245 FERN GAZ. 17(5). 2006 British Pteridological Society wil Royal Botanic Garden Edinburgh Linnean Society of London Ferns for the 21st Century Proceedings of the International Pteridophyte Symposium at the Royal Botanic Garden Edinburgh, Scotland, UK 12-16 July 2004 Part 3 Editors: M. Gibby, A. Leonard & H. Schneider MISSOURI BOTANICAL ROYAL BOTANIC GARDEN EDINBURGH 246 FERN GAZ. 17(5). 2006 FERN GAZ. 17(5): 247-257. 2006 247 PHYLOGENETIC SYSTEMATICS AND EVOLUTION OF THE GENUS HYMENOPHYLLUM (HYMENOPHYLLACEAE: PTERIDOPHYTA) S. HENNEQUIN'|, A. EBIHARA’, M. ITO’, K. IWATSUKI & J.-Y. DUBUISSON! No 'Laboratoire de Paléobotanique et Paléoécologie, Université Pierre et Marie Curie, | rue Cuvier, F-75005 Paris, France. (Email: sabine.hennequin@snv.jussieu.fr) "Department of System Sciences, Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan. The Museum of Nature and Human Activities, Hyogo, Yayoigaoka 6, Sanda 669-1546, Japan. Key Words: Hymenophyllum, Hymenophyllaceae, phylogeny, morphological evolution, rbcL, rps4-trnS, rbcL-accD, morphology, cytology. In this study we address the phylogenetic relationships within the genus Hymenophyllum. Our sampling includes the segregate monotypic genera Cardiomanes, Serpyllopsis, Rosenstockia, and Hymenoglossum, representatives of the five subgenera proposed for Hymenophyllum by Morton, and of the section Microtrichomanes. Using morphology, cytology, and nucleotide sequences (rbcL, rps4-trnS, rbcL-accD), we obtained a fully resolved topology with several clades well supported. We confirm the monophyly of two clades within the Hymenophyllaceae. Serpyllopsis and Rosenstockia are nested in Hymenophyllum within a derived clade, while Cardiomanes and Hymenoglossum are positioned within a basal grade. Although some of the phylogenetic associations that were previously proposed within Hymenophyllum are supported, many traditionally defined infrageneric taxa are not resolved as monophyletic: subg. Hymenophyllum and Sphaerocionium are paraphyletic, and the broad subg. Mecodium, whose homogeneity had never been questioned, appears polyphyletic. INTRODUCTION The debate surrounding the systematics of the Hymenophyllaceae originates mainly from the existence of intermediate shapes between the bivalved sorus typically described for Hymenophyllum Sm. and the tubular one characterising Trichomanes L. (Iwatsuki, 1977), as well as from the existence of very peculiar taxa. Consequently, many authors refuted the bigeneric system and proposed a number of genera ranging from 6 to 42 (Copeland, 1938, 1947; Morton, 1968; Pichi Sermolli, 1977; Iwatsuki, 1984, 1990). Recent phylogenetic studies (Pryer ef al., 2001; Ebthara et al., 2002: Hennequin ef a/., 2003) have suggested the existence of two clades, one corresponding to Trichomanes s.\., the other to Hymenophyilum s.1. This latter notably includes the "intermediate" taxa treated as separate genera even by advocates of an oligogeneric system (Morton, 1968; Iwatsuki, 1984, 1990) : Cardiomanes C.Presl, Serpyllopsis Bosch, Rosenstockia Copel. and Hymenoglossum C.Presl. It also includes many species of the genus Microtrichomanes sensu Copeland (1938, 1947), a taxon previously 248 FERN GAZ. 17(5): 247-257. 2006 considered closer to Trichomanes than to Hymenophyllum by most authors (Ebihara et al., 4 In addition to the issue of the circumscription of Hymenophyllum, specialists of the family encountered another problem when studying the systematics of the taxon, even in a strict sense. The genus consists of about 350 species, which grow predominantly as epiphytes. In comparison with the Trichomanes lineage, the Hymenophyllum lineage appears quite homogeneous morphologically and ecologically with all species having long creeping rhizomes with thin roots and mostly pendant leaves. A thorough study reveals, however, that many features are quite variable. These are, for example, the sorus morphology (Iwatsuki, 1977), but also the indumentum, the stele anatomy, and the chromosome numbers (Tryon and Tryon, 1982). Copeland (1937) thus noted that "Hymenophyllum is even less homogeneous than Trichomanes". Nevertheless, this variability does not provide reliable characters for the systematics of the genus, and Copeland (1937) deplored "the absence of single conspicuous criteria for the recognition of the natural groups within Hymenophyllum". Several groups have however been proposed for Hymenophyllum since the 19" century. The four major 20" century classifications of the genus are reported in Table 1. The delimitation of the taxa proposed does not vary much among these classifications. The main conflicts lie in the taxonomic rank attributed to the taxa and in the affinities suggested among them. Using a combination of two chloroplastic regions, Ebihara et a/. (2002), Ebihara et al. (2003) (rbcL + rbcL-accD) and Hennequin et al. (2003) (rbcL + rps4-trnS) provided insights into the relationships within Hymenophyllum. By sampling 25 species of the Hymenophyllum \ineage, Hennequin et al. (2003) obtained three main clades: 1) Hymenophyllum s.s., corresponding globally to the subg. Hymenophyllum Sm. proposed by Morton (1968) and including the genera Rosenstockia and Serpyllopsis, and the subg. Craspedophyllum C.Presl and Hemicyatheon Domin.; 2) a clade including several species of the subg. Mecodium Copel.; and 3) a clade composed of species of subg. Sphaerocionium (C.Presl.) C.Chr. including species of Microtrichomanes (Mett. ex Pantl) Copel. Cardiomanes reniforme (G.Forst.) C.Presl and Hymenoglossum cruentum (Cav.) C.Presl were retrieved at the base of the tree along with one species of Mecodium. Nevertheless, several clades lacked bootstrap support. In this study, we combine and extend the taxonomic sampling used by Ebihara er al. (2002), Ebihara ef al. (2003) and Hennequin ef al. (2003) and add rbcL-accD or rps4-trnS data for all species. In addition, we combine molecular with morphological and cytological data. MATERIALS AND METHODS Taxonomic sampling We adopt here Morton's (1968) classification for the purpose of presenting the results. The name of the section sensu Morton (1968) will thus be placed in parenthesis when needed. In addition to the taxonomic samplings used by Ebihara er al. (2002), Ebihara et al. (2003) and Hennequin ef al. (2003), our sampling (Table 2) includes H. marginatum Hook. & Grev., the type species of Morton's (1968) subgenus Craspedophyllum, three species of Mecodium (H. australe Willd., H. demissum (G.Forst.) Sw. and H. polyanthos (Sw.) Sw. and H. (Sphaerocionium) hirsutum (L.) Sw. This sampling represents all but four out of the 11 genera proposed by Copeland (1938) for Hymenophyllum s.1., and all but three out of the 10 sections proposed by Morton (1968) for the genus. We first performed an analysis based on rbcl using a broad sampling including five 7richomanes taxa and five non-Hymenophyllaceae as 249 HENNEQUIN et al.: THE GENUS HYMENOPHYLLUM jeoido.noayed “Ld ‘[eoidonuegd ‘g {pueyeaz MAN “ZN ‘BIUOpaye MON ‘ON ‘RoIdONOIN “LN ‘ueywpodowisoy ‘D ‘aplyD “yO ‘eipeysny ‘sny ‘eisy “y Seunuadry “Sry suoNNquysip (2) ‘eypyjaqvy.7 MwBU JeUOIDAS Paoe|dunN sy) Jopun saupuoyoidy Ul papnyaul (g) ‘saupwopidasy UOIdAS SAUDWIOYILAT UL papnyoul (¢) tapploupuolpav.y A{iwieJ-qns ayi Jo aAneRyuasaidas onbiun (p) : wnpAydojdig, uoidas-qns dy) Surpnypoul (¢) SUOdIS-qns vB SB WIIPOIAaP UL papnyoul (Z) ‘UONdaS-qns eB se WNIUOLIO4IDYdGs Ul papnyoul (7) id 6 (s) f1-AIOLDI (9) { NALOADIYN ZN I (p) SauDWwOIpaDy ~—- SauDuOIpaDy SOUDUOIPAD) SIUDUOIPAD ZN ‘sny ré staajdosajdp stiajdosajdp stuajdosajdp siiajdosajdp d OL wniuowosavyds wniuotosavydy = uniuotosapydy = wniuotsosapydy — uiniuotoo4avyds wniuotosapyds B1V-YD I uniwuolojdaT wniuo1oj}daT (1) wniuolojdayT ZN ‘sny c uoayjDAIIUAT] UOaYyIDAIIWAL] UOIYIDAINUATL] uoayJOAINUALT Vv I uopolsAy UOpOlIAy UOPOLlAW UOpolIAPy a OL wnipojlyy) wnipojlyD WNIBULLAY wnppAydoy rad WNIBULAIN LN ¢ pisang pisang pisang pisang V b wnasajdiydup wnasajdiydup (Z) wnypAydouauAy] wniajdiydup 31Vv-Y4D b wnulyoadny wnutjoadny ay €¢ wnyaAydouaudyy unpjAydouamAp wnpAydouawudy —uinyjAydouauAyy — wing AydouauAy] wnjjAydouaumapy —wungjAydouaudAy] ZN ‘shy Z DUOJAYID| wnjjAydopadspay wnypjAydopadsp4y avIDBNAMOD ZN ‘sny t unyrydojdiq d OOI < wniporay wniposay wniporapy (¢ ‘Z) wniposapy wuNIpOIaW wNIpOIIW uinipo2aumopnasd DSOUN] gq ON I DIYIOJSUISOY DIYIOISUASOY D1YIOISUISOY DIYIOISUASOY BIV-YD I sisdopjAdias sisdopjAduag sisdopjAdsag sisdopjAduag yo I WNSSO]BOUIUAT] —UINSSO]/BOUIUAT] WNSSO[BOUIUAPY UNS SO] BOUIUAT] SUOT]IIS piouas-qnsg vIoudy PIOUdN) SUOT}Iag B1guad-qns plguan RIOUdN) (L) soloads jo (0661861) (LL61) (8961) (Lrol “8¢ol) uolNLsig JoquinN ryns}ea| TJOULIS 1ydI1q uo}LIO| purjadoy UUM “(0661 “UONNGLYSIP dy) puv ‘sUOT}OdS PaIpnys dy} UI JoquINU satoads “P861) Fynsyemy pur (L161) HIOWLIg 1YydId (3961) VOLOW “(Lr6I “8E61) PUL|adoD Jo suoNKoYtsse[d oy) Jo UosLIVdWOD “] a1qUL 250 FERN GAZ. 17(5): 247-257. 2006 Table 2. Names and sources of material sequenced, with GenBank numbers. LPP = Laboratoire de Paléobotanique et Paléoécologie, H = Hymenophyllum,; T = Trichomanes. Cardiomanes reniforme (G. Forst.) C. Presl - rbcL U30833, rbcL-accD AB083290 (Ebihara 011222-07 New Zealand (TI, CHR)), rps4-trnS AY095132 (Rumsey s.n., cult. RBG Kew); H. acanthoides (Bosch) Rosenst. - Ebihara Kinabalu 030, Malaysia (TI), rbcL AB064291, rbcL-accD AB064303, rps4-trnS DQ364196; H. apiculatum Mett. ex Kuhn - Dubuisson HV 1997-23 Venezuela (LPP, Duke), rhcL AF275642, rbcL-accD AY 775438, rps4-trnS AY095131; H. armstrongii (Baker) Kirk - Smith 2610 New Zealand (UC), rbcL AY095109; rbcL-accD AB162691 (Ebihara 011219- 09 New Zealand (TI)), rps4-trnS AY095128; H. australe Willd. — T. A. Ohsawa 001125-03, Australia (TI), rhbcL AB191439; rbcL-accD AB191439; rps4-trnS AY775412; H. bailevanum Domin.- Streimann s.n. Australia (UC), rbcL AF275643, rbcL-accD AB191441 (Ebihara 010909-02, Australia (TI)), rps4-trnS AY095129; H. barbatum Baker - Ebihara 000319-01 Japan (TI) rbcL AB064287, rbcL-accD AB064299, rps4-trnS AY095124 (Munzinger & Engelmann 297 Laos (Duke)); #. deplanchei (Mett.) Copel.- Ebihara 001224-03 New Caledonia (TI), rbcL AB064288, rbcL-accD AB064300, rps4-trnS AY095136 (Munzinger 367 New Caledonia (P)); H. demissum (GForst.) Sw. - Glasgow B. G. 830, cult. RBG Edinburgh, rbcL AY775402, rbcL-accD AY775441, rps4-trnS AY775416; H. dilatatum (G.Forst.) Sw. - Brownsey & Birchard New Zealand (Duke), rbcL AY095111, rbcL-accD AB191444 (Ebihara 011219-06, New Zealand (TI)), rps4- trnS AY 095138; H. dimidiatum Mett. - Ebihara 001225-08 New Caledonia (TI), rbcL AB064289, rbcL-accD AB064301, rps4-trnS DQ364197; H. ferrugineum Colla - Taylor 6074 Chile (UC), rbcL AF275644, rbcL-accD AB191445 (Ebihara 021224- 02, Chile (TI), rps4-trnS AF537124; H. flabellatum (G.Forst) Sw. - Unknown collector 42 Tahiti (UC), rbcL AY775403, rbcL-accD AY775442, rps4-trnS AY 775417; H. fucoides (Sw.) Sw. - Dubuisson HV-1997-9 Venezuela (Duke), rbcL U20933; rbcL-accD AY 775449, rps4-trnS AY095142; H. fuscum (Blume) Bosch - Ito 2000 0210-16 Java (TI), rbcL AB064292, rbcL-accD AB064304, rps4-trnS AY775408; H. hirsutum (L.) Sw.- Dubuisson HR-1999-6 La Réunion (LPP, Duke), rbcL AY775407, rbcL-accD AY775450, rps4-trnS AY775432; H. hygrometricum (Poir.) Desv. - Dubuisson HR-1999-13 La Réunion (LPP, Duke), rhcL AY095113; rbcL-accD AY775451, rps4-trnS AY095118; H. lanceolatum (Hook.& Arn.) Copel. - O'Brien s.n. Hawaii (UC), rbcL AF275646, rbcL-accD AY775452, rps4-trnS AY 095119; H. oligosorum Makino - Ebihara 001105-01 Japan (TI), rbeL AB064293, rbcL-accD AB064305, rps4-trnS AY775422; H. paniense Ebihara et al. - Ebihara 001225-02 New Caledonia (P, TI, KYO, NOU), rbcL & rbcL-accD 001225-02, rps4- irnS AY775410; H. pectinatum Cav. - Wedin H41, Chile (UC), rbcL AY095115, rbcL-accD AB191450 (Asakawa 2017, Chile), rps4-trnS AY095 134; H. polyanthos (Sw.) Sw. - Ebihara 991122-01 Japan (TI), rbcL AB064295, rbcL-accD AB064307, rps4-trnS AY 775423; H. polyanthos (Sw.) Sw. — Dubuisson s.n. La Réunion (LPP), rbcL AY775405, rbcL-accD AY775445, rps4-trnS AY775424; H. scabrum A.Rich. - Ebihara 011223-05 New Zealand (TI), rbcL AB083278, rbcL-accD AB083278, rps4-trnS AY775428; H. secundum Hook. & Grev. - Taylor 6075 Chile (UC), rbcL AF275648, rbcL-accD AY775437, rps4-trnS AY095125; H. sibthorpioides Mett.- Dubuisson HR-1999-1 La Réunion (LPB, Duke), rbcL AY095117, rbcL-accD HENNEQUIN et al.: THE GENUS HYMENOPHYLLUM 251 AB192688, rps4-trnS AY095127; H. subdimidiatum Rosenst. — Ebihara 001226-01 New Caledonia (TI), rhcL AB064290, rbcL-accD AB064302, rps4-trnS AY 095140; H. tenellum Kuhn. - Dubuisson HR-1999-27 La Réunion (LPP, Duke), rbcl AY095116, rbcL-accD AB191453, rps4-trnS AY095126; H. tunbrigense (L.) Sm. - Dubuisson NV. 2.1 France (LPP, Duke), rhcL Y09203?, rhcL-accD AY 775436, rps4- trnS AY 095123; H. wrightii Bosch - Ebihara 000901-01 Japan (TI), rhcL AB064294, rbcL-accD AB064306, rps4-trnS AY775430; Hymenoglossum cruentum (Cav.) C. Presl - Wedin H38 Chile (LPP), rbcL AY095107, rbcL-accD AB191455 (Ohsawa 2015, Chile), rps4-trnS AY 095133; Rosenstockia rolandi-principis (Rosenst.) Presl - van der Werff 16045 New Caledonia (UC), rbcL AY095110, rbcL-accD AB064286/AB04298 (Ebihara 001225-11 New Caledonia (TI)), rps4-trnS AY095143; Serpyllopsis caespitosa (Gaudich.) C.Chr. - Taylor 6076, Chile (UC), rbcL AF275649, rbcL-accD AB191456 (T. A. Ohsawa 2014 Chile (TI)), rps4-trnS AY095130; T. digitatum Sw. - Dubuisson HR-1999-11' La Réunion (LPP, Duke), rbcL AY095114, rbcL-accD AB162676, rps4-trnS AY 095120; T. javanicum Blume - Hennequin 2001-7 cult. Indonesia (LPP), rhcL Y09195, rbcL-accD AY 775453, rps4- trnS AY095141; T. rigidum Sw.- Dubuisson HV-1997-3 Venezuela (Duke), rbcl AY095108, rbcL-accD AY775447, rps4-trnS AY095137; T. taeniatum Copel. - Matsumoto 01-948 Vanuatu (TNS, TI)), rhcL AB162681, rbcL-accD AB162681, rps4-trnS AY095121 (Game 86/08, Cook Islands (UC)); T. tamarisciforme Jacq. - Dubuisson HR-1999-32 La Réunion (LPP, Duke), rbcL Y09202°, rbcL-accD AY775448, rps4-trnS AY 095135; out-groups, as in Hennequin ef al. (2003), to confirm the monophyly of the two clades within the family (results not shown). We then reduced the out-group to three species of Trichomanes as in Hennequin et al. (2003). All three markers were sequenced for the whole sampling, so that 44 sequences are newly produced. Morphological and cytological characters The coding of morphological and cytological character state changes is based on Hennequin (2003) and we used in this study the same matrix as in Hennequin (2004). Morphological and cytological characters states changes were reconstructed using MacClade 3.0. (Maddison and Maddison, 1992). All characters were treated as unordered and plotted onto the topology recovered in the maximum parsimony combined analysis (results not shown but used in the discussion). We performed and compared both ACCTRAN and DELTRAN optimization options. DNA sequencing All procedures for DNA extraction, amplification and sequencing follow Hennequin et al. (2003). We used primers rbcL1195F (5'-TTCTACAGTTCGGTGGTGG-3';, newly designed) and accD816R (5'-CCATGATCGAATAAAGATTCA-3'; Ebihara ef ail., 2003), and newly designed internal primers HIF3 (5'-TGTCAGGTTCTAAC- ATGTGATTG-3') and HIR3 (5'-CCTATACCTGTTTGAACAGCATC-3') to amplify and sequence rbcL-accD, respectively. Phylogenetic analyses We treated indels as binary characters following Barriel’s (1994) method. MrBayes 3.0 (Ronquist and Huelsenbeck, 2003) allows the integration of data other than nucleotide Pin F FERN GAZ. 17(5): 247-257. 2006 or protein sequence, so that morphological and cytological characters as well as characters resulting from the treatment of indels were integrated both in parsimony analyses ("MP"; run with PAUP*4.0b10; Swofford, 2001) and in likelihood analyses ("ML"; run with MrBayes 3.0). For MP, we conducted unequally weighted analyses as described in Pryer et al. (2001) and Hennequin et al. (2003). All searches used a heuristic approach (TBR branch-swapping, 10 replicates of random sequence addition, MulTrees option on). The robustness of each branch was assessed by bootstrap analysis (100 replicates; Felsenstein, 1985). For likelihood analyses, we used ModelTest 3.06 (Posada and Crandall, 2000) to determine the nucleotide substitution model that best fits our data. We performed ML analyses using a random tree and a GTR + I + G model. Clade credibility values were estimated by calculating the posterior probability for each node using a Bayesian procedure as implemented in MrBayes 3.0. 10,000 trees were sampled and the consensus tree was computed (PAUP*4.0b10; Swofford, 2001) on the last 9,450 trees, excluding the 550 trees found in the "burn-in period". RESULTS AND DISCUSSION Delimitation of Hymenophyllum s.1. In both the MP strict consensus tree and the majority—rule consensus tree inferred from the Bayesian estimation of posterior probabilities (Figures 1 and 2), the four segregate genera Serpyllopsis, Rosenstockia, Cardiomanes and Hymenoglossum are embedded within Hymenophyllum s.\. These results are in agreement with previous phylogenetic studies and question all the classifications that were proposed for the family. They nevertheless confirm some hypotheses of affinity proposed between taxa and the group successively treated as genus "Hymenophyllum s.\." (Pryer et al., 2001), "Hymenophyllum group", or tribes Hymenophylleae (Schneider, 1996) or Hymenophylloideae (Presl, 1843; Iwatsuki, 1984, 1990). The following taxa had been considered close to, or treated as included in, Hymenophyllum s.1.. Hymenoglossum (Presl, 1843; Pichi Sermolli, 1977; Schneider, 1996 who included it in his tribe Hymenophylleae), Serpyllopsis (Presl, 1843; van den Bosch, 1861; Schneider, 1996), Rosenstockia (Tryon et Tryon, 1982; Iwatsuki, 1984, 1990), and Microtrichomanes (Copeland, 1938; Pichi Sermolli, 1977). On the other hand, no author had suggested such a broad group, and even less the inclusion of Cardiomanes reniforme and T. (Pleuromanes) pallidum Blume in Hymenophyllum. Hymenophyllum 5.1. systematics Morton (1968) proposed five subgenera for Hymenophyllum: Mecodium, Craspedophyllum, Hymenophyllum, Hemicyatheon and Sphaerocionium (Table 1). The subg. Hymenophyllum is further divided in five sections: Hymenophyllum, Ptychophyllum, Eupectinum, Myriodon and Buesia. Representatives of the first three only were available for this study. This subgenus globally corresponds to the clade Hymenophyllum s.s. retrieved in this study. Hymenophyllum s.s. (bootstrap support (BS) = 96%, posterior probability (PP) = 0.56) also includes taxa placed in the subgenera Hemicyatheon, Craspedophyllum, Mecodium (H. oligosorum and H. fuscum), and the species of the segregate genera Serpyllopsis and Rosenstockia. This clade is supported by several morphological apomorphic character states: the presence of a common hair type on fronds, a margin denticulation and five soral characters, the most conspicuous being the presence of a small thick base (Hennequin, 2004). This clade has a cosmopolitan distribution, with a few species observed in Europe and probably in North HENNEQUIN et al.: THE GENUS HYMENOPHYLLUM 253 America as well. Within the Hymenophyllum s.s. clade, relationships are poorly resolved and weakly supported, except for a very robust clade (H. barbatum - R. rolandi-principis, BS = 100%, PP = 1) already retrieved by Ebihara er al. (2002) and named "H. acanthoides clade". This exclusively Australasian clade includes many taxa of the section Ptychophyllum and is supported by the basic chromosome number x = 21. It may be broadened to include the two species of Hemicyatheon (BS = 95%, PP = 1). The inclusion, in this clade, of H. fuscum (Blume) Bosch, type species of the genus Amphipterum sensu Copeland (1938), was already obtained by Ebihara er al. (2003). Amphipterum thus appears close to species of the section Ptychophyllum sensu Morton (1968), in agreement with Copeland (1938), Pichi Sermolli (1977) and Iwatsuki (1984, 1990). Our results, on the other hand, refute Morton (1968)'s treatment of Amphipterum H. (Hy phyllum) barb H (Me bas ig H. ( dt ) th. a, H. aatie ene? ‘USCUI H. (Ptychophyllum) eas taaieie H. (Ptychophyllum) dimidiatum A aiaaiaiel sisilieicbdiciad H. ( Hymen cosmopolitan oan Australasia H. tomaythec) baileyanum and Southern South America) H. ( = 11,12, 13, 14, 1 8, 21 H. (ychopnyium fenaltan 400° H. (Craspedophyllum) armstrongii H. (Crasped se marginatum Serpyllopsis caespitosa = H. oe ee —— H. rics meres fucoides sate H. (Eupectinum) pectinatum a thos clade 0-H. (Mecodium) po sa anthos Japan paryar Ps ics + temperate areas re{L Hi (Mecodium) wri ia a 100, H. ( . (Mecodium) sadvalitbias La Réunion 1oor_ 7. (Microtrichomanes) digitatum 81 E Sa taeniatum Sphaerocionium s.I. Tropics + South 100-7 —(#H. eee hirsutum La Réunion temperate areas H. (Sp sasteuesienis laid um ‘= H. ( metri H H. (Meco australe 1 93° 5 aft Ses ca flabellatum T. (Pleuromanes) pallidum basal g 95 __sfl he ‘oups Hymenophylium paniense Au rol + Southern South America H. (Mecodium) scabrum 6 j qecancanete aiid I aa iomanes reniforme T. (Cephalomanes) java : bagel: cabelas famansotome Figure 1. Single most parsimonious tree recovered from unequally weighted analysis of combined rbcL + rbcl-accD + rps4-trnS and morphological and cytological characters; tree length = 6056.93 steps, CI = 0.6178, RI = 0.6105 (Farris, 1989). Numbers at nodes are bootstrap values > 50%. 