Phylogenetic Relationships Within the Haplolepideous Mosses

Author(s): Catherine La Farge, Brent D. Mishler, John A. Wheeler, Dennis P. Wall, Kirsten Johannes, Steffan
Schaffer, and A. Jonathan Shaw
Source: The Bryologist, 103(2):257-276.
Published By: The American Bryological and Lichenological Society, Inc.
DOI: http://dx.doi.org/10.1639/0007-2745(2000)103[0257:PRWTHM]2.0.CO;2
URL: http://www.bioone.org/doi/full/10.1639/0007-2745%282000%29103%5B0257%3APRWTHM%5D2.0.CO
%3B2

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by
nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s
Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or
rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries,
and research funders in the common goal of maximizing access to critical research.
 The Bryologist 103(2), pp. 257–276
Copyright 䉷 2000 by the American Bryological and Lichenological Society, Inc.

Phylogenetic Relationships Within the Haplolepideous Mosses

CATHERINE LA FARGE
Department of Botany, Duke University, Durham, NC 27708, U.S.A. Current address: Department of Biological
Sciences, University of Alberta, Edmonton, AB, T6G 2E9, Canada

BRENT D. MISHLER, JOHN A. WHEELER, DENNIS P. WALL, AND KIRSTEN JOHANNES

University Herbarium and Department of Integrative Biology, University of California, Berkeley, CA 94720-2465,
U.S.A.

STEFFAN SCHAFFER
Institute for Biotechnology (IBT 1), Research Centre Juelich, PF 1913, D-52425 Juelich, Germany

A. JONATHAN SHAW
Department of Botany, Duke University, Durham, NC 27708, U.S.A.

Abstract. Mosses with haplolepideous peristomes form a major lineage within the arthrodon-
tous taxa, the Dicranidae. Relationships among lineages within the Dicranidae are explored using
three cpDNA regions: rbcL, rps4, and the region spanning trnL(UAA)- trnF(GAA). Maximum
parsimony analyses of combined data sets support robust clades that correspond to traditionally
recognized families. Phylogenetic relationships of 71 exemplar taxa rooted with five outgroup taxa
identify 13 major clades within the Dicranidae. Predominantly the monophyly of these groups is
strongly supported, although relationships among the clades are ambiguous. The phylogenetic
implications of the current taxon sampling include 1) the Dicranidae are monophyletic, 2) the
Pottiales are polyphyletic with the Calymperaceae and Octoblepharaceae cladistically distant from
the Pottiaceae, and these two former families should be transferred to the Dicranales, 3) the
Dicranales are polyphyletic, with Ditrichaceae and Rhabdoweisiaceae more closely related to the
Pottiaceae than other members of the Dicranales, 4) the Fissidentaceae are nested within the
Dicranales, 5) the Rhachitheciaceae and Erpodiaceae are nested within the Dicranidae and sister
to the Pottiaceae, 6) Amphidium is sister to the pottiaceous clade, 7) the Seligeriaceae form a
well supported sister relationship to the Grimmiales, and 8) Scouleria aquatica, Bryoxiphium
norvegicum, Ditrichum flexicaule, and Timmiella crassinervis appear to form basal lineages with-
in the Dicranidae.

