University of Alberta

CIRCUMSCRfPTION AND PHYLûûENETiC ïRENDS IN THE ORTHOTRI~ (BRYOPSTDA)

Bernard Goffi.net O

A thesis submitted to the Faculty of Graduate Studies and research in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Department of Botany

Edmonton, Alberta Spring, 1997 Acquisiüons and Acquisitions et . Bibliographk ~~ - senices biiraphiques

The author has ganted a non- L'auteur a accordé une licence non exclusive licence ailowing the exc1usk permettant à la National Liirary of Canada to Bibiiothèque natiode du Canada de reproduce, toan, hiuteor selI rrprodiiirr, prêter, distribuer ou copies ofhismer thesis by any means vendre des copies de sa îhèse de and in any form or format, mahg ~uel~uemanière et sous quelque this thesis available to interested forme que ce soit pour mettre des persons. exemplaires de cette thèse a la disposition des personnes intéressées.

The author retains ownenhip of the L'auteur consewe la propriété du copyright in hider thesis. Neither droit d'auteur qui protège sa thèse. Ni the thesis nor substantial extracts la thèse ni des extraits substantiels de fkom it may be printed or othenvise celle-ci ne doivent être imprimés ou reproduced with the author's autrement reproduits sans son permission. autorisation.

With over 500 species, the Orthotnchaceae () are one of the largest fdes of . Twenty-fivc genera are currentiy included in the Orthotrichaceae, with Macromitium and Onhotrichunr accounting for most of the taxonomie diversity. Together with the Erpodiaceae, Rhacbitheciaceae, Microtheciellaceae and the Helicophyllaceae, they forrn the Orthotrichales. The limits and the relationships of the family and the order are here examined. The Microtheciellaceae are now excluded from the Orthotrichaies on the basis of their pleurocarpy. The Helicophyllaceae are characterized by a unique combination of gametophytic characters, and are tentatively excluded, although with no clear affinities to other groups. Cornparisons of nucleotide sequences of the chloroplast gene rbcL suggest that the and the Erpodiaceae belong to the Haplolepideae, and should therefore be excluded from the Orthotrichaies. The genera Meioweisiopsis, and Trigonodicryon, as well as Octogonella and Uleasnum are excluded from the Orthotrichaceae based on cornparison of morphological characten, and transferred to the Ditrichaceae, Grimmiaceae, and Rhachitheciaceae, respectively. Pleurozygodontopsis is placed in synonymy with Zygudon. Cladistic analysis of 37 rbcL nucleotide sequences fuahennore lead to the exclusion from the Onhotrichaceae of two gymnostomous genera, namely Arnphidium, and Dnunmondia, boîh with afnnities to the Haplolepideae. Twenty-two genera, including two new genera, Bryomaltaea and Maneria, as well as the reinstated genus Codonoblepharon, are now accepted within the Orthotrichaceae. These genera are disaibuted between four tribes belonging to two subfamilies. The sister-group to the Orthotrichales, now reduced to the Orthotrichaceae, remains somewhat arnbiguous but is hypothesized, based on morphological characters to be the Bryales, rather than the Splac hnales. 1am very gratefd to Dale Vitt for giving me a chance to pursue my interests in Bryology by takkg me on as a graduate student. His passion for mosses has been a tremendous source of inspiration. Our interactions continuously stimuiated my bryological curiosity. His trust gave me the confidence and the -dom to go beyond the initial project and examine evolutionary hypotheses in related groups of mosses. 1 am also gratefd to Randall Bayer who joint Dale in supe~singmy thesis. Randall Bayer gave me the opportunity to explore the world of DNA sequencing. I thank him for accepting the difficult task of initiating me to molecular systematics, and for dl his support and help throughout this study.

This research was financially supported through grants fiom the Natural Sciences and Engineering Research Council of Canada to Dale Vitt (A-6390) and to Randall Bayer (A- 3797). Field work in Panama and the West Indies was made possible by a financial support nom the Nationai Geographic Society to Dale Vitt and myself. 1dso wish to thank the University of Alberta for a Dissertation fellowship as well as the former Department of Botany for financial support during the summer of 1991, 1992, and 1993. Mark Dale, Brian Jones, WiiîÎed Schofield, and Mark Wilson are thanked for their help at various stages of this research and for their constructive criticism on an earlier draft of the thesis. Throughout this study 1benefited fkom continuous interaction with Catherine La Farge- England, a dear fiiend with whom 1 shared my failures and successes, my passion for bryology and my drearns of a perfect gel! The niendship of Trevor Goward, Marie- Noëiie Hindryckx, and Paul Boxus have over the last six years proven more than once to be the best medicine (besides Belgian chocolate) for Lifting my spirits, in difficult times. 1 wish to thank Ross Hastings for sporadic yet always interesting discussions on mosses. 1 also thank Brett Purdy, and Sarah Hambleton for their help in the laboratory, Alex Espinosa and Pascal Bourgeois for their assistance in the field, as well as Linda Halsey for "computer time". Fially 1wish to express my dearest thanks to Sandi Vitt for ber patience and her help (particularly the 24 hours computer side assistance). 1 am grateful to Douglas and Pamela Soltis (WSU) and their former students (Leigh Johnsson, Gregory Plunkett) as well as Russeil Chapman and Debra Waters (LSU) for teaching me DNA sequencing techniques while visiting their laboratory. William Buck (NY), Celina Matteri (BA), and Heinar Streimann (CANB) kindly send me fiesh material of Cardoriefia quinquefuria, Macrocoma papillosa, and Desmotheca apicdata for DNA extraction, for which 1am grateful.

My th& are extended to William Buck (NY), Patricia Geissler (G), and Robert Magili (MO) for making older literature available to me. 1am also thankful to Marshall Crosby for promptly responding to several nomenclanirai questions. This work would not have been possible had 1not had access to critical herbarium collections. 1am gratefid to the curators of BM, FH,G, H, J, MO, NY, and S for the loan of specimens over the last six years. My interest for both mosses and lichens surfaced during a field excursion at the ScientSc Station "Mont-Rigi" (University of Liege). 1am very gratehil to Emmanuël Sérusiaux and René Schumacker for their support and patience in the early hours of my Lichenological and bryological endeavors. Enfin je souhaiterai remercier Anne dont la patience a été Zi la hauteur de ma passion pour ce travail. Puisse l'avenir auquel cette thèse nous conduit être sa récompense. Et puis il y a Louis, qui hit ma bouée de secours ancrée dans l'autre monde, celui de tout les jours, loin des mousses mais phdu Monde. Finalement mon rêve n'aurait pO se réaliser si je n'avait pas bénéficié d'un soutien quasi inconditionel de mes parents. Je vous remercie pour votre confiance. Merci aussi à François, Roger, et tous les autres pour votre soutien. Chapter 1: Intmduction ...... 1 Objectives ...... 3 Literaîurecïi ...... 6 Chapter 2. A recoarideration of the affidties of Kleioweisiopsis. Pleuro~ygodontopsis.Trigonodictyon. and the WcrothecieUaceae (Bryopsida. Orthotrichales) ...... -9 KIeio weisiopsis Dixon ...... 10 Zygodon Hook & Tay1 ...... 12 Tj+godictyon Dixon & P. & la Varde ...... -14 MicmtheciellaDixon ...... 16 Literanirecited ...... 21 Chapter 3. The Rhacbitheciaceae: revised generic circumscription and ordinal affinities ...... : ...... -23 Rhachitheciopsis P .de la Vade ...... 25 Rhachithecim Le Jolis ...... -27 UZea~trumBuck ...... -29 Zandericr Goffiaet ...... 30 Jonesiobryum Bizot & P&s ex Man & PurseJi ...... -32 Artificial key to the Rhacbitheciaceae ...... 33 Systematic affinities of the Rhachitheciaceae ...... 33 Literature cited ...... -44 Chapter 4. Cardotiella elimbuta (Thér.) GotXinet comb . nov., and the signifïiennce of dorsdy uniseriate endostome segments in the Orthotricbaceae ...... -47 Cardotielh elwu(Thér.) Gofiet ...... -47 Literatunz cited ...... -35 Chapter 5. Circumscciption and phylogeny of the Orthotrichales inferred from rbcL sequence anaiyses ...... -56 Materialandmethods ...... 6û DNA extraction and KR-DNA amplification ...... 60 Sequencing and alignment ...... -61 Sequence and phylogemtic analysis ...... -61 .,... Results ...... ,... 62 Discussion ...... , ...... ,., RbcL sequence data ...... 64 Circumscription of the Orthotrichales ...... 65 Ordinal relationship of the Orthotrichdes ...... 67 Subfamilial phylogenetic &üioaships ...... 4 Phylogenetic relationships of Chthoaichaceous genera ...... 71 Affinities of excluàed taxa ...... -74 Phylogenetic conclusions ...... 75 ...... Litemture cited ...... -87 Chapter 6. Circumscription. and phylogenetic trends in the subfdlies of the OrthoMchaceae interd from morphdogy ...... 95 Materhis~dmthods ...... 97 Selectionoftaxa ...... 97 Dataanalysis ...... 99 Results ...... 100 Character def"iaitionand coding ...... 100 Charactersexcludtd ...... 110 Phylogenetic relationships ...... -113 Discussion ...... 114 The ûrthotrichaceae: morphological chanctesizatïon and hfkafhmiliai relationships...... 114 Circumscription and relationships within the Zygodontoideae ...... 118 CircumScnption and relationships witbin the Orthotrichoideae ...... 122 CircumScnption and nlationships within the Macmmitrioideae ...... 124 Phyiogenetic conclusion ...... 129 Literaturecited ...... 139 Chapter 7. Reoised gewric classification of the Orthotricbaceae ...... -146 Artinciai key to the genera of the Orthotrichaceae ...... -159 Literature cited ...... 160 Chapter 8. Conclusion ...... 162 Ciwllscription of the Onhotrichaies ...... -162 Circumscription of the Orthotrichaceae ...... -163 Major phylogenetic relatimhips ...... -163 Literature cited ...... 165 Appendir one. RbcL gene sequences of 37 arthrdontous mosses ..... -168 LIST OF TABLES Table 1.1. Cunent generic classification of the Orihotrichaceae (updated from Vitt 1982a) and the number of pieshclude in each genus ...... 2 Table 12Systematic position of the Orthotrichaceae and putative ~Iatedfdes in major classification schemes ...... 5 Table 3.1. Morphological chanrtexhtics sepacating the genera of the Rhachitheciopsis andRhachithecium ...... 42

Table 3.2. Morphologicai characteristics separating the genera ZiurderUlr and Uleamir... -A

Table 5.1. Synthetic primers (5'03') used for sequencing the rbcL gene in mosses .. -8 1 Table 5.2. Taxa for which the nbcL gene was obtahed in this sbidy (dl vouchers deposited at ALTA unless ottiezwise noted) ...... 82 Table 5.3. Dishiiution and fkquency (46) of constant and phyIogenetically informative sites over all taxa and over the ûrthoaichaceae only ...... 84 Table 5.4. PairWise cornparison of rbcL nucleotide sequences within the Onhotnchaceae.

Table 6.1. List of representative specixnens examined ...... 136 Table 6.2. Morphological characters scored for 37 species representing 3 outgroup taxa and 34 Orthotrichaceae ...... -..... 137 Figure 2.1. Kleioweisiopsisdrnticulato ...... 18 Figure 2.1. Pleurozygudontopsis decurrem ...... 19 Figure 2.3. Tngonodictyon idicum ...... -20 Figure 3.1. RluichitheciopsL tisserantii ...... -36 Figure 3.2. Morphology of Rhachiihciup~tisserantii and Rhachitrheeum perpusillm 27 ...... LI...... LI...... *.....*...... *...... ~~ Figure 3.3. Morphology of the peristome teeth in the Rkhitheciaceae ...... 38 Figure 3 -4. Rhachithecim papiltosum ...... 39 Figure 3.5. Menaoctoblepharis ...... 40 Figure 3.6. Zmtderia octoblepharis ...... -41 Figure 4.1. Carhtiella elimbata ...... 32 Figure4.2. Carhtiellaelimbata ...... -53 Figure 4.3. Architecture of endosiorne segments in Cardotiella elimbata, Florschuetziella steerei, and PIeurotthotnkhum chilense ...... -54 Figure 5.1. Distribution of phylogenetically informative characters, the average consistency index, and the average number of steps per phylogenetically uiforrnative characm dong the rbcL sequence...... 77 Figure 5.2. Phylogenetic tree reconstructed using the neighbor-joining methoci, with Tamura's distance parameter including both transitions and tramversions ...... -78 Figure 5.3. Swmuyconsensus trees of most parsimonious Fitch trees found in a heurimc search using rbcL sequence data with Sphagnm palustre and Andreaea nipestns as designated outgroups ...... -79 Figure 5.4. One of 39 most parsimonious Fitch mes found based on rbcL sequence data and using Sphagnum and Andreaea as outgroups...... 80 Figure 6.1. Strict conseasus tree of 56 most parsimonious mes obtahed from a cladistic analysis including 54 morphological characters...... 130 Figure 6.2. 50% majority deconsensus tree of 56 most parsimonious trees obtained from a cladistic analysis including 54 morphological characters...... 13 1 Figure 6.3. Phylograrn of tree number the ...... 132 Figure 6.4. Phylogram of tree numkr three showing the character transformations supporthg the monophyly of the heotrichacuie. as weil as the relationships between the subfdesand genera within the Zygodontoideae ...... 133 Figure 6.5. Phylogram of tree nwnber three showing the character transformations supporiingthe monophyly of the ûrthotrichoideae and the relationships within the subfamily ...... 134 Figure 6.6. Phylognun of tree number three showing the character traasformations supporthg the monophyly of the Macromitrioideae, and the nlationships within the subfamily ...... 135 Chapter one: L

Chapter one

Introduction

The ûrthotrichaceae, with over MO species, represent one of the largest families of mosses. The farnily is fourid in mosf major terrestrial biomes, hmthe arctic tundra to tropical rainfo~ests,and hmsubantarctic sea shores to subalpine heaths in the tropics. Adaptation to xemphytism (see Vitt 1981) is thought to have kn the major driving force in the diversification of the family, or at lemt of such speciose genera as Macromirn*unt(Vitt and Ramsay 1985), or Orthotrichwn (Vin 1971). Similar trends axe al& evident in two famiiies considered only distantly related to the ûrthotrichaceae (Vin 1984),namly the Gtimmiaceae (Churchill 1981), and the Poniaceae (Zander 1993). Both families are mostiy temcolous or saxicoious, whereas the Oahotrichaceae are predominantly epiphytic, aad within the genus Onhotrrchum, only the most denved species gmw on rocks (Vin 1971).

Twenty seven genera are currently included in the family, and of these, nineteen have 10 piesor fewer (Table 1.1). The taxonomie diversity in the Orthotrichaceae is concentrated in three genera, Orthotri'chum Hedw., Macromiîrium Bnd., and Zygodon Hook. et Tayl., that serve as type genera for three subfamilies. A fou& subfamily accommodates the small genus Dlummondia. The cumnt circumscription differs hmVin's (1984) concept of the family, mainly by the inclusion of Amphidium, Meioweisiopsis, and Uleustrum. Vitt (1973) argued against affinties of Aniphtdim with the ûtthotrichaceae and transfened the genus to the , whereas Lewinsky (1976) supponed her argument in favor of an orthotrichaceous origin of Amphidim by similarities in the anatomy of the um, particuiarly the anatomy of the trabeculae. Zander (1993) recently transferreà the genus Kleiwehiopsis from the Pottiaceae to the Orthoaichaceae, based on the similarities between K. denticulata and A. cyaihicarpwn. Zander (1993) aIso excluded UIea~trtmzfrom tbe Pottiaceae, and placed it in the Orthotrichaceae, "a disposition making the least necessaty emendation of recognized famiiy limits". Arnphidium and Kleiweisiopsis are gymnostomous (lack a peristome) whereas UIeasmmt has a peristome that a priori is incompatible with any of the four major peristome-types. Addnssing the familial Chapter one: 2 relationships of these genera may thus nly on the analysis of an independent set of data, such as DNA nucleotide sequences. Fleischer (1920). who was the nSt to IWO- the distinctiveness of the Onhotnchaceae at higher ranks. included in the Orthotrichaa &ides the Orthoeichaceae, also the Erpodiaceae (Table 1.2). The Rhachitheciaceae were later segregated hmthe Orthotnchaceae (Robinson 1964) and the Micmtkieilaceae were established by Miller and Hdgton(1977) for the erpodiaceous taxon, Microtkciella kecrei. The HelicophyUaceae. traditiondy coosidered with affinities to Racopilum in the Bryala sensu lato, were transferred by Crosby (1980) to the Orthotrichaceae. In the most recent classification (Vitt 1984) the Oahotrichales are thus circumscri'bed by five families: the ûrthoridhaceae, Erpodiaceae, Rhachitheciaceae, Microtheciellaceae, and Helicophyilaceae. the latter four of which are small families including fewer 20 pies. The relationsbips arnong these families are obscured by the great variation in gametophytic and spomphytic characters, and this heterogeneity, for example in the peristomial architecture (see Edwards 1979,1984) may be indicative of the polyphyly of the order. Many of the early species now recognized in Ma~rorni~um,Schlotheimia, and Ulota, were initially describeci as spies of Orthonchum. Amott (1825) may not have recognized the great taxonomie heterogeneity of this genus sensu Hedwig, but he was the first to consider Orthomchum sufficiently distinct to deserve its own family (as order "ûrthotrichoideae"). Bri&1(1826), in his extensive classification of mosses based on the distribution of the fedegametangia and the extent of the peristome, recognued orders and families but included Onhotnéhum and related genera in the large "order" Amphistomi including peristomate acrocarpous taxa Müller (1851) recognued the Orthotrichaaae (as a submbe) and placed it within the haplolepideous Poniaies (as tribe Pottioideae), between what is now considered the Pottiaceae and the Grinrmiaceae. Schimper (1856) segregated th Zygodontaceae hmthe Orthoeichaceae. and placed both families between the Encalyptaceae (also haplolepideous) and the Grimmiaceae. Iaeger (1874) too, considered the Orthotrichaceae and the Grimmiaceae to be related, and Brothem (1909) saw the @nhotric&aceae marking the transition from the Haplolepideae to the Diplolepideae. Fleischer (1920) based his classification primarily on the architecture of the peristom, and refognbd that the Orthotrichaceae were more closely related to other diplolepideous mosses than to the haplolepideae. He placed the Orthotrichaceae, together with the Erpodiaceae in the ûrthotrichineae, in the Chapterone: 3

Isobryales. Fleischer's (1920) classi£ication was adopted by Brothenis (1924- 1925) in his monumental treatise of the Bryopsida Subsequently flable 12.), the Chthotrichineae have mainly ken recognized at the ordinal taak (Dixon 1932, Crosby 1980, Walther 1983. Vitt 1982a). The most modem classification of mosses (Vitt 1984) divides the Bryopsida primarily on the basis of the architecture of the peristome foliowing Fleischer's (1920) system, which in tum is inspired hmPhilikrt's (1884-90)studies on the peristome. The arthrodontous pristome is composed of articulate teeth, and four distinct types can be recognized (Via 198 1). Three of these typicaiiy have two rows of teeth (diplolepideous peristome), with the endostom either opposite (Funnria-type;Schwartz 1994, Shaw, Anderson and Mishier 1989) or altemate to the exostom. The Orthotrichales and the Bryales smu hoshare an alternate arrangement of teeth (LeWinse 1989, Shaw 1985, Vitt 1981). but ciiffer by the presence of cilia (appendages in between the segments of the endostome) in the Bryaies (Shaw, Anderson and Mishler 1989). The fourth peristome-type (haplolepideous or Dicranum-type peristome; Shaw, Misider, and Anderson 1989) is chmcterized by a single row of teeth with asymmetric divisions occurriug in the innermost peristomiat layer. The direction of peristome evolution is not agreed upon (Crosby 1980, Vitt 198 1, Shaw and Roiver 1985). although most ment evidence (see Vitt, Gofïïnet, and Hedderson 1997) wggest that the FwnrU-type represents the ancestral type of pcristome, and that the Dicmnum-type diverged prior to the evolution of both the O~hot~chumand the Bryum-types. Whether the Onhonichum-type and the Brym-type actuaüy &fine natural groups of equal rank is subject of debate: the uniqueness of the Onhotriëhum-type peiistome (Vitt 198 1). may indwd not be sufficient to exclude a denvation hma Bgum-type peristome as suggested by Shaw (1985).

Objectives. The curent circumscription of the Onhotrichales and the Orthotrichaceae serves convenience rnorc than phylogenetic rationale. The purpose of this study is to examine the afflnities of the genera of the Orrhotrichaceae and the Olthooichales based on comparative morphology and cladistic analyses of morphological and molecular data. This study wiîl address the foiiowing questions: 1) are the Oabotrichaceae and the Orhotrichales monophyletic groups?, anci 2) what are the major phylogenetic relationships among the taxa that compose the Orthotrichaceae. Chapter one: 4

Table 1.1. Current generic classincation of the Orthotricbaceae (updated ftom Vin 1982a) and the number of species includcd in each genus

Taxon number of pies Zygodontoideae Brotherus Arnphidium Schimp. Kleioweisiopsis Dixon Leptodontiopsis B roth. Stenomim'um (Mitt*) Broch. Zygodon Hook. & Tayl. approx. 50 PIeurozygodontopsis Duon 1 Orthotricboideae B CO&. Bryoduonia Sainsb. 1 Ceuthotheca Lewiasky 1 Muelleriella Dusén 4 Orthomitrium Lewinsky-Haapasaari & Crosby 1 Orthotrichum Hedw. 118 Pleurorthommchum Broth. 1 Stoneobryum Norris & Robinson 2 Ulota Mohr approx. 35 Dnimmondoideae Vitt Dnrmrnondia Ho&. Macromitrioideae Broth. Cardotiella Vitt t Desmorheca Lindb. Florschuerziella Vitt Groutiella S teere approx. 10 Leiomimùm Mitt* 1 Leratiu Broth. 1 Macrocorna (CMüU.) Grout 10 Macromitrïum Brid. approx. 250 Schlotheimia Brid. approx. 50 Uncertain affinities Octogonella Dixon Trigonodictyon Dixon & P de la Vade Uleustrtun Buck

1 ais0 placed in its own subfamily by Brotherus (1925), Walther (1983), and Crum (1987) Table 1.2. Systematic position of the Orthotrichaceae and putative related families in major classification schemes (families included in the Orthotrichales by Vitt 1984 are in bold)

Jaeger (1874) Brolherus ( 1909) Flcischer ( 1920) Broihcrus (1924) Dinan (1932)

Tribe Grimmiaccue Grimmiaceoe Bryalcs lsobryalcs Orthotrichalcs Zygodonteae Orthotrichaceae lsobryalcs Orthotrichineae Erpodiaceae Orthatrichcae Funariaceoc Orthotrichineae Erpodiaceae Piychamiiriaccoe Erpodiaceae Orthotrichaceae Orthotrichaceae Orthotrichaceoe Rhacopilineae Eubryales Rhacopilineac Htlicophyllaccae Rhacopilinae Helicophyllaceae Hclicophyllaceae

Robinson (197 1) Crosby ( 198 1) Walther (1 983) Viti (1982b) Viii 0984)

Bryales Orthotrichalcs Grimmicks Bryidac Bryales Rhachithecirceac Orthotrichaceac Erpodiaccae Orthotrichales Orthotrichincac Erpodiaceae Erpodiaceat Orthotrichales Orthotrichaccac Orthotrichaccac Helicophyllaceac Rhachitheciaccae Orthotric haccae Erpodiaceae Erpodlaceae Orthotrichaccae ~icrothccicllaccae~ Rhachitheciaccac Rhachitheciaccae Rhrchlthcciaccac .,. Helicophyllaccae Hclicophyllaceae Microthccitllaccac Microthccicllaceic Heclwigiacese Hclicophyllaceat Helicophyllaceae Htlicophyllaceae Cryphaceae Wardiaceac

I çstsblishcd by Robinson (1964) * estrblishcd by Harrington and Miller (1977) a P, 'C1 1 Chapterom: 6

Litemtm cited Amott, GA. Walker-. 1826. Nouvelle disposition m&hodique des es@ces de mousses exactement co~ues.4-- M6moïres de la Société d'Histoire Naturelle de Paris 2: 249-320 (+ 412-414; prrpublication: 1825). Bride1 (-Brideri), S.E. 1826. Bryologia Universa sui Systematica ad Novam Methodm Dispositio, Historia et Descnptio moaium Musconmi Frondoso~n Hucusque cognitom cum synonymia exAucton'bus Pmbatissimis. Volumn hum. joan. Ambros. Bd,Lipsiae. Brothems, V. F. 1909. Musci (Laubmoose). III. Unterklasse Bryales: II Spezielier Teil. In A Engler's and K. PrantI [eds.], Die Natürlichen Wanzeafamilien 1 (3.1, II), 277-1246. W. Engelmann, Leipzig. 1925. Musci (Laubmoose) III. Unterklasse Bryales: II Spezieller Teil. In A. Engler's and K. Prantl [eds.], Die NatUrIicben Pnanzenfamilien, Band 10, 1. Hualfte, 143-478; Band 11,2. HHLfte, 1-542. W. EngelmaM, Leipzig. Churchill, S.P. 1981. A phylogenetic analysis, classification and synopsis of the geuera of the Grimmiaceae (Musci). In: V.A. Funk and D.R. Brooks [eds.], Advances in Cladistics. Roceedings of the fhtannuai meeting of the Willy Hennig Society, 12-79144. The New York Botanical Garden, New York. Crosby, M.R. 1980. The diversity and nlationships of mosses. In: R.J. Taylor, and A.E. Leviton [eds.], The Mosses of , 115-129. Califomia Acaâemy of Science, San Franscico. Dixon, & N. 1932. Classincation of mosses. In: F. Verdoom [ed.], Manual of Bryology, 397-412. Martius Nijhoff, The Hague. Jaeger, A.. and F. Sauerbeck. 1874. Genera et species musconun systematice disposita seu adumbratio florae musconun totis orbis terrarum. Bericht der SZ. Gaflischen Naturwissenscch4fchen Gesellsch@ 1872- 1873: 62-236. Lewksky, J. 1976. On the systematic position ofhphidiwn Schimp. Lindbergia 3: 227-231. 1989. Does the orthotrichaceous type of peristome exist? A study of penstome evolution in the genus Orthonchum Hedw. with possible denvation of the haplolepideous peristome. Joumal of the Hatlori Botmical Laboratory 67: 335-363. Miller, H.A., and Hanington, A.J. 1977. Microtheciellaceae, fam.nov. Journal of Bryology 9: 5 19: 524. Chapter one: 7

Müller, C. 1851. Synopsis Muscorum Frondosonim Maium Hucusque Cognitonun Pars Secunda, Musci Vegetationis Pleurocarpicae. Nb. Foerstner, Beroiini. Philibert, H. 18û4-189û. De l'importance du péristome pour les aff?nit& naturelies des mousses. Revue Bryologique 11: 49-52.65-72. Études sur le péristome, Revue Bryologique 11: 8 l-87,12: 67077.8 1-85; 13: 17-27,8 1-86; 14: 9-11,81-90; 15: 6-12,24-28,3744,50-56,56-6û, 6549,9043; 16: 1-9, 39-4467-77; 17: 8-12,25-29,3842. Robinson, H. 197 1. A revised classification for the orciers and famiLies of mosses. Phytdogia 2 1: 289-293. Schwartz, O. M. The development of the peristome-fohg layers in the Funariaceae. Intemations2 Journal of Plmrt Science 155: 640657. Shaw, J. 1985. Peristome structure in the Orthotrichaceae. Joumal of the Hanori Botmicul Laboratory 60: 119-136. , ,and B. D. Mishler. 1989a Peristome development in mosses in relation to systematics and evoiution. III. Funaria hygrometrica, Bryum pseuàocapillare, and B. bicolor. Systematic Botmy 14: 24-36. ,B. D. Mishler, and L. E. Anderson. 1989b. Peristome development in mosses in relation to systematics and evolution. IV. Haplolepideae: Ditrichaceae and . The Brydogist 92: 3 14-325. ,and J. R RohRr. 1984. Endostomial architecture in diplolepideous mosses. Joudof the Huttori Botanicd Laborutory 57: 41-6 1. Schimper, W.P. 1856. Corollarium. Bryologiae Europeae. Schweizerùart. Stuttgart. Vitt, D. H. 197 1 (1972). The infiageneric evolution, phylogeny, and of the genus Orfho~chum(Musci) in North Arnerica Nova Hedwigia 21: 683- 71 1. 1973. A revision of the genus Onhmichm in North Arnenca, north of Mexico. Bryophytomm Bibliutheca 1: 1-208. 198 1. Adaptive modes of the sporophyte. The Bryologist 84: 166- 186. 1982a The genera of the Orthotrichaceae. Beiheft zu Nova Hehuigiu 71: 261-268. Botanical Laborutory 54: 1-94. Chapter one: 8

1982b. Sphagnopsida and Bryopsidê In: SP. Parker [ed.], Synopsis and Classification of Living Organisms, vol. 1,307-336. 1984. Classification of the Bryopsida In R M. Schuster [ed], New Manual of Bryology, Vol. 2. Hattori Botanical Laboratory, Nichinan. Japan. ,aad H. P. Ramsay. 1985. The M(u:romimÙm complex in Austrdasia (Orthotrichaceae: Bryopsida). Part II. Distribution, ccology, and paleogeography. Jmmd of the Hattori Bomical Laborutory 59: 453-468. ,B. Goffinet, and T.A. Hedderson. 1997. The ordinal classification of the mosses: Questions and answers for the 1990's. Roceedings of the Centenary Meeting of the British Bryological Society. (in press). Walther, K. 1983. Bryophytina. LaubmooseseIn I. Gerloff, and J. Palt [eds.] A. Engler' s Syllabus der Manzenfamilien. Aufi. 13, V(2): 1-108. Gebrüder Bomtriiger, Berlin. Zander, KHI 1993. Genera of the Pottiaceae: Mosses of harsh environments. Bulletin of the BMalo Society of Nutural Sciences 32: 1-378. Chapter two: 9

A reromidemtion of the PiAnities of 1IJCiowciswpsis Ple~~~~godontopsis,Trigollodictyon, and the Micrdhedellriceae (Bryopsith, OrtbotdcLales)

The Oahotrichaceae is a cosmopolitan family, and with over 500 species is one of the most diverse families of artbraiontous mosses (Vitt 1982). The ûrthotrichaceae are found in a Fatvariety of habitats raaging fiom canopy branches in tropical upper montane forests (Macromim*umsp.; Vin & Ramsay 1985) to rocks nasea level on subantarctic islands (Muelleriella sp.; Vin 1976). Vin (1982) iisted 14 genera and characterized the family by a peristome of aiternating endostomial segments and exostomial teeth, a thick OPL (accounting for the recuwed teeth). an endostome lacking a basal membrane and cilia, costate leaves with neariy isodiametric, thick-waüed, papillose upper ceils, and no differentiated ahceiis. and by terminal perichaetia with additional growth by lateral innovations. Via (1984) later added Cardoriella Vitt, Octogonella Dixon, Pleurotygodonfopsis Dixon, Trigonodictyon Dixon & P de la Varde, and reinstated Leptodontiopsis Broth., Leratiu Broth., Leiornitrkm Mitt., and thus accepted 2 1 genera in the ûrthotrichaceae. Recently Nomis and Robinson (1987). Lewinsky (1994), and Lewinsky and Crosby (1996) established additional new genera Stoneobryum, Ceufhotheca, and Onh~mit~wn,respectively, whereas Zander (1993) transfemd the genera Uleastnun Buck and Kleioweisiopsis Dixon from the Pottiaceae to the Orthotrichaceae. With the controversial inclusion of the gymnostomous genus Amphidium (Lewinsky 1976; see also Vitt 1982) the family cumntly hcludes 27 genera, of which 15 are composed of no more than thtee species. The Oahotrichales sensu Vin (1984) include in addition to the type family, also the Erpodiaceae Broth., Helicopbyllaceae Broth., Microtheciellaceae Miller & Harrington, and Rhachitheciaceae Robinson. The cumnt circumscription of both the order and the family serves convenience more than phylogenetic rationale, as no single autapomorphy or combination of unique characters states can ci priori k found for defining either timon. While morphological heterogeneity can result hmsevere reductionary trends with loss of mical character states (peristome features in gymaostomous taxa) or from reversal to plesiomorphic Chapter two: 10 characters States within a monophyletic lineage. heterogeneity may also be indicative of the polyphyly of a taxon. Discriminating between these alternative sources of heterogeneity is not always obvious and îherefore the monophyly of a given taxon should idedy be investigated using phylogenetic methods without any O priori judgment In mosses, paraileihm in gametophytic feanires cm be severe even above the ordinal level (eg., Calontnion; Vitt 1995; Waters et ai. 1996) but for cladistic rationaie to solve such cases of ambiguous etieswould require the inclusion of distantly related putative sister-taxa In the case of the Orthotrichales, alternative affinities for some of the taxa may lie, as argued for below, with the Dicranales, Grimmiales, and Leucodontales, and maybe even the Seligeriales (see chapter 3). Including potential sister-groups ftom these orders. particularly if these are &rived within their lineage (and thus with few plesiotypic characters) may introduce more phylogenetic noise and obscure relationships, rather than solve them. Taxa whose affhities clearly are outside the ingroup should be transferred whenever possible. pnor to a cladistic analysis, as is common practice in taxonomic monographs where the phylogenetic reconstmction is preceded by the taxonomic treatment This study examines the systematic affinities of Meioweisiopsis, Trigonodic~on,and Microthecielkuceae, and the taxonomic status of Pleurozygodontopsis. glewweisiopssis Dixon, Smithsonian Misc. Coll. 72(3): 18.1920 Type: Kleioweisiopsis denticulata Dixon, Kenya, Aberciare Mts., Ailan 3951 (holotype: BM!; isotype: US). (Fig. 2.1)

In the protologue of Kleioweisiopsis DUon (1920) suggested placing his new genus in the Pottiaceae based on overall similarity with Astomwn and Hynicnostomum. Zander (1993) in his thorough monopph of the Pottiaceae excluded KIeioweisiopsis and tentatively msfemd the genus to the Onhotnchaceae, near Amphidim. KZeioweisiupsis denticdata, the sole species of the genus (Zander 1993), is characterized by its subulate perichaetial leaves, a very short seta seemingly caducous with the capsule, lack of stomata, and a differentiated, yet mostly persistent operculum (Fig. 2.1). At the base of the operculum the cells are weakly differentiated and may not be Miyfunctional. The exothecium is cornposed of ffagile cells that disintegrate ailowiug for the spores to be dispersed. The type specimen is annotated by Saito as a combination in Pleuridiwn Brid. in the Diuichaceae. KIeioweisiopsis exhibits a combination of feahues indeed charactenstic of the Ditrichaceae sensu Brothenis (1924; e.g., undifferentiated aiar cells, Chapter two: 11 costa with median guide cells and few ventral stereids, and mostly smooth laminal cells; Fig. 2.1) and should therefore k included in this family. Within the Ditrichaceae, imrnersed capsules are cornmon to 8 genera, confhed to the Ditrichoideae. This subfamily is divided in two groups based on the presence of a differentiated operculum (Brothems 1924, Buck and Snider 1992). Because the opetcuium is not My differentiated and fiiactionai, Meioweisiiopis canwt be assigned to either group of taxa unambiguously. The flmt grwp, lacking a differentiated opedum, includes four genera: Pleuridium, Cladostomum C. MU,Crumuscus Buck & Snider, and Sporledera Hampe. Pleuditun, Cladostomum, and Crrunuscscur merfrom KZeioweisiopsis by the presence of a weil developed cauline central strand, mitrate calyptrae (except P. papillosm Ma@; Buck and Snider 1992). Cnauscw can furthemore be distinguished based on the parriaily bistratose lamina (Buck and Snider 1992). Sporledera lacks a centrai strand in the stem, but ciiffers by the ovate to oblong leaf base, the presence of stomata, and the mamillate capsule. Compared to the taxa developing a dehiscent operculum on an irnmemd capsule, Meioweisiiopis mers hmPringieella Card. by the lack of an annulus, and cucullate calyptrae, fbm Garckecr C. MW.by the gymnostomous and spheric capsule, nom Astomiopsis C. Miill. by the flexuose leaves, the la& of an annulus, fkom Eccremidium Hook. f. & Wils. subg. Eccremidiwn by its flexuose leaves, and fkom subg. Pseudo-Pleuridium Broth. by its costate leaves. Kleioweisiopsis appears to share sufncient chanicten with the Ditrichaceae to be included in this family, where it should remain a distinct genus.

Plants gregarious, to 5-6 mm high. Stems orthotropic, outer cortical cells weakly differentiated, central strand lacking. Leaves 2.0-2.5 mm long and 0.25-0.35 mm wide at base, erect to flexuose, lanceolate, entire, denticdate to semilate toward apex. Costa ending within a few celis of apex, with median guide cells, and adaxial (sub-)stereids and seemingly a single abaxial stereid. Basal laminal ce& hyaline, nctanguîar, thin to moderately thick-wailed, 24-90 pm long, 8-24 pm wide. Upper laminal celis, short rectangular, quantate to oblate, thick-walied, 10-24 pm long, 8- 14 pm wide. Acrocarpous, autoicous. Perichaetia with paraphyses, archegonia 320 plong. Perichaetial leaves to 4.5 mm long, erect and soxnewhat fiexuo~,linear-lanceolate, serrulate near apex. Perigonia subapical. paraphyses present, antheridia to 140 pm long. Perigoniai leaves short, lanceolate. Serae 0.3 mm long, outer cells slightly thick-walled, inner cells thin-walled, mostly coliapsed, central strand differentiated. Capsule sphenc, Chapter two: 12 exothecial cells very thin, disintegrating in old capsules, stomata absent Operculum long, obliquely msttate, rosmim to 0.4 mm long. Calyptrae cuculiate, covering at least half the capsule, about 0.7 mm long. Spores 18-22 pm, granulose.

Key to the Mtrichaceae with immersed capsuies @itricboideae). 1. Operculom not differentiated ...... 2 1. Operculum dinerentiated ...... 6 2. Stomata present ...... -.-..3 2. Stomata absent ...... a 3. Cauline centrai strand present ...... Pleuridium 3. Cauline cenaal straiid absent ...... Sporledera 4. Lamina partiaiiy bistratose ...... ,...... Cmmwcus 4. Lamina unistratose ...... 5 5. Leaves ovate, calyptra mitrate, camecentral strand present ...... Cladostomum 5. Leaves lanceolate, calyptrae cucullate, cauline central strand absent .. .KZeioweisiopsis 6. Peristome present ...... Garckea 6. Peristome absent ...... *.....7 7. Caiyptrae mitrate ...... - 7. Calyptrae cucullate or covering only the operculum ...... 8 8. Annulus aot differentiated ...... Kleioweisiopsis 8. Annuius differentiated ...... , ...... ,...... 9 9. Operculum with long oblique rosmim ...... Atomio@ 9. Opeccuium mamillate, or short erect rostnim ...... Eccremidium

Zygodon Hook. & Tayl., Musc. Brit. 70. 18 18. PIeurozygodontopsis Dixon, Annales Bryologici 12: 5 1. 1939. syn. nov. Type: Pleurozygodontopsis decurrens Dhn Zygodon reinwIUdtÜ (Homsch.) Braun in B.S.G. Pleurozygodontopsis decurrens Dixon. Annales Bryologici 12: 5 1. 1939. syn. nov. Type: Sumatra, Gunong Losir, Atjeh. Gajolanden, 2940 m. 5 Feb. 1937, van Steenis 10159 Oectotype chosen here: BM-herb. Dixon 4044); ibidem, von Steenis 10161 (syntype: BM-herb. Dixon 404 1!). (Fig. 2.2)

Dixon (1939) proposed the new genus Pleurozygodontopsis to accommodate two collections similar in habit and foliation, to slender forms of Leptodontium (Pottiaceae) or Chapter two: 13

more mbust fomof Zygodon. Gametophytically these two genera are similar in habit but can be readily segregated by the costa anatomy (Zander & Vitt 1979). Dixon (1939) placed the new genus in the Orthotrichaceae, where "the position of the inflorescence, the decunent leaves. and the nmwly cyhdric. almost fiusiform, plicate capsule" would distinguish the genus from related genem Dixon interpreted the gametangia to be either basal or lateral on the autoicous individuais. yet on closer examioatiofi it is clear that the gametophyte produces temiinai gametangia and fesumes growth by subapical innovations. When a terminal perîchaetium is produced, a subapical perigonium cm be developed. Monoicy, decumnt Ieaves and cyiindricai capsules are characters hiown fiom the genus Zygodon (Maita 1926). When compared to miesof section Zygodon (i.e., species with papiilose upper laminal cells) P. decurrens appears identical to the pantropical Z reinwardtii (Homsch.) Braun in B.S.G., except for the organization of the gametangïa. Both spacies have broad leaves with undulate margins, apical marginal teeth, papiilose (sornetimes oniy weakiy so) laminal cells cove~gthe abaxial surface of the costa near the apex, and are monoicous (Fig. 22). Zygodon reinwardtii has been reported to be synoicous (Via 1993), "synoicous and synoicous or dioicous" (Fleischer 1904). or synoicous, but "pure male and femak idorescences present too" (Malta 1926). Both specimens of P. decurrens show no indication of any antheridia present in the penchaetia, but rather separate male and fernale gamctangia, on a single individual. In P. decurrens the costa ends below the apex. On a single stem, the apical ceiis of the leaves can be either papillose or more often smooth. The smooth ceils are often somewhat elongate and may give the appearance of the costa king excurrent. In Z reinwardtii the costa has been described as vanishing below the apex (Grout 1946, Vin 1993), and percuneat (Malta 1926). excurrent (Lewinsky 1990), or shortiy excurrent (Fleischer 1904). 1have examined specimeas of Z reinwardtii fiom various paris of the neotropics and one specimen fiom New Guinea, and found the papillosity of the apical cells to be variable even among leaves hma single individual. In some leaves, where the costa distinctly vanishes below the apex, the unistratose apex cm be composed distally by smooth and elongate ceils and proximally by papiiiose ceiis. When these subapical papillose celis are lacking the elongaie ceUs seem to represent the distai portion of the costa, yet because the apex aiways appears unistratose 1feel that the costa of Z reinwardtii cannot be described as excunent, and is therefor similar to that of P. decurrens. With regards to the papiilose celis on the abaxial surface of the apex of the costa of 2 reUnvordtii Witt 1993), 1found it to be variable too, ranghg from virnially smooth to conspicuously papillose, even on a single stem of Z reinwardtii. The costa in Chapter two: 14

Z reinwardtii and P. decurrens can be either smooth ail the way to the apex, or be covered ventraily in the apical region. None of the characters presented by Dixon (1939) to suppoa the distinction of P. decurrens, witùstand critical comparison with Z reinwardtii and the two species shodd therefore be considered synonymous.

Trigonodictyon Duon & P. de la Varde. Annales de Czyptogamie Exotique 1: 40.1928. Type: TBgonodictyon indicm Dixon & P. de la Varde (Fig. 2.3)

Dixon and P. de la Varde (in P. de la Varde 1928) characterized thei.new genus by the basal juxtacostal cells that form clear triangles iike the cancellinae in the Calymperaceae. Because the abundant matenal of type and only collection lacks sporophytes systematic placement remained tentative. They decided to tentatively uiclude Trigonodictyon in the Orthotrichaceae on the basis of the homogenous costa anatomy and the ventral position of the two guide cells. Trigonodictyon shms with some ûrthotrichaceae, narnely the Orthotnchoideae and the Zygodontoideae, terminal perichaetia (unlike the original authon 1kmd some perichaetia; holorype BM) and the ventral guide cells in the costa. Trigonoàictyon diffea, however, nom all Orthoaichaceae by the laminal cell architecture. The distal basal cells are distinctly sinuose to irregularly thick-walled (Fig. 2.3). a character distinctive for the Grimmiaceae. The Grirnmiaceae sensu Churchill (198 1) are composed of 11 genera distributed among three subfamilies. The genenc and suprageneric taxonomy is mostly based on sporophytic characters. The only gametophytic characters, besides the sporeling type, are the shape of the leaf, the differentiation of basal and upper laminal cells, and the anatomy of the costa. Rhabdogriltytliu and Guernbelia mer hmTrigonodicf~on by the long acuminate apex of the leaf. The shap of the leaf and the differentiation of the laminal ceiis in two types, suggest afnnities with the Corinodontoideae. The costa of the Grimmiales is characterized by a type-C architecture (Churchili 198l), that is the adaxial, median and abaxial cells are ail clearly distinct ftom one another (Kawai 1968). Traasfonnations of this typical costa anatomy in the Grimmiales lead either to a loss of complexity (homogenous costa: type-A; or a costa with a differentiation of the adaxial celis ody: type-B), or to an increased complexity, with the differentiation of guide cclls in the median portion of the costa and ventral stereid cells (type-E; Chunihiil 198 1). The dorsal or abaxiai costal cens of Trigonodictyon have walls that are not strongly thickened (Fig. 2.3), unlike true stereids that have very incmsate wails and very nanow lumens. In a tramverre section of the costa, the substereids of Trigonodictyon can hardly be Chapter two: 15

differentiated hmthe surroundhg ceus, and the costa is thus of the type-A. Kawai (1968) defined the various types of costal anatomy based on transverse section only, and may in many cases have reiied solely on observations by previous authors. Indeed, when examined in surface view, taxa considered with typ-A costai anatomy (Pleurorth~~chum;Kawai 1966) can be interpreted as having a type-B costa (see chapter 6). Examination of the costa of Tn'gonodiccryon in surface view reveals that the abaxial ceiis, except for the most proximal ones. are mostly short, and chlomphyllose, whereas the adaxial cells, are long, hyaline and with straight walls in the proximal portion of the costa, and chlorophyllose and with sinuose walls above. The median ceils can also be examined in surface view, and they are mostly dis~ctlynarrower than the adjacent ceiis, and hyaiine. Similar patterns, pazticularly the sinuose adaxial cells have been seen in various representatives of the Grimmiales, but never in the Onhotrichales. The above characters seem to justify transferring Trigonodictyon from the ûrthotrichales to the Grimmiaies, a hypothesis diat had alrrady been raised by Dixon and P. de la Varde (1928), who had been misled by the costa anatomy as seen in transverse section. Based on Gangulee (1972), Trigonodic~anwould key out as Racomitrium snicr~oiiurn(Min.) laeg., a species recently transferred to Grimmia by Deguchi (1980). Trigonodictyon differs from G. m*ctifoiiaby the unistratose margin, and the lack of a central strand in the stem. Until sporophytes are found, Trigonodictyon should be retained distinct within the Grimmiales, with uncertain systernatic affinities. Considering my different interpretation of the costa anatomy and the discovery of the fernale gametangia, I am here providiilg a revised and updated description based on the only collection known, the holotype.

