Proceedings of the lith International Diatom Symposium San Francisco, Califormia
12-17 August 1990
Edited by JOHN PATRICK KOCIOLEK
Califomma Academy of Sciences
arr
i ‘imal
hy, he.
ied
res el av hate
rr
+ ar
us
" See Se
‘.
m sob bs
pon
i"
Proceedings of the 11th International Diatom Symposium
San Francisco, California
12-17 August 1990
Edited by John Patrick Kociolek
Diatom Collection, California Academy of Sciences
Published by the California Academy of Sciences
San Francisco, California
Memoirs of the California Academy of Sciences, Number 17
SCIENTIFIC PUBLICATIONS COMMITTEE:
Thomas F. Daniel, Editor-in-Chief Ann Senuta, Managing Editor
John Patrick Kociolek, Volume Editor Katie Martin, Editorial Assistant Robert C. Drewes
Wojciech J. Pulawski
Adam Schiff
Gary C. Williams
© 1994 by the California Academy of Sciences, Golden Gate Park, San Francisco, California 94118
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without permission in writing from the pub- lisher.
Library of Congress Catalog Card Number 94-070113 ISBN 0-940228-34-3
Proceedings of the 11th International Diatom Symposium
12-17 August 1990
Table of Contents
ORC WOLG ore he ee eee RIT Lee ine ete AO M ee SiGe conan clierene Rittens
Announcements
Chandra, A. Excerpts from the will of the late (Mrs.) Savitri Sahni (SOD ETO RSS es setrcvsnetenrsae tate seeatters arecuricat a micmae on one tanecoya a ok miro set wane
Morphology, Taxonomy, and Systematics Crawford, R.M. Transmission electron microscopy and diatom research ........
Lee, J.J. Diatoms, or their chloroplasts, as endosymbiotic partners for Roraminiferay ai: tesscactaistaewists rain cig esasra sa sideec at ee ya ora a aisle yaad ate
Rivera, P.S. and H. L. Barrales. Asteromphalus sarcophagus Wallich and other Species.o& the genusioft the coastiof Chile’ ace. setae siecdeers = viele ele ele
Loseva, E. Are both Rhizolenia curvirostris Jousé and R. barboi Brun found in Pleistocene sediments of northeastern Europe? .<. 24.s0 02 qecms ee ease + sare vee
Ferrario, M. E. and E. A. Sar. Valve morphology of Coscinodiscus janischii Schindt(Bacillariophy cea) Mocs cicqowte a ele wherein oiesvhct ee arteleliste cfetel earers m ohei are
Gallagher, J.C. Genetic structure of microalgal populations. I. Problems associated with the use of strains as terminal taxa .............. 20.00.0000 0e
McCartney, K., J. Ernissee, and D. E. Loper. Mathematical modeling of Silicofla- gellate skeletal morphology and implications concerning skeletal latticeworks
Conley, D.J., P. V. Zimba, and E. Theriot. Silica content of freshwater and Marine sbenthiGrdratOMs. acteurs wi.5 cous loess cwigyererwieuaeeaieie eras 5 aedersushenarne swear
Makarova, I. V. The morphology of marine genera of the family Thalassio- Siraceae’ Lebourvemend! Hasle 2.5 5 sacar e ees bis so Oe ews sows nes
Moreno, L. and S. Licea. Morphology of three related Coscinodiscus Ehrenberg taxa from the southern Gulf of Mexico and coastal North Pacific of Mexico
Reichardt, E. Two new species of the genus Surirella from Lake Turkana, CASte ATI Cale wet citer seen eee ced ceee ae A ede i Oe ie le neh cea Sts has Saallaler er sca. s
Kilham, S.S. Do diatom "superspecies" ever evolve? ..............0 eee e ee
Fukushima, H., T. Ko-Bayashi, H. Ohtsuka, and S. Yoshitake. Morphological variability of Navicula recens (Lange-Bertalot) Lange-Bertalot ...............
Czarnecki, D.B. The freshwater diatom culture collection at Loras College, Dubuques LO Wa Manas sen vacoceoa ve cohardencve teria oreo ce bite at eliiere GR D-gie Peers 6m tee eoent ens
Compére, P. Diatoms from hypersaline inland waters of Egypt ...............
John, J. Mastogloia species associated with active stromatolites in Shark Bay, Westcoast Of Australiay, cerevsrencis sore eePecrei seh. «sdf ete notecegeasie mimuarens Guba es heaenine, iclanans
87
Johansen, J. R., S. Lacognata, and J. P. Kociolek. Examination of type material
Of Denticula rainierensts SOVETELOM sees ohote te acts aneveyoicie © ratte acest nate eee 211 Hargraves, P. E. and A. M. Schmid. Morphology, cytology, and growth
characteristics of the diatom Planktoniella sol (Wall.) Schiitt .................221 Kociolek, J.P., L. Sicko-Goad, and E. F. Stoermer. Cytoplasmic fine structure of
tWOrEnceyOnema SPECIES: iiss ares © 6 oie = on eo loo enere iets easrecw ahi ene eucueiebm torent 235 Williams, D.M. Ontogeny and phylogeny in the genus Tetracyclus ........... 247
Khursevich, G. Evolution and phylogeny of some diatom genera of the class Centrophyceae:- .aek.cccaiers ce 2S trees nt cate thehetiece, 2 spon rotons aed aus eene Seale are sae eee PASS |
Khursevich, G. Morphology and taxonomy of some centric diatom species from the Miocene sediments of the Dzhilinda and Tunkin hollows ................. 269
Kobayasi, H., S. Kobori, and $. Sunaga. Taxonomy and morphology of two forms Ofithe NitzsGhia sinuataiGOmMplex mec arr elite reelected cheese ie Se eee 281
Ognjanova-Rumenova, N., D. Temniskova-Topalova, and M. Valeva. Ultrastructure of Fragilaria Lyngbye species (Bacillariophyta) from neogene sediments in Bulgaria. c2.5.5 ceo e.tic sake pac cuesc ease mr evegt tac, tile heey. Gaghts (Ab aes aleridlaes ateaatieh fleets atone SRN 291
Temniskova-Topalova, D., N. Ognjanova-Rumenova, and M. Valeva. Nonmarine biostratigraphy of some genera of the class Centrophyceae from southern
Bulge arta critea.: ates io dotnet aon 5aSe1the te suey cals: sca tye) Beare vale whiasoy ale taualoha ania sat alev dN tafe rsans 301 Licea, S. Thalassiosira species from the Southern Gulf of Mexico ............ 311 Solliday, J.D. Morphological variations in fossil diatoms from Mono Lake ..... 337
Gronlund, T. Lagoonal diatom flora of the Holocene Baltic Litorina Sea in comparison with the Eemian Baltic sea flora .......... eee eee 349
Marine Palaeoecology Loseva, E. Marine Pleistocene diatom assemblages of northeastern Europe ..... 359
Risberg, J. and P. Sandgren. Shore displacement in the Stockholm area, Sweden, during the early Holocene as recorded from diatom and magnetic analysis— Acpreliminany.re porte .eeee cts: one tee ie eee eet esi tenets Se
Williams, K. M. and S.L. Forman. Diatoms and sediment of a modern glacial fjord environment, western Spitsbergen 222.55. ..e9. 202s sosee ees ose ee ee On
Fourtanier, E. and R. Oscarson. Ultrastructure of some interesting and strati- graphically significant diatom taxa from the Upper Palocene to Lower Eocene sediments of ODP site, 752; caster IndianvOcCeam cai ctor rc tetre otenieretneaerees os 399
White, L.D. Diatom biostratigraphy of chert intervals in the Miocene Monterey formation at Pt. Reyes, Pt. Afio Nuevo, Mussel Rock, and Lions Head, California: accord wats. ccinw arena wn 4 eee sencesaet eee aude ie aah oe ova evalekane coleh nail ohehe 411
Brackish and Marine Ecology
Leskinen, E. and G. Hillfors. Dynamics of epiphytic diatoms on Cladophora glomerata in the Tvérminne Archipelago, northern Baltic Sea ................ 425
V1
Fryxell, G.A. Planktonic marine diatom winter stages: Antarctic alternatives to TESEINPUSPOLES.< eyes secestche fersusne See dass ab talie@isuate ahd Dremintle sh tuous ey ae a atnstenvet meres 437
Marshall, H. Spatial and temporal diatom relationships in the lower James River, VAT DUT ae IN SUA re rs We cassie apednseokoh ane nates Rem UEP Poe eesete es he) odotgogeys stay onayaane se ielats 449
Freshwater Palaeoecology
Aasheim, S. Diatom stratigraphy and environmental history of a small lake basin
in;southwest Norway. ls Mhedast 4000) years) peace ce eerie Gee are oe ore 459 Main, S.P. Diatoms in alkaline peat: Preservation and extraction ............. 465 Kovari-Gulyas, E. Well-preserved diatoms in limnic opals of Hungary ......... 473 Harper, M. A. Did Europeans introduce Asterionella formosa Hassall to New
Ze AAG) Se secs x captecn sca yore (ona ang cove Weyer ay aun clo syava yaad aystsve m GMaie ue De aid coe Tiayela ae Peuk Ba 479 Krebs, W.N. The biochronology of freshwater planktonic diatom communities
UMW ESECLIMIN Ofte AUMELI CA areas neue vetore a cetene ute fore egenete gape ss Gin eaueleta ey ole uellsuels) Gila ene) 485
Metcalfe, S. and P. Hales. Holocene diatoms from a Mexican crater lake—La ISG Nag C Oya iil Amen eee tne ea anetet ee ore sue seed escheat spniceieiare ere enacusre ci tes 501
Wasell, A. Diatom-stratigraphy in the sediment of "Skua Lake," Horseshoe Island, Antarctica. Preliminary report ........... 0. eee cee eee eee eee eee S17
Whitmore, T., M. Brenner, and X. L. Song. Environmental implications of the late Quaternary diatom history from Xingyun Hu, Yunnan Province, China ....525
Anderson, N.J. Inferring diatom palaeoproduction and lake trophic status from fossilidiatompasscmblases yee acim ee tiiiiee eerie ee tere eae ere DOS
Shero, B.R. Diatom assemblages of the past 11,000 years from the Yellowstone Waker basing Wyoming. WS. Ay Sate sc aise oaigua a sno eieaieis og acctecaas aicies = sures never 549
Freshwater Ecology
Rogers, C. E. and L. Sicko-Goad. Effects of chemical fixation on flow cytometric properticssol diatoms, 26 oie cee cas tae ac iwes oclueuies soMe wad nee me DOT
Watanabe, T. and K. Asai. Numerical estimation of organic pollution based on the attached diatom assemblage in Lake Biwa and its inflows ................ 567
Kiss, K.T. and G. Pajak. Seasonal change of diatoms in the plankton of the Vistula River, above and below the Goczalkowice Reservoir, Poland ..........583
Burton, T. M., M. P. Oemke, and J. M. Molloy. Effects of grazing by the Trichopteran, Glossoma nigrior, on diatom community composition in the Ford Ray Stee VINCI Sate es peueecncecters spect ous feneter act oney oy on stsl nse tcvetyskenet snsyia clans si ellerepese sneao sch aises 599
Burton, T. M., M. P. Oemke, and J. M. Molloy. Effects of stream order and alkalinity on the composition of diatom communities in two northern Michigan GIVE ISVSLEUIS we yeueyatreyetoeattuene ere since Stays emyineyer excise cece tela te, shin Arey atsispsr avers wecrsytnale ty sie 609
Carney, H.J., D. A. Hunter, and C.R. Goldman. Seasonal, interannual and long- term dynamics of planktonic diatoms in oligotrophic Lake Tahoe ............. 621
Vii
Gell, P. A. and F. Gasse. Relationships between salinity and diatom flora from someyAustralian saline: Lakes! cjeerseistuiere eis eres) tet stetere eieeeaesene ener) neta 631
Huttunen, P. and J. Turkia. Diatoms as indicators of alkalinity and TOC in lakes: Estimation of optima and tolerances by weighted averaging .................. 649
Katoh, K. Spatial variation of diatom assemblages in Minami-Aizu moors,
Fukushima Prefecture, Japan: The effect of pool size ..................0000, 659 Yoshitake, S. and H. Fukushima. Estimation of recovery of water quality in urban :water system iby diatom: flora 22. gece cee ee eee wleis sos eis) eee ia ere 665
Viil
11th DIATOM SYMPOSIUM 1990
Foreword
The 11th International Diatom Symposium was held 12-17 August 1990, in San Francisco, on the campus of San Francisco State University, and it was sponsored in part by the California Academy of Sciences. The meeting was officially convened by Dr. John Barron with the support of the Bay Area Diatomists Group. The Local Organizing Committee, composed of Dr. Barron, Al Mahood, Margaret Hanna, Dr. Elisabeth Fourtanier, Dr. Lisa White, Dr. Eileen Hemphill-Haley, Ray Wong, and James Fidiam, provided organizational, logistical and other practical support, and without them the meeting would not have been as smooth as it was. The committee insisted the main goal of the San Francisco meeting was to make the meeting truly an international one. This original intent was realized as over 200 diatomists from over 30 countries participated in the meeting.
It was my hope in producing this Proceedings volume that the diversity seen in the attendance of the San Francisco meeting would be mirrored by the submissions to the volume. This goal has been realized in part as evidenced by the large number of submis- sions and the diversity of contributors. There is also diversity in the wide range of taxonomic groups studied, and a diversity of habitats, including fossil and Recent in both the freshwater and marine environments. The wide range of subjects, localities and con- tributors suggests to me that, like the goal for the meeting, the intent to make this Proceed- ings reflect the international spirit and flavor of the San Francisco meeting has been attained.
I would like to acknowledge the leadership, encouragement and friendship of John Barron, for without his enthusiasm and hard work the meeting would never have happened. Special thanks are due to all the members of the Local Organizing Committee for their diligence, planning and enthusiasm for the meeting. Members of the Department of Invertebrate Zoology and Geology of the California Academy of Sciences helped in many ways to make the meeting a success; the efforts of Mary Alice Tatarian and Patricia Dal Porto require special recognition. The reviewers of the contributors provided excellent and espe- cially prompt peer review. The patience and good humor of the contributors throughout what has probably seemed an eternity before their scientific efforts see publication are particularly appreciated. Finally, Ann Senuta, managing editor of the Scientific Publications Department of the California Academy of Sciences, has shown great skill and patience in taking my edited/revised copy and producing this high-quality volume. I am indebted to her for all of her hard work.
J. Patrick Kociolek
11th DIATOM SYMPOSIUM 1990
Excerpts from the Will of the Late (Mrs.) Savitri Sahni (1902-1985)
by Anil Chandra
Birbal Sahni Institute of Palaeobotany, Lucknow 226007, India
“It is my desire that in token of my love and devotion to that fine and great spirit of my noble husband late Professor Birbal Sahni and in the pride of my illustrious husband’s unique lofty ideals, a portion of my ashes be strewn in the flower beds around his ‘Samadhi’ at the Birbal Sahni Institute of Palaeobotany, Lucknow.
‘To my palaeobotanist colleagues the world over whom I hold in very high regard and affection, I offer my deep gratitude for their cooperation in the cause of science and towards fulfillment of our ideal, which has been a source of great encouragement to me. I express for them my unbounding affection and my all blessings follow them ever. I realize what great it is in life to have been given the great gift of receiving such unstinted affection from all over the palaeobotanical world, and I am beholden to them.”’
With such noble feelings, the late Padamsri (Mrs.) Savitri Sahni founded Birbal-Savitri Sahni Foundation with the following programs:
(1) International Research Collaborative Programmes for exchange of scientists between the Birbal Sahni Institute of Palaeobotany, Lucknow, and palaeobotanical organizations abroad; this would be known as the Birbal-Savitri Sahni Collaborative Research Programme.
(2) Birbal-Savitri International Fellowships would be awarded to young palaeobotanists/earth scientists for carrying out research in any specialized branch of palaeobotany in India.
(3) Birbal-Savitri Sahni International Awards would be given in alternative years to an outstanding scientist excelling in palaeobotanical and allied fields; awards would carry a cash prize of 25,000 rupees and a plaque in gold and silver. Nominations for such award may be forwarded to the Foundation between April 10th and April 26th every year.
(4) Savitri Sahni Samman, founded by friends and admirers of Mrs. Savitri Sahni for her dedication to the cause of palaeobotany, would carry a cash prize of 10,000 rupees, including a medal that is to be given annually on January 22nd to a palaeobotanist for outstanding research work.
(5) Savitri Sahni Smarak Lecture, instituted with cost donations from well-wishers of Mrs. Savitri Sahni, carries a token honorarium of 5,000 rupees for an invited lecture every
3
CHANDRA, ANIL
year on September 19th in any specialized field of palaeobotany. The lecture would be published as a monograph under the auspices of the Birbal-Savitri Sahni Foundation. These programs are meant to promote palaeobotanical and allied sciences the world over, for which Mrs. Savitri Sahni donated every bit of her belongings to the nation and entrusted to Birbal-Savitri Sahni Foundation. The entire residence of the Sahni’s, situated at the banks of the river Gomti, is planned to be converted into a museum-cum-guest house, and a ‘‘palaeogarden”’ is also being planned to develop at the site where her last remains were consigned to flames.
For further details one may contact: Dr. Shyam C. Srivastava, Secretary, Birbal-Savitri Sahni Foundation, 686 Birbal Sahni Marg, Post Bag No. 1, New Hyderabad Post Office, Lucknow 226 007, India.
11th DIATOM SYMPOSIUM 1990
Transmission Electron Microscopy and Diatom Research
by Richard M. Crawford
Alfred-Wegener-Institut fiir Polar und Meeresforschung, Postfach 120161, Columbusstrasse, D-2850 Bremerhaven, Germany
with 2 plates
Abstract: Many of the most important questions concerning diatom structure and biology have been answered using transmission electron microscopy. Some of the major advances are reviewed here. Unresolved problems are identified, and new questions are raised with a plea for a more balanced approach to the microscopy of diatoms.
Introduction
Throughout its history, diatom microscopy has been somewhat restricted in its field of interest. For the first 150 years, light microscopists concentrated on the morphology of the major components of the frustule—the valves—and neglected the ‘‘minor’’ components of the cingulum and the soft parts of the cells such as the nuclei and plastids. In the subsequent 45 years or more, electron microscopists have covered much the same ground, albeit in greater detail and with the same general purpose, i.e., to arrive at a classification that received common agreement and reflected as closely as possible the natural phylogeny of the group.
Furthermore, most published papers included information from scanning electron micros- copy (SEM) and relatively little from transmission electron microscopy (TEM). This is regrettable. Just as valuable information from light microscopy has been seriously over- looked (Cox 1981) so too has the abundance of morphological markers within the diatoms by all but a few workers. Such markers—manifestations of an explosive radiation of evolution—have been the source of important information on morphogenesis as they appear in strict sequence following the events of cytokinesis under the control of the nucleus.
The early electron microscopy of diatoms was, in fact, performed using TEM, and the first pore plates were seen in the remarkably high-quality micrographs of direct prepara- tions by Kolbe and Golz (1943). Hustedt (1945) followed, and then Desikachary (e.g., 1954), Helmcke and Krieger (1953-1977) and, during most of the 1950s, Okuno (see Okuno 1962 for a list of many of his works). Most of these studies used direct preparations, and sections were not obtained until the early 1960s.
CRAWFORD, RICHARD M.
These early pioneers would have had some concept of the wealth of information awaiting discovery but little idea that TEM studies would be overtaken, at least in terms of usage, by SEM. Paradoxically, the leap in potential resolution may have been too great for the peace of mind of many light microscopists (see, for example, the misgivings of Hendey 1959). Certainly the preparative techniques of SEM proved much simpler for the majority of us as we continued to collect comparative morphological data with the SEM.
Rather than tackle the formidable task of assessing the progress made in electron micros- copy as a whole, I will briefly focus on the considerable achievements of a small number of workers who have turned to some neglected questions, largely with TEM, and then I will attempt to establish what problems remain to be addressed.
(1) How is the cell wall morphology faithfully reproduced in successive generations?
This question is complex, and a number of different aspects are best considered separately.
THE ROLE OF THE NUCLEUS AND ASSOCIATED STRUCTURES IN CELL DIVISION AND NEW WALL FORMATION
Nuclear division precedes cell division and, as we know from light microscopy, is a prerequisite of half-frustule (valve + hypocingulum) formation. Indeed, nuclear migration occurs in many genera, and even though cell wall formation can begin before the nuclei are in position, the nucleus is generally found close to the developing valve. Few com- prehensive series of sections have been made, and fixations have often been poor, but several of the features associated with the nucleus seem to be directly involved in wall morphogenesis.
Some of the structures associated with the nucleus were observed in the classical light microscope studies on Surirella by Lauterborn as early as 1896, and this work has been beautifully corroborated by Pickett-Heaps et al. (1984) using TEM. Other studies using the TEM include those of Manton et al. (1969a, b; 1970a, b) on mitosis and meiosis in Lithodesmium, Drum and Pankratz (1964), again on Surirella; and Crawford (1973) on Melosira. More recently, detailed studies of mitosis and wall morphogenesis in a number of genera by Pickett-Heaps and co-workers have provided a much clearer picture for diatoms and perhaps a better understanding of the behavior of the mitotic spindle in general (Pickett-Heaps 1987; Pickett-Heaps et al. 1991). Furthermore, the structure and behavior of the spindle, microtubule center, and polar complex promises valuable phylogenetic information (e.g., Pickett-Heaps 1983) even if only some of these structures are involved in cell division and successive events. Nuclear migration is affected by microtubules under the control of the microtubule center (MC), which disappears at metaphase only to reappear later in mitosis. Its role in valve morphogenesis of centric diatoms is ‘‘less easy to estimate’’ (Pickett-Heaps et al. 1991), and Schnepf et al. (1980) demonstrated little effect of microtubule inhibitors on the formation of Atrheya frustules. Schmid (1980) and Blank and Sullivan (1983) have shown that the distribution of tube processes in some centrics and the development of the raphe in some pennates is upset but the general wall morphology is not. Nevertheless, the MC is closely involved with the SDV in pennate diatoms, for example, Nitzschia, Hantzschia, and Pinnularia. In the former two the MC moves with the SDV towards an off-center position or to the developing raphe. However, the MC is not concerned with movement of the SDV, because in Ach- nanthes the MC and subsequently the SDV moves away from the site of raphe development and the raphe slit is filled in (Boyle et al. 1984) The silica deposition vesicle itself is always present and is the focus of events throughout silica wall formation.
11th DIATOM SYMPOSIUM 1990
The functions of the microtubules also appear to differ. They are intimately associated with the developing raphe in pennates, more so in some genera than in others (Pickett- Heaps et al. 1991) and with developing setae in some centrics. The raphe fibre is another feature closely involved in controlling wall morphogenesis, although it is unclear whether it is done so to shape the valve as, for example, in Pinnularia (Pickett-Heaps et al. 1979), or to prevent silica deposition (Edgar and Pickett-Heaps 1984), or both (Pickett-Heaps et ala199ili):
Similar material to the raphe fibre has been found beneath developing rimoportulae of one species each of Ditylum, Chaetoceros, and Stephanopyxis and two species of Odontella by Li and Volcani (1984, 1985a, b, c,). Crawford and Schmid (1986) have proposed that such material may, in general, prevent silica deposition wherever it occurs. Pickett-Heaps et al. (1991) echo this proposal and illustrate similar granular material within the SDV and the lengthening hollow tube of the rimoportule in Ditylum as well as granular material subtending the fultoportule.
This area of research would surely repay further study with improved fixation techniques. Perhaps, for example, freeze substitution may show this material to be more widespread than we know at present. Both the labiate process (rimoportule) apparatus (LPA) of the centrics and the raphe fibre disappear on completion of that part of the valve. Li and Volcani (1985b) proposed that the rimoportule region is the primary silicification site for valve formation in centric diatoms. Crawford and Schmid (1986) objected to this proposal on the grounds of Li and Volcani’s (1985c) interpretation of events during morphogenesis of Odontella sinensis, but there are other reasons for caution. These include the large number of rimoportulae in the valves of some genera, for example, Podosira, Stictocyclus, and Coscinodiscus ; the complete lack of tube processes of any kind in some genera, for example, Chrysanthemodiscus, and the lack of a counterpart for the rimoportula as a focal point for the development of elements of the cingulum. As Pickett-Heaps et al. (1991) point out, the relationships of the rimoportule to some pattern centers or annuli (von Stosch 1977) as in Ditylum is coincidental.
The existence of a cytoskeleton that breaks down during meiosis and oogamy and reconstitutes after the auxospore of Thalassiosira eccentrica reaches maturity has been investigated by Schmid (1984a), and from reports of the many processes mentioned above it is clear that such an organisation is complex among diatoms. Further examination throughout the group is warranted.
THE ROLE OF CYTOPLASMIC VESICLES, MEMBRANES, ETC., IN NEW WALL FORMATION
The silica deposition vesicle (SDV) was discovered by Reimann and by Drum and Pankratz in 1964, and its membrane was given the name ‘‘silicalemma’’ in 1966 (Reimann et al. 1966). Stoermer et al. (1965) demonstrated close involvement of the SDV with the develop- ing cell wall. Also appearing at this time was the series of micrographs of thin sections by Drum et al. (1966), which still repay examination. All of these observations eventually paved the way for the work of Dawson (1973) and Chiappino and Volcani (1977) on pennate diatoms and Schnepf et al. (1980) and Schmid on centrics (1984a, b 198Sa, b). There have also been numerous recent contributions by Pickett-Heaps and co-workers that referred to the above.
Before, or just after, completion of cytokinesis a SDV appears beneath the plasmalemma of each daughter cell. Derived by the fusion of small vesicles (Schmid and Schulz 1979), which are presumed to be dictyosome derived, the vesicle expands centrifugally. In the pennates, this expansion is essentially horizontal (Chiappino and Volcani 1977; Pickett-
7
CRAWFORD, RICHARD M.
Heaps 1983), but development is asymmetric and manifest, for example, in Voigt discon- tinuities (Mann 1983). In the centric Coscinodiscus wailesii, Schmid (1986a, b) showed a complex horizontal development of threads of the SDV that circumscribe the areola before expanding vertically and then again horizontally to close off the loculae or chambers in the developing valve. By very careful control of the osmotic potential of the fixative, Schmid demonstrated the involvement of a number of types of vesicles and even the mitochondria, dictyosomes, and cytoskeletal elements in moulding the valve shape and the areolae size. Schmid (1984b) also proposed control on a finer scale by vesicles from the dictyosome that lie directly beneath the pore plates, but she also pointed out that the SDV does not have a moulding capacity, thereby encouraging us to consider two separate concepts here—namely, the silica deposition itself and the moulding of the structure.
Since the communication of the present paper, Pickett-Heaps et al. (1991) have published their review of valve formation in which they distinguish between ‘‘membrane-mediated morphogenesis’’ and ‘‘macro-morphogenesis.’’ The former is brought about within the SDV. More careful studies like those of Schmid’s are needed before we can confidently explain the moulding and pattern reproduction. The more complex frustules of centric diatoms are likely to be more rewarding here, but the enigmatic question ‘‘How does the diatom reproduce its discrete pattern?’’ may prove elusive for some time. Part of the answer, of course, may come from investigation of the third aspect.
THE DEPOSITION OF THE SILICA AND ITS CONTROL
Even among those workers sectioning diatoms, few have examined the developing siliceous component at high magnification. Borowitzka and Volcani (1978) and Schmid and Schulz (1979) were the first to show that silica is added in the form of particles, but there is still uncertainty regarding the establishment of the fundamental pattern of the valve (Pick- ett-Heaps et al. 1991). Polysaccharides and proteins have been separately proposed for a template for the pattern.
Somewhat later in the process there appears to be variation in the way the valve is built up. For example, Li and Volcani (1984, 1985a) reported a change in Ditylum brightwellii from a smooth basal layer through microfibrillar addition to one of hexagonal columns. On the inside surface however, hexagonal columns are absent. In Odontella species, only the basal layer and microfibrillar form is found. In most of the centric diatoms (Ditylum may be an exception here; see Li and Volcani 1984, Figs. 11a, b) addition of the silica occurred on the biological outside of the wall component in direct contrast to the situation in Amphipleura pellucida (Stoermer et al. 1965), in Navicula cuspidata (Edgar and Pick- ett-Heaps 1984), and in Pinnularia (Pickett-Heaps et al. 1979). There the silica is added on both sides of a kind of middle lamella. In Odontella (Crawford and Schmid 1986) and Striatella (Roth and De Francisco 1977), development appears to be only towards the biological inside of the valve, but more species need to be investigated before any firm statements are made. Just as importantly, the araphid group must also be examined.
