Circumscription and phylogeny of the Orthotrichales (Bryopsida) inferred from RBCL sequence analyses

Bernard Goffinet; Randall J. Bayer; Dale H. Vitt

Mosses of the subfamily Orthotrichoideae represent one of the main components of the cryptogam epiphytic communities in temperate areas. During the last two decades, this taxonomical group has undergone an extensive revision that has led to its rearrangement at the generic level. However, their phylogenetic relationships and inferences on the evolutionary patterns that have driven the present diversity have little advanced. In this study, we present a dated molecular phylogenetic reconstruction at the subfamily level, including 130 samples that represent the 12 genera currently recognized within the subfamily, and the analysis of four molecular markers: ITS2, rps4, trnG, and trnL-F. We also analyze 13 morphological characters of systematic value to infer their origin and diagnostic utility within the subfamily. The phylogenetic reconstruction yields three main clades within the subfamily, two of which correspond to the tribe Zygodonteae, and one to Orthotricheae. Within Zygodonteae,...

Phylogenomic data increase the possibilities of resolving the evolutionary and systematic relationships among taxa. This is especially valuable in groups with few and homoplasious morphological characters, in which systematic and taxonomical delimitations have been traditionally difficult. Such is the case of several lineages within Bryophyta, like Orthotrichaceae, the second most diverse family of mosses. Members of tribe Orthotricheae are common in temperate and cold regions, as well as in high tropical mountains. In extratropical areas, they represent one of the main components of epiphytic communities, both in dry and oceanic or hyperoceanic conditions. The epiphytic environment is considered a hostile one for plant development, mainly due to its low capacity of moisture retention. Thus, the diversification of the Orthotrichaceae in this environment could be seen as striking. Over the last two decades, great taxonomic and systematic progresses have led to a rearrangement at the ...

American Journal of Botany 85(9): 1324–1337. 1998. CIRCUMSCRIPTION AND PHYLOGENY OF THE ORTHOTRICHALES (BRYOPSIDA) INFERRED FROM RBCL SEQUENCE ANALYSES1 BERNARD GOFFINET,2 RANDALL J. BAYER,3 AND DALE H. VITT Department of Biological Sciences, The University of Alberta, Edmonton, Alberta, Canada T6G 2E9 The affinities as well as the circumscription of the Orthotrichaceae (Bryopsida), one of the most diverse families of mosses, have been the focus of a controversy for much of the last century. We obtained rbcL sequences for 37 arthrodontous mosses, including 27 taxa of the Orthotrichales. The sequences were analyzed using maximum parsimony and neighbor joining in order to (1) test the monophyly of the Orthotrichales and the Orthotrichaceae; (2) determine their phylogenetic relationships; and (3) test the current subfamilial classification within the Orthotrichaceae. Both analyses suggest that the Orthotrichales are polyphyletic. The Erpodiaceae and the Rhachitheciaceae as well as Amphidium and Drummondia, two genera of the Orthotrichaceae, are shown to be of haplolepideous affinity. The Splachnales, the Bryales sensu lato, and the Orthotrichales form a monophyletic clade sister to the Haplolepideae. Both neighbor joining and maximum parsimony also suggest that the Orthotrichaceae are composed of two major lineages dominated either by acrocarpous or cladocarpous taxa. The monophyly of the family is, however, only well supported by Tamura’s distances. The genera Macrocoma, Macromitrium, Orthotrichum, Ulota, and Zygodon all appear to be artificial assemblages. This study illustrates the contribution of rbcL sequence data to bryophyte systematics and, particularly, in determining the affinities of taxa lacking a peristome, whose characters are central to the classification of mosses. Key words: bryophytes; evolution; mosses; Orthotrichales; Orthotrichaceae; phylogeny; rbcL; systematics. The central concept in the Hennigian phylogenetic approach is the use of appropriate outgroups, which allows adequate polarization of character state transformation in the ingroup. A reconstruction of the evolutionary history of the Orthotrichaceae, one of the most diverse families of mosses, based on morphology has been hampered by the lack of a peristome in certain taxa, the absence of gametophytic characters that are informative at higher taxonomic levels, and consequently the absence of a consensus regarding the putative sister group and other outgroups to the order. Within the Bryophyta sensu lato, phylogenetic reconstructions using nucleotide sequence data have focused exclusively on the monophyly and the relationships of the division and its classes (e.g., Mishler et al., 1992, 1994; Waters et al., 1992; Capesius, 1995; Bopp and Capesius, 1996; Hedderson, Chapman, and Rootes, 1996) rather than on the evolutionary history of arthrodontous mosses. 1 Manuscript received 3 August 1997; revision accepted 26 January 1998. This study is part of a Ph.D. dissertation presented by the senior author to the Faculty of Graduate Studies of the University of Alberta. This study was supported by Grants A-3797 and A-6390 from the Nature Sciences and Engineering Research Council of Canada to RJB and DHV, respectively. Field work was made possible through a grant by the National Geographic Society to DHV and BG. We are grateful to William Buck (NY), Celina Matteri (BA), and Heinar Streimann (CANB) for kindly sending us fresh material of Cardotiella quinquefaria, Macrocoma papillosa, and Desmotheca apiculata, respectively, as well as Ross Hastings (PMAE) for the loan of Ptychomitrium gardneri for DNA extraction. Finally we thank G. Zurawski for the set of sequencing primers. 2 Author for correspondence, current address: Department of Botany, Box 90339, Duke University, Durham, NC 27708. 3 Present address: CSIRO, Division of Plant Industry, Australian National Herbarium, Centre for Plant Biodiversity Research, Molecular Systematics Lab, GPO Box 1600, Canberra, ACT, 2601, Australia. In arthrodontous mosses (Bryopsida; sensu Vitt, Goffinet, and Hedderson, 1997), the peristome teeth that line the mouth of the capsule and regulate spore dispersal are made of cell wall remnants. 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 the pattern of cell divisions occurring in these layers, particularly in the IPL, are central to the classification of arthrodontous mosses. Vitt (1984) recognized five types of articulate peristomes (i.e., four diplolepideous and one haplolepideous type), which characterize the subclasses of the Bryopsida (see also Edwards, 1979, 1984). Vitt, Goffinet, and Hedderson (1997) recently argued that morphological characters used to define these major peristome types are not phylogenetically informative and that higher level phylogenetic reconstruction of the Bryopsida will rely on independent sources of data such as ontogeny and gene sequences. The development of the funariaceous, bryaceous, and dicranaceous peristome types has been studied recently (Shaw, Anderson, and Mishler, 1989; Shaw, Mishler, and Anderson, 1989; Schwartz, 1994), but unless similar data are available for the orthotrichaceous peristome, polarizing the developmental pathways and thus defining homologies and monophyletic groups based on peristome architecture remain impossible. The peristome of the Orthotrichales is diplolepideous and is characterized by the following features: ‘‘an endostome with segments that alternate with exostome teeth; lack of basal membrane and with segments, which are not or are rarely keeled; and an exostome that has a thick, outer layer and a thin, inner layer’’ (Vitt, 1982a). Crosby (1980) and Vitt (1981a) were the first to suggest that the ancestral arthrodontous peristome was diplolepideous. Crosby (1980) and later Shaw and Rohrer 1324 August 1998] GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY (1984) considered that the Bryum type represented the most primitive among the diplolepideous types and consequently that the orthotrichaceous peristome was derived through reduction, a hypothesis also supported by Hedenäs (1994). Vitt (1984) argued that ‘‘the evolution of the peristome has not been uni-directional’’ and that the bryaceous peristome ‘‘is a totally separate evolutionary line than either the orthotrichaceous or haplolepideous divergences’’ from an ancestral funariaceous type (see also Vitt, 1981a). More recently, Lewinsky (1989) argued in favor of a distinct orthotrichaceous peristome type that is ‘‘most likely to have evolved from a peristome with a formula of 4:2:4 with complete alignment of the cells in the IPL,’’ thus the funariaceous type, but unlike Vitt (1984) she considers the orthotrichaceous peristome as possibly pivotal to the evolution of the dicranaceous and the bryaceous types. The Orthotrichaceae sensu Vitt (1984; and not sensu Churchill and Linares, 1995) include over 500 species, distributed among 22 genera (see Goffinet, 1997a, b), with Orthotrichum, and Macromitrium accounting for more than two-thirds of the species richness (Vitt, 1982a). The family is cosmopolitan in distribution and is particularly prevalent in tropical montane forests. The Orthotrichaceae are traditionally characterized by a unique peristomial architecture (see above), as well as ‘‘small, papillose upper leaf cells; no differentiated alar cells; large, usually mitriform calyptrae; terminal perichaetia with additional growth by lateral innovations’’ (Vitt, 1982b). These characters are, however, by no means constant within the family, resulting in alternative hypotheses of familial circumscription. The controversy has focused particularly on the haplolepideous vs. the diplolepideous affinity of Amphidium, which lacks a peristome, and has atypical cell ornamentation (Brotherus, 1909 vs. 1925; Vitt, 1973 vs. Lewinsky, 1976), or of taxa whose peristome is reduced (Drummondia: Shaw, 1985 vs. Vitt, 1972). The current infrafamilial classification (Vitt, 1982a) reflects the four possible combinations of two characters, position of the female gametangia and ‘‘shape’’ of the calyptrae, and