Phylogeny, classification and species delimitation in the ... · PDF file1010 Aranda & al. • Phylogeny of Odontoschisma TAXON 63 (5) • October 2014: 1008–1025 Version of Record

1008

TAXON 63 (5) • October 2014: 1008–1025Aranda & al. • Phylogeny of Odontoschisma

Version of Record (identical to print version).

Received: 17 Jan 2014 | returned for first revision: 2 Jul 2014 | last revision received: 1 Sep 2014 | accepted: 1 Sep 2014 | not published online ahead of inclusion in print and online issues || © International Association for Plant Taxonomy (IAPT) 2014

INTRODUCTION

The growing body of molecular data fuelled by ongoing large-scale barcoding projects, coupled with the fast develop-ment of bioinformatics tools, offer a unique context for objec-tive species delimitations (Monaghan & al., 2009; Fujita & al., 2012). Probably one of the most widely used approaches to date, the generalized mixed Yule-coalescent (GMYC) method, aims at delimiting independently evolving species by identifying the nodes that define the transitions between inter- and intra-specific processes (Fujisawa & Barraclough, 2013; Talavera & al., 2013). This method has been applied in a wide range of

situations, for example when rapid (Barley & al., 2013) and/or recent (Brewer & al., 2012) divergence is not accompanied by morphological change, and in complex patterns of micro-endemism (Vuataz & al., 2013). Since taxonomic issues in describing and understanding biodiversity patterns culminate as organisms decrease in size and observable morphological complexity (Whittaker & al., 2005), techniques such as GMYC have been shown to be especially useful in taxa with reduced morphologies such as amphipods (Murphy & al., 2013) and, among plants, in algae (Payo & al., 2013) and ferns (Chang & al., 2013). In bryophytes, cryptic speciation has been repeat-edly reported (e.g., Heinrichs & al., 2010; Kreier & al., 2010;

Phylogeny, classification and species delimitation in the liverwort genus Odontoschisma (Cephaloziaceae)Silvia C. Aranda,1,2* S. Robbert Gradstein,3* Jairo Patiño,4 Benjamin Laenen,4,5 Aurélie Désamoré4,5 & Alain Vanderpoorten4

1 Departamento de Biogeografía y Cambio Global, Museo Nacional de Ciencias Naturales (CSIC), C/José Gutiérrez Abascal 2, 28006 Madrid, Spain

2 Azorean Biodiversity Group (GBA, CITA-A) and Platform for Enhancing Ecological Research & Sustainability (PEERS), Universidade dos Açores, Departamento de Ciências Agrárias, Rua Capitão João d’Ávila, 9700-042 Angra do Heroísmo, Azores, Portugal

3 Muséum National d’Histoire Naturelle, Dept. Systématique et Evolution, Case Postale 39, 57 rue Cuvier, 75231 Paris cedex 05, France

4 Institute of Botany, University of Liège, 4000 Liège, Belgium5 Institut für Systematische Botanik, Zollikerstrasse 107, 8008 Zürich, Switzerland* contributed equally to this paperAuthor for correspondence: Silvia C. Aranda, [email protected]: SCA, http://orcid.org/0000-0001-6373-8924

DOI http://dx.doi.org/10.12705/635.12

Abstract A species-level phylogeny of Odontoschisma (Jungermanniidae: Cephaloziaceae) was produced to revisit the infrageneric classification and delimit the species within the genus. New methods of species delimitation have been used to explicitly contrast taxonomic hypotheses and test the relevance of the morphological traits traditionally used in this group. The results confirm previous evidence suggesting that the circumscription of Odontoschisma needs to be enlarged to include Iwatsukia and Cladopodiella, and further indicate that a third genus, Anomoclada, is nested within it. Twenty-three molecular entities were recognized based the results of generalized mixed Yule-coalescent (GMYC) analyses. These entities partly conflicted with traditionally defined species that were shown to belong to different, not necessarily closely related entities, adding to the growing body of evidence calling for an extensive revision of species delimitations in taxa with reduced morphologies like leafy liverworts, using an integrative taxonomic approach. A fully revised classification of Odontoschisma into five sections, twenty species, and three subspecies is presented. While the species reported from Europe, Asia and North America were of polyphyletic origins, all Neotropical species were resolved as monophyletic, which could result from a combination of fast speciation rates and reduced dispersal in the Neotropics, and potential extinction in other areas, especially sub-Saharan Africa.

Keywords GMYC; infrageneric classification; integrated taxonomy; liverworts; molecular phylogeny; Neotropics; polyphyletic species

Supplementary Material Electronic Supplement (Table S1; Fig. S1) and alignment are available in the Supplementary Data section of the online version of this article at http://www.ingentaconnect.com/content/iapt/tax

1009

Aranda & al. • Phylogeny of OdontoschismaTAXON 63 (5) • October 2014: 1008–1025


Orzechowska & al., 2010; Ramaiya & al., 2010). In the absence of subsequent morphological re-evaluation, however, cryptic species have only rarely been formally described, calling for the increasing need of an integrative taxonomic approach in the group (Medina & al., 2012, 2013; Heinrichs & al., 2013).

Here, we apply the GMYC model in the context of the phylogeny of the widespread, tropical-holarctic liverwort genus Odontoschisma (Dumort.) Dumort. The last compre-hensive treatment of the genus, which dates back to the early 20th century (Stephani, 1908), included 30 species that are now considered to belong to five genera in four families, including Chiloscyphus Corda (Lophocoleaceae), Lethoco-lea Mitt. (Acrobolbaceae), Notoscyphus Mitt. (Geocalyceae), Odonto schisma and Anomoclada Spruce (Cephaloziaceae) (e.g., Fulford, 1968; Schuster, 1974, 2002; Gradstein & Costa, 2003; So, 2004; Gradstein, 2013). The genus has been differ-ently classified in Cephalozi aceae, Adelanthaceae or its own family, Odontoschismataceae (see Schuster, 1974, for discus-sion). Recent molecular-phylogenetic studies have unequivo-cally recovered the genus as a member of Cephaloziaceae (Forrest & al., 2006; Vilnet & al., 2012). Odontoschisma is defined by having rigid stems without a hyalodermis, leafless ventral branches, undivided leaves, leaf cells with conspicu-ous trigones, underleaves reduced and with slime cells, and antheridial walls made up of irregularly arranged, non-tiered cells (Schuster, 1974, 2002). The circumscription of Odonto-schisma was, however, recently challenged by molecular analyses showing that Iwatsukia jishibae (Steph.) N.Kitag. and the two species of Cladopodiella H.Buch are nested within Odontoschisma (Vilnet & al., 2012), from which they differ by their bifid instead of unlobed leaves. Furthermore, Anomo-clada portoricensis (Hampe & Gottsche) Steph., only species in the genus Anomoclada Spruce, is morphologically very similar to Odonto schisma (Stephani, 1908; Gradstein & al., 2001; Schuster, 2002), but their relationships have not been analysed yet.

A first infrageneric classification of Odontoschisma was proposed by Schuster (1974), who arranged the holarctic species into three sections: sect. Odontoschisma (O. grosseverruco-sum Steph., O. prostratum (Sw.) Trevis., O. sphagni (Dicks.) Dumort.) with only ventral branching, bordered leaves and no gemmae; sect. Denudata R.M.Schust. (O. denudatum (Nees) Dumort., O. elongatum (Lindb.) A.Evans) with lateral and ven-tral branching, unbordered leaves and with gemmae; and sect. Macounia R.M.Schust. (O. macounii (Austin) Underw.) with the same characters as sect. Denudata but lacking secondary pigmentation. A few tropical species were tentatively added by Schuster (2002), but a classification on a worldwide basis has not been attempted yet.

In the present study, we produced a species-level phy-logeny of Odontoschisma based on three plastid markers. We first revisited the infrageneric classification and delimitated the species within the genus, explicitly contrasting traditional taxonomic hypotheses. We then used the phylogeny to test the relevance of morphological traits previously used for species delimitations. Based on these results, a fully revised classifica-tion of the genus is proposed.

MATERIALS AND METHODS

Taxonomic sampling and molecular protocols. — Two sampling strategies were employed to circumscribe Odonto-schisma. The first was carried out at the level of Cephalozi-aceae, considering a large sample of the genera included in the family. The second sampling was performed to define species within the genus and determine their phylogenetic relationships. The first set of analyses (Analysis I) included 16 Odontoschisma specimens, each representing one of 18 spe-cies currently included in the genus (Fulford, 1968; Schuster, 1974, 2002; Gradstein & Costa, 2003; So, 2004; Váňa & al., 2013; Gradstein, 2013, unpub.). Cladopodiella fluitans (Nees) Jørg., C. francisci (Hook.) Jørg., Iwatsukia jishibae, I. bifida (Fulford) Schust. and Anomoclada portoricensis were also included in the sampling to test their possible relations with Odontoschisma. Species of all other genera of Cephaloziaceae as circumscribed by Crandall-Stotler & al. (2009) and Frey & Stech (2009), except Haesselia Grolle & Gradst., Metahygro-biella R.M.Schust. and Trabacellula Fulford, were employed as outgroups, including Nowellia curvifolia (Dicks.) Mitt., Fusco cephaloziopsis lunulifolia (Dumort.) Váňa & L.Söderstr., Cepha lozia bicuspidata (L.) Dumort., Schiffneria hyalina Steph., Hygrobiella laxifolia (Hook.) Spruce, Pleurocladula albescens (Hook.) Grolle, Schofieldia monticola J.D.Godfrey and Alobiellopsis parvifolia (Steph.) R.M.Schust. In the sec-ond set of analyses (Analysis II), multiple accessions of each Odonto schisma species were sampled whenever possible to test their monophyly; 46 accessions were analysed in total (Appen-dix 1). We were unable to sequence the rare Odontoschisma purpuratum Herzog (Malaysia), O. soratamum Fulford (Colom-bia) and Iwatsukia spinosa (Fulford) R.M.Schust. (Venezuela) due to the lack of sufficiently recent specimens for DNA stud-ies. In particular, O. soratamum and I. spinosa are only known from the type specimen.

DNA extraction was conducted using the DNeasy Plant Minikit (Qiagen Benelux B.V., Venlo, The Netherlands). In Analysis I, sequences of the trnL-F region were produced for the ingroup taxa and downloaded from GenBank (http://www .ncbi.nlm.nih.gov/genbank) for the outgroups. In Analysis II, the trnL-F region, the atpB-rbcL intergenic spacer and the rps4 gene were sequenced for each accession. We initially used the universal primers described by Shaw & al. (2003), but due to amplification problems with a considerable num-ber of accessions, we re-designed specific primer pairs for the atpB-rbcL spacer (atpB_odonto 5′-GTTCCTARAGATA GAAGATACATTCTC-3′ and rbcL_odonto 3′-AGTCTCCGT TTGTGGTGACA-5′) and trnL-F region (trnC_odonto 5′-GAT TGTTTCCATATTCAGGG-3′ and trnF_odonto 3′-CCTCTG CTCTACCRACTGA-5′).

The three loci were amplified by polymerase chain reac-tion (PCR) in a 20 µl volume containing 10 µl RNase-free demineralized H2O, 4 µl reaction buffer 5×, 0.8 µl dNTPs (1 mM each), 1.5 µl of 50 mM MgCl2, 0.8 µl of each primer (10 µM), 0.8 µl bovine serum albumin (BSA), 0.8 µl dimethyl sulfoxide (DMSO), 0.5 µl of Go Taq DNA polymerase (5 Uµ–1) and 5 µl DNA. PCR cycling consisted of 5 min denaturation at

1010



94°C, followed by 30 cycles of 30 s denaturation at 94°C, 30 s annealing at 52°C, 50 s extension at 72°C and finally 4 min at 72°C. PCR products were sent to Macrogen (http://www .macrogen.com/) for purification and DNA sequencing.

Sequences were assembled and edited using Sequencher v.4.01 (Gene Codes, 1998) and contigs were aligned with MAFFT v.7 (Katoh & Standley, 2013). Manual refinement of the aligned sequences was performed with SeaView v.4.4.0 (Gouy & al., 2010), inserting gaps where necessary to preserve positional homology. Indels were scored as binary characters regardless of their length using the single indel coding (SIC) algorithm in SeqState v.1.4.1 (Müller, 2005).

Phylogenetic analyses and topological hypothesis test-ing. — Phylogenetic trees were estimated using Bayesian inference (BI) and maximum likelihood (ML) performed in MrBayes v.3.2.2 (Ronquist & al., 2012) and RAxML v.8.0 (Stamatakis & al., 2008), respectively. The three loci were concatenated into a combined dataset after checking visually the congruence of independent analyses for each locus. Indels were added to a separate binary character matrix. The general time reversible (GTR) with a gamma distribution to take varia-tion among sites into account was selected as the best-fit model based on the minimum Akaike information criterion (AIC) with jModelTest v.2.1.4 (Darriba & al., 2012). A separate model with identical forward and reverse transition rates was applied to the indel matrix. For BI analyses, all model parameters were unlinked and estimated independently across data partitions except topology and branch length. Four independent Markov Chain Monte Carlo (MCMC) chains were run for 5 million generations, with default uniform priors, and the posterior distribution was sampled every 1000th generations to ensure independence from successive sampling. Burn-in was pruned and effective sample size (ESS) estimation was used to assess MCMC convergence in Tracer v.1.5 (Rambaut & Drummond, 2007). For ML analyses, support for branches was assessed by a non-parametric bootstrap analysis with 100 replicates

Constrained analyses were performed to test the likeli-hood of the topology resulting from the Bayesian inference (i.e., null hypothesis) against three alternative hypotheses forc-ing traditional classifications of (1) Cladopodiella fluitans and C. francisci as sister species, (2) Odontoschisma sphagni and O. prostratum as different species, and (3) the monophyly of O. grosseverrucossum accessions. The clades corresponding to sect. Odontoschisma, sect. Macounia and sect. Denudata were kept constant in both the null and alternative hypotheses, following the recommendations of Bergsten & al. (2013). Model likelihoods were estimated with the stepping-stone method (Xie & al., 2011) using 196,000 MCMC steps sampled every 500th generations for each of 50 b-values between 1 (posterior) and 0 (prior) after discarding the first 196,000 generations as initial burn-in set by default. The contribution to the marginal likelihood from each step is estimated from a sample-size of 392. Analyses were run for four independent MCMC chains from which the arithmetic mean of marginal likelihoods was estimated for each model to calculate Bayes factors (BF).