Haplolepideous mosses represent one of four haplolepideous peristome is homologous with the
subclasses of Bryopsida and have been formally endostome of the diplolepideous peristome, show-
classified as the Dicranidae (Vitt et al. 1998). Peri- ing that Philibert had compared non-homologous
stome structure has provided the basis for bryo- structures. Edwards developed a formula that des-
phyte classification, since Philibert’s pivotal work ignates the number of cells in 1⁄8 of the capsule cir-
during the 1880’s (Taylor 1962). Philibert distin- cumference that contribute to peristome formation.
guished two types of arthrodontous peristomes, The three innermost amphithecial layers that typi-
haplolepideous and diplolepideous. He distin- cally form the peristome in arthrodontous mosses
guished the haplolepideous peristome by its single are (outside ⬎ inward) the Outer Peristomial Layer
column of cell wall plates on the outer surface of (OPL), Primary Peristomial Layer (PPL), and Inner
the tooth, whereas the outer surface of the outer Peristomial Layer (IPL). Haplolepideous peri-
tooth of the diplolepideous peristome has two col- stomes are derived from the PPL and IPL, with a
umns of cell wall plates. Philibert’s reference to sin- peristomial formula of 2:3. Contribution from the
gle and double plates has often been misinterpreted OPL to peristome formation is uncommon and ru-
as single and double peristomes. Blomquist and dimentary in haplolepideous mosses. Despite the
Robertson (1941) described peristome development generalized haplolepideous cell formula, the 2:3
in Aulacomnium heterostichum and enumerated the pattern of the PPL and IPL layers is variable within
amphithecial layers that give rise to the diplolepi- some haplolepideous taxa. Reduction or amplifica-
deous peristome. tion can result in variation of the IPL layer (1–8
In a comprehensive review of the mature peri- cells) in the mature peristome (e.g., Venturiella,
stome, Edwards (1979) emphasized that the single, Goffinet 1998; some species of the Dicranaceae and