deviates little from Brotherus’s concept early this century (Brotherus, 1925). In Vitt’s (1982a) phylogenetic arrangement, the Zygodontoideae (Schimp.) Broth. (acrocarpous and cucullate calyptrae) are basal, followed by the Drummondioideae Vitt (cladocarpous and cucullate calyptrae), which are sister to a clade composed of the Orthotrichoideae Broth. and the Macromitrioideae Broth. (acrocarpous and cladocarpous, respectively, and both with mitrate calyptrae). Addressing the phylogenetic relationships of the Orthotrichales and its delineation is essential for reconstructing the evolutionary history of orthotrichaceous genera. Goffinet (1997b) recently excluded the Microtheciellaceae from the Orthotrichales based on their pleurocarpy. Goffinet (1997a) also transfered two genera from the Orthotrichaceae to the Rhachitheciaceae. The peristomial architecture of the Rhachitheciaceae (Goffinet, 1997a) and of the Erpodiaceae Broth. (Crum, 1972; Edwards, 1979) suggests possible affinities to the Haplolepideae, and particularly the Seligeriales. The monophyly of the Orthotrichaceae and the Orthotrichales has recently also been questioned by De Luna (1995) while examining the systematic affinities of the Hedwigiaceae. Based on a 1325 phylogenetic analysis using morphological characters, De Luna (1995) concluded that the Orthotrichales merely represent an evolutionary grade, suggesting that a clade of predominantly cladocarpous Orthotrichaceae (Macromitrioideae, Drummondioideae) and the putatively related Erpodiaceae Broth., Microtheciellaceae Harrington and Miller, and Helicophyllaceae Broth.(see below) was sister to the pleurocarpous mosses, and separated from the acrocarpous Orthotrichaceae and their possible sister group, the Rhachitheciaceae Robins. by the Hedwigiaceae. Affinities between the Hedwigiaceae and the Orthotrichaceae 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 (acrocarpy, lack of costa, and presence of pseudoparaphyllia), the affinities of Hedwigia are unlikely to be resolved with traditional morphological characters. Severe reduction trends in morphological characters, particularly with regard to the peristome, are thus the source of the controversies regarding the delineation of the Orthotrichales and its type family, the Orthotrichaceae. Ambiguous homologies between characters defining alternative peristome types hamper resolving the phylogenetic affinities of these taxa. Variation in the nucleotide sequence of the chloroplast gene rbcL, encoding for the ribulose 1,5 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 reconstructing the phylogeny of all orthotrichaceous genera based on morphology and for addressing the character evolution in this taxonomically and morphologically diverse family. Thus, the objectives were to use rbcL sequence data (1) to test the monophyly of the Orthotrichales and the Orthotrichaceae; (2) to determine 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 subfamilies proposed by Vitt (1982a). MATERIALS AND METHODS Taxon sampling—Following the exclusion of five genera from the Orthotrichaceae (Goffinet, 1997a, b) the family now consists of 22 genera. While material has been seen for all taxa except for the recently described Orthomitrium (Lewinsky-Haapasaari and Crosby, 1996), extractions were not attempted for taxa known only from the type specimen (Ceuthotheca Lewinsky-Haapasaari, Leiomitrium Mitt., Leratia Broth.), or for which only very little material was available (Stoneobryum Norris and Robinson). DNA extractions were attempted for the remaining 17 genera, and for the most diverse genera, such as Macromitrium and Orthotrichum, several species belonging to morphologically distinct infrageneric taxa were tentatively included. DNA extractions were also attempted for representatives of all related families in the Orthotrichales. Furthermore, ten outgroup taxa were chosen, representing the Funariaceae, Splachnaceae, Encalyptaceae, Hedwigiaceae, the Haplolepideae (Ptychomitriaceae and Rhabdoweisiaceae), and the ciliate alternate peristomate mosses (Mniaceae and Thuidiaceae). DNA extraction and PCR-DNA amplification—DNA was extracted, following a modification of Doyle and Doyle (1987), from apices of stems and branches, or single sporophytes sampled from dried herbar- 1326 TABLE 1. a AMERICAN JOURNAL OF BOTANY [Vol. 85 Synthetic primers (59–39) used for sequencing the rbcL gene in mosses. Forward primers 427a 997 GCTTATTCAAAAACTTTCCAAGGCCCGCC GGTAAACTTGAAGGAGAACG Reverse primers 295R 515R 678Ra 895Ra 1081R CTAATGGGTAAGCAACATAAGC CATCCTAATAATGGACGACC GATTTCGCCTGTTTCGGCTTGTGCTTTATAAA ACCATGATTCTTCTGCCTATCAATAACTGC CCCAGTCTTGAGTGAAGTAAATACC Marks primers designed and provided by G. Zurawski; others were designed by authors. ium or fresh collections. Gametophytic samples were cleaned of contaminating material in distilled water, then air-dried overnight and lyophilized prior to the extraction. One to 10 mg dry mass of each sample were placed in a spot plate, covered with liquid N2, and ground to a fine powder once the liquid N2 was completely evaporated. The powder was then suspended in 250 mL of 2X CTAB (hexadecyltrimethylammonium bromide)-1% beta-mercapto-ethanol extraction buffer heated at 608C and the solution transferred to a microcentrifuge tube. The tubes were then placed in a 608C water bath. After 45 min, 250 mL of CI (chloroform-isoamyl 24:1) were added and the tubes gently inverted for 10 min. Cell debris were then isolated by centrifuging the tubes for 10 min at maximum speed (e.g., 14 000 rpm). The aqueous phase was transferred to a fresh tube (the CI extraction can be repeated if the aqueous solution is not clear). Then 350 mL of 95% ethanol were added and the DNA allowed to precipitate at 48C. After at least 2–3 h, the tubes were centrifuged for 5 min at maximum speed again to pellet the DNA. The ethanol was discarded and the pellet washed gently by adding 350 mL of 70% ethanol. The tubes were centrifuged again for 5 min at 7000 g. The ethanol was discarded and the pellet dried in a vacuum centrifuge. The DNA was then suspended in 100 mL of TBE (Tris-boric acid-EDTA) at 658C for 10 min and kept at 2208C subsequently. The rbcL gene was amplified via the polymerase chain reaction (PCR) using Taq Polymerase (Promega Corporation, Madison, WI). The amplification reaction was performed in 50 mL volume including 50 mmol/L KCl, 10 mmol/L Tris-HCl, 0.1% Triton, 5% glycerol, 2.5 mmol/L MgCl2, 20 mmol/L of each dNTP, 0.2 mmol/L of primers Z1 and 1351R (Wolf, Soltis, and Soltis, 1994) and 0.2–30.0 ng of template DNA. The samples were exposed to the following temperature profiles using a Grant thermal cycler: one cycle of 948C for 3 min and 858C for 4 min, during which 1 unit of Taq DNA Polymerase would be added to each tube, and 30 cycles of 948C for 1 min, 528C for 1 min, 728C for 2 min, and finally one segment of 728C for 10 min. The doublestranded product subsequently served as the template for the amplification of single-stranded target DNA. This second PCR followed the same profiles as before, but the reaction solution included only one of the two PCR primers: Z1 for amplification of the forward strand and 1351R for the reverse strand. The single-stranded product was precipitated with 20% PEG/2.5 mol/L NaCl, washed first with 70% EtOH and then with 95% EtOH, before being suspended in 7 mL of Tris bufferEDTA (Morgan and Soltis, 1993). Sequencing and alignment—Sequencing the single-stranded template followed the dideoxy chain termination method (Sanger, Nicklen, and Coulson, 1977) using the Sequenaset version 2.0 T7 DNA polymerase following the manufacturer’s instructions (Amersham, Canada). The sequence of the primers used is given in Table 1. The sequencing products were electrophoresed on a 6% polyacrylamide gel (0.4 mm thickness; 1X TBE buffer), at 2400 V for 4 h. The gels were fixed in 10% acetic acid, washed with distilled water, and dried in an oven at 658C for 30 min and autoradiographed for 36–48 h. The sequences were entered in MacClade (version 3.03) and aligned manually against available sequences of Sphagnum palustre and Andreaea rupestris (GenBank accession numbers L13485 and L13473, respectively; Manhart, 1994). The first and last 30 nucleotide sites, corresponding to the sequences of the PCR primers, were excluded from parsimony analyses. Sequence and phylogenetic analysis—RbcL sequence variation was analyzed by neighbor-joining (NJ; Saitou and Nei, 1987), and maximum parsimony (MP; Fitch, 1971). Neighbor-joining analyses were performed using MEGA version 1.0 (Kumar, Tamura, and Nei, 1993) using Tamura’s distance parameter, following the authors’ ‘‘Guidelines for choosing distance measures,’’ and including both transitions and transversions. A ‘‘bootstrap confidence level’’ for the NJ tree was calculated over 100 replicates. Fitch parsimony analyses of nucleotide data were performed with PAUP version 3.1 (Swofford, 1993) on a MacIntosh PowerPC 7200/90 using the heuristic search with the following options in effect: keep all characters, multistate taxa interpreted as uncertainties, tree-bisection-reconnection branch swapping, steepest descent, collapse zero-length branches, and addition sequence ‘‘as is.’’ In an attempt to locate additional islands of shortest trees (Maddison, 1991), the search strategy further followed the steps recommended in Pryer, Smith, and Skog (1995) except that the search was only replicated 100 times. Bootstrap analysis (Felsenstein, 1985; Hillis and Bull, 1993) was performed with 100 bootstrap replicates of the heuristic search with the same set of options in effect. Relative support for branches was further determined by the decay analysis (Bremer, 1988; Donoghue et al., 1992) following the converse constraint method (Baum, Sytsma, and Hoch, 1994). The ACCTRAN (accelerated transformation optimization) option of PAUP was applied for calculations of branch lengths. The phylogenetic signal present in the rbcL sequence data was estimated by calculating the g1 statistic of the distribution of tree length of 500 000 random trees produced with PAUP using the ‘‘random tree’’ option (Hillis and Huelsenbeck, 1992). Consistency (CI) and retention indexes (RI) and f values were calculated for all MPTs using PAUP. PAUP was also used to generate a matrix of absolute and mean distance between sequences. Average unit character consistencies (AUCC) were calculated for competing phylogenies in an attempt to select the tree with the strongest asymmetric distribution of homoplastic characters in the matrix (Sang, 1995). Costs in terms of number of additional steps required for alternative phylogenies were explored using the enforce constraint command during heuristic searches (Swofford, 1993). The same set of options as in the unconstrained searches was applied. RESULTS The rbcL gene was successfully sequenced for 22 taxa of the Orthotrichaceae distributed among ten genera, for representatives of two of the related families as well as for all selected outgroup taxa (Table 2). Amplification products were not obtained for Florschuetziella Vitt, Leptodontiopsis Broth., Muelleriella Dusén, Pleurorthotrichum Broth., and Stenomitrium (Mitt.) Broth., Helicophyllum Brid. The sequences obtained from PCR fragments generated using the PCR primers Z1 and 1351R August 1998] TABLE 2. GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY 1327 Taxa for which the rbcL gene sequence was obtained in this study (all vouchers deposited in ALTA unless otherwise noted). Taxon ORTHOTRICHALES ORTHOTRICHACEAE Zygodontoideae Zygodon pungensa C. Müll. Zygodon obtusifolius Hook. Zygodon intermedius B.S.G. Zygodon reinwardtii (Hornsch.) Braun Amphidium lapponicum (Hedw.) Schimp. Voucher GenBank accession number La Farge-England 8097 Vitt 38301 Vitt 29262 Goffinet 636 Vitt 33854 AF005534 AF005535 AF005532 AF005533 AF005543 Orthotrichoideae Orthotrichum obtusifolium Brid. Orthotrichum anomalum Hedw. Orthotrichum lyellii Hook. & Tayl. Ulota lutea (Hook. f. & Wils.) Mitt. Ulota obtusiuscula C. Müll. & Kindb. Bryodixonia perichaetialis Sainsb. Vitt 33870 Goffinet 4115 Goffinet 3162 Fife 8042 Goffinet 3161 Fife 8083 AF005537 AF005538 AF005536 AF005540 AF005539 AF005541 Drummondioideae Drummondia prorepens (Hedw.) Britt. Vitt 26711 AF005542 Vitt 27485 Schäfer-Verwimp 9686 Schäfer-Verwimp 6902 Buck 26230 Goffinet 2764 Goffinet 1173 Matteri 6521 AF005522 AF005520 AF005521 AF005523 AF005527 AF005526 AF005525 Breedlove 69342 Streimann 49345 Goffinet 656 Goffinet 2648 Vinas 96-4 AF005524 AF005528 AF005529 AF005530 AF005531 ERPODIACEAE Aulacopilum hodgkinsoniae (C. Müll.) Broth. Venturiella sinensis (Vent. in Rabenh.) Müll. Hal. Vitt 28261 Vitt 34842 AF005545 AF005546 RHACHITHECIACEAE Uleastrum palmicola (C. Müll.) Zander Vitt 21162 AF005547 Vitt 27234 Priddle 1408 AF005514 AF005513 ENCALYPTACEAE Encalypta procera Bruch Vitt 37966 AF005548 HEDWIGIACEAE Hedwigia ciliata (Hedw.) P. Beauv. Goffinet 3324 AF005517 MNIACEAE Mnium thomsonii Schimp. Vitt 35884 AF005518 PTYCHOMITRIACEAE Ptychomitrium gardneri Lesq. Ireland 7038 (PMAE) AF005549 RHABDOWEISIACEAE Rhabdoweisia crenulata (Mitt.) Jameson Vitt 36707 AF005544 SPLACHNACEAE Tayloria lingulata (Dicks.) Lindb. Splachnum sphaericum Hedw. Schofield 98443 Goward 95-1470 AF005515 AF005516 THUIDIACEAE Abietinella abietina (Hedw.) Fleisch. Goffinet 4106 AF005519 Macromitrioideae Schlotheimia brownii Schwaegr. Schlotheimia tecta Hook. f. & Wils. Schlotheimia trichomitria Schwaegr. Cardotiella quinquefaria (Hornsch.) Vitt Groutiella apiculata (Hook.) Crum & Steere Groutiella chimborazense (Mitt.) Florsch. Macrocoma papillosa (Thér.) Vitt Macrocoma tenuis (Hook. & Grev.) Vitt subsp. sullivantii (C. Müll.) Vitt Macromitrium incurvifolium (Hook. & Grev.) Schwägr. Macromitrium longifolium (Hook.) Brid. Macromitrium richardii Schwaegr. Desmotheca apiculata (Dozy & Molk.) Lindb. OUTGROUP TAXA FUNARIACEAE Funaria apophysata (Tayl.) Broth. Funaria hygrometrica Hedw. a This specimen is tentatively identified as Zygodon pungens C. Müll., a species hitherto not known from Africa (Malta 1926), but may represent a new species. 1328 AMERICAN JOURNAL OF BOTANY [Vol. 85 TABLE 3. Distribution and frequency (%) of constant and phylogenetically informative sites over all taxa and over the Orthotrichaceae only; characters with ambiguous or missing data are included. All taxa Codon position Constant Phylogenetically informative First Second Third Total Orthotrichaceae (total) 368 (83.6) 30 (6.1) 407 (92.5) 18 (4.1) 148 (33.6) 109 (24.7) 923 (69.9) 157 (11.9) 1158 (87.8) 108 (8.2) were, as expected, 1320 bases long (690% of the total length of the gene). Alignment of the sequences with known sequences of Sphagnum pallustre and Andreaea rupestris did not require the inclusion of gaps. Phylogenetically informative characters are preponderant at the third codon position followed by the first and the second codon site (Table 3). RbcL sequence variation yields 108 sites that are potentially phylogenetically informative within the Orthotrichaceae. Informative characters are distributed fairly evenly across the sequence (Fig. 1). The distribution of 500 000 random trees using all characters had a skewness index g1 of 20.55; excluding the first and second codon position yielded a similar index (g1 5 20.54), while trees generated solely from characters of these two first codon sites had a higher skewness index (g1 5 20.39). Fig. 1. Distribution of phylogenetically informative (PI) characters, the average consistency index (CI), and the average number of steps per phylogenetically informative character along the rbcL sequence based on tree shown in Fig. 3. Each class contains 50 nucleotide sites, except for the first class that includes only sites 31–50. Cross-lined bars 5 total number of phylogenetic informative characters; solid bars 5 average CI multiplied by 10; hollow bar 5 total number of steps per total number of phylogenetically informative characters. Using the maximum parsimony criterion with Sphagnum and Andreaea as the outgroup yields 39 most parsimonious trees (MPT) distributed among three islands of size 21, 6, and 12 (Fig. 2). 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 between 13 306 and 19 198. Both methods of analysis agree on the polyphyly of the Orthotrichales and place the Erpodiaceae and the Rhachitheciaceae as well as the orthotrichaceous genera Amphidium and Drummondia in a monophyletic clade with Ptychomitrium and Rhabdoweisia (Figs. 3–4). Constraining the search for the inclusion of the Erpodiaceae and Rhachitheciaceae in the Orthotrichales results in trees that are 17 and 15 steps longer than the shortest unconstrained trees, respectively. Including Amphidium and Drummondia in the Orthotrichaceae also increases the length of the most parsimonious topology (125 steps), and furthermore these two genera remain more closely related to each other than they do to any other genus of the Orthotrichaceae. The NJ tree agrees with the strict consensus tree over all MPTs in the following relationships (Figs. 3–4): (1) Funaria occupies a basal position among arthrodontous mosses; (2) the Orthotrichales and the Splachnales form a monophyletic clade sister to the ciliate mosses; and (3) the Haplolepideae, including Amphidium, Drummondia, Venturiella, and Uleastrum, form a monophyletic group sister to derived Diplolepideae. Hedwigia is sister to the two ciliate mosses in 36 MPTs (Fig. 2) as well as in the NJ tree (Fig. 4) and basal in the diplolepideous clade (excluding the Funariaceae) in the remaining three trees (Fig. 2). Compared to MP, NJ provides similar or slightly higher bootstrap values for major lineages (Haplolepideae, combined Splachnales and Orthotrichales, Orthotrichaceae, and ciliate mosses), but like MP, NJ fails to yield strong support for their relationships (Figs. 3, 4). The remaining genera of the Orthotrichaceae (thus excluding Amphidium and Drummondia) form a strongly supported monophyletic family in the phylogenetic reconstruction using Tamura’s distance parameter (Fig. 4). Among the 39 MPTs found in the cladistic analysis (Fig. 2), 15 trees (38%), distributed between island 1 and 3 (each with nine and six trees, respectively), support the monophyly of the family. The trees of these two islands differ mainly by the position of Encalypta (Fig. 2). In island 3 (mean f values: 13 943), Encalypta is sister to the Haplolepideae (Fig. 2B), while in islands 1 and 2 it occupies a position basal to the dichotomy between Haplolepideae and the derived Diplolepideae (Fig. 2C; mean f value overall: 16 634). Comparison of the distribution of homoplasy among characters from two selected trees contrasted mainly by the position of Encalypta (island 1 vs. island 3), yields an AUCC value of 0.71 for both, August 1998] GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY suggesting no difference in the distribution of homoplastic characters (Sang, 1995). All other indexes being equal, the tree with the lowest f value (see Farris, 1972) is shown (Fig. 3). The Orthotrichaceae are composed of two major lineages: one including all Macromitrioideae (Cardotiella, Desmotheca, Groutiella, Macrocoma, Macromitrium, and Schlotheimia) as well as Zygodon obtusifolius, and one uniting the remaining species of Zygodon with the Orthotrichoideae (Bryodixonia, Orthotrichum, and Ulota). Large genera such as Macromitrium, Orthotrichum, and Zygodon appear paraphyletic, independent of the method of analysis. In the NJ analysis, as well as in all MPT of island 2, Desmotheca is basal in a clade centered around Macromitrium, while the other MPT suggest a derived position, sister to Macromitrium richardii. DISCUSSION RbcL sequence data—RbcL sequence data are routinely used to reconstruct evolutionary histories among (Chase et al., 1993; Hasebe et al., 1995) or more rarely within (e.g., Haufler and Ranker, 1995) genera of vascular plants. With 30% variable sites, including 12% of sites with changes that may be phylogenetically informative, the variation in the nucleotide sequences of the rbcL gene among 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 vascular plants, i.e., with no obvious hot spot detectable (Olmstead and Sweere, 1994). The distribution of randomly generated trees has a lefthand skewness with a g1 value (20.55) significantly smaller than the critical value (20.10 or 20.08 for 100 or 250 characters, respectively; P 5 0.01) furnished by Hillis and Huelsenbeck (1992), suggesting that our rbcL data set is more structured than a random data set of equal size and that it may contain significant phylogenetic signal (Hillis and Huelsenbeck, 1992). Unlike Conti, Fischbach, and Sytsma’s (1993) observation that the third codon position introduces ‘‘noise,’’ a lower g1 value is observed when the matrix is restricted to the third codon position (20.54 vs. 20.39), suggesting that the changes at the third codon position are more structured. 