition package. The two- choice chamber was placed in an incubator (Percival, model 135LL), and a recirculating water bath allowed independent heating or cooling of the seawater. An animal was initially placed in the shallow water, and the change in its body temperature was followed for 12 h as it shuttled between air and water. A variety of water and air temperature combi- nations were offered, separated by differing increments. Regulation of bod\ temperature in the field Regulation of body temperature was assessed in freshly collected crabs in the field. Crabs were fitted with thermo- couples (Physitemp IT18) on a 2-m lead (n = 5). Each crab was released on a falling high tide and allowed to settle; body temperature was recorded at half-hour intervals until the following high tide, using a BAT 12 digital thermometer (Physitemp Instruments). At the same time, air temperatures were recorded 5 cm above the rock surface, and seawater temperature was recorded at the low tide, using a Physitemp IT 14 thermocouple calibrated against a mercury thermom- eter. Experiments were repeated on days when air temper- atures were higher or lower than the ambient seawater temperature. Results Rate of change of body temperature An increase or decrease in water temperature of about 10 C resulted in a rapid change in body temperature (Fig. 2). Body temperature equilibrated with the surrounding water, within 2-3 min. In air, body temperature changed more slowly, and heat loss from the body was more rapid than heat gain. The body temperature took 25 min to equilibrate to a 10 C drop in air temperature, but it failed to reach equilibrium with the surrounding air within the 60-min experimental period when the temperature was raised by 10 "C. Although body temperature reached 90% of the final temperature within 20 min. it increased slowly thereafter. 42 I. J. McGAW Table 1 Tliennal preference o/Hemigrapsus nudus with incrcusmx temperature Air temperature (C) Water temperature I C) at emigration* Crabs acclimated to 10 C Crabs acclimated to 16 "C 5 20 35 25.5 0.81 27.4 061 27.9 0.6d 25.3 0.68 24.7 0.39 25.7 0.53 * Mean ( standard error of the mean) upper temperature at which crabs (n = 25 ) emigrated from water into air with a temperature of 5 C, 20 C, or 35 C as the temperature of the water was raised. Critical thermal maximum anil minimum temperatures In water, the CTMax of 31.1 C a standard error of the mean (SEM) of 0. 16 C for crabs acclimated to 10 C was significantly lower than the CTMax of 33.6 0.11 C for crabs acclimated to 16 C (Student's ? test = -2.32, P = 0.02). The difference between the two acclimation groups was greater in air. Crabs acclimated to 10 C had a CTMax of 33.2 0.34 C, which was significantly lower than the CTMax of 35.3 0.5 C for 16 C acclimated animals (t test = -3.45, P = 0.001). In addition, the CTMax values in water were significantly lower than those in air ( ANOV A, F ---- 7.55, P = 0.007). Acclimation to either 10 C or 16 C also affected the critical thermal minimum temperature. The CTMin in water of 3.5 0.14 C for crabs acclimated to 10 C was signif- icantly lower than the 4.82 0.14 C for 16 C acclimated crabs (/ test = -6.71. P < 0.001). A similar trend was observed in air, with CTMin values of 3.44 0.15 C and 3.99 0.12 C, for 10 C and 16 C acclimated crabs, respectively (nest = -2.89, P = 0.005). As with CTMax, there was a significant effect associated with the medium: the CTMin values in air were significantly lower than those in water (ANOVA, F = 10.41. P = 0.002 1. Temperature preference When 40 crabs (again, acclimated to either 10 C or 16 C) were placed randomly in a thermal gradient of 7 C to 30 C, there was considerable movement within the first 30 min. Temperature selection appeared to be complete after 3 h. with very little movement in the gradient thereafter. Although a small percentage of the crabs selected the ex- treme temperatures of 7 C C or 30 C. most were distributed between 1 1 C and 24 C. The mean preference range of 17.01 C 0.65 C SEM for 10 C acclimated crabs was significantly higher than the 14.60 'C 0.78 "C selected by 16 C acclimated crabs (t test = 2.37. P = 0.02). Control experiments were carried out for the two acclimation tem- peratures, with no thermal gradient. Control crabs did not show a preference for any area of the gradient tank. A similar effect of acclimation on temperature preference was observed in the temperature-avoidance experiments. When the temperature of the water was gradually increased, crabs exited from under the shelters and started to become active between 19-21 C. but did not leave the water at this temperature. Although there were three different air tem- peratures that crabs could emigrate into, the air temperature had no significant effect on emigration temperatures from water (Table 1) (ANOVA, F = 2.47. P = 0.088). Since air temperature had no effect on behavior, data for the three air temperatures was pooled. There was a significant behav- ioral effect based on acclimation: crabs acclimated to 10 C had a mean emigration temperature of 26.94 0.24 C; this was significantly higher than the mean emigration temper- ature of 25.25 0.19 C for crabs acclimated to 16 C (ANOVA. F = 10.47. P = 0.002). All crabs had left the water when the temperature reached 34 C. Although all crabs left the water when the temperature was raised, this was not the case when the water temperature was lowered. Only 45<7r of the crabs acclimated to 10 C and 30% of the crabs acclimated to 16 C emigrated from the water. The rest of the crabs remained in the water even though the temperature was reduced below their CTMin. incapacitating them. Statistical results for the animals that exhibited emigration behavior are given in Table 2. Air temperature had no significant effect on the emigration temperature of the crabs (ANOVA, F = 0.14, P = 0.87). Although the mean emigration temperature of 4.95 0.31 C for 10 C acclimated animals was lower than the 5.79 0.42 C for 16 C acclimated crabs, this difference was statistically insignificant (ANOVA. F = 2.58.P = 0.1 13). Therefore, in water, 10 C acclimated crabs have a pref- erence range between 4.95 C and 26.94 C, with a mean preference of 17.1 C. The crabs acclimated to 16 C have a mean temperature preference of 14.60 C with a narrower preference range between 5.79 C and 25.25 C. Acclimation to either 10 C or 16 C had a similar effect on temperature avoidance in air. When the temperature of Table 2 Tlifrnial preference of Hemigrapsus nudus with decreasing temperature Air temperature (C) Water temperature (C) at emigration* Crabs acclimated to 10 C Crabs acclimated to 16 "C 5 20 35 4.72 0.53 5.23 0.57 4.91 0.66 5.87 0.51 5.24 0.61 6.26 0.89 * Mean ( standard error of the mean) temperature at which crabs (;i - 25) emigrated from water into air with a temperature of 5 C, 20 C. or 35 "C as the temperature of the water was lowered. THERMOREGULATION IN HEM1GRAPSUS NUDUS 43 Table 3 Shiittliiif; behavior of Hemigrapsus nudus from water into air u-hen both media were maintained at 10 C 20 C or 30 "C Parameter* Temperature (C) Number of shuttles Duration of each shuttle (mini Percent time spent in air/24 h II) 20 30 26.4 5.1 18.0 4.3 27.3 6.2 8.6 0.9 21.2 8.1 28.5 7.5 16.6 11.3 24.5 23.5 64.3 16.5 : Values are the mean ( standard error of the mean) response for 8 crabs. the air was gradually raised, the crabs exited from under tiles in an attempt to escape. The mean exit temperature for crabs acclimated to 10 C was 27.39 0.62 C; this was significantly higher than the exit temperature of 24.17 0.58 C recorded for 16 C acclimated crabs (t test = -3.8, P < 0.001 ). Consistent with lower emersion temperatures in water, not all crabs exited from under shelters as the air temperature was gradually reduced. Five of the 25 crabs acclimated to 10 C remained under the shelters, while 10 of the crabs acclimated to 16 C did not exit. Statistical anal- ysis includes only the animals exhibiting this exit behavior. Although the mean exit temperature of 4.36 0.47 'C recorded for 10 C acclimated crabs was lower than the 5.85 0.7 C recorded for 16 C acclimated animals, this difference was statistically insignificant (t test = -1.82, P = 0.077). Therefore, the temperature preference range in air for 10 C acclimated crabs was 4.36 C to 27.39 C, which was broader than that for 16 C acclimated animals (5.85 C to 24.17 C). Shuttling behavior The shuttling movement of 16 C acclimated crabs be- tween air and water was studied during a 24-h period to determine the number and duration of excursions into air (Table 3). There was no significant difference in the number of shuttles between air and water as a result of ambient temperature (ANOVA, F = 0.96, P = 0.4). There was a trend towards an increase in the average duration of each excursion into air as the temperature increased, but it was not statistically significant (ANOVA, F = 3.01. P : 0.05). When the percentage of time (per 24-h period) that each crab spent in air was considered, a significant pattern emerged (Table 3). As the temperature increased, the crabs spent a significantly greater total percentage of time in air (ANOVA, F ---- 14.72, P < 0.001). The percent of time that the crabs spent in air at 30 C was significantly higher than time spent in 10 C and 20 C conditions, but there was no significant difference between 10 C and 20 C (Tukey test, q = 1.18, P = 0.684) Behavioral control of hod\ temperature When offered a choice of 20 C water with 8 C air (trial 5), the crabs remained in the water most of the time: the mean body temperature of 20.3 C 0.2 C SD was not significantly different from the water temperature (Fig. 3). The body temperature in experimental trial 5 was signifi- cantly higher than in the other trials (Tukey test. P < 0.05). In experimental trial 4 ( 14 C water and 24 C air), the crabs also remained in the water; again, the mean body tempera- ture of 14.2 0.3 C was not significantly different from that of the water (Fig 3). In all other trials, the crabs maintained the body temperature at levels between the temperature of the air and the water. In the shallow water of the chamber, periodically, a crab either raised or submerged itself to control its body temperature between mean values of 7.7 0.9 C and 14.6 1.5 C. Examination of the body temperature of individual crabs shows the thermoregulatory responses in more detail (Fig. 4a-d). When the water was held at 4-5 C and the air at 33-34 C (Fig. 4a), the crab spent the first 2 h shuttling O 40 - 35 - 25 - 20 - 10 - 5 - Water temperature Air temperature \ Body temperature I 1 . :' i Experimental Trial Figure 3. Mean body temperature ( SD) of Wf'.i;ra/>.Mf.\ niuhis ( n = 8) when offered a choice between air and water, maintained at different temperatures relative to one another. Hashed bars represent water temper- ature, and solid bars represent air temperatures. 44 I. J. McGAW a a a> o m 30- 20- 10- AIR WATER 30- 20- I m 10- 6 Time (h) AIR K 12 WATER 6 Time (h) 12 _. 30- o 8. E TJ O m 20- 10- WATER AIR 6 Time(h) 12 _ 30 o 20 o CO 10 AIR WATER 6 Time(h) 12 Figure 4. Representative examples of body temperatures of individual Hi'mixnipsiix nudux in a two-choice chamber with the ability to shuttle between air and water of different temperatures, (a) Water of 4-5 "C and air of 33-34 "C. (b) Water of 33-34 C and air of 4-5 C. (c) Water of 6-7 C and air of 30-31 "C. (d) Water of 8-9 C and air of 20-21 C. between air and water, after which it raised or submerged itself in the water to maintain a body temperature of about 8-13 C. When the temperatures of the air and water were reversed (4-5 C air and 33-34 C water), the crab (Fig 4b) was still able to maintain a body temperature between 8 and 13 C for most of the 12-h experimental period. When the difference between air and water was decreased (8-9 C water and 29-30 C air), body temperature fluctuated some- what during the first 2 h when the crab was active; there- after, body temperature was maintained between 10 C and 17 C (Fig. 4c). When air and water temperatures (20-21 'C air and 8-9 C water) approached limits within the animal's preference range (Table 1.2). the crab tended to shuttle back and forth between air and water, spending extended periods of time in either medium, where body temperature equilibrated with the medium (Fig. 4d). Bil\ temperature in the field Changes in body temperature of H. nndiis were recorded in the field during an intertidal period (Fig. 5). Body tem- perature was monitored on a cold day (Fig 5a), when the air temperature dropped below that of the seawater. Although air temperatures fell from 18 C to 10.7 C, body temper- atures decreased only slightly. The mean body temperature of the crabs (n = 5) dropped from 17.65 0.28 C SEM when initially emersed, down to 14.92 0.44 C at the end of the intertidal period. This was only a 16% drop in body temperature, compared to a drop of 41% in the temperature of the surrounding air. Body temperature increased rapidly when the crabs were re-immersed, reaching 16.5 0.27 C as it equilibrated with the seawater. On a warm day (Fig. 5b) the air temperature quickly rose THERMOREGULATION IN HKMIGKAPSUS NVOVS 45 24 "- 20 - 3 ro 16 s. seawater temp. 8 J r 1700 1900 2100 2300 0100 Time 25 U _ 20 o 3 | a I 15 10 -" Time Figure 5. Changes in body temperature (mean SEM) of 5 specimens of Hemigrapsus midiis (solid line) in the intertidal zone. Crabs were released during a falling tide and monitored until the following high tide. Seawater temperatures (dotted line, solid symbols) and air temperatures (dashed line, open symbols! were also recorded during this time. Times of emersion and immersion of the crabs, as well as low tide (LT). are indicated on the graphs. Recordings were made on (al 8 July 200 1. when surrounding air temperatures were lower than ambient seawater tempera- tures and (b) 23 July 2001. when air temperature was higher than seawater temperature. from 11 C in the morning to 22-24 C by early afternoon. Despite this 12 C rise in air temperature, the body temper- atures of the crabs did not change as rapidly, and reached only 16.86 0.51 C by the end of exposure period in air. The change in body temperature was similar to the observed increase in seawater temperature during the day (Fig. 5b). When the crabs were re-immersed, their body temperatures quickly equilibrated with the seawater. Discussion The observed rates of change in body temperature (Fig. 2) were similar to those reported previously for Hemigrapsus nudus (Greenaway et al., 1996). In lobsters, heat loss in air is more rapid than heat gain ( Whiteley el al.. 1995): this was also observed here for H. muhts (Fig. 2), probably due to the evaporative heat loss in air. During these experiments, sev- eral of the crabs regurgitated frothed fluids from the stom- ach, smeared this over the ventral carapace with the chelae, and raised their body above the substrate. This foaming behavior has been reported for a number of crab species (Lindeberg, 1980; Maitland. 1990) and can used to reduce body temperature (Jansen, 1970). However, there was no evidence to suggest that the H. nudus specimens were using this method to slow their rate of heating. The relative humidities of 60%-70% used during the experiments, which mimicked conditions measured in the field, could have reduced the effectiveness of (but not eliminated) evap- orative cooling (Edney, 1961). Although foaming behavior was not observed in the field, it is possible that it could reduce heating rates on warm days with low relative hu- midity (Thurman. 1998). Fiddler crabs (Uca species) are able to maintain a body temperature below that of the surrounding air by changing posture, blanching, and evap- orating water from the body surface (Wilkens and Finger- man, 1965; Smith and Miller, 1973; Thurman. 1998). In the present study, there was no difference in heating or cooling rates when comparing live and dead animals (not shown), suggesting that there is no active mechanism that allows H. nmlus to control the rate of heat gain or loss from the body. When H. nudus was acclimated to different temperatures, an increase in the upper survival limits occurred as a result of the higher acclimation temperature; this has been re- ported previously for H. nudus (Todd and Dehnel, 1960). as well as for other species of crustaceans (Mundahl and Benton. 1990; Lagerspetz and Bowler. 1993; Korhonen and Lagerspetz, 1996; Cuculescu el al.. 1998; Stillman and Somero, 2000). However, much less is known about the critical thermal minima. In the present study, acclimation to a lower temperature extended the CTMin. Acclimation to a wider range of temperatures has also been shown to extend the CTMin range in other crustaceans (Layne et al., 1987; Stillman and Somero, 1996). Survival limits in air as a function of temperature have not been investigated previously for H. nudus. Interestingly, both the CTMax and CTMin were greater in air than in water. The heating and cooling rate in the incubator (0.5 C/min) was adequate to allow equalization of the body with the surrounding air (unpubl. data; Fig. 2). These results are somewhat surprising: an animal would already be phys- iologically challenged by the switch in ventilatory media (Greenaway el al., 1996). because an increase in air tem- perature decreases the oxygen-carrying capacity of the 46 I. J. McGAW hemocyanin and results in thermal acidosis (Morris et al., 1996b). The CTMax was determined close to the body of the crab rather than by using internal temperature probes, which tended to tangle around the legs when the crabs were turned over, affecting the ability of the animal to right itself. Even though relative humidities in the incubator were high (60%-80^:), a degree of evaporative cooling could have kept the body temperature a degree or so cooler than the surrounding air (unpubl. data; Fig. 2). This would suggest that the upper lethal limits in air were probably similar to those measured in water. However, if evaporative cooling reduced the body temperature, then the CTMin in air would also be expected to occur at a higher temperature than in water. This did not happen in the present study. The thermal preference behavior of crabs acclimated to 10 C and 16 C, ascertained in a thermal gradient and by temperature-aversion experiments (Table 1 ), did not reflect their temperature tolerances: in both cases. 10 C accli- mated crabs had a higher temperature preference than those acclimated to 16 C. In the temperature-aversion experi- ments, the oxygen tension was maintained at constant lev- els, so the emigration from water was a direct consequence of temperature. Indeed, aquatic hypoxia is not an impetus for emersion in this species (Moms et ai, 1996c). The air temperature that the crabs could exit into did not affect the exit temperature from the water (Table 1 ). This was unex- pected, since acute exposure to higher air temperature (>15 C) is costly and is associated with thermal acidosis and compensatory increases in cardiac output to maintain ade- quate oxygen uptake (Morris et ai, 1996a, b). Acclimation to 10 C or 16 C also influenced aversion behavior in air. When the air temperature was raised, crabs acclimated to 10 C exited from under stones at a higher temperature than did 16 C acclimated crabs. The adaptive significance of this behavior is unclear, since 16 C acclimated crabs are more tolerant of higher temperatures (CTMax values). Thus, the effect of acclimation on behavior is apparently the opposite of its effect on survival regimes. In other reports on ther- moregulatory behavior, lobsters that are acclimated to warm water choose wanner temperatures than do cold-acclimated individuals, possibly to maintain an optimal thermal regime for metabolic activities (Crossin ft nl.. 1998). When Astiiftis astacus, a crayfish, is acclimated to cold or warm water, this also directly affects thermal preference (Kivivuori, 1994). Acclimation to either 15 C or 25 C has no effect on the emersion response of the shore crab Cardans macnas. which exits into air when the water temperature reaches 28 C (Taylor and Wheatly, 1979). Likewise, acclimation to differing temperatures has no effect on the temperature preference of Procambants clarkli. the red swamp crayfish (Espina ft ai, 1993). In contrast to these responses, when the crayfish Orconectes immunis is acclimated to cold wa- ter, it tends to choose higher temperatures than animals acclimated to warm water, yet no explanation is given for this paradox (Crawshaw, 1974). Several factors can be eliminated as causes for the unex- pected behavior observed in the present study. ( 1 ) The crabs were not responding to temperature increases of a particular magnitude, as occurs in lobsters (Cooke-Schreiber et ai, 2001 ). (2) The warming rate of the water (0.5 C/min) was slow enough to allow the body temperature to equilibrate with the surrounding medium (Fig. 2). (3) Although the crabs were introduced into the apparatus in groups of five, this was unlikely to have a substantial effect on their be- havior: they were roughly equal in size and there was ample shelter both of these factors would reduce aggressive in- teractions between animals (Jacoby, 1981). (4) Although Carcinus maenas exhibits a behavioral hypothermia when exposed to hypoxic conditions (DeWachter et ai, 1997), this is probably not the case for H. mtdus because oxygen levels were maintained during experiments and this species does not modify its behavior in response to hypoxia (Morris et ai, 1996c). (5) Finally, the acclimation period of 2 weeks should have been long enough for an increased temperature tolerance (Layne et ai, 1987: Cuculescu et ai, 1998). Indeed, rapid acclimation to thermal zones is an advantage for intertidul organisms: H. mtdus, which acclimatizes within 48 h (Todd and Dehnel, 1960), is no exception. When considering factors that could have influenced this overt behavior, it is worth noting that H. mtdus could not be acclimated to temperatures greater than 16 C without in- ducing widespread molting. Because the entire molting pro- cess can take several weeks (O'Halloran and O'Dor, 1988). it is possible that the 16 C acclimated crabs were just starting to molt (D, or D 2 stage) without visible signs. Early stages of the molting process are associated with biochem- ical and physiological changes (see Chang, 1995) and make H. midus less tolerant of high temperatures (Todd and Dehnel, I960). The crabs acclimated to 10 C would not undergo molting and could be expected to be more tolerant of the higher temperature regimes than the 16 C acclimated crabs, which would avoid warmer temperatures. In addition, activity levels of cold-acclimated Astacus astacus decrease when these crayfish are warmed in water (Lehti-Koivunen and Kivivuori, 1994); if this were the case here for H. mtdus, then cold-acclimated crabs would be less active and would not exhibit an escape response until a higher temper- ature. Though all animals showed avoidance behavior when the temperature was increased, this was not the case when temperature was lowered. Only 30%-45% of the crabs emigrated from water and 20%-40% remained under shel- ters in the air. As expected, crabs acclimated to 10 C appeared to emigrate at a lower temperature than those acclimated to 16 C. However, since only animals that emigrated from the water or from under shelters were used in the analyses, this difference was not statistically signiri- THERMOREGUl.ATION IN HEM1GRAPSUS NUDUS 47 cant (Table 2). The reason that not all the animals exited when the temperature decreased becomes apparent when their long-term survival in extreme temperatures is consid- ered. The crabs also did not recover after a few minutes of exposure at CTMax; they did, however, recover from cool- ing, even after several hours of exposure below CTMin. The same result is reported foiAstacus axtacnx (Lehti-Koivunen and Kivivuori, 1994). Since H. nudus can survive exposure to low temperatures these crabs would not benefit from leaving the water or a protective shelter, where they would become vulnerable to predation. The results of the shuttling experiments between air and water (Table 3) correspond to the behavioral patterns ob- served in the avoidance experiments. In cold water ( 10 C), H. nudus individuals made fewer excursions into air; there- fore the total time spent in air was also less (Table 3). At higher temperatures, the crabs were more active, making a greater number of excursions and spending a greater amount of time in air. In air, H. nudus is able to take up sufficient oxygen via an increased cardiac output (Morris et ai, 1996a, b). However, this is not without cost, especially at higher temperatures, where hemocyanin affinity and pH are affected to a greater degree, suggesting that oxygen delivery to the tissues declines when H. nudus breathes air at warm temperatures (Morris et ai. 1996a, b). Given these factors, the opposite behavior with respect to temperature may have been expected. However, as temperature increases in both air and water, so does oxygen uptake. The possible advan- tages of emigration from warm water into warm air could be a reduction in oxygen demand, as a consequence of evap- orative cooling across the gills (Taylor and Wheatly. 1979). In addition, the CTMax values showed that the crabs toler- ate higher temperatures in air than in water, which may explain why they spend more time in air at higher temper- atures. In the present study, only two or three animals (tested at the 30 C regime) spent longer than 5 h emersed (Table 3), whereas Greenaway et ai (1996) found that H. nudus can remain emersed for up to 8 h. These workers were using colder water temperatures (10-13 C) than this animal is normally exposed to in summer (Gosselin and Chia, 1995). Thus, the crabs were probably moving into the warmer air (19-22 C) due to a thermal preference rather than to the selection of a particular medium. To test this hypothesis, the behavior of H. nudus was investigated as crabs shuttled between air and water of differing tempera- tures, to determine if they were able to maintain the body temperature within a preferred range. Although purple shore crabs are unlikely to encounter such extreme differences in air and water temperatures as shown in Figure 3, the results obtained suggest that they possess well-developed thermosensory mechanisms. The crabs tended to migrate to the air-water interface; although a slight microhabitat may have existed there, they exhibited a complex series of behaviors that suggested they were using the thermal properties of both media to control body temperature (Fig. 3). The crabs raised or submerged their bodies in the shallow water of the chamber, thus gaining the benefits of evaporative cooling from the gills (Taylor and Wheatly, 1979) without the imbalances in pH and hemocy- anin affinity caused by longer emersion in adverse temper- ature regimes (Morris et ai, 1996a. b. c). Maintenance of an optimal body temperature, rather than selection of a partic- ular medium, appeared to be most important factor. In support of this conclusion, when water temperatures of 4-5 "C and air temperatures of 33-34 C were offered (Fig. 3, trial 1 ). the crabs were able to maintain a body temperature of about 8-12 C, independent of the two media (Fig. 4a). When air and water temperatures were reversed (Fig. 3, trial 7), the body temperature was still maintained within similar limits (Fig. 4b). It was also important to investigate the thermoregulatory behavior of H. nudus in the field, since this species displays different behaviors in its natural environment (McGaw. 2001 ). Greenaway et ai ( 1996) suggest that H. nudus may routinely experience 10 C differences in body temperature in the field. Certainly, the porcelain crab Petrolisthes cinc- tipes. which occupies a similar niche, may be exposed to temperatures under rocks in excess of 20 C (Stillman and Somero, 1996). And even though H. nudus voluntarily exits into air in the laboratory (Greenaway et ai, 1996; Burnett and McMahon, 1987), this was not observed in any of the experimental animals in the field. They remained under rocks or deep in crevices during the intertidal period. This behavior has adaptive significance in that it keeps crabs in close contact with cover, thus avoiding the threat of preda- tion (Low, 1970; Duly, 1981: McGaw, 2001). Indeed, H. nudus prefers to shelter underneath larger boulders, which provide the added advantages of heating or cooling more slowly (Stillman and Somero, 1996). Thus, using subtle movements within this thermal microhabitat, the crabs were able to maintain their body temperature independent of the surrounding air (Fig. 5). Additionally, the prevailing weather conditions can have a profound effect on the mi- crohabitat and behavior of animals (Stillman and Somero, 1996). I have observed crabs active at low tide on humid or dull days; clearly other factors, in combination with tem- perature, play a role in emersion behavior and deserve further investigation. H. nudus is well adapted for an existence in the intertidal zone (Morris et ai, 1996a. b, c; Greenaway et ai, 1996). The present study demonstrates that this species is able to detect differences in its thermal environment and use the thermal properties of both water and air to control its body temperature within a fairly narrow range. This study extends the work on thermoregulatory behavior in aquatic crusta- ceans (Crawshaw, 1974; Reynolds and Casterlin, 1979a, b, c, d: Lewis and Roer. 1988; Mundahl and Benton, 1990; Espina et ai. 1993; Kivivuori, 1994; Lehti-Koivunen and 48 I. J. McGAW Kivivuori, 1994; Crossin ct ai, 1998) by examining the responses of an amphibious species during exposure to temperature change in both aquatic and aerial environments. Acknowledgments I thank the Director and staff of the Bamfield Marine Sciences Centre for the use of facilities. 1 also thank Drs. Winsor Watson III (Univ. of New Hampshire) and Steven Jury (SUNY) for helpful discussion. This work was sup- ported by a UNLV SITE and New Investigator Award. Literature Cited Burnett. I.. E., and B. R. McMahon. 1487. Gas exchange, hemolymph acid-base status and the role of branchial water stores during air exposure in three littoral crab species. Physio/. Zool. 60: 27-36. Chang. E. S. 1995. Physiological and biochemical changes during the molt cycle in decapod crustaceans: an overview. J. Exp. Mar. Biol. Ecol. 193: 1-14. Cooke-Schreiber, S. M., S. H. Jury, and W. H. Watson III. 2001. Seasonal difference in behavioral thermoregulation in the lobster, Ho- iniinix ameriiaiuix. Am. Zoo/. 40: 478A. Crawshaw, I,. I. 1974. Temperature selection and activity in the craw- hsh Orconectes immnnis. J. Comp. Ph\siol. 95: 161-172. Crossin, G. T., S. Abdulazziz AI-Ayoub. S. H. Jury. W. H. Howell, and W. H. Watson III. 1998. Behavioral thermoregulation in the Amer- ican lobster Homarux ninericannx J. &/>. Biol. 201: 365-374. Cuculescu, M.. D. Hyde, and K. Bowler. 1998. Thermal tolerance of two species of marine crab. Cancer pagitrus and Carcinux inaenax. ./. Therm. Biol. 23: 107-1 10. Daly, G. P. 1981. Competitive interactions among three crab species in the inlcrtidal /one. Ph.D. Thesis. University of Oregon. deFur. P. L., and C. P. Mangum. 1979. The effects of environmental \ariahles on the heart rates of invertebrates. Comp. Bun-hem. P/nsiol. 62A: 2X3-294. Dehnel, P. A. 1960. Effect of temperature and salinity on the oxygen consumption of two intertidal crabs. Biol. Bull. 118: 215-249. DeWachter, B., and B. R. McMahon. 1996. Temperature effects on heart performance and regional hemolymph How in the crab Cancer magister. Comp. Biochem. Ph\siol. 114(1): 27-33. DeWachter. B.. F. J. Sartorius, and H. O. Portner. 1997. The anaer- obic end product lactatc has a behavioural and metabolic signalling function in the shore crab Carcinus inaenax. J. E.\p. tiiol. 200: 1015- 1024. Edney, E. B. 1961. The water and heat relationships of fiddler crabs (Ucn spp.). Trans. K. Sue. S. Afr. 36: 71-41. Espina, S.. F. I). Herrera, and L. F. Buckle. 1993. Preferred and avoided temperatures in the crawfish. Procunthanix clarkii. ./. Therm. Biol. 18: 35-34. Fredrich, M., B. DeWachter, F. J. Sartorius, and H. O. Portner. 2000. Cold tolerance and the regulation of cardiac performance and hemo- lymph distribution in Maja xaninailo (Crustacea: Decapoda). Pln-xiol. Kiochcin. Zonl. 73: 406-415. Gosselin, L. A., and F. S. Chia. 1995. Characten/ing temperate rocky shores from the perspective of an earh juvenile snail: the main threats to survival of newly hatched Nucella emargiiitita. Men. Hiol. 122: 625-635. Greenaway, P., S. Morris, K. R. McMahon, C. A. Farrelly, and K. L. Gallagher. 1996. Air breathing by the purple shore crab Heinigrup- \nx iiiuhis (Dana). I. Morphology, behaviour and respiratory gas ex- change. I'lnxio/. /.*/ 69: 7X5-X05. Hutchison, V. H., and J. D. Maness. 1979. The role of behavior in temperature acclimation and tolerance in ectotherms. Am. Zool. 19: 367-384. Jacoby, C. A. 1981. Behavior of the purple shore crab Hemigrapsus iniilin Dana. 1X51. J. Cmxtac. Biol. 1: 531-544. Jansen, P. 1970. Eco-physiological studies on the "posing" behaviour of Uca tangeri. Forma Functio 2: 58-100. Jury, S. H., and W. H. Watson HI. 2000. Thermosensitivity of the lobster. Homarus americamis. as determined by cardiac assay. Biol. Bull. 199: 257-264. Kivivuori. L. A. 1994. Temperature selection behaviour of cold and warm acclimated crayfish (AvMiw axtacnx}. J. Therm. Biol. 19: 291- 247. Korhonen. I. A., and K. Y. H. Lagerspetz. 1996. Heat shock response and thermal acclimation in Axel/itx aquaticus. ./. Therm. Biol. 21: 44-56. Lagerspetz, K. Y. H., and K. Bowler. 1993. Variation in heat tolerance of individual Asel/us ai/naticnx during thermal acclimation. J. Therm. Biol. 18: 137-143. Layne, J. R., D. L. Claussen, and M. L. Manis. 1987. Effects of acclimation temperature, season and time of day on the critical thermal maxima and minima of the crayfish Orconectes ruxiicnx. ./. Therm. Biol. 12: 1X3-187. Lehti-Koivunen, S. M., and L. A. Kivivuori. 1994. Effect of tempera- lure acclimation in the crayfish Asruciis axtacux (L.) on the locomotor activity during a cycle of temperature change. J. Therm. Biol. 19: 244-304. Lewis, D. H., and R. D. Roer. 1988. Thermal preference in distribution of blue crabs, Callinectex xapultix. in a power plant cooling pond. J. Cmxtac. Biol. 8: 2X3-2X4. Lindeberg, W. J. 1980. Behaviour of the Oregon mud crab Hemigrapxux oregonenxix (Dana) (Brachyura: Grapsidae). Crustaceana 39: 263- 281. Low, C. .1. 1970. Factors affecting the distribution and abundance of two species of beach crab. Hemigrapxux oregonensix and Hemigrapsus multix. MSc Thexix. University of British Columbia, BC, Canada. Maitland, D. P. 1990. Carapace and branchial water circulation and water related behaviours in the semaphore crab Heloecius coni/formix (Decapoda: Brachyura: Ocypodidae). Mar. Biol. 105: 275-286. McGaw, I. J. 2001. Impacts of habitat complexity on physiology: Purple shore crabs tolerate osmotic stress for shelter. Extuurine Coastal Shelf Sci. 53: 865-876. Morris, S., P. Greenaway, and B. R. McMahon. 1996a. Air breathing by the purple shore crab Hemigrapsus nitdus (Dana). II. Respiratory gas and acid-base status in response to emersion. Physiol. Zool. 69: X06-83X. Morris, S., P. Greenaway, and B. R. McMahon. 1996b. Air breathing by the purple shore crab Hemigrapsus niidns (Dana). III. Haemocyanin (unction in respiratory gas transport. P/ivxiol. Zool. 69: 834-863. Morris, S., P. Greenaway, and B. R. McMahon. 1996c. Air breathing by the purple shore crab Hemigrapxux niuhis (Dana). IV. Aquatic hypoxia as an impetus for emersion? Oxygen uptake, respiratory gas transport, and acid-base state. Phvsiol. Zool. 69: X64-886. Mundahl. N. D., and M. J. Benton. 1990. Aspects of the thermal ecology of the rusty crayfish Orconectes riisticus (Girard). Oecologia 82: 210-216. O'Halloran, M. J., and R. K. O'Dor. 1988. Molt cycle of male snow crabs Chionocetes opi/io from observations of external features, setal changes and feeding behavior. J. Cmxtac. Biol. 8: 164-176. Payette, A. L., and I. J. McGaw. 2001. Can ovigerous crayfish (Pro- ccimhamx cltirkii) control their brooding environment through behav- ior'' Ari:ona-NevaJa Acaci. Sci. 36: I2A. Pirro, M. D., S. Cannicci, and G. Santini. 1999. A multi-factorial experiment on heart rate variations in the intertidal crab Pacln-grapxux marmonitns. Mar. Biol. 135: 341-345. THERMOREGULATION IN HEMIGRAPSUS NUDUS 49 Ramirez, L. F. B.. F. D. Herrera, F. C. Sandoval. B. B. Sevilla. and M. H. Rodriguez. 1994. Deil thermoregulation of the crawfish Pro- lamhanis c/arkii (Crustacea. Camhridae). ./. Therm. Biol. 19: 419- 422. Reynolds, \V. H., and M. E. Casterlin. 1979a. Behavioral thermore- ulution and activity in Homarus americanus. Comp. Biocliem. Plivsiol. 64A: 25-28. Reynolds, VV. H., and M. E. Casterlin. 1979b. Thermoregulatory be- havior of the pink shrimp Penaeitx Jiioranini Burkenroad. Hn/mbio- logia 67: 179-182. Reynolds, W. H., and M. E. Casterlin. 1979c. Behavioral thermoreg- ulation in the grass shrimp Palaemonetes vulgaris (Say). Rev. Can. Biol. 38: 45-46. Reynolds, W. H., and M. E. Casterlin. 1979d. Behavioral thermoreg- ulation in the spiny lobster Panulirus argus. H\drohiologia 66: 141- 143. Sastry, A. N. 1983a. Pelagic larval ecology and development. Pp. 213- 282 in The Biology of Crusraceu, Vol. 7: Behavior and Ecolog\, F. J. Vernberg and W. B. Vernberg. eds. Academic Press. New York. Sastry, A. N. 1983b. Ecological aspects of reproduction. Pp. 179-270 in The Biology of Crustacea. Vol. 8: Environmental Adaptations. F. J. Vernberg and W. B. Vernberg. eds. Academic Press. New York. Schmitt, \V. I. 1921. The marine decapod Crustacea of California. Univ. Calif. Publ. Zool. 23: 1-470. Smith, W. K., and P. C. Miller. 1973. The thermal ecology of two South Florida tiddler crabs Uca rapax (Smith) and Uca pugilator (Bosc). Physiol. Zoo/. 46: 186-207. Stillman. J. H., and G. N. Somero. 1996. Adaptation to temperature stress and aerial exposure in congeneric species of intertidal porcelain crabs (Genus: Petrolisthes): correlation of physiology, biochemistry and morphology with vertical distribution. ./. v/>. Binl. 199: 1845- 1 855. Stillman, J. H., and G. N. Somero. 2000. A comparative analysis of the upper thermal tolerance limits of eastern Pacific porcelain crabs. Genus Petrolisthex: influences of latitude, vertical zonation. acclimation and plnlogeny. Pliyxiol. Biocliem. Znol. 73: 200-208. Taylor, E. VV. 1981. Some effects of temperature on respiration in decapodan crustaceans. J. Therm. Biol. 6: 239-248. Taylor, E. W., and M. G. Wheatly. 1979. The behaviour and respira- tory physiology of the shore crab, Carcinus maemns (L.) at moderately high temperatures. J. Comp. Physiol. 130: 309-316. Thurman, C. L. 1998. Evaporative water loss, corporal temperature and the distribution of sympatric crabs (Uca) from South Texas. Comp. Biochem. Phyxiol. 119A: 279-286. Todd, M. E., and P. A. Dehnel. 1960. Effect of temperature and salinity on heat tolerance in two grapsoid crabs Hemigrupsits midux and Hem- igrapsus oregonensis. Biol. Bull. 118: 150-172. Truchot. J. P. 1983. Regulation of acid-base balance Pp. 431-457 in The Biology of Crustacea, Vol 5: Internal Aiuitomv and PhvMnlnftical Regulation. L. H. Mantel, ed. Academic Press. New York. Vernberg, W. B., and F. J. Vernberg. 1972. Environmental Phyxiologv of Marine Organisms. Springer. New York. Whiteley, N. M., A. H. Al-Wassia, and E. W. Taylor. 1995. The effects of sudden changes in temperature on aquatic and aerial respiration in the lobster Homam.i gainmarns (L.). Mar. Frexlm: Behav. Phvsiol. 27(1 1: 13-27. \\ilkens, J. L., and M. Fingerman. 1965. Heat tolerance and temper- ature relationships of the fiddler crab, Uca pugilator. with reference to body coloration. Biol. Bull. 128: 133-141. Reference: Biot. Bull. 204: 50-56. (February 2003) 2003 Marine Biological Laboratory Synthesis of a High-Density Lipoprotein in the Developing Blue Crab (Callinectes sapidus) ANNA WALKER 1 , SEICHI ANDO 2 . AND RICHARD F. LEE 1 * ' Department of Patholog\, Mercer University School of Medicine, Macon, Georgia 31207: Lipoprotein Research Laboratory, Department of Fisheries Science, Kagoshima University, 4-50-20 Shimoarata, Kagoshima 890-0056, Japan: and Skida\ra\ Institute of Oceanography, Savannah, Georgia 31411 Abstract. An important lipoprotein in the hemolymph of crustaceans is Lpl. It transports lipid to peripheral tissues and also has a role in crustacean immune recognition. We employed a monoclonal antibody specific for the Lpl pep- tide to demonstrate by ELISA, western blot and immuno- histochemistry the appearance of Lpl during development of Callinectes sapidus, the blue crab. Lpl was first found in stage 5 embryos and appeared to be synthesized by lateral basophilic cuboidal cells that demonstrated cytoplasmic im- munoreactivity for Lpl at their interface with the yolk mass. The embryonic cuboidal cells bore a strong cytologic re- semblance to the hepatopancreas cells of later stages (zoea, megalopae, adults), which were also immunoreactive for Lpl. Introduction The hemolymph of male and female decapod crustaceans contains a high-density lipoprotein (Lpl) with concentra- tions ranging from 1.1 to 2.0 mg/1 (Lee and Puppione, 1988; Spaziani, 1988; Lee. 1991; Spaziani and Wang, 1991; Stratakis et ai. 1992; Tom et ai. 1993; Yepiz-Plascencia ct ill., 1995; Ruiz-Verdugo et ul.. 1997). It plays an important role in transporting lipids from the hepatopancreas to pe- ripheral tissues such as muscle, and functions as a j8-l,3- glucan-binding protein in crustacean immune recognition (Khayat ct ai, 1994; Hall ct ai. 1995; Kang and Spaziani, 1995). Embryos of Callinectes sapidus. the blue crab, develop in ess sacs through a series of 10 staues (Table I) over a Received 25 July 2002; accepted 8 November 2002. * To whom correspondence should be addressed skio.peachnet.edu E-mail: dick(S' period of 16-23 days. At stage 10, they emerge from the egg sacs as swimming zoea larvae; these metamorphose into megalopae. then into juvenile crab forms, and ultimately become adult crabs. Until they emerge from egg sacs, embryos are nutritionally dependent on lipids and lipo- vitellin stored within the eggs. Lipovitellin (LpII) is a high-density lipoprotein that differs from Lpl in density, sediment coefficient, and peptide components (Lee and Puppione, 1988; Lee and Walker, 1995). In adult blue crabs, Lpl is composed of phospholipids (45%). cholesterol (2%). triacylglycerols (3%), and one peptide (49%, molecular mass 1 12 kD) (Lee and Puppione. 1988). Although Lpl was reported in juvenile and adult blue crabs, it has not been previously reported in crab oocytes or embryos. We employed a monoclonal antibody specific for the Lpl peptide to demonstrate by ELISA. western blot, and immunohistochemistry the appearance of Lpl during blue crab development. In addition, we offer immunohisto- chemical evidence that the developing hepatopancreas is the site of Lpl synthesis in embryonic and larval stage blue crabs, and remains so in the adult. Materials and Methods Collection of crabs, isolation of Lpl. and purification of Lpl peptide Blue crabs were collected by trawling in the estuaries near Skidaway Island, Georgia (USA). Hemolymph was collected with a 5-nil disposable syringe from the base of the swimming leg and centrifuged in a low-speed centrifuge (3 C) for 10 min at 2000 X g to remove clotted materials and cells. Hemolymph lipoproteins were separated from other hemolymph proteins by adjusting the density of the 50 LIPOPROTEIN IN DEVELOPING BLUE CRAB 51 Table 1 Description of embryo stages o/Callinectes sapidus Stage Description Elapsed Time (hours at 27 C) 1 Fertilization T Early cleavage; morula 12 (random mass of yolk cells) 3 Late cleavage; blastula 36 (mass of undifferentiated yolk cells) 4 Embryonic naupliar stage; transparent 85 embryo above the yolk 5 Early appendage formation; embryo 111) invading ventral portion of yolk 6 Embryonic eye; eye appears as scarlet 160 crescent; elongating appendages 7 Presence of beating heart; pigmented ISO appendages 8 Oval, pigmented eye; 50% of yolk utilized; 210 clear appendages 9 Compound eye with dark pigmentation; 230 only small amounts of yolk 10 Protozoeae stage ready for hatching into 2X0 free-swimming zoea hemolymph and then centrifuging it (Beckman L5-40 ultra- centrifuge. 40.3 rotor). Salt solutions used to adjust the solutions densities were prepared according to the methods outlined by Lindgren (1975). Solution densities were veri- fied by refractometry using an Abbe refractometer (Bausch and Lomb). Consistent with an earlier study (Lee and Pup- pione, 1988), lipoproteins with densities less than 1.063 g/ml were not detected in blue crab hemolymph. Thus, blue crab high-density lipoprotein (Lpl) was isolated by adjust- ing the density of hemolymph to 1.21 g/ml with solid potassium bromide, followed by 40 h of centrifugation at 1 17,000 X g. The floating layer of high-density lipoprotein was removed and dialyzed for 24 h at 4 C against 0.22 M NaCl containing 1 mM EDTA and 2 mM sodium azide. After dialysis, lipoproteins were run on vertical slab gels (7% polyacrylamide. 0.1% sodium dodecyl sulfate (SDS). 0.8% mercaptoethanol), following the procedures of Laemmli ( 1970). The protein band (apoLpI) was visualized with 0.3 M copper chloride. The apoLpI was cut from the gel and eluted by electrodialysis (Electro-Eluter model 422. Bio-Rad). The eluted apoLpI was dialyzed against 0.22 M NaCl. SDS- polyacryamide electrophoresis of the purified peptide was carried out to verify its purity. Purified apoLpI was used as the antigen for the preparation of monoclonal antibodies. Monoclonal antibody production Four female BLB/c mice were immunized with 50 fil of apoLpI (0.1 mg/ml) mixed with Freud's complete adjuvant. The injections were repeated twice (4 weeks and 6 weeks after the original injection) with 25 /u,g of apoLpI in Freud's incomplete adjuvant. Three days after the last injection, the mice were sacrificed and spleens removed. The mouse spleen cells were fused with a mouse myeloma strain Sp2/(), using polyethylene glycol as the fusing agent, as described by Galfre and Milstein (1981). After fusing, cells were plated in hypoxanthine/aminopterin/thymidine selection medium in microplates on a feeder layer consisting of mouse peritoneal macrophages. The wells were screened by indirect ELISA (enzyme-linked immunosorbent assay) for antibodies to apoLpI. The positive hybridomas were cloned by limiting dilution. The hybridomas were grown on RPMI 1640 medium (Gibco) supplemented with 10% fetal calf serum and 0.01% streptomycin and penicillin. Indirect ELISA assa\ fur apoLpI An indirect ELISA assay was used to test antibodies, using the procedures described by Lee and Walker ( 1995), and absorbance was measured at 410 nm with an ELISA microplate reader (model EL307C. Bio-Tek Instruments). Indirect competitive ELISA for apoLpl A criss-cross serial dilution analysis was carried out to determine the optimal concentrations of apoLpI and anti- body (Hornbeck el i>ln>m ct-cnipiii: support of yolk formation and preparation tor (light. Insect Hinclicin. 21: 653-663. Tom, M., O. Shenker, and M. Ovadia. 1993. Partial characterization of three hemolymph proteins of Pcmiciis semisulcatus de Haan (Crusta- cea. Decapoda. Penaeidae) and their specific antibodies. Ciunp. Bio- chcm. Physio/. 104B: Sll-SK.. Van der Horst, D. J. 1990. Lipid transport function ol lipoprotems in flying insects. Biochim. Biophys. Ada 1047: 195-21 I. Wheeler. D. E., and J. K. Kawooya. 1990. Purification and character- ization of honey hee vitcllogenin. Arch. Insect Biochem. Physio/. 13: 253-267. Yepiz-Plascencia. G. M., R. Sutelo-Mundo, L. Vasques-Morena, R. Zieglt'r, and I. Higurera-Ciapara. 1995. A non-sex specific hemo- lymph lipoprotein from white shrimp Pi-nucns ninnaniei Boone. Iso- lation and partial characterization. OWH/>. Biuchem. Phy.iiol. 11 IB: 181-187. Yepiz-Plascencia, G., F. Vargas-Albcires, F. Jimenez-Vega, L.M. Ruiz- Verdugo, G. Romo-Figueroa. 1998. Shrimp plasma HDL and (3-glucan binding protein (BGBP): comparison of biochemical charac- teristics. Comp. Biochem. Ph\\inl. 121B: 309-314. Reference: Biol. Bull. 204: 57-67. (February 2003) 2003 Marine Biological Laboratory Reproductive Biology of Hemiramphus brasiliemis and H. balao (Hemiramphidae): Maturation, Spawning Frequency, and Fecundity RICHARD S. McBRIDE* AND PAUL E. THURMAN Florida Marine Research Institute, Florida Fish and Wildlife Conservation Commission, 100 8th Avenue SE, St. Petersburg, Florida 33701-5095 Abstract. Analyses of life-history data show that both the size-specific batch fecundities and the age-specific spawn- ing frequencies differ for two halfbeak species, Hemiram- phus hrasiliensis. the ballyhoo, and H. halao, the balao. Halfbeak ages were determined from sectioned otoliths; histological data was used to describe oocyte development and estimate spawning frequency; and batch fecundity was measured from counts of whole oocytes in final maturation. Hemiramphus hrasiliensis lived longer (4 versus 2 years) and had a higher survival rate ( 14.9% versus 7.5% annually) than H. halao did. Of the two species the larger and longer- lived congener, H. hrasiliensis, reached sexual maturity at a larger size (fork length 198 versus 160 mm). The spawning period of age-0 females was strongly related to season, whereas spawning by older females occurred throughout the year. Reproduction by both species peaked during late spring or early summer, and all mature females were spawn- ing daily during April (H. hrasiliensis) or June (H. halno). This is the first demonstration of iteroparity for the family Hemiramphidae. H. hrasiliensis had a lower batch fecundity (about 1164 versus 3743 hydrated oocytes for a 100-g female) than H. halao did. Such low batch fecundities are typical of the order Beloniformes, but quite different from those of other fishes that live in association with coral reel habitats. H. halao' s higher batch fecundity is consistent with the life-history theory that predicts higher numbers of eggs for shorter-lived species; this is possible because H. halao produces smaller hydrated oocytes than H. hrasilien- sis (modal diameter about 1.6 versus 2.4 mm). The high spawning frequency of Hemiramphus species compensates Received 21 June 2002; accepted 6 November 20O2. * To whom correspondence should be addressed. E-mail: richard. mcbride@fwc. state. fl. us for their low batch fecundity. The annual fecundity of both species is similar to that of other reef fish species, after adjusting for body size and spawning frequency. The life- time fecundity of H. halao was very similar to that of H. hrasiliensis, after accounting for the differences in survival for each species. This suggests a fine tuning of different reproductive traits over the entire life cycle that results in roughly equivalent lifetime fecundity for both species. Introduction Two pelagic halfbeak species. Hemiramphus hrasiliensis and H. halao, are conspicuous and abundant elements of the Atlantic Ocean's coral reef fauna ((Toilette, 1965; Ny- bakken, 1997, p. 368; McBride el al.. 2003). These conge- ners are similar in size and shape (about 30 cm maximum length: McBride el al, 1996) but differ in both habitat use and diet. Both halfbeak species intermingle above coral reef habitats; otherwise, H. hrasiliensis is found only inshore of reef habitats and H. balao is found only offshore of reefs (McBride et al., 2003). H. hrasiliensis preys on zooplankton and grazes on seagrasses, whereas H. halao is a planktivore (Berkeley and Houde. 1978). Berkeley and Houde (1978) also characterized both species as oviparous summer- spawners with low batch fecundities (i.e., the number of eggs released per spawning event: Hunter et al., 1985). and they reported that H. hrasiliensis lived longer but had a lower batch fecundity than H. halao. These life-history patterns (i.e., age, reproduction, and mortality) are particu- larly intriguing because such patterns suggest a trade-oft between survival and reproductive output. Comparing life-history traits within species and between morphologically similar species in different habitats is a powerful method for understanding life-history evolution 58 R. S. McBRIDE AND P. E. THURMAN (Partridge and Harvey, 1988). If the life-history patterns of fishes evolve largely in response to their environment, it is striking that H. hrasiliensis and H. balao have much larger eggs but lower batch fecundities (i.e.. egg diameter > 1 mm and thousands of eggs per batch: Berkeley and Houde. 1978) than other coral reef fishes (e.g., see Thresher, 1984). These large eggs and low batch-fecundity values may sim- ply reflect the evolutionary history of hemiramphids. Aver- age batch fecundities for other oviparous hemiramphids range from one hundred (Silva and Davies, 1988; Coates and Van Zwieten, 1992) to a few thousand eggs per female (Talwar, 1962, 1967). In contrast, an average female coral reef fish with a body size similar to that of a Hemiramphus species produces about 100,000 eggs (Thresher, 1984). Such low fecundities for Hemiramphus species imply either high fertilization success, high survival rates, or the produc- tion of multiple batches of eggs. Multiple spawning is a common life-history trait among marine fishes and can greatly increase lifetime reproductive output. Although multiple spawning has been suspected to occur in several hemiramphids, it has never been demonstrated conclusively (e.g.. Ling. 1958; Talwar. 1967; Coates and Van Zwieten, 1992). In this study, multiple spawning is demonstrated for both H. hmsiliensis and H. balao. and new measurements are made of other life-history variables, namely age, mortality, size at maturity, egg size, batch fecundity, and spawning frequencies. Such detailed measurements demonstrate the interaction of phenotypic traits that determine fitness in two congeneric hemiramphids. We also compare these traits for Hemiramphus species with those of other hemiramphids and other coral reef fish species to evaluate the importance of evolutionary history in constraining allocation of repro- ductive effort. Materials and Methods Fishes were collected in the coastal waters of southeast- ern Florida (approx. 26.0 N, 80.0 W to 24.5 N, 82.2 W). Hemiramphus brasiliensis and H. balao were collected to- gether near the surface in association with coral reefs. H. brasiliensis alone was collected in other inshore habitats such as bank habitats in nearby Florida Bay, and so it is more numerous in our collections overall. From July 1997 to October 1998, 100 to 200 fish were subsampled. on each of 4 days per months from the catch of commercial fishing operations. Additional specimens were collected indepen- dent of the commercial fishery for a target number of 4 additional trips per month and a sample size of 12 fish per trip. Fish were kept on ice and brought to the laboratory for processing. Fish lengths and weights were measured in the laboratory. Fork length (FL) was measured to the nearest millimeter from the tip of the upper jaw to the fork of the tail. Whole body weight was recorded to the nearest O.I g. Ages of halfbeaks were determined by examining annual increments deposited on otoliths. For each trip in the months from July 1997 to June 1998, 12 fish were selected, at random, for aging; their sagittal otoliths were removed and stored dry. A low-speed saw was used to cut multiple 500-/xm-thick sections along the transverse plane through the otolith core. Otoliths were cut only from fish larger than 200 mm in FL, because otoliths of smaller fish are known to be age-0 (Berkeley and Houde, 1978). Sectioned otoliths were mounted to coded glass slides and examined, usually at 40 X, with reflected light under a dissecting microscope. The annuli were counted as a measure of fish age. in years, by two readers. If the two independent counts did not agree, then a third reading was conducted, with both readers work- ing together. Only 5% of the otoliths were so difficult to evaluate by both readers that they were rejected (/; = 61 ). The frequency of annulus formation was confirmed as an- nual by a marginal increment analysis. In such an analysis, the percentage of age-1 ballyhoo with an opaque margin, which was interpreted as a second annulus, was calculated for each month; monthly frequencies were checked for periodicity of annulus formation. Annual survival estimates (S) for each species were de- rived using the estimator from Robson and Chapman s= where x is the coded age class (0 = youngest age [in years] fully vulnerable to fishing), / v is the number of fish per age-class .v, and A is the oldest age class observed. The data for this analysis were only from the period October-May, because age-0 fish are not fully vulnerable to the sampling gear during the summer months (Berkeley and Houde, 1978). Gonads from 1 2 randomly selected fish in each collection were removed during the period of July 1997 to October 1998 and prepared for histology. Ovarian tissue was ini- tially fixed in 10% buffered formalin; a section of tissue was then transferred to ethanol, embedded in glycol methacry- late, sectioned along the transverse plane, stained with the periodic acid-Schiff (PAS) reaction, iron-hematoxylin, and counterstained with metunil yellow (Quintero-Hunter et ul., 1991). Gonads were assigned a stage based on the most advanced stage of oocyte development, namely perinucleo- lar. cortical alveolar, vitellogenic, nucleus migration, or nucleus breakdown. Cellular atresia such as postovulatory follicles (POP) and PAS-positive melano-macrophage cen- ters were also noted. Characterization of POFs follows the descriptions of Hunter and Macewicz ( 1985). Identification of PAS-positive bodies follows the descriptions in Grier and Taylor (1998; pp. 531. 539-540) and McBride et al. (2002). MARINE HALFBF.AK REPRODUCTION 59 Table 1 Maturity categories for female Hemirumphus ,v/i/>. Maturity Category Most Advanced Oocyte Stage Atresia and POFs'' Immature Perinucleolar stage Maturing Cortical alveolar (CA) Little or no atresia Little or no atresia Mature Vitellogenic or FOM h stage may be POFs; PAS+ bodies 1 Regressed Perinucleolar or CA stage PAS + bodies'" Females were scored according to their most advanced oocyte stage, and past spawning was inferred based on the presence of postovulatory follicles (POFs) and PAS-positive melano-macrophage centers (see text for details). The dashed line separates immature from mature stages, the division used for calculating size at 50% maturity. '' Postovulatory follicles (POFs) were observed as newly collapsed struc- tures after dusk. They were readily observed for about 24 h, after which they became more compact and darker. h Final oocyte maturation (FOM) began with migration of the nucleus, continued with breakdown of the nucleus ( = hydration), and ended with ovulation of eggs. L PAS-positive (PAS + ) melano-macrophage centers appeared as com- pact, bright purple bodies when our staining technique was used and are similar to yellow or brown bodies when other stains were used. The presence of vitellogenic oocytes was the primary indication of maturity (Table 1 ). Vitellogen, a protein se- creted by the liver and endocytosed by oocytes, accumulates in yolk globules that appear in the cytoplasm of oocytes during the spawning season (Wallace and Selman. 1978). Mature females with regressed ovaries were distinguished from immature virgin or maturing virgin females by the presence of PAS-positive melano-macrophage centers. Such PAS-positive bodies are involved in focal tissue deg- radation, and their presence increases during and after go- nad regression (Grier and Taylor, 1998). Size at 50% ma- turity was calculated as the inflection point of a logistic equation modeling the percent frequency of mature females: maturity = 1/(1 + exp(-/UFL - #])), where A -- the instantaneous rate of increase at the origin and B = the inflection point or the point where 50% of the individuals are mature. Model parameters were estimated by the logistic procedure of SAS software (SAS, 1990). Spawning frequency was estimated by the "post-ovula- tory follicle method" of Hunter and Macewicz (1985). We assumed that POFs became indistinguishable from other atretric bodies after about 24 h, so their presence as col- lapsed structures with identifiable thecal and granulosa lay- ers indicated that an individual female had spawned during the previous day. Batch fecundities and oocyte diameters were determined from examination of whole oocytes. Batch fecundities were estimated for 41 specimens of H. brasiliensis and 3 of H. balao collected in March 1997, May 1997, February-April 1998, and March 1999. About 1 g of tissue was removed from the anterior and posterior sections of the left and right ovaries, blotted dry. and weighed to the nearest 0.001 g. After the tissue was washed, teased apart, and placed in a solution of 33% glycerin:67% water, the number of hy- drated oocytes was counted. Batch fecundity was estimated according to the total weight of the ovary, following the methods of Hunter ft ul. (1985). To increase sample size, these data are presented together with data from Berkeley and Houde ( 1978), who used a similar method for estimat- ing batch fecundity and obtained similar results. Whole oocyte diameters, for at least 300 oocytes per female, were then measured to the nearest micrometer with the aid of a video system and image-analysis software. Results Age, growth, and xun'h'iil Sectioned otoliths revealed an alternating pattern of opaque and translucent bands (Fig. 1 ). The darker, translu- cent areas represented the periods of faster growth during the summer; the whitish, opaque bands reflected the periods of slower growth during the winter. Marginal increment analysis of Hemiramphus brasiliensis otoliths showed that a single annulus was formed each year and that annulus formation was complete by June of each year (Fig. 2). Although too few specimens of H. balao were available for the marginal increments of this species to be similarly analyzed, we chose June as the biological hatchdate for both species. Our aging results indicated that H. brasiliensis lives longer than H. balao (4 versus 2 years; Fig. 3). At a given age, individuals of H. brasiliensis were also larger on av- erage than those of H. balao, and females of each species were larger than male conspecifics. Of the 1022 specimens of H. brasiliensis aged, the largest was 294 mm FL, whereas of the 132 H. balao specimens aged, the largest was 251 mm FL. Annual survival of H. brasiliensis during the study period averaged 14.9% (95% confidence limits: 12.2%- 17.6%) and was double that of H. balao (7.5% on average; 2.01%-13.0%, 95% c.l.). Reproduction Ovaries of both halfbeak species were composed of two cylindrical lobes, roughly equal in size. During initial mat- uration of the specimens we examined, and again during spring recrudescence, these lobes increased in girth, devel- oped a pinkish color, and extended anteriorly along the coelomic cavity. Each lobe was a hollow sac with oocytes arranged in lamellae that extended into a central lumen (Fig. 4A-C). Virgin females lacked any PAS-positive melano- macrophage centers and had a thin gonad wall (Fig. 4A, C); mature, regressed females had PAS-positive melano-mac- rophage centers and a thick gonad wall (Fig. 4B). Oocytes 60 R. S. McBRIDE AND P. E. THURMAN Figure 1. Photomicrographs of sectioned utolilhs representing various age classes for Hemiramphus //liii* hni.\ilicnxix (ballyhoo), with a whitish, opaque annulus on the margin of the sectioned otolith, by month. Otoliths were collected from July 1997 to June 1998 (n = 1019). and fresh postovulatory follicles could be observed (Figs. 4J. M; 5C). Vitellogenic oocytes not in the final stages of oocyte maturation were observed together with oocytes that had migrating nuclei and with hydrated oocytes (Fig. 4N), which suggested a fairly rapid turnover of oocytes. The patterns of oocyte development for the two species did not appear to differ, except that the modal diameter of hydrated oocytes was only about 1 .6 mm for H. balao compared to about 2.4 mm for H. brasiliensis (Fig. 5B, D). In addition, females of H. halao matured (size at 50% maturity = 160 mm FL) at a smaller size than females of H. brasiliensis (i.e.. size at 50% maturity == 198 mm FL; Fig. 6). Both species became mature as age-0 fish (i.e., young-of-the- year). There were distinct seasonal trends in maturation, and the seasonal patterns of POP occurrence indicated prolonged, albeit slightly staggered, spawning periods for both species (Figs. 7, 8). During autumn, the incidence of immature and maturing individuals increased because age-0 fish were more frequently caught in the sampling gear; and the gonads of older fish were regressing as winter approached. Spawn- ing frequency peaked in April for H. brasiliensis and in June for H. balao. but spawning by some females was evident year-round based on the continued presence of POFs in older fish. Spawning frequencies were clearly age-specific, even for such short-lived species. All age classes of both species were spawning on a daily, or near-daily, basis, during spring or summer (Fig. 8). In other seasons, age-1 females spawned more frequently than age-0 females did, but age-2 ballyhoo females did not necessarily spawn more frequently than age-1 ballyhoo females. In terms of envi- ronmental associations, spawning activity peaked when the photoperiod was longest (i.e., June), and most juveniles were growing when temperatures were highest (/ e August' Fig. 9). Batch fecundity was higher for H. balao than for H. brasiliensis across the range of mature sizes (Fig. 10). Batch fecundity also increased more rapidly with increasing fish size in H. balao than in H. hrasiliensis. The average batch size for a 100-g female of H. halao (3743 oocytes) was more than three times that of a 100-g female of H. hrasil- ifiisis ( 1 164 oocytes). At a common size of 200 g, the batch fecundity of H. halao (8346 oocytes) was more than live times that of H. hrasiliensis (1538 oocytes). Discussion We report here, for the first time, the evidence that Hemiramphus species are multiple-spawners with group- synchronous oocyte development. Many general life-history patterns are similar for the two species of halfbeak. Both are short-lived, fast-growing, gonochoristic, and oviparous. Still, there were interspecific life-history differences, some of which varied in a manner predicted by life-history theory. The results of this study show the interplay between differ- ent phenotype-based life-history traits that are balanced out once lifetime reproductive output is calculated. Maximum ages for both species in our study were I or 2 years older than the ages determined in a previous study, but our results do not alter the conclusions that these are fast- growing and short-lived species (Berkeley and Houde, 1978). Berkeley and Houde (1978) reported that the maxi- mum age of 1 100 specimens of H. brasiliensis was age-2 and the maximum age of 135 specimens of//, balao was age- 1 . They read annuli on fish scales, and some researchers have noted that this method tends to underestimate ages compared to the use of otoliths (e.g., Lowerre-Barbieri et id.. 1994). Our results may also differ by mere chance from 1000 - 100 - 10 1 300 275 H. brasiliensis D H. balao |> 250 0) ^ 225 ; o "- 200 175 H brasiliensis -females H brasiliensis -males O H balao -females D H. balao -males 01234 Age (years) Figure 3. Number of Hcmirwnphus /j/W/mvu (balhhool and H. h,i/,i,i ( halao), collected by age class (upper panel I and the si/es ol each age class, by species and sex (lower panel; mean 95% confidence limits). 62 R. S. McBRIDE AND P. E. THURMAN ;" . .-.; j Figure 4. Hislological features of ovarian and oocyte development for Hcmirumplms hrasiliensis (bally- hoo). (A) whole cross section from a virgin, immature female; (B) partial cross section from a regressed, mature female; (C) whole cross section from a virgin, maturing female; (D) enlargement of detail of a late-stage "yolk vesicle" oocyte with cortical alveoli; (E) an even later-stage "yolk vesicle" oocyte; (F) a very early-stage "yolked" or vitellogenic oocyte; (G) a slightly more advanced vilellogenic oocyte; (H) a vitellogenic oocyte with yolk globules (probably a cell undergoing nucleus migration); (I) a vitellogenic oocyte with a migrating nucleus. MARINE HALFBEAK REPRODUCTION 63 0) CL D H balao, n-65 I H brasiliensis, n=522 0.00 0.50 1.00 1.50 2.00 2.50 3.00 Oocyte Diameter (mm) 3.50 Figure 5. Oocyte diameter frequencies for Hemimmfilws brasiliensis (ballyhoo) (A-C) and H. balao (balao) (D). in the final stages of oocyte maturation. The largest mode in each figure represents a developing batch of oocytes in final oocyte maturation, whereas the smallest mode represents a reservoir of primary growth oocytes and vitellogenic oocytes prior to nucleus migration. A third, middle mode of oocytes represents a batch of oocytes just entering final oocyte maturation. Measurements were made from (Ala mature H. brasiliensis whose most advanced oocyte stage was nucleus migration, (B) a hydrated H. brasiliensis whose most advanced oocyte stage was nucleus breakdown. (C) a spawning H. brasiliensis with ovulated eggs, (D) a hydrated H. balao whose most advanced oocyte stage was nucleus breakdown. At least 300 oocytes were measured per female. those of Berkeley and Houde; for H. brasiliensis we found only three age-3 individuals and one age-4; for H. balao we found only a single age-2 individual. Sample sizes in both E 9 iuu f 80- O 8 60- 'c 1 40 " 1 20- n 1! 1 Fork length (mm) Figure 6. Percent frequency of mature (vitellogenic) female Hcmir- tiniplni.t hrusilienxis (ballyhoo) and H. balao (bulao). Values are calculated by fork length intervals of 10 mm from fish collected during the peak spawning season (March-August), n = number of fish. studies were larger for H. brasiliensis because this species is more commonly caught and is found in a wider range of habitats than H. balao (McBride et al., 2003). At least one other hemiramphid, Reporhamphus inclainn'/iir, lives longer and grows larger (7 years, 380-mm FL; Ling, 1958). So even among hemiramphids, H. brasiliensis and H. balao are short-lived and grow to only modest lengths. Sizes of age-1 fish (mean = 248-mm FL for H. brasil- iensis and 226-mm FL for H. balao) were significantly different for the two species and were generally larger than previously reported. Berkeley and Houde (1978) reported that size at age-1 can vary between years for H. brasiliensis (1974 == 216-mm FL, 1975 = 230-mm FL) and that at age-1, H. balao was smaller (209-mm FL) than H. brasil- iensis. The sizes of both species overlapped (McBride et at.. 1996), and the mean sizes may vary from year to year naturally. Efforts to estimate the growth rates by nonlinear models failed because the short life spans of both species made it impossible to reasonably fit a growth model to the data. H. brasiliensis. however, spawns about 2 months earlier than H. balao (i.e., April versus June modes), so their growth rates may be fairly similar on a daily or monthly basis. We also found, as did Berkeley and Houde (1978), that females were longer than males at a common age. Survival rates differed between the two species. In our study, H. brasiliensis had a higher annual survival rate than H. balao did: 14.9% and 7.5%. respectively. We found the indicating the beginning of final oocyte maturation; (J) a vitellogenic oocyte with a migrating nucleus next to a fresh postovulatory follicle (POP); (K) a late-stage vitellogenic oocyte with the nucleus positioned against the chorion; (L) a fully hydrated oocyte after nucleus breakdown; (M) three hydrated oocytes. one without a follicle and adjacent to a fresh POF; and (Nl juxtaposition of a vitellogenic oocyte prior to nucleus migration, a vitellogenic oocyte undergoing nucleus migration, and a hydrated oocyte following nucleus breakdown. The major stages of oocyte development are indicated by capitalized letters: Perinucleolar (P). cortical alveolar (C) ( = yolk vesicle), vitellogenic (V) ( = yolked), hydrated (H). Other features include the chorion (ch), cytoplasm (cy), fibrils (fi), follicle (fo), lamellae (la), lumen (lu). nucleus (nu), periodic acid-Schiff reaction-positive (pas + ) melano-macrophage centers, post-ovulatory follicle (pot"), tunic (tu) (=gonad wall), yolk formation (yo). and yolk vesicles (yv). All scale bars, 0.250 mm. 64 R. S. McBRlDE AND P. E. THURMAN 100 80 60 40 c g '^ 20 20 1 10 o S. 80 60 ' 40 nd M A M J A S O N D FJ Immature Maturing I Mature I Regressed Figure 7. Seasonal reproductive cycles of (Al female Hemiramphus /'/iiw/ic/i.u.i (ballyhoo) and (B) female H. htiltio, (balao) based on histo- logical criteria for fish collected from July 1997 to October 199K. Values were calculated from mature tish sizes only (fork length S: 198 mm for H. i -H.W.V and >160 mm for H. balao). Nd = no data. was 2.2 mm. The modal size of hydrated H. balao oocytes ( 1.6 mm) was also very close to the size of the H. balao egg (1.5 mm) illustrated by Rass (1972). These independent descriptions and illustrations of mature oocytes and eggs confirm that interspecific differences in egg sizes exist for these two congeners. Sizes at maturity for both species were smaller than sizes attained by the first winter, and we agree with Berkeley and Houde ( 1978) that both species mature in their first year. In our study, the size at 50% maturity was 63.3% of the maximum body size for H. brasiliensis (313 mm FL) and 58.6% for H. balao (273 mm FL; McBride et al., 1996). To our knowledge, no study of hemiramphids has determined size at maturity with the precision that we achieved here, but other studies have reported the size of the smallest females with hydrated oocytes. Size at maturity in these other spe- cies is 56.8% maximum body length for Hyporhamphas iiiclanochir (Ling, 1958), 58.2% for Hemiramphus liinhntiix (Silva and Davies. 