0007-2745/00/257–276$2.15/0
258 THE BRYOLOGIST [VOL. 103

Ditrichaceae, Shaw et al. 1989). Rudimentary pro- of the collaborators’ research interests. Representative ex-
stomial thickenings (OPL) modify the formula to a emplars from additional families within the Dicranales,
Grimmiales, Seligeriales, and Pottiales are also included.
4:2:3 cell pattern in scattered taxa of the Dicranales, Specimen voucher information and Genbank accession
Seligeriales, and Grimmiales (e.g., Churchill 1981; numbers are provided in Table 2.
Edwards 1979; La Farge-England 1998). The hap- Genome sampling and alignment.—Three regions of
lolepideous peristome remains the primary morpho- the cpDNA genome were sequenced for molecular char-
acters: two coding regions, rps4 and rbcL, and the intron
logical character defining the Dicranidae. and exon 2 of trnL(UAA), and spacer between the trnL(UAA)
The most recent classification of the Dicranidae and trnF(GAA) (hereafter referred to as the trnL region).
recognized four orders: Pottiales, Dicranales (in- Most sequences were taken from previous studies or Gen-
cluding the Fissidentales), Grimmiales, and Seli- bank, however 77 sequences were newly generated for
geriales (Vitt 1984; Vitt et al. 1998). The distinction this paper (see Table 2). Fresh collections or herbarium
material (up to 30 years old) were used for DNA extrac-
between these orders is based on details of peri- tion, using a modification of Doyle and Doyle (1990). For
stome structure (e.g., division of teeth and orna- details on amplification, purification, and sequencing re-
mentation). The circumscription and recognition of action methods see Buck et al. (2000). The primers and
suprageneric taxa within the Dicranidae has varied primer sequences used for amplification of the trnL, rps4,
and rbcL chloroplast regions are shown in Table 3. The
during this century. Brotherus (1924) recognized program Se-AL v 1.0 (Rambaut 1995) was used to help
four orders with 13 families, Dixon (1932) recog- align the trnL region. A total of 258 trnL characters were
nized five orders with 10 families, Walther (1983) deleted from the analyses because of missing data and/or
included five orders and 19 families, and Vitt alignment ambiguity. The rps4 characters were restricted
(1984; Vitt et al. 1998) recognized four orders with to the coding region, excluding the intergenic spacer
(rps4-trnS) at the 3⬘ end (38 bases), excluding a total of
22 families (Table 1). The major differences among 69 characters. The initial 30 base pairs of the rbcL were
these treatments include the circumscription of the also excluded due to missing data for some taxa. The final
Pottiales, the placement and recognition of the Se- data set included 2,335 aligned characters.
ligeriaceae, and the distinction of new families. Phylogenetic analyses.—Tree searches were performed
using PAUP* 4.0.0b2 (Swofford 1998). Constraint trees
Recent advances in bryophyte systematics have were generated using MacClade 3.03 (Maddison & Mad-
resulted from molecular analyses of relationships dison 1992). Decay analyses were facilitated using
among the major clades of Bryopsida. Based on AutoDecay 4.0 (Eriksson 1998) and PAUP*. Fitch parsi-
evidence from nuclear (nr), chloroplast (cp), or mi- mony analyses were performed on three character com-
tochondrial (mt) genes, the Dicranidae have an am- binations for 76 taxa: trnL-rps4, rps4-rbcL, and trnL-
rps4-rbcL. Heuristic parsimony analyses of unweighted,
biguous position within the Bryopsida. Conflicting unordered characters were conducted with the following
results from several analyses suggest that the Di- settings: 100 random addition replicates, TBR branch
cranidae are either 1) sister to the Encalyptales and swapping, ‘collapse’ option in effect when branch lengths
together form a sister clade to Timmiales-Funari- were equal to zero, and ‘steepest descent’ off. Bootstrap
searches (Felsenstein 1985) consisted of 100 replicates
ales lineage (nad2 & nad5 mtDNA, Beckert et al. with 10 random additions of the data per bootstrap repli-
1999); 2) part of an unresolved clade including the cate. A maximum of 200 trees was saved for each random
Encalyptales and Diplolepideae, but excluding the replicate. Decay analyses were also run as a measure of
Funariales, based on rbcL (Goffinet et al. 1998), or support for the clades (Bremer 1994). Different cpDNA
3) sister to a clade consisting of the Encalyptales character combinations were analyzed to examine their
utility to resolve phylogenetic relationships at various hi-
and Diplolepideae, excluding Scouleria and Tim- erarchical levels. A constraint analysis was performed to
mia, based on 18S nrDNA, trnL, and rps4 cpDNA test traditional ordinal and familial concepts within the
(Cox & Hedderson 1999). For relationships within Dicranidae.
the Dicranidae, Stech (1999a) has explored the use
of the trnL region to circumscribe the Dicranaceae. RESULTS
The major objectives of our study are to 1) identify
the major clades within the Dicranidae using three Molecular characters.—Three character sets
cpDNA regions, and 2) examine relationships (trnL-rps4, rps4-rbcL, and trnL-rps4-rbcL) are an-
among well-supported lineages within these clades. alyzed that differ in character consistency indices
These data provide a framework to evaluate phy- and number of informative characters, maximum
logenetic relationships and evaluate the evolution- parsimony (MP) trees, and unresolved clades (Table
ary implications of a number of ecological and bi- 4). Twenty-eight percent of the trnL characters is
ological features within the Dicranidae. parsimony informative, 40% for rps4, and 25% for
rbcL. A summary of the strict consensus trees for
MATERIALS AND METHODS each analysis is presented for comparison (Fig. 1).
All three analyses have minor differences among
Taxon sampling.—Our analyses include a selection of
71 taxa from the Dicranidae, plus five out-group taxa. The
the character consistency indices. The rps4-rbcL
taxa represent a concentration from the Calymperaceae, and trnL-rps4-rbcL analyses show the best resolu-
the Dicranaceae, and Rhabdoweisiaceae, that form a focus tion (three versus four MP trees, respectively).
 TABLE 1. Classifications of ordinal and familial taxa within the Dicranidae. Families with representative exemplars in the current analyses are indicated in bold. The ordinal 2000]
classification for families within the Dicranidae suggested by our data is presented as Current Hypothesis. The families are arranged in alphabetical order, not an inferred
phylogenetic order. Quotation marks (‘‘’’) indicate the provisional status of these orders. Families included in the Dicranidae with uncertain placement are indicated by incertae
sedis (see text for references). Amphidium and Dicranella are not given familial status, and therefore are not classified in the table. Our inferred ordinal classification of families
within the Dicranidae is based on data from this study and data from the following, Cox et al. (2000); Eddy (1990); Goffinet and Cox (2000); Hedderson et al. (1999); Reese
and Zander (1987); Vitt (1984); and Vitt et al. (1998).

