Page 1
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofitpublishers, academic institutions, research libraries, and research funders in the common goal of maximizing access tocritical research.
Unraveling the evolutionary history of the liverworts(Marchantiophyta): multiple taxa, genomes and analysesAuthor(s): Laura L. Forrest, E. Christine Davis, David G. Long, Barbara J.Crandall-Stotler, Alexandra Clark, and Michelle L. HollingsworthSource: The Bryologist, 109(3):303-334. 2006.Published By: The American Bryological and Lichenological Society, Inc.DOI: 10.1639/0007-2745(2006)109[303:UTEHOT]2.0.CO;2URL:http://www.bioone.org/doi/full/10.1639/0007-2745%282006%29109%5B303%3AUTEHOT%5D2.0.CO%3B2
BioOne (www.bioone.org) is an electronic aggregator of bioscience research content,and the online home to over 160 journals and books published by not-for-profit societies,associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associatedcontent indicates your acceptance of BioOne’s Terms of Use, available atwww.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercialuse. Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder.
Page 2
Unraveling the evolutionary history of the liverworts
(Marchantiophyta): multiple taxa, genomes and analyses
LAURA L. FORREST
Department of Plant Biology, University of Southern Illinois,
Carbondale, IL 62901, U.S.A.
e-mail: [email protected]
E. CHRISTINE DAVIS
Department of Biology, Duke University, Box 90338,
Durham, NC 27708-0338, U.S.A.
e-mail: [email protected]
DAVID G. LONG
Royal Botanic Garden, Edinburgh, 20A Inverleith Row,
Edinburgh, EH3 5LR, Scotland, U.K.
e-mail: [email protected]
BARBARA J. CRANDALL-STOTLER
Department of Plant Biology, University of Southern Illinois,
Carbondale, IL 62901, U.S.A.
e-mail: [email protected]
ALEXANDRA CLARK AND MICHELLE L. HOLLINGSWORTH
Royal Botanic Garden, Edinburgh, 20A Inverleith Row,
Edinburgh, EH3 5LR, Scotland, U.K.
e-mails: [email protected] ; [email protected]
ABSTRACT. Nucleotide sequence data from three chloroplast genes (rbcL, rps4 and psbA), one
nuclear gene (the ribosomal LSU) and one mitochondrial gene (nad5) were assembled for
173 species in 117 genera of liverworts, making this the largest molecular phylogeny of the
group to date. Analyses of these data provide support for the monophyly of the liverworts,
and for previously resolved backbone relationships within the Marchantiophyta. The earliest
divergence involves the ‘‘simple thalloid’’ taxa of the Haplomitriaceae and Treubiaceae. A
Blasiaceae/complex thalloid clade is resolved as sister to all remaining liverworts. The leafy
liverworts do not resolve as monophyletic. The separation of the Aneuraceae/Metzgeriaceae
from all other simple thalloids and their placement within the ‘‘leafy’’ clade as sister to the
enigmatic leafy genus Pleurozia, as suggested in earlier molecular phylogenies, is also
supported by this far larger data set.
KEYWORDS. Liverworts, Marchantiophyta, multi-genome phylogeny, molecular phylogeny,
rps4, rbcL, pbsA, nad5, LSU.
^ ^ ^
THE BRYOLOGIST 109(3), pp. 303–334
Copyright �2006 by the American Bryological and Lichenological Society, Inc.
0007-2745/06/$3.35/0
Page 3
Liverworts hold a pivotal position in early land plant
evolution, with growing evidence that they represent
the earliest diverging lineage of embryophytes (e.g.,
Groth-Malonek & Knoop 2005; Groth-Malonek et al.
2005; Pruchner et al. 2002; Qiu et al. 1998;
Steinhauser et al. 1999; Wolf et al. 2005). The group
also has the oldest fossil record of any bryophyte, with
some evidence that it dates to at least 475 million
years before present (Wellman et al. 2003). However,
until the start of this millennium surprisingly few
hypotheses of phylogenetic relationships within this
ancient and morphologically heterogeneous group
had been tested with molecular data (Forrest &
Crandall-Stotler 2005). Recently, enormous leaps in
our understanding of liverwort evolution have
occurred, with the publication of a number of key
multi-gene, multi-taxon analyses (e.g., Crandall-
Stotler et al. 2005; Davis 2004; Forrest & Crandall-
Stotler 2004, 2005; He-Nygren et al. 2004, 2006).
However, all but the last of these studies had their
independent foci on specific clades within the
Marchantiophyta Stotler & Crand.-Stotler, with
Forrest and Crandall-Stotler (2004, 2005) and Cran-
dall-Stotler et al. (2005) focusing on the simple
thalloid liverworts (Jungermanniopsida Stotler &
Crand.-Stotler subcl. Metzgeriidae Barthol.-Began),
and Davis (2004) and He-Nygren et al. (2004)
focusing on the leafies (subcl. Jungermanniidae Engl.).
No one of these multi-gene analyses has included a
broad level of sampling across the entire March-
antiophyta and, further, an entire class, the complex
thalloids (Marchantiopsida Cronquist, Takht. & W.
Zinm.), has effectively been neglected since the single-
locus study of Boisselier-Dubayle et al. (2002).
Estimates of the total number of liverwort species
are in the order of 4500–5000 (http://bryophytes.
plant.siu.edu) in 376 genera and 74 families (Cran-
dall-Stotler & Stotler 2000). The vast majority of
species belong to the Jungermanniopsida subcl.
Jungermanniidae or leafy liverwort group (Schuster
1984). Although the leafy liverworts contain most of
the taxonomic diversity of the phylum, the most
significant morphological diversity is expressed within
the thalloid liverwort groups. Thalloids currently
comprise around 500 species of simple thalloids
(Jungermanniopsida subcl. Metzgeriidae) and 350
species of complex thalloids (Marchantiopsida).
Liverwort taxon sizes are representative of the ‘‘hollow
curve’’ distribution of Willis (1922); of the currently
accepted liverwort genera, 148, i.e., ca. 40%, are
monospecific, and relatively few contain more than
100 species, with the two largest genera being
Frullania Raddi (over 300 species), and Plagiochila
(Dumort.) Dumort. (over 400 species). Likewise,
eight families are monospecific, and 37 contain fewer
than 10 species. There are 14 families with over 100
species, including one far outlier, the Lejeuneaceae
Casares-Gil, which alone is estimated to contain
nearly 1000 species in 91 genera (Gradstein et al.
2003). However, to address whether these disparate
taxon sizes are a natural phenomenom, or if this
‘‘hollow curve’’ distribution is an artificial construct,
as suggested by Clayton (1974) and Walters (1986),
discussion must be framed in terms of clades rather
than in terms of traditional arbitrary taxonomic units
like families and genera. This is particularly important
as evidence mounts that polyphyly and paraphyly are
rife in traditional taxonomies of the group (e.g.,
Forrest et al. 2005b; Long et al. 2000; Schaumann et al.
2005).
Recent phylogenetic studies have identified the
major backbone clades within the group to be 1) the
Haplomitriopsida Stotler & Crand.-Stotler, named by
Stotler and Crandall-Stotler in 1977 and here defined
to include the Haplomitriaceae Dedecek and Treu-
biaceae Verd. [¼ Treubiopsida Stech., J.-P. Frahm,
Hilger & W. Frey as delimited in He-Nygren et al.
(2006)], with 20 recognized species, 2) the March-
antiopsida (defined by He-Nygren et al. (2006) to
include the Blasiales (R. M. Schust.) Stotler & Crand.-
Stotler), with around 350 species, and 3) the
Jungermanniopsida Stotler & Crand.-Stotler, with
around 4000–4500 species (Crandall-Stotler et al.
2005; Davis 2004; Forrest & Crandall-Stotler 2004,
2005; Forrest et al. 2005a; Heinrichs et al. 2005; He-
Nygren et al. 2006).
To provide a broader sampling across the
diversity of liverworts, we have assembled a data
matrix with 189 liverwort accessions representing all
but two of the 32 suborders and including approx-
imately 2.5–4% of all liverwort species. The Brevian-
thineae J. J. Engel & R. M. Schust., containing two
genera (Brevianthus J. J. Engel & R. M. Schust. and
Chonecolea Grolle) and four species, and the mono-
304 the bryologist 109(3): 2006
Page 4
specific Monocarpineae R. M. Schust., were not
available for sampling. Although our sampling is
focused on resolving the major lineages within the
group, and under-represents diversity in the highly
speciose leafy liverworts, it represents the largest
taxonomic sampling of liverworts to date, with 117 of
the estimated 376 genera (over 30%) represented.
Twenty-five of the 32 genera of complex thalloids
(78%), 30 of the 38 genera of simple thalloids (79%),
and 62 of the 306 genera of leafies (20%) are included.
MATERIAL AND METHODS
Taxon sampling. Sequences were assembled for
189 accessions of hepatics, with representation as
follows: eight species, three genera, both suborders of
the Haplomitriopsida; 36 species, 25 genera, seven of
the eight suborders of the Marchantiopsida; 49
species, 27 genera, all eight suborders of the
Jungermanniopsida subcl. Metzgeriidae; and 80
species, 62 genera, 15 of the 16 suborders of the
Jungermanniopsida subcl. Jungermanniidae. Five
hornwort, ten moss species and two vascular plants
were also included, and initial topologies (Maximum
Parsimony) were rooted using four algal taxa. Non-
liverwort sequences were mostly obtained from
GenBank. For GenBank numbers and voucher
information, see Table 1. Classification throughout
the paper follows Crandall-Stotler and Stotler (2000),
unless explicitly stated.
Molecular methods. Molecular methods differ
according to laboratory, with sequences being pro-
duced from three different labs (see Table 1).
Protocols for sequences generated at Southern Illinois
University follow Forrest and Crandall-Stotler (2004,
2005); methodologies from Duke University follow
Davis (2004), and from the Royal Botanic Garden,
Edinburgh, follow Schill et al. (2004).
Data analyses. Sequences were aligned in PAUP*
version 4.0b10 (Swofford 2002). Regions of ambig-
uous alignment and incomplete data (i.e., the
beginnings and ends of sequenced regions) were
identified and excluded from further analyses. The
data matrix is available in NEXUS format at
TreeBASE (www.treebase.org; study accession num-
ber S1519; matrix accession number M2725).
Missing data.—A total of 137 sequences were
missing, out of 1,050 possible sequences (ca. 13%). Of
these, 29 were from the nr LSU, 34 were from nad5,
29 from rbcL, 24 from rps4 and 21 from psbA (see
Table 1). Also, nad5 sequences generated at Duke
University only included the region between primers
K and Li, while those generated at Southern Illinois
University included the region between primers K2 or
K and L (Beckert et al. 1999). Despite this, the entire
K2-L region was included in the analyses, due to its
usefulness in resolving backbone relationships (Forr-
est et al., unpublished data; Groth-Malonek et al.
2005).
Analyses were conducted at Duke University,
Southern Illinois University and the Royal Botanic
Garden, Edinburgh. Maximum parsimony (MP)
phylogenetic analyses were performed using PAUP*
4.0b10 (Swofford 2002), mounted on an Apple
Macintosh G4 [SIU]. Bayesian inference (BI) analyses
were performed using MrBayes 3.0b4 (Huelsenbeck &
Ronquist 2002) and MrBayes 3.1 (Huelsenbeck &
Ronquist 2005), mounted on an Apple Macintosh G4
[SIU], G5 [E] or Unix running Linux Centos 3
[Duke]. Maximum parsimony analyses were run
under Fitch parsimony, using 1000 random addition
replicates, with TBR, saving 15 trees per replicate. The
most parsimonious trees (MPTs) found were then
input to a second round of TBR, with no limit on the
number of trees saved. Bootstrapping was performed
with 1000 replicates, using an heuristic search strategy
(five random addition replicates, saving five trees per
replicate). Prior to the combined analyses, separate
parsimony analyses were conducted for each of the
five loci individually, to identify and eliminate
problem sequences.
MrModeltest version 1.1b (Nylander 2002) was
used to establish the model of DNA evolution with
the best fit to the data for the combined DNA matrix,
and also for each locus individually, using the
topology of one of the trees resolved by MP analysis
(Table 2). Posterior probabilities for clades were
generated with at least two independent runs per
analysis, using the models described in Table 2, with
one million generations run for the homogeneous
analysis (BI1). The homogeneous analysis used the
General Time Reversible model of nucleotide sub-
stitution (Yang 1994) plus a gamma distribution of
rate variation among sites and invariant sites
(GTRþIþR) as selected by Akaike’s Information
Forrest et al.: Liverwort phylogeny 305
Page 5
Tab
le1.
Co
llec
tio
nd
etai
lsan
dG
enB
ank
acce
ssio
nn
um
ber
sfo
rta
xain
clu
ded
inth
isst
ud
y;as
teri
sks
den
ote
seq
uen
ces
no
tge
ner
ated
by
the
auth
ors
.A
x¼
esta
bli
shed
axen
iccu
ltu
re.
Un
der
Lab
,
Du
ke¼
Du
keU
niv
ersi
ty,
SIU¼
Sou
ther
nIl
lin
ois
Un
iver
sity
,E¼
Ro
yal
Bo
tan
icG
ard
ens,
Ed
inb
urg
h.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Acr
olej
eun
eafe
rtil
is(R
ein
w.e
tal
.)Sp
ruce
exSt
eph
.In
do
nes
ia,B
ali,
Sch
afer
-Ver
wim
p17
009
[acc
esse
dfr
om
Gen
Ban
k]
——
*AY
6849
29—
—
Acr
omas
tigu
mex
ile
(Lin
den
b.)
A.E
van
sK
enya
,Ch
ua-
Pet
iot
Mb2
8(N
Y)
AY
6081
92—
—A
Y60
8041
AY
6079
19D
uke
Ad
elan
thu
sli
nd
enbe
rgia
nu
s(L
ehm
.)M
itt.
Arg
enti
na,
Dre
hw
ald
9700
83(N
Y)
AY
6081
93A
Y60
8262
DQ
4396
74A
Y60
8042
AY
6079
20D
uke
All
ison
iaco
ckay
nii
(Ste
ph
.)R
.M.S
chu
stN
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r44
70(A
BSH
)A
Y68
8679
AY
6887
42A
Y50
7389
AY
5074
32A
Y50
7472
SIU
An
astr
oph
yllu
mm
ich
auxi
i(F
.Web
er)
H.B
uch
U.S
.A.,
No
rth
Car
oli
na,
Sarg
ent
s.n
.(A
BSH
)(A
x)A
Y68
8680
AY
6887
43A
Y50
7390
AY
5074
33A
Y50
7473
SIU
An
eura
pin
guis
(L.)
Du
mo
rt.
Ru
ssia
,Bak
alin
&B
akal
in[a
cc.1
](N
Y)
AY
6081
95A
Y60
8263
DQ
4396
75A
Y60
8043
AY
6079
21D
uke
An
eura
pin
guis
(L.)
Du
mo
rt.
U.S
.A.,
Wis
con
sin
,Hat
cher
cult
ure
#12
(AB
SH)
(Ax)
AY
8773
66A
Y68
8744
AY
5073
91—
AY
5074
74SI
U
An
thel
iaju
lace
a(L
.)D
um
ort
.C
anad
a,Sc
hof
ield
&Sc
hof
ield
1020
42(D
UK
E)
AY
6081
96—
DQ
4396
76A
Y60
8044
AY
6079
22D
uke
Apo
met
zger
iafr
onti
pili
s(L
ind
b.)
Ku
wah
.&J.
J.E
nge
lA
rgen
tin
a,D
reh
wal
d97
0084
(NY)
AY
6081
97A
Y60
8264
DQ
4396
77A
Y60
8045
AY
6079
24D
uke
Apo
met
zger
iapu
besc
ens
(Sch
ran
k)K
uw
ah.
Sco
tlan
d,M
idlo
thia
n,L
ong
2952
3(E
)D
Q26
8876
DQ
2689
06D
Q26
8967
DQ
2689
82D
Q26
8994
SIU
Apo
treu
bia
nan
a(S
.Hat
t.&
Ino
ue)
S.H
att.
&M
izu
t.N
epal
,Lon
g30
462
(1)
Lon
g30
451
(2)
(E)
DQ
4600
30(1
)D
Q26
8907
(2)
AY
8773
89(2
)D
Q26
8983
(2)
AY
8773
97(2
)SI
U/E
Asc
idio
tabl
eph
arop
hyl
laC
.Mas
sal.
ssp
.ala
skan
a
Stee
re&
R.M
.Sch
ust
.
U.S
.A.,
Lew
is51
4(F
)A
Y60
8198
AY
6082
65D
Q43
9678
AY
6080
46A
Y60
7925
Du
ke
Ast
erel
laaf
rica
na
(Mo
nt.
)A
.Eva
ns
Po
rtu
gal,
Mad
eira
,Sch
ill
18(E
)D
Q26
5770
DQ
2689
08D
Q28
5999
DQ
2206
73—
E
Ast
erel
lagr
olle
iD
.G.L
on
gC
hin
a,Q
ingh
aiP
rov.
,Lon
g27
203
(E)
DQ
2657
71D
Q26
8909
DQ
2860
00D
Q22
0674
DQ
2657
44E
Ast
erel
law
alli
chia
na
(Leh
m.&
Lin
den
b.)
Gro
lle
Nep
al,L
ong
3025
1( E
)D
Q26
5772
DQ
2689
10D
Q28
6001
DQ
2206
75D
Q26
5745
E
Ath
alam
iah
yali
na
(So
mm
erf.
)S.
Hat
t.A
ust
ria,
Sch
ill
60(E
)D
Q26
5773
DQ
2689
11D
Q28
6002
DQ
2206
76D
Q26
5746
E
Ath
alam
iapi
ngu
isF
alc.
Ind
ia,U
ttar
anch
al,L
ong
3088
9(E
)D
Q26
5774
DQ
2689
12D
Q28
6003
DQ
2206
77D
Q26
5747
E
Au
stro
foss
ombr
onia
aust
rali
s(M
itt.
)R
.M.S
chu
st.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4610
(AB
SH)
AY
6886
81A
Y68
8745
AY
5073
92A
Y50
7434
AY
5074
75SI
U
Au
stro
foss
ombr
onia
peru
vian
a(G
ott
sch
e&
Ham
pe)
Sto
tler
,Cra
nd
.-St
otl
er&
A.V
.Fre
ire
Ven
ezu
ela,
Stot
ler
&C
ran
dal
l-St
otle
r41
87(A
BSH
)A
Y68
8682
AY
8773
84A
F53
6230
AY
5074
35A
Y50
7476
SIU
Bal
anti
opsi
sd
iplo
phyl
la(H
oo
k.f.
&T
aylo
r)M
itt.
Au
stra
lia,
Stre
iman
n59
554
(NY)
AY
6081
99A
Y60
8266
DQ
4396
80A
Y60
8047
AY
6079
26D
uke
Baz
zan
iaaf
fin
is(L
ind
enb
.&G
ott
sch
e)T
revi
s.Ja
mai
ca,D
avis
146
(DU
KE)
AY
6082
00A
Y60
8267
DQ
4396
79A
Y60
8048
AY
6079
27D
uke
Bla
sia
pusi
lla
L.(
1)U
.S.A
.,H
icks
&D
avis
on23
92(D
UK
E)
AY
6082
01A
Y60
8268
—A
Y60
8049
AY
6079
28D
uke
Bla
sia
pusi
lla
L.(
2)U
.S.A
.,N
ort
hC
aro
lin
a,R
enza
glia
s.n
.(A
BSH
)(A
x)A
Y68
8683
AY
6887
46A
F53
6232
AY
5074
36A
Y50
7477
SIU
Bry
opte
ris
fili
cin
a(S
w.)