Phylogenetic species delimitation. — We used the gen-eralized mixed Yule-coalescent (GMYC) model to delimit

putative species on the basis of phylogenetic information (for review, see Fujisawa & Barraclough, 2013). This method diag-noses species-level clades by identifying the shift in diversifica-tion rate associated with the change from reticulate evolution (a neutral coalescent model) to divergent evolution (a Yule pure birth model). The null model assumes that the entire sample derives from a single population undergoing a single coalescent process, whereas the GMYC model classifies the observed branching time intervals into two categories, as the result of either inter- or intraspecific processes of lineage sorting. A like-lihood ratio test is then used to identify the best-fit model. The method was extended to allow multiple thresholds to account for variation in the timing to speciation (Monaghan & al., 2009). Although the multiple threshold version may show a cer-tain tendency to oversplitting, its use has been recommended to explore delimitation changes when the assumptions of the single threshold needs to be relaxed and when, for instance, the sampling of individuals across a representative occupancy range may be achieved (Fujisawa & Barraclough, 2013).

In order to apply the GMYC technique, we first dated the obtained phylogeny by performing a relaxed-clock analysis in BEAST v.1.7.5 (Drummond & al., 2012). We employed four independent MCMCs to sample a prior distribution of absolute substitution rates derived from the analysis of the entire Liver-wort Tree of Life calibrated with 21 fossils (unpub. results). This distribution had a mean of 4.453 × 10–4 and a standard deviation of 1.773 × 10–6 substitutions per site per Ma. Each MCMC was run for 100 million generations, with sampling every 10,000th generations under a birth-death and a Yule speciation model, respectively. The latter was selected against the former based on Bayes factors and employed in subsequent analyses. Con-vergence and mixing of the four chains was checked using the program Tracer v.1.5. One thousand trees were discarded as burn-in and the remaining trees from each of the four runs were combined.

The 50% majority-rule chronogram derived from the dat-ing analysis was then used to apply the GMYC model. Follow-ing Powell & al. (2011), we ran the analyses without removing the outgroups. The GMYC output also produced a lineage-through-time plot, which we visually evaluated for changes in branching rate. Single and multiple-threshold GMYC models were run using the SPLITS package in R, with default options (Ezard & al., 2009). After estimating likelihoods for single (Pons & al., 2006) and multiple thresholds (Monaghan & al., 2009), a modified Akaike information criterion corrected for small sample size (AICc; for details see Powell, 2012) was used to identify the best GMYC model.

Morphological analysis. — Each accession was scored for 20 morphological characters that have been used in the clas-sification of Odontoschisma (Appendix 2). Ancestral character state reconstructions were performed using a reverse-jump Markov Chain Monte Carlo (hereafter RJ-MCMC) (Pagel & Meade, 2006), as implemented in BayesTraits v.2.0 (Pagel & al., 2004). Character evolution was mapped on the sample of trees from the posterior probability distribution produced by the MrBayes analysis (see below) after pruning of the out-groups. At each iteration, the chain samples a tree, one of

1011



four transition models (including one or two rate models, and models wherein either the forward or the backward rate is 0), and values of rate parameters. The resulting combination of those parameters is then accepted or rejected depending on the Metropolis-Hastings term. When a combination of tree and rate parameters is accepted, the rate values are used to derive the set of ancestral states at each internal node simultaneously (“global approach”; Pagel, 1999). The chain was run for 10 mil-lion generations, discarding 25% of burn-in and sampling every 10,000th generation. To circumvent the issue associated with the fact that not all of the sampled trees necessarily contain the internal nodes of interest, reconstructions were performed using a most recent common ancestor (MRCA) method. This method identifies, for each tree in the Bayesian sample, the MRCA to a group of species and reconstructs the state at the node, then combines this information across trees. In order to contrast alternative hypotheses regarding the ancestral state at key nodes of the phylogeny, we also implemented the “local” approach, wherein the significance of the reconstruction is explicitly tested at each node of interest (Pagel, 1999). For that purpose, we successively fixed each of the nodes of interest at one of the states it can possibly take. Then, the RJ-MCMC was re-run as described above. A second, independent chain was run to sample rate parameters and derive overall likelihoods of the reconstructions when the node was fixed at its alternative state(s). Bayes factors were used to determine the support for alternative states.

RESULTS

Circumscription and intra-generic classification of Odon-toschisma. — The 50% majority-rule consensus of the trees sampled from the posterior probability distribution generated by the MrBayes analysis of the trnL-F matrix (Analysis I) is presented in Fig. 1. All Odontoschisma species sampled were resolved within a clade supported with a posterior probability (PP) of 100 and a ML bootstrap support (BS) of 100%. The two species of Iwatsukia, the two species of Cladopodiella, and Anomoclada were nested within that clade.

In the trnL-F, rps4, and atpB-rbcL matrix (Analysis II), 27.8% of the 1604 sites were variable within Odontoschisma. The four independent MCMC MrBayes runs showed good mix-ing with large effective sample sizes (ESS > 200) and converged successfully to the stationary phase, as shown by an average standard deviation of split frequencies of 0.006 and average potential scale reduction factor for parameter values of 1. The 50% majority-rule consensus of the 6622 trees sampled from the posterior probability distribution is presented in Fig. 2.

Odontoschisma sect. Odontoschisma (including O. pros-tratum, O. sphagni and O. grosseverrucosum) and sect. Denu-data (including O. macounii, O. denudatum, O. sandvicense (Ångstr.) A.Evans, O. subjulaceum Austin, O. naviculare (Steph.) Grolle, O. cleefii Gradst. & al., O. longiflorum (Taylor) Trevis., O. variabile (Lindenb. & Gottsche) Trevis., O. engelii Gradst. & Burghardt, O. brasiliense Steph., O. portoricense,

0.05

88/-

98/100100/100

99/-

99/-

58/-

99/89

100/96

100/100

70/67

69/-

100/-

99/52

88/-

100/99

99/-

82/-

O. prostratum O4

O. sphagni OD6O. grosseverrucosum OG3

O. fluitans CFLU

O. bifidum IWAO. jishibae

O. macounii OD7O. denudatum OD8O. longiflorum O3

O. engelii OEN1O. variabile OV4

O. denudatum subsp. sandvicense O11

O. brasiliense OBRA

O. portoricense O8

O. denudatum subsp. naviculare ON3

Alobiellopsis parvifolia

Nowellia curvifolia

Fuscocephaloziopsis lunulifolia

Cephalozia bicuspidata

Schiffneria hyalinaHygrobiella laxifolia

Pleurocladula albescensSchofieldia monticola

O. elongatum OE1

O. francisci CFRA

Fig. 1. Fifty per cent majority-rule consensus of the trees sampled from the posterior prob-ability distribution generated by a Bayesian analysis of the trnL-F region showing the relationships of Odonto schisma, Iwatsukia, Cladopodiella and Anomoclada species in Cephaloziaceae. Numbers above branches correspond to Bayesian posterior prob-abilities (left) and non-parametric bootstrap proportions (right) from the maximum likelihood analysis. Accession identifiers are given for newly generated sequences. Species names follow the new nomenclature (see below).

1012



0.04

Now

ellia

cur

vifo

lia

O. d

enud

atum

OD

18

O. s

phag

ni O

SP

14

O. f

luita

ns C

FLU

O. f

ranc

isci

CFR

A

O. d

enud

atum

OD

8

O. j

ishi

bae

O. z

hui D

EN

A2

O. z

hui O

N2

Sch

iffne

ria h

yalin

a

O. v

aria

bile

OV

4

O. p

rost

ratu

m O

4

O. d

enud

atum

sub

sp. n

avic

ular

e O

N3

O. b

rasi

liens

e O

BR

A

O. d

enud

atum

OD

17

O. p

orto

ricen

se O

PO

R

O. m

acou

nii O

M2

O. e

long

atum

OE

1

O. d

enud

atum

OD

13

O. d

enud

atum

OD

12

O. e

long

atum

OS

P10

O. e

long

atum

OE

2

O. c

f. lo

ngifl

orum

OV

2O

. eng

elii

OE

N1

O. g

ross

ever

ruco

sum

OG

3

O. s

phag

ni O

D6

O. c

leef

ii O

9

Fusc

ocep

halo

ziop

sis

lunu

lifol

ia

O. p

rost

ratu

m O

6

O. d

enud

atum

sub

sp. s

andv

icen

se O

11

O. p

seud

ogro

ssev

erru

cosu

m O

G2

O. z

hui D

EN

A1

O. p

orto

ricen

se O

8

O. m

acou

nii O

M1

O. s

phag

ni O

SP

1

O. p

rost

ratu

m O

PR

1

O. p

rost

ratu

m O

PR

2

O. v

aria

bile

OV

3

O. p

seud

ogro

ssev

erru

cosu

m O

7

O. b

ifidu

m IW

A

O. m

acou

nii O

D7

O. s

phag

ni O

D1

O. p

rost

ratu

m O

PR

11

O. m

acou

nii O

M3

O. v

aria

bile

DP

AR

O. l

ongi

floru

m O

3

O. s

phag

ni O

SP

4

O. d

enud

atum

OD

15

O. v

aria

bile

OV

5

Cep

halo

zia

bicu

spid

ata

100/

100

100/

100

100/

98 98/6

5

93/6

8

100/

100

77/9

9

100/

94

51/-

100/

100

78/6

2

97/6

6

98/-

100/

100

100/

98

100/

100

100/

98

82/7

8

98/7

5100/

100

100/

100

99/7

5

100/

92

72/-98

/-

100/

99100/

94

100/

96 100/

100

100/

100

100/

- 84/8

4

100/

8973

/-

83/-

100/

-

100/

100

99/9

098/-

88/7

5

sect

. Odo

ntos

chis

ma

sect

. Den

udat

a

sect

. Iw

atsu

kia

sect

. Nee

sia

sect

. Cla

dopo

diel

la

A

B

C

DE

F

G

H

I

JKL

M

N OP

Q R

S

T

U

V

Fig

. 2. P

hylo

gram

of t

he 5

0% m

ajor

ity-r

ule

cons

ensu

s of t

he tr

ees s

ampl

ed b

y th

e M

rBay

es a

naly

sis b

ased

on

atpB

-rbc

L, tr

nL-F

and

rps4

sequ

ence

s with

in O

dont

osch

ism

a. T

he p

hylo

gene

tic

posit

ion

of O

. jis

hiba

e (d

ashe

d br

anch

) is o

nly

base

d on

trnL

-F se

quen

ce. N

umbe

rs a

bove

bra

nche

s cor

resp

ond

to B

ayes

ian

post

erio

r pro

babi

litie

s (le

ft) a

nd n

on-p

aram

etric

boo

tstra

p pr

opor

-tio

ns (r

ight

) fro

m th

e m

axim

um li

kelih

ood

anal

ysis

. Bra

nche

s in

bol

d re

pres

ent a

cces

sions

gro

uped

with

in th

e sa

me

mol

ecul

ar e

ntity

by

the

GM

YC

anal

ysis

. Let

ters

indi

cate

the

node

s fo

r ch

arac

ter s

tate

reco

nstr

uctio

n (s

ee E

lect

r. Su

ppl.:

Tab

le S

1).

1013



O. zhui Gradst. & al., O. elongatum and O. pseudogrossever-rucosum Gradst. & al.) were resolved with PP/BS values of 100/100 and 72/-, respectively. Section Odontoschisma showed a synapomorphic acquisition of bordered leaves and a transi-tion of leaf surface from concave to ± flat, while the absence of lateral branches was homoplastic (see, respectively, characters 6, 12 and 3 in Appendix 2). The inclusion of O. portoricense (= Anomoclada portoricensis) within sect. Denudata was sup-ported by lateral branches, unbordered leaves and presence of gemmae, and its sister relationship with O. brasiliense by a synapomorphic increase in gametophyte size (character 1, Appendix 2). Anomoclada-type branches, previously consid-ered exclusive to O. portoricense, were observed in several unrelated species of sect. Denudata (node M, Fig. 2; Table S1), so that the evolution of this trait was homoplastic (charac-ter 4, Appendix 2). Section Macounia was nested within sect. Denudata; some of its characteristic features were indepen-dently acquired by several other members of sect. Denudata. Thus, mid-leaf cell size (character 10) showed a synapomorphic increase independently in sect. Macounia and in the clade of O. portoricense and O. brasiliense (nodes G and R in Fig. 2, respectively), and the cuticle (character 16) shifted indepen-dently from finely verruculose to smooth in sect. Macounia and O. elongatum (Appendix 2). Odontoschisma francisci (Hook.) L.Söderstr. & Váňa (= Cladopodiella francisci) was sister to sect. Denudata, whereas O. fluitans (Nees) L.Söderstr. & Váňa (= C. fluitans) and the position of the two Iwatsukia species was unresolved outside of these two main sections. Morpho-logically, Iwatsukia spp. and O. fluitans are unique within Odonto schisma in having leaves divided to 1/3 of their length (Appendix 2). The polyphyletic origin of the two Cladopodi-ella species was further supported by the constrained analy-ses, where forcing the two species into monophyly resulted in a substantial decrease in the marginal log-likelihood (BF = 26.21). Accordingly, traits previously identified as diagnostic for Cladopodiella, including the presence of nodular thicken-ings on alternate long walls in the capsule epidermis (character 20) and bifid leaves (character 7), were reconstructed as having evolved at least twice independently (Table S1).

Within sect. Denudata, the first dichotomy resolved a clade with a posterior probability of 72 but lacking ML bootstrap support, comprised of two accessions of O. pseudogrossever-rucosum on the one hand, and the remaining species of the section on the other, which were characterized by the syn-apomorphic lack of underleaves and appressed to squarrose gemmiparous leaves (characters 17 and 18 in Appendix 2). The South American accessions of the section formed a fully supported clade (node O, Fig. 2), whose most recent common ancestor was dated at 29 (21–39) Ma (Electr. Suppl.: Fig. S1). The sister clade (node J, Fig. 2) was formed by accessions of O. denudatum from Europe, Australasia and Hawaii, which were morphologically characterized by the shared presence of Anomoclada-type branches (character 4, Appendix 2). Within clade O, the relationship between O. longiflorum, O. engelii, and O. variabile was supported by the synapomorphic transi-tion towards elongate leaves with a length to width ratio of up to 1.5 : 1 (character 8, Appendix 2).

Phylogenetic species delimitation within Odontoschisma. — The single-threshold GMYC was identified as the best-fit model (AICc = 134.05) over models with multiple thresholds. Under the single-threshold model, the null hypothesis (i.e., all specimens belong to a single species) could be significantly rejected (P = 0.011). Twenty-three molecular entities were identified by the GMYC model within the genus (Fig. 2), 9 as distinct clusters of haplotypes and 14 as singletons.

