sign in sign up
<<
Home , Adelanthaceae, Anastrophyllum, Bazzania, Bryopteris, Calycularia, Calypogeia

Org Divers Evol (2016) 16:727–742 DOI 10.1007/s13127-016-0284-4

ORIGINAL ARTICLE

A phylogeny of () based on nuclear and chloroplast DNA markers

Kathrin Feldberg1 & JiříVáňa2 & Johanna Krusche1 & Juliane Kretschmann1 & Simon D. F. Patzak1 & Oscar A. Pérez-Escobar1 & Nicole R. Rudolf1 & Nathan Seefelder1 & Alfons Schäfer-Verwimp3 & David G. Long 4 & Harald Schneider5,6 & Jochen Heinrichs1

Received: 15 February 2016 /Accepted: 27 April 2016 /Published online: 7 May 2016 # Gesellschaft für Biologische Systematik 2016

Abstract Cephaloziaceae represent a subcosmopolitan line- but the bicuspidata complex and the Cephalozia age of largely terrestrial leafy liverworts with three-keeled hamatiloba complex require further study. A Neotropical perianths, a reduced seta, capsules with bistratose walls, fila- clade of originates from temperate ancestors. mentous sporelings, large, thin-walled cells, and vegetative Odontoschisma yunnanense is described as new to science. distribution by gemmae. Here we present the most compre- hensively sampled phylogeny available to date based on the Keywords Cephaloziineae . . nuclear ribosomal internal transcribed spacer region and the Integrative . . Liverworts chloroplast markers trnL-trnF and rbcL of 184 accessions representing 41 of the 89 currently accepted and four of the five currently accepted subfamilies. Alobielloideae are Introduction placed sister to the remainder of Cephaloziaceae. Odontoschismatoideae form a sister relationship with a clade Molecular phylogenies have greatly improved our knowledge consisting of Schiffnerioideae and Cephalozioideae. of liverwort evolution. Studies incorporating molecular evi- Cephalozioideae are subdivided in three genera, dence have led to numerous adjustments of morphology- Fuscocephaloziopsis, Cephalozia,andNowellia, the last two based family and circumscriptions (Hentschel et al. in a robust sister relationship. Most morphological species 2006a;Feldbergetal.2009; Engel et al. 2010;Heand circumscriptions are supported by the molecular topologies Glenny 2010; Heinrichs et al. 2012b;Wangetal.2014; Shaw et al. 2015; Bechteler et al. 2016; Patzak et al. 2016; Villarreal et al. 2016) and the recognition of ignored species Electronic supplementary material The online version of this article (doi:10.1007/s13127-016-0284-4) contains supplementary material, (Rycroft et al. 2004; Heinrichs et al. 2010, 2015; which is available to authorized users. Vanderpoorten et al. 2010; Ramaiya et al. 2010;Renner et al. 2013b;Arandaetal.2014; Bakalin and Vilnet 2014). * Jochen Heinrichs These studies demonstrated the need for integrative studies [email protected] combining evidence from multiple sources (Dayrat 2005; Padial et al. 2010). They also showed that differences between 1 Department of Biology I and GeoBio-Center, Ludwig Maximilian morphologically semicryptic species “are virtually impossible University, Menzinger Str. 67, 80638 Munich, Germany to apprehend independent of molecular data corroborating 2 Department of , Charles University, Benátská 2, 128 01 Praha their significance” (Renner 2014). Accordingly, comprehen- 2, Czech Republic sive species-level phylogenies with extensive population sam- 3 Mittlere Letten 11, 88634 Herdwangen-Schönach, Germany pling are needed to estimate the actual liverwort diversity 4 Royal Botanic Garden, Edinburgh EH3 5LR, UK (Fuselier et al. 2009; Kreier et al. 2010). 5 School of Life Sciences, Sun Yatsen University, Jungermanniales are one of the main lineages of leafy liv- Guangzhou 510275, Guandong, China erworts. They include more than 3000 generalistic species 6 Department of Life Science, Natural History Museum, London SW7, growing mainly on soil, rock, or trunk bases rather than as 5BD, UK true epiphytes (Heinrichs et al. 2005; Feldberg et al. 2014; 728 K. Feldberg et al.

Söderström et al. 2016). He-Nygrén et al. (2006) accepted use a wide genus concept of Cephalozia with four suborders of Jungermanniales. One of these is the Fuscocephaloziopsis, Metahygrobiella, , Cephaloziineae, a cosmopolitan clade characterized by the Pleurocladula,andSchofieldia as its heterotypic synonyms. presence of usually succubous, undivided or two-lobed Here we present the most comprehensively sampled phy- , very small or absent underleaves, scattered rhizoids, logeny of Cephaloziaceae based on three molecular markers frequent presence of ventral branches, and usual- (nrITS1-5.8S-ITS2 region, cp trnL-trnF, rbcL) and discuss ly enclosed by a calyptra (Crandall-Stotler et al. 2009). Its different approaches to resolving monophyletic genera. We families , , and Cephaloziaceae test current species circumscriptions by including multiple were subject to several molecular phylogenetic studies, lead- accessions and examine whether the recovered phylogenetic ing to numerous generic and subgeneric refinements (e.g., relationships correspond to or conflict with these morpholog- Feldberg et al. 2010;Vilnetetal.2010, 2012;Heinrichs ically circumscribed entities. et al. 2012a); however, the classification of Cephaloziaceae is still controversial (Vilnet et al. 2012; Potemkin and Sofronova 2013;Váňaetal.2013). Materials and methods Cephaloziaceae have three-keeled perianths with the third keel ventral, a reduced, thin seta, ovoid-ellipsoidal capsules Taxon sampling, outgroup selection, and morphology with bistratose walls, filamentous sporelings, and usually large, thin-walled cells. Vegetative distribution by gemmae Taxon sampling was based on Schuster (2002), Crandall- on ascending flagelliferous shoots is common. Until recently, Stotler et al. (2009), Váňaetal.(2013), and Söderström et al. some 15 genera were accepted in this family (Gradstein et al. (2016). Initial species identification relied on the treatments of 2001; Crandall-Stotler et al. 2009) of which several were Fulford (1968), Váňa(1988), Schuster (1974, 2002), Piippo paraphyletic or polyphyletic in molecular phylogenies. (1984), Paton (1999), Bednarek-Ochyra et al. (2000), Vilnet et al. (2012) and Aranda et al. (2014) showed that Damsholt (2002), Engel and Glenny (2008), and Potemkin H.Buch and Iwatsukia N.Kitag. are nested in and Sofronova (2013). Taxa used in the molecular study, in- Odontoschisma (Dumort.) Dumort., and Váňaetal.(2013) cluding GenBank accession numbers and voucher details, are and Gradstein et al. (2014) transferred its two and four respec- listed in Supplementary file 1: Table S1. Depending on the tive species to the latter genus. The circumscription of availability of material, ingroup taxa were selected to represent Cephalozia (Dumort.) Dumort. proved to be even more diffi- the morphological variation and geographical distribution of cult. Vilnet et al. (2012) and Feldberg et al. (2013)demon- Cephaloziaceae. Two species of Adelanthaceae were chosen as strated that representatives of the genera Metahygrobiella outgroup based on Forrest et al. (2006) and Feldberg et al. R.M.Schust., Nowellia Mitt., Pleurocladula Grolle, and (2013). Species resolved as non-monophyletic in the presented Schofieldia J.D.Godfrey nest in Cephalozia. To keep the ge- molecular phylogenies were revised. nus Nowellia in a framework of monophyletic genera, Vilnet et al. (2012)restrictedCephalozia to include only those taxa DNA extraction, PCR amplification, and sequencing that were earlier treated as Cephalozia sect. Cephalozia (Schuster 1974). These authors transferred other Cephalozia tissue was isolated from herbarium collections housed at elements to Pleurocladula, thereby overlooking that the herbaria E, GOET, JE, and M. Total genomic DNA was Schofieldia hadbeenestablishedalready2yearsearlier. purified using Invisorb Spin Plant Mini Kit (Invitek, Berlin, Váňaetal.(2013) pointed out this discrepancy and proposed Germany) prior to amplification. Protocols for polymerase to extend the small Neotropical genus Fuscocephaloziopsis chain reaction (PCR) were carried out as described in previous FulfordtoincludePleurocladula, Schofieldia, and those publications: nrITS1-5.8S-ITS2 region from Feldberg et al. Cephalozia species outside subg. Cephalozia.Thistreatment (2004); rbcL gene from Hentschel et al. (2006a), trnL-trnF was based on the Cephaloziineae chronograms presented by region from Feldberg and Heinrichs (2006). Since published Feldberg et al. (2013) who resolved the generitype of protocols and primer combinations did not allow amplifica- Fuscocephaloziopsis, Fuscocephaloziopsis pulvinata tion of all accessions, we modified the protocols and designed (Steph.) Fulford, in a clade with Pleurocladula albescens new primers for rbcL and the nrITS1-5.8S-ITS2 region (Hook.) Grolle and Cephalozia crassifolia (Lindenb. & (Table 1). Gottsche) Fulford. Metahygrobiella was likewise transferred PCR was performed in a total volume of 50 μlcontaining to Cephalozia because two of its species nested in a clade with 1 μmtemplateDNA,38.7μl double-distilled water, 1 μlfor- the generitype, Cephalozia bicuspidata (L.) Dumort. ward and 1 μl reverse primer (Eurofins Genomics, Ebersberg, Potemkin and Sofronova (2013) pointed to the limited taxon Germany), 10 μl MyTaq-reaction-buffer, and 0.3 μl MyTaq- sampling of the available phylogenies and the weak morpho- polymerase (Bioline GmbH, Luckenwalde, Germany). The logical support for the new classification. They proposed to PCR program was carried out as follows: initial denaturation A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 729

