AMERICAN FERN es JOURNAL ioe QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Editor Alan R. Smith Department of Botany, University of California, Berkeley, CA 94720 Associate Editors Gerald J. Gastony, Department of Biology, Indiana University, Bloomington, IN 47401 Christopher Haufler, Department of Botany, University of Kansas, Lawrence, KS 66045 David B. Lellinger, U. S. National Herbarium NHB-166, Smithsonian Institution, Washington, D.C. 20560 Terry R. Webster, Biological Sciences Group, University of Connecticut, Storrs, CT 06268 The American Fern Society Council for 1985 TERRY R. WEBSTER, Biological Sciences Group, University of Connecticut, Storrs, CT 06268. President FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. Vice-President W. CARL TAYLOR, Milwaukee Public Museum, Milwaukee, WI 53233. Secretary JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, Knoxville, TN 37916. ‘tans DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Beiiacis. VT 0540 Pe ‘Teakives ALAN R. SMITH, Dept. of Botany, University of California, — a 94720. Journal Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, D Memoir Editor DENNIS Wm. STEVENSON, Dept. of peng Science, ley see Columbia University, New York, NY 100 Fiddlehead Forum Editor The “American Fern Journal” (ISSN 0002-8444) is an illustrated quarterly devoted to the general study of ferns and fern allies. It is owned by the American Fern Society, published at the Pringle Herbarium, University of Vermont, Burlington, VT 05405, and printed by Allen Press, Inc., 1041 New Hampshire St., Lawrence, KS 66044. Second-class postage paid at Burlington, VT, and addi- tional entry point. Subscriptions $12.00 gross, $11.50 net if paid through an agency (agency fee $0.50); sent free to members of the American Fern Society (annual dues $10.00 + $4.00 mailing surcharge beyond U.S.A., Canada, and Mexico; life membership $200.00 Orders for back issues should be addressed to Dr. James D. Montgomery, Ecology III, R.D. 1, Berwick, PA 18603. Back volumes 1910-1978 $5.00 to $6.25 each; single back numbers of 64 pages or less, $1.25; 65-80 pages, $2.00 each; over 80 pages, $2.50 each, plus shipping. Back volumes 1979 et seq. $8.00 each; single back numbers $2.00 each, plus shipping. Ten percent discount on orders of six volumes or more. Fiddiehead Forum The editor welcomes contributions from members and non-members, including miscellaneous notes, offers to exchange or purchase materials, personalia, horticultural notes, and reviews of non- technical books on ferns. Spore Excha Mr. Neill D. Hall, 1230 Northeast 88th Street, Seattle, WA 98115, is Director. Spores exchanged and lists of available spores sent on request. Gifts and Bequests Gifts and bequests to the Society enable it to expand its services to members and to others inter- ested in ferns. Botanical books, back issues of the Journal, and cash or other gifts are always wel- comed, and are tax-deductible. Inquiries should be addressed to the Secretary. Table of Contents (A list of articles arranged alphabetically by author) BARRINGTON, Davin S., The present evolutionary and taxonomic status of the fern genus Polyst Geet ee ee Boom, BRIAN M., Ethnopteridology of the Chacobo Indians in Amazonian Bolivia ......... Barrron, D. Mi (ome: Moth 5 oe ss rng es BURKHALTER, JAMES R., A new station for Dicranopteris flexuosa in Bay County, Florida .... Prrsancstand, BOEYIE 1 tame UE EL 5 or i es ee ee Cox, PAUL ALAN AND P. B. ToMLINSON, Relationships between ecological — and branch- in the tree fern Lophosoria quadripinnata in Veracruz, Wasiod) 32 os DUCKETT, JEFFREY G., Wild gametophytes of Equisetum sylvatioum. ......-.---5+-+6+2++*-> Esteves, LucIANO M., Git M. FELIPPE, AND THEREZINHA S. MELHEM, Germination and mor- phology of spores of Trichipteris corcovadensis ......--..---++---+srrssre tert Pesreeu, Gu, Mi ieee ve EOL) i ee sicnuns, Finney Gen Golcer Ol) 26 re ee GRANT, PAUL (see een eee ee Hasxmm. Mutant L. (see Werth 00h) 2 en Hauke, RICHARD L., Ontogeny of the commissure ee . Gametophytes of Equisetum giganteum .......------------rs Henan, Pevee BE ee WN eae LDER, SAMUEL W. (see ee ee oe Huipurt, AKKE (see Weeth et G1) <0. oo ee Iwareuns, Kumuc (eed Rath rr rr . New records for longevity of Marsilea sporocarps .....-.------+---s:s2rttt KaTo, MASAHIRO, an and Kunio IwaTsuKI, An unusual submerged aquatic ecotype of Asplenium Kort, L. S., and D. M. BRITTON, ies of morphological characteristics of leaves and the spo- ial region in the taxono y of Isoetes in northeastern Ni North America ........- LANDRY, GARRIE P., and SAMUEL W. a Marsilea macropoda new to Louisiana....... LELLINGER, DAVID B., Nomenclatural notes on some ferns of Costa Rica, Panama, and Colom- FF Fy ——_—,, review of Arkansas ferns ce ee ree eee eee review of Ferns of Jamaica, a guide to ra pteridophytes .........-------+++++7- LUCANSKY, TERRY W., Anatomical studies of Sphaeropteris and Cnemidaria (Cyatheaceae) RE McDona _p, J. W. i. a MELHEM, THEREZINHA S. . (see Esteves et a Ea ee MickEL, JOHN T., Three new anemias from northern South America .......-.--------+--°> {in scien ee ee Prck, Carot J. (see Peck, James H., et Oi ee ee PECK, — H., STEVE L. ORZELL, ERIC SUNDELL, and Carot J. Peck, Dryopteris ludoviciana and D. D. x australis new Me ee Rury, PHILuip M., New locations for Isoétes tegetiformans in Georgia .....------ += SmitH, ALAN R., review of Ferns to ker and GFW (in ———., review of Index filicum. Supplementum quintum: pro annis 1961-1975 ......-..--- , review of Guide de fougéres et plantes alliées .................................. SMITH, Davip K., Polystichum aleuticum from Adak Island, Alaska, a second locality for the speci SPICER, Peaieaay A., RoByN J. BURNHAM, PAUL GRANT, and Harry GLICKEN, Pityrogramma calomelanos, the primary, post-eruption colonizer of Volcan Chichonal, Chiapas, CO ee PS ee ee eee ee” SUnNoeLL Hesc lose Pack, alee OT eal) 6 er a aa THomas, R. DALE, A newly proudeuret habitat for Isoétes melanopoda in Louisiana ........ Tora, R. E., JR., B. H. Mars, S. K. PERKINS, J. W. MCDONALD, and G. A. PETERS, en cine Wacner, Davin H., JR. i vin 98 Se ee ee spores WERTH, CHARLES R., MELANIE L. snes wee AKKE HULBURT, Osmunda cinnamomea forma epeennr at Mountain Liebe, Virgie Soa es cc i a ee Volume 75, Number 1, pages 1-32, issued 2 August 1985 Volume 75, Number 2, pages 33-72, issued 5 August 1985 Volume 75, Number 3, pages 73-104, issued 6 December 1985 Volume 75, Number 4, pages 105-136, issued 5 March 1986 AMERICAN FERN =e JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Editor Alan R. Smith Department of Botany, University of California, Berkeley, CA 94720 Associate Editors Gerald J. Gastony, Department of Biology, Indiana University, Bloomington, IN 47401 Christopher Haufler, Department of Botany, University of Kansas, Lawrence, KS 66045 David B. Lellinger, U. S. National Herbarium NHB-166, Smithsonian Institution, Washington, D.C. 20560 Terry R. Webster, oe Sciences Group, University of Connecticut, Storrs, CT 06268 The American Fern Society Council for 1986 FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. President JUDITH E. SKOG, Biology Dept., George Mason University, Fairfax, VA 22030. Vice-President W. CARL TAYLOR, Milwaukee Public Museum, Milwaukee, WI 53233. Secretary ES D. CAPONETTI, Dept. of Botany, University of Tennessee, Knoxville, TN 37916. Treasurer DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Burlington, VT 05405. Records Treasurer JAMES D. MONTGOMERY, Ecology III, R.D. 1, Berwick, PA 18603. Back Issues Curator ALAN R. SMITH, Dept. of Botany, University of California, Berkeley, CA 94720. _Journal Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, DC 20560. Memoir Editor DENNIS Wm. STEVENSON, Dept. of Biological Sciences, Barnard College, Columbia University, New York, NY 10027. Fiddlehead Forum Editor The “American Fern Journal” (ISSN 0002-8444) is an illustrated quarterly devoted to the general study of ferns. It is owned by the American Fern Society, and published at the Pringle Herbarium, University of Vermont, Burlington, VT 05405. Second-class postage paid at Burlington, VT, and additional entry point. Claims for missing issues, made 6 months (domestic) to 12 months (foreign) after the date of issue, and orders for back issues should be addressed to Dr. James D. Montgomery, Ecology III, R.D. 1, Berwick, PA 18603. Changes of address, dues, and applications for membership should be sent to the Records Trea- rer er. General inquiries concerning ferns should be addressed to the Secretary. Subscriptions $12.00 gross, $11.50 net if paid through an agency (agency fee $0.50); sent free to members of the American Fern Society (annual dues, $10.00 + $4.00 mailing surcharge beyond U.S.A., Canada, and Mexico; life membership, $200.00). Back volumes 1910-1978 $5.00 to $6.25 each; single back numbers of 64 pages or less, $1.25; 65-80 pages, $2.00 each; over 80 pages, $2.50 each, plus shipping. Back volumes 1979 et seq. $8.00 each; single back numbers $2.00 each, plus shipping. Ten percent discount on orders of six volumes or more. Fiddiehead Forum The editors (Dennis Wm. and Jan Wassmer Stevenson) welcome contributions f; b non-members, including miscellaneous notes, offers to exchange or purchase materials, personalia, horticultural notes, and reviews of non-technical books on ferns. a Mr. Neill D. Hall, 1230 Northeast 88th Street, Seattle, WA 98115, is Director. Spores exchanged and lists of available spores sent on request. Gifts and bequests to the Society enable it to expand its services to members and to others inter- ested in ferns. Botanical books, back issues of the Journal, and cash or other gifts are always wel- comed, and are tax-deductible, Inquiries should be addressed to the Secretary. Table of Contents (A list of articles arranged alphabetically by author) ALVERSON, Epwarp R. and JosEPH ARNETT, Pellaea brachyptera new to Washington Anstey, omprys (ea Avera ee ees BARRINGTON, Davin S., CaTHy A. Paris, and THoMas A. RANKER, Systematic inferences from re and stomate size in the ferns 6 a Bus, ELEANOR M., A second eastern North American occurrence for forked spleenwort, plenium brane: po ee Se CARTER, ears AYNE R. FaIRCLOTH, Osmunda cinnamomea forma frondosa in the astal Saal an of Georgia and Florida CUSICK, rage W., Cystopteris tennesseensis in West Virginia ........................- FAIRCLOTH, WAYNE R. (see Gorter] 0 ee ee AUFLER, CHRISTOPHER Hi toee Windies) ee ee HIcKEy, Bap soétes megaspore surface morphology: Nomenclature, variation, and sys- ti HICKOK, baat mC. (eee Ware) Liovn, Rowent Mises Skog)... a LUEBKE, NEIL T. Lge EN a i ee ee Moran, Rossin C., The neotropical fern genus Olfersia............. 0.0... e eevee eee Moyroup, RICHARD (see Nauman) ........ 6.6 eer tee ttt eens NAUMAN. CLIFTON E.. Prichomanes i Ploride ©. 2.6 es ——— and RICHARD Moyroup, A new substrate for Ophioglossum palmatum in Florida ... Paris, CATHY A. (see Barrington et al.) ee PECK, JAMES H, Second locality for Dryopteris sopra in Arkansas 2.3 ICE, M. G., review of . Porth Ol TO i wee neste ne Race Tuomas A. and CHARLES R. WERTH, an enzymes from herbarium specimens: Electrophoresis as an afterthought .............---. +... 2+ 1s erect es as Hi WI ne ee eee ee re ee ne ee Skoc, JupITH E. and Rosert M. Luoyp, review of Biology of pteridophytes ............... SmrrH, ALAN R., Revision of the neotropical fern genus Cyclodium....................... So.tis, Douc.as E. and Pameta 8. Soutis, Active enzymes from megaspores of Marsilea and hee ee ee eee Soave. Pasian G. tape Sole, Of os es re ne er ee eee TAYLOR W. Carl, review of A pie manual of the ferns & fern-allies of the United States and Canada a —, (sce Wagner, W. H., et al)... .- seer eee ee re tenes es Tayor, W. Cart and Net T. Lueske, Germinating spores and growing sporelings of aquatic Vonetos 0 2h a ee Trron, ROLLA, Some new names and combinations in Plermdacene 2. a Wactien, F. 8. (see Wagner, W. H., et al) ........- 55 e rere eer rer tener ess PU Woda WH)... ee Wacner, W. H., Jr., F. S. WAGNER, and W. Cari Taytor, Detecting abortive sp in herbarium specimens o! of sterile hybrids eee ar ee a en ice and , Three new spec f ts (Botrychi bg. Botrychium) endemic in wpoatorn North America... 2. so ee ee eee WOM eT GA Ware cs ce rhe ee eee ee WARNE, Tuomas R., GARY L. WALKER, and Lestie G. HICKOK, A novel method for surface- sterilizing and sowing fern spores ...........----------ssssseeereser tt Me a Whittier) eee eee WertH, CHares R. ay WuiTTIER, DEAN P. and Terry R. WEBSTER, Gametophytes of Lycopodium lucidulum from counties ee eo en ae we WINDHAM, MIcHAEL D., The use of fern herbarium specimens in biosystematic research: In- Ge i A a a ——— and CuristopHer H. HAuFLer, Biosystematic uses of fern gametophytes derived from herbarium specimens Volume 76, Number 1, pages 1-32, issued 28 May 1986 Volume 76, Number 2, pages 33-100, issued 21 August 1986 Volume 76, Number 3, pages 101-160, issued 23 October 1986 Volume 76, Number 4, pages 161-192, issued 14 January 1987 SSE ce Oe Spee ome ee ane a ar aR ne ey % iii 2 orca er ean cutee bh Te rg hegre an Saal pete Sa Passa cent ippele er 9 ees | Eel Medea th PONE Pane” Ex ag ORI wre epaNe Erte MRI ee Neg orcas 8 AMERICAN a FERN meow JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Pityrogrammea calomelanos, the Primary, Post-eruption Colonizer of of Volcan Chicho Mexico Robert A. Spicer, Robyn J]. Burnham, Paul Grant, and Harry Glicken i Bilateral Spores in New World Grammitid Ferns Florence S. Wagner 6 The Atomic aa iti nf Cr. J eheie. oc Randy Wayne and Peter K. Hepler 12 Ethnopteridology of the Chacobo Indians in Amazonian Bolivia Brian M. Boom — Special Report The Present Evolutionary and Taxonomic Status of the Fern Genus Polystichum David S. Barrington 22 ee ee as te rors 8 ie Fores of Kansas and Missouri : David M. Johnson New Records for Longevity of Marsilea Sporocarps : —— No! ideal Nee Fe of Costa Rica, On oe ee Panama, and Colombia. — David B. Lellinger ae The American Fern Society Council 985 TERRY R. WEBSTER, Biological Sciences ioe aoe pole of Connecticut, Storrs, CT 06268 Preeidest FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109 Vico-Presiden: MICHAEL I. COUSENS, Faculty of Biology, University of West Florida, Pensacola, FL 32504. Secretary JAMES D. CAPONETTI, a. of Botany, University of Tennessee, Knoxville, TN 37916. bean: DAVID S. BARRINGTON, Department of Botany, University of Vermont, ee ete VT 0 ALAN R. SMITH, Dept. of Botany, University of California, Berkeley, CA 94720. Journal Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, DC 20560. gpa Editor DENNIS Wm. STEVENSON, Barnard College, Columbia University, eget ——- NY lehead sel Editor American Fern Journal RUA ee ee Dept. of Botany, University of California, Berkeley, CA 94720 ASSOCIATE EDITORS GERALD J. GASTONY ............. Dept. of Biology, ae a, Bloomington, IN 47401 CHRISTOPHER MAI. 8S ea of Botany, University of Kansas, ee. wrence, KS 66045 DAVID & LELLINGER ............... U.S. Nat'l Herbarium NHB-166, Smithsonian Institution, ashington, DC 20560 TERRY R. WEBSTER ____ Biological Sciences —-— University of Connecticut, Storrs, CT 06268 fe The S ccrmoecage Fern Jounal” (ISSN 0 } is an n iustrated quarterly devoted to the general study of s barium, . ‘Unive of Vermont podcast sige aencmmbuale y, and pomge paid at arama gan oe Claims fo issing issu ade 6 ths (d ic) 12 inobithe (focetea) offer the date al tous and orders for back iss hould be addressed to Dr. James D. Montgomery, Ichthyological Asso- — ciates, R.D. 1, Berwick, PA 18603 address, d fi bership should be sent to Dr. David S. Bar- 05401. : American Fern Journal 75(1):1-5 (1985) Pityrogramma calomelanos, the Primary, Post-eruption Colonizer of Volcan Chichonal, Chiapas, Mexico ROBERT A. SPICER Life Sciences Department, Goldsmiths’ College, Creek Road, London SE8 3BU, England ROBYN J. BURNHAM Department of Botany KB-15, University of Washington, Seattle, WA 98195 PAUL GRANT Geology Department, Imperial College, London SW7 2BP, England Harry GLICKEN U.S. Geological Survey, Cascades Volcano Observatory, 5400 MacArthur Blvd., Vancouver, WA 98661 The southern Mexican Volcan Chichonal, otherwise known as “El Chichén,” (17°20'N, 93°12'W) erupted violently on 28 March 1982 at 2330 (local time), send- ing a column of ash and sulfur aerosol 16.8 km into the atmosphere. Further intermittent activity culminated in two major explosive eruptions on 3 April (1930 local time) and 4 April 1982 (0530 local time). During the final eruptions, column collapse gave rise to three major pyroclastic surges and flows that devastated 154 km? of paratropical rain forest (sensu Wolfe, 1979) and agricultural land (Fig. 1). More complete descriptions of the eruptions are to be found in Varekamp et al. (1982), SEAN Bulletin (1982), Duffield et al. (1984) and Sigurdsson et al. (1985). The pre-eruption vegetation of the area consisted of a patchwork of second- growth rain forest and local crops, which included cocoa, bananas, coffee, and maize. Numerous epiphytic pteridophytes were to be found in the area, with Selaginella aff. schizobasis Bak. (Burnham 001, BM, MEXU), Blechnum occi- dentale L. (Burnham 09, BM, MEXU), and Pteris grandifolia L. (Burnham 065, BM, MEXU) as frequent components of the ground cover. The eruptions produced no lava flows, but trachyandesitic air-fall ash and pyroclastic surges and flows inundated a roughly circular area within a radius of approximately 6 km from the crater. The nature of the volcaniclastic deposits is described in detail in Sigurdsson et al. (1985). A particular feature of the eruptions was the large amount of sulfur ejected into the atmosphere (Thomas, 1982; Mroz & Sedlacek, 1982; Rampino & Self, 1984) and in the tephra (Varekamp & Luhr, 1982; Luhr et al., 1982). Chemical analyses of the tephra are given in Hoffer et al. (1982) and Cochemé and Demant (1983). High local annual precipitation (>200 cm) (Rzedowski, 1983) subsequently produced numerous erosional gulleys that have cut down, in some cases through several meters of volcanic debris, to the 1982 pre-eruption soil horizon. In February 1984 we spent four weeks in the Chichonal area. The purpose of our visit was to study the formation of potential plant fossil deposits associated 2 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) eerces Mexico Peer Volcan hichonalg =~ 5: Fic. 1. Map of the Volcan Chichonal area. The dotted line indicates the approximate extent of the — devastated area. Sites of Figs. 2-4 are indicated. with the eruptions. This work involved studying plant and volcanic sediment . distribution in the area, and it was immediately apparent that Pityrogramma 7 1 1 (T ink ‘D. Ll 055, 058, BM, MEXU) was the primary colonizer~ 7 wherever devastation of the pre-eruption vegetation had been intense. By Feb- ruary 1984, fertile fronds of P. calomelanos, up to 1 m long, were observed to 2 have regenerated from rhizomes in the 1982 pre-eruption soil horizon. Further- more, this regrowth had apparently stabilized the soil against further erosion : (Fig. 2). Regrowth from buried rhizomes accounted for the largest plants of Pit- _yrogramma that were observed, but these only represented a small proportion — of the overall population in the devastated area. By far the greatest number ag individuals occurred as young sporophytes, typically 5-7 cm high, at densities of - approximately 40 per m? on stable, pumice-strewn slopes to within 1-2 km the crater (Fig. 3). Young sporophytes were usually rooted in tephra the size of R. A. SPICER ET AL.: PITYROGRAMMA CALOMELANOS 2 4 7 / “ lanos. 2. Plants regenerating from buried rhizomes in erosional I gullies 5.3 km from crater. 3. Young sporophytes on pumice-strewn slopes 2 km from crater. 4. Plants Fics. 2-4. Pityrogramma calome Mm crevasses 200 m from crater rim. 4 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) fine to medium sand between pumice blocks averaging 5-10 cm in maximum diameter. With decreasing distance from the crater, the sporophyte density de- creased and became more patchy. However, sporophytes with fronds of 15 cm or so long were observed in small rock crevasses approximately 200 m east of the new crater rim (Fig. 4). Other notable occurrences were at the edge of eroded sulfur crusts surrounding non-active fumarolic sites on the distal lobes of large pyroclastic flows and on post-eruption flood deposits in the Rio Magdelena river valley downstream from Ostuacan. Pityrogramma is a genus native to tropical America. It is common in open habitats and is a colonizer of disturbed ground. Consequently it frequently oc- curs as a weed in palm and banana plantations (Tryon & Tryon, 1982). Although some authors consider the genus to be native to Africa and Madagascar (Tryon & Tryon, 1982), others consider it as introduced in the Old World, from where it has spread rapidly throughout the wet tropics (Schelpe, 1975). Its natural hab- © itats include gravel bars in streams and rivers, rocky cliffs, and landslide areas, and it was noted as an early colonizer of laval flows on Mount Lamington, Papua (Schelpe, 1975). Schelpe (1975) stated that Pityrogramma was recorded on Krak- — atoa in 1906; he suggested that its arrival there was due to long-distance dispersal of spores by wind. Because of the high sulfur content of the plume from Volcan Chichonal it is — evident that P. calomelanos is tolerant of high sulfur concentrations in the sub- — strate. Buried rhizomes were undoubtedly exposed to leached sulfurous com- — pounds from the overlying volcaniclastics, and both prothalli and young sporo- : phytes have been able to survive high local concentrations of sulfur in close — proximity to fumaroles. In February 1984, sulfurous gases were still seeping from _ some pyroclastic flows, and large quantities of hydrogen sulfide were being given _ off from the crater. It would seem, therefore, that Pityrogramma is also tolerant — of high concentrations of atmospheric sulfur and acidic rain. Seed plants are recolonizing the devastated area, but none so prolifically as — Pityrogramma. It is easy to see why this is so. The rhizomes are capable of | surviving violent burial and some degree of thermal shock, and are able to produce fertile fronds relatively quickly. As with most homosporous pterido- — phytes, vast numbers of wind-dispersed spores are produced, any one of which © has the capacity to produce a prothallus. Even if prothalli are spatially separated — so as to prevent outcrossing, self-fertilization can give rise to a new sporophyte — in a relatively short time. On the other hand, seed plants in the same situation — may take longer to mature, flower, and set seed. Seeds are not produced in such — large numbers as spores and are usually not dispersed so far. : High relative concentrations of monolete fern spores have recently been re- ported from coals immediately overlying an iridium-rich clay at the Cretaceous/ Tertiary boundary in New Mexico and Colorado (Pillmore et al., 1984). This fern — spike” is interpreted as representing the initial floral recovery phase following a possible terminal Cretaceous bolide impact as suggested by Alvarez et al. (1980). _ fe present environment surrounding El Chichon is in some ways analogous [0 ~ that hypothesized for the immediate post-impact terrestrial Tertiary ecosystem. _ We thank the following for their kind assistance: Servando de la Cruz, UNAM; ~ R. A. SPICER ET AL.: PITYROGRAMMA CALOMELANOS 5 Mario Sousa and the staff of the Herbario Nacional de México; Sr. Isidro Cha- vez; the people of Ostuacan; and The Royal Society. LITERATURE CITED ALVAREZ, L. W., W. ALVAREZ, F. ASARO, and H. V. MICHEL. 1980. Extraterrestrial cause for the CocHEME, J.J. and A. DEMANT. 1983. Naturaleza y composicién del material emitido por el Volcan Chichonal, Chiapas (marzo- i 1982). In El Volcan Chichonal. Mexico, D.F.: Universidad Nacional Autonoma de Méx DUFFIELD, W. A., R. I. TILLING, and R Cun 1984. Geology of El Chichén volcano, Chiapas, Mexico. J. Volc. Geotherm. Res. 20:117-132 HOFFER, . G. P. FILIBERTO, and P. MUELA. 1982. Erupiion of El Chichén volcano, Chiapas, Mexico, March to 7 April 1982. Science 218: ae Lunr, J., . 4 E. CARMICHAEL, and J. C. VAREKAMP. Minerology and petrology of ejecta from - March-April saan, eruptions of El USE ag Chiapas, Mexico. Eos. Trans. Amer. eophys. Union 63:1126-1127. Mroz, as J. and W. A. SEDLACEK. 1982. Stratospheric aerosols from El Chichén. Eos. Trans. Amer. e PILLMoreE, C. L., R. H. TicHaby, C. J. Ortu, J. S. Gitmore, and J. D. KNniGHT. 1984. Geologic fram hones of nonmarine Ceatecebun-Tertiaey boundary sites, Raton Basin, New Mexico and Colorado. wigager = 1180-1183. RampPIno, M. R. and S. SELF . The atmospheric effects of El ppt Sci. Amer. 250:34—43. RZEDOWSKI, J. 1983. ae i: México. Mexico, D.F.: Editorial Lim SCHELPE, E.A.C.L.E. 1975. Obse eae on the spread of the American pee Pityrogramma calo- melanos. Fern Gaz. 1:101- SEAN BULLETIN, Smithsonian pies 1982. Volcanic events: El] Chichon. 7:2 oe H., S. N. Carey, and J. M. EspInDoLa. 1985. The 1982 eruptions of El oan volcano, xico: Stratigraphy of pyroclastic deposits. J. Volc. Geotherm. Res. {in press). THoMas, P E. 1982. Satellite observations of the El] Chich6én volcanic aerosol in the stratosphere. os. Trans. Amer. Geophys. Union 63:8: TRYON, R. pe and A. F. Tryon. 1982. Ferns se allied plants with special reference to tropical America. New York: Seirinnne Vers oo J.C. and J. Lunr. 1982. The 1982 eruption of El eo “eeu Nature of the eruption, and volatiles. Eos. Trans. Amer. Geophys. Union 63:1 ————,, K. PRESTEGAARD, and J. CANEPA. 1982. The se ais of El Chichén ves Chisues. Mexico. Part 1. Stratigraphy, volume is agen element characteristics o ash fall deposits. Geol. Soc. Amer. Abstr. Prog. 1 Wo re, J. A. 1979. Temperature parameters of humid to a forests of eastern Asia and relation o forests of other regions of the northern hemisphere and Australasia. U.S. Geol. Surv. Prof. Paper 1106:1-37. American Fern Journal 75(1):6-11 (1985) Bilateral Spores in New World Grammitid Ferns FLORENCE S. WAGNER Department of Botany, University of Michigan, Ann Arbor, MI 48109 Most of the grammitid ferns, now placed by many authors in a separate family, Grammitidaceae, or a subfamily, Grammitidoideae, of the Polypodiaceae, were at one time classified as members of the genus Polypodium. In 1967 Morton wrote, “Evidence ... has been accumulating in recent years that there are two major groups of species—those centering around the true Polypodium, as typi- fied by P. vulgare L., and those belonging to Grammitis Swartz and some closely | related small genera.’’ Ching (1940) created a family, Grammitaceae, to accom- modate these dwarf polypodies. In 1947 Holttum recognized the family, which — he revised to Grammitidaceae, and later wrote (1955), “There are no interme- | diates between them and any true Polypodiaceae, and a close relationship seems very doubtful.” After treating them, in 1947 and previously, as members of the Polypodiaceae, Copeland came to believe the group should have family status : (1951). In her revision of Grammitis in New Guinea, Parris (1983) placed the genus in Grammitidaceae and gave a history of the family classification. Most authors have accorded the grammitid ferns family status, but W. Wagner (1973) - retained them in a subfamily, the Grammitidoideae of the Polypodiaceae, and : Tryon and Tryon (1982) placed them in a tribe, the Grammitideae of the Poly- 1 podiaceae. : The degree of relationship between the grammitid and the polypodioid ferns still remains, however, a matter of argument. Several presumably clear-cut char- — acters separate the two groups. These have been discussed by de la Sota (1960) ; and Morton (1967). Evans (1969) gave a comparison of the Grammitidaceae with : the Polypodium pectinatum-plumula complex, and Price (1983) contrasted Poly- podium, Pecluma, and Ctenopteris, the first two in the Polypodiaceae s.s., the | last in the Grammitidaceae. | Briefly the distinctions between the polypodioid and the grammitid ferns are ~ these: The Polypodiaceae s.s. have in general creeping rhizomes with scattered © fronds, the grammitids have in general upright rhizomes and tufted fronds. Fronds © of polypodies are often articulate, of grammitids, non-articulate. Scales of the — former tend to be peltate and commonly borne on the leaf blades; those of the . latter are basally attached and found only on the rhizome. The polypodioids — bear scales and various types of hairs while the grammitid indument is typically — composed of bristle-like setae often borne in pairs or multiples. Venation in the x a & B et = Fics. 1-6. 1. Sporangium and spores of Ctenopteris longa (Buchtien 5252, MICH) x 80. 2. Sporan- i agner 78528, MICH) x 80. 3-4. Scanning © electron micrographs of spores of G. lanigera var. lanigera (Wagner 77062A, MICH) x 2500. Arrow — n micrograph of G. lanigera var. stella (Little — 9356, US) x 2500. 6. G. lanigera var. lanigera (Wagner 77512, MICH) x . . Mie te aE DAE ee tas Ue RP ania eens ae re ce F. S. WAGNER: GRAMMITIS SPORES 8 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) polypodioids is characteristically reticulate with free included veinlets; in gram- mitids the veins are usually free, or if reticulate, without included veinlets. The sporangial stalk in the polypodioids is composed of two or three rows of cells, in the grammitids only one. Spores of polypodioids are bilateral, monolete, and without chlorophyll; spores of grammitids are tetrahedral to globose, trilete, and have chlorophyll. Although exceptions may be cited to many of the above characters, spores have traditionally provided a dependable difference: ‘‘Possi- bly the most uniform and easily observable difference is in the spores, those of Polypodium being monolete and those of Grammitis trilete . . .” (Morton, 1967). Exceptions to this rule are the subjects of this paper. Between 1977 and 1983 collections made in Costa Rica in connection with a cytological survey of the ferns of that country, included several specimens of Grammitis lanigera (Desv.) Morton var. lanigera [Ctenopteris lanigera (Desv.) Copel.]. In all characters this fern appears to be an incontrovertible member of the Grammitidoideae. However the correlation of characters is not perfect. The exception is the spores. These are conspicuously bilateral (Figs. 2-4). They are green with tuberculate surface and a single long laesura, which commonly has a characteristic bend or slight angle near the center (Fig. 3). In some spores the laesura forms an abbreviated fork (Fig. 4, arrow). The sporangium bears setae and has a stalk composed of a single row of cells for most of its length (Fig. 2). The plants are tufted epiphytes with a short rhizome (Fig. 6). Spreading, stiff, one-celled hairs are borne on the stipe and blade, and the rhizome scales are basally attached and setose. Grammitis lanigera var. lanigera has been recorded from Costa Rica to Bolivia and has also been found in Martinique and Hispaniola (Morton, 1967). Our collections are from the fol- lowing localities in Costa Rica (all are W. Wagner numbers deposited in MICH): Alajuela, Road N of San Ramon, 77062A; Cartago, Estrella, 79036; Tapanti, 83012; Heredia, Cerro Vueltas, 77008, 77013; Slopes of Volcan Barba above Porrosati, 77512; Zurqui, 83019; Puntarenas, Monteverde, 83004B; San José, Cascajal, 78524, 78528. Other grammitid ferns suggesting relationship to Grammitis lanigera were subsequently examined to determine their spore shape. One of these is Ctenop- teris stella Copel., which Morton (1967) considered to be a variety of G. lanigera. A specimen from Colombia (Huila, 15 km SE of Garzon, Little 9356, US) and one from Bolivia (Cochabamba, King and Bishop 9677, MO) were found to have bilateral green spores, but with a different laesural pattern. In these the majority have a modified trilete scar (Fig. 5). Furthermore, the sporangial stalk differs from that of the type variety, and from nearly all grammitid ferns, in having two rows of cells for most of its length. Thus, Morton has combined as varieties in one species, two ferns with very different sporangial stalks. Similar confusion has been found in plants named Grammitis cultrata (Willd.) Proctor from Guatemala (Quezaltenango, Aguas Jorgines, Skutch 904, MICH) and Mexico (Chiapas, La Independencia, Breedlove 33641, MICH: Volcan de Tacana, Matuda 2912, MICH) all of which have a single-rowed sporangial stalk, while other specimens from the last locality (Matuda S-225 and S-229, MICH) have two-rowed sporangial stalks. These differ also in the shape of the pinnae F. S. WAGNER: GRAMMITIS SPORES 9 which are shorter and more rounded in the last two collections. All of them have green bilateral spores and setose sporangia. The same combination of characters, bilateral green spores and a two-rowed stalk together with a setose sporangium is found in two other grammitids, Ctenopteris sericeolanata (Hook.) Copel., which Morton thought to be synony- mous with Grammitis lanigera var. lanigera, and G. senilis (Fée) Morton [Cten- opteris senilis (Fée) Copel.] from Central America, northern South America, and the Antilles, which Stolze (1981) placed in synonymy with G. cultrata. Whether or not a single cell is to be found at the base of the two-rowed sporangial stalk in these ferns has not been determined. In Prosaptia contigua (Forster) Pres] (Grammitidaceae) and in Dictymia J. Smith (Polypodiaceae) the sporangial stalk of two rows of cells is supported by a single cell. These are anomalies of uncertain significance in both families (Wilson, 1959). As a result of these observations, a broader consideration of polypodioid characters in gram- mitid ferns will be treated in a future paper. Morton listed Polypodium alternifolium Hook. and P. longum C. Chr. [both are Ctenopteris longa (C. Chr.) Copel. according to Copeland, 1956] as two other species synonymous with Grammitis lanigera var. lanigera. A specimen of P. alternifolium (Ecuador, Sodiro, Aug. 1875, MICH), however was found to have tetrahedral spores. Ctenopteris longa (Bolivia, hacienda Simaco, camino a Tip- uani, Buchtien 5252, MICH), on the other hand, has bilateral green spores like those found in our Costa Rican specimens (Fig. 1). Copeland (1956) in placing P. longum under Ctenopteris wrote: “Spores of South American specimens oblong; of Costa Rican, probably tetrahedral, but very few seen.” He also described, as a New species, Ctenopteris fabaespora from Panama, basing the epithet on its bilateral spores. I have not seen a specimen, but according to David Lellinger (pers. comm.) Morton considered this fern to be the same as Grammitis lanigera var. lanigera. References, such as Copeland's, to bilateral spores in grammitid ems are rare and have usually referred to the genus Loxogramme. Bilateral spores are known in Loxogramme in about half of the species, and although the genus has been thought by many to be a grammitid fern, evidence now seems to indicate that its relationships lie more closely with the polypodioid ferns (M. Price, pers. comm.). Another suggestion that bilateral spores have been found in grammitids is that of Nayar and Devi (1965) who described the spores of Ctenopteris brevivenosa (v.A.v.R.) Holtt. as trilete-tetrahedral but added that many are monolete-bilat- eral. They also found Prosaptia contigua to have monolete-bilateral spores. Specimens of both these species (Singapore, Holttum 23319, MICH, and Su- matra, H. H. Bartlett 7963, MICH, respectively) were examined by me and neither was found to have bilateral spores. Thus there may be different spore types within the same species, or, perhaps, confusion about spore shape in cer- tain Grammitidoideae may arise because of the widespread phenomenon of intrasporangial germination that occurs in this group of ferns (Stokey and Atkin- son, 1958). At least three of the spores visible in my Figure 1 have divided. porangia are seen frequently with the entire contents at various stages of spore germination. Minute two-celled gametophytes, often with the spore wall still 10 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) TABLE 1. Selected Grammitid (G) and Polypodioid (P) Characters in Certain Grammitid Ferns. Spore Sporangium Color Shape Laesura Capsule Stalk Grammitis lanigera var. green® bilateral? monolete’ with _setose® 1 row® of lanigera a bend cells Grammitis lanigera var. green® bilateral” modified’ setose® 2 rows? of stella trilete cells Ctenopteris longa green® bilateral? monolete’ with —setose® 1 row® of cells Ctenopteris sericeolanata _— green® bilateral” monolete?” setoseS 2 rows? of Grammitis cultrata green® bilateral” monolete? setose® 2 rows? of cel Grammitis “cultrata”’ green® bilateral” monolete’ with —setose® 1 row® of cel Grammitis senilis green® bilateral? monolete? setoseS 2 rows? of cel Polypodium alternifolium _green® tetrahedral® trileteS setose® 1 row® of globose cells remaining, may appear oblong, when, in fact, they may have been formed from tetrahedral spores, and also swelling prior to division may make bilateral spores appear to be globose. Variability in spore type and the occurrence of the two types of spores have been discussed by W. Wagner (1974). Within the same genus or very closely related genera, both bilateral and tetrahedral spores are known, for example, in Gleichenia, Antrophyum, Vittaria, and Lindsaea, as well as in Loxogramme. e situation in the grammitid ferns reported here is of special interest with regard to the relationships of the Grammitidaceae with the Polypodiaceae. All of the taxa discussed above have typical grammitid characters with important exceptions (Table 1). The occurrence of polypodioid characters suggests the pos- sibility that the two groups of ferns may be more closely related than many authors have surmised. This study suggests, furthermore, that a re-evaluation of the taxonomy of this group of grammitid ferns is needed. Taxa that have been treated as synonymous or as varieties differ in characters that have usually been considered fundamental and conservative in the general taxonomy of ferns. I am grateful to the following people for help in many ways: David Bay, Bruce Donahoe, Luis Diego Gomez, David Lellinger, and Warren H. Wagner. Material used in this study was made available through NSF (INT 7819909), the Omens zation for Tropical Studies, and three herbaria, MICH, MO, and US. LITERATURE CITED CHING, R. C. 1940. On natural classification of the family “Polypodiaceae.” Sunyatsenia 5: 201-288 COPELAND, E. B. 1947. Genera filicum. Waltham, Mass.: Chronica Botanica. F. S. WAGNER: GRAMMITIS SPORES 11 : pi saa Philipp. J. Sci. 80:93-271. —————. 1956. Ctenopteris in America. Philipp. J. Sci. 84:381-475. Evans, A. M 1969. In eta Ps relationships in the Polypodium pectinatum-plumula complex. n. Missouri Bot. Gard. 55:193-293 Wines “i E. 1947. A revised classification of leptosporangiate ferns, J. Linn. Soc., Bot. 53:123- ee 1955. A revised flora of Malaya. Vol. I. Ferns of Malaya. Singapore: Government Office. Morton, C. V. 1967. The genus Grammitis in Ecuador. Contr. U.S. Natl. Herb. 38:85-123 Nayak, B. K. and S. Biches 1965. Spore Lorighenatepe ot Indian ferns. Grana Palynol. 6:121- 127. PARRIS, B. ve 1983. Filicales) ew pi een Blumea 29:13-222. PRICE, ei as 1983. Pe —- a new tropical American fern genus. Amer. Fern J. 73:109-116. Sota, E. R. DELA. 1960. ehietinores y¢ Gramenitidacene Argentines: Opers Lillesne 5:5- 228. STOKEY, e me and L. R. peeping 958. 1-403 STOLZE, R 6. 1881 Ferns and fern allies of Guatemala. Part II. Polypodiaceae. Fieldiana, Bot. n.s. TRYON, : ce pp A. F. TRYON. 1982. Ferns and allied plants. New York: Springer-Verlag. WacNER, W. H. 1973. Some future challenges of fern systematics and phylogeny. Bot. J. Linn. Soc. ss ap 1):245-256. 1974. Structure of spores in relation to fern phylogeny. Ann. Missouri Bot. Gard. 61:332- Witson, K. A. 1959. Sporangia of the fern genera allied with Polypodium and Vittaria. Contr. Gray Herb. 185:97-127 REVIEW “Ferns to know and grow,” by F. Gordon Foster. xiv + 228 pp. ISBN 0-917304- 98-5. Portland, Oregon: Timber Press. 1984. $29.95 (hardbound). ird revised and somewhat enlarged edition of The gardener’s fern book, reviewed in Amer. Fern J. 56:81, 1966. This edition is in a larger page format and additional species are treated. Unfortunately, the quality of reproduction of the photographs has suffered. It may be ordered from Timber Press, P.O. Box 1631, Beaverton, OR 97075.—A.R.S. American Fern Journal 75(1):12-18 (1985) The Atomic Composition of Onoclea sensibilis Spores RANDY WAYNE and PETER K. HEPLER Department of Botany, University of Massachusetts, Amherst, MA 01003 It is generally accepted that fern spores contain all the elements that are needed for germination because spores germinate on distilled water in the ab- sence of added nutrients (Raghavan, 1980). Here we present the first quantitative data on the atomic composition of Onoclea spores that permit one to know which elements are present in the spore. The presence of an element does not imply its essentiality as a nutrient; rather, it allows for the possibility that this element could be involved with development. One element of particular interest to us is calcium, which is essential for spore germination (Wayne & Hepler, 1984a). Ninety percent of the calcium present in Onoclea spores resides in their walls, whereas 10% is intracellular (Wayne & Hepler, 1984b). By selectively removing calcium from the spores, we show that the wall-associated calcium is required for Onoclea germination in calcium-free media. MATERIALS AND METHODS Mature sporophylls of Onoclea sensibilis L. were collected in Amherst and Pelham, Massachusetts in January, 1981 and February, 1982 and stored in plastic bags in the freezer at —15°C. Prior to an experiment, sporangia were wetted with a 0.1% solution of Aerosol O. T. (Fisher Scientific Co., Pittsburgh, PA), sterilized for 1 min with 1 1 of a 20% (v/v) solution of commercial bleach (5.25% NaOCl), and rinsed with 500 ml of sterile water in a manner modified from Stockwell and Miller (1974). For germination assays, approximately 2 mg of ster- ilized spores were sown on either 10 ml of the standard medium containing 1 _ mM Ca(NO,),, 0.81 mM MgSO,, and 3.45 mM KNO, (pH 5.2), or on 10 ml of 2.5 mM EGTA (ethylene glycol-bis(@-amino-ethy] ether)-N,N,N’,N’-tetraacetic acid, © PH 6.8) for 24 h then transferred to 10 ml of the standard medium plus or minus — Ca({NO,), one hour prior to irradiation. Spores were irradiated to induce germi- nation for 5 minutes with broad band red light (energy fluence rate = 2.4 J-m™" s~’). All manipulations were carried out in sterile plasticware in the dark or under a dim green safelight (Wayne & Hepler, 1984a). Percent germination was determined 48 h after irradiation by the acetocar-_ mine-chloral hydrate method of Edwards and Miller (1972). Data are expressed — as the mean + two standard errors of the mean. 4 In order to analyze the elements contained in the spores, sterilized spores - were dried at 105°C for 2 h and weighed immediately. For the determination 0! calcium, magnesium, potassium, cobalt, nickel, iron, copper, manganese, sodium, — zinc, mercury, cadmium, chromium, silver, molybdenum, vanadium, and lead, — WAYNE & HEPLER: ONOCLEA SPORES 13 spore samples were dissolved in hot concentrated sulfuric acid (1 ml) plus 10 drops of concentrated nitric acid, diluted with water to a 25-50 ml volume de- pending on weight (usually 10 mg), and then analyzed for metals using a Perkin- Elmer 403 Atomic Absorption Spectrophotometer following the Perkin-Elmer standard methods guide. Carbon, hydrogen, and nitrogen analyses were per- formed on a Perkin-Elmer 240 Elemental Analyser according to the modified Pregl-Dumas technique. Sulfur analysis was done by classical BaSO, titration using Thorin indicator after Schéniger oxygen flask combustion. Phosphorus was analyzed by forming phosphomolybdic acid from orthophosphate after conver- sion by hot H,SO, and HNO, digestion. Phosphomolybdic acid was then mea- sured spectrophotometrically at a wavelength of 882 nm. Oxygen analysis was done by a modified Unterzaucher procedure. Chlorine was analyzed as Cl- after a Schéniger oxygen flask combustion followed by a coulometric titration with silver. Culcita coniifolia, Cycas taiwaniana, Nothoscordum bivalve, Psilotum nudum, Selaginella kraussiana, and Tradescantia virginiana were grown in the green- house of the University of Massachusetts, Amherst. Voucher specimens are on deposit in the University of Massachusetts Herbarium, Amherst. Funaria hygro- metica was grown in Laetsch’s medium under continuous white light (Laetsch, 1967). RESULTS AND DISCUSSION The atomic composition of spores of Onoclea sensibilis is shown in Table 1. The spores contain all the macro- and micronutrients needed for higher plant growth (Clarkson & Hanson, 1980). The relative atomic composition of Onoclea spores is similar to that reported for leaves of a variety of ferns (Héhne & Richter, 1981) and corn plants (Epstein, 1972). The high levels of minerals found in On- oclea spores contrast with the very low levels found in corn microspores (Pfahler & Linskins, 1973). The difference in the mineral content of corn microspores and Onoclea spores may reflect the functional differentiation of fern spores as prop- agules. Onoclea spores may carry sufficient nutrients to ensure the proper de- velopment of the gametophyte, whereas the style provides the nutrition for an- giosperm microgametophytes. Consistent with their role as propagules, the spores can withstand considerable desiccation (cf. Barker & White, 1964) as evidenced by their low water content. In Onoclea, the dry weight accounts for 95.43 + 1.02% of the fresh weight of the spore, whereas the dry weight of a plant cell may be only 7% of the fresh weight (Burling & Jackson, 1965; Ray, 1962). Carbon, hydrogen, and oxygen are the most common elements in the spores. The carbon/hydrogen/oxygen molar ratio of 1/1.6/0.3 is a result of the abun- dance of storage lipids and proteins in the spores (DeMaggio & Stetler, 1980; Towill & Ikuma, 1975). This ratio remains nearly constant for 24 h following hydration. The carbon/hydrogen/oxygen ratio changes to 1/1.3/0.9 in week-old gametophytes indicating that in this later developmental stage carbon is mainly Present as carbohydrate (Table 2). tassium, magnesium, sodium, and calcium are the major cations in the spores. 14 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) TABLE 1. The Atomic Composition of Onoclea sensibilis Spores. Number of atoms Element Percent dry weight nmol/mg dry weight __ relative to calcium Cc 58.59 + 2.00 48,784.35 977.64 Oo 21.25 + 1.90 13,281.75 H 7.76 + 0.55 76,942.16 1541.93 N 4,59 + 0.28 3277.00 65.67 P 0.82 + 0.18 254.74 3 K 0.70 + 0.00 179.04 3.59 Ss 0.53 + 0.07 163.76 3.28 Mg 0.34 + 0.07 139.89 2.80 Na 0.23 + 0.07 100.04 2.00 Ca 0.20 2.0.07 49.90 1.00 Cl 0.110 + 0.020 31.027 0.622 Co 0.040 + 0.037 6.787 0.136 Fe 0.024 +0. 4.297 0.086 Ni 0.011 + 0.020 1.874 0.038 Pb 0.010 + 0.018 0.483 0.010 Mn 0.008 + 0.002 1.456 0.029 Zn 0.006 + 0.003 0.918 0.018 Cu 0.005 + 0.003 0.787 0.016 Cr 0.002 + 0.000 0.385 0.008 Ag 0.002 + 0.001 0.185 0.004 Cd 0.0004 + 0.0006 0.035 0.0007 Hg 0.00001 + 0.00001 0.0005 0.00001 Vv <0.001 <0.196 <0.004 Mo <0.0007 <0.073 <0.001 Unknown 4.8 Total 100.0 Calcium has been shown to be essential for the normal spore development. It is _ required for germination (Wayne & Hepler, 1984a), rhizoid growth (Miller et al., — 1983}, and normal protonema development (Cooke & Racusen, 1982). Calcium is — also required for archegonium, antheridium, and sporophyte formation in Wood- — sia obtusa (Bryan & O’Kelley, 1967), and spermatogenesis in Marsilea (Wick, | 1979). In Onoclea, the amount of calcium in the spores is similar to the levels of — calcium found in the fiddlehead, expanded leaf, rachis, and empty sporangia; — however, there is considerably more calcium found in gametophytes growing in — a calcium-sufficient medium (Table 3). The average calcium content in the var- ious tissues of Onoclea is 0.24% which is similar to that found in a variety of greenhouse-grown plants (Table 3}. Calci : , primarily localized in spore walls, i ssary for germination (Wayne & Hepler, 1984b). Onoclea spores, for example, germinate readily when placed : | in deionized water plus MgSO, and KNO,, 4 EGTA, a chelating agent that removes the wall bound calcium. The ability to — germinate can be regained by adding exogenous calcium (Table 4), indicating e importance of external calcium. The potential of unwashed spores to ger- minate in calcium-free media depends on the spore density. We have found that ~ but not if they are prewashed with WAYNE & HEPLER: ONOCLEA SPORES 15 TABLE 2. The Carbon, Hydrogen, and Oxygen Ratios of Several Developmental Stages of Onoclea. Atomic content Ratio with re- Developmental stage Element (umol/mg dry wt) spect to carbon Dry spores Cc 48.78 + 1.66 1 H 76.94 + 5.46 1.6 Oo 13.28 + 1.19 0.3 Hydrated spores (dark) Cc 48.40 + 1.77 1 H 81.56 + 3.37 a7 Oo 15.00 + 0.13 0.3 Hydrated spores (5 min red) Gc 49.07 + 1.03 1 H 81.61 + 1.49 pO Oo 13.91 + 1.44 0.3 Gametophytes (1 week old) Cc 34.03 + 3.36 1 H 44.55 + 2.56 1.3 Oo 29.75 + 9.75 0.9 germination is promoted by high spore density, probably as a result of pooling sufficient calcium from many cell walls (Wayne & Hepler, 1984a). Other studies also show that plants are capable of adding nutrients to “nutrient free,” media thereby masking the requirement for a given nutrient (Miller et al., 1983: Pfahler & Linskins, 1973; Saunders & Hepler, 1983; Wayne & Hepler, 1984a). The concentrations of the transition metals cobalt, nickel, lead, chromium, silver, and copper are relatively high (cf. Gauch, 1972). These trace elements may accumulate in the spores in order to provide a reservoir for the growing TABLE 3. The Calcium Content of Plant Tissues. Calcium content Species Tissue or organ (% dry weight) Onoclea sensibilis gametophyte 0.46 empty sporangia 0.26 rachis 0.22 0.20* expanded leaf 0.17 fiddlehead 0.11 Onoclea sensibilis average of all parts 0.24 ilotum nudum frond 0.06 Culcita coniifolia frond 0.12 Selaginella kraussiana shoot 0.42 Cycas taiwaniana leaf 0.47 Nothoscordum bivalve leaf 0.52 Funaria hygrometica protonema 0.64 Tradescantia virginiana leaf 1.10° * The range is 0.1-0.4% dry weight. *The high calcium content of T. virginiana is a consequence of the presence of calcium oxalate crystals, 16 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) TABLE 4. The Effect of Washing with EGTA on Germination of Onoclea. Treatment Percent germination Control 89.00 + 1.00 EGTA-washed 0.75 + 1.50 EGTA-washed + 1 mM Ca(NO,), 87.25 + 4.50 gametophyte and young sporophyte before a sufficiently large root system can be produced. Alternatively, the accumulation of these heavy metals may be fortuitous but not essential (especially in the case of cobalt), indicating that On- oclea may be a bio-accumulator of heavy metals from the environment and thus may serve as a bio-indicator. Other ferns have been shown to accumulate heavy metals from the environment (see Puckett & Burton, 1981, for a review; Hunter, 1953). This would provide yet a new way in which a fern can be utilized as an environmental monitor (Klekowski & Poppel, 1976; Petersen et al., 1980; Petersen & Francis, 1980). Interestingly, many of these bivalent metals can act as calcium antagonists (Hagiwara, 1973) and thus as inhibitors of germination (Petersen et al., 1980; Petersen & Francis, 1980; Wayne & Hepler, 1984a). ; The concentration of chloride is too low for it to serve as the only counter ion for potassium. The positive charges are probably neutralized by organic acids — such as malic acid, which would account for the low intracellular pH we have observed in Onoclea. : The high level of phosphorus is probably due to the presence of phytic acid — (inositol hexaphosphate) in the spores (DeMaggio et al., 1983). Inositol polyphos- 4 phate is usually considered to be a storage form of phosphate, although recently — it has been shown to interact with calcium ions in triggering cell responses — (Rasmussen, 1981). : Silicon is present in high concentrations in the spores of Selaginella (Tryon & Lugardon, 1978) and the leaves of many true ferns (Héhne & Richter, 1981). It Pia also be part of the spores of Onoclea. We, however, were unable to assay or it. : Here we have presented the natural abundance of elements in the spores of — Onoclea. We have shown how the C/ H/O ratio in particular reveals the devel- _ opmental state of the cells and that there is sufficient calcium associated with — the cells to facilitate germination. By removing the wall-bound calcium, we have _ ment. | Elemental analyses were performed by Dr. G. Dabkowski, Director of the a Microanalysis Laboratory, University of Massachusetts, Amherst. Support for this research was provided in part by NSF grant PCM 840 2414 to P.K.H. anda Sigma Xi grant in aid of research to R.W. WAYNE & HEPLER: ONOCLEA SPORES 17 LITERATURE CITED BARKER, W. G. and R. G. WHITE. 1964. puieaige: of Ma posmag in lyophilized spores of the fiddlehead rm, Matteuccia pensylvanica. Amer. Fern J. 54:87-89. BASSEL, \ og C. C, KUEHNERT, and J. H. Mili. 1981. gears migration and asymmetric cell division in Onoclea sensibilis spores: an ultrastructural and cytochemical study. Amer. J. Bot. 68:350-360 BurLING, E. and W. T. JACKSON. 1965. Cha cough in bistacans eae in cell walls during elongation of oat coleoptile cells. Pl. Physiol. 40:138-1 CLARKSON, D. T. and J. B. HANSON. 1980. The wee nutrition of higher plants. Annual Rev. PI. 98. Cooke, T. au and R. H. RACUSEN. 1982. Cell rages in the filamentous gametophyte of the fern Onoclea sensibilis L. Planta 155:449-4 DeEMaccio, A. E. and D. STETLER. 1980. SoD products in spores of Onoclea sensibilis L. Amer. J. Bot. 67:452-455 , T. TEMPLEMAN, L. Sotpatis, and D. A. STETLER. 1983. Ultrastructure and biochemistry of jek spore germination: protein and phytic acid. (Abstr.} Pl. Physiol. 72(Suppl.)}:99. Epwarps, M. F. and J. H. MILLER. i Growth regulation by reads in fern gametophytes. III. Inhibition of spore germination. Amer. J. Bot. 59:458-46 EpsTEIN, E. 1972. Mineral nutrition a plants: principles and RE New York: John Wiley & Sons. Gaucu, H. G. 1972, Inorganic plant nutrition. eis a Pa.: Dowden, Hutchinson, & Ross. GIWARA, S. 1973. Ca ie a aeane es 1-102 HOuneg, H. and B, RICHTE Unt aii ber den Mineralstoff- und Stickstoffgehalt von Farnkrautern. ae ne ac 10. Huckasy, C. S. and J. H. MILLER. 1984. Spore germination and rhizoid differentiation in Onoclea sensibilis. Pl. Physiol. 74:656-662. Hunter, J. G. 1953. The boy tras of bracken: some major- and trace-element constituents. L Sci. Food ss le -20. KLEKOwskI, E. J., JR. an ns as PopPEL. 1976. Ferns: potential in situ bioassay system for aquatic borne se agersiorane Amer. Fern J. 66:75-79. LAETSCH, i M. 1967. Ferns. In Methods in developmental biology, eds. F. H. Wilt and N. K. essells. New York: Thomas Y. Crowell. SRO; poole of the sensitive fern, Onoclea sensibilis. Bull. Environm. Contam. Tox- icol. 24:489-495. —— ae. F Ay: FRANCIS. 1980. Diftorential eons of fern and moss spores in response to ercuric chloride. Amer. Fern J. 66 PFAHLER, P. L. and H. F. LINSKINS. 1973. rate percentage and mineral content of maize (Zea mays L.) pollen and ns Theor. Appl. Genet. 45:32-3 Puckett, K. J. and M. A. S. BURTON. 1981. The effect i ace elements on lower plants. Pp. 213- 238 in Effect dé. heavy metal pollution on plants, vol. 2, ed. N. W. Lepp. London: Applied Science Publishers. RaGHavaN, V. 1980. Cytology, physiology and biochemistry of germination of fern spores. Int. Rev. hee 62:69-118. SSEN, H. 1981. Calcium and cAMP as synarchic messengers. New York: Wiley. Ray, P.M. 1962. Cell wall synthesis and cell elongation in oat coleoptile tissue. Amer. ]. Bot. 49: SAUNDERS, M. L 2 K. HEPLER. 1983. Calcium antagonists and calmodulin inhibitors block cytokinin-induced bud formation in Funaria. Developm. Biol. 99:41-49, Stockwett, C. R. and J. H. Miter. 1974. Regions of cell wall expansion in the protonema of a fern. Amer. J. Bot. 61:375-378. 18 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) TowiL, L. R. and H. Ikuma. 1975. Photocontrol of the germination of Onoclea spores. IV. Metabolic changes during germination. Pl. Physiol. 56:468-473. Tryon, A. F. and B. LuGARDON. 1978. Wall structure and mineral content in Selaginella spores Pollen & Spores 20:315-340. VOGELMANN, T. C., A. R. BAsseL, and J. H. MILLER. 1981. Effects of microtubule-inhibit ] migration and rhizoid differentiation in germinating fern spores (Onoclea sensibilis). Pro- toplasma 109:295-316. . ——— and J. H. Miter. 1980. Nuclear migration in germinating spores of Onoclea sensibilis: the th and kinetics of movement. Amer. J. Bot. 67:648-652. Wayne, R. and P. K. HEPLER. 1984a. The role of calcium ions in phytochrome-mediated germination of spores of Onoclea sensibilis L. Planta 160:12-20, ——— and P. K. HEPLer. 1984b. Red light stimulates an increase in intracellular calcium in the spores of Onoclea sensibilis. Pl. Physiol. 77:8-11. Wick, S. M. 1979. The localization and possible role of calcium ions during mitosis. Dissertation, Stanford University. REVIEW “Gametophytes of Ophioglossaceae,” by D. D. Pant, D. D. Nautiyal, and D. R. Misra. Phyta Monograph 1:1-111, 1984. i This is a report on the gametophytes and young embryos of the Ophioglossa- ceae in India. Six species of Ophioglossum, 4 species of Botrychium, and Hel minthostachys were studied from extensive collections of gametophytes. A help- ful introduction to the literature initiates this conventional morphological study. Observations on the structure and development of these gametophytes and em- bryos expand our knowledge about the gametophytes from India. New and in- teresting information presented includes: meristems of the gametophyte lobes of Helminthostachys that suggest that the lobes are short lateral branches; slight lignification of the central strand of elongated parenchyma cells in the cylindri- cal gametophytes of Ophioglossum and H elminthostachys; absence of a ventral. canal cell in Ophioglossum archegonia; and endoscopic embryo development in O. nudicaule. The concluding section of the monograph summarizes the new observations and incorporates them into a useful review of the literature. Per tinent information from the literature is also tabulated in a 12 page table at the end of the text. There is an abundance of illustrations (over half of the mono graph). The line drawings are excellent but the halftones are in need of improve ment. The quality of the halftones is not a serious detraction from the pape! because much of the illustrated material is presented as both line drawings and : American Fern Journal 75(1):19-21 (1985) Ethnopteridology of the Chacobo Indians in Amazonian Bolivia BRIAN M. Boom New York Botanical Garden, Bronx, NY 10458 The Chacobo are a small tribe of some 400 Indians belonging to the Panoan language family who live in an isolated rainforest region in northern Bolivia. During a six-month ethnobotanical study of the Chacobo living in the vicinity of the village Alto Ivon, a special effort was made to collect all the pteridophytes and to obtain indigenous names and information on use. The village of Alto Ivon (11°45’S, 66°02’W) is located some 60 air miles south of Riberalta in the Bolivian department of Beni. The forest is Tropical Moist with a distinct dry season, generally from about June to October. The elevation is about 200 m and the terrain is level. The forest canopy is at about 25 m, and the species composition appears to be typical of southwestern Amazonia. The five ecologically most important tree families, as revealed by a 1 ha. inventory, are Moraceae, Myristicaceae, Palmae, Leguminosae sensu lato, and Vochysi- aceae. This tree flora is not especially species-rich; only 94 species with a DBH greater than 10 cm were encountered in the hectare. As with the trees, the pteridophyte flora is not particularly rich. Between Nov. 1983 and Apr. 1984, 13 genera and 25 species of pteridophytes were collected. If the species diversity was not impressive, what was quite interesting was the high degree of fern utilization. The Chacobo have uses for 16 species, or 65 percent of the total encountered in their territory. The uses are all medicinal in nature and generally involve the making of a decoction of fronds or rhizomes. No instances of ferns used as food were encountered. In Table 1 an accounting is given of the ferns collected, the indigenous names, and the manner in which they are employed. Numbers listed are in the author’s series, and vouchers are deposited at NY. Specimens were collected in the forest surrounding Alto Ivon, brought back to the village, and presented to Chacobo informants to elicit names and uses. Ten principal informants were used: Pae Chavez, Caco Soria, Yari Vargas, Tani Chavez, Cana Antelo, Pae Ortiz, Pae Davalos, Maria Soria, Caco Ortiz, and Rabi Ortiz. The last individual was the tribal chief at the time of the study and was responsible for transcribing my tape recorded interviews. When more than one name is given for a species, this represents the name elicited from a second informant and not an alternative name provided by the same informant. A guide to the pronunciation of names is given in Appendix 1. I could not detect any general term comparable to “fern” in English. Chacobo fern names often encompass several botanical species (Table 2). Most of the names appear to be complex terms made up of common terms having other applications in the Chacobo language. Thus, cashimétsisi can be broken down into cashi (= bat, the mammal) and métsisi (= claws); perhaps this alludes to the 20 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) TABLE 1. Ferns of the Chacobo Indians. Adiantopsis radiata (L.) Fée: mitaisa (5048), not used. Adiantum glaucescens KI.: mitaisi (4813), ee decoction used to bathe children who have high ever; xé€qui nihijahéhua (4997), n Adiantum lucidum (Cav.) Sw.: mitaisi nd. tod decoction used to bathe children who have high fever; mitahitsa (5010), not used; xéquitaxo (4090), frond decoction drunk to cure rheuma- tism. Adiantum obliquum Willd.: xéquihuitaxo (4076), pce decoction drunk to cure diarrhea; xéquitaxo 85), frond decoction drun tism. Adiantum petiolatum Desv.: xéquihuitaxo ud smi decoction drunk to cure diarrhea. Adiantum tomentosum K1.: mitaisa (5061), no Asplenium pearcei Baker: mitaisa jahéhua veins sii used; jonotasi (4127), frond decoction drunk to e stomachache. Ctenitis protensa (Afz.) Ching: oe (4812), rhizome shavings boiled and cooled decoction k to cure appendici Ctenitis submarginalis (Langsd. & ene C. Chr.: toriahuitaxo (4998), rhizome shavings placed di- on skin over appendix to relieve pain of appendicitis; jinaristi (4124), frond decoction fae to othe headache. Cyathea sp.: éjiquéréxe (4088), frond decoction drunk to seep pain of appendicitis. Hecistopteris ce (Spreng.) J. Smith: no Chacobo name (4403), not used. Lindsaea divaricata K1.: toriahuitaxo (4730), rhizome shavings beled — cooled decoction drunk to dicitis. Lindsaea lancea (L.) Brade var. lancea: xéqui jahuéhua (4911), frond decoction used to bathe hyper- active children Lindsaea portoricensis Desv.: mitaisa (5049), n Lomagramma guianensis (Aubl.) Ching: no est name (4743, 4799}, not used. Lomariopsis japurensis (Mart.) J. Smith: toriahuitaxo (5020), not used; cashimétsisi (4089) two drops of stipe exudate mixed with water and this solution lea to alleviate stomachache. Metaxya rostrata (Willd.) Presl: xéqui jahué (5047), not use Polypodium bombycinum Maxon: jihui ratsamica nishi (5036), rhizome scrapings cere picide in cold water and resulting paste is put directly on skin over appendix to cure appendici Polypodium si Satta aff. brevifolium Link: jonotasi (4128), frond decoction ie to cure stomachache. Polypodium (Phlebodium) decumanum Willd.: jonotasi (4889), no’ Polypodium (Microgramma) corse drema Desv.: ciaague (4019), frond decoction drunk to alleviate stomachach Polypodium Sp: mitaisa (5050, xéqui jahéhua (4380), not used. ) Proctor, vel aff.: xéqui jahuéhua (5014), not used. Trichomanes Sonal ede. jorojina (4640), not used; mitahuisma (4130), frond decoction drunk cure diarrhea. Trichomanes vittaria Poir.: jorojina (4023), reportedly used as a medicinal, but informant could not ore specific. small curved scales on the rhizome. Other names, like mitaisi, appear to be simple, proper nouns applying only to those particular species. Of course, it is possible that mitaisi has a non-fern meaning unknown to me or to the linguists of the Summer Institute of Linguistics who have studied the language and have prepared a dictionary (unpublished). A translation of name components is given in Table 2. Appreciation is extended to Drs. John T. Mickel and Alan R. Smith for iden- tifying the specimens, to Dr. Brent Berlin for reviewing the manuscript, and B. M. BOOM: ETHNOPTERIDOLOGY OF CHACOBO INDIANS 21 TABLE 2. Chacobo Fern Names. capétéjiquéréxé [capéte (cayman); ji (line, tail); quéréxé (arrow head with several barbs)]—Cy- athea sp. cashimétsisi te (bat, the ne ee meétsisi (claws, fingernails)|—Lomariopsis japurensis; Polypo- dium rogramma) m ophyllum jihui ratsamica pee [jihui a peas (to climb, grasping tightly); nishi (liana)]|—Polypodium ycinum nese va a tail)]—Ctenitis submarginalis tasi— Asplenium serie ei; Polypodium (Campyloneurum) aff. brevifolium; Polypodium (Phlebo- ee het di cuma jorojina [jina (line, tail] —Trichomanes pinnatum; Trichomanes vittaria mitaisa—Adiantopsis radiata; Adiantum tomentosum; cpa portoricensis,; Polypodium sp. mitaisa jahéhua [jahéhua (mother)|—Asplenium pearc mitaisi—Adiantum glaucescens; Adiantum lucid ee itis protensa; Ctenitis submarginalis; Lindsaea divaricata; Lomariopsis japurensis equi huitaxo [xéqui (maize)|—Adiantum obliquum; Adiantum petiolatum xéqui jahué [xéqui (maize); jahue (thing, something)]— Metaxya rostrata that xéqui eae [xéqui (maize); jahuéhua (that which is larger than . . .]]|—Lindsaea lancea var. lancea; Polypodium sp.; Tolan Arhanes tome eras xéqui nihi jahéhua [xéqui (maize); nihi (forest); j Adiantum glaucescens xéqui taxo [xéqui (maize)|—Adiantum hicdehoee Adiantum pie especially to the Edward John Noble Foundation for a generous grant that made fieldwork amongst the Chacobo possible. APPENDIX 1. Pronunciation of Chacobo names. After studying the Chacobo language, the Summer Institute of Linguistics de- vised a written form employing symbols that were as similar as possible to Span- ish. Under their system there are 21 symbols in the Chacobo alphabet. These are listed below with my interpretation of their Spanish or English equivalent sound. 4 = Spanish a. b = Spanish b. c = Spanish c. ch = Spanish ch. é = no exact equivalent; intermediate in sound to the Spanish u and e; unrounded midcentral vowel as in the English word hurt. h = no exact equivalent; similar in sound to a “hard” English h, executed with a glottal stop. hu = no exact equivalent; similar to the English sound wh. i = Spanish i. j = Spanish j. m = Spanish m. n= Spanish n. 0 = Spanish o. p = Spanish p. qu = no exact equivalent; similar in sound to the English kw. r= Spanish single-trill r. s = Spanish s. sh = no exact equivalent; similar in sound to the English sh. t = Spanish t. ts = no exact equivalent; similar in sound to the English ts, executed ny a strong oral explosive. x = no exact equivalent; similar in sound to the English sh. y = Spani American Fern Journal 75(1):22-28 (1985) Special Report The Present Evolutionary and Taxonomic Status of the Fern Genus Polystichum: The 1984 Botanical Society of America Pteridophyte Section Symposium DavipD S. BARRINGTON Department of Botany, University of Vermont, Burlington, VT 05405 The fern genus Polystichum has presented major problems in definition and circumscription of species since its description by Roth in 1799. Part of the prob- lem is the vast diversity within the genus: the number of species is reported as more than 175 by Copeland (1947). In addition, hybridization is extremely com- mon (Knobloch, 1976), and agamospory has been reported among experimental plants from Europe (Vida & Reichstein, 1975). There is also substantial evidence of phenotypic and ontogenetic variability within species of the genus. Good progress has been made in solving evolutionary and taxonomic problems in the north-temperate and boreal regions (Manton, 1950: Manton & Reichstein, 1961; — Kurata, 1964; Sleep & Reichstein, 1967; Daigobo, 1972: W. Wagner, 1973; D. Wag- ner, 1979); however, little taxonomic and virtually no evolutionary work has been done on the genus in tropical or austral regions. Hence this large genus of dryop- teroid ferns is in need of substantial attention from systematic and evolutionary biologists, especially in tropical regions. Christopher Haufler at the University of Kansas organized a symposium on Polystichum for the American Institute of Biological Sciences meetings in Fort Collins, Colorado, during August of 1984. As Haufler noted in his opening re- © marks, the purpose of the symposium was to draw attention to a complex and © MORPHOLOGICAL VARIATION AND EVOLUTION IN POLYSTICHUM Warren H. Wagner, Jr.—University of Michigan The systematic problems in the genus Polystichum are best placed in the — context of the structural variation evident at different taxonomic levels. The 4 D. S. BARRINGTON: POLYSTICHUM 23 indusium abaxial and peltate; spores bilateral and monolete; perispore well- developed; haploid chromosome number of 41 or a multiple thereof. Transfor- mation of diagnostic features (such as loss of the peltate indusium) has led to problems in definition of Polystichum as well as the description of various seg- regate genera, which may or may not represent monophyletic groups. In Polys- tichum, as in many fern genera, definition of a group and its component species has been based on subjective arguments such as those from tradition, authority, and consistency—rather than from arguments based on objective documentation and analysis of the variation in structure throughout the group. Several of the genera segregated from Polystichum are in need of critical re- evaluation. Plecosorus, known from Mexico to Panama, differs only in lacking a true indusium (common in Polystichum) and having a modified, recurved margin protecting the sporangia (false indusium). Papuapteris of New Guinea shares both of the critical features of Plecosorus, presumably as homoplasies. New World taxa placed in Phanerophlebia or Cyrtomium have broad lamina segments provided with irregularly anastomosing veins (not with regularly anas- tomosing veins with included veinlets as in true Old World Cyrtomium) but are otherwise typical members of Polystichum. Hybrids of other genera and Polys- tichum are now known. Most dramatic is the hybrid between Dryopteris gol- diana and Polystichum lonchitis from Ontario, on which the University of Mich- igan pteridologists are currently working. Polystichum and its allied genera constitute an array of evolutionary groups that have maintained at least partial genetic homology in spite of isolation and independent evolution. Hybridization and polyploidization commonly generate reticulate complexes in Polystichum. However, a series of plants that appear to be hybrids may ac- tually be the products of divergence. One newly documented complex from Hawaii comprises a group of three bipinnate species that are similar in leaf dissection and general aspect, but differ in chromosome number (n = 41, 82, and 164), leaf and segment shape, habitat preference, and scale structure. Within species, structural features can be highly variable. Numerous trivial forms of the well-known eastern North American Polystichum acrostichoides have been reported and described. A peculiar environmental factor has been pinpointed as responsible for the derivation of forma incisum (Gray) Gilbert, which has variously pinnatifid pinnae instead of typical serrulate pinnae on leaves produced late in a season on damaged plants bearing otherwise typical foliage (W. Wagner et al., 1970). Such striking short-term alterations demonstrate that, in Polystichum, variations may be superficial and should be tested. POLYSTICHUM IN TEMPERATE AND BOREAL NORTH AMERICA David H. Wagner—University of Oregon Polystichum is depauperate in eastern North America (four species, one of which is endemic), but more diverse in the West (13 species, nine of which are endemic). A hypothesis for a reticulate complex involving 11 of the species is 2d on an analysis of the sexual species and seven known sterile hybrids (W. Wagner, 1973; D. Wagner, 1979). From the seven basic diploid species, five tet- 24 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) raploids and one hexaploid have been generated in nature. All but one of the six needed diploid (or tetraploid) progenitors are known. Though fragmentary specimens presumed to represent the sixth, a second progenitor for P. ander- sonii, are on hand, better documentation (complete plants, cytological analysis, or hybrid synthesis) is needed. Despite the current good resolution of the complex, individual specimens are frequently difficult to identify from diagnostic keys. The difficulty may be better understood with reference to Vavilov’s ““Law of Homologous Series in Variation” (Vavilov, 1922). Based on exhaustive studies of crop plants, Vavilov concluded that closely allied species are characterized by similar and parallel series of hereditary variations. In Polystichum, parallel series of hereditary variations are evident in ontogenetic series among the species with divided fronds. For ex- ample, a juvenile frond of P. braunii (fully bipinnate at maturity) is less divided — than a mature frond of P. setigerum, which is merely bipinnatifid at maturity. When plants are under environmental stress, leaf development is arrested in a juvenile, stunted condition, although sori may still be produced. For example, the tetraploid Polystichum scopulinum growing in exposed sites at high eleva- tions closely mimics the tetraploid P. kruckebergii, with which it shares a set of chromosomes contributed by the diploid P. lemmonii. Hybrid taxa, whether — fertile species or sterile hybrids, approach parental extremes in variation when — growing in a habitat typical of one parent. The most nearly parallel series of variations is found in the two sibling species, P. munitum and P. imbricans. Although consistently ecologically distinct, the two are often found in contiguous habitats and hybridize frequently. The hybrids are sterile, as shown by cytological analysis. Continuing studies of these two species should be of great interest, since our understanding of divergent specia- tion in the ferns lags behind our understanding of polyploidization as a route to new species. POLYSTICHUM IN THE WEsT INDIES John T. Mickel—New York Botanical Garden West Indian Polystichum is the most distinctive element in the New-World Polystichum flora. Nevertheless, it has received attention only twice: Maxon’s | (1909, 1912) classic revisionary work and Morton’s (1967) study of the Cuban species. Seventeen of the 30 known species are once pinnate, and 14 are prolifer- ous; both of these traits are rare on the mainland (though common in Japan and ~ the Himalayas). Cuba, Jamaica, and Hispaniola constitute the center for the West — Indian group: 25 of the 30 species are endemic there, whereas only three of the - species reach Puerto Rico and two the Lesser Antilles. Only two species, P. muricatum and Fr. platyphyllum, are found both in the West Indies and in con- — Latin America. Both are widespread species of low and middle eleva- ons. Many of the West Indian species are limited to calcareous rock and soils, — which is doubtless a critical determinant in their geographic distribution. : Rigo ged Polystichum has been studied in detail, hybridization has been shown to play an important role in speciation. Preliminary study suggests that ~ Se ea eM RE EON Tg ne Pa ee ewe TEE Fl Be, Ne ape Ree earthy ae D. 8. BARRINGTON: POLYSTICHUM 25 the West Indian taxa are no exception. Chromosome counts have been made for only six species (Walker, 1966). However, analysis of gross structural features of the sporophytes, as well as of spore-surface structure and spore-shape irregu- larities, when taken together with the few cytological records, demonstrates that there are at least six sterile hybrids and four tetraploid nothospecies among West Indian Polystichum. POLYSTICHUM IN MExIco Alan R. Smith— University of California, Berkeley Recent floristic work has provided the first modern treatments of Polystichum for the southern limits of the North American continent (Smith, 1981; Stolze, 1981). The genus in northern Latin America comprises about sixteen species, mostly confined to southern Mexico. Four species groups can be discerned. The largest group, comprising nine species (all of which are bipinnate, indusiate, and grow at middle to high elevations) has its affinities with the P. setiferum group (section Metapolystichum of Daigobo, 1972). The second group is of two north- temperate species at the southern limits of their ranges: P. munitum from Isla Guadelupe and P. acrostichoides from Nuevo Leon. The third group has a single representative, the rare P. munchii, closely allied to the West Indian P. echin- atum. Like many of its West Indian allies, P. munchii is limited to limy substrates. A fourth group of four species has affinities farther south in continental Latin America. Two, P. polyphyllum and P. speciosissimum, are exindusiate species from high elevations; the other two, P. platyphyllum and P. mickelii, occur in low to mid-elevation rain forests. Seven of the Mexican species have been ex- amined cytologically. Four counts are diploid (n = 41); P. fournieri and P. poly- phyllum are tetraploid (n = 82); and P. platyphyllum, probably a species com- plex, has both cytotypes. In northern Latin America, it appears that species infrequently grow together, and interspecific hybridization has not been docu- mented. POLYSTICHUM IN CENTRAL AND SOUTH AMERICA David S. Barrington—University of Vermont Polystichum in Central and South America can be divided into two geograph- ical and evolutionary groups, one tropical and the other austral. The austral group has received some attention (Christ, 1893, 1905; Looser, 1968), but the tropical one has not until quite recently (Barrington, 1985). The situation in Costa Rica is best known, where the genus is represented by six species above 2800 m. Two are at the southern limit of their range (diploid P. speciosissimum and tetraploid P. fournieri) and two others are common throughout the mountains of the New World tropics (diploid P. lehmannii and tetraploid P. polyphyllum). In addition, two species are apparently endemic, an undescribed diploid and an undescribed tetraploid. Three hybrids have been 26 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) found, including a sterile triploid (with 41 pairs and univalents) between the two endemic species and a sterile triploid between the endemic tetraploid and some unknown diploid progenitor allied to P. polyphyllum. The two endemic species appear to be an allotetraploid and one of its diploid progenitors based on chro- mosomal evidence, morphology, and preliminary electrophoretic analysis (C. H. Haufler, pers. comm.}. Secondary interaction in the Sierra Talamanca of Costa Rica appears to be promoted by human disturbance, because the hybrids all occur in disturbed sites, and the progenitor species are found in proximity most often along roadcuts and in pastures and disturbed cloud forest and alpine areas. Preliminary work in northern South America suggests that there are few species and that hybrids are not common. At high elevations, two broad-ranging species are common: P. lehmannii (along open streambanks) and P. polyphyllum (throughout the paramo). In upper-elevation cloud forests, there is a suite of locally endemic and mostly undescribed species (mostly from the moist margins of small streams). At middle elevations, the broad-ranging P. platyphyllum grows with the locally common P. {or Cyrtomium) dubium. In one mixed population of the two in Ecuador, there are plants with an array of features intermediate between P. dubium and P. platyphyllum that resemble the type of P. dictyo- phyllum. Spores of these intermediates are irregular, allowing the suggestion that at least some plants assignable to P. dictyophyllum are hybrids between P. dubium and P. platyphyllum. Further study of this trio should provide better understanding of the relationship of the anomalous P. dubium to the rest of Polystichum. At low elevations, the broad-ranging P. muricatum is found as far south as Ecuador. This diploid species is peculiar in that sori often abort before maturity, though regular spores are produced during meiosis. Reniform indusia at some sori are a second unusual feature of P. muricatum. SUMMARY In the New World, there are five recognizable elements in Polystichum: the highly reticulated boreal-north temperate element (affinities circumboreal); the Mexican element (affinities warm-temperate Asian); the Antillean element (af- finities in Southeast Asia and the western Pacific); the Andean element (affinities unclear}; and the Austral or South Andean element (affinities austral, including New Zealand and Australia). All of these but the Austral element received some attention in this symposium; all need further scholarly attention. The symposium speakers noted in the discussion that followed the presentations that multiple origin of analogous character states is particularly common in Polystichum: for example, the true indusium has been lost in three of the five elements of the flora, the Antillean, the Andean, and the Austral. Polystichum species, defined on both morphological and cytological grounds, have characteristic habitats and elevational preferences. Hybridization typically yields sterile hybrids and new species via polyploidization, both in north-tem- perate areas and in montane tropics. However, in the tropics, hybridization ap- pears to be localized: the Greater Antilles and Costa Rica are rich in hybrids, whereas other parts of the Neotropics have not yet yielded many hybrids. Either siehianien iii inisaiasiaasiamainits: seerens heii inmiasitecaenmasesn: Sonica ia & i y =_— oii ec stricta scat ecneaanaciunge ine ie pee Pee Pe See ea eae ee Reap RMS PS eS OR aie TOES Fee eee TEN Re es ad heme D. 8. BARRINGTON: POLYSTICHUM ot Students of Polystichum have before them the substantial task of providing monographic treatments for tropical American groups. In addition, there remains the task of identifying phylogenetic groups above the species level among New World Polystichum. LITERATURE CITED BARRINGTON, D.S. 1985. The morphology and origin of a new Polystichum hybrid from Costa Rica. Syst. Bot. 10:199-204. Curist, H. 1893. Les différentes formes de Polystichum aculeatum (L. sub Polypodio), leur groupe- ment et leur dispersion, y compris les variétés exotiques. Ber. Schweiz. Bot. Ges. 3:26-48 ~ 1905. Uber die australen Polystichum-Arten. Ark. Bot. 4(12):1-5. CHRISTENSEN, C. 1905-6. Index filicum. Copenhagen: Hagerup. CopeLanp, E. B. 1947. Genera filicum. Waltham, Mass.: Chronica Botanica. Daicozo, S. 1972. Taxonomical studies on the fern genus Polystichum in Japan, Ryukyu, and Taiwan. Sci. Rep. Tokyo Bunrika Daigaku, Sect. B. 15(224):57-80. Hooker, W. J. 1862. Species filicum. Vol. 4. London: W. Pamplin. Knostocu, I. W. 1976. Pteridophyte hybrids. Publ. Mus. Michigan State Univ. Biol. Ser. 5:277-352. Kurata, S. 1964. On the Japanese ferns belonging to the Polystichum polyblepharum group. Sci. Rep. Yokosuka City Mus. 10:17-41 Looser, G. 1968. Los helechos del genero Polystichum Roth en Chile: notas preliminares. Anales del Museo de Historia Natural 1:49-58. Manton, I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge: Cambridge iv. Press. _ 8nd T. REICHSTEIN. 1961. Zur Cytologie von Polystichum braunii (Sp ) Fée und seiner Hybriden. Ber. Schweiz. Bot. Ges. 71:370-383. : Maxon, W. R. 1909. A revision of the West Indian species of Polystichum. Studies in Tropical American Ferns 2. Contr. U.S. Natl. Herb. 13:25-38. ~~ 1912. Further notes on the West Indian species of Polystichum. Contr. U.S. Natl. Herb. 249-51, Morton, C. V. 1967. Studies of fern types. I. Contr. U.S. Natl. Herb. 38:29-83. A.W. 1799. Auszug eines Briefes von Hrn. Prof. Mertens in Bremen an den Herausgeber. Strep a rob Bot. (Leipzig) 2(1):103. ae A. and T. REIcHsTEIN. 1967. Der Farnbastard Polystichum meyeri hybr. nov. = Polystichum trey (Spenner) Fée x P. lonchitis (L.) Roth und seine Cytologie. Bauhinia 3:299-309, 363- OF ee SMTi, A.R. 1981. Part I Pteridophytes. In Flora of Chiapas, ed. D. E. Breedlove. San Francisco: ae ~aitornia . of Sciences. Srouze, phe 1981. Ferns and fern allies of Guatemala. Part 2. Polypodiaceae. Fieldiana, Bot. n.s. — -522 eee N.I. 1922. The law of homologous series in variation. J. Genet. 12:48-89. VIDA, G. and T. REICHsTEIN. 1975. Taxonomic problems in the fern genus Polystichum caused by 28 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) hybridization. Pp. 126-135 in European floristic and taxonomic studies, ed. S. M. Walters. Faringdon: E. W. Classey. Wacner, D. H. 1979. Systematics of Polystichum in western North America and north of Mexico. Pteridologia 1:1-64. Wacner, W. H., Jr. 1973. Reticulation in Holly Ferns (Polystichum) in the western United States and adjacent Canada. Amer. Fern J. 63:99-115. , D. R. FarrAR, and B. W. McALPIN. 1970. Pteridology of the Highlands Biological Station area, southern Appalachians. J. Elisha Mitchell Sci. Soc. 86:1-27. WaLkeER, T. G. 1966. A cytotaxonomic survey of the pteridophytes of Jamaica. Trans. Royal Soc. Edinburgh 66:169-237. SHORTER NOTES Marsilea quadrifolia and M. vestita in the Floras of Kansas and Missouri.— — In the fern floras of both Kansas and Missouri two species of water-clover ferns, — Marsilea, have been recognized: a native species, Marsilea vestita Hook. & Grev., and the introduced Eurasian M. quadrifolia L. M. vestita is common in central and western Kansas, while M. quadrifolia has been reported from only two counties in the southeastern part of the state (Petrik-Ott, Nova Hedwigia Beih. 61:1-332, 1979). In Missouri, both species have been reported from a single coun- — ty each: M. vestita (as the synonym M. mucronata A. Braun) from Barton County | in southwestern Missouri and M. quadrifolia from Platte County in northwestern Missouri (Steyermark, Flora of Missouri, Iowa State University Press, Ames, 1963). On the basis of specimens known at present, however, Marsilea quadri- — folia should be deleted from the flora of Kansas and M. vestita should be deleted — from the flora of Missouri. The latter is particularly of note as this species has — been listed as endangered in Missouri. Petrik-Ott (op. cit.) reported M. quadrifolia from Kansas on the basis of two specimens, both from the herbarium of the University of Kansas (KANU): Kol- _ stad & Harms 1581 from Cherokee Co., and Holland 1993 from Neosho Co. The reasons for this are unclear as both sheets are annotated as M. vestita by Petrik- Ott, and the only mention of M. quadrifolia on either sheet is on the original ; label of Holland 1993. An additional specimen (Harms 1137, KANU) from Cher- ; okee Co. was not cited by Petrik-Ott but was annotated by her as M. quadrifolia; — this specimen has characters of sporocarp pedicel length (7 mm), position of — attachment of this pedicel (4 mm above the petiole base), and glabrate sporocarps ~ that are characteristic of M. quadrifolia. The specimen also has, however, asym- ’ metrical abaxially hairy leaflets that are longer than wide and have concave inner margins and slightly crenulate terminal margins. In addition, the rhizomes — are 1.0 mm or less thick and lack roots in the internodes. These are all characters - of M. vestita, which normally has short-pedicelled (2-3 mm), basally attached - * hairy sporocarps. FE a ENS ATP aac AMER re RIEL ek Al Og Saucer as MN SHORTER NOTES 29 The long sporocarp pedicel attached above the petiole base and the glabrous sporocarps in this plant are perhaps due to the development of the sporocarp under water, which has been shown to modify pedicel length and sporocarp indument (Bhardwaja, Trop. Ecol. 8:17-21, 1967). Sporocarps are normally pro- duced only when these amphibious plants are emersed and dry, but occasionally the sporocarps develop under shallow water. In such instances the pedicel is attenuated and both the pedicel and sporocarp are almost glabrous. If the petiole is also attenuated, the pedicel, which is in fact a portion of that same leaf rather than the petiole of a separate leaf, may appear above the base of the petiole rather than at the base. In Missouri, Marsilea quadrifolia was reported from Platte Co. by Gier (Amer. Fern J. 45:64-65, 1955) and was recollected there by Dunlap in 1962 (specimen in SMS). Steyermark (op. cit.) reported M. vestita from 1.5 miles north of Milford, Barton Co., on the basis of Palmer 53101. While I have not seen this particular specimen, I have seen others from the same locality: Palmer 54495 (MICH), Palmer 53949 (GH, KANU), Dunlap s.n. in 1961 (NLU, SMS), and Dunlap s.n. in 1963 (SMS). Of these, Palmer 54495 and the Dunlap specimen from 1961 at NLU are fertile. Both bear the paired or triple (occasionally single) long-stalked Sporocarps attached above the petiole base that characterize M. quadrifolia; in addition both these specimens and the sterile ones have symmetrical glabrous leaflets as wide as long or wider, with convex or straight inner margins and entire terminal margins, and rhizomes 1.0-1.3 mm thick bearing internodal roots, further confirming this identification. A specimen of Marsilea from yet a third locality in Missouri was included among specimens sent from SMS: Boone Co., Watkins pond southwest of Mid- Way, 12 June 1963, Dunlap s.n. Although this specimen is of a sterile, floating- leaved form, in which both species would have glabrous symmetrical entire- Margined leaflets on long flexuous petioles, its stouter rhizomes and internodal Toots indicate that it is also M. quadrifolia. The criteria presented by Petrik-Ott (op. cit.) for distinguishing the two species are fairly reliable as long as adequate fertile material and leaves formed on land are present in a collection; the additional characters mentioned here are sup- plementary ones that can be used with scanty fertile or even sterile material, regardless of whether it grew on land or in water. Caution should be exercised in applying these characters outside the Great Plains region, however, as some of them, e.g., internodal roots, are found in extralimital species as well. Marsilea quadrifolia appears to be spreading westward from northeastern orth America, and collectors should continue to seek it both from additional localities in Missouri and from Kansas. M. vestita, although it is not expanding its range, is probably locally dispersed by migrating waterbirds and may even- tually thus be brought into Missouri from nearby stations in eastern Kansas and ebraska, _ {thank the curators of GH, KANU, MICH, NLU, and SMS for making spec- mens available for this study, and R. Brooks, W. H. Wagner, Jr., and an —— ymous reviewer for comments on the manuscript.—Davip M. JOHNSON, Division of Biological Sciences, University of Michigan, Ann Arbor, MI 48109. j 30 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) New Records for Longevity of Marsilea Sporocarps.—In 1952, Allsopp (Nature 169:79-80) reported the viability of Marsilea sporocarps 61.5 and 68 years old. These sporocarps were obtained from collections made by C. G. Pringle in Mex. — ico in 1890 and by W. N. Suksdorf in Washington state in 1883. Recently, I successfully cultivated sporophytes from herbarium material 99-100 years old. In fall of 1982, sporocarps of Marsilea crenata Presl collected by D. H. Camp- — bell in the Philippines in 1911 (from the herbarium of the University of Michi- gan-MICH) were scarified and placed in water for a class demonstration. Young sporophytes from these 71 year-old sporocarps were apparent a few days later. Success with these sporocarps prompted me to try to obtain plants from sporo- carps that were still older. In December 1983 I was successful in producing — sporophytes of M. vestita Hook. & Grev. from a specimen on loan from Field Museum (F) collected by K. Brandegee in California in 1906 (77 years old) and of M. mollis Robinson & Fernald from a specimen (also on loan from F) collected by C. G. Pringle in Oaxaca, Mexico, in 1894 (89 years old). Finally, in March — 1984 I obtained young plants from a collection on loan from Gray Herbarium — (GH) of M. oligospora Goodding made by W. N. Suksdorf in 1883/1884. At 99 to 100 years of age these are the oldest sporocarps yielding sporophytes yet recorded in Marsilea. Voucher specimens of plants grown from these collections : will be deposited in MICH and at the institutions that provided the viable spore material. In view of reports of apomictic production of sporophytes in Marsilea (Bhard- — waja & Abdullah, Nova Hedwigia 21:521-528, 1972), I thought it possible that i these sporophytes were being produced directly from megagametophytes with- out fertilization, but isolated megaspores from the Brandegee, Pringle, and Suks- dorf collections yielded no sporophytes while the controls did, and micro- and megagametophyte development appeared normal. aximum ages that viable Marsilea sporocarps are capable of reaching | are yet to be determined, and it is thus valuable to have the abundant and widely | distributed collections of such collectors as Pringle and Suksdorf for periodic sampling. Already, however, there is differential behavior among specimens. Only sporocarps from the Suksdorf collection at GH produced sporophytes; spo _ rocarps from other herbaria, while they opened normally after scarification, showed neither gametophyte nor sporophyte development. This may represent natural loss of viability but may also be viability loss due to treatment with microwaves for fumigation (Hill, Taxon 32:614-615, 1983). As Hill (op. cit.) point ed out, having viable propagules present on herbarium sheets often makes it possible to obtain live plants of species otherwise difficult or expensive to collect. For Marsilea there is the added advantage that plants grown from specimens with cauanentis sporocarps yield improved material for description and study; spec: imens with mature sporocarps usually have leaves in poor condition. : Recently, Bhardwaja (Aspects of Plant Sciences 3:39-62, 1980) reported the production of gametophytes (but not sporophytes) from M. burchellii A. Brau? sporocarps 130 years old. In time perhaps this finding can be matched by the production of sporophytes from specimens equally old. : I thank the curators of F, GH, and MICH for making t! imens available Tia erg Psd acne melee ol are cide wil ee Ral ha al hacia det es alate SHORTER NOTES 31 on loan, and W. H. Wagner, Jr., for comments on the manuscript.—Davip M. JOHNSON, Division of Biological Sciences, University of Michigan, Ann Arbor, MI 48109. Nomenclatural Notes on some Ferns of Costa Rica, Panama, and Colombia.- II.—This is a continuation of the series begun a few years ago (Amer. Fern J. 67: 58-60. 1977) to record changes of names pertinent to ongoing floristic projects. Alsophila imrayana var. basilaris (Christ) Lellinger, comb. nov.—Cyathea bas- ilaris Christ, Bull. Herb. Boissier II, 4:949. 1904.—LEcToTyPE (chosen by Gastony, Contr. Gray Herb. 203:127, 1973): Costa Rica, Wercklé (P not seen; authentic material NY, US). Glyphotaenium trifurcatum (L.) Lellinger, comb. nov.—Polypodium trifurcatum L., Sp. Pl. 2:1084. 1753.—Type: Based on t. 138 in Plumier’s “Tractatus,” which illustrates a specimen collected by Plumier on Martinique. Thelypteris frigida (Christ) Smith & Lellinger, comb. nov.—Aspidium frigidum Christ, Bull. Herb. Boissier II, 6:160. 1906.—Type: Costa Rica, Pcia. Car- tago, Volcan Turrialba, Wercklé (P not seen). Davip B. LELLINGER, Dept. of Botany, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560. REVIEW “The ferns and fern allies of Southern Africa,” by W. B. G. Jacobsen. 542 pp. ISBN 0-409-09836-1. Durban, South Africa: Butterworth Publishers Ltd. 1983. R70.40 (ca $60.00). The plan of this book is much like that of a college textbook, with parts, numbered sections, and shaded tables. There are two parts: Part I deals with ecology, environment, and distribution; Part II covers the pteridophytes in sys- tematic order. The area covered includes not only South Africa, but all of south- ern Africa south of the 17th parallel. The first part of the book is excellent—I have seen ecology textbooks with less detailed discussion of factors affecting distribution. Pteridophyte examples are used throughout this discussion of factors such as topographic (elevation), edaph- i; (geomorphology and soils), climatic (rainfall, temperature, slopes, light, wind), and life forms. For those of us in the northern hemisphere it helps to remember t southern Africa is sub-tropical since the tip is at 35° S (about the same as Los Angeles}, and that north-facing slopes are warmer and drier. plant communities are described in a fairly long chapter. There is a list 32 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 1 (1985) of localities, dominant trees, undershrubs, vines, and in tabular form the rare, occasional, and frequent pteridophytes to be found in each. The photographs j vary in quality, some giving a good representation of the forest type, fern, etc, ] others are not as good. The second part, dealing with the southern Africa Pteridophyta, begins witll 1 some general descriptive material on collecting and preserving specimens, and a glossary (an annoying place for this, in the middle of the book). Since I am not | familiar with the fern flora, this part of the review concentrates on the more mechanical aspects of the book. There is both a technical key to the families, and a more descriptive key for families or genera. The keys throughout (includ. ing the keys to species of each genus) are a nuisance to use, and this is the only ] really objectionable part of the book. They are of the indented type, but not indented. So, for the first key, to make comparison “1,”’ one reads through a. vertical column of numbers: 1, 2, 3, 3’, 2’, 4, 4’, 2” (help!), 1’, 5, 6,.... It is difficult enough to find and read the paired comparisons without having o WOITy about | whether there is a third statement embedded in the column someplace. Fortu- nately there are not third choices very often as far as I could tell, but it takes a lot of unnecessary work to find each pair of statements. The characteristics are clear, and the statements are generally well contrasted. There are keys to the genera within families and to species within genera. I tried a few where 1 had material available, and found no problems, but my material was from cule ' tivated plants and not extensive enough to really evaluate the keys : For each species there is a description, discussion of distribution and ecology, a photograph of an herbarium specimen, and a nice distribution map. There a . a few line drawings, which are more helpful than the photographs. I found a few errors, but none of any real consequence: map 81 should be map 8 (p. 138), Mickel is several times spelled Mickels, Torres Street should be Torres Strait (p. 382), and a few others of even lesser consequence. : With the single exception of the keys, I liked the book. It would certainly be helpful if one were planning a trip to this fascinating part of the world, ani 1 enjoyed reading the ecological section and thinking about going. The parts on ecology of pteridophytes, although referenced to southern Africa, have wid general application and make a nice introduction to this neglected area of pter idology.