All accessions of O. sphagni and O. prostratum but one (OD1) were identified as a single species morphologically char-acterized by the synapomorphic transition from obtuse or acute to long acuminate or ciliate female bract apices (character 19, Appendix 2). The respective accessions of these two species did not form monophyletic groups, and constraining the acces-sions of O. sphagni on the one hand, and of O. prostratum on the other, to monophyly, resulted in a substantial decrease in log-likelihood (BF = 34.19). Species represented by a single accession, including O. fluitans, O. francisci, O. bifidum, O. brasiliense and O. engelii were not recognized as closest relatives to each other. Multiple accessions of O. longiflorum, O. macounii, O. portoricense, O. elongatum and O. variabile were resolved as monophyletic with high supported values of and were recognized as distinct species by the GMYC analysis (Fig. 2). The three accessions of O. grosseverrucosum (labelled as O. pseudogrosseverrucosum and O. grosseverrucosum in Fig. 2) were resolved as polyphyletic, two accessions from Japan forming a fully supported clade that was sister to the remainder of sect. Denudata and one accession from China that was included within sect. Odontoschisma with full support. Consistent results were shown by constraint analyses enforc-ing the three accessions together, which did not support the monophyly hypothesis (BF = 119.11). Odontoschisma denudatum as previously circumscribed was polyphyletic. One accession from Colombia (labelled as O. cleefii in Fig. 2) was resolved within a clade also comprised of O. longiflorum and O. engelii. Three monophyletic accessions from China initially labelled as O. zhui in Fig. 2 were recovered as a distinct species (PP/BS values of 100/100). The group was supported by a synapomor-phic transition towards appressed gemmiparous leaves (char-acter 18, Appendix 2). The remaining accessions of O. denu-datum were resolved as paraphyletic. The core group of four European accessions was recognized by the GMYC model as a distinct species sister to a clade comprised of two accessions from Hawaii (node L in Fig. 2), one of O. denudatum (= O. sub-julaceum) and the other of O. denudatum subsp. sandvicense (= O. sandvicense). These two clades were sister to a clade comprised of two accessions (node N, Fig. 2), one of O. denu-datum from Japan and one of O. denudatum subsp. naviculare (= O. naviculare), each recovered as a distinct species.

DISCUSSION

Circumscription and infrageneric classification of Odon-toschisma. — The present analyses confirm previous evidence suggesting that the circumscription of Odontoschisma needs to be enlarged to include Iwatsukia and Cladopodiella (Vilnet

1014



& al., 2012) and further indicate that a third genus, Anomo-clada, is nested within it. In contrast to recent taxonomic treat-ments identifying Anomoclada as a separate genus (Fulford, 1968; Gradstein & al., 2001; Gradstein & Costa, 2003), the present results indicate that Anomoclada needs to be synony-mized with Odontoschisma, adding to the growing body of evidence challenging the taxonomic status of the large number of monotypic liverwort genera described from South America (Gradstein & al., 2001). The inclusion of Anomoclada into the synonymy of Odontoschisma is consistent with the general morphological similarity between the two taxa, which both have prostrate stems with ventral stolons and without hyaloder-mis, succubous undivided leaves, thick-walled leaf cells with distinct trigones and rudimentary underleaves with slime cells. The two main features that led Spruce (1876) and subsequent taxonomists to recognize Anomoclada at the genus level were the abundant secretion of slime by the underleaves and the development of the leafy branches from the dorsal side of the stem, so-called Anomoclada-type branches (Crandall-Stotler, 1969). Evans (1903), however, showed that mucilage secretion on underleaves also occurs in Odontoschisma, yet to a much lesser degree. The close morphological similarity of the two genera is further supported by the presence of Anomoclada-type branches in O. denudatum (Schuster, 1974) and newly reported here in O. longiflorum. Ancestral character state reconstructions indicated that evolution of this trait has been homoplastic and cannot serve as a criterion that warrants rec-ognition of Anomoclada as a distinct genus.

Two of the three sections identified within Odontoschisma (Schuster, 1974), namely sect. Odontoschisma and sect. Macou-nia, were resolved as monophyletic, but the recognition of the latter would render sect. Denudata paraphyletic. This suggests that sect. Macounia should be merged with sect. Denudata, which is consistent with the sharing of morphological traits between O. macounii and other members of sect. Denudata, such as presence of lateral branching, concave leaves with-out border, mid-leaf cells 25–40 µm, rounded to acute api-ces of female bracts, and asexual reproduction by gemmae. The circumscription of sect. Denudata needs to be further expanded to include O. portoricense, O. brasiliense, O. longi-florum, O. variabile, O. engelii, O. sandvicense, O. naviculare, O. macounii and three new species (see below). The inclu-sion of O. brasiliense within sect. Denudata is at odds with its strongly bordered leaves and apparent lack of gemmae, which are characteristic of sect. Odontoschisma. The species deviates from members of sect. Odontoschisma, however, by produc-ing lateral branches (even though sparingly), which are always absent in the species of that section.

The inclusion of Iwatsukia spp., O. fluitans, and O. fran-cisci in Odontoschisma, but outside of sect. Odontoschisma and sect. Denudata, makes it necessary to accommodate these taxa in new sections. The two species formerly included in Iwatsukia form a clade and are accommodated within the new section Iwatsukia. Conversely, the two species of the former genus Cladopodiella are resolved as polyphyletic and are therefore accommodated into two new Odontoschisma sec-tions, sect. Cladopodiella comb. nov. and sect. Neesia sect.

nov., respectively. The polyphyly of the two species previously included in Cladopodiella indicates that the sharing of morpho-logical traits by these two species, such as bifid leaves, tiered antheridial wall cells (Schuster, 1974), and the tendency for nodular thickenings in the capsule epidermis to be developed on all longitudinal walls instead of on alternate longitudinal walls only, is homoplastic. The development of nodular thicken-ings on alternate walls is a characteristic and constant feature of several liverwort families, including Calypogeiaceae, Cepha-loziaceae (excluding Hygrobiella Spruce), Lepidoziaceae, and Pleurozi aceae (Schuster, 1984; Crandall-Stotler & al., 2009), so that the independent acquisition of this trait in O. fluitans and O. francisci was highly unexpected. The presence of species with undivided or bilobed leaves in a single genus was hitherto unknown in Cephaloziaceae. Nevertheless, genera including species with either undivided or bilobed leaves are not rare in leafy liverworts such as Kymatocalyx Herzog (Cephaloziell-aceae), Marsupella Dumort. (Gymnomitriaceae), Plagiochila (Dumort.) Dumort. (Plagiochilaceae), Syzygiella Spruce (Jame-soniellaceae), and various genera of Lepidoziaceae and Lopho-coleaceae, even though leaves in these genera may sometimes be shallowly bifid only. In leafy liverworts, leaves usually pos-sess two or more growing points, each with its own embryonic apical cell, so that developing leaves are initially typically lobed (Schuster, 1984). Undivided leaves could therefore theoretically be considered as ontogenetically derived, although phyloge-netic evidence is equivocal (contrast, e.g., the order of acquisi-tion of undivided vs. bilobed leaves in Plagiochila, Groth & al., 2002; Roivainenia Perss. [= Syzygiella], Feldberg & al., 2010; and Lophocolea (Dumort.) Dumort., Hentschel & al., 2007).

Species delimitation within Odontoschisma. — Twenty-three molecular entities were recognized based on the results of the GMYC analyses. These entities partly conflicted with traditionally defined species that were shown to belong to dif-ferent, not necessarily closely related entities. This adds to the growing body of evidence showing the need for extensive revi-sion of species delimitations in taxa with reduced morpholo-gies like bryophytes using an integrative taxonomic approach (Sukkharak & al., 2011; Dong & al., 2012; Hutsemékers & al., 2012; Medina & al., 2012, 2013, Hedenäs & al., 2014).

Odontoschisma cleefii sp. nov. (Fig. 3), previously tenta-tively identified as a subspecies of O. denudatum on morpho-logical grounds, is the most common species of the genus in the páramos of the northern Andes and currently only known from Colombia but expected to occur in neighbouring countries as well. The new species is characterized by its concave leaves densely covered by large papillae (ca. 4–15 µm long and 4–10 µm wide), which are present all over the inner leaf surface. Such large papillae are observed only in the eastern Asian species O. grosseverrucosum and O. pseudogrosse verrucosum. Fur-ther characteristics of O. cleefii include the frequent presence of a whitish leaf border consisting of collapsed leaf-margin cells, the presence of gynoecia in the upper portion of the stem, while they occur in the lower portion of the stem in other species, and the broadly rounded to somewhat obtuse apices of the female bracts and bracteoles, which are never acute-acuminate like in the other species of the genus. The plants are typically small in

1015



size (0.8–1.5 mm wide) and dark purplish, which is presumably an adaption to the strong irradiation in the high tropic-alpine environment. Because of their dark pigmentation, most collec-tions were previously identified as O. atropurpureum Steph., a name given in the past to all small, dark purplish pigmented páramo plants of the genus including O. cleefii, O. denudatum and O. variabile, to which the type of O. atropurpureum is now assigned (Gradstein, 2013). Odontoschisma cleefii is separated from O. denudatum and O. variabile by its coarsely papillose leaf surfaces. In O. denudatum and O. variabile, the papillae are usually very small, less than 2.5 µm long and only well visible at the leaf margin, the leaf surfaces appearing smooth. Odontoschisma variabile, moreover, differs by having flat, not concave leaves.

The polyphyly of O. grosseverrucosum as previously cir-cumscribed indicates that the focus on the presence of con-spicuous papillae to identify one unique Asian species (Hattori, 1962; Gao & Cao, 2000; So, 2004) was misleading and that they belong to two distinct species in different sections. The existence of two species within O. grosseverrucosum s.l. was noted by Hattori (1950), who described plants belonging to O. pseudo grosseverrucosum as “O. grosseverrucosum” and the true O. grosseverrucosum as O. lutescens S.Hatt. In later publications, however, Hattori (1962) did not maintain this dis-tinction, and So (2004) showed that O. lutescens is a synonym of the true O. grosseverrucosum. The accession from China, resolved within sect. Odontoschisma, fully matches the mor-phological characters of the type, whereas the two accessions from Japan are described as O. pseudogrosseverrucosum sp. nov. (Fig. 4). Re-study of herbarium material showed that the two species clearly differ in plant size, stem epidermis, branch-ing, presence of asexual reproduction and leaf characters. Odontoschisma pseudogrosseverrucosum is a smaller plant than O. grosseverrucosum, measuring only 0.5–0.65(–0.8) mm in width (0.8–1.6 mm in O. grosseverrucosum), the dorsal epi-dermis cells are smaller, subquadrate and thick-walled (rect-angular and rather thin-walled in O. grosseverrucosum), the leaves are distinctly concave and never bordered (rather flat and bordered in O. grosseverrucosum) and lateral branches and gemmae may be present (absent in O. grosseverrucosum). The two species have also different geographical ranges, O. pseudo-grosseverrucosum occurring in temperate eastern Asian (cen-tral Japan, South Korea, Russian Far East), whereas O. gros-severrucosum is a subtropical species ranging from southern Japan to northern Thailand and Vietnam.

The specimens assigned here to O. zhui sp. nov. (Fig. 5) were previously identified as O. denudatum on morphologi-cal grounds, but were resolved in the molecular analysis as a separate lineage that is sister to O. elongatum. Odontoschisma zhui is morphologically similar to O. denudatum but differs from the latter by smaller leaf cells (13–22 µm long in mid-leaf), asymmetrical, curved-subfalcate leaves, and appressed leaves on gemmiparous branches. In its peculiar leaf shape and gemmi parous shoots O. zhui is also similar to O. naviculare from tropical SE Asia and New Caledonia, but it differs from the latter by smaller leaf cells, absence of a dorsal leaf-free strip, and predominantly ventral branching.

Within the O. denudatum clade, which includes O. denu-datum, O. sandvicense and O. naviculare, the GMYC model identified five molecular entities: a core European O. denuda-tum (4 accessions), O. naviculare (New Caledonia; 1 accession), O. sandvicense (Hawaii; 1 accession), O. subjulaceum (Hawaii; 1 accession), and one further accession of O. denudatum from Japan sister to O. naviculare. As a result, O. denudatum is para-phyletic in our analysis. One taxonomic interpretation would be to strictly follow the GMYC criterion and indeed recognize five species. We refrain, however, from doing so because the accessions of O. denudatum and O. subjulaceum were morpho-logically identical, perfectly matching the morphological cir-cumscription of the species. We believe that recognizing a truly cryptic species in a genus where all other species display clear morphological differentiation would not be sound and therefore consider O. subjulaceum as a synonym of O. denudatum. A second option would be to recognize three species, including the core European O. denudatum, O. sandvicense (including the O. denudatum subsp. sandvicense accession and the Hawaiian O. denudatum), and O. naviculare (including the O. denudatum subsp. naviculare accession and the Japanese accessions of O. denudatum). In fact, the paraphyly of O. denudatum and the strong geographic partitioning of genetic variation is consistent with the phylogenetic patterns expected under budding specia-tion (Funk & Omland, 2003). This solution, however, would render O. sandvicense and O. naviculare heterogeneous and indistinguishable at the morphological level. Since these two species can each be distinguished from O. denudatum by one morphological trait, O. sandvicense differing by flat leaves (concave in O. denudatum) and O. naviculare by curved-falcate leaves (straight in O. denudatum) (see Appendix 1), we treat O. sandvicense and O. naviculare as subspecies of O. denu-datum. The use of subspecies has intimately been associated with the idea of geographic separation (the “Rassenkreis” or “geographical subspecies” in Meikle, 1957) and evolutionary biologists traditionally used the subspecies category to diag-nose geographically distinct populations that were thought to be in the early stages of speciation (Mayr, 1963). Attribution of the subspecies rank to O. sandvicense and O. naviculare is further warranted by the fact that subspecies are unlikely to meet the criterion of monophyly (Crandall & al., 2000) because they represent an incipient stage of differentiation prior to reciprocal monophyly (Zink, 2004).

The GMYC model failed to resolve O. sphagni and O. prostratum as distinct. Constrained analyses further indi-cated that a monophyletic origin of accessions assigned to each species can be rejected. Morphologically, O. prostratum differs from O. sphagni mainly by smaller leaf cells and plant size. The size difference is especially pronounced in eastern North America, where both species occur. Moreover, the two species differ in habitat preference, O. prostratum usually occurring on soil or rock, whereas O. sphagni typically grows on peaty soil (Schuster, 1974). In spite of their different ecological pref-erences, O. sphagni and O. prostratum show much overlap in their morphology. European populations of O. sphagni may be as small as those of O. prostratum, with leaves measuring only 0.7–0.8 mm long. These plants can only be separated from

1016



O. prostratum by their somewhat larger leaf cells, measuring 20–35 µm long in O. sphagni and 15–22(–25) µm in O. prostra-tum. The trigones in O. sphagni are usually medium-sized and the lumina are rounded, whereas O. prostratum typically has small trigones and subquadrate lumina. Specimens of O. pros-tratum with medium-sized trigones and rounded lumina can, however, be observed. Except for their habitat and slightly smaller cells, these plants are inseparable from O. sphagni. The leaf border can be slightly wider in O. prostratum (1–4 cells wide) than in O. sphagni (1–2(–3) cells wide) (Schuster, 1974). However, the leaf border is often not well-defined, its width varies and clearly shows much overlap in both species, so that this character does not allow for separating the two taxa. In the light of the polyphyletic origin of accessions of both species and the continuum in morphological variation in traits assumed to be diagnostic, O. prostratum is reduced to synonymy under O. sphagni, as proposed by Vilnet & al. (2012).