Table 1 Primers used in the present study. New primers designed for Cephalozioideae sequencing in bold

Method Primer name Primer sequence Publication

ITS1-5.8S-ITS2 Nested PCR 1. PCR Hep2-F 5′-GAG TCA TCA GCT CGC GTT GAC - 3′ Groth et al. (2003) HepC-R 5′-TCT CCA GAC TAC AAT TCG CAC A-3′ 2. PCR and Hep3-F 5′-CGG TTC GCC GCC GGT GAC G-3′ sequencing HepA-R publ. 5′-CGC CGC TAC TAC GGA AAT CCT A-3′ PCR and sequencing Hep4-F 5′-CGT TGT GAG AAG TTC ATT AAA CC -3′ New HepD-R 5′-CCG CYT AGT GAT ATG CTT AAA CTC-3′ Additional primers for Bryo5.8S-F 5′-GAC TCT CAG CAA CGG ATA-3′ Hartmann et al. (2006) sequencing (ITS5.8SA-F) ITS5.8S-R 5′-GAC GCT CAG GCA GGC AT-3′ Feldberg et al. (2004) rbcL Nested PCR 1. PCR and rbcL-1Pl-F (Pl1por-F) 5′-ATG TCA CCA CAA ACA GAA ACT AAA Wilson et al. (2004) sequencing GCA AGT-3′ M1390-R 5′-CTT TCC AWAYTT CRC AAG CAG CRG-3′ 2. PCR and rbcL210-F 5′-TGG ACT ACG GTT TGG ACT GA-3′ Gradstein et al. (2006) sequencing rbcL1200-R 5′-TGY CCY AAA GTT CCA CCA CC-3′ Wilson et al. (2004) Additional primers for rbcL680-F 5′-GCY GAA ACT GGT GAA ATTAAA G-3′ sequencing rbcL700-R 5′-GTC CTT TAATTT CAC CAG TTT C-3′ Nested PCR 1. PCR and rbcL-1Pl-F (Pl1por-F) 5′-ATG TCA CCA CAA ACA GAA ACT AAA Wilson et al. (2004) sequencing GCA AGT-3′ rbcL1340-R 5′-CAC GAA TAA CCT CAT TAC CTT CAC New GAG C-3′ 2. PCR and rbcL110-Ceph-F 5′-GGC AGC ATT TCG TAT GAC TCC-3′ New sequencing rbcL1227-Ceph-R 5′-TGR CCT AAR GTT CCA CCA CCA New AAT-3′ Additional primers for rbcL627-Ceph-F 5′-ACC RTT TAT GCG TTG GAG AGA-3′ New sequencing rbcL894-Ceph-R 5′-TGC ATY GCA CGG TGA AT-3′ New trnL-trnF PCR and sequencing trnLlej-F 5′-GGT AGA CGC TAC GGA CTT AA-3′ Feldberg and Heinrichs (2006) trnF-R (Taberlet 5′-ATT TGA ACT GGT GAC ACG AG-3′ Taberlet et al. (1991) f-R A50272) at 95 °C for 2 min, followed by 30 cycles of denaturation at Bidirectional sequences were generated by an ABI 3730 48 95 °C for 1 min, annealing at 52 °C (51–53 °C) for 0:50 min, capillary sequencing machine using the BigDye Terminator elongation at 72 °C for 1:30 min. Final elongation was carried v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster out in one step of 10 min at 72 °C. For rbcL, the primers rbcL- City, CA, USA). Newly generated sequences were assembled 1Pl-F and M1390-R or rbcL1340-R were used for the first and edited with CodonCode Aligner v5.0.2 software PCR. If no product was detected, a second (nested) PCR (CodonCode Corporation; http://www.codoncode.com/ was carried out with the primer pairs rbcL210-F and aligner/download.htm). rbcL1200-R or rbcL110-Ceph-F and rbcL1227-Ceph-R. Sequencing primers were the same as employed for PCR Phylogenetic analyses and an additional inner primer pair was also used. If a product was obtained in the first PCR step, rbcL210-F and rbcL1200- Two hundred and forty-one newly generated sequences and RorrbcL110-Ceph-F and rbcL1227-Ceph-R were used. If a 202 published sequences (He-Nygrén et al. 2004; Vilnet et al. nested PCR was necessary, rbcL680F and rbcL700-R or 2012;Feldbergetal.2010, 2013) were aligned manually in rbcL627-Ceph-F and rbcL894-Ceph-R were used. For Bioedit version 7.0.5.2 (Hall 1999). Positions including gaps nrITS, the primers Hep2-F and HepC-R were used for the first were excluded from all alignments and lacking data were cod- PCR. If no product was detected, the primers Hep3-F and ed as missing. Maximum parsimony (MP) analyses were car- HepA-R were used for a nested PCR. Sequencing primers ried out with PAUP* version 4.0b10 (Swofford 2000). MP were Hep3-F and HepA-R in addition to Bryo5.8S-F and heuristic searches were conducted with the following options ITS5.8S-R. Alternatively, the primers Hep4-F and HepD-R implemented: heuristic search mode, 100 random-addition- were used. PCR products were purified with the sequence replicates, tree bisection-reconnection (TBR) branch InvisorbPCRapace purification kit (Invitek, Berlin, Germany). swapping, MULTrees option on, and collapse zero-length 730 K. Feldberg et al. branches off. All characters were treated as equally weighted F, 303, nrITS1-5.8S-ITS2, 581. Of a total of 2012 character and unordered. Non-parametric bootstrapping values sites, 1295 were constant, 132 variable but parsimony unin- (Felsenstein 1985) were generated as heuristic searches with formative, and 585 parsimony informative (Table 2). 500 replicates, each with ten random-addition replicates. The Maximum parsimony analyses of the combined dataset number of rearrangements was restricted to ten millions per yielded >300.000 maximally parsimonious trees of 2363 replicate. Bootstrap percentage values (BV) ≥70 were steps, consistency index (CI) 0.42, and retention index (RI) regarded as good support (Hillis and Bull 1993). Where more 0.89. Since the strict consensus of these trees largely resem- than one most parsimonious tree was found, trees were sum- bled the optimal ML phylogeny and the tree based on marized in a strict consensus tree. Bayesian inference, only the latter is depicted (Figs. 1, 2, 3, The three genomic regions were first analyzed separately to and 4). MP and ML bootstrap percentage values are displayed check for incongruence. The strict consensus trees of the non- on the Bayes-topology in addition to Bayesian posterior prob- parametric bootstrap analyses were compared by eye to iden- abilities (order BPP/ML-BV/MP-BV). tify conflicting nodes supported by at least 70 % (Mason- Alobielloideae are represented by a clade including Alobiella Gamer and Kellogg 1996). The trees gave no evidence of husnotii (Spruce) Schiffn. and three accessions of Alobiellopsis incongruence. Hence the datasets were combined. R.M.Schust. (1.00/100/97); they are placed sister to the remain- Maximum likelihood (ML) inference was carried out using der of Cephaloziaceae (1/95/70). The Odontoschismatoideae RAxML 8.1.2 (Stamatakis 2014; http://sco.h-its.org/exelixis/ (1.00/100/100) split into two main lineages. One includes ac- web/software/raxml/#documentation)asimplementedin cessions of Odontoschisma jishibae (Steph.) L. Söderstr. & raxmlGUI 1.5b1 (Silvestro and Michalak 2012; http:// Váňa, Odontoschisma fluitans (Nees) L. Söderstr. & Váňa, sourceforge.net/projects/raxmlgui/). JModelTest 2.1.7 Odontoschisma pseudogrosseverrucosum Gradst., S.C.Aranda (Guindon and Gascuel 2003; Darriba et al. 2012)was & Vanderp., and Odontoschisma sphagni (Dicks.) Dumort. employed to choose a nucleotide substitution model for the (0.65/69/55). The other comprises Odontoschisma denudatum combined dataset. The Akaike information criterion (AIC) as (Mart.) Dumort., Odontoschisma elongatum (Lindb.) A.Evans, well as the Bayesian information criterion (BIC) supported the Odontoschisma engelii Gradst. & Burghardt, Odontoschisma TIM3 + Γ + I model. This model is not available in RAxML; francisci (Hook.) L.Söderstr. & Váňa, Odontoschisma hence, the best-fitting overparametrized available model, grosseverrucosum Steph., Odontoschisma longiflorum GTR + Γ (Rodríguez et al. 1990), was employed. This ap- (Taylor) Trevis., Odontoschisma variabile (Lindenb. & proach follows the suggestion given by Posada (2008). Gottsche) Trevis., and a so far undescribed entity, Analyses were run using the thorough-bootstrap-option, 100 Odontoschisma yunnanense, sp. nov. (see Taxonomic treat- replicate runs, bootstrap replicates estimated with the ment) (1.00/74/72). Neotropical accessions of Odontoschisma autoMRE option and saving branch lengths. are placed in a derived clade. Schiffnerioideae are placed sister Bayesian inference was undertaken with MrBayes 3.2.6 to Cephalozioideae (1.00/-/-). The Cephalozioideae (1.00/68/ (Ronquist and Huelsenbeck 2003; http://mrbayes. 58) split in two main lineages. One includes Nowellia in a sister sourceforge.net/index.php) on the CIPRES Science Gateway relationship to Cephalozia (1.00/96/98). The other includes ac- (Miller et al. 2010; https://www.phylo.org/portal2/login!input. cessions of Fuscocephaloziopsis (0.90/60/62). Two accessions action). A single partition with the GTR + Γ + I substitution of Nowellia wrightii (Gottsche ex Spruce) Steph. ex Duss are model was used. The analyses were set up with two parallel placed sister to a clade with nine accessions of Nowellia runs with four chains, 50,000,000 generations, and sampling curvifolia (Dicks.) Mitt. (1.00/100/100). Two accessions of the every 5,000th tree. Stationarity and convergence of runs and generitype F. pulvinata nest in a robust Fuscocephaloziopsis estimation of burnin were checked using Tracer 1.6 (http:// subclade including also Fuscocephaloziopsis macrostachya tree.bio.ed.ac.uk/software/tracer/). The trees of the run with (Kaal.) Váňa & L.Söderstr., Fuscocephaloziopsis catenulata the best likelihood were summarized with TreeAnnotator 1. (Huebener) Váňa&L.Söderstr.,Fuscocephaloziopsis gollanii 8.0 (BEAST package; http://beast.bio.ed.ac.uk/)and visualized using FigTree 1.4.2 (http://tree/bio.ed.ac.uk/ software/figtree). Bayesian posterior probability (BPP) confi- Table 2 Distribution of constant and phylogenetically informative sites dence values were regarded as significant when BPP ≥0.95 for aligned positions of the three genomic regions (Larget and Simon 1999). rbcL trnL-F ITS1-5.8S-ITS2 Total