254 FERN GAZ. 17(5): 247-257. 2006 as a subsection in the subg. Mecodium. The remaining species of Hymenophyilum s.s. display a very variable soral morphology and various chromosome numbers. The acquisition of new cytological data, currently under study (Ebihara and Hennequin) may provide further insights into the relationships of these taxa. With more than 100 species, Mecodium is a diverse putative subgenus whose monophyly had never been questioned (apart from the position of Amphipterum), mainly because it was considered to exhibit a high morphological homogeneity. Nevertheless, the subgenus appears characterised rather by the absence, than by the presence, of peculiar characters, such as margin denticulation or hairs on fronds. Our results confirm that Mecodium is polyphyletic as suggested by Hennequin et al. (2003). Several Mecodium species form a derived well supported (BS = 100%, PP = 1) clade, sister to Hymenophyllum s.s. (BS = 98%, PP = 1). This clade, named H. polyanthos reer H. (Amphipterum) fuscum H. (Pty hy, yohy HH } AA, Pry H. (Pty h, 10 hy Hi )) +h, H. (Hymenophyllum) barbatu H. AA, i ) ti, H. (Ptychophyllum) dimidiatum Rosenstockia rolandi-principis i Yy p Hymenophyllum s.s. Hi (Hemicvathecn\ bail cosmopolitan (mainly Australasia i: 7 at and Southern South America) H. (1 Pt hy, Avi } jAth. Ai H. (Ptychophyllum) tenellum H. (Cras, ~phyllum) armstrongii H. (Hymenophyilium) tunbrigense H. (Eupectinum) pectinatum 0.997 H. (Mecodium) polyanthos Japan 1 H. (Mecodium) wrightii H. (Mi Adis ) 4 as H. polyanthos clade Tropics + temperate areas x = 28 H. (Mecodium) polyanthos La Réunion tum T. (Microtrichomanes) taeniatum Sphaerocionium s.1. H. (Sphaerocionium) lanceolatum Tropics + South temperate areas x= 36 g > : 2. > = 3 v 3 ¢ 3 g H. (Sphaerocionium) hygrometricum H. AA, i ) v9 I 1 eee H. (Mi iio 1p- H. (Mecodium) flabellatum 0.99 L__serL T. (Pleuromanes) pallidum | basal groups ae Australasia + Southern South America Os. x= 36 : 1 Hymenoglossum cruentum Hymenophyllum paniense H. (AAG mH ) Ps H. (Me, oti ) Ailatat, Cardiomanes reniforme Te (Canhal, wi T. (Pachy A eiceanrilapenmopaijan area J, T. (Pachyc. Figure 2. Majority-rule consensus tree of the 9,450 trees sampled during the Bayesian analysis of combined rbcL + rbcL-accD + rps4-trnS and morphological and cytological characters. Numbers at nodes are Bayesian posterior probabilities > 0.90. cirorme haetum) rigidum HENNEQUIN et al.: THE GENUS HYMENOPHYLLUM 255 clade based on one of the type species proposed for Mecodium, H. polyanthos (Sw.) Sw. (Morton, 1968; Pichi Sermolli, 1977; Iwatsuki, 1984, 1990), is supported by one exclusive apomorphic character state, the basic chromosome number x = 28. The H. polyanthos clade is distributed throughout the Neotropics, the Paleotropics and into temperate regions (Japan). Another species of subg. Mecodium, H. oligosorum, is nested within Hymenophyllum s.s. This position is in agreement with Iwatsuki (1984) and is notably supported by cytological data (x = 21; Tatuno and Takei, 1969). The remaining species of Mecodium included in our sampling are retrieved at the base of the tree in a grade comprising also the genera Cardiomanes and Hymenoglossum. These basal taxa can be distinguished from the H. polvanthos clade by characters that appear plesiomorphic for the genus, i.e. the chromosome number x = 36, also present in Sphaerocionium s.|., a stele where internal xylem is reduced in comparison to the massive protostele (reduced or dorsi-ventral stele) but not subcollateral as in the other clades of the genus, and a tendency to display a rougher habit than the other species of Hymenophyllum (thicker rhizomes, large fronds, and for some species a thicker (2-4 cells thick) lamina). In addition, these basal species have a typical austral distribution (Chile, New Zealand, Australia, New Caledonia). Species of the subg. Sphaerocionium are all retrieved in a clade named Sphaerocionium s.1. (BS = 90%, PP = 1), which includes in addition some species of the problematic section Microtrichomanes. This clade appears sister to Hymenophyllum s.s. + H. polyanthos clade (BS = 71%, PP = 0.65). The monophyly of Sphaerocionium was not questioned, as it was characterised by distinctive stellate hairs at the lamina. Morton (1968) divided his subgenus in two sections: Sphaerocionium and Apteropteris, and according to analyses not shown here (Hennequin, 2004) the latter is embedded in Sphaerocionium s.1. Microtrichomanes was placed by several authors in Trichomanes but Copeland (1938) suggested an affinity between this taxon and Sphaerocionium. Our results corroborate his hypothesis and they are discussed in more detail in a related study (Ebihara er al., 2004). Sphaerocionium s.1. is a pantropical clade, with some taxa expanding in southern temperate areas. It is supported by several apomorphic character states, the most conspicuous being the presence of marginal setae, with stellate marginal hairs in Sphaerocionium and unicellular marginal hairs in Microtrichomanes. It has been proposed that the latter hair type is the result of a reduction from stellate hairs (Copeland, 1938; Holttum, 1963). Craspedophyllum and Hemicyatheon are small subgenera, made up of only two species. The monophyly of both subgenera is confirmed here, but there is no support for their treatment at the subgeneric level. A treatment as sections, as realised by Iwatsuki (1984, 1990), appears more appropriate. CONCLUSIONS AND PROSPECTS The results obtained in this study corroborate those previously obtained on a reduced sampling and using less data (Ebihara ef al., 2002; Hennequin, 2003; Hennequin et al., 2003: Ebihara er al., 2003). New results include the monophyly of the two species of subg. Craspedophyllum and the confirmed polyphyly of subg. Mecodium. The analyses based on three molecular markers combined with morphological and cytological data provide much better support for the clades previously obtained, although robust support is still lacking for the clades retrieved within Hymenophyllum s.s. and for the basal relationships within the Hymenophyllum lineage. By exploring the branch-length distribution, we recognized the short distances between clades in the poorly supported 256 FERN GAZ. 17(5): 247-257. 2006 parts of the phylogeny. This could be either explained as the result of rapid radiations or through low mutation rates. Hopefully, by broadening the taxonomic sampling and adding more molecular, morphological and cytological data, we will be able to improve our understanding of the complex history of the genus. REFERENCES BARRIEL, V. 1994. Phylogénies moléculaires et insertions-délétions de nucléotides. C. R. Acad. Sci., Sciences de la Vie, Paris 317: 693-701. COPELAND, E.B. 1937. Hymenophyllum. Philip. J. Sci. 64: 1-196. COPELAND, E.B. 1938. Genera Hymenophyllacearum. Philip. J. Sci 67: 1-110. COPELAND, E.B. 1947. “Genera Filicum”, Chronica Botanica, Waltham, Mass., USA. EBIHARA, A., IWATSUKI, K., KURITA, S. & ITO, M. 2002. Systematic position of Hymenophyllum rolandi-principis Rosenst. or a monotypic genus Rosenstockia Copel. (Hymenophyllaceae) endemic to New Caledonia. Acta Phytotax. Geobot. 53: 35-49. EBIHARA, A., IWATSUKI, K., OHSAWA T.A. & M. ITO. 2003. Hymenophyllum paniense Cigmenopiivilanes), a new species of filmy fern from New Caledonia. Syst. Bot. 28: 228-235. EBIHARA A., HENNEQUIN S., IWATSUKI K., BOSTOCK P.D., MATSUMOTO S., JAMAN R., DUBUISSON, J.-Y. & ITO. M. 2004. Polyphyletic origin of Viruntdhoncnws (Prantl) Copel. (Hymenophyllaceae) with a revision of the species. Taxon 53: 935-948 FARRIS, J.S. 1989. The cui index and the rescaled consistency index. Cladistics 5: 417-419. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791. HENNEQUIN, S. 2003. Phylogenetic relationships within the fern genus Hymenophyllum s.l. (Hymenophyllaceae, Filicopsida): contribution of morphology and cytology. Comptes Rendus Biologies 326: 599-611. HENNEQUIN, S. 2004. Le genre Hymenophyllum Sm. (Hymenophyllaceae, Filicopsida) : systématique phylogénétique, évolution morphologique et histoire biogéographique. Thése de doctorat de l'Université Pierre et Marie Curie. ENNEQUIN S., EBIHARA A., ITO M., IWATSUKI K. & DUBUISSON J.-Y. 2003. Molecular systematics of the fern genus Hymenophyllum s.1. (Hymenophyllaceae) based on chloroplastic coding and noncoding regions. Mol. Phylogenet. Evol. 27: 283-301. HOLTTUM, R.E. 1963. Flora of Malaysia, serie II Pteridophyta, Vol. 1 part 2. Government Printing Office, Singapore. IWATSUKI, K. 1977. Studies in the systematics of filmy ferns: III]. An observation on the involucres. Bot. Mag. Tokyo 90: 259-267. IWATSUKI, K. 1984. Studies in the systematics of filmy ferns: VII. A scheme of classification based chiefly on the Asiatic species. Acta Phytotax. Geobot. 35: 165-179. IWATSUKI, K. 1990. Hymenophyllaceae. Jn Kramer U. K. and Green P. S. (eds.) “The families and genera of vascular plants”, Vol. I: Pteridophytes and Gymnosperms, 157-163. Springer-Verlag, Berlin, Germany. MADDISON, W.P., & MADDISON, D.R. 1992. MacClade: analysis of phylogeny and character evolution. Sinauer, Sunderland, Mass. HENNEQUIN et al.: THE GENUS HYMENOPHYLLUM Ve # MORTON, C.V. 1968. The genera, subgenera and sections of the Hymenophyllaceae. Contrib. U.S. Nat. Herb. 38: 153-214. PICHI SERMOLLI, R.E.G. 1977. Tentamen Pteridophytorum genera in taxonomicum ordinem redigendi. Webbia 31: 313-512. POSADA, D. & . CRANDALL, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818. PRESL, K.B. 1843. Hymenophyllaceae. Eine botanische Abhandlung. Prag. PRYER, K.M., SMITH, A.R., HUNT, J.S., & DUBUISSON, J.-Y. 2001. rbcL data reveal two monophyletic groups of filmy ferns (Filicopsida: Hymenophyllaceae). Am. J. Bot. 88: 1118-1130. RONQUIST, F., & HUELSENBECK, J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572-1574. SCHNEIDER, H. 1996. Vergleichende Wurzanatomie der Farne. Shaker, Aachen, SWOFFORD, D.L. 2001. PAUP*, phylogenetic analysis using parsimony (* and other methods), ver.4. Sinauer Associates, Sunderland, Massachusetts. TATUNO, S., & TAKEI, M. 1969. Karyological studies in Hymenophyllaceae I. Chromosomes of the genus Hymenophyllum and Mecodium in Japan. The Botanical Magazin Tokyo 82: 121-129. TRYON, R.M., & TRYON, A.F. 1982. Ferns and allied plants, with special reference to tropical America. Springer-Verlag, New Yor VAN DEN BOSCH, R. 1861. Hymenophyllaceae Javanicae. Verhandelingen der Koninklijke Akademie van Wetenschappen Amsterdam. 258 FERN GAZ. 17(5). 2006 FERN GAZ. 17(5): 259-261. 2006 259 MICRO-FUNGAL PTERIDOPHYTE PATHOGENS S.. HELPER Royal Botanic Garden Edinburgh, 20A Inverleith Row, Edinburgh, EH3 5LR Scotland (Email: s.helfer@rbge.org.uk) Key words: Uredinales, rust fungi, pteridophytes. ABSTRACT Of the 225 genera of pteridophytes listed in Kubitzki (1990) there are 131 (58%) genera with no known fungal association, according to the most comprehensive fungal database. The remaining 94 genera are represented by 524 taxa at the species and subspecies level which form about 1848 mainly parasitic interactions with 822 fungal taxa. Around 450 of these interactions are parasitic associations with rust fungi (Uredinales, Basidiomycetes), which are represented by four genera (and two form genera) and around 130 species and subspecies. Fungal synonymies have been resolved as far as possible, however, for this presentation pteridophyte synonymies have only partly been resolved, due to my lack of experience with ferns. This paper examines the taxonomic distribution of fern - fungus interactions in general and the importance of the fern rusts in particular. Examples of interactions are illustrated with the aim of raising awareness among pteridologists and mycologists. INTRODUCTION Pteridophytes are not normally associated with plant disease problems, and, whilst the current information is far from complete, it appears that many genera are relatively free from fungal pathogens. The rust fungi (Uredinales) are the most important pathogens affecting ferns (only one record of rust for Lycopodiatae, none for Equisetatae), with around 100 species in two rust families (Cummins & Hiratsuka, 2003). The largest genus is Milesina with 34 recognised species and 22 species of the form genus Uredo (asexual form only) believed to belonging to this genus. Together with Uredinopsis (26 species and 2 form species) and Hvalopsora (7 species and 5 possible form species) it belongs to the family Pucciniastraceae and uses species of Abies as alternate hosts (Hiratsuka, 1958). The fourth genus, Desmella, is a member of the Uropyxidaceae, infecting ferns in South and Central America. It is only known from its uredinial and telial stages (no known alternate host). MATERIALS AND METHODS For this study the herbarium records and database of the US National Fungus Collection at Beltsville (BPI) (Farr et al. 2004) and other online herbaria were searched, and selected specimens were examined using light and scanning electron microscopy. Over 2500 records were processed, their synonymies cleared as far as possible, and results plotted against relevant host data. 260 FERN GAZ. 17(5): 259-261. 2006 Table 1. The twenty most common genera of microfungi on ferns and fern allies. Table 2. The 20 pteridophyte genera, most commonly attacked by microfungi. Fungal genus Fern hosts Pteridophyte genus Fungi Milesina 118 Pteridium 116 Uredinopsis 69 Athyrium 100 Uredo 60 Dryopteris 96 Hyalopsora ao Equisetum ie Cercospora 47 Cibotium 58 Mycosphaerella 42 Cyathea 58 Desmella 29 Pteris 54 Taphrina 28 Selaginella 47 Phyllosticta 27 Polypodium 44 Rhizoctonia 25 Adiantum 43 Pleospora 24 Lycopodium 38 Pythium 24 Asplenium a2 Phaeosphaeria 22 Osmunda a9 Leptosphaeria 21 Polystichum 33 Trichopeltheca 18 Blechnum 29 Pseudocercospora 16 Nephrolepis 28 Clathrospora 14 Dicksonia A Prytoptor 14 Rumohra af Dasyscyphus 13 Platycerium 20 Fusarium 13 Matteuccia 19 HELFER: MICRO-FUNGAL PTERIDOPHYTE PATHOGENS 261 RESULTS Tables | and 2 show the 20 most common genera of microfungi found on ferns and fern allies, and the 20 pteridophyte genera most commonly attacked by microfungi respectively. The most commonly recorded fern pathogens are the rust genera Milesina, Uredinopsis and Hyalopsora, together with the anamorphic form genus Uredo. Several fern families are host to all the fern rust genera, whereas others are restricted in their susceptibility, such as the Vittariaceae, Lomariopsodaceae (only one rust recorded respectively) and Schizaeaceae (probably all restricted to Desmella), Aspleniaceae and Davalliaceae (restricted to Milesina). Of particular interest is the record of Uredo vetus on Selaginella sp. in China, the only record of a rust on Lycopodiatae (Hennen, 1997). Conversely, there appears to be no obvious pattern in the distribution of fern families as host plants of the four rust genera. Key to the fern rusts: l only urediniospores present Uredo 1’ teliospores present 2 2 teliospores external, 2-celled on pedicells Desmella 2’ teliospores formed inside host epidermal cells 3 3 urediniospores more or less lanceolate, smooth or with a few lines of coglike warts Uredinopsis 3’ urediniospores usually echinulate or verrucose, orange pigment present in cytoplasm Hyalopsora "as above, no pigment present Milesina CONCLUSIONS At present, there is only a limited database on the distribution and frequency of parasitic microfungi on ferns. Further collecting should be encouraged and pteridologists will play a crucial role in this endeavour. REFERENCES CUMMINS, GB. & HIRATSUKA, Y. 2003. Illustrated Genera of Rust Fungi. APS Press, St Paul, Minnesota. FARR, D.F., ROSSMAN, A.Y., PALM, M.E., & MCCRAY, E.B. (no date) Fungal Databases, Systematic Botany & Mycology Laboratory, ARS, USDA. Retrieved 004, from http://nt.ars-grin.gov/fungaldatabases/ HENNEN, J.F. 1997. Uredo vetus sp. nov., the first record of a rust on Selaginella, and the use of the name Uredo. Mycologia 89:801-803. HIRATSUKA, N. 1958. Revision of taxonomy of the Pucciniastreae. Contributions from the Laboratories of Phytopathology and Mycology, Faculty of Agriculture, Tokyo University of Education. 31. 167pp. KUBITZKI, K. (ed.) 1990. The Families and Genera of Vascular Plants. Volume 1. Pteridophytes and Gymnosperms. Springer-Verlag Berlin etc. 404pp. 262 FERN GAZ. 17(5). 2006 FERN GAZ. 17(5): 263-270. 2006 263 PHENOLOGICAL ASPECTS OF FROND PRODUCTION IN ALSOPHILA SETOSA (CYATHEACEAE: PTERIDOPHYTA) IN SOUTHERN BRAZIL J.L. SCHMITT! & P.G. WINDISCH” 'Centro Universitario Feevale, Novo Hamburgo (RS) and PPG-Botanica/UFRGS, Porto Alegre (RS).(Email: jairols@feevale.br) °PPG-Biologia/Universidade do Vale do Rio dos Sinos - UNISINOS. 93022-000 Sao Leopoldo — RS, Brazil (Email: pgw@bios.unisinos.br) Key words: phenology, frond production, growth rates, ecology, spore production. ABSTRACT Two populations of Alsophila setosa Kaulf. in secondary semi-deciduous subtropical forest remnants in the State of Rio Grande do Sul, Brazil were studied with attention to frond formation, expansion and senescence rates, as well as to phenology of sporangia formation and spore release, during a 15 month period. Plants of various sizes were marked at a site at Morro Reuter (45 plants) and another at Sapiranga (48 plants) municipalities. The average frond production rates were 5.51 fronds/year at Morro Reuter, and 4.14 fronds/year at Sapiranga. After frost occurrence in early winter, all the exposed young croziers were irreversibly damaged with necrosis of the tissues. A new set of croziers was formed in October (spring), with all the croziers uncoiling almost simultaneously, 84.4% of the specimens in Morro Reuter and 66.7% in Sapiranga presenting one or more croziers in the initial expansion stages. The senescence rates were 6.97 fronds/year at Morro Reuter, and 4.33 fronds/year at Sapiranga. Low temperatures (including the occurrence of frost) and low rainfall during winter coincide with the highest frond senescence, with some plants losing all the fronds. The species presents the capacity to compensate for the occasional loss of all the young fronds in a short period of time, keeping the number of fronds relatively stable at a given development stage. The data indicate ecological limits to the occurrence of this species in Southern Brazil. Spore production occurred only in a few plants, which were at least 2.5m tall. Spore formation is seasonal and maturation gradual to irregular even in a single frond. INTRODUCTION Information on reproductive aspects and growth patterns of ferns are relevant for a better understanding of their role in forest formations. Ferns may set patterns in early succession stages as well as influence the establishment of other species in regeneration processes. Interesting data discussed by Coomes ef al. (2005) suggest that the competition of the ground layer herbs may reduce regeneration opportunities for some seed plants. However, tree ferns and other caudex forming ferns may act as substrate for seed germination and initial establishment of certain woody angiosperms and gymnosperms, thus having an important role in their regeneration. Biological data are also important for the conservation effort for endangered species, such as the tree ferns which are subject to commercial (even if illegal) 264 FERN GAZ. 17(5): 263-270. 