1988). and 58.0% for Zenarchoincnts kampcni (Coates and Van Zwieten, 1992). Size at maturity is very close to 60% maximum body size for a number of hemiramphids, and this percentage value may be useful for predicting the size at maturity of hemiramphid that have not been studied. Both H. hrasiliensis and H. halao spawn frequently, even daily, for at least a few months of the year. Although a same pattern when we calculated survival rates using data from Berkeley and Houde ( 1978), the only previous study available: estimates of H. brasiliensis survival varied be- tween the collection years ( 1974 = 23.8%. 1975 = 17.3%); these values were higher than that for H. balao (1974 = 12.4%). Even if the use of scales by Berkeley and Houde resulted in truncated ages (not identifying fish older than age-2), the number of fish aged was so large that the estimates of survival should not have been biased (Murphy, 1997). The more likely cause of the interannual variation is that survival rates may have declined since the mid-1970s or that violations of steady-state assumptions (variable recruit- ment of age-0 fish) may have biased the estimates of annual survival as calculated here. The macroscopic appearance and development of Henii- nunpluts species gonads were similar to those of Hypo- I'luunplnis (Reporhamphus) melanochir (Ling, 1958). The group-synchronous nature of oocyte development and large egg sizes have been noted for other hemiramphids (Ling, 1958; Taiwan 1967). The pattern of final oocyte maturation follows the diel cycle, and both species spawn at dusk (McBride ct al.. 2003). At sunset, the modal diameter of ovulated H. hrtmilienM'.'i eggs was 2.8 mm. Berkeley and Houde ( 1978) illustrated an H. hni.\ilicnsi.\ embryo that was about 2.5 mm in diameter and a near-hydrated oocyte that o c cr 0) C ra.w//tvi.us (ballyhoo), and //. luiliin (halao). by age class. Spawning frequency is based on the percent frequency of females with fresh postovulatory folli- cles (POFs). A value of 100% means that all females in a particular age class were spawning every day of that month. ;; = number ol tish. MARINE HALFBEAK REPRODUCTION 65 40 T Salinity, n=250 Temperature, n=252 20 2 13:00 - x Miami Harbor - Key West 10:00 M O N D Figure 9. Monthly average salinity and temperature (upper panel) measured at the sea surface at the time of fish collection. Error bars are 95% confidence limits. Seasonal change in photoperiod for Miami Harbor entrance and Key West (lower panel: data source: National Oceanic and Atmospheric Administration Tide Tables). prolonged reproductive season has been noted for other hemiramphids (Ling, 1958; Coates and Van Zwieten. 1992). this study provides the first conclusive evidence of multiple spawning within a year for any hemiramphid. Moreover, our age-specific analysis demonstrates that older fish spawn more frequently and for longer periods than age-0 fish do. In 10,000- o o 9,000 ' >s 8,000 ' ... -^ o ..'' c 7,000 - 13 o # ,-'"" CD 6,000 ' " ^ <$> -g 5,000 ' to o CQ 4,000 ' 3,000 ' ..'' o 2,000 ' o / Heniiramphus as the average number of eggs produced in the next generality b\ each ffiihiU 1 in the present generation Species A (years) 1, d, f, w. m, l,m x H. brasiliensis 1.01)1)1)1) 151 1 1 17.3 87.3 168,719 168,719 1 0.14S70 240 1264.6 126.7 303.507 45.133 2 0.02211 240 1353.2 150.4 324,757 7.181 3 0.00329 240 1493.4 187.9 358.419 1 . 1 74 4 0.00049 240 1445.1) 1 74.9 346.809 170 i 1,111, = 222,381 H. baliin 1.00000 124 2124.41 64.8 274.114 274.114 1 0.07527 240 4226.33 1 10.5 1.014,319 76,347 2 0.00567 240 2787.75 79.2 669,060 3,791 X 1,111, = 354.251 Variables are calculated by age classes in years (,v). Survival by age class (I,) was determined by the results of the Robson-Chaprnan survival estimate (see text for details). The number of days spawning by age class (d v ) was generalized from Figure 8. The batch fecundity by the average size female in each age class (f t ) was estimated from Figure 10. and the average weight of females in each age class (w,) was calculated from length-weight relationships (McBride, unpuhl.). Annual fecundity by age class (m,) is the product of the spawning frequency (d,) and batch fecundity (t,). and the expected reproductive output (I,m,) is the expected contribution of eggs produced by each age class after accounting for survival. fact, histological examination shows that some spawning by H. brasiliensis occurred year-round. Altogether, our data demonstrate that H. balao is a smaller, shorter-lived fish than H. brasiliensis. but it ma- tures at a smaller size and produces more, albeit smaller, eggs per batch. Morphological constraints of body size probably lead to this inverse relationship between size and number of eggs produced per spawning event (Elgar. 1990). and the larger batch size of H. Inilao is consistent with the life-history theory that predicts larger batch size for shorter- lived species (Stearns. 1976). H. balao is also more com- mon offshore of the reef tract, where food may be patchier than it is inshore; and various models predict that this would select for more numerous but smaller eggs ( Wootton, 1994). Ultimately, the life-history traits that typically represent trade-offs in evolutionary terms (i.e., survival and growth rates, age-specific spawning frequency, size-specific batch fecundity) are balanced so that the lifetime egg production of both Hemiramphns species is the same order of magni- tude; Table 2). Certainly the environment is shaping some elements of the reproductive traits of these Hemiraniphus species, but the influence of their phylogenetic history is clearly evident. Although Hemiramphus species are not structure-oriented, they are associated with coral reef habitat and thus are exposed to environmental cues similar to those encountered by coral reef fish. (<'.,?., warm temperature, high salinity. open coastal hydrodynamics, and tropical weather distur- bances). However, compared to most marine teleosts, and particularly other coral reef fishes. H. hrasiliensis and H. balao both have very large eggs, high spawning frequen- cies, and low batch fecundities U'.t,'., Thresher, 1984; Gross, 1987). Their reproductive style is typical of the order Be- loniformes, with its large eggs, low fecundity, multiple spawning events, and embryos that attach to floating vege- tation (Berkeley and Houde, 1978; (Toilette et ai, 1984). In fact. Hemiramphidae is a particularly interesting family to study while exploring trade-offs in reproductive traits be- cause there is remarkable variation in the egg size, fecun- dity, and reproductive mode of its species. For example, the diameters of hemiramphid eggs range from 1.3 to 3.5 mm, and reproductive modes include producing demersal eggs, buoyant pelagic eggs, or precocious young (Wourms, 1981; (Toilette et at., 1984; Meisner and Burns, 1997). Although many of the life-history traits of H. brasiliensis and H. balao are not shared by other coral reef fishes that have similar habitats, the annual fecundities of these two species might well be in line with those of other coral reef fishes. To illustrate this point, annual individual fecundities of H. brasiliensis and H. balao range from 169,000 to 1,014,000 eggs per year (Table 2), and these estimates are within the range of annual fecundity for a relatively small lutjanid, Rhomboplites aunmihcnx (140,000-3,000,000 eggs; Cuellar et rthern Anchovy, Engraulis mordax. R. Lasker. ed. NOAA Tech. Rep. NMFS 36. U.S. Dept. Commerce. Washington. DC. Hunter, J. R., N. C. H. Lo, and R. J. H. Leong. 1985. Batch fecundity in multiple spawning fishes. Pp. 67-78 in An Egg Production Method fur Estimating Spawning Biomass of Pelagic Fish: Application in the Northern Anchovy, Engraulis mordax, R. Lasker, ed. NOAA Tech. Rep. NMFS 36. U.S. Dept. Commerce, Washington, DC. Ling, J. K. 1958. The sea garfish. Reporhamphus melanochir (Cuvier & Valenciennes) (Hemiramphidae I. in South Australia: breeding, age determination, and growth rate. Atixt. ./. Mar. Freshw. Res. 9: 60-1 10. Lowerre-Barbieri, S. K., M. E. Chillenden, Jr., and C. M. Jones. 1994. A comparison of a validated otolith method to age weakfish, Cvnoscion rcgulis. with the traditional scale method. Fish. Hull. I US) 92: 555-568. McBride, R. S., L. Polisher, and B. Mahmoudi. 1996. Florida's hall- beak, Hemiramphus spp.. bait fishery. Mar. Fish. Rev. 58: 29-38. McBride, R. S., F. Stengard, and B. Mahmoudi. 2002. Maturation and diel reproductive periodicity of round scad (Carangidae: Decapterus puiictiitiix). Mar. Biol. 140: 713-722. McBride, R. S., ,). Styer, and R. Hudson. 2003. Spawning cycles and habitats for ballyhoo and balao {Hemiramphidae: Hemiramphus) in South Florida. Fixh. Bull. fUSI 101(3) (in press). Meisner, A. D., and J. R. Burns. 1997. Viviparity in the halfbeak genera Dermogenys and Nomorhamphus (Teleostei: Hemiramphidae). J. Morphol 234: 295-317. Murphy, M. D. 1997. Bias in Chapman-Robson and least-squares esti- mators of mortality rates for steady-state populations. Fish. Bull. (US) 95: 863-868. Nybakken, J. VV. 1997. Marine Biology: an Ecological Approach. 4lh ed. Addison Wesley Longman. Menlo Park. CA. 481 pp. Partridge. L., and P. H. Harvey. 1998. The ecological context of life history evolution. Science 241: 1449-1455. Quintero-Hunter, I., H. Grier, and M. Muscato. 1991. Enhancement of histological detail using metanil yellow as counterstain in periodic acid Schiff s hemotoxylin staining of glycol methacrylate tissue sec- tions. Biotech. Histochem. 66: 169-172. Rass, T. S. 1972. On the occurrence of ichthyoplankton in Cuban waters: pelagic eggs. Tr. Inst. Okeanol. Akml. Nauk. SSSR 93: 5-41. (In Russian). Roberts, C. M. 1997. Connectivity and management of Caribbean coral reefs. Science 278: 1454-1457. Robson, D. S., and D. G. Chapman. 1961. Catch curves and mortality rates. Traits. Am. Fish. Soc. 90: 181-189. SAS Institute Inc. 1990. SAS/STAT User's Guide. Vols. I and II. SAS Institute. Cary. NC. 1686 pp. Silva, E. I. L., and R. W. Davies. 1988. Notes on the biology of Hemiramphus limhatus (Hemiramphidae: Pisces) in Sri Lanka. Trap. Freshw. Biol. 1: 42-49. Stearns, S. C. 1976. Life-history tactics: a review of the ideas. Q. Rev. Biol 51: 3-47. Talwar, P. K. 1962. A contribution to the biology of the halfbeak, Hyporhamphus georgii (Cuv. & Val.) (Hemirhamphidae). Indian J. Fish. 9: 168-196. Talwar, P. K. 1967. Studies on the biology of Hemirhainphiix mar- ginatus (Forskal) (Hemirhamphidae-Pisces). J. Mar. Biol. Axxoc. India 9: 61-69. Thresher, R. E. 1984. Patterns in the reproduction of reel fishes. Pp. 343-388 in Reproduction in Reef Fishes, T.F.H. Publications, Neptune, NJ. Wallace, R.. and K. Selman. 1978. Oogenesis in Funilulux heiemclitiis. Dcv. Biol. 62: 354-369. Wootlon, R. J. 1994. Life histories as sampling devices: optimum egg size in pelagic fishes. J. Fish Biol. 45: 1067-1077. Wourms, J. P. 1981. Viviparity: the maternal-fetal relationship in fishes. Am. Zoo]. 21: 473-515. Reference: Binl. Bull. 2(14: 6S-80. (February 2003) 2003 Marine Biological Laboratory Collection and Culture Techniques for Gelatinous Zooplankton KEVIN A. RASKOFF 1 '*. FREYA A. SOMMER 2 , WILLIAM M. HAMNER 3 , AND KATRINA M. CROSS 4 Monterey Bay Aquarium Research Institute, Moss Landing, California 95039-9644: 2 Hopkins Marine Station. Pacific Grove, California 93950-3094: 3 University of California. Los Angeles, California 90095-1606: and 4 Monterey Bay At/iiariuni, Monterey. California 93940-1085 Abstract. Gelatinous zooplankton are the least under- stood of all planktonic animal groups. This is partly due to their fragility, which typically precludes the capture of intact specimens with nets or trawls. Specialized tools and techniques have been developed that allow researchers and aquarists to collect intact gelatinous animals at sea and to maintain many of these alive in the laboratory. This paper summarizes the scientific literature on the capture, collec- tion, and culture of gelatinous zooplankton and incorporates many unpublished methods developed at the Monterey Bay Aquarium in the past 15 years. Introduction Gelatinous zooplankton is a generic term for transparent and delicate planktonic animals with mesoglea-like internal (issues that aid in regulating buoyancy. These animals in- clude some radiolarians and foraminifera. as well as medu- sae, siphonophores, ctenophores, chaetognaths. pteropods, heteropods. appendicularians, salps, doliolids. and pyro- somes (e.g.. Hamner et y+ K. SK. PK Cassiopei ti Mimucliuna Anemia. Lighting 24-27 1 y+ RF Mastigias papua Anemia, Lighting 27-29 3 mo + K. PK. HP Hydrozoa Aequorea victoria Anemia. Rotifers (hydroid s). Juvenile Aurelia, l-'uit'iunu 10-15 6 mo + K. PK Eutonina indicans Artemiii. Rotifers 10-15 3 mo + K. PK Polyorchi x i>eiiicilltitn.\ Artciiihi 10-15 3 1110 + PK . K ( 'raspeda cn\iti sowerbii Wild Ires hwater plankton. Fro/en Daphnia Freshwater 27 <3 mo K. RT 1 Itlltl foil nosa Artemiii. Rotifers 24-27 6 mo + PK. K Data summarized from Sommer (IW2. 1943) for the Monterey Bay Aquarium. * K = Kreisel: PK = Pseudokreisel: SK = Stretch kreisel; RF = Reverse How; HP = Hon/onlal pseudokreisel; RT = Rectangular tank COLLECTION AND CULTURE OF GELATINOUS ZOOPLANKTON 75 Table 2 Selected culture techniques of non-cnidarian gelatinous zooplankton Species Diet Temperature (C) Lit'espan Tanks Reference Ctenophores Pleiinihrachia hachci Artemia 10-15 2 mo + K, PK Sommer, 1992 Wild-caught zooplanklon 15 nd Large jars Hirota, 1972 Pleurobrachia pileti.\ Wild-caught copepods 15 S 1110 + Modified PK Greve, 1970 Bolinopsis infundibulutn Artemia 10-15 8 mo + K. PK Sommer. 1992 Wild-caught zooplankton 16 7 mo Modified PK Greve. 1970 Mnemiopsis mccrailyi Wild-caught zooplankton 21-31 <2 mo 30 1 tanks Baker and Reeve. 1974 Beroe spp. Ctenophores, gelatin 10-15 <3 mo K. PK Sommer. 1992 Beroe gracilis Pleurobrachia pilcux 15 6 mo Modified K Greve, 1970 Beroe cuciimi\ Bolinopsis infundibulum 15 1 mo+ Modified K Greve, 1970 Molluscs Clionc limacina None 10-15 1 1110 + PK. K This study Wild-caught pteropods 12-14 4 1110 + Dishes Conover and Lalli. 1972 Cliopsis kmhni None 10-15 6 mo + PK, K This study Chaetogranths Sagitta hispida Copepods 17-31 2 mo + 30-1 tanks Reeve. 1970; Reeve and Walter. 1972 nd nd nd Modified K Greve. 1968 Pelagic Tunicates Oikoplt'iira dioicti Cultivated phytoplankton 13 8-12 days RJ Paffenhofer. 1973 Fritillaria boreal i\ Cultivated phytoplankton 12 nd RJ Paffenhofer and Hams, 1979 Thalia democratica Phytoplankton nd 8-20 days Large jars Heron, 1972 nd = no data. * K = Kreisel; PK = Pseudokreisel; RF = Reverse flow; RJ = Rotating jars. (such as Chrysaora fuscescens), since they will tend to congregate in the center of the tank, away from the walls (Tables 1 and 2). Any rectangular tank can be modified into a "pseudo- kreisel." but care must be taken to ensure that the height and width of the tank are about equal, or the water in the tank will not be able to rotate in a perfect circle and will create areas within the tank of limited flow where the animals may accumulate and contact the sides. Rectangular tanks are modified by glueing a screen across the upper corner at an angle of about 30-40 from vertical in front of the over- flow (Fig. 4). Water enters the tank through a perforated tube positioned so that the flow sweeps across the screen down towards the bottom of the tank. It is important that the tube is positioned so that the flow is parallel to the screen and covers the entire screen so that specimens are swept away rather than drawn against it. Curved plastic or vinyl inserts are glued with silicone into the bottom corners to round them into a more circular shape. Friction-fitting stiff screens can also be used to round the corners, although this option makes the tank more difficult to clean and maintain than one with solid corners. Water Several water quality issues are important for the suc- cessful culture and rearing of gelatinous organisms. Tem- perature and salinity must be kept within a range appropri- Screen \ Downwelling Inlet Outflow Inlet - \ Screen - ^Outflow Inlet Figure 3. Stretch kreisel design showing the two rotating Langmuir cells set up by the placement of the side and downwelling inlets. 76 K. A. RASKOFF ET AL. Inlet Inlet \ -Outflow Screen Outflow Plastic' or Vinyl Figure 4. Pseudokreisel design made from a standard tank. Bottom corners are filled in with silicone and solid pieces of plastic or vinyl. Outflow is separated from the tank by the inlet and screen. ate for the species being reared. The water must be relatively clean and filtered, especially if the animals are to be used for any display purpose. Small particles in the water will quickly clog the outflow screens. Filtering the water with 20-jum pleated cartridge tillers is usually sufficient; however, some cultures that are very sensitive to biological fouling (such as many hydroid species) may need additional filtration to the 3 jam level. Although air bubbles can be helpful in the culture of many small gelatinous animals by increasing water circulation, they can be detrimental to larger adult sizes (>3 cm). The bubbles can be ingested and collect in the gut and radial canals of medusae and cteno- phores, causing the animals to become positively buoyant, disrupting their normal swimming and feeding behaviors. A more serious problem is that these bubbles will slowly work themselves through the mesoglea. which can lead to infec- tion. A degassing system for the water may be needed if the incoming water tends to be supersaturated. A degassing tower in which the water trickles down through small plastic- balls or other material serves to degas the water before it enters the tank. Deep-sea animals may be sensitive to the high oxygen concentrations of surface waters. Reducing the oxygen concentration in tank water by bubbling nitrogen gas has been used in the past with some success, although it does not appear to be critical for most deep-sea species. Maintenance Throughout the course of feeding and rearing, tanks ac- cumulate debris that should be removed regularly. The use of pipettes, small brushes, basters, and siphons for removing larger debris, including waste, uneaten Anemia, and other food items, will help keep the tanks clean and discourage fouling growth. Pipettes of any size and type can be used to gently lift and collect debris. Kitchen basters work well for removing larger items because of their large reservoir vol- ume and wide bore. Siphons are best constructed from small-bore acrylic tubes with flexible plastic tubing at- tached, so that the tubing may be pinched to stop flow if an animal gets too close to the suctioning tip. Additionally, siphoning the "waste water" into a temporary container allows for the retrieval of any specimen that might inadver- tently be removed. To protect the insides of the tanks from scratches, it is helpful to dip the end of the acrylic tube into liquid plastic, available from most hardware stores; alterna- tively, a small ring of Nalgene tubing may be placed on the end of the siphon tube. Floating layers of lipid-rich mate- rials can be removed by skimming with small jars or beakers or fine-meshed nets, or by absorbing the material onto paper towels floated on the surface of the water. The sides of the tanks can be cleaned by wiping with brushes (firm paint brushes work well) or non-abrasive pads. For larger tanks (>75 1), painting or scrub pads can be covered with nonabrasive nylon mesh fabric and attached to poles for cleaning hard-to-reach areas of the tanks. The wood or metal handles of these scrubbers can be covered in plastic tubing to reduce the adherence of tentacles. Flow to the tanks can also be temporarily shut off and the animals allowed to collect on the bottom of the tank during cleaning. Also, tanks can be cleaned just after the animals have been fed, when tentacles are typically retracted and less apt to become ensnared (C. Widmer, Monterey Bay Aquarium, per. comm.). Screens in the tanks collect debris quickly and need to be scrubbed and cleaned at regular intervals. When screens become clogged, organisms are more likely to stick to them, possibly with fatal results. Even with proper cleaning and filtration, biofouling in COLLECTION AND CULTURE OF GELATINOUS ZOOPLANKTON 77 culture and rearing tanks can become a serious problem. In some cases of diatom and algae fouling, reducing the light that shines on the tank can help reduce growth, but typi- cally, scrubbing the tanks eventually becomes necessary. When diatom, hydroid, or other fouling organisms cannot be satisfactorily removed by any of the means discussed previously, bleaching is necessary. This can be especially useful on the screens, pumps, and waterlines, which can be very difficult to clean by other means. The entire tank system may need to be bleached every 1-6 months, depend- ing on the size and fouling rate. During bleaching, the occupants of the tank must be removed and transferred to a holding facility. The longer the tanks and lines are allowed to bleach, the more complete the fouling kill will be. Over- night is preferred, but bleaching for even an hour kills most fouling organisms. As a rule of thumb. 1 1 of standard 3%-6% sodium hypochlorite (NaOCl) bleach will treat about 200 1 of water ( = 1 gallon bleach/800 gallons of water), but this amount can be increased or decreased de- pending on the severity of the fouling and the time available to let the tank bleach. The water level in the tank should be dropped so that there is no overflow when the bleach is added. If the tanks have self-contained pumps, these should be run at a high flow rate to mix the bleach and flush it into the pump housings. To complete the process, the bleach must be neutralized. This can be accomplished by adding about 60 g of sodium thiosulfate (Na^S^O,) per liter of bleach used (=1 cup/ gallon). The sodium thiosulfate crystals may be dissolved in a bucket of water prior to adding to the tank. When the color of the water in the aquarium changes from yellow-green to clear, sufficient thiosulfate has been added for neutraliza- tion. Allow the thiosulfate several minutes to run through the entire tank and pumps. The treated water is then drained from the tank and discarded. While draining, thoroughly rinse out the tank with freshwater. Stubborn growth can be removed at this time by scrubbing. After all debris and treated water is removed, begin to refill the tank with seawater. minimizing turbulence and bubbles during the refilling since bubbles will stick to the walls of the tank and will have to be removed before gelatinous animals are returned. Discussion The use of the techniques described herein for the cap- ture, culture, and rearing of gelatinous zooplankton has allowed researchers to address many important biological issues. Historically, these contributions were limited pri- marily to the disciplines of systematics, developmental bi- ology, and evolution. More recently, new advances in our understanding of behavior, physiology, ecology, and ocean- ographic processes from the sea surface to the abyssal depths have also been possible. Through the use of culture methodologies, laboratory- based experimentation on salps and larvaceans has begun to address important ecological questions about the role these animals play in the nutrient cycling of the oceans and their impact on the ecosystem. These organisms have some of the fastest generation times and largest nutrient turnovers in the world, and their fecal pellets and associated "marine snow" are important sources of carbon transport into the deep sea (e.g., Alldredge, 1972; Silver et ai. 1998). Recent laboratory studies have shown that some species of medusa have chemically-regulated feeding behaviors (Arai, 1991, 1997; Tamburri et ul., 2000), with several different chemical stimuli controlling the feeding and swim- ming of both hydrozoan and scyphozoan medusae. Tank- based studies on the vertical migration of medusae (Mackie c? nl.. 1981; Mills, 1983) and on their swimming and feeding behaviors (e.g., Costello and Colin, 1995; Suchman and Sullivan, 2000) have provided much information on the physiological and behavioral components of medusa loco- motion as it relates to prey selection and capture. The interactions between gelatinous zooplankton and hu- mans are increasing, whether from envenomation (Burnett, 2001 ); blooms that clog power plant intakes (Masilamoni et ul.. 2000); interactions, both positive and negative, with fisheries (Mutlu et ul.. 1994: Mutlu. 1999; Mills. 2001; Purcell and Arai, 2001): or the general increase in gelati- nous zooplankton populations in perturbed or eutrophic environments (Mills. 1995. 2001; Arai. 2001). The oppor- tunities for scientific studies of gelatinous zooplankton are vast and largely untouched. We hope researchers can use some of the techniques presented here to expand the re- search being done on these important but poorly understood marine organisms. The public's fascination with and appreciation of gelati- nous zooplankton is growing rapidly. What were once con- sidered nasty animals that might sting or otherwise disturb beachgoers are now a major attraction in public aquaria all over the globe. The time and money spent by the aquarium industry to provide compelling exhibits on gelatinous zoo- plankton is a testament to their appeal. Over 3.4 million people visited the Monterey Bay Aquarium during the tem- porary "Planet of the Jellies" exhibit in 1992 and 1993 (Powell, 2001; J. Tomulonis, Monterey Bay Aquarium, pers comm.). Jellyfish and ctenophores were given permanent starring roles in the Outer Bay Wing, and in a new tempo- rary exhibit, "Jellies: Living Art." Aquarists in the United States and elsewhere are responsible for many of the tech- niques discussed in this paper. Aquariums around the world provide the bulk of the layperson's information on gelati- nous zooplankton. and we hope that the rising public ap- preciation of these important and beautiful animals may lead to increased financial and societal support for their continued study. 78 K. A. RASKOFF ET AL Acknowledgments F. Boero, A. Case. M. Coates, J. Connor, J. Costello, R. Hamilton, C. Harrold, G. Matsumoto. S. McDaniel, C. Priewe, K. Reisenbichler. B. Robison, R. Sherlock, J. To- mulonis, B. Upton, B. Utter. G. VanDykhuizen, C. Widmer. and D. Wrobel provided information and support for this review. J. Connor. B. Robison. G. Matsumoto, M. LaBar- bera, and two anonymous reviewers provided valuable com- ments on this manuscript. This work was supported by the David and Lucile Packard Foundation through MBARI/ MBA Joint Projects Committee. Literature Cited Abe, Y., and M. Hisada. 1969. On a new rearing method of common jellyfish, Aiirelia aiirita. Bull Mar. Biol Stn. Asamiishi 13: 205-209. Acuna, J. L., D. Deibel, and S. Sooley. 1994. A simple device to transfer large and delicate planktonic organisms. Limnol. Oceanogr. 39: 2001-2003. Alldredge, A. L. 1972. Abandoned larvacean houses: a unique food source in the pelagic environment. Science 177: 885-887. Alldredge, A. L., and W. M. Hamner. 1980. Recurring aggregation of zooplankton by a tidal current. Estncii: Coast. Mar. Sci. 10: 31-37. Arai, M. N. 1991. Attraction of Aiirelia and Aequorea to prey. H\dn>- biologia 216-217: 363-366. Arai, M. N. 1997. .4 Functional Biology of Scv/i/io:oa. Chapman and Hall. London. Arai, M. N. 2001. Pelagic coelenterates and eutrophication: a review. Hytlrohiologia 451: 69-87. Baker, A. C. 1963. The problem of keeping planktonic animals alive in the laboratory. J. Mar. Biol. Assoc. UK 43: 291-294. Baker, L. D., and M. R. Reeve.1974. Laboratory culture of the lobate ctenophore Mnemiopsis mccradyi with notes on feeding and fecundity. Mar. Biol. 26: 57-62. Brewer, R. H. 1976. Larval settling behavior in Cvanea capillata (Cni- daria: Scyphozoa). Biol. Bull. 150: 183-199. Brewer, R. H. 1978. Larval settlement behavior in the icllytish Aiirelia aiirita (Linnaeus) (Scypho/.oa: Semaeostomeae). Estuaries 1: 120- 122. Brewer, R. H. 1984. The influence of the orientation, roughness, and wettability ot solid surfaces on the behavior and attachment of planulae of Cyanea (Cnidaria: Scyphozoa). Biol. Bull. 166: 11-21. Brewer, R. H., and J. S. Feingold. 1991. The effect of temperature on the benthic stages of Cyanea (Cnidaria: Scyphozoa), and their seasonal distribution in the Niantic River estuary. Connecticut. J. Exr>. Mar. Biol. Ecul. 152: 49-6(1. Bullock, T. H. 1943. Neuromuscular facilitation in scyphomedusae. ./. Cell. Com/1. Phvsiol. 22: 251-272. Burnett, J. W. 2001. \K-dical aspects of jellyfish envenomation: patho- genesis. case reporting and therapy. Hvdrobiologia 451: 1-9. Calder, 1). R. 1974. Strohil.Mion of the sea nettle, Clnysaora aiiim/ue- cirrha, under field conditions. Hi,*/. Bull. 146: 326-334. Cargo, D. G. 1975. Comments on the laboratory culture of Scyphozoa. Pp. 145-154 in Culture oj Marine invertebrate Animal-,: Proceedings, W. L. Smith and M. H. Chanley, eds. Plenum Press, New York. Childress, J. J.. and K. V. Thuesen. 1993. F: fleets of hydrostatic pres- sure on metabolic rates of six species of deep-sea gelatinous zooplank- ton. Limnol. Oceanogr. 38: 665-670. Childress, .(. J., A. T. Barnes. L. B. Quetin, and B. H. Rohison. 1978. Thermally protecting cod ends tor the recovery of living deep-sea animals. Deep-Sea Res. 25: 419-422 Conover, R. J., and C. M. Lalli. 1972. Feeding and growth in Clione limacina (Phipps), a pteropod mollusc. J. Ev/>. Mar. Biol. Ecol. 9: 279-302. Costello, J. H., and S. P. Colin. 1995. Flow and feeding by swimming scyphomedusae. Mar. Biol. 124: 399-406. Custance, D. R. N. 1964. Light as an inhibitor of strobilation in Aiirelia aiirita. Nature 204: 1219-1220. Dawson, M. N. 2000. Variegated mesocosms as alternatives to shore- based planktonkreisels: notes on the husbandry of jellyfish from marine lakes. J. Plankton Res. 22: 1673-1682. Gibson, D. M., and G. A. Paf'fenhiif'er. 2000. Feeding and growth rates of the doliolid, Dolioletta gegenbauri Uljanin (Tunicata, Thaliaceu). J. Plankton Res. 22: 1485-1500. Gilmer, R. W. 1972. Free-floating mucus webs: a novel feeding adap- tation for the open ocean. Science 176: 1239-1240. Greve, W. 1968. The "planktonkreisel". a new device for cultunng zooplankton. Mar. Biol. 1: 201-203. Greve, W. 1970. Cultivation experiments on North Sea ctenophores. Helgol. Meeresunters. 20: 304-317. Greve, W. 1975. The "Meteor Planktonktivette": a device for the main- tenance of macrozooplankton aboard ships. Aaiiacnltnre 6: 77-82. Groat, C. S., C. R. Thomas, and K. Schurr. 1980. Improved culture of Aiirelia aiirita scyphistomae for bioassay and research. Ohio J. Sci. 80: 83-87. Haddoek, S. H. D., T. J. Rivers, and B. H. Robison. 2001. Can coelenterates make coelenterazine'' Dietary requirements for luciferin in cnidarian bioluminescence. Proc. Nat/. Acad. Sci. USA 98: I 1 148- I I 151. Haeekel, E. 1893. Planktonic studies: a comparative investigation of the importance and constitution of the pelagic fauna and flora (Eng. transl.) Appendix 6. Re. Comm. for 1889-1891. U.S. Connn. Fish, and Fish- eries. Hamner, W. M. 1975. Underwater observations of blue-water plankton: logistics, techniques, and safety procedures for divers at sea. Limnol. Oceanogr. 13: 1045-1051. Hamner, W. M. 1985. The importance of ethology for investigations of marine /ooplankton. Bull. Mar. Sci. 37: 414-424. Hamner, W. M. 1990. Design developments in the planktonkreisel, a plankton aquarium for ships at sea. J. Plankton Res. 12: 397-402. Hamner, W. M., and D. Schneider. 1986. Regularly spaced rows of medusae in the Bering Sea: Role of Langmuir circulation. Limnol. Oceanogr. 31: 171-177. Hamner, W. M., L. P. Madin. A. L. Alldredge, R. W. Gilmer, and P. P. Hamner. 1975. Underwater observations of gelatinous zooplankton: sampling problems, feeding biology, and behavior. Limnol. Oceanogr. 20: 407-917. Harbison, G. R., G. I. Matsumoto, and B. H. Robison. 2001. Lam- pocteis crnentiventer gen. nov.. sp. nov.: a new mesopelagic lobate ctenophore. representing the type of a new family (Class Tentaculata, Order Lobata. Family Lampoctenidae. fani. nov.). Bull. Mar. Sci. 68: 299-31 I. Harris. R. P.. P. H. Wiebe. J. Lenz. H. R. Skjoldal, and M. Huntley, eds. 2000. ICES Zooplankton Methodology Manual. Academic Press. San Diego. CA. Heine, J. ed. 1986. Blue Water Diving Guidelines. California Sea Grant College Publ. T-CSGCP-014, La Jolla, CA. Helland, S., G. V. Triantaphyllidis, H. J. Fyhn, M. S. Evjen, P. Lavens, and P. Sorgeloos. 2000. Modulation of the free amino acid pool and protein content in populations of the brine shrimp Anemia spp. Mar. Biol. 137: 10(15-101(1. Hemleben, C., and H. Kitazato. 1995. Deep-sea foraminifcra under long time observation in the laboratory. Deep-Sea Res. 42: 827-832. Hensen, V. 1887. Uber die Bestimmung des Planktons oder des mi Meere treibenden Materials an Pflanzen und Thieren. Konn. ItV.w Untersitch. Deutsch. Meere 5: 1-107. COLLECTION AND CULTURE OF GELATINOUS ZOOPLANKTON 79 Heron, A. C. 1972. Population ecology of a colonizing species: the pelagic tunicate Thalia democratica. I. Individual growth rate and generation time. Oecologia 10: 269-293. Hirota, J. 1972. Laboratory culture and metabolism of the planktonic ctenophore, Pleurobrachia hachei A. Agassiz. Pp. 465-484 in Biolog- ical Oceanography of the Northern North Pacific Ocean, A. Y. Tak- enouti, ed. Idemitsu Shoten. Tokyo. Keen, S. L. 1987. Recruitment of Aureliu aurita (Cnidaria: Scyphozoa) larvae is position-dependent, and independent of conspecific density, within a settling surface. Mar. Ecol. Prog. Ser. 38: 151-160. Kramp, P. L. 1965. The hydromedusae of the Pacific and Indian Oceans. Dana-Rep. Car/sberg Found. 63: 1-162. Larson, R. 1991. Why jellyfish stick together Caribbean thimble jellies travel in the same circles. Nat. Hist. 66-71. Larson, R. J., G. I. Matsumoto, L. P. Madin, and L. M. Lewis. 1992. Deep-sea benthic and benthopelagic medusae: recent observations from submersibles and a remotely operated vehicle. Bull. Mar. Sci. 51: 277-286. Mackie, G. O. 1960. The structure of the nervous system in Vclella. Q. J. Microsc. Sci. 101: 119-131. Mackie, G. O., R. J. Larson, K. S. Larson, and L. M. Passano. 1981. Swimming and vertical migration of Aurelia aurita (L) in a deep tank. Mar. Behav. Physiol. 7: 321-329. Madin. L. P. 1974. Field observations on the feeding behavior of salps (Tunicata: Thaliacea). Mar. Biol. 25: 143-147. Martin, V. J., and R. Koss. 2002. Phylum Cnidaria. Pp. 51-108 in Atlas of Marine Invertebrate Larvae, C. M. Young, ed. Academic Press, San Diego, CA. Marlindale, M. Q. 1987. Larval reproduction in the ctenophore Mnemi- opsis mccnidyi (order Lobata). Mar. Biol. 94: 409-414. Masilamoni, J. G., K. S. Jesundoss, K. Nandakuniar. K. K. Satpathy. K. V. K. Nair, and J. Azariah. 2000. Jellyfish ingress: a threat to the smooth operation of coastal power plants. Curr. Sci. (Bangalore) 79: 567-569. Matsumoto, G. I., and G. R. Harbison. 1993. In situ observations of foraging, feeding, and escape behavior in three orders of oceanic ctenophores: Lobata, Cestida. and Beroida. Mar. Biol. 117: 279-287. Matsumoto, G. I., and B. H. Robison. 1992. Kiyohimea usagi, a new species of lobate ctenophore from the Monterey Submarine Canyon. Bull. Mar. Sci. 51: 19-29. Mayer, A. G. 1910. Medusae of the World. Vol. I: The Hydromedusae. Carnegie Institution of Washington, Washington, DC. Miglietta, M. P., L. Delia Tommasa, F. Denitto, C. Gravili, P. Pagliara, J. Bouillon, and F. Boero. 2000. Approaches to the ethology of hydroids and medusae (Cnidaria. Hydrozoa). Sci. Mar. 64: 63-71. Mills, C. E. 1983. Vertical migration and diel activity patterns of hy- dromedusae: studies in a large tank. J. Plankton Res. 5: 619-635. Mills, C. E. 1995. Medusae, siphonophores. and ctenophores as plank- tivorous predators in changing global ecosystems. ICES J. Mar. Sci. 52: 575-581. Mills, C. E. 2001. Jellyfish blooms: Are populations increasing globally in response to changing ocean conditions? H\drohiologia 451: 55-68. Mills, C. E., and M. F. Strathmann. 1987. Reproduction and develop- ment of marine invertebrates of the northern Pacific coast: data and methods for the study of eggs, embryos, and larvae. Pp. 44-71 in Reproduction and Development of Marine Invertebrates of the North- ern Pacific Coast: Data and Methods for the Stud\ of Eggs, Embi-yos. and Lan i ae, University of Washington Press. Seattle. Mills, C. E., F. Boero, A. Migotto, and J. M. Gili, eds. 2000. Trends in H\Jro~oan Biology-IV, Sci. Mar. 64 (Supl. 1). Mutlu, E. 1999. Distribution and abundance of ctenophores and their zooplankton food in the Black Sea. II. Mnemiopsi.t leidvi. Mar. Biol. 135: 603-613. Mutlu. E., F. Bingel, A. C. Gucu, V. V. Melnikov, U. Niermann, N. A. Ostr, and V. E. Zaika. 1994. Distribution of the new invader Mne- iiiii'/i.w'.v sp. and the resident Aurelia aurita and Pleurobrachia pileus populations in the Black Sea in the years 1991-1993. ICES J. Mar. Sci. 51: 407-421. Olmon, J. E., and K. L. Webb. 1974. Metabolism of "M in relation to strobilation of Aurelia aurita L. (Scyphozoa). J. Exp. Mar. Biol. Ecol. 16: 113-122. Orlov, D. 1996. Observations on the settling behaviour of planulae of Clava multicomis Forskal (Hydroidea, Athecata). Sci'. Mar. 60: 121- I2S. Paffenhtifer, G.-A. 1970. Cultivation of Calanus he/go/andicus under controlled conditions. He/gol. Meeresunters 20: 346-359. Paffenhiifer. G.-A. 1973. The cultivation of an appendicularian through numerous generations. Mar. Biol. 22: 183-185. Paft'enhofer, G.-A., and R. P. Harris. 1979. Laboratory culture of marine holozooplankton and its contribution to studies of marine plank- tonic food webs. Adv. Mar. Biol. 16: 21 1-308. Pagliara, P., J. Bouillon, and F. Boero. 2000. Photosynthetic planulae and planktonic hydroids: contrasting strategies of propagule survival. Sci. Mar. 64: 173-178. Pantin, C.F.A. 1952. The elementary nervous system. Proc. R. Soc. Loud. Ser. B 140: 147-168. Powell, D. C. 2001. A Fascination for Fish: Adventures of an Under- water Pioneer. UC Press/Monterey Bay Aquarium. Berkeley. Purcell, J. E. 1997. Pelagic cnidarians and ctenophores as predators: selective predation. feeding rates, and effects on prey populations. Ann. lust. Oceanogr. 73: 125-137. Purcell, J. E., and M. N. Arai. 2001. Interactions of pelagic cnidarians and ctenophores with fish: a review. Hydrobio/ogia 451: 27 H. Purcell, J. E., D. L. Breitburg, M. B. Decker, W. M. Graham, M. J. Youngbluth, and K. A. Raskoff. 2001. Pelagic cnidarians and ctenophores in low dissolved oxygen environments. Pp. 77-100 in Coastal Hypoxia: Consequences for Living Resources and Ecosystems. N. N. Rabalais and R. E. Turner, eds. American Geophysical Union, Washington, DC. Ragulin. A. G. 1969. Underwater observations on krill. Tr. Vses. Nauclmo-Iss/ed. Inst. Morsk. Ryhn. Kho:, Okeanogr. 66: 231-234 (Engl. transl.). Raskoff, K. A. 2001. The impact of El Nino events on populations of mesopelagic hydromedusae. Hydrobiologia 451: 121-129. Raskoff, K. A. 2002. Foraging, prey capture, and gut contents of the mesopelagic narcomedusa. Solmissus (Cnidaria, Hydrozoa). Mar. Biol. 141: 1088-1107. Rees, J. T. 1979. Laboratory and field studies on Eutonina indicant (Coelenterata: Hydrozoa). a common leptomedusa of Bodega Bay. California. Wti.inuinn J. Biol. 36: 201-209. Rees, W. J., and F. S. Russell. 1937. On rearing the hydroids of certain medusae, with an account of the methods used. J. Mar. Biol. Assoc. UK 22: 61-82. Reeve, M. R. 1970. Complete cycle of development of a pelagic chaetognath in culture. Nature 227: 381. Reeve, M. R. 1981. Large cod-end reservoirs as an aid to the live collection of delicate zooplankton. Limnol. Oceanogr. 26: 577-580. Reeve. M. R., and M. A. Walter. 1972. Conditions of culture, food-size selection, and the effects of temperature and salinity on growth rate and generation time in Sagitta hispida Conant. J. E.\p. Mar. Biol. Ecol. 9: 191-200. Robison, B. H. 1993. Midwater research methods with MBARI's ROV. Mar. Technol. Soc. J. 26: 32-39. Robison. B. H., K. R. Reisenbichler, R. E. Sherlock, J. M. B. Silguero, and F. P. Chavez. 1998. Seasonal abundance of the siphonophore. Nanomia bijuga. in Monterey Bay. Deep-Sea Res. II 45: 1741-1751. Russell, F. S. 1953. The Medusae of the British Isles: Anthomednsae. 80 K. A. RASKOFF ET AL. Leptomedusae, Limnomedusae, Trachymedusae and Narcomedusae. The University Press, Cambridge. Sameoto, D., P. H. Wiebe, J. Runge. L. R. Postel. J. Dunn, C. Miller, and S. Coombs. 2000. Collecting zooplankton. Pp. 55-81 in ICES Zooplankton Methodology Manual. R. P. Harris, P. H. Wiebe, J. Lenz, H. R. Skjoldal. and M. Huntley, eds. Academic Press, San Diego, CA. Sato, R., J. Yu, Y. Tanaka. and T. Ishimaru. 1999. New apparatuses for cultivation of appendicularians. Plank. Biol. Ecol. 46: 162-164. Sato, R., Y. Tanaka. and T. Ishimaru. 2001. House production by Oikoplenra dioica (Tunicata, Appendicularia) under laboratory condi- tions. ./. Plankton Res. 23: 415-423. Schmahl, G. 1985. Bacterially induced stolon settlement in the scy- phopolyp of Aurelia aitrita (Cnidaria, Scyphozoa). Helgol. Meere- suntcrs. 39: 33-42. Siefker. B., M. Kroiher, and S. Berking. 2000. Induction of metamor- phosis from the larval to the polyp stage is similar in Hydrozoa and a subgroup of Scyphozoa (Cnidaria. Semaeostomeae). Helgoland Mar. Res. 54: 230-236. Silver, M. W.. S. L. Coale, C. H. Pilskaln, and D. R. Steinberg. 1998. Giant aggregates: importance as microbial centers and agents of ma- terial flux in the mesopelagic zone. Limnol. Oceanogr. 43: 498-507. Sommer, F. A. 1992. Husbandry aspects of a jellyfish exhibit at the Monterey Bay Aquarium. Pp. 362-369 in American Associations of Zoological Parks ami Ai/iiuriiuns Annual Conference Proi -ceding. Toronto. Sommer, F. A. 1993. Jellyfish and beyond: husbandry of gelatinous zooplankton at the Monterey Bay Aquarium. Pp. 249-261 in Proceed- ings of the Third International Ai/iianiini Congress. Boston, MA. Spangenberg, D. B. 1965. Cultivation of the life stages of Aurelia aurita under controlled conditions. J. Ev/>. Zoo/. 159: 303-309. Spangenherg, D. B. 1971. Thyroxine induced metamorphosis in Aure- lia. J. E\p. Zoo/. 178: 183-194. Strathmann. M. F. 1987. Reproduction and Development of Marine Invertebrates of the Northern Pacific Coast: Data anil Methods for the Study of Eggs, Emhryus. and Lamie. University of Washington Press. Seattle. Suchman, C. L., and B. K. Sullivan. 2000. Effect of prey size on vulnerability of copepods to predation by the scyphomedusae Aurelia aurita and Cyanea sp. J. Plankton Res. 22: 2289-2306. Sugiyama, K., and O. R. Anderson. 1997. Experimental and observa- tional studies of radiolarian physiological ecology. 6. Effects of sili- cate-supplemented seawater on the longevity and weight gain of spon- giose radiolarians. Spongaster telras and Dictyocoryne trimcatuin. Mar. Micropalcontol. 29: 159-172. Tamburri, M. N., M. N. Halt, and B. H. Robison. 2000. Chemically regulated feeding by a midwater medusa. Limnol. Oceanogr. 45: 1661- 1666. Thuesen, E. V., and J. J. Childress. 1994. Oxygen consumption rates and metabolic enzyme activities of oceanic California medusae in relation to body si/.e and habitat depth. Biol. Bull. 187: 84-98. Totton. A. K. 1965. ,4 S\nopsis of the Siphonophora. Br. Mus. (Nat. Hist.), London. Utter, B. D. 2001. Culturing the umbrella jelly, Eutonina iiulican.i (Ro- manes, 1876). Drum and Croaker 32: 38-44. Ward. W. W. 1974. Aquarium systems for the maintenance of cteno- phore and jellyfish and for hatching and harvesting of brine shrimp (Anemia salina) lanae. Chesapeake Sci. 15: 116-118. Youngbluth. M. .1. 1984. Water column ecology: in situ observations of marine zooplankton from a manned submersible. Pp. 45-57 in Divers. Submersible! and Marine Science. N. C. Fleming, ed. Memorial Uni- versity of Newfoundland, Occasional Papers in Biology. Vol. 9. Reference: Bioi Bull. 204: Sl-95. (February 2003) 2003 Manne Biological Laboratory Branchial Musculature of a Venerid Clam: Pharmacology, Distribution, and Innervation LOUIS F. GAINEY. JR. 1 *. JAMES C. WALTON 1 , AND MICHAEL J. GREENBERG 2 1 Department of Biological Sciences, University of Southern Maine, Portland, Maine 04104: and ' The Whitne\ Laboratory' of the University of Florida, 9505 Ocean Shore Blvd.. St. Augustine, Florida 32086 Abstract. This study was meant to analyze the neural control of the branchial muscles of the clam Mercenaria mercenaria. Gills isolated from the animal contract in re- sponse to 5-hydroxytryptamine (5HT). dopamine (DA), and acetylcholine (ACh); but the ACh contraction occurred only it" the gills had been pretreated with the cholinesterase inhibitor eserine. The 5HT antagonists cyproheptadine and mianserin blocked the contractile effects of all of the ago- nists. However, gills exposed to the 5HT antagonists and eserine relaxed in response to ACh. The DA antagonist SCH-83566 inhibited the effects of DA. but had no effect on contractions induced by 5HT and ACh. The ACh antagonist hexamethonium inhibited both the excitatory and inhibitory effects of ACh. but had no effect on contractions induced by 5HT and DA. 5HT and DA in gill tissue were visualized by using immunohistochemistry. Within each gill filament are dorsoventral neurons running adjacent to the epithelium and containing immunoreactive 5HT and DA. A complex net- work of 5HT-positive fibers is associated with the septa, blood vessels, and muscles, whereas DA-positive fibers are restricted to the septa. We propose that 5HT is the excitatory transmitter to the gill muscles, and that DA and ACh exert their excitatory effects by stimulating 5HT motor nerves. ACh may also be an inhibitory transmitter of the muscles. Introduction In most clams, the water current that supports respiration and feeding is driven through the gills by the beat of the lateral cilia. But the diameter and shape of the passages Received 8 July 2002; accepted 12 December 2002. * To whom correspondence should be addressed. Department of Bio- logical Sciences. University of Southern Maine. PO Box 9300. Portland. ME 04104-9300. E-mail: gainey@usm.maine.edti through the gill and thus the flow of water are controlled by contractions of the branchial musculature. These two fundamental activities of the branchial pump ciliary and muscular are regulated and coordinated by transmitters and modulators that are released at synapses by neurons that constitute an extensive network in the gill. The neural control of bivalve gill cilia has been extensively studied: this paper focuses on the musculature. The gills of the venerid clam Mercenaria mercenaria are eulamellibranch and plicate (Kellogg. 1892). That is. the filaments are connected to adjacent filaments via tissue junctions, and the descending and ascending lamellae are connected to each other and thrown into a series of folds (the plicae) by interlamellar septa (Fig. 1 ). The dorsoventral spaces within the gill, defined by adjacent septa and the intervening plicae, are the water tubes. The plicae exist in two conformations (Fig. 1C): either their contours are smooth the "primary folds" of Kellogg, or the "major plicae" of Eble (2001) or smaller depressions appear at the apexes of the plicae, giving rise to "secondary folds" (Kellogg), or "minor plicae" (Eble). We have seen the plicae alternate between these two conformations. Dorso- ventral blood vessels lie at the apex of each plica, within each septum, and within each filament (Kellogg, 1892; Eble, 2001). The blood channels of the branchial filaments are connected with the dorsoventral and septal blood vessels by a meshwork of horizontal blood vessels that are actually interlamellar abfrontal extensions of the filaments (see fig. 34 of the venerid Tapes unreus in Ridewood, 1903; fig. 4.20 in Eble, 2001; fig. 1 in Medler and Silverman, 2001). The horiz.ontal meshwork of vessels (collectively called the "subfilamentar tissue"; see Ridewood, 1903) lines the water tubes (Fig. ID). That bivalve "ills contain muscle fibers and are capable si 82 L. F. GAINEY ET AL cerebral ganglion visceral ganglion siphon pedal ganglion descending lamella ascending lamella foot exhalent (hyperbranchial) chamber visceral mass outer demibranch inner demibranch shell mantle cavity -gill mantle I) 5HTn water flow water flow gf Figure 1. Diagramatic anatomy of Mercenaria mercenaria: adapted from various sources on the basis of our own observations. (A) A clam on the halt shell. (B) Cross section of a clam. (C) Cross section of a water tube: with the musculature relaxed (left), and contracted (right). The water tube muscles are within the walls of the horizontal blood vessels; the vessels are not shown here. (D) Details of a relaxed demibranch (as in C, left). This cross section is slightly out of the horizontal plane; thus, the filaments to the left of the dashed line are at the level of the intern lamentar tissue junctions that contain the longitudinal muscles; whereas the filaments to the right of the line are at the level of the ostia and horizontal blood vessels. The walls of the horizontal blood vessels contain both the water-tube muscles and a dense network of serotonergic nerves; neither of these is shown. Abbreviations: bv = blood vessel; gf = gill filament; hbv = horizontal blood vessel; 1m = longitudinal muscle; o = ostium; s = septum; wt = water tube; wtm = water tube muscle; 5HTn = serotonergic neuron. of muscular activity has been known for over a hundred years since Kellogg (1892) published his observations on branchial anatomy and movement in a variety of bivalves, including Merci'miriu. Longitudinal muscle fibers have been described in the intertilament tissue junctions and septa of Mercenaria (Kellogg, 1892; Ridewood, 1903; Eble, 2001) and many other bivalve gills. In addition to these longitudinal muscles, called "horizontal muscles" by Atkins (1943) and others, Medler and Silverman (1997) noted the presence of a diffuse network of muscle fibers in the water- tube epithelium of the non-plicate gills of Dreissena poly- inorpha. The plicate gills of Mercenaria lack a water-tube epithelium, but they contain a similar network of muscle fibers in the walls of the horizontal blood vessels (Fig. 1 ). Neural elements occur within the filaments of both fili- branch and eulamellibranch gills (e.g., Setna, 1930; Aiello. 1990), but they also occur in the gills of eulamellibranchs (like Mercenaria) in association with the septa, blood ves- sels, and interfi lamentar muscles; structures that, by defini- tion, do not occur in tilibranch gills. Indeed, neurons have been reported in association with the longitudinal muscles in a unionid mussel, Ligiuniu subrostrata (Dietz et at.. 1985), in Merccmiria (Gainey et ai, 1999a), and in an oyster, Crassostrea virginica (Nelson, 1960). Nerves have GILL MUSCLE PHARMACOLOGY AND ANATOMY 83 also been observed in the interlamellar septa in Solen mar- ginatus and Ensis siliqua (Atkins. 1937), Mercenaria (Gainey et at., 1999a). and Crassostrea (Nelson. 1960, Galtsoff, 1964); and in the water-tube muscles and ostia of Mercenaria (Candelario-Martinez et ai, 1993). An extensive literature indicates that the flow of water through bivalve gills varies continuously within wide limits in response to both physical and biological factors (summa- rized by Dame. 1996; Jorgensen. 1996; Bayne, 1998). But the lateral cilia in both Mercenaria and Mytilus edulis beat only within a relatively narrow range of frequencies (about 10-25 beats/s) (Aiello, 1960; Catapane, 1983; Gainey etui.. 1999a), so the stimulatory and inhibitory motor nerves to the cilia seem to be activating a simple on-off switch. Medler and Silverman (2001 ) found, in Mercemiria. that the geometry of the water tubes changed, and their diameters decreased, in response to 5-hydroxytryptamine (5HT; sero- tonin). Such changes would tend to modify flow (Grunbaum et a!.. 1998). so changes in the tone of the branchial mus- culature might well be participating in the continuously variable regulation of water flow through the gill. Although the branchial muscles have the potential to modulate water flow through the gills, and neural elements are clearly present, the pharmacology and neural control of these muscles has received relatively little attention. In brief, acetylcholine (ACh) contracts the gill muscles in both Dreissemi polymorpha and Corbicula fluminea (Snow et ai. 1995; Medler and Silverman. 1997), whereas 5HT relaxes the gill muscles of Lignmia subrostrata (Gardiner et ai. 1991) and contracts those of Mercenaria (Gainey et ai, 1998; Medler and Silverman. 2001 ). In addition, the peptide FMRFamide contracts the gill muscles of Dreissemi (Medler and Silverman. 1997). The relationships between the effects of possible neurotransmitters on gill muscles, the distribution of these agents in identifiable neural networks, and the interactions among the elements of the networks are at present unexplored. We have been using the gill of the quahog Mercenaria mercenaria to study the neural control of branchial water flow. In a previous study, we found that 5HT and dopamine (DA), respectively, switch the activity of the lateral cilia on and off, and that YFAFPRQamide, an SCP-like peptide endogenous to Mercenaria. modulates the effects of DA (Gainey et al., 1999a). Now we report on the pharmacology of the branchial muscles, focusing especially on the actions of 5HT. DA, ACh. and their antagonists. We have also investigated the distribution of the branchial muscles and their innervation by immunoreactive serotonergic and do- paminergic nerves, expecting the findings to be consistent with our pharmacological observations. Preliminary results of this study have been presented to the Society for Inte- grative and Comparative Biology (Gainey et ai. 1998. 2001). Materials and Methods Animals Quahogs (Mercenaria mercenaria L.) that had been dug from various locales along the northeast Atlantic coast were purchased from Harbor Fish, Portland, Maine. The animals were held at 10 C in natural seawater (30 ppt) on a 12-h light/dark cycle. Individuals were held a minimum of 3 days before use. G/7/ preparation and apparatus Gills were dissected away from the body wall and sepa- rated into demibranchs, and the branchial nerves removed (Fig. 1 ). Muscular contractions were recorded as changes in the length of the anterior-posterior axis of the isolated demibranchs. Contractions of the branchial muscles were recorded in either of two ways: ( 1 ) Isolated demibranchs were sus- pended in organ baths and attached with thread to isometric force transducers (Grass FT03 and UFI 1030) equipped with springs; the resulting contractions were therefore semi-iso- tonic. The transducers were interfaced to Biopac DA 100 amplifiers and a Biopac MP100 analog-to-digital converter. (2) Ultratrasonic crystal transceivers (Sonometrics) were tied to the ends of demibranchs with thread. One end of the demibranch was pinned to a piece of rubber band that was glued with rubber cement to the bottom of a plastic petri dish (4.7-cm diameter); the petri dishes were placed on a cooling plate to maintain temperature. Under these condi- tions, the muscles were unrestrained and contracted against virtually no external load. The isotonic contractions were measured with a digital ultrasonic measurement system (Sonometrics TRX series 8). In both cases, the magnitude of the contractions was measured with AcqKnowledge version 3.5 (Biopac Systems). All experiments were earned out at 10 C in aerated artificial seawater (ASW; recipe in Welsh et ai. 1968). To retard the oxidation of dopamine (DA), the water was buff- ered with an ascorbic acid buffer as described by Malanga ( 1975); this buffered seawater was used in all of the exper- iments. Production and analyses of dose-response cnn'es Our initial experiments were performed with force trans- ducers; but prolonged contraction against the load of the springs used with these devices caused the gill muscles to fatigue. Consequently, we exposed each demibranch only once to a single concentration of agonist, and the dose- response curve was constructed from these individual re- sponses. In later experiments with the Sonometrics digital ultrasound measurement system, no external force was ap- plied to the muscles. No evidence of fatigue was observed. 84 L. F. GA1NEY ET AL. so a single demibranch could be used to construct an entire dose-response curve. Because the response to serotonin (5HT) and DA has a seasonal component (Gainey, pers. obs.). the dose-response data reported here were collected between November and July. All contractions and relaxations, measured in millimeters, were expressed as a percentage of the initial length of each demibranch. Regression lines were fitted with a logistic function of the form: response = n/1 + exp(/3 () + j3,* log( agonist)), where a is the asymptotic value of the max- imal contraction, and j3 and /3, are intercept and slope parameters. Initially, all three parameters were estimated using nonlinear regression (Systat, v 9); later, a was fixed in the model, reducing the error estimates of the remaining parameters. The concentrations of agonist giving half-max- imal responses (EC MI ) were estimated according to the following formula: EC ?0 = 1(> A ( - /3,//3, ). Effects of antagonists Each of the tour demibranchs from the same clam were suspended in an organ bath and attached to a force trans- ducer. After 15 min of relaxation, each of the demibranchs was exposed to an agonist at a standard concentration: 5HT = 2 X 1(T 5 M; DA and acetylcholine (ACh) = 5 X 1CF 5 M. After the resulting contractions had stabilized, the baths were flushed, and an antagonist at 10~ 4 M was added to three of the four demibranchs. After 60 min. the standard dose of agonist was reapplied to all four demibranchs. with the antagonist still present on the three demibranchs. The total number of demibranchs treated with a specific antag- onist is given in the data tables. The effect of the antagonist was expressed as the ratio between the second and first agonist-induced contraction (contraction ratio). Analysis of the contraction ratios of untreated controls with a Kolmogorov-Smirnov one-sample test revealed that these data were not normally distributed (P < 0.001, two-tailed, n = 139). The contraction ratios were therefore normalized by a logarithmic transformation, and the normality of this transformation was checked as above (Kolmogorov-Smirnov: P 0.614). The In trans- formed ratios of the controls were tested against a mean of (since In 1 =0) with a one-sample / test. This is mathe- matically equivalent to a paired / test because the contrac- tions used to construct the ratios were from the same demi- branch. Since the contraction ratios of the controls for 5HT, DA, and ACh were all significantly greater than 1, the normal- ized contraction ratios of the antagonists were compared to the normali/cd contraction ratios of the appropriate agonist control using post hoc paired Tukey HSD tests after an initial one-way ANOVA. But some of the antagonist con- traction ratios were 0. thus these ratios become undefined by a logarithmic transformation. To overcome this limitation. 0. 1 was added to all contraction ratios prior to the logarith- mic transformation. Although the statistical tests were per- formed on the In-transformed data, tabular data are pre- sented in the Results section untransformed for clarity. The P values reported for these tests are one-tailed probabilities: P values less than 0.05 were considered significant. In some of the experiments e.g., ACh after exposure to cyprohep- tadine or mianserin the gills relaxed rather than con- tracted; these data are coded in the tables as negative values. The concentration of antagonist that produced 50% inhi- bition (IC 5() ) was calculated using the experimental design described above, except that the demibranchs were exposed to lower concentrations of antagonists. Contraction ratios i.e.. the ratios of the second to the first contractions were regressed against the log of the concentration of antagonist. Because the contraction ratios were significantly greater than 1 for all of the controls, the IC ?(I was calculated by solving the regression equation for a contraction ratio that was 50% of the mean contraction ratio of the control. Branchial anatomy For relaxed specimens, isolated demibranchs were kept overnight, at 5 C, in isotonic MgCI 2 in ASW (7.6% MgCl : in distilled water added to an equal volume of ASW). For contracted specimens, the isolated demibranchs were placed in [Q M 5HT immediately after dissection. To observe the inner face of the water tubes, we cut dorsoventrally along several septa with fine scissors, separating a section of the demibranch into two layers. One of these was removed, and the remainder of the demibranch was then pinned to the bottom of a small petri dish, which had been coated with Sylgard. Fixation always carried out at 5 C varied with the object to be observed (e.g., muscle. 5HT. DA) and is described below. Because mammalian antibodies were used for the immunohistochemistry, subsequent rinses and solu- tions were made with mammalian phosphate-buffered saline (PBS). Crysostat sections were prepared as follows. After fixa- tion and a 15-min rinse in PBS (0.1 M sodium phosphate, 140 mM NaCl; pH 7.3). the demibranchs were placed in a solution of 30% sucrose/PBS overnight at 5 C. Pieces of demibranch were then placed in Tissue Tek OCT com- pound, frozen, and sectioned at 12 /urn. Sections were placed on gel-coated slides and stored at 20 C until used. Thick sections were prepared as follows. After fixation and three 15-min rinses in PBS. pieces of demibranch were placed in a plastic mold and covered with 12% Type A pigskin gelatin in 0. 1 M PBS that had been heated to 50 C. After the gelatin had cooled, the tissue was sectioned at 100 /j.m with a vibratome. The sections were heated briefly at 50 C on gel-coated slides to melt the excess gelatin. GILL MUSCLE PHARMACOLOGY AND ANATOMY 85 We usually processed whole mounts and sections simul- taneously and therefore followed a schedule designed for whole mounts. All of the steps in this protocol were carried out at 5 C. After fixation, rinse four times (1 h for each rinse) in PBS (O.I M. pH 7.3); or for DA, in 0.05 M PBS with 1% sodium metabisulfite. Incubate overnight in blocking solution (0.25% goat serum/ 1' BS A/PBS): for DA. in- clude \'7c sodium metabisulfite. Incubate overnight in primary antibody diluted appro- priately with PBS. Rinse four times in PBS (two 30-min rinses, one over- night rinse, one 30-min rinse). Incubate overnight in secondary antibody, phalloidin. or both, the reagents diluted appropriately in PBS. Rinse three times in PBS (1 h. overnight. 1 h). Mount the specimens under coverslips in 60% glyc- erol-1% n-propyl gallate/PBS. Muscle. The branchial muscles were visualized with phalloidin conjugated to the fluorescent probe Alexa Fluor 488 (Molecular Probes. Eugene. Oregon), the conjugate used in a concentration of 1 unit/100 jul in 0.1 M PBS. For single-stained preparations, whole mounts were fixed for 1 h in 4<7r formaldehyde with 0.01 M PBS (pH 7.3; 530 mM NaCl). rinsed twice, and then stained overnight. To double- label immunochemically stained preparations, the phalloi- din was added to the tissues at the same time as the sec- ondary antibody. 5HT and YFAFPRQamide. Pieces and sections of demi- branch were fixed overnight in 4% paraformaldehyde in 0.01 M PBS (pH 7.3; 530 mM NaCl); the fixative was prepared as described in Gainey et al. ( 1999a). For 5HT, the primary polyclonal antiserum was raised in rabbit to 5HT conjugated to BSA with paraformaldehdye (Diasorin. Still- water, Minnesota). For YFAFPRQamide. the primary poly- clonal antiserum was raised in rabbit to the peptide conju- gated to thyroglobulin (custom synthesis, etc.. by SynPep. Dublin. California). In both cases, the secondary antibody was raised in goat to rabbit IgG conjugated to Alexa Fluor 594 (Molecular Probes). Dopamine. Pieces and sections of demibranch were fixed for 2 h in 57c glutaraldehyde/19r sodium metabisulfite/PBS (0.01 M: pH 7.3; 530 mM NaCl). The primary polyclonal antiserum was raised in rabbit to DA conjugated to BSA with glutaraldehyde (Diasorin). The secondary antibody was raised in goat to rabbit IgG and conjugated to Alexa Fluor 594 (Molecular Probes). For negative controls, the primary antibodies were omitted from a slide in each series of preparations. Confocal images of 5HT distribution were made with a Leica LSCM SP2 microscope at the Whitney Laboratory. 5 min Figure 2. Traces of contractile activity recorded from three untreated demibranchs taken from a single clam. (A) Quiescent. (B) Occasional, spontaneous contraction. (Cl Arrhythmic, spontaneous contractions. Con- tractions were recorded with force transducers. St. Augustine, Florida. Fluorescent images were made with a Nikon Eclipse TE200 microscope equipped with a Spot RT digital color camera (Diagnostic Instruments). Images were prepared for publication with Adobe Photoshop. Drugs All chemicals were purchased from Sigma-Aldrich, St. Louis, Missouri, or ICN Pharmaceuticals. Costa Mesa, Cal- ifornia. The specificities of the antagonists listed in the tables were obtained from the Cell Signaling & Neuro- science catalog (2000/2001 ed.) of Sigma/RBI. Results Activity of isolated gills Most of the isolated demibranchs were quiescent in the organ baths (Fig. 2a), but occasionally gills would contract spontaneously and relax (Fig. 2b). and on rare occasions they would beat arrhythmically (Fig. 2c). All three demi- branchs in Figure 2 were from the same clam; the fourth, not pictured, was also quiescent. Of the hundreds of prepara- tions we have observed, only a handful showed the sponta- neous, arrhythmic contractions seen in Figure 2c. Pharmacology of branchial muscles Agonists. 5HT, DA, and ACh contracted the gill muscles in a dose-dependent manner, but the response to ACh was observed only in gills pretreated for 15 min with 10~ 4 M eserine (Fig. 3). The responses to all three agonists were 86 L. F. GAINEY ET AL LQG5HT(M1 RESPONSE -7 -6 -S -4 -3 -2 5 min Figure 3. Traces of contractions in response to increasing concentra- tions of 5HT; successive doses were added at the arrows. Each response is from a separate demibranch; data were recorded with force transducers. indistinguishable: the gills contracted tonically and, after 30 s to several minutes, reached their maxima. Dose-response curves for the three agonists were pre- pared (Fig. 4), and their characteristics are listed in Table 1. The rankings of the EC 5() values are ACh < 5HT < DA, and since the 95% confidence intervals do not overlap, the values are statistically different. The maximal contractions in response to 5HT and DA are equal, and both are signif- icantly larger than the maximal contractions induced by ACh. The comparative data noted above were independent of the method used to record the contractions. However, if we consider each agonist separately, then its EC 50 is signif- icantly less, and its induced contractions were larger (except for DA) when the contractions were recorded with the digital ultrasound system rather than with force transducers (Table I). The following neurotransmitters, all applied to the tissue at 1()~ 4 M. neither contracted nor relaxed the gills: ATP, GABA, histamine. and octopamine. Furthermore, the fol- lowing three peptides all found in Mercrmiriti and all applied at 10~ 6 M neither contracted nor relaxed the gills: FMRFamide, AMSFYFPRMamide, and YFAFPRQamide. Previously, we found that YFAFPRQamide modulates the effects of DA on the lateral cilia and those of 5HT on the frontal cilia (Gainey et til.. 1999a). Therefore, to determine whether the peptide would modulate the effects of 5HT or : Ultrasound recorder : Foree transducers : I : i * log |5HT(M) | -(. -5 -4 log|ACh(M)| Figure 4. Dose-dependent muscle contractions (as percentages of the resting length) in response to 5HT, DA, and ACh. Solid circles and lines: data recorded with force transducers; each datum is the response of a separate demihranch. Open circles on dashed lines: data recorded with an ultrasound system from either two (5HT, ACh) or five (DA) demibranchs; each preparation was exposed to increasing concentrations of agonist. Demibranchs exposed to ACh were pretreated with 10 4 M eserine. GILL MUSCLE PHARMACOLOGY AND ANATOMY 87 Table I Summary of dose-response effects for 5HT. DA. and ACh Agonist Type* EC 5() (M)t (95% CDI Cmax (%) (95% CI) 4 : 5HT II 1.1 x 1(T 4 (0.5-1.8 x 1(T 4 ) 20 (17-23) us 2.1 x 10~ 5 (1.9-2.5 x 10~ 5 ) 33 (28-38) DA ft 5.9 x 10~ 4 (2.3-9.5 x 1(T 4 ) 21 (14-29) us 1.4 x 1CT 4 (1.2-1.7 x 10' 4 ) 32 (17-48) ACh ft 1.5 x 10~ 5 (0.34-3.3 x 10" 5 ) 2 (1-3) us 8.6 X 10~ 6 (0.2-1.3 x 10~ 5 ) 9 (7-11) * f t = force transducers used to measure contractions; us = ultrasound used to measure contractions t Concentration of agonist giving a half maximal response. t 95% confidence intervals associated with the estimates. S Maximal predicted contraction. DA on the muscle, we applied YFAFPRQamide to the demibranchs before exposing them to 2 X 1CT 5 M 5HT or DA. At concentrations ranging from 1CT 9 to 1CT 6 M (5HT) or 10~ s to 10~ h M (DA), and exposures ranging from 15 min to 1 h (5HT) or 1 h (DA), the peptide had no effect upon contractions induced by either 5HT or DA. Antagonists. Because the three effective agonists contract the gill, and since the mechanical responses to 5HT, DA, and ACh are indistinguishable, we asked whether the mus- cles have receptors for each of the agonists, or whether one or more of the agonists are acting indirectly by stimulating the release of another agonist from motor nerves. To test these possibilities, antagonists were sought for each agonist, and these agents were cross-tested against the other ago- nists. Controls. In control experiments, each gill received two consecutive, equal doses of the same agonist. For each of these agonists, the second contraction in re- sponse to the same concentration was usually larger than the first, and the contraction ratios were signifi- cantly greater than 1 (Table 2). Moreover, when Table 2 The effect of antagonists on the actions of 5HT. DA, and ACh Antagonist Type* Agonistt Mean contraction ratio SD (n)t />$ None (control) 5HT 2.20 1.86 (72) < 0.001 ''' None (control) DA 2.18 1.74 (46) <0.001* p None (control) ACh 3.21 3.65 (18) 0.00 l* p Cyproheptadine 5HT, 5HT 0.324 0.495 (11) -o.oo ] DA 0.076 0.162 (10) 0.001 ' ACh -1.19 2.10 (9) *r Mianserin 5HT, 5HT 0.475 0.193(8) 0.002*' DA 0.494 0.490(8) <0.001*' ACh -0.449 0.701 (9) *r SKF-83566 DA, DA 0.483 0.300 (9) 0.003*' 5HT 0.869 0.233 (9) 0.32 ACh 0.881 0.530(3) 0.22 Hexamethomum ACh n ACh 0.215 0.262(5) 0.001 '" DA 1.05 0.593 (6) 0.49 5HT 2.14 0.850(8) 0.50 * The primary type of mammalian receptor blocked by the antagonist. t Agonist concentrations: 5HT = 2 x 10~ 5 M; DA & ACh = 5 x 10~ 5 M. Antagonist concentrations were 10' 4 M. t Contraction ratio: height of contraction after exposure to the antagonist/height of contraction before exposure to the antagonist. S P values are one-tailed probabilities: *i = significant inhibition; *p = significant potentiation; *r = relaxation, the second contraction was coded as a negative value. 