In all developing valves examined to date, completion is recognized by the smooth ap- pearance of the silica on all surfaces. This leads one to ask three further questions. How is the process started? How is the deposition controlled? And how is completion achieved? The first question is likely to be the easiest to answer insofar as the process will ultimately have been triggered by nuclear division leading to the production of the SDV and the supply of Si and other wall components. The deposition itself may either be controlled by the time for which the process is allowed to continue, or by means of its rate (Pick- ett-Heaps et al. 1979). Intuitively, one would argue that the simpler would be through
11th DIATOM SYMPOSIUM 1990
maintaining a constant rate of deposition that was allowed to continue longer in thicker parts of the frustule. Control of the rate is likely to be through the organic template or matrix, which is considered by several workers to exist in the SDV (e.g., Nakajima and Volcani 1969), but Lee and Li suggest (pers. comm.) that the rate is controlled by the supply of membrane, which is itself determined by the production of glycoproteins.
Concerning the third question; clearly part of the answer is through genetic signals passed through the membrane system. Schmid has observed a switch in dictyosome production from silica transport vesicles to dark vesicles that she maintains may provide organic material to prevent further deposition. There are several reports from diatoms and other silica metabolising plants and animals that this organic material provides protection against dissolution. However, since the face of the siliceous element is rough during development (see above), a smoothing of the silica has to occur before this organic ‘‘sealing’’ process takes place in order to achieve the completely even surface found in many electron micrographs. Compare, for example, Li and Volcani’s Figures 66 and 75 of Ditylum (Li and Volcani 1985a). It appears most likely that the smoothing of the frustule is brought about by control of the final stages of silicification before the switch of vesicle production, and it is the answer to how this is achieved that may prove the most elusive. It is difficult to imagine how a change from one vesicle type to another can bring all of the silica up to the same level before silicification ends.
THE FATE OF THE MEMBRANES INVOLVED IN CELL WALL FORMATION
This final step in wall morphogenesis has received attention from a number of workers (see references in Schulz 1984), but it was Schmid (1986a, b) who investigated this problem in depth and presented four possible models for the fate of these membranes. One of these (M3) seems unlikely (Schmid 1986a) as it proposes de novo synthesis of plasmalemma in the ‘‘new’’ half of the cell. Nevertheless, this strategy was favored by Schulz (1984). Two other models (M1, M2) involve fusion at the valve edge of the inner profile of the SDV in the ‘‘new”’ half of the cell with the plasmalemma in the ‘‘old”’ half. One of these models sees the outer profile of membrane, SDV, and plasmalemma lost; the other sees them incorporated as organic matrix into the valve. The fourth model, (M4, and that favored by Schmid), envisages reticular fusion of SDV and plasmalemma at the center outside of the valve. Some membrane is lost to the outside, but retraction occurs at the valve edge and some of the membrane on the outside during maturation of the valve (plasmalemma and distal silicalemma) is internalized. At the same time, the incorporation of dense vesicles to the proximal silicalemma causes it to become plas- malemma-like. Schmid (1986b) calculates the area of membrane saved by this method and addresses the problems of timing of valve release and incorporation of organic matter into the frustule. It seems inconceivable to the author that much variation would exist in this fundamental process throughout the diatoms, but more work is called for to establish that this so.
(2) What is the function of the different cell wall components?
This question has seldom been asked, if at all, by light microscopists and only indirectly by electron microscopists. As the organelles such as rimoportulae, fultoportulae, ocelli, pseudocelli, rays, conopea, and partecta have been described, so possible functions have been proposed for them. However, we are still a long way from a complete explanation, and much needs to be done here, but I wish to confine my remarks to the two major wall components: the girdle bands and the valves.
CRAWFORD, RICHARD M.
The bands of the cingulum have been relatively neglected compared to the valves, but their function can be considered in the light of proposals that they, like the valves them- selves, originated from siliceous scales (Round and Crawford 1981). The more obvious functions are, firstly, to ensure protection for the developing protoplast as the two parent valves pull apart from each other prior to cytokinesis and, secondly, to allow penetration of the sperm in oogamous species (Drebes 1977). The third possible function is to provide the means by which the dimensions of the cells can be controlled through variation of the number of components in the cingulum. This has obvious significance in the planktonic situation and has parallels in the contro] of filament length through alteration of the separa- tion valve index in Aulacoseira granulata (Davey and Crawford 1986). Clearly, the greater the proportion of the frustule that the cingulum assumes, the more important it will be that the structure of the areolae resemble those of the valves. It is notable that in some genera where the cingulum is a relatively major component, similar vela have been shown by them. Good examples may be found in the Rhizosoleniaceae (Sundstrom 1986) in Eucampia and in Acanthoceras (Round et al. 1990). However, there are other genera where this is not obviously the case, for example, Ditylum and Dactyliosolen (Round et al. 1990).
As I have indicated above, we have come to regard the valves as the important part of the frustule simply because they are generally the largest. They may be the largest simply because they are polar (see Round and Crawford 1981; Mann and Marchant 1989 for evolutionary considerations), but perhaps we should pay greater attention to the cingulum and take account of its role in diatom biology.
That the valve itself provides a robust envelope that presumably deters at least some grazing predators while permitting light to penetrate to the plastids is usually taken for granted as is the fortuitous fact of nature for transmission electron microscopists that silica is opaque to electrons while being transparent to light. A function that has been investigated by a number of authors (Ross et al. 1977; Crawford 1979; Fryxell and Medlin 1981) is the provision of linking mechanisms on the outside of the valves in those species that do not exist as unicells, but this has been investigated more through SEM studies. The work of Schmid concerning the inner surface is an exception. Using a number of techniques, she has proposed that the inner apertures of the fultoportulae act as anchorage points for the cytoskeletal system (Schmid 1984a).
Prior to the publication of the first TEM micrographs of pore plates, most concepts of the areolae saw them as more or less simple holes in the frustule. We now know that there are indeed such holes in some genera, for example, Fragilaria (Crawford et al. 1985), Corethron (Crawford and Round 1989), and occasional simple pores scattered among the areolae in a number of other genera, for example, Eucampia (Round et al. 1990). True vela are, however, chiefly covered by plates of silica, the cribra, which may themselves have tiny pores or may only have areas of extremely weak silicification as Schmid and Schultz (1979) point out for Thalassiosira. Our investigations of a number of diatoms suggest the situation may be more complex.
Thin sections through the valve of the marine centric Actinocyclus subtilis (Andersen et al. 1986) have revealed an elaborate loculate valve with very fine pores on the outside surface that open inwardly, eventually, to a large chamber. This chamber is closed on the inner surface by fine siliceous projections that are supported by organic material as shown by sections treated with dilute hydrogen fluoride, which dissolves silica but leaves organic matter intact. A similar component was illustrated by Drum et al. (1966) in Aulacoseira (Melosira) granulata, and is shown here to be siliceous by hydrogen fluoride treatment. (Figs. 1-6). Interestingly, the organic layer extends under the whole of the mantle surface,
10
11th DIATOM SYMPOSIUM 1990
and siliceous projections extend upwards generally rather than to individual areolae. This component has also been found in the related species Aulacoseira ambigua, but what is interesting about both genera is that this organic/siliceous complex only occurs beneath porous areas of the valve (between arrowheads in Fig. 3), that is, beneath each areola in Actinocyclus but beneath the entire valve mantle in Aulacoseira. There are, no doubt, good physiological reasons why this complex is necessary, but if so, then we should ask whether it is present in many other diatoms. We should also ask how much information of comparative morphological importance have we been missing by acid cleaning, which gives the type of picture of the inside surface of the valve shown in Fig. 2? Similar questions may also be asked of the diatotepum—a wholly organic component reported from a number of diatoms by von Stosch (1981, but see also Schulz et al. 1984), supported by thin sections, and illustrated here from Odontella and Triceratium. In Odontella there are several layers of what appear to be helicoidally arranged fibrils that recall in some respects the micrographs of Subsilicea by von Stosch and Reimann (1970), whereas in Triceratium there is a single layer with perforations beneath the areolae. These perforations resemble those beneath the ocellus in Odontella (von Stosch 1981, Fig. 19). Whether such perforations affect the passage of materials through the wall or not has yet to be established, and we do not yet know how widespread they are taxonomically. Nonetheless, these two examples reflect some of the variation existing in the diatotepum, particularly in centric diatoms. More thin sectioning is required here to obtain a fuller picture, but TEM studies have shown that more circumspect cleaning procedures may ensure more information from the SEM where, in the past, it has literally disappeared down the drain.
(3) How does the cell move?
After the problem of valve pattern, this is perhaps the most frequently posed and the most difficult to answer. However, with the elegant work of Edgar and others before her (Drum and Pankratz 1965; Hopkins and Drum 1966; Drum 1967), we may have come as close to a full explanation as we shall for some time. A combination of histochemical staining, fluorescence microscopy, and thin section work has demonstrated that in Navicula cuspidata actin filaments control the movement of mucopolysaccharide filaments that are produced by secretory vesicles along the length of the raphe slit. The cell is thus moved in the opposite direction (Edgar and Pickett-Heaps 1982, 1983; Edgar and Zavortink 1983). Similar filaments have been shown in other pennate diatoms, and Pickett-Heaps et al. (1991) propose a more general role for them in valve morphogenesis. Nevertheless, it remains to be shown how the fine control needed for change of direction and the high degree of synchrony so beautifully shown by Bacillaria paxillifer can be achieved, but an answer is more likely to come from physiological experiments than from microscopy.
(4) What are the phylogentic relationships of the major groups?
The phylogenetic relationships of the centrics, the araphid pennates, and the raphid pen- nates have been determined throughout the history of microscopy, but recent TEM studies of morphogenesis have made, and will continue to make, a significant contribution (Pick- ett-Heaps et al. 1991). For example, the position of the monoraphid group has been fun- damentally changed from an exclusive position to one within the biraphid families in a recent classification based largely on studies of the developing ‘‘raphe’’ in the non-raphid valve mentioned above (Round et al. 1990). Much more needs to be done, but a major unanswered question concerns the position of the araphid group. Meaningful data is likely to come from molecular biology studies, but morphogenetic work is still needed on the genera.
CRAWFORD, RICHARD M.
(5) How has the diatom group as a whole evolved?
This final question has been partly answered through TEM work. The group is certainly monophyletic. Virtually all individuals replace a siliceous valve and a series of girdle bands at each cell division. They possess a unique method of locomotion in raphid genera and, astonishingly, the flagellum of the male gamete has been found to lack the central pair of microtubules in two genera, Lithodesmium (Manton and von Stosch 1966) and Biddulphia (Heath and Darley 1972). These findings need verification in other genera to establish how widespread the feature is within the centrics (motile gametes are absent from the raphid pennates). The proposals of Round and Crawford (1981), largely supported but significantly modified by Mann and Marchant (1989), that the diatoms evolved from scaly ancestors leaned heavily on the similarity of auxospore scales of a number of centric species to the simple structure of ribs radiating from an annulus in some developing centric valves. More thin sections of auxopores of both centrics and pennates will help answer question four, but other aspects of this question are more likely to come from molecular genetics.
In conclusion, despite the relatively small numbers of workers in the field, transmission electron microscopy and particularly thin sectioning has allowed us to make considerable advances and answer many of the questions posed above. Nevertheless, some questions remain, and new ones have been raised, but these are less likely to be resolved using one technique alone. There is clearly a greater need than ever for a holistic approach that combines whatever microscope technique may be needed with those of physiology, biochemistry, molecular genetics, and even those of ecology.
Acknowledgements
I am grateful to Rebecca Linstead for several of the micrographs from her investigation of Aulacoseira. This is AWI publication No. 460.
Literature Cited
ANDERSEN, R. A., L. K. MEDLIN AND R.M. CRAWFORD. 1986. An investigation of the cell wall components of Actinocyclus subtilis (Bacillariophyceae). J. Phycol. 22:466— 484.
BLANK, G.S. AND C. W. SULLIVAN. 1983. Diatom mineralisation of silicic acid. VII. Influence of microtubule drugs on symmetry and pattern formation in valves of Navicula saprophila during morphogenesis. J. Phycol. 19:294-301.
BOROWITZKA, M. A. AND B. E. VOLCANI. 1978. The polymorphic diatom Phaeo-dac- tylum tricornutum: ultrastructure of its morphotypes. J. Phycol. 14:101-121.
BoyLe, J.A., J.D. PICKETT-HEAPS, AND D.B. CZARNECKI. 1984. Valve mor- phogenesis in the pennate diatom Achnanthes coarctata. J. Phycol. 20:563-573.
CHIAPPINO, M. L. ANDB. E. VOLCANI. 1977. Studies on the biochemistry and fine struc- ture of silica shell formation in diatoms. VII. Sequential cell wall development in the pennate Navicula pelliculosa. Protoplasma 93:205—221.
Cox, E.J. 1981. The use of chloroplasts and other features of the living cell in the taxonomy of naviculoid diatoms. Pp. 115-133 in Proceedings of the sixth symposium on recent and fossil diatoms, R. Ross, ed. O. Koeltz, Koenigstein.
11th DIATOM SYMPOSIUM 1990
CRAWFORD, R.M. 1973. The protoplasmic ultrastructure of the vegetative cell of Melosira varians C. A. Agardh. J. Phycol. 9:50-61.
1979. Filament formation in the diatom genera Melosira C.A. Agardh and Paralia Heiberg. Nova Hedwigia. Beih. 64:121-133.
CRAWFORD, R. M., H. M. CANTER, AND G.H.M. JAWORSKI. 1985. A study of two morphological variants of the diatom Fragilaria crotonensis Kitton using electron micros- copy. Ann. Bot. 55:473-485.
CRAWFORD, R. M. AND F. E. ROUND. 1989. Corethron and Mallomonas—some striking morphological similarities. Pp. 295-305 in The Chromophyte Algae, problems and perspec- tives, B. S. C. Leadbeater, J.C. Green and W. A. Diver, eds. Oxford University Press.
CRAWFORD, R.M. AND A-M. SCHMID. 1986. Ultrastructure of silica deposition in diatoms. Pp. 291-314 in Biomineralisation in lower plants and animals. Systematics As- sociation Special Volume No. 30, B. S. C. Leadbeater and R. Riding, eds. Oxford Univer- sity Press.
DAVEY, M. C. AND R. M. CRAWFORD. 1986. Filament formation in the diatom Melosira granulata. J. Phycol. 22:144—150.
Dawson, P. A. 1973. Observations on the structure of some forms of Gomphonema parvulum Kiitz. IIL. Frustule formation. J. Phycol. 9:353-365.
DESIKACHARY, T. V. 1954. Electron microscope study of diatom wall structure. V. Trans. Amer. Microsc. Soc. 73:274-277.
DREBES, G. 1977. Sexuality. Pp. 250-283 in The Biology of Diatoms, D. Werner, ed. Blackwell, Oxford.
DRUM, R. W. 1967. Motility in Algae. III (abstract). J.Phycol. 3 (suppl.).
Drum, R. W. AND H. S. PANKRATZ. 1964. Post mitotic fine structure of Gomphonema parvulum. J. Ultrastuct. Res. 10:217-223.
. 1965. Locomotion and raphe structure of the diatom Bacillaria. Nova Hed- wigia 10:315-317.
DRuM, R. W., H. S. PANKRATZ, AND E. F. STOERMER. 1966. Electron microscopy of diatom cells. Jn Diatomeenschalen im Elektronenmikroskopischen Bild. Teil VI, J—G. Helmcke and W. Krieger, eds. Cramer, Lehre.
EpGaR, L. A. AND J. D. PICKETT-HEAPS. 1982. Ultrastructural localisation of polysac- charides in the motile diatom Navicula cuspidata. Protoplasma 113:10-22.
1983. The mechanism of diatom locomotion. I. An ultrastructural study of the motility apparatus. Proc. R. Soc. London B, 218:331-343.
. 1984. Valve morphogenesis in the pennate diatom Navicula cuspidata. J. Phycol. 20:47-61.
EDGAR, L. A. AND M. ZAVORTINK. 1983. The mechanism of diatom locomotion II. Identification of actin. Proc. R. Soc. London B, 218:345-348.
FRYXELL, G. A. AND L. K. MEDLIN. 1981, Chain—forming diatoms: evidence of parallel evolution in Chaetoceros. Crypt. Algol. 2:3-29.
HEATH, I. B. AND W. M. DARLEY. 1972. Observations on the ultrastructure of the male gamete of Biddulphia levis Ehrenberg J. Phycol. 8:51—59.
13
CRAWFORD, RICHARD M.
HELMCKE, J—G. AND W. KRIEGER. 1953-1977. Diatomeenschalen im elektronenmik- roskopischen Bild. Parts 1-10. Cramer, Vaduz.
HENDEY, N.I. 1959. The structure of the diatom cell wall as revealed by the electron microscope. J. Quekett Microsc. Soc. 5:147-175.
HOPKINS, J.T. ANDR.W.DRUM. 1966. Diatom motility: an explanation and a problem. Br. phycol. Bull. 3:63-67.
HustTebT, F. 1945. Die Struktur der Diatomeen und des Elektronenmikroskops fur ihre Analyse. Archiv. f. Hydrobiologie. 41:315—332.
KOLBE, R. W. AND E. GOLz. 1943. Elektronenmikroskopische Diatomeenstudien. Ber. d. Deutsch. Bot. Ges. 61:91-98.
LAUTERBORN, R. 1896. Untersuchungen ber Bau, Kernteilung und Bewegung der Diatomeen. W. Engelmann, Leipzig. 165 pp.
Li, C-W. AND B.E. VOLCANI. 1984. Aspects of silicification in wall morphogenesis of diatoms. Phil. Trans. Roy. Soc. Lond., B 304:519-528.
1985a. Studies on the biochemistry and fine structure of silica shell formation in diatoms. VIII. Morphogenesis of the cell wall in a centric diatom Ditylum brightwellii. Protoplasma 124:10-29.
1985b. Studies on the biochemistry and fine structure of silica shell formation in diatoms. IX. Sequential valve formation in a centric diatom Chaetoceros rostratum. Protoplasma 124:30-41.
1985c. Studies on the biochemistry and fine structure of silica shell formation in diatoms. X. Morphogenesis of the labiate process in centric diatoms. Protoplasma 124:147-156.
MANN, D.G. 1983. Symmetry and cell division in raphid diatoms. Ann. Bot. 52:573- 581.
MANN, D. G. AND H. J. MARCHANT. 1989. The origins of the diatom and its life cycle. In The Chromophyte Algae: problems and perspectives, B. S.C. Leadbeater, J.C. Green and W. A. Diver, eds. Oxford University Press, Oxford.
MANTON, I. AND H. A. VON STOSCH. 1966. Observations on the fine structure of the male gamete of the marine centric diatom Lithodesmium undulatum. J. Roy. Microsc. Soc. 85:119-134.
MANTON, I., K. KOWALLIK, AND H. A. VON STOSCH. 1969a. Observations on the fine structure and development of the spindle at mitosis and meiosis in a marine centric diatom (Lithodesmium undulatum) I. Preliminary survey of mitosis in spermatogonia. J. Microsc. 89:295—320.
1969b. Observations on the fine structure and development of the spindle at mitosis and meiosis in a marine centric diatom (Lithodesmium undulatum) I. The early meiotic stages in spermatogenesis J. Cell Sci. 5:271-298.
1970a. Observations on the fine structure and development of the spindles at mitosis and meiosis in a marine centric diatom (Lithodesmium undulatum) Il. The later stages of meiosis I in male gametogenesis. J. Cell Sci. 6:131-157.
11th DIATOM SYMPOSIUM 1990
1970b. Observations on the fine structure and development of the spindle at mitosis and meiosis in a marine centric diatom (Lithodesmium undulatum). TV. The second meiotic division and conclusion. J. Cell Sci. 7:407-443.
NAKAJIMA, T. AND B.E. F. VOLCANI. 1969. 3,4-Dihydroxyproline: a new amino acid in diatom cell walls. Science 164:1400-1401.
OKUNO, H. 1962. Electron microscopical study on fine structures of diatom frustules. XIX. Bot. Mag. Tokyo 75:119-126.
PICKETT-HEapsS, J.D. 1983. Valve morphogenesis and the microtubule centre in three species of the diatom Nitzschia. J. Phycol. 19:269-281.
1987. Diatom mitosis: Implications of a model system. /n Algal Development (Molecular and Cellular Aspects), W. Wiessner, D.G. Robinson, and R.C. Starr, eds. Springer-Verlag, Heidelberg.
PICKETT-HEaps, J.D., A-M. M. SCHMID, AND L. A. EDGAR. 1991. The cell biology of diatom valve formation. /7 Progress in Phycological Research Vol. 7, F. E. Round and D. J. Chapman, eds. Biopress, Bristol.
PICKETT-HEaps, J.D., A-M.M. SCHMID, AND D.H. Tippit. 1984. Cell division in diatoms. A translation of part of Lauterborn’s Treatise of 1896 with some modern con- firmatory observations. Protoplasma 120:132-154.
PICKETT-HEapsS, J. D., D. H. TipPIT, AND J. A. ANDREOZZI. 1979. Cell division in the pennate diatom Pinnularia. IV. Valve morphogenesis. Biol. Cellulaire 35:199-206.
PICKETT-HEAPS, J. D., R. WETHERBEE, AND D.R. A. HILL. 1988. Cell division and morphogenesis of the labiate process in the centric diatom Ditylum brightwellii. Protoplas- ma 143:139-149.
REIMANN, B.E.F. 1964. Deposition of silica inside a diatom cell. Exp. Cell Res. 34:605-608.
Ross, R., P. A. SIMS, AND G.R. HASLE. 1977. Observations on some species of the Hemiauloideae. Nova Hedwigia, Beih. 54:179-213.
RoTtu, L. E. AND A. DE FRANCISCO. 1977. The marine diatom Striatella unipunctata. II. Siliceous structures and the formation of intercalary bands. Cytobiologie 14:207-221.
ROUND, F.E. AND R.M. CRAWFORD. 1981. The lines of evolution of the Bacil- lariophyta. I. Origin. Proc. Roy. Soc. London B 211:237-260.
ROUND, F. E., R. M. CRAWFORD, AND D.G. MANN. 1990. The Diatoms, Biology and Morphology of the Genera. Cambridge University Press. 747 pp.
SCHMID, A-M.M. 1980. Valve morphogenesis in diatoms: a pattern-related filamentous system in pennates and the effect of APM, cholchicine and osmotic pressure. Nova Hed- wigia 33:811-847.
1984a. Wall morphogenesis in Thalassiosira eccentrica: comparison of auxospore formation and the effect of MT inhibitors. Pp. 47-70 in Proceedings of the 7th International Diatom Symposium, D.G. Mann, ed. O. Koeltz, Koenigstein.
1984b. Schalenmorphogenese in Diatomeen in Communications of Instit. Lightweight Structures. 28:300-317.
CRAWFORD, RICHARD M.
1986a. Organization and function of cell structures in diatoms and their mor- phogenesis. Pp. 271-292 in Proceedings of the 8th International Diatom Symposium, M. Ricard, ed. O. Koeltz, Koenigstein.
1986b. Wall morphogenesis in Coscinodiscus wailesii Gran et Angst. II Cytoplasmic events of valve morphogenesis. Pp. 293-314 in Proceedings of the 8th In- ternational Diatom Symposium, M. Ricard, O. Koeltz, Koenigstein.
SCHMID, A-M. M. AND D. SCHULTZ. 1979. Wall morphogenesis in diatoms: deposition of silica by cytoplasmic vesicles. Protoplasma 100:267-288.
SCHNEPF, E., G. DEICHGRABER, AND G. DREBES. 1980. Morphogenetic processes in Attheya decora (Bacillariophyceae, Biddulphiineae). Pl. Syst. Evol. 135:265-277.
SCHULZ, D. 1984. Das Schicksal von Silicalemma und Plasmalemma wahrend der Zellwandbildung bei Coscinodiscus granii Gough und Thalassiosira eccentrica (Ehrenb.) Cleve. Ber. d. Deutsch. Bot. Ges. 97:411-419.
SCHULZ, D., G. DREBES, H. LEHMANN, AND R. JANK-LODWIG. 1984. Ultrastucture of Anaulus creticus Drebes & Schultz with special reference to its reduced ocelli. Eur. J. Cell Biol. 33:43-S1.
STOERMER, E. F., H.S. PANKRATZ, AND C.C. BOWEN. 1965. Fine structure of the diatom Amphipleura pellucida. 11. Cytoplasmic fine structure and frustule formation. Am. J. Bot. 52:1067-1078.
StoscH, H. A. VON. 1977. Observations on Bellerochea and Streptotheca including descriptions of three new planktonic diatom species. Nova Hedwigia, Beih. 54:113-166.
. 1981. Structural and histochemical observations on the organic layers of the diatom cell wall. Pp. 231-252 in Proceedings of the 6th Symposium on Recent and Fossil Diatoms, R. Ross, ed. O. Koeltz, Koenigstein.
StToscH, H. A. VON AND B.E. F. REIMANN. 1970. Subsilicea fragilarioides gen. et sp. nov., eine Diatomee (Fragilariaceae) mit vorwiegend organischer Membran. Nova Hed- wigia, Beih. 31:1-36.
SUNDSTROM, B.G. 1984. The marine diatom genus Rhizosolenia. A new approach to the taxonomy. Doctoral dissertation (Lund BS/NBB-1008). 196 pp.
11th DIATOM SYMPOSIUM 1990
Explanation of Plates
Plate 1, Figs. 1-4
FIGURES 1-4. Aulacoseira granulata. FIGURE 1. Separation valve illustrating straight lines of pores on mantle and a non-porous valve face. x 3800. FIGURE 2. Inner surface of two sibling linking valves showing curved rows of pores. Arrows indicate internal apertures of two rimoportulae. Note clear image of areolae. See text for further explanation. x 3300. FIGURE 3. Thin section through mantle. Organo-silica complex extends from the ‘‘Ringleiste’’ (R) to the corner of the valve at the edge of the valve face, between the arrowheads. x 17,000. FIGURE 4. Section of interfacing sibling valves. Note point at which the complex fuses with valve face (arrows). x 10,000.
Plate 2, Figs. 5-8
FIGURES 5, 6. Aulacoseira granulata. FIGURE 5. Detail of valve mantle overlain above by two series of girdle bands. Note fine branches of silica extending from the organic layer (arrow) to the pore plates (vela) of the areolae (arrowhead). There are also indications of silica and organic material outside the vela (double arrow). x 43,000. FIG- URE 6. Similar section to Figure 5 but treated with dilute HF. x 48,000. FIGURE. 7. Odontella sp. Thin section of valve with diatotepum consisting of at least four layers of microfibrillar material apparently organized in a helicoidal arrangement. x 50,000. FIG- URE. 8. Section of valve of Triceratium sp. Diatotepum extends underneath the valve surface but is thickest beneath areolae where pores are regularly found (arrows). x 41,000.
CRAWFORD, RICHARD M.
11th DIATOM SYMPOSIUM 1990
19
11th DIATOM SYMPOSIUM 1990
Diatoms, Or Their Chloroplasts, As Endosymbiotic Partners For Foraminifera
by John J. Lee
Department of Biology, City College of CUNY, Convent Avenue and 138th Street, New York, New York 10031
and
Department of Invertebrates, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024
with 5 plates
Abstract: Several families of larger foraminifera have representatives that harbor diatoms or the chloroplasts of diatoms as endosymbionts. The diatoms involved in these associations, all pennate taxa, are identified. Factors affecting the distributions of these symbiotic species and host-symbiont interactions are described. Foramineriferal symbioses with diatoms are compared to those involving other algal taxa.
Introduction
Foraminifera are a major group (220 families and 40,000, possibly as many as 60,000 fossil-described species but only 4,000 living species) of marine amoebae characterized by their distinctive granular reticulopodia (Fig. 1). Most have multichambered agglutinated or calcareous tests. Throughout their long evolutionary history (Cambrian to present) there have been families (~ 40) of ordinary sized foraminifera (SO-500 [im) that have given rise to much larger descendants (1 mm-—1l cm) that are given the general name larger foraminifera. The evolutionary occurrence of these protistan giants was not random but generally corresponded to polytaxic episodes in the geologic record (see Lee and Hallock 1987, and references cited therein), which were periods of global warming, reduced oceanic circulation, raised sea levels, and diminished planktonic productivity as a consequence of reduced nutrient input into the sea.