—JAMes D. Montcomery, Ichthyological Assoc., Inc., Berwick, PA 18603. cD ERE ee aM ee Ta ee ine as ae emeet ge lap eee Re Weenies PARSE, NT Ie Wee epee a See REE GN eB Team MS Suey BIEN Pie i ye MDT Ser a ees ee RE ME Me Mon ROR MES ke eT Lee ae See pen : scale and reference to latitude and soaps in each map. INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable), including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 812 by 11 inch paper of good quality, not “erasable” paper. Double manuscripts throughout, including title, authors’ names and addresses, text (including heads and keys}, literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author’s name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. Use S.1. (metric) units for all mea- sures (e.g., distance, elevation, weight} unless quoted or cited from another source (e.g., specimen citations). For nomenclatural matter {i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- breviate titles of serial publications according to Botanico-Periodicum-Huntian- um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as ots of nomenclatural matter are not included in lit- - erature cited. For shorter notes and reviews, put all references paren Illy — in text. Use Index herbariorum (Regnum Veg. 106:1-452. 1981) for designations of herbaria. Illustrations should be proportioned to fit page width with caption on the same Page. Provide margins of at least 25 mm on all illustrations. For continuous-tone . illustrations, design originals for Aegina without i mniform — amount. In compote blocks, ee t pI t g! .phs . Avoid com- a ne and linwees cbc kt cu se Ahlen Coordinate _ sequence and numbering of figures toot of Tables) with viel of ence —— 2 . ain scales and symbols in - ves, not tude a A splendid four-color volume... Including the Ferns, the Horsetails, and the C lub Mosses Written and illustrated by ELFRIEDE ABBE This handsome volume is a reproduction of a now-rare limited edition of 150 copies which first appeared in 1981—the magnificent accomplish- ment of one gifted person, Elfriede Abbe, who was the author, illustrator, designer, typesetter, printer, and binder. The text describes 27 members of the Filicales, covering their habitat and distribution, culture, history, and use Collectors of rare books, herbalists, fern enthusiasts, and other horti- culturalists will want to own this fine volume. About the author: _Elfriede Abbe was artist-in-residence in the Botany Department at Cornell before she retired. She has continued her career as an nd hibited internationally. 70 wood engravings and drawings, includ- ing 20 full-color, full-page plates. $35.00 Also of interest — GARDEN SPICE AND WILD POT-HERBS An American Herbal By W.C. MUENSCHER and MYRON A. RICE With illustrations by ELFRIEDE ABBE $12.95 paper ORNEL s ‘UNIVERSITY PRESS _ 20 Box 250, —— pee York 14851 SERRATE CMH ENE ee NS aE ee a ES Se PSE ee ep “ AMERICAN oe FERN oe April-June 1985 | QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Three New Anemias from Northern South America John T. Mickel 33° Sporopollenin Content of the Spore Apparatus of Azolla R. E. Toia, Jr., B. H. Marsh, S. K. Perkins, J. W. McDonald, and G. A. Peters 38 Role of Morphological Char. Characteristics of Leaves and the Sporangial Region in the Tax- onomy of Isoetes in Northeastern North America _L. S. Kott end D. M. Britton 44 A ie ah tislypick hes Pee ae = George R. Proctor Shorter Notes Dryopteris ludoviciana and D. x australis New to Arkansas | James H. Peck, Steve L. Oral Eric Sundell, and Carol J. Peck ne Se inom Sees Anh Sls: Ahi, Sete Seng br Sen eer oa S David K- Suaith oa Pa AG els Sate 5 Saat aca ales aa bali The American | Fern Society _ ‘TERRY R. WEBSTER, Biological Sciences apse le of Connecticut, Storrs, CT 06268. President FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48 VieePresde MICHAEL I. COUSENS, Faculty of Biology, University of West Florida, Pensacola, FL 3 meee JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, Knoxville, TN 37916. Treasurer DAVID S. BARRINGTON, Department of Botany, University of Vermont, Burlington, VT 05401 Records Treasurer ALAN R. SMITH, Dept. of Botany, University of California, a CA 94720. rite Editor DAVID B. LELLINGER, EPO: an Institution, Washington, DC 2 moir Editor DENNIS Wm. STEVENSON, Barnard College, Columbia aa oor York, nt se Fiddlehead ene Editor —— — Journal POR Oe ee Dept. of Botany, eS of California, Berkeley, CA 94720 ASSOCIATE EDITORS recat oo. aa Dept. of Biology, Indiana University, Bloomington, IN 47401 R HAUFLER Kansas, or es Dept. of Botany, patient: ‘DAVIO GD, LELLINGER -... 2s... U.S. Nat'l Herbarium NHB-166, Smithsonian “Tnstitatiods eee ashington, DC 20560 _ TERRY R. WEBSTER | _ Biological Sciences Group, University of Connecticut, Storrs, CT 06268 is The “ “American Fern jounat” GSEN 0 seebonane is an illustrated geen devoted to the general _ ern Society, and published at the Pringle Hi ———— - Univer. ve 2 ty of Vermont, B tinsinn VA 06401 Bacned-cl t id ditional See eo : ‘ * o rs 4 ~~ i 7 ea a ae ae = z ee ree. os Canker ae s,s < e a PEG Bens st. fe the Hatin nf teat y Ce a ce a << e % hould be add: d to Dr. _— Db Montgomery, Ichthyological Asso- : | Shae, 2 ‘Berwick, PA 18603. ns aC HE b k t to Dr. David S. me ” American Fern Journal 75(2):33-37 (1985) Three New Anemias from Northern South America JOHN T. MICKEL New York Botanical Garden, Bronx, NY 10458 Of the over 100 species of Anemia, about 65 are found in southern and eastern Brazil with a secondary center of speciation in Mexico (20 species). Central America and the western and northern portions of South America have provided relatively few species, those being mostly very widespread taxa. Apparently this is a misleading picture since we here report three new species from northern South America—one from northern Brazil and two from essentially a single locality near Puerto Ayacucho in southern Venezuela. All three species are known from single collections and were found among granite rocks. It would seem that these areas have not been well collected, and since Anemia generally favors grassy and/or rocky habitats, there may be additional taxa yet to be discovered in southern Venezuela and Colombia and northern Brazil. Anemia antrorsa Mickel, sp. nov. (Figs. 1G-I).—Type: Brazil, Amazonas, lower slopes of Pico Rondon, Perimetral Norte Highway Km 211, 3 km from Km 211, granite rock outcrop, 2 Feb 1984, Prance et al. 28739 (holotype NY; isotype UC). Anemiam tomentosam var. australem Mickel forma laminae simulans sed pinnis fertilibus brevibus antrorsis falcatis et laminae pilis contortis discreta. (L., antrorsus, directed upward, referring to the habit of the fertile pinnae.) Rhizome horizontal, compact, ca. 6 mm diam., with brownish-yellow hairs, 5- 8 mm long; fronds erect, sterile and fertile fronds alike, 28-34 cm long; stipe slightly less than half the sterile frond length and slightly more than half the fertile frond length, densely hirsute with orange hairs 3-5 mm long; blade del- tate-ovate, 15-17 cm long, 8-10 cm broad, bipinnate-pinnatifid, lamina papyra- ceous, dull, pilose on both surfaces with multicellular hairs 0.5-1.0 mm long twisted on abaxial surface), densely tomentose on rachillae, lower surfaces and rachis; sterile pinnae 9-12 pairs, subsessile (petiolules 1-2 mm), upper pinnae Perpendicular to the rachis, lower pinnae slightly ascending, 5-7 pinnules per Pinna, the lower pinnae slightly more exaggerated basiscopically, lobes acute to ” , Margin entire to crenulate; veins free, evident; fertile pinnae remote from the sterile pinnae, ca. half as long as the adjacent sterile pinnae, ascending at sal angle to strongly upwardly falcate, short-petiolulate (3-4 mm), the ultimate isions with narrow laminar tissue; spores tetrahedral, vertically compressed, striate, the ridges broad with narrow grooves between, 69-79 (av. 76.1) ym diam. (Figs. 2A, B). This species bears a strong resemblance to A. tomentosa var. australis Mickel, but A. antrorsa is quite distinct with its short, suberect pinnae, strongly hairy stipe and rachis, and twisted abaxial laminar hairs. Such habit for fertile pinnae 's otherwise well known in the genus, occurring in all three subgenera: subg. IZ et Ma, ) _@ i 1 } i} | i J. T. MICKEL: ANEMIA - Oe ait amma Pe reel "canal KG Spores of Anemia. A, B. Proximal and distal faces of A. antrorsa, x 467. C, D. Proximal and Fic. 2 di istal faces of A. ayacuchensis, x 600. E, F. Proximal and distal faces of A. porrecta, 467. 36 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985} Coptophyllum (sect. Anemibotrys: A. aspera, A. perrieriana; sect. Trochopteris; A. elegans, A. eximia), subg. Anemiorrhiza (A. colimensis), and subg. Anemia (sect. Adetostoma: A. brandegeea, A. clinata, A. intermedia, A. salvadorensis). | This is the first report for sect. Tomentosae of subg. Coptophyllum. Anemia ayacuchensis Mickel, sp. nov. (Figs. 1A-C).—Type: Venezuela, Ama- ~ zonas, 8 km §S of Puerto Ayacucho and just S of the Rio Carinagua, | savannas between the Rio Orinoco and the road, growing at base of — boulders, 90 m, 31 Oct 1971, Davidse 2740 (holotype UC; isotype MO, © not seen). Ab. A. jaliscana Maxon sporis spinosis atque rhizomatis pilis cupreis differt. Rhizome horizontal, compact, 6 mm diam., hairs reddish-orange; fronds erect, sterile fronds 10-16 cm tall, fertile frond 24 cm tall; stipe 0.5 mm diam., 4-"4 of ' the sterile frond length, ca. °{ of the fertile frond length, stramineous, glabres- cent; blade oblong (7 cm long, 2.8-3.6 cm broad), once pinnate, texture papyra- ceous; pinnae 5-6 pairs, apical pinna obtuse with one (less commonly two) large — lateral lobes; pinnae opposite, oblong, truncate to cuneate at base, hirsute on | both surfaces, obtuse at apex; margin minutely denticulate; veins free; fertile pinnae erect, long-petiolulate, approximate to the sterile pinnae, far surpassing the sterile blade in height, the ultimate divisions lacking lamina; spores tetra- hedral-globose, striate, the ridges narrow with clavate spines, 65-70 (av. 67.8) hm diam. (Figs. 6, 7). This is amazingly similar to A. jaliscana Maxon of western Mexico (see Mickel, _ 1982), but is distinct in its spore ridges being spiny (rather than smooth) and the rhizome hairs reddish orange (rather than brownish yellow). r 5 Anemia porrecta Mickel, sp. nov. (Figs. 1D-F).—Type: Venezuela, Amazonas, — along road from Puerto Ayacucho to Samariapo near crossing with Rio Cataniapa, savanna and granite dome, on wet granite rock in dense shade of other rock, 31 July 1967, Wessels Boer 1922 (holotype NY). Lamina deltata, segmentis rotundatis vel ovatis, et pinnis fertilibus angulo 45 rigide adscendentibus insignis. (L., porrectus, stretched out in a straight line, alluding to the habit of the fertile pinnae.) | Rhizome horizontal, compact, 5-6 mm diam., hairs brownish orange; fronds erect, sterile and fertile fronds alike, 9-21 cm tall; stipe 0.6-0.8 mm diam., ca. 2 of the frond length, stramineous, hirsute; blade deltate, 5-10 cm long, 3.3-8.0 cm broad, pinnate to bipinnate-pinnatifid, texture papyraceous; pinnae 5-7 pails, . slightly ascending, apex pinnatifid, pinnae opposite to subopposite, short-petiol _ ulate (1-2 mm), anadromous, 4-6 pairs of pinnules, pinnules and their lobes somewhat rounded or ovate, acute to obtuse, margin shallowly crenulate, hirsute — on both surfaces, most densely so on lower surface of costules; veins free; fertile pinnae held rigidly at 45° angle, short-petiolulate (3-5 mm), remote from the — - : — pinnae, nearly as long as the longest (basal) sterile pinnae, the ultimate ivisions with narrow lamina; spores tetrahedral, vertically compressed, striate, ne ae : ene ree neh te gre tgte: saute Fe au aaa gee SEL CRE PEA ae MR ee ee J. T. MICKEL: ANEMIA 37 the ridges broad and smooth with narrow grooves between, 79-99 (av. 89.9) um diam. (Figs. 2E, F). Anemia porrecta apparently has no close relatives and probably belongs to subg. Coptophyllum sect. Tomentosae. The stomata are attached, not floating. The epidermal cells are highly contorted, much like those of A. elegans of sect. Trochopteris, but the habit is quite unlike members of that group. It is readily identified by its rounded segments and rigidly held fertile pinnae at a 45° angle. This study was supported in part by a grant from the National Science Foun- dation, DEB 82-09956. I am indebted to Donald Black for the SEM photomicro- graphs and Bobbi Angel for the drawings. LITERATURE CITED MickEL, J. T. 1982. The Mexican species of Anemia (Schizaeaceae). Brittonia 34:388-413. REVIEW “Arkansas ferns and fern allies,” by W. Carl Taylor. 1984. 262 pp. Milwaukee Public Museum. ISBN 0-89326-097-5. $29.95. This excellent treatment stands, along with R. H. Mohlenbrock’s “The Illus- trated Flora of Illinois: Ferns” and T. M. C. Taylor's “Pacific Northwest Ferns and Their Allies,” among our most beautiful and useful pteridophyte floras. The introductory material on fern morphology, life history, names, and Arkansas geology and fern distribution makes the book entirely comprehensible to neo- phytes to fern study. The key to genera is thoughtfully illustrated with an ex- ample of each genus, and so it is virtually impossible to go wrong at this critical point in making an identification. Seventy-two species in 31 genera are treated. Each species has a short description, brief list of synonyms, statement of habitat and range, and useful notes. The notes include cytological and hybridization information and, in the case of rare or newly discovered species, some infor- mation about the plants’ discovery. Distribution maps (one dot per county) are included for each taxon, as are very brief specimen lists that provide documen- tation for the maps. The illustrations, by Paul W. Nelson, are among the best published for United States pteridophytes. Many seem to have been drawn from life. All are reproduced clearly and in large size. The book concludes with an ample glossary, literature cited, a checklist of Arkansas pteridophytes, and an index to common and scientific names. The book is available from the Publi- Cations Division, Milwaukee Public Museum, 800 W. Wells St., Milwaukee, WI 93233. Everyone interested in the pteridophytes of the south-central United States or who enjoys excellent illustrations of pteridophytes will want to own a opy.—Davip B. LELLINGER, Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560. American Fern Journal 75(2):38-43 (1985) Sporopollenin Content of the Spore Apparatus of Azolla’ R. E. Tora, JR., B. H. Marsh, S. K. PERKINS, J. W. MCDONALD, and G. A. PETERS Battelle-C. F. Kettering Research Laboratory, 150 East South College Street, Yellow Springs, Ohio 45387 In 1928 Zetsche and Huggler used the term sporopollenin to describe the exine (outer coat material) obtained from either spores or pollen grains (Shaw, 1971). © This material, which is comprised of oxidative, cross-linked polymers of carot- enoids and/or carotenoid esters (Brooks & Shaw, 1968; Gooday et al., 1974), is i resistant to ultraviolet irradiation, desiccation, biological decay, and non-oxida- — tive chemical attack (Shaw, 1971; Strohl et al., 1977). The extensive work of — Zetsche and co-workers (1928-1937), and the later studies of Shaw and co-work- . ers (1964-1974}, summarized by Brooks and Shaw (1971) and Shaw (1971), have — provided substantial information on the chemistry, biochemistry, and geochem- — istry of sporopollenin. Sporopollenin has a wide taxonomic distribution, occurring in myxobacteria, fungi, algae, pteridosperms, gymnosperms, and angiosperms (Brooks, 1971; At kinson et al., 1972; Gooday et al., 1974; Strohl et al., 1977; Honegger & Brunner, 1981). Sporopollenin in sporocarps of Azolla—a heterosporous, free-floating fern that occurs worldwide in freshwater habitats (Moore, 1969)—was inferred his- tochemically (Lucas & Duckett, 1980), but not chemically defined. Therefore, we undertook this effort to verify the presence of sporopollenin in Azolla spores, — and gain some insight into its contribution to the structure of spore apparatus. — We have now identified this compound in Azolla spore apparatus by infrared — spectroscopy (IR) following acetolysis. The percent (by weight) sporopollenin of — the spore apparatus has been determined, and its contribution to spore apparatus — architecture demonstrated with scanning electron microscopy prior to and fol- — lowing acetolysis. : MATERIALS AND METHODS Azolla mexicana Pres] was used as the source of sporocarps. The original — population was collected from the Graylodge State Waterfowl Area, Butte Coun- ty, California, during the summer of 1974 by S. Ela. The culture we utilized was obtained from D. W. Rains and co-workers of the University of California-Davis — in 1978, and subsequently maintained at this laboratory. Cultures were grow? — ys nitrogen-free IRRI medium (Peters et al., 1980) and maintained under am | illumination of 100-200 uE/m*/s, with a 16/8 hr, 26/18°C light-dark regime Voucher specimens are maintained at the New York Botanical Garden. Mature ; * Contribution No. 865 from the Battelle-C.F. Kettering Research Laboratory. TOIA ET AL.: AZOLLA SPOROPOLLENIN 39 megasporocarps, microsporocarps, and microsporangia were harvested from the frond material by agitation in large volumes of water, and separated from the majority of the frond debris by sieving. Individual spore types were isolated by centrifugation on Percoll discontinuous density gradients at 16,000 x g for 5 min. Microsporangia banded at the 15-30% Percoll interface, microsporocarps band- ed in the 30% Percoll region, megasporocarps sedimented into the 45% Percoll layer, and the remaining debris formed a pellet on the bottom of the centrifuge tube. The spore types were removed from the gradients and repeatedly washed with water. They were then air dried at room temperature for 48-72 hr, and stored desiccated at room temperature. The sporopollenin of Lycopodium spores has been well characterized by chemical studies (Brooks, 1971; Shaw, 1971) and was employed as a reference material in this study. The Lycopodium spores were kindly provided by W. Doyle, University of California-Santa Cruz; the material was originally obtained from Carolina Biological Co. as Lycopodium powder, and, as such, may repre- sent a mixture of spores from more than one Lycopodium species. Acetolysis of the dried material was based on the method of Atkinson et al., 1972. The acetolysis residues were ground into a homogenous powder with KBr as a matrix substance, and pelleted at high pressure for use in infrared spec- troscopy. Spectra of these samples were recorded on a Beckman IR-20A Infrared Spectrophotometer. A Perkin-Elmer Model 240 Elemental Analyzer, equipped with an MC-341 Microjector from Control Equipment Corp. for sample automation, was used to analyze dried, pulverized samples of spore material for carbon, hydrogen, and nitrogen. For scanning electron microscopic (SEM) observations, acetolysis residue ma- terial was mounted on aluminum SEM stubs with double-sided tape, and sputter coated with 20 nm platinum. Fresh spore material, which was used for compar- ison, was fixed in 2% glutaraldehyde in 50 mM PO,, pH 7.2, dehydrated in a graded acetone series, and critical point dried using liquid CO, as the transition uid. This material was then mounted and coated as indicated above. Exami- nation was performed in an ISI DS-130 scanning electron microscope. RESULTS AND DISCUSSION Sporopollenin can be defined at the practical level as acetolysis-resistant ma- terial (Brooks, 1971; Atkinson, 1972). IR spectra of acetolysis-resistant material obtained from Azolla spore apparatus compare very well with those obtained for Lycopodium spores carried through the same procedure (Fig. 1), as well as with published spectra for Lycopodium sporopollenin (Atkinson et al., 1972; Honegger & Brunner, 1981). The minor differences between the spectra are prob- ably due either to the fact that sporopollenin preparations exhibit variability since their precursors (carotenoids and/or carotenoid esters) vary among ssa el isms (Strohl et al., 1977), or to technical difficulties in grinding the acetolysis- resistant material into a homogenous powder in preparation for IR analysis {Ho- negger & Brunner, 1981). The band at approximately 2350 cm~* could be due to 40 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) A ei ie | , T Ul | | Ul T | ' T ] tT T T l Zz =) L di copodium o Lycopodium P Z2 10 bos | cS 4 ’ ww 4 i . ee o 30 + H T = : H } 4 i i ‘ 2S bd T = ’ = 20 — = a e vr = 10° . f g d 0 = (a rans ue os co N a — Oo — = cf. Ss 2 N ni — — = - a < = = Oo = a a. = = =< n oO > =< —_ < wn = -— a ae" = — — oO i ” oa = = ce a tu _ oO oO te o =< = ce a pad =z 9 a ae = =) = oO ui li lu foal lJ ine x = = Fic. 1. Comparison of leaf length of Isoetes based on population means. Number of populations sampled indicated parenthetically for each taxon. of a corm is therefore virtually impossible to determine with accuracy. Karrfalt (1977) speculated on the age of Isoetes corms; however, his studies did not con- sider tissue previously shed, and therefore his results cannot be conclusive. Roots emerging from the base of the corm have a large root cap (Peterson et al., 1979). The roots branch dichotomously and have a stigmarian type of anat- omy, possessing a large central air cavity surrounded by cortex, with the vascular _ tissue eccentrically placed. While roots are young and actively growing, they appear white. Later, only the root tip remains white, while the body of the root becomes a uniform brown. The gross morphology of roots does not provide — characteristics of taxonomic value because the roots are morphologically very | constant and generally behave in the same manner in all species considered in i The leaves provide all of the taxonomically important vegetative information | ipsa ore ; 4 4 KOTT & BRITTON: ISOETES ~ TaBLE 1. Variability of Leaf Length and Number within Sample Populations of Isoetes spp." Leaf length Leaf number Mean (cm) SD CV Mean SD CV I. macrospora 5.6 1.3 22.2 11.6 3.1 26.4 6.8 2.0 28.7 21.7 4.0 18.6 I. tuckermanii 8.9 2.6 28.7 11.1 3.0 26.7 9.1 a7 18.5 13.4 3.1 22.7 I. riparia 14.8 2.8 19.1 11.0 3.7 34.0 10.5 1.3 12.6 11.3 4.2 37.5 I. acadiensis e223 3.0 13.5 17.8 8.8 49.8 9.1 2.7 29.7 28.0 12.6 45.0 I. echinospora 9.6 3.1 32.0 15.5 4.3 29.9 7.6 1.5 20.0 14.2 2.9 20.2 I. eatonii 14.5 2.7 18.9 34.0 7.2 21.1 47.4 7a 15.8 46.0 49.8 108.3 I. engelmannii 11.4 1.8 15.6 21.4 27 12.6 18.8 1.3 6.9 16.0 9.2 57.6 ‘ Isoetes hieroglyphica plants were too few to analyze statistically. about the species. Leaves are awl-shaped, but vary from very fine and thread- like throughout to thick and stout, flaring at the base and with an abrupt point at the tip. Leaf color is typically bright green, but varies from yellow-green, to dark green, and occasionally to reddish-green. Those parts of the leaves that are buried in the substrate are opaque white or brown. Leaf bases of some species that occur outside the study area are chestnut-brown and persistent as in I. melanopoda Gay & Dur. Internally the leaves have four longitudinal air chambers that are traversed by thin diaphragms at regular intervals. The vascular tissue is located in the central tissue of the leaf. Leaf Characteristics.—Leaf measurements have been used as supporting diag- nostic evidence in several keys (Eaton, 1900, 1908; Pfeiffer, 1922) but data pre- sented here indicate that leaf length, at least in part, is variable and perhaps strongly dependent on ecological conditions and the vigor of the plant. Mea- surements of leaf length yielded results seen in Figure 1. The standard error is large for taxa that have large leaf length ranges reflecting the great sensitivity of the leaf length to environmental conditions. Isoetes engelmannii and I. eatonii had the longest leaves, with means of 26 and 23 cm respectively, but both also had considerable range of variability as by the high standard errors. Isoetes riparia had a grand mean of 15 cm with population means of 33 cm at the upper end of its range. The remaining species had leaf length grand means between 8 and 10 cm. Even the leaf length ranges are similar for these last species. This observation indicates that leaf length is not a good taxonomic character for most of these species. The variability of leaf length within some sample populations is shown in Table 1 as coefficients of variability (CV). A high CV value indicates a high 48 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) TABLE 2. Variability of Leaf Length and Number between Populations within Species of Isoetes. Leaf length Leaf number Grand Grand mean Range mean Range (cm) z SE CV (cm) > SE CV I. macrospora 7.6 3-17 2.7 34.4 19.1 4-53 11.6 60.6 I. hieroglyphica* 8.0 5-11 a ste 13.0 7-16 ~ — I, tuckermanii 9.2 4-23 3.6 39.1 22.4 10-45 8.5 37.9 I. riparia 14.6 6-33 6.2 42.5 14.8 5-36 7.0 47.3 I. acadiensis 9.3 2-21 5.0 53.9 25.6 9-35 15.2 59.6 I. echinospora 9.6 4-25 4.4 46.2 14.5 7-26 4.4 30.4 I. eatonii vit tg 8-46 13.5 59.3 46.6 12-98 31.2 67.1 I. engelmannii 26.0 10-50 11.8 45.6 27.2 9-61 474 63.1 * Isoetes hieroglyphica populations too few to analyze statistically. degree of variability, whereas a low CV value indicates a low degree of vari- ability. All CV values for leaf length were above 6.9, generally most were over 18, and the highest being 32 (Table 1). Between populations, leaf length CV values were even higher than within populations. These high values, which ranged from 34 to 59 (Table 2), reflect the effect of environment on individual plants during the time of leaf development. Leaf number per plant was recorded because this number has been used as diagnostic or supportive taxonomic evidence in the past. Results of the means and the ranges of leaf number per plant are shown in Figure 2. Isoetes eatonii had the highest grand mean at 47 leaves per plant, as well as the largest range of population means (12-98). Isoetes acadiensis, I. tuckermanii and I. engelman- nii had grand means in the 20s, while the remaining species had grand means below 20 leaves per plant. The extremely large variation in leaf number (=range) within species strongly suggests that leaf number is a function of factors such as age and vigor of plants. Habitat perhaps may play an important role since plants inhabiting shallow waters may have a reduced leaf number due to wave or ice action, while those in deep, undisturbed waters may exhibit higher leaf numbers since older leaves can accumulate on the corm. Variability in leaf number was compared among plants of a few sample pop- half being 30 or higher. Older plants tend to have more leaves because they — have a larger corm, while it takes new plants several years to build up to an average number of leaves. Because a population consists of both young and old — individuals, leaf number is not reliable diagnostic feature. The CV values than within popul population. also tend to be higher within species (i.e., between populations) un Populations (Table 2). They range from 30 to 67. Such values show — great variability of leaf number from plant to plant and from population to _ ! | : om cea I mage KOTT & BRITTON: ISOETES 1407 1304 1204 mean population leaf number T i ' — ' — range of population means oo | pale i of highest leaf umber of an invidual > a ae ee ome of ow ww wed standard error — grand mean Cee aoe Se EATONiI (42) I, (48) (16) ENGELMANNII i, ACADIENSIS (41) (129) (99) TUCKERMANII ECHINOSPORA MACROSPORA RIPARIA e (127) (8) HIEROGLYPHICA Fic. 2. Comparison of leaf number of Isoetes based on population means. Number of populations sampled indicated parenthetically for each taxon. 50 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) TABLE 3. Tendencies in Gross Leaf Characteristics in Isoetes spp.’ (Expressed as a Percentage]. Leaf habit % Leaf texture %? Leaf color %$ Slight- Soft Stiff Yel- lyre- Re- Total and and _é WNo. low- Dark Red- No. Erect curved curved pops. slender stout pops. green green green pops. I. macrospora 39 27 34 106 35 65 26 6 91 3 33 I. tuckermanii 63 19 18 ae 8 91 9 56 7a 14 14 14 I. riparia 100 87 95 5 42 79 eat 14 I. acadiensis 25 25 50 a2 100 12 75 25 4 I.echinospora 73 14 14 = 103 e447 30 43.20 29 8 I. eatonii 100 40 92 9 26 100 10 I. engelmannii 98 2 41 80 20 15 100 7 1 Isoetes hieroglyphica had too few specimens for reliable data. ? Includes only those plants that do not fall into “typical” texture for genus. * Includes only those plants that were not “typical” bright green. Isoetes eatonii generally had erect leaves with a soft and lax leaf texture, yet a few specimens were erect with very stout leaves. Isoetes tuckermanii exhibited erect, fine leaves, but many populations also had leaves either slightly or severely recurved. Isoetes riparia had erect or straight leaves that often appeared to have been trained by the water current. Isoetes macrospora and I. acadiensis had a mixture of all three leaf types but exhibited the recurved habit half of the time. While I. acadiensis leaves were primarily fine-textured, those of I. macrospora were most often stout and rigid. The com state for Isoetes echinospora leaves was erect, but leaves were occasionally slightly recurved or reflexed. There was a tendency for the leaves to be fine. Isoetes engelmannii leaves were also gen- erally erect and usually somewhat lax. Qualitative characters such as leaf habit and texture are often quite arbitrary and do not easily conform to classes. An effort was made to score leaf form as a combination of leaf habit and leaf texture. Table 3 is not intended to show the actual numerical representation of leaf form for each taxon, but instead, to show prevailing trends in each taxon. “Leaf texture” (Table 3) includes only those populations in each species that did not fall into the “typical” texture state, and — is a score of those plants with leaf characters tending either toward fine and soft or thick and stout. These atypical populations included about one third of those | scored Of approximately 600 populations scored for leaf color, only 90 were atypical and fell into categories of yellow-green, dark green and red-green. The typical color was bright green. Isoetes eatonii was yellow-green or bright green, I. tuck- | ermanii mostly yellow-green or green with some tending toward reddish-green. Isoetes riparia ranged from a high number of yellow-green and bright green populations to some that were dark green. Isoetes acadiensis and I. macrospord — both tended towards the dark green leaf color, occasionally being reddish-green. — The color of I. echinospora and I. engelmannii leaves were typically bright — green, but both tended towards yellow-green (Table 3). ‘ PIB No RIS N ma eccte p gala me seco KOTT & BRITTON: ISOETES 51 Fic. 3. Coloration of sporangial walls. Bar = 50 um. a, Pale brown scattered cells of Isoetes riparia sporangium. b, Groups of brown cells give a patchy appearance to I. riparia sporangium. Pfeiffer (1922) and earlier workers refer to the presence or absence of bast bundles (fibers) in the peripheral tissues of the leaves. These tissues probably act as strengthening tissue and occur in plants as a result of specific ecological conditions. Pfeiffer found that all taxa in northeastern North America usually lacked bast bundles, except occasionally in emergent specimens of I. eatonii and I. engelmannii. All the other taxa are aquatic and the formation of bast bundles in the leaves never occurs. Because this feature is uncommon in the taxa under study, it has no taxonomic value. However this may be a useful feature in ter- Testrial and amphibious species. Of 62 taxa that Pfeiffer reviewed, all terrestrial species had peripheral bast bundles while those truly aquatic species mostly had none. Amphibious species, as expected, had few, weak bundles. Eaton (1900) and Clute (1905) also agree that the amount of bast depended on the exposure of the plant and that this character should not have very much importance in aquatic species. Stomata are probably formed in developing leaves that are exposed to air. In Most species studied here this does not occur regularly, because during early leaf development in spring, water is at a high level and plants are well covered with water, although they may become emergent later. This feature, like the bast 52 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) TaBLE 4. Characters of the Sporangial Region in Isoetes spp. Sporangium Velum Ligule Length Width Length Length (mm) (mm) Color of mature sporangia (mm) (mm) I. macrospora 3-5 3-4 clear to brown-spotted 1-1.5 to 2 I. hieroglyphica 5 3 usually clear 1 — I. tuckermanii 3-5 3 clear to tan i to 2 I. riparia 3-7 2-4 clear to very pale tan to 2 to 3 groups or cells in (gen. less) streaks I. acadiensis 3-5 2.5-3 clear to single dark or 1-1.5 to 3 pale tan cells scat- tered over sporangi- um I. echinospora 4-10 3 clear, or patches of tan 1-2.5 to 2.5 cells, or evenly tan I. eatonii 6-12 3.4-5 clear or tan 1-2 to 3.5 I. engelmannii 3-10 2-4 clear 1-3 to 3.5 bundles, appears to be correlated with growing conditions. Species such as he macrospora that grow in deeper water never have stomata. Those growing near- er to the shoreline occasionally show the presence of some stomata on leaves, as in I. riparia and I. tuckermannii, but in these it is not a constant feature (Eaton, 1900; Pfeiffer, 1922). The length of membranous leaf margins that extend some distance up from the base of the leaf tends to be related to the length of the leaves and therefore indirectly to the vigor of the plants. This feature showed a great degree of vari- ability from plant to plant within a given taxon. The maximum distance that this tissue reached on the leaves of most taxa was between 18 and 23 cm from the leaf base; however in most cases this distance was much shorter. Characteristics of the Sporangial Region.—Sporangia vary in size and color depending on maturity or growing conditions. Young sporangia are unpigmented and appear transparent, while mature sporangia may or may not have some cells that have acquired a brown pigment. The number of cells affected by the pig- ment, their distribution, and intensity of the pigmentation determined whether the sporangium was scored as clear (no pigment), spotted (individual cells that are brown) (Fig. 3a), patchy (groups of brown cells in patches) (Fig. 3b), or ta? (all cells slightly colored). Table 4 shows the sporangium coloration for all species, but data for I. hieroglyphica are based on relatively few specimens. Matthews and Murdy (1969) found sporangial color patterns extremely vari- able in two southeastern species. This character tends to be constant for most specimens within species of the northeast, although occasionally there are strik- ing exceptions to the otherwise typical patterns. Isoetes eatonii, I. engelmannii, — I. tuckermanii, and I. macrospora had sporangia that were usually unspotted — but a few specimens of I. tuckermanii had light brown or tan sporangia, and otherwise typical specimens of I. I times had sporangia that were : ici eciiataiaie i , l KOTT & BRITTON: ISOETES 53 a. I. ECHINOSPORA b. I. EATONII c. I. RIPARIA d. 1. TUCKERMANIT gi f. I. ACADIENSIS —— e. I. MACROSPORA OS4G Fic. 4. Representative ligules of Isoetes species. brown-spotted. Isoetes echinospora and I. riparia typically had colored cells in sag sparsely or closely placed, which gave the sporangia a coarsely Tabl aspect. Isoetes acadiensis usually had a tan or spotted sporangium wall. Phen 4 shows data collected on sporangium size for each taxon. Although I. the ah I. engelmannii, and I. echinospora appear to have the largest sporangia, ast generally had sporangia much smaller than 10 mm in length. Because Isoetes eatonii and I. engelmanni had large sporangia of generally the same size, 54 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) and the remaining taxa had smaller sporangia of comparable size, the character of sporangium size does not serve to differentiate individual taxa. e velum is a thin layer of tissue that covers the upper regions of the spo- rangium on its adaxial side. This covering may be complete, almost complete, partial, or very short as in the species studied here. Velum lengths, measured from the base of the ligule downward towards the center of the sporangium, varied from less than 1 mm to 3 mm, but were generally 1.0-1.5 mm long. Therefore in these taxa this feature appears to be of no diagnostic value. Pfeiffer (1922) used a relative measure to determine velum length by relating it to spo- rangium length. Palmer (1896) found that vela varied from % to % of sporangium length on different leaves of one plant of I. riparia. Because the sporangia varied a great deal in size due to age of the leaf, while the vela generally retained a constant length within a given plant, this mode of measuring the velum length was considered unacceptable here. Ligules—minute, fine flaps of tissue lying just above the sporangium on the leaf—were considerably damaged on dried specimens. Even ligules from living plants, which yielded the data for Table 4 and drawings for Figure 4, were s0 ephemeral that it was not certain whether the whole ligule was always observed. Although exact characterization of ligule shape and size is difficult, the ligules of Isoetes echinospora (Fig. 4a) and I. macrospora (Fig. 4e) were generally del- toid, being as long as wide; those of I. riparia (Fig. 4c) and I. tuckermanii (Fig. 4d) appeared to be triangular but slightly elongate, while ligules of I. eatonii (Fig. 4b) and I. acadiensis (Fig. 4f) usually appeared considerably elongate. Lig- ules of I. eatonii generally attained a larger size than the ligules of other species. CONCLUSIONS Most vegetative features of Isoetes are highly plastic and strongly reflect the growing conditions of the plants. This is also true for many other aquatics. Leaf length and leaf number have been statistically shown to have great variability within species and therefore are not diagnostic. Similarly the presence of stomata and bast bundles is a reflection of exposure to air during early leaf development. Perhaps leaf curvature and coloration may also be a product of environmental influences although this is difficult to prove. Nevertheless, these features show such variability within species that they cannot be of taxonomic or even diag- nostic value. Only sporangial coloration seems to generally correlate with other species characteristics (Kott & Britton, 1983). Although sporangium size appears to be somewhat constant within species, the size is often so similar among the taxa that it is not a valuable diagnostic aid. Both the velum and ligule do not yield further distinguishing information about these species. It therefore must be concluded that vegetative features in Isoetes species of northeastern North America are either so constant (such as corm lobing and : nature of roots) or so variable as to be of no taxonomic value. Species identifi- cation must rely primarily on spore characteristics (ornamentation and size), which agree well with other species specific characteristics such as chromosome _ KOTT & BRITTON: ISOETES 55 number, chromatographic spot patterns, ecological preferences, and behavior of the species. LITERATURE CITED Cute, W. 1905. What constitutes a species in the genus Isoetes? Fern pe 13:41-47. Eaton, A. 1900. The genus Isoetes in New England. Fernwort Papers 2:1-16. ——. 1908. Isoetaceae. Pp. 58-61 in Gray’s new manual of botany, ed. New York: American Cc Boo : FERNALD, M. L. 1950. Gray’s manual of botany, 8th ed. New York: American Book Company. Kort, L., and D. M. Brirron. 1980. mosome numbers for Isoetes in northeastern North Amer- ica. Canad. J. Bot. 58:980-984. and . 1982a. Comparison of cl tographic spot patt f North Amer- ican Isoetes ae gore Fern J. 72:15-18. and —————. 1982b. A comparative study of spore germination of some Isoetes species of northeastern is America. Canad. J. Bot. 60:1679-1687. and —————. 1983. Spore morphology and taxonomy of Isoetes in northeastern North 63. Mattuews, J. F., and W. H. Murpy. 1969. A study of Isoetes common to the ge outcrops of the southeastern Piedmont, United States. Bot. Gaz. (Crawfordsville) 130: PALMER, T. C. sy Notes on Isoetes riparia and Isoetes saccharata. Bot. Gaz. Ceaunioan 21: 218-22 PAOLILLO, D. J., “4 1963. The apse anatomy of Isoetes. Illinois Biol. Monog. 31:1-130. PETERSON, R. L., M. Scott, and L. Kort . Root cap structure in Isoetes macrospora Dur. Ann. Bot. (London), n.s. | 44:739-744 PFEIFFER, N. 1922. Monograph on the Isoetaceae. Ann. Missouri Bot. Gard. 9:79-232. American Fern Journal 75(2):56-70 (1985) New Species of Thelypteris from Puerto Rico GEORGE R. PROCTOR Department of Natural Resources, P.O. Box 5887, Puerta de Tierra, PR 00906 During the course of field work in Puerto Rico since July 1983, many new fern records have been discovered, some of them representing entities new to science. Preliminary to publication of a book about Puerto Rican ferns, a series of short papers has been initiated in order to describe new taxa and discuss other relevant matters. A previous paper (Proctor, 1984) described a new species of Trichomanes; the present one concerns the genus Thelypteris. All the newly discovered Puerto Rican thelypterids belong to subgenera Amauropelta and Goniopteris. In considering Amauropelta, the sectional assign- ments follow those of Smith (1974). There is no satisfactory modern classification of Goniopteris; therefore the species of this group are not assigned to sections in the present paper. In citing my own collections other than types, the first set of specimens is deposited in the herbarium of the Department of Natural Re- sources, San Juan, Puerto Rico (SJ); the second and third sets are deposited at the United States National Herbarium (US) and the Institute of Jamaica (IJ), respectively. More than 100 species of Thelypteris occur in the Antilles, and nearly half of these belong to subgenus Amauropelta. The importance of these islands as cen- ters of speciation for the latter taxon is emphasized by the recent discovery of three new Puerto Rican species, now being described. Two of these belong in section Amauropelta, and thus are allied to the species bearing sessile, reddish- resinous glands characterized many years ago by Morton (1963). SUBG. AMAUROPELTA SECT. AMAUROPELTA Thelypteris namaphila Proctor, sp. nov. (Fig. 1)—Type: Puerto Rico, Municipio de San German, Maricao State Forest, just S of Road 120 at approx. km 16.5, ca. 820 m, along banks of a small brook, 23 Nov 1983, Proctor 39834 (holotype US; isotype IJ, SJ). A T. piedrensi (C. Chr.) Morton squamis rhizomatis dense setulosis, planta multo minore et graciliore in omnibus partibus, lamina plene 2-pinnata cum numerosis pinnulis liberis nec adnatis nec decurrentibus, pinnulis auriculatis et _ inaequilateralibus et indusiis glandulis minutis albocapitatis exornatis differt. Rhizome erect, slender (0.3-0.4 cm diam.), tightly invested with stipe bases: scales lustrous brown, narrowly deltate-acuminate, up to 4mm long and 0.8-12 mm broad near base, minutely but abundantly setulose. Fronds erect, up to 49 cm long, of stiff texture; stipes 1.5-7 cm long, 0.5-1 mm diam., bearing near base a few spreading scales like those of rhizome, and minutely setulose on adaxial side and groove only, otherwise glabrous. Rhachis rather densely setulose on all , sides throughout. Blades 2-pinnate, narrowly elliptic, up to 39 cm long, (4-) 7- | | | | | | 4 : aa sh ht acta a ft ae ee OM re eo PROCTOR: THELYPTERIS 57 Dapartamenta Se Recussas Natareies arto FLORA DE PUERTO RICO papery a ace ; 2 3 a3 | elev. c. 820 m. On moist banks of anall streee in montane woodland: rhizome erect; - fronds of stiff texture. eas George #. Praeter. Ne eesM Fecna Nev. 23, 1963 Fic. 1. Thelypteris namaphila, holotype, Proctor 39834. — ceed middle, gradually narrowed toward base with 7-8 pairs of Mintle 3: creasingly widely separated pinnae (the lowest mere auricles), acu- at the sli — Pinnae sessile, the longest linear-oblong, truncate or nearly so slightly inequilateral base, acuminate at apex, 0.5-1 cm broad, fully pin- 58 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985} nate with up to 8 pairs of non-decurrent pinnules below the deeply pinnatifid apex; pinnules oblong-inequilateral, usually auriculate at base on acroscopic side, the margins otherwise entire to lightly crenulate, mostly up to 5 mm long and 1-1.5 mm wide, with 4-6 pairs of usually simple, very oblique veins, these strongly raised on both sides. Vascular parts and tissue minutely but harshly strigillose on adaxial side, the costae setulose beneath, the abaxial tissue glabrous and bearing scattered sessile reddish-resinous glands. Sori submarginal, con- fined to apical half of pinnules and segments; indusium small but persistent, ciliate with stiff acicular hairs, and also bearing a few minute whitish-capitate glands; sporangia glabrous. It appears that Thelypteris namaphila was one of the elements originally in- cluded in Aspidium sanctum var. portoricense Kuhn (Dryopteris sancta var. portoricensis (Kuhn) C. Chr., Smithsonian Misc. Collect. 52:380. 1909). Of two i types at US one (Sintenis 403) is a mixture of T. sancta (L.) Ching and T. namaphila. The other isosyntype (Sintenis 5956), representing an isolectotype of var. portoricense, is entirely T. sancta. In spite of its superficial similarity to T. sancta, T. namaphila is more nearly related to T. piedrensis, but differs from the latter in its more densely setulose rhizome scales, in being smaller and more slender in all its parts, in having fully 2-pinnate blades with many free pinnules neither adnate nor decurrent, with pinnules auriculate and inequilateral, and in its indusium with minute whitish- capitate glands (vs. indusium with much larger, sessile, globular, reddish-resi- nous glands). As its name implies, T. namaphila typically grows in wet soil beside brooks and streams, while T. piedrensis grows on stony slopes or shaded banks in montane forest. Thelypteris namaphila appears to be confined to an area of serpentine substrate, but T. piedrensis is not thus limited in distribution. The delicate cutting of the blades of T. namaphila somewhat resembles that of T. sancta (for which it has been mistaken), but this new species clearly differs from T. sancta in its narrow, setulose rhizome-scales, its stiff, harsh texture, and in its submarginal sori with more prominent indusium. Paratypes: Puerto Rico. Municipio de Maricao: Monte Allegrillo, 900 m, 3 Apr 1913, N. L. Britton et al. 2612 (US); Maricao State Forest, 700 m, 9 July 1963, Liogier 9823 (US); Maricao State Forest, along Rio Maricao above Criadero de Peces, 450-600 m, 18 Apr 1984, Proctor 40468. Thelypteris rheophyta Proctor, sp. nov. (Fig. 2}—Type: Puerto Rico, Municipio de Ponce, Barrio Anén, along Rio Inabén toward base of high falls, 500- 700 m, among boulders beside stream, 21 Jan 1984, Proctor 40042 (ho- lotype US; isotypes IJ, SJ). A T. consanguinea (C. Chr.) Proctor squamis rhizomatis fere deliquis, stipitis squamis late ovato-bullatis et cellulatis isodiametris tenuiter clathratis, pinnis paucioribus longius distantibus, pinnis maximis basi pinnulis duabus elongatis instructis, venis pluribus differt. Rhizome erect or sometimes decumbent, 0.3-0.5 cm diam., closely invested with stipe bases, and nearly devoid of scales (those produced attached only to stipe bases). Fronds fasciculate, up to 45 cm long (often less than 25 cm); stipes ia a a aca a PROCTOR: THELYPTERIS 54 Thelypsteris rheophrta Procter, sp. nov. enn rey compra CUTTRAL: Berrie be te Oml 234 5 yor Remy iia eg desid phineme erect. Ano Col. George ®. Proctor, Nc SOOM? Pune Jam, 22, 1984 Fic. 2. Thelypteris rheophyta, holotype, Proctor 40042. black at base, otherwise greenish-stramineous, 3-9.5 cm long, glabrous or nearly So, at first bearing toward base a few scattered, deciduous, broadly ovate-bullate, glabrous, delicately clathrate scales, the cells approximately isodiametric. Rhachis like the stipe, except for a line of minute incurved puberulence along adaxial 60 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) side. Blades oblanceolate, up to 33 cm long, 5-12 cm broad above the middle, gradually narrowed toward base (the lowest pinnae usually 0.5-1 cm long), long- acuminate at apex. Minute aerophores present at abaxial basiscopic base of pinnae. Pinnae usually 15-18 pairs, opposite, subopposite or alternate, inequi- lateral, usually ascending and somewhat falcate, with both proximal divisions (pinnules) longer than the rest, the basiscopic one usually at an angle divergent from the others on that side; longest pinnae typically 1.5 cm apart, narrrowly deltate-acuminate, mostly 2.5-6 cm long, 1-1.8 cm wide at base; segments ap- proximately 7-12 on a side, oblong-falcate, acute at apex, mostly 1.2-1.5 mm wide, all but the free proximal pinnules narrowly joined at base with up to about 6 pairs of simple veins (to 7 in proximal pinnules, with basal veins in these usually forked). Costae hispidulous on adaxial side, the costules and tissue gla- brous or with a few scattered hairs; all parts glabrous beneath, the tissue bearing few to rather numerous sessile, reddish-resinous glands. Sori supramedial to submarginal; indusium often asymmetric with one side slightly prolonged down- ward along vein, glabrous or sparsely hispidulous, the irregular margin usually beset with globular, reddish-resinous glands; sporangia glabrous. Thelypteris rheophyta, as its name implies, shares with T. consanguinea and other related forms (including the recently described T. fluminalis A. R. Smith of Ecuador) the habit of growing along the rocky courses of streams, where the plants frequently become submerged in flowing water during periods of flooding. The group as a whole needs further taxonomic study. The Puerto Rican material now being described has usually been misidentified as T. sancta, but it is actually more nearly related to T. consanguinea. I first thought these plants might represent a form of T. consanguinea, but further study has revealed an apparently consistent suite of characters distin- guishing the Puerto Rican populations. Thelypteris rheophyta differs from T. consanguinec in the following characters: 1) near absence of scales on the rhi- zome, those at base of stipe being broadly ovate-bullate and delicately clathrate with nearly isodiametric cells, vs. obvious presence of an apical tuft of rhizome scales, these mostly deltate-lanceolate, flat, and with the clathrate cells consis tently much larger, and longer than broad; in addition, the latter scales have resinous-glandular margins; 2) longer stipes (3-9.5 cm vs. 2-4 cm); 3) fewer pil nae (15-18 vs. 18-35) spaced farther apart (largest ones typically 1-1.5 cm apart vs. 0.6-0.8 cm); 4) presence of a pair of free pinnules at base of largest pinnae {as well as on reduced pinnae toward base), the basiscopic one often longer tha? the acroscopic, and diverging at an angle from the other (and much shorter) divisions of the same side; in T. consanguinea, the proximal divisions of the longest pinnae are not free, and the basiscopic one is short and not diverget! from the other segments on the same side (only on reduced pinnae toward base of blade are they longer and divergent); and 5) usually more numerous veins (e. 7 vs. 3-5 pairs). a ite oan wastes is a plant of more delicate texture and cutting than i a rat eC iia linear-oblong segments are usually more widely sepa : as ore _their own width) by broad sinuses, and an ind ge cture is absent or rudimentary. The scales at base of stipes are, however PROCTOR: THELYPTERIS 61 very similar. Thelypteris rheophyta occurs only along stream courses, but T. sancta is much more widely distributed. From T. resinifera (Desv.) Proctor, which often grows in similar habitats, T. rheophyta differs in 1) the delicate, nearly isodiametric clathrate cell structure of the bullate-ovate scales near base of stipe (vs. elongate subclathrate cells of flat lanceolate scales, more like those of T. consanguinea but coarser); 2) fewer pinnae (15-18 pairs, vs. usually 30-40 pairs, seldom less); 3) presence of free proximal pinnules (vs. all divisions of the pinnae adnate); and 4) lamina glabrous (or nearly so) abaxially (vs. lamina pilosulous on vascular parts beneath). Paratypes: PUERTO Rico. Without definite locality, May 1883, Eggers 733 (US). Municipio de Rio Grande: Sierra de Luquillo, Caribbean National Forest: Slopes of El] Yunque, 28 May 1944, W. H. Wagner s.n. (US); Road 191, km 12.3, ca. 680 m, 8 Aug 1983, Proctor 39373; Road 191, km 9.9 at Quebrada Juan Diego, 480-500 m, 10 Oct 1983, Proctor 39603; Quebrada Sonadora above crossing of Road 186, 300-450 m, 21 Feb 1984, Proctor 40172. Municipio de Arecibo: Barrio Esperanza, vicinity of Observatorio de Arecibo, beside Rio Tanama, 160-180 m, 26 Feb 1984, Proctor 40250, 40256. SUBG. AMAUROPELTA SECT. UNCINELLA Thelypteris inabonensis Proctor, sp. nov. (Fig. 3)—Type: Puerto Rico, Municipio de Ponce, Cordillera Central, Toro Negro State Forest, along headwa- ters of Rio Inab6én above high falls, 1120-1250 m, 22 Jan 1984, Proctor 40069 (holotype US; isotypes JJ, SJ). A T. rustica (Fée) Proctor magnitudine multo minore, squamis setulosis, pilis plerumque acicularibus in rachi et abaxialiter in costis differt. A T. frigida (Christ) Smith & Lell. et T. funckii (Mett.) Alston squamis setulosis in stipite et rachi, pilis abaxialiter in costa crassioribus et numerosioribus et indusio dense longiciliato differt. A T. frigida stipitibus brevioribus, rachi pilis longis crassis pluricellular- ibus carenti, superficie adaxiali densius strigillosa differt. A T. funckii costa squamis carenti differt. Rhizome erect, slender (ca. 0.5 cm diam.) and tightly invested with stipe bases, bearing toward apex numerous dark lustrous brown, narrowly lance-linear, densely setulose scales up to 4 mm long and 0.2-0.4 mm wide near base, the tips often contorted-filiform. Fronds erect-arching, up to 60 cm long; stipes 5-10 cm long, very densely hispidulous with grayish acicular (rarely hamate) hairs of variable length (mostly 0.2-0.4 mm long}, and also beset with numerous spread- ing scales like those of rhizome; rhachis likewise clothed with similar hairs and scales throughout. Blades narrowly elliptic, up to 55 cm long, mostly 9-11 cm broad near middle, gradually narrowed toward base with 8-10 pairs of reduced, slightly deflexed pinnae (the lowest often less than 1 cm long), long-acuminate at apex. Pinnae mostly 25-30 pairs, sessile, the longest linear-oblong, truncate at base, acutish to short-acuminate at apex, 1.3-1.7 cm wide, pinnatifid with mostly 12-14 pairs of segments; aerophores absent. Segments oblong, slightly falcate, up to 6 mm long and 2-3 mm wide, rounded to subacute at apex, with up to 7 Pairs of simple veins. Vascular parts and tissue densely strigillose on adaxial Side; costae and costules densely hispidulous abaxially but lacking scales, the 62 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) FLORA DE PUERTO Rico SepetaT ants Se Recursos Matures Sar Juan, Peete mee Monicfpie de Ponce is * Thelreteris inshonensis Procter, sp. nov. CORDILLERA CENTRAL: Tore Negro State Forest. Along headwaters of Rfo Inabén abere bigh falls. SS — elev. 1120-1250 =. er eS es Meist eroding earth bank of stress in partial shat Fors Jan. 22, 1984 Fic. 3. Thelypteris inabonensis, holotype, Proctor 40069. tissue with numerous short, erect, acicular (rarely hamate) hairs and ee glands. Sori small, supramedial to submarginal, often somewhat concealed the narrowly reflexed margins, with persistent, densely long-ciliate indusiu™ sporangia glabrous. PROCTOR: THELYPTERIS 63 This species seems related only to T. rustica in the West Indian area, but clearly differs from that Lesser Antillean species in its much smaller size, setu- lose scales, and in having mostly acicular (rather than hamate) hairs on rhachis and abaxially on costae. It more closely resembles T. frigida and T. funckii of Costa Rica and northwestern South America. From both of these it differs in its setulose scales, its stouter and more abundant abaxial costal hairs, and in its densely long-ciliate indusium. From T. frigida it further differs in its shorter stipes, in the intermediate and variable nature of the rhachis hairs (but none long, stout, and obviously pluricellular), and in being much more densely strigil- ose on the adaxial surface. From T. funckii it also differs in wholly lacking scales on the costae beneath. Thelypteris inabonensis so far is known to occur only along the banks of a single stream, where it is uncommon. It should be considered a rare and endan- gered species. SuBG. GONIOPTERIS Thelypteris cordata (Fée) Proctor var. imitata (C. Chr.) Proctor, comb. et stat. nov.—Dryopteris imitata C. Chr., Kongl. Svenska Vetenskapsakad. Hand. III, 16(2):29, t. 6, figs. 1-4. 1937.—Type: Haiti, Ekman H 3724, isotype IJ. Range: Andros, Hispaniola, and Puerto Rico. The small group of ferns allied to T. cordata presents a somewhat intricate problem in defining taxa. Christensen (1937) described Dryopteris imitata as “Near D. cordata (Fée) C. Chr. with similar pinnae but it is pinnate nearly to the apex, not proliferous, and with very conspicuous, reniform, grey indusia make it very distinct.” However, sul quent ] of more material from a wider range tends to reduce the distinctiveness of these two forms. Populations in Cuba and Jamaica have numerous rather long, simple hairs on the rhachis (mixed with much shorter stellate ones) and a very minute indusium with simple and forked hairs; plants of the Bahamas (Andros), Hispaniola, and Puerto Rico bear only short stellate hairs on the rhachis and have a much larger indusium likewise covered with tiny stellate hairs only. Plants from all areas vary greatly in size (the Jamaican ones most consistently large), and there seem to be no other clear-cut morphological distinctions. Since neither typical T. cordata nor T. im- itata are ever proliferous, Christensen’s statement in this regard seems irrelevant xcept in relation to the allied T. reptans (J. F. Gmel.} Morton, which occurs throughout the same range. In my opinion, T. cordata and T. imitata should be considered only varietally distinct from each other. : Thelypteris cordata var. imitata is quite common on sheltered limestone cliffs in Puerto Rico, where its identity was first recognized by Weatherby on the basis of material collected by Chrysler in 1947 (US). Maxon (1947, p. 126) had previ- ously misidentified other material (Sargent 3273, US) as Dryopteris asplenioides (Sw.) Kuntze. Thelypteris abdita Proctor, sp. nov. (Fig. 4)—TyPe: Puerto Rico, Municipio de Utuado, Rio Abajo State Forest, ca. 1.6 km due WSW of Campamento Crozier, 320-340 m, in sheltered or hidden crevices of moist vertical limestone cliff, 24 Jan 1984, Proctor 40099 (holotype US; isotypes Tj, SJ). AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985} FLORA DE PUERTO RICO Monicfpio de Utuade The)yoteris abdita Proctor, sp. nev- RIO ARAJD STATE FOREST: ¢. 1.6 kom. due W.S.¥. ef Caspanente Crozi . ban er, elev. 320-340 = yp cmensstearemers omermancariyag —S ey, Ocm] 2 3 4 5 = mettres crevices of moist vertical limestone 40009 to Jam. 24, 1984 Fic. 4. Thelypteris abdita, holotype, Proctor 40099. A T. cordata var. imitata (C. Chr.) Proctor differt magnitudine multo minor / superficie stellato-puberula, venis simplicibus vel 1-furcatis, indusio glandulos? cum multis pilis simplicibus vel raro furcatis. | Rhizome erect, suberect, or decumbent, very small (0.2-0.4 cm diam.}, shot i PROCTOR: THELYPTERIS 65 closely invested with stipe bases, at apex bearing a tuft of glabrous or sparingly stellate-puberulous lustrous brown scales, these narrowly deltate, acuminate, up to 4 mm long and 0.3-0.8 mm broad at base. Fronds few (usually 4-7), laxly fasciculate and spreading or decumbent, 6-19 (-23) cm long, forming a loose rosette; stipes stramineous, 2-7 cm long, 0.3-0.6 mm diam., lightly puberulous chiefly on adaxial side with minute stellate (sometimes a few merely forked) hairs. Rhachis more densely stellate-puberulous than stipe, on all sides. Blades approximately linear (rarely narrowly lance-linear), 5-16 (-19) cm long, 1-2 (-3) cm broad, pinnate nearly throughout, slightly narrowed at base end and at the blunt, shortly pinnatifid apex. Pinnae 10-20 pairs, mostly alternate, apart or sometimes approximate, stalked, rounded-oblong (rarely almost rotund) or very bluntly deltate-oblong, 0.5-1 (1.5) cm long, 0.3-0.6 cm wide, the margins entire, with 3-6 pairs of veins, these all simple or the basal 1-3 pairs forked, all tips free and ending in minute elliptic hydathodes short of the margin (seen adaxi- ally). Vascular parts, margins, and tissue on both sides puberulous with few to numerous minute stellate hairs. Sori supramedial; indusium erect at sporangial maturity, irregularly roundish, with glandular margin and bearing numerous long, simple (rarely forked) hairs; sporangia glabrous. This diminutive species has usually been mistaken in the past for a small form of Thelypteris cordata (Maxon, 1947, p. 127), and in fact grows with or near T. cordata var. imitata in many localities. However, many observations in the field have clearly shown that T. abdita is an independent, uniformly distinct, and much smaller entity, and bears little resemblance to T. cordata var. cordata as I have long known it in Jamaica. Comparison of T. abdita with T. cordata var. imitata shows that besides its much smaller size (mature fronds mostly 6-19 cm long and 1-2 cm wide, seldom larger, vs. up to 50 cm long and up to 5.5 cm or more wide), T. abdita differs in its densely to sparsely stellate-puberulous (vs. glabrous) tissue, simple or at most 1-forked veins, these always free (vs. many 1-2-forked veins, some often anastomosing), and moderate-sized, glandular in- dusium with many long, simple or rarely forked hairs (vs. a more prominent, non-glandular indusium with many much smaller, always stellate hairs). T. ab- dita is endemic to Puerto Rico; small forms of T. cordata var. cordata and T. cordata var. imitata seen from Cuba and Hispaniola respectively, though rather Similar in appearance, are apparently juvenile developmental stages or else environmentally stunted, and are always distinct on the basis of one or more of the characters stated above. I am satisfied that T. abdita as described above represents the mature stage of this species. Paratypes: Puerto Rico. Municipio de Utuado: Utuado, 25 Mar 1887, Sintenis 6588 (US); Utuado, 350 m, 12 Nov 1943, Sargent 3272 (US). Municipio de Lares: Road 129 [old route}, km 20, 7 May 1966, Woodbury s.n. (SJ). Municipio de Barrio Charcas, near end of Road 437 (Finca Laboy}, ca. 220 m, 15 Oct 1983, Proctor 39625. Municipio de Arecibo: Barrio Esperanza, gorge of Rio Tanamé in vicinity of Observatorio de Arecibo, 200-270 m, 26 Feb 1984, Proctor 40241. Thelypteris verecunda Proctor, sp. nov. (Fig. 5)—TyPe: Puerto Rico, Municipio de Quebradillas, Barrio Charcas, near end of Road 437 (Finca Laboy}, ca. 220 m, on moist shaded limestone ledge near base of cliff, 5 Oct 1983, Proctor 39581 (holotype US). AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) . b @ee © © o6 $@ @. *“@ %~e ee S £9 8Oe Be | *s*e©e08 66 s° f FLORA DE PUERTO Rico F See naan Rte, Sn Recuenes Mataraine itt - 5 Muniefpic Quebradillas i ort hh. ; Thelypteris yerecunda Proctor, sp. nov. MONTANAS GUARIONEX: Barrio Charceas. Near end of i Read 437, elev. c. 220 =, ' | Omi eS 4 y Om moint shaded limestone ledge near base of cliff. 4 Fic. 5. Thelypteris verecunda, holotype, Proctor 39581. A T. reptanti (J. F. Gmel.) Morton magnitudine minore, venis liberis plerum — que simplicibus, indusii pilis longis simplicibus differt. T. abditae Proctor mag | nitudine et textura similis, sed ab illa frondibus dimorphis ubique pilis stellatis — et simplicibus indutis, soris inframedialibus differt. : Rhizome creeping or decumbent, slender (ca. 0.2-0.3 cm diam.)}, bearing 4 PROCTOR: THELYPTERIS 67 apex a small tuft of brown, minutely stellate-puberulous, deltate-acuminate scales ca. 1 mm long and 0.5 mm wide at base. Fronds decumbent-spreading, dimor- phic, clothed throughout with minute stellate (occasionally forked) and numerous much longer simple hairs, the latter mostly 0.3-0.6 mm long; stipes 1-1.5 (-2) cm long, 0.4-0.5 mm diam. Sterile blades oblong, 2.5-4 cm long, 1.5-2 cm broad, truncate at base, rounded at the broadly lobed apex, with 2-4 pairs of short- stalked, round-oblong, entire pinnae, these 0.8-1 cm long and 0.4-0.6 cm wide; veins simple or 1-forked, all free. Fertile blades linear-attenuate, 13-15 cm (or more) long, 1.2-1.8 cm broad, truncate at base, the rhachis bearing a minute proliferous bud well below the apex; pinnae 15-20 pairs (or more), somewhat variable, mostly rounded-oblong to oval, 0.5-0.9 cm long. 0.3-0.4 cm wide, entire, short-stalked (or rarely a few sessile and somewhat adnate). Sori small, infra- medial; indusium minute, erect, bearing a tuft of long, white simple hairs; spo- rangia glabrous. The dimorphic fronds (the elongate ones proliferous), indument, and the in- framedial position of the sori suggest comparison with T. reptans, but the small size, entirely free and mostly simple veins, and the long, simple indusial hairs indicate closer affinity with T. abdita. Perhaps it is a hybrid of these two species, both of which occur in the same general area. However, it is clearly distinct from both, and moreover appears to have normal, fertile sporangia. Thelypteris verecunda so far is known only from the type collection. Thelypteris hildae Proctor, sp. nov. (Fig. 6)—TyPE: Puerto Rico, Municipio de Utuado, Rio Abajo State Forest, ca. 1.6 km due WSW of Campamento Crozier, 320-340 m, in crevices of moist shaded limestone cliffs, 24 Jan 1984, Proctor 40100 (holotype US; isotypes IJ, SJ). A T. guadalupensi (Wikstr.) Proctor habito cremnophilo, frondibus monomor- phis stipitibus longioribus laminis basi truncatis, superficie glabra differt. Plants always growing in crevices of shaded calcareous or non-calcareous cliffs or rock ledges. Rhizome short, erect or ascending, mostly 0.4-0.6 cm diam. (excluding the closely-investing stipe bases), at apex bearing a tuft of lustrous brown, glabrate or sparingly and minutely stellate-puberulous scales, these chief- ly deltate-attenuate, up to 6 mm long and mostly 0.5-1 mm broad near base. Fronds few (usually 4-6), monomorphic, loosely fasciculate, spreading, up to 50 cm long (usually much less); stipes stramineous, (4-) 9-18 (-22) cm long (up to 45% of length of entire frond), 0.6-0.9 (-1.2) mm diam., minutely stellate-puberu- lous or sometimes glabrous. Blades narrowly lance-deltate or occasionally lan- ceolate, 12-28 (-32) cm long, 2.5-4.5 (-6.5) cm broad near base or below middle, truncate or very slightly narrowed at base, acuminate at apex, pinnatifid through- out (or sometimes a single pair of free pinnae, rarely more, at base) except toward the subentire apex. Segments 15-22 (-25) pairs, cut 1-2/5 (-%) to the rhachis with mostly acute sinuses, oblong to narrowly deltate, blunt to acute at apex, with 8-17 pairs of veins, the majority of these (except in very large plants) simple and free (or a few 1-forked, these rarely anastomosing}, the proximal pair from adjacent segments usually joined in the tissue and sending an excurrent veinlet 68 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985} Thelvypteris hildae Procter, sp. nov, —— RIO ARAIO ovary t TREST: ¢. 1.6 km. due W.S.¥. of (Om) 234 5 | Cenpanente Crevier, wlan, 320-340 = eis isang tn crevices c ist shaded limestone cliffs. 40100 1984 Fic. 6. Thelypteris hildae, holotype, Proctor 40100. to the sinus. Adaxial side of blade glabrous except for minute stellate hairs along the rhachis-groove and sometimes a few on costae; abaxial side of blade wit! rhachis minutely stellate-puberulous to glabrous, and with scattered minute stel-_ late hairs on costae and veins, the tissue glabrous; margins likewise with a few PROCTOR: THELYPTERIS 69 stellate hairs especially at sinuses. Sori usually about medial, sometimes paired; moderately large indusium present, always minutely stellate-puberulous; spo- rangia glabrous. The group of T. guadalupensis in Puerto Rico involves two common and dis- tinct species usually lumped under one name. Typical T. guadalupensis (together with its somewhat enigmatic variant ‘forma portoricensis’’) is an essentially terrestrial species (sometimes also on mossy boulders, rarely on ledges), well known and widespread in the Greater and Lesser Antilles. It is common in Puerto Rico. Also in Puerto Rico occurs a related entity that is always found in the crevices of shaded cliffs and ledges, never in soil on the ground, and which is also morphologically distinctive. Although probably a close relative of T. gua- dalupensis, this entity, now described as T. hildae, differs consistently in 1) its habitat (always in crevices of ledges and cliffs, vs. nearly always terrestrial in soil or sometimes on mossy stones); 2) its monomorphic (vs. more or less dimor- phic) fronds, these never proliferous at the tip; 3) its much longer, more slender stipes; 4) its blades truncate or but slightly reduced at base with seldom more than 1 pair of separated pinnae (if any). In T. guadelupensis the blades are long- decrescent downward, usually having two to many pairs of separate pinnae or widely-separated segments. In addition, T. hildae has glabrous (vs. minutely stellate-puberulous) tissue, and less divided, mostly free veins. This newly-rec- ognized species is apparently endemic to Puerto Rico, where it has a wide dis- tribution in suitable habitats. Unlike many species found on limestone rocks, T. hildae has also been found on cliffs of igneous origin. Thelypteris hildae is named for Hilda Diaz Soltero, former Secretary of Nat- ural Resources, under whose regime the present study of Puerto Rican ferns was initiated, and whose enthusiastic encouragement I have much appreciated. aratypes: PurRTo Rico. Municipio de Utuado: Probably near type locality, 7 Dec 1943, Sargent 3298 (US); 22 Apr 1947, Chrysler 6728 (US); 3 May 1983, Acevedo 46 and 57 (SJ). Municipio de Rio Grande: Sierra de Luquillo, Caribbean National Forest, Quebrada Juan Diego above Road 191, km 9.9, 480-500 m, 10 Oct 1983, Proctor 39606. Municipio de Ponce: Cordillera Central, Barrio Anon, along Rio Inabén toward base of high falls, 500-700 m, 21 Jan 1984, Proctor 40066. Municipio de : Barrio Frontén, Road 140, km 49, ca. 320 m, 12 July 1984, Proctor 40556. Municipio de Arecibo: Barrio Esperanza, vicinity of Observatorio de Arecibo, 250-300 m, 26 Feb 1984, Proctor 40238. Municipio de Hatillo: Barrio Bayaney, Cueva Clara de Empalme, near intersection of Roads 134 and 455, 250-280 m, 25 Feb 1984, Proctor 40188; Cueva de la Catedral, near intersection of Roads 129 and 134, 200-250 m, 12 Nov 1983, Proctor 39716; same locality, 11 Aug 1984, Proctor 40676. Municipio de Quebradillas: Barrio Charcas, near end of Road 437 (Finca Laboy), ca. 220 m, 5 Oct 1983, Proctor 39562; gorge of Rio Guajataca, E side of river along bottom of gorge, ca. 130 m, 15 Oct 1983, Proctor 39610. Thelypteris hastata (Fée) Proctor var. heterodoxa Proctor, var. nov. —Type: Puerto Rico, Municipio de Hatillo, Barrio Bayaney, Cueva de la Catedral, near intersection of Roads 129 and 134, 220-250 m, on steep wooded slope in limestone ravine, 11 Aug 1984, Proctor 40678 (holotype US; isotype S]). A T. hastata var. hastata lamina deltata vel ovato-deltata, pinnis latioribus Marginibus lobatis differt. 70 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) Although identical in minute details, this entity differs from typical T. hastata in its deltate or ovate-deltate blades, the basal or next to basal pair of pinnae being the longest. This contrasts with var. hastata, whose blades are oblanceolate to narrowly oblong-obovate (rarely elliptic), being widest above the middle and normally having up to six gradually reduced pairs of pinnae toward base. Fur- ther, the pinnae of var. heterodoxa are usually 2.5 cm or more in width, with distinct marginal lobing, whereas those of var. hastata are seldom over 2 cm wide and are usually serrulate or crenate-serrate, or rarely shallowly lobed. The group of Thelypteris hastata in Puerto Rico includes T. hastata itself (common), its var. heterodoxa (rare), T. leptocladia (Fée) Proctor, and several apparent hybrids involving these and other species. These forms require further investigation. However, the only other goniopterid fern in Puerto Rico with blades of somewhat similar outline to T. hastata var. heterodoxa is T. abrupta (Desv.) Proctor (recently discovered in this island for the first time, 26 Aug 1984, Proctor 40762}, but the latter species clearly differs in details of indument, in the nar- rowly cuneate bases of the lower pinnae, in having a distinct, persistent indu- sium, and in the sporangial stalks (instead of the heads) being setulose. It is probably not closely related. ACKNOWLEDGMENTS eral persons have provided greatly appreciated assistance thus far in my Puerto Rican fem work. Among these I would especially like to thank Dr. Juan “Tito” Cordero and his friends Miguel “Papo” Vives and William Estremera, of Quebradillas; and Dr. Horst Haneke and his delightful family, of Ponce. Among departmental colleagues, the field assistance of Barbara Cintrén, Oscar Diaz, Rubén Padrén, Benito Pinto. and Richard Stud js} a. y ‘. DE tated. | thank Dr. Alain Liogier for assistance in preparing the Latin diagnoses, and Dr. José Vivaldi for reviewing the manuscript. LITERATURE CITED Proctor, G. R. 1984. A new filmy fern from Puerto Rico. Amer. Fern J. 74:7-8. SmiTH, A. R. 1974. A revised classification of Thelypteris subgenus Amauropelta. Amer. Fern }- 64:83-95. Se eee SHORTER NOTES Dryopteris ludoviciana and D. australis New to Arkansas.— Until recently, Arkansas was known to have three species and one hybrid of Dryopteris (Taylor & Demaree, Rhodora 81:503-548, 1979; Taylor, Arkansas ferns and fern allies, 1984), of which only D. marginalis (L.) Gray occurs frequently across the state. D. spinulosa (O. F. Muell.) Watt and D. x leedsii Wherry are each known from one locality, while D. celsa (Palmer) Small is known from three localities. Con- sequently, the discovery in 1984 of a species and a hybrid new to the state constituted significant additions to the fern flora of Arkansas and furnished im- portant phytogeographic data on the distribution of Dryopteris taxa in eastern North America (Carlson & Wagner, Contr. Univ. Mich. Herb. 15:141-162, 1982). Dryopteris ludoviciana (Kunze) Small, Louisiana Log Fern, was discovered for the first time in Arkansas at Warren Prairie State Natural Area (Bradley Co.), located 20 km west of Monticello (Peck & Peck 84641, LRU, MICH, MIL; Sundell & McIntyre 2864, UAM). The fern grows with Lorinseria areolata (L.) Pres in the moist, lowland Quercus phellos L. woods. This Coastal Plain locality is 350 km north of St. Mary Parish, Louisiana, and Hardin Co. and Tyler Co., Texas, sites of the nearest previously reported populations (Thomas, Wagner & Messler, Castanea 38:269-274, 1979; Correll & Correll, Aquatic and wetland plants of Southwestern United States, 1972). The Arkansas population is the most north- western in North America, and only the fourth population known to occur west of the Mississippi River. Dryopteris x australis (Wherry) Small, Southern Log Fern, the backcross hy- brid between D. celsa and D. ludoviciana, was discovered for the first time in Arkansas at a wooded acid seep located at the southern periphery of the Ouachi- ta Mts. in Garland Co., some 30 km southwest of Hot Springs (Orzell 1429, UARK; Peck 84680, LRU, MICH, MIL). The hybrid occurs with one parent, D. celsa, but not with D. ludoviciana. The closest known population of the latter species is at Warren Prairie, Bradley Co., Arkansas, 200 km to the southeast, making this hybrid population another example of “hybridization by remote control” (Wagner, Amer. Fern J. 33:71-73, 1943). The nearest hybrid populations occur in Louisiana, some 500 km to the southeast (Wagner & Musselman, Cas- tanea 47:182-190, 1982). The Arkansas population is the most northwestern of the eight known populations of this North American hybrid. It is the third dis- Covered and currently the only extant population west of the Mississippi River. We thank W. C. Taylor, Milwaukee Public Museum, and W. H. Wagner, University of Michigan, for verification of our identifications. This research was Sponsored, in part, by a faculty research grant from Office of Research and Sponsored Programs and the College of Science, Office of Research, Science and Technology, at University of Arkansas at Little Rock.—James H. Peck, De- partment of Biology, University of Arkansas at Little Rock, Little Rock, AR 72204; L. OrzELL, Arkansas Natural Heritage Commission, Little Rock, AR 72201; ERIC SUNDELL, Department of Natural Sciences, University of Arkansas-Monti- Cello, Monticello, AR 71655; and CAROL J. PECK, Arkansas State Plant Board, Little Rock, AR 72203. 72 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 2 (1985) _| Polystichum aleuticum from Adak Island, Alaska, a Second Locality for the — Species:—Field work conducted by E. Hultén and W. J. Eyerdam in 1932 along — the Aleutian Island chain resulted in many vascular plant collections reported — in the Flora of the Aleutian Islands (Hultén 1937; ed. 2, 1960). Polystichum — aleuticum C. Chr. in Hultén (Svensk Bot. Tidskr. 30:515, 1936) was collected by — Eyerdam from Atka Island in the Central Aleutians. Until now, Eyerdam’s col- — lection has represented the only known occurrence of the species (Wagner, Pter- 7 idologia 1:1-64, resus he certainly must rank among the most restricted and rarest _ ferns of North Am In 1975, while sillectig bryophytes on Adak Island, I discovered a substantial — population of this rare Aleutian fern. The specimens were examined and an- — notated by Wagner (ORE), who has indicated that they compare favorably with _ type materials. Specimens have been deposited in several herbaria: Alaska, — Aleutian Islands, Adak Island, Mt. Reed, rock ledges below summit, 51°49'N, — 176°44'W, 400 m, 19 Sept 1975, D. K. Smith UT-54678 (ALA, ORE, TENN). Christensen (in Hultén, Svensk Bot. Tidskr. 30:515-528, 1936; Amer. Fern J. — 28:110-113, 1938) suggested a strong relationship exists among P. aleuticum, Hi- — malayan P. prescottianum (Wall.) Moore, western and central Chinese species — P. shensiense Christ, P. sinense Christ, P. moupinense (Franch.) Bedd, and P. — lachnense (Hook.) Bedd. He noted, however, it is distinctive by “... its peculiar — erose-dentate, greenish flat indusia and . . . castaneous, thick stipes a older leaves — ” Christensen further noted that the only substantial difference between P. — Iocbugies and P. aleuticum was the entire scales (vs. dentate scales) of the latter. _ Careful examination of the scales of Adak materials reveals that the scale ma- _ gins are quite denticulate to shortly ciliate. Further studies are needed to clarify | the relationship of the Aleutian fern to its Himalayan-Chinese counterparts. __ Contributions from the Botanical Laboratory, University of Tennessee, n.s. No. — 998.—Davip K. SmiTH, Department of Botany, University of Tennessee, Knox: — ville, TN 37996-1100. BOOKS ON FERNS ee An extensive list of floras, monographs, and horticultural manuals is avail- : . > able for two first-class stamps. I also buy used fern books; please send list | — of authors, hei and dates. Myron Kimnach, 1600 Orlando Rd, San Mar- — ino, CA 911 AMERICAN mie FERN ste July-September 1985 JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY An Unusual Submerged Aquatic Ecotype of Asplenium unilaterale Masahiro Kato and Kunio Iwatsuki A Newly Discovered Habitat for Isoétes melanopoda in Louisiana _R. Dale Thomas A *, : Io. Be tc..k a . > Se * iT a! £ 7 wate, Be. : Terry W. Lucansky Germination and Morphology of Spores of Trichipteris corcovadensis Luciano M. Esteves, Gil M. Felippe, and Therezinha S. Selhem New Stati ¢. N: ae a 2. DB. Pes ae wietiueds bE y WOUEEYs New Locati: for Isoét teg tif in Georgi Phillip M. Rury — James R. Burkhalter _ The American Fern Society Council for 1985 TERRY R. WEBSTER, Biological Sciences Group, University of Connecticut, Storrs, CT 06268. President FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. Vice-President W. CARL TAYLOR, Milwaukee Public Museum, Milwaukee, WI 5323 Secretary JAMES D. Dept. of Botany, University of Tennessee, AST TN — Treasurer DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Burlington, VT 0540 R eco ae Tr reasurer ALAN R. SMITH, Dept. of Botany, University of California, ne CA 94720. = Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, DC 2: Editor DENNIS Wm. STEVENSON, Dept. of Biological Science, Barnard aye Columbia University, New York, NY 10027. Fiddlehead Forum Editor American Fern Journal EDITOR Pade ee re ee Dept. of Botany, spate: of California, - Berkeley, CA 94720 oe ASSOCIATE EDITORS GERALD J.GASTONY............. Dept. of Biology, ees eonessan Bloomington, IN 47401 URMIG OPTION FAUFLER .. 2 pt. of Botany, Unive rsity of Kansas, 1 ee wrence, KS 66045 — cathe B.LBLLINGER 3. ee, U.S. Nat'l Herbertom NHB-166, prema ai Institution, hington, DC 20560 TERRY R. WEBSTER ____ Biological Sciences Group, University of Sento Storrs, CT 06268 The “American Fern ieee (ISSN 0002-8444) is an illustrated quarterly devoted to the general — Herbarium, — seus hoe Bec, owned by the American Fern Society, and published at the _ University of Vermont, Retain, VT 05405. Second-class postage paid at Burlington, VT. = _ additional entry point. - a, oe £ i iit tick — sede or bac ues shold be dred t Dt James D. Montgomery, Ecology III, R.D. 1. ‘Berwick, PA 18603. Oh. g e : “Cone ng once sh be addressed to the S a ee m3 49 +1 fs aw of issue, : Bes addr due and npplicat fi bership should be sent tothe Records ret American Fern Journal 75(3):73-76 (1985) An Unusual Submerged Aquatic Ecotype of Asplenium unilaterale MASAHIRO KATO Botanical Gardens, Nikko, Faculty of Science, University of Tokyo, Nikko 321-14, Japan KUNIO IWATSUKI Botanical Gardens, Faculty of Science, University of Tokyo, Tokyo 112, Japan Although many ferns inhabit moist places, extremely few are actually deserv- ing of the term aquatic. There are the floating miniferns Azolla and Salvinia, and others that usually root in mud with a fluctuating water level, e.g., Ceratop- teris, Acrostichum, and Marsileaceae. Rheophytes, subject to floods in st bed are periodically aquatic. Some plants are moistened with constant spray from waterfalls (Giesenhagen, 1892; Mickel, 1972; Iwatsuki, 1975; Page, 1979). A species of Stenochlaena lives in water pockets formed by Pandanus leaf bases (Price, 1982, p. 200). This report of Asplenium unilaterale Lam. makes it the second known fern that lives as a genuine submerged aquatic. The other such is Bolbitis heudelotii (Fée) Alston of Africa, sometimes cultivated in tropical aquaria, although it is merely a form of the species that is most often rheophytic (Hennipman, 1977, p. 236-240). Asplenium unilaterale is already known to be highly variable in cy- tology (Lovis, 1973), reproduction (Murakami & Iwatsuki, 1983) and habitat and morphology (Iwatsuki, 1975), with some plants (var. udum) having an extremely thin (bistratose) lamina lacking intercellular spaces. The submerged form was collected in Seram (Ceram) Island, the Moluccas, Indonesia, during field trips in 1983 and 1984-85. On the island, ecologically and reproductively different forms of Asplenium unilaterale also are terrestrial or on dry or wet rocks in montane forests, as they are in the entire range. The new form grows in a spring named “Air Mata Makariki” at 350 m above sea level on the southern slope of Murkele Mountain Ridge at the central part of the island. is area is abundant in limestone rocks, and river water occasionally under- spring (Fig. 1). Smaller populations grow more shallowly submerged on roots 74 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) 5m Fic. 1. Diagram of the spring “Air Mata Makariki.” A, B: two largest water-flowing mouths. C: the deepest point (ca. 2 m deep). F: two trees of Ficus sp. M: Wae (=River) Makariki. Solid circle: individuals of the submerged aquatic ecotype of Aspleni ilaterale. A indicate water flow. from trees on the slope, or among rocks at the bottom of the spring. The smallest populations, which consist of a few individuals, grow on tree roots and in rock crevices, both beside the spring, slightly above the water surface. The submerged form differs from the terrestrial only in the following combi- nation of characters: rhizomes are rather long-creeping, branching, and bearing long slender branched roots, altogether constituting an entangled rhizome-rool system. Detached leaves or pinnae are trapped by the rhizome-root system, and new plants anchor and grow in the system (Fig. 2). Anatomical features of the roots do not differ from those of the terrestrial forms. Gemmae are profusely borne and develop into young plants on pinnae, whereas the terrestrial forms are usually not gemmiferous. However, plants that were transplanted in pots @ a greenhouse in the University of Tokyo Botanical Gardens in 1983 have so fat not produced any gemmiferous leaves. : : In the aquatic form, the apices of pinnae are broadly obtuse as compared with subacute apices of the land forms. Anatomically the pinnae consist of an upp — epidermis, a mesophyll of spongy tissue and a lower epidermis with stomata, a in the land forms, but 10-20% of stomata are incomplete (Fig. 3). Newly devel- oped leaves of two-year-old plants cultivated in soil in pots retain this feature. ty Each sporangium, a typical leptosporangium of the terrestrial Asplenium unilat- erale, produces approximately 32 spores, some of which are abortive, with the a KATO & IWATSUKI: ASPLENIUM UNILATERALE 75 Fics. 2,3. Submerged aquatic Asplenium unilaterale. Fic. 2. Leaf bearing young plants developed from gemmae. Fic. 3. Surface view of a lower epidermis. Asterisks indicate incomplete stomata. same exospore ornamentation as that of the terrestrial form. The number sug- gests an ability to be apogamous. In contrast to species of aquatic ferns that are distinct morphologically and taxonomically, the submerged Asplenium unilaterale retains the morphology for land life and is considered to be an ecotype or a local form at the early stage of quantum speciation (Grant, 1981). The peculiarly submerged aquatic life seems to be facultative in the presumably recently formed habitat. This assumption is on the small morphological differences such as incomplete development of some stomata between the submerged and land forms, and the recent origin of the spring which is suggested by the occurrence of the Ficus sp. trees in the spring. Another noteworthy aspect is reproduction of the submerged form. The veg- etative reproduction by pinna-borne gemmae that grow into new plants anchor- ing in the rhizome-root system may be the most suitable means for the fern in question to remain completely submerged in the flowing water. It is a substitute for reproduction by spores and gametophytes which, even if apogamous, might not be possible for this fern, because freely dispersed spores are easily washed away in flowing water. Since such vegetative reproduction occasionally takes place in the terrestrial Asplenium unilaterale in Seram Is., it is a phenotypic variable that is merely strongly expressed by the aquatic form. A parallel de- velopment is the possible induction of adventitious buds in a saturated atmo- sphere in the filmy fern Trichomanes minutum (Yoroi & Iwatsuki, 1977). The origin of the submerged form remains to be clarified, but it may be sug- gested that when the spring arose, the aquatic ecotype was derived from one of the possibly apogamous land forms that occur on Seram Island. Biosystematic analysis of interrelationships among the infraspecific variants of Asplenium uni- laterale on the island is needed. This study was financially supported by Grant-in-Aid for Overseas Scientific 76 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985} Field Research from the Ministry of Education, Science and Culture, Japan, nos. 58042006, 59041021 and 60043021. LITERATURE CITED GIESENHAGEN, K. 1892. Ueber hygrophile Farne. Flora 76:157-181. GRANT, V. 1981. Plant speciation. 2nd ed. New York: Columbia University Press. HENNIPMAN, E. 1977. A 2 ph f the f, g Rolhitis (I i ie id ) Leiden Bot. Series 2:1-331. The Hague: Leiden University Press. IWATSUKI, K. 1975. Taxonomic studies of Pteriodophyta X. Acta Phytotax. Geobot. 27:39-55. Lovis, J. D. 1973. A biosystematic approach to phylogenetic problems and its application to the Aspleniaceae. Pp. 211-228 in The phylogeny and classification of the ferns, ed. A. C. Jermy et al., J. Linn. Soc. Bot. 67 (Suppl. 1). MickEL, J. T. 1972. A “filmy fern” in the genus Cystopteris. Amer. Fern J. 62:93-95. MurakaMI, N. and K. IwaTsuki. 1983. Observation on the variation of Asplenium unilaterale in Japan with special reference to apogamy. J. Jap. Bot. 58:257-262. Pace, C. N. 1979. The diversity of ferns. An ecological perspective. Pp. 9-56 in The experimental biology of ferns, ed. A. F. Dyer. London: Academic Press. Price, M. G. 1982. The ferns of Steer and Harrington. Contr. Univ. Michigan Herb. 15:197-204. Yorol, R. and K. Iwatsuki. 1977. An observation of Trichomanes minutum and allied species. Acta Phytotax. Geobot. 28:152-159. REVIEW “Index filicum. Supplementum quintum: pro annis 1961-1975,” by F. M. Jat rett, with collaborators. 1985. 245 pp. New York: Clarendon Press, Oxford Uni- versity Press. ISBN 0-19-854579-7, $39.95. Students of ferns are familiar with the great value of this index, begun by that most illustrious of pteridologists, Carl Christensen, in 1905. It is the only com prehensive source for obtaining bibliographic citations (name and place of pub- lication, plus basionym for new combinations and locality for new species) for all Latin names, generic rank and above. Thus, we are forever grateful to the compilers of this and all preceding volumes. A new and welcome feature of the fifth supplement is the inclusion of names of fern allies. Corrections from past indices are denoted by an asterisk. As helpful as this index will be, it will not satisfy all needs: continuing the tradition of past volumes, infraspecific names are not indexed. In evaluating genera with which I am most familiar, I found only a few minor errors and omissions (to whom should I send them?) that will no doubt be corrected in subsequent volumes. One parting lament—one wishes that this index were more current: in the last ten years, I extrapolate that enough new names have accumulated to fill about 165 pages of a new supplement!— ALAN R. SmiTH, Department of Botany, University of California, Berkeley, C4 _ 94720. ' | | | | American Fern Journal 75(3):77-79 (1985) A Newly Discovered Habitat for Isoétes melanopoda in Louisiana R. DALE THOMAS Herbarium, Department of Biology, Northeast Louisiana University, Monroe, Louisiana 71209-0502 Isoétes melanopoda Gay & Dureau, Blackfoot Quillwort, ranges from Georgia to Texas and north to South Dakota and New Jersey. Old collections are known from East Baton Rouge, Avoyelles, and Rapides parishes in Louisiana (Brown & Correll, 1942). More recent collections have been made from Calcasieu (Brooks & Maples, 1971) and Sabine (Carroll & Thomas, 1981) parishes. However, this species has been considered rare in the state (Thieret, 1980) and until the present finds the only extant population known in Louisiana was from a low woods in Sabine Parish. Recently, plants of I. melanopoda were found to be locally abun- dant and quite common in wet, low areas of unplowed soybean and cotton fields. The new habitat for I. melanopoda in Louisiana was discovered on 8 May 1984. A low wet area in an unplowed soybean field in Richland Parish provided the first find. All the surrounding parts of the field had been treated with her- bicides to kill the winter weeds in preparation for spring planting. The low area was too wet for tractors to negotiate so it had been left unsprayed. Close ex- amination of the area revealed that literally tens of thousands of plants of Isoétes melanopoda were thriving in the field. Many plants in the treated areas showed no ill effects of the herbicide, although all seed plants were dead. A week later the entire field, including the low area, was disced and planted in soybeans. During the period of a month I found new populations in Richland, Ouachita, Morehouse, Caldwell, Catahoula, West Carroll, Bossier, and Caddo parishes. All these populations contained plants that were much larger, more numerous, and less scattered than those previously collected from Sabine Parish (Carroll & Thomas, 1981). All populations found in the nine new parishes were either in cultivated fields (soybean—West Carroll, Richland, Catahoula; cotton—More- ouse; oat—Franklin; or open wet areas like pastures or hayfields—Caldwell, Ouachita, Bossier, Caddo). The areas might be flooded during heavy winter rains but in no areas were plants found completely submerged in water. The areas are dominated by Juncus and Cyperaceae with a few dicotyledons. The most abundant Juncus species are Juncus biflorus, J. bufonius, J. diffusissimus, J. dud- leyi, J. scirpoides, J. validus, and rarely J. effusus or J. coriaceus. Cyperaceae is represented by Carex frankii, Cyperus iria, C. pseudovegetus, F imbristy lis ~~ tumnalis, F. miliacea, F. vahlii, Scirpus koilolepis, and undetermined species of Eleocharis. Dicotyledons included Tillaea aquatica, Plantago hybrida, P. virgin- ica, Callitriche peploides, Cerastium glomeratum, Ranunculus pusillus, Gratiola neglecta, G. virginca, Lindernia anagallidea, and Linaria canadensis. The sterile low specimens of Eleocharis and Juncus make the Isoétes specimens very dif- ficult to locate at first. Concentrated searching reveals that Isoetes melanopoda 78 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) leaves are more pea-green than those of the seed plants. Leaves of Isoétes are splayed in the open; this character is lost in thick vegetation. Current cultivation practices for cotton and soybeans seem to favor the spread of Isoétes melanopoda in the South. When a farmer begins to prepare his fields for planting in the spring, low, wet areas are left to be disced later after further drying. By the time these low areas are dry, spores of Isoétes are mature. The spores spend the summer in the ground and apparently germinate only after fall rains begin. The corms of old plants are also dormant during the summer and new leaves appear again in November. Denser populations of I. melanopoda occur in cultivated areas than in uncultivated ones. In the Franklin Parish pop- ulation, plants were in distinct rows following the furrows left by individual discs. All populations in Ouachita, West Carroll, and Franklin parishes checked in December, 1984 had thousands of plants from old corms bearing leaves about four inches tall. These leaves grew to about 10 inches in length by April 1985 and were withered and dead by 1 June 1985. Almost all cotton and soybean fields in this area lay fallow from harvest time in October and November until the following April or early May. Although Isoétes is a perennial plant, its sum- mer dormancy allows it advantages similar to those of winter annuals. Recent improvements in laser technology have made it possible for farmers to have their fields leveled and sloped to prevent low areas. No populations of Isoétes were found in these fields, perhaps because they can now be disced earlier and thus before the spores have had a chance to mature. Collections made during 1984: LoutsiaNA.—Richland Parish: Unplowed soybean field S of La. 15 on Richland Plantation, 1.5 miles W of Rhymes and La. 133, Sec. 10, T16N, R5E, 8 May, Thomas 88366. Franklin Parish: Edge of oat field W of dirt road, S of La. 132 and Old Mixon School near Murphy woods, Sec. 15, T16N, R8E, 8 May, Thomas 88417. Sabine Parish: Wet woods beside U.S. 171 at Bayou San Patricio, about 2.8 miles NW of Noble and La. 483, Sec. 33, T9N, R13W, 15 May, Thomas 88666 & Taylor 6956. Ouachita Parish: Low wet area of pasture on Layton Farm, S of US. 80 at Rifle Range Road E of La. 139 and Sicard, E of Monroe, Sec. 71, coin R4E, 23 May, Thomas 88835. Morehouse Parish: Wet, low unplowed area in soybean field beside paved road N of Little Lake Lafourche and La. 595 and S of Gum Ridge, Sec. 21, T18N, R6E, 23 May, Thomas 88846. Parish: shallow, temporary pool at gas pipe line beside La. 133, 4.4 miles N of La. 847 N of Hebert, Sec. 31, T15N, R5E, 24 May, Thomas & Earthman 88879. Catahoula Parish: Wet area 4! edge of field at ball park beside La. 124, S edge of Harrisonburg, Sec. 38, T9N, R6E, 24 May, se & Earthman 8886. Parish: Low end of soybean field S of La. 2 at Redwing n Colewa Bayou W of Oak Grove, Sec. 13, T21N, R9E, 25 May, Thomas 88904. Bossier Parish: be Lom edge of hay field in creek bottom N of La. 2 at branch of Cypress Bayou, 1.1 miles E of La. 157 and E of Plain Dealing, Sec. 12, T22N, R13W, 1 June, Thomas 88994 & Taylor 7166. Caddo ; Low, open wet area dominated by Juncus near a small stream on Kendrick nie W of La. 1 and 2.7 miles N of Vivian, Sec. 11, T22N, R16W, 1 June, Thomas 89029 & Taylor 720 Voucher specimens of all collections are on doitch i in Northeast Louisiana University Herbarium (NLU). Duplicates of some of the collections are availa or exchange. LITERATURE CITED Brooks, shig and R. S. Maptes, JR. 1971. A recent find of Isoetes in Louisiana. Amer. Fern J- e THOMAS: ISOETES MELANOPODA 79 Brown, C. A. and D..S. CorRELL. 1942. Ferns and fern allies of Louisiana. Baton Rouge: Louisiana State University Press. CarROLL, A. N. and R. D. THOMAS. 1981. I ] poda in Sabine Pari h, Louisiana. Phytologia 48:274-275. TutErET, J. W. 1980. Louisiana ferns and fern allies. Lafayette: Louisiana Natural History Museum. SHORTER NOTE A New Station for Dicranopteris flexuosa in Bay County, Florida.—While per- forming a routine botanical field survey at Bay Point Resort near Panama City, Florida, on 7 November 1984, I discovered a small colony of Dicranopteris flexuosa (Schrad.) Underw. (forked fern or net fern) in a shallow empty drainage ditch passing through a slash pine forest southeast of the junction of Delwood Beach Road and Magnolia Beach Road. The plants were situated at the upper edge of the ditch slope, were rooted in soil consisting of whitish sand and dark organic particles, and grew with Lycopodium cernuum and Woodwardia areo- lata. This habitat is similar to other sites where D. flexuosa has previously been found in the southeastern United States (Wherry, The Southern fern guide. 