Brotherus 1924 Dixon 1932 Walther 1983 Vitt 1984; Vitt et al. 1998 Current hypothesis
Musci Musci Bryophytina Bryopsida Bryopsida
Bryales Bryales Bryopsida Dicranidae Dicranidae
Eubryinales Arthrodonteae Bryidae
Haplolepideae ‘‘POTTIALES’’
Erpodiaceae
FISSIDENTALES FISSIDENTALES DICRANALES POTTIALES Ditrichaceae
Fissidentaceae Archefissidentaceae Ditrichaceae Pottiaceae Pottiaceae
Fissidentaceae Bruchiaceae Calymperaceae Rhabdoweisiaceae
DICRANALES Archidiaceae Bryobartramiaceae Rhachitheciaceae
Archidiaceae GRIMMIALES Pleurophascaceae
Ditrichaceae Grimmiaceae Dicranaceae DICRANALES ‘‘DICRANALES’’
Bryoxiphiaceae Leucobryaceae Dicranaceae Calymperaceae
Seligeriaceae DICRANALES Dicnemonaceae Leucobryaceae Dicranaceae
Dicranaceae Archidiaceae Aongstroemiaceae Rhabdoweisiaceae Eustichiaceae
Dicnemonaceae Dicranaceae Wilsoniellaceae Nanobryaceae Fissidentaceae
Leucobryaceae Dicnemonaceae Ditrichaceae Leucobryaceae
Pleurophascaceae Pleurophascaceae FISSIDENTALES Viridivelleraceae Nanobryaceae
Leucobryaceae Fissidentaceae Bryoxiphiaceae Octoblepharaceae
POTTIALES Bryoxiphiaceae Eustichiaceae Phyllodrepaniaceae
Calymperaceae SYRRHOPODONTALES Eustichiaceae Sorapillaceae Pleurophascaceae
Encalyptaceae Syrrhopodontaceae Phyllodrepaniaceae Sorapillaceae
Pottiaceae POTTIALES Dicnemonaceae Viridivelleraceae
POTTIALES Pottiaceae Pleurophascaceae
LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES

GRIMMIALES Pottiaceae Fissidentaceae GRIMMIALES

Grimmiaceae SYRRHOPODONTALES Ptychomitriaceae
Calymperaceae Grimmiaceae
Seligeriaceae
SELIGERIALES Incertae sedis:
Leucophanaceae Seligeriaceae Archidiaceae
Bryoxiphiaceae
GRIMMIALES GRIMMIALES Drummondiaceae
Grimmiaceae Ptychomitriaceae Schistostegiaceae
Ptychomitriaceae Grimmiaceae Scouleriaceae
Erpodiaceae Serpotortellaceae
Ephemeraceae
Splachnobryaceae
Wardiaceae
259
 TABLE 2. A list of the exemplar taxa used for the analyses. Taxon, DNA voucher number, and Genbank accession number or publication reference is given. The subclasses

260
are arranged following Vitt et al. (1998); the families (cf. Vitt 1984) and species are arranged alphabetically within their respective taxa. Ten exemplars are composite (cmp)
taxa, indicating that the three sequences were taken from different specimens. The sequences represent either two species of the same genus or two distinct specimens of the
same species. Abbreviations for the contact person for DNA voucher are as follows: BG–Bernard Goffinet, BM–Brent Mishler, CLF–Catherine La Farge, GB–Genbank, SS–
Steffan Schaffer. Regions not sampled for a species are indicated by n/a. Bold accession numbers indicate new sequences for this study.