Biogeographic patterns in Odontoschisma. — The syn-onymy of O. prostratum and O. sphagni, and the polyphyly of different accessions from North America, Europe, and Azores, is reminiscent of similar evidence for recurrent trans-Atlantic migrations across the North Atlantic in moss species (Szöevenyi & al., 2008; Stenøien & al., 2011). Conversely, mounting evi-dence indicates that amphi-Atlantic bryophyte species some-times correspond to complexes of semi-cryptic endemic species (e.g., Heinrichs & al., 2011; Hutsemekérs & al., 2012; Medina & al., 2012, 2013). Altogether, these observations suggest that previous assessments of the patterns of amphi-Atlantic disjunc-tions in bryophytes based on check-lists, with 43% of the spe-cies of mosses found in North America also found in Europe, and 70% of the moss species in Europe also found in North America (Frahm & Vitt, 1993), need to be revised.

While the species reported from Europe, Asia and North America were of polyphyletic origins, all of the Neotropical species were resolved in a fully supported clade and only one species of the group, O. variabile, has a disjunct Afro-American range. Three hypotheses could be proposed to explain this pat-tern. First, the Neotropics appear as a center of diversification for liverwort diversity (Gradstein & al., 2001; Vanderpoorten & al., 2010b), potentially triggered by the Miocene uplift of the Andes (Heinrichs & al., 2005a), and the recent origin of Neo-tropical species explains their endemicity. This hypothesis is, however, at odds with the timing of diversification of the Neo-tropical Odontoschisma clade, 30 Ma. Second, the Neotropical species have a low long-distance dispersal capacity owing, for instance, to the poor tolerance of their spores to the drought and frost conditions that prevail in high-altitude air currents. This hypothesis is supported by the fact that spores of the páramo endemic O. cleefii show a very poor drought resistance, losing their capability of germination after one day of drying (Van Zanten & Gradstein, 1988). Third, the Neotropical species of Odontoschisma could be paleo-endemics. In particular, the com-paratively low levels of Afro-American disjunctions (about 5% of the flora of tropical America; Gradstein, 2013) as compared to the North Atlantic disjunction, could be interpreted in terms of the influence of extensive drought periods in Africa during the Pleistocene, which led to an extensive decline of moist forests

and possibly to the extinction of parts of the local hepatic flora (Heinrichs & al., 2005b). These hypotheses are, however, not mutually exclusive, and the high levels of Neotropical endemic diversity observed in Odontoschisma, also reported in other liverwort genera such as Diplasiolejeunea (Spruce) Schiffn. (Dong & al., 2013), Leptoscyphus Mitt. (Vanderpoorten & al., 2010a), and Plagiochila (Heinrichs & al., 2006), could result from a combination of fast speciation rates, reduced dispersal capacities and potential extinctions in other areas.

TAXONOMY

Based on the analyses presented here, the following clas-sification of Odontoschisma is presented, with citation of accepted name, basionym and type. Full synonymies, species descriptions, keys to species, geographic distribution and eco-logical preferences are given in the forthcoming monograph of the genus (Gradstein & Ilkiu-Borges, in press). Three species not available for molecular analysis (O. purpuratum, O. sorata-mum, O. spinosum) are marked by an asterisk.

Odontoschisma (Dumort.) Dumort., Recueil Observ. Jungerm.: 19. 1835 ≡ Pleuroschisma sect. Odontoschisma Dumort., Syll. Jungerm. Europ.: 68. 1831 – Type: Odontoschisma sphagni (Dicks.) Dumort. (≡ Jungermannia sphagni Dicks.).

= Anomoclada Spruce in J. Bot. 14 [= n.s. 5]: 133. 1876, syn. nov. – Type: Anomoclada mucosa Spruce (= Odontoschisma porto ricense (Hampe & Gottsche) Steph.).

I. Odontoschisma sect. OdontoschismaBranching exclusively ventral-intercalary. Leaves undi-

vided. Leaf margins bordered. Mid-leaf cells 10–30 µm long, with trigones. Cuticle smooth or papillose. Gemmae lacking. Apices of female bracts long acuminate. Antheridial jacket cells irregularly arranged.

1. Odontoschisma sphagni (Dicks.) Dumort., Recueil Observ. Jungerm.: 19. 1835 ≡ Jungermannia sphagni Dicks., Fasc. Pl. Crypt. Brit. [1]: 6. 1785 – Holotype: England, Surrey, “in palustribus, Sphagno palustri frequenter adhaerens prope Croydon Surriense” (BM n.v. [coll. Dickson; fide Grolle, 1976]; isotypes: S Nos. B29108!, B29109!, B29110!, B29111!, UPS-THUNB n.v. [fide Grolle, 1976]).

= Odontoschisma prostratum (Sw.) Trevis. in Mem. Reale Ist. Lombardo Sci., Ser. 3, Cl. Sci. Mat. 4: 419. 1877 ≡ Jungermannia prostrata Sw., Prodr.: 143. 1788 – Holo-type: JAMAICA. “ad radices arborum in montibus”, Swartz s.n. (S No. B98897!; isotypes: S Nos. B98896! [hb. Lehmann] & B98895! [hb. Lehmann ex hb. Weber], STR! [coll. Nees]).

2. Odontoschisma grosseverrucosum Steph. in Bull. Herb. Boissier, sér. 2, 8: 593. 1908; Sp. Hepat 3: 377. 1908 – Lecto-type (designated by So in J. Hattori Bot. Lab. 95: 252. 2004): TAIWAN. Mt. Taitum, Faurie 22 (G No. 000640 [barcode G00112865]!; isolectotype: BM).

1017



II. Odontoschisma sect. Iwatsukia (N.Kitag.) Gradst., S.C. Aranda & Vanderp. in Phytotaxa 162: 232. 2014 ≡ Iwatsukia N.Kitag. in J. Hattori Bot. Lab. 27: 178. 1964 – Type: Iwat-sukia exigua N.Kitag. (= Odontoschisma jishibae (Steph.) L.Söderstr. & Váňa).Branching exclusively ventral-intercalary. Leaves bifid,

apices acute to long acuminate. Leaf margins unbordered. Mid-leaf cells 10–30 µm long, walls evenly thickened, trigones absent. Cuticle along the middle lamella of the cell walls densely and finely papillose. Gemmae lacking. Apices of female bracts long acuminate. Arrangement of antheridial jacket cells unknown.

3. Odontoschisma jishibae (Steph.) L.Söderstr. & Váňa in Phyto taxa 112: 13. 2013 ≡ Cephalozia jishibae Steph., Sp. Hepat. 6: 437. 1924 ≡ Iwatsukia jishibae (Steph.) N.Kitag. in Acta Phytotax. Geobot. 21: 114. 1965 – Lectotype (des-ignated here): JAPAN. Shinano, Mt. Yaku, July 1908, Jishiba 203 (G barcode G00061127!; isolectotype: JE!).

= Iwatsukia exigua N.Kitag. in J. Hattori Bot. Lab. 27: 178. 1964 (fide Gradstein & al., 2014) – Holotype: MALAYSIA. Sabah, S slope of Mt. Kinabalu below Paca Cave, Iwatsuki 1070 (NICH!; isotype: JE!).

4. Odontoschisma bifidum (Fulford) Gradst., S.C.Aranda & Vanderp. in Phytotaxa 162: 232. 2014 ≡ Cladomastigum bifidum Fulford in Acta Bot. Venez. 2: 80. 1967 ≡ Iwat-sukia bifida (Fulford) R.M.Schust. in Bull. Natl. Sci. Mus. Tokyo 11: 314. 1968 – Holotype: VENEZUELA. Bolivar, Auyan-tepui, along Río Churún, SE of the “sec-ond wall”, 1690 m, 21 May 1963, Steyermark 93300 (NY barcode 00713605!).

*5. Odontoschisma spinosum (Fulford) Gradst., S.C.Aranda & Vanderp. in Phytotaxa 162: 232. 2014 ≡ Cladomastigum spinosum Fulford in Mem. New York Bot. Gard. 23: 840. 1972 ≡ Iwatsukia spinosa (Fulford) R.M.Schust. in Trop. Bryol. 2: 249. 1990 – Holotype: VENEZUELA. Bolivar, summit of Meseta de Jáua, Steyermark 98009 (CINC n.v.).

III. Odontoschisma sect. Neesia Gradst., S.C.Aranda & Van-derp., sect. nov. – Type: Odontoschisma fluitans (Nees) L.Söderstr. & Váňa (≡ Jungermannia fluitans Nees).Branching exclusively ventral-intercalary. Leaves bifid,

apices rounded. Leaf margins unbordered. Mid-leaf cells large, 30–60 µm long, without trigones. Cuticle smooth. Gemmae lacking. Apices of female bracts rounded to acute. Antheridial jacket cells arranged in tiers.

6. Odontoschisma fluitans (Nees) L.Söderstr. & Váňa in Phytotaxa 112: 12. 2013 ≡ Jungermannia fluitans Nees in Funck, Krypt. Gew. Fichtelgeb. 29: specimen 593. 1823 ≡ Cladopodiella fluitans (Nees) Jørg. in Bergens Mus. Skrift., n.s., 16: 276. 1934 – Holotype: CZECH REPUB-LIC. Riesengebirge, Schneekoppe, “in stehenden Wässern der Weißwiese”, Funck s.n. (STR!).

IV. Odontoschisma sect. Cladopodiella (H.Buch) Gradst., S.C.Aranda & Vanderp., comb. & stat. nov. ≡ Cladopo-diella H.Buch in Memoranda Soc. Fauna Fl. Fenn. 1: 89. 1927 – Type: Odontoschisma francisci (Hook.) L.Söderstr. & Váňa (≡ Jungermannia francisci Hook.).Branching exclusively ventral-intercalary. Leaves bifid,

apices subacute. Leaf margins unbordered. Mid-leaf cells 20–30 µm long, with small trigones. Cuticle smooth. Gemmae present. Apices of female bracts narrowly rounded to subacute. Antheridial jacket cells arranged in tiers.

7. Odontoschisma francisci (Hook.) L.Söderstr. & Váňa in Phytotaxa 112: 12. 2013 ≡ Jungermannia francisci Hook., Brit. Jungermann.: pl. 49. 1816 ≡ Cladopodiella francisci (Hook.) Jørg. in Bergens Mus. Skr., n.s., 16: 274. 1934 – Lectotype (designated by Váňa & al. in Phytotaxa 112: 12. 2013): England, Hampshire, 1812, Lyell s.n. (BM barcode BM001146107!).

V. Odontoschisma sect. Denudata R.M.Schust., Hepat. Antho-cerotae N. Amer. 3: 833. 1974 (“Denudatae”) – Type: Odonto schisma denudatum (Nees) Dumort. (≡ Junger-mannia denudata Nees).

= Odontoschisma sect. Macounia R.M.Schust., Hepat. Antho-cerotae N. Amer. 3: 848. 1974 (“Macouniae”), syn. nov. – Type: Odontoschisma macounii (Austin) Underw. (≡ Sphagnoecetis macounii Austin).Branching lateral-intercalary and ventral-intercalary.

Leaves undivided or emarginate. Leaf margins unbordered, occasionally bordered. Mid-leaf cells 10–40 µm long, with trigones. Cuticle smooth or papillose. Gemmae usually present. Apices of female bracts rounded to acute to short acuminate. Antheridial jacket cells irregularly arranged.

8. Odontoschisma brasiliense Steph., Sp. Hepat. 3: 369. 1908 – Lectotype (designated here): BRAZIL. Rio de Janeiro, Glaziou 11760 (G barcode G00112855!; isolectotypes: BM!, NY!).

9. Odontoschisma cleefii Gradst., S.C.Aranda & Vanderp., sp. nov. – Holotype: COLOMBIA. Cundinamarca, páramo Cruz Verde, laguna El Verjón, on humid soil in grass páramo, 3300 m, 26 May 2012, Laura V. Campos 717 (COL!; isotype: PC barcode 0703249!). — For drawings from the holotype, see Fig. 3.Characterized by purplish color, concave leaves covered

by large papillae, frequent presence of a whitish leaf border of collapsed cells, occurrence of gynoecia in upper portions of shoots and broadly rounded to somewhat obtuse apices of female bracts and bracteoles.

Plants prostrate to ascending, usually purplish in color, older stem portions colorless, growing in small mats or creeping loosely among other bryophytes. Leafy shoots 0.8–1.5 mm wide, leafy branches ventral-intercalary or ventrolateral-intercalary, stolons frequent, short or long, sometimes branched. Stems 0.15–0.2 mm in diameter, in cross section made up of 18–20 thick-walled epidermal cells surrounding larger and almost

1018



thin-walled medullary cells; dorsal epidermis cells in surface view quadrate to shortly rectangular, narrowly and evenly thick-ened arranged in straight rows. Leaves concave, contiguous to imbricate, lamina standing upwards to spreading, insertion line reaching dorsal midline, dorsal leaf-free strip absent; leaves shortly ovate to subquadrate to suborbicular, ca. 0.5–0.7 mm long, about as wide as long to slightly longer than wide, slightly asymmetrical, apex rounded to truncate, dorsal margin curved, dorsal base shortly decurrent, ventral margin curved and dis-tinctly narrowed at the base, leaf sometimes with a whitish bor-der. Leaf cells isodiametrical-hexagonal to slightly elongate, 20–30 µm long in mid-leaf, trigones medium-sized to large, light brown to violet, darker pigmented than lumen, lumen rounded to stellate, middle lamella indistinct; leaf margin cells sometimes whitish-decolorate, especially along ventral and apical margin, with a conspicuously thickened outer wall (5–12 µm thick), form-ing a whitish leaf border; oil-bodies not observed; cuticle covered by large, low, colorless papillae, papillae rounded to slightly elongate, 4–15 µm × 4–10 µm, 1–10 per cell, present all over leaf surface but most prominent in lower half of leaf. Underleaves

rudimentary, broad scale-like, sometimes bifid with 2 short, narrowly lanceolate lobes, with a few slime cells at margins and on surface. Gemmae present or lacking, pale green, 1–2-celled, ± thick-walled, produced in whitish clusters on tips of stout, upright, flagelliform shoots with appressed, scale-like leaves and underleaves. Androecia very small, in whitish catkin-like spikes, with 4–10 pairs of bracts. Gynoecia on short-ventral branches in upper portion of stem, whitish to violet, bracts and bracteoles in 4 series, ± free, inner bracts 1.2 × 0.6 mm, unequally bifid to 1/3, apex of longest lobe obtuse, of shortest lobe broadly rounded, margins entire or with 1–2 remote laciniate teeth, surface of bracts low substriate-papillose by large papillae, margins with a white border of thicker-walled cells; inner bracteole as long as bracts, very shortly bifid and with rounded lobes, margins entire or with a short, laciniate tooth. Perianth large, cylindrical, 5 mm long, uniformly deep violet in color, cells smooth, mouth crenate. Sporophyte not observed.