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 first two codon sites carry more homoplasies than the third position, which may thus be more informative. Alternatively, the taxon sample may be too disparate for variability at the more conserved positions 1 and 2 to be mostly phylogenetically informative. The significance of the difference between the g1 value based on our data set and that of a random matrix may, as a result, be due to differences in the frequency of character states rather than in the congruence among characters in both data sets (Källersjö et al., 1992). This second hypothesis most likely applies to our data since our restricted taxon sample represents ma- 1329 jor lineages of arthrodontous mosses with a long evolutionary history (Frey, 1977, 1990). An indication of the accuracy of our sequences as well as for the presence of pseudogenes may be obtained from the comparison of amino acid sequences with the distribution of active sites (Kellog and Juliano, 1997). All amino acid residues of the active site found in spinach by crystallography (Andersson et al., 1989) are conserved among studied bryophyte taxa, except for position 404 (amino acid numbering), which is scored as polymorphic (Arg in addition to the conserved Gly) for six taxa, and thus excluded from phylogenetic analyses in MP. This observation may provide some preliminary support for the accuracy of the sequences obtained. Circumscriptions of the Orthotrichales—The Orthotrichales are currently composed of four families: the Erpodiaceae, Helicophyllaceae, Orthotrichaceae, and Rhachitheciaceae (Goffinet, 1997b). RbcL sequence data analyzed using either the neighbor joining or the maximum parsimony criterion indicate that the Orthotrichales form a polyphyletic taxon. Both the Erpodiaceae (Aulacopilum hodgkinsoniae, Venturiella sinensis) and the Rhachitheciaceae (Uleastrum palmicola) are indeed placed in a clade with haplolepideous taxa (Figs. 3, 4). This clade also includes the ‘‘orthotrichaceous’’ genera Amphidium and Drummondia. The monophyly of this haplolepideous clade is strongly supported (MP: decay index [DI] of 4, bootstrap value [BV] 5 77%; NJ: BV 5 94%). Constraining these taxa to a relationship within a monophyletic Orthotrichales is furthermore very costly in terms of parsimony, requiring up to 25 additional steps. Variation in the nucleotide sequence of the rbcL gene is thus congruent with earlier hypotheses based on morphology and cytology, suggesting haplolepideous affinities of these taxa (for Amphidium see Anderson and Crum, 1958; Vitt, 1970, 1973; Drummondia: Shaw 1985; Erpodiaceae: Edwards, 1979; Rhachitheciaceae: Goffinet, 1997a). More extensive sampling of haplolepideous taxa would be needed before ordinal affinities of Amphidium and Drummondia are elucidated (see also below, affinities of excluded taxa). De Luna (1995) recently argued that the Orthotrichales represented an evolutionary grade, reached by two lineages. The cladocarpous lineage (including, e.g., Erpodiaceae, Macromitrioideae) was sister to the Leucodontales and separated from the acrocarpous line (e.g., Rhachitheciaceae, Orthotrichoideae) by the Hedwigiaceae. Affinities between the Hedwigiaceae and the Orthotrichaceae have also been proposed by other authors (Walther, 1983; Frey et al., 1995). Constraining the heuristic search to include Hedwigia ciliata in the Orthotrichaceae (as defined in Figs. 3 and 4, and with no further relationships specified within this clade) yielded 12 trees that not only are 16 steps longer than the MPTs in the unconstrained search, but also share a monophyletic Orthotrichaceae. The incongruence between De Luna’s morphological study (De Luna, 1995) and the present molecular analysis may be due to inconsistent and inadequate taxon sampling with the absence of representatives of the Leucodontineae from the molecular analysis and of bryalean taxa from the morphological analysis. The latter analysis further includes taxa that are here shown 1330 AMERICAN JOURNAL 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 and the Orthotrichaceae cannot be excluded, but if the Hedwigiaceae were closely related to the Orthotrichaceae, rbcL data suggest that the Orthotrichaceae would most likely remain a natural group. This phylogenetic hypothesis seems to be supported by 18S sequence data analyses (C. Cox and T. Hedderson, personal communication, University of Reading, U.K.). The branch that supports the Hedwigiaceae–cladocarpous Orthotrichales–Leucodontales lineage is supported in De Luna’s analysis (1995) by three synapomorphies, namely plagiotropic growth, presence of pseudoparaphyllia, and differentiated perichaetial leaves. Two of these characters are reversed in the cladocarpous Orthotrichales, leaving a single character that actually supports a relationship of these taxa with the Hedwigiaceae and Leucodontales, namely plagiotropic growth. This character is, however, homoplastic among many families and genera of mosses (Meusel, 1935) and may thus not be truly informative at the ordinal level. Analysis using the distance method yields a single tree (Fig. 4) 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 analysis. Käss 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 results of both analyses were fairly congruent. Kim, Rohlf, and Sokal (1993) critically examined the accuracy of the NJ method under different constraints on a random data set and found that this method ‘‘has the highest accuracy overall.’’ Russo, Takezaki, and Nei (1996) recently compared the efficiency of various tree-building methods in recovering a known phylogeny and concluded that NJ ‘‘gives as good a result as the more time consuming. . .methods.’’ The congruence between the NJ tree and 15 MPTs is therefore here interpreted as supporting the monophyly of the Orthotrichaceae. With the exclusion of the Erpodiaceae, Rhachitheciaceae, and Hedwigiaceae from the Orthotrichales, the order is now reduced to the Orthotrichaceae and the Helicophyllaceae. In the absence of rbcL data, the phylogenetic relationship between these two families remains dubious due to unique gametophytic characters, such as reduced ventral and dorsal leaves, and lateral leaves that are inrolled when dry, and the absence of a peristome in Helicophyllum torquatum, the sole species of the Helicophyllaceae (Vitt, 1982c). Buck and Vitt (1986) argued in favor of a close relationship with the Racopilaceae (ciliate diplolepideae), as suggested by earlier authors (Brotherus, 1909, 1925; Fleischer, 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 pleurocarpous lineage (La Farge-England, 1996). Other loci are currently being tested for addressing the affinities of the Helicophyllaceae. A putative phylogenetic relationship between the Orthotrichaceae and the Cryphaeaceae and the monotypic Wardiaceae (Walther, 1983) was not OF BOTANY [Vol. 85 Fig. 2. Summary consensus trees of most parsimonious Fitch trees found in a heuristic search using rbcL sequence data with Sphagnum palustre and Andreaea rupestris as outgroups. (A): strict consensus of all three islands (39 MPTs; tree 964 steps; CI: 0.390; RI: 0.624); (B): 50% majority rule tree of island 3 (12 MPTs; mean f value: 13 943 6 656); (C): 50% majority rule tree of islands 1 and 2 (27 MPTs; mean f value: 16 633 6 1 308); (D): 50% majority rule tree of all 15 MPTs showing the Orthotrichaceae monophyletic (mean f value: 16722 6 2015). examined in the present study. Wardia hygrometrica should be excluded from the Orthotrichales based on well differentiated alar cells and prosenchymateous cells. Combined with the acrocarpous condition (Welch, 1943), these characters may indicate a bryalean origin, a hypothesis recently examined using 18S gene sequences (Hedderson, Cox, and Gibbings, 1997). The Cryphaeaceae have traditionally been placed among the Leucodontineae (Vitt, 1984; Buck and Vitt, 1986). A relationship of this family to the Hedwigiaceae, and thus the Orthotrichales, 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, 1996). Ordinal relationship of the Orthotrichales—The affinities of the Orthotrichales, here restricted to the Orthotrichaceae, as indicated by comparisons of the rbcL sequences, are within a diplolepideous lineage including the Splachnales and the ciliate mosses. This clade shares a common ancestor with the Haplolepideae, and together they form a sister group to the Funariales (Figs. 3–4). This topology is congruent with Vitt’s hypothesis (Vitt, 1981a) that the opposite diplolepideous peristome is primitive among arthrodontous mosses and that the haplolepideous peristome is derived from such an ancestral type (Vitt, 1984). Alternative hypotheses proposed by Lewinsky (1989; Orthotrichales basal to dichotomy between haplolepideae and ciliate diplolepideae) or by Shaw and Rohrer (1984) and Crosby (1980; ciliate diplo- August 1998] GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY 1331 Fig. 3. One of 39 most parsimonious Fitch trees found based on rbcL sequence data and using Sphagnum palustre and Andreaea rupestris as outgroups. The dotted lines identify branches not present in the strict consensus tree. Bootstrap values (% of 100 replicates) 50% or higher are given below the branch, and decay indexes are presented above the branch. Taxa excluded from the Orthotrichales are in boldface. Fig. 4. Phylogenetic tree reconstructed using the neighbor-joining method, with Tamura’s distance parameter including both transitions and transversions. Bootstrap values (% of 100 replicates) 50% or higher are plotted on the tree. Taxa excluded from the Orthotrichales are in boldface. Abbreviations for generic names in the Orthotrichaceae follow those of Table 4. lepideae are the most primitive mosses) both require five additional