Family/subclass Taxon DNA voucher # trnL Genbank rps4 rbcL

Funaridae
Funariaceae Funaria hygrometrica Hedw. (cmp) GB n/a AF023776 n/a
Funariaceae Funaria hygrometrica Hedw. (cmp) BM 49 AF231175 n/a AF226818
Orthotrichidae
Orthotrichaceae Macromitrium incurvifolium (Hook. & Grev.) Schwaegr. BM 202 AF231162 AF226754 AF226802
Orthotrichaceae Orthotrichum lyelli Hook. & Tayl. (cmp) GB n/a AF023814 n/a
Orthotrichaceae Orthotrichum pumilum Sw. (cmp) BM 83 AF231176 n/a AF226819
Bryidae
Bartramiaceae Bartramia halleriana Hedw. (cmp.) GB n/a n/a AF231090
Bartramiaceae Bartramia stricta Brid. (cmp.) GB AF023756 AF023799 n/a
Mniaceae Mnium hornum Hedw. (cmp) GB n/a AF023796 n/a
Mniaceae Mnium cuspidatum Hedw. (cmp) BM 82 AF231177 n/a U87082
Dicranidae
Bryoxiphiaceae Bryoxiphium norvegicum (Brid.) Mitt. CLF 131 AF231260 AF231267 AF231294
Calymperaceae Arthrocormus schimperi (Dozy & Molk.) Dozy. & Molk. BM 244 AF231157 AF226750 AF226797
Calymperaceae Calymperes afzelii Sw. (cmp) BM 165 AF231152 AF226744 AF226788
Calymperaceae Calymperes afzelii Sw. (cmp) BM 179 n/a n/a AF226789
Calymperaceae Calymperes tahitense (Sull.) Mitt. BM 123 AF231149 AF226741 AF226785
Calymperaceae Calymperes tuamotuense Bartr. BM 119 AF231154 AF226746 AF226793
THE BRYOLOGIST

Calymperaceae Exodictyon incrassatum (Mitt.) Card. BM 261 AF231189 AF226776 AF226833

Calymperaceae Exostratum blumei (Hampe) Ellis BM 243 AF231160 AF226753 AF226800
Calymperaceae Leucophanes albescens C. Müll. BM 136 AF231158 AF226751 AF226798
Calymperaceae Leucophanes glaucum (Schwaegr.) Mitt. BM 241 AF231159 AF226752 AF226799
Calymperaceae Mitthyridium constrictum (Sull.) H. Rob. BM 122 AF231143 AF226734 AF226778
Calymperaceae Mitthyridium junquilianum (Mitt.) H. Rob. BM 219 AF231144 AF226735 AF226779
Calymperaceae Mitthyridium obtusifolium (Lindb.) H. Rob. BM 121 AF231142 AF226733 AF226777
Calymperaceae Octoblepharum albidum Hedw. BM 125 AF231155 AF226747 AF226794
Calymperaceae Octoblepharum albidum Hedw. BM 242 AF231156 AF226748 AF226795
Calymperaceae Syrrhopodon banksii C. Müll. BM 124 AF231148 AF226740 AF226784
Calymperaceae Syrrhopodon confertus Sande Lac. BM 240 AF231151 AF226743 AF226787
Calymperaceae Syrrhopodon fimbriatulus C. Müll. BM 247 AF231150 AF226742 AF226786
Calymperaceae Syrrhopodon gardneri (Hook.) Schwaegr. BM 164 AF231146 AF226737 AF226781
Calymperaceae Syrrhopodon incompletus Schwaegr. BM 163 AF231181 AF226767 AF226824
Calymperaceae Syrrhopodon tortilis Hampe BM 160 AF231147 AF226738 AF226782
Dicnemonaceae Dicnemon calycinum (Hook.) Schwaegr. BM 144 AF231170 AF226763 AF226813
Dicnemonaceae Eucamptodon muelleri Hampe & C. Müll. CLF 122 AF231261 AF231280 AF231319
Dicranaceae Arctoa falcata (Hedw.) Loeske CLF 110 AF231234 AF231265 AF231292
Dicranaceae Arctoa fulvella (Dicks.) Bruch & Schimp. CLF 119 AF231235 AF231266 AF231293
[VOL. 103
 TABLE 2. Continued. 2000]