Number of sites in matrix 1128 303 581 2012 Results Constant 798 179 318 1295 Variable but parsimony 41 18 41 132 Alignment of the three genomic regions resulted in a com- uninformative Parsimony informative 257 106 222 585 bined dataset with the following: rbcL, 1128 positions, trnL- A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 731

Fig. 1 Condensed majority rule consensus tree recovered in stationary„ phase of Bayesian search. Dotted branches have Bayesian Posterior Probabilities <0.95. Pruned clades (Figs. 2, 3,and4) are indicated

(Steph.) Váňa & L.Söderstr., Fuscocephaloziopsis loitlesbergeri (Schiffn.) Váňa & L.Söderstr., Fuscocephaloziopsis crassifolia (Lindenb. & Gottsche) Váňa & L.Söderstr., and Fuscocephaloziopsis connivens (Dicks.) Váňa & L.Söderstr. (1.00/99/95). Another robust subclade includes Fuscocephaloziopsis monticola (J.D.Godfrey) Váňa& L.Söderstr., Fuscocephaloziopsis leucantha (Spruce) Váňa& L.Söderstr., Fuscocephaloziopsis affinis (Steph.) Váňa& L.Söderstr., Fuscocephaloziopsis lunulifolia (Dumort.) Váňa & L.Söderstr., Fuscocephaloziopsis albescens (Hook.) Váňa& L.Söderstr., Fuscocephaloziopsis pachycaulis (R.M.Schust.) Váňa & L.Söderstr., and Fuscocephaloziopsis pleniceps (Austin) Váňa & L.Söderstr. (1.00/82/86). Multiple accessions of most species form monophyletic lineages. Accessions of Cephalozia hamatiloba Steph. are found in three independent lineages. Cephalozia ambigua C.Massal., Cephalozia badia (Gottsche) Steph., Cephalozia crossii Spruce, Cephalozia grandifolia Steph., and Cephalozia drucei (R.M.Schust.) Váňa nest in C. bicuspidata.

Discussion

Supraspecific classification

We were able to include four of the five currently accepted subfamilies (Söderström et al. 2016) in our molecular sam- pling: Alobielloideae, Cephalozioideae, Odontoschismatoideae, and Schiffnerioideae (Fig. 1). The small Neotropical Trabacelluloideae (Haesselia Grolle & Gradst., two species; Trabacellula Fulford, one species) were not available for molecular study. All subfamilies got high statistical support, with Alobielloideae placed sister to the re- mainder of Cephaloziaceae. Alobielloideae is a mainly tropi- cal subfamily. It comprises the monospecific type genus of the subfamily, Alobiella, which is restricted to the Neotropics, and the genus Alobiellopsis, which contains two Neotropical, one Japanese, and two South African species. All species are rath- er small and generally occur as pure patches on exposed min- eral soils and soil-covered rocks or banks (Schuster 2002). Alobiella husnotii is nested in a clade with three accessions of Alobiellopsis but this position lacks support (Fig. 2). A decision on the taxonomic value of Alobiellopsis should thus await an extended sampling and inclusion of the type species Alobiellopsis acroscypha (Spruce) R.M.Schust. The recently proposed inclusion of Anomoclada, Cladopodiella Spruce, and Iwatsukia in Odontoschisma (Váňaetal.2013;Arandaetal.2014; Gradstein et al. 2014) renders Odontoschismatoideae monogeneric. Gradstein and 732 K. Feldberg et al.

Fig. 2 Majority-rule consensus tree of trees recovered in stationary phase of Bayesian search; includes Alobielloideae and Odontoschismatoideae. Bayesian posterior probabilities (bold), MP- (italics), and ML-bootstrap percentage values are given at branches A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 733

Fig. 3 Majority-rule consensus tree of trees recovered in stationary phase of Bayesian search; includes Nowellia and Cephalozia. Bayesian posterior probabilities (bold), MP- (italics), and ML-bootstrap percentage values are given at branches 734 K. Feldberg et al.

Fig. 4 Majority-rule consensus tree of trees recovered in stationary phase of Bayesian search; includes Fuscocephaloziopsis. Bayesian posterior probabilities (bold), MP- (italics), and ML-bootstrap percentage values are given at branches A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 735