Nee
sB
oli
via,
Chu
rch
ill
etal
.198
55(N
Y)
AY
6082
02A
Y60
8269
DQ
4396
81A
Y60
8051
AY
6079
30D
uke
Cal
ycu
lari
acr
ispu
laM
itt.
Jap
an,F
uru
kis.
n.(
AB
SH)
AY
6886
84A
Y68
8747
AY
5073
95A
Y50
7437
AY
5074
78SI
U
Cal
ypog
eia
mu
elle
rian
a(S
chif
fn.)
Mu
ll.F
rib
.U
.S.A
.,D
avis
130
(DU
KE)
AY
6082
03—
—A
Y60
8052
AY
6079
31D
uke
Cav
icu
lari
ad
ensa
Step
h.
Jap
an,N
ish
imu
ras.
n.(
AB
SH)
—D
Q26
8913
DQ
2689
68D
Q26
8984
AY
5074
79SI
U
Cep
hal
ozia
cate
nu
lata
(Hu
eben
er)
Lin
db
.U
.S.A
.,A
nd
erso
n26
795
(DU
KE)
—A
Y60
8270
—A
Y60
8053
AY
6079
32D
uke
306 the bryologist 109(3): 2006
Page 6
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Cep
hal
ozie
lla
vari
ans
(Go
ttsc
he)
Step
h.
New
Zea
lan
d,E
nge
l21
964
(F)
mix
edw
ith
Jam
eson
iell
aco
lora
ta(L
ehm
.)St
eph
.
AY
6082
22—
DQ
4396
89A
Y60
8074
AY
6079
53D
uke
Cep
hal
ozie
lla
hir
ta(S
tep
h.)
R.M
.Sch
ust
.A
ust
rali
a,St
reim
ann
5979
3(N
Y)
AY
6082
04A
Y60
8271
DQ
4396
82A
Y60
8054
AY
6079
33D
uke
Cer
atol
ejeu
nea
coar
ina
(Go
ttsc
he)
Step
h.
Bra
zil,
Zar
tman
1235
.1(D
UK
E)
AY
6082
05A
Y60
8272
AY
6080
26A
Y60
8055
AY
6079
34D
uke
Che
ilol
ejeu
nea
trif
aria
(Rei
nw
.,B
lum
e&
Nee
s)M
izu
t.B
oli
via,
Gra
dst
ein
9951
A[a
cces
sed
fro
mG
enB
ank]
——
*AY
5480
94—
—
Chi
losc
yph
us
poly
anth
os(L
.)C
ord
aU
.S.A
.,U
tah
,Sto
tler
&C
ran
dal
l-St
otle
r43
56(A
BSH
)D
Q26
8877
DQ
2689
14D
Q26
8969
—D
Q26
8995
SIU
Con
ocep
hal
um
con
icu
m(L
.)D
um
ort
.s.la
t.U
.S.A
.,Il
lin
ois
,SIU
gree
nh
ou
se,S
totl
ers.
n.(
AB
SH)
DQ
2688
78A
Y68
8748
AY
6887
78A
Y68
8791
DQ
2689
96SI
U
Con
ocep
hal
um
japo
nic
um
(Th
un
b.)
Gro
lle
Nep
al,L
ong
3072
4(E
)D
Q26
5776
DQ
2689
15D
Q28
6005
DQ
2206
79D
Q26
5749
E
Con
ocep
hal
um
sale
bros
um
Szw
eyko
wsk
iet
al.
Ind
ia,U
ttar
anch
al,L
ong
3097
5(E
)D
Q26
5775
DQ
2689
16D
Q28
6004
DQ
2206
78D
Q45
9560
E
Cor
sin
iaco
rian
dri
na
(Sp
ren
g.)
Lin
db
.P
ort
uga
l,M
adei
ra,S
chil
l9
(E)
DQ
2657
77D
Q26
8917
DQ
2860
06D
Q22
0680
DQ
2657
50E
Cry
ptom
itri
um
him
alay
ense
Kas
hya
pN
epal
,Lon
g30
559
(E)
DQ
2657
78D
Q26
8918
DQ
2860
07D
Q22
0681
DQ
2657
51E
Cry
ptot
hal
lus
mir
abil
isM
alm
b.
Bel
giu
m,B
oud
ier
s.n
.[O
ct20
01]
(CO
NN
)A
Y60
8207
AY
6082
74D
Q43
9683
AY
6080
57A
Y60
7936
Du
ke
Cya
thod
ium
tube
rosu
mK
ash
yap
Nep
al,L
ong
3055
8(E
)D
Q26
5779
DQ
2689
19D
Q28
6008
DQ
2206
82D
Q26
5752
E
Cyc
lole
jeu
nea
con
vexi
stip
a(L
ehm
.&L
ind
enb.
)A
.Eva
ns
Bra
zil,
Zar
tman
1232
.1(D
UK
E)
AY
6082
08A
Y60
8275
—A
Y60
8058
AY
6079
37D
uke
Dip
loph
yllu
mal
bica
ns
(L.)
Du
mo
rt.
Fra
nce
,Vit
t&
Fra
hm
3670
5(F
)A
Y60
8210
AY
6082
77—
AY
6080
60A
Y60
7939
Du
ke
Dip
loph
yllu
mob
tusi
foli
um
(Ho
ok.
)D
um
ort
.U
.S.A
.,O
rego
n,W
hee
ler
s.n
.[re
f.n
o.3
51]
(AB
SH)
AY
6886
85A
Y68
8749
AY
5073
97A
Y50
7439
AY
5074
80SI
U
Dre
pan
olej
eun
eabi
ocel
lata
A.E
van
sE
cuad
or,
Gra
dst
ein
1005
3(G
OE
T)
[acc
esse
dfr
om
Gen
Ban
k]
——
*AY
5480
97—
—
Du
mor
tier
ah
irsu
ta(S
w.)
Nee
sU
.S.A
.,G
offi
net
6462
( DU
KE)
AY
6082
11A
Y60
8278
—A
Y60
8061
AY
6079
40D
uke
Du
mor
tier
ah
irsu
ta(S
w.)
Nee
sM
exic
o,V
erac
ruz,
Lon
g,G
arci
a&
De
Lu
na
2962
7(E
)D
Q26
5780
DQ
2689
20D
Q28
6009
DQ
2206
83D
Q26
5753
E
Exo
rmot
hec
apu
stu
losa
Mit
t.P
ort
uga
l,M
adei
ra,S
chil
l8
(E)
DQ
2657
81D
Q26
8921
DQ
2860
10D
Q22
0684
DQ
2657
54E
Fos
som
bron
iaan
gulo
sa(D
icks
.)R
add
iSp
ain
,Ten
erif
e,U
.&E
.Dre
hw
ald
s.n
.
[ref
.no
.371
](A
BSH
)
AY
6886
86A
Y68
8750
AY
5073
98A
Y50
7440
AY
5074
81SI
U
Fos
som
bron
iafo
veol
ata
Lin
db
.U
.S.A
.,M
assa
chu
sett
s,D
.&S.
Wil
liam
ss.
n.(
AB
SH)
AY
6886
87—
AY
5073
99A
Y50
7441
AY
5074
82SI
U
Fos
som
bron
iapu
sill
a(L
.)D
um
ort
.Sa
rgen
tli
vin
gcu
ltu
reco
llec
tio
n,B
erke
ley
(UC
)A
Y60
8212
AY
6082
79D
Q43
9684
AY
6080
62A
Y60
7941
Du
ke
Fos
som
bron
iasp
.in
ed.[
Ark
ansa
s]U
.S.A
.,A
rkan
sas,
Stot
ler
&C
ran
dall-
Stot
ler
3940
(AB
SH)
AY
6886
88A
Y68
8751
AY
5074
00A
Y50
7442
AY
5074
83SI
U
Fru
llan
iaal
bert
iiSt
eph
.E
cuad
or,
Dav
is29
5( D
UK
E)
AY
6082
13A
Y60
8280
DQ
4396
85—
AY
6079
42D
uke
Fru
llan
iaeb
orac
ensi
sL
ehm
.U
.S.A
.,Il
lin
ois
,Sto
tler
80–4
354
(AB
SH)
DQ
2688
79D
Q26
8922
AY
6887
79—
AY
6888
27SI
U
Fru
llan
iam
onil
iata
(Rei
nw
.,B
lum
e&
Nee
s)M
on
t.Ja
pan
,Miz
uta
ni
s.n
.(A
BSH
)(A
x)D
Q26
8880
AY
6887
52A
Y50
7401
—A
Y50
7484
SIU
Gac
kstr
oem
iaw
ein
dor
feri
(Her
zog)
Gro
lle
Au
stra
lia,
Stre
iman
n58
531
(NY)
AY
6082
14A
Y60
8281
AY
6080
28A
Y60
8063
AY
6079
43D
uke
Goe
beli
ella
corn
iger
a(M
itt.
)St
eph
.N
ewZ
eala
nd
,von
Kon
rat
2000
/197
(F)
——
—A
Y60
8064
AY
6079
44D
uke
Gre
eneo
thal
lus
gem
mip
aru
sH
asse
lA
rgen
tin
a,T
ierr
ad
elF
ueg
o,L
ong
3183
1( E
)A
Y87
7367
DQ
2689
23A
Y68
8780
AY
6887
92A
Y68
8828
SIU
Gym
nom
itri
onco
nci
nn
atu
m(L
igh
tf.)
Co
rda
Can
ada,
Dav
is42
4(D
UK
E)
AY
6082
15—
DQ
4396
86A
Y60
8065
AY
6079
45D
uke
Forrest et al.: Liverwort phylogeny 307
Page 7
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Gyr
oth
yra
un
der
woo
dia
na
M.H
ow
eC
anad
a,V
anco
uve
rIs
lan
d,F
orre
st59
3(A
BSH
)D
Q26
8881
DQ
2689
24D
Q26
8970
DQ
2689
85D
Q26
8997
SIU
Hap
lom
itri
um
and
inu
m(S
pru
ce)
R.M
.Sch
ust
.G
uad
elo
up
e,G
offi
net
1904
(CO
NN
)A
Y60
8216
AY
6082
82A
Y60
8029
AY
6080
66A
Y60
7946
Du
ke
Hap
lom
itri
um
blu
mei
(Nee
s)R
.M.S
chu
st.
Mal
aya,
Fu
ruki
384
(AB
SH)
AY
8773
68—
AY
5074
02A
Y50
7443
AY
5074
85SI
U
Hap
lom
itri
um
gibb
siae
(Ste
ph
.)R
.M.S
chu
st.(
1)N
ewZ
eala
nd
,En
gel
&vo
nK
onra
t24
919
(F)
—A
Y60
8283
AY
6080
30A
Y60
8067
AY
6079
47D
uke
Hap
lom
itri
um
gibb
siae
(Ste
ph
.)R
.M.S
chu
st.(
2)N
ewZ
eala
nd
,Mt.
Art
hu
r,St
otle
r&
Cra
nd
all-
Stot
ler
4541
(AB
SH)
AY
8773
69A
Y68
8753
AY
6887
81A
Y68
8793
—SI
U
Hap
lom
itri
um
hoo
keri
(Sm
.)N
ees
U.S
.A.,
Sch
ofie
ld95
224
( DU
KE)
*AY
3304
37A
Y60
8284
—A
Y60
8068
AY
3129
04D
uke
Hap
lom
itri
um
hoo
keri
(Sm
.)N
ees
Sco
tlan
d,P
erth
shir
e,L
ong
2804
2(E
)D
Q26
8882
DQ
2689
25—
—A
Y87
7398
SIU
Hap
lom
itri
um
mn
ioid
es(L
ind
b.)
R.M
.Sch
ust
.Ja
pan
,Fu
ruki
7408
(AB
SH)
DQ
2688
83D
Q26
8926
DQ
2689
71A
Y50
7444
AY
5074
86SI
U
Har
pan
thu
ssc
uta
tus
(F.W
eber
&D
.Mo
hr)
Spru
ceU
.S.A
.,R
isk
etal
.103
41(D
UK
E)
AY
6082
17—
DQ
4396
87A
Y60
8069
AY
6079
48D
uke
Hat
tori
anth
us
erim
onu
s(S
tep
h.)
R.M
.Sch
ust
.&In
ou
eJa
pan
,Fu
ruki
s.n
.(A
BSH
)D
Q26
8884
AY
6887
54A
Y50
7403
AY
5074
45A
Y50
7487
SIU
Her
bert
us
old
fiel
dian
us
(Ste
ph
.)R
od
way
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4580
(AB
SH)
AY
6886
89D
Q26
8927
AY
5074
04A
Y50
7446
AY
5074
88SI
U
Her
bert
us
saku
rai
(War
nst
.)S.
Hat
t.ss
par
ctic
us
Ino
ue
&St
eere
Ala
ska,
Stee
re74
(2)8
82(N
Y)
-Sa
rgen
tli
vin
gcu
ltu
re
colle
ctio
n,B
erke
ley
(UC)
AY
6082
16A
Y60
8282
AY
6080
31A
Y60
8066
AY
6079
46D
uke
Her
bert
us
subd
enta
tus
(Ste
ph
.)F
ulf
.E
cuad
or,
Dav
is34
2( D
UK
E)
AY
6082
19A
Y60
8286
DQ
4396
88A
Y60
8071
AY
6079
50D
uke
Her
zogo
bryu
mte
res
(Car
rin
gto
n&
Pea
rso
n)
Gro
lle
New
Zea
lan
d,B
ragg
ins
92/1
02(F
)A
Y60
8220
——
AY
6080
72A
Y60
7951
Du
ke
Hym
enop
hyt
onfl
abel
latu
m(L
abil
l.)D
um
ort
.N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r45
11(A
BSH
)A
Y87
7370
AY
6887
55A
Y50
7406
AY
5074
48A
Y50
7489
SIU
Hym
enop
hyt
onle
ptop
odu
m(H
oo
k.&
Tay
lor)
Step
h.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4492
(AB
SH)
DQ
2688
85D
Q26
8928
AY
5074
05A
Y50
7447
AY
5074
90SI
U
Isot
ach
isly
alli
i(R
.M.S
chu
st.)
Mit
t.N
ewZ
eala
nd
,En
gel
2182
5(F
)A
Y60
8221
AY
6082
87A
Y60
8032
AY
6080
73A
Y60
7952
Du
ke
Isot
ach
ism
ult
icep
s(L
ind
enb
.&G
ott
sch
e)G
ott
sch
eP
anam
a,St
otle
r&
Cra
nd
all-
Stot
ler
3478
(AB
SH)
AY
8773
71A
Y87
7385
AY
5074
07A
Y50
7449
AY
5074
91SI
U
Jen
sen
iaco
nn
iven
s(C
ole
nso
)G
roll
eN
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r45
35( A
BSH
)A
Y73
4736
AY
7347
48A
Y68
8782
AY
5074
50A
Y68
8829
SIU
Jen
sen
iasp
inos
a(L
ind
b.&
Go
ttsc
he)
Gro
lle
Ven
ezu
ela,
Mer
ida,
Vil
larr
eal
793
(AB
SH)
AY
7347
38A
Y73
4750
AY
7346
89A
Y73
4698
AY
7347
07SI
U
Jubu
lahu
tich
insi
ae(H
ook.
)D
um
ort
.ssp
.jav
anic
aSt
eph
.Ja
pan
,Kod
ama
s.n
.(A
BSH
)(A
x)D
Q26
8886
DQ
2689
29A
Y50
7408
AY
6887
94A
Y50
7492
SIU
Jubu
lape
nn
sylv
anic
a(S
tep
h.)
A.E
van
sU
.S.A
.,R
isk
1100
5(D
UK
E)
AY
6082
23A
Y60
8288
—A
Y60
8075
AY
6079
54D
uke
Jubu
lops
isno
vae-
zela
ndia
e(H
odgs
.&A
rnel
l)R
.M.S
chu
st.
New
Zea
lan
d,v
onK
onra
t99
/Feb
.#16
(F)
AY
6082
24A
Y60
8289
AY
6080
33A
Y60
8076
AY
6079
55D
uke
Jun
germ
ann
iacr
enu
lifo
rmis
Au
stin
U.S
.A.,
Ris
k11
014
(DU
KE)
AY
6082
26A
Y60
8290
—A
Y60
8078
AY
6079
57D
uke
Jun
germ
ann
iaex
sert
ifol
iaSt
eph
.ssp
.cor
dif
olia
(Du
mo
rt.)
Van
a
Can
ada,
Dav
is43
1(D
UK
E)
AY
6082
25—
DQ
4396
90A
Y60
8077
AY
6079
56D
uke
Jun
germ
ann
iale
ian
tha
Gro
lle
U.S
.A.,
Illi
no
is,S
totl
er&
Cra
nd
all-
Stot
ler
107
( AB
SH)
—A
Y68
8756
AY
5074
09A
Y50
7451
AY
5074
93SI
U
Jun
germ
ann
iasu
bell
ipti
ca(L
ind
b.e
xK
aal.)
Lev
ier
U.S
.A.,
Sch
ofie
ld11
1132
( DU
KE)
(mix
edco
llec
tio
n
wit
hB
leph
aros
tom
atr
ich
oph
yllu
m(L
.)D
um
ort
.)
——
—A
Y60
8050
AY
6079
29D
uke
Lei
ocol
eah
eter
ocol
pos
(Th
ed.e
xH
artm
.)H
.Bu
chU
.S.A
.,U
tah
,Sto
tler
&C
ran
dal
l-St
otle
r43
57( A
BSH
)D
Q26
8887
DQ
2689
30D
Q26
8972
DQ
2689
86D
Q26
8998
SIU
Lej
eun
eacl
adog
yna
A.E
van
sU
.S.A
,Dav
is63
(DU
KE)
——
—A
Y60
8079
AY
6079
58D
uke
308 the bryologist 109(3): 2006
Page 8
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Lep
icol
eaat
ten
uat
a(M
itt.
)St
eph
.N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r45
86(A
BSH
)D
Q26
8888
DQ
2689
31A
Y50
7410
AY
5074
52A
Y50
7494
SIU
Lep
icol
eaoc
hro
leu
ca(L
.f.e
xSp
ren
g.)
Spru
ceC
hil
e,D
reh
wal
d97
0188
(NY)
——
—A
Y60
8080
AY
6079
59D
uke
Lep
icol
eara
ra(S
tep
h.)
Gro
lle
Pap
ua
New
Gu
inea
,Nor
ris
6657
5(D
UK
E)
AY
6082
27A
Y60
8291
DQ
4396
91A
Y60
8081
AY
6079
60D
uke
Lep
idol
aen
acl
avig
era
(Ho
ok.
)D
um
ort
.N
ewZ
eala
nd
,En
gel
2305
2(F
)A
Y60
8228
AY
6082
92—
AY
6080
82A
Y60
7961
Du
ke
Lep
idoz
iare
ptan
s(L
.)D
um
ort
.Sa
rgen
tli
vin
gcu
ltu
reco
llect
ion
,Ber
kele
y(U
C)
AY
6082
29A
Y60
8293
—A
Y60
8083
AY
6079
62D
uke
Let
hoc
olea
glos
syph
ylla
(Sp
ruce
)G
roll
eE
cuad
or,
Dav
is25
9(D
UK
E)
AY
6082
30A
Y60
8294
DQ
4396
92A
Y60
8084
AY
6079
63D
uke
Lob
atir
icca
rdia
loba
ta(S
chif
fn.)
Fu
ruki
New
Zea
lan
d,G
len
ny
&K
inse
r45
81;d
et.
Cra
nd
all-
Sto
tler
(AB
SH)
AY
8773
73A
Y68
8757
AY
5074
21A
Y50
7462
AY
5075
07SI
U
Lop
hoco
lea
appa
lach
ian
us
R.M
.Sch
ust
.[C
hilo
scyp
hus
appa
lach
ian
us
(R.M
.Sch
ust
.)J.