Etymology. – The new species is dedicated to Professor Antoine M. Cleef, world specialist on the flora and vegetation of the páramo, who made numerous collections of the new species.

Distribution and ecology. – Only known from the high Andes of Colombia where it is common in páramo vegetation of the eastern, central and western cordilleras, at elevations between 2300 and 4000 m, growing on humid soil in grass páramo, bamboo páramo and in Sphagnum bogs. In the eastern cordillera the species was exclusively found above 3300 m, in the central and western cordilleras also at lower elevation in azonal páramo vegetation and on low summits (e.g., Chocó: summit of Cerro del Torrá, 2770 m. Huila: páramo del Río la Candelaria, 2300 m).

Further specimens. – COLOMBIA. Arauca: Sierra Nevada del Cocuy, Quebrada El Playon, 3350 m, Cleef 9178 (U). Boyacá: Vado Hondo, páramo between Peña Arnical y Alto de Mogetes, Cleef 9245, 9258, 9489 (U); páramo de la Sarna entre Sogamoso and Vado Hondo, Cleef 9523a (U); Péna de Arnical N de Vado Hondo, Cleef 9489 (U); páramo de la Rusia, 3820 m, Cleef 7476 (U). Casanare: páramo de Pisba, 3500 m, Aguirre & al. 2905 (COL, U). Chocó: San José del Palmar, Cerro del Torrá, ca. 2750 m, Silverstone-Sopkin & al. 1830 (U). Cundinamarca: páramo de Palacio, headwaters of Río Negro, 3360 m, Cleef 3819 (U), ibid., 1.5 km S of Lagunas de Buitrago, 3560 m, Cleef 4127 (COL, U); páramo Cruz Verde, 3325–3600 m, Cleef 2885, 3086, 3243, 3304 (COL, U), 3089, 3244, 3277 (U); páramo between Cogua and San Cayetano, 3660–3700 m, Cleef 6248, 6269 (COL, U), 6221, 6372, 6497 (U); páramo de Chingaza, Santana & Aguirre 337, 358, 381, 382, 388 (COL); páramo de Sumapaz, 3500–4000 m, Cleef 1637a, 8403 (U). Huila: La Plata, headwaters of Río La Candelaria, 2300 m, Aguirre & al. 6556 (COL, U), ibid., Van Zanten & al. 815, as O. atropurpureum (Van Zanten & Gradstein, 1988) (U). Meta: páramo de Sumapaz, 3600 m, Cleef 1197 (COL).

10. Odontoschisma denudatum (Nees) Dumort., Recueil Observ. Jungerm.: 19. 1835 ≡ Jungermannia denudata Nees in Martius, Fl. Crypt. Erlang.: XIV. 1817 – Holotype: GER-MANY. Erlangen, Martius s.n., c. gemmis (STR! [coll. Nees]; isotype (?): S No. B73168!).

Fig. 3. Odontoschisma cleefii (from the holotype). A, habit, dorsal view; B, stem cross section; C, habit, ventral view; D, dorsal leaf-free strip; E, leaf; F, mid-leaf cells; G, leaf margin. — Scale bars: A, C = 500 µm; B = 50 µm; D = 200 µm; E = 250 µm; F, G = 25 µm. — Drawn by A.-L. Ilkiu-Borges.

1019



10a. Odontoschisma denudatum subsp. denudatum= Odontoschisma subjulaceum Austin in Bull. Torrey Bot. Club

6: 303. 1879, syn. nov. – Lectotype (designated by So in J. Hattori Bot. Lab. 95: 257. 2004): [U.S.A.] Hawaii, Maui, on soil, 2600 m, Dec 1875, Baldwin 233, hb. Levier 1439 (G No. 20698!; isolectotypes: FH!, NY!).

10b. Odontoschisma denudatum subsp. naviculare (Steph.) Gradst., S.C.Aranda & Vanderp., comb. & stat. nov. ≡ Jamesoniella navicularis Steph., Sp. Hepat. 6: 101. 1917 ≡ Odontoschisma naviculare (Steph.) Grolle in Acta Bot. Fenn. 125: 71. 1984 – Holotype: New Caledonia, Mt. Dogny, 1050 m, Jul 1909, Mme L. Le Rat 208 (G No. 13519 [barcode G00128131]!; isotypes: L!, PC!).

10c. Odontoschisma denudatum subsp. sandvicense (Ångstr.) Gradst., S.C.Aranda & Vanderp., comb. & stat. nov. ≡ Sphagnoecetis sandvicensis in Öfvers. Kongl. Vetensk.-Akad. Förh. 29(4): 22. 1872 (“Sphagnaectis”) ≡ Odon-toschisma sandvicense (Ångstr.) A.Evans in Trans. Con-necticut Acad. Arts Sci. 8: 256. 1892 (“1891”) – Holotype: [U.S.A.] Hawaii, Honolulu, Jun 1852, Andersson s.n. (S No. B47419; isotypes: G No. 22402 [barcode G00265225]!, S No. B47418!). So (2004) designated B47419 in S as the lectotype of

O. sandvicense. However, since this is the only specimen from the Ångström herbarium (its duplicate, S-B47418, is from the Möller hb.), since Ångström holotypes are in S and since the label is in Angström’s handwriting, it should be considered the holotype.

11. Odontoschisma elongatum (Lindb.) A.Evans in Rhodora 14: 13. 1912 ≡ O. denudatum f. elongatum Lindb. in Helsing-fors Dagblad 45: 2. 1874 – Lectotype (designated by Grolle in Trans. Brit. Bryol. Soc. 6: 263–64. 1971): SWEDEN. Lappland, Lycksele, Ångström s.n., Hep. Eur. Exsicc. 440 (ed. Gottsche & Rabenhorst) as Sphagnoecetis communis (H-SOL!; isolectotypes: G!, H!, JE!, S!, etc.).

12. Odontoschisma engelii Gradst. & Burghardt in Fieldi-ana, Bot., n.s., 47: 194. 2008 – Holotype: ECUADOR. Loja, Cordillera Oriental, Parque Nacional Podocarpus, Cajanuma, lower margin of páramo, 3150 m, on the verti-cal, slightly overhanging and partially shaded side of a large, crystalline boulder in the páramo, at the junction of the trail leading to the lakes and the “mirador”, forming dense, low, brownish mats, 19 Sep 2006, Gradstein & al. 10178 (GOET!; isotype: QCA!).

13. Odontoschisma longiflorum (Taylor) Trevis. in Mem. Reale Ist. Lombardo Sci., Ser. 3, Cl. Sci. Mat. 4: 419. 1877 ≡ Sphagnoecetis longiflora Taylor in London J. Bot. 5: 281. 1846 – Holotype: JAMAICA. s. coll. (FH n.v.; isotypes: S Nos. B47414! & B47415!).

14. Odontoschisma macounii (Austin) Underw. in Bull. Illinois State Lab. Nat. Hist. 2: 92. 1884 ≡ Sphagnoecetis macounii

Austin in Bull. Torrey Bot. Club 3: 13. 1872 – Holotype: CANADA. Ontario, Lake Superior, 25 miles north of Michepicoten and near Otter Head, Macoun s.n. (MANCH No. 20340 n.v. [cf. Grolle, 1976]).

15. Odontoschisma portoricense (Hampe & Gottsche) Steph. in Hedwigia 27: 269. 1888 ≡ Sphagnoecetis portoricensis Hampe & Gottsche in Linnaea 25: 343. 1853 (“1852”) ≡ Anomoclada portoricensis (Hampe & Gottsche) Váňa, in Bryologist 92: 344. 1989 – Holotype: Puerto Rico, Schwanecke s.n. (BM!).

16. Odontoschisma pseudogrosseverrucosum Gradst., S.C. Aranda & Vanderp., sp. nov. – Holotype: JAPAN. Honshu: Hiroshima Province, Hiroshima-shi, Aki-ku, Ayno-cho, Mt. Egesan, 35°19′10″ N, 132°32′22″ E, ca. 550–570 m, on rock, 15 Apr 2012, T. Katagiri 3525, c. gyn. (PC barcode PC0703247!; isotype: HIRO Nr. 1015793!). — For drawings from the holotype, see Fig. 4.Differs from O. grosseverrucosum by smaller plant size,

Fig. 4. Odontoschisma pseudogrosseverrucosum (from the holotype). A, habit, dorsal view; B, leaf apex; C, habit, ventral view; D, mid-leaf cells; E, stem cross section; F, underleaf; G, leaf; H, portion of shoot showing dorsal leaf-free strip. — Scale bars: A, C, H = 500 µm; B, D = 25 µm; E, F = 50 µm; G = 250 µm. — Drawn by A.-L. Ilkiu-Borges.

1020



subquadrate dorsal epidermis cells with thick walls, concave, unbordered leaves, and occasional presence of lateral branches and gemmae.

Plants prostrate, green to dark green to brown, in flat mats, small, well-developed leafy shoots 0.5–0.65(–0.8) mm wide, 1–1.5 cm long, little and irregularly branched, leafy branches ventral-intercalary and occasionally lateral-intercalary (in material from Russian Far East), ventral stolons frequent, stems sometimes terminating in a long stolon or microphyllous shoot. Stems ca. 0.15 mm in diameter, stem surface smooth, epider-mis cells thick-walled (rarely thin-walled), small, quadrate to subrectangular, 15–25 µm long × 12–20 µm wide. Leaves undi-vided, distinctly concave with margins upcurved, obliquely to widely spreading, distant to imbricate, insertion line oblique to substransverse, not extending to dorsal midline, dorsal leaf-free strip 1–4 cells wide; leaves ovate, small, ca. 0.35 mm long × 0.25 mm wide when well developed, often smaller, slightly longer than wide, subsymmetrical with ventral margin slightly more strongly curved than dorsal margin, apex broadly rounded to asymmetrically emarginate, ventral base narrowed, dor-sal base narrowed or not. Leaf cells small, 10–20 µm long in mid-leaf, isodiametrical to somewhat elongate, slightly larger to base, 20–25 µm, not smaller to margin; trigones small to large, cell-lumen varying from quadrate to stellate, middle lamella distinct or indistinct; leaf border absent; oil-bodies rounded to elongate, finely granular-papillose, 2–4 per cell, rather small, 4–9 × 4–6 µm; cuticle covered by large, rounded to elongate papillae in lower half of leaf lamina (in large leaves) or over whole leaf lamina (in small leaves), papillae 3–10(–15) × 3–10 µm, 1–5(–8) per cell. Underleaves vestigial or rather well-developed, lanceolate to subquadrate, to 0.15 × 0.1 mm, apex undivided or shortly bifid, margins sometimes with a tooth; slime papillae not seen. Gemmae present, scarce, pale green, ovate, 2-celled, 20–25 × 13–16 µm, produced in dense pale-green clusters on upright, swollen shoots apices or at tips of short microphyllous branches, gemmiparous leaves densely imbricate and appressed, underleaves on gemmipa-rous shoots much smaller than leaves. Dioicous. Androecia not seen. Gynoecia whitish, on short-specialized ventral branches, bracts and bracteoles in 3 series, made up of smooth cells; inner bracts ovate, concave, shallowly and unequally bifid, apices acute, margins entire to irregularly crenate; inner bracteole as long as bracts, apex rounded or shortly bifid. Perianths rare, cylindrical, bluntly trigonous. Sporophytes not seen.

Distribution and ecology. – Temperate and subtropical East Asia: Eastern Russia, Korea and Japan (Honshu, Kyushu, southwards to Yakushima Island), ca. 100–1650 m. On moist limestone rock and cliffs, also on decaying wood, in evergreen forest.

Further specimens (all originally annotated as “Odonto-schisma grosseverrucosum”). – RUSSIA. Far East: Khasan-sky Distr., Kedrovaya Pad’ State Reserve, Bakalin P-3-26-7 (PC, VLA). SOUTH KOREA. Kyong Nam: Chiri Mt., Bakalin Kor-12-3b-09, Kor-12-10b-9 (PC, VLA). JAPAN. Honshu: Segari, Kobe-shi, Hyogo-ken, Kodama s.n., Hep. Japon. Exsicc. 1036 (ed. S. Hattori & M. Mizutani) (GOET, L, PC); Nagano-ken, Shiokawa valley, Katagiri 889 (HIRO); Hatsukaichi-shi,

Miyajima I., Y. Inoue 174 (HIRO, PC). Kyushu: Miyazaki, Faurie 1296 (G, syntype of O. grosseverrucosum); Kuma-moto, Hitoyohi, Mayebara s.n., Hep. Japon. Exsicc. 108 (ed. S. Hattori) (BM, GOET, PC, S, U); Yakushima, ca. 1500 m, Takaki & Mizutani s.n., Hep. Japon. Exsicc. 793 (ed. S. Hattori) (PC, U).

*17. Odontoschisma purpuratum Herzog in Trans. Brit. Bryol. Soc. 1(4): 297. 1950 – Holotype: MALAYSIA. Sarawak, ridge of Mt. Dulit, on tree trunk near ground, in mossy for-est on exposed peak, ca. 1400 m, 5 Oct 1932, P.W. Richards 2162, c. andr. (JE!; isotypes: BM!, L!).

*18. Odontoschisma soratamum Fulford in Mem. New York Bot. Gard. 11: 338. 1968 – Holotype: COLOMBIA. Ama-zonas-Vaupes, Rio Apoporis, Soratama, 250 m, Schultes & Cabrera 12883 (FH!).

19. Odontoschisma variabile (Lindenb. & Gottsche) Trevis. in Mem. Reale Ist. Lombardo Sci., Ser. 3, Cl. Sci. Mat. 4: 419. 1877 ≡ Sphagnoecetis variabilis Lindenb. & Gottsche in Gottsche & al., Syn. Hepat.: 688. 1847 – Holotype: MEXICO. Oaxaca, Tepinapa, “ad terram”, Liebmann 315a (W n.v.; isotypes: C!, S Nos. B47422! & B47423!).

20. Odontoschisma zhui Gradst., S.C.Aranda & Vanderp., sp. nov. – Holotype: CHINA. Zhejiang Prov., Longquan City, Fengyangshan Nature Reserve, Yangtiancao, ca. 1500 m, on moist sandstone rock in evergreen forest in deep river valley, in shade, 3 Dec 2012, S.R. Gradstein & R.-L. Zhu 12384 (PC barcode PC0703248!; isotype: HSNU!). — For drawings from the holotype, see Fig. 5.Characterized by leaves tending to become curved-subfal-

cate, small leaf cells (13–22 µm long in mid-leaf), and a smooth cuticle. Moreover, branching is predominantly ventral-interca-lary and the leaves and underleaves of specialized gemiparous shoots are appressed.