2006 exploitation. Some species are the source of fibrous material formed by the adventitious root mass covering the caudex which is used as substrate for growing orchids and aroids, while others have plants removed from the field and used in landscaping, decoration or as shelter building material. Even with the adherence to the CITES agreements restricting the trade, local use still represents a major pressure on some fern populations Peclbatedl data for Neotropical ferns are mostly found dispersed in the taxonomic and floristic literature and generally only of descriptive nature. There is a lack of studies as to their reproduction, growth and development. Among the groups that need urgent attention are the tree ferns (Dixit, 1986). Alsophila setosa Kaulf. (Cyatheaceae), one of the species currently extracted from primary and secondary semi-deciduous subtropical forests in the State of Rio Grande do Sul, Brazil (Windisch, 2002; Schmitt & Windisch, 2005), is the object of the present study. In southern Brazil Alsophila setosa is found in the mixed humid forests with Araucaria as well as in semi-deciduous forests. The removal of plants is not only a problem for the conservation of this species but also leads to the alteration of the forest formations. In contrast, Dickonia sellowiana Hook. is not under protection by any local or national legislation. Even if not used for tree-fern fibre extraction, its aesthetic properties make it an object of commercial exploitation for ornamentation, including church decoration for weddings in certain localities (Windisch, 2002). The caudexes support several epiphytic species, which themselves may also face conservation problems due to the disappearance of the sporophyte. In the present study the frond formation of A. setosa has been followed in situ in order to establish frond production, expansion and senescence rates, as well as the phenology of sporangial production and spore release. MATERIAL AND METHODS Two populations were studied, one in the municipality of Morro Reuter at 29° 32’ S and 51° 04’ W, 700m alt., another in Sapiranga at 29° 38° S and 51° 00’ W, 570m alt., from May 2000 to August 2001. These localities present secondary semi-deciduous seasonal forest formations. The regional climate is subtropical with the absence of drought. Average compensated monthly temperatures are normally under ist during four months of the year, this cold period being a determinant for physiological seasonality (Teixeira ef al., 1986). The closest Meteorological Station (Ivoti, 127m alt, data for the years 2000- 2001) indicated 13.3 C as the average of the coolest month (July) and 25. 1 C for the warmest (February), absolute minimum of —1.0 C (July) and maximum of ‘57 C (September and January). At the Morro Reuter, local residents registered —5°.0 C in the winter of 2000. During the period of the present study, precipitation data for Ivoti indicated 2138mm yearly rainfall, with less during May (average 59.4 mm), and maximum in September (232.2mm), October (303.9 mm) and January (308.6mm). Data for an extended period of time could not be obtained. Frost may occur several times during some winters. Monthly visits were made in order to follow the frond production phenology of 45 plants in Morro Reuter and 48 plants in Sapiranga. The plants were selected and marked by closely placed numbered stakes in the ground. During the 15-month period crozier formation, frond maturation and senescence were observed on every visit. At the beginning of the survey the youngest crozier was marked by a loose loop of plastic line SCHMITT & WINDISCH: FROND PRODUCTION IN ALSOPHILA SETOSA — 265 (not interfering with the crozier development), and the marked crozier serving as a reference in relation to the existing fronds and subsequently produced ones. In mature plants sporangia formation and spore release were recorded. The selection of two populations was decided upon in order to guarantee results in case of predatory extraction in one of the sites. The 15 months observation period ensured data for a complete seasonal growth-year. Soil samples were collected for NPK macronutrients analysis, with phosphorous and potassium measured by atomic emission spectrophotometry and nitrogen by standard nitrometric methods (Greenberh, 1992; Page ef a/., 1982). Statistical analyses were performed using methods described by Vieira (1980) and Watt (1998) and through the SPSS 9.0 software program (SPSS Inc., Chicago IL, USA) at the Universidade do Vale do Rio dos Sinos data processing facility. RESULTS AND DISCUSSION Alsophila setosa is an understory arborescent plant, forming an erect caudex up to 10m tall and up to ca. 10cm wide, covered by the spiny remnants of frond bases (and dry frond remnants at the distal part), with a crown of up to 3m long, and fronds which are tripinnate-pinnatissect at the base. The stipe is ascending, with blackish spines, presenting 2-4 pairs of aphlebia (with laminar tissue) at the basal portion. orro Reuter soil samples showed an average of 0.35 (+0.20)% nitrogen, 4.5 (£1.36) ppm phosphorous, and 175.16 (+51.09) ppm potassium. The average values for the Sapiranga samples were 0.36 (+0.15)% nitrogen, 3.8 (£0.69) ppm phosphorous and 229.33 (492.81) ppm potassium. Soil analyses demonstrated heterogeneity in macronutrients composition (NPK) among sites of each locality, but the application of t-test for independent homogenized samples indicated that contents of nitrogen (P = 0.981), phosphorous (P = 0.440) and potassium (P = 0.282) in the soil are statistically equal in both localities. Although the macro-nutrients (NPK) did not show significant differences, considering the probability of different histories of soil usage in the two sites, differences in micronutrients may be expected between the study sites. The average yearly frond production in the samples was 5.51 (+3.55) fronds/plant for Morro Reuter and 4.14 (41.93) fronds/plant for Sapiranga, the difference being statistically significant (P = 0.024). After the occurrence of frost between May and the beginning of July 2000, all young croziers at the apex of the caudex were damaged, with necrosis of the tissues, keeping the frond production at zero until August (Fig. 1) in both populations. With the arrival of the spring season, new croziers were formed and fronds started to expand, almost simultaneously, with 84.4% (Morro Reuter) and 66.7% (Sapiranga) of the specimens presenting one or more croziers in the initial phases of expansion. The largest frond production occurred during the period of higher rainfall (Fig. 1), with the peak in October (4.4+3.80 in Morro Reuter and 1.89+1.90 fronds/plant in Sapiranga). These observations agree with Luederwaldt (1923), who commented that in Southeastern Brazil new fronds of Alsophila setosa (cited as Hemitelia setosa) started to bud in the spring, approximately by the end of October. Shreve (1914) observed that Cyathea pubescens Mett. ex Kuhn, in humid mountain forests of Jamaica, formed new fronds during the winter and spring. Tanner (1983) working in Jamaica observed that the increase in the frond production of C. pubescens occurred in October, November and December after an increase in rainfall. Durand & Goldstein (2001) observed the native species of genus Ciborium Kaulf. in Hawaii produced most of their fronds 266 FERN GAZ. 17(5): 263-270. 2006 between February and April, characterizing a marked seasonality, which does not seem to be related with the rainy season. Seiler (1981) registered that the frond production in A. salvinii Hook. was not synchronous, but occurred at the end of the dry season and the beginning of the humid season. Data on the frond production of Dicksonia blumei Moore from Java was presented by Jaag (1942), but unfortunately not covering a complete growth year. The same author marked plants of A. glauca J. Sm, evaluating the production and life-span of fronds, with the complete renewal of the crown occurring between 182 and 254 days. The 55 croziers marked at Morro Reuter expanded on average 5.41 cm/day between October and November, 0.93 cm/day between November and December. From December to January 2001, 31 fronds continued to increase in length, but at a much slower rate (0.083 cm/day). In the fourth month only 11 fronds still presented some expansion (0.07 cm/day average). The maximum expansion rate was 7.48 cm/day between October and November 2000. The 31 croziers marked at Sapiranga expanded on average 5.38 cm/day between October and November, 1.24 cm/day between November to December. From December to January 22 fronds increased in length, with an average of 0.092cm/day. From January to February 2001 only eight fronds continued expansion with an average of 0.037cm/day. These data agree with the observations by Shreve (1914), who described different growth/expansion rates in different stages. The highest expansion rate presented in his study, 4.94 cm/day for Cyathea pubescens growing in the humid mountain forests of Jamaica, comes close to the values recorded in the present study. The average total number of fronds varied throughout the year. In Morro Reuter the smallest average was of 3.04 mature fronds/plant (October 2000), and the maximum 7.46 mature fronds/plant (May 2000). Comparing the average number of fronds in May | > 350 Frond production to Rainfall (mm 0 +o——o J J ——#— Morro Reuter —-o— Sapiranga ---a--- Rainfall | | 7 J ee | Figure 1. Average monthly frond production of A/sophila setosa populations in Morro Reuter and Sapiranga compared with precipitation (June 2000 to May 2001). SCHMITT & WINDISCH: FROND PRODUCTION IN ALSOPHILA SETOSA 267 2000 with that in May 2001 (5.77 mature fronds/plant) a statistically significant difference can be observed (P < 0.0001, n = 45). In Sapiranga, the smallest average was of 3.68 mature fronds/plant (October 2000), while the highest value was 5.95 mature fronds/plant (December 2000). Comparing the average number in June 2000 (5.62 mature fronds/plant) with that of June 2001 (5.33 mature fronds/plant), the values are statistically equivalent (P = 0.644, n = 48). The average yearly senescence rate of fronds in Morro Reuter was 6.97 fronds/plant, being superior to the new frond production rate. In Sapiranga, the plants presented an average yearly senescence rate of 4.33 fronds/plant, about the same as the new frond production rate (Fig. 2). Our field observations indicate that with the intense cold and frost occurrences, the fronds initially presented loss of photosynthetic surface, by partial damage to parts of the laminar tissue, followed by the total drying out of the fronds (Schmitt & Windisch, 2001). Reduction of the mature fronds was gradual, with eight individuals in the total sample (both localities) losing all of their fronds. n the Sapiranga population, the similar yearly average production and senescence of fronds indicates the maintenance of a constant number of fronds from one year to the next. The significant reduction of the total number of mature fronds in Morro Reuter, comparing data from May 2000 and May 2001, may be due to different local conditions including exposure to frost. The canopy of the forest in Morro Reuter is more open than that of Sapiranga, increasing the degree of exposure. These data may also indicate ecological limits for the occurrence of this species in Southern Brazil. A/sophila setosa presents the capacity to compensate for the occasional loss of all the young fronds ina short period of time, and so the number of fronds is kept relatively stable at a given development stage, although in a few cases no new fronds were produced. The production of new fronds and total number of fronds are similar to those found in the literature for other tree-ferns (Table 1) in other parts of the world. Our data for Number of fronds oo Morro Reuter Sapiranga ‘ON New y onde ‘BSe Senescent fronds ] eee pia J Figure 2. Average sti sotid perny ad senescence in the Morro Reuter and Sapiranga populations of Alsophila setosa (standard deviation indicated). 268 FERN GAZ. 17(5): 263-270. 2006 Alsophila setosa seem to be the first record of total loss of fronds. Low temperatures (including frost occurrence) and low rainfall during the cold season coincide with the loss of fronds. Seasonality in the loss of fronds was also observed in other tree ferns such as A. salvinii (Seiler, 1981), Cibotium glaucum (Sm.) Hook. (Walker & Aplet, 1994), while it was less pronounced in Cyathea pubescens (Tanner, 1983) and C. hornei (Baker) Copel. (Ash, 1987). Herbivory was observed in some fronds, but always partial and with preference to young fronds. No total loss of fronds due to herbivory was observed. It was not possible to identify the herbivore. Ants of the genus /ridomyrmex Mayr found on the plants probably only use cavities at the stipe bases for nesting. In Morro Reuter 8.88% of the plants produced fertile fronds, but only 2.08% in Sapiranga. All the fertile plants had caudexes at least 2.5m in length. Most of the fertile fronds presented developing sporangia between February and March, while in April the liberation of spores had already started. Spore maturation and liberation was gradual and irregular even with respect to the position on a single frond. Rosenstock (1907) observed that A/sophila setosa plants less than 8m tall are normally sterile. The low fertility values observed are related to the age structure of the populations, as there is a predominance of younger plants. Similar observation was made by Young & Leon (1989) in the Peruvian Amazon region, where only two specimens of Trichipteris nigra (Mart.) R. M. Tryon, from a total of 25 plants, were fertile and those were at least 5.5m tall. In Alsophila setosa vegetative reproduction by underground structures (Schmitt & Windisch, 2005) seems to be quite effective to allow for a low production of fertile Table 1. Production of new fronds and total number of fronds in tree-ferns. Locality Species Fronds/plant | Fronds/year Java (Tjibodas)* Alsophila glauca 6-12 6-13* El Salvador” Alsophila salvinii 6 pe Jamaica‘ Cyathea pubescens 7 8 Fiji® Cyathea hornei 3-11 3-9 Hawaii® Cibotium glaucum 5-16 3-5 Brazil (Morro Reuter, RS)‘ Alsophila setosa 0-21 0-14 Brazil (Sapiranga, RS)‘ Alsophila setosa 0-17 0-11 ; Jaag 1942, Seiler 1981, 1995, Tanner 1983, 4Ash 1987, Walker & Aplet 1994, present study, *data from 10 month observation period. SCHMITT & WINDISCH: FROND PRODUCTION IN ALSOPHILA SETOSA 269 fronds. Page (1979) pointed out that spore production might be low when a plant is in vegetative growth or high when the plant is under more severe ecological conditions, competing with other plants. Furthermore, Sato (1982) suggested that the cold climate has a restrictive effect on frond expansion as well as in the spore production period. The present data for A. setosa seem to agree with that suggestion. The gradual spore liberation over a period of time may be a positive aspect for species survival, allowing for higher chances of dispersal to new recently exposed microhabitats (Ranal, 1995). ACKNOWLEDGEMENTS This study was made possible by support from the Brazilian National Research Council - CNPq, State Foundation for Research of Rio Grande do Sul — FAPERGS, Universidade do Vale do Rio dos Sinos (Sdo0 Leopoldo), and Centro Universitario E (Novo Hamburgo). Cristina L. J. Schmitt and Lucas Schmitt provided welcome help in the field. Reviewers provided stimulating suggestions for the improvement of the manuscript. REFERENCES ASH, J. 1987. Demography of Cyathea hornei (Cyatheaceae), a tropical tree-fern in Fiji. Austral. J. Bot., 35 (3): 331-342. COOMES, D. A., ALLEN, R. B., BENTLEY, W. A., BURROWS, L. E., CANHAM, C. D., FAGAM, L., FORSYTH, D. M., GAXIOLA-ALCANTAR, A., PARFITT, R. L., RUSCOE, W. A., WARDLE, D. A., WILSON, D. J. & WRIGHT, E. F. 2005. The hare, the tortoise and the crocodile: the ecology of angiosperm dominance, conifer persistence and fern filtering. J. Ecology, 93: 918-935. DIXIT, R. D. 1986. Tree ferns: an urgent need of conservation. /ndian Fern J., 3: 42- 45. DURAND, L. Z. & GOLDSTEIN, G. 2001. Growth, leaf characteristics, and spore production in native and invasive tree ferns in Hawai. Amer. Fern J. 91 (1): 25-35. GREENBERT A. R. (ed.). 1992. Standard methods for examination of water and wastewater — 18*4 ed. American Public Health assoc., American Water Works assoc. & Water Environment Federation, Washington, D. C. JAAG, O. 1942. Untersuchungen iiber den rhythmus der lauberneuerung, die lebensdauer der blatter und den epiphytenbefall bei einigen farnen in den tropen. Mitteilungen der Naturforschenden Gesellschaft Schaffhausen, XVIII (6): 205-257. LUEDERWALDT, Von H. 1923. Die Cyathaceen aus der Umgebung der Stadt S. Paulo. Zeitschriff Deutscher Verein fiir Wissenschaft und Kunst, Sao Paulo. p. 83- 118. PAGE A. L., MILLER R. H., KEENEY D. R. (eds.). 1982. Methods of soil analysis, part.2, Chemical and microbiological properties, 2.ed. SSSA, Madison, 1982. American society of agronomy, Inc & Soil Science society of America, Inc, Madison, Wisconsin USA. PAGE, C. M. 1979. Experimental aspects of fern ecology. In: DYER, A. F. The experimental biology of ferns. Academic Press, London. p. 552-581. RANAL, M. A. 1995. Estabelecimento de pterid6fitas em mata mesofila semidecidua do Estado de Sao Paulo. 3. Fenologia e sobrevivéncia dos individuos. Rev. Brasil. Biol., 55 (4): 777-787. ROSENSTOCK, E. 1907. Beitrage zur Pteridophytenflora Siidbrasiliens. I]. Hedwigia, 46: 57-69. 270 FERN GAZ. 17(5): 263-270. 2006 SATO, T. 1982. Phenology and wintering capacity of sporophytes and gametophytes of ferns native to Northern Japan. Oecologia, 55: 53-61. SCHMITT, J. L. & WINDISCH, P. G 2001. Prejuizos causados pela geada no desenvolvimento de A/sophila setosa Kaulf. (Pteridophyta, Cyatheaceae). Revista de estudos / Centro Universitario Feevale, 24: 79-88. SCHMITT, J. L. & WINDISCH, P. G. 2005. Aspectos ecologicos de Alsophila setosa Kaulf. (Cyatheaceae, Pteridophyta) no sul do Brasil. Acta Bot. Bras., 19: 859-865. SEILER, R. L. 1981. Leaf turnover rates and natural history of the Central American tree fern Alsophila salvinii. Amer. Fern J., 71: 75-81. SEILER, R. L. 1995. Verification of estimated growth rates in the tree fern A/sophila salvinii. Amer. Fern J., 85: SHREVE, F. 1914. A montane vipeiuade a contribution to the physiological plant geography of Jamaica. Washington, D.C. Carnegie Institution of Washinhton. p. 51- 39 TANNER, E. V. J. 1983. Leaf demography and growth of tree-fern Cyathea pubescens Mett. Ex Kuhn in Jamaica. Bot. J. Linn. Soc., 87: 213-227. TEIXEIRA, M. B., COURA NETO A. B., PASTORE, U. & RANGEL FILHO, A. L. R. 1986. Vegetacao. In: Levantamento de recursos naturais. Rio de Janeiro: IBGE. v.33. p. 541-620. VIEIRA, S. 1980. Introdugdo a Bioestatistica. 3. ed. Rio de Janeiro, Campus. 196p. WALKER, L. R. & APLET, G. H. 1994. Growth and fertilization responses of Hawaiian tree ferns. Biotropica, 26 (4): 378-383. WATT, T. A. 1998. Introductory statistics for biology students. 2nd, Chapmann and Hall. WINDISCH, P. G. 2002. Fern conservation in Brazil. Fern Gaz., 16: 295-300. YOUNG, K. R. & LEON, B. 1989. Pteridophyte species diversity the Central Peruvian Amazon: importance of edaphic specialization. Brittonia, 41 (4): 388-395. FERN GAZ. 17(5): 271-278. 