Ilkiu-Borges (2015) accepted 21 Odontoschisma species of (Masuzaki et al. 2010). Metzgeriopsis was resolved as an which 13 occur in the Neotropics, 8 in the Holarctic, and 8 element of and treated as a synonym of in the Paleotropics and the pacific region. The species gener- (Spruce) Steph. (Gradstein et al. 2006). In ally grow on organic material like humus, peat, decaying both cases, the morphology of the generative structures wood, and on tree bases from the lowlands to more than supported the re-classification, paralleling the 4000 m NN. Use of the nrITS region in addition to chloroplast morphology-based treatment of Schiffneria. markers enhances the phylogenetic resolution compared to Schiffnerioideae was resolved in a sister relationship with Aranda et al. (2014); however, several basal nodes of Cephalozioideae; however, analyses of nrITS alone placed Odontoschismatoideae do not get posterior probabilities it sister to Fuscocephaloziopsis, albeit without strong sta- ≥0.95 or bootstrap percentage values ≥70 (Fig. 2). One of tistical support. Studies with an extended sampling includ- the most surprising findings of previous molecular research ing further nuclear markers are desirable to test the current on Odontoschismatoideae was the observation that Iwatsukia phylogenetic position. [now Odontoschisma sect. Iwatsukia (N.Kitag.) Gradst., Cephalozioideae is a subcosmopolitan clade with a cen- S.C.Aranda & Vanderp.; Aranda et al. 2014]nestsin ter of diversity in the northern hemisphere and includes Odontoschisma (Vilnet et al. 2012). Iwatsukia differs from rather small, soft-textured that are usually closely Odontoschisma in the traditional sense by its bifid leaves, attached to the substrate. They occur in a wide variety of evenly thickened cell walls, nodular thickenings on all longi- habitats and grow on decaying wood, trunk bases, soil, tudinal cells of the capsule wall, and the presence of peat, or intermingled with Sphagnum spp. Several studies surface wax (Gradstein and Ilkiu-Borges 2015). It has alterna- identified incongruences of morphology-based genus cir- tively been aligned with Cephaloziaceae subfam. cumscriptions and molecular phylogenies (Vilnet et al. Alobielloideae (Schuster 2002) or placed in a family of its 2012; Feldberg et al. 2013). These studies demonstrated own, Cladomastigaceae (Fulford 1968). In our phylogeny, that elements of Fuscocephaloziopsis, Metahygrobiella, O. (sect. Iwatsukia) jishibae is placed in the unsupported main Nowellia, Pleurocladula,andSchofieldia nest in clade of Odontoschisma, sister to a clade with O. fluitans, Cephalozia. To establish monophyletic genera, Váňa O. pseudogrosseverrucosum,andO. sphagni.Arandaetal. et al. (2013) introduced a largely revised classification. (2014)resolvedtwoIwatsukia accessions in a basal polytomy They accepted Nowellia and recircumscribed Cephalozia to with other Odontoschisma elements. The position of include only taxa of C.sect.Cephalozia as well as members of Iwatsukia inside Odontoschisma is thus not yet stable and a Metahygrobiella. The other elements of Cephalozia were sister relationship of Iwatsukia and Odontoschisma cannot yet transferred to an extended genus Fuscocephaloziopsis that be safely rejected. This part of the phylogeny deserves further also includes Pleurocladula and Schofieldia. Potemkin and study based on an extended marker and taxon sampling. The Sofronova (2013) pointed to the weak morphological differ- position of the taxon within Odontoschismatoideae however ences of the newly circumscribed genera Cephalozia and is unambiguous. Fuscocephaloziopsis and proposed to include all Schiffnerioideae was set up to include a single Asian Cephalozioideae taxa in a single genus, Cephalozia s.l. species, Schiffneria hyalina Steph. (Söderström et al. Our phylogeny (Figs. 3 and 4) includes a comprehensive 2016). This species is renowned for its subthallose mor- sampling of Cephalozioideae and exemplars of the type spe- phology that separates it from all other elements of cies of all three relevant genera (C. bicuspidata, F. pulvinata, Cephaloziaceae, yet its fits well within this N. curvifolia). Cephalozioideae are split in two main lineages family (Grolle and Piippo 1984;Schuster2002). It was of which one includes elements of Nowellia and Cephalozia alternatively placed in a familiy Schiffneriaceae; however, sensu Váňaetal.(2013) in a robust sister relationship (Fig. 3). it has been identified as a member of Cephaloziaceae in The other lineage contains elements of Fuscocephaloziopsis several molecular phylogenies (Heinrichs et al. 2005; in the circumscription of Váňaetal.(2013)(Fig.4). In our Forrest et al. 2006 ; He-Nygrén et al. 2006; Vilnet et al. opinion, genera should be recognizable based on morpholog- 2012; Feldberg et al. 2013). Subthallose or thallose liver- ical evidence to be encodable in morphology-based floras. worts are rare within the “leafy liverwort” clade Nowellia can easily be recognized by its concave leaves that Jungermanniidae and have been placed in families of their usually form a ventral water-sac whereas the separation of own. Examples include the Asian genera Furuki Cephalozia and Fuscocephaloziopsis is far less clear-cut. & Z.Iwats. in family Mizutaniaceae of the simple thalloid Fuscocephaloziopsis species usually have a leaf-free zone Metzgeriidae (Furuki and Iwatsuki 1989)and on the dorsal side of the stem whereas leaf bases of Metzgeriopsis K.I.Goebel in Metzgeriopsidaceae (Engel Cephalozia reach the dorsal stem midline. Further differences 1982). Molecular phylogenies identified Mizutania as an are seen in the stem anatomy. Stem cell walls of Cephalozia element of Jungermanniidae rather than Metzgeriidae and species are equally thickened in cross section while the cell were suggestive of an affiliation with wall thickness increases towards the inner portions of the stem 736 K. Feldberg et al. in most members of Fuscocephaloziopsis. In general, Holarctic and the Neotropics (Fig. 3). The short branches in Fuscocephaloziopsis stems appear to be more robust than many species crown groups provide evidence for a low mo- those of Cephalozia and produce only sometimes lateral- lecular variation and could result from rather recent range terminal branches (frequently present in Cephalozia). expansions, possibly in combination with bottleneck events Considering the presented evidence, it is in general possible triggered by the Pleistocene glaciations (Hewitt 1996, 2000). to separate Cephalozia and Fuscocephaloziopsis based on Range contractions and expansions are not uncommon in tem- morphology. For the time being, we therefore maintain the perate and polar species and have been suggested for concept of Váňa et al. (2013) and accept the three genera tamarisci (L.) Dumort. (Heinrichs et al. 2010), Nowellia, Cephalozia, and Fuscocephaloziopsis yet admit (Dumort.) Dumort. species (Heinrichs et al. that the concept of Potemkin and Sofronova (2013)maybe 2012a), and Holarctic accessions of pubescens better suited to identify Cephalozioideae species in identifica- (Schrank) Raddi (Fuselier et al. 2011). tion books and field guides. Fuscocephaloziopsis splits in two Most currently accepted species concepts such as phyloge- robust main clades. Clade I (Fig. 1) includes species without netic species concepts and the unified species concept (de terminal branching, very obliquely to horizontally inserted Queiroz 2007) consider reciprocal monophyly as the explicit leaves with isodiametric cells, deeply divided female bracts criterion of species. Incongruence of morphological taxa with and a strongly ciliate perianth mouth. Clade II includes species DNA sequence-based phylogenetic results is considered as with rare lateral-terminal (“Frullania-type”) branching in ad- evidence for limitations of species concepts relying exclusive- dition to ventral-intercalary branching and a perianth mouth ly on the diagnostics of morphological features (Hutsemekers that is usually denticulate rather than strongly ciliate. The leaf et al. 2012). However, molecular phylogenies allow us to re- insertion is subject to variation. consider the morphology of non-monophyletic species and to establish new circumscriptions for the different lineages, e.g., Species circumscriptions and ranges for independent lineages of the Neotropical liverwort Lepidolejeunea involuta (Gottsche) Grolle (Heinrichs et al. Molecular studies of morphologically circumscribed species 2015) and the Australasian buccinifera (Hook.f. & incorporating variable molecular markers and a comprehen- Taylor) Gottsche, Lindenb. & Nees (Renner et al. 2013b). In sive sampling do not necessarily question the taxon’smono- other cases, morphology is insufficient to identify lineages phyly but often point to an internal structure with geographi- that are well separated in molecular phylogenies, e.g., in the cally separated subclades. Examples include the leafy liver- liverwort Nees (Kreier et al. 2010) and the worts cordaeana (Huebener) Moore and Porella Sphagnum subsecundum Nees s.l. (Shaw et al. 2008)and platyphylla (L.) Pfeiff. with separate European and North Rhynchostegium riparioides (Hedw.) Cardot (Hutsemekers American lineages (Heinrichs et al. 2011), Radula demissa et al. 2012). M.A.M.Renner and Radula tasmanica Steph. with separate Several such examples are present in Cephalozia s.str. New Zealand and Tasmanian lineages (Renner et al. 2013a), C. hamatiloba is paraphyletic and includes Cephalozia