Family/subclass Taxon DNA voucher # trnL Genbank rps4 rbcL

Dicranaceae Brothera leana (Sull.) C. Müll. BM 260 AF231186 AF226773 AF226830
Dicranaceae Campylopus areodictyon (C. Müll.) Mitt. CLF 134 AF231237 AF231268 AF231295
Dicranaceae Campylopus argyrocaulon (C. Müll.) Broth. CLF 113 AF231238 AF231269 AF231315
Dicraaceae Campylopus atrovirens De Not. CLF 133 AF231239 AF231270 AF231316
Dicranaceae Campylopus eberhardtii Par. BM 138 AF231172 AF226765 AF226815
Dicranaceae Campylopus umbellatus (Arn.) Par. BM 137 AF231171 AF226764 AF226814
Dicranaceae Cynodontium jenneri (Schimp.) Stirton CLF 118 AF231240 AF231271 AF231318
Dicranaceae Dicranella heteromalla (Hedw.) Schimp. CLF 99 AF231241 AF231272 AF231296
Dicranaceae Dicranodontium denudatum (Brid.) Britt. CLF 135 AF231242 AF231273 AF231317
Dicranaceae Dicranodontium pulchroalare Broth. BM 259 AF231187 AF226774 AF226831
Dicranaceae Dicranoweisia cirrata (Hedw.) Lindb. CLF 112 AF231243 AF231274 AF231297
Dicranaceae Dicranum condensatum Hedw. CLF 98 AF231244 AF231275 AF231298
Dicranaceae Dicranum muehlenbeckii Bruch & Schimp. CLF 155 AF231245 AF231276 AF231299
Dicranaceae Dicranum scoparium Hedw. CLF 97 AF231246 AF231277 AF231300
Dicranaceae Holomitrium vaginatum (Hook.) Brid. BM 141 AF231168 AF226761 AF226811
Dicranaceae Kiaeria blyttii Bruch & Schimp. CLF 121 AF231252 AF231283 AF231306
Dicranaceae Leucoloma humbertii P. Varde CLF 66 AF231253 AF231284 AF231307
Dicranaceae Leucoloma rutenbergii (Geh.) Wright CLF 64 AF231254 AF231285 AF231308
Dicranaceae Leucoloma serrulatum Brid. CLF 150 AF231255 AF231286 AF231309
Dicranaceae Oncophorus wahlenbergii Brid. CLF 101 AF231256 AF231287 AF231310
Dicranaceae Orthodicranum fulvum (Hook.) Cas. Gil. CLF 36 AF231257 AF231288 AF231311
Dicranaceae Paraleucobryum enerve (Thed.) Loeske BM 255 AF231184 AF226770 AF226827
Dicranaceae Paraleucobryum longifolium (Hedw.) Loeske BM 256 AF231185 AF226771 AF226828
Ditrichaceae Ditrichum flexicaule (Schwaegr.) Hampe CLF 129 AF231247 AF231278 AF231301
Ditrichaceae Ditrichum pallidum (Hedw.) Hampe CLF 196 AF231248 AF231279 AF231302
Ditrichaceae Pleuridium acuminatum Lindb. CLF 197 AF231262 AF231289 AF231312
Erpodiaceae Aulocopilum hodgkinsoniae (Hampe & C. Müll.) Broth. BG 541 AF231236 AF222897 AF005545
Fissidentaceae Fissidens dubius P. Beauv. CLF 193 AF231249 AF231281 AF231303
LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES

Fissidentaceae Fissidens philonotulus Besch. BM 131 AF231167 AF226759 AF226809

Fissidentaceae Fissidens subbasilaris Hedw. CLF 192 AF231250 AF231282 AF231304
Grimmiaceae Grimmia pulvinata (Hedw.) Sm. BG 990 AF231251 AF222900 AF231305
Leucobryaceae Leucobryum antillarum Besch. BM 177 AF231165 AF226757 AF226806
Leucobryaceae Leucobryum sanctum (Schwaegr.) Hampe BM 258 AF231183 AF226769 AF226826
Leucobryaceae Leucobryum tahitense Ångström BM 126 AF231166 AF226758 AF226808
Leucobryaceae Schistomitrium robustum Dozy and Molk. BM 253 AF231182 AF226768 AF226825
Pottiaceae Pseudosymblepharis schimperiana (Par.) Crum BM 156 AF231164 AF226756 AF226805
Pottiaceae Stegonia latifolia (Schwaegr.) Broth. BG 779 AF231263 AF222901 AF231314
Pottiaceae Timmiella crassinervis (Hampe) Koch BM 155 AF231173 AF226766 AF226816
Pottiaceae Tortula obtusissima (C. Müll.) Mitt. (cmp.) BM 37 AF231180 n/a AF226823
Pottiaceae Tortula ruralis (Hedw.) G.M.S. (cmp) GB n/a AF023831 n/a
261
262 THE BRYOLOGIST [VOL. 103