Plants prostrate, light green to dark green to reddish-brown or purplish, forming loose to dense mats, leafy shoots 1–1.3 mm wide, leafy branches usually ventral-intercalary, rarely ventro-lateral-intercalary, ventral stolons simple or branched, present near base of stem, occasionally higher up stem, leafy shoot sometimes terminating in a long stolon. Stems ca. 0.15–0.16 mm in diameter, rigid, epidermis composed of small, quadrate to rectangular, thick-walled cells, dorsal epidermis cells in almost straight rows. Leaves concave with margins ± upcurved (espe-cially dorsal margin), obliquely to widely spreading, becoming curved-subfalcate, subimbricate, insertion line usually reach-ing dorsal midline, dorsal leaf-free strip 0(–1) cell wide; leaves shortly ovate, ± asymmetrical, 0.5–0.8 × 0.5–0.6.5 mm, as wide as long or longer than wide (1–1.3 : 1), apex rounded, dorsal mar-gin almost straight to curved and slightly narrowed at base, dor-sal base not or very slightly decurrent, ventral margin broadly arched and narrowed at base, margins entire or slightly crenulate by projecting margin cells. Leaf cells small, isodiametrical to elongate, in mid-leaf 13–18(–22) µm long and 11–18 µm wide, cells slightly smaller towards leaf margin and slightly larger towards

1021



leaf base; trigones small to large and swollen, contiguous, cell-lumen rounded to stellate, middle lamella indistinct; leaf border lacking; oil-bodies not seen; cuticle smooth or slightly verrucu-lose. Underleaves lacking (except on gemmiparous shoots and in gametoecia). Gemmae present, whitish-green, 1–2-celled, rounded to ellipsoid, thick-walled, walls ca. 2–8 µm thick, pro-duced on tips of upright, flagelliform shoots with appressed to obliquely spreading leaves and underleaves, underleaves becom-ing progressively larger towards apex of gemmiparous shoot. Dioicous. Androecia very small, in short, whitish, catkin-like spikes near stem base, with 4–5 pairs of bracts. Gynoecia on short ventral branches near stem base, whitish-green, bracts and bracteoles in 2 series, inner bracts unequally bifid, apex of lobes acute, margins entire or remotely crenate, surface of bracts finely striate-papillose; inner bracteole about as long as bracts, bifid and with acute lobes. Perianth cylindrical, ± 3 mm long, whitish above, green below, deeply plicate, cells smooth, mouth contracted, crenate. Sporophyte not seen.

Etymology. – The new species is dedicated to Professor Rui-Liang Zhu, specialist of Lejeuneaceae and liverworts of China.

Distribution and ecology. – Subtropical East Asia (central and southeastern China, southern Japan), from near sea level to about 1600 m. On sandstone rock and rotten wood in deep river valleys, in subtropical evergreen forest.

Further specimens. – CHINA. Guanxsi: Maoershan Nature Reserve, Longtrangjiang, 480 m, Zhu & al. 20040908-47 as O. grosseverrucosum (HSNU). Huan: Yizhang Co., Mt. Mangshan, Koponen & al. 50768 as O. denudatum (GOET, H), 51488, 55589 as O. denudatum (H, S); Yanling Co., Taoyu-andong, Koponen & al. 55308 as O. grosseverrucosum (H, TNS), Zhejiang: Fengyangshan Nature Reserve, Zhu & Wei 20110418-37 as O. denudatum (HSNU), Gradstein & Zhu 12385, 12386 (c. andr.), 12387 (c. gyn.) (HSNU, PC). JAPAN. Honshu: Nigata Pref., Sado I., Homma 4580 as O. grosseverruco-sum (NICH); Toyama Pref., Kurobe valley, Mizutani s.n. as O. grosse verrucosum (NICH); Fukui Pref., Mt. Fujikura, Mizutani 4373 as O. grosseverrucosum (NICH); Shiga Pref., Kanzaki-gun, Echigawa valley, Kodama 27161 as O. grosse-verrucosum (NICH). Shikoku: Kochi Pref., Kitagawa-mura, Hashimoto & al. 1603002 as O. grosseverrucosum (NICH), ibid., Tai-mura, H. Inoue 643 as O. denudatum (TNS). Kyushu: Yakushima I., Hattori 6972, 7677 as O. grosseverrucosum forma (NICH).

ACKNOWLEDGEMENTS

The authors thank three anonymous referees for their construc-tive comments on the manuscript. Thanks are also due to the curators of the herbaria cited in the text for the loan of specimens. We further-more express our warmest gratitude to Anna Luiza Ilkiu-Borges for preparing the drawings of the three new species of Odontoschisma and to the following colleagues who kindly sent us samples for the molecular study: V.A. Bakalin (Russia), L.V. Campos (Colombia), P.G. Davidson (U.S.A.), K. Georgson (Sweden), T. Hallingbäck (Swe-den), T. Katagiri (Japan), J. Larrain (Chicago), E. Lavocat-Bernard (Guadeloupe), H. van Melick (The Netherlands), D. Pinheiro da Costa (RB), B. Shaw (U.S.A.), T. Tyler (Sweden), Yong K.T. (Malaysia) and R.-L. Zhu (China). The second author (SRG) is very grateful to L. Ellis (BM), L. Hedenäs and I. Bisang (S), M. Hoff (STR), S. Léon-Yánez (QCA), M. Price (G), M. Stech (L) and J. Uribe (COL) for making facil-ities available and help during his visits to these herbaria, to T. Katagiri and R.-L. Zhu for help with literature, and to R.-L. Zhu for hosting his visit to China. Work of SRG in the Stockholm herbarium and of SCA, BL and AD at Leiden University was supported by Synthesys grants. SCA was supported by a Spanish JAEPre grant from CSIC. AV acknowledges financial support from the Fonds Léopold III.

LITERATURE CITED

Barley, A.J., White, J., Diesmos, A.C. & Brown, R.M. 2013. The challenge of species delimitation at the extremes: Diversification without morphological change in Philippine sun skinks. Evolution 67: 3556–3572. http://dx.doi.org/10.1111/evo.12219

Bergsten, J., Nilsson, A.N. & Ronquist, F. 2013. Bayesian tests of

Fig. 5. Odontoschisma zhui (A, D–F, I, J, L from the holotype; B, C, G, H, K, M from Gradstein & Zhu 12387). A, habit, ventral view; B, habit, dorsal view; C, ibid., with gynoecia; D–E, underleaves; F–G, stem cross sections; H–I, leaf margin; J, leaf; K–L, mid-leaf cells; M, gemmiparous shoot, dorsal view. — Scale bars: A, B, M = 500 µm; C, J = 250 µm; D–G = 50 µm; H, I, K, L = 25 µm. — Drawn by A.-L. Ilkiu-Borges.

1022



topology hypotheses with an example from diving beetles. Syst. Biol. 62: 660–673. http://dx.doi.org/10.1093/sysbio/syt029

Brewer, M.S., Spruill, C.L., Rao, N.S. & Bond, J.E. 2012. Phyloge-netics of the millipede genus Brachycybe Wood, 1864 (Diplopoda: Platydesmida: Andrognathidae): Patterns of deep evolutionary history and recent speciation. Molec. Phylogen. Evol. 64: 232–242.

http://dx.doi.org/10.1016/j.ympev.2012.04.003Chang, Y., Li, J., Lu, S. & Schneider, H. 2013. Species diversity and

reticulate evolution in the Asplenium normale complex (Aspleni-aceae) in China and adjacent areas. Taxon 62: 673–687.

http://dx.doi.org/10.12705/624.6Crandall, K.A., Bininda-Emonds, O.R.P., Mace, G.M. & Wayne,

R.K. 2000. Considering evolutionary processes in conservation biology. Trends Ecol. Evol. 15: 290–295.

http://dx.doi.org/10.1016/S0169-5347(00)01876-0Crandall-Stotler, B. 1969. Morphology and development of branches

in the leafy Hepaticae. Beih. Nova Hedwigia 30: 1–261.Crandall-Stotler, B., Stotler, R. & Long, D. 2009. Phylogeny and clas-

sification of the Marchantiophyta. Edinburgh J. Bot. 66: 155–198. http://dx.doi.org/10.1017/S0960428609005393Darriba, D., Taboada, G.L., Doallo, R. & Posada, D. 2012. jModelTest

2: More models, new heuristics and parallel computing. Nature, Meth. 9: 772–772. http://dx.doi.org/10.1038/nmeth.2109

Dong, S., Schäefer-Verwimp, A., Meinecke, P., Feldberg, K., Bom-bosch, A., Pócs, T., Schmidt, A.R., Reitner, J., Schneider, H. & Heinrichs, J. 2012. Tramps, narrow endemics and morphologi-cally cryptic species in the epiphyllous liverwort Diplasiolejeunea. Molec. Phylogen. Evol. 65: 582–594.

http://dx.doi.org/10.1016/j.ympev.2012.07.009Drummond, A.J., Suchard, M.A., Xie, D. & Rambaut, A. 2012.

Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molec. Biol. Evol. 29: 1969–1973. http://dx.doi.org/10.1093/molbev/mss075

Evans, A.W. 1903. Odontoschisma macounii and its North American allies. Bot. Gaz. 36: 321–348. http://dx.doi.org/10.1086/328416

Ezard, T., Fujisawa, T. & Barraclough, T.G. 2009. SPLITS: SPecies’ LImits by Threshold Statistics. R package version 1.0-18/r45. http://R-Forge.R-project.org/projects/splits/

Feldberg, K., Váňa, J., Long, D.G., Shaw, J., Hentschel, J. & Hein-richs, J. 2010. A phylogeny of Adelanthaceae (Jungermanniales, Marchantiophyta) based on nuclear and chloroplast DNA markers, with comments on classification, cryptic speciation and biogeog-raphy. Molec. Phylogen. Evol. 55: 293–304.

http://dx.doi.org/10.1016/j.ympev.2009.11.009Forrest, L.L., Davis, E.C., Long, D.G., Crandall-Stotler, B.J., Clark,

A. & Hollingsworth, M.L. 2006. Unraveling the evolutionary his-tory of the liverworts (Marchantiophyta): Multiple taxa, genomes and analyses. Bryologist 109: 303–334.

http://dx.doi.org/10.1639/0007-2745(2006)109[303:UTEHOT]2.0.CO;2

Frahm, J.-P. & Vitt, D. 1993. Comparisons between the mossfloras of North America and Europe. Nova Hedwigia 56: 307–333.

Frey, W. & Stech, M. 2009. Marchantiophyta, Bryophyta, Anthoceroto-phyta. Pp. 1–257 in: Frey, W. (ed.), Syllabus of plant families – A. Engler’s Syllabus der Pflanzenfamilien, part 3, Bryophytes and seedless vascular plants, 13th. ed. Stuttgart: Borntraeger.

Fujisawa, T. & Barraclough, T.G. 2013. Delimiting species using single-locus data and the generalized mixed Yule coalescent approach: A revised method and evaluation on simulated data sets. Syst. Biol. 62: 707–724. http://dx.doi.org/10.1093/sysbio/syt033

Fujita, M.K., Leache, A.D., Burbrink, F.T., McGuire, J.A. & Moritz, C. 2012. Coalescent-based species delimitation in an integrative taxonomy. Trends Ecol. Evol. 27: 480–488.

http://dx.doi.org/10.1016/j.tree.2012.04.012Fulford, M.H. 1968. Manual of the leafy Hepaticae of Latin America.

Mem. New York Bot. Gard. 11: 277–392.Funk, D.J. & Omland, K.E. 2003. Species-level paraphyly and poly-

phyly: Frequency, causes, and consequences, with insights from

animal mitochondrial DNA. Annual Rev. Ecol. Evol. Syst. 34: 397–423. http://dx.doi.org/10.1146/annurev.ecolsys.34.011802.132421

Gao, C. & Cao, T. 2000. Flora Yunnanica, vol. 17. Beijing: Science Press.

Gouy, M., Guindon, S. & Gascuel, O. 2010. SeaView Version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Molec. Biol. Evol. 27: 221–224.

http://dx.doi.org/10.1093/molbev/msp259Gradstein, S.R. 2013. Afro-American hepatics revisited. Polish Bot. J.

58: 149–177.Gradstein, S.R. & Costa, D.P. 2003. The Hepaticae and Anthocerotae

of Brazil. Mem. New York Bot. Gard. 86: 1–317.Gradstein, S.R. & Ilkiu-Borges, A.L. In press. A taxonomic revision

of the genus Odontoschisma (Marchantiophyta: Cephaloziaceae). Nova Hedwigia.

Gradstein, S.R., Churchill, S.P. & Salazar-Allen, N. 2001. Guide to the bryophytes of tropical America. Mem. New York Bot. Gard. 86: 1–577.

Gradstein, S.R., Aranda, S.C. & Vanderpoorten, A. 2014. Notes of early landplants today: Transfer of Iwatsukia to Odontoschisma (Cephaloziaceae, Marchantiophyta). Phytotaxa 162: 232–233.

http://dx.doi.org/10.11646/phytotaxa.162.4.6Grolle, R. 1976. Verzeichnis der Lebernmoose Europas und benach-

barter Gebiete. Feddes Repert. 87: 171–279. http://dx.doi.org/10.1002/fedr.19760870303Groth, H., Helms, G. & Heinrichs, J. 2002. The systematic status of

Plagiochila sect. Bidentes Carl and Caducilobae Inoue (Hepaticae) inferred from nrDNA ITS sequences. Taxon 51: 675–684.

http://dx.doi.org/10.2307/1555022Hattori, S. 1950. Contributio ad Floram Hepaticarum Yakusimensem,

IV. J. Hattori Bot. Lab. 4: 49–70.Hattori, S. 1962. Hepaticae Japonicae Exsiccatae, ser. 1–12. Alphabeti-

cal list of species. J. Hattori Bot. Lab. 25: 163–185.Hedenäs, L., Désamoré, A., Laenen, B., Papp, B., Quandt,

D., González-Mancebo, J.M., Patiño, J., Vanderpoorten, A. & Stech, M. 2014. Three species for the price of one within the moss Homalothecium sericeum s.l. Taxon 63: 249–257.

http://dx.doi.org/10.12705/632.16Heinrichs, J., Lindner, M., Groth, H. & Renker, C. 2005a. Distribu-

tion and synonymy of Plagiochila punctata (Taylor) Taylor, with hypotheses on the evolutionary history of Plagiochila sect. Arrec-tae (Plagiochilaceae, Hepaticae). Pl. Syst. Evol. 250: 105–117.

http://dx.doi.org/10.1007/s00606-004-0220-5Heinrichs, J., Lindner, M., Gradstein, S.R., Groth, H., Buchbender,

V., Solga, A. & Fischer, E. 2005b. Origin and subdivision of Pla-giochila (Jungermanniidae: Plagiochilaceae) in tropical Africa based on evidence from nuclear and chloroplast DNA sequences and morphology. Taxon 54: 317–333.

http://dx.doi.org/10.2307/25065360Heinrichs, J., Lindner, M., Groth, H., Hentschel, J., Feldberg, K.,

Renker, C., Engel, J., von Konrat, M., Long, D. & Schneider, H. 2006. Goodbye or welcome Gondwana? – Insights into the phy-logenetic biogeography of the leafy liverwort Plagiochila with a description of Proskauera, gen. nov.(Plagiochilaceae, Jungerman-niales). Pl. Syst. Evol. 258: 227–250.

http://dx.doi.org/10.1007/s00606-006-0411-3Heinrichs, J., Hentschel, J., Bombosch, A., Fiebig, A., Reise, J.,

Edelmann, M., Kreier, H.-P., Schaefer-Verwimp, A., Caspari, S., Schmidt, A.R., Zhu, R.-L., von Konrat, M., Shaw, B. & Shaw, A.J. 2010. One species or at least eight? Delimitation and distribution of Frullania tamarisci (L.) Dumort. s.l. (Jungerman-niopsida, Porellales) inferred from nuclear and chloroplast DNA markers. Molec. Phylogen. Evol. 56: 1105–1114.

http://dx.doi.org/10.1016/j.ympev.2010.05.004Heinrichs, J., Kreier, H.-P., Feldberg, K., Schmidt, A.R., Zhu,

R.L., Shaw, B., Shaw, A.J. & Wissemann, V. 2011. Formalizing morphologically cryptic biological entities: New insights from

1023



DNA-taxonomy, hybridization and biogeography in the leafy liv-erwort Porella platyphylla (Jungermanniopsida, Porellales). Amer. J. Bot. 98: 1252–1262. http://dx.doi.org/10.3732/ajb.1100115

Heinrichs, J., Dong, S., Schäfer-Verwimp, A., Pócs, T., Feldberg, K., Czumaj, A., Schmidt, A.R., Reitner, J., Renner, M.A.M., Hentschel, J., Stech, M. & Schneider, H. 2013. Molecular phy-logeny of the leafy liverwort Lejeunea (Porellales): Evidence for a neotropical origin, uneven distribution of sexual systems and insufficient taxonomy. PLoS ONE 8: e82547.

http://dx.doi.org/10.1371/journal.pone.0082547Hentschel, J., Zündorf, H.-J., Hellwig, F.H., Schäfer-Verwimp, A.