Trigonodictyon indicum Dixon & P. de la Vasde (Fig. 2.3). Type: India, Pambar Tomnt, Kodaüranal, Pulney Hills, févr. 1927, R.P. Foreau (holotype, BM!; isotype. NY!). to 2.5 (-3.0) cm long. Stems prostrate (-decumbent ?), outer cortex composed of 2-3 layers of small, thick-walled, orange-red, celis, inner cortex of wide, moderately thick-wailed, centrai strand lacking. Branches few (?) and mostly basitonous. Leaves 1.3-1.6 mm long and 0.3-0.6 mm wide, erect appressed, lanceolate to ovate-lanceolate, apex acute, margin entire, plane. Costa strong, ending a few ceils below apex, with abaxial ceiis short rectangular ?O subquadrate chlorophyilose almost to the base, median substereids and adaxial cells hyaline and rectangular with straight walls in proximal portion of the costa, and chorophyliose and with sinuose waiis above. Basal laminal ceifs Chapter two: 16 rectanguiar, srnooh, evenly thick-walled, walls straight, 12-60 pm long, 8-12 pm wide. Median ceils, rectangular. smooth, with siouose thick walls. Upper laminal cells. short quadrate but mostly irregular, smooth, thick wailed 10-24 pm long, 8-14 wide. Maryhairs, with two quadrate, hyaline basal ceiis. Acrocarpous. dioicous. Penchaetia lacking paraphyses. archegonia 600 pm long, theto five per perichaetium. Penchaetid leaves to 2.2 mm long and 0.8 mm wide, erect, lanceolate to ovate- lanceolate. apex acute to acuminate, margin entire, refiexed in lower half. Perigonia and sporophyte unlmom.

Mkrothecicûà Dixon, J. Bot, Lond. 69: 1 & pl. 595, fig. 1, 1931. Type: Microtheciella kerrei Dix., "On branches of Rhabdia lycioides, on open grave1 bed in river, Ban Tmg. Langsuan, Siam, c. 50 m alt.; 17 Feb. 1927, leg. A.F. Kerr (195)" (holow-BM!). This monotypic genus is lmown only hmtwo collections, one nom Thailand and the second one fkom Laos (Dixon 193 1,1936). Dixon placed his new genus in the Erpodiaceae where it rernained until recently (Cm1972). Miller and Harrington (1977) considered Microtheciellà anomalous withia the Erpodiaceae fiom which it differed by its costate leaves. differentiated marginal cells, and specialized motor cells, and therefore proposed a new family, the Microtheciellaceae, to accommodate it (see Miller & Hacfington 1977, for excelient illustrations). Despite some parallels with the Neckeraceae or the Hookeriales, Miller & Harrington (1977) retained the Microthecieilaceae within the Orthotrichales. Its affinities within the order remained obscure Witt 1982). Examination of the type specimen (BM) revealed that the perichaetia in Microtheciella kerrei are produced on shon lateral innovations. Vegetative leaves similar to those found on sterile anis are not developed on these laterai innovations, instead only juvenile leaves an found below the perichaetim nie mode of pericheatim production is thus reminiscent of the pleurocarpous type but could also represent a "derived" type of cladocarpy with severe reduction of the sterile portion of the perichaetium-bearing axis, as in Molendou (Pottiaceae; La Farge-England 1996). Both types can be disthguished based on the morphology of the juvenile leaves and on the width of the axis~bearingthe female gametangia (La Farge-England 1996). In Microtheciella keerei the juveniles leaves below the pericheatium are broader than those developed on sterile branches, but the width of the innovation appears sirniiar to that of sterile axes. At present, and until more matenal becomes available 1prefer to consider Microtheciella keerei as Chapter two: 17 pleumcarpous, and shouici thenfore be excludcd hmthe ûrthotrichaceae. The lack of pseudoparapbyliia in Microtheciella kzerei (Miller & Harrington 1977) is wt incompatible with affbities with pleurocarpous taxa, since pseudoparaphyilia have ken lost repcatedly duRng the evolution of various pleumcarpous taxa (Buck and Vin 1986, Hede- 1995), including in the Neckeraceae (HomaIodendron Fleisch., Enroth 1993). Gametophytic features such as leaf shape. and cell areolation, are reminescent of the Neckeraceae (Miller & HacfiIIgton 1977). but the exostome in this family is never reduced (Enroth 1989) as it is in Microtheciek (Miiler & Hanington 1977). Hedeniis (1995) recentiy demonstrated that the gametophytic chatacters are ofien less informative in addressing higher level relationships ammg pleurocarpous mosses. The systematic affiaities of Microtheciella among pleurocarpous heages will need to be huther critically examined. For POW the family should be retained distinct and placed near the Neckeraceae. Chapter two: 18

Figure 2.1. Kleioweisiopsis &dmticuhtu (hoIotype-BM). a Calyptrae. - b. Upper lamiad ceiis. - c. Median laminal ceus. 6 Vegetative leaf. - e. Perichaetid le&. - f. Operculate capsule. - g. Basal laminal ceiis. - h. Apex of vegetative leaf. - 1. Transverse section of the costa in median pomon of the vegetative leaf. Scale bar = 0.2 mm (d e), 81 w(a,f, hl, [email protected], g, 0- Cbapter two: 19

Figure 2.2. Pleurozygodontopsis decurrens (hoIotype-BM). a Vegetative leaves. - b. Upper laminal celis. - c. Transverse section of upper lamina ceus. - d Basai laminal cells. - e. Transverse section of costa in median portion of vegetative Leaf. f. - Apex of vegetative leaves. - g. Transverse section of stem. - h. Inoer (left) and outer (right) perichaetial leaves. - 1. Stomata. - j. Capsule habit. - k. Transverse section of the seta. Scale bar = 057 mm (j), 0.2 mm (a, h), 81p(f), 20p@, c, d. e. g, i, k) Chapter two: 20

Figure 2.3. Trigonodictyon indicm (holotype-BM). a. Vegetative leaves. - b. Upper laminai cells. - c. Median laminal ceus. - CLBasal laminal cells. - e. Transverse section of the costa in median portion of vegetative leaf. - f. Perichaetiai leaves. - g. Transverse section of the stem. Scaie bar = 0.2 mm (a, f), 20p@, c, d, e, g) Chapter two: 2 1

fiterature cited Bmtherus, V. F. 1924. Musci (Laubmoose) m. UnterHasse Bryales: II Spezieller Teil. h A. Engler's and K. Rantl [&.], Die Natürlichen Pnamenfamilien, 2nd. edition, vol. 10, 143-478. W. Engelmann, Leipzig. Buck, W.R. and Vitt, D.H. 1986. Suggestions for a new classification of pleurocarpous mosses. Taron 35: 21-60. and Snider LA. 1992. Crt~~tlwcusvitah gen. et sp. nov. (Ditrichaceae). Conîn*bun'onsfrurnthe University of Michigan Herhriwn 18: 39-4 1. Churchill, S.P. 1981. A phylogenetic analysis, classification and synopsis of the genera of the Grimmiaceae (Musci). In V.A.Funck & D1. Bmoks (eds.), Advances in Cladistics. Roccedings of the htmeeting of the Willi Hennig Society. New Yoprk Botanical Garden, New York 127-144. Deguchi, H. 1980. Note sur quatre espèces ûimalayennes de la famille des Grimrniaceae (Musci). Hikobia 8: 259-268. Dixon, H.N. 1931. New genera of Asiatic mosses. Joumal of Botany, London 69: 1-7. 1936. On a collection of mosses hmLaos. Annales of Bryology 9: 61-72. 1939. High alpine mosses fiom Sumatra. Annales Bryologici 12: 48-56. Enroth, J. 1989. Bryophyte flora of the Huon Peninsula, Papua New Guinea. MNII. Neckeraceae (Musci). Acta Botanica Fennica 137: 4 1-80. 1993. Notes on the Neckeraceae (Musci). 18. Description of Curvicladium, a new genus from southern and southeastem Asia. Annales Botanici Fennici 30: 109- 117. Fleischer, M. 1904. Die Musci der Flora von Buitenzorg (zugleich Laubmossfïora von Java). Erster Band. Sphagnales; Bryaks (Arthrodontei [Haplolepideae]). Leiden, Bïill. 1-XXXI + 1- 1103. Gangulee, H.C. 1972. Mosses of Eastern India and adjacent regions. fasc. 3. Syrrhopodontales, Pottiales, & Grimmiales. 567-830. Calcutta. Grout, A.J. 1946. North Amencan Fiora, vol. 15 part 1. Bryales, Orthotrichaceae. -62 pp. New York. Hedenh, L. 1995. Higher taxonomie level relationships among diplolepideous pleunxrarpous mosses -a cladistic overview. Journal of Bryology L 8: 723-78 1. Kawai, 1. 1968. Taxonomie studies on the midrib in Musci (1). Significance of the midrib in Systematic Botany. nie Science Reports of Kanazawa University 13: 127- 157. Chapter two: 22

LaFarge-England, C. 1996. Growth fom, branching pattern, and perichaetial position in mosses: Cladocarpy and pleumcarpy redefind The Bryologist 99: 170-186. Lewiasky, J. 1976. On the systematic position of Amphidium Schimp. Lindbergia 3: 227-23 1. 1990. Zygodon Hook. & Tay1 in Australasia: a üuonomic revision including SEM-studies of puistomes. Lndbergia 15: 109-139. 1994. The genus Pleurorthotrichunt Broth. Lindbergiu 19: 11-24. ,and MRCrosby 1996. Orthominiwn tuberculatum (Orthotnchaceae), a new genus and species hmGuizhou, China. Novon 6: 1-5. Malta, N. 1926. Die Ga- ZyghHo& et Tayl.. Eine monographische Studie. Lumijas Universitates Botaniskzi D4M Darbi 1: 1- 184. Miller, &A., aud Hamhgton, A.J. 1977. Microthecieilaceae, fam. nov. Journal of Bryology 9: 5 19: 524. Norris, D.H., and Robinson, H. 198 1. Stoneobryum, a new genus of of Oahotrichaceae fiom South Africa and Southem Queensland. me Bryologist 84: 95-99. Potier de la Varde, R. 1928. Musci novi indici. Annales de Cryptogamie Exotique L : 39-47. Vit& D.H. 1976. A monograph of the genus Muelleriella Dusén. Journal of the Hanori Botanical krborutory 4& 9 1- 113. 1982. The genera of the Orthotrichaceae. Beiheft zur Nova Hedwigia 7 1 : 26 1- 268. 1984. Classification of the Bryopsida. In R M. Schuster [ed.], New Manual of Bryology. Vol. 2,696-759. Hattori Botmical Laboratory, Nichinan, Japan. 1993. ûrthotrichaceae. A. J. Sharp, & A. Cm,and P. M. Eckel [eds.], The Moss Fiora of Mexico, 590-657. New York Botanical Garden, New York, 1995. The genus Calodon (Bryopsida): Taxoaomy, phylogeay, and biogeography. The-Bryologist 98: 338-358. ,and H. P. Ramsay. 1985. The Macromitrium complex in Australasia (Orthotrichaceae: Bryopsida). Part 1. Taxonomy and phylogenetic relationships. Jouml ofthe Hattori Botanical Luboratory 59: 325-45 1. Waters, D-,B. Goffinet, DH.Via, and R. Chapman. 1996. The systematic afnnties of the Calomniaceae (Bryopsida) based on 18s nrDNA nucleotide sequences. American Journal of Botany (Supplement) 83: 27 Zander, R.H. 1993. Genera of the Pottiaceae: Mosses of harsh environmenm. Bullerin of the B@do Society of Naturul Sciences 32: 1-378. Chapter three: 23

The Rhachitheciaceae: revlsed generic dricpmserlptbn and ordinai aninities

The Rhachitheciaceae is a smaii fdyestablished by Robinson (1964) to accommodate two former orthotrichaceous genera: Rhachithecim Le lol. and IwauuLi & Noguchi. Rhachitheciopsis P. de la Varde, considered by Iwatsuki (1957) to k closely related to Hypnodonopsis, was reduced to a subgenus of Rhachithecim by Robinson (1964). Recently, Allen and Pursell(l99 1) ümsferred the genus Jonesiobry~iMen & Ruseil hmthe Funariaceae to the Rhachitheciaceae, and Zander (1993) excluded the genus Tisserantielk P. de la Varde hmthe Pottiaceae and suggested that it too, "clearly belongs to the Rhachitheciaceae". AU 13 species, except Rhachitheciuin perpusillunt sensu Crum (1956), are narrowly distributed in tropical and subtropical regions of both hemisphens, whem they grow on trees. The Rhachitheciaceae have been characterized by spathulate to lingulate, ofien apiculate vegetative leaves, differentiation of laminal celis into hyaüae rectangular basal cells and chiorophyllose, isodiametric upper cells, a single costa extending to the upper portion of the leaf, the lack of a differentiated cauline epidermis, terminal perichaetia on the stem, a simple peristome composed of 16 teeth fused into eight pairs (Cm1993; Zander 1993). As part of a generic revision of the Chthotrichaceae, the genera Octogonella Dixon and Uleastrum Buck were studied. Both genera share several sporophytic feanires?that are a priori incompati%le with typical Orthotrichaceae, but are instead reminiscent of the Rhachitheciaceae. The circumscription of this family is here revised and its systematic affinities to the Orthotrichaceae are discussed.

1. RljlA~HtTHEcIOpsfsP. de la Varde, Bull. Soc. bot, France 73: 74. 1926.

Potier de la Varde (1926) proposed this new genus to accommodate three Rhachithecium-Wre specimens hmAfiica, that diffeted nom the symmetric R. perpusillum (as R nanswaliense C. MU.)among others by the seta that is completely curved downward when young or moist and strongly curved into an "s"-like shape when dry, and poorly differentiated perichaetial leaves (Potier de la Varde 1926). Iwatsuki (1957) retained Rhachirheciopsis as distinct and suggested close afhities with Chapter three: 14

Hypnodontopsis based on the mitrate caiypttae. Potier de la Varde (1926) tentatively descni the calyptra of RhachitJzeciopsU as mitrate, and Iwatsuki (1957) interpreted this feature as an indication of close affinities between Hypnodontopsis and Rhachitheciopsis. 1have seen several calyptrae in the type specimen. and aithougb not fully developed, they appear clearly cucuiiate. Further, the genus Hypnodontopsis now includes a second species lK me;xr'canu (Thér.) Robinson, thin mers hmthe type species, by several charoctus, including the cucuilate calyptme (Robinson 1964). The shape of the caiyptrae thus fails to be informative in addnssing the generic relationship of Rhuchitheciopsis. Iwatsuki (1957) Mecemphasizeâ that in both Rhachitheciopsis and Hypnodontopsis the anatomy of the seta, immediately below the capsule is strongly asymmetric, due to heavy excennic waU thickening on one side of the seta (Iwaeuki 1957; Fig. 3.li). 1have examined the anatomy of the seta immediately below the uni in Rhachiaheciwn perpusillum (Chang Jinkun 1748-3-ALTA) and R. popillosum (holotype-NY) and in both cases the thickenings an *O unilateraily heavier, suggesting that this character may be shared by taxa with strongly curved seta (see dso under Uleasrnun). The inner surface of the peristome teeth of R. trjserantii lacks the palisade-like markings typical of Hypnodontopsis (Figs. 3.2a & b and 3.3e & f). in their architecture and omamentation the peristome teeth of Rhochitheciopsis (Figs. 3.2a & b) are very similar to those of Rhnchithecim (Fig. 3.k & d). Scattered additional divisions in the outer layer resultiog in three or four cells (Fig. 3.2a) per tooth pair have also ken observed in species of Rhachithecium. as weli as in other taxa such as Glyphominium (Brothenis 1924), and do not seem to be taxonomicaiiy informative. Potier de la Varde (1926) described the teeth as muiutely papiilose. SEM observation, however, reveals that the teeth are smooth on the outer surface and at most miautely comgate on the inner surface, as they are in Rllochithecim (Fig. 3.3~& d). Iwatsuki (1957) desfribed the costa of Hypnodontupsis apiculatus as "in midleaf with two laycrs of stereid cells and 1 layer of large, thin-waiied ventral cells, at the basal portion of the leaf with ~WOlayen of thick-walled celis, ..." I have seen only one specimen of H. apiculatus (Musci Japonici Exsiccati 1425-ALTA). Here the costa defnitely bas adaxial substereids, as in H. rnexzèana (Amène 4793-NY; Robinson 1964). The costa of Rhaichitheciopsis lacks these ventral stereids and is more reminiscent of the costa of Rhachithecim (see below). The capsule of Rhachitheciopsis and Rlvrchithecim is globose to cup-shaped whereas in Hypnodontopsis the capsule is more elongate and cyhdrical. This suite of evidence thus supports Robinson's (1964) hypothesis that Rhachitheciopsis is more closely related to Rhachithecium than to Hypnodontopsis. Chapter the: 25

Robinson (1964) considered Rhuchitheciopsis and Rhrichithecium indistinguishable at the genenc level, and rejected a distinction based on the shapc of the ope~ulumand the differentiation of the perichaetial ieaves, two cbaracters he considered "uausually subject to variation in this group". As a dthe transferred Rhachitheciopsis tisserantii to Rhachithecium, where he placed it in its own subgenus distinct fmm subgenus Rhachithecium by the smooth caiyptra, the "not noticeably differentiated perichaetial leaves and the nearly flat operculum. 1have seen three specirnens of Rhachitrhcciopssis and in aii the, the opeda were slightly convex and smooth (i-e., without a rostnun or a mammillae), and the perichaetial leaves were not sheauiing the seta despite king as long as the seta. Furthemore the calyptra is definitely smooth in Rhachitheciopsis compared with a scabrous surface in Rhuchithecim (Wiiams 19 14; Cm1956; IwaauLi 1957). This study also revealed additional characters that separate R tisserantii nom species of Rhclchitheciwn sensu stncto. The costa in Rhachithecium is weak, with only two bands of substereids, while in Rhachitheciopsis tisseruntii the costa is strong due to the presence of several layers of substereid bands (Fig. 3. ld; see also Robinson 1964; Iwatnilri 1957). The annulus of Rhachiiheciopsis is composed of two layea (and not one as descriid in the protologue; Fig. 3.lg). whereas in Rhachithecim the annulus is simple (Wiiams 1914; Iwatniki 1957: Cnun 1993; but see also under R. welwitschii) except maybe for R braziliense (Brotherus 1924; synonymized with R. perpusillum by Crum 1956). Finally, the spore surface in R tisserantii is conspicuously striate (Fig. 3.2c), while that in R perpusillm and R. papilloswn appears pitted (Fig. 3.2d), perhaps resulting nom the proliferation of striation rather than fiom the formation of actual ph. Considering the above differences in both the gametophyte (differentiation of perichaetial leaves, costa anatomy. surface of calyptra, spore wail omamentation) and the sporophyte (shape of the operculum, complexity of the annulus; Tab. 3.1) 1feel that R tisserantii is best retained in its own genus. In the light of additionai characters discovered since its original description 1am here presentiag a more complete description of the species.

Rhachithecbpsis &semntil P. de la Varde. Buil. Soc. Bot. France 73: 74. 1926.(Figs. 3.1-2) Type: Central Afncan Republic: Plateau à 10 km d'Ippy, sur branches d'Albiuia brownii, 6.7.1929., Tisserant 133 (Lectotype here âesignated-PC!, herb. P. de Varde 2047). Rhacluhecium tisserantii (P. de la Varde) Robins. The Bryologist 67: 449. 1964. Chapter the: 26

Plants to 3 mm high, green, oahotropic, acrocarpous. Stem sympodially braaching, composed of wide lumened, thick-wded celis throughout, epidernial cells not dinerentiated, central strand lacking. Leaves spnading when moist, erect to spreading when dry, lingulate-spathulate, acute, to 1.4 mm long, 0.5 mm wide, margins plane, entire. Costa strong, ending 1/10 to 1/4 below apex, in transverse section with large, thin-waJied, hyaline laminal cells, covering several abadlayers of substerrids. Basal ceb differentiated in lower third or haif of the leaf, hyaline. subquadrate to mostly long recmgular, 26-64 pin long, 1626 pm wide, moderately thick-waiied, smooth. Upper cells green, short, subquantate to isodiametric, hexagonal, 10-40 pm long, 10-24 pm wide, rather thin-wded, flat, smooth. Gemmae not seen (but see Iwatsuki 1957). Autoicous, perichaetia terminai, paraphyses present Perichaetial Ieaves weakly differentiated, Iiagulate to lanceoiate, with incurved margins near apex. apiculate, with costa filhg most of the apiculus, to 1.7 mm long. Basal celis rhomboidelongate, smooth, hyaline. Upper cells differentiated oniy near apex, isodiametric, smooth,green. Perigonia axillary, gemmiform, paraphyses absent. Perigonial leaves dinerentiated, shortly ovate-oblong, strongly concave, to OSmm long, costa lacking or weak. Seta twisted to right, curved downward when moist or capsule young, erect and kinked upward whea dry, 1.5 mm long, composed of a cylinder of large, wide lumened thick- wailed cells, smunded by small thick-wailed epidermal ceUs arranged in a single row except below capsule, epidermal layers asymmetric, biseriate on one side and uniseriate on the opposite side; central strand present. Capsules 0.8 mm long, 0.7 mm wide, eight- nbbed, broadly cupshaped, exothecial cells 2: 1, those forming ribs, yeiiowish, with very thick longitudinal anticihal walls, stomata phanemporic, few, resoicted to neck. Operculum slightiy convex, lacking rosmor mamilla Peristome single, of 8-fused teeth; teeth lanceolate, 160 pm long, to 110 pm wide near base, incurved when moist, ncurved when dry, yeilowish, smwth on both sides, outer layer of a single tooth uniseriate dotsaily, with some additional divisions occurring irrguiarly, trabeculae swng and projecting; base of tooth covered with thin, smooth biseriate membrane; inner surface of twth smooth, composed of half a ceil column. Calyptrae smooth, cuculiate. Annulus bistratose, with both rows uniseriate, the outer row to f 30 pm high, inwr row It 50 pm high, caducous to persistent. Spore 20-28 pm in diameter, papilîose. Specirnens emmined: Central Afncan Republic: 6 km S d'Ippy, sur arbre dans les rochers du K. Lekpowa (?), 26 Fév. 1927. Tisserant 247 (PC); 60 km N Bambari, rochers près Wamice, sur vieux bois, 27.7.1927, Tisserant 392 (PC). Chapter three: 27

2. RaAc&ITBECIUM Le Job, MhSoc. Sc. Nat Math. Cherbourg 29: 305. 1895. a RIioch~eciumweIwitsckii (Duby) Zander, Bull. Buffalo Soc. Nat. Sci. 32: 277. 1993. Zygodon wehvirschii Duby, M6m. soc. Phys. Hist Nat. Gen6ve 21: 44,1871. Type: "ad basin de Mom de Zopollo 5100 alt. in prov. Huilla regni Angol. det. cd. Welwitsch." (holotype-G!). Lnea welwitschii (Duby) Broth. in Par., In. Bryol. ed 55:96. 1906.

UZea welwitschii Mers hmother species of mec1 (now Uleastnmt Buck) by the strongly ribbed capsule, the curved seta (except. CI. octoblepharis (Jaeg.) Zander) and was therefore recently transfeçred to the genus Rhachithecium (Zander 1993). Zander (label annotation) hesitated to synonymize this species with the pantropical R. perpusillum hmwhich he considered it "doubdully different"". The holotype consists only of a few, but complete (except for the calyptrae) individuals embedded between mica slides. The pexistome teeth are identicai to that of Rhachitheciopsis or Rhachitheciwn: the teeth are fused into eight pairs, and the dorsal surface of each tooth pair is biseriate and bears swng trabeculae, whereas the huer surface is uniseriate and smooth. Duby (1871) described the peristome as made of 16-32 teeth, and his drawing suggests that the paired teeth are split nearly to the base. The type daes, however, clearly show 16 teeth fwd into 8 pairs, split only in the upper portion, as in aii peristomate species of Rhuchithecim (see below). Duby also described the perichaetial leaves as "angustelanceolatis enervis celiulis peilucidis elongato-parailelogrammicis", and thus differentiated nom the vegetative leaves, and indicative of afnnities with Rhachitheciunr The operculum too, is reminiscent of that in Rhachithecium: it is distinctly conic and shortiy rostrate. The annulus was descrikd and illustrated by Duby as biseriate, suggesting a relationship with Rhachitheciopsis. In Rhachitheciwn the annuius is simple (Williams 1914; Iwatzuki 1957; Cnun 1993) except maybe for R. braifiense (Broth.) Broth. 1have not seen the type of the latter, but Brothems' (1924) ülustrations clearly show a compound annuius. Cnun (1956) synonymized R. bradliense with the widespread R. perpusillum, a species he illustrates has having a single annulus (Cm1956, 1993). One of the four slides of the type of R. welwirschii inchdes portions of an annulus, detached hmthe capsule. The celis are elongate and not infîated, as in Rhuchithecium perpusiiIum (Cm1993). The hgments of the annuius are clearly unistratose. In Rhachitheciopsis the annulus detaches nom the capsule rim and both layers of the annulus separate when pressure is applied on the slide. Whether the annulus of Chapter three: 28

Rhachithecium welwitschii is acWybistratose as desfnbed by Duby, but has fragmented into separate layers during the preparation of the siide canmt be determuied. The slide with the annuius holds ail other portions of the deoperculate capsule, and one can assume that aii portions of the annulus are present Together aU the hgments of the annuius only maice up once the length of the cîrcdezence of the capsule mouth, suggesting that the mulus is composed of a single row of cells, and not two. The other capsules of the type are old and have lost their anaulus. iWuzchithecium wehvitschii thus has ciear aSities with Rhachitkcium as suggested by Zander (1993). Rhacirithecim welwitschii is currently hown fkom only the type collected in Angola, whereas Rhachithecium perpusiZlum is reporteci for most of western Afiica (O'Shea 1995). 1agree with Zander (in lin) that R welwitschii is almost identical to R petpuillm but unless the Mcanmaterial of R pepusi1lu.m is revised, the clifferences in the shape of the annular cek should be taken as supporthg the taxoaomic distinction between both species.

b. ~RitheciumpupiUosum (Williams) Wijk & Marg., Taxon 9: 52. 1960.(Figs. 3.3~- d, 4) Octogomdla scabrjcofia Dixon, J. Botany 74: 1 (+ plate 610). 1936, syn. nov. Type: "On uee, Kasadi. near Simla, India, 2929 (?); cou. P. Tiwary; comm. R.P. Foreau (1Ib)" (holotype-BM!)

Among the specirnens of Indian mosses distributed by R. P. Foreau, Dixon found a collection of a few individuais "clearly allied to Rhochithecium but well marked in the coarsely tubedate areolation"and the somewhat different peristome. When Dixon (1936) placed this ttew species in its own genus, Octogodïiz, Rhachithecim included only R. perptcsillum, a species with smooth upper leaf ceUs and 16 peristome teeth fused into 8 pairs split near the apex. Robinson (1964) faüed to include Octogonella in the Rhachitheciaceae and the genus remained in the Orthotrichaceae (Chopra 1975; Cm 1987). As part of a generic revision of the Orthotrichaceae, I had the opportunity to examine the type of the only species of the genus, Octogonella scabriJolia. This species presents characters shdbetween the Rbachitheciaceae and two subfamilies of the ûrthotrichaceae, namely the Orthotrichoideae and Zygodontoideae (e-g., the unicostate leaves, dimorphic proximal and distal celis, papiiiose upper leaf cells. and terminal pezichaetia on stem) but its single peristome of smooth teeth with strong dorsal trabeculae Chapter three: 29

is known, among onhotrichalean generis, only hmthe Rhachitheciaceae, namely Rhrrchitheciopsis and Rhochithecim. The strongiy differentiated penchaetial leaves of 0.sc&#olia suggest a relationship with Rhuchithecim. The upper leaf ceiis of Rhchithecim can k either smooth (Rpe'pusillum, and R welwitschii), mamillos (R nipponicm) or unipapillose (Rpapîllosum). The strong papiilae of upper leaf ceils of O. scabri$olia are sMar in shape and sïze to those of R papillosm (w-iarns) Wijk & Marg. (Fig. 3.4; Williams 19 14 and Iwatsuki & Sharp 1976 for illustrations of R papillosm). Dixon (1936) also noticed that the peristome teeth in OctogoneUa are not split near the apex, and used this character to mersupport bis new genus. Rhochithecim pefp~csilllwnand R welwitschii have 16 peristome teeth that an fwdin pairs except at tips (Cnm 1956). but in R papilloszun the teeth appear to be either split (WiUiams 19 14) or entire (Iwatsuki & Sharp 1976). The degree of fusion of the teeth thus fails to support both generic and specific distinction. Dixon (1936) also reported that "the ribs of the capsule are irregularly and minutely, but often quite markedly comgate or papillose". yet, based on the only capsule 1 saw, the ribs appear to be smooth. Octogonella scabrifolia is identical in gametophytic as well as sporophytic characters with R papillosum. The latter was for long known only fiom the type locality @ the Philippines until Iwatsuki and Sharp (1976) reported it nom eastem India The presence of this species is northem India (type of Octogoneila) is thus not surprishg even though it represents a noticeable range extension. This species is epiphytic and in the collection nom northern India, the few stems of R. papilloswn are found intermked with a species of Sernatophylium.

3. ULEXSTRUMBuck, Candollea 40: 203. 1985. The genus UIeastrum Buck (=Ulea C. MUI.) was recently transfened by Zander (1993) fkom the Pottiaceae to the Orthotnchaceae. Chen (1941) had recommended excluding LIleastmm (then Lnea C. MW.)hm the Poniaceae on the basis of its epiphytic habitat and peristome architecture, and Mersuggested placing it near Rhachithecim (then in the ûrthotrichaceac). Gametophytic characters, sucb as leaf shape, la& of differentiated cauline epidermis, and differentiated penchaetial leaves point indeed toward a relationship with the Rhachitheciaceae, rather than the Orthotrichaceae, and 1therefore propose removing Uteastnmi fiom the ûrthotnchaceae and include it in the Rhachitheciaceae. Robinson (1964) refrained nom hcluding Uleastnm (then Uleu) in the Rhachitheciaceae, because of the smooth capsule and the Chapter three: 30

straight seta. With the inclusion in the Rbachitheciaceae of Jmesiobryum and Tisserantiella (Ailen & Purseil 1991; Zander 1993). two genera with exclusively smooth capsules. ribbed ums no longer define the family. Furthemore. variation of tbis feattire occurs withïn Cneustnun: U. octoblepharis bas a curved seta, whereas the other species have a straight seta. UIeastrton is c~entlycomposed of four viesdistributeci in two groups (Brothenis 1924): one including Uleastnmr nitidm (Th&.) Zander, CI. palmicola (C. Mü11.) Zander, and U*puraguense (Besch.) Buck, and the other, restricted to W. octoblepihare (Jaeg) Zander. Brothems (1924) separated U. octoblephare on the bais of its dioicy, smooth laminal cells, obtuse leaves. curved seta, smooth calyptrae. undinerentiated annulus, smaller spores. split peristome teeth, and temcolous habitat. In addition, U. octoblephre mers from congedc species by the presence of costal stereids, reddish-brown, isodiametric exotùecial ceils with strongly ihickened distal periclinal walls, heavier thickening on the IPL than on the PPL, and 2:2 peristome formula (Tab. 3.2). AU of these dinerencesjus* establishg a new genus for CI. octoblepluir. Since the generic type of Uleastrum is U. paraguense (Buck 199 1). 1 propose the new genus 2anderia to accommodate U. octoblephare.

4. ZANDEU Goffinet gen. nov. Type: 2anden'a octoblephoris (Spruce ex. kg)Goaiaet comb. wv. Spruceella C. MW., Gen. Musc. Fr. 396, 1900, hom. illeg. non Pierre, 1890. Type: Spnrceellla octoblepharis CC.MU. Planta terricola, dioica, 3.0 mm alta. Folia iinguiata ad spathulata. Ceiiulae basilares subquadratae ad rectangulae lucidae, superio~ssubquadratae, laevis. Capsulae exsertae. cylindricae. laeviae. Penstomium simplen Cellulae exothecii rubrae, quadratae ad hexagonae. tenuis. Sporae 68jm latae. grandosa.

Zanderirr octoblepitaris (Jaeg.) &&et comb. nov. (Figs. 3.5-6) Ponia octoblepharis Jatg.. Ber. S. Gall. Nam.Ges. 18714872: 343, 1873. Ulea octoblephans (Jaeg) C. MU.. Hedwigia 37: 234, 1898. Type: "Amerika austral. in terra arenosa rubra ad flora Amazonum p. Santarern in locis umbrosis. Spruee 163" (holotype-NY !; isotypes-NY!). Uleostncm octoblephre (Jaeg.) Zaader, Bd.Buffalo. Soc. Nat. Sci. 32: 277, 1993.

Plants to 3 mm high, green, orthotropic, acrocarpous. Stem sympodially braoched, composed of wide lumeaed, thick-walied cells throughout, epidennal cells not Chapter three: 3 1 differentiated, cenûal strand lacking. Leaves spreading when moist, erect to spreadùig, crisped when dry, linguiste-spathdate, acute to obtuse, 0.5-1.5 mm long, 0.2-0.6 mm wide, with largest leaves rosette-iike at apex, srnalier leaves below, margins incurved clasping in bottom third, plane above, entire. Costa ending 1/10 to 114 below apex, in transverse section with adaxial guide celis, median stereids, and abaxial substereids. Basal ceils differentiated in iower third or hdf. hyaline, subquadrate to mostly long rectmguiar, 24-60 jm long, 15-28 pm wide. with thick-wails. smooth. Upper cells green, short, subqiradrate to isodiametric, hexagonal, b16 p.m long. 644 pm wide, thick- wailed, flat, srnoah. Dioicous. pichaetia terminal, paraphyses present, pcrichaetial leaves spathdate, to 1.5 mm long and 0.6 mm wide, enct crisped when dry and erect spreading when moist, the inner ones lanceolate, erect sheaihiog, to 0.7 mm long, celi architecture as in vegetative leaves. Perigonia not seen. Seta not twisted, curved downward when dry and erect when moist, 1.5 mm long, composed of a cylinder of wide lumened cells with evenly thickened walls, outer cells with eccentric thickening, lumen very nmw, confined in transverse section to proximal end of ce& central strand present. Capsules 0.8 mm long, 0.3 mm wide, smooth, cylhdric; exothecial cells quadrate to hexagonal, with thin anticlinai walls and very thick outer periclinal wall; stomata superficial. few, restricted to the base of um, neck not dinerentiated. Operculum conical, long and obliquely rostrate. Peristome single, of 16, lanceolate endostome teeth fused into eight pairs, split at apex, erect when moist, hcurved when dry, yeilowish, 160-180 plong, to 5060 pm wide near base, thickening heavier on the PLthan on the PPL, with arachnoid omamentation on both surface, smooth and thin dong median vertical walls on either side, IPL thickenîng fiiaher thiiining in center of tooth, giving fenestrate appearmce. Calyptrae smooth, cucullate. Dinerentiated annuius lacking. Spore 6-8 p in diameter, finely grandose.

The new genus is named in hoaor of Dr. Richard Zander for suggesting the genenc distinction of UIeastmm octoblephons ((Zander 1993).

The closest systematic affinities of &deria are not clear. Considering the number of differences with Uleastrum, the sister group of Zanderia may need to be sought for elsewhere in the Rhacùitheciaceae. The subquadrate exotheciai cells are known besides in Zonden'u only from Tisseruntiella pulchella (nier. & Hilp.) Zander, which also lacks a differentiated annulus. This Afncan species, however cliffers by its papillose laminal celis, a rather globose to cup-shaped uni, pale exothecial cells, and a gymnostomous Chapter three: 32 capsule (hoIotype-PC!). A weak costa similar to that of Zanden'a is also present in Rhachithecim, which also includes a species Mth smooth laminal ceils. The latter Mers however by the shape of the um, of the exothecial cells, and peristome architecture. The excentnc thickening of the outer ceh of the seta separates Zanderia from Uleastwn and is furiher found in Hypnodntopsis, &chitheciopsis, Rhnchithecim (see above). Fraûm and Frey (1987) studied the twisting mechaaism in the cygneous seta of the genus Campylopus, and concludeci that the rotation of the seta upon moistening or during drying, was due to the presence of thme distinct layers in the cell wd. The curved seta of ande~may similady k conrlated to the presence of epidermal cells with asymmetric thickening in the distal portion of the seta, considering that species of Uleasm~i,that aU have straight setae, have outer ceils with evenly thickened wds. Whether the differential thickening of epidenrial cells below the um, common to the Rhachitheciaceae with cwed seta, iadicates coamon aocestry remaius speculative. Tùe genera of the Rhachitheciaceae express a wide range of gametophytic and spotophytic character transformations, and even within the genus Tisserantiella sensu Zmder (1993), the divergence between species is such that the monophyly of the genus could be senously questioned. Differentiating between patristic and homoplastic distances separating the genera, is fiirthermore bampered by the uncertainty regardhg the sister-group of the Rhachitheciaceae. Except for R. perpusillum, ail species currently known ftom the Rhachitheciaceae are narrow endemics. with Little or no overlap in their ranges. The transatlantic (Jonesiobryum), transpacinc (Hypnodontopsis), or pantropical (Rhachitheciwn) distribution of the genera, conelated with high phenetic distances between hem, may be a strong indication of an ancient origin and early diversification of the famiy.

5. JONESIOBRWMBizot & P&s ex Allen & bu,The Bryologist 94: 441 (1991) Jonesiobry~lsphaerocarpum Buot ex Allen & Purseil The genus Jonesiobry~lcomprises three species, of which J. sphuerocarp~lis the only paleotropical species. The genus has recently been transferred from the Funariaceae to the Rhachitheciaceae (Men & PurseU 1991). It is easily distinguished hmrelated genera by immersed capsules. While studying specimens of Rhuchitheciopsis tisserantii, 1came across a collection with imwrsed capsules and scarcely differentiated penchaetid leaves, characters reminiscent of the genus Jonesiobryum. Based on Vital (1983) and Menand Rirsell(l991), the specimen keys to J. sphaerocarpum. This species was previously thought to be restricted to a single locality in Nigeria, West Afnca The Chapter three: 33 present specimen was collected early this cenniry £hmthe Central Afiican Republic, and represents only the second collection of this species. This major westward range extension across the continent may suggest that the species is actually more cornmon across the West African plains. The habitat of this second collection agrees weil with that of the type (Bizet et al. 1974): both specimens were collected on trees in savanna woodlands. SpecVnen examine& Central Mcan Republic, "darbre, savanne boisée, environs des Sabangas, 80 km N Bambari, 20 juin 1927. Tissemnt 334 (PC-herb. Potier de la Varde).

Artifîcial key to the seven genera of the Rbachitheciaceae. 1. Capsule immersed, seta shorter than urn ...... Jonesiobryum 1. Capsule emergent or exserted seta longer than uni ...... 2 2. Peristome lacking ...... Tisserantiella 2. Peristome present ...... -3 3. Penstome with dorsal trabecuiae ...... -4 3. Penstome without trabeculae ...... S 4. Stem with centrai strand, penchaetid leaves conspicuously differentiated, sheathing, operculum short-rostrate, spores pitted ...... Rhachithecium 4. Stem lacking central strand, perichaetiai Ieaves undifferentiated, spreading, operculurn flat, spores striate ...... Rhachitheciopsis 5. Longitudinal thickenings present on outer surface of peristome teeth .. Hypnodontopsis 5. Longitudinal thickening lacking, outer surface smooth ...... 6 6. Lamind cells smooth, dioicous, run reddish-brown, and seta curved ...... Zànderïa 6. Laminal cells papdose, autoicous, uni pale, seta erect ...... Uleastnm

Systematic affinities d the Rhacbitheciaceae. The Rhachitheciaceae have ken placed histoncally in or war the Orthoaichaceae (Brothem 1924; Robinson 1971; Crosby 1980; Vitt 1982, 1984; Walther 1983; Cm 1993). The peristome of the Rhachitheciaceae differs fiom the typical orthotrichaceous peristome (Shaw 1985; Lewinsky 1989) in that it is at the most, composed of a single row of teeth. Except in anderia, the inner layer of cells contributhg to the teeth is composed of only eight cells, while 16 cells compose the outer layer. This 2: 1 peristome is dinicult to reconcile with a 4:2 exostome architecture found in al1 arthrodontous mosses (Edwards 1984; Shaw et 1989a; Schwartz 1994) includhg those species that lack an exostome at Chapter three: 34 maturity (Edwacds 1979; Shaw et ai. 1989b). Altematively, if the single peristome of the Rhacbitheciaceae is an endostome, the outer iayer composed of 16 ceils wodd be consistent with the pattern observed for the PPL (Edwards 1979,1984). In Hypnodontopsis mexicana, Rhachitheciopsis tissermtii, as weH as Uleastrum palmicola, the basal portion of the outer surfaces of the teeth are covered by a smooth membrane (Figs. 3.2a; 3.3a c) that is hardly visible under the light mimscope. This outer membrane is composed of 4 ceils per tooth pair. If the teeth are derived from the PLand PPL, then the formula becornes 42: 1, and the outer b'xnemb~e"represents a nduced OPL. Diplolepideous peristomes reduced to the endostome are hown hmother mosses (e.g., Dichelodontim nindum, Ma@ 1987). Wbether the endostome of the Rhachitheciaceae represents a reduced diplolepideous or haplolepideous peristome cannot be determined based on the peristome formula, because the 4:2: 1 cell pattern would be atypical for either the diplolepideous or the haplolepideous peristome. When Edwards (1979) critically examined cell patterns of haplolepideous peristome teeth, he found that "a aumber of haplolepideous species ..... possess a distinctive type of double peristome." The Seligeria-type peristome as he called it, is characterized by an endostome that has strong dorsal trabeculae, and an exostome that "adheres to the margins of these as a thin membrane." This endostome architecture is very sirnilar to that found in Rhachitheciopsis and Rhachithecium, except that in the Seligeria-type peristome the PLis composed of 24 cells. instead of 8. This haplolepideous peristome type is further defined by little or no thickening on the ventral layer, a situation also present in Rhachitheciopsis, Rhuchithecim, and Uleasttlon, where the horizontal divisions of the PLappear very fault in light rnicroscopy. Except for lacking the reduced OPL, the peristome architecture of andena (is.. a x:2:2 pattem) is simiiar to that of Ventunella (Erpodiaceae), a genus Edwards (1979) tentatively also linked to the group defined by the Seligeria-type peristome. This peristome type is fiirtber found in Glyphonzitrium (Ptychomitnaceae), a genus (particularly the Asian species G. humillimum (Min) Card., and G. minutissimum (Okarn,) Broth.) with considerable resemblance to the Rhachitheciaceae (Robinson 1964). Robinson (1964) refrained to include Glyphomitnum in the Rhachitheciaceae because it lacks the bbdistin~tiveseta or plications in the capsule", an argument that no longer holds considering the cumnt circumscription of the family (see above). Churchill (198 1) excluded Glyphiniwn from the Grimmiaceae (inchhg the Ptychomitriaceae), but was unable to ascertain an alternative systematic position. Whether the Seligena-type-penstome defines a naturd group or not remaias to be addressed. Peristome features alone may be misleadhg in addressing familiai Chapter three: 35 relationships (Buck 1991). but considering that the Rhachitheciaceae and the Seligeriaceae occupy very different habitats (&ces versis rocks) it is uniikeiy that different selection pressures associated with these habitats would have dtedin nearly identical peristom architectures. The peristomt of the Rhachitheciceae, which is atypical arnong diplolepideous moses. is very similar to that of certain haplolepideous taxa may be seen as a sûong indication of the haplolepideous naaire of the Rhacbïtheciaceae. Tbis hypotbesis is cumntly king tesied using molecuiar data- Chapter three: 36

Figure 3. la-j. Rhachitheciopsis risserantii. a Vegetative Leaves. - b. Upper laminal cells. - c. Basal laminal cells. - d, Transverse section of costa at midleaf. -e. Transverse section of stem. - f. Caiyptrae. - g. Annuius. - h. Perichaetial laves. - i. Transverse section of seta immediately below the m.- j. Capsule. habit when dry (Tisserant 247- PC). Scale bar = 0.2 mm (a, f, h), 20p@, c, ci, e, g, i) and 70p.m (j). (lectotype-PC unless otherwise noted) Figurc 3.2. a-c. Morpholog>'ol'RJ~(~~-lri~l~~,~iopsis fisxermitii and Rlm-liirlirc-irrirc pt.rprisillrotr a-c. Rlicrt-l~irkuc-iopsisrissrrntrrii ( lectoty pe-PC ) . - a. Outcr surfilcc OS tooth pair. - b. Inner surface of rwth pair. - c. Spore. - d. Rliïrcblii~lieciiirirpurprisill~irtt ( C hmg Ji nkun 1748-3 -A LTA ) ;spn. Chüptcr thrcc: 38

Figure 3.3. a-f. Morpholog). olperistomc teeth in the Rhüchitheçiaceiie. a-b. Ulem-rrrrrtr pdiricoln ( Vitr 2 1 161-ALTA). - a. Outcr surface of tooth pair. - b. Inncr surtkc of iooth pair. c-d. Rlm~hitlrrcirrrtip

Figure 3 A. Rhuchithecium papiüusum (OctogonelZu scabnfolia: holotype-B M). - a. Vegetative leaf. - b. Upper laminai cells. - c. Basal laminai ceiis. - d. Longitudinal view of upper laminal cells. e. Innemixt perichaetial leaf. - f. Perigonial Ieaf. - g. Transverse section through stem. Scale bar = 0-2 mm (a, e, f), and 20p(b, c, d, g). Chapter three: 40

Figure 3.5. Zartde~~octobiepharis(holotpye-NY). a. Vegetative leaves (left to nght: fiom base to apex of stem). -b. Transverse section of the leaf. - c. Upper laminal ceiis, - d. Blisal laminal cells. - e. Transverse section of the stem- - f. Exothecial cells. - g. Calyptrae. - h. Transverse section of exothecial celis (outer surface-right). - i. Perichaetial laves. - j. Transverse section of seta. - h. Capsule. habit. Scale bar = 0.2 mm (a, g, i, k), and 20p(b, c, d, e, f, h, j.). Chaptcr thrcc: 4 1

figure 3.6. a-d. Zlrplr

Table 3.1. Morphological characteristics separating the genera Rhpchitheciopsis and Rhachithecium.

Cauline central strand absent present Costal abaxial ceiis to 5 layers 2 layers Peric haetial leaves erect-spreading erect-sheathing Calyptrae smwth papillose Annulus biseriate unisenate ' Spores striate pitted ODercuium fiat short rostrate ------p. 1Except for R. brazilienre wbkh is iiiustrafed by Brothems (1925, p. 16) with a double Chapter three: 43

Table 3.2. Morphological characteristics separatiag the genera Zanderia and Uleastmm.