1964; Lakela & Long, Ferns of Florida. 1976). This gleicheniaceous fern, which is native to Mexico, South America, and the West Indies, was first collected in the United States in 1913 in Mobile County, Alabama, but did not persist there (Small, Ferns of the Southeastern States. 1938; Dean, Ferns of Alabama, rev. ed. 1969). Likewise, in 1947 it was discovered in a transient colony in Osceola County, Florida, and in 1955 a colony was found in Hillsborough County, Florida, that was still extant in 1964 (Wherry, op. cit.). Lakela and Long (op. cit.), however, reported no recent collections of this fern. The present account is thus the first documented report of D. flexuosa in the United States in 20 years, and the Bay County station constitutes an extension of 400 kilometers northwestward from the last previously documented site in Hillsborough County. Dicranopteris flexuosa is also hereby verified as an extant natural element of the pteridophyte flora of Florida and the United States. Owing to its obvious rarity, its inclusion on Florida’s rare and endangered biota lists is certainly warranted. Herbarium specimens (Burkhalter 9784) have been depos- ited at UWFP, FSU, and FLAS. Dr. Michael I. Cousens, a pteridologist from the University of West Florida, visited the Bay County Dicranopteris colony on 19 January 1985 and collected frond material for gel electrophoretic analyses. He also discovered a number of Dicranopteris gametophytes at that time, which possibly indicates that this small isolated colony of this interesting tropical fern is reproducing in the seemingly nonsalubrious climate of northwest Florida—JaMEs R. BURKHALTER, The Her- barium, Building 58, Room 77, University of West Florida, Pensacola, FL 32914. American Fern Journal 75(3):80-91 (1985) Anatomical Studies of Sphaeropteris and Cnemidaria (Cyatheaceae)}' TERRY W. LUCANSKY Department of Botany, University of Florida, Gainesville, FL 32611 Despite a renewal of interest in the tree ferns (Tryon, 1970, Gastony, 1973, 1974, 1981; Gastony & Tryon, 1976; Stolze, 1974), relatively little is known ana- tomically about these largest of ferns. Previous anatomical studies have dealt primarily with mature paleotropical species (Bower, 1912; Ogura, 1927, 1972; Godwin, 1932; Mehra & Singh, 1955) and have shown that similarities between genera occur in both the vascular anatomy and nodal patterns. Recent studies of mature members of the neotropical Cyatheaceae also have shown striking similarities among the genera (Lucansky, 1974b, 1976b, 1977; Lucansky & White, 1974) Tryon (1970) revised the classification of the family Cyatheaceae and recog- nized six genera and three principal evolutionary lines among the squamate genera. Sphaeropteris with its undifferentiated (conform) scales represents an evolutionary line and occurs at the base of the squamate genera, while Cnemi- daria with its marginate scales is found at the top of another evolutionary line. Sphaeropteris supposably contains elements allied to the other two major evo- lutionary groups in the family (Tryon & Tryon, 1982). On the basis of scale characters, Sphaeropteris subgenus Sphaeropteris is related to Alsophila and subgenus Sclephropteris is allied with Trichipteris, Cyathea and Cnemidaria. Holttum and Edwards (1983) believe that this latter alliance (together with cer- tain species of Tryon’s Sphaeropteris) forms a natural group which needs a new subdivision. Holttum (1963, 1964, 1965) also has intensively studied the Cyatheaceae, but recognized only a single genus Cyathea for the same taxa in the family. He subdivided the genus on the basis of characters of the stipe scales and associated characters of indusia, hairs and venation, and felt that the only sharp subdivision within the genus is between the subgenus Sphaeropteris and the subgenus Cy- athea (Holttum, 1982; Holttum & Edwards, 1983). His subgenus Sphaeropteris is characterized by peculiar marginal setae on the stipe scales and certain indusial and venation characters, while Tryon (1970) regarded the uniformity of cells of the stipe scales as the most important character for delimiting the genus Sphae- ropteris. Although Tryon (1970) promoted subgenus Sphaeropteris to generic rank, Holttum & Edwards (1983) feel that it is not a natural group and contains species that do not belong in the group. Sphaeropteris sensu Tryon is a genus of approximately 120 species found in both the New and Old World. Its distinctive feature is the presence of conform stipe scales with undifferentiated or poorly differentiated cellular construction ‘Publication No. 6244 of the Florida Agricultural Experiment Station. LUCANSKY: ANATOMY OF CYATHEACEAE 81 (Tryon, 1970). In some respects (e.g. leaf morphology, position of trichomes and indusium) this genus is similar to the other squamate genera in the Cyatheaceae. Cnemidaria sensu Tryon, is a genus of 25 species confined to the American tropics. Its closest relationship is with Cyathea (Tryon & Tryon, 1982), and Holt- tum even includes it in subgenus Cyathea of the genus Cyathea (Holttum & Edwards, 1983). According to Tryon (1970) this genus constitutes a strong evo- lutionary line, and several characters, including the lack (rarely present) of tri- chomes on the adaxial side of the costae, contrast markedly with the other squa- mate genera. Based upon spore morphology, venation, leaf architecture, habit and lack of trichomes adaxially on the costae, Cnemidaria forms the most natural and distinctive genus in the Cyatheaceae (Stolze, 1974). Holttum & Sen (1961) reinforced the distinctiveness of the genus by pointing out the peculiar character of the spores. It also is the most advanced genus in the family with its complex venation and reduced laminar architecture being end products of its adaptive development. The principal characters of Cnemidaria rarely, if ever, occur in the other squamate genera. The present study is part of a broad investigation of the neotropical Cyathea- ceae (Tryon, 1970; Tryon & Tryon, 1982). The anatomy of representative species of Sphaeropteris and Cnemidaria is elucidated to provide data that can be used to determine the taxonomic status and phyletic position of these two genera. Previous workers (Holttum & Sen, 1961; Sen, 1964; Lucansky & White, 1974) have already shown the importance of anatomy and morphology to the study of this group of plants. MATERIALS AND METHODS The following species were examined in this study: Sphaeropteris elongata (Hook.) Tryon, S. senilis (K1.) Tryon, Cnemidaria mutica var. mutica (Maxon) Stolze and G. mutica var. grandis (Maxon) Stolze. Previous workers had consid- ered these varieties of C. mutica to be distinct species (Maxon, 1912; Christensen, 1938; Tryon, 1970). Developing shoot tips were collected along roadsides or in montane and lowland rainforests in Costa Rica and Venezuela. Voucher speci- mens are on file in the herbarium of Duke University (DUKE). Plant materials were killed and fixed in formalin-acetic acid-alcohol (FAA) and sectioned on a macrotome (Lucansky, 1976a). Sections (slices) were parti- tioned into manageable sizes, dehydrated in a tertiary-butyl alcohol series and embedded in paraffin (Johansen, 1940). Sections (8 um) were made and stained with safranin-fast green. Stained sections were photographed with a 35 mm Nikon M35 S camera, and entire slices were photographed with a 35 mm single- lens reflex camera. Unless otherwise noted, the results given are based upon the combined data of all species. RESULTS AND DISCUSSION Although members of Sphaeropteris are arborescent, species of Cnemidaria are characterized by an acaulescent habit. The caudices of some species of 82 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) Cnemidaria, however, many reach 1.5 meter in length (Stolze, 1974). The stipe scales of Sphaeropteris are conform, whereas the scales of Cnemidaria are marginate. Certain species of Sphaeropteris may possess one or more dark apical setae on a scale, while the scales of Cnemidaria lack apical setae (Tryon, 1970, Tryon & Tryon 1982). The costae of Sphaeropteris (and other squamate genera] are pubescent adaxially, whereas trichomes, if present, occur only on the abaxial surface of the costae (and other axes) of Cnemidaria. A simple leaf architecture, costal areoles and a whitish arachnoid scurf also characterize Cnemidaria, but are lacking in Sphaeropteris. Spores of Cnemidaria with their three large equa- torial pores are unique in the family Cyatheaceae (Stolze, 1974). The stelar pattern in the mature stems of both Sphaeropteris and Cnemidaria is a dictyostele with overlapping leaf gaps and consists of individual vascular bundles (meristeles), each surrounded by sclerenchymatous tissue (Figs. 1-4). The number of meristeles varies from 3-5, depending upon the species and length of the leaf gaps. A single-layered epidermis composed of variously shaped, thick-walled cells typically is sloughed in mature sporophytes, although remnants of this layer may persist. In S. elongata the epidermal cells are thin-walled and occasionally fill with tannins. Previous workers had reported that the outer walls of the epider- mal cells may be thickened or cutinized (Sen & Mittra, 1966; Sen, 1968). The outermost layer of the stems of both genera studied is typically a hypodermis composed of two zones that are variable in thickness (Figs. 5, 6). The outer zone typically is partially sloughed and composed of variously shaped, thick-walled parenchyma cells filled with tannins and/or starch grains. The inner zone con- sists of sclerified, thick-walled parenchyma cells that closely resemble fibers (Fig. 5, 6). These cells are formed by secondary sclerosis of the walls with retention of the nucleus (Sen, 1968), and generally possess tannins and/or starch grains. The inner zone in S. elongata may be quite extensive and possess large, Tal- domly-distributed mucilage-sac cells—singly or in group of 2-7 cells (Fig. 10): Distinctive cubical cells (to be discussed later) are noted between the hypodermis and the cortex in both species of Cnemidaria and S. senilis (Figs. 5). In S. senilis the hypodermis consists of only a single zone of sclerified parenchyma cells filled with tannins. Previous workers (Ogura, 1972; Mehra & Singh, 1955) also reported a single hypodermal zone of fibers, but such a homogenous layer may be due to the loss of the outer zone of parenchyma cells or the age (young) the plant. A two-zoned hypodermis represents a unifying character in the group and has been previously reported for the other genera (except Metaxya) in the family (Lucansky, 1976b, 1977, 1982). Although Sen (1964) found that a band of sclerenchyma tissue may occur 2 between the cortical layers in Dicksonia and Culcita, the cortex in the species in this study consists solely of large, thin-walled parenchyma cells filled with — tannins and/or starch grains. Large mucilage-sac cells are randomly distributed in the cortex, singly or in a group of 2-10, and form an articulated laticiferlike — system (Figs. 7, 8). Similar idioblasts have been reported for Lophosoria (Luca®” — sky, 1982), the other squamate genera (Lucansky, 1976b, 1977) and Dicksoni¢ © (Williams, 1925). Schiitze (1906) called these idioblasts excretion containers, rath-_ LUCANSKY: ANATOMY OF CYATHEACEAE 83 3 4 Fics. 1-4. Transections of tree-fern stems. Fic. 1. Sphaeropteris elongata, x8. Fic. 2. Sphaeropteris senilis, x1.5. Fic. 3. Cnemidaria mutica var. mutica, 1.7. Fic. 4. Cnemidaria mutica var. grondie, *1.3. cb = cortical bundle, m = medullary bundle, me = meristele, s = external stelar sheath, s’ = internal stelar sheath, x = primary xylem. er than secretion cells, and reported that they contain fatty acids or tannins, whereas Ogura (1972) found that they contain slime. Numerous small localized areas of sclerified parenchyma cells, filled with tannins, occur within the cortex of both species of Cnemidaria and S. senilis (Fig. 8), but are lacking in S. elon- 84 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 83 (1985) Fics. 5-8. Anatomical details of tree-fern stems. Fic. 5. Two-zoned hypodermis of Cnemidaria mutica var. grandis, x 124. Note cubical cells. Fic. 6. Two-zoned hypodermis of Sphaeropteris elon- gata, X130. Fic. 7. Mucilage-sac cells in cortical zone of S. senilis, x 126. Fic. 8. Localized areas of sclerified parenchyma in cortex of C. mutica var. grandis, x144. c = cortex, cc = cubical cell, ep = epidermis, h, hypodermis, ms = mucilage-sac cell, st = sclerified (parenchyma) tissue. LUCANSKY: ANATOMY OF CYATHEACEAE 85 gata. Cubical cells occur between these sclerified areas and the cortex in these species (Fig. 9). Similar areas of sclerenchyma tissue have been noted in the cortex of Cystodium (Sen & Mittra, 1966), Lophosoria (Lucansky, 1982) and some species of Trichipteris and Cyathea (Lucansky, 1977}, but do not occur in Alsoph- ila and Nephelea (Lucansky, 1976b). Cortical bundles are a characteristic feature of the Cyatheaceae (Lucansky, 1974b, Ogura, 1972), and are not found in any other group of ferns. Cortical bundles undergo divisions and fusions, end blindly in the cortex, fuse with meristeles and leaf traces, or infrequently proceed directly to the petiole as a leaf trace (Lucansky 1974b). Schiitze (1906) believed that they are possibly in- volved with the movement and deposition of starch in the stem. In the present study cortical bundles occur in S. elongata and in S. senilis (Fig. 10), but are lacking in both species of Cnemidaria. These bundles may be large and form a conspicuous feature in S. elongata (Figs. 1, 10), but are small and relatively indistinct in S. senilis (Fig. 2). Cortical bundles have previously been found in some species of Nephelea, Alsophila, and Trichipteris (Lucansky 1976b, 1977), but are lacking in Cyathea, Metaxya, and Lophosoria (Lucansky 1977, 1982). Both Stolze (1974) and Tryon (1970) believe that Cnemidaria has probably arisen from a line of Cyathea, and the lack of cortical bundles in both genera supports this contention. These accessory bundles in Sphaeropteris may possess a partial sheath of sclerified parenchyma cells, although Schiitze (1906) reported that cor- tical bundles usually lack a sheath. They are surrounded by a distinct single- layered endodermis that may be filled with tannins and possess Casparian strips (Fig. 10). A parenchymatous pericycle (1-3 layers) encircles the primary phloem, which consists of both sieve cells and phloem parenchyma. No tangential cells are found in the phloem of the cortical bundles, as previously reported for Nephelea, Alsophila, and Trichipteris (Lucansky 1976b, 1977). Small bundles have only xylary element in their centers, whereas large bundles may have one or more parenchymatous areas within the xylary mass. The primary xylem con- sists of tracheids with scalariform wall thickenings, with xylem parenchyma filled with tannins interspersed among these xylary elements (Fig. 10). Xylem maturation is mesarch. In all species a meristele is surrounded by an external and internal stelar sheath composed of sclerified parenchyma (or sclerenchyma) cells filled with tannins. Both stelar sheaths arise from localized areas of sclerified cells that undergo fusion to form a continuous sheath (Lucansky & White, 1976). The pres- ence of sclerified tissue around the individual meristeles is a characteristic fea- ture of the Cyatheaceae, and had been previously noted in other cyatheoid genera (Lucansky 1976b, 1977, 1982; Ogura, 1972). Both stelar sheaths typically are delimited externally and internally by a single layer of cubical cells. These distinctive cells are greatly thickened on three walls (wall proximal to thin- walled parenchyma cells remains thin-walled), and each cell contains a single, large solitary crystal (Fig. 9). The crystalloid structure in each cell is insoluble in H,SO, (Holttum & Sen, 1961) and is thought to be composed of silica (Sen, 1968). Cubical cells have been found in other genera of neotropical Cyatheaceae (Lu- cansky 1976b, 1977, 1982), and occur in the cortex of certain dicksonioid species 86 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) psttatalNsPeareirayy : . ; LAY e: HY : : « Fics. 9-12. Anatomical details of tree-ferns stems. Fic. 9. Cubical cells with crystals in SphaeroP teris senilis between external stelar sheath and parenchymatous zone of a meristele, x 560. Fic. ey Large cortical bundle in S. elongata, x 123. Fic. 11. Meristele of S. elongata showing tangential cells, x57. Fic. 12. Meristele of Cnemidaria mutica var. mutica showing tangential cells, x 120. cc = cell, cr = crystal, e = endodermis, ms = mucilage-sac cell, p = primary phloem, pe = pericycle, P2 ss parenchymatous zone, t = tangential cell, x = primary xylem. LUCANSKY: ANATOMY OF CYATHEACEAE 87 (Sen, 1964). Although Ogura (1972) thought that these cells were sclerenchyma cells, their living protoplast, wall morphology, position and resemblance to parenchyma cells in young stems indicate that they are thick-walled paren- chyma. Sen (1964) also believed that they were not sclerenchyma cells based upon their rate of cell division and cellular inclusions. A parenchymatous zone composed of large, thin-walled parenchyma cells filled with starch grains and tannins separates the stelar sheaths from each meri- stele (Figs. 1, 12). Large mucilage-sac cells occur singly or in groups of 2-7 within this zone (Fig. 11). Schiitze (1906) believed that these parenchymatous zones may function in the conduction and storage of carbohydrates. Each meristele in an amphicribral bundle delimited by a distinct endodermis filled with tanniferous substances (Figs. 11, 12). Distinct Casparian strips are lacking in the walls of these cells. A pericycle of 2-3 rows of thin-walled paren- chyma cells completely encircles the primary phloem. Although Ogura (1927) reported that the protophloem was usually compressed, thick-walled, and swol- len, the protophloem and metaphloem typically are indistinguishable in material studied. The latter tissue is composed of sieve cells, phloem parenchyma cells filled with tannins, and distinctive tangential cells (Figs. 11, 12). The phloic tan- gential cells are large and elongate tangentially in transection (Fig. 13) and form a characteristic feature of the Cyatheaceae. They are more numerous in Cnemi- daria than in Sphaeropteris, and their position and pattern of arrangement is similar to that reported for the other cyatheoid genera (Lucansky 1976b, 1977, 1982). They represent specialized sieve cells that are devoid of nuclei, possess sieve areas on their lateral walls and accumulate callose (Sen, 1964), although they have been variously referred to as false sieve tubes (Schiitze, 1906) or mu- cilage cells (Ogura, 1927, 1972). According to Ogura, these distinctive cells may be partially or entirely replaced by mucilage cells or longitudinally elongate cells. The primary xylem is primarily composed of tracheids with scalariform wall thickenings, with xylem parenchyma cells filled with tannins and starch grains interspersed among these tracheary elements (Figs. 11, 12). Although earlier studies (Ogura, 1927; Sen, 1964) reported that protoxylem is usually absent in the primary xylem of mature stems, tracheids with spiral wall thickenings were infrequently seen in the present study. Each meristele is composed predomi- nantly of metaxylem and xylem maturation is mesarch in all species studied. The pith contains large, thin-walled parenchyma cells that frequently anaes starch grains and tannins. Large mucilage-sac cells occur randomly in the pith, either singly or in groups of 2-5. In all species studied, numerous medullary bundles are scattered in the pith (Fig. 14) and represent another characteristic feature of the Cyatheaceae. They have been found in the other squamate genera in the family (Lucansky, 1974b, 1976b, 1977; Ogura, 1972}, but are lacking in Metaxya and Lophosoria (Lucansky, 1974a, 1982). Adams (1977) reported that the medullary bundle system of Cyathea fulva consists of a central network and a peripheral one composed of three types of bundles. Medullary bundles arise de novo in the pith (Ogura, 1927, Lucansky, 1974b) or are separated from the leaf-gap margins (Godwin, 1932; Lucansky, 1976b). They undergo fusions and 88 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) Ex; ee er w4 e. Fics. 13-16. Anatomical details of tree-ferns stems. Fic. 13. Tangential cells in the primary eae of meristele of Sphaeropteris elongata, x139. Fic. 14. Medullary bundle with partial sheath pret pith region of Cnemidaria mutica var. mutica, x581. Note mucilage-sac cells. Fic. 15. ee eee er f diarch root of S. senilis, x67. Fic. 16. Transection of petiole strand of S. senilis, *71. mucilage-sac cell, mx = m etaxylem, p protoxylem, s = sheath, t = tangential cell, x = primary xylem. LUCANSKY: ANATOMY OF CYATHEACEAE 89 divisions, end blindly in the pith, fuse with a meristele or leaf trace or proceed directly to the petiole as a leaf trace (Lucansky, 1974b; Lucansky & White, 1974; Adams, 1977). The medullary bundles are identical in cellular composition to cortical bundles. Although phloic tangential cells have been found in medullary bundles of certain species of Alsophila and Nephelea (Lucansky, 1976b), no phloic tangential cells occur in these bundles in the species studied. Small med- ullary bundles typically have only xylary elements in their centers, whereas larger bundles may have 1-several parenchymatous areas within the xylary mass. Each bundle may be partially or totally surrounded by a partial sheath composed of sclerified parenchyma cells (Fig. 14), although Ogura (1972) reported that med- ullary bundles usually lack a sheath. Only those bundles located along the in- ternal stelar sheath normally lack such tissue in the species in this study. In addition to these sheaths, variously sized localized areas of sclerified paren- chyma cells are randomly scattered in the pith of both Sphaeropteris and Cne- midaria. Cubical cells occur between these areas and the pith proper. Localized areas of sclerified tissue also occur infrequently in Nephelea and Trichipteris, but commonly occur in Cyathea (Lucansky 1976b, 1977). Transections of the adventitious roots show similar anatomical features in all species studied (Fig. 15). The epidermis is typically sloughed in mature roots, and the outer cortex, composed of thick-walled parenchyma cells, forms the outer boundary of the organ. The epidermis, if present, is composed of thick- walled cells filled with tannins. The outer cortical cells also are frequently filled with tannins and may be partially sloughed. The inner cortex, composed of thicker-walled, sclerified parenchyma cells with tannins, typically forms the bulk of the cortex. Previous workers (Schiitze, 1906; Sen, 1968) found the position of these two cortical zones reversed, while other investigators (Ogura, 1927, 1972, Lucansky 1976b, 1977, 1982) indicated a similar arrangement for the zones thai comprise the cortex in the cyatheoid genera. In S. senilis the cortex may be composed of three zones of parenchyma cells, based upon wall thickness. A distinct endodermis composed of a single layer of cells filled with tanniferous substances delimits the stele (Fig. 15). A pericycle (1-2 cells) composed of large, thin-walled, tanniferous parenchyma cells surrounds the vascular tissue. The primary phloem consists of sieve cells and phloem parenchyma, while the pri- mary xylem is composed primarily of scalariform pitted metaxylem and some spiral and reticulate—scalariform protoxylem. The xylem is diarch with exarch maturation (Fig. 15). Vascular parenchyma cells filled with tannins occur be- tween the primary phloem and xylem. Root traces originate either from a meristele or from the base of leaf traces and pass obliquely through the cortex. Leaf traces arise at successive levels in a leaf gap and proceed to the petiole to form a much-dissected vascular pattern that is similar for all species studied (Lucansky & White, 1974). Both types of traces are identical in cellular composition to the accessory bundles. Sphaeropteris and Cnemidaria. A single-layered epidermis composed of small thick-walled, tanniferous cells forms the outer boundary of a petiole. A two- zoned hypodermis occurs beneath the epidermis. The outer zone consists of 90 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) large, thick-walled parenchyma cells, whereas the inner zone is typically more extensive and composed of small, sclerified parenchyma cells. Both zones fre- quently contain tannins, and mucilage-sac cells infrequently occurs in the inner zone in S. elongata. Ground tissue, composed of thin-walled parenchyma cells with starch grains and/or tannins, comprises the bulk of a petiole. Large muci- lage-sac cells are randomly scattered in this tissue in both genera—either singly or in groups of 2-3 (Figs. 16). Larger groups (4-10) of these idioblasts are infre- quently found in Cnemidaria. Similar idioblasts were noted in the petioles of the other cyatheoid genera, but are lacking in Metaxya (Lucansky 1976b, 1977, 1982). In addition to mucilage-sac cells, small localized areas of sclerified paren- chyma cells occasionally occur in the ground tissue of both species of Cnemi- daria, but are absent in Sphaeropteris. These areas arise de novo, and are scattered in this tissue. A layer of cubical cells surrounds each area, and also occurs between the ground tissue and inner hypodermal zone in C. grandis. Each petiole strand is surrounded partially or totally by a sheath of thick- walled, sclerified parenchyma cells (infrequently some cells become fibers) (Fig. 16). A layer of cubical cells may occur on both sides of a given sheath in Cnemi- daria. Cellular composition and arrangement of the stele of each petiole strand is similar to a meristele or accessory bundle of the stem. Tangential cells infre- quently occur in the primary phloem in both species of Cnemidaria, but are not found in Sphaeropteris. The primary xylem is U- or V-shaped, with the proto- xylem found in a median position on the concave side of the vascular tissue (Fig. 16). Typically a single, large intercellular space is produced schizolysigenously adaxial to the protoxylem pole in both genera. Previous studies have reported that the protoxylem may partially disintegrate and form a cavity with tyloses (Schutze, 1906; Ogura, 1927), although no cavities were noted in the petioles of other cyatheoid genera (Lucansky 1976b, 1977, 1982). Based upon the species studied, Sphaeropteris and Cnemidaria are similar in many anatomical characters, but are significantly different in certain morpho- logical and anatomical features to warrant their generic separation in the Cy- atheaceae. Anatomically they are closely related and form a natural group with the other squamate genera in the family, and should not be widely separated in any phyletic scheme proposed for the family. LITERATURE CITED ADAMS, Davin C. 1977. Ciné an ly ae ee l Tl ee . Cyathea fulva. Amer. Fern J. 67:73-80. Cs v Bower, F.O. 1912. Studies in the phylogeny of the Filicales. II. Lophosoria, and its relationship '° the Cyatheoideae and other ferns. Ann. Bot. (London) 26:269-323. Cc ; & 1938. Filicinae. In Manual of Pteridology, ed. F. Verdoorn. The Hague: Martinus Gastony, G. J. 1973. A revision of the fern genus Nephelea. Contr. Gray Herb. 203:81-148. "1974. Spore morphology in the Cyatheaceae. I. The perine and sporangial capacity: general considerations. Amer. J. Bot. 61:672-680. - 1981. Spore morphology in the Dicksoniaceae. I. The genera Cystodium, Thyrsop and Culcita. Amer. J. Bot. 68:808-819. eat R. M. Tryon. 1976. Spore morphology in the Cyatheaceae. II. The genera Lophsorie. etaxya, Sphaeropteris, Alsophila, and Nephelea. Amer. J. Bot. 63:738-758. ' | LUCANSKY: ANATOMY OF CYATHEACEAE 21 Gopwin, H. 1932. Anatomy of the stele of Cyathea medullaris Sw. New Phytol. 31:254-264. Ho.ttruM, one E. 1963. Cyatheaceae. In Flora Malesiana, Series II, Pteriodophyta, Vol. 1 (2):65-176. —_.. . The tree ferns of the genus Cyathea in Australasia and the Pacific. Blumea 12:241- en ———. 1965. Tree-ferns of the genus Cyathea in Asia (excluding Malesia). Kew Bull. 19:463- 487. . 1982. Species of Cyathea in Western Pacific related to C. multiflora Sm. and allies in America. Kew Bull. 37:383-3 —— igh ve SEN. 1961. planes al classification of the tree ferns. Phytomorphology 11: 406- — oes P: a Edwards. 1983. The tree-ferns of Mount Roraima and neighboring areas of the Guayana Highlands with comments on the family Cyatheaceae. Kew Bull. 38:155-188. JOHANSEN, D. A. 1940. Plant microtechnique. New York: McGraw Hill. LucaNnsky, T. W. 1974a. Comparative studies of the nodal and vascular anatomy in the neotropical Cyatheaceae. I. Metaxya and Lophosoria. Amer. J. Bot. 61:464-471. . 1974b. Comparative studies of the nodal and vascular anatomy in the neotropical Cy- preg II. The squamate genera. Amer. J. Bot. “ 72- 480. Tide Sees a new approach for th g of large pl peci Stain Sartes 51:199-201. 1976b. jpucaict studies of the neotropical Cyatheaceae. I. Alsophila and Nephelea. Kole. Fern J. 66:93-101. 1977. Anatomical studies of Cyathea and Trichipteris (Cyatheaceae). Amer. J. Bot. 64: 253- 259, . 1982. Anatomical studies of the neotropical Cyatheaceae. I]. Metaxya and Lophosoria. Amer. Fern J. 7. : Se ind R.A: ae 1974. Comparative mutes of - nodal and mesial anatomy in the neotropical Cyatheaceae. III. Nodal and p clusions. Amer. . Bot. 61:818-8 and R. A. WHITE. 1976. Comparative — poured in young sporophytes of tree ferns. I. A primitive and an advanced tax er. J. Bot. 63:463-472. Maxon, ges R. 1912. The tree ferns of North pei Se Inst. Annual Rep. 1911:473- MEHRA, P ‘ ree . P. SINGH. 1955. Observations of the anatomy of Alsophila glabra Hook. Sci. & Cult. Ocura, Y. 8, “Comparative anatomy of the Japanese Cyatheaceae. J. Fac. Sci. Univ. Tokyo, Sect. Bite gee: ig yr neancutiea anatomy of vegetative organs of the pteridophytes. 2nd ed. in K. Linsbauer’s Handbuch der Pflanzenanatomie. Bd. 7. T. 3. Berlin: Gebriider Borntraeger. Scuiirze, W. 1906. Zur physiologischen Anatomie einiger tropischer Farne, besonders der Baum- farne. Beitr. Wiss. Bot. (Stuttgart) 5:329-376. Sen, U. 1964. Importance of anatomy in the clyloaees of tree ferns and their allies. Bull. Bot. Soc. Bengal 18:26-34. “———._ 1968. Anatomy of Culcita macrocarpa. Canad. J. Bot. 46:43-46. and D. Mirrra. 1966. The anatomy of Cystodium. Amer. Fern J. 56: StoLze, . ms 1974. A taxonomic revision of the genus Cnemidaria ee tat Fleldiane, Bot. 1-98. Tryon, R te 1970. The classification of the Cyatheaceae. Contr. Gray Herb. 200:1-53. ~~ and A. F. TrYon. 1982. Ferns and allied plants. New York: Springer-Verlag. eae S. 1925. Some points on the anatomy of Dicksonia. Proc. Roy. Soc. Edinburgh. 45:286- American Fern Journal 75(3):92-102 (1985) Germination and Morphology of Spores of Trichipteris corcovadensis LucIANO M. EsTEVEs and GIL M. FELIPPE Departamento de Fisiologia Vegetal, Instituto de Biologia, C. P. 6109, Universidade Estadual de Campinas, 13100 Campinas, Brasil THEREZINHA S. MELHEM Segao de Dicotiled6éneas, Instituto de Botanica, C. P. 4005, 01000 Sado Paulo, Brasil Several papers have dealt with the physiology of germination of pteridophyte spores. Published data show that spores of the majority of fern spores are light- sensitive. They also germinate in a wide range of temperatures, from 1°C (Pterid- ium aquilinum) to 35°C in the case of Onoclea sensibilis (Miller, 1968). To our knowledge, no research has been published in Brazil on the physiology of ferns, with the exception of the germination of spores of Cyathea delgadii (Marcondes-Ferreira & Felippe, 1984). In this paper a study was made of ger- mination of Trichipteris corcovadensis (Raddi) Copel. This species is a tree fern that occurs in the Serra do Mar area of Southern Brazil (Tryon, 1970), at eleva- tions from 250 to 2100 meters (Barrington, 1978). A study of the spore morphology of several specimens was also carried out. Spore morphology of the species has been studied previously as Trichopteris corcovadensis (Kremp & Kawasaki, 1969; Erdtman & Sorsa, 1971). MATERIALS AND METHODS Material of Trichipteris corcovadensis was collected in the Reserva Biolégica do Parque Estadual das Fontes do Ipiranga, Sao Paulo, Brazil, and identified by Dr. Paulo Windish (UNESP, Sao José do Rio Preto). Spores were collected from five different specimens gathered in March of different years. These collections were designated: A (1982 and 1983); B (1980, 1981, and 1982); C (1982 and 1983); D (1982); and E (1982). The spores were stored in closed bottles in the dark at 4°C. Some material from specimen B was also stored at 25°C. Spores from the five specimens were always stored separately, and the experiments were carried out with all five specimens for comparison. Germination.—Germination studies were carried out using 25 ml of Mohr's solution, as modified by Dyer (1979), in an Erlenmeyer flask. The flasks with medium were autoclaved at 120°C for 20 minutes, and nystatin (50 units-ml~’) was then added as recommended by Dyer (1979). Spores were scattered on the surface of the medium with the aid of a spatula. For each treatment, three flasks were used, and three microscope slides were prepared from aliquots from each flask. Five fields from each slide were examined through a microscope. Germi- nation was considered as the protrusion of the rhizoid. € experiments were carried out in growth cabinets (Forma Scientific model ESTEVES ET AL.: GERMINATION IN TRICHIPTERIS 93 24). Temperature was maintained at 25°C and light irradiance at 320 u.W-cm~? (continuous white fluorescent light). In some experiments the effect of constant temperature was examined (5, 10, 15, 20, 25, 30, 35, 40, and 45°C) as well as alternate temperatures of 5-25, 10-25, 15-25, 20-25, 30-25, 35-25, 40-25, and 45- 25 (12 hours at each temperature in a 24 hour cycle, starting temperature 25°C). The effect of photoperiod was also studied and photoperiods of 1, 8, 12, 16 and 24 hours were used. The effect of irradiance was studied using light irradiances of 220, 320, 900, 1400, 1900, and 2500 1.W-cm~’. The growth regulators gibberellic acid (GA,), indole-3-acetic acid (IAA), 6-benzyladenine (6-BA) and 2-chloro- ethylphosfonic acid (CEPA) were used at the following concentrations: 0, 5, 25, 50, and 100 ug-m1~' (the growth regulators were added to the flasks after auto- claving). Viability of the spores was tested with acetocarmine according to Manton (1950). Spores were placed on a microscope slide with a drop of acetocarmine, covered with a cover slip, and pressed to break the spore walls (acetocarmine did not penetrate walls of intact spores). Spores with broken walls were counted 36 hours later, whether stained or not (data are presented as percentage of stained spores in relation to total spores with broken walls; 15 replicates were used). Germination data are presented as the angular value: arc sin \/%. The per- centage value is the number of germinated spores in relation to the total number of spores in each field. Statistical analysis was performed with data transformed into angular values. Where necessary the confidence interval (at 95%) and coef- ficient of variation are presented. Analysis of variance was also performed, and when F.,, was significant the LSD,., was determined by Tukey’s method as modified by Snedecor (1962). Lower case letters in tables compare the values within columns, taking into account the LSD,... Lipids and protein.—Total lipids were determined according to Gemmrich (1977), using 50 mg of spores in each extraction. Each specimen was analyzed in triplicate. Data are presented as mg lipid-100 mg~' of spores. Soluble protein was determined according the dye-binding method of Bradford (1976), using 200 mg of spores in each determination and three replicates per treatment. Data are Presented as mg protein-100 mg~' spores. Morphology of spores.—The spores were submitted to acetolysis according to Erdtman (1969). To observe the perine, spores were treated with sodium carbon- ate at 3% before acetolysis (Morbelli, 1974). The Wodehouse (1935) method was used to determine the presence of cellular content in the spores. At least five slides were prepared for each sample. Measurements were made with 25 acetolysed spores, distributed on at least three slides (Salgado-Labouriau et al, 1965). Spores were measured not later than seven days from the prepa- ration of the slide, to avoid swelling of spores (Melhem & Matos, 1972). The Percentage of different spore shapes was determined in 45 fields, distributed on at least three slides (the same for the observation of the perine). To obtain the Photomicrographs, an Olympus Vanox photomicroscope with an automatic cam- era was used. Kodak Panatomic-X, was used with a green filter to obtain details of the spores, For specific details, the scanning microscope was also used, as described by 94 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) [-) > mo G o za ° a .- z = oO prey © DAYS Fic.1. Germination of spores of specimen A of Trichipteris corcovadensis at 25°C in constant white fluorescent light nae uW-cm?) and darkness over a 25 day period. Open circles: light; closed circles: darkness. Confidence intervals (5%) are shown. Cruz & Melhem (1984). Photographs were obtained from a scanning electron microscope GO, model JSM P15, with Kodak PX-135 (ASA 100 film). The measurements presented were subjected to an analysis of variance, and when F,., was significant, the LSD,., (Tukey) was determined (Snedecor, 1962}. TABLE 1. Effect of Temperature on the Germination of Spores of T. corcovadensis (Specimen A) Kept in Continuous White Light. Temperature* (°C) Germination (angular value) 20 sre? 25 16.6° 30 10.4° LSD,,, (Tukey) 48 25-20 12.72 25-25 16.6 13.0* LSD,,, (Tukey) 2.4 et a ee es) Ce a ene, * No germination occurred at 5, 10, 15, 35, 40, 45°C and in the alternate pairs 25-5, 25-10, 25-15, 25-35, 25-40 and 25-45°C. ** See Materials and Methods; a, b, and c, compare the values within columns. eee ESTEVES ET AL.: GERMINATION IN TRICHIPTERIS 95 TaBLeE 2. Effect of Light scene and Photoperiod on the Germination of Spores of T. corcova- densis (Specimen A) Kept a Irradiance Germination Photoperiod Germination (uW-cm~?) (angular value) h) (angular value) 220 18.0° 0 0.0 320 18.9° 1 0.0 900 13.8 8 17.6 1400 13.4> 12 18.5 1900 12.9 16 17.7 2500 13.6° 24 18.9 LSD,., (Tukey) 3.0 F,,, not significant for photoperiods of 8, 12, 16, and 24 h. RESULTS AND DISCUSSION In all experiments dealing with germination, spores were obtained only from specimen A Figure 1 shows the results of one experiment in which spores were kept at 25°C either in continuous white fluorescent light or in darkness. No germination occurred in darkness and this was observed in all experiments. In the light, germination started on day 6. This proceeded at a fast rate until day 14 when a plateau was reached and maintained up to day 25. The maximum germination value was around 20 (angular value). In subsequent experiments germination was always counted 14 days after the start of the experiment. No germination occurred at 5, 10, 15, 35, 40, and 45°C, even when these tem- peratures were alternated daily with 12 hours at 25°C. Germination occurred a 20, 25, and 30°C and with the alternate pairs 20-25 and 25-30°C (Table 1). At all temperatures tested, the spores germinated only in the light. Germination was higher with the two lower light irradiances used (Table 2) and germination oc- curred only (with the photoperiods tested) with photoperiods of 8 h and above (Table 2). In all the other experiments mentioned i in this Cap age as light irradi ance used was 320 1 W-cm~ and light w g Pp in which temperature was tested shown 1 previously). TABLE 3. Effect of Growth Regulators on the Germination of Spores of Specimen A of Trichipteris corcovadensis, in Continuous Fluorescent White Light, at 25°C. a ee Germination (angular value) Concentration (ug-ml-) IAA GA, 6-BA = 0 18.8" 18.57 16.9°* 17.9 5 14.1 5g 7 0.0 25 3.6" 10.2° 0.0 - 50 0.0 a7" = AE 100 a SA eee a LSD... (Tukey) 2.2 1.8 _ “ ee mein * Student test. 96 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) ® Fic. 2. Schematic representation of spores of Trichipteris corcovadensis showing the two extreme shapes. A = shape I; tl = trilete length; ts = trilete scar. B = shape II Spores of T. corcovadensis are positively photoblastic like the majority of pteriodophytes: Miller (1968) in his review showed that in 75 species studied only 4 germinated in darkness. Cyathea delgadii occurs with T. corcovadensis in the same region, shows germination over a wider range of temperatures (15 to 30°C), and is also a photoblastic positive species (Marcondes-Ferreira & Fe- lippe, 1984). All growth regulators tested had an inhibitory effect upon germination (Table 3) and were also unable to change the photoblasticity of the spores. In all experiments the maximum values for germination were always low. This can also be seen in Table 4, in which the effect of period of storage at 4°C on germination is shown. After 135 days, germination is reduced. This was vali for spores of specimen A which were collected both in 1982 and 1983. TABLE 4. Effect of Storage on the Germination of — of Trichipteris corcovadensis (Specimen A) in Continuous Fluorescent White Light at 25°C oo Storage at 4°C Germination (days) (angular value) 15 19.17 67 17.0° 92 17.18 96 18.9" 102 18.4" 108 19.2° 134 16.9" 135 18.7° 238 11.5" 239 9.0" 254 9.2» ESTEVES ET AL.: GERMINATION IN TRICHIPTERIS 97 TasLeE 5. Shape Variation of Spores of Five Specimens of Trichipteris corcovadensis Collected in Different Years. Specimen—year Shape I Shape II Intermediate shapes A—1982 100.0 0.0 0.0 B—1980 98.1 0.0 1.9 B —1981 96.4 0.0 3.6 B—1982 28.6 27.1 44.3 C—1982 30.5 22.9 44.6 D—1982 28.5 14.3 57.2 E —1982 31.5 22.4 46.1 Spores collected from specimens B, C, D, and E never germinated. No treat- ment was able to change this. This lack of ability to germinate was observed with both freshly collected spores and stored spores. For example, spores of specimen B were stored at 4°C and 25°C and germination was tested from 15 to 960 days in regular intervals, but germination never occurred. After acetolysis two extreme spore shapes could be seen: I (Fig. 2A) and II (Fig. 2B) and intermediate shapes between I and II. All spores of specimen A belonged to shape I (Table 5). Spores of specimen B collected in 1980 and 1981 belonged to shape I (around 95%) and intermediate shapes; specimens C, D, and E presented shapes I, II, and intermediates. Viability was studied by staining the spores with acetocarmine. Acetocarmine stains the protoplasm and in so doing suggests that the spore is living or viable, but it does not signify that the stained spore can germinate. In this paper the term viable means living as defined by Bewley & Black (1978). For example, immature embryos of Annona crassiflora are viable or living, but they do not germinate for at least 200 days after harvest (Rizzini, 1973). All spores tested for viability were from specimens collected in 1982. After breaking the walls me- chanically the spores were stained with acetocarmine. The relationship between stained and nonstained spores with broken walls was independent of spore shape. Results are shown in Table 6. In all cases the percentage of stained spores was above 70%. The majority of spores exhibited cellular content when they were subjected to the Wodehouse method (Fig. 3). Thus, all specimens had viable (living) spores, regardless of spore shape. TABLE 6. Viability of Fresh! Five Specimens of Trichipteris corcovadensis, y Collected Spores of Five Specim: P - cted in 1982. Spores Were Stained with Acetocarmine After Having Their Walls Broken Me- eS en ae Specimen Viability (7%) — lee lc = 88 98 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) Fics. 3-7. Spores of Trichipteris. Fic. 3. Specimen C treated by the Wodehouse method, showing the cellular content. Fic. 4. Surface view of the spore of specimen E, scanning electron photomicro- ESTEVES ET AL.: GERMINATION IN TRICHIPTERIS 99 TaBLE 7. Total Lipid and Soluble Protein in Spores of Five Specimens of Trichipteris corcovadensis Collected in Different Years. mg:100 mg‘ spores Specimen—year Lipid* Protein* A—1982 44.2 1.14 B—1981 47.5 1.01 B —1982 48.8 0.93 C—1982 50.1 gS —1983 45.0 a D—1982 — 0.98 E —1982 a 1.02 * F,,, not significant for lipids and protein. The content of total lipids and soluble protein was determined in spores col- lected in different years. About 47% of the weight of the spores is lipid and only 1% is protein (Table 7). No differences could be seen in lipid and protein content in the different specimens studied. Gemmrich (1977, 1982) showed similar values for lipids (over 59% by weight) with spores of Anemia phyllitidis. According to Cran (1979), 20% by weight of spores of Dryopteris pseudomas is lipid. Measurements of the equatorial diameter and the trilete length (see Fig. 2A) of spores after acetolysis are shown in Table 8. Spores from specimen A showed the largest diameter, but this diameter was not statistically different from that of spores from specimen B. In all cases the length of the trilete is about 1.6 times smaller than the equatorial diameter independent of the spore. TaBLe 8. Equatorial ina and Trilete Length of Five Specimens of Trichipteris corcovadensis Collected in Different Equatorial diameter Trilete length X pm xX pm Specimen—year Shape I Shape II Shape I Shape II A—1982 73.1° — 46.0° oe —1980 57.84 57.0% 35.9° 34.8° B—1981 71.5" 57.84 46.3* 45.0 B —1982 64.2% 34 41.6° 37.6 C—1982 65.5° 57.44 40.5 37.4° D—1982 68.5°° 61.1% 43.9% 37.3° E —1982 68.5" 62.3° 44.0% 37.7" (Tukey) 3.2 3.6 Coefficient of variation 5.4% _ ii ee eo ae Staph. Fic. 5. Surface view of specimen D, scanning electron phot ph. Fic. 6. Surface view of specimen B showing the perine, scanning electron photomicrograph Fic. 7. et of specimen B after treatment with sodium carbonate, as shown in the optical microscope. 100 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985} TABLE 9. Main Aspects of the Surface of Spores of Five Specimens of Trichipteris corcovadensis. Specimen Main sculpture Pits near trilete scar A granulate ? evident B granulate ? or pitted, interca- evident lated with small granules C psilate, clearly pitted smaller near the trilete scar D psilate, delicately pitted absent E psilate, with small and few pits evident The surface of the spores was observed through the optical and the scanning electron microscopes. The sculpture is variable and variations occur not only among the five specimens but also in the same sample. An idea of this variation is given in Table 9, but even so it was possible to define the main sculpture pattern of the spores from each specimen. For example, spores from specimen A show a granulate sculptine while specimen D exhibits a psilate pattern. Spec- imen E shows the trilate scar demarcated by pits (Fig. 4), which are not observed in spores from specimen D (Fig. 5). According to Gastony (1974), the perine is lacking in many species of Trichip- teris. In the present case, perine was observed under both the scanning (Fig. 6) and optical microscopes (Fig. 7). Data presented in Table 10 were obtained wi the optical microscope (spores treated with sodium carbonate before acetolysis). It can be seen in Table 10 that 94% of spores of specimen A showed immature granulate perine. Perine at different stages of development was present in spores independent of their shape. According to Gastony & Tryon (1976) perinous sculptine is attained quite late in spore maturation. They state that “when dried, very immature spores will collapse as the result of incomplete development of mature wall structure and such spores will appear pinched in microscopy”. Maturation is a continuum, and even though the mature wall structure will preclude the pinched appearance upon drying, the late development in the final perine is not attained yet. Spores 9 10. Frequency of Spores with Perine of Trichipteris corcovadensis Collected in Different ears. Specimen—year Perine (%) A—1982 94° B—1980 79¢ B—1981 86> 1982 7g0 LSD,,, (Tukey) 44 ESTEVES ET AL.: GERMINATION IN TRICHIPTERIS 101 not quite fully mature will present different stages of perine deposition (Gastony & Tryon, 1976). They rely on their experience to judge immature, partially ma- ture, or fully mature perines. Taking into account the work of Gastony (1979), all spores of specimens B, C, D, E and most of spores from A are immature. None of the spores of specimen A present the perine as hair-like processes, considered by Gastony (1979) as a feature of mature perine, which identifies a mature spore. However, some of the immature spores of specimen A germinate, showing that even though they are not fully mature morphologically, they have reached maturity from a physiolog- ical point of view. The majority of spores were viable (i.e., stainable), but only the almost mature ones (morphologically) of specimen A actually germinated. L.M.E. thanks Fundagao de Amparo a Pesquisa do Estado de Sao Paulo for grant 80/462:2. The authors thank Dr. L. Sodek for correcting the English and Mrs. Luzia M. Kiyono for the typing. LITERATURE CITED BaRRINGTON, D. S. 1978. A revision of the genus nari a Conte: paid meine — Bew ey, J. D. and M. BLack. 1978. Physiology J 8 Berlin: Springer-Verlag. BRADFORD, M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochem. 72:248-254. Cran, D. C. 1979. The ultrastructure of fern gametophyte cells. In The experimental biology of ferns, ed. A. F. iach London: Academic Press. Cruz, M. - V. and T. S. MELHEM. 1984. Estudos polinicos em Sapindaceae. Revista Brasil. Bot. 7: -25. Dyer, A. : 1979. The culture of f tali tigation. In The experimental biology of ferns, ed. A F. Dyer. ‘London: Academ ERDTMAN, G. 1969. Handbook of palynology rycen ae tampon An introduction to the study of pollen grains and spores. Copenhagen: M ~—— and P. Sorsa. 1971. Pollen and spore morphology/plant Ends Stockholm: Almqvist & Wiksell. Gastony, G. J. 1974. Spore morphology in the Cyatheaceae. I. The perine and sporangial capacity: general considerations. Amer. J. Bot. 61:672-680. — 1979, ae morphology in the Cyatheaceae. III. The genus Trichipteris. Amer. J. Bot. 66: 1238-12 Gad R. roi Paves, 1976. Spore morphology in the Cyatheaceae. II. The — Lophosoria, Metaxya, Sphaeropteris, Alsophila and Nephelea. Amer. J. Bot. 63:738 litidis L. Pl. Sci. Lett. 9:3 - 1982. Effect of red light ps gibberellic acid on lipid metabolism in germinating spores of K fern Anemia ioc Physiol. Pl. (Copenhagen) 54: »O. W. on WASAKI. 1969. The spores of the pteridophytes. Tokyo: Hirokawa Publishing Compan Manton, I. 1950, pat ale of cytology and evolution in the Pteridophyta. Cambridge: Cambridge University Press. Manco NDES-FERREIRA, W. M. and G. M. FELIPPE. 1984. eee 8ermination of spores of Cyathea delgadii. Revista Brasil. Bot. 7:53 MELHEM, T. §, and M. E. R. Matos. 1972. Variabilidade de i pn de eee Crassipes Benth—Labiatae. Hochnes 2:1-10. Miter, J. H. 1968. Fern terial. Bot. Rev. (Lancaster) 34:361-440. ta 102 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985} Morse tl, M. oi 1974. eng — en hibridos interespecificos del genero Blechnum L. Bol. Soc. Argent. Bot. 1 Rizzini, C. T. 1973. — in atl of Annona crassiflora. J. Exp. Bot. 24:117-123. SALGADO-LABOURIAU, M. L., P. E. VANZOLINI, and T. S. MELHEM. 1965. Variation of polar axes and equatorial ‘dunce | in pollen grains of two species of Cassia. Grana hes fa 6:166-176. SNeDEcOR, G. N. 1962. Statistical methods. Ames: The Iowa State University Pre Tryon, R. M., JR. 1970. The classification of the Cyatheaceae. Contr. Gray Herb. “a 53. WopeHousg, R. P. 1935. Pollen grains. Their structure, a and significance in science and medicine. New York: McGraw-Hill Book Com SHORTER NOTE New Locations for Isoétes tegetiformans in Georgia.—Since its discovery in one pool on Heggies Rock, Columbia County, Georgia, in 1976, and its descrip- tion in 1978 (Rury, Amer. Fern J. 68:99-108), fourteen more pools of I. tegetifor- mans Rury have been found in Columbia, Greene, Hancock, and Putnam Coun- ties, Georgia (Fig. 1). James R. Allison of Lawrenceville, Georgia, discovered these populations at six new sites during 1978 and 1979, while exploring some 150 flatrock exposures in four states. In May 1978, Dr. James G. Bruce and I visited the first of these new localities (Greene Co.) with Mr. Allison and confirmed the presence of I. tegetiformans in five different pools. In a December 1979 letter to me, Mr. Allison reported —— GREENE 33°32" 32" 4 82°15'05" W 1 ot = {soétes tegetiformans in Georgia. Distribution map modified from J. R. Alec's new localities; numbers in parentheses s indicate the cies of populations (pools) saat ne a sven locality. Solid otiwlos J granite la ‘Scbichedl SHORTER NOTE ve Fics. 2 and 3. Tsoétes ee 2, Type population at Heggies Rock. Southern aspect of this soars pool shows the low rim or “spillway” that minimizes soil and water retention. Only the north — of the “peg is seloediaa by mats of Polytrichum, Selaginella tortipila A. Braun and Andropogon hice = Jan 1978). 3, The most densely populated pool at the Greensboro flatrock locality. astern aspect of pool nd gr 2, Note the lack of encroaching & mosses and grasses along the rim. (Jim Allison and Jim Bruce, left to right, 13 May 1978.) 104 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 3 (1985) having found another nine pools of I. tegetiformans from five new flatrock sites in Columbia, Hancock, and Putnam Counties, and sent specimens that confirmed his discoveries. Since then, I have been unable to locate Mr. Allison, who to my knowledge has not published these data. Therefore, this note was prepared for the benefit of others interested in the distribution of this unique, mat-forming Isoétes. Isoétes tegetiformans is restricted to shallow, summer-dry pools on porphyritic granitic flatrocks. Moreover, populations may occur sympatrically with the more widespread I. piedmontana (Pfeiffer) Reed, which typically is confined to deeper soils in seepage areas and in open woodland communities along the margins of these same flatrocks. Photographs of the type population at Heggies Rock (Fig. 2) and a pool of I. tegetiformans at the Greensboro flatrock (Fig. 3) are presented to aid its recognition in the field. The following locations and the distribution map (Fig. 1) for all known pop- ulations of I. tegetiformans are based primarily on Mr. Allison’s 1979 letter. Grorcia: Columbia Co.: two pools at quarry site near Little Kiokee Creek, S of Ga. Rte. 232 and ca. 2.4 km S of Heggies Rock, Appling, 106 m, Allison et al. 949 . one pool on flatrock ca. 1.07 km W of Little Kiokee Creek site, elev.?; Two pools on flatrock ca. 2.14 km W of the Little Kiokee Creek site, elev.?. Greene Co.: five pools at “Greensboro flatrock” ca. ee km SSE of Greensboro and SE of the junction of State Road 925 and County Road 56 (33°28'30’N, 83°08’W), 185 m. Allison et al. 2 (GA); Allison 647 (GA); Jones et al. 22868 (GA). Hancock Co.: two pools on granite flatrock * pi Boone property, at end of dirt road off Ga. Rte. 15 ca. 7a hen BE of parte elev. ?, Allison (NCU); Alli Murphy 1310 (GA). Putn n Lake part ca. 16 km SW of the Greensboro flatrock site and ca. 1.28 km W of | the junction of Greene, Hancock, and Putnam Counties, 139 m Allison & sss 1132 (GA). PuiLtip M. Rury, Harvard University Herbaria, 22 Divinity Avenue, Cam- bridge, MA 02138. REVIEW “Guide de fougéres et plantes alliées,” by Rémy Prelli. 1985. 199 pp. Paris: Lechevalier. ISBN 2-7205-0516-1. (hardbound). A general discussion of the biology and classification of ferns and fern allies, followed by an illustrated floristic account of French pteridophytes. INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable), including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 812 by 11 inch paper of good quality, not “erasable” paper. Double space manuscripts throughout, including title, authors’ names and addresses, text (including heads and keys}, literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures}. Arrange parts of manuscript in order just given. Include author's name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. he S.1. (metric) units for all mea- sures (e.g., distance, elevation, weight) unless quoted ‘ source {e.g., specimen citations). For nomenclatural matter fi. e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- _ breviate titles of serial publications according to Botanico-Periodicum-Huntian- — um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). _ References cited only as part of nomenclatural matter are not included in lit- erature cited. For shorter notes and reviews, put all references parenthetically ; in text. Use Index herbariorum (Regnum Veg. 106: 4-452. 1981) for — a of herbaria. Illustrations should be proportioned to fit page: width with caption « on ‘de same : 3 nie. tone o page. Provide margins of at least 25 mm on a illustrations, design originals for ithout reduction by a uniform. — Fe ee : amount. In composite blocks, abut edges , of adi ent photographs Avoid com- : ‘continuous-tone and line-c le illustrations or blocks. | Coordinate sequence oe pes of figures (and of tables) — citation in text. oe scale and eres to aia and longitu de in ez oe _ Proofs and reprint order forms are sent to we iste the printer. Authors: > oS — send corrected proofs. to the editor and seems orders t to the e printer. AMERICAN FERN Mtr October-December 1985 JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Relationships between Ecological Pattern and Branching in the Tree ern Lophosoria quadripinnata in Veracruz, Mexico Paul Alan Cox aly P. B. Tomlinson Ontogeny of the C i f Equiset Richard L. Hauke Wild Gametophytes of Equisetum sylvaticum Jeffrey G. Duckett Osmunda cinnamomea forma frondosa at Virginia Charles R. Werth, Melanie L. Haskins, and Akke Hulburt _ Shorter Notes Gametophytes of Equisetum giganteum Richard L. Hauke Marsilea macropoda New to Louisiana Garrie P. Landry and Samuel W. Holder Review EACLE Schelpe (1924-1905) 133 2s The American Fern Society or 1985 TERRY R. WEBSTER, Biological Sciences isipsnatly University of Connecticut, Storrs, CT 06268. ident FLORENCE 8S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109 Vice-Preskleat W. CARL TAYLOR, Milwaukee Public — Milwaukee, WI 5 Secretary JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, os eiyiie: TN 37916. Treasurer DAVID S. BARRINGTON, Dept. of Botany, thes of Vermont, Burlington, VT 05405. ecords Treasurer ALAN Dept. of Botany, University of California, oo. CA — Journal Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, DC Memoir Editor DENNIS Wm. STEVENSON, Dept. of Biological Science, Barnard sae Columbia University, New York, NY 10027. Fiddlehead Forum Editor American Fern Journal EDITOR MUA Obie ee. Dept. of Botany, pigeon vn California, ry, CA 94720 ASSOCIATE EDITORS GERALD J.GASTONY............. Dept. of Biology, Indiana University, Bloomington, IN 47401 Speer PARTI ces Dept. of Botany, University of Kansas, ae wrence, KS 66045 DAVID B. LELLINGER oa ........ US. Nat'l Herbarium NHB-166, Smitheontan inal : oo Washington, DC 20560 TERRY R. WEBSTER .... Biological Sciences Group, University of Connecticut, Storrs, CT 06268 ie “American Foss A apinceel {ISSN 0002-8444) is an illustrated quarterly devoted to the general . ed by the American Fern Society, and published at the Pringle H Univers of Vermont Burlington, VT 05405. Second-class postage paid at Burlington, VT, and additional entry point. Claime fe toa} ; ty ai. PRS Sr pr, tiie Uk ens Sintes ewes mde Smuts Gomes) ‘aes ts etcieasy. "cology 1h RES a he CEE Se ue £ membership should be sent to the Records Ty . Sere rn a z A criptions gross, $11.50 net if paid through an sa anno Canoes fee $0.50); vont he oS members of the American Fern Society (annual dues, $10.00 + $4.00 mailing surcharge beyond USA. Canada, and Mex: ie membership $200.00 00). | volumes 1910-1978 $5.00 to $6.25 each: single back numbers of 64 pages or less, $1.25; 65-80 rts $200 ech tn Sip plus shipping. Back volumes 1979 et sea. met oe a0 satiate poco ozs ne erat ramen mbar Z or . Ans ; es Pies ae $ Deo es 7 American Fern Journal 75(4):105-110 (1985) Relationships between Ecological Pattern and Branching in the Tree Fern ia quadripinnata in Veracruz, Mexico L a PAUL ALAN Cox Department of Botany and Range Science, Brigham Young University, Provo, UT 84602 P. B. TOMLINSON Harvard Forest, Harvard University, Petersham, MA 01366 Pteridophytes often play important roles in determining the nature of plant communities (Watt, 1947). This role is particularly pronounced in tropical cloud forests where the diversity and variety of growth forms in ferns often is very great. In this paper we describe an investigation between the relationship of branching and patterns of shoot distribution in a tree fern. We chose to study the distribution pattern of shoots of Lophosoria quadripin- nata (J. Gmelin) C. Chr. (Lophosoriaceae), a small tree fern that occurs throughout the Neotropics (Tryon & Tryon, 1982), and attempted to relate the observed pattern of shoot distribution to the morphology and branching pattern of the species. The shoot morphology of this species has been described superficially by earlier authors. Bower (1926) illustrated a short segment of the axis and showed that the aboveground stems arise from a rhizome. Lucansky (1982) limited his comparison of the anatomy of Lophosoria and Metaxya to the erect stems. We therefore sought to understand the very prominent role Lophosoria quadripin- nata plays in cloud forest vegetation by futher investigating its morphology and ecology. METHODS Previous workers (Tryon et al., 1973) had made extensive maps of pteridophyte vegetation occurring in the cloud forest at Pas de Enriques in the state of Ve- racruz, Mexico, an area that occurs on the steep slopes of the escarpment of the Sierra Madre Oriental at an elevation of about 1350 meters. The shoot distri- bution of Lophosoria quadripinnata at Pas de Enriques was initially studied by mathematical analysis of these maps; subsequently, field investigations of Lo- Phosoria quadripinnata were made and material for morphological analysis was Collected. The original maps compiled by Tryon et al. (1973) were sampled with a paper quadrat scaled so as to yield data comparable with a transect of 32 contiguous Meter-square quadrats. In each quadrat, the presence or absence of Lophosoria quadripinnata was recorded. Ten transects of this type were analyzed for spatial “eTogeneity of shoots using a one-dimensional pattern analysis, a variation of the two-dimensional technique described by Greig-Smith (1957). First, a variance Was calculated for the original transect by using the formula 106 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) y x - #) ae ss Bat where x is the value for Lophosoria quadripinnata (1 for present, 0 for absent) in the ith quadrat, x is the mean value for all quadrats, and n is the number of quadrats. Since this variance is calculated for samples taken using a meter- square quadrat, the value of the variance is plotted against a 1 meter block size. The 32 quadrat transect is then reanalyzed by considering it to represent 16 contiguous 2 meter by 1 meter quadrats. Presence and absence are then recorded for each of these new rectangular quadrats, and a new variance calculated. The value of this new variance is then plotted on the same graph against a 2 meter block size. The original 32 quadrat transact is then considered to consist of 8 contiguous 4 meter by 1 meter quadrats. Presence and absence are calculated for each of these 4 meter by 1 meter quadrats, and the new variance is then plotted against an 8 meter block size. This procedure is then repeated for a 16 meter block size. The resulting graphs (Fig. 1) yield significant information con- cerning shoot distribution, for a peak in the graph indicates significant spatial heterogeneity or pattern at that particular block size (Greig-Smith, 1957). Subsequent to the analysis of the maps, field studies of a population of Lo- phosoria quadripinnata were conducted at Pas de Enriques. Erect shoots were not found to be uniformly distributed in the vegetation, but instead were grouped into distinct clumps or patches, each clump appearing in the vegetation as 4 dense mass of leaves. The maximum dimensions parallel and perpendicular to the fall line for 10 such patches were recorded. In addition, a large patch was completely excavated and a map of the rhizome system indicating all erect and horizontal shoots was drawn. The unearthed rhizome system then was sectioned, color-coded for reassembly, washed, photographed, preserved in FAA, and ana- lyzed in the laboratory. Approximately two weeks later the rhizomes were sequentially sectioned and filmed in frame-by-frame analysis. Each sawn surface was photographed by means of a cine camera after it was planed on a radial-arm saw in a modification of the basic cinematographic technique described by Tomlinson (1970) and Zim- mermann (1976). The resulting motion picture was analyzed for details of mor- phology, including branching of axes, and anatomy, including stelar patter? characterizing leaf and branch insertion. RESULTS Eight of the ten transects indicated significant ecological pattern as show? in Figure 1 which consists of graphs of transect variances plotted against block sizes. The mean block-size at which significant pattern occurs was found to be 4.25, meters. Of the remaining two transects, one did not exhibit pattern at any block — size for Lop hosoria and the other did not include any Lophosoria trunks. _ The rhizomatous growth of Lophosoria appears to be one of the major factors : generating the observed ecological pattern. Although large patches measuring — COX & TOMLINSON: BRANCHING IN LOPHOSORIA 107 variance 16 124 block size (m) pe 1. Analysis of pattern of Lophosoria quadripinnata at Pas de Enriques, Veracruz, Mexico, for ight ransects. For each graph block size is plotted along the x axis while variance is plotted against axis, Up to 19 meters in diameter and occupying over 280 square meters were record- ed, the mean diameter of a Lophosoria patch as measured in the field (Table 1) . 9.97 meters, a value that would be detected in the pattern analysis as a peak at a block size of either 4 or 8 meters. Our morphological analysis confirms that L. quadripinnata has a dimorphic shoot system consisting of underground plagiotropic axes (rhizomes) that bear 108 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) TABLE 1. Diameters of Lophosoria quadripinnata Clones on a Vertical Projection. Mean = 5.97 m; s.d. = 5.18 m. Clone # Diameter CO ON OOF WN FB S bo SJ 3 3 co only scale leaves and aboveground orthotropic axes (trunks) that bear foliage leaves (Fig. 2). The scale leaves have no indication of a rachis or a blade. Each rhizome originates as a branch from an existing rhizome, grows horizontally for an extended period, and then becomes erect as a leafy axis that emerges as the visible trunk of the fern. The sympodium is proliferated by branch rhizomes that arise at intervals along the parent rhizome. Branches develop from buds located abaxial to the scale leaves on the rhizome, but the frequency of bud development does not follow any regular pattern. However, branches occur most often at the distal end of rhizomes in the region where the axis becomes ortho- tropic. The erect aerial stem does not itself branch. Branch and leaf insertion can be distinguished readily in transverse section. The leaf trace of a scale leaf makes a characteristic U-shaped leaf gap in the vascular cylinder of the stem, while the branch trace forms no gap, the branch being completely solenostelic from its level of initiation. Phyllotaxis in the rhizome is spiral with a mean angular divergence of 135.4 degrees (n = 69, s.d. = 28.5) but may change in axes that exhibit transition from plagiotropic to orthotropic growth. DISCUSSION The analysis of pattern is a valuable technique for detecting non-randomnes in the spatial arrangement of vegetation. The determination of the scale oF block size at which significant pattern occurs allows hypotheses to be made concerning factors that control that pattern. In our study of Lophosoria quadripinnata, found significant pattern to exist at a mean block size of 4.25 meters. Since ne features of the terrain or soils of the site were known to exhibit a pattern at! al level and since none of the other pteridophyte species recorded by Tryon “G (1973) showed pattern at that level, we hypothesized that exogenous factors wel? not responsible for the pattern detected in L. quadripinnata. Our hypothess- instead was that an endogenous factor, specifically a rhizomatous gro habit : was responsible for the detected pattern. However recent workers (Lucansky: : 1974, 1982; Lucansky & White, 1974) had not studied decumbent axes in L. quad: : COX & TOMLINSON: BRANCHING IN LOPHOSORIA 109 as FhHizome ) erect shoot . A7 35 O 36 2 22 4 1 | 54cm ASUSUGRGSOGEOGSEUSSODCRORORGRUSSOR SESE AOSRESEST ORE SSEDEAATOSSORSOCIOSOREOROSSOSOSSOAOROS EAGAN SEAS eRea TERT = 3 oO: + LA 5 ~ % Saute) Pd a ~, A - % = CY oe =: es 3 = .S “i ae €) = AS = Pog aS s re Fic. 2. Diagrammatic map of rhizome system excavated of Lophosoria quadripinnata. Distances given in centimeters. Angles of branching st stylized and do not rey t actual diverg branches. ripinnata although Bower (1926, p. 286; 1912, p. 293) was aware of their existence. Therefore field studies were necessary to test our hypothesis that rhizomatous growth is a prominent feature of L. quadripinnata growth. Our studies confirmed the existence of a rhizomatous growth pattern in L. quadripinnata while the measurement of patches revealed patch size to be consistent with the scale of pattern detected in our pattern analysis. It therefore appears likely that rhizoma- tous growth is the major factor determining ecological pattern of Lophosoria quadripinnata at Pas de Enriques. However it is also possible that patch size and pattern reflect some unknown feature of sexual reproduction (Conant & Cooper-Driver, 1980); this alternative cannot be tested with our data. Rhizomes are known from other tree fern genera such as Cyathea sternbergii Pohl (Brade, 1971) and Alsophila marriana (Hook.) Tryon (Hallé, 1966} while Cyathea parvula (Jenm.) Dunim produces negatively geotropic branches that root when they reach the soil (Tryon & Tryon, 1982). It is not known how these features affect their ecological pattern. : Pattern analysis has not been used previously in ecological studies of pteri- dophytes with the exception of Anderson’s (1961) work on Pteridium aquilinum in North Wales. That study is notable in that it examines the overall pattern of clonal development of fern rhizomes (cf., Bell & Tomlinson, 1980). Certainly Pattern analysis has proven useful in our study of Lophosoria quadripinnata and 110 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) we commend its use to ecologists and morphologists as a means of generating hypotheses concerning the relationship of ecology and morphology in ferns. ACKNOWLEDGMENTS We thank R. Tryon and A. Tryon for useful discussions and two anonymous reviewers for useful criticisms of our manuscript. M. Moore assisted in data collection. During this study Cox was funded by a Danforth Fellowship and a National Science Foundation Graduate Fellowship. Field work in Mexico was supported by the Atkins Garden Fund, Harvard University. LITERATURE CITED ANDERSON, D. 1961. The structure of some upland plant communities in Caernarvonshire. I—The pattern shown by Pteridium aquilinum. J. Ecol. 49:369-377. Be.t, A. D. and P. B. oe 1980. Adaptive architecture in rhizomatous plants. J. Linn. Soc., Bot. 80:125- Bower, F.O. 1912. ihe in the phylogeny - the hn eg IV. Lophosoria and its relation to the Cyatheoidea and other ferns. Ann. 2 : og a Jorns: Vol. II. London: Cambridge University Press. Brabe, A.C. “forma” de Cyathea sternbergii Pohl. peti 10:73-76. Conant, D. S. and G. Cooper-Driver. 1980. Autogamous allohomoploidy in Alsophila and Nephe- E speciation in homoploid homosporous ferns. Amet. Bot. “67: 1269-1288. J. P. 4057. ——— pant eapleny. Jondon: ater wort: HALLE, F. 966. Bovis } tes. Adansonia “Oo foes 4. peer T. W. 1974. Comparative studies of the nodal and ase anatomy in the neotropical yatheaceae. I. Metaxya and Lophosoria. Amer. J. Bot. 61:472-480. : ae Anatomical studies on the neotropical Cyatheaceae. us Metaxya and Lophosoris. Amer. Fern J. 72:19-29. ——— and R. A. WuiTe. 1974. Comparative studies of the nodal and vascular anatomy in the Egypt hypo Cyatheaceae. III. Nodal and petiole patterns: summary and conclusions. Amer 61:818- TOMLINSON, P. B. 1970. generere towards an understanding of their morphology and anat- oo ances in botanical research, Vol. 3, R. D. Preston. Lor ic Academ: Tryon, R., B. VOELLER, A. TRYON, and R. Ripa. 1973. Fern biology in Mexico. BioScience 23:28-32. ~—— and A. Tryon, 1982. Ferns and allied plants, with special reference to tropical America. New York: Springer-Verlag eee ina piece hen es the ecology of bracken. IV. The structure of the community. New 86 ud M.H. 1976. The study of vascular patterns in higher plants. Pp. 221-235 in Transport ~ T processes in plants, eds. I. E. Wardlaw and J. B. Passioura. New York: Academic American Fern Journal 75(4):111-119 (1985) Ontogeny of the Commissure of Equisetum RICHARD L. HAUKE Department of Botany, University of Rhode Island, Kingston, RI 02881 The genus Equisetum is characterized by having whorled leaves fused into a nodal sheath. This sheath consists of a variable number of segments, or leaves, three in E. scirpoides to 50 or more in E. giganteum. Each segment has a single vascular bundle, continuous with a vascular bundle of the internode below it, and is terminated by a tooth, or free leaf tip. Between adjacent segments there is an anatomically distinct commissure. This commissure was first described by Duval-Jouve (1864) as consisting of transversely elongated cells at the bottom of the commissure bordered by oblique to irregular cells, as opposed to the vertically elongate cells of the sheath epi- dermis. Milde’s (1867) description was a little more complete. There is a row of transversely elongated cells lying one above another, which have strongly thick- ened walls. Their ends bend a little downward, giving them a half-moon shape. Neighboring cells overlap them, and this overlapping may be so strong as to completely cover the cells of the commissure. Miiller (1888) provided an ex- tended discussion of the commissural appearance. He named the transversely oriented row of cells “anchor cells” (ankerzellen) because they anchor the ad- jacent sheath segments. With the adjacent obliquely oriented cells they form parabolic curves he compared to “chain lines” (kettenlinien), which are formed by suspending a chain between two upright supports. With descriptions, math- ematical formulae, and polarized light analyses of cell walls, he supported his theory that the characteristic form of the commissure is the result of rapid elon- gation of the sheath teeth stretching the still meristematic cells of the commis- sural region. He described the appearance of the commissure in cross-section and the vari- ation between species. The anchor cells have swollen ends and thick walls. They are underlain by small, thick-walled cells of the inner epidermis. In some species the commissure is shallow, in others deep, and in some the sides of the com- missure overhang the furrow and partly to completely occlude it. ; Mature anchor cells bend at the ends transversely outward. The wall is thick, but thinnest at the bend. According to Miiller this makes th i stre able. The sheath functions to protect the growing point and developing sheaths, and its stretchability allows them to elongate through it, and provides flexible Protection and mechanical strength for the intercalary meristems at the base of internode above the sheath. Miiller (1888) described the ontogeny of the anchor cells in E. hyemale. At commissure the young cells are 2-4 layers thick, descended from one or two Meristem cells. Over them lies a single, large, somewhat 6-sided cell, which belongs to the outer epidermis. The inner wall is about twice as long as the outer wall. The lateral neighbors of the anchor cell are meristematic and divide, so 112 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) that the sides overgrow the anchor cell, until the commissural groove is a mere slit (Fig. 15). The outer wall of the anchor cell becomes as big as the inner wall. Just under the anchor cell a cell is cut off, either from the underlying cells or from the anchor cell itself, which may divide once or more, and which forms a pillow for the anchor cell. Finally the ends of the anchor cell grow out and gradually bend around, forming the anchor shape. Then the walls become thick, and thus is established the deceptive appearance of an anchor cell lying deep in the inside of a closed tissue complex. The nucleus of the young anchor cell is larger than those of the adjoining cells and is elliptical. As the cell ends grow out, the nucleus remains right in the middle of the lengthwise axis, coming to lie against the outside wall and becoming spindle-shaped. According to Miiller the rhizome sheaths have no such distinctive commissure. Schellenberg (1895) disagreed with Miller, and attributed the characteristic appearance of the commissure to unequal growth at the base rather than teeth enlargement. He pointed out that the teeth have elongated before the commis- sure forms. The lower part of the sheath segment is more actively growing than is the lower part of the anchor cell row. Upper cells of the anchor cell row accommodate this unequal growth by stretching into parabolic curves. Sadebeck (1902) described the establishment of the anchor cell row by division from the neighbor cells. As the sheath grows, it forms new anchor cells, but only from below. He cited Miiller’s explanation of tooth growth causing stretching of cells. More recent authors have not described the commissure. DeBlock (1923) speaks of the sheath segments being “concrescent,” Eames (1936) describes the leaves as ‘fused laterally,” Smith (1955) says they are “laterally united,” Bierhorst (1971) calls them “connate,” and having “‘a single row of hook-shaped cells that seem to clamp the adjacent leaves together,” Foster and Gifford (1974) say they are “united,” and Bold et al. (1980) “fused.” I (1979) observed variations in appearance of the commissures of species of subgenus Equisetum, and have since studied the anatomy and ontogeny of the commissure of Equisetum. The distinctive row of cells in the middle of the commissure will be called the C-cells here, short for commissure, rather than anchor cells, which implies a function which may be misleading. MATERIALS AND METHODS Buds of Equisetum arvense, Equisetum diffusum, E. hyemale, and E. telmateia were killed and fixed in FAA, transferred to 70% EtOH, dehydrated in a tertiary butyl alcohol series, embedded in Paraplast, (mp 56-57°C) and sectioned on 4 rotary microtome at 10 wm (longitudinal sections), or 20 um (transverse sections}. These were stained with aqueous ferric chloride and tannic acid (the reverse of the usual sequence of Foster, 1934), counterstained with .01% toluidine blue ™ 95% EtOH, and mounted in Diaphane. Some fixed buds were cleared with 10% KOH followed by chloral hydrate, stained with tannic acid and ferric chlorid®, teased apart in Diaphane and mounted on slides. The photos were taken Kodak Plus-X or Technical Pan film with a Nikon Microflex UFX camera mount: ed on a Zeiss microscope. Sources of plant material are as follows: R. L. HAUKE: COMMISSURE OF EQUISETUM 113 Equisetum arvense.—Moat in front of main library, University of Rhode Is- land campus, Kingston, 31 Aug 1974 (Voucher: Hauke 514, KIRI). E. diffusum.—Pot in University of Rhode Island, Botany Department green- house, Kingston, 10 Jan 1985. Plant originally collected near Simla in Himchal Pradesh state, India (Voucher: Hauke s.n., KIRI). E. hyemale.—Northeast corner of Galilee Escape Rd. and Great Island Rd., Narragansett, Rhode Island, 31 Dec 1984 (Voucher: Hauke 515, KIRI). E. telmateia.—South side of Claremont Ave., just E of Alvarado St., Oakland, California, 4 Dec 1980 (Voucher: Hauke C1, KIRI). RESULTS From the cleared material of E. diffusum, it was apparent that the sheath teeth enlarge and mature before there is much development of the lower portion of the sheath (Fig. 1). Such clearings also showed dramatically the mature form of the commissure, with its row of C-cells and oblique lines of neighboring cells (Figs. 2, 3). Longitudinal sections confirm what the clearings suggested, that the three youngest sheaths are primarily growing by tooth development, that the fourth and fifth youngest sheaths show rapid growth of the basal, fused part of the sheath and the commissures, and that the sixth and older sheaths have maturing or matured C-cell rows. The youngest sheaths do not show any median line of cells (Fig. 4). The earliest appearance of a median line is 4 large cells appearing approximately square in tangential sections (Fig. 5). The number in- creases until there are 10 or more square to slightly rectangular cells (Fig. 6). In cross-section of the sheath (Fig. 7) these cells can be seen as trapezoidal, so that an abaxial longitudinal section would appear square, whereas the adaxial sec- tion would appear rectangular. I could not determine whether the C-cell row originated by one or more of the pre-commissure sheath cells assuming a more median position, by longitudinal division of pre-commissure cells to create a median line of cells, or in some other manner. Once the row is present, the number of cells in it greatly increases by transverse division (Fig. 9) and the cells ome more rectangular to tangentially elongated. That the increase in number of C-cells is by transverse division throughout the row rather than by basal initial activity is also indicated by the relation between the C-cells and the adjacent cells. At the cuboidal stage (Fig. 6), the two are about the same height. In the rectangular stage there are often two or more C-cells for each neighbor cell (Fig. 8). In the tangential stage, three or more C-cells may abut a single neighbor (Fig. 10). Whereas the G-cells are dividing only transversely, the neighboring cells are dividing obliquely (Fig. 8, arrow) and thus establishing the oblique orientation of the adjacent cells. Adaxial to the C-cell row is an inner epidermis of vertically elongated cells. There may be two or more side-by-side under a given C-cell (Fig. 6, ar row), and each may be longer vertically than two C-cells (Figs. 11, 12). As the C-cells mature, they bend outward at the ends (Fig. 13) to achieve the shape of a letter C in transverse section. This bending causes a deep furrowing of the commissure in E. diffusum (Fig. 3). AMERICAN FERN JOURNAL: VOLUME 75 NUMB ER 4 (198 9) Fics. 1 . 1-3. E . quiset : precocious t um diffusu eeth el iJusum, cleari a ongation. Scale bar a of sheaths. Fic. 1. B 0 um. Fic. 2. aan Hg — a gt ihe of C-cells and adj i jacent oblique ] furrow: I* e ba * oO Ss Wi fi Scale ba r= 200 C-cell r um. Fic. 4 pe al dor sa pee ame pex of the sh m diffusum eath is tow , tangential i ard the lef section, young sh 2 left. Scale ba g sheath before initia r= 20 pm. R. L. HAUKE: COMMISSURE OF EQUISETUM 115 Fic.5. Equisetum diffusum, tangential section, earliest visible C-cell row, of 4 ae cells. Scale = 20 um. Fic. 6. Equiesiom: diffusim, ene tel 6 potion, C-cell row of more than 10 rectangular cells. aw indicates the Is, taller than the C-cells and two abreast Scale bar = 50 um. Fic ime tam ane mae, eroam-enttion. sheath with trapezoidal C-cell (arrow) underlain by adaxial a cells. Seale ber = 50 um. Fic. 8. Equisetum telmateia, tangenti Section, C-cell row of a rectangular cells. Arrow marks oblique division figure in a neigh- boring cell, Scale bar = AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) ie bar = : : : ae : Fic. 9. Equisetum telmateia, tangential section, C-cell undergoing transverse division. = sc 20 um. Fic. 10. Equisetum arvense tangential section, commissure with C-cells tangentially e pate Fics. 11, 12. Equisetum diffusum, tangential section, two optical sections : erti- the C-cell row of tangentially elongated cells underlain by adaxial epidermal cells which are V' cally elongated and narrower than the C-cells. Scale bar = 20 pm. R. L. HAUKE: COMMISSURE OF EQUISETUM 117 ys Fic. 13. Equisetum diffusum, cross-section, C-cell with prominently bent ends. Scale bar = 20 um. Fic. 14. Equisetum telmateia, cross-section, commissure with adaxial epidermis proliferated by peri- clinal division. Scale bar = 20 um. Fic. 15. Equisetum hyemale, cross-section, commissure of C-cells occluded by growth of neighboring cells. Scale bar = 50 um. Fic. 16. Equisetum diffusum, tangential Section, rhizome commissure showing less-ordered structure. Scale bar = 80 um. . The ontogeny described here for the commissure of E. diffusum was also seen in E. telmateia. In that species, however, periclinal division of the adaxial layer (Fig. 14) in the basal part of the commissure made it multilayered. Equisetum arvense, also studied for comparison, does not develop as deeply furrowed a commissure as do the above two. Because of its occluded commissure, E. hye- male does not show the distinct C-cell row with oblique adjacent cells charac- teristic of E. diffusum, and was not studied developmentally. At maturity (Fig. 15) it does appear to be occluded by overgrowth of adjacent cells, as described saree (1888) and to have formed “pillow” cells by periclinal division of the -cell. 118 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) The nucleus of the C-cell is particularly noteworthy. It is always prominent and remains in the center of the cell (Figs. 8, 9, 14). As the cell changes from cuboidal to rectangular to tangential, the nucleus becomes more extremely el- lipsoidal. As the cell folds outward, the horizontally attenuated nucleus remains in the center portion (Fig. 13). The distinctive commissures described are from aerial stems, or branches. The rhizome sheath commissures are apparently much less regular in their devel- opment (Fig. 16), for the C-cells do not form a single distinct, uniform row, nor do the adjacent cells form “chain lines.” DISCUSSION Although Miiller’s (1888) description of the commissure, with its distinctive median row of abaxial epidermal cells, the C-cells (his ankerzellen) was accu- rate, his “chain line” explanation for the oblique appearance of adjacent cells by stretching caused by growth of the sheath teeth, was not. Schellenberg (1895) attributed them to unequal growth at the base, which is more nearly correct. Actually, the characteristic appearance is the result of control of plane of divi- sion. In the median row, cell division is always transverse, and there is little subsequent elongation of daughter cells. Hence from a row of a few initially cuboidal cells is produced a longer row of rectangular cells, and eventually a very long row of tangentially elongate cells. This transverse division activity occurs throughout the C-cell row, and there is no basal initial, as Sadebeck (1902) implied. The cells adjacent to the commissural row divide in various plants, a yee obliquely, to produce the parabolic curves with C-cells at the apex. ctive outward bending of the C-cell, and the equally distinctive ap- Rec and position of the nucleus cannot be explained here. Since the sheath grows as a unit from an initial ring produced just below the apex, descriptions of the sheath segments as “‘concrescent,” “fused,” or “united” are misleading. These terms could be understood to indicate ontogenetic rather than phylogenetic fusion. The term “connate” is preferable, since it means lit- erally “born together” and indicates congenital fusion. LITERATURE CITED rst, D. W. 1971. i. of vascular plants. New York: The Macmillan Co. oud HG. G5 Arpiontian cad T Daisies pe 1 ts and New York: DELEVoRYAS. 1980. Morphology of plants and fungi. romana 1923. Contribution a étude des Equisétacées. Thesis. Faculty of Medicine & Pharmacy, Roe ak J. 1864. Histoire naturelle des Equisetum de France. Paris: Bailliere et Fils. . The use of tannic acid and iron chloride for staining cell walls in me eristematic tissue. Stain Tech. 9:91-92. —— and E. se GirrorD. 1974. Comparative morphology of vascular plants. 2nd ed. San Fran- cisco: Hauke, R. L. 1979. A Goaaeale mon edwigia ograph of Equisetum subgenus Equisetum. Nova Hi 30:385-455. (1978) na ae R. L. HAUKE: COMMISSURE OF EQUISETUM 119 Mitpe, J. 1867. Monographia Equisetorum. Nov. Actorum Acad. Caes. Leop.-Carol. German. Nat. Cur. 32(2):1-605 + 35 pl. Mier, C. 1888. Uber den Bau der Commissuren der Equisetenscheiden. Jahrb. Wiss. Bot. 19: 497-575. SADEBECK, R. 1902. Equisetaceae. Pp. 520-548 in Die naturlichen Pflanzenfamilien, Vol. 1(4). eds. A. Engler and K. Prantl. Leipzig: Verlag Engelmann. SCHELLENBERG, H.C. 1895. Zur Ent juiset heiden. Ber. Deut. Bot. Ges. 13:165-174. o SmitH, G. M. 1955. Cryptogamic botany II. Bryophytes and pteridophytes. 2nd ed. New York: McGraw Hill. REVIEW “Ferns of Jamaica, a Guide to the Pteridophytes,” by George R. Proctor. 1985. vi + 631 pp. British Museum (Natural History). ISBN 0-565-00895-1. £50.00. The publication of a complete new pteridophyte Flora for a tropical country is always a major event. Jamaica, being a small island (11,424 km’; 4411 mi*) with high mountains (up to 2250 m elevation) and varied topography and soils, sup- ports an unusually large pteridophyte flora (609 taxa, 82 (13.3%) of them endem- ics) for its size. The Ferns of Jamaica is unique in that the author resided in Jamaica for many years. He made extensive field surveys and collections and saw almost all the taxa as living plants. This contrasts greatly with other Floras, which are by necessity based largely on herbarium specimens with occasional trips to the field by the authors. Thus, the habitat data given by Proctor are unusually complete and helpful, and even serve to amplify the distinctions be- tween closely related species like Pityrogramma williamsii Proctor (on calcar- eous earth banks and limestone boulders) from the similar appearing P. sulphu- tea (Swartz) Maxon (on non-calcareous earth banks). The well-written keys, brief descriptions, and habitat and frequency statements make this Flora easy and pleasant to use; one could read it cover-to-cover for the pleasure of discov- ering the many original ecological observations. Some of the species of each genus are illustrated, often by reproducing graceful line drawings found in 19th century works on ferns. The taxonomic concepts are broad at all levels, but subcategories are freely employed. The book is the capstone and monument to the productive years that Proctor spent in Jamaica, but it is by no means a culmination: he is currently preparing a Flora of the pteridophytes of Puerto Rico and the Virgin Islands, to which he has added dozens of species new to € Flora as a result of intensive fieldwork and a consummate knowledge of Antillean Pteridophyta.—Davip B. LELLINGER, Department of Botany, National eg of Natural History NHB-166, Smithsonian Institution, Washington, DC American Fern Journal 75(4):120-127 (1985) Wild Gametophytes of Equisetum sylvaticum JEFFREY G. DUCKETT School of Biological Sciences, Queen Mary College, Mile End Road, London E1 4NS, U.K. Though generally stated to be rare in nature, large populations of Equisetum gametophytes can sometimes be found on bare mud on the exposed margins of lakes and rivers in Britain. A detailed analysis of the reproductive biology and dynamics of such wild populations is given by Duckett and Duckett (1980). Apart from these littoral sites, gametophytes are occasionally seen on bare damp soil by field bryologists (Appleyard, 1981; Crundall, 1983). Elsewhere in the world Equisetum gametophytes have been recorded on bare mud, particularly that exposed on river banks, following the monsoon in the Indian subcontinent (Mo- han-Ram & Chatterjee, 1971) or seasonal rains and the melting of winter snow in North America (Walker, 1921, 1931, 1937; Mesler & Lu, 1977; Duckett & Duck- ett, 1980). The frequent occurrence of hybrid horsetails with distributions indicative of multiple origin, rather than fragmentation, and long distance dispersal of single clones (Duckett, 1979), indicates that sexual reproduction in this group of pteri- dophytes may not be as infrequent as normally assumed. The apparent rarity of the gametophytes is related to the production of short lived spores over a short period during each growing (Duckett, 1970a) and narrow habitat tolerance in situations which only appear sporadically: bare mud which is neither too nor too wet must be available for colonization when the spores are liberated. Even after establishment the gametophytes are poor competitors with bryophytes and vascular plants and are apparently excluded from the immediate vicinity of the parent plants by allelopathic compounds (Milton & Duckett, 1985). Detailed scrutiny of the literature on wild gametophytes reveals that the ma- jority of observations have been made on very few species. Only those of E. arvense L., E. palustre L. and E. fluviatile L. have been found in Britain. I have failed to find those of E. telmateia Ehrh. and E. variegatum Schleich ex Weber and Mohr despite intensive search on many occasions of apparently suitable habitats in areas of water seepage and in wet depressions in sand dunes 1 spectively. Attempts to establish wild gametophytes of E. telmateia on carefully marked sites where spores had been sown artificially were also unsuccessful (J. N. B. Milton, unpubl. data). North American records are mainly for species in subgenus H ippochaete [E- laevigatum A.Br. (Walker, 1921, 1931), E. scirpoides Michx. (Walker, 1937; Feig- ley, 1949) and E. hyemale (Mesler & Lu, 1977)] plus records for E. arvense (Wa er, 1931; Matzke, 1941; Hauke, 1967) and a single locality with E. telmateia (Walker, 1931). Descriptions from the Indian subcontinent are all referable to E. soreargenes Desf. (Mohan-Ram & Chatterjee, 1971). Eetanrrigumany E. pratense Ehrh. have never been described and ee desctiption ord for E. sylvaticum L. dates from Bischoff (1853). From om provided, plus the lack of morphological information on culture? J. G. DUCKETT: EQUISETUM GAMETOPHYTES 121 gametophytes at that time it is far from certain that these gametophytes can be referred to E. sylvaticum. Whilst collecting bryophytes in late August 1984 on the normally inundated margin of the southern shore of Loch an Daimh, Glen Lyon, Mid West Perth- shire, Scotland, Professors R. Brown, B. Lemmon (University of Louisiana) and I discovered a population of Equisetum gametophytes. Since E. sylvaticum was the only horsetail present in abundance in the vegetation surrounding Loch an Daimh, and the gametophytes differed in sexuality from those of other species of Equisetum previously seen in Britain, it seemed likely that the wild gameto- phytes were also of this species. Identification was subsequently confirmed from comparison with E. sylvaticum gametophytes grown in axenic culture. This paper compares the morphology and sexual behavior of the wild game- tophytes of E. sylvaticum with (1) gametophytes of E. sylvaticum in axenic cul- ture (2) wild gametophyte populations of other species of Equisetum. Features of the E. sylvaticum habitat are compared with those of other wild gametophyte localities. Nomenclature follows Corley and Hill (1981) for bryophytes and Clapham, Tutin, and Warburg (1962) for vascular plants. OBSERVATIONS The site.—Loch an Daimh is a reservoir constructed in the 1950s in the upper part of Glen Lyon. The elevation of the high water level is approximately 400 meters. In August 1984 the water was some 10 meters below this, having fallen steadily since the spring (D. Long, pers. comm.) during one of the driest seasons on record in the Scottish Highlands. The vegetation bordering the south shore of the Loch is referable to the Cal- lunetum vulgaris, Vaccineto-Callunetum and Molinieto-Callunetum associations of McVean and Radcliffe (1962), but with the addition of abundant Equisetum sylvaticum (20-50 stems per m2). A few stems of E. palustre and E. fluviatile were found by the outflow stream below the dam and E. variegatum and E. arvense grew in small quantities on the sides of streamlets entering the Loch. In comparison with E. sylvaticum it is unlikely that these four species make a significant contribution to the spore rain. The Equisetum gametophytes were restricted to a layer of sandy peat, pH 5.2, about 1 meter above the low water level along a 200 meter stretch of shore from grid reference 27/507462 to 27/502464. They were growing on both horizontal and vertical surfaces of the peat and around the remains of Calluna plants which predated the construction of the dam. In the immediate vicinity of the gameto- phytes only about 5% of the surface of the reservoir margin was covered with Vegetation. In consisting of a normally inundated reservoir margin, this site closely resem- bles other localities in Britain where Equisetum gametophytes are of regular occurrence. However, with one exception (Ladybower, Yorkshire with a value of 5.5) the PH at Loch an Daimh (pH 5.2) was significantly lower than at the other sites (6.0-7.5). The gametophytes were associated with 8 species of pha- 122 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) nerogams (highly depauperate compared with up to 80 at other sites), 23 liver- worts, and 39 mosses. Most abundant were Festuca ovina, Isolepis setacea, Pellia epiphylla, Jungermannia gracillima, Pogonatum aloides, Ditrichum heteromal- lum, Dicranella subulata, D. rufescens, Campylopus paradoxus, Ephemerum ser- ratum, var. serratum and Pohlia nutans. These species are indicative of a sub- strate of much lower nutrient status than that of Pennine reservoirs in Northern England (Duckett & Duckett, 1980; Duncan & Dalby, 1960). These observations on pH and associates suggest that the gametophytes of E. sylvaticum may have different ecological requirements from those of other horsetails. Gametophyte morphology and sexual behavior.