J.E
ngel
&R
.M.S
chus
t.
U.S
.A.,
Dav
ison
&H
icks
2934
( DU
KE)
AY
6082
06A
Y60
8273
—A
Y60
8056
AY
6079
35D
uke
Loph
ocol
eahe
tero
phyl
la(S
chra
d.)
Du
mor
t.¼
Chi
losc
yphu
s
prof
undu
s(N
ees)
J.J.
En
gel&
R.M
.Sch
ust
.
USA
,IL
,Sto
tler
&C
ran
dal
l-St
otle
rs.
n.(
AB
SH)
DQ
2688
89D
Q26
8932
DQ
2689
73D
Q26
8987
DQ
2689
99SI
U
Lu
nu
lari
acr
uci
ata
(L.)
Lin
db
.M
exic
o,D
istr
ito
Fed
eral
,Lon
g29
553
(E)
DQ
2657
82D
Q26
8933
DQ
2860
11D
Q22
0685
DQ
2657
55E
Lu
nu
lari
acr
uci
ata
(L.)
Lin
db
.Sc
otl
and
,Eas
tL
oth
ian
,For
rest
520
(AB
SH)
DQ
2688
90D
Q26
8934
AY
6887
83A
Y68
8795
AY
6888
30SI
U
Mak
inoa
cris
pata
(Ste
ph
.)M
iyak
eC
hin
a,St
otle
r&
Cra
nd
all-
Stot
ler
4047
(AB
SH)
AY
8773
74A
Y87
7386
AY
8773
90A
Y87
7393
AY
8773
99SI
U
Man
nia
and
rogy
na
(L.)
A.E
van
sP
ort
uga
l,M
adei
ra,S
chil
l20
(E)
DQ
2657
83D
Q26
8935
DQ
2860
12D
Q22
0686
DQ
2657
56E
Man
nia
frag
ran
s(B
alb
is)
Fry
e&
L.C
lark
Swit
zerl
and
,Sch
ill
34(E
)D
Q26
5784
DQ
2689
36D
Q28
6013
DQ
2206
87D
Q26
5757
E
Mar
chan
tia
chen
opod
aL
.M
exic
o,V
erac
ruz,
Lon
g29
660
(E)
DQ
2657
85—
DQ
2860
14D
Q22
0688
DQ
2657
58E
Mar
chan
tia
infl
exa
Nee
s&
Mo
nt.
Mex
ico
,Ver
acru
z,L
ong
2962
8(E
)D
Q26
8891
DQ
2689
37D
Q26
8974
—D
Q26
9000
SIU
Mar
chan
tia
pale
acea
Ber
tol.
Mex
ico
,Ver
acru
z,L
ong
&G
arci
a29
683
(E)
DQ
2657
86D
Q26
8938
DQ
2860
15D
Q22
0689
DQ
2657
59E
Mar
chan
tia
poly
mor
pha
L.
[acc
esse
dfr
om
Gen
Ban
k]*A
F22
6020
*M68
929
*U87
079
*NC
0013
19*N
C00
1319
Mar
supe
lla
emar
gin
ata
(Eh
rh.)
Du
mo
rt.v
ar.a
quat
ica
(Lin
den
b.)
Du
mo
rt.
Fra
nce
,Fra
hm
90/5
89(F
)—
——
AY
6080
87A
Y60
7966
Du
ke
Mar
supi
diu
mla
tifo
liu
mR
.M.S
chu
st.
Co
sta
Ric
a,D
auph
in29
20(N
Y)
AY
6082
33A
Y60
8298
AY
6080
34A
Y60
8088
AY
6079
67D
uke
Mas
tigo
leje
un
eaau
ricu
lata
(Wil
son
&H
oo
k.)
Sch
iffn
.B
oli
via,
Chu
rch
ill
2127
5[a
cces
sed
fro
mG
enB
ank]
——
*AY
5480
70—
—
Mas
tigo
phor
ad
icla
dos
(Bri
d.e
xF
.Web
er)
Nee
sT
anza
nia
,Poc
s89
119/
V(F
)—
——
AY
6080
89A
Y60
7968
Du
ke
Met
zger
iaco
nju
gata
Lin
db
.U
.S.A
.,Il
lin
ois
,Hen
son
1247
( AB
SH)
AY
6886
90D
Q26
8939
AY
5074
11A
Y50
7453
AY
5074
95SI
U
Met
zger
iafr
uti
cola
Spru
ceE
cuad
or,
Dav
is36
1(D
UK
E)
AY
6082
34—
AY
6080
35A
Y60
8090
AY
6079
69D
uke
Met
zger
iafu
rcat
a(L
.)D
um
ort
.U
.S.A
.,N
ort
hC
aro
lina,
For
rest
&V
illar
real
620
(AB
SH)
DQ
2688
92D
Q26
8940
DQ
2689
75D
Q26
8988
DQ
2690
01SI
U
Met
zger
iali
ebm
ann
ian
aL
ind
enb
.&G
ott
sch
eV
enez
uel
a,M
erid
a,F
orre
st57
5(A
BSH
)D
Q26
8893
DQ
2689
41D
Q26
8976
DQ
2689
89D
Q26
9002
SIU
Moe
rcki
abl
ytti
i(M
oer
ch)
Bro
ckm
.U
.S.A
.,O
rego
n,W
agn
ers.
n.(
AB
SH)
—D
Q26
8942
AY
5074
12A
Y50
7454
AY
5074
96SI
U
Moe
rcki
afl
otov
ian
a(N
ees)
Sch
iffn
.U
.S.A
.,N
ewY
ork
,Kin
ser
&Sc
hu
ette
709
(AB
SH)
DQ
2688
94A
Y68
8758
AY
5074
13A
Y68
8796
AY
5074
97SI
U
Mon
ocle
ago
ttsc
hei
Lin
db
.M
exic
o,S
haw
1015
1(D
UK
E)
AY
6082
35A
Y60
8299
DQ
4396
94A
Y60
8091
AY
6079
70D
uke
Forrest et al.: Liverwort phylogeny 309
Page 9
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Mon
ocle
ago
ttsc
hei
Lin
db
.M
exic
o,V
erac
ruz,
Lon
g,G
arci
a&
De
Lu
na
2963
7(E
)D
Q26
5787
DQ
2689
43D
Q28
6016
DQ
2206
90D
Q26
5760
E
Mon
ocle
ago
ttsc
heiL
ind
b.ss
p.e
long
ata
Gra
dst
.&R
.Mu
esP
uer
toR
ico
,Sto
tler
&C
ran
dal
l-St
otle
r89
–60
(AB
SH)
AY
6886
91A
Y87
7388
AY
5074
14A
Y50
7455
AY
5074
98SI
U
Mon
osol
eniu
mte
ner
um
Gri
ff.
Ger
man
y(C
ult
.,G
ott
inge
nB
ota
nic
Gar
den
),
Gra
dst
ein
s.n
.(E)
DQ
2657
88D
Q26
8944
DQ
2860
17D
Q22
0691
—E
Myr
ioco
lea
irro
rata
Spru
ceE
cuad
or,
Gra
dst
ein
etal
.100
33B
[acc
esse
dfr
om
Gen
Ban
k]
——
*AY
5480
73—
—
Nar
dia
inse
cta
Lin
db
.U
.S.A
.,D
avis
432
( DU
KE)
(mix
edco
llec
tio
nw
ith
Lop
hoz
iasp
.in
det
.)
AY
6082
32A
Y60
8296
DQ
4396
93A
Y60
8086
AY
6079
65D
uke
Nar
dia
scal
aris
Gra
yU
.S.A
.,D
avis
438
( DU
KE)
AY
6082
36—
DQ
4396
95A
Y60
8092
AY
6079
71D
uke
Neo
hod
gson
iam
irab
ilis
(Per
ss.)
Per
ss.(
1)N
ewZ
eala
nd
,So
uth
Isla
nd
,Mal
colm
s.n
.(E)
DQ
2657
89D
Q26
8945
DQ
2860
18D
Q22
0692
DQ
2657
61E
Neo
hod
gson
iam
irab
ilis
(Per
ss.)
Per
ss.(
2)N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r45
42(A
BSH
)A
Y68
8692
AY
6887
59A
Y50
7415
AY
5074
56A
Y50
7499
SIU
Neo
tric
hoc
olea
biss
etii
(Mit
t.)
S.H
att.
Jap
an,I
nou
es.
n.(
F)
AY
6082
37A
Y60
8300
DQ
4396
96A
Y60
8093
AY
6079
72D
uke
Not
eroc
lad
aco
nfl
uen
s(H
oo
k.f.
&T
aylo
r)Sp
ruce
Arg
enti
na,
Tie
rra
del
Fu
ego
,Lon
g31
768
(E)
DQ
2688
95D
Q26
8946
DQ
2689
77D
Q26
8990
DQ
2690
03SI
U
Not
eroc
lad
aco
nfl
uen
s(H
oo
k.f.
&T
aylo
r)Sp
ruce
Ven
ezu
ela,
Mer
ida,
For
rest
566
(AB
SH)
AY
8773
76A
Y68
8760
AY
6887
84A
Y68
8797
AY
5075
00SI
U
Now
elli
acu
rvif
olia
(Dic
ks.)
Mit
t.U
.S.A
.,R
isk
&G
ross
1222
0(D
UK
E)
AY
6082
38A
Y60
8301
DQ
4396
997
AY
6080
94A
Y60
7973
Du
ke
Od
onto
leje
un
ealu
nu
lata
(Nee
s)Sc
hif
fn.
Bra
zil,
Zar
tman
1235
.2(D
UK
E)
AY
6082
39A
Y60
8302
DQ
4396
98A
Y60
8095
AY
6079
74D
uke
Od
onto
sch
ism
ad
enu
datu
m(N
ees)
Du
mo
rt.
U.S
.A.,
Hor
n18
09(D
UK
E)
AY
6082
40A
Y60
8303
AY
6080
36A
Y60
8096
AY
6079
75D
uke
Pal
lavi
cin
ialy
elli
i(H
oo
k.)
Car
ruth
.U
.S.A
.,A
rkan
sas,
Mar
shs.
n.(
AB
SH)
AY
7347
42A
Y68
8761
AY
5074
16A
Y68
8798
AY
5075
01SI
U
Pal
lavi
cin
iaru
bris
tipa
Sch
iffn
.A
ust
rali
a,C
argi
ll&
Vel
la50
7(A
BSH
)A
Y73
4744
AY
7347
53A
Y73
4693
AY
7347
02A
Y73
4711
SIU
Pal
lavi
cin
iaxi
phoi
des
(Ho
ok.
f.&
Tay
lor)
Tre
vis.
New
Zea
lan
d,N
ort
hIs
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4514
( AB
SH)
AY
7347
43A
Y73
4752
AY
7346
92A
Y73
4701
AY
7347
10SI
U
Pel
lia
appa
lach
ian
aR
.M.S
chu
st.
U.S
.A.,
Ala
bam
a,D
avis
on35
79(A
BSH
)D
Q26
8896
AY
6887
62A
Y68
8785
AY
6887
99A
Y68
8831
SIU
Pel
lia
end
ivii
foli
a(D
icks
.)D
um
ort
.Ja
pan
(SIU
gree
nh
ou
se),
Hig
uch
is.
n.(
AB
SH)
DQ
2688
97A
Y68
8763
AY
6887
86A
Y68
8800
AY
6888
32SI
U
Pel
lia
epip
hyll
a(L
.)C
ord
a(1
)U
.S.A
.,N
ort
hC
aro
lin
a,St
otle
r&
Cra
nd
all-
Stot
ler
4414
(AB
SH)
AY
6886
93A
Y68
8764
AY
6887
87A
Y50
7457
AY
5075
02SI
U
Pel
lia
epip
hyll
a(L
.)C
ord
a(2
)U
.S.A
.,R
isk
&G
ross
1223
1( D
UK
E)
AY
6082
42A
Y60
8305
——
—D
uke
Pel
tole
pis
quad
rata
(Sau
t.)
Mu
ll.F
rib
.N
orw
ay,P
rov.
Sogn
og
Fjo
rdan
e,L
ong
etal
.313
60(E
)D
Q26
5790
DQ
2689
47D
Q28
6019
DQ
2206
93—
E
Pet
alop
hyl
lum
ralf
sii
(Wil
son
)N
ees
&G
ott
sch
e
exL
ehm
.
U.K
.,J.
Pro
skau
ercu
ltu
revi
aD
.Mu
elle
r-T
AM
Uto
AB
SH(1
),to
Sarg
ent,
toB
erke
lycu
ltu
reco
llec
tio
n
(2)
( AB
SH)
AY
8773
77(1
)A
Y60
8306
(2)
AY
5074
17(1
)A
Y50
7458
(1)
AY
5075
03(1
)SI
U/D
uke
Ph
yllo
thal
lia
niv
icol
aE
.A.H
od
gs.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4537
(AB
SH)
AY
6886
94A
Y68
8765
AY
5074
18A
Y50
7459
AY
5075
04SI
U
Pla
gioc
has
ma
rups
etre
(J.R
.Fo
rst.
&G
.Fo
rst.
)St
eph
.P
ort
uga
l,M
adei
ra,S
chil
l5
(E)
DQ
2657
91D
Q26
8949
DQ
2860
20D
Q22
0694
DQ
2657
63E
Pla
gioc
has
ma
wri
ghti
iSu
ll.
Mex
ico
,Ver
acru
z,L
ong
etal
.296
36(E
)D
Q26
5792
DQ
2689
50D
Q28
6021
DQ
2206
95D
Q26
5764
E
310 the bryologist 109(3): 2006
Page 10
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Pla
gioc
hil
aau
stin
iiA
.Eva
ns
U.S
.A.,
Ris
k10
849
(DU
KE)
AY
6082
44—
DQ
4396
99A
Y60
8099
AY
6079
78D
uke
Ple
uro
zia
purp
ure
aL
ind
b.
U.S
.A.,
Sch
ofie
ld10
2815
(DU
KE)
—A
Y60
8307
AY
6080
37A
Y60
8100
AY
6079
79D
uke
Ple
uro
zia
purp
ure
aL
ind
b.
Bh
uta
n,L
ong
2886
8(E
)D
Q26
8899
DQ
2689
51A
Y87
7391
—A
Y87
7401
SIU
Pod
omit
riu
mm
alac
cen
se(S
tep
h.)
Cam
pb
.Si
nga
po
re,T
ans.
n.(
AB
SH)
—D
Q26
8952
DQ
2689
79D
Q26
8991
DQ
2690
05SI
U
Pod
omit
riu
mph
ylla
nth
us
(Ho
ok.
)M
itt.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4517
(AB
SH)
AY
7347
40A
Y68
8766
AY
5074
19A
Y50
7460
AY
5075
05SI
U
Por
ella
nav
icu
lari
s(L
ehm
.&L
ind
enb
.)P
feif
f.U
.S.A
.,C
alif
orn
ia,S
totle
r&
Cra
ndal
l-St
otle
r34
10(A
BSH
)A
Y68
8695
AY
6887
67A
Y50
7420
AY
5074
61A
Y50
7506
SIU
Por
ella
pin
nat
aL
.U
SA,G
offi
net
4744
(DU
KE)
AY
6082
45A
Y60
8308
—A
Y60
8101
AY
6079
80D
uke
Pre
issi
aqu
adra
ta(S
cop
.)N
ees
U.S
.A.,
Sch
ofie
ld10
5579
(DU
KE)
AY
6082
46A
Y60
8309
AY
3129
35A
Y60
8102
AY
6079
81D
uke
Pre
issi
aqu
adra
ta(S
cop
.)N
ees
Sco
tlan
d,S
ou
thE
bu
des
,Lon
g33
268
(E)
DQ
2657
93D
Q26
8953
——
—E
Pti
lid
ium
cili
are
(L.)
Ham
pe
Can
ada,
Sch
ofie
ld10
3486
(DU
KE)
AY
6082
47A
Y60
8310
AY
6080
38A
Y60
8103
AY
6079
82D
uke
Rad
ula
com
plan
ata
(L.)
Du
mo
rt.
Ru
ssia
,Bak
alin
[acc
.8]
(NY)
—A
Y60
8311
—A
Y60
8104
AY
6079
83D
uke
Rad
ula
perr
otte
tii
Go
ttsc
he
exSt
eph
.Ja
pan
,Miz
uta
ni
1503
0(F
)A
Y60
8248
AY
6083
12D
Q43
9700
AY
6081
05A
Y60
7984
Du
ke
Reb
ouli
ah
emis
phae
rica
(L.)
Rad
di
U.S
.A.,
Illi
no
is,F
orre
st53
1( A
BSH
)D
Q26
8900
AY
8773
83A
Y68
8788
AY
6888
01A
Y87
7340
2SI
U
Ric
card
iaca
pill
acea
(Ste
ph
.)M
een
ks&
C.D
eJo
ng
Ven
ezu
ela,
Mer
ida,
For
rest
558
(AB
SH)
AY
8773
78A
Y68
8768
AY
8773
92A
Y68
8802
AY
6888
33SI
U
Ric
card
iafu
coid
es(S
w.)
Sch
iffn
.Ja
mai
ca,D
avis
151
(DU
KE)
AY
6082
49A
Y60
8313
—A
Y60
8106
AY
6079
85D
uke
Ric
card
iam
ult
ifid
a(L
.)G
ray
Can
ada,
Van
cou
ver
Isla
nd
,For
rest
&B
adco
ck60
1(A
BSH
)—
DQ
2689
54D
Q26
8980
DQ
2689
92—
SIU
Ric
cia
cili
ifer
aL
ind
enb
.Sw
itze
rlan
d,S
chil
l37
(E)
DQ
2657
94D
Q26
8955
DQ
2860
22D
Q22
0696
DQ
2657
65E
Ric
cia
flu
itan
sL
.U
.S.A
.,Io
wa,
Cad
ys.
n.(
E)
DQ
2657
95D
Q26
8956
DQ
2860
23D
Q22
0697
DQ
2657
66E
Ric
cia
hu
eben
eria
na
Lin
den
b.s
sp.s
ull
ivan
tii
(Au
stin
)
R.M
.Sch
ust
.
U.S
.A.,
Illi
no
is,S
totl
er&
Cra
nd
all-
Stot
ler
92–1
75(A
BSH
)
—D
Q26
8957
AY
5074
22A
Y50
7463
AY
5075
08SI
U
Ric
cioc
arpo
sn
atan
s(L
.)C
ord
aA
ust
rali
a,V
icto
ria,
Sepp
elt
2046
7( E
,AD
T)
DQ
2657
96D
Q26
8958
DQ
2860
24D
Q22
0698
DQ
2657
67E
Rie
lla
hel
icop
hyl
la(B
ory
&M
on
t.)
Mo
nt.
Spai
n,A
lica
nte
,P
uch
e&
Gim
eno
(BM
)D
Q26
8901
DQ
2689
59D
Q26
8981
——
SIU
Sau
teri
aal
pin
a(N
ees)
Nee
sA
ust
ria,
Sch
ill
64(E
)D
Q26
5797
DQ
2689
60D
Q28
6025
DQ
2206
99D
Q26
5768
E
Scap
ania
nem
orea
(L.)
Gro
lle
(1)
U.S
.A.,
Dav
is12
4(D
UK
E)
AY
6082
50A
Y60
8314
AY
6080
39A
Y60
8108
AY
6079
87D
uke
Scap
ania
nem
orea
(L.)
Gro
lle
(2)
U.S
.A.,
Illi
no
is,S
totl
er&
Cra
nd
all-
Stot
ler
s.n
.