& Heinrichs, J. 2007. Taxonomic studies in Chiloscyphus Corda (Jungermanniales: Lophocoleaceae) based on nrITS sequences and morphology. Pl. Syst. Evol. 262: 125–137.

http://dx.doi.org/10.1007/s00606-006-0463-4Hutsemekérs, V., Vieira, C.C., Ros, R.M., Huttunen, S. & Vander-

poorten, A. 2012. Morphology informed by phylogeny reveals unexpected patterns of species differentiation in the aquatic moss Rhynchostegium riparioides s.l. Molec. Phylogen. Evol. 62: 748–755. http://dx.doi.org/10.1016/j.ympev.2011.11.014

Katoh, K. & Standley, D.M. 2013. MAFFT Multiple sequence align-ment software version 7: Improvements in performance and usabil-ity. Molec. Biol. Evol. 30: 772–780.

http://dx.doi.org/10.1093/molbev/mst010Konstantinova, N.A. 2004. Iwatsukia jishibae (Steph.) Kitagawa

(Cephaloziaceae, Hepaticae) in Russia. Arctoa 13: 203–209. http://dx.doi.org/10.15298/arctoa.13.16Kreier, H.-P., Feldberg, K., Mahr, F., Bombosch, A., Schmidt, A.R.,

Zhu, R.-L., von Konrat, M., Shaw, B., Shaw, A.J. & Heinrichs, J. 2010. Phylogeny of the leafy liverwort Ptilidium: Cryptic specia-tion and shared haplotypes between the Northern and Southern Hemispheres. Molec. Phylogen. Evol. 57: 1260–1267.

http://dx.doi.org/10.1016/j.ympev.2010.10.002Mayr, E. 1963. Animal species and evolution. Oxford: Oxford Univer-

sity Press. http://dx.doi.org/10.4159/harvard.9780674865327Medina, R., Lara, F., Goffinet, B., Garilleti, R. & Mazimpaka, V.

2012. Integrative taxonomy successfully resolves the pseudo-cryp-tic complex of the disjunct epiphytic moss Orthotrichum consimile s.l. (Orthotrichaceae). Taxon 61: 1180–1198.

Medina, R., Lara, F., Goffinet, B., Garilleti, R. & Mazimpaka, V. 2013. Unnoticed diversity within the disjunct moss Orthotrichum tenellum s.l. validated by morphological and molecular approaches. Taxon 62: 1133–1152. http://dx.doi.org/10.12705/626.15

Meikle, R.D. 1957. What is the subspecies? Taxon 6: 102–105. http://dx.doi.org/10.2307/1217753Monaghan, M.T., Wild, R., Elliot, M., Fujisawa, T., Balke, M.,

Inward, D.J.G., Lees, D.C., Ranaivosolo, R., Eggleton, P., Barraclough, T.G. & Vogler, A.P. 2009. Accelerated species inventory on Madagascar using coalescent-based models of spe-cies delineation. Syst. Biol. 58: 298–311.

http://dx.doi.org/10.1093/sysbio/syp027Müller, K. 2005. SeqState: Primer design and sequence statistics for

phylogenetic DNA datasets. Appl. Bioinf. 4: 65–69. http://dx.doi.org/10.2165/00822942-200504010-00008Murphy, N.P., Adams, M., Guzik, M.T. & Austin, A.D. 2013. Extraor-

dinary micro-endemism in Australian desert spring amphipods. Molec. Phylogen. Evol. 66: 645–653.

http://dx.doi.org/10.1016/j.ympev.2012.10.013Orzechowska, M., Siwinska, D. & Maluszynska, J. 2010. Molecular

cytogenetic analyses of haploid and allopolyploid Pellia species. J. Bryol. 32: 113–121.

http://dx.doi.org/10.1179/037366810X12578498136075Pagel, M. 1999. The maximum likelihood approach to reconstructing

ancestral character states of discrete characters on phylogenies. Syst. Biol. 48: 612–622.

http://dx.doi.org/10.1080/106351599260184Pagel, M. & Meade, A. 2006. Bayesian analysis of correlated evolution

of discrete characters by reversible-jump Markov chain Monte Carlo. Amer. Naturalist 167: 808–825.

http://dx.doi.org/10.1086/503444Pagel, M., Meade, A. & Barker, D. 2004. Bayesian estimation of

ancestral character states on phylogenies. Syst. Biol. 53: 673–684. http://dx.doi.org/10.1080/10635150490522232Payo, D.A., Leliaert, F., Verbruggen, H., D’Hondt, S., Calumpong,

H.P. & De Clerck, O. 2013. Extensive cryptic species diversity and fine-scale endemism in the marine red alga Portieria in the Philip-pines. Proc. Roy. Soc. London, Ser. B, Biol. Sci. 280: 2012–2660.

http://dx.doi.org/10.1098/rspb.2012.2660Pons, J., Barraclough, T.G., Gomez-Zurita, J., Cardoso, A., Duran,

D.P., Hazell, S., Kamoun, S., Sumlin, W.D. & Vogler, A.P. 2006. Sequence-based species delimitation for the DNA taxonomy of undescribed insects. Syst. Biol. 55: 595–609.

http://dx.doi.org/10.1080/10635150600852011Powell, J.R. 2012. Accounting for uncertainty in species delineation

during the analysis of environmental DNA sequence data. Methods Ecol. Evol. 3: 1–11. http://dx.doi.org/10.1111/j.2041-210X.2011.00122.x

Powell, J.R., Monaghan, M.T., Oepik, M. & Rillig, M.C. 2011. Evo-lutionary criteria outperform operational approaches in producing ecologically relevant fungal species inventories. Molec. Ecol. 20: 655–666. http://dx.doi.org/10.1111/j.1365-294X.2010.04964.x

Ramaiya, M., Johnson, M.G., Shaw, B., Heinrichs, J., Hentschel, J., von Konrat, M., Davison, P.G. & Shaw, A.J. 2010. Morpho-logically cryptic biological species within the liverwort Frullania asagrayana. Amer. J. Bot. 97: 1707–1718.

http://dx.doi.org/10.3732/ajb.1000171Rambaut, A. & Drummond, A.J. 2007. Tracer, version 1.5. http://tree

.bio.ed.ac.uk/software/tracer/Ronquist, F., Teslenko, M., Van der Mark, P., Ayres, D.L., Darling,

A., Höhna, S., Larget, B., Liu, L., Suchard, M.A. & Huelsen-beck, J.P. 2012. MrBayes 3.2: Efficient Bayesian phylogenetic inference and model choice across a large model space. Syst. Biol. 61: 539–542. http://dx.doi.org/10.1093/sysbio/sys029

Schuster, R.M. 1968. Studies on Hepaticae XLV. On Iwatsukia Kitagawa. Bull. Natl. Sci. Mus. Tokyo 11: 309–317.

Schuster, R.M. 1974. The Hepaticae and Anthocerotae of North Amer-ica East of the Hundredth Meridian, vol. 3. New York: Columbia University Press.

Schuster, R.M. 1984. Comparative anatomy and morphology of the Hepaticae. Pp. 760–1070 in: Schuster, R.M. (ed.), New manual of bryology, vol. 2. Nichinan: The Hattori Botanical Laboratory.

Schuster, R.M. 2002. Austral Hepaticae, part II. Nova Hedwigia Beih. 119: 1–606.

Shaw, A.J., Cox, C.J. & Boles, S.B. 2003. Polarity of peatmoss (Sphag-num) evolution: Who says bryophytes have no roots? Amer. J. Bot. 90: 1777–1787. http://dx.doi.org/10.3732/ajb.90.12.1777

So, M.L. 2004. Studies on the type specimens of Odontoschisma (Hepaticae) in Asia and the Pacific. J. Hattori Bot. Lab. 95: 249–261.

Spruce, R. 1876. On Anomoclada, a new genus of Hepaticae, and on its allied genera, Odontoschisma and Adelanthus. J. Bot. 14: 230–235.

Stamatakis, A., Hoover, P. & Rougemont,J. 2008. A rapid bootstrap algorithm for the RAxML Web-Servers. Syst. Biol. 75: 758–771.

http://dx.doi.org/10.1080/10635150802429642Stenøien, H.K., Shaw, A.J., Shaw, B., Hassel, K. & Gunnarsson, U.

2011. North American origin and recent European establishments of the amphi-Atlantic peat moss Sphagnum angermanicum. Evolu-tion 65: 1181–1194. http://dx.doi.org/10.1111/j.1558-5646.2010.01191.x

Stephani, F. 1908. Species Hepaticarum: Odontoschisma. Bull. Herb. Boissier, ser. 2, 8: 582–595.

Sukkharak, P., Gradstein, S.R. & Stech, M. 2011. Phylogeny, taxon circumscriptions, and character evolution in the core Ptychan-thoideae (Lejeuneaceae, Marchantiophyta). Taxon 60: 1607–1622.

Szöevenyi, P., Terracciano, S., Ricca, M., Giordano, S. & Shaw, A.J. 2008. Recent divergence, intercontinental dispersal and shared

1024



Appendix 1. Voucher information and GenBank accession numbers for the specimens of Odontoschisma and outgroup taxa used in this study. Sequences in bold where obtained from GenBank.