—The gametophytes of E. syl- yvaticum were more or less circular cushions varying from 0.5-2.0 mm in height and dark, dull green in color. They ranged in diameter from 2-9 mm (mean 4,7 mm). All the 49 individuals collected were female. None bore sporophytes. The overall morphology of the gametophytes was identical to that of other species in the wild. Each comprised a basal cushion 10-20 cells thick bearing abundant colorless rhizoids ventrally [the morphologically similar Fossombronia fimbriata, also found at Loch an Daimh, has purple rhizoids (Paton, 1974)] and closely packed photosynthetic lamellae on the dorsal surface (Fig. 1). Archegonia (Fig. 2) are produced by the marginal meristem of the cushion and at maturity lie at the base of the lamellae. Earlier attempts (Duckett, 1970, 1973) at growing spores of E. sylvaticum re- sulted solely in production of male gametophytes. However, by using Parker medium (Klekowski, 1969) on which gametophytes grow more rapidly than on the Knop and Beijerinck agars previously employed, a sexual behavior pattern similar to that in all other species of the subgenus Equisetum is produced. Iden- tical results to those set out in Fig. 3 for spores from North America (Deerfield, Massachusetts) have also been obtained from British material. The majority of the spores grow into male gametophytes which produce an- theridia indefinitely. The remainder are initially female but subsequently pro- duce antheridia. Following flooding of the latter, sporophytes (from 1-6 per 84 metophyte) are produced by self-fertilization. Assuming that the wild gametophytes at Loch an Daimh derived from spores germinating in May, when collected in late August they would be approximately 125 days old. Thus their sexuality is very different from that of cultured game tophytes. Aside from the predominance of males, in culture virtually all the females have subsequently produced antheridia by this time. It = noteworthy that Rumberg (1932), using mineral nutrient agars as culture medium, found that female gametophytes of E. sylvaticum subsequently pre- duced antheridia at much the same rate as recorded here (Fig. 3), whereas tos? grown by Buchtien (1887) on sand and soil remained female. S Mon fev cscagees of the lamellae and archegonia of E. sylvaticum —— men igures 1 and 2. A detailed description of the antheridia is given © (1973). Lamellae in both the cultured and wild individuals of E. sylv@ ticum are of Plate type characteristic of subgenus Equisetum (Duckett, 1973) and a rea gt greatly in size and shape. However, examination . t apices which are obtusely lobed and flat predominate — ].G. DUCKETT: EQUISETUM GAMETOPHYTES 123 Fic. 1, Lamellae from female gametophytes of Equisetum sylvaticum. A-C from cultured gameto- Phytes. D-F from wild gametophytes. All x 35 except, E, x 100. AMERICAN F ERN JOURNAL: VOLUME 7 5 NUMBE R4 (1985) Fic. 2 . Arche goni gametophytes. All x : : Aagage tne : syivaticum. A, B fro m cultured gametophytes. C- E fr J. G. DUCKETT: EQUISETUM GAMETOPHYTES 125 100 q q T T T T T T T T T T 90} go | = oO ‘e) ‘al n 70 oO O QO O re) w = 60 o vegetative 5 °o male @ 50 e female Baal A female,then bisexual 40 : . 32 x bisexual, with sporophytes 30F { x 20 seca 3 10 x oe ra @) O—_ py l i 1 Ll aria 1 ade 60 80 100 120 140 160 i80 Days from sowing spores Fic. 3. Sexual behavior of gametophytes of Equisetum sylvaticum grown in single spore cultures on Parker medium (Klekowski, 1969). 300 spores, collected from Deerfield, Massachusetts, were cultured as set out in Duckett (1970). Arrows indicate the approximate age of the wild gametophytes. (cf. the funnel-shaped apices of E. arvense). The apical cells are usually short whereas those in E. telmateia and E. fluviatile are elongate. In gross morphology the lamellae of E. sylvaticum are closest to those of E. palustre. The latter species however possesses apical cells that are usually less than 30-40 um wide whereas those of E. sylvaticum mostly fall into the size range 30-60 um. In common with other members of the genus, archegonia of E. sylvaticum Possess protruding necks composed of three tiers, each of four cells. At ma- turity the cells of the terminal tier elongate considerably (up to six times their original length), separate from each other, and bend away from the neck canal. ; is maturational divergence appears to be more variable in E. sylvaticum than in other species of Equisetum. Prior to divergence the terminal neck cells are slightly twisted both in cultured (Fig. 2A) and wild (Fig. 2D) specimens. The subterminal neck cells of E. sylvaticum are much shorter than those of E. flu- Viatile and E. arvense (Duckett & Duckett, 1980). DISCUSSION On the basis of morphology, sexuality, and availability of spores in the adja- cent vegetation, there seems to be little doubt that the wild gametophytes from the shore of Loch an Daimh are referable to E. sylvaticum. Although the pe morphology of these wild gametophytes is the same as that of other species in nature, they differed in two major respects; sexuality and the absence of spo- es. Wild gametophyte populations of E. arvense, E. palustre, and E. fluviatile in 126 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) late spring comprise mixtures of male and female individuals. But, by the end of each season few or no males can be found. However, the presence of spo- rophytes on the vast majority of the females which remain attests to their earlier occurrence. Thus it would seem likely that no male gametophytes of E. sylvati- cum were ever present at Loch an Daimh in 1984. It may also be significant in this context that the wild gametophytes of E. sylvaticum (mean diameter 4.7 mm) were significantly larger than females of E. arvense, E. fluviatile, and E. palustre (all less than 3 mm) of comparable age. This suggests that conditions were par- ticularly favorable for the growth of E. sylvaticum gametophytes, a situation that would act against the production of males. On the basis of size data now available for E. arvense, E. sylvaticum, E. fluviatile, E. palustre, and E. telmateia (Walker, 1931) it appears that wild ga- metophytes of species in subgenus Equisetum rarely if ever achieve dimensions of between 10 mm and 35 mm, recorded in Hippochaete (Feigley, 1949; Mohan- Ram & Chatterjee, 1971; Mesler & Lu, 1977). Although the occurrence of wild gametophytes described here did not result in production of sporophytes of sexual origin, this incident establishes the po- tential for sexual reproduction in E. sylvaticum. At some similar site, but where E. pratense grows alongside E. sylvaticum one may readily envisage the origin of E. x mildeanum Rothm., the hybrid involving the two species in subsection Subvernalia (Duckett, 1979). When grown axeni- cally under identical culture regimes each species of Equisetum has a distinct pattern of sexual behavior (Duckett, 1970, 1972). Specific differences in the pro- portions of males and females in culture are paralleled in mixed species g@- metophyte populations found in the wild. Two hybrid horsetails, E. arvense * E. fluviatile and E. fluviatile x E. palustre have already been found to arise from such situations (Duckett & Duckett, 1980). Under the culture regime used in this investigation 27% of the spores of E. sylvaticum produced female gametophytes but the corresponding figure for E. pratense was only 8% (Duckett, unpubl. data). A similar difference in sexuality in the wild would clearly favor the production of hybrid sporophytes. LITERATURE CITED oo. J. 1981. The annual meeting, 1980, Bristol. Bull. Brit. Bryol. Soc. 37:7-14. —— ies . 1853, Bermerkungen zur Entwicklungsgeschichte der Equisetum. Bot. Zeitung (Berlin) 11:97-109. BUCHTIEN, Oo, 1887. Ratha gsg , ee Aine Prothallium von Equisetum. Biblioth. Bot. 21-49. CiaPHaM, A. R.,.T.G. Tummy, and E. F. WaRBURG. 1962. Flora of the British Isles. 2nd ed. Cambridge: Cambridge University Press. Corey, M. F. V. and M. O. Hut. 1981, Distribution of bryophytes in the British Isles. A cen pe of their occurrence in Vice Counties. Cardiff: Briti is eaclaay Duckerr, 5 ee oi sour meeting, 1982, Penrith. Bull. Brit. Bryol. Soc. 15-0 S ees . . - of the ee : y tum: ]. n. Bot. 63:327-352. . genus Equisetum, subgenus Equise . are — behaviour of the genus Equisetum, subgenus Hippocheete. J. Linn. Soc., Bot. ].G. DUCKETT: EQUISETUM GAMETOPHYTES 127 ——. 1973. eae. RELI: pa the gametophytes of the genus Equisetum, subgenus Equise Soc . 1979. i A aaakel Lae a ey teow reg and hybridization in ee ‘ipa and North American species of the genus Equisetum. J. Linn. Soc., Bot. 79:205- and A. R. DucKETT. 1980. Ba see pol biology and population dynamics of ey atic phytes of Equisetum. J. Linn. Soc., Bot. 80:1-40. DuncaN, J. E. AND M. DALBy. 1960. The Ceecten of Swinsty and Fewston reservoirs 1957 to 1959. Naturalist, Hull 874:81-88. FEIGLEY, M. 1949. An occurrence of gametophytes of Equisetum in Cheboygan County, Michigan. Amer. Fern J. 39:106-109. Hauke, R. L. 1967. Sexuality in a wild population of Equisetum arvense gametophytes. Amer. Fern J. 57:59-66. KLEKOWSKI, E. Jr. 1969. Pa biology of the Pteridophyta. III. A study of the Blechnaceae. nn. Soc., Bot. 62: McVEaNn, D. N. and D. A. Ye at 1962. Plant communities of the Scottish Highlands. A study of Scottish mountain moorland and forest vegetation. London: Her Majesty's Stationery Office. Matzke, E. B. 1941. Gametophytes of Equisetum arvense. Torreya 41:181-187. MESLER, es R. and K. L. Lu. mE Mein gametophytes of Equisetum hyemale in northern Califor- ia. Amer. Fern J. 67:9) MILTon, 1 ‘N. B. and J. G. stir 1985. Potential allelopathy in Equisetum. Proc. Roy. Soc. inburgh 86:468-469. Monan-Ram, H. Y. age - CHATTERJEE. 1970. Gametophytes of Equisetum ramosissimum subsp. ramosissim I. Sexuality and its modification. SMa T an tia! 20:151-172. PaTON, J. A. 1974. GF cabiere fimbriata sp. nov. J. Bryol. 4. RUMBERG, J. 1932. Entwicklungsgeschichte der Prothallien von rae sylvaticum und E. palus- tre. Planta 15:1-42. WaLKER, E.R. 1921. The gametophytes of Equisetum laevigatum. Bot. Gaz. (Crawfordsville) 71:378- 391. ———. 1931. The gametophytes of three species of Equisetum. Bot. Gaz. (Crawfordsville) 92:1-22. “——. 1937. The gametophytes of Equisetum scirpoides. Amer. J. Bot. 24:40-43. American Fern Journal 75(4):128-132 (1985) Osmunda cinnamomea forma frondosa at Mountain Lake, Virginia CHARLES R. WERTH Department of Botany, University of Kansas, Lawrence, KS 66045 MELANIE L. HASKINS Department of Biology, University of Richmond, Richmond, VA 23173 AKKE HULBURT Mountain Lake Biological Station, University of Virginia, Pembroke, VA 24136 Pronounced dimorphism of sterile and fertile pinnae is a conspicuous mor- phological attribute of all species of Osmunda. Sporangia are borne on the mar- gins of highly reduced segments that lack chlorophyll at maturity. The green spores initially give these fertile pinnae a greenish cast, but after release of the spores, they become wholly cinnamon-brown or gray-brown and soon wither. Of the three Osmunda species occurring in eastern North America, two bear fertile pinnae on otherwise undifferentiated sterile fronds: in Royal Fern, O. regalis L., these fertile pinnae occur at the leaf tip, while in the Interrupted Fern, O. claytoniana L., two to six fertile pinnae occur in the central portion of the frond. In the Cinnamon Fern, O. cinnamomea L., the dimorphism is more com- plete, as sporophylls consist entirely of fertile pinnae and become completely achlorophyllous. n our arrival at Mountain Lake Biological Station in the summer of 1984, we were surprised to discover a large number of O. cinnamomea plants bearing partially fertile fronds. These ranged continuously from those which were almost normal, i.e., with only a small portion of sterile blade tissue (usually at the tip) to those which were almost totally sterile, bearing only a few fertile segments, or even simply a few sporangia on sterile segment margins. The distribution of fertile pinnae on the fronds was highly irregular and unpredictable. The most frequent pattern was that with the lower portion of the frond fertile and inter- grading into an upper sterile portion (Fig. 1A, B), but numerous fronds were found with a reverse pattern (Fig. 1C), or with sterile pinnae both above and below fertile ones (Fig. 1D). Numerous specimens were taken and have been deposited in the University of Richmond Herbarium. By midsummer, normal fertile fronds of O. cinnamomea wither and fall to the ground or lie draped over sterile leaves. However, the sterile portions of the intermediate leaves remained turgid throughout the summer, making them & pecially conspicuous (a few of the nearly completely fertile fronds withered]. This allowed us to take note of their distribution during our travels in the Mou” tain Lake area (Giles and Montgomery counties, Virginia) in the summers © 1984 and 1985. It became apparent that the intermediate fronds were Very i calized; we found them only in the immediate vicinity of the Biological Station and on Bald Knob and Bean Field Mountain near Mountain Lake (about 2 0" southwest of the Biological Station). In these areas were patches where the 0 WERTH ET AL.: ONMUNDA CINNAMOMEA 129 10 cm Fic. 1. Osmunda cinnamomea forma frond porophylls collected at Mountain Lake in ~~ wet of 1984, showing a portion of the range of variation in distribution of fertile pinnae. A pee Most frequent variant with predominantly fertile lower portion intergrading into predominantly Sterile upper portion. C. Leaf with sterile base and fertile tip. D. Leaf with fully fertile area in middle of frond, and with one side more fertile than other. 130 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) termediate fronds predominated, although normal sporophylls were always found with them, sometimes on the same plant. Virtually all leaves produced during the summer of 1985 were normal, although one fertile-tipped intermediate was noted on the Biological Station grounds. Plants with such leaves were originally accorded varietal status as O. cinna- momea var. frondosa (Torrey, 1840), but the later reduction to forma (Britton, 1890) represents a much more satisfactory treatment. This leaf form is an envi- ronmentally induced abnormality and in this respect is analogous to numerous other trophophyll-sporophyll intermediates known in normally dimorphic ferns such as Onoclea sensibilis forma obtusilobata (discussed by Beitel et al., 1981). In some ferns with dimorphic or subdimorphic fronds, production of inter- mediate fronds is frequent or even characteristic. For example in Asplenium platyneuron, intermediate leaves can be found on most individuals. Develop- mental control appears to be exceedingly rigid in O. cinnamomea, however, since forma frondosa has rarely been reported (Torrey, 1840; McLouth, 1897; Owen, 1901; House, 1933; Kittredge, 1941; Margaret Wren, pers. comm.; Murray Evans, pers. comm.). The cause of this rare phenomenon has never been fully explained, despite its possible significance in understanding the reproductive biology and/or morphogenesis of leaves in O. cinnamomea. A complete expla- nation will require further study, but for the present the following observations are relevant. It is quite possible that production of sporophylls in O. cinnamomea is depen- dent upon an elevated nutritional level in the shoot apex. Production of sporo- phylls in nature is well known to occur primarily in open areas or under partial canopies. Enhanced sucrose concentrations favor development of excised leaf primordia into sporophylls in vitro (Sussex & Steeves, 1958; Harvey & Caponetti, 1972, 1973). Also, presence of ammonium ion is required for in vitro formation of fertile leaves (Harvey & Caponetti, 1974). The expression of frondosa leaves often seems to be linked to environmental trauma (similarly to formation of partially fertile Onoclea sensibilis fronds—see Beitel et al., 1981). McLouth (1897) found frondosa fronds emerging in the fall from plants whose leaves had been destroyed by fire. House (1933) reported finding frondosa leaves in July on plants surrounded by road tar, though it is not made clear whether the tar had been applied the same or the previous summer. Owen (1901) reported frondosa leaves following a late spring frost. Margaret Wren (pers. comm.) collected frondosa in the fall of 1980 and in the spring 9 1981 at a road construction site in Myrtle Beach, South Carolina, which had sem disturbed by Hurricane David in the fall of 1979. Leaves are determined as being sterile or fertile at the latest by the summer preceding their emergence (Steeves & Wetmore, 1953). It is reasonable to hypothesize that injury to tropho- phylls, which would lower nutrient flow to the shoot apex, might effect a partial reversal in the determination of developing leaf primordia. These altered leaves would then emerge late in the same season or the following spring. Steeves (1959) hypothesized that a reversal from sterile to fertile condition resulted in the predominantly fertile-tipped fronds observed by himself and by WERTH ET AL.: OSMUNDA CINNAMOMEA 131 Torrey (1840), based on the assumption that determination is directed from base to tip of the leaf primordium (Steeves & Briggs, 1958). Extending this hypothesis would lead us to predict that the leafy-tipped fronds observed by us and by Kittredge (1941) resulted from a reversal from fertile to sterile. However, the highly varied distribution of fertile pinnae we observed casts some doubt on this simplified explanation. It is difficult to conceive what sort of stimulus could cause reversals in both directions. We are currently unable unequivocally to identify an environmental stimulus which might have triggered the formation of the frondosa leaves we observed in 1984 at Mountain Lake. A weather station is located on the Biological Station grounds only a few meters from some of the plants in question. Records from this station over several years show no highly unusual conditions, although a cold snap did occur in late May of 1984 with low temperatures recorded at 31°F (—1°C). We would like to suggest the possibility that grazing by deer may have been a cause. Substantial grazing on O. cinnamomea was observed in the Moun- tain Lake area in 1982 (W. H. Wagner, pers. comm.). It is possible that instances of severe grazing during 1982 or 1983 may have resulted in the highly localized pattern of frondosa occurrence we observed in 1984. The possibility that injury is a cause of this interesting phenomemon suggests that it may be an adaptive response. A reduction in the amount of photosynthetic area caused by injury could be partially compensated by conversion of pre- sumptive sporogenous tissue into photosynthetic tissue. It would be interesting to attempt to induce this response experimentally by removing leaves from O. cinnamomec plants at various times during the growing season. We are grateful for facilities and weather records provided by the Mountain Lake Biological Station and would like to thank James Caponetti for helpful comments on the manuscript and Margaret Wren for providing details of her findings in South Carolina. LITERATURE CITED BEITEL, J. M., W. H. WacNneR, JR., and K. S. WALTER. 1981. Unusual frond development in sensitive fern Onoclea sensibilis L. Amer. Mid]. Naturalist 105:396-400 BritTon, N. L. 1890. Catalogue of plants found in New Jersey. : Harvey, W. H. and J. D. Caponetti. 1972. In vitro studies on the induction of sporogenous tissue on leaves of cinnamon fern. I. Environmental factors. Canad. J. Bot. 50:2673-2682. 1973. In vitro studies on the induction of sporogenous tissue on leaves of ism. Canad. J. Bot. 51:341-349. us tissue on leaves of SteEves, T. A. 1959. An interpretation of two forms of 230. and W. R. Briccs. 1958. Morphogenetic studies on Osmunda cinnamomec L.: The origin and early development of vegetative fronds. Phytomorphology 8:60-72. 132 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) — and R. H. WETMoRE. 1953. Morphogenetic studies on Osmunda cinnamomea L.: Some aspects of the general morphology. Phytomorphology 3:339-354. Sussex, I. M. and T. A. STEEVES. 1958. Experiments on the control of fertility of fern leaves in sterile culture. Bot. Gaz. (Crawfordsville) 119:203-208. Torrey, J. 1840. Report on the Botanical Department of the Geological Survey of New York. As- sembly, No. 50:113-197. SHORTER NOTES Gametophytes of Equisetum giganteum.—lIn a recent article, Duckett and Pang (J. Linn. Soc., Bot. 88:11-34, 1984) claimed that sexual behavior of Equisetum giganteum L. was similar to that of the other species of Equisetum subg. Hip- pochaete, with homosporous spores producing either antheridial or archegonial gametophytes. The archegonial gametophytes as they age begin to produce an- theridial lobes and eventually cease archegonial production. The antheridial gametophytes remain so, or with age and by lamellar proliferation also produce archegonial branches. I had earlier described E. giganteum gametophytes as different from those of any other species because they were normally bisexual, that is, simultaneously hermaphroditic, with both gametangia being borne upon the same meristem, rather than having separate antheridial lobes as the sequen- tially hermaphroditic species do (Hauke, Beih. Nova Hedwigia 8:1-123, 1963; Bull. Torrey Bot. Club 96:568-577, 1969). On this basis I had classified E. gigan- teum as a distinct section Incunabula, and theorized on evolution of heterothal- lism in Equisetum. Duckett and Pang speculated that my culturing procedure, using mass cultures, resulted in a very rapid shift from female to male, which I then interpreted as simultaneous hermaphroditism, rather than rapid sequential hermaphroditism. They cited a plant in cultivation in the Botanic Gardens, Edinburgh as the source of the spores from which they grew their Equisetum gametophytes. I wrote to the Botanic Gardens at Edinburgh requesting a specimen from the plant in ques- tion. Dr. C. N. Page was gracious enough to send me a fine specimen of t plant, and noted that it was originally collected in Mexico. Although it is super ficially much like E. giganteum, the Edinburgh plant is E. myriochaetum Schlecht. and Cham. That species has stomata in a single line on each side of the inter nodal grooves, rather than rows 2 to 4 stomata wide. The branch ridges viewed in profile show a sawtooth pattern rather than a blocky pattern. The stem sheath segments are flattened and green, rather than rounded and light brown. Se far as I have seen, all large branched Equisetum collections from Mexico are either E. myriochaetum or the hybrid E. x schaffneri, and E. giganteum does not occur in that country. As E. myriochaetum rather than E. giganteum, the gametophytes described by Duckett and Pang are in agreement with my (1969) descriptions of that species, and their challenge to my classifications of Equisetum subg. Hippochaete and sali Se o of evolution of heterothallism in the genus Equisetum is not soe ported.—RicHarp L. HauKE De , iversi ode Islan’ ae partment of Botany, University of Rh SHORTER NOTES 188 Marsilea macropoda New to Louisiana—During February, 1982, one of us (SH) discovered a population of Marsilea growing within the city limits of New Or- leans. We identified the plants as Marsilea macropoda Engelm. ex A. Braun. The population occurs in an old abandoned lot. The situation is suggestive of persistence after cultivation; however, local residents could not substantiate this. The lot is littered with gravel and concrete rubble, indicating that the plants might also have been introduced through disturbance and may have persisted at the locality. The population is large and vigorous, and at the time of collection most plants were fertile. Specimens from the population (Landry 7895) are de- posited at LAF, SMU, US, and NY. Marsilea macropoda is readily distinguish- able from other species known to occur in Louisiana, M. uncinata and M. vestita, by its markedly more robust stature, the lack of a prominent tooth on the spo- rocarps, and the frequent presence of two or more sporocarps per peduncle. Marsilea macropoda is otherwise known from south-central Texas, where it previously has been regarded as a Texas endemic (Correll & Johnston, Manual of the Vascular Plants of Texas, Texas Research Foundation, Renner, TX, 1970). The Louisiana population represents a range extension of approximately 400 miles for the species.—GARRIE P. LANDRY, Department of Biology, University of Southwestern Louisiana, Lafayette, LA 70504; and SAMUEL W. HOLDER, Jefferson Parish Environmental and Development Control Department, 3600 Jefferson Hwy., Jefferson, LA 70121. E. A. C. L. E. Schelpe (1924-1985) Professor Edmund André Charles Lois Eloi (‘Ted’) Schelpe died after a sudden heart attack at his home on 12 October 1985. He was born in Durban, South Africa, on 27 July 1924 and completed a BSc with distinction in Botany at Natal University in 1943; at the same University he obtained an MSc (Class I) in 1946. In 1952, he was awarded the DPhil degree of Oxford University (Wadham Col- lege). He held the following posts at the University of Cape Town: Lecturer in Botany, 1953-1954; Senior Lecturer and Curator of the Bolus Herbarium, 1954- 1958; Associate Professor and Curator of the Bolus Herbarium, 1968-1973, the title Curator being changed to Director in 1970; Professor (ad hominem) and Director of the Bolus Herbarium, 1973-1985. Professor Schelpe was an elected Fellow of the Royal Society of South Africa, the Linnean Society of London, and the University of Cape Town. From 1964 he was a member of the Committee on Pteridophyta of the International Association of Plant Taxonomists. Among Voluntary societies, he played a valuable role in helping found the Orchid So- ciety of Southern Africa and was President of the Botanical Society of Southern He made major contributions to the systematics of Pteridophyta in southern Africa. A few days before he passed away, he finished checking the final proofs of the volume on Pteridophyta for the Flora of Southern Africa. This was a 134 AMERICAN FERN JOURNAL: VOLUME 75 NUMBER 4 (1985) spectus Florae Angolensis. He published over 100 scientific papers and books, which besides Pteridophyta covered topics in orchids, bryophytes, plant ecology, and phytogeography. A complete bibliography of Professor Schelpe’s publica- tions will appear in Bothalia, volume 16, part 2.—From material provided through the kindness of A. V. HALL, University of Cape Town, South Africa. 1986 AIBS MEETING—CALL FOR PAPERS The American Fern Society and the Botanical Society of America will meet with the AIBS Annual Meeting at the University of Massachusetts, Amherst, on 10-14 August 1986. Those members of the American Fern Society wishing to present a paper or poster and who have not received abstract forms may obtain forms from the program chairman of the American Fern Society: Dr. FLORENCE S. WAGNER, Department of Botany, University of Michigan, Ann Arbor, MI 48109. Referees, 1985—I thank the Associate Editors and referees listed below for their valuable assistance in the review process. Their evaluations of manuscripts submitted to American Fern Journal have aided authors, made my job easier, and contributed to the quality of our journal.— ALAN R. SMITH David S. Barrington James D. Montgomery Brent Berlin Richard Mueller David E. Bilderback Clifton E. Nauman Donald M. Britton James H. Peck Ralph E. Brooks Michael G. Price Raymond B. Cranfill V. Raghavan Jeffrey G. Duckett P. Mick Richardson Ernest M. Gifford, Jr. Rudolf Schmid Richard L. Hauke David Seigler R. James Hickey Darlene Southworth David M. Johnson Robert G. Stolze Donald R. Kaplan W. Carl Taylor Jon E. Keeley Alice Tryon Terry W. Lucansky Rolla Tryon Michael R. Mesler David H. Wagner John T. Mickel Warren H. Wagner, Jr. John H. Miller Index to Volume 75—1985 Classified entries: botanical names (new names in boldface); major subject headings, including from titles; reviews (grouped, listed by last name of first author of work reviewed). Names words of authors, follo llowed by titles of artic] f Iphabetically in Table of Contents, iii-iv. €s or references to first authors, are listed alp tsa te ex ns _ ips Anemia: subg. Coptophyllum, 36, 37; A. antrorsa, — emidaria 33; A. ayacuchensis, 36; A. elegan: ns, 37; A. jalis- is B Cyatheaceae 80; Bg. — ——? entosa val- uisetum 111; Sphaeropteris, 80 INDEX TO VOLUME 75 Asplenium: A. subnormale, 73; A. —— 73 Atomic composition, of Onoclea s Ba Azolla: A. mexicana, 38; oe content Bolbitis -opceg agen Cnemi 7G. ve. 81; anatomy, 8 Cecipteris 6; C. brevi sa, 9; C. hax 9; C. lanigera, 8; C. any 9; C. senilis, 9; C. ciate. . Cc. stella, 8 Culcita coniifolia, 13 Cyatheaceae: Cnemidaria, 80; Sphaeropteris, 80; ric ee 92 Cyrtomium. ici . xuosa Dryopteris: D. 71; D. goldiana, 23; D. ludoviciana, 71 Ecology: aquatic ecotype of Asplenium unilat- erale, 73; branching patterns in Lophosoria, 105; colonization of Volcan Chichonal by Pityro- amma, 1 uisetum: E. arvense, 112; E. diffusum, 112; E. a 112; E. telmateia, 112; pi of E. giganteum, 132; . is icum, 120; on- togeny of commissur eae of ees Indians, 19 Evolution: Polys Gametophytes: oneceniere 79; Equisetum gi- ganteum, 132; E. sylvaticum, 120 G trifurca’ Grammitidaceae, 6 Grammitis, 6; G. cultrata, 8; G. lanigera, 8; G. Senilis, 9 Hybridization, in Polystichum, 22 Indians, Chacobo, 19 Isoétes: I. acadiensis, 44: I. eatonii, 44; I. echi- Macrospora, 44; I. melan riparia, 44; I. — 102; I. tuckerman- ii, 44; morphology Loxogramme, 9 Lophosoria ore branching pattern, 105 Lycopodium, 39 ea: M. crenata, 30; M. macropoda, 133; M. ata, 28; M. mollis, 30; M oligospora 30; M. quadrifolia, 28; M. vestita, 28; longevity of Sporocarps, 30 : Equisetum, commissure, 111; Iso- étes, leaves nee sporangial region, 44; ache teris, spores, 9: i ELA, ae L. E. Schelpe (1924-1985), 133 clea sensibilis, atomic composition of Spores, 12 cinnamomea f. Papuapteris, 29 ea f. frondosa, 128 135 ecluma, enamel Pityr _ wile 1 Plecosorus, 2 Polypeiiiegas pean 6: ue see omiase 9; P. longum, 9; P. tum-plumula complex pecti pasate Symposium, ee taxonomy, 22; sect. Metapolystichum, 25; P. aleuticum, 72; P. ac serra: 23, 25; P. braunii, 24; P. dictyo- BP, ae SO 26; P. echinatum, 25; imbricans, 24; P krucke- murica pie 24, 25; P. polyphyllum, 25; P. scopulinum, 4; P. setigerum, 24; P. speciosissimum, 25 um Pteridium aquilinum Range sstonbona/ dale Alaska, — Is- chum Ar land, Polystic Dryopteris ludoviciana, D. x australis, 71; rl ida, Dicranop' flexuosa, 79; Georgia, Marsilea cmos M. vestita, 28; Louisiana, Isoétes melanopoda, 77; Marsilea macropoda, 133 Referees, 1985, 1 Reviews: Foster, . Gordon, Ferns to know and W. B. G., The ferns and Africa, 31; Jarrett, F. metophytes of Ophioglossaceae, 18; Prelli, Remy, Guide de fougeres et oe alliees, 104; Proctor, George R., Ferns of Jamaica, a guide to the peidphytes 119; _ . Cal Ar- k rns and fern allies, Schelpe, ey A.C. LE. faa obituary, 133 Selaginella: kraussiana, 13, 40; aff. schizobasis, 1 rags S. elongata, . snahneey, 00 pores: Anemia, 35; Azol 8; Grammitis, 6; Ly- copodium, 39; A oe Oncon. 12: ace reduaia 6; Tri- chipteris, 92 Sporopollenin, of Azolla, 38 Symposium, Polystichum, 22 Thelypteris: subg. Amauropelta, 56; subg. Gon- iopteris, 63; T. abdita, 63; T. cordata var. imi- tata, 56; T. rheophyta, 58; T. verecun- da, 65 Trichipteris corcovadensis, 92 Volcan Chichon 7 STATEMENT OF OWNERSHIP. MANAGEMENT AND CIRCULATION Required by 39 U.S.C. 3685) 1A. TITLE OF PUBLICS MERICAN FERN JOURNAL 3A. NO. OF pact a SUBSCRIPTION 3. FREQUENCY OF ISSUE ANNUALLY PRIC! quarterly 4 wi Ka $9 subserx NOWN OFFICE OF City, County, Pringle Herbarium, University of Vermont, Burlington, VT 05405 Pringle Herbarium, University of Vermont, Burlington, VT 05405 6. FULL NAMES AND COMPLETE MAILING ADDRESS OF PUBLISHER, ED! PUBLISHER (Name and Complete Mailing Address) AMERICAN FERN SOCIETY, INC. (see 5.) Alan R..Smith, Dept. of Botany, Univ. of California, Berkeley, CA 94720 none 7. OWNER (If owned by a corporation, its name and addr: fated and also immediately thereunder the names and addresses of stockholders owning or hoiding | percent or more of total amount akoak sooptory owned by a corporation, the names and addresses of ~ progeny owners must be given. ee firm, its name and address, as well as that of each individual m in. If the publica- tion is published by a nonprofit organization, its name and address must be stated.) (Item must be completed.) FULL NAME ADDRESS a AMERICAN FERN SOCIETY, INC. (see 5.) ee | | [ ee | | 8. KNOWN BONDHOLDERS, MORTGAGEES, AND OTHER SECURITY HOLDERS OWNING OR HOLDING 1 PERCENT OR MORE OF TOTAL AMOUNT OF BONDS, MORTGAGES OR OTHER SECURITIES (If there are none, so state) FULL NAME ADDRESS none 6. — COMPLETION ing NONPROFIT ORGANIZATIONS AUTHORIZED TO MAIL oa SPECIAL RATES (Section 423.12 DMM only) one) urpose, function, and nonprofit status of this organization and the exempt status for Federal income tax purposes (Check NAS NOT CHANGED DURING Ha —- HAS CHANGED DURING If changed, [x] PRECEDING 12 MONTHS PRECEDING 12 MONTHS Snowe with this statement.) “4 LNO. COPIES OF SING EXTENT AND NATURE OF CIRCULATION i cae: | mmr ED NEAREST TO (See instructions on reverse side) — at cy MON A ame ISSUE PUBLING DATE A. TOTAL NO. COPIES (Net Press Run) 1400 1400 B. PAID AND/OR REQUESTED CIRCULATION 1. Sales through dealers and and carriers, street vendors and counter sales 0 9 2. Mail Subscription (Paid and/or requested) 12 o4 1096 c. TOTAL PAID — REQUESTED CIRCU (Sum of 10B1 and 10B2) Arn 1204 1096 D. FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS SAMPLES, COMPLIMENTARY, AND OTHER FREE COPIES A 3 E. TOTAL DISTRIBUTION (Sum of Cand D) 1205 1099 = 2 ee A F. COPIES NOT DISTRIBUTED ee 1. Office use, left over, unaccounted, spoiled after Printing 19 5 3 OL Glasnen fern Al gents re) (8) Be: G. TOTAL (Sum of FE Fi and 2 chouit * a as 1400 1400 | certify that the statements made by SIGNATURE AND T}TL “PUBLISHER, BUSINESS M Gen” RS : me above are correct and complete | Alan R. Smith, editor — PS Form 3526, July 1984 a (See instruction on reverse) INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable}, including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 8% by 11 inch paper of good quality, not:“erasable” paper. Double space manuscripts throughout, including title, authors’ names and addresses, text (including heads and keys), literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author’s name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed - pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. Use S.I. (metric) units for all mea- sures (e.g., distance, elevation, weight) unless quoted or cited from another source (e.g., specimen citations). For nomenclatural matter {i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- breviate titles of serial publications according to Botanico-Periodicum-Huntian- _ um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as part of nomenclatural matter are not included in lit- erature cited. For shorter notes and reviews, put all references paren enthetically in text. Use Index herbariorum (Regnum Veg. 106:1-452. 1981) for a of herbaria. Illustrations should be proportioned to fit page width with caption « on the meme Page. Provide margins of at least 25 mm on all illust is-tone illustrations, design ori ithout a ae gape | ao In com bu tt of acent RES: com- posite ‘blocks, abu edges adj : “sequence and rubering of Sgr ad of wie with Sehgal "ste e and reference to latitude aiid: longitude in ¢ a Proofs and Se ee he ee eae . _ Should send corrected proofs to the editor and SS. ee - —o will be assessed t chee for extensive alterations eee dD idee aa co AMERICAN Ge FERN Noe JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Isoétes Megaspore Surface Morphology: Nomenclature, Variation, and Systematic Im- portance R. James Hickey 1 Active Enzymes from Megaspores of Marsilea and Regnellidium Douglas E. Soltis and Pamela S. Soltis 17 Germinating Spores and Growing Sporelings of Aquatic Isoétes W. Carl Taylor and Neil T. Luebke 21 Shorter Notes Pellaea brachyptera New to Washington Edward R. Alverson and Joseph Arnett 25 A Second F. Minsk A . om rs rr... c.l page ry . tentrionale i Eleanor M. Bush 26 Second Locality for Dryopteris carthusiana in Arkansas James H. Peck 28 ae Information for Authors oe , oe 2 Cover 3 ae The American Fern Society Council for 1986 TERRY R. WEBSTER, Biological Sciences Group, University of Connecticut, Storrs, CT 06268. President FLORENCE S$. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. Vice-President w. CASL TAYLOR, Milwaukee Public Museum, 5 sigorpmaans WI 53233. D. CAPONETTI, t. of Botany, University of Tennessee, Knoxville, TN 37916.Treasurer DAVID S. BARRINGTON, Dept. of Botany, University of Vetiuet Burlington, VT 05405. Records reasurer ALAN R. SMITH, Dept. of Botany, University of California, gospels CA 94720. Bp cone Editor Smithsoni DAVID B. LELLINGER, ian Institution, Washington, DC ir Editor DENNIS Wm. STEVENSON, Dept. of igi Science, Barnard sents Columbia University, New York, NY 1 Fiddlehead Forum Editor Hees Fern Journal EDITOR AiR OMe ee Dept. of Botany, agp! of California, ey, CA 94720 ASSOCIATE EDITORS GERALD J.GASTONY............. Dept. of Biology, Indiana University, Bloomington, IN 47401 Cae ETI PIGAIP OR ee ee Dept. of Botany, University of Kansas, : Lawrence, KS 66045 _ DAVID B. 2 soma Se Saas U.S. Nat'l Herbarium NHB-166, Smithsonian Institution, Roe Was! on, DC 20560 TERRY R. WEBSTER ... . Biological Sciences Group, University of Connecticut, Sess CT 06268 ; ‘The “American Fern ern Journal” (ISSN 0002-8444) is an illustrated quarterly devoted to the general study of ferns. It is owned by the American Fern Society, and published at the Pringle Herbarium, — University of Vermont, Burlington, VT 05405. Second-class postage paid at Burlington, VT, and additional entry point. : x iI,g be ¢f. . 1 at +h date of issue. and orders for back issues should be addressed to Dr. James D. Montgomery, Ecology Ill, RD. 1 Berwick, PA 18603 _ Changer of adres, , and sche tions fe bership should be sent to the Records Trea- Subscript mn pe a re net if paid through an agency oe fee $0.50); sent free to : : ‘USA, Canada, and Meri Fern Society (annual dues, $10.00 + $4.00 mailing surcharge beyond USA, C Sone mane ip 00 oe e back £6 or less, $1.25; 65-80 : ‘ hageor di ymnpee ieee ei ere ‘AMERICAN Fann Journat. Dept. of Botany, University a 2 American Fern Journal 76(1):1-16 (1986) Isoétes Megaspore Surface Morphology: Nomenclature, Variation, and Systematic Importance R. JAMES HICKEY Botany Department, Miami University, Oxford, OH 45056 The Isoétaceae is a small, cosmopolitan family of heterosporous pteridophytes within the lycopsid line of evolution. Based on a number of shared derived characters, such as a ligule and heterospory, it appears to be most closely related to the extant family Selaginellaceae. Unlike the latter, members of the Isoétaceae are aquatic or are associated with seasonally inundated habitats. The family is monotypic: the genus Isoétes comprises approximately 150 species (Tryon & Tryon, 1982). The exact number of species is difficult to establish because of discrep- ancies among authorities regarding species delimitation and because Isoétes in many parts of the world is still poorly known. Systematic studies in the genus have been greatly hampered by both the lack of collections and the tremendous degree of phenotypic plasticity; as stated by Kott (1980) “perhaps the only constant in this group is its variation.” Although a thorough study of the genus leaves no doubt as to the validity of most species (Boom, 1982; Kott & Britton, 1983; Hickey, 1985), identification of taxa is difficult because of the paucity and variability of characters. As a result of these prob- lems, and the small number of collections per species, workers have placed special emphasis on one or two characters and have erected infrageneric clas- sification systems around them. The first such classification was proposed by Braun (fide Fuchs, 1962) and was based on habitat preference, primarily as inferred from leaf morphology. In general this classification system was widely accepted; subsequent treatments (e.g. Baker, 1880; Motelay & Vendryes, 1882; Eaton, 1900) included some hierarchical changes, but the underlying theme re- mained unchanged. As a result of better specimen quality and more exact field data, it soon became apparent that this classification system was both unnatural and simplistic (Engelmann, 1882; Eaton, 1900). West and Takeda (1915) made the first break with implied habitat characteristics by stressing anatomical differ- €nces. While their emphasis on primary, observable characters is noteworthy, it should be realized that their characters are essentially the same ones that Braun (1864) used to distinguish between terrestrial and aquatic habitats. In 1922, Pfeiffer classified members of the family according to magnepore Surface ornamentation (Table 1). Her reason for rejecting previous classifications 'SNot directly stated but seems to stem from the unreliability of habitat data and T understanding of evolutionary changes in megaspore surface mor- : Phology, Pfeiffer recognized four basic megaspore surface patterns and, based on these, established sections Tuberculatae, Echinatae, Cristatae, and Reticu- : - Despite criticisms regarding the reliability of megaspore morphology (En- 8elmann, 1882; Eaton, 1900; Pfeiffer, 1922; Duthie, 1929), and the lack of corre- 2 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) TABLE 1. Infrageneric Classification System in Isoétes by Pfeiffer (1922), as Modified by Fuchs (Fuchs 1962; Fuchs-Eckert 1981). Section Spore type Isoétes (=Cristatae Pfeiffer) cristate Terrestres A. Br. (=Reticulatae Pfeiffer) reticulate alustres A. Br. (=Tuberculatae Pfeiffer) tuberculate Echinatae Pfeiffer echinate Laevis Fuchs levigate spondence between megaspore morphology and other characters (Berthet & Lecocgq, 1977), her classification system has been almost universally accepted. In 1981, Fuchs-Eckert added a fifth section, Laevis, which is characterized by smooth megaspores. During revisionary studies on Neotropical species (Hickey, 1985), it became obvious that there were additional problems associated with this reliance on megaspore surface morphology. In particular, use of this single character set has obscured phylogenetic relationships (Hickey, 1985, 1986a). The problems of this one-character taxonomy are amplified by overly simplistic and non-discrimi- natory megaspore character state assessments and by a failure to look for cor- relations with other characters (and its corollary, reasons for a lack of such correlations). Multiple character analyses are necessary to determine the limits and affinities of Isoétes species (Hickey, 1985, 1986a, b). Within such analyses, megaspore morphology remains one of the most significant sources of phyloge- netically informative characters. It is essential, however, that the particular states be accurately and completely described. The primary purpose of this paper is to present a more precise, expanded nomenclature for the character states of Isoétes megaspore ornamentation. In addition, the related problems of character variation, homoplasy, and the systematic importance of megaspore data are dis- cussed. While all of these facets are interconnected, it is simpler to discuss terminology, spore maturity, polymorphism, and convergence separately. TERMINOLOGY Surface morphology.—While Pfeiffer (1922) recognized only four formal cat- egories of megaspore ornamentation, her discussions suggest that she was aware of - greater diversity of forms. For example, within sect. Tuberculatae she rec ognized warty, tuberculate, and papillate types as well as spores in which ie ornamentation “fail{s) to appear, leaving a smooth surface.” Because of the dif- ficulty in describing this diversity she relied largely on photographs of the spores; unfortunately these photographs were poorly reproduced and therefore of rel- atively little use. Even with the addition of a fifth spore type (smooth or levigate, F uchs-Eckert, 1981) the diversity of surface morphology far exceeded the NUM ber of recognized types. Although the use of only five spore types simplified the identification of ornamentation patterns, the inclusiveness of the terminology would not allow for discrimination between otherwise similar taxa. As an aller R. J. HICKEY: ISOETES MEGASPORES 3 TaBLE 2. Types of Megaspore Surface Morphology in Isoétes: Descriptions, Previous Classification, and Specific Examples. Tuberculate: Spores with moderately broad, radially symmetrical projections, as high as wide, with ight, inclined sides and an acute apex. Examples—I. bradei, I. luetzelburgii, I. melan- opoda. (Figs. 1a, b, 4a) Pustulate: Spores with broad, radially symmetrical projections, as wide as high or wider, with convex sides and an obtuse, rounded apex. Previously classified as tuberculate. Examples—I. cu- bana, I. triangula. (Figs. ic, d) Levigate: Spores without apparent ornamentation, smooth. Previously classified as tuberculate or levigate. Examples—I. lechleri, I. mexicana, I. gunnii. (Figs. 1e, f) Saccate: Spores with radially to bilaterally symmetrical, bulbous projections which appear deflated and collapsed. Previously classified as tuberculate. Example—I. gigantea. (Fig. 2a) Clavate: Spores with radially symmetrical projections which are approximately as high as wide, with obtuse, expanded apex. Previously classified as tuberculate. Examples—I. ovata, a sub- clavate condition is seen in I. clavata. (Fig. 2b) Verrucate: Spores with radially symmetrical projections, as high as or slightly higher than wide, with straight parallel sides and an obtuse apex. Previously classified as tuberculate. Example— I. gardneriana. (Fig. 2c Baculate: Spores with narrow, radially symmetrical projections 112 or more times as high as wide, with straight, parallel sides and an obtuse, rounded apex. Previously classified as tubercu- late. Examples—I. panamensis, I. baculata. (Figs. 2d-f) Echinate: Spores with narrow, radially symmetrical (or nearly so) projections, at least 11 times as high as wide, with straight, inclined sides and an attenuate, pointed apex. Previously clas- sified as echinate. Examples—I. echinospora, some specimens of I. storkii and I. andina. (Figs. 5d, e) Rugulate: Spores with broad, distinct muri, as wide as or wider than tall, with a broadly rounded est; the muri not forming distinct areoles. Previously classified as cristate. Examples—I. savatieri, I. herzogii. (Figs. 3a—c, 5b, c) Cristate: Spores with thin, distinct muri, taller than broad, usually with an apically serrate crest; not forming distinct areoles. Previously classified as cristate. Examples—I. storkii, I. ulei, I. andina. (Figs. 3d, 5 Reticulate: Spores with thin, distinct muri, taller than broad, usually with an apically serrate crest; forming distinct areoles. Previously classified as reticulate. Examples—I. engelmannii, I. martii, I. novo-granadensis. (Figs. 3e, f) Retate: Spores with broad, distinct muri, as wide as or wider than tall, with a broadly rounded crest; forming distinct areoles. Previously classified as reticulate. Examples—I. durieui, I. ste- vensii. (See fig. Va in Berthet & Lecocq, 1977, and fig. 7 in Croft, 1980) native, I propose a more discriminatory megaspore terminology circumscribing twelve defined morphological types (Table 2). While the inherent variability and intergradation in surface morphology precludes an all-encompassing classifica tion, the additional seven categories are distinct enough to merit recognition, and will allow for more rapid and accurate identification of taxa. Definitions (Table 2) and scanning electron micrographs (Figs. 1-5) are included to provide a for standardization and stability of nomenclature. Spore types were initially characterized from Neotropical eer and the resultant States compared with spores of extra-Neotropical species. A 1 Hterature survey and an examination of Isoétes specimens from A, GH, and NY revealed only one additional ornamentation type. This retate spore pattern is Similar in morphology to the more common reticulate spore type. However, the 4 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) two types differ in their evolutionary origins: retate spores are the result of anastomosing rugae or tubercles while reticulate spores are the result of anas- tomosing cristae. Retate spores are found in I. durieui Bory (fig. Va in Berthet & Lecocgq, 1977), I. stevensii Croft (fig. 7 in Croft, 1980), and several other Old World species. Even with the recognition of an increased number of megaspore types, some spores are intermediate in morphology and do not fit neatly into one of the twelve categories. Such spore patterns are best described by means of hyphen- ated terms, e.g. echinate-cristate or cristate-reticulate. Secondary ornamentation.—In addition to gross surface morphology, other aspects of the spore surface have systematic application. One such character is the change in density and discreteness of ornamentation on the distal spore surface immediately adjacent to the equatorial ridge. Changes in this region may take the form of a decrease (Figs. 1b, 5c), an increase (Figs. 3d-f, 5e, f), or the total loss of ornamentation (Figs. 4e, 5a, b). This region of differential ornamen- tation, the girdle, is useful in the identification of I. acadiensis, I. hieroglyphica A. A. Eaton, I. macrospora (Kott & Britton, 1983), I. weberi Herter, I. montezu- mae A. A. Eaton, I. savatieri Franchet, and I. storkii Palmer (Hickey, 1985). A second suite of characters, which is associated with the variation in form of the equatorial and proximal ridges, has rarely been used as a source of systematic characters. However, a number of tropical species do possess stable and unusual ridge features that are of systematic value (Hickey, 1985). For example, I. gigan- tea U. Weber is readily distinguished from all other Neotropical species by its extremely narrow, knife-like proximal ridges (Fig. 2a). In morphologically similar species pairs such as I. cubana Baker and I. triangula Weber, the cross-sectional shapes of the equatorial and proximal ridges (i.e. whether rounded or triangular) are diagnostic (Hickey, 1985). Isoétes gardneriana Mett. and, to a lesser extent, I. panamensis Maxon & Morton are characterized by an unusual extension of the equatorial ridge that projects outwardly at the junctures with the proximal ridges (Figs. 2c, d). This distinctive flange is also found in I. melanotheca Alston DAE SA Soi ae aS arene a Fic. 2, Megaspore surface morphology in Isoétes. Saccate form: a, I. gigantea, Luetzelburg 15057 {M). Subclavate form: b, I. clavata, Martyn 379 (M)}. Verrucate form: c, I. gardneriana, Gardner 3563 (G). Baculate form: d, I. panamensis, Woodson et al. 1685 (MO). e, I. panamensis, Woodson et . 1685 (US). f, I. panamensis, Gémez 17530 (CR). Bar = 200 um. Fic. 3. Megaspore surface morphology in Isoétes. Rugulate form: a, I. herzogii, pace t pee ig! (AAU); b, I. weberi, Herter 95840 (NY); c, I. spec. nov., Chavez 2323 (GH). aan a d, I. ulei, Ule 3533 (CORD). Reticulate form: e, I. martii, Sehnem 14987 (Fuchs’ herba Ze spec. nov., Spannagel 301 (S)}. Bar = 200 pm. : R. J. HICKEY ISOETES MEGASPORES AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) R. J. HICKEY: ISOETES MEGASPORES z = m4 fad ies) 2 = Z R ia 2 = hed ) vd < Z a4 = o Z 24 = Z < iS) R. J. HICKEY: ISOETES MEGASPORES 9 of Africa. In contrast, the sinuous equatorial ridge (Figs. 2b, d-f) often seen in I. panamensis, I. cubana, and I. acadiensis is variable and unreliable as a species identifying character. SPORE MATURITY While an extensive study of spore development has not been undertaken, a number of developmental changes have been observed which are worthy of discussion. Many of these changes have been previously reported (Boom, 1982; Kott & Britton, 1983) but are important enough to be repeated. As Isoétes mega- spores mature, the form and consistency of the perispore (sensu Robert et al., 1973) undergo distinct changes. Initially the perispore is soft and mealy and, when touched, portions of it tend to flake off. With age, the perispore solidifies into a uniform, integral structure. This change in texture has been regarded as the “test” of spore maturity (Boom, 1982; Kott & Britton, 1983). Spore maturity has not been a major problem in Isoétes taxonomy and I know of only one species, I. saccharata Engelm., which has been described on the basis of im- mature spore characters. The real difficulty associated with spore maturity lies in the subsequent identification of specimens. While an ornamentation pattern is apparent fairly early in spore development (Figs. 4a, b), it may be drastically different from that of the mature spore. Developmental changes in ornamenta- tion result from either an initial lack of ornamentation or an obscuring of initial perispore patterns by subsequent wall deposition. This is particularly evident in I. andicola (Amstutz) Gomez (Fig. 4) where young spores are weakly pustulate (Figs. 4a, b) and older spores vary from pustulate, to rugulate, to levigate. Because the developing perispore is soft, the surface morphology of spores from plants collected early in the growing season is often distorted when pressed for herbarium specimens. Distortion is also possible in specimens where the Perispore deposition is complete but where the wall has not completely solidi- fied. In such cases, the ornamentation is evident, but the various protuberances and muri become flattened or bent out of position (Figs. 1f, 2b). Once the peri- spore hardens, it is brittle and easily damaged (Figs. 2e, 3d). On some spores, such damage can be so extensive as to make it difficult or impossible to interpret clearly the original surface pattern. POLYMORPHISM Prior to Pfeiffer’s (1922) monograph, megaspore surface morphology was con- sidered so variable as to “furnish no legitimate grounds for classification (En- Bee tense SA asc Cuca es ae en emmme ose Rea OEE acca = : Fic. 4. Megaspo: 5 icola: a and b, c and d, and e and f are re surface morphology in Isoetes andicola: a an : phs of megaspores taken from successively older sporophylls of Hutchinson - _— oor (GH-spirit material). Photos b, d, and { are high magnification views of the sp oe ® respectively. Bars in a, c, and e = 200 um; bars in b, d, and f = 30 wm. R. J. HICKEY: ISOETES MEGASPORES 11 gelmann, 1882). However, since the publication of Pfeiffer’s monograph, vari- ability in spore morphology has been largely ignored. In an organism where so many morphological characters have been excluded because of phenotypic plas- ticity, it is surprising that the variability in the primary character has not been more openly discussed. The following discussions present a brief overview of the more common patterns of infraspecific variation observed in Isoétes mega- spore ornamentation. Such variation falls into two major categories: clinal and non-clinal. Clinal variation.—Perhaps the first recorded example of the recognition of clinal variation is that given by Eaton (1900) in a discussion of changes in mega- spore ornamentation in I. engelmannii along a north-south transect. Recently, Croft (1980) interpreted the species diversity (and megaspore surface morphol- ogy) in Papuasian Isoétes to be a result of speciation along an ancient east-west cline. Extensive geographic variation was noted in I. riparia by Proctor (1949), but was not interpreted as clinal. The presence of conflicting taxonomic inter- pretations (Proctor, 1949; Reed, 1965; Kott & Britton, 1983) of this group suggests that additional study is still needed. Clinal variation in any character, including megaspore surface ornamentation, is neither surprising nor problematic. Thorough documentation of such variation could help us understand modes of speciation in addition to the origins and relationships of spore ornamentation patterns. The generally restricted ranges and the relatively few collections of most Isoétes species account, in part, for the lack of documentation of otherwise obvious clinal variation. Clines have also been overlooked due to rather narrow species circumscriptions and the lack of accompanying discussions of taxonomic relationships. A careful analysis of the morphology of these narrowly circumscribed species frequently leads to a rein- terpretation of their identity as members of subtle continua (Hickey, 1985). In the Neotropics, geographic and elevational clines are common. Isoetes stor- kii provides an excellent example of an elevational/ecological cline. This species is endemic to Costa Rica and is known from only three localities, each of which iffers in mean temperature, mean daily solar radiation, and elevation (Cox & Hickey, 1984). Individuals from these three populations show intergradation om spore morphology from weakly echinate-cristate, to echinate, to sub-levigate. Megaspore diameter shows a parallel progression from large to small spores. In the southern Andes, megaspores of I. savatieri vary from rugulate, rarely tuberculate, in the south, to rugulate and rugulate-retate in the north (Figs. 5a- ¢). Isoétes andina and I. killipii of the northern Andes (Ecuador to Venezuela) also show clinal variation (Figs. 5d-f). In the former, megaspores of southern Plants are typically densely echinate while those from more northerly popula- ee Pig. 5. Infras: ific Gos 5 culate form, Die pt pints b. Saas form, Borge s.n. (GH); c, Weakly rugulate form, Castellanos 114242 (AA). I. andina: d, S ly echinate form, @llgaard & Balslev 8137 (AAU); e, Densely echinate form, Oligaard & Balslev 8782 (AAU); f, Densely cristate form, Oligaard & Balslev 8943 — 200 um. 12 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) tions have a low, crowded, cristate ornamentation. Spores of I. killipii from Venezuela are weakly reticulate to reticulate-cristate while those from Colombia and Ecuador are strongly and distinctly reticulate. In the latter two species there is parallel variation in velum development superimposed upon the variation in megaspore morphology. Non-clinal variation.—Perhaps more striking, and certainly more difficult to deal with, is the degree of non-clinal infraspecific variation within individual populations, plants, and sporangia. While Pfeiffer (1922) recognized the presence of such variation, she considered it to be too minor to negate the use of mega- spore surface morphology as the primary systematic character. Duthie (1929}, like Engelmann (1882) and Eaton (1900), was not impressed with the importance of spore morphology and commented on the expression and systematic impli- cations of spore morphology in two new African species, I. capensis and I. stephansenii. The megaspores of these species were described (Duthie, 1929) as ranging from tuberculate, to rugulate, to retate (as reticulate) within a single population; similar variation within sporangia and on the various faces of indi- vidual spores was also described. Spores of the Australian I. muelleri A. Br. exhibit a range of surface morphology from tuberculate, to rugulate, to retate (Marsden, 1976). In the Neotropics, examples of infraspecific and intrasporangial variation are common (I. montezumae, I. mexicana Underw., and I. ei U. Weber). Perhaps the most striking example is that seen in populations of I. andicola from west-central Bolivia where a single sporangium may contain pus- tulate, rugulate, and levigate megaspores. This species could be assigned to either sections Isoétes, Laevis, or Palustres! Intrasporangial variation may also include differences in spore shape, pres ence or absence of cellular contents, and spore fusion. These types of variation are often associated with sterile hybrids (Taylor et al., 1985) or plants with un- balanced genomes (Marsden, 1976; Pant and Srivastava, 1962, 1965; Mondal, 1978). While such variation is important, it is beyond the scope of the current paper and the reader is referred to the citations above. A final pertinent aspect of polymorphism is that dealing with variation in ornamentation on a single spore. It has long been recognized that the distal and proximal faces of a megaspore may exhibit different surface patterns (Boom, 1982; Duthie, 1929; Proctor, 1949; Williams, 1943). For example, megaspores of I. herzogii are usually either levigate or tuberculate on the proximal faces but are — commonly rugulate distally. Such variation presents both descriptive and taxonomic problems. In describing the megaspore surface morphology, the pat- terns observed on both the proximal and distal faces must be compared. A simple description, such as “spores rugulate,” is insufficient. Traditionally, when such spores were encountered only the distal pattern was communicated, a situation which often led to errors in identification. Even with spores showing homoge ssetescn amentation, the description must be more explicit so as to remove any ambiguity, e.g. “all faces reticulate.” The taxonomic consequences of this Pro” lem are similar to those encountered with species producing several megaspore types. Despite a lack of biological justification, it has generally been assum that the distal spore face represents the “official” character state and it is 0 the R. J. HICKEY: ISOETES MEGASPORES 13 basis of its ornamentation alone that species have been classified. A species like I. eshbaughii could be assigned to either sect. Isoétes or Palustres on the basis of distal spore morphology, whereas using the proximal ornamentation it could be included only in sect. Palustres. CONVERGENCE OF MEGASPORE SURFACE FEATURES A high degree of convergence in surface morphology is suggested by a con- sideration of the form and distribution of megaspore characters as discussed above. Further support for this conclusion, reviewed below, is derived from character compatibility, morphocline analysis, parsimony assessment, infraclade analysis, and hybridization studies (Hickey, 1985). The presence of morphologically similar megaspore types in many unrelated species is strong evidence that many ornamentation patterns are convergent. For instance, levigate megaspores are found in I. andicola, I. hypsophila Hand.- Mazz., I. hopei Croft, in the I. montezumae complex, the I. herzogii complex, the I. ecuadoriensis complex, and others. Comparative data from geographic distribution, subula shape, velum coverage, labium and ligule morphology, sto- mate distribution, scale and peripheral fiber development, microspore ornamen- tation, habitat data, and cuticle development indicate that these species and species groups represent evolutionarily distinct lineages. This conclusion is fur- ther supported by an examination of the levigate spore surface. In the I. mon- tezumae and I. herzogii species complexes, the spore wall is made up of a thick, evenly deposited exospore with little or no perispore externally whereas in the I. ecuadoriensis and I. lechleri complexes the exospore layer is thin and the perispore is well-developed. Similar examples of homoplasy are found among species with rugulate megaspores (I. savatieri, I. capensis, I. andicola, I. hiero- glyphica, etc.), reticulate megaspores (I. martii A. Br. ex Kuhn, I. killipii, I. engelmanii, etc.), and echinate megaspores (I. echinospora Dur., I. storkii, and I. andina). Convergence as an explanation for the presence of similar megaspores among various taxa is also supported by the repeated, parallel occurrences of morpho- logically distinct megaspore types within species, within individual plants, and within individual sporangia. The tuberculate-rugulate-retate morphocline, for example, is seen in I. savatieri, I. capensis, I. stephansenii, and I. muelleri, and a tuberculate-rugulate-levigate morphocline in I. andicola, I. montezumae, an I. herzogii. Additional examples of repeated morphoclines in apparently unre- ted taxa are abundant. It is highly improbable that taxa sharing identical or even similar patterns of megaspore polymorphism evolved from the same poly- morphic ancestors. Such a scenario would demand an untenable degree of con- vergence, parallelism and reduction in virtually every other character. lack of correlation between megaspore character states and other mor- Phological characters was first demonstrated by Berthet and Lecocq (3977). ies Owed a lack of significant correlation between megaspore and microspore ornamentation. In my own studies, similar comparisons between megaspore Or- tation and both vegetative and microspore characters failed to show sig- 14 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) nificant correlations. The strongest correlations found have been between tu- berculate megaspores and either echinate microspores or the absence of a well- developed velum. Not only are these correlations weak, all three of the character states in question are primitive and hence point to a grade level relationship only. There is no evidence to suggest that taxa sharing these primitive states represent an independent evolutionary lineage. Additional evidence for convergence is found among the diversity of mega- spore types produced by species within known monophyletic assemblages (as defined by uniquely derived character conditions). For example, the peculiar phyllopodia of certain African and European species represents a derived char- acter state relative to the sclerified scale (Hickey, 1986a). Species posessing this derived condition show tuberculate, rugulate, and retate patterns of ornamen- tation. It is postulated (Hickey, 1985, 1986a) that the rugulate and retate spore types evolved subsequent to the divergence of this clade and that similar spore types in other species alliances are not homologous. Even more striking in its megaspore variation is the Neotropical clade (which includes, among others, I. panamensis, I. gardneriana, and I. montezumae} characterized by a unique, trigonal subula (Hickey, 1985). The megaspore variation in this lineage ranges from tuberculate as the primitive state, to the more derived baculate, rugulate, levigate and pustulate conditions. As in the preceding example, the presence of similar spore types in other evolutionary lineages is considered to be the result of convergence. A final piece of evidence for “convergence” of spore morphology comes from recent hybridization experiments. Boom (1980) and Taylor et al. (1985) have shown that there are no sterility barriers between several species of certain sections of Isoétes. Their data suggest that species from different sections of Isoétes (as defined by spore morphology) are not as isolated as was previously believed. Furthermore, Taylor et al. (1985) suggest that the cristate megaspore pattern seen in I. eatonii Dodge, I. gravesii A. A. Eaton, and perhaps in I. riparia, was derived through hybridization between the reticulate-spored I. engelmannil and the echinate-spored I. muricata. Their hypothesis is supported by distribu- tion, cytological, enzymatic, and artificial hybridization data. Previously pub- lished chromatographic analyses of these taxa (Kott & Britton, 1982) are also compatible with their hypothesis. Considering the large number of polyploid taxa in Isoétes it seems likely that hybridization is a common source of mega- spore variation, in particular, as a source of intermediate spore morphologies. CONCLUSIONS Previous classifications of Isoétes have been based on habitat or inferred hab- itat preferences or on megaspore surface tation. The former system was found to be unacceptable because of the tremendous plasticity in leaf morphol- ogy and the lack of habitat specificity. Acceptance of the megaspore-based | sification has been due more to the lack of a viable, alternative classification and to a failure to evaluate critically infraspecific variability and character een tion than to any inherent strengths of that system. The variation described in R. J. HICKEY: ISOETES MEGASPORES 15 this paper indicates that the Isoé€tes megaspore is not as conservative as some have suggested (Kott & Britton, 1983). To the contrary, evidence is presented which suggests that megaspore variation is at least as plastic as vegetative char- acters. The reliability of any character is supported by corroboration with other char- acters that circumscribe the same taxa (that is, character correlation and its cor- ollary, the lack of incompatibility with other characters). Little consideration has been given to character compatability studies in Isoétes as a means of supporting or refuting the evolutionary relationships among species as indicated by mega- spore morphology. The isolated attempts to examine a wider assemblage of char- acters are especially noteworthy because they have shed light on the role of megaspore morphology in classification. The works of Wagner (1964), Evans (1968), and Boom (1982) indicate that megaspore characters, when used in con- cert with phenology and habitat preference, are useful for delimiting natural species alliances. Likewise, Hickey has shown that multiple character analyses using vegetative features alone or in concert with either habitat data or mega- spore morphology provide a basis for lower (1986b) and higher (1986a) taxonomic delimitations. Patterns of variation in megaspore ornamentation indicate that surface mor- phology is extremely liable to convergence. The evolutionary causes for such convergence are unknown but may be the result of: a) parallel responses to particular habitat requirements or environmental stimuli (such as shortened growing seasons), b) inherently limited genetic potential (which canalizes mor- phological potential), c) selection for synaptosporic dispersal mechanisms (see Kramer, 1977), d) stochastic events, or e) hybridization. At present, there are insufficient data to advance an infrageneric classification alternative to that based on megaspore morphology. It is first necessary to con- tinue work on a regional level by circumscribing natural species groups and species alliances. Such analyses should examine the variation, polarity, and evo- lution of the definitive characters. Individual character states must be strictly defined and, whenever possible, convergences identified. The reclassification of megaspores presented in this paper represents a first step in this process by clearly defining the primary states of this character and describing the major sources and patterns of variation associated with those states. ACKNOWLEDGMENTS T thank Gregory J. Anderson, Anne Bruneau, Jerry Gastony, Carl Taylor, and Rolla and Alice Tryon for their suggestions and criticisms during the preparation of this manuscript. This study was LITERATURE CITED Baker, J.G. 1880. A synopsis of the species of Isoetes. J. Bot. 18:65-70, 105-110. S : +P. and M. LEcoca. 1977. Morph } gi a : p > ¢ ses du gen Isoetes L. Pollen & Spores 19:329-349. 16 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) Boom, B. M. 1980. Intersectional hybrids in Isoétes. Amer. Fern J. 7 ——_——. 1982. Synopsis of Isoétes in the southeastern United ssi paetcon 47: 59. BRAUN, A. 1864. Les espéces d’Isoétes de I’Ile Sardaigne. Ann. Sci. Nat. Bot. V, 2:306-377 Cox, P. A. and i aa Hickey. 1984. Convergent megaspore evolution and Isoetes. Acne Naturalist 124:43 Crort, J. R. wat ee taxonomic revision of Isoetes L. (Isoetaceae) in Papuasia. Blumea 26:177-190. Durtuie, A. V. 1929. The species of Isoetes found in the Union of South Africa. Trans. Roy. Soc. South Africa 17:321-334. Eaton, A. A. 1900. The genus Isoetes in New England. Fernwort Papers 2:1-16. ENGELMANN, G. 1882. The aE Isoétes in North America. Trans. St. Louis Acad. Sci. 4:242-280. Evans, A. M. 1968. Isoétace: 8-10 in Manual of the ees flora of the Carolinas, ed. A. Radford et al. aed ain: Univedsliy of North Carolin Fucus, H. P. 1962. Nomenklatur, Taxonomie und a cradil ‘hee Gattung Isoétes Linneaus in geschichtlicher Betrachtung. Beih. Nova Hedwigia 3:1-117. Fucus-Eckert, H. P. 1981. i. Palmeri H. P. Fuchs, eine neue Isoétes-Art des Paramo. Proc. on. Ned. Akad. Wetensch. C84:165-17 Hickey, R. J. 1985. Revisionary studies of Sot Isoétes. Ph.D. dissertation, The University ITs. . 1986a. The early evolutionary and morphological diversity of Isoétes, with descriptions of two new Neotropical species. Syst. Bot. 11:309-321 ———.. 1986b. On the identity of Isoétes triquetra. Taxon (in press). Korr, L. S. 1980. The taxonomy and biology of the genus Isoetes in northeastern North America. Ph.D. thesis, einai of a rea Kort, L. S. and D. M. Britton. rison 2 chromatographic spot patterns of some North American — Sag ge ea Fore: and Spore morphology taxonomy of Isoetes in northeastern North America. Cone : Bot. 61:3140-3163 KRAMER, K. U. 1977. Synaptospory: a hyp thane A eae function of spore sculpture in pteri- dop . Straits Settlem. 30:79-8 MARSDEN, c R. lee . Morphological variation and ae of Isoetes Muelleri A. Br. J. Adelaide Bot. Gard. 1 Monpat, M. S. 1978. Sian morphology of Isoetes coromandelina L. from Kalyani. Bull. Bot. Soc. Bengal 32:8-12 Moretay, L. and A. VENDRYES. 1882. Monographie des Isoéteae. Actes Soc. Linn. Bordeaux 36:309- 06, t. 8-17. Pant, D. D. and G. K. ne 1962. The genus Isoetes in India. Proc. Natl. Inst. Sci. India, Pt. B, Biol. Sci. 28: and logia 30:239-251. PFEIFFER, N. E. 1922. Monograph of the waren Ann. Missouri Bot. Gard. 9:79-232. ‘ - R. 1949. Isoétes riparia and its variants. Amer. Fern J. 39:110-121. EED, C. F. 1965. Isoetes in southeastern United States. Phytologia 12:369-400. RoserT, D., F. FoLAND-HEYDACKER, J. LAROCHE, P. FOUGEROUX, and L. DAVIGNON. 19 La sei mégasporale de I’ Isoetes setacea Bosc ex Delile. Etude en microscopies photonaue et élec- troniques localisation et nature de la silice entrant dans sa constitution. Adansonla ia, sér. 2 13:313-332. Taytor, W. Ae N. T. LUEBKE, ooo B. SMITH. 1985. poeciaicn and hybridisation in North Amer = Roy. Soc. Edinburgh 86B:25 aces R M. pian Tao 1982. Ferns and allied ai 2 agian York: Springer-Verlag. a W. H., JR. 1964. Isoétaceae. Pp. 40-41 in Guide to the vascular flora of the Carolinas, ed- i. A. Radford et i Chapel Hill: University of North Carolina. 2, EST, naire sis i 1915. On Isoétes japonica A. Br. Trans. Linn. Soc. London, Bot., Ser. pon Cytology and reproduction of some Indian species of Isoetes. Cyto- wale Ee 8. pr On Isoetes australis $. Williams, a new Asie from western Australia. Part I Genera morphology. Proc. Roy. Soc. Edinburgh 62B American Fern Journal 76(1):17-20 (1986) Active Enzymes from Megaspores of Marsilea and Regnellidium Douctas E. SOLTIs and PAMELA 8. SOLTIS Department of Botany, Washington State University, Pullman, WA 99164-4230 Herbarium specimens have, of course, long been invaluable sources of mor- phological data. More recently, they have also been utilized in chemosystematic studies (e.g. Anderson & @vstedal, 1983; Giannasi & Niklas, 1981). Secondary compounds, such as flavonoids, are stable and are not degraded following des- iccation. Hence, herbarium specimens provide a wealth of potential data in such chemical analyses. However, herbarium specimens apparently have not been utilized in applications of the electrophoretic technique because living material is a prerequisite. The Marsileaceae offers an unusual opportunity to utilize her- barium specimens in electrophoretic analyses. For example, Johnson (1985) dem- onstrated that 99- to 100-year-old sporocarps of Marsilea oligospora still produced functional megaspores and microspores. To determine whether sporocarps also provide a potential source of isozyme data, we initiated an electrophoretic in- vestigation. MATERIALS AND METHODS Sporocarps were obtained from specimens (WS) of Marsilea vestita and Reg- nellidium diphyllum, scarified with a razor blade, and soaked in water to effect germination. Sporocarps were obtained from the following specimens: Ownbey $.n., greenhouse material (Regnellidium diphyllum); Old s.n., Harney Co., Ore- gon (Marsilea vestita); Davis 212, Grant Co., Washington (Marsilea vestita). After 24 hours the sorophore {in the case of Marsilea) and sporocarp wall were re- moved and the spore suspension filtered through a Buchner funnel under suc- tion. Megaspores (25-50) were then scraped into a small porcelain mortar. In the Case of Regnellidium, care was taken to ensure removal of all gelatinous material from the megaspores. Several sporophytes of Marsilea vestita were also exam- ined electrophoretically and the banding patterns obtained compared to those observed for megaspores. The megaspores from a single sporocarp were ground with a glass rod in a small drop of the tris-HCl grinding buffer-PVP solution of Soltis et al. (1983). Sporophytes of Marsilea vestita were prepared as previously outlined (Soltis et , 1983). For both sporophyte tissue and megaspores, the ground plant material was absorbed onto filter paper wicks and starch gel electrophoresis conducted following the general methods of Soltis et al. (1983). The following enzymes were Phate dehydrogenase ([NAD]G3PDH}, isocitrate dehydrogenase (IDH)}, leucine aminopeptidase (LAP), malate d (M YY 18 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) phoglucoisomerase (PGI), phosphoglucomutase (PGM)}, 6-phosphogluconate de- hydrogenase (6-PGD), shikimate dehydrogenase (SkDH), and triosephosphate isomerase (TPI). For all enzymes examined, a 12.5% starch gel was employed. ADH, AAT, EST, FE, LAP, ME, PGI, and TPI were resolved using the following modification of gel and electrode buffer system 8 of Soltis et al. (1983): electrode buffer com- posed of 0.039 M lithium hydroxide, 0.263 M boric acid, pH 8.0; gel buffer composed of 0.033 M tris, 0.005 M citric acid, 0.004 M lithium hydroxide, 0.030 M boric acid, pH 8.0. Gel and electrode buffer system 9 was used to resolve PGM and MDH. IDH, [NAD]G3PDH, and 6-PGD were resolved on system 1. All enzymes except ADH and LAP were stained following Soltis et al. (1983); ADH and LAP were stained following Soltis (1986). RESULTS AND DISCUSSION For all enzymes examined, the observed bands migrated anodally. Clear banding was observed for megaspores of Marsilea vestita for ADH, EST, FE, [NAD]G3PDH, IDH, MDH, ME, PGI, PGM, 6-PGD, and TPI. No activity was observed for SkDH; very faint activity was occasionally observed for AAT and LAP. Especially clear and intense banding was obtained for ADH, IDH, MDH, ME, PGI, 6-PGD, and TPI. The oldest sporocarp of M. vestita from which active enzymes were obtained was taken from a 32-year-old herbarium specimen. Only a small amount of sporocarp material of Regnellidium was available; hence, it was not assayed for all enzymes. Clear banding was achieved for ADH, ME, PGI, and TPI. The sporocarp of Regnellidium analyzed was taken from a 30-year-old specimen. Marsilea vesitata was investigated most intensively and results for each en- zyme follow (Figs. 1-7). For both sporophyte tissue and megaspores, similar banding patterns were observed. For ADH, three bands were expressed: two darkly staining, closely associated bands, and a more slowly migrating fainter band (Fig. 4). For EST and FE, two and three bands, respectively, were observed: Two bands were observed for each of the following enzymes of glycolysis and the pentose-phosphate pathway: [NAD]G3PDH, PGI (Fig. 5), PGM (Fig. 6), and 6-PGD. Four bands were consistently observed for MDH (Fig. 7). For TPL @ closely associated group of three bands, as well as a more slowly migrating band, poy aot (Figs. 1, 2). A single band was observed for both IDH and ME ig. 3). The major purpose of this investigation is to document that active enzymes — be derived from sporocarps (some of which may be as old as 32 years). It is also interesting to note, however, that the results for most enzymes are consistent with the number of loci expected in diploid seed plants (Gottlieb, 1981, 1982). It is now well established that, despite high chromosome numbers, ferns are i80- zymically diploid (see Haufler and Soltis, 1986, for review). The chromosome numbers reported for Regnellidium diphyllum (2n = 38) and species of Marsilea (n = 20) are low, however, compared to most homosporous ferns (Love et al., 1977; Tryon & Tryon, 1982). For PGI, PGM, 6-PGD, and [NAD]G3PDH, all en SOLTIS & SOLTIS: ENZYMES FROM MEGASPORES 19 f ot. 4 7 5 Fics. 1-7. graph g ing enzy ing f gas] f Marsilea vestita and Regnellidium diphyllum. Putative loci are numbered sequentially with the most anodally mi- i numbered 1. Fic. 1. Regnellidium diphyllum. Fics. 2-7. Marsilea vestita. Fics. 1-2. TPI, arrows between loci 1 and 2 indicate presumed interlocus heterodimer. Fic. 3. ME. Fic. 4. ADH. Fic. 5. PGI. Fic. 6. PGM. Fic. 7. MDH. zymes of glycolysis or the pentose-phosphate pathway, two isozymes were evi- dent in Regnellidium and Marsilea. Two isozymes are also observed for these enzymes in diploid angiosperms and gymnosperms (Gottlieb, 1981, 1982). For ADH and MDH, three and four bands, respectively, were observed in both sporophytes and megaspores. These observations are also consistent with the number of isozymes typically observed for these enzymes in diploid seed plants (Gottlieb, 1981, 1982). The only exception to diploid isozyme expression is TPI, which appears to have three, rather than two isozymes (the number typical of diploid seed plants) in both Marsilea vestita and Regnellidium diphyllum. Sim- ilar results for TPI have also been reported for most other ferns (Gastony & Gottlieb, 1985; Haufler, 1985; Haufler & Soltis, 1986). However, some ferns, such as species of Botrychium, do have the expected diploid number of two isozymes for TPI (Soltis & Soltis, 1986). The observation that active enzymes can be obtained from megaspores of Marsileaceae is comparable to the results of Gastony and Gottlieb (1982) and Haufler and Soltis (1984) who found that it is possible to obtain enzyme banding from individual gametophytes of homosporous ferns. These findings indicate that genetic analyses of Marsileaceae, as well as other heterosporous pteridophytes, such as Selaginella and Isoetes, can easily be conducted for heteromeric en- zymes by isolating megaspores from individual sporocarps and examining these pooled megaspores using standard electrophoretic procedures. pe Active enzymes were obtained from a 30-year-old sporocarp of Regnellidium diphyllum and from sporocarps of Marsilea vestita as old as 32 years. That active enzymes can be obtained from sporocarps of such age indicates that the Mar- sileaceae offers unusual opportunities for systematic study. Herbarium speci- Mens could provide a source of “living” material for electrophoretic investiga- tion. For example, herbarium specimens could be used to increase the number of populations surveyed. In addition, herbarium specimens could serve as an important source of rare taxa, or taxa that are difficult to obtain (such as Reg- nellidium). The Marsileaceae also affords unusual opportunities from a popu- 20 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) lation biology perspective. It might be possible, for example, to analyze temporal variation in allele frequencies in a single population. This would necessitate the availability of numerous collections from a single site spanning many years. ACKNOWLEDGMENT We thank Dr. Amy Jean Gilmartin for the use of herbarium specimens. LITERATURE CITED ANDERSON, O. M. and D. O. @vsTEDAL. 1983. Anthocyanin content in Saxifraga svalbardensis and some allied species. Biochem. Syst. Ecol. 11:239-241. dc G. J. and L. D. Gortiies. 1982. Evidence for genetic heterozygosity in a homosporous rn. Amer. J. Bot. 69:634-637. si ——. 1985. Genetic variation in the homosporous fern Pellaea andromedifolia. Amer. J. sesh 72:257-267. GiannasI, D. E. and K. J. NIKLAS. si open paleobiochemistry of some fossil and extant agaceae. Amer. J. Bot. 68:7 ibaa . pe 1981. Electrophore bn eae and plant populations. Prog. Phytochem. 7:1-46. . Conservation and duplication of isozymes in plants. Science 216:373-380. Flies c a 1985. Enzyme variability and modes of evolution in the fern genus Bommeria. Syst. Bot. 10:92-104. and D. E. Soxtis. 1984. Obligate segue in a homosporous fern: Field confirmation of a laboratory ee Amer. J. Bot. 71:878-881. Foal ———. 1986. Genetic evidence sei that rot gages ferns with high chro- me OES are diploid. Proc. Natl. Acad. Sci.: in JOHNSON, Dh M. 1985. New records for longevity of Marsilea sporocarps. Amer. Fern J. 75:30 Love, A., D. Love, and R. E. G. Picut SERMOLLI. 1977. Ciieiinintiiniiel atlas of the halen * uz: J. Cramar. SOLTIs, D. 5 1986. Isozyme number and enzyme compartmentalization in Equisetum. Amer. J. Bot.: in press and P. s. SOLTIS. 1986. arming oan of inbreeding in the fern Botrychium virginianum (Ophiogl r. J. Bot.: , C. H. Haurter, D. C. sida si G. J. Gastony. agree Starch gel electrophoresis of ferns: A compilation of grinding buffers, gel and electrode buffers, and staining schedules. Amer. Fern J. 73:9-27. Tryon, R. M. and A. F. Tron. 1982. Ferns and allied plants. New York: Springer-Verlag. American Fern Journal 76(1):21-24 (1986) Germinating Spores and Growing Sporelings of Aquatic Isoétes W. Car. TAYLOR and NEIL T. LUEBKE Botany Section, Milwaukee Public Museum, Milwaukee, WI 53233 Several techniques have been described recently for germinating spores of aquatic species of Isoétes which grow in lakes, ponds, and streams in eastern North America. Boom (1980) used glass vials containing sterile sand and pond water in which he successfully raised sporelings of I. engelmannii, I. flaccida, I. macrospora, and I. riparia. Kott and Britton (1982) germinated spores of I. acadiensis, I. echinospora, I. macrospora, I. riparia, and I. tuckermanii in small vials half filled with distilled water. However, many of their vials became con- taminated with fungi that destroyed the viability of the megaspores. Sam (1982) germinated megaspores of I. engelmannii in an inorganic nutrient medium sup- plemented with Nystatin to inhibit fungi. Webster (1979) described a technique for germinating spores of Selaginella that is similar to the following procedure we use for Isoétes. We have been culturing plants from spores of Isoétes acadiensis, I. echino- spora, I. engelmannii, I. hieroglyphica, I. macrospora, I. riparia, and I. tucker- manii for use in germination and hybridization experiments for several years. The aquatic species of Isoétes are ideal pteridophytes for germination and hy- bridization experiments. Their sporangia are large and contain many spores. Megaspores and microspores are borne in different sporangia on separate leaves. If necessary, unopened sporangia can be excised from sporophylls intact and surface cleansed to eliminate extraneous spore contamination. Thus, megaspores and microspores can be completely isolated for controlled breeding experiments. Also, their spores readily germinate in demineralized water and gametophytes and sporophytes develop normally for several months without supplemental nutrients. Spores are obtained from plants collected in September and October, since plants harvested earlier in the year may not have mature, viable spores. Isoétes hieroglyphica and I. macrospora retain their spores through the winter, so ma- ture spores of these species can also be gathered in the spring. | Megaspores are cleansed of microspores in a sieving apparatus made of plastic tubing approximately 10 cm long and 2.5 cm in diameter. An 8 cm square of 0.27 mm sifter mesh (available from Carolina Biological Supply Company) is fitted to one end of the tubing with a rubber band. The sieving apparatus, containing Megaspores, is attached to a water faucet using plastic strapping tape. Spores are washed with a steady flow of cold water for 30 minutes, followed by a one minute rinse with sterile, demineralized water. Although megaspore surfaces appear clean after the pened sporangia are preferred for crossing experiments chance of contamination from microspores. Intact sporangia above treatment, un- because there is less can be excised from 22 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) sporophylls using a sharp scalpel and fine forceps. Outer whorls of sporophylls contain the most mature spores. By cutting on the abaxial side along its point of attachment to the corm, a sporophyll can be removed without rupturing the sporangium. Leaf tissue is carefully peeled away from the sporangium using forceps (Fig. 1). Excised sporangia are cleaned with a sieving apparatus in the manner described above for megaspores. Sporangial walls of I. riparia, I. mac- rospora, and I. hieroglyphica are relatively strong, making their sporangia easy to remove without rupture. Sporangial walls of I. echinospora, I. engelmannii, I. tuckermanii, and I. acadiensis weaken rapidly when their spores are mature. This makes extraction of intact sporangia more difficult. If unopened sporangia of these species are desired, plants should be collected and their sporangia extracted by mid-September. Plants of all aquatic species may be kept for a week or more by rinsing, draining, and sealing the plants in plastic bags which are kept cool until the sporangia are removed. Using aseptic technique, spores or sporangia are transferred with forceps to plastic tissue culture dishes. We use 35 mm x 10 mm size dishes (Corning 25000) each containing 4 ml of sterile, demineralized water. Intact megasporangia are opened and their contents dispersed. Generally, 30 to 60 or more megaspores are put in each dish. Microspores may be added to the cultures by briefly switl- ing part of a ruptured microsporangium in each dish. Dishes are sealed with 12 mm X 100 mm strips of parafilm to prevent evaporation of the water. Cultures are vernalized in the dark for 100 days at 2°C prior to their incubation at 20°C with a 12L:12D photoperiod using 20 watt fluorescent grow lights 25 cm above the cultures. Experiments indicate that only I. echinospora requires this vernalization. Although spores of the other species germinate without cold treat- ment, vernalization accelerates germination and sporophyte formation. When the above procedure is followed, the first megagametophytes can be expected within 10 days after the start of incubation. A few days later, arche- gonia can be recognized by brown pigmented quartets of neck cells on the megagametophytes (Fig. 2). Cultures containing both megaspores and micro- spores normally develop sporophytes within 10 to 30 days depending on the species (Fig. 3). In our cultures, I. echinospora, I. engelmannii, and I. riparia usually form gametophytes and sporophytes sooner than I. acadiensis, I. hier0- glyphica, I. macrospora, and I. tuckermanii. Initially, we used the fungicide Nystatin in our cultures at 100,000 units/liter of sterile, demineralized water, but we have found it unnecessary in subsequent experiments. Using the above procedure we have had little problem with fun contamin ation. It is also unnecessary to add nutrients to the cultures. Indeed, — the addition of nutrients can increase the growth of any contaminating orgen isms. Megaspores of Isoétes contain sufficient food reserves to support devel- witht oo pees —. bel siennaae techniques in Isoétes. Fic. 1. Excising megasporangium from sporophyll. Fic. - : mel tophyt imately 120 days old; bar = 1 ™™- a ehaias ¢ Sat 3. Sporophytes, approximately 90 days old. TAYLOR & LUEBKE: ISOETES 23 24 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) oping gametophytes and sporophytes for several months. By that time sporo- phytes are old enough to be transferred to a sand, gravel, and peat substrate in larger dishes. We have had the most success when a young sporophyte is transferred shortly after the emergence of its first root. At this stage we move the sporophytes to glass storage dishes (Corning 3250) containing a 2.5 cm layer of substrate. The substrate is a mixture of 4 parts fine sphagnum peat, 3 parts fine quartz sand, and 3 parts medium quartz sand. Sporophytes are placed on top of this mixture and covered with a thin layer of medium quartz sand. The additional layer of sand anchors young plants to the substrate and promotes rooting. This is impor- tant for continued sporophyte development because floating plants slowly de- teriorate. Demineralized water is added to a level approximately 1 cm above the sand. After water is added and any floating plants anchored, each dish is covered with a glass lid to reduce evaporation. The young sporophytes are cultured at 18°C with a 15L:9D photoperiod using 20 watt fluorescent grow lights 20 cm above the dishes. As the plants continue to grow water depth is gradually increased. Within a year the sporophytes are transferred to 21” (6.4 cm) square plastic pots contain- ing the substrate described above. This substrate is covered with a thin layer of quartz gravel to keep the peat from floating as the potted plants are submerged in 20 gallon aquaria. Nutrients for plant growth are provided by several small fish living in each aquarium. May through October, the aquaria are maintain at approximately 18°C with a 15L:9D photoperiod using four 30 watt fluorescent lights above each tank. December through March, the aquaria are kept at 2°C and each day receive 8 hours of very low light from two fluorescent ceiling lights. April and November are transition months during which the temperature and daylength are gradually adjusted to the above conditions. Sporophytes of aquatic Isoétes, collected from their native habitats, are grown under the same conditions. The methods described in this paper for germinating spores and growing sporelings of Isoétes are simple. Using these procedures, large numbers of 8@- metophytes, and sporophytes of selected parentage, are easily produced. Ger- mination studies and crossing experiments can be set up without contamination from unwanted spores and resulting sporophytes can be grown for further study. LITERATURE CITED Korr - sae oper — hybrids in Isoétes. Amer. Fern J. 70:1-4. a _D. M. Brirron. 1982. A comparative study of spore germination of some I species of northeastern North America. Canad. ]. Bot. 60:1679-1687. shin S.J. 1982. A germination method for Isoétes. Amer. Fern J. 72:61. EBsTER, T.R. 1979. An artificial crossing technique for Selaginella. Amer. Fern J. 69:9-13. SHORTER NOTES Pellaea brachyptera New to Washington.—On 9 July 1984, the Sierra cliff brake, Pellaea brachyptera (Moore) Baker, was encountered while conducting a botan- ical survey of the Lake Chelan-Sawtooth Wilderness for the Washington Native Plant Society. Vouchers were collected (Alverson 799, MICH, ORE, UC, US, WS, WTU) from a large colony above the Lakeshore Trail #1247, 1 km north of Prince Creek, in the Wenatchee National Forest of Chelan County. The Wash- ington populations are restricted to a small area along the upper reaches of Lake Chelan, an 88 km long fiord-like body of water that occupies a glacial trough extending from the Cascade Range on the northwest to the arid Columbia Basin on the southeast. This cliff brake was previously known to range only from northern California (Placer and Lake counties) north to southwestern Oregon (Lane County) (Munz, A California flora and supplement, 1973; D. H. Wagner, pers. comm.). The Chelan County populations represent a northerly outlier some 500 km disjunct from the nearest population in southwestern Oregon. Thus, P. brachyptera is a new addition to the flora of Washington, as well as to the Pacific Northwest flora as delineated by Hitchcock and Cronquist (Flora of the Pacific Northwest, 1973). Actually, P. brachyptera was first collected in Washington by Ralph and Dor- othy Naas in 1971, along the Prince Creek trail about one mile up from Lake Chelan. The specimen, #938b (deposited in the North Cascades National Park herbarium), was misidentified as Pellaea glabella var. simplex, a species found on shaded basalt cliffs in the nearby Columbia Basin. We observed the greatest concentration of P. brachyptera plants on south- to west-facing slopes between Prince Creek and Rattlesnake Creek, at elevations of 330 to 600 m, in the south half of Section 29, Township 31 N, Range 19 E. Total number of plants in this population, spread over at least one hundred hectares, probably exceeds one thousand. Further surveys along the Lakeshore Trail revealed the presence of four smaller populations, extending as far uplake as Round Mountain, about 11 km northwest of Prince Creek. Field sighting forms giving the precise locations of these additional populations have been filed with the Washington Natural Heritage Program in Olympia, WA, which has added P. brachyptera to the list of sensitive plant species in Washington. Along Lake Chelan, Pellaea brachyptera grows on open, dry, south- to west- facing slopes that support sparse stands of Pinus ponderosa and Pseudotsuga menziesii. Associates include Purshia tridentata, Agropyron spicatum, Bromus tectorum, Balsamorhiza sagittata, Achillaea millefolium, Hieracium albertinum, and Collomia grandiflora. Pellaea brachyptera grows in rocky soil and around the base of boulders, a habitat shared with the equally abundant Aspidotis densa. The soils are derived from the regional bedrock, Swakane Biotite Gneiss, as well as glacial deposits such as till and unsorted outwash (Cater & Wright, U.S.G.S. 8eological map, 1967). The exposed slopes serve to accentuate the effects upon the vegetation of the regional climate, which is characterized by hot, dry sum- mers and cool, snowy winters. Individual plants of P. brachyptera are often 25 26 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) quite robust, in some cases forming clumps up to 1 m in diameter, with hundreds of fronds. The habitat has been degraded as a result of many decades of grazing by sheep and horses, and the possibility of this population having originated as an accidental (or even deliberate) introduction must be considered. However, the likelihood of this having happened is slight. The Washington populations appear to be both too remote, and too well established, to have originated within the last century. A more plausible interpretation is that they are remnants of a more continuous distribution along the eastern base of the Cascade Range. Interestingly, several flowering plants exhibit a pattern of disjunction similar to that of P. brachyptera. Arctostaphylos patula (Ericaceae), Spiranthes porri- folia (Orchidaceae), Pectocarya pusilla (Boraginaceae), and Githopsis specula- rioides (Campanulaceae) are among plants of southern distribution that occur in disjunct populations along the upper reaches of Lake Chelan. In general, plants of southern affinities migrated northward and reached their northern range lim- its during the hypsithermal interval that followed the close of the last glacial period (Detling, Bull. Mus. Nat. Hist. Univ. Oregon 3:1-57, 1968). Most probably, Pellaea, Arctostaphylos, Spiranthes, Pectocarya, and Githopsis extended their ranges northward to Lake Chelan during this warmer period. As the climate subsequently cooled to present conditions, these plants were able to persist lo- cally along Lake Chelan, while connecting populations to the south were extir- pated. The great diversity of habitats and vegetation along Lake Chelan is an indication that the area is particularly suitable for the retention of relicts. The influence of Lake Chelan in moderating cold winter temperatures is likely to have been a factor as well. Thus, the Chelan County populations of Pellaea brachyptera are interpreted here asa northern outlier of a once more or less continuous series of populations. This is in contrast to previously documented fern disjunctions, such as Polysti- chum scopulinum (Wagner & Rouleau, Amer. Fern J. 74:33-36, 1984), and Pellaea wrightiana (Wagner, J. Elisha Mitchell Sci. Soc. 81:95-103, 1965), in which long distance spore dispersal is suggested as the most probable means of origin. How- ever, it should be noted that these disjunctions follow the prevailing (west !0 east) direction of upper atmosphere air movement. The Washington populations of Pellaea brachyptera, in contrast, are located north of the species primary range, and south to north atmospheric air currents are relatively uncommon. The authors thank Ron Taylor and Joe Miller of the Washington Native Plant Society for originating and supporting the Chelan-Sawtooth field survey, and Ralph and Dorothy Naas for providing information on their original sighting © Pellaea brachyptera.