2006 271 CONSERVATION OF TWO ENDANGERED FERNS, ARCHANGIOPTERIS SOMAI AND A. ITOI (MARATTIACEAE: PTERIDOPHYTA), BY PROPAGATION FROM STIPULES W.L. CHIOU'”, Y.M. HUANG! & C.M. CHEN' ‘Division of Forest Biology, Taiwan Forestry Research Institute Author for correspondence Key words: Archangiopteris itoi, Archangiopteris somai, conservation, Marattiaceae, endangered plants, stipule propagation, Pteridophyte, reproduction. ABSTRACT Archangiopteris somai Hayata and A. itoi Shieh are ferns endemic to Taiwan and are categorized as endangered and critically endangered species respectively. Five fresh stipules were removed from each of 10 sporophytes of A. somai and A. itoi growing in Wu-lai, northern Taiwan. After rinsing in clean water and placing on medium (4:1, soil: peat moss) 50 stipules of each species were cultured at room temperature with 12 hr fluorescent light each day. After one year plantlets were produced by 40% of A. somai stipules and 90 % of A. itoi stipules. Within each species, the mean sprouting rate and sprouting time of stipules from stems of different sizes did not differ significantly. Sprouting and non-sprouting stipules were not significantly different in size. The relationship between average sprouting time and stipule size was very weak (4. somai) or non-existent (4. ito’). The growth of the mother plants from which stipules were stripped was not significantly different from their growth in the previous year, nor did it differ from the growth of control plants. This simple method of propagation from stipules provides an effective means of propagating these two species for horticulture, ex sifu conservation and in situ restoration. INTRODUCTION Archangiopteris Christ & Gies. is recognized as one of the ancient lineages of pteridophytes. Eleven species of Archangiopteris have been found in southeast China, northern Vietnam and Taiwan (Ching, 1958). Most taxa are endemic to these areas. The origin of Archangiopteris can be traced back to the Middle Jurassic period in the fossil record (Hill & Camus, 1986). This genus is phylogenetically closely related to Angiopteris Hoffm., another marattialean genus endemic to Southeast Asia, and Protomarattia Hayata, which is restricted to northern Vietnam (Hayata, 1919; Chang, 1975). Extant species of these genera represent relics of an ancient lineage that evolved through several glaciation and vicariance events. As a relic taxon, Archangiopteris provides information invaluable for tracing the evolutionary history of eusporangiate ferns, although until recently species of Archangiopteris have received little attention (Hsu et al., 2000; Chiang et al., 2002). Furthermore, many species of Archangiopteris are rare or endangered. Two species of Archangiopteris, A. somai Hayata and A. itoi Shieh, have been documented in Taiwan (DeVol & Shieh, 1994). Both species are endemic. Archangiopteris somai is an endangered species and A. ifoi is critically endangered (Kuo, 1997: Moore, 2001). Only two populations of each species have been reported, Z12 FERN GAZ. 17(5): 271-278. 2006 one of each in Wulai, in northern Taiwan, and one of each in Lienhwachi, in central Taiwan. However, the population of A. ifoi in Lienhwachi is known only from the original collection on which the species description is based, with no recent records from this site. Population sizes of both species are very limited, with c. 1000 individuals of A. somai and less than 100 of A. itoi, and in the field young sporophytes are very rare. Archangiopteris somai gametophyte growth is very slow and after 2.5 years in culture only about 1% of gametophytes produced sporophytes (Chiou and Huang, unpublished). For A. ifoi, spore germination and gametophyte culture have never been ocumented. In addition, there are no reports of tissue culture of the sporophytes of either species. Thus, conservation of these two endangered ferns is critical. Species within the Marattiaceae produce stipule buds and although expansion of these is rarely seen in the field (for example in Danaea wendlandii, Sharpe & Jernstedt 1991) some horticulturists have attempted to propagate marattialean ferns from stipules (Hoshizaki & Moran, 2001; Jones, 1987). To aid their conservation, an attempt was made to propagate the two species of Archangiopteris from stipules. Sprouting rate and sprout time were documented and the effect of stem and stipule size on sprouting analyzed. The morphology of young fronds was studied to assist further investigation in the field. Propagation from stipules proved to be an easy and effective method of propagating both species for conservation and horticulture. MATERIALS AND METHODS Because Archangiopteris somai and A. itoi are so rare, great care has to be exercised in using any of these plants or their parts for experimentation and propagation. First we tried to remove above ground stipules, but found they were firmly attached to the stems. To avoid damaging the “mother plants”, we removed underground stipules, which were not firmly attached to the stems, from sporophytes of the population in Wulai. For each species, five stipules were removed from each of 10 sporophytes. As soon as they were harvested, stipules were sealed in plastic bags to prevent dehydration. In the laboratory the stipules were rinsed for three minutes with clean water and then placed on, and half- covered by, medium (4:1, soil: peat moss) in plastic boxes. All cultures were maintained under white fluorescent illumination (24 umole m°s"', 12 h/d) at 20 to 28°C. The effects of mother plant size (stem diameter) and stipule size (width x length) on sprouting were analyzed (Table 1). Stipules were examined for sprouts every month for one year. In addition, the plants from which stipules had been removed were monitored throughout the experiment to determine whether stipule removal affected plant growth. RESULTS The stems and stipules of Archangiopteris itoi were significantly larger than those of A. somai. Plantlets began to sprout from 4. itoi stipules after three months in culture and sprouting peaked at 90 % after seven months. For A. somai, plantlets first sprouted after four months in culture. Cumulative sprouting increased slowly throughout the year to 40 % (Fig. 1). The stipules of A. itoi had a higher sprouting rate and shorter sprouting time than stipules of 4. somai (Table 1). Within each species, stem size did not affect sprouting rate significantly. However, the sprouting rate of stipules taken from the small stems of 4. somai was somewhat higher than the sprouting rate of stipules removed from large stems. Similarly, stem size did not affect sprouting time significantly, but there was a tendency for stipules from small stems of 4. somai to produce plantlets more quickly than stipules from large CHIOU et al.: CONSERVATION OF TWO ENDANGERED FERNS a Table 1. Archangiopteris somai and A. itoi stem diameter, stipule size, and sprouting rate and time. A. somai A. itoi t-test!’ Diameter of stem(cm) 4.1+0.9 Sotl9: = Stipule size Width (cm) 1.740.3 Pie sh dl Length (cm) 2204 34505 Width x Length 3.94+1.3 nao lal Sprouting rate (%) 40.24+21.1 90.0+4.1 ** Sprouting time (mo) 7.242.4 4.6+1.1 ** 'All comparisons between these two species were significantly different (P < 0.01). 50 - 100 - 90 | 80 = = 60 2 2 E © 50 & S ” = 40 g [e) _ a 30 a E 20 @ 10 0 Months in culture Figure 1. Sprouting rates (left axis, vertical bars) and cumulative sprouting rates (right axis, curves) of Archangiopteris somai (grey) and A. itoi (black) stipules cultured for one year. 274 FERN GAZ. 17(5): 271-278. 2006 Table 2. Mean sprouting rates and times for stipules taken from Archangiopteris somai and A. itoi stems of different sizes. Archangiopteris somai Diameter of stem (cm) 3 (n=3) 4(n=2) 5 (n=5) Sprouting rate (%)! 60 30 32 (40-80) (0-60) (0-60) Mean sprouting time (mo)' 6.1 7.0 8.4 (4-12)? (6-8) (6-11) Archangiopteris itoi Diameter of stem (cm) 6(n=1) 7(n=1) 8 (n=4) 9(n=2) 10(n=1) 13 (n=1) Sprouting rate (%)' 100 60 95 90 80 100 Mean sprouting time (mo)' 3.8 = 4.6 4.6 4.8 5.2 ‘None of the data in the same row were significantly different (t-test). “Numbers in parentheses are the range of values. stems (Table 2). For both species, stipules that gave rise to plantlets did not differ significantly in size from stipules that did not produce plantlets (Table 3). For A. somai there was a significant, weak, negative correlation between stipule size and sprouting time (Fig. 2). For A. itoi the correlation was not significant. The majority of first fronds emerging from 4. somai stipules were simple fronds (75%), but most first fronds emerging from stipules of A. itoi had one to three pairs of pinnae (93 %) (Table 4). None of the plants from which stipules were removed for the study exhibited obvious signs of injury or damage during the next year. For each plant, growth, the number of new fronds, and the timing of spore production were similar to the year before and the year after stipules were removed. In addition, the growth and other characteristics of plants from which stipules were removed did not differ significantly os the growth of control plants from which no stipules had been removed (data not shown). | DISCUSSION During the study we observed no expansion of stipule buds for either species, 4. somai or A. itoi, in the field. However, stipules have been used to propagate offspring of CHIOU et al.: CONSERVATION OF TWO ENDANGERED FERNS 275 Table 3. The relationship between sprouting status and stipule size in Archangiopteris somai and A. itoi. Species Archangiopteris itoi A. somai Sprouting status! a b t-test a b e t-test Stipule size (cm*) 9.3+3.0 7.6+1.7 ns 3.5+1.1 4441.4 3.7415 ns Stipule number = 45 5 20 24 6 'a = sprouting stipules; b = non-sprouting stipules that died in culture; c = non- sprouting stipules alive after one year in culture. several species in this family. One method for propagation of some marattialean species is to use the stipule buds (Jones, 1987). However, each sporophyte of 4. somai and A. itoi produces only 2 to 5 fronds per year so removing entire fronds could injure or kill these rare ferns. In this study, we removed underground stipules that did not subtend a living frond. Many of these stipules produced plantlets, and their removal did not damage the living fronds or affect the phenological characteristics of the mother plant in the following year. The sprouting rate and sprouting time for stipules removed from stems of different sizes were not significantly different within the range of stem sizes used in this study. However, given the small population size of both species and our concerns about their conservation, only 10 plants of each species were used in this study. A larger sample size may yield somewhat different results. This is most likely for A. somai, where the mean sprouting rate of stipules from 3 cm stems was about twice the rate for stems of other sizes, and sprouting time was about one month shorter. Thus, to propagate A. somai, we recommend using stipules from 3 cm stems, especially where parent plants are limited. In contrast, the sprouting rate of A. itoi reached 90%, sprouts formed in three to seven months, and the effect of stem size was negligible. Underground stipules from stems of any size appeared to be suitable for propagating this species. Stipule size had no significant effect on the sprouting rate or time of either species. Therefore, it appears not to be a factor in the vegetative propagation of these ferns. Table 4. Frequency of different types of first fronds emerging from stipules of Archangiopteris somai and A. itoi. A. somai A. itoi (%) (%) Simple frond 75 7 : Frond with one pair of pinnae 20 51 Frond with two pairs of pinnae 5 31 Frond with three pairs of pinnae 0 1] 276 FERN GAZ. 17(5): 271-278. 2006 i y = -0.5291x + 9.0484 12 + . r=0.23 p<0.01 ee ® — 10+ co) e @ — gt ee ® ~ ae & Ss 6+ ry ® ® © = ® Aa 4+ e ee e 2 t t t t me l 2 3 4 5 6 Stipule size (cm2 ) & pan y = 0.0784x + 3.8905 i + r=0.22 p-0.05 e 6 F @ @ ee e 5 Sprouting time (mo) ee) T e @ g $ e & stipule size (cm” ) Figure 2. Correlation between stipule size (width x length) and the time for plantlets to sprout from stipules of Archangiopteris somai (top) and A. itoi (bottom) CHIOU et al.: CONSERVATION OF TWO ENDANGERED FERNS zat Ferns can be propagated from spores under natural conditions and in the laboratory or greenhouse. When spores are limited, ferns can be propagated vegetatively, most commonly from frond buds, but also by tissue culture. Archangiopteris somai spores will produce gametophytes. However, the gametophytes grow very slowly and after 2.5 years only 1% had produced sporophytes in multi-spore cultures (unpublished). There are no published reports of spore germination for A. itoi and we failed to get the spores of this species to germinate. Propagation of the two species by tissue culture was also unsuccessful (Gen Chang, personal communication). Stipule culture appears to be a feasible and efficient method of propagating fern plantlets for conservation, both in situ and ex situ, e.g., in a botanical garden (Ranker, 1994). This method also could facilitate the propagation and conservation of other rare species of Marattiaceae. Another advantage of stipule culture over spore culture is that plants mature earlier. Usually, the first frond of sporophytes produced from gametophytes is simple. About 25% of A. somai plantlets produced from stipule cultures had one or two pairs of pinnae and 93 % of the A. itoi plantlets had one to three pairs of pinnae. Plantlets with more than one pinna grew faster and reached maturity earlier than plantlets with only a single frond, which is very important for horticulture, ex sifu conservation and in situ restoration. The primary disadvantage of vegetative propagation is that populations of ferns derived from these plants will have limited genetic diversity. However, based on a study of the afpB-rbcL intergenic spacer of chloroplast DNA, the genetic diversity of populations of 4. somai and A. itoi is surprisingly high (Chiang et al., 2002). Thus, propagating plantlets from the stipules of a number of different plants will help maintain a large proportion of the genetic diversity of each population. A further limitation of stipule culture is that the sporophytes of 4. somai and A. itoi produce only two to five new fronds each year (unpublished) and, consequently, only two to five new stipules each year. However, we do not know how long stipules survive. Clearly, the number of stipules is limited and care must be exercised when determining when and how many stipules should removed from a plant. REFERENCES CHANG, C. 1975. Morphology of Archangiopteris Christ & Giesenhagen. Acta Phytotax. Sinica 15: 235-247. CHIANG, T.Y., CHIANG Y.C., CHOU H., CHENG Y. P. & CHIOU W.L. 2002. Phylogeography and conservation of Archangiopteris somai Hayata and 4. ifoi Shieh based on nucleotide variation of the atpB-rbcL intergenic spacer of chloroplast DNA. Fern Gaz. 16(6): 335-340. CHING, R.C. 1958. A revision of the fern genus Archangiopteris Christ & Giesenhagen. Acta Phytotax. Sinica 7: 201-224. DEVOL, C.E. & SHIEH, W.C. 1994. Marattiaceae. In: Editorial Committee of the Flora of Taiwan, 24 edition (Eds.) Flora of Taiwan I, 24 edition, pp. 74-79. Editorial Committee of the Flora of Taiwan, 2"4 edition, Taipei. HAYATA, B. 1919. Protomarattia, a new genus of Marattiaceae, and Archangiopteris. Bot. Gaz. 67: 84-92. HILL, C.R. & CAMUS, J.M. 1986. Evolutionary cladistics of marattialean ferns. Bull. British Mus., Nat. Hist. (Bot.) 14: 219-300. HOSHIZAKI, B.J. & MORAN, R.C. 2001. Fern grower’s manual. Timber Press, Inc., Portland, Oregon. 604 pp. 278 FERN GAZ. 17(5): 271-278. 2006 HSU, T.W., MOORE, S.J. & CHIANG, T.Y. 2000. Low RAPD polymorphism in Archangiopteris itoi, a rare endemic fern in Taiwan. Bot. Bull. Acad. Sinica 41: 15- 18. JONES, D.L. 1987. Encyclopaedia of ferns. Timber Press, Inc., Portland, Oregon. 433 pp. KUO, C.M. 1997. Archangiopteris somai Hayata. In: LU, S.Y., CHIOU, W.L., CHENG, Y.P. & CHEN, C.W. (Eds.) Rare and endangered plants in Taiwan. II, pp, 7-8. Council of Agriculture, Taipei. MOORE, S.J. 2001. Archangiopteris itoi Shieh. In: LU, S.Y., & CHIOU, W.L. (Eds.) Rare and endangered plants in Taiwan. VI, pp, 9-10. Council of Agriculture, Taipei. RANKER, T.A. 1994. Evolution of high genetic variability in the rare Hawaiian fern Adenophorus periens and implications for conservation management. Biol. Cons. 70: 19-24, SHARPE, J.M. & JERNSTEDT, J.A. 1991. Stipular bud development in Danaea wendlandii (Marattiaceae). Amer. Fern J. 81: 119-127 FERN GAZ. 17(5): 279-286. 2006 279 FILICALEAN FERNS FROM THE TERTIARY OF WESTERN NORTH AMERICA: OSMUNDA L. (OQSMUNDACEAE : PTERIDOPHYTA), WOODWARDIA SM. (BLECHNACEAE : PTERIDOPHYTA) AND ONOCLEOID FORMS (FILICALES : PTERIDOPHYTA) K.B. PIGG', M.L. DEVORE” & W.C. WEHR™ 'School of Life Sciences, PO Box 874501, Arizona State University, Tempe, AZ 85287-4501, USA; "Department of Biological & Environmental Sciences, 135 Herty Hall, Georgia College & State University, Milledgeville, GA 31061, USA; *Burke Museum of Natural History and Culture, University of Washington, PO Box 353010, Seattle, WA 98195-3010, USA "deceased, 12 April, 2004 Key words: Blechnaceae, fossil fern, Osmunda, Osmundaceae, Tertiary, Wessiea, Woodwardia ABSTRACT Recently discovered frond remains assignable to Osmunda wehrii Miller (Osmundaceae), as well as several new records of Woodwardia (Blechnaceae), and a new onocleoid fern are reported from the Tertiary of western North America. Pinnule morphology of QO. wehrii supports the inclusion of this species in Osmunda subgenus Osmunda, as originally proposed by Miller and suggests a close affinity to O. regalis L. and O. japonica Thunb. New occurrences of the Woodwardia aerolata clade are noted for the Late Paleocene of western North Dakota and of a highly reticulate-veined form from the Miocene of western Washington. Re-evaluation of specimens of W. deflexipinna H. Smith (Succor Creek, Miocene) confirms its close affinity to W. virginica J. Smith. A fern with onocleoid anatomy is recognized from the middle Eocene Clarno Nut Beds of Oregon. Together, these examples demonstrate that the presence of critical taxonomic features, even in fragmentary remains, can increase our knowledge of filicalean fern evolution, biogeography and ecology in the Tertiary. INTRODUCTION Despite a widely held belief that the fossil record provides little information about them, filicalean ferns, including derived forms, are a relatively common component of many Tertiary floras. However, the impact of the fossil record in deciphering Tertiary fern evolution has been limited for several reasons. The focus for many Tertiary researchers has been on the collection and comparison of angiosperm remains, although ferns are typically included and occasionally highlighted in published treatments of floras (e.g., Pabst, 1968). Often paleobotanists focus on investigating major groups of plants during what they perceive as key moments of evolutionary times. In the case of ferns, which have been thought by many to undergo their major radiations primarily during the Paleozoic and Mesozoic, the pulses and subsequent evolution of many filicalean ferns in the Tertiary previously have been ignored. Fragmentary remains, particularly of nonfertile fronds, have not been studied in detail, because of the perceived difficulty of relating fossil material to modern taxa. Occasionally, florules 280 FERN GAZ. 17(5): 279-286. 