Upper laminal celis smwth papillose Guide cells adaxial median Costal stereid median not differentiated Sexual condition dioicous autoicous Exotheciai ceils: shape imgular-isodiametric rectangular anticlinal walls thin-waüed thick-walled outer penchai waU eveniy thickened coiIenchymatous colour red-brown paie brown Peristome 2:2 2: 1 teeth splitnng at apex teeth fused to apex 'Yenestrate" due to thinner areas on IPL thickening continuous IPL>PPL IPLCPPL Spore cl0 pm >15p Seta curved erect Epidermal cells of seta thickening excentric even d around Annulus absent present Habitat terricolous epiphytic Chapter three: 44

Literature Cited Men B. & -11 R A. 1991. A reconsideration of the systematic position of Jonesiobryum. Ine Bryologist 94: 438-442. Bizot M., R B. Pierro, and T. P&s . 1974. Trois genres nouveaux de Muscinées. Revue BryologQue et Lichénobgique 40: 25-3 1. Brothem, V. F. 1901-09. Musci (Laubrnoose). ID. Unterkiasse Bryaies: II Spezieller Teil. In A. Engier's and K. Prantl [eds.], Die Natiiriichen Wauzenfamilien 1 (3,I, II), 277-1246. W. Engelmann, Leipzig. 1925. Musci (Laubrnoose) m. UnterHasse Bryales: II Spezieller Teii. In A. Engler's and K. Rantl [eds.], Die Naaalichen Pflanzenfamilien, 2nd. edition, vol. 11, 1-542. W. Engelmana, Leipzig. Buck, W.R. 1991. The basis for familial classification of pleumcarpous mosses. Advurtces in Bryology 4: 169-185. Chen P.-C. 1941. Studien über die ostasiatischen Arten der Pottiaceae. 1. Hedwigia 80: 1-76. Chopra R. S. 1975. The taxowmy of Indian mosses (An Introduction). Botanicd Monograph 10: 1-63 1. Ciosby, M. R. 1980. The diversity and relationships of mosses. In R. J. Taylor and A. E. Leviton [eds.], The Mosses of North Amcrica, 1L5- 129. California Academy of Science, San Francisco, California Cm&A. 1956. Notes on Hypnodon, a genus of Orthotrichaceae new to North America. nie Bryologist 59: 2634. 1972. A taxonomie account of the Erpodiaceae. Nova Hedwigia 23: 201-224. 1993. Rhachitbeciaceae. In A.J. Sharp, H.A. Cm,and P.M. Eckel [eds.] The MOSSFlora of Mexico. Memoirs of the New York Botanicol Garden 69: 588-590. Churchill. S. P. 198 1. A phylogenetic analysis, classification and synopsis of the genera of the GNnmiaceae (Musci). In: V. A. Funk and D.R. and Brooks [eds.], Advances in Cladistics. New York Botanical Garden, New York. 127-144. Dixon, H. N. 1932. Classification of mosas. In F. Verdoorn [ed.], Manual of Bryology, 397-412. Martius Nijhoff, The Hague. . 1936. Degas genenun novonmi musconm. Journal of Botany 74: 1-10. Duby, J.-E. 187 1. Choix de cryptogames exotiques nouvelles ou peu connues. 1. Mousses (4&* suite). Mémoires de in Société de Physique et d'Histoire Naturelle de Genève 2 1: 425-444 + Tab 1-V. Chapter three: 45

Edwards, S. R. 1979. Taxonomie implications of celi patterns in haplolepideous moss peristomes. In G. S. C. Clarke and J. O. Duckett [eh.], Bryophyte Systematics, 3 17- 346. Academic Press, London. 1984. Homologies and inter-relations of moss peristomes. In R. M. Schuster (ed.], New Manuai of Bryology, vol. 2,658695. Haîton Botaaical Laboratory, Nichinan. Frahm. M.,and W. Frey. 1987. The twist mechanism in the cygneous seta of the genus Campylopur. Morphology, stmcture and huiftion. Nova Hedwigia 44: 29 1-304. Iwatsuki, 2. 1957. The genus Hypnodon and its allies. nie Bryologist 60: 299-3 10. ,Sharp AJ. 1976. Notes on Rhachithecim in Assam, Iudia. Miscellania Bryologica et Lichenologica 7: 105- 107. Lewinsky, J. 1989. Does the OrthomChum type of peristome exist? A study of peristome evolution in the genus Orthotn'chum Hedw. with possible derivation of the haplolepideous peristome. Journal of the Hatton' Botanical Loboratory 67: 335-363. Magill, RE. 1987. On the endostomid nature of the Dichelodontium (Ptychomniaceae) peristome. Memoirs of the New York Botanical Garden 45: 87-94. 0'shea, B. Checklist of the mosses of sub-Saharan Africa Tropical Bryology 10: 9 1- 198. Potier de la Varde, R. 1926. Rhachitheciopis P. de la V., genre nouveau d'Orthotrichacées de rAfrique tropicale. Bulletin de la Société Botanique de France 73: 74-76, + plate 1. . 1941. Tisserentella P. de la V.genus novunt f

, .and B. D. Mishler. 1989a Peristome development in mosses in relation to systematics and evolution. III. Fullork hygrometn'ca, Bryum pseudocapillare. and B. bicolor. Systematic Bot

CardotieIh elhk&z ('I'hér.) GoWt coino. nov., and the signiticance of dody adseriate endostome segments in the Ortûotrichaceae1

The flora of Madagascar and the Mascarenes includes nearly 1100 moss taxa (Kis 1985) and includes many endemics. Among orthotrichaceous gemra, three are particulady well represented: Cardotiefla Vitt, Macromiîrinan Bride. and Schlotheimia Brid. While the latter two genera have diverufied throughout the tmpics, only one species of CardotieUa is known outside south-eastern Afkica and East Afncan Islands. The predorninantly neotropical genus Groutiella Steere is represented on Madagascar by two species - G. laxo-torqmta (Besch.) Wijk & Marg. and G. elùnbata (Thér.) O'Shea (Crosby et al. 1983). The latter specîes was initiaily descnid in 1929 by Thénot as Micromitrim elimhhun and subsequentiy traasferred to Groutiella (a new name for the iiiegitimate homonym Mi~romi~mSchimp.) by OShea (1995) as previously suggested by Crosby et al. (1983). Examination of the holotype, the only known collection of Micromitrizun elimbarunt. indicates that this species differs nom congeneric taxa not only by the la& of a weli developed border in the lower portion of the leaf, but also by the presence of conspicuous decurrencies of large. hyaline cells, small upper leaf celis, and non-plicate cdyptrae (Fig. 4.le, h, & 1). Ail of these characters suggest an afnnity with species of Cardotiella, a genus to which M. elimbatum Thér. is therefore transfemd.

CARDOTIELLAELCMBATA on&.) Goffinet, comb. nov. (Figs. 4.1-3) Microminium elimbah~nTh& Recueil. Publ. Soc. Havraise Études Diverses. 1929-1930: 113. 1929. Type: "Madagascar. Province de Farafmgana. Ifandana, sur souche d'arbre, en forêt. leg. R. Decary," 8.9.1926 (holotype, PC!; isotype. S). Groutiella elimbata (Thér.) O'Shea, Trop. Bryol. 10: 95. 1995.

Plants reddish brown, plagioaopic, cladocarpous. Stem monopodiaily branching, pentagonal in transverse section. composed of a cylinder of wide. thick-wded cells surrouaded by a bi- to tnstratose epidermis of narrow thick-walled ceils, central smd lacking. Rhizoids orangeied, smooth, abundantly branched. commoaly originating from the abaxial surface of costa of both stem and branch Ieaves. Stem leaves ovate-

' A modified version of this chapter has ken accepted for publication in The Bryologist 99 (4). 1996. Chapter four: 48

laaceolate, widest just above insertion, 1.û&1.45 mm long, 0.40-0.65 mm wide, flexuose when dry, wi&ly spreading when moist, decurreat, weaidy apiculate, costa with an adaxial row of guide cek and several abaxial rows of stereids, prominent, orange-&, ending below apex, margin recurved in lower haif, plane above. Upper and median laminal celis subquadrate, pentagonal, or slightly elongate. srnooh except for scatterrd unipapillose juxtacostai ce&, thick-walIed, in irreguiar rows. Basal laminal cells isodiameûic to elongate, smooth except for a few marginal unipapiiiose ceils above decumncy, thick-walled. Decunencies composed of up to six rows of hyaüne, moderately thick-walled cells, with outer ones strongly bulging. Branch leaves ovate- lanceolate, widest above insertion, 1.30- 1.75 mm long, 0.40-0-50 mm wide, flexuose- contorted when dry. widely spreading when moist, decmnt, weakly apiculate, costa with adaxial row of guide cells and several abaxial rows of stereids, prominent, reddish, ending well below apex, margin namwly recurved at least in lower half, plane above. Upper and median laminal celis subquadrate, pentagonai, or slightly elongate thmughout except for Ioblate ceiis near base, 5-10 pm in diameter, srnooth except for some, mostly adaxiaily, unipapiilose cells particuiarly dong costa, evenly thick-walled, mostly in irregular rows. Upper marginal ceils forming discontinuous and indistinct border of elongate cells twice as long as wide. Basal lamiod cells rectanguiar to elongate, 10-45 pm long, 6-10 .pmwide, smooth except for a few marginal unipapillose cells above decurrency, thick-walled. Decunencies composed of up to six rows of hyaliw, quadrate. rectangular, or elongate, moderately thick-walled cells, outer ones strongly bulging. Axillary hairs two pet leaf axil, to 170 pm long, basal ceils quadrate. brown, smooth; upper 2-5 cells rectangular, hyaiine, smooth to vemcose. Dioicous? Perichaetia terminal on enct, lateral branches, penchaetid leaves narrowly lanceolate, 1A2.O mm long, 0.30-0.35 mm wide, flexuose from erect base. Seta to 11 mm long, erect, slightly twisted to the left, composed of narrow epidermal ceils grading into wider central celis, walls yeiiow, thick throughout except near center, thin-walled cells often collapsed; differentiation of centrai straad seemingiy absent at the level of vaginuia. Vaginula naked or with few paraphyses and archegonia in upper half, composed of bistratose layer of nmw, thick-walied epidermal cells surroundhg a multistratose layer of thick-walled wide lumened cells. Oprculum conic. long rostrate, to 1.3 mm long. Capsule erect, eight-ribbed, yellow to browa; urn ovoid, to 2 mm long; neck well dinerentiated, to 1 mm long. Exothecial ceils thick-wded more or less rectangular, to 3 times longer than wide except for oblate cells forming rim and elongate celis dong the edges of ribs. Stomata superficial, few, resnicted to neck. Peristome double (4:2:4); exostome Chapter four: 49

consistiirg of 8, long, lanceolate teeth splitting near apex, often truncate, yellow, to 400 pn long, strongly ncwed when dry, erect when mois&OPL thick, densely papillose proximally and striate distally, PPL thin, horizontaliy striate-papillose proximally and verticdy striate-papillose distaily; endostome of 16, slender, linear. hyaline segments, these to 400 plong and 40 pwide near base, erect, inflexed, PLhork~ntally striate on the contiguous basai ceiis, verticaily striate above, PPL thin, smooth thmughout, waU remnants shortiy protniding or embedckd in the PPL thickening. Spores isomorphic, 14- 18 pn in diameter, irreguiarly-striate papillose. Caiyptra smooth, loM, to 2.8 mm long, naked or sparsely hairy; hairs erect, uni- or biseriate, smooth, composed of rectangular, thick-walled ceUs.

The genus Cardotiella now includes six species of which four are endemic to the East Afncan islands (Crosby et al. 1983; Vitt 1981). Cardotiella is weii marked hmother orthotrichaceous gencra by the leaf decutrencies composed of "intlated, bulging, tuberculate, hyaline cells" aad the lobate, non-plicate calyptrae (Vitt 1981) with the latter characters suggesthg a close relationship to Schlothehia (Vin et al. 1995). Additionally, all species have papillose upper leaf ceUs except for C. renuuiàii (Broth.) Vitt, a nanow endemic known fkom oniy the type specimen collected on Madagascar. Overall, Cardoriella elimbata is similar to C. renuuldii, particulariy in the predominantly smooth laminal cells, the indistinct border in the upper part of the leaf, and the flexuose rather than secund branch laves, but differs in at least two gametophytic characters - scattered papiliae on both surfaces of the upper lamina, and naked vaginulae. In C. renauldii, the laminal cells are smwth throughout the leaf (except for the basal marginal cells) whiie scattered cells with a single papiilae. mostly on their adaxial surface oniy, are consistently found almg the costa and acmss the lamina in the upper portion of the leaf in C. elimbata (Fig. 4.k, 0. Cardoriella renauldii merdaers from C. elimbata by the abundant multiseriate hairs on the vapinuia. The type collection of C. elimbuta is abundantly fhiting and the vaginulae are naked, except for the presence of a few paraphyses and archegonia in the upper haif. Cardotiella elirnbata also fesembles C. subappendiculatu, which can however, be distinguished by its unifody unipapiflose cells, plane matgin, larger spores. and distinctly biseriate dorsal surface of the endostome (see below). When Vitt (198 1) proposed the genus Cardotiella. the sporophyte was hownfor only a single species, C. suboppendiufata (Broth.) Vin. In tbis species, the peristome is composed of eight teeth and eight segments. Recently, Van Rooy and Van Wyk (1992) Chapter four: 50 described the peristorne for another species, C. secd(C. MU.) Vin, and contirmed the peristomiai architecture for the genus, except that in that species the endostome is composed of 8 to 16 segments. In both species the segments alternate with the teeth and possess on their PPL surface the median vertical waii characteristic of diplolepidous altemate penstomes. In C. elimbutu, the presence of the median vertical wall as well as the horizontal walls cm be demonstrated only with the use of the SEM. The PPL layer is thinner than the IPL and the anticlinal waii nmnants of the PPL hardly emerge above the generai surface of the PPL and are even most often not pmtniding at all (Figs. 4.2e, 3% b). Segments with a smooth PPL surface lacking anticIina1 wall remnants bave also been reported in Florschuetziella Vitt (Vitt 1979, 198 1) and Pleurorthotrichwn Broth. (LewinsLy 1994; Shaw 1986). In Florschuettiella scabe~m(Broth.) Vitt, the pnsence of "partial outer vertical and horizontal waiis" on the lowermost portions oniy, of the outer suffixe of the som segments suggests "a reduction in the outer layer" in the median and upper portions of the segments (Vin 198 1). In F. steerei Vitt, light microscopical observation suggests the complete absence of anticbal wail remnants on the dorsal surface of the segments, while SEM reveals the presence of a median vertical line and altemating horizontal lines (Fig. 4.3~).These lines are not apparent throughout the dorsal surface of the segment. Wds associated with these lines hardly protmde above the surface; the lack of liaes in some portions of the segments is interpreted as a strong resorption of the wall remnants and their embeddhg in the PPL thickening. A lateral view of the segments clearly reveds the presence of a PPL thickeaing, that is much thinner than that of the IPL (Fig. 4.M). Shaw (1986) interpreted the endostomial architecture in Pleurorthotrihm chilense Broth. as possibly resdting from "a reduced number of divisions in the primary peristornial layer, or a displacement of the anticlinal wdthat fomthe vertical linet'. The first explmation would nquire the loss of every PPL division since none of the 16 segments has a median vertical wall on the dorsal surface. nit loss of anticlinal divisions matching the plane of division of the OPL would have to be compensated by the presence of other divisions. Examination of the ventral surface of the exostome teeth of the holotype (H-Broth!) reveais the presence of 16 divisions in PPL. The plane of each division is aiigned with the vertical plane of the division of the OPL that separates two consecutive teeth. For the PPL divisions to be displaced, yet still be alined with the marginal vertical waîis of each tooth, would require the breakdown between two teeth to be displaced by one cell. While these modifications would result in the lack of a median vertical waii on the dorsal surface of the segments, Chapter four: 5 1 ihey would also lead to an opposite placement of the sesegent, which is cleatly not the case. My own observations, based on a weii developed pristome on half a capsule stu attached to the gamtophyte, suggest tbat a vertical wall is present on the PPL surface of the segment. The walis, however, appear as faint lines uoder light microscopy and the use of the SEM reveals that they do not, or only scarcely, extend above the thin PPL thickening of the dorsal surface of the segment (Figs. 4.3e, f). In orthotrichaceous taxa witb weii developed endostomes, the endostomid PPL thickening is typicaiiy thinner than the IPL one. A strong resorption of anticlinal wali remnants to or klow the surface of the PPL couid resdt in the wds being undetected in iight microscopy. but not in SEM. In C. elimbata, F. steerei, and P. chilense the dorsal surface of the segments appears uniseriate unless they are examined in SEM, 1 therefore conclude that endostome segments in C. elimbata, F. steerei? and P. chilense are characterized by a strong resorption of anticlinal PPL wds, leaving these remuants to siightiy emerge nom the PPL thickening or even be completdy immersed in it* The endostome of C. elimbanun has some occasional abetfations in that certain segments are forked hma uniseriate base (Y-shaped) or fused and ending in a unisenate portion (inverted Y-shape). These observations do not sam to have any taxonomie or developmental significance. The vertical anticlinal divisions on the IPL surface of the basal membrane are somewhat displaceci and are not aligned with the PPL or the OPL planes of division. The segments are supported by either a single celi or are raised? straddling above two celis of the basal membrane. Apparentiy the cells between the segments are somewhat narrower han ceils supporthg the segments. Such asymmetry is well documented for several species of Onhonichum (Lewinsky 1993) and is found in other Orthotrichaceous genera as weii. The spores of C. elimhta are similar to those of C. secdaand C. subappendiculata: they are isomorphic, and irrrgularly-striate papillose (Fige4.2d). They are smaller than those of C. subappendiculata (14-18 pm vs 20-25 pm. Vitt 198 1) and within the range of those of C. sec&, the spores of which are about 15-23 pin diameter (Van Rwy 1990). Chapter four: 52

Figure 4.1. Curdotiella elimbata (holotype-PC). -a Stem leaves. -b. Branch leaves. -c. Partial transverse section of branch le&- d. Basal ceils of branch leaf. - e. Decurzent celis of branch leaf. -f. Longitudinal view of upper jmta costal cells of branch le& -g. Upper marginal celis of branch leaf. -h. Median upper laminal cells of branch leaf. -i. Partial transverse section of stem. -j. Stomata -. k. Partial tramverse section of seta. -1. Caiyptra. (scaie for a & b: -2 mm;for c-k: 20 pm; for 1: .3 -1. Figure 4.2. Cirr

Figure 4.3. Architecture or endostome segments in Cmcscrmcnt. PPL on ii~o. Chapter four. 55

Literature cited Crosby M. R., U. Schultze-Motel, & W. Schuitze-Motel. 1983. Katalog der Laubmoose von Madagascar und den umliegenden Inseln. Willdenowia 13: 187-255. Een, G. 1978. Mosses hmthe Mascatenes - 2. Mer@ 4: 2 19-223. Es, G. 1985. Mosses of South-east tropical Mc& An annotated list with dism'butioa data Institute of Ecology and Botany of the Hungarian Academy of Sciences. Budapest. Lewinsky, J. 1993. Monographie saidies on Onhotnihum (Musci). Bryobrothera 2: 1- 59 Lewinsky-Haapasaari. l. 1994. The gcnus Pleurorthotnkhum Broth. Lindbergia 19: 11- 24. OIShea, B. 1995. Checklist of the mosses of sub-Saharan Afkica Tropical Bryology 10: 9 1-198. Shaw, J. 1986. Peristome structure in the Oahotrichaceae. Journal of the Hutton' Botanical Lizboratory 60: 1 19-136. Thénot, M. 1. 1929. Septihe contribution à la flore bryologique de Madagascar. Recueil des Publications de la Société Hmtraise dgtudes Diverses 1929-1930: 99-122. Van Rooy, J. 1990. A taxonomie revision of the Macromitrioideae (Orthotrichaceae: Musci) in Southern Afnca. M Sc. thesis. University of Pretoria, South Anica. & A. E. Van Wyk. 1992. A conspectus of the subfamily Macromitrioideae Qryopsida: Onhotrichaceae) in Southern Africa. me Bryologist 95: 205-2 15. Vin, D. H. 1979. New taxa and new combinations in the Orthotrichaceae of Mexico. The Bryologist 82: 1- 19. 198 1. The genera LeiomitnWumand Cardotiellu gen. nova (ûrthotrichaceae). Jouml of the Hattori Botanicai kboratory 49: 93-1 13. ,T. Koponen, & D. H. Noms. 1995. Bryophyte flora of the Huon Peninsula, Papua New Guiwa LV. Demotheca, Groutiella, Mucrocoma and MacromitrÏum(Orthotrichaceae, Musci). Acta Botanica Fennica 154: 1-94. Chapter five: 56

Chapter five

In arihrodontous mosses, the peristome teeth that line the mouth of the capsule and regulate spore dispersai, arr made of ceii wall remnaats. Three concentric layers, namely the outer (OPL), the primary (PPL) and the inner (IPL) peristomial layer contribute to the teeth (Blomquist and Robertson, 1941). The number and tâe pattern of cell divisions occurring in these layers, paaicularly in the PL,are central to the classincation of artbrodontous mosses. Vitt (1984) recognized five types of articulate peristomes; four of these are diplolepideous and one is haplolepideous. The latter (or dicranaceous penstome type) is chanrterized by 1) a single row of teeth made from the proximal walls of the PPL and the distal walls of the PL,and 2) by a smngiy asymmetric '"nrrt-late" (sensu Schwartz, 1994) division in the PL leading to a (4):2:3 cell pattern per eight of the peristome circumference (0PL:PPL:IPL; Edwards, 1979; Shaw, Mishler, and Anderson, 1989). The typical diplolepideous peristome consists of an additional row of teeth composed of proximal OPL walis and distal PPL walls. The flanged (or heterolepideous) peristorne, pady composed of whole cells, and of additional intermediate exostome teeth (Edwards, 1984), characterizes the Encalyptineae, and is considered of diplolepideous affinities by Via (19û4). Tii the Funana-type peristome aii divisions are aligned and the peristome fodais 4:2:4 (Shaw, Anderson, and Mishler, 1989; Schwartz, 1994). The bryaceous type is characterized by additional divisions in the IPL leading to the development of cilia in between the segments, the latter of which lie altemate to the exostome teeth (Shaw,Anderson, and Mishler. 1989). In the orthomchaceous peristome, the thickening on the outer surface (OPL)of the exostome teeth is much heavier than on the inner surface (PPL), resuliing in the teeth being recurved when dry. The divisions are typically aligned but the segments altemate with the teeth; the peristome formula is 4:2:4 (Lewinsky, 1989; Vitt, 1981a). Crosby (1980) and Vin (198 la) were the ktto suggest that the haplolepideous peristome derived from a diplolepideous ancestor, but they disagreed with respect to the identity of this primitive diplolepideous peristome. Crosby (1980) and later Shaw and Rohrer (1984) suggested that the ancestral peristome was of the Bryum-type. The Chapter five: 57 orthotrichaceous peristome, was consequentiy derived through reductioa, a hypothesis aiso supported by Hedeniis (1994). Vitt (1984) argwd that "the evolution of the peristome has not ban uni-directional" and that the bryaceous pristome "is a totally separate evolutionary line than eitkr the orthotrichaceous or haplolepideous divergences" ftom an ancesttal fiinariaceas type (see also Vitt 19%la). More recently, Le* (1989) argwd in favor of a distinct orthotrichaceous peristome type that is "rnost 1üEely to have evolved from a peristom with a formula of 4:2:4 with complete digrment of the celis in the PL" - thus the hpriaceous type - but ualike Vitt (1984) she considers the orthotrichaceous peristome as possibly pivotal to the evolution of the dicranaceous and the bryaceous types. Vin, Goffinet, and Hedderson (1997) recently argued that morphological chanrters. used to define these major peristome types, are not phylogenetically informative, and tbat phylogenetic rezoastruction of the Bryopsida at the ordinal (or subordinal) level wül rely on independent sources of data such as ontogeny and gene sequences. The development of the fuaariaceous, bryaceous, and dicranaceous peristome types have ken studied recently (Shaw, Mishler. and Anderson, 1989; Shaw, Anderson, and Mishler, 1989; Schwartz 1994), but unless similar data are available for the orthotnchaceous peristome, polarizing the developmental pathways, and thus definhg homologies and monophyletic groups based on peristome architecture remains impossible. The Orthottichales include in addition to the Orthotrichaceae sensu lato, the Erpodiaceae Broth., Helicophyllaceae Broth., and Rhachitheciaceae Robinson. The Microtheciellaceae Harrington & MiUer were recently excluded by Goffi.net (cbapter two) on the basis of the pleurocarpous distribution of the fernale gametangia, and placed in the Isobryales. The Rhachitheciaceae are reminiscent in their gametophytic characten of the Zygodontoideae (ûrthotrichaceae). The peristome is always single and except for Uleustrwn octoblephuris, each tooth pair is biseriate dorsally and uniseriate ventraiiy; the peristome formula is thus 2:l (chaptcr the). The Erpodiaceae have traditionaiiy been placed in the Orthotrichdes (Fleischer; 1920; Brothems, 1925; Dixon, 1932; Reimea, 1954; Robinson, 197 1; Crosby, 1980; Vitt 1982a, 1984). Among erpodiaceous taxa, only species Venturiella and Wildiu are peristomate. As in U. octublephrire of the Rhachitheciaceae, the teeth are uniseriate on both surfaces (Cm1972; Edwards, 1979), and the peristome architecture follows a 2:2 pattern (chapter three). These peristome formulas would be atypical for either an exostome or an endostome of any major pristome type, and are therefore uninformative in determinhg the ordinal mtiesof these taxa. Edwards (1979) studied the peristome of Venturiella and reported a Chapter five: 58

"rudimentary unthickened basai exostome" reminiscent of that found in the Seligeriaceae. SimiIar thickenings were &O obsemd in the Rhachitheciaceae (chapter three). Walther (1983) consequently tra~sfemdthe Erpodiaceae to the Haplolepideae. The aflinities HelicophyII~(~forquatum. the sole species of the HeiicophylIaceae, remain dubious too, due to the absence of a peristome and the unique gametophyte with reduced ventral and dorsal leaves, and lateral leaves that are inrolled when dry. Buck and Vitt (1986) argwd in favor of a close relationship with the Racopilaceae (ciliate diplolepideae), as suggested by earlier authors (Bmthems, 1909, 1915; Reischer, 1920; Dixon, 1932; Reimers. 1954). The perichaetia are, however, produced terminally on the main axis (De Luna, 1995). a condition that is a priori incompatible with a placement in a pleumcarpous lineage (La Farge-Engiancl, 1996). Walther (1983) had by far the broadcst concept for the Orthotichales, including in addition to the Orthotrichaceae, Hedwigiaceae, and Rhachitheciaceae, the Cryphaceae, and the monotypic Wardiaceae, two families traditionaily considered pleurocarpous (Vin 1984; but see also La Farge-England, 1996). The circumscription of the ûrthotrichales, based on morphological characters, may thus range from a single to seven families. The Orthotrichaceae sensu Vitt (1984; and not sensu Churchill and Linares 1995) is amo~gthe most diverse families of mosses including over 50species. distributed among 22 genera (see chapter two, and three), with Orthotrichum, and Macromitnùm accoun~gfor more than two thirds of the spccies richness (Vitt, 1982b). The family is cosmopolitan in distribution and is particularly prevalent in tropical montane forests. The Orthotrichaceae are traditionally characterized by "small, papillose upper leaf cells; no differentiated aiar ceiis; large, usudly mitriform calyptrae; terminal perichaetia with additional gmwth by lateral innovations; and sporophyte with a diplolepideous peristome having the foiiowing feanins - an endostome with segments which altemate with exostome teeth; lack of basal membrane and with segments, which are not or are rarely keeled; and an exostome which has a thick. outer layer and a thin, inner layei' (Vitt, 1982b; but see also Via 198 lb). Investigations into the evolutionary history of the fdy have been hampered by the lack of monographie revisions for large genera (except Onhotnchtim Lewinsky 1993) and unstable familial circumscription. The controversy focused panicularly on the baplolepideous versus the diplolepideous affiriity of gymnostomous taxa (Amphidim: Brothenis, 1909 versus 1925; Vit& 1973 versus Lewinsky, 1976), or of taxa whose peristome is reduced (Dnrmmondia: Shaw, 1985 versus Vitt, 1972). The cmnt infrafamilial classification (Vitt, 1982b) reflects the four possible combinations of two characters - position of the female gametangia and "shape" Chapter five: 59

of the calyptrae - and deviates litüe hmBrothem' concept early this cenhuy (Brothems, 1925). In Vitt' s (1982b) phylogenetic arrangement, the Zygodontoideae (Schimp.) Broth. (acrocarpous and cuculiate calyptrae) are basal, followed by the Dnimmonàioideae Vitt (clad~~arpousand cucullate calyptrae) wbich are sister to a clade composed of the Oahonichoideae Broth. and the Macromitnoideae Bmth. (acrocarpous and cladocarpous respectively, and both with mitrate calyptrae). This phylogenetic arrangement is based upon a Funario-type ancestor îhat is acmcarpous and has a cucullate calyptra. Zygodon sect. Bryoides Malta mershares smooth laminal ceils with the Funariaceae, and may thus represent the most primitive mon in the family. The monotypic Desmothecoideae accepted by Bmthents (1925 as Pseudo-Macromitrioideae), Walthers (1983) and Cnun (1987), wen synonyrnized with the Macromitrioideae by Vin (1990) who felt that Demothecu may be patristicaiiy distantiy related to Macromihiwn but because it shares "several basic features that characterize the MacromiÛioideae....the genus is best kept in this subfamily". The monophyly of the ûrthotrichaceae and the Orthotrichales has recently been questioned by De Luna (1995) while examining the systematic affinities of the Hedwigiaceae. Based on a phylogenetic analysis using morphological characters, De Luna (1995) concluded that the Onhotrichales merely represent an evolutionary grade, suggesting that a clade of predominantly cladowpous ûrthotnchales (Macromitrioideae, Dnunmondioideae, Erpodiaceae, Microthecieilaceae. and Helicophyliaceae) was sister to the pleumcarpous mosses, and separated from the more primitive acrocarpous taxa of the order (acrocarpous ûrthotrichaceae and Rhachitheciaceae) by the Hedwigiaceae. Gffinities behwen the Hedwigiaceae and the ûrthotrichaceae have been proposed by other authors (Walther, 1983; Frey et al., 1995). Because of the gymnostomous nature of the Hedwigiaceae and a unique combination of gametophytic characters (acmarpy, lack of costa, and presence of pseudoparaphyllia), the atfiities of Hedwigia are unlikely to be resolved with traditional morphological characters. The central concept in the Hcnnigian phylogenetic approach is the use of appropriate outgroups, that dows adequate polarkation of cbaracter state transfomation in the ingroup. A reconstruction of the evolutionary history of the ûrthotrichaceae based on morphology has ken hampered by the lack of a peristome in certain taxa, the absence of gametophytic characters that are informative at higher taxonomie levels, and consequentiy the absence of a consensus regarding the putative sister-group and other outgroups to the order. Within the Bryophyta sensu Vin (1984), phylogenetic reconstructions using nucleotide sequence data have focused exclusively on the Chapter five: 60

monophyly and the relaîionships of the division and its classes (e.g., Mishler et al., 1992 and 1994; Waters et al., 1992; Capesius, 1995; Heddemn, Chapman, and Rootes 1996) rather than on the evolutionary history of arthrodontous mosses. Variation in the nucleotide sequence of the cblomplast gene &L, encoding for the ribulose 1,s biphosphate carboxylase. has found a wide application in reconstructing the evolutionary histories within suprageneric taxa of vascular plants (Chase et al, 1993; Hasebe et al., 1995). The present study aims at setting the foundations for reconstmcting the phylogeny of all orthotrichaceous geneni based on morphology and for addressing the character evolution in this taxonomicaily and morphologically diverse family. Thus, the objectives are to use hcL sequence data 1) to test the mcnqhyly of the Orthotrichales and the Orthotrichaceae; 2) to detennine the phylogenetic position of the Orthotrichales within the arthrodonteae, and thus identify its putative sister-group; and 3) to test the monophyly and the phylogenetic relationships of the subfamiles proposed by Vin (1982).

Materid and Methods

Taon sampfing. Following the exclusion of five genera from the Orthotrichaceae (chapter two and the) the family now consists of 22 genera. While materiai has been seen for aii taxa except for the recently described Orthominiwn (Lewinsky-Haapasaari and Crosby 1996), extractions were not attempted for taxa known only from the type specimen (Ceuthotheca Lewinsky-Haapasaafi, Leiminiwn Mitt., Lemtia Broth.), or for which only very littie material was avaiiable (Stoneobryum Norris and Robinson). DNA extractions were attempted for the remaiillng 17 genera, and for the most diverse-genera such as Macr~mi~wnand Orthotrichm, several species belongîng to morphologicaiiy distinct infrageneric taxa were tentatively included. DNA extractions were aïs0 attempted for representatives of ail related families in the OrChotrichales. Furthemore ten outgroup taxa were chosen, representing the Funariaceae, , Encalyptaceae, Hedwigiaceae, the Haplolepideae (Rychomirriaceae and Rhabdoweisiaceae), and the ciliate aiternate pristomate mosses (Mniaceae, and Thuidiaceae). DNA extraction and PCR-DNAampiification. DNA was extracted followhg a modification of Doyle and Doyle (1987; see Goffinet and Bayer, unpubl. for fdi protocol). The rbcL gene was amplified via the polymerase chah reaction (PCR) using Taq Polymerase (Promega). The amplification reaction was performed in 50 pL volume including H) mM KI, 10 mM Tris-Ha, 0.1% Triton, 5% glycerol, 2.5 mM MgC12,20 Chapter five: 6 1 mM of each dNTP, 0.2 rnM of primers 21 and 1351R (Wolf, Soltis, and Soltis, 1994) and 0.2 to 30.0 ng of template DNA. This soiution was overlaid with 30 of minerai oil. The samples were exposed to the following temperature profiles using a Grant thermal cycler: one cycle of WC for 3 min and 85% for 4 min during which 1 unit of Taq DNA Polymerase (Promega) wouid k added to each tube, and 30 cycles of 94T for one min, 52% for one min, 72'C for two min, and fhaliy one segment of 72'C for ten min. The double-stranded product subsequently semd as the template for the amplification of single-stranded target DNA. This second PCR followed the same profiles as before but the reactions solution included only one of the two PCR primers: 21 for amplincation of the forward smdand 1351R for the reverse strand. The single stranded product was precipitated with 20% PEGl2.5 M NaCl, washed ntst with 70% EtOH, then with 95% EtOH before king suspended in 7mL of TE (Morgan and Soltis. 1993). Sequencing d Pllgnmeat. Sequencing the singie-stranded template foiiowed the dideoxy chah termination method (Sanger, Nicklen. and Coulson. 1977) using the sequenase@Version 2.0 T7 DNA Polymerase foiiowing the mmufacturers instructions (Amersham, Canada). The sequence of the primers used are given in Table 5.1. The sequenchg products were electrophoresed on a 6 % polyacrylamide gel (0.4 mm thickness; 1X TBE bufler), at 2400 V for four hours. The gels were fixed in 1096 acetic acid, washed with distilled water, and dned in an oven at 65'C for 30 min. and autoradiographed for 3648 hours. The sequences were entered in MacClade (version 3.03) and aligned against available sequences of Sphagnum palustre and Andreaea rupestris (GenBaak acc. no. LI3485 and L13473, respectively; Manhart 1994). The ht and last 30 nucleotide sites, comesponding to the sequences of the PCR primers, were excluded from parsimony analyses. Sequence and Phyiogenetic siulysis. RbcL sequence variation was arialyzed by Neighbor-Joining (NI; Saitou & Nei, 1987), and Maximum Parsimony (MP; Fitch 197 1). Neighbor-joining aaalyses were performed using MEGA v. 1.0 (Kumaf, Tamura, and Nei, 1993) usiag Tamuni's distance parameter, foUowing the authors' "Guidelines for choosing distance measures", and iocluding both transitions and transversions. A ''bootstrap confidence level" for the NJ tree was calculated over LOO replicates. Fitch parsimony analyses of nucleotide data were performed with PAUP v. 3.1 (Swofford, 1993) on a MacIntosh PowerPC 7200190 ushg the heuristic search with the followiag options in effect: keep aii characiers, multistate taxa interpreted as uncertain~es,tree- bisection-recomection branch swapping, steepest descent, collapse zero-length branches, and addition sequence "as is". In an attempt to locate additional islaods of shortest trees Chapter five: 62

(Maddison, 199 1). the search strategy Merfollowed the steps recommended in Pryer, Smith. and Skog (1995) except that the search was ody ~phted100 thes. Bootstrap anaiysis (Felsenstein, 1985; HillÛ and BullT1993) were performed with 100 bootstrap replicates of the heuristic search with the same set of options in effcct. Relative support for branches was fiutber detefmjaed by the Qcay analysis (Bremer 1988; Donoghue et al. 1992) following the converse constraint method developed by Baum, Sytsma, and Hoch (1994). The AC- (accelerated transformation optimization) option of PAUP was applied for calcuiations of branch lengths. The phylogenetic signal ptesent in the rbcL sequence data was estimated by calcdating the g, statistic of the distriiution of tree length of 500.000 random trees produced with PAUP using the "random tree" option (Hiliis and Huelsenbeck, 1992). Consistency (CI) and retention indexes (RI), and f- values are calculated for all MPTs using PAUP. PAUP was aiso used to generate a matrix of absolute and mean distance betwan sequences. Average unit character consistencies (AUCC) are calculated for competing phylogenies in an attempt to select the me with the strongest asymmetric distribution of homoplastic characters io the matrix (Sang. 1995). Costs in te- of number of additional steps required for alternative phylogenies were explond using the enforce constraint command durhg heuristic searches (Swofford, 1993). The satm set of options as in the unconstrained searches was appiied. Results

The rbcL geoe was successfhlly sequenced for 22 taxa of the Orthotrichaceae distributed among 10 genera, for representatives of 2 of the related families as well as for aii selected outgroup taxa (Table 5.2; Appeadix 1). Amplincation products were not obtained for Heficophyllm. The sequences obtained from PCR fhgments generated using the PCR primers Z 1 aiid 135IR, were as expected 1320 bases long (approximately 90% of the total length of the gene). Alignment of the sequenoes with known sequences of Sphagnum palme and Andreaea rupesrris did not require the inclusion of gaps. Phylogeneticdy informative characten are preponderant at the third codon position foiiowed by the nrSt and the second codon site (Table 5.3). RbcL sequence variation yield 108 sites that are potentially phylogenetically informative within the Orthotrichaceae. Informative characters are distributed fairly evenly across the sequence (Fig. 5.1). The distribution of 500.000 random trees ushg all characters had a skewness index g, of -0.55; excludîng the first and second codon position yielded a similar index Cbapter five: 63

(g,= -0.54). while trees generated solely hmcharacters of these two first codon sites had a hi* skewness index (g, = -0.39). Using the maximum parsimony criterion with Sphagnm and Andreaea as the outgroup yields 39 most parsimonious trees 0distri'buted arnong three islands of size 2 1,6 and 12 rrspectively (Fig. 5.3). The MPTs are 964 steps long, and have a CI of 0.390 and a RI of 0.624, The f-value of the trees varies ûetween 13306 and 19198. Both rnethods of analysis agne on the polyphyly of the Orthotric~esand place the Erpodiaceae and the Rhachitheciaceae as weii as the orthotrichaceous genera Amphidiwn and Dtumnondia in a monophyletic clade with Ptychomitk and Rhabdoweisia (Figs. 5.2-4). ConsOaining the swchfor the inclusion of the Erpodiaceae and Rhachitheciaceae in the ûrthotrichales results in trees that are 17 and 15 steps longer than the shoitest unconstrained trees. Including Amphidium and Dnunmonàia in the Orthotrichaceae also increases the length of the most parsimonious topology (+ 25 steps), and fwthermore these two genera remain more closeiy related to each other than they are to any other genus of the Orthorrichaceae. The NT tree apeswith the strict consensus tree over al1 MPTs in the following relationships (Figs. 5.2 & 4): 1) Funaria occupies a basai position amo~gartbrodontous mosses; 2) the Orthomchales and the form a monophyletic clade sister to the ciliate mosses; and 3) the haplolepideae, including Amphidim. Dmmrnondia, VentUriella and Uleastrm, form a monophyletic group sister to derived Diplolepideae. Hemvigia is sister to the two ciliate mosses in 36 MPTs (Fig. 5.3) as weil as in the NJ-tree (Fig. 5.2). and basal in the diplolepideous clade (excluding the Funariaceae) in the ~rnainingthree trees (Fig. 5.3). Compared to M.,NJ provides similar or slightly higher bootstrap values for major Lineages (Haplolepideae, combined Splachnaies and Orihotrichales, Orthotrichaceae, and ciliate mosses), but like MP, NJ fails to yield strong support for their relationships (Figs. 5.2 & 4). The remaining genera of the Orthotrichaceae (thus excluding Amphidium and Dnunmondia) form a saongly supported monophyletic family in the phylogenetic construction ushg Tamura's distance parameter (Fig. 5.2). Among the 39 MPTs found in the cladistic analysis (Fig. 5.3), 15 trees (38%)' distributed between island 1 and 3 (each with 9 and 6 t~esrespectively), support the monophyly of the family. The trees of these two islands ciiffer mainly by the position of Encalypta (Fig. 5.3). In island 3 (mean f-values: 13943), Encalypta is included in the haplolepideae (Fig. 5-34), whiie in islands 1 and 2 it occupies a position basal to the dichotomy between haplolepideae and the derived diplolepideae (Fig. 5.3-C; mean f-value overall: 16634). Cornparison of the distribution of homoplasy among characters fkom two selected trees contrasted maùily by Chapter five: 64

the position of Encalypta (islaud 1versus island 3), yield an AUCC value of 0.7 1 for both, sugges~gno diffe~ncein the distniution of homoplastic characters (Sang. 1995). AU other indexes being equal. the tree with the lowest f-value (see Fams 1972) is shown (Fig. 5.4). The Onhoaichaceae are composed of two major lheages: one including all Macrominioideae (Cardotiella, Desmotheca, Groutiellh, Mucrocornu, Macromim*tun. and Schlotheimia) as weil as Zygodon [email protected] one uniting the rrmainuig species of Zygodon with the Orthotrichoideae (Brydixonia, Orthohichum, and Ulota). Large genera such as Ma~romit~um.Orthohichum, and Zygodon appeas paraphyletic, independent of the method of analysis. In the NJ-aaalysis, as well as in ail MIT of island 2, Desmotheca is basal in a clade centered arouod Macr~rnit~um,while the other MPT suggest a derived position. sister to Macromim~umrichardii.

Discussion

RbcL sequence data. RbcL sequence data are routhely used to reconstruct evolutionary histories among (Chase et al. 1993; Hasebe et al. 1995) or more rarely within (Haufier and Ranker, 1995) genera of vascular plants. With 30% variable sites. inciuding 1296 of sites with changes that rnay be phylogeneticdy idormative, the variation in the nucleotide sequences of the rbcL gene anmg moss taxa may yield sufficient characters for reconstructing the evolutionary history of the taxa included in this study. The distribution of informative characters appears rather uniform (Fig. 1). and is similar to that observed in vascdar plants, that is with no obvious hot-spot detectable (Olmstead and Sweere, 1994). The distribution of randomly generated trees has a left- hand skewness with a g, value( -0.55) significantiy derthan the critical value (-0.10 or -0.08 for 100 or 250 characters respectively; p=0.01) fimished by Hillis and Huelsenbeck (l992), suggesting that our nbcL &ta set is more structured than a random data set of equal size and that it may contain significant phylogenetic signal Ofillis and Huelsenbeck 1992). Uniîke Conti, Fischbach, and Sytsma's (1993) observation that the third codon position introduces "noise", a lower g, value is observed when the matrix is restricted to the third codon position (-0.54 vs -0.39) suggesting that the changes at the third codon position are more smictured. Analyses of a data set restricted to the third codon positions only, yield topologies congruent with those obtained with the complete data set, whereas a search excluding changes at the third codon position results in a consensus tree incongruent with monophyletic concepts of either the ciliate mosses, or the haplolepideae (results not shown). This observation would suggest that the fmt two Chapter five: 65 codon sites carry more homoplasies than the third position, which may thus be more informative. Altematively, the taxa sample may be too disparate for variability at the more conserved positions one and two to be mostly phylogeneticaily informative. The significaace of the clifference between the g, value based on our data set and that of a random rnatrix may as a result be due to di&rences in the hquency of characar States rather than in the congruence among characiers in both data sets (Kaiiersjo et al.. 1992). This second hypothesis most likely applies to our data since our restncted taxon sample represents major lineages of arthrodontous mosses with a long evolutionary history (Frey, 1977,1990). Cornparison of the variation in the nucleotide sequence against the secondary structure of the protein (as commonly doue with nrDNA genes) is not possible, but an indication for the accuracy of our sequences may be obtained nom the preservation of active sites in the enzyme. AU amino acid residues of the active site found in spinach by crystallography (Andersson et al.. 1989) are conserved among studied bryophyte tana, except for position 404 (amino acid mmbe~g)which is scored as polymorphic (Arg in addition to the conserved Gly) for six taxa, and thus excluded fkom phylogenetic analyses in Mi?. This observation may provide some prelùninary support for the accuracy of the sequences obtained.