and the simple thalloid (L.) Corda with a (Metahygrobiella) albula Steph., Cephalozia cf. neesiana, North American lineage and a European lineage, the latter, Cephalozia conchata (Grolle & Váňa) Váňa, and the however, including a few North American accessions C. bicuspidata complex. It exhibits a considerable morpho- (Fuselier et al. 2009). Such topologies indicate genetic differ- logical variation but current evidence is insufficient for a re- ences related to geographical isolation over longer periods of vised classification. A possible solution is the recognition of time and are suggestive of cryptic or near cryptic speciation several independent entities yet morphological boundaries (Shaw 2001; Heinrichs et al. 2009). Similar examples hardly still need to be established. Revisionary studies also need to occur in our Cephaloziaceae dataset yet an accession of consider the presence of different chromosome numbers in F. crassifolia from the Azores is placed sister to two C. hamatiloba s.l. (Váňa 1988). Similarly problematic is the Neotropical accessions and an accession of F. connivens from circumscription of the generitype C. bicuspidata. Extension of São Tomé sister to several Holarctic accessions (Fig. 4). the sampling of Vilnet et al. (2012) demonstrates that acces- Several monophyletic species occur on two or more conti- sions of C. ambigua, C. badia, C. grandifolia, C. crossii,and nents, in some contrast to a general trend towards the accep- C. drucei nest in the robust C. bicuspidata clade. Váňaand tance of narrower distribution ranges of species Long (2011) consider species concepts within Cephalozia (Hedenäs 2008; Ramaiya et al. 2010;Renneretal.2013b). (s.l.) “in a state of considerable uncertainty” and identify the Examples include O. sphagni with European, Macaronesian, C. bicuspidata complex as one of the most problematic North American, and Neotropical accessions (Fig. 2), groups. This view is clearly confirmed in our study. F. connivens with Holarctic and African accessions, According to Váňa and Long (2011) “all species of the com- F. catenulata with Holarctic and West Indian accessions plex exhibit phenotypic plasticity under different ecological (Fig. 4), and N. curvifolia with representatives from the conditions”; however, this hypothesis has not yet been tested A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 737 in a phylogenetic context. Problems in identifying biological Gradstein and Ilkiu-Borges 2015). Molecular investigations of species within the C. bicuspidata complex may also be related tropical liverworts tend to identify a larger number of unrec- to the occurrence of populations with different chromosome ognized entities (Dong et al. 2012;Renneretal.2013b), pro- numbers (n = 9, 18, 27). Forms with n = 9 have been recog- viding some evidence that the liverwort diversity within cold nized as C. ambigua; C. bicuspidata s.str. is considered to and temperate regions is better known than that of the tropics. have n =18and Cephalozia lammersiana (Huebener) Spruce This hypothesis, however, needs to be tested by additional n = 27; the latter two taxa however have no differentiating species-level studies with extensive population sampling. morphological characters (Paton 1999). Chromosome num- bers of the investigated DNA vouchers are unknown; hence, Origin of tropical Cephaloziaceae diversity we feel unable to identify the clade with C. bicuspidata s.str. This species was described by Linnaeus (1753: 1132) to occur Cephaloziaceae represents a subcosmopolitan lineage with “in Europae umbrosis humidis” yet eastern Holarctic acces- numerous species occurring in tropical regions (Gradstein sions are present in several crown group clades of et al. 2001; Gradstein and Ilkiu-Borges 2015). The tropics C. bicuspidata s.l. Cephalozia badia is considered to represent have been considered as a museum of diversity but also the a (sub-)antarctic species that differs from C. bicuspidata by its cradle of temperate lineages, especially for genera with a cen- relatively poorly developed hyalodermis, strong brownish- ter of diversity in the tropics (Jablonski et al. 2006;Marshall purplish pigmentation, and absence of gemmae; its sex distri- 2006; McKenna and Farell 2006). This hypothesis is in accor- bution is unclear (Bednarek-Ochyra et al. 2000;Schuster dance with a global phylogeny of the liverwort genus 2002). Six accessions of C. badia form a monophyletic line- whose temperate lineages nest in tropical clades age within C. bicuspidata s.l. We included four accessions (Heinrichs et al. 2013). Subcosmopolitan liverworts that occur from the type locality South Georgia and thus consider our in tropical highlands rather than lowlands may show different material correctly identified. C. crossii and C. grandifolia are patterns including colonization of the tropics from temperate two Neotropical taxa that nest in the C. bicuspidata complex. regions. A good example is the leafy liverwort Scapania The autoicous C. crossii has been separated from whose tropical lineages nest in temperate clades (Heinrichs C. bicuspidata by its whitish green color, long ciliate perianth et al. 2012a). According to our present sampling, mouth, and narrow leaf lobes (Schuster 2002). C. grandifolia Cephaloziaceae fits the latter pattern, being in accordance with has been considered as dioicous (Fulford 1968) but neverthe- its tropical center of diversity at the higher altitudes. A less treated as a synonym of C. bicuspidata by Váňa(1988). Neotropical clade of Odontoschisma (O. engelii, The two New Zealand accessions of C. drucei have earlier O. longifolium, O. variabile) is resolved in a derived position been considered to belong to a different genus, and seems to originate from Holarctic ancestors. Neotropical Metahygrobiella (now considered a synonym of Cephalozia, accessions of N. curvifolia nest in a Holarctic clade as do Váňaetal.2013). Lumping all these taxa into a morphologi- Neotropical representatives of the C. bicuspidata complex. cally variable C. bicuspidata would allow identification of the taxon based on morphological evidence yet would hardly ex- Perspectives plain the evolutionary processes that led to the current diver- sity. A key to a deeper understanding of the complex lies in the Different chromosome numbers are known from several tax- study of accessions of which the chromosome numbers are onomically difficult jungermannialean liverworts including known, and involvement of variable marker systems including Raddi (Buczkowska et al. 2015)and a larger set of loci. Such studies are available for several liv- Corda (Hentschel et al. 2006b). Studies involv- erworts including the Reboulia hemisphaerica (L.) Raddi ing chromosome counts and variable marker systems complex (Boisselier-Dubayle et al. 1998) and Raddi (Buczkowska et al. 2004) may shed light on the evolutionary (Odrzykoski et al. 1996). Another promising candidate for history of such species complexes and allow consideration of such a study is F. pleniceps. This species is split into two phenomena such as polyploidy as an explanation for the ob- robust lineages with overlapping geographical ranges. served diversity (Buczkowska et al. 2012). Our investigation With the exception of the above examples, our phylogenet- identifies the C. bicuspidata complex as a promising candi- ic hypothesis is in good accordance with current date for such studies and tags several taxa that have not yet morphological-typological species concepts as well as current been considered as members of this group. Micro-evolution estimates of species diversity (Söderström et al. 2016). Our will possibly become a prime topic of future bryology, espe- molecular investigation of Cephaloziaceae led to the recogni- cially when comprehensive global species-level phylogenies tion of only a single so far undescribed entity, O. yunnanense, are available. sp.nov.(seeFig.4 and Taxonomic treatment), clearly demon- Although we were able to extend the taxon sampling of strating that the taxonomy of Odontoschisma has already been Cephaloziaceae to 41 of the 89 currently accepted species improved by several integrative studies (Aranda et al. 2014; (Söderström et al. 2016), we must admit that several critical 738 K. Feldberg et al. species still need to be added to the datasets, e.g., Haesselia Taxonomic treatment roraimensis Grolle & Gradst. and Trabacellula tumidula Fulford. Extension of the sampling will permit further Odontoschisma yunnanense K.Feldberg, Váňa, D.G.Long balancing of the different generic concepts of & Heinrichs spec. nov. (Fig. 5). Type: China. Yunnan Cephaloziineae and better understanding of morphological Province. Fugong County, Lumadeng Xiang, Yaping evolution and biogeographical patterns within this lineage. Cun, E slope of Gaoligong Shan (Nu Jiang catchment), Our study identifies the genera Alobiella and Alobiellopsis Burma/Yunnan border ridge, above lake south of “Yaping as a target for follow-up studies and points to some uncer- Pass”, 27° 12′ 17.2″N, 98° 41′ 48.6″E. Steep alpine slope tainties with the systematic position of the enigmatic genus with block scree, low cliffs and dwarf Rhododendrons; in Schiffneria. Extension of the nuclear marker sets will allow to humid mossy crevices of granite outcrop. Alt. c. 3660 m, explore if some nuclear and chloroplast genomes of 12.08.2005, D. G. Long 34657 (holotype, E 00250469; Cephalociaceae include conflicting phylogenetic signals. isotype M).