Because all the modern species of larger foraminifera are hosts for endosymbiotic algae, and because they are found in shallow tropical and semitropical seas where algal en- dosymbiosis is a common phenomenon (e.g., corals, giant clams, sea anemones), we infer that symbiosis was the driving force in the evolution of larger foraminifera (Lee and Hallock 1987; Lee et al. 1979a). The evidence to extend this hypothesis from the present
21
LEE, JOHN J.
to extinct groups may be very difficult to obtain. While the tests of host foraminifera preserve well in the fossil record, their organic contents usually do not. Sporadic quick burial conditions during the fossilization procedure may hold the key to this question. A thin section and small samples of a phosphoric chert contained the remains of a paleozoic fusulinid larger foraminifer, Pseudoschwagerina montanensis, and hollow spherical remains that might have endosymbiotic algae (Lee and Hallock 1987). The whole question of which algal group (or groups) was (were) endosymbiotic in the extinct Paleozoic (Pen- nsylvanian -> Permian) fusulinids is intriguing.
Modern larger foraminifera as a group are the hosts for a wide diversity of algal types (reviewed in Lee and McEnery 1983, Lee and Anderson 1991). One family, Archaiadae, has species that are the hosts for chlorophytes (Chlamydomonas spp.). The foraminifera in another family, Peneroplidae, have endosymbiotic rhodophytes (Porphridium pur- pureum), and foraminifera in the family Soritidae bear endosymbiotic dinoflagellates (Sym- biodinium spp. and Amphidinium sp.). Two families, Alveolinidae and Calcarinadae, which arose in the Cretaceous, and two other families Nummulitidae and Amphisteginidae, which arose in the Paleocene-Eocene, are hosts for endosymbiotic diatoms.
DIATOMS AS ENDOSYMBIONTS
The individual diatom endosymbionts are located just below the expanded pore-rims in perforate foraminifera (Leutenegger 1977, 1983; Hansen and Burchardt 1977; McEnery and Lee 1981, Figs. 2-5). Several experiments, one using the vital dye neutral red, and the other, H'4CO3, have indicated that the pores are physiologically active even though they have an organic lining (Berthold 1976; Leutenegger and Hansen 1979).
Although fine structural observations made as long ago as 197] (Deitz-Elbrachter 1971) showed that diatoms were endosymbionts in some species of larger foraminifera, they could not be identified. Endosymbiotic diatoms within their hosts, in common with other foraminiferal endosymbionts, are naked protoplasts with highly reduced cell envelopes. Subsequent fine structural studies of diatom-bearing larger foraminifera confirmed the original observation that diatoms were endosymbionts and found some diversity in pyrenoid structure that suggested there might be more than one species of diatom involved in the symbiotic phenomenon (Leutenegger 1977, 1983, 1984; Hansen and Burchardt 1977; Berthold 1978; Schmalljohann and Rottger 1978). The abundance of larger foraminifera in modern seas and their contribution to deposition of CaCO3 is not generally appreciated. On Kudaka Jima (Okinawa prefecture, Japan), living ‘‘star sands’’ (Calcarinids; Figs. 11, 13) are produced at arate of 6 x 102g CaCO3 M~*Y™ (Sakai amd Nishihira 1981). Deposition rates are as high as kg M~“Y7 in close proximity to coral reef margins (Hallock 1981). Fortunately the diatom endosymbionts were isolated in culture using methods that worked well with salt marsh diatoms (Lee et al. 1975), and they formed identifiable frustules (Lee et al. 1979b; Lee et al. 1980a, b). Since diversity of diatom symbiont species in the same hosts was recognized from the first isolations, many questions were raised. Were the endosymbiotic diatoms sequestered from those encountered by their hosts during their grazing? If that was the case, there should be some reflection of the internal diatom population in the external diatom assemblage. Are the endosymbiotic diatoms adaptive for hosts living at different depths or different seasons? Over the course of the past decade more than 3,000 specimens of larger foraminifera have been collected, carefully and asep- tically brushed, washed, and then crushed, so that the diatom endosymbionts could be liberated, cultured, and identified. Results reported to date are from isolations from six species of larger foraminifera: Amphistegina lessonii, A. lobifera, Heterostegina depressa, Borelis schlumbergerii (Fig. 14), Operculina ammonoides (Fig. 12), and Calcarina calcar
D9.
11th DIATOM SYMPOSIUM 1990
(Fig. 11), mainly from the Gulf of Elat on the Red Sea, but also some specimens from the Indian Ocean, Hawaii, Palau, and the Great Barrier Reef (Lee et al. 1979, 1980a, b, 1989; Lee and Reimer 1983; Reimer and Lee 1984, 1988). Work on isolations from the Pacific ‘‘star sands’’ is still in progress.
Of 20 species of endosymbiotic diatoms isolated, all are small (< 10 im), pennate species belonging to the genera Fragilaria, Navicula, Nitzschia, Amphora, Achnanthes, Cocconeis, and Protokeelia (Figs. 6-10). All of the endosymbiotic species are rare in the natural communities in which the foraminifera feed or have never been found other than as symbionts (Lee et al. 1989). Half of the endosymbionts found were new species or varieties. One species, Navicula muscatinei, has a very unusual life cycle for a diatom (Lee and Xenophontos 1989). It is the only one described thus far that has an autospore stage (Fig. 15). Nitzschia frustulum var. symbiotica was the most commonly isolated endosymbiont; it was found in 33.7% of the six hosts examined (Lee et al. 1989). Also very common in five of the hosts were Nitzschia panduriformis var. continua (14.5%), Fragilaria shiloi (9.9%), Nitzschia laevis (8.7%), and Amphora roettgerii (6%).
At every depth sampled, significant numbers of hosts harbored N. frustulum var. sym- biotica, N. laevis, and N. panduriformis var. continua. In contrast, F. shiloi was rarely isolated from hosts at depths greater than 25 m. Achnanthes macenerae, Protokeelia hot- tingeri, and Amphora sp., on the other hand, were only recovered in deeper waters (25 m). As a general rule (71% of the isolations), individual foraminifera usually hosted only one species of diatom at a time. In about 25% of the isolations, a second species was found; rarely three or more were isolated from the same host. Some of the endosymbiotic diatom species tended to be more common in particular hosts rather than others. Nitzschia laevis and F. shiloi were the most common endosymbionts in Borelis schlumbergerit; N. laevis, and A. macenerae were most common in Operculina ammonoides, and N. frustulum var. symbiotica was the most common in the population of Calcarina calcar from the Indian Ocean. No seasonal correlations of hosts and symbionts were found (Lee et al. 1989).
Several experimental studies attempted to answer the same questions by using a different approach. Specimens of Amphistegina lessonii were rendered nearly aposymbiotic by in- cubating them in tissue culture flasks containing | x 10°M DCMU anchored in a plastic raft at a shallow depth (5 m) in the Gulf of Elat. After five days the hosts had bleached and aliquots were: (1) fixed for fine structural studies (Koestler et al. 1985), (2) used in primary production studies, or (3) distributed into flasks containing mixtures of different species of endosymbiotic or free-living diatoms (Lee et al. 1983, 1986). Each mixture contained two endosymbiotic species and one non-symbiotic species. Ten different com- binations of species were used. The flasks were then incubated in rafts anchored at 10-m and 20-m depth on the sea floor. The results suggested that some of the diatom species were selected (or were the most competitive) over others during the rebrowning of the hosts. None of the diatom species that were free-living isolates survived intracellularly in the foraminifera. By comparing the results of different combinations, Nitzschia valdestriata and N. laevis were the most successful endosymbionts and F. shiloi the least.
Since very few small pennate diatoms, out of a potential of several hundreds of small species found in the habitat of larger foraminifera, are involved in the endosymbiosis phenomenon, what are the ‘‘special properties’’ of these species that allow them to establish and maintain the relationship with their hosts? One of the more logical places to look for cell recognition properties is the cell envelope. Some initial progress along these lines has been reported (Lee et al. 1988a). Polyvalent antisera against the frustule fraction of each of three species of endosymbiotic diatoms, N. panduriformis, F. shiloi, and A. ten-
23
LEE, JOHN J.
nerima, were raised in rabbits. The antisera were then reacted with the same and other species of endosymbiotic diatoms, with free-living diatoms, and with the frustule-less diatoms from freshly crushed diatom-bearing hosts.
By reacting the antisera first with one species and then with a second species, it was possible to show that the antisera contained antibodies that fell into that groups: (1) those that seemed to react with all the diatoms tested, (2) those that reacted only with the endosymbiotic species tested, and (3) those that reacted only with the particular species that was used as the antigen to raise the antisera in the rabbit. Goat anti-rabbit antibody conjugated with FITC was used to visualize the results in an epifluorescent microscope. The qualitative evidence that demonstrates that the endosymbiotic diatoms share some common surface antigens only piques curiosity and demands more quantitative and detailed probing by molecular techniques.
Nutritional studies of a number of endosymbiotic clones of N. panduriformis, N. laevis, N. frustulum, N. valdestriata, A. tenerrima, Navicula reissii, and F. shiloi suggested that all of the isolates required thiamine because the clones failed to grow on the second or third transfer to media without it (Lee et al. 1986). Biotin stimulated the growth of six of the eight clones tested. Only one clone (N. frustulum) was stimulated by vitamin B)2. Concentrations of NO3 that supported optimal growth of the diatom clones spanned a range from 2 ttm (N. panduriformis) to 2 mM (N. valdestriata and F. shiloi), At the depths in the Gulf of Elat, where most of the foraminifera were collected, the concentration of NO3 rarely exceeds | tg at 1-1 (Levanon-Spanier et al. 1979), a condition suggesting that even the foraminifera that host the diatoms with the lowest NO3 requirements must sequester and tightly recycle bound nitrogen, or that the systems are constantly nitrogen limited. The requirements for PO4-3 gave a similar picture. Nitzschia valdestriata and A. tenerrima had the lowest requirements for optimal growth (1 um) and N. panduriformis, N. laevis, and N. reissii the highest (100 ttm). At the same depths the level of POg-3 rarely exceeds 0.3 pg at 17.
Field observations, field and laboratory experiments (Lee et al. 1980a, d; Zmiri et al. 1974) and growth experiments (Hallock 1981; Rottger 1972, 1976) of intact host/symbiont systems suggested that there were optimum ranges of light intensities and spectral qualities for growth, calcification, and primary production. Respirometry was used to test the responses of four species of diatom endosymbionts: N. laevis, N. panduriformis, N. valdestriata, and F. shiloi (Lee et al. 1982). The clones tested were photoinhibited at high light intensities and did well at moderately low light (175 u4W cm). Photocompensation points of these clones were at a light level that approximated 2% of the light measured at 1m in the spring at Elat. If the algae were free (that is, not inside the foraminiferal shell) photocom- pensation depth for the clones tested would be reached between 40-50 m.
The depth ranges of diatom-bearing larger foraminifera are quite interesting. Some species (e.g., Heterostegina depressa and Operculina ammonoides) seem to occupy wide ranges of depth (Leutenegger 1983). However, when this is examined more closely, the statement needs qualification. For example, at shallow depths H. depressa is found on the shady sides of tide pools (R6ttger 1972) but in the Gulf of Elat it is found in deeper water (30-40 m). Several simple behavioral studes have shown phototaxis in diatom-bearing larger foraminifera (Zmiri et al. 1974; Lee et al. 1980a). Amphistegina lessonii was phototaxic at photonic fluxes between 10!!-105 photons cm ~~ S! and unresponsive at lower light levels. The action spectrum for this response peaked near 500 nm (Zmiri et al 1974). Amphistegina lobifera was positively phototaxic at an incident illumination be- tween 0.1-I klx and negatively phototaxic at higher light levels (~4 klx) (Lee et al. 1980a). These phototaxic responses help explain the distribution of some species, for example
24
11th DIATOM SYMPOSIUM 1990
Operculina ammonoides, just below the surface of the sediment at 35—40 m in Elat and on the surface of the sediment when it is shaded by patches of Halophila stipulacea. Other diatom-bearing larger foraminifera seem to have more restricted depth ranges (reviewed in Leutengger 1983; Reiss and Hottinger 1984). Some diatom-bearing species (e.g., Baculogypsina sphaerulata and Calcarina calcar) are found only in shallow waters; other species (e.g., Heterocyclina tuberculata and Cyclopeus carpenteri) are only found in deeper ( 70 m) waters.
The growth rates of diatom-bearing species of larger foraminifera, which have been studied in the laboratory, are correlated with light (ROttger 1972, 1974; Muller 1978). None of the species tested will grow in the dark, even if they are well fed. Heterostegina depressa was the first of these diatom-bearing species to be studied in this respect (ROttger 1972, 1974). This species survives and grows well in the absence of any obvious concentration of food if it is incubated in the light. Both H. depressa and Amphistegina lessonii grew best at 600-800 an (~ 12-16 WE m® S"!) (Réottger et al. 1980). Slightly higher light levels (14-40 LE m2 S°!) gave the maximum growth rate for other Hawaiian clones of A. lessonii (Hallock et al. 1986). The light saeueH level for clones of A. gibbosa from the Caribbean was also between 14 and 40 LE mS". Populations grown at 6.4 and 3.9 wE mS"! had lighter, thinner, and flatter tests than those at higher light levels (Hallock et al. 1986).
Primary production, = measured by 4C oe in Megas te lessonii and i lobifera was 1.95 x 10° mg 4C hr! foraminifer “! and 2.9 x 10 mg!4C h'! foraminifer"! respectively (Hallock 1981; Muller 1978). The role of symbiont-fixed carbon in the overall carbon budgets of larger foraminifera seems to vary widely. As mentioned earlier, sym- biont-fixed carbon seems to satisfy the overall carbon budget of the diatom-bearing species Heterostegina depressa (Rottger 1972, 1974). Feeding by these species seems to supply fixed nitrogen, phosphorus, and, perhaps, micronutrients and vitamins to the foraminiferan/symbiont system. The same may also be true for Amphistegina spp. (Lee et al. 1988b; Kuile et al. 1987). This aspect of other species of diatom-bearing larger foraminifera is unknown. However, studies of symbiont-bearing planktonic foraminifera and dinoflagellate-bearing larger foraminifera show that feeding is the major source of fixed carbon for these types of associations (Bé et al. 1981; Lee and Bock 1976).
The transfer of symbiont-fixed carbon to their hosts has not been studied in detail. Fine structural studies of a number of species of chlorophyte-bearing and dinoflagellate-bearing hosts have shown starch grains packed in symbionts and distributed in the host’s cytoplasm (Leutenegger 1977, 1983; Miiller-Merz and Lee 1976). How they get from one to the other is inferential. They may be released from digested or autolyzed cells, or they may be exocytosed. With the exception of the red algal symbionts, all symbionts in the foraminifera are surrounded by their own cell membrane and a symbiosome (or symbiont vacuole). The starch would have to pass through both. Fine structural studies of Peneroplis and its endosymbiotic red alga, Porphyridium purpureum, seem to suggest that polysaccharide sheath fibrils are passed from the symbiont directly into the cytoplasm of the host (Lee 1990). Fine structure cannot give an easily grasped, quantitative, dynamic answer to the transfer question. Chromatographic techniques were used to separate and identify 4C tracer- labeled photosynthates in six foraminiferal/algal associations (Kremer et al. 1980). A large percentage of the label in the three diatom-bearing species tested, A. lobifera, A. lessonii, and H. depressa, was found in lipids (51, 3] and 33% respectively) and glycerol (6, 5 and 11%). Further work along these lines with careful separation of symbionts from host cytoplasm by differential centrifugation could be quite useful in characterizing path- ways of carbon flow. Host digestion of symbionts and/or symbiont autolysis are possible pathways of energy flow from the symbionts to the hosts that have not been carefully
25
LEE, JOHN J.
examined. Evidence for flow of carbon from C tracer-labeled food to the symbionts in Amphistegina lobifera and Amphisorus hemprichii was clearly demonstrated in radioautographs (Lee et al. 1988b).
Endosymbiots isolated from a variety of algal/invertebrate systems are stimulated to release metabolites in the presence of host homogenates (reviewed by Trench 1979). A sterile homogenate from crushed whole Amphistegina was also effective in stimulating the release of #C labeled photosynthate from the log phase axenic endosymbiotic diatom clones tested (N. valdestriata, N. laevis, N. panduriformis, N. frustulum var. symbiotica, A. tener- rima, F. shiloi, and Navicula hanseniana). The increase of release ranged from 190 to 9,000% (25-76% total carbon fixed, not corrected for respiration; Lee et al. 1984). The host homogenate also stimulated a control, a clone of Amphora sp. that was isolated as an epiphyte in a salt marsh. The authors (Lee et al. 1984) were quite cautious about interpreting their results without further separation and categorization of the active fac- tor(s). In varying degrees, the host homogenate had other effects on the growing and dividing diatoms; it interfered with formation of new frustules (Figs. 16-18). Fragilaria shiloi was the most affected species (Figs. 16, 18). New cells were protoplasts with little or no vestiges of frustules. Since the endosymbionts are all protoplasts in their hosts (Fig. 3) this seems a very promising topic for future study of host/symbiont interactions.
DIATOM CHLOROPLASTS AS ENDOSYMBIONTS
A review of the relationships between diatoms and foraminifera must consider the diatom- chloroplast husbanding (or sequestering) foraminifera (Figs. 19, 20). Some members (per- haps many, or all) of three families, Elphidiidae, Nonionidae and Rotaliellidae, are known to retain functional chloroplasts of some of the algae they partially consume. The nonionids and elphidiids are morphologically very complex. Their apertures, the principal opening from which most foraminiferal pseudopodia emerge, are very reduced. In the elphidiids the sutures between chambers have funnel-like openings, fossae, through which the pseu- dopods emerge (Fig. 19). The fossae are lined with tooth- or comb-like projections that seem to act like sieves. Microalgae trapped by the reticulopodial web are transported back to those sieved, depressed openings where digestion takes place (Lee et al. 1991). Here frustules remain as the chloroplasts are extracted from partially digested diatoms. The fossae are openings of an elaborate canal system that is connected to every chamber (diagram in Lee and Hallock 1987). The relationship (if there is any) of the canal systems in elphidids to the chloroplast husbandry phenomenon has not yet been rigorously studied. The number of chloroplasts husbanded by the nonionids and elphiidids is quite high. For example, several species, Elphidium williamsoni, E. excavatum, and Haynesina germanica, collected from the shallow waters of Limfjorden (Denmark), sequestered respectively 9.7 (+ 4.9) x 103, 1.2103, and 5.2 (+ 1.6) x 10%) chloroplasts per organism (Lopez 1979). Similar numbers have been reported for members of the same families collected at Plymouth, England; Falmouth, Massachusetts; Southampton, New York; Elat, Israel; and Mombasa, Kenya (Lee et al. 1988c; Lee and Lee 1990). Primary production calculated on an organic dry weight basis is as high in chloroplast-sequestering species as it is in larger foraminifera with whole diatoms (Lopez 1979; Lee et al. 1988b). For example, Elphidium crispum and Heterostegina depressa were both collected by hand in the same sample at 38m at Wadi Taba on the Red Sea. Specimens of E. crispum fixed 1.5 bg C mg foraminiferal dry-weight! 48 h"! while specimens of Heterostegina depressa, a diatom- bearing species, fixed slightly less (1.28 ug C mg foraminiferal dry-weight! 48 hr!) (Lee et al. 1988c).
Fine structural studies and photosynthetic pigment analyses have provided further infor- mation on the nature of chloroplast donors and the relative preservation state of the
26
11th DIATOM SYMPOSIUM 1990
chloroplasts and their pigments (Lopez 1979; Knight and Mantoura 1985; Lee et al. 1988c). The chloroplasts are in various states of preservation but most appear to be normal diatom chloroplasts surrounded by a host vacuole. HPLC and two-dimensional chromatography showed typical diatom pigments; chlorophylls a and c, fucoxanthin, diatoxanthin and, as could be expected, some phaeophytins a and c (Lopez 1979; Knight and Mantoura 1985). Several studies have attempted to measure the life of sequestered chloroplasts, factors that affect chloroplast longevity, and selectivity in chloroplast donors (Lopez 1979; Lee and Lee 1990). Results suggested that the phenomenon needs to be studied on a case-by- case basis. For example, Lopez (1979) calculated that under normal light/dark conditions E. williamson must eat at least 65 chloroplasts individual! h”! while H. germanica needed to consume only 20 individual"'h!. When juvenile H. germanica were fed a mixture of Navicula menisculus and Amphora tenerrima, they sequestered a steady state population of 80 chloroplasts in the light and half as many when incubated in the dark (Lee and Lee 1990). Some quality aspects of chloroplast retention have been shown in experiments in which only one algal species was available. The chloroplasts of Chlorella were sequestered at very low numbers by Elphidium crispum from Drake’s Island in Plymouth harbor, England. On the other hand, the number of chloroplasts sequestered by populations fed only Cocconeis placentula or Amphora sp. were approximately 160-180 individual! (Lee and Lee 1990). The initial chloroplast loss of starved elphidiids maintained in the light was rapid (Ty2 2 weeks) and then became more gradual (Ty2 5 weeks; Lee and Lee 1990).
While much is known about the diatom-bearing and diatom chloroplast sequestering foraminifera, in truth, almost every aspect of the phenomena needs deeper probing. Among the more interesting aspects I look forward to working on in the next few years are the following questions: Can we detect the differences between small pennate diatoms that are endosymbionts and those that are not? Are there any antigenic similarities between sequestered chloroplasts and endosymbiotic envelopes? What are the metabolites passed from the diatoms to their hosts? How are diatom population numbers regulated in their hosts? I hope others will join us in examining these interesting and unusual diatom-involved phenomena.
Literature Cited
BE, A. W.H., D. A. CARON, AND O. R. ANDERSON. 1981. Effects of feeding fre- quency on life processes of the planktonic foraminifer Globigerinoides sacculifer in laboratory culture. J. Mar. Biol. Assoc. U. K. 61:257-277.
BERTHOLD, W.U. 1976. Ultrastructure and function wall perforations in Patellina cor- rugata Williamson, Foraminifera. J. Foraminiferal Res. 6:22-29.
1978. Ultrastrukturanalyse der endoplasmatischen Algen von Amphistegina lessonii d’Orbigny, Foraminifera (Protozoa) und ihre systematische Stellung. Arch. f. Protistenk. 120:16-62.
DIETZ-ELBRACHTER, G. 1971. Untersuchungen tiber die Zooxanthellen der Fora-mini- fere Heterostegina depressa d’Orbigny 1826. Meteor Forsch. Ergeb. 6:41-47.
HALLOCK, P. 1981. Light dependence in Amphistegina. J. Foraminiferal Res. 11:40-46.
HALLOCK, P., L. B. FORWARD AND H. J. HANSEN. 1986. Influence of environment on the test shape of Amphistegina. J. Foraminiferal Res. 16:224-231.
HANSEN, H.J. AND B. BUCHARDT. 1977. Depth distribution of Amphistegina in the Gulf of Elat. Utrecht Micropaleont. Bull. 1:225-239.
LEE, JOHN J.
KNIGHT, R. AND R. F.C. MANTOURA. 1985. Chlorophyll a and carotenoid pigments in foraminifera and their symbiotic algae: analysis by high performance liquid chromatog- raphy. Mar. Ecol. Prog. Ser. 23:241-249.
KOESTLER, R.J., J.J. LEE, J. REIDY, R.P. SHERYLL AND X. XENOPHOTOS. 1985. Cytological investigation of digestion and re-establishment of symbiosis in the larger ben- thic foraminifer Amphistegina lessonii. Endocytobiosis Cell Res. 2:21—54.
KREMER, B. P., R. SCHMALJOHANN AND R. ROTTGER. 1980. Features and nutritional significance of photosynthates produced by unicllular algae symbiotic with larger foraminifera. Mar. Ecol. Prog. Ser. 2:225-228.
KUILE, B. TER, J. EREZ AND J.J. LEE. 1987. The role of feeding in the metabolism of larger symbiont-bearing foraminifera. Symbiosis 4:335—350.
LEE, J.J. 1990. Fine structure of the rhodophycean Porphyridium purpureum in situ in Peneroplis pertusus (Forskol and P. acicularis (Batsch) and in axenic culture. J. Foraminiferal Res. 20:162-169.
LEE, J.J. AND O.R. ANDERSON. 1991. The biology of foraminifera. Academic Press, London.
LEE, J.J. AND W.D. BocK. 1976. The importance of feeding in two species of soritid foraminifera with algal symbionts. Bull. Mar. Sci. 26:530-537.
LEE, J. J., Y. CHAN, AND A. LAGZIEL. 1988a. An immunofluorescence approach toward the identification of endosymbiotic diatoms in several species of larger foraminifera. Pp. 230-241 in Immunochemical approaches to coastal estuarine and oceanographic questions, C.M. Yentsch, F.C. Mague and P. K. Hogan, eds. Springer-Verlag, NY.
Leg, J. J., J. EREZ, B. TER KUILE, A. LAGZIEL AND S. BURGOS. 1988b. Feeding rates of two species of larger foraminifera Amphistegina lobifera and Amphisorus hemprichii, from the Gulf of Eilat (Red Sea). Symbiosis 5:61—102.
LEE, J. J., J. EREZ, M. E. McENERY, A. LAGZIEL AND X. XENOPHOTOS. 1986. Experi- ments on persistence of endosymbiotic diatoms in the larger foraminifer Amphistegina lessonii. Symbiosis 1:211-226.
LEE, J. J., W. W. FABER, JR. AND R.E. LEE. 1991. Granular reticulopodal digestion—a possible preadaptation to benthic foraminiferal symbiosis? Symbiosis, in press.
LEE, J. J. AND P. HALLOCK. 1987. Algal symbiosis as the driving force in the evolution of larger foraminifera. Ann. N. Y. Acad. Sci. 503:330-347.
Leg, J.J., E. LANNERS, AND B. TER KUILE. 1988. Retention of chloroplasts by the foraminifer Elphidium crispum. Symbiosis 5:45-60.
Leg, J.J. AND R.E. LEE. 1990. Chloroplast retention in elphidids (Foraminifera) Pp.
215-220 in Endocytobiology IV, P. Nardon, V. Gianinazzi-Pearson, A.M. Grenier, L. Margulis and D.C. Smith, eds. Institut National de la Recherche Agronomique, Paris.
LEE, J.J. AND M.E. McENERY. 1983. Symbiosis in foraminifera. Pp. 36-68 in Algal Symbiosis, L. J. Goff, ed. Cambridge Univ. Press, Cambridge.
Leg, J.J., M. E. McENERY, AND J. GARRISON. 1980a. Experimental studies of larger foraminifera and their symbionts from the Gulf of Elat on the Red Sea. J. Foraminiferal Res. 10:31-47.
11th DIATOM SYMPOSIUM 1990
LEE, J. J., M.E. McENERY, E. KAHN, AND F. SCHUSTER. 1979a. Symbiosis and the evolution of larger foraminifera. Micropaleontology 25:118—140.
LEE, J. J., M. E. McENERY, R. L. KOESTLER, M. J. LEE, J. REIDY, AND M. SHILO. 1983. Experimental studies of symbiont persistence in Amphistegina lessonii, a diatom-bearing species of larger foraminifera from the Red Sea. Pp. 487-514 in Endocytobiology II, H. E. A. Schenk and W. Schwemmler (eds.). Walter de Gruyter and Co., Berlin.
LEE, J. J.. M.E. McENERY, B. TER KUILE, J. EREZ, R. ROTTGER, R. F. ROCKWELL, W. W. FABER, JR., AND A. LAGZIEL. 1989. — Identification and distribution of endosym- biotic diatoms in larger foraminifera. Micropaleontology 35:353-366.
LEE, J. J.. M. E. McENERY, M. J. LEE, J. REIDY, J. GARRISON, AND R. ROTTGER. 1980b. Algal symbionts in larger foraminifera. Pp. 113-124 in Endocytobiology 1, W. Schwem- mler and H. E. A. Schenk, eds. Walter de Gruyter and Co., Berlin.
LEE, J. J.. M. E. McENERY, R. ROTTGER, AND C. W. REIMER. 1980c. The isolation, cul- ture and identification of endosymbiotic diatoms from Heterostegina depressa d’ Orbigny and Amphistegina lessonii d’ Orbigny (larger foraminifera) from Hawaii. Bot. Mar. 23:297- 302.