88 L. F. GAINEY ET AL Table 3 .\iitiixniii.\t.\ thai Inn/ an we//Hni>itli\ Ostia F Gardiner el til.. IWI Ligumia subrostrata Ostia F (I Gardiner el ul .. Wl HETERODONTA Dreissemi polymurpha V/TR F Snow el til.. 1995 V/Ostia + Medler & Silverman . 1997 Corbicula ftuminea IFD/Ostia F Medler & Silverman . 2(101 Mei'centiriti mercenaria TR M This study * DVO. direct visual observation of isolated gills. IFD. intertilument distance, recorded on videotape. Ostia. change in diameter of ostia measured. TR, direct measurement with a transducer of movement or force development. V. measurement of length or area changes recorded on video tape. t M. marine; F. fresh water. Habitat is defined in terms of salinity. Note that the species listed here as "marine" are all at least moderately euryhaline (5-15%c to 30-40%r). The criterion for designation of habitat as "fresh water" (F) is the ability of animals to live and reproduce (or survive prolonged immersion) in fresh water. t +. excitation [increased tone (or ostia increased in diameter): or increased rate, regularity, or amplitude of contractions]. , inhibition [relaxation (or decreased diameter of ostia), or reduced rate, amplitude, or regularity of contractions]. 0, no response observed. The predominant responses of the tissues to each transmitter are listed. bearing the gill cilia, and in nerves running out to that region (Gainey et at.. 1999a). We have certainly not identified all of the transmitters in the innervation of the filaments, but the morphological restriction of YFAFPRQamide to the fila- ments, and its physiological restriction to effects on cilia, suggests that innervation of the branchial filaments may be exclusively in the service of the cilia, and that the remaining two neural divisions regulate the muscles. These consider- ations also support our hypothesis that the proposed syn- apses of dopaminergic and cholinergic neurons onto sero- tonergic neurons will be found in the septa. Coordination between the ciliary pump and the branchial muscles. Two video endoscopic observations suggest that the lateral cilia and the gill muscles act in a coordinated fashion. First, when the gills of a unionid, Pygunodon cataracta. stop pumping, the water tubes constrict, but re-open when pumping resumes (Tankersley, 1996). Sec- ond, when the valves of Mercenaria are closed, the lateral cilia are immobile, and the gills are tonically contracted, both longitudinally and dorsoventrally (Baker, pers. comm.). When the clam is actively pumping, we expect that se- rotonergic stimulation of the muscles is reduced and the muscles are relaxed. Under these circumstances, the ostia, water tubes, and blood vessels would be open, so the flow of water and hemolymph would be maximized. When the clam closes, the dopaminergic innervation would become active, switchini; the lateral cilia off and stimulating the serotoner- gic plexus. The longitudinal and dorsoventral muscles and the water-tube muscles would then constrict, closing the ostia and constricting the water tubes and blood vessels. Comparative aspects of branchial muscle pharmacology Although there is an extensive literature on the pharma- cology of bivalve muscles, it is largely focused on the anterior byssus retractor muscle of Mytilns and isolated ventricles of a variety of bivalves including that of Merce- naria. In contrast, the pharmacology of branchial muscles has been studied in relatively few species of bivalves, in part because the branchial musculature is not an advantageous model for the study of muscle cells per se. Branchial mus- cles are small and are embedded in a complex organ; thus they cannot be directly attached to a recording apparatus. Furthermore, their neural supply is complex, and the inner- vation of specific muscles is not readily accessible. How- ever, the pharmacology of these muscles has been studied by those interested in the physiology of bivalve gills; the available data are summarized in Table 4. The effect of 5HT on the muscle varies with species. The gills of Mytilns and Mercenaria are contracted by 5HT. whereas those of Dreissena polymorphic Anodonta gruiulis, and Ligumia subrostrata are relaxed. There is no taxonomic order in these data; but 5HT contracted the gills of the two marine species and relaxed those of the three freshwater species. Because the sample size is miniscule, however, the 94 L. F. GAINEY ET AL apparent relationship between 5HT action and habitat may be coincidental. ACh had a net excitatory effect on five of the six species on which it was tested; one species (Ligitmia subrostrata) showed no effect, but the gills were not pretreated with eserine. The effect of ACh on the gills of Mercenaria was revealed only after pretreating them with eserine; in addi- tion, the inhibitory effect of ACh became evident only when the gills were exposed to a 5HT antagonist and eserine. The relaxing effect of ACh has not been seen in any of the other gills tested, but then the pharmacological analysis reported here was not used in the other studies. Painter and Greenberg (1982) examined the effects of 5HT and FMRFamide on the ventricles of 50 species of bivalves and remarked that "the responses were strikingly diverse, varying qualitatively with dose as well as species." In their analysis, however, clear taxonomic relationships were discernable. In comparison to ventricles, gills are much more complex and interact directly with the environ- ment. For example, sodium transport in the gills of fresh- water mussels appears to be regulated by a serotonergic neural mechanism (data summarized in Dietz el ai, 1985). groundwork for studies of the integrated control of gill function. Acknowledgments We thank David Chicoine, Robert Pirone, and Kelly J. Vining (University of Southern Maine) for their assistance in performing the pharmacology experiments. In addition, David Chicoine helped assemble Figures 6 and 7. The following individuals at the Whitney Lab provided assis- tance: Dr. Dimitri Budko provided the basic protocol used for whole mounts; Dr. Paul J. Linser provided technical assistance with immunohistochemistry and confocal mi- croscopy and took the picture used for Figure 6E; Leslie Van Ekeris made the cryostat sections and provided tech- nical assistance with the immunohistochemistry; M. Lynn Milstead drew and prepared Figure 1 for publication. Fi- nally, Seth Tyler (University of Maine) advised on the phalloidin staining of muscle. Support was provided to LFG by grants from the Maine Science and Technology Foun- dation and the Bioscience Research Institute of Southern Maine (at the University of Southern Maine). The odd response of the control .y/Y/.v /;; experiments with antagonisls When two successive equal doses of any agonist (i.e., 5HT. DA, or ACh) were applied to the control gills, the second contraction was typically larger than the first, and this result was initially inexplicable. Later, however, we discovered that the gills produce nitric oxide (NO) in re- sponse to 5HT, and that NO potentiates gill muscle con- tractions (Gainey el ai, 1999b). This mechanism may also explain another experimental observation: that ultrasonic transducers record higher maximal contractions than force transducers, and they produce dose-response curves with lower EC 50 s. Thus, when force transducers were used, demibranchs could be exposed only to a single dose of agonist, so the individual contractions constituting the dose- response curves were not potentiated by NO. In contrast, when ultrasonic transducers were used, the demibranchs could be exposed to a set of increasing doses of agonist, so NO was produced, the contractions were potentiated, and the resulting dose-response curves were steeper. Summary The gills of Mercenuria are equipped with an array of muscles and four distinct sets of cilia, and the activity of these effectors coordinated by a complex neural net- work transports water and particles in support of respira- tion and feeding. This paper and a previous one on the modulation of ciliary activity (Gainey et /.. 1999a) lay the Literature Cited Aiello, E. L. 1%0. Factors affecting ciliary activity on the gill of the mussel A/vri'/ii.s etlulix. Pliysiol. Zoo/. 23: 120-135. Aiello, E. I,. 1970. Nervous and chemical stimulation of gill cilia in bivalve molluscs. Physinl. Zool. 43: 60-70. Aiello, E. L. 1990. Nervous control of gill ciliary activity in M\iilns eilitlis. Pp. 189-208 in Neiirahioloxy o/Mytilus edulis. G. B. Stefano, ed. Manchester University Press, Manchester, U.K. Atkins, D. 1937. On the ciliary mechanisms and interrelationships of lamellihranchs. Part I: New observations on sorting mechanisms. Q. J. Microxc. Sci. 79 (New Series): 181-308. Atkins, D. 1943. On the ciliary mechanisms and interrelationships of lamellibranchs. Part VIII: Notes on gill musculature in the microcilio- branchia. Q. J. Micmsc. Sci. 84 (New Series): 187-256. Bayne, B. L. 1998. The physiology of suspension feeding by bivalve molluscs: an introduction to the Plymouth "TROPHEE" workshop. J. Ev/>. Mar. Bio/, few/. 219: 1-19. t'andelario-Martinez, A., D. M. Reed, S. J. Prichard, K. E. Doble, T. D. Lee, W. Lesser, D. A. Price, and M. J. Greenberg. 1993. SCP- related peptides from bivalve mollusks: identification, tissue distribu- tion, and actions. Bio/. Bull. 185: 428-439. Catapane, E. J. 1983. Comparative study of the control of lateral ciliary in marine bivalves. Comp. Biochem. Pliysiol. 75C: 403-405. Dame, R. F. 1996. Ecology of Marine Bivalves: an Ecosystem Approach. CRC Press. Boca Raton, FL. Dietz, T. H., W. L. Steffens, W. T. Kays, and H. Silverman. 1985. Serotonin localization in the gills of the freshwater mussel, Ligiimia MihroMrala. Can. J. Zool. 63: 1237-1243. Eble, A. F. 2001. Anatomy and histology of Mercenaria inercenaria. Pp. 1 1 7-220 in Biologv of The Hani Clam, J. N. Kraeuter and M. Caslagna, eds. Elsevier Science. Amsterdam. Gainey, L. F., Jr., K. J. Vining, and M. J. Greenberg. 1998. Pharma- cology of gill muscle libers in Mercenaria inercenaria. Am. Zool. 38: 122A. Gainey, L. F., Jr., K. J. Vining, K. E. Doble, J. M. Waldo, A. Cande- GILL MUSCLE PHARMACOLOGY AND ANATOMY 95 lario-Martinez, and M. J. Greenberg. 1999a. An endogenous SCP- related peptide modulates ciliary healing in the gills of a venerid clam. Merceiuiria merceiitiriti. Biol. Bull. 197: 159-173. Gainev, L. F., Jr.. R. T. Pirone, and M. J. Greenberg. 1999b. Nitric oxide potentiates gill muscle contraction in Meneiuuiu mcrcenaria. Am. Zoo/. 39: 71 A. Gainey, L. F.. Jr., J. Walton, and M. J. Greenberg. 2001. Neuromus cular anatomy of clam gills. Am. Zoo/. 41: 448 (Abstract). Galtsoff, P. 1964. The American oyster. Chapter VII: The gills. Fish. Bull. US Fixh Wikll. Sen: 64: 121-151. Gardiner. D. B., H. Silverman, and T. H. Dietz. 1991. Musculature associated with the water canals in freshwater mussels and response to tnonoamines in vitro. Biol. Bull. 180: 453-465. Grunbaum, D., D. Eyre, and A. Fogelson. 1998. Functional geometry of ciliated tentacular arrays in active suspension feeders. J. Exp. Biol. 201: 2575-2589. Jorgensen, C. B. 1975. On gill function in the mussel. A/vri/ii.v CI/H//.V L. Ophelia 13: 1X7-232. Jorgensen, C. B. 1996. Bivalve filter feeding revisited. Mar. Ecol. Prog. Ser. 142: 287-302. Kellogg, .1. L. 1892. A contribution to our knowledge of the morphology of lamellibranchiate molluscs. Bull. U.S. Fix/:. Coinm. 10: 389-436. Malanga, C. J. 1975. Dopaminergic stimulation of frontal ciliary activ- ity in the gill of Mytilus editlis. Comp. Biochem. Phvsio/. 51C: 25-34. Medler. S., and H. Silverman. 1997. Functional organization of intrin- sic gill muscles in /.ebra mussels, Dreissena polyinarpha (Mollusca: Bivalvia). and response to transmitters in vitro. Invertehr. Biol. 116: 200-212. Medler. S., and H. Silverman. 2001. Muscular alteration of gill geom- etry in vitro: implications for bivalve pumping processes. Biol. Bull. 200: 77-86. Nelson, T. C. 1960. The feeding mechanism of the oyster. II. On the gills and palps of Ostrea eJulis, Crassostrea virginica and C. angulata. J. Morpliol. 107: 163-203. Painter, S. D., and M. J. Greenberg. 1982. A survey of the responses of bivalve hearts to the molluscan neuropeplide FMRFamide and to 5-hydroxytryptamine. Biol. Bull. 162: 311-332. Ridewood, W. G. 1903. On the structure of the gills of the Lamelli- branchia. Philox. Trans. R. Sot: Loiul. 195: 147-284. Roop, T., and M. J. Greenberg. 1967. Acetylcholinesterase activity in Crassostrea virgiuica and Mercenuriti mercenaria. Am. Zoot. 7: 737- 738. Setna, S. B. 1930. The neuromuscular mechanism of the gill of Pecten. Q. J. Mit-rost: Sci. 73: 365-392. Snow, V., J. Duncan, and J. L. Ram. 1995. Contractile responses of zebra mussel gills. Ev/>. Biol. 95: A640. Abstract #3711. Tankersley, R. A. 1996. Multipurpose gills: effect of larval brooding on the feeding physiology of freshwater unionid mussels. Invenehr. Biol. 115: 243-255. Welsh, J. H., R. I. Smith, and A. E. Kammer. 1968. Uihorutorv Exercises in Invertebrate Physiology, 3rd ed. Burgess, Minneapolis, MN. Reference: Biol. Hull. 2(14: Wi-l().v il-'ebruar> 2 2003 Marine Biological Laboratory Salinity Tolerance of Larval Rapana venosa: Implications for Dispersal and Establishment of an Invading Predatory Gastropod on the North American Atlantic Coast ROGER MANN* AND JULIANA M. HARDING Department of Fisheries Science, Virginia Institute of Marine Science, College of William and Mary. Gloucester Point. Virginia 23062 Abstract. The lack of quantitative data on the environ- mental tolerances of the early life-history stages of invading species hinders estimation of their dispersal rates and estab- lishment ranges in receptor environments. We present data on salinity tolerance for all stages of the ontogenetic larval development of the invading predatory gastropod Rapana venosa, and we propose that salinity tolerance is the dom- inant response controlling the potential dispersal (= inva- sion) range of the species into the estuaries of the Atlantic coast of the United States from the current invading epicen- ter in the southern Chesapeake Bay. All larval stages exhibit 48-h tolerance to salinities as low as 15 ppt with minimal mortality. Below this salinity, survival grades to lower values. Percentage survival of R. venusu veligers was sig- nificantly less at 7 ppt than at any other salinity. There were no differences in percentage survival at salinities greater than 16 ppt. We predict that the counterclockwise, gyre-like circulation within the Chesapeake Bay will initially distrib- ute larvae northward along the western side of the Del- MarVa peninsula, and eventually to the lower sections of all major subestuaries of the western shore of the Bay. Given the observed salinity tolerances and the potential for dis- persal of planktonic larvae by coastal currents, establish- ment of this animal over a period of decades from Cape Cod to Cape Hatteras is a high probability. Introduction The Norway/United Nations Conference on Alien Spe- cies considers alien invasive species as the second most Receded 2 July 21)02; accepted 4 November 2002. ' ! To whom correspondence should be addressed. E-mail: rmann@vims.edu important threat, after habitat destruction, to indigenous biodiversity (Sandlund ct / various ages 10-27 days post-hatch) exposed to 20 veligers per treatment Veliger age (days post-hatch) Salinity (ppt) 2 4 6 9 11 13 7 76.9(3.8) 65.5 (5.6) 51,0(5.4) 20.5 (6.6) 62.3(6.8) 74(12.4) 29.4(8.0) 10 74.8(5.3) 70.9(5.5) 76.0(3.4) 67.1 (4.3) 72.1 (5.0) 92.4(1.8) 42.6(2.9) 13 77.9(2.4) 79.3 (2.S) 59.5(6.5) 75.9(9.7) 75.7 (6.5) 90.9(3.0) 69.2(2.6) 16 80.7(2.4) 80.7(3.1 1 90.7(3.1) 82.2(3.0) 83.8 (3.9) 89.2(1.7) 83.5 (2.2) 19 78.9(3.4) 82.6(3.4) 97.1 (0.8) 90.9(2.6) 77.3(3.7) 86.9(5.2) 83.1 (3.1) 22 85.7(3.0) 80.9(5.4) 97.8(1.4) 90.4(2.2) 69.5(3.2) 88.7(2.4) 87.8(1.8) 25 83.5(6.7) Sl.412.4) 94.7(2.8) 94.9 (0.9) 74.9(7.3) 93.5(1.5) 87.6(1.9) 32 59.5 (3.8) 90.5(4.1) 89.3 (1.4) 97.3 (1.4) 79.6(0.9) 85.8(6.5) 84.3(4.1) 15 17 19 21 -j ^ 25 27 7 2.3(2.3) 23.3 (14.4) 58.4(6.7) 95.8(2 1) 75.9(1.7) 97.2(2.8) 92.5(3.8) ID 42.5(12.2) 69.4(3.9) 82.2(4.2) 93.3(3.5) 66.4(4.5) 97.0(3.0) 88.7(8.0) 13 59.9(8.0) 54.6(14.4) 83.0(0.7) 88.9(5.9) 75.0(0.9) 89.9(1.4) 79.5 (8.8) 16 74.3(3.5) 59.7(5.9) 87.0(1| 93.9(6.1) 67.7(6.4) 97.2(2.8) 82.6(11.5) 19 68.8(12.5) 61.4(4.2) 84. 1 (2.1) 92.0(5.4) 74.8(4.1) 97.4(2.5) 95.2 (4.8) 22 78.8(4.3) 68.8 (4.9) 78.7(5.4) 86.1 (4.3) 81.0(1.9) 87.1 (6.5) 100(0.0) 25 71.5(2.6) 73.0(4.1) 72.7(3.8) 96.7(1.7) 80.9(1.0) 98.0(2.0) 98.7(1.3) 32 56.3(10.3) 67.6(3.3) 73.7(4.2) 94.5 (2.S) 69.9(2.4) 93.7(3.4) 96.4 ( 1.8) unclear. The limited observations from the Kerch Strait region suggest that an upstream limit of 12-13 ppt in the Chesapeake Bay is possible, and that low winter tempera- tures will not exclude Rapana from regions that infre- quently experience winter ice. Wu ( 1988) reports that in its native range, R. venosa can exploit estuarine regions that have warm summer temperatures and avoid possible surface freezing in winter by migrating into deeper water in these regions. Larvae of R. venoaa exhibit considerable plasticity in the duration of their planktonic development under experimen- tal conditions of temperature and salinity that mimic the summer conditions in the Chesapeake Bay, and they do not require specific metamorphic cues to complete the transition to the crawling, benthic post-larval phase (Harding and Mann, unpubl. data). Laboratory-cultured individuals can exploit a variety of native bivalves as prey, including the Table 2 Suiiiniars of two-factor AN() \'.\ n'elr^cr a^e salinity) used Jo \alpin.\ has no pelagic larval stage. Juveniles of Urosalpinx hatch and crawl away from the substrate-at- tached egg masses. Urosalpinx has been recolonizing its former Bay habitat over the past three decades by crawling up the Bay bottom over "islands" of suitable substrate. In the absence of an invader, the temporary displacement of Urosalpinx is but a minor perturbation in evolutionary time; however, the introduction of Rapana adds a new and op- portunistic component to this reestablishment process. There arguably now exists a race to reoccupy this tempo- rarily vacated niche; a race that may favor the invader because of the sequence of events that temporarily displaced the native species. Vermeij (1996) theorized that physiological tolerances are evolutionarily conservative parameters contributing to the determination of the range of survival. In this context we predict that, as a result of the counterclockwise, gyre-like circulation within the Chesapeake Bay, pelagic larvae of Rapana venosa originating from parents in the Hampton Roads region will initially be distributed northward along the western shore of the DelMarVa peninsula, and will eventually reach the lower sections of all the major subestu- aries of the western shore of the Bay. This entire region is within the salinity tolerance of the larval forms (compare Table 1 with Fig. 2). The potential for long-distance dis- persal within a single generation remains to be determined, although recent collections of small (<75 mm in length) adults on the Virginia Bay shore of the peninsula suggest that a distance of tens of kilometers per generation is pos- sible. Dispersal onto and along the coastal shelf outside of the Bay mouth may be influenced by both northward- and southward-flowing residual current. The effects on dispersal depend on depth, wind conditions, and time within the known egg laying period of the invader in the southern Chesapeake Bay. Establishment over a period of decades by natural dispersal in estuaries and coastal regions from Cape Cod to Cape Hatteras was considered a high probability by Mann and Harding (2000). This prediction still stands and is supported by the essentially continuous distribution of mol- lusc species suitable as prey in shallow waters throughout this range (for examples, see Theroux and Wigley, 1983). The time frame may, however, be considerably reduced by dispersal of larval forms in ballast water during intra-coastal maritime trade, a suggestion reinforced by the tolerance of the larval form (this study) and the location of both the Norfolk, Virginia. U.S. Naval base and an international 102 R. MANN AND J. M. HARDING container terminal within the extant adult range of invasion in Hampton Roads. If, as Vermeij (1996) suggests, factors such as "the presence of competitors, predators, or disease organisms, or the absence of a critical host, food, or sym- biotic species" prevent a species from extending its range, it is unlikely that Ru/wmi will be further restricted within the projected range. Large individuals of R. venoxu appear admirably equipped to compete with large native gastropods and have few obvious predators in the Middle Atlantic coastal region when they are full grown. We can find no reports of diseases of R. venosa in any of its native or introduced ranges. Finally, the only notable parasite of R. venosa in both its Black Sea and Chesapeake Bay popula- tions are shell-boring polychaetes of the genus Polydora (Gutu and Marinescu, 1979; Mann and Harding. 2000). The actions of Palydoni appear to have little, if any, detrimental effect on infected individuals in either location; may be limited to some individuals of R. venosa that forage epifau- nally; and may be terminated by burial of the host whelks as they grow and shift to an infaunal habit. Indeed, observa- tions on rapa whelk biology and physiological tolerances in the Chesapeake Bay strongly suggest that this animal is capable of successful colonization and establishment of viable populations within estuarine habitats up and down the East Coast of the United States. Acknowledgments Support for this project was provided by Virginia Sea Grant (R/MG-98-3), the Department of Fisheries Science, Virginia Institute of Marine Science, and partial support to RM by the National Science Foundation (OCE-98 10624). Special thanks are extended to local watermen and seafood processors who donated adult Ru/wmi to our research col- lection. We thank D. Bryn Jones, Dario Savini, Melissa Southworth, Rhonda Howlett, Peter Kingsley-Smith, Erica Westcott, Stephanie Haywood, and Catherine Ware for as- sistance in maintenance of adult brood stock and larval cultures. This manuscript is dedicated to the late Professor Ruth Dixon Turner, whose enthusiasm for the larval ecol- ogy of marine molluscs remains as an inspiration to us all. This is Contribution Number 2506 from the Virginia Insti- tute of Marine Science. Literature Cited Bombace, G., G. Kabi, L. Fiorentini. and S. Speranza. 1994. Analysis nl I he efficacy of artificial reefs located in five different areas of the Adriatic Sea. Fifth International Conference on Aquatic Habitat En- hancement. Bull. Mm. Sci. 55(2-3): 559-580. Carlton, ,|. 1996. Pattern, process, and prediction in marine invasion ecology. Binl. Con.scrv. 78: 97-106. Carlton, J. 1999. Molluscan invasions in marine and estuarine commu- nities. Malacologia 41: 439-454. Chukchin, V. 1984. Ecology of the (lastropo,! Molluscs ,>/ the Black Sea. Acad. Sc. USSR, Naukova Dumka. Kiev. 175 pp. (in Russian). Chung, E.V, S. V. Kim, and V. G. Kim. 1993. Reproductive ecology of the purple shell, Rapana venosa (Gastropoda: Muricidae). with special reference to the reproductive cycle, deposition of egg capsules and hatching of larvae. Korean J. Malacol. 9(2): 1-15. Cucaz, M. 1983. Rapana venosa (Valenciennes, 1846) vivente nel Golfo di Trieste. Bull. Malacol. 19(9-12): 261-262. Drapkin, E. 1963. Effect of Rapana he-oar Linne ( Mollusca, Muricidae ) on the Black Sea fauna. Dnkl. Akad. Nauk. SSSR. 151: 700-703. Ghisotti, F. 1971. Rapana llioinasiana Crosse, 1861 (Gastropoda, Muri- cidae) nel Mar Nero. Conc/iig/ie (Milan) 7: 55-58. Ghisotti, F. 1974. Rapana venosa (Valenciennes), nuova ospite Adri- atiea'.' Conciliate I Milan) 10: 125-126. Golikov. A. N. 1967. Gastropoda: Pp. 79-91 in Animals and Plants <>/ Peter rite Great Bay. Nauka, Leningrad. Green, R. 20(11. Morphological variation of three populations of the veined rapa whelk. Rapana venosa, an invasive predatory gastropod species. Master's thesis. College of William and Mary, Gloucester Point. VA. Gutu, M., and A. Marinescu. 1979. PolvJora ciliata Polychaeta perfo- rates the gastropod Rapana thomasiana of the Black Sea. Trav. Mtts. Hist. Nat. "Grigore Antipa" 20: 35-42. Hanna, G. I). 1966. The introduced mollusks of Western North America. O/ Malacology, Republic of China. 7: 27-32. Luckenbuch, M., R. Mann, and J. E. Wesson, eds. 1999. Oyster Reef Habitat Restoration: A Synopsis of Approaches. Virginia Institute of Marine Science. Gloucester Point. VA. 366 pp. Mann, R. 200(1. Restoring oyster reef communities in the Chesapeake Bay: a commentary. ./. Shellfish Res. 19: 335-340. Mann, R. 2001. Restoration of the oyster resource in the Chesapeake Bay. Bull. Aauaeul. Assoc. Can. 101: 38-42. Mann, R., and J. M. Harding. 2000. Invasion of the North American Atlantic coast by a large predatory Asian mollusc. Binl. Invasions 2: 7-22. Marinov, T. M. 1990. The Zoohenthos From the Bulgarian Sector of the Black Sea. Acailenn of Science Publications Sofia. 195 pp. |ln Bulgar- ian]. Mel, P. 1976. Sulla presen/a di Rapana vcnosa (Valenciennes) e di Charonia variegala seaiien-ae (Ar. & Ben.) nell'Alto Adriatico. Con- chiglie. (Milan} 12(5-6): 129-132. Morton, 1$. 1994. Prey preference and method of attack by Rapana bc-oar (Gastropoda: Muricidae) from Hong Kong. Pp. 309-325 in The Malacofaiina of Hong Kong anil Southern China III, B. Morton, ed. Hong Kong University Press. Hong Kong. Pastorino, G. P., A. Penchaszadeh, L. Schejler, and C. Bremec. 2000. Rapana vcnosa (Valenciennes) (Mollusca: Muricidae): a new gastro- pod in South Atlantic waters. J. Shellfish Res. 19: 897 900. Rennie, S., and B. Niel.son. 1994. Chesapeake Ba\- Atlas. Virginia Institute of Marine Science. Gloucester Point. VA. SALINITY TOLERANCES OF LARVAL RAPANA 103 Kinaldi, E., 1985. Rapana vcnosa (Valenciennes) spiaggiata in notevole quantity sufla spiaggia di Rimini (Fo). Boll. Malacol. 21: 318. Knl. insliii in, I. G., and V. I. Hiznjak. 1998. Stocks of Rapana thoina- suina in the Kerch Strait. Ryhn. A7i; 1: 34-41 (in Russian). Sandlund, O. T., P. J. Schei, and A. Viken. 1999. Introduction: the many aspects of the invasive alien species problem. Pp. 1-7 in Invasive Species and Biodiversity Management, O. T. Sandlund. P. J. Schei, and A. Viken. eds. Kluwer Academic Publishers, Dor- drecht. 413 pp. Savini, D., J. M. Harding, and R. Mann. 2003. Rapa whelk Rapana venosa (Valenciennes, 1846) predation rates on hard clams Mercenaria mercenaria (Linneaus. 1758). J. Shellfish Res. 21(2). (In press). Stroup, K., and R. Lynn. 1963. Atlas of Sulinnv and Temperature Distributions in Chesapeake Bay 1952-61 and Seasonal Averages WV-rt/. Graphical Summary Report 2. Report 6,-f-/. The Chesapeake Bay Institute, The Johns Hopkins University. Terrcni. G. 1980. Molluschi poco conosciuti dell'Arcipelago Toscano: 1-Gasteropodi. Boll. Malacol. 16: 9-17. Theroux, R. B., and R. L. Wigley. 1983. Dixtrihiition and Abundance o/ East Coast Bivalve Mollusks Based on Specimens in the National Marine Fisheries Service Woods Hole Collections. NOAA Technical Report NMFS SSRF-76X. 172 pp. Tsi, C. Y., X. T. Ma, Z. K. Lou, and F. S. Zhang. 1983. Illustrations of the Fauna of China (Mollitscal Vol. 2, plates I-IV. Science Press, Beijing. 150 pp. Vermeij. G. J. 1993. A Natural History of Site/Is. Princeton University Press. Princeton, NJ. Vermeij, G. J. 1996. An agenda for invasion biology. Bio/. Conserv. 78: 3-9. Williamson, M. 1996. Biological Invasions. Chapman and Hall, London. 244 pp. Wu, Y. 1988. Distribution and shell height-weight relation of Rapanu renosa Valenciennes in the Laizhou Bay. Marine Science/Hanani; Kexue 6: 39-40. Zolotarev, V. 1996. The Black Sea ecosystem changes related to the introduction of new mollusc species. Mar. Ecol. 17: 227-236. Reference: Biol. Bull. 204: 104-108. (February 2(1113) 2003 Marine Biological Laboratory Short-Distance Spawning Migration of Tropical Freshwater Eels JUN AOYAMA 1 -*. SAM WOUTHUYZEN 2 . MICHAEL J. MILLER 1 , TADASHI INAGAKI 1 . AND KATSUMI TSUKAMOTO 1 ' Ocean Research Institute, The University of Tokyo, 1-15-1 Minamidai. Nakano. Tokyo 164-8639. Japan: and : Research Center for Oceanography, Indonesian Institute of Sciences, Anchor. Jakarta, Indonesia The freshwater eels have fascinated biologists because of their spectacular long-distance migrations between their freshwater habitats and their spawning areas far out in the ocean (1. 2, 3). Recent progress on the molecular phytogeny of freshwater eels suggests that the\ originated in the trop- ics (4), and information on the reproductive ecologv and recruitment of tropical species will provide new insight into the evolution of the spawning migration of the freshwater eels (5, 6). However, the larvae (leptocephali) of the intinv sympatric tropical species are morphologically similar (7). so they are impossible to identify, and their spawning areas are thus virtually unknown. Recently, however, we have collected small leptocephali from around Sulawesi, Indone- sia, and have used species-specific genetic markers to iden- tify them as laiTae of Anguilla celebesensis and A. borne- ensis, which provides the first definitive information about the general spawning areas of these tropical eels. More- over, the discovery of a spawning area of A. celebesensis in Toinini Bay and the presence of small specimens of two species in the Celebes Sea indicate that, in contrast to the long migrations made b\ temperate eels, tropical eels make much shorter migrations to spawn in areas near their fresh- water habitats. This difference in migratory behavior may reflect an evolutionary dine among freshwater eels that extends from tropical to temperate regions. Early in the lust century, the Danish oceanographer Jo- hannes Schmidt succeeded in collecting small anguillid leptocephali in the Sargasso Sea thousands of kilometers away from their growth habitats in Europe and North Amer- ica, which indicated that the two species of Atlantic fresh- Received 20 June 2002: accepted 4 December 2002. * To whom correspondence should he addiessed. iaoyama@ori.u-tokyo.ac.jp [.-mail: water eels make remarkably long spawning migrations ( 1 ). After this finding, he and his colleagues shifted their efforts to search for the spawning areas of freshwater eels in the Indo-Pacin'c region where most of the species in the genus are found. They successfully collected more than 1400 leptocephali in the Indo-Pacific region (7) during the Carls- berg Foundation's oceanographic expedition around the world from 1928 to 1930. However, most of these lepto- cephali were relatively large, and the overlap in their mor- phological characters made it difficult to identify them exactly (7, 8). Since then, the spawning areas of the Indo- Pacitic anguillid species have remained a mystery, except for the Japanese eel, Anguilla japonica, which was found to spawn to the west of the Mariana Islands in the western North Pacific (3). Molecular phylogenetic research on the genus Anguilla has recently stimulated interest in the spawning migrations of tropical freshwater eel species by suggesting that the tropical eel. Angnilla borneensis. which is endemic to Borneo, is the most basal species, and that the genus radi- ated out from the tropics to colonize temperate regions (4). Because of these studies, species-specific genetic markers can be used to identify all species of anguillid leptocephali (9). Molecular techniques have, for the first time, allowed studies on the distribution of tropical anguillid leptocephali that can reveal the location of eel spawning areas and the nature of their migrations. In this study, we have collected the smallest tropical eel leptocephali ever reported and used molecular genetic methods to identify species of anguillid leptocephali from the Celebes Sea and Tomini Bay on the east side of Sulawesi. We provide the first definitive infor- mation on the spawning areas of tropical freshwater eels. A cruise of the R/V Hakuho Maru (Ocean Research 104 SHORT MIGRATIONS FOR TROPICAL EELS 105 Institute, University of Tokyo) in the western Pacific. Celebes, and Sulu Seas was conducted from 14 January to 10 March 2000 (Fig. I). A subsequent cruise of the R/V Banina Java VII (Research Center for Oceanography. In- donesian Institute of Sciences), made in the waters around Sulawesi from 8 to 30 May 2001, partly overlapped the sampling area of the Hakuho Muni cruise (Fig. 1). Lepto- cephali were collected during both cruises using identical Isaacs-Kidd midwater trawls with 8.7-irr mouth openings and 0.5-mm mesh. The collections usually consisted of 60-min oblique or step tows within the upper 300 m. Aboard ship, the leptocephali were tentatively identified on the basis of morphological characteristics (7, 8), but these character- istics could not always indicate one species. Total length and other measurements were recorded, and the specimens were preserved in 95% ethanol. In the laboratory, the spec- imens were identified by comparing their mitochondria! 16S rRNA gene sequences with those of morphologically well- identified adult specimens, as has been previously reported (9). Briefly, total genomic DNA was extracted from each leptocephalus according to a standard protocol (9). A por- tion of the mitochondria! 16S ribosomal RNA gene (about 500 base pairs) was amplified by the polymerase chain reaction (PCR) using two oligonucleotide primers, H2510 and H3058 (9). Amplification parameters were 30 cycles of denaturation at 94 C for 30 s, annealing at 55 C for 30 s, and extension at 72 C for 60 s. The PCR products were sequenced according to the manufacturer's protocol (Ap- plied Biosystems Inc.) on a 373A DNA sequencer (Applied Biosystems Inc.). Sequences were determined from the light strand only. The sequence data obtained from the leptoceph- ali were directly compared, without any alignments, to homologous data for anguillids in the Pacific region (A. japonica, A. aiistralis. A. borneensis, A. celebesensis, A. dieffenbachii, A. hicolor. A. megastoma, A. inarmontta, A. reinhanltii. A. obscura, and A. interioris). deposited in DDBJ/ EMBL/GenBank under accession numbers AB021748, AB021751-AB021754. AB021757. AB021758, AB021760- AB021762, AB021764. Within species the sequences were the same or had only one or two site differences, but among species the differences were more than threefold (6-74 sites). All of the sequences determined in the present study will appear in the DDBJ/EMBL/GenBank nucleotide se- quence databases with the accession numbers: AB097700- AB097767, in sequence. During the two research cruises, we collected 67 lepto- cephali and one glass eel (the transparent early juvenile stage of eels) of the genus Anguilla (15 leptocephali from the Hakuho Mam cruise and 53 from the Barium Java VII cruise, which included the glass eel). Genetic species iden- tification of the leptocephali clearly distinguished 12 A. marmoruta (34.0-50.7 mm in total length and one glass eel of 47.8 mm), 41 A. celebesensis (13.0-47.8 mm), 3 A. borneensis (8.5. 13.0, 35.4 mm). 4 A. hicolor (42.6-49.2 mm), and I A. interioris (48.9 mm), all from the waters around Sulawesi (Fig. 1). Also identified were 3 A. bicolor (31.3-46.0 mm) from the waters to the north of New Guinea, and 2 A. marmorata (28.0, 36.5 mm), 1 A. obscura (36.7 mm), and 1 A. aiistralis (47.0 mm) from the western South Pacific (their distributions are not shown). This is the first description of the distribution of anguillid leptocephali identified using genetic markers from Indone- sian waters, and these data suggest that as many as five species of the genus Anguilla may use the Indonesian waters as an area for spawning and larval development. Jespersen (7) reported on collections of relatively large anguillid lep- tocephali from many of these same areas, but could not make precise species identifications. Another more recent study used the same molecular genetic techniques used in this study to identify anguillid leptocephali as small as 16.3 mm in the western North and South Pacific, but did not make collections in the Indonesian Seas (9). Of particular interest in our study were the small leptocephali of A. borneensis that were 8.5 and 13.0 mm (indicating an age of about 16 and 26 days after hatching [10, I 1 1), which were collected in the Celebes Sea to the east of Borneo, and the specimen (35.4 mm) that was collected to the south in Makassar Strait, where water from the Celebes Sea is trans- ported (12). The freshwater growth habitat of A. borneensis is limited to the east-central part of Borneo (5. 6). which strongly suggests that this species spawns in the Celebes Sea and then migrates back to its growth habitat adjacent to the spawning area (Table 1 ). Another tropical anguillid species, Anguilla celebesensis, has a wider distribution that extends from Luzon of the Philippines to across Sulawesi (5, 6). Interestingly, the small leptocephali of this species collected about 25 days after hatching (10, 11) were found in two different seasons and in two different areas separated by the northern arm of Sulawesi: a I2.3-mm specimen was found at Station 50 in the Celebes Sea in February, and a 13-mm specimen was found at Station 21 in Tomini Bay in May (Fig. 1 ). Further, the collection of nine A. celehesensis leptocephali at Station 21 in Tomini Bay ranged in total length from 13 to 48.9 mm (fully grown larval stage [10, 1 1 1). These facts indicate that individuals of A. celebesensis inhabiting the watershed of Tomini Bay spawn over a relatively long period and that their leptocephali are retained in Tomini Bay because it is semi-enclosed and its waters apparently do not mix much with those of other areas (13). Therefore, these eels are probably geographically isolated from those in the Celebes Sea. The findings reported here indicate that freshwater eels living in tropical areas may have life-history characteristics that differ markedly from those of their temperate relatives, which have a single spawning site for each species, long spawning migrations in both the North Atlantic (1,2) and North Pacific Ocean (3), and distinct spawning seasons (Table I ). This distinction is supported by recent analyses of 120 E 125 E 120 E 125 E Pacific Ocean 10N 5N 5S A. celebesensis 10N 5N 5S A. borneensis A. marmorata A. bicolor A. interioris Figure 1. I In- study area, showing the sampling stations and the locations where the various species of anguillid leptocephali were collected. Upper left: Stations during cruises of the R/V Hakiiho Mam (circles) in February 2000 and the R/V Biininu Jnyu \'ll (squares) in May 2001. Only a few station numbers are shown to indicate the order of sampling or to identify those mentioned in the text. Upper right: Solid symbols indicate sampling stations where A. < r/< />( uv/w.v leptocephali were collected and open symbols indicate negative stations. Lower left: Collection locations of .A. hitrnecnsis. Lower right: Collection locations of ,4. marmorata. A. bicolor. and A. interioris. Symbols same as lor upper right panel. SHORT MIGRATIONS FOR TROPICAL EELS Table 1 Comparison of presumed spawning areas, ranges, and distant:? of migration of tempi-rate and tropical eels, (genus Anguilla) 107 Species Presumed spawning area Approximate distance to spawning area (6) Approximate latitudinal range (5) Nearest Endmost European eel .4. anguilla Sargasso Sea (2) 25N, 60W 28N-68N 4000 km Azores. Cape Verde Is. 8000 km Norway. Mediterranean American eel A. rostrata Sargasso Sea (2) 25N. 60W 10 : N-62N 900 km Greater Antilles 5500 km Iceland Japanese eel .4. japtmica West of Mariana Is. (3) 15N. 142E 18N^t3N 2000 km Taiwan 3500 km Northern Japan A. borneensis Celebes Sea 3N, 122E Equutor-7N 480 km Tawau, Borneo 650 km Mahakam Riv.. Borneo A. ce/ebesensis* Tomini Bay r j S. 121 : E ].4S-0.5N 80km Coastal areas around the bay 300 km Numbers in parentheses are reference citations; see Literature Cited. * For a part of a species or population found in the present study. otolith microstructure which showed that tropical anguillids in the Indonesian region may spawn (14) and recruit to freshwater (15) throughout much of the year. Catadromous freshwater eels have been suggested to have originated in the tropical waters near present-day Indonesia sometime around the late-Cretaceous to Eocene, and the endemic tropical species A. borneensis has been found to be the most likely basal catadromous eel species (4). Various character- istics of the migrations of present-day anguillid species clearly show at least a partial geographic cline (Table 1 ). For example, the most likely basal species, A. borneensis, is distributed narrowly over about 7 degrees of latitude (Equa- tor to 7 N) and spawns nearby in the Celebes Sea at a distance of only 480-650 km; in contrast, the growth hab- itat of the European eel extends widely over 40 degrees of latitude, and the distance to its spawning area in the Sar- gasso Sea ranges from about 4000 to 8000 km (Table 1 ). This and the short spawning migration of A. celebesensis in Tomini Bay suggest that freshwater eels of the genus An- guilla originally had migrated only short distances to local spawning areas in the warm waters surrounding their fresh- water growth habitats in the tropics. But following the passive, long-range dispersion of their leptocephalus stages by currents, freshwater eels may have had to evolve long- distance migrations to return from their temperate growth habitats to their tropical spawning grounds. The Atlantic species of freshwater eels are often used, even in basic biological textbooks, as a classic example of a species with a spectacular long-distance migration. How- ever, our findings provide the first evidence that this long- distance migration is an adaptation by eels that colonized temperate regions. Therefore, a new era of research on the ecology and behavior of tropical eels has begun, and it promises to unveil the mystery of the origin and evolution of the catadromous migrations of the genus Anguillci. Acknowledgments We thank all of the scientists who participated in the eel cruises and who worked together discussing the sampling design, operating the nets, and sorting samples. We wrote this paper on behalf of all the scientists aboard, and we also thank all the crew of the Hakuho Muni and the Baruna Jaya VII for their help during the cruise. This work was supported in part by Grants-in-Aid Numbers 07306022. 07556046. 08041139, 08456094, 10460081. and 11691177 from the Ministry of Education, Science, Sports and Culture. Japan, and by grant Numbers JSPS-RFTF 97L00901 from the "Research for the Future Program" of the Japan Society for the Promotion of Science. KT was supported by the Re- search Foundation "Touwa Shokuhin Shinkoukai" and the Eel Research Foundation "Noborika." Literature Cited 1 . Schmidt, J. 1922. The breeding places of the eel. ./. Philox. Trans. R. Sm: R 211: 174-208. 2. McCleave, J. D., R. C. Kleckner. and M. C'astonguay. 1987. Re- productive sympatry of American and European eels and implications for migration and taxonomy. Am. Fish. Soc. Symp. 1: 286-297. 3. Tsukamoto. K. 1992. Discovery of spawning area for the Japanese eel. Nature 356: 789-791. 4. Aoyama, J., M. Nishida, and K. Tsukamoto. 2001. Molecular phylogeny and evolution of the freshwater eels, genus Anguilla. Mol. Phylogenet. Evol. 20: 450-459. 5. Ege, V. 1939. A revision of the genus Anguilla Shaw. Dana-Rep. 16: 1-256. 108 J. AOYAMA ET AL 6. Tesch, F. W. 1977. The Eel: Biology ana 1 Management of Anguillid Eelx. Chapman and Hall. London. 7. Jespersen, P. 1942. Indo-Pacific leptocephalids of the genus An- guilla. Dana-Rep- 22: 1-128. 8. Castle, P. H. J. 1963. Anguillid leptocephali in the southwest Pa- cific. Zoo/. Pnbl. Victoria Univ. 33: 1-14. 9. Aoyama, J.. N. Mochioka, T. Otake, S. Ishikawa, Y. Kawakami, P. H. J. Castle, M. Nishida, and K. Tsukamoto. 1999. Distribution and dispersal of anguillid leptocephali in the western Pacific revealed by molecular analysis. Mar. Ecol. Prog. Ser. 188: 193-200. 10 Aral, T., J. Aoyama, S. Ishikawa, M. J. Miller, T. Otake, T. Inagaki. and K. Tsukamoto. 2001. Early life history of tropical Anguilla leptocephali in the western Pacific Ocean. Mar. Biol. 138: 887-895. 1 1. Ishikawa, S., K. Suzuki, T. Inagaki. S. Watanabe, Y. Kimura, A. Okamura, T. Otake, N. Mochioka, Y. Suzuki, H. Hasumoto, M. Oya, M. J. Miller, T. W. Lee, H. Fricke, and K. Tsukamoto. 2(101. Spawning time and place of the Japanese eel. Anguilla japonica, in the North Equatorial Current of the western North Pacific Ocean. Fish. Sci. 67: 1097-1103. 12. Miyama, T., T. Awaji, K. Akitomo, and N. Imasato. 1995. Study of seasonal transport variations in the Indonesian Seas. J. Geophys. Res. 100: 20.517-20,541. 13. Hatayama, T., T. Awaji, and K. Akitomo. 1996. Tidal currents in the Indonesian Seas and their effect on transport and mixing. J. Genphy.f. Re*. 101: 12.353-12.373. 14. Arai, T., D. Limbong, T. Otake, and K. Tsukamoto. 2001. Re- cruitment mechanisms of tropical eels Anguilla spp. and implications for the evolution of oceanic migration in the genus Anguilla. Mar. Ecol. Prog. Ser. 216: 253-264. 15. Sugeha, H. Y., T. Arai, M. J. Miller, D. Limbong, and K. Tsuka- moto. 2001. Inshore migration of the tropical eels, Anguilla spp.. recruiting to the Poigar River estuary on Sulawesi Island. Mar. Ecol. Prog. Ser. 221, 233-243. THE BIOLOGICAL BULLETIN 2003 Subscription Rates Volumes 204-205 *Paid Subscriptions include both print and electronic subscriptions at: www.biolbull.org Institutional* Individual* One year subscription (6 issues - 2 volumes) $280.00 $105.00 Single volume (3 issues) $140.00 $52.50 Single Issues $ 50.00 $20.00 *Surface delivery included in above prices. For prompt delivery, we encourage subscribers outside the U.S. to request airmail service. Airmail Delivery Charge U.S. and Canada: $ 25.00 Mexico: $ 60.00 All other locations: $ 1 00.00 Orders Must Be Prepaid in U.S. Dollars, Check Payable to The Marine Biological Laboratory About The Biological Bulletin ISSN: 0006-3185 Frequency: Bimonthly Number of issues per year: 6 Months of Publication: February, April, June, August, October, December Subscriptions entered for calendar year Volume indexes contained in June and December issues Annual report of the Marine Biological Laboratory contained in August issue Most back issues available Claims handled upon receipt No agency discounts Internet: www.biolbull.org Please address orders to: Wendy Child Subscription Administrator The Biological Bulletin Marine Biological Laboratory 7 MBL Street Woods Hole, MA 02543-1015 U.S.A. Fax: 508-289-7922 Tel: 508-289-7402 Email: wchild@mbl.edu Published by the Marine Biological Laboratory Woods Hole, Massachusetts, 02543 U.S.A. THE BIOLOGICAL BULLETIN (www.biolbull.org) 2003 SUBSCRIPTION FORM (VOLUMES 204-205, 6 ISSUES) All subscriptions run on the calendar year; price includes both print and online journals. (please print) NAME: INSTITUTION: ADDRESS: CITY: STATE: POSTAL CODE: COUNTRY: TELEPHONE: FAX: E-MAIL ADDRESS: Please send me a 2003 subscription to The Biological Bulletin at the rate indicated below: O Individual: $105.00 (6 ISSUES) O Institutional: $280.00 (6 ISSUES) O Individual: $ 52.50 (3 ISSUES) O Institutional: $140.00 (3 ISSUES) Check one: G February, April, June or O August, October, December Please send me the following back issue(s): G Individual: at $20.00 (PER ISSUE) O Institutional: at $50.00 (PER ISSUE) Delivery Options Surface Delivery (Surface delivery is included in the subscription price.) Air delivery (Please add the correct amount to your payment.) a U.S. and Canada: $25.00 O Mexico: $60.00 a All other locations: $100.00 Payment Options Enclosed is my check or U.S. money order for $ payable to The Marine Biological Laboratory Please send me an invoice. (Note: Payment must be received before subscription commences.) Please charge my H VISA, O MasterCard O Discover Card $ Account No.: Exp. Date: Signature: Date: Return this form with your check or credit card information to: Marine Biological Laboratory Subscription Office The Biological Bulletin 7 MBL Street Woods Hole, MA 02543-1015 ARINE IYESOURCES CENTER MARINE BIOLOGICAL LABORATORY WOODS HOLE. MA 02543 (508)289-7700 WWW.MBL.EDU/SERVICES/MRC/INDEX.HTML Animal and Tissue Supply for Education & Research 150 aquatic species available for shipment via online catalog: . http://www.biolbull.org ONLY BX61 MOT RESEARCH MICROSCOPE. Brighter multi-color fluorescence - exquisitely balanced - achieved easier and faster than ever thought possible. DIG with the optimal balance of contrast and resolution. Rock solid stability imazing 3D and ' 'aging. frame with push- button controls and Z-Drive. Automatically moves the stage up or down in increments as fine as 0.01pm. MOTORIZED reflected light illuminator. Automatic control of 6 filter cubes accommodates multi-color fluores- ice with speed OTORIZED sextuple revolving nosepiece with slider slot for Nomarski DIG. MOTORIZED filter wheels. Exchange excitation, emission and transmitted light filters with the press of a button. MOTORIZED condenser. Automatically adjusts condenser for bright- field, darkfield, phase contrast and DIG applications. It's all about imaging. And motorization lets you do things you've never been able to do before with greater ease, speed and affordability. . . ^^..B ROCKET SCIENCE: 800 446 5967 olympusamerica.com/microscopes Cover The background imae on the cover of this issue shows an area ( 1-2 nr) of intertidal microbial mats, 0.5 to 2 cm thick, adjacent to the largest salt works in North America Exportadora de Sal located in Guerrero Negro, Baja California Sur, Mexico. Hy- persaline ponds belonging to the salt works harbor extensive cyanobacterial mats, which are extraordi- narily diverse, complex, and highly organized eco- systems. On p. 160 of this issue, David J. Des Marais reviews his own and others' comprehensive studies of these mats. He asks how such systems respond in a coordinated fashion to cyclical or transient envi- ronmental changes and how they influence sedi- mentation and produce gases. Microbial mats are laminated, and the component microorganisms in the community are localized to layers at specific depths. Layering is visible in the inset on the cover a Nomarski image of a section through the upper 2 mm of a mat. The productivity of the mat is dominated by Microcoleus cltthono- plastes, a filamentous cyanobacterium (the yellow- green region near the top of the section). During the day, photosynthesis by Microcoleus and many other cyanobacteria is intense, and the oxygen generated diffuses downward. But the light is strongly ab- sorbed, and the oxygen is rapidly consumed by heterotrophs. so an aphotic, anoxic zone develops, beginning at a depth of only about 1 .5 mm (the dark region near the bottom of the section). Meanwhile, anaerobic organisms generate H,S, which diffuses upward, but is consumed by photosynthetic bacte- ria (which function in very dim, infrared light) and chemoautotrophs like Beggiatou (the beaded filaments visible adjacent to the dark region). The dark layer thus marks the interface between the diminished concentrations of oxygen and sulfide. At night, photosynthesis ceases, the upper levels of the mat become sulfidic as the oxygen con- centration falls, and motile organisms may move upward. Thus, the position of a microorganism in the mat is determined by many factors, including the steep gradients of light, oxygen, and sulfide, physiological adaptations to changes in those gra- dients, trophic mechanisms, and relationships with other organisms at higher and lower levels in the system. In a related paper in this issue of The Biological Bulletin, John R. Spear and colleagues from the laboratory of Norman R. Pace (p. 168) report on molecular approaches to identifying the compo- nents of mat communities in Guerrero Negro and thus quantifying the extent of diversity. They also characterize, partially, a novel, relatively simple, laminated microbial community that occurs in crys- talline gypsum; this finding documents further the enormous diversity of microorganisms at this site. The articles by Des Marais and Spear el al. are both part of a workshop entitled Outcomes of Genome- Genome Imeractions (p. 155). This meeting was meant to establish links among biogeochemical fac- tors, microbial metabolic processes, maintenance of microbial population structures in nature, and mi- crobial symbioses with multicellular hosts. The workshop was held at Woods Hole, Massachusetts (May 1-3, 2002) and was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory (MBL). The large, background image of the microbial mats //; situ was taken by John R. Spear (University of Colorado, Boulder), and the Nomarsky image was taken by Jack D. Farmer (Arizona State University, Tempe). The composite picture on the cover was produced by Beth Liles at the MBL. THE BIOLOGICAL BULLETIN APRIL 2003 Editor Associate Editors Section Editor Online Editors Editorial Board Editorial Office MICHAEL J. GREENBERG Louis E. BURNETT R. ANDREW CAMERON CHARLES D. DERBY MICHAEL LABARBERA SHINYA INOUE. Inia^ini; and Microscopy JAMES A. BLAKE. Keys to Marine Invertebrates of the Woods Hole Region WILLIAM D. COHEN, Marine Models Electronic Record and Compendia PETER B. ARMSTRONG JOAN CERDA ERNEST S. CHANG THOMAS H. DIETZ RICHARD B. EMLET DAVID EPEL KENNETH M. HALANYCH GREGORY HINKLE NANCY KNOWLTON MAKOTO KOBAYASHI ESTHER M. LEISE DONAL T. MANAHAN MARGARET MCFALL-NGAI MARK W. MILLER TATSUO MOTOKAWA YOSHITAKA NAGAHAMA SHERRY D. PAINTER J. HERBERT WAITE RICHARD K. ZIMMER PAMELA CLAPP HINKLE VICTORIA R. GIBSON CAROL SCHACHINGER WENDY CHILD The Whitney Laboratory. University of Florida Grice Marine Laboratory. College of Charleston California Institute of Technology Georgia State University University of Chicago Marine Biological Laboratory ENSR Marine & Coastal Center. Woods Hole Hunter College, City University of New York University of California, Davis Center of Aquaculture-IRTA. Spain Bodega Marine Lab., University of California, Davis Louisiana State University Oregon Institute of Marine Biology, Univ. of Oregon Hopkins Marine Station. Stanford University Auburn University, Alabama Millennium Pharmaceuticals, Cambridge, Massachusetts Scripps Inst. Oceanography & Smithsonian Tropical Res. Inst. Hiroshima University of Economics, Japan University of North Carolina Greensboro University of Southern California Kewalo Marine Laboratory, University of Hawaii Institute of Neurobiology, University of Puerto Rico Tokyo Institute of Technology, Japan National Institute for Basic Biology, Japan Marine Biomed. Inst.. Univ. of Texas Medical Branch University of California, Santa Barbara University of California, Los Angeles Managing Editor Staff Editor Editorial Associate Subscription & Advertising Administrator Published by MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASSACHUSETTS http://www.biolbull.org Cool L Fast. Reliable. ..., 'Hfl ' ii Launch a Microway Linux cluster and solve your problems in life sciences, structure design, weather, astronomy . . . Run cool. Microway's CoolFlow 1U chassis is designed to provide the best cooling in the industry. Low temperatures result in higher CPU and memory reliability, plus longer cluster life. Take control. Monitor and control your cluster remotely via our secure web-based GUI. Microway's NodeWatch hardware and MCMS cluster management software report on temperatures, voltages, and chassis fans, and let you set thresholds for notification and/or failsafe shutdown if problems occur. Today's HPC market is wilder than ever. Microway's expert design team will guide you through the Whitewater of hardware, software, and storage solutions. Our clusters are competitively priced and delivered on time and their proven reliability yields low total lifetime cost. What's more, our tech support is legendary. Call us to test drive NodeWatch online, arrange a benchmark, and discuss your next Linux cluster with people who speak your language. Visit our website for HPC Times technical news. CoolRak Cabinet Features Xeon XBIade Technology 1.056 Teraflops per 44U cabinet running 3.06 GHz Xeons NodeWatch remote control and monitoring CoolFlow proprietary 1U and 4U chassis Myrinet, GigE, or InfiniBand connectivity ^MDU^ ^ Opteron 'Microway Scientists have been counting on us for over 20 years. SM ^ 508.746.7341 www.microway.com CONTENTS VOLUME 204. No. 2: APRIL 2003 RESEARCH NOTE Maruyama, Tadashi, Euichi Hirose, and Masaharo Ish- ikura Ultraviolet-light-absorbing tunic cells in didemnid as- cidians hosting a symbiotic photo-oxygenic pro- karyote, Prochloron 109 DEVELOPMENT AND REPRODUCTION Carpizo-Ituarte, Eugenio, and Michael G. Hadfield Transcription and translation inhibitors permit meta- morphosis up to radicle formation in the serpulid polychaete Hydroid.es elf gam Haswell 114 NEUROBIOLOGY AND BEHAVIOR Garm, A., E. Hallberg, and J. T. H0eg Role of maxilla 2 and its setae during feeding in the shrimp Palaemon adspersus (Crustacea: Decapoda) . . . . 126 PHYSIOLOGY AND BIOMECHANICS Clode, Peta L., and Alan T. Marshall Variation in skeletal microstructure of the coral Gal- axea fascicularis: effects of an aquarium environment and preparatory techniques 138 Clode, Peta L., and Alan T. Marshall Skeletal microstructure of Galaxea fascicularis exsert septa: a high-resolution SEM study 1 46 OUTCOMES OF GENOME-GENOME INTERACTIONS Sogin, Mitchell, and Diana E. Jennings Introduction 159 Des Marais, David J. Biogeochemistry of hypersaline microbial mats illus- trates the dynamics of modern microbial ecosystems and the early evolution of the biosphere 1 60 Spear, John R., Ruth E. Ley, Alicia B. Berger, and Norman R. Pace Complexity in natural microbial ecosystems: the Guerrero Negro experience 168 Vallino, Joseph J. Modeling microbial consortiums as distributed met- abolic networks 174 Edwards, Katrina J., Wolfgang Bach, and Daniel R. Rogers Geomicrobiology of the ocean crust: a role for che- moautotrophic Fe-bacteria 180 Teske, Andreas, Ashita Dhillon, and Mitchell L. Sogin Genomic markers of ancient anaerobic microbial pathways: sulfate reduction, methanogenesis, and methane oxidation 186 Fuhrman, J. A., and M. Schwalbach Viral influence on aquatic bacterial communities. . . 192 Polz, Martin F., Stefan Bertilsson, Silvia G. Acinas, and Dana Hunt A(r)Ray of hope in analysis of function and diversity of microbial communities 196 Foster, Jamie S., Robert J. Palmer, Jr., and Paul E. Kolenbrander Human oral cavity as a model for the study of ge- nome-genome interactions 200 Amaral Zettler, Linda A., Mark A. Messerli, Abby D. Laatsch, Peter J. S. Smith, and Mitchell L. Sogin From genes to genomes: beyond biodiversity in Spain's Rio Tin to 205 Cast, Rebecca J., David J. Beaudoin, and David A. Caron Isolation of symbiotically expressed genes from the dinoflagellate symbiont of the solitary radiolarian Thalassicolla nucltata 210 Bonfante, P. Plants, mycorrhizal fungi and endobacteria: a dialog among cells and genomes 215 Wernegreen, Jennifer J., Patrick H. Degnan, Adam B. Lazarus, Carmen Palacios, and Seth R. Bordenstein Genome evolution in an insect cell: distinct features of an ant-bacterial partnership 221 LIST OF PARTICPANTS. . 232 THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is published six times a year by the Marine Biological Laboratory. 7 MBL Street. Woods Hole. Massachusetts 02543. Subscriptions and similar matter should be addressed to Subscription Administrator. THE BIOLOGICAL BULLETIN. Marine Biological Laboratory. 7 MBL Street. Woods Hole, Massachusetts 02543. Subscription includes both print and online journals. Subscription per year (si\ issues, two volumes): S280 for libraries; SI 05 for individuals. Subscription per volume (three issues): $140 for libraries: $52.50 for individuals. Back and single issues (subject to availability): $50 for libraries: $20 for individuals. Communications relative to manuscripts should be sent to Michael J. Greenberg. Editor-in-Chief, or Pamela Clapp Hinkle, Managing Editor, at the Marine Biological Laboratory. 7 MBL Street. Woods Hole, Massachusetts 02543. Telephone: (508) 2S9-7I49. FAX: 508-289-7922. E-mail: pclapp@mbl.cdu. http://www.biolbull.org THE BIOLOGICAL BULLETIN is indexed in bibliographic services including Index Medicus and MED- LINE. Chemical Abstracts, Current Contents, Elsevier BIOBASE/Current Awareness in Biological Sciences, and Geo Abstracts. Printed on acid free paper, effective with Volume 180, Issue 1, 1991. POSTMASTER: Send address changes to THE BIOLOGICAL BULLETIN, Marine Biological Laboratory, 7 MBL Street, Woods Hole, MA 02543. Copyright 2003, by the Marine Biological Laboratory Periodicals postage paid at Woods Hole, MA, and additional mailing offices. ISSN 0006-3185 INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts outstanding original research reports of general interest to biologists throughout the world. Papers are usually of intermediate length (10-40 manuscript pagesl. A limited number of solicited review papers may be accepted after formal review. A paper will usually appear within four months after its acceptance. Very short, especially topical papers (less than 9 manuscript pages including tables, figures, and bibliography) will be published in a separate section entitled "Research Notes." A Research Note in The Biological Bulletin follows the format of similar notes in Nature. It should open with a summary paragraph of 150 to 200 words comprising the introduction and the conclusions. The rest of the text should continue on without subheadings, and there should be no more than 30 references. References should be referred to in the text by number, and listed in the Literature Cited section in the order that they appear in the text. Unlike references in Nature, references in the Research Notes section should conform in punctuation and arrangement to the style of recent issues of The Biological Bulletin. Materials and Methods should be incorpo- rated into appropriate figure legends. See the article by Loh- mumi et al. (October 1990, Vol. 179: 214-218) for sample style. A Research Note will usually appear within two months after its acceptance. The Editorial Board requests that regular manuscripts con- form to the requirements set below: those manuscripts that do not conform will be returned to authors for correction before review. 1 . Manuscripts. Manuscripts, including figures, should be submitted in quadruplicate, with the originals clearly marked. ( Xerox copies of photographs are not acceptable for review pur- poses.) If possible, please include an electronic copy of the text of the manuscript. Label the disk with the name of the first author and the name and version of the wordprocessing software used to create the file. If the file was not created in some version of Microsoft Word, save the text in rich text format (rtf). The sub- mission letter accompanying the manuscript should include a telephone number, a FAX number, and (if possible) an E-mail address for the corresponding author. The original manuscript must be typed in no smaller than 12 pitch or 10 point, using double spacing (i/icli/diiii; figure legends, footnotes, bibliography, etc.) on one side of 16- or 20-lb. bond paper, 8 by 11 inches. Please, no right justification. 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 I inch (2.5 cm). Manuscripts should conform to the Council of Biology Edi- tors Sr\'le Manual, 5th Edition (Council of Biology Editors. 1983) and to American spelling. LInusual 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. Abstract (of no more than 200 words). Introduction. Materials and Methods. Re- sults, 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. Title page. The title page consists of a condensed title or running head of no more than 35 letters and spaces, the manuscript title, authors' names and appropriate addresses, and footnotes listing present addresses, acknowledgments or contribution num- bers, and explanation of unusual abbreviations. 3. Figures. The dimensions of the printed page. 7 by 9 inches, should be kept in mind in preparing figures for publication. We recommend that figures be about 1 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 illustra- tion 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 and cited, in order, in the text. The author's name and an arrow indicating orientation should appear on the reverse side of all figures. Digital art: The Biological Bulletin will accept figures sub- mitted in electronic form; however, digital art must conform to the following guidelines. Authors who create digital images are wholly responsible for the quality of their material, including color and halftone accuracy. Format. Acceptable graphic formats are TIFF and EPS. Color submissions must be in EPS format, saved in CMKY mode. Sojhvare. Preferred software is Adobe Illustrator or Adobe Photoshop for the Mac and Adobe Photoshop for Windows. Spe- cific instructions for artwork created with various software pro- grams are available on the Web at the Digital Art Information Site maintained by Cadmus Professional Communications at http:// cpc.cadmus.com/da/ Resolution. The minimum requirements for resolution are 1200 DPI for line art and 300 for halftones. Size. All digital artwork must be submitted at its actual printed size so that no scaling is necessary. Miiltipunel figures. Figures consisting of individual parts (e.g., panels A. B, C) must be assembled into final format and submitted as one file. Hard cop\. Files must be accompanied by hard copy for use in case the electronic version is unusable. Disk identification. Disks must be clearly labeled with the following information: author name and manuscript number; for- mat (PC or Macintosh); name and version of software used. Color: The Biological Bulletin will publish color figures and plates, but must bill authors for the actual additional cost of printing in color. The process is expensive, so authors with more than one color image should consistent with editorial concerns, especially citation of figures in order combine them into a single plate to reduce the expense. On request, when supplied with a copy of a color illustration, the editorial staff will provide a pre-publi- cation estimate of the printing cost. 4. 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 high cost of setting tabular material in type, authors are asked to limit such material as much as possible. Tables, with their headings and footnotes, should be typed on separate sheets, num- bered with consecutive Arabic numerals, and placed after the Literature Cited. Figure legends should contain enough informa- tion to make the figure intelligible separate from the text. Legends should be typed double spaced, with consecutive Arabic numbers, on a separate sheet at the end of the paper. Footnotes should be limited to authors' current addresses, acknowledgments or contri- bution numbers, and explanation ot unusual abbreviations. All such footnotes should appear on the title page. Footnotes are not normally permitted in the body of the text. 5. Literature cited. In the text, literature should be cited by the Harvard system, with papers by more than two authors cited as Jones el al. 1980. Personal communications and material in prep- aration or in press should be cited in the text only, with author's initials and institutions, unless the material has been formally accepted and a volume number can be supplied. The list of references following the text should 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 bio- logical 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 using 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 addi- tional 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 Vfsinclafjelags Islendinga without abbreviation. F. All single word journal titles in full (e.g.. Veliger, Ecol- ogy. Brain). G. The order of abbreviated components should be the same 6. Reprints, page proofs, and charges. Authors may pur- as the word order of the complete title (i.e., Proc. and Trans. chase reprints in lots of 100. Forms for placing reprint orders are placed where they appear, not transposed as in some BIOLOGICAL sent with page proofs. Reprints normally will be delivered about 2 ABSTRACTS listings). to 3 months after the issue date. Authors will receive page proofs of articles shortly before publication. They will be charged the H. A few well-known international journals in their preferred current cost of printers' time for corrections to these (other than forms rather than WORLD LIST or USASI usage (e.g.. Nature. corrections of printers' or editors' errors). Other than these charges Science, Evolution NOT Nature, Land., Science, N.Y.; Evolution, for authors' alterations. The Biological Bulletin does not have Lancaster. Pa.) page charges. Reference: Bio/. Bull. 204: KW-113. (April 2003) 2003 Marine Biological Laboratory Ultraviolet-Light-Absorbing Tunic Cells in Didemnid Ascidians Hosting a Symbiotic Photo-oxygenic Prokaryote, Prochloron TADASHI MARUYAMA K *-t, EUICHI HIROSE 2 , AND MASAHARU ISHIKURA 1 ' Marine Biotechnology Institute. Kamaishi Laboratories, Heita 3-75-1, Kainuishi. Iwate 026-0001, Japan: and 'Department of Chemistry, Biology, and Marine Sciences, Faculty of Science, University of the Ryukyus. Nishihara, Okinawa, 903-0213, Japan Coral reef invertebrates that host phototrophic svmbionts arc thought to protect themselves and their symbionts with mycosporine-like amino acids (MAAs) UV-absorbing sub- stances that act as sunscreens (Dunlap. W. C., and J. M. Stuck. 1998. J. Phycol. 