Fig. 5 O. yunnanense,sp.nov.A Shoot in dorsal view. B Shoot in ventral view. C Leaves. D Underleaves. E Marginal leaf cells. F Cross section of stem [all from holotype] A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 739

Plants up to ca. 20 mm long, 1.5 to 1.7 mm wide with laterally compressed branch leaves, lack of crenulate leaf mar- leaves, prostrate to ascending from a rhizomatous base of gins, and marginal leaf cells without considerably thickened geotropic stolons and old axes; green to yellowish green, with- outer walls. Gemmiparous shoots (as present in O. zhui)have out reddish or purplish secondary pigmentation, older shoots not been observed. becoming brownish with age. Shoot apices sometimes taper- ing into geotropic stolons with subsequently smaller leaves Ecology that give way to numerous rhizoids; ventral- (or ventro- lateral) intercalary branches often present, in part without bas- The type of O. yunnanense was collected on the “Biotic al leaves, others with ascending sectors with small leaves in Survey of Gaolingshan” survey close to the border of the lower part and well-developed leaves in the upper part, the Yunnan with Myanmar. The area was a steep mountain slope latter branches often producing a stolon near base. Stem ca. with scree and rock outcrops, dominated by dwarf 0.25–0.3 mm broad, soft-textured, green to brownish; epider- Rhododendron shrubberies and dense, liverwort-rich mats. mis cells mostly rectangular in surface view, just below a leaf Associated species included the lepidozioides insertion elongated, ca. 2–2.5 times as long as wide; in cross S.Hatt. & Inoue, and the liverworts alpinum section with two rows of small thick-walled cortical cells, Steph., Anastrophyllum joergensenii Schiffn., subsequent cells less thick-walled and gradually merging into pearsonii Steph., crispula Mitt., Gymnomitrion larger, thin-walled medula cells. Rhizoids numerous, scattered revolutum (Nees) H.Philib., Gymnomitrion rubidum (Mitt.) along stem, often present up to shoot apex; on main shoots Váňa, Crand.-Stotl. & Stotler, Gymnomitrion verrucosum only ventral, on stolons on all sides of the stem. Leaves ca. W.E. Nicholson, Plicanthus hirtellus (F.Weber) R.M. 0.4–1.0 mm broad, ca. 0.6–1.0 mm long, succoubous and Schust., Solenostoma subacutum (Herzog) Váňa, Crand.- imbricate on main axes, distant on branches, obliquely Stotl. & Stotler, and quinquedentata (Huds.) inserted, not decurrent; insertion line extending to stem mid- H.Buch. line dorsally (two to three cells overlap); leaves of well- developed shoots laterally spreading, on branches often later- Acknowledgments We thank the directors and curators of the ally compressed, often distally directed near the shoot apex; Göttingen University Herbarium (GOET), the Herbarium Haussknecht “ ” (JE), the Bavarian State collection for Botany (M), and the Herbarium strongly concave ( bowl-shaped ), short ovate (ca. 0.8 to 1.2 of the Royal Botanic Garden Edinburgh (E) for the loan of specimens and times as long as broad), dorsally orientated, somewhat asym- the permission for destructive sampling, and the staff members of the metric with ventral margin more strongly curved than dorsal LMU Genomics Service unit for support. Parts of this study were sup- margin; branch leaves smaller, flatter and more elongated ported by the German Research Foundation (DFG grant HE 3854 / 4). (1.2–1.3 times as long as broad); well-developed leaves often with one to two small teeth (1/2–1superposedcellterminated by a slime papilla) at ventral and less frequently dorsal margin References near leaf insertion point. Often one to four additional slime papillae borne near the ventral leaf insertion, dorsally with Aranda, S. C., Gradstein, S. R., Patiño, J., Laenen, B., Désamoré, A., & – Vanderpoorten, A. (2014). Phylogeny, classification and species de- zero to two slime papillae. Cells near leaf insertion ca. 17 limitation in the liverwort genus Odontoschisma (Cephaloziaceae). 22 μm broad, ca. 25–44 μm long, elongate (ca. 1.5 to 2.0 Taxon, 63,1008–1025. times as long as broad), soon becoming +/− isodiametric and Bakalin, V. A., & Vilnet, A. A. (2014). Two new species of the liverwort ca. 17–20 μm long, 16–20 μm broad, with distinct genus Hygrobiella Spruce () described from the North Pacific based on integrative taxonomy. Plant Systematics subnodulose, sometimes confluent trigones; cells of the leaf and Evolution, 300,2277–2291. margin not eminently thickened, not forming a distinct border; Bechteler, J., Lee, G. E., Schäfer-Verwimp, A., Pócs, T., Peralta, D. F., cell surface smooth. Underleaves small and irregularly Renner, M. A. M., et al. (2016). Towards a monophyletic classifica- shaped, ca. 0.1–0.25 mm long, ca. 0.07–0.1 mm broad; with tion of Lejeuneaceae IV: reinstatement of Allorgella, transfer of Microlejeunea aphanella to Vitalianthus and refinements of the slime-papillae, sometimes shallowly bifurcated with short subtribal classification. Plant Systematics and Evolution, 302, blunt lobes or acute lobes bearing a slime papilla; margins 187–201. sometimes with teeth of one to two superposed cells bearing Bednarek-Ochyra, H., Váňa, J., Ochyra, R., & Lewis Smith, R. I. (2000). slime papillae; cells of the underleaves without trigones, walls The liverwort flora of Antarctica. Cracow: Polish Academy of Sciences. evenly and moderately thickened. Oil-bodies, gemmae, and Boisselier-Dubayle, M. E., Lambourdiere, J., & Bischler, H. (1998). Taxa generative structures not observed. delimitation in Reboulia investigated with morphological, cytologi- cal, and isozyme markers. The Bryologist, 101,61–69. Differentiation Buczkowska, K., Odrzykoski, I. J., & Chudzińska, E. (2004). Delimitation of some European species of Calypogeia Raddi (Hepaticae, Jungermanniales) based on cytological characters O. yunannense is morphologically similar to Odontoschisma of oil bodies and multienzyme phenotype. Nova Hedwigia, 78, zhui Gradst., S.C. Aranda & Vanderp. yet differs by its often 147–163. 740 K. Feldberg et al.