LEE, J. J.. M. E. McENERY, M. SHILO, AND Z. REISS. 1979b. Isolation and cultivation of diatom endosymbionts from larger foraminifera (Protozoa). Nature 280:57-S8.
LEE, J.J. AND C.W. REIMER. 1983. Isolation and identification of endosymbiotic diatoms from larger foraminifera of the Great Barrier Reef, Australia, Makapuu Tide Pool, Oahu, Hawaii, and the Gulf of Elat, Israel, with the description of three new species Amphora roettgerii, Navicula hanseniana, and Nitzschia frustulum variety symbiotica. Pp. 327-343 in Proceedings of the 7th International Diatom Symposium, D. G. Mann, ed. O. Koeltz, Koenigstein.
LEE, J. J., C. W. REIMER, AND M.E. McENERY. 1980d. The identification of diatoms isolated as endosymbionts from larger foraminifera from the Gulf of Elat (Red Sea) and the description of 2 new species, Fragilaria shiloi, sp. nov. and Navicula reisii sp. nov. Bot. Mar. 23:41-48.
LEE, J. J., N. M. SAKS, F. KAPIOUTOU. S.H. WILEN, AND M. SHILO. 1984. Effects of host cell extracts on cultures of endosymbiotic diatoms from larger foraminifera. Mar. Biol. 82:113-120.
LEE, J. J. AND X. XENOPHOTOS. 1989. The unusual life cycle of Navicula muscatinei. Diatom Res. 4:69-77.
LEE, M.J., R. ELLIS, AND J.J. LEE. 1982. A comparative study of photoadaptation in four diatoms isolated as endosymbionts from larger foraminifera. Mar. Biol. 68:193—197.
LEUTENEGGER, S. 1977. Ultrastructure de foraminiferes perfores et imperfores ainsi que de leurs symbiotes. Cah. Micropaleont. 3:1—52.
1983. Specific host-symbiont relationship in larger foraminifera. Micropaleon- tology 29:111-125.
1984. Symbiosis in benthic foraminifera: specificity and host adaptations. J. Foraminiferal Res. 14:16—35.
LEUTENEGGER, S. AND H.J. HANSEN. 1979. Ultrastructural and radiotracer studies of pore-function in foraminifera. Mar. Biol. 54:11-16.
29
LEE, JOHN J.
LEVANON-SPANIER, I., E. PADAN, AND Z. REISS. 1979. Primary production in a desert- enclosed sea—the Gulf of Elat (Aqaba), Red Sea. Deep Sea Res. 26:673-685.
Lopez, R. 1979. Algal chloroplasts in the protoplasm of three species of benthic foraminifera: taxonomic affinity, viability and persistence. Mar. Biol. 53:201—211.
McENERY, M.E. AND J.J. LEE. 1981. Cytological and fine structural studies of three species of symbiont-bearing larger foraminifera from the Red Sea. Micropaleontology 27:71-83.
MULLER, P.H. 1978. !4Carbon fixation and loss in a foraminiferal-algal symbiont sys- tem. J. Foraminiferal Res. 8:35—-41.
MULLER-MERZ, E. AND J.J. LEE. 1976. Symbiosis in the larger foraminiferan Sorites marginalis (with notes on Archaias spp.). J. Protozool. 23:390-396.
REIMER, C. W. AND J.J. LEE. 1984. A new pennate diatom: Protokeelia hottingeri gen. et sp. nov. Proc. Acad. Nat. Sci. Phila. 136:194-199.
1988. New species of endosymbiotic diatoms (Bacillariophyceae) inhabiting larger foraminifera in the Gulf of Elat (Red Sea), Israel. Proc. Acad. Nat. Sci. Phila. 140:339-351.
REISS, Z. AND L. HOTTINGER. 1984. The Gulf of Aqaba. Springer-Verlag, NY.
ROTTGER, R. 1972. Die Bedeutung der Symbiose von Heterostegina depressa (Foraminifera, Nummulitidae) fur hohe Siedlungsdichteund Karbonatproduktion. Abh. Deut. Zool. Gesch. 65:43-47.
1974. Larger foraminifera: reproduction and early stages of developement in Heterostegina depressa. Mar. Biol. 26:5-12.
. 1976. Ecological observations of Heterostegina depressa (Foraminifera, Num- mulitidae) in the laboratory and in its natural habitat. Maritime Sed. Spec. Pub. 1:75—80.
ROTTGER, R., A. IRWAN, R. SCHMALJOHANN, AND L. FRANZISKET. 1980. Growth of the symbiont-bearing foraminifera Amphistegina lessonii D’Orbigny and Heterostegina depressa D’ Orbigny (Protozoa). Pp. 125-132 in Endocytobiology 1, W. Schwemmler and H. E. A. Schenk, eds. Walter de Gruyter and Co., Berlin.
SAKAI, K. AND M. NISHIHIRA. 1981. Population study of the benthic foraminifera Baculogypsina sphaerulata on an Okinawan reef flat and preliminary estimation of its annual production. Proc. Fourth Int. Coral Reef Symp. 2:763—766.
SCHMALJOHANN, R., AND R. ROTTGER. 1978. The ultrastructure and taxonomic identity of the symbiotic algae of Heterostegina depressa (Foraminifera: Nummunlitidae). J. Mar. Biol. Assoc. U. K. 58:227-237.
TRENCH, R.K. 1979. The cell biology of plant-animal symbioses. Ann. Rev. Plant Phys. 30:485-532.
ZMIrI, A., D. KAHAN, S. HOCHSTEIN, AND Z. REISS. 1974. Phototaxis and thermotaxis in some species of Amphistegina (Foraminifera). J. Protozool. 21:133-138.
30
11th DIATOM SYMPOSIUM 1990
Explanation of Plates
Plate 1, Figs. 1-5
FIGURE 1. Granulareticulopodia characteristic of a foraminiferan. LM, Phase contrast. 1000x. FIGURE 2. Endosymbiotic diatoms (D) in the pockets formed by pore rims inside test (shell) of a foraminifera. The organic liner of the pore (P) remains after test has been dissolved. TEM. Scale bar = 10 um. FIGURE 3. Endosymbiotic diatom in a pore rim. Symbiosome membrane (arrow) surrounds cell; note absence of a frustule. TEM. 24,000x. FIGURE 4. Fragment of a perforate foraminiferan test showing pores (arrow) in cross section and cup-like inner pore rims. SEM. Scale bar = 10 fm. FIGURE 5. Fragment of perforate foraminiferan test illuminated and inclined at an angle to emphasize cup-like apsects of pore rims. SEM. Scale bar = 20 Lum.
Plate 2, Figs. 6-10
FIGURES 6-10. Representative endosymbiotic diatoms. FIGURE 6. Nitzschia frus- tulum var. symbiotica (left); Fragilaria shiloi (right). Scale bar = 4 um. FIGURE 7. Cocconeis andersonii. Scale bar = 2 um. FIGURE 8. Navicula muscatinei. Scale bar = 2 um FIGURE 9. Protokeelia hottingeri. Scale bar = 4 um. FIGURE 10. Nitzschia valdestriata. Scale bar = 4 um. All figures SEM.
Plate 3, Figs. 11-13
FIGURES 11-13. Representative foraminifera hosts of endosymbiotic diatoms. FIGURE 11. Pacific ‘‘star sand,’’ Calcarina calcar. Scale bar = 400 um. FIGURE 12. A modern nummulite (coin-shaped foraminiferan) from the Red Sea, Operculina ammonoides. Scale bar = 1 mm. FIGURE 13. Indo-Pacific ‘‘star sand,’’ Calcarina spengleri. Scale bar = 400 um. FIGURE 14. A “‘roller-shaped’’ modern imperforate foraminifer, Borelis schlumbergii. Scale bar = 200 um. All figures SEM.
Plate 4, Figs. 14-18
FIGURES 15-18. Endosymbiotic diatom species. FIGURE 15. Putative autospore of Navicula muscatinei. From study by Lee and Xenophontos (1989). Scale bar = 10 Lm. FIGURES 16-18 are from log phase cultures treated with sterile homogenates of host cells (Lee et al. 1984). FIGURE 16. Fragilaria shiloi with a partially formed frustule. Scale bar = 4 um. FIGURE 17. Nitzschia frustulum var. symbiotica with partially formed frustule. Scale bar = 4 um. FIGURE 18. F. shiloi that has divided twice during incubation with host homogenate. Note no new frustule elements have been formed during this in- terval. Scale bar = 4 um. All figures SEM.
Plate 5, Figs. 19-21
FIGURES 19-21. Chloroplast-sequestering species of foraminifera. FIGURE 19. El- phidium crispum from the Indian Ocean at Mombasa, Kenya. SEM. Scale bar = 200 jum. FIGURE 20. Haynesina germanica from the Greater Sippewisett salt marsh, Falmouth, Massachusetts. SEM. Scale bar = 100 um. FIGURE 21. Sequestered chloroplasts (c) in a section of a chamber of E. crispum from the Indian Ocean at Mombasa, Kenya. TEM. 11,500x.
LEE, JOHN J.
11th DIATOM SYMPOSIUM 1990
33
LEE, JOHN J
11th DIATOM SYMPOSIUM 1990
LEE, JOHN J
36
11th DIATOM SYMPOSIUM 1990
Asteromphalus sarcophagus Wallich and Other Species of the Genus Off the Coast of Chile
by P. S. Rivera and H. L. Barrales
Department of Botany, University of Concepcion, P.O. Box 2407, Ap. 10, Concepcion, Chile
with 2 tables and 4 plates
Abstract: Asteromphalus Ehrenberg is the only genus of the family Asterolampraceae found along the Chilean coast. The most common species reported from Chile, both from coastal and oceanic waters, is A. heptactis. Six other species have been found, however, although very rarely, including A. arachnae (Bréb.) Ralfs, A. elegans Grev., A. hookeri Ehrenberg, A. robustus Castr., A. roperianus (Grev.) Ralfs and A. shadboltianus Castr. Some of these have been reported as fossils only, while others as both fossil and living.
This is the first report of the occurrence of Asteromphalus sarcophagus Wallich in Chile. The species was found only in northern waters. Great variability of valve shape found in this species led us to synonymize the forms obovatus and pandorae described by Thorrington-Smith, as well as A. hustedtii (Kolbe) Thorrington-Smith and A. petterssonii (Kolbe) Thorrington-Smith, with A. sarcophagus.
The structure of the cingulum of A. heptactis is described. This is the first time the distal end of the hyaline rays of this species is shown to be occluded by a thin layer of silica with a small central perforation. The external opening of the labiate process is at the basal end of the ray. Previous observations suggest presence of the occluding layer in other genera of the Asterolampraceae as well.
Using data from the literature, a key for the differentiation of all species of Asteromphalus reported from the Chilean coast is given.
Introduction
The family Asterolampraceae H. L. Smith is comprised, according to Gombos (1980), of the following genera: Asterolampra Ehrenberg, Asteromphalus Ehrenberg, Bergonia Tempére, Rylandsia Greville, and Discodiscus Gombos. Simonsen (1972, 1979) also in- cluded Brightwellia Ralfs in the family. However, Ross and Sims (1973) and Gombos (1980) did not agree with Simonsen’s suggestion. Ross and Sims proposed that Brightwellia be placed within the Coscinodiscaceae, while Gombos considered it an independent branch of the Coscinodiscaceae and not in the main developmental line leading to the
37
RIVERA, P.S.
Asterolampraceae. Presently the members of the Asterolampraceae are defined by partially areolated valve surfaces, marginal labiate processes, and hyaline rays opened to the interior of the valve through ray-slits and to the exterior through holes at the marginal ends.
Asterolampra and Asteromphalus are the only members of the Asterolampraceae present in modern oceans. Asteromphalus, with seven species, is the only genus reported either from fossil deposits or recent samples along the Chilean coast, and the species include A. arachne (Bréb.) Ralfs, A. elegans Greville, A. hookeri Ehrenberg, A. robustus Castracane, A. roperianus (Grev.) Ralfs, A. shadboltianus (Grev.) Ralfs, and A. heptactis (Bréb.) Ralfs (Rivera 1983; Rivera et al. 1990).
Asteromphalus differs from the other genera of the family principally by having one hyaline ray narrower, conferring to the valve a bilateral symmetry. Although A. heptactis is the most common species of the genus reported from both coastal and oceanic waters along the Chilean coast, it is always scanty, except in some samples collected south of 40.S. In spite of the fact that this species has been studied with electron microscopy by several workers in the last 15 years (Okuno 1951, 1964; Fryxell and Hasle 1973; Gerloff and Helmcke 1974; Takano 1983; Ricard 1987; and others), some of its morphological features, among them the structure of the cingulum, are still practically unknown. In general, these studies have been done on only a few frustules. Because of this, it has not been possible to compose a complete image of its morphological variability. Some of the features are likely to exhibit a wide range of variation, even within the same species (number of hyaline rays, position of the central area, etc.).
Continuing our attempt to enhance the knowledge of the morphology and taxonomy of the marine diatom flora off the Chilean coast, and with the specific intention of making a study on the species of the genus Asteromphalus, a considerable number of samples deposited at the Diatom Collection of the University of Concepcion, Chile, was reviewed.
Materials and Methods
Samples studied were collected along the Chilean coast between 18.20°S and 45.52’S. Material from the Indian Ocean was collected during the A. Bruun Expedition in 1963 and was made available to us by Prof. J. Gerloff (Botanisches Institut Berlin Dahlem). All samples are deposited in the Diatom Collection, Department of Botany, Universidad de Concepcion, Chile. (DIAT-CONC). Samples for both light and electron microscopy studies were treated for removal of organic matter according to the method described by Hasle and Fryxell (1970). A Zeiss Photomicroscope II was used for light microscopy. For electron microscopy, cells were dried following the method of Anderson (1951). Electron microscopy studies were made at the Laboratory of Electron Microscopy, Univer- sidad de Concepcion. Photographs were taken on an ETEC Austoscan U-1 scanning electron microscope, and transmission electron micrographs were taken by means of a Philips EM 200. The terminology used for the Asterolampraceae is that suggested by Anonymous (1975), Ross et al. (1979), and by Gombos (1980), in particular.
Observations Asteromphalus sarcophagus Wallich (Figs. 1-30)
Wallich 1860:47, Pl. 2, Fig. 12. Rattray 1890:666. De Toni 1894:1, 417.
SYNONYMS: Asteromphalus sarcophagus f. obovatus Thorrington-Smith 1970:821, Pl. 1, Fig. 3. Asteromphalus sarcophagus f. pandorae Thorrington-Smith 1970:821, Pl. 1, Fig. 2. Aster- omphalus hustedtii (Kolbe) Thorrington-Smith 1970:822, Pl. 1, Fig. 4 (Liriogramma hustedtii
38
11th DIATOM SYMPOSIUM 1990
Kolbe 1955:173, Fig. 6; Pl. 2, Fig. 21). Asteromphalus petterssonii (Kolbe) Thorrington-Smith 1970:822, Pl. 1, Fig. 5 (Liriogramma petterssonii Kolbe 1954:40, Fig. 9; Pl. 4, Figs. 46a, b).
MATERIALS: Pacific Ocean: DIAT-CONC 2186, MarChile VIII, St. 12, 25.08. 1972, 18.20’S, 74.16°W. DIAT-CONC 3900, MarChile VIII, St. 12, 08.08.1972, 18.20’S, 74.16°W. Indian Ocean: DIAT-CONC 1668, A. Bruun, N. 143, 1963, 01.54’S, 79.52’E.
AMENDED DIAGNOSIS: Valves flat, 18-111 {um length, 9-23 um width, oblong or linear- oblong through ovoid to ovoid-lanceolate, usually more or less constricted near one (Figs. 2-5, 16, 23-26) or both rounded extremities, giving the valve, in the latter cases, a sar- cophagus-like shape (Figs. 1, 10-15). Central hyaline area, formed by the central expansion of the rays, about one half the length of the transapical axis and located centrally (Figs. 9, 28-30) to markedly eccentric (Figs. 1-5, 10-20, 24-27), with separating lines smooth (Figs. 1-3). Five broad hyaline rays, symmetrically located on the valve, appear elevated on the external valve surface (Fig. 4). Coinciding with the apical axis, a broad hyaline ray is always found opposite a sixth narrower hyaline ray. Depending on the shape of the valve and on the position of the central area, all of the broad hyaline rays are either of approximately the same or different length. All the rays extend only to near the edge of the valve and thus appear to have rounded ends (Fig. 6). The rays open to the exterior through a hole located at the marginal end and to the interior by a linear to undulate ray-slit (Figs. 3, 5). At internal marginal ends the rays have a well-differentiated labiate process. The labiate processes lack external tubes but their simple circular opening can be recognized from the outside at the basal side of the ray-hole (Fig. 6). Internally, the labiate processes are stalked and each longitudinal slit, surrounded by two lips, is curved (Fig. 7), particularly in the case of the narrow hyaline ray (Fig. 8). Areolae usually form radial striae but varying to distorted or irregular apical rows in different specimens, 6—11 in 10 {tm near the apical axis (inner margin of the areolated segment) and smaller toward the valve margin, 11-12 in 10 [um (Figs. 1-5). Areolae are subcircular in outline, with internal foramen and external cribra with circular to elongated uniformly distributed pores, about 4-5 in | um (Figs. 6, 8). This diagnosis is based on the descriptions of A. sar- cophagus, A. sarcophagus f. pandorae, A. sarcophagus f. obovatus, Liriogramma hustedtii, L. petterssonii, and on our own material from the Pacific and Indian oceans.
Data from literature reveals that Asteromphalus sarcophagus, including the taxa we con- sider as synonyms, have always been found in tropical and subtropical zones. The taxon is known for the Atlantic Ocean between the Gulf of Mexico and 07.8’S (Kolbe 1955; Thorrington-Smith 1970), for the Indian Ocean between 10.N and 25.09’S (Wallich 1860; Heiden and Kolbe 1928; Kolbe 1956; Taylor 1966; Thorrington-Smith 1970; Simonsen 1974), and for the Pacific Ocean. In the Pacific Ocean the furthest north (6.44’N- 129.28’ W) and the furthest south (2.52’S—89.50°W) locations have been recorded by Kolbe (1955) as Liriogramma hustedtii and L. petterssonii (Kolbe 1954) respectively. However, our samples reveal presence of A. sarcophagus as far as 18.20’S-74.16’W. The species was found infrequently very scanty and only in two vertical plankton samples in the same locality. Frustules of A. sarcophagus appear to be extremely fragile, and many of the specimens observed were found eroded or only as fragments.
External cribrum of areolae is convex and slightly raised above the valve surface, with regularly distributed pores. The shape of areolae is relatively constant, and basically there were no differences in the number of areolae in the material studied when compared with data from literature or from measurements made on illustrations available (Table 1).
Thorrington-Smith (1970) has emphasized that one of the principal differences between A. hustedtii or A. petterssonii and A. sarcophagus is the presence of puncta in the former
39
RIVERA, P.S.
spua payomnsuos
aqqeuea wn
suojqo 0} (sant -WOXd ,Santwuanxe pepunol) popunol wuojuAd sopts ‘sapIs SaTWadXa SA\TWAXO Javouod aABOUOS pepunol ‘sapts yora Iau UIA, yy Buo[qo DABOUOD YA PIOAO 0} UOHINSUOS 1YSI]s SOATBA o1quioyl o1quioyl -IRouT] 218 A0GO oIquols WIOJIAO WLIOJIAO a1eTOIUR] yt Suojqo —s- Jo adeys arjound avloore Jo SMol panojstp pouoistp Noy ypevo}sip =. |eotdesury avjoor jo 0} yeIpel ,[eIpel pur [eipel +[etpel 4[etpel ‘repnBoL oO} yeiper —s pure rorde — vonewoHG Clall +01-6 +678 d +678 é é é — UISIVU dAlPA = = 9 Ol OI-9 Ol OI = av[ooly IT-8 #87L *L é poe) é é é — UIs TeU! JOUUT 81-6 xe CCC LI LI-91 cl 60.91 OT CITI oT (,,6000°0) TZ (wir) stxe jeordesuery, 8E-77— eS E807 TI1-98 Oe (09) TE-0T Of 9C VE 9S (,8100'0) L’Sp — (wiri) stxe jeordy paopuvd snjDAaogo snépydooavs ‘J sn&pydooavs *y sn8pydorips “J snépydoosvs Ipaisny iypaisny — tuosssayjad snpoyduoiasy snjoydwo.asy snjoydwosaisy — snjpydwodsaisy puiupsso1.r7 DIULUDASOLAIT Apmis se se sP se se se waseld pL6T WesuoUls OL6T Ws voy suLoy Ty CS6I PGIOM = PS6T PION O98T YONIRAL “Kpnys quasoid oy) wos pue oinjesoy] oy) WOIy sMdpydoouvs snpoyduosaisy uo eyep s1s0joydiow “| ATAVL
40
11th DIATOM SYMPOSIUM 1990
>
‘UONne SNOT WoO, poule}go UONRUUOJUT, sura0Q ueIpu] yinos sueaoQ sted pue ura0Q ura0Q ues0Q ue20Q pure onueny ortor ueIpu] URIPU] UeIPuy YNOS URIpUy YINoS ur20Q yinos ues0(—) yinos yinos yinos pue yon pue yuoN onuepy ynos pue YON oyloeg yinosg ukd9Q URIpUy uONNqmsIq R[OaIe asIv] eB L[OdIe ISIE] v L[OAIe dBIE] B UI SO]BUTLIO} UT SO}BUTUIA] UT SayRUTULTA) sArd Aer ourpeAy = Avs outpeAy — Aes ourpeAy aurpeAy MOLIRU MOLIeU AOLU Jo pua ‘Aer ourpeAy = Aer ourpeAy = Aer ourpeAy jeursiew yore JO peolq yore JO peolq yora jo yora pua qe outds pua ye outds puo ye ourds ye UO — yeursiew jeursrew yeursiew = - - — sassaoo.1d 212 1qe'] ouua909 A[payreu =, EUa09a dINUaI9a 0} A[payreut 0} == 9tyUIa99a A[payreu vam [BQUa. dtuacsa = oLUa009 Anystys dUUa999 0} LUI. ouuac.a 4olua.9— o1uas ouquas9a «JO UOTRI07T 9 9 9 9 9 9 9 9 ° skel ourtpReAy Joquiny piopupd snjDAogo sn&pydoosvs ‘y sn&vydooavs *y snépydooaps *y sn&vydoo.vs 1ypajsny Ijpaisny — iuosssajjad snjpydwosajsy snjpydiuosajsy snjpydwuosajsy snjpydiuosajsy puauvssoinT puiupsso1sT Apnys se se se se se se
Waseag pL 6] uesuowrs
OFLC IAEA SD Oa UELLOUNS SC6I POM = PSOT PATON O98T YONA
“ponuluood “|[ ATAV EL
41
RIVERA, P.S.
species and areolae in the latter. In the Chilean material only one A. hustedtii-like valve was found (Fig. 18), and because of this, it was impossible to prepare it for SEM. The general appearance of the individual corresponds to typical A. sarcophagus cells. The perforations appear brighter, however, and this may be due to a higher content of silica or simply that one is dealing with an anomalous cell. It should be mentioned that Kolbe (1955) in his description of Liriogramma hustedtii Kolbe referred to these perforations specifically as ‘‘areolae.”’
In the Chilean material the length/breadth ratio of the cells varies from 1:2.2 to 1:3. Great variability in the degree of concavity of the sides at each end of the valve was observed. This range of variation is shown in Figures 9-30 and represents a continuum of forms from one extreme to the other. This, together with the fact that no real differences exist in relation to the number of hyaline rays and areolae, and that all these forms share the same geographical distribution, has led us to consider all forms previously described as different taxa, as synonyms of A. sarcophagus. We include Asterophalus hustedtii (Kolbe) Thorrington-Smith and A. petterssonii (Kolbe) Thorrington-Smith as synonyms of A. sar- cophagus. The occurrence of ‘‘central markings’’ and of ‘‘irregular areolation’’ observed by Kolbe in both taxa were considered by him as primitive characters and were the basis for his creation of the new genus Liriogramma. Based on our observations we cannot detect fundamental differences in the morphology of A. hustedtii, A. petterssonii, and A. sarcophagus. Furthermore, in some of our specimens an undulated narrow hyaline ray, similar to that illustrated by Kolbe for his L. petterssonii and by Thorrington-Smith (1970) for his form pandorae, can be discerned (Figs. 2, 3, 5, 25). Finally, we are in agreement with Simonsen (1974) in the sense that Liriogramma species are ‘‘malformations of some sort.”
Asteromphalus heptactis (Brébisson) Ralfs (Figs. 31-42)
Ralfs in Pritchard 1861:838, Pl. 8, Fig. 21. Okuno 1964:23, Pl. 443. Takano 1983, sheet 128. BASIONYM: Spatangidium heptactis Brébisson 1857:296, Pl. 3, Fig. 2.
SYNONYMS: Asteromphalus reticulatus Cleve 1873:5, Pl. 1, Fig. 2. Asteromphalus ralfsianus (Norman) Grunow in Schmidt et. al. 1876, Pl. 38, Figs. 6, 7. Asteromphalus ornithopus Karsten 1905:90, Pl. 8, Fig. 13.
MATERIALS: Pacific Ocean: DIAT-CONC 3901, MarChile VIII, St. 12, 18.20°S- 74.16'W, 08.08.1972. DIAT-CONC 3634, MarChile VII, St. 13, 18.96°S—74.47°W, 19.08.1972. DIAT-CONC M-1287, Exp. Downwind, St. 29, 22.38’S—72.00’ W, 02.01.1958. DIAT-CONC M-2165, Bahia Concepcion, 36.43’S—73.00°W, 14.06.1965. DIAT-CONC M-2169, Bahia Concepcién, 36.43’S—73.05°W, 09.06.1965. DIAT-CONC M-1356, 38.00’S—73.50'W, 22.11.1978. DIAT-CONC 2282, 40.40’S—74.07°W, 15.11.1970. DIAT- CONC M-1260, 41.48’S-73.05’W, 29.04.1977. DIAT-CONC 2555, 41.49°S—73.05’W, 05.08.1977. DIAT-CONC 2575, 41.49°S-73.05’W, 08.10.1977. DIAT-CONC 2360, 45.12’S-75.28'W, 16.11.1978. DIAT-CONC 2286, 45.52’S—75.40’W, 16.11.1970.
Cells are single, in girdle view rectangular to drum-shaped with slightly convex and radially undulated valves (Fig. 31). Pervalvar axis 14-30 ttm. One broad and homogeneously silicified band in each cingulum, 5.6—7.3 [um width (Figs. 31, 32). Valves subcircular to oval, 29-50 um in diameter. Central hyaline area circular to polygonal, eccentric, about 1/3 the diameter of the cell; separating lines jagged (Figs. 33, 34). Six to seven broad hyaline rays (2.8-4 um width), plus a narrow one, are elevated on the external valve
42
11th DIATOM SYMPOSIUM 1990
surface and extend to edge of valve. The rays appear square-ended externally, but with rounded ends in internal valve view (Figs. 35, 36). Narrow hyaline ray (0.9-1.5 {um width) is usually slightly longer and only somewhat raised. If six broad hyaline rays are present, three of them are arranged on one side of the valve and the others on the opposite side (separated by the narrow hyaline ray; Figs. 34, 35); when seven broad hyaline rays are present they retain the same position and the additional ray follows in line with the narrow hyaline ray (Fig. 33). Each hyaline ray opens inward by a somewhat linear ray-slit (Figs. 35, 36). At the marginal external end the ray is occluded by a slightly depressed and delicate layer of silica, oriented perpendicular to the valvar plane (Figs. 32, 37-41). This layer has a central, circular hole of ca. 0.2—0.4 [tm in diameter, bordered by one or two hyaline rings of silica (Figs. 37-39). These rings usually appeared eroded. In the basal part of the distal end of the hyaline rays there is a second hole, of approximately the same diameter, which corresponds to the external opening of the labiate process (Figs. 37-40).