[ref
.no
.265
](A
BSH
)
AY
8773
79A
Y68
8769
AY
5074
23A
Y50
7464
AY
5075
09SI
U
Sch
iffn
eria
hya
lin
aSt
eph
.C
hin
a,Y
un
nan
,Lon
g33
706
( E)
DQ
2689
02D
Q26
8961
——
—SI
U
Sch
isto
chil
aap
pen
dic
ula
ta(H
oo
k.)
Du
mo
rt.e
xT
revi
s.N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r44
55(A
BSH
)A
Y68
8696
AY
6887
70A
Y50
7424
AY
5074
65A
Y50
7510
SIU
Sch
isto
chil
ale
hm
ann
ian
a(L
ehm
.&L
ind
enb
.)
Car
rin
gto
n&
Pea
rso
n
Au
stra
lia,
Stre
iman
n58
554
(NY)
AY
6082
51—
DQ
4397
02A
Y60
8109
AY
6079
88D
uke
Spha
eroc
arpo
ste
xan
us
Au
stin
(1)
U.S
.A.,
Gof
fin
et64
76( D
UK
E)
AY
6082
52—
—A
Y60
8110
AY
6079
89D
uke
Spha
eroc
arpo
ste
xan
us
Au
stin
(2)
U.S
.A.,
Illi
no
is,S
totl
er&
Cra
nda
ll-St
otle
r15
15(A
BSH
)A
Y68
8697
AY
6887
71A
Y50
7425
AY
5074
66A
Y50
7511
SIU
Sten
orrh
ipis
mad
agas
cari
ensi
s(S
tep
h.)
Gro
lle
Mad
agas
car,
Poc
s94
46/A
Q(F
)A
Y60
8253
AY
6083
15—
AY
6081
11A
Y60
7990
Du
ke
Forrest et al.: Liverwort phylogeny 311
Page 11
Tab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
Sym
phyo
gyn
aas
pera
Step
h.
Pan
ama,
Vil
larr
eal
668
( PM
A)
DQ
2689
03D
Q26
8962
AY
6887
89A
Y68
8803
AY
6888
34SI
U
Sym
phyo
gyn
abr
ongn
iart
iiM
on
t.E
cuad
or,
Dav
is34
1(D
UK
E)
AY
6082
54—
—A
Y60
8112
AY
6079
91D
uke
Sym
phyo
gyn
ah
ymen
oph
yllu
m(H
oo
k.)
Nee
s&
Mo
nt.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4452
(AB
SH)
AY
6886
98A
Y68
8772
AY
5074
26A
Y50
7467
AY
5075
12SI
U
Sym
phyo
gyn
au
nd
ula
taC
ole
nso
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4463
(AB
SH)
AY
8773
80A
Y68
8773
AY
6887
90A
Y68
8804
AY
6888
35SI
U
Tar
gion
iah
ypop
hyl
laL
.P
ort
uga
l,M
adei
ra,S
chil
l11
(E)
DQ
2657
98D
Q26
8964
DQ
2860
26D
Q22
0700
DQ
2657
69E
Tar
gion
iah
ypop
hyl
laL
.U
.S.A
.,C
alif
orn
ia,S
totle
r&
Cra
ndal
l-St
otle
r33
97(A
BSH
)—
DQ
2689
65A
Y50
7427
AY
6888
05A
Y50
7514
SIU
Tel
aran
eapu
lche
rrim
a(S
tep
h.)
R.M
.Sch
ust
.A
ust
rali
a,St
reim
ann
5955
4(N
Y)
AY
6082
56A
Y60
8316
DQ
4397
04A
Y60
8114
AY
6079
93D
uke
Tem
nom
apu
lche
llu
m(H
oo
k.)
Mit
t.ex
Bas
tow
New
Zea
lan
d,B
ragg
ins
92/8
7(F
)A
Y60
8257
AY
6083
17D
Q43
9705
AY
6081
15A
Y60
7994
Du
ke
Tet
ralo
phoz
iase
tifo
rmis
(Eh
rh.)
Sch
ljak
ov
Can
ada,
Dav
is42
2(D
UK
E)
AY
6082
58A
Y60
8318
—A
Y60
8116
AY
6079
95D
uke
Tre
ubi
ala
cun
osa
(Co
len
so)
Pro
sk.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4561
(AB
SH)
AY
6886
99A
Y68
8774
AY
5074
28A
Y50
7468
—SI
U
Tre
ubi
apy
gmae
aR
.M.S
chu
st.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4582
(AB
SH)
DQ
2689
05A
Y68
8775
AY
5074
29A
Y50
7469
AY
5075
15SI
U
Tri
and
roph
yllu
msu
btri
fid
um
(Ho
ok.
&T
aylo
r)
Fu
lfo
rd&
Hat
cher
Ven
ezu
ela,
Ric
ard
iet
al.9
730/
T(F
)A
Y60
8259
AY
6083
19D
Q43
9706
AY
6081
17A
Y60
7996
Du
ke
Tri
choc
olea
tom
ento
sa(S
w.)
Go
ttsc
he
Ecu
ado
r,D
avis
368
(DU
KE)
AY
6082
61A
Y60
8320
AY
6080
40A
Y60
8119
AY
6079
98D
uke
Tri
tom
aria
quin
qued
enta
ta(H
ud
s.)
H.B
uch
U.S
.A.,
Sch
ofie
ld10
6093
(DU
KE)
AY
6082
61A
Y60
8321
DQ
4397
07A
Y60
8119
AY
6079
98D
uke
Ver
doo
rnia
succ
ule
nta
R.M
.Sch
ust
.N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r46
02(A
BSH
)A
Y68
700
AY
6887
76A
Y50
7430
AY
5074
70A
Y50
7516
SIU
Wie
sner
ella
den
uda
ta(M
itt.
)St
eph
.N
epal
,Lon
g30
337
(E)
DQ
2657
99D
Q26
8966
DQ
2860
27D
Q22
0701
—E
Xen
oth
allu
svu
lcan
icol
aR
.M.S
chu
st.
New
Zea
lan
d,S
totl
er&
Cra
nd
all-
Stot
ler
4580
(AB
SH)
AY
8773
81A
Y68
8777
AY
5074
31A
Y50
7471
AY
5075
17SI
U
Cha
raco
nn
iven
sSa
lzm
.ex
A.B
rau
n[a
cces
sed
fro
mG
enB
ank]
—*A
F40
8200
*AF
0971
6—
—
Chl
amyd
omon
asre
inh
ard
tii
Dan
g.[a
cces
sed
fro
mG
enB
ank]
*AF
1834
63—
*BK
0005
4—
*U17
357
Cho
rell
avu
lgar
isB
eij.
[acc
esse
dfr
om
Gen
Ban
k]—
—*A
F49
9684
—*A
B00
1684
Nep
hro
selm
isol
ivac
eaF
.Ste
in[a
cces
sed
fro
mG
enB
ank]
—*A
F11
0138
*NC
0009
27*N
C00
0927
*NC
0009
27
An
dre
aea
roth
iiF
.Web
er&
D.M
oh
rU
.S.A
.,Sh
aw11
565
(DU
KE)
AY
3128
61A
Y31
2862
AY
6080
25A
Y31
2866
AY
3128
63D
uke
An
dre
aeob
ryu
mm
acro
spor
um
Stee
re&
B.M
.Mu
rray
[acc
esse
dfr
om
Gen
Ban
k]*A
Y33
0426
—*A
F23
1059
*AF
3069
53*A
Y31
2893
Den
droh
ypop
tery
giu
mar
busc
ula
(Bri
d.)
Kru
ijer
Ch
ile,
Hol
z&
Fra
nza
rin
gC
H00
–80
(NY)
AY
6082
09A
Y60
8276
AY
6080
27A
Y60
8059
AY
6079
38D
uke
Hoo
keri
alu
cen
s(H
edw
.)Sm
.U
.S.A
.,B
uck
3771
4(N
Y)
AY
3304
39*Z
9896
9*A
Y31
2931
AJ2
5131
6A
Y31
2906
Du
ke
Mn
ium
hor
nu
mH
edw
.[a
cces
sed
fro
mG
enB
ank]
*AY
3304
41*A
Y31
2879
*AF
2268
20*A
F02
3796
*AY
3129
08
Pol
ytri
chu
mpa
llid
iset
um
Fu
nck
[acc
esse
dfr
om
Gen
Ban
k]*A
Y33
0445
*AY
3128
84*A
Y31
2934
*AF
3069
56*A
Y31
2914
Spha
gnu
mcu
spid
atu
mE
hrh
.ex
Ho
ffm
.[a
cces
sed
fro
mG
enB
ank]
*AY
3095
06*A
Y30
9560
*AF
2318
86*A
F47
8246
*AY
3096
08
Spha
gnu
mpa
lust
reL
.[a
cces
sed
fro
mG
enB
ank]
*AY
3314
51*A
Y31
2888
*AF
2318
87*A
F23
1892
*AY
3129
20
Tak
akia
cera
toph
ylla
(Mit
t.)
Gro
lle
U.S
.A.,
Ala
ska,
Sch
ofie
lds.
n.[
31A
ug
1994
](A
BSH
)D
Q26
8904
DQ
2689
63—
DQ
2689
93—
SIU
Tak
akia
lepi
dozi
oid
esS.
Hat
t.&
Ino
ue
[acc
esse
dfr
om
Gen
Ban
k]*A
F19
7061
*AY
3128
89*A
F24
4565
*AF
3069
50*A
Y31
2921
An
thoc
eros
angu
stu
sSt
eph
.[a
cces
sed
fro
mG
enB
ank]
——
*AB
0874
49*A
B08
7443
*AB
0874
34
312 the bryologist 109(3): 2006
Page 12
Criterion and hierarchical Likelihood Ratio Tests. For
the heterogeneous analysis data partitions were made
according to gene (five partitions; BI5), and the
analyses were run for two million generations.
Sequences were also partitioned by codon, resulting in
a 14-partition model (BI14); analyses using this model
were run for 5 million generations. The burnin for
each run was determined by visualizing the curve of
likelihood versus generations in Microsoft Excel. Trees
from the burnin were disgarded and a majority rule
consensus of the remaining trees was computed using
PAUP*. Posterior probabilities (PP) for all runs for
each of the Bayesian analyses were compared to check
that convergence was reached. We regard Bayesian
support values as significant when they are 95% (or
0.95) or higher (e.g., Cox et al. 2004), although lower
values have been accepted by some authors (e.g.,
Nickrent et al. 2005, who used a value of 0.90).
RESULTS
Parsimony analysis found 128 MPTs 22,302 steps
long, consistency index 0.2310 (0.2000 excluding
uninformative characters), retention index 0.6717 and
rescaled consistency index 0.1552. Bootstrap support
(BS) values greater than 50% were generated for 181
of the 208 potential nodes (i.e., 87%), while 129 nodes
had BS values of 80% or above.
The results are robust to different sampling
methods; i.e., the topologies resolved by MP and
Bayesian analyses are mostly congruent, although
there are some slight differences in the placements of
Sphaerocarpales Cavers, Neohodgsonia Perss., Cya-
thodium Kunze, Allisonia Herzog, Cryptothallus
Malmb., Verdoornia R. M. Schust. and Jungermannia
L. In most cases the alternate placements have PPs of
less than 95%. (It should be noted that our use of
genus rather than species names here and in the
discussions that follow is simply for convenience and
does not imply monophyly of these genera, a
consideration that in most instances is beyond the
scope of this paper.)
Liverworts are resolved as monophyletic with
100% support in MP BS and all Bayesian analyses
(Figs. 1, 3). Relationships resolved among outgroup
taxa are comparable to those resolved by other
analyses focused on these groups (e.g., see Cox et al.
2004; Duff et al. 2004; Wolf et al. 2005) (Fig. 3). ThreeTab
le1.
Co
nti
nu
ed.
Tax
on
Vo
uch
erin
form
atio
nN
rL
SU
na
d5
rbcL
rps4
psb
AL
ab
An
thoc
eros
pun
ctat
us
L.
[acc
esse
dfr
om
Gen
Ban
k]*A
F22
6035
—*U
8706
3—
—
Lei
ospo
roce
ros
du
ssii
(Ste
ph
.)H
asse
l[a
cces
sed
fro
mG
enB
ank]
—*A
Y89
4803
*AY
4630
52—
—
Meg
acer
osfl
agel
lari
s(M
itt.
)St
eph
.N
ewZ
eala
nd
,Sto
tler
&C
ran
dal
l-St
otle
r44
78(A
BSH
)A
Y87
7375
AY
8773
87*A
Y46
3040
—A
Y87
7400
SIU
Ph
ymat
ocer
osbu
lbic
ulo
sus
(Bro
t.)
Sto
tler
,W.T
.Do
yle
&C
ran
d.-
Sto
tler
U.S
.A.,
Cal
ifo
rnia
,Doy
les.
n.[
14A
pr
2004
](m
ixed
wit
hP
hae
ocer
ospe
arso
nii
(M.H
ow
e)P
rosk
.)
(CA
NB)
DQ
2688
98D
Q26
8948
DQ
2689
78—
DQ
2690
04SI
U
Equ
iset
um
telm
atei
aE
hrh
.[a
cces
sed
fro
mG
enB
ank]
—*A
J130
749
*AF
3135
80—
—
Psi
lotu
mn
ud
um
(L.)
P.B
eau
v.[a
cces
sed
fro
mG
enB
ank]
—*A
J012
794
*AP
0046
38*A
P00
4638
*AP
0046
38
Forrest et al.: Liverwort phylogeny 313
Page 13
major lineages are supported within the liverworts
(Figs. 1, 2), in agreement with previous multigene
studies (Crandall-Stotler et al. 2005; Davis 2004;
Forrest & Crandall-Stotler 2004, 2005; Forrest et al.
2005a; He-Nygren et al. 2006), as well as some single
locus studies (e.g., Heinrichs et al. 2005). These
correspond to the classes Haplomitriopsida (including
the Haplomitriales Schljakov and Treubiineae Stotler
& Crand.-Stotler) (99% MP BS; 100% BI1, BI5, BI14
PP), Marchantiopsida (including the Blasiales) (100%
MP BS; 100% BI1, BI5, BI14 PP) and Jungerman-
niopsida (100% MP BS; 100% BI1, BI5, BI14 PP).
Table 2. Model selection results for each partition using the hierarchical likelihood ratio test and Akaike’s Information Criterion as
implemented in the programme MrModeltest (Nylander 2002). GTR: General time-reversible, Rodriguez et al. (1990); Base: Base
frequencies (A, C, G); Rmat: Rate matrix (AC, AG, AT, CG, CT); Shape: value of a in gamma distribution; Pinvar: proportion of
invariant characters.
Partition No.
bases
Model -ln likelihood Optimised parameters
26S 914 GTRþIþG 9697.6152 Base¼(0.2622 0.2261 0.3260) Rmat¼(0.6895 2.0192 0.5979 0.6695 7.9856)
Rates¼gamma Shape¼0.4493 Pinvar¼0.3285
nad5 1650 GTRþIþG 19033.2910 Base¼(0.2624 0.2156 0.1950) Rmat¼(2.0648 3.9320 0.4061 1.1277 6.2290)
Rates¼gamma Shape¼0.7947 Pinvar¼0.1056
nad5 codon 1 356 GTRþG 2580.6216 Base¼(0.2835 0.1663 0.2862) Nst¼6 Rmat¼(6.1922 4.1587 0.9291 1.9858 58.6202)
Rates¼gamma Shape¼0.5026 Pinvar¼0
nad5 codon 2 355 GTRþG 2550.1533 Base¼(0.1987 0.3175 0.1657) Nst¼6 Rmat¼(1.1166 3.9638 0.2874 2.3679 17.9541)
Rates¼gamma Shape¼0.4080 Pinvar¼0
nad5 codon 3 356 GTRþG 5306.3105 Base¼(0.2572 0.1717 0.1520) Nst¼6 Rmat¼(2.7770 6.7791 0.2012 0.9236 12.4265)
Rates¼gamma Shape¼1.5147 Pinvar¼0
nad5 intron 583 GTRþG 7242.5244 Base¼(0.3030 0.2221 0.2218) Nst¼6 Rmat¼(1.1745 2.0723 0.3858 0.5732 2.0765)
Rates¼gamma Shape¼0.8244 Pinvar¼0
rbcL 1308 GTRþIþG 37593.5625 Base¼(0.3074 0.1426 0.1275) Rmat¼(1.4295 6.5097 0.4708 2.3036 8.0060)
Rates¼gamma Shape¼0.7429 Pinvar¼0.3483
rbcL codon 1 436 GTRþIþG 6400.7085 Base¼(0.1984 0.2228 0.3699) Nst¼6 Rmat¼(3.8412 1.5426 0.8986 1.1002 9.9526)
Rates¼gamma Shape¼0.4093 Pinvar¼0.3814
rbcL codon 2 436 GTRþIþG 3722.9866 Base¼(0.2728 0.2612 0.1935) Nst¼6 Rmat¼(1.8161 2.7315 2.0852 6.9783 8.7538)
Rates¼gamma Shape¼0.4441 Pinvar¼0.3869
rbcL codon 3 436 GTRþIþG 26206.2598 Base¼(0.3218 0.1187 0.0834) Nst¼6 Rmat¼(0.9391 8.6187 0.2608 0.8943 6.8754)
Rates¼gamma Shape¼2.6177 Pinvar¼0.0225
rps4 554 GTRþIþG 22272.8828 Base¼(0.3950 0.1352 0.1493) Rmat¼(1.5579 7.1590 0.2418 1.8934 5.7926)
Rates¼gamma Shape¼0.9136 Pinvar¼0.1500
rps4 codon 1 185 GTRþIþG 5637.0337 Base¼(0.4084 0.1907 0.1739) Nst¼6 Rmat¼(1.9834 6.5365 0.2095 1.2738 4.2443)
Rates¼gamma Shape¼1.1690 Pinvar¼0.1914
rps4 codon 2 185 GTRþIþG 4156.6079 Base¼(0.3201 0.2109 0.1530) Nst¼6 Rmat¼(1.4178 8.1437 0.4568 1.8675 6.2222)
Rates¼gamma Shape¼0.6139 Pinvar¼0.1751
rps4 codon 3 184 GTRþIþG 12118.2607 Base¼(0.4211 0.0860 0.1342) Nst¼6 Rmat¼(1.5660 7.3912 0.1875 2.9779 8.5439)
Rates¼gamma Shape¼3.6568 Pinvar¼0.0056
psbA 1072 GTRþIþG 20777.5352 Base¼(0.2586 0.1902 0.1554) Rmat¼(1.0597 11.0235 1.7899 0.8227 17.6730)
Rates¼gamma Shape¼0.7975 Pinvar¼0.4675
psbA codon 1 358 GTRþIþG 2747.6267 Base¼(0.2426 0.1604 0.3071) Nst¼6 Rmat¼(5.6606 7.8798 1.0402 1.3272 36.2863)
Rates¼gamma Shape¼0.3280 Pinvar¼0.4837
psbA codon 2 357 GTRþIþG 1439.5475 Base¼(0.2223 0.2417 0.1909) Nst¼6 Rmat¼(1.1284 2.6137 1.6465 6.4758 9.6332)
Rates¼gamma Shape¼0.7564 Pinvar¼0.5285
psbA codon 3 357 GTRþIþG 16174.3984 Base¼(0.2739 0.1854 0.0581) Nst¼6 Rmat¼(0.3012 11.9740 0.6004 0.1405 5.7104)
Rates¼gamma Shape¼1.8966 Pinvar¼0.0904
combined 5498 GTRþIþG 114648.9922 Base¼(0.2943 0.1837 0.1556) Rmat¼(1.2999 6.6361 0.5492 1.4833 7.0032)
Rates¼gamma Shape¼0.6438 Pinvar¼0.3107
314 the bryologist 109(3): 2006
Page 14
The earliest divergence involves a split between
Haplomitriopsida and all other liverworts. Haplomi-
triopsida comprise three extant genera, with Treubia
and Apotreubia forming a monophyletic Treubiaceae,
sister to Haplomitrium Nees (Fig. 4).