Taxon, origin, voucher (herbarium), Lab. ID, atpB-rbcL, trnL-F, rps4

O. bifidum (Fulford) Gradst., S.C.Aranda & Vanderp. (= Iwatsukia bifida), Venezuela, Halling 5568 (NY), IWA, KJ620667, KJ620713, –; O. brasiliense Steph., Brasil, Rio de Janeiro, Santos & Costa 568 (RB), OBRA, KJ620655, KJ620692, KJ620747; O. cleefii Gradst., S.C.Aranda & Vanderp., Colombia, Campos 717 (COL, PC), O9, KJ620649, KJ620698, KJ620746; O. denudatum (Nees) Dumort., Azores, Terceira, Vanderpoorten TER4 (LG), OD8, KJ620644, KJ620682, KJ620738; O. denudatum (Nees) Dumort., Sweden, Georgson s.n. (PC), OD12, KJ620646, KJ620680, KJ620737; O. denudatum (Nees) Dumort., France, Gradstein s.n. (PC), OD13, KJ620643, KJ620681, KJ620735; O. denudatum (Nees) Dumort., The Netherlands, Van Melick 214610 (PC), OD15, KJ620645, KJ620683, KJ620736; O. denudatum (Nees) Dumort., Japan, Hiroshima, Miyauchi 1331 (HIRO, PC), OD17, KJ620647, KJ620693, KJ620750; O. denudatum (= O. subjulaceum Austin), Hawaii, Frahm 472 (hb. Frahm), OD18, KJ620642, KJ620696, KJ620733; O. denudatum subsp. naviculare (Steph.) Gradst., S.C.Aranda & Vanderp. (= O. naviculare), New Caledonia, Larrain s.n. (hb. Larrain, LG), ON3, KJ620648, KJ620695, KJ620739; O. denudatum subsp. sand-vicense (Ångstr.) Gradst., S.C.Aranda & Vanderp. (= O. sandvicense), Hawaii, Patino s.n. (LG), O11, KJ620668, KJ620697, KJ620734; O. elongatum (Lindb.) A.Evans, Sweden, Hallingbäck 4259 (hb. Hallingbäck), OE2, KJ620663, KJ620702, KJ620728; O. elongatum (Lindb.) A.Evans, Sweden, Hallingbäck 44271 (hb. Hallingbäck), OE1, KJ620664, KJ620701, KJ620729; O. elongatum (Lindb.) A.Evans, U.S.A., Alaska, B. Shaw 7389 (DUKE), OSP10, KJ620665, KJ620711, KJ620730; O. engelii Gradst. & Burghardt, Ecuador, Burghardt 6990 (PC, QCA), OEN1, KJ620659, KJ620686, KJ620755; O. fluitans (Nees) L.Söderstr. & Váňa (= Cladopodiella fluitans), Belgium, Vanderpoorten K626-6 (LG), CFLU, KJ620666, KJ620712, KJ620754; O. francisci (Hook.) L.Söderstr. & Váňa (= Cladopodiella francisci), Belgium, Sotiaux 37952 (hb. Sotiaux), CFRA, KJ620641, KJ620710, KJ620753; O. grosseverrucosum Steph., China, Zenjiang, Gradstein 12388 (PC), OG3, KJ620634, KJ620676, KJ620723; O. jishibae (Steph.) L.Söderstr. & Váňa, EU791722, –; O. longiflorum (Taylor) Trevis., Ecuador, Gradstein 12112 (PC), O3, KJ620654, KJ620684, KJ620740; O. cf. longiflorum (Taylor) Trevis., Ecuador, 1990 m, Burghardt 7076 (PC), OV2, KJ620656, KJ620685, KJ620741; O. macounii (Austin) Underw., Iceland, Vanderpoorten 4470 (LG), OD7, KJ620635, KJ620704, KJ620724; O. macounii (Austin) Underw., Sweden, Hallingbäck 2031 (hb. Hallingbäck, PC), OM2, KJ620638, KJ620705, KJ620726; O. macounii (Austin) Underw., Russian Far East, Bakalin Mag-12-11-10 (VLA), OM1, KJ620636, KJ620706, KJ620725; O. macounii (Austin) Underw., Russian Far East, Bakalin S-32-28a-06 (VLA), OM3, KJ620637, KJ620707, KJ620727; O. portoricense (Hampe & Gottsche) Steph. (= Anomoclada portoricensis), Suriname, Allen 20342 (MO), O8, KJ620657, KJ620690, KJ620749; O. portoricense (Hampe & Gottsche) Steph. (= Anomoclada portoricensis), Suriname, Allen 20379 (MO), OPOR, KJ620658, KJ620691, KJ620748; O. prostratum (Sw.) Trevis., U.S.A., North Carolina, B. Shaw 12994 (DUKE, PC), O4, KJ620631, KJ620669, KJ620722; O. prostratum (Sw.) Trevis., U.S.A., North Carolina, Vanderpoorten s.n. (LG), O6, KJ620628, KJ620671, KJ620714; O. prostratum (Sw.) Trevis., U.S.A., Alabama, Davidson 8150 (PC), OPR2, KJ620629, KJ620673, KJ620716; O. prostratum (Sw.) Trevis., U.S.A., North Carolina, B. Shaw 12993 (DUKE, PC), OPR1, KJ620630, KJ620672, KJ620721; O. prostratum (Sw.) Trevis., U.S.A., Connecticut, B. Shaw 7184 (DUKE), OPR11, KJ620633, KJ620677, KJ620718; O. pseudogrosseverrucosum Gradst., S.C.Aranda & Vanderp., Japan, Katagari 3525 (HIRO, PC), O7, KJ620639, KJ620708, KJ620751; O. pseudogrosseverrucosum Gradst., S.C.Aranda & Van-derp., Japan, Y. Inoue 174 (HIRO, PC), OG2, KJ620640, KJ620709, KJ620752; O. sphagni (Dicks.) Dumort., Azores, S. Jorge, Vanderpoorten & Patino (LG), OD6, KJ620626, KJ620670, KJ620719; O. sphagni (Dicks.) Dumort., Azores, Terceira, Frahm Az-110 (hb. Frahm), OSP1, KJ620627, KJ620675, KJ620715; O. sphagni (Dicks.) Dumort., Azores, Pico, Vanderpoorten & Patino OD1 (LG), OD1, KJ620632, KJ620679, KJ620717; O. sphagni (Dicks.) Dumort., Belgium, Gradstein 12380 (PC), OSP4, KJ620624, KJ620674, KJ620720; O. sphagni (Dicks.) Dumort., U.K., Scotland, Long 37024 (E), OSP14, KJ620625, KJ620678, KJ620756; O. variabile (Lindenb. & Gottsche) Trevis., Colombia, Campos 718 (COL, PC), DPAR, KJ620650, KJ620694, KJ620742; O. variabile (Lindenb. & Gottsche) Trevis., Ecuador, 3035 m, Schäfer-Verwimp 31776 (hb. Schäfer-Verwimp), OV3, KJ620651, KJ620689, KJ620744; O. variabile (Lindenb. & Gottsche) Trevis., Ecuador, 2250 m, Burghardt 7047 (PC), OV4, KJ620652, KJ620687, KJ620743; O. variabile (Lindenb. & Gottsche) Trevis., Ecuador, 2400 m, Schäfer-Verwimp 31991 (hb. Schäfer-Verwimp), OV5, KJ620653, KJ620688, KJ620745; O. zhui Gradst., S.C.Aranda & Vanderp., China, Zenjiang, Gradstein 12385 (PC), DENA1, KJ620660, KJ620700, KJ620731; O. zhui Gradst., S.C.Aranda & Vanderp., China, Zenjiang, Gradstein 12384 (PC), DENA2, KJ620661, KJ620699, –; O. zhui Gradst., S.C.Aranda & Vanderp., China, Hunan, Koponen 50933 (H), ON2, KJ620662, KJ620703, KJ620732; Alobiellopsis parvifolia (Steph.) R.M.Schust., JX630020, –; Cephalozia bicuspidata (L.) Dumort., JX630025, JX308561; Fuscocephaloziopsis lunulifolia (Dumort.) Váňa & L.Söderstr., JX629958, AM398315; Hygrobiella laxifolia (Hook.) Spruce, JX630054, –; Nowellia curvifolia (Dicks.) Mitt., JX629997, AY608094; Pleurocladula albescens (Hook.) Grolle, JX629992, –; Schiffneria hyalina Steph., AY463585, JF513488; Schofieldia monticola J.D.Godfrey, JX629988, –

polymorphism are shaping the genetic structure of amphi-Atlantic peatmoss populations. Molec. Ecol. 17: 5364–5377.

http://dx.doi.org/10.1111/j.1365-294X.2008.04003.xTalavera, G., Dinca, V. & Vila, R. 2013. Factors affecting species

delimitations with the GMYC model: Insights from a butterfly survey. Methods Ecol. Evol. 4: 1101–1110.

http://dx.doi.org/10.1111/2041-210X.12107Van Zanten, B.O. & Gradstein, S.R. 1988. Experimental dispersal

geography of neotropical liverworts. Beih. Nova Hedwigia 90: 41–94.

Váňa, J., Söderström, L., Hagborg, A. & von Konrat, M. 2013. Notes on Early Land Plants Today. 41. New combinations and synonyms in Cephaloziaceae (Marchantiophyta). Phytotaxa 112: 7–15.

http://dx.doi.org/10.11646/phytotaxa.112.1.2Vanderpoorten, A., Schäfer-Verwimp, A., Heinrichs, J., Devos, N.

& Long, D.G. 2010a. The taxonomy of the leafy liverwort genus Leptoscyphus (Lophocoleaceae) revisited. Taxon 59: 176–186.

http://www.jstor.org/stable/27757061Vanderpoorten, A., Gradstein, S.R., Carine, M.A. & Devos, N.

2010b. The ghosts of Gondwana and Laurasia in modern liverwort distributions. Biol. Rev. (Cambridge) 85: 471–487.

http://dx.doi.org/10.1111/j.1469-185X.2009.00111.x

Vilnet, A.A., Konstantinova, N.A. & Troitsky, A.V. 2012. Molecular phylogeny and systematics of the suborder Cephaloziineae with special attention to the family Cephaloziaceae s.l. (Jungerman-niales, Marchantiophyta). Arctoa 21: 113–132.

http://dx.doi.org/10.15298/arctoa.21.11Vuataz, L., Sartori, M., Gattolliat, J.-L. & Monaghan, M.T. 2013.

Endemism and diversification in freshwater insects of Madagascar revealed by coalescent and phylogenetic analysis of museum and field collections. Molec. Phylogen. Evol. 66: 979–991.

http://dx.doi.org/10.1016/j.ympev.2012.12.003Whittaker, R.J., Araujo, M.B., Paul, J., Ladle, R.J., Watson, J.E.M.

& Willis, K.J. 2005. Conservation biogeography: A<ssessment and prospect. Diversity & Distrib. 11: 3–23.

http://dx.doi.org/10.1111/j.1366-9516.2005.00143.xXie, W., Lewis, P.O., Fan, Y., Kuo, L. & Chen, M.-H. 2011. Improving

marginal likelihood estimation for Bayesian phylogenetic model selection. Syst. Biol. 60: 150–160.

http://dx.doi.org/10.1093/sysbio/syq085Zink, R.M. 2004. The role of subspecies in obscuring avian biologi-

cal diversity and misleading conservation policy. Proc. Roy. Soc. London, Ser. B, Biol. Sci. 271: 561–564.

http://dx.doi.org/10.1098/rspb.2003.2617

1025



Appendix 2. Matrix of 20 morphological characters in Odontoschisma s.l. (including Anomoclada, Cladopodiella and Iwatsukia). Missing char-acters are scored as “ ? ”. Characters based on morphological study of the vouchers (Appendix 1) and additional herbarium material, as well as on literature (Fulford, 1968; Schuster, 1968, 1974; Konstantinova, 2004; So, 2004).

Characters1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

O. bifidum IWA 0 2 1 1 1,2 2 2 2 0 1 3 0 0 1 1 1 2 2 1 ?O. brasiliense OBRA 2 0 1 1 2 0 0 0 0 2 0 1 1 0 0 0 0 2 ? ?O. cleefii O9 1 0 0 1 0 2 0 0 0 1 1 0 0,1 0 0 2 0 0 0 ?O. denudatum OD8 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum OD12 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum OD13 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum OD15 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum OD17 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum OD18 1 0 0 0 0 2 0 0 0 1 1 0 0 0 0 1 0 1 1 1O. denudatum subsp. naviculare ON3 1 0 0 1 1 2 0 0,1 1 2 1 0 0 0 0 1 0 0 ? ?O. denudatum subsp. sandvicense O11 1,2 0 0 1 1 0 0 0 0 2 1 0 0 0 0 1 0 0 1 ?O. elongatum OE1 1 1 0 1 0 2 0 0 0 1 1 0 0 2 0 0 0 1 1 1O. elongatum OE2 1 1 0 1 0 2 0 0 0 1 1 0 0 2 0 0 0 1 1 1O. elongatum OSP10 1 1 0 1 0 2 0 0 0 1 1 0 0 2 0 0 0 1 1 1O. engelii OEN1 0 0 0 1 0 2 0 2 0 1 1 0 0 0 0 0 0 0 1 ?O. fluitans CFLU 1 0 1 1 0 0 2 1 0 2 3 0 0 0 0 0 1 2 0,1 0O. francisci CFRA 0 0 1 1 0 2 1 0 0 1 3 0 0 0 0 0,1 1 0 0,1 0O. grosseverrucosum OG3 1 2 1 1 1,2 1 0 0 0 0 0 1 0 0 0 2 0 2 ? ?O. jishibae JIS 0 2 1 1 1 2 3 0,1 0 1 3 0 0 1 1 1 2 2 1 0O. longiflorum O3 1 2 0 0 2 0 0 0,1 0 2 0 1 0 0 0 1 0 1 1 ?O. cf. longiflorum OV2 1 0 0 1 1 1 0 0,1 0 1 1 0 0 0 0 1 0 1 1 1O. macounii OD7 1 2 0 1 0 2 0 0 0 2 2 0 0 1 0 0 1 0 1 1O. macounii OM2 1 2 0 1 0 2 0 0 0 2 2 0 0 1 0 0 1 0 1 1O. macounii OM1 1 2 0 1 0 2 0 0 0 2 2 0 0 1 0 0 1 0 1 1O. macounii OM3 1 2 0 1 0 2 0 0 0 2 2 0 0 1 0 0 1 0 1 1O. portoricense O8 2 0 0 0 0 0 0 2 0 2 1 0 0 1 0 1 0 0 1 ?O. portoricense OPOR 2 0 0 0 0 0 0 2 0 2 1 0 0 1 0 1 0 0 1 ?O. prostratum O4 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. prostratum OPR11 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. prostratum O6 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. prostratum OPR1 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. prostratum OPR2 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. pseudogrosseverrucosum O7 1 1 0 1 1,2 1 0 0 0 0 0 0 0 0 0 2 0 0 ? ?O. pseudogrosseverrucosum OG2 1 1 0 1 1,2 1 0 0 0 0 0 0 0 0 0 2 0 0 ? ?O. sphagni OD6 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. sphagni OSP1 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. sphagni OD1 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. sphagni OSP4 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. sphagni OSP14 1 0 1 1 2 1 0 0,1 0 1 0 1 0 0 0 1 0 2 2 1O. variabile DPAR 1 0 0 1 1 1 0 0,1 0 1 1 0 0 0 0 1 0 1 1 1O. variabile OV3 1 0 0 1 1 1 0 0,1 0 1 1 0 0 0 0 1 0 1 1 1O. variabile OV4 1 0 0 1 1 1 0 0,1 0 1 1 0 0 0 0 1 0 1 1 1O. variabile OV5 1 0 0 1 1 1 0 0,1 0 1 1 0 0 0 0 1 0 1 1 1O. zhuii DENA1 1 0 0,1 1 0 2 0 0 1 0 1 0 0 0 0 0,1 0 0 1 ?O. zhuii DENA2 1 0 0,1 1 0 2 0 0 1 0 1 0 0 0 0 0,1 0 0 1 ?O. zhuii ON2 1 0 0,1 1 0 2 0 0 1 0 1 0 0 0 0 0,1 0 0 1 ?1. Plant width: less than 1 mm wide (0), 1–2 mm wide (1), 2–4 mm wide (2); 2. Pigmentation: plants with reddish or purplish pigmentation (0), pigmented but never with reddish or purplish (1), not pigmented (2); 3. Lateral branches: present (0) or absent (1); 4. Anomoclada-type branches: present (0) or absent (1); 5. Leaf-free strip: 0–1 cell wide (0), 1–2(–3) cells wide (1), 3–8 cells wide (2); 6. Leaf lamina: surface and margins flat (0), surface flat to slightly concave, dorsal margin upcurved (1), surface always concave, margins upcurved (2); 7. Leaf apex: undivided (0) or bifid to 1/5 (1), 1/3 (2) or 1/2 (3); 8. Leaf elongation: leaf about as wide as long (0), elongate, up to 1.5 : 1 (1), strongly elongate, 1.5–2.5 : 1 (2); 9. Leaves: straight (0), curved-falcate (1); 10. Mid-leaf cell size: 9–16(–20) µm (0), 16–25(–30) µm (1), 25–40 µm (2); 11. Trigones: small to medium-sized (0), medium-sized to large (1), very large (2), indistinct, walls evenly thickened (3); 12. Leaf border: lacking (0), present (1); 13. Outer wall of leaf margin: thickened to ca. 5 µm (0), very strongly thickened, to 12 µm (1); 14. Middle lamella of leaf cells: indistinct (0), distinct, colorless (1), distinct, brown (2); 15. Transverse walls of leaf cells ornamented by numerous verruculae (sometimes interpreted as plas-modesmata): absent (0), present (1); 16. Cuticle: completely smooth (0), finely verruculose (1), coarsely verrucose (2); 17. Underleaves: absent or rudimentary (0), present but small (1) well developed, to 1/2× leaf length (1); 18. Gemmiparous leaves appressed (0), appressed to squarrose (1), lacking (2); 19. Female bract apices rounded to obtuse (0), obtuse to acute to short acuminate (1), long acuminate to ciliate (2); 20. Capsule epider-mis: with nodular thickenings on all long walls (0), on alternate long walls (1).

Documents

Phylogeny, classification and species delimitation in the ... · PDF file1010 Aranda & al. • Phylogeny of Odontoschisma TAXON 63 (5) • October 2014: 1008–1025 Version of Record