34: 418-430). However, the histo- logical distribution of MAAs in the host tissues has not vet been visualised. We have localized the UV-absorbing sub- stances in the tissues of t\vo colonial didemnid ascidians Lissoclinum patella and Diplosoma sp. that contain the symbiotic photo-oxygenic prokaryote Prochloron sp. Cross-sections of unfixed tissue from these ascidians were examined by UV-light microscopy at 320 or 330 nm, wavelengths at which UV light is absorbed bv MAAs. Within the tunic, the gelatinous integument of the colony, UV light was exclusively absorbed by a particular type of cell, the tunic bladder cell. Tunic bladder cells with strong UV absorption were denser in the upper tunic, which lies over a colony's zooids. than in the basal tunic underlying the zooid. In the upper tunic, those cells with strong UV absorption were most dense near the surface. The tunic bladder cell is highly vacuolated. and the vacuole contains strong acid, which destabilizes MAAs. Furthermore, the UV-absorbing portion of tunic bladder cells seemed to be cup-shaped, indicating thai the MAAs arc not localized in the vacuole. but in the cvtoplasm. Received 21 June 2002; accepted 22 January 2002. * To whom correspondence should be addressed. E-mail: tadashim@jamstec.go.jp $ Current address: Marine Ecosystem Research Department. Japan Ma- rine Science and Technology Center (JAMSTEC). Natsushima 2-15, Yo- kosuka, Kanagawa 237-0061. Japan. These results strongly suggest that didemnid ascidians accumulate MAAs in tunic bladder cells as a protection against UV radiation. Ultraviolet (UV) radiation poses severe problems for organisms in tropical marine environments, because the path for sunlight through the atmosphere is shorter ( 1 ) and the seawater is clearer (2). Since the discovery of UV-absorbing substances in marine organisms in the Great Barrier Reef (3), it has been thought that many coral reef invertebrates hosting symbiotic microalgae or photo-oxygenic prokaryotes protect themselves and their symbionts by using, as sunscreens, mycosporine-like amino acids (MAAs) which absorb UV (for review see 4, 5, 6). The MAAs (A max 310-360 nm) have strong absorption in the UVA (320-400 nm) and UVB (280-320 nm) range, and some, porphyra-334 (7) and shinorine (8), have been shown to be photochemically stable to UV irradiation. Un- der ultraviolet radiation (UVR), they neither fluoresce nor produce radical intermediates (9). The MAAs are transpar- ent to visible light (i.e.. photosynthetically active radiation, PAR), which is required by the phototrophic symbionts of many invertebrates whose tissues are also exposed to the UVR. UVR induces an increase in the MAA content of zoo- xanthellate corals, microalgae, and seaweeds (for review see 6). Corals and some red macroalgae living in shallow water have greater MAA concentrations than those from deeper or shadowed waters (for review see 4, 6). The photosynthetic ability of symbionts isolated from their host tissues is se- verely damaged by UVR, whereas that of the symbionts within the host is more resistant (10, 11; for review see 6). 109 10 T. MARUYAMA ET AL The MAAs are thought to be synthesized by the shikimie acid pathway, which has not been reported in metazoan cells (12, 13; for review see 6). It is thought that the host invertebrates acquire MAAs from their diet or from their symbionts (13; for review see 4). In summary, MAAs, which may be synthesized by the symbiont upon stimulation by UV radiation, function as a sunscreen for the symbionts as well as the host. In tropical waters, some didemnid ascidians host the symbiotic photo-oxygenic prokaryote Prochloron (for re- view see 14). The symbiont cells are usually sequestered within the host colony under an integument, the ascidian tunic, which is transparent to visible light (PAR), but which absorbs maximally at about 320 nm (10). In a symbiotic didemnid, Lissoclimun patella. 93%-98% of UV light (312 nm) was reduced by a slice of its integument tunic, thick- ness 0.5-1.0 mm. whereas 83%-90% of visible light was passed through the slice (10). The tunic of L. patella also contains MAAs, such as shinorine, mycosporine-glycine, and palythine (10, 15, 16). Although the tissue content and chemistry of MAAs have been studied intensively in L. patella and a variety of other organisms (5). the detailed distribution of these MAAs in the ascidian tu- nic whether they are localized in certain tunic cells or in the tunic matrix remains to be determined. To ap- proach this problem, we studied the histological distri- bution of UV-absorbing substances in the tunic of two photo-symbiotic didemnid ascidians hosting Prochloron; we used UV-light microscopy at 320 or 330 nm. the wavelengths at which MAAs typically show maximal absorption. A sheet-like colony of didemnid ascidians usually con- sists of three layers: a layer of zooids and cloacal cavity (ZC in Fig. 1 ) sandwiched between two layers of tunic (UT and BT in Fig. 1 ). Both the upper and basal layers of tunic are transparent to visible light (Fig. 2A; also see fig. 3 in reference 10). In photo-symbiotic didemnid ascidians. the Prochloron cells are exclusively distributed within the clo- acal cavity, i.e., outside the host tissue (CC in Fig. 1 ). In one of the host ascidians, Diplosoma sp., UV microscopy re- vealed a strong layer of UV absorbance in the upper tunic and a similar layer with less UV absorption in the basal tunic (Fig. 2B). The strength of the UV absoiption seemed to depend upon the density of UV-absorbing objects in the layer. Moreover, the UV absorption of each object seemed to be stronger in the layer of the upper tunic than in the basal tunic (Fig. 2B). Similar UV-absorbing layers were also observed in the other host ascidian studied, L. patella (data not shown). These results agree well with a previous report that MAA concentration is higher in the upper tunic than in the basal tunic in L. patella ( 10). Prochloron cells occupy- ing the cloacal cavity in the middle layer of Diplosoma sp. (ZC layer in Fig. 2) also absorbed UV light. Observation of BA UT ZC BT : Prochloron o : tunic bladder cell Figure 1. Schematic drawing of a cross-section of a colony of a symbiotic didemnid ascidian. BA. branchial aperture; BT. basal tunic layer; CA, cloacal aperture; CC. cloacal cavity; T, tunic (lightly shaded area); UT. upper tunic layer: Z, zooid (more darkly shaded area); ZC, layer of zooid and cloacal cavity. the upper tunic at higher magnification revealed that each UV-absorbing object was round or cup-shaped, with a di- ameter (of the openings of the cup) of 30 13 /urn (Fig. 3). When cross-sections of fixed, embedded, and stained ascid- ian tissues were observed, the tunic was seen to contain many bladder cells, about 50 jam in diameter; moreover, these cells correspond to the UV-absorbing objects (Fig. 4 A). This obviously indicates that the MAAs are not local- ized in the matrix of the tunic, but in the tunic bladder cells. The bladder cell is characterized by a large vacuole occu- pying most of the cytoplasm (Fig. 4B). Such bladder cells are common in the tunic of many ascidians in the family Didemnidae (17), and their vacuoles usually contain a strongly acidic fluid (18). The cup shape of the UV-absorb- ing objects (Fig. 3) indicates that the UV-absorbing sub- stance is contained in the cytoplasm of the bladder cell, and that the large vacuole lacks this substance. Because the vacuoles of the bladder cell are known to contain strong acid (17. 18), and because MAAs are unstable in acidic conditions (19). it is reasonable that these MAAs are local- ized in the cytoplasm. We have previously reported that the tunic of L. patella contains shinorine. one of the most common MAAs, as a major UV-absorbing substance, together with two minor MAAs, mycosporine-glycine and palythine (10); in Austra- lia, however, L. patella was reported to contain shinorine and mycosporine-glycine (15), or only mycosporine-glycine ( 16). In this study, we freeze-dried Diplosoma sp., extracted the material with a methanol-tetrahvdrofuran (4:1) mixture. UV-ABSORBING CELLS IN DIDEMNID ASCIDIANS HOSTING PROCHI.ORON UT ZC BT UV-UT UV-BT Figure 2. Visible- and ultraviolet-light micrographs of a cross-section of a living colonial ascidian. Dipln.tiima sp. The ascidian sample was collected at Akajima, an island in the Ryukyu Archipelago, Japan. Thin slices of living specimens were cut with a razorblade by hand. They were placed on a quartz slide and immersed in tillered seawater surrounded by glycerol, covered with a quartz coverslip. and observed under a light microscope with Nikon Fluor objectives (10X NA = 0.5, 20x NA = 0.75). The glass eyepiece and condenser lens were removed, because they absorb UV light. An interference filter (bandpass) for 319 II nm (median 50% maximum transparency) or for 331 5 nm (median 50% maximum transparency) was inserted between a reflex camera body (Olympus OM-2) mounted on the microscope and the objective lens. The light source was a UV lamp (Vilber-Lourmat T15-N, France: or Toshiba F6T5 UV-B lamp, Japan). Black-and-white film (Fuji Neopan F) was used to record the images. Because the focal plane for the UV light is different from that for visible light, it was determined before the experiment by making test exposures. For visible-light microscopy, the interference UV bandpass filter was removed from the light path, and an UV-opaque filter was inserted between the light source and the specimen. (A) Visible-light micrograph; (B) UV-light micrograph (320 nm). UT, upper tunic layer: BT. basal tunic layer; ZC, zooid and cloacal cavity layer; UV-UT, UV-light-absorbing zone in the upper tunic; UV-BT, UV-light-absorbing zone in the basal tunic. Scale. 200 fum. and analyzed the extracts for MAAs by reversed-phase liquid chromatography according to Dionosio-Sese ci al. (10); the MAAs from another ascidian, Halocyntliiu mretzi, were used as references (20). The major MAA (about 94%) was shinorine. and the minor MAA was palythine. Because these substances are soluble in water (21 ), MAAs probably exist as soluble components in the cytoplasm of the bladder cells. Isolated Prochloron cells from L. patella contained shinorine. which was also dominant in the host tunic (10). MAAs are thought to be synthesized by the shikimic acid pathway, which has not been reported in metazoans cells (12, 13). In symbiotic didemnid ascidians, the MAAs are probably synthesized in the symbiont algal cells, then trans- ferred to the host tissue, and accumulated in the cytoplasm of the bladder cells mostly in the upper tunic. Because the UV-absorbing layer in the upper tunic is transparent to PAR but absorbs UVR. the underlying host tissues and the symbionts are protected from the UVR, but they can still receive PAR needed for photosynthesis. The Figure 3. Higher magnification UV micrograph (320 nm): cross-sec- tion of a UV-absorbing zone in the upper tunic of a living colony of Diplosoma sp. The cytoplasm of tunic bladder cells was observed as dark, round or cup-shaped. UV-absorbing objects (arrowheads). Scale. 100 /xni. For UV microscopy, see the legend of Figure 2. 112 T. MARUYAMA ET AL. UT ZC BT Figure 4. (A) Light micrograph of a cross-section of a colony of Diplo.wma sp. ( 10% formalin-seawater fixation; I0-/u,m-thick paraffin section stained with hematoxylin and eosin). Scale, 200 ,uin. (B) Higher magnification light micrograph of a tunic bladder cell (2.5% glutaraldehyde prefixation and 1% OsO 4 postfix- ation; l-/Min-thick resin section stained with toluidine blue). Scale. 10 /j.m. (C and D) Schematic representations of A and B. respectively. CC, cloacal cavity; N, nucleus of tunic bladder cell; TB. tunic bladder cell; TC, other tunic cells; V, vacuole of tunic bladder cell; Z. zooid; small structures (dots) in CC, cells of Prochloron sp. mechanisms underlying translocation of MAAs and their uptake by the tunic bladder cells remain to be elucidated. Acknowledgments We are grateful to Prof. R. A. Lewin for discussion. Staff members at Akajima Marine Science Laboratory are ac- knowledged for their help during sample collection. This work was performed as a part of The Industrial Science and Technology Project, Technological Development of Biolog- ical Resources in Bioconsortia, supported by the New En- ergy and Industrial Technology Development Organization (NEDO). The present study is partly supported by a JSPS Grant-in- Aid. Literature Cited 1. Barker, K. S., R. C. Smith, and A. E. S. Green. 1980. Middle ultraviolet radiation reaching the ocean surface. Photochem. Photobiol. 32: 367-374. UV-ABSORBING CELLS IN DIDEMNID ASCIDIANS HOSTING PROCHLORON 113 2. Fleischmann, E. M. 1989. The measurement and penetration of ul- traviolet radiation into tropical marine water. Limnol. Oceanogr. 34: 1623-1629. 3. Shibata, K. 1969. Pigments and a UV-absorbing substance in corals and a blue-green alga living in the Great Barrier Reef. Plant Cell Physio/. 10: 325-335. 4 Dunlap, W. C., and J. M. Shick. 1998. Ultraviolet radiation-absorb- ing mycosporine-like amino acids in coral reef organisms: a biochem- ical and environmental perspective. J. Pliycol. 34: 418-430. 5. Karentz, D. 2001. Chemical defenses of marine organisms against solar radiation exposure: UV-absorbing mycosporine-like amino acids and scytonemin. Pp. 481-520 in Marine Chemical Ecology, J. B. McClintock and B. J. Baker, eds. CRC Press, Boca Raton. FL. 6. Shick, J. M., and W. C. Dunlap. 2002. Mycosporine-like amino acids and related gadusols: biosynthesis, accumulation, and UV-protec- tive functions in aquatic organisms. Anna. Rev. Physiol. 64: 223-262. 7. Conde, F. R., M. S. Churio, and C. M. Previtali. 2000. The pho- toprotector mechanism of mycosporine-like amino acids. Excited-state properties and photostability of porphyra-334 in aqueous solution. J. Photochem. Photobiol. B Biol. 56: 139-144 8. Adams, N. L., and J. M. Shick. 1996. Mycosporine-like amino acids provide protection against ultraviolet radiation in eggs of the green sea urchin Strongylocentrotus droebachiensis. Photochem. Photobiol. 64: 149-158. 9 Shick, J. M., W. C. Dunlap, and G. R. Buettner. 2000. Ultraviolet (UV) protection in marine organisms II. Biosynthesis, accumulation, and sunscreening function of mycosporine-like amino acids. Pp. 215- 228 in Free Radicals in Chemistry, Biology and Medicine, T. Yo- shikawa, S. Toyokuni. Y. Yamamoto, and Y. Naito eds. OICA Inter- national, London. 10. Dionisio-Sese, M. L., M. Ishikura, and T. Maruyama. 1997. UV- absorbing substances in the tunic of a colonial ascidian protect its symbiont, Prochloron sp., from damage by UV-B radiation. Mar. Biol. 128: 455-461. 11. Ishikura, M., C. Kato, and T. Maruyama. 1997. UV-absorbing substances in zooxanthellate and azooxanthellate clams. Mar. Biol. 128: 649-655. 12. Favre-Bonvin, J., ,|. Bernillon, N. Salin, and N. Arpin. 1987. Biosynthesis of mycosporines: mycosporine glutaminol in Trichoth- ecium rosenm. Phytochemistry 26: 2509-2514. 13. Shick, J. M., S. Romaine-Lioud, C. Ferrier-Pages, and J.-P. Gat- tuso. 1999. Ultraviolet-B radiation stimulates shikimate pathway- dependent accumulation of mycosporine-like amino acids in the coral Stylophora pistillata despite decreases in its population of symbiotic dinoflagellates. Limnol. Oceiinogr. 44: 1667-1682. 14. Lewin, R. A., and L. Cheng, eds. 1989. Prochloron: A Microbial Enigma. Chapman and Hall, New York. 15. Dunlap, W. C., and Y. Yamamoto. 1995. Small-molecule antioxi- dants in marine organisms: antioxidant activity of mycosporine-gly- cine. Com/). Binclwin. Plmiol. 112B: 105-114. 16. Lesser, M. P., and W. R. Stochaj. 1990. Photoadaptation and protection against active forms of oxygen in the symbiotic procaryote Prochloron sp. and its ascidian host. Appl. Environ. Microbiol. 56: 1530-1535. 17. Hirose. E. 2001. Acid containers and cellular networks in the ascid- ian tunic with special remarks on the ascidian phylogeny. Zoo/. Sci. 18: 723-731. 18 Hirose, E., H. Yamashiro, and Y. Mori. 2001. Properties of tunic acid in the ascidian Phallusia nigra (Ascidiidae, Phlebobranchia). Zoo/. Sa. 18: 309-314. 19. Lemoyne, F., J. Bernillon, J. Favre-Bonvin, M. L. Bouillant, and N. Arpin. 1985. Occurrence and characteristics of amino alcohols and cyclohexenone, components of fungal mycosporines. Z. Natur- forsc/i. 40c: 612-616. 20. Nakamura, H., J. Kobayashi, and Y. Hirata. 1982. Separation of mycosporine-like amino acids in marine organisms using reversed- phase high-performance liquid chromatography. J. Chromatogr. 250: 113-118. 2 1 . Ito, S., and Y. Hirata. 1977. Isolation and structure of a mycospo- rine from the zoanthid Palythoa tuberculosa. Tetrahedron Lett. 28: 2429-2430. Reference: Biol. Bull. 204: 1 14-125. (April 21 10?) 2003 Marine Biological Laboratory Transcription and Translation Inhibitors Permit Metamorphosis up to Radiole Formation in the Serpulid Polychaete Hydroides elegans Haswell EUGENIO CARPIZO-ITUARTE 1 2 AND MICHAEL G. HADFIELD Kewalo Marine Laboratory and Department of Zoology. University of Hawaii. 41 Almi St.. Honolulu. Hawaii 96813 Abstract. Settlement and metamorphosis in most well- studied marine invertebrates are rapid processes, triggered by external cues. How this initial environmentally mediated response is transduced into morphogenetic events that cul- minate in the formation of a functional juvenile is still not well understood for any marine invertebrate. The response of larvae of the serpulid polychaete Hydroides elegans to inhibitors of mRNA and protein synthesis was examined to determine if metamorphosis requires these molecular pro- cesses. Competent larvae of H. elegant were induced to metamorphose by exposing them to a bacterial film or a 3-h pulse of 10 mM CsCl in the presence of the gene-transcrip- tion inhibitor DRB (5,6-dichloro-l-/3-D-ribofuranosylben- zimida/ole) or the translation inhibitor emetine. When in- duced to metamorphose in the presence of either inhibitor, larvae of H. elegans progressed through metamorphosis to the point at which branchial radicles start to develop. DRB and emetine inhibited the incorporation of radiolabeled uri- dine into RNA and radiolabeled methionine into peptides, respectively, indicating that they were effective in blocking the appropriate syntheses. Taken together, these results in- dicate that the induction of metamorphosis in H. elegans does not require de novo transcription or translation, and that the form of the juvenile worm is achieved in two phases. During the first phase, larvae respond to the inducer by attaching to the substratum, secreting a primary tube, resorbing the prototroch cilia, undergoing caudal elonga- Received 16 April 2002; accepted 4 December 20(12 1 Present address: Institute de Investigaciones Oceanologicas. UABC. Apdo. Postal 453, Ensenada. B.C. Mexico 22800. To whom correspondence should be addressed. E-mail ecarpizo@uabc.mx Abbreviations: DRB. 5,6-dichloro- 1 -/3-D-nboturanosylbenzimida/ole. tion. and differentiating the collar: once the collar is formed, they begin secreting the secondary, calcified tube. During the second phase, the small worm develops branchial radi- oles and begins to grow, requiring new mRNA and protein syntheses. Introduction Complex life histories in which a larval stage undergoes metamorphosis to achieve a juvenile form are common throughout the animal kingdom, including at least 1? phyla of marine invertebrates (Strathmann. 1493). Well-known examples of this postembryonic transformation are amphib- ians and insects, which have been used as model systems for vertebrates and invertebrates for decades (see: Gilbert et , 2 min). and a 5()-/il sample of the supernatant fluid was recovered to measure the ra- dioactivity remaining in it. The remaining supernate was removed from the tube, and the larvae were washed once with ASW and resuspended in 500 /id of 50 mM sodium phosphate buffer. pH 7.2. The larval tissue was lysed by sonication for 30 s in a Branson Sonifier (model 450. Branson Ultrasonics Corporation). The sample was re- moved from the sonicator every 10 s and put on ice for 20 s to prevent overheating of the tissue. TCA precipitation was carried out according to the protocol described by Ausubel ct ul. (1987) to measure incorporation of "S-methionine into newly synthesized proteins. The sonicated sample was centrifuged ( 15.120 x #, 2 min) at RT, and the supernatant fluid was transferred to a clean 1.5-ml microtube (Eppen- dorf). To a 50-/U.1 sample of this supernate was added 0.5 ml of bovine serum albumin (0.1 mg/ml) containing 0.02% NaN, followed by 0.5 ml of ice-cold 20% TCA, and the suspension was incubated on ice for 30 min. After incuba- tion, the suspension was filtered under vacuum onto a mi- crofiber filter (Gelman, Type E), washed twice with ice-cold METAMORPHOSIS AND GENE EXPRESSION 117 10% TCA and twice with 100% ETOH. and the filters were allowed to dry for 30-45 min at RT. To quantify the incorporation of 35 S-methionine. radioactivity was mea- sured by introducing the fiberglass filters with the TCA precipitate into vials containing 5 ml of scintillation fluid (ScintiSafe, Econo 2; Fisher Scientific). Radioactivity in the vials was measured in a Beckman LS 7000 liquid scintilla- tion system or a Packard liquid scintillation analyzer 1900 TR. Results were expressed as DPM/fj,g of protein. The protein concentration in the remaining sample (250-450 jul) was measured using the spectrophotometric method described by Whitaker and Granum (1980). Statistical analysis of metamorphosis The proportion of larvae that underwent metamorphosis was determined, and these values were arcsine-transformed to estimate statistical differences among treatments using one-way ANOVAs or Kniskal-Wallis ANOVAs of ranks when equal-variance tests failed. Pairwise multiple compar- isons were tested using the Student-Newman-Keuls test or Dunnett's method when all treatments were compared to each other, or Bonferroni's method when treatments were compared to a control. All statistical tests were conducted with the aid of SigmaStat software (SPSS Inc.). Results Effects of the transcription inhibitor DRB on metamorphosis When competent larvae of Hydroides elegans were in- duced to metamorphose with a bacterial biofilm or a 3-h pulse of CsCl in the presence of DRB. they completed metamorphosis of the larval body after 3-4 h but did not develop branchial radioles (Fig. 1 ). The minimum effective concentration of DRB to prevent branchial radicle forma- tion in all the larvae tested with either inducer was 10 ^M (Figs. 2, 3; Bonferroni test, P < 0.05). Lower concentra- tions of DRB tested either had no effect (0. 1 ;u,M, Figs. 2. 3) or showed only a slight, but significant (Bonferroni test, P < 0.05), reduction in the proportion of larvae that com- pleted formation of branchial radioles (1 juA/, Fig. 2). Preincubation of competent larvae of H. elegans in 10 H.M DRB for up to 9 h before addition of an inducer (biofilm or CsCl) did not inhibit metamorphosis (Figs. 4, 5). In fact, larvae responded significantly faster (Student-Newman- Keuls test. P < 0.05) to biofilm in the presence of DRB than in its absence (Fig. 4A). a difference that was reduced after 14 h of incubation when the experiment was stopped (Fig. 4B). Preincubation in 10 p.M DRB for up to 9 h did not stop the larvae from metamorphosing in response to a pulse of CsCl, but a significant reduction in the percentage that metamorphosed was observed (Bonferroni's method, P < 0.05, Fig. 5). Effects of the protein-synthesis inhibitor emetine on metamorphosis Competent larvae of H. elegans metamorphosed when exposed to a biofilm in the presence of 1 ju,A/ emetine (Fig. 6), with no significant reduction in the proportion of larvae that metamorphosed after 3 h compared with induction by a biofilm alone. The larvae progressed in metamorphosis to the point at which branchial radioles begin to develop: concentrations of emetine as low as 250 nM were effective in preventing the formation of these structures. Preincuba- tion for 1 h in 1 p.M emetine prior to exposure to a biofilm did not curtail metamorphosis (Fig. 7), but the proportion of larvae that had metamorphosed after 3 h was slightly re- duced. Results similar to the ones observed when a biofilm was used as inducer were obtained when larvae were stimulated to metamorphose with a 3-h pulse of 10 mM CsCl in the presence of emetine (Fig. 8). The proportion of larvae that metamorphosed was similar in the presence and absence of 1 iJiM emetine, and the proportions that metamorphosed in these treatments were similar to those obtained in response to a bacterial biofilm. Labeling RNA and proteins in vivo Incorporation of [5,6 3 H]uridine into RNA by competent larvae of H. elegans in the presence of DRB varied with concentration and period of exposure. In comparison to incorporation without an RNA-synthesis inhibitor, [5.6 H]uridine incorporation was reduced up to 80% after an incubation period of 2.25 h in 70 /u.M DRB (Fig. 9A). Even with the lowest concentration of DRB tested (10 ;uM), incorporation of [5,6 3 H]uridine was reduced by 60% (Fig. 9A). When larvae were exposed to 10 fj.M DRB for 24 h before addition of [5,6 3 H]uridine, incorporation of [5.6 'Hjuridine was reduced up to 88% in comparison to that in larvae incubated in the absence of DRB (Fig. 9B). A sub- stantial decrease in incorporation of [5.6 ^Hjuridine, up to 42%, was evident after the first 3 h of exposure to DRB (Fig. 9B). Spontaneous metamorphosis was 3.2% and 6% in the two experiments in control larvae kept in conditions equivalent to those of larvae incubated in the presence of DRB. Synthesis of new proteins in larvae of H. elegans exposed to emetine was dependent on the concentration and length of exposure to the inhibitor. When larvae were exposed to different concentrations of emetine for 4 h, a decrease in ^S-methionine incorporation was observed. A reduction of 41% in incorporation of 35 S-methionine into proteins 118 E. CARPIZO-ITUARTE AND M.G. HADFIELD R I" iM^-'iP C/ Figure 1. Metamorphosis in Hydroides elegans in the absence or presence of an inhibitor of macromolecular synthesis. (A) A worm newly metamorphosed in response to a bacteria] biotilm. (B) A worm in which metamorphosis was induced in the presence of the translation inhibitor emetine. (C) The posterior section of the tube of the worm shown in B. Newly metamorphosed worms exposed to induction cues in the presence of the trancription inhibitor DRB have the morphology shown in B. br. branchial radicles; c, collar, pt. primary tube; t, secondary calcareous tube. Scale bar = 20 fj.m for all micrographs. occurred with exposure to 250 nM emetine, and a reduction of up to 87% occurred in response to 1 jj.M emetine (Fig. 10A). Exposure of larvae to 1 /j.M emetine for 1 h was sufficient to reduce incorporation of 35 S-methionine into newly synthesized protein by 79%. An increase in the length of exposure to 1 juA/ emetine did not substantially reduce the incorporation of the radiolabeled umino acid. A maxi- mum reduction of 81% in 35 S-methionine incorporation was detected after 6 h of exposure to 1 ^M emetine (Fig. 10B). An increase in the length of exposure of larvae to 250 nM emetine resulted in lower levels of incorporation of 35 S- methionine. Periods of exposure to the inhibitor of 6 and 9 h reduced '^S-methionine incorporation by 68% and 74% respectively, in comparison to controls where no emetine METAMORPHOSIS AND GENE EXPRESSION 119 ox 100 80 60 - 40 20 - NBR ' - BR . /T t ~"k *~ *-~^X N 0.1 1 10 DRB (uM) A. Figure 2. Effects of the transcription inhibitor DRB on metamorphosis and differentiation of branchial radicles in Hydroides elegans. Larvae were induced to metamorphose with a marine biofilm in the presence (0.1. 1. and 10 jjjVft and absence (0) of DRB for 3.5 h. Points indicate mean percent- ages of larvae 1 SE (n = 4 replicates/treatment) that were swimming (S). metamorphosed without branchial radicles (NBR). or metamorphosed with branchial radioles (BR) at the end of the experiment. c 80 -+- FSW BQ BIO .= p. 60 BIO+DRB -+- DRB 93 40 ** Ol } I S 20 ^- ^ t A ^ ^ * u Controls 369 B. Pre-incubation in DRB (h) Treatments 100 *. ov oe => . _j * ^K K N ^ i , ~~ T Controls 3 ( i 9 1:111 n i in DRB(nM) + Cs Treatment Figure 3. Effects of the transcription inhibitor DRB on metamorphosis and differentiation of branchial radioles in Hydroides elegans. Larvae were exposed to a 3-h pulse of 10 mMCsCl in the presence (treatments 1 and 10) and absence (treatments and BIO) of DRB. After the cesium pulse, the solution was replaced with fresh filtered seawater (FSW) containing 1 or 10 /xA/ DRB in the treatments where DRB was included, or only FSW in the treatments where DRB was absent. Points indicate mean percentages of larvae 1 SE (n = 4 replicates/treatment) that were swimming (S), metamorphosed without branchial radioles (NBR). or metamorphosed with branchial radioles (BR) 24 h after initiation of the cesium pulse. in DRB (h) Treatments Figure 4. Effects of preincubation in DRB on metamorphosis of Hydroides elegans. Larvae were preincubated in 10 \iM DRB for 3. 6. or 9 h before transfer to a biofilmed petri dish. In the treatments with DRB. the inhibitor was present throughout the experiment. Points indicate per- centages of larvae that metamorphosed 1 SE (n = 4 replicates/treatment) after being transferred to the petri dish. (A) 3 h afer tranfer to the inducing biofilm. larvae had metamorphosed to the pre-radiole stage in the treat- ments where DRB was present. (B) 14 h after transfer to the biofilmed dish, small branchial radioles began to develop in the treatment where DRB was present. BIO, substratum coated with a marine biofilm; FSW. seawater filtered through a 0.22-^.m filter; in DRB control, no biofilm was added. was present (Fig. 11 A). In contrast, the lowest level of incorporation of 35 S-methionine into newly synthesized protein (97% reduction) was observed when the larvae were induced with CsCl after being incubated for 2 h in 250 nM emetine (Fig. 11B). Interestingly, induction of larvae to metamorphose with CsCl alone showed a reduction in 120 E. CARPIZO-ITUARTE AND M.G. HADFIELD BIO 100 Cs+DRB V5 ^Cs V5 DRB o jj 80 ^^ FSW p. J O 60 5 ^^ | ^ ^^ -2 40 ^* 1 20 s== =-^ i x BIO 369 Time of pre-incubation (h) Treatment Figure 5. Effects of preincubation in DRB on metamorphosis of Hydroides elegans. Larvae were preincubated in 10 /n.A-7 DRB for 3, 6. or 9 h before being induced to metamorphose with a 3-h pulse of filtered seawater (FSW) containing 10 mM CsCl and 10 fiM DRB. After the CsCl pulse, solutions were replaced with FSW, and metamorphosis was allowed to progress for 21 h. Points indicate mean percentages of larvae that metamorphosed I SE (n = 4 replicates/treatment) 24 h after initial exposure to CsCl. DRB, 10 pM DRB; Cs, 10 mM CsCI/ 3 h; BIO, substratum coated with a marine biofilm; FSW, seawater filtered through a 0.22-ju.m filter. In the treatment where Cs was present with DRB (DRB + Cs), larvae metamorphosed to the pre-radiole stage. incorporation of S-methionine of 82% compared to a control where no CsCl or emetine were present (Fig. 1 IB). Discussion Results from the present study demonstrate two important elements of metamorphosis in Hydroides elegans. First, settlement, attachment, secretion of the primary and second- ary tubes, velar loss, collar formation, and caudal elongation proceed in the presence of inhibitors that drastically reduce transcription and translation. Second, in the presence of transcription and translation inhibitors, metamorphosis stops at the point when branchial radicles begin to grow. When competent larvae of H. elegans were induced to metamorphose with a bacterial biofilm or CsCl in the pres- ence of effective concentrations of DRB, they completed metamorphosis to the point at which branchial radioles begin to develop. Development of the radioles can be con- sidered the initiation of early juvenile development in H. elegans, after all larva! structures have been lost and the juvenile shape has been realized. These changes include loss of the prototroch, differentiation of the collar, caudal elon- gation, and secretion of primary and secondary tubes. Sim- ilar results were reported initially for the tropical nudi- o o. o 100 80 60 20 emet+BIO BIO emetine EtOH FSW Treatment Figure 6. Effects of the translation inhibitor emetine on metamorpho- sis in Hydroides elegans. Larvae were exposed to a marine biofilm in the presence (emet + BIOl or absence (BIO) of 1 ^M emetine. Bars indicate percentages of larvae that metamorphosed I SE (H = 4 replicates/ treatment) after 3 h of exposure to the biofilm. BIO, substratum coated with a marine biofilm; emetine, emetine 1 fj.M control; EtOH, equivalent con- centration of ethanol used to dissolve emetine; FSW, seawater filtered through a 0.22-p.m filter. In the treatment where the biofilm was present with emetine (emet + BIO), larvae metamorphosed to the pre-radiole stage. 100 -I I/I '55 80 o & 60 \ S 40 ^ 20 emet+BIO BIO emetine EtOH FSW Treatment Figure 7. Effects of preincubation in the translation inhibitor emetine on metamorphosis in Hydroides elegans. Larvae were incubated for I h in 1 IJ.M emetine in a clean petri dish and then transferred to a biofilm-coated petri dish, still in the presence of emetine (emet + BIO). Bars indicate percentages of larvae that metamorphosed 1 SE (n = 4 replicates/ treatment) after 3 h of exposure to the biofilm. Emetine, emetine 1 /J.A-/ control; EtOH. equivalent concentration of ethanol used to dissolve eme- tine; FSW. seawater filtered through a 0.22-ju.m filter. In the treatment where the biofilm was present with emetine (emet + BIO), larvae meta- morphosed to the pre-radiole stage. branch Phestilla sibngae by Hadfield (1978) and later confirmed in the same species by B. J. McCauley (1997) and Del Carmen and Hadfield ( 1999, 2000). In the presence METAMORPHOSIS AND GENE EXPRESSION 121 100 35 80 o .c a fe 60 | 40 ^ 20 Cs+emetine Cs emetine BIO FSW Treatment Figure 8. Effects of the translation inhibitor emetine on induction ot metamorphosis in Hydnrides elegans. Larvae were incubated for 3 h in 1 \M emetine in filtered seawater (FSWl that also contuinted 10 mM CsCl to induce metamorphosis. Solutions were replaced with FSW after 3 h. and the larvae were allowed to complete metamorphosis for another 16 h. Bars indicate percentages of larvae that metamorphosed I SE (;i = 4 repli- cates/treatment). Cs, 10 mM CsCl/3 h: emetine. 1 /xM emetine; BIO, substratum coated with a marine biofilm; FSW, seawater filtered through a 0.22-ju.A/ filter. of DRB, transcription and translation were drastically re- duced without inhibiting induction of metamorphosis in competent larvae of P. sibogae. Further experiments showed that activation of the receptor for the metamorphie signal is not affected by reduction in transcription, but new transcription appears to be necessary for the final, elonga- tion phase of metamorphosis in P. sibogae (Del Carmen and Hadfield. 2000). The observation that metamorphosis in H. elegans progresses in the presence of DRB, together with the dem- onstration that DRB effectively inhibits RNA synthesis, indicates that the initiation and early phases of metamor- phosis in this species are independent of new RNA synthe- sis. However, the fact that we were never able to completely inhibit RNA synthesis leaves open the possibility that syn- thesis of some mRNAs was resistant to the action of DRB, and thus that new transcripts may play an essential role during metamorphosis for the first hours after induction. Transcriptional resistance to DRB in HeLa cells was re- ported by Zandomeni el al. (1982). The observation that development in the presence of DRB stops at the point when branchial radicle development begins indicates that new transcripts are necessary for the growth of the radicles and that these transcripts are among the more than 80% of RNA syntheses inhibited by DRB. Even a 40% reduction in synthesis of RNA, measured by incorporation of [5,6 3 H]uridine, was sufficient to prevent the formation of branchial radicles. It is not surprising that active growth of new structures, such as the branchial radi- cles, is dependent on transcription, because a broad spec- trum of new tissues including nerves, muscles, and secre- tory epithelium must be generated, each requiring many new proteins. Experiments on the hydrozoan Hydractinia echinata revealed that morphogenesis of the adult polyp can be influenced by pulse applications of a-amanitin or cordycepin. These compounds, which block mRNA tran- scription at different levels, inhibit the formation of tenta- cles and stolons in H. ec/iinata when applied during late gastrulation or 3 h after induction of metamorphosis (Eiben, 1982). When larvae of Hyilroides elegans were induced to meta- A. 1.4e+6 1.2e+6 '5 l.Oe+6 Q. 8 - 0e+5 61 i 6.0e+5 S Q. 4.0e+5 O 2.0c+5 0.0 B. II 10 30 50 DRB(|aM) 70 o EL SI E Q. O 4e+6 3e-f6 - 2e+6 le+6 036 12 Incubation (h) 24 Figure 9. Incorporation of [5-6 'H|uridine into newly synthezised RNA by larvae of Hydroides elegans. In (A), larvae were incubated for 2.15 h in the presence of the different concentrations of DRB indicated on the .v axis. In (B), larvae were incubated in 10 p.M DRB in FSW for the different periods of exposure indicated on the .v axis. In both (A) and (B). larvae were allowed to incorporate [5-6 3 H]uridine during the last 45 min. Points are means of uridine incorporation 1 SE (n = 2 replicates/ treatment). 122 A. E. CARPIZO-ITUARTE AND M.G. HADFIELD 80x10' 60.x 10 3 - D, 40x1 3 2 a. 20.x 10 1 - 250 500 1000 Concentration of emetine (nM) B. SOxlO 5. 60x1 O 3 - E 40x10' n. -a 20x10-' - no emetine emetine 1 uMolar O O 1 4 6 Period of incubation (h) Figure 111. Incorporation of 33 S-methionine into newly synthesized proteins by larvae of HyJmitles elegans in the presence of the translation inhibitor emetine. In (A), larvae were incubated in varying concentrations of emetine, shown on the _v axis, for 4 h with incorporation of "S- methionine during the last 1.5 h. Points are means of methionine incorpo- ration 1 SE (H = 2 replicates/treatment). In (B). larvae were incubated in 1 fj.M emetine for the intervals indicated on the .v axis and allowed to incorporate "S-methionine for an additional 45 min. Points are mean methionine incorporation 1 SE. morphose with either a bacterial biotilm or a pulse of CsCl in the presence of the translation inhibitor emetine, just as with DRB, they completed metamorphosis and stopped developing before the outgrowth of the branchial radioles. Neither the highest concentration of emetine used ( 1 juM) nor an extended preincubation period in emetine stopped initiation of mctamorphic morphogenesis in response to a bacterial biofilm or cesium as inducers. These results are consistent with the hypothesis that initiation of metamor- phosis in H. elegim* is also highly independent of synthesis of new proteins, a result that is concordant with previous studies of other species, especially certain molluscs (Fen- teany and Morse, 1993; Del Carmen and Hadtield. 1999). Larvae of the archaeogastropod Haliotis ntfescens (Fenteany and Morse, 1993) and the nudibranch Pliestilla sibogae (B. J. McCauley, 1997; Del Carmen and Hadfield, 1999) initiate metamorphosis in the presence of translation inhibitors. When competent larvae of H. nifescens were induced to metamorphose in the presence of concentrations of emetine or anisomycin sufficient to block protein synthe- sis, settlement and plantigrade attachment still occurred, indicating that these premetamorphic processes do not require de novo protein synthesis (Fenteany and Morse, 1993). The authors suggest that the induction of settlement A. 1 40\ 1 0' '3 lOOxlO 3 p . XO.xlO 3 WO .9- 60x1 3 E 40.x 10' 20x10' 9h 6h no emetine B. Period of incubation in emetine Treatment incubation in emetine before induction with CsCl Treatment Figure II. Incorporation of 35 S-methionine into newly synthesized proteins by larvae of Hydroides elegans after different periods of incuba- tion in 250 nM emetine. In (A), larvae were incubated in emetine at the concentrations indicated on the v axis and allowed to incorporate 35 S- methionine during the last 2 h. Points are means of methionine incorpo- ration 1 SE (n = 2 replicates/treatment, except for 6 h. which represents 1 sample). In (B). larvae were first incubated in 1 fj.M emetine in filtered seawater (FSW) before being induced to metamorphose with a 3-h pulse of CsCl ( 10 mA/) and allowing larvae to incorporate methionine for 45 min after the CsCl pulse and in the presence of emetine. Points are mean methionine incorporation I SE (H = 4 replicates/treatment for 2, 1. and h, and 2 replicates/treatment for CsCl and not induced (no CsCl)). METAMORPHOSIS AND GENE EXPRESSION 123 and plantigrade attachment in the presence of protein-syn- thesis inhibitors is consistent with the notion that these behavioral responses are controlled by chemosensory mech- anisms mediated by the nervous system. Similar results were reported for the larvae of P. sibogae: experiments with emetine showed that metamorphosis can be initiated in the presence of the translation inhibitor and that development stops just before final elongation of the body (B. J. McCau- ley, 1997). It is particularly interesting that protein synthesis in lar- vae of Hvdroides elegans was reduced even further when CsCl was added to induce metamorphosis (Fig. 11B). Kroiher el at. (1991) also reported decreased incorporation of amino acids into proteins when larvae of the hydrozoan H\dractinia echinata were exposed to CsCl, attributing it to a reduction in the Na + concentration of the medium when CsCl is added. Transport of amino acids across the cell membrane is known to be a function of a Na + -transport in marine invertebrates (Stephens, 1988). Inhibition of development of the branchial radioles by emetine at concentrations at which protein synthesis was reduced by only 40% is consistent with the data obtained with DRB. Thus, unsurprisingly, both new mRNA tran- scripts and their translation products are necessary for pro- duction of the branchial radioles. Taken together, the data on RNA and protein synthesis in metamorphosing Hvdroides elegans strongly support the conclusion that metamorphosis and juvenile development occur in two phases one independent of new gene tran- scription and translation, and a second in which these two molecular processes are necessary for development and growth to continue. During the first phase, larvae respond rapidly to the bacterial inducer, change their swimming behavior, attach, secrete primary and secondary tubes, lose the prototroch cilia, differentiate the collar, and elongate the caudal region (Carpizo-Ituarte and Hadfield, 1998). During the second phase, the metamorphosed worm develops branchial radioles, a process that requires new RNA and protein synthesis. Results reported for larvae of other ma- rine invertebrates reinforce the idea that metamorphic com- petence in many marine invertebrate larvae is a develop- mental stage primed to metamorphose and in which the activation and near completion of metamorphosis requires neither de novo synthesis of mRNA nor proteins (Hadfield, 2000). Since larvae of H. elegans survive through metamorpho- sis with very reduced synthesis of new mRNAs or proteins, cell proliferation, which requires both processes, probably does not play an essential role during metamorphosis. Pre- vious studies in other invertebrate species are consistent with this hypothesis. In Phestilla sibogae, labeling of divid- ing cells with an antibody against a proliferating cell nuclear antigen (PCNA) decreases when larvae reach competency and is detected again only after the larvae have completed metamorphosis and begun early juvenile development (B. J. McCauley, 1997). In embryos of Hydractinia echinuta. the index of cell proliferation measured by the BrdU method decreased to 4% by 78 h after fertilization. In larvae of H. echinata kept at low temperatures, the cell proliferation index went to zero after 60 to 70 days without affecting the ability of the larvae to undergo metamorphosis (Plickert el ul, 1988). Preliminary observations on Hydroides elegans using PCNA indicated that cell proliferation decreases sharply when larvae become competent and increases nota- bly in the developing branchial radioles of newly settled juveniles (Carpizo-Ituarte. unpubl. results). Metamorphosis in most well-studied insects and amphib- ians is a relatively slow process wherein hormonal tran- scription factors mediate essential cascades of "de novo" transcription and translation to regulate metamorphic mor- phogenesis (Gilbert era!., 1996; Tata. 1996, 1999; Shimizu- Nishikawa el /., 2002). In contrast, the experimental results obtained with the polychaete Hydroides elegans support a generalization noted by Hadfield et al. (2001): for most well-studied marine invertebrates, the competent larva car- ries within it a well-formed juvenile body, which metamor- phosis need only liberate from larval structures. In this case, metamorphosis should require little or no de novo transcrip- tion or translation, processes that will be necessitated when juvenile growth begins. This may be the result of adaptation to undergo rapid metamorphosis and make an extremely vulnerable period that of a planktonically adapted larva residing on the benthos as brief as possible (Hadfield. 2000). The more precisely we define the molecular events of metamorphosis in Hydroides elegans and other marine in- vertebrates, the better we are able to make comparisons with model systems among insects and amphibians. So far. the capacity for metamorphosis to proceed almost indepen- dently of new gene expression in many marine invertebrates points toward different mechanisms to initiate postlarval development. To what extent these mechanisms are differ- ent from those known in insects and amphibians awaits further investigation. Knowledge of the signal-transduction pathways that are activated during induction of metamor- phosis and how the initial triggering event orchestrates the genetic machinery to turn a larva into a juvenile in a wide variety of marine invertebrates will help us to understand how the widespread phenomenon of metamorphosis in ma- rine invertebrates evolved. Acknowledgments Contributions of colleagues in the Hadfield lab at the University of Hawaii's Kewalo Marine Laboratory were many, and we are grateful for all of them. This research was supported by ONR grants NOOO 14-95-1015 and N00014- 124 E. CARPIZO-ITUARTE AND M.G. HADFIELD 95-1-0196. EC-I was the recipient of a scholarship for graduate studies from CONACyT, Mexico. Literature Cited Ausubel, F. M., R. Brent. R. E. Kingston, D. D. Moore, J. G. Seidmann, .]. A. Smith, and K. Struhl. 1987. Section VI. Specialized Applica- tions. Unit 10. IS. Biosynthetic Labeling of Proteins. Pp. 10. IS. I - 10. IX. 1 -) In Ciinrnl Protocols in Molecular Biolo K \. Vol II. John Wiley. Boston. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidmann, J. A. Smith, and K. Slruhl. 1992. Section I. Quamitution of Pro- teins. Unit 10.1. Colorimetric Methods: Bradford Method. Pp. 10.5- 10-6 in Short Protocols in Molecular Biolog\. 2nd ed. John Wiley. Boston. Baxter, G. T., and D. E. Morse. 1987. G protein and diacylglycerol regulate metamorphosis of planktonic molluscan larvae. Proc. Null. Acad. Sci. USA 84: 1867-1870. Berking, S. 1998. Hydrozoa metamorphosis and pattern formation. Pp. 81-131 in Current Topics in Developmental Biology, R. A. Pedersen and G. P. Schatten, eds. Academic Press, New York. Biggers, W. J., and H. Laufer. 1999. Settlement and metamorphosis of Capitella larvae induced by juvenile hormone-active compounds is mediated by protein kinase C and ion channels. Biol. Bull. 196: 187-198. Broun, M., S. Sokol, and H. R. Bode. 200(1. Oi,i;.vi-, a homologue of goosecoid. participates in the patterning of the head, and is expressed in the organizer region of Hydra. Development 126: 5245-5254. Cariolou, M. A., and D. E. Morse. 1988. Purification and characteriza- tion of calcium-binding conchiolin shell peptides from the mollusc, Haliolis mfescens. as a function of development. ./. Comp. Plivsiol. B 157: 717-729. Carpizo-Ituarte, E., and M. G. Hadfield. 1998. Stimulation of meta- morphosis in the polychaete Hvdroidcs elet>wm Haswell (Serpulidae). Biol. Bull. 194: 14-24. Cavanaugh, G. M. 1956. Formulae ami Methods I'/. Mamie Biological Laboratory, Woods Hole. MA. Pp. 62-&y. Clare, A. S. 1996. Signal transduction of barnacle settlement: calcium re-visited. Biofouling 10: 141-159. Clare, A. S., R. F. Thomas, and D. Rittschof. 1995. Evidence for involvement of cyclic AMP in the pheromonal modulation of barnacle settlement. ./. Exp. Biol. 198: 655-664. Degnan, B. M., and D. E. Morse. 1993. Identification of eight ho- meobox-containing transcripts expressed during larval development and at metamorphosis in the gastropod mollusc Haliotis mfescens. Mot. Mar. Biol. Biotech. 1: 1-9. Degnan, B. M., and I). E. Morse. 1995. Developmental and morpho- genetic gene regulation in Haliotis rufcsccns larvae at metamorphosis. Am. Zool. 35: 391-398. Degnan, B. M., J. C. Groppe, and D. E. Morse. 1995. Chymotrypsm inRNA expression MI digestive gland amocbocytes: cell specification occurs prior to m, ti phosis and gut morphogenesis in the gastropod. Haliotis mfescens. :, \,ch. Dev. Biol. 205: 97-1(11. Del Carmen, K. A., and A Hadfield. 1999. Metamorphosis in a marine gastropod: the rol .inscription and translation. Am. Zool. 39: 20A. Del Carmen, K. A., and M. G. ll.idtield. 2000. Metamorphosis in a marine gastropod: the role of transcription. P. 37A in Larval Biolof20) tend to lie in semicircles around the bundle of modified cilia (Fig. 4E) and fall into two types. The majority are rather electron- lucent, but two to four of them are electron-dense and contain many vacuoles (Fig. 4E, asterisk). The innermost sheet cell contains a scolopale (Fig. 4F) and forms a den- dritic sheet, which encircles the cilia from just below the basal part of the seta to the distalmost 20 /u,m. In the region of the scolopale, the innermost sheet cell makes a projection that suiTounds one of the sensory cells with a large rootlet (Fig. 4F). The extracellular matrix within the lumen of the dendritic sheet is electron-dense. The next five to eight sheet cells send projections into the lumen of the setal shaft (Fig. 4D), and one of these projections continues to within at least 20 ju,m below the tip (Fig. 4C). We could not identify which sheet cell makes this projection. An extracellular matrix encircles both the ciliary compartment and the cell projec- tions in the setal lumen. The extracellular matrix is most prominent at the base and barely detectable near the tip (Fig. 4C). It contains both electron-dense and electron-lucent areas. The cuticle of the seta is rather thick, and in the distal part it is divided into six distinct layers (Fig. 4C). The outermost thin epicuticle and the ca. 1 -/nm-thick, electron- lucent second layer are the only ones present in the out- growths of the seta (Fig. 4C). The third layer is electron- dense and contains radially projecting canals that seem to be composed of the same material as the second layer. The cuticle of the convex side of the seta is more granular and without canals. Due to an oblique cross-section of the seta, the third layer varies between 1 and 2 /im in thickness. The fourth layer is about 0.5 /u,m thick, fibrous, and changes gradually into the fifth layer, which is similarly fibrous but very electron-dense and about 0. 1 jum thick. A sixth mem- branous layer encircles the lumen (Fig. 4C, insert: ML). In the proximal two-thirds of the shaft, the second and third layers have gradually merged to form one homogeneous electron-dense layer 2-4 jum thick (Fig. 4D). The fourth layer has expanded to a thickness of about 1 /im. The fifth layer is not detectable, but the innermost membranous layer remains unchanged. Type 2 (Fig. 5) is innervated by 18-25 sensory cells, which give rise to 18-25 modified cilia (mean = 23.25. n = 12) (Fig. 5G). Some of these cilia branch, and there are 28-30 ciliary branches in the distal end of the lumen (Fig. 5D, F). The cilia continue all the way to the tip of the seta, where they protrude through the terminal pore, which lies in the extension of the setal lumen (Fig. 5, Cl. C2). All the sensory cilia are alike and contain rather few microtubules. in the distal part only 4-10 strands. Two of the sensory cells have a large rootlet; the rest have a small fragmented rootlet. The sensory cells are again surrounded by numerous semi- circular sheet cells (>20), which are dimorphic in their Figure 6. Canal system in the third layer of cuticle of seta type 2. (Al Cross-section of type 2 seta cu. 5 /urn from the tip. Section is in a rather poor condition because it is the first in a series. (B) Close-up of section in A. showing cuticular canals (CuCl interconnected by circular canal. (C) Close-up of section in A, showing the pores (arrows) where cuticular canals connect with external environment (EE). Cu2-Cu5 = cuticular layers 2-5, SC = sensory cilia, SS = scale-like setule. FUNCTIONS OF MX 2 IN P. ADSPERSUS 133 electron density and arranged in a variable pattern. About 10 of the sheet cells send projections into the setal lumen. Midway up the setal shaft, the number of cell projections is reduced to three to four, and at the tip only the modified sensory cilia are present within the lumen (Fig. 5D-F). The innermost sheet cell forms the dendritic sheet, which runs from just below the base of the seta all the way to the tip (Fig. 5D, insert) and thus encloses the sensory cilia in a sparse electron-dense extracellular matrix. The innermost sheet cell also contains a rather rudimentary scolopale (Fig. 5H). From the base of the seta, there is also an extracellular matrix in the setal lumen outside the ciliary compartment, which narrows distally and is absent at the tip. The cuticle is of the same thickness and structure as described for type 1 setae, but it is not granular, and the canals in the electron- dense third layer are better developed. The second layer is also thinner in the distal part (Fig. 5D). The tube of the terminal pore appears to be formed from the third layer, but it does not contain any canals (Fig. 5. Cl. C2). The canals in type 2 setae are connected to the external environment (Fig. 6). The radial canals are interconnected by a circular canal, which runs on the outside of the third cuticle layer (Fig. 5D, arrowhead; Fig. 6B). This canal in turn is con- nected to the exterior via a pore (Fig. 6C, arrow). These pores are made from infoldings of the epicuticle and second cuticular layer (Fig. 5D, arrow). Unfortunately, the series of sections is not complete, and we are therefore not able to show if the openings are really pores or if they continue and expand further distally. Type 3 setae (Fig. 7) are innervated by four to nine sensory cells (mean = 6. 1 . n = 11), each giving rise to one modified cilium (Fig. 7E, F). These cilia continue un- branched to the tip (Fig. 1C). One or two of the sensory cells have a large rootlet; the rest have fragmented rootlets. The type 3 seta has fewer (<20) enveloping cells than the two other types (Fig. 7E). At least six of them send projections into the lumen, and one proceeds to at least 20 ^m below the tip (Fig. 7C). The innermost cell forms the dendritic sheet and contains a well-developed scolopale. In the region of the scolopale. the innermost sheet cell encircles one or two of the sensory cells. As in the other types, the dendritic sheet forms the ciliary compartment, which contains the modified cilia and a sparse extracellular matrix. The extra- cellular matrix in the setal lumen is similar to that found in the two other types. The cuticle is also similar to what is found for type 2, except that the tube is lacking in type 3. Discussion We found that maxilla 2 of Palaemon adspersus performs stereotypical movements when the animal is handling food items. The medial rims of basis 1 and 2 of maxilla 2 touch the food items gently and were never seen to directly manipulate them. This area of maxilla 2 carries three types of setae, each with different external and internal morphol- ogy. Type 1 is serrate, robust, and has 5-9 sensory cells. Type 2 is serrulate, slender, and has 18-25 sensory cells; it also has a terminal pore from which sensory cilia protrude into the external environment. Type 3 is serrulate, short, and possesses four to nine sensory cells. In all three types, the sensory cells fall into two distinct morphological groups. Putative sensory properties of the three types are summa- rized in Table 1. Our results on the activity and morphology of maxilla 2 and its setae in Palaemon adspersus lead us to conclude that the basis of maxilla 2 collects chemical information from a dense population of gustatory setae before the food item is either eaten or rejected. The ultrastructure indicates that the three types of setae have distinct and separate functions. Type 1 and type 3 setae are innervated by four to nine sensory cells, one or two of which have a well-developed rootlet in close contact with a scolopale in the innermost sheet cell. A scolopale is an electron-dense structure made of closely packed strands of microtubules bound together with accessory proteins. This indicates that these cells are mechanoreceptors, since it is believed that desmosomal connections between the scolo- pale and the cilia rootlet are necessary for transduction of the mechanical signal (Schmidt and Gnatzy, 1984; Derby, 1989; Crouau. 2001). The remaining sensory cells have a small fragmented rootlet that is not in contact with the scolopale; they lack dynein arms in the ciliary region, and there are few strands of microtubules in the paraciliary region. This suggests that they are chemoreceptor neurons, but this assumption is based on the absence of mechanore- ceptive structures (Schmidt and Gnatzy, 1984; Derby, 1989; Gleeson et al.. 1996; Hallberg and Hansson, 1999). We therefore conclude that type 1 and type 3 setae are bimodal sensors capable of detecting both mechanical and chemical stimuli. The existing ultrastructural studies on decapod se- tae suggest that this is the most common type of sensory innervation (Altner et al., 1983; Schmidt and Gnatzy, 1984; Derby. 1989; Gate and Derby, 2001, 2002). We emphasize that our conclusions are based on behavior and morphology alone, and that no certainty about the sensory properties can be obtained without employing other methods such as elec- tro-physiological recordings from the sensory cells. Al- though both type 1 and 3 setae are probably bimodal, their size, shape, and arrangement indicate that they have differ- ent functions. The long, robust, and almost straight type 1 setae seem well suited to act as "guard setae," protecting the rather fragile type 2 setae from too much mechanical stress when food objects are being handled. A similar system is found on the first antennae of many decapod crustaceans, including the spiny lobster, Paniilirns argits, where the aesthetascs (fragile chemosensory setae) are protected by long robust guard setae (Steullet et al.. 2000b; Cate and Derby, 2001). In the case of the spiny lobster, the ultrastructure of these setae is still unknown, and the protective function is as- 134 A. GARM ET AL. Figure 7. Type 3 seta. (A) Scanning electron micrograph of type 3 seta in dorsal row. Planes indicate approximate area of transmission electron micrographs. (B) Close-up of the tip. (C) Semi cross-section ca. 10 ;u.m from tip. The cuticle is divided into six distinct layers (EC, Cu2-Cu5, ML not shown). Cu3 has prominent canals (CuC). Insert = close-up of lumen with ciliary compartment (CC) enclosed by the dendritic sheet (DS), one cell extension (CE). and a sparse extracellular matrix (ECM). (D) Cross-section in nonserrulate part of seta. Cuticle layers 2-5 have fused to form one more or less homogenous layer. Prominent lumen with large amount of extracellular matrix, eight to nine cell extensions, and a ciliary compartment with nine sensory cilia. (E) Cross-section just below the base of the seta. The ciliary compartment is enclosed by a dendritic sheet and 1 5-20 semicircular sheet cells of two types (SCI, SC2). The sheet cells are surrounded by extracellular matrix. Insert = close-up of ciliary compartment. (F) Cross-section ca. 50 /j,m after E. One cell has a large rootlet (LR); the other cells are sectioned in the region of the basal body (BB) or in the ciliary region (CR) or the paraciliary region (PR). A weak scolopale is seen (Sc). Arrowhead indicates extrusion of sheet cell enclosing sensory cell. (G) Cross section ca. 5 (urn after F. The putative chemosensory cells contain a fragmented rootlet (FR). The mechanosensory cell is in contact with the scolopale via desmosomes (arrowheads). Arrow indicates large ciliary rootlet. sumed from their arrangement and external morphology. The guard setae on P. argus are not serrate but simple (no outgrowths). Most other systems with crustacean mechano- sensitive setae studied so far are sensitive to movements of the surrounding water (Wiese, 1976; Heinisch and Wiese, 1987; Derby, 1989); but this does not seem to be the case with the type 1 setae since the socket is reduced, they have no outgrowths suitable to receive water movements, and they are themselves moved all the time. Due to the activity and position of type 1 setae in Palaenwn adspersus, we FUNCTIONS OF MX 2 IN P. ADSPERSUS Table 1 Suintihiiy of the structures and suggested functions for setti tvpe I, 2, 3 of Palaemon adspersus * Plus indicates presence; - indicates absence. 135 Type of seta Length in fim Pore* Neurons/cilia Scolopale* Modality Primary function Type 1 Type 2 Type 3 Serrate Serrulate Serrulate 250-300 200-250 + 150-200 4-9 18-25 4-9 + Bimodal + Bimodal + Bimodal Protection Gustation Grooming believe that they contain mechanoreceptor neurons that function as proprioceptors, providing information on the amount of pressure the animal puts on the food object and thus ensuring that it does not break the setae. The chemo- sensory function of type 1 setae is not clear but might be gustatory. Due to their position, curvature, and small size, the ser- rulate type 3 setae cannot be guard setae. They are more likely used for grooming the ventral side of the basis of maxilla 1, which they scrape every time these mouthparts pass each other. This explanation is consistent with previous studies on decapods in which serrulate setae have been found to be involved in grooming (Martin and Felgenhauer, 1986; Bauer, 1989, 1999; Nickell et ai, 1998), and the scale-like setules serve to remove debris. The type 2 setae are also bimodal sensors, and their primary function is probably gustatory, as indicated by their many chemosensory cells and by the observation that the sensory cilia protrude through the terminal pore. This also indicates that even though type 1 and 3 setae may be chemosensory, the type 2 setae are probably the primary chemosensors of Mx2, and their mechanosensory function may again be proprioceptive. The type 2 seta is the first decapod seta to be reported in which cilia protrude from a terminal pore. Terminal pores are rather common in decapod setae, but so far they have been studied only with SEM (Altner e!al.. 1986; Finn et til.. 1999; Garm and H0eg, 2000; Coelho and Rodrigues, 2001 ). This might fail to reveal protruding cilia, because the treat- ment during dehydration and critical-point drying may cause shrinkage of the soft tissue, thereby retracting the sensory cilia. Naked sensory cilia of presumed chemosen- sors of other crustaceans, Hutchinsoniella macracantha (Cephalocarida) and Pachypygus gibber (Copepoda), have been shown by TEM, and in the former case also by SEM, although the documentation is not of the highest quality (Hipeau-Jacquotte, 1986; Elofsson and Hessler, 1994). In both cases, these setae are suggested to be gustatory, but they differ structurally from what is described here for type 2 setae of Palaemon adspersus in that they are small (<30 jam), have a very thin cuticle, and are innervated by only two to three sensory cells. They are also presumed to be unimodal gustatory setae (no mechanosensory function) (Hipeau-Jacquotte, 1986; Elofsson and Hessler, 1994), which has also been suggested for setae of the amphipod Anonyx lilljeborgi (Steele and Steele, 1999). The number and arrangement of the sheet cells resemble earlier reports from decapod setae (Wiese, 1976; Ball and Cowan. 1977; Hallberg et ai. 1992). The division into two types of sheet cells might have a functional explanation. In the mysid Neomysis integer, the sheet cells are arranged in populations, with each creating a separate part of the cuticle of the setae (Guse, 1980). Our results show a tendency toward an outer population of electron-dense cells and an inner population of electron-lucent cells, the latter sending cell extensions into the lumen of the seta. Although we have no direct proof, we suggest that the electron-dense cells form the basal part of the cuticle of the seta and the electron- lucent cells form the distal part. A projection of the inner- most sheet cell that encircles the mechanosensory cells has not been reported before for any crustacean setae. We are not sure of the function of this arrangement, but it could help to anchor the sensory cells and thereby improve the transduction of the mechanosensory signal. In all three types of setae, the arrangement of the cuticle is extremely complex, with the differentiation throughout the shaft giving the distal part as many as six distinct layers (EC, Cu2-Cu5, ML) and water-filled canals. This could indicate that the distal parts of these setae are experiencing great mechanical stress, which they overcome by having several kinds of cuticle with different mechanical proper- ties. The highly specialized water-filled canals have not been reported before in crustacean setae. One functional explanation could be that some setae involved in handling food objects must be flexible in their distal parts. This means that the cuticle must be able to fold, stretch, or change volume. Having a water-filled canal system gives the tissue the capability of locally changing the volume by moving water. The opening to the exterior also makes it possible to change the volume of the entire canal system; however, the minute size of the opening may limit the speed of this process. As mentioned earlier, cross-sections of the distalmost part of the canal system are missing, and it is possible that the openings expand further distally. Some earlier studies found similar electron-lucent canals in the cuticle (Crouau, 1997; Matsuura and Nishida, 2000; Cate and Derby. 2002) but did not discuss them in detail, and no connection to the exterior was described. 136 A. GARM ET AL The activity of maxilla 2 in Pulaeiiion mlspersus is very stereotypic. We never observed the Mx2 to mechanically influence or hold the food; it merely touched it frequently (3-5 Hz). This behavior supports the idea that some of the setae on the medial rims of basis 1 and 2 of maxilla 2 have gustatory properties like those proposed for maxilla 2 of other decapods (Schembri. 1982: Garm and H0eg. 2001). Mechanical functions of maxilla 2 are, however, described for these other decapods. In the anomurans Mimida sarsi and M. tenuiiuami, the basis of maxilla 2 is additionally used to reorient small food objects (Garm and H0eg, 2001 ). In the hermit crab Pugiinis ntbricatiis, the basis of maxilla 2 assists in filtering sediment before eating (Schembri, 1982). In the case of Hoinarns gammants, maxilla 2 is described as having mechanical functions, but the observa- tions are not very detailed (Barker and Gibson, 1977). In other decapods for which detailed observations of mouth- part movements have been made, no account is given for the movements of maxilla 2 (Hunt et al.. 1992; Stemhuis et /!<> /< wv 101: 17-38. Schmidt, M., and W. Gnatzy. 1984. Are the funnel-canal organs the 'campaniform sensilla' of the shore crab Carcinus maenas (Decapoda: Crustacea) 1 ' II. Ultrastructure. Cell Tissue Res. 237: 81-97. Steele, V. J., and D. H. Steele. 1999. Cellular organization and fine structure of type II microtrich sensilla in gammaridean amphipods (Crustacea). Can. J. Zool. 77: 88-107. Steinbrecht, R. A. 1997. Pore structure in insect olfatory sensilla: a review of data and concepts. Int. J. Insect Morphol. Embryol. 26: 229-245. Steinbrecht, R. A. 1998. Bimodul thermo- and hygrosensitive sensilla. Pp. 405-422 in Microscopic Anatomy of Invertebrates. Insecta. F. W. Harrison and M. E. Rice. eds. Wiley-Liss, New York. Stemhuis, E. J., B. Dauwe, and J. J. Videler. 1998. How to bite the dust: morphology, motion pattern and function of the feeding append- ages of the deposit-feeding thalassinid shrimp Callianassa subterra- nea. Mar. Biol. 132: 43-58. Steullet, P., and C. D. Derby. 1997. Coding of blend ratios of binary mixtures by olfactory neurons in the Florida spiny lobster, Panulinis argus. J. Comp. Pliysiol. A 180: 123-135. Steullet, P., H. S. Gate, and C. D. Derby. 2000a. A spatiotem- poral wave of turnover and functional maturation of olfactory receptor neurons in the spiny lobster Panulirus argus. J. Neurosci. 20: 3282- 3294. Steullet, P., H. S. Cate, W. C. Michel, and C. D. Derby. 2000b. Functional units of a compound nose: aesthetasc sensilla house similar populations of olfactory receptor neurons on the crustacean antennule. J. Comp. Neural. 418: 270-280. Wiese, K. 1976. Mechanoreceptors for near-field water displacement in crayfish. J. Neurophysioi 39: 816-833. Reference: Biol. Bull. 204: 138-145. (April 2003) 2003 Marine Biological Laboratory Variation in Skeletal Microstructure of the Coral Galaxea fascicularis: Effects of an Aquarium Environment and Preparatory Techniques PETA L. CLODE AND ALAN T. MARSHALL* Anah'tical Electron Microscopy Laboratory, Department of Zoologv, La Trobe Unirersitv, Melbourne, Victoria 3086. Australia Abstract. To compare the crystalline microstructure of exsert septa, polyps of the scleractinian coral Galaxea fas- cicularis were sampled from shallow reef flat colonies, from colonies living at a depth of 9 m, and from colonies kept in a closed-circuit aquarium. Septal crystal structure and ori- entation was markedly different between corals in the field and in aquaria. In samples collected from deep water, acic- ular crystals were composed of conglomerates of finer crys- tals, and skeletal filling was considerably reduced when compared with samples collected from shallow water. Com- parisons were also made between septa prepared in sodium hypochlorite (commercial bleach), sodium hydroxide (NaOH). hydrogen peroxide (H^O,), and distilled water (dH 2 O). Commercial bleach was the most effective solvent for tissue dissolution in investigations of skeletal structure. Samples prepared in NaOH commonly displayed crystalline artefacts, while the use of dH,O and H 2 O 2 was labor inten- sive and often resulted in unclean preparations. Fusiform crystals were seen only on G. fascicularis septa prepared in bleach and on Acropora formosa axial corallites prepared in either bleach or dH-,0. We suggest that the mechanical agitation and additional washing necessary for samples pre- pared in dH^O. NaOH. or H^O, resulted in the loss of fusiform crystals from these preparations. Introduction The use of aquaria to grow and maintain corals for scientific experimentation has become increasingly popular Received 6 June 2002; accepted 31 January 2003. * To whom correspondence should be addressed. E-mail: /ooam(s' 7.00. latrohe.edu.au (see Carlson, 1999) and, with the degradation of coral reefs across the globe, it may soon be necessary for experimental corals to be constantly maintained in this way. In the past, keeping scleractinian corals alive under artificial conditions for long periods of time was exceedingly difficult, but advances in filtration, lighting, and water systems have made the propagation of these corals in aquaria much easier. However, little is known about the impact of artificial en- vironments upon scleractinian coral calcification, behavior, growth, and reproduction, with direct comparisons between field and aquarium-maintained corals rare. Studies of the crystalline and overall skeletal structure of scleractinian corals necessitate the removal of the surround- ing epithelia to visualize the CaCO, skeleton underneath. Many treatments have been used to dissolve the epithelial tissue, including HoOo (Jell, 1974), freshwater (Wainwright, 1963; Johnston, 1979). NaOH (Johnston. 1979; Isa, 1986). and commercial bleach (sodium hypochlorite) (Sorauf. 1972. 