The Blasiaceae H. Klinggr., which include two
monospecific genera, Blasia L. and Cavicularia Steph.,
are sister to all the Marchantiopsida or complex
thalloids (Fig. 5). Within the complex thalloids as
traditionally defined, levels of parsimony bootstrap
support are low, although there are significant
posterior probabilities. The branching order of the
earliest divergences in this lineage (Sphaerocarpales/
Neohodgsonia/Lunularia Adans.) varies according to
which analytical criterion is used, with the Sphaer-
ocarpales involved in the earliest divergence under
MP (Fig. 2) and BI1, while Neohodgsonia is involved
under more model-based approaches (BI5 and BI14)
(Fig. 5). However, support for either topology is low
(55% MP BS; 52% BI1 PP for Sphaerocarpales
diverging first; 93% BI5, 97% BI14 PP for Neo-
hodgsonia first). A Marchantia L./Preissia Corda
lineage is supported in all analyses as sister to all
remaining complex thalloids (93% MP BS; 100% BI1,
BI5, BI14 PP). Monoclea Hook., which is nested
within the crown group, receives varying amounts of
support as sister to Dumortiera Nees (55% MP BS;
100% BI1, BI5, BI14 PP).
The level of molecular evolution within the
complex thalloids is lower than for any other lineage,
with the notable exception of the crown group taxon
Cyathodium, which occurs on a comparatively long
branch (Fig. 2). Cyathodium is resolved as sister to
Exormotheca Mitt./Corsinia Raddi in most analyses
(without MP BS support; 100% PP BI1, 99% BI5), but
is sister to Monosolenium Griff. in the 14-partition
Bayesian analysis (91% PP).
The Jungermanniopsida are split into two
supported lineages (Figs. 1, 2), one comprising most
of the simple thalloid liverworts (Fig. 6), including all
taxa of the Pelliineae Schljakov, Fossombroniineae
Stotler & Crand.-Stotler, Phyllothalliineae R. M.
Schust. and Pallaviciniineae R. M. Schust. except
Verdoornia (99% MP BS; 100% BI1, BI5, BI14), and
the other comprising the leafy liverworts, and the
simple thalloids of the Metzgeriineae Schljakov plus
Verdoornia (89% MP BS; 100% BI1, BI5, BI14). This
leafy/Metzgerialean lineage contains two supported
lineages, one comprising the Metzgeriaceae H.
Klinggr., Aneuraceae H. Klinggr., Verdoornia and the
leafy liverwort Pleurozia Dumort. (Fig. 7) (99% MP
BS; 100% BI1, BI5, BI14), and the other, all remaining
taxa of the Jungermanniidae, i.e., leafy liverworts
(93% MP BS; 100% BI1, BI5, BI14).
Within the Jungermanniidae, several clades are
supported under parsimony, likelihood and Bayesian
approaches. The clade designated as Leafy I in Davis
(2004) is supported as monophyletic and excludes a
Ptilidium Nees/Neotrichocolea S. Hatt. clade in
Bayesian analyses (Fig. 8) (100% PP in BI1, BI5,
BI14). Under a parsimony criterion, the position of
this Ptilidium/Neotrichocolea clade is equivocal (e.g.,
unsupported as sister to Leafy I in Fig. 2). Other
lineages that are not supported as part of Leafy I
under a parsimony bootstrap criterion, although are
part of the clade in the MP strict consensus tree (not
Figure 1. Diagram of relationships of major clades. Numbers above the branches are Maximum Parsimony BS values/homogeneous
Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Forrest et al.: Liverwort phylogeny 315
Page 15
Figure 2. Phylogram (one of 128 equally parsimonious trees), with major clades (as in Fig. 1) highlighted. Taxa specifically discussed
in the text are annotated. Thinner branches are absent from the strict consensus tree. This tree is also available in NEXUS format
from TreeBASE (see methods section).
316 the bryologist 109(3): 2006
Page 16
shown), are a Porella L./Ascidiota C. Massal. clade
(100% MP BS, BI1, BI5, BI14 PP), a Jubulopsis R. M.
Schust./Lepidolaena Dumort./Gackstroemia Trevis.
clade (98% MP BS, 100% BI1, BI5, BI14 PP) and
Goebeliella Steph. The rest of the Leafy I taxa form a
large, well-supported monophyletic group comprising
several representatives of the Lejeuneaceae, Jubula
Dumort., Frullania and Radula Dumort. (81% MP
BS; 100% BI1, BI5, BI14).
The remaining leafy liverworts are supported in a
single large clade designated as Leafy II (Davis 2004)
(Figs. 1, 2, 9) (89% MP BS; 100% BI1, BI5, BI14).
Although the Neotrichocolea/Ptilidium clade resolves
with Leafy II in Bayesian analyses, this position is not
strongly supported (88% BI1, 95% BI5, 88% BI14
PP). The first well-supported divergence from the
main clade in Leafy II, then, involves Schistochila
Dumort., with Temnoma Mitt. forming the second
divergence. Trichocolea Dumort., Plagiochila, Lopho-
colea (Dumort.) Dumort., Chiloscyphus Corda, Tri-
androphyllum Fulford & Hatcher, Mastigophora Nees,
Herbertus Gray, Lepicolea Dumort., Telaranea Spruce
ex Schiffn., Lepidozia (Dumort.) Dumort., Bazzania
Gray and Acromastigum A. Evans form a mono-
phyletic group that is referred to herein as Clade A
(Fig. 9) (72% MP BS; 100% BI1, BI5, BI14). Their
sister group (82% MP BS; 100% BI1, BI5, BI14) is
split into two lineages; one, designated as Clade B,
includes Adelanthus Mitt., Odontoschisma (Dumort.)
Dumort., Schiffneria Steph., Nowellia Mitt., Cepha-
lozia (Dumort.) Dumort., Herzogobryum Grolle,
Stenorrhipis Herzog, Cephaloziella (Spruce) Schiffn.,
Tetralophozia (R. M. Schust.) Schjakov, Anastrophyl-
lum (Spruce) Steph., Tritomaria Loeske, Scapania
(Dumort.) Dumort. and Diplophyllum (Dumort.)
Dumort. (55% MP BS; 100% BI1, BI5, BI14), and the
other, Clade C, includes Marsupidium Mitt., Leth-
ocolea Mitt., Balantiopsis Mitt., Isotachis Mitt., Ca-
Figure 3. Outgroup relationships (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are
Maximum Parsimony BS values/homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Figure 4. Haplomitriopsida (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are
Maximum Parsimony BS values/ homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Forrest et al.: Liverwort phylogeny 317
Page 17
lypogeia Raddi, Anthelia (Dumort.) Dumort., Leioco-
lea (Mull. Frib.) H. Buch, Jungermannia, Harpanthus
Nees, Gyrothyra M. Howe, Nardia Gray, Marsupella
Dumort. and Gymnomitrion Corda (58% MP BS;
100% BI1, BI5, BI14).
DISCUSSION
The recent extremely rapid progress in our
understanding of liverwort phylogeny through the
application of molecular systematic methods has
resolved many questions regarding early divergences,
and allowed the reinterpretation of morphological,
ultrastructural and biochemical data in the light of
robust evolutionary hypotheses. By combining data
from some of these pivotal studies, as well as new
data, into a supermatrix, we have been able to solidify
these separate pieces of phylogenetic evidence into
one robust framework. Many terminal clades, how-
ever, still lack internal resolution. Some of the more
noteworthy relationships that were either confirmed
or newly resolved within the major lineages are
discussed individually below.
Confirmation of previous results. Many of the
clades supported in the analyses reported herein have
also been resolved in previous studies (e.g., Boisselier-
Dubayle et al. 2002; Crandall-Stotler et al. 2005; Davis
2004; Forrest & Crandall-Stotler 2004, 2005; Hein-
richs et al. 2005; He-Nygren et al. 2004, 2006; Long et
al. 2000; Wheeler 2000). Some of the relationships
revealed by these earlier studies were unexpected and
Figure 5. Marchantiopsida (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are
Maximum Parsimony BS values/ homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
318 the bryologist 109(3): 2006
Page 18
have remained somewhat controversial. For example,
until the studies of Davis (2004) and He-Nygren et al.
(2004), the alliance of Pleurozia with the Metzger-
iineae had never been postulated. Nonetheless, this
unexpected relationship has subsequently been re-
solved in several other analyses (Crandall-Stotler et al.
2005; Forrest et al. 2005a; Heinrichs et al. 2005) and is
further supported in our current analyses (Fig. 7). In
addition, the position of a Treubia and Haplomitrium
clade as the first diverging lineage of the extant
liverworts (Figs. 1, 4), suggested in Forrest and
Crandall-Stotler (2004), cemented in Forrest and
Crandall-Stotler (2005) and Crandall-Stotler et al.
(2005), and taxonomically formalized in Heinrichs et
al. (2005), was almost unprecedented in liverwort
classification. A relationship between Haplomitrium
Figure 6. Simple thalloids clade I (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are
Maximum Parsimony BS values/ homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Figure 7. Simple thalloids clade II (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are
Maximum Parsimony BS values/homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Forrest et al.: Liverwort phylogeny 319
Page 19
and Treubia, although not reflected in contemporary
classifications, was first suggested by Goebel (1898).
Upon review, this relationship is supported by several
fundamental morphological and ultrastructural fea-
tures that have been reconstructed as plesiomorphic
for hepatics (Crandall-Stotler et al. 2005). These
include mucilage-secreting epidermal cells, massive
blepharoplasts, scattered gametangia and tetrahedral
apical cells. The assertion of He-Nygren et al. (2006)
that tetrahedral apical cells are a derived feature of
this clade has not been supported here, or in previous
analyses (Crandall-Stotler et al. 2005; Forrest &
Crandall-Stotler 2004, 2005; Heinrichs et al. 2005).
The alliance of the Blasiales with the complex
thalloids, as resolved by Davis (2004), Forrest and
Crandall-Stotler (2004, 2005), Crandall-Stotler et al.
(2005), Forrest et al. (2005a), Heinrichs et al. (2005)
and He-Nygren et al. 2006, although suggested by the
ultrastructural work of Pass and Renzaglia (1995), was
also not indicated in any of the morphologically based
classifications of the group. The nesting of Monoclea
well within the complex thalloids was first suggested
by the analyses of Lewis et al. (1997) and has since
been supported by Wheeler (2000), Boisselier-Du-
bayle et al. (2002), Davis (2004), He-Nygren et al.
(2004, 2006), Forrest and Crandall-Stotler (2004,
2005), Crandall-Stotler et al. (2005) and Heinrichs et
al. (2005) (Fig. 5). This phylogenetic placement is in
sharp contrast to its traditional morphologically based
placement as an isolated lineage at the base of the
complex thalloids (Crandall-Stotler & Stotler 2000;
Schuster 1984, 1992). The large genus Radula, tradi-
tionally postulated to represent an isolated lineage
within the leafy liverworts, is instead nested within the
Porellales (R. M. Schust.) Schljakov emend. Stotler &
Crand.-Stotler (Fig. 8), as also shown in the
phylogenies of Ahonen (2004), Davis (2004) and He-
Nygren et al. (2004, 2006).
At lower taxonomic levels, many genera have
been resolved both here and in earlier studies as
paraphyletic, accentuating the need for substantial
taxonomic revision in many lineages, including
Asterella P. Beauv. in the complex thalloids (Long et
al. 2000), Fossombronia Raddi, Pallavicinia Gray and
Symphyogyna Nees & Mont. in the simple thalloids
(Forrest & Crandall-Stotler 2004, 2005), and Junger-
mannia in the leafy liverworts (this study, Fig. 9).
Paraphyly is also evident in many of the large,
morphologically defined families, including the
Marchantiaceae (Bisch.) Lindl., Pallaviciniaceae Mig.,
Jungermanniaceae Rchb., Geocalycaceae H. Klinggr.
and Lophoziaceae Cavers. Future comprehensive
molecular analyses that sample widely within these
larger genera and families are needed to satisfactorily
resolve their within-group relationships.
Complex thalloids. Based on patterns of short
branch lengths, rates of molecular evolution in loci
from both the chloroplast and the mitochondrion
Figure 8. Leafy clade I (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are Maximum
Parsimony BS values/ homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
320 the bryologist 109(3): 2006
Page 20
appear to be far slower in the complex thalloid
liverworts than in the other lineages (Fig. 2). These
reduced rates, which correspond to a lower number of
variable characters, are probably responsible for the
low numbers of supported nodes in this group.
Studies using only the marchantiopsid taxa rooted on
the Haplomitriopsida (Long et al. 2005, in prep.)
produce congruent topologies, and the morphological
evolution of the group is discussed in more detail
therein. Regarding the relative contributions of the
five different loci to the complex thalloid topology,
the nr LSU region, which was utilized by Boisselier-
Dubayle et al. (2002) and Wheeler (2000), although
contributing many of the parsimony-informative
characters, actually performed comparatively poorly
in resolving the relationships of taxa within the group.
For example, morphology strongly supports the
Ricciaceae Rchb. as a natural taxon (Bischler 1998).
Figure 9. Leafy clade II (based on the topology from the 5 partition Bayesian analysis). Numbers above the branches are Maximum
Parsimony BS values/ homogeneous Bayesian PPs/5-partition Bayesian PPs/14-partition Bayesian PPs.
Forrest et al.: Liverwort phylogeny 321
Page 21
However, based purely on nr LSU data (Boisselier-
Dubayle et al. 2002: fig. 2) they form a polyphyletic
assemblage, in which Ricciocarpos Corda is well
separated from Riccia L. and sister to Monosolenium.
Only when Boisselier-Dubayle et al. (2002) added
morphological data to the analysis did they recover a
clade of Riccia/Ricciocarpos/Oxymitra Bisch. ex Lin-
denb. (65% MP BS). Data from the rbcL gene alone
(this study; not shown) give an unambiguously
monophyletic Ricciaceae, as does analysis of the five
loci combined (Fig. 5) (84% MP BS; 100% BI1, BI5,
BI14 PP).
Early divergences within the complex thalloid
lineages remain equivocal in our analyses, although
they always include Sphaerocarpos Boehm. and Neo-
hodgsonia. Conflicting results for the branching order
of these early branches, albeit with far lower levels of
taxon sampling, were also apparent in Forrest and
Crandall-Stotler (2004, 2005) and Crandall-Stotler et
al. (2005). Despite similarly equivocal resolution of
these taxa in the analyses of He-Nygren et al. (2006)
these authors nonetheless treat Sphaerocarpos as a
separate order, the Sphaerocarpales, because of its
distinctive morphology, while implying a nested
position for Neohodgsonia in the Marchantiales.
Clearly, further work is needed to resolve the
relationships of these taxa.
Often classified in its own order, the Monocleales
(Nees) A. B. Frank, Monoclea has traditionally been
viewed as the link between the complex thalloid and
simple thalloid liverworts because of its blend of
characters from both (Schuster 1984). Its alliance with
the complex thalloids was deemed remote. Although
its pattern of involucral development (Crandall-
Stotler & Stotler 2000; Johnson, 1904; Leitgeb 1877),
Conocephalum-like antheridial receptacle and march-
antioid antheridial ontogeny (Schuster 1984), idio-
blastic oil cells and monoplastidic meiosis (Renzaglia
et al. 1994) link Monoclea to the complex thalloids,
sporophyte characters in particular suggested a
primitive position for the taxon within the complex
thalloid lineage.
Wheeler (2000) produced the first molecular
phylogeny for the group, which included 16 complex
thalloid genera. He resolved Monoclea as part of a
crown polytomy within the complex thalloids, thereby
supporting Johnson’s (1904) conclusion that the
absence of air chambers and ventral scales in
Monoclea was an example of morphological reduction
due to the plant’s semi-aquatic habitat.
Although there has been some discussion re-
garding the nested position of Monoclea within the
main complex thalloid clade (e.g., Forrest & Crandall-
Stotler 2004), what has not been considered is the
identity of its closest relative. Neither Wheeler (2000)
nor Boisselier-Dubayle et al. (2002) have any support
for a sister relationship between it and any other
genus. Although Forrest and Crandall-Stotler (2005)
and Davis (2004) obtained support for a sister
relationship between Monoclea and Dumortiera, this
was based on very limited sampling of complex
thalloid genera. Only now, with 25 of the 32 complex
thalloid genera included, can we feel reasonable
confidence in the placement of Monoclea as sister to
Dumortiera (Fig. 5) (55% MP BS; 100% PP in BI1,
BI5 and BI14). Despite the rather low support for this
topology in parsimony-based analysis, which may well
correspond to low rates of molecular evolution
providing few parsimony-informative characters
within the complex thalloid lineage, this topology is
resolved in all analyses. In fact, there are morpho-
logical characters that unite these genera. While
Dumortiera has a complex archegoniophore, with an
8–10-lobed receptacle on a long stalk, both taxa have
almost sessile rounded male receptacles. Thallus
anatomy is very similar in the two, with neither
having air chambers. In fact, Gradstein et al. (2001:
214) stated that Monoclea often grows in association
with Dumortiera, and that sterile material of the two
genera can be difficult to separate. More significantly,
both taxa have similar slightly lobed sporocytes and
undergo monoplastidic meiosis (Shimamura et al.
2003). Although monoplastidic meiosis also occurs in
other taxa of the complex thalloid lineage, including
Wiesnerella Schiffn., Lunularia, some species of
Marchantia and Blasia, slightly lobed sporocytes occur
only in Monoclea and Dumortiera (Shimamura et al.
2003). It is interesting that as long ago as 1951,
Proskauer concluded from comparative systematic
studies of the two genera that Monoclea ‘‘may indeed
belong to the dumortieroid line’’ (1951: 265). Our
molecular evidence supports his view.
In general, traditional classifications of the
complex thalloids placed far greater weight on
322 the bryologist 109(3): 2006
Page 22
reproductive characters than on vegetative thallus
morphology; therefore the presence/absence of an
archegoniophore outweighed the shared absence of air
chambers when it came to classifying these genera.
The fact that both genera occupy similar habitats,
even growing intermixed, means that convergence
towards a common morphology in what we have
historically regarded as a directly environmentally
responsive feature seems an obvious explanation.
Although it remains possible that air chambers were
lost independently in Monoclea and Dumortiera,
parsimony suggests a single loss in the common
ancestor of this lineage with subsequent reduction of
the archegoniophore and elaboration of the spor-
ophyte in Monoclea.
Unfortunately the wider relationship of these two
genera is not resolved here, although two of the
Bayesian analyses support a sister grouping with
Targionia L./Wiesnerella (100% BI1, 97% BI14 PP), as
also resolved by Long et al. (in prep.).
Within the ‘‘core’’ Marchantiales, neither Du-
mortiera nor Neohodgsonia resolve within the
Marchantiaceae, where they were placed by Bischler
(1998). Within the Marchantiaceae, Marchantia is
paraphyletic with Preissia nested within it. Both the
Aytoniaceae Cavers and Cleveaceae Cavers resolve as
monophyletic, supporting the traditional view of
them as morphologically well-supported families. In
the Aytoniaceae the results of Long et al. (2000) are
supported in that the largest genus, Asterella, is
paraphyletic. In the Cleveaceae, Peltolepis Lindb.,
which was excluded from the family by Bischler
(1998), falls clearly within it. This follows its tradi-
tional placement based on strong gametophytic and
spore ornamentation similarities.
The remaining crown group of complex thalloid
genera falls out as a major unresolved polytomy (Fig.