Although broad and narrow hyaline rays are similar in these structural features, they differ in other aspects of their structure. The broad hyaline rays have an external marginal end section that is circular in shape (ca. 1.9-2.5 um in diameter), with its basal portion formed by two linear segments arranged at approximately 90 degrees (Figs. 37-39). The narrow hyaline ray, however, has an elongated (Fig. 42), or sometimes horseshoe-shaped external marginal end section with its internal basal and lateral sides with areolae-like perforations (Fig. 41). The internal part of each labiate process is well differentiated and the slit is curved. Areolated segments taper towards the center of the valve. Areolae are somewhat hexagonal, 6-8 in 10 um, usually uniform in size and only slightly smaller in the valve margin, forming lines parallel to the marginal prolongation of the hyaline rays as well as tangential lines (Figs. 33-35). External cribra have subcircular pores uniformly distributed, 4-5 in | um, and circular to subcircular internal foramen, 1.2—1.5 um in diameter. Areolae are connected by passage pores (Fig. 36).
Asteromphalus heptactis, the most common species of the genus, is a cosmopolitan plankton taxon. Simonsen (1974) pointed out some reservations about its designation as oceanic as proposed by Cupp (1943), Hendey (1964), and Sournia (1968). Material studied in this paper was collected along the Chilean coast between 18.20’S and 45.52’S, and the species was particularly abundant in coastal samples south of 40.S. The species had been previously reported for the Chilean coast between 20.39°S and 51.35’S (Krasske 1939, 1941; Balech 1962; Rivera 1969; Fenner et al. 1976; Uribe et al. 1982). Okuno (1951, 1964), Fryxell and Hasle (1973), Gerloff and Helmcke (1974), and Ricard (1987) have previously studied the species by electron microscopy. Morphological data of A. heptactis compiled both from the literature and from the Chilean material examined in this paper are summarized in Table 2.
Asteromphalus heptactis is very closely related to A. parvulus Karsten (1905:90, Pl. 8, Fig. 14). Priddle and Fryxell (1985) found it impossible to distinguish between the two in routine analysis of samples. However, A. parvulus has a larger central area (about one half the valve diameter) located symmetrically on the valve, usually fewer broad hyaline rays, smaller areolae, and has been found mostly in subantarctic and antarctic waters. The individual shown by Fenner et al. (1976, Pl. 4, Fig. 2) as Asteromphalus hookeri Ehrenberg has an oval valve, jagged separating lines, and a markedly eccentric central area, and these features would assign it to A. heptactis.
The following key has been developed from our observations as well as from data in the literature, to separate the eight taxa of Asteromphalus reported so far from Chilean samples.
43
RIVERA, P.S.
The characteristics used are discernible by light microscopy.
1 Valves elongated, rhombic with concave to oblong sides with variably constricted
TKO} UE YOU =) (0b meyer ees ra ae a Ce rears Poe Cael eer A. sarcophagus
1’ Valves circular to suboval ..................
2 Central area located symmetrically on the valve ...... 3
2 Centralvarea, eccenthiG=s «mane so.) 4 aye seas ee ee 5
3 Separating lines smooth, sometimes bifurcated. Marginal external end of broad hyalinetrays: round-ended sy) 3 ...cnse ener rae kon eee A. hookeri
3’ Separating lines jagged. Marginal external end of broad hyaline rays square-ended 4 Areolae 5-8 in 10 um. Valve diameter usually <60 um . A. heptactis
4 Areolae 12-16 in 10 um. Valve diameter usually 80 um . A. roperiatus
5 Central area small ca. 1/6-1/7 the valve diameter and markedly eccentric. Usually 4
broad hyaline rays with rounded marginal external ends ... A. arachne
5°’ Central area about 1/4-1/2 the valve diameter ....... 6
6 Areolae 5-8 in 10 [im. Five to seven broad hyaline rays with square external mar- inal Ends: . ace Sv eeate. jaye wees Sane: Goan, ee OR ey 4 ne ee A. heptactis
6’ Areolae usually more than 10 in ]O um .......... i
7 Ten to twelve broad hyaline rays. Valves circular. Central area about 1/3 with respect to the valve diameter ................0.. A. elegans
7° Four to eight broad hyaline rays... ...........4. 8
8 Valves 25-50 um in diameter. Areolae 12 in 10 um ... A. robustus
8 Valves larger. Areolae up to 14 in 10 Um ..... 2... A. shadboltianus
Asteromphalus arachne (Bréb.) Ralfs was reported from Arica (ca. 18.28’S—70.20’W) by Schmidt 1886 in Schmidt et al. (1874— 1959). Asteromphalus elegans Greville was ob- served in material from Mejillones (ca. 23.06°S—70.27’W) by Boyer (Rivera and Gebauer 1989) and off Antofagasta (ca. 20.30°S—69.00’W) by Hendey (1937). Belyaeva (1972) reported this taxon from the South Pacific Ocean between O.S. and 32.S. Asteromphalus hookeri Ehrenberg was found off the coast of Mejillones by Hendey (1937) and in the Strait of Drake (ca. 58.54°S-66.08°W) by Meyer (1966). Asteromphalus robustus Castracane was reported from Tripoli of Tiltil (ca. 33.05’°S—70.56’ W) and Tripoli of Mejil- lones (ca. 23.06’S—70.27’W) by Frenguelli (1949), and also from Mejillones by Tempére and Peragallo (1907) as A. brookei Grunow. Asteromphalus roperianus (Grev.) Ralfs was found in Tripoli of Mejillones (ca. 23.06°S—70.27'W) by Rattray (1890), and by Tempére and Peragallo (1907), and by Frenguelli (1949). Asteromphalus shadboltianus (Grev.) Ralfs has been reported only from Tripoli of Mejillones (Rattray 1890; Méller 1891; Tempere and Peragallo 1907; Frenguelli 1949).
Discussion
A salient feature of these results is the great variation in the morphology of the species under study. It appears that this situation is likely to be encountered in all members of the genus Asteromphalus, as well as in those belonging to Asterolampra, following Simon- sen (1974).
The main criteria for distinguishing among species of Asteromphalus traditionally have been the number of broad hyaline rays, size, and position of the central area on the valve face and the number of areolae. Recognizing that these characteristics, as well as others found in the present study, exhibit a wide range of variation within a taxon, we agree with Simonsen (1972) in the sense that one has to make use of them with the utmost
44
11th DIATOM SYMPOSIUM 1990
UONeI}sN|[! Woy paulejgo UONRULOJU y
peao eL-9¢ S-+ 8-9 €/I ouuaace 8-L 0S - 67 Oy dejnostogns ~——Apnys juasaug
#[PAO ~ 7 ¥£/1 - P/L #U9299 +L = 01 aeINosOqQns L861 Preory S861 1exAd4 ~ ~ age] ouua9 L=Q cs — 0€ JB [NDII9 29 2|PPud = Les €/1 — p/T ~9U)U9999 «L OOT — 8E +Jb[Nodogns £86 Ourye] 9L6l ~ L-9 €/I 99999 «L Sp - IE [BAO 0} Je[NoII9 "[e 1a JouUDF x[BAO £L6I PISPH ~ ~ ~ 49U]U9999 «Ll = 0} Jepnostogns ¥ [ex b L-9 €/1 — b/I 9uUaa9a L-+ OL 89 qejnoutoqns $961 CUNO As L-9 £/1 — b/1 90U3099 i£=9 06 — OF Je[NIIOQns Is6l OunyxO = a £/I -— p/I é ouUaa L OOL— Os Ploosip $961 Aapuay
9UU2999 ~ $9 -9 €/1 — b/I Apysiys L OOI~ 8€ sJP[NOIIOGNs €r61 ddn5 a = €/1 - P/1 oujuaaca L OlL- +9 [RAO 0} ploosip —_L€6 1 AapuaH > (EES r/1 91U2999 ic Sil — Ge [BAO OL IBNIIID —OEHT IPIISNH (win | ur) tuniqus Ja awWeIp pueq ul suoneiopiad Joquinu DATRA O} BIR Pose [P1]Ud9 sXes aurypeAy SOATRA yora JO YIP jo Joquiny, jo Joquinyy aRjouy [Puss jo oney jo uonisog JO Jaquinu [e]O], Ja}OWRIP sATRA, jo adeys
tee
“Apnjs quasaid ay) woay pue ainjesayt] ay) Woy syoviday snjpydwosaisy uo wep a30joydIow “7 ATAVL
45
RIVERA, PS.
discrimination and, following Priddle and Fryxell (1985), not singly but in conjunction with other relevant features.
Asteromphalus sarcophagus serves as an excellent example to illustrate variation of the general outline of the valves. Although in this species the elongated shape of the valves differs considerably with the generalized circular outline in the other members of the genus, the range of variation observed in the Chilean material allowed us to synonymize the forms obovatus and pandorae described by Thorrington-Smith (1970) with the species of Wallich, as we also did with A. hustedtii and A. petterssonii. In A. heptactis, variation in valve shape is not as notorious, and is usually reported as subcircular to oval, except by Hendey (1964) and by Priddle and Fryxell (1985), who described it as discoid or circular, respectively. We consider the systematic value of this characteristic of little im- portance.
The position of the central area in Asteromphalus sarcophagus was observed to be also highly variable, from centric to markedly eccentric. In A. heptactis, however, it always retains an eccentric position, as reported in the majority of previous studies. A similar situation is found regarding the number of broad hyaline rays, which is constant in A. sarcophagus and quite variable in A. heptactis. In both taxa length of the hyaline rays varies depending on the shape of the valve, and this is particularly so in the case of A. sarcophagus, as shown in the present study. In summary, characterization of Asteromphalus taxa should be attempted only after a fairly large number of individuals have been exam- ined.
As far as we know, the structure of the girdle in the genus Asteromphalus has not been observed or described previously by means of electron microscopy. This assumption seems also to be valid for the rest of the members of the family Asterolampraceae, as no references are made to it in recent papers (Okuno 1964; Fenner et al. 1976; Takano 1983; Gombos 1980; Ricard 1987). Only single valves were found of A. sarcophagus, and thus the structure of its cingulum is still unknown. The cingulum has been observed only in A. heptactis. In this species, the characteristics of these hyaline bands were very constant in all individuals observed. These bands are completely different from those in Rylandsia biradiata (Gombos 1980, Fig. 56), where vertical lines of perforations are distinct. Perhaps after a thorough and detailed examination of the structure of the cingulum in the Asterolampraceae, its characteristics could be used as additional criteria to distinguish between the various genera or species, as done in other centric diatoms, e.g., in the genus Thalassiosira (Fryxell et al. 1981).
One of the most interesting findings of this study was the recognition, in Asteromphalus heptactis, of a thin wall of silica with a small central opening, occluding the external marginal end of both broad and narrow hyaline rays. Previous information about the Asterolampraceae described the hyaline rays as opening to the exterior through a large hole positioned at their marginal ends (Fryxell and Hasle 1973; Ross and Sims 1973; Gombos 1980). Although Fryxell and Hasle (1973), as well as Ricard (1987), referred to this opening as a ‘‘pseudonodule,’’ such a designation was considered erroneous by Simon- sen (1975) and by Gombos (1980), who simply used the words ‘‘hole’’ and ‘‘ray hole,”’ respectively. In all samples prepared for SEM by critical point drying, presence of the silica layer could be recognized. However, when samples were treated to remove the organic matter, the silica layer did not stand up to this processing; after treatment the ends of the hyaline rays appear simply as large holes.
Although this is the first time that this peculiar structure is described for the genus Asterom- phalus, it should be pointed out that it can be recognized in figures obtained by other
46
11th DIATOM SYMPOSIUM 1990
authors. Gombos (1980, Fig. 46), in a marginal view of Asterolampra marginata aff. var. curvilineata Brun, shows two hyaline rays where their external holes appear occluded by a membrane. This same layer can also be clearly discerned in his Figure 49, corresponding to Discodiscus tetraporus (Brun) Gombos. Similarly, the silica layer is also recognizable in Asteromphalus heptactis, as presented by Takano (1983) in his Figure B of sheet 128. However, in none of these figures is it possible to detect either the presence of the central perforation or that corresponding to the labiate process. Considering these points, it seems plausible to assume that the presence of this layer is a characteristic feature of the Asterolampraceae. The general appearance of this layer, with the exception of a closely positioned labiate process, is very similar to the operculate pseudonodule described for Actinocyclus circellus Watkins by Watkins and Fryxell (1986). It should be recalled, however, that pseudonodules, typical of the family Hemidiscaceae, are known to occur only one per valve, and this is not the case in Asteromphalus.
It is expected that research now in progress, using the critical point drying method, will contribute additional information about the relevance of the silica layer as a distinctive feature of the Asterolampraceae.
Literature Cited
ANDERSON, T. F. 1951. Techniques for the preservation of three dimensional structure in preparing specimens for the electron microscope. Trans. N. Y. Acad. Sci. II, 13:130-134.
ANONYMOUS. 1975. Proposals for a standardization of diatom terminology and diag- nosis. Nova Hedwigia, Beih. 53:323-354.
BALECH, E. 1962. Tintinoidea y Dinoflagellata del Pacifico segtin material de las ex- pediciones Norpac y Downwind del Instituto Scripps de Oceanografia. Revista del Instituto de Investigaciones de Ciencias Naturales, Buenos Aires, Zoologia 7:1—253.
BELYAEVA, T. V. (1972). Distribution of large diatom algae in the Southeastern Pacific. Okeanologiya 12:475-483.
BREBISSON, A. de. 1857. Refermant la description de quelques nouvelles Diatomees observées dans le Guano de Pérou, et formant le genre Spatangidium. Bull. Soc. Linnéenne de Normandie 2:292-298.
CLEVE, P.T. 1873. On diatoms from the Arctic Sea. Bihang Kongl. Svenska Akad. Handl. 1:1-28.
Cupp, E.E. 1943. Marine plankton diatoms of the west coast of North America. Bull. Scripps Inst. Ocean. 5:1-238.
DE TONI, J.B. 1891-1894. Sylloge algarum omnium hucusque cognitarum, vol. II, Bacillarieae; sectio I, Raphideae, pp. 1490 (1891); sectio II, Pseudoraphideae, pp. 491- 817 (1892); sectio IIH, Cryptoraphideae, pp. 818-1556 (1894). Typis Seminarii, Patavii.
FENNER, J., H. J. SCHRADER, AND H. WIENICK. 1976. III. Diatom phytoplankton studies in the southern Pacific Ocean, composition and correlation to the Antarctic convergence and its paleoecological significance. Initial Reports of the Deep Sea Drilling Project 35:757-8 13.
FRENGUELLI, J. 1949. Diatomeas fosiles de los yacimientos chilenos de Tiltil y Mejil- lones. Darwiniana 9:97—157.
47
RIVERA, P.S.
FRYXELL, G. A. AND G.R. HASLE. 1973. Coscinodiscineae: Some consistent patterns in diatom morphology. Nova Hedwigia, Beih. 45:69-96.
FRYXELL, G. A. G. F. HUBBARD, AND T. A. VILLAREAL. 1981. The genus Thalas- siosira: variations of the cingulum. Bacillaria 4:41—63.
GERLOFF, J. AND J-G. HELMCKE. 1974. Tafel 715-824 in Diatomeenschalen im eleck- tronenmikroskopischen Bild. Teil VII. J-G. Helmcke, W. Krieger and J. Gerloff, eds. J. Cramer, Lehre.
GomBOs, Jr., A.M. 1980. The early history of the diatom family Asterolampraceae. Bacillaria 3:227-272.
HASLE, G.R. AND G. A. FRYXELL. 1970. Diatoms: cleaning and mounting for light and electron microscopy. Trans. Amer. Microsc. Soc. 89:469-474.
HEIDEN, H. AND R. W. KOLBE. 1928. Die marinen Diatomeen der deutschen Sud-Polar Expedition 1901-1903. Deut. Sud-Polar Exp. 8:450—714.
HENDEY, N.I. 1937. The plankton diatoms of the Southern Seas. Discovery Rep. 16:151-364.
. 1964. An introductory account of the smaller algae of British coastal waters. Part V: Bacillariophyceae (Diatoms). Her Majesty’s Stationery Office, London, 317 pp.
HUSTEDT, F. 1930. Die Kieselalgen in Kryptogamen-Flora von Deutschland, Oester- reich und der Schweiz, L. Rabenhorst, ed. 7:1—920.
KARSTEN, G. 1905. Das Phytoplankton des Antarktische Meeres nach dem Material der deutschen Tiefsee-Expedition 1898-1899, in Wissenschaftliche Ergebnisse der Deutschen Tiefsee-Expedition auf dem Dampfer Valdivia 1898-1899, C.M. Chen, ed. 2:3-136.
KOLBE, R.W. 1954. Diatoms from equatorial Pacific cores. Rep. Swedish Deep-Sea Exp. 1947-1948, 6:1-49.
. 1955. Diatoms from equatorial Atlantic cores. Rep. Swedish Deep-Sea Exp. 1947-1948, 7:151-184.
. 1956. Diatoms from equatorial Indian Ocean cores. Rept. Swedish Deep-Sea Exp. 1947-1948, 9:1-S0.
KRASSKE, G. 1939. Zur Kieselalgenflora Siidchiles. Arch. Hydrobiol. 35:349-468.
. 1941. Die Kieselalgen des chilenischen Kistenplanktons. Arch. Hydrobiol. 38:260-287. MEYER, R. 1966. Contribucion al estudio del fitoplancton del Paso de Drake. Cuaderno 1, Ciencias del Mar, Universidad Cat6lica de Valparaiso 41-82.
MOLLER, J.D. 1891. Lichtdrucktafeln hervorragen sciiner und vollstandiger Miil- ler’scher Diatomaceen-Praparate, 1891, Verzeichniss der in den Lichtdrucktafeln MOllerscher Diatomaceen-Praparate enthaltenen Arten. Wedel.
OKUNO, H. 1951. Electron microscopical study on antarctic diatoms. J. Jap. Bot. 26:305-310.
. 1964. Fossil diatoms. Pl. 414-513 in Diatomeenschalen im elektronenmikros- kopischen Bild, Teil V. J-G. Helmcke and W. Krieger, eds. J. Cramer, Lehre.
48
11th DIATOM SYMPOSIUM 1990
PRIDDLE, J. AND G. FRYXELL. 1985. Handbook of the common plankton diatoms of the southern ocean. Centrales except the Genus Thalassiosira. British Antarctic Survey, Cambridge. 159 pp.
PRITCHARD, A. 1861. A _ history of the infusoria, including the Desmidiaceae and Diatomaceae, British and foreign. Whittaker and Co., London. 968 pp.
RATTRAY, J. 1890. A revision of the genus Coscinodiscus Ehrenberg and some allied Genera. Proc. R. Soc. Edinburgh, 24:449-692.g
RICARD, M. 1987. Diatomophycees. /n Atlas du Phytoplancton Marin, Vol. 2. A. Sour- nia, ed. Centre National de la Recherche Scientifique, Paris. 297 pp.
RIVERA, P. 1969. Sinopsis de las Diatomeas de la Bahia de Concepcion, Chile. Gayana, Bot. 18:1-112.
. 1983. A guide for references and distribution for the class Bacillariophyceae in Chile between 18.28’S and 58.S. Bibl. Diatomologica 3:1—386.
RIVERA, P. AND M. GEBAUER. 1989. Diatomeas chilenas en las Colecciones de Boyer, Cleve and Mdller, Schulze y Smith, depositadas en la Academia de Ciencias Naturales de Filadelfia, Estados Unidos. Gayana, Bot. 46:89-116.
RIVERA, P., M. GEBAUER, AND H. BARRALES. 1990. A guide for references and dis- tribution for the class Bacillariophyceae in Chile between 18.28’S and 58.S. Part II. Data from 1982 to 1988. Gayana, Bot. 46:155—198.
Ross, R. AND P. A. SIMS. 1973. Observations on family and generic limits in the Centrales. Nova Hedwigia, Beih. 45:97— 130.
Ross, R., E. J. Cox, N. I. KARAYEVA, D. G. MANN, T. B. B. PADDOCK, R. SIMONSEN, AND P. A. SIMS. 1979. An amended terminology for the siliceous components of the diatom cell. Nova Hedwigia, Beih. 64:513-533.
SCHMIDT, A. 1874-1959. Atlas der Diatomaceen-Kunde. R. Reisland, Leipzig. Taf. 1- 472.
SIMONSEN, R. 1972. Ideas for a more natural system of the centric diatoms. Nova Hedwigia, Beih. 39:37—54.
——. 1974. The diatom plankton of the Indian Ocean Expedition of R/V *‘Meteor”’ 1964-65. ‘‘Meteor’’ Forschungsergebnisse, ser. D, 19:1—107.
. 1975. On the pseudonodulus of the centric diatoms, of Hemidiscaceae recon- sidered. Nova Hedwigia, Beih. 53:83—-94.
1979. The diatom system: ideas on phylogeny. Bacillaria 2:9-71.
SOURNIA, A. 1968. Diatomees planctoniques du Canal de Mozambique et de L’Ile Maurice. Memoire Office de Recherche Scientifique et Technique Outre-Mer 31:120.
TAKANO, H. 1983. Sinopsis of red-tide organisms. Working Party on Taxonomy in the Akashiwo Kenkyukai, ed. Fisheries Agency, Japanese Government, Sheet 128.
TAYLOR, F. J.R. 1966. Phytoplankton of the south west Indian Ocean. Nova Hedwigia 12:433-766.
TEMPERE, J. AND H. PERAGALLO. 1907. Diatomées du Monde Entier. Ed. 2, Arcachon, Grez-sur-Loing (S.-et -M.) 1:1-16.
49
RIVERA, P.S.
THORRINGTON-SMITH, M. 1970. Some new and little-known planktonic diatoms from the west Indian Ocean. Nova Hedwigia, Beih. 31:815-835.
URIBE, E., S. NESHYBA, AND T. FONSECA. 1982. Phytoplankton community composi- tion across the west wind drift off South America. Deep-Sea Res. 29:1229-1243.
WALLICH, G.C. 1860. The siliceous organisms found in the digestive cavities of the Salpae, and their relation to the flint nodules of the chalk formation. Trans. Microsc. Soc. N. S. 8:36-55.
WATKINS, T.P. AND G. A. FRYXELL. 1986. Generic characterization of Actinocylus: consideration in light of three new species. Diatom Res. 1:291-312.
Explanation of Plates
Plate 1, Figures 1-8
FIGURES 1-8. Asteromphalus sarcophagus. FIGURES 1-2. LM. Areolae usually forming radial striae, but distorted or irregular rows may be present. Number of broad hyaline rays (5) very constant in the species. FIGURES 3, 5. SEM. Internal valve views. Rays open to the interior by a linear to undulate ray slit. FIGURE. 4. SEM. Hyaline rays appear elevated on external valve surface. FIGURE 6. SEM. All rays extend only to near edge of valve and open to exterior through a hole located at marginal end. FFGURES 7-8. SEM. Internal view of labiate processes, located at the marginal end of hyaline rays; longitudinal slit curved, particularly in the case of the narrow hyaline ray (Fig. 8).
Plate 2, Figures 9-30
FIGURES 9-30. Range of cell variability in Asteromphalus sarcophagus and in allied taxa. Figures 9-10 according to Thorrington-Smith 1970; Figures 11-12, 14-15 according to Simonsen 1974; Figures 13, 16-18, 22-26 present study. FIGURES 19-20. Liriogram- ma_ hustedtii. FIGURE 21. Asteromphalus sarcophagus f. obovatus. FIGURE 27. Asteromphalus hustedtii. FIGURE 28. Liriogramma petterssonii. FIGURE 29. Asterom- phalus petterssonii. FIGURE 30. Asteromphalus sarcophagus f. pandorae.
Plate 3, Figures 31-36
FIGURES 31-36. Asteromphalus heptactis. FIGURES 31-32. SEM. Each cingulum is composed of a broad and homogeneously silicified band. FIGURES 33-34. LM. Valve shape varies from subcircular to oval; the central hyaline area is always eccentric and the number of hyaline rays is variable. FIGURE 35. SEM. Internal valve view. Labiate process is located at marginal end of each ray. FIGURE 36. TEM. Internal view of a valve showing ray slit of hyaline rays and the external cribrum of areolae.
Plate 4, Figures 37-42
FIGURES 37-42. SEM Asteromphalus heptactis. FIGURES 37-39. Broad hyaline rays have an external marginal end section circular in shape, with its basal portion formed by two linear segments arranged at approximately 90 degrees. Central perforation of the occluding layer is bordered by one or two rings of silica, which usually appear eroded. External opening of the labiate process is evident in the basal part. FIGURES 40-41. An occluding marginal layer with central and basal perforations is also present in the narrow hyaline ray; the internal basal and lateral sides have areolae-like perforations. FIGURE 42. Marginal end section of the narrow hyaline ray is normally elongated, sometimes horseshoe-shaped.
11th DIATOM SYMPOSIUM 1990
an . wen? ie
x prey ae _— Dit 7 », "i Kes | ‘ - ae. de
ON oa
51
RIVERA, PS.
+ DS
11th DIATOM SYMPOSIUM 1990
53
RIVERA, P.S.
11th DIATOM SYMPOSIUM 1990
Are Both Rhizosolenia curvirostris Jousé and R. barboi Brun Found in Pleistocene Sediments of Northeastern Europe?
by Emma Loseva
Komi Scientific Center, Ural Department, Academy of Sciences of Russia, Syktyvkar, 167610, Russia
with 3 plates
Abstract: Marine diatom assemblages from bore-hole 1-Ja, located north of the Pechora River lowland, were examined. Rhizosolenia valves, formerly determined as R. curvirostris and R. barboi, were observed with light and scanning electron microscopy, and it was determined that only R. curvirostris was present in the assemblages. Rhizosolenia curvirostris is the index species for the middle Pleistocene in the Pacific. In northeastern Europe, R. curvirostris is characteristic of lower and middle Pleistocene sediments. This species became extinct in the region about 130 Kyr.
Introduction
Rhizosolenia curvirostris Jousé and R. barboi Brun are known to be index species for the Pleistocene and Pliocene, respectively, in the North Pacific. Jousé (1968) described R. curvirostris from sediments of the Sea of Okhotsk and the North Pacific. She restricted its upper age limit to the middle Pleistocene and emphasized that this species was absent in modern plankton and surface sediments of the Pacific.
The valve of R. curvirostris resembles a hollow, cylindrical, more or less curved tube with a short and wide triangular spine on the curved portion of the valve. Jousé (1971) later described R. curvirostris var. inermis, a taxon that differed from the nominate variety by its lack of a spine on the curved part of the valve. Since this variety has an age range of late Pliocene—-early Pleistocene, Jousé considered it the ancestor of R. curvirostris var. curvirostris.
Later studies have shown that R. curvirostris var. inermis is a synonym of R. barboi Brun (Atlas of Microorganisms 1977). Akiba and Yanagisawa (1986) determined that the only difference between R. curvirostris and R. barboi is presence of a spine on the curved portion of the valve in the former species and its absence in the latter species. Jousé (1961) previously distinguished cold-water and warm-water diatom assemblages in core section 3361 in the North Pacific, over six horizons from middle Pleistocene to Holocene. Cold-water assemblages mainly contained arctoboreal and northern boreal species, while warm-water assemblages contained southern boreal species. Rhizosolenia curvirostris was
35
LOSEVA, EMMA
characteristic of all cold horizons from the middle (horizons VI and IV) to upper (horizon II) Pleistocene.
According to the stratigraphic scheme of the North Pacific, the duration of the Quaternary is 1.8 m.y. Within this interval three diatom zones are distinguished (Donahue 1970, Koizumi et al. 1980) upwards: Actinocyclus oculatus, Rhizosolenia curvirostris and Den- ticula seminae. Rhizosolenia curvirostris is the index species for middle Pleistocene of the North Pacific (Jousé 1969, Schrader 1973, Koizumi 1977) and northern Chukotka (Polijakova 1986), but the R. curvirostris zone extends up to the late Pleistocene in the North Pacific. This corresponds to the Brunhes magnetic epoch and is equal to the duration of the whole Pleistocene by the USSR schemes. The duration of the Quaternary is 0.8 m.y., and the durations of its subdivisions are—Holocene: 0-10 Kyr; late Pleistocene: 10-110 Kyr; middle Pleistocene: 110-380 Kyr; and early Pleistocene: 380-800 Kyr (Kras- nor and Zarrina 1987).
Stepanova (in Slobodin and Stepanova 1986) first found both R. curvirostris and R. barboi in upper Cenozoic sediments from the northern part of the Taimir peninsula and distin- guished the R. barboi zone in middle Pliocene in East Siberia. Later Dzinoridze registered both R. curvirostris and R. barboi in two sections in the northern part of the Pechora River lowland. She identified R. barboi by the absence of the spine on the valve curve and the straighter valve outlines than in R. curvirostris. In this way these two species were used for correlation between Pechorian and Siberian sections (Slobodin et al. 1986), resulting in the age of sediments of northeastern Europe to be considered more ancient than late Pliocene—early Pleistocene.