—Epwarp R. ALVERSON, 12530 SE 47th PI., Bellevue, WA 98006, and JoOsEPH ARNETT, Biology Dept., Western Washington University, Bel- lingham, WA 98225. A Second Eastern North American Occurrence for Forked Spleenwort, AS plenium septentrionale.— An unusually interesting group of North i eee ferns are those that have their main occurrence in western North America SHORTER NOTES en one or a few outliers hundreds of kilometers to the east (Wagner, W. H., Jr., Ann. Missouri Bot. Gard. 59:203-207, 1972). They include such species as Cheilanthes castanea, Pellaea wrightiana, Aspidotis densa, Polystichum scopulinum and the forked spleenwort, Asplenium septentrionale (L.) Hoffm. The last of these was considered until recently to be entirely a western species. Its easternmost local- ities were in southwestern South Dakota, Colorado, western Oklahoma, and Texas (Lellinger, D. B., A field manual of the ferns and fern allies of the United States and Canada. Smithsonian Inst. Press, 1985), until this unusual spleenwort was discovered far to the east near Waiteville, Monroe County, West Virginia (Emory, D. L., Amer. Fern J. 60:129-134, 1970). On 9 June 1985, while botanizing along an unpaved road in Hardy County, West Virginia, I encountered a small stand of plants that I thought from a dis- tance to be a tall cluster of mosses. Looking closer on the rocky ledge, I recog- nized immediately that I had discovered an additional population of Asplenium septentrionale for West Virginia. This find represents the second stand some 2000 kilometers from the nearest known station in western Oklahoma and is approximately 225 kilometers northeast of the Monroe County locality. At the time of this discovery the Brooks Bird Club was conducting a foray in Hardy County, so that several members were able to substantiate my identifi- cation. Later Rodney Bartgis, William Grafton, and Robert Richardson checked the surrounding territory within a radius of 45 meters from my first colony of the forked spleenwort, and they located 120 living plants, but all of the rock outcrops were not explored and later additional plants were found in September. Twenty dead plants were also noted. My second trip to the locality was on 29 September, when I was accompanied by W. H. Wagner, Jr., F. S. Wagner, and Kyle Bush. We not only explored more outcrops but we also recorded additional associated plants. The topography of the Hardy County site varies from sandstone rock outcrops to loose crumbling shale that predominates the area. Rocks are of the Hampshire formation, a series of red non-Devonian shales and sandstones (Cardwell, D. H., Geologic history of West Virginia. Geological and economic survey, 1975, Pp. 64). Unlike the plants at the Monroe County locality, which grow on loose chips of Shale (Emory, 1970, fig. 1), ours grow on much firmer rock, usually in horizontal cracks in otherwise continuous, unbroken rock surfaces. The forked spleenwort Population occurs at an elevation of 730 to 850 meters. The rock outcrops face southward and the plants tolerate various exposures from 0 percent where the tock ledges are at the edge of the woods to 20-70 percent in the woods. In order to define the habitat as well as possible, the following list of associated species was prepared (those marked with asterisks are especially prominent). Woody plants include Amelanchier arborea, Crataegus uniflora, P arthenocissus *Quercus coccinea, *Q. ficinale, * Danthonia sp., Draba ramosissima, Eupatorium rugosum, Fragaria vir- 8iniana, Heuchera sp., Hieracium trailii, Houstonia longifolia, Penstemon 28 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) laevigatus, Phlox subulata, Saxifraga virginiensis, Silene pensylvanica, Sisym- brium altissimum, and Solidago caesia. The plants of forked spleenwort at this locality are not only much more nu- merous but are much larger than those at the Monroe County locality. Fronds reach to over 12 cm and the largest tufts measure 10 cm in diameter with hundreds of fronds. Practically all of the fronds are forked only once and only near the tip in the apical 1 cm in contrast to the form of the species that forks several times and much lower on the rachis (e.g., Page, C. N., The ferns of Britain and Ireland. Cambridge Univ. Press, 1982, fig. 13). Other ferns found on the rock outcrops are much less common. Asplenium platyneuron is the most numerous, but most plants of that grow on soil; only a few are found upon the rock surfaces. Asplenium montana, A. trichomanes, Dennstaedtia punctilobula, Polypodium virginianum, and Woodsia obtusa are all found on the same outcrops as A. septentrionale, but except for Woodsia are very rare in the area, and all of them are more or less dwarfed. A hybrid involving A. septentrionale with possibly A. platyneuron, A. trichomanes, or A. montana was encountered first by Rodney Bartgis and later by Florence Wagner. It is currently being analyzed and will be reported in the near future. Emory (1970) noted that it is entirely possible that other stations of this rare fern would be discovered, as there are numerous similar habitats. His prediction has now come true. It is also possible that other stations will be discovered between the Monroe County and the Hardy County stations and beyond. The list of associates given here should help in recognizing potential habitats. At the same time I emphasize Emory’s statement regarding protection of this plant. I urge the greatest caution in maintaining these plants in their natural state without disturbance or vandalism. Vouchers and exact locality data are on record in West Virginia University Herbarium, Morgantown, WV.—ELEANOR M. Bush, 5 Bush Avenue, Philippi, West Virginia. Second Locality for Dryopteris carthusiana in Arkansas.—The spinulose wood fern, Dryopteris carthusiana (Villars) H. P. Fuchs [D. spinulosa (O. F. Muell.} Watt], was first reported for Arkansas by Lesquereux (A catalog of the plants of Arkansas. Pp. 346-399 in Owen, Second report of a geological reconnaissance of the middle and southern counties of Arkansas made during the years 1 and 1860, 1860) from “woods.” Based on this report, it was included in state lists by Harvey (Bot. Gaz., Crawfordsville 6:188-190, 1881) and Branner and Coville {A list of the plants of Arkansas. Pp. 155-242, in Branner, Annual report of the geological survey of Arkansas for 1888, Vol. IV, 1891). Buchholz (Amer. Fern J. 14:33-38, 1924) expressed doubt about the presence of this species in Arkansas, because he could not locate voucher material. Based on a discovery by M. Moore in 1924, Buchholz and Palmer (Trans. Acad. Sci. St. Louis 25:91-158, 1926) reported this species from the north side of Mt. Magazine, Loga? Co. Arkansas. Recent efforts to locate a voucher or plants at that locality were as : unsuccessful (Taylor & Demaree, Rhodora 81:503-548, 1979; Taylor, Arkans0s ferns and fern allies, 1984, p. 106). The only verifiable population of D. car" — SHORTER NOTES 29 siana in Arkansas was discovered by D. M. Moore on 7 August 1960, in Stone Co., Arkansas, where approximately a dozen plants occur at the entrance of Rowland Cave. This population represents the most extreme southwestern pop- ulation of this species in eastern North America (Carlson & Wagner, Contr. Univ. Michigan Herb. 15:141-162, 1982). On 5 October 1985, while surveying the status of Woodsia scopulina D. C. Eaton on Mt. Magazine, I located 4 plants of D. carthusiana on the northside of the mountain, near its summit, in the vicinity of Brown’s Spring. This population, associated with three other fern species also occurring as peripheral populations [Dennstaedtia punctilobula (Michx.) Moore, Dryopteris marginalis (L.) Gray, and Woodsia scopulina D. C. Eaton], is most probably the population initially dis- covered by Moore in 1924. Verification of the occurrence of a Mt. Magazine population extends the known range of D. carthusiana 300 km to the southwest of the Stone Co. population. The occurrence of this northern species in Arkansas appears to be related to “northern” environmental factors provided by elevation (860 m) at the top of the tallest mountain in Arkansas and by moderated, cool, moist air blowing from a cave entrance. Based on the known locations of D. carthusiana in Arkansas, it is most improbable that Lesquereau ever saw this species in Arkansas during his travels; the earlier attributions of this species in the Arkansas flora must be considered spurious. This research was sponsored, in part, by the State of Arkansas’ Nongame Preservation Program Committee, by a faculty research grant from Office of Research and Sponsored Programs, and by the College of Science, Office of Research, Science and Technology, at the University of Arkansas at Little Rock.— JAMEs H. Peck, Department of Biology, University of Arkansas at Little Rock, Little Rock, AR 72204. REVIEWS “A field manual of the ferns & fern-allies of the United States and Canada,” by David B. Lellinger with photographs by A. Murray Evans. 1985. ix + 389 Pp. + 45 pp. of color photographs. Washington, D.C.: Smithsonian Institution Press. ISBN 0-87474-602-7, $45.00 (hardcover). ISBN 0-87474-603-5, $29.95 (pa- Perback). This manual, containing keys, descriptions, and photographs, provides a means to identify the ferns and fern-allies of the contiguous United States and Canada. our hundred and six native and naturalized species, subspecies and varieties are treated. Three hundred and forty-one of these are illustrated by color pho- tographs. The well written introduction includes valuable information on collection, Classification, distribution, ecology, structure, life history, and culture of pteri- dophytes. Sections on classification and paleontology are particularly good. Statements on culture are useful to those who grow ferns and fern-allies. 30 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) Keys and descriptions constitute most of the text. A key to families directs the user to family descriptions which are followed by keys to genera. These keys guide one to a generic description and a key to the species of that genus. For each species there is a scientific and common name, a description of diagnostic characters, and statements about habitat, abundance, distribution, and culture. Species are numbered uniformly in the keys, descriptions, and photographs to make comparisons easy. Following the species descriptions, there is a unique section titled “Hybrid Complexes” where the author graphically presents results from biosystematic studies of species complexes in the form of hypothetical pedigrees. Systematists will find this part particularly interesting. The color photographs in this work are impressive not only for their quality but for their number. Anyone who has attempted to photograph ferns and fern- allies can appreciate the effort by Dr. Evans. Users will find these pictures helpful in confirming their determinations. The book concludes with an excel- lent, illustrated glossary, an extensive bibliography, a checklist of species, lesser taxa, and hybrids, and separate indices to common and scientific names. The keys, constructed of well chosen, contrasting characters, are a pleasure to use. In addition, keys are numbered to facilitate backtracking if the wrong state- ment of a couplet is chosen. With this book and a minimum of experience, nearly anyone can successfully identify a pteridophyte. Drs. Lellinger and Evans have combined their talents to produce a scholarly, informative, and very usable text- book on ferns and fern-allies that certainly should be in the library of every professional and amateur pteridologist. Even those who do not yet know and love ferns could be enticed to our ranks by this fine volume.—W. Cart TAYLOR, Botany Section, Milwaukee Public Museum, Milwaukee, WI 53233. “Biology of pteridophytes,” edited by A. F. Dyer and C. N. Page. 1985. 474 pp. Proc. Royal Soc. Edinburgh, Section B, Volume 86. Edinburgh: The Royal Society of Edinburgh. ISSN-0308-2113. This volume presents the proceedings of an international symposium held at the University of Edinburgh in September, 1983. The purpose of the symposium was to consider pteridophytes as living organisms as demonstrated both through field-oriented and laboratory studies and to maintain the momentum of the cur- rent interest in pteridophyte biology. Contributions cover an extremely wide range of topics from cell and molecular biology, genetics, water relations, and development, through ecology and diversification, modes of speciation, life-his- tory studies, conservation, and the biology of economic ferns. There are 57 af pers and 37 abstracts of poster demonstrations by 75 authors from 22 countries. Almost all of the contributions are short and leave the reader with a great desi for more complete information. However, they represent the papers given and reflect the current state of pteridophyte research as of 1983, and many conta!” a wealth of literature citations, Pteridophytes in the broad sense were the first colonizers of land and today they are the second most diverse group of vascular plants after the angiosper™S REVIEWS 31 An important fact emerging from current research is the realization that pteri- dophytes exhibit many basic processes which, with more complete understand- ing, will provide phylogenetic meaning for features found in higher vascular plants, and that the key to understanding the life-cycle of seed plants lies in understanding facets of the life-cycle of the modern day remnants of their distant ancestors. For example, gametophyte/sporophyte antagonism, perhaps induced by their interaction, may represent the initial process leading to incompatibility in flowering plants. The fixation and retention of mutational load, promoted by the apical meristem of leptosporangiate ferns, may relate to selection for larger and more complex meristems in other plants providing increased buffering against disadvantageous mutations, and, indeed, even to outbreeding mechanisms. An- theridiogens are a part of this process and it is suggested that products of me- tabolism in ferns have become phytohormones in which a gibberellin-like sub- stance acts as a signal for morphogenesis and that these substances are the ancestors of the gibberellin hormones which influence seed plant development. The co-evolution of insects and plants, which has culminated in the pollination relationships of flowering plants, may have had its earliest beginnings with spore- eating and transport by Devonian and Carboniferous arthropods. Chemical com- ponents and biochemical pathways of spore germination are similar to those observed in seeds and represent evolutionary experiments that took place during the evolution of the seed. There are a number of other fascinating studies detailed in these proceedings which stimulate hypotheses on the phylogeny and evolutionary success of pteri- dophytes. Tannins and phenols are common in ferns. Ferns apparently have a much more limited array of anti-herbivore and anti-microbial chemicals com- pared to angiosperms, although many classes do have such activity. To date, however, there is only negative evidence in terms of a direct chemical response to herbivory, and insects seem to be responding to ferns rather than ferns to insects. The multi-layered structure of the male gamete bears remarkable resem- blance to a flagellar root structure found in the motile phase of Coleochaete-like algae. Pteridophytes, bryophytes, gymnosperms, and angiosperms all have sim- ilar changes in cytoplasmic organelles and cytoplasmic ribosomes during the gametophyte-sporophyte transition. However, within pteridophytes, the varia- tion and interaction is much greater than in any other plant group. Comparative sperm morphology reveals similarities between Selaginella, bryophytes, and Marsilea, which are profoundly different from the similarities shown by Lyco- Podium, Equisetum, Ginkgo, homosporous ferns, and cycads. This suggests that Lycopodium and Selaginella are phyletically distantly related and that homo- Sporous and heterosporous ferns are unrelated. o Recent studies in physiology and water relations are providing new clues by which we may gain a greater understanding of pteridophyte ecology. Ability ” withstand vegetative desiccation is more common in pteridophytes than in seed Plants in both the sporophyte and gametophyte generations. This is a sayin shared with bryophytes and algae. In addition, sporophytes and gametophytes are genotypically shade-tolerant and can grow at very low photon flux naperoawed (a feature that might explain the abundance of fern spores immediately above 32 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 1 (1986) the Iridium layer at the Cretaceous/Tertiary boundary). In addition, photosyn- thesis of sporophytes is not saturated by normal atmospheric carbon dioxide levels, even at low oxygen tension, a feature perhaps adaptive at the time of the initial radiation of land plants. Although their rates of efficiency suggest that pteridophytes are capable of growing under stress, their low potential growth rate results in a general inability to become effective annual plants. Thus, they are not often capable of exploiting frequently disturbed habitats. Other studies suggest that hybrids are common in Isoétes, Equisetum, Botrych- ium, and Lycopodium, and that outbreeding in taxa with underground gameto- phytes must be more prevalent than commonly believed. Pteridophyte hybrids in the tropics occur most frequently in disturbed habitats but their relative fre- quency appears to be considerably less than in temperate regions. Two new approaches in fern research show great promise for clarifying many of our current controversies in systematics and evolutionary biology. These in- clude molecular advances in chloroplast DNA hybridization and protein enzyme analysis and life-history strategies combined with basic physiology. The molec- ular approaches hold great promise for clarifying systematic and phyletic rela- tionships as well as uncovering patterns of population genetic diversity. Studies of breeding systems and life-histories will do much to increase our understand- ing of the ecology of pteridophytes as a whole. This review only touches upon the new research and ideas presented in this volume. Taken together, the papers indicate that current research in pteridology is important to and is of interest to the botanical public as a whole. Those of us who study this group of organisms will benefit greatly by the breadth of infor- mation contained within.—JupirH E. Skoc, Department of Biology, George Ma- son University, Fairfax, VA 22030; and Ropert M. Lioyp, Department of Botany, Ohio University, Athens, OH 45701. INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable), including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 812 by 11 inch paper of good quality, not-“erasable” paper. Double space ut, including title, authors’ names and addresses, text {including “sal and keys), literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author’s name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. Use S.I. {metric} units for all mea- sures (e.g., distance, elevation, weight) unless quoted or cited from another source (e.g., specimen citations). For nomenclatural matter {i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- — breviate titles of serial publications according to Botanico-Periodicum-Huntian- um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as part of nomenclatural matter are not included in lit- erature cited. For shorter notes and reviews, put all references parenthetically ogee Use I ndex herbariorum (Regnum Veg. 106:1-452. 1981) for designations of herbaria Hhistrations should be proportioned to fit page width with caption on thesame Page. Provide margins of at least 25 mm on all illustrations. For continuous-tone illustrations, design originals for reproduction with amount. In composite blocks, abut edges of ‘adjacent photographs “Avoid com- bining | Sequence and n b J of ables: ith order of citati in text. Explain scales and ping in figures themselves, not in | captions. Include . ‘Scale and reference to latitude and longitude in each map co __ Proofs and reprint order forms are sent to authors . es printer. Authors a | Send corrected proofs: to the editor and — — to the — : Author will be ass gi nsive alterations n Bool oo "For other matters a fot or ee es nsult recen = ot The Grae sonal a ra BOOKS ON FERNS An extensive list of floras, monographs, and horticultural manuals is avail- able for two first-class stamps. I also buy used fern books; please send list of authors, titles, and dates. Myron Kimnach, 1600 Orlando Rd, San Mar- ino, CA 91108. _ BRITISH PTERIDOLOGICAL SOCIETY ie Open to all who are interested in growing and studying ferns and fern- ae | allies. Full members receive THE FERN GAZETTE and the BU 4 _ OF THE BRITISH PTERIDOLOGICAL SOCIETY. Membership subscrip- - tons are £7 for fll members, £5 for ordinary members (not receiving the | ! ITE}, and £ 5 for student members (under 25 years of age)- For | s, U.S. senna should apply to Dr. David S. Barrington, Pringle _ artment of Botany, University of Vermont, Burlington, VT oe sidents should apply to Hon. Gen. Sec., % British Pte — oneees History), Cromwell Road, Lom: AMERICAN ao! FERN eae JOURNAL et Ay fom thy et ty ne a Ae ee eae QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY Three New Species of Moonworts (Botrychium subg. Botrychium) Endemic in Western . North America W. H. Wagner, Jr.and F.S.Wagner 33 : Gametophytes of Lycopodium lucidulum from Axenic Culture Dean P. Whittier and Terry R. Webster 48 Revision of the Neotropical Fern Genus Cyclodium Alan R. Smith 56 Shorter Notes A New Substrate for Ophioglossum palmatum in Florida Clifton E. Nauman and Richard Moyroud 55 ; Cystopteris tennesseensis in West Virginia Allison W. Cusick $9 Corrections to Index Filicum — 47 . Information for Authors SS oc Coes The American Fern Society Council fo FLORENCE S. WAGNER, Dept. of Botany, anes Ne cs Ann Arbor, MI 48109. President JUDITH E. SKOG, Biology Dept., George Mason University, sou Ly 22030. Vice-President W. CARL TAYLOR, Milwaukee Public Museum, Milwaukee, WI 5 Secretary JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, - au T™ sf? oe DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Burlington, VT 0 pena Trae JAMES D. MONTGOMERY, Ecology III, R.D. 1, Berwick, PA 18603. Back Issues Curator ALAN R. SMITH, Dept. of Botany, University of California, goes vi 94720. Chat Editor DAVID B. LELLINGER, Smithsonian Institution, Washington, DC 2 emoir Editor DENNIS Wm. STEVENSON, Dept. of Biological Science, Barnard pea Columbia vial New York, NY 10027. Fiddlehead Forum Editor Soraiagee’ oe Journal Pe ORE Dept. of Botany, —— of California, Berkeley, CA 94720 ASSOCIATE EDITORS Dept. of Biology, a University, Bloomington, IN 47401 AES ee a ea Dept. of Botany, University of Kansas, _ ce, KS 66045 ee .,.. U.S. Nat’] Herbarium NHB-166, ee Institution, ashi 20560 _ TERRY R. WEBSTER ____ Biological Sciences Group, University of Connecticut, ark CT 06268 aie ‘American Fern Journal” ose Haein is an illustrated quarterly devoted to ny — a n Fern Society, and published at the Pringle H Lane of some. 1 Bingo, V1 VT 05405. Second-class postage paid at Burlington, “VI. and : : MUSSING 1s Je 6 rn ths fd tic) 419 fter th date of issue, oe ca a ae ee . Ce ee ee oe back 3 sl e addressed t to Dr. James D. Montgomery, Ecology III, R.D. 1, Berwick, PA 18603. a fed address, dues, and pplic tions f bership should be sent to the Records Trea- Qr "Ce ginning os should be addressed to the ce s gross, -50 net if paid through an agency ssi fee $0.50); sent free to — "members of the American Fem ociety herent ey $10.00 + $4.00 mailing beyond USA, A. S.A., Canada, and Mexico; life membership, $200. Bn hues 00-178 nS oc sil ck mmr of page lo 1.5 65-80 ee ot ee $8.00 each; BE eet tern om enters of - (ER os : Fen N JOURNAL, Dept. of Botany, University ee 3 American Fern Journal 76(2):33-47 (1986) Three New Species of Moonworts (Botrychium subg. Botrychium) Endemic in Western North America W. H. WAGNER, JR. and F. S. WAGNER Department of Biology, The University of Michigan, Ann Arbor, MI 48109 SE ee a a eo Ree Probably no group of modern ferns or fern-like plants is as poorly understood as the Ophioglossales. Interpretations of the “fertile segment” (sporophore) and the unusual stele (eustele?) have troubled botanists for many years. Even the relationships of the order as a whole are in doubt. At the genus and species level, there has been little agreement, and the taxonomy is still up in the air. E. P. St. John (1949) grappled with some of these problems and gave a theory to explain them, based upon the subterranean gametophyte, which he believed inhibits hybridization and outcrossing. He stated that “hybridization, or even cross-breeding within the species, is almost impossible” and that as a result, the species tend to be very widely distributed. “One of the outstanding peculiarities of this group of plants,” he wrote, “is that while they are of world-wide distri- bution the number of species is very small.” Local endemics may be geographic variants of other species. According to St. John, the species of the Ophioglossa- ceae are closely related, and the course of evolution is usually indicated by transitional and juvenile forms. Through selfing of the subterranean gameto- phytes a trivial variation may form a colony, and such divergent forms might persist so long as they meet minimum requirements for survival. Now, over a third of a century after St. John’s interesting proposals, we ques- tion a number of his points on the basis of many new data. We find no evidence that subterranean gametophytes inhibit hybridization and outcrossing. Indeed, __ 17 hybrids are known in Lycopodium, 1 in Psilotum, and 12 in Botrychium : (Wagner et al., 1985) plus two additional Botrychium hybrids described below— ___ alltaxa with subterranean gametophytes. The idea that the species of Ophioglos- sales are generally widely distributed is not supported by the data; only a few have truly broad distributions, such as Botrychium lunaria, B. virginjanum, and Ophioglossum nudicaule. Most species have narrow ranges, e.g, B. subbifoliatum, endemic to Hawaii, B. lunarioides, southeastern United States, B. oneidense, northeastern North America, B. echo, southwestern United States. The true ranges Were obscured by the crude taxonomies of a third. gfe tentury ago. Botrychium Oreale became much delimited geographically when it was found that the Pa- cific Northwest plant so-named belongs to another species, B. pinnatum. We used to believe that B. ternatum occurred both in Asia and in North America, but we now know that the American plant is a distinct species, B. rugulosum, a narrow endemic, known only in the Great Lakes region from Montreal to Min- _ ‘Resota (Wagner & Wagner, 1982). : : It is not true that members of the Ophioglossales lack diversity. St. John men- | tions Helminthostachys and Cheiroglossa as distinctive genera, but what about Such distinctive elements within genera as Ophioglossum bergianum or O. pen- een Se eae sea a i om aa ai SES 1 ee NO A oe Se fee et Se ee Tae. eee airy Rages = EL Spake Mee te ane | BOTAN APT FR 34 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) dulum, or Botrychium lunarioides, B. lanuginosum, and B. paradoxum? Even major variations, like the dissected form of B. dissectum, may form a welter of intermediates. The dissected and undissected forms may run together in the same habitat; there is no evidence that inbred colonies are formed. Research in Japan by Norio Sahashi of Toho University has uncovered remarkable diversity in evergreen grapeferns, Botrychium subg. Sceptridium, and research by us in western North America has revealed the same in the moonworts, Botrychium subg. Botrychium. In contrast to the six well known species of western Europe and eastern North America, the species of western North America have been poorly collected, and even now some are reported from only several localities. Because of new ample population samples, we now know that western North America is the metropolis for moonworts. Including the three to be described here, there are 13 known moonwort species now in the west, as well as a number of sterile hybrids, of which the remarkable and apparently self-reproducing B. x watertonense is the best known (Wagner et al., 1984). Still other species probably exist in the grasslands and mountains of western North America, and we urge field botanists to make a special effort to locate them. The three species and two hybrids described below are all very rare. For all three species we have been fortunate in finding excellent localities, so that we can base our descriptions on hundreds of plants. One, the Prairie Moonwort, has been collected, until recently, mainly in Canada. The Stalked Moonwort has also been taken mainly in Canada (Saskatchewan to British Columbia); only one United States locality has as yet been discovered. The Upward-lobed Moonwort has, so far, a distinctive distribution, from Nevada to the Yukon and northern Ontario. Future field explorations will unquestionably reveal additional locali- ties and populations. We urge collectors to search for “Botrychium Havens” such as the Wallowa Mountains of Oregon, where no fewer than 10 species and two hybrids, including those described here, are now known (cf. Mason, 1975). A key is given below to all of the moonwort species known in western North America, an area defined as west of a line from Winnipeg, Manitoba to Dallas, Texas. There is some evidence of “leakage” of certain western disjuncts east of this line, especially along Hudson Bay and western Great Lakes (Wagner & Wagner, unpublished). (The western taxon, B. pumicola, has not been included in the key because its status is at present in question; according to research in progress it appears to us to be a form of the extremely variable B. simplex.) The key is based upon large, well developed plants of each of the species. To deter- mine young or depauperate plants, we must use associations with large plants. luster and color, spore size (all members of this group produce spores, even when they first appear above ground), and the cutting, which is usually a sim- plified form of the adult cutting. Key To BOTRYCHIUM suUBG. BOTRYCHIUM IN NorTH AMERI 1. Trophophore (sterile segment) absent, replaced by a second sporophore (fertile segment); sporophore branches short, mostly less than 5 mm (ex- cept the lowest branches in very large plants) B. paradoxum oe eee een eer eee ee WAGNER & WAGNER: BOTRYCHIUM 35 1. Trophophore present with sporangia scattered along margins, or (usually) ee absent altogether; sporophore branches longer, mostly more than 5 be bo B techoro narrow, completely ringed along the margins with spo- rangia; spores irregular in size and shape ........... B. X watertonense . Trophophore with a few sporangia at base (very rarely extra sporo- phores at base) or (usually) no sporangia; spores regular in size and shape. 3. Lower pinnae linear, lanceolate, or ovate, with a central or at least basal midrib. 4. Trophophore broadly deltate, usually subsessile; pinnae linear; sporophore divided near base into two or more major axes ... Be Pelee oe pn oes eo el enc eae ee B. lanceolatum 4. Trophophore mostly oblong to ovate, subsessile to long-stalked; segments linear to oblong to spatulate; sporophores with only one major axis or, if more, the laterals arising well above the base. 5. Trophophore stalk long, 30 (10-50) percent of trophophore length; blade outline mostly triangular-ovate; lower pinnae stubby, rhomboidal and asymmetrical; common stalk (old plants) commonly with a single narrow brown stripe below the trophophore: 2... 0 a: B. pedunculosum 5. Trophophore stalk shorter, 5-25 percent of trophophore length; blade outline mainly oblong, or triangular by exaggeration of basal pinna pair; lower pinnae more elongate, oblong to lin- ear; common stalk lacking a single narrow brown stripe be- low the trophophore. . Pinnae usually well separated, linear to oblanceolate with pointed tips; basal pinnae mostly cleft into a smaller lower segment and larger upper segment; lamina shiny green in BO ee B. echo . Pinnae usually approximate, oblong to ovate with aaa or blunt tips; basal pinnae not cleft into two segments; lamina luster various. 7. Pinnae with few lobes, these mainly on the basal side; lowest pinnae mostly exaggerated, ascending and com- monly subclasping; pinnae broadly adnate, strongly asymmetrical; lamina dull gray-green . life gece cise ees B. hesperium 7. Pinnae with numerous lobes, these roughly equal in number on upper and basal sides; lowest pinnae con- form, mostly equal in length to those above; pinnae nar- rowly adnate, nearly symmetrical; lamina ae bright green in life 4... oe ee a 2p aeg 3. Lower pinnae or pinnules lunulate, fan-shaped, wedge-sha Square, with flabellate venation, a central midrib lacking. 8. Apex of trophophore usually not deeply divided, commonly con- jo?) fo?) 36 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) cave (in life); upper pinnae or lobes tending to be somewhat irregularly fused; sporophore arising at various positions, from ground level to near the apex; trophophore sessile to long-stalked. Leaves herbaceous to succulent; trophophore strongly lami- nate, lobed to 1-2-pinnate, oblong (in young and smaller plants) to deltate (in older and larger plants); segments round- ed, asymmetrical, fan-shaped to spatulate, the outer margins ik ewe taco ini nese B. simplex . Leaves succulent; trophophore more or less skeletonized, usu- ally merely shallowly lobed, linear to oblong, the segments angular, commonly square or oblong, the outer margins usu- ally irregularly and coarsely toothed ............ B. montanum 8. Apex of trophophore usually deeply divided, mostly flat (in life); upper lateral pinnae and lobes regularly separated; sporophore attachment high on the common stalk; trophophore sessile to short-stalked. 10. Pinnae lunate to broadly cuneate, the sides of the lower pinnae at a 90°-180° angle, remote to overlapping. 11. Pinna pairs (3)5-6(8), approximate to overlapping (except in shaded sites), mostly entire or nearly so; texture leath- ery, color (in life) shiny dark gray-green ........ B. lunaria 11. Pinna pairs (2)3-4(5), usually remote, mostly crenulate to dentate; texture thinly herbaceous, color (in life) shiny Hee Veleneteen 8 ee B. crenulatum 10. Pinnae cuneate to oblong, the sides of the pinnae at a 0°-90° angle; usually remote. 12. Sporophore usually only slightly longer than the tropho- phore; segments linear to oblong to spatulate, the largest strongly asymmetrical and bifid; spore diameter 34-39 ein Ur Flavors sontion} ©... ss B. campestre Sporophore usually 1.3-2.0 times longer than the tropho- phore; segments cuneate or narrowly cuneate, the larg- est only moderately asymmetrical, usually not bifid; spore diameter 40-50 um. 13. Outer pinna margins entire or undulate, not deeply dentate; pinnae usually only slightly ascending; lam- ina thick, not veiny, dull gray-green ... B. minganense . Outer pinna margins deeply and irregularly dentate or serrate; pinnae strongly ascending; lamina thin, veiny, shiny yellow-green .........----- B. ascendens od © oa tS ood oo ascendens W. H. Wagner, sp. nov. (Figs. 1, 2).—Type: U.S.A., Ore- gon, Wallowa Co., § of Enterprise, ca. 1.5 mi from beginning of Hur- ricane Creek Trail, Wagner 83363 (MICH). Segmenta cuneata, valde ascendentia; textum tenuis, venis prominentibus: margines valde dentati et plus minusve lobati: chromosomatum numerus 2 ~ WAGNER & WAGNER: BOTRYCHIUM F — see G1. Botrychium ascendens leaves, Wagner 83363. Note pinna outlines, orientation, and margins. Plants exclusive of roots reaching to 20 cm tall, averaging 10-12 cm tall, the common stalk 5 cm (2—12) tall, sporophore one-third to one-half again as long 8 the trophophore; trophophore bright yellow-green and shiny in life, the veins °nspicuous and the lamina surface somewhat crinkled; trophophore stalk 1-30 Percent of total trophophore length; blade outline narrowly oblong-triangular, "P to 6 cm long and 2 cm wide; segments up to six pairs, strongly ascending, me 38 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) Fic. 2. Botrychium trophoph tli d venation patterns. A. B. ascendens, Wagner 81127. C. B. campestre, Wagner 85016. P. B. pedunculosum, Wagner 81124. upper ones extremely ascending; lower pinnae with somewhat curved posterior margins and straight anterior margins, commonly with marginal sporangia; larg- est pinnae 12 mm long, 10 mm wide; veins dichotomous, 2 or 3 major veins entering the pinna base, 14-17 vein endings along outer margins of larger pinnae; margins deeply and sharply dentate and serrate, often two- to five-lobed; spo- rophore branches mostly simple except for lower ones with 1-3 branchlets; spores finely verrucose, (40)44-47(54) wm in diameter; chromosomes n = 90. Paratypes: Canapa. Alberta: Waterton Lakes National Park, Snowshoe Trail, ca. 1 mi NW of Red Rock Canyon Parking Area, Aug 1978, R. Campbell s.n. (MICH). British Columbia: Festuca Pass and lower SW slopes of Pipe Organ Peak, Taylor et al. 5984, 5985 (UBC—mixed with B. lunaria). Ontario: Vicinity of mouth of Severn River, Moir 1444 (CAN). Yukon Territory: Dawson, Porsild 2040 (CAN— mixed with B. lunaria). U.S.A. Montana: Lewis and Clark Co., Sun River Primitive Area, E slope of China Wall near N end, 27 Jul 1948, Hitchcock (WTU) Ja: Clark Co., Charleston Mts. [now Spring Mts.], Rainbow Falls, Clokey & Clokey 7462 (UC—mixed with B. lunaria), Clokey 8215 fall mixed with B. lunaria, DS, NY, UC, WTU), Wagner 81150, 82115 (MICH); Charleston Peak, Clokey 7463 (all mixed with B. lunaria, DS, NY, UC, WTU), Clokey & Bean 7464 (UC—mixed with B. ia). The epithet refers to the strongly ascending pinnae. This species is most similar to B. crenulatum Wagner (Wagner & Wagner, 1981) from which it differs in its narrowly cuneate segments (vs. broadly flabellate segments), sharply serrate margins (vs. shallow to strongly crenulate or dentate), a tendency for supernu- merary sporangia on the lower pinnae (vs. no sporangia on the lower pinnae), more dilated and ascending terminal segments (vs. more skeletonized vas spreading), relatively taller and more robust sporophore with more upright branches (vs. usually shorter sporophore with thinner and more spreading branches}, and chromosome complement of n = 90 (vs. n = 45). A living pes is illustrated in Lellinger, 1985, figure 109. From B. lunaria and B. minganense, B. ascendens differs in a number of characters (cf. Wagner & Wagner, ttt figure 2: A = ascendens, C = crenulatum, L = lunaria, and M = minganense). It is interesting that there is no spore size increase with polyploidy; the tetraploid WAGNER & WAGNER: BOTRYCHIUM 39 N | Fic. 3. Presumed hybrid botrychiums with abortive spores. A. B. pedunculosum x pinnatum. OR: Wallowa Mts., Lostine River Valley, Wagner 83361a. B. B. ascendens x crenulatum, Wagner 83363a. B. ascendens has spores only slightly different in size from the diploid B. cren- ulatum (see below, under B. campestre, the more striking contrast in spore size between that species and B. minganense). Botrychium ascendens has a widely scattered range in western North Amer- ica, but it is local and rare. In the type locality, in the Wallowa Mountains of Oregon, however, it occurs in hundreds along Hurricane Creek Trail in grassy fields interspersed with spruces. Growing with it are B. crenulatum, B. lunaria, and B. minganense. Only a single hybrid (Wagner 83363a) was found. The other parent was de- duced to be B. crenulatum. The pinnae of the hybrid average wider than those of B. ascendens, approaching the width of B. crenulatum (Fig. 3). The lowest pinnae are fan-shaped like those of B. crenulatum, but the remainder are cu- neate as in B. ascendens. As expected, the spores of the hybrid turned out to be abortive, According to our present knowledge of the chromosomes of these plants, the hybrid should prove to be a sterile triploid. Botrychium campestre W. H. Wagner & D. R. Farrar, sp. nov. (Figs. 2, 4, 5).— Type: U.S.A., Iowa, Plymouth Co., Five Ridge Prairie and Banks Tract, Wagner 85016 (MICH). Caulis subterraneus corporibus minutis fere sphaericis botryoideis munitus; Segmenta parva linearia oblonga vel spatulata plerumque remota, segmentis eeamis valde asymmetricis bifidisque; lamina oblongo-linearis, plerumque la- tissima ultra medium; sporophorum pro parte maxima trophophorum aequans vel parum superans: sporae parvae 34-38 wm diametro; chromosomatum nu- Merus n = 45, Plants exclusive of their roots 6(12) cm tall; common stalk usually 5(10) cm AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986} Fic. 4. the grasses. Botrychium campestre. Living plant in its native habitat, difficult to see among Minnesota, Pipestone Co. WAGNER & WAGNER: BOTRYCHIUM 41 Fic.5. Botrychium campestre leaves. Note short sporophores, narrow, remote, deeply incised pin- hae, and broad midrib. Upper row, Farrar s.n.; lower row, Eilers s.n. long; underground stems with tiny, nearly spherical bodies 0.4-0.8 mm in di- ameter, in grapelike clusters; sporophores stubby, usually equal to or slightly longer than trophophore but sometimes one-half or more longer; trophophore dull, whitish green, fleshy, the veins submersed; trophophore sessile (usually) to short-stalked, the stalk up to 10 percent of the total trophophore length; the stalk and midrib fleshy and 2-4 mm broad; blade outline oblong to linear oblong, commonly widest above the middle, up to 35 x 12 mm; segments linear to oblong 'o spatulate, the largest strongly asymmetrical and typically bifid, the lower lobe one-third to two-thirds the length of the upper lobe; number of pinna pairs few, Usually less than six, these mostly remote; segments ascending mostly 30°-50°; st pinnae reaching 7 mm long and 4 mm wide, served by two major veins 42 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) terminating on outer margin with 8-12 veinlets; the lowest pinnae generally narrower and less complex than the 1-3 distal pairs; outer margins mostly shal- lowly crenate or dentate; sporophore with all axes fleshy and flattened; spores finely verrucate, small, (30)34-38(44) um, chromosomes n = 45. Paratypes: CANADA. Alberta: Ft. Saskatchewan, near eastern edge, Turner 7546 (ALTA); Ft. Sas- katchewan, Turner 8199 (ALTA—mixed with B. minganense). Saskatchewan: Beechy, Ledingham & Hudson 992 (DAO); Dinsmore, Ledingham & Hudson 1448 (USAS); 15 mi ENE of Fox Valley, Ledingham & Jones 5676 (USAS). U.S.A. Iowa: Monona Co., Loess Hills Wildlife Area, 15 May, and 2 June 1982, Eilers s.n. (ISC, MICH); Wagner 83207, 83208 (MICH). Plymouth Co., E of Banks Tract, Farrar s.n. (MICH). Woodbury Co., Sioux City, Briar Cliff College, Loess Hill Prairie, Wagner 83205, 83206, 85014 (MICH); Stone State Park, Wagner 85015 (MICH). Minnesota: Lac Qui Parle Co., An- telope Hills, Wagner 85018 (MICH); Marietta, Lyon 889 (MIN). Lincoln Co., Hole in the Mountain Prairie Preserve, Wagner 85017 (MICH). Pipestone Co., Aetna Preserve, 2 mi NE of Holland, Wagner 85019 (MICH). Nebraska: Brown Co., Niobrara Valley Preserve, Freeman 1517 (KANU). North Dakota: McHenry Co., McHenry Sandy Prairie, Stevens 1530 (US, UDA). McLean Co., 15 mi SSW of Velva, NE slope strip mine piles, 18 June 1974, Disrud & Weber s.n. (MICH). The epithet refers to the occurrence of this species in grasslands and plains. Botanists, including ourselves, were much confused as to the identity of this moonwort, as shown by their annotations on herbarium sheets, viz., B. boreale var. obtusifolium, B. dusenii, B. lunaria, B. lunaria var. minganense, B. matri- cariifolium, B. matricariifolium var. hesperium, and B. simplex. Botrychium campestre is actually very distinctive once its characters are recognized. It re- sembles most B. minganense, with which it grows in at least one locality, and from which it can be separated by its much smaller, narrower, more dissected pinnae, fewer pinna pairs (in comparable-sized leaves), more fleshy and robust axes, diploid chromosome number, and spores 20-25 percent smaller than those of B. minganense. The peculiar round bodies attached to the stem are currently under investigation by D. R. Farrar and ourselves. They are evidently gemmae capable of propagating the plant; although root proliferations are known, spe- cialized gemmae have never before been reported in Ophioglossales. Except for one questionable report, no vegetative reproduction has heretofore been known in Botrychium. Another feature, the early seasonal disappearance of the leaves, is different from all other known species of North American moonworts; in Iowa the leaves appear in April and turn brown and mostly die by mid-June. Botrychium campestre was first discovered in lowa by Theodore Van Brugge? and Lawrence Eilers in 1982. They called the plant to the attention of Donald R. Farrar, who, recognizing that it was unusual, referred it to us. Borrowed Botrychium specimens from the plains areas of the United States and Canada revealed that the prairie botrychium possesses a wide range, shown in Fig. 6. All localities so far are in prairies or prairie-like habitats, either in open grass OF Beno shrubs. Data on labels read “moist sandy saline meadows,” “dry west- facing loess prairie bluff,” “sandy prairies,” “gravelly shores of alkaline slough.” The northwesternmost locality is a “north-facing, nearly bare, sandy side © south ditch of Canadian National Railroad, near the eastern edge of Fort Sas tchewan about 15 mi NE of Edmonton.” Since many prairie plants occur along railroad tracks, collectors are urged to search for new localities in such places: The Prairie Moonwort is exceedingly rare and inconspicuous and takes m persistence and patience to find. WAGNER & WAGNER: BOTRYCHIUM 43 Fic. 6. Botrychium campestre. Range indicated by dots of known collections. Botrychium pedunculosum W. H. Wagner, sp. nov. (Figs. 2, 7).—Type: U.S.A., Oregon, Wallowa Co., Lostine River Valley, 1.6 mi S of ranger cabin, 13.3 mi S of Lostine Church, grassy second-growth area along road, Wagner 83361 (MICH). Stipes communis maturus plerumque vitta angusta pallido-brunnea e tropho- Phoro decurrenti instructa; stipes trophophori longus, 10-50% totae longitudinis ttophophori; lamina trophophori triangularis vel ovato-triangularis; pinnae late uxae, plerumque approximatae, breves et ‘angulares, grosse dentatae et plus minusve acutae; chromosomatum numerus n = 90. Plants exclusive of roots (5)11-13(25) cm tall; common stalk usually with nar- tow pale brown stripe running down from the trophophore, (4)6-7(15) cm tall; Sporophore (1.5)2(2.5) times as long as the trophophore; trophophore (in life) gray- green, dull, leathery, with veins submersed; trophophore stalk (10)30(50) percent of total trophophore length; blade {in life) commonly shallowly rolled, spoon- like, triangular to ovate-triangular, widest at the base, av. 4cm long, 2.5 cm wide; pinnae somewhat ascending commonly approximate, (2)3(6) pairs, broadly at- tached, the lower ones asymmetrical, rhomboidal, angular, the tips usually more oF less Pointed, with 1-6 entire to coarsely crenate lobes, the upper ones becom- 'Ng narrowly flabellate: largest pinnae up to 15 mm long and 10 mm wide; AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) Fic.7. Botrychium pedunculosum, Wagner 81124. Note long trophophore stalks, and angular, deep- ly incised lower pinnae. sporophore stalk (20}50(70) percent of sporophore length, sporophore branches in large plants with two major ascending lateral branches up to 70-100 percent as long as whole sporangial cluster; spores finely verrucate (36)37-41(43} »™ diameter; chromosomes n = 90 . Paratypes: CANapa. British Columbia: E of Quesnel on road to Barkerville, Calder et al, 14152 WAGNER & WAGNER: BOTRYCHIUM 45 (DAO—mixed with B. pinnatum); Rte. 26, near Quesnel, Wagner 83339 (MICH); Rte. 26, 6.2 mi E of Victoria Creek, Wagner 83347 (MICH). Saskatchewan: Cypress Hills, Hudson 1926 (DAO); Cypress Hills (Central Block), Wagner 83309, 83111 (MICH). U.S.A. Oregon: Wallowa Co., Lostine River Valley, 1.2 mi N of Bowman Trail, Wagner 81130 (MICH); type locality, Wagner 81124 (MICH). The epithet refers to the long stalk on the trophophore. The few collections of this species known outside of the Wallowa Mountains were identified as Bo- trychium boreale var. obtusifolium and B. matricariifolium. Mature plants of B. pedunculosum can be told readily from these taxa, as well as from B. hesperium and B. echo (cf. Wagner & Wagner, 1983a) by the usual presence on older plants of a well-defined, narrow, brown stripe running down the common stalk from the trophophore, by the unusually long trophophore stalk, and by the triangular- ovate blade with angular stubby asymmetrical segments. At the type locality, there are at least 200 large, healthy plants, growing together with B. multifidum (frequent), B. lanceolatum (abundant), B. lunaria (rare), B. pinnatum (frequent), B. minganense (frequent), and B. simplex (rare). A living plant is illustrated by Lellinger (1985, fig. 114). Where we have seen it in British Columbia, it occurs with the first four species in the foregoing list. Although it is widespread in western Canada, it is evidently very local and rare. The hybrid, Botrychium pedunculosum = pinnatum (Fig. 3) was found among plants of B. pedunculosum, growing close to B. pinnatum. It merges the char- acters of the parents nicely: The plant is not so robust and broad as B. pinnatum nor so slender as B. pedunculosum. The trophophore of the hybrid shows the segmentation and dissection of B. pinnatum together with the more pointed pinnae and long stalk of B. pedunculosum. The sporophore is relatively short, as in B. pinnatum, but is divided into three major branches, as in B. peduncu- losum. The single hybrid plant was readily recognized by its morphology and was found to have pronounced spore abortion as is usual in sterile interspecific crosses in this genus. DISCUSSION With the description of the three taxa given above, we are able now to gain a better view of the endemism in the moonworts (subgenus Botrychium). It should be remembered that the common species of Europe and/or eastern North Amer- ica, and their variations, are well known by hundreds of collections in the major herbaria of the world. They include B. lunaria, B. lanceolatum, B. boreale, B. simplex, B. matricariifolium, and B. minganense. Previously poorly known or unknown, the western North American taxa are well represented by ample collections for the first time. We thus have a substantial basis for taxonomic interpretation that we lacked before. And, more often than — oS of these taxa co-exist (Wagner & Wagner, 1983b) confirming their taxonomic dis- tinctions. Accordingly, we can estimate the level of endemism. Table 1 summarizes what we know about endemism in the moonworts. In Spite of St. John’s implication that the species tend to have world-wide distri- -Dutions, there is only one, B. lunaria, about which that can be said. It is — not only throughout the boreal region but turns up again in Australia and New 46 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) TABLE 1. Endemism in Botrychium subg. Botrychium (based on Wagner & Wagner, 1981, 1983a; Wagner et al., 1984; and present paper). Circumboreal and Amphitropical Western North America B. psa (L.) Swartz B. ascendens Wagner Circumbor B. campestre Wagner & Farrar B. lanceolatum fa ) Angstr. B. crenulatum Wagner B. simplex H B. echo Wa B. Northern and pee ek hesperium (Maxon & Clausen) i Wagner & Lellinger di Le Amphiatlantic (Europe, e. N.A.} B. pedunculosum Wagner B. matricariifolium A. Br. B. pinnatum St. Jo Northern North America B. montanum Wagner inganense Victorin B. paradoxum Wagner North Central North America B. x watertonense Wagner’ o Wagner ‘Included because of its apparent ability to reproduce. Zealand. Also the plant known as B. dusenii (Christ) Alston from southern South America may be the same as B. lunaria or a variety of it. As shown in the table, most of the species are limited to less than half of a continent. Some of them, like B. campestre, have much narrower ranges; and at least one, B. mormo, has a very small range. It is interesting to note that more than half of the world’s species of moonworts are known only from western North America. ACKNOWLEDGMENTS This study was funded by National Science Foundation Grant BSR-82-2768, Monographic Studies of the genus Botrychium. The Iowa State Avisory Board of Reserves aided us in our field work in the vicinity of Sioux City in 1983. We are grateful to a = = Speen who gave us assistance, especially Edward Alverson, W. R. Anderson, George Argus, David Bay, — a William Blan- kenship, J. H. Hudson, and Donald R. Farrar. D. B. em! loa D. H. Wagner made helpful ts on the manuscript. The authorities “ the ala Hills Provincial Park, epee LITERATURE CITED os D. B. 1985. A field manual of the ferns and fern-allies of the United States and Canada. ashington, D.C.: Smithsonian Inst. Press. Mason, y ahs Guide to the plants of the Wallowa Mountains of northeastern Oregon. Eugene: iv. Oregon Mus. Nat. Hist. Spec. Publ. St. JOHN, E P. 1949. The evolution of the Ophioglossaceae of the eastern United States. Quart. J- Florida Acad. Sci. 12:207-219. WAGNER, Ww. we Jr., and F. S. WAGNER. 1981. New species of moonworts, Botrychium subg. Botrych- ‘um (Ophioglossaceae) from North America. Amer. Fern J. 71:20-30. and —_—_——. i992 1 a newly recognized species of ergreen grapefern in the Great Lakes area of North America. ti Univ. Michiga® Herb. 15:315-324. WAGNER & WAGNER: BOTRYCHIUM a7 and —————. 1983a. Two moonworts of the Rocky epee Botrychium hesperium and a new species poy confused — it. sp roe }. 7: and ————. 1983b. y tool in the study of New World Botrychium (Op edamasnal Taxon 32:51-63. 2 ER ara dj. M. Berret. 1985. Evidence for ific hybridi pt id phyt with subterranean — igo aise Soc. peer 86B: 273-285. —_, (. SIAVUPLER, AND: J, ERSON. 1984. A new nothospecies of moonwort Sit ba lnesacane: Hotyhteat aa J. Bot. 62:629-634. CORRECTIONS TO INDEX FILICUM Index filicum is now produced by the Royal Botanic Gardens, Kew, and as the new editor I am keen to correct the errors in the original and subsequent supplements. I am aware of a number of these but am sure that others are as yet undetected. I very much welcome information on any omissions and mistakes known to my colleagues. Infraspecific names are to be included in the next supplement, which will run from 1976-1985. From 1986, new names of pteri- dophytes will be published annually, as an appendix of an annual Index Kew- €nsis, but Index Filicum will continue to be published at five- or ten-year inter- vals.—B. S. Parris, The Herbarium, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, England. American Fern Journal 76(2):48-55 (1986) Gametophytes of Lycopodium lucidulum from Axenic Culture DEAN P. WHITTIER Department of General Biology, Vanderbilt University, Nashville, TN 37235 TERRY R. WEBSTER Department of Ecology and Evolutionary Biology, The University of Connecticut, Storrs, CT 06268 Lycopodium gametophytes are known from about 7% of the species (Bruce & Beitel, 1979). It is often difficult to find the subterranean gametophytes of Ly- copodium in nature. One way of increasing the number and kinds of subterra- nean gametophytes for study is to grow them from spores in axenic culture. Two types of subterranean Lycopodium gametophytes, carrot- and disc-shaped, have been grown in axenic culture on nutrient media containing minerals and sugar. The structure of these gametophytes was essentially the same as that for game- tophytes from nature (Whittier, 1977, 1981). Lycopodium lucidulum represents a third type of subterranean gametophyte. Its gametophytes are axial, white, and fleshy (Spessard, 1922) and do not resem- ble the disc- or carrot-shaped gametophytes of L. obscurum and L. digitatum respectively which have been previously grown in culture (Whittier, 1977, 1981). The main goal of this study was to grow this third type of subterranean Lyco- podium gametophyte in axenic culture, and to compare its morphology to ga- metophytes of the same species from nature. MATERIALS AND METHODS Plants of Lycopodium lucidulum Michx. bearing sporangia were collected from various places in the northeastern United States and Tennessee in early fall. Vouchers are on deposit in the Vanderbilt University Herbarium (VDB). Spores were obtained by allowing fertile branches to dry over sheets of paper- The spores were surface sterilized with 20% Clorox by the method of Whittier (1964) and were sown on 15 ml of nutrient medium in culture tubes with screw caps that were then tightened. The nutrient medium was composed of Knudson's solution of mineral salts, FeEDTA, 0.5% sucrose, and 0.6% agar. The spores were cultured at 24 + 1°C in darkness or in light, 1250 lux, from Gro-lux flu- orescent lamps for 12 of every 24 hours. The gametophytes studied were first photographed and then fixed. Fixation from culture was accomplished with a 1:1 mixture of 4% glutaraldehyde and 10% acrolein which were both in 0.07 M phosphate buffer at pH 6.8. Gameto- phytes from nature were fixed in FAA. After fixation, the gametophytes and pieces of gametophytes were embedded in paraffin and sectioned by conven tional techniques (Johansen, 1940}. The sections were stained with Heidenhain’s hematoxylin, safranin, and fast green. Hand sections of unfixed gametophytes WHITTIER & WEBSTER: LYCOPODIUM GAMETOPHYTES 49 were stained with alcian blue (1% in 3% glacial acetic acid) to demonstrate mucilage (Pearse, 1968). RESULTS Efforts to observe the early stages of spore germination after the spores had been in culture for 6-12 months were unsuccessful. Cultures that had been in the dark for 2 or more years were the source of gametophytes for this investi- gation. In these older cultures, one to a few gametophytes were usually found in most tubes. No gametophytes were found in cultures maintained in the light. The gametophytes of L. lucidulum from axenic culture are axial structures (Figs. 1, 3). However, they are slightly flattened with distinctive dorsal and ven- tral surfaces (Figs. 2, 4, 9). The dorsal and ventral surfaces are separated by indentations along the lateral surfaces of the gametophyte (Figs. 3, 9). The dorsal surface is covered with short paraphyses which obscure the sex organs (Figs. 2, 3, 4, 5, 9). Although the paraphyses may be completely uniseriate, often the basal portions are biseriate where they attach to the gametophyte (Fig. 5). If the pa- raphyses have a biseriate base, that portion branches to form two uniseriate hairs (Fig. 5). The ventral surface has rhizoids which are coated with a mucilag- inous material. The mucilage stains with alcian blue (Fig. 6), indicating that the rhizoid mucilage is composed of an acid mucopolysaccharide. The youngest (smallest) gametophytes from culture have the dorsiventral growth habit (Figs. 8, 9). Even with these small gametophytes, the indentations on the lateral surfaces extend for almost their whole length. Thus, it appears that the mature growth habit is established at a small size and early age in the gameto- phytes of L. lucidulum. . The meristem is located in a groove on the lower surface of the apical region (Figs. 4, 8, 9, 10). The derivatives to the upper side of the meristem form the dorsal tissues including sex organs and paraphyses (Figs. 4, 10). The immature dorsal tissues overarch the meristematic region and cause the meristematic groove to be on the lower surface of the apical region (Figs. 4, 10). The meristematic derivatives to the lower side form the ventral portion of the gametophyte eat cluding unicellular rhizoids. The meristematic groove is continuous with the indentations on the lateral surfaces of the gametophyte (Fig. 9). The apical region of the gametophyte is usually about the same width as the nature portion of the gametophyte, especially the mature region immediately basipetal to the apex. Thus, it appears that the meristem has limited activity on its lateral margins in the apical region. Gametophytes (Fig. 4) with expanding apical regions appear to be ones which are preparing to branch. A few branched Sametophytes were found in the cultures. ae € gametangia are initiated in the apical region from recent derivatives of the meristem (Fig. 10). In mature gametophytes, a gradation of gametangial de- Velopmental stages from initiation to maturity exists in the dorsal tissue of the apical region. Stages in the development of antheridia are illustrated in Figure 10. The antheridia are mature once their position has shifted to the top of the dorsal lip. The mature antheridia are large and sunken and have only a few AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) 50 f n surface 0 Fics. 1-9. Gametophytes of L. lucidulum from axenic culture. Fic. 1. Gametophyte © nutrient medium, x1.5. Fic. 2. Dorsa : : : x 600. Fic. yses, 120. Fic. 6. Rhizoid tips with mucilaginous sheaths stained with alcian blue, WHITTIER & WEBSTER: LYCOPODIUM GAMETOPHYTES 51 jacket cells at the gametophyte surface (Fig. 10). It appears that during fixation sperm are occasionally released and masses of sperm can accumulate on the surface of cultured gametophytes (Fig. 10). The position and development of the archegonia have the same relationship to the apex as the antheridia. The archegonial necks are not obvious on the whole gametophytes because of the dorsal paraphyses. The necks are of a me- dium length for Lycopodium (Figs. 5, 11). Most of the archegonia have 4 neck canal cells and 4 tiers of exposed neck cells (Fig. 11). Rarely are the antheridia and archegonia interspersed along the gametophytes from culture. Young gametophytes have only antheridia, and older gametophytes eventually form archegonia. Transition regions do occur between areas with all archegonia or all antheridia in which both gametangia are present. However, even in these regions there is little intermixing of the gametangia. Usually in these transition regions one type of gametangium has a median distribution and the other type has a more lateral position. It appears that the change in type of gametangium formation occurs in the middle of the apical meristem first and spreads laterally. The internal tissues of the gametophytes from culture do not have the mycor- thizal fungus (Figs. 7, 10) which is found in the ventral tissues of gametophytes from nature (Fig. 12). The internal tissues are composed of numerous layers of isodiametric cells (Figs. 7, 13) and no palisade or vertically elongated cells occur. Although the internal tissues of most of the cultured gametophytes are rather uniform (Fig. 7), some gametophytes contain cells with distinctive contents (Fig. 13). In these cases the central cells contain large amounts of a fine grain, blue- staining, condensed tannin (Fig. 13). In addition, these cells have large amounts of starch, although it should be noted that the other internal cells do contain Some starch. The surface cells in the gametangial region and 5 or 6 layers of the most ventral cells lack the condensed tannins (Fig. 13). DISCUSSION AND CONCLUSIONS The Percentage of spore germination was very low, less than 0.1%, and the early stages of germination were not observed in these cultures. There may have been so few germinating spores that they were missed when the cultures were sampled at 6 and 12 months. However, it is possible that the spores of L. lucidu- take longer than a year to germinate. Bruchmann (1910) has reported that Spores of some species of Lycopodium take 3 or more years to germinate. A delayed germination would help to explain why the gametophytes of L. lucidu- lum were only found in 2-3 year old cultures. oo Ee ‘ Cross Section of gametophyte with antheridia and paraphyses above the lateral ee and = 20ld-bearing surface below, x50. Fic. 8. Lateral view of young gametophyte with overarching Paraphyses-bearing dorsal surface, *5. AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) 52 2 Sat * one Se! ene PE Pian Os te oe, gs eS eS ae 7 ey fot - Ye PY et reo Fics. 10-13. Structural details of gametophytes of L. lucidulum. Fic. 10. Sagittal i apical region of gametophyte with meristem (arrow) and numerous developmental stages sapere idia in the dorsal tissues. Sperm on dorsal surface (arrowheads), 65. Fic. 11. Section seen urf ith arc} ium havi 1 four neck canal cells, x 125. Fic. 12. Section of gametop The germination of spores of Lycopodium species with subterranean ane’ tophytes has been poor in axenic culture. For the three species, L. a digitatum, and L. obscurum, that have been germinated in culture (Whittier, - 1981}, germination has not exceeded 0.1%. Experimental studies are ~~ on spore germination in other species of Lycopodium. Establishing pees for accelerating germination and/or increasing the percentage of germinatio” would greatly enhance studies on the subterranean gametophytes of Lycop? dium in axenic culture, ed with The gametophytes of L. lucidulum from axenic culture were compar gee the descriptions of gametophytes from nature by Spessard (1922) and Bruce WHITTIER & WEBSTER: LYCOPODIUM GAMETOPHYTES 53 Beitel (1979). In addition these cultured gametophytes were compared with field- collected gametophytes from a wooded area in Storrs, Connecticut, one of the sites from which spores were obtained. These comparisons show that the basic organization of the cultured gametophytes is the same as that for the gameto- phytes from nature. A major difference between gametophytes from the two sources is the absence of a mycorrhizal fungus from the cultured gametophytes. These gametophytes develop on a nutrient medium containing minerals and sucrose. Thus, the cul- ture medium supplies the necessary nutrients, especially the organic material, which the gametophytes from nature obtain from the mycorrhizal fungus. Like other subterranean, pteridophyte gametophytes grown in axenic culture (Whit- tier, 1983; Whittier & Peterson, 1984), gametophytes of L. lucidulum develop without the mycorrhizal fungus if a carbon source is supplied. The gametophytes from nature and culture have a thickened, axial morphol- ogy with the possibility of apical branching. The apices of the gametophytes from both sources are the same, with the meristematic groove being overarched by a lip of immature dorsal tissue. In culture the dorsal and ventral surfaces are separated by an indentation along the lateral surfaces of the gametophyte. Al- though there appears to be no growth associated with the lateral indentations on the gametophytes from culture, they are continuous with the meristematic groove in the apical region. The gametophytes from nature, especially the larger ones, are partially to almost completely boat-shaped. It would appear that growth along the sides of these gametophytes, if only in the apical region, causes the ventral surface to be displaced upward along the lateral margin. The thickness of this marginally uplifted ventral tissue produces the boat-shape. This shape has not been observed in the cultured gametophytes, possibly because they are younger and smaller than those from nature. The morphology of the sex organs and their placement is the same on game- tophytes from both sources. The dorsal surface of the cultured gametophytes also have the paraphyses reported for gametophytes from nature (Spessard, 1922; Bruce & Beitel, 1979). The paraphyses are apically uniseriate but usually arise in pairs from biseriate bases. Spessard (1922) suggested that these paraphyses are mucilaginous. The paraphyses on the cultured gametophytes give no indi- cation of mucilage production. There is no mucilage formed by cells on the dorsal surface of the cultured gametophytes, although the rhizoids on the ventral surface have mucilaginous sheaths. The only viscid material found on the dorsal urface of some of the cultured gametophytes was sperm released during fixa- on. The internal structure of the cultured gametophytes is typical for the 1, ganas The cells in the ventral region are isodiametric, which is similar to the condition for gametophytes from nature. The radially elongated i Bs se ae mycor- thizal zone of the gametophytes of L. digitatum (Bruce, 1979) and L. -(Spessard, 1922) its not si in the gametophytes of L. lucidulum. The my- Corrhizal zone in L. lucidulum is similar to that found in L. selago (Bruchmann, 1910), in which the fungus occurs in the 5 or 6 layers of cells just internal to the ventral surface cells. In most of the sections of cultured gametophytes, the internal cells were sim- 54 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) ilar to each other. However, in some gametophytes there were differences in the contents of the internal cells. Five or six layers of cells in the ventral region lacked the large amounts of condensed tannins present in the more central and dorsal tissue. Except for differences in the amounts of the condensed tannins, the tissues of these gametophytes appear normal. Why all the gametophytes did not exhibit this condition is unknown. However, these sections appear to dem- onstrate that there is a potential difference in the ability of the internal cells to synthesize polyphenolic material or condensed tannins. The distribution of the condensed tannins in these gametophytes is of interest because the cells lacking condensed tannins are those which in gametophytes from nature contain the mycorrhizal fungus. Additional studies are needed to determine under what conditions the differences in the polyphenolic contents of the cells occur and to determine if the high amounts of polyphenolic materials usually correlate with the non-mycorrhizal portions of the gametophyte The studies on Lycopodium gametophytes from axenic culture, which now include three types of the subterranean gametophytes, demonstrate that a my- corrhizal fungus is not necessary for gametophyte growth. As long as the nutrient medium contains sugar, the gametophytes develop in culture without a mycor- rhizal fungus. The morphology and anatomy of these gametophytes from culture are essentially the same as for those from nature. Even though no fungus is present, the mycorrhizal zone differentiates in L. obscurum and L. digitatum and, in those gametophytes of L. lucidulum with large amounts of condensed tannins, the mycorrhizal zone is distinctive. These observations indicate that the internal and external structure of Lycopodium gametophytes is an inherent characteristic of the gametophytes and not dependent on the mycorrhizal fungus. ACKNOWLEDGMENTS The authors thank Dr. William G. Eickmeier and Dr. Dominick J. Paollilo for spore collections and Dr. R. James Hickey for gametophytes from nature. This study was supported in part by the 1 derbilt University Research Council. LITERATURE CITED Bruce, J. G. 1979. Gametophyte of Lycopodium digitatum. Amer. J. Bot. 66:1138-1150. Iara hen sts BEITEL. 1979. A community of Lycopodium gametophytes in Michigan. Amer. :33-41. hia A 1910. Die Keimung der Sporen und die ssp - Prothallien von Lyco- um clavatum L., L. annotinum L., und L. selago L. Flora 1 cae D. A. 1940. Plant microtechnique. New York: ile eae penta PEARSE, tee E. 1968. Histochemistry theoretical and applied. 3rd ed. Vol. 1. made Williams Wilkins Co. Sieceiits 9 A. 1922. Prothallia of Lycopodium in America II. L. lucidulum and L. obscurum Vér- io ideum. Bot. Gaz. (Crawfordsville) 74:392-413. year e palpi, The effect of sucrose on apogamy in Cyrtomium falcatum Presl. Amer. Fem . 54:20-25 - 1977. Gametophytes of Lycopodium obscurum as grown in axenic culture. Canad. J. Bot. 55:563-567. WHITTIER & WEBSTER: LYCOPODIUM GAMETOPHYTES 55 . 1981. Gametophytes of Lycopodium digitatum (formerly L. complanatum var. flabelli- forme) as grown in axenic culture. Bot. Gaz. (Crawfordsville) 142:519-524. . 1983. Gametophtyes of Ophioglossum engelmannii. Canad. J. Bot. 61:2369-2373. and R. L. PETERSON. 1984. Gametophytes of Botrychium lunarioides and their mucilage- coated rhizoids. Canad. J. Bot. 62:2854-2860. SHORTER NOTE A New Substrate for Ophioglossum palmatum in Florida.—Ophioglossum pal- mutum L. occurs on a variety of substrates throughout most of its geographic range (Mesler, Amer. Fern J. 64:33-39, 1974, and references therein). In Florida, the species has been reported by all previous authors to occur exclusively as an epiphyte among leaf bases of the Cabbage Palm [Sabal palmetto (Walt.) Lodd. ex Schultes]. We are here reporting its occurrence on another species, Saw Pal- metto [Serenoa repens (Bartr.} Small]. The plants were first found by Steve Farns- worth in Martin County (T39S, R40E, secs. 28-33). The site is a mixture of Pine flatwoods and Cabbage Palm hammock with scattered portions of Oak hammock. A unique combination of factors has allowed O. palmatum to colonize the Saw Palmetto including: 1) A large nearby population of O. palmatum growing on Cabbage Palms; this indicates proper climatic and site-specific conditions and Serves as an unusually sizable spore source for the species. 2) The presence of an unusual arborescent form of the Saw Palmetto, with trunks to nearly 3 m tall and an ample number of leaf bases available for colonization. 3) The absence of fire for probably 20-30 years. The site is bordered by a road to the south, a canal to the north, and by orange groves and houses to the west and east—all impediments to fire. This absence of fire is probably the most important factor. It contributes, at least in part, to the existance of both the spore source popula- tions and the arborescent form of the Saw Palmetto. The site is on private property and currently being cleared for cattle grazing. We understand that burning is not planned; therefore the plants may persist at the site for some time. The ferns were present on at least six separate Saw Palmettos and more than 20 Cabbage Palms within a single one acre tract. One of us (R.M.) has transplanted several Saw Palmettos and Cabbage Palms (that had otherwise been marked for cutting) bearing O. palmatum to his nursery. The occurrence of O. palmatum on a new substrate does not appear to be a sign of recovery by this rare and endangered species. Saw Palmettos ordinarily occur in habitats with a natural fire cycle. Since fire effectively eliminates both the plants of O. palmatum and their microhabitat, we do not expect this phe- nomenon to be frequent in Florida. A decrease in natural fires in areas near uman habitation or agricultural lands may slow the decline in numbers wet O. tum has been experiencing, though it is doubtful that this will offset _ due to direct habitat destruction by humans.—C.iFTON E. NAUMAN, Fairchil Tropical Garden, 10901 Old Cutler Rd., Miami, FL 33156, and RicHarD Moyroup, 202 Grove Way, Delray Beach, FL 33444. American Fern Journal 76(2):56-98 (1986) Revision of the Neotropical Fern Genus Cyclodium ALAN R. SMITH Department of Botany, University of California, Berkeley, CA 94720 In recent years, the tropical American genus Cyclodium Presl (Dryopterida- ceae) has been regarded as comprising two species (Christensen, 1905-1906; Copeland, 1947), a subgroup of Dryopteris comprising five species (Morton, 1939), or congeneric with Stigmatopteris (Kramer, 1978; Tryon & Tryon, 1982), with about 20 species. It has been variously allied to Phanerophlebia (Copeland, 1947), Ctenitis and Tectaria (Tryon & Tryon, 1982, p. 520), Soromanes (a segre- gate of Polybotrya; Christensen, 1938; Pichi Sermolli, 1977), and Thelypteris subgenera Goniopteris and Meniscium (Morton, 1939). A study of the constituent species, as well as allied genera, convinces me that none of these circumscriptions is tenable. In this paper, I admit six additional known species within its circumscription and describe two new ones. Thus de- fined, Cyclodium comprises 10 species and has its closest alliance with neither Dryopteris nor Stigmatopteris but probably with Polybotrya, also an exclusively American genus. The relationship of Cyclodium meniscioides, the type of the genus, with certain other species, particularly C. calophyllum, C. heterodon, and C. varians, is so close that one has some difficulty separating them specifically. There is also some evidence that hybridization is taking or has taken place to produce both sterile hybrids and possibly fertile allopolyploid derivatives. The more dissected and free-veined species (C. trianae, C. seemannii) are connected to those with anastomosing veins through C. inerme and agree with the type in microscopic details of indument, blade architecture, indusium, spores, and ve- TAXONOMIC HIsTORY Cyclodium was described in 1833 by Presl, who admitted three species: C. glandulosum (Blume) Presl [type from Java; =Pronephrium glandulosum (Blume) Holttum, a member of the Thelypteridaceae]; C. meniscioides (Willd.} Presl; and C. confertum (Kaulf.) Presl [=C. meniscioides]. He later recognized a fo species, C. abbreviatum Pres] [based on Aspidium abbreviatum Schrader]. Moore (1857-1862) transferred two additional species to Cyclodium: GC. heterodon (Schrader) Moore and C. cumingianum (Presl) Moore [=Anisocampium cum- ingianum (Presl) Moore, an athyrioid genus]. Hooker (1862) and Hooker and Baker (1867) treated Cyclodium as a section of Aspidium and included also a hookeri Baker [=Amphineuron opulentum (Kaulf.) Holttum, a thelyp- eroid]. A. R. SMITH: CYCLODIUM 57 with species now regarded as belonging to Arachniodes, Polystichopsis, and Lastreopsis. Morton (1939) realized the relatively close relationship between C. menis- cioides, with imparipinnate laminae, and species with a confluent, pinnatifid apex (varians, calophyllum), but he also included a species of Thelypteris subg. Goniopteris (Dryopteris clypeata Maxon & Morton) in this alliance. Kramer (1978) and Tryon and Tryon (1982) combined Cyclodium and Stig- matopteris in a single genus and used the later name Stigmatopteris in antici- pation of its conservation. This proposal (by Tryon & Tryon, 1981) was rejected by the Committee for Pteridophyta (Pichi Sermolli, 1982). No one appears to have questioned the artificiality of Stigmatopteris as de- fined by Christensen (1909, 1913), even though Christensen recognized two dis- tinct groups (Eustigmatopteris and Peltochlaena) based on lamina texture and the presence or absence of indusia. It is my belief that these two groups have no close relationship and should not be considered congeneric. Stigmatopteris is construed here in a narrow sense to comprise those species 12-25 of Chris- tensen’s (1913) revision, along with a few others more recently described. MATERIALS AND METHODS This revision is based almost entirely on the study of approximately 1200 collections (including duplicates) representing about 475 gatherings from 26 her- baria. I am grateful to curators and staff of the following institutions for loans of specimens: AAU, B, BM, BR, C, CAY, COL, F, FI, G, GB, GH, K, M, MICH, MO, NY, P, PORT, RB, U, US, VEN, W, Z. Peter Edwards has kindly sent fixed chromosome material from specimens of Cyclodium grown at Kew. He has also provided important information on the habit and distribution of one of the new species, C. akawaiorum, which was collected on a recent expedition to Mt. Roraima. ae Spore samples for SEM were mounted on aluminum stubs using television tube coat. Mounted samples were allowed to dry for 24 hours before being coated with approximately 25 nm gold using a Polaron V sputter coater. SEM was done on an ISI DS-130 at an accelerating voltage of 8 or 10 kV. Facilities were provided by the Electron Microscope Laboratory, Life Sciences Building, University of California, Berkeley. Nelson Barton performed the sample prep- aration and microscopy. nes I thank Colleen Sudekum for executing most of the drawings of individual species. Charlotte Hannon provided the drawings of Cyclodium akawaiorum and aided in preparation of the distribution maps. Correspondence with Robbin Moran, Illinois State Natural History Survey, on Cyclodium and related genera resulted in discussion and insight into intergeneric relationships. MorRPHOLOGY Rhizome and scales.—Rhizomes of Cyclodium species are woody, on eae and long- to short-creeping. Internode length is correlated with habit, with those 58 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) at att wt ro. lavit d Pol p i y ybotrya. A. Rhi of Cyclodium inerme, Steyermark et al. 122169, UC. B. Rhizome of Polybotrya serratifolia (Fée) Klotzsch, Smith 922, UC. C. Stipe base of Cyclodium guianense, Liesner & Gonzdlez 10375, UC. s = sclerenchyma; vb = vascular bundle; vt = ventilation band; bs = brachysclereid; lt = leaf trace. species that are hemiepiphytic (C. meniscioides, C. akawaiorum, C. varians) having longer creeping rhizomes and more distant stipes than those that are terrestrial. Individual meristeles are not surrounded by a blackened sclerenchy- matous sheath, are more variable in size, and are arranged in a more irregular ring (Fig. 1A) than in species of Polybotrya (Fig. 1B). Numerous patches of black- ened sclerenchymatous cells (brachysclereids) dot the stele throughout, at least in older rhizomes of Cyclodium (Fig. 1A). The rhizomes bear densely overlapping, linear-lanceolate scales at the apex, but these become less visible and abraded back from the apex. In general, the scales are brown with a denticulate margin; sometimes they have a slightly darkened midband. They are subentire to denticulate along the margin (Figs. 6E, 7F, 9H). These sca ] hat similar to those in some species of Polybotrya and are readily distinguished from scales of most Dryopteris and Stigmatopteris secs which have scales that tend to be broader, less rigid, and more uniform in color. Axis vasculature and sulcation.—In cross section, stipes of Cyclodium have numerous (generally 8-16) round to oval meristeles arranged in a ring, with the two largest bundles being at the adaxial side of the ring (Fig. 1C). This pattern is typical for dryopteroid ferns and contrasts markedly with the pattern observ in athyrioid and thelypteroid ferns, where there are generally only two vascular bundles in the stipe. The outer portion of the stipe consists of a band of scleren- chyma that is interrupted only by two relatively narrow ventilation bands along the adaxial flanks. Fronds of Cyclodium are anadromous {i.e., the lamina is polystichoid; Chris- tensen, 1920), with the acroscopic segment arising before the basiscopic one (Fig. 13C, D, F, G). This is true of both proximal and medial pinnae. In the case of simply pinnate species, the first main lateral vein arises acroscopically (Figs. 7K: 11B, C, F). Anadromous architecture is the primitive condition in the close related genera Polybotrya (Moran, in litt.) and Maxonia, and also in the less closely related genera Polystichopsis and Arachniodes. Axes of Cyclodium are slightly to deeply grooved adaxially, and the grooves of the costae are more or less continuous with grooves of the main rachis. This A. R. SMITH: CYCLODIUM 59 pattern is like that of most dryopteroid genera and differs from that found in tectarioid genera such as Ctenitis and Tectaria, which have axes raised or flat- tened adaxially. Lamina.—The texture of the lamina in Cyclodium is usually chartaceous to subcoriaceous and rather firm. In this feature it is similar to Polybotrya, Olfersia, and Maxonia, but it differs from Stigmatopteris, which generally has thinner laminae. Cyclodium also lacks the internal glands (translucent areas in the lam- ina) characteristic of most Stigmatopteris species. Laminae vary from 1-pinnate to 2-pinnate or more divided. More dissected species (C. seemannii, C. trianae) greatly resemble certain species of Polybotrya (e.g., P. caudata). Even the less divided ones, such as C. meniscioides, look much like some simply pinnate Polybotrya species (e.g., P. serratifolia) and are sometimes determined as such in herbaria. This resemblance in lamina cutting strongly suggests a near rela- tionship of the two genera. In Cyclodium, species with a pinnate lamina, like C. meniscioides, have a subconform apical “‘pinna” (Fig. 7G) instead of a pinnatifid apex, and the lamina is thus imparipinnate. This particular lamina shape is often correlated with loss of dissection and can be seen in many unrelated genera, e.g., Thelypteris, Cyathea, Cnemidaria, and Diplazium. Those species of Cyclo- dium with the least divided laminae also show increased tendency toward di- morphism (e.g., C. akawaiorum, Fig. 6A; C. calophyllum, Fig. 7A, B). This may be in part related to the development of a hemiepiphytic habit, which is often correlated with dimorphy in many unrelated genera, e.g., Lomariopsis, Polybot- tya, Maxonia, Lomagramma, and Bolbitis. Venation.—On the basis of outgroup comparison, the general and primitive condition in Cyclodium is for the veins to be simply pinnate and unbranched in the ultimate segments. This pattern occurs in C. inerme (Fig. 11H), C. gui- anense (Fig. 9D), C. rheophilum, C. trianae (Fig. 13D, G), and G. seemannii. With oss of dissection and increased webbing, there is the almost inevitable conse- quence of anastomosing of veins, from weakly united (C. heterodon var. ab- breviatum, Fig. 11F), to one or two pairs united at a more oblique angle (C. Varians, Fig. 9K), to several pairs more strongly united between the costa and Margin (C. heterodon var. heterodon, Fig. 11C; C. calophyllum, F ig. 7A, B; C. akawaiorum, Fig. 6C), to many pairs regularly united (veins meniscioid: C. me- Niscioides, Fig. 7G, I, K). These differences in venation were used to distinguish Cyclodium in the restricted sense, comprising one or two species. However, this character by itself is inadequate to delimit Cyclodium, as shown by the transi- tional venation types found in C. calophyllum, C. heterodon, and C. varians. ny genera of ferns contain both free-veined and reticulate-veined species, &8. Thelypteris, Polybotrya, Polypodium, and Pteris, and in general this char- acter is a poor one with which to circumscribe genera. : : Vein endings in Cyclodium are not usually visible without partial sop of the lamina, but in general they reach nearly to the margin. There is a margin commi 1 vein in a few species, e.g., GC. akawaiorum (Fig. 6C), a feature also found in Polybotrya polybotryoides (Baker) Christ and Olfersia. Vein endings in Stigmatopteris are usually easily visible (viewed from the adaxial surface}, enlarged or clavate at the tip, and terminate well back of the margin. AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) ce ae ‘ or r ; ee Se oe a a ~ eae ae ; Spores of Cyclodium. A, B. C. guianense, Poole 1958, NY; bar = 7.5 um. C, D. C. guianense, Davis 180, GH; bar = 14.5 and 2.3 um, respectively. E. C. varians hybrid, Jenman s.n., NY; bar ~ 7-9 um. F, G. C. rheophilum, de Granville 2586, F; bar = 7.5 and 1.9 um, respectively. Indument.—Cyclodium species are devoid of acicular hairs (as Thelypter's has), but bear short, spreading, blunt, yellowish or hyaline hairs 0.05-0.2(0.3) mm long. These hairs are most obvious in the grooves of the rachis and costae adax- ially and at the base of the costae abaxially. Such hairs are most numerous an A. R. SMITH: CYCLODIUM 61 Fic. 3. Spores of Cyclodium. A, B. C. inerme, Tillett et al. 44942, NY; bar = 14.7 and 7.6 um, respectively. C. C. inerme, Maas & Westra 3990, U; bar = 29 um. D. C. trianae var. chocoense, Killip & Cuatrecasas 38913, US; bar = 7.7 um. E. C. trianae var. trianae, Killip & Smith 28540, US; bar = 74 um. F. C. heterodon var. abbreviatum, Neowied s.n., BR; bar = 19.4 um. longest in C. trianae and GC. seemannii, but at least a few can usually be found in all species. Scattered, uniseriate (basally 2 or 3 cells wide in some species), brownish, appressed, hairlike scales occur on the costae abaxially in most species. Very similar hairs (both the spreading hyaline type found on axes and the brownish appressed hairlike scales) occur in some species of Polybotrya, but are acking or are not readily apparent in Stigmatopteris. Cyclodium guianense has broader, lanceolate, lighter colored scales on the costae abaxially (Fig. 9F). Most Cyclodium species lack glands, but C. seemannii bears yellowish, resin- ous, sessile glands on the lamina abaxially (Fig. 13B). A few specimens of C. guianense and C. trianae have similar glands (Fig. 13E). All species have scat- tered, minute (0.1-0.3 mm), appressed, septate, reddish or brownish hairs (also reduced scales?) on the leaf tissue and veins abaxially. Indusia.—With the exception of C. trianae and C. seemannii, which have orbicular-reniform indusia with a narrow sinus (Fig. 13B, D), Cyclodium species ve peltate indusia (Figs. 7D, H; 9E; 11). It seems likely that the peltate con- AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) bar = Fic. 4. Spores of Cyclodium. A, B. C. meniscioides var. meniscioides, Lleras et al. P17441, U; - = pm. 7 and 2.0 um, respectively. C. C. Sema var. prumgnarie Schulz 8695, U; bar = 1 eek D. C. npSinesieee K.E.R. 167, pt, UC; bar = 14.7 um. E. C. meniscioides var. rigidissimum, 2.3 wm, 167, pt, UC; = 7.5 pm. FG. ibsice sorbifolia, Glaxo piney C; bar = 7.5 and respectively. dition is derived in the genus, and that orbicular-reniform indusia represent an ancestral condition. Both character states are known in other dryopteroid guess Closest relatives Polybotrya and Olfersia are exindusiate, while Maxonia - orbicular-reniform indusia. More distant relatives are either exindusiate (Stig A. R. SMITH: CYCLODIUM 63 matopteris) or have orbicular-reniform indusia (Polystichopsis, Dryopteris, Arachniodes). In Cyclodium, indusia are nearly always caducous: consequently they may not be present on mature fronds and many herbarium specimens. This feature distinguishes the genus from many other dryopteroid genera, where at least some indusia are retained even after spores are shed (as in Dryopteris). SPORES Spores of Cyclodium are bilateral with a monolete scar. A variously folded and ornamented perispore surrounds a relatively smooth exine (Fig. 2A). The spores are mostly 44-65 wm long and 30-44 um wide. The most common pattern is exemplified by spores of C. inerme (Fig. 3A, B), which has coarse rugae (in- flated folds) with fine papillate surface projections. Similar spores occur in C. trianae var. trianae (Fig. 3D, E). A more intricate pattern of rugae occurs in C. guianense (Fig. 2, B-D). In C. rheophilum, the folds are less prominent, but the surface ornamentation is denser (Fig. 2F, G). Spores of Cyclodium meniscioides have a perispore with relatively few folds but bear a dense, anastomosing net- work of surface projections (Fig. 4A, B, E). There appears to be considerable variation in the surface pattern in this species (Fig. 4C), but this may be a function of spore maturity. Anomalous spores, without the usual pattern of ridges and ace ornamentation, were seen in C. heterodon (Fig. 3F) and C. akawaiorum (Fig. 4D). These spores were also considerably larger than usual, 100 wm long or More. Spores of a putative hybrid involving C. varians were mostly strongly malformed, but a few were densely echinate (Fig. 2E). In some samples, a rather high proportion of the spores seemed collapsed (Fig. 3C). This is like the situation found in Bolbitis by Hennipman (1977), where Presumably fertile and sexual species produce a high percentage of malformed Spores. In overall aspect, spores of Cyclodium are rather similar to spores of many other dryopteroid ferns (compare, e.g., illustrations of Didymochlaena truncatula (Swartz) J]. Smith and Cyc lopeltis i lata (Swartz) J. Smith in Tryon & Tryon, 1982, pp. 518, 486). They are especially similar to spores of some Polybotrya species, e.g., P. fractiserialis (Baker) J. Smith (Tryon & Tryon, 1982, p. 541) and P. caudata Kunze (Moran, ms. in prep.). They also greatly resemble spores of Olfersia cervina (L.) Kunze (Tryon & Tryon, 1982, p. 541, as Polybotrya cervina). Stigmatopteris nothochlaena (Maxon) C. Chr. [=S. jamaicensis (Desv.} Proctor] (Tryon & Tryon, 1982, p. 522) has similar spores, but they lack the fine surface omamentation of most Cyclodium species. Other Stigmatopteris species, e.8., : Prionites (Kunze) C. Chr. and S. contracta (Christ) C. Chr., seem also to lac © surface ornamentation (A. Tryon, ms. in prep.). There is somewhat a resemblance to spores of Polybotrya sorbifolia Mett. ex Kuhn (Fig. 4F, G). 3 Seneral, Cyclodium species lack the finely and densely echinate surface of many Polybotrya species (see, e.g., figs. 80.24, 80.28, 80.29 in Tryon & Tryon, 1982; Moran, ms. in prep.). Because of parallelism, convergence, and the extent of Variation within genera, it seems doubtful that spore surface morphology alone can be used to characterize genera of dryopteroid ferns. 64 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) CHROMOSOME NUMBER Only two previous counts, both for C. meniscioides, have been made in Cy- clodium. Walker (1973) reported n = ca. 42 for a plant from Guyana [Evemy (=Every) & Burgess 141, cultivated at Kew], while Silva Araujo (1976) reported 2n = 60 for a plant from Manaus, Amazonas, Brazil (Araujo 244, INPA). The latter is most likely an erroneous count or a misidentification; I have not seen the voucher. I have obtained additional counts on cultivated material of C. meniscioides that establish the base number for this genus as x = 41. Every & Burgess 57 (Kew accession no. 681-69-06186), K!, from Mataruma, Kirikava, Guyana, and Every & Burgess 258 (063-70-00585), K!, from Essequibo, Guyana, showed un- equivocally 2n = 41 II at meiotic metaphase. Similarly, Leppard 1278 (479-74- 04441), K, from Brazil, showed 2n = 41 II. The closest relative, Polybotrya, Olfersia, and Maxonia, have all been re- ported as n = 41 (Walker, 1966), so this number in Cyclodium is not surprising. Stigmatopteris s.s. also has the same base number (Walker, 1966), as do nearly all dryopteroid ferns. HYBRIDS Several herbarium specimens suggest that hybridization occasionally takes place in Cyclodium. The primary area of sympatry, and thus the most likely area for hybridization to occur, is in the Guianas. Jenman s.n. (NY), Guyana, Pome- roon River, combines features of several species, particularly C. varians, C. me- niscioides, and C. inerme. The long-creeping rhizome suggests that it is a hemi- epiphyte, and the pinna lobing is very irregular. A few spores (Fig. 2E) are seemingly well-formed, but most are decidedly irregular. Several collections from Malali, Guyana (Jenman s.n., NY) are also suggestive of hybridization. De la Cruz 2880 (F, GH, NY, UC), from Kamakusa, Guyana, is also very irregular in pinna lobing and may represent a hybrid between C. meniscioides and C. varians. A collection from Para, Brazil (Spruce 26*, K) appears intermediate between C. inerme and C. meniscioides. Another collection from Pernambuco, Brazil (Gardner 1218, K), appears intermediate between C. heterodon and C. meniscioides. All of these putative hybrids are apparently rare and are thus poorly known, all but one lack mature spores. Little more can be said about their origin and parentage until apparently hybrid plants can be studied in the field or until cytological materials can be obtained. ECOLOGY AND DISTRIBUTION Species of Cyclodium occur in relatively undisturbed lowland and lower mon- _ tane rainforests in the American tropics, from Trinidad and Panama throughout northern South America to Amazonian Bolivia, northern Argentina, and Para- guay. The preferred habitat, as judged from herbarium labels, appears to be i? A. R. SMITH: CYCLODIUM 65 rain forests and gallery forests, along streams, and, in the case of at least C. meniscioides, in waterlogged soils and swamps. Most Cyclodium species occur between 0 and 800 m elevation. Exceptionally, C. guianense, C. meniscioides var. meniscioides and C. trianae var. chocoense are found at elevations up to 1450 m. The only species regularly known from middle elevations is C. akawaiorum, all collections of which have been made from 1000 to 1500 m. The center of diversity is the Guianas, where six of the ten species occur; three others are centered in western Venezuela, Colombia, and Panama; an additional species occurs in coastal Brazil. The most wide-ranging species, C. meniscioides, occupies a range nearly coextensive with that of the genus; it is absent only from Panama and western Colombia. Five species (C. akawaiorum, calophyllum, theophilum, seemannii, and varians) and several varieties have very restricted ranges and are represented by relatively few collections in herbaria. A few species, particularly C. inerme, C. guianense, C. meniscioides, and C. trianae var. chocoense, seem to be common understory ferns in portions of their ranges; in view of the widespread nature of many neotropical lowland ferns, it is a little surprising that these cyclodiums do not range into Central America, the Antilles, and many parts of South America. Although most species of Cyclodium are terrestrial, several show a tendency to hemiepiphytism. The most notable in this regard is C. akawaiorum, which has been recorded as climbing five meters into the trees. Many collections of C. meniscioides have been noted as “epiphytes” on herbarium labels, but in gen- eral this species does not appear to climb more than one or two meters up trunks; it and other species of the genus are apparently always rooted in the ground. Data on C. varians is meager, but it probably also climbs trees in certain situa- tions. One species, C. rheophilum, is a probable rheophyte; others, as C. gui- anénse, may occasionally occur on rocks, but probably do not grow in stream beds. The remaining species are known as occurring only terrestrially. RELATIONSHIPS Interspecific.