5) and further resolution of this clade is a priority for
the future. It includes both carpocephalate (e.g.,
Conocephalum Hill, Wiesnerella) and acarpocephalate
(e.g., Targionia, Riccia) genera, confirming the view
that reduction in gametophytic complexity (partic-
ularly of the reproductive structures) has been rife at
all levels in the complex thalloids. However, within
this polytomy a number of relationships are apparent,
such as that of Monoclea and Dumortiera discussed
above, as well as Corsinia and Exormotheca. These
relationships will be discussed in greater depth in a
forthcoming paper (Long et al., in prep.).
In regard to our treatment of Conocephalum, we
are aware of a recent publication by Szweykowski et al.
(2005) that effectively restricts Conocephalum conicum
(L.) Lindb. to Europe, placing samples from the
U.S.A. and India in a new species, C. salebrosum
Szweykowski, Buczkowska & Odrzykoski. We have,
however, retained the epithet conicum for our Illinois
sample because it does not resolve with C. salebrosum
in parsimony analyses that we have conducted in
conjunction with GenBank psbA accessions (from
Kim et al. 2001, as cited by Szweykowski et al. 2005).
Sterile thalli from Illinois populations match the
morphology assigned to C. salebrosum rather than that
of C. conicum s.str., but they contain a seven base pair
insertion in the trnK-psbA intergenic spacer region
that is missing in all C. salebrosum sequences, but
present in all other sampled C. conicum s.l. accessions.
We feel that the diversity within C. conicum s.l. needs
further investigation. Indeed, Szweykowski et al.
(2005) stated that there are six cryptic species within
this taxon; possibly the Illinois material will fall more
naturally into one of the species/races that has not yet
been addressed taxonomically. On the other hand, the
Indian material fits C. salebrosum both in terms of
morphology and of psbA sequence data.
What is clear from our analyses (Fig. 5) is that the
traditional subdivision of the complex thalloids into
the orders Monocleales, Marchantiales Limpr. and
Sphaerocarpales (Bischler 1998; Grolle 1983; Schuster
1984) is now convincingly demolished; more recent
classifications, such as that of Crandall-Stotler and
Stotler (2000), will require major overhaul. However,
a robust classification of complex thalloids is still
some way off without better resolution of both the
early-diverging lineages and the large crown group.
The monophyly of the Blasiales and its placement
as the earliest divergence from the complex thalloid
lineage are well supported in all analyses (Figs. 1, 5);
given the high degree of morphological similarity
between the two genera (Blasia and Cavicularia), this
is not an unexpected result. Previous results suggest-
ing a paraphyletic Blasiales (Forrest & Crandall-
Stotler 2004; Heinrichs et al. 2005) are due to
problems with DNA contamination, as discussed
further in the section ‘‘Benefits of collaboration.’’
Forrest et al.: Liverwort phylogeny 323
Page 23
Simple thalloids. The significance of the para-
phyly of the simple thalloid liverworts has been
discussed in several recent publications (e.g., Cran-
dall-Stotler 2005 et al.; Davis 2004; Forrest &
Crandall-Stotler 2004, 2005; Frey & Stech 2005;
Heinrichs et al. 2005; He-Nygren et al. 2004). In
general, there has been agreement that with the
reassignment of Haplomitrium and Treubiaceae to the
Haplomitriopsida and the Blasiales to the March-
antiopsida, the Metzgeriidae comprise two main
clades. These have been designated as Simple Thalloid
I (Fig. 6) and Simple Thalloid II (Fig. 7) (Davis 2004);
the former includes the majority of genera tradition-
ally classified in the Metzgeriidae (Crandall-Stotler &
Stotler 2000) and the latter, the genera of the suborder
Metzgeriineae, Verdoornia and the leafy liverwort,
Pleurozia. He-Nygren et al. (2006) relegated the taxa
of Simple Thalloid I to a new subclass, the Pelliidae
He-Nygren et al., and retain only the taxa of Simple
Thalloid II, excluding Pleurozia, in the Metzgeriidae.
Given the ambiguity that still exists in the resolution
of many lineages within the liverworts, however, we
feel that it is premature to translate any molecular
topology into a classification scheme. For example,
the classification of Pleurozia itself is problematic. He-
Nygren et al. (2006) placed Pleurozia with leafy
liverworts in the subclass Jungermanniidae, despite
the fact that in their Bayesian analysis it is resolved in
the Metzgeriaceae/Aneuraceae clade (PP 100%), in
agreement with our analyses. Only in their NONA
analysis (He-Nygren et al. 2006) is Pleurozia resolved,
with very low support, as sister to the rest of the leafy
liverworts. In our analyses and in those of He-Nygren
et al. (2006), there are no polytomies along the
backbone of the Simple Thalloid I clade, in contrast to
those illustrated in the diagrammatic scheme of
relationships presented in Frey and Stech (2005: fig.
1). A Pellia Raddi/Noteroclada Taylor ex Hook. &
Wilson clade is strongly supported as a member of the
earliest divergence within the Simple Thalloid I group,
with the Fossombroniineae sister to the Pallavicinii-
neae in the remaining lineage. Calycularia Mitt. is
weakly supported as sister to the Fossombroniineae/
Pallaviciniineae lineage, while Allisonia, which is
traditionally classified in the same family as Calycu-
laria (Crandall-Stotler & Stotler 2000), is resolved
within the Fossombroniineae clade. This topology has
also been resolved in previous studies focused on the
Metzgeriidae (Crandall-Stotler et al. 2005; Forrest &
Crandall-Stotler 2005), but differs from that resolved
in Heinrichs et al. (2005), particularly as regards the
earliest divergence in the lineage and the positions of
Makinoa Miyake and Noteroclada. The absence of
Pellia in the analyses of Heinrichs et al. (2005),
however, may be responsible for their resolution of
Makinoa in the earliest divergence of Simple Thalloid
I and the nesting of Noteroclada in the Fossombro-
niineae. It has been demonstrated that the absence of
a long-branch taxon can, in fact, drastically alter the
topology of early divergences (e.g., Crandall-Stotler et
al. 2005; Soltis & Soltis 2004); for example, when
Treubia is excluded, a complex thalloid/Blasia clade is
resolved as the earliest diverging lineage from the
main body of the liverworts and Haplomitrium is
nested in the Simple Thalloid I clade (Forrest &
Crandall-Stotler 2004) or its position on the liverwort
backbone phylogeny becomes highly unstable (Davis
2004). Since the relationships resolved within the
Simple Thalloid I clade have been discussed in detail
in earlier publications (Crandall-Stotler et al. 2005;
Forrest & Crandall-Stotler 2004, 2005), our discus-
sions here will be restricted to taxa whose placements
have been considered problematic (Frey & Stech
2005), namely, Makinoa, Phyllothallia E.A. Hodgs.
and Noteroclada.
In the analyses herein Makinoa is resolved with
strong BS and PP support as sister to a Fossombro-
niineae/Allisonia clade (Fig. 6), in congruence with
previous analyses by He-Nygren et al. (2004) and
Crandall-Stotler et al. (2005). In traditional classi-
fications (Crandall-Stotler & Stotler 2000; Schuster
1992), Makinoa was aligned with the Pallaviciniineae.
This placement was based primarily on shared
features of perichaetial organization and sporophyte
anatomy. As in Symphyogyna, the cluster of arche-
gonia is protected by an overarching, large posterior
perichaetial scale, a pseudoperianth is lacking and the
capsule is cylindric with a bistratose capsule wall. The
sporophyte is enclosed in a fleshy, scale-covered shoot
calyptra and the capsule opens along one or two slits.
Thallus anatomy, on the other hand, suggests a closer
affinity to Allisonia, as also suggested in our analyses
(see also Crandall-Stotler et al. 2005). Thalli in both
taxa lack strands of hydrolyzed cells, and bear
324 the bryologist 109(3): 2006
Page 24
uniseriate, ventral slime hairs that are distributed in
two rows near the thallus apex, and red-brown to
purple rhizoids. Like Makinoa, Allisonia fails to
elaborate a pseudoperianth after fertilization, enclos-
ing the sporophyte only in a calyptra that is subtended
at its base by perichaetial scales; in contrast, however,
sporophytic capsules in Allisonia are globose and
dehiscence is irregular. Anatomical studies of peri-
gonial organization and sporophyte development are
needed to clarify this relationship. As discussed in
previous publications (Forrest & Crandall-Stotler
2004, 2005), Verdoornia, which was originally aligned
with Makinoa by Schuster (1984), is well supported in
both molecular (e.g., Fig. 7) and morphological
analyses (Crandall-Stotler et al. 2005) as a member of
the Aneuraceae (Frey & Stech 2005).
The position of Phyllothallia as sister to the large
clade comprising most of the Pallaviciniineae has been
resolved in both previous (Crandall-Stotler et al. 2005;
Forrest & Crandall-Stotler 2004, 2005; Heinrichs et al.
2005; He-Nygren et al. 2006) and the current analyses
without ambiguity (Fig. 6) (BS¼ 85%, PP for all
Bayesian analyses¼100%). Placed in its own suborder
by Schuster (1967, 1984), the genus was later
tentatively assigned to the Treubiales by Crandall-
Stotler and Stotler (2000). In particular, the occur-
rence of mature spheroidal capsules that dehisce
irregularly in P. fuegiana R. M. Schust. (Hassel de
Menendez 1971) seems discordant with the elongate,
valvate capsules that are diagnostic of the Hymeno-
phytaceae R. M. Schust. and Pallaviciniaceae. Ana-
tomical studies have shown, however, that there are
ontogenetic transformations from spheroidal to
elongate capsule morphologies within this clade
(Forrest & Crandall-Stotler 2004). When these find-
ings are considered in concert with gametophytic
characters of Phyllothallia that are suggestive of
Symphyogyna, the placement of Phyllothallia in a
shared ancestry with the Pallaviciniineae does not
seem problematic (see Forrest & Crandall-Stotler 2004
for further discussion).
Since the morphological studies of Schiffner
(1911), various authors have aligned Noteroclada with
Pellia (e.g., Evans 1939; Schuster 1984; Crandall-
Stotler & Stotler 2000). Despite the differences in
habit, leafy in Noteroclada and thalloid in Pellia, the
taxa share several major anatomical features. These
include antheridia individually sunken into ostiolate
chambers, spheroidal capsules with basal elatero-
phores and similar patterns of endosporic spore
germination. In the 8-locus analyses of Forrest and
Crandall-Stotler (2005) and Crandall-Stotler et al.
(2005) Noteroclada is resolved as sister to Pellia, with
BS support of 70% and 85%, respectively. In contrast,
earlier 5-locus analyses that included an incomplete
rbcL sequence for Noteroclada (AF536228) were
ambiguous for both Noteroclada and Pellia (Forrest &
Crandall-Stotler 2004). The position of Noteroclada in
phylogenetic schemes has been further confounded in
the analyses of He-Nygren et al. (2004), which
resolved it in a clade with Metzgeria Raddi and
Pleurozia, Schaumann et al. (2005), where it is
resolved, without support, in a clade comprising the
Metzgeriineae, and Heinrichs et al. (2005) in which
two GenBank accessions resolve it in separate clades,
the former in the Fossombroniineae and the latter as
sister to Aneura Dumort.
To test these various hypotheses we have
included two geographically separated collections of
this species, one from Chile and one from Venezuela
(Table 1), in our current analyses. They resolve as
sister to the four Pellia collections under MP (63%
BS) and Bayesian (100% BI1, BI5, BI14 PP) criteria
(Fig. 6). In addition, unpublished data for rbcL and
rps4 generated by Forrest place a second Venezuelan
collection [Freire & Crandall-Stotler 4189 (ABSH)], and
a collection from the Ecuadorian Andes [Weiss &
Schwerdtfeger s. n. (ABSH)] as monophyletic with the
Noteroclada collections included herein, while He-
Nygren et al.’s (2004) Noteroclada sequences
(AY462318, AY462377) resolve well within an Aneura
pinguis (L.) Dumort./A. maxima (Schiffn.) Steph.
clade (Forrest & Crandall-Stotler, unpublished anal-
ysis). These results strongly suggest that our reso-
lution of Noteroclada sister to Pellia is the most
reliable topology for this taxon and that its resolution
in the Aneuraceae is based on laboratory errors, as will
be discussed further in the section ‘‘Benefits of
Collaboration.’’ Indeed, He-Nygren et al. subse-
quently (2006) excluded Noteroclada from their
analyses due to uncertainties about its sequence data.
The topology of the Simple Thalloid II clade (Fig.
7) resembles that resolved in several previous analyses
(e.g., Crandall-Stotler et al. 2005; Davis 2004; Forrest
Forrest et al.: Liverwort phylogeny 325
Page 25
& Crandall-Stotler 2005), with a monophyletic
Metzgeriaceae sister to a monophyletic Aneuraceae as
defined by Frey and Stech (2005) to include
Verdoornia. Increased sampling reveals Metzgeria and
Apometzgeria Kuwah. to be paraphyletic, while the
three species of Riccardia Gray are monophyletic. The
resolution of Lobatiriccardia (Mizut. & S. Hatt.)
Furuki sister to an Aneura/Cryptothallus clade sup-
ports its recognition as a genus distinct from Aneura,
as proposed by Furuki (1991). Sampling within this
clade should be greatly increased in future studies, not
only in the species-rich Metzgeria and Riccardia, but
also in Aneura, which, in fact, shows comparatively
large levels of molecular divergence within A. pinguis
alone (Fig. 2; Forrest, unpublished data).
Leafy liverworts. The separation of the leafy
liverworts into two main clades (Leafy I and II), as
first suggested by Davis (2004) and He-Nygren et al.
(2004), remains supported. Many sister group rela-
tionships within these clades have been newly resolved
or gained further support in our analyses. However,
the position of the Ptilidium group in relation to these
two clades remains unresolved (Figs. 2, 9).
The ‘‘Leafy I’’ clade, first named by Davis (2004),
and also resolved in the single-locus studies of
Heinrichs et al. (2005) and Knoop (pers. comm.), is
also resolved in the present analyses (Figs. 1, 2, 8).
This clade has received varying support in the past,
with a PP of 63% and 97% in the homogeneous
Bayesian analyses of Heinrichs et al. and Knoop,
respectively, but posteriors of 100% (homogeneous
and two heterogeneous partition models) in the more
comprehensive analyses of Davis. However, Leafy I
lacked MP and ML BS support in Davis’ and Knoop’s
studies. Again, in our study, Bayesian analyses
(homogeneous and heterogeneous) support this
clade, while MP and ML BS failed. Likely, these
discrepancies in support values are due to sensitivity
in the placement of Ptilidum within the leafy liverwort
lineage.
Davis (2004) first showed that Ptilidium is
associated with Neotrichocolea; however, the position
of this Ptilidium/Neotrichocolea clade within the leafy
liverwort clade was unresolved in all her analyses. In
the recent analyses of He-Nygren et al. (2006), as well
as in our analyses, the position of the Ptilidium/
Neotrichocolea clade is also unsupported. In parsi-
mony-based analyses it is part of a basal leafy
polytomy; in Bayesian analyses it falls, without (or
with barely) significant PPs (Fig. 9) (88% in BI1 and
BI14, although 95% in BI5), with ‘‘Leafy II,’’ in
agreement with Davis (2004) and Knoop (pers.
comm.). Additional topologies that have been re-
ported, but which were not supported by our analyses,
include resolution of Ptilidium within a Porellales/
Radulales (R. M. Schust.) Stotler & Crand.-Stotler
clade in some of the POY analyses of Ahonen (2004)
and He-Nygren et al. (2004), and as sister to the
‘‘Leafy I’’ clade (however, with non-significant PP) in
Heinrichs et al. (2005) and He-Nygren et al. (2006).
Despite the lack of support for the phylogenetic
placement of Ptilidium, both Heinrichs et al. (2005)
and He-Nygren et al. (2006) proposed recognizing
this ‘‘Leafy IþPtilidium’’ clade as an order, Porellales.
The following characters are cited by Heinrichs et al.
(2005) as justification for this placement: mostly trifid
leaves, absence of ventral branching and production
of pinquisanes as secondary metabolites. According to
He-Nygren et al. (2006: 20) ‘‘the only non-homo-
plasious morphological synapomorphy of the clade . . .
is the endogenous spore germination,’’ which is,
however, absent in Ptilidium. While many of the taxa
resolved in Leafy I have previously been associated on
morphological grounds as Porellales (for discussion
see Crandall-Stotler & Stotler 2000), Radula, the
Lepidolaenaceae Nakai and Ptilidium (which has been
placed with Mastigiophoraceae R. M. Schust. and
Chaetophyllopsidaceae R. M. Schust. in suborder
Ptilidiineae R. M. Schust. (see Ahonen 2004)) have
not previously been taxonomically allied with this
lineage.
The placement of the Ptilidium/Neotrichocolea
clade bears directly on the interpretation of ventral
lobule and water sac evolution in the leafy liverworts.
All taxa resolved in Leafy I in this study (Fig. 8) have
ventral lobules that are smaller than the dorsal lobe of
the leaf, while in Leafy II (Fig. 9), all lobulate taxa
have dorsal lobules that are smaller than the ventral
lobe (e.g., Scapania, Diplophyllum, Schistochila, Now-
ellia). The elaboration of lobules into water sacs can
occur in both lineages although it is more common in
Leafy I (e.g., Nowellia in Leafy II, but all genera of the
Lepidolaenineae and Jubulineae (Spruce) Mull. Frib.
in Leafy I). Neotrichocolea does produce ventral
326 the bryologist 109(3): 2006
Page 26
lobules that are sometimes elaborated into water sacs,
but only on leaves of secondary to quartenary
branches (Evans 1905). The water sacs of Neo-
trichocolea are galeate like those of the Lepidolaeni-
neae and Frullaniaceae of Leafy I, but are elaborated
from the ventralmost lobe of the 3-lobed leaf, rather
than from the middle lobe as is the case in these Leafy
I taxa (Evans 1905; Schuster 1972). If Neotrichocolea is
a lineage within Leafy II, the ventral water sac may be
parsimoniously interpreted as a pleisiomorphy for the
leafy liverworts. Alternatively, as suggested by their
different ontogenetic origins, ventral water sacs in
these two lineages may represent homoplasies in
response to common environmental pressures. It
seems that answering the question of where Ptilidium
and Neotrichocolea truly belong will require more
sampling, both of sequence characters and taxa;
clearly, questions about the direction of morpholog-
ical evolution within the leafy liverwort clades cannot
be fully addressed until the position of this lineage is
resolved.
The unexpected inclusion of the Lepidolaenaceae
and Jubulopsidaceae (Hamlin) R. M. Schust. with
taxa traditionally comprising the Porellales and
Radulales in a well-supported lineage, i.e., Leafy I
(Fig. 8), was first resolved by the analyses of He-
Nygren et al. (2004) and Davis (2004). Several
morphological features of the leafy gametophytes,
such as incubous leaf insertions, ventral lobules or
watersacs on at least some leaves and/or underleaves,
lateral (never ventral) branching, and precocious,
endosporic spore germination, support this relation-
ship (see also Davis 2004; He-Nygren et al. 2004;
Heinrichs et al. 2005). In addition, taxa in this clade
are predominantly epiphytes, lack mycorrhizae-like
mutualistic associations with fungi (Kottke & Nebel
2005; Nebel et al. 2004), and bear their perigonia on
either spicate or capitate lateral branches. There is,
however, significant disparity within the clade in the
organization of the sporophyte and its associated
gametophytic investments, with well-developed peri-
anths and spheroidal to ovoid capsules characterizing
the Porellineae R. M. Schust., but scale-covered
perigynia and long cylindric capsules in Lepidolae-
naceae and Jubulopsidaceae. In the Radulales a partial
stem perigynium-perianth complex encloses the
sporophyte (Schuster 1984: fig. 79) and the capsule is
ovoid to cylindric. The widely separated positions
accorded to these three groups in morphologically
based classifications (e.g., Crandall-Stotler & Stotler
2000; Evans 1939; Schuster 1984) are reflective of
these differences. Indeed, it seems that the assumption
in hepaticology that anatomical features of the
sporophyte and its associated investments are the best
predictors of phylogenetic relationship is not true of
many lineages, as also illustrated by the Phyllothallia/
Pallaviciniineae, Makinoa/Fossombroniineae and
Monoclea/Dumortiera clades. This may hold even
more widely across the bryophytes, as Buck (1991)
reported a similar phenomenom in the mosses.