Results and Discussion
I had the opportunity to investigate five samples from bore-hole 1-Ja, north of Pechora River, which were provided through the courtesy of V. Zarchidze. Both light and scanning electron microscope observations were made. Observations revealed that valves are often split longitudinally, so that only one of the halves has the triangular (sometimes two-apex) spine while the other half lacks a spine (Figs. 1-12b). Therefore, halves without the spine can resemble R. barboi but originate from R. curvirostris frustules. Moreover, the degree of curvature is of no significance. This fact was not mentioned previously, although Jousé (1968) demonstrated R. curvirostris valves without the spine. Thus, positive identification of both species is possible only when one can see the whole valve. In my opinion, R. curvirostris was present only in the diatom assemblages in the samples from bore-hole 1-Ja. Hence, there is no reason to consider the sediment age to be more ancient. In the last few years single R. curvirostris valves or their fragments were registered in some marine early and middle Pleistocene diatom assemblages in northeastern Europe (Loseva, 1992). The upper age limit of this species together with Thalassiosira nidulus (Temp. and Brun) Jousé in this region is the shklovskoje interglacial of middle Pleistocene. Both of these species had become locally extinct before the moskovskoje glaciation, about 130 Kyr, while in the North Pacific their upper age limit is 276 Kyr (Sancetta and Silvestri 1984).
Thus, R. curvirostris is found in early and middle Pleistocene marine diatom assemblages of northeastern Europe, and its presence cannot be used to consider the age of sediments more ancient.
56
11th DIATOM SYMPOSIUM 1990
Literature Cited
AKIBA, F, AND V. YANAGISAWA. 1986. Taxonomy, morphology and phylogeny of the Neogene diatom zonal marker species in the middle-to-high latitudes of the North Pacific. Init. Repts. Deep Sea Drilling Project, 87:483-553.
ATLAS OF MICROORGANISMS IN BOTTOM SEDIMENTS OF THE OCEANS. 1977. "Nauka." Moscow. 32 pp.
DONAHUE, J.G. 1970. Pleistocene diatoms as climatic indicators in North Pacific sed- iments. Mem. Geol. Soc. Amer. 126:121-138.
JousE, A. P. 1961. Diatoms and their role in explaining ocean history. USSR Academy of Sciences, Geography Ser. 2:13-20
. 1968. New diatom species in bottom sediments of the Pacific and the Sea of Okhotsk. News Syst. Plant. Non Vasc. 5:12-21.
. 1969. Diatoms in Pleistocene and Late Pliocene sediments in the Boreal Pacific. Pp. 5-27 in Micropaleontology and organogenous sedimentation in the oceans. VI Congres INQUA. ‘‘Nauka,’’ Moscow. 288 pp.
. 1971. New and interesting species of diatoms from bottom sediments of the Pacific. News Syst. Plant. Non Vasc. 8:12-18.
KoIzuMI, I. 1977. Diatom biostratigraphy in the North Pacific region. Proc. Ist Con- gress. Pacific Neogene Stratigr. Tokyo, 235— 253.
KoIzuMI, I., J. BARRON, AND H. HARPER. 1980. Diatom correlation of Legs 56 and 57 with onshore sequences in Japan. Init. Reports. Deep Sea Drilling Project, 56-57: 687-693.
KRASNOV, I. I. AND E. P. ZARRINA. 1987. The interregional stratigraphic scheme of Quaternary sediments of east-European platform and its comparison with other geochronological scales. Pp. 19-23 in Cenozoic sediments origin and_ structural geomorphology of the USSR. XII Congress INQUA. Leningrad.
LOsEVA, E.I. 1992. Atlas of marine Pleistocene diatoms of northeastern European part of the USSR. "Nauka." Leningrad.
POLUAKOVA, E. I. 1986. Ecological and systematical characteristic of upper Cenozoic diatom flora of northern Chukotka. Pp. 79-83 in Urgent questions of modern palaeoalgogy. Kiev.
SANCETTA, C. AND S. SILVESTRI. 1984. Diatom stratigraphy of the Late Pleistocene (Brunhes) Subarctic Pacific. Mar. Micropaleontolgy 9:263—274.
SCHRADER, H-J. 1973 Cenozoic diatoms from the northeastern Pacific. Init. Reports. Deep Sea Dnilling Project 18:673-797.
SLOBODIN, V. Ya., N. I. DRUZHININA, AND G. V. STEPANOVA. 1986. Correlation of the Neogene marine sediments in the Arctic from biostratigraphical data. Pp. 196-198 in Reports of the 10th micropaleontological conference of the USSR. Zonal Stratigraphy based on Microorganisms and Methods of its Development. Leningrad.
SLOBODIN, V. Ya. AND G. V. STEPANOVA. 1986. Stratigraphy and paleogeography of the Cenozoic of Northern Taimir. Pp. 110-113 in Cenozoic of the shelf and islands of the Soviet arctic. Leningrad.
LOSEVA, EMMA
Explanation of Plates
Plate 1, Figures 1-8 FIGURES 1-8. Rhizosolenia curvirostris. FIGURES 1-6. LM x 1000. FIGURES 7, 8. SEM. External view. FIGURES 7a, 8b. x 1000. FIGURES 7b, 8a. x 3300
Plate 2, Figures 9a-f FIGURES 9a-f. Rhizosolenia curvirostris. SEM. Broken frustule. FIGURES 9b, d, e. x 1000. FIGURE 9a. x 2000. FIGURES 9c, f. x 3300.
Plate 3, Figures 10-12 FIGURES 10a—12b. Rhizosolenia curvirostris. SEM. Internal view. FIGURES 10a, b, lla, 12a. x 1000. FIGURES 11b, c. x 3300. FIGURE. 12b. x 4700.
11th DIATOM SYMPOSIUM 1990
59
LOSEVA, EMMA
60
11th DIATOM SYMPOSIUM 1990
Ti
61
11th DIATOM SYMPOSIUM 1990
Valve Morphology of Coscinodiscus janischii Schmidt (Bacillariophyceae)
by Martha E. Ferrario and Eugenia A. Sar
Division Ficologia, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata, Paseo del Bosque s/n 1900, La Plata, Argentina
with | plate
Abstract: }Coscinodiscus janischii Schmidt from Puerto Madryn (Chubut) and Puerto Brown (Tierra del Fuego), Argentina, was studied with light and scanning electron microscopy. Clear differences between C. janischii and similar species C. gigas Ehrenberg and C. wailesii Gran and Angst were observed. Our results show that C. janischii fits the description of Coscinodiscus sensu stricto, although some of the areolae have distinctive characteristics not previously observed in the genus.
Introduction
The genus Coscinodiscus Ehrenberg, with about 400 validly described taxa (Van Landin- gham 1968), has been revised extensively in the last two decades, based upon observations made with the scanning electron microscope. The result of these revisions has been that many Coscinodiscus species have been placed in other genera. In particular, Thalassiosira has been expanded through many published studies, including the early works by Hasle (1968) and Fryxell and Hasle (1974).
Simonsen (1975, 1979) established three morphologic groups within the genus Coscinodis- cus, based on the pattern of labiate processes. Rapid progress in developing a concept of the genus took place following publication of Simonsen’s hypotheses. In accordance with Ross and Sims (1973), Fryxell (1978) proposed Coscinodiscus argus Ehrenberg as the generictype. Further studies resulted in the splitting of some species of this genus into several others, including: Psammodiscus Round and Mann (Round and Mann 1980), Thalassiosiropsis Hasle (Hasle and Syvertsen 1985), Stellarima Hasle and Sims (Hasle and Sims 1986a), and Azpeitia M. Peragallo (Fryxell et al. 1986). Subsequent to the above studies, Hasle and Sims (1986b) emended the diagnosis of the genus Coscinodiscus. Final- ly, and with the purpose of introducing more precise limits among the species, Fryxell and Ashworth (1988) included additional ultrastructural diagnostic characters to the ones classically employed.
In this paper we study the valvar morphology of Coscinodiscus janischii Schmidt in order to analyze the position of this species in the genus C. sensu stricto. Comparisons are made with the nearest species, C. gigas Ehrenberg and C. wailesii Gran and Angst.
63
FERRARIO, MARTHA E.
Materials and Methods
Marine neritic phytoplankton was collected in 50 fm mesh plankton nets from two sites in neritic Argentinian waters: Puerto Madryn, Chubut, 29 Sept. 1981 (sample no. 3621), and Puerto Brown, Tierra del Fuego, 24 Feb. 1963 (sample no. 2396). The samples were preserved in buffered formalin. They were subsequently cleaned of organic matter (Hasle and Syvertsen 1980), and subsamples were mounted in Hyrax for study in a Wild M20 phase contrast light microscope. Other subsamples were mounted on glass coverslips at- tached to aluminum stubs, sputter coated with gold, and examined under a JEOL JSM T100 scanning electron microscope. The material was added to the collection of the Division Ficologia, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata as ‘‘Diatoms from Chubut’’ and ‘‘Diatoms from Tierra del Fuego.’’
The above material was compared with slides 91, 148, 182, 551, and 848 from the Tempére and Peragallo Collection, 2nd Edition (Tempére and Peragallo 1915) and with slide series 167 from the Frenguelli Collection.
Results
Coscinodiscus janischii Schmidt, 1878 (Pl. 64, Figs. 3, 4)
Cells solitary, coin shaped, 160-240 um in diameter, 35-40 [1m deep about the pervalvar axis. Valves flat, sharply elevated in the marginal area. Areolae pattern radial with spirall- ing secondary rows more or less evident (Fig. 1). Central area small, hyaline, without any aperture, sharply delimited in the interior part with randomly distributed siliceous granules and bundles of linear markings radiating from its margin (Figs. 3, 6).
Areolae ordered in complete and incomplete rows, the latter originating towards the mid- radius. Central areolae elongated in radial direction (Figs. 2, 3), 2.5-3.5 in 10 um. Areolae diminishing in size towards the margin, 44.5 in 10 um, becoming loculated and hexagonal only near the periphery (Figs. 4, 9). External cribra delimited by a solid siliceous covering except in the mantle where they are continuous (Figs. 7, 10). Pores of the cribra closed by very delicate sieves, the cribrella (Figs. 3, 8), which are often destroyed during cleaning.
Valve mantle vertical, possessing two rings of areolae with continuous cribra cover and siliceous ridge-like irregularities around their margins (Fig. 10). Microrimoportulae placed in one ring on the mantle, 24 areolae distant, 2 in 10 um (Fig. 12). Two macrorimopor- tulae, 120-135 ° apart, located in or close to the ring of microrimoportulae (Fig. 11).
Data about the distribution of this species along the coast of Argentina are reported else- where (Ferrario 1981).
COMPARISONS WITH OTHER MATERIAL
Coscinodiscus janischii from the Tempére and Peragallo Collection was scarce and general- ly broken. Specimens were markedly similar to the cells observed along the Argentine coasts, particularly with respect to the following: diameter, general view of the valve with a narrow outstanding marginal area with hexagonal areolae, areolae pattern, density and shape, and type of central area.
Comparison with the Frenguelli material (Frenguelli 1928) allowed us to establish clear differences with respect to general view of the valve, type of central area, and size and shape of the areolae in different parts of the valve.
64
11th DIATOM SYMPOSIUM 1990
Discussion and Conclusions
The analysis of some classical works, such as Schmidt (1878, pl. 64, Figs. 3, 4), Rattray (1890:543), Peragallo and Peragallo (1897-1908:432, pl. 118, Fig. 4), and Hustedt (1928:459, Fig. 257) allowed us to establish similarities between the observations on our material and the descriptions and illustrations of these authors. The comparison was made following the classic criteria that emphasized areolae size and array, presence or absence of a particular structure in the central area, and diameter and general aspect of the valves.
As can be seen from our results, we assume the material from Puerto Madryn and Puerto Brown to be conspecific with the material in the Tempére and Peragallo Collection. Based on this conclusion we consider the differences with Hustedt’s illustrations to be a difference in drawing, since this author takes into account the material from the Tempere and Peragal- lo Collection.
In accordance with the criteria of Rattray (1890) and Takano (1976) among other authors discussed, we have noticed that Coscinodiscus janischii easily can be confused in the light microscope with other large species, such as C. gigas and C. wailesii. Because of the similarities we think it useful to establish the most outstanding differences between these taxa.
Coscinodiscus gigas does not have the narrow and clear marginal area observable in the light microscope. Areolae are smaller in the center and increase in size towards the periphery. Microrimoportulae are closer to one another and located in the transition be- tween the mantle and valve surface (Takano 1976:134, Figs. 5, 6). Macrorimoportulae are relatively smaller and less coiled (Takano 1976, Figs. 14-16); Fryxell and Ashworth 1988, Fig. 28), and the mantle is deeper with a lattice pattern of hexagonal areolae (Takano 1976, Figs. 5, 6; Fryxell and Ashworth 1988, Fig. 24).
Coscinodiscus wailesii was recently examined by Schmid and Volcani (1983). According to their paper C. wailesii does not have a narrow marginal area (LM). Areolae are smaller, almost regular in size throughout the valve, and always hexagonal (Schmid and Volcani 1983:388, Figs. 2, 3). There are no solid siliceous coverings enclosing the external cribra of the areolae (Schmid and Volcani 1983, Fig. 30). The central area is larger with bundles of siliceous thickenings reaching the margin (Schmid and Volcani 1983, Fig. 2). Microrimoportulae have a very different distribution pattern, being scattered on the valvar surface as well as being present in a ring at the juncture of the mantle and valve and in an additional ring in the free border of the mantle (Schmid and Volcani 1983, Fig. 10).
The results obtained on the valvar morphology of C. janischii allow us to include this species in the genus Coscinodiscus sensu stricto, taking into account the limits established by Hasle and Sims (1986b) for the emended genus. However, the presence of poroid-like areolae (Fig. 5) on the surface of the material studied suggests the need to widen the limits of variability of areolar structure in the genus.
Acknowledgements
We are grateful to Dr. A.M. Schmid for her active and generous participation in the comparison between our material and C. wailesii.
65
FERRARIO, MARTHA E.
Literature Cited
FERRARIO, M.E. 1981. Diatomeas Centrales de la Ria de Puerto Deseado, Santa Cruz, Argentina. V. Fam. Coscinodiscaceae y Biddulphiaceae. Lilloa 35:115-123.
FRENGUELLI, J. 1928. Diatomeas del Oceano Atlantico frente a Mar del Plata (Republica Argentina). An. Mus. Nac. Hist. Bernardino Rivadavia 34:497—-572.
FRYXELL, G. A. 1978. Proposal for the conservation of the diatom Coscinodiscus argus Ehrenberg as the type of the genus. Taxon 27:122-125.
FRYXELL, G. A. AND T. K. ASHWORTH. 1988. The diatom genus Coscinodiscus Ehren- berg: characters having taxonomic value. Bot. Mar. 31:359-374.
FRYXELL, G. A. AND G.R. HASLE. 1974. Coscinodiscineae: some consistent patterns in diatom morphology. Nova Hedwigia Beih. 45:69—-96.
FRYXELL, G. A., P. A. SIMS, AND T. P. WATKINS. 1986. Azpeitia (Bacillariophyceae): related genera and promorphology. Syst. Bot. Monogr. 13:1-74.
HASLE, G.R. 1968. The valve processes of the centric diatom genus Thalassiosira. Nytt. Mag. Bot. 15:193-201.
HASLE, G. R. AND P. A. SIMS. 1986a. The diatom genera Stellarima and Symbolophora with comments on the genus Actinoptychus. Br. Phycol. J. 21:97-114.
1986b. The diatom genus Coscinodiscus Ehrenberg Bot. Mar. 29:305-318.
HASLE, G. R. AND E. E. SYVERTSEN. 1980. The diatom genus Cerataulina: morphology and taxonomy. Bacillaria 3:79-113.
1985. Thalassiosiropsis, a new diatom genus from the fossil records. Micropaleontology 31:82-91.
HustTeDT, F. 1928. Die Kieselalgen Deutschlands, Osterreichs und der Schweiz mit Beriicksichtigung der ubrigen Linder Europas sowie der angrenzenden Meeresgebeite. Pp. 273-478 in Kryptogamen-Flora von Deutschland, Osterreich und der Schweiz, L. Raben- horst, ed. Akad. Verlags., Leipzig.
PERAGALLO, H. AND M. PERAGALLO. 1897-1908. Diatomées Marines de France et des Districts Maritimes Voisins. Micrographe Editeurs, a Grenz-sur-Loing. 491 pp.
RATTRAY, J. 1890. A revision of the genus Coscinodiscus Ehrb., and of some allied genera. Proc. R. Soc. Edinburgh 6:449-692.
Ross, R. AND P. A. SIMS. 1973. Observations on family and generic limits in the Centrales. Nova Hedwigia Beih. 45:97-121.
ROUND, F. E. AND D. G. MANN. 1980. Psammodiscus nov. gen. based on Coscinodiscus nitidus. Ann. Bot. 46:367-373.
SCHMID, A.M. AND B.E. VOLCANI. 1983. Wall morphogenesis in Coscinodiscus wailesii Gran and Angst. I. Valve morphology and development of its architecture. J. Phycol. 19:387-402.
SCHMIDT, A. 1874-1944. Atlas der Diatomaceen-Kunde. R. Reisland, Leipzig.
SIMONSEN, R. 1975. On the pseudonodulus of the centric diatom, or Hemidiscaceae reconsidered. Nova Hedwigia Beih. 53:83-94.
—. 1979. The diatom system, ideas on phylogeny. Bacillaria 2: 9-71.
66
11th DIATOM SYMPOSIUM 1990
TAKANO, H. 1976. Scanning electron microscopy of diatoms. III. Coscinodiscus gigas Ehrenberg. Bull. Tokai Reg. Fish. Res. Lab. 88:133-141.
VAN LANDINGHAM, S.L. 1968. Catalogue of the fossil and recent genera and species of diatoms and their synonyms. Part II. Bacteriastrum through Coscinodiscus. J. Cramer, Vaduz. 494-1086 pp.
Explanation of Plate
Plate 1, Figs. 1-12
FIGURES 1-12. Coscinodiscus janischii. FIGURE 1. Valvar surface showing elevated marginal area and radial areolae pattern with spiralling secondary rows (internal view). Scale bar = 50 um. FIGURE 2. Central area showing randomly distributed siliceous granules and bundles of linear markings radiating from its margin (internal view). Scale bar = 10 um. FIGURE 3. Areolae around central area elongated in radial direction, showing polygonal cribral pores closed by cribrella (internal view). Scale bar = 5 bm. FIGURE 4. Marginal area with areolae becoming hexagonal only near the periphery. Scale bar = 10 um. FIGURE 5. Broken valve showing structure of areolae. Scale bar = 2m. FIGURE 6. Central area (external view). Scale bar = 5 um. FIGURE 7. Cribra of circular areolae delimited by solid siliceous covering (mid-radius, external view). Scale bar = 2 um. Figures 8-12: Scale bar = 5 um. FIGURE 8. Circular areolae (mid-radius, internal view). FIGURE 9. Detail of the loculate areolae in marginal region. FIGURE 10. Mantle showing a microrimoportula opening surrounded by continuous cribral coverage. FIGURE 11. External opening of a macrorimoportula surrounded by con- tinuous cribral coverage. FIGURE 12. Interior of mantle showing a macrorimoportula and the neighboring microrimoportulae.
67
FERRARIO, MARTHA E
ae
68
11th DIATOM SYMPOSIUM 1990
Genetic Structure of Microalgal Populations. I. Problems Associated With the Use of Strains as Terminal Taxa
by Jane C. Gallagher
Department of Biology, City College of C.U.N.Y. Convent Ave. at 138th St., New York, New York 10031
with 8 figures and 3 tables
Abstract: Two approaches to sampling have been developed for the examination of the distribution of characters for phylogeny reconstruction and the classification of diatoms and other microalgae. The first is the traditional microbiological method where a small number of strains isolated from widely spaced environments are compared. This approach is used mostly for physiological and molecular characters. The levels of divergence among individual strains are then interpreted in relation to differences in similar parameters obtained from terrestrial organisms. The alternative approach is the population method where large numbers of strains are examined from both contiguous and widely spaced environments. This approach is commonly used for studies of morphology, but rarely for molecular data. Thus, the difference between morphological and molecular studies of phylogeny is not simply due to different types of characters, but is also due to different sampling strategies. The effect of these two sampling strategies on interpretations of evolutionary relationships was explored by analysis of a data set available for the temporal and spatial genetic structure of the Narragansett Bay populations of Skeletonema costatum (Grev.) Cleve. Cladistic and distance measures were compared. The results demonstrate that if all 457 strains are compared in an analysis using the strains as terminal taxa, then it is not possible to recover the patterns of seasonal genetic variation. It is only when groups of strains are compared that patterns can be discerned. Recom- mendations are made for sampling strategies for molecular evolutionary studies of diatoms.
Introduction
The genetic structure of microalgal populations and species is the result of the combined effects of selection, drift, migration, and mutation, which are mediated by the historical and biological factors particular to each taxon. The pattern of genetic structure determines, in part, the taxon’s responses to changes in its environment (Loveless and Hamrick 1984).
Over the last 30 years intraspecific variation has been documented for almost all parameters hypothesized to govern the ecological success of organisms (see Gallagher 1986; Wood 1987; Brand 1990, for reviews). The results of virtually all of these investigations support the hypothesis that the genetic diversity within and among species of microalgae is greater than that found within and among species of higher organisms (Olsen 1990). For example, 30-60% of enzyme loci tested by gel electrophoresis are monomorphic (invariant) among
69
GALLAGHER, JANE C.
species within genera of higher organisms (Avise 1975; Ferguson 1988). However, no monomorphic enzyme loci have ever been detected among strains within species of microalgae.
Although the existence of intraspecific variation is well-documented for every type of microalga, the spatial and temporal scales of differentiation have rarely been quantified (see Gallagher 1986; Brand 1990, for reviews). This lack of information is largely the result of sampling methods developed for microorganisms. Traditional microbiological sampling methods examine one or two strains isolated from several distantly spaced environments. These strains are tested for one or more parameters, and the results are usually extrapolated to the entire population present in the ecosystem from which the strain was obtained. This approach treats strains as terminal taxa that are assumed to be representative of the population from which they were isolated or of a phylogenetic group. Natural selection is usually invoked to account for differences among the putative groups that the strains are supposed to represent. In some cases the pattern of differences among strains is interpreted to indicate species level differences and new taxa are described. In other cases, the differentiation among strains shows no discernable pattern, and these data are used to cast doubt on morphological classifications (e.g., Hayhome et al. 1989). The presence of intragroup diversity confounds such conclusions. For example, consider the case in which a single, genetically diverse population extends over a wide area. The presence of genetic differences between two strains isolated from different locations might lead to the conclusion that there are two distinct populations. In this case, the discontinuity among small numbers of strains obtained using a microbiological sampling strategy can force interpretations of ecological or evolutionary discontinuities where none exist.
A sampling strategy based on the use of single strains as terminal taxa is fundamentally different from the population methods used for morphological analyses. The use of single strains as terminals is propelled by the current interest in the use of molecular methods in systematics and population biology because many of these methods are too expensive and time-consuming to apply to large samples. Thus, the difference between morphological and molecular studies is not simply one of comparing different types of characters, but is also one of comparing different types of sampling strategies. This difference can create serious problems in the interpretation of data. The goal of this paper is to illustrate some of these problems and to suggest some solutions.
The effects of large versus small sample sizes and the use of populations versus strains as terminal taxa can be illustrated by an reexamination of the data set available for allozyme diversity in the Narragansett Bay populations of Skeletonema costatum (Grev.) Cleve. The banding patterns of 457 strains of this species were examined, and it was demonstrated that the frequency of different types of strains varied with season (Gallagher 1979; 1980). The extreme genetic differentiation between the summer and winter forms is mediated by the existence of genetic and physiologically intermediate populations that are most abundant in the interbloom period (Gallagher 1982). A recent reexamination of some unpublished data also demonstrates that space is also important in determining population structure and that these two variables are not independent (Gallagher, in prep.). These studies used populations as the terminal taxa and the patterns of differentiation among them are shown in Figure | for a cladistic analysis of the data set. The effect of using single strains as terminal taxa can be determined by resampling the individual ermine if some or all of the patterns shown in Figure | can be recovered.
70
11th DIATOM SYMPOSIUM 1990
OUTGROUP
SrB7os1
WB76s2
WB77s4
WPB77s2
WPB76s2
OS°O=14 > SS O=19) Cll =the,
SPB/6s2
SPB75s3
SB/5s | 9876s 1a SB/76s2 SPB75sZa SB79sZ 5B/5sla
SB75s2a
L SB75s3
FIGURE 1. Cladistic analysis of populations of Skeletonema costatum in Narragansett Bay, R.I., using the independent allele model and presence/absence coding of dif- ferent alleles in a sample. Sample coding: SPB = summer prebloom samples; WPB = winter prebloom samples; SB =summer bloom samples; WB = winter bloom samples; 75 = 1975; 76 = 1976; 77 = 1977; sl, sla, s2, s2a, s3 =stations 1 through 3 from north to south within the bay. For example, SPB75s1 is the summer prebloom sample for 1975 at station 1. Strict consensus tree of 10 equally parsimonious trees obtained with successive weighting.
71
GALLAGHER, JANE C.
Materials and Methods
The details of sampling the strains and electrophoresis are described in detail in Gallagher (1980, and in prep). Briefly, individual strains were isolated by micropipet from unenriched natural seawater samples and were each placed in separate test tubes of f/2 medium (Guillard and Ryther 1962). The sampling strategy was to obtain sample sizes of ap- proximately 50 strains from a mid-bay station two weeks prior to and during the winter and summer blooms of S. costatum for two consecutive years starting in the summer of 1975. During the summers of 1975 and 1976, additional samples were taken in the bay at different locations to ascertain the effect of spatial differentiation on population structure. Great care was taken to treat all isolates in an identical manner and for most samples the survival of isolates was very high. This high survival rate validated the extrapolation of frequencies of different types of strains in the samples to the natural populations from which they were isolated.
Strains were categorized according to their allozyme banding patterns at five enzyme loci: phosphoglucose isomerase (PGI), malate dehydrogenase (MDH), glutamate dehydrogenase (GDH), and two loci of superoxide dismutase (TO1 and TO2). Strains were grown for electrophoresis under identical conditions regardless of the initial conditions at the time of isolation in order to eliminate environmental effects on gene expression. In addition, a selection of genetically different strains was grown under a variety of conditions to determine if banding patterns changed with environment. Only invariate loci were used. Under these specific conditions, the differences in banding patterns among strains were assumed to reflect underlying genetic differences. Banding patterns were interpreted as alleles at each locus and were converted into genotype by assuming that S. costatum is diploid. A previous analysis of this data set divided the strains into broad groups and the terminology is retained here for convenience (Gallagher 1982). The WH and Whet groups have the most common alleles present in winter populations at all four seasonally variant loci. These groups are also minor components of summer populations. The SH group has the most common summer bloom alleles at all four loci that show changes in frequency with season. The “‘Mixed’’ and ‘‘Rare’’ groups are heterogeneous groups that have com- mon summer alleles at some loci and winter alleles at the remainder. They are physiological and genetic intermediates between the extremes represented by the summer and winter blooms (Gallagher 1982).
For cladistic analyses, each band (allele) was coded as a separate, individually segregating character whose presence or absence was scored without assigning it to a locus. Homology among bands found in different organisms and samples was based on identical mobilities (Rf values) and reactivity using a particular staining protocol. This coding approach has been termed the Independent Allele Model (Mickevich and Johnson 1976; Mickevich and Mitter 1983; Buth 1984). Forms of this coding method have also been used in many phenetic interpretations of banding patterns in microalgae (e.g., Soudek and Robinson 1983; Hayhome and Pfiester 1983; Cembella et al. 1988; Hayhome et al. 1989). Cladograms were calculated using the Hennig86 program (Farris 1988). Trees were rooted using an all zero outgroup. This method maximizes the overall congruence of characters in the ingroup. The fit of the input data to the cladogram is measured by the consistency index (CI) which is calculated as:
Ch = Si/Pi
where Sj 1s the number of steps for character 1 implied from the data and Pji is the number of steps in the tree. The overall CI for the tree is the sum of character ranges divided by
11th DIATOM SYMPOSIUM 1990
the total number of steps in the tree (Farris 1989). The retention index (RI) is a measure of the degree to which a character is synapomorphic and is calculated as:
RI]j = (hi-si)/(hi-li)
where hj is the maximum number of steps for character i, s is the observed number of steps in the tree and | is the minimum number of steps (Farris 1989). A character that is an autapomorphy (unique to one taxon) has an RI of 0 and one that is perfectly synapomor- phic and has no homoplasy has an RI of 1.0.