Circumscnptions of the OrthoMcôaies. RbcL sequence data aoalyzed using either the wighbor-jobhg or the maximum parsimony criterion indicates that the Onhotrichales sensu Vitt (1984) are a polyphyktic mon. Both the Erpodiaceae (Venturiella sinensis) and the Rhachitheciaceae (Uleasfrzun palmicola), as weii as the oahotrichaceous genera Amphidim and Drullunondia are indeed placed in a clade with haplolepideous taxa (Figs. 2 & 4). The monophyly of this haplolepideous clade is strongly supported @P:decay index [Du of4, boots~apvalue (BVJ of 77%; NJ: BV&4%). Constraining these taxa to a relationship within a monophyletic Onhotrichales is fkthermore very costly in tenns of parsimony, requûing up to 25 additionai steps. Variation in the nucleotide sequence of the rbcL gene is thus congruent with eariier hypotheses based on morphology and cytology, suggesting haplolepideous affities of these taxa (Anderson and Crum, 1958; Vitt, 1970 and 1973; Edwards, 1979; Shaw, 1985; Gofiet chapter three). More extensive sampling of haplolepideous taxa wouid be needed before ordinal affiaities are addressed mer(see also affinities of excluded taxa). Chapter five: 66

De Luna (1995) recentiy argued that the ~otrichaiesrepnsented an evolutionary grade, reached by two limages. The cladocarpous lineage (including, e.g., Erpodiaceae, Macromitrioideae) was sister to the Leucodontales and separated hmthe acrocarpous Iine (e.g., Rhachitheciaceae, Orthotrichoideat) by the Hedwigiaceae. Constraining the heuristic search to include Hehuigia citiuta in the ûrthotrichaceae (as defined in Fig.5.2 & 5.4, and with no fiderrelationships specified within this clade) yielded 12 mes that not oniy are 16 steps longer than the MPTs in the unconsûained search, but also share a monophyletic Orthotrichaceae. me incongmence between De Luna's morphological study (De Luna 1995) and the present molecular analysis may be due to inconsistent and indequate taxon sampling with the absence of reprcsentatives of the Leucodontineae fiom the rnolecular analysis and of brydean taxa hmthe morphological analysis. The latter analysis merincludes taxa that are here shown to be unrelated to the diplolepideae (i-e., Erpodiaceae, Rhachitheciaceae) and might be affected by the subsequent misinterpretations of the nature of the peristome of these families (scored as an exostome instead of an endostome). A closer relationship between the Hedwigiaceae aad the Orthotrichaceae cannot be excluded, but if the Kedwigiaceae were closely related to the Orthotrichaceae, rbcL data suggest that the Orthotrichaceae would most Likely remain a natwal group. The branch that supports the Hedwigiaceae-cladocarpous Orthotrichales-Leucodontales Lineage is supported in De Luna's analysis (1995) by 3 synapomorphies, namely: plagiotropic growth. presence of pseudoparaphyuia, and diffenntiated penchaetid leaves. Two of these characters are reversed in the cladocarpous Orthotrichales, leaving a single character that actuaily supports a relationship of these taxa with the Hedwigiaceae and Leucodontaies, aamely plagiotropic growth. This cbaracter is, however, homoplastic among many families and genera of mosses (Neusel, 1935) and may thus not be truly informative at the ordinal level. Analysis using the distance method yields a single tree (Fg. 5.2) that shows strong support for the monophyly of the Orthotrichaceae, a phylogeny congruent with that of 15 of the 39 MPTs found in the cladistic aaalysis. Kass and Wink (1995) and Barker, Linder, and Harley (1995) also used both methods (MP, and NJ) for phylogenetic reconstructions based on rbcL sequence data, and the rrsuits of both analyses were fairly congruent. Km, Rohlf. and Sokal(1993) criticaily examined the accuracy of the NJ method under different constraints on a random data set, and found that this method "has the highest accuracy overali". Russo, Takezaki, and Nei (1996) recently compared the effciency of various tree-building methods in recovering a known phylogeny, and concluded that NJ "gives as good a result as the more thne consuming... methods". The Chapter five: 67 congruence between the NJ tree and 15 MPTs is therefore hem interpreted as supporthg the monphyly of the OOrihotrichaceae. With the exclusion of the Erpodiaccae, Rhachitheciaceae, and Hedwigiaceae nom the Orthotrichales, the order is now reduced to the Oahotrichaceae and the Helicophyllaceae. In the absence of molecuiar data, the phylogenetic relationship between these two families remains dubious due to the unique combination of morphological characters of Helicophyllum (Vitt 1982a), and will most Likely ndto be.resolved using molecular characters. A phylogenetic relationship between the Orthotrichaceae and the Cryphaeaceae adthe monotypic Wardiaceae was not examimd in the present study. Wardia hygrometica should be excluded fmm the Orthotrichales based on well differentiated alar cells and its prosemihymateous cells. Combined with the acrocarpous condition (Welch, 1943). these characters may indicate a bryalean origin, a hypothesis recentiy confirmed ushg 18s gene sequences (Hedderson et al. unpubl.). The Cryphaeaceae have traditionaily been placed among the Leucodontineae (Vitt 1984; Buck and Vitt, 1986). A relationship of this family to the Hedwigiaceae, and thus the Orthotrichaies, has been considered and rejected (De Luna 1995), but may need to be reexamined in the light of the discovery of cladocarpous species of Cryphaea sensu lato (La Farge-England., 19%).

OrdiaPl rehtiomhip of the Orthotrichales--The affhities of the Orthotrichales - here restricted to the ûrthotrichaceae - as indicated by cornparisons of the rbcL séquences, are within a diplolepideous lineage including the Splachnales and the ciliate mosses. This clade shares a cornmon ancestor with the Haplolepideae, and together they form a sister group to the Funariales (Figs. 5.24). This topology is congruent with Vitt's hypothesis (Vitt, 198 la) that the opposite diplolepideous peristome is primitive among arthrodontous mosses, and that the haplolepideous peristome is denved hmsuch an ancestral type (Vin 1984). Alternative hypotheses proposed by Lewinsu (1989; Orthoûichales basal to dichotomy between haplolepideae and ciliate diplolepideae) or by Shaw and Rohrer (1984) and Crosby (1980; ciliate diplolepideae are the most primitive mosses) both nquire 5 additional steps. The phylogeny obtaiwd by either NJ or MP analysis (Figs. 2 & 4) also agrees with Vin (1984) with regard to the monophyly of a lineage composed of the Splachnales, Orthotrichales and the ciliate mosses. The relationships among these lineages nmain however, ambiguous (Fig. 5.3). The most parsimonious senario points toward the Splachdes and the Orthotrichaies behg sister-groups. Vitt (1984) by contrast proposed Chapter five: 68 that the Orthoaichaies are sister to the ciliate mosses, a topology that would only require one additional step based on our rbcL &ta set. Koponen (1977,1983) considered the genus Bra~hymi~umto be the most primitive extant member of the Splachaaceae. Unlike in related genera, the thickening on the PPL is heavier than on the OPL, and the PPL also bas strong trabeculae (Koponen, 1977,1982). Botà these features are shared with the FC(IUIMand the Bryum pdtome ~lpc(Shaw and Rohrer, 1984). If the Ortbotricbales and the Splachaales formed a monophyletic group, both orders would most likely remain naniral orden, with the Orthoaichales defmed by thick-wailed laminai cells. The heavier thickenhg on the OPL would have arisen independently in both limages. In terms of peristome evolution, these liwages would form a plesiomorphic sister group to the ciliate mosses, the latter one king defined by the asymmetric division of the PLLeading to the development of ciüa If we consider the alternative scenario where the Splachnales are sister to a clade composed of the Oahotrichales and the ciliate rnosses, the most parsimonious topology based on our data suggesu that within the latter, the ciliate mosses remain monophyletic. An extensive mon sampling in the Bryales is, however, needed before their evolutionary relationship can be addressed more critically. Both the Orthotnèhales and the Bryales sensu lato taxa share one plesomorphic state with the Splachnales: either completeiy aligned ceil divisions (Orthoaic hales; Lew insky , 1989) or heavier thickening of the PPL and ventral exostomiai trabeculae (typical ciliate mosses). The ciliate mosses have traditionally been dehdby a strongly asymmetric division in PLleading to the development of cilia. While some "ciliate" mosses actuaüy lack cilia (Buck and Vitt, 1986), have a thickened OPL (Eucamptodontopsis, A Newton pers. corn), and others even have opposite peristomes (Garovaglia div. sp. Dwing 1977, Nishimura and Watanabe. 1992), none has yet been found to have a nist-late symmetric division in the PL. An asymmetric dirision is elsewhere found oniy among haplolepideous taxa Considering that the Rhachitheciaceae are here show to be of haplolepideous affinity, their penstome with a 2: 1 formula most Wrely results fkom reduction, through the loss of the asyrnmetric first-late division. The asymmetric division of the ciliate mosses may not be homologous to that of the haplolepideae (Shaw, Mishler, and Anderson, 1989). Assuming that the genetic complexity behllid the asymmetnc division in both group is similar, we camot exclude the possibility that a loss of it or a reversal to a symmetric division may be possible among ciliate mosses and may even occur in such genera as Miefichhoferia with a peristome formula of 424(Shaw and Rohrer, 1984; Shaw and CN~,19û4). Since the possibility of a reversal to a symmetric Chapter five: 69 division cannot be excludeci, a bryalean ongh of the Onhotsichaceae will need to be furtber examïned in cornparison with a broad sanipie of ciliate taxa Hedeniis (1994) considered Schlotheinrio to be pleurocarpous, and based on this interpretation suggested that the ûrthotrichaceae "are rather close to the clade where most pleurocarpous mosses belong". He Merhypothesized that "the transition to plewocarpy mut have been graduai", and in this scenario the Ortbotrichaceae would occupy an intermediate position. The Macmmitnoideae dinet however hmtypical pleumcarps by severai characters and should rather be considerrd cladocarpous (La Farge-Englancl, 1996). In genera where both cladocarpy and acmcarpy occur, the former seerns to be nstrîcted to terminal taxa suggesthg that the trend is hmacfocarpy to cladocarpy with no obvious cases of revenais (La Farge-England, 1996). Such a general trend may suggest that in the ûrthooichaceae too, the primitive condition is acrocarpy. Thus if the Orthotrichales are indeed reduced ciliate mosses, their putative sister group would most likely belong to a group of acrocarpous Eubryales.

Subf- phylogenetic relationships. The ten orthotrîchaceous genera remabhg in the present analysis are distributed among two clades that are moderately to strongly supported by either method of analysis, with bootstrap values (BV)ranging from 77 to 94% and decay indexes (DI)of 2 and 3. The orthotrichoid-clade combines the Zygodontoideae (Zygodon sections Zygodon and Bryoides) and the Orthoaichoideae (Bryoàixoniu, Orthotrichuni, UIora),while the macromitrioidclade includes ail the Macromitrioideae (Cardorielia, Desmotheca, Groutiella, Macrocorna, Macrornitrium, and Schlotheimiu) as well as Zygodon obtus$oLius (Zygodon sect. obtuicijiolii). Zygodon obtrrsi$olius has retained severai charactea considered here plesiomorphic, such as acrocarpy, smooth, cucullate calyptrae, but exhibits aiso some features reminiscent of the Macromitrioideae (strongly bulging cells, coarse papillae, undifferentiated basai cells, etc.; see chapter SU). If 2. obtusifolius is to be retained within the Zygodontoideae, the unexpected relationship with the Macromitrioideae as pmposed by rbcL sequence data may be an artifact due to a long branch attraction (Hendy and Penny, 1989) or an indication of hybridization involving chloroplast capture from a macromitrioid taxon (Soltis and Kusoff, 1995). In the latter case, the topology obtaîned here may represent the "correct" gew me, but deviate fiom the mie phylogeny of the taxa (Doyle, 1992). This hypothesis is currently king tested by comparing sequence data of the nuclear gene 18s. Chapter five: 70

With the exclusion of the Dnimmondioideae, Vin's (1982b) phylogenetic arrangement of subfamilies would have the Zygodontoideae (cucullate calyptrae) basal to a dichotomy between the ûrthotrichoideae and the Macromitrioideae (both typically or exclusively have large mitrate calyptrae). The sister-group relatiomhip between the orthotnchoid and the macromitrioidclades proposeà hem deviates kmVitt's (1982b) phylogenetic concept of the family by the inclusion of the Zygodontoideae in the former clade, and parsimony would need to be reldby 6 steps for bis concept to be satisfied (without comtraiaing the affinities of Z obtus~oli~d~).The monophyly of Zygodon and thus Zygodontoideae is not supported by rbcL sequence data either, but the relatiomhip of sections Zygodon and Bryoides to the Orthotrichoideae remain unresolved. If Zygodon is indeed paraphyletic, section Zygodon would most likely be the most primitive taon, given that it shares the plesiomorphic state "smooth lamina1 ceils" with either a splachnaceous or a bryaceous ancestor. The two sections of Zygodon differ on average by 40 mutations, while oniy 28 changes separate section Zygodon fiom the Orthotrichoideae, compitnd to 43 changes becwmi section Bryoides and the Orthotrichoideae. Such divergences may be indicative of the monophyly of a clade composed of section Zygodon and the ûrthotrichoideae, and support a single orîgin of papillae that would defhe this clade. With regards to the Macmmitrioideae, both methods of analysis yield two strongly suppocted monophyletic clades (Fig. 5.2 & 4): one composed of all thespecies of Schlotheimia, the other including 1remaining Macromitnoideae (is., Desmotheca, Groutiella, Macrocorna, and Macr~nti~um).Vitt, Koponen, and Noms (1995) suggested that Schlotheimia and Cardotellla should k placed in a separate subfamily, and even if they did not explicitly state it, this hypothesis was based on the distinct lobate calyptra RbcL data do not support such relationship as 10 more steps are needed to unite Schlotheimia and Curdotieliu into a monophyletic clade. The species of Schlotheimia differ on average hmother Macrornitrioi&ae (including Cardotielia)by 47 bases. By conuast, the average distance between these other macromitrioid genera is only 23 bases. If rates of molecular evolution are assumed to be simila among these taxa (a reasonable assumption considering both clades are prqdominaatly phyiiodioicous, and epiphytic in tropical montane forests; see Britten 1986). Merences in nucleotide sequence may indicate a relative ancient divergence betweea Schlotheimia and the other Macromitrioideae. Schlorhemia differs fkom other Macromitrioideae also by several morphological characters, most of which are unique within the Orthotrichaceae (muitistratose calyptrae, color of leaves, diagonal arrangement of cells; Goffinet unpubl.). Chapter five: 7 1

and thus not informative phylogeneticaliy. The pleisiomorphic state "smooth laminal ceiis" may, as was argued in the case of Zygodon sect. Bryoides, be indicative of the primitive position of Schlotheimia in the evolution of the Macromitrioideae. conside~g both the morphologicai aiid molecular divergences between Schlotheimia and the remaining Mactomitrioideae. and the strong support for the monophyly of both clades on molecular grouds. accommodating Sclilotheimia in its own subfamiy, as suggested by Vitt, Koponen, and Noms (1993). may better reflext the evolutionary relationship between these clades. Brotherus (1925, as Pseudo-Macromitrioideae), Walther (1983) and Cnun (1987) isolated Desmotheca hmother Macromitrioi&ae on the bais of its dimorphic steriie and fertile branches. Vitt (1990) however argued that excluding Desmotheca from the Macromitrioideae wodd most certainly result in the paraphyly of the later subfamüy. Based on variation in the rbcL sequences, the phylogenetic affinities of Desmotheca clearly Iay with Macromitium sensu [ato, but =main ambiguous with regard to its sister group within this clade. Whether Desmotheca should be retained in its own subfamily, and placed sister to the Macromitrioideae (excluding Schlotheimia), or be inserted within the latter, is not clear. In 33 MPTs (including the 15 that share a monophyletic Orthotrichaceae) Desmotheca occupies a derived position (Fig. 5.4). nested between two Macromifrium species. In the remaining 6 MPTs as well as in the NJ tree (Figs. 5.4 & 2), Desmotheca is sister to the Macromitrioideae (Schlotheimia excluded). The phenetic distance between Desmotheca and other macromitnoid genera (Table 5.4; excluding Schlotheimia) are on average similar to those between Orthonichwn and Ulota (20 bases), in the ûrthotrichoideae, suggesting that placing Desmotheca in a distinct subfamily may not be appropriate. Besed on our phylogenetic analyses of the rbcL sequence variation, the Orthotrichaceae could be regarded as composed of two subfamiles, the Onhotrichoideae and the Macromitrioideae with the latter Merdiviàed into two tnks "Macromiaiae" and "Schlotheimiae". Aitematively, the two subfarniiies may deserve recognition at the family level as suggested by Churchill and Linares (1995) and the Macromitriaceae would then be composed of two subfdes, the Macromitrioideae and the "Schlotheimioideae". The statu of the Zygodontoideae, and thus the relationship of the two main subgenera to the Orthotrichoideae needs merstudy.

Phylogenetic relationsbips of OrthoMchac~onsgenera. Analysis of nbcL sequence using either the MP or the NJ method suggests that four of the five larger Chapter five: 72

genera, namely Ma~rornit~um,Orthofrichun, LllotaT and Zygodon, are parapbyletic. The paraphyly of Zygodon bas been addressed eariier. Bqruduonia is a monotypic genus endemic to New Zealand Sainsbury (1945) argued for a generk distinction Born Ulma on the basis of "the highly differentiated and coaspicuous perichaetial bracts and the diminutive caiyptra" and the immersed capsule. In addition, the perichaetial leaves have prorate basal lamina1 ceUs (Gofiet unpubl.), a feature othemir unknom hmthe Orthotrichoideae. Btyodbonia does, however, share many of the characters found in the genus Wlota, such as very thick-walled cauluie cells, the differentiated marginal celis of the lamina, îhe Dexuose to crisped leaves. BryodLxonià may thus be patristically very denved; however, in cladistic terms it may not deserve taxonomie recognition at the generic level. RbcL sequences of Ulotu lutea and Bryodkonia pen'chaetialis meronly by 6 mutations, a degree of divergence that is similar to that found between species of Groutiella (4) or Schlotheimia (8-16) but moreover. it is lcss than the divergence recorded between the two species of Ulota (14; Table 5.4). Immersed capsules arc characteristic of many mosses that are taxonomically unrelated but share a xerophytic habitat (Vitt, 198 la). Within the Orthotrichaceae completely immersed capsules are characteristic for Schlotheimia sect. Stegotheca, and a shortening of the setae leading ultimately to an Mmersed capsule occurs in Onhotichum subg. Gymnoponrs sect. Leiocarpa, subg. Pulchella sect. Rivufario and sect Diaphana, as well as in subg. Or?hotrkhum (Lewinsky, 1993). Considering the low degree of morphological and molecular divergence of Brywlixonia, segregation at the generic level does not seem appropriate. The monophyly of Ulota (including B. perichetialis) is compromised by O. lyellii in the NJ tree and in 13 MPT (in 13 other MPTç their relatioaship is not resolved). Sirong affurities of O. lyelüi for Ulota (MP: 99% BV and DI of 5) and LI. obtusiuscula in pariicular (13 MPTs) rnay indicate that among the different lineages of Orthotrichum, subg. Gyninopoms (Braithw.) Limpr. is the most closely related to Ulm. Among the characters that subg. Gymnoporus sect Leiocurpa (see description in Lewinsky, 1993) shares with Ulota, only the long flexuose vegetative leaves may be derived wit.the Orthotrichoideae and thus be indicative of common ancestry. The genus Ufota (even if including Brydixonia) is morphologicaily well defined hmOrthotrz'chum, suggesting that the paraphyly of the Onhotrichum, if confirmeci, would need to be resolved by dividing the genus into discrete entities rather than broadening the concept of Orthotrichum by including Ulota. The relationships of the subg. Or~hophyllumDelogn. (O. obtusifolium) and subg. Onhomkhiun (O. anomalwn) are not uuambiguously Chapter five: 73

resolved either: they form sister *w in 13 MPTs as well as the W me, whiie in 13 other MPTs, 0.ob~olium is sister to a clade composed of the remaining Orthotnchoideae. Orthomkhm ob~oliumhad ken placed together with the related 0.gymnostomum Brid. in the genus, NyholmielIa (sek LeMnsky 1993 for history), based on "the obtuse leaves with plane or incwed leaf matgin and incrassate leafkells with a stout central papUae on each side" (Lewinsky 1993). Patterns in peristome omamentation are similar to those observeci elsewhere in the genus (Lewinsky 1993) and it may therefore be more parsimonious to retain subg. Onhophyllm in Orthohichurr Within the macmmitrioidclade (Fig. 5.4) the nlatiouships rpmain ambiguous too (except for Schlotheimia see above), either because Macromitrim and Mucroco111(~euly are paraphyletic, or because of an insufficieat taxon sample. The genus Ma~romi~umis, with over 250 species (Vin 1982) by far the most speciose genus of the ûrthotrichaceae (Vin 1982). Mitten (1869) divided the genus in four sections, to which Buck ( 199 1) recently added sect. Reverberatum. Two of these have recently ben raised to the genus level, narnely Mi~rorni~um(now Groutiella Steere) and Macrocoma Grout. The genus Macromitrium remains however morphologicaUy extremely diverse in terms of size of the , degree of differentiation of the basal ceils, shape of the um,laminal cell shape, orientation and omamentation. The relatively high cost in terms of parsimony, for a monophyletic Macrominiurn (9 steps) may be seen as just one other indication that the genus as currently defined is stül a heterogeneous assemblage. Groutiella differs fiom M~crorni~umby the marginal limbidium of hyaline elongate cells, and a short caiyptra cove~gody the upper portion of the m. Except for G. tomentosa. Groutiella is restricted to the Neotmpics. where 10 species occur. The sister species in ail shonest trees is Macromitrium longi$ioliwn, a neotmpical endemic, rather than any of two paleotropicd taxa (M. richardii is known nom Afiica and the Americas [van Rooy and van Wyk, 1992; Vitt 19931 and belongs to M. ligulare-group of the Old World; M. incurvifolium occurs throughout the Pacinc ûcean [Vin & Ramsay, 19851). Whether these putative affinities of Groutieiiu for neotmpical Macromitria indicates a common ancestry with a distinct neotropical lineage of Macrornitrhm. needs to be further investigated. The genus Macrocoma is composed of two subgenera, subg. Truchyphyflum(M. papillosa) and subg. Macrocoma (M. tenuis), that ciiffer by a series of characters, but particulariy by the weii developed peristome of the former (Vitt 1980). Our molecular data suggest that these two taxa too, form an artificial group; three additional steps are needed to restore the monophyly of Macrocoma. Macrocornu papillosa is found in a Chapter five: 74 basal position among macfomitrîoid taxa (excluding Schlotheimia) in 33 MPTs (Fig. 5.4). while in the rexnaining 6 it is found in a more derived position with both species of Groutieila and Macromimùm longijiiliran and Macrocorna tenuis. While the Macromitrioideae are typicaliy cladocarpous (i.e., with theu perichaetia terminal on lateral branches) and have dimorphic leaves between stem and branches, M. papillosu is clade and acrocarpous (Le, with penchaetia terminal on lateral branches and on the stem; Gofnnet unpubl.) and the leaves are not differentiated into stem and brancb leaves. True acrocarpy also occurs in subg. Mucrocorna (e.g., M. brudiensis wtt] Vitt), wMe other taxa of this subgenus arc strictly cfadocarpous (e.g. M. tenuis [Hook. & Grev.] Vitt). Acrocarpy and cladocarpy bave not kenreported before hmthe same mon, not to mention nOm the same individuai (see La Farge-England, 1996). The combination of plesiomorphies such as undinerentiated stem and branch leaves, terminal cauline gametangia and the complete double peristome may be a strong indication that subg. Trachyphyllum is a primitive clade not only when compared to subg. Macroconin (Vitt 1982), but maybe even with regard to the evolution of the Macromitrioideae.

Aninities of excluded taxa. Critidy addressing the relationships of the taxa here excluded fiom the Orthotrichaceae is beyond the scope of the present study, and would need a broader sampluig of haplolepideous taxa A suite of unique characters combined with the lack of chanicters that are phylogeneticaiiy crucial rnay always hamper detemiining sister group relationships based on morphology only. Crum (1987) already suggested that gametophytic characters may not sufnce to resolve the phyiogenetic relationship of Druunondia. The same opinion prevails with regards to Amphidim. Brotherus (1929, Anderson and Cnun (1958). and Vitt (1973.1982a, 1984) placed Amphidiun near Rhubdoweisiu. Our nsults, though preliminary, do not suggest that these genera an closely related and friture study may need to consider alternative relationships, as for example with Glyphomizriwn, due to the overall similarity of Amphidiwn lapponicwn-Glyphomitrium daviesii. Presence of a peristome aiiows for a more explicit hypothesis to be made regarding the systematic position of the Erpodiaceae and the Rhacbitheciaceae. Edwards (1979), as part of a review of the haplolepideous peristome, exarnïned the peristome architecture of Venturiellu, and concluded that the teeth are "strongly dorsally mbeculate, and also have a rudimentary unthickened basal exostome", and that 'these characters are of a haplolepideous peristome dthough not of the dicraaaceous type" but of a distinct type. the Seligeria-type. This peristome type is characterized by little ventrai Chapter five: 75

thickening, strong dorsal trabeculae, and an exostome reduced to a thùi, smooth membrane adhering to the trabeculae. "ibis combination of characters has also been observeci in the Rhachitheciaceae, and has ken interpreted as a possible hdication of haplolepideous aninties of the fdy(GOffi.net unpubl.). Molecular data thus tend to contirm these hypotheses, and consequently the peristome of the Rhachitheciaceae and the Erpodiaceae should k regardeci as derivecl, through reduction, from a typical haplolepideous peristome. Though the monophyly of a group of taxa sharing the Seligeria-type peristome has not kxnCnticaily examined, the nearly identical pewtomes of Rhachithecim. Giyphornim*m(Ptychornitriaceae), and Blindia (Seligeriaceae), may be seen as an indication of close phylogenetic relationships, despite gametophytic ciifferences. Altematively. the Rhachitheciaceae may be more closely related to Rhabdoweisia in the Dicranaies. considering the similarities in the overall habit. leaf shape, and celi shape and differentiation.

Phytogenelic conclusiolls-Sequence data of the chloroplast gew rbcL are usehl in circumscnbing the ûrthoaichales, particularly with regard to the systematic position of taxa lacking peristome features that are cenaal to the classification of mosses. Analyses of the variation in the aucleotide sequence using either the parsimony or the distance method strongly suggests that the Orthotrichales are polyphyletic, and that the Erpodiaceae and the Rhachitheciaceae are of haplolepideous affinities as suggested by their Seligeria-type peristome. The ûrthotrichaceae tm, are show to be an aaincial assemblage due to the cumnt inclusion of Anrphidiwn and Dnmtmondia, two genera better placed among the Haplolepideae. The Orthotrichaceae are only distantly related to the latter clade and should rather k considered a member of a derived diplolepideous clade. The relationship to the Splachnales and the ciüate mosses remain unsettled, but at present all three üneages are best considered monophyletic. Molecular data do not support Vin's (1982b) subfdalphylogeny. and instead suggest that the Orthotrichaceae are composed of two iineages. The ktconsists of the Zygodontoideae. and the Orthotnchoideae, while the second includes ail Macromitrioideae. The monophyly of the Zygodontoideae remains ambiguous, and it is proposed that if the paraphyly suggested by rbcL sequence data is confirmed by future studies that Zygodon sect. Bryoides. hcluding the taxa with smooth laminal celi and often weil-developed peristome, may repmsent the most primitive clade within this predominantly acrocarpous clade. Schloiheimia would occupy a similar position in the evolutionary history of the Macromitrioideae. Chapter five: 76

The Orthotrichaceae are now composed of 20 genera: Bryodkonia, Cardotieiia, Ceuthotheca, Desmotheca, Florschue~eila,Growicla, Loiùmitntnwn,Leptudon~~opsis, Leratia, Macroconta, Mactomitrium, Muelleriefla, 0rthomimtnm,Ortho~fikhum, Pleurorthotnchum, Schlotheimiu, Stenomimbn, Stoneobryunr, Wlota, and Zygodon Examination of cladistic relationships and associateci pbcnetic distances suggests that the monotypic genus Brydixonia may k better regarded as a paaistically derïved species of Ulota. Our molecular data furthemore reveal that larger genera such as Mact~mi~um and Zygodon may merely rrpresent evolutionary grades. Gene data obviously have proven to provide a significant contribution in resolving the circumscnption of the ûrthotrichaceae, and the reiationship of the main Lineages. Above all, however, this molecuiar study has laid the foundation for criticaily reexamining the morphological characters that are central to the generic concept used in the Orthotrictiaceae. Chapter five: 77

Figure S. 1: Distribution of phy logenetically informative charaders, the average consistency index, and the average number of steps per phylogeneticdy informative character dong the rbcL sequence. Each class contains 50 sites, except the first one which inchdes sites 3 1-50. e:total number of phylogenetic informative characters (PL). m: average CI value multiplied by 10 O:Total number of steps per total number of phylogeneticaily informative characters Chapter five: 78

G, chimborazensis G. apiculata Mt. longifolium quinque faria papillosa tenuis incurvifolium richardii apicula ta Z, obtusifolius S. brodi S. tecta S. triehomi tria

0, obtusifolium O. anomalum obtusiuscula lyellii lutea perichae tialis

Figure 5.2. Phylogenetic tree reconstnicted using the neighbor-joining rnethd with Tamura's distance parameter including both transitions and transversions. Bootstrap values (100 replicates) 50%or higher are plotted on the tree. Taxa excluded from the Orthotrichales are in bold. Chapter five: 79

- Macromitnoid-clade Macromitrioid-clade r Orthotrichoid-cIade rn Orthotrichoid-dade - Splachnaceae - - Sptachaaceae Hedwig ia Hedwigiiz I Bryales (ciliate mosses) Bryales (ciliate mosses) m Haplolepideae - HapIoIepideae rn A Encalypta - Encaiypta Funariaceae Funariaceae

r Andreuea I Andteaea SP~P~0 Sphagnum

Macromitrioid-clade Orthotrichoid-clade L Splachnaceae Hedwigia Hedwigia Bryales (ciliate mosses) Bryales (ciliate mosses) Haplolepideae Haplolepideae EncaCyptu Eeypta Funariaceae 1- 1- Andreaea 1- Andreaea

Figure 5.3. Summary consensus trees of most parsimonious Fitch mesfound in a heuristic search ushg rbcL sequence data with Sphagnum piustre and Andreaea rupesfrisas designated outgroups. A: strict consensus uee of all3 islands (39MPTs; mes964 steps; CI: 0.390; RI: 0.624); B: 50% majority rule tree of island 3 (1- (mean f-value: 13943 f 656); C: 50% majority nile tree of islands 1-2 (27MPTs) (mean f-value: 16633 f 1308); D: 50% major@ nile tree of ail 15 mes showing the Orthotrichaceae monophyletic (mean f-value: 16722 f 20 15) Chapter five: 80

S, tecta S , trichosni tria S, brownii

tenuis ckimborazensis apiculata longi folirun incurvifofium richardii apiculata quinquefaria -2, obtusifolius Z, intermedius 2, reinwardtii O. obtusifolium O. anomalum O, lyellii W. obtusiuscufa U, lutea B, perichaetialis

- 9l L Hedwigia 55 -

Vbatririella Aulacopf lttni P tychomi tri um Ulaartm Rhabdoweisia Eucalyp ta Funaria h- Funaria a. Andreaea Sphagnum Figure 5.4. One of 39 most parsimonious Fitch txees found based on rbcL sequence data and using Sphagnum and Andrea- as outgroup. The dotted lines identw branches not present in the strict consensus tree. Bootstrap vaiues (96 of 100 replicates) higher than 50% are given below the branch, and decay indexes are presented above the branch. Taxa excluded nom the Orthotrichales are in bold. (Abbreviations for the Orthotrichaceae follow those of table 5.4) Chapter five: 8 1

Table 5.1. Synihetic p"nm (5'03') used for sequencing the rbcL gene in mosses (* marks primers designed and pmvided by O. Zurawski; othen designed by authors).

Reverse primers 295R CTAATGGGTAA-CATAAGC 5 1SR CATCCTAATAATGGACGACC 678R* GATTTCGCCTGTTTCGGCTTGTGCTMIATAAA 895R* ACCATGATIICTTCTGCCTATCAATAACTGC 108 1R CCCAGTCTTGAGTGAAGTAAATACC Chapter five: 82

Table 52. Taxa for which the rbcL gene sequence was obtained in thb snidy (all vouchers deposited in ALTA uniess otherwise noted). Taxon Vouchet

Zygdontoklone Zygodm pungens' C. Miiii. Lu Forge-Englmtd 8097 Zygodon obtus~oliusHook. Vitt 38301 Zygodon intemedius B.S.G. Viiî 29262. Zygodon reinwardn'i (Homsch.)Braun Gofinet 636. Amphidium lapponicum (Hedw .) Schirnp. Vitt 33854. Orthotrichddeae Orthotrichm obtusifolium Bnd. Vitt 33870. Orthotrichum anontalm Hedw. Gofinet 41 15. OrrlotnChum lyellii Hook. & Tayl. Gofinet 3 162. Ulota lutea (Hook. f. & Wh.)Mitt. Fifie 8042. Uhta obtusiwcula C. MU. & Kindb. Gofinet 3161. Bryodtkonia penchaetiafis Sainsb. Fi$e 8083. Drummondioideae Dnunmondia prorepens (Hedw.) Brin Macromitrioideae Schlotheimia brownii Schwaegr. Vin 27485. Schlorheiinia tecta Hook. f. & Wils. Schdfer-Venvimp 9686. Schlotheimia trichornitria Schwaegr. Scwer-Verwimp 6902. Cardotieila quinqut$aria (Homsch.) Vitt Buck 26230. Groutieflaapiculatu (Hook.) Cm& Steere Gfliner 2764. Groutiella chimbororense Wtt.) Florsch. Gofinet 1 173. Macrocoma pupillosa (Th&.) Vitt Matteri 652 1. Macrocoma tenuis (Hook. & Grev.) Vit Breedlove 69342. subsp. sullivuntii (C. MU.)Vin Macromitrttrtumincurv~oliwn. Streimann 49345. Macromitrium iongifoiium (Hook.) Brid Gofinet 656. Macromitrium richrdii Schwaegr. Gofinet 2648. Desmotheca upiculara (Dow & Mok) Lindb. Vinas 96-4 s specimn is tentatively identitied as Zygodon pungens C. MUI. a species hitherto not known fkom Afnca (Maita 1926); but may reprisent a new species. Chapter five: 83

Table 5.2 (cont). Taxa for which the rbcL gene sequence was obtained in this study (ail vouchers deposited in ALTA dessotherwiSe noted). Taxon Voucher ENCALYPTACEAE Enculypta pro~~uBmch Vitt 37966-

ERPODUCEAE Aulacopiluni hodgkinssniae (C. MUL) Broth. Vitt 2826 1 FUNARLACEAE Funaria apophysata (Tayl.) Broth. Vitt 27234 Funada hygrumemca Hedw. Pn'ddZe 1408.

HEDWIGIACEAE Hedwigia cilùito (Hedw.) P. Beauv. 1MNIAcEAE Mnim thomsonii Schimp. Vitt 35884. PTYC~OM~IACEAE Prychomitrium gardneri Lesq. Irerand 7038 (PMAE).

Rhabdoweisia crenulata (Min) Jameson Vin 36707.

Uleastnun palmicola (C. MU.) Zander Vitt 21 162. SPLACHNACEAE (Dicks.) Lilidb. Schofeld 98443. sphen'cwn Hedw. Goward 95- 1470.

'I'HmDIACEAE Abietinella abietim (Hedw.) Fleisch. Gofinet 4106. Chapter five: 84

Table 5.3. Distribution and fiequency (96) of constant and phylogeneticdiy Motmative sites over ali taxa and over the ûrthotrichaceae only; chanrters with ambiguous or rnissing data an included.

ALLTAXA nrSt second thüd total cqdUn position 1 1 Constant Phylogenetically informative Table 5.4. Pairwise cornparison of rbcL nucleotide sequence within the Orthotrichaceae Below diagonal: absolute I distance; above diagonal: average distance. Distance values used for intrageneric cornparisons (see text) are in bold.

1 S. tecta 2 S. trichamitria 3 S. brownii 4 C. quinquefaria 5 M. tenue 6 M. papillosa 7 G. chimborazense 8 G. apiculata 9 Mt. incurvifolium 10 Mt. longifolium 11 Mt. richardii 12 D. apiculata 13 Z. intermedius 14 2, reinwardii 15 2. pkngens 16 2. obtusifolius 17 O, lye2lAi 18 0. obtusifolium 19 0. anomalwn 20 U. obtusiuscula 21 U. lutea 22 B. perichaetialis

Macromitrium; S = ~chlotheimia;O = Orthotrichum; U = Ulota; Z = Zygadon) Table SA (cont.). Pairwise cornparison of rbcL nucleotide sequence within the Orthotnchaceae Below diagonal: absolute I distance; above diagonal: average distance. Distance values used for intrageneric cornparisons (see tent) are in bold.

- - 1 2 3 '4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 1 S. tecta 0.031 0.032 0.035 0.045 0.033 0.045 0.039 0.042 0,041 0.047 0.047 2 S. trichornitria 0.033 0.032 0.033 0.041 0.030 0.041 0,034 0.038 0.036 0.042 0.042 3 S. bromii 0.029 0.035 0.035 0.041 0.031 0.041 0.034 0.038 0,038 0.043 0,043 4 C. quinquefaria 0,018 0.038 0.036 0.045 0.035 0.048 0,039 0.043 0.045 0,049 0.049 5 M. tenue 0.017 0.036 0.037 0.042 0.031 0.044 0.033 0.039 0.039 0.044 0,044 6 M. papillosa 0,020 0.036 0.036 0.042 0.030 0.044 0.036 0.041 0.039 0.044 0,084 7 G. chimborazense 0.020 0.042 0.042 0.045 0.034 0.049 0.040 0.042 0.045 0,049 0,049 8 G. apiculata 0.020 0.045 0.044 0.045 0.037 0.051 0.042 0.044 0.046 0.049 0,051 9 Mt. incurvifolium 0.017 0.036 0.034 0.044 0.032 0.047 0.038 0.044 0.042 0.047 0.047 10 Mt. longifolium 0.029 0,048 0.049 0.050 0.042 0.055 0.045 0,048 0.050 0,053 0,055 11 Mt. i-ichardfi 0.013 0.033 0.035 0.045 0.032 0.048 0.037 0.045 0.045 0,050 0.050 12 D. apiculata - 0.037 0.035 0.045 0.030 0.045 0.036 0.042 0.041 0.046 0.046 13 Z. intermedius 49 - 0.009 0.031 0.034 0.024 0.018 0,023 0.023 0.026 0,026 14 2. reinwardi i 46 12 - 0.030 0.036 0.021 0,014 0.019 0.020 0.023 0.021 15 2. pungens 59 41 39 - 0,037 0.036 0.027 0,032 0.033 0,036 0,035 16 2, obtusifolius 39 45 47 49 - 0.042 0.035 0.037 0.037 0,042 0,042 17 0. lyellii 60 31 18 47 55 - 0.015 0.017 0,009 0.012 0,009 18 0. obtusifolium 47 24 18 35 46 20 - 0.011 0.017 0.020 0,017 19 0. anomalum 56 31 25 42 49 22 15 - 0,017 0.018 0.017 20 CI. obtusiuscula 54 30 36 49 49 12 12 23 - 0.011 0.009 21 U. lutea 60 34 30 47 55 16 36 24 14 - 0,005 22 B. perichaetialis 60 34 20 46 55 12 11 32 12 6 - B 4 '(B = Bryodixonia; C = Cardotielle; D = Desmotheca; G = Groutidla; M = Macrocorna; Mt = ip" ~acromitrium;S = Schlotheimia; O = Orthotrichum; U = Ulota; Z = Zygodon) Chapter five: 87

Literature cited Anderson, L. E., and & Cnun 1958. Cytotaxonomic smdies on mosses of the Canadian Rocky Mountains. Bulletin of the Natio~iMuseum of Canada 160: 1-89. Andersson, I., S. Knight, G. Schneider, Y. Lindqvist, T. Lundqvist, C.4. Briinden, and G. H. Lorimer. 1989. Crystal SUU- of the active site of ribulose-bisphosphate carboxylase. Nature 337: 229-234. Barker, N. P., H. P. Luider, and E. H. Harley. 1995. Polyphyly of Anindiaoideae (Poaceae): Evidence from rbcL sequence data. S'stmtcltic Botany 20: 423435. Baum, D. A., K. J. Sytsma, and P. C. Hoch. 1994. A phylogenetic analysis of Epilobiwn (Onagraceae) based on nuclear ribosomal DNA sequeaces. SystekcBotany 19: 363-388. Blomquist, H. L., and L. L. Robertson. 194 1. The development of the peristome in Aulacomnium heterostichum. Bulletin of the Torrey Botanical Club 68: 569-584. Bremer, K. 1988. The iimits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 46: 1801 - 18 1% Britten, R. J. 1986. Rates of DNA sequence evolution differ between taxonomie groups. Science 23 1: 1393-1398. Brotherus, V. F. 1901-09. Musci (Laubmoose). m. Unteddasse Bryales: KI Spezieiler Teil. In A. Engler's and K. Prantl [eds.], Die Natürlichen Pflanzenf'en 1 (3, I, ïI), 277- 1246. W. Engeùnaon, Leipzig. 1925-25. Musci (Laubmoose) III. Unterklasse Bryales: II Spezieller Teil. In A. Engkr's and K. Prantl [eds.], Die Natürlichen Wanzeofamilicn, 2nd. editioa, vol. 1 1, 1-542. W. Engelmann, Leipzig. Buck, W. R 1991 Contributions to the moss flora of Guyana Memoirs of the New Botanical Garden 64: 1û4- 196. ,and D. H. Vitt. 1986. Suggestions for a new familial classification of pleurowpous mosses. Taon 35: 2160. Capesius, 1. 1995. A molecular phylogeny of Bryophytes based on ihe nuclear encoded 18s rRNA genes. Journal of Plant Physiology 146: 59-63. Chase. M. W., D. E. Soltis, R. G. Olmstead, D. Morgan, D. H. Les, B. D. Mishler, M. R. Duvall, R. A. Rice, H. G. Hills, Y.-L.Qiu, ICA. Kron, J.H. Rettig, E. Conti, J. D. PalmerJ. R. Manhart, K.J. Systma, H. J. Michaels, J. W.Kress, K. G. Karol, W. D. Clark, M. Hedrén, B. S. Gaut, R. K. Jansen, K. 4.Kim, C. F. Wimpee, J. F. Smith, G. R. Fumier, S. H. Strauss, Q.-Y. Xiang, G. M. Plunkett, P. S. Soltis, S. M. Swensen, S. E. Wiiliams, P.A. Gadek, C. J. Quian, L.E. Eguiarte, E. Golenberg, G. H. Lem Ir$. Chapter five: 88

W. Graham, S. C. H. Barrett, S. Dayanandan, and V. A Albert. 1993. Phylogenetics of seed plants: an anaiysis of nucleotide sequences hmthe plastid gene rbcL. Annuls ofthe Missouri Boranical Garden. 80: 528-580. Churchill, S. P., and E. L. Linares C. 1995. Prodromus Bryologiae Novdiranatensis. Introduction a la Rora de Musgos de Colombia. Parte II. Bibliotheca "Jose Jeronimo Triam" 12: 455-924. Conti, E., A. Fischbach, and KJ. Sytsma. 1993. Tnirelationships in Onagraceae: impLications hmrbcL sequence data, Annufs of the Missouri Botmical Garden 80: 672-685. Crosby. M. R. 1980. The diversity and relationships of mosses. In R J. Taylor and A. E. Leviton [eds.], The Mosses of North America, 1 15-129. Caiifotnia Academy of Science, San Francisco, California, Cm,H. 1972. A taxonomie account of the Erpodiaceae. Nova Hedwigia 23: 20 1-224. 1987. Leratia and its place in the ûrthotrichaceae, subfdy Macromitrioideae. Mmoirs of the New York Botunical Garden 45: 602606. De Luna, E. 1995. The ckumscription and phylogenetic relationships of the Hedwigiaceae (Musci). Systenzutic Botany 20: 347-373. Dixon, H. N. 1932. Classification of mosses. In F. Verdoorn [ed.], Manuai of Bryology, 397-4 12. Martius Nijhoff, The Hague. Donoghue, M. J., R G. Olmstead, J. F. Smith, and J. D. Palmer. 1992. Phylogenetic relationships of Dipsacales based on rbcL sequences. Mals ofthe MLrsout+ Botanical Garden 79: 333-345. Doyle, J. J. 1992. Gene mes and species mes: molecular systematics as oneîharacter taxonomy. Systemutic Bormy 17: 144- 163. ,and J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of khleaf tissue. Phytochemical Bulletin 19: 1 1-15. During, H. J. 1977. A taxonomical revision of the Garovaglioideae (Pterobryaceae, Musci). Bryophytom Bibliotheca 12: 1-244. Edwards. S. R. 1979. Taxonomie implications of ceii patterns in haplolepideous moss peristomes. In G. S. C. Clarke and J. G. Ducken [eds.], Bryophyte Systematics, 3 17- 346, Academic Press, London. 1984. Homologies and inter-relations of moss peristomes. In R. M. Schuster [ed.], New Manuai of BryoIogy, vol. 2,658-695. Hattori Botanical Laboratory. Nic ban. Chapter five: 89

Fds,J. S. 1972. Estimating phylogenetic trees fiom distance matrix. The American Naturalist 106: 645-668. Felsenstein, 1. 1985. Conndcnce lirnits on phylogenies: An appmach ushg the bootstrap. Evoluron 39: 783-791. Fitch, W. M. 1971. Toward defining the course of evolution: Minimum change for a specifk trec topology. Systematic Zoology 20: 406416. Fleischer, M. 1920. Natütiisches System der Laubmoose. Hehigia 61: 390-400. Frey, W. 1977. New Vorstellungen über die Verwmdschaftengruppen und die Stammesgeschichte der Laubmoose. In W. Frey, H. Hurka, and F. Obervwinkler [eds.], Beitage zur Biologie der aiederen Wanzen, 117-1 39. Stuttgah , 1990. Genoelemente ps-angiospennen Urspmgs bey Bryophyten. Botanischer Jahresbericht der Systematik 11 1: 433456. ,J.-P. Frahm, E. Fischer, and W.Lobin. 1995. Die Mws- und Fampflanzen Europas. Kleine Kryptogamenfîora Band IV XII + 426 p. Gustav Fischer Verlag, Stuttgart. Hasebe, M., P.G. Wolf, K.M. Ryer, K. Ueda, M. Ito, R. Sano, G. J. Gastony, J. Yokoyarna, J.R. Manhart, N. Murakami, E. H. Craue, C. H. Hader, and Hauk, W.D. 1995. Fern phylogeny based on rbcL nucleotide sequences. American fem Journal 85: 134-181. Haufier, C. H., and T. A. Ranker. 1995. RbcL sequeoces provide phylogenetic insights among sister species of the fem genus Polypodium. American Fem Jouml85: 36 1- 374. Hedderson, T. A., R. C. Chapman, and W. R. Rootes. 1996. Phylogenetic relationships of bryophytes infemd hmnucleartncoded rRNA gene sequences. Plant System tic Md Evolution 200: 2 13-224. HedenaS, L. 1994. The basal pleufocarpous diplolepideous mosses-A cladistic approach. The Brydogisî 97: 225-243. Hendy, M. D., and 0. Penny. 1989. A framework for the quantitative study of evolutionary trees. Systematic ZooIogy 38: 297-309. HUS,D. M.,and J. P. Huelsenbeck. 1992. Signal, noise, and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189-195. ,and J. JBuiI. 1993. An empirical test for bootstrapping as a method for assessing confidencc in phylogenetic analysis. Systenatic Biology 42: 182-192. Ka11ersj0, M., J.S. Famis, A. G. Kluge, and C. Bult 1992. Skewness and permutation. Cladistic 8: 275-287. Chapter £ive: 90