2006 have been recovered that contain fern remains in essentially monotypic stands (e.g., Smith, 1938; Crabtree, 1988); however, these localities have been rarely studied. In recent years, several whole plant reconstructions of fossil ferns have been produced (e.g., Rothwell and Stockey, 1991; Pigg and Rothwell, 2001) and new emphasis has been focused on fern biogeography and ecology (Collinson, 2001, 2002; Page, 2002). With this new attention accorded to the fossil ferns of the Tertiary, the potential now exists for a better understanding of their evolutionary, biogeographic, and ecological significance. Our recent work in the middle Miocene, permineralized Yakima Canyon flora of central Washington, USA has provided information on three fern genera that occur there: Osmunda wehrii Miller (Osmundaceae); Woodwardia virginica J. Smith (Blechnaceae), and a small fern with onocleoid vegetative anatomy, Wessiea yakimaensis Pigg & Rothwell (Miller, 1982; Pigg and Rothwell, 2001). In this contribution we report new occurrences of related taxa in several Tertiary localities of western North America. Together, these examples show that even fragmentary fern remains, when demonstrating critical taxonomic features, can provide considerable new information about filicalean fern evolution in the Tertiary. RESULTS AND COMMENTS Osmunda wehrii (Osmundaceae) Osmunda wehrii was described from permineralized rhizomes and leaf bases by Miller (1982) from Yakima Canyon, central Washington, USA, where it occurs intermingled with taxodiaceous foliage, stems and cones, the abundant foliage and smaller rhizomes of Woodwardia virginica and rhizomes and stipes of Wessiea yakimaensis (Pigg and Rothwell, 2001). Several pinnules of Osmunda wehrii have recently been recognized in attachment or closely associated to O. wehrii stipes (Figs. 1, 2). Fronds are bipinnate and the most extensive segments are up to around 2.8 cm long and represent portions of pinnae and individual pinnules. The pinnules are up to 8 mm long and 3.5 mm wide and each has an oblique base (Fig. 1, 2). Venation is dichotomous with the initial forking occurring close to the point of divergence of lateral veins from the pinnule midvein with a second and third order of dichotomies occurring in the basal part of pinnules (Fig. 2). Based on these morphological features, Osmunda wehrii pinnules are assignable to the subgenus Osmunda, which includes the extant royal fern, O. regalis, and the two Asian species O. japonica and O. lancea. Pinnules are most similar to those of O. regalis (Fig. 3) and O. japonica both of which share with O. wehrii the features of an oblique pinnule base and multiple vein dichotomies, features that O. Jancea lacks (Hewitson, 1962). Anatomical details still under investigation will provide additional information about this middle Miocene member of Osmunda subgenus Osmunda. Woodwardia aerolata (Blechnaceae) A fern with possible affinities to W. aerolata is recognized from Late Paleocene Beicegel Creek flora of western North Dakota, USA (Figs. 4, 7). This flora is very similar to the Almont flora of central North Dakota (Crane, et al., 1990), both in composition and preservational type. However, in comparison to the Almont flora, material from the Beicegel Creek flora tends to have a greater percentage of permineralized specimens and provides an unparalleled opportunity to critically examine both the anatomy and morphology of Late Paleocene taxa (DeVore et al., 2003). Previous to our studies no ferns have been described from Almont or Beicegel Creek, however several fragmentary fern pinnules that represent at least two different PIGG et al.: FILICALEAN FERNS OF WESTERN NORTH AMERICA 281 Figures 1,2. Osmunda wehrii pinnules from Yakima Canyon, Middle Miocene, Washington, USA, Figure 1. Overview of pinnae attached to rachis, x 4. Figure 2. Detail of pinnule to show dichotomising venation, x 4.3. Figure 3. Osmunda regalis pinnule for comparison, x 5. Figures 4,7. Woodwardia pinnule from Beicegel Creek, Late Paleocene, North Dakota, USA. Note elongate, pinnatifid pinna and diamond- shaped aeroles characteristic of extant W. aerolata. Figure 4. Detail of pinnule, x 3.8. Figures 5,6,8. (left), 9-11. Woodwardia deflexipinna vegetative and fertile pinnae, Middle Miocene Succor Creek, Oregon, USA. Figures 5,6. Overviews of disarticulated pinnae, x 1.5, and x 2.7, respectively. Figure 7. Overview of Beicegel Creek specimen, x 2.5. Figure 8. Sori with flap-like indusia, at left: fossil W. deflexipinna, x 15; at right: SEM of extant W. virginica, x 16. Figure 9. Rhizome with crozier (at arrow) previously illustrated by Smith (1938) and Graham (1965), x 1. Figures 10,11. Fertile pinnae of W. deflexipinna showing indusiate sori (arrows), X 3, and x 4.3, respectively. 282 FERN GAZ. 17(5): 279-286. 2006 types of ferns recently have been recovered. The Woodwardia specimen is 3.5 cm long x 2.7 cm wide, and is comprised of a pinnatifid, possibly apical, portion of the frond showing several pinnule lobes (Fig. 4, 7). The venation includes relatively coarse elongate, diamond-shaped aeroles 4-6 mm long x 6-8 mm wide and is marginally freely dichotomising, confirming its identification as Woodwardia rather than Onoclea (Collinson, 2001). A second taxonomically important character, the presence or absence of marginal teeth could not be determined from available material. A second group of Woodwardia taxa with an excellent fossil record are those of the W. virginica clade (=Anchistea sensu Cranfill and Kato, 2001). Woodwardia deflexipinna was originally described from the Miocene Succor Creek of Oregon based on material collected from a small florule where these ferns occurred in a nearly monotypic stand. Fronds, rhizomes, crosiers and fertile remains were described and briefly illustrated by H. Smith (1938) and later by Alan Graham (1965). Reinvestigation of this material confirms that the vegetative and fertile remains are remarkably like those of both fossil and extant W. virginica (Figs. 5, 6, 8-11: Pigg and Rothwell, 2001). Frond remains are pinnatifid pinnae preserved for up to 6.5 cm long wide with up to 10 pinnule segments (Figs. 5, 6). Pinnules are 11mm long x 4 mm wide. Like W. virginica, this species has a single series of areoles that parallel the midvein, and relatively few anastomoses in the lateral part of the pinnule. Meshes are 3-4 mm long x 0.5-0.7 mm across (Fig. 6). Fertile frond fragments consist of elongate sori around 2 mm long x 0.5 a RAT CMMs cM ol OE 14) oat ol am ‘ t. 7 Sf na, es, ‘ EES, ; { ’ yt _ Figure 12. Woodwardia Miocene, Vasa Park, Washington, USA. Vegetative frond segment showing highly reticulate pattern of venation, x 1.3. Figures 13-15. Onocleoid fern, Middle Eocene Clarno Nut Beds, Oregon, USA. Figure 13. Transverse section of rhizome showing helically arranged, attached stipe bases, x 1.5. Figure 14. External view of rhizome showing numerous attached stipes and adventitous roots, x 0.9. Figure 15. Detail of stipe trace (arrows), x 5. PIGG et al.: FILICALEAN FERNS OF WESTERN NORTH AMERICA 283 mm wide, each partly covered by an elongate, flap-like indusium that hinges along its outer side like those of W. virginica (Figs. 8, 10, 11). A crozier-bearing rhizome, previously illustrated by Smith (1938) and Graham (1965) is around 9 cm long and 1.2 cm thick (Fig. 9). Spores have not been recovered. A third new occurrence of fossil Woodwardia is from the Miocene Vasa Park locality near Seattle. Only a few fragments have been discovered to date. The largest is up to 3.8 cm long x 3.4 cm wide and includes several pinnules. Pinnules are 1.8 cm long x 0.9 cm across, and veins are highly reticulate, resulting in short, polygonal meshes 2-3 mm long x 0.8-1.0 mm (Fig. 12). In comparison with extant species, these remains mostly closely resemble those species with numerous, short, polygonal reticulations in their leaves, such as W. fimbriata, a species from western North America today. OID fern A permineralized fern with onocleoid anatomy has been discovered in the middle Eocene Clarno Nut Beds of central Oregon. This locality is well known for its beautifully preserved fruits and seeds (Manchester, 1994). Rhizomes are preserved for up to 7.2 cm in length and 1.5 - 2 cm in diameter and produce numerous stipes (Fig. 14). The stipes extend up to 1.9 cm in length from their attachment to the rhizome and are around 0.3 cm across. Anatomical structure is similar to that of Wessiea yakimaensis Pigg & Rothwell, from Yakima Canyon and Makotopteris from the middle Eocene Princeton chert of southern British Columbia (Stockey, et al. 1999; Pigg and Rothwell, 2001). In this pattern, paired hippocampiform frond traces are helically arranged and are produced without leaf gaps (Fig. 13, 15). In contrast, adventitious root traces are produced from the stipe bases, and leave behind a root gap. This pattern is exhibited by onocleoid ferns and is common to a large number of higher filicalean genera (Stockey et al., 1999; Pigg and Rothwell, 2001). As additional vegetative and hopefully fertile remains become available for these ferns, their affinities will be more closely resolved. DISCUSSION & CONCLUSIONS Although such factors as angiosperm bias and fragmentary preservation have often limited a detailed description of Tertiary filicalean ferns, the careful study of their remains may help flesh out the record of their evolution and diversification into habitats they occupy today. Whole plant reconstructions (¢.g., Onoclea sensibilis, Rothwell and Stockey, 1991; Woodwardia virginica, Pigg and Rothwell, 2001) have previously demonstrated that essentially modern filicalean fern species were established in present day habitats and persist in similar sites today. The examples provided in this study are mostly fragmentary in nature but still possess critical features that can be used for identification and provide additional new data to the record. The Osmundaceae have the most extensive fossil record of the filicalean ferns, extending back to the Permian based on characteristic vegetative anatomy of rhizomes and leaf bases (Serbet and Rothwell 1999; Collinson, 2001). Of the three genera in the family, Osmunda is the best represented and most diverse. The genus is traditionally divided into three subgenera, Osmunda, Osmundastrum and Plenasium. Molecular studies generally support this subgeneric classification, however the genus Osmunda appears to be paraphyletic to Zodea and Leptopteris (Yatabe et al., 1999). The subgenus Osmunda is well represented in Cretaceous and Tertiary localities by both anatomically preserved rhizomes and compressed foliage. Permineralized stems include O. pluma (North Dakota, Paleocene), O. oregonensis (Oregon, Eocene), O. iliaensis Romania (Miocene-Pliocene), and QO. nathorstii (Spitzbergen, Eocene; Miller, 1967, 1982). Most 284 FERN GAZ. 17(5): 279-286. 2006 recently, O. shimokawaensis, a fern very similar to O. wehrii has been described from the Miocene of Japan (Matsumoto & Nishida, 2003). Foliage of subgenus Osmunda is known from several localities in the Late Cretaceous and Tertiary (Miller 1967; Collinson, 2001). The discovery of pinnules of Osmunda wehrii supports Miller’s earlier (1982) placement of the species into subgenus Osmunda and the fossil material is similar to O. regalis and O. japonica (Hewitson, 1962). The Osmunda subgenus is well established by the Miocene, so its presence at Yakima Canyon is no surprise. Evolution and paleoecology of the Osmundaceae are, however, interesting in light of the recent resurgence of interest in the relationship of angiosperms to the Tertiary radiation of the higher filicalean ferns (Lovis, 1977; Rothwell, 1987; Schneider, et al. 2004). Unlike the higher filicalean ferns, the Osmundaceae has been around since the Permian, and thus predate angiosperm influences by over 140 million years. Osmundaceous ferns are often associated in the fossil record in conifer-dominant environments. Osmunda regalis often occurs as the dominant fern in Zaxodium swamps of southeastern North America today, as did O. wehrii in the middle Miocene Yakima Canyon flora of ashington state. Presumably, osmundaceous ferns evolved in relation not to angiosperms, but to conifers. In contrast, the first evidence of Woodwardia is in the Late Cretaceous (Collinson, 2001). This genus was an important component of the circumboreal flora of the Northern Hemisphere Paleocene (Kvaéek, 1994; Collinson, 2001). Although the fossil record of Woodwardia is in need of revision, it is now evident that the occurrence of several clades sensu Cranfill and Kato (2001) among woodwardioid ferns can be recognized in the Tertiary. The basalmost clade, represented by W. aerolata (=Lorinseria aerolata sensu Cranfill and Kato, 2001) appears in the Late Cretaceous and Paleocene. The next clade is represented by W. virginica (=Anchistea virginica sensu Cranfill and Kato, 2001) and related taxa. It is interesting to note that although this clade is estimated to be ancient, it does not appear in the fossil record until the Miocene. The Miocene, however, is the heyday for the W. virginica clade with not only permineralized W. virginica remains in Washington at Yakima Canyon (Pigg and Rothwell, 2001), but the very similar W. deflexipinna at contemporaneous localities in the Succor Creek flora of Oregon (Smith, 1938; Graham,1965). Asian Miocene occurrences of the Woodwardia virginica clade are also known, as are those of central and western Europe (some fossils of W. munsteriana, W. maxonii; Hurnick, 1976; Pigg and Rothwell, 2001). Of the other, apparently more derived clades that may also have a fossil record, the Vasa Park site may document an early occurrence for one of the more highly reticulate-veined forms such as W. fimbriata, a species native to western North America today. Lastly, three onocleoid ferns with characteristic stipe and root trace anatomy are now known: the Princeton chert Makotopteris and the Clarno Nut Beds fern, both from the middle Eocene, and Wessiea from the middle Miocene. These ferns cannot be placed into modern genera. Makotopteris cannot be named to a genus because although fertile remains are known, some critical taxonomic characters could not be gleaned from the fossil and comparative anatomical information of remains known in the fossil is not available for all pertinent living relatives. Wessiea and the Clarno fern are presently known only from their rhizome and stipe anatomy. Nevertheless they demonstrate that middle Miocene and Eocene floras, respectively, hosted a variety of highly derived filicalean ferns. PIGG et al.: FILICALEAN FERNS OF WESTERN NORTH AMERICA 285 ACKNOWLEDGEMENTS We thank Jon Hager, Seattle for donating collections from Vasa Park to the Burke Museum and thereby making them available for study; Robyn J. Burnham, University of Michigan Museum of Paleontology, Ann Arbor and Patrick Fields, Michigan State University, East Lansing, for allowing us to examine material of Woodwardia deflexipinna; Steven R. Manchester, Florida Museum of Natural History, Gainesville, for allowing us to study the Clarno Nut Beds fern, and Ray Cranfill, Jepson Herbarium, Berkeley for information on extant Osmunda and Woodwardia. We acknowledge funding from National Science Foundation EAR 9980388 and EAR 0345838 and a Travel Grant, School of Life Sciences, Arizona State University to KBP; National Science Foundation EAR 0345569 and a Faculty Research & Development Grant, Georgia College & State University to MLD; and the Wesley C. Wehr Paleobotanical Endowment, University of Washington, to WCW. REFERENCES COLLINSON, M. E. 2001. Cainozoic ferns and their distribution. Brittonia 53 (2): 173-235. COLLINSON, M. E. 2002. The ecology of Cainozoic ferns. Review of Palaeobotany & Palynology 119 (1-2): 51-68. RABTREE, D. R. 1988. Mid-Cretaceous ferns in situ from the Albino Member of the Mowry Shale, southwestern Montana. Palaeontographica 209B: 1-27. CRANE, P. R., MANCHESTER, S. R., & DILCHER, D. L. 1990. A preliminary survey of fossil leaves and well-preserved plant reproductive structures from the Sentinel Butte Formation (Paleocene) near Almont, North Dakota [USA]. Fieldiana Geol 20: 1-64 CRANFILL. R. & KATO, M. 2001. Phylogenetics, biogeography, and classification of the woodwardioid ferns (Blechnaceae). In: Chandra, S. & Srivastava (Eds.), Pteridology in The New Millennium, pp. 25-48, Kluwer Academic Publishers, The Netherlands. DEVORE, M.L., PIGG, K. B., & MANCHESTER, S. R. 2003. Late Paleocene Plants from North Dakota: New localities and insights from the Almont flora. Paper 240-5, Geological Society of America, Abstracts. GRAHAM. A. 1965. The Sucker Creek and Trout C reek Miocene floras of southeastern Oregon. Kent State University Bulletin, Research Series IX, Kent State University Press 53: 1-147. HEWITSON, W. 1962. Comparative morphology of the Osmundaceae. Annals of the Missouri Botanical Garden 49 (1-2): 57-93. HURNIK, S. 1976. Die fossilen Arten der Gattung Woodwardia Smith, 1793 und ihre Vertretung um nordbohmischen Tertiar. Sbornki Narodniho Muzea V. Praze. 32B(1): 15-44 KVACEK, Z. 1994. Connecting links between the Arctic Palaeocene and European Tertiary floras. In: BOULTER, M. C. & FISHER. H. C. (Eds.) Cenozoic plants and climates of the Arctic, pp.251-266 NATO ASI Series, C. 127. Springer-Verlag, Berlin. LOVIS. J. D. 1977. Evolutionary patterns and processes in ferns. In R. D. Preston and H. W. Woolhouse (Eds.). Advances in Botanical Research 4: 229-440. Academic Press, London. MANCHESTER. S. R. 1994. Fruits and seeds of the Middle Eocene Nut Beds flora, 286 FERN GAZ. 17(5): 279-286. 2006 Clarno Formation, North Central Oregon. Palaeontographica Americana 58: 1-205. MATSUMOTO, M. & NISHIDA, H. 2003. Osmunda shimokawensis sp. nov. and Osmunda cinnamomea L. based on permineralized rhizomes from the Middle Miocene of Shimokawa, Hokkaido, Japan. Paleontological Research 7 (2): 153-165. MILLER, C. N., Jr. 1967. Evolution of the fern genus Osmunda. Contributions to the Museum of Paleontology, University of Michigan 21 (8): 139-203. MILLER, C.N., Jr. 1982. Osmunda wehrii, a new species based on petrified rhizomes from the Miocene of Washington. American Journal of Botany 69 (1): 116-121. PABST, M. B. 1968. The flora of the Chuckanut Formation of northwestern Washington. The Equisetales, Filicales, and Coniferales. University of California Publications in Geological Sciences 76, University of California Press, Berkeley & Los Angeles. PAGE, C. N. 2002. Ecological strategies in fern evolution: a neopteridological overview. Review of Palaeobotany & Palynology 119(1-2): 1-33. PIGG, K. B. & ROTHWELL, G. W. 2001. Anatomically preserved Woodwardia virginica (Blechnaceae) and a new filicalean fern from the middle Miocene Yakima Canyon flora of central Washington, USA. American Journal of Botany 88(5):777-787. ROTHWELL, G. W. 1987. Complex Paleozoic Filicales in the evolutionary radiation of ferns. American Journal of Botany 74(3): 458-461. ROTHWELL, G W. & STOCKEY, R. A. 1991. Onoclea sensibilis in the Paleocene of North America, a dramatic example of structural and ecological stasis. Review of Palaeobotany & Palynology 70(1-2): 113-124. SCHNEIDER, H., SCHUETTPELZ, E., PRYER, K. M., CRANFILL, R., MAGALLON, S. & LUPIA, R. 2004. Ferns divested in the shadow of angiosperms. Nature 428 (6982): 553-557. SERBET, R. & ROTHWELL, G. W. 1999. Osmunda cinnamomea (Osmundaceae) in the Upper Cretaceous of western North America: additional evidence for exceptional species longevity among filicalean ferns. International Journal of Plant Sciences 160 (2): 425-433. SMITH, H. V. 1938. Some new and interesting Late Tertiary plants from Sucker Creek, Idaho-Oregon boundary. Bulletin of the Torrey Botanical Club 65 (8): 557-564. STOCKEY, R. A., NISHIDA, H., & ROTHWELL, G. W. 1999. Permineralized ferns from the middle Eocene Princeton chert. I. Makotopteris princetonensis gen. et sp. nov. (Athyriaceae). International Journal of Plant Sciences 160 (5): 1047-1055. YATABE, Y., NISHIDA, H., & MURAKAMI, N. 1999. Phylogeny of Osmundaceae inferred from rbcL nucleotide sequences and comparison to the fossil evidences. Journal of Plant Research 111 (1108); 397-404. FERN GAZ. 17(5): 287-291. 