1974; Gladfelter, 1982, 1983; Brown et al., 1983; Hidaka, 1988, 199 la, b; LeTissier, 1988, 1990, 1991; Con- stantz, 1989; Hidaka and Shirasaka, 1992). Unfortunately, the possible effects of these chemicals upon the underlying crystal structure have been ignored. Comparisons between studies are complicated by the variety of methods used to prepare samples and by the broad range of species studied. In this study, we describe differences in crystal structure, orientation, and patterns of deposition between corals col- lected from field conditions and closed-circuit aquaria. In addition, we evaluate and compare the suitability of several chemical treatments commonly used to prepare scleractin- ian coral skeletons for examination with a scanning electron microscope. 138 VARIATION IN CORAL SKELETAL MICROSTRUCTURE 139 Materials and Methods Collection ami maintenance of Galaxea fascicularis Colonies of brown and green color morphs of the scler- actinian coral Galaxea fascicularis were collected at Heron Reef in the Capricorn Bunker Group of the Great Barrier Reef, Australia. Morphs of both colors were obtained at low tide from the reef flat; brown morphs only were collected by scuba divers from a depth of 9 m. Specimens to be maintained under field conditions were transported in buckets of seawater to Heron Island Research Station, where they were kept in well-aerated, flow-through aquaria in natural seawater at 24-25 C. Colonies collected from the reef flat were kept in outdoor, sunlit aquaria, with light levels equivalent to those experienced on the reef flat: colonies collected at depth were kept in shaded aquaria. with light levels similar to those at a depth of 9 m in the field. Specimens to be maintained in an aquarium environment were sealed into plastic bags with a small amount of sea- water, and the bags were put into small insulated boxes for transport to La Trobe University. Melbourne. These colo- nies were kept in a closed-circuit marine aquarium. The aquarium contained natural, well-aerated seawater at 25 C and was subjected to an illumination cycle of 14 h of light and 10 h of darkness; the light was provided by metal halogen lamps (photosynthetic photon flux density 150 photons m s ). Conditions in the aquarium were monitored regularly: 25% of the water was changed twice weekly; pH was kept between 8.1 and 8.2 and salinity from 900 to 1200 mosmol kg" 1 . Corals were fed a mixture of brine shrimp and fish weekly. Sample preparation Colonies were allowed to recover for at least 2 days after collection before being used in experiments. Individual polyps of G. fascicularis were separated from colonies (see Marshall and Wright. 1991) at 1200 h. Five polyps were placed in either commercial bleach (12% NaOCl) or NaOH (5 N) at 60 C for 30 min; H 2 O : (10%) for 3 h at room temperature (RT); or dH 2 O at RT for at least 24 h, to remove the overlying soft tissue. The resultant corallites were rinsed well in running water and then several times in dH 2 O. Any soft tissue remaining upon the exsert septa was removed by agitation and pipetting of dH ; O onto the sam- ple. The corallites were then dried at 60 C for 24 h. Aquarium-maintained corals were kept in the closed-circuit aquarium at La Trobe University for 26 weeks before polyps ( n - 5 ) from each of two colonies were sampled at 1200 h and prepared in either bleach, NaOH, or dH,O. Electron microscopy Exsert septa were removed from G. fascicularis corallites under a dissecting microscope and mounted flat using car- bon tape. Samples (n == 50+) were coated with 5 nm platinum and previewed in a JEOL JSM 840A scanning electron microscope at 10 kV. High-resolution imaging was subsequently conducted on a JEOL 6340-F field emission scanning electron microscope at 2 kV. In addition to being mounted flat, septa from colonies growing in shallow field conditions (/; = 5) and the aquarium (/; = 9) were secured upright in an appropriate substage, with indium foil placed between the sample and the stage to improve conductivity and provide flexible compression. Samples were coated with platinum, then fractured to provide a cross-sectional area. Following fracturing, the samples were re-coated and viewed by high-resolution field emission scanning electron microscopy (FESEM). Acropora formosa White-tipped branches (which possess few symbiotic zooxanthellae) were sampled from A. formosa colonies at low tide from Heron Reef at 1 200 h and immediately placed into bleach (n = 5) for 30 min or into d^O at ambient temperature (/; = 5) for 24 h, for investigation of the axial corallite. Branches were washed in running water for 24 h before being rinsed in dH 2 O and dried at 60 C for 24 h. Branch tips were secured upright in hollow stubs, using partially polymerized araldite, so that the axial corallite extended about 3 mm above the upper surface of the stub. Polymerization was completed at 60 C for a further 30 h. Conductivity of the corallites was improved by overlaying the araldite with conductive silver epoxy (ProSciTech) and coating with 10 nm platinum. Samples were viewed in a JEOL JSM 6340-F field emission scanning electron micro- scope at 1 kV or 2 kV. Results Comparison of field and aquarium samples Marked differences in crystal orientation and deposition were detected between septa from polyps collected from their natural environment (shallow reef flat) and those sam- pled from polyps that had been maintained in a closed- circuit aquarium for 26 weeks. When sampled from the natural reef environment, septa possessed acicular crystals that remained highly ordered and extended to the outer edge of the growing septa (Fig. 1A). In contrast, septa sampled from aquaria-maintained polyps had small, unordered crys- tals deposited with little uniformity (Fig. IB). This pattern of unordered skeletal deposition extended to a depth of 1 0.06 jum (H = 25). Little variation in this pattern of deposition was observed within or between colonies. 140 P. L. CLODE AND A. T. MARSHALL Structural features typically associated with normal Gal- axea fascicularis septa, such as acicular and fusiform crys- tals and well-defined fasciculi (Fig. 2 A), were rarely dis- cerned on septa sampled from polyps maintained in the closed-circuit aquarium. Instead, much of the septal surface of aquarium-maintained polyps was covered with small (<100 nm in diameter) spherulitic crystals (Fig. 2B). G. fascicularis polyps sampled from the closed-circuit aquar- ium were also typically paler than corals living on the reef flat, and they appeared to have fewer symbiotic zooxanthel- lae, although this was not quantified. In contrast, the crystal structure of septa from deep field- collected G. fascicularis colonies was only marginally dif- ferent from that of septa from shallow field-collected colo- nies. Acicular crystals of septa sampled from colonies collected from the shallow reef flat appeared as solid, broad crystals tightly packed together (Fig. 3 A). However, acicu- lar crystals upon septa collected from deeper water tended to be composed of conglomerates of finer, elongated crys- tals that were loosely associated (Fig. 3B). This loose as- sembly gave the appearance of increased porosity, with skeletal filling somewhat reduced. Little structural differ- ence in the other crystal types nano, lamellar, and fusi- form was noted between septa taken from shallow reef- flat corals and those living in deeper water. Septum length also varied between polyps from the two field conditions. Septa sampled from polyps growing at depth were considerably longer than those from corallites living on the shallow reef flat (results not shown). We are unsure whether this feature was related to depth or was simply natural morphological variation. No other structural differences were observed between septa sampled from corallites collected from the reef flat and from deeper water. Chemical treatments Commercial bleach was the best chemical treatment for digesting soft tissue from corallites of the scleractinian corals G. fascicularis and A. formosa; dissolution occurred in less than 30 min at 60 C. Washing samples well in running water, followed by brief rinses in dH 2 O, ensured that the septa on the resultant corallites were free of residual tissue (Fig. 4). No intensive cleaning or agitation was nec- essary. Skeletons prepared in bleach were rapidly cleaned, so the crystalline structure of the septa could be readily observed with SEM. All of the crystal types described by Clode and Marshall (2003) (nano, acicular, lamellar, and fusiform) were clearly observable upon septa prepared in this manner. An example of fusiform crystals is shown in Figure 5. NaOH was also very effective in removing soft tissue from G. fascicularis corallites; tissue was rapidly digested at 60 C. However, large, crystalline structures regularly ob- served on septa prepared in NaOH (Fig. 6), were never seen in samples prepared in any other chemical treatment; thus they are probably artefactual. The only way to prevent the appearance of such structures was to ensure that the coral- lites were washed well after treatment in NaOH. with ad- ditional agitation and washing of samples prior to drying. This resulted in septa free from crystalline artefacts. In such cases, nano, acicular, and lamellar crystals were all clearly visible upon the septal surface; however, fusiform crystals were never observed on septa prepared in this manner. In contrast, the use of dH 2 O to remove the overlying soft tissues from G. fascicularis septa was difficult, time con- suming, and often left corallites unclean. The crystalline surface was masked by overlying material and thus could not be viewed with SEM. Furthermore, polyps immersed into dH 2 O secreted excess mucus, which further compli- cated tissue removal. As a result, many G. fascicularis septa prepared in dH 2 O were coated in an amorphous material of unknown composition, which completely obscured the crys- talline structures underneath (Fig. 7). Clean septa free of tissue remnants and this coating could be obtained by ex- tensive rinsing and pipetting of dH 2 O onto the sample. Like septa prepared in NaOH, G. fascicularis septa prepared in dHnO did not possess fusiform crystals. The perforate nature of A. formosa skeletons made treat- ment with dH 2 O much easier and more successful than it was for G. fascicularis. Although gentle agitation was nec- essary to completely remove the soft tissue overlying the axial corallite. the process was considerably less stringent than the rinsing protocol necessary to prepare G. fascicu- laris septa. A. formosa axial corallites prepared in dH 2 O were clean and free of epithelial remnants following gentle agitation, and fusiform crystals were regularly observed along the primary septa extending into the calyx (Fig. 8). Tissue dissolution by H 2 Oi was both difficult and highly ineffective (Fig. 9). G. fascicularis septa, despite being covered by only a thin layer of soft tissue, took several days to prepare, and the process was exceptionally tedious and labor intensive. Extensive additional agitation and washing was required to produce clean septa, so that the crystalline structure could be seen with SEM. With this treatment, like dH^O. excessive secretion of mucus added to the difficulty of obtaining suitable preparations. All crystal types, except fusiform crystals, could be readily observed upon septa that had been adequately prepared in H 2 O 2 and extensively cleaned with dH 2 O. Discussion A comparison of the crystalline structure of exsert septa between colonies growing under aquarium conditions and natural, field conditions has revealed that drastic changes to crystal deposition and skeletal formation may occur in VARIATION IN CORAL SKELETAL MICROSTRUCTURE 141 _ **.. B^r. ^f^f^j^ij^ *y >[ ' '*//'' /' / 3 ^ ' *1 $,4y r ">- -T '>rf*#V*J w-jy * ' <- ' T v>-%c Figure 1. Electron micrographs of cross-fractured CXSLTI sepia from Giilti\\'u fascicularis coiallitcs samplctl (A) directly from the reef flat: and (B) after 26 weeks in a closed-circuit aquarium. Mucus or tissue residue (R) can also be seen in (A) adjacent to the skeleton (Sk). Scale bars: A = 1 /xm; B = 500 nni. Figure 2. Electron micrographs of typical growth surfaces of Ga/axea fascicitlaris exsert septa sampled (A) from the reef flat: and (B) after maintenance in a closed-circuit aquarium for 26 weeks. Scale bars: A = 500 nm; B = 300 nm. Distinct clusters of similarly oriented acicular crystals, known as fasciculi, are clearly evident in septa sampled from the reef flat lA: clusters numbered 1-5) but were not observed on septa sampled from the aquarium (B). Figure 3. Electron micrographs detailing the different nature of acicular crystals at Ihc growing edge of exsert septa of Galaxea fasciculuris sampled (A) on the reef flat: and (B) at a depth of 9 m. Scale bars: A = 250 nm; B = 400 nm. 142 P. L. CLODE AND A. T. MARSHALL Figure 4. Electron micrograph detailing the crystalline structure of a Galaxeafascicularis exsert septum that was readily observed following tissue digestion in commercial bleach. Scale bar = 10 nm. Figure 5. Electron micrograph of fusiform crystals (*) evident along the lateral edge of a Gulaxea fascicularis exsert septum prepared in commercial bleach. Scale bar = 1 /im. Figure 6. Electron micrograph of crystalline artefacts commonly observed on Galuxi'ii fascicitlurix exsert septa prepared in NaOH. Scale bar = 1 juni. F'igure 7. Electron micrograph of amorphous remnants that regularly covered the crystalline surface of Galaxeafascicularis exsert septa prepared in distilled water. Scale bar = 5 fj.m. Figure 8. Electron micrograph of an Acropora formosa axial corallite prepared in distilled water, showing fusiform crystals (*) upon a primary septum. Scale bar = I jum. Figure 9. Electron micrograph of tissue-like remnants that regularly remained attached to Gulaxeu fascicu- laris exsert septa after preparation in hydrogen peroxide. Scale bar = 5 nm. VARIATION IN CORAL SKELETAL MICROSTRUCTURE 143 corals maintained for long periods in closed-circuit aquaria. In addition, preparing such samples in commercial bleach may result in the complete loss of fusiform crystals from the skeletal surface. Comparison of field and aquarium samples It is evident that calcification in scleractinian corals can be severely affected by changes in the surrounding environ- ment. Much recent attention has focused upon how major changes in environmental conditions such as atmospheric CO 2 levels and water temperature affect coral calcification (Done. 1999; Kleypas el ai, 1999; Pittock, 1999). Our results indicate that even relatively small modifications to the surrounding environment, particularly those that are artificially generated, may significantly alter the pattern of crystal deposition and the rate of calcification in captive corals. Variability in the growth rate of aquarium-kept cor- als may be influenced by, and dependent upon, what are often considered minor details, such as water quality and movement, food availability (Mortensen, 2001), and subtle changes in light intensity and wavelength (see Carlson, 1999). Acicular crystal growth on the exsert septa of Galaxea fascicularis is probably continuous, as evidenced by their lack of crystal substructure (Clode and Marshall, 2003). However, septa sampled from aquarium-maintained polyps did not exhibit this pattern of crystal deposition. Instead, small, randomly oriented, spherulitic crystals covered the entire septal surface. As a result of the discontinuous, un- ordered growth, skeletal porosity was likely to increase. When compared with those growing in their natural envi- ronment, a variety of scleractinian corals show a significant reduction in skeletal porosity following containment within an aquarium (see Carlson, 1999). The fact that the region of disoriented growth in G. fascicularis extends to an approximate depth of only 1 /xm in septa sampled from corals kept in aquaria for 26 weeks suggests that calcification and growth rates of these corals may be significantly reduced, with drastic changes in calci- fication patterns also expected. In many branching sclerac- tinian corals, gross colony morphology changes dramati- cally following maintenance in aquaria (see Carlson, 1999), with colonies developing unnatural morphological traits. In addition to directly affecting the physical processes of coral calcification, suboptimal conditions may also invoke changes indirectly through stress responses. Colonies main- tained under aquarium conditions appeared to lose a large proportion of their symbiotic algae. Loss of zooxanthellue from gastrodermal cells in times of stress is common to many zooxanthellate scleractinian corals (Hoegh-Guldberg and Smith, 1989; Brown el al., 1995), and calcification rates can be greatly reduced through the induced expulsion of these symbionts (Goreau, 1959). Biomineralization anoma- lies also occur in symbiont-bearing foraminifera exposed to stressful environmental conditions (Toler and Hallock, 1998). These results raise concerns about the validity of exper- iments, particularly those investigating calcification rates, etc., with corals kept in closed-circuit aquaria, where arti- ficially modified or unnatural environments may have in- duced atypical patterns of calcification. Marked differences in skeletal organization between corals sampled from natu- ral and artificial environments highlight the importance of accurately imitating natural conditions in closed-circuit aquaria where experimental corals are to be maintained. Comparison of the crystal structure of corals living under shallow- and deep-water field conditions revealed only mi- nor differences in septal microstructure. The principal dif- ference in crystal structure between polyps from the two different water depths was that acicular crystals of samples from 9 m were much more finely structured and appeared to have a higher porosity than their counterparts from the reef flat. Skeletal "filling," or secondary thickening (Gladfelter, 1982), was also clearly reduced in septa sampled from corallites collected from the deeper water. A reduced calci- fication rate because of lower light intensities would ac- count for the differences between septa sampled from the two environments. Further research into the calcification rates of such colonies would help to clarify the cause of the differences, as would an accurate determination of porosity. Chemical treatments Despite large variation in skeletal preparatory methods, commercial bleach has been the solvent used most often for digestion of scleractinian coral tissue. Our results confirm that commercial bleach is highly effective in dissolving the soft tissue of the corals G. fascicularis and A. formosa, particularly when it is heated to 60 C. Excellent skeletal preparations for SEM can be rapidly obtained with rela- tively little effort and without discernible deleterious effects upon skeletal microstructure, even if corallites are incubated in bleach for up to 12 h (P. Clode, S. Howe, and A. Marshall, La Trobe University, unpubl. data). We initially believed that the fusiform crystals we saw on G. fascicularis corallites prepared in bleach were artefacts, because they were not present in other preparations. How- ever, the finding that A. formosa axial corallites prepared in dH 2 O also possessed these crystals ruled out the possibility of a chemically induced origin. We conclude that the addi- tional cleaning of corallites with jets of water and mechan- ical agitation, which was necessary for corallites prepared in either NaOH, H 2 O 2 , or dH 2 O, removed the fusiform crystals from exposed growing edges. 144 P. L. CLODE AND A. T. MARSHALL In A. fonnosci corallites prepared in both bleach and dH,O, fusitbriii crystals were observed along the primary septa extending into the calyx, areas protected by the wall of the coralhte. Furthermore, the widely spaced skeletal ele- ments and highly perforate nature of the skeleton, which is typical of white-tipped A. formosa axial corallites (Oliver, 1984), made the removal of soft tissue easier and the need for extensive cleaning unnecessary. Fusiform crystals ob- served on axial spines of Acropora cervicornis (Gladfelter, 1982) were not seen in A. formosa. It is possible that the fusiform crystals on these exposed areas were lost during preparation. The suitability of NaOH for dissolving soft tissue from G. fascicularis corallites is unquestioned, with activity at 60 C highly effective. However, generation of artefacts and the loss of fusiform crystals from skeletal preparations, which was also noted by Isa (1986) in Acropora hebes axial corallites prepared in NaOH, limits the applications of this chemical for coral skeleton preparation. The use of dH^O to remove soft tissue was thought to be ideal for maintaining crystal structure, due to the low water solubility of CaCO 3 and the unlikelihood of artefact gener- ation. However, we found the method to be time-consuming and difficult, and thus impractical for use with many species of corals. For perforate corals such as A. formosa. dH-,0 can be effectively applied to remove much of the soft tissue. However, without extensive rinsing, many G. fascicularis septa prepared in dH 2 O remained coated in a film of un- known material that completely obscured the skeletal struc- ture beneath. The lack of any discernible features in this film suggest that it was not epithelial in nature, but was perhaps either a thin layer of mucus that was subsequently removed in well-washed preparations or a remnant of the mesogloea, as suggested by Hidaka (199 la). Similarly, the use of H^O, to investigate the skeletal microstructure of G. fascicularis exsert septa was inappro- priate and labor intensive. As with dH^O and NaOH, addi- tional cleaning and agitation of samples resulted in the loss of fusiform crystals from the skeletal surface. An additional disadvantage is that HiO-, may begin to actively dissolve the CaCO, skeleton of scleractinian corals (Mitsuguchi ct ai. 2001). Our investigations provide an overview of chemical treat- ments and their suitability for dissolving soft tissue from coral skeleton. We have found that the suitability of these chemicals is species-specific. For example. NaOH is not suitable for use with the azooxanthellate scleractinian corals Tubastrea faulkncri and Deihlrophvllia sp. or the zooxan- thellate scleractinian coral Seriatopora h\stri\ (A. Marshall and P. Clode, La Trobe University, unpubl. data). These findings may help to explain why several different tech- niques have been used in the past and may also account for some of the contradictory results that can be found in the literature on the skeletal and crystal structure of corals. Acknowledgments This research was conducted with the assistance of an Australian Research Council grant to ATM. All samples were collected under Great Barrier Reef Marine Park Au- thority permits to ATM. We wish to thank the staff at Heron Island Research Station for their services and Ms. (Toilette Bagnato for her assistance with sample collection. Literature Cited Brown, B. E., R. Hewit, and M. D. LeTissier. 1983. The nature and construction of skeletal spines in Pocillopora damicornis (L.). Coral Reefs 2: 81-89. Brown, B. E., M. D. A. LeTissier, and J. C. Bythell. 1995. Mechanisms of bleaching deduced from histological studies of reef corals sampled during a natural bleaching event. Mar. Biol. 122: 655-663. Carlson, B. A. 1999. Organism responses to rapid change: what aquaria can tell us about nature. Am. Zoo/. 39: 44-55. Clode, P. L., and A. T. Marshall. 2003. Skeletal microstructure of Galaxea fascicularis exsert septa: a high-resolution SEM study. Biol. Bull. 204: 146-154. Constantz. B. R. 1989. Skeletal organization in Caribbean Acropora spp. (Lamarck). Pp. 175-199 in Origin. Evolution and Modern Aspects of Biomineralization in Plants and Animals. R. E. Crick, ed. Plenum Press, New York. Done, T. J. 1999. Coral community adaptability to environmental change at the scales of regions, reefs and reef zones. Am. Zool. 39: 66-79. Gludfelter, E. H. 1982. Skeletal development in Acropora cervicornis I. Patterns of calcium carbonate accretion in the axial corallite. Coral Reefs 1: 45-51. Gladfelter, E. H. 1983. Skeletal development in Acropora cenicomis II: Diel patterns of calcium carbonate accretion. Coral Reefs 2: 91-100. Goreau, T. F. 1959. The physiology of skeleton formation in corals. I. A method for measuring the rate of calcium deposition by corals under different conditions. Biol. Bull. 116: 59-75. Hidaka. M. 1988. Surface structure of skeletons of the coral Galuxea fascicularis formed under different light conditions. Pp. 95-100 in Proceedings of the Sixth International Coral Reef Symposium, Vol. 3, J. H. Choat et ai. eds. 6th International Coral Reef Symposium Executive Committee. Townsville, Australia. Hidaka, M. 1991a. Deposition of fusiform crystals without apparent diurnal rhythm at the growing edge of septa of the coral Galuxea tasciciilaris. Coral Reefs 10: 41-45. Hidaka, M. 1991h. Fusiform and needle-shaped crystals found on the skeleton of a coral. Galaxea fascicularis. Pp. 139-143 in Mechanisms iiihl Phytogeny ofMinerali:atinn in Biological Systems. S. Suga and H. Nakahara, eds. Springer Verlag, Tokyo. Hidaka, M., and S. Shirasaka. 1992. Mechanism of phototropism in young corallites of the coral Galaxea fascicularis. J. Exp. Mar. Biol. Ecol. 157: 69-77. Hnegh-Guldherg, O., and G. J. Smith. 1989. The effect of sudden changes in temperature, light and salinity on the pollution density and export of zooxanthellae from the reef corals Stylophora pistillata Esper and Seriatopora hystrix Dana. J. Exp. Mar. Biol. Ecol. 129: 279-303. Isa, V. 1986. An electron microscope study on the mineralization of the skeleton of the staghorn coral Acropora hebes. Mar. Biol. 93: 91-101. VARIATION IN CORAL SKELETAL MICROSTRUCTURE 145 Jell, J. S. 1974. The microstructure of some scleractinian corals. Pp. 301-320 in Proceedings of the Second International Coral Reef Sym- posium, A. M. Cameron, ed. Great Barrier Reef Committee, Brisbane. Australia. Johnston. I. S. 1979. The organization of a structural organic matrix within the skeleton of a reef-building coral. Scanning Electron Mi- crosc. 1979 II: 421-431. Kleypas, J. A., R. W. Buddemeier, D. Archer, J.-P. Gattuso, C. Lang- don, and B. N. Opdyke. 1999. Geochemical consequences of in- creased atmospheric carbon dioxide on coral reefs. Science 284: 118-119. LeTissier, M. D. 1988. Diurnal patterns of skeleton formation in Pocil- lopora Jainicornis (Linnaeus). Coral Reefs 7: 81-88. LeTissier, M. D'A. A. 1990. The ultrastructure of the skeleton and skeletogenic tissues of the temperate coral Caryophyllui viiiihi. ./. Mar. Biol. Assoc. UK 70: 295-310. LeTissier, M. D'A. A. 1991. The nature of the skeleton and skeletogenic tissues in the Cnidaria. Hydrobiologiu 216/217: 397-402. Marshall, A. T., and A. Wright. 1991. Freeze-substitution of sclerac- tinian coral for confocal scanning laser microscopy and X-ray micro- analysis. J. Microsc. 162: 341-354. Mitsuguchi, T., T. Uchida, E. Matsumoto, P. Isdale, and T. K:)\vana. 2001. Variations in Mg/Ca, Na/Ca, and Sr/Ca ratios of coral skeletons with chemical treatments: implications for carbonate geochemistry. Geochim. Cosniocliim. Ada 65: 2865-2874. Mortensen, P. B. 2001. Aquarium observations on the deepwater coral Lophelia pertusa (L., 1758) (Scleractinia) and selected associated in- vertebrates. Ophelia 54: 83-104. Oliver, J. K. 1984. Intra-eolony variation in the growth of Acropora fonnosa: extension rates and skeletal structure of white (zooxanthcllae- free) and brown-tipped branches. Coral Reefs 3: 139-147. Pillock, A. B. 1999. Coral reefs and environmental change: adaptation to whai'Mw, Zool. 39: 10-29. Sorauf, J. E. 1972. Skeletal microstructure and microarchitecture in Scleractinia (Coelenterata). Palaeontology 15: 11-23. Sorauf, J. E. 1974. Observations on microstructure and biocrystulliza- lion in coelenterates. Biomineralization 7: 37-55. Toler, S. K., and P. Hallock. 1998. Shell malformation in stressed Amphistegina populations: relation to biomineralization and paleoen- vironmental potential. Mar. Micropaleontol. 34: 107-115. VVainwright, S. A. 1963. Skeletal organisation in the coral. Poci/lopora itamiconiis. Q. J. Microsc. Sci. 104: 169-183. Reference: Biol. Bull. 204: 146-154. (April 2003) > 2003 Marine Biological Lahoratoi\ Skeletal Microstructure of Galaxea fascicularis Exsert Septa: A High-Resolution SEM Study PETA L. CLODE AND ALAN T. MARSHALL* Analvtical Electron Microscopy Lahorator\, Department of Zoology, La Trobe University. Melbourne, Victoria 3086, Australia Abstract. The deposition of four crystal types at the growth surface of the septa of several color morphs of the coral Gala.\ea fascicularis was investigated over a 24-h period. Results suggest that nanocrystals. on denticles at the apices of exsert septa, may be the surface manifestation ot centers of calcification. These crystals were also found on the septa of the axial corallite of Acropora formosa. The deposition of nanocrystals appears to be independent ot diurnal rhythms. Internally and proximal to the septal api- ces, distinct clusters of polycrystalline fibers originate from centers of calcification and form fanlike fascicles. Upon these fascicles, acicular crystals grow and extend to form the visible fasciculi at the skeletal surface. Deposition of aragonitic fusiform crystals in both G. fascicularis and A. fonnosa occurs without diurnal rhythm. Nucleation of fusi- form crystals appears to be independent of centers of cal- cification and may occur by secondary nucleation. Forma- tion of semi-solid masses by fusiform crystals suggests that the crystals may play a structural role in septal extension. Lamellar crystals, which have not been reported as a com- ponent of scleractinian coral skeletons before, possess dis- tinct layers of polyhedral plates, although these layers also do not appear to be associated with daily growth increments. The relationship of lamellar crystals to other components of the scleractinian coral skeleton and their involvement in skeletal growth is unknown. Introduction The microstructural components of the CaCO, skeleton from a wide range of scleractinian corals have been well documented (see Wainwright. 1963; Vahl, 1966; Sorauf. 1970, 1972. 1974. 1980; Wise. 1970, 1972; Chevalier, Received 6 June 2002; accepted 31 Jan 2003. *To whom correspondence should he addressed. E-mail: /ooanii" /on. laliohe.edu. an 1974; Jell. 1974; Constantz. 1986, 1989). However, descrip- tions of skeletal microstructure are inconsistent, reflecting differences in both interpretation and structural variation. The relationships between the various crystalline micro- structures found on the surface and within the interior of the skeleton are not fully understood, although this is funda- mental to an understanding of the origin of crystal forma- tion, deposition and growth. The basic structure of the coral polyp is a tubelike skel- eton, or corallum, divided by longitudinal and horizontal partitions. Sitting in the top of this tube is the living polyp. The key elements of the corallum are the longitudinal divi- sions, the septa, which are joined laterally by the wall (thecal of the corallum. Those septa that extend above the top of the theca are referred to as exsert septa. Exsert septa are one of the primary sites of CaCO, deposition and skeletal extension in the scleractinian coral Galaxea fas- cicultiris (Marshall and Wright. 1998). These elongated septa protrude upward from the wall of the corallite and encircle the oral disc (see Fig. 1 ). This arrangement allows for individual septa to be easily detached from the corallite, without significant damage, for subsequent investigation ot the crystalline microstructure with scanning electron mi- croscopy (SEM). The internal structure of the coral skeleton has been primarily studied by light microscopy. It was established early (Ogilvie, 1896) that the internal structure of septa is composed of arrays of vertically elongated centers of cal- cification from which polycrystalline fibers extend radially to form fascicles. These fanlike systems are regarded as the basic building blocks of the skeleton. On the external sur- face, two major crystal types have been described. These are clusters of acicular crystals that form fasciculi and spindle- shaped crystals referred to as fusiform crystals. The latter have been suggested to be deposited with a diel periodicity and to be the site of nucleation of fasciculi (Gladtelter, 146 MICROSTRUCTURE OF EXSERT SEPTA 147 Figure 1. A corallite of Galaxea fascicularis, showing exsert septa (*) protruding from the wall of the corallite and encircling the oral disc. Scale bar = 2 mm. Figure 2. Granular nanocrystals located upon the distal growth edge of a Gulnxeu fcucicularis exsert septum. Scale bar = 100 nm. Figure 3. Granular nanocrystals observed upon a septum of an axial corallite of Acropora formosa. Scale bar = 100 nm. Figure 4. Highly ordered acicular crystals at the distal edge of a Galaxea fascicularis exsert septum. Scale bar = 500 nm. Figure 5. Fasciculi, composed of distinct clusters of similarly oriented acicular crystals, constituting much of the distal surface of a Galaxea fascicularis exsert septum. Scale bar = 5 jum. Figure 6. Lamellae crystals located proximal to the distal growth edge of a Galaxea fascicularis exsert septum. Scale bar = 1 /jm. 1982, 1983). Fusiform crystals (Gladfelter, 1983) have been In this study we have investigated, over a 24-h period, the suggested to be calcite, in contrast to the bulk of the skel- crystalline microstructure at the growth surface of exsert eton, which is formed from aragonite. septa from the reef coral G. fusciciilaris. Structural charac- 148 P. L. CLODE AND A, T. MARSHALL teristics of four crystal types, including one crystal type not previously reported in scleractinian corals, are described at a new level, with magnifications greater than 50,000 X achievable by low voltage, high-resolution field emission (FE) scanning electron microscopy. We find no evidence of rhythmic deposition of any crystal types. We also show by X-ray microanalysis that the composition of fusiform crys- tals does not appear to differ from the remainder of the aragonite skeleton. Materials and Methods Coral collection and maintenance Green, yellow, and brown color morphs of colonies of the reef coral Galaxea fascicularis L. (different from the Japa- nese morphs described by Hidaka and Yamazato (1985)), and white-tipped branches of the reef coral Acropora for- mosa (Dana) were collected at low tide from the reef flat at Heron Reef in the Capricorn Bunker Group of the Great Barrier Reef, Australia. The corals were transported in buckets of seawater to the Heron Island Research Station, where they were maintained in sunlit, well-aerated flow- through aquaria in natural seawater at 24-25 C. Following collection, the corals were allowed to recover for at least 2 days before being used for experimentation. On occasion, individual G. fascicularis polyps and A. formosa axial branches were sampled directly from the reef flat and im- mediately placed into the appropriate chemical treatment. G. fascicularis polyps of the green color morph were rou- tinely used for all experiments. Sample preparation for field emission scanning electron microscopy Individual G. fascicularis polyps were sampled from colonies over a 24-h period (0600, 1200, 1800, and 2400 h; n = 5 for each time period). Individual polyps were easily separated from colonies with forceps, as the fragile co- enosteum joining individual polyps could be removed with- out damaging the polyp itself. Axial tips of A. formosa branches were sampled at 1200 h (n = 6) and 2400 h (n = 6). All samples were placed in 12% NaOCl (commercial bleach) at 60 C for 30 min, and the resultant corallites were rinsed well in running water and then in distilled water (dH,O) several times. Any tissue remaining on the corallite was removed by gentle agitation and pipetting of dH 2 O onto the sample, before being dried at 60 C for 24 h (Clode and Marshall. 2003b). G. fascicularis exsert septa (four septa from each polyp) were removed from the corallite with forceps, under a dissecting microscope, and mounted flat using carbon tape. Septa were coated with 5 nm platinum and previewed in a JEOL JSM 840A scanning electron microscope at 10 kV. High-resolution imaging was conducted on a JEOL 6340-F field emission (FE) scanning electron microscope at 2 kV. A. formosa branch tips (n = 12) were secured upright in hollow stubs with partially polymerized araldite, so that the axial polyp extended about 3 mm above the upper surface of the stub. Polymerization was then completed at 60 C for a further 30 h. Conductive silver epoxy (ProSciTech) was used to improve the conductivity of the upright corallite, before it was coated with 10 nm platinum. Axial polyps were viewed in a JEOL JSM 6340-F FE scanning electron microscope at 1 kV and 2 kV. All size measurements were obtained using the computer software package UTHSCSA Image Tool ver. 1.23 (Uni- versity of Texas). All statistical analyses were performed using the computer software package JMP ver. 3.1.6 (SAS Institute, Inc). X-ray microanalysis For comparative elemental analyses of fusiform crystals and typical skeleton, G. fascicularis septa were mounted flat as described above, coated with 200 A Al. and analyzed by X-ray microanalysis in a JEOL JSM 840A SEM fitted with a Link exL X-ray analyzer (Oxford Instruments). The ana- lyzer was equipped with an LZ5 light element detector with a takeoff angle of 40. Selected area analyses were con- ducted at 15 kV and a beam current of 2 X 10 10 A, from an area of 1 /xm X 1 /u,m, for 100 s livetime. Element concentrations were calculated against microprobe refer- ence standards (BioRad) using the PhiRhoZ model (Oxford Instruments) (Marshall, 1982; Marshall and Condron, 1987), and element ratios calculated. Because the X-rays from elements of interest could be generated from a depth of up to 2 /am at 15 kV, only large fusiform crystals were analyzed, reducing the likelihood that extraneous X-rays would be derived from skeleton below the crystal itself and affect the element ratios. The areas selected for analysis were horizontal relative to the X-ray detector. Transverse slices Small G. fascicularis polyps were rapidly frozen in liquid propane (-180 C) and freeze-substituted in a mixture of 10% acrolein in diethyl ether, according to the protocol outlined by Marshall and Wright (1991). Transverse slices of freeze-substituted material, 400 jam thick, were prepared with a diamond saw (see Marshall and Wright, 1991), attached to glass slides with araldite, and then polished with aluminium oxide. The polished slices were rinsed in dH 2 O and air-dried. Samples were viewed unmounted on a Zeiss Axioskop microscope with polarized light. Results Crystal types Using low voltage, high-resolution FESEM, we identified four principal crystal forms on the growth surface of exsert septa sampled from polyps of Gala.xea fascicularis. These MICROSTRUCTURE OF EXSERT SEPTA 149 four crystal forms were nuno, aeicular, lamellar, and fusi- form. All four types were common to septa collected at 0600, 1200. 1800. and 2400 h, and all appeared consistently similar in structure across the four sampling periods. Sim- ilarly, no notable differences were observed in the crystal structure of septa sampled from different color morphs, nor were there differences between corallites sampled directly from the reef flat and those allowed to recover in aquaria at Heron Island Research Station for 2 days. Granular nanocrystals were commonly observed at the actively growing distal tip of G. fascicitlaris exsert septa predominantly on denticles. These crystals appeared as small, clustered groups of rounded crystals that exhibited little order in orientation or pattern of deposition (Fig. 2). Nanocrystals were also observed on septal surfaces of A. fonnosa axial corallites (Fig. 3). Nanocrystals were highly variable in size: the smallest resolvable crystals averaged 19 0.8 nm (n = 12) in diameter, and the largest were about 400 nm in diameter. Acicular crystals were the predominant crystal form on the surface of G. fascicularis exsert septa. These crystals were evident over much of the septal surface and extended perpendicular to the plane of the skeletal surface. Acicular crystals were typically large, solid crystals elongated along the c axis (Fig. 4), although smaller and more needlelike crystals were also observed. In contrast to nanocrystals, individual acicular crystals were elongated and exhibited a high degree of order and orientation. Groups of acicular crystals growing parallel to each other extended from an unseen origin, which was presumably the underlying fasci- cles. This arrangement resulted in the appearance of distinct clusters of similarly oriented acicular crystals termed fas- ciculi, visible at the skeletal surface (Fig. 5). No notable growth increments were evident within individual acicular crystals, suggesting a pattern of continuous growth. Lamellar structures were typically observed in positions proximal to the extending distal edge of G. fascicularis septa. These crystals were similar to acicular crystals in that initially, at low magnification, they appeared to be large, elongate crystals extending perpendicular to the c axis. However, at high magnification it became evident that these were not single crystals, but layers of polyhedral plates resembling tabular crystals, which formed lamellar-like stacks (Fig. 6). These crystal stacks were distinctly different from acicular crystals: the apparent continuous nature of acicular crystal growth contrasted with the formation of distinct layers and the obvious discontinuous pattern of crystal deposition in lamellar stacks. Individual crystal lay- ers within these stacks were less than 100 nm in thickness, whereas the crystal stacks themselves were highly variable in both height and diameter. Fusiform crystals were observed principally along the lateral edges of G. fascicularis exsert septa (Fig. 7) and upon A. fonnosa primary septa extending into the calyx of axial corallites (Fig. 8). These crystals were regularly ob- served on all coral samples, regardless of time of sampling. Fusiform crystals appeared as large, tapered structures that were usually clustered together to form a semisolid. crys- talline mass along the lateral edges of the septa (Fig. 7). In G. fascicularis. these crystals averaged 4.6 0.2 /im in length and 2.3 0.1 jiim in width ( 28). Fusiform crystals observed on A. fonnosa axial corallites were sig- nificantly shorter (3.7 0.2 /xm; n == 28: P < 0.01: Student's rtest) and narrower (1.6 0.1 /nm: n = 28; P < 0.0001 : Student's ; test) than those on G. fascicularis septa. Using high-resolution FESEM, we determined that fusi- form crystals were not monocrystalline, but were instead composed of small, polycrystalline aggregates 21 0.8 nm (n = 20) in diameter (Fig. 9). These spherical crystals upon the surface of fusiform crystals were not dissimilar to the nanocrystals at the growth edge. Needlelike crystals were never observed upon, or seen to be extending from, the surface of fusiform crystals in either G. fascicularis or A. fonnosa. The ratios of the three major skeletal elements, Ca, Sr, and Mg, present in individual fusiform crystals and typical skeleton of G. fascicularis exsert septa, determined by X- ray microanalysis. revealed that fusiform crystals were of a very similar elemental composition to the main skeletal component. Differences (P > 0.05; Student's t test) ob- served in the Ca:Mg ratio, the Sr:Mg ratio, and the Ca:Sr ratio between fusiform crystals and skeleton were highly insignificant (Table 1 ). No differences were observed, with respect to microstruc- ture or chemical composition, between color morphs of G. fascicularis or between corals processed immediately on collection from the reef and corals processed after being kept in aquaria for 2 days after collection. Freeze-substiluted transverse slices The major structural components of septa were clearly visible in transverse sections of whole freeze-substituted G. fascicularis polyps visualized with polarized light (Fig. 10). Centers of calcification, which appeared as distinctly darker (denser) regions, were evident along the midline of each septum. The closeness of the centers to each other and the thickness of the section made it difficult in a single focal plane to resolve the centers as separate structures: this was possible, however, when the plane of focus was changed. The centers possessed a granular substructure, but again, because of the thickness of the section, this was difficult to illustrate photographically. These centers of calcification extended along the central region of each septum, ceasing just short of the lateral edges. From these centers of calci- fication, highly ordered fascicles with distinct orientations radiated outwards to form fanlike systems (Fig. 10). Bun- dles of acicular crystals that form fasciculi at the septal surface cannot be visualized. 150 P. L. CLODE AND A. T. MARSHALL #** ' ; ,._*'*-- *"' '- Figure 7. Clusters of fusiform crystals forming a semisolid crystalline mass along the lateral edge of a Galaxe a fascicularis exsert septum. Scale bar = 1 jiun. Figure 8. Fusiform crystals (*) upon a primary septum extending into the calyx of an Acropora formosa axial corallite. Scale bar = 1 fjm. Figure 9. An isolated fusiform crystal from a Calaxea fascicularis exsert septum, shown in increasing magnifications. At high magnification, the surface of the tapered polycrystallite is seen to be covered in small, spherulitic crystals. Scale bars: A = 1 /nm: B = 200 nm: C = 100 nm. Figure 10. Transverse section through a freeze-substituted Galaxea fascicularis septum, viewed with polarized light, showing granular centers of calcification (C) and centrically arranged fascicles (*) radiating from these centers of calcification to form fanlike trabeculae. An array of trabeculae forms the entire septum. Scale bar = 100 jiim. Discussion The most significant finding of this investigation is the presence of nanocrystals at the major growth points the denticles on the skeletal septa of Galaxea fascicitlari*. These nanocrystals may represent nucleation sites for the deposition of acicular crystals that ultimately form fanlike systems, or fascicles, of polycrystalline fibers, which are the major building blocks of the skeleton. Also found on septa were clusters of acicular crystals forming fasciculi, fusiform MICROSTRUCTURE OF EXSERT SEPTA 151 Table 1 Comparison of the ratios of the primary skeletal elements present in individual fusiform crystals and typical skeleton from Galaxea fascicularis exsert septa, as determined by X-ray microanalysis n Ca:Mg Sr:Mg Ca:Sr Skeleton 6 143:1 2:1 76:1 Fusiform crystals 7 134:1 1.9:1 76:1 P Value* 0.77 0.82 0.99 * Student's / test: n = number of analyses. crystals, and a novel lamellar type of crystal. None of these crystals appeared to be deposited in a diurnal rhythm. The structure of the exsert septum and the structure and location of the various crystal types is summarized in Figure 1 1 . No differences were observed between different color morphs or between corals processed immediately on collection and corals maintained in aquaria for 2 days before processing. Crystal types Granular nanocrystals as small as 19 nm in diameter were observed by FESEM on the apical denticles of G. fascicu- laris septa. These nanocrystals have not been previously described. They may be the basic elements of centers of calcification, since denticles are the terminations of trabec- ular axes, which are extended centers of calcification in the septa (Ogilvie, 1896). Nanocrystals were also observed on the septa of the axial corallites of A. fonnosa. The nano- crystals appear to have some similarity to the crystals form- ing the nuclear packets described by Constantz (1989) and to the granulated crystallites upon the surface of "spherular crystals" noted by Isa (1986). Constantz suggested that Figure 11. Diagrammatic representation of an exsert septum of Galaxea fasicu/aris showing the internal structure and distribution of crystal types. The internal structure is based partially on Ogilvie (1896). The diagram represents the upper part of the exsert septum with a transverse slice removed. The exsert septum is orientated so that the interior of the corallum would be to the left. Centers of calcification are depicted in black and terminate at denticles. Fibers form systems of trabecullae around the centers of calcification. New centers of calcification appear as the exsert septum extends. The different crystal types are drawn approximately to scale relative to each other, and their typical locations are indicated. Granular nanocrystals ( 1 ) are located at the growth edge (see Fig. 2). Acicular crystals (2) are widely distributed on the septal surface (see Figs. 4 and 5). Fusiform crystals (3) are found at the lateral edge (see Fig. 7), and lamellar crystals (4) are located close to the distal growth edge of the exsert septum (see Fig. 6). 152 P. L. CLODE AND A. T. MARSHALL nuclear packets were centers of calcification, describing them as small clusters of tiny crystals less than 100 nm in size that existed in high frequency near rapidly growing regions. The mechanisms involved in the formation and deposi- tion of the nanocrystals, so that they may act as nucleating centers for future crystal growth, remains unknown. There is evidence, however, that small nascent crystals of CaCO, may develop upon a fibrillar organic matrix, which is evi- dent within small pockets formed between calicoblastic ectodermal cells and the pre-existing skeleton (Clode and Marshall. 2003a). Clusters of acicular crystals form the distinctive fasciculi, which are visible on the surface of septa in G. fascicularis. The surface of coral skeleton is frequently characterized by groups of nearly parallel acicular crystals termed fasciculi (Wise. 1972). Depending upon the orientation of the crys- tals within the fasciculi, the skeletal surface may appear to be granular or relatively smooth. The acicular crystals pre- sumably nucleate and extend from the apical edges of fascicles. Fascicles are fanlike systems of polycrystalline fibers radiating from centers of calcification (Ogilvie, 1896). The relationship between the smaller fasciculi and the larger underlying fascicles is not clear, since individual fasciculi cannot be readily recognized below the skeletal surface (Jell, 1974). Presumably, fasciculi give rise to the underly- ing fascicles; however, fascicles are also present in corals that do not have fasciculate skeletal surfaces (Wise. 1972). Fusiform crystals were predominantly observed along the lateral edges of the septa. The term "fusiform" was first coined by Gladfelter ( 1982, 1983) to describe large, tapered crystals found on the growing surface of Acropora cen'i- cornis axial corallites. Hidaka (1988) also employed the term to describe similar crystals observed on G. fascicularis exsert septa. Our observations on the size and shape of fusiform crystals from the septa of both A. fonnosa and G. fascicularis are consistent with these studies. The signifi- cant size differences evident between the fusiform crystals of G. fascicularis and A. fonnosa are likely to reflect dif- ferences in polyp size, with G. fascicularis polyps consid- erably larger than those of A. fonnosa. Earlier studies have also reported similar crystals, but these were described as "equant" crystals (see Constantz, 1989), while Isa (1986) preferred to use the term "spindle-shaped crystals." Le- Tissier (1988) also reported fusiform crystals upon the surface of Pocillopora damicornis corallites, but these lacked the characteristic tapered ends and may be a different crystal type. Fusiform crystals on G. fascicularis septa were typically observed at the lateral edges where centers of calcification do not persist (Cuif and Dauphin, 1998). This is consistent with the suggestion of Constantz (1989) that centers of calcification were not required for nucleation and growth of fusiform crystals. Fusiform crystal formation may result from secondary nucleation. which can occur due to the presence of already existent crystal structures (Simkiss, 1986). Upon the surface of fusiform crystals it was possible to resolve small spherical nanocrystals that were, on average, 21 nm in diameter. Isa (1986) reported that the surface structure of spindle-shaped (fusiform) crystals was com- posed of clusters of small, rounded crystals less than 50 nm in size, indicating that fusiform crystals were polycrystal- line in nature. Isa (1986) also found that the fusiform crystals were hollow; however, the preparations had been treated with osmium tetroxide, which will react with CaCO, to cause dissolution and recrystallization. We observed no evidence of acicular crystal growth upon individual fusiform crystals, contrary to the proposal of Gladfelter ( 1982. 1983) that clusters of needlelike crystals extended from fusiform crystals to ultimately form fascic- uli. Instead, large clusters of fusiform crystals were typi- cally cemented together to form a semisolid crystalline mass, a feature also noted by Hidaka (1991b). which bore little resemblance to the distinctive fasciculi. In addition, fasciculi, which were common to the entire septal surface, may be spatially isolated from fusiform crystals, which were typically confined to the distal (Hidaka, 1991a) or lateral edges. Hidaka (1991b) also recognised this paradox and suggested that fasciculi may form in several different ways. Variability in the reported distribution of fusiform crys- tals on septa (Gladfelter, 1982, 1983; Hidaka, 1991a,b; Hidaka and Shirasaka, 1992) has made interpretation and understanding of crystal deposition and skeletal extension in corals difficult. Reasons for these reported differences are unknown, but preparatory techniques and environmental conditions may have significant effects upon skeletal mi- crostructure (Carlson. 1999; Clode and Marshall, 2003b). To our knowledge, lamellar crystals have not been re- ported as a component of any recent scleractinian coral skeleton. There is some suggestion that lamellar structures are existent in hydrozoans and tabulate and rugose antho- zoans (see Wendt, 1990), although these appear to refer more to the orientation of fibrillar-type crystals than to true crystalline stacks of polyhedral plates. Lamellar stacks com- posed of polyhedral plates are very common in molluscs (Watabe and Dunkelberger, 1979), particularly in Nautilus shell nacre (Gregoire, 1987). While molluscan lamellar structures may be either aragonite or calcite, lamellar crys- tals on mature G. fascicularis skeletons are likely to be aragonitic, as calcite persists only in the developing skeletal elements of coral larvae. The function of these lamellar structures in scleractinian corals is unknown, as is their relationship to other crystal types and their involvement in the overall extension and growth of skeletal elements. Compositional analysis of fusiform crystals X-ray microanalysis of individual fusiform crystals sug- gests that fusiform crystals are identical to skeleton in MICROSTRUCTURE OF EXSERT SEPTA 153 element composition; therefore, they are aragonitic and not calcitic in nature. Constantz (1989) also suggested that fusiform crystals were likely to be composed of aragonite. although he provided no supporting evidence. Gladfelter (1982). using X-ray microanalysis of large areas of the skeleton of Acropora cen'icornis, found that Mg concen- trations were higher in areas where fusiform crystals were common than in other regions of the skeleton. Since calcite has a higher proportion of Mg than aragonite, it was sug- gested that fusiform crystals were composed of calcite. However, under these circumstances, it would be impossi- ble to determine exactly what was analyzed, with the pres- ence of fusiform crystals in each region of analysis not confirmed. likely to represent growth increments; however, as no in- termediate stages of deposition were observed over the 24-h sampling period, these layers do not appear to be associated with a daily pattern of crystal deposition and growth. Acknowledgments This research was conducted with the assistance of an Australian Research Council grant to ATM. All samples were collected under Great Barrier Reef Marine Park Au- thority permits to ATM. We wish to thank the staff at Heron Island Research Station for their services, Ms (Toilette Bag- nato for her assistance with sample collection and prepara- tion and Mr Alan Jacka for polishing sliced material. Diurnal rli\tluus All four crystal types found on the exsert septa of G. fascicularis were present and remained similar in structure and disposition, regardless of time of sampling over a 24-h period. Apparent diurnal rhythms of crystal deposition have been reported in Plesiastrea versipora (Howe and Marshall, 2002), Acropora cen'icornis (Gladfelter, 1983), Pocillo- pora damicornis (LeTissier, 1988), and Manicina areolata (Barnes, 1972). Hidaka (1988) initially reported a diurnal pattern of fusiform crystal deposition in the exsert septa of G. fascicularis corallites, but he later retracted this interpre- tation in favor of crystal deposition being without rhythm (Hidaka, 199 la). Similarly, we report that A. formosa axial corallites, whether sampled at 1200 or 2400 h, possessed fusiform crystals along the primary septa extending into the calyx. This finding is not in accordance with that of Glad- felter (1983). who only observed fusiform crystals upon axial corallites of A. cen'icornis branches sampled in dark- ness. Gladfelter (1982. 1983) hypothesized that fusiform crys- tals form a loose scaffolding on the surface of exsert septa at night and that acicular crystals nucleate on the fusiform crystals during the day, ultimately giving rise to fasciculi. This diel cycle of deposition of fusiform crystals was pro- posed to account for skeletal extension in zooxanthellate corals at night. However, diel deposition of fusiform crys- tals is apparently not a universal phenomenon (e.g., Hidaka, 1991a), and such crystals are not present in all corals (e.g., Howe and Marshall. 2002). The universal presence of acicular crystals as a predom- inant component of scleractinian coral skeletons during both day and night, in combination with their lack of discernible substructure, suggests that the growth of these crystals is continuous. Whether the rate of crystal extension and growth varies throughout the day is unknown, but diurnal variations in skeletal extension have been reported (Barnes and Crossland, 1980). In contrast, the presence of distinct layers within lamellar stacks suggests an intermittent, highly regulated process of crystal deposition. Each layer is Literature Cited Barnes, D. J. 1972. The structure and formation of growth ridges in scleractinian coral skeletons. Proc. R. Soc. Land. B 182: 331-350. Barnes, D. J., and C. J. Crossland. 1980. Diurnal and seasonal varia- tions in the growth of a staghorn coral measured by time lapse pho- tography. Limnol. Oceanogr. 25: 1 1 13-1 1 17. Carlson, B. A. 1999. Organism responses to rapid change: what aquaria can tell us about nature. Am. Zoo/. 39: 44-55. Chalker, B. E. 1976. Calcium transport during skeletogenesis in herma- typic corals. Comp. Biochem. Physio/. 54A: 455-459. Chevalier, J. P. 1974. On some aspects of the microstructure of recent scleractinia. Pp 345-35 1 in Proceedings of the Second Internationa/ Coral Reef Symposium, A. M. Cameron, ed. Great Barrier Reef Com- mittee. Brisbane, Australia. Clode, P. L., and A. T. Marshall. 2003a. Calcium associated with a fibrillar organic matrix in the scleractinian coral Galaxea fascicularis. Protoplasma 220:153-161. Clode, P. L., and A. T. Marshall. 2003b. Variation in skeletal micro- structure of the coral Gala.\ea fascicularis: effects of an aquarium environment and preparatory techniques. Biol. Bull. 204: 138-145. Constantz, B. R. 1986. Coral skeleton construction: a physiochemically dominated process. Palios 1: 152-157. Constantz, B. R. 1989. Skeletal organization in Caribbean Acropora spp. (Lamarck). Pp. 175-199 in Origin. Evolution and Modern Aspects of Biomineraliz&tion in Plants and Animals, R. E. Crick, ed. Plenum Press, New York. Cuif, J. P., and Y. Dauphin. 1998. Microstructural and physico-chem- ical characterization of 'centers of calcification' in septae of some recent scleractinian corals. Palaeomol. Z. 72: 257-270. Gladfelter, E. H. 1982. Skeletal development in Acropora cervicornis I. Patterns of calcium carbonate accretion in the axial corallite. Coral Reefs 1: 45-51. Gladfelter, E. H. 1983. Skeletal development in Acropora cen'icornis II: Diel patterns of calcium carbonate accretion. Coral Reefs 2: 91-100. Gregoire, C. 1987. Ultrastructure of the Nautilus shell. Pp. 463-486 in Nautilus: The Biology and Pa/eobio/ogv of a Living Fossil, W. B. Saunders and N. H. Landman, eds. Plenum Press, New York. Hidaka, M. 1988. Surface structure of skeletons of the coral Galaxea fascicularis formed under different light conditions. Pp. 95-100 in Proceedings of the Sixth International Coral Reef Symposium, Vol. 3, J. H. Choat et al.. eds. 6th International Coral Reef Symposium Executive Committee, Townsville, Australia. Hidaka, M. 1991a. Deposition of fusiform crystals without apparent diurnal rhythm at the growing edge of septa of the coral Galaxea fascicularis. Coral Reefs 10: 41-45. Hidaka, M. 1991b. Fusiform and needle-shaped crystals found on the skeleton of a coral, Galaxea fascicularis. Pp. 139-143 in Mechanisms 154 P. L. CLODE AND A. T. MARSHALL and Phytogeny of Mineralization in Biological Systems. S. Suga and H Nakahara, eds. Springer Verlag, Tokyo. Hidaka, M., and S. Shirasaka. 1992. Mechanism of phototropism in young coralhtes of the coral Galaxea fascicularis. J. Exp. Mar. Bio/. Ecol. 157: 69-77. Hidaka. M.. and K. V'amazato. 1985. Color morphs of Galaxea fas- cicularis found in the reef around the Sesoko Marine Science Center. Galaxea 4: 33-35. Howe, S. A., and A. T. Marshall. 2002. Temperature effects on calci- fication rate and skeletal deposition in the temperate coral, Plesiasirea rersipora (Lamarckl. J. Exp. Mar. Biol. Ecol. 275: 63-81. Isa, Y. 1986. An electron microscope study on the mineralization of the skeleton of the staghorn coral Acropora hebes. Mar. Biol. 93: 91-101. Jell, J. S. 1974. The microstructure of some scleractinian corals. Pp. 301-320 in Proceedings of the Second International Coral Reef Sym- posium, A.M. Cameron, ed. Great Barrier Reef Committee, Brisbane, Australia. LeTissier, M. D. 1988. Diurnal patterns of skeleton formation in Pocil- lopora damicornis (Linnaeus). Coral Reefs 7: 81-88. Marshall, A. T. 1982. Application of (pzl curves and a windowless detector to the quantitative X-ray microanalysis of frozen-hydrated bulk specimens. Scanning Electron Microsc. 1982 I: 243-260. Marshall, A. T., and R. J. Condron. 1987. A simple method of using 4> (/):) curves for the X-ray microanalysis of frozen-hydrated biological bulk samples. Micron Microsc. Acta 18: 23-26. Marshall, A. T., and A. Wright. 1998. Coral calcification: autoradiog- raphy of a scleractinian coral Galaxea fascicularis after incubation in 45 Ca and I4 C. Coral Reefs 17: 37-47. Marshall, A. T., and O. P. Wright. 1991. Freeze-substitution of scler- actinian coral for confocal scanning laser microscopy and X-ray mi- croanalysis. J. Microsc. 162: 341-354. Ogilvie, M. M. 1896. Microscopic and systematic study of madreporar- ian types of corals. Phil. Trans. 187: 85-345. Simkiss, K. 1986. The processes of biomineralization in lower plants and animals an overview. Pp. 19-37 in Biomineralization in Lower Plants and Animals, B. S. C. Leadbeater and R. Riding, eds. Published for Systematics Association by Clarendon Press. Oxford. Sorauf, J. E. 1970. Microstructure and formation of dissepiments in the skeleton of the recent Scleractinia (hexacorals). Biomineralization 2: 1-22. Sorauf, J. E. 1972. Skeletal microstructure and microarchitecture in Scleractinia (Coelenterata). Palaeontology 15: 11-23. Sorauf, J. E. 1974. Observations on microstructure and biocrystalliza- tion in coelenterates. Biomineralization 7: 37-55. Sorauf, J. E. 1980. Biomineralization. structure and diagenesis of the coelenterate skeleton. Acta Palaeontol. Pol. 25: 327-343. Vahl, J. 1966. Sublichtmikroskopische Untersuchungen der Kristallinen Grundbauelemente und der Matrixbeziehung zwischen Weichkorper und Skelett an Caryophyllia Lamarck 1801. Z. Morphol. Okol. Tiere 56: 21-38. Wainwright, S. A. 1963. Skeletal organisation in the coral. Pocillopora damicornis. Q. J. Microsc. Sci. 104: 169-183. Watabe, N., and D. G. Dunkelberger. 1979. Ultrastructural studies on calcification in various organisms. Scanning Electron Microsc. 1979 II: 403-415. Wendt, J. 1990. Corals and coralline sponges. Pp. 45-66 in Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends, J. G. Carter, ed. Van Nostrand Reinhold. New York. Wise, S. W. 1970. Scleractinian coral skeletons: surface microarchitec- ture and attachment scar patterns. Science 169: 978-980. Wise, S. W. 1972. Observations of fasciculi on developmental surfaces of scleractinian exoskeletons. Biomineralization 6: 160-175. OUTCOMES OF GENOME- GENOME INTERACTIONS Proceedings of a workshop sponsored by THE CENTER FOR ADVANCED STUDIES IN THE SPACE LIFE SCIENCES AT THE MBL 1 to 3 May 2002 J. Erik Jonsson Center for the National Academy of Sciences, Woods Hole, Massachusetts Funded by THE NATIONAL AERONAUTICS AND SPACE ADMINISTRATION under Cooperative Agreement NCC 2-1266. Reprints of this symposium can be obtained from the Center for Advanced Studies in the Space Life Sciences, 7 MBL Street, Woods Hole, MA 02543. CONTENTS Outcomes of Genome-Genome Interactions Sogin, Mitchell, and Diana E. Jennings Introduction . 159 Des Marais, David J. Biogeochemistry of hypersaline microbial mats illus- trates the dynamics of modern microbial ecosystems and the early evolution of the biosphere 160 Spear, John R., Ruth E. Ley, Alicia B. Berger, and Norman R. Pace Complexity in natural microbial ecosystems: the Guerrero Negro experience 168 Vallino, Joseph J. Modeling microbial consortiums as distributed met- abolic networks 1 74 Edwards, Katrina J., Wolfgang Bach, and Daniel R. Rogers Geomicrobiology of the ocean crust: a role for che- moautotrophic Fe-bacteria 180 Teske, Andreas, Ashita Dhillon, and Mitchell L. Sogin Genomic markers of ancient anaerobic microbial pathways: sulfate reduction, methanogenesis, and methane oxidation 186 Fuhrman, J. A., and M. Schwalbach Viral influence on aquatic bacterial communities. . . 192 Polz, Martin F., Stefan Bertilsson, Silvia G. Acinas, and Dana Hunt A(r)Ray of hope in analysis of function and diversity of microbial communities 196 Foster, Jamie S., Robert J. Palmer, Jr., and Paul E. Kolenbrander Human oral ca\ity as a model for the study of ge- nome-genome interactions 200 Amaral Zettler, Linda A., Mark A. Messerli, Abby D. Laatsch, Peter J. S. Smith, and Mitchell L. Sogin From genes to genomes: beyond biodiversity in Spain's Rio Tinto 205 Cast, Rebecca J., David J. Beaudoin, and David A. Caron Isolation of symbiotically expressed genes from the dinoflagellate symbiont of the solitary radiolarian Thalassicolla nucleata 210 Bonfante, P. Plants, mycorrhizal fungi and endobacteria: a dialog among cells and genomes 215 Wernegreen, Jennifer J., Patrick H. Degnan, Adam B. Lazarus, Carmen Palacios, and Seth R. Bordenstein Genome evolution in an insect cell: distinct features of an ant-bacterial partnership 221 LIST OF PARTICPANTS . . 232 157 Reference: Bio/. Bull. 204: 159. (April 2003) 2003 Marine Biological Laboratory Introduction MITCHELL SOGIN 1 AND DIANA E. JENNINGS- l The Josephine Ba\ Paul Center for Comparative Molecular Biology and Evolution and 'The Center for Advanced Studies in the Space Life Sciences, Marine Biological Laboratory; Woods Hole, Massachusetts For more than 3.5 billion years, microbes of untold diversity have dominated every comer of our biosphere. For example, the cyanobacteria Synechococcus and Prochlorococcus, with a global biomass of approximately 1 billion metric tons, are responsible for 10% to 50% of the ocean's primary productiv- ity. Microorganisms are also responsible for key processes in geochemical cycling, biodegradation, and the protection of entire ecosystems from environmental insult. Thus, they con- trol global utilization of nitrogen through nitrogen fixation, nitrification, and nitrate reduction; and they drive the bulk of carbon, sulfur, iron, and manganese biogeochemical cycles. At higher trophic levels, rarely studied bacterial mutualists pro- vide essential nutrients and other compounds to diverse plant and animal hosts, and thus have a pervasive impact on the distribution, productivity, and diversification of multicellular organisms. Therefore, although Earth's early history teaches us that microbial life can thrive in diverse environments devoid of multicellular organisms, the continued survival of later evolv- ing multicellular plants and animals is completely dependent upon interactions with microorganisms. Although microorganisms have a central role in shaping planetary environments, little is known about how they function in consortia, i.e., extraordinarily diverse, complex, and highly organized communities. And even less is known about the responses of these microbial communities to cy- clic and transient environmental shifts. Clearly, the activi- ties of the varied microbes constituting a consortium must be coordinated, and these structured populations must also be able to detect and respond to their ever-changing envi- ronments; but we lack comprehensive descriptions of the biochemical and genetic mechanisms underlying these obligatory relationships. The control of microbial growth and biogeochemical activity in all ecosystems must include the coordinated expression of multiple genomes from dif- ferent organisms. Moreover, such genome-genome interac- tions should scale from thousands of species of microbes that function within consortia to maintain or shape the environment, to relatively simple binary interactions be- tween species (i.e., a symbiont or a pathogen and its host). In recent years, advances in molecular biology have revo- lutionized the life sciences. In the near future, microbial ecologists and evolutionary biologists will decipher the molecular language that coordinates the expression and evolution of microbial genomes and entire ecosystems. This workshop entitled Outcomes of Genome-Genome Interactions derives from previous discussions between Mitchell Sogin and John Hobbie of the Marine Biological Laboratory (MBL), Andreas Teske of the Woods Hole Oceanographic Institution, and David Stahl of the Llniver- sity of Washington. The discussions were about how to link biogeochemical measurements with metabolic processes and microbial population structures existing in natural set- tings. This meeting is meant to start a continuing conver- sation among a variety of investigators microbiologists, biogeochemists, ecosystem experts, molecular phylogeneti- cists, and molecular ecologists who are united by their interest in complex microbial processes on Earth. The ulti- mate objective is to foster novel interdisciplinary studies in ecosystems biology and evolution that are relevant to the interests of the National Aeronautics and Space Adminis- tration (NASA) in fundamental space biology and the ex- ploration of life beyond planet earth. This workshop was sponsored by the Center for Ad- vanced Studies in the Space Life Sciences (CASSLS) and the Josephine Bay Paul Center for Comparative Molecular Biology and Evolution both at the MBL. CASSLS was established in 1995 through a cooperative agreement be- tween the MBL and the Life Sciences Division of NASA. The Center acts as an interface between NASA and the basic science community, promoting interactions and dis- cussion in areas of mutual interest. The Josephine Bay Paul Center sponsors research that integrates the powerful tools of genome science, molecular phylogenetics, and molecular ecology. The aims are to advance our understanding of the relationships among living organisms, to quantify and as- sess biodiversity, and to identify molecular mechanisms of biomedical importance. 