Kass, E., and M. Wink. 1995. Molecular phylogeny of the Papilionoideae (family Leguminosae): RbcL gene sequences venus chexnical iaxonomy. Botunica Acta 108: 149-162. Kim, J., F. I. Robif..and R R Sokal. 1993. The accuracy ofphylogenetic estimation using the neighbor-joiaiag mthod. Evolution 47: 471-486. Koponen, A. 1977. The peristome and spores in Splachnaceac and theirs evolutionary and systematic signincance. Bryophyom Bibüotheca 13: 535-567. 1982. On the stnictwe and function of the peristome in Splachnaceae. Joumai of the Huttori Botmicul Laborutory 53: 73-98. 1983. Studies on the generic concept in the classification of the moss family Splachnaceae. Publicationfiom the Department of Botany, University of Helsinki 1 1: 1-48. Kumar, S., K. Tamura, and M. Nei. 1993. MEGA: Molecular Evolutionary Genetic Analysis, version 1.O. Cornputer program distributed by Pe~sylvaniaS tate University, University Park, Pennsylvania. LaFarge-Englanà, C. 1996. Growth-fonn, braaching piittem, and penchaetid position in mosses: Cladocarpy and pleurocarpy redefmed. The Bryologist 99: 170- 186. Lewinsky, 1. 1976. On the systematic position of Amphidium Schimp. Lindbergia 3: 227-23 1. 1977. The genus Onhotrichum. Morphological snidies and evolutionary remarks. Joumal of the Hattori Botanical Laboratory 43: 3 1-6 1. 1989. Does the orthotrichaceous type of peristome exist? A study of peristome evolution in the genus Orthomchwn Hedw. with possible denvation of the haplolepideous peristome. Joumal of the Hattori Botanical Laborarory 67: 335-363. 1990. Zygodon Hook. & Tayl. in Australasia: a taxonomie revision including SEM studies of peristomes. Lindbergh 15: 109439. 1993. A synopsis of the genus Orthotnchwn Hedw. (Musci, Orthotrichaceae). Bryobrothera 2: 1-59. ,and M.R. Crosby 1996. Orth~rni~umtuberculatum (Orthotrichaceae), a new genus and species fiom Guizhou, China. Novon 6: 1-5. Maddison, D. R 1991. The discovery and importance of multiple islaads of most- parsimonious trees. Systematic Zoology 40: 3 15-328. Manhart, J. R 1994. Phylogenetic analysis of green plant rbcL sequences. Moleculur Phylogenetics mrd Evolution 3: 1 14- 127. Chapter five: 9 1

Meusel, H. 1935. Wuchsformen und Wuchstypen der europaischen Laubmoose. Nova Acta Leopoidina Neue Folge 3: 123-277. Mishlef, B. D., P. H. Tbrall. I. S. Hopple, Jr., E. De Luna, and R. Viigaiys. 1992. A molecuiar approach to the phylogeny of bryophytes: Cladistic andysis of chloroplast- encoded 16s and 23s ribosomal RNA genes. The Bryufogist95: 172- 180. ,L. A. Lewis, M. A. Buchhehn, K. S. Remigiia, D, JT Garbary, C. F. Delwiche, F. W. Zechman, T. S. btz,and R. L. Chapman. 1994. Phylogenetic relationships of the "green algae" and "bryophyte". Annuls of the Missouri Botunical Garden 81: 451483. Miîten, W. 1869. Musci Austro-americani. Jouml of the Linnoon Society of Botuny 12: 1-659. Morgan, D. R., and D. E. Soltis. 1993. Phylogenetic relaaonships arnong members of S-gaceae sensu lato based on rbcL sequence data. Annals of the Missouri Botmical Garden 80: 63 1-660. Nishimura, NaTand K Watanabe. 1992. SEM obsemations on the penstome teeth of mosses, 2. Isobryales. nie Natuml Environmental Science Research 5: 83-101. Olmstead R. G., and J. A. Sweere. 1994. Combining data in phylogenetic systematics: An empirical approach using themolecular data sets in the Solanaceae. Systematic Biology 43: 467-48 1. mer, K. M., A. R. Smith, and J. E. Skog. 1995. Phylogenetic relationships of extant fersn based on evidence from morphology and rbcL sequences. Amerkm Fern Journal 85: 205-282. Reimers, H. 1954. Bryoptiyfa. In H. Melchior and E. Wedermam [eds.], A. Engler's Syllabus der Wanzenfden, 12th. edition, 218-268. Gebctirder Bomtriigger, Berlin. Robinson, H. 197 1. A revised classification for the orders and families of mosses. Phytologia 2 1: 289-293. Russo. C. A. M., N. Takezaki, and M. Nei. 1996. Efficiencies of dinercnt geaes and different tree-building methods in recovering a knwon vertebrate phylogeny. Molecular Biology adEvolurion 13: 525-536. Sainsbury, G. O. K. 1945. New and criticai species of New Zealand mosses. Transaction of the Royal Society of New Zealand 75: 169-185. Saitou. N., and M. Nei. 1987. The Neighbor-joining method: a new method for reconstructing phylogenetic mes. Molecular Biology and Evolution 4: 406-425. Chapter five: 92

Sang. T. 1995. New measurements of distribution of homoplasy and reliability of parsimonious cladograms. Tmon 44: 77-82. Sanger, F., S. Niclclen, and A. R Coulson. 1977. DNA sequencing with chain- terminahg inhibitors. Proceedings of the National Academy of Sciences, USA 74: 5463-5467. Schwartz, O. M. The development of the peristome-formiog layers in the Funariaceae. Intematiod Joumal of PhtScience 155: 640-657. Shaw, f. 1985. Peristome structure in the Orthotrichaceae. louml of the Naftori Botanical Laboratory 60: 119-136. ,and J. R. Rohrer. 1984. Endostomid architecture in diplolepideous mosses. Journal of the Hatton Botanical Liboratory 57: 41-61. ,and H. Cnun. 1984. Peristome homology in Mielichhoferia and a taxonomic account of North American species. Joumal of the Hattori Botanical Laborutory 57: 363-38 1. ,B. D. Mishler, and L. E. Anderson. 1989. Peristome development in mosses in relation to systematics and evolution. IV. Haplolepidae: Ditrichaceae and Dicranaceae. The Bryologist 92: 3 14-325. ,L. E. Anderson, and B. D. Mishlec. 1989. Peristome development in mosses in dation to systematics and evolution. m.Fumria hygrometri'ca, ~rym pseudocapillare. and B. bicolor. Systematic Botany 14: 24-36. Soltis, D. E., and R K. Kuzoff. 1995. Discordance between nuclear and chloroplast phylogeneis in the Heuchera group (Saxifhgaceae). Evotution 49: 727-742. Swofford, D. L. 1993. PAUP: Phylogenetic Analysis Using Parsimony, version 3.1.1. Compter Rogram ditributed by the nlinois Naturai History Swey, Champaign, IUSnois. van Rooy, J.. and A. E. van Wyk. 1992. A conspectus of the subfamily Macromitrioideae (Bryopsida: ûrthoaichaceae) in Southern Afnca. nie Bryologist 95: 205-215. Vitt, D. H. 1970. The fdyOrthotricbceae (Musci) in North Amenca, north of Mexico. Ph.D. thesis. University of Michigan, AM Harbour. 1972. A monograph of the genus Drununondia. Canadiun Joumal ofBaany 50: 1191-1208. 1973. A revision of the genus OrthomChum in North America, nonh of Mexico. Bryophytorum Bibliotheca 1: 1-208. Chapter five: 93

1980 (1981). The genus Macroconur 1. Typification of names and taxonomy of the species. The Bryologist 83: 405-436. 198la Adaptive modes of the moss sporophyte. The Bryologist 84: 166-186. 198lb. The genera Lcio&um and Cardotiellu gen. nov. (Orthotrichaceae). Journal of the Hattori Botanicd Laboratoty 49: 93- 113. 1982a. Sphagnopsida and Bryopsida In S. P. Parker Ed.] Synopsis and Classification of Living Organisms, vol. 1,307-336. McGraw-Hill, New-York. 1982b. The genera of the ûrthotrichaceae. Beihefr u< Nova Hedwigk 7 1: 26 1- 268. 1982c. Populational variation and speciation in austral mosses. Journal of the Hattori Botmical Laboratory 52: 153-159. 1984. Classification of the Bryopsida. In R M. Schuster [ed.], New Manuai of Bryology ,Vol. 2. Hat& Botanical Laboratory, Nicbinan, Japan. 1990. Desmotheca (Orthotrichaceae): Gondwanan fragmentation and the origin of a Southeast Asian genus. Tropical Bryology 3: 79-88. 1993. Onhotrichaceae. In A. J. Sharp, H. A. Cnim, and P. M. Eckel [eds.], The Moss Flora of Mexico, 590-657. New York Botanical Garden, New York. ,and H. P. Ramsay. 1985. The Macr~mi~umcomplex in Australasia (Oahotrichaceae: Bryopsida). Part II. Distribution, ecology, and paleogeography. Journal of the Hanon' Botmicaal Laboratory 59: 453-468. ,T. Koponen. and D. H. Noms. 1993. Bryophyte flora of the Huon Peninsula, Papua New Guinea. Lm. Uluta and Schlotheimiu (ûrthotrichaceae, Musci). Acta Botanica Fennica: 148: 5-25. , , 1995. Bryophyte flora of the Huon Peninsula, Papua New Guinea. LV. Desmotheca, Grouriella, Macrocoma and MacromirnUm (Orthorichaceae, Musci). Acta Butanica Fennica: 154: 1-94. ,B. Goffi.net, and T.A. Hedderson. 1997. The ordinal classification of the mosses: Questions and answers for the 1990's. In: N.W. Ashton, LW. Bates, and J.G. Duckett. Bryology for the Second Century. (in press). Walther, K. 1983. Bryophytina. Laubmoose. In J. Gerloff, and J. Poelt [eds.] A. Engler's Syliabus der Pflanzenfamiiien. Ad. 13, V(2): 1- 108. Gebrüder Borntriiger, Berlin. Waters, D. A., M. A. Bucbheim, R. A. Dewey, and R. C. Chapman. 1992. Preliminary inferences of the phylogeny of bryophytes fkom nuclear-encoded ribosomal RNA sequences. American Journal of Botany 79: 459-466. Chapter five: 94

Welch, W. H. 1943. The systematic position of the genera Wardia,Hydropogon. and Hydrupogonella. The Bryologisr 46: 2546. Wolf, P. G.. P. S. Soltis, and D. E. Soltis. 1994. Phylogenetic relationsbips of Dcnnstaedtioid fems: Evidence from rbcL sequences. Moleculor Phylogenetics and Evolution 3: 383-392. Zander, R H. 1993. Genera of the Pottiaceae: Moses of harsh envhnments. Bulletin of the Bu#ialo Society of Natural Sciences 32: 1-378. Chapter six: 95

Chapter 51.

CiFcuisiscriptioa, aad phylogenetic trends in the mbf'of the OrtboMchaœae inferred thin morphobgy

The ûrthotrichaceae are a cosmopoiitaa famüy, with nearLy 600 species (Vitt 1982a). Typical peristomate Onhotrichaceae sbue 1) a double peristome of aitemating exostome teeth and endostome segments, 2) the la& of diaas well as additional divisions in the PL,and 3) a heavily thickened exostomid OPL (Lelvinsky 1989, Shaw 1985, Vitt 1982a). Gametophytic feanires alone can be misleadhg in addrrssing systematic afnnities of gymnostomous taxa or those taxa with a ~ducedperistome (e-g., Amphidhm and Drummondiu, see chapter 5). but most if not aiI ûrthooichaceae can nevertheless be characterized by the foiiowing combination of gametophytic characters: cauline central strand lacking, upper laminal celis isodiametric and papiliose, alar cells not differentiated fkom innet basal cells, and perichaetia terminal on stem or lateral branches. The family now includes 20 genera (chapter two, three, and five) that are distributed among three subfamiles: the Orthotrichoideae, the Zygodontoideae (Schimp.) Broth., and the Macromitrioideae Broth. These subfamilies have traditionaily been deked based on the position of the female garnetaagia (acrocarpy versus cladocarpy) and the shape of the calyptrae (cucullate versus mitrate; Brotherus 1925, Vitt 1972). Acrocarpous taxa with mitrate and typically plicate caiyptrae. namely Bryudixronia Sainsb., Muelleriella Dusén, Orthotrrèhum Hedw., Orthornitrhm Lewinsiry-Haapasaari & Crosby, Stoneobrywn Noms & Robinson, and Uloto Mohr., as well as Pleur~rtho~chumwith its smooth cucullate caiyptrae, have traditionaily been included in the Orthotrichoideae (Brotherus 1925, Vin 1982a Lewhsky 1994, Lewinsky-Haapasaati and Crosby 1996). The Macromitrioideae (i.e.. Card&& Vitt, Ceuthotheca Lewinsky, Florschuetziella Vitt, Groutiella Cnun & Steere. Lei~rni~umMitt., Leratiu Broth., Macrocoma [C. MU.] Grout, Macromitrum Brid., and Schlotheirnia Brid.) share features of the calyptrae with the former subfamily, but differ fiom the Orthotrichoideae in their gametangia that develop fkom the apical meristem of lateral branches (except Leratia, see Cnim 1987). Demtheca Lindb. was segregated in its own subfamily by Brotherus (1925) and Cnun ( 1987). but retained within the Macromitrioideae by Vitt (1990). The Zygodontoideae (Leptodontiupsis Broth.. Sten~rni~um(Mitt.) Broth., Zygodon Hook. & Tayl.) differ from the other subfamiiies (except Pleur~rtho~chunt,and Macromitnùm sect- Reverba~nBuck) by tûeir invariably smooth and cuculiate caiyptrae. Vin (1982a) pmposed a specuiative phylogenetic anangement in which the Zygodontoideae are basal to a dichotorny between the Orthotrichoideae and the Macromitrioideae. Such an evolutiolliiry scenario is baseci, even though this is not expiicitly stateâ, upon a acrwarpous rmcestor with cucullate calyptrae. Moleculas data. (chapter five), support the bypothesis of an acfocarpous ancestor rather than pleurocarpous affinities of the famüy (or portion thereof) as suggested by Hedeniis (1994). Furthemore, interpretation of the gene-based phylogeny, leads to the hypothesis that this ancestor to the Orihotrichaceae, as weIl as the primitive taxa within the family, had smooth lamina1 celis. Alternative hypotheses to Vitt's (1972) classification have emerged with ongoing taxonomie snidies, but were never fodypcoposed. In 1979, Vitt suggested that Florsch~e~ella,Macrocorna and Leiomitrium (then including species now placed in Cardotiella) may be closely related and represent a distinct taxon. More recently, Vitt, Koponen and Noms (1993) considered a separation of Schlutheimia and Cardotiella fiom the Macromitrioideae on the basis of the lobate versus the entire to lacerate calyptrae (see Vitt 198 Ia). In connast to these hypotheses which share a monophyletic concept of the Orthotrichaceae, Churchill and Linares (1995) felt that the Orthotrichaceae sensu Vitt (1984) were defïned only by plesiomorphic character-states and that the acrocarpous (Orthotnchoideae and Zygodontoideae) and the cladocarpous (Macromitrioideae) lineages did not share a common evolutionary history. They thenefore proposed recognuing the Macmmitrioideae as a distinct family, the Macromitriaceae S.P. ChurchiU. De Luna (1995), too, considered the Orthotnchaceae to represent an evolutionary grade, with the Macromitrioideae king more closely related to the Leucodontales, but bis study did not include many peristome characters, which may prove critical when addressing the monophyly of the Orthotrichaceae or the Orthotrichales. A phylogenetic nconstmction of aahrodontous mosses based on 18s gene sequences (Hedderson et al. unpubl.), however, suggests that the acrocarpous and cladocarpous heages of the Orthotrichaceae are closely related and form a monophyletic family. RbcL sequence data are congruent with a monophyletic concept of the family, albeit the most parsimonious trees should be lengthened by one step (see discussion chapter five). The molecular phyiogeny based on the chioroplast gene does, however, agree with Churchill and Linares' (1995) hypothesis regardhg the monophyly of a combined Zygodontoideae and Orthotrichoideae clade, sister to the Macromitrioideae. Chapter six: 97

Only a few phyiogenetic relatiomhips hypothesized witbin the Bryopsida have been proposed or tested using modem phylogenetic appraaches with morphological data (e.g., Koponen 1968, Vitt 1971, Churchill 1981), and only recently has cladistic rationale become of cornmon practice in morphology-based evolutionary stuclies of mosses (De Luna 1995, He&& 1994,1995, Zander 1993, Vitt 1995). This study is the nrSt attempt to circumsctibe supragenenc taxa in the Orihotrichaceae and identify their phylogenetic relationships based on a cladistic analysis of morphological characters. The Orthotrichaceae are a large family and, except for OrthotrrChum (Lcwiasky 1993) and Zygodon (Malta 1926). speciose genera such as MacrmibiMt, Schlotheimia, and Ulota have not yet ken criticaily monographed and the evolutionary history of their species remains obscure. Although gehc~lationships WU be addressed, this study does not pretend to solve the phylogeny within the Orihotrichaceae. Instead it should be viewed as 1) a study examining the suitability of the characters and the states commody used when addressing systematic relationships in the ûrthotrichaceae, and 2) an attempt at defining broad phylogenetic relationships that may serve as a basis for future phylogenetic reconstructions withgenera or groups of putatively related genera.

Msteriai and Metbods

Selection of iun. Following the amended cin:umscription proposed in chapters two, theand five, 20 genera, now considered the terminal taxa of the Orthotrichaceae. are included in this study (Table 6.1). The monophyly of some genera, such as Macromitrium and Zygodon, bas been questioned on molecular grounds (chapter five). Instead of compartmentaliziag such taxa for the present analysis (Mishlet 1994), severai species thought to represent distinct infiageneric evolutionaty trends have ken included for Macrocomu, Macromimmum,Onhotrichum. Schlotheimia, Ulota, and Zygodon (Table 6.1). The range of taxa used for the present analysis was aügned merwith that included in the molecular analysis (chapter five) in order to have overlapping taon samples. Because sporophytes are unknown for C. quinquejlaria {Vitt 198 la), a second species of Cardotiella, C. elimbuta (?M.)Gof5.net (see chapter tbree) was added. in total, 34 taxa representing 20 orthoQichaceous genera are included in the analysis. Character states were determined by direct examination of herbarium material, except when available matenal was incornpiete, in which case character-states were-identifiecl based on pubiished descriptions. The one collection of the recently desfnbed monotypic Chapter six: 98 genus Orthomim*umwas not studied and all entries are based on the protologue (Lewiasky-Haapassari & Crosby 1996). The systematic relationships of the ûrthotrichales within the diplolepideous mosses, and particularly with regatd to the Bryales sensu ho,remain umesolved (chapter five). Analyses of moldardata frwi either the chIoroplast gene rbd. (chapter five) or the nuclear gene encoding for the 18s rRNA (Hedderson et al. unpubl.), have provided evidence for the basal position of the Fmariales among arthmdontous mosses (Vitt, Goffinet, and Hed&mn 1997). The Orthotrichales belong to a clade also including the Splachnales and the Bryales sensu lato. Sequence data of the 18s gene strongly suggest that the Orthoûicbales are closely related to the Bryaies sensu ho. Whether the ciüate mosses form an evolutionary grade or a nadgroup, and whether the ûrthotrichales are sister to all Bryales or only a portion of these, needs to be Meraddressed critically. In either case, molecuiar evidence (particularly hmthe 18s; Heddersson et ai. unpubl.) combined with general evolutionary trends in morphological charactea such as branching pattern and distribution of perichaetia (La Farge-England 1996) suggest that the Orthotrichaceae are derived fiom an acrocarpous (see discussion in chapter 5) and not a pleurocarpous ancestor (as suggested by Hedeniis 1994). A single representative of the Funariales, Splachnales, and Bryales has ken included in the present analysis (Table 6.1). The Orthouichales are composed of two families, the Orthotrichaceae and the Helicophyllaceae. The Helicophyllaceae are monotypic, and have dubiu uncertain affinities with the Orthotrichaceae: both families share a priori only one apotypic character, namely papillose laminal cells. Helicophyllm is gymaostomous, and addressing phylogenetic relationships of apenstomate taxa based on gametophytic characters only can be misleaâing (see De Luna 1995). The strongly dimorphic, if not trimorphic, leaves found on a single axis (or module) are in sharp contrast with isomorphic leaves in the Orthotrichaceae (in the Orthotrichaceae, leaf dimorphism applies to leaves of distinct modules; see character 5 below). At present the Helicophyllaceae are best considered outside the ûrthotrichaceae, and should therefore not be included in the pnsent analysis. Representatives of the haplolepideae, a group sister to the Splachnales-Onhotnchales- Bryales clade, were not included in the present anaiysis. The peristome of the haplolepideae is an opposite peristome (Vitt, GofE.net & Hedderson 1997): the exostome is typically missing in the haplolepideae, but if deveiopmental data were extrapolated to a mature state, the exostome teeth would indeed be opposite the endostomial segments. Chapter six: 99

Such peristome architecture û indicative of an early derivation in the evolution of arthrodontous mosses (Vitt, Goffinet, and Hedderson 1997), a hypotbesis supported by analyses of nucleotide sequences of both the nuclear 18s (Hedderson et al. unpubl.) and the plastid gene rbcL (chapter five). Both studies suggest that the haplolepideae are deriveà fiom a Funaria--type ancestor and are sister to a clade composed of the Splachnales, Orthotrichales, and BryaIes smu luto. The relationships within the haplolepideae are poorly understood, and mne of the taxa exhibits clear gametophytic plesiomorphks similar to those fomd in the Funariales and the Splachnales. Instead maay ma(e.g., Grimmiaceae and Pottiaceae) are clearly homoplastic with the ûrthotrichaceae in gametophytic features such as thick-walied cells, and large mitrate and plicate calyptrae. Although all putative sister groups should ideally be included in a phylogenetic analysis, the haplolepideae were expected to introduce phylogenetic noise, rather than contribute to polaDPng character-states, and were not considered in this study. Data analysis. Fitch parsimony analyses were pedormed with PAUP (version 3.1, Swofford 1993) on a Power MacIntosh 7200/90 using the heuristic search with the following options in effect: keep al1 characters, multistate taxa interpreted as polymorphic, steepest decent, and collapse zero length branches. Funaria hygr~me~ca, Bra~hyni~umjamesonii, and Mnium thomsonii wen used simultaneously in the aaalysis, but the trees were rwted via the outgroup methoci (Watmus and Wheeler 198 1) with Fwraria as the sole outgroup. None of the outgroups included exhibits plesiomorphic states (sensu Miller 1979 and Vitt 1984; see also Shaw and Rohrer 1984) for al1 characters. Character transformations were therefore left reversible, multistate cbaracten unordered, and polyphyletic origins of derived character states allowed. Furthemore, character-states of binary or multistate characters were given the same weight. A preliminary search was performed with 100 random additions, foilowed by a second analysis of 500 random replicates but with swapping restricted to trees at the most one step longer than the most parsimonious trees (of length X') found in the preliminary search and on a maximum of 25 trees that are X+2 steps long (see Maddison 1991, Pryer et al. 1995). A strict and a 50% majority detree were constructed using PAUP. Consistency and retention indexes, as well as f-value (Farris 1972) were calculated for all most parsimoaious trees using PAUP. Character evolution was reconstructed using the accelerated transformation criterion in MacClade version 3 (Maddison and Maddison 1992). Relative support for branches was determined by decay analysis (Bremer 1988, Donoghue et al. 1992) foilowing the method developed by Baum, Sytsma, and Hoch Chapter six: 100

(1994) and using LOO replicates in the heuristic search. Enforcing topoiogical constraints in heucistic searches, using the same set of options as in the above analysis, ailowed for the costs of alternative hypotheses to be determineci.

Character definition and coding. Sixty-eight characters were initially examineci for ail taxa, and are reviewed below. Invariable, and autapomorphic characters were retained for the anaiysis, but 14 characters were excluded because of the difficulty in either observing the actual character or denning distinct character-states. The ma* is presented in Table 6.2. 1. Stem orientation: orthotropic (O), plagiotmpic (1). The life fom of a moss results fiom the cumulative effect of its pwthform, branchhg pattern and the position of the female gametangia (La Farge-England 19%). Generai ands observed among mosses suggest that certain combinations of character states of these three charaftea appear more commonly than others, and therefore that these characters may not be totdy independent. Closer examination reveals however that aü 12 possible combinations of characters states are found in mosses (La Farge-England 199@, and therefore di three characters should be considered independent and be included in the analysis (see characters 22 and 23). In the Orthotrichaceae, the stems are either orthotropic (e.g., Zygodon) or plagiotropic (e-g., Macrornitrbn). The basal portion of the stems of O. lyeUii or P. chilenre. become pmstrate as the stem grows, but the distal portion of this main axis stiu grows upright. Orthotrichum lyeliii is therefore considered orthotropic. 2. Anatomy of axial cadine tek: celis differentiated into a ciuster of narrow, thin- walled ceils that are often collapsed, leaving a central cavity in the stem (O), or ceils not differentiated (1). In ail three outgroup taxa, Fwarià, Bru~hymi~um,and Mn&, the central ceiis are weii differentiated hmthe smounding parenchymatous ceils, wbereas in the Orthotrichaceae no such differentiation has been observed (Héôant 1977; but see Lewinsky 11977 1). 3. Anatomy of outermost disd tbe stem: ceils thin-walled, fomllng a distinct hyalodermis (O), or hyalodennis not differentiated (1). In Funaria and Brachyminiwn, the outer cells are dimorphic: ceiis immediately adjacent to the parenchyma ceils have thick, red ceii walls; the cells outside to this differentiated cortex have clear and thin walls, and form a unistratose layer called a hyalodermis. Such differentiation is known from a variety of mosses but is lacking in the Orthotrichaceae. Chapter six: 10 1

4. Anatomy of innet cortical cells: thin-wded (O), thick-walled (1) or multilayered (2). The parenchyma cells of the stems are thin-walled in Fm&, Brc~chymirnüm,and Mnium. In most Orthotriochaceae, the walls are thick, i.e., in Iight microscopy, the middle lamellae separating two adjacent cells is clearly visible. In some taxa, such as in many miesof Uha, the ceil wall appears to be composed of many, concentric layes 5. Leaf morpho1ogy: stem and branch leaves identical (O), or stem and branch leaves dimorphic. in taxa with onhoa~pic,and sympodially branching stems, the leaves of the branches are identical to those of the preceding module (except for the juvenile leaves; . Mishler & De Luna 1991, and La Farge-England 1996). In taxa with monopodially branching stems, the branch leaves may be identicai to (Florschuetziella) or differ from the stem leaves (Macrominlm). In the latter case, the stem leaves are always smaiier than the bmch leaves. The bmch leaves of taxa with dimorphic leaves, such as Ma~romi~m,are similar to the stem leaves of orthotropic, sympodially branching taxa with monomorphic leaves, such as Onhotrichm. Dimorphism is here considered to be the derived con&tion, suggesthg that the stem Ieaves differentiated fiom leaves that were similar if not identical to the branch leaves. Consequently compatisons of celi shapes, omamentations, etc., between species with monomorphic and dimorphic leaves are based on stem and branch leaves, respectively. 6. Leaf base morphology: lower half of leaf more or less as wide as upper half (O), or leaf base wide and contracted into the linear upper lamina (1). The vegetative leaves of Bryodixonia, and many species of Ulota, are unique within the Orthotnfhaceae by their ovate leaf base, that is nmwed into an aciiminate upper lamina. 7. Leaf attachment: base of leaf not decumnt or decunencies composed of chlorophyliose cells (O), or decurrencies conspicuous, composed of infiated hyaline ceils (1). Whereas many taxa of the Orthotrichaceae can have decurrent leaves (e.g.. Zygodon, Malta 1926), decurrencies composed of several rows of hyaline, bulging. and sometimes papiiiose to tuberdate ceils, are only lmown fiom the pnus Cardotiella. 8. Anatomy of eosfs: guide cells differentiatted (O) or not differentiated (1). unknown (?). Kawai (1968) ncognized three basic layers in the anatomy of the costa: the adaxial (a), abaxial (b), and the median (c) layer. Based on the degree of differentiation of three celi layers and the pnsence of additional layen, Kawai (1968) distinguished between five types of costal anatomy. Kawai (1968) examined several taxa of the Orthotrichaceae. He interpreted the costa of Leratia, Orthutrichun, and Pleurorthotnchwn as homogenous, thus with no obvious dinerentiation between the layers (type-A), whereas in Macromiinum he considered the adaxial cells to be clearly distinct (type-B). Lewinsky Chapter six: 102

(1994) and Maita (1926) too described the costa of Orthotrz'chwn and Zygodon, respectively, as homogenous. In transverse section the differentiatioa of the costa is often ambiguous: ali cells are ihick-walled and while the width of the cells appears to decrease hmthe adaxial to the abaxial surfLace of the costa, the demarcation of distinct layers is often impossible. In suface view however, the adaxial and the abaxial ceils are often of distinctive shapes. Adaxial ceUs (excluding chlorophyllose cells, e-g., Muelleriella) are always (except for Leraîia) broadly rectangular, hyaline, and with moderately incrassate waiis, and tk end walls are always £lat By contrast the abaxial ceils (again excludhg chlorophyiiose cells) are typicaily linear and pointed (e.g., Macromitiwn), and rarely imgularly rectangular with fiai ends (Onhotrkhhwn; see character 10). The rectangular adaxial celis are here intqreted as guide ceiis. This character is not scored for Orthornitrim. In the prot~ioguethe anatomy is not described except for the absence of stereids (Lewinsky-Haapasaari and Crosby 1996). 9. Lamiad ceils covering the adaxial daceof the cm:present (O) or absent (1). Laminal cells (morphologically similar to the cells of the lamina, and thus chlorophyllose) cove~gthe ventrai surface of the costa have been found in all three outgroups. Within the ûrthotrichaceae, chlorophyllose cells covering the ventral surface of the costa, are known only nom two taxa ofMueleriella, including the putative primitive taxon M. crassifolia (Vin 1982b). 10. Sbnpe of abaxisl costal ceW. with flat ends (O), or pointed ends (1). unknown (?). In aN the ûrthotrichaceae examined, with the exception of the genus Urthonichum, the abaxial celis are rather elongate to Illieu, with pointed ends. These ceils are best considered substereids. In Orthotrichum. the abaxial ceils are imguiarly rectanguiar, with flat ends, and are hereinterpreted as distinct from substereids. In Orthominium, a monotypic genus with potential afnnities with either the ûrthotrichoideae or the Macromitrioideae. the costa is described as lacking stereids (Lewinsky-Haapasaari and Crosby 1996). This character-state would be congruent with that found in the Orthotrichum (but not aI.i other genra of the subfamily), but is incompatible with the anatomy of the costa in the Macromitrioideae. Since no material was seen, this character is better scored "unknown". 11. Abdsurface of the casta not or ody partiy covered by laminal ceUs (O), or laminal cells covering the abaxial surface of the costa almost to the base (1). unknown (?). In some taxa the abaxial layer of the costa is composed of chlorophyllose ceiis, often sunilar in shape and omamentation to the adjacent laminal cells (e.g., Florschuetziella). Laminal cells covering the costa are hown from various taxa in the Onhoeichaceae, Chapter six: 103

particularly in Zygodon (Malta 1926, Vin 1993). In Z reinwardtii, the laminal cells are reported to cover the apical portion of the costa, but the extent of thb laminal sheet is variable witbin a population (chapter two), but is always restricted to the upper half of the leaf. In Lerutia and Florschuemélla these laminal cells extend almost invariably to the base. In species of Orth~~chum,the abaxial layer of the costa is also composed of chlorophyllose ceh, albeit aot as conspicuously as in the former genera. 0rth0mi~um is scored as unlcnown. 12. Upper buniddi dace: flat (O), or buiging (1). Both penchai waiis of upper laminai ceUs in many Macromiüioideae are distinctly bent outward whereas in most taxa these walls are plane. 13. Ornamentation of upper iamhiceii: smooth (O), papaose (1), not applicable (-). The Orthotrichaceae have traditiondy been characterized by their papiilose laminal ceils, but some taxa such as Schlotheimiu or Zygdon sect. Bryoides are exclusively composed of smooth ceiied species. In Muelleriella the laminal are smooth, but the lamina is also bistratose. Because a link between the two characters could not be excluded, the omamentation is, for now, scored as "not applicable" in Muelleriella. 14. Morphology of paplllae "single" (O), or in pairs and apparentiy shortiy bifid, or c-shaped (1), not applicable (-). The papiliae in Zygodon sect. Zygodon appear xattered, whereas in Zygodon obtus~oliw,as well as in species of the Macromitrioideae sensu lato, the papiilae appear in pairs (not ail of them), and may appear branched or narrowly c-shaped (see also Malta 1926). UnlüEe the branched papiliae found in Onhomiehum however, they never are dis~ctlystalked and bifurcated. 15. Anatomy of basai laminai ceUs: cells with equally thickened anti- and penclinal wails (O), or cells with penclinal waUs thicker than anticlinal wails (l), or "not applicable" (0). In transverse section, the cens of Stenomihium and Pleuronhohichum differ f?om other taxa in the Orihotrichaceae by their celIs having pencha1 walls that are much thicker than the longitudinal anticlinai walis. AU other taxa examined have evedy thickened was. This charactet was scored as "not applicable" for taxa with smooth cells. 16. Anstomg of upper lamina1 edls: cells-wall thin (O), or thick (1). Upper laminal cells of the Orthotrichaceae are typically thick-walled (middle lamellae visible in light microscospy), w hereas Funaria and Bruchymim~umhave thin-w alied cells. 17. Shape of basai ceb: celis isodiametric (O), oblate (11, or elongate-rectangular (2). The laminal cells of the ûrthoaichaceae are typicaily differentiated into rectanguiar basal cells and isodiametric upper ceUs (e.g.. MacrominiMi). In some taxa this Chapter sù: IOJ

differentiation is not obvious and the lamina appears to be unifonaly composed of isodiamaic ceiis, except for the most pmximal ceils (e.g., Macrocorna). In Groutiellu and Curdotietla, the basal celis are not strictly isodiametric, but an rather oblate (with tranmersely elongate lumens). 18. Thickness 0flPminn: unistratose (O), or bistratose (1). In the Orthotrichaceae, bistratose laminas an typicai ody for Muellen'efla (Vitt 1976), but also occur scaîtered in other genera (Orthotrrchum haflii 19731, Macr~rni~wndiuphmrum Witt and Ramsay 19851). 19. Morphology of bdcek: ceils monomorphic (O), or ceils dimorphic. Whereas in most Orthotrichaceae, the basai cells are rathet dorm,except for larger juxtacostal cells in some taxa (e.g.. Groutiella), in Leptodontiopsis, and particularly in Pleuronhom'chum and Stenomihiun, the basal ceils are of two types with Som cells having very incrassate walls forming distinct yellowish longitudinal bands alternating with hyaüne, moderately tbick-wded cells. 20. Morpbobgy of basal mwghof the k.I: composed of undifferentiated cells (O), or ceils differentiated fodga mostly uniseriate band of hyaline cells that are larger than the adjacent inner cells, and have thinnet and bulging outer periclinal wak (1). These differentiated cells occur in many Macromitrioideae. 2 1. Morphology of basai matginai ceüs of the Lmlna: cells not Merentiated from inner lamina1 cells (O). or cells hyaline and rhombic (L), cells, hyaline, and quadrate, with their anticlinal walls much thicker than the periclinal wds(2), or cells hyaline, linear (3). Subquadrate marginal cells are typical of Ulmand B~oduonia.Linear, hyaline (hm-) marginal ceils extending hmthe base upward, are known only hmthe genus Groutiella. Other specics have elongate to hear maiginal cells, but these are always chlorophyilose and restricted to the upper portion of the lamina (e.g.. Macromitri*um ulophyllwn. Goninet 1993). These linear ceiis typical of Groutiella are different fiom the basal hyaline &ginal cells referred to in character 20: in Groutiella the outer most basal marginal ceils are rectanguiar to elongate. and with ihin, outer periciinal waiis, whereas the inner cells adjacent to this basal margin are bear and fom a band that extends beyond the uniseriate basal margin of large celis. 22. Position of peiichaeüa: tenninal on main axis (O), or terminal on lateral branches (1). The variation in distribu?ioîiof perichaetiai position has been examined extensively by La Farge-England (1996). Within the Orthotrichaceae, female gametangia (the perichaetia) are terminal either on the main axis (acrocarpy) or on lateral branches (cladocarpy). The main axis of acrocarpous taxa such as 0. anomlum may resume Chapter six: 105 growth by produchg 1 or 2 subapical innovations. These innovations are considered to be modules of the same hierarchial levels as the module fiom which they orighated, thus these branches are considered "new"stems, and the plant is therefore an acrwarp. In Ma~romit~iimr,the stem produces lateral innovations even though the apical menstem is still gmwing. These lateral branches are generally terminaicd by a gametangium; the apical cauline meristem does not produce temünal perichaetia and the plant is considered cladocarpous. in Macrocorna papiflosa, the stem produces lateral innovations, which produce terminai gametaagia, but the main axis, too, produces terminal gametangia; M. papillosa is thus considered polymorphic - both acrocarpous and cladocarpous. 23. Bmching of the main axis: sympodial(0) or monopodial (l),not applicable (?). The branching mode is independent hmthe distn'bution of perichaetia on the axes (Mishler and De Luna 199 1; La Farge-Engiand 1996). Sympodial branching refers to a succession of modules of the same hierarchiai level by subapical innovation; whereas monopodial branchhg implies the co~ectioaof modules of different ranls; a stem, module of the first degree, produces laterai branches, modules of the second degree. Strictly acrocarpous Orthoaichaceae aIways branch sympodially, whereas cladocarps have monopodially branching stems. In cladocarpous taxa, the secondary branches that are tenninated by a gametangium may resume growth by subapical innovations, and thus branch sympodiaiiy. ûrthotropic branches may become prostrate and fuoction as a stem, and thus branch monopodially. Cleady the "homology" of these axes needs to be addressed mer,but in the present study, the braaching pattern was compared between functional stems, regardless of whether the stem is a modified bmch or not (see also Men and Crosby 1986 for similar ambiguities in the Pterobryaceae). In Brachymifrium no branch development has been observed and this characters is scond as "unlcnown". 24. Sexual coiidition: monoicous (O), dioicous (1). or phyllodioicous (2). Individual plants cm bear either both sexes (archegonia and antheridia) or only one. In dioicous taxa. male plants can be monomorphic with the fernale plants, or be reduced to dwarf plants that seemingly grow exclusively on the fernale plant. Such sexual dimorphism is cailed phyUodioicy, and is hown witbïn the Orthotrichaceae from oniy the Macromiaioideae. Ramsay and Vitt (1986) hypothesized, based on cytological data, that withh Ma~romi~um,monoicy is derived through polyploidization of haploid dioicous taxa, and that the sexual conditions merevolved to phyllodioicy through aneuploidy. 25. Morphology of paraphyses: paraphyses - panly - composed of inflated to globose cells (O), or of rectangular cells (1). Paraphyses in Funaria and Brachymim are composed of infiated, rather thin-wded hyaline cells, whereas in aII the Chapter six: 106

Orthotrichaceae studied, the paraphyses are made ofrectangular, rarather thick-walled cells. 26. Spore morpbology: unicellular (O). or multicelluiar (l), unknown (?). Except for Orthotn*chumsteerei (Lewinsky 1993), multicellular spores are known only fiom Macrocoma subg. Traciayphyllum (this study), Muelleriella Witt 1976), and Orth~rni~unt(Lewinsky and Crosby 1996). The sporophyte of Cardotieilu quinquefaBa is unlmown, and ail the remaining characters were scored as "unknown". 27. Caîyptra sprlaœ: smooth (O), or plicate (l), unkwwn (?). The calyptra of mosses is generaliy smooth. In many Orthoüichaceae, however, the calyptra is distinctly plicate, with the plication composed of two cells at the base and a uniseriate crest Calyptrae of Brachymitrim were not seen, and characters pertaïning to the calyptrae are scored as unlaiown (except character 28). 28. Calyptra shape: cucullate (O), or mitrate (1). unluiown (?). Cucullate calyptrae are defined by a long single slit. In the ûrthoaichaceae, both types occur, and the shape of the caiyptrae is conserved within most genera. Ail outgroup taxa have a cucullate caiyptrâ 29. Caiyptta outpwth: absent (O), or present (I), unknown (?). The calyptrae of the Orthotrichaceae are typically omarnented with hairs. In taxa with piicate calyptrae the hain originated from the plicae (Janzen 1917). The ontogeny of hairs in taxa with smooth cdyptrae (e.g., Pleicrorthotn~chumchilense, and Schlotheimia tn'chomitrù) has not ken investigated, but is tentatively considered homologous to the former. 30. Outiine of nuface cebof calyptrae in transverse section: rectanguiar (O), isodiametric (1), unlcnown (?). In Macr~mi~umand Or?honichm, amongst others, the calyptra is composed of a single layer of thick-wailed cylindricai ceiis. In Zygodon and Schlotheimia, the ooter ceiis are rectanguiar in transverse section. In Funaria and Mniwn, the caiyptrae is unistratose and composed of broadly rectangular ceiis. 3 1. Thickness obdyptra: unistratose (O), uni- to bistratose (Zygodon-type), (uni-) bistratose to trismatose (Schlotheimia-type), unknown (?). In Zygodon the calyptrae are composed of one to two layers. The outer layer is composed of rather nmwceils, and the inner ceUs are broader. In Schlothemia, the inner ceUs tend to be broader than the outer cells too, but the outer ceils are mostly isodiametric rather than mtanguiar in transverse scetion. In the other taxa, the caiyptrae are mostly unistratose (except for the plication) and composed of namw - in transverse section isodiametric - ceils. 32. Outliw of base of caigptra: entire or at most slightly dissected (O), lobate (l), or fkinged and Iacerate (2). or deeply lobate (3). unknown (?). The genus Schlotheimia has Cbapter six: 107 traditionaliy been characterized by its distinctly lobate calyptrae. The lobes, typically five of them, are trapezoidal in shape, and th sinuses do not extend beyond the unistratose base of the calpyaic. In Lllou, the base is also lobate, but the lobes an rather irreguiar, and the sinuses extend to the mediau portion of the calyptrae; this situation is considered a distinct state. 33. Caiyptm aizc: covering the um completely (O), only the upper half of the um (l), or only the operculum (2). unknown (?). The size of the calyptra, in comparison with the length of the urn, is variable within the Orthoaichaceae. In most taxa the calyptrae reach the base of the un, whereas in others, particularly in Groutiella they cover only the upper half of the m.In taxa with immersed capsules, the calyptra may only cover the opercuium (Demotheca). 34. Caïyptra apex smooth (O), or scabmse due to pmjecting cell ends (l), UlZkIlown (?)- The apex of the calyptra in Zygodon obtus~olius,kratia neocaledonica (and others), is roughened by prorate ceils. This situation differs hmpapillose cdyptrae found in Orthozrichuna obtusifolium or Florschuetziella steerei. where the ceiis produce papillae over their lumen. 35. OpemuIum shape: flat (O), or conical (l), unknown (?). In the Orthotrichaceae, the operculum is either flat or nearly so (Desmotheca), or conspicuously convex, and conic. Opercula were lacking in all the material of Pleurorthohichum examined, coosequently characters 35 to 38 are scored as unknown for this mon. 36. Rostrum: absent (O), present (l), unknown (?). Stoneobvm is the only Orthotrichaceous taxon where the operculum is not terminated by a rosmim. 37. Rostriuii length: long (O) or short (l), unlaiown (?). The length of the rostm is here considered with regard to the width of the opercuium. Short rostra, i.e., rostra shorter thea the width of the base of the opercuium, occur in various taxa of the Orthotrichaceae (e.g., Orthonichum, Florschuetziella). 38. Rostnim orientation: perpndicular to the base of the operculwn (O), or oblique or curved (1), unknown (?). The rostmm in 0rthotrichwn is always straight, whereas in Zygodon, the rostnun is oblique. 39. Uni surface: smooth (O), or nid(1). unknown (?). The exothecial cells in many Orthouichaceae are dinerentiated into longitudinal bands. The cells forming the nbs differ hmthose ktwen the ribs by the heavily thickened outer periclinal and axial anticlinal walls. The ribs alternate with the exostome teeth, and the differentiation of ribs may be correlated, at least in some taxa with a weil developed peristome. This needs to be merstudied. Chapter six: IO8

40. Width of the mouth of the am:mouth not constsicted (O), mouth coastncted when dry (1), U1SII1own (?). In some Orthotnchaceae the mouth of the capsule is constricted even in the moist state, and thus is independent of the presence of ribs extendhg to the mouth. Malta (1926) hypothesized a coneIation of the constriction at the mouth and the duction of the peristome in Zygodon. Similar tends are obvious in Mucroconm. More study is needed to assess this phenornenon. 41. Sfomata: absent (O), or present (1). unknown (?). Stomata are present in most mosses. In the Orthotrichaceae, they have been lost in Schlotheimiu tectu and S. trichornitria. The loss of stomata is considered derived in mosses (Miller 1979). 42. Number of gunrd cek: 1 (O), or two (l),unknown (?). Except for Funaria hygrometrrèa, which has a single "donut"-shaped guard celi per stomata, aiI other taxa examined have two guard cells. 43. Stomaw expowue: phaneropore (O), or cryptopore ( 1), hown(?). Stomata with their guard celis immersed and partiy or nearly covered by subsidiary celis occur in various umlated taxa of mosses (Vitt 198 1). Within Orthonichurn sensu lato, Lewinsky (1977) hypothesized that ùnwrsed stomata had arisen only once, whereas Vitt (1976) argued for a polyphyletic ohgin of csrptoporic stomata and thus for the generic distinction of Muelleriella. 44. Stomata distribution: conûned to the neck or the lowermost portion of the um (O), to the lower balf and median portion of the um (1). or the upper half of the uni (2), unlaiowo (?). 45. Anatomy of the PaLl aüs of the seta: cells diffcrentiated into a central saand (O), or not differentiated (1). AU taxa examined, except Muelleriellu and Stoneobryum, have the axial ceils of the seta differentiated: the^ wdls are thinner and often coiiapse, resuiting in an axial cavity. 46. Anatomy of the œksumunding the centrai stmd in the seta: cells not differentiated from other cortical ceils (O), or differentiated and forming a distinct ring around the central strand (1). In Funaria, as weli as in some Orthotrichaceae (Onh~~chumlyellii; see drawings in Lewinsky 1994) the celis surrounding the central strand are differentiated hmthe adjacent cortical ceils and fom a more or less distinct ring around the central strand. 47. Twisting of the dkW portion of the seta: dextrom (O), or sinistrorse ( 1) , unkIlown (?). These character-states are defined based on a view from the inside of the spiral. In many taxa the twisthg in the distal portion of the seta differs from that in the basal portion. Here only the distal portion of the seta, including the neck of the capsule is Chapter six: 109 considered. This character was not scored for taxa with immersed or emergent capsules (e.g., Stoneobyun mina) 48. Preperistome @rastome): absent (O), or pnsent (1). A preperistome (it., a thickened layer composed of the outer OPLwalls and adjacent wds-Outer Peristomial Layer, Blomquist and Robertson 1941-) was thought to k nstricted to the Orthotrichoideae, and particuiarly OnhotrChum sensu lato (LewinsLy 1977), but have recentiy been reported ais0 hmFlorschuetzielfa (Vin 1979) and Leratia (this study). 49. Dcgrc of huion of exostome tee* teeth frre (O), hised into 8 fke pairs (1). or joined ail around (2). In maay Chhoalchaceae the exostome teeth an hwd into eight pairs. Lack of fusion is known hmvarious taxa (e-g., Lcratia, Schlotheimia), whereas reduction of the exostome leading to a continuous "membrane" has occumd only within the Macromihioideae (e.g., Groutiella, Macrocorna, Macrominim). 50. Tbickness of the exosforniai OPL: OPL as thick (O), thinner (1). or thicker (2) than the exostomial PPL. The orthotrichaceous peristome ciiffers fiom other diplolepideous peristome types in the exostomial OPL-thickening being heavier thm the PPL-thickening of the exostorne tooth. For taxa with the exostome reduced to a membrane, this character was considered as "not applicable": The exostome in these taxa has equaiiy thick outer peristomial and pcimary peristomial layers, and this reduction in the thickening is considered linked to the reduction in size. 5 1. Morphology of exosfornial PPL: PPL with smgtrabeculae (O), or trabeculae absent (1). Trabeculae comspond to the thickened anticlinal wali remnants of the exostomial PPL. Such thickenings are missing from aii orthotrichaceous taxa examined, but are present in Funaria, Brachymitrïum, and Mnim. 52. MorphoIogy of cndostomirl PPL: PPL without (O), or with a median vertical wall, not ap~licable(0). In the opposite diplolepideous peristome, the outer layer of the endostome segments is composed of a single row of PPL cells, whereas in the altemate peristomes (Orthotrichaceae and Mnim, in ihis study) the segments are made of two haif cells on their PPL surface. The endostome is reduced or xnissing in various taxa, and this character is treated as not applicable. 53. Width of endostome segments: segments as wide as exostomial teeth (O), or nmwer (1). Peristome nduction has taken place in many orthoaichaceous taxa that have adapted to xenc habitats (Vin 1981b). In Zygodon reinwardtii, the mouth of the urn is constricted, and the peristome is reduced to a tudimentary exostome and short endostome segments. The segments, though reduced, are stili almost connected at their Chapter six: 1 10 base, and therefore the endostome segments cire considered to be as wide as the teeth, since they are as wide as they could k,given the consaiction of the mouth. 54. EndO6fomiPl ckpzesent (O), or not (1). In the diplolepideous altemate Bryum-type peristome the IPL undergoes additional divisions leadhg to a 4:2:8 pattern. Breakdown dong the longitudinal wds of these ceiis leads to the formation of narrow appendages, the cilia. Cilia are never developed in the Oahotrichaceae, and additional divisions in the IPL when not found in any of the taxa studied.