Buczkowska, K., Sawicki, J., Szczecińska, M., Klama, H., & Forrest, L. L., Davis, E. C., Long, D. G., & Crandall-Stotler, B. J. (2006). Baczkiewicz, A. (2012). Allopolyploid speciation of Calypogeia Unraveling the evolutionary history of the liverworts sphagnicola (Jungermanniopsida, Calypogeiaceae) based on iso- (Marchantiophyta): multiple taxa, genomes and analyses. The zyme and DNA markers. Plant Systematics and Evolution, 298, Bryologist, 109,303–334. 549–560. Fulford, M. H. (1968). Manual of the leafy Hepaticae—part III. Memoirs Buczkowska, K., Hornik, B., & Czolpińska, M. (2015). Two ploidy levels of the New York Botanical Garden, 11,277–392. of genetically delimited groups of the Calypogeia fissa complex Furuki, T., & Iwatsuki, Z. (1989). Mizutania riccardioides, gen. et spec. (Jungermanniopsida, Calypogeiaceae). Biodiversity Research and nov. (Mizutaniaceae, fam. nov.): a unique liverwort from tropical Conservation, 39,1–6. . Journal of the Hattori Botanical Laboratory, 67,291–296. Crandall-Stotler, B., Stotler, R. E., & Long, D. G. (2009). Phylogeny and Fuselier, L., Davison, P. G., Clements, M., Shaw, B., Devos, N., classification of the Marchantiophyta. Edinburgh Journal of Botany, Heinrichs, et al. (2009). Phylogeographic analyses reveal distinct 66,155–198. lineages of Metzgeria furcata and M. conjugata () in Damsholt, K. (2002). Illustrated flora of Nordic liverworts and and . Biological Journal of the Linnean . Lund: Nordic Bryological Society. Society, 98,745–756. Darriba, D., Taboada, G. L., Doallo, R., & Posada, D. (2012). JModelTest Fuselier, L., Shaw, B., Engel, J. J., von Konrat, M., Costa, D. P., Devos, 2: more models, new heuristics and parallel computing. Nature N., et al. (2011). The status and phylogeography of the liverwort Methods, 9,772. genus Apometzgeria Kuwah. (Metzgeriaceae). The Bryologist, 114, – Dayrat, B. (2005). Towards integrative taxonomy. Biological Journal of 92 101. the Linnean Society, 85,407–415. Gradstein, S. R., & Ilkiu-Borges, A. L. (2015). A taxonomic revision of De Queiroz, K. (2007). Species concepts and species delimitation. the genus Odontoschisma (Marchantiophyta: Cephaloziaceae). – Systematic Biology, 56,879–886. Nova Hedwigia, 100,15 100. Dong, S., Schäfer-Verwimp, A., Meinecke, P., Feldberg, K., Bombosch, Gradstein, S. R., Churchill, S. P., & Salazar-Allen, N. (2001). Guide to the of tropical America. Memoirs of the New York Botanical A., Pócs, T., et al. (2012). Tramps, narrow endemics and morpho- – logically cryptic species in the epiphyllous liverwort Garden, 86,1 577. Diplasiolejeunea. Molecular Phylogenetics and Evolution, 65, Gradstein, S. R., Wilson, R., Ilkiu-Borges, A. L., & Heinrichs, J. (2006). 582–594. Phylogenetic relationships and neotenic evolution of Metzgeriopsis (Lejeuneaceae) based on chloroplast DNA sequences and morphol- Engel, J. J. (1982). Hepaticopsida. In S. P. Parker (Ed.), Synopsis and ogy. Botanical Journal of the Linnean Society, 151,293–308. classification of living organisms, Vol. 1 (pp. 271–304). New York: Gradstein, S. R., Aranda, S. C., & Vanderpoorten, A. (2014). Notes on McGraw-Hill. early land plants today. 47. Transfer of Iwatsukia to Odontoschisma Engel, J. J., & Glenny, D. (2008). A flora of the liverworts and hornworts (Cephaloziaceae, Marchantiophyta). Phytotaxa, 162,232–233. of New Zealand. Vol. 1. Monographs in Systematic Botany from the Grolle, R., & Piippo, S. (1984). Bryophyte flora of the Huon Peninsula, Missouri Botanical Garden, 110,1–897. Papua New Guinea. V. subfam. Zoopsoideae and Engel, J. J., He, X., & Glenny, D. (2010). Studies on Cephaloziaceae subfam. Schiffnerioideae (Hepaticae). Annales XXII. The systematic position of Amphilophocolea R.M. Schust. Botanici Fennici, 21,299–307. together with comments on the status of Tetracymbaliella Grolle – Groth, H., Lindner, M., Wilson, R., Hartmann, F. A., Schmull, M., and Lamellocolea R.M. Schust. Phytotaxa, 9,41 52. Gradstein, S. R., et al. (2003). Biogeography of Feldberg, K., & Heinrichs, J. (2006). A taxonomic revision of (Hepaticae): natural species groups span several floristic kingdoms. (Jungermanniidae: ) in the Neotropics based on nuclear Journal of Biogeography, 30,965–978. and chloroplast DNA and morphology. Botanical Journal of the Guindon, S., & Gascuel, O. (2003). A simple, fast and accurate method to – Linnean Society, 151,309 332. estimate large phylogenies by maximum likelihood. Systematic Feldberg, K., Groth, H., Wilson, R., Schäfer-Verwimp, A., & Heinrichs, J. Biology, 52,696–704. (2004). Cryptic speciation in Herbertus (Herbertaceae, Hall, T. A. (1999). BioEdit: a user-friendly biological sequence alignment Jungermanniopsida): range and morphology of Herbertus sendtneri editor and analysis program for Windows 95/98/NT. Nucleic Acid inferred from nrITS sequences. Plant Systematics and Evolution, Symposia Series, 41,95–98. – 249,247 261. Hartmann, F. A., Wilson, R., Gradstein, S. R., Schneider, H., & Heinrichs, Feldberg, K., Hentschel, J., Bombosch, A., Long, D. G., Váňa, J., & J. (2006). Testing hypotheses on species delimitations and disjunc- Heinrichs, J. (2009). Transfer of grollei, G. patoniae tions in the liverwort (Jungermanniopsida: and to Solenostoma based on chloro- Lejeuneaceae). International Journal of Plant Sciences, 167, plast DNA rbcL sequences. Plant Systematics and Evolution, 1205–1214. 280,243–250. He, X., & Glenny, D. (2010). Perssoniella and the genera of Feldberg, K., Váňa,J.,Long,D.G.,Shaw,A.J.,Hentschel,J.,& : a new classification based on molecular phyloge- Heinrichs, J. (2010). A phylogeny of Adelanthaceae nies. Australian Systematic Botany, 23,229–238. (Jungermanniales, Marchantiophyta) based on nuclear and chloro- Hedenäs, L. (2008). Molecular variation and speciation in Antitrichia plast DNA markers, with comments on classification, cryptic speci- curtipendula s.l. (Leucodontaceae, Bryophyta). Botanical Journal ation and biogeography. Molecular Phylogenetics and Evolution, of the Linnean Society, 156,341–354. 55,293–304. Heinrichs, J., Gradstein, S. R., Wilson, R., & Schneider, H. (2005). Feldberg, K., Heinrichs, J., Schmidt, A. R., Váňa, J., & Schneider, H. Towards a natural classification of liverworts (Marchantiophyta) (2013). Exploring the impact of fossil constraints on the divergence based on the chloroplast gene rbcL. Cryptogamie Bryologie, 26, time estimates of derived liverworts. Plant Systematics and 131–150. Evolution, 299,585–601. Heinrichs, J., Hentschel, J., Feldberg, K., Bombosch, A., & Feldberg, K., Schneider, H., Stadler, T., Schäfer-Verwimp, A., Schmidt, Schneider, H. (2009). Phylogenetic biogeography and taxonomy A. R., & Heinrichs, J. (2014). Epiphytic leafy liverworts diversified of disjunctly distributed bryophytes. Journal of Systematics and in angiosperm-dominated forests. Scientific Reports, 4, 5974. Evolution, 47,497–508. Felsenstein, J. (1985). Confidence limits on phylogenies: an approach Heinrichs, J., Hentschel, J., Bombosch, A., Fiebig, A., Reise, J., using the bootstrap. Evolution, 39,783–791. Edelmann, M., et al. (2010). One species or at least eight? A phylogeny of Cephaloziaceae (Jungermanniopsida) based on nuclear and chloroplast DNA markers 741