2006 287 GROWTH IMPAIRMENT OF HUMAN CELLS BY FERN SPORE EXTRACTS S.E. SIMAN & E. SHEFFIELD® Faculty of Life Sciences, G30B Stopford Building, University of Manchester, Oxford : Road, M13 9PL.UK author for correspondence (Email: L.Sheffield@manchester.ac.uk) INTRODUCTION Our review of the literature (Siman er a/.1999) posed a question — are there human health risks from fern spores?. Our conclusion was that there may be. Carcinogenesis in humans caused by tissues of ferns is well established (e.g. Alonso-Amelot & Avendaiio, 2002) and spores of some ferns cause allergic reactions and contact dermatitis in some people. We first obtained evidence for DNA-damage and carcinogenicity from experiments with spores of a single species (Preridium aquilinum - bracken) fed to mice. Similar experiments have since established that spores of five fern taxa (including northern and southern hemisphere bracken) induce DNA adducts in upper gastrointestinal tissue of mice (Siman et a/. 2000a). Experimental administration of spores to whole humans is not practical, but experiments with human cells strengthen the conclusion that fern spores can cause DNA damage. Administration of extracts of spores of Dicksonia antarctica, Pteris vittata, Sadleria pallida, Anemia phyllitidis and Pteridium aquilinum to human premyeloid leukaemia cells induces breaks in their genomic DNA (Siman er al. 2000b). DNA damage is strongly correlated with carcinogenic events (e.g. Fairbairn ef al. 1995) but even if such damage was caused in whole humans after inhalation or ingestion of spores, it is of course possible that repair mechanisms would mend the breaks before tumourigenesis was triggered. The aim of the research reported here was to establish whether fern spore extracts prevent the growth and proliferation of human cells. MATERIALS AND METHODS Fifty mg of spores of six ferns: Anemia phyllitidis (L.) Swartz, Cyathea arborea Sm.. Dicksonia antarctica Labill., Drvopteris filix-mas (L.) Schott, Preridium aquilinum (L.) Kuhn and Pteris vittata (L.) were ultrasonicated and extracted with 0.5 ml sterilised distilled water overnight. P. aquilinum and D. filix-mas were examined because of the established toxicity of their vegetative tissues, the remaining species were included to provide breadth of taxonomic differences and/or horticultural significance. The HaCaT keratinocyte human cell line used adopts a fibroblast morphology in culture. Cells were seeded at 2-4 x 104 cells mI”! in 1.5 cm diameter wells in 24-well dishes and fed 1 ml complete Dubecco’s modified Eagle medium (DMEM) containing 5% fetal calf serum. After 24 h (start of “day 0”) the medium was replaced by complete DMEM containing only 0.5% serum. (This reduction in serum sensitises the cells.) At the same time the cells were fed 50 ml spore extract. “Vehicle controls” were fed 50 ml sterilised, distilled water, “controls” were given no treatment. Cell number in a fixed amount of liquid extracted was counted in a haemocytometer 288 FERN GAZ. 17(5): 287-291. 2006 every 24h for 5d. The average of 3 counts per well for 3 wells per day formed the basis for a calculation of the mean of means. The mean final number of cells mI”' on day 5 of each treatment, and the growth rate between day 3 and 5, were compared with the controls using ¢ tests for unmatched samples. RESULTS Over the five days of the experiment cell number increased in control treatments at a rate typical of this cell line, by a factor of x10-15. Control cell counts increased by means of x12.4 and 12.6, respectively. Cell number after treatment with extracts of A. phyllitidis, C. arborea, P aquilinum and P. vittata increased less than those of controls (see Figure 1). Cell cultures exposed to P. aquilinum extracts showed an increase in cell number of x2.0 - highly significantly less than that of both the controls and vehicle controls (tf = 5.29 for control, 6.28 for vehicle control, df=4, ps0.01). The final concentration of cells in wells treated with P. aquilinum extracts was about six times lower than those of the controls. Extracts of 3 other species: P. vittata, C. arborea and A. phyllitidis, impaired the growth of the cultured cells - increase factors of x6.4, 7.4 and 10, respectively. (Not significantly different from either of the controls at the 99% confidence level, but C. arborea treatment was significantly impaired at 95% confidence level.) D. filix-mas extract had no detectable effect on the growth of the cultured cells (increase factor x12.4). Cells treated with extract of the tree fern D. antarctica increased more than any others (x15) over the five days of the experiment, but this was not significantly different from the controls. Between days 2 & 3 cultures treated with P. vittata extracts experienced a significant inhibition of cell proliferation (9.6, df=4, ps0.01,see Figure 2). From day 3 these cultures started to recover. After day 3 all growth rates stabilised. Between days 3 & 5 5 ek oe [S: o r 7 & eC fins day O sday 5 =r lene oes es = S Es 500 & a 400 Oo Oo tony 300 o = S 200 E s S 100 Figure 1. HaCaT cells treated with fern spore extracts then cultured for 5 days. Column marked (*) differs from controls at ps0.01. SIMAN et al.: GROWTH IMPAIRMENT OF HUMAN CELLS 289 the vehicle control cells and the cells treated with D. antarctica and D. filix-mas extracts had a higher proliferation rate than cells of any of the other treatments. Growth rate was highest in D. antarctica-treated cells (=2.3, df=4, ps0.1). The growth rate of P. aquilinum-treated cells was lower than that of the vehicle controls (#=2.3, df=4, ps0.1). DISCUSSION Cytotoxicity. Spore extracts of 4 of the 6 ferns tested impaired the growth of the human cells used in the conditions of this study. Cytotoxicity was strongest in P. aquilinum treatments (the results indicate that the cells in these cultures would eventually have died). P. vittata is a common glasshouse weed, a popular garden ornamental, and a potential phytoremediation treatment (Ma er a/. 2001). P. vittata extract treatment generated the second lowest cell number and can be said to have had a cytostatic effect. C. arborea-treament generated the third lowest final cell density. This growth impairment was not mirrored by the other tree fern species used, however, (the horticultural favourite, D. antarctica) which, if anything, appeared to promote cell growth. Vegetative tissues of the garden ornamental D. filix-mas have been used to kill tapeworms and insects. Perhaps spores are not as toxic as vegetative tissues, or non-mammalian cells respond differently to the toxins, since cells treated with spore extracts of this species were the ones least different from the controls. Explanation for cell growth impairment. Extracts of P. aquilinum, D. antarctica, P. vittata, and A. phyllitidis induce DNA strand-breaks in human cells (Siman ef al. 2000b). In the present study, only P. aquilinum and P. vittata extracts showed a pronounced adverse effect on the cells — those of D. antarctica did not. Growth impairment may therefore be due to something other than damage resulting from DNA strand-breaks in some or all ferns, or experimental differences in the study reported here perhaps reduced the strand-breaks 600 “& control Dicksonia antarctica ® vehicle control Ee > Anemia phyilitidis ahdele Carkiol ee Bes Jp coe sonia ic i = OD be Gabacerceel Dryopteris filix-mas ato ~~ Pteridium aquilinum = a Pteris vittata Anemia phyllitidis x 8 300 Cyathea arborea ° g Pteris vittata E 200 3 = S o E 100 Pteridium aquilinum Day 0 Day 1 Day 2 Day 3 Day 4 Day 5 Figure 2. HaCaT cell number after treatment with fern spore extracts. 290 FERN GAZ. 17(5): 287-291. 2006 caused by some species and not others. One difference between the experiments is that spore extracts used in the study of DNA strand-breaks were made with the solvent dimethyl sulphoxide (DMSO), rather than water. Damaging compounds differ in solubility between media, and as ferns differ markedly in biochemical constituents, this could explain the difference in effects. Of course neither DMSO or water constitute realistic solvents for the human cell surfaces that encounter spores. It could be argued, however, that water constitutes the milder solvent, so deleterious effects that are seen with both water and DMSO provide strong evidence for toxicity of the spores in question. Another explanation for growth impairment is the formation of adducts (which form when compounds bind to DNA). Adducts may be repaired by the cell, or can initiate events leading to the creation of a cancer cell. Povey ef al. (1996) established that spores of Pteridium can induce DNA adducts in tissues of mice, and we have shown that spores of P vittata and A. phyllitidis share this property (Siman et al., 2000a).The main Preridium carcinogen, ptaquiloside, can cause adduct formation (Smith & Seawright 1995), is water soluble (Ojika er al. 1987), and is found in bracken tissues including spores (Schmidt et a/. 2005; Rasmussen, Schmidt & Sheffield unpubl.) so it is possible that this compound is responsible for the cytotoxicity of P. aquilinum spore extracts. Human health implications? This in vitro study strengthens the extensive in vivo evidence of the toxicity of Pteridium. Health warnings about walking through bracken stands during the sporing season may therefore be reasonable, although far from all Preridium stands are fertile (Wynn er a/. 2000), and there is no way to relate likely numbers of spores inhaled to potential risk. Bracken remains the most damaging fern examined to date, in all the studies reviewed, but the evidence presented herein suggests we should not be complacent about other ferns. Toxins may of course differ between species, and there is no defensible way to extrapolate from effects on starved cells in vitro to risk for entire healthy humans (either in terms of effects seen, or numbers of spores used). However, humans are not good at dealing with health risks. For example, there is probably no safe dose of ionising radiation, yet people choose to live in areas dominated by granite, and therefore experience high levels of natural background radiation. Smokers do not start “complaining” about their health until long after taking the first puff (cf. Hainsworth 2000). If health risk may be incurred from inhaling spores of ferns other than bracken, there is no reason to “never touch a fern again” (cf. Hainsworth 2000). Anyone regularly exposed to large numbers of spores need spend only a few pence on a mask (available from most DIY stores) which excludes particles. Our advice to those involved with caring for sporing ferns, or harvesting spores is simple — better safe than sorry! ACKNOWLEDGEMENTS We are very grateful to Drs A. Povey, D. Lovejoy, G. Brunner and A. Ludlow who helped in various ways during the course of this study. REFERENCES ALONSO-AMELOT, M E & AVENDANO, M. 2002. Human carcinogenesis and bracken fern: a review of the evidence. Current Medicinal Chemistry 9:675-86. SIMAN et al.: GROWTH IMPAIRMENT OF HUMAN CELLS 291 FAIRBAIRN, D.W., OLIVE, P.L. & O’NEILL, K.L. 1995 The comet assay: a comprehensive review. Mutation Research 339: 37-59. HAINSWORTH. P.H. 2000 Human health risks from fern spores? — a review. Pteridologist 3:111. , L.Q., KOMAR, K.M., TU, C., ZHANG W., CAI, Y & KENNELLY, E.D. 2001. A fern that hyperaccumulates arsenic. Nature 6820: 579. OJIKA, M., WAKAMATSU, K., NIWA, H & YAMADA, K. 1987. Ptaquiloside, a potent carcinogen isolated from bracken fern Pteridium aquilinum var. latiusculum: structure elucidation based on chemical and spectral evidence and reactions with amino acids, nucleosides and nucleotides. Tetrahedron 43: 5261-5274. POVEY, A. C., EVANS, I.A., TAYLOR, J.A. & O’CONNOR, PJ. 1996. es post-labelling analysis of DNA adducts formed in the upper gastrointestinal tissue of mice fed bracken extract or bracken spores. 1996. British Journal of Cancer 74: 1342-1348. SCHMIDT, B., RASMUSSEN, L.H., SVENDSEN, G.W., INGERSLEV, F. & HANSEN, H.C.B. 2005 Genotoxic activity and inhibition of soil respiration by ptaquiloside, a bracken fern carcinogen. Environmental Toxicology and Chemistry 24: 2751-2756. SIMAN, S.E., POVEY, A.C. & SHEFFIELD, E. 1999. Human health risks from fern spores? - a review. Fern Gazette 15:275-287. SIMAN, S.E., POVEY, A.C., WARD, T.H., MARGISON, GP. & SHEFFIELD, E. 2000a. The genotoxicity of fern spores. In Bracken fern: biology, toxicity & control. eds. Smith R. & Taylor J.A. IBG Special Publication. 99-105. SIMAN, S.E., POVEY, A.C., WARD, T.H., MARGISON, GP. & SHEFFIELD, E. 2000b. Fern spore extracts can damage DNA. British Journal of Cancer. 83:69-73. SMITH, B.L. & SEAWRIGHT, A.A. 1995. Bracken fern (Pteridium spp) carcinogenicity and human health — a brief review. Natural Toxins 3: 1-5. WYNN, J.M., SMALL, J.L., PAKEMAN, R.J. & SHEFFIELD, E. 2000. An assessment of genetic and environmental effects on sporangial development in bracken (Preridium aquilinum (L.) Kuhn) using a novel quantitative method. Annals of Botany 85:113-115. Vian Fs FERN GAZ. 17(5). 2006 FERN GAZ. 17(5): 293-302. 2006 293 RESPONSES OF PTERIDOPHYTE SPORES TO ULTRAFREEZING TEMPERATURES FOR LONG-TERM CONSERVATION IN GERMPLASM BANKS D. BALLESTEROS, E. ESTRELLES & A.M. IBARS Jardi Botanic-ICBiBE, Universitat de Valencia. Quart, 80. 46008 Valencia. Spain. (Email: ana.ibars@uv.es) Key Words: fern spores, long term conservation, ultra freezing, viability. ABSTRACT There are many unresolved questions around the loss of viability of pteridophyte spores and the most suitable conditions for long term conservation. The effects of humid and dry conditions, different temperatures, and the short exposure of spores to liquid nitrogen have been occasionally studied by various authors. The work presented here is the first result of a project focussed on long-term conservation of spores of pteridophytes. Using species from different ecological habitats, we show the effects of ultra-freezing, at -80°C and -196°C (LN) for six months of storage, on the germination process as well as on the development of the gametophyte until it reaches sexual maturity. We analyze and comment on the results obtained for the final germination percentage and the germination rate, the final percentage of gametophytes that reach the laminate developmental phase, and of gametophytes that attain the sexual phase under the two conditions. All these data are referenced to the initial viability of the samples used as well as to a control of spores stored at room temperature (approx. 25°C). INTRODUCTION Conserving biodiversity involves maintaining the genetic variability of the different groups of living species in all aspects. The pteridophytes are the oldest group of vascular plants on Earth. As far as shapes, sizes and species are concerned, it was a very varied and diverse group in days gone by. Nowadays, even though they are just a shadow of that former abundance, they greatly contribute to the richness of plant biodiversity, not only because of their floral significance but also for their phylogenetic value A large percentage of pteridophytes tends to be associated with ecosystems that are particularly sensitive to degradation (mature forest formations, humid areas and riparian habitats, etc.), some of which are considered by present-day legislation as natural habitats of community importance. Some of these taxa are included in the lists of species of community interest which require strict protection. Other than contributing to the conservation of species of a greater scientific interest and also to the maintenance of biological biodiversity, in all the aspects this entails, their protection and conservation also contribute to ensuring that these ecosystems are indeed maintained. In view of the clear need of conserving and regenerating ecosystems under threat, and given that pteridophytes are very sensitive to environmental changes, it is essential that their spores should be included in a germplasm bank in which material is to be conserved for the biological-experimental study of the species as well as for the conservation of biodiversity on a long-term basis. 294 FERN GAZ. 17(5): 293-302. 2006 Germplasm banks play an important role in the long-term ex situ conservation of plant species. Seed conservation in germplasm banks is a reality supported by numerous studies (Ellis et al., 1985; Dickie et al., 1990; Gomez-Campo, 2001). Nonetheless, no conclusive studies exist as far as the conservation of pteridophytes is concerned to guarantee a long-term conservation methodology. Various studies verify the relatively rapid loss of spore viability under environmental conditions. This loss occurs in a few days (Equisetum sp.), a few months (Osmunda regalis) or even a year or so (Onoclea sp. and Matteuccia sp.) in the case of green spores; in a few months (Gleicheniaceae and Thyrsopteris elegans), between | year and a decade (the majority of species), and in exceptional cases, several decades (Pellaea sp., Asplenium serra, Marsilea sp.) where non-green spores are concerned (Lloyd & Klekowsky, 1970; Dyer, 1979; Page, 1979; Windham et al., 1986; Lindsay et al. 1992) The different studies that have been conducted on this theme indicate causes of this rapid viability loss: biochemical and metabolic factors, such as the deficiency of respiratory substrates, the lack of membrane integrity, the inactivity of growth enzymes and substances in non-green spores (Beri & Bir, 1993), or genetic factors such as chromosomal mutations (Page ef al., 1992). It has been indicated that the loss of viability in green spores may be due to their high respiratory rate (Lloyd & Klekowsky, 1970), or to the loss of the capacity to recover photosynthetic activity after drying (Lebkuecher, 1997). However, impacts of these causes are not completely clear because experimentation has been conducted on this theme. the other hand, most of the few studies that actually deal with the conservation of pteridophytes have focussed on analysing which conditions are optimum to maintain spore viability during storage. Different temperatures in germplasm banks have been tested (4°C, 5°C, -12°C, -20°C, -70°C), with both humid and dry conservation methods. Even cryoconservation techniques using liquid nitrogen have been tested. Some studies indicate that the autecology of the species may be a significant factor when establishing conservation protocols. Other factors, such as the ploidy level, have been analysed in different studies (Windham et al., 1986: Lindsay et al. 1992; Agrawal et al., 1993; Simabukuro e¢ a/., 1998: Pence, 2000; Constantino et al., 2000: Quintanilla et a/., 2002; Aragon & Pangua, 2004). Extending spore viability up to 2 or 3 years in green spores has been achieved in the Equisetum genus (Jones & Hook, 1970). Furthermore, spores from species such as Woodwardia radicans, Culcita| macrocarpa, Athyrium filix-femina, Phyllitis scolopendrium and many others, remained in storage and suffered no viability loss over more than 12 or 24 months (Lindsay er al. 1992; Quintanilla et a/., 2002; Aragon & Pangua, 2004); also spores immersed in liquid nitrogen are still viable for at least 75 months (Pence, 2000). Generally speaking however, few quantitative and long-term data exist that determine the best and most lasting way to conserve spores. The study by Page er al. (1992) indicated that when spores are being stored, a need exists to maintain not only their viability, but also both their growth capacity and genetic integrity. In this respect, and as a novelty with regard to the rest of existing works undertaken in this particular field, the effects of tested conditions regarding germination and early development, the effects of late gametophyte development, along with their capacity of completing the life cycle, will be analysed. This is the main objective since what is being dealt with here is the discovery of which conditions are optimum for the long-term storage of fern spores in germplasm banks. For this reason BALLESTEROS et al.: RESPONSES OF PTERIDOPHYTE SPORES TO ULTRAFREEZING — 295 the spores need to be viable and capable of producing sporophytes, which are to be subsequently