Chamcters exciuded 1. Anatomy of outer cortical cauline cells. The outer ceUs of the stem are typically smaüer than the parenchyma celis and colored (yellowish to deep orange rather than hyaline). In some taxa, this differentiation aects several layers of cells, iastead of just the outermost cells. By contrast in Z pungens, the superhcial cells are poorly differentiated. In most taxa, the differentiation may affect only the color of the walls, whereas in others, the walls of adjacent parenchyma cells are also colored. This character needs Merstudy- 2. Axillary hairs. Maryhairs were found of taxonomie use in the Pottiaceae (Saito 1975) and within pleurocarpous lineages (Hedeniis 1989). Within the Orthotrichaceae these hairs seems to be caducous as they can only be found with certainty in the apical region of either the stem or branches. Ali orthotrichaceous taxa examined have axillary hairs with at a least a single brown and quacirate cell, that is well differentiated from the remaining celis that are hyaline and rectangular. In some taxa, most hairs have been found with two such basal ceils (Pleur~rtho~chumchilense; Cordotiella quinquefaria) and in Lemtia, both cells that compose the hah are short, and brownish. The hairs can be very short and composed of only two cells (BryodUrmia) to very long and with many cells (Sten~mi~um).1have found variation within the number of cells in severai taxa, and on single stem the number of ceiis can be variable. This chatacier is best excluded until the variation is more criticaliy examiaed. 3. Penchaetial leaves. Ali the Orthoaichaceous taxa have differentiated perichaetial leaves. The degree of differentiation however varies considerably. Perichaetial leaves that ciiffer in shape and size fkom the vegetative leaves are known fiom various taxa (e.g., Pleurorthotnchum, Ceuthotheca, Schloîheimia subg. Stegotheco). In general, however, the differentiation is less conspicuous, and is often restricted to the cell areolation and ornamentation. In BryodUronia, for example the basal celis of the penchaetial leaves are prorate, whereas those of the vegetative leaves are smooth. tu Grouriella all leaves are Chapter six: L 11

simiiar in shape, but the penchetid leaves have basal cells that are more lax. elongate, rhomboid, hyaline, and the differentiation into chlorophyllose, isodiametric or oblate ceUs is restricted to the distal portion of the leaf. The differentiation of the penchaetid leaves needs to be examined critically kforc discrete charâcter states can be defmed, and before this character can yield signincant phylogenetic information. 5. ûutline of stem transverse section and ranking of leaves. Ln Stenomitnùm and Pleurorîhomcizm the vegetative leaves are distinctly fanked in five rows. In Florschueaiella too the leaves are five-radced, but less conspicuously. Examining the outhe of the transverse section of the stem, revealed that in those taxa the section is distinctly pentagonal. hi most other taxa however, the transverse section is also pentagonal, or marly so. and rarely round. The character was excluded fiom the anaiysis, because it could not be broken dom into discrete character states. 6. Leaf habit and shape: The habit of the leaf has ken used to distinguish between most species of Orfhom'chm and Mota. The habit of the leaf seems to be correlated with the size and shape of the leaves, and perhaps even to the anatomy of the costa and cell areolation. The lack of stereids, for example. seems correlated to erect leaves, in taxa with short, or broadly oblong leaves (&thotichum ob~ilsijioliurn.or Zygodon obtus~olius). In Onhofrichum lyellii, the leaves are flexuose, yet their Costa lacks stereids, too. In this case the twisthg of the leaf may be made possible because the leaves are much longer, and rather linear-lanceolate, as well as by a weakening of the costa in the upper lamina. These characters have to be considered with great caution, particularly if possibly linked characters are included. Furthemore, the shape of the leaves is variable within genera (see iliustratiom in Malta [1926, for Zygodon], Malta 11927, 1933. for Ulota], or Lpwinsky [1993] and Vitt Cl9731 for variation in Orthotrichum), and identification of discrete states is not readily obvious. These characters are therefore bettcr excluded from phylogenetic analyses, at least at the generic Ievel and above. 7. Peristome featws. The omamentation of the penstome in the ûrthotrichaceae has ken snidied by Lewbslry (1977) and Shaw (1985). Though definite trends may be apparent within genera (towards a more elaborate omamentation in Orth~~churn, Lewinslcy 1977). the ornamentation appears very much linked to the degree of reduction of the pristome. Taxa with weil developed peristomes (e.g., Schlotheimia brownii, Macromitrïum longifolium, or Stenominiwn pentustichum) have saongly striate-reticulate and even lamellate OPL surfaces, whereas reduced teeth in Macrocorna or Groutiella, are covered with low, scatterrd papiiiae. Taxa with intermediate types of penstome Chapter six: 1 12

(moderately reduced) have teeth that are papillose, striate-papillose, with the striation chmging orientation upward dong the teeth. The use of this charstcter in phylogenetic analyses, may have to await furthet study. The number of PLdivisions is another character thai may prove significant in identifjhg primitive taxa within the fdy. Within the genus Orthohichum, Lewinsky (1993) argued that the primitive taxa have 32 ceils composiug the PL. Subsequent evolution led to a reduction and a disp1acement of the PL,resulting in only 16 ceils in the PL,with theU anticlinai walIs not alignecl with the planes of divisions in the outer layer (Lewinsky 1993). The number of ceils in the IPL can be determined either from the segments alone or fkom the basal comecting membrane. In many taxa, however, half or aii segments as weil as the basal membrane are lost, and the pertinent characters are consequently missing. Ushg this character with a reduced mon sampie covering such a large family as the Onhomchaceae, and including several taxa with reduced endostomes, rnay fdto yield an accurate phylogenetic signal and lead to excessive homoplasy. It is therefore excluded from the present analysis. 8. Alignment of the proximal cells of the operculuai. Hedeniis (1994, 1995) considered this character in the phylogenetic reconstruction of pleurocarpous taxa. In the Orthotrichaceae, the ceils of the proumal rows of the operculum, are aligned in more or lcss distinct vertical rows. In Funaria hygrometrïca. these cells form lines that are spirally nMsted amund the axis of the operculum, an observation that led to the hypothesis that the position of the cdof the opercuium may be conelated to the oatogeny of the peristome. The peristome is formed hmthree layers of ceiis, below the presumptive zone of the opercuium. h Funaria hygrometrica the peristome teeth are strongly sinuose, similar to the pattern of cells of the operculum. It is thus possible that the orientation of the ceUs of the operculum is dependent on the habit of the peristome. Within the Orthotrichaceae, the stroag reductionary trends in the evolution of the peristome (as hypothesized in Orthmchum; Vin 1971, Lewinsky 1993) may result in the loss of alignment of the cells of the opercuium. Such comiations need to be tested before the character is used in a phylogenetic anaiysis. 9. Chromosome numbers. Though the number of chromosomes is known for repnsentatives of several genera of the Orthotrichaceae (see summazy in Fritsch 199 1, Ramsay 1993, Ramsay, Streima~,and Vin 1995), no counts have been made for the smaller genera. Furthemore, the nurnber of chromosomes varies within gencra (Fritsch 1991, Ramsay and Vitt 1984, 1986). Ramsay and Vitt (1986) hypothesized that polyploidy (X=6 to X= 12) is foilowed by aneuploid reduction through chromosome loss Chapter six: 1 13 or hision. It is however impossible to ascertain the origin of aneuploids (fusion or loss of chromosome). Assumiiig homology for a given chromosome number shared between taxa may result in false synapomorphies, and therefore this characters has been excluded from the analysis.

Phylogenetic reiaüonships. The heuristic search, ushg Funaria hygmmetrica as the outgroup, found a singk island composed of 56 equaiiy parsimonious trees. The mes are 191 steps long and have a coasistency and a retention index of 0.414 and 0.685, respectively. The f-value varies between 3720 (tree three) and 4568 (mes 18-20). A strict consensus as well as a 50% majcrïty detree are presented (Figs. 6.1 and 6-21. Furthemore, the tree with the lowest f-value (tree th)is chosen for exarnining character transformations in the evolution of the Orthotnchaceae (Figs. 6.3-6). In ail mes Bra~hyrni~unris sister to a dichotomy between Mnim and the ûrthotrichaceae (Fig. 6.1). The onhotrichaceous taxa are distributed ammg three major clades, except for Leraria which is basal within the farnily (Fig. 6.1). The 56 most parsimonious trees are congruent with regard to the monophyly of a zygodontoid-clade (Lepodontiopsis, Pleurorthotn'chum. Stenonitrïum, Zygodon), an orthotrichoid-clade (Bryodironia, Muelleriella, Onhominium, Otthotrichum, Stuneobry~,ad Ulota), and a macromitrioid-clade (Cardotiella, Ceuthotheca, Desmotheca, Florschue~iella, Groutiella, Leiomitrium, Macrocoma, Macromitrim, Schlotheimia). This analysis confirms strong affinities of Ceurhorheca for the Macmmitrioideae, and Orth~mi~umfor the Orthotrichoideae, as well as only a distant relationship between Orthotrichoideae and Pieurorthotrichwn, which is here placed in the Zygodontoideae. Within the latter clade, Zygodon appears paraphyletic, with species of section Zygodon more closely related to a derived clade composed of kptodontiopsis, Pleutortho~chum,and Stenomrmr~*m,than to the other sections of Zygodon (Fig.6.4). The monophyly of Orthohichum and Ulota within the Onhotrichoideae is not supported by any trre. The parapbyly of Ulota resuits from LI. obnc~iusculaand W. lutea being more closely related to Btyodixonia than to U. magellmica. OOrotrichum appears either paraphyletic (e-g., tree 1, not show) or polyphyletic (me3, Fig. 6.5), due to Muelleriella which is nested within Orthotrichum, or to closer affinities of O. lyellii with the clade composed of Ulota, Stoneobrywn, and Orthomihium. Within the Macmmitrioideae. too, not ail currently accepted genera appear monophyletic. Both Macroconui and Macromitrium are represented as paraphyletic or polyphyletic assemblages (Fig. 6.6). The relationships within the Macromitrioideae remain ambiguous for many taxa as shown in the strict consensus tree Chapter six: 114

(Fig. 6.1). AU most parsimonious however. agree on most taxa with short basal cells king less derived, forming a parapbyfetic sister pupto the clade composed of taxa with long rectangular basai cells.

The OrthoMchaceae: morphoiogical fhnincterlzation and infrafamilial ~iationship.A cladistic analysis using 54 morphological chai?u:ters yields a strongly supported monophyletic Orthotrichaceae. Among the seven characters that support the monophyly of the Orthotrichaceae (Fig. 6.4) foin can be considered invariable within the family: the lack of cauline ceneal sttand (charecter 2), thick-wailed upper laminal cells (16), OPL thickeaing heavier than PPL thickening (50) and absence of trabecuiae on the exostomid PPL (5 1). The latter two charecter states are invariable among taxa with well developed peristomes only. The absence of a cauline central strand is considered denved within mosses (Millet 1979) and so are cells with iacrassate walls (Vitt 1984). a thick OPL and a lack of dorsal exostomid trabecuiae (Via Goffhet, Hedderson 1997.; chapter five). The costa is always composed, in the plane of the laminal cells, by two ventral guide ceils. The abaxiai cells are always clearly dinerentiated into substereids (e.g., Zygodon). except in some taxa, mostly those with weak costae (e.g., Orthonichum). In Orrhotichum, the anatomy of the costa has been interpreted as homogenous (Lewinsky 1994). yet the abaxial cells cliffer clearly fiom the the adaxial guide cells, although their differentiation into substereids is aot complete, and the cells are irreguiar in shape. The Orthotrichaceae, except for two species of Muellerielfa,can be defined Merby the absence of chlorophyllose cells on the adaxial surface of the costa (9). The remaining character-states, namely papillose laminal celis (13). and outer cells of the calyptrae with a short rectanguiar outline in transverse section (30) are not constant within the famiiy (Figs. 6.4-6). but are here suggested the plesiotypic character-states of the family. The fusion of the exostome teeth into eight pairs defines the monophyletic group sister to Leratia. that is all three major lineages within the Orthotrichaceae (Fig.6.1). If a more denved position is considered for Lemtiu (sec below), the hision of teeth wouid then becorne pleisiotypic, and synapomorphic for the Orihotrichaceae. The parenchyma cells of the stem, are tb-wded in ail thme outgroups, whereas in the Orthotrichaceae (except for Stoneobryni) the walls are thick or even very incrassate (e.g., Ulota). The transformation of this character is considered ambiguous by MacClade, because of the basal position of Leratia which has very thick-walled parenchyma cells. Considering that Chapter six: 115 all outgroup taxa have thin-walled cells, a transformation fkom thin to thick-, to very thick-wailed cells appears more Iürely than having very thick-wded cells arising hm thin-walled cells, and later evolving to moderately thick-wailed cells. s-, In s-, the Oahotricbaceae sensu Vitt (1984; and considering amenciments in chapters two, three, and five) are thus phylogeneticdly clearly debed by apotypic character-states such as: lack of a cauüne central sirand, moderately thick-wailed cauhe parenchyma cells, thick-walled and papillose upper laminal ce&, exostome teeth fused into eight pairs and with an OPL thicker than the PPL, lack of trabecuiae on the exostomial PPL. Whether ail of these character-states would remain synapomorphic in a broader phylogenetic context is ciifficuit to ascertain since the mie sister-group to the Orthoaichaceae is not Imown. Thick-wded upper laminal cells are known ammg acrocarpous Bryales (the putative sister-group to the Orthotrichaceae; chapter five) fkom the Bartramiaceae, a group coasidered by some authors a distinct order, sister to the remaining Bryaies sensu fat0 (e-g., Koponen 1982). Cornparisons of 18s gene sequences, however, suggests that the Bartramiaceae are nested within the Bryales (Hedderson et al. unpubl.). The possibility for the incrassate upper cells to be plesiotypic in the evolution of the Orthotrichaceae, due a shedancestry with the Bartramiaceae, can therefore be rejected, and thick-walled upper laminal cells remains a good apomorphy for the Orthotrichaceae. Churchill and Linares (1995) considered the Orthotrichaceae to be defiwd by plesiotypic character-states ody, and to represent an evolutionary grade reached by two distinct heages: one composed of the acrocarpous taxa (Zygodontoideae and Orthotrichoideae) and the second one, restricted to the Macromitrioideae. A polyphyly of the fdywould wcessitate parailellism in six character-states defining the Orthouichaceae, as well as in thne additional characters pertaining to the calyptrae, that defmed a combined Orthotnchoideae-Macromitrioideae clade (see below). Incrassate cells have evolved, and cauline central strand been lost in various clades (e.g., Bartramiaceae, Orthotxichaceae, Pottiaceae) and calyptrae sharing same sbape, ornamentation, etc., have &sen independently in distantly related taxa (e.g., Orthotrichaceae, Grimmiaceae; Janzen 19 17). The possibility for a polyp hyletic origin of the Orthotnchaceae thus rests on the likelihood of pardelism (venus shared ancestry) in peristomial features among putatively distantly related taxa Thick OPL and loss of dorsal exostomial trabeculae are known fiom the Pterobryaceae Kuidb. (Nishimura and Watanabe 1992) and the Hookeriaceae Schimp. (Vin 1981a). two pleurocarpous, but othenvise unrelated taxa (Vitt 1984). Hedeniis (1994) included Schlotheimia in his phylogenetic analysis of pleurocarpous mosses. He argued for a rather close relationship Chapter six: 116

between the Orthoaichaceae and "a clade where most pleurocarps belong". The various types of disüibution of female gametangia in mosses have been cntically reexamuied by La Farge-England (1996) who considered Schfotheimia cladocarpous and not pleurocarpous as Haden& did (1994). Molecular data, both hmthe chloroplast gene rbcL (chapcr five) and the nuclear gene 18s (Hedécrson et al. unpubl.) converge with respect to the monophyly of the ûrthobichaceae. and based on morphological data available, this systematic hypothesis still represents the most parsimonious evolutionary scenario (Vitt, Goffinet & Heddersoa 1997).

Thm large subfamilies have ken ~cognizedtraditionally withlli the Onhotrichaceae based on the shape of the calypttae and the distribution of the fende gametangia: the Oahotnchoideae, Macromitrioideae and the Zygodontoideae (e.g ., Brotherus 1924, Vin 1982a). The distinction of these three major iineages within the ûrthotrichaceae is supported by ail 56 most parsimonious trees, but suppon for their monophyly is weak (Fig. 6.1). The relationships among these iineaga is Mercongruent with Vitt's (1982a) hypothesis that the Zygodontoideae an sister to a clade combining the Orthotrichoideae and the Macromitrioideae (Fig. 6.1). The Zygodontoideae, as weli as its sister-group are defmed by th= putative synapomorphies al1 pertaiouig to the calyptrae. Vitt (1984) considered the mitrate calyptrae, as found in the Orthoaichoideae and the Macromitroideae, the plesiotypic state in mosses. The cucullate calyptrae of the Zygodontoideae is invariably smooth, whereas in the Orthoaichoideae and the Macromitroideae the calyptrae are plicate, a state a priori considered denved in nosses. The polarization of the transfonnatioas of the shape (cuculiate versus mitrate) and the sUnace of the caiyptrae (smooth versus plicate) in the Onhotrichaceae is ambiguous, and cannot be solved without a better understanding of its potential sister-group. The nam>wly rectanguiar ceils composing the outer layer of the calyptrae in this subfarnily may be seen as a reduction of the bmadly rectangular cells in the outgroups, with fûrther reduction leading to the in transverse section, isodiametric cells found in the Otthoaichoideae and Macromitrioideae. The oblique orientation of the rostnun may be an adaptation to facilitate shedding of the cuculiate calyptrae or dternatively the calyptrae rnay have becow be cucullate under the constraint of an oblique rostnun. In either case, the oblique orientation of the rostnun seems linked to the cucuilate shape of the calyptrae, and may thus be coasidered plesiotypic, too. As a result, the Onhotrichoideae-Macromitnoideae clade appears more denved than the Zygodontoideae. The shape and number of papiilae per ceii were seen by Vitt (1972, 1982a) as Chapter six: I 17 discrimantory characters for the Zygodontoideae, many of which have up to six, small, clavate papïliae per cell. The number and size of the papillae is, however, variable Malta 1926, %der & Vitt 1979, Vitt 1972,1982a) and some Macromitrioideae have numrous sdpapi11ae (e-g., M. incun@ioZlm),too. Vitt (1983) and later Vitt and Ramsay (1985) argued that the primitive Australasian MaCrornitria have smooth laminal ceUs and that papiLLosity has ariscn independently in several groups. The small papillae in the Macromitrîoideae are most iikely not stnctly homologous to those of the Zygodontoideae. A distinction of papîiia types between these taxa is difficuit to rationalize based on morphology alone, and may have to await shidies of their ontogeny. Considering that papillae in the Macromitrioideae are most ofien if not always associated, with buiging rather than flat cells, as in most Zygodontoideae, papillae may actually undergo a different developmental sequence in the Macromitrioideae and the Zygodontoideae. The relationships between the thRe subfdies as obtained flom morphological data are weakly supporteci (Fig. 6.1) and merfiom the nucleotide based phylogeny (chapter five), by the ûrthotrichoideae king sister to the Macromitrioideae rather than forming a monophyletic group with the Zygodontoideae. Tramferring the Oithotrichoideae to a sister-group position with the Zygodontoideae, a relationship congruent with the molecdar phylogeny, and keeping all internai topologies to the subfamilies identical to those obtahed in the most parsimonious me, not ody requires three additional steps, but also results in the Orthotrichoideae-ZygodoDtoideae clade king defhed strictly by plesiomorphic character States. Olmstead (1995) recently argued in favor of "plesiospecies", Le., species that lack uniquely derived characters. Extending this concept, should dow for "plesiosubfamilies" to be, at least tentatively accepted as well, as resulting nom the diversification of an ancestral plesiospecies. Plesiotaxa may actually lack autapomorphies ody at the time of the actual speciation event of the sister apotaxon. Ultimately, however, genetic change may occur in the plesiotaxon (gradua1 evolution), and Iead to autapomorphic character-states. A plesiotaxon may thus only be an artifact of a panicular study based on a given set of characters avaüable. Future studies rnay reveal morphological or molecuiar character-states unique to a 44Zygodontoideae+Ortù~trichoideae"clade. Therefore the absence of synapomorphic characters for this combined clade may not represent an obstacle for accepting the monophyly of tbis iineage. Considering that most branches collapse when the most parsimonious scenario is relaxed by a single step, the alternative phylogeny, though it necessitates 3 additional steps, caanot be rejected based on the data presented here. Chapter six: 1 18

Circumscription and relationsbips within the Zygodontoideae. When Brothems (1924) proposed the Zygodontoideae he included next to the type genus, also Rhachiihocim Le Jolis, a genus later placed in its own famiiy (Robinson 1964) and recently excluded nom the ûrthomchales (chapter theand five). Malta (1926) arranged the nearly 70 species of Zygodon into four sections. Section Bryoides included aii taxa with smooth lamina1 ceb. Papülose taxa were distributed arnong three sections. Sections Obtus@ioIiiand Stenomimmùmare both monotypic and are defhed by obtuse and five-ranked leaves, respectively. The type section. the largest of all four, included ail remaining papillose taxa Malta (1926) considered section Zygodon. and particularly the synoicous species, the Z reinwctdtii-group to represent the most primitive members of the genus. Leptodontiopsisjkagil~oliawas initially placed by Brothems (19 11) in the Pottiaceae (see also Bmthem 1924). and later described as Zygodonfragi~î$oZiusBroth. (in Malta 1926) and included in section Zygodon. Zander & Vitt (1979) too, suggested that Leptodontiopsis "is probably best considered as a synonym of Zygodon'" (see also Vitt 1982a). Leptodontiopsis currently includes four piesali endemic to high mountains in East Mcaor Bomeo (Brotherus 1910, Dixon 1934, 1938). The monotypic genus Stenomitrium, which is endemic to western South America (Robinson 1975). was segregated nom Zygodon by Brothcnis (1909) whereas Mitten (1869). and Malta (1926) considered it synonymous w ith Zygodon. Vitt (1984) tentatively retained Leptodontiopsis and the monotypic Stenonzim*umdistinct nom Zygodon. The relationships arnong the Zygodontoideae are weii resolved and consistent in all trees, but support is weak (Fig. 6.1). Sten~nU~umand Leptodontiopsis have clear affinities with Zygalon, particuiarly section Zygodon, resulting in the paraphyly of the genus Zygodon sensu S~C~O,in ail most parsirnonious tms (Fig. 6.4). The paraphyly seems even more accentuated by the inclusion of Ple~rorthot~chum,a monotypic genus endemic to central Chile (Lewinsky 1994) and its close relatiooship with Stenumim*urn and Leptodontiopsis (Fig. 6.4). L,ewïnsky (1994) recently argued that the affinities of Pleurorthomchtun couid be with either the Macromitrioideae or the Orthotrichoideae, and that some characters are even incompatible with either placement. Pleuror?hotichum chilense. easily distinguished within the Onhotrichaceae by the narrowly lanceolate and long acuminate perichaetial leaves (Lewinsky 1994), actually shares several character-states (including plesiomorphic ones) with the Zygodontoideae, and particularly Stenomihium. The perichaetia are terminal on the stem. and not on short lateral branches as suggested by LewinsQ (1994). The stem resumes growth through a Chapter six: 1 19 single lateral innovation to a centimeter below the apex. This axis is clearly identifiable as a branch by the presence of jwenile leaves at its base (see La Farge-England 1996). Not aii branches are developed following the perichaetium formation on the main fis, but these appear to be basitonous, rather than acrotonous as in the Macromitrioideae. Male gametangia were unknown for PleurorthomChum and phyllodioicy (a character state hitherto known only hmthe Macromitrioideae) was suspected (Lewhsicy 1994). Pengonia were found however in at least two coIlections (Cd& Inga Skonsberg 211- H-Br.; Muhu IOII9-N), on plants monomorphic with the perichaetium bea~gplants. The calyptra is cucullate and in transverse section ~vealsa similar cell outhe as in other species of the Zygodontoideae. The hairy calyptrae, reminiscent of the Orthotrichoideae or Macromitrioideae, are not incompatible with the Zygodontoideae, since they occur also in the African species, Z nichornitri-us Hook. & Wils., a species which also has differentiated penchatid leaves (Malta 1926). The chlorophyllose cells are distinctly papiiiose, and whereas the upper laminal cells bear only one or two papillae (see Lewinsky 1994) the proximal chiorophyllose cells are ornamented by up to six, albeit smd, papillae, similar to most species of section Zygodon. The transfer of Pleurorthonichwn fiom the Orthotrichoideae to the Zygodontoideae is thus well justified based on morph01ogical characters. Pleuronhotnchm fonns with Stenomitnwn and Leptodontiopsis a monophyletic group seemiogly distinct fiom Zygodon sensu stricto by dimorphic basal ceils. The proximal portion of the leaf is composed of alternathg bands of clear, evenly thick- walled cells and ceUs with strongly incrassate, nodose to prose, yeilow walls. Malta (1926) reports yellow-colored basal cell walls from various andean species of Zygodon (e.g., 2 nivalis Hampe). and nodose walls fiom a southem South Amencan species, Z bartramiouies @usen) Mah Zygodan bamamioides had been included in (section) Stenominiwn by Brothems (1925), but excluded by Malta (1926). Both authors may have disagreed with regard to the affhities of Z bartramioides, but both rejected recognizing Sten~mi~unzat the generic level because Z bartrmioides was seen as the obvious iink bewnStenornitritrium and Zygodon, since it shared five ranked leaves with the fonner and acrocarpy with the latter (Brothems 1925, Mdta 1926). Grifîii (1990). also placed Stenomitrium pentastichzun in Zygodon, but unlike previous authors, described 2. burtrumioides as having creeping stems with erect branches. Material of Z bammioides was not available for study, but may be crucial in understanding the affinities of Stenomitrim. Leptodontiopsis, Pleurorthotrichum, and Stenominiwr each exhibit at lest one unique charactcr that is a priori difficult to reconcile with the genus Chapter six: 120

Zygodon or even only with section Zygodon. L+eptodontiopsis is the only taxoa in the Zygodontoideae to have smooth ums. Plewrrthorn'chum has differentiated cells surrounding the central strand in the seta, a uniquely reduced peristome (withh the Zygodontoideae; chapter four) and by fat the tallest plants in the subfamily, with a height of up to ten centimeters (L.ewinsky 1994). Sten~mi~umis eady distinguished by its cladocarpy and prostrate growth. The last two taxa funber Merhm section Zygodon by the seta which is twisted to the left in the distal portion. Whether Leprodontiopsis, Pleurorthom'chunr, and odten~rni~wnwarrant generic status based on these patristic distances, depends,within a phyiogenetic fkamework, more on the monophyly of its immediate sister-groups (Le., section Zygodon) than on the cbaracter-states themselves. Section Zygodar is composed of two major groups, distinguished by their sexual condition. Malta (1926) considered the dioicous taxa derived and the monoicous species composing the Zygodon reinwardtii group the most primitive within the genus. This hyphesis is supported by the pFesent data but cornes within a paraphyletic concept of ihe section, with 2 intemedius sister to the dioicous genera Lqtodontiopsis, Pleurorthomkhum, and Sten~mi~um.Constrauiiag 2. intennedius and 2. reinwardtii to be sister-taxa, requires only one additional step. To satisfy the monophyly of the genus Zygodon, parsimony would need only to be relaxed by two steps. Obviously the relationships among sections of Zygodon are aot robust. Expandhg the current concept of Zygodon to accomodate StenominiMi, Leptodontiopsis and Pleurorthotrichurn, to solve the paraphyly of Zygodon may not be appropriate, and wodd a prion' obscure evolutionary trends within the Zygodontoideae. as it would result in a highly variable and ill-defined genus. The remaining papiiiose taxon in Zygdon, nameiy 2. obtusi$olius, is sister to a clade combining ôoth sections Bryoides and Zygodon. as well as the Stenomitriumi:lade. Zygodon obtusif~lius,the sole species in section Qbtusifolii (Malta 1926) ciiffers corn congenenc taxa by its short, obtuse. erect-appressed leaves, bulging laminal cells, short basal cells, chlorophyllose cells cove~gmost of the abaxial surface of the costa, prorate celis at the apex of the calyptrae (also seen in 2. pungens), and "bifïd to c-shaped papiiiae (Malta 1926, and pers. obs.) Of these characters, only the latter four were considered in the present analysis (see "msuits"), and only two of these, namely the laminal cells covering the costa and the rather isodiametric basal cells of the leaf phylogenetically define Z obtusïjiolius. The prorate apical celis of the calyptrae are interpted as a plesiotypic state. due to the "basai" position of Lrratia, which also has these Werentiated ceils. Furthemiore, ail taxa with "bifid papillae also have simple Chapter six: 121 papillae, and ail these taxa were rored polymorphic. Since Florschuetziella, 2. dtusfoliw, and Letaria which ail share ihis polymorphism, are **basal1'in their respective clade or in the family, the character transformation is lefi ambiguous, and thus not phylogenetically informative. kratia has been traditionally placed in the Macmmitrioideae (Brotherus 1925 [as Lemrieh Broth. & Syd.]; Cnun 1987, Via 1982a) a relationship based on a "Mizcroniimmum-iikeperistome" with deep-set endostome and a short, and imgularly disseaed membrane (Cm1987). While a relationship with the Orthotrichoideae has been argued against by Cm(1987), affinites to the Zygodontoideae have never been hypothesized, even tùough Leratia has a non-plicate cucuiiate caiyptrae. Lratia may indeed be better considered closely related to Z obtusiJ01iu.s. The only character that is a priori in conflict with an inclusion in the Zygodontoideae, is the presence of a preperistome (Goffiet unpubl.). This additional outer reduced peristome has not bem hitherto reliably reported hmthe Zygodontoideae (Lewinsky 1977). Zygodon obtus~oliwis a widespread species knowa f?om Asia. Afnca, South America, and Australasia (Lewinslry 1990, Vitt 1993). In Austrdasia, it is hown nom severai localities in New Zealand, but oniy one fkom Australia (Tasmania; Lewinsky 1990). Section Obhc~ifdiiis cmntly considered monotypic (Malta 1926), but examination of populations hmThailand, Mexico, and Australasia reveals variation in peristome ornamentation (results not shown) that may be indicative of cladogenesis. The Australsian populations are considered conspecific (Lewinsky 199û), and their disjunct distribution between New Zealand and Tasmania as well as their overall scattered distribution in New Zealand may suggest poor dispersal capabiiities. Within such scenario and considering that Leratia is endemic to New Caledonia, fiom where 2. obnrsifolliur has not ken reporteci yet (Ruseu aad Reese 1982), it is easily conceivable that Leratia was derived fkom 2. obtus~oliusthrough long distaace dispersal or represents a vicariat of the latter. A basal position of Leratiu within the Oahotnchoideae appears anornabus, as it leaàs to an early loss -reversal- of features indicative of adaptations to xuic enviromeats (buiging ceii-walis and "bifid" papillae; see Schofield 198 1) in a fdywhere major evolutionary trends are considered to lead to increased drought tolerance (Vitt 1972, 1982a, 1983, Vitt and Ramsay 1985). Leratia is thus best considemi a rnemkr of the Zygodontoideae, with close affinities with 2. obtusifolius. Section Obfusifioliimay now again be characterîzed by its distinct papillae, as suggested by Malta (1926) as weil as by the bulging ceiis, both unique features within the Zygodontoideae. Chapter six: L22

With both sections Ob~oliiand Zygodon, now denved through the acquisition of papiilae, section Bryoides, defined by smooth laminai ceils, may be regarded as the most primitive mernber of the Zygodontoideae. Such a scenario would be against Malta's (1926) hypothesis. but be more congruent with general trends in mosses (see Milier 1979, and Viti 1984) as well as with the molecular based hypothesis that the Onhotrichaceae would be deriwd fbm an acrocarpous ancestor with a cucullate calyptrae and smwth laminal cells-

Circumscription and datiollships witbh the OrthoMchoideae. The Orthotrichoideae have been traditionaily definexi by terminal caulùle gametangia and a large mitrate calyptra (Brotherus 1925, Vin 1972). Neither of these character-states is however apomorphic for the subfamily in the present phylogenetic scenario: acrocarpy is the plesiotypic state to the family, and miirate calyptrae evolved in the ancestor to the Orthotrichoideae and the Macromioiâeae. Derived character-states distinguishing the Orthotrichoideae hmthe Macromitrioideae (and also hmthe Zygodontoideae) can hstead be found in the sporophyte: the stomata are found in the lower portion of the um, above the neck (41) and the distal portion of the seta is sinisaorse (47; Fig. 6.5). Six genera share this set of characters and compose the ûrthotrichoideae (Fig. 6.5): Bryodhonia, Mrcelleriella, Orthominium, Orthohichum, Stoneobryum, and Ulota (the position of the stomata is however variable withui the latter; Malta 1927, 1933). The adaxial cells of the costa are either well differentiated fiom the adaxial guide ceiis, into substereids (Ulota spp.. Muelleriella. Stoneobrywn) or only weakly so and the Costa appears homogenous in transverse section (e.g., Orthotrichum). The absence of distinct substereids in Onhomèhum, may be intepted as a result on heterochrony, with the differentiation of the dorsal celis pmmaturely stopped. The paraphyly of the two most speciose genera, Orthotnchum and Ulota as suggested based on rbct sequence cornparisons (chapter five), is supported by morphological data. A molecular cornparison of BryodUronia perichaetialis with two species of Mora revealed that Ulotu lutea is more closely related to Bryodixonia than to CI. obîusiuscula (U.nzagellunica not included in molecular analysis) suggesting that a generic distinction of Bryodixonia is not appropriate. Close afnnities between Ulota and Bryodixunia are very much apparent hmthe morphological aoalysis: both genera share weii differentiated basal marginal cells with theh thicker anticlinal walls (2 112). inner cortical cauline ceils with very thick (and deeply colored) walls (4/2; except in U. mgellanica), and a deeply lobed but not lacerate calyptrae (323). Bryodhonia diffea from Ulota by Chapter six: 123 the "highiy differentiated and conpicuous penchaetid bracts and the diminutive calyptrae" (Sainsbury 1945) as well as prorate basai cells in the penchaetid leaves (unique within the Orthoaichaceae, results not shown), and a flat operculum. The taxonomy of the European species of Cnota has been weii established (see Frey et al. 1995) and the genus has been reviscd for Ausaalia (Malta 1927). South-America (Malta 1933), and recentiy forNorth America (Vin 1972) and Japan (Iwatsuki 1959). Despite this gdtaxonornic hmewo& evolutionary trends within the genus appear obscure and no innageneric classincation of the appmximately 50 taxa (Vitt, Koponen, and Norris 1993) has ever been proposeci. This apparent homogeneity within the genus may suggest that the monophyly of Wuta should be rrstored by inciudiag Bryodixonia rather than splitting muta Ulto several, a priori ili defhed genera. The genus Orthoinchum is, with 11%species (Lewiasky 1993, 1995, 1996) by far the most speciose genus of the Onhotrichoideae. The species are currently distributed among seven subgenera (Lewinsky 1993). of which heare repnsented in this study: subg. Orthophyiium (O. obtusifiiIium), subg. Onhonchum (O. momalum) and subg. Gymnopow (O. lyeilii). The genus appears para- or even polyphyletic in ai3 most parsimonious ms,due to the inclusion of MuellerieDa or to the stronger affinities of O. lyeilii for the clade composed of 0nhomimmum,Stoneobrym, and Uha sensu lato. Muelleriella, Stoneobryum, and Orthomitrium, are al1 recent segregates of Orthotrichum, and are composed of four, two, and one species, respectively Witt 1976, Noms and Robinson 198 1, LewinsLy-Haapasaari and Crosby 1996). The most parsimonious scenacio restoring the monophyly of the Orthotrichwn complex is only one step longer than the unconstrained tree, but retains the paraphyly of Onhonichum sensu Orthohichum obncsifoliimi and 0.~omalum (the type of the genus) are closely nlated (Fig. 65) and together form a sister-group to Muelleriella, a relatiomhip defined by the flat opeaulurn and the pnsence of a preperistome. Muellenefla differs hmtypical Orthotrichunt species by a bisîratose lamina, a more "complex" costal anatomy, lack of a central strand in the diminutive seta (capsule immersed), mdticeliular spores, exothecial cells with stmngly thickened axial anticlinal wails (thickening hcreases towards the outer surface where the thickening of both waiis meet, resuiting in a V-shape pattern in transverse section), and smooth capsules (Via 1976, ihis study). Such high patristic distance may indicate high evolutionary rates in MuelIen'efIa or an early divergence fiom Orthomkhum. Lewinsky (1977)argued agaiast a segregation of MuelIerieIIa fkom Orthotrichum as both taxa share cryptoporic stomata and a prepenstome, two States previously known, within the Orthotrichaceae, oniy from Orrh~~~hwn.Cryptoporic Chapter six: 124 stomata occur however aiso in another segregate, Stoneobrym, which does not appear closely related to Orthu~chum.and preperistomes have been ddbedrecently f?om Florschueaiella (Mmmitrioideae; Vitt 1979) and Lerutka (Zygodontoideae; this shidy). Muelferieflais composed of four species; thne of these are nanow high elevation island endemics that are thought to have been derived hma widespread lowland species, M. cr~~sijiolia(Via 1976). The muiticelluiar spores may a prion' be seen as an argument against long distance dispersai (van Zanten and Pks 198 1) and thus ment speciation within the genus. Morphometac analysis of M. crass@ioolioover its range reveals iittie variation in ailcharacters snidied (Vitt 1976). If strong selective pressure prevent morphological divergence between long isolatecl populations of this species (Vitt 1982b), the "signincaat" morphological differences berween Orthomchum and Muelleriella may be indicative of an ancient divergence from a common ancestor shared with Orthotriihum (sensu Vitt 1984), rather than a ment split from a derived (sensu Lewinsky 1994) group of Onhotrichia (ir., with cryptoporic stomata). The extent of morphological divergence may of course vary with the mode of cladogenesis, and Muelleriella may have accumulated aIl the above changes rapidly afier speciation, and have remained rather static in its morphology ever since then. Reraining Muellerielia as a distinct genus, may necessitate redefining Onhotichum, and consider additional segregate genera. If Onhotn'chum represents an evolutionary grade rather than a natutal group. distinct evolutionary lineages may ody be identified if iheir potential sister taxon (e-g.. Ulota) is included in the analysis. The aFfinities between Orthomitrium and Stoneohyum mg- 6.5) are most Wrely aa artifact of the data, as both taxa are paaisticaily only very distantly related (1 1 states separates these two taxa in tree the, Fig. 6.2) and should be reconsidered more specificaliy when addressing the evolution of the Orthonichuni-complex. A phylogenetic reconsmiction of the genus Orrhotn'chm is in progress (tewinsky pers.com.), and should allow for the circumscription of Orthonichum and its relationships to its "satelitte" genera to be addressed more criticaily.