—Relationships within Cyclodium can be deduced from overall morphological similarities and differences (Fig. 5). Clearly, C. inerme is most closely related to a trio of species, C. guianense, C. heterodon var. abbreviatum, and C. trianae var. chocoense; characters indicating these affinities are the free venation and the pinnate-pinnatifid laminae. The last two taxa are allopatric with respect to C. inerme, and it seems likely that ecogeographic speciation has Occurred. Varieties abbreviatum and chocoense appear to intergrade with var. heterodon and var. trianae respectively. Variety trianae is that member of the Senus that most closely resembles Polybotrya and can serve as a possible evo- lutionary link to that genus; it is also closely related to the highly local and rare Cyclodium seemannii. The origin of C. meniscioides, with highly anastomosing venation and subentire pinnae as probable derived character states, Is uncertain, but it Seems most closely related to and possibly derived from C. heterodon var. heterodon. The highly local C. akawaiorum may be derived from C. menis- 66 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) rheophilum akawaiorum varians meniscioides var. rigidissimum guianense ane ? calophyllum¢€ : meniscioides inerme heterodon _ heterodon var. abbreviatum ~ var. heterodon trianae var. chocoense Polybotra ores Maxonia trian 1. var. tranae——— —> seemannil Fic. 5. Hypothetical relationships in Cyclodium. Arrows indicate direction of evolution. cioides var. rigidissimum, but also shows some resemblance to C. varians. Cy- clodium varians and C. calophyllum are both very restricted in range and show intermediacy in several characters between more wide-ranging species; it seems possible that they are relatively recent hybrid derivatives. Further investigation of this hypothesis is desirable. Cyclodium rheophilum is most closely related to and probably derived from C. guianense through ecological specialization. e rooting of a phylogenetic tree depicting the relationships in Fig. 5 is un- certain, but it seems most likely that C. trianae var. trianae and C. seemannil represent the most primitive elements in the genus. Character states that have been used to arrive at this hypothesis are indicated in Table 1. In Fig. 5, I have indicated evolutionary direction only where morphological specializations pro- vide good evidence. Suprageneric.—The affinities of Cyclodium are without doubt with other dryopteroid genera. Characters uniting this assemblage are the usual presence of round-reniform or peltate indusia; bilateral spores with a variously wing Perispore; pinnate or often more dissected lamina; midribs of penultimate se8- ments grooved, with the rachis-groove usually open to admit the pinna-groove: pesmi: vascular bundles in the stipe bases; dictyostelic, scaly rhizome; and x = _ More specifically, Cyclodium is most closely allied to Polybotrya, differing from that in the presence of indusiate sori, the sporangia always in discrete, le A. R. SMITH: CYCLODIUM 67 TaBLE 1. Character State Polarities in Cyclodium. Primitive Derived abit Terrestrial Hemiepiphytic, rheophytic Dissection Decompound Pinnate-pinnatifid, 1-pinnate Apex Pinnatifid Con Veins Free Anastomosing Areoles 1 or 2 series >2 series Submarginal vein Absent Present Laminar glands Absent Present Indusia Sinus present Peltate round sori (never acrostichoid); monomorphic or weakly dimorphic fronds; lack of a strong hemiepiphytic habit; rhizome lacking black sclerchymatous sheaths surrounding the meristeles; and in the spores not densely and sharply echinate. The monotypic genus Maxonia is also closely related but differs in being strongly dimorphic, having a more dissected sterile lamina (to 3-pinnate-pinnatifid), pos- sessing a dorsiventral rhizome, and having densely echinate spores. Olfersia, also monotypic, differs in having parallel venation with vein tips united by an intramarginal vein, a conform apical “pinna,” strongly dimorphic fronds, acros- tichoid fertile pinnae, and the absence of indusia; rhizome and stipe base scales are also lighter in color. Stigmatopteris, with which Cyclodium has traditionally been combined or associated, is a more distant relative. It differs from Cyclodium primarily in being exindusiate, having pellucid-punctate glands on the lamina (most easily seen with transmitted light), pinnae with serrate apices, generally thinner laminae, lighter brown, thin, ovate-lanceolate scales on the stipe base and costae abaxi- ally, and in having veins ending in a clavate tip well back from the margin (as Seen on the adaxial side of lamina). Stigmatopteris species are always terrestrial or epipetric and have a short-creeping or suberect rhizome. Some Cyclodium species have been thought to possess punctate laminar glands (e.g., by Christen- sen, 1913, p. 74), but I have been unable to verify this. Sometimes the stomata (guard cells) appear lighter green than the surrounding lamina when viewed at about 30 times magnification and could easily be confused with pellucid dots of the kind seen in the lamina of Stigmatopteris. Dryopteroid genera with which Cyclodium has a more remote relationship include Phanerophlebia, Dryopteris, Arachniodes, and Polystichopsis. The first of these differs from Cyclodium in having spinulose margins, 1-pinnate fronds, and the absence of minute hairs in the costal grooves. Dryopteris usually has a Catadromic lamina distal to the proximal pinnae, reniform or round-reniform indusia, a suberect rhizome, and more ovate, lighter-colored stipe base scales. Arachniodes usually has a much more dissected lamina and round-reniform indusia. One species generally placed in Arachniodes, A. macrostegia (Hook.} Tryon & Conant, may have affinity to Cyclodium, particularly to C. trianae. It agrees with Cyclodium in having an anadromous lamina, a somewhat creeping izome, and in being found at low elevations, but differs in having a more dissected lamina, round-reniform indusia, and in lacking minute costal hairs 68 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) adaxially. Polystichopsis differs in having round-reniform indusia (or is exin- dusiate), axes densely long-pilose (some hairs acicular), generally thinner lami- nae, broader, more flaccid stipe base and rhizome scales, and ultimate segments that are usually mucronate at the tip. One species, P. ochropteroides (Baker) Morton, has a large greatly dissected lamina (4-pinnate) and approaches Arach- niodes macrostegia in many characteristics. It was synonymized with that species by Tryon and Tryon (1982, p. 501) and also by Proctor (1985), but differs in the long-pilose axes and occurrence at higher elevations. In assessing the relationship of Cyclodium, two New World tectarioid genera (for characterization of this group, see Holttum, 1984, and references cited there- in), Ctenitis and Lastreopsis, have also been considered as possible allies but rejected on the basis of several characters: a lamina in which the axes of the penultimate segments are not grooved adaxially (often slightly raised), with cte- nitoid hairs (short, multicellular hairs lacking an acicular tip, often with reddish cross walls, the cells collapsing upon drying) on this axis. Several relatives of Cyclodium show a tendency toward anastomosing veins; Polybotrya, Stigmatopteris, and Phanerophlebia all have some members with this condition. Only Phanerophlebia agrees with most cyclodiums in having pel- tate indusia. Several genera are similar to Cyclodium in being primarily (or primitively) anadromous: Polybotrya, Maxonia, Polystichopsis, and at least some Arachniodes. Strong dimorphism is found in Polybotrya, Maxonia, and Olfersia, but several cyclodiums show a tendency toward dimorphism. It seems likely that the common tendencies shown by various dryopteroid genera are a reflection of common ancestry and perhaps even reticulate evolution. These shared charac- ters make generic delimitation in this group especially difficult. TAXONOMIC TREATMENT, \ a4 Cyclodium Presl, Tent. Pterid. 85. 1836.—Aspidium sect. Cyclodium (Pres) Hooker, Sp. Fil. 4:36. 1862.—Lectorype (chosen by J. Smith, Hist. Fil. 203. 1875): Aspidium confertum Kaulf. [= Cyclodium meniscioides (Willd.) Pres] var. meniscioides]. Hooker referred to infrageneric taxa (denoted by the symbol §) in Aspidium as subgenera or sections (p. 5: 6), but called Aspidium § Cyrtomium (p. 37) a section; hence I adopt sectional rank for his Aspidium § Cyclodium. The type of Cyclodium has generally been considered to be Aspidium meniscioides (Christen- sen, Ind. Fil. xxv. 1906; Copeland, 1947; Index Nominum Genericorum 1:470. 1979; Tryon & Tryon, 1982, p. 519); however, John Smith (1875) selected A. confertum, one of the original species, and I see no reason to reject his choice. Dryopteris subg. Stigmatopteris (C. Chr.) C. Chr. group Peltochlaena Fée ex C. Chr., Kongel. Danske Vidensk. Selsk. Skr., Naturvidensk. Afd. VIL. 10: 74. 1913.—Lectorypr (chosen here): Dryopteris subobliquata (Hook.) Kuntze [=Cyclodium inerme (Fée) A. R. Smith. The other syntypes are Dryopteris varians (Fée) Kuntze and D. sancti-gabrieli (Hook.) Kuntze. A. R. SMITH: CYCLODIUM 69 Peltochlaena Fée, Gen. Fil. (Mém. Fam. Foug. 5) 289. 1852. nom. provis. et ille- it—T ype: “Peltochlaena nephrodiiformis Fée.” Fée clearly adopted this as a provisional name (‘Nous indiquons parmi les cyclodiées .. . un genre que nous n’osons pas constituer définitivement, la description que nous donnons de |’espéce type, la seule que nous connaissions, le fera suffisamment connaitre a titre provisoire.”); thus it has no taxonom- ic standing [ICBN, Art. 34.1(b)]. (Pichi Sermolli, Webbia 32:351. 1978). Plants terrestrial or hemiepiphytic; rhizome dictyostelic, creeping, clothed at and near apex with lanceolate or linear-lanceolate, usually castaneous scales, these entire or denticulate; trophopods lacking; stipes non-articulate, with 8-16 vascular bundles arranged in a ring; laminae monomorphous to often slightly or decidedly dimorphous (the fertile fronds more erect, longer-stiped, and with contracted fertile segments), usually pinnate to 2-pinnate-pinnatifid (rarely sim- ple), anadromous with the segments or pinnules on the acroscopic side of a pinna generally longer and arising first (if simply pinnate, then an acroscopic lateral vein arising first); laminae gradually reduced to a pinnatifid apex or imparipin- nate; rachis and costae adaxially grooved, the grooves + confluent from one axis to the next; grooves adaxially and sometimes rachis and costae abaxially with uniformly short, blunt-tipped hairs 0.05-0.2(0.3) mm long; costae abaxially with uniseriate (basally to 3 cells wide in a few species), brownish or reddish-brown, appressed hairs (actually reduced scales); veins free and the basal pair of a group meeting the margin at or near the sinus, or connivent below the sinus, or regu- larly anastomosing (uniting in pairs) to form a series of areoles, these united veins often producing an excurrent veinlet; lamina chartaceous to subcoriaceous or coriaceous, not translucent, lacking pellucid-punctate glands; indusia peltate (8 spp.) or round-reniform (2 spp.), caducous often before the sporangia mature, glabrous or glandular-ciliate on the margin and sometimes the surface; sporangia long-stalked, the stalk 0.4-0.5 mm long, capsules often abscissing at the tip of the stalk; spores bilateral, monolete, with a folded, winglike perispore; x = 41. KEY TO SPECIES AND VARIETIES 1. Veins regularly anastomosing; laminae monomorphous to strongly di- morphous; indusia peltate. 2. Laminae with a + conform, entire or crenate apical segment. 3. Lateral pinnae ovate-lanceolate, up to 11 cm long and 2.3 cm wide; costae abaxially lacking scales; indusial margin entire; areoles 1- or 2-seriate between costa and margin. ....------- + © abawaiorum ios) c 0] “ “ i) pour 2 PT =) =) ie}) ic?) ®, 5 5 3 bo a i?) = 2 7-seriate. i ile pi long, 3cm wide ..... 4. Pinnae serrulate; sterile pinnae to 11 cm long, 3. pene 6b. C. meniscioides var. paludosum 70 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) 5. Veins strongly raised on adaxial surface; pinnae of sterile fronds approximate and overlapping ..................... je eee eee . 6c. C. meniscioides var. rigidissimum 5. Veins not strongly raised on ages surface; pinnae distant, not overlapping .......... a. C. meniscioides var. meniscioides 2. Laminae with a confluent, joewey teal segment. 6. Costae and rachis abaxially persistently scaly ..... 2. C. calophyllum 6. Costae and rachis abaxially lacking scales. 7. Rhizome long-creeping (sometimes hemiepiphytic?), with dark purplish-brown linear scales; laminae subdimorphous, the fer- tile with pinnae 0.8-1.5 cm wide; (Guyana, Trinidad) ......... OS eS eee 10. C. varians 7. Rhizome short-creeping (plants terrestrial), with brownish, lin- ear-lanceolate scales; laminae monomorphous, pinnae usually more than 2 cm wide. 8. United veins 2 or 3 pairs; pinnae sinuate ................. es... 4a. C. heterodon var. heterodon 8. United veins 1 pair; pinnae pinnatifid, with a cuneate base SS ea ere 4b. C. heterodon var. abbreviatum 1. Veins all free; laminae monomorphous; indusia peltate or orbicular-re- niform. 9. Indusia peltate; laminae pinnate to pinnate-pinnatifid. 10. Pinnae linear-lanceolate, less than 1 cm wide, crenate; plants rheophytic; (French Guiana) .................. 7. C. rheophilum 10. Pinnae lanceolate, greater than 1 cm wide, crenately lobed to pinnatifid. ibs ipa daha cells wide on costae abaxially; pinnae (15)20- Bee cs ee ck ew 3. C. guianense ot. ee lacking on costae abaxially; pinnae 9-15(24) pairs. 12. Rhizome scales 3-5(7) mm long, relatively inconspic- uous; sori often impressed (lamina embossed adaxially) Re ee ce es wees 5. C. inerme 12. Rhizome scales ca. 1 cm long, often conspicuous; sori not or only slightly impressed ..............---++-+: ee SS Oe 4b. C. heterodon var. abbreviatum 9. Indusia orbicular-reniform, with a narrow sinus; laminae pinnate- pinnatifid to bipinnate-pinnatifid. 13. Lamina abaxially with yellowish to orangish, — — L SRO Re, WC mIIR ee ROR Tee. ge moa a Sa fe tg seemannu 13. Lamina lacking such glands. 14. Laminae bipinnate throughout most of their length, largest pinnules pinnatifid, constricted at their base and short- simran 6 9a. C. trianae var. trianae 14. Laminae shallowly to usually deeply pinnate-pinnatifid, if — the pinnules entire to crenate, adnate and de- ee 9b. C. trianae var. chocoense A. R. SMITH: CYCLODIUM 71 re 1 Cyclodium akawaiorum A. R. Smith, sp. nov. (Fig. 6).° WY LO ta oe Gi LL = Ae ag \N\ EL; => (EZ. WO SE NNNO SS SN cot Xa J oe ae aN AN ( =\ Fic. 6. Cyclodium akawaiorum, Persaud 110, NY. A. Habit. B. Juvenile frond. C. Sterile pinna. D- of rhizome apex. Sori and indusia. E. Scale i A. R. SMITH: CYCLODIUM 73 C. akawaiorum. Those few collections in herbaria are often identified as Poly- botrya, probably because of the climbing rhizome and general similarity to some simply pinnate Polybotrya species. Most collections seen lack fertile fronds. Although there is as yet inadequate information concerning several species of Cyclodium, C. akawaiorum is apparently the only species in the genus that is a high-climbing hemiepiphyte; C. meniscioides is sometimes a low-climbing hemi- epiphyte, as may be C. varians. Notes with the type collection indicate that plants were common but mostly sterile, growing “from bases of bryophyte covered trees, under trees amongst buttress roots, and also on some palm stems, climbing to 15 ft., stems with conspicuous roots to only 6” above ground level... . Spore- lings and young plants common in thick humus on forest floor .... Stock soon becomes creeping, and climbs a convenient support.” All Venezuelan collections were also noted as climbing or “epiphytic” by their collectors. Peter Edwards (in litt.) indicates that the Persaud collection was probably made from the same population as K.E.R. 76. The species epithet honors the local Akawaio Indians who, according to Edwards, were essential in conducting the Kew Expedition to Roraima (K.E.R.); one individual in particular was the first to spot the new Cyclodium. 2. Cyclodium calophyllum (Morton) A. R. Smith, comb. nov. (Fig. 7, A-D).— 48}5 Dryopteris calophylla Morton, Bull. Torrey Bot. Club 66:49. 1939.— Type: Colombia, Dept. Santander, Magdalena Valley, vicinity of Bar- ranca Bermeja, 100-500 m, Haught 1353 (US, 3 sheets!; isotypes GH! MICH). Rhizomes creeping; fronds to ca. 1.5 m long, somewhat dimorphous, the sterile shorter and with wider pinnae than the fertile; stipes ca. “2-4 the lamina length, stramineous to tan, short-pubescent with hairs ca. 0.2 mm long or less, also with scattered, persistent, linear-lanceolate scales to 6 mm long, these brown, dentic- ulate; rachises with scales similar to those of stipes; laminae ca. 60-80 cm long, up to 30 cm wide, pinnate with (2}15-18 pairs of lateral pinnae, these subabruptly reduced distally to a shallowly pinnatifid apex; pinnae subentire, crenulate (ster- ile), or shallowly lobed (fertile) 1% or less the distance to costa (lobes as wide as or wider than long), up to 25 cm long, the fertile 2.0-2.5 cm wide, the sterile 3.0- 3.5 cm wide, lowermost stalked up to 10 mm, rounded to truncate at base, acute to acuminate at apex, + equilateral or acroscopic side slightly wider and with a small auricle at base; basal anterior veinlets free, next 3 or 4 pairs of veinlets anastomosing to form 3-4 series of areoles between costa and margin, with (30}40- 60 pairs of lateral veins per pinna, these 3.5-5.0 mm apart; costae {and — obviously costules) abaxially with numerous filiform to linear-lanceolate, light brown, twisted scales, these uniseriate at apex and leaving a reddish spot after dehiscence; laminae chartaceous to subcoriaceous; sori in 3-5 ~ regular rows between costa and margin, discrete, with relatively large. brownish, peltate, gla- brous, entire indusia. Distribution.—Known only from Depts. Santan Colombia, and Edo. Zulia, Venezuela (Fig. 8). All c vations, 100-500 m. der and Norte de Santander, ollections are from low ele- 74 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) = —)-— Mett., Ann. Sci. Nat. Bot., V. 2:243. 1864.—Nephrodium trianae (Mett.) (70% 4 Baker, Syn. Fil. 286. 1867.—Dryopteris trianae (Mett.) Kuntze, Rev. Gen. Pl. 2:814. 1891.“Typr: Colombia, Prov. de Barbacoas [Narifio], via de Tuquerres, Triana “32” (B!; isotype BM! photos UC! US!). a4 egy Nephrodium firmifolium Baker, Syn. Fil., ed. 2. 501. 1874.— Dryopteris firmifolia (Baker) Kuntze, Rev. Gen. Pl. 2:812. 1891.“Type: Peru [Dept. San Mar- tin], Mt. Guayapurima, prope Tarapoto, Spruce 4662 (K! isotypes BM! BR! P). Stipes + equalling or only slightly exceeding lamina length; laminae 2-pin- nate-pinnatifid at base, apex deeply pinnatifid to pinnatisect; larger pinnae 4-6 cm wide, with pinnules constricted to a narrow stalk at the base, not strongly decurrent; pinnules {at least the largest ones) shallowly to deeply pinnatifid, more deeply so along acroscopic margin, sometimes with a nearly free anterior basal segment; lamina texture chartaceous to subcoriaceous; indusia glabrous on surface, ciliate along margin. Distribution.—Known from Panama (Darién), Colombia, Ecuador, and Peru (Fig. 12); low elevations (100-700 m) in dense, primary rain forest. Discussion.—When additional and more complete collections can be made, it may prove difficult to draw even varietal boundaries between var. trianae and var. chocoense. Nevertheless, var. trianae, the more dissected of the two, has been found in Ecuador and Peru (primarily on the Amazonian side), areas where var. chocoense is not yet known. Furthermore, var. chocoense is common along the Pacific coast of Colombia, while var. trianae is known from this region only from the type. Sterile specimens of C. trianae, especially var. trianae, look remarkably like Cowen bipinnate species of Polybotrya, and as discussed in the section on fe lationships, this species may provide the closest extant link to that genus. Spang: collections: Panama (1 coll. seen). Darién (1): Along headwaters of Rio Tuquesa, © 2 3 from Continental Divide, Croat 27128 (US). Cotomaia (1). Narifio: only the type. ECUADOR (12). 539 ma Ree son de Concepcién, Playa Rica, Mexia 8452 (BM, UC). Morono-Santiago (1): eae m NNW of military camp, Brandbyge & Asanza 31930 (AAU). Pastaza (10): Neat anelos, Mexia 6894 (GH, UC, US); Finca El Valle de Muerte on Rio Curaray, ca. 10 km E ae A. R. SMITH: CYCLODIUM 93 Curaray (Jesus Pitishka), Harling & Andersson 17617 (GB); Curaray, near posto militar, Harling & Andersson 17534, 17387 (GB); Curaray, NE of Destacamento, Holm-Nielsen et al. 21985 (AAU)}, 21987 pee ob. Cyclodium trianae var. chocoense A. R. Smith, var. nov. (Fig. 13G).—Typk: Colombia, Choco, Corcovada region, upper Rio San Juan, ridge along Yeracii Valley, 200-275 m, Killip 35287 (US! isotype COL). Stipes quam lamina usque ad 2.2 plo longiorus; laminae profunde pinnato- pinnatifidae vel bipinnatae; pinnulae integrae vel crenatae, decurrentes; pinnae fertiles (2)3-4(5) cm latae; laminae subcoriaceae vel coriaceae; costae infra ple- rumque pilis numerosis hyalinis 0.1-0.2 mm longis; indusia ad marginem ciliata. Distribution.—Known only from Panama (Darién) and the Pacific coast of Colombia (Fig. 12); low elevations (0-200 m, rarely to 1450 m) in dense, primary tain forest. Only five (out of 41) collections are from above 250 m (Lellinger & de la Sota 211, 253: Duke 15522; Alverson 140, 143). Discussion.—This variety is clearly closely allied to C. inerme, and most spec- imens of var. chocoense have been so determined (as Dryopteris guianensis, a misapplied synonym) in herbaria. From C. inerme, var. chocoense differs in the generally larger fronds with wider, more deeply pinnatifid to pinnate pinnae, in the more numerous and longer hyaline hairs on the costae abaxially, in the veins often forked in the ultimate segments, and in the round-reniform indusia (with 4Narrow sinus) that are often ciliate on the margin and ti on the surface. Another close relative is C. heterodon var. abbreviatum. Cyclodium trianae is allopatric with respect to both C. heterodon and C. inerme. Additional collections: PANAMA (4 coll. seen). Darién (2): Hills of Sperdi, near Puerto Obaldia, Pittier 4413 (GH, US), 4415 (GH). Panama (1): Cerro Jefe, van der Werff & van Hardeveld 6986 (UC). San Blas (1): El Llano-Carti Road, de Nevers & Herrera 4001 (UC). Cotomesia (41). Antioquia (2): Vicinity of Planta Providencia, 28 km SW of Zaragoza, valley of Rio Anori, Alverson et al. 140 (MO), 443 (MO, NY). Cauca (2): Rio Micay, at Guayabal, Cuatrecasas 14130 (US); Timbiqui, Lehmann B.T. 433 (US). Choco (19): Rio San Juan, hill in front of Palestina, Cuatrecasas 21421 (F, US); Hydro Camp No. 8, peak over Rio Curundu, Duke 15522 (US); Rio Serrano, 4-6 km above Guayabal, Forero et al. 1342 (COL, NY); Quebrada La Sierpe, Forero & Jaramillo 4450 (COL, US); Road Andagoya- Condoto, near Andagoya, Forero et al. 5269 (COL, US); Pan American Hwy, ca. bes km W of Las imas, Gentry & Renteria 24057 (US); near jct. Rio Condoto and Rio San Juan, Killip 35084 (COL, US); Bahia Solano, near Ciudad Mutis, Killip & Garcia 33622 (COL, US); near Punta San Francisco Solano, ca. 10 km NE of Puerto Mutis, Lellinger & de la Sota 82 (COL, US); NW side of Alto del cy, Lellinger & de la Sota 211, 253 (COL, US); 1.0-1.5 km NW of El Valle, Lellinger & de la Sota 343 (COL, US); Mojarras de Tad6, 8.5 km E of Istmina, Lellinger & de la Sota 417 (COL, US); Rio Salto, 9 km W of Andagoya, Lellinger & de la Sota 478 (COL, US}; Upper Rio Truando, 5 km NE Confluence of Rio Nercua, Lellinger & de la Sota 562 (US); Upper Rio Truando, 2 km SSW A Confluence of Rio Nercua, Lellinger & de la Sota 590 (US). Narifio (2): Quebrada La Toma, os Telembi, above Barbacoas, Ewan 16861 (BM, US); E side of Gorgon casa “a or 1446 (G); Rio Naya, Puerto Merizalde, Cuatrecasas 14078 (US); Rio Yurumangui, Veneral, Cuatre- 15810 (F, US); Rio Calima, La Trojita, Cuatrecasas 16541 (F, US); Rio Cajambre, Quebrada del c Cuatrecasas 17730 (F, US), 17731 (UC); Bahia de Buenaventura, Quebesde de Aguadulce, V aSas 19981 (F, US); Rio Calima, Quebrada de la Brea, Cuatrecasas 21199 (F); Cérdoba, Dagua alley, Killip 5283 (NY, GH, US); Dagua Valley, along Rio Calima, Killip 11786 (GH, US); Agua u 94 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) Clara, along Hwy. Buenaventura-Cali, Killip & Cuatrecasas 38913 (F, US); ca. 18 km E of Buenaven- tura, Killip & Garcia 33246 (COL, US); along Rio Dagua, ca. 20 km E of Buenaventura, Killip & Garcia 33320 (COL, US); Cérdoba on Rio Dagua, Lehmann 3802 (B, G); Rio Dagua, near Buenaventura, Lehmann 4435 (US); Cérdoba, Dagua Valley, Pittier 529 (US). a) ' 10. Cyclodium varians (Fée) A. R. Smith, comb. nov. (Fig. 9, G-k).-£1 ephrodium varians Fée, Hist. Foug. Antill. (Mém. Fam. Foug. 11) 88, pl. 24, figure (5934 2. 1866.—Dryopteris varians (Fée) Kuntze, Rev. Gen. Pl. 2:814. 1891.— 2494" Stigmatopteris varians (Fée) Alston, Kew Bull. 1932:309. 1932.—The- \39@ 9 lypteris varians (Fée) Reed, Phytologia 17:323. 1968.—Cyclodium vari- Aug ans (Fée) Morton ex Vareschi, Fl. Venez. 1:368. 1969, nom. nud.TyPe: ee Trinidad, 1864, Germain s.n. (P, not seen; cited by Christensen, 1913, as seen at Paris). Rhizomes moderately long-creeping, with ca. 1 cm between stipe bases, ca. 5- 8 mm diam., at apex with dark purplish-brown (with age nigrescent), shining, conform, linear scales 4-5 mm long, 0.3-0.5 mm wide, these + entire or with a few marginal teeth, somewhat twisted at apex; fronds 45-75({105) cm long, some- what dimorphous, the sterile ones shorter and with broader pinnae than the fertile; stipes equalling or slightly shorter than laminae, stramineous to usually — tan, glabrescent, 2-5 mm diam., with persistent scales only at the very base; laminae to 50 cm long, 15-20 cm wide, with 9-17 pairs of lateral pinnae and a subconform, hastate terminal segment that tends to be more lobed and wider at base than the lateral pinnae; fertile pinnae subentire, sinuate, crenate, or shal- lowly lobed (lobes twice as wide as long), (5)10-15 cm long, 0.8-1.5 cm wide, the lowermost stalked to 5 mm, cuneate at the base, acute or acuminate at the tip; sterile or juvenile pinnae up to 12 cm long, 1.5-2.5 cm wide, often sharply serrate at the tip, truncate at the acroscopic base, cuneate basiscopically; veins anasto- mosing to form 1 or 2 series of areoles between costa and margin; costae and costules abaxially with scattered, + appressed, reddish-brown, septate, hairlike scales up to 1 mm long; yellowish or translucent glandular hairs 0.1 mm long along adaxial rachis groove and abaxially at pinna bases, some similar glands also sometimes along costae abaxially; sori in (1)2-3 rows between costa and mare, discrete or occasionally subconfluent where veins join, with a large, reddish-brown, peltate, short-ciliate indusium up to 1.3 mm diam. Distribution.—Known only from Trinidad (see Jenman, 1887) and Guyana (Fig. 8). A report from Surinam (Kramer, 1978, citing Stahel BW 7218) is based on @ misidentification of C. guianense. Vareschi (1969, p. 368) listed it as to be ex pected in Venezuela, but I have seen no collections. All collections appear have been made near sea level. eS A. R. SMITH: CYCLODIUM 95 have had an allopolyploid origin. This idea is reinforced by the unusual vari- ability in cutting of some specimens (hence the specific epithet varians), even on the same frond. Jenman (1909), Christensen (1913), and Fée in his original de- scription all commented on this variation, which can be readily seen in Parker s.n., from Guyana, and Purdie s.n., from Trinidad. In these two specimens, some pinnae are incised three-fourths or more of the distance to the costa, while other pinnae on the same frond are merely sinuate or shallowly crenate. Little is known of the habitat of this species. Jenman (1909, p. 204) described it as “as often erect in growth as horizontal.” The rather long-creeping rhizome shows little or no evidence of being buried in the soil. It seems possible that this species is a low-level hemiepiphyte, at least in some situations. Several collec- tions show the fronds directed toward the rhizome apex, as if the rhizome axis were more or less vertical against or leaning upon other vegetation. Additional collections: Guyana. Essequibo River, Moraballi Creek, near Bartica, near 0 m, Rich- ards 22 (BM, K); Mazaruni River, 1899, Jenman s.n. (K, NY, US); Macourie Creek, Essequibo River, 1895, Jenman s.n. (NY); Essequibo, 1862, Appun 27 (B, BM, P—not seen, W); Malali, Demerara River, Jenman 6853 (GH), s.n. (K, NY); Demerara, 1830, Parker s.n. (K, 2 sheets); TRINIDAD. Aripo Savanna, Britton & Britton 2941 (US); Aripo Savanna, Othmer 116 (BM, M); Aripo, Othmer 117 (M, 2 sheets); 1903, Othmer s.n. (P), Prestoe s.n. (BM); Purdie s.n. (BM); Aripo, Mar 1851, collector not legible (K); Trinidad Bot. Gard. Herb. 1225 (US). EXCLUDED NAMES Nephrodium acutum Hook., Sp. Fil. 4:147, pl. 271. 1862, non Presl (1825). =Las- treopsis acuta (Hook.) Tindale.—Synrypes: Brazil, Sellow s.n. (K); Peru, near Tarapoto, Spruce 4662 (K). Implicitly lectotypified by Baker, Syn. Fil., ed. 2. 501. 1874, when he selected Spruce 4662 (K) as type of Ne- phrodium firmifolium Baker [=Cyclodium trianae (Mett.) A. R. Smith]. Hooker's description of Nephrodium acutum is in agreement with the Sellow collection, while his illustration (pl. 271) is of the Spruce collec- tion. Dryopteris clypeata Maxon & Morton, Bull. Torrey Bot. Club 66:52. 1939. = Thelypteris clypeata (Maxon & Morton) Kramer, Acta Bot. Neerl. 18: 141. 1969.—Stigmatopteris clypeata (Maxon & Morton) Lellinger, Proc. Biol. Soc. Washington 89:730. 1977.—T PE: Panama, hills back of Puerto Obaldia, San Blas Coast, 50-200 m, Pittier 4309 (US, 2 sheets). This is a member of subg. Goniopteris and is closely allied to Thelypteris hol- odictya Kramer from Surinam. 2 Aspidium extensum Fée, Gen. Fil. 294. 1852, non Blume (1828). = Ctenitis mela- nosticta (Kunze) Copel—Thought by Christensen (Ind. Fil. 73. 1905) to be the same as Dryopteris trianae [=Cyclodium trianae (Mett.) A. R. Smith]. Nephrodium abe Hook., Sp. Fil. 4:86, pl. 242A. 1862 = Stigmatopteris rotundata (Willd.) C. Chr.—The specimen (Spruce 2153) cited as this species by Fée (Aspidium imrayanum (Hook.} Fée, Crypt. Vasc. Brés. 1:133. 1869) is the type collection of Polypodium sancti-gabrieli Hook. 96 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) [= Cyclodium guianense (Klotzsch) A. R. Smith]. Nephrodium imray- anum is based on a collection from Dominica, Imray s.n. (K). LITERATURE CITED CHRISTENSEN, C. 1905-1906. Index filicum. Copenhagen: Hagerup. 1909. On Stigmatopteris, a new genus of ferns with a review of its species. Bot. Tidsskr. 29:291- 304 ———. 1913. 4 monograph of the genus Dryopteris. Part I. The tropical American pinnatifid- gonna species. Kongel. Danske Vidensk. Selsk. Skr., Naturvidensk. Afd. (7)10:55-282. monograph of the genus Dryopteris. Part II. The tropical American bipinnate- ———. 1938. Filicinae. Pp. 522-550 in Manual of pteridology, ed. F. Verdoorn. The ff. Nijho: CopELAND, E. B. 1947. =— seem genera oF | rie —— sae: eure Botanica. ea wena a reel g r eiden Bot. Series Hourruss, £ me 1984. Studies in the fern-genera sg to Tectaria, I. A commentary on recent ————, and J. G. BAKER. 1867. Synopsis filicum. cme Robert Hardwicke. JENMAN, G. S. 1887. The ferns of Trinidad. J. Bot. 188 ————. 1909. The ferns and fern allies of the Oa a a nelies and Guiana. [issued in parts, pt. 12, beginning p. 203, oe in Feb 1908]. Ed. J. H. Hart. Trinidad: Government Printing Office, Port-of-S Kramer, K. U. 1978. The hae dite of Suriname. Uitgaven Natuurw. Studiekring Suriname Ned. Antillen 93:1-198 — T. 1857-1862. Index F ‘ibian: London: Williams & Norgat Morton, C. V. 1939. On the genus Cyclodium. Bull. Torrey Bot. Club 66:47-52. Pag subun R. E.G. 1977. Tentamen pteridophytorum genera in taxonomicum ordinem redi- gendi. Webbia 31:313-512. 982. Third report of the Committee for Pteridophyta. Taxon 31: 315- 316. Proctor, G. R. 1985. Ferns of Jamaica. London: British Museum (Natural History). Siva Araujo, I. J. 1976. Polypodiaceae. P. 483 in IOPB chromosome volte reports LIII, ed. A. Love. Taxon 25:483-500. TrYON, R. M. and A. F. TRYON. 1981. (58 1) Proposal to conserve Stigmatopteris C. Chr. 1909 against Cyclodium Presl, 1836, and Peltochlaena Fée, 1852 (Pteridophyta). Taxon 30:349-350. - 1962. Ferns and allied plants, with special reference to tropical America. New York: Springer-Verlag. WALKER, T. G. 1966. A cytotaxonomic . ‘ Soc : survey of the pterido es of Jamaica. Trans. Roy. 90¢- Edinburgh 66:169-237. oat la Evidence from cytology in the classification of ferns. Pp. 91-110 in The phylogeny classification of the ferns, ed. A. C. Jermy et al. J. Linn. Soc., Bot. 67, Suppl. 1 APPENDIX: List OF ExSICCATA The following is a complete list of all numbered collecti ee listed ed assigned taxa in the taxonomic treatment. For mix pet Re eomeniersay Type collections are indicated by a boldface T; they are not Ep rwioacge e type of the accepted n erson et al. 140 (9b); 143 (9b). Appun 27 (10); 166 eg 1176 (6c). Austin Argent et al. 6455 (6a). et al. 7214 (6a). Aymard & Stergios 358 (6a). Aymard et 954 (5); 955 (5); 4196 (6c). 325 (6a). Benoist 715 (6a). Berlin 949 (6a). asta 2957 (5); 7245 (5); 7283 (6a). Bierhorst nT agitate el etal eee eg A. R. SMITH: CYCLODIUM 97 — (6a). Billiet & Jadin 1677 (5); 1923 (5). Bischler 1446 (9b); 2461 (2); 2592 (2); 2628 (3); Black 54- 6942 (6a). Blanchet 2208 (4b); 2251 (4b); 2468 (4b); 2486 (4b); 2519 (6a). Boom & Mori 1574 (3); 1894 ma 1990 (3). Boom et al. 762 (4b). Brade 15360 (6a). Brad eee 18130 (4a). Brandbyge & Asanza 31930 (9a). Britton & Britton 2941 (10). Broadway 5758 (6a); 6 6a). oS 1641 (3). Conant 914 (6a); 1007 (6a); 1008 (6a); on a ). Cow Soderstrom 2157 (5 Cremers 5389 (5); 5390 (5); 5567 (6a); 5696 (5); 5697 (5); 5766 (5); 6084 (6a); poe (3); 6389 (3); 6559 pa 6608 ag 7019 (6a); 7370 (3); 7411 (5); 7724 (5); 7828 (5). Croat 18383 (6a); 18464 (6a); ~~ (6a); 18710 (6a); 18771 (6a); 20311 (6a); 20967 (6a); 27128 (9a). Croizat 782 (3). Cuarto 154 (4a). Cuatrecasas 6996 (6a); 10563 (6a); 14078 (9b); 14130 (9b); 15810 (9b); 16541 (9b); 17730 (9b); 17731 (9b); ads pont 21199 (9b); 21421 (9b Daniéls & Jonker 744 (3); 852 (3); 1259 (6a); 1300 (6a). Davidse & Gonzdlez 14294 (6a); 21735 (3). Davis 180 (3). Dawson 14893 (6a). de Granville C.123 (3); C.155 (5); 239 (5); 560 (5); 777 (5); 886 (5); 1445 (6a); 1753 (5); 1761 (6a); 1860 (3); 2123 (3); 2580 (5); 2586 (7-T); 2626 (5); 2752 (3); 2894 (6a); 3078 (6a); 3093 (3); 3288 (6a); 3530 (5); 4149 (3); 4730 (6a); 5243 (5); 5301 (3); 5302 (5); 5481 (6a); 5888 (5); 6455 (5). de la Cruz 1174 (6a); 1177 (6a); 1921 (6a); 3119 (6a); 3488 (6a); 3844 (6a); 3847 (6a); 4302 (6a). Delascio et al. 9604 (5). de Nevers & Herrera 4001 (9b). de Santos et al. R.1486 (6a). Duarte 6756 ards 1136 (5). Eiten et al. 2125 (6a). Ellenberg 2803 (6a); 2868 (6a); 2870 (6a). Engel 231 (3). Erskine 204 (6a). Every & Burgess 52 (6a); 258 (6a). Ewan 16861 (9b). Garefa-Barriga 14336 (6a); 14347 (6a). Gardner 1218 (4b/4 x 62); 2759 (6a); 5671 (4a). Geijskes 1012 (3). Gentry & Renteria 24057 (9b); 24421 (9b). Gentry et al. 15824 (6a). Glaziou 5392 (4a); 7322 (6a); 10180 (6a); 12373 (5); 12374 (6c-T). Gleason 183 (6a). Glocker 10 (6a). Gonggrijp & Stahel 601 (5); 974 [= BW974] (6a). Gottsberger 34-11067 (6a). Guppy 6095 (6a); 6135 (5); 6195 (5); 6212 (5); 6213 (5); 6249 (6a); 6250 (6a); 7418 (6a). ed 17173 (4b). Harling & Andersson 17387 (9a); 17534 (9a); 17617 (9a). Hart 6791 (6a). Hassler 83 (6a); 9049 (6a); 10131 (6a); 10131a (6a); 11625 (6a). Haught 1353 (2-T). Hertel et al. 15280 (6a). osha 17124 (6a). Holm-Nielsen et al. 21985 (9a); 21987 (9a); 22029 (9a); 22061 (9a). Holst & van der Werff 2573 (6a). Holt & Blake 602 (3). Hombersley 132 (6a). Hostmann 15 (5-T). Huber 753 (5). Irwin & Soderstrom 5739 (6a); 5881 (6a). Irwin & Westra 47168 (6a). Irwin et al. 10139 (6a); 16303 (6a); 54523 (5); 54923 (3). Jaramillo & Coello 3648 (9a). Jenman 1481 (5); 1419 (5); 1600 (6a); 4213 (5); 6850 (6a); 6853 (10). Killip 5283 (9b); 11786 (9b); 35084 (9b); 35287 (9b-T). Killip & Cuatrecasas 38913 (9b). Killip & Garcia 33172 (9b); 33246 (9b); 33320 (9b); 33622 (9b). Killip & Smith 26165 (6a); 26566 (6a); 28540 (9a); 30316 (6a); 30331 (6a); 30335 (5). Klug 241 (6a); 396 (6a). Kramer 2364 (5). Kramer & Hekking 2401 (6a); 3075 abe (5). sna et oe 3254 (6a). Krukoff 4924 (6a). 4 ee re oe Se ouw & Li 7 (5); 251 Jouw & Lindeman 543 (5); 2364 (5); 2389 (6a); 2457 (5); aby: 258 {Ob}; 343 (9b} 417 (9b) (5); 131 (6a); 188 ‘ e: . 124 (6a). Lleras et al. P17441 6a). i 119 (4b). — or Meyer 6219 (6a); 6280 (6a). Monsalve 429 (0b) . Moron losén 3941 (6a). Moss 26 (6a). Muiller & r % oo. ae Sa). llgaard et al. 34511 (6a); 34528 (9a); 94742 (6a); 35360 (6a). Othmer 116 (10} 10). 98 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) Pennell 3723 (6b-T). Persaud 110 (1). Philcox et al. 3710 (6a); 3911 (6a). Pinas 8354 [= LBB 8354] (6a). Pittier 529 (9b); 4413 (9b); 4415 (9b). ease 2452 (6a). Plowman et al. 9532 (5). Poole 1822 (5); 1922 (3); 1930 (3); 1958 (3). Potter 5281 (6a). Prance & Pennington 1825 (3). Prance et al. 9452 (3); ~ ofl nes 19947 (3); 22686 (3); 59224 (6a). Prestoe ae (3). Pulle 386 (6a). os & Sousa 157 (6a). Ratter et al. 1022 (6a); R2134 (6a). Richards 15 (6a); 22 (10); 799 (6a); 6591 ve pes (6a). Richards et al. R504 (6a). Riedel 43 (4b); 47 (4b); 82 (6a). Ruiz-Terdn & Lopez-Palacios 11473 (1). Rusby & Squires ote 9 379 (6a Sandwith 1473 (5). Sastre 3017 (6a). Bh ociurek 25 (3/5); 135 . 316 (6a); 1157 (3-T); 16174 (6a). Schubert 2224 (5). Schultes ote (3); 5661 (3). Schultes & Lopez 9427 (3). Schultes & Cabrera 13474 (6a); 13928 (3); 16647 (6a); 17354 (6a); 19329 (3). Schultes & rab Boer a ia 10175 (3). Schulz 8043 (5); 8695 [= LBB 8695] (6a); 9943 [= LBB 9943] (6a). Schunke 5610 (6a). Schwacke 200 (6a). Smith, A. C. 2745 (3); 2896 (5). Soejarto & Cardozo 718 (6a); 719 (6a); 740 (6a). Solomon 3565 (6a); 3573 (6a); 6186 (6a). Spruce 25 (6a); 26 (6a); 36 (5-T); 379a (6a); 379b (6a); 2153 (3-T); 2159 (3); 2775 (6a); 2776 (6a); 4662 (9a-T); 4689 (6a). Stahel 463 [= BW 7218] (3); 496 [= BW 7202] (6a). Stahel & Gonggrijp 601 [BW pooidiggsiank BW 2642]. Steyermark 57958 (3); 58077 (3); 87654 (3); 88601 (5); 89104 (5); 89529 (5); 90656 (3). Steyermark & Bunting 102582 (3). Steyermark & Nilsson 9 (1); 124 (1). Steyermark et al. 92644 (1); 122169 (5); 130318 (5). Thomas & Plowman 3157 20 Tillett et al. 44942 (5). Tresling B496 (5). Triana 32 (9a-T). Tryon & Tryon 5205.5 (6a). Tutin 420 (6a Ulbricht 121 (6a). Ule 9128 ie van der Werff 604 (6a). van der Werff & Gonzdlez 5207 (6a). van der Werff & van Hardeveld 6986 (9b). van Donselaar 3740 (6a); Vareschi 6904 (6a); Vargas [Univ. Cuzco 11476] (6a). Versteeg 308 (6a). Warrington et al. K.E.R.76 (1-T); K.E.R.167 (6c); K.E.R.1207 (1). Williams, L. 11381 (3). Williams, R. S. 1191 (6a); 1202 (6a). Wullschlaegel 1710 (3). Wurdack & Monachino 41294 (6a). Sy SHORTER NOTE Cystopteris tennesseensis in West Virginia.—Cystopteris tennesseensis Shaver (Tennessee Bladder-fern) is a fertile allopolyploid that apparently originated from an ancestral cross between C. bulbifera (L.) Bernh. and C. protrusa (Weath.)} Blasdell. The relationships of C. tennesseensis to other Cystopteris species have recently been described by Haufler (Proc. Roy. Soc. Edinb. 86B:315-323, 1985). Tennessee Bladder-fern is widely distributed over the eastern and central United States from Maryland, Virginia, North Carolina, and Georgia, west to Kansas and Oklahoma (Blasdell, Mem. Torrey Bot. Club 21:1-102, 1963; Cranfill, Ferns and fern allies of Kentucky, 1980; Moran, Amer. Fern J. 72:93-95, 1982). While C. tenneseensis seems to be rare or local in the easternmost part of its range, it is difficult to assess accurately its total geographic distribution and frequency. Many regional manuals do not include this fern or mention it only peripherally. Tennessee Bladder-fern also is under-reported since it may grow in mixed pop- ulations with C. bulbifera or the Cliff Fragile-fern, Cystopteris tenuis (Michx.) esv. This latter species has long been known as C. fragilis (L.) Bernh. var. mackayi Lawson (Moran, Castanea 48:224-229, 1983). Collectors may fail to sur- vey the entire population and thus collect only one of the species present. The problem is further compounded by hybridization between species (Moran, 1983). Consequently, C. tennesseensis often is overlooked or misidentified. Tennessee Bladder-fern has not previously been reported from West Virginia, though Strausbaugh and Core (Flora of West Virginia, p. 20, 1978) suggested that it likely would be found in the state. Cystopteris tennesseensis has been collected tom five counties in West Virginia. The specimens examined are listed below. West Vircinia: Brooke Co.: Shaley talus, Cross Creek, S of McKinleyville, 17 Oct 1983, Cusick 23209 (NCU, WVA). Grant Co.: Greenland Gap Rd, E of Scherr, 30 July 1978, Rader 2091 & White (WVA). Pleasants Co.: Near St. Marys, 8 May 1937, Brooks & Margolin s.n. (WVA); 1 mi S of Eureka, 1 June 1945, Bartholomew s.n. (WVA); sandstone, Middle Island Creek, E of St. Marys, 6 Oct 1983, Cusick 23180 & Ortt (MICH). Putnam Co.: Bridge Creek Rd., 10 June 1983, Denison s.n. (WVA). Co.: Sandstone cliff, Otter Creek, base of Green Mountain, 12 Oct 1957, Rossbach 1170 (WVA). West Virginia is a logical part of the total range of C. tennesseensis. Among adjacent states this species is frequent only in Kentucky and Ohio. The only known Virginian population is in Montgomery County (Moran, Jeffersonia 13: 30-34, 1982). The species has been collected once in Frederick County, Maryland agner, Amer. Fern J. 34:125-127, 1944). The North Carolina populations of Tennessee Bladder-fern represent the extreme southeastern limit of its known distribution. These sites, in Craven and Jones counties, are the only known occurrences on the Atlantic Coastal Plain (Blasdell, 1963). Reports of C. tennes- seensis from Pennsylvania (Wherry et al., Atlas of the flora of Pennsyly en 10, 1979) need confirmation; they may be based in part upon C. laurentiana (Cc. ifera x fragilis). Tennessee Bladder-fern should be sought in southcentral and southwest Pennsylvania and in western Maryland, as well as in additional West Virginian counties. Cystopteris tennesseensis is easily mistaken for its putative parents. Most of 100 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 2 (1986) the West Virginian specimens examined had been misidentified as C. bulbifera. Moran (1982) reported similar difficulties in identification of C. tennesseensis in Illinois. The first two Pleasants County collections cited above are the basis for the Strausbaugh and Core report of C. bulbifera in that county. These two species grow in similar rocky habitats. They seldom are found on stream terraces or in deep soils, the typical habitat of C. protrusa. Cystopteris tennesseensis often grows in more xeric situations than C. bulbifera. Although C. tennesseensis is most frequent and abundant on limestone in most of its range, in Ohio and West Virginia it is found variously on limestone, dolomite, shale, and sandstone. Sub- strate is recorded for only three of the eight West Virginia collections cited above. Tennessee Bladder-fern apparently is rare or local in West Virginia. However, more thorough collecting is needed to determine accurately its status in the flora. I thank Dr. Roy B. Clarkson and Linda Rader, Department of Biology, West Virginia University, for their courtesy and cooperation. Robbin Moran, Illinois Natural History Survey, Champaign, kindly shared with me his thoughts on Cystopteris. Marilyn Ortt, Ohio Department of Natural Resources, provided valuable field assistance.—ALLISON W. Cusick, Division of Natural Areas & Pre- serves, Ohio Department of Natural Resources, Fountain Square, Columbus, OH 43224. INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable}, including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 812 by 11 inch Het of good quality, not “erasable” paper. Double space m t, including title, authors’ names and addresses, text (including Sante and keys), i dbire cited, tables (separate from text), and figure captions (grouped as Consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author’s name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. Use S.I. (metric) units for all mea- __ sures (e.g., distance, elevation, weight) unless quoted or cited from another source (e.g, specimen citations). For nomenclatural matter {i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). read : breviate titles of serial peiggenrs according to Botanico-Periodicum-Huntian: um (Lawrence, G. H, al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as ae of nomenclatural matter | are not masa in lit- erature cited. For shorter notes and reviews, put all theti . text. Use Index herbariorum (Regnum Veg. 106: 1-452. 1981) for designations herbaria. — should be proportioned to fit page width with caption on the same Page. Provide margins of at oo 25 mm on all illus os a illustrations, design originals fo t ith : amount. In composite blocks, af ie edges ¢ o of adjacent cum “Avoid | com= iu bining continuous-tone and line- ocks. Sequence and numb f figures (and sith rer ef cation oe a scales and symbols in figures themselves, not in . scale and reference to latitude and longitude in each. map. ___ Proofs and reprint order forms are sent to authors by the printer. a. Authors : : : = a pe olga corrected proofs to the editor and reprint orders to the printer. oe aes mater of form ¢ or se, oe nsu u sites be assessed changes for extensive ae: patic — ae : 7 Z AMERICAN FERN nae July-September 1986 JOURNAL QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY _ The Use of Fern Herbarium Specimens in Biosystematic Research: Introduction Michael D. Windham Active Enzymes +. , 3 pee 8 + Sp s rl pk eae” an Aftertt ght z Thomas A. Ranker and Charles R. Werth Biosystematic Uses of Fern G * Pane Ne i¢£ De Re oe 2. nt "Michael D. Windham and Christopher H. Hauler Detecting Abortive Sp Soe Bas teens Oe £ Ceavitle Uvhrid WH Cae awa aw Carl Taylor Factors Affecting Prolonged Spore Viability in Herbarium Collections of Three Species of Pellaea Michael D. Windham, Paul G. Wolf, and Thomas A. Ranker cf 'y tic Inf; a 3 Os. es. pe Se. ees ‘Davi 5, Barvigton, Cathy A Pars ond Thomas A. Ranker 4a The American Fern Society Council for 1986 FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. President angen E. SKOG, Biology Dept., George Mason University, Fairfax, — 22030. Vice-President W. CARL TAYLOR, Milwaukee Public Museum, Milwaukee, WI 532 Secretary JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, a TN 37916. Treasurer DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Burlington, VT 05405. Records Treasurer JAMES D. eoakigenguages: Ecology III, R.D. 1, Berwick, PA 1860 Back Issues Curator ALAN R. SMITH, Dept. of Botany, University of California, Svea CA 94720. Journal Editor DAVID B. ssesigacsindeaeaiste neste Institution, Washington, DC 20560. Memoir Editor DENNIS Wm. STEVENSON, Dept. of Biological Sciences, Barnard Cillees. Columbia t University, New York, NY 10027 Fiddlehead Forum Editor iasicen Fern Journal EDITOR DLA me ee ee a S. Dept. of Botany, University of California, Berkeley, CA 94720 ASSOCIATE EDITORS “GERALD |.GASTONY ......-. Dept. of Biology, Indiana University, Bloomington, IN 47401 Seer AUPE Dept. of Botany, University of | : Lawrence, KS sated DAVID B. PERINGER _. U.S. Nat'l Herbarium NHB-166, oo Institution : n, DC 20560 _ TERRY R. WEBSTER _... Biological Sciences Group, University of Connecticut, ee CT 06268 The “American Fern Journal” (ISSN 0002-8444) is an illustrated quarterly devoted to the general _ Study of ferns. It is owned by the American Fern Society, and published at the Pringle Herbarium, : University of Vermont, Bertingion, VT 05405. Second-class postage paid at Burlington, VT, and entry point. aes Pe . eo aL ( } Sate 42 months (foreign) after the date of issue, addressed to Dr. James D. Montgomery, Ecology III, RD. 1, coi Cee decry arals \ wriae for membership should be sent to the Records Trea- surer. ct ss Comins d to the Secretary. Siberipions $12.00 gow, $11.50 net if paid through an agency {agency fee $0.50); sent free to : members of the American Fern ee ond _US.A- Canada, and Mexico; life membership, $200 ™ 1 " f GA pages or less, $1.25; 65-80 = Pago : : : mus ick or pee 50 ch sing Back volumes 1979 et seq. $8.00 each; oe Soe ek ee es, See Wie Storing Ten percent discount on orders of six yolumper _MASTE ti tne ac om Jounsat, Dept. of Botany, University of American Fern Journal 76(3):101 (1986) The Use of Fern Herbarium Specimens in Biosystematic Research: Introduction MICHAEL D. WINDHAM Department of Botany, University of Kansas, Lawrence, Kansas 66045 Dried specimens have been an indispensable part of botanical science since the first half of the sixteenth century (Morton, 1981). This is especially true for descriptive botany and classical taxonomy, which provide the foundation for modern studies of plant evolution. In recent years, the focus of evolutionary investigations has shifted from the relatively stable end products (species) to the dynamic process itself. This has led to the belief that the most accurate recon- struction of relationships is achieved through an amalgamation of morphological, genetic, biochemical, physiological, and ecological data. As a result, classical morphological taxonomy is being supplanted by biosystematics, an integrated approach that emphasizes the use of live plants (Solbrig, 1970). As the popularity of biosystematics has increased, the perceived importance of herbarium specimens has declined. Lagging interest in preserved materials has, in turn, led to a restructuring of budgetary priorities within the biological sciences. Many herbaria show signs of serious financial neglect, including cur- tailment of hours (e.g., CAN; see Anon., 1985), reduced availability of loans (eg. UT; see Holmgren et al., 1981), and lower rates of specimen accession. This is an alarming trend that, if continued, will eventually undercut public support for botanical research. Admittedly, there are many facets to the budgetary crisis in botany, but the shift from classical taxonomy to biosystematics should not be a contributing factor. As will be shown in the following articles, most of the tech- niques commonly used in biosystematic research can be applied successfully to herbarium specimens. Many of these methods are non-destructive (i.e., accept- able to curators), and they can yield valuble information on relationships, breed- ing systems, hybridization, and other aspects of evolution. In the papers that follow, we begin to explore the potential of herbarium collections for biosyste- matic research among pteridophytes. LITERATURE CITED . AA f NJ 1 Sci As § st. ANONYMoUs, 1985, Major cutbacks at Canadian National S80C. SY Coll. Newsl. 13:44. -Houcren, P., W. Keuken, and E. SCHOFIELD. 1981. Index herbariorum. Part I. 7th ed. Regnum Veg. 106:1-452. : Morton, A. 1981. History of botanical science. London: Academic Press. ane Ps _ Sotaric, 0. 1970. Principles and methods of plant biosystematics. Toronto: Macmillan WeSSOUR! BOTANICAL NOV 6 1986 American Fern Journal 76(3):102-113 (1986) Active Enzymes from Herbarium Specimens: Electrophoresis as an Afterthought THoMAsS A. RANKER and CHARLES R. WERTH Department of Botany, University of Kansas, Lawrence, Kansas 66045 Electrophoretic analyses of enzyme variation in plants routinely employ fresh leaf tissue. Extracts of such material are analyzed immediately or the leaves or extracts are preserved through freezing or lyophilization. The possibility that active enzymes may be obtained from dried pressed specimens for electropho- retic analyses has not been considered by population biologists or systematists. In the present study, however, we demonstrate that, in certain cases, electro- phoretic investigations of pteridophytes can be extended to include dried spo- rophytic material. This expansion of the utility of pressed specimens from her- baria greatly enhances the potential of such material to contribute to biosystematic research. Numerous studies on the effects of dehydration on the anatomy, biochemistry, and physiology of desiccation-tolerant and desiccation-intolerant plant species have been reported (e.g., see reviews by Ijin, 1957; Todd, 1972; Crowe & Clegg, 1978; Bewley, 1979; Bewley & Krochko, 1982). These effects are complex, vary among species, and often depend upon the particular conditions of dehydration and rehydration, even within a taxon. In general, however, the studies have shown that desiccation-tolerant plants are capable of quickly restoring photo- synthesis and respiration upon rehydration and, under certain conditions, des- iccation-intolerant species may also resume these functions. Thus, it has been established that some dehydrated plants can preserve enzymes that are capable of continuing physiological activity upon rehydration. The present study was initiated following the observation that electrophoretic expression of enzymes had been obtained from leaf tissue of Asplenium platy- neuron plants that inadvertently had been allowed to desiccate in plastic bags at 4°C (Figs. 1, 2). This led to a more extensive survey of dried pressed material of varying ages and from a variety of taxa. MATERIALS AND METHODS Small quantities of leaf or rhizome material (5-15 mg) were removed from pressed specimens (see Table 1 for location of vouchers) and ground in the phosphate crushing buffer of Soltis et al. (1983). The grinding was done in the wells of porcelain spot plates using the rounded ends of small test tubes a8 pestles. The exudate was absorbed onto rectangular wicks of Whatman 3 MM chromatography paper. The wicks were inserted immediately into 12.8% starch gels for electrophoresis. zymes were surveyed using buffer systems 6, 8, and 11 as outlined by Soltis et al. (1983; modified in Haufler, 1985) and the continuous tris-citrate buffer (gel RANKER & WERTH: ENZYMES 103 a oy “a _— a as ie ee Pics. 1-7, Photographs of gels. Anode is at the top of each figure. Fics. 1, 2. Electrophoretic banding patterns of Asplenium platyneuron sporophytes. Bands at and to the right of the arrows are from desiccated leaves; all others are from fresh leaves. Fic. 1. PGI. Fic. 2. PGM. Fics. 3-7. Comparisons of electrophoretic banding patterns from fresh and dry sporophytic material, arranged in pairs of dry followed by fresh of same individual, except for last pair (m-n) which is fresh-dry. Lanes in all five figures correspond to those labeled in Fig. 5. Lane a is marked for reference in all figures. a-b: Dryopteris carthusiana (dry material 0.2 yrs old); c-d, e-f: Woodsia oregana (0.5); g-h, i-j, k-l: W. mexicana (0.5); m-n: W. obtusa (0.5). Fic. 3. PGM. Fic. 4. PGI. Fic. 5. TPI. Fic. 6. LAP. Fic. 7. 6-PGD. PH 8.0) of Selander et al. (1971). Esterase (EST), amino aspartate transaminase (AAT), and phosphoglucomutase (PGM) were assayed on system 6; hexokinase (HK) and leucine aminopeptidase (LAP) on system 8; acid phosphatase (AcPH) on system 11 or the tris-citrate; isocitrate dehydrogenase (IDH), malate dehydro- genase (MDH), 6-phosphogluconate dehydrogenase (6-PGD), and shikimate de- hydrogenase (SkDH) on system 11; phosphoglucose isomerase (PGI) and triose- osphate isomerase (TPI) on systems 6 and 8. Standard staining schedules were employed (e.g., see Soltis et al., 1983; Werth, 1985). RESULTS AND DISCUSSION The results are summarized in Table 1. They show that while active enzymes Were obtained from a variety of pressed materials, there was considerable vari- ation in the quality of the results according to the species, age of material, and enzyme system assayed. Most specimens exhibited some activity for at least 104 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) several enzymes. In a number of cases, however, discrete bands could not be resolved. In such cases the enzyme molecules may have been so degraded that attempts to obtain greater resolution by the use of different electrophoretic con- ditions may not be productive. Thus, the most promising cases for further in- vestigation are those for which discrete bands could be identified. Many such cases were observed (Table 1) and some will be discussed in greater detail below. In many of the examples, bands could be identified that were not clearly re- solved or which seemed to be accompanied by more bands than would be ex- pected at a locus, given a knowledge of enzyme structure (Gottlieb, 1981) and intracellular compartmentalization of isozymes (Gastony & Darrow, 1983; Soltis, 1986). Extra bands may be due to post-translational modification, inappropriate treatment of material causing the production of artifacts (Harris & Hopkinson, 1976), or changes in molecular conformation related to desiccation (Todd, 1972). These “ghost” bands sometimes may be eliminated or reduced by changes in experimental technique such as the use of different gel and electrode buffers or of a different grinding buffer (Harris & Hopkinson, 1976; Soltis et al., 1980). Banding patterns from dried specimens that are not completely resolvable may still be of value if they can be interpreted when directly compared to results from living plants of the same species (see below). The ability to preserve enzyme integrity for relatively short periods does not seem to be related to the living plant’s ability to tolerate desiccation. With one exception (Bommeria pedata), all specimens less than one year old showed resolvable enzymatic activity at most or all loci assayed. These included species of Dryopteris (Figs. 3-7), Hemionitis, Lycopodium, Polypodium, and Woodsia. The best overall results among these taxa were obtained from the Woodsia specimens (Figs. 3-7). These species grow in mesic habitats and have delicate, easily dehydrated leaves, yet clearly resolved isozyme bands were obtained from six-month-old pressed leaves. Isozymic expression can often be obtained from pressed material that is com- parable to that from live plants. Because the Woodsia leaves were collected from plants under cultivation, we had the opportunity to compare isozymic patterns from dry material directly with those of fresh leaves of the same individuals. Similarly, for three species of Pellaea, comparisons could be made to live g@- metophytes grown from viable spores obtained from the same herbarium spec- imens (see Windham & Haufler, 1986; Windham et al., 1986). In this way W® could check for qualitative and quantitative changes in electrophoretic expres sion upon dehydration. It can be seen from Figs. 3-7 that for most enzymes there were only slight decreases in the quality of the bands from the dried material and that the quantity of bands remained the same. One notable exception to the latter in Woodsia was for LAP where all of the dried specimens expressed an extra band not found in living plants (Fig. 6). Results comparable to those from live gametophytes were obtained in dried leaves of two-year-old P. ternifolia and eight-year-old P. wrightiana for LAP, TPI, 6-PGD, SkDH, IDH, and HK. Partial or complete loss of expression was observed in PGI, MDH, and PGM. Dried 23-year-old P. truncata showed resolvable bands only for TPI; however a three-year-old specimen of this species had clear bands at most loci examine RANKER & WERTH: ENZYMES 105 These results are similar to those obtained from wilted wheat leaves (Todd, 1972) where some isozymes showed only qualitative changes upon wilting while others showed both qualitative and quantitative shifts in expression. These changes apparently are due, directly or indirectly, to the effects of desiccation (Todd, 1972; Bewley & Krochko, 1982). These results have two important impli- cations: 1) some enzyme systems may not be scored reliably using pressed spec- imens, depending upon the purpose of the study; and, 2) results from pressed material should be compared to those from live material of the same species for accurate assessment of banding patterns. The ability to preserve enzyme activity beyond about a year may be related to the desiccation tolerance of a species and/or to the particular conditions of drying or storage. For example, a two-year-old specimen of Polypodium poly- podioides (Table 1) produced resolvable bands at all loci examined from both rhizomes and leaves. These results are not surprising given the recurrence of dehydrating conditions in the life history of this epiphyte. Stuart (1968) showed that leaves of this species could lose up to 97% of their water content and resume normal photosynthesis and respiration upon rehydration. In contrast, though, a conspecific one-year-old specimen (Table 1) exhibited clear bands for all loci from rhizome extracts but for only three of eight enzyme systems from the leaves. Both of our specimens were collected as green, apparently viable, individuals. Thus, there may have been differences in the conditions of drying (see below) or storage between these specimens that affected enzyme activity. More data are needed to assess accurately the effects of these variables. In this study particular attention was given to the ability of Selaginella species to preserve active enzymes since many species of this genus are known for their extreme tolerance of desiccation. An initial assay indicated that specimens of xerophytic species were more likely to be of value electrophoretically than were those of mesophytes (e.g., compare S. apoda, a mesophytic species, with other taxa, all xerophytes; Table 1); hence, subsequent analyses only included ts cerhal sentatives of the former. The oldest specimen of Selaginella to provide evidence of enzymatic activity was one of S. wallacei, collected 64 years ago. Specimens of other species up to about 25 years old provided resolved banding patterns for most loci. Several species were sampled across decades (most thoroughly in S. densa and S. rupestris) over which a general increase in activity and resolution was observed with a decrease in age (Figs. 8-13). Previous studies have reported on the ability 0 desiccated Selaginella speci- mens to resume growth and various physiological functions. Tryon (1955) re- Ported that a six-month-old dry specimen of S. densa var. densa resumed gro upon rehydration. Webster and Steeves (1964) found that a mat of S. densa stored ty for two years and nine months showed signs of growth when watered for several weeks. Eickmeier and coworkers (Eickmeier, 1979, 1980, 1983; ra et al., 1982) have done extensive work on photosynthetic recovery of desiccat a S. lepidophylla. They have shown that individuals that have been desiccat for at least two years may fully recover photosynthetic and wean! pune Upon rehydration. Uphof (1920) studied the anatomy of several xerophytic pole " Selaginella and described a number of anatomical features that may con 106 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) Taste 1. Summary of Electrophoretic Analyses of Dried Plant Specimens. of ma- pee, Enzyme system Taxon years Source’ AAT ° AcPH EST Adiantum concinnum 1 Ranker 786 +++ Asplenium b yi 4 Magrath et al. 12156 teas Asplenium bradleyi 17 Matthews et al. s.n. Sa A. monant 4 Ranker 7 ++ A. montanum 5 Werth biting Umass A. montanum 13 Lafferty 13 +4+4+ A. pinnatifidum 4 Magrath et a 12147 ++ . pinnatifidum 31 McGregor 37 ae 2 A. platyneuron 2 McGregor ee +++ . platyne 17 Johnson 1306 ma 8 A. rhizophyllum 5 Brooks & Harris 15155 +++ A. rhizophy 5 Werth 81819 +++ Aspidotis de. 1.5 Ranker ss al. 756 At Bommeria pedata* 0.5 Ranker et al. 801 7 > sm Cheilanthes i 2.5 Ranker as Simmons 794 er C. eatonii 38 Weber 3. C. eatonii 24 Wilson ed a C. eatonii 15 Stephens & ‘eacks 43000 a . wootonii 18 Brooks 493 ez Cystopteris diaphana* 1 Ranker 780 - ++? ris carth 0.2 Windham et al. 688 D. i 0.5 Werth KF bs a Hemionitis elegans* 0.5 Ranker et al. ++ - et ‘ 1 Ranker et al. 727 ae Lycopedium dendroideum 0.5 Ranker 838 “e L. lucidulum 0. Ranker 837 if L. reflexum 1 Ranker et al. 769 a = a L. sp. 1 Ranker et al. 783 a = zi L.. sabinifolium 0.5 Ranker et al. 840-2 = + 7% Notholaena formosa** 1.5 Windham et al. 539 N. limitanea** 1.5 Windham et al. 575 N. pallens** 1.5 Windham et al. 527 Pellaea atropurpurea 39 s.n. - P, 9 Brooks & Hauser 12238 e P. mucro 28 an 1281 = P. ternifolia** 2 Weber & Bajakiewicz rh? 16507 (COLO) lf? P. truncata** 23. Martin 5146 (UNM) rh one If P. wr ightiona* . 19 —_- Bozeman et al. 45152 Ed wrightiana’ 8 _—_ Reeves et al. 6340 (ASU) rh lf Pityrogramma tartarea 1 Ranker 770 - = 2 P. triangularis 1.5 Ranker & Wolf 754 ++7 Polypodium adelphum 1 Ranker 815 - +++ = . qureum* 2 Ranker et al. 717 rh oe If 2 californicum* 5 Ranker & Artus 96 = Te a ses 2 Ranker et al. 720 rh : If P. glycyrrhiza 5 Ranker & Artus 203 & 267. rh +++ o ) fo ee RANKER & WERTH: ENZYMES ie TABLE 1. Extended. Enzyme system HK IDH LAP MDH 6-PGD PGI PGM SkDH TPI ca + i aa ++ + +e ++ + aah . e ; ab . x R 3 ;: e : 7 + - + ~ + _ + + . ae ++ +++ + ~ + - + _ pti cielo + — S +++ +4 Pere ae + a a ++ — ina a +++ ++ +++ 7 - +++ +++ a z - - - + - - ++ ae se - + + + - pe ++ = + ~ + - - att + ++ ++ + ++ ~ + +++ att + ++ +++ ++ ++ a ++ +++ +4 . ] . + o - ae ++ +t ++ ++ +++ + adea ++ . + ++ - ++ > ++ ++ +++ - ++ S aa _ tee one + +40 +e +t ++ va ++ + ++ - = a? 2 + ++ z . oe AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) 108 TABLE 1. Continued. Age of ma- terial Enzyme system in Taxon years Source’ AAT AcPH EST P. polypodioides 2 Ranker et al. 723 + P. polypodioides 1 Ranker et al. 761 - P. virginianum* 0.5 Ranker et al. 839 = aa. a3 Pteridium oT 1 Ranker et al. 758 it P. 0.5 Ranker 816 _ att Selaginella apoda 70 Seymour 32 = me S. apoda 36 — — oe a s. apoda 14 Thom x f 3 ones : he 81-347 = s S. arizonica 18 Keil & Lehto a = S. bigelovii 28 Hartman ek Re 5 S. densa 31 Taylor 7394 a S. densa 2. OE as > S. densa 17 ce ss Brooks 23262 aes = S. densa 12 Stephens 66679 Bad S. densa 4 Rohde : _ 2961 4+ S. densa 3 Warner 1 Te = S. lepidophylla 19 Wilson is nae S. lepidophylla 2 Villarreal et al. 2270 aie S. pallescens a Ranker et al - a S. rupestris 53 mpson s.n ” oa S. rupestris 39 McGregor 966 a a S. rupestris 29 Lathrop 1993 + is S. rupestris 20 arms 2759 ia < S. rupestris 18 Brooks 455 AS a S. rupestris 9 Churchill 7643 qs Ee s. ris 5 H & § 5332 be es S. siberica 10 Taylor 19764 7 ae S. tortipila 35 Sargent s.n. sad S. tortipila 15 _Pittillo et al. 3164 id S. underwoodii 17 Stephens & Brooks 21869 age S. underwoodii 10 Brooks & Hauser 10737 aged S. underwoodii a Brooks 17210 +* S. underwoodii*** 1 Brooks 17185 aie S. wallacei 64 i - a S. weatherbiana 23 ee S. weatherbian 15 Stephens & Brooks 42983 ahs Woodsia ilvensis 0.5 Ranker et al. 832 ae W. mexic 0.5 Windham & Czech 591 W. mexicana** 0.5 Windham & Haufler W. mexicana** 0.5 Windham & ers 781 W. oregana** 0.5 Windham oregana** 0.55 Windham & ng 652 eee Explanation f y ‘4 ntry eS ie ity det ect ed; +, activity sual ncaa oe Ee yed, a ,no Sena ae ity ecballe ds; ++ mccoy isozymes or artifacts present; +++, . resolved * Pressed specimen compared ent individual. ** Pressed specimen compared to living material of same individual. o resolved, p' bands. to living material (sporophytes or gametophytes) of same species: RANKER & WERTH: ENZYMES ue TABLE 1. Extended. Enzyme system HK IDH LAP MDH _ 6-PGD PGI PGM __ SkDH TPI +++ ++ +++ ++ ++ +++ ++ ++ +++ +++ ++ +++ ++ ++ ++ ++ +++ +++ ++ +++ ++ ++ +++ ++ ++ +++ +++ ~ +++ _ a es a Me ned +++ +++ ++ ++ +++ - “+ e = 3 a 4 fe + + ue a ag 2+ +++ ++ ae +++ +++ . as Boe: + ++ ++ ++ > +++ + + - _ + +++ + + - + +++ ++ ++ +++ ++ ++ +++ +++ ++ +++ aoe ++ - ++ +++ of +++ ++ ++ ~ ++ +++ + Be ng ae +++ + +++ +++ ++ ++ + _ — ++ +++ ++ ae a+ ++ - ++ +++ Aas ++ ++ ++ ++ +++ - if = - + + - ++ - + a - + - ++ S) <0 ++ ++ ++ - +++ ie acts > ++ ++ +++ +++ - ++ +++ +++ ae et ote+ i et: +++ +++ ++ +++ +++ . + +++ ++ ++ +++ +++ + + + +++ - ieee 2 ++ + “ +++ +++ +++ +++ ++ — +++ +++ * ++ +++ ++ “= +++ +++ xe e+ +++ ++ _ ++ +++ ++ +++ +++ ++ _ +++ +++ c - ~ + oe + - > ++ ++ + = +++ +++ + oe de ++ ++ = +++ +++ + ++ ++ ++ +++ ott ++ tea ee a et ee + +++ +++ +++ +++ +++ +++ is ++ ++ +++ +++ +4++ +++ +++ (eae ++ aa ++ nha ++ +++ +++ +++ +++ +++ +++ - em, originally labeled S. weatherbiana but electrophoretically identified as S. under- Woodii. : Voucher specimens deposited at KANU, unless otherwise noted. th = rhizome extract; lf = leaf extract. 110 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) : 2 gt ae bs i 3 ott Mia A 4 ew Fics. 8-13. Photographs of gels. Anode is at the top of each figure. Lane a is marked for reference in all figures. The same individuals are represented in Figs. 8-11. Species (age of specimen in years) are: a, b, c: Selaginella densa (23, 12, 4, respectively); d, e: S. weatherbiana (23, 15, respectively); f, g, h, i: S. underwoodii (1, 17, 10, 1, respectively). Fic. 8. TPI. Fic. 9. LAP. Fic. 10. EST. Fic. 11. MDH. Figs. 12 and 13 are of the same individuals. Species (age) are: a-e: S. rupestris (29, 20, 18, 9, 5 respectively); f: S. wallacei (64); g: S. bigelovii (28); h: S. arizonica (18); i: S. densa (17); j: S. siberica (10); k: S. densa (3). Fic. 12. TPI. Fic. 13. MDH. ute to the ability of these taxa to withstand desiccation. While the potential of the specimens we studied to resume in vivo physiological function was not assessed, our results show that the extreme desiccation tolerance of these species allows, at least, for the consistent preservation of some enzymes for more than 25 years. The observation that identical isozyme bands were expressed in con- specific individuals sampled at approximately 10-year intervals over 20- to 40- year periods, suggests that no changes in molecular confirmation occurred at most loci. Thus, while comparisons with live material should be done in any study that incorporates dried specimens, this may be less crucial with species of this genus. : The ability to recover well-resolved isozymic patterns from herbarium specl- mens also provides an opportunity to use electrophoretic markers to check the classification of specimens of questionable identity. This may be useful partic- ularly in the taxonomically difficult genus Selaginella due to the cryptic nature of most of the characters used in distinguishing species (Tryon, 1955). An eX ample of this was provided inadvertently in the present study. One of the one- year-old herbarium specimens of S. underwoodii (Table 1) actually was labeled S. weatherbiana. Electrophoretically, however, it was found to be quite distinct from the other specimens of the latter species and shared most of its allozymes with S. underwoodii (Figs. 8-11, lane f). Reinspection of the specimen reveal that, in fact, it had been misidentified. Larger sample sizes are needed to deter RANKER & WERTH: ENZYMES 111 mine whether the genomic markers are absolute (i.e., mutually exclusive) but these data show that the use of electrophoresis may be of value in identifying morphologically confusing individuals. CONCLUSIONS There are two situations where the inclusion of dried pressed specimens in electrophoretic investigations appears most promising: 1) when recently collect- ed material is available (i.e., six months or less) and 2) when studying certain xerically adapted taxa. The practical implications of the first situation are many. For example, if it is difficult, not desirable, or impossible to collect live whole plants, leaves may be returned to the laboratory in a plant press for subsequent electrophoresis. The results from Woodsia and Pellaea suggest that this may be possible for a wide variety of taxa. Certainly, press-drying is not the method of choice in preserving leaves to be used for electrophoretic studies, but it does provide a viable and, perhaps, heretofore unappreciated alternative. If leaves are to be preserved by pressing, the method of drying should be considered since it is likely that enzyme activity and integrity will be preserved better under ambient temperature drying than under drying with heat (see Todd, 1972; Eick- meier, 1979). For example, the leaves of Woodsia and Pellaea were press-dried under ambient temperatures whereas those of Bommeria pedata were dried with the aid of a portable electric plant dryer (compare these taxa in Table 1). While the effects of microwave drying on enzymatic activity were not investi- gated, Bacci et al. (1985) showed a significant loss of integrity of the internal leaf anatomy in several angiosperm species when so treated. Since such destruction of the cellular environment might disrupt enzyme form and function, a study of the effects of microwave drying on enzymatic activity is of interest. In the second situation, where the taxa under study are xerophytes, electro- Phoretic sampling could be conducted over a much broader geographic a than that which could be visited by an investigator in a limited period. The taxonomic scope of an investigation also could be expanded to include taxa not collected for various reasons. Herbarium specimens would be indispensable in this regard (see Windham & Haufler, 1986, for a further discussion). Since such small quantities of material are necessary (i.e., 5-15 mg}, the damage a ope ed specimens is negligible and, in our view, far outweighed by the information gained. We suggest, however, that removal of material from a borrowed herbar- ium specimen should be done only with the explicit permission of the appro- priate curator. : In taxa that exhibit particularly strong preservative qualities (e.g., Selaginella), there is the possibility of adding a temporal element to populational studies. For example, the use of herbarium specimens could provide investigators with the °pportunity to evaluate the dynamics of the genetic structure of . taxon gre given region through time. Such data could potentially be used to — . Power of studies on mating systems, genetic me mens effects, gene flow, an migration, among others (see Windham & Haufler, 1986). The results of ihe present study are preliminary but suggest that pressed spec- 112 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) imens of various groups of pteridophytes may be used to expand the taxonomic and geographic scope of biosystematic studies. Other fern biologists are strongly encouraged to explore further this potential in their particular groups of interest. We are grateful to Ralph Brooks of KANU and to the curators of ASU, COLO, and UNM for permission to remove small quantities of leaf material and spores from pressed specimens. Genie Trapp and Chris Haufler kindly provided helpful comments on the manuscript. We thank Dr. Gerald J. Gastony and an anonymous reviewer for their valuable suggestions, Craig Martin for helping with literature references, and Alan Smith for checking the identity of several specimens. LITERATURE CITED Bacci, M., A. CHeccucct, G. CHEccucci, and M. R. PALANDRI. 1985. Microwave drying of herbarium imens. Taxon 34:649-653. BercTRoM, G., M. SCHALLER, and W. G. EICKMEIER. 1982. Ultrastructural and biochemical bases of mee in the drought-tolerant vascular plant, Selaginella lepidophylla. J. Ultrastruct. 78:269-282. BEWLEY, : D 1979. Physiological aspects of desiccation tolerance. Annual Rev. Pl. Physiol. 30:195- tee J. E. KrocHKo. 1982. Desiccation-Tolerance. Pp. 325-378 in Encyclopedia of plant physiology, New Series Volume 12B, Physiological plant ecology II, eds. O. L. Lange et al. Berlin-Heidelberg: Springer-Verlag Crowe, J. H. and J. S. CLecc. 1978. Dry biclowind! systems. New York: Academic Pre EICKMEIER, W. G. 1979. Photosynthetic recovery in the resurrection plant Selaginella pee after wetting. Oecologia 30:93-106 . 1980. Photosynthetic recovery af resurrection spikemosses from different hydration re- gimes. Oecologia 46:380-385 ————. 1983. Photosynthetic recovery of the resurrection plant Selaginella lepidophylla aes ov Grev.} Spring: effects of prior desiccation rate and mechanisms of desiccation damag Oecologia 58:115-120. Gastony, G. J. and D. C. Darrow. 1983. Chloroplastic and _— isozymes of the homosporous fern Athyrium filix-femina L. Amer. J. Bot. 70:1409-1 Gortuizs, L. D. 1981. Electrophoretic evidence and plant sh in te Prog. Phytochem. 7:1-46. Harris, H. and D. A. Hopkinson. 1976. Handbook a. enzyme electrophoresis in human genetics. Amsterdam: North-Holland Publishing Co Haur er, C. H. 1985. Enzyme variability ies of evolution in the fern genus Bommeria. Syst. Bot. 10:92-104. Iyjin, W. 3 1957. Drought resistance in plants and physiological Annual Rev. PI. Physiol. 7-274. Ls * Re R. K., M. H. Smrru, S. Y. YANG, W. E. JOHNSON, AND J. B. Gentry. 1971. IV. Biochemical Spins Set Ey errenatice in Te genus Peromyscus. I. Variation in the old- field mouse (Peromyscus polionotus). Studies in genetics VI. Univ. Texas Publ. 7103:49-90. Sottis, D. E. 1986. Genetic evidence for diploidy in Equisetum. Amer. J. Bot. 73:908-913. Paap eg G. J. Gastony. 1980. Detecting enzyme variation in the fern genus Bommeria: an analysis of methodology. Syst. Bot. 5:30-38. ,D. C. Darrow, and G. J. Gasrony. 1983. Starch gel electrophoresis of ferns: compilation of grinding buffers, gel and electrode buffers, and staining schedules. Amer Fern. J Stuart, T. 8. 1968. Revival of oe and photosynthesis in dried leaves of Polypodium poly- podioides. Planta 83:185-206 FOND, Ge po 1972. Water deficits and enzymatic activity. Water Deficits and Plant Growth 3: 177- Acelsain R Mt Jr. 1955. Selaginella rupestris and its allies. Ann. Missouri Bot. Gard. 42:1-99- RANKER & WERTH: ENZYMES 113 Upuor, J. C. TH. 1920. Physiological anatomy of xerophytic Selaginellas. New Phytol. 19:101-131. Wesster, T. R. and T. A. eaarere 1964. Observations on drought resistance in Selaginella densa ydb. Amer. Fern J. 54:1 WertH, C. R. 1985. implementing an isozyme laboratory at a field station. Virginia J. Sci. 36:53-76. WINDHAM, M. D. and C. H. HAUFLER. 1986. Biosystematic uses of fern gametophytes derived from h 128. , P. G. WotF, and T. A. RANKER. 1986. Factors affecting prolonged spore viability in her- barium collections of three species of Pellaea. Amer. Fern J. 76:141- American Fern Journal 76(3):114-128 (1986) Biosystematic Uses of Fern Gametophytes Derived from Herbarium Specimens MICHAEL D. WINDHAM and CHRISTOPHER H. HAUFLER Department of Botany, University of Kansas, Lawrence, Kansas 66045 Ferns and fern allies differ from other vascular land plants in having a com- pletely independent gametophytic generation. Although relatively small and eas- ily overlooked, fern gametophytes have their own set of morphological features, growth conditions, and habitat preferences. Among the ferns, it is the gameto- phytes that are responsible for both sexual reproduction and initial habitat se- lection, and knowledge of their physiological tolerances and breeding system characteristics is indispensible when trying to understand patterns of dispersal migration, and speciation. Thus, although physically dwarfed by the sporophytes, fern gametophytes play a key role in the evolutionary process and any “biosys- tematic” study should include analyses of the gametophytic generation. Unfortunately, fern gametophytes are seldomly collected and, even when they are, many of the characteristics that make them biologically important cannot be preserved. Therefore, pteridophyte biosystematists must study living cultures of gametophytes. Such cultures are generally initiated by collecting fresh spo- rophytes, harvesting their spores, and sowing these spores on soil or agar sub- strates. This approach has yielded a wealth of data, but it has also reinforced the perception that viable spores can only be obtained from recently collected materials. This, in turn, has discouraged the use of fern gametophytes in bio- systematic studies since it is difficult to obtain representative samples of widely — species and living plants of many narrow endemics may be unavail- able. Although fern biologists interested primarily in sporophyte morphology and anatomy usually depend upon herbarium specimens to circumvent these sam- pling problems, it was thought that this option was not available to those working with gametophytes. However, recent studies indicate that the spores from her- barium collections of a variety of fern species remain viable for decades. Johnson (1985) reported that the spores of certain Marsilea species were capable of get- mination after 99-100 years, and Windham et al. (1986) germinated 50-year-old spores of Pellaea truncata. Spore longevity of this magnitude may actually be oe are ferns, especially in taxa occupying periodically xeric habitats. Pray (in Lloyd & Klekowski, 1970) reported that most cheilanthoid fern spores germinated after 10 years, and spores from a specimen of Notholaena standley! at KANU (Brooks 489) remained viable after 19 years in the herbarium (Wind- ham, unpubl. data). Although spores of some mesic species such as Cystopteris protrusa and Woodsia obtusa lose their ability to germinate after a few yeals: prolonged spore viability has been reported in several ferns inhabiting moist environments. Spores of the tropical species Asplenium serra germinated after 48 years (Fischer, 1911), and those of Polypodium glycyrrhiza remain viable for at least five years (T. Ranker, pers. comm.). WINDHAM & HAUFLER: GAMETOPHYTES 115 Aside from the genetic constitution of the plant, the most important factor controlling spore longevity in herbarim collections appears to be the type of treatment used to reduce insect infestation (Windham et al., 1986). Although spore viability declines rapidly in specimens fumigated with methyl bromide, Dowfume or heat, exposure to PDB or low temperatures (— 20°C) seems to have little adverse effect. To date, most surveys of spore longevity have focused on materials from larger herbaria using the more toxic fumigants. Thus, there is reason to believe that prolonged spore viability is more prevalent among pteri- dophytes than the literature would indicate. This raises the possibility that spores from pressed specimens might be used for large scale analyses of gametophytes, which would improve both the scope and quality of biosystematic investigations in ferns. The present paper explains how herbarium specimens can be employed to circumvent sampling limitations in biosystematic studies by providing a ready source of gametophytic material. Examples from ongoing research using herbar- ium-derived gametophytes demonstrate ways in which these plants may improve and expand the data base of evolutionary and phylogenetic investigations. Al- though there are many possible uses for such gametophytes (Miller, 1968; Dyer, 1979), the present discussion focuses on morphological, chromosomal, and ge- netic evidence that can be obtained from pressed plants. The techniques de- scribed herein have proven especially valuable in Pellaea and other closely related genera, but it is apparent that they can be useful even in species where spore viability does not exceed five years. MATERIALS AND METHODS Spores were sown and gametophyte cultures maintained using the methods outlined by Windham et al. (1986). Samples showing low germination percent- ages (<1%) were excluded from the analyses to minimize the possibility of contamination. Morphological features were routinely studied through a Wild issecting microscope, though a Zeiss phase contrast microscope was used for detailed observations of gametophytes cleared with a saturated solution of chlo- tal hydrate. For chromosomal investigations, gametophytes with a well-devel- oped meristem (6-12 weeks old) were submerged in a 0.2% colchicine solution for 24 hours prior to being placed in Farmer's fixative (3 parts ethanol : 1 part glacial acetic acid). After 24 hours in the fixative, gametophytes were rinsed in distilled water and hydrolized in full-strength Glusulase (snail cytase} for three hours. Quite fragile at this point, they were then carefully rinsed in two changes of distilled water and stored in fresh fixative at 20°C. Chromosome slides were Prepared by excising the meristematic regions of several cor een a aera ating them in 1% iron acetocarmine, and squashing in a 1:1 mixture 0 ee Carmine and Hoyer’s medium. Cells containing chromosomes cones bs ne staphed using the equipment and materials described by Haufler ned al. { : | Enzymes for electrophoretic analyses were extracted by crus = g eae “A phytes in a plastic weigh boat with a glass rod after adding 2-5 drops of the Phosphate grinding buffer-polyvinylpyrrolidone solution of Soltis et al. (1983). 116 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) Gametophytes were ground individually for progeny analyses of meiotic segre- gation or crushed in groups of 8-10 to determine the isozyme phenotype of the pressed parental plant. The modified gel-electrode system number 11 described by Haufler (1985a) was used to run isocitrate dehydrogenase (IDH)}, malate de- hydrogenase (MDH), phosphoglucomutase (PGM), and shikimate dehydrogenase (SkDH). The enzymes aspartate aminotransferase (ATT), hexokinase (HK), leu- cine aminopeptidase (LAP), phosphoglucoisomerase (PGI), and triosephosphate isomerase (TPI) were run using Haufler’s (1985a) modification of gel-electrode system number 8. The standard staining recipes of Soltis et al. (1983) were fol- lowed for AAT, LAP, MDH, and PGM. Agarose overlay stains (Soltis et al., 1983) were used to assay for IDH, HK, PGI, SkDH, and TPI. Photographs of stained gels were taken with a Pentax 35-mm camera using a red filter and Kodak Technical Pan film. MORPHOLOGY Although classifications and phylogenetic hypotheses in ferns have tradition- ally focused on attributes of the sporophyte, it is clear that morphological evi- dence from the gametophytic generation can make a major contribution to our understanding of pteridophyte evolution (Stokey, 1951; Nayar & Kaur, 1971; At- kinson, 1973). However, a much wider sampling of taxa will be necessary to understand evolutionary trends in the haploid generation. Herbarium collections form a logical starting point for surveys of gametophyte morphology since they provide an extensive sample of both geographic and taxonomic diversity. Two examples drawn from studies of cheilanthoid ferns illustrate the potential value of this approach. During research on the dark-stiped members of Pellaea section Pellaea, Pray (1968a) germinated spores from 10-15-year-old herbarium specimens in order to augment gametophyte samples derived from freshly collected sporophytes. By this means, he was able to obtain living materials of all species in the group and produce one of the best descriptive studies of cheilanthoid gametophytes avail- able. Investigations of the light-stiped members of section Pellaea could benefit similarly from analyses of prothalli derived from spores of herbarium specimens. For example, Pray (1968b) observed unusual growth patterns in a collection of P. andromedaefolia from Coronado Island and suggested that insular populations might produce gametophytes that are morphologically distinct from those of mainland populations. Unfortunately, he was unable to obtain recently collected materials from other island localities, and most of California’s Channel Islands are now off limits to botanical collecting. However, his hypothesis could still be tested by germinating spores from pressed plants collected when insular pop¥ lations were more accessible. Ongoing studies of the genus Notholaena provide another indication of the value of herbarium specimens for analyses of gametophyte morphology. As not ed by Tryon and Tryon (1982), the prothalli of several species of N otholaena exhibit glandular trichomes that produce a farinose exudate similar to that ob- served on the sporophytes. This character could be of great value in clarifying WINDHAM & HAUFLER: GAMETOPHYTES 117 evolutionary relationships within this taxonomically contentious group of chei- lanthoid ferns. However, a general survey of Notholaena based solely upon fresh material is not feasible given the highly restricted distribution of some species (Tryon, 1956). An investigation of this group using a combination of herbarium specimens and live plants seems most appropriate and, to date, more than half of the species in the genus have been examined. Glandular gameto- phytes appear to be characteristic of all species included in the section Notho- laena by Tryon and Tryon (1982). On the other hand, farinose trichomes are absent from the prothalli of species in section Argyrochosma, providing further evidence that this species cluster represents a separate evolutionary line. These and other examples (Atkinson, 1973; Nayar & Kaur, 1971) amply dem- onstrate the value of the independent data set that can be obtained from the study of gametophyte morphology. Aside from the natural populations them- selves, herbarium collections provide the most extensive sample of fern diversity available. The successful culture of gametophytes derived from the spores of herbarium specimens obviates the perceived requirement for fresh material and prothalli obtained in this manner can (and should) figure prominently in future surveys of morphological variation. CHROMOSOMAL STUDIES As indicated by Walker (1973), ferns have been the subject of a tremendous amount of cytogenetic research during the past thirty years and, as a result, our understanding of phylogenetic relationships has greatly improved. Many of the species for which chromosomal data are currently lacking occupy relatively in- accessible regions of the world, and gametophytes derived from herbarium col- lections may be of value in these situations. For example, the first chromosome counts for Bommeria knoblochii, a taxon endemic to the rugged mountains of Chihuahua, Mexico, were obtained from the gametophytic progeny of plants collected as herbarium specimens (Gastony & Haufler, 1976). This approach may Prove especially valuable in the genus N otholaena where endemism is well- developed and few species have been investigated cytogenetically (Tryon & Tryon, 1982). Although it will be useful to ascertain the chromosome numbers of all fern species, other important goals can be achieved using gametophytes derived from herbarium specimens. Recent cytogenetic studies have revealed an increasing number of cytotypes, cryptic species, and undetected hybrids (eg. Smith & Foster, 1984; Walker, 1985), and further investigations along these lines are likely to have a significant impact on studies of fern evolution. The assumption that a single chromosome count is characteristic of each morphologically defined taxon has proven false in many cases (e.g., Wagner et al., 1965; Tryon, 1968; Haufler Necessary. As with most organisms, fern cytotypes appear ' Netic distinctness through ecological or geographic separation. Thus, a Specimens collected throughout the known range of a taxon are especially we Suited to an investigation of intraspecific chromosomal variability. 118 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) Since there is a positive correlation between chromosome number and spore size in most ferns (Barrington et al., 1986), cytogenetic work on herbarium spec- imens with unusual spore measurements may be the most efficient way to detect new cytotypes. It was this approach that led to the discovery of a diploid form of Woodsia obtusa (Windham, unpubl. data), a species that Brown (1964) consid- ered an ancient autotetraploid derived from W. oregana. This example supports Walker’s (1985) assertion that correlations between spore size and ploidy level must be based upon a sufficient number of intraspecific chromosome counts. Brown (1964) did not detect cytotypes during a limited cytogenetic survey of Woodsia, but his spore measurements were derived from a large sample of herbarium specimens that was almost certain to include chromosomal variabil- ity. In this case, the failure to restrict spore measurements to cytological vouchers led to the exceptional (and erroneous) conclusion that spore length and ploidy level were not correlated in Woodsia. Brown’s approach to what was then a cytologically unknown group is not unusual, and many cytologists do not mea- sure spores from the same plant that provided the chromosome count. This reflects the fact that sporophytic materials collected for meiotic chromosome studies rarely exhibit mature spores. However, given the amount of intraspecific chromosomal variability detected in recent years, it is clear that spore measure- ments included in correlative studies must be derived from the actual plant that yielded the count. This relationship is most easily maintained by doing chro- mosome counts from herbarium specimens where the mature spores themselves provide the gametophytic material used to determine chromosome number. Although different species often show parallel correlations between spore size and ploidy level, interspecific comparisons must be approached cautiously (Steb- bins, 1950). A disjunct population of Pellaea breweri from southern Nevada exhibited unusually large spores averaging 54.3 um in length. However, a chro- mosome count of n = 29 (Fig. 1) based on gametophytes from a 10-year-old herbarium specimen agreed with the only previous report for the species (Tryon & Britton, 1958). In this case, spore size comparisons with closely related species would falsely indicate that the southern Nevada population of P. breweri was polyploid, concealing the non-chromosomal nature of spore length variation wi in this taxon. Since the species in many fern genera are poorly defined and difficult to identify, it might be wise to develop separate correlations for each major geographic region currently included in the distribution of a single species. Herbarium collections, which represent the summation of distributional knowl- edge, would be an especially valuable resource in this regard. Once correlations between spore size and chromosome number are firmly established in a species complex, it becomes possible to identify cytologically unknown collections (Bar- rington et al., 1986) and accurately map cytotype distributions using pr specimens (e.g., Tryon, 1968; Smith, 1975). Basic analyses of this sort provide the foundation for additional studies with far-reaching evolutionary implications. One of the most neglected aspects of fern cytogenetics is the subject of karyo- type analysis. As recently as 1985, fern cytologists were still trying to develop 4 standardized system of data presentation that would facilitate comparative stu¢- ies of fern karyotypes (Walker, 1985). However, the potential rewards of this WINDHAM & HAUFLER: GAMETOPHYTES 119 1- : ~ * + > - % - = s 4 * . *y * . on : : * ‘aS ~X . ord a og a 8 5! Sete Sew Ay ee oO gi a. J a a #* am? a4 x Fa gl si Pm pr Fics. 1, 2. Chromosomal squashes obtained from meristematic cells of gametophytes originating from the spores of herbarium specimens. Fic. 1. Pellaea breweri (Fisher 1786, UNLV), 1000. Note that a single haploid chromosome set (n = 29) is present although the spores from this collection are unusually large. Fic. 2. Pellaea truncata (Nash, McGill, Hensel, Lehto, and Pinkava 10158, ASU}, x 2000. Thi re Py abe Peet + s Se ean oD og Ky : information on ferns. Chromosomes are highly condensed, sister chromatids are evident, and chro- mosomal morphology is variable. approach easily overshadow the difficulties posed by terminology, technique, and high chromosome base numbers. For the first time, Walker (1985) has dem- onstrated concrete differences in chromosome morphology between closely re- lated fern species. Although Walker’s analyses were based entirely upon root tip preparations, there are advantages to using gametophytic materials for karyo- types. Prothallial meristematic cells contain half the number of chromosomes EVOLUTIONARY GENETICS 1985b) illustrated how electrophoretic analyses of : In a recent review, Haufler ( ' our understanding of pteridophyte 120 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) tissue. For this reason, herbarium specimens have seldom been considered an appropriate source of material for studies of plant isozymes. However, the dis- covery that some fern spores remain capable of germination for several decades (Windham et al., 1986) suggests that pressed specimens may provide an important data base for future electrophoretic investigations. Studies of geographic variability.—The power of the electrophoretic tech- nique is directly dependent upon the quality of the available sample. This is especially clear when attempting to characterize a species isozymically. Biosys- tematic analyses of mixed populations frequently identify potential genomic markers, but these cannot be construed as mutually exclusive until a represen- tative sample of all closely related species has been examined. Growing and studying gametophytes derived from herbarium specimens may be an efficient means to conduct the necessary geographic surveys of enzyme variability. This approach is currently being used to investigate genetic variability and phylogenetic relationships among three members of the Pellaea truncata com- plex in which long-term spore viability has already been documented (Windham et al., 1986). Pellaea truncata (syn. P. longimucronata) is a diploid species con- fined to the southwestern United States and adjacent Mexico (Tryon, 1957). Pellaea ternifolia ranges from southern Chile to southern Arizona and includes both diploid and tetraploid cytotypes (Tryon, 1968). Although currently separated by at least 100 km (Windham, unpubl. data), populations of P. truncata and diploid P. ternifolia apparently hybridized in the past to produce the allotetra- ploid species P. wrightiana (Wagner, 1965). Ongoing electrophoretic studies of these species indicate that freshly collected and herbarium-derived materials provide complementary data sets that can be very useful when investigating isozyme variability. For example, the majority of sporophytes in P. truncata are homozygous for a single isozyme at the cathodal locus of PGI (PGI-2). However, electrophoretic analyses of multiple gametophytes from herbarium specimens collected throughout the range of the species occasionally yielded faster (Fig. 6F) and slower (Fig. 6S) bands. Apparently, these isozymes have localized distribu- tions, and they were rare or absent at the localities where living population samples were collected. Some populations of the allotetraploid species P. wrigh- tiana show a fixed heterozygotic pattern at the PGI-2 locus which includes a band identical to the faster isozyme seen in P. truncata (Figs. 3B, C, D). Since similar bands have not been detected in the P. ternifolia parent, it seems likely that this rare P. truncata genotype was involved in at least one of the hybridiza- tion events that initiated P. wrightiana. Thus, a geographic survey of P. truncata herbarium specimens has improved our understanding of evolutionary processes in the entire species complex. Once isozyme markers have been established on the basis of extensive elec- trophoretic sampling, they can be used to identify specimens in which critical morphological features are atypical, easily misinterpreted, or simply lacking. This eS lead to the recognition of cryptic species and interspecific hybrids, both of which provide valuable information concerning phylogenetic relationships. Even supposedly “sterile” interspecific crosses can be studied in this manner because most hybrid ferns produce a small percentage of spores capable of WINDHAM & HAUFLER: GAMETOPHYTES 121 ope sa Fics. 3-8. Zymograms of gametophytes grown from spores of herbarium dividual gametophytes from different sporophytes of Pellaea wrightiana. 