used in both habitat restoration and for research purposes. MATERIAL AND METHODS This study has been carried out with 10 pteridophyte species present in the Mediterranean area. The species and populations collected are shown in Table 1. In order to obtain a significant sample of the variability in the natural population, work was conducted with spores of at least 20 individuals per species. For spore obtaining, collection sheets are prepared in glossy paper for the fronds from the individuals collected in the field. These sheets are placed in a dry, aired place under light pressure, and are left at room temperature for a week. After this time, the spores which had fallen on the paper were collected and stored in glass vials once they had been sifted with a 0.074 mm sieve in order to eliminate any remains of sporangia and paleae. The spores were placed into Eppendorf tubes, approximately | mg/spores per tube. The tubes were stored in the dark at 25°C (laboratory temperature), -80°C and —196°C (liquid nitrogen) for 6 months. After a fast defrost at 40°C for 5 minutes, 1 mg spores was suspended in | ml of liquid Dyer culture medium and dispensed 5 drops with a micropipette among seven 5.5 mm petri dishes on culture medium with 1% agar (Dyer, 1979). The dishes were sealed thoroughly with Parafilm and incubated at 12 h light photoperiod (daylight fluorescent tubes, photon irradiance 25-50 mol m-2 s-1 in the 400-700 nm regions) at 20°C. In order to observe the start of germination and how it developed, the percentage of germinated spores was noted daily after 10" day, afterwards every three days. The final spore viability was checked by analysing the germination percentage after 30 days, by randomly counting 100 spores per dish and by considering those spores with either a primary rhizoid or a first chlorophyllous cell as having germinated. The percentage of gametophytes that reached the laminar developmental phase was also analysed after 30 days. The laminar phase was taken as the initiation of the 2D growth (the transition stage), that is, the first division in a perpendicular plane of the prothallium cells. This date was chosen since it was the time when all the gametophytes in the controls taken at zero time reached this phase. Subsequently gametophytes were transferred to soil in order to check the correct appearance of the sexual phase in each replica of the three treatments after 120 days. An arcsine transformation was applied to the percentage data, and a one-way ANOVA or a t-test was used to analyse them. The Tukey test was used (p>0.05) on the means to identify homogeneous groups. All statistical tests were carried out with the SPSS program, version 11.0. RESULTS The obtained results (Table 2) show a 24-72 hour delay in the start of the germination and in reaching T5g among the spores conserved at room temperature (25°C) with respect to those that were ultrafrozen. This delay does not occur in the species whose germination began after 6 or more days, as is the case of Polystichum lonchitis, Notholaena marantae and Ceterach officinarum, species from high mountainous terrain and xerophilous environments, respectively. If we look at Figure 1 we see that a significant viability loss has been revealed in all the species used in the study (ANOVA, p-value <0.05) when they were kept at room FERN GAZ. 17(5): 293-302. 2006 296 ‘pasn saisads ay} Jo sjieiap UONDAT[09 "| ATAVL "BIOUDTRA ‘Ioqoseusg wnsDUulgyjo yovsiajay £00C/0 1/70 WOOL 69fXSO¢ OO] op OouRLeEY ‘eployy “iodsq £007/60/S | W OCHC [E35 ONRES SOS WY .p PUIOD sutyoUuo] WNY ISAO "BIDUQTRA “SOARTRSSEIA £002/80/€1 woe ECILALOE plOA NEY STR stasnyod staajdapay J ‘OTJOAISED “INsoKY] [Op e]JoqQeisiA, £00C/L0/81 WOOT ICAALOE OPEL [op 104 WNJDIJNID UNYINSA]Og “O][AISED “WINSOKP] [OP PI[IqQuISTA €£00C/L0/L1 WOOF 9TMALOE Jeuesuryy [op ooureg punuaf-xipif WnldAy] ‘OT[AISBD ‘TENsOKPY [OP B][OqRISTA £007/LO/LI W O8El OTMALOE OpinZ Jap josey spu-xipif St4ajdoaaq ‘eIoUg TBA “BSUOTETIIA £00C/L0/£0 WwW Q97 OefASO’ JOJeS ek] Op OouRUeg DIDIJIA SMA ‘OT[AISBD ‘JesoKP\ [OP P]]IQeISTA £007/90/TI W OSE! ITAALOE OpinZ [ap lose syisv.4f s14ajdojsay ‘Q][AISBD “O9A Op BIPNITV €£00C/S0/£T we os9 TCMASOE snpurliusg op souRsIO ADJUDADUL DUADJOYJON "BUOIID “ID IP BADIA eT €007Z/70/07 W009 PIDGLIE [Blajseg Jop osjequiq siiaidouo wniuajdsp eq apnyoly Wn AQBIO'T saiodg BALLESTEROS et al.: RESPONSES OF PTERIDOPHYTE SPORES TO ULTRAFREEZING — 297 Storage temperature Wi control 25 wi -80°C Wi -196°C 100 — 7 — ee 90 — 80 — 70 60 — Germination percentage o Oo | % E 2 & 2 Ss : = es ¢ 3 38 S _ — ~ s &§ § 8 &§ € $8 $$ gs 3B Q = £ & & c 3 3 £ g 2 x So ov = o ra = » Q S = = = 0 S s g - 2 pr S g = w 3 3 a 3 E E Re Q L& c =| — 3 oA 3 3 S 2 Q ® = s x S 5 = ¢ D 9 r 3 9 3 = ® > o > 7 r= Q $ = 16) $ rs) o Q o > S) ~ < i 0 = & i Species Figure 1. Germination percentage of the different species after 6-month of storage at 25°C, -80°C and —196°C in comparison with the initial value. FERN GAZ. 17(5): 293-302. 2006 LIL e 70 F £66 9 b 096 I- soupy LiL e b'0 + £66 g's v DoI8- stuaidocugy Lit t £0 + £66 L ¢ ez LiL e b'0 F 0°66 CL ¢ 096 I- s1p1Spaf LIL 2 = SOF £'86 oa ¢ Do08- stuaidopsy LiL eB b'0 F £86 8 9 Suse LIS v Col Fiz $6 L 096 I> wnapurryffo LIE e 61 FETS 56 L a8 yoosala) LIS e b'8 F698 01 L Do$Z LiL e Sl F0'S6 S 7 096 1- aaa ee puluaf-x1]t UL RB 9°0 + 9°S6 ¢ v Jo08° eu : ih B 81 F916 8 9 DoSTZ LiL e 70 + 9°66 9 b 096 I> Sie st4ajdouo LIL e L'0 + 1°66 9 b Jo08° wnquajdsp LIL BLOF F'86 69 ¢ DoSZ . sayAydojoues (9 eee le aa , seed a 05] Avp suruursoq ae ag uOXxE |, YIM 8a}eId Hur saykydoyauey illite ieiies eae 298 BALLESTEROS et al.: RESPONSES OF PTERIDOPHYTE SPORES TO ULTRAFREEZING 299 ‘ounqesaduiay pur soisads yora 10} UMOYs die ‘saXydojaued yeNxas YIM saysip JO Joquinu oy} sv [Jom se ‘sep QQ Jaye saiAydojowesd [PUOISUdUIP-OM) sWOdEq PRY IY) Sosods payeurwUas Jo adeyuaoIad au “(¢~'Q=%” AdYON} 189) sdnosB snosuadowioy ‘OS | ‘Aep Suruulsaq s,uoneUIUay *7 ATAVL 07 + £06 LLPors vitoiG 6074 7'£6 9'0 + 9°86 £0 + £66 b'0 + 186 60+ 1'S6 RIFL 6 9'0 + £86 9'0 + 186 £0 + 1°86 90 + 9°96 80+ 1756 90 + 0'L6 va) \o i oO stasnjod sia dayay | DIDJIA SUA styouo] wnyIUSA]Od wnjpajnov uNYIUNSAJOg apjuDADUL puavjoyujoN 300 FERN GAZ. 17(5): 293-302. 2006 temperature (25°C), except for Ceterach officinarum, which maintains the same germination percentage after a 6-month storage as those spores either stored at —80°C or at —196°C in liquid nitrogen do. These germination percentages of those spores stored in an ultra frozen state were no different to the controls taken after collecting plant material, so therefore no viability loss was observed. Furthermore, those spores that do not germinate were also seen to not swell in the same way as those that do germinate, that is, there is no proper imbibition. As for late gametophyte development, no negative effects to it were observed at any of the temperatures tested in the study, as Table 2 shows, since more than 90% of germinated spores in all the species (except Ceterach officinarum) and treatments reach the two-dimensional gametophyte stage after a 30-day culture. Furthermore, sexual structures develop in the gametophytes in virtually all the dishes after 120 days of culturing, grow normally after transference to soil and completed their development. This point is of great relevance when the aim is to obtain mature plant for populational reinforcements. DISCUSSION AND CONCLUSIONS Agrawall et al. (1993) and Pence (2000) were pioneers in observing the effects that both cryoconservation and long-term storage at a temperature of —196°C (reached with liquid nitrogen) have on spores. Their results are positive since they not only established that cryofreezing does not kill spores, but also that spores remain viable after being stored at this temperature for up to 75 months. However, the results from these studies are qualitative and/or only focus on germination. Therefore, the real effect that cryoconservation has on spore viability and on the subsequent gametophyte development remained unknown. Whittier (1996) also studied the effects that ultrafreezing at —70°C had on Equisetum hyemale green spores by observing that the viability of these spores could be prolonged. After 16-month storage, more than a quarter of the spores stored remained viable, when these do not survive more than 3 or 4 weeks at room temperature. After a six-month work period, our results provide the first quantitative data as well as data regarding gametophyte development after ultrafreezing (-80°C and —196°C) in non-green spores. We notice that not only does ultrafreezing not affect spore viability after 6 months, but also that the gametophyte develops in perfect conditions, with no sign of it being negatively affected. We could conclude a priori from this that ultrafreezing is seen to be a good conservation method for both spore viability and genetic integrity of the conserved material since no effects are provisionally observed on its development, that is, on its phenotype. This is coherent with what is expressed in the bibliography on the conservation of biological material at cryogenic temperatures (cryopreservation), a process able to suppress biological reactions completely, therefore permitting this material to be stored indefinitely. Thanks to cryobiology, cryopreservation processes may be controlled in such a way that the damage cells suffer is the minimum, therefore meaning that recovery is the maximum, providing the process is conducted properly. By contrast, a considerable loss of viability exists in those spores stored at room temperature (25°C) in all the species tested, except for Ceterach officinarum. Furthermore, a delay was seen in the start of those species whose germination initiation is under 6 days. We have also seen that the non-germinated spores do not swell in the same way as those that do germinate, that is, there is no proper imbibition, just as Beri BALLESTEROS et al.: RESPONSES OF PTERIDOPHYTE SPORES TO ULTRAFREEZING — 301 & Bir (1993) demonstrated with Preris vittata. The viability percentage of some species has dropped by more than 30%, as is the case with Asplenium onopteris, Dryopteris filix-mas, Cystopteris fragilis, Polystichum lonchitis, Athyrium filix-femina and Notholaena marantae. There are also some species whose germination percentage has dropped less than 15%, as is the case of Polystichum aculeatum, Pteris vittata and Thelypteris palustris. As far as these differences are concerned, no relation of any kind has been observed as to them being able to influence the autecology of the species, the lesure-type spores, the ploidy level, nor the taxonomic group to which the different species belong. In all the species where a viability loss occur after storage at 25°C however, the late gametophyte development of germinated spores occurs correctly without any differences with regard to those conserved at ultra low temperatures. This provisionally indicates that the viability loss after 6 months storage is produced by the alteration to germination. However, there appears to be no damage in the genetic integrity of the material. It will be necessary to await further data with longer storage times in order to confirm this, and to also check the anomalous gametophyte proportion (digitate or callus-like forms) that is produced in the two-dimensional phase, just as Smith & Robinson (1975) point out. We have selected these parameters as a development measure because those that measure early gametophyte development only indicate whether a delay in gametophyte development exists (see for example Beri & Bir (1993) and Camloh (1999)). However, we needed to record whether gametophyte development is completed correctly because it is the data relating to germplasm bank spore conservation that interest us. The fact that no delay to the germination initiation is noted in spores that germinate from the sixth day, even though they lose viability at 25°C, could be related to the ecology of the species, since these species correspond to environments with water deficiencies in the three cases observed: high mountainous terrain and xerophilous environments. Without going into physiological questions, even though it would be an interesting field to analyse, this might occur because the spore displays some germination-delaying mechanism, even when imbibed in such a way that it increases the chance of water availability during the subsequent gametophyte development. Therefore, the delay may assume the slower imbibition which was pointed out by Beri & Bir (1993). This behaviour could be a dormancy status as a survival strategy in extreme xeric environments (Kornas, 1985). ACKNOWLEDGEMENTS The authors thank two anonymous reviewers for very helpful comments. The work was supported by the Ministry of Science and Technology, project REN 2002-03697. REFERENCES AGRAWAL. D. C., PAWAR, S. S. & MASCARENHAS, A. F. 1993. Cryopreservation of spores of Cyathea spinulosa Wall. ex Hook. f. An endangered tree fern. J. Plant Physiol. 142: 124-126. ARAGON, C. F. & PANGUA, E. 2004. Spore viability under different storage conditions in four rupicolous Asplenium L. taxa. Amer. Fern J. 94(1): BERL A. & BIR, S. S. 1993. Germination of stored spores of Pteris vittata L. Amer. Fern J. 83(3): 73-78. CAMLOH, M. 1999. Spore age and sterilization affects germination and early 302 FERN GAZ. 17(5): 293-302. 2006 gametophyte development of Platycerium bifurcatum. Amer. Fern J. 89: 124-132. CONSTANTINO, S., SANTAMARIA, L. M. & HODSON, E. 2000. Storage and in vitro germination of tree fern spores. Botanic Gardens Micropropagation News 2(4): 58-60. DICKIE J. B. , R. H. ELLIS, H. L. KRAAK, K. RYDER AND P. B. TOMPSETT. 1990. Temperature and seed storage longevity. Annals of Botany 65: 197-204. DYER, A. F. 1979. The culture of fern gametophytes for experimental investigation. In: DYER A. F., ed. The experimental biology of ferns, pp. 253-305. Academic Press, London. ELLIS, R.; HONG, T.D. & ROBERTS, E.H. 1985. Handbook of seed technology for genebanks. Vol. I. Principles and methodology. International Board for Plant Genetic Resources. Roma GOMEZ-CAMPO, C. 2001. La practica de la conservacion de semillas a largo plazo. In: Conservacion de especies vegetales amenazadas en la regidn mediterranea occidental, pp. 255-266. Ed. Centro de Estudios Ramon Areces. Madrid KORNAS, J. 1985. Adaptative strategies of African pteridophytes to extreme environments. Proceedings of the Royal Society of Edinburgh 86B: 391-396. LEBKUECHER, J. G 1997. Desiccation-time limits of photosynthetic recovery in Equisetum hyemale (Equisetaceae) spores. Amer. J. Bot. 84(6): 792-797. LINDSAY, S., WILLIAMS, N., DYER, A. F. 1992. Wet storage of fern spores: unconventional but far more effective!. In: IDE, J.M.; JERMY, A.C. & PAUL, A.M. (Eds) Fern Horiculture: past, present and future perspectives, pp. 285-294. Intercept, Andover. LLOYD, R. M. & KLEKOWSKY, E. J. 1970. Spore germination and viability in Pteridophyta: evolutionary significance of chlorophyllous spores. Biotropica 2(2): 129-137. PAGE, C. N. 1979. Experimental aspects of fern ecology. In DYER A.F. (Ed). The experimental biology of ferns, pp. 551-589. Academic Press, London. PAGE C. N., DYER A. F., LINDSAY, S. & MANN, D. G. 1992. Conservation of pteridophytes: the ex situ appreoach. In: IDE, J. M., JERMY, A. C. & PAUL, AM EDS. Fern horticulture: past, present and future perspectives, pp.269-278. Intercept, Andover. PENCE, V. C. 2000. Survival of chlorophyllous and nonchlorophyllous fern spores through exposure to liquid nitrogen. Amer. Fern J. 90(4): 119-126. QUINTANILLA, L. G; AMIGO J.; PANGUA E. & PAJARON S. 2002. Effect of storage method on spore viability in five globally threatened fern species. Annals of Botany 90(4): 461-467. SIMABUKURO, E. A., DYER, A. F. & FELIPPE, G M. 1998. The effect of sterilization and storage conditions on the viability of the spores of Cyathea delgadii. Amer. Fern J. 88(2):; 72-80. SMITH, D. L. & ROBINSON, P. M. 1975. The effects of storage age on germination and gametophyte development in Polypodium vulgare L. New Phytol. 74: 101-108. WHITTIER, D. P. 1996. Extending the viability of Equisetum hyemale spores. Amet. Fern J. 86(4): 114-118. WINDHAM, M. D.; WOLF, P.G. & RANKER, T. A. 1986. Factors affecting prolonged stored spore viability in herbarium collections of three species of Pellaea. Amet. Fern J. 76(3): 141-148. FERN GAZ. 17(5): 303-309. 2006 303 HERBIVORY ON EPIPHYTIC FERNS OF A MEXICAN CLOUD FOREST K. MEHLTRETER', K. HULBER? & P. HIETZ’ ‘Departamento Ecologia Funcional, Instituto de Ecologia, A.C., km 2.5 antigua carretera a Coatepec No. 351, Congregacion El Haya, Xalapa 91070, Veracruz, México Institut ftir Okologie und Naturschutz, Universitat Wien, Althanstr. 14, 1091 Wien, Austria ’Botanisches Institut, Universitat fiir Bodenkultur, Gregor-Mendel-Str. 33, 1180 Wien, Austria Key words: cloud forest, epiphytes, leaf age, fertility, herbivory, Mexico ABSTRACT The often-stated hypothesis that ferns are attacked less by herbivores than are angiosperms has not been confirmed for terrestrial ferns. Several authors reported for terrestrial ferns and angiosperms the same number of insect pest species, and similar leaf damage of 5-38 percent, depending on species, leaf age, and type of vegetation. We studied five epiphytic species: Pleopeltis crassinervata (Fée) T. Moore, Polypodium furfuraceum Schltdl. & Cham., P. plebeium Schltdl. & Cham., P. polypodioides (L.) Watt, and P. rhodopleuron Kunze , ina Mexican cloud forest to test the hypothesis that epiphytic ferns have less leaf damage than terrestrial ferns. For each species we tagged 14-30 sections of tree branches and marked each fern leaf individually. For each leaf, herbivory was estimated as leaf area loss for each pinna, using a scale of seven damage classes (0%, less than or equal to 10%, < 25%, < 50%, $ 75%, < 100%, 100%), in February 2003 and February 2004. In 2004, we counted the number of marked and unmarked new leaves to calculate leaf life-span. Leaf damage depended strongly on species and leaf life-span, but generally did not differ from the values reported for three terrestrial fern species in the same forest site (5.8- 11.1%). P. furfuraceum and P. rhodopleuron were the least damaged species in both years with 8.4 - 10.7 % mean leaf area loss, while P. plebeium had the highest leaf area losses of 21.2 - 22.0 %. The highest leaf damage in P. plebeium might be a consequence of its longer leaf life-spans of 29.5 + 3.4 months, while P. rhodopleuron, the least damaged species, had the shortest leaf life-span of less than 12 months. INTRODUCTION Although the hypothesis that ferns are attacked less and damaged less by insects than are angiosperms is long established (Schneider 1892), it did not receive much attention until the publication of the classical article of Ehrlich and Raven ( 1964), who stated that “In fact, very few insects feed on ferns at all, a most surprising and as yet unexplained fact with no evident chemical or mechanical basis”. As possible reasons, the lack of food sources such as flowers and fruits and the consequently lower co-evolution between insects and ferns, and the probable better biochemical defences were discussed. Biochemical analyses revealed tannins, cyanogenic glycosides, phytoecdysones and other toxic or deterring substances in ferns. However, their effect as defending mechanisms against insects was not stronger than that of biochemical compounds of angiosperms (SooHoo & Fraenkel, 1964; Southwood, 1973; Cooper-Driver, 1976; Lawton, 1976). 304 FERN GAZ. 17(5): 303-309. 