159 Reference: Biol. Bull. 204: 160-167. (April 2003) 2003 Manne Biological Laboratory Biogeochemistry of Hypersaline Microbial Mats Illustrates the Dynamics of Modern Microbial Ecosystems and the Early Evolution of the Biosphere DAVID J. DBS MARAIS Ames Research Center, Moffett Field California 94035 Photosynthetic microbial mats are remarkably complete self-sustaining ecosystems at the millimeter scale, yet they have substantially affected environmental processes on u planetary scale. These mats may be direct descendents of the most ancient biological communities in which even oxygenic photosynthesis might have developed. Photosyn- thetic mats are excellent natural laboratories to help us to learn ln>w microbial populations associate to control dy- namic biogeochemical gradients. Light sustains both oxygenic and anoxygenic photosyn- thesis; in turn, photosynthesis provides energy, organic sub- strates, and oxygen to the community (Fig. 1). Although photosynthetic bacteria might dominate the biomass and productivity of the mat, many aspects of the emergent properties of this ecosystem ultimately reflect the activities of the associated nonphotosynthetic microbes, including the anaerobic populations. These nonphotosynthetic processes constitute the ultimate biological filter on chemical biomar- kers (e.g., porphyrins, hopanes, isoprenoids, and other bio- genie hydrocarbons), and also on isotopic and geologic biosignatures that subsequently enter the fossil record. Also, the transformation of photosynthetic productivity by the microbial community can contribute diagnostic "biosigna- ture" gases that might represent examples of search targets for remote detection of astronomical life (e.g., Des Marais et a!.. 2002a). To understand the overall structure and function of mat communities, it is thus critical to determine the E-mail: David.J.DesMarais@nasa.gov The paper was originally presented at a workshop titled Outcome* of Genome-Genuine Interactions. The workshop, which was held at the J. Erik .Innssoii Center of the National Academy of Sciences, Woods Hole. Massachusetts, from 1-3 May 20(12. was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266 nature and extent of interaction between photosynthetic and nonphotosynthetic, including anaerobic, microbiota. Both the diversity of biota and the functional complexity within the mats, coupled with the highly proximal and ordered spatial arrangement of microorganisms, offer the potential for a staggering number of interactions. The prod- ucts of each group can affect the responses of other groups in both positive and negative ways. For example, cyanobac- teria generate organic matter (a potential growth and energy substrate for other organisms) but also oxygen (a toxin for many anaerobic processes). Anaerobic activity recycles nu- trients to the phototrophic community, but it also generates potentially toxic sulfide (Van Gemerden. 1993). Accord- ingly, microorganisms have developed strategies to cope with the daily oscillation between extremes of eutrophy and toxicity. A study of subtidal cyanobacterial mats growing in the hypersaline seawater evaporation ponds of the salt producer Exportadora de Sal, S.A. (ESSA) is ongoing at Guerrero Negro, in Baja California, Mexico. This study is furnishing examples of the steep environmental gradients experienced by mat microorganisms, and providing preliminary evi- dence for intimate interactions between these populations. These observations indicate that future studies of genome- genome interactions will contribute substantially to our understanding of the origins, environmental impacts, and biosignatures of photosynthetic microbial mats. The Microenvironment Within Photosynthetic Microbial Mats To understand the function of a microbial mat commu- nity, the physical and chemical microenvironment in which the microbes live must be known well and in detail. At Guerrero Negro, the environment within the mat differs substantially from that in the overlying water column. The 160 BIOGEOCHEMISTRY OF HYPERSALINE MICROBIAL MATS 161 Day Night CO 2 O S-Gases Water column f chemohthotrophic S bacteria Mineral Phases Representative Microsensor Measurements Organic and Mineral Biomarkers Figure 1. Schematic of a cyanobacterial microbial mat with associated depth-related light and chemical gradients. Flow diagram at center is modeled after Fenchel and Finlay ( 1995). Boxes denote functional groups of microorganisms, and arrows denote flows of chemical species into or out of microorganisms. S mlermed indicates sulfur in intermediate oxidation states. Schematic at left depicts vertical gradients of O, and sultide during the day and at night. Oxygen concentrations are shown decreasing to zero at a depth of 2 mm during the day, and just below the mat surface at night. The vertical bars at upper left represent the relative depths of penetration of blue (B). green (G), yellow (Y), orange (O), and red (R) light. Such chemical gradients and light penetration profiles of both filamentous and unicellular mats are qualitatively similar, although the depth scale (mm) of such profiles tends to be greater for unicellular mats. community just beneath the mat surface typically experi- ences steep vertical gradients of light intensity and redox conditions that change markedly during the diel cycle. The intensity and spectral composition of the light that penetrates the mat is changed by absorption and scattering. Motile photosynthetic organisms optimize their position with respect to the resultant light gradient; some biota harvest light in the infrared spectral range (J0rgensen el al, 1987; J0rgensen and Des Marais, 1988). When oxygenic photosynthesis ceases at night, the upper layers of the mat become highly reduced and sulfidic (J0rgensen et ai, 1979; J0rgensen, 1994). Counteracting gradients of oxygen and sulfide shape the chemical environment and provide daily- contrasting microenvironments that are separated on a scale of a few millimeters (Fig. 1; Revsbech et ai, 1983). Radi- ation hazards (UV, etc., Garcia-Pichel, 2000) as well as oxygen and sulfide toxicity elicit motility and other physi- ological responses. This combination of benefits and haz- ards of light, oxygen, and sulfide promotes the allocation of the various essential mat processes to the periods of light and dark periods (Friind and Cohen, 1992; Bebout et al., 1994) and to various depths in the mat. Light microenvironment The light flux penetrating the mat can be measured both as downward irradiance (the total down-welling light that passes through a horizontal plane) and as scalar irradiance (the sum of all light that converges upon a given point within the mat). Due to the high density of photosynthetic 162 D. J. DES MARAIS organisms, bacterial mucilage, and mineral particles in mats, light absorption is dominated by the light-harvesting pigments of the phototrophic bacteria, and light is strongly scattered. Because absorption and scattering of light are quite substantial within the mat, scalar irradiance can dif- fer substantially from downward irradiance (J0rgensen and Des Marais. 1988). Because scalar irradiance measures the total light actually available at a given location, it consti- tutes the most meaningful description of the environment of a microorganism. Measurements of scalar irradiance were obtained both for a microbial mat that was dominated by a filamentous cya- nobacterium, Microcoleus chthonoplastes, and for a mat that grew at higher salinity and was dominated by unicel- lular cyanobacteria (Jorgensen and Des Marais, 1988). A strong decline in intensity and a marked change in spectral composition of the light are both typically observed with depth in the dark olive mat, dominated by Microcoleus cyanobacteria. Minima in the spectra correspond to the absorption maxima of the photosynthetic pigments of cya- nobacteria. Chlorophyll a (Chi n) absorbs at wavelengths of about 430 and 670 nm, phycocyanin at about 620 nm, and various carotenoids in the range of 450 to 500 nm. In contrast, the mat that was dominated by unicellular cya- nobacteria had a lower density of cells, a more gelatinous texture, and a light orange-tan color. Light penetrated more deeply into the unicellular cyanobacteria] mat. although blue light was strongly attenuated. Carotenoids achieved most of the light absorption in this mat. In both the Microcoleus and unicellular mats, longer-wavelength light, particularly longer than 900 nm. penetrated farthest into the mat (Fig. 1 ). Such studies illustrate how the mat matrix affects the penetration of light and the physiology of the biota. For example, mat cyanobacteria that use light that has been filtered by overlying diatoms exhibit greatest photosynthetic activity at wavelengths between 550 and 650 nm (Jorgensen et al., 1987), a region that lies between the absorption maxima of Chi a. In contrast, planktonic cyanobacteria exposed to a broader spectrum of light in their natural environment show significant activity at wavelengths cor- responding to the absorption maxima of Chi a (Jorgensen and Des Marais, 1988). Chemical gradients The high rates of oxygenic photosynthesis that occur in the narrow photic zone of the mat create steep and variable gradients (Revsbech et al., 1983) in pH and in concentra- tions of dissolved inorganic carbon (DIC) and O 2 (DO). The oxic zone reflects a dynamic balance between photosyn- thetic O 2 production and O 2 consumption by a host of sulfide-oxidizing and heterotrophic bacteria. Using microelectrodes, the depth distribution of |DO], sulfide concentrations, and pH was determined in a mat dom- inated by M. chthonoplastes (Jorgensen and Des Marais. 1986). These depth profiles are typical for these mats (e.g., Canfield and Des Marais. 1993). Extremely high rates of oxygenic photosynthesis create DO levels that are nearly five times the value of air-saturated brine, yet this O 2 has a residence time of only 2 min. Oxygen production can be- come negligible at a depth of only 0.5 mm, due to light limitation (Fig. 2a). However, O 2 diffuses farther down to a point at which it overlaps with sulfide diffusing up from below. This interval is typically inhabited by abundant green nonsulfur phototrophic bacteria (e.g., Chloroflexus) and by Beggiatoa (Fig. 3b). As sunset approaches, the oxic zone collapses quickly, and the oxic-anoxic boundary ap- proaches the mat surface (Canfield and Des Marais, 1993). Accordingly, conditions alternate between O 2 supersatura- tion and millimolar concentrations of sulfide. Remarkably, diverse microbiota have apparently become well adapted to these conditions. The relative abundances of photoautotrophic bacteria (e.g., purple sulfur bacteria and green nonsulfur bacteria) and chemolithotrophic sulfide-oxidizing bacteria can be af- fected by the amount of light that reaches the chemocline (J0rgensen and Des Marais. 1986). Light levels as low as 1% of the incident near-infrared radiation (800 to 900 nm) are sufficient for Chromatium, a phototrophic purple sulfur bacterium, to dominate the chemocline (Fig. 2b). Thus the balance between the penetration of O ; and the penetration of light into the sulfide-rich zone determines the balance in the populations of sulfur bacteria. Biogeochemical Cycling of Key Elements and Their Compounds The waters that host well-developed microbial mats are typically depleted in the basic nutrient elements (Javor, 1983), yet microbial mats are highly productive aquatic ecosystems. This remarkable productivity reflects the effi- cient recycling of key nutrients within the mat ecosystem. The cycling of carbon, oxygen, sulfur, and nutrients has been studied in mats dominated by M. chthonoplastes (e.g., Canfield and Des Marais. 1993: Bebout et al., 1994). The potential for substantial coupling among populations arises through cyanobacterial production of hydrogen and small organic acids (Fig. 1; Stal et al.. 1989; Van Der Oost et al., 1989: Stal, 1991). Such substances can serve as substrates for energy and growth for a broad array of mi- croorganisms. Bacterial production of low-molecular- weight nitrogen and sulfur compounds is also important. These compounds lie at the center of energy and electron flow in anaerobic ecosystems and thus are potential basis for microbial interactions. For example, interspecies transfer of hydrogen facilitates many well-studied anaerobic consortia (e.g.. Ferry and Wolfe, 1976; Mclnerney et al., 1979). Hydrogen can represent not only an agent of electron trans- fer but also an important thermodynamie control with the potential for significantly altering the metabolic function of BIOGEOCHEM1STRY OF HYPERSALINE MICROBIAL MATS 163 '''">- "~\ > ' : "'" ' "&?- .--A.- V ^:/44f\ -0.5 (b) |O.M 2 min- 1 ^M O 2 min- 1 400 800 1200 100 200 300 400 2004006008001000 100 200 300 400 -0.5 Depth, mm 0.0 0.5 1.0 \ Oxygen-sulfide interface 0.0 0.5 1.0 1.5 2.0 Oxygen-sulfide interface Sulfide-rich zone 100 200 300 400 500 H.M H 2 S 7 7.5 8 8.5 9 9.510 P H Figure 2. Depth gradients in O 2 oxygenic photosynthesis, sulfide, pH, and microbiota. (a) At left is a schematic vertical section of the topmost 2 mm of subtidal mat dominated by Microcoleus chthonoplastes cyanobacteria. Letters along the right margin indicate the following: A: diatoms; B: Spirulina spp. cyanobacteria; C: Oscillaloria spp. cyanobacteria; D; Microcoleus chthonoplastes cyanobacteria; E; nonphotosynthetic bacteria; F: unicellular cyanobacteria; G: fragments of bacterial mucilage; H: green nonsulfur bacteria; I: Beggiatoa spp.; J: metazoans (e.g., nematodes); K: abandoned cyanobacterial sheaths. Also shown are depth profiles for key chemical constituents in the Pond 5 mat, as follows: photosynthetic O 2 production rates (horizontal barsl, and concentrations of O : and sulfide, and pH (data from J0rgensen and Des Marais, 1986). each partner. Virtually every member of the anaerobic com- munity is subject to such effects (Schink, 1988; Zinder, 1993); therefore, the participation of cyanobacteria in the cycling of hydrogen and organic acids could substantially affect biogeochemical function and community composition (Hoehler et al.. 2001). Similarly, anaerobes consume these thermodynamically sensitive end products and thus can provide an important feedback on fermentation and nitrogen fixation by cyanobacteria at the levels of both enzyme and gene regulation. Carbon, oxygen, and sulfur budgets Several general observations can be made about the cycling of carbon, oxygen, and sulfur (Canfield and Des Marais, 1993; Des Marais et al., 2002b). During the day, most of the O-> produced is recycled within the mat by O 2 respiration and some sulfide oxidation. At night, CK is consumed principally by sulfide oxidation near the mat- water interface. Microbial sulfate reduction is the principal source of DIC at night. Although abundant Chloroflexus- type (anoxic phototroph) filaments are visible microscopi- cally at the O 2 -sulfide interface, anoxygenic photosynthesis accounts for less than 10% of the total carbon fixation rate. A careful comparison of the relative O 2 and DIC fluxes across the mat-water interface reveals that, during the day, more DIC diffuses into the mat than O 2 diffuses out (Can- field and Des Marais, 1993; Des Marais et al.. 2002b). At night, more DIC diffuses out of the mat than CK diffuses in. However, both the net CK and the net DIC fluxes are balanced over the full 24-h cycle. This budget indicates that, during the day, carbon having an oxidation state greater than zero is incorporated into the mat, and carbon having a similarly high oxidation state is liberated at night. The chemical nature of this carbon is unknown. Although all of the key processes are strongly influenced by temperature, their rates scale with temperature by 164 D. J. DBS MARAIS roughly the same amount (Canfield and Des Marais, 1993: Des Marais et ai, 2003). Over a 24-h period, the overall impact of these very high metabolic rates is that the net accumulation of carbon is low. Apparently this mat is a closely coupled system in which high rates of photosyn- thetic carbon fixation fuel high rates of carbon oxidation. This efficient oxidation of organic components regenerates nutrients that, in turn, maintain high rates of primary pro- duction. Gas production The high productivity associated with photosynthetic mi- crobial mats, coupled with their proximity to the atmo- sphere and prominent role in ancient coastal environments, indicates that such mats probably influenced the early at- mosphere substantially. Cyanobacteria and diatoms provide large quantities of photosynthate to anaerobes in the mat. In the Microcoleus mats at night, O 2 is consumed by sulfide oxidation at the mat surface and lowermost water column (Canfield and Des Marais, 1993), thus the entire mat be- comes anoxic (Fig. 1). Accordingly, mat cyanobacteria must ferment to obtain energy at night, and they probably produce an array of reduced low-molecular-weight com- pounds. At Guerrero Negro, Hoehler et al. (2001) observed that subtidal Microcoleus mats generated CO, CH 4 , and significant quantities of H : . Rates of emission of CO cor- related with rates of photosynthesis, implicating cyanobac- teria, diatoms, or both as sources. Emission rates of H 2 were greatest at night, consistent with fermentation under anoxic conditions. Methane emission rates were unchanged during the diel cycle, indicating a source beneath the zone in the mat that becomes oxygenated during the day. Abundant organic photosynthates apparently interact also with sulfides in mat pore waters to produce dimethyl sulfide and other organosulfur gases, some of which escape the mat ( Visscher et al., 2003). These fluxes of reduced gases are significant for at least three reasons. First, microorganisms that inhabit cyanobac- terial mats benefit from abundant products of photosynthe- sis. Therefore, the advent of oxygenic photosynthesis bil- lions of years ago perhaps triggered a profound evolutionary transformation and diversification within the anaerobic mi- crobial world. Second, the proximity of photosynthetic mats to the atmosphere allows a substantial fraction of reduced gases to escape biological recycling and to enter and pro- foundly alter atmospheric composition (Hoehler et al, 2001 ). Early in Earth history, atmospheric reduced biogenic gases such as methane and organosulfides might have been important both as greenhouse gases and as substrates for energy and growth of other biota, including those that were geographically distant from the sources of these gases. Third, if analogous microbial ecosystems indeed exist on habitable planets orbiting other stars, they should influence the compositions of their atmospheres. The closest of these planets might soon be observable by astronomers (e.g., Des Marais et al., 2002a). Specific Microbe-Microbe Interactions The mat ecosystem depends upon intimate interactions between key groups of bacteria. Oxygenic photosynthesis by cyanobacteria and diatoms maintains a "food chain"- that is, a flow of both photosynthetic products and their derivatives that nourishes a vast consortium of mat micro- organisms (e.g., Fig. 3). The specific details of this flow of reduced species are so important that they might account for several key unanticipated observations; two such examples are described below. Figure 3. Transmission electron micrographs of biota in the subtidal Microciileux mat (Elisa D'Antoni D'Amelio, Ames Research Center, un- puhl.). |3A) Community at about 0.2 mm depth, showing several Micro- fu/ri/.v (M) trichomes situtated within a common exopolymer sheath (S). Nearby are Phormidium cyanobactena (P) and anoxygenic phototropic bacteria (PB). possibly green nonsult'ur bacteria. Scale bar at lower left equals I jum. (3B) Community at about 1.4 mm depth, just below the O, sulfide chemocline. Large Beggiaroa filaments (B) are accompanied by photosynthetic green nonsulfur bacteria exhibiting their characteristic in- tracellular photosynthetic bodies, chlorosomes (arrows). Scale bar at lower left equals 1 fj.m. BIOGEOCHEMISTRY OF HYPERSALINE MICROBIAL MATS 165 Association between cyanobacteria and anoxygenic phototrophic bacteria An unidentified filamentous phototrophic bacterium has been described which actually lives inside the sheaths of viable Microcoleus cyanobacteria (D'Amelio et al., 1987). Its occurrence inside cyanobacterial sheaths is interesting because O-, inhibits anoxygenic photosynthesis. D'Amelio ct at. (1987) proposed that, because light levels and DO vary during the day, this bacterium alternates between pho- toheterotrophic growth (at high light levels; using organic matter excreted by the cyanobacteria) and sulfide co-metab- olism with cyanobacteria (at relatively low light levels, where O, production is minimal). This anoxygenic pho- totroph might even assist cyanobacteria by consuming sul- fide after sunrise and thus relieving sulfide inhibition of oxygenic photosynthesis. Aerobic sulfate reduction Sulfate-reducing bacteria are quantitatively important consumers of dissolved organic matter. Furthermore, the sulfide they produce sustains a wide variety of phototrophic and chemotrophic bacteria. The highest rates of sulfate reduction occur in the shallowest part of the subtidal Mi- crocoleus mat, close to the photosynthetic source of fresh organic matter (Canfield and Des Marais, 1993). Although O 2 is typically an effective inhibitor of bacterial sulfate reduction, the highest reduction rates actually occur within the mat's aerobic zone during the daytime (Canfield and Des Marais, 1991 ). A thorough search was made for anaer- obic microenvironments within the aerobic zone that might serve as refugia for the sulfate-reducing bacteria, yet none were found. The specific factors that attenuate this O 2 inhibition of sulfate reduction are not known. However, fermentation products are probably abundant in the vicinity of the cyanobacteria, and their roles as chemical reductants might offset the toxic effects of oxidants such as O 2 . Future Research The studies described above have been performed prin- cipally on subtidal marine hypersaline cyanobacterial mats. Although intertidal and supratidal cyanobacterial mats also have received some attention, the level of effort summarized in this review must continue to be applied to other mat types. Examples of such mats include those dominated by unicellular cyanobacteria, eukaryotic algae, and nonphoto- synthetic bacteria such as sulfide-oxidizing bacteria in deep- sea communities (e.g., Ward et al., 1992; Stal and Cau- mette, 1994). Mats growing in low-sulfate environments such as lakes, streams, and thermal springs also merit more attention. Studies of biogeochemical cycling in mats should be broadened to include additional populations of mat micro- organisms (e.g., heterotrophs, methanogens, and novel bac- teria) that probably contribute substantially to the commu- nity. We must better understand how key nutrients such as nitrogen and phosphorus are regenerated and retained by the various types of mats. Mats that coexist with active mineral precipitation (e.g., with calcium carbonate: Fouke et al.. 1999; Reid et al., 2000) merit study to help us understand the roles of the microbes in the precipitation of minerals and the impact of mineral formation upon mat biogeochemistry. Perhaps most promising is the use of gene sequences and gene expression studies to understand the ecology of mi- crobial mat communities. Methods for the identification and interpretation of 16s RNA and other macromolecules are improving rapidly and hold great promise. These phylo- genic investigations should be combined with studies of biogeochemistry and gene expression to elucidate the key linkages between microbial populations, processes, and the emergent products and environmental impacts of microbial mats. Acknowledgments The preparation of this manuscript was supported by a grant from NASA's Astrobiology Institute. I thank Expor- tadora de Sal, S. A., for continuing field support at Guerrero Negro, Baja California Sur, Mexico. Literature Cited Bebout, B. M., H. W. Paerl, J. E. Bauer, D. E. Canfield, and D. J. Des Marais. 1994. Nitrogen cycling in microbial mat communities: the quantitative importance of N-fixation and other sources of N for primary productivity. Pp. 265-272 in Microbial Mars. L.J. Stal and P. Caumette, eds. NATO ASI Series, Vol. G 35. Springer- Verlag, Berlin. Canfield, D. E., and D. J. Des Marais. 1991. Aerobic sulfate reduction in microbial mats. Science 251: 1471-1473. Canfield, D. E., and D. J. Des Marais. 1993. Biogeochemical cycles of carbon, sulfur, and free oxygen in a microbial mat. Geochim. Cosmo- chim. Acta 57: 3971-39X4. D'Amelio, E. D., Y. Cohen, and D. J. Des Marais. 1987. Association of a new type of gliding, filamentous, purple phototrophic bacterium inside bundles of Microcoleus chthonoplastes in hypersaline cyanobac- terial mats. Arch. Microhiol. 147: 213-220. Des Marais, D. J., M. Harwit, K. Jucks, J. F. Kasting, J. I. Lunine, D. Lin, S. Seager, J. Schneider, W. Traub, and N. Woolf. 2002a. Remote sensing of planetary properties and biosignatures on extrasolar terrestrial planets. Astrt>hii>lo/>y 2: 153-181. Des Marais, D. J., D. Albert, B. Bebout, M. Discipulo, T. Hoehler, and K. Turk. 2002b. Biogeochemical contrasts between subtidal and intertidal-to-supratidal hypersaline microbial mats (abstract). Astrobi- ology Science Conference 2002: 188. NASA Ames Research Center, Moffett Field; CA. Des Marais, D. J., B. M. Bebout, S. Carpenter, M. Discipulo, and K. Turk. 2003. Carbon and oxygen budgets of hypersaline cyanobacte- rial mats: effects of diel cycle and temperature (abstract). American Society of Limnology and Oceanography, 2003 Annual Meeting, Salt Lake City. Fenchel, T., and B. J. Finlay. 1995. Ecology and Evolution in Anoxic Worlds. Oxford University Press, Oxford. Ferry J. G., and R. S. Wolfe. 1976. Anaerobic degradation of benzoate to methane by a microbial consortium. Arch. Microhiol. 107: 33-40. 166 D. J. DES MARAIS Fouke, B. W., J. D. Farmer, D. J. Des Marais, L. Pratt, N. C. Sturchio, P. C. Burns, and M. K. Discipulo. 1999. Depositional facies and aqueous-solid geochemistry of travertine-depositing hot springs (Angel Terrace, Mammoth Hot Springs. Yellowstone National Park, USA). J. Sediment. Res. 70: 565-585. Friind, C.. and V. Cohen. 1992. Diurnal cycles of sulfate reduction under oxic conditions in cyanobacterial mats. Appl. Environ. Micro- bio/. 58: 70-77. Garcia-Pichel, F. 2000. Cyanobactena. Pp. 907-929 in Encyclopedia of Microbiology. 2nd ed., Vol. 1. J. Lederberg. ed. Academic Press, San Diego. CA. Hoehler, T. M., B. M. Bebout, and D. J. Des Marais. 2001. The role of microbial mats in the production of reduced gases on the early Earth. Nature 412: 324-327. Javor, B. 1983. Nutrients and ecology of the Western Salt and Expor- tadora de Sal Saltern brines. Pp. 195-205 in Sixth Internationa/ Sym- posium on Salt, B. C. Schreiber and H. L. Hairier, eds. Salt Institute, Alexandria, VA. Jargensen, B. B. 1994. Sulfate reduction and thiosulfate transformations in a cyanobacterial mat during a diel oxygen cycle. FEMS Microbiol. Ecol. 13: 303-312. Jergensen, B. B., and D. J. Des Marais. 1986. Competition for sulfide among colorless and purple sulfur bacteria in cyanobacterial mats. FEMS Microbiol. Ecol. 38: 179-186. Jergensen, B. B., and D. J. Des Marais. 1988. Optical properties of benthic photosynthetic communities: fiber optic studies of cyanobac- terial mats. Limnol. Oceanogr. 33: 99-1 13. J0rgensen, B. B., N. P. Revsbech, T. H. Blackburn, and Y. Cohen. 1979. Diurnal cycle of oxygen and sulfide microgradients and micro- bial photosynthesis in a cyanobacterial mat. Appl. Environ. Microbiol. 38: 46-58. Jorgensen, B. B., Y. Cohen, and D. J. Des Marais. 1987. Photosyn- thetic action spectra and adaptation to spectral light distribution in a benthic cyanobacterial mat. Appl. Environ. Microbiol. 53: 879-886. Mclnerney, M. J., M. P. Bryant, and N. Pfennig. 1979. Anaerobic bacterium that degrades fatty acids in syntrophic association with methanogens. Arch. Microbiol. 122: 129-135. Reid, R. P., P. T. Visscher. A. W. Decho, J. Stolz, B. M. Bebout, I. G. Maclntyre, H. W. Paerl, J. L. Pinckney, L. Prufert-Bebout, T. F. Steppe, and D. J. Des Marais. 2000. The role of microbes in accretion, lamination and early lithification of modern marine stroma- tolites. Nature 406: 989-992. Revsbech, N. P., B. B. Jergensen, and T. H. Blackburn. 1983. Micro- electrode studies of the photosynthesis and O 2 , H,S, and pH profiles of a microbial mat. Limnol. Oceanogr. 28: 1062-1074. Schink B. 1988. Principles and limits of anaerobic degradation. Pp. 771-846 in Biology of Anaerobic Microorganisms, A. J. B. Zehnder. ed. John Wiley, New York. Stal, L. J. 1991. The metabolic versatility of the mat-building cyanobac- teria Microco/eus chthonoplastes and Oscillatoria limosa and its eco- logical significance. Algol. Stud. 64: 453-467. Stal, L. J., and P. Caumette. 1994. Microbial Mats: Structure, Devel- opment and Environmental Significance. Series G: Ecological Sci- ences, Springer- Verlag, Berlin. Stal, L. J., H. Heyer, S. Bekker, M. Villbrandt. and W. E. Krumbein. 1989. Aerobic-anaerobic metabolism in the cyanobacterium Oscil/a- toria limnosa. Pp. 255-276 in Microbial Mats. Physiological Ecology of Benthic Microbial Communities, Y. Cohen and E. Rosenberg, eds. American Society for Microbiology, Washington. DC. Van Der Oost, J., B. A. Bulhuis, S. Feitz, K. Krab, and R. Kraayenhof. 1989. Fermentation metabolism of the unicellular cyanobacterium Cnmothece PCC 7822. Arch. Microbiol. 152: 415-419. Van Gemerden, H. 1993. Microbial mats: a joint venture. Mar. Geolog\ 113: 3-25. Visscher, P. T., L. K. Baumgartner, D. H. Buckley, D. R. Rogers, M. Hogan, C. Raleigh, K. Turk, and D. J. Des Marais. 2003. Dimethyl sulfide and methanethiol as biogenic signatures in laminated microbial ecosystems. Appl. Environ. Microbiol. (In pressl. Ward, D. M., J. Bauld, R. W. Castenholz, and B. K. Pierson. 1992. Modern phototrophic microbial mats: anoxygenic, intermittently oxy- genic/anoxygenic, thermal, eukaryotic and terrestrial. Pp. 309-324 in The Proterozoic Biosphere: A Multidiscipttnary Study. J. W. Schopf and C. Klein, eds. Cambridge University Press. New York. Zinder S. H. 1993. Physiological ecology of methanogens. Pp. 128-206 in Methanogenesis: Ecology. Physiology, Biochemistry and Genetics. J. G. Ferry, ed. Chapman and Hall, New York. Discussion QUESTION: As you mentioned, Dave, cylindrical sheaths made of exopolymeric material have within them two populations of organisms that you have proposed might maintain a relationship responsive to diurnal phenomena. Do you think that the mat structure is very heterogeneous, containing clusters of organisms that would, based upon our current understanding, not be expected to co-exist? For example, do some associations between popula- tions protect organisms from harmful products? DES MARAIS: Yes, I included these examples of associations between populations of organisms to illustrate this point. I think also that this is one of the reasons why Mitch Sogin had me talk early in this meeting; that is, to show that there is, potentially, a long menu of important ecological phenomena that could be ad- dressed by genomic studies. Again, in the example I showed, sulfide is removed in the morning by the anoxygenic phototrophic bacteria, which benefits the cyanobacteria. This example also includes the cross feeding by the cyanobacteria that benefits the anoxygenic phototrophs. Evidence for cross feeding has actually been documented in Yellowstone Park by David Ward. Using isotopic labeling, he observed that photosynthate does flow from cyanobacteria to green non-sulfur bacteria. I think we have dem- onstrated that sulfide inhibits at least some types of oxygenic photosynthesis in these mats. Also, anoxygenic phototrophs can reduce sulfide levels in natural environments. Another example is that the surface-dwelling microbial populations screen the deeper ones from UV and shorter wavelengths that would be injurious to their photosynthetic apparatus. Dick Castenholz, Ferran Garcia- Pichel and others have documented this extensively. QUESTION: I am wondering whether other organisms are seques- tered in a way that enhances certain processes. You presented BIOGEOCHEMISTRY OF HYPERSALINE MICROB1AL MATS 167 measurements of very high rates of sulfate reduction. Are the high rates possible because sulfate-reducing bacteria are protected by other organisms that you are just not observing directly? DES MARAIS: I will defer to Dave Stahl to discuss current studies of bacterial sulfate reduction that occurs in the presence of molecular oxygen. But, some ten years ago, Don Canfleld and I probed extensively, with electrodes, for anoxic micro-environ- ments within the oxygenated photic zone of cyanobacterial mats. In these mat pore waters, oxygen diffuses some 50 to 100 microns in just a few seconds, and so we spaced our electrode sampling profiles about one mm apart. But we just couldn't find anoxic microenvironments. However, even though oxygen is a strong oxidant, the rate at which it oxidizes other substances is not as fast as the rates of some other oxidants, such as radicals, etc. Perhaps the deleterious oxidation reactions involving oxygen are mitigated by faster reactions carried out by reductants such as hydrogen. Relatively fast-acting reducing compounds in the mats might con- fer the advantage that the sulfate-reducing bacteria need to main- tain these very high rates in the presence of oxygen. So, the relative rates of reactions, as well as mechanisms of physical protection, might play key roles in these mat communities. QUESTION: What about migration? There is another dynamic with biota moving in response to chemical gradients. DES MARAIS: What I find interesting is that some populations migrate, whereas others don't even under circumstances where migration would seem to benefit both of these populations. For example, microorganisms can migrate vertically within cyanobac- terial mats in response to changes in the depth of oxygen and light penetration, which can vary diumally by a few mm. Beggiatoa are nonphotosynthetic bacteria that oxidize sulfide with oxygen, and so it is advantageous for them to occupy zones where both sulfide and oxygen coexist. But, there are different diameters of Beggia- toa, and I recollect that the ones having smaller diameters tend to be much more mobile than the big ones. Also, some photosynthetic populations, particularly purple sulfur bacteria, have been ob- served to migrate. How does the community select between pop- ulations that migrate as conditions change, versus the more sta- tionary populations? There are examples of two environmental cues for migration that appear be in conflict with each other. What causes Beggiatoa to migrate downward after sunrise? Is it the light or is it the oxygen and sulfide chemical gradients? It can't be just an avoidance response to high oxygen concentrations. When Beggiatoa is still at the surface in the morning, bathed in sunlight, oxygen production is starting beneath it. At some point, Beggiatoa must dive down through an oxygen rich zone to get down to the darker sulfide- oxygen interface that is most suitable during the daytime. So the question "What are all of the cues that drive these organisms to migrate?" is both an important question and one that is still in search of definitive answers. Reference: Biol. Bull. 204: 168-173. (April 2003) 2003 Marine Biological Laboratory Complexity in Natural Microbial Ecosystems: The Guerrero Negro Experience JOHN R. SPEAR, RUTH E. LEY, ALICIA B. BERGER, AND NORMAN R. PACE* Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, 347 UCB, Boulder, Colorado 80309 The goal of this project is to describe and understand the organismal composition, structure, and physiology of mi- crobial ecosystems from hypersaline environments. One collection of such ecosystems occurs at North America 's largest saltworks, the Exportadora de Sal, in Guerrero Negro, Baja California Sur. There, seawater flows through a series of evaporative basins with an increase in salinity until saturation is reached and halite crystallization begins. Several of these ponds are lined with thick (JO cm) micro- bial mats that have received some biological study. To determine the nature and extent of diversity of the microbial organisms that constitute these ecosystems, we are conduct- ing a phylogenetic analysis using molecular approaches, based on cloning and sequencing of small subunit AA + 1.9 H 2 O NO,- + 10 H + + 0.58 C 6 H 12 O 6 + 0.034 SO 4 2 - + 8.9 e~ AA + 4.9 H,O NH 4 + + 3 H : O - NO,- + 10 H + + 8 e~ '/2 N 2 + 4 FT + 0.58 C 6 H I2 O 6 + 0.034 SO 4 -~ + 3.9 e~ - AA + 1.9 H 2 O 2 H,O -> O, + 4 H + + 4 e" O 2 + 4 H + + 4 e" 2 H 2 O AA + (Protein) n O (Protein) n + 1 + H 2 O 4 AA + 6.8 C 6 H I2 O 6 + Mg- + + 46.2 H + + 48.6 e" Chi + 42.3 H,O + 0.14 SO 4 : 8 C 6 H n O 6 + P0 4 3 " + 92 FT + 89 e~ O C 48 H l)( ,O 6 P + 46 H,O C 6 H, 2 0~ 6 + (C 6 H,,,O s ) n (C h H,,,0 5 )n + , + H 2 4 AA + 2 PO 4 -'- O DNA + 0.73 C n H 12 O 6 + 0.14 SO 4 : " + 1.9 H 2 O + 1.9 H + + 7.5 Abbreviations: AA. amino acids (C, ,H 7 ,0, 7 NS,, U , 4 ""' + ); Chi, chlorophyll (C,,H 72 O 5 N 4 Mg): Protein (C, 5 H 5 |O, I7 NS,,,,, 4 ""' + ): DNA (C 96 H ]4 O X N 4 P 2 ). Reactions are either reversible = or irreversible ( ). decay of redox potential via a distributed metabolism that forms the cornerstone of our approach. We focus on micro- bial systems because these systems exhibit the greatest degree of metabolic capacity, are responsible for the major- ity of biogeochemistry on earth, display fast dynamics that allows for practical experiments, and are less susceptible to loss of diversity. Thermodynamically Constrained Metabolic Biogeochemical Model A traditional reductionist biogeochemical model would include differential equations for growth of each microbial functional group, equations for Monod-type growth kinet- ics, and numerous conditional statements to direct the use of Figure 1. Conceptualization of the metabolic biogeochemistry model. Half reactions lead to production of protein (and other building-block constituents), which is then allocated to reactions governed by an optimization function. Abiotic reactions are incorporated with standard kinetics. MODIIINIi MK'KORIAI CONSORTUMS 177 Table 2 Governing equations for the thennod\niimiL'ull\ constrained biogeochemislrt- model Constraint Equation Free energy Redox Biomass composition Reaction kinetics Enzyme allocation Photosynthesis LP Problem S,r,AG, 2 ,/-,?,' = C:N L " < " '' , < C:N Ur|x ' r r, E,f,(S) V; dE, dE, dE, S,, = E T and (// PAR(r) 1 |Chl] h AG^ [Chi] + A\ a , Maximize S^ft'V',.^) + R^r^E,} Definitions: r, are reaction rates (Table 1); e, are electron-pair transfers associated with reaction /'; , is concentration of enzyme allocated to reaction i, and E T is the total protein concentration; R c and R N are the subset of reactions (/,) that lead to C and N accumulation in living biomass, respectively;//^) are functions that describe uptake kinetics of substrate 5. such as Michaelis-Menten; PAR(r) is the photosynthetic active radiation; h is water column depth; and r p is the rate of photosynthesis (Table 1 ). Solution of the linear programming (LP) problem gives ;, and , at time t, which are used in standard C and N conservation equations to obtain solution of state variables over time. See Vallino el al. (1996) for a more thorough development. the various electron donor and acceptor resources that exist within the environment (Koelmans et al.. 2001 ). While this approach has been useful for well-defined laboratory exper- iments with a few species growing on a limited number of well-defined substrates, it is not practical for extension to more diverse microbial ecosystems with numerous or ill- defined substrates. Reductionist-based biological models fail to incorporate the governing laws that define living systems (Lawton, 1999); the models are based solely on empirical observations. Consequently, these models are brittle and often fail as the system's state changes signifi- cantly over time and space (Vallino, 2000). To develop a robust model that can predict microbially governed biogeochemistry in spatially and temporally di- verse environments, a more holistic, systems-based perspec- tive must be taken. Our governing philosophy is that living systems synthesize and allocate metabolic capability in such a way as to optimally utilize available resources in the environment as governed by NET. What we seek to deter- mine is the nature of the objective function that living systems tend to follow, and what causes living systems to diverge from this function. This optimization-based approach was first developed in a thermodynamically constrained metabolic framework to examine bacterial utilization of dissolved organic matter (Vallino et at., 1996). However, this model still uses an (a) S 3 - (C) 18 16 14 J12 I 10 8 6 4 1 2 Time(d) 1 2 Time (d) (b) 5 * S 3 ) 2 1 i 1 2 Time (d) 1 2 Time (d) Figure 2. accumulation Metabolic model. Changes in resource concentrations of (a) ammonium and (b) nitrate, and of biological structure (c) protein and (d) chlorophyll over the course of the simulation. 178 (a) 5 0.6 3. I 0.4 O 0) !0.2 N C LU (c) J. J. VALLINO NH 4 Uptake (b) NO 3 Uptake 8 6 =0 4 aj re or ! l 1.5 ^5 1 5 a re 0.5 X or NH; + 0.58C 6 H, 2 O 6 + 0.9e -> AA + H 2 O / i j " ' f i 1.2 4 i i a 3 0.8 o a) 'a re 2 c c 0.4 X X NO,+ 10H' + 0.58Glc + 9e - . AA + H 2 O / 0123 Time (d) N 2 Fixation (d) 10 0.2 8 S 3 ^ tiO.15 6 ~ g .! i" X *" 2 K 0123 Time(d) Chlorophyll Synthesis 1 IO.B 6 0.6 O 0.4 gO.2 HI .'/2N 2 + 4H* + 0.58GIC + 4e -> AA + H ; O. .1 1 i Glc + AA + 46. 2H* + 48. 6e" - Chi 0123 Time (d) 1 2 Time (d) 3 Figure 3. Reaction rates (solid line) and allocation of protein (enzyme concentration, dashed line) for reactions involving (a) ammonium uptake, (b) nitrate uptake, (c) N, fixation, and (d) chlorophyll synthesis during the course of the simulation. Abbreviations: Glc, glucose; AA, amino acids, Chi, chlorophyll. organismal approach, in that the model tracks bacterial biomass. We increase the applicability of our modeling approach by removing emphasis on synthesizing bacterial biomass and placing it instead on synthesis of metabolic capability exhibited by the whole ecosystem. The model consists of a set of metabolic half-reactions that represents the major metabolic capability of a planktonic ecosystem (Table 1 ). But instead of synthesizing bacteria, reactions produce protein, chlorophyll, storage compounds, and other fundamental building materials observed in living systems (Fig. 1 ). These building materials represent those summed over all organisms in the ecosystem, not any one particular organism. Indeed, organisms are not directly modeled. Newly synthesized protein is then allocated to those meta- bolic reactions that optimize the specified objective criteria, while enzymes no longer in use can be degraded back into constituent amino acids (Fig. 1 ). A linear programming (LP) problem is used to determine the reaction rates (r,) and enzyme concentrations (",) that maximize a given objective function, subject to fundamental constraints, such as energy, redox, composition, kinetics, and light-capturing capabili- ties (Table 2). Although the model does not distinguish species in a classic sense, it does from a functional perspec- tive. As environmental conditions change, so do allocations of resources to metabolic reactions. Real systems accom- plish this same objective via relative changes in species abundances and magnitude of gene expression. As an example of the model, we simulate a marine phytoplankton bloom, where metabolic reactions associated with ammonium (NH 4 + ) and nitrate (NO ? ~) uptake, N r fixation, carbon dioxide (CO->) fixation, and biomass syn- thesis (protein and chlorophyll) are included in the model (Table 1). The optimization goal chosen was maximizing the rate of biomass synthesis, though others could be for- mulated. Resources made available were 5 ju,A/NH 4 + , 5 fj,M NO 3 ", atmospheric NT, and light. The model simulation proceeds by preferentially consuming NH 4 + over NO," (Fig. 2a,b), which is evident by the allocation of protein (in the form of enzyme) to NH 4 + uptake (Fig. 3a). but not to NO 3 ~ uptake nor N 2 fixation (Fig. 3b,c). There is also a strong initial allocation of protein to chlorophyll synthesis (Figs. 2d, 3d), but this protein is rapidly reallocated after 0.5 d due to diminished returns on the investment in light harvesting capacity (i.e., chlorophyll), which saturates at high chlorophyll concentration (Fig. 3d). As NH 4 + becomes exhausted (Fig. 2a). protein is reallocated from NH 4 * to NO_,~ uptake (Fig. 3a. b). Subsequently, as NO, be- comes depleted (Fig 2b), protein is allocated to N : fixation (Fig. 30. MODHUNCi MICROBIAL CONSORTIUMS 179 Conclusions If nonequilihrium thermodynamics governs biogeochem- istry, our metabolic modeling approach represents a more direct means of capturing ecosystem dynamics than classic, organismal-based approaches. The approach also predicts how whole-system genomic transcription and translation should proceed, which can be compared to actual systems using techniques currently being advanced in molecular biology. Because the metabolic ecosystem model is based on fundamental governing equations, it should prove more robust and have a greater operating range than organismal- based models. Acknowledgments This research has been supported by NSF grants OCE- 9726921 and DEB-98 15598 and the Mellon Foundation. Literature Cited Allen, P. M. 1985. Ecology, thermodynamics, and sell'-organi/ation: towards a new understanding of complexity. Can. Bull, Fish. Aauat. Sci. 213: 3-26. Amend, J. P.. and K. L. Shock. 2001. Energetics of overall metabolic reactions of thermophilic and hyperthermophilic Archaea and Bacteria. FEMS Microhiol. Rev. 25: 175-243. Boetius. A., K. Ravenschlag, C. J. Schuber, D. Rickert, F. Widdel, A. Gieseke, R. Amann, K. K. Jorgensen, U. Witte, and O. Pfannkuchc. 2000. A marine microhial consortium apparently mediating anaero- bic oxidation of methane. Nature 407: 623-626. Caldwell, D. E., G. M. Wolfaardt, D. R. Korber, and J. R. Lawrence. 1997. Do bacterial communities transcend Darwinism? Adv. Microh. Ecol. 15: 105-141. Choi, J. S., A. Mazumder. and R. I. C. Hansell. 1999. Measuring perturbation in a complicated, thermodynamic world. Ecol. Model. 117: 143-158. Galloway, J. N., and E. B. Cowling. 2002. Reactive nitrogen and the world: two hundred years of change. Ambio 31: 64-71. Hoehler, T. M., M. J. Alperin, D. B. Albert, and C. S. Martens. 1998. Thermodynamic control on hydrogen concentrations in anoxic sedi- ments. Geochim. Cosiiiochiin. Ada 62: 1745- 1756. Houghton, J. T., Y. Ding, IX J. Griggs, M. Noguer, P. J. van der Linden, and I). Xiaosu. 2002. Cliimite Change 21X11: The Scientific Basis: Conlrihittinn of Working Group / to the Third Assessment Report nl the Intergovernmental Panel on Climate Change (IPCCl Cambridge University Press. Cambridge. UK. Pp. 1-444 Jackson, B. E., and M. J. Mclnerney. 2002. Anaerobic microbial metabolism can proceed close to thermodynamic limits. Nature 415: 454-456. Jakobsen, R., and D. Postnia. 1999. Redox zoning, rates of sulfate reduction and interactions with Fe-reduction and methanogenesis in a shallow sandy aquifer, Romo, Denmark. Geochim. Cosmochim. Acln 63: 137-151. Johnson, L. 1988. The thermodynamic origin of ecosystems: a tale of broken symmetry. Pp. 75-105 in Entropy, Information, and I-ivohiium: New Perspectives on Physical and Biological Evolution. B. H. Weber, D. J. Depew, and J. D. Smith, eds. MIT Press, Cambridge. MA. Jorgensen, S. E. 1994. Review and comparison of goal function in system ecology. Vie Milieu 44: 1 1-20. Jorgcnsen, S. E., B. C. Patten, and M. Slraskraba. 2000. Ecosystems emerging: 4. growth. Ecol. Model. 126: 244-284. Koelmans, A. A., A. Van Der Hei.jde, L. M. Kni.jt'f. and R. H. Aalderin. 2001. Integrated modelling ol entiophication and organic contami- nant fate and effects in aquatic ecosystems. A review. Water Res. 35: 3517-3536. Lawton, J. H. 1999. Are there general laws in ecology 1 ' Oikos 84: 177-142. Madigan. M. T., J. M. Martinko, and J. Parker. 2000. Brock Biology of Microorganisms, 4th ed. Prentice Hall, Upper Saddle River. N.I. Margalef, R. 1968. I'erspcctives in Ecological Theory. University of Chicago Press, Chicago. Morowitz, H. J. 1968. Energy Flow in Biology: Biological Organ/ration Ax a Problem in Thermal Physics. Academic Press, New York. Nielsen, S. N. 1995. Optimisation of exergy in a structural dynamic model. Ecol. Model. 77: I I 1-122. Nielsen, S. N., and R. E. Ulanowicz. 2000. On the consistency between thermodynamical and network approaches to ecosystems. Ecol. Model. 132: 23-3 1 . Ocluin. H. T. 1971. Environment. Power and Society. John Wiley. New York. Odum, H. T. 1983. Systems Ecology. John Wiley. Toronto. Odum, H. T., and R. C. Pinkerton. 1955. Time's speed regulator: the optimum efficiency for maximum power output in physical and bio- logical systems. Am. Sci. 43: 321-343. Onsager, L. 1931. Reciprocal relations in irreversible processes. Phvs. Rev. 37: 405-426. Prigogine, I. 1955. Introduction to Thermodynamics of Irreversible Pro- cesses. John Wiley, New York. Prigogine, I. 1978. Time, structure, and fluctuations. Science 201: 777- 785. Prigogine, I., and G. Nicolis. 1971. Biological order, structure and instabilities. Q. Rev. Biopliyx. 4: 107- I4X. Schlesinger, W. H. 1997. Biogeochemistry: An Analysis of Global Change, 2nd ed. Academic Press. San Diego. CA. Schneider, E. D. 1988. Thermodynamics, ecological succession, and natural selection: a common thread. Pp. 107-138 in Entropy. Informa- tion, and Evolution: New Perspectives on Physical and Bii>lognat Evolution, B. H. Weber, D. J. Depew. and J. D. Smith, eds. MIT Press. Cambridge, MA. Schneider, E. D., and J. J. Kay. 1994. Complexity and thermodynam- ics: towards a new ecology. Futures 26: 626-647. Schriidinger, E. 1944. What Is Life'.' Cambridge University Press, Cam- bridge. UK. Schiilz. H. N., T. Brinkholf, T. G. H. M. M. Ferdelman, A. Teske, and B. B. Jorgensen. 1999. Dense populations of a giant sulfur bacterium in Namibian shelf sediments. Science 284: 443-495. Toussaint, O., and E. D. Schneider. 1998. The thermodynamics and evolution ot complexity in biological systems. Camp. Bioclicm. Physiol. A 120: 3-4. Ulanowicz, R. E. 1986. tlrowlh anil Development: Ecosystem.': Phenom- enology. Springer- Verlag. New York. Ulanowicz, R. E. 1997. Ecology, the Ascendent Perspective. Columbia University Press, New York. Vallino, J. J. 2000. Improving marine ecosystem models: use of data assimilation and mesocosm experiments. J. Mar. Res. 58: 1 17-164. Vallino, J. J., C. S. Hopkinson, and J. E. Hobble. 1996. Modeling bacterial utilization of dissolved organic matter: optimization replaces Mound growth kinetics. Limnol. Oceanogr. 41: 1541-1604. Wiley, E. O. 1988. Entropy and evolution. Pp. 173-188 in Entropy. Information, and Evolution: New Perspectives on Physical and Bio- logical Evolution. B. H. Weber. D. J. Depew. and J. D. Smith, eds. MIT Press. Cambridge, MA. Reference: Biol Bull. 204: 1X0-185. (April 2003 1 2003 Marine Biological Laboratory Geomicrobiology of the Ocean Crust: A Role for Chemoautotrophic Fe-Bacteria KATRINA J. EDWARDS*, WOLFGANG BACH, AND DANIEL R. ROGERS Geomicrobiology Group, Department of Marine Chemistry and Geochemistry, McLean Lab, MS#$, Woods Hole Oceanographic Institution. Woods Hole, Massachusetts 02536 The delicate balance of the major global biogeocheinical cycles greatly depends on the transformation of Earth ma- terials at or near its surface. The formation and degradation of rocks, minerals, and organic matter are pivotal for the balance, maintenance, and future of many of these cycles. Microorganisms also play a crucial role, determining the transformation rates, pathways, and end products of these processes. While most of Earth 's crust is oceanic rather than terrestrial, few studies hare been conducted on ocean crust transformations, particularly those mediated by endo- lithic (rock-hosted) microbial communities. The biology and geochemistry of deep-sea and sub-seafloor environments are generally more complicated to study than in terrestrial or near-coastal regimes. As a result, fewer, and more tar- geted, studies usually homing in on specific sites, are most common. We are studying the role of endolithic microor- ganisms in weathering seafloor crustal materials, including basaltic glass and sulfide minerals, both in the vicinity of seafloor hydrothermal vents and of/'-a.\is at un.sedimented (young) ridge flanks. We are using molecular phylogenetic surreys and laboratory culture studies to define the size, diversity, physiology, and distribution of microorganisms in the shallow ocean crust. Our data show that an unexpected diversity of microorganisms directly participate in rock weathering at the sea/loor. and imply that endolithic micro- bial communities contribute to rock, mineral, and carbon transformations. *To whom correspondence should ho addressed. E-mail: kedwardsls 1 whoi.edu The paper was originally presented at a workshop titled Outcome* of (.ifnoim'-Gcnomc Interaction*. The workshop, which was held at the J. Erik Jonsson Center ot the National Academy of Sciences, Woods Hole. Massachusetts, from 1-3 May 2002. was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-12dh Weathering reactions in Earth's near-surface environ- ments play pivotal roles in balancing chemical exchange between the lithosphere. hydrosphere, and atmosphere. Micro- organisms at and beneath the surface affect the transforma- tion rates, mechanisms, and pathways of exchange for many elements. Over 70% of the Earth's exposed crust is oceanic, and most of this material occurs well below the euphotic upper ocean regime. The oceanic crust is composed largely of basalt ( -50 wt 7c SiO 2 ; 4-15 wt 7r each of MgO, FeO, CaO, AKO,), although metal sulfide minerals (e.g.. pyrite. FeS,) are prominent at seafloor hydrothermal ridge axes. After their formation at seafloor spreading centers, ocean crust rocks undergo weathering, either by reaction with seawater while exposed at the ocean floor or from fluid-rock interaction below the seafloor. While the effects of crustal weathering and fluid-rock interaction are well documented (Alt, 1995), the role of microorganisms in these processes is poorly understood. To date, investigations of sub-seafloor endolithic (rock-hosted) microbial communities have been largely limited to exam- ination of textural features, such as channels and pits, which appear in petrographic thin sections and other microscopic preparations, and are thought to indicate microbial activity (e.g., Fisk et al., 1998). However, the physiological func- tions of members of sub-seafloor ecosystems have not been elucidated, precluding our ability to place these hypothe- sized communities into proper global biogeocheinical con- text. Here, we first briefly review the physical and chemical weathering regime of the ocean crust, and then discuss our recent findings regarding the physiological activities and phylogenetic relationships among prokaryotes that actively participate in crustal alteration at the ocean floor. The uppermost 200-500 m of basaltic ocean crust is char- acterized by high permeabilities (10 12 10 15 m 2 ) that facil- itate the circulation of large quantities of seawater (Fisher, 180 (ii OMICKORKII (Ki'i ()! 1111 OC1 AN CKIi.ST 181 1998: Fisher and Becker, 2000). Chemical reactions between seawater and seafloor rocks change the compositions of botli the oceans and the aging ocean crust (Hart and Staudigel, 1986; Mottl and Wheat, 1994; Alt et al.. 1996; Staudigel et uL 1996; Elderfield et al., 1999; de Villiers and Nelson, 1999; Wheat and Mottl, 2000). Basaltic rocks react with oxygenated deep-sea water to form secondary hydrous alteration minerals, including a variety of Fe-oxyhydroxides and micas, and clay minerals such as celadonite (K(Mg,Fe)(Fe,Al)Si 4 O m (OH) 2 ) and smectite (e.g.. montmorillinite, ( l/2Ca,Na)(Al,Mg,Fe) 4 (Si,Al) s O 20 (OH) 4 *nH 2 O)(Honnorez, 1981; Alt, 1995). These minerals replace glass and primary minerals such as olivine ((Mg, Fe) 2 SiO 4 ), sulfides (e.g., pyrite), and to lesser extents plagioclase (NaAlSi 3 O s ) and clinopyroxene ((Ca,Mg,Fe,Al)i (Si.AI ),()(,), and they till fractures and void space in the crust. The extent of oxidation associated with low-temperature alter- ation is extremely variable at different spatial scales. Age dating of celadonite suggests that the products of oxidative alteration persist in crust of significant age (10 K) million years; Peterson et al., 1986; Hart and Staudigel, 1986; Galla- han and Duncan, 1994): these products can be detected and analyzed and the data used to inteipret past fluid-rock interac- tions in ocean crust. The chemical reactions that occur from fluid-rock inter- action in the sub-seafloor not only change the mineralogy of the ocean crust, but also remove many important seawater constituents, such as magnesium and sulfate, while other constituents, such as calcium, are added. These reactions are thus generally responsible for maintaining the chemical composition of the oceans over geological time frames, and they participate in controlling the balance of the greenhouse gases such as CO 2 . Because the entire volume of the ocean is circulated through the ocean crust roughly once every million years, we must have a fundamental understanding of the rates, mechanisms, and pathways of ocean crust water- rock reactions so that we may better predict feedbacks such as those between climate change and seawater-crust ex- change. Many of the low-temperature water-rock reactions we have mentioned release energy, yet are kinetically sluggish; consequently, where conditions are otherwise suitable (ap- propriate temperature, availability of nutrients, etc.). this chemical energy could be used by microorganisms for met- abolic growth. Textural observations (Fisk et al., 1998; Torsvik et nl.. 1998; Furnes and Staudigel, 1999) and highly variable carbon isotope measurements (Furnes et al.. 2001 ) have indeed suggested that microbial activity is present in the ocean crust. These textural criteria which include rec- ognition and interpretation of "pit-textures," "sponge-tex- tures," and "zoned palagonite" can be easily and rapidly applied to a large number of samples for qualitative initial screenings and surveys, but they cannot provide definitive evidence of specific microbial activity in the crust. Further- more, studies of crust-hosted microbial communities have not yet elucidated how they might function physiologically. Hence, the ocean crust is an understudied, yet potentially vast, microbial habitat. Because sample access, contamina- tion, preservation, low biomass, and activity are problem- atic in the deep sea, many of the usual methods of detecting microbial communities and measuring their activities are not practical. Consequently, the actual fraction of ocean crust that is microbially altered is difficult to estimate. Textural signatures in the alteration products of certain rocks suggest that up to 75% of the uppermost crust is microbially altered (Furnes and Staudigel, 1999), whereas such features in other samples suggest that most alteration is probably abiogenic (Alt and Mata. 2000). Studies of micro- bial crust alteration have been infrequent, so we cannot conclusively assess the extent and importance of microbial activity within the ocean crust. Hence, the physical, chem- ical, and energetic regimes of young upper ocean crust must be considered with special care, so that specific predictions may be made and tested for use in focused environmental studies. To address the inferences made from previous geochemi- cal studies, we have explored microbial weathering reac- tions that may occur in the upper ocean crust during early- stage (crust <10 million years old) oxidative alteration of basaltic glass and sulfide minerals. In July of 2000, in situ weathering and colonization experiments were initiated at the Juan de Fuca Ridge axis off the northwestern coast of America (Edwards et ill.. 2003a). A variety of naturally occurring sulfide minerals were reacted for 2 months at the seafloor at low-temperature (^3C), ambient seafloor con- ditions. Upon collection, the sulfide surfaces were heavily colonized by bacteria and densely encrusted with weather- ing products that were largely composed of Fe oxides (Fig. 1 ). The mineralogical and fluorescent hybridization data (FISH; Fig. 1) suggest that these Fe oxide minerals owe their presence to the activity of neutrophilic Fe-oxidizing bacteria within surface pits (Edwards et al., 2()03a). Surface pits are ideal colonization sites for aerobic Fe-oxidizing bacteria because, once biofilms have formed and Fe oxide crusts have been produced, the bacteria are partially pro- tected from free advective (initially) and diffusive exchange with well-oxygenated deep-sea water. This allows pit-colo- nizing, Fe-oxidizing bacteria to more readily compete with very rapid abiotic oxidation kinetics of ferrous iron so that they may harness this oxidation energy for growth. Both culture-independent (molecular phylogenetic) and culture-dependent studies have been used to further explore the occurrence and diversity of mineral-oxidizing microor- ganisms. For both types of studies, we examined prokary- otic populations associated with weathering habitats in the deep sea (Fig. 2). These habitats include the surfaces of brec- ciated sulfide rubble material that derives from collapsed chimney, flange, or other structures common in hydrother- mal vent environments. We also examined metalliferous 182 K. J. EDWARDS ET AL. X%+Z- ,-v .'"U:- /iA;^P~ r-^w. Figure 1. Patlcrns ol cell .UK! oxide accumulation on the surface of a seafloor reacted sultidc mmeial. i A) DAPI-slained image of surface Ihat was heavily colonized by Bacteria (determined hy fluorescent in situ hybridizations; FISH); scale = 50 /jm. Cells are bright white dots and patches, predominately colonizing pores and pits on sample, the outlines of which are dark and coincide with the edges of the cell masses. I B.C I scanning electron micrographs of the same surface, showing thick Fe oxide accumulations over top of (B) or within (C) pits and pores on surface; scale 100 juni. (D) Fe oxides in C at higher magnification; scale = 10 urn. In B. Fe oxides form an effectne cap over the pits (arrows). In C. the oxides (arrows) are less well formed and are growing inside pits. We hypothesi/e that the more unformed Fe oxides in C are the precursors to the massive forms seen in B. sediments that accumulate in the vicinity of seafloor hydro- thermal sites as the result of plume events or the collapse of sulfide-anhydrite structures. Our culture-independent studies are based on compara- tive analysis of 16S rDNA sequences from uncultured or- ganisms present in environmental samples, and restriction fragment length polymorphism (RFLP; Hugenholt/ ct til., 1WK). RFLP analyses indicate that all of the weathering habitats examined are characterized by very low microbial diversity. In most cases, the microbial community is dom- inated by only one to three operational taxonomic units (OTUs; Moyer ct til.. IM94) or phylogenetically coherent taxonomic groupings as defined by RFLP analysis (D.R. Rogers, C.M. Santelli. and KJ. Edwards, unpubl. data). Comparative analyses of 16S rDNA sequences obtained from these uncultured organisms indicate that sulfur(S)- oxidizing bacteria that are to varying degrees related to the genus Thiomicrospira represent one dominant microbial group in these samples (Rogers and Edwards, unpubl. data). These observations are consistent with findings from other studies (both culture-dependent and culture-independent) that have been conducted at seatloor hydrothermal vent sites (e.g., Wirsen ct nl., 1993), and they support previous infer- ences that minerals play an important role in supporting the GEOM1CROBIOLOGY OF THE OCEAN CRUST 183 High-temperature vent (-250-400C) Low-temperature vent (-5-40C) Brecclated rubble Figure 2. Schematic cross-section of a submarine hydrothermal vent system, emphasizing sulfide weathering habitats. Mineral surfaces exposed to oxygenated water are favorable environments for aerobic lithoautolrophs that can oxidize the minerals to obtain metabolic energy. Such environ- ments include the surfaces of hydrothermal chimneys, brecciated rubble resulting from the collapse of extinct chimneys, and metalliferous sedi- ments formed by particulates settling out of the vent plume. In some instances, mounds of sediments and brecciated rubble are infiltrated by low-temperature hydrothermal fluids formed by mixing of high-tempera- ture fluids and seawater. providing aerobic environments at moderate temperatures (~25-40C). Figure modified after McCollom (2000) with permission from Elsevier. growth of S-oxidizing prokaryotes over geological time scales, long after hydrothermal activity dissipates (Eberhard et al.. 1995). In contrast to our FISH studies and microscopic obser- vations (Fig. 1), our culture-independent phylogenetic ap- proaches failed to support our findings that suggest that Fe-oxidizing microorganisms are present in low-tempera- ture weathering deposits. We did not identify any 16S rDNA sequences bearing similarity with gene sequences from any known Fe-oxidizing prokaryotes. This lack of sequence-based support for the presence of Fe-oxidizing microorganisms within environmental samples is not un- usual in studies of microbial weathering in the deep sea. For example, Thorseth et al. (2001 ) used both scanning electron microscopy and 16S rDNA sequence analysis on seafloor basalt glass to study microbial weathering at the seafloor. Although their SEM studies reveal Fe oxide particles re- markably similar to those observed in environmental and culture studies of neutrophilic Fe-oxidizing bacteria, their culture-independent, phylogenetic analyses failed to pro- duce any 16S rDNA sequences related to known Fe-oxidiz- ing species. In contrast to our culture-independent studies, our cul- ture-dependent studies have revealed a wide diversity of novel, autotrophic, Fe-oxidizing bacterial strains that previ- ously had no known Fe-oxidizing or autotrophic relatives represented in pure culture (Edwards et al., 2003b). Our culture techniques are based on the FeS gradient-tube method originally devised by Kucera and Wolfe ( 1957) and modified by Emerson and Moyer (1997). Using the same samples as were used for our culture-independent studies (above) for inoculum, we first initiated enrichment cultures on an organic carbon-free artificial seawater (ASW; modi- fied after Jannasch et al.. 1985; Edwards et al., 2003b) medium, using sulfide minerals such as pyrite as the sole energy source. Following initial enrichment, pure cultures of Fe-oxidizing bacteria were obtained by successive serial dilutions of enrichments to extinction in FeS gradient-tubes (Kucera and Wolfe, 1957). Putative isolates of Fe-oxidizing bacteria were determined to be clonally pure using RFLP analysis (Edwards et al.. unpubl. data). Physiological anal- yses have revealed that all of these isolates are obligate lithotrophs. capable of growth with Fe 2 + , but not with any alternate electron donors tested so far (sulfide, thiosulfate, Mn 2 + ). Autotrophy has been confirmed by determining 14 C fixation in FeS, FeS 2 , and basalt (~10 wt. % FeO)-based gradient tubes (Edwards et al., 2003b). Table 1 shows the phylogenetic affiliations among some of our Fe-oxidizing isolates. Most of our strains fall within the alpha- or gamma-subdivisions of the Proteobacteria and have moderately close relatives within broad groups of known heterotrophic bacteria, the Hypomicrobia and Mari- Tahle 1 Phylogenetic affiliations among axenic Fe-o.xidi'ing strains Strain number Bacterial division BLAST database match (% related) Metabolism inferred from closest relative FO1 a-Proteobacteria Hyphdinonu* jannaschiana (81%) Heterotrophy FO3 a-Proteobacteria Uncultured Marine bacterium SCRIPPS_94 (95%) Heterotrophy FO4 y-Proteobacteria Uncultured Marinobacter sp. PCOB-2 (94%) Heterotrophy FO6 y- Proteobacteria Uncultured Marinobacter sp. PCOB-2 (95%) Heterotrophy FO8 y-Proteobacteria Uncultured Marinobacter sp. ME108 (99%) Heterotrophy FOK) y-Proteobacteria Uncultured DCM-ATT-12 (90%) Unknown 184 K. J. EDWARDS ET AL. nobacter, respectively. If these sequences had not been derived from pure cultures of Fe-oxidizing lithoautotrophs. they (and thus their in situ physiological activity) would probably have been classified as heterotrophic on the basis of their phylogenetic relationships with known physiologi- cal groups. Our findings indicate that Fe-oxidizing autotrophs may be overlooked in culture-independent studies in the deep sea (if not other habitats as well) due to their close phylogenetic affiliations with physiologically distinct (heterotrophic) spe- cies. Our studies and others (e.g.. Emerson et /., 1999; Emerson and Moyer. 1997, 2002) clearly indicate that neu- trophilic Fe-oxidizing bacteria harbor unexpected diversity, which is just now becoming appreciated. This has important implications for how we study deep-sea microbial commu- nities within the context of their ecological and geochemical functions, and suggests critical shortcomings in the most commonly applied approaches based on 16S rDNA, com- parative gene sequence analysis. The occurrence of Fe-oxidizing capacity among an ever- broadening range of bacteria may have important ecological and evolutionary ramifications. For example, in relatively isolated environments, such as the oligotrophic, low-carbon, bare-rock oceanic crustal habitat, diverse groups of micro- organisms may converge on functions that are well suited to the particular environment. Alternatively, the capacity for Fe-oxidation may have been lost by some species after they occupied more carbon-rich habitats. Such occurrences among marine lithoautotrophs may be supported by an alpha-Proteobacterium that possesses genes for both the large and small subunits of ribulose-l,5-bisphosphate car- boxylase/oxygenase (RubisCO). yet the heterotrophic strain has not been shown to fix carbon (Francis et ai, 2001). Indeed, although heterotrophic Fe- and Mn-oxidizing bac- teria have long been recognized in the environment, the physiological purpose of this oxidation is often unknown (Ghiorse, 1984. and references therein). Rather than serving some explicit biological role, this oxidation in some species may be an evolutionary holdover that no longer has physi- ological relevance. As one final explanation, Fe-oxidation among groups that we would typically classify as hetero- trophic may be an example of a multifunctional metabolism that allows them to adapt to a rapidly changing environ- ment. Further studies, both laboratory and field-based, are needed to explore the implications of microbial activity within the ocean crust. Studies on Fe-oxidizing bacteria should provide critical information about sub-seafloor com- munities and biogeochemical processes. Useful laboratory investigations of Fe-oxidizing bacteria would include the following. ( 1 ) Studies to determine the mechanism of Fe- oxidation among known strains of Fe-oxidizing bacteria: knowledge of the pathway and key genes and enzymes for Fe-oxidation could be used to develop molecular diagnos- tics for this activity, and these could be applied in environ- mental settings. (2) Isotopic studies both to define the car- bon fixation pathways and associated stable carbon isotope fractionations and to determine any stable Fe isotope frac- tionation that occurs during oxidation in pure cultures: these studies are key to developing geochemical diagnostics that can be applied to rocks long after activity diminishes. Acknowledgments Funding for this work was provided by NSF grants OCE- 0096992 and EAR-0073998 to KJE. Special thanks to present and past members of the Edwards lab for various contributions to this work, and to two anonymous reviewers for their helpful comments and insights. WHOI contribution number 10876. Literature Cited Alt, J. C. 1995. Sub seafloor processes in mid-ocean ridge hydrothermul systems. Pp. 85-114 in Seafloor Hydrolhennal Systems, S.E. Humphris. R.A. Zierenberg. L.S. Mullineaux. and R.E. Thomson, eds. American Geophysical Union. Washington, DC. Alt, J. C., and P. Mata. 2000. On the role of microbes in the alteration of submarine basaltic glass: A TEM study. Earth Plane!. Sci. Leu. 181: 301-313. Alt, J. C.. D. A. Teagle, C. Laverne. D. A. Vanko, W. Bach et al. 1996. Ridge flank alteration of upper oceanic crust in the eastern Pacific: A synthesis of results for volcanic rock of Holes 504B and 896A. Pp. 435-450 in Proceedings of the Ocean Drilling Program Scientific Result.', I4S. J. C. Alt, H. Kinoshita. L. B. Stokking, and P. J. Michael, eds. Ocean Drilling Program. Texas A & M University. College Sta- tion. TX. de Villiers, S., and B. K. Nelson. 