While the monophyly of the Porellaceae Cavers is
strongly supported, its position relative to other
lineages within the clade is equivocal (Fig. 8). It
resolves sister to a clade comprising the Goebeliella-
ceae Verd., Lepidolaenaceae and Jubulopsidaceae, but
without significant support in most analyses (PP .
95% only under homogeneous Bayesian criterion).
The position of Goebeliella Steph., a genus with
sporophytes and perianths like Porella but distinctive
gametophytes that bear pairs of galeate water sacs on
their leaves, has varied among analyses. It is resolved
as sister to Radula in Heinrichs et al. (2005), nested
between the Porellaceae and a Radula/Frullania/
Jubula/Lejeuneaceae clade in He-Nygren et al. (2006)
and sister to the Lepidolaenaceae/Jubulopsidaceae
clade in Davis (2004), He-Nygren et al. (2004) and
this analysis. The Lejeuneaceae resolve as sister to the
Jubulaceae and comprise the crown group of Leafy I.
The Frullaniaceae are placed sister to the Jubulaceae/
Lejeuneaceae clade, with strong support (Fig. 8).
These three families share the occurrence of spores
with exine rosettes (Slageren 1995; Weis 2001),
sporophyte development entirely within the calyptra,
a reduced foot and fixed elaters (Spruce 1884–1885).
Within the Lejeuneaceae, more focused studies
indicate that some generic circumscriptions may
change when molecular data are taken into account
(e.g., Heinrichs et al. 2005; Wilson et al. 2004).
‘‘Leafy II,’’ first named by Davis (2004), is a well-
supported clade in all our analyses (Figs. 1, 2, 9) (but
see discussion of Ptilidium/Neotrichocolea, above).
The resolution of Schistochilaceae H. Buch as an early
diverging lineage, sister to the remaining taxa of Leafy
II, is also well established, in agreement with Davis
Forrest et al.: Liverwort phylogeny 327
Page 27
(2004), He-Nygren (2004, 2006) and Heinrichs et al.
(2005). The remaining taxa resolve into three major
clades, A, B, and C (Fig. 9), as designated in Davis
(2004). Clade A is well supported in our analyses and
confirms the alliance of the monophyletic families
Lepidoziaceae Limpr., Lepicoleaceae R.M. Schust. and
Herbertaceae Mull. Frib. (including Mastigophora).
These families are morphologically united by incu-
bously inserted leaves that are mostly 2- or 4-parted,
only rarely trifid, as well as by endogenous branches
(or flagellae). The sister group relationships among
these three families, however, are unsupported in our
own or previously published analyses. The inclusion
of Trichocolea within Clade A is well supported by our
analyses, although its position within it is not; based
on the 12-gene analyses of Davis (2004) it likely is part
of the Lepidoziaceae/Lepicoleaceae/Herbertaceae lin-
eage. The present analyses continue to support the
sister group relationship between these families and
some elements of the succubous-leaved Geocalycaceae
H. Klinggr. [Chiloscyphus and Lophocolea of the subf.
Lophocoleodeae Jorg.] plus Plagiochila clade, as
previously reported in multiple studies. It should be
noted, however, that the Geocalycaceae are para-
phyletic, with Harpanthus Spruce of the subf. Geo-
calycoideae H. Klinggr. resolved sister to Gyrothyra M.
Howe in clade C, albeit with weak support.
Our continued use of the name Lophocolea (e.g.,
Fig. 9, clade A) is in contrast to authors who consider
the genus synonymous with Chiloscyphus (Engel &
Schuster 1985). Although molecular phylogenetics has
been brought to bear on this issue (He-Nygren &
Piippo 2003), and reveals that species from Chiloscy-
phus s.str. nest within Lophocolea, low sampling levels
(with only seven out of over 300 recognized species)
and low clade support values mean that the issue
should not be considered resolved. Indeed, given the
range of morphological variation within Chiloscyphus
s.l. (Engel & Schuster 1984), this appears to be a
lineage in which increased molecular sampling,
coupled with morphological character mapping,
could be highly rewarding in isolating morphologi-
cally quantifiable taxa.
Clade B is solidly supported for the first time in
the present study, and is united by succubously or
transversely inserted leaves. Our analyses confirm the
alliance of Scapaniaceae Mig./Lophoziaceae Cavers as
sister to the Cephaloziellaceae Douin. Herzogobryum,
traditionally classified close to Gymnomitrion in the
Gymnomitriaceae Limpr. based on numerous shared
morphological characters (Schuster 2002), is nested
between this clade and that comprising the Cepha-
loziaceae Mig., thereby rendering the Gymnomitria-
ceae paraphyletic. The monophyly of the
Cephaloziaceae and its sister relationship to the
Scapaniaceae/Lophoziaceae/Cephaloziellaceae clade
are well supported, in agreement with numerous
previous studies. The inclusion of Adelanthus in Clade
B, as suggested by classifications (e.g., Crandall-Stotler
& Stotler 2000) as well as previous phylogenetic
results (Davis 2004; Heinrichs et al. 2005) has good
support in our analyses.
In Clade C the sister group relationship between
Nardia and the Gymnomitriaceae (excluding Herzo-
gobryum) is supported for the first time in the present
study. The four sampled species assigned to Junger-
mannia (s.l.) are polyphyletic within Clade C. Three
of the species form a supported monophyletic group
that is sister to the Nardia/Gymnomitriaceae clade in
all but the 14-partition Bayesian analysis, but
Jungermannia leiantha Grolle is resolved in a separate
clade (Fig. 9). In fact, our analyses suggest a
relationship between J. leiantha and Leiocolea hetero-
colpos (Thed.) H. Buch [[ Lophozia heterocolpa
(Thed.) M. Howe], another gemma-producing taxon
with beaked perianths. The status of J. leiantha has
long been controversial. Some authors (e.g., Muller
1957; Schuster 1969) have recognized Jungermannia
to comprise only J. leiantha and one or two closely
related species that also possess beaked perianths, with
the remaining species comprising the genus Solenos-
toma Mitt. This segregation of Solenostoma from
Jungermannia has not been generally accepted in
recent works (e.g., Grolle 1966; Grolle & Long 2000;
Stotler & Crandall-Stotler 2000; Vana 1996). How-
ever, the resolution of J. leiantha in a separate clade
from the other Jungermannia species supports the
hypothesis of two genera, although this should be
tested with increased sampling from within the genus/
genera. To complicate matters, if two genera are,
indeed, to be recognized, the name Liochlaena Nees
must be applied to the J. leiantha segregate since
Jungermannia has been typified by J. atrovirens
Dumort. (Vana 1973), a species recognized by
328 the bryologist 109(3): 2006
Page 28
Schuster (1969) as a Solenostoma; Solenostoma sensu
Schuster would, in turn, become Jungermannia. These
findings also support the conclusions of Yatsentyuk et
al. (2004) based on trnL-trnF sequences that Leiocolea
should be recognized as distinct from Lophozia
(Dumort.) Dumort., which is resolved nested in the
Scapaniaceae of Clade B in several other analyses
(Heinrichs et al. 2005; He-Nygren et al. 2004, 2006;
Yatsentyuk et al. 2004).
The sister group relationship of Balantiopsida-
ceae H. Buch and Acrobolbaceae E. A. Hodgs.,
suggested in the rbcL and rps4 single locus analyses of
Heinrichs et al. (2005) and Knoop (pers. comm.),
respectively, also has support in some of our analyses.
In contrast to the alignment of Gyrothyra with the
Balantiopsidaceae by Schuster (1972, 1984), however,
Gyrothyra is resolved sister to Harpanthus, a rela-
tionship proposed by Crandall-Stotler (1978) based
on similarities in leaf ontogeny, perigonial organiza-
tion, antheridial anatomy and perigynial develop-
ment. Although no one morphological character
defines all the taxa currently resolved in Clade C, most
have succubous leaves (Calypogeia being a noteable
exception) and some level of pendent perigynium or
marsupium development.
Of the major lineages, leafy liverworts (Figs. 8, 9)
and complex thalloids (Fig. 5) are the groups with the
lowest amount of resolution in these analyses,
containing several polytomies under maximum par-
simony and non-significant nodes in Bayesian anal-
yses. While the low resolution in the complex
thalloids appears to be the result of reduced molecular
evolution, that in the leafy liverworts is more likely
due to the age of the group. Leafies comprise
approximately 4,000 species (http://bryophytes.plant.
siu.edu) and appear to be one of the most recent
lineages in the phylum, with some rapidly evolving
taxa. Thus, taxon sampling sensitivity may be
especially problematic here, and loci that are appro-
priate for the other liverwort lineages may be less
appropriate. By contrast the sister group to the leafy
liverworts, Metzgeriidae (simple thalloid II, Fig. 7),
comprising an estimated 220 species, is fully resolved
into supported nodes. Although this study has
resolved relationships among the major lineages
within the liverworts, much more sampling, both of
taxa and loci in the leafy liverworts, will be required to
satisfactorily resolve a phylogeny that is equally robust
for all lineages.
Benefits of collaboration. During the course of
this collaborative project, we have discovered the
extent to which several problems have confounded
previous studies. Collaborative studies such as these
are extremely important at this stage of synthesis,
when researchers need to assess the state and stage of
data collection in order to proceed with relevant
questions to future studies. These problems became
clear to us when data were compared, specimens were
reexamined, and generally, we obtained a better feel
for liverwort phylogeny. They include misidentifica-
tion of taxa, differential amplification of contami-
nants from mixed specimens/DNA extractions, and
perpetuation of these errors in the GenBank database.
Following are examples and suggestions for remedies.
The cases of Cavicularia and Noteroclada (as
mentioned earlier) are examples of problems that can
occur post-extraction, and perpetuation of errors
through GenBank. Contamination of a DNA sample
meant that some erroneous sequences for Cavicularia
were published in Forrest and Crandall-Stotler (2004),
leading to the erroneous suggestion of a paraphyletic
Blasiales. The problem sequences were later deter-
mined to belong to Pellia. Although this was corrected
in Forrest and Crandall-Stotler (2005), Crandall-
Stotler et al. (2005) and in our analyses here, an rbcL
sequence remained in GenBank long enough for
another study to be published using it (Heinrichs et
al. 2005), thus perpetuating the error. In the case of
Noteroclada, He-Nygren et al. (2004) were not able to
recognize their sequences as problematic because they
included only one accession of this monospecific
genus. Only by comparisons with additional data was
it made apparent that the sequences had the incorrect
name associated with them. These sequences remain
in GenBank and the rbcL sequence has subsequently
been used in another phylogenetic study (Heinrichs et
al. 2005), underlining the importance of early
correction of database errors, and of critical review
and verification of GenBank sequences in future work,
as done in a forthcoming study by Knoop (pers.
comm.).
Mixed specimen collections are a common
feature in hepaticology, and can present serious
problems for phylogenetic reconstruction. For exam-
Forrest et al.: Liverwort phylogeny 329
Page 29
ple, detailed examination of a collection of Lophozia
sudetica (Nees) Grolle (Davis 432 (DUKE)) by Crandall-
Stotler revealed fragments of Nardia insecta Lindb.,
identified by the presence of small triangular under-
leaves. This contamination resulted in the anomalous
position of Lophozia as sister to Nardia scalaris Gray
in Davis (2004). Comparison with the position of
Lophozia in other analyses (e.g., Heinrichs et al. 2005;
He-Nygren et al. 2004, 2006), however, highlighted
the problem and led to reexamination of our material.
In cases of mixed collections, a number of
scenarios are possible. For example, the unintended
taxon can be pulled for extraction (differentially or
randomly); both taxa can be included in the DNA
extraction but unequal preservation of the material
can mean that the DNA from one is degraded and
fails to amplify in PCR, or cells in one taxon may lyse
more readily to release DNA into solution than in the
other; mutations in primer sites can cause one taxon
to amplify over the other; or both taxa can amplify
(giving rise to multiple peaks in sequencing electro-
pherograms). In many ways the worst problems, at
least theoretically, can be caused by differential
preservation of two taxa in a collection that contains
very few contaminants (as in the Lophozia/Nardia
example)—and this is not an entirely unlikely
scenario given the probability that different taxa have
different levels of drought tolerance (or that one
produces dispersal organs such as gemmae). It is thus
possible that a very small fragment of a contaminant
in a collection can amplify differentially for a whole
series of sequenced loci. Differential primer response,
on the other hand, is not likely to apply to every locus
in a multi-gene analysis, and so can be identified by
incongruencies in the placement of taxa for different
data partitions, while if both taxa amplify, problems
become apparent at the sequencing stage. A very
mixed collection, while presenting opportunity for
errors during the initial DNA extraction, is more
likely to be identified when the voucher is rechecked
after the anomalous phylogenetic placement is
observed. However, even when the correct taxon is
carefully pulled from a mixed collection, there is still
potential for error if the contaminant produces
gemmae, as these can be caught between the speci-
men’s leaves or lamellae. Indeed, PCR of separate
DNA extractions using gemmae and thalli of Xen-
othallus R. M. Schust. is far more successful for the
gemmae (Forrest, unpublished data).
The point we intend to make here is that
herbarium-based methodology, while suitable for
reconciling disparities in morphological observations
and verifying identities, cannot resolve problems that
occur post-DNA extraction. All researchers, careful as
they may be, are vulnerable to these pitfalls, and only
by using a comparative method do problems become
apparent. Thankfully, now that the overall picture of
liverwort phylogeny is emerging, we can better judge
when relationships are anomalous and deserve a
second look.
Herbarium-based methodology is, however, es-
sential in rectifying the other common factor that can
confound phylogenetic estimates: the correct identi-
fication and isolation of the material being studied.
For many liverwort taxa this is not a trivial exercise.
Herbarium material is often fragmentary, and collec-
tions, as noted above, are frequently mixed. Further,
annotations, even by acknowledged experts in the
organisms, are sometimes incorrect or incomplete. In
the Lophozia/Nardia example, the bulk of the
collection is Lophozia sudetica. However, similarities
in gross morphology of the sterile dried material—for
example, both taxa with curving bifid leaves—make it
extremely easy to mistake the mixture for a pure
collection. This stresses the importance for phyloge-
netic studies within the hepatics to be performed in
collaboration with taxonomic experts on the group, to
verify both the initial and the reciprocal-illumination
plant identities.
Future research. A robust phylogeny for the
hepatics, as well as providing a framework for
focusing morphological and developmental studies,
provides potential for the dating of major events in
liverwort evolution and their correlation with major
events in Earth history. There has been speculation,
for example, that the radiation within the leafy
liverworts (perhaps particularly the Lejeuneaecae, a
largely epiphytic/epiphyllous group) correlates with
the expansion of angiosperm forests, and the
associated expansion in niches (e.g., Ahonen et al.
2003). In fact, there is phylogenetic evidence to
support this hypothesis, suggesting that the major
lineages of Lejeuneaceae diversified in the early-mid
Cenozoic (41.34–61.68 Mya) in line with the devel-
330 the bryologist 109(3): 2006
Page 30
opment of complex angiosperm forest habitats
(Newton et al. 2006). The idea that ancient lineages
exploited new epiphytic habitats to diversify alongside
flowering plants, rather than prior to, or in the
shadow of, angiosperms, is an idea gaining support
among plant evolutionary biologists (e.g., Schneider
et al. 2004). To test these hypotheses, phylogenies
must have dated nodes. The study by Newton et al.
(2006) is exciting in that it represents the first
systematic attempt to date major bryophyte lineages.
However, as the focus of their paper is pleurocarpous
mosses, the implications of the dating of liverwort
nodes are not discussed. That said, the paper includes
34 liverwort accessions across all three classes, and
certainly does provide us with a first approximation
for clade ages.
However, more in-depth studies are required,
and are a logical next step from the analyses presented
herein. Liverworts are unique among bryophytes in
having a long fossil record, including well-studied
macrofossils (e.g., Krassilov & Schuster 1984; Oos-
tendorp 1987), as well as the putative liverwort spores
described by Wellman et al. (2003) from 475 million
years ago. This offers real possibilities in terms of
putting date ranges onto nodes within the liverwort
phylogeny instead of using an external date, as is the
case in Newton et al. (2006). However, there is an
important caveat: extreme care must be taken when
considering the evolution of the complex thalloids, a
group that shows very different rates of molecular
evolution from the rest of the liverworts (Fig. 2), and
thus has the potential to confound age estimates.
Earlier in the paper we pointed out that liverwort
taxa fall into a classic ‘‘hollow curve’’ distribution
(Willis 1922), but noted that this may be an artifact of
classification within the group, indicating more about
how taxonomists delimit taxa than about evolu-
tionary pattern (Guyer & Slowinski 1993). Low
sampling levels among the leafy liverworts still do not
allow us to address issues of clade size in any depth.
However, it is apparent that hepatic topology is highly
pectinate, with small lineages progressively sister to far
larger ones (e.g., Haplomitriopsida/all other liver-
worts; Blasiaceae/complex thalloids; Pleuroziales/
simple thalloids II—see Fig. 1). Systematically,
pectinate topologies are explained by differential
evolutionary rates such as those produced by key
innovations and/or extinction rates in different
lineages, leading to the concept of ‘‘success’’ (resist-
ance to extinction or propensity for speciation)
(Pearson 1999). Although not every case of lineage
‘‘success,’’ or lack of it, will have a biological
explanation, we remain hopeful that detailed inves-
tigations of the distribution of morphological char-
acters across our phylogeny will elucidate at least
some of the key factors responsible for the present-day
distribution of diversity within the liverwort lineages.
The ability to use our estimate of liverwort
phylogeny to test hypotheses regarding morphological
evolution, radiations and key innovations, as well as
potential coevolution, for example with fungal
associates, within the group, opens up exciting new
opportunities for hepaticology. We have identified
several key areas of the tree, however, that still need
further phylogenetic focus to stabilize nodes, and
these should become priority areas for further
research.
ACKNOWLEDGMENTS
We thank the following people: Angela Newton (BM) and Neil
Bell (BM) for Riella DNA; Daniela Schill (E) for many of the
complex thalloid collections and DNAs; Raymond E. Stotler
(SIU) for help with literature, vouchers and authorities and for
proofreading the final draft; Dylan Kosma (SIU) for maintenance
of the culture collection; Sedonia Sipes (SIU) for access to
sequencing facilities; Michael Moller (E) for running one of the
Bayesian analyses; Cymon Cox (BM) for assistance with analyses;
Rosemary Wilson and Jochen Heinrichs (GOET) for comments
on an earlier draft; Volker Knoop (BONN) for access to his pre-
publication rps4 manuscript; Juan Carlos Villarreal A. (SIU) and
Bernard Goffinet (CONN) for proofreading and comments on
earlier drafts; Jon Shaw (DUKE) for the invitation to present this
paper at the XVII International Botanical Congress in Vienna,
and for providing the impetus behind the manuscript; Jon Shaw
(DUKE) and John Engel (F) for reviewing this paper; and
definitely not least, the many other people listed as collectors in
Table 1, who made this research possible. Funding sources: ECD
would like to acknowledge grant no. EF-0531730 (to A. J. Shaw).