Cladograms were calculated using both unweighted characters and with the successive weighting protocol of Hennig86 in order to determine which technique gave greater resolu- tion (Farris 1989). The latter technique has the advantage of providing a means of basing groups on more reliable characters without making prior decisions on weighting. Weights are calculated from the best fits, as determined by the product of the character consistency and retention indices and are then scaled to lie in the range 0-10 (Farris 1988).
Distance coefficients were calculated between pairs of strains using Jaccard’s coefficient (Sneath and Sokal 1973). This is a simple matching coefficient calculated by:
Sj = [a/(a + u)] x 100
where u is the number of mismatches and a is the number of matches between the banding patterns. This was converted to a distance coefficient by subtracting it from 100.
The distance matrix was used as input to a distance tree calculated by the Fitch-Margoliash method (program FITCH in the PHYLIP package) (Fitch and Margoliash 1967; Felsenstein 1985). This method constructs a tree by successive addition of each sample and then attempts to optimize the overall fit of the distances among samples in the tree to the distances in the input matrix by rearranging the branches. The fit is assessed by minimizing the sum of squares of the differences between the distances represented in the tree and the input distances. The original Fitch-Margoliash method allowed the tree to be fit with negative branch lengths; the Felsenstein implementation uses a minimum branch length of zero. The position of a midpoint root, the point midway between the two most distant taxa, (Farris 1972) was calculated for each tree.
Results
A companion paper (Gallagher, in prep.) provides the details of a comparison of methods of data analysis using groups of strains (populations) as terminal taxa and a variety of coding methods. This analysis showed that cladistic analyses using the Independent Allele Model and successive weighting was effective in analyzing the data. Fitch-Margoliash trees had the best fits for the distance analyses and were used to build trees here. Phylogenetic systematics offers a better system for the analysis of different types of data and is recommended. However, the distance analyses were included here in order to demonstrate that the problems inherent in the use of single strains as terminal taxa are common to both methods of analysis.
Several sets of analyses were run using subsamples of the Narragansett Bay data set. Seven are shown here to illustrate the general pattern of results.
Figure 2 illustrates a cladogram of the most parsimonious tree calculated using 13 randomly chosen strains from the summer prebloom sample of 1975 as input. Successive weighting improved resolution of the tree. Two major groups were identified. The larger of the
3
GALLAGHER, JANE C.
L=129, CI=0.87, RI=0.88
—=(—=OUTGROUP
25 5==ITA49 SPB Mixed —=23 i 2—— Te As SPB Mixed (ey a paere SPB Mixed 11—TIIA9 SPB Rare pres —rrass SPB Mixed pa SPB Rare (ee eee SPB Mixed — Ji PeStiass SPB Mixed poe SPB Mixed 21 (—1—I13 SPB Mixed 17 a SPB Mixed
Eis] l(GmeS SPB WH
ey gles 355 SPB WH
FIGURE 2. Most parsimonious tree calculated using strains from the summer prebloom of 1975 from station 2a and successive weighting. The designation beside each strain is the season of isolation (SPB) and the genetic group of each strain according to the classification of Gallagher (1982) (see text).
L=138, CI=0.93, RI=0.93.
poe E=18 ; 27 BS
1 wy 2==3B1 (—I=13 1 glen O==WR2 5 poe 15 1=IIAl
kL, sfle—ue2
11—1B3 I ca 135; —4—HB1 —=6—=UP 1 —7]=—=ED1
FIGURE 3. Strict consensus of two equally parsimonious trees obtained using the first stain isolated from each seasonal sample as input and successive weighting.
74
11th DIATOM SYMPOSIUM 1990
L=163, €I=0.80, RI=0.90
-=0—=OUTGROUP -=425 5==11A49 SPB Mixed Aa ij 10—=I1A69 SPB Mixed gol s—3n1 SH SH
FIGURE 4. Most parsimonious tree obtained using the strains included in Figures 2 and 3 and successive weighting.
two clades subdivides at nodes 21 and 22 into two subgroups. The Mixed strains were found in all of the clades.
Figure 3 shows a cladogram of the consensus tree calculated using successive weighting and the first strain isolated from each seasonal sample as input. Two main clades emerge from the analysis. The first at node 17 is from the summer bloom and the second is comprised of mixed and WH types. The strains are clearly separated by genotype, but not by time of isolation. Strains with the most common winter genotypes are also present in summer samples, and two of them were included in this analysis by chance. Strains 13 and WR2 were included in both of the analyses depicted in Figures 2 and 3.
Figure 4 illustrates a cladogram of all of the strains used in the first two analyses. The complete data matrix for this analysis is shown in Table 1. In this cladogram, two main groups again appear. However, the topologies are not simple additions of the first two cladograms. The position of various strains changes dramatically. The nature of the changes can be illustrated by the positions of strains I3 and WR2. In Figure 3 these strains were
75
GALLAGHER, JANE C.
S60 =1e8 ‘1I90=19 ‘001 = WUT qndur se jas wep Avg NasuvseueN aus oy) pue SUNYSIIM JAISsQooNS Susn poureiqo saaq} snoruowts -ied Ajjenba gg] jo sns -uasuod IMIG “¢ JUNO
HS
HS
HS HS HS
HS
HM
€SdN—=b 2c IT alla
9a1I=001>] san=66> gtas—=86= 0SLS=9 b= exede—=6 b= 9TXIWI=6 PS tias=crs z7san—=2bS expavZ—=1bS 9zx0Tav7—=0b> 2X6 TLS=6E5 €x9zaz7—=8E zWI=LES gav7—=9E> ZIXPlae—=sSe= 8bXPWI=beS Sxlde—=f £5 Tsan—=ze= 2xLiis=tc= execaz—=0F] LIXZ71d@—=6 75 8hdn—=87=>
2x9 bdS=L7S '94S=97=5 exzzaz7—=S7S 6pVII=r>
Zbaz—=C= EIT isaz=zl | Teqa—=Ls sTx0Tda—=955 stags Tpas=ps> 9taa=Es> Z1aW=2S> sSdS=9=-b TIS sas—=s=! 6 bpHH=LOTS] €7vII=095 abAN—=6S= suas )9 TT ZuM=1— gSvII=b6=-LIIS s¥II=0s— | exEsda—=1spST TS ssag—=6 b= 2x Tpak—=0 TT} bbaN—=601S TTAN==8 01> €TAN=L6S excaI—=96= Lz7qH#=s6> ZXZTAN==€6>|
€Las=z64 7*6Z2IN=165
HM bz2x61da—=06 9*0z7da=68=
bSWII=885
Lpqa==Le>
pextw SXLSGH=9 85 eXLAN=S8
9fda—=b B=
EXSTIN=C 8S
z7SaH—=z8=
HM 89xz71dg—=18 sx1dg—=08
z7eaH=6L>
plas=8l=
€baH=LLS
HM LX6TEH=9 1S LEGH=SLS
€7aH—=b LS
ZXOTHH=E1S
Stah=2L>
ZX0SVII=1L>
£TaN—=0L>
sTas—=6954
HM 8exTIaI—=B95 qeum ETXLTGI=LI= 8SEH=99=|
2%6 bIN—=S9S|
2XLETN—=b9>
HM 9x9qg—=f9> faH=295
LdI=—19> TdS==8 4 O0¢@LS==L a OVAN See) 0eLs=1?7S eSdH—07> cI=61>= bbds—=81S ExCONC LI) 7xGSVII=9 Sal LXx9GH=S TS bpAN=PIS 6aH=€ ra €x9EWII=21I> 6bdN—=11S 7XZ2EAN—=01S beqH—6> LSis=85 Tav7=1>
anowoLno—=90—4
76
11th DIATOM SYMPOSIUM 1990
Allele F 0 0 0
1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0
3 x — ~ c vo uo) 2 oO vz =I = o xo =) of = a =) = = = = S =) S 8 S v = S o~4 Lv 3 n Ge io} a = Ss G =) a =) io} 2 n > od c s oy s AI uo} ay o = & ia 5 S = Ss aI Ss Qa
5
iS = 4 a |] & £9$$9498S8e9ececc pape
fe > << 3 < <HrOre Ne
a | 4 ens SSSSSSSs3S 8585590558 =
77
GALLAGHER, JANE C.
TABLE 2. Inferred genotypes of strains used in the distance analyses. Each letter refers to an allele at each locus.
Locus Strain PGI MDH GDH TO TO2 IIA1 BB KK HH UU RR 2B3 AA MM GG XX PP NY1 BB KK GG UU RR HB1 BB KK GG UU RR SF1 AB KK FF UU RR UPI BB KK GG UU RR ED1 BB KK GG UU RQ MF2 BB KK EE UU RR 13 AA KK Il TU PR WR2 AA KK FiFi UU RQ 3B1 AA MM GG VV PE. 1B3 BB KK II UU RR IIA4 AA MM GG WwW PP ITA9 BB JK HH XX PP IIA10 AA MM GG WW PP IA11 BB KK GG UU RR IIA12 AA MM GG WW RE
at the base of the second clade, whereas in Figure 4 they are the unresolved sister group of the WH strains. Similar changes in branching order can be seen by the change in position of strain I[A36 and IIASO. In none of these three cladograms can patterns of change with season be clearly delineated.
Figure 5 depicts the consensus cladogram of all of the 457 strains using successive weight- ing. This cladogram must be regarded as an approximate result because so many equally parsimonious trees were generated that the capacity of the program was exceeded. For this analysis categories of strains with more than one representative were tallied together. For example, one strain had a banding pattern identical to that of strain NY32 and both strains are represented as NY32x2 at node 10. The total strains fell into 112 categories, 46 categories had more than one strain and 19 categories had more than five strains. The total analysis had very little resolution, and many of the groups of strains in the most common winter category appeared as unresolved sister taxa. Several strains with only a single representative grouped together. Most the the common summer bloom categories (SH) also appeared as unresolved sister taxa at node 113. Although the summer and winter categories show a weak separation, there is no clear evidence of seasonal changes in frequencies. The consistency index declined from 0.87 in Figure 2 to 0.61 in Figure ae
Figures 6-8 show distance trees calculated for successively increasing numbers of strains. The genotypes of the strains and the input distance matrix are listed in Tables 2 and 3, respectively. Figure 6 depicts the tree for the first 10 strains on the list; Figure 7, the first 12 strains and Figure 8, the entire data set. The average percent standard deviations for the three trees are 10.3, 13.1, and 24.8, respectively. The same strains were used in the distance analysis shown in Figure 7 and the cladistic analysis in Figure 3. The two analyses illustrate different patterns of relationships that can be seen by comparing the
78
11th DIATOM SYMPOSIUM 1990
I]
i
II
Il
I]
ss Ol OF Ol Ol 08 Ot a Of 0 0 08 0c 9 Or Of O€ Of rs 0 08 0c aa 08 0c ao OO CIV IV OV 6V pV cdl Ide CUM tI cAW Idd IdN 14S 1qdH TAN td IVII
aT
“AINSLALU DOUP}SIP BO} 1! UAAUOD 0} YO] WOIJ payors JUdIdIJJa09 s,pswdo"L SuIsn XWeW JURISIP paye[No|eD,
€ ATAVL
79
GALLAGHER, JANE C.
EB Del UP | NY | HB | MF 2
2BS
IlA|
_ _!0_ ,
FIGURE 6. Distance tree obtained using the first 10 strains classified in Figure 3. The position of the midpoint root is designated by the arrow. Thescale bar represents 10 % divergence.
position of strain SF1 to the other strains. The difference in results is due to the calculation of distance coefficients compared to allelic distributions.
Figure 8 illustrates the addition of the first five strains isolated from the summer 1975 prebloom sample to the strains analyzed in Figure 7. Comparisons of Figures 6-8 illustrate the changes in topology among trees due to the addition of taxa. This can be illustrated by examining the relationship of strain SFI in all three analyses. In Figure 6, it is the sister taxon to 2B3; in Figure 7, it is the sister to 2B3 and 3B1. In Figure 8, its position has shifted to a location basal to the branch leading to WR2 and [3. Comparisons of the tree distances between taxa also change with the addition of taxa. For example, the input distance from strain SF1 to strain 2B3 is 30 (Table 3). The tree distances are 38, 40, and 72 for figures 6-8, respectively. The input distance between SF1 and ED1 is 40. The tree distances are 46, 37, and 35, respectively, for the three analyses. The positions of the calculated midpoint roots for all three distance trees are shown. In all cases, the root falls between strain 2B3 and the next adjacent taxon, with strains isolated in the summer on both sides of the root. Discussion
None of the analyses using single strains as terminal taxa recovered the seasonal patterns detected in the original analysis (Gallagher 1980). In both the cladistic and the distance analyses, the addition of taxa to the tree changed topologies among all of the taxa. This was an expected result. The pattern of relationships revealed by both cladistic and distance analyses is based on the overall fit of all of the taxa in the matrix to the analysis. Adding taxa changes the relationships.
80
11th DIATOM SYMPOSIUM 1990
[ IB3
3B |
ED | HB | Oar
NY | IAI
FIGURE 7. Distance tree obtained using all of the strains shown in Figure 3.
In the cladistic analysis, the increase in sample size was also associated with an decrease in the consistency index. The pattern of decreasing CIs with the addition of taxa to an analysis is common to all cladistic studies (Sanderson and Donoghue 1989). Above the species level, this phenomenon has been attributed to increased detection of parallelisms and convergences (homoplasy) in the data set (Sanderson and Donoghue 1989). Here, at the population level, increasing the number of strains increases the probability of detecting the reticulate relationships among characters caused by interbreeding. As was noted in Gallagher (1982), out of the 457 strains isolated in the Bay, 103 fell into the Mixed and Rare categories of strains that have common summer alleles at some loci and winter alleles at other loci. Another 41 strains fell into the Whet group that is heterozygous at one locus for some common winter alleles. Therefore, 32% of the strains isolated showed evidence of genetic introgression between the summer and winter extremes. Increasing the sample sizes improves the possibility of including these strains in a data matrix. A simple parsimony analysis results in unresolved consensus trees when strains are treated as terminal taxa because a large number of equally parsimonious trees are generated to account for the multiple associations of alleles at different loci. Each allele must evolve multiple times on different branches. This accounts for the increase in the detected homoplasy of the data set and the decline in CI with larger sample sizes. Successive weighting improves the resolution because it gives the characters that co-occur most fre- quently higher weights with each round of analysis until the tree stabilizes. Co-occurrence of characters at the population level is caused by linkage disequilibrium. This is the non-random association of particular alleles at some loci and alleles at other loci. In diatoms, the combination of asexual reproduction and sexual reproduction with the pos-
81
GALLAGHER, JANE C.
IAQ ILAIO HAI2 HAF 3B|
FiGuRE 8. Distance tree obtained using all of the strains in Figure 7 plus the first five strains isloated in the summer prebloom of 1975.
sibility of both outbreeding and selfing can create linkage disequilibrium in three ways: (1) linkage on the same chromosome, (2) non-random association of gametes, and (3) differential asexual growth of different genotypes. Successive weighting is then able to resolve some of these groups. However, even though the differentiation among the common summer and winter forms is extreme, there is some heterogeneity within these groups. Locus TO] has three common summer alleles and three different common winter alleles. The GDH locus is diverse at all times of year and no seasonal differences in frequencies are present. This intragroup diversity then causes the summer and winter groups to frag-
82
11th DIATOM SYMPOSIUM 1990
ment into different common subgroups when the strains and alleles are examined inde- pendently. This prevents resolution of clear seasonal patterns in the data set.
There is a disturbing aspect of these analyses. When very small numbers of strains are used in the analysis, the probability of identifying the reticulate nature of relationships is reduced. The cladograms based on these small sample sizes thus appear to be very well- supported relationships and appear to reveal patterns, and the addition of larger sample sizes gives the impression of confounding the relationships. In other words, it appears that increasing the sample size degrades the results rather than improves them. In fact, this is really not the case. Larger samples simply reveal the reticulate nature of the character distributions of S. costatum and support the conclusion that the summer and winter blooms are simply opposite extremes of a genetic continuum rather than separate species.
The problem of decreasing resolution with increasing sample size is not a problem with cladistics. Distance analyses also show decreasing resolution. Although it is not possible to calculate a distance tree with the PHYLIP program for all of the taxa, it is possible to run enough analyses to determine that adding taxa creates a ‘“‘bushy’” structure with calculated branch lengths that deviate from the lengths in the input distance matrix. Many branches with zero lengths are encountered as taxa are added and interpretation of pattern becomes increasingly more difficult. Distance trees suffer a disadvantage compared to cladistic analyses with successive weighting. A distance tree is based on pairwise cal- culation of distance coefficients and the allelic distributions are not used directly. For example, in Table 2, strains SF] and WR2 have a distance coefficient of 0.40, as do strains I3 and 1B3, but each of these pairs of strains has different alleles. There is also no way to weight different branches. Although different distance coefficients exist that take into account frequencies of alleles at different loci, these coefficients are meant to be applied to populations, not individual genotypes as is used here. In any case, they all share the problem of discarding character information. Therefore, if the only way to resolve recurrent groups is to use linkage disequilibrium, then a distance analysis cannot succeed. An additional problem with distance trees is determination of a root. For many diatoms, determination of an outgroup is unclear. Some investigators have used a midpoint root as an alternative (e.g., Cembella et al. 1988). The validity of this assumption is based on equal amounts of divergence among branches of the tree. In the current analysis the position of the root is on a long branch and there is no reason to assume equal rates of evolution. No pattern is discernable because strains isolated in the summer occur on both extreme ends of the tree.
It has been suggested that the summer and winter populations of S. costatum be divided into different species based on the large genetic distance between them compared to higher organisms (Gallagher 1980). The current analysis shows that that conclusion is not supported by the allozyme data nor morphological differences (Gallagher, 1980). Large amounts of genetic diversity at the population-level have been reported for bacteria also (McArthur et al. 1988) and for other protist groups. Clearly, empirically derived coeffi- cients of taxonomic divergence developed for higher organisms do not apply to unicellular ones.
Could the addition of more molecular characters improve the resolution of the tree and result in the division of the Narragansett Bay populations of S. costatum into different species or varieties? Potentially, they could. However, these characters would have to be chosen with caution. One of the strengths of enzyme electrophoresis is that it examines variation in the nuclear genome at randomly chosen loci. Many molecular techniques, such as sequencing, usually examine variation only at a single locus. At lower phylogenetic levels, particularly in the presence of linkage disequilibrium, patterns of variation in a
83
GALLAGHER, JANE C.
single gene can give a false impression of divergence. This has been termed the “‘gene tree vs. species tree’’ problem (Pamilo and Nei 1988). For example, consider the problem of examining variation in chloroplast DNA (cpDNA) as an additional set of characters. For most organisms, cpDNA is inherited uniparentally and is therefore not subject to recombination. This means that it behaves genetically as if it were a single gene. Therefore, the differences among cpDNA patterns obtained from different strains are estimates of the divergence in the chloroplast DNA lineages, and are not necessarily representative of the entire genome of the taxa in question. Consequently, the pattern of cpDNA differences must be evaluated relative to markers developed for nuclear genes in order to determine the true patterns of differences. A similar problem can be found for sequencing multicopy genes, such as 16s-like ribosomal DNA. These genes do not evolve independently, but evolve together in a process called concerted evolution (Appels and Honeycutt 1986). When a mutation is introduced in one of the copies of the gene, it can propagate through the other copies to homogenize the gene family. Similarly, heterogeneous copies of the gene that are combined in one genome as a result of mating genetically distinct individuals will eventually homogenize. The rate at which this homogenization process occurs is not known. Microheterogeneity among copies of 16s-like genes in a position that is highly conserved in other organisms was noted in one strain of S$. costatum by Medlin et al. (1988) and could possibly be additional evidence for introgression. However, as concerted evolution progresses over time, the microheterogeneity at that site may be lost.
What recommendations can be made, then, for solutions for these problems? The first is to use groups of strains as the terminals. This approach is described in detail for the Narragansett Bay data set in a companion paper (Gallagher, in prep.). These groups need not be large enough to generate allele frequencies, presence/absence data for the groups seems to be sufficient and were used to generate the cladogram shown in Figure | (Galla- gher, in prep). Estimates of the minimum number of strains per group could be determined by a simulation based on resampling the S. costatum data set at random to attempt to recover the seasonal patterns. For other analyses, the characteristics of the groups could be defined by hypotheses of relationships, i.e., for a study of morphological divergence groups should be composed of strains with different morphologies. For a biogeographical study, groups of strains from different regions could be compared. For all of these studies a cladistic approach to the data will have the greatest utility because it has the best methodological and philosophical foundation for incorporating different types of data into a single analysis or series of related analyses.
In order to optimize a sampling strategy for labor intensive molecular analyses, a hierar- chical approach of using low resolution methods with larger sample sizes to determine sample sizes for higher resolution techniques can be taken. For example, the pattern of restriction fragment polymorphisms for a larger sample size can be used to select strains for more intensive sequencing studies.
It is necessary that phycologists work together to generate enough data sets so that potential benchmarks for species-level and population-level amounts and patterns of molecular divergence for different algal groups can be developed. It is clear that empirical benchmarks developed for higher organisms do not apply to microalgae. For example, in higher plants, chloroplast DNA (cpDNA) rflps usually vary only at the generic level. In diatoms, they vary at the population level (Stabile et al. 1990).
Finally, diatomists need to develop a molecular equivalent of type specimens. One of the great challenges of molecular biology is that new techniques are being developed at a rapid pace. It is extremely difficult to relate the characters used by different studies to each other without applying them to the same set of specimens. Therefore, wherever
84
11th DIATOM SYMPOSIUM 1990
possible it is desirable to include universally available strains in an analysis to provide reference points. Many of these strains can be obtained from the various worldwide culture collections. Strains can also be deposited in the collections. When this is not possible, it is recommended that DNA from the strains be saved for analysis by others. It is hoped that these recommendations would simply make the philosophy and approaches taken to molecular data more similar to that taken to morphological data.
Acknowledgements
This study was supported, in part, by National Science Foundation grants DEB80-21744 and OCE-8809484 and by PSC-CUNY grants 665155 and 669175 from the Research Foundation of the City University of New York. Bruce Huber helped with PHYLIP, read the manuscript, and provided valuable comments.
Literature Cited
APPELS, R. AND R.L. HONEYCUTT. 1986. rDNA: Evolution over a billion years. Pp. 81-137 in DNA Systematics, S. K. Dutta, ed. CRC Press. Boca Raton.
AVISE, J.C. 1975. Systematic value of electrophoretic data. Syst. Zool. 23:465-481.
BRAND, L.E. 1990. Review of genetic variation in marine phytoplankton species and the ecological implications. Biol. Oceanogr. 27:236-245.
BuTH, D.G. 1984. The application of electrophoretic data in systematic studies. Ann. Rev. Syst. Ecol. 15:501-522.
CEMBELLA, A. D., F.J.R. TAYLOR, AND J.C. THERRIAULT. 1988. Cladistic analysis of the electrophoretic variants within the toxic dinoflagellate genus Protogonyaulax. Bot. Mar. 31:39-51.
FARRIS, J.S. 1972. Estimating phylogenetic trees from distance matrices. Am. Nat. 106:645-668.
—. 1988. Hennig86 Reference. Version 1.5.
1989. The retention index and the rescaled consistency index. Cladistics 5:417—
420.
FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791.
FERGUSON, A. 1988. Isozyme studies and their interpretation. Pp. 184-201 in Prospects in Systematics, D. L. Hawksworth, ed. Systematics Association, Clarendon Press, London. FITCH, W. M. AND E. MARGOLIASH. 1967. Construction of phylogenetic trees. Science 155:279-284.
GALLAGHER, J.C. 1979. Population genetics of Skeletonema costatum in Narragansett Bay. Ph.D. thesis. U. Rhode Island. 232 pp.
. 1980. Population genetics of Skeletonema costatum in Narragansett Bay. J. Phycol. 16:464—474.
1982. Physiological variation and electrophoretic banding patterns of geneti- cally different seasonal populations of Skeletonema costatum (Bacillariophyceae). J. Phycol. 18:148-162.
GALLAGHER, JANE C.
1986. Population genetics of microalgae. /n Algal Biomass Technologies: An Interdisciplinary Perspective, W.R. Barclay and R. P. McIntosh, eds. Nova Hedwigia 83:6-14.
Genetic structure of microalgal populations. Il. Time-space interactions in Skeletonema costatum (Bacillariophyta) in Narragansett Bay, R. I. (submitted).
GUILLARD, R.R.L. AND J.H. RYTHER. 1962. Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran. Can. J. Microbiol. 8:229-239.
HAYHOME, B.A. AND L. A. PFIESTER. 1983. Electrophoretic analysis of soluble en- zymes in five freshwater dinoflagellate species. Am. J. Bot. 70:1,165-1,172.
HAYHOME, B. A., D. M. ANDERSON, D. M. KULIS, AND D. J. WHITTEN. 1989. Variation among congeneric dinoflagellates from the northeastern United States and Canada. Mar. Biol. 101:427-435.
LOVELESS, M.D. AND J. L. HAMRICK. 1984. Ecological determinants of genetic struc- ture in plant populations. Ann. Rev. Syst. Ecol. 15:65-95.
MCARTHUR, J. V., D. A. Kovacic, AND M.H. SMITH. 1988. Genetic diversity in natural populations of a soil bacterium across a landscape gradient. Proc. Nat. Acad. Sci. 85:9621-9624.
MEDLIN, L. K., H.J. ELwoop, S. STICKEL, AND M.L. SOGIN. 1988. The charac- terization of enzymatically amplified eukayotic 16S-like rRNA-coding regions. Gene 71:491499.
MICKEVICH, M. F. AND M.S. JOHNSON. 1976. Congruence between morphological and allozyme data in evolutionary inference and character evolution. Syst. Zool. 25:260—270.
MICKEVICH, M. F. AND C. MITTER. 1983. Evolutionary patterns in allozyme data: a systematic approach. Pp. 169-176 in Advances in Cladistics, Vol. 2. Proceedings of the 2nd Meeting of the Willi Hennig Society, N. I. Platnick and V. A. Funk, eds. Columbia U. Press. New York.
OLSEN, J.L. 1990. Nucleic acids in algal systematics. J. Phycol. 26:209-14.
PAMILO, R. AND M. NEI. 1988. Relationships between gene trees and species trees. Molec. Biol. Evol. 5:568-583.
SNEATH, P. H. H. AND R. R. SOKOL. 1973. Numerical Taxonomy. W.H. Freeman, San Francisco.
SANDERSON, M.J. AND M.J. DONOGHUE. 1989. Patterns of variation in levels of homoplasy. Evolution 43:1781-1795.
SOUDEK, D. AND G.G.C. ROBINSON. 1983. Electrophoretic analysis of the species and population structure of the diatom Asterionella formosa. Can. J. Bot. 61: 418-433.
STABILE, J. E., J.C. GALLAGHER, AND E. T. WURTZEL. 1990. Molecular analysis of
intraspecific variation in the marine diatom Skeletonema costatum. Biochem. Syst. Ecol. 18:5-9,
Woop, A.M. 1987. Molecular biology, single cell analysis and quantitative genetics:new evolutionary genetic approaches in phytoplankton ecology. Pp. 41-71 in Immunochemical Approaches to Estuarine, Coastal and Oceanographic Questions. C .M. Yentsch, F. C. Mague, and P. K. Horan, eds. Elsevier. New York
86
11th DIATOM SYMPOSIUM 1990
Mathematical Modeling of Silicoflagellate Skeletal Morphology and Implications Concerning Skeletal Latticeworks
by K. McCartney University of Maine at Presque Isle, Presque Isle, Maine 04769 J. Ernissee
Department of Geology and Geography, Clarion University, Clarion, Pennsylvania 16214
and D. E. Lopeér
Geophysical Fluid Dynamics Institute, Florida State University, Tallahassee, Florida 32306
with 3 figures
Abstract: Silicoflagellates are one of three siliceous microfossil groups that have skeletons made up of a rigid latticework of rods. Mathematical modeling of silicoflagellate skeletal design suggests that an important factor influencing skeletal shape is the minimization of apical surface area, while the reduction of skeletal material appears to have secondary importance. Presumably, minimization of area is also important for the radiolarians and ebridians, which are the other groups having latticework skeletons. The reason why these three groups use such a skeletal design, and why this morphology is not found among non-siliceous groups, remains a mystery.