1999. Detection of low-temperature hydrothermal fluxes by seawater Mg and Ca anomalies. Science 285: 721-723. Eberhard, C., C. O. Wirsen. and H. W. Jannasch. 1995. Oxidation of polymetal sulfides by chemolithoautotrophic bacteria from deep-sea hydrothermal vents. Geomicmhiol. J. 13: 145-164. Edwards, K. J., T. M. McCollom. H. Konishi, and P. R. Buseck. 2003a. Seafloor bio-alteration of sultide minerals: results from I'M situ incuba- tion studies. Geochim. Cosmochim. Ada. (In press). Edwards, K. J., D. R. Rogers, C. O. Wirsen, and T. M. McCollom. 2003h. Isolation and characterization of novel psychrophilic, neutro- philic. Fe-oxidizing chemolithoautotrophic alpha- and gamma-Pro- teobacteria from the deep-sea. Appl. Em-iron. Microbiol. (In press). Elderfield, H., C. G. Wheat, M. J. Mottl, C. Monnin, and B. Spiro. 1999. Fluid and geochemical transport through oceanic crust: A transect across the eastern flank of Juan de Fuca Ridge. Earth Planet. Sci. Lett. 172: 151-165. Emerson, D., and C. L. Moyer. 1997. Isolation and characterization of novel iron-oxidizing bacteria that grow at circumneutral pH. Appl. Environ. Microbiol. 63: 4784-4792. Emerson, D., and C. L. Moyer. 2002. Neutrophilic Fe-oxidizing bac- teria are abundant at the Loihi seamount hydrothermal vents and play a major role in Fe oxide deposition. Appl. Environ. Microbiol. 68: 3085-3093. Emerson, IX, J. V. Weiss, and J. P. Megonigal. 1999. Iron-oxidizing bacteria are associated with ferric hydroxide precipitates (Fe-plaque) on the roots of wetland plants. Appl. Environ. Microbiol. 65: 2758- 2761. GEOMICROBIOLOGY OF THE OCEAN CRUST 185 Fisher, A. T. 1998. Permeability within basaltic ocean crust. Rev. Geo- phys. 36: 143-182. Fisher, A. T., and K. Becker. 2000. Channelized fluid flow in oceanic crust reconciles heat-flow and permeability data. Nature 403: 71-74. Fisk, M. R., S. J. Giovannoni, and I. H. Thorseth. 1998. Alteration of oceanic volcanic glass: textural evidence of microbial activity. Science 281: 978-980. Francis, C. A., E. M. Co, and B. M. Tebo. 2001. Enzymatic manganese (II) oxidation by a marine alpha-proteobacterium. Appl. Environ. Mi- crobiol. 67: 4024-4029. Furnes, H., and H. Staudigel. 1999. Biological mediation in ocean crust alteration: how deep is the deep biosphere? Earth Planet. Sci. Lett. 166: 97-103. Furnes, H., K. Muehlenbachs, T. Torsvik, I. H. Thorseth, and (). Tumyr. 2001. Microbial fractionation of carbon isotopes in altered basaltic glass from the Atlantic Ocean, Lau Basin, and Costa Rica Rift. Chem. Geol. 173: 313-330. Gallahan, W. E., and R. A. Duncan. 1994. Spatial and temporal vari- ability in crystallization of celadonites within the Troodos ophiolite. Cyprus: implications for low temperature alteration of the oceanic crust. J. Geophys. Res. 99: 3147-3161. Ghiorse, W. C. 1984. Biology of iron-and manganese-depositing bacte- ria. Anna. Rev. Microbiol. 38: 515-550. Hart, S. R., and H. Staudigel. 1986. Ocean crust vein mineral deposition: Rb/Sr ages, U-Th-Pb geochemistry, and duration of circulation of DSDP sites 261, 462. and 516. Geochim. Cosmochim. Acta 50: 2751-2761. Honnorez, J. 1981. The aging of the oceanic crust at low temperature. Pp 525-587 in The Sea. C. Emiliani, ed. Wiley, New York. Hugenholtz, P., C. Pitulle, K. L. Hershberger, and N. R. Pace. 1998. Novel division level bacterial diversity in a Yellowstone hot spring. J. Bacterial. 180: 366-376. Jannasch, H. W., C. O. Wirsen, D. C. Nelson, and L. A. Roberson. 1985. Thiomicrospira crunogena sp. nov., a colorless, sulfur-oxidiz- ing bacterium from a deep-sea hydrothermal vent. Int. J. Syst. Bacte- rial. 35: 422-424. Kucera, S., and R. S. Wolfe. 1957. A selective enrichment method for Gallionella. ferruginea. J. Bacterial. 74: 344-349. McCollom, T. M. 2000. Geochemical constraints on primary productivity in submarine hydrothermal vent plumes. Deep-Sea Res. I 47: 85-KM. Mni 1 1. M. J., and C. G. Wheat. 1994. Hydrothermal circulation through mid-ocean ridge flanks: fluxes of heat and magnesium. Geochim. Cosmochim. Ada 58: 2225-2237. Moyer, C. L., F. C. Dobbs, and D. M. Karl. 1994. Estimation of diversity and community structure through restriction fragment length polymor- phism distribution analysis of bacterial 16S rRNA genes from a mi- crobial mat at an active, hydrothermal vent system, Loihi Seamount, Hawaii. Appl. Environ. Microhiol. 60: 871-879. Peterson, C., R. Duncan, and K. F. Scheidegger. 1986. Sequence longevity of basalt alteration at Deep Sea Drilling Project Site 597. Pp. 505-515 in Initial Reports DSDP 92M. D. K. Leinen and E. A. Rea, eds. U.S. Government Printing Office, Washington, DC. Staudigel, H., T. Plank, B. While, and H.-U. Schmincke. 1996. Geo- chemical fluxes during seafloor alteration of the basaltic upper oceanic- crust: DSDP Sites 4 1 7 and 4 1 8. Pp 1 9 -38 in Subduction Top to Bottom. G.E. Bebout. S.W. Scholl, S.H. Kirby. and J.P. Platt, eds. American Geophysical Union, Washington, DC. Thorseth, I. H., T. Torsvik, V. Torsvik, F. L. Daae, R. B. Pedersen, Keldysh-98 Scientific Party. 2001. Diversity of life in ocean floor basalt. Earth Planet. Sci. Lett. 194: 31-37. Torsvik, T., H. Furnes, K. Muehlenbacks, I. H. Thorseth, and O. Tumyr. 1998. Evidence for microbial activity at the glass-alteration interface in oceanic basalts. Earth Planet. Sci. Lett. 162: 165-176. Wheat, C. G., and M. J. Mottl. 2000. Composition of pore and spring waters from Baby Bare: global implications of geochemical fluxes from a ridge flank hydrothermal system. Geochim. Cosmochim. Acta 64: 629-642. Wirsen, C. O., H. W. Jannasch, and S. J. Molyneaux. 1993. Chemo- synthetic microbial activity at Mid-Atlantic Ridge hydrothermal vent sites. J. Geophys. Res. 98: 9693-9703. Reference: Biol. Bull. 204: 186-191. (April 2003) 2003 Marine Biological Laboratory Genomic Markers of Ancient Anaerobic Microbial Pathways: Sulfate Reduction, Methanogenesis, and Methane Oxidation ANDREAS TESKE'-*'t ASHITA DHILLON 2 . AND MITCHELL L. SOGIN 2 l Biology Department. Woods Hole Oceanographic Institution. Woods Hole. Massachusetts 02543: and 'Josephine Bay Paul Center for Comparative Molecular Biologv and Evolution. Marine Biological Laboratory, Woods Hole. Massachusetts 02543 Genomic markers for anaerobic microhial processes in marine sediments sulfate reduction, methanogenesis, and anaerobic methane oxidation reveal the structure of si'l- fate-reducing, methanogenic, and methane-oxidizing micro- hial communities (including uncultured members): thev al- low inferences about the evolution of these ancient microhial pathways: and the\ open genomic windows into extreme microbial habitats, such as deep subsurface sedi- ments and hydrothermal vents, thai arc analogs for the early Earth and for extraterrestrial microbiota. Sulfate reduction and methanogenesis are two terminal anaerobic bioremineralization pathways that convert low- molecular-weight products of other bacterial processes (degradation of polymers, fermentation) to CO 2 and meth- ane. Sulfate-reducing bacteria are physiologically and phy- logenetically highly diverse (Castro et a!.. 2000; Widdel and Bak. 1992): they oxidize a wide variety of low-molecular- weight compounds (short-chain fatty acids, alcohols, al- kanes, aromatic compounds, acetate) to CO,. In marine sediments, the range of sulfate-reducing bacteria is limited by sulfate availability. When sulfate is depleted, methano- t Current Address: Deparlmem of Marine Sciences. CB # 3300. Venable Hall 12-1. University of North ( mlma at Chapel Hill, Chapel Hill, NC 27599. *To whom correspondence shouki be addressed. E-mail: teske@email. unc.edu The paper was originally presented at a workshop titled Outcome* of Genome-Genome Inicrafiinnx. The workshop, winch was held at the J. Erik Jonsson Center of the National Academy of Sciences. Woods Hole. Massachusetts, from 1-3 May 2002. was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266. genie archaea become the dominant anaerobic microbial population. Autotrophic methanogens utilize hydrogen as energy source for the reduction of CO, to methane; special- ized genera of methanogens are also capable of inter- and intramolecular disproportionate of C t and C-, carbon com- pounds (methanol. methylamines, acetate) to methane and CO 2 (Boone et a/., 1993). Where methane and sulfate co- exist (for example, at the interface of sulfate-reducing and methanogenic sediment layers, or at marine methane seeps and vents), sulfate-dependent anaerobic methane oxidation takes place; methane of biogenic origin is oxidized to CO-, with sulfate as the terminal electron acceptor (Valentine and Reeburgh. 2000). As proposed originally on the basis of biogeochemical field data and thermodynamic consider- ations (Hoehler et al.. 1994), anaerobic methane oxidation is carried out by syntrophic consortia of methanotrophic archaea and sulfate-reducing bacteria, in which the sulfate- reducing partner catalyzes the electron transfer from meth- ane to sulfate and assimilates a portion of the methane oxidation products (Boetius ct al.. 2000; Orphan et al., 20()lb). Methanotrophic archaea of different phylogenetic affiliation can form dense, highly ordered clusters with sulfate-reducing syntrophs. or may occur in less tight asso- ciations (Orphan et al.. 2002). The sulfate-reducing, methanogenic, and methane-oxi- dizing microbial populations that are found in anaerobic marine microbial ecosystems today are the modern descen- dants of ancient microbiota whose isotopic imprints are pervasive in the carbon- and sulfur-isotopic record, from the present back to the Archaean-Proterozoic transition (Knoll and Canfield, 1998). The only oxidants that these pathways require are CO : or carbonate, and sulfate, which existed before the photosynthetic oxygenation of the Earth's bio- 186 GENOMIC MARKERS OF ANAEROBIC MICROBIAL PATHWAYS 187 sphere and the appearance of free oxygen in the atmosphere and the marine water column. Isotopic evidence for widely expressed microbial sulfate reduction, in the form of 34 S- depleted sedimentary sulfides. goes back to the middle and early Proterozoic. 2.2 to 2.3 billion years ago (Canfield et til.. 2000). The mineralization of organic matter by metha- nogenesis, followed by methane oxidation, may even pre- date the onset of sulfate reduction. The carbon isotopic imprint of this process, in the form of highly ' ^C-depleted kerogen (5 U C > -60%c), is found in late Archaean and early Proterozoic kerogens, 2.8 billion years old (Strauss and Moore, 1992). This isotopic record was originally in- terpreted as evidence for widespread aerobic methane oxi- dation (Hayes, 1994). Anaerobic methane oxidation is more likely, since evidence for the stepwise and pervasive oxy- genation of the proterozoic biosphere begins to appear only at a later time, about 2.2 billion years ago (Des Marais et ul.. 1992). Key Genes for Sulfate Reduction and Anaerobic Methane Cycling The antiquity and evolutionary significance of these mi- crobial pathways is shown in the high degree of phyloge- netic conservation of their key genes and key enzymes. In sulfate-reducing prokaryotes, the aps gene codes for the key enzyme adenosine-5'-phosphosulfate reductase, which cat- alyzes the activation and subsequent reduction of sulfate to sullite (Friedrich, 2002). A second key gene of dissimilatory sulfate reduction, dsrAB, codes for the alpha and beta sub- units of the enzyme dissimilatory sulfite reductase, which catalyzes the reduction of sulfite to sulfide (Wagner el ul., 1998). The dsrAB and aps genes are phylogenetically con- served in several deeply branching phyla of bacterial and archaeal sulfate reducers. When specific gene transfer events are taken into account, the dsrAB and aps genes allow a simultaneous phylogenetic and metabolic identifi- cation of sulfate-reducing prokaryotes (Klein et ul.. 2001; Friedrich, 2002). Coenzyme M methyl reductase is the key enzyme of methanogenesis; it catalyzes the terminal and highly exer- gonic step of the methanogenesis pathway, the reduction and release of the coenzyme-M-bound methyl group as free methane. The Coenzyme M methyl reductase gene (mrcA) is found in methanogenic archaea; it is sufficiently con- served and consistent with 16S rRNA phylogenies to allow the identification of methanogenic archaeal lineages in en- vironmental samples (Springer et ul.. 1995; Lueders et ul.. 2001; Ramakrishnan etal.. 2001 ). At present it is not known whether anaerobic methane-oxidizing archaea are using a version of this enzyme for the activation and reoxidation ot methane. If anaerobic methane oxidation by archaea could proceed through a reversal of classical methanogenesis pathways, the Coenzyme M methyl reductase reaction would be the most difficult and energy-demanding step to reverse (Hoehler and Alperin. 1996). Current full-genome sequencing efforts using purified ANME-1 and ANME-2 archaea from environmental samples are testing whether the genomes of these methanotrophic archaea carry coenzyme M methyl reductase genes (Orphan el ul.. 2002). The Guaymas Basin Hydrothcrmal Vent Sites as Model System To search for deeply branching and (possibly) ancestral representatives of sulfate-reducing, methanogenic, and methane-oxidizing microorganisms and their key genes in modern environments, we focus on hydrothermal vent hab- itats. Hydrothermal vents represent some of the earliest and best protected microbial habitats that may have survived repeated environmental disturbances in the surface bio- sphere; vents can in principle occur on every planet with oceans and active plate tectonics or volcanism. On Earth, hydrothermal vents sustain complex microbial ecosystems that utilize inorganic energy sources (such as sulfide, hy- drogen, and reduced metals) and geothermal sources of carbon (such as methane. CO^, and geothermally synthe- sized low-molecular-weight organic compounds) (Kelley et ul.. 2002). The hydrothermally active sediments of the Guaymas Basin (Gulf of California. Mexico) provide a relatively well-studied model system for the complexity of the microbial communities that are involved in sulfate re- duction, methanogenesis, and methane oxidation. Cultiva- tions, lipid biomarker analyses, 16S rRNA. and functional gene sequencing are beginning to reveal unusually complex microbial communities that include sulfate-reducing pro- karyotes, methanogenic archaea, and anaerobic methanotro- phic archaea and their sulfate-reducing syntrophs (Fig. 1 ). Specifically, results of the Guaymas Basin survey (Teske et ul.. 2002; Dhillon et ul.. unpubl.) will also help to identify novel sulfate-reducing. methanogenic. and methane-oxidiz- ing microorganisms in deep subsurface sediments, where these processes are predominant (D'Hondt et ul., 2002). These anoxic environments represent analogs to subsurface life under extraterrestrial conditions in which an inhospita- ble surface environment might have driven microbial life underground or never allowed its evolution within a pho- totrophic. oxygenated biosphere. Guaymas Basin Microbial Communities Sulfate-reducing bacteria and archaea are dominant ter- minal oxidizers of organic matter in the Guaymas Basin, as shown by high rates of sulfate reduction measured over wide temperature ranges up to about 100C (Jorgensen el ul.. 1990. 1992; Elsgaard ct ul.. 1994; Weber and Jorgensen, 2002). Hyperthermophilic, autotrophic, or mixotrophic ar- chaea of the genus Archueoi>lobus were found by cultiva- tion (Burggraf et ul.. 1990) and 16S rRNA sequencing 188 A. TESKE ET AL Beggiatoa spp., filamentous sulfide- oxidizing bacteria at the oxic/anoxic water/sediment interface sediment - water anaerobic methane-oxidizing consortia archaea (red), bacteria (green) H Sulfate reducers, here the hyperthermophilic archaeon Archaeoglobus profundus H 2 S + NO 3 > H 2 O -> S0 4 2 H 2 S + 2O 2 -->SO 4 2 +2H + Prokaryotes involved in sulfur and methane cycling in the Guaymas Basin CH 4 + SO 4 2 "> HCO 3 '+ HS'+ H 2 O SO 4 2 + 4 H 2 --> H 2 S + 2 OH" SO 4 2 + CH 3 COOH + H + -> HS> 2 CO 2 + 2 H 2 O 2 H 2 O t CH 4 from vent fluid, and pyro- lysis of organic substrates Methanogenic archaea, here the hyperthermophile Methanopyrus kandleri C0 2 + 4 H 2 --> CH 4 + 2 H 2 O CO 2 H 2i acetate, from vent fluid and pyrolysis of organic compounds Figure 1. Simplified scheme of microorganisms and their reactions (sulfate reduction, methanogenesis, methane oxidation, sulfide oxidation ) in the methane and sulfur cycles in the Guaymas Basin hydrothermal vents. Clockwise, sulfate-reducing bacteria and archaea (Archaeoglobus profimdiis), methanogenic archaea (Melh- niKipynis kandleri), methane-oxidizing consortia, and sultide-oxidizing bacteria (Beggiatoa spp. I. Fluorescence in situ hybridization image of anaerobic methane-oxidizing consortium, courtesy K. Knittel and A. Boetius (Max Planck Institute for Marine Microbiology. Bremen, Germany). The bacterial and archaeal species shown here are representatives of metabolically and phylogenetically diversified functional classes of prokaryotes. It has to be noted that sulfate reduction and sulfate-dependent methane oxidation are almost certainly uncoupled in the Guaymas sediments. The highly diversified sulfate-reducing prokaryotic community can oxidize a wide range of substrates (H : and acetate are just the simplest examples), whereas methane-oxidizing syntrophs are most likely restricted to methane oxidation intermediates provided by their archaeal partners. (Teske et al., 2002). Moderately thermophilic or mesophilic fatty acid oxidizing sulfate reducers have been cultured from Guaymas (Rueter et al.. 1994). Surveys with 16S rRNA detected predominantly members of the propionate- oxidizing, acetate-producing family Desulfobulbaceae (Teske et al., 2002) and members of the acetate-oxidizing family Desulfobacteriaceae (Dhillon et fragment length (nucteotides) D Grown 120 hours with viruses Grown 120 hours without viruses Figure 1. Analysis, by terminal restriction fragment length polymorphism, of bacterial community compo- sition in seawater inocula (filtered to remove protists) grown in seawater that had been filtered to remove all cellular organisms and viruses ("without viruses") or cellular organisms only ("with viruses"). Each bar represents an operational taxonomic unit, and its height is proportional to the amount of amplified DNA in that particular fragment. Error bars are one standard error of the mean. ronments are not characterized, and we know very little about the overall process. Virus effects on bacterial evolution can stem from both negative and positive interactions. Negative interactions include infection-mediated mortality that can provide selec- tive pressure. Resistance of hosts and viral adaptation to resistance is probably an ongoing process it can be viewed as a "war," with measures and counter-measures and con- stant adaptation of both partners. It would seem that viruses are unlikely to lead to extinctions, because as the host population size got smaller, the viruses would have more and more difficulty in "finding" a host by diffusion. Nev- ertheless, a local reduction in population size might cause an evolutionary bottleneck. Positive interactions include viral conversion whereby viral genetic material codes for new host capabilities. Also, bacteria may use viruses as a nutri- tion source, maybe even using "decoy" receptors to lure "unsuspecting" viruses to attempt infection (Fuhrman, 1999). There are obviously additional evolutionary effects of genetic exchange mediated by viruses. On evolutionary time scales, this process helps to homogenize genomes, especially within close relatives, which may help to keep a bacterial "species" together. Acknowledgments This work was supported by NSF grants OCE9906989 and DEB0072770. We also thank Lita Proctor, Rachel No- ble, Robin Wilcox. and Ian Hewson for helpful conversa- tions and assistance with experiments. Literature Cited Ackermann, H.-W., and M. S. DuBow. 1987. Viruses of Prokaryotes. Vol. I. General Properties of Bacteriophages. CRC Press. Boca Raton. FL. Avaniss-Aghajani, .., K. Jones, D. Chapman, and C. Brunk. 1994. A molecular technique for identification of bacteria using small subunit ribosomal RNA sequences. Bioteelmiques 17: 144-149. Bergh, O., K. Y. Barsheim, G. Bratbak, and M. Heldal. 1989. High abundance of viruses found in aquatic environments. Nature 340: 467-468. Fuhrman, J. A. 1999. Marine viruses: biogeochemical and ecological effects. Nature 399: 541-548. Fuhrman, J. A., and R. T. Noble. 1995. Viruses and protists cause similar bacterial mortality in coastal seawater. Limnol. Oceanogr. 40: 1236-1242. Gonzalez, J. M., R. Simo, R. Massana, J. S. Covert, E. O. Casa- mayor, C. Pedros-Alio, and M. A. Moran. 2000. Bacterial com- munity structure associated with a dimethylsulloniopropionate-pro- ducing North Atlantic algal bloom. Appl. Environ. Microbiol. 66: 4237-4246. Jiang, S. C., and J. H. Paul. 1998. Gene transfer by transduction in the marine environment. Appl. Environ. Microbiol. 64: 2780-2787. Noble, R. T., and J. A. Fuhrman. 2000. Rapid virus production and removal as measured with fluorescently labeled viruses as tracers. Anpl. Environ. Microbiol. 66: 3790-3747. Proctor, I,. M., and J. A. Fuhrman. 1990. Viral mortality of marine bacteria and cyanobacteria. Nature 343: 60-62. Riemann, L., and M. Middelboe. 2002. Viral lysis of marine bacterio- plankton: implications for organic matter cycling and bacterial clonal composition. Ophelia 56: 57-68. Ripp, S., O. A. Ogunseitan, and R. V. Miller. 1994. Transduction of a freshwater microbial community by a new Pseudomonas aeruginnsa generalized transducing phage, Utl. Mol Ecol 3: 121-126. BACTERIA-VIRUS INTERACTIONS 195 Steward, G. F., J. Wikner, D. C. Smith, W. P. Cochlan, and F. Azam. 1992. Estimation of virus production in the sea: I. Method develop- ment. Mar. Microb. Food Webs 6: 57-78. Thingstad, T. F., and R. Lignell. 1997. Theoretical models for the control of bacterial growth rate, abundance, diversity and carbon de- mand. Aquat. Microb. Ecol. 13: 19-27. Waterbury, J. B., and F. W. Valois. 1993. Resistance to co-occurring phages enables marine Synechococcus communities to coexist with cyanophages abundant in seawater. Appl. Environ. Microbiol. 59: 3393-3399. Wilcox, R. M., and J. A. Fuhrman. 1994. Bacterial viruses in coastal seawater: lytic rather than lysogenic production. Mar. Ecol. Prog. Ser. 114: 35-45. Wilhelm, S. W.. and C. A. Suttle. 1999. Viruses and nutrient cycles in the sea viruses play critical roles in the structure and function of aquatic food webs. Bioscience 49: 781-788. Wilhelm, S. W., S. M. Brigden, and C. A. Suttle. 2002. A dilution technique for the direct measurement of viral production: a compari- son in stratified and tidally mixed coastal waters. Microb. Ecol. 43: 168-173. Wommack, K. E., and R. R. Colwell. 2000. Virioplankton: viruses in aquatic ecosystems. Microbiol. Mol. Bio/. Rev. 64: 69-114. Reference: Biol. Bull 204: 196-199. (April 2003) 2003 Marine Biological Laboratory A(r)Ray of Hope in Analysis of the Function and Diversity of Microbial Communities MARTIN F. POLZ*. STEFAN BERTILSSONt, SILVIA G. ACINAS. AND DANA HUNT Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 The vast majority of microorganisms in the environment remain uncultured, and their existence is known only from sequences retrieved by PCR. As a consequence, our under- standing of the ecological function of dominant microbial populations in the environment is limited. We will review microbii.ll diversifv studies and show that these may have moved from an extreme underestimation to a potentially severe overestimation of diversity. The latter results from a simple PCR-generated artifact: the cloning of heteroduplex molecules followed by Escherichia coli mismatch repair, which may generate an exponential increase in obsen-ed sequence diversity. However, simple modifications to cur- rent PCR amplification protocols minimize such artifuctnal sequences and may bring within our reach estimation of bacterial diversity in environmental samples. Such esti- mates may spur new culture-independent approaches based on genomic and microarray technology, allowing correla- tion of phvlogenetic identitv with the ecological function of unculturable organisms. In particular, we are developing a DNA microarray that enables identification of individual populations active in utilization of specific organic sub- strates. The array consists of 16S and 23S rDNA-targeted oligonucleotides and is hybridized to RNA extracted from samples incubated with n C-\abe\ed organic substrates. Populations that metabolize the substrate can be identified h\ the radiolabel incorporated in their rRNA after only one * To whom correspondence should be addressed. E-mail: mpolzfs' mit.edu tPresent address: Department of Evolutionary Biology, Limnology, Uppsala University. Norbyvagen 20, SE-75236. Uppsala, Sweden. The paper was originally presented at a workshop titled Outcomes of Genome-Genome Interactions. The workshop, which was held at the J. Erik Jonsson Center of the National Academy of Sciences. Woods Hole. Massachusetts, from 1-3 May 2002, was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266. to two cell doublings, ensuring realistic preservation of community structure. Thus, the microarray approach may provide a powerful means to link microbial community structure with in situ function of individual populations. The last two decades have seen a radical shift in our understanding of microbial diversity. Previously the number of bacterial and archaeal species had been estimated to be in the thousands. It is now generally accepted that the number may actually be as high as several million (Torsvik et al., 2002). This change was brought about by a gradual replace- ment of diversity estimates based on pure-culture isolation of strains with a determination of diversity based on co- occurring gene sequences, largely ribosomal RNA (rRNA) genes (Head et al., 1998). Although early attempts were made to screen diversity by shotgun cloning of environmen- tal DNA, with subsequent detection and sequencing of rRNA gene inserts (Schmidt et al.. 1991), large-scale ap- plication of the molecular approach was dependent on PCR protocols that allow the enrichment of rRNA genes from genome mixtures using universal primers (Head et al., 1998). Today, the assessment of the entire diversity of rRNA sequences (ribotypes) coexisting within specific mi- crobial communities has become a realistic possibility due to the ease of PCR implementation and the increased avail- ability of high-throughput sequencing facilities. Although the exact magnitude of microbial diversity still remains an open question, the PCR-based approach has led to the retrieval of large numbers of sequences from almost any environment examined (Hugenholtz et al., 1998). Thus extensive comparative databases are now available from which patterns of microbial community structure are begin- ning to emerge. For example, studies of bacterioplankton diversity in the ocean, which represents one of the best- studied environments, have shown that, surprisingly, the major phylogenetic groups in the open ocean and the coastal ocean are similar, despite marked differences in trophic 196 FUNCTION AND DIVERSITY OF MICROBIAL COMMUNITIES 197 state and habitat quality between these environments (Gio- vannoni and Rappe, 2000). Such observations have pro- vided important insights; yet they also highlight the major problem of the molecular diversity approach: because very few of the retrieved sequences have closely related cultured representatives available, the ecological role of an organism in question cannot even be guessed (Hugenholtz et al.. 1998). Furthermore, even when closely related cultured organisms exist, they can display quite significant genomic, physiological, or metabolic differences (Gray and Head, 2001). New alternatives for diversity studies, such as anal- ysis of large genome fragments retrieved from the environ- ment (Beja et al.. 2000) or gene cassette PCR for recovery of complete open reading frames from environmental DNA (Stockes et al., 2001), can enhance our understanding of uncultured organisms. Nonetheless, elucidation of structure- function relationships or niche differentiation of populations within microbial communities remains one of the big chal- lenges in microbial ecology. During the last few years, molecular diversity studies have been augmented with tracer techniques that allow assignment of biogeochemical function to uncultured mi- crobial populations [recently reviewed by Gray and Head (2001 )]. Most notably, combined microautoradiography and in situ hybridization (STAR- or MICROFISH) (Lee et al.. 1999; Ouverney and Fuhrman, 1999; Cottrell and Kirch- man, 2000) or stable isotope probing (Boschker et al.. 1998; Radajewski et al., 2000) allow identification of microbial populations responsible for the metabolism of specific or- ganic compounds. In both cases, environmental samples are incubated with isotopically labeled substrates. In STAR- or MICROFISH, microautoradiography and in situ hybridiza- tion are carried out on the same microscope slide with the goal of matching uptake of radiochemicals with phyloge- netic identification on the single-cell level. In stable isotope probing, either lipid biomarkers (Boschker et al.. 1998) or DNA (Radajewski et al., 2000) are extracted from commu- nities incubated with 13 C-labeled compounds. If cells grow on the added compounds, their pool of macromolecules will be isotopically heavy compared to those of metabolically inactive organisms. This makes it possible to identify the organism in one of two ways: ( 1 ) by mass spectrometry of labeled "signature" lipids (Boschker et al., 1998); or (2) by separation by ultracentrifugation of community DNA ac- cording to mass differences, followed by identification of rRNA genes in the isotopically heavy DNA pool by PCR, cloning, and sequencing (Radajewski et al., 2000). The above approaches have already produced interesting insights into the ecological roles of uncultured Bacteria and Archaea. For example, using MICROFISH, it was demon- strated that low-temperature Archaea, which represent a dominant group in deep ocean water but are currently known only from rRNA gene clone libraries, readily take up amino acids at low ambient concentrations (Ouverney and Fuhrman, 1999). In another study, incubation of anaerobic sediments with l3 C-acetate yielded signature lipids of gram- positive bacteria rather than those of the more readily iso- lated delta Proteobacteria sulfate-reducing bacteria (Bosch- ker et al., 1998). However, each of the techniques has distinct drawbacks. Stable isotope probing requires very high substrate concentrations and long incubation times, so that the procedure actually resembles enrichment cultures (Radajewski et al.. 2000); MICROFISH involves labor- intensive hybridization and microautoradiography, which limits the number of populations whose metabolism can be explored. We are currently developing a combination of DNA microarrays and radiotracer incubations, the "functional di- versity array," as a high-throughput complement to the above methods (Bertilsson and Polz, 2001). DNA microar- rays can carry hundreds to thousands of specific nucleic acid probes, which are arrayed in discrete spots. During the hybridization process, these probes capture their specific target from mixtures of templates. If these are either radio- actively or fluorescently labeled, the presence or, to a cer- tain extent, the quantity of all specific templates for which probes have been spotted can be ascertained. The applica- tion of DNA microarrays for screening and monitoring microbial community structure by arraying rRNA-specih'c oligonucleotide probes is being explored by a number of laboratories (Cho and Tiedje, 2001; Small et al.. 2001: Koizumi et al., 2002). In the functional diversity array, diversity screening is combined with detection of popula- tions responsible for specific transformations in the com- munity (Fig. 1 ). Samples are spiked with l4 C-labeled com- pounds, leading to incorporation of radionuclides into rRNA of populations that actively metabolize the compound of interest (Bertilsson and Polz, 2001 ). RNA is subsequently extracted, fluorescently labeled, and hybridized to the mi- croarray, which contains oligonucleotide probes specific for each "ribotype" in the community. Radioactivity in each spot due to hybridization can be determined by either mi- croautoradiography or phosphor-imaging, so that in combi- nation with the fluorescent signal, a specific activity can be estimated for each population (Fig. 1). For the functional diversity array to be generally appli- cable, differentiation of populations by the arrayed probes is not, by itself, sufficient. In addition, several critical ques- tions must be evaluated. First, what is the detection limit for 14 C-labeled rRNA hybridized to the array? Second, can realistic substrate concentrations be used, and what are the kinetics of rRNA synthesis after uptake of label under environmental conditions'? Third, to what extent can an entire microbial community be represented on the array? Below, we evaluate these questions, with special emphasis on approaches for studying rRNA gene diversity in micro- bial communities as a necessary precondition for determin- ing biogeochemical activity of previously unidentified pop- ulations. Quantification of the radioactive signal on arrays shows 198 M. F. POLZ ET AL Environmental rRNA clone library Incubation with 14 C-substrates Extraction of target rRNA 14 C UV/EtBr 250 200 150 100 - 50 - # Unique ribotypes # Screened clones 400 800 1200 Probe synthesis & Immobilization on array I * ,' t I . H *i> >'*>* > : 3&S8tlS&S&&S^ : *"******* ' '*' ^iJiWW^wUr^??^ Hybridization -Detection of captured rRNA -Identification of (active) ribotypes Figure 1. Outline of the experimental approach used for the functional diversity array. that the approach is sensitive enough to detect populations at naturally occurring levels. The use of phosphor-imaging screens, to which the entire array is exposed, allows 14 C hound to each spot to be quantified. We have experimentally determined the detection limit for 14 C on these screens to be 0.1 DPM for a spot 150 jam in diameter, which consists of about 10 8 oligonucleotide probes. Assuming that the rRNA molecules fragment to about 300 bp and that only 1% of the oligonucleotide probes will be bound to templates after hybridization, the detection limit is between 10 2 and 10 3 cells. This estimate is based on a cellular rRNA content between 1,000 and 10,000 molecules, which is in the range of slow- and fast-growing cells, respectively. Using this procedure, we have been able to specifically detect and differentiate rRNA from sulfate-reducing strains grown on l4 C-labeled lactate (Klepac and Polz, unpubl. data). For the array to represent the actual microbial populations responsible for metabolism of a specific compound, realistic substrate concentrations must be used in the incubations to avoid introducing a major bias in community structure due to selective growth of specific populations. In coastal wa- ters, which we are using as a model ecosystem, we found that even low additions of 14 C-labeled organic substrates (representing <3% of the total organic carbon) produced highly radiolabeled rRNA after 7 h incubation at in situ conditions. This incubation time is similar to the average generation time for the entire bacterial community (9 h), and both the uptake rate and the growth yield on the labeled substrates were linear during the incubation, suggesting that there were no major shifts in the microbial community. These tests also showed that the proportion of labeled C allocated to rRNA was strongly dependent on the quality of the substrate (e.g., 12%- 19% for adenine. 1.1%-1.3% for acetate). In addition, tests with exponentially growing bac- teria in pure culture showed that a constant fraction of the total cellular I4 C (average 8% for Vibrio cholera and 17% for E. coli) could be recovered in rRNA after about one cell doubling. Thus, the major advantages of rRNA detection are the linearity of the labeling process and the possible limi- tation to few cell doublings, which ensure that community structure will be only minimally biased. The third question, whether rRNA diversity can be as- certained with realistic effort, requires a reexamination of the PCR-based approach. We have recently presented the hypothesis that a simple, PCR-induced artifact may lead to severe overestimation of diversity of rRNA genes (Thomp- son et nl., 2002). During the co-amplification of homolo- gous templates with universal primers, a significant fraction (up to 50%) of products may be present as heteroduplices. These were increasingly prevalent as template diversity increased or primer availability became limiting (Thompson ft nl.. 2002). After cloning, heteroduplex molecules may FUNCTION AND DIVERSITY OF MICROBIAL COMMUNITIES 199 become subject to mismatch repair by the E. coli MutHLS system. This can theoretically lead to independent repair of each mismatched position since the repair system, in vivo, is directed by hemimethylation (Modrich. 1987). which is absent in PCR products. A model exploring the effects of heteroduplex repair demonstrated that the undirected repair process might be responsible for large overestimation of rRNA diversity (Thompson et ai, 2002). For example, a simple system of 2 sequences with 3 shared mismatched positions can result in 8 sequence permutations; for 4 se- quences with 10 shared mismatched positions, the number increases to 6136 (Thompson et al. 2002). Although this is a dramatic example, the potential contribution of heterodu- plex repair to sequence diversity can easily be avoided by "reconditioning PCR." a low-cycle-number reamplification of a 10-fold diluted, mixed-template PCR product. Although the exact contribution of heteroduplex repair to diversity estimates is still being analyzed in our laboratory, we have used the modified amplification protocol (recondi- tioning PCR) to estimate bacterial diversity in a coastal bacterial community. We generated a clone library from amplified 23S rRNA genes, then assessed sequence diver- sity in the library by a combination of rarefaction analysis and Chao-1 estimators, which are based on capture-recap- ture statistics (Hughes et ai. 2001). The results demon- strated that diversity was relatively moderate, with the num- ber of coexisting sequence types remaining in the low 100s (Acinas. Hunt. Bertilsson, and Polz. unpubl. data). This ongoing analysis is currently complemented with rarefac- tion of a 16S rRNA gene library derived from the same sample, demonstrating that it may indeed be possible to represent entire communities with reasonable effort on DNA microarrays. Acknowledgments This work was supported by grants from NSF and Sea Grant. We are also grateful to Mitch Sogin for providing access to the MBL sequencing facility, and to Byron Crump and John Hobbie for help during sampling. Literature Cited Beja, O., M. T. Suzuki, E. V. Koonin, L. Aravind, A. Hadd, L. P. Nguyen, R. Villacorta, M. Amjadi, C. Garrigues, S. B. Jovanovich. R. A. Keldman. and E. F. DeLong. 2000. Construction and analysis of bacterial artificial chromosome libraries from a marine microbial assemblage. Environ. Microbiol. 2: 516-529. Bertilsson, S. A., and M. F. Polz. 2001. Application of a diversity array to study specific substrate utilization in individual populations of het- erotrophic bacteria. P. 503 in Abstracts. American Society for Micro- biology. 101 st General Meeting. Orlando, FL. Boschker, H. T. S., S. C. Nold, P. Wellsbury, D. Bos, W. de Graaf, R. Pel, R. .1. Parkes, and T. E. Cappenberg. 1998. Direct linking of microbial populations to specific biogeochemical processes by ' 'C- labeling of biomarkers. Nature 392: 801-805. Cho, J.-C., and J. M. Tiedje. 2001. Bacterial species determination from DNA-DNA hybridization by using genome fragments and DNA microarrays. Appl. Environ. Microhinl. 67: 3677-3682. Cottrell. M. T.. and D. I.. Kirthmun. 2000. Natural assemblages ot marine proteobacteria and members ot the Cytophagft-Flavobacter cluster consuming low- and high-molecular-weight dissolved organic matter. Appl. Environ. Microhiol. 66: 1692-1697. Giovannoni, S.. and M. Rapue. 2000. Evolution, diversity, and molec- ular ecology of marine prokaryotes. Pp. 47-84 in Microbial Ecology o\ the Oceans. D. L. Kirchman, ed. Wiley-Liss, New York. Gray, N. D.. and I. M. Head. 2001. Linking genetic identity and function in communities of uncultured bacteria. Environ. Microbiol. 3: 481-492. Head, I. M., J. R. Saunders, and R. W. Pickup. 1998. Microbial evolution, diversity, and ecology: a decade of ribosomal RNA analysis of uncultivated microorganisms. Microb. Ecol. 35: 1-21. Hugenholtz, P., B. M. Goebel. and N. R. Pace. 1998. Impact of culture independent studies on the emerging phylogenetic view of bacterial diversity. / Bacteriol. 180: 4765-4774. Hughes, J. B.. J. J. Hellmann, T. H. Ricketts. and B. J. M. Bohannan. 2001. Counting the uncountable: statistical approaches to estimating microbial diversity. Appl. Environ. Micrabiol. 67: 4399-4406. Koizumi, Y., J. J. Kelly, T. Nakagawa, H. I'rakavva, S. EI-Fantroussi, S. AI-Muzaini, M. Fukui, Y. Urushigawa, and D. A. Stahl. 2002. Parallel characterization of anaerobic toluene- and ethylbenzene-de- grading microbial consortia by PCR-denaturing gradient gel electro- phoresis, RNA-DNA membrane hybridization, and DNA microarray technology. Appl. Environ. Microbiol. 68: 3215-3225. Lee. N.. P. H. Nielsen, K. H. Andreasen, S. Juretschko, J. L. Nielsen, K.-H. Schleifer, and M. Wagner. 1999. Combination of fluorescent in situ hybridization and microautoradiography a new tool for struc- ture-function analyses in microbial ecology. Appl. Environ. Microbiol. 65: 1289-1297. Modrich. P. 1987. DNA mismatch correction. Aniui. Rev. Biochem. 56: 435-466. Ouverney, C. C., and J. A. Fuhrman. 1999. Combined microautora- diography 16S rRNA probe technique for determination of radioiso- tope uptake by specific microbial cell types in situ. Appl. Environ. Microbiol. 65: 1746-1752. Radajewski, S., P. Ineson, N. R. Parekh, and J. C. Murrell. 2000. Stable-isotope probing as a tool in microbial ecology. Nature 403: 646-649. Schmidt, T. M., E. F. DeLong, and N. R. Pace. 1991. Analysis of marine picoplankton community by 16S rRNA gene cloning and se- quencing. J. Bactcriol. 173: 4371-437S. Small, J., D. R. Call, F. J. Brnckman, T. M. Straub, and D. P. Chandler. 2001. Direct detection of I6S rRNA in soil extracts by using oligonucleotide microarrays. Appl. Environ. Microbiol. 67: 4708-4716. Stokes, H. W., A. J. Holmes, B. S. Nield, M. P. Holley, K. M. H. Nevalainen, B. C. Mabbutt, and M. R. Gillings. 2001. Gene cas- sette PCR: sequence-independent recovery of entire genes from envi- ronmental DNA. Appl. Environ. Microbiol. 67: 5240-5246. Thompson, J. R., L. A. Marcelino. and M. F. Polz. 2002. Heterodu- plexes in mixed-template amplifications: formation, consequences and elimination by 'reconditioning PCR.' Nucleic Acids Res. 30: 2083- 2088. Torsvik, V., L. Ovreas, and T. F. Thingstad. 2002. Prokaryotic diver- sity: magnitude, dynamics, and controlling factors. Science 296: 1064- 1066. Reference: Bio/. Bull. 204: 200-204. (April 2003) 2003 Marine Biological Laboratory Human Oral Cavity as a Model for the Study of Genome-Genome Interactions JAMIE S. FOSTER, ROBERT J. PALMER, JR., AND PAUL E. KOLENBRANDER* Oral Infection and Immunity Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Building 30, Room 310, 30 Convent Drive MSC 4350. Bethesda, Man-land 20892-4350 The enormous diversity of culturable bacteria within the oral microbial community coupled with experimental acces- sibility renders the human oral cavity a valuable model to investigate genome-genome interactions. The complex in- teractions of oral bacteria result in the formation of biofilms on the surfaces of the oral cavity. One mechanism thought to be important in biofilm formation is the coaggregation of bacterial partners. In this paper, we examine the role of coaggregation in oral biofilms and develop protocols to elucidate the spatial organization of bacterial species re- tained within oral biofilms. To explore these issues, we have employed two experimental systems: the saliva-coated flow- cell and the retrievable enamel chip. From flowcell studies, we have determined that coaggregation can greatly influ- ence the ability of an oral bacterial species to grow and he retained within the developing hiofilm. To examine the spatial architecture of oral biofilms. fluorescent in situ hybridization protocols were developed that successfully target specific members of the oral microbial community. Together, these approaches provide insight into the devel- opment of oral biofilms and expand our understanding of genome-genome interactions. The human oral cavity contains more than 500 species of bacteria that interact among themselves and with their host tissues (Kroes et ai, 1999; Paster et ai. 2001). These complex interspecies associations result in the formation of * To whom correspondence should be addressed. E-mail: pkolenhrander@dir.nidcr.nih.gov The paper was originally presented at a workshop titled Outcomes of Genome-Genome Interactions, The workshop, which was held at the J. Erik Jonsson Center of the National Academy of Sciences. Woods Hole. Massachusetts, from 1-3 May 2002. was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266 microbial biofilms on the hard and soft tissues of the oral cavity (Gibbons and Hay, 1988; Hallberg et ai, 1998). Within these oral biofilms, numerous molecular and bio- chemical exchanges result in communication between dis- tinct genomes (here called genome-genome interactions). To explore the nature of genome-genome interactions in the oral cavity, it is necessary to understand the composition of the microbial community, the mechanisms by which oral bacteria associate, and the spatial arrangement of the com- munity. LIntil recently, description of plaque community composition relied on culture-dependent techniques (Moore and Moore, 1994; Socransky and Haffajee, 1994). Culture- independent methods have identified previously unknown and uncultured community members. These uncultured oral bacteria constitute a low percentage of the total bacterial numbers compared to the high percentage of uncultured bacteria in other natural environments (Kroes et ai. 1999; Paster et ai, 2001). Although these studies provide infor- mation on species composition, they do not address the spatial organization of the oral community. The juxtaposi- tion of different bacteria in three-dimensional oral biofilms such as dental plaque probably contributes to and may direct metabolic cross-feeding symbioses and transcriptional sig- nal exchange between organisms. Examining the precise architecture of oral biofilms may provide a clearer under- standing of the role each organism plays in the overall community structure and in genome-genome interactions. One mechanism through which oral bacteria may com- municate and facilitate genome-genome interactions is Co- aggregation. In coaggregation, oral bacterial cells bind to specific bacterial partners (Cisar et al.. 1979). To date, all oral bacteria tested coaggregate with at least one other bacterial species (Whittaker et al.. 1996; Andersen et ai. 1998; Kolenhrander et ai, 2002). Coaggregation occurs 200 ORAL MICROBIAL COMMUNITIES 201 between genetically distinct bacteria and is mediated by protein adhesins on one cell type that recognize comple- mentary carbohydrate receptors on the partner cell type. This phenomenon has been hypothesized to be essential to the formation of oral biofilms (Kolenbrander et ui. 2002), although little direct work has been performed on the con- sequences of coaggregation interactions. Several outcomes of these pairwise interactions are conceivable: (a) only one organism benefits, (b) one organism is detrimentally influ- enced, (c) both organisms benefit, (d) both organisms suffer, and (e) neither organism is influenced by the presence of the other. Although these scenarios are an oversimplification of what can occur in multispecies environments such as the oral cavity, they serve as starting points for assessing the consequences of coaggregation in vivo. Furthermore, it is becoming clear that coaggregation interactions exist outside the oral cavity in freshwater biofilms (Rickard et ul.. 2000, 2002). Therefore, the significance of contact-based cell-cell interactions in the bacterial world has probably been under- estimated, and the outcomes of these genome-genome in- teractions are likely to be a universal driving force in biofilm development. Several experimental designs have been developed to examine the formation of oral biofilms (Wilson, 1999; Wimpenny, 2000). These model systems often rely on the flow of nutrients over a surface on which bacteria are able to attach and grow. In our laboratory we use two experi- mental models, a saliva-coated flowcell (Kolenbrander et ai. 1999) and a retrievable enamel chip (Palmer et ai. 200 la). Each method has its own advantages for the exam- ination of oral microbial communities. The flowcell enables biofilms to form under defined conditions of species and nutrients. The basic design of a flowcell is a microscope slide and coverslip separated by a two-channel molded silicone gasket (Kolenbrander et a/., 1999). Inlet and outlet ports enable saliva, the sole nutrient source, to coat the glass surfaces with a salivary conditioning film of host proteins. After the salivary conditioning film is established, bacterial strains are injected into the flowcell chamber. As the biofilm forms within the flowcell, colonization and growth can be examined noninvasively by confocal laser microscopy (CLM). In comparison, the enamel chip facilitates the un- derstanding of natural biofilms that form within the human oral cavity. Enamel chips cut from human third molars are placed into two acrylic appliances worn intraorally by vol- unteers. The chips are then recovered and examined using microscopy (Palmer et al., 2001a). Initial studies on the outcomes of coaggregation interac- tions have been conducted in the flowcell model with three primary colonizers of the tooth surface: Streptococcus gor- donii DL1, Actinomyces naeslundii T14V, and Streptococ- cus oralis 34 (Palmer et al., 2001b). Each bacterium can coaggregate with the other two. The behavior of the strains as monocultures was assessed by examining their abilities to grow planktonically (as liquid cultures) in saliva, and to grow as biofilms in saliva. In planktonic culture, S. gordonii reproducibly reached a cell density of 10 7 cells per milliliter of saliva and was transferable (i.e.. growth was maintained over three transfers). A. naeslundii numbers consistently tapered off within 18 h after the initial transfer to saliva. S. oralis behaved inconsistently: growth occurred, but the maximum cell density varied between 10' and 10 h cells per milliliter of saliva, and cultures were not always transfer- able. These behaviors were duplicated in the flowcell sys- tem: monoculture biofilms of S. gordonii grew reproduc- ibly, those of A. naesliimlii never grew, and those of S. oralis grew only once in six experiments. Once behavior as monocultures was assessed, the out- come of pairwise interactions between the strains in bio- films was investigated (Palmer et al.. 2001b). The first strain was inoculated into the flowcell and allowed to adhere for 20 min; nonadherent cells were then washed out and the second strain was introduced and allowed to adhere for 20 min. After the subsequent washout of nonadherent cells, the coculture biofilm was examined immediately (time 0), after 4 h, and after overnight growth with flowing saliva. When the initial strain was S. gordonii. combination with either of the other two strains produced identical results: S. gordonii grew as it did in monoculture, and the partner strain (A. naeslundii or S. oralis) failed to grow (Fig. 1 A). Cells of the partner strain were retained within the S. gordonii biofilm. but biomass of either partner strain was clearly reduced over the course of the experiment. Thus, the growth of S. gor- donii was apparently unaffected by the presence of A. naeslundii or S. oralis. In marked contrast to these interac- tions, when 5. oralis and A. naeslundii were combined in a biofilm, both bacteria grew luxuriantly (Fig. IB). Growth as a coculture biofilm of these two organisms (neither of which could grow reproducibly as a monoculture biofilm) was much greater than that of S. gordonii. which grew indepen- dently under identical culture conditions. This is a clear example of a mutualism in which both organisms benefit from the interaction. Such interactions may be important in establishment of regional heterogeneity in oral biofilms in vivo, and we are currently using the retrievable enamel chip system to relate the results of our in vitro investigations to the situation in vivo. The study of bacterial symbiotic interactions described above was conducted by using antibodies to identify the organisms and give spatial information on the oral commu- nity. To complement these studies, we have begun to em- ploy fluorescence i';i situ hybridization (FISH) coupled with the flowcell and enamel chip models. The advantages of using FISH are that uncultured bacteria can be detected and that development of the probes is more rapid than produc- tion and characterization of antibodies. Fluorescently la- beled oligonucleotide probes designed to the 16S rRNA sequence of different oral bacteria were hybridized in situ 202 J. S. FOSTER ET AL S. gordonii (green) combined with A. naeslundii (red) timeO overnight B S. oralis (green) combined with A. naeslundii (red) time overnight 1. (_'i.ullure biotilms showing multiple outcomes of coaggregation. (A) Strcptni-ncciix xnnlonii interaction with .4r//mwvco iuii:\lundii. S. gonlimii was introduced first, followed by A. nucslnnilii. Biotilnis were examined with contocal microscopy at time (((immediately after washout of nonadherent cells, left panels) and after overnight growth on saliva (right panels). Lower panels of each vertical pair are 3x zooms of the center of the corresponding upper panels. S. gun/onii (green) was detected by constitutive green fluorescent protein fluorescence; ,4. ncicslnntlii (red) was detected by secondary immunofluorescence. (B) iV/v/voriKvi/.s oralis interaction with A. mieslinulii. Details as above, except that S. oralis (green) was detected b\ primary immunofluorescence. All scales. 25 fi\n. ORAL M1CROBIAL COMMUNITIES 203 Figure 2. Confocal micrographs of oral biofilms examined with fluorescence in xiin hybridization. (A) Mixed-species biofilm containing Streptococcus gonJonii DL1, Actinomyces naeslundii PKl(a), and Fiisohnc- terium inicleatiim PKI594(f) grown on saliva in a flowed] for 4 h. The biofilm was stained with an FITC-labeled oligonucleotide probe (green) targeted to 5. gordonii and counterstained with the nucleic acid stain Syto 59 (red ). Colocalized stains are yellow. Scale. 25 ju.m. (B, C) Monospecies biofilms inoculated with S. f-urd<>iiii DLI and grown on enamel chips for 4 h. Biofilms were stained with an oligonucleotide probe specific for S. gordonii and the nucleic acid stain DAPI. Scale, 10 /j,m. (B) High-magnification image of enamel chip showing total number of cells visualized by DAPI stain. (C) Same location on enamel chip as in B. but only detecting cells labeled with oligonucleotide probe. with the growing biofilm. thus enabling bacterial species to be located without biofilm disruption (Fig. 2). Flowcells were consecutively inoculated with cultures of S. gordonii DLI, A. nticslundii PK19 (each an early colonizer), and Fusobacterium nucleunini PK1594 (late colonizer). After 4 h of saliva flow, biofilms were probed with a fluorescently labeled oligonucleotide designed to target streptococci (5'- GCTGCCTCCCGTAGGAGT-3'; JF20) as well as with a general nucleic acid stain to detect all cells (Fig. 2A). Based on distinctive morphologies of S. gordonii (coccus shaped). A. iKicsInndii (rod shaped), and F. nuclecituin (slender rods with tapered ends), all cell types could be visualized within the biofilm by staining with the nucleic acid marker Syto 59 (Molecular Probes, Eugene, OR). However, labeling by the fluorescent oligonucleotide probe was visible only in S. gordonii cells; the other two organisms did not bind the streptococcal probe (JF20) (Fig. 2A). RNA levels in \itii are frequently very low, and therefore detection can be prob- lematic. To test the detection efficiency of FISH probes compared with general nucleic acid stains, monospecies biofilms containing 5. gordonii were grown in saliva in vitro on enamel chip surfaces and exposed to a streptococcal- specific probe (Fig. 2B, C). The biofilms were also stained with the nucleic acid stain diamidino-2-phenylindole dihy- drochloride (DAPI) to fluorescently mark all cells within the biofilm. All of the S. gordonii cells could be stained with the oligonucleotide probe without high background fluores- cence (Fig. 2C). Taken together, these results suggest that our FISH protocols can be employed in conjunction with the flowcell and enamel chip systems. This work represents just a few of the approaches used to study oral bacterial interactions. Despite the complexity and diversity of organisms present within the human oral cavity, experimental model systems such as these can be used to explore the consequences of the interactions between dis- tinct genomes and to elucidate common underlying mech- anisms of communication within microbial biofilms. Acknowledgments The authors thank Paul Egland and David Blehert for their useful comments on this manuscript. Literature Cited Andersen, R. N., N. Ganeshkumar. and P. E. Kolenbrander. 1998. Hvlicobacter pvlori adheres selectively to Fusobacterium spp. Oral Microhiol. [mmunol, 13: 51-54. Cisar, J. O., P. E. Kolenbrander, and F. C. Mclntire. 1979. Specificity of coaggregiition reactions between human oral streptocci and strains of Actinomvces viscosux or Actinoiiiycc.t nai'.\liindii. Infect. Iininun. 24: 742-752. Gibbons, R. J., and I). I. Hay. 1988. Absorbed salivary proline-rich proteins as bacterial receptors on apatitic surfaces. Pp. 143-163 in Molecular Mechanisms of Microbial Adhesion, L. Switalski, M. Hook, and E. Beachey, eds. Springer-Verlag, New York. Hallberg, K., K. J. Hammarstrom. E. Falsen, G. Uahlen, R. J. Gibbons, D. I. Hay, and N. Stromberg. 1998. Actinomyces naeslundii geno- species 1 and 2 express different binding specificities to N-acetyl-beta- D-galactosamine. whereas Actiimiinrc* odontolyticus expresses a dif- ferent binding specificity in colonizing the human mouth. Oral Microbial. Immunol. 13: 327-336. Kolenbrander, P. E., R. N. Andersen, K. Kazmerzak, R. Wu, and R. J. Palmer, Jr. 1999. Spatial organization of oral bacteria in biofilms. Methods Enrynwl. 311): 322-332. Kolenbrander, P. E., R. N. Andersen, 1). S. Blehert, P. G. Egland, J. S. Foster, and R. J. Palmer, Jr. 2(102. Communication among oral bacteria. Microhiol. Mol. / 10). Our research ranges from environments like the warm (45 C), acidic (pH 2.7) Nymph Creek in Yellowstone National Park to tem- perate alkaline lakes in the Sandhills region of western Nebraska. The focus of this report is the acidic, heavy- metal-rich Rio Tinto in southwestern Spain. The Rio Tinto flows 100 km through the world's largest pyritic (FeS->) belt. The river gets its red color from the high levels of iron dissolved in its acidic waters (pH 2.0). Ferric hydroxide t t ii.\ alhitlus Rhudolorula Spi~elloni\ces acitmtnuJits -Morlierellti polycephala Gigaspora margdrila RTSiin3 Scvpha ciliala Mnemiopsis leid\i Diaphanoeca granais ' Ichthyophonus hoferi RTSiinW RT5iinl6 Nucleariasp. Ascomycota Basidiomycota Chytridiomycota Zygomycota Gracilaria sp. Chondrus crispus RTlnl4cul RTlnl Chlamyilnmonas noclieama Unknown Chlamvdomonas reinhardtn Volvox carleri RT5iin2 RT5iin49 Chlorella minulissima Nanochlorum eitcarvotum RTSiinlO RTSiinS RT5in45 Z\gnerna circumcannata Cylindrocystis brebissonii Klebsormidium flacci ditm Zea mavs " Orvza saliva Acantnamoeba caslellann - Thaumatqmonas sp. Heteromila slobosa RTSiinl9 RTSiinlO Pauline/la chromatophora -Euelvnha rotunda mdentified eukaryote (UEU 1 30856) RT3nl9 Cercomonas ATCC 50316 Massisreria marina RTSiin4 RT5iin35 RT5in4 RT5in36 -Polerioochromonas malhainensis - Ochromonas danica -Hibberdia magna -RTln9 .. Fungi Animals Red Algae Green Algae Land Plants Testate Pilose Amoebae Cercomonads RT7in48 Eolimna minima Pseuao-nitzschia pungens Unknown Cafeteria roenbergensis Proteromonas lacertae _ RT5iin25 Lab\rinthuloides minula ,-, , , Glaucoma chaltom m< rColpidium campylum Vorticella microstoma Opislhonecta hennegu\i Stramenopiles O.\\tricha graniilifera RT7iinl Toxoplasma gondii Sarcocystis muris Alexandrium lamarense Prorocenlrum micans Alveolates Dictyostelium discoideum Gephyramoeba sp. Physarum polycephalum RTSiin21 RT5iin44 Filamoeba nolandi Masligamoeba (Phreatamoeba) balamuthi Entamoeba histolytica . Paravahlkamnfia usliana 5in38 RTSi Naegleria gruberi Euglena mutabilis RT8n7cul Euglena gracilis . 0.05 substitutions/site 21 IS AMARAL ZETTLER ET AL. not exhaustive, we surmise that there are still more undis- covered novel lineages in the river. Despite our growing knowledge of the Tinto's eukaryotic diversity, we know little about the role eukaryotes play in shaping the varied ecosystems that occur along the river. For example, we do not know if these biofilm communities have microenvironments that enhance survival of their members. Could fungal metal sequestration protect nontol- erant species'? Furthermore, we know little about how these organisms have evolved adaptations to extreme concentra- tions of acid and metals. To explore these questions, we have been isolating or- ganisms from the river for e.\ situ physiological experi- ments. We have established monocultures of Chlamydomo- nas sp., Euglena cf. mutabilis, Chlorella sp., and Vannella sp. isolated from enrichments of river water and are cur- rently exploring the physiology of these protists from ex- treme environments. We have initiated our physiological studies on an acido- philic species of a chlamydomonad alga isolated from the river Chlamydomonas sp. Our first question about the physiology of the Tinto acidophiles was the nature of the cytosolic pH (pH,). There are published reports of acido- philes from all domains of life with internal pH values that deviate from neutral these include the archaebacterium Picniphilus oshimac, pH, = 4.6 (van de Vossenberg et ai. 1998); the eubacterium Bacillus acidocaldarius, pH; = 5.6- 5.8 (Thomas et al., 1976); and the eukaryotic alga Euglena wntabilis. pH, = 5.0-6.4 (Lane and Burris. 1981). Using the fluorescent H + indicator BCECF, we determined that our acidophilic chlamydomonad isolate maintains an aver- age internal pH of 6.6 at an external pH of 1 (M. A. Messerli, L. A. Amaral Zettler. S.-K. Jung. P. J. S. Smith, and M. L. Sogin. unpubl.). Our other isolates await similar measurements. Given that there is a 40,000-fold difference in hydrogen ion activity between the inside and the outside of these cells, we propose the existence of active transport mechanisms that help these organisms regulate their internal pH. We hypothesize that novel diversity in H + -ATPases may ex- plain the ability of different protist species to thrive in the low pH, high-metal Rio Tinto environment. There are two major families of H + -ATPases: the V/F/A-ATPases and the P-type-ATPases. The V-type ATPases can occur in the plasma membrane of eukaryotes (but are more commonly associated with vacuolar membranes) and consist of at least 1 1 subunits and a molecular mass approaching 10 6 Da. In contrast, eukaryotic P-type ATPases consist of either mo- nosubunits (as with H + -ATPases) or a hetero-subunit (alpha and beta, as found in the Na '/K ' -ATPases and H + /K + - ATPases); have a molecular weight of about 100 kDa; and form a phosphorylated intermediate during the course of ATP hydrolysis. Indirect evidence of novel ATPases comes from studies of the protozoan parasite Leishnuinia dono- nini. which has the ability to switch between living in a neutral environment, pH 7.5. as a promastigote (flagellated stage) and in an acidic environment, pH 5.0, as an amasti- gote (nonrlagellated stage) (Meade et al., 1989). The plasma membrane of this organism contains a P-type ATPase that has two isoforms with slightly different sequences. Isoform la is expressed in both promastigotes and amastigotes, whereas isoform Ib is expressed more abundantly in the amastigotes (Meade et al.. 1989). This difference suggests the use of a sequence change to accommodate the acidic condition. Modifications to ion regulatory machinery might be reflected by convergent amino acid substitution patterns or by accelerated rates of change in acidophilic protist lineages, as revealed in phylogenetic analyses. For example, portions of membrane-bound V- and P-type ATPases that are exposed to the acidic external environment may display different amino acid substitution patterns than do domains that face the cytoplasm. We are currently using degenerate primers designed against two conserved regions, the phosphorylation site and the ATP-binding site, to amplify members of the P-type superfamily of ion transporters. Thus far, all of our clones fall into the heavy-metal P-type class but may represent different metal transporters. We have found more diverse sequences in the acidophilic Chlamydomonas than in the neutrophilic C. reinhardtii. We are screening additional clones for H + -transporting ATPases. Once we obtain ion-transporter sequence information from these acidophiles, we will focus on correlating the expression of these transporters in space and time to bio- geochemical characteristics in the river. This will bring us beyond the study of biodiversity in the river to questions at the heart of potential genomic interactions between mem- bers of the microbial consortia. With this kind of approach, we may also be able to determine whether symbiotic inter- actions are occurring in this environment. Acknowledgments This work was supported by the National Science Foun- dation's Lexen Program DEB-0085486. the NASA Astro- biology Program NCC2-1054, and an NIH:NCRR 01395. The authors wish to acknowledge the support of the Amils lab at the Autonomous University of Madrid and the tech- nical assistance of Brendan Keenan and Erik Zettler. Literature Cited Amaral Zettler. I,. A., F. Gomez, E. R. Zettler, B. G. Keenan, R. Amils, and M. L. Soj-in. 21)02. Eukaryotic diversity in Spain's River of Fire. Nuiiiri- 417: 137. Duran, C"., I. Marin, and R. Amils. 1999. Specific metal sequestering acidophilic fungi. Pp. 521-530 in Biohydrometallurgy and the Envi- i-HiiiiH'iit T\mrilx the Mining of the 21st Century. Proceedings of the International Biohydrometallurgy Symposium, IBS '99, R. Amils and A. Ballester, eds. Elsevier, Amsterdam. Hl'YONI) WnniVI RSII V IN RIO TIN K ) 209 Gonzalez-Toril, E., F. Gomez, N. Rodriguez, D. Fernandez-Remolar, J. Zuluaga, I. Marin, and R. Amils. 2001. Geomicrobiology of the Tinto River, a model of interest for biohydrometallurgy. Pp. 639-650 in Biohydrometallurgy: Fundamentals, Technology, and Sustainable Development, V. S. T. Ciminelli and O. Garcia, eds. Elsevier, Amster- dam. Lane, A. E., and J. E. Burris. 1981. Effects of environmental pH on the internal pH of Chlorella pyrenoidosa. Scenedesmus quadricauda, and Euglena mutabilis. Plant Physiol. 68: 439-442. Lehlanc, M., J. A. Morales, J. Borrego, and F. Elbaz-Poulichet. 20(10. 4,500 year-old mining pollution in southwestern Spain: long-term implications for modern mining pollution. Econ. Ceol. 95: 655-662. Lopez-Archilla, A. I., and R. Amils. 1999. A comparative ecological study of two acidic rivers in southwestern Spain. Microb. Ecol. 38: 146-156. Lopez-Archilla, A. I., I. Marin, and R. Amils. 1993. Bioleaching and interrelated ucidophilic microorganisms from Rio Tinto, Spain. Geomi- crohiol J. 11: 223-233. Lopez-Archilla, A. I., D. Moreira, 1. Marin, and R. Amils. 1994. El rio Tinto. un curso de agua vivo pero con mala fama. Quercas (Septem- ber): 19-22. Lopez-Archilla, A. I., I. Marin, and R. Amils. 1995. Microbial ecology of an acidic river: biotechnological applications. Pp. 63-73 in Biohy- tlrometallnrgical Processing, C. A. Jerez. T. Vargas. H. Toledo, and J. W. Wiertz, eds. University of Chile, Santiago. Lopez-Archilla, A. I., 1. Marin, and R. Amils. 2001. Microbial com- munity composition and ecology of an acidic aquatic environment: the Tinto River. Spain. Microb. Ecol. 41: 20-35. Meade, J. C., K. M. Hudson, S. L. Stringer, and J. R. Stringer. 1989. A tandem pair of Leishmania donovani cation transporting ATPase genes encode isoforms that are differentially expressed. Mol. Biochem. Parasitol. 33: 81-91. Porter, K. G., E. B. Sherr, B. F. Sherr, M. Pace, and R. W. Sanders. 1985. Protozoa in planktonic food webs. ./. Protozool. 32: 409-415. Thomas, J. A., R. E. Cole, and T. A. Langworthy. 1976. Intracellular pH measurements with a spectroscopic probe generated in situ. Fed. Proc. 35: 1455 van de Vossenberg, J. L., A. J. M. Drissen, W. Zillig, and W. N. Konings. 