Circumscription and relationsblps witbin the Macromitrioideae. The Macromitroideae are composed of over 300 species distributed among nine genera, all nstncted to south temperate. montane tropical or subtropical climates (Vin 1982a). The subfamily has ken debed ttaditionaiiy by large mitrate calyptrae, creeping stems. and gametangia terminal on lateral branches (Vitt 1972,1982a). In the present phylogenetic scenario (Figs. 6.56) mitrate calyptrae are shared with the Orthotrichoideae. Chapter six: 125

Florschue~iellais sister to ail remaining taxa of the subfamily in aU most parsimonious mes (Fig. 6.1). Because Florsd~e~elUois polymorphic with regard to the position of the perichaetia (a single gametophyte can bear terminal perkhaetia on stems or branches), the ongin of cladocarpy in the subfamily is ambiguous: is cladocarpy plesiotypic in the Macromitrioideae or has it arisen independently in FlorschuefzteIla and in the ancestor to the remaining members of the subfdy? If the distribution of the perkhaetia is controllcd by a single gene. polymorphism may be indicative of gene duplication or autopolyploidy with subsequent mutations or allopolyploidy. Genome duplication can be suspected early in the evolution of Mucromimmum(Ramsay and Vitt 1984,1986) and may have even occurred in the ancestor to the subfamily. Genetic change required for the character transformation to cladocarpy may be "minimal",since cladocarpy occurs in Orthotrichoideae and the Zygodontoideae, too. Though autopolyploidy seems iikely, ailopolyploidy bas recently been demonstrated in mosses, and may acWybe a common mechanism of speciation in bryophytes (Boisselier- Dubayle and Bischler 1996, Wyatt et al. 1988, 1993). Florschuetziella shares several character-states wiîh 2. obtusifolius (and Lcrutia), and species of Ma~romi~um,and may have origiuated through hybridization between distantly related progenitor species (e.g., belonging to distinct subfarnilies). If hybridization is confirmed for this species, the basal position of Florschueaiellu within the Macromitrioideae may reflect more its mode of speciation than its actual phyiogenetic relationships with other species of the Macromitrioideae (MacDade 1992). Until the systematic affinities of Florschuetziella are elucidated, cladocarpy should be regarded as synapornorphic for the Mac romitrioideae . Within the Onhotrichaceae, the stem branches either sympodiaiiy, through subapical innovations, or moaopodiaily. involving quasi continuous acrotonous development of lateral branches dong a growing stem (see La Farge-England 1996). The genetic basis or proximal causation for a switch to monopodial growth in the Macromitrioideae or other Orthotrichaceae (see above), is not known. The ultimate causation for changing the mode of braaching is however most WEely to be sought for in the parailel occurrence of cladocarpy, where perichaetia an produced ai the apex of lateral branches. Consequentiy, monopodial bmching shouid, iike cladocarpy (see above) be regarded as a synapomorphy for the Macromitroideae. The occurrence of polymorphism may have genetic causes sunilar to those advanced for polymorphism in the distribution of the perichaetia, and needs to be investigated mer. Chapter six: 126

Constraining the character ttansformation to satisfy the parsimony criterion fier results in the plesiotypic state for the shape of the basal cek to be isodiametric (character 17/2) rather than nctanguiar (17/0) as in Schlotheimiu, or most species of Macromitrium (Fig. 6.6). Short basal ceîis are characteristic of severai genera (i.e., Cardorielfa, Florschuettiella, Groutiella, Leiomitrium, Macrocom4) as weU as MacromimUm sect. Cometium Min. (Brothem 1925). These cells, except the most proximal rows, are always chlorophyllose, and thus very sMarto the idametric upper laminal ceiis. The basai cells of most Macrornitria are isodiameaic and chlorophyiiose in the juvenile Ieava, and later undergo elongation and lose the chlomphyli. The short cells found at rnanuity in Groutiellu and other taxa, may have arisen thtough heterochny (paedomorphosis sensu McNamara 1986). The ecological sipificance of short chlorophyilose basal cells, is difficult to ascertain in the absence ofempirical data. Based on field observations and herbarium label information, species with uniforrn, chlorophyliose ceiis in the lamina appear to grow predomhantly in xerophytic (micro- )habitats, whereas taxa with long basal cells seem to have broder amplitudes dong a moist to dry gradient. The genus Groutiella, for example, which is geographically sympatric with Macr~mi~um,shares with the latter an epiphytic habitat in tropical forests. Groutiella is however restricted, in Central America and the West Indies, to lower montane forests, or to disturbed habitats in montane forests (trees on road sides and in pastures), but is WNally absent hmthe moss flora of the upper montane and elfin fonsts, whereas Macromitim is often a dominant component of epiphytic communities at these higher elevations. Groutiello differs superficially hmMacromifrium only by the Merentiated hyaline rnargin (character 21; Fig. 6.6) and has most Uely evoived from a Macrornitim-iike ancestor (Fig. 6.6). If such correlation between basal celï shape and habitat can be confirmed and demonstrated for Florschuetziella, Cardoriella, LeiornihiMi, adMacrocom (both M. papillusa and M. tenuis), then short basai cells may not be plesiotypic in the famiiy (as suggested in Fig. 6.6), and are better considered to have evolved independently fiom an ancestor which retained the plesiotypic state to the family, nawly re.ctan@ar basal cells. AU Macromitrioideae are characterired by prostrate growth (Fg. 6.6). with stems growing paralle1 to the surface of the substratum. This change in the direction of growth of the stem, fiom orthotropic to plagiotropic. is most easily interpreted, in the Macromitrioideae as weii as in the cladocarpous Zygodontoideae and Orthoeichoideae, as a consequence of monopodial growth. U&e in the latter two subfdes, the functional Merentiation of stem and branches. has been accompanied in most Chapter six: 127

Macromiaioideae by the development of dimorphic leaves (Fig. 6.6, character 5). The stem leaves in Macromitium and Schlothehia for example are typicdy acuminate bm an ovate base, and much derthan the branch leaves, with the latter behg similar in ceil areolation, outlim, size, etc., to the stem leaves of the Orthohichoideae. Such leaf dimorphism is thus most iikely due to changes in the stem rather than the branch leaves. The plesiotypic condition where stexns and branches bear identical leaves, may not be identical to the situation in the acrocaqous ancestor shared with the Orthotrichoideae. In an acrocarpous plant, the stem is terminateci by a perichaetium, and so will the subsequent subapical innovations. Fnnn a fiinctional point of view, these modules are identical, yet hma dewlopmental standpoint, the first axis of the plant is the stem and the lateral innovation is a braiich. If acrocarpy and cladocarpy are homologous (and thus exclusive in a haploid organism, see above), the cladocarpous ancestor to the Macromitrioideae, would have lost the ability to produce a teminal perichaetium on the stem. In a cladocarpous individual, the stem and the branch assume different fùnctions: the stem insures continuous growth, whereas the branch develops the perichmtium. A perichaetium bearhg branch, may "resume" growth by means of a subapical innovation. which in tum will assume the bction of perichaetium production. The leaves of this subapical branch are identical to those of the branch below. Thus if we compare two consecutive modules of the same order, thus with the same function (production of the perichaetium) in Orth~~chumand Mucromitn*um,the leaves would be interpreted as isomorphic in both taxa; yet if modules were defined based on ontogeny (primary stems vems primary branches) a cornparison within the same two taxa wouid lead to the leaves found in Mucromitn*umto be considered dimorpbic. If modules sharing an identical function shouid be compareci, the stem of most Macromitrioideae couid not be compared to any module in the acrocarpous Orthotrichaceae, since the stems of the latter always produce a terminal penchaetium. Bryophyte systematics is based on cornparisons of iunctional modules (e-g., stolon, stem, branch, and flagella in Squomidim. Pterobryaceae; AUan & Crosby 1986). If such a critenon is acceptable for reconstnicting the phylogeny of modular organism, than the weight of alternative fuactions remains to be determined Attributhg more significance to the function of perichaetia production rather than perenniality of the plant, may a priori appear justified, since the former may contribute more to the pereaniality of the species rather than just the individual. Applied to the Orthotrichaceae, this would Lead to the considerarion of a distinct character, morphology of the leaves of sterile stems. This character would only be applicable to cladocarpous taxa, since stems in acrocarps are ultimately fertile (produce a Chapter six: 128

perichaetium), rather than sue. The phylogenetic consequences of such strategy may be that it wouid allow for a denved position of taxa with isomorphic leaves within the Macromitrioideae, since this monaation wouid be Wuenced no longer by the character-state in the sister-group, the acrocarpous Orthotrichaceae. The phylogenetic relatiomhips among the various genera composing the Macromitnoideae are obviously very much dependent on the eariy character transformations. Including additional characters, such as pristomid features, may conüiiute to solving generic affinities. The genus Schlotheimia is the only large genus whose monophyly has withstood both the morphoogical and molecular analyses. Its siter-group relationship to the remaining Mammitrioideae, though snongly supported in the latter analyses (ctiapter five), is absent hmall most parsimonious scenarios. Species of Schlotheimia all have a well developped peristome, including an endostomial comecting that may be synplesiomorphic for the farnily (Lewinsky 1993). Furthemore, al1 Schlotheimiae have smooth lamina1 ceils, and smooth calyptrae, both feanins that would Link it to what has been argued above to k the least derived acrocarpous taxon in the family, namely Zygodon sect Bryouies. A basal position of this large genus is therefore not to be dismissed. The polyphyletic origins suggested for Macrocontu and MucromihiMI, are likely to withstand future rnalysis. Macrocorna is a smd genus including I 1 species, arranged in two subgenera, subg. Macrocom and subg. TrachyphylIum (Bmth.) Vitt (Vitt 1980). In addition to the differences described by Vin (198O), the bitypic subg. TrcchypSlIum is also distinct by its multicellular spores. short and oblique mtnun, and given the oblique rosmun, the calyptra which is campanulate as in congenenc species, but has a major slit on one side, may be best considered cucullate. These ciifferences may k regafded as sunicient to regard subg. Trachyphyllm as a distinct genus. Whether the polymorphism in the branching pattern and the distribution of the perichaetia, is indicative of a hybrid ongui of the ancestor to the subgenus, needs to be investigated fiuther. M.cromitium is the largest genus in the ûrthotrichaceae with approximately 250 species (Vitt 1982a) and is suggested to be, based on morphological and molecular data, of polyphyletic origin. The species taxonomy has been revised in many parts of the world (van Raoy and Wijk 1992, Vitt 1983,1993, Vin and Ramsay 1985, Vitt et al. 1995), but still has to withstand critical study for most of South Amenca, the center of diversity of Ma~romi~um(Vin 1982a). Completion of the world monograph of Macromitnum should ailow for examplar taxa of major evolutionary trends to be identified and for the hem prrsented phylogenetic hypotheses of this morphologically diverse species complex to be addressed Mer. Chapter six: 129

Phyiogenetic wnclrrsion A phylogenetic anaiysis of 54 morphologicai characters supports the recognition of the three subfamilies traditionally accepted wîthin the Oahotrichaceae: the Zygodontoideae, ûrtûotrichoideae, and Macromitroideae. The relationships ammg these subfamilies are denwd by fea- of the caiyptrae and on the dism'brtion of the perichaetia, and are congruent with Via's (1982a) tentative phylogenetic arrangement. Five genera are cmntly accepted withh the Zygodontoideae (ir., Leptodontiopsis, Pleurortlotn'hum, Stenomim0m,Zygodon, as weii as Leroria), and the recognition of the sections of Zygodon at the generic levei appears more appropnate &an bmadening the generic concept of Zygodon to accommodate aii of the variation in the subfamily. Within the Zygodontoideae, section Bryoides, including taxa with srnooth laminal ceils, is argued to be the least derived. The Orthotrichoideae are composed of Muelleriella. Ortlomitntnum,Orthichm, Stotteobryum, and Uloto. The monotypic genus Bryoduonia is best considad synonymous with Ulotu. The recent segregate genera of Orth~~chum(Muelhielfa, Onhomi~um,and Stoneobrywn) are retained at the generic level, until the possible polyphyly of Orthonichum is mer examined. Nine genera compose the largest of the tbsubfamilies, the Macromitrioideae: Curdotiefla, Ceuthotheca, Desmotheca, Florschuertiella, Grouriella, Leiomitn~m,Macrocomu, Macromitrium, and Schlotheimio. The relationships among these taxa remain obscure, possibly as a result of hybridization, or analogous character- States. Chapter six: L30

Gmutiella chimborazensis GmuüWla apiwlata Desmothew apiculata

T Ceudhotheca ctyptocarpa 1 Macrocoma tenuis Mawmitnüm incurviolium Macmmitn'um dchardii Macromitnüm Iong~oIium Schlotheimïa brownr'i 1 - Schlodheim~âtecta Schlodheimia trichornitria M8cr000ma pa~r7bS Cardbüella quiinquefana Cardotiella elimbata

Flo~sdruetP'ell.steerei -Btytxiikonia perid,aetr;alis 1 I Ulota lutea Ulota obtusiuscuIa 1 UIota magellanica Orthomitrium tuberculatum Stoneobryum mimm 1 Orihotn'chum lyellii Orthotn'chum anomalum Orthotn'chum obtusifolium Muelleriella crassifolia Stenomitnirm pentastichum Pleurorthotn'cnum drilense Leptodontiopsis fragilifolia Zygodon intemedius Zyg&on reinwardtii I Z'odbn pungens

A Zygodbn obtusifolius

Mnium thomsonii Bmchymitrium jamesonii

Figure 6.1. Strict consensus tree of 56 most parsimaiious trees (MP'ï) obtaiaed fran a cladistic adysis including 54 morpho10gïcai characters. MPTs 19 1 steps long, CI= 0.41414; values above branch indicate the number of steps oeeded for the branch to collapse. Chapter six: l3 1

Gruutiel& chrnborazensk Gmut&I& apiculata 71 Desmooheca apiculata Ceumhotheca L-l cryptocarpa

Schlotheimîà brownii Schlofheimia tecta Schlotheimia tnèhomitria

Bryod'on& perièhaetialis Ulota lutea Ulota obtusiuscula Ubta mage1lanE;ca Olalomitn'um tuberculatum Stoneobryum miium

Stenomitnùm pentastichum Pleurvtfhotrictium chilense Leptudontiopsis fragilifolia Zygochn intennedius 100 Zyg-n rein wardtii 100 Zygwâon pungens Zygodon obtusifolius

Figure 6.2. 50% majority rule consensus tree of 56 most parsimonious trees (MPT) obtained hma cladistic analysis including 54 morphological characters. MPTs 19 1 steps long, CI= 0.414; values above branch indicaie the percentage of most parsimonious trees that share the topology. Chapter six: 132

GmuMla chimborazensis

Desmodheca apiwlata

I Ceidhothem cryptocarpa Schlotheimia brownir' ; SchIodheimia tnchomitria

9 SctrIoIheimia tecta Macrocorna tenuis MacIOrniMurn ïncumMfolium œ - , ,Macmmitn'um richard17 Mammitnum Iongifolium - Macmma papiIIa~a Ca&üe/la qurirquefad

rn Carcba'ella elimbata a Leiomitnum plicatum - Florschuetaela steerei Bryodlxonk~ penchaeüalr's UIota lutea - UIota obtusiuscula Ulota mageIIanka I Orthomitrium tuberculatum c st~neobry~mtnr'nrm 0rtho~'churn lyellr'i I Otthotn'èhum anomalum Orthotn'chum obtusifolium Muelle&& crassifolia - Stenomitnirm pentastkhum - Pleum~otriCnum chiIens8 = 5 changes Leptodontiopsis fmgilifolr'a intermedius Zygodon rehwerdtii ZYgodbn Pwwns Z'odm obtusifolius Lemtia neocaledonica - Mnium thomsonii

Figure 6.3. Phylogram of me number three; tree with lowest f-value among aii most parsimonious trees obiauied fiom a cladistic analysis includiag 54 morphological characters. Tree 191 steps long, CI= 0.414. Zygodon pungens

Figure 6.4. Phylograrn of tree number thne showing the character transformations supporting the monophyly of the Orthotrichaceae, as well as the relationships between subfarnilies and the genera within the Zygodontoideae. For character numbers see text . Chapter six: L 34 Chapter six: 135 Chapter six: 136

Table 6.1. List of representative specimens examinecl (all deposited in ALTA unless otherwise noted) * refers to taxa for wbich enaies are baseâ on publishiag descriptions.

OUTGROUP Fu~riahygromemca Hedw. Pri'ddle 1408 Brachynltntnwnjmesonii TayL Grifln 955 Mnium thomsonii Schimp. Schofield 94579 & Hedderson ORTHOTRICHACEAE Z ygodoatoideae LeptodontiopsrS ~gil@WizBroth. Lectotype (H--Broth.) Stenomitrium pentastichurn (Mont.) Broth. Mahu 10685 Zygodon intemedius B.S.G. Vin 29362 Zygodon obtusi$olius Hook. Vin 38301 1 Zygodon pungens C. MueU. Zygodan reinwardtii (Homsch.) B .S .G. Orthotrichoideae Bryodkonia petidmenalis Sainsb. Vin 29862 Muelleriella crossi$iolia (Hook. f. & Wiis.) Dusén Engel 7718 subsp. acuta (CMull.) Vitt Orthomimmumtuberculahun Lewinslfy-Haapasaari *Lewinsky-Haopasauri and Crosby & Crosby ( 1996) Orthotrichurn anomaium Hedw. Gofinet 41 15 OrthomChum lyellii Hook. & TayI. Gofinet 3 162 Orrhonichum obtusijioliwn Bnd. Golyinet 4137 Pleurorthotrichum chilense B rotb. Holotype (H-Broth.) Stuneobrywn mirtun (Ltewinslcy) Noms & Robinson MagilZ7054 (NO) Ulota lutea Mitt. Fife 8042 UZota magellanica (Mont,)Jaeg. Schüfer-Verwimp & Verwimp 7970 UZota obtusiuscuia Mac. & Kindb. Goffinet 3161 Macromitrioideae Cardotiella elimbata (Thér.) Gofiet Holotype (PC; Gofinet 1996) Cardotiella quinquefaria (Homsch.) Vitt Vital & Buck 11836 Ceuthotheca cryptocarpa (Bart.) Lewinsky-Haapasaari Holotype (FH) Desrnotheca apicufata @ozy & Mok) Lhdb. Vinas 96-4 Florschuetziella steerei Vin Holotype Groutiellu apicuka (Hook.) Ccum & Steere Gofinet 2764 Groutielia chimborc~ensis(Mitt.) Florsch. Gofinet 1173 Leiomitrium plicatum Mitt. Holotype (NY) kratiu neocaledonica B rotb. Lectotype (H-Bro th .) Macrocoma papillma (Th&.) Vitt Syntype (Hickens 1921-J) Macrocoma tenuis (Hook. & Grev.) Vitt Schüfier-Venvimp 8 150 subsp. sufliwanti~(CMÜU.) Vitt - Macromitrium incu~ifolium(Hook. & Grev.) Schwaegr. Streimann 42622 Macromitn'um longijioliwn Hook. Gofinet 656 Macromitrium richardii Schwaegr. Gofinet 1689 Schlotheimia brownii Schwaegr. Vitt 27485 Schlotheimia recta Hwk. & Wds. Schdfer- Verwimp 9686 Schlotheimia trichornitri. Mitt. Schüfer- Verwimp 6902 L This specimen is tentatively identified as Zygodon pungens CMuil., but this neotropical species is hithexto not known from Afnca (Malta 1926). Chapter six: 137

00000d0000C I-tOdddddP4d4I- dF44r(dd@- t d4F 000000@-I OOC C(drldrlr(*Orir(r OOOO44P-dd4r 0000000000C d0d00d0400C mOOOOOOOmm~ 000000~000C

000 000 000 000 000 O00 III 000 000 rldd Chapter six: 138

0 1010 1 1000$ f O 1OOC 4 IO 1d 1 Iddo 1 l O 100C 4 1414 l 1444l l d l ddv O I 1 1 O I 10000001 OOC Chapter six: 139

Literature cited Man, BH.. and UR. Crosby 1986. Revision of the genus Squmnidim (Musci: Meteoriaceae). Joumal of the Hattori Botunicul Luboratory 6 1: 423476. Baum, D. A., K. J. Sytsma, and P. C. Hoch. 1994. A phylogenetic analysis of Epildh (Onagraceae) based on nuclear n'bosomd DNA sequences. Systemtic Botany 19: 363-388. Blomquist, H. L., and L. L. Robemon. 1941. The development of the peristome in Aulacomnium hetermtichum. Bulletin of the Torrey Botmrial Club 68: 569-584- Boisselier-DuBayle, M.C., J. Lambourdière, and H. Bischler. 1996. Rogeny analysis by isozyzne markers in the polyploid liverwort Plagiochenna rupestre. Canadian Journal of Botmy 74:521-527. Bremer, K. 1988. The limits of amino-acid sequence data in angiosperm phylogenetic reconstruction. Evolution 46: 1801- 18 17. Brotherus, V. F. 1909. Musci (Laubmwse). III, Unterklasse Bryales: II Spezieller Teil. In A. Engler's and K. Rantl [eds.], Die Natürlichen Wanzenfden 1 (3,I, II), 277- 1246. W. Engelmann, Leipzig. 19 10. Musci. Wissenschaftliche Ergebnisse der Deutschen Zentral-Afrika- Expedition 1907- 1908 2: 1925. Musci (Laubrnwse) III. Unterklasse Bryales: II Spezieiler Teil. In A. Engler's and K. Prantl [eds.], Die Natürlichen Paanzenfamilien. 2nd. edition, vol. 1 1, 1-542. W. Engelmann, Leipzig. Churchill, S.P. 198 1. A phylogenetic analysis, classification and synopsis of the genera of the Grimmiaceae (Musci). In: V.A. Funk and D.R. Brooks [eds.], Advances in Cladistics. Roceedings of the fht annual meeting of the WiUy Hennig Society, 127- 144. The New York Botanical Garden, New York. and E. L. Linares C. 1995. Rodromus Bryologiae Novo-Gtaaatensis. Introduction a la Flora de Musgos de Colombia. Parte II. Bibliorheca "Jose Jeronimo T- " 12: 455-924. Cm,H. 1987. Leruria and its place in the Orthouichaceae, subfamily Macrornitrioiàeae. Mernoirs ofthe New York Botanical Garden 45: 602-606. De Luna, E. 1995. The circumscription and phylogenetic relationships of the Hedwigiaceae (Musci). Systematic Borany 20: 347-373. Dixon. H. N. 1934. A contribution to the moss flora of Borneo. Linnean Journal of Botany 50: 57-140. Chapter six: 140

1938. Tropical East Afiican mosses-Pari 1. Acrocarpi. Journal of Botany 76: 249-26 1. Donoghue, M. J., R. G. Ohtead, J. F. Smith. and J. D. Palmer. 1992. Phylogenetic ~lationsbipsof Dipsacales baseci on rbcL sequences. Msof the Missouri Botmiicci Garden 79: 333-345. Fanis, J. S. 1972. Estimating phylogenetic trees hmdistance mat&. Ihe American Naturaikt 1106: 645-668. Fitch, W. M. 197 1. Toward defining thc coune of evolution: Minimum change for a specific tree topology. Systemc~ticZoology 20 406416. Frey. W.,J.-P. Frahm, E. Fischer, and W. Lobin. 1995. Die Moos- und Fampflanzen Europas. Kleine Kryptogamenflora Band IV: W + 426 p. Gustav Fischer Verlag, Stuttgart. Fristch, R 1991. Index to bryophyte chromosome counts. Bryophytonan Bibliotheca 40: 1-352. Goffinet, B. 1993. Taxonomie and floristic notes on neotropicd Macromitrioideae (Orthotrichaceae). Tropical Bryology 7: 149-1%. Griffin, Iü, D. 1990. Two new pentastichous species of Zygodon fiom high elevation in Venezuela Cryptogamie, Bryoi., Lichénol. 11 : 163-168. Hedderson, T. A., R C. Chapman,and W.R. Rootes. 1996. Phylogenetic relationships of bryophytes inferred fiom nuclear-encoded rRNA gene sequences. Plant Systematic and Evolution 200: 213-224. HedenB, L. 1989. Axillary hairs in pleufocatpous mosses -a comparative study. Lindbergia 15: 166-180. 1994. The basal pleurocarpous diplolepideous mosses-A cladistic approach. The Bryologist 97: 225-243. 1995. Higher taxonomie level relationships among diplolepidous pleufocarpous mosses-a cladistic overview. Journal of Bryology 18: 723-781. Iwatsuki, 2. 1959. A revision of the japanese species of the genus UIotu. Journal of the Hattori Botunical Imboratory 2 1: 138-156. Janzen, P. 1917. Die Hauk der Laubmoose. Hedwigia 58: 158-280. Koponen, T. 1968. Generic revision of Mniaceae Min (Bryophyta). Ann. Bof. Fenn. 5: 117-151. 1982. Generic and family concepts in the Mniaceae. Beihefi zur Nova Hedwigia 7 1: 249-260. Chapter six: 14 1

La Farge-Englanà, C. 19%. Growth-form, branching pattern, and perichaetial position in mosses: Cladocarpy and plewoçaxpy redefined. nie Bqdogirn 99: 170- 186. Lewinsky, J. 1977. The genus OrthomChum. Morphological studies and evolutionary remarks. Joumf of the Hanori Botdcal Laborutory 43: 3 1-6 1. 1989. Does the orthotrichaceous type of peristome exist? A study of peristome evolution in the genus Onhomichuin Hedw. with possible derivation of the haplolepideous peristom. Jounzal of the Hanon Botmical Luboratory 67: 335-363. 1990. Zygdon Hodc & Tayl. in AusÛalasia: a taxonomie revision hcludhg SMstuciies of peristomes. Lntdbergia 15: 109-139. 1993. A synopsis of the genus Onh~~chtunHedw. (Musci, Orthotrichaceae). Bryobrothera 2: 1-59. 1994. The genus P le~rorthot~chumBroth. Lindbergia 19: 11-24. 1995. Orfhohl'chumnotabile Lewinsky-Haapasaari, a new moss species hm Sichuan, China Lindbergia 20: 102-105. 1996. OrihoWzkhum holmenii, a new corticolous species nom bzakhstan with comments on Onh~~chwnhnlüi in Asia. The Bryologisat 99: 1-5. ,M.R. Crosby 1996. Orthomitrim fuberculuilwn (Orthotrichaceae), a new genus and species from Guizhou, Chiria Nuvon 6: 1-5. Maddison, D. R. 1991. The discovery and importance of multiple istands of most- passimonious trees. Systematic Zoology 40: 3 15-328. Maddisoa. W.P., and D.R Maddison. 1992. MacClade. Analysis of Phylogeny and Character Evolutioa. Sinauer, Sunderland, Malta, N. 1926. Die Gaüung Zygodon Hook. et Tayl. btv~asUnivers. Bot. D&a Darbi 1: 1-185. 1927. Die Ulota-Arien Sud-Amerikas. Acta Horti Botanici Universitaris Lutviensis 1: 165-208. 1933. A survey of the Australasian species of Ulota. Acta Horti Bot. Univ. ht~.7: 1-24. McDade, L. 1992. Hybnds and phylogenetic systematics. II. The Impact of hybrids on cladistic analysis. Evolution 46: 1329- 1346. McNamara. W.A guide to the nomenclature of heterochrouy. Journal of Paleontology 60:4-13. Meusel, H. 1935. Wuchsforme~und Wuchstypen der euopaischen Laubmoose. Nova Acta Leopoldina Neue Folge 3: 123-277. Chapter six: 142

Mishler, B. D. 1994. Cladistic analysis of molecular and morphological data. American luunial of PhysicaI Anthropology 94: 143-156. ,and E. De Luna 1991. The use of ontogenetic data in phylogenetic analyses of mosses. Advunces in Bryology 4: 121-167. Miller, H.A. 1979. The phylogeay and distribution oftùe Musci. In: G.S.C. Clarke and I.G. Duckett [eàs.]. Bryophyte Systematics, 11-39. Academic Press, London. Minen, W. 1869. Musci Aus-americani. Joumal ofthe Linnean Sociery of Botany 12: 1-659. Nishimura, N., and K. Watanabe. 1992. SEM obse~ationson the peristome teeth of mosses, 2. Isobryales. nit Natural Environmental Science Research 5: 83-101. Noms, D.H., and H. Robiiwn. 1981. Stoneobryum, a new moss genus of ûrthotrichaceae from South Africa and Southem Queensland. nie Bryologist 84: 95- 99. Olmstead. R G. 1995. Species concepts and plesiomorphic species. Systematic Botany 20: 623-630. Pryer, K. M., A. R Smith, and I. E. Skog. 1995. Phylogenetic relationships of extant fersn based on evidence nom morphology and rbcL sequences. American Fern Journal 85: 205-282. huseli, R.A., and W.D.Reese. 1982. The mosses reporied from New Caledonia Journal of the Hattori Botanical Lobaratory 53: 449482 Ramsay. H.P., and D.H. Vitt. 1984. Distribution, cytotaxonomy, and sexuality of Macronitrium in the PaciGc. Journal of the Hanon' Botmical Luborarory 55: 23-33. ,and . 1986. The Macrmitrium-complex in Ausnalasia (Orthotrichaceae: Bryopsida). Part m. Cytotaxonomy. Journal of the Hanori Botmical Luboratory 6 1: 1-43. ,H. Streimann, and D.H. Vitt. 1995. Cytological studies on mosses fiom Papua New Guinea. 1. Introduction and the family Orthotrichaceae. Tkopical Bryology 11: 151-160. Reimels. H. 1954. Bryophyta. In H. Melchior and E. Wedermaan [eds.], A. Engler's Syllabus der Wanzenfamilien, 12th. edition, 2 18-268. Gebrürder Borntriigger, Berlin. Robinson, H. 1964. New taxa and new records of bryophytes from Mexico and Central America. Tlim Bryologist 67: 446-458. 197 1- A revised classification for the orders and families of mosses. Phymlogia 2 1: 289-293. Chapter six: 143

Sainsbii~y,G. O. K 1945. New and critical species of New Zealand mosses. Transacîion of the Royal Suciew of Néw Zealand 75: 169- 185. Saito, K. 1975. A monograph of Japanese Poniaceae (Musci). Journal of the Hanori Botanical Luboratory 39: 373-537. Schofield, WB. 198 1. Ecological sigaincaace of morphological characters in moss gametophyte. ï7w Brydogist M 149-165. Schwartz, O. M. The development of the petistome-forming layers in the Funariaceae. Intemationsl Joumal of PhtScience 155: 640-657. Shaw, J. 1985. Peristorne structure in the Orthotrichaceae. Journal of the Hanon Botanical Lubomtory 60: 1 19-136. ,and J. R. Rohrer. 1984. Endostorniai architecture in diplolepideous mosses. Joumd of the Hattori Botunical hboratory 57: 41-6 1. Swofford, D. L. 1993. PAUP: Phylogenetic Analysis Using Parsimony ,version 3.1.1. Computer Program ditributed by the moisNatural History Swey, Champaign, Illinois. van Rooy, J., and A. E. van Wyk. 1992. A coaspectus of the subfamily Macromitrioideae (Bryopsidâ ORhotrichaceae) in Southern Africa The Bryologist 95: 205-215. Vitt, D. H. 1970. The family Orthotncbaceae (Musci) in North America, north of Mexico. W .D.thesis. University of Michigan, Ann Harbour. 197 1 (1972). The innageneric evolution, phylogeny, and taxonomy of the genus 0nhotrichum (Musci) in Noah America Nova Hedwigiu 2 1: 683-7 1 1. 19'72. A monograph of the genus Dnunmundia. Canadiun Joumal of Botmy 50: 1191-1208. 1973. A revision of the genus Orrhotnchum in North America, noah of Mexico. Bryophytonrm Bibliotheca 1 : 1-208. 1976. A monograph of the genus Muelleriella Dusén. Joumal of the HattoB Botanical Lubomtory 40: 9 1- 1 13. 1979. New taxa and new combinations in the Orthoûichaceae of Mexico. The Bryologist 82: 1- 19. 1980 (1981). The genus Macrocorna 1. Typincation of names and taxonomy of the species. nie Brydogist 83: 405436. 198 la. The gemra Leiomitriwn and Cardotiella gen nov. (Orthotrichaceae). Joumal of the Hattori Botunical Laboratory 49: 93-1 13. 1981b. Adaptive modes of the moss sporophyte. me Bryologist 84: 166-186. Chapter six: 144

1982a The genera of the Orthotrichaceae. Beiheft zu Nova Hedwigia 7 1: 26 1- 268. 1982b. Popdationai variation and speciation in austral mosses. Journal of the Hatton Botanicaf Luborarory 52: 153-159. 1983. The New Zedand species of the pannapical genus Macromifrium (Orthoaicbaceae: Musci): Taxonomy, phylogeny and phytogeography. Joudof the Hanori Botanical Laboratory 54: 1-94.. 1984. Classification of the Bryopsida. In R. M. Schuster [ed.], New Manual of Bryology, Vol. 2. Hattori Botanical Laboratory, Nichinan, Japan. 1990. Desmotheca (ORhoaichaceae): Gondwanan fkagmentation and the ongin of a Southeast Asian genus. Tropical Bryology 3: 79-88. 1993. Orthotricbaceae. In A. J. Sharp, H. A. Cnun, and P. M. Eckel [eds.], The Moss Fiora of Mexico, 590657. New York Botanical Garden, New York. 1995. The genus Calomnion (Bryopsida): Taxonomy, phylogeny, and biogeography. nie Bryologist 98: 338-358. ,and H. P. Ramsay. 1985. The Macrornitrhï complex in Australasia (Orthotrichaceae: Bryopsida). Part 1. Taxonorny and phylogenetic relationships. Journal of the Hattori Botanical Laboratory 59: 32545 1. ,T. Koponen, and D. H. Noms. 1993. Bryophyte flora of the Huon Peninsula, Papua New Guinea. Lm. Ulota and Schlotheimia (Orthotrichaceae, Musci). Acta Botanica Fennl'co: 148: 5-25. ,B. Goffinet, and T.A. Hedderson. 1997. The ordinal classification of the mosses: Questions and answers for the 1990's. Roceedings of the Centenary Meeting of the British Biyological Society. (in press). Walther, K. 1983. Bryophytina. Laubmoose. In J. Gerloff, and J. Poelt [eds.] A. Engler's Syllabus der Paanzenfarnilien. Ad. 13, V(2): 1- 108. Gebrüder Borntriiger, Berlin. Watrous, LE., and Q.D.Wàeeler. 1981. The out-group cornparison method of character analysis. Sysr. W.30: 1- 11. Wyatt, R-,I.J. Odnykoski, A. Stonebumer, H.W.Bass, and O.A. Galau. Allopolyploidy in bryophytes: Multiple origins of . Proceedings of the National Academy of Sciences, USA 85: 5601-5604. Wyatt, Re, IJ. Odrzykoski, A. Stonebumer, H.W. 1993. Isozyme evidence regarding the origins of the allopolyploid moss Plagiomnium curvulatum. Lindbergia 18: 49-58. Chapter sut: 145

Zander, R.H. 1993. Genera of the Pottiaceae: Mosses of harsh environments. Bulletin of the BI(aa10 Society of Natural Sciences 32: 1-378. Zander, R.H., and DH. Vitt. 1979. Gametophytic distinctions of Zygodon (Orthotrichceae) and Anoectangium and kptodonti' (Pottiaceae) and the status of Anoectmgium rubrigenunium of Hawaü. CoMdm Journal of Botuny 57: 292-2%. Zantea van, B.O., and T. P&s. 198 1. Distribution and dispersai of Bryophytes. Advances in Bryology 1: 479-562. Chapter seven: 146

Chapter seven

Bevised generic ciassi6caaOn of the Orthotrichaceae.

The goal of phylogenetic studies is to elucidate trends in cbaracter evolution, in order to bettes understand the relationships atmmg taxa sharing particular charactem. Since characters, morphological or others, form the basis for arr~ulpingtaxa in groups of higher ranls, classincation should reflect the evolutionaq history of the organisms under consideration @onoghue and Cantino 1988). Classincations, and particularly classification of the Bryopsida, have until recently predominmtly relied on comparative morphological studies (see Vin 1984). With the advent of moleculai techniques (see Hillis and Moritz 1990), new sets of characters, a prid thought to occur in large numbers (but see Doyle 1992), and for which homology in more easily established (see Wagele 1995), have become available. Their use in phylogenetic reconsmictions often yield phylogenies connictïng with those obtained fiom morphology (Hillis 1987), but these disc~panciesare often just as striking as the incongruencies between different morphology-based phylogeaies or different gene trees (Patterson. Williams, and Humphries 1993). Two approaches have been proposed for retrieving phylogenetic information hmsuch incongruent data sets (see de Que@ Donoghue, and Kim 1995, for review). The nrst method, taxonomie congruence, is based on the cornparison of alternative mes by constructing a consensus tree (Adams 1972) of ail alternative trees. This me- has been criticized 1) because incongruence may ofien resuit fiom procedural problems (taxon sample, errors in assessing homology, etc.; Patterson, Williams, and Humphries 1993) 2) because of unequa1 rates of evolution (Buil et al. 1993), 3) because it is subjective and not testable (Tehler 1995), and 4) because the consensus tree itself is not a phylogeny, but only a summary of congruence of alternative trees (SwoRord 199 1). Hillis (1987) also argued that a characm set is never informative at all systematic levels, and therefore suggested that data sets should be combined in an attempt to ~solvethe phylogeny based on "total evidence". Combining data sets may however yield strongly unnsolved phylogenies if the data sets combined are of different size (e.g., uneven tamm sample), and the final matrix includes many missing data (see Wihson 1995). In the present study, the generic representation of the Orthotrichaceae is complete in the morphological analysis, whereas only 10 of the 19 genera were Chapter seven: 147 included in the molecular study due to difficulties in obiaining good quality DNA, if any, for many of the srnail, ofien monotypic genera known only fiom the type collection. A combined analysis was therefore consi&& premanue, and will have to await obtaining rbcL sequences for at lest some of the rnissing taxa. The most parsimonious phylogenies obtained hmboth data sets are partly incongruent with regards to the subfamilial and genenc circMptiom withia the Oahotrichaceae. Presenting a tevised classification of the fdybased on conflicthg mes may a prion' appear subjective and thus not scientinc flehler 1995), but is not les objective than the initial choice of the taxa or the characters used to infer the phylogenies. Neither phylogeny is a priori better than the other. Molecuiar characters, single nucleotides in the sequence, derhm their lack of complexity and thus a higher likelihood of homoplasy (Wagele 1995). Furthennore, characters retrieved nom a single gene may not satism the criterion of independence, because not ail changes may be equally probable at a particular site (e.g., in a case of base composition bias) or dong the sequence (functional comtraints) and therefore yield trees incongruent with the actual taxon phylogeny (Doyle 1992). Compared to molecular characters, establishing the homology of morphological characters may be more ciifficuit (see discussion chapter six). Furthermore, the high level of homoplasy found in the most parsimonious scenario (chapter six) may simply be indicative of the inadequacy of the characters iocluded. The subfamüial classification adopted here is based on the rnolecuiar phylogeny (chapter five), whereas the hypotheses derived from the analysis of morphological characters serve as a frarnework for circumscribing the genera and aibes. The Onhotrichaceae are here divided into two subfamilies, the Orthotrichoideae (including the Zygodontoideae) aad the Macromitroideae. Morphological characters suggest that the Zygodontoideae are basal to the dichotomy between the Macromitrioideae and the Orthotrichoideae sensu stricto. The distinction between the Zygodontoideae and its sister-group is based solely on characters of the calyptrae (shape, and anatomy). Though the calyptra of the Zygodontoideae is clearly distinct fiom that found in other orthoaichaceous taxa, the gewtic complexity of this cbaracter may not be very high, since smooth calyptrae do occur in some species 0f0dzo~chm(Lewinsky 1993). and cucullate calyptrae are known fkom Macrorninium leprieuni (Goffinet 1993). Complex charactcrs may be given more weight in phylogenetic aaaiysis (Wagele 1995, see also Kadereit 1994), but given the rather simple anatomy and morphology of mosses, establishing a gradient .of complexity of characters, particularly gametophytic characters is hazardous, at present. A combined Zygodontoideae-Orthotrichoideaewould in the most parsimonious scenario Chapter seven: 148 lack any synapomorphies (see chapter six), and be characterized by such synplesiomorphies as acmcarpy, orthotropic sympodially branching stems, whereas the Macromitroideae include mostiy taxa with apotypic traits such as plagiotropic, monopodially branched stems with prichaetia termiad on lateral branches. The phylogenetic signal carrieâ by two of these tbne characm is compromiscd by the presence, within the Macromitrioideae, of "basal" taxa that are polymorphic for the latter two characters. A possible hybrid nature of these taxa may be at the base of the incongruence regardiog the subfamiiial atiesbetween the two trees (McDade 1992). This hypothesis is expected to be addressed som. The genenc circdptionfollows the discussion presented in chapter six. It is hypothesized that most speciose genera of the Orthoinchaceae (Le., Macromitium, Orthonichum, Zygodon), as weil as the small genus Macrocorna, are cumntly representing evolutionary grades. The fmt systematic implications of this study are proposed hem. with the recognition ai the generic level of all four sections composing the genus Zygodon sensu Malta (1926), as well as the two subgenera of Macrocorna (Vin 1980). The remaining two putative paraphyletic assemblages, Mucromimmumand Orthotnchum, are retained in a conservative sense. pending fiuther study.

ORTEIOTRICHACEAE~niott, Disp. Méth. esp. mousses: 13. 1825 (1826?, fide Margadant 1968)

Zygodoataceae Schimp., Coroil. Bryol. Eur. 39. 1856; Type: Zygodon Hook. & Tayl. Macromitnaceae S.P. Churchül, Bibl. J.J. Trima 12: 588. 1995, syn. nov. Type: Macrornitnum Brid,

Plants pale or dark green to msty brown, 1 to lOcm tail. Stems erect or plagiotropic, branched sympodiaily or monopodialiy; outer cortical ceils typically with very thick and colored walls, parenchyma cells thick-waiied and hyaline to yellowish, to rarely very thick-walled and orange-red in color, central strand lacking. Rhizoids, red, thick-walled. smooth to papiiiose, sparsely to abundantly branched. Axillary hiars with one or two differentiated basai brown cells, and no to several, hyaline, rectangular, thick-wailed ceiis. Leaves erect, appressed or variously flexuose when dry. spreading when moist, spirally inserted or in five ranks, plane, rugose to undulate, flat or keeled, ovate, oblong, lanceolate, to linear. apex acute, acuminate or obtuse, rarely retuse. Costa always Chapter seven: 149

present, single, reacbing apex, to excunent, composed of two ventral guide tells, dorsal cells typically differentiated into substereids, nirely undinerentiated and ~Morophyilose; chlorophyllose cells covering the adaxial daceof the guide ceus typicdy absent. Basal laminal ceils differentiated, rectanguiar and hyaiine to isodiameîric and chlorophyllose, thick-wailed, rarely nodose or porose, flat, smwth to papülose or tuberdate. Upper laminal ceils thick-wded, isodiametric to short iectanguiar, rately long rectangular, flat to bulging, smooth, uaipapillose to pluripapillose, papillae clavate to conical, simple, bifid Dioicous. monoicops (autoicous, farely synoicous or paroicous), or phyiiodioicous. Penchaetia terminal on stem or branches, paraphyses always present, long, hem, ceUs rectangular, apical cell pointed Perichetial leaves weakly to strongly differentiated, oblong-laaceolate to linear-lanceolate; costa percurrent to strongly excurrent and fonning arista; basal cells dinerentiated and often extending to the upper lamina, rectangular to rhombic, hyaline, smooth, tuberculate or prorate, upper cells chlorophyllose. isodiamemc, smooth to papillose. Perigonia axillary or terminal, bud-like. Perigonial leaves snongly differentiated, short, concave; costa often weakly ditferentiated, ending well below the apex; basal cells rliombic, rather thin to moderately thick-walied, hyaline to yellowish; chlorophyllose cells, restricted to apex. Monosetous, seta short or long, sinistrorse to dextrwse at the apex, outer cortical cells thick-wailed, parenchyma cells thick-walled, central strand present to rarely lacking. Vaginula glabrous to ha@, ochrea well developed or absent. Capsule immersed, emergent, to exserted, neck weil differentiated to almost lacking. Stomata with two guard cells, in neck, lower or rarely upper portion of the uni, pore elongate, phaneroporic to cryptoporic, and subsidiary cells weakly to strongly coverhg guard cells. Exothecial ceiis, rectangular, anticlinal and outer periclinal walls moderately to very thick, rarely strongly collenchymateous, inner periclinal wali moderately thick, ceils often differentiated into longitudinal bands, forming ribs, ceils at mouth short, isodiametric to oblate. Annulus typically absent or poorly differentiated. Um cylindric to globose, mouth wide or constricted. Penstome double, simple or lacking. Prostome of 64 cells present or absent. reduced to base of tooth, if weU developed, caducous and peeling. Exostome teeth, 16, free or fused into eight pairs, or continous membrane, long acuminate to mincate, or even absent, erect, reflexed or recurved; OPL of 32 symmtric cells, heavily thickened, papillose, striate, reticulate or lamellate; PUof 16 symmetric cells, moderately thickencd on the exostome, here smooth, papillose, striate, rarcly reticulate. Endostome segments, 16,8 or lacking, inflexed, flat or keeled, lanceolate to lincar, or truncate, he or iùsed into continous membrane, comecting membrane oae to thrre cells high, or Chapter seven: 150

lacking, cilia absent; PPL smwth, thin to very thin; PLof 16 to 32 synunetric to strongly asymmetric cells, papillose to striate papillose, moderately thick. Calyptrae cucuiiate or mitrate, narrowly cylindric, broadly campanulate, to conic, covering the um competly or ody the opercuiwn, smooth to plicate, entire, lobate or modetately to deeply lacerate, thick-wailed ceils unistratose to tristratose, isodiametric to rectangular, smwth, papillose or prorate, hairs lacking or present, miseriate to foliose, cells thick-wallecl, smwth, papillose or prorate. ûpe~:uimflat to rnostiy conic, rostrum short to long, rarely absent, erect or cwed, straight or oblique. Exosporic or endosporic, spores granulose, reticulate or pitted, isomorphic to strongly dimorphic.

1. Orthotrichoideae 1 (see .pllot.tioir p. 154) Zygodontoideae Broth., Nat. W. 1 l(2): 11. 1925. syn. wv.

Plants pale to deep green. Stem typicaUy orthotropic, and sympodially branched, rarely plagiotropic and monopodially branched Leaves erect appressed, twisted, flexuose or crispe& plane. Abaxial costal substereids weli or poorly differentiated, chlorophyllose cells covering adaxial surface typically absent. Laminal cells flat, rarely bdging, smooth, uni- or plurïpapillose, papiilae clavate, conical, simple to bifid, small or coarse. Dioicous, monoicous (autoicous, rarely synoicous or paroicous), acrocarpous, rarely cladocarpous. Sem dextrose or sinistrose. Peristome, double, simple or lacking. Exostome teeth fiee or fused, never reduced to membrane. Endostome fite, comectiog membrane present or absent, never reduced to a continuous membrane. Calyptrae cucullate or mitrate.

A. Zygodonteue (Broth.) Goffiet Basal laminal cells mono- or dirnorphic, evenly thick-walied, rarely nodose, or prose. Chlorophyilose laminai cells flat, rarely bulging, smooth to papiiiose, papiUae one to six, smail cfavate to coarse, simple to bifid at base, aever bifd hma staik. Seta dextrorse or sinistmrse. Calyptrae cucullate, entire, smooth, cove~gupper haif of uni, thick-wailed ceils cells rectangular, and bistratose.

3 1.Codonoblepharon Schimp. , Spec. Musc. Suppl. 2(1): 142. 1824. 4 2. Bryomaltaea Gofiet Chapter seven: 15 1

3. Leratia Broth., 0fters. Finska VetensL-Soc. FW.51A(17): 14.1909. 4. Zygodon Hwker & Taylor, Musc. Brït. 70. 18 18. 5 S. Leptodontiopsis Bmth. in Mildbc. ,Wiss. Ergebn. Deutsch. Zentr. Arf. Exp. 2: 146. 1910. 6. Pleurorthotrichum Broth., &Vers. Finska Vetensk-Akad. Hancil. 47: 1. 1905. 7. Srenomifrium (Mitt.) Brotb., Nat, Pfl. l(3) 464.1902.

B. Orthohicheae offi in et' Basal cells monomorphic, even thick-walled to aodose, never porose. Chiorophyllose lamirial cells nat, never bulging, smooth to papillose. papiliae I or two, rarely more, conical, mely srnali and clavate, simple to bratlched hma stalk. Seta sinistrose. Calyptrae mitrate, entire to deeply lobate, plicate, dysmooth, covering um completely or only operculum, thick-wded cells cells isodiameaic, and unistratose.

1. OrthoiTichum Hedw., Spec. Musc.: 162,1801. 2. Muelleriella Dusén, Bot. NOL 1905: 304. 1905. 3. Orthomitrium Lewinslry-Haapasaari & Crosby, Novon 6: 2,1996. 4. Stoneobryum Noms & Robinson, The Bryologist 84: 96. 1981. 7 5. Ulota Mohr ,Am. Bot. 2: 540. 1806.

II. Macromitrioideae Bmth., Nat. W.1 l(2): 25. 1925 Pseudo-Macmmitrioideae Broth., Nat. M.1 l(2): 49. 1925. Desmothecoideae Crum, Mem. New Yrok Bot. Grad. 45: 603. 1987.

Plants paie to deep green, to mty brown. Stem plagiotropic, creepng to rately pendulous. monopodiaily, rareIy aiso sympodiaily branched. Branches erect, sympodially branched. Leaves erect appressed, twisted, flexuose or crisped. plane to rugose or undulate. Abaxial costal substereiâs weii differentiated, chlorophyllose cells covering adaxial surface absent. Laminal celis fïat, bulging, smooth, uni- or pluripapillose, or tuberdate, papiiiae coaical, or cylinâric, simple to bifid, small or coarse, basal cells porose or not. Dioicous, monoicous (autoicous), mostly phyUodioicous, cladocarpous, rarely also acrocarpous. Seta dextrorse. Peristome, double, simple or lacking. Exostome teeth fke, fkdinto pairs, basal margin Chapter seven: 152 contiguous, or teeth reduced to membrane. Endostome fkee, comec~gmembrane present or absent, qe11 developed to trunctae of reduced to a continuous membrane. Caiyptrae mitrate. rarely cucullate.

8 A. Schlotheimieae Goffi.net Plants deep green to mty brown. Basal laminal cells, smooth or prorate. Upper laminal cells aiways flat and smooth. Cladocarpous. Calyptrae mitrate, smooth, cylindric to conic, lobate base unistratose, bi to tristratose above.

1. Schlotheimia Brid. Sp. Musc, 2: 16. 18 12.

Plants green to deep green. Basal laminal cells, smooth or tuberculate, not prorate. Upper laminal ceUs flat or bulging, smooth, uni- to pluripapillose, papillae simple or bifid at base. Cladocarpous, rarely also acrocarpous. Calyptrae mitrate, rarely cuculiate. plicate, rarely smooth, campaaulate to conic, entire to deeply lacerate, unistratose, thick-wailed cells isodiametric.

1. Macromitrium Brid. Mant. Musc.: 132. 1819- 2. Leiomitrium Min., Phil. Trans. Roy. Soc. Lond. 168: 390, 1879. 3. Cardotiella Vitt, J Hattori Bot. Lab. 49: 10 1, 198 1. 4. Macrocorna (C. Müil.) Grout, The Bryologist 47: 4, 1944. 10 5. Matteria Gofiet 6. Florschuetziella Vin, The Bryologist 82: 16. 1979. 7. Groutiella Sieere, The Bryologist 53: 145. 1950. 8. Ceuthotheca Lewinsky-Haapasaari, Lindbergia, 19: 18. 1994 9. Desmotheca Lindb. J. Lin*Soc. Bot. 13: 184. 1872.

Annotations and nomenclaturaJ changes

1. The subfamilies Orthotrichoideae and Zygodontoideae were described by Brotherus in the same publication, with the Zygodontoideae preceding the Orthotrichoideae (Brotherus 1925). According to article 19.4 of the International Code of Botanical Nomenclature (Greuter 1994), the name Orthotrichoideae is to be preferred Chapter seven: 153 over Zygodontoideae because the Orthotrichoideae contains the type of the famüy: OrthotrrChum Hedw.

2. Zygoàontcic (Broth.) Gofniiet amb. wv. Zygodontoideae Broth., Nat. W. 1l(2): 11. 1925. Type: Zygodon Ho& & Tayl.

3. Codowb~ephuonSchwaegr. Spec. Musc. Suppl. 2(1): 142,1824. Type: C. menziesii Schwaegr. Spec. Mus. Suppl. 2(1): 142,1824. Zygodon Hook. & Tayl. sect. Codonoblepha~n(Schwaegr.) C. Müli. Linnea 18: 669,1845. Thyridim Min. sect CodonOblephanun (Schwaegr.) C. MW.in Jaeg. Ber. S.Gai.i. Naturw. Ges. 1877-1878: 414,1880 (Ad. 2: 678). Zygodon sect. Bryoides Malta, Acta Univers. Lam. 6: 28 1. 1923. syn. nov.