2006 Table 1. Insect families with highest numbers of fern feeding species (summarized from Balick et al. 1978). Insect family Insect order No. of fern feeding species Aphididae Homoptera 104 Curculionidae Coleoptera 85 Tenthredinidae Hymenoptera 48 Noctuidae Lepidoptera 44 Miridae Heteroptera e Diaspididae Homoptera ad Anthomytidae Diptera 20 Aleyrodidae Homoptera 18 Lygaeidae Heteroptera 18 ixlidae Auchenorrhyncha 14 Lecaniidae Homoptera 4 Pyralidae Lepidoptera 12 Pseudococcidae Homoptera 11 Tortricidae Lepidoptera 11 Balick er al. (1978) conducted the first preliminary field studies on this topic and found no evidence that ferns are damaged less than angiosperms. Hendrix (1980) concluded that some insect groups might be under-represented on ferns. Both authors ascertained that research data are too incomplete for a final conclusion. Balick’s review lists data for 75 insect families feeding on ferns, with dominance of insects with sucking (vs. chewing) mouth parts. Homoptera, Coleoptera, Hymenoptera, Lepidoptera, and Heteroptera comprise the highest number of fern feeding species (Table 1). Most herbivorous species were reported for the cosmopolitan weed Preridium aquilinum due to the search for biological control organisms (Kaplanis et al., 1967; Carlisle & Ellis, 1968; Weiczorek, 1973; Lawton, 1976; Hendrix, 1977), while for other fern genera data obviously reflect our lack of knowledge, but not the real numeric interactions between insect herbivores and ferns (Table 2). Only 128 interactions refer to fern species, and not to genera or entire groups (Balick et a/. 1978). Hendrix’s (1980) review lists 465 fern-feeding insect species, but few ferns were identified to species level. Other field studies by Hendrix and Marquis (1983) and Mehltreter and Tolome (2003) estimated damages to three terrestrial fern species, the first ones at La Selva (Costa Rica), the later ones at the same forest site as the present study in Mexico. Both studies confirmed that leaf damage on ferns is similar to that on angiosperms (5-15 %), although these may be caused by a smaller number of insect species. Until now, leaf damage was studied only on terrestrial ferns. Consequently, our objective was to investigate whether epiphytic ferns present lower levels of leaf damage than terrestrial ferns, because we supposed that the limited availability of water and nutrients would result in lower growth rates and consequently better defence mechanisms to avoid leaf damage. Additionally, we investigated whether leaf age, leaf fertility and leaf life-span are positively correlated with leaf damage. MEHLTRETER etal: HERBIVORY ON EPIPHYTIC FERNS 305 Table 2. Fern genera with highest numbers of attacking insect species (summarized from Balick et a/. 1978). Fern genus No. of insect species Pteridium 119 Asplenium 50 Cibotium ai Dryopteris 26 Adiantum 26 Polypodium Pe Pteris 23 Nephrolepis 20 ‘yathea 17 Athyrium 16 Polystichum 15 Blechnum 14 METHODS The study site is a tropical lower montane forest 2.5 km south of Xalapa, Veracruz (19°30’N, 96°57’ W), at 1300 m elevation. Climatic conditions are seasonal with a dry period from November to April, a mean annual temperature of 18°C and a mean annual precipitation of 1500 mm. We studied herbivory on five epiphytic fern species: Pleopeltis crassinervata (CR, 22 individuals), Polypodium furfuraceum ( FU, 19), Polypodium plebeium (PL, 30), Polypodium polypodioides (PO, 14), Polypodium rhodopleuron (RH, 25). Each leaf was tagged, and classified as young (not fully expanded) or mature (fully developed), and as sterile or fertile. Finally we checked each leaf for the presence of miners, and estimated herbivory as leaf area loss for each pinna, using a scale of seven damage classes (0%, less than or equal to 10%, <=25%, <=50%, <=75%, <100%, 100%). Herbivory data were estimated during the dry season in February 2003 and one year later. Estimations for each pinna were averaged for each leaf and individual using the median of each damage class. Data were analysed with ANOVAS on ranks. Paired t-tests comparing individuals/branch sections were applied for each species to test for differences between years. T-tests were applied for differences between leaf ages and differences between sterile and fertile leaves. Leaf life-span was calculated as the mean leaf number per plant (of both years) divided by leaf production and multiplied by the number of months of the observation period. Statistical analyses were performed with SIGMASTAT (1995). RESULTS AND DISCUSSION In 2003, Polypodium plebeium had the highest leaf damage of 21.2 + 1.9 %, while the other four species had similar and significantly lower levels of damage, 8.4 — 9.3 % (ANOVA on ranks, H = 37.15, df = 4, P < 0.001, Dunn’s Method, P < 0.05)(Figure 1). In 2004 damage for Pleopeltis crassinervata and Polypodium polypodioides increased significantly (paired t-test, t = -4.00, df = 20, P< 0.001, and t = -2,28, df= 11, P= 0.031, respectively) to 17.9 + 2.2 % and 16.0 + 2.3 %, respectively, and was similar to that recorded for P. plebeium (22.0 + 2.2 %). Polypodium rhodopleuron had the lowest 306 FERN GAZ. 17(5): 303-309. 2006 SCR OFU mPL GPO RH. Mean leaf damage (%) | 2003 2004 | | Year Figure 1. Mean leaf damage (+ | SE) of five epiphytic fern species in 2003 and 2004. Different letters indicate significant differences (ANOVA on ranks, P < 0.05) among fern species within the same year. CR = Pleopeltis crassinervata (n = 22 individuals in 2003, 21 in 2004), FU = Polypodium furfuraceum (19, 17), PL = Polypodium plebeium (30, 30), PO = Polypodium polypodioides (14, 12), RH = Polypodium rhodopleuron (25, 24) ‘Olyoung leaves H mature leaves _ & < oO 20 | oD 13] EF 15- © go | ‘ws 10 - & a a Jj feb] | | = | CR FU PL PO RH | Species Figure 2. Mean leaf damage (+ | SE) of young and mature leaves of pooled data of 2003 and 2004. Differences between leaf ages were significant for all species (t-test, P < 0.001). CR = Pleopeltis crassinervata (n = 93 young leaves, 331 mature leaves), FU = Polypodium furfuraceum (31, 190), PL = Polypodium plebeium (27, 347), PO = Polypodium polypodioides (41, 244), RH = Polypodium rhodopleuron (6, 374). MEHLTRETER etal.: HERBIVORY ON EPIPHYTIC FERNS 307 1 Ofertile leaves M sterile leaves 35 30 a Mean leaf damage (%) s p CR FU PL PO RH Species Fic. 3. Mean leaf damage (+ 1 SE) of fertile and sterile leaves of pooled data of 2003 and 2004. Different letters indicate significant differences (t-test, P < 0.05) between leaf types of the same species. CR = Pleopeltis crassinervata (n= 36 fertile leaves, 388 sterile leaves), FU = Polypodium furfuraceum (133, 88), PL = Polypodium plebeium (25, 349), PO = Polypodium polypodioides (21, 264), RH = Polypodium rhodopleuron (176, 204). levels of damage of 9.7 + 1.5 %. High leaf damage on P. plebeium may be explained by a higher number of feeding insect species or a higher number of individuals of the same species. The leaf damage to three terrestrial fern species at the same site (5.8-11.1 %, Mehltreter & Tolome, 2003) was similar to or slightly lower than that on the epiphytic species, but this may be a consequence of their shorter life-span (K. Mehltreter, pers. obs.). By contrast, leaf damage of epiphytic orchids and bromeliads in the same forest was about an order of magnitude lower than in ferns (Winkler et al., 2005). We found no significant correlation between leaf damage and leaf life-span, possibly because of the low number of studied species. However, Polypodium plebeium had the greatest damage and the longest leaf life-span of 29.5 + 3.4 months while the leaves of Polypodium rhodopleuron had the shortest life-span of 12.0 + 0.0 months and the less damage. In fact, the leaf life-span of this species is probably less than 12 months, because all leaves appear to be shed in the dry season, but an annual data survey cannot resolve shorter time-spans. The other three species had intermediate leaf life-spans of 20.5 + 2.9 months (Polypodium polypodioides), 20.6 + 2.0 months (Polypodium furfuraceum) and 20.8 + 3.3 months (Pleopeltis crassinervata) and intermediate levels of leaf damage (Figure 1). On all species mature leaves were significantly more damaged than young leaves (t- test, P < 0.001, Figure 2), which indicates that herbivores do not feed exclusively on young leaves of understory plants as reported by Coley & Aide (1991), but continue feeding on mature leaves, as we also directly observed. Sterile leaves showed greater damage than fertile leaves in Pleopeltis crassinervata (t = -2.489, df = 422, P = 0.013) and Polypodium rhodopleuron (t = -5.021, df = 378, P < 0.001), but in the other species there was no difference (Figure 3). These results 308 FERN GAZ. 17(5): 303-309. 2006 confirm the observations of Mehltreter & Tolome (2003) on terrestrial ferns that fertile leaves are not exposed to a higher selective pressure of herbivores, especially if we suppose that both leaf types of species with monomorphic leaves possess similar life- spans (Mehltreter & Palacios-Rios, 2003). Miners affected four of the five fern species, and were found on 0.9 % (Pleopeltis crassinervata), 2.4 % (Polypodium plebeium), 3.6 % (Polypodium furfuraceum) and 17.4 % (Polypodium rhodopleuron) of the leaves. Polypodium polypodiodes was not damaged by miners. Polypodium rhodopleuron was the species with the shortest leaf life-span, but the heaviest attack by miners. Short-living leaves normally possess fewer biochemical defences, and consequently might be more likely to be infested with miners. CONCLUSIONS We cannot confirm our hypothesis that epiphytic ferns are better protected against herbivores and consequently less damaged than terrestrial ferns. However, we are able to conclude that: . Leaves are attacked by herbivores during their entire life-span and not only during their early development. 2. Leaf life-span may be one of the most significant factors to interpret results on herbivory, and should be registered within herbivory studies. 3. Damage of fertile leaves is similar to sterile leaves. ACKNOWLEDGEMENTS Javier Tolome and José Luis Gonzalez Galvez helped during laboratory work. This research was supported by the Instituto de Ecologia, A. C. (902-17-796 to K.M.) and the Austrian Science Fund (FWF grant number P14775 to P.H.). REFERENCES BALICK, M.J., D.G. FURTH, & COOPER-DRIVER, G. 1978. Biochemical and evolutionary aspects of arthropod predation on ferns. Oecologia 35:55-89. CARLISLE, D.B. & ELLIS, P.E. 1968. Bracken and locust ecdysones: Their effects on molting in the desert locust. Science 159:1472-1474. COLEY, P.D. & AIDE, T.M. 1991. Comparison of herbivory and plant defenses in temperate and tropical broad-leaved forests. In: PRICE, P.W., LEWINSOHN, T.M, FERNANDES, GW. & BENSON, W.W. (Eds.) Plant-animal interactions: evolutionary ecology in tropical and temperate regions, pp. 25-49. John Wiley & Sons, New York. COOPER-DRIVER, G. 1976. Chemotaxonomy and phytochemical ecology of bracken. Bot. J. Linn. Soc. 73: 35-46. EHRLICH, P.R. & RAVEN, P.H. 1964. Butterflies and plants: a study in coevolution. Evolution 18: 586-608. HENDRIX, S.D. 1977. The resistance of Pteridium aquilinum (L.) Kuhn to insect attack by Trichoplusia ni (Hiibn.). Oecologia 26:347-361. HENDRIX, S.D. 1980. An evolutionary and ecological perspective of the insect fauna of ferns. Am. Nat. 115:171-196. HENDRIX, S.D. & MARQUIS, R. J.1983. Herbivore damage to three tropical ferns. Biotropica 15:108-111. KAPLANIS, J.N., THOMPSON, M.J., ROBBINS, W.E. & BRYCE, B.M. 1967. Insect MEHLTRETER etal: HERBIVORY ON EPIPHYTIC FERNS 309 hormones: Alfa ecdysone and 20-hydroxyecdysone in bracken fern. Science 157:1436-1438. LAWTON, J. 1976. The structure of the arthropod community on bracken (Preridium aquilinum (L.) Kuhn). Bot. J. Linn. Soc. 73:187-216. MEHLTRETER, K. & TOLOME, J. 2003. Herbivory of three tropical fern species of a Mexican cloud forest. In: CHANDRA, S. & SRIVASTAVA, M. (Eds.) Pteridology in the new Millennium, pp. 375-381. Kluwer, Dordrecht. MEHLTRETER, K. & PALACIOS-RIOS, M. 2003. Phenological studies of Acrostichum danaeifolium (Pteridaceae, Pteridophyta) at a mangrove site on the Gulf of Mexico. J. Trop. Ecol. 19:155-162. SCHNEIDER, G. 1892. The book of choice ferns, Vol. 1. Gill, London. SIGMASTAT 1995. Sigmastat, statistical software, version 2.0. Jandel Scientific Software, San Rafael. SOO HOO, C. & FRAENKEL, G. 1964. The resistance of ferns to the feeding of Prodenia eridania larvae. Ann. Entomol. Soc. Am. 57:788-790. SOUTHWOOD, T.R.E. 1973. The insect-plant relationship — an evolutionary perspective. Symp. Royal Entomol. Soc. Lond. 6: 3- WEICZOREK, H. 1973. Zur Kenntnis der Adlerfarninsekten: Ein Beitrag zum Problem der biologischen Bekimpfung von Preridium aquilinum (L.) Kuhn in Mitteleuropa. Ann. Angew. Entomol. 72:337-358. WINKLER, M., HULBER, K., MEHLTRETER, K., GARCIA FRANCO, J. & HIETZ. P. Herbivory in epiphytic bromeliads, orchids and ferns in a Mexican montane forest. J. Trop. Ecol. 21:147-154. 310 FERN GAZ. 17(5). 2006 a1) FERN GAZ. 17(5). 2006 BIODIVERSITY AND CHOROLOGY OF PTERIDOPHYTES FROM BUENOS AIRES PROVINCE, ARGENTINA J.P RAMOS GIACOSA,' E.R. DE LA SOTA! & GE. GIUDICE'! 'Catedra de Morfologia Vegetal, Facultad de Ciencias Naturales y Museo, UNLP, Paseo del Bosque s/n°, B1900FWA La Plata, Argentina. (Email: jpramosgiacosa@hotmail.com) Key words: Biodiversity, Chorology, Pteridophytes, Buenos Aires, Argentina. Buenos Aires Province is situated between 33° 16’- 41° 02° S and 56° 39’- 63° 23° W. It has an area of 307.571 km? and is the biggest and the most populated province in Argentina. Studies of pteridophytic diversity and chorology for this region were carried out as a first step for the conservation of species. In the studied area 20 families, 41 genera and 87 specific and infraspecific taxa are recorded. Pteridaceae is the most diversified family. The biodiversity of Buenos Aires is less rich than that of northwestern and northeastern Argentina but richer than the boundary provinces. For the chorological study, 0° 10° x 0° 15” latitude x longitude squared maps were elaborated and the presence of taxa was represented by dots. The analysis of the maps shows that the species are concentrated in three areas of biodiversity: 1) Tandilia hills 2) Ventania hills 3) La Plata estuary environs. Between these three areas there is a notable decrease in the number of squares occupied by Pteridophytes, with the only records being for Azollaceae and Marsileaceae. The squares situated at 38° S have the richest biodiversity of the province and in these areas conservation should be prioritised. 312 FERN GAZ. 17(3). 2005 INSTRUCTIONS FOR AUTHORS PAPERS should not usually exceed 20 printed pages and are generally expected to be considerably shorter. Review articles, as well as reports of original research, are encouraged. Short notes are acceptable e.g. new records. The senior author should supply a fax and email address to facilitate correspondence. MANUSCRIPTS should be submitted in English (British) in electronic format (preferably) or hard copy (two copies), in 10-point Times New Roman font and gales spaced. Electronic versions of text and tables should be compatible with WORD, w figures as pdf or jpg files, and sent as email attachments or CDroms. All hemor will be refereed THE TITLE should reflect the content of the paper and be in BOLD CAPITALS (11- point) and centrally aligned. Generic and specific names should be in italics and any title containing a generic or specific name must be followed by the family and Pteridophyta in brackets e.g. TRICHOMANES SPECIOSUM (HYMENOPHYLLACEAE: PTERIDOPHYTA) IN SOUTHERN SPAIN AUTHOR ABBREVIATIONS should follow Pichi Sermolli's (1996) Authors of scientific names in Pteridophyta, Royal Botanic Gardens, Kew. MAIN HEADINGS: should be in BOLD CAPITALS (10-point) and centrally aligned. SUBSIDIARY HEADINGS: should be in bold, the first letter of each word in capitals, the rest in lower case and left-aligned. AUTHORS' NAMES AND FULL ADDRESSES: follow the title and are centrally aligned. KEY WORDS: up to ten. ABSTRACT: should reflect the content of the paper. FIGURES: there is no distinction between photographs and line drawings in numbering. All should be presented in a form ready for reproduction, ideally in JPG format (please contact editor with queries), with a scale bar where appropriate. Lettering or numbers (Arabic) should be in the bottom left using uppercase Times Roman and be sufficiently large to be legible if reduction is necessary during printing. The number of photographs allowed in any one issue is limited by cost. Figure captions should be on a separate shee TABLES: can be printed in either portrait or landscape format. Authors should consider this when preparing tables. Authors should ensure that tables fit the printed page size in a legible form. MEASUREMENTS: should follow the metric system. CHECKLISTS: should follow the format of Baksh-Comeau, Fern Gaz. 16(1, 2): 11- 22, REFERENCES: should follow the style of a recent issue of The Fern Gazette, e.g.:- HOOKER, W.J. 1864. Species Filicum, 5. Dulau & Co., London. MORTON, C.V. 1947. The American species of Hymenophyllum, section Sphaeroconium. Contr. U.S. Natl. Herb. 29(3): 139-201. STEVENSON, D.W. & LOCONTE, H. 1996. Ordinal and familial relationships of pteridophyte genera. In: CAMUS, J.M., GIBBY, M. & JOHNS, R.J. (Eds) Pteridology in perspective, pp. 435-467. Royal Botanic Gardens, Kew. JOURNAL ABBREVIATIONS: should follow Botanico Periodicum Huntianum & Supplements Alterations from the original text at proof stage will be charged for unless they are minor points of detail. Twenty-five offprints will be supplied free to the senior author. THE BRITISH PTERIDOLOGICAL Registered Charity No. 1092399 Patron: HRH The slim of Wales . March Officers and Committ President: Dr A.F. Dyer, 499 ek Road West, Balerno, Edinburgh EH14 7AL ae _ E-mail: pieg sien aia Om Vice-Presidents: H. Rickard, Prof. B.A.Thomas © Honorary General Secretary: Dr Y.C. Golding, 7 Grane pm: pores Derbyshire SK17 SNH _ E-mail: Secretary@eBPS.org.uk Treasurer: Mrs G Smith, Rookwood, | Prospect Road, ie Broad, Lowestoft, Suffolk es NR32 3PT; E-mail: Treasurer@eBPS.org.uk — ae Membership Secretary: M.S. Porter, y West Avenue, Wi; ston, Cumbria CAT Reo | : E-mail: Mem Meetings Secretary: _ e PJ. Acock, 13: Star Lane, St Lares Cray, Conservation Officer; Dr HS. McHaffie, 180 Granton n Roa, Eainburgh EHS TAH : Conservation Officer: — ‘Dr EJ Rumsey Dept. of Botany, N tural H saoky Mess eh & Recorder aS . asta oad, L ae E-mail: Co ion(@eBPS. g ik/Reck der@eB Editor of the ee Miss A.M. Paul, Dept. of B Botany, The! tur. Cromwell pr green SW7 SBD; E-mail: Bu Editor of The Fern Gazente: Prof. M. Gibby, Royal Botanic Gai dinbur Inv erleith Row, Edinburgh EH3 SLR; E-mail: FernGazette@eBPS.org.uk Editor g Preridologist: Dr J.W. Merryweather, ‘The Whins’, Auchtertyre, by Kyleof ochalsh, Wester Ross Iv40 8EG; E-mail: Pteridologist@eBPS.org.uk eS Editor of BPS WWW Site-» www.eBPS.org.uk: _-A.C. Pigott, Kersey’s Farm, Mendlesham, . —— Suffolk IP14 SRB; E-mail: Webmaster@eBPS. org.uk : Elected Commiiee Members: -R.G Ackers, A.R. Busby, Dr M. oe Dr SD. Martinelli, F. McGavigan, B.D. Smith, R.W. Sykes Booksales Organiser: S.J. Munyard, 234 Harold Road, Hastings, East Sussex TN35 5NG E-mail: Booksales@eBPS. org.uk Horticultural Maylene, Officer A.R. Busby. 16 Kirby Comer Road, Canley, Coventry CV : Archiv E-mail: Horti uk / A Si che es Bo Br oad Loesoh: Suffolk NR32 3PT: E-mail: Merchandise@eBPS.org.uk Plant Exchange Organiser: ecg sg re Seis Cottage, Tredilion, Abergavenny, Gwent 8BB; E-mail: PlantExchange@eBPS.org. uk | aoe Exchange Organisers: Mr B. & Mrs A. x Wright, 130 Rriace vega pide Tockwitt h ; York Y' pete Trustees of Greenfield & Centenary Funds: _ Dr A.F. Dyer, Dr ¥.C. ¢ sol ‘The BRITISH PTERIDOLOGICAL SOCIETY was founded in 1891 and today continues as seas for THE FERN GAZETTE VOLUME 17 PART 5 2006 CONTENTS MAIN ARTICLES . Pes inecke « sci and ‘evolution of the —— Hymenophyllum - ce ep asaconaes Pteridophyta) ‘5S. J poacinaaee Ebihara, M. Ito, K. Iwatsuki & J.-Y. Dubuisson 247-257 Ss Helfer ee oe oS 259-261 henologice of frond prpleciai in Alsophila setosa yatheaceae: Pteridophyta) i in southern Brazil ae Schmit & & PG hoamieas os 263-270 rom m stipules : ie iM. Huang & c. M. Chen ns from the tertiary 0 of western North America: Osmu