Sequences generated at Southern Illinois University were funded
by NSF grant no. DEB-997796, those at the Royal Botanic
Gardens, Edinburgh were funded by RBGE Molecular Phylo-
genetics Projects, and those at Duke University were funded by a
plant systematics training grant from the Mellon Foundation.
LITERATURE CITED
Ahonen, I. 2004. Molecular phylogeny of the liverwort order
Porellales (Marchantiophyta, Jungermanniopdsida). In B.
Goffinet, V. Hollowell & R. Magill (eds.), Molecular
Forrest et al.: Liverwort phylogeny 331
Page 31
systematics of bryophytes. Monographs in Systematic
Botany from the Missouri Botanical Garden 98: 168–188.
———, M. Muona & S. Piippo. 2003. Inferring the phylogeny
of the Lejeuneaceae (Jungermanniopsida): a first appraisal of
molecular data. The Bryologist 106: 297–308.
Beckert, S., S. Steinhauser, H. Muhle & V. Knoop. 1999. A
molecular phylogeny of bryophytes based on nucleotide
sequences of the mitochondrial nad5 gene. Plant Systematics
and Evolution 218: 179–192.
Bischler, H. 1998. Systematics and evolution in the genera of the
Marchantiales. Bryophytorum Bibliotheca 51: 1–201.
Boisselier-Dubayle, M.-C., J. Lambourdier & H. Bischler. 2002.
Molecular phylogenies support multiple morphological
reductions in the liverwort subclass Marchantiidae (Bryo-
phyta). Molecular Phylogenetics and Evolution 24: 66–77.
Buck, W. R.. 1991. The basis for familial classification of
pleurocarpous mosses. Advances in Bryology, 4: 169–185.
Clayton, W. D. 1974. The genus concept in practice. Kew
Bulletin 38: 149–153.
Cox, C. J., B. Goffinet, A. J. Shaw & S. B. Boles. 2004.
Phylogenetic relationships among the mosses based on
heterogeneous Bayesian analysis of multiple genes from
multiple genomic components. Systematic Botany 29: 234–
250.
Crandall-Stotler, B. 1978. Morphogenesis and anatomy of the
gametophyte phase of Gyrothyra underwoodiana Howe.
Nova Hedwigia 29: 257–279.
——— & R. E. Stotler. 2000. Morphology and classification of
the Marchantiophyta. Pages 21–70. In A. J. Shaw & B.
Goffinet (eds.), Bryophyte Biology. Cambridge University
Press, Cambridge.
———, L. L. Forrest & R. E. Stotler. 2005. Evolutionary trends
in the simple thalloid liverworts (Marchantiophyta, Junger-
manniopsida subclass Metzgeriidae). Taxon 54: 299–316.
Davis, C. 2004. A molecular phylogeny of leafy liverworts
(Jungermanniidae, Marchantiophyta). In B. Goffinet, V.
Hollowell & R. Magill (eds.), Molecular systematics of
bryophytes. Monographs in Systematic Botany from the
Missouri Botanical Garden 98: 61–86.
Duff, R. J., D. C. Cargill, J. C. Villarreal & K. S. Renzaglia. 2004.
Phylogenetic relationships of the hornworts based on rbcL
sequence data, novel relationships and new insights. In B.
Goffinet, V. Hollowell & R. Magill (eds.), Molecular
systematics of bryophytes. Monographs in Systematic
Botany from the Missouri Botanical Garden 98: 41–58.
Engel, J. J. & R. M. Schuster. 1984 [1985]. An overview of the
genera of Geocalycaceae subfamily Lophocoleoideae (Hep-
aticae). Nova Hedwigia 39: 385–463.
Evans, A. W. 1905. Hepaticae of Puerto Rico V. Ceratolejeunea.
Bulletin of the Torrey Botanical Club 32: 273–290, pls. 19,
20.
———. 1939. The classification of the Hepaticae. Botanical
Review 5: 49–96.
Forrest, L. L. & B. J. Crandall-Stotler. 2004. A phylogeny of the
simple thalloid liverworts (Jungermanniopsida, subclass
Metzgeriidae) as inferred from five chloroplast genes. In B.
Goffinet, V. Hollowell & R. Magill (eds.), Molecular
systematics of bryophytes. Monographs in Systematic
Botany from the Missouri Botanical Garden 98: 119–140.
——— & ———. 2005. Progress towards a robust phylogeny
for the liverworts, with particular focus on the simple
thalloids. Journal of the Hattori Botanical Laboratory 97:
127–159.
———, ——— & D. G. Long. 2005a. [Abstract] Multiple
phylogenetic analyses of data from multiple loci support
placement of Haplomitrium Nees and Treubia K. I. Goebel as
the earliest diverging lineage in liverwort phylogeny. Page 32.
In Abstracts, scientific meeting, Botany 2005. Botanical
Society of America, Austin, Texas.
———, S. W. Schuette, B. J. Crandall-Stotler & R. E. Stotler.
2005b. A molecular study of the simple thalloid liverwort
Jensenia (Marchantiophyta, Pallaviciniaceae). The Bryologist
108: 204–211.
Frey, W. & M. Stech. 2005. A morpho-molecular classification of
the liverworts (Hepaticophytina, Bryophyta). Nova Hedwi-
gia 81: 55–78.
Furuki, T. 1991. A taxonomic revision of the Aneuraceae
(Hepaticae) of Japan. Journal of the Hattori Botanical
Laboratory 70: 293–397.
Goebel, K. 1898. Organographie der Pflanzen, Vol. 1, Abteil.
2(1) Spezieller Teil. Fischer, Jena.
Gradstein, S. R., M. E. Reiner-Drehwald & H. Schneider. 2003. A
phylogenetic analysis of the genera of Lejeuneaceae (Hep-
aticae). Botanical Journal of the Linnean Society 143: 391–
410.
———, S. P. Churchill & N. Salazar Allen. 2001. Guide to the
bryophytes of tropical America. Memoirs of the New York
Botanical Garden 86: 1–577.
Grolle, R. 1966. Die Typisierung von Jungermannia L. und J.
lanceolata L. Taxon 15: 186–191.
———. 1983. Nomina generica Hepaticarum; references, types
and synonymies. Acta Botanica Fennica 121: 1–62.
——— & D. G. Long. 2000. An annotated check-list of the
Hepaticae and Anthocerotae of Europe and Macaronesia.
Journal of Bryology 22: 103–140.
Groth-Malonek, M. & V. Knoop. 2005. Bryophytes and other
basal land plants: the mitochondrial perspective. Taxon 54:
293–297.
———, D. Pruchner, F. Grewe & V. Knoop. 2005. Ancestors of
trans-splicing mitochondrial introns support serial sister
group relationships of hornworts and mosses with vascular
plants. Molecular Biology and Evolution 22: 117–125.
Guyer, C. & J. B. Slowinski. 1993. Adaptive radiation and the
topology of large phylogenies. Evolution 47: 253–263.
Hassel de Menendez, G. G. 1971. Datos adicionales de la
extraordinaria Phyllothallia fuegiana Schust. (Hepaticae).
Boletın de la Sociedad Argentina de Botanica 13: 290–299.
Heinrichs, J., S. R. Gradstein, R. Wilson & H. Schneider. 2005.
332 the bryologist 109(3): 2006
Page 32
Towards a natural classification of liverworts (March-
antiophyta) based on the chloroplast gene rbcL. Cryptoga-
mie, Bryologie 26: 131–150.
He-Nygren, X., I. Ahonen, A. Juslen, D. Glenny & S. Piippo.
2004. Phylogeny of liverworts—beyond a leaf and a thallus.
In B. Goffinet, V. Hollowell & R. Magill (eds.), Molecular
systematics of bryophytes. Monographs in Systematic
Botany from the Missouri Botanical Garden 98: 87–118.
———, A. Juslen, I. Ahonen, D. Glenny & S. Piippo. 2006.
Illuminating the evolutionary history of liverworts (March-
antiophyta)—towards a natural classification. Cladistics 22:
1–31.
——— & S. Piippo. 2003. Phylogenetic relationships of the
generic comples Chiloscyphus-Lophocolea-Heteroscyphus
(Geocalycaceae, Hepaticae): insights from three chloroplast
genes and morphology. Annales Botanici Fennici 40: 317–
329.
Huelsenbeck, J. P. & F. Ronquist. 2002. MrBayes version 3.0B.
Available from the authors: http://morphbank.ebc.uu.se/
mrbayes3/info.php.
——— & ———. 2005. MrBayes version 3.1. Available from
the authors: http://morphbank.ebc.uu.se/mrbayes3/info.
php.
Johnson, D. S. 1904. The development and relationship of
Monoclea. Botanical Gazette 38: 185–205.
Kim, H. N., E. Nitasaka, I. J. Odrzykoski & T. Yamazaki. 2001.
Phylogenetic relationships among taxa of the liverwort
Conocephalum conicum (Conocephalaceae) revealed by psbA
sequence. Genes and Genetic Systems 76: 279–288.
Kottke, I. & M. Nebel. 2005. The evolution of mycorrhiza-like
associations in liverworts: an update. New Phytologist 167:
330–334.
Krassilov, V. A. & R. M. Schuster. 1984. Paleozoic and Mesozoic
fossils. Pages 1172–1194. In R. M. Schuster (ed.), New
Manual of Bryology. Hattori Botanical Laboratory, Ni-
chinan, Japan.
Leitgeb, H. 1877. Untersuchungen uber die Lebermoose. Heft
III. Die Frondosen Jungermannieen. O. Deistung’s Buch-
handlung, Jena.
Lewis, L. A., B. D. Mishler & R. Vilgalys. 1997. Phylogenetic
relationships of the liverworts (Hepaticae), a basal embry-
ophyte lineage, inferred from nucleotide sequence data of
the chloroplast rbcL gene. Molecular Phylogenetics and
Evolution 7: 377–393.
Long, D. G., L. L. Forrest, M. L. Hollingsworth, A. Clark & D. B.
Schill. 2005. [Abstract] The complex thalloids—evolving
simply. Page 450. In Abstracts, XVII International Botanical
Congress, Vienna, Austria.
———, M. Moller & J. Preston. 2000. Phylogenetic relation-
ships of Asterella (Aytoniaceae, Marchantiopsida) inferred
from chloroplast DNA sequences. The Bryologist 103: 625–
644.
Muller, K. 1957 [1956]. Gattung Jungermannia, Gattung
Solenostoma. Pages 809–811, 812–839. In K. Muller. [Die
Lebermoose Europas, Vol. II] Dr. L. Rabenhorst’s Krypto-
gamen-Flora von Deutschland, Osterreich und der Schweiz,
3 Auflage, VI. Band, 2 Abteilung. Geest & Portig, Leipzig.
Nebel, M., H.-P. Kreier, M. Preussing, M. Weiss & I. Kottke.
2004. Symbiotic fungi associations of liverworts are the
possible ancestors of mycorrhizae. Pages 339–360. In R.
Agerer, M. Piepenbring & P. Blanz (eds.), Frontiers in
Basidiomycete Mycology. IWH-Verlag, Etching.
Newton, A. E., N. Wikstrom, N. Bell, L. L. Forrest & M. Ignatov.
2006. Dating the diversification of the pleurocarpous
mosses. In A. E. Newton, R. Tangney & E. De Luna (eds.),
Pleurocarpous mosses: Systematics and Evolution. CRC
Press. (In press.)
Nickrent, D.L., A. Blarer, Y. L. Qiu, R. Vidal-Russell & F. E.
Anderson. 2005. Phylogenetic inference in Rafflesiales: the
influence of rate heterogeneity and horizontal gene transfer.
BMC Evolutionary Biology 2004(4): 40.
Nylander, J. A. A. 2002. MrModeltest – version 1.1b. Available
from the author. http://www.ebc.uu.se/systzoo/staff/
nylander.html.
Oostendorp, C. 1987. The bryophytes of the Palaeozoic and the
Mesozoic. Bryophytorum Bibliotheca 34: 1–112.
Pass, J. M. & K. S. Renzaglia. 1995. Comparative microanatomy
of the locomotory apparatus of Conocephalum conicum
(Hepaticae, Conocephalaceae). Fragmenta Floristica et
Geobotanica 40: 365–377.
Pearson, P. N. 1999. Apomorphy distribution is an important
aspect of cladogram symmetry. Systematic Biology 48: 399–
406.
Proskauer, J. 1951. Notes on Hepaticae II. The Bryologist 54:
243–266.
Pruchner, D., S. Beckert, H. Muhle & V. Knoop. 2002. Divergent
intron conservation in the mitochondrial nad2 gene,
signatures for the three bryophyte classes (mosses, liver-
worts, and hornworts) and the lycophytes. Journal of
Molecular Evolution 55: 265–272.
Qiu, Y. L., Y. Cho, J. C. Cox & J. D. Palmer. 1998. The gain of
three mitochondrial introns identifies liverworts as the
earliest land plants. Nature 394: 671–674.
Renzaglia, K. S., R. C. Brown, B. E. Lemmon, J. G. Duckett & R.
Ligrone. 1994. The occurrence and phylogenetic significance
of monoplastidic meiosis in liverworts. Canadian Journal of
Botany 72: 64–72.
Rodriguez, F., J. L. Oliver, A. Marin & J. R. Medina. 1990. The
general stochastic model of nucleotide substitution. Journal
of Theoretical Biology 142: 485–501.
Schaumann, F. W. Frey, T. Pfeiffer & M. Stech. 2005. Molecular
circumscription, intrafamilial relationships and biogeogra-
phy of the Gondwanan liverwort family Pallaviciniaceae
(Hepaticophytina, Bryophyta). Studies in austral temperate
rain forest bryophytes 27. Plant Systematics and Evolution
252: 27–48.
Schiffner, V. 1911. Zur Morphologie von Noteroclada. Oester-
reische Botanische Zeitschrift 61: 325–332.
Forrest et al.: Liverwort phylogeny 333
Page 33
Schill, D.B., D. G. Long, M. Moeller & J. Squirrell. 2004.
Phylogenetic relationships between Lophozialceae and
Scapaniaceae based on chloroplast sequences. In B. Goffinet,
V. Hollowell & R. Magill (eds.), Molecular systematics of
bryophytes. Monographs in Systematic Botany from the
Missouri Botanical Garden 98: 141–149.
Schneider, H., E. Schuettpelz, K. M. Pryer, R. Cranfill, S.
Magallon & R. Lupia. 2004. Ferns diversified in the shadow
of angiosperms. Nature 428: 553–557.
Schuster, R. M. 1967. Studies on antipodal Hepaticae, IX.
Phyllothalliaceae. Transactions of the British Bryological
Society 5: 283–288.
———. 1969. The Hepaticae and Anthocerotae of North
America east of the hundredth meridian, Vol. II. Columbia
University Press, New York and London.
———. 1972. Phylogenetic and taxonomic studies on Junger-
manniidae. Journal of the Hattori Botanical Laboratory 36:
321–405.
———. 1984. Evolution, phylogeny and classification of the
Hepaticae. Pages 892–1070. In R. M. Schuster (ed.), New
Manual of Bryology. Hattori Botanical Laboratory, Ni-
chinan, Japan.
———. 1992. The Hepaticae and Anthocerotae of North
America east of the hundredth meridian, Vol. V. Columbia
University Press, New York and London.
———. 2002. Austral Hepaticae Part II. Nova Hedwigia Beiheft
119: i–viii, 1–606.
Shimamura, M., Y. Mineyuki & H. Deguchi. 2003. A review of
the occurrence of monoplastidic meiosis in liverworts.
Journal of the Hattori Botanical Laboratory 94: 179–186.
Slageren, M. W. van. 1985. A taxonomic monograph of the
genera Brachiolejeunea and Frullanoides (Hepaticae), with a
SEM analysis of the sporophyte in the Ptychanthoideae.
Mededelingen van het Botanisch Museum en Herbarium
van de Rijksuniversiteit te Utrecht 544: 243–302.
Soltis, D. E. & P. S. Soltis. 2004. Amborella not a ‘‘basal
angiosperm?’’ Not so fast. American Journal of Botany 91:
997–1001.
Spruce, R. M. 1884–1885. Hepaticae of the Amazon and the
Andes of Peru and Ecuador. Transactions and Proceedings
of the Botanical Society of Edinburgh 15: 1–589, pls. i–xiii.
Steinhauser, S., S. Beckert, I. Capesius, O. Malek & V. Knoop.
1999. Plant mitochondrial editing. Journal of Molecular
Evolution 48: 303–312.
Stotler, R. & B. Crandall-Stotler. 1977. A checklist of the
liverworts and hornworts of North America. The Bryologist
80: 405–428.
Swofford, D. L. 2002. PAUP*. Phylogenetic analysis using
parsimony (* and other methods). Version 4. Sinauer
Associates, Sunderland, Massachusetts.
Szweykowski, J., K. Buczkowska & I. J. Odrzykoski. 2005.
Conocephalum salebrosum (Marchantiopsida, Conocephala-
ceae)—a new Holarctic liverwort species. Plant Systematics
and Evolution 253: 133–158.
Vana, J. 1973. Studien uber die Jungermannioideae (Hepaticae)
2. Jungermannia Subg. Jungermannia. Folia Geobotanica et
Phytotaxonomica 8: 255–309.
———. 1996. Notes on the Jungermanniineae of the world.
Anales del Instituto de Biologia de la Universidad Nacional
Autonoma de Mexico, Serie Botanica 67: 99–107.
Walters, S. M. 1986. The name of the rose: a review of ideas on
the European bias in angiosperm classification. New
Phytologist 104: 527–546.
Weis, G. 2001. Morphologische und anatomische Untersu-
chungen der Sporophyten bei den Jubulaceae Klinggr. und
Lejeuneaceae Casares-Gil (Hepaticae) und deren system-
atische Bedeutung. Bryophytorum Bibliotheca 57: 5–302.
Wellman, C. H., P. L. Osterloff & U. Mohiuddin. 2003.
Fragments of the earliest land plants. Nature 425: 282–285.
Wheeler, J. A. 2000. Molecular phylogenetic reconstructions of
the marchantioid liverwort radiation. The Bryologist 103:
314–333.
Willis, J. C. 1922. Age and Area. Cambridge University Press,
Cambridge.
Wilson, R., S. R. Gradstein, J. Heinrichs, H. Groth, A. L. Ilkiu-
Borges & F. A. Hartmann. 2004. Phylogeny of Lejeuneaceae:
a cladistic analysis of chloroplast gene rbcL sequences and
morphology with preliminary comments on the mitochon-
drial nad4–2 spacer region. In B. Goffinet, V. Hollowell & R.
Magill (eds.), Molecular systematics of bryophytes. Mono-
graphs in Systematic Botany from the Missouri Botanical
Garden 98: 189–202.
Wolf, P. G., K. G. Karol, D. F. Mandoli, J. Kuehl, K.
Arumuganathan, M. W. Ellis, B. D. Mishler, D. G. Kelch, R.
G. Olmstead & J. L. Boore. 2005. The first complete
chloroplast genome sequence of a lycophyte, Huperzia
lucidula (Lycopodiaceae). Gene 350: 117–128.
Yang, Z. 1994. Estimating the pattern of nucleotide substitution.
Journal of Molecular Evolution 39: 105–111.
Yatsentyuk, A. P., N. A. Konstantinova, M. S. Ignatov, J.
Hynonen & A. V. Troitsky. 2004. On phylogeny of
Lophoziaceae and related families (Hepaticae, Jungerman-
niales) based on trnL-trnF intron-spacer sequences of
chloroplast DNA. In B. Goffinet, V. Hollowell & R. Magill
(eds.), Molecular systematics of bryophytes. Monographs in
Systematic Botany from the Missouri Botanical Garden 98:
150–165.
ms. received October 20, 2005; accepted May 15, 2006.
334 the bryologist 109(3): 2006