Delimitation and distribution of (L.) Dumort. Larget, B., & Simon, D. L. (1999). Markov chain Monte Carlo algorithms (Jungermanniopsida, ) inferred from nuclear and chloro- for the Bayesian analysis of phylogenetic trees. Molecular Biology plast DNA markers. Molecular Phylogenetics and Evolution, 56, and Evolution, 16,750–759. 1105–1114. Marshall, C. R. (2006). Fossil record reveals tropics as cradle and muse- Heinrichs, J., Kreier, H.-P., Feldberg, K., Schmidt, A. R., Zhu, R. L., um. Science, 313,66–67. Shaw, B., et al. (2011). Formalizing morphologically cryptic biolog- Mason-Gamer, R. J., & Kellogg, E. A. (1996). Testing for phylogenetic ical entities: new insights from DNA taxonomy, hybridization, and conflict among molecular data sets in the tribe Triticeae biogeography in the leafy liverwort (Gramineae). Systematic Biology, 45,524–545. (Jungermanniopsida, Porellales). American Journal of Botany, 98, Masuzaki, H., Shimamura, M., Furuki, T., Tsubota, H., Yamaguchi, T., 1252–1262. Majid, H. M. A., et al. (2010). Systematic position of the enigmatic Heinrichs, J., Bombosch, A., Feldberg, K., Kreier, H.-P., Hentschel, J., liverwort Mizutania (Mizutaniaceae, Marchantiophyta) inferred Eckstein, J., et al. (2012a). A phylogeny of the northern temperate from molecular phylogenetic analyses. Taxon, 59,448–458. leafy liverwort genus Scapania (Scapaniaceae, Jungermanniales). McKenna, D.D. & Farell, B.D. (2006). Tropical forests are both evolu- Molecular Phylogenetics and Evolution, 62,973–985. tionary cradles and museums of leaf beetle diversity. Proceedings of Heinrichs, J., Dong, S., Yu, Y., Schäfer-Verwimp, A., Pócs, T., Feldberg, the National Academy of Science USA, 103,10947-10951. K., et al. (2012b). A 150-year old mystery solved: transfer of the Miller, M.A., Pfeiffer, W., & Schwartz, T. (2010). Creating the CIPRES rheophytic endemic liverwort irrorata to . Science Gateway for inference of large phylogenetic trees. In Phytotaxa, 66,55–64. Proceedings of the Gateway Computing Environments Workshop Heinrichs, J., Dong, S., Schäfer-Verwimp, A., Pócs, T., Feldberg, K., (GCE), 14 Nov. 2010, New Orleans, Louisiana, USA (pp. 1-8). Czumaj, A., et al. (2013). Molecular phylogeny of the leafy liver- IEEE. ń wort Lejeunea (Porellales). Evidence for a Neotropical origin, un- Odrzykoski, I. J., Chudzi ska, E., & Szweykowski, J. (1996). The even distribution of sexual systems and insufficient taxonomy. PLoS hybrid origin of the polyploid liverwort Pellia borealis. – ONE, 8, e82547. Genetica, 98,75 86. Heinrichs, J., Feldberg, K., Bechteler, J., Scheben, A., Czumay, A., Pócs, Padial, J. M., Miralles, A., De la Riva, I., & Vences, M. (2010). The – T., et al. (2015). Integrative taxonomy of Lepidolejeunea (Porellales, integrative future of taxonomy. Frontiers in Zoology, 2010,7 16. Jungermanniopsida): ocelli allow the recognition of two neglected Paton, J. A. (1999). The liverwort flora of the .Colchester: species. Taxon, 64,216–228. Harley Books. Hentschel, J., Wilson, R., Burghardt, M., Zündorf, H.-J., Schneider, H., & Patzak, S., Renner, M.A.M., Schäfer-Verwimp, A., Feldberg, K., Heinrichs, J. (2006a). Reinstatement of Lophocoleaceae Heslewood, M.M., Peralta, D. F., et al. (2016). A phylogeny of (Jungermanniopsida) based on chloroplast gene rbcLdata:explor- Lophocoleaceae--Brevianthaceae and a revised clas- ing the importance of female involucres for the systematics of sification of Plagiochilaceae. Organisms Diversity and Evolution, Jungermanniales. Plant Systematics and Evolution, 258,211–226. doi 10.1007/s13127-015-0258-y. Piippo, S. (1984). Bryophyte flora of the Huon Peninsula, Papua New Hentschel, J., Zündorf, H. J., Hellwig, F. H., Schäfer-Verwimp, A., & Guinea. VI: Lepidoziaceae subfam. Lepidozioideae, Heinrichs, J. (2006b). Taxonomic studies in Chiloscyphus Corda Calypogeiaceae, Adelanthaceae, Cephaloziaceae subfam. (Jungermanniales: Lophocoleaceae) based on nrITS sequences and Cephalozioideae and subfam. Odontoschismatoideae and morphology. Plant Systematics and Evolution, 262,125–137. (Hepaticae). Annales Botanici Fennici, 21,309–334. He-Nygrén, X., Ahonen, I., Juslén, A., Glenny, D., & Piippo, S. Posada, D. (2008). jModelTest: phylogenetic model averaging. (2004). Phylogeny of liverworts—beyond a leaf and a thallus. Molecular Biology and Evolution, 25,1253–1256. Monographs in Systematic Botany from the Missouri Botanical Potemkin,A.D.,&Sofronova,E.A.(2013).Taxonomicstudyof Garden, 98,87–118. Cephalozia in Russia and circumscription of the genus. Arctoa, He-Nygrén, X., Juslén, A., Ahonen, I., Glenny, D., & Piippo, S. (2006). 22,173–206. Illuminating the evolutionary history of liverworts Ramaiya, M., Johnston, M. G., Shaw, B., Heinrichs, J., Hentschel, J., von (Marchantiophyta)—towards a natural classification. Cladistics, – Konrat, M., et al. (2010). Morphologically cryptic biological species 22,1 31. within the liverwort . American Journal of Hewitt, G. M. (1996). Some genetic consequences of ice ages, and their Botany, 97,1707–1718. role in divergence and speciation. Biological Journal of the Linnean Renner, M. A. M. (2014). Radula subg. Radula in Australasia and the – Society, 58,247 276. Pacific (Jungermanniopsida). Telopea, 17,107–167. Hewitt, G. (2000). The genetic legacy of the Quaternary ice ages. Nature, Renner, M. A. M., Devos, N., Brown, E. A., & von Konrat, M. J. (2013a). – 405,907 913. New records, replacements, reinstatements and four new species in Hillis, D. M., & Bull, J. J. (1993). An empirical test of bootstrapping as a the Radula parvitexta and R. ventricosa groups method for assessing the confidence in phylogenetic analysis. (Jungermanniopsida) in Australia: cases of mistaken identity. Systematic Biology, 42,182–192. Australian Systematic Botany, 26,298–345. Hutsemekers,V.,Vieira,C.C.,Ros,R.M.,Huttunen,S.,& Renner, M. A. M., Devos, N., Patiño, J., Brown, E. A., Orme, A., Elgey, Vanderpoorten, A. (2012). Morphology informed by phylogeny re- M., et al. (2013b). Integrative taxonomy resolves the cryptic and veals unexpected patterns of species differentiation in the aquatic pseudo-cryptic Radula buccinifera complex (Porellales, moss Rhynchostegium riparioides s.l. Molecular Phylogenetics Jungermanniopsida), including two reinstated and five new species. and Evolution, 62,748–755. Phytokeys, 27,1–113. Jablonski, D., Roy, K., & Valentine, J. W. (2006). Out of the tropics: Rodríguez, F., Oliver, J. L., Martin, A., & Medina, J. R. (1990). The evolutionary dynamics of the latitudinal diversity gradient. general stochastic model of nucleotide substitution. Journal of Science, 314,102–106. Theoretical Biology, 142,485–501. Kreier, H.-P.,Feldberg, K., Mahr, F., Bombosch, A., Schmidt, A. R., Zhu, Ronquist, F., & Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian R.-L., et al. (2010). Phylogeny of the leafy liverwort Ptilidium: phylogenetic inference under mixed models. Bioinformatics, cryptic speciation and shared haplotypes between the Northern and 19, 1572–1574. Southern Hemispheres. Molecular Phylogenetics and Evolution, 57, Rycroft, D. S., Groth, H., & Heinrichs, J. (2004). Reinstatement of 1260–1267. Plagiochila maderensis (Jungermanniopsida: Plagiochilaceae) 742 K. Feldberg et al.

based on chemical evidence and nrDNA ITS sequences. Journal of Váňa, J., & Long, D. G. (2011). Cephaloziaceae (Marchantiophyta) of the Bryology, 26,37–45. Sino-Himalayan region. Nova Hedwigia, 93,153–164. Schuster, R. M. (1974). The Hepaticae and Anthocerotae of North Váňa, J., Söderström, L., Hagborg, A., & von Konrat, M. (2013). Notes America east of the hundredth meridian (Vol. 3). New York and on Early Land Plants Today. 41. New combinations and synonyms London: Columbia University Press. in Cephaloziaceae (Marchantiophyta). Phytotaxa, 112,7–15. Schuster, R. M. (2002). Austral Hepaticae Part II. Beihefte Nova Vanderpoorten, A., Schäfer-Verwimp, A., Heinrichs, J., Devos, N., & Hedwigia, 119,1–606. Long, D. G. (2010). The taxonomy of the leafy liverwort genus Shaw, A. J. (2001). Biogeographic patterns and cryptic speciation in (Lophocoleaceae) revisited. Taxon, 59,176–186. – bryophytes. JournalofBiogeography,28,253 261. Villarreal, J. C., Crandall-Stotler, B. J., Hollingsworth, M. L., Long, D. Shaw, A. J., Boles, S., & Shaw, B. (2008). A phylogenetic delimitation of G., & Forrest, L. L. (2016). Divergence times and the evolution of “ ” the Sphagnum subsecundum complex (Sphagnaceae, Bryophyta). morphological complexity in an early land plant lineage – AmericanJournalofBotany,95,731 744. () with a slow molecular rate. New Phytologist, ň Shaw, B., Crandall-Stotler, B., Vá a, J., Stotler, R. E., von Konrat, M., 209,1734–1746. Engel, J. J., et al. (2015). Phylogenetic relationships and morpho- Vilnet, A. A., Konstantinova, N. A., & Troitsky, A. V. (2010). Molecular logical evolution in a major clade of leafy liverworts (phylum insight on phylogeny and systematics of Lophoziaceae, Marchantiophyta, order Jungermanniales): suborder – Scapaniaceae, and . Arctoa, Jungermanniineae. Systematic Botany, 40,27 45. 19,31–50. Silvestro, D., & Michalak, I. (2012). RaxmlGUI: a graphical front-end for Vilnet, A. A., Konstantinova, N. A., & Troitsky, A. V. (2012). Molecular RaxML. Organisms Diversity and Evolution, 12,335–337. phylogeny and systematics of the suborder Cephaloziineae with Söderström, L., Hagborg, A., von Konrat, M., Bartholomew-Began, S., special attention to the family Cephaloziaceae s.l. Bell, D., Briscoe, L., et al. (2016). World checklist of hornworts and (Jungermanniales, Marchantiophyta). Arctoa, 21,113–132. liverworts. PhytoKeys, 59,1–828. Stamatakis, A. (2014). RaxML Version 8: A tool for phylogenetic analy- Wang, J., Gradstein, S. R., Shi, X.-Q., & Zhu, R.-L. (2014). Phylogenetic position of Trocholejeunea and a new infrageneric classification of sis and post-analysis of large phylogenies. Bioinformatics, doi: 10. – 1093/bioinformatics/BTU033. Acrolejeunea. Bryophyte Diversity and Evolution, 1,31 44. Taberlet, P.,Gielly, L., Pautou, G., & Bouvet, J. (1991). Universal primers Wilson, R., Gradstein, S. R., Heinrichs, J., Groth, H., Ilkiu-Borges, A. L., for amplification of three non-coding regions of chloroplast DNA. & Hartmann, F. A. (2004). Phylogeny of Lejeuneaceae: a cladistic Plant Molecular Biology, 17, 1105–1109. analysis of chloroplast gene rbcL sequences and morphology with Váňa, J. (1988). Cephalozia (Dum.) Dum. in Africa, with notes on the preliminary comments on the mitochondrial nad4-2 spacer region. genus (Notes on some African Hepatic Genera 10). Beihefte Nova Monographs in Systematic Botany from the Missouri Botanical – Hedwigia, 90,179–198. Garden, 98,189 202.