Major clades.—A comparison of the strict con-

sensus trees from each analysis shows that 13
AF226821

AF231313
AF226812
AF005543
AF005544

AF005547

AF226822
AF226817
clades are consistent, despite the large variation in
rbcL

n/a

n/a

n/a
n/a
numbers of characters between the data sets (Table
4, Fig. 1). These clades are 1) Calymperaceae, 2)
Octoblepharaceae, 3) Leucobryaceae, 4) Fissiden-
taceae, 5) Dicranaceae, 6) Dicranella, 7) Potti-
aceae, 8) Rhachitheciaceae-Erpodiaceae, 9) Ditri-
Genbank rps4

AF231290
AF023779

AF226762
AF222896

AF231291

AF023780

AF023701
chaceae, 10) Rhabdoweisiaceae, 11) Amphidium,
AF22898

12) Grimmiales, and 13) Seligeriaceae. For the spe-

n/a

n/a

n/a

n/a
cies composition of these 13 clades, the exemplar
taxa are indicated in the trnL-rps4-rbcL analysis
(Fig. 2). Scouleria aquatica, Ditrichum flexicaule,
Timmiella crassinervis, and Bryoxiphium norvegi-
cum form independent basal lineages within the Di-
AF231178

AF231258
AF231169
AF231233

AF231264

AF231179
AF231174
AF231259

cranidae, except Timmiella in the trnL-rps4 analy-

trnL

n/a

n/a

n/a

n/a

sis (Fig. 1, see legend).

The topologies of the trnL-rps4-rbcL and rps4-
rbcL strict consensus trees are essentially the same
for the 13 major clades (Fig. 1). The exception is
clade four (Fissidentaceae), which is split in the
DNA voucher #

rps4-rbcL analysis. Another difference is the reso-

BM 143
BG 577

BG 987
BG 533
BG 765
BG 545
BG 749

lution of the relationships of the 13 clades. In the

SS/BM
SS/BM
BM 99

rps4-rbcL analysis these clades have a resolved re-

GB

GB

lationship, whereas in the trnL-rps4-rbcL the clades

form two major polytomies. The third difference is
the resolution of the basal lineages of the Dicrani-
dae; in trnL-rps4-rbcL they are resolved, whereas
they form a basal polytomy in rps4-rbcL. The re-
Ptychomitrium drummondii (Wils.) Sull. & Lesq. (cmp).

sults of the trnL-rps4 analysis suggest a compara-

tively different topology of the 13 clades. The ma-
jor discrepancy is with the placement of clades
Rhabdoweisia crenulata (Mitt.) Jameson (cmp)

Uleastrum palmicola (C. Müll.) Zander (cmp)

three and 10 in a more basal position within the

Rhabdoweisia crispata (With.) Lindb. (cmp)

Uleastrum paraguense (Besch.) Buck (cmp)

tree. The other difference is the inclusion of Tim-

Amphidium californicum (C. Müll.) Broth.
Amphidium lapponicum (Hedw.) Schimp.