3 = PGI, 4=S DH, 5= DH. a = Windham & Haufler 639, KANU. b-e = Windham & Haufler 629, KANU. The variant, and e originated eS ance specimens. Fics. 3-5. In- k independently. Sporophyte d cannot be distinguished from c. Fic. 6. PGI ett were produced by separate gametophytes. Fic. 7. from a “sterile,” triploid, interspecific hybrid between Pellaea trun & Windham 434, ASC). Note that the electromorphs are monomorphic an component alleles of this fixed-heterozygote are not equal. Fic. 8. IDH. A set of individual gameto- of Pellaea truncata (Cottam 14452, HUH) showing segregation & truncata (Craig s.n., UNM). c = fixed heterozygosity among the gametophytes of the allotetra- Ploid P. wrightiana (Clark s.n., UNM). These plants are also notable for the remarkable longevity 9 6 Sh 9 S 1 of their spores. “a” was collected in 1956, “b” in 1936, and “c with an unbalanced combination can often be identified solely by (Haufler et al., 1985; Gastony, germination (DeBenedictus, 1969). Individuals : genomes (such as allopolyploid backcrosses) pana in the relative dosages of isozymes Among temperate species of Pellaea, hybridization appears to be most com- 122 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) mon between the diploid P. truncata and the tetraploid P. wrightiana (Windham, unpubl. data). Pray (1971) reported signficant spore germination in one of these hybrids and, during the present study, spores were sown from several herbarium specimens that were morphologically intermediate between P. truncata and P. wrightiana. Although most of the spores from these plants were malformed and failed to germinate, the larger ‘basketball’ spores did produce prothalli. Elec- trophoretic progeny analyses using individual gametophytes from one of these collections revealed a strong dosage effect in IDH (Fig. 7). The cathodal and heterodimeric bands were much darker than the anodal band, suggesting that two of the three genomes in the parental sporophyte coded for the cathodal isozyme. This cathodal band is found in both P. truncata and P. wrightiana, though in the latter species it is always combined with another isozyme to form a fixed heterozygotic pattern (Fig. 8C). The plant in question probably arose through cross fertilization between gametophytes with the IDH genotypes shown in Figures 8B and 8C. Thus, electrophoretic analyses of gametophytes derived from this herbarium specimen support its identification as a triploid hybrid be- tween P. truncata and P. wrightiana. At least 43% of homosporous ferns are polyploid (Vida, 1976), and electro- phoretic analyses can provide a great deal of information concerning the origin and subsequent evolution of these species. Polyploid taxa normally exhibit fixed heterozygotic banding patterns for a number of enzyme loci (Werth et al., 1985a) and, in most cases, the gametophytes obtained from them have the same complex patterns as the parental sporophyte and are genetically identical to one another. Thus, electrophoretic surveys of enzymes can be used to estimate the ploidy level of gametophytes from herbarium specimens without recourse to chromo- somal techniques. For example, progeny arrays from heterozygous sporophytes of the diploid species Pellaea truncata show segregation of isozyme bands in IDH (Fig. 8A). Gametophytes from the allotetraploid P. wrightiana, on the other hand, exhibit fixed heterozygotic banding patterns because the prothalli them- selves are diploid and each contains two divergent genomes (Fig. 8C). The ob- servation of fixed heterozygosity in the gametophytic progeny of a hybrid be- tween P. truncata and P. wrightiana (Fig. 7) also identifies this plant as a polyploid. The absence of isozymic variability among these gametophyt ts that most viable spores are the result of non-reductive sporogenesis. Once a polyploid is identified, isozyme evidence can help to determine wheth- er the taxon originated through interspecific hybridization (allopolyploidy) or intraspecific chromosome doubling (autopolyploidy). An allopolyploid should contain genomic markers from each species involved in its hybrid origin (Werth et al., 1985a). The detection of a polyploid containing the markers of a single species provides circumstantial evidence that the plant or population arose through autopolyploidy (Haufler et al., 1985). In addition, analyses of gameto- phytes can be used to distingiush disomic from tetrasomic inheritance and there- by identify potential autopolyploids. Since autopolyploids are often morpholog- ically indistinguishable from their diploid ancestors, electrophoretic analyses of her barium collections may facilitate their recognition and thus allow biosyste- matists to assess the evolutionary importance of autopolyploidy in the ferns. a WINDHAM & HAUFLER: GAMETOPHYTES 123 Recent studies involving the Appalachian Asplenium complex have demon- strated remarkable congruity between enzyme data and morphologically based hypotheses of relationship (Werth et al., 1985a, 1985b). Each diploid species contained a unique set of isozymic markers, and most hybrid individuals exhib- ited anticipated patterns of electrophoretic additivity. However, these studies also revealed unexpected evidence of “orphan” alleles, selective gene silencing, and multiple origins among the allopolyploid members of the complex. “Or- phan” alleles are those isozymes found in an allopolyploid species which have not been identified in any of the putative diploid progenitors. On the other hand, the absence of one or more ancestral markers from a known allopolyploid is often viewed as evidence of gene silencing. Orphan alleles and apparent gene silencing could both result from incomplete sampling of enzyme variability in the diploid progenitors. The former may represent rare isozymes whose distri- bution was restricted to a few diploid plants at the locality where the original allopolyploid was formed. The faster migrating PGI-2 isozyme found in some plants of Pellaea wrightiana (Figs. 3B, C, D) might have been classified as an orphan allele if its presence at low frequency in P. truncata had not been revealed by an electrophoretic survey of gametophytes from herbarium speci- mens (Fig. 6F). “Silenced” genes may be interpreted as extremely common iso- zymes which are not found in some members of the hybridizing diploid popu- lation and, therefore, cannot be considered absolute genomic markers. It is true that the putative orphan and silenced isozymes might result from mutation, but causal investigations should not be undertaken until the possibility of sampling error is ruled out. Although the true nature of orphan alleles and silenced genes has not been resolved in ferns, there can be little doubt that many allopolyploids are the result of several, independent hybridization events (Werth et al., 1985b; Haufler, 1985b). Most of the allopolyploid species examined to date show two or more fixed isozyme patterns which represent the amalgamation of genes from different pa- rental populations. Gametophytes from five plants representing two different Populations of the allotetraploid Pellaea wrightiana serve to illustrate the phe- nomenon of multiple polyploid origins. In this small sample, there are two PGI- 2 genotypes (Fig. 3), two SkDH genotypes (Fig. 4), and three IDH genotypes (Fig. 5). Considering all loci, plants c and d are identical to one another, but each of the other individuals shows a unique genotype. Thus, if there has been no ge- netic exchange among genotypes in P. wrightiana since their origin, the species must have arisen through at least four different hybridization events. Even if 8enetic exchange and mutation have occurred, at least two separate origins would required to explain the genetic variability observed in P. wrightiana. oe Investigations of multiple origins in allopolyploid ferns may shed new light on the biogeographical aspects of speciation and migration. Werth et al. (1985b) Stated that “areas now occupied by single genotypes may circumscribe centers of allopolyploid origin.” Under this hypothesis, allopolyploid populations show- Mg a mixture of isozyme phenotypes would result from range expansion and Secondary contact. However, it is equally plausible that regions showing a di- Versity of polyploid phenotypes represent the point of origin, and that genetically 124 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) uniform, outlying populations result from subsequent migration of single geno- types. The relative merits of these opposing hypotheses cannot be judged solely on the basis of data from the allopolyploid, but the question could be resolved through a careful examination of the diploid progenitors. If rare allopolyploid isozymes can be traced back to their diploid source through electrophoretic sur- veys of herbarium specimens, it may be possible to determine the approximate point of origin for each enzymatic phenotype observed in the polyploid taxon. This type of research might provide critical information concerning the early stages of allopolyploid speciation and establishment. Breeding system analyses and crossing programs.—Large-scale geographic surveys of enzyme variability can also improve our understanding of evolution- ary mechansims in diploid homosporous ferns. Breeding systems have a pro- found effect on gene flow, dispersability, and other factors involved in the spe- ciation process. Although Klekowski and Baker (1966) hypothesized that homosporous ferns are predominantly inbreeding, recent studies of populational variability in Bommeria (Haufler & Soltis, 1984), Cystopteris (Haufler, 1985b), and Pellaea (Gastony & Gottlieb, 1985) challenge this interpretation. The breed- ing system of most pteridophyte taxa remains to be established, and gameto- phytes derived from herbarium specimens can be useful in several ways. Since the inbreeding potential of some fern species varies from population to popu- lation (Cousens, 1979), any attempt to characterize the breeding system should incorporate data from a variety of habitats and localities. If an adequate sample taken throughout the range of a genetically variable diploid species reveals little genetic differentiation between populations, this provides circumstantial evi- dence that the species is largely outcrossing (Hamrick et al., 1979). If, on the other hand, genetic variants are geographically localized, it is likely that the species is primarily inbreeding. In the latter situation, it may be possible to investigate the rates and patterns of gene flow accompanying divergence be- tween populations because herbarium collections provide a sample of both geo- graphic and temporal variability. Artificial crossing experiments were considered an integral part of ‘‘biosyste- maty” by those who originally defined the term (Camp & Gilly, 1943), and most systematic studies of seed plants devote a great deal of time and effort to inves- tigations of crossability. Unfortunately, the same cannot be said for ferns, in which the gametophytes necessary for crossing experiments mature slowly and require special growth conditions to insure sporophyte production (Cousens, 1979). Most artificial crosses reported in the fern literature involve interspecific hy- bridization, a technique routinely used by European pteridologists (Lovis, 1968) but rarely applied in other regions of the world. Crossability has also been used as a measure of intraspecific genetic divergence in Bommeria hispida (Haufler, 1979) and Blechnum spicant (Cousens, 1979). Information gained from intra- an interspecific crossing experiments can be especially critical in groups such as the ferns where cryptic species are common and the number of phylogenetically useful morphological characters is limited. Interspecific crosses between genetically different parental populations yield hybrids with distinctive isozymic (Werth et al., 1985b), chromosomal (Wagner, WINDHAM & HAUFLER: GAMETOPHYTES 125 1971), and morphological (Barrington, 1985) phenotypes. For this reason, evolu- tionary hypotheses should not be based solely upon the results of a single hy- bridization experiment or event. Without a doubt, the greatest diversity of hy- brids can be obtained by using herbarium collections as the primary source of gametophytes. Lovis (1968) reported that certain natural hybrids, such as As- plenium ~ alternifolium, are extremely difficult to synthesize in the laboratory. This could result from genetic incompatibility between diploid strains used in the hybridization experiments. Thus, attempts to “recreate” a natural allopoly- ploid might be more successful and informative if the selected diploids were genetically similar to the supposed parents involved in the original cross. Again, electrophoretic surveys of gametophytes derived from a diverse array of pressed plants may help to locate putative parental populations. This approach can also enhance the utility of intraspecific crossing experiments, which have tradition- ally used morphological differences or geographic isolation as measures of ge- netic distance. Although morphology and geography often reflect the underlying genetic structure of a species, genetic distances can be calculated more precisely through surveys of enzyme variability. Following this, phenotypic, genotypic, and distributional data can be used to design a crossing program yielding the maxi- mum amount of information on intraspecific genetic diversity. The temporal component.—Considering all of the potential uses of herbarium specimens in biosystematic research, perhaps the most intriguing involves anal- yses of genetic change through time. In natural populations that are not subject to strong artificial selection, evolutionary changes are rarely observed during the tenure of most studies (or botanists). Instead, past evolutionary events are in- ferred from the differential distribution of morphological or genetic traits among modern populations. This approach may be a valid way to detect large-scale evolutionary changes, but it reveals little about the mechanisms involved. Since € evolutionary process is often equated with changes in allele frequencies within a population (Mayr, 1982), mechanistic investigations are probably best initiated at this level. Through the use of electrophoretic analysis of isozymes, gametophytes from herbarium specimens can generate valuable data concerning genetic change within a population. The only prerequisite for such studies is a representative sample of living plants on which to base estimates of allele frequencies for the modern population. Once this requirement is met, valid comparisons can be Made to either isolated plants or population samples previously collected at the same geographic locality. If statistical evaluations of changes in gene frequency are to be undertaken, gametophytes must be available from two or more plants collected at approximately the same time. Thus, the chosen locality must be Popular with collectors or visited by a botanist who collected a large number of duplicates. In either case the collections must have been deposited in herbaria where spores are likely to remain viable for long periods (e.g., COLO or UNM; See Windham et al., 1986). This is a highly improbable chain of events, and most analyses of genetic change using fern herbarium specimens will probably take the form of simple probability statements comparing isolated individuals from the past with the modern population. The strongest case for genetic change 126 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) would result from the detection of an allele in the pressed plant which was not found in the modern population. Although this is an unlikely situation, powerful probability statements can be generated for any genetic locus (or combination of loci) showing strongly skewed allele frequencies. For example, if a plant collected 50 years ago is homozygous for an allele which has a frequency of 5% in the modern population, the probability is very low (0.0025) that this individual was randomly selected from a genetically identical population. Since the her- barium specimen and modern sample belong to the same spatial population (i.e., they were collected at the same locality), it is likely that they represent different temporal populations. In other words, the probability is high that allele frequen- cies at this genetic locus have changed during the last 50 years as the population has evolved. A temporal series of isolated specimens from a single locality could be espe- cially informative when analyzed in conjunction with environmental data (such as those recorded in tree rings at the collection site). If such a series suggested a major change in allele frequencies during a certain interval, it might be pos- sible to correlate this change with an evolutionary bottleneck caused by drought, brush fire, or some other local catastrophe. Causal investigations of this sort have rarely been undertaken in wild populations, but they may be important to our understanding of evolutionary processes. There are potential limitations to this approach, most of which are related to the time span available for study. Pellaea and Marsilea show continued spore viability after 50 and 100 years respectively (Windham et al., 1986; Johnson, 1985), but substantial changes in gene frequency are not likely to be observed within this time frame without strong selection or some major perturbation of the environment. The probability of detecting such changes will be greatest in short-lived diploid species occupying an unstable habitat, and initial studies should probably focus on populations with these char- acteristics. CONCLUSIONS The fern life cycle comprises two completely independent generations, both of which must be investigated if we are to understand the important aspects of pteridophyte biology. Unfortunately, practical limitations have reduced the im- pact that analyses of the gametophytic generation have made on evolutionary studies. Difficulty in obtaining representative samples of living gametophytes has been the major problem, but this may be overcome in some ferns by germinating spores from herbarium specimens. Studying the gametophytes derived from these spores could improve our sampling of morphological variability and provide a new set of characters for use in phylogenetic analyses. Cytogenetic studies of such gametophytes can be used to document intra- and interspecific differences in chromosome number and may be the most reasonable approach to karyotypic research. Electrophoresis of herbarium-derived gametophytes can expand our otherwise limited sample of isozymic variability, can supply information on the origins of polyploid taxa, and can be used to search for “orphan” alleles. Through crossing programs, gametophytes originating from a diverse array of herbarium WINDHAM & HAUFLER: GAMETOPHYTES 127 specimens can be useful in documenting genetic discontinuities and recognizing cryptic species. They can also provide a window to the past for analyzing changes in gene frequency over time. Preliminary studies indicate that the usefulness of herbarium collections is not limited to investigations of sporophytic morphology, ecology, and biogeography. Through studies of gametophytes derived from spores, herbarium specimens of ferns can provide a wealth of information on pterido- phyte biology and evolution. At present, the only serious barrier to studies of this sort is our lack of knowledge concerning prolonged spore viability. It is hoped that the ideas presented herein will stimulate additional research on spore longevity in a variety of pteridophyte genera. Only then will we be able to assess fully the value of herbarium specimens for biosystematic research on ferns. LITERATURE CITED ATKINSON, L. 1973. The gametophyte and family relationships. Pp. 73-90 in The phylogeny and classification of the ferns, ed. A. Jermy et al., J. Linn. Soc., Bot. 67 (Suppl. 1). BARRINGTON, D. 1985. The oe and origin of a new Polystichum hybrid from Costa Rica. yst. Bot. 10:199-204 , C. Paris, and T. RANKER. 1986. Systematic inferences from spore and stomate size in the —1 Brown, D. 1964. A monograph of the ins genus Woodsia. Nova Hedwigia 16:1-154. Camp, H. and C. Gity. 1943. The structure and origin of species. Brittonia 4:323-385. Cousens, M. 1979. Gametophyte ontogeny, sex expression, and genetic load as measures of pop- tion divergence in Blechnum spicant. Amer. J. Bot. 66:116-132. DeBeNeocr, V. 1969. Apomixis in ferns with special reference to sterile hybrids. Unpublished D. dissertation, University of Michigan. Dyer, : 1979. The culture of gametophyes for experimental eerie Pp. 253-305 in The exper angie biology of ferns, ed. A. Dyer. London: Academ Fischer, H. 1911. Licht- und Dunkelkeimung bei cher Beih. a idee 27:60. ASTONY, G. 1986. mentee evidence for the origin of fern species by unreduced spores. Amer. J. Bot. 73:1563-1569 ‘ ane |. reba 1985. Genetic variation in the homosporous fern Pellaea andromedifolia. Amer. J. Bo ooo and G: tee 1976. Chromosome numbers and apomixis in the fern genus Bommeria. a er ee Ize. 1 . L + 2 ts a ica : Hamarick, J., Y. LINHART, and J. MiTTon. 1979. Rel electrophoretically detectable genetic variation in . plants. Ann. Rev. Bool ao 10:173-200. Haurter, C. 1979. A biosystematic revision of Bommeria. J. Arn Arnold r. me en os Enzyme variability and modes of evolution in Bommeria mevnretintalt Syst. eee 1985. Preridophyte evolutionary biology: the electrophoretic approach. Proc. Roy. Soc. Edinburgh 86B:315-323. >. and D. Soins. ce Obligate outcrossing in a homosporous fern: Field confirmation of a rspiwtade prediction. Amer. J. Bot. 71:878-881. SS NDHAM, D. BritTon, and S. ROBINSON. 1985. Triploidy and its evolutionary signif- in Cystopteris protrusa: Canad. J. Bot. 63:1855-1863. os, D. pad New records for longevity for Marsilea Kuskowski, E. and H. BAKER. 1966. Evolutionary significance Science 153: sens #6 Loyn, R. and E. KLEk 1970. Spore germination and viability in Pteridophyta: Evolutionary po significance of 1 ehioconhy tious spores. Biotropica 2:129-137. . Amer. Fern J. 75:30-31. of polyploidy in the Pteridophyta. 128 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) Lovis, J. rage Fern hybridists and fern hybridising. II. Fern hybridising at the University of Leeds. Brit. Fern Gaz. 10:13- Mayr, E. oe The growth of A sca! thought. Cambridge: Harvard University Press MILLER, J. 1968. Fern gametophytes as experimental material. Bot. Rev. (Lancaster) 34:36 51-44 Nayar, B. and S. Kaur. 1971. Gametophytes of homosporous ferns. Bot. Rev. (Lancaster) 37: 295- 396. Pray, T. 1968a. The gametophytes of Pellaea section Pellaea: Dark-stiped series. Phytomorphology 18:113-142. . 1968b. Interpopulational variation in the gametophytes of Pellaea andromedaefolia. Amer. J. Bot. 55:951-960. . 1971. The gametophytes of natural hybrids in the fern genus Pellaea. Amer. Fern J. 61: 128-136. RANKER, T. and C. WERTH. 1986. Acti\ f herbarium specimens: Electrophoresis as an ought. Amer. Fern J. 76:102-1 13, SmitH, A. 1975. The California species of Aspidotis. Madrofio 23:15-24. and M. Foster. 1984. Chromosome numbers and ecological observations of ferns from El Tirol, Paraguay. Fern Gaz. 12:321-329. Soxtis, D., C. HAurLer, D. DaRRow, and G. Gastony. 1983. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. Amer. -27 StTessins, G. 1950. Variation and evolution in plants. New York: Columbia University Press. STokey, A. 1951. The contribution by the gametophyte to classification of the homosporous ferns. Phytomorphology 1:39-58. Tryon, A. 1957. A revision of the fern genus Pellaea section Pellaea. Ann. Missouri Bot. Gard. 44: 1 . Ee . 1968. Comparisons of sexual and apogamous races in the fern genus Pellaea. Rhodora 70:1-24. and D. BRITTON. 1958. Cytotaxonomic studies on the fern genus Pellaea. Evolution 12: 7-145. TRYON, R. 1956. A revision of the American species of Notholaena. Contr. Gray Herb. 179:1-106. ———. A. TRYON. 1982. Ferns and allied plants. New York: Springer-Verlag. Vina, G. po The role of polyploidy in eis: ze 267-294 in Evolutionary biology, ed. V- ovak and B. Pacltova. Praha: Czech. A WAGNER, - F 1965. ages arene) in North sone and the question of its origin. J. Elisha hell Sci. Soc. 8 : ae Evolution 2 Psaeete in relation to the Appalachians. In The distributional history of the biota of the southern Appalachians. Part 2: Flora, ed. P. Holt. Va. Polytech. and State Univ., Res. Div. Mono. 2:147-192. , D. Farrar, and K. CHEN. 1965. A new sexual form of Pellaea glabella var. glabella from ions issouri. —. Kate 55: pile thes WALKER, T. 1973. E logy in th tion of ferns. Pp. 91-110 in i phylogeny and asa of the ferns, ed A. Jermy et al., J. Linn. Soc., Bot. 67 (Su 1985. Cytotaxonomic studies of the ferns of Trinidad 2. The cytology ~~ taxonomic implications. Bull. Brit. Mus. (Nat. Hist. ) ace a scosd 200. WertH, C., S. GuTrMan, and W. ESHBAUGH. 1985a. id f reticulate evolution in the Renee Asplenium poeapiet. Syst. Bot. 10: 184-182. a ——— J 0eon d species in Asplenium. oC uo Saas Race goa WINDHAM, M., P. WoLF, AND T. RANKER. 1986. Factors affecting pr rolenard score viability in her- barium specimens of three species of Pellaea. Amer. Fern J. 76:1 American Fern Journal 76(3):129-140 (1986) Detecting Abortive Spores in Herbarium Specimens of Sterile Hybrids W. H. WAGNER Jr. and F. S. WAGNER Department of Botany and Herbarium, University of Michigan, Ann Arbor, Michigan 48109 W. CARL TAYLOR Botany Section, Milwaukee Public Museum, Milwaukee, Wisconsin 53233 Uses of herbarium specimens for biosystematic research are being extended by new techniques, and those described here have been exploited only in the past few decades. Not only do dried and pressed pteridophyte specimens pro- vide such classical gross data as geographical, ecological, and morphological, but microscopic and chemical data as well. Specimens on herbarium sheets often yield hypotheses of hybridity by morphology alone, based on the phenomenon of intermediacy (W. Wagner, 1983), in which the characters of higher plant hy- brids tend to have expressions lying between those of the parental species: If one has ovate leaves, and the other lanceolate leaves, the hybrid has more or less ovate-lanceolate leaves. The hybrid’s characters are not necessarily in the middle; some may be closer to one parent, some closer to the other. Probably most investigations of hybrid pteridophytes begin in the herbarium, where specimens showing interspecific intermediacy are encountered. We call such specimens “‘intertaxa” because they do not fit either presumed parent. Whenever they are found we examine their spores. Interspecific pteridophyte hybrids fall into three categories as regards spore production, namely (a) sterile hybrids that have defective spores (e.g., Fig. 1), (b) fertile hybrids (either intro- gressant or, much more commonly, allopolyploid) that have normal viable spores, and (c) apogamous hybrids that produce unreduced spores. We are not con- cerned here with fertile hybrids, except for those that are apogamous (e.g., in the United States, Pteris x hillebrandii and Asplenium x heteroresiliens). Oblig- ately apogamous hybrids can usually be recognized because some or most of the sporangia have only 32 spores that are viable; those with 64 spores show abortion (various combinations are also possible; cf. Manton, 1950). Apogamous intertaxa involve one apogamous and one sexual parent. Non-hybrid taxa with normally 32 or 16 spores per sporangium might be confused with apogamous hybrids but these can be recognized because the reduced spore number is regularized, char- acteristic of all sporangia, and the spores are non-abortive. Examples of 32- or 16-spored sporangia in normal sexual species are found in a wide diversity of 8enera (e.g., Ceratopteris, Cheilanthes, Lindsaea, Alsophila, Sadleria). Allopo- lyploid sexual hybrids behave like ordinary species, even though they are ac- tually hybrids, and many of these are frequent or common; they are recognized by their spore size, usually larger than that of their parents. es Not all spore abortion is associated with hybridity (see below under “Addi- tional Perspectives”). Nearly all spore samples from normal show at least Some deviant spores. It is not unusual to find as much as 10-20 percent of spores 130 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) of normal species to be abortive. The causes for this are not known: it may be genetically determined or produced by environmental conditions. A most inter- esting group of non-intertaxa with abortive spores is that involving crosses be- tween different ploidal levels of the same species. These are often very hard to recognize on the basis of gross morphological characters. Fortunately plants of this type appear to be uncommon in nature, but it may simply be that they are undetected. Sterile intertaxa with defective spores are much more common than any other types of hybrids, and it is with these that we are mainly concerned here. When we consider the high incidence of sterile nothospecies among pteridophytes, the use of spore abortion as a means of identification can become extremely impor- tant, especially where the parental species are subtly differentiated on the basis of gross morphology. In the scouring rushes, for example, Hauke (1963) showed that several of the species combinations (E. x trachyodon [E. hyemale x vari- egatum], E. x ferrissii [E. hyemale x laevigatum], and E. X nelsonii [E. laevi- gatum X variegatum]) are surprisingly common in herbaria. However, they are distinguished by subtle vegetative characters, so that abortive spores provide one of the best aids for their identification. There are different kinds and degrees of hybrid spore abortion, depending upon the case, and so far as we know, no one has brought these together and classified them. The most complete study of hybrid abortion in ferns is by DeBenedictus (1969). Involved are the relative development of sporangia, de- posits of materials on the inner capsular wall, sizes of spores (both the outer and inner diameter of the perispore), spore surfaces, relative wall development (es- pecially the perispore), forms of the laesurae, spore pigmentation, extent of wall collapse, presence and condition of the protoplasts, and germination. Various combinations of these different abortive expressions produce a diversity of syn- dromes. Indeed, we believe that future studies may show that the nature of spore abortion is just as characteristic of a given hybrid combination as details of its morphology, cytology, or chemistry. TECHNIQUES The goal is to remove spores without damaging the herbarium specimen and to prepare the spores so that they may be observed and studied readily. The techniques described here apply to leptosporangiate ferns, but they may be readily adapted to other pteridophytes. One must beware of occasional contamination from other specimens; an interesting sterile hybrid may be contaminated wi good spores from adjacent sheets or specimens collected at the same time. Also, specimens with partly immature spores can be confusing; the mixture from 4 half-developed sorus of very young with nearly mature or mature spores may resemble true abortion. Given an uncontaminated and mature herbarium spec- imen that we want to test for presence of abortive spores, the procedure is aS follows: 1. Under dissecting microscope, examine specimen for areas of leaves or stems where there are numerous sporangia that are mature and well-developed. Find WAGNER ET AL.: SPORE ABORTION 131 sporangia that have not opened or that have opened only partially. Old, wide- open sporangia may have but few spores remaining, and young, tightly closed sporangia may have only incompletely formed spores, although these can be useful sometimes when abortion shows up early in development. 2. Place drop of Hoyer’s Solution (Anderson, 1954) on the middle of a slide and dip the point of a dissecting needle into the solution to wet the tip in order to provide a “glue” for picking up spores. 3. Gather material on tip of needle. For leptosporangiate ferns simply amass a number of spore-filled sporangia on the wet needle-tip. For eusporangiate ferns and other pteridophytes, dig into one or more sporangia so that the needle- tip becomes covered with spores. 4. Transfer spores or sporangia to the drop of Hoyer’s Solution on the micro- scope slide and stir in order to spread the spores evenly. Repeat (3) and (4) as necessary to accumulate roughly a hundred or more spores in the drop. 5. Pass needle-tip through flame of alcohol lamp and burn off all of the re- maining spores, sporangial debris, and Hoyer’s Solution. Clean the burned tip with paper tissues to prevent contamination when using the same needle on another herbarium collection. 6. Cover spores in the drop of Hoyer’s Solution by slowly lowering the cover glass over them. Tap gently on the top of the cover glass with the needle to spread the spores around uniformly and to flatten any sporangial or other soral debris that may be holding up the cover glass. Other media, e.g., diaphane, silica gel, glycerin jelly, may be used for mounting the spores, but in publication this should be stated, so that repeated measurements will be comparable. 7. Under the compound microscope, survey slide looking for evidence of abor- tion, as described below. If in doubt, make as many additional slides as neces- sary, following steps (2) through (7). 8. If slide proves to be interesting (as in Fig. 1) and of sufficient significance to retain, allow it to dry along the edges of the cover glass, after removing excess Hoyer’s Solution. When the edges become sufficiently dry (after a couple of weeks}, ring the cover glass with Glyptal (General Electric Co. product), and Permit the sealant to dry. Most herbaria prefer to have good, well-labeled spore Slides kept together with the specimen from which the sample came. The com- pletely dried slide may be placed in a pocket directly on the herbarium sheet with the specimen. (Such slides can be broken only by rough handling, not likely to be encountered in a well-kept herbarium.) SIGNS OF SPORE ABORTION Abortion can be characterized as significant variation from the normal prod- Ucts of sporangio- and sporogenesis. The different manifestations in spore abor- tion are interrelated, probably because they are traceable to the same morpho- 8enetic upsets. For example, spore samples showing exaggerated variability in Spore size (Fig. 2B) will show all or most of the symptoms given below. Spore collapse may go hand in hand with loss of the protoplast. Unusually massive Perisporial development often occurs together with reduction of the size of the AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) a fr a \" mS) ARTE Ms q Yen Fic. 1. Light microscope photograph of abortive spores of Botrychium ascendens * crenulatum Oregon, MICH. Note deviations in spore size, wall co apse, absence of protoplast (evidenced by black appearance). Dwarf and giant spores are especially conspicuous. WAGNER ET AL.: SPORE ABORTION 133 CO) Oo x AY Ono awe ae Fic. 2. Outline drawings (diagrammatic) of prominent spore features as seen under the light micro- scope. A, Methods of measuring longest diameter. A,, Bilateral spores, protoplast measurement. me Ti laesurae. G Diplosp 1 ing r dispersum, Florida). H, Nearly opaque, highly irregular, darkly pigmented abortive spores in which the separate wall layers and protoplast are not visible. I, Collapsed and twisted spores showing concave walls and strongly curved outlines. J, Optical effect of “air pockets” in spores that have lost their protoplasts. Protoplasts as compared to the parents. Major variables that characterize abor- tion are the following. Sporangium size.—In some hybrids, the sporangia are smaller than normal, approximately two-thirds or less the size of those of the parents (Wagner & Chen, 1965). Apparently the processes that upset spore development influence the mat- uration of the whole sporangium. The dwarfed sporangia fail to develop the annulus and stomium completely, and the sporangia do not dehisce. However, mixed with the dwarfed unopened sporangia there may be apparently normal Sporangia, even though the spores within are abortive. Thus there is not perfect correlation between abortion of the spores and abortion of the sporangia (De- 134 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) Benedictus, pls. 19, 20). Abortion of sporangia seems to be especially frequent in leptosporangiate ferns. One of the most extreme cases of sporangial abortion we have observed occurs in a very rare natural hybrid between Dryopteris and Polystichum in which the spores are completely deformed. In some years, even at the end of the growing season, no sporangia develop to full size; all are small, closed, and somewhat “crumpled.” The capsule walls have not fully differen- tiated; the annular cells are either entirely unthickened or only sporadically thickened, and their number is considerably reduced from the numbers in annuli normally found in the parents (Wagner & Wagner, unpubl.). Even more severe sporangial abortion was found in Osmunda claytoniana x regalis (O. x ruggii), to the extent that we could not recognize individual spores in the sporangium, the capsule containing coalesced masses of thin-walled cells, and the sporangial walls undifferentiated and fleshy (Wagner et al., 1978). Deposits on capsule walls.—Sterile hybrids of ferns with typical massive peri- spores (e.g., Dryopteris, Polystichum, Asplenium, Blechnum) often show an ir- regular development of the perispore in abortive spores, correlated with a de- posit of globules of dark material resembling in color the perisporial substance (DeBenedictus, 1969, pl. 22). Although no sporangial abortion per se is detected, this phenomenon usually indicates that something is unusual. It is particularly useful as an indicator in samples of hybrid ferns collected late in the growing season that have discharged all or most of their spores. Spore size.—Size of spores has two components: dimension of protoplast and dimension of the wall layer. Abortive hybrid spores show much greater vari- ability of size than do normal spores (Figs. 1, 2B). For measuring spores we use a single value, the longest diameter. Tetrahedral spores with very shallow, ru- dimentary perispore appear triangular under the microscope, the longest dimen- sion being the distance from angle to angle (Fig. 2A,). In bilateral spores, the longest dimension is the distance running from end to end as defined by the exospore (Fig. 2A,). The latter is complicated, however, by the presence in many bilateral spores of a jagged perispore of considerable thickness. The perispore, however, is usually transparent, so that the measurements can be taken of the outlines of the exospore which is more uniform. The irregular outlines of most perispores lead to an erroneous impression of variability when measured be- cause of the prominent tubercles and ridges (Fig. 2A,). In many abortive hybrid spores, unfortunately, it is sometimes difficult to see the exospore (Fig. 2H) and the overall length of the spore must then be used. In this case, comparative studies that include the parents demand that all the spores be measured the same way. Some trilete spores with cristae or equatorial flanges may be difficult to measure and may require measurement of the inner exospore wall. Data from spore sizes may be conveniently plotted as simple frequency histograms (cf. Wagner & Chen, 1965, fig. 4). Of special interest are the “giant” or “basketball” spores (Figs. 1, 2B;) studied extensively by DeBenedictus (1969); these are unreduced spores that have been found to be viable and to produce gametophytes. They are usually circular oF elliptical in outline and often lack laesurae. Giant spores are believed to be WAGNER ET AL.: SPORE ABORTION 135 unreduced spore mother cells produced by interruptions of meiosis, and often occur only 16 per sporangium in a genus where 64 is the normal number. Perhaps related cytokinetically are the occasional “double” spores (Taylor et al., 1985, pl. 2d) with two units connected by a narrow isthmus. If derived from bilateral spores, the two units, though parallel, are usually attached at one end (De- Benedictus, 1969, pls. 23, 26). If derived from tetrahedral spores, the two units commonly form a centrally constricted structure like a dumbbell or hourglass. Surface.—The investigation of surface characters received an enormous boost from the advent of the scanning electron microscope (SEM). Unfortunately, in both light microscopic and SEM studies, the observer has to use subjective judg- ment to pick out and illustrate “representative” or “average” spores. Students are often astonished by the amount of variation that exists in a whole field of spores. For that reason we utilize simple line tracings to illustrate variation in certain spore features (see below). The study of surface patterns in abortive hybrid spores is still in its infancy. Most students of pteridophyte spores have confined themselves to non-abortive spores of normal species. The time is ripe to make detailed SEM studies of hybrid spores to determine the morphogenetic expressions between the parental spore surfaces, i.e., the inheritance pattern of parental spore surfaces in hybrid taxa. For example, what are the spore sculp- tures of hybrids between species with spinose vs. cristate spores, or verrucate vs. pustulate spores? Examples of hybrid spore surfaces are given by Brooks (1982) and Taylor et al. (1985). The problem with most samples of abortive spores is their extreme variability in all respects, including their outer surfaces. A “typ- ical” or “average” spore surface is thus more difficult to judge than in normal spores. Only by illustrating the range of variations can the spore surfaces of a given intertaxon be completely documented. Perispore thickness. SEM does not show the extent of the perispore in com- parison with the protoplast, since the latter must be determined by the outline of the exospore which is visible only by light microscopy. Not only do normal species differ in the extent of their respective perisporial thickness and pattern (even between closely related taxa such as Cystopteris fragilis and C. dickieana), but nothospecies may differ from either of their parents in a transgressive expres- sion of exaggerated perispore material which may or may not be accompanied by a reduction in the size of the protoplast (Fig. 2C vs. 2D). Estimates of the extent of protoplast reduction and perispore exaggeration may be made in SEM only with broken spores. The best way of showing relative perispore/ protoplast thickness is by simple line drawings (such as those in W. Wagner, 1966, 1973; Wagner & Boydston, 1958; Fig. 2, A-F, J) which are tracings made with a drawing tube attachment on the compound light microscope. The outer lines of the trac- ~~ represent perispores and the inner the approximate borders of the proto- hvt ] liff t forms: side of a bilateral spore, producing diate laesura of a tetrahedral spore, producing a trilete spore (Fig. 2F). a.—The so-called “scar” of pteridop the single straight laesura running along one what is called a monolete spore; and the trira made up of three rays 120° from each other, 136 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) There is, in general, close correlation between spore type and the genus of pteridophytes (W. Wagner, 1974). Thus, Psilotum has monolete spores and Bo- trychium has trilete spores. In normal species, spores intermediate between these two types are rare (F. Wagner, 1985). In sterile hybrids, as well as apoga- mous hybrids, however, intermediate laesural forms occur frequently. In Hawaii, the nothospecies Pteris x hillebrandii, intermediate between P. cretica (apoga- mous) and P. irregularis (sexual) inherits the apogamous characters of P. cretica; the hybrid spores, like those of P. cretica, are much more diverse in shape than those of the sexual P. irregularis. Most interesting are their laesurae: Those of the sexual species are symmetrical and uniform, but those of the apomicts are asym- metrical and variable, running the gamut from trilete to nearly monolete, with the former prevailing (W. Wagner, 1974, fig. 6; this paper, Fig. 2E vs. 2F). Ap- parent hybridization between Polypodium atrum and P. plumula has produced the widespread neotropical P. dispersum. Possessing an unusual apogamous life cycle, P. dispersum is a triploid with 111 unpaired chromosomes but it produces viable spores that are more or less globular with highly variable scar configu- rations, including monolete and trilete and all forms between (Evans, 1969, fig. 14; this paper, Fig. 2G). Similar spherical and ovoid giant spores in many other hybrids show laesural irregularities similar to those described. Pigmentation.—Some hybrid abortion is revealed in herbarium specimens simply by scanning the sori with the naked eye. In hybrids they are often black- ish or dark brown, in contrast to normal relatives with light brown or tan sori. Further examination under the microscope reveals that the capsules have dark deposits and the spores have exaggerated dark perispores, so strongly pigmented that it is impossible to detect the outlines of the protoplast within (Fig. 2H). Good examples occur in the Appalachian Asplenium intertaxa (e.g., Wagner et al., 1973) in which the sori are deep chestnut in color because of the deposits. During the latter half of the growing season, the sori of normal, fertile species often expand to several times their normal width because most of the sporangia are wide open and extend in all directions, and the sori themselves appear to over- lap. The sori in most hybrids, in contrast, have so much sporangial abortion that they remain closely contiguous, and this, plus the dark color, distinguishes them from sori of normal species. Spore collapse.—Collapse of the spore takes the form of major concavities or isting of the entire exospore, and the whole spore may have a prune-like surface due to depressions (Figs. 1, 21). Some forms of collapse seem to be as- sociated with absence of the protoplast. Good illustrations of this condition in samples from dried herbarium specimens are seen in the Hawaiian hybrid whiskfern, Psilotum complanatum x nudum (W. Wagner, 1968, p. 117, fig. 1) and in the Michigan hybrid moonwort, Botrychium matricariifolium x simplex (W. Wagner, 1980, fig. 3). In the Psilotum, spores of the parents are regularly bean-shaped (except for a few abortive ones) and contain protoplasts with stor- age droplets. The hybrid has many spores with deep concavities, over half of them being empty shells. Spore collapse is probably associated with death of the protoplast because turgor is lost in final development. Drying of the specimen during pressing likely exacerbates the deformity. The technique of staining used WAGNER ET AL.: SPORE ABORTION 137 traditionally to test for absence of protoplast in pollen of angiosperms does not work well in most pteridophytes because the spore walls are too thick. Protoplast.—The protoplast may be influenced in various ways in hybrid spo- rogenesis. The amount and nature of food storage may be modified, or the entire protoplast, as noted above, may die and be absorbed. In Ly lium sect. Com- planata [=Diphasiastrum], Wilce (1965) discovered that certain interspecific crosses could be recognized by reduced amounts of oil in their protoplasts. Such reduction in oils was correlated with other signs of abortion, e.g., empty or mis- shapen spores. The massive resistant spore walls had to be broken before she could stain the protoplasts. Protoplast size varies greatly in abortive spores, some- times reaching notably small dimensions (Fig. 2B,). Such minute protoplasts are produced by unequal chromosomal segregation during meiosis, with certain tel- ophase nuclei receiving only a few chromosomes. When abnormal cytokinesis occurs, on the other hand, whole chromosome complements plus the cytoplasm may be doubled, producing giant spores. Some giant spores with huge protoplasts may be more than doubled, perhaps becoming 4x or even 8x. Generally, spores lacking a protoplast can be recognized by the absence of any materials inside the exospore; the spore is “blank,” as in the Psilotum no- thospecies discussed above. In herbarium specimens the empty spores are com- monly filled with air, producing optically a conspicuous black or dark gray area within the exospore (Figs. 1, 2J). We can say for certain that those spores that lack protoplasts are inviable. We cannot, however, be sure that those spores that have normal-appearing contents are all viable; this can be determined only by experimental germination. Germinability of spores.—The ultimate test of abortion is viability. Even spores that appear peculiar may germinate if they have not been killed by chemicals or heat for herbarium fumigation (Windham et al., 1986). In some hybrids, the Percentage of spores from dried specimens that germinate may be unexpectedly high (DeBenedictus, 1969; Whittier & Wagner, 1971). Some Dryopteris hybrids (for example, D. celsa x marginalis and D. clintoniana x cristata) have germi- nation of up to one-third of the spores sowed, in spite of great variability in size and other signs of abortion (cf. Whittier & Wagner, 1971, table 1). Among spleen- Worts such intertaxa as Asplenium pinnatifidum x platyneuron (A. x kentuck- iense) and A. montanum x pinnatifidum (A. x trudellii) are notable for their re- markable amounts of germination, probably from giant spores (Wagner & Wagner, unpubl.). DeBenedictus (1969) made a detailed study of a number of “fertile Sterile hybrids, and found that unreduced spores formed gametophytes and that these produced sporophytes apogamously. We believe, therefore, that in spite of their mostly abortive spores, many “sterile” hybrids are capable of forming at least limited populations. SOME ADDITIONAL PERSPECTIVES _Abortive spores of the types described here are best known in homosporous Pleridophyte hybrids, especially ferns. But abortion of these or other types also °ccurs in non-hybrid pteridophytes and in hybrids and non-hybrids of hetero- 138 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) sporous pteridophytes and seed plants. Genetic or environmental factors may produce the spore abortion found in some herbarium specimens, the plants in- volved being normal species and not sterile intertaxa. For example, we are currently studying a wild mutant of Asplenium platyneuron with a level of spore abortion that exceeds that of the normal form that grows with it. The mutant also has a tendency for 32-spored sporangia (Wagner & Wagner, unpubl.). Spore abortion also within species is seen in the occasional triploids in which the process of meiosis is confused by the presence of three rather than two chro- mosome sets. However, Haufler et al. (1985) determined that in triploid Cysto- pteris protrusa numerous normal-appearing spores occur, in addition to spores that are deformed, making a triploid diagnosis difficult. Hybrids between diploid and tetraploid varieties of the same species produce abortive spores, e.g., triploid Asplenium trichomanes in the eastern United States and Canada. The use of normal vs. abortive spores and small vs. large spores in herbarium material was nicely illustrated by Moran (1982), who was able to describe the geographical distributions of the 2x, 3x, and 4x forms of A. trichomanes, the first with small spores, the second with abortive spores, and the third with large spores. Another non-hybrid source of abortion may be cold or drought shocks during spore development. Wilce (1965) stated that “In Lycopodium alpinum, L. sitch- ense, and L. complanatum, the relatively high rate of abortion is perhaps ac- counted for by the northern distribution of these species, for abortive spores and pollen are not uncommon in other plants of high latitudes.” She also suggested that some spore abortion in what appear to be normal species may be due to backcrossing that produces plants so closely resembling one of the parents that they cannot be distinguished. In a monograph of the genus Pyrrosia, Hovenkamp (1986) reported collections with abortive spores in nearly three dozen taxa, all of which also have collections with normal spores. That unusual patterns of abortion occur in hybrids of certain heterosporous pteridophyte groups is nicely illustrated by the hybrids of the quillwort genus Isoétes. Extensive recent studies by one of us (Taylor) indicate that hybridization is much more common than previously realized. The hybrids are recognized by their production of flattened and polymorphic megaspores of which less than 1% germinate in culture. Hybrid Isoétes show two types of megaspore abortion. In diploid hybrids, megaspores are seen to be relatively uniform in size and shape but they appear small and flattened (especially on their proximal sides). In hybrids of higher ploidy, spore abortion is characterized by the production of polymorphic megaspores. While there continues to be production of numerous flattened megaspores like those found in the diploids, there are also many larger. irregularly shaped megaspores. Some of these larger megaspores are hemispher- ical, others are found to be fusiform, and a few are even dumbbell-shaped when two spores are linked by an isthmus. Occasionally, an entire tetrad will be found that is interconnected by isthmuses. Further, the triradiate ridges that typically mark the proximal side of the megaspore are usually irregularly formed. In addition to variation in spore shape and triradiate ridge form, megaspore di- mensions may range from less than 100 um to over 1000 um within a single sporangium. WAGNER ET AL.: SPORE ABORTION 139 Although some of the larger quillwort megaspores open and exude a pad of gametophyte tissue, the few spores that germinate to produce gametes with nor- mal-looking archegonia are not the abnormally large “basketball” spores de- scribed for homosporous pteridophytes but globose, average-sized megaspores which occur in very low numbers. It appears that the globose, more normal- looking megaspores possess an ornamentation pattern intermediate between pu- tative parents. Microspore abortion is also evident in hybrid taxa. Abortive mi- crospores are irregular in shape and most appear to be devoid of cytoplasm. We should mention here the situation so common in flowering plants, in which the pollen grains may be quite uniform in size and shape but lack protoplasts. Thus it is possible to take masses of pollen grains, stain them, and determine which grains have and which have no protoplasts. Those that do not take stain lack protoplasts. Seed plant hybrids tend to have high percentages of empty pollen grains, and normal species have low percentages. We believe that one of the primary functions of surveys of herbarium collec- tions of pteridophytes for spore abortion is to form testable hypotheses regarding hybridity. The coincidence of interspecific intermediacy together with abortive spores is a sure sign of hybridity at least in homosporous pteridophytes; this combination of circumstances has occurred so commonly in biosystematic re- Search over the past thirty years that likelihood of any other explanation is exceedingly small. However, abortion may sometimes be due to other factors than hybridity, as pointed out here. We hope that in the future more will be learned about the comparative aspects of abortion. For example, are the patterns of abortion different in different intertaxa? Is hybrid abortion different from intraspecific? The fact that we find such strong differences between groups such as homosporous and heterosporous pteridophytes and betweer non-seed plants and seed plants suggests that there are important differences in sporogenesis that characterize these groups. We can learn about these questions only by acquiring more and better data than we have now. LITERATURE CITED ANDERSON, L. W. 1954, Hoyer’s solution as rapid permanent mounting medium for bryophytes. Bryologist 57:242-244. : : DEBENEDICTUS, V. M. M. 1969. Apomixis in ferns with special reference to sterile hybrids. Ph.D. Dissertation, University of Michigan. Brooks, R. E. 1982. A new Woodsia hybrid from Kansas. Amer. Fern J. 72:79-84. ANS, A.M. 1969. Interspecific relationships in the Polypodium pectinatum-plumula complex. Ann. Missouri Bot. Gard. 55:193-283. foe : Haurter, C. H., M. C. WinpHam, D. M. Britton, and S. J. ROBINSON. 1985. Triploidy and its evolutionary significance in Cystopteris protrusa. Canad. ]. Bot. 63:1855-1 Hauke, R. L. 1963, A taxonomic monograph of the aie subgen Nova Hedwigia 8:1-123, tables 1-12, graph 1-3, pl. 1-22. : : HoveNkamp, P. 1986, A monograph of the genus ba Brill/Leiden Univ. poe es NTON, I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge: us Hippocheete. Beih. Moran, R. C. 1982. The Asplenium trichomanes complex in the United States and adjacent Can- ada. Amer. Fern J. 72:5-1 140 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) TAYLOR, C., N. T. LUEBKE, and M. B. SmitH. 1985. ee SE and hybridization in North Amer- an Quillworts. Proc. Royal Soc. Edinburgh 86B:259-263. WAGNER, my S. 1985. ro . spores in New World sta ferns. Amer. Fern J. 75:6-1 Wacner, W. H. Jr. 1966. New data on North American oak ferns, Gymnocarpium. Rhodora 68: 121-134. . 1968. Hybridization, taxonomy, and evolution. oo 9 in Modern methods in plant taxonomy, ed. V. H. Heywood. London: Academic . 1973. iain of holy ed (Polystichum) in —. ani United States and adjacent Canada. . 1974. Structure a spores in se to fern phylogeny. Ann. Missouri Bot. Gard. 61:332- : 1980. A probable new hybrid paper Botrychium matricariifolium x simplex, from central pane oe omy Bot. as sisi . 1983. Reti 3 4} 1 h eee Sere eee Ly 4 +3 Chapt. 8 in Petite in cladistics. Vol. 2, eds. N. I. Platnick and V. A. Funk. New York: Columbia Univ. Press. and K. E. BoypsTon. 1958. A new hybrid spleenwort from artificial cultures at Fernwood per its relationship to a peculiar plant from West Virginia. Amer. Fern J. 48:146-159. and K. L. CHEN. 1965. mn of spores and sporangia as a tool in the detection of Dryopteris hybrids. Amer. Fern J. 55:9-29 , F. S. Wacner, J. A. LANKALIs, and J. F. Marriews. Ants. Asplenian montanum neuron. A new aang member of the Appala J. Elisha Mitchell Sci. Soc. soning ——.,,C.N. Said, Jr., and D. H. WAGNER. 1978. New observations on the royal ern hybrid, Osmunda x ruggii. Rhodora 80:92-106. sc D. P. and W. H. Wacner, Jr. 1971. The variation in spore size and germination in Dryopteris taxa. Amer. Fern J. 61:123-127. Witce, J. H. 1965. Section Complanata of the genus Lycopodium. Beih. Nova Hedwigia 19:i-ix, 1- 233, pl. : WINDHAM, M., P. G. Wotr, and T. A. RANKER. 1986. Factors affecting exiniggee spore viability in herbarium collections of three species of Pellaea. Amer. Fern J. 76:1 American Fern Journal 76(3):141-148 (1986) Factors Affecting Prolonged Spore Viability in Herbarium Collections of Three Species of Pellaea MIcHAEL D. WINDHAM, PAuL G. Wo tr, and THomas A. RANKER Department of Botany, University of Kansas, Lawrence, Kansas 66045 Among lower vascular plants, the establishment of new individuals and pop- ulations is largely dependent on the transport of spores by prevailing winds (Tryon, 1970; Wagner, 1972). Since long distance dispersal may require prolonged exposure of spores to unfavorable conditions, spore longevity has a major impact on species migration and distribution. The degree to which spores remain viable for long periods also determines the applicability of many biosystematic tech- niques to herbarium specimens (Windham & Haufler, 1986). Although an un- derstanding of spore longevity is thus critical to investigations of several biolog- ical phenomena, little is known about the factors affecting prolonged spore viability in ferns. Previous studies have shown that the spores of various pteridophyte species differ greatly in their ability to germinate after storage at ambient temperatures (Dyer, 1979). Two major groups of ferns have been identified on the basis of spore longevity (Lloyd & Klekowski, 1970): those with green spores containing active chlorophyll (mean viability of 48 days), and those with non-chlorophyllous spores (mean viability of 2.8 years). Most studies of the factors affecting spore longevity have compared these two groups. For example, Lloyd & Klekowski (1970) suggested that high water content, elevated respiratory rates, and the ab- Sence of a desiccation-resistant spore wall might account for the short viability of chlorophyllous spores. Since viability declines rapidly in Equisetum spores even when they are protected from desiccation (Hauke, 1963), the intensity of respiratory activity seems to be the limiting factor in this case. Although these Studies may explain the lack of long-term viability in green spores, they shed little light on factors controlling longevity in the majority of fern taxa with non- seen spores. oo Depending on the species, non-chlorophyllous fern spores remain viable for Periods ranging from several months (Lloyd & Klekowski, 1970) to nearly 100 years (Johnson, 1985). Both temperature and humidity are known to influence longevity (Okada, 1929), but a variety of other factors may also be involved. Although differences in respiratory rate seem relatively minor in this group (Oka- da, 1929), they may prove significant when calculated in terms of years or de- cades. Thus, ploidy level could be an important factor because polyploids tend to show lower respiratory rates (Levin, 1983). In addition, the lower surface/ volume ratios of polyploid spores would reduce the relative exposure of cyto- plasm to the unfavorable environment surrounding the spore. For herbarium specimens traditionally used in spore longevity studies, this unfavorable envi- Tonment may include a variety of physical and chemical treatments used for Preservation and insect control. All of these factors have the potential to affect 142 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) spore longevity, and their relative importance should be assessed through com- parative studies of germination in a variety of non-chlorophyllous fern spores. The fern genus Pellaea provides a good starting point for such investigations. Many members of this genus are xerophytic, and Pray (1968) indicated that the spores of most specimens remain capable of germination 10-15 years after press- ing. In addition, Mickel (1962, pers. comm.) reported that a spore collection of Pellaea atropurpurea (L.) Link yielded viable gametophytes after 34 years, one of the oldest germination records among homosporous ferns. Several taxa (and ploidy levels) are common in the United States, where they have been heavily collected through the years and are well-represented in both large and small herbaria. The present study examines the germination ability of spores derived from herbarium specimens of three closely related species: P. truncata Goodd., P. wrightiana Hook., and P. ternifolia (Cav.) Link. Germination patterns are then compared with data on ploidy level, specimen age, and herbarium treatment to elucidate some of the factors affecting spore longevity in Pellaea. MATERIALS AND METHODS Collections from the following ten herbaria were included in the study after obtaining explicit permission from the respective curators (abbreviations are those of Holmgren et al., 1981): ARIZ, ASU, COLO, MICH, MO, MSC, NY, UC, UNM, and US. For each herbarium, one sporulating specimen of Pellaea trun- cata (diploid) and P. wrightiana (tetraploid) was chosen to represent each decade from the 1930s to the 1980s. When collections of these species were unavailable, the appropriate cytotypes of P. ternifolia were substituted. Spores from all sam- ples were measured at 400 using an ocular micrometer mounted on a Leitz compound microscope, and ploidy levels were then assigned on the basis of established correlations between chromosome number and spore length (Tryon, 1968; Windham, unpubl.). The effect of ploidy level on prolonged spore viability was tested using the paired collections of haploid and diploid spores from each herbarium. Although it would have been desirable to base this comparison solely on the cytotypes of a single autopolyploid species (such as P. ternifolia), this approach was rendered impossible by the small number of specimens available. The allotetraploid P. wrightiana was derived through hybridization between P. truncata and P. ternifolia (Wagner, 1965), and it was hoped that genetic similar- ities among these taxa would minimize the effect of interspecific differences on the paired ploidy level comparisons. Spores were cultured on an inorganic medium enriched with Parker’s mac- ronutrients and Thompson’s micronutrients and solidified with 1% agar (Kle- kowski, 1969). The medium was autoclaved and poured into sterilized plastic petri dishes (35 mm diameter) which were kept inside clear plastic boxes throughout the experiment to prevent excessive water loss. Plants chosen 4s spore sources were lightly dusted with compressed air prior to sowing to reduce the possibility of contamination by stray spores. Spores were sown by carefully miverting a sporulating section of an herbarium specimen directly over a petri dish and lightly tapping the sheet to deliver the desired quantity of spores (ca. WINDHAM ET AL.: SPORE VIABILITY 143 1000 per dish where available). The cultures were maintained under Sylvania Gro-Lux bulbs on a 12 hour photoperiod at an irradiance of 73 microeinsteins/ m’/sec. (PAR) and a temperature of 22 (+2)°C. Germination was defined as the first emergence of the gametophyte from the ruptured spore wall since this is the most reliable measure of physiological activity in the absence of ultrastructural or biochemical data (Miller, 1968). The condition of the spores was checked every three days for most collections. Pre- liminary data gathered at three-day intervals suggested that specimen age might be positively correlated with the amount of time required to germinate. Since COLO and UNM provided the most complete temporal series of germinating spores, collections from these herbaria were resown and observed daily to de- termine the strength of the correlation. Sample size was increased in this analysis by incorporating several specimens that were not part of the original study. Spearman’s rank correlation coefficients (r; Spearman, 1904; Zar, 1972) were calculated to test for significant associations between age of specimen and num- ber of days from sowing to germination. A spore collection was considered viable when more than 1% of the spores in a petri dish showed signs of gametophytic emergence. Collections showing less than 1% germination were considered dead since the possibility of contamination by extraneous spores could not be ruled out. All germination ceased during the fifth week after sowing, and observations were discontinued after seven weeks. Percent germination was estimated at the end of the experiment by determining whether viable gametophytes outnum- bered ungerminated spores. A complete list of the specimens used in this study is available from the senior author upon request. RESULTS Several patterns emerge from the spore germination data (Table 1}, the most obvious of which involve differences in viability between herbaria. All speci- mens from ASU and COLO, and all but one from UNM, produced spores ca- pable of germination. All collections from MO, UC, and US failed to germinate, and germination was observed in only a single specimen from MICH. At the Temaining three herbaria (ARIZ, MSC, and NY), only relatively recent collec- tions (post 1940s} germinated. Low germination percentages in five of the six oldest collections provided additional evidence of the predicted negative cor- relation between spore viability and specimen age. However, the rate of decline in germination ability was much lower than expected. Eleven of the 46 speci- Mens (24%) collected prior to 1960 showed spore germination, and one plant of ellaea truncata collected in 1936 (Craig s.n., UNM} still had viable spores. Although spores from both COLO and UNM remained viable for nearly 50 years, the two herbaria showed clear differences in germination rate. All spores from COLO germinated in six to ten days, whereas samples from UNM required five to 25 days to germinate. In the sample from COLO, the number of days to germinate was not associated with the age of the specimen (r, ae 0.470, 0.10 < P < 0.15). In contrast, there was a highly significant positive association between time to germination and age of specimen in cultures from UNM (r, = 0.969, P < AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) 144 TABLE 1. Variability in Spore Germination Among Herbarium Specimens of Pellaea. Her- ba 1980s 1970s 1960s 1950s 1940s 1930s r ium 2x 4x 2x 4x 2x 4x 2x 4x 2x 0 4x BX Cax ARIZ + + ++ ++ —1 - - — ASU ++ ++ ++ +4 ++ ++ + COLO +1 ++ ++ ++? ++ +4 os - MICH we hott = is fe 0 bo MO Pin | ENE, | patna | pre | ee nerd Diem bes ra MSC ae _ ++ ++ ++ 1 +4 - ++ +4 - - - — - = = = UC _ - - = a = & UNM ~ ++ + ++ + ++ + + + = US _ ie - be ~ 0.9, fora 2 x 2 contingency table of ploidy [haploid or diploid] vs. germination [yes or no]). DISCUSSION Ploidy level.—At the outset of this study, it was predicted that spore longevity should be more prevalent at higher ploidy levels since polyploids tend to show lower respiratory rates and have larger spores with lower surface/volume ratios. Of the seven fern taxa previously reported to germinate after 25 years (Allsopp: 1952; Fischer, 1911; Johnson, 1985; J. Mickel, 1962, pers. comm.), two (Asplenium serra Langsd. & Fisch. and Pellaea atropurpurea (L.) Link) are definitely poly- ploid (Live et al., 1977). The remaining five species belong to the genus Marsilea, for which few chromosome counts are available. Even if all of the Marsilea species were diploid, they would not be comparable to diploid homosporous ferns due to the large size of the megaspores and the presence of a protective sporocarp. Thus, on the basis of these limited data, it might be argued that the WINDHAM ET AL.: SPORE VIABILITY 145 spores of polyploids do show greater longevity. However, analyses of the 90 specimens included in the present study do not support the predicted relation- ship between ploidy level and spore longevity. Although diploid plants from UNM tend to show lower germination percentages than tetraploids (Table 1), this pattern was not observed in other herbarium collections and long-term vi- ability was apparently unaffected. Overall, the oldest germinating spores were derived from the diploid species Pellaea truncata (Table 1), but the chi-square test verified that germination ability was independent of ploidy level. Specimen age.—In herbaria showing some germination activity, percent ger- mination tended to decline through time, eventually culminating in a complete loss of viability (Table 1). Occasional deviations from this trend probably can be explained in light of the unique history of each specimen. For example, the tetraploid 1950s collection from MSC showed greater than 50% germination whereas the diploid sample from that decade failed to germinate entirely. It therefore seems likely that one of these plants did not follow the normal se- quence of events involved in curation at MSC. Further investigation has shown that the tetraploid specimen (Cobean s.n.) was derived from a plant in the Uni- versity of California Botanical Garden which was still alive and providing chro- mosome counts in the mid-1960s (Lloyd, 1966). Thus, although this plant was collected in 1956, it probably wasn’t pressed and incorporated into the herbarium until the 1960s, and the germination response reflects this fact. If the history of each specimen included in this study were known, it is quite possible that most of the anomalies in the germination records would disappear. Regardless of such individual idiosyncrasies, the overall pattern of germination remains relatively clear and interpretable. The prolonged retention of spore viability in these Pellaea collections was somewhat unexpected. Twenty-four percent of specimens older than 25 years Temained capable of germination, and the spores of one plant were viable after 90 years. The 1930s were arbitrarily chosen as the early limit of this investigation based upon a small pilot study which yielded no germination prior to 1936. However, most of the pre-1930 specimens included in that study were derived from MO and US, which showed no evidence of germination even in recent collections. Therefore, it is quite possible that further research using older spec- imens from ASU, COLO, or UNM will produce new records for spore longevity in Pellaea. Se Herbarium treatment.—The number of days from sowing to germination in- creased as the age of the specimen increased in most herbarium collections, and UNM provides a clear example of this trend. The only clear exception is COLO, which shows no association between time to germination and date of collection. This may be explained by the fact that the University of Colorado herbarium is Situated in a cool, dry environment and has never been forced to use poisons or insect repellents (W. Weber, in litt.). Thus, instead of being a natural — quence of aging, slower germination may be caused (or at least accentuated) by ‘Treatments designed to keep the herbarium free of destructive insects. In fact, it appears that most of the differences in germination among herbaria may be a result of different curatorial techniques. 146 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) All collections from MO and US were subject to repeated or continuous ex- posure to methyl bromide or Dowfume, both of which are listed by the Environ- mental Protection Agency as highly toxic (Category I) poisons (Schofield & Cri- safulli, 1980). Judging from the fact that specimens from MO and US showed no germination, it is likely that these two insecticides are just as deadly to fern spores as they are to insects and humans. In the late 1970s, NY discontinued the use of Dowfume and substituted pyrethrum, which is effective against insects but harmless to mammals (Schofield & Crisafulli, 1980). The pronounced increase in spore viability at NY during this period may be a consequence of this change in curatorial practices. Chemical treatments do not represent the only threat to spore longevity in herbarium collections. All specimens at UC were heat-treated at least once at 60°C for 24 hours, apparently resulting in a complete loss of viability. MICH and ARIZ have used a combination of insecticides through the years, including heat, para-dichlorobenzene (PDB), and (at the latter herbarium) carbon bisulfide or carbon tetrachloride. This may account for the sporadic ger- mination patterns observed in collections from these herbaria. Until 1980, the sole insecticide used at ASU was PDB, after which time all incoming and out- going specimens were frozen for about four days at —20°C. These techniques appear to have little or no effect on spore viability since all collections from this herbarium germinated. UNM and MSC used identical treatments (PDB only), but the latter herbarium showed reduced spore longevity. Higher humidity at Michigan State University could be responsible for this loss (see Dyer, 1979), and it may be significant that spore viability increased at MSC in the 1960s when the herbarium was first air-conditioned. Percentage and rate of germination were highest at COLO, where prevailing environmental conditions make insec- ticides unnecessary. Botanists engaged in biosystematic studies of ferns have traditionally under- estimated the research value of herbarium specimens. Now that this potential is being realized, investigators should do everything possible to insure that her- barium resources are not depleted or misused. Since viable spores are especially useful in biosystematic research (Windham & Haufler, 1986), efforts should be made to identify and implement curatorial techniques that maximize spore lon- gevity in herbarium collections. Obviously, most herbaria cannot follow the ex- ample of COLO, or they soon would be overrun with insects. However, freezing of specimens has proven effective in insect control, and exposure to temperatures of —20°C seems to have no ill effects on spore viability. Although long-term impacts cannot be assessed at this time, it appears that the insecticide pyre (in the form of Dione powder) may have little impact on viability. It is clearly preferable to the use of Dowfume, methyl bromide, and heat treatments, which quickly result in loss of viability and should be discontinued if at all possible. None of the herbaria included in this study use microwave radiation, but this should also be added to the list of treatments that are likely to reduce spore longevity. Although effective against insects, microwave exposure disrupts the internal cell structure of pressed plants (Bacci et al., 1985) and has been shown to reduce seed viability in herbarium specimens of Malvaceae (Hill, 1983). Aside from encouraging curators to experiment with less destructive means of WINDHAM ET AL.: SPORE VIABILITY 147 insect control, individual botanists can increase the future research value of their own collections by depositing duplicates in herbaria where such treatments are already in use. A comprehensive list of treatments used by each herbarium could be included in future editions of Index Herbariorum and would be useful to collectors and biosystematists alike. For their part, researchers intending to use herbarium specimens for biosystematic studies should fully describe their tech- niques to the curator, obtain permission in advance, and carefully follow all curatorial instructions. Through close cooperation among collectors, herbarium personnel, and biosystematists, the research value of herbarium collections can be greatly enhanced. We thank the following curators for granting permission to use their specimens in the germination study and for providing information concerning curatorial techniques: Dr. John Beaman (MSC), Dr. David Lellinger (US), Dr. William Martin (UNM), Dr. Charles Mason (ARIZ), Dr. John Mickel (NY), Dr. Nancy Morin (MO), Dr. Donald Pinkava (ASU), Dr. Alan Smith (UC), Dr. Warren H. Wagner, Jr. (MICH), and Dr. William Weber (COLO). We also thank Dr. Christopher Haufler for reading the original manuscript and providing helpful comments and criticisms. : LITERATURE CITED ALLsopp, A. 1952. Longevity of Marsilea sporocarps. Nature 169:79-80. Bacci, M., A. CHeccucct, G. CHEccucct, and M. PALANDRI. 1985. Microwave drying of herbarium specimens. Taxon 34:649-653. : Dyer, A. 1979. The culture of gametophytes for experimental investigation. Pp. 253-305 in The experimental biology of ferns, ed. A. Dyer. London: Academic Press FiscHer, H. 1911. Licht- und Dunkelkeimung bei Farnsporen. Beih. Bot. Centralbl. 27:60. Hauke, R. 1963. A taxonomic monograph of the genus Equisetum, subgenus Hippochaete. Nova Hedwigia 8:1-123. Hit, S. 1983. Microwave and the herbarium specimen: potential dangers. Taxon 32:614-615. Hotmcren, P., W. KEUKEN, and E. SCHOFIELD. 1981. Index herbariorum. Part I. 7th ed. Regnum Veg. 106:1-452. JOHNSoN, D. 1985. New records for longevity of Marsilea sporocarps. Amer. Fern J. 75:30-31. Kirkowsx1, E. 1969. Reproductive biology of the Pteridophyta III. A study of the Blechnaceae. J. inn. Soc., Bot. 62:361-377. Levin, D. 1983. Polyploidy and novelty in flowering plants. Amer. Naturalist 122:1-25. Luoyp, R. 1966. Pteridaceae. In IOPB chromosome number reports VIII, ed. A. Léve. Taxon 15: 282-283. i : ~———— and E. KLEKowski. 1970. Spore germination and viability in Pteridophyta: Evolutionary significance of chlorophyllous spores. Biotropica 2:129-137. Love, A., D. Lov and R. Pasa Sean. ee. Cytotaxonomical atlas of the Pteridophyta. Vaduz: er. Micke, J. 1962. A monographic study of the fern genus Anemia, subgenus Coptophyllum. Iowa State Coll. J. Sci. 36:349-482. agg MILter, J. 1968. Fi tophvt tal material. Bot. Rev. (Lancaster) 34:361-440. erm asi = ig eae alte i . * : Okapa, Y, 1929, otes on the germination of the spores of -~ apse with special regard to their viability. Sci. Rep. Tahoku Imp. Univ., Ser. 4, Biol. 4:1 ~182. Pray, T. 1968. The gametophytes of Pellaea section Pellaea: Dark-stiped series. Phytomorphology 18:113-142. : : ae ScHortetp, E. and S. CrisAFuLL. 1980. A safer insecticide for herbarium use. Brittonia 32:58 62. 148 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) SPEARMAN, C. 1904. The proof and measurement of association between two things. Amer. J. Psy- chology 15:72-101. Tryon, A. 1968. Comparisons of sexual and apogamous races in the fern genus Pellaea. Rhodora re 70: Tryon, R. 1970. Development and evolution of fern floras of oceanic islands. Biotropica 2:76-84. WacneR, W. Jr. 1965. Pellaea wrightiana in North Carolina and the question of its origin. J. Elisha pt —-.. . Disjunctions in homosporous vascular plants. Ann. Missouri Bot. Gard. 59:203-217. WINDHAM, Ma snc C. HAuFLER. 1986. Biosystematic uses of fern gametophytes derived from her- imens. rae Fern J. 76:114-128. ZAR, J. 1972. F aenttia aia testing of the Spearman rank correlation coefficient. J. Amer. Statist. American Fern Journal 76(3):149-159 (1986) Systematic Inferences from Spore and Stomate Size in the Ferns Davip S. BARRINGTON and Catny A. Paris Botany Department, University of Vermont, Burlington, Vermont 05405-0086 THOMAS A. RANKER Botany Department, University of Kansas, Lawrence, Kansas 66045 Size of equivalent cells has traditionally been assumed to be constant within species and variable between species. Systematists have commonly measured cells with constant form, such as spores and stomates, as a means of distinguish- ing species and hybrids in polyploid complexes, since the best known factor determining cell size is ploidy level (Stebbins, 1950). Preliminary hypotheses about polyploid evolution can be generated directly from herbarium specimens based on spore and stomate measurement data. The first evidence of a relation between cell size and polyploidy in the ferns was Lawton’s 1932 demonstration that prothallial cells, lower epidermal cells, and stomates in induced apogamous races of Dryopteris marginalis and Wood- wardia virginica vary directly with ploidy level. Butters and Tryon’s 1948 work on Woodsia x abbeae included epidermal cell and annulus cell measurements that corroborated data from spore germination experiments to document an in- stance of somatic autopolyploidization. More recently a considerable body of evidence has been assembled support- ing the contention that spore size is related to ploidy level in the ferns. Manton (1950) noted that Vancouver Island Cystopteris fragilis (n = 84) has smaller spores n Swiss Cystopteris alpina (n = 126). Hagenah (1961) postulated a polyploid Series in species of Cystopteris based on spore-size measurements. In a mono- graph of Cystopteris, Blasdell (1963) concluded that it was possible to infer ploidy level from spore size in that genus. Blasdell assigned ploidy levels to cytologically unknown components of polyploid complexes based only on spore size (see Lovis, 1977 for a critique). Wagner (1966) demonstrated that there were two varieties of Gymnocarpium dryopteris: a larger-spored typical variety, which is tetraploid, and a smaller-spored variety disjunctum, which is diploid. Schneller (1974) recently provided an analysis of ploidy level and spore size in the Dryo- Pleris filix-mas group of Europe. He concluded that ploidy and spore size are Positively correlated for diploid to pentaploid cytotypes in a‘closely knit phylo- 8enetic group. In an exhaustive analysis of spore features of northeastern North American Isoetes, Kott and Britton (1983) provided new insight into spore-size variability. Teas megaspore variation within sporangia of a single plant, between spo- Tangia of a single plant, and between plants of a population was found to be More or less equivalent and negligible, variation between populations of Sapecte was demonstrated to be twice that between plants of a population. Microspores were found to be slightly less variable. Kott and Britton (1983) suggested that 150 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) spore size can be used to characterize species so long as a sample of at least 20 spores be measured from single plants of a representative set of populations. They demonstrated that, between species, both microspore and megaspore sizes varied directly with ploidy level. Some spor t analyses have yielded confusing results. Brown (1964) found that spores of the tetraploid Woodsia oregana var. cathcartiana were the same size or smaller than those of var. oregana, which is diploid. His results, however, are apparently a consequence of having mixed diploid and tetraploid cytotypes of variety oregana (see Windham & Haufler, 1986). More intriguing are the apparently inconsistent results for Dryopteris. Manton (1950) found it odd that spores of diploid Dryopteris abbreviata and tetraploid D. filix-mas were almost identical in size. Britton (1968) concluded from an analysis of eastern North American Dryopteris species that spore size and ploidy level are not closely correlated in the genus, based on documentation of relatively large spore sizes of the diploid species D. fragrans and D. dilatata (=D. expansa). He found that mean spore size of D. dilatata (=D. expansa) was indistinguishable from that of the tetraploids D. campyloptera and D. spinulosa. Wagner (1971) provided spore-length data for an expanded set of Dryopteris species in his survey 0 Appalachian wood ferns. His data corroborate Britton’s observation that there is no simple relationship between spore length and ploidy level in Dryopteris. The observations on Dryopteris by Manton, Britton, and Wagner suggest that sizes of diploid spores within a genus can be quite different. This variation in spore size among diploid species depends on at least three factors. 1) Size varies with adaptation for dispersal. Small size increases the likelihood of dispersal outside the region of the parent, but spore size increases on islands so that propagules can remain on the island rather than disperse into terrain that cannot host gametophyte development and function (Carlquist, 1966). 2) Spore size is also thought to increase for nutritional reasons (as in the evolution of mega- spores). 3) Cox and Hickey (1984) provided evidence that environmental param- eters strongly affect spore size in Isoetes storkii of Costa Rica. Their investigation of three populations whose habitats differed in mean temperature, mean daily solar radiation, and altitude demonstrated that plants in colder, shadier, higher habitats had smaller spores. This paper suggests that analyses of cell size vari- ation among species should, when possible, include a representative sample of populations from a set of characteristic habitats for each species. Hence, spore size is determined not only by ploidy level but by the reproductive biology and environmental regime of a taxon. Stomate size (usually as length of the guard cells) has also been correlated with ploidy level. Wagner (1954) noted that tetraploid members of the Appala- chian Asplenium complex had larger stomates than the diploids. Lovis an Reichstein (1968) provided data on stomate length for two sterile diploid hybrids and each of their derived fertile tetraploids. In each case the tetraploid stomates =e longer than those of the corresponding diploid, and there is little overlap of diploid and tetraploid stomate samples. Schneller (1974) reported stomate lengths for members of the D. filix-mas complex in Europe. He demonstrated a corre- lation between ploidy level and stomate size for diploid to pentaploid cytological Pee sir Mn RE ee a a eine Ammar arange Se ea Re ee Ce ee ee BARRINGTON ET AL.: SPORE AND STOMATE SIZE 151 races. Stomate measurements are particularly useful in assessing sterile hybrids, since irregular spores preclude spore comparisons. Counterintuitive results of stomate-size analysis have also been reported. Bar- rington (1986) reported stomate lengths of diploid Polystichum acrostichoides, tetraploid P. braunii, and their hybrid, which is known as both triploid and tetraploid. The order of these taxa based on stomate length (smallest to largest) was: tetraploid progenitor, triploid hybrid, diploid progenitor, tetraploid hybrid. Barrington suggested that P. braunii was derived from diploid Asian species with much smaller stomates than those of P. acrostichoides, but could not explain the anomalous guard-cell sizes of the hybrids. We provide an analysis of the anom- alous stomate sizes in this hybrid below. Experimental error in measurement has been a problem in the work with spore and stomate size. Factors affecting the measurement of cells include: 1) calibration of microscope; 2) characteristics of mounting medium; 3) definition of cell boundary; and 4) choice of cells for measurement. Lovis (1964) encoun- tered a 30% discrepancy between his data sets and those of Meyer in spore-size measurements of the European Asplenium trichomanes complex. Difference in mounting medium accounted for 6% of the difference: Lovis found that his spores, in gum chloral, yielded measurements 6% higher than Meyer’s spores of the same taxon, which were measured dry. Lovis also determined that Meyer's Measuring all spores versus his measuring only those whose long axis was per- pendicular to the line of sight accounted for a further 1.5-3.5% discrepancy. Nevertheless, he was unable to account for a residual discrepancy of over 20%, which is equivalent to the difference between a diploid and a tetraploid taxon. Lovis’s efforts are evidence that caution must be used in comparing spore mea- surements made by different investigators or those made with different equip- nt. Analysis of data on spore and stomate size has not usually included statistical analysis. Selling (1944) applied statistical techniques to problems of spore size, but his excellent example was not followed until recently. Analyses have seldom included simple descriptive statistics such as means and standard deviations, let alone tests of the significance of spore-size differences among related taxa. € provide new spore and stomate measurement data for three groups of ferns in three different families in order to assess the value of cell measurements in making systematic and evolutionary inferences from herbarium specimens. The methods are not equivalent, nor are the analyses: these data sets are inde- Pendently brought to bear on the general problem of cell size and fern system- atics. In Adiantum, we document herbarium specimens of a new tetraploid taxon from eastern North America. In Polypodium, we explore the problem of allo- Polyploid versus autopolyploid origin of tetraploid P. californicum. In Polysti- chum, we suggest a possible candidate for a missing diploid progenitor in a new tropical montane reticulate complex. ADIANTUM IN NORTHEASTERN NORTH AMERICA Recent systematic work on the Adiantum pedatum complex in eastern North America has demonstrated that there are three divergent entities there: A. pe- 152 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) TABLE 1. Spore Size (um) in Adiantum pedatum. No. Taxon Mean length (s.d.) Observed limits N sporophytes Woodland 2n 37.03 (2.494) 31.8-42.0 200 8 Serpentine 2n 42.96 (4.416) 32.0-52.8 385 16 Tetraploid 51.36 (5.162) 35.2-68.8 549 22 datum ssp. pedatum, a diploid of rich, deciduous woodlands; ssp. calderi, a diploid of serpentine substrates (formerly var. aleuticum; see Cody, 1983); and an unnamed tetraploid taxon of serpentine outcrops in north-central Vermont (Paris, 1986; Paris & Windham, in prep.). Although these entities have several diagnostic electrophoretic markers, they are not easily distinguished by their structural characters, most of which overlap considerably between taxa. Phe- notypic plasticity further obscures species boundaries. Especially difficult to dis- criminate are the diploid and the tetraploid plants on serpentine, in part because the tetraploids, when growing in exposed areas, resemble their serpentine dip- loid progenitors quite closely. Because tetraploid plants have consistently larger spores than diploids, spore size can be used effectively to differentiate the two serpentine taxa. The predictable relationship between spore size and ploidy level in A. pedatum has made it possible to survey herbarium collections from serpentine areas within the Appalachian Mountain ultramafic belt in order to define the range of the new tetraploid taxon. Materials and methods.—Spores were mounted in Hoyer’s medium on glass slides and measured at 400 using an ocular micrometer. Twenty-five spores per sporophyte were measured, each in its longest dimension as it was oriented on the slide. The perispore, thin and unornamented in this group, was included in the measurement. An analysis of variance was performed on the data using the BMDP7D program (Dixon, 1983). Independence of spore size and ploidy level was tested using the G-test of Sokal and Rohlf (1981). To test whether A. pedatum spores swell in Hoyer’s medium over time, a sample of 25 spores from each of two sporophytes was mounted in Hoyer’s medium and measured after two days and again after 85 days. Significance of the difference in sample means was tested using Student's t-test. ; Results.—Mean spore sizes for the three taxa were 37.03 um for the woodland diploid, 42.96 um for the serpentine diploid, and 51.36 um for the new tetraploid (Table 1). A G-test of independence of spore size and ploidy level, using spores from cytologically or electrophoretically documented sporophytes, yielded evi- dence that spore size and ploidy level are not independent in this group (Table 2). Thus spore size may be used as a reliable indicator of ploidy in the A. pedatum complex. An analysis of variance (Table 3A) demonstrated that the means among taxa are significantly different (P < 0.001). Not only are both dip- loids different from the tetraploid, but the diploids differ significantly from one another (P < 0.01), as demonstrated by a GT-2 test for multiple comparisons among pairs of means (Sokal and Rohlf, 1981; Table 3B). Non-parametric tests BARRINGTON ET AL.: SPORE AND STOMATE SIZE 153 ABLE 2. Test of Independence of Spore Size and Ploidy for Adiantum Taxa on Serpentine (N = 6 coal 25 spores/sporophyte). Spores = 47 um Spores > 47 um Row totals Diploid 125 25 150 Tetraploid 50 100 150 Column totals 175 125 300 G Statistic (with Williams’ correction for a 2 x 2 table) = 80.97*** f= P< 0.001. that do not assume homoscedasticity show equally significant differences. A t- test of the difference in the mean size of spore samples mounted in Hoyer’s medium for 2 days and for 85 days was insignificant. POLYSTICHUM IN CosTA RICA AND ANDEAN SOUTH AMERICA Evidence of a tropical montane polyploid complex reminiscent of well-known North Temperate complexes in the genus Polystichum now exists (Barrington, 1985a). The complex includes two unnamed species endemic to Costa Rica, an allotetraploid and one of its diploid progenitors. Morphologically, the best can- didate for the second progenitor is P. polyphyllum, a common polymorphic species of the Andes, Costa Rica, and Mexico. However, P. polyphyllum is tetraploid on counts from Costa Rica, Ecuador (Barrington, unpublished data), and Mexico (Smith in Barrington, 1985b). Spore length and width measurements were made to establish the relationship between spore length and ploidy level in Polystichum and search for a diploid cytotype of P. polyphyllum in northern Latin America. Included in the study set were four entities: 1) the endemic diploid from Costa Rica (ten sporophytes, documented to be diploid); 2) the TABLE 3. Tests Comparing Mean Spore Length for Adiantum pedatum Taxa. ee A. Analysis of variance Degrees of e Source of variation freedom MS —_Pource of variation Among taxa 2 17,694.1541 857.83*** Within taxa 1131 20.6267 B. GT-2 test for differences between means (pairwise comparisons) Difference between means (in um) Pair ; Lee 8 dates Woodland 2n vs. serpentine 2n Serpentine 2n vs. tetraploid call —_____ Woodland 2n vs. tetraploid 14.325 P< 6.01. "= P < 0.001. 154 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 3 (1986) TABLE 4. Spore Measurements (in um) for Polystichum Species (30 spores/sporophyte. Standard deviation is for sample means from grand mean.) Taxon Spore length (s.d.}) Spore width (s.d.) Spore volume (s.d.}) No. plants Endemic diploid 38.52 (1.41) 28.75 (1.02) 16,940 (1590) 10 Endemic tetraploid 46.20 (3.05) 33.27 (2.50) 27,620 (5430) 10 Costa Rica polyphyllum 40.14 (0.59) 29.20 (0.85) 18,110 (1270) 5 Andean polyphyllum 41.67 (5.18) 30.25 (3.36) 20,870 (6730) 8 endemic tetraploid from Costa Rica (ten sporophytes, documented to be tetra- ploid); 3) P. polyphyllum from Costa Rica (four sporophytes, documented to be tetraploid); and 4) P. polyphyllum from the northern Andes from Venezuela to Ecuador (ploidy unknown). Materials and methods.—Spore measurements of Polystichum species were made using a phase contrast microscope with a calibrated ocular micrometer at 400. Measurements comprised length and width of exospore wall of those spores whose long axis was parallel to the slide surface. Damaged spores and irregular spores were excluded from measurement. To assess the differential effects of mounting media on Polystichum spores over time, 20 spores from a single sporophyte (Barrington 1275, VT) of the en- demic tetraploid were mounted in Hoyer’s medium, Permount, and lactic acid on glass slides and were measured seven times over a two-week period. Spores in Hoyer’s medium swelled 15% in two weeks, but spores in the other two media remained unchanged. To characterize spore size for each of the taxa in the study group, 30 spores from each sporophyte mounted in Permount were measured. Volume of each spore was computed using the formula for an ellipsoid, V = LW’/6 (where L is the longest dimension of the exospore and W is width measured at right angles to L). Descriptive statistics were developed using the Minitab Statistical Package (Ryan et al., 1976). To test the hypothesis that spore length could serve as a predictor of ploidy level, independence of spore size and ploidy level was tested using the G-test of Sokal and Rohlf (1981). Results.—The mean spore lengths for the four sets of plants were 38.52 um for the endemic diploid, 46.20 um for the endemic tetraploid, 40.14 wm for P. polyphyllum in Costa Rica, and 41.67 um for Andean P. polyphyllum (Table 4). The mean volumes for these same taxa were 16,940 um*, 22,630 um’, 18,110 pm”, TABLE 5. Test of Independence of Spore Size and Ploidy for Polystichum Species (N = 10 spe phytes/taxon, 30 spores/ sporophyte). Oe Spores S 42.36 um Spores > 42.36 um Row totals an Rw Costa Rica diploid 277 23 300 Costa Rica tetraploid 50 250 300 Column totals 327 273 600 G Statistic (with Williams’ correction for a 2 x 2 table) = 393.24843*** 7