1998. Bioenergetics and cytoplasmic membrane stability of the extremely acidophilic. thermophilic Archaeon Picrophilus oshimae. Extremophiles 2: 67-74. Reference: Biol. Bull. 204: 210-214. (April 2003) 2003 Marine Biological Laboratory Isolation of Symbiotically Expressed Genes From the Dinoflagellate Symbiont of the Solitary Radiolarian Thalassicolla nucleata REBECCA J. CAST 1 '*, DAVID J. BEAUDOIN 2 . AND DAVID A. CARON' 1 Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543; 2 Marine Biological Laboratory, Woods Hole, Massachusetts 02543; and ^University of Southern California, Los Angeles, California Symbiotic associations are fundamental to the survival of many organisms on Earth. The ability of the symhiont to perform ke\ biochemical functions often allows the host to occupy environments that it would otherwise find inhos- pitable. This can have profound impacts upon the diversi- fication and distribution of the host. Cellular organelles (chloroplasts and mitochondria} represent the fined stages of integration of endosymbionts. These organelles were of critical importance to the evolution and success of eukary- otic lineages on our planet because they allowed the host cells to harness light energy and to thrive in the presence of oxvgen. The marine photosymbiotic associations that we study represent an earlier stage in the process of symbiont integration one in wliich the photobiont can still be re- moved from the host and exist on its own. These systems are of interest to us for two reasons. First, they are ecologically important in the marine environment where they occur. These organisms form zones of photosynthetic production in oceanic regions typically low in nutrients. Second, investi- gation of these interactions may shed light on the molecular and evolutionary mechanisms involved in the integration of cells and their genomes. Associations between microbial symbionts and their hosts are fundamental trophic relationships that occur in * To whom correspondence should be addressed. E-mail: rgast@whoi.edu The paper was originally presented at a workshop titled Outcomes of Genome-Genome Interactions. The workshop, which was held at the J. Erik Jonsson Center of the National Academy of Sciences, Woods Hole, Massachusetts, from 1-3 May 2002, was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266. both terrestrial and aquatic environments and depend upon regulated genomic communication between the interacting organisms. We are particularly interested in the photosym- biotic relationships that involve microscopic (protistan) al- gae in association with a variety of "invertebrate" hosts (protistan and metazoan); these relationships are fairly com- mon in the marine environment and contribute significantly to nutrient flow in the nutrient-poor regions of the ocean where they occur. The symbiosis is essential to the survival of the host organisms, although the algal symbiont can exist in the free-living state. Besides the conspicuous invertebrate photosymbioses that most people are familiar with (anemones and corals), a variety of protistan species also harbor cytoplasmic algal symbionts. Most notably, protozoa within the subphylurn Sarcomastigophora (the foraminifera, radiolaria, and acan- tharia; collectively referred to as "sarcodine protozoa") exist in symbioses with a wide array of algae (Anderson, 1983; Lee and McEnery, 1983; Caron and Swanberg, 1990; Michaels, 1991). These sarcodine protozoa are often large for single-celled organisms. Solitary cells often attain sizes more than 1 mm in diameter (some benthic foraminifera form discs more than 1 cm in diameter), and the presence of complex pseudopodial networks can increase the effective size of these organisms up to a few centimeters (Fig. 1). Some colonial species of radiolaria even form gelatinous pseudopodial matrices several centimeters in diameter and more than I m in length. These organisms are important contributors to the biological assemblages in all tropical and subtropical oceans of the world. Sarcodines form conspic- uous zooplankton assemblages in oligotrophic surface wa- ters, where they contribute significantly to primary produc- tivity (by virtue of their intracellular symbionts), predation. 210 SYMBIOTICALLY EXPRESSED GENES IN DINOFLAGELLATE SYMBIONTS Figure 1. Solitary planktonic radiolurian Tlicilu.ixicolUi niiclcutu. The central capsule of the organism is the dark circle in the center; it is surrounded by the rhizopodia containing the symbionts (small dots around the central capsule). The entire organism is about a millimeter in diameter. and the vertical flux of organic material and crystal miner- als, such as calcium carbonate and strontium sulfate (Caron and Swanberg. 1990; Swanberg and Caron. 1991; Caron et Hi, 1995; Michaels et al.. 1995). Our interest in the regulation of photosymbiotic relation- ships arose while working on the identification of the sym- bionts in planktonic foraminifera and radiolaria to deter- mine whether co-evolution of host and symbiotic alga occurred (Cast and Caron, 1996; Gast et al., 2000). We molecularly identified several taxonomically different algae that form symbioses in the planktonic sarcodines. Dinoflagellates are the most common of these, but we also found prymnesiophyte and prasinophyte algae. Further- more, it appears that most algal symbionts are actually genetically more similar to non-symbiotic algae than they are to other symbiotic algal lineages. This led us to wonder about what made very different algae suitable as symbionts, sometimes in the same host (Gast and Caron, 2001 ), and to hypothesize that symbiotic algae have something in com- mon, which can be identified genetically, that makes them acceptable as symbionts. We have undertaken comparative studies of gene expression in a symbiotic alga to investigate the mechanisms of symbiotic communication. We are studying the symbiont-host relationship between the dinoflagellate alga Scrippsiella niitricula and the solitary planktonic radiolarian Thalassicolla nucleata. The radiolar- ian system allows us to compare the genes expressed in the symbiotic state of the algal symbiont with those expressed in the free-living state of the alga in culture. The informa- tion that we obtain from this system can be used to identify the genes necessary for maintaining the interaction, deter- mine the timing of gene expression, examine the genomic characteristics associated with symbiotic-competence (why one alga is a good symbiont and another is not), and examine the mechanisms of secondary and higher level endosymbiotic events. We chose to begin our work of isolating algal genes expressed in the symbiotic state by studying the dinoflagel- late symbiont (Scrippsiella niitriciila) of the solitary radio- larian T. nucleata (Fig. 1). To identify genes being ex- pressed in the symbiosis, we required RNA from both the free-living and symbiotic states of the alga. We currently have the different algal symbionts as free-living laboratory cultures, but the intact association must be collected from the environment. T. nucleata generally harbors thousands of symbiotic algae, it can be collected from the environment fairly easily, and, most importantly, the host and symbiont can be separated from each other with minimal host RNA contamination of the sample. The host nucleus, organelles. and ribosomes are separated from the symbionts by the central capsule (pp. 90-94, Anderson, 1983). We believe that very little host mRNA will be present in the extracap- sular material, and we are careful not to destroy the central capsule when stripping away the desired extracapsular ma- trix. Thalassicolla nucleata was collected by plankton net tows in the Sargasso Sea and the San Pedro Channel. Individuals were transferred to sterile seawater in multiwell culture dishes and placed in lighted incubators or on work- benches at ambient temperatures. This allowed the hosts to clear themselves of prey items prior to recovery of the symbiont. All hosts were examined for prey before the extracapsular material was stripped away from the central capsule. Studies involving host incorporation of I4 C through the uptake of symbiont-produced photosynthate indicated that the interaction continued to function normally when the host was held in the laboratory (Anderson. 1978; Gast. unpubl. results). We believe that the symbiotic pattern of gene expression is also maintained. Stripped extracapsular material was immediately placed in RNALater (Ambion) until processed for RNA extraction (RN Aqueous, Ambion). At least 200 individual hosts were collected and processed for each subtraction analysis. Due to the relatively small amount of symbiotic-state RNA available for our project, we used a subtraction method that would permit the analysis of very small amounts of starting material. Through the use of two PCR enrichment amplifications, suppression subtractive hybrid- ization (SSH; Diatchenko et al.. 1996) allows as little as 25 ng of mRNA (or 50 ng of total RNA) to be analyzed. We used the Clontech PCR-Select cDNA Subtraction Kit to analyze both total RNA and mRNA from the dinoflagellate symbionts (Fig. 2). To date, we have completed five sub- tractions one from total RNA that failed, three from total RNA that were successful but yielded mostly host and bacterial ribosomal genes, and one from mRNA that was successful. We switched to analyzing mRNA because we recovered as many ribosomal clones as symbiosis clones when we analyzed total RNA. Contaminating ribosomal fragments were not unexpected since they would not be 212 R J. CAST ET AL B Figure 2. Results from suppression subtractive hybridization. (A) cDNA synthesis from symbiotic (lanes 1 & 2). free-living (lanes 3 & 4). and control (lane 5) RNA. M = 1-kh ladder. (B) Final PCR products from subtracted samples. Lane 1 symbiotic, subtracted: lane 2 symbiotic unsubtracted; lane 3 free-living sub- tracted; lane 4 free-living unsubtracted; lane 5 control subtracted; lane 6 control unsubtracted. M = l-kb ladder. cDNAs were generated and subtracted by using the PCR-Select cDNA Subtraction kit from Clontech. Total RNA was isolated from the algae in the symbiotic state and the free-living state using either the Ambion RNAEasy kit or the Epicentre MasterPure RNA Purification kit. and mRNA was recovered using the Qiagen Oligotex mRNA kit. In both instances, the RNA from each state was separately reverse transcribed using MMLV-reverse transcriptase (Superscript II) to generate double-stranded cDNAs. Both sets of cDNAs were restriction digested with Rsa\ to generate fragments with blunt ends, and the digested cDNAs from the symbiotic state were divided into two aliquots. Each aliquot was ligated to one of two different adaptors supplied in the kit. The strategy of the forward subtraction is to identify cDNAs up-regulated or unique to the symbiotic state of the alga. In this process, an excess of cDNAs from the free-living state (without ligated adapters) was mixed in two separate reactions with the adaptor-ligated cDNAs from the symbiotic state. The two tubes were separately heated to 95C to denature the cDNAs, and then each was allowed to hybridize at 68C for 8 h. This step promotes hybridization between the cDNAs expressed in common in the symbiotic and free-living states. To further enhance the subtraction of common cDNAs, the contents of the two tubes were then mixed together without a second denaturation step, combined with an additional excess of heat-denatured cDNAs from the free-living state, and hybridized overnight at 68C. PCR primers specific to the two adapters (Clontech) were used to amplify cDNAs created during the subtraction that possessed one DNA strand ligated to adapter 1 and the other DNA strand ligated to adapter 2R. Only cDNAs from the symbiotic state of the alga should be able to amplify in this step, so we are enriching for the differentially regulated transcripts. The subtracted and amplified cDNAs were cloned into pGEM-T (Promega) and used to transform JM109 competent cells (Pro- mega). A reverse subtraction was also performed in which the adaptor-free cDNA was from the symbiotic state and the adaptor-ligated cDNAs had been isolated from the free-living state. These subtracted cDNAs are therefore enriched in genes specific to the free-living state. removed by subtraction with cultured dinoflagellate RNA. We identified 23 clones as being potentially up-regulated in the symbiotic state of the alga. Nucleotide Blast searches (Altschul et /., 1997) were conducted for sequences from each of the putatively up- regulated clones that were isolated. Of the 23 selected clones, 8 turned out to be ribosomal genes from the host, bacteria, or fungi. Twelve of the remaining clones (C4 group) were very similar to each other, varying primarily by small differences in length. These clones had no significant sequence match in the database (Table 1 ). Of the remaining three clones, only F7 showed significant homology to data- base sequences. The homology between F7 and Cyplasin S is the result of three 30-base repeats in the clone sequence that are also present in the gene. This 30-base repeat seems to occur in other proteins as well, and may represent a shared structural region. There are generally very few algal or protist cDNA sequences in GenBank, so the overall lack of homology to sequences in the database is disappointing but not unexpected. We are continuing to screen clones from the final PCR products of our mRNA subtraction to obtain more genes unique to the symbiotic state of the alga. Because of the small amount of RNA available from our symbiotic state, it was not possible to confirm differential gene expression using northern blot. Instead, we used cDNA generated from the free-living and symbiotic state RNAs in a dot blot format (Fig. 3). Clone F9 was included as a control for positive hybridization to the free-living SYMBIOTICALLY EXPRESSED GENES IN DINOFLAGELLATE SYMBIONTS Table 1 List of clones recovered and analyzed with suppression subtractive hybridization 213 Clone Number of Size of group BLASTn results E value clones fragment (bpl C4 Human DNA clone RP11-100A16 0.049 12 278 F7 Aplysia punctata mRNA for Cyplasin S le-60 1 161 E7 Klebsiella pneumoniae SL032 plasmid 0.003 1 635 B8 Human chromosome 16 clone RP1 1-407623 0.095 1 367 0.5 ug free- living cDNA 1 ug free- living cDNA 1 ug A3 plasmid DNA lugE7 plasmid DNA ft ft ft .0.5 ug symbiont cDNA ug symbiont cDNA lugBS plasmid DNA ft ft PROBE B8 C4 E7 F7 F9 Figure 3. Confirmation of clone expression in symbiotic cDNA. To determine if the subtracted library contained clones up-regulated in the symbiotic state, 96 colonies were randomly chosen, and the cDNA insert from each was PCR amplified with the adapter-specific primers. Amplified products were denatured by addition of an equal volume of 0.6 N NaOH and then spotted onto Hybond NX nylon membranes and baked at 80C for 1 h. The membrane was then hybridized to either a forward-subtracted or reverse-subtracted cDNA probe. Probes were generated by incorporating biotin-labeled nucleotides into either the secondary PCR products or the original cDNAs. Membranes were prehybridized for 1 h at 65 or 72C in hybridization buffer (5XSSC, 0.1% N-lauroylsarcosine, 0.02% SDS, 1% blocking powder). The denatured probe was added to the prehybridization and allowed to hybridize overnight at 65 or 72C. Membranes were washed twice, for 5 min each, in 2XSSC, 1 .0% SDS at room temperature. The next two washes, 15 min each, were in 2XSSC, 0.1% SDS at the hybridi/ution temperature. Hybridization was detected using the CDP-Star Southern Star chemiluminescent method. Only those inserts that showed hybridization to the forward-subtracted probe but not to the reverse-subtracted probe were examined further. To confirm that these clones were differentially regu- lated, inserts from these positive clones were PCR amplified and digested with Rsa\. Sma\, and Eagl to remove adaptors. Samples (0.5 and 1 mgl of the unsubtracted cDNAs from the symbiotic and free-living states, along with the plasmid clones A3, E7 and B8, were denatured, spotted onto membranes, and probed with the labeled inserts. The results for the symbolically up-regulated clones (B8, C4, E7, and F7) are shown, along with one clone that is down-regulated in the symbiotic state (F9|. cDNA. This and other clones represent potentially down- regulated genes in the symbiotic state of the alga. We have not begun to analyze these clone sequences, but we recog- nize that these are potentially informative as well. The dot blots allowed us to confirm that our differentially expressed clones were not expressed, or were expressed at a reduced level, in the free-living cDNA populations. Suppression subtractive hybridization analysis generates fairly small cDNA fragments, so rapid amplification of cDNA ends (RACE) analyses must be used to obtain the full cDNA sequence. We began our RACE analysis with the B8 clone because we knew that we had the 3' end of the transcript. Currently we have about 800 basepairs of se- quence for this clone. It has two open reading frames that are interrupted by a single stop codon in each. Neither of these reading frames has a start codon, and because the few dinoflagellate cDNAs that have been examined indicate the presence of fairly long 3' untranslated regions (300-500 basepairs), we may still not have reached the 5' end of the transcript. New primers are being designed to try and re- cover more of the 5' end. Our other clones are less defined as to whether they are at the 5' or 3' end of the cDNA, and we are currently extending from both ends. In this decidedly ambitious project, we have been suc- cessful in recovering several genes that appear to be up- regulated in the symbiotic state of the dinoflagellate sym- biont S. nutricula. We know of only one other study using SSH in a non-mammalian context. That study, on the sexual stages of diatoms (Armbrust, 1999), confirmed only 10 up-regulated clones, so our low number of recovered clones may not be unusual, especially since our RNA quantities were so limiting. We plan to acquire the full-length cDNA sequences for the clones we have; to pursue the construction of a cDNA library for the dinoflagellate symbiotic-state RNA; and to develop microarrays to aid in screening for differentially expressed genes. Acknowledgments We thank the anonymous reviewers of this manuscript for their helpful comments. This work was supported by the NASA Astrobiology Program through grant number NCC 2-1054 to the MBL Astrobiology Center. WHOI contribu- tion number 10815. 214 R. J. CAST ET AL. Literature Cited Altschul. S. F., T. L. Madden, A. A. Schaffer. J. Zhang, Z. Zhang, VV. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389-3402. Anderson, O. R. 1978. Fine structure of a symbiont-bearing colonial radiolarian. Collosphaera globularis and 14C isotopic evidence for assimilation of organic substances from its zooxanthellae. J. Ultra- struct. Res. 62: 181-189. Anderson, O. R. 1983. Radiolaria. Springer- Verlag, New York. Armbrust, E. V. 1999. Identification of a new gene family expressed during the onset of sexual reproduction in the centric diatom TluiUis- siosira weissflogii. Appl. Environ. Microbiol. 65: 3121-3128. Caron. D. A., and N. R. Swanberg. 1990. The ecology of planktonic sarcodines. Rev. Aqitat. Sci. 3: 147-180. Caron. D. A., A. F. Michaels. N. R. Swanberg, and F. A. Howse. 1995. Primary productivity by symbiont-bearing planktonic sarcodines (Acantharia, Radiolaria, Foraminifera) in surface waters near Bermuda. J. Plankton Res. 17: 103-129. Diatchenko, L., C. Y.-F. Lau, A. P. Campbell, A. Chenchik, F. Mo- qadam. B. Huang, S. Lukyanov, K. Lukyanov, N. Gurskaya, E. D. Sverdlov, and P. D. Siebert. 1996. Suppression subtractive hybrid- ization: A method for generating differentially regulated or tissue- specilic cDNA probes and libraries. Pnn: Null. Actiil. Sci. 93: 6025- 6030. Cast, R. J., and D. A. Caron. 1996. Molecular phylogeny of symbiotic dinoflagellates from Foraminifera and Radiolana. Mul. Bin/. Enil. 13: 1192-1197. Cast, R. J., and D. A. Caron. 2001. Photosymbiotic associations in planktonic foraminifera and radiolaria. Hvdrobiologiu 461: 1-7. Cast, R. J., T. A. McDonnell, and D. A. Caron. 2000. srDNA based taxonomic affinities of algal symbionts from a planktonic foraminifer and a solitary radiolarian. J. Plivcol. 36: 172-177. Lee, J. J., and M. E. McEnery. 1983. Symbiosis in foraminifera. Pp. 37-68 in Algal Symbiosis. E. L. Goff, ed. Cambridge University Press. Cambridge. Michaels, A. F. 1991. Acanthanan abundance and symbiont productivity at the VERTEX seasonal station. J. Plankton Res. 13: 399-418. Michaels, A. F., D. A. Caron, N. R. Swanberg, F. A. Howse, and C. M. Michaels. 1995. Planktonic sarcodines (Acantharia. Radiolaria. Fo- raminifera) in surface waters near Bermuda: abundance, biomass and vertical flux. J. Plankton Res. 17: 131-163. Suanherg, N. R., and D. A. Caron. 1991. Patterns of sarcodine feeding in epipelagic oceanic plankton. J. Plankton Res. 13: 287-312. Reference: Biol. Bull. 204: 215-220. (April 2003) 2003 Marine Biological Laboratory Plants, Mycorrhizal Fungi and Endobacteria: a Dialog Among Cells and Genomes P. BONFANTE Dipartimento di Biologia Vegetate dell' Universita di Torino and Istituto di Protezione delle Piante, Sezione di Torino, Viale Mattioli 25. 10125 Torino, Italv This review focuses on mycorrhizas, which are associa- tions between fungi and the roots of 90% of terrestrial plants. These are the most common svmbioses in the world; the\ involve about 6000 species of fungi distributed through all the fungal phvla and about 240,000 species of plants, including forest and crop plants. Thanks to mycorrhizal symbiosis and nutrient exchanges, regulated b\ complex molecular signals, the plant improves its vegetative growth, while the fungus accomplishes its life cvcle. Molecular and cellular analvses demonstrate that during colonization the cellular organization of the t\vo eukaryotes is completelv remodeled. For example, in cortical cells, structural modi- fications involve both the host and the microbiont. Recent studies revealed that in arbuscular mycorrhizas (AM), sys- tem complexity is increased bv the presence of a third svmbiont: a bacterium living inside the fungus. The pres- ence of this resident genome makes the investigation of the molecular dialogues among the symbiotic partners even more complex. Molecular analysis showed that the bacte- rium has genes involved in the acquisition of mineral nu- trients. The experimental data support the current view that mycorrhizal symbioses are often tripartite associations. Endosymbioses are excellent systems with which to in- vestigate the dialog among genomes and cells. The aim of this short report is to demonstrate that mycorrhizas are of particular interest in an evolutionary context, because they E-mail: paola.bonfante@unito.it, p. bonfanteCs'ipp. cnr.it The paper was originally presented at a workshop titled Outcomes of Genome-Genome Interactions. The workshop, which was held at the J. Erik Jonsson Center of the National Academy of Sciences, Woods Hole. Massachusetts, from 1-3 May 2002, was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266. offer a fine example of interactions between plants, fungi, and bacteria in the rhizosphere (Fig. 1 ). Defining Mycorrhizas Mycorrhizas are complex interactions comprising fungi belonging to different taxonomic groups and about 90% of land plants (Smith and Read, 1997). Mycorrhizas are suc- cessful in colonizing diverse environments, and their eco- logical success reflects a high degree of diversity in the genetic and physiological abilities of the fungal endophytes. About 6000 species in the Zygo-, Asco-, and Basidiomyco- tina have been recorded as mycorrhizal, but the advent of molecular techniques is increasing the number, since new mycorrhizal species as well as new associations are de- scribed on the basis of their DNA fingerprinting (Bidar- tondo et at.. 2002). The taxonomic position of plant and fungal partners defines the types of mycorrhiza, the main distinction of which is between endomycorrhiza and ecto- mycorrhiza. Generally speaking, the fungal hyphae in en- domycorrhiza penetrate the root cells to establish an intra- cellular symbiosis, whereas in ectomycorrhiza the hyphae remain extracellular. However, various patterns of coloni- zation are adopted by mycorrhizal fungi during their inter- actions with the host, mostly among endomycorrhizal fungi. In addition to arbuscular mycorrhizal (AM) fungi, which will be discussed in greater detail, ericoid mycorrhizal fungi colonize the root cells of their host, producing an infection unit that involves a single host cell without spreading to the neighboring root cells. In orchid mycorrhizas, intracellular coils in cortical cells are usually produced by Basidiomy- cetes during both the protocorm and root colonization; in this case, the infection unit comprises a larger number of host cells. Ectomycorrhizal fungi have a more recent evo- lutionary history; they do not penetrate the host cell wall, and they complete their colonization through two major 215 216 P. BONFANTE events: the production of a tissue-like structure (the mantle) covering the root surface, and the development of a laby- rinthine, extracellular hyphal network within the root tis- sues, termed the Hartig net (Bonfante, 2001 ). This report focuses on arbuscular mycorrhizal (AM) fungi, which have been recently classified in a new taxon, the Glomeromycota (Schiissler et a/., 2001). Fossil and molecular data suggest that roots and AM fungi have shared a cooperative life since Devonian times (Simon et ill., 1993). The success of mycorrhizas in evolu- tion is mainly due to the central role that AM fungi play in the capture of nutrients from the soil in almost all ecosys- tems (Smith and Read, 1997), and in phosphate uptake in particular (Smith and Barker, 2002, for a review). As a consequence, they are crucial determinants of plant biodi- versity, ecosystem variability, and plant community produc- tivity (van der Heijden et ai, 1998). AM fungi are not only an essential feature of the biology and ecology of most terrestrial plants, they also interact with different classes of bacteria during their life cycles. In fact, AM fungi establish interactions both with bacteria living in the rhizosphere (Fig. 1) during their extraradical phase and with endosym- biotic bacteria that live in the cytoplasm of some fungal isolates (Perotto and Bonfante, 1997; Bonfante el ai, 2001 ). To understand these multiple interactions and to apply them in low-chemical-input agricultural systems is one of the most exciting challenges of current research in the field of molecular microbe-plant interactions. Plant-Fungal Interactions: Cells, Genes, and Signals The impressive diversity of the plant and fungal taxa involved in mycorrhizal symbiosis has resulted in their anatomical description in many hosts since the early twen- tieth century (Smith and Smith, 1997, for a review). The characterization of mycorrhizal phenotypes has led to a well-defined picture of the colonization by AM fungi, the main aspects of which are summarized on the web site (http://www.bioveg.unito.it/lotus.htm). The availability of plant mutants with impaired symbiotic capabilities has demonstrated that colonization is a multi- step, genetically regulated process that is under the control of specific loci (Bonfante et ai, 2000; Novero et ai, 2002). As a consequence of this process, the cellular organization of the two eukaryotes is completely remodeled. A detailed analysis of cell-to-cell interactions between host and myco- biont identifies the cell walls, membranes, and cytoskeleton of both partners as the structures where crucial changes occur (Bonfante, 2001). However, analyzing the molecular bases of the dialogue Figure 1. The scheme illustrates the multiple cellular interactions established among AM fungi, root cells, and bacteria in the rhizosphere. The magnification of the spore (insert at top left) shows the endobacteria living in the cytoplasm. MYCORRHIZAL SYMBIOSIS 217 between the two partners is not an easy task. Other symbi- otic systems are more advanced: for example, the interac- tion between Rlu-obiiim and legumes is finely regulated by signal molecules, which are perceived by receptors that activate a signal transduction cascade, eventually leading to the activation of target genes and to the production of a nodule (Long. 1996). In AMs, knowledge of these steps is in its infancy; potential signal molecules have been found but not characterized (Buee et ai, 2000). Attention is there- fore mostly focused on the genes that code for proteins responsible for the functional traits in AM fungi; for exam- ple, the phosphate transporter (Harrison and van Buuren, 1995) or a metallothionein gene (Lanfranco et at., 2002). Further challenges are posed by other, obscure aspects of AM: their obligate biotrophic status, their multinuclear con- dition, and an unexpected level of genetic variability (Hijri et ai. 1999: Lanfranco et ai. 1999: Kuhn et al. 2001 ). The genome of AM fungi is in fact huge, ranging from 0.3 pg to 1.12 pg/DNA for the nucleus, depending on the species (Gianinazzi-Pearson et ai, 2001). For all these reasons, genomic projects have not been attempted. On the plant side, other problems are encountered. Ara- bidopsis thaliana, the first plant genome to be sequenced. does not establish any symbiotic interactions (the Arabido- pis Genome Initiative, 2000). However, rice does establish mycorrhizal associations, and genomic data accrued for rice will provide significant information (Goodman et ai. 2002; Sasaki et ai, 2002). Other plants are also being investigated. Breakthroughs from recent research on genomics. involving plants such as Medicago tnincatula and Lotus japonicus, have led to the availability of about 100.000 expressed sequence tags (ESTs). Moreover, the availability of mutants that are impaired in their symbiotic properties has recently led to the discovery of plant genes that code for proteins that are essential to the dialogue between plants and both sym- biotic microbes AM fungi and rhizobia (Stracke et ai, 2002; Endre et ai, 2002). The genes NORK and SYMRK belong to a large class of plant and animal genes that code for receptor complexes (Kistner and Parniske, 2002). Both SYMRK and NORK present a repeated leucine-rich motif (RLM) in their extracellular domain and an intracellular receptor like-kinase domain (RLK). The discovery of these genes which encode plant receptor kinases required for fungal and bacterial symbioses opens the way for detailed analysis of signal perception and downstream signaling pathways that are associated with microbial recognition (Spaink, 2002; Kistner and Parniske, 2002). Plant-Fungal-Bacterial Interactions In addition to the well-known interactions between plants and fungi, mycorrhizal roots offer excellent ecological niches for other microbes; some rhizosphere bacteria adhere tightly to fungal hyphae, whereas others are directly asso- ciated with the root surfaces (Bianciotto et ai, 2001; Bian- ciotto and Bonfante. 2002). In addition, mycorrhizal fungi may host bacteria that complete their life cycles within fungal cells. As opposed to many other eukaryotic cells, which show some level of integration with bacteria and are increasingly appreciated by ecologists and evolutionary bi- ologists for their huge diversity (Moran and Wernergreen, 2000), fungi offer only a limited number of examples. One of the best known is Geosiphon pyriforme, a zygomycete closely related to Glomales. It can host cyanobacteria in- side characteristic bladders in the apical hyphal region (Schiissler and Kluge. 2001). AM fungi are unique in hosting bacteria in their cyto- plasm. Intracellular structures very similar to bacteria and bacteria-like organisms (BLOs) were first described in the 1970s (Scannerini and Bonfante. 1991 for a review). Ultra- structural observations clearly revealed their presence in many field-collected fungal isolates. Further investigation of these BLOs. including the demonstration of their pro- karyotic nature, was long hampered because they could not be cultured. Only a combination of morphological observa- tions (electron and confocal microscopy) and molecular analyses allowed us to identify BLOs as true bacteria and to start unraveling their symbiotic relationship with AM fungi (Bianciotto et ai, 1996). Isolate BEG 34 of the fungus Gigaspora margarita con- tains a large number of BLOs that can be easily detected by staining with fluorescent dyes that are specific for bacteria and can distinguish between live and dead ones. On the basis of the 16S rDNA sequences, the bacterial endosym- bionts living in G. margarita (BEG 34) were first identified as belonging to the genus Burkholderia (Bianciotto et ai, 1996). As a further step, on the basis of the 16S rDNA amplified from isolates of Scutellospora persica, S. casta- nea, and G. margarita, a strongly supported clade was obtained, which contained all endosymbiotic bacteria so far sequenced in Gigasporaceae. It was located close to the genus Burkholderia, as well as to the genera Ralstonia and Pandorea. A new bacterial taxon was therefore proposed: Candidatns Glomeribacter gigasporarum (Bianciotto et al., 2003). The results demonstrate that endobacteria are wide- spread in Gigasporaceae, and suggest that they represent a stable cytoplasmic component. Preliminary results showing that bacteria move along with the fungi from one generation to the next, following a vertical transmission mechanism (V. Bianciotto and G. Becard. unpubl.), provide a first experi- mental confirmation of the statement. A number of morpho- logical observations showing bacteria living inside Glomus spores and hyphae (Scannerini and Bonfante, 1991) might suggest that endobacteria are not limited to the Gigaspora- ceae. However, attempts to obtain ribosomal sequences and to identify these endophytes on the basis of their DNA sequences have been so far unsuccessful, suggesting that 218 P, BONFANTE these bacteria are limited in number or, if present, belong to a mixed population. The functional significance of AM fungal endobacteria is not clear; many attempts to cultivate them have been un- successful. The finding that a genomic library developed from G. margarita spores also has bacterial sequences (van Buuren el a/.. 1999) helped us to identify some genes belonging to Candidatus Glomeribacter gigasporarum. Among the bacterial genes so far identified, the most inter- esting are those involved in nutrient uptake (i.e. a putative phosphate transporter operon. pstY, in colonization events by bacterial cells (vac); and in chemotaxis (Ruiz-Lozano and Bonfante, 1999, 2000; Minerdi et al., 2002). A DNA region containing putative nitrogenase coding genes (nif operon) was also found (Minerdi et al., 2001), but these genes have not yet been demonstrated to belong to the Candidatus Glomeribacter genome. Conclusions In conclusion, the analysis of the multiple interactions established by AM fungi with plant and bacterial cells offers new keys for understanding the complexity of AM symbi- osis. In addition to the still-open question about the signal molecules produced by AM fungi and recognized by poten- tial receptors, the widespread presence of bacteria inside or specifically associated with AM fungi suggests that many AM symbioses are tripartite associations. This possibility will lead to the definition of new parameters in the design of mixed inocula, while the identification of fungal strains that contain endosymbiotic bacteria with important genetic traits opens up new strategies for the practical use of AM fungi. Acknowledgments P.B. thanks Dr. Diana Jennings for her support during the meeting and for her careful revision of the text. Many thanks are due to the AM symbiosis group (L. Lanfranco, A. Genre, A. Faccio, M. Novero, I. Martini. M. T. Delia Beffa). and to the Glomeribacter group (V. Bianciotto, D. Minerdi. E. Lumini, S. Abba). The research illustrated in this review has been funded by the EU GENOMYCA project, QLK5- CT-2000-01319. and by the Italian National Council of Research. Literature Cited Arabidopsis Genome Initiative. 2000. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796-815. Bianciottn, V., and P. Bonfante. 2002. Arbuscular mycorrhizal fungi: a specialized niche for rhizospheric and endocellular bacteria. Antonie Leeuwenhuek 81: 365-375. Bianciotto, V., C. Bandi. D. Minerdi, M. Sironi. H. V. Tichy, and P. Honl'ante. 1996. An ohligately endosymbiotic fungus itself harbors obligately intracellular bacteria. Appl. Environ. Microbiol. 62: 3005- 3010. Bianciotto. V., S. Andreotti. R. Balestrini, P. Bonfante, and S. Perotto. 2001. Mucoid mutants of the biocontrol strain Pse iidomonus fluore- scens CHAD show increased ability in biofilm formation on mycorrhi- zal and nonmycorrhizal carrot roots. Mol Plant-Microbe Interact. 14: 255-260. Bianciotto, V., E. Lumini. P. Bonfante, and P. Vandamme. 2003. 'Candidatus Glomeribacter gigasporarum'. an endosymbiont of arbus- cular mycorrhizal fungi. Int. J. Syst. Erol. Microbiol. 53: 121-126. Bidartondo. M. I., D. Redecker. I. Hijri, A. Wiemken, T. D. Bruns, L. Dominguez, A. Sersic, J. R. Leake, and D. J. Read. 2002. Epi- parasitic plants specialized on arbuscular fungi. Nature 419: 389-392. Bonfante. P. 2001. At the interface between mycorrhizal fungi and plants: the structural organization of cell wall, plasma membrane and cytoskeleton. Pp. 45-91 in Mycota. IX: Fungal Associations. B. Hock, ed. Springer Verlag. Berlin. Bonfante, P., A. Genre, A. Faccio, I. Martini, L. Schauser, J. Stou- gaard, J. Webb, and M. Parniske. 2000. The Lotus japonicus LjSym4 gene is required for the successful symbiotic infection of root epidermal cells. Mol. Plant-Microbe Interact. 13: 1109-1120. Bonfante, P., V. Bianciotto, J. M. Ruiz-Lozano, D. Minerdi, E. Lumini, and S. Perotto. 2001. Arbuscular mycorrhizal fungi and their en- dobacteria. Pp. 323-337 in S\mbiosis. ). Seckbach. ed. Kluwer, Dor- drecht. Buee, M., M. Rossignol, A. Jauneau. R. Ranjeva, and G. Becard. 2000. The presymbiotic growth of arbuscular mycorrhizal fungi is induced by a branching factor partially purified from plant root exudates. Mol. Plant-Microbe Interact. 13: 693-698. Endre, G., A. Kereszt, Z. Kevei, S. Mihacea, P. Kalo, and G. B. Kiss. 2002. A receptor kinase gene regulating symbiotic nodule develop- ment. Nature 417: 962-966. Gianinazzi-Pearson, V., D. van Tuinen, E. Dumas-Gaudot, and H. Dulieu. 2001. Exploring the genome of Glomalean fungi. Pp 3-17 in Mycota, IX: Fungal Associations. B. Hock. ed. Springer Verlag. Berlin. Goodman R. M., R. Naylor, H. Tefera, R. Nelson, and W. Falcon. 2002. The rice genome and the minor grains. Science 296: 1801-1804. Harrison, M. J., and M. L. van Buuren. 1995. A phosphate transporter from the mycorrhizal fungus Glomus versiforme. Nature 378: 626- 629. Hijri, M., M. Hosnv, D. van Tuinen, and H. Dulieu. 1999. Intraspecific ITS polymorphism in Scutellospora castanea (Glomales, Zygomycota) is structured within multinucleate spores. Fungal Genet. Biol. 26: 141-151. Kistner, C., and M. Parniske. 2002. Evolution of signal transduction in intracellular symbiosis. Trends Plant Sci. 7: 51 1-518. K ii h ii. G., M. Hijri, and I. R. Sanders. 2001. Evidence for the evolution of multiple genomes in arbuscular mycorrhizal fungi. Nature 414: 745_748. Lanfranco, L., M. Delpero, and P. Bonfante. 1999. Intrasporal vari- ability of ribosomal sequences in the endomycorrhizal fungus Giga- spora margarita. Mol. Ecol. 8: 37 15. Lanfranco, L., A. Bolchi, E. Cesale Ros, S. Ottonello, and P. Bonfante. 2002. Differential expression of a metallothionein gene during the presymbiotic versus the symbiotic phase of an arbuscular mycorrhizal fungus. Plant Physial. 130: 58-67. Long. S. R. 1996. Rhizobium symbiosis: nod factors in perspective. Plant Cell 8: 1885-1898. Minerdi, D., R. Fani, R. Gallo, A. Boarino, and P. Bonfante. 2001. Nitrogen fixation genes in an endosymbiotic Biirkholtteria strain. Appl. Environ. Microbiol. 67: 725-732. Minerdi, D.. R. Fani, and P. Bonfante. 2002. Identification and evolu- tionary analysis of putative cytoplasmic A/i/>A-like protein in a bacte- rial strain living in symbiosis with a mycorrhizal fungus. J. Mol. Evol. 54: 815-824. MYCORRHIZAL SYMBIOSIS 219 Moran, N. A., and J. J. VVernegreen. 2000. Lifestyle evolution in symbiotic bacteria: insights from genomics. Tree 15: 321-326. Novero. M., A. Faccio, A. Genre. J. Stougaard, K. J. Webb, L. Mulder, M. Parniske, and P. Bonfante. 2002. Dual requirement of the LiS\m4 gene for mycorrhizal development in epidermal and cortical cells of Lotus japonicus roots. New Phytol. 154: 741-750. Perotto, S., and P. Bonfante. 1997. Bacterial associations with mycor- rhizal fungi: close and distant friends in the rhizosphere. Trends Mi- cmbiol. 5: 496-503. Ruiz-Lozano, J. M., and P. Bonfante. 1999. Identification of a putative P-transporter operon in the genome of a Burkholderia strain living inside the arbuscular mycorrhizal fungus Gigaspora margarita. J. Bacterial. 181: 4106-4109. Ruiz-Lozano, J. M., and P. Bonfante. 2000. Intracellular Burkholderia of the arbuscular mycorrhizal fungus Gigaspora margarita possesses the vacB gene, which is involved in host cell colonization by bacteria. Microb. Ecol. 39: 137-144. Sasaki, T., T. Matsumoto, K. Yamamoto, K. Sakata, et al, 2002. The genome sequence and structure of rice chromosome 1. Nature 420: 312-316. Scannerini, S., and P. Bonfante. 1991. Bacteria and bacteria like ob- jects in endomycorrhizal fungi (Glomaceae). Pp. 273-287 in Symbiosis as Source of Evolutionary Innovation: Specialion and Morphogenesis, L. Margulis and R. Fester, eds. The MIT Press. Cambridge. MA. Schiissler, A., and M. Kluge. 2001. Geosiphon pyrifonne. an endocy- tosymbiosis between fungus and cyanobacteria. and its meaning as a model system for arbuscular mycorrhizal research. Pp. 151-161 in Mycota. IX: Fungal Associations, B. Hock, ed. Springer Verlag, Berlin. Schussler. A., H. Gehrig, D. Schwarzott, and C. Walker. 2001. Anal- ysis of partial Glomales SSU rRNA genes: implications for primer design and phylogeny. Mycol. Res. 105: 5-15. Simon, L., J. Bousquet. R. C. Levesque, and M. Lalonde. 1993. Origin and diversification of endomycorrhizal fungi and coincidence with vascular land plants. Nature 363: 67-69. Smith, F. A., and S. E. Smith. 1997. Tansley Review No. 96. Structural diversity in (vescicular)-arbuscular mycorrhizal symbioses. New Phy- tol. 137: 373-388. Smith, S. E., and S. J. Barker. 2002. Plant phosphate transporter genes help harness the nutritional benefits of arbuscular mycorrhizal symbi- osis. Trends Plant Sci. 7: 189-190. Smith, S. E., and D. J. Read. 1997. Mycorrhizal Symbiosis. Academic Press, London. Spaink, H. P. 2002. A receptor in symbiotic dialogue. Nature 417: 910-911. Stracke. S., C. Kistner, S. Yoshida, L. Mulder, S. Sato, T. Kaneko, S. Tabata, N. Sandal, J. Stougaard, K. Szczyglowski, and M. Par- niske. 2002. A plant receptor-like kinase required for both bacterial and fungal symbiosis. Nature 417: 959-962. van Hum i'ii. M., L., Lanfranco, S. Longato. D. Minerdi, M. J. Harri- son, and P. Bonfante. 1999. Construction and characterization of genomic libraries of two endomycorrhizal fungi: Glomus versiforme and Gigaspora margarita. Mycol. Res. 103: 955-960. van der Heijden, M. G. A., J. N. Klironomos, M. Ursic, P. Moutoglis, R. Streitwolf-Engel, T. Boiler, A. Wiemken, and I. R. Sanders. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature 396: 69-72. Discussion QUESTION: Can you cure the fungal symbiont of the bacterial symbiont and then look at the differential effect of the fungus with and without the bacterial symbiont or the plant? BONFANTE: This is a crucial question. We are facing the prob- lem together with Dr. G. Becard from Toulouse. Dr. Becard's group has demonstrated that the bacteria are not sensitive to many antibiotic treatments. In fact, some antibiotics seemed to increase bacterial growth. QUESTION: Can you grow the fungus in nitrogen-free medium? BONFANTE: Yes, arbuscular mycorrhizas fungal spores usually germinate in water. It is well known that the first growth steps (non-symbiotic phases) proceed without other nutrients. COMMENT (JOHN HOBBIE): In Europe, scientists seem to appre- ciate much more than in the U.S. the importance of the my- corrhizal fungi in incorporating nutrients into shrubs and trees. A recent book points out that, in forests, most of the nitrogen and phosphorus enters the trees by way of the mycorrhizal fungi. It is also well known that most of the forests of the world are nitrogen limited. Most research until now has involved the uptake of phos- phorus by mycorrhizae. Plants grown with mycorrhizae clearly grow much better than those grown without, but in many cases nitrogen could have had an effect as well. What the fungi can do. of course, is to get at the organically bound nitrogen, which is most of the nitrogen in the soil. This organically bound nitrogen is not available to the plant without the microbial "mineralization" to ammonium and nitrate, or without an enzymatic breakdown to amino acids, uptake, and transport to the roots by the mycorrhizal fungi. The fungi obtain sugars from the tree roots and provide nitrogen, phosphorus, and even water to the tree or shrub. Up to 30% of the carbohydrates fixed in photosynthesis can be trans- ported under ground and respired by the fungi. The importance of the symbiosis between plant roots and my- corrhizal fungi is just beginning to be investigated. Ecologists want to know about the regulation of the fungal breakdown of proteins and other nitrogen compounds, and how much of the total pool of organic nitrogen in soil is available. BONFANTE: While the role of AM fungi in phosphate uptake is largely acknowledged, their role in nitrogen uptake is a recent discovery. Fitter's laboratory recently published an interesting article in Nature, which shows that AM fungi use an organic N source 1 . The role of ectomycorrhizal fungi in nutrient cycles is 1 Hodge. A., C. D. Campbell, and A. H. Fitter. 2001. An arbuscular mycorrhizal fungus accelerates decomposition and acquires nitrogen di- rectly from organic material. Nature 413: 297-299. 220 P. BONFANTE well demonstrated, and molecular mechanisms are being explored. known to accompany that in the roots. This has been well de- The most recent views on these topics can be found in the book by scribed in ectomycorrhizae: early colonizers and later colonizers van der Heijden and Sanders". have been identified in many plant communities. The development of molecular probes is now providing new views and opened new QUESTION: Is there any evidence yet that, in a forced succession, questions on the identification of spatial and temporal factors there might be a concomitant succession of fungi in the soil? underlying community structures 1 . The new approaches of molec- ular ecology will allow us to directly monitor the fungal succession BONFANTE: Under natural conditions, a fungal succession is under m i crocO smal conditions. - van der Heijden. M. G. A., and 1. R. Sanders, eds. 2002. Mycorrhizal ' Dahlberg, A. 2001. Community ecology of ectomycorrhizal fungi: an Ecology. Ecological Studies, Vol. 157. Springer, New York. advancing interdisciplinary field. New Phytol. 150: 555. Reference: Bid. Bull. 204: 221-231. (April 2003) 2003 Marine Biological Laboratory Genome Evolution in an Insect Cell: Distinct Features of an Ant-Bacterial Partnership JENNIFER J. WERNEGREEN.* PATRICK H. DEGNAN, ADAM B. LAZARUS, CARMEN PALACIOS, AND SETH R. BORDENSTEIN Josephine Bay Paul Center for Comparative Molecular Biology and Evolution, Marine Biological Laboraton; 7 MBL Street, Woods Hole. Massachusetts 02543 Bacteria that live exclusively within eukaryotic host cells include not only well-known pathogens, but also obligate mutualists, many of which occur in diverse insect groups such as aphids. psyllids, tsetse flies, and the ant genus Camponotus (Buchner, 1965; Douglas, 1998; Moran and Telang, 1998; Baumann et al., 2000; Moran and Baumann, 2000). In contrast to intracellular pathogens, these primary (P) endosvmbionts of insects are required for the sunival and reproduction of the host, exist within specialized host cells called bacteriocytes, and undergo stable maternal transmission through host lineages (Buchner, 1965; McLean and Houk, 1973). Due to their long-term host associations and close phylogenetic relationship with well- characterized enterobacteria (Fig. 1), P-endosymbionts of insects are ideal model systems to examine changes in genome content and architecture that occur in the context of beneficial, intracellular associations. Since these bacteria have not been cultured outside of the host cell, they are difficult to study with traditional genetic or physiological approaches. However, in recent years, molecular and com- putational approaches have provided important insights into their genetic diversity and ecological significance. This review describes some recent insights into the evolutionary genetics of obligate insect-bacteria symbioses, with a par- ticular focus on an intriguing association bet\veen the bac- terial endos\mbiont Blochmannia and its ant hosts. * To whom correspondence should be addressed. E-mail: jwernegreen@ mbl.edu The paper was originally presented at a workshop titled Outcomes of Genome-Genome Interactions. The workshop, which was held at the J. Erik Jonsson Center of the National Academy of Sciences. Woods Hole, Massachusetts, from 1-3 May 2002, was sponsored by the Center for Advanced Studies in the Space Life Sciences at the Marine Biological Laboratory, and funded by the National Aeronautics and Space Adminis- tration under Cooperative Agreement NCC 2-1266 The specific functions of most insect P-endosymbionts remain unknown. However, the typically unbalanced diets of their hosts suggest these bacteria play a nutritional role. In fact, nutritional functions are well documented in certain P-endosymbionts such as Buchnera aphidicola, which pro- vides essential amino acids that are lacking in the plant sap diet of its aphid host (Sandstrom et al., 2000), and Wiggles- worthia glossinidia, which provisions its tsetse fly host with B-complex vitamins lacking in vertebrate blood (Nogge, 1981; Aksoy, 2000). The bacterium SOPE (Sitophilus oryzae primary endosymbiont) oxidizes excess methionine consumed by the weevil (Gasnier-Fauchet and Nardon. 1986) and increases mitochondrial enzymatic activity of the host by providing vitamins such as panthothenic acid and riboflavin (Wicker and Nardon, 1982; Heddi et al.. 1991, 1999). Blattobacterium, a symbiont that lives within fat bodies of cockroaches, may be involved in tyrosine biosyn- thesis (Goldberg and Pierre, 1969) and apparently recycles uric acid of the host, as evidenced by elevated uric acid levels in hosts from which the bacteria have been eliminated (Cochoran, 1985). The biosynthetic abilities of these bacte- ria allow hosts to exploit food sources and habitats that would otherwise be inadequate; therefore, symbiont acqui- sition can be viewed as a key innovation in the evolution of these insect lineages (Moran and Telang, 1998). Indeed, sister genera lacking P-endosymbionts typically utilize a different, more nutrient-rich dietary resource (Nardon and Grenier, 1991). These insect-bacteria relationships are re- ciprocally beneficial, as the highly specialized bacteria rely on the host cellular environment for their survival. We are studying the evolutionary outcome of an obligate association between ants and a bacterial endosymbiont re- cently designated Candidutus Blochmannia gen. nov. (Blochmannia) (Sauer et al., 2000). This bacterium lives 221 222 J. J. WERNEGREEN ET AL within host cells that are typically adjacent to or within the midgut epithelium, and forms an obligate association with at least three closely related genera in the ant subfamily Formicinae: Polyrhachis, Colobopsis, and Camponotus (Dascheffl/., 1984; Schroder el al., 1996; Sameshima et al.. 1999). Camponotus is the second largest ant genus; it in- cludes 931 species and is found in almost every biogeo- graphic region (Bolton, 1995). Although Blochmannia is the best characterized ant endosymbiont. we know little about its gene content and nothing about its functional role. Build- ing upon relatively rich molecular datasets currently avail- able for Bitchnera and Wigglesworthia, we are using com- parative approaches to explore the impact of endosymbiosis on sequence and genome evolution in Blochmannia, includ- ing rates and patterns of DNA and protein sequence evolu- tion and changes in genome size and content. In future work, genomic comparisons and experimental approaches promise to shed light on the physiological and ecological roles of this bacterium. Forces Shaping Sequence Evolution in Endosymbionts Phylogenetic analysis of insect mutualists Analyses of 16S rDNA genes show that many insect endosymbionts group with the y-3 subdivision of Pro- teobacteria, along with Escherichia coli and related entero- bacteria(Fig. 1) (Munson et al., 1991; Schroder et al., 1996; Sameshima et al., 1999). Furthermore, many previous stud- ies suggest that Buchnera, Wigglesworthia, and Blochman- nia share a very close phylogenetic relationship and may be monophyletic (e.g., Schroder et a/., 1996; Spaulding and von Dohlen, 1998; Sauer et al., 2000). However, these phytogenies were often estimated using maximum parsi- mony or distance approaches, which may be highly biased by the fast evolutionary rates and strong AT bias of endo- symbiont sequences. In contrast, maximum likelihood anal- ysis (Fig. 1) strongly suggests that Buchnera is phyloge- netically distinct from Wigglesworthia and Blochmannia, but that the last two are closely related to each other. This aspect of the phylogeny agrees with that proposed by Charles et al. (2001). Multiple, independent origins of en- dosymbiotic lifestyles are not entirely surprising, since en- dosymbiotic bacteria have arisen in many classes of bacte- ria, including cv-Proteobacteria, /3-Proteobacteria, and flavobacteria (Douglas, 1989; Moran and Telang. 1998). Indeed, multiple origins of endosymbiosis within the y-Pro- teobacteria makes this group an ideal model to explore (i) phylogenetically independent transitions to an endosymbi- otic lifestyle (e.g., Buchnera versus Wigglesworthia and Blochmannia). and (ii) adaptation of closely related bacte- rial species to quite different host associations (Wiggles- worthia versus Blochmannia). The phylogenies of P-endosymbionts within a single host system are generally congruent with their hosts' species. indicating that the symbiont origin traces back to a single, often ancient, infection event within each host group. This pattern of host-symbiont cospeciation has been demon- strated for the aphid- Buchnera symbiosis (Munson et al.. 1991; Clark et al., 2000; Funk et al., 2000), the tsetse fiy-Wigglesworthia symbiosis (Chen et al.. 1999), and the psy\\id-CarsonelIa symbiosis (Thao et at., 2000). Thus, in contrast to facultative symbionts or pathogens that can transfer horizontally to new hosts, P-endosymbionts have been vertically inherited over long evolutionary timescales. These ancient insect-bacterial associations date back as far as 150-200 MYA for the aphid-Buchnera association (Mun- son et al.. 1991) and about 50-100 MYA for tsetse fly- Wigglesworthia (Moran and Wernegreen, 2000). Phyloge- netic congruence between Camponotus and Blochmannia also holds true across a large number of host species (Sameshima et al., 1999; Sauer et al., 2000; P. H. Degnan et al., unpubl. data) as well as within a single Camponotus species (A. B. Lazarus et al.. unpubl. data), indicating the ant-bacterial association is evolutionarily stable and at least as old as the genus Camponotus (>20 MY; Wilson. 1985) if not even older. Population bottlenecks and genetic drift Maternal transmission of P-endosymbionts is thought to impose a population bottleneck that reduces the number of bacteria passed on from mother to offspring (Mira and Moran. 2002). Successive bottlenecks throughout the evo- lution of these ancient associations are expected to reduce the effective population size (N e ) of the bacterial partner. Consequently, endosymbiont population sizes may be de- termined by insect host population sizes, which are orders of magnitude smaller than the extremely large populations of free-living bacteria (N e ~ 10 q for species of enterobacteria; Selander et al., 1987). Models of nearly neutral evolution predict that reduced N e will lower the efficacy of natural selection and will elevate rates of fixation of deleterious mutations through random genetic drift (Ohta, 1973). Over time, the accumulation of deleterious mutations may nega- tively affect the fitness of the symbiont and host (Rispe and Moran. 2000; Wernegreen and Moran, 2000). Exacerbating the effect of genetic drift in P-endosymbionts is their ap- parent lack of recombination among genetically distinct lineages (Funk et al., 2000; Wernegreen and Moran, 2001 ). This strict asexuality contrasts with recombination in free- living bacterial strains, and may produce an effect known as Muller's ratchet (Mullen 1964, Moran, 1996) in which genetic drift in small populations is increased because wild- type genotypes cannot be introduced through recombina- tion. Several studies show that the repair of slightly deleterious mutations is important in shaping sequence evolution of P-endosymbionts. Evidence for drift includes fast rates of OUTCOMES OF BACTERIA-INSECT SYMBIOSES 223 sequence evolution, changes that destabilize the 16S rRNA secondary structure, elevated rates of amino acid substitu- tions, and higher ratio of nonsynonymous to synonymous substitutions (dN/dS) compared to free-living bacteria (Moran, 1996; Brynnel et ai, 1998; Lambert and Moran. 1998; Wernegreen and Moran, 1999; Clark et at., 1999). Similarly, protein-coding genes of Blochmannia show ac- celerated rates of evolution and elevated dN/dS, suggesting this ant symbiont may also experience strong genetic drift (unpubl. data). Pervasive acceleration of protein evolution across the genome is not easily explained by relaxed or positive selection, which is expected to act at individual genes. Nor can elevated mutation alone explain the ob- served rate increase, since mutation would affect dN and dS equally, with no expected change in dN/dS. Finally, popu- lation genetic analyses of Buchnera associated with two aphid species show low levels of sequence polymorphism consistent with population bottlenecks, and an excess of rare alleles and nonsynonymous polymorphisms that suggest strong effects of genetic drift (Funk et ai. 2001; Abbot and Moran, 2002). In total, the pervasive rate elevation across genes, elevated dN/dS, and intraspecific polymorphism data are most consistent with reduced efficacy of selection across the genome due to genetic drift in small populations. Intra- cellular pathogens also show elevated rates of protein evo- lution that suggest genome degradation through genetic drift (Andersson and Andersson, 1999). On a related note, a recent study attributed accelerated protein evolution in Buchnera to mutational bias alone (Itoh et a/.. 2002). The authors argued against any role of genetic drift, on the basis that Buchnera presumably did not show an increase in dN/dS. However, the dN and dS values used to calculate this ratio were derived from different studies of different Buchnera loci. Thus, the apparent lack of elevated dN/dS must be weighed against the results of several studies that show a significant elevation of this ratio when dN and dS are calculated from the same dataset (Moran, 1996; Brynnel et at., 1998; Wernegreen and Moran, 1999; Clark et al, 1999). Itoh et at. (2002) also raise the intriguing sug- gestion that if slightly deleterious mutations are fixed in populations over time, these mutations will eventually ren- der all genes functionless. However, mutations that severely impair or eliminate functions of necessary genes are not slightly deleterious; rather, these mutations would have high selective coefficients and would be eliminated even from small endosymbiont populations. For example, in the AT- rich Buchnera genome, high-expression genes (e.g., chap- eronins and ribosomal proteins) have distinct amino acid usage patterns compared to genes with putatively low ex- pression levels (Palacios and Wernegreen, 2002). High- expression genes tend to use amino acids that are less aromatic and are encoded by relatively GC-rich codons, suggesting strong selection against aromatic amino acids and against amino acids with AT-rich codons. Thus, while AT mutational bias and genetic drift influence amino acid usage in Buchnera, selection at high-expression genes is sufficiently strong to attenuate the effects of mutational bias on amino acid content. We could generalize that, despite strong effects of genetic drift, selection still constrains del- eterious amino acid changes in Buchnera, especially at high-expression loci. Mutational bias in endosymbionts: random or adaptive? Along with genetic drift, intracellular mutualists and pathogens also experience strong mutational pressure that, over time, can severely alter the base composition of their genomes. In contrast to the moderate base compositions of the enterics, sequences of intracellular bacteria are charac- terized by extremely low percentage content of GC (Fig. 2). A 37-kb fragment of the Carsonetla ruddii genome was recently found to be just 19.9% GC, making this psyllid symbiont the most AT-rich bacterial genome yet character- ized (Clark et al., 2001). Analysis of six kilobases of Bloch- mannia sequences (unpubl. data) corroborates earlier evi- dence of low GC content for this bacterial genome (~30% GC; Dasch. 1975). This AT bias has a strong impact on the amino acid composition of Buchnera, Carsonella, and Wigglesworthia proteins, which are highly biased toward amino acids encoded by AT-rich codons (Clark et al., 1999, 2001; Akman et at.. 2002; Palacios and Wernegreen. 2002). Two main hypotheses have been proposed to explain base compositional biases observed in most obligately intracel- lular mutualists and pathogens. First, AT richness may reflect strong mutational bias resulting from the loss of DNA repair genes by random genetic drift (Eisen and Hanawalt, 1999; Moran and Wernegreen, 2000). According to this hypothesis, an underlying AT mutational bias is repaired less efficiently in small intracellular genomes that lack certain repair functions. Second, AT bias may be an adaptive feature of an intracellular lifestyle, explained by the high energetic cost and lower accessibility of GTP and CTP compared to ATP and UTP (Rocha and Danchin, 2002). ATP, for example, plays a significant role in cellular metabolism and is the most abundant nucleotide. Under this hypothesis, a nucleotide pool biased toward UTP and ATP and a corresponding AT mutational bias would be more efficient and thus would allow intracellular bacteria to ex- ploit the host cell more effectively. Selection on each GC AT mutation would be miniscule, but selection might favor larger changes (such as gene loss) that contribute to an overall mutation bias. Consistent with this adaptive hypoth- esis, the AT contents of other intracellular elements (e.g., plasmids, phage, and insertion sequences) are also generally higher than those of their hosts, and base composition of phage corresponds to infection type, as virulent phages are more AT-rich than temperate ones (Rocha and Danchin. 2002). 224 J. J. WERNEGREEN ET AL 59 100 Antonina crawii S-endosymbiont Wigglesworthia-Glossma austeni P Wigglesworth/a-Glossina brevipalpis P Blochmannia-Camponotus festinatus P Blochmannia-Camponotus pennsylvanicus P ' Colobopsis nipponicus P-endosymbionl Polyrhachis lamellidens P-endosymbiont Dysmicoccus neobrevipes S-endosymbiont Glossina pallidipes S-endosymbiont -J Plagiolepis pigmaea endosymbiont ' Sitophilus oryzae P-endosymbiont i Uroleucon astronomus S-endosymbiont (U type) 78 I Yamatocalhs tokyoensis S-endosymbiont 100 Yersinia pestis Arsenophonus-Tr/atoma infestans S 99 L Arsenophonus-Nasonia vitripennis S Haemophilus influenzae - Proteus vulgaris - Erwinia herbicola Formica fusca endosymbiont Enterobacter asbunae Klebsiella pneumoniae Salmonella typhimunum 100 :100 Eschenchia coli CBuchnera-Acrythosiphon pisum P Buchnera-Schizaphis grammum P -Pseudomonas aeruginosa - 05 substilutions/site Figure 1. Phylogenetic relationships among insect endosymbionts (boldface) and related y-Proteobacteria. estimated from the 16S rDNA gene. Both maximum likelihood (ML) and Bayesian analyses give the tree topology presented. Values above nodes (in boldface) are bootstrap values for maximum likelihood analysis, and values below nodes are posterior probability values generated by the Bayesian analysis. Branch lengths reflect genetic distance under the maximum likelihood model used. This phylogeny strongly supports the following hypotheses: (i) a single origin of endosymbionts in the ancestor of the ant genera Camponotus, Colobopsis, and Polyrhachis. (ii) independent origins of symbiosis in the ants Formica and Plageolepis, and (iii). that Biichnera is a phylogenetically distinct lineage from Wigglesworthia and Blochmannia, which are closely related. I6S rDNA sequence data: Most 16S rDNA sequences were obtained from GenBank (with the exception new Blochmannia, obtained as described below). Nucleotide sequence accession numbers for other I6S rDNA sequences used in phylogenetic analysis are as follows: Antonina crawii S-endosymbiont AB030020; Buchnera- Acyrthosiphon pisum (P-endosymbiont) NC002528: Arsenophonus-Triatoma infestans (S-endosymbiont) U91786; Colobopsis nipponicus endosymbiont ABO 18676; Dysmicoccus neobrevipes S-endosymbiont AF476104; Emembacter ashuriae AB004744; Eschenclua coli NC0009I3; Em-inUi herbicola AB004757: Formica fusca endosymbiont AB018684; Wigglesworthia-Glossinia austeni (P-endosymbiont) AF022879; Wigglesworthia-Glossinia brevipalpis (P-endosymbiont) L3734I; Glossinia pallidipes S-endosymbiont M99060; Haemophilus inflneiiMC NC000907; Klebsiel/a pneumoniae AB004753; Arsenophonus-Nasonia vit- ripennis (S-endosymbiont) M90801; Plagiolepis pigmaea endosymbiont ABOI8683: Pseudomonas aeruginosa NC0025I6: Polyrhachis lame/liilens P-endosymbiont AB018680; Proteus vu/Kuris J01874: Biichnera -Schi:a- phis f>raiintim (P endosymbiont) L18927; Sitophilux oryzae P-endosymbiont AF005235; Salmonella typhi- murium NC003197; Uroluecon astronomus S-endosymbiont (U type) AF293623; Yersinia pestis NC003143; Yamatocallis tokyoensis S-endosymbiont AB064515. Obtaining Blochmannia I6S rDNA data: Genomic DNA was extracted from individual Camponotus festinatus and C. penns\ivanicus workers using the DNeasy tissue kit (Qiagen) according to the manufacturer's instruc- tions. This DNA was used as template for PCR reaction using SL and SR universal eubacterial 16S rDNA primers (Schroder et a/.. 1996). The single 1.6-kb band PCR product was then cleaned up using a column purification kit (Qiagen) and sequenced on an ABI 3700 automated sequencer. SL. SR. and two internal primers were used to sequence the PCR product. The resulting sequences were assembled and edited using PHRED. OUTCOMES OF BACTERIA-INSECT SYMBIOSES Notable exceptions to the link between an intrucellulur lifestyle and AT bias include the relatively GC-rich ge- nomes of the pathogen Mycobacterium leprae (57.8% GC; Cole et ill.. 2001); the weevil endosymbiont. SOPE (54% GC; Heddi et ai, 1998); and the /3-Proteobacterial subdivi- sion endosymbiont of mealybugs. Treinhlnyii (57.1% GC across a 35-kb region; Baumann et al., 2002). Interestingly, both SOPE and M. leprae also differ from most other obligately intracellular bacteria in having relatively large genome sizes (-3 Mb and 3.27 Mb, respectively) (Fig. 2). The genome size of Tremblaya is unknown, but its close relative Burkholderia pseudomallei has a very large genome (7.25 Mb; unpubl. data of the B. pseudomallei Sequencing Group at the Sanger Institute; http://www.sanger.ac.uk/ Projects/B_pseudomallei/). These larger genomes may re- tain DNA repair functions that are missing from small, AT-rich genomes of most intracellular bacteria. Under- standing the physiology of M. leprae. in particular, may help to distinguish whether the AT mutational bias of most intracellular bacteria is due to a loss of DNA repair functions by genetic drift or to selection for energetic efficiency. This pathogen experiences elevated mutation rates that drive genome deterioration, yet is relatively GC-rich. The hypothesis of adaptive mutational bias would predict that M. leprae has access to more energetic resources and competes less severely with its host for nutrients. Genome Evolution in Endosymbionts: Size Matters Full genome sequences of endosymbionts have provided new insights into the mechanisms and consequences of genome reduction. Recently published endosymbiont ge- nomes include those of Buchnera aphidicola associated with the pea aphid Ac\rthosiphon pisitm ( Ap) (Shigenobu et ill.. 2000). the greenbug Schizaphis graminum (Sg) (Tamas et nl.. 2002), and the gall-forming aphid Baizongia pistacea (van Ham et al., 2003); and Wigglesworthia glossinidia of the tsetse fly Glossina brevipalpis (Akman et al., 2002). Additional endosymbiont genomes, including that of Blocli- mannia, are being sequenced. Comparative genome analyses illustrate striking parallels between P-endosymbionts of insects and intracellular patho- gens. As described previously, pathogens and insect mutu- alists are characterized by reductive genome evolution, a syndrome that includes severely reduced genome size com- pared to free-living bacterial relatives, elevated rates of sequence evolution, and low genomic GC contents (Anders- son and Kurland, 1998) (Fig. 1). The small chromosome sizes of Buchnera (450-650 kb; Charles and Ishikawa, 1999; Wernegreen et al.. 2000; Gil et al., 2002) and Wiggleswor- tliia (698 kb; Akman et al., 2002) imply substantial gene loss since their divergence from the enteric bacteria (4.5-5.5 Mb genome size range for E. coli; Bergthorsson and Ochman, 1995). Because most bacterial genomes contain primarily coding DNA, genome reduction in endosymbionts PHRAP. and CONSED. These 16S rDNA genes of Blochmannia-C. pennsylvanicus and Blochmannia-C festinatus are assigned GenBank accession numbers AY196850 and AY 19685 1, respectively. Phvlogenetic analysis methods: Alignments were created using the Ribosomal Database Project II sequence aligner (Maidak et al. 2001), then manually edited in MacClade v. 4.05 (Maddison and Maddison, 2002). Maximum likelihood parameters were identified according to the Akaike information criterion (AIC) of Modeltest v. 3.06 (Posada and Crandall, 1998). The most likely model was a general time reversible (GTR) model in which invariant sites and the gamma distribution were estimated from the data. The optimized parameters (Rmat = J0.8676 4.6744 2.0447 1.0516 7.4521, shape of gamma distribution = 0.5500, and proportion of invariant sites = 0.5115) were used for all ML searches. The tree topology presented is the consensus of 100 separate heuristic ML searches, each starting from random trees, using PAUP v. 4.0blO; (Swofford, 2002). ML bootstrap values were determined from 100 bootstrap replicates, with each replicate starting from 10 random trees. Replicates were performed in parallel on a Beowulf cluster using the clusterpaup program (A.G. McArthur. http://jbpc.mbl.edu/mcarthur). Bayesian analysis was performed on the same data matrix (MrBayes ver. 2.01 ; Huelsenbeck and Ronquist, 2001 ) by running four simultaneous chains for 300.000 generations, sampling every 100 generations. Stationarity in likelihood scores was determined by plotting the -InL against the generation. All trees below the observed Stationarity level were discarded, resulting in a 'burnin" of 5000 generations. The 50% majority-rule consensus tree was determined to calculate the posterior probabilities for each node. The selected model for Bayesian analysis was the GTR. using empirical base frequencies, and estimating the shape of the gamma distribution and proportion of invariant sites from the data. The Bayesian tree with the best likelihood score was identical to the ML tree presented, and the parameter values across this tree were virtually identical to those obtained in the ML analysis. Limited sequence data (<570 bp of 16S rDNA gene) were available for four taxa included in