INTRODUCTION TO LATTICEWORK SKELETONS
A wide variety of different skeletal types are used in the organic world (see Rief and Thomas 1975). Skeletal latticeworks, in which linear or curved rods are interconnected to form a rigid net that encloses most of the organism, are among the less common, being used only by the radiolarians, ebridians, and silicoflagellates. These microorganisms are siliceous and planktic; indeed, the only group of siliceous organisms that do not have
87
McCARTNEY, K.
latticework skeletons are the diatoms. Rigid latticeworks are not found among macroor- ganisms or microorganisms that are benthic or have calcareous skeletons.
The precise function of the latticework skeleton is enigmatic, but the occurrence of lat- ticeworks among three distinctly different organism groups suggests some functional con- vergence. Latticework skeletons have obvious advantages and disadvantages over those skeletons that have continuous or overlapping surfaces that surround most of the cell, such as found in foraminifera, coccoliths, and diatoms. The most obvious advantage of the skeletal latticework is its light weight, an important attribute for organisms that must remain in suspension. Use of geodesic domes in building design suggests that rigid lat- ticeworks offer considerable strength for less weight, although no research has been done on the strength of latticeworks in the context of organisms or on the reasons why such microorganisms might need strong skeletons. The large openings between the skeletal elements suggest that the skeleton would not serve well as a defense against predators.
McCartney and Loper (1989) have proposed three reasons why skeletal latticeworks are found exclusively among siliceous rather than carbonate organisms: (1) the tensile strength of quartz is higher than that of calcite, which may offer advantages for skeletal latticeworks; (2) the greater surface area of the latticework skeleton may not favor calcite, which dis- solves more easily in seawater than does silica; and (3) since silica is undersaturated in modern oceans, siliceous organism prefer latticeworks because they use less skeletal material than skeletons that surround the organism with continuous surfaces. Together these reasons suggest that latticework skeletons offer a lightweight construction for which silica is the best suited material available to the organism.
MATHEMATICAL MODELING OF LATTICEWORK SKELETONS
Mathematical models of silicoflagellate skeletal morphologies have recently been con- structed in order to learn more about skeletal latticeworks and determine factors influencing silicoflagellate skeletal design. Silicoflagellates are ideal for preliminary modeling of this type because their simple geometric designs and lack of internal skeletal elements make the mathematical modeling of the three-dimensional framework much easier than for more complex organisms, such as radiolarians. In addition, presence of a basal ring in silicoflagellates provides an easy basis for orienting the skeleton and provides a flat basal surface to further facilitate construction of mathematical models.
Models of silicoflagellate morphology reviewed here are based on optimization mathe- matics, in which one dependent variable is minimized over all possible values on one or more independent variables subject to certain constraints. Optimization morphologies are found throughout nature and are commonly seen as polygonal patterns such as the colonial structure of corals, the venation of the insect wing, or the cracks produced by the desic- cation of mud. These patterns are the result of close-packing conditions and are efficient because the material needed for the walls between individual ‘‘cells’’ is minimized while still allowing the volume of within each polygon to be relatively large.
The latticework skeletons of siliceous microorganisms have many characteristics similar to the polygonal structures above, including intersections formed by the juncture of three skeletal elements with angles commonly close to 120 degrees. However, the skeletal lat- ticework is not the result of close-packing since the siliceous microorganisms are not colonial and the polygons formed by these elements are not the surface expression of planar structures that run through the organism. Indeed, latticework skeletons found among siliceous microorganisms differ considerably from the structures mentioned previously in that they are forms of intersecting rods rather than planes.
88
11th DIATOM SYMPOSIUM 1990
That latticework skeletons tend to have a spherical shape suggests that they result from area minimization. This has been previously proposed for some radiolarian skeletons (Thompson 1942; Almgren 1982), and considerable optimization study has been done with soap bubbles (see Almgren and Taylor 1976). A surface of minimal area results when some physical effect, such as surface tension, acts to minimize the area of the surface given some constraint, such as constant enclosed volume. The sphere is the min- imum surface enclosing a specified volume, provided that no other constraints apply.
Silicoflagellates, however, differ from radiolarians and ebridians in their relatively large and flat basal area, which gives hemispherical rather than spherical shape to the skeleton. The basal opening is also undivided so that it is much larger than any of the apical openings. The differences between apical and basal surfaces strongly suggest that these surfaces play different roles, that the silicoflagellate skeleton is not the result of simple minimization of total area, and that area minimization applies to the apical surface. Presence of paired silicoflagellates with connected basal rings suggests that the basal surface is important to the reproduction of the organism.
The general morphology of silicoflagellate skeletons suggests that two factors influence skeletal design. The hemispherical shape of the apical surface indicates that reduction of apical surface area is an important factor while the simple skeletal morphologies suggest that reduction of skeletal material is also important. Neither of these, however, are expected to be important to the exclusion of all others; silicoflagellate skeletons are probably a compromise between these and other factors. If the minimization of apical surface area was all-important, then the silicoflagellate would have a more radiolarian appearance, with many skeletal elements and smaller open spaces so that apical structure would closely approximate a spherical surface. If the reduction of skeletal material was the all-important factor then the skeleton would be of extremely simple shapes, if indeed it had a skeleton at all.
MODELING OF SILICOFLAGELLATE MORPHOLOGY
In modeling silicoflagellate morphologies through optimization mathematics we seek to construct formulae that minimize one variable with respect to others and to determine the shape of the resulting skeletal design. Four external variables are of special interest, these being the apical surface area (Aa), the basal surface area (Ab), the volume (V) enclosed between these surfaces, and the total length (S) of the skeletal elements. Because four variables are difficult to present mathematically and graphically, we have reduced these to three variables by dividing the basal area into each of the others to produce the dimen- sionless variables A*, V*, and S* (see McCartney and Loper 1989, for formulae). Dimen- sionless variables have the additional advantage of being unitless.
The apical area is determined by dividing the apical surface into triangular facets (see McCartney 1988a, b) and summing their total area. The volume is measured by computing the area between each triangular facet and the plane of the basal ring, and summing this total. The procedure used for modeling silicoflagellate morphology is to optimize a single dependent variable with respect to V*, within the skeleton. Formulae have been derived (see McCartney and Loper 1989) that minimize A* or S*; these formulae were constructed to include a number on independent, dimensionless, variables from which we could figure the shape of the optimal morphology. Examples of internal variables include length of the apical bridge and shape of the apical and basal rings. These internal variables are allowed to vary to optimal values, for the internal variables can be specified if we wish to study a particular shape, or see how other variables change if one variable is specified.
89
McCARTNEY, K.
A &
CORBISEMA
& 9
FOUR-SIDED DICTYOCHA
e) f) FIGURE 1. Silicoflagellate skeletal morphologies as modeled by op- timization mathematics: a) Cor- bisema with straight basal sides, b) Corbisema with bowed sides, ¢) Dic-
tyocha with bridge parallel to minor
4-SIDED 6-SIDED axis, d) Dictyocha with bridge paral-
lel to major axis, e) four-sided Dis-
DISTEPHANUS DISTEPHANUS tephanus, f) six-sided Distephanus.
Optimization models for four basic silicoflagellate skeletal morphologies have been con- structed. The four studied morphologies (Fig. 1) are the three-sided Corbisema, the four- sided Dictyocha, and the four- and six-sided Distephanus. The internal variables within each model were varied in order to produce a wide variety of forms within each group, such as Dictyocha with apical bridge parallel to the long or short axis (Figs. Ic, d, respec- tively), or Corbisema with bowed sides (Fig. 1b). Thus, different silicoflagellate con- figurations can be modeled and compared to determine their efficiency at reducing A* or S* for given V*.
a)
We have found that for each of the four models, minimization of apical surface area (A*) consistently produce easily recognizable silicoflagellate configurations, while minimization of skeletal material (S*) commonly results in configurations that are rare or not found in silicoflagellates. We (McCartney and Loper 1989) conclude that silicoflagellate skeletons
90
11th DIATOM SYMPOSIUM 1990
FIGURE 2. Plot showing A* values for given V* for four silicoflagellate morphotypes. The six-sided Distephanus morphology has the lowest A* values of those tested.
91
McCARTNEY, K.
tend to reduce the apical surface area to a minimum and that this factor is more important than reduction of skeletal material. However, comparison between models shows that some silicoflagellate designs are more efficient at minimizing A* than others. Mor- phologies with a greater number of basal sides and more complex apical structures min- imize A* better than simpler skeletal designs. For the four silicoflagellate groups modeled thus far, the six-sided Distephanus best minimizes A* while the three-sided Corbisema has the highest A* values (Fig. 2).
The occurrence of relatively simple silicoflagellate skeletal configurations such as Cor- bisema suggest that other factors besides the minimization of A* are also important to the organism. A strong candidate for a second factor, as mentioned previously, is the reduction of skeletal material. Silica is undersaturated in the oceans and may be a limiting nutrient to silicoflagellates (Tappan 1980). Reduction of skeletal material also reduces the weight of the skeleton and, thus, settling velocity. S* values for silicoflagellate configura- tions that minimize A* are shown in Figure 3. The simpler morphologies utilize less skeletal material for all values of V*.
The simple morphology of Corbisema lends itself to a more detailed model (McCartney and Loper, 1992) in which the basal sides can be bowed outwards. Tests of this model show that the most efficient Corbisema morphology for minimizing A* has basal sides that bow outwards so that the ring itself would be round (see Fig. 1b), while the morphology that uses least skeletal material has sides that bow outwards to a lesser degree. The mor- phologies that have less skeletal material are more typical of fossil Corbisema skeletons, suggesting again that silicoflagellate skeletal material is an important factor. These data support the idea that silicoflagellate skeletal morphology is a compromise between reduc- tion of skeletal material and reduction of apical surface area.
Discussion
Perhaps the most important conclusion resulting from the optimization modeling of silicoflagellates is the apparent importance of apical surface area. McCartney and Loper (1989) have presented several hypotheses for why this is important and believe that the most reasonable answer is that silicoflagellates minimize apical surface area in order to minimize surface energy. This conclusion suggests that surface tension plays a very im- portant role for the silicoflagellate organism. Unfortunately, there is very little literature on the relationship between surface energy and organisms, particularly those of very small size, and this hypothesis remains to be tested.
The importance of minimizing apical surface area should also apply to radiolarians and ebridians. The skeletons of these organisms, however, have internal elements and lack the basal ring that gives a hemispherical shape to the skeletons of silicoflagellates. Whether these differences are caused by some functional adaptation or simply reflect differences in growth processes or evolutionary history is one of many questions that remain un- answered. That the most abundant and diverse group of siliceous organisms, the diatoms, use a completely different skeletal design suggests that the reduction of silica usage may be less important or that defense against predators is more important with the diatoms than with the other siliceous groups.
The geographic distribution of silicoflagellate skeletal morphologies is interesting in that, throughout the Cenozoic, more complex morphologies are found at higher latitudes while simpler morphologies are found at lower ones. For example, six-sided Distephanus predominate at high latitudes in modern oceans while four-sided Dictyocha predominate at lower latitudes. In the Paleogene Dictyocha predominated at high latitudes while Cor-
92
11th DIATOM SYMPOSIUM 1990
FIGURE 3. Plot showing V* versus S* for the four silicoflagellate morphotypes. The S* values are obtained for a morphology with minimum A*. For any value of V*, the six-sided Distepahnus uses the most silica while the four-sided Dictyocha uses the least. The numbers and letters shown in the inset represent measured silicoflagellate specimens. A = Dictyocha with bridge parallel to minor axis. F = Dictyocha with bridge parallel to major axis. Z = Dictyocha with diagonal bridge. C = Corbisema. Numbers indicate a Distephanus with that number of basal sides. From McCartney (1988a).
McCARTNEY, K.
bisema were more common at lower ones. This suggests that the minimization of apical surface might be more important in colder waters than near the equator, while the im- portance of reducing skeletal material increases towards the equator. Thus, there might exist environmental reasons that influence each of these factors, and silicoflagellate mor- phology in general.
Acknowledgements
We gratefully acknowledge assistance, constructive comments, and suggestions made by R. D. K. Thomas and financial support from the Division of Math/Science, University of Maine at Presque Isle. This is Contribution No. 343 of the Geophysical Fluid Dynamics Institute, Florida State University, Florida State University, Tallahassee, FL 32306.
Literature Cited
ALMGREN, F. J., JR. 1982. Minimal surface forms. Mathematical Intelligencer 4:164— PZ:
ALMGREN, F. J., JR. AND J. E. TAYLOR. 1976. The geometry of soap films and soap bubbles. Sci. Am. 235:82-93.
McCARTNEY, K. 1988a. Modeling silicoflagellate skeletal morphology. Ph.D. Disser- tation, Florida State University, Tallahassee. 217 pp.
1988b. SILICO: A computer program for the three-dimensional measurement of silicoflagellate skeletons. Computers & Geosciences 14:99-111.
McCaRTNEY, K. AND D.E. LopPER. 1989. Optimization of the silicoflagellate genera Dictyocha and Distephanus. Paleobiology 15:283-298.
1992. Optimization modeling of the silicoflagellate genus Corbisema. Micropaleontology 38:87-93.
REIF, W-E. AND J. A. ROBINSON. 1975. Geometrical relationships and the form-function complex: animal skeletons. Neues Jahrb. Geol. Paleont. Monats. 3:184—-191.
TAPPAN, H. 1980. The Paleobiology of Plant Protists. W. H. Freeman & Co., San Fran- cisco. 1028 pp.
THOMPSON, D. W. 1942. On Growth and Form. Cambridge University Press (2nd ed.), Cambridge, England. 1116 pp.
94
11th DIATOM SYMPOSIUM 1990
Silica Content of Freshwater and Marine Benthic Diatoms
by Daniel J. Conley
Horn Point Environmental Laboratory, Center for Environmental and Estuarine Studies, University of Maryland System, P. O. Box 775,
Cambridge, Maryland 21613
Paul V. Zimba
Department of Fisheries and Aquatic Science, University of Florida Gainesville, Florida 32606
and Edward Theriot
Academy of Natural Sciences of Philadelphia, 19th and The Parkway, Philadelphia, Pennsylvania 19103
with | figure and 2 tables
Abstract: The silica content of benthic diatoms was determined from cultures of 12 freshwater and six marine clones. Marine benthic diatom silica content per unit of biovolume (0.0199 + 0.0301 pmol jim”) was similar on average to that obtained for freshwater benthic diatoms (0.0179 + 0.0256 pmol m™) and marine benthic diatoms contained on average significantly more silica per unit of biovolume than marine planktonic species (0.000502 + 0.000466 pmol um, Conley et al. 1989). Although limited, our data demonstrate that marine benthic diatoms can be as heavily silicified as freshwater benthic and planktonic diatoms. Therefore, differences in silica content per unit of biovolume between marine planktonic diatoms and other diatoms cannot necessarily be solely at- tributed to salinity. Our results coupled with a reported reduction in the silica content of marine diatoms from the geologic record implies that evolutionary changes have occurred in marine planktonic diatom silica content. We conclude that marine planktonic diatoms probably have reduced their silica content downward and have evolved to species that are better adapted to ambient environmental conditions as dissolved silica concentrations have decreased in the world’s oceans.
CONLEY, DANIEL J.
By contrast, benthic diatoms are able to remain heavily silicified without losing a competitive advantage to more lightly silicified diatoms.
Introduction
Diatoms are known to adjust their silica content for a variety of reasons including growth rate, nutrient limitation, salinity, and temperature. A recent analysis that examined the silica content of diatoms as it related to cell size demonstrated that the amount of silica contained in the frustule of a planktonic diatom could be predicted from the biovolume of the diatom (Conley et al. 1989). Furthermore, Conley et al. (1989) have shown that freshwater planktonic diatoms have one order of magnitude more silica contained within the frustule per unit of biovolume than do marine planktonic species. Significant differences were not observed between freshwater pennate and freshwater centric diatom silica content when corrected for biovolume.
Casual observation has suggested that benthic diatoms are more heavily silicified than their planktonic counterparts. In this paper we will compare the silica content of marine and freshwater benthic diatoms to those determined for planktonic diatoms in a previous study (Conley et al. 1989) and address the question of whether the silica content of freshwater benthic diatoms are different from marine benthic diatoms. In addition, we will evaluate observed differences in diatom silica content in the context of the evolution of diatoms.
Materials and Methods
Twelve clones of freshwater benthic diatoms and six clones of marine benthic diatoms were grown in batch culture. Cultures were grown at 20°C and 100 ptEinsteins m2? s! illumination on a 16:8 L/D cycle. Diatoms were harvested during mid-exponential log- phase for determination of silica content, cell biovolume, and abundance. Freshwater diatom cultures were obtained from the Diatom Culture Collection of Loras College’s Department of Biology (Czarnecki 1987), and isolates from Lake Lochlossa and Lake Okeechobee, Florida. Taxa used from the Loras College Diatom Collection included Coc- coneis placentula var. lineata (Ehrenberg) V. H., Cymbella minuta Hilse ex Rabh., Gopho- nema acuminatum v. pusilla Grun., Navicula cryptocephala Kiitz., Nitzschia sigma (Kiitz.) W. Sm., Pinnularia viridis (Nitz.) Ehrenberg, and Surirella ovata Kitz. Taxa isolated from Lake Okeechobee included Achnanthes c.f. minutissima Kitz., Navicula c.f. menis- culus, Nitzschia palea (Kiitz.) W. Sm, and Nitzschia subacicularis (Hustedt) and Achnan- thes c.f. minutissima Kiitz. was isolated from Lake Lochlossa. The freshwater benthic diatom cultures were grown in volcanic ash media (Czarnecki 1987) and Woods Hole MBL media (Stein 1973).
Marine diatom cultures were obtained from the Provasoli-Guillard Center for Culture of Marine Phytoplankton, Bigelow Laboratory for the Ocean Sciences, West Boothbay Har- bor, Maine. Species analyzed included Achnanthes brevipes (Ag.) Cl., Amphiprora paludosa vy. duplex Donkin, Amphora sp., Nitzschia frustulum (Kiitz.) Grun., Stauroneis amphoroides Grun, and an unidentified pennate diatom. The marine benthic diatom cultures were grown in f/2 media (Stein 1973). All clonal designations are reported in Tables 1 and 2.
Diatom silica content was determined by filtering replicate subsamples (30-50 mL) from cultures onto Nuclepore polycarbonate membrane filters (47 mm diameter, 0.4 [tm porosity) at vacuum pressures less than 100 mm Hg. Filters were digested using a wet alkaline extraction (0.1 N NaOH) in a boiling water bath, neutralized, and the extract
96
11th DIATOM SYMPOSIUM 1990
measured for dissolved silica on a Technicon AutoAnalyzer II (Paasche 1980a; Krausse et al. 1983). Dissolved silica also was determined on the filtrates to ensure that cultures were silica sufficient. Silica content values determined in this study, therefore, contain both frustule and intracellular silica. However, the cell wall accounts for the majority of the silica measured from a cell (Werner 1977) with < 10 % of the silica retained on the filter as intracellular silica (Paasche 1980b; Taylor 1985).
Subsamples also were collected for cell density and biovolume measurements. Cell counts were made using the Utermohl sedimentation method as modified by Venrick (1978). Biovolume was determined by measuring 10 cells per culture and relating the taxa to specific geometric shapes. All statistical analysis was carried out using Statistical Analysis System for Personal Computers (SAS-PC).
3
Benthic Diatoms
© - Freshwater mg - Marine
Log [Si Content (pmol Si)]
1 2 3 4 5 6 Log [Biovolume (um )]
FIGURE 1. Relationship of biovolume to silica content of freshwater benthic diatoms (open circles) and marine benthic diatoms (closed squares). The regression lines are from freshwater planktonic diatoms (solid line) and marine planktonic diatoms (dashed line) determined in Conley et al. (1989).
OF
CONLEY, DANIEL J.
TABLE |. Silica content (pmol cell = ) reported with standard deviation, biovolume ( um? ), and silica content per unit of biovolume (pmol yum) of freshwater benthic diatoms.
Species Silica content Biovolume Si/Biovolume
Achnanthes c.f. minutissima OKEE 0.595 + 0.564 (n=2) 112 0.00531 Achnanthes c.f. minutissima LOCH 0.207 + 0.204 (n=2 144 0.00144 Cocconets placentula vy. lineata A-83 233 3400 0.0685 Cymbella minuta 023 31.8 420 0.0757 Gomphonema acuminatum v. pusilla A-67 81.3 + 78.8 (n=2) 6210 0.0131 Navicula cryptocephala 38/03/A 42.8 7220 0.00593 Navicula c.f. menisculus OKEE 1.30 + 0.310 (n=2) 144 0.00903 Nitzschia palea OKEE 1.67 + 1.19 (n=2) 358 0.00467 Nitzschia sigma L-7 66.3 5100 0.0130 Nitzschia subacicularis OKEE 5.58 + 2.29 (n=2) 1490 0.00375 Pinnularia viridis A-53 50.2 + 44.5 (n=2) 5395 0.00930 Surirella ovata A-55 21.2 + 18.8 (n=2) 4000 0.00530 Results
The silica content of both freshwater benthic diatoms (Table 1) and marine benthic diatoms (Table 2) varied by about a factor of two within clones harvested on different dates; silica content has been reported to vary by an order of magnitude within a single clone (Taylor 1985). Reproducibility in benthic diatoms is more problematic than in planktonic diatoms because benthic diatoms often clump together and adhere to the culture vessel wall.
The amount of silica per unit of biovolume in marine benthic diatoms averaged 0.0199 + 0.0301 pmol j1m°3 and is several orders of magnitude greater than determined for mar- ine planktonic species (0.000502 + 0.000466 pmol um”, Conley et al. 1989). The amount of silica per unit of biovolume in marine benthic diatoms also was similar on average to those obtained from freshwater benthic diatom cultures (0.0179 + 0.0256 pmol jum’). For the most part, freshwater benthic diatoms contained a similar amount of silica per unit of biovolume as their planktonic counterparts (Fig. 1), although two (Cocconeis placentula v. lineata and Cymbella minuta) of the 12 diatom species analyzed contained significantly more silica per unit of biovolume than the others (p < 0.0001). A significant log-log linear relationship was obtained between freshwater benthic diatom silica content and biovolume (r2 = 0.76, p < 0.0002, n= 12):
logio [silica content (pmol cell"!)] =
(1.18, S. E. = 0.210)log10 [biovolume (1m3)] - (2.60, S. E.= 0.662). (1)
98
11th DIATOM SYMPOSIUM 1990
TABLE 2. Silica content (pmol cell -!) reported with standard deviation, biovolume ( 1m?), and silica content per unit of biovolume (pmol ji nv) of marine benthic diatoms.
Species Clone Silica Content Biovolume —_Si/Biovolume Achnanthes brevipes WAT7 1.76 1680 0.00105 Amphiprora paludosa v. duplex 73M 90.6 + 116 (n=4) 1152 0.0786 Amphora sp. WTAM 6.16 252 0.0244 Nitzschia frustulum 13M 0.504 + 0.271 (n=3) 108 0.00467 Stauroneis amphoroides 11M 30.0 + 50.3 (n=4) 2956 0.0101 Unidentified pennate WTMB 1.42 2518 0.000654
This relationship was not significantly different from that obtained for freshwater planktonic diatoms by Conley et al. (1989). A significant relationship was not obtained between marine benthic diatom silica content and biovolume (r? = 0.16, p< 0.44, n=6).
Discussion
Our results demonstrate that benthic diatoms generally contain significantly more silica per unit of biovolume than do planktonic diatom species and that marine benthic diatoms can be as heavily silicified as freshwater benthic diatoms. Salinity differences between the two environments cannot account for the variation observed in silicification between marine and freshwater planktonic diatoms (Conley et al. 1989). Previous research on the effects of salinity on diatom silica content are equivocal. Olsen and Paache (1986) found that cells of Thalassiosira pseudonana had a higher silica content at lower salinities. Conversely, McMillan and Johansen (1988) found that valves of 7. decipiens were less heavily silicified at lower salinities.
The leading hypotheses as to why freshwater planktonic diatoms contain relatively more silica than marine planktonic diatoms relate to differences in sinking strategy and dif- ferences in ambient dissolved silica concentrations between freshwater and marine en- vironments, factors that generally are not a concern in the ecology of benthic diatoms. It is interesting to note that the more lightly silicified benthic species often are observed in plankton samples. For example, two of more lightly silicified marine diatoms, Achnanthes brevipes and Nitzschia frustulum, are both species commonly found in the plankton of Chesapeake Bay (Wilderman 1984). In freshwaters, many planktonic diatoms are tychoplanktonic and spend a portion of their life cycle in the benthos, often in resting stages, such that there are very few true freshwater diatoms that spend their entire life history in a planktonic phase. For marine planktonic species it may be an advantage to have a lower silica content, making it less likely to sink out of the photic zone, since, in the oceans, once a diatom is lost from the upper mixed layer reentry into the photic zone is difficult.
Relative dissolved silica availability between marine and freshwaters also might select for more lightly silicified marine planktonic diatoms. In general, there is a connection between silicification and ambient dissolved silica concentrations (Paasche 1980b), which
99
CONLEY, DANIEL J.
might explain the differences in silicification between marine planktonic diatoms and other diatoms. Although freshwaters may have low dissolved silica concentrations (usually during summer), dissolved silica concentrations are often well above limiting values during the seasons of optimal growth. By contrast, dissolved silica concentrations in most regions of the world oceans are extremely low, except in polar waters where the most heavily silicified marine diatoms are found.
Differences in the amount of silica per unit of biovolume in marine planktonic diatoms support the assertion that the silica content of marine planktonic diatoms has been reduced with time. Barron and Baldauf (1989) suggest that Paleogene (65-23 Ma) marine diatoms and radiolarians appear to possess considerably thicker-walled siliceous frustules and tests than their Neogene (23 Ma-present) counterparts. Moore (1969) reported that Quaternary radiolarians have one-fourth less silica than Eocene radiolarians and that radiolarian silica content has linearly decreased downward from 50 Ma to present. Barron and Baldauf (1989) suggest that a substantial evolutionary turnover occurred in marine diatoms and radiolarians during the early Oligocene as relative dissolved silica availability changed in the paleo-oceans. The changes in silica content may reflect an increased competition for dissolved silica during the late Cenozoic leading to more lightly silicified furstules in Neogene diatoms. Prior to the evolution of organisms requiring dissolved silica for growth, concentrations in the Precambrian ocean were much higher and controlled by chemical processes, specifically inorganic reactions such as the sorption of dissolved silica by various clays and precipitation of authigenic mineral phases such as opal-CT (Siever 1992). The newly evolved sponges and radiolarians in the Cambrian modified the Si biogeochemical cycle and began to utilize large quantities of dissolved silica. On the land angiosperms evolved, and with them a variety of plants that deposit biogenic silica, especially the grasses (Kaufman et al. 1983). However, it was the evolution of the diatoms beginning in the late Jurassic that drastically reduced dissolved silica concentrations and depressed concentrations to their present low level (Siever 1991).
Another explanation for the lighter silicification is physiological in that perhaps marine planktonic diatoms have become more proficient at constructing frustules. The formation of a frustule is an energy-requiring process (Werner 1977), and if heavy silicification is not necessary for survival it would seem that selective pressures would allow for diatoms to become more lightly silicified with time. If the oceans reached a new silica equilibria after the origin of radiolarians and diatoms or after a change in ocean circulation, then a stepwise reduction in silica content per cell would be expected, not a monotonic continuous decrease as exhibited by radiolarians (Moore 1969).
In short, we need to know the nature (direction, pattern, and amount) of change in diatom silica content and dissolved silica availability in the paleo-oceans before we can determine the controlling factors of diatom silica content. We conclude that marine planktonic diatoms probably have reduced their silica content downward and have evolved to species that are better adapted to ambient environmental conditions. By contrast, benthic diatoms are able to remain heavily silicified without losing a competitive advantage to more lightly <