BARRINGTON ET AL.: SPORE AND STOMATE SIZE 159 LANG, F. A. 1971. The Polypodium vulgare complex in the Pacific Northwest. eR 21:235-254. LAWTON, E. 1932. Regeneration and induced polyploidy in ferns. Amer r. J. Bot. 1 4, LioypD, R. M. and F. A. LANG. 1964. The Polypodium vulgare complex in North eR Brit. Fern Vg G Lovis, J. D. 1964. The taxonomy of Asplenium trichomanes in Europe. Brit. Fern Gaz. 9:147-160. ———. 1977. Evolutionary patterns and processes in fern s. Advances Bot. Res. 4:229-415. and T. REICHSTEIN. 1968. Die zwei diploiden Ailes trichomanes x viride-Bastarde und ihre Fahigkeit ; zur spontanen Chromosomenverdoppelung. Bauhinia 4:53-63. Manton, I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge: Cambridge Univ. Press 957. The problem of Polypodium virginianum. Amer. Fern J. 47:129-134., and M. Suivas. 1953. Two sear eee forms of Polypodium virginianum in eastern North America. Nature mbes s10- 411 Paris, C. A. 106. Ab igation of the Adiant dat plex i stern North America. M.S. The esis, University « of Vermont, Burlington on. RANKER, T. A. aan. The Polypodium vulgare complex in Northern California. M.A. Thesis, Hum- boldt State peeipiuig picigge California. and M.R R. MESLER. 1982. The origin of tetraploidy in bic peice californicum: an au- topos ome Bot. Soc. America, Misc. Publ. 162:77 (abstra ct). Ryan, T. A. Jr., L. Jomner, and B. F. RYAN. 1976. Minitab siailets handbook. Boston: PWS Pub- lish ishers. _ J. 1974. Untersuchungen an einheimischen sagsies insbesondere der Dryopteris filix- as-Gruppe. 1 Teil. Ber. Schweiz. Bot. Ges. 84:195-21 SELLING, 0 H. 1944. Studies in the recent and fossil seen ae Hehe with particular reference o their aioe avian iets — pppoe 16: 1-1 SHIvAs, te G. 1961. Contributions to the cytol ogy of species of Polypodium in Europe and America. II. Taxonomy. J. Lin inn. Soc., Bot. 58 7-38. SINNOTT, me W. 1960. Plant morphogenesis. New York: MeCraw Hl KAL, R. R. and F. J. ROHLF. 1981. Bio ometry. San Francisco: W. H. Freeman & Co. STeBBINs, G. L. 1950. Variation and evolution in plants. New York: Columbia Univ. Press. Wacner, W. H. Jr. 1954. Reticulate evolution in the Appalachian aspleniums. Evolution 8:103-118, oo 1966 1971. Evolution of Dryopteris in relation to the Appalachians. Pp. 147-191 in The dis- tribestional history of the biota of the southern Appalachians. Part II. Flora, ed. P. C. Holt. Research Division Monograph 2. Virginia Polytechnic Institute and State University. Blacks- urg, Virginia. WINDHaM, M. D. and C. H. HAUFLER. 1986. Biosystematic uses of fern gametophytes derived from ium specimens. Amer. Fern J. 76:114-128. APPENDIX To calculate the predicted length eae for the hybrids, haploid length values for the progenitor genomes were derived by multiplying the observed guard-cell lengths by 1/1.26 (for the diploid Species) and 1/1.26? (for the tetraploid cua 1/1.26 is the ratio of diameters of two spheres, the first having half the volume of the second. Then the peated cnn was computed by multiplying the = of the haploid values for h nome—{a +b + b)/4 for the tetraploid cytotype and (a + b + b)/3 for the triploid cyto type— —by a eset he increase in diameter ©xpected for a given ploidy level of the hybrid (1.26”? for the triploid, 1.26? for the tetraploid). INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable), including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 812 by 11 inch paper of good quality, not “erasable” paper. Double space manuscripts throughout, including title, authors’ names and addresses, text (including heads and keys), literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author's name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed pages. Avoid footnotes and do not break words at ends of lines. Make table headings and figure captions self-explanatory. Use S.I. (metric) units for all mea- sures (e.g., distance, elevation, weight) unless quoted or cited from another source (e.g., specimen citations). For nomenclatural matter (i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- breviate titles of serial publications according to Botanico-Periodicum-Huntian- um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as part of nomenclatural matter are not included in lit- erature cited. For shorter notes and reviews, put all references parenthetically in text. Use Index herbariorum (Regnum Veg. 106:1-452. 1981) for designations of herbaria. Illustrations should be proportioned to fit page width with caption on the same page. Provide margins of at least 25 mm on all illustrations. For continuous-tone illustrations, design originals for reproduction without reduction or by a uniform amount. In composite blocks, abut edges of adjacent photographs. Avoid com- bining continuous-tone and line-copy in single illustrations or blocks. Coordinate sequence and numbering of figures (and of tables) with order of citation in text. Explain scales and symbols in figures themselves, not in captions. Include a scale and reference to latitude and longitude in each map. Proofs and reprint order forms are sent to authors by the printer. Authors should send corrected proofs to the editor and reprint orders to the printer. Authors will be assessed charges for extensive alterations made after type has been set. For other matters of form or style, consult recent issues of American Fern Journal and The Chicago manual of style, 13th ed. (1982. Chicago: Univ. Chicago Press). Occasionally, departure from these guidelines may be justified. Authors are encouraged to consult the editor for assistance with any aspect of manuscript preparation. Papers longer than 32 printed pages may be sent to the Editor of Pteridologia (Memoir Editor, see cover 2). AMERICAN a FERN se October-December 1986 JOURNAL ee else QUARTERLY JOURNAL OF THE AMERICAN FERN SOCIETY The Neotropical Fern Genus Olfersia Robbin C. Moran Trichomanes in Florida Clifton E. Nauman Some New Names and Combinations in Pteridaceae Rolla Tryon A Novel Method for Surface-sterilizing and Sowing Fern Spores Thomas R. R. Warne, Gury L. Walker, and Leslie G. Hickok Shorter Note Osmunda cinnamomes fi frondosa in the Coastal Plain of ia and Florida Richard Carter and Wayne R. Faircloth Review Referees, 1986 Index to Volume 76—1986 161 187 The American Fern Society Council for 1986 FLORENCE S. WAGNER, Dept. of Botany, University of Michigan, Ann Arbor, MI 48109. President JUDITH E. SKOG, Biology Dept., George Mason University, clei A 22030. Vice-President W. CARL TAYLOR, Sativesukes Public Museum, Milwaukee, WI Secretary JAMES D. CAPONETTI, Dept. of Botany, University of Tennessee, Kn ville T™N separ oe rer DAVID S. BARRINGTON, Dept. of Botany, University of Vermont, Burtiagion, VTO aes r JAMES D. MONTGOMERY, Ecology III, R.D. 1, Berwick, PA 1860: Back Issues Curator ALAN R. SMITH, Dept. of Botany, University of California, Beka, = 94720. Journal Editor DAVID B. LELLINGER, Smithsonian Institution, Washington Memoir Editor DENNIS Wm. STEVENSON, Dept. of sores Sciences, “ta Cae Columbia University, New York, NY 1 Fiddlehead Forum Editor aa Fern Journal EDITOR ee ee Dept. of Botany, ape of Ca pee erkeley, CA 94 BORE : ASSOCIATE EDITORS 1 (GERALD ].GABTONY ......>.....: Dept. of Biology, Indiana amare iad Bloomington, IN 47401 RIPPER IAUPLER 8 Dept. of Botany, — of Kansas, : as : Be oe . = ae ashington, DC 20560 . TERRY R. WEBSTER . . Biological Sciences Group, University of PRD Storrs, CT 06268 — The “American Fern journal fe eranns 0002-8444) is an illustrated quarterly devoted to the general study of ferns. It y the American Fern Society, and published at the Pringle Herbarium, Ur VT 05405. Second-class postage me at Burlington, VT, and AG sh. (£. aoe +1, date of ieaite it A sere and orders for back issues should be addressed to Dr. James D. Montgomery, Ecology Ill, R.D. 1. _ Berwick, PA 18603. — of address, dues, and applications for membership should be sent to the Records Trea- : Subanon 200 pom $11.50 net if paid through an jaseny tases fee $0.50); sent free to : the American Fern Society {annual dues, $10.00 + $4.00 mailing surcharge beyond 110-1978 $5.0 to $6.25 each single back numbers of 68 pags or ls, $1.25: 68-00 : 80 00 ach pla shipring. Ten pee Back volumes 1979 et seq. $8.00 each; — res fo Anica Fe Joon Dept. of Botany, Tbe 9 ae Ten Percent discount on cedens of sie vohenen ae : American Fern Journal 76(4):161-178 (1986) The Neotropical Fern Genus Olfersia ROBBIN C. MORAN Illinois Natural History Survey, 607 E. Peabody Dr., Champaign, Illinois 61820 Olfersia Raddi (Dryopteridaceae) consists of a single species, O. cervina (L.) Kunze, which is widely distributed in forests of the neotropics. This species differs from all other dryopteroid ferns by its numerous, fine, parallel veins that connect at their tips by a submarginal vein. Other features that distinguish Ol- fersia from many dryopteroid ferns are its strongly differentiated sterile and fertile leaves, simply pinnate sterile lamina, entire pinnae, and conform apical pinna. Olfersia is morphologically most similar and cladistically most closely related to Polybotrya, a genus that also has strong sterile-fertile leaf dimorphism. Most recent pteridologists have placed Olfersia in Polybotrya; however, Olfersia differs by the characters noted above and lacks the unique stem anatomy of Polybotrya. This paper, an outgrowth of my work on Polybotrya (Moran, 1986), is based on a study of about 550 herbarium sheets that represent approximately 335 in- dividual collections from 29 herbaria (see Acknowledgments). I spent seven months collecting Olfersia and Polybotrya in Costa Rica, Ecuador, Peru, and Venezuela and was able to study living plants of Olfersia in each of those coun- tries. Important observations were made about the biology of Olfersia that would have been impossible to discern from herbarium specimens, such as the duration of fertile leaves on the stem, the orientation of sterile vs. fertile leaves, and how often the plants were terrestrial or became scandent. Herbarium, cytological, and anatomical material were also collected during field work. TAXONOMIC HISTORY Olfersia cervina has received various treatments from pteridologists. It was first described by Linnaeus (1753), who placed it in Osmunda, at that time a polyphyletic genus. His specific epithet cervina means “deer,” and he may have been alluding to the fancied resemblance of the velvet ona deer s antlers to the brown sporangia that cover the fertile pinnules of Olfersia. Linnaeus never saw a specimen of this species, but knew of its existence from the excellent engrav- ings of Plumier (1705) and Petiver (1712). Swartz (1806) placed this species in Acrostichum because he recognized that the sporangial capsules have a vertical, _* bowlike annulus (his group “Gyratae”), unlike true Osmunda, which has only a _ lateral patch on one side of the capsule wall (his group “Spurie Gyratae”). Swartz's genus Acrostichum, consisting of all ferns having the sori spread across the abaxial surface of the leaf and lacking an indusium, was also a large, polyphy- letic group. Raddi (1816) described the genus Olfersia, honoring Dr. Ignaz Fraz epee von Olfers (1793-1871), a botany professor from Rio de Janeiro sign . - probably met during his travels with the Austrian Expedition to Brazil (1817- MISSOURI BOTANICAL FEB 5 1987 162 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) 1818). While in Rio de Janeiro, Raddi collected the type species of his new genus, O. corcovadensis, at Mt. Corcovado. But Raddi never compared Olfersia with the closely related genus Polybotrya or with A. cervinum, suggesting that he was unaware of their close relationship. He distinguished Olfersia from other fern genera by its fertile pinnae that are entire and have sporangia on both surfaces. Kaulfuss (1824), apparently unaware of Olfersia corcovadensis, placed A. cer- vinum in Polybotrya, a genus described by Willdenow (1810). He made this decision because Polybotrya was said to have sporangia that cover both surfaces of the pinnules—the condition that was thought to occur in A. cervinum. Based on this criterion, Sprengel (1827) put O. corcovadensis in Polybotrya, alongside A. cervinum. Apparently, prior to Sprengel’s work it was not realized how closely these two plants were related. Link (1833) also thought that O. corcovadensis should be placed in Polybotrya, but provided the name P. raddiana—a super- fluous name for P. corcovadensis (Raddi) Sprengel. Pres] (1851) carefully studied the fertile leaves of O. cervina and concluded that, unlike O. corcovadensis, the sporangia do not occupy both surfaces of the fertile pinnules. Thus, he created the genus Dorcapteris to accommodate O. cervina. Unfortunately, he included in Olfersia many unrelated species of Elaphoglossum (Lomariopsidaceae) and Stenochlaena (Blechnaceae). Throughout the first half of the 1800s, pteridologists maintained cervina and corcovadensis as distinct. Finally, Hooker (1864) observed that intermediates exist in the shape of the fertile pinnules, the main character that had been used to separate the two plants. Accordingly, he placed the two plants in synonymy, using the name A. cervinum. Although later pteridologists have agreed that cervina and corcovadensis are the same species, they have differed in their generic placements. Most have followed Christensen (1905), who placed O. cer- vina in Polybotrya, but Smith (1875), Brade (1971), and Pichi Sermolli (1977) have maintained Olfersia as a distinct genus. GEOGRAPHY Compared with most other dryopteroid ferns, Olfersia has an extensive range. Polybotrya, for example, with its 35 species, is essentially coextensive with O. cervina—both occur from Oaxaca/ Veracruz, Mexico, southward through Cen- tral America, from northern South America to Bolivia and Rio Grande do Sul, Brazil, and from Cuba to Grenada in the West Indies (Fig. 1; Moran, 1986). Olfersia inhabits wet, shaded, tropical forests from steamy rain forests at sea level to cool cloud forests up to 2000 m. It is most common in primary forests and rarely occurs in disturbed habitats. Its distribution correlates generally with the moist montane regions of tropical America. F, ew collections have been made in areas of less topographic relief, such as the Matto Grosso, the Brazilian Am- azon, and the Guyanan Highlands. Olfersia occurs on only one oceanic island—Cocos Island in the Pacific Ocean, about 500 kilometers (310 miles) southwest of Costa Rica. This finding supports Tryon’s (1970) contention that the ferns of oceanic islands tend to be those that R. C. MORAN: OLFERSIA 163 Fic. 1. Geographic distribution of Olfersia cervina. are widespread beyond the source area for the island (the source area for Cocos Island is defined by Tryon as Central America and Colombia). Widely distrib- uted species have a broader ecological amplitude and therefore establish them- selves more often on oceanic islands because of their ability to grow under a wide range of environments (Tryon, 1970). COMPARATIVE MORPHOLOGY AND ANATOMY Stem.—The stem of Olfersia is short-creeping with internodes generally 1-3 cm apart. Although Olfersia is frequently terrestrial, it has a tendency to grow on logs and to climb up tree trunks for distances up to 1 m. Since its stem is of limited growth, it never becomes high-climbing like most species of Polybotrya. Consequently, it makes an excellent pot plant (Walker, 1985a). Stem anatomy is the most important difference between Olfersia and Poly- 164 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) botrya. The meristeles of Olfersia are not readily seen in cross section because each meristele is surrounded by whitish, parenchymatous cells of the ground tissue and not by a sheath of dark sclerenchyma fibers as in Polybotrya (Fig. 2a, b). The arrangement of meristeles around the center of the stem is irregular, not even and circular as in Polybotrya. The meristeles also have irregular shapes that vary from round to oblong to slightly lobed. The epidermis and hypodermis consist of dark, thick-walled fibers that in cross section give the appearance of a black ring encircling the stem. This ring con- trasts sharply with the whitish parenchyma of the underlying ground tissue, which contains numerous amyloplasts. Embedded in the ground tissue are myr- iad clusters of brachysclereids, each having dark cell walls with numerous pits. These darkened clusters also contrast sharply with the whitish ground paren- chyma and are perhaps the most conspicuous feature of a freshly cut stem. Internal glandular hairs such as those that occur in the ground tissue of stems and petioles of many dryopteroid ferns (Mehra & Mittal, 1961) are apparently lacking in Olfersia. Stem scales.—Golden or yellowish scales, which usually darken upon drying, densely cover the stems of Olfersia. They lie slightly appressed to the stem and are linear with an attenuate tip. The abruptly truncated base is attached along a narrow, slightly curved line, not at a single point. The scales are thin and membranaceous, and the cell outlines are easily visible. The margins are usually denticulate, although some scales appear entire, except for a few minute teeth near the apex. The teeth are always composed of the projecting end walls from two adjacent cells (Fig. 2e). The scales of Olfersia are similar to those of Poly- botrya osmundacea and P. polybotryoides. Roots.—Olfersia produces numerous, dark, tough, fibrous roots from all sides of the stem; these bear no positional relationship to the leaf bases. Anatomically, the roots of Olfersia are like those described for Polybotrya by Moran (1986). Petiole.—A cross section of an Olfersia petiole reveals vascular bundles ar- ranged in a mushroom-like shape, with the base of the mushroom oriented adax- ially (Fig. 3b). A similar pattern also occurs in Cyclodium trianae (pers. obs.), Polybotrya (Moran, 1986), Maxonia (Walker, 1972), and in some species of Lo- mariopsidaceae (Holttum, 1978). Further study of petiole anatomy in dryopteroid genera may provide helpful information for assessing relationships. Scales occur on the petiole bases of Olfersia. Although similar to those on the stem, these scales are usually wider and shorter. The petiole base in some ferns is specialized to accumulate food and to persist as a storage organ long after the rest of the leaf to which it was attached has withered and decayed. These specialized petiole bases, called “‘trophopods,” are a new source of systematic data in ferns (Wagner & Johnson, 1983). Olfersia, however, lacks trophopods, and the food-storing function is performed by the — which contains abundant starch grains in the cells of its ground paren- chyma. Rhachis and costae.—In fresh material, the rhachis and costae of Olfersia are rounded or shallowly grooved adaxially. Two linear, yellowish, aerophores occur on the sides of the rhachis and petiole. When dry, the large, parenchymatous R. C. MORAN: OLFERSIA 165 0.5mm Fic. 2. Various features of Olfersia cervina. a, cross section of the stem; b, cross section of a Polybotrya stem for comparison; c, fertile leaf with entire pinnae; d, fertile leaf with pectinate pinnae; e, close-up of stem scale margin; f, reduced, uniseriate scales from the costa and laminar tissue. a, d, e, f: Moran 2181 (F). b: Moran 2182 (F). c: Brackenridge s.n., U.S. Exploring Expedition (PH). AP, aerophore; LT, leaf trace; MS, meristele; RT, root trace; SC, sclereid cluster; SH, scleren- chyma sheath surrounding meristele. 166 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Fic. 3. Sterile leaves of Olfersia cervina. a, mature fertile leaf; b, cross section of petiole base owing vascular bundles arranged in a m shape; c, close-up of rhachis-pinnae jucture; d, e: juvenile plants with precociously fertile leaves: f, pinnae showing venation, note submarginal connecting strand. a, b, c, f: Moran 2181 (F). e, d: Antonio 3715 (MO). R. C. MORAN: OLFERSIA 167 cells in the interior of the rhachis collapse, indenting the adaxial surface of the rhachis. The two aerophores on each side of the rhachis and petiole (Fig. 3b) also collapse, since they are composed of soft walled, parenchymatous cells. As a result, the rhachis and costae become trisulcate. The costae, when dry, may remain rounded or become shallowly grooved. These grooves, however, fade out before they reach the rhachis so that the costal grooves are not continuous with the rhachis grooves. Since in other genera of dryopteroid ferns the grooves are decurrent into one another (Holttum, 1984), the non-decurrent grooves of Olfersia are interpreted as derived. Leaf apex.—Olfersia has a conform apex similar to the lateral pinnae (Figs. 2c, d, 3a). Polybotrya, however, has a pinnatifid apex even in its simply pinnate species. Two species of Polybotrya, P. espiritosantensis and P. polybotryoides, have subconform apices, but these species are clearly derived within the genus and do not indicate a relationship with Olfersia (Moran, 1986). The leaf apex, therefore, affords a convenient distinguishing trait between the two genera. Venation.—Olfersia has numerous, long, fine, parallel veins connected at their tips by a submarginal strand (Fig. 3f). The veins usually branch once or twice near the costa and then continue unbranched to the connecting strand. All the veins are approximately the same width and do not form distinctly pinnate groups. The only species of Polybotrya having venation remotely similar to that of Olfersia are P. fractiserialis and P. sorbifolia, but they lack a submarginal connecting strand, have more branched veins, and form distinctly pinnate groups. The venation of Olfersia is not only unlike that of Polybotrya but unlike that of any dryopteroid genus. This dissimilarity emphasizes the distinctness of Olfersia. Hairs.—The leaf of Olfersia is usually described as glabrous; however, the abaxial surface is sparsely beset with inconspicuous, slender “hairs” 0.1-1.2 mm long (Fig. 2f). Usually seen on young, incompletely expanded leaves and along the costa of mature leaves where the lamina has been protected from abrasion, these tiny, uniseriate structures are “hairs” by definition only; developmentally they are scales. Evidence for their scalelike nature comes from comparison with such other dryopteroid genera as Polybotrya, Dryopteris, and Arachniodes, in which the “hairs” intergrade with scales. In these genera, the “hairs” at the apex of the costa are tiny, less than 0.1 mm long, two to fifteen celled, mostly appressed to the surface, and have reddish cross walls. Each cell is slightly flattened and the apical cell of the hair is conform, unlike the true hairs of dryopteroid ferns. Towards the middle of the costa these hairs elongate by increasing the number of cells, and some become two to three seriate near the base, that is, they become a scale. On the rhachis and base of the costa, the scales become large and polyseriate. Every intermediate form can be found; hence the evidence that the hairs are actually reduced scales. I suggest that these tiny, uniserlate scales be called ‘‘proscales.” : : Like most species of Dryopteris (Christensen, 1913}, Olfersia lacks true hairs. True hairs in dryopteroid ferns do not intergrade with scales and are erect or spreading—never appressed. True hairs are whitish, cylindrical, and thicker walled than the proscales. Their apical cells are usually of a different shape than the rest of the cells, often being acicular or acute (Moran, 1986, fig. 10). 168 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Fertile leaves.—Olfersia produces strongly differentiated sterile and fertile leaves. The fertile leaves are like a skeleton of the sterile leaves because the green, photosynthetic tissue has been reduced to the axes as a narrow wing. Such extreme dimorphy is termed “holodimorphy” (Wagner & Wagner, 1977) and involves many characters, not merely the reduction of the green lamina. In Olfersia, the sterile and fertile leaves also differ in size, dissection, orientation, venation, and duration on the stem. Since Olfersia has the same syndrome of foliar dimorphy characters as that described for Polybotrya by Moran (1986), only the most conspicuous differences will be mentioned. The sterile and fertile leaves differ in size, with the laminae of the fertile leaves smaller than those of the sterile. The total length of the fertile leaves, however, is greater than that of the sterile because of the elongated petiole, which is generally twice the length of the lamina. The two terrestrial species of Polybotrya are like Olfersia in having taller fertile leaves. The climbing species of Polybotrya, however, have fertile leaves that are shorter than the sterile leaves. Presumably, taller fertile leaves are advantageous for terrestrial plants since the sporangia are borne at a higher position where drying occurs more readily and the spores are more likely to be picked up by air currents before they hit the ground. The orientation of sterile and fertile leaves also differs. Sterile leaves, which do most of the photosynthesis, recline with age, orienting themselves perpen- dicularly to the sun’s rays. This position would be advantageous in a shaded, tropical forest where sunlight is reduced. Fertile leaves remain erect, however, which presumably promotes drying of sporangia. Differences in the orientation of sterile and fertile leaves typically occur in ferns that exhibit dimorphy (Wag- ner & Wagner, 1977). This correlation suggests that strong selective pressures exist for differences in sterile-fertile leaf orientation. After shedding their spores, fertile leaves quickly wither. Rapid senescence probably contributes to the plant's economy since large quantities of carbohydrates and mineral nutrients are not spent to maintain a structure that lacks appreciable photosynthetic surface. As a corollary to their ephemeral character, the fertile leaves are of “cheaper” construction, that is, they have little of the rigid support collenchyma found in sterile leaves. Seasonal production of fertile leaves panies sterile-fertile leaf dimorphy in some ferns, but seasonality does not appear to be present in Olfersia. On the Atlantic coastal plain of Costa Rica, I saw populations of Olfersia that contained plants in various stages of fertile leaf production; some plants had fertile leaves in the crozier stage, others had fertile leaves in the spore-releasing stage, and in some individuals the fertile leaves had died and withered. This observation Suggests continuous rather than seasonal production. Dr. Henk van der Werff, working in the state of Falcon, Venezuela, also found that Olfersia produced fertile leaves throughout the year (pers. comm.}. The question of seasonality in fertile leaves, however, needs further study. Olfersia produces fertile leaves freely on both terrestrial and climbing stems. In contrast, the climbing species of Polybotrya almost always produce fertile leaves on the scandent portion of the stem. Production of fertile leaves on the R. C. MORAN: OLFERSIA 169 terrestrial stem of a climbing Polybotrya is extremely rare; I know of only three examples among the thousands of plants I have seen in the field and in herbaria. This variability in the site of fertile leaf production in Olfersia, compared to its inflexibility in Polybotrya, is an important behavioral character differentiating the two genera. Fertile pinnae.—Two extreme forms of fertile pinnae occur in Olfersia. The first type is pectinate (Fig. 2d), and sporangia are produced only on the abaxial surface of the leaf. When dry, however, the adaxial surface retroflexes so that sporangia appear to occupy all sides of a now cylindrical pinnule. The second type of pinnae is entire and has sporangia on both surfaces (Fig. 2c; amphiacrosti- choid). The pectinate type prevails throughout most of the range, but the am- phiacrostichoid type is common in southeastern Brazil, where it accounts for approximately 80% of the specimens. Although the two types of pinnae differ greatly in appearance, they may occur on the same plant and therefore only represent forms. Accordingly, the amphiacrostichoid type, which has been named O. corcovadensis, is placed in synonymy. Diplodesmic veins.—The sorus of Olfersia contains a special vascular supply that lies just beneath the sporangia but above the normal veins, so that a cross section of the sorus reveals two sets of veins at two levels (Fig. 4c, d). This condition, termed ‘“‘diplodesmic,” is also found in Polybotrya. Diplodesmic veins, in contrast to the normal veins, are very delicate, obscure, and abundantly anas- tomosed. The extensive anastomoses connect infrequently with the normal veins, which are usually thick, prominent, and rarely branched. Diplodesmic veins are apparently homologous with the veins of the sterile leaf. This homology can be seen in transitional (part sterile, part fertile leaves) where the tips of the sterile veins grade into the diplodesmic veins in the fertile parts (Fig. 4c, d). The leaves need to be cleared to see this gradation because the receptacle darkens upon drying and obscures the faint diplodesmic veins. The normal veins sometimes anastomose irregularly in transitional part sterile, part fertile leaves (Walker, 1985a). I interpret this phenomenon as an early manifestation of the develop- mental pathways that convert the sterile veins into the reticulated diplodesmic veins. Sporangia.—The sporangia of Olfersia resemble those described by Moran (1986) for Polybotrya with the exception of the occurrence in Olfersia of a single- celled gland near the base of the paraphysis. Plants from southeastern Brazil lack these glands, which occur in plants from other portions of the range (Fig. 4f, g). The function of these glands is unknown. Spores.—The spores of Olfersia are monolete and generally 46-57 pm, rela- tively long compared with the spores of most other dryopteroid ferns (Tryon & Tryon, 1982). The exospore is smooth and overlain by a prominently folded perispore with a sparsely echinate surface. This sculpturing resembles that of P. caudata, although most species of Polybotrya have densely echinate perispores (Moran, 1986). Tryon and Tryon (1982, fig. 80.30) have a photograph taken with the scanning electron mi of an O. cervina spore from Guatemala. Spores of plants from other Latin American countries show the same pattern of sculp- turing (pers. obs.). 170 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Fic. 4. Various anatomica eas I characteristics of Olfersia cervina. a, abaxial epidermis; b, odeaial OF dermis; ¢, d, e, diplo ins of fertile pinnules, c and e show how the vein tips t 1c I q € show how the |S ges to form the diplodesmic veins; g, h, sporangial stalks with glandular cell and paraphyses; 1, J ae gt rangial stalks from southeastern Brazil lacking paraphyses and glandular cell. c: Jurgens s.n. (P). b, d: Moran 2184 (F). e: Glaziou 450 (P). f, g, h: Smith 1284 (UC). i, j: Mexia 4096 (UC). CHROMOSOME NUMBERS Three cytological studies report Olfersia as a sexual diploid. Walker (1966, fig. 40) first noted the chromosome number for Olfersia cervina as n = 41 from a plant collected in Jamaica. Smith and Mickel (1977) also found n = 41 for a plant collected on Trinidad and observed micronuclei present at meiotic teleophase R. C. MORAN: OLFERSIA 171 as in Elaphoglossum. Eight more plants of Olfersia from Trinidad were cytolog- ically examined by Walker (1985b), and all showed n = 41. This base chromo- some number, along with morphological and anatomical features, relates Olfer- sia to the dryopteroid ferns. RELATIONSHIP TO POLYBOTRYA After studying the drastic changes in venation of what he called “anomalous” leaves, Walker (1985a) recently concluded that Olfersia is a simplified, derived member of Polybotrya. These leaves are cut to various degrees between the ample sterile leaves and the skeletonized fertile leaves. As the lamina becomes more reduced and divided, that is, more like the fertile leaves, the veins shorten and anastomose, forming a reticulum (see Walker’s figs. 3-5). The pattern of this reticulum is similar to the venation of Polybotrya subgenus Soromanes, and Walker’s suggestion that O. cervina is derived from Polybotrya is based upon this similarity. Walker's “anomalous” leaves, however, actually represent leaves that are part sterile and part fertile. This transitional nature is the key to understanding why reticulate veins develop, as well as their taxonomic significance. The fertile leaves of Olfersia contain a highly specialized vascular supply to the sporangia— the diplodesmic veins, which form an elaborate network of delicate veins anas- tomosing irregularly beneath the sporangia (Fig. 4c, d, e). This special venation is derived from the veins of the sterile leaf, that is, the vein tips of the sterile leaf are converted into the diplodesmic veins (Fig. 4c, e). Thus, in a develop- mental sense, the veins of the sterile leaf and the diplodesmic veins of the fertile leaf are “homologous.” This homology refutes the hypothesis that reticulate veins in part sterile, part fertile leaves of Olfersia indicate a relationship with Poly- botrya subgenus Soromanes. Instead, reticulate veins in part sterile, part fertile leaves merely express a stage (albeit precocious) in the formation of reticulate, diplodesmic veins. Pteridologists have often placed Olfersia in Polybotrya because both possess holodimorphic sterile and fertile leaves and creeping stems that are densely covered by scales. My research revealed additional similarities between the two genera (Table 1). But despite these similarities, important differences exist. Per- haps most significant is that Olfersia lacks the unique stem anatomy of Polybo- trya. Moreover, Olfersia differs in two features of the lamina: the conform ter- minal pinna and venation unlike that of any dryopteroid fern. These differences justify keeping Olfersia distinct from Polybotrya. Keeping these two genera distinct, however, does not resolve the question of their cladistic relationship. Two cladistic hypotheses seem probable (Fig. 5, Table 2). Hypothesis B, that Olfersia was derived from Polybotrya, requires that ~ unique stem anatomy of Polybotrya—the scl hyma sheath g each meristele and the even, circular arrangement of meristeles—evolved and then was lost (Fig. 5, characters 9 & 10). It also requires a reversal in stem habit, from terrestrial and short creeping in the ancestor to long ee. Polybotrya, back to short creeping and partially terrestrial in Olfersia (Fig. 5, character 8). 172 TABLE 1. AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Comparison of Polybotrya and Olfersia. SIMILARITIES Stem indument: Vascular bundle arrange- ment in the petiole: Sterile-Fertile leaves: ‘araphyses: Sporangial stalks: Indusium: Densely scaly Mushroom-shaped Holodimorphic Present, unbranched, uniseriate Two-rowed except three-rowed immediately below capsule Absent Laminar scales: Ha : Fertile and sterile leaf length: Fertile leaf production: Seasonality of fertile leaves: Perispore sculpturing rounding meristele: Stem habit: Proscales and polyseriate scales Generally present Fertile leaf shorter than the sterile, but longer in the two terrestrial species Only on scandent portion of em Seasonal and continual Usually spiny, rarely smooth Circular Present High-climbing (two spp. ter- restrial) Diplodesmic veins: Present Chromosome number: = 41 DIFFERENCES Polybotrya Olfersia Leaf apex: Pinnatifid Conform, similar to lateral pin- nae Venation: Various, but not as in Olfer- Numerous, fine, long and par- sia allel _ Submarginal connecting Absent (present in P. polybo- Present and: tryoides) Proscales only Absent Fertile leaf longer than sterile leaf due to lengthened peti- ole On scandent or terrestrial stems Continual throughout the year Smooth Irregular Absent Short-creeping, terrestrial or climbing for short distances In short, hypothesis B requires three reversals. Hypothesis A, that the two genera evolved from a common ancestor that lacked the unique stem anatomy but had dimorphic leaves, requires no reversals. Therefore, hypothesis A is favored. Moran (1986) presented evidence that Cyclodium is the most likely ancestor of oth genera. TAXONOMIC TREATMENT Olfersia Raddi, Opusc. Sci. Bologna 3:283 tab. 11., fig. b. 1819.—Acrostichum — Olfersia (Raddi) Hooker, Sp. Fil. 5:254. 1864.—Type: O. corcovadensis Dorcapteris Presl, Epim. Bot. 166. 1849.—Type: Osmunda cervina L. A monotypic genus, with the characteristics of its single species. R. C. MORAN: OLFERSIA 173 Olfersia Olfersia Polybotrya +} lnti ip of Olfersia and Polybotrya. ; d character states sh in Table 2. The open rectangles in hypothesis B represent character state reversals. Fic. 5. Alternative hypoth Th h d 7 Olfersia cervina (L.) Kunze, Flora 7:312. 1824. Osmunda cervina L., Sp. Pl. 2: 1065. 1753.—Acrostichum cervinum (L.) Swartz, Syn. Fil. 14. 1806.—Poly- botrya cervina (L.) Kaulfuss, Enum. Fil. 55. 1824.—Dorcapteris cervina (L.) Presl, Epim. Bot. 167. 1851.—LectoryPe: (chosen by Proctor, Fl. L. Antill. 2: 223. 1977): Plumier, Tract. Fil. Amer., pl. 154 (1705), illustrating a plant from inique. Plate 154 is mounted on herbarium sheets at F and GH Olfersia corcovadensis Raddi, Opusc. Sci. Bol. 3:283, tab. 11, fig. b. 1819.—Poly- botrya corcovadensis (Raddi) Spreng., Syst. Veg. 4:33. 1827.—Polybotrya raddiana Link, Hort. Berol. 2:134. 1833 (nom. superf. for P. corcovadensis).— TYPE: Brazil, Rio de Janeiro, Mt. Corcovado, Raddi (not seen, but Raddi’s tab. 11 certainly represents this species). Polybotrya cervina (L.) Kaulfuss forma transitoria Rosenstock, Hedwigia 46:156. 1907.—TypeE: Brazil, Santa Cruz, Piccado Nova, Jurgens & Stein 69 (NY!, UC)). Stem short-creeping, primarily terrestrial, occasionally short climbing, in cross- Section with numerous scattered dark sclereids, the meristeles irregularly shaped and arranged, not surrounded by a dark sclerenchyma sheath; scales linear, to 25 X 2 mm, yellowish to golden, often turning brown upon drying, the margins denticulate to nearly entire, the base truncate, attached along its length. Sterile leaves subdistant, usually 0.5-1.2 x 0.3-0.5 m; petiole shorter than the lamina, 20-50 cm long, scaly at base, with scales similar to those on the stem; lamina pinnate with a conform terminal pinna, thick, subcoriaceous, apparently gla- brous, but actually with minute, <0.1 mm long, appressed proscales; pinnae ovate-lanceolate to lanceolate, commonly 15-30 x 3-8 cm, 4-12 pairs, short-stalked, continuous (not articulate) to the rhachis, the margins entire, the base unequal, with the basiscopic side excavate and the acroscopic side broadly rounded, the apex acute to acuminate; veins forking near or at the base, parallel, ca. 1 mm 174 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) TaBLE 2. Characters and Character States Used in the Cladistic Analysis (Fig. 5) of Olfersia and Polybotrya. Character State Polarities were Determined by Comparison to Closely Related Outgroup Genera, such as Arachniodes, Cyclodium, Dryopteris, Maxonia, and Polystichopsis. Character Ancestral condition Derived condition 1. Sterile & fertile leaves: Monomorphic Dimorphic 2. Leaf dissection: Decompound (3-pinnate) 1-pinnate 3. Apex shape: Pinnatifid Conf. 4. Vein groups: Pinnate groups Not pinnate 5. Veins: Free Connected at tips 6. Sporangial stalks: Glabrous Paraphysate 7. Diplodesmic veins: Absent esen 8. Stem habit: Terrestrial and short-creeping Climbing and long-creeping 9. Sclerenchyma sheath sur- rounding meristeles: Absent Present 10. Meristele arrangement: Irregular Circular apart, slightly oblique to the costae, connected by a submarginal connecting vein {at times difficult to see in dried material); axes nearly glabrous, but with slender, inconspicuous, appressed, multicellular hairs (actually reduced scales), the cells often flattened with reddish cross walls, these hairs intermixing and intergrading with appressed, narrow, scattered scales; grooves shallow, glabrous, not decur- rent. Fertile leaves 1- or 2-pinnate, produced freely on terrestrial as well as scandent stems, more erect than the sterile leaves and taller due to the length- ened petiole; petiole longer than the petioles of the sterile leaves, usually 40-85 cm long; pinnae pectinate, linear, 7-20 cm long; pinnulus 0.5-1 x 0.1-0.3 mm, set at right angles to the costa, broadly adnate and joined by a narrow wing of tissue; sori exindusiate, spreading along the axes; sporangial stalks paraphysate and with a single glandular cell, the paraphysis hairlike, 3-6 celled, the apical cell clavate-glandular, but the stalks in plants from southeastern Brazil are eglan- dular and non-paraphysate; spores monolete, echinulate, (40)46-57(60) um. n = 41. N of Ixtlan de Juarez, trail E of Rte 175 at Campamento Vista Hermosa toward Ladi. Mickel 6384 (NY), 6399 (NY). Veracruz: Mpio. Hidalgotitlan, afluente del Rio Las Cuevas, Wendt et al. 3846-A (NY). Beize. Middlesex, Schipp 402 (F, GH, MICH, NY, UC, US, Z); Stann Creek, Mt. Cow Valley, Gentle 3528 (MICH). Guatema a. Alta Verapaz: Cubilquitz, von Tirckheim 418 (P, US) 4031 (GH, LIL}, 8044 (GH). Izabal: Near Puerto Barios, Standley 72823 (F). Petén: Dolores, km 85 W of Machaquila Road, Contreras 2648 (US). Quiché: Finca Chaila, “Zona Reyna,” Skutch 1805 (P); no locality, Skutch 1805 (GH). Prov. Unknown: no locality, Godman & Salvin 99 (BM). HONDURAS. co _— Ceiba, Dyer 208 (NY, US). Cortés: along Lake Yojoa, ca. 6 km N of Rancho Agua 155 (MO). Cocos Island: Chatham Bay, Jiménez 3157 (CR, F, NY. US). Limon: 7 km § of Bibri, Gémez llas José: 2.7 km SE of bridge over Rio Pacuare, 10 km SE of San Isidro, Wagner & Gémez 79637 (CR). Prov. Unknown: Navarito, Lankester 728 (F, US). Panama. i: vicinity of Gualaca, ca. 8.5 miles from Planes de Hornito, La Fortuna on road to damsite, Antonio 5068 (MO). é: Los Ped- R. C. MORAN: OLFERSIA 175 regales, ridge between Rio Blanco del Norte and Rio Cano Sucio, Knapp & Dressler 3786 (MO). Darien: Serrania de Pirre, 9-10 km due N of Alto de Nique, gi 37859 (MO). Veraguas: valley of Rio as Bocas on road between Alto Piedra (above Santa Fe) and Calovebora, Croat 27421 (MO). Cusa. Oriente: Montes United Fruit, Nicaro, Mayari, Acuna & — 19620 (US). Las Villas: banks of Yayabo River, Bauao Hills, Leon & Clement 3975 (NY). Santiago: Santiago, Linden 2188 (P). Jamaica. Portland: John Crow Mts., Crosby et al. 975 (MICH). St. Catharine: Mt. Diabolo, Hespen- heide 1396 (GH). St. Thomas; mountain trail between House Hill and Cuna Cuna Gap, Maxon 8918 (NY), 8976 (NY). Westmoreland: Cho-Cho Gully, 4.5 miles ESE of Bluefields, Wilson & Webster 507 (GH). Puerto Rico. Vicinity of Hollymount, Mt. D Diabolo, Maxon 2220 (NY); El Yunque, Scamman 6540 (GH); Luquillo National Forest, La Mina Falls trail, Howard et al. 15570 (GH). Harri. Massif du Nord, Morne Rhesnau, Ekman 4394 (US). Dominican ReEpus.ic. Barahona: Montiada Nueva, SE of Polo, Howard & Howard 8538 (BM, GH, NY, US). San Francisco de Macoris: Lo Bracito, Abbott 2197 (US). Santiago: Loma Bajita, District San José. Valeur 1006 (NY, US). Santo Domingo: Cordillera Central, La Cumbre, Arroyo Los Guananitos, Ekman 12 12364 (US). MakrTINIQUE. Pilon du Vauchoy, Plee s.n. (P); no oar Belanger 25 (F, P); Calebasse, 1868, Hahn 2 (BM); forests of Camp Colson, Duss 1502 (NY, P US). Dominica. Lisdara, Hodge 100 (NY, US); Mt. Diablotiri, Lloyd 894 (NY); Waterfall, Roseau Valley. Lloyd 802 (NY, US). GUADELOUPE. Cascade Vauchelet, epee 6738 (P); Duss 872 (P); forests of mapigoa Duss 4140 (NY, P); no Supe L’ age or sath NY, P). St. Lucia. Great Piton, Ramage s.n. (BM); Roseau-Milette Ridge, Box 5 (BM, US). Sr. me hi Wieienlats tops, Smith & cca « s.n. (BM); Eggers 6766 (F); Mts. above i beniliclade River, Morton 5250 (US); along ap enema River, Morton 5412 (US); Cumberland Mts., Morton 5909 (US). GRENADA. No locality, Sherr ng 177 (NY, US). St. CHRIsToPHER. Mt. ridge N of the Crater, Box 288 (BM, US); Nine Turn Gut, ot 361 (BM, US); 1889, Berkeley s.n. (NY). Montserrat. Shafer 377 (F, NY, US); Shafer 763 (NY, US). TRinap. 2.75 miles up Aripo Valley, S side of N Range, Jermy 2625 (BM); 3 miles up Caura Road, by river, Jermy 2101 (BM); Maracas Falls trail, Fay 224 (BM). FRENCH GUIANA. Arataye River, tributary of Approuague, Oldeman 2991 (CAY, NY); région de Paul Isnard, Massif du Décou Décou, Cremers 7927 (CAY, Z); Massif des Emerillons, De Granville 3901 (CAY); Sommet Tabulaire, ca. 45 km SE de — cane 6393 (CAY, Z); entre Saut Mais et Pic Matécho, 30 mt ENE de Saul, de Granville 3095 de Saul, de Granville 3286 (CAY). Surinam. Nassau Mts., Maguire & Maguire 40703 (NY); Nassau Mts., Marwijne River, Maguire et al. 39085 (NY); a Mts., SW plateaus, Lindeman et al. 465 (NY, Z); region of Mt. Raywa, et s.n. (NY); Dist. Brokopondo, Nat. Res. Brownsberg, Tjon-Lim Sang & Wiel 51 (US). Guyana. Cuyun River, Arawak esr bia 380 (BM). VENEZUELA. Aragua: prope coloniam Tovar, Fendler me (GH, MO). Barinas: Dtto. Bolivar, 1 km E of La Soledad, Smith 1374 (UC). Bolivar: Sierre de Lema, Ciaciiee de Rio haa 80 km (en linea recta) al SW de El Dorado, Steyermark 89382 (VE Falcon: Sierra de San Luis, Montafia de ook guariba, approxamadamente 5 km E del Hotel Parador, Steyermark 99319 (NY, VEN). Lara: Dtto. Palavencino, road to Parque Nacional Terepaima, ca. 10- 11 km SE of town of Rio Claro, Smith 1284 (UC). Mérida: Dtto. Sucre, Estanquez, Paramo Las Coloradas, Ortega & app camer 2112 (PORT). Miranda: Sta. Teresa-Altagracia de Orituco, Ar- isteguieta 1761 yh bles fee va Esparta: Isla Margarita, Johnston 152 (F, GH, NY, UC). Sucre: Dtto. Marino & mendi, Pentisdle de Paria, Steyermark et al. 121554 (MO). Tachira: vicinity of Las Minas, 16 aaleye SE of Santa Ana, Steyermark & a. 118892 (MO, VEN). Trujillo: Dtto. Bocono, 33.5 km SE of Bocono, road to Guaramacal, Smith 1542 (MO, PORT, UC). Yaracuy: Dtto. Bruzual, Montafia de Maria Lionza, Steyerm air, Aricuaisa, Liesner & Gonzdles 1317. iat del Rio Casta, Pittier 14042 (F, VEN). CoLomsia. Boyaca: : Agu a San Benito, oom 1412 (COL), 1413 (COL), 1420 (COL), 1423 (COL). Choco: NW side of ty del Buey, Lellinger & de la Sota 287 fap psabe fa Caqueta: Castill Sechules 6680: (6H) Guajira: Serrania de Macuira, Cerro . Snag Ee Sean ae (COL, F, MICH, MO, NY, PH, UC, VT). Narifio: Municipio de Roan entre La Planada y Pialapi, Mora 4021 (COL). Ecuapor: Carchi: Penias Blancas, . km below Maldonado on the Rio San Juan, Madison et al. 4616 (F). Morona-Santiago: nonce ssi small mountain ridge ca. 8 km SE of Mision Bomboiza, Holm-N ielsen et al. 4387 say amppEn, Baeza-Lago Agrio, S of Finca La Cascada, ca. 92 km from Lago Agrio, ot at SOG77 (AAU, 176 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Q). Peru. Amazonas: Prov. Bagua, 43 km (by road) NE of Chiriaco, Barbour 4514 (MO). Cuzco: Prov, La Convencion, Dudley 10324 (GH). Huanuco: ridge E of Tingo Maria, Allard 22301 (GH, US), 22304 (US), 22523 (US); Prov. Leonicio Prado, Dtto. Rupa Rupa, cerca Cerro Quemado, E of Tingo Maria, Schunke 10491 (MO). Madre de Dios: Pantiacolla, serrania across Rio Alto Madre de Dios from Shintuya, Gentry et al. 27366 (MO). Puno: Prov. Carasaya, Vargas 18936 (GH). Boutvia. La Paz: Rio Chimate, Tate 513 (NY); Ipurima, Williams 1201 (GH, NY, US); Hacienda Simaco sobre el camino a Tipuani, Buchtien 5233 (F, MO, NY, US, Z). Cochabamba: Prov. de Chapare, road to San Onofre, 3.3 km N of road to Villa Tunari at point 97.5 km from Cochabamba, Foster 79-140 (UC). Brazi. Espiritu Santo: “Voyage d’Auguste de Saint-Hilaire,” no collector #342 (P); Santa Barbara de Ca- parao, near waterfall above village, Mexia 4096 (US, UC). Mato Grosso: Gorge of Veu da Noiva, Chapado do Guimaraes, Prance et al. 19114 (NY). Para: Serra dos Carajas, “Azul,” near camp at Alston 9009 (BM). Rio Grande do Sul: Sta. Cruz. Jiirgens 69 (NY, UC). Santa Catarina: Santo Antonio, Rohr 1035 (LIL). Sao Paulo: Iguape, Morro dos Pedras, Brade 2708 (PH), 7708 (NY, UC). Prov. Unknown: no locality, Martius 375 (MO). EXCLUDED TAXA See Christensen (1906) for a complete list of excluded taxa, most of which are species of Elaphoglossum and Stenochlaenaa placed in the genus by Pres] (1851). ACKNOWLEDGMENTS I thank Dr. Kenneth R. Robertson (Illinois Natural History Survey, Champaign) for his help and encouragement with all aspects of this study. Most of the work for this monograph was done in the t the Illinois N | History Survey, Champaign, and I am greatly indebted to the Survey for its support and use of its facilities. Mr. Bill N. McKnight, also of the Survey, assisted with the reproduction of the illustrations. work was done while I was a doctoral student in the Department of Plant Biology, University of Illinois, Urbana. I thank the Department for its financial support. My first field trip, to Costa Rica, was made on funds awarded by the Tinker Foundation. A Doctoral Dissertation Improvement Grant (#83-06990) from the National Science Foundation provided finan- cial support for three additional collecting trips in Latin America and for a semester of study at Harvard University. Correspondence and discussion with the following pteridologists gave insight into various aspects of the biology of Olfersia: Dr. Alan R. Smith (University of California, Berkeley), Mr. Robert G. Stolze (Field Museum of Natural History, Chicago, IL), Dr. Rolla M. Tryon and Dr. Alice F. Tryon sag University, Cambridge, MA), and Dr. Warren H. Wagner, Jr. (University of Michigan, Ann r). I am grateful to the directors and curators of the following herbaria for the large quantities of valuable material they made available to me, often for a considerable period of time: A, AAU, BM, CAY, COL, CR, F, G, GH, L, LIL, LPB, MICH, MO, NY, P, PH, PORT, Q, QCA, RB, LITERATURE CITED Brave, A.C. 1971. O género Polybotrya no Brazil. Bradea 1:57-67. CHRISTENSEN, C. 1905-1906. Index Filicum. Copenhagen: H. Hagerup. : 1913. A monograph of the genus Dryopteris, part 1. Kongel. Danske Vidensk. Selsk. Skrift, Naturvidensk. Afd., VIL. 10:55-282. Hottrum, R. E. 1978. Lomariopsis group. Flora Malesiana, II. 1:255-330. SRN Md MRE aT Se Sr SR he ee eT ee ee Ea ae na R. C,. MORAN: OLFERSIA 177 . 1984. Studies of fern genera orig to Tectaria, I. A commentary on recent schemes of classification. Fern Gaz. 12:313- Hooker, W. J. 1864. Species filicum, vol. : London: dees Pamplin. Kautruss, G. F. 1824. Enumeratio filicum. Leipzig: Cnobloch. Link, H. F. 1833. Hortus regius botanicus —— Z: = en Reimer. eign C. 1753. Species plantarum. Stockholm: Salvi ar N. — C. MITTAL. 1961. pianificeca : tities glands in relation to filicin. Pl. Med. 189 Moran, i GC. ae Monograph of a ilo ig fern genus Polybotrya. Ph.D. dissertation, Univ. of Illinois at Urbana-Cham PETIVER, J. 1712. Pterigraphia americana. London. CHI SERMOLLI, R. E.G. 1977. Tentamen pteridophytorum genera in taxonomicum ordinem redi- gendi. PLUMIER, C. 1705. Traité des fougéres de I’Amerique. Paris: Imprimerie royale. PrEsL, K. B. 1851. Epimeliae botanicae. Prague: Haase. PROCTOR, G. R. 1977. Pterid singe neg scat the Lesser Antilles. vol. 2, ed. R. A. Howard. Jamaica Plain, Massachusetts: Arn RaDDI, J. ein Synopsis filicum = saab Conus i. Bol. MITH, A. R. and J. T. MicKEL. 1977. Chromosome commie for Mexican ferns. Brittonia 29:391-398. S . 1875. Historia filicum. London: MacMillan SPRENGEL, C. 1827. Caroli Linnaei, systema veaniebiiliasns: ed. 16. Gottingen: Dieterich. Swartz, O. 1806. Synopsis filicum. Kiel: Bipliop. Nov. Acad. TRYON, “é apy 1970. Devclntnns and evolution of fern flowers of oceanic islands. Biotropica 2: and F. TRYON. 1982. F: d allied plants, with special ref to tropical America York: Springer-Verl ag. WAGNER, w. HL, Jr. and D. M. JOHNSON. 1983. am St a commonly looked storage structt of udu a value in ferns. Taxon 32:51-63. and F.S. WAGNER. 1977. Fertile-sterile leaf dicseephy in ferns. Gard. Bull. Straits Settlem. WaLkerR, T. G. 1966. A cytotaxonomic survey of the pteridophytes of Jamaica. Trans. Roy. Soc. Edinburgh 66:169-237 . 1972. se anatomy - Maxonia apiifolia: a climbing fern. Fern Gaz. 10:241-250. ————. 1985a. Anomalous fronds and venation in Polybotrya nals Fern Gaz. 13:13-16. 1985b. Cytotaxonomic studies of the ferns of Trinidad 2. The cytology and taxonomic implications. Bull. Brit. Mus. (Nat. Hist.) Bot. 13:149-249. igccos, C. L. 1810. Caroli a Linné species plantarum, ed. 4, vol. 5. Berlin: Nauk. APPENDIX: LIST OF EXSICCATA The following is a complete list of all numbered collections seen; unnumbered collections are not listed. Abbott 1951, 1980, 2197. Acufia & Zayas 19620. Alain 3236. Allard 22301, 22304, 22523. Alston 5825, 9009. Anderson & Sternberg 3321. Antonio 3715, 5068. Aristeguieta 1761. Armond 155 Barbour 2625, 4514. Horner « ny 428, 505. Belanger 25. Blomquist 11928. Box 288, 361, 514. Brade 373, 2708, 7708. Brenes 5768, 16236, 16411, 33122. Britton 780, 2249. Britton & Bruner 7592. Britton & Cowell 2210. Britton & jolaerzg 1081. Britton & foes’ 2066. Britton & Wilson 5280. Britton et al. 1242, 5246, 2599. Buchtien 1052, 2240, 5233. Burch 1357. Burger & Antonio 11115, 11272. Burger & Burger 8016. Burger atta 4170, 4554. Burger & tha ee hore ie esis & Herrera 1683. Chrysler & Roever 5240. Clement 982, 6666. Clute 296. Con- treras 2648. Cook & Doyle 12. Cook & Griggs 567. Cremers 6393, 7927. Croat 764, 23219, 25627, 26530, 27421, 33285, 35812, 36794, 37859. Crosby et al. 975 178 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) Dale 23. Daly et al. 1803. Daniels & Jonker 1044. Donnell-Smith 6938. Dudley 10324. Dusén 78, 8185, 13642. Duss 872, 1502, 4140. Dyer 208 Eggers 1040, 5207, 6766. Ekman 4356, 4394, 6759, 12364. Ernst 2055. Evans & Bowers 3303. Fay 187, 393. Fendler 59, 426. Fielder 156. Fisher 42. Folsom 5916. Folsom & Edwards 3377. Foster 97-140. Foster & Foster 1290. Fournier 309. Fuertes 944. en 253. Gastony et al. 387, 648. Gaudichaud 48. Gentle 3528. Gentry & Glewell 6958, 6959. Gentry et al. 27366. Glaziou si Godman & Salvin 99. Gémez 3351, 6895, 7123, 19439, 22034, 20343. eat de 3095, 3286, 390 art 202, 228, seo espenheide 1396. Hess 240, 7040. Hioram 249, 2422. Hodge 100, 101. rode Hodge 1715, 1785, 1976, 2874, 3563. Holdridge 5148. Holm-Nielsen et al. 4387. Hom bersley 105. Howard 11152, 12215. Howard & Howard 8538. Howard et al. 15570. eal ae 2625, 2848, 10857, 10988. Jiménez 1377, st 3157. Johnston tt Jurgens 69. Kennedy 1961. Klawe 1546. Knapp 5342. Knapp & gre 3786. Koptur Lankester 728, 7925. Laurito 8802. Lellinger 396, 550. Lellinger & de la a 287. Lent 1424. Lent et al. nis Leon 3230. Leon & sot 3975, 6611. . & Roca 7991. Leon & Seifriz 78212. Leon et al. 10421. L’Herminier 19. Liesner & Gonzdlez 13172. Lindeman et al. 465. Linden 2188. Liogier 10080. nna 802, 894. Lutz 1450. ok 7080. Madison et al. 4616. Maguire & Maguire 40703. Maguire et al. 39085. Martius 375. Maxon 2220, ee 2511, 4179, 8918, 8976, 9037, 9045. Maxon & Killip 452, 754. McCallum 105. Mexia 4096. Mickel 3425, 3468, 3043, 6384, 6399. Mora 4021. Moran 2181, 2184. Morely & Whiteford 637. Morton 5250, 5412, 5909. Murillo 1412, 1413, 1420, 1423. Nee 9793. Neill 4481. Ocampo 1082, 1996. Ohaus 98. Oldeman 2991. Ollgaard et al. 35677. Ortega & Aymard 1562. Ortega & Gonzalez 355. Ortega & Marcano-Berti Pablo 36. Pabst 4771. Pipoly 5302, 5916. Pittier 1824, 12369, 14042, 16400. Pollard et al. 236. Prance et al. 19114. Proctor 3968, 18946, 21730. Reitz & Klein 3081. Rodriguez 6738. stbignes 235, 1035. Rose & Russel 2 Saunders 1117. Scamman 5982, 5983, 6540, 6544, 7145, 7481, 7719, xis 8149. Schipp 402, 932. Schmalz 183. Schultes 5660. Seifriz 322. Shafer 377, 763, 3229, 3506, 3570, 4441, 8620, 8819. Sherring 177. ios sth Skutch 1805. Smith, A. R. 1284, 1374, 1542. Smith, H. H. 1075. Smith, J. D. 6938. mith, L, rade 2207. Snodgrass & Heller 959. Solomon et al. 6581. Sota, de la 758. Spannagel 326. ee a 7957, 72823. Standley & Valerio 47000, 48521. bogotsahe Ray 461, 470. Stevens & Krukoff 6706, 7 101, 12214. Steyermark 75597, 89159, 89382, 993 9. Steyermark & Agostini 52931. Steyermark & Liesner 118892. Steyermark & Rabe 7 71722, shaees 96371. Steyermark et al. 95827, 121554, 125015. Stimson 3891. Sugden 30. Svensen 342. Tate 513. Tjon-Lim Sang & Wiel 51. Tschudi 163. Turckheim, von 418, 4031, 8044. Tutin 380. Underwood 1249, 1257, 2716, 2888 Valerio 1084, 33123. Valeur 360, 770, 1006. Vareschi 5771. Vargas 18936. Wagner & Gomez 97637. Watt 96. Weber 6129. Weddel 650. Wendt et al. 3846. Werff, van der & titel Williams 1201. Williams & Molina 17787. Wilson 106. Wilson & Webster 448, 507. t 784. Yunker 18544. American Fern Journal 76(4):179-183 (1986) Trichomanes in Florida CLIFTON E. NAUMAN Fairchild Tropical Garden, 10901 Old Cutler Road, Miami, FL 33156 The genus Trichomanes L. is represented in Florida by five species in two subgenera. Their increasing rarity and a lack of ecological information on these species warrants bringing renewed attention to the genus in Florida. Toward that end, I have provided diagnostic and distributional information as well as a key for the Florida taxa. This article is intended to update the previous literature and to subject the information to public critique prior to its inclusion in the pteridophyte treatment for the flora of Florida. For synonymy the reader should consult Wessels Boer (1962) and Proctor (1977, 1985); for the subgeneric treatment of Trichomanes see Morton (1968). The key and distributions were compiled from field populations and representative specimens at FLAS, FTG, NY, and USF. KEY TO THE FLORIDA SPECIES . False veins (those not connected to other veins) absent; lamina without marginal, castaneous, acicular trichomes in stellate clusters, mostly pin- natifid with broad segments, never simple, entire leaf usually more than %.cm. long (subs. Achomanes Presi) ....-. .. es ae: T. holopterum p= 1. False veins present; lamina with marginal castaneous, acicular trichomes in stellate clusters, simple or pinnatifid with narrow segments, entire leaf usually less than 5 cm long (subg. Didymoglossum) ..................... - & Veins pinnate: lamina pinnate ooo sc hei ek Vide vs T. krausii 2. Veins flabellate: lamina simple or lobed =... . 00. -. 0. ce cece tee enews 3 3. Midrib extending to apex of the lamina; involucres fully immersed; lips immersed, indistinct from lamina tissue, but flaring and undulate, usually without a.dark line of margingl Cells... ¢ <. <6. 25sec veer oe T. petersii . Midrib not extending past the middle of the lamina, or absent; involucres fully immersed or not; lips not immersed, distinct from lamina tissue, usually with a dark line of marginal cells cette eee ee sere eee e teste tenes 4. Veins of uniform width to their apices ....T. punctatum ssp. floridanum 4. Veins widening at their apices ..........-----+++++-e0 T. lineolatum oo Trichomanes holopterum Kunze—In addition to the key characters, this species can be distinguished from the other taxa by its erect to ascending rhizomes, generally exserted receptacles, and fully immersed involucres lacking distinct ips. ie ei and habitat.—Monroe and Collier Counties, Florida, and the An- tilles. Epiphytic plants on trees, logs, and stumps; generally in moist, relatively protected habitats; Florida populations are known only from cypress swamps. Sporulates all year. 180 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) This species was first collected in Florida in 1964 by R. and R. Stone in the Big Cypress Swamp, Collier County (Delchamps, 1966). Lakela and Long (1976) cited only Monroe County for the species. The occurrence of the species in two large counties is perhaps misleading since the stations for it in Collier and Mon- roe are relatively close to their common border. Visits to the populations in early 1986 indicated that plants are still present in relatively large numbers. Available habitat was abundant nearby and it is reasonable to expect more stations to be found with continued field work. The present populations compare well with the description given by Farrar and Wagner (1968), indicating a fairly stable existence (at least locally). Farrar and Wagner described the gametophytes as more common and covering a larger area than the sporophytes; this is true for the present populations. It is possible that the gametophytes are more protected from periodic dry conditions by being more appressed to the substrate than are the emergent sporophytes. This sug- gestion requires more field study, but may explain the greater abundance of the gametophytes at these localities. The gametophytes are frequently embedded in moss and liverwort colonies; any effects on the populations due to competiton with the latter are unknown. Trichomanes krausii Hooker & Greville—Trichomanes krausii resembles T. holopterum in having pinnatifid leaves, but can be distinguished by its narrower segments, castaneous, acicular, marginal (often reflexed) trichomes, and a dis- tinct involucre lip that is not completely immersed in the leaf tissue. Distribution and habitat.—Dade County, Florida, and the Antilles south to continental tropical America; epiphytic or rarely epipetric plants in pockets of humus on limestone, in moist forested habitats. Sporulates all year. Trichomanes krausii was first collected in Florida in 1903 by J. K. Small and J. J. Carter in Dade County. Plants were subsequently collected by Small from Nixon-Lewis Hammock, Camp Longview, and Silver Palm, and collected by A. A. Eaton south of Cutler. The plants are absent now from most of these localities, probably as a result of habitat disturbances. The Nixon-Lewis Hammock for example, has been bulldozed (P. Adams, pers. comm.). Plants are currently known to occur in Castello and Fuchs hammocks, at most only a few miles from the Silver Palm location reported by Small (1938). The population at Castello Hammock is represented by a single colony of about three Square decimeters. A search was not made for gametophytes, but they may occur apart from the sporophytes at these stations and possibly in at least one adjacent hammock (Ross Hammock) in which the sporophytes are apparently absent. If the gametophytes are persistent and as numerous as those of T. holopterum, it is possible that new locations for the plants may be found. Trichomanes lineolatum (Bosch) Hooker in Hooker & Baker. Distribution and habitat.—Dade County, Florida (now apparently extirpated), the Antilles, Venezuela, and Brazil; epipetric or rarely epiphytic plants in moist forested habitats. Wessels Boer (1962) reported the species only on limestone. Sporulates all year. C. E. NAUMAN: TRICHOMANES IN FLORIDA 181 Lellinger (1985) treated T. lineolatum as extirpated in Florida. This appears to be the case, but I have included the species here in order to facilitate its redis- covery by other workers. The species is easily confused with T. punctatum ssp. floridanum, which is extant in Castello Hammock and adjacent Ross Hammock. It is possible that T. lineolatum has been overlooked in this locality because of its similarity to T. punctatum ssp. floridanum. (The key characters easily separate the two taxa, but require a 10 power handlens.) This species was first collected in Florida in 1906 by J. K. Small and J. J. Carter in Dade County. Small (1931, 1938) knew of this species only from Ross Ham- mock near Silver Palm. Searches by myself and others as recently as 1981 have failed to locate plants at this site. There has been little disturbance at Ross Hammock that would explain its apparent absence. To my knowledge, no one has yet searched for gametophytes. Trichomanes petersii A. Gray—This species can be distinguished from the other simple-leaved taxa by a midrib that extends to the leaf apex, a fully immersed involucre, and flaring, undulate involucre lips that are otherwise little distinct from the lamina tissue. Distribution and habitat.—Citrus, Hernando, and Sumter counties, Florida, northward in the Southeastern United States from Georgia, Alabama, and Lou- isiana north to North Carolina and Tennessee, and south to Mexico and Gua- temala; epipetric on sandstone or other acidic rocks or epiphytic in moist hab- itats, often near water. Sporulates in summer. Trichomanes petersii was first collected in Florida in 1936 by E. St. John in Hernando County. The species appears to be restricted to acid substrates in spite of reports to the contrary (Brass, 1955; Duncan, 1955). Wessels Boer (1962) re- ported it as epiphytic or epipetric on sandstone and limestone rocks. He appar- ently based his inclusion of limestone as a substrate on herbarium material, including a collection by E. St. John from Brooksville. Wherry (1955) reported that a visit to this site revealed plants occurring on only two boulders, both of which were chert, not limestone. This strengthened Wherry’s contention that the species occurs only on acid substrates. The occurrence of the species as an epiphyte seems to be more prevalent in southern portions of its range, and occurrence on rocks is more common in northern portions of its range. In relatively warmer regions the species is known largely as an epiphyte (Stolze, 1976; Wessels Boer, 1962). The American Institute of Biological Sciences field trips in 1986 revealed plants in Sumter County oc- curring as epiphytes on oaks in a swamp system near the Withlacoochee State Forest. Farther north, T. petersii occurs on wet, acid rock substrates in protected crevices, usually near running water (e.g., Shaver, 1954, and references therein; Wyatt, 1984). Lower temperature regimes may be critical in determining whether it grows epiphytically or epipetrically. Trichomanes punctatum Poiret in Lam. subsp. floridanum W. Boer c Distribution and habitat.—Citrus, Sumter, and Dade counties, Florida; endemic; epiphytic or epipetric on limestone in moist forested habitats. Other subspecies 182 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) occur throughout the American tropics. Lellinger (1985) reported only Dade and Sumter counties as the extant range for this species. Sporulates all year, but mostly in spring and summer northward. This subspecies was first collected in Florida in 1901 by J. K. Small in Dade County. Plants were first collected in Citrus and Sumter counties in 1936 by R. St. John. Small (1931, 1938) listed Snapper Creek Hammock, and several pine- land hammocks as far southwest as Camp Longview and Royal Palm Hammock. Additionally, the plant has been collected south of Cutler, and at Hattie Bauer, Colwell, Sykes, and Ross hammocks by A. A. Eaton, J. K. Small, and others, The plants are still known to occur in Fuchs and Castello hammocks in Dade County and are apparently absent at the other stations. Small (1918a, 1918b, 1931) used the name T. punctatum, but in 1938 he mis- applied the name T. sphenoides Kunze to the Citrus and Sumter County popu- lations. The misapplication was followed by Wherry (1964). Both authors were apparently convinced that the disjunction of the two populations was partly due to their being different taxa, or at least was sufficient to judge them to be different taxa. Yet, these same authors were willing to accept the disjunct populations of T. petersii in Florida. Perhaps their greater familiarity with populations of T. petersii in more northern states allowed them greater confidence in applying the name to the disjunct Florida populations. A lack of familiarity with T. sphe- noides, its reported wide distribution in the American tropics (especially in the Antilles), and Eaton’s (1906) misapplication of the name to Dade County popu- lations, made it a logical possibility for the northern stations of T. punctatum subsp. floridanum. CONSERVATION STATUS All five taxa have at least one population represented within a Federal, State, or local park or preserve. This does not however, guarantee their preservation. The taxa inhabiting swamps, T. holopterum and T. petersii, may undergo some recovery if habitats are not modified in those areas under protection. Taxa oc- curring in hammocks on limestone, T. krausii, T. punctatum ssp. floridanum, and T. lineolatum, are in danger of extirpation, despite protective efforts. These species appear to require high relative atmospheric humidities. Sinkholes with year-round standing water and dense canopies once maintained such high rel- ative humidities in the hammock habitats. Drainage efforts in Florida since the early 1900s, however, have sufficiently lowered the water table that the sinkholes no longer contain standing water for any significant period of time. Additionally, human disturbance has opened many hammocks to the desiccating effects of wind and sun. As a result, one species (T. lineolatum) may have already been lost, and two others are in immediate danger of extirpation. EXCLUDED SPECIES Trichomanes boschianum Sturm—This plant was listed by Small (1931, 1938) as questionably occurring in western Florida, stating in 1938 (p. 57), “The Flor- C. E. NAUMAN: TRICHOMANES IN FLORIDA 183 ida station for it appears to be lost to science.” The species has been ex- cluded from Florida by all subsequent authors. Since I know of no docu- mented records in Florida, I am excluding it. Trichomanes sphenoides Kunze (=T. punctatum ssp. sphenoides (Kunze) W. Boer)—Misapplied to T. punctatum ssp. floridanum by Eaton (1906), Small (1938), and Wherry (1964). This subspecies occurs in the Greater Antilles, Central, and northwestern South America (Wessels Boer, 1 1962). Trichomanes punctatum ssp. sphenoides (Kunze) W. Boer—Treated as a syn- onym of ssp. floridanum by Lakela and Long (1976). Trichomanes membranaceum L. —Mistakenly cited for Florida by Mickel (1979). Mickel has confirmed this (pers. comm.). ACKNOWLEDGMENTS I thank Dr. R. P. Wunderlin, Univ. of South Florida, ae A. M. Evans, Univ. of Tennessee, Knoxville, and reviewers for their comments and cri itique LITERATURE CITED Brass, L. J. 1955. Report of the southern Florida field trip. Amer. Fern J. 45:48-54. DELCHAMps, C. A. 1966. Binion holopterum—a filmy fern new to the United States. Amer. Fern J. 56:138-13 DUNCAN, W. H. 1955. nes records for Georgia ferns. Amer. Fern J. 45:1-10. EaTon, A. A. 1906. Pteridophytes observed during three excursions into southern Florida. Bull. Farrar, D. R. and W. H. WAGNER, ie 1968. The gametophytes of Trichomanes holopterum Kunze. Bot. Gaz. ee po 210-219 LakELA, O. and R. W. Lonc. 1976. Ferns of Florida Miami: Banyan LELLINGER, D. B. 1985. A field sa of ta ferns & fern allies of the United States & Canada. Washington, DC: Smithsonian Institution MIcKEL, J. T. 1979. How to know the ie and fern allies. Dubuque, Iowa: W. C. Brown C Morton, C. V. 1968. The genera, subgenera, and sections of the Hymenophyllaceae. Contr. U. S. Natl. Herb. 38:153-214. Proctor, G. R. 1977. Pteridophyta. Vol. 2 haah Flora of the Lesser Antilles, ed. R. A. Howard. sic Plain, Mass.: Arnold Arbore’ . Ferns of Jamaica. London: Sauk Museum (Nat. Hist.). SHAVER, I. te 1954. Ferns of the eastern central states with special reference to Tennessee. New York: Dover Publ. SMALL, J. ws 1918a. Ferns of Royal Palm Hammock. New York: Publ. by the author. —————.. 1918b. Ferns of Tropical Florida. New York: Publ. by the author. : ey Ferns of Florida. New York: Science Press. re es - 1938. Ferns of the Southeastern States. ene nn: Stouze, R.G. 1976. Fee and fern allies of Guatemala, Part I. Ophioglossaceae through Cyathea- ceae. Fieldiana, Bot. 39:1-130. WESSELS bas J. G. 1962. The new world species of Trichomanes sect. Didymoglossum and Mi- eerl. 11:277 Wherry, E. pp sgen ge ler of Thich petersii. Amer. Fern J. 45:93-95. 1964. The southern fern guide. Garden City, New York: Doubleday & Co. Wyarr, R. 1984. A new station for Trichomanes petersii in Alabama. Amer. Fern J. 74:30. American Fern Journal 76(4):184-186 (1986) Some New Names and Combinations in Pteridaceae ROLLA TRYON Harvard University Herbaria, 22 Divinity Avenue, Cambridge, Massachusetts 02138 Although Pteridaceae Reichb. f. has been considered to be an illegitimate name (Pichi Sermolli, 1982), Article 63.3 of the International Code of Botanical Nomenclature was altered at Sydney and the name is consequently legitimate. It may, however, be antedated by Pteridaceae Gaud. or Pteridaceae (S. F. Gray) Gaud. and for this reason Panigrahi (1986) has proposed the name for conser- vation with appropriate authorship. The name was used by Tryon and Tryon (1982) and it will be used in “The Families and Genera of Vascular Plants” in anticipation of its conservation. Adopted in a broad sense, the Pteridaceae in- clude major groups of ferns that are currently recognized, sometimes as tribes and sometimes as families. In order to be consistent with other treatments of fern families in “The Families and Genera of Vascular Plants,” it is necessary to place them as subfamilies. The nomenclature of these follows with the new status provided when necessary. Pteridaceae subfam. Adiantoideae (Presl) R. Tryon, stat. nov.—Adianteae, as tribe Adiantaceae Presl, Tent. Pterid. 139. 1836. Pteridaceae subfam. Ceratopteridoideae {J. Smith) R. Tryon, stat. nov.—tribe Ceratopterideae J. Smith, Hist. Fil. 170. 1875. Pteridaceae subfam. Cheilanthoideae (J. Smith) R. Tryon, stat. nov.—tribe Chei- lantheae J. Smith, Hist. Fil. 277. 1875. Pteridaceae subfam. Platyzomatoideae (Nakai) A. F. Tryon, Amer. J. Bot. 51:942. 1964. Pteridaceae subfam. Pteridoideae. Pteridaceae subfam. Taenitidoideae (Presl) R. Tryon, stat. nov.—tribe Taeniti- deae Presl, Tent. Pterid. 222. 1836 Sections of Syngramma.—Three sections are recognized in the genus: sect. Syngramma and sect. Toxopteris, both with a simple lamina, and sect. Cras- pedodictyum with a ternate, pedate, or palmate lamina. Well developed leaves of the six species of sect. Craspedodictyum have 4 3-foliolate to 5-(rarely 6-)foliolate lamina and represent a distinct group within the genus. In a study of Syngramma and related genera, Holttum (1975) recog- nized that Craspedodictyum did not merit generic rank and he placed the species in sect. Syngramma in a group “with pedate fronds.” sect. Craspedodictyum (Copel.) R. Tryon, stat. nov.—Craspedodic- tyum Copel., Philipp. J. Sci., C. 6:84. 1911. Species of Cerosora.—A third species is added to the genus Cerosora, along with C. chrysosora (Baker) Domin of Borneo and the recently described C. su" matrana Holtt. of Sumatra. This is the previously misunderstood Gymnogramina microphylla or Pityrogramma microphylla (Hook.) Domin of the Himalayas. ee, ee eee aS R. TRYON: PTERIDACEAE 185 The characters of the two previous species provide a clear separation of Cero- sora from other genera of the Pteridaceae: stem creeping, with rigid, dark tri- chomes or bristles, petiole with one vascular bundle near the base, veins all free, exindusiate sori along the veins with few-celled, simple trichomes among the sporangia, and spores with a prominent equatorial flange. The species C. micro- phylia has these characters and is clearly a member of the genus. F 65° * Cerosora microphylla (Hook.) R. Tryon, comb. nov.—Gymnogramma micro- phylla Hook., Icon. Pl. 10:t. 916 (Cent. Ferns t. 16). 1854. Cheilanthes smithii—The species Pellaea smithii, with unusual long and nar- row ferti g t d ponding long, modified marginal indusia, is placed in Cheilanthes. It was included, along with related species of eastern Asia, in Mildella, a segregate genus from Cheilanthes by Hall and Lellinger (1967). As in Cheilanthes species, the lamina of C. smithii is more or less gradually reduced apically and the ultimate segments are mostly joined or a few are adnate. Pellaea species differ in their imparipinnate lamina and stalked to sessile ultimate seg- ments. (p05 | Cheilanthes smithii (C. Chr.) R. Tryon, comb. nov.—Pellaea smithii C. Chr., Acta Horti Gothob. 1:84. 1924. The genus Paraceterach.—Paraceterach is enlarged to include five Asiatic species often treated in Gymnopteris or Hemionitis, along with the two Austra- lian species previously included in the genus. It is characterized by the following: lamina 1- or 2-pinnate, on the lower surface densely invested with indument which conceals the sporangia (especially the young ones), exindusiate soral lines on unmodified, free, or occasionally anastomosing, veins, and spores that are cristate, echinate or plain (the surface deposit incomplete). This suite of char- acters serves to distinguish the genus from related ones such as Coniogramme, Bommeria, and Hemionitis. Three groups may be recognized in the genus. The Australian species, P. muelleri and P. reynoldsii, have a moderately creeping stem bearing mature scales with a dark, sclerotic center and lighter margins, and subclathrate, con- colorous lamina scales. An Asiatic group of two species, P. delavayi and P. marantae, (the latter also westward to Europe) has short-creeping, multicipital stems, with light brown to somewhat reddish, narrow, concolorous scales, and the lamina indument is similar to that in the Australian species. The other Asiatic group, composed of P. bipinnata, P. sargentii, and P. vestita, differs only in having the lamina indument of very long, appressed and matted trichomes. These groups represent three lineages that can be accommodated within Paraceterach as are distinctive elements within Cheilanthes, Notholaena, Pellaea, and Hem- ionitis. Hemionitis is an American genus of seven species, differing from Paraceter- ach especially in the indument on the lower surface of the segments. There are few or no scales, and the few to numerous trichomes are short (longer ones are also present in two species with dimorphic pubescence}, acicular, and erect or patent, and the soral lines of sporangia are evident. In addition, some species 186 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) have an erect stem, and most of them have an entire to deeply lobed lamina. Hemionitis and Paraceterach evidently represent separate, convergent lineages. Isozyme studies will be especially useful for depicting relations of these genera. ‘“Hemionitis” arifolia is reported by Manton & Sledge (1954) as an apogamous tetraploid species. The chromosome number of n and 2n = 120 clearly relates it to the Pteridaceae and especially to the subfam. Cheilanthoideae. It is a mor- phologically isolated species of uncertain affinity and likely represents a derived alloploid. Giannasi and Mickel (1979) identified seven flavonoids common to Hemionitis arifolia and H. elegans, and also reported four flavonoids unique to H. arifolia. It is recognized that Acrostichum marantae L., here included in Paraceterach, is sometimes considered to be the type of Notholaena (Pichi Sermolli, 1983, pp. 112-117). However, John Smith (1875, p. 278) was the first to typify Notholaena and his choice, Pteris trichomanoides L. [Notholaena trichomanoides (L.) Desv.] cannot be supersede The species of Paraceterach are the following: Paraceterach bipinnata (Christ) R. Tryon, comb. nov.—Gymnopteris bipinnata Christ, Notul. Syst. (Paris) 1:55. 1909. Paraceterach delavayi (Baker) R. Tryon, comb. nov. —Gymnogramma delavayi Baker, Ann. Bot. (London) 5:484. 1891. —Gymnopteris delavayi (Baker) Un- derw Paraceterach marantae (L.) R. Tryon, comb. nov.—Acrostichum marantae L., Sp. Pl. 1071. 1753.—Gymnopteris marantae (L.) Ching. Paraceterach muelleri (Hook.) Copel., Gen. Fil. 75. 1947. Paraceterach reynoldsii (F. Muell.) Tindale, in J. M. Black, Fl. South Australia, ed. 3, by J. P. Jessop, 1:53. 1978. Paraceterach sargentii (Christ) R. Tryon, comb. nov.—Gymnopteris sargentii Christ, Bot. Gaz. (Crawfordsville) 51:355. 1911. Paraceterach vestita (Hook.) R. Tryon, comb. nov. —Gymnogramma vestita Hook., Icon. Pl. t. 115. 1837.—Gymnopteris vestita (Hook.) Underw. LITERATURE CITED Giannasi, D. E. and J. T. MICKEL. 1979, Systematic Onno of flavonoid pigments in the fern genus Hemionitis oe Brittonia 31:40 : Hatt, C. je ~ D. B. LELLin 1967. A revision of fhe fern genus Mildella. Amer. Fern J. 57: Ho.itrum, R. E. 1975. A comparative account of the fern-genera Syngramma J. Sm. and Taenitis Willd. Kew Bull. 30:327-343. MANTON, I. and W. A. SLEDGE. 1954. Observat; cytolo; - f the pteridophyte flora of Ceylon. Philos. Trans. Ser. B. 238: 427-185. et ; PANIGRAHI, G. 1986. Proposal to conserve Pteridaceae Reichenbach (Pteridophyta). Taxon 35:385- Pic SeRMouts, R. E.G. 1982. A further contribution to the nomenclature of the families of Pteri- ophyta. Webbia 35:223-337. 1983. Fragmenta Pteridologiae, VIII. Webbia eg 111-140. Smitu, J. 1875. Historia filicum, — acmillan & Tryon, R. M. and A. F. Tryon. 1982. Ferns and at vlitina with special reference to tropical America. New York: Springer-Verlag. American Fern Journal 76(4):187-188 (1986) A Novel Method for Surface-sterilizing and Sowing Fern Spores THOMAS R. WARNE, Gary L. WALKER, and LESLIE G. HICKOK Department of Botany, University of Tennessee, Knoxville, Tennessee 37996-1100 Numerous methods are available for the sterilization of fern spores (see Dyer, 1979). Many of these methods are inefficient in terms of time and spore loss and require bulky equipment such as a centrifuge or vacuum filtration units. We present an easy and effective method for spore sterilization that requires no specialized tools or equipment, other than a sterile area. We have used the method presented here routinely to sterilize Ceratopteris richardii Brongn. spores. In addition, this technique has worked with spores of Lygodium, Pteridium, and other species of Ceratopteris (unpublished data). The basic innovation in this method involves the removal of liquid from a spore suspension in a conical centrifuge tube by means of a Pasteur pipet while leaving the spores in place. This method requires the following supplies: 1) sterile Pasteur pipets (Fisherbrand® borosilicate glass, 5%4 inches) with square-cut ends that are not cracked or chipped; 2) sterile pipet bulbs (natural rubber dropper bulb); and 3) sterile conical centrifuge tubes (Pyrex® No. 8060, 15 ml). To accom- plish spore removal, insert a pipet with attached bulb into a conical centrifuge tube that contains a liquid and spores and suspend the spores in solution by bubbling air into the liquid. While continuing to bubble air, gently, but securely, seat the pipet tip onto the base of centrifuge tube. A slight rotation of the pipet may assist seating, but excess downward or lateral force will crush the pipet tip. Squeeze the pipet bulb to force additional air out of pipet, then slowly release the bulb to aspirate liquid into the pipet; most spores should collect at the base of the centrifuge tube and not be drawn into the pipet. This procedure has been incorporated into the following generalized spore sterilization and sowing method. Spore preparation.—These procedures can be performed outside of a sterile area. Transfer spores into a conical centrifuge tube and presoak them with dis- tilled water for 16 hours. We routinely dry-sterilize (120°C, 2 hours) centrifuge tubes to eliminate cross-contamination of spore stocks and minimize contami- nation from the tube. If the spore source is exceptionally dirty or contaminated, wetted spores may be transferred with a Pasteur pipet to a clean sterile centri- fuge tube immediately prior to sterilization. This significantly reduces the amount of coarse debris that may harbor and shield contaminants during sterilization. We prepare and sterilize from 2 to 250 mg of Ceratopteris spores per tube, depending on the desired sowing density and number of cultures to be sown. Spore sterilization.—All procedures for spore sterilization and spore sowing should be performed in a sterile area using standard sterile technique. All ma- terials that come into contact with spores should be sterilized. To avoid contam- ination, use a clean sterile pipet and pipet bulb for each step. To sterilize spores, remove the presoak solution using the method described 188 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) above and discard the pipet used. Add ca. 2.0 ml or one full Pasteur pipet of freshly prepared 0.875% sodium hypochlorite (1 part commercial bleach, 5.25% sodium hypochlorite, to 5 parts distilled water). To insure adequate sterilization of spores and centrifuge tube interior, add this solution so that it washes from the lip of the centrifuge tube and suspend the spores in this solution by bubbling air through a clean, sterile pipet. Remove the sodium hypochlorite solution with a clean, sterile pipet using the technique described above. Immediately rinse the spores three times. For each rinse, add one full pipet of sterile distilled water to the tube and remove water with a clean, sterile pipet using the above method. For sterilization, we routinely expose Ceratopteris spores to the sodium hypochlorite solution for 21 minutes. Spore sowing.—For spore sowing, add a known amount of sterile distilled water to the sterilized spores and uniformly suspend spores by bubbling air in solution. Dispense drops of the spore suspension into the appropriate culture vessel containing liquid or agar-solidified medium, soil, etc. For agar-solidified medium, spores can be uniformly distributed across the agar surface with a sterile wire bent into a smooth cornered triangle (coat hanger shape). We routinely use 5 Pasteur pipet drops of spore suspension (ca. 0.146 ml) per culture vessel. Controlled-drop pipets may make sowing more precise. For con- trolled sowings it is necessary to predetermine the number of spores per unit weight and to correct for maximum percentage germination of a given spore stock. For example, we have determined that in 1.0 mg dry weight there are ca. 1250 C. richardii spores and that maximum percentage germination may range from 60 to 95% for different sources. Modifications for spores of other taxa.—Adjustments that may be needed in presoak and sterilization times and solutions, depending upon the spores used, can be empirically determined. Sterilization of spores smaller than those of Ceratopteris, which are ca. 86-129 um diameter (Lloyd, 1974), may require care- ful selection of both Pasteur pipets and centrifuge tubes and careful practiced technique. ACKNOWLEDGMENTS This research was supported by grant DCB-85-11273 from the National Science Foundation to L. Hickok and T. Warne. LITERATURE CITED — Dyer, A. F. 1979. The culture of fern gametophytes for experimental investigation. Pp. 253-305 in The experimental biology of ferns, ed. A. F. Dyer. London: Academic Press. Lioyp, R. M. 1974. Systematics of the genus Ceratopteris Brongn. (Parkeriaceae) II]. Taxonomy. Brittonia 26:139-160. SHORTER NOTE Osmunda cinnamomea forma frondosa in the Coastal Plain of Georgia and Florida.—The article of Werth et al. (Amer. Fern J. 75:128-132, 1985) reporting the discovery and observations of Osmunda cinnamomea L. forma frondosa (Torrey & Gray) Britton at Mountain Lake, Virginia, has provoked interest in this puzzling morphology. Typically, O. cinnamomea sporophylls are strictly dimor- phic with respect to spore production; however, forma frondosa produces spo- rophylls that bear both sterile and fertile pinnae. This unusual form of Cinnamon Fern has not been reported previously from Georgia (McVaugh & Pyron, Ferns of Georgia, The University of Georgia Press, 1951) or Florida (Lakela & Long, Ferns of Florida, Banyan Books, Miami, 1976). However, there are at VSC three records of Osmunda cinnamomea forma frondosa from the Coastal Plain of Georgia and Florida: Georcia: Brantley Co.: % mi E of ITT-Rayonier Satilla Forest Headquarters along Hwy GA-32 between Browntown and Needmore, 9 Sept 1982, Faircloth 8711. Lowndes Co.: Slough S$ of Valdosta, along Loch Laurel Road, 4.3 mi S of Hwy GA-31, 24 Apr 1986, Carter 4807. FLoripa: Leon Co.: Disturbed sandy ground just E of St. Marks River and 2 mi W of Jefferson County line on Hwy US- 27, 1 Nov 1979, Carter 2318. The Brantley County population consisted of seven plants that were located on the backslope of highway GA-32 and numerous plants that extended into the adjacent pine flatwoods. Only those plants located on the backslope had fron- dosa-type sporophylls. Development of laminar and sporogenous pinnae on a frond was highly variable among the seven plants and even among fronds on individual plants. There was evidence that the plants on the backslope had been mown over earlier in the year by Rayonier personnel. At the Lowndes County site only a single plant of forma frondosa was found in a population of about one hundred typical plants. This plant was located near the edge of woods and was conspicuously smaller than its typical neighbors. Data on population size and other pertinent notes relating to site disturbance is lacking for the Leon County record. Our observations also suggest that this phenomenon is environ- mentally induced but we have no additional explanations to offer beyond those reviewed and proposed by Werth et al. (1985). The phenomenon probably has little genetic or taxonomic significance; however, we feel it is worthy of contin- ued study to elucidate a clearer explanation of the factors underlying morpho- logical expression of leaf dimorphism. We do suspect that forma frondosa is of more frequent and widespread occurrence than has been reported previously.— RICHARD CARTER and WayNE R. FAIRCLOTH, Department of Biology, Valdosta State College, Valdosta, GA 31698. 190 AMERICAN FERN JOURNAL: VOLUME 76 NUMBER 4 (1986) REVIEW “Polypodiaceous Ferns of India,” by C. K. Satija and S. S. Bir. 1985. vi + 132 pp. including 9 photographic plates. Aspects of Plant Sciences, Volume 8. New Delhi: Today & Tomorrow’s Printers and Publishers. Distributed in U.S.A. and Canada by: Scholarly Publications, 7310 El Cresta Drive, Houston, TX 77083. $19.00. This is a listing of Indian Polypodiaceae in the strict sense, with also the Grammitids, Loxogramme, and Dipteris. For each species there are selected references, ecological comments, and a statement of distribution, but no keys or descriptions. Following the main text are two addenda and an appendix by S. S. Bir, a very incomplete three pages of errata, and an index. Some species that were included solely on the basis of literature reports are apparent misidentifications or mislocalizations, whereas others that I believe to be reliably on record from India are unmentioned. I also noted five species originally described from India that were omitted: Grammitis cuspidata Zenker, Plantae Indicae 1, t.2. 1835 (a Loxogramme); Grammitis attenuata Kunze, Linnaea 24:251. 1851 (a true Grammitis); Polypodium decurrenti-adnatum Ro- senstock, Repert. Spec. Nov. Regni Veg. 12:248. 1913; Drynaria meeboldii Ro- senstock, ibid.; and Arthromeris jarrettiae Sastry & Chowdhury, Bull. Bot. Surv. India 11:442. 1972. Four combinations are proposed in Phymatopteris but two of these were al- ready published by Pichi Sermolli in Webbia 28 (1973) in a paper cited several times elsewhere in the book. As a genus, Phymatopteris Pic. Ser. can hardly stand as it merges with the earlier Crypsinus Presl.—M. G. Price, Herbarium, North University Building, University of Michigan, Ann Arbor, MI 48109. Referees, 1986 I thank the Associate Editors and referees listed below for their valuable assistance in the review process. Their evaluations of manuscripts submitted to American Fern Journal have aided authors, made my job easier, and contributed to the quality of our journal ALan R. SMITH Donald M. Britton R. James Hickey Douglas E. Soltis Howard Calkin David M. Johnson Dennis Wm. Stevenson David S. Conant Robert M. Lloyd W. Carl Taylor Gillian Cooper-Driver Michael R. Mesler David H. Wagner Michael Cousens John T. Mickel Warren H. Wagner, Jr bara Ertter James D. Montgomery James W. Wallace A. Murray Evans Robbin C. Moran Charles R. Werth Donald R. Farrar Suzanne Morse Dean P. Whittier Arthur C. Gibson Michael G. Price Michael D. Windham Judith E. Gordon George R. Proctor George Yatskievych Leslie D. Gottlieb Thomas Ranker American Fern Journal 76(4):191 (1986) Index to Volume 76—1986 Classified entries: botanical names (new names in boldface); major subject headings, including key words from titles; reviews (grouped, listed by first author of work reviewed). Names of autho isilowed by titles of articles or references to first authors, are listed alphabetically in Table of Contents, iii-iv. Adiantum pedatum complex, 151 Anatomy: Cyclodium, 57; Olfersia, 163 plenium septentrionale, 26 Botrychium, 33: in western North America, 33: subg. Botrychium, 33; B. ascendens, 36; B. as- cendens x crenulatum, 39; B. campestre, 39; B. pedunculosum, 43; B. pedunculosum x pinnatum, 45 Cerosora, 184; C. age 185 —— oe + te 1; 4a . 38; B. campestre, 42; B. pedunculosum, 43; Cyclo- dium meniscioides, 64; Pellaea breweri, 118; Pellaea truncata, 119 Conservation: Trichomanes in Florida, 182 nipica Index filicum iques: Isoétes spp., 21; Lycopo- sae hos any Cyclodium, 56: C. akawaiorum, 71; C. calophyl- lum, 73; ianense, 75; C. heterodon var. heterodon, 79; C. heterodon var. abbrevia- tum, 80; e, 82; C. meniscioides var. caentactolties. 87; C. meniscioides var. palu- 87; C. meniscioides var. rigidissimum, 88; C. rheophilum, 88; C. seemannii, 89; C. trianae var. Tare 92; C. trianae var. cho- varians, 94 Cyatopteri imi 99 usiana, 28 Ecology, Cyclodium, 64 Electrophoresis: in pteridophytes 102; Marsilea, q7: a dium Endem cap OEE 45 Hiccane in herbarium specimens, 102; Marsil- ea, 17; Regnellidium, 17 Gametophytes: biosystematic uses of, 114; Iso- étes spp., 21; Lycopodium lucidulum, 48 Geographic variation: in Pellaea, 120 Herbarium specimens: in biosystematic re- 149; spore viability ir in, 141; stomate size in, 149 Hemionitis, 185; H. arifolia, 186 Hybridization: Botrychium, 39, 45; Cyclodium, 64; heme 121 ex filicum, corrections, 47 i oe morphology, 1; spore termi- nology eine lucidulum, 48 Marsilea vestita, 17 Megaspores: Isoétes, 1; Marsilea, 17; Regnelli- lum, 17 Morphology: Cyclodium, 57; Olfersia, 163 Olfersia, 161; O. cervina, 161 Ophioglossum palmatum, 5 Osmunda cinnamomea f. frondosa, 189 Paraceterach, 185: P. bipinnata, 186; P. delavayi, 86; P. maran tae, 186; P. muelleri, 186; P. reyn- oldsii, 186; P. sargentii, 186; P. vestita, 186 Pellaea: P. atropurpurea, 142; P. brachyptera, 25; P. breweri, 118; P. ternifolia, 142; P. truncata, podium: spore size, 155; stomate size, 155; P. californicum, 155; P. Polystichum: spore size, 153; P. perastichiolden mp P. braunii, 157; P. heohetbam: 153; P. x ri, 158 eaves 184: subfam. Adiantoideae, 184; subfam. oe 184; subfam. Cheilanthoideae, 184; subfa atyzomato- ideae, 184; subfam. Picvickloas, 184; subfam SY 184 Range ectenatona Arkansas, Dryopteris carthu- — 26; West Virginia, Cystopteris esse soon o bafevees Pacey iphyum 17 Reviews: Dyer, A. F. and C. N. Page, eds., Biology of EATEN 30; Lellinger, D. B., A field 7 2 sk. pgfan Pe ohms Tiz fab | Se ee | States & Canada, 29; Satija, C. K. and S. S. Bir, Polypodiaceous ferns of India, soa Spores: abortion, 129; size in Adian atum aia 152; viability in Plas 3 141; Cyclo- sae eine 1, 21; Olfersia, 161 es, Spore orrelis tion: Isoétes, 21 Spore PACT and culture, 187 Stigmatopteris, 56 181; T. punctatum subsp. floridanum, 181 STATEMENT OF OWNERSHIP MANAGEMENT AND CIRCULATION Required by 39 U.S.C. 3685) 1A. TITLE OF PUBLI B. PUBLICATION NO. 2. DATE OF FILING AMERICAN FERN JOURNAL o1o1o;2/8/4i4 4 18 Nov 1986 REQUENCY UE 3A. NO. OF A ANNUAL ne Ga ANNUALLY 4 10 m “quarterly: te subede KNOWN OFFICE OF City, Cou Code) (Not printers) Pringle Herbarium, University of Vermont, Burlington, VT 05405 Pringle Herbarium, University of Vermont, Burlington, VT 05405. nee re rea 6. FULL NAMES AND COMPLETE MAILING ADDRESS OF PUBLISHER EDITOR NOT be biank) PUBLISHER (Name and Complete Mailing Address) AMERICAN FERN SOCIETY, INC. (see 5.) Dr. Alan R. Smith, Dept. of Botany, Univ. of California, Berkeley, CA 94720 none 7. OWNER (If owned by a corporation, its name and address must be stated and also im immediately thereunder the names and addresses of stockholders owning or holding 1 percent or more of total amount of stock. If not owned by a corporation, the names and addresses a the individual owners be given. If owned by a partnership or other ny cts gi firm, its name and address, as well as that of each individual must be given. If the publice- tion is published by a nonprofit organization, its name and address must be stated.) (Item must be completed.) mneece (see 5.) FULL NAME American Fern Society, Inc. oe on 8. — Lip eapueriemnnginn MORTGAGEES, AND OTHER SECURITY HOLDERS OWNING OR HO: LDING 1 PERCENT OR MORE OF TOTAL |T OF BONDS, MORT GAGES OR OTHER SECURITIES (If there are none, so state) mec none 9. FOR COMPLETION BY NONPROFIT ORGANIZATIONS AUTHORIZED TO MAIL AT SPECIAL RATES (Section 423. bee a nt. The purpose, function, and nonprofit status of this is organization and the exempt status for Federal income t sh purpoess iek ) as (2) jae NOT CHANGED DURING HAS CHANGED DURING If chang ublish bmit explanation of x] PRECEDING 12 MONTHS PRECEDING 12 MONTHS Saar rn pe apeiseanun og — 10. A AcT OPIES OF SINGLE EXTENT AND NATURE OF CIRCULATION ISSUE DURING Por ECEOING: ISSUE PUBLISHED NEA NEAREST TO te MONTHS _ FILING DATE __ A. TOTAL NO. COPIES (Net Press Run) 1350 1400 8. PAID 1D CIRCULATION Sales through dealers and carriers, street vendors and counter sales 16) 0 2 ee 2. Mail i Subscription 1214 1222 €. TOTAL PAID CIRCULATION (Sum of 10B1 and 10B2) 1214 i222 O. FREE DISTRIBUTION BY MAIL, CARRIER OR OTH SAMPLES, COMPLIMENTARY, AND OTHER FREE comes 1@) 0 E. TOTAL DISTRIBUTION (Sum of C and D) 1214 1222 F. Copies NOT OISTRIBUTED . Office use, left over, unaccounted, iled af inting 136 178 ed 2. Return from News Agents fe) (@) es G. TOTAL (Sum of E, Fl and 2-should a ee 1350 1400 id iW 3 EA , oad ~ SIGNATURE ee LE OF EDITOR, PUBLISHER OR OWN me pein are correct and complete : Alan R. Smith, editor oe PS Form : July 1982 3526 See tn ) INFORMATION FOR AUTHORS Authors are encouraged to submit manuscripts pertinent to pteridology for publication in the American Fern Journal. Manuscripts should be sent to the Editor. Acceptance of papers for publication depends on merit as judged by two or more referees. Authors are encouraged to contribute toward publishing costs; however, the payment or non-payment of page charges will affect neither the acceptability of manuscripts nor the date of publication. Authors should adhere to the following guidelines; manuscripts not so pre- pared may be returned for revision prior to review. Submit manuscripts in trip- licate (xerocopies acceptable), including review copies of illustrations. Do not send originals of illustrations until they are requested. Use standard 8% by 11 inch paper of good quality, not “erasable” paper. Double space manuscripts throughout, including title, authors’ names and addresses, text (including heads and keys), literature cited, tables (separate from text), and figure captions (grouped as consecutive paragraphs separate from figures). Arrange parts of manuscript in order just given. Include author's name and page number in upper right corner of every sheet. Provide margins of at least 25 mm all around on typed — pages. Avoid footnotes and do Bot break words at ends of lines. Make table headings and figure cap tory. Use S.I. (metric) units for all mea- sures (e.g., distance, elevation, weight) unless ‘quoted or cited from another source {e.g., specimen citations). For nomenclatural matter [i.e., synonymy and typifi- cation), use one paragraph per basionym (see Regnum Veg. 58:39-40. 1968). Ab- breviate titles of serial publications according to’ Botanico-Periodicum-Huntian- um (Lawrence, G. H. M. et al., 1968, Pittsburgh: Hunt Botanical Library). References cited only as part of nomenclatural matter are not included in lit- erature cited. For shorter notes and reviews, put all references parenthetically in text. Use Index herbariorum (Regnum Veg. 106: 1-452. = for a of herbaria. L Illustrations should be proportione 1 to fit - page. Provide margins of at least 25 mm on all | illustrations. For continuous-tone illustrations, design origit pe amount. In composite blocks, abut edges of adjacent photographs. Sein S se width with caption on the: same : sequence and numbering of figures s (and of of tables) with order of citation in text. oo Explain ot a themselves, not in capti tic _ Include : map. a J +, in each scale and reference to nd longitud thors by one hashes cackeca o the wir aoe reprint ‘orders 1 to - printer. - a for extensive alterations made a 9e has ction without reduction or by a uniform