Blindia acuta (Hedw.) Bruch & Schimp.

miella within the Pottiaceae, clade seven. The re-

Ptychomitrium gardneri Lesq. (cmp)

lationships among the 13 major clades are resolved

in this analysis. The reason for the comparatively
Taxon

Scouleria aquatica Hook. (cmp)

Scouleria aquatica Hook. (cmp)

large number of MP trees (24 versus three or four)

Ptychomitrium gardneri Lesq.

is three regions of the tree with polytomies among

species or genera (Fig. 1).
Support (bootstrap estimates (BV) ⱖ 50% and
decay indices (DI)) for the major clades is sum-
marized for each analysis in Table 5. These lineages
form robust clades (BV ⱖ 70%, DI ⬎ 3) with few
exceptions. Given the topological discrepancies of
the three analyses, support values for the 13 clades
indicates that the trnL-rps4-rbcL analysis has the
best overall values for these identified clades (Table
5) and incorporates the maximum number of char-
Continued.

acters. Therefore, the remainder of the paper will

concentrate on this analysis.
Rhabdoweisiaceae
Rhabdoweisiaceae
Rhabdoweisiaceae
Rhabdoweisiaceae
Rhachitheciaceae
Rhachitheciaceae
Ptychomitriaceae
Ptychomitriaceae
Ptychomitriaceae

Clades one and two are unequivocally robust, in-

Family/subclass

Scouleriaceae
Scouleriaceae

dicating support for both the Calymperaceae and

Seligeriaceae
TABLE 2.

Octoblepharaceae (sensu Eddy 1990) given the cur-

rent taxon sampling (Fig. 2). The sister relationship
of these clades is robust as well (Table 5). The Oc-
toblepharaceae, represented by a single species,
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 263

TABLE 3. Primers and Primers Sequences. Publication or contact person is indicated.

Primer
region Primer Primer sequence Reference
trnL-F
trnC (forward) 5⬘ CGA AAT CGG TAG ACG CTA CG 3⬘ Taberlet et al. 1991
trnF (reverse) 5⬘ ATT TGA ACT GGT GAC ACG AG 3⬘ Taberlet et al. 1991
rps4
rps5 (forward) 5⬘ ATG TCC CGT TAT CGA GGA CCT 3⬘ Souza-Chies et al. 1997
trnS (reverse) 5⬘ TAC CGA GGG TTC GAA TC 3⬘ Souza-Chies et al. 1997
rbcL
M007F (forward) 5⬘ CCA CAA ACG GAG ACT AAA GC 3⬘ Wheeler, per. comm.
M665F (forward) 5⬘ GGA GAG ATC GTT TCG TAT TTG TAG C 3⬘ Wheeler, per. comm.
M745R (reverse) 5⬘ CTT CAC AWG TAC CTG CRG TAG C 3⬘ Wheeler, per. comm.
M985 (forward) 5⬘ GGT ACT GTW GTA GGT AAA C 3⬘ Lewis et al. 1997
M1244R (reverse) 5⬘ CTC GAT TAG CAA CTG CAC CAG G 3⬘ Wheeler, per. comm.
M1390R (reverse) 5⬘ CTT TCC AWA YTT CAC AAG CAG CAG 3⬘ Lewis et al. 1997
MtrnRR (reverse) 5⬘ GCT CTA ATC CAC TGA GCT AC 3⬘ Wheeler, per. comm.

have a long branch (Fig. 3). Clade three (Dicrana- included. Clade 11 (Amphidium) is strongly sup-
ceae) has low bootstrap support (BV ⫽ 62), but a ported as a distinct clade, although its relationship
strong decay index (DI ⫽ 4). Clade four (Fissiden- is not robustly resolved. Clade 12 (Grimmiales) has
taceae) including Fissidens philonotulus, is poor, robust support. Clade 13 (Seligeriaceae) represent-
which may represent a potential long branch prob- ed by a single exemplar species, has robust support
lem (Fig. 3). Otherwise, the clade has robust sup- as the sister group to clade 12, Grimmiales. The
port for the two remaining exemplars. Clade five majority of the 13 clades within the Dicranidae
(Leucobryaceae) has low support (Fig. 2, Table 5). have robust support (Table 5, Fig. 2). Decay indices
With the exclusion of the basal taxon, Campylopus along the ‘‘backbone’’ of the consensus tree are
eberhardtii, the Leucobryaceae clade is robust. weak (DI