Schwaegerichen (1824) included a single species in Codonoblepharon and defined the genus by the double peristome, composed of 16, paireci, reflexed exostome teeth, and erect arcuate segments, unisexual plants with terminal capituiate gametangia, and cuculiate calyptrae. Muer (1845,1848) regarded Codonoblepharon as a section of Zygodon, and iacluded four taxa. Jaeger (1874) adàed two new species and transferred 7 more hmZygodon. Should Codonoblepharon be later retained at the sectional level. the name Zygodon sect. Codonoblephron (Schwaegr.) CMüil. (1845) has priority over Zygodon sect. Bryoides Malta (1926). Though the name Codonoblepharon was first used by Schwaegrichen (1824), the concept used here for the genus foilows that of Malta for Zygodon sect. Bryoidese. Codonoblephron thus inchdes only species with smooth laminal ceils Only one species fonnerly placed in Codonoblephron is retained here, in addition to C. menziesii: Cudonoblephuron pungens (C-Milil.) Jaeg. (Ber. St. Gall. Naturw. Ges. 1872- 1873: 119. 1874). The type material of the following taxa has not been studied but it is presumed that these tw, belong to Codonobfepham: Zygodon comulensis Lorentz, Bot. Zeit. 24: 187, 1866. 2. gracillimus Fieisch., Musci Flo. Buitz. 2: 392, 19024904. Z. humilis Thw. & Mitt., J. Lin. Soc. 13: 304, 1873. 2. menziesii var. angustifolius Malta, Acta Univers. Latv. 10: 3 17, 1924. 2. microtheca Dixon ex Mdta, Acta Univers. Latv. 10: 3 15, 1924. Chapter seven: 154

2. minutus C. Müll. & Hampe, Linnea 28: 209,1856. 2. parvulus Geheeb & Hampe, humer. Musconun brasiliens. 23,1897.

Zygodon forsten' (Dicks.) Min (Am. Mag. of natu. hist. 2. ser. 8: 32 1, 185 1) included by Malta (1926) in 2. sect. Bryoides, may however not belong to Codonoblepharon. Zygodonforsen' is the only species of this group to acur in the northem temperate zone, whereas all othet species are mostly distributed in subtropical to tropical regions or in the southem temperate zone (Malta 1924). Thus, Z forsteri could have denved hman ancestor bdonging to Zygodon sensu and have lost the papiiiae. Kindberg (1897; Eur. N. Am, Bryin. 2: 3 14315) placed 2. forsteri in own inhgenenc taxon (dnot clear). Material of Z forsten' was not readily available for study. The group, sect. Bryoides sensu MaIta (1926) is in critical need of revision.

4. Bryomaitaea Gomet gen. nov.

Zygodon Hook. et Tayl. afnnis. Planta epiphytica, acrocarpa, autoica Folia Ligulata, ovata-lanceolata, obtusa, erecta-appressifolia. Cellulare superiores subquadratae, papiliosae. Calyptrae cucuilatae, laevis, glabrae. Peristomium duplex. Type: Bryomaitaea obtusifoiia (Hook)Gofnwt comb. nov. Zygodon obt~s~liusHo& Musci Exo.: 159.1820. Type: New Zealand; Knight (lectotype: BM; isotype: H)

Plants sd,to 1.5 cm tall. Stems orthotropic, sympodially branched. Leaves erect- appressed, ovate-oblong, obtuse. Costa ending below apex, with two ventral guide cells, covered on the abaxial surface by rather short to isodiametric chlorophyliose ceiis. Basal cells weakly or not differentiated from upper laminal ceils. Upper laminal celis isodiametric, thick-walled, strongly buiging, coarsely papiilose, papillae two or three, corse, bifid (or in pairs). Acmcarpous, Autoicous, or dioicous. Penchaetid leaves weakly differentiated. Capsule exserted, urn ribbed, stomata phaneroponc, in neck. Peristome double, altemate. Exostome of 16 teeth hised into eight pairs. Endostome of 8 segments. Annulus none. Operculum conic, with a short oblique, rostnim. Calyptrae cucullate, smooth, with prorate cens at apex. Bryomaltaea is easiiy disthguished from Zygodon by its erect app~ssedand obtuse leaves. One specimen from Tiiailand (Touw 8473, ALTA) Mers from typical B. obtusifolia by king dioicous. Whether B~omaltaeuis indeed monotypic as suggested Chapter seven: 155

by Malta (as 2. sect. ob~ofius,1926) needs to be criticaiiy reexamined (see Lewinsky 1990).

5. Leptodontkpsis Broth. Leptodontiopsis was descnkd by Brothem for L.fragiIï$olia Broth., a species endemic to the high elevations in East Africa (Brothenis 19 11). Leptodontiopsis Mers fiom Zygodon sect. Zygodon mainly by the smooth capsule, and dimorphic basai ceus. Leptodontiopsis may a priori not deserve generic distinction but considering a possible a common evolutionary history with Sten~mit~um,and Pleurorthotnfhm(chapter 6) it should tentatively be retained as a distinct genus. Malta (1926) described Zygodon frcigili$olius based on two collections from Mount Kilimanjar0 and attributed the species to Brothem. His description fits that of the type of Leptodontiopsisfragilifolia by Brothem (19 1 1): both have fiagile leaf apices, and dimorphic basai cells. 1have not yet seen the type of Z.fragiiif0Iiu.s Broth. ex Malta, but if this specimen is identical to Leptodonti~psisfragi&io~iaBroth., the latter name would have prïority, if Leprodontiopsis should be synonymized with Zygodon.

6. Oitbotricheae Goffinet trib. nov. Planta acrocarpa. Calyptrae mitriformae. Type: Onhotrichum Hedw.

7. nota Mohr Bryodixonia Sainsb. Trans. Roy. Soc. New Zeaiand 75: 177. 1945. sga nov. Ulotu perichuetialis (Sainsb.) Goffiet mmb. nov. Bryodixonia perichaetialis Sainsb.. Trans. Roy. Soc. New Zealand 75: 177. 1945. Type: New Zealand, "on bark of subalpine scnib, Mount Egmont, ca. 4000 feet; cou. G.O.K.Sainsbury, January 1945, no. 6005".

1 have not yet seen the type of this species, but have however examined the collections kept in ALTA. This species is very distinctive, by its Ulota-like habit, immened capsules and well differentiated perichaetîal leaves.

8. Scblotheimieae Goffhet trib. nov. Planta cladocarpa Calyptrae miaiformae, laevis. Type: Schlothemia Brid. Chapter seven: 156

A sister group relationship of Schlotheiinia with regard to the other Macromitrioideae was strony supported basad on analyses of rbd. sequence data. Though this relationship was not suppoRed in the most parsimonious scenario usiag morphological characters, Schlothemtia was the only large genus the monophyly of which withstood both analyses. Moxphologically the genus is weil differentiated Born the Ma~rorni~um- complex, and the Uiva&tbly weii dcveloppcd peristome suggests that Schloiheimio is the lest derived in the Uacromitmideae.

9. Macmmitrieae Goffiet trib. nov. PIanta cladocarpa vel clado- et acrocarpa Calyptrae mitrifomae, plicatis. Type: Macromitnum Brid.

10. Matteria Goffiet gen. nov. Macroconia (Broth.) Gmtafnnis. Planta epiphytica, autoica, acro- et cladoc- Folia patula (patens) y squarrosa. Ceiiulae costaiis adaxialis eiongata. Peristomü duplex. Sporae multicelluiae. Type: MatteRa gracillima (Besch.) GoEi.net MacromihiM, subg. Trachyphylllum Broth. in Engl. & Pr., Nat. W. l(3): 478,1902- Macrocomu (Hornsch. ex. C. MW.)Grout subg. Trachyphyifum(Broth.) Via, The Bryologist 83: 433, 1980.

a Matteria gracillima (Besch.) Gofiet comb. nov. Schlotheimia gracillim Besch., Buii. Soc. Bot. France 32: LM. 1885. Type: "Patagonie occidentale, île Weilington: Port Eden, 24 janvier 1879 (Dr. Smtier, no. 1838 e. p.)" (holotype-BM-Besch.). Macromitrium grucillimum (Besch.) Broth. in Engl. & Pr., Nat. W. l(3): 478, 1902. Macrocorna gracillima (Besch.) Vin, The Bryologist 83 (4): 433, 1980. Specimen axmined Chile: Llanquihue: Volcan Osorno, 10.5 km from junction of road Ensenada Yerbas Buenas, on road to ski slopes, f 940 m. 41°05'S,72'33' W.on fallen branch, James and Andrea Solomon 4602 (ALTA); Valvidia: Bog of Astelia, Donatia and Gaimardia with scattered patches of Sphagnm at summit of El Mirador, ca. 100 m. southwest of refugio, Cordiliera Pelada, near road between LA Union and Punta Hueicolia, alt. 1000 m., 40°07'S, 73'33 W, on tree in Fitzroya dominated woods, 19 Chapter seven: 157

January 1976, Crosby 12874 (ALTA); Crumao, Cordillera Pelada, 1ûûû m, 29 J. 1934, P. A. Hollennayer (I);Cordillera Pelada (S. Juan), lûûû m, 30.1.35, P. Ath. Wemayer(J); Ida Magdalena; Grenzfluss., 13.12.39, G.H. Schwaae no. 33k (0;Cordillere Pelado, Febnüu 1935, A. Hollermclyer (S).

b. Maiteria papillosa (Thér.) Gofï5.net comb. nov. Macr~mi~umpapiI10sum Thér. in Hen., Arch. Esc. Fam. Fac., Cienc. Med. Cordoba 7: 50, 1939. Type: "Prov. Chiloé, Dep. Llanquihé: Pétrohué, leg. C-C- Hosseus (n. 52 1)" (lectotype-PC-Thér.; isoîype-J!). Macrocoma papillosa (Thér.) Vitt, The Bryologist 83 (4): 433, 1980. Specimen examine& Chile: Parque Nac. Nahuel Huapi. Neuquén. Puerto Cihtaros. Bosque costero bordeando el brazo Blest con Nothofagus dombeyii, Saxegothaea conspicua, hurelia philippiara, &ara lanceolata, etc. 4l0û4'S,7l049W. 750 mm, 13/3/95, on small branches of Saxegothaea, Matteri 652 1, Schiavone (Musci Patagonici Exsiccati-ALTA); Puerto Moutt, Saltos de Petrohue on Guerina auellana, bark, shaded, 150 m, laauary 1979,O. van Zanten 7-901143 (ALTA); DC Region, Nationalpark Viilarrica, Nothofaguswald, on Nothofagus; 920 mm, 9. Januar 1987, Schafr-Venvimp & Verwimp 8144 (ALTA).

In addition to the characters presented by Vitt ( 1980), Matteria (as M. subg. Trachyphyllum) is easily disiinguished nom Macrocona (as M. subg. Mucrocorna) by the muhicellular spores. The caiyptra of Matteria is subentire (i.e., laceration dong the plications, are restrïcted to the basal portion of the calyptra) but has one major slit expancihg beyond the middle. This situation has also been observed in species of Macrocoma (e.g., M. orthotrichoides Pedersen 13393-ALTA). Whether the calyptrae are actualiy cucullate is difficult to ascertain, and more material may be needed to cl* this. The cucullate nature of the caiyptrae rnay however not be surprising considering that the rostrum is somewhat oblique and that in other genera of the Orthotrichaceae, a cucdate calyptra is often found in species with oblique rostra.

Key to the genera of the Orthotrichaceoe

1. Stems orthotropic, sympodiaily branched ...... 2 1. Stems plagiotropic, monopodiaily branched, rarely also sympodiaily branched ...... 13 Chapter seven: 158

2. Laminal cells smooth ...... 2. Laminai ceus papillose ...... 6 3. Calyptrae cuculiate, seta dextrorse below urn, capsule ewerted ...... Codonoblephron 3. Calyptrae mit-, seta sinisûorse or capsule immcrsed ...... -4. 4. Lamina bitratose, um smooth, spores multicellular ...... C.CC.C.CMuelle 4. Lamina Wtratose, if bistratose, um nbkd and spores unicellular ...... 4 5. Dorsal ceils of costa lïnear, with pointed ends, leaves fiexuose ...... Stoneobryum 5. Dorsal ceils of costa imgularly rectangular, not linear, leaves ereft appressed. rarely

flexuose ...... , ~..~.-.~.~.~-~-~C.~*~*--*~*.e~~.~*~-*~~.~-e-~---~CC...... Orth0mrnch~m 6. Basal marginal celis differentiated hminner cells, hyaline, quadrate, with strongly thickened, horizontal anticlinal wails ...... Ulota 6. Basal marginal ceils not differentiated eom inner cells ...... 7 7. Basal ceiis dimorpbic, strongly incrassate, nodose or porose, yeilowish cells altemating with evenly thickened hyaline cells ...... 8 7. Basal cells monomorphic, if nodose, then cells not fomiing alternathg bands of yeilowish and hyaline celis ...... 9 8. Urn smooth, long exserted, gymnostomous, perichaetial leaves weakly differentiated, endemic to hi& elevatiom in East AErica and Borneo ...... Leptodontiopsis 8. Um ribbed, ernergent, peristome double. perichaetial leaves strongly differentiated, endemic to Central Chile ...... Pleuronhomchum 9. Lamina1 cells strongy bulghg ...... IO 9. Laminal celis flat ...... -11 10. Leaves green, short, ovate oblong, broadly obtuse, um nbbed ...... Bryomaltaea 10. Leaves orange-brown to green, ovate lanceolate, narrowly obtuse, urn smooth ...... kratia 11. Dorsal ceils of costa linear, with pointed ends, leaves ffexuose, cdyptrae cucullate, seta dextrorse ...... Zygodun 1 1. Dorsal cells of costa irregularly rectangular, not linear, leaves erect appressed, rarely flexuose, calyptrae mitrate, seta sinistrorse...... 12 12. Capsule globose, constricted at mouth, stomata cryptoporic, restricted to upper half of ...... Orthomrtrrum*. 12 Capsule cylindric to ovate, if globose, than not conaicted at mouth, stomata phaneropric or cryptoporic, restricted to the lower haif of the uni ...... ,...... 0rthotntnchum 13. Calyptrae lobate at base ...... !. 4 Chapter seven: 159

13. Calyptrae entire to lacerate ...... 16 14. Basal marginal ceiis hyaline, quacirate, with strongly tùickened, horizontal anticlinal walls, forming multiseriate margin, cdyptrae unistratose ...... Ulota 14. Basal marginal cells not dinenntiated, or if differentiatd, than not quadrate and . not in several mws, calyptrae bi- to trisiratose ...... 15 15. Leaves decmnt, deCumencies composed of two to thme rows of large, infiated hyaline, papülose to tuberdate cells, dioicous ...... Cardotiella i5. Leaves not decurrent, and hyaline basal marginal ceIis not papillose or tuberculate, -. phyliodioicous ...... *...... -...... ~....SchIothetm~~~ 16. Stem and branch leaves monomorphic (only shape is considered not the size) ..... 17 16. Stem and bnuich kaves conspicuously dimorphic ...... 2 1 17. Basal marginal cells hyaline, quadrate, with strongly thickened, horizontal anticlinal walls, forming multiseriate masgin, calyptrae unistratose ...... Ulotu 17. Basal marginal ceils not dinerentiated, or if differentiated, then not quadrate and not in several rows, calyptrûe bi- to tristratose ...... 18 18. Abaxiai layer of the costa composed of stmngly bulging, papillose, isodiametric ceils almost to the base ...... Fïorschuteziella 18. Abaxiai layer of costa composed of substereiâs (ir, flat, smooth, linw celis) ...... 19 19. Basai ceils rectangular, dimorphic, strongly hcrassate, nodose, porose, yellowish cells aitemating with evenly thickened hyaline celis, calyptrae cucullate ...... C...... I,,...... ,...... Stenom~tnum 19. Basai cells, except for the most proximsl ones, isodiametric, monomorphic ...... 20 20. Leaves squarrose-mcurved, to widely spreading, abaxial surface of costa composed of elongate ceils into apex, spores muiticelluiar ...... Mutteria 20. Leaves ercct-appressed, abaxial surface of costa covered with isodiametnc cells, spores uniceiiuiar ...... - 2 1. Marginal ceiis linear and hyaline, forming a distinct border at least in the lower third of the le& ...... Groutiella 21. Marginal celis not differentiated into a border of linear, hyaline cells ...... 22 22. Laminal cells smooth ...... 23 22. Laminal cells papillose ...... 26 23. Laminal ceiis fiat ...... Schlothetmta*. 23. Laminal cells bulging ...... 24 24. Perichaetial leaves aristate, capsule immersed ...... Ceuthothecu 24. Pericbaetial leaves not with long excurnnt costa, if aristate, capsule exserted ...... 25 Chapter seven: 160

25. Leaves decurrent, decumncies composed of two to three rows of large, inflated hyaline, papillose to tuberdate ceUs, basai ceiis not differentiaied, short, isodiametric, oblate, chlorophyllose ...... Cordotieila 25. Leaves not decunent, if decurrent thea cells liwar and chlorophyllose, or decurrency composed of a single row of hyaüec, bulging cells, basai cells differentiated hm . * upper cells, rectangular, hyaline ...... Macromttnum 26. Branch leaves of swile and perichaetium bearing branches dimorpbic ...Desmotheca 26. Brmch leaves moaomorphic ...... - 27. Basal laminal ceIls long nctanguiar, hyaline ...... Ma~r~mttn~m 27. Basal laminal ceiis, except the most proximal ones, short, isodiametric to oblate, chlorophyUose ...... 28 28. Leaves decumnt, decurirncies composed of two to ihree rows of large, iaflated hyaline, papiliose to tubernilate cells, basal cells uot differentiated, short, isodiametric, oblate, chlorophyliose ...... 28. Leaves not decurrent, if decunrnt then celis linear and chlorophyllose, or decurrency composed of a single row of hyaline, bulging ceus, basal cells differentiated £tom upper cells, rectangular, hyahe ...... A 29. Capsule ribbed, dioicous ...... 2 29. Capsule smooth ,phyllodioicous ...... Macromitrium subg. Cometium Mitt.

Literature cited Adams, EN. 1972. Consensus techniques and the cornparison of taxonomie trees. Syst. Zool. 21: 390-397. Bull, J.J., JS. Huelsenbeck, C.W. Cunnigham, DL. Swofford, amd P.J. Waddeiî. 1993. Partitionhg and combining data in phylogenetic analysis. Syst. Biof.42: 383-397. de Queiroz, A., M.J. Donoghue, and J. Kim. 1995. Separate versus combined analysis of phylogenetic evidence. Am. Rev. Ecol. Syst. 26: 657-68 1. Donoghue, M.J., and P.D. Cantho. 1988. Paraphyly, ancestors, and the goals of taxonomy: a botanical defense of cladism. Tlre Botanical Review 54: 107- 128. Doyle, J.J. 1992. Gene trees and species mes: molecular systematics as one-character taxonomy. Syst. Bot. 17: 144-163. Gofiet, B. 1993. Taxmomic and fioristic notes on neotropical Macromitrioideae (Orthotrichaceae). Tropical Bryofogy 7: 149- 154. Greuter, W. (chairman) 1994. International Code of Bosanical Nomenclature (Tokyo Code). Koeltz, Konigstein. Chapter seven: 16 1

Hillis. D.M. 1987. Molecular versus rnorphological approaches to systematics. AM. Rev. Ecol. Sysr. 18: 23-42. Hillis. DM.,and C. Moria. 1990. Molecular Sy stematics. Sinauer. Sunderland. Kadereit, J. W. 1994. Molecules and morphology .phylogene tics and genetics. Bot. Acta 107: 369-373. Lewinsky, J. 1993. A synopsis of the genus Orthotnchum Hedw. (Musci, Onhouicbaceae). Bryobrothera 2: 1-59. Malta, N. 1926. Die Gatiung Zygalon Ho&. et Tayl. Latvijas Univers. Bot. Ddna Durbi 1: 1-185. Margadant, W.D. 1968. Early Bryological Literature. Mededelingen van her Botanisch Museum en Herbarium van der Rîjkrntiversiteit te Umcht 283: xi + VIII + 1-278. McDade, L. 1992. Hybrids and phylogenetic systematics. II. The Impact of hybrids on cladistic analysis. Evolution 46: 1329-1346. Patterson, C.. DM.Williams, and CJ. Humphries. 1993. Congruence btewenn moiecular and morphological phylogenies. hn. Rev. Ecol. Syst. 24: 153- 188. Tehler, A. 1995. Morphological data, molecular &ta, and total evidence in phylogenetic aiialysis. Canadian Journal of Botany 73 (Suppl. 1): S667-S676. Vin, D.H. 1980 (198 1). The genus Macrocorna 1. Typification of names and wonomy of species. nie Bryologisr 83: 405436. 1984. Classification of the Bryopsida. In R M. Schuster [ed.], New Manual of Bryology. Vol. 2,696-759. Hatton Botanical Laboratory. Nichinan, Japan. Wilkinson. M. 1995. Coping with abundant missing entries in phylogenetic inference using parsirnony. Syst. Biol. 44: 501-5 14. Chapter eight: 162

Cbaptet eight

Conclusion

The Bryopsida (sensu Vin 1984) with over 9000 species (Schofield 1985), are as diverse as the Fiiicopsida (sensu Mickel1982) and form the third largest class of land plants afkr the Magnoliopsida and Lilicopsida (sensu Cmnquist 1982). Although the relationships of mosses to other gnxn plants have been the focus of many studies over the last decade (e.g., Capesius, 1995; Hedderson, Chapman, and Rootes 1996; Krantz et al. 1995, Manhart 1994; Mishler and Churchill 1984,1985; Mishler et al. 1992.1994; Waters et al. 1992),the monophyly and the phylogeny of the Bryopsida has drawn little attention since the advent of molecular techniques. Addressing the evolution of mosses ai the ordinal and faniilid level, necessitates the identification of major moaophyletic groups and the selection of appropriate exemplar taxa (Mishler 1994). The Bryopsida are currently divided into 16 orclers (Vitt 1982; 14 of these as suborders in Vin 1984), of which 14 share an arthrodontous peristome. Vitt (198 1) recopized four major types of aahrodontous peristomes: the Funaria-type, Orthotrï'chum-type, Bryzun-type, and the Dicranum-type. The direction ofperistomial character transfonnations ammg these types is not agreed upon (Crosby 1980, Lewinsky 1989, Shaw and Rohr 1984, Vitt 1984.). The Oahotnchales as currently dehed (Vitt 1984) are heterogeneous with regards to the architecture of the peristome (see Edwards 1979,1984, Vin 198 1, Shaw 1985, Lewinsky 1989), and their monophyly needs to be critically reexamined before the order and its type family cmbe adequately represented in any higher level phylogenetic study.

Circumscription of the Ottûo~chaies.The ûrthotrichales sensu Vitt are composed of five fdes:the Erpodiaceae, Helicophyllaceae, Microtheciellaceae, ûrthotrichaceae, and Rhachitheciaceae. Except for the Orthotrichaceae, these families are composed of gymnostomous taxa or taxa with a peristome that a prion' is incompatible with any of the four main types. The family is shown io be of polyphyleac ongin (chapters thrw and five). The Microtheciellaceae are excluded based on morphology: the only species. M. kerrei, is clearly pleurocarpous. The family may be related to the Neckeraceae. aud is therefore tramfermi to the Leucodontales. Critical examination of the peristome of the Erpodiaceae (Edwards 1979; pers. observations) and of the Rhachitheciaceae (chapter three), suggests that these families may be oniy distantly related to the ûrthotrichaceae. Tbese hypotheses wen subsequently suppoacd by cornparisms of nucleotide sequences of the chloroplast gene, rkL, and these families, the Rhachitheciaceae and the Erpodiaceae, are tramferrad to the baplolepideae, with affhities to the Seligeriales (chapter thne and five). The monotypic and gymnostomous Helicophyilaceae are charactetized by a unique combination of morphologicai characters, and are cmidered ody distantly related to the Oahotrichaceae- Though m01ecuIar data are not yet available for this taxon, the fdyis tentatively excluded hmthe order. The ûrthotüchales are thus here considend nstncted to the Orthotrichaceae smnc Vitt (1984).

Clrcumscription of the Orthotrichaceae. The Orthotrichaceae, one of the largest families of anbrodontous mosses, were prior to this study, composed of 26 genera (Lewinsky 1976, Lcwinsky 1994, Lewinsky-Haapasaari and Crosby 1996, Via 1984, Zander 1993) distributad among four subfamilies (Vin 1972). The heterogeneity of the family, as expressed by the great variation in rnorphological characters, is best explaineci by its polyphyletic ckumscription (chapters two and three). Critical examination of morphological characters based on type material leads to 1) the synonymy of the monotypic Pleurozygodontopsis with Zygodon (2.reinwardtii, chapter two), and 2) the exclusion hmthe ûrthotrichaceae of Kleioweisiopsis (Diaichaceae, chapter two), Tn'gonoàic~on (Grirnmiaceae, chapter two), OctogoneUo and UIea~tnrm(Rhachitheciaceae, chapter three). The systematic afnaties of gymnostomous taxa such as Amphidicm and Drununondia were addressed using molecular techniques. Based on a cladistic anaiysis of rbcL nucleotide sequences, both genera are shown to be of haplolepideous origin and only distantly relateci to the ûrthotricbaceae (chapter five). The Orthotrichaceae are thus restricted to 19 of the original genera.

M@or phylogenetic reiationships. Cladistic analyses using either molecular or morpho1ogicaI characters provide smng support for the monophyly of the ameuded Orthotrichaceae, L'ather than for a polyphyletic origin of the family (Churchill and Linares 1995, De Luna 1995). The relationships among the 19 genera are, however obscure. Based on rbcL sequence data, the Zygodontoideae and the (hihotrichoideae form a monophyletic clade sister to the Macromitrîoideae, wheteas morphological characters support a close relationship between the latter two subfdes with the Zygodontoideae sister to this combined clade. It is argueci that uncertainties about the homology of certain morphological character-states may be at the ongin of the discrepancies between both Chapter eight: 164

analyses. nie subfamilial classification of the Orthotrichaceae therefore follows the gene phylogeny, and the family is divided in two subfdes, the Orttiotrichoideae and the Mmmitrioideae (côapter seven). The four most speciose genera of the Oahotricbaceae, namely M~crorni~um, Orthomchm, (Ilotu, d ZygorlOn, as weil as the small genus M~u=rocomaare suggested to merely represent evolutionary grades (chapter five and six). The monophyly of Ulota is restored with the inclusion of the monotypic genus Bryudixonia (chapter seven). The heterogeneity of Orthotrichum and MixrominiMI nads Merstudy, whereas new genera are proposed to accommodate the variation found in Zygodon and Macrocomu (chapter seven). The Orthoûichaceae are now considered to be composed of 22 genera The relationships between the genera remain mostiy ambiguous. Moiphologicai stasis (see Williamson 1987, and Lanon 1989) associated with decoupled molecular and morphological evolution may lead, in a classincation system based solely on comparative morphology, to the recognition of grades rather ihan natural groups (e-g., Zygodontoideae and Onhom'chum). Similarly, the inniosic nature of the molecular character, that is the simplest and thus least robust character available, may result in higher levels of homoplasy (Wagele 1995) and thus yield false phylogenies due to long branch effects (Hendy and Penny 1989). A revised classincation of the ûrthotrichaceae is presented (chapter seven). It iricorporates relationships strongly supported by at least one of the data sets (e.g., sister- &rwp relationship between Schlothehia and the odrr M2~c~omitrioideae)as weil as "default" aff?nities(e.g., monophyly preferred to paraphyly of the Zygodon-complex as suggested by rbcL data). The Orthoüichaceae are thus considered composed of two subfarnilies, each including two tribes: the Zygodonteae and the Orthotricheae (Orthotrichoideae) and the Schiotheimieae and the Mmmitrieae (Macromitrioideae).

This study is the first attempt to examine the genenc circumscription of an order and a large family of mosses, using molecular data. DNA sequences are shown to offer a unique opportwuty to solve the systematic affinities of taxa lacking characters that are centrai to the classification of mosses. Carefiil morpbological cornparisons of taxa remains nevertheless essentiai to phylogenetic studies because molecules a priori tell us something about relationships but nothing about character evolution, and kause most hypotheses are rnpaainghil oniy if placed in a morphological context. The results presented here together with those obtained in a parallel study using 18s gew sequences (Heddersoa et al. unpubl.) formed the basis for a nMsed classification of the Bryopsida (Vitt, Goffinet, and Hedderson 1997). Chapter eight: 165

Literature cited Capesius, L 1995. A molecuiar phylogeny of Bryophytes based on the nuclear encoded 18s rRNA genes. Joumal of PhtPhysiology 146: 59-63. Churchiil, S. P., and E. L. Linares C. 1995. Prociromus Bryologiae Novo-Granaiensis. Iauoduccion a la Flora de Musgos de Colombia Parte IL Biblioùeca "JoseJeronimo Trirara" 12: 455-924. Cronquist, A. 1982. Magnoliophyta. In S. P. Parker [ed.] Synopsis and Classification of Living Orgmisms, vol. 1,357-487. McGraw-HiU.New-York. Crosby, M. R. 1980. The diversity and relationships of mosses. In R. J. Taylor and A. E. Leviton [eds.], The Mosses of North Amenca, 115-129. Califomia Academy of Science, San Francisco, California De Luna, E. 1995. The circumscription and phylogenetic relationships of the Hedwigiaceae (Musci). Systematic Botany 20: 347-373. Edwards, S. R. 1979. Taxonomie imphcations of cell patterns in haplolepideous moss peristomes. In G. S. C. Clarke and J. G. Duckett [eds.], Bryophyte Systematics, 3 17- 346. Academic hss,London. 1984. Homologies and inter-relations of moss penstomes. In R. M. Schuster [ed.], New Manual of Btyology, vol. 2.658695. Hattori Botanical Laboratory, Nicbinan . Hedderson, T. A., R. C. Chapman, and W. R. Rwtes. 1996. Phylogenetic relationships of bvophytes infened from nuclearacoded rRNA gene sequences. PIant Systematic and Evolution 200: 2 13-224. Hendy, M. D., and D. Penny. 1989. A f!iamework for the quantitative study of evolutionary trees. Systematic Zoology 38: 297509. Larson, A. The klationship between speciation and morphological evolution. In: D. One, and I.A. Endler [eds.]. Speciation and its consequences, 579-598. Sinauer, Sunderland., Lewinsky, J. 1976. On the systematic position of Amphidium Schimp. Linàbergia 3: 227-23 1. 1989. Does the orthotrichaceous type of peristome exist? A study of peristome evolution in the genus Orthotnchum Hedw. with possible derivation of the haplolepideous penstome. JoumI of rk Hattori Butanical Labotatory 67: 335-363. 1994. The genus Pleumnhotrichm Broth. Lindbergia 19: 1 1-24. Chapter eight: 166

MA.Crosby 19%. Orthumitim Nberculatzun (Orihotnchaceae), a new

genus and species kmGuizhou, China. Novon 6: 1-5. Krantz, ED., D. MiLs, M.-L.Siegler, 1. Capesius, C.W. Sensen, and V.A.R. Huss. 1995. The origin of land plants: Phylogenetic relationships among charophytes, bryophytes, and vascularJ. plants infeffed fkom Complete small-subunit nbosomd RNA gene sequences. Md EvoL 41: 74-84. Manbart, J. R 1994. Phylogenetic analysis of green plant rbcL sequences. Molecular Phylogenetics and Evolution 3: 114-127. Mickel, J.T. 1982. Filicophyta. In S. P. Parker Ed] Synopsis and Classification of Living Organisms, vol. 1,341-347. McGraw-Hill, New-York. Mishler, B. D. 1994. Cladistic aiidysis of molecular and morphological data. Amerïcan Journal of Physical hthrqdqy94: 143-156. ,and S.P. Churchill. 1984. A cladistic approach to the phylogeny of bryophytes. Brittonia 36: 406424.

t 1985. Transition to a land flora: phylogenetic relationship of the green algae and bryophytes. Cladistics 1: 305-328. , P. H. Thrall, I. S. Hopple, Ir., E. De Luna, and R. Vigalys. 1992. A mloecular approach to the phylogeny of bryophytes: Cladistic analysis of chioroplast- encoded 16s and 23s ribosomal RNA genes. The Bryologist 95: 172-180. , L. A. Lewis, M. A. Buchheim, K. S. Renzagiia, D,JT Garbary, C. F. Delwiche, F. W. Zechman, T. S. Kantz, and R L. Chapman. 1994. Phylogenetic relationships of the "green algae" and '%ryophytetl.Annals of the Missouri Botanical Garden 8 1: 45 1-483. Schofield, WB. 1985. Introduction to Bryology . Macmillan Publisbing Company, New York. Shaw. J. 1985. Peristome stmcture in the Otthotrichaceae. Jouml of the Hattori Botmical Laborutory 60: 1 19-136. ,and J. R. Rohrer. 1984. Endostomial architecture in diplolepideous mosses. Journal of the Hattori Botmical kibomtory 57: 4 1-61. Vitt, D. H. 1970. The family ûrthotrichaceae (Musci) in North America, north of Mexico. Ph.D. thesis. University of Michigan, AM Harbour. 1972. A monograph of the genus Drt(lllltlondiu. Canadian Joumal of Botany 50: 1191-1208. 1973. A revision of the genus Onhom*chumin North Amenca, north of Mexico. Bryophytorum Bibliotheca 1: 1-208.

Chapter eight: 167

198 1. Adaptive modes of the moss sporophyte. The Bryoiogist 84: 166-1 86. 1982a. Sphagnopsida and Bryopsida In S. P. Parker [ed.] Synopsis and Classification of Living Organisms, vol. 1,307-336. McGraw-HiU, New-York. 1982b. The genera of the Orthotrichaceae. Beihejt zu Nova Hedwigia 7 1: 261- 268. 1984. Ciassification of the Bryopsida. In R M. Schuster [ed],New Manual of Bryology, Vol. 2,696759. HattoriJ. Botanical Laboratory, Nichinan, Japan. ,B. Goffinet, and T.A. Hedderson. 1997. The ordinal classification of the mosses: Questions and answers for the 1990's. In: N.W. Ashton, I.W. Bates, and J.G. hiclett. Bryology for the Second Century. (in press). Wagele, J.W. 1995. On the information content of characters in comparative morphology and molecular systematics . Zool. Sys. Evol. Research 33: 4247. Waters, D. A., M. A. Buchheim, R A. Dewey, and R. C. Chapman. 1992. Preliminary inferences of the phylogeny of bryophytes hmnuclear-encodeci nisoma1 RNA sequences. Aniericm loudof Botany 79: 459466. Williams, P.G. 1987. Selection or constraint?: a proposal on the mechanism of stasis. In: K.S.W. Campbell, and M.F. Day [eds.], Rates of Evolution, 129- 142. Aiian & Uwin, London. Zander, R. H. 1993. Genera of the Pottiaceae: Mosses of harsh environments. Bulletin of the B@do Society of Naturai Sciences 32: 1-378. Appendix 1. RbcL gene sequences of 37 arthrodontous mosses.

10, 20> 30> 40, 50, 60, 70, 8Or Funaria hygrometrica ??????????????????????????????GCPCGA~~AAAGATTACCGATTAACrPA'PPACACPCC Funaria apophysata ??????????????????????????????.T...t~~.~.~~~,m...... *9v.***qA...... ??????????????????????????????.Te...... ,.v...... ~...~.. TA.,,,,,..^^..,.,..,. Tayloria lingulata ??????????????????????????????.T...... TA.,,,^^.,,^..,,^..,^ Heàwigia ciliata ??????????????????????????????.T...... A..,....,...... ~ mium thmonii ??????????????????????????????...... G....G.. .. *.A..*...... Abietinella abietiena ...... A...... Schlotheimia tecta ??????????????????????????????.T~...... *.A.. *...A.....*...... Schlotheimia trichornitria ??????????????????????????????1T...... ~....~...... A,...~..~...... Schlotheimia brownii ??????????????????????????????~T~....I...... ~....l..A...... 1...~..1 Cardotiella quinquefaria ??????????????????????????????.Ttt,,...,,..,...~~,~,~,.~~,..~~A.~~~.~.~..~...... ~acxocanatenue ??????????????????????????????.T,...... ,..~.....~ ...... A.,.~~.~~.~,~,~,.~., Macrocma papillosa ??????????????????????????????.T...... A,,,,.,~.,...~...... Groutiella chimborazense ??????????????????????????????~T..1...I....~...~.~..~...... A...~~.~~...~,....., Groutiella apiculata ??????????????????????????????.T...... A.....,,...... ~ Macromitrium involutifolium ??????????????????????????????*Te*****o***s+~~~e+e*v~e*.ev-*A~*m~e~s~*+~~v~v.*~~ Macromitrium longifolium ??????????????????????????????.T,,,,.,..,...,.,...... ~....A~,.,~,,,...~~.~~~~~ Macromitrium richardii ...... A~.~~..A.~.....~~~~. Desmotheca. apiculata ??????????????????????????????~T.1.1....,...... ~~...~.,..A,,,..,~~~~~.~,~,~,, Zygodon intedius ??????????????????????????????.T...... ~...,...~...... A...,..A...~~~~~~~~~ Zygodon reinwardii ??????????????????????????????.T.,,...~~.~~~~~~~.~~~~~.~~~..A~,~~~.~,..~.....~.~ Zygodon pungens ...... A.,...~.~.~,.~.,...~ Zygodon obtusifolium ??????????????????????????????.T.,...... ,.....,.,...... ~,~..A~,,.,,,,...... ~~~ Orthotrichum lyellii ...... A,.,*...... *. Orthotrichum obtusifolium ??????????????????????????????.T....,...,...... A.,~~~~~.~,~~.,.~.,. Orthotrichurn anornalm ??????????????????????????????.T...... ,...... ~...... A,..,.,.~...... ~.~.~ Ulota obtusiwcula ??????????????????????????????.Tg....~.~..~~~~..~~*v~.e~~~.~A~~o~..q~.~.~~~v~.~v Ulota lutea ??????????????????????????????IT~,,...,..,.~...... ,.~....A,,,.,~.,.~~....~~~~ Bryodixonia perichaetialis ??????????????????????????????.T~.,,,.~.,.~,.,..,..,....,...A.,,.,., *....*..**.. Dnmmondia prorepens ??????????????????????????????~T,.....,..*..*.....*...... ,,A,* .,....,...... *. Anphidium lapponicum ??????????????????????????????.T.~...... A...~...... RhaWoweisia crenulata ??????????????????????????????.T.~...... ,~...... A...... ,...... Aulacopilum hogdkinsoniae ??????????????????????????????.T...... A...... ,~~.~..... Venturiella sinensis ...... A,.,,*., ...... Uleastrum palmicola ??????????????????????????????.T...... ,.....,...... ~~..~A.,~,.,....~~~.~~~~. Encalypta procera ??????????????????????????????.T...... A..,.....,...... etychomitrium gardnerii ??????????????????????????????.T...... ,.,,.,,,.,,,A,....,A...,.,P..,.. Appendix one: 169

...... *-.-..-.m...... O::::: ...... Appendix one: 170

..C C**..- .CI -..:? : ..-...a.. m....-

: : : : : :?Y! ...... d:::::::: u...... :*. :+q++w ...... cl: : : : : : ...... --*..-.-...... Appendix 1 (cont.). RbcL gene sequences of 37 arthrodontous mosses.

Funaria hygrometrica Funaria apophysata Splachnum sphaericm Tayloria lingulata Hedwigia ciliata mium thomsonii Abietinella abietiena Schlotheimia tecta Schlotheimia trichdtria Schlotheimia brownii Cardotiella quinquefaria Macrocoma tenue Macrocorna papillosa Groutiella chimborazense Groutiella apiculata Macromitrium involutifolium Macromitrium longifolium ~acromitriurnrichardii ûesmotheca, apiculata Zygodon intermedius Zygodon reinwardii Zygodon pungens Zygodon obtusifolim Orthotrichum lyellii Orthotrichum obtusifolium Orthotrichum anomalun Ulota obtusiuscula Ulota lutea Bryodixonia perichaetialis ~trmmondiaprorepens Amphidiun lapponicum RhaWoweisia crenulata Aulacopilm hogdkinsoniae Venturiella sinensis Uleastrum palmicola Encalypta procera ~tychomitriumgardnerii Appendix one: 172

...... :?:::::: ...... v::::::: Appendix one: 173

...... : :" : : : : ...... : :v=!:?Y???=! : : : : : : ...... ::??:::: g* *...... O...... ~.-~.'~~~.*.~~~~.~'.~~..~~.~,~ "1:::::::::::::::::::::::::::::::...... Bg:::::::::::::::::::::::::::::::::: ...... SE!:::;::::::::::::::::: AppendVt one: 174

...... C--.*....-.C.II...... *.-.-..*.L-.-.*..,..*...... -.LI... Appendix one: 175

...... -.*..-CI...... +<<<.:;<;€; ...... *..*. <.':y: ...... +~U{+{++ ...... CI- :Y:::: ...... ::?Y::::: ...... CI C .m...... <++++{++...... r'++<$

Âk...... Appendix one: 177

...... -..*-..o......

...... S.. C......

...... Appendu one: 178

.*.- Y::: ..o.

y : : m......

:::=!:::::? <*:++++dd{< a::::::::::...... % :+e;~+++-.. 6;;i+i;< &:::::G:::: Appendix one: 179

Appendix one: 18 1 Appendix 1 (cont.). RbcL gene sequences of 37 arthrodontous mosses.

-ria hygrometrica Funaria apophysata ..T...... A...~.~..~*~*.~~.~~~.~*.~**********~~ Splachnwn sphaericum ..T....T...... T.,...... ~,T...... ,T.....A..A....~,....~,.~T~~.~~,,~T.~A~,..,. ~aylorialingulata ..Tt...T...... T.~C~~.~~..~T~~~~~~~~T~~~~.A~~A~~..~.~.~.~...T.~~..~~..~~~~~~~~~ Heàwigia ciliata ..T....T...~~...... ~....T*.*~....T.....A..A~.~..~~....~..T.**....*T...... *. mium thomsonii ..T....T...... T.IC.1...C1.TttC....~~...... Abietinella abietiena ..T. ...T.., .....T....,...... T...... T..?.1A..A.....C~~~~~.~.T~.~.....T.....~..~ Schlotheimia tecta ..T....T.~..,.I.TI~~~,..I...T,t.....,Tt.?~~A~~A...... T.~.e....T.,l..~.~e Schlotheimia trichdtria ..T..,~T...ll...... C....~~.1T..1...1,T.~~~.A~~A~..~....~~....T.~.I~~~~T~~~~e~o~~ Schlotheimia brownii ..T. ...T...... T.,*..t.,?**T*...... T.....A.~A..~.l.l~.e~~t.T~e~~~~~,T....~~~.. Cardotiella quinquefaria ..T...lT....~.~.T.I....v....T...... T.....A..A...~~~~.l.....T...... ~T..~...... Macrocana tenue ..TI...T...... T.~...... T..~...~.T~,.~oA..A..~.~..,.~eA~~T~~.~e~~~T~ee....~. Macrocma papillosa ..T...l..l..e*****Ittte*. **mT...1.t*IT.?..A..A...vr*tr...*.*T***.e...T**s*e**~v Groutiella chiniborazense ..Tt...T...... T...~...... T...~...~T.,?.~A..A~...... T...... ~.T..~~.~~~~ ~routfellaapiculata ..T.~..T...... T..,,.....,...~...... T~~~..A..A~..,....~...~.T.~~.~~~~T~,,~~~~~~ Macrdtrium involutifolium ..T....T.***e+*.TI**IeC ..a* ~T.....~..T~~?~~A..A~*~~~*~~~~~*~~T~o~~*o~*T~~~.~v~~~ Macromitrium longifolium ..T....C...... ~T...... ,C.~T....,..~T~.,.~A.~A.~~~..~~~~~.~~T~~~~...,T.~.....C, Macromitrium richardii .*T*...T***.....TI....C.....T*....***T**.*.A*.A...****..*...*T*.******T**.***.** ksmotheca. apiculata ..T....T...... T...... T...~~.1.T.~~..At.A~t.~.~~...1...T...... T.....~~.~ Zygodon interniedius ..T....T....,...T.FC~~t~....T.....l,,.,... Zygodon reinwardii ..T....T~..~....T~~C~~,..C..I.....~~~T~.~~~A~~A~...~.~~~~~T.~T~~~~~~~~T~~~~~~~~. Zygodon pungens ..T...... eT~IC~~~~e~~q~~~~1~~t.T~1.~1A~1Av~1.~I~~~~IT~~Tt~~~~~t~T~.v~~C~vv Zygodon obtusifolium ..T....T.....,..T~tC,~,,~~.tT~t.1..1,Tt,?~,A.,At,~,.I...~.~..T...~~...T...... orthotrichwn lyellii ..T....T...... Tt~Co+.I..1.T'PII~IeIIT*v**eA.mA~...~~*+~1~T*.sI..*v*ttT*e**e*C** Orthotrichum obtusifolium ..T....T..,....~T~.CI..~....T...1..~1T.1?..A,~A~,...,.~~~eT.IT.~.~....T~..~..... Orthotrichum anmlum ..T..,.T~~~.....T~,Ct...1.~,T.1.~1~~.T.~.~~A~~A~~e..~.....T~.T~.....~.T.~~~~~~~~ Ulota obtusiuscula ..TI*..T ...... T.~CC~~.1~..T~...... T.~?.*A1.AI.t*o.~~.1~T*t~1~.~.~~1Tv....Ct. Ulota lutea ..T....T....,.~.T..C...~~~~.T~.~..~.~T..?~.A~~A.~.~..,~~~.T.~~~.....~.T.~~...C.. Bryodixonia perichaetialis ..T....T.,...,,.T..C...,....T...... T.....A..A...... ,.T....,...... T,,,.,,C,, Drumrondia prorepens ..T....T...... C.C.....,T,...... T.,...A,,A~,i...... T...r.r..T.,.,tt.,. Pnlphidium lapponicum ...... T.,...... T..C...... T.,....~,.,~..~,. Rhabdoweisia crenulata ..T....T..I....IT..C..C...... T...... i.T.~...AttAC...... ~...... T...... ~ulacopilm hogdkinsoniae ..T. ...T...,,.,*T, ...... T,..~~~~,T~,.~~A~~A~~~~~~~.~..T..T~~...... ~~~~.~.~~~ Venturiella sinensis . T..~.T.~~,~~~~T.~~~.....~~T.~..~...T.....A~~A~~..~~~~.~~T~.T..~...~...... Uleastrum palmicola ..T....T.,.t....T..C,,.,...,T,,~.....T,....A,.A~...... ,.T..T.,.~.,.,....,~~~~. Encalypta procera ..T...... C..~.....T.....I...~~...... ~tychomitriumgardnerii ..T....T...... ,T,,...... T...,,...T..?..A..A...... ,,T.,T...... T...... Appendix one: 183

...... - o...... -..... 4$1:::::::::::::: .O...... -..m.. .m.-.*m.. o.,.. o..... -..-.-C.. *. . :5 :

...... *A ...... -.."...."-' a ::::?<:::::::::::::: $: ...... Ba:::::::::::::::'?:::: ...... o.. ...-..S.. *. . *. . *. . -...... m.

o...... Appendùr one: 184