steps. The phylogeny obtained by either NJ or MP analysis (Figs. 3, 4) also agrees with Vitt (1984) with regard to the monophyly of a lineage composed of the Splachnales, Orthotrichales and the ciliate mosses. The relationships among these lineages remain however, ambiguous (Fig. 2). The most parsimonious scenario points toward the Splachnales and the Orthotrichales being sister groups. Vitt (1984), by contrast, proposed that the Orthotrichales are sister to the ciliate mosses, a topology that would only require one additional step based on our rbcL data set. Koponen (1977, 1983) considered the genus Brachymitrion to be the most primitive extant member of the Splachnaceae. Unlike in related genera, the thickening on the PPL is heavier than on the OPL, and the PPL also has strong trabeculae (Koponen, 1977, 1982). Both these features are shared with the Funaria and the Bryum peri- stome type (Shaw and Rohrer, 1984). If the Orthotrichales and the Splachnales formed a monophyletic group, both orders would most likely remain natural orders, with the Orthotrichales defined by thick-walled laminal cells. The heavier thickening on the OPL could have arisen independently in both lineages or be a synapomorphy for both, with subsequent reversal in Brachymitrion. In terms of peristome evolution, these lineages would form a plesiomorphic sister group to the ciliate mosses, the latter one being defined by the asymmetric division of the IPL leading to the development of cilia. If we consider the alternative scenario where the Splachnales are sister to a clade composed of the Orthotrichales and the ciliate mosses, the most parsimonious topology based on our data suggests that within the latter, the ciliate mosses remain monophyletic. An extensive taxon sampling in the Bryales is, however, needed before their evolutionary relationship can be addressed more critically. The Ortho- 1332 AMERICAN JOURNAL trichales and the Bryales sensu lato each share one plesiomorphic state with Brachymitrion (Splachnales): either completely aligned cell divisions (Orthotrichales; Lewinsky, 1989) or heavier thickening of the PPL and ventral exostomial trabeculae (typical ciliate mosses). The ciliate mosses have traditionally been defined by a strongly asymmetric division in IPL leading to the development of cilia. While some ‘‘ciliate’’ mosses actually lack cilia (Buck and Vitt, 1986), have a thickened OPL (Eucamptodontopsis, A. Newton, personal communication, Smithsonian), and others even have opposite peristomes (Garovaglia div. sp.; During, 1977; Nishimura and Watanabe, 1992), none has yet been found to have a first-late symmetric division in the IPL. An asymmetric division is elsewhere found only among haplolepideous taxa. Considering that the Rhachitheciaceae are here shown to be of haplolepideous affinity, their peristome with a 2:1 formula most likely results from reduction, through the loss of the asymmetric 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 behind the asymmetric division in both groups is similar, we cannot 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 Mielichhoferia with a peristome formula of 4:2:4 (Shaw and Crum, 1984; Shaw and Rohrer, 1984). Since the possibility of a reversal to a symmetric division cannot be excluded, a bryalean origin of the Orthotrichaceae will need to be further examined in comparison with a broad sample of ciliate taxa. Hedenäs (1994) considered Schlotheimia to be pleurocarpous and, based on this interpretation, suggested that the Orthotrichaceae ‘‘are rather close to the clade where most pleurocarpous mosses belong.’’ He further hypothesized that ‘‘the transition to pleurocarpy must have been gradual,’’ and in this scenario the Orthotrichaceae would occupy an intermediate position. The Macromitrioideae differ, however, from typical pleurocarps by several characters and should rather be considered cladocarpous (La Farge-England, 1996). In genera where both cladocarpy and acrocarpy occur, the former seems to be restricted to terminal taxa, suggesting that the trend is from acrocarpy to cladocarpy with no obvious cases of reversals (La Farge-England, 1996). Such a general trend may suggest that in the Orthotrichaceae 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. Subfamilial phylogenetic relationships—The ten orthotrichaceous genera remaining 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 Orthotrichoideae (Bryodixonia, Orthotrichum, and Ulota), while the macromitrioid clade includes all the Macromitrioideae (Cardotiella, Desmotheca, Groutiella, Macrocoma, Macromitrium, and Schlotheimia) as well as Zygodon obtusifolius (Zygodon OF BOTANY [Vol. 85 sect. Obtusifolii). Zygodon obtusifolius has retained several characters considered here plesiomorphic, such as acrocarpy, smooth, cucullate calyptrae, but exhibits also some derived features reminiscent of some Macromitrioideae (strongly bulging cells, coarse papillae, undifferentiated basal cells, etc.; see Goffinet, 1997c). If Z. obtusifolius is to be retained within the Zygodontoideae, the unexpected relationship with the Macromitrioideae as proposed 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 Kuzoff, 1995). In the latter case, the topology obtained here may represent the ‘‘correct’’ gene tree, but deviates from the true phylogeny of the taxa (Doyle, 1992). This hypothesis is currently being tested by comparing sequence data of the nuclear gene 18S. With the exclusion of the Drummondioideae, Vitt’s (1982a) phylogenetic arrangement of subfamilies would have the Zygodontoideae (cucullate calyptrae) basal to a dichotomy between the Orthotrichoideae and the Macromitrioideae (both typically or exclusively have large mitrate calyptrae). The sister group relationship between the orthotrichoid and the macromitrioid clades proposed here deviates from Vitt’s (1982a) phylogenetic concept of the family by the inclusion of the Zygodontoideae in the former clade, and parsimony would need to be relaxed by six steps for his concept to be satisfied (without constraining the affinities of Z. obtusifolius). The monophyly of Zygodon and thus Zygodontoideae is not supported by rbcL sequence data either, and the relationship of sections Zygodon and Bryoides to the Orthotrichoideae sensu stricto remain unresolved. If Zygodon is indeed paraphyletic, section Bryoides would most likely be the most primitive taxon, given that it shares the plesiomorphic state of ‘‘smooth laminal cells’’ with either a splachnaceous or a typical bryaceous ancestor. The two sections of Zygodon differ on average by 40 mutations, while only 28 changes separate section Zygodon from the Orthotrichoideae, compared to 43 changes between section Bryoides and the Orthotrichoideae. Such divergences may be indicative of the monophyly of a clade composed of section Zygodon and the Orthotrichoideae and support a single origin of papillae that would define this clade. With regards to the Macromitrioideae, both methods of analysis yield two strongly supported monophyletic clades (Figs. 2, 4): one composed of all three species of Schlotheimia, the other including all remaining Macromitrioideae (i.e., Cardotiella, Desmotheca, Groutiella, Macrocoma, and Macromitrium). Vitt, Koponen, and Norris (1993) suggested that Schlotheimia and Cardotiella should be placed in a separate subfamily based on the smooth and distinctly lobate calyptra. RbcL data do not support such relationship as ten more steps are needed to unite Schlotheimia and Cardotiella into a monophyletic clade. The species of Schlotheimia differ, on average, from other Macromitrioideae (including Cardotiella) by 47 bases. By contrast, the average distance between these other macromitrioid genera is only 23 bases. If rates of molecular evolution are assumed to be similar among these taxa (a reasonable assumption considering both clades are predominantly phyllodioicous, and epiphytic in tropical montane forests; see Britten, 1986), dif- 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 B. 5 Bryodixonia; C. 5 Cardotiella; D. 5 Desmotheca; G. 5 Groutiella; M. 5 Macrocoma; Mt. 5 Macromitrium; S. 5 Schlotheimia; O. 5 Orthotrichum; U. 5 Ulota; Z. 5 Zygodon. a Phylogenetic relationships of Orthotrichaceous genera—Analysis of rbcL sequence using either the MP or the NJ method suggests that four of the five larger genera, namely Macromitrium, Orthotrichum, Ulota, and Zygodon, are not monophyletic. The paraphyly of Zygodon has been addressed earlier. Bryodixonia is a monotypic genus endemic to New Zealand. Sainsbury (1945) argued S. tecta S. trichomitria S. brownii C. quinquefaria M. tenue M. papillosa G. chimborazense G. apiculata Mt. incurvifolium Mt. longifolium Mt. richardii D. apiculata Z. intermedius Z. reinwardii Z. pungens Z. obtusifolius O. lyellii O. obtusifolium O. anomalum U. obtusiuscula U. lutea B. perichaetialis ferences in nucleotide sequences may indicate a relatively ancient divergence between Schlotheimia and the other Macromitrioideae. Among the morphological characters that separate Schlotheimia from other Macromitrioideae, those related to the calyptrae (shape, anatomy, outline) are shared only with Cardotiella (Vitt, 1981b), whereas the remaining ones (cell areolation, color of leaves) are plesiomorphic or autapomorphic (Goffinet, 1997c) and thus phylogenetically uninformative. The pleisiomorphic state of ‘‘smooth laminal cells’’ 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. Considering both the morphological and molecular divergences between Schlotheimia and the remaining Macromitrioideae, and the strong support for the monophyly of both clades on molecular grounds, accommodating Schlotheimia in its own subfamily may better reflect the evolutionary relationship among these clades. Brotherus (1925, as Pseudo-Macromitrioideae), Walther (1983), and Crum (1987) isolated Desmotheca from other Macromitrioideae on the basis of its dimorphic sterile and fertile branches. Vitt (1990), however, argued that excluding Desmotheca from the Macromitrioideae would most certainly result in the paraphyly of the later subfamily. Based on variation in the rbcL sequences, the phylogenetic affinities of Desmotheca clearly lay with Macromitrium sensu lato, but remain 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. 3), nested between two Macromitrium species. In the remaining six MPTs as well as in the NJ tree (Figs. 3, 4), Desmotheca is sister to the Macromitrioideae (Schlotheimia excluded). The phenetic distance between Desmotheca and other macromitrioid genera (Table 4; excluding Schlotheimia) are, on average, similar to those between Orthotrichum and Ulota (20 bases), in the Orthotrichoideae, suggesting that placing Desmotheca in a distinct subfamily may not be appropriate. Based on our phylogenetic analyses of the rbcL sequence variation, the Orthotrichaceae could be regarded as composed of two subfamilies, the Orthotrichoideae and the Macromitrioideae, with the latter further divided into two tribes ‘‘Macromitriae’’ and ‘‘Schlotheimiae.’’ Alternatively, the two subfamilies may deserve recognition at the family level as suggested by Churchill and Linares (1995) and the Macromitriaceae would then be composed of two subfamilies, the Macromitrioideae and the ‘‘Schlotheimioideae.’’ The status of the Zygodontoideae, and thus the relationship of the two main subgenera to the Orthotrichoideae, needs further study. 1333 — 0.006 0.012 0.035 0.036 0.035 0.040 0.040 0.036 0.049 0.033 0.031 0.032 0.035 0.045 0.033 0.045 0.039 0.042 0.041 0.047 0.047 8 — 0.012 0.036 0.036 0.033 0.040 0.040 0.038 0.047 0.036 0.033 0.032 0.033 0.041 0.030 0.041 0.034 0.038 0.036 0.042 0.042 16 16 — 0.030 0.031 0.030 0.033 0.033 0.033 0.041 0.031 0.029 0.035 0.035 0.041 0.031 0.041 0.034 0.038 0.038 0.043 0.043 46 48 39 — 0.014 0.015 0.017 0.017 0.017 0.025 0.016 0.018 0.038 0.036 0.045 0.035 0.048 0.039 0.043 0.045 0.049 0.049 48 47 41 19 — 0.012 0.011 0.013 0.015 0.020 0.015 0.017 0.036 0.037 0.042 0.031 0.044 0.033 0.039 0.039 0.044 0.044 46 44 39 20 16 — 0.016 0.017 0.018 0.024 0.020 0.020 0.036 0.036 0.042 0.030 0.044 0.036 0.041 0.039 0.044 0.044 53 53 44 22 15 21 — 0.003 0.014 0.016 0.019 0.020 0.042 0.042 0.045 0.034 0.049 0.040 0.042 0.045 0.049 0.049 53 53 44 22 17 23 4 — 0.017 0.016 0.022 0.020 0.045 0.044 0.045 0.037 0.051 0.042 0.044 0.046 0.049 0.051 48 50 43 23 20 24 19 23 — 0.023 0.014 0.017 0.036 0.034 0.044 0.032 0.047 0.038 0.044 0.042 0.047 0.047 64 62 54 33 26 31 21 21 30 — 0.030 0.029 0.048 0.049 0.050 0.042 0.055 0.045 0.048 0.050 0.053 0.055 44 47 41 21 20 26 25 29 18 40 — 0.013 0.033 0.035 0.045 0.032 0.048 0.037 0.045 0.045 0.050 0.050 41 43 38 24 23 27 26 26 23 38 17 — 0.037 0.035 0.045 0.030 0.045 0.036 0.042 0.041 0.046 0.046 42 42 46 50 48 47 55 59 48 64 43 49 — 0.009 0.031 0.034 0.024 0.018 0.023 0.023 0.026 0.026 46 44 46 47 49 48 56 58 45 65 46 46 12 — 0.030 0.036 0.021 0.014 0.019 0.020 0.023 0.021 60 54 54 60 56 56 59 59 58 66 59 59 41 39 — 0.037 0.036 0.027 0.032 0.033 0.036 0.035 44 40 41 46 41 40 45 49 42 55 42 39 45 47 49 — 0.042 0.035 0.037 0.037 0.042 0.042 60 54 54 63 58 58 65 67 62 72 64 60 32 28 47 55 — 0.015 0.017 0.009 0.012 0.009 51 45 45 52 44 47 53 55 50 60 49 47 24 18 35 46 20 — 0.011 0.017 0.020 0.017 56 50 50 57 52 54 56 58 58 63 60 56 31 25 42 49 22 15 — 0.017 0.018 0.017 54 48 50 59 52 52 59 61 56 66 60 54 30 26 43 49 12 22 22 — 0.011 0.009 62 56 56 65 58 58 65 65 62 70 66 60 34 30 47 55 16 26 24 14 — 0.005 62 56 56 65 58 58 65 67 62 72 66 60 34 28 46 55 12 22 22 12 6 — ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 GOFFINET TABLE 4. Pairwise comparison of rbcL nucleotide sequences within the Orthotrichaceae. Below diagonal is absolute distance; above diagonal is average distance. Distance values used for intrageneric comparisons (see text) are in boldface.a August 1998] 1334 AMERICAN JOURNAL for a generic distinction from Ulota on the basis of ‘‘the highly differentiated and conspicuous perichaetial bracts and the diminutive calyptra’’ and the immersed capsule. In addition, the perichaetial leaves have prorate basal laminal cells (Goffinet, 1997c), a feature otherwise unknown from the Orthotrichoideae. Bryodixonia does, however, share many of the characters found in the genus Ulota, such as very thick-walled cauline cells, the differentiated marginal cells of the lamina, and the flexuose to crisped leaves. Bryodixonia may thus be patristically very derived; however, in cladistic terms it may not deserve taxonomic recognition at the generic level. RbcL sequences of Ulota lutea and Bryodixonia perichaetialis differ only by six mutations, a degree of divergence that is similar to that found between species of Groutiella (4) or Schlotheimia (8–16), but, moreover, it is less than the divergence recorded between the two species of Ulota (14; Table 4). Immersed capsules are characteristic of many mosses that are taxonomically unrelated but share a xerophytic habitat (Vitt, 1981a). Within the Orthotrichaceae completely immersed capsules are characteristic for Schlotheimia sect. Stegotheca, and a shortening of the setae leading ultimately to an immersed capsule occurs in Orthotrichum subg. Gymnoporus sect. Leiocarpa, subg. Pulchella sect. Rivularia, and sect. Diaphana, as well as in subg. Orthotrichum (Lewinsky, 1993). Considering the low degree of morphological and molecular divergence of Bryodixonia, segregation at the generic level does not seem appropriate. The monophyly of Ulota (including B. perichaetialis) is compromised by O. lyellii in the NJ tree and in 13 MPT (in 13 other MPTs their relationship is not resolved). Strong affinities of O. lyellii for Ulota (MP: 99% BV and DI of 5) and U. obtusiuscula in particular (13 MPTs) may indicate that among the different lineages of Orthotrichum, subg. Gymnoporus (Braithw.) Limpr. is the most closely related to Ulota. Among the characters that subg. Gymnoporus sect. Leiocarpa (see description in Lewinsky, 1993) shares with Ulota, only the long flexuose vegetative leaves may be derived within the Orthotrichoideae and thus be indicative of common ancestry. The genus Ulota (even if including Bryodixonia) is morphologically well defined from Orthotrichum, suggesting that the paraphyly of Orthotrichum, if confirmed, 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. Orthophyllum Delogn. (O. obtusifolium) and subg. Orthotrichum (O. anomalum) are not unambiguously resolved either: they form sister taxa in 13 MPTs as well as the NJ tree, while in 13 other MPTs, O. obtusifolium is sister to a clade composed of the remaining Orthotrichoideae. Orthotrichum obtusifolium had been placed together with the related O. gymnostomum Brid. in the genus Nyholmiella (see Lewinsky, 1993, for history), based on ‘‘the obtuse leaves with plane or incurved leaf margin and incrassate leaf-cells with a stout central papillae on each side’’ (Lewinsky, 1993). Patterns in peristome ornamentation are similar to those observed elsewhere in the genus (Lewinsky, 1993), and it may therefore be more parsimonious to retain subg. Orthophyllum in Orthotrichum. Within the macromitrioid clade (Fig. 4) the relation- OF BOTANY [Vol. 85 ships remain ambiguous too (except for Schlotheimia see above), either because Macromitrium and Macrocoma truly are paraphyletic, or because of an insufficient taxon sample. The genus Macromitrium is, with over 250 species, by far the most speciose genus of the Orthotrichaceae (Vitt, 1982a). Mitten (1869) divided the genus in four sections, to which Buck (1991) recently added sect. Reverberatum. Two of these have recently been raised to the genus level, namely Micromitrium (now Groutiella Steere) and Macrocoma Grout. The genus Macromitrium remains, however, morphologically extremely diverse in terms of size of the plant, degree of differentiation of the basal cells, shape of the urn, and laminal cell shape, orientation, and ornamentation. The relatively high cost in terms of parsimony, for a monophyletic Macromitrium (nine steps) may be seen as just one other indication that the genus as currently defined is still a heterogeneous assemblage. Groutiella differs from Macromitrium by the marginal limbidium of hyaline elongate cells and a short calyptra covering only the upper portion of the urn. Except for G. tomentosa, Groutiella is restricted to the Neotropics, where ten species occur. The sister species in all shortest trees is Macromitrium longifolium, a neotropical endemic, rather than any of two paleotropical taxa (M. richardii is known from Africa and the Americas [van Rooy and van Wyk, 1992; Vitt 1993] and belongs to M. ligulare group of the Old World; M. incurvifolium occurs throughout the Pacific Ocean [Vitt and Ramsay, 1985]). Whether these putative affinities of Groutiella for neotropical Macromitria indicate a common ancestry with a distinct neotropical lineage of Macromitrium needs to be further investigated. The genus Macrocoma is composed of two subgenera, subg. Trachyphyllum (M. papillosa) and subg. Macrocoma (M. tenuis), that differ by a series of characters, but particularly by the well-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. Macrocoma papillosa is found in a basal position among macromitrioid taxa (excluding Schlotheimia) in 33 MPTs (Fig. 4), while in the remaining six it is found in a more derived position with both species of Groutiella, Macromitrium longifolium, and Macrocoma tenuis. While the Macromitrioideae are typically cladocarpous (i.e., with their perichaetia terminal on lateral branches) and have dimorphic leaves between stem and branches, M. papillosa is clado- and acrocarpous (i.e, with perichaetia terminal on lateral branches and on the stem) and the leaves are not differentiated into stem and branch leaves (Goffinet, 1997c). True acrocarpy also occurs in subg. Macrocoma (e.g., M. braziliensis [Mitt] Vitt), while other taxa of this subgenus are strictly cladocarpous (e.g., M. tenuis [Hook. & Grev.] Vitt). Acrocarpy and cladocarpy have not been reported before from the same taxon, not to mention from the same individual (see La FargeEngland, 1996). The combination of plesiomorphies such as undifferentiated 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. Macrocoma (Vitt, 1982c), but maybe even with regard to the evolution of the Macromitrioideae. August 1998] GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY Affinities of excluded taxa—Critically addressing the relationships of the taxa here excluded from the Orthotrichaceae is beyond the scope of the present study and would need a broader sampling of haplolepideous taxa. A suite of unique characters combined with the lack of characters that are phylogenetically crucial may always hamper determining sister group relationships based on morphology only. Crum (1987) already suggested that gametophytic characters may not suffice to resolve the phylogenetic relationship of Drummondia. The same opinion prevails with regards to Amphidium. Brotherus (1925), Anderson and Crum (1958), and Vitt (1973, 1982b, 1984) placed Amphidium near Rhabdoweisia. Our results, though preliminary, do not suggest that these genera are closely related, and future study may need to consider alternative relationships, as for example with Glyphomitrium, due to the overall similarity of Amphidium lapponicum-Glyphomitrium daviesii. Presence of a peristome allows for a more explicit hypothesis to be made regarding the systematic position of the Erpodiaceae and the Rhachitheciaceae. Edwards (1979), as part of a review of the haplolepideous peristome, examined the peristome architecture of Venturiella and concluded that the teeth are ‘‘strongly dorsally trabeculate, and also have a rudimentary unthickened basal exostome,’’ and that ‘‘these characters are of a haplolepideous peristome although not of the dicranaceous type’’ but of a distinct type, the Seligeria type. This peristome type is characterized by little ventral thickening, strong dorsal trabeculae, and an exostome reduced to a thin, smooth membrane adhering to the trabeculae. This combination of characters has also been observed in the Rhachitheciaceae and has been interpreted as a possible indication of haplolepideous affinities of the family (Goffinet, 1997a). Molecular data thus tend to confirm these hypotheses, and consequently the peristome of the Rhachitheciaceae and the Erpodiaceae should be regarded as derived, through reduction, from a typical haplolepideous peristome. Though the monophyly of a group of taxa sharing the Seligeria-type peristome has not been critically examined, the nearly identical peristomes of Rhachithecium, Glyphomitrium (Ptychomitriaceae), and Blindia (Seligeriaceae) may be seen as an indication of close phylogenetic relationships, despite gametophytic differences. Alternatively, the Rhachitheciaceae may be more closely related to Rhabdoweisia in the Dicranales, considering the similarities in the overall habit, leaf shape, and cell shape and differentiation. Phylogenetic conclusions—Sequence data of the chloroplast gene rbcL are useful in circumscribing the Orthotrichales, particularly with regard to the systematic position of taxa lacking peristome features that are central to the classification of mosses. Analyses of the variation in the nucleotide sequence using either the parsimony or the distance method strongly suggest that the Orthotrichales are polyphyletic and that the Erpodiaceae and the Rhachitheciaceae are of haplolepideous affinities as suggested by their Seligeria-type peristome. The Orthotrichaceae too, are shown to be an artificial assemblage due to the current inclusion of Amphidium and Drummondia, two genera better placed among the Haplolepideae. The Orthotrichaceae are only distantly related to the latter 1335 clade and should rather be considered a member of a derived diplolepideous clade. The relationship to the Splachnales and the ciliate mosses remains unsettled, but at present all three lineages are best considered monophyletic. Molecular data do not support Vitt’s (1982a) subfamilial phylogeny and instead suggest that the Orthotrichaceae are composed of two lineages. The first consists of the Zygodontoideae and the Orthotrichoideae, while the second includes all Macromitrioideae. Putative basal taxa within these clades may be represented by Zygodon sect. Bryoides and Schlotheimia, respectively, which are characterized by smooth laminal cells. The Orthotrichaceae are now composed of 20 genera: Bryodixonia, Cardotiella, Ceuthotheca, Desmotheca, Florschuetziella, Groutiella, Leiomitrium, Leptodontiopsis, Leratia, Macrocoma, Macromitrium, Muelleriella, Orthomitrium, Orthotrichum, Pleurorthotrichum, Schlotheimia, Stenomitrium, Stoneobryum, Ulota, and Zygodon. Examination of cladistic relationships and associated phenetic distances suggests that the monotypic genus Bryodixonia may be better regarded as a patristically derived species of Ulota. Our molecular data furthermore reveal that larger genera such as Macromitrium and Zygodon may merely represent evolutionary grades. Gene data obviously have made a significant contribution in resolving the circumscription of the Orthotrichaceae and the relationship of the main lineages. Above all, however, this molecular study has laid the foundation for critically reexamining the morphological characters that are central to the generic concept used in the Orthotrichaceae. LITERATURE CITED ANDERSON, L. E., AND H. CRUM. 1958. Cytotaxonomic studies on mosses of the Canadian Rocky Mountains. Bulletin of the National Museum of Canada 160: 1–89. ANDERSSON, I., S. KNIGHT, G. SCHNEIDER, Y. LINDQVIST, T. LUNDQVIST, C.-I. BRÄNDEN, AND G. H. LORIMER. 1989. Crystal structure of the active site of ribulose-bisphosphate carboxylase. Nature 337: 229– 234. BARKER, N. P., H. P. LINDER, AND E. H. HARLEY. 1995. Polyphyly of Arundinoideae (Poaceae): evidence from rbcL sequence data. Systematic Botany 20: 423–435. BAUM, D. A., K. J. SYTSMA, AND P. C. HOCH. 1994. A phylogenetic analysis of Epilobium (Onagraceae) based on nuclear ribosomal DNA sequences. Systematic Botany 19: 363–388. BLOMQUIST, H. L., AND L. L. ROBERTSON. 1941. The development of the peristome in Aulacomnium heterostichum. Bulletin of the Torrey Botanical Club 68: 569–584. BOPP, M., AND I. CAPESIUS. 1996. New aspects of Bryophyte taxonomy provided by a molecular approach. Botanica Acta 109: 368–372. BREMER, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 46: 1801–1817. BRITTEN, R. J. 1986. Rates of DNA sequence evolution differ between taxonomic groups. Science 231: 1393–1398. BROTHERUS, V. F. 1909. Musci (Laubmoose). III. Unterklasse Bryales: II Spezieller Teil. In A. Engler’s and K. Prantl [eds.], Die Natürlichen Pflanzenfamilien 1 (3, I, II), 277–1246. W. Engelmann, Leipzig. . 1925. Musci (Laubmoose) III. Unterklasse Bryales: II Spezieller Teil. In A. Engler’s and K. Prantl [eds.], Die Natürlichen Pflanzenfamilien, 2d ed., vol. 11, 1–542. W. Engelmann, Leipzig. BUCK, W. R. 1991 Contributions to the moss flora of Guyana. Memoirs of the New Botanical Garden 64: 184–196. , AND D. H. VITT. 1986. Suggestions for a new familial classification of pleurocarpous mosses. Taxon 35: 21–60. CAPESIUS, I. 1995. A molecular phylogeny of Bryophytes based on the 1336 AMERICAN JOURNAL nuclear encoded 18S rRNA genes. Journal of Plant Physiology 146: 59–63. CHASE, M. W., ET AL. 1993. Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580. CHURCHILL, S. P., AND E. L. LINARES C. 1995. Prodromus Bryologiae Novo-Granatensis. Introduccion a la Flora de Musgos de Colombia. Parte II. Bibliotheca ‘‘Jose Jeronimo Triana’’ 12: 455–924. CONTI, E., A. FISCHBACH, AND K. J. SYTSMA. 1993. Tribal relationships in Onagraceae: Implications from rbcL sequence data. Annals of the Missouri Botanical 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, 115–129. California Academy of Science, San Francisco, CA. CRUM, H. 1972. A taxonomic account of the Erpodiaceae. Nova Hedwigia 23: 201–224. . 1987. Leratia and its place in the Orthotrichaceae, subfamily Macromitrioideae. Memoirs of the New York Botanical Garden 45: 602–606. DE LUNA, E. 1995. The circumscription and phylogenetic relationships of the Hedwigiaceae (Musci). Systematic Botany 20: 347–373. DIXON, H. N. 1932. Classification of mosses. In F. Verdoorn [ed.], Manual of bryology, 397–412. 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. Annals of the Missouri Botanical Garden 79: 333–345. DOYLE, J. J. 1992. Gene trees and species trees: molecular systematics as one-character taxonomy. Systematic Botany 17: 144–163. , AND J. L. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11– 15. DURING, H. J. 1977. A taxonomical revision of the Garovaglioideae (Pterobryaceae, Musci). Bryophytorum Bibliotheca 12: 1–244. EDWARDS, S. R. 1979. Taxonomic implications of cell patterns in haplolepidous moss peristomes. In G. S. C. Clarke and J. G. Duckett [eds.], Bryophyte systematics, 317–346. Academic Press, London. . 1984. Homologies and inter-relations of moss peristomes. In R. M. Schuster [ed.], New manual of bryology, vol. 2, 658–695. Hattori Botanical Laboratory, Nichinan. FARRIS, J. S. 1972. Estimating phylogenetic trees from distance matrix. American Naturalist 106: 645–668. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. FITCH, W. M. 1971. Toward defining the course of evolution: minimum change for a specific tree topology. Systematic Zoology 20: 406– 416. FLEISCHER, M. 1920. Natürlisches System der Laubmoose. Hedwigia 61: 390–400. FREY, W. 1977. Neue Vorstellungen über die Verwandschaftengruppen und die Stammesgeschichte der Laubmoose. In W. Frey, H. Hurka, and F. Obervwinkler [eds.], Beiträge zur Biologie der niederen Pflanzen, 117–139. Stuttgart. . 1990. Genoelemente prä-angiospermen Ursprungs bei Bryophyten. Botanischer Jahresbericht der Systematik 111: 433–456. , J.-P. FRAHM, E. FISCHER, AND W. LOBIN. 1995. Die Moos- und Farnpflanzen Europas. Kleine Kryptogamenflora Band IV: XII 1 426 p. Gustav Fischer Verlag, Stuttgart. GOFFINET, B. 1997a (1998). The Rhachitheciaceae: revised circumscription and ordinal affinities. Bryologist 100: 425–439. . 1997b. A reconsideration of the affinities of Kleioweisiopsis, Pleurozygodontopsis, Trigonodictyon, and the Microtheciellaceae (Bryopsida, Orthotrichales). Journal of Bryology, in press. . 1997c. Circumscription and phylogeny of the Orthotrichales. Ph.D. dissertation. University of Alberta, Edmonton, Canada. HASEBE, M., P. G. WOLF, K. M. PRYER, K. UEDA, M. ITO, R. SANO, G. J. GASTONY, J. YOKOYAMA, J. R. MANHART, N. MURAKAMI, E. H. CRANE, C. H. HAUFLER, AND W. D. HAUK. 1995. Fern phylogeny based on rbcL nucleotide sequences. American Fern Journal 85: 134–181. HAUFLER, C. H., AND T. A. RANKER. 1995. RbcL sequences provide phylogenetic insights among sister species of the fern genus Polypodium. American Fern Journal 85: 361–374. HEDDERSON, T. A., R. C. CHAPMAN, AND W. R. ROOTES. 1996. Phylo- OF BOTANY [Vol. 85 genetic relationships of bryophytes inferred from nuclear-encoded rRNA gene sequences. Plant Systematic and Evolution 200: 213– 224. , C. J. COX, AND J. G. GIBBINGS. 1997. Phylogenetic relationships of the Wardiaceae (Musci) inferred from 18s rRNA and rps4 gene sequences. American Journal of Botany 84 (Supplement): 25. HEDENÄS, L. 1994. The basal pleurocarpous diplolepideous mosses—a cladistic approach. Bryologist 97: 225–243. HENDY, M. D., AND D. PENNY. 1989. A framework for the quantitative study of evolutionary trees. Systematic Zoology 38: 297–309. HILLIS, D. M., AND J. P. HUELSENBECK. 1992. Signal, noise, and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189–195. , AND J. J.BULL. 1993. An empirical test for bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology 42: 182–192. KÄLLERSJÖ, M., J. S. FARRIS, A. G. KLUGE, AND C. BULT. 1992. Skewness and permutation. Cladistics 8: 275–287. KÄSS, E., AND M. WINK. 1995. Molecular phylogeny of the Papilionoideae (family Leguminosae): rbcL gene sequences versus chemical taxonomy. Botanica Acta 108: 149–162. KELLOG, E. A., AND N. D. JULIANO. 1997. The structure and function of RuBisCo and their implications for systematic studies. American Journal of Botany 84: 413–428. KIM, J., F. J. ROHLF, AND R. R. SOKAL. 1993. The accuracy of phylogenetic estimation using the neighbor-joining method. Evolution 47: 471–486. KOPONEN, A. 1977. The peristome and spores in Splachnaceae and theirs evolutionary and systematic significance. Bryophytorum Bibliotheca 13: 535–567. . 1982. On the structure and function of the peristome in Splachnaceae. Journal of the Hattori Botanical Laboratory 53: 73–98. . 1983. Studies on the generic concept in the classification of the moss family Splachnaceae. Publication from the Department of Botany, University of Helsinki 11: 1–48. KUMAR, S., K. TAMURA, AND M. NEI. 1993. MEGA: Molecular Evolutionary Genetic Analysis, version 1.0. Computer program distributed by Pennsylvania State University, University Park, PA. LA FARGE-ENGLAND, C. 1996. Growth-form, branching pattern, and perichaetial position in mosses: cladocarpy and pleurocarpy redefined. Bryologist 99: 170–186. LEWINSKY, J. 1976. On the systematic position of Amphidium Schimp. Lindbergia 3: 227–231. . 1989. Does the orthotrichaceous type of peristome exist? A study of peristome evolution in the genus Orthotrichum Hedw. with possible derivation of the haplolepideous peristome. Journal of the Hattori Botanical Laboratory 67: 335–363. . 1990. Zygodon Hook. & Tayl. in Australasia: a taxonomic revision including SEM studies of peristomes. Lindbergia 15: 109– 139. . 1993. A synopsis of the genus Orthotrichum Hedw. (Musci, Orthotrichaceae). Bryobrothera 2: 1–59. LEWINSKY-HAAPASAARI, J., AND M. R. CROSBY. 1996. Orthomitrium tuberculatum (Orthotrichaceae), a new genus and species from Guizhou, China. Novon 6: 1–5. MADDISON, D. R. 1991. The discovery and importance of multiple islands of most-parsimonious trees. Systematic Zoology 40: 315–328. MALTA, N. 1926. Die Gattung Zygodon Hook. et Tayl. Eine monographische Studie. Latvijas Universitates Botaniská Dárza Darbi 1: 1– 184. MANHART, J. R. 1994. Phylogenetic analysis of green plant rbcL sequences. Molecular Phylogenetics and Evolution 3: 114–127. MEUSEL, H. 1935. Wuchsformen und Wuchstypen der europäischen Laubmoose. Nova Acta Leopoldina Neue Folge 3: 123–277. MISHLER, B. D., L. A. LEWIS, M. A. BUCHHEIM, K. S. RENZAGLIA, D. J. GARBARY, C. F. DELWICHE, F. W. ZECHMAN, T. S. KANTZ, AND R. L. CHAPMAN. 1994. Phylogenetic relationships of the ‘‘green algae’’ and ‘‘bryophyte.’’ Annals of the Missouri Botanical Garden 81: 451–483. , P. H. THRALL, J. S. HOPPLE, JR., E. DE LUNA, AND R. VILGALYS. 1992. A molecular approach to the phylogeny of bryophytes: cladistic analysis of chloroplast-encoded 16S and 23S ribosomal RNA genes. Bryologist 95: 172–180. August 1998] GOFFINET ET AL.—ORTHOTRICHALES: CIRCUMSCRIPTION AND PHYLOGENY MITTEN, W. 1869. Musci Austro-americani. Journal of the Linnean Society of Botany 12: 1–659. MORGAN, D. R., AND D. E. SOLTIS. 1993. Phylogenetic relationships among members of Saxifragaceae sensu lato based on rbcL sequence data. Annals of the Missouri Botanical Garden 80: 631– 660. NISHIMURA, N., AND K. WATANABE. 1992. SEM observations on the peristome teeth of mosses, 2. Isobryales. Natural Environmental Science Research 5: 83–101. OLMSTEAD, R. G., AND J. A. SWEERE. 1994. Combining data in phylogenetic systematics: An empirical approach using three molecular data sets in the Solanaceae. Systematic Biology 43: 467–481. PRYER, K. M., A. R. SMITH, AND J. E. SKOG. 1995. Phylogenetic relationships of extant ferns based on evidence from morphology and rbcL sequences. American Fern Journal 85: 205–282. REIMERS, H. 1954. Bryophyta. In H. Melchior AND E. Wedermann [eds.], A. Engler’s Syllabus der Pflanzenfamilien, 12th ed., 218–268. Gebrürder Bornträgger, Berlin. ROBINSON, H. 1971. A revised classification for the orders and families of mosses. Phytologia 21: 289–293. RUSSO, C. A. M., N. TAKEZAKI, AND M. NEI. 1996. Efficiencies of different genes and different tree-building methods in recovering a known vertebrate phylogeny. Molecular Biology and Evolution 13: 525–536. SAINSBURY, G. O. K. 1945. New and critical 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 trees. Molecular Biology and Evolution 4: 406–425. SANG, T. 1995. New measurements of distribution of homoplasy and reliability of parsimonious cladograms. Taxon 44: 77–82. SANGER, F., S. NICKLEN, AND A. R. COULSON. 1977. DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, USA 74: 5463–5467. SCHWARTZ, O. M. 1994. The development of the peristome-forming layers in the Funariaceae. International Journal of Plant Sciences 155: 640–657. SHAW, J. 1985. Peristome structure in the Orthotrichaceae. Journal of the Hattori Botanical Laboratory 60: 119–136. , L. E. ANDERSON, AND B. D. MISHLER. 1989. Peristome development in mosses in relation to systematics and evolution. III. Funaria hygrometrica, Bryum pseudocapillare, and B. bicolor. Systematic Botany 14: 24–36. , AND H. CRUM. 1984. Peristome homology in Mielichhoferia and a taxonomic account of North American species. Journal of the Hattori Botanical Laboratory 57: 363–381. , B. D. MISHLER, AND L. E. ANDERSON. 1989. Peristome development in mosses in relation to systematics and evolution. IV. Haplolepidae: Ditrichaceae and Dicranaceae. Bryologist 92: 314–325. , AND J. R. ROHRER. 1984. Endostomial architecture in diplolepideous mosses. Journal of the Hattori Botanical Laboratory 57: 41–61. SOLTIS, D. E., AND R. K. KUZOFF. 1995. Discordance between nuclear and chloroplast phylogeneis in the Heuchera group (Saxifragaceae). Evolution 49: 727–742. SWOFFORD, D. L. 1993. PAUP: phylogenetic analysis using parsimony, 1337 version 3.1.1. Computer Program distributed by the Illinois Natural History Survey, Champaign, IL. VAN ROOY, J., AND A. E. VAN WYK. 1992. A conspectus of the subfamily Macromitrioideae (Bryopsida: Orthotrichaceae) in Southern Africa. Bryologist 95: 205–215. VITT, D. H. 1970. The family Orthotrichaceae (Musci) in North America, north of Mexico. Ph.D. dissertation. University of Michigan, Ann Arbor, MI . 1972. A monograph of the genus Drummondia. Canadian Journal of Botany 50: 1191–1208. . 1973. A revision of the genus Orthotrichum in North America, north of Mexico. Bryophytorum Bibliotheca 1: 1–208. . 1980 (1981). The genus Macrocoma I. Typification of names and taxonomy of the species. Bryologist 83: 405–436. . 1981a. Adaptive modes of the moss sporophyte. Bryologist 84: 166–186. . 1981b. The genera Leiomitrium and Cardotiella gen. nov. (Orthotrichaceae). Journal of the Hattori Botanical Laboratory 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, NY. . 1982b. The genera of the Orthotrichaceae. Beiheft zu Nova Hedwigia 71: 261–268. . 1982c. Populational variation and speciation in austral mosses. Journal of the Hattori Botanical Laboratory 52: 153–159. . 1984. Classification of the Bryopsida. In R. M. Schuster [ed.], New manual of bryology, vol. 2. Hattori Botanical Laboratory, Nichinan, Japan. . 1990. Desmotheca (Orthotrichaceae): Gondwanan fragmentation and the origin of a Southeast Asian genus. Tropical Bryology 3: 79–88. . 1993. Orthotrichaceae. In A. J. Sharp, H. A. Crum, AND P. M. Eckel [eds.], The moss flora of Mexico, 590–657. New York Botanical Garden, New York, NY. , B. GOFFINET, AND T. A. HEDDERSON. 1997. The ordinal classification of the mosses: Questions and answers for the 1990’s. In J. G. Duckett, N. W. Ashton, and J. W. Bates [eds], Bryology for the second century. Leeds: W. S. Mary and the British Bryological Society, in press. , T. KOPONEN, AND D. H. NORRIS. 1993. Bryophyte flora of the Huon Peninsula, Papua New Guinea. LIII. Ulota and Schlotheimia (Orthotrichaceae, Musci). Acta Botanica Fennica: 148: 5–25. , AND H. P. RAMSAY. 1985. The Macromitrium complex in Australasia (Orthotrichaceae: Bryopsida). Part II. Distribution, ecology, and paleogeography. Journal of the Hattori Botanical Laboratory 59: 453–468. WALTHER, K. 1983. Bryophytina. Laubmoose. In J. Gerloff AND J. Poelt [eds.], A. Engler’s Syllabus der Pflanzenfamilien. Aufl. 13, V(2): 1–108. Gebrüder Bornträger, Berlin. WATERS, D. A., M. A. BUCHHEIM, R. A. DEWEY, AND R. C. CHAPMAN. 1992. Preliminary inferences of the phylogeny of bryophytes from nuclear-encoded ribosomal RNA sequences. American Journal of Botany 79: 459–466. WELCH, W. H. 1943. The systematic position of the genera Wardia, Hydropogon, and Hydropogonella. Bryologist 46: 25–46. WOLF, P. G., P. S. SOLTIS, AND D. E. SOLTIS. 1994. Phylogenetic relationships of Dennstaedtioid ferns: evidence from rbcL sequences. Molecular Phylogenetics and Evolution 3: 383–392.

RELATED PAPERS

RELATED TOPICS

Log In

Circumscription and Phylogeny of the Orthotrichales (Bryopsida) Inferred from RbcL Sequence Analyses

Circumscription and Phylogeny of the Orthotrichales (Bryopsida) Inferred from RbcL Sequence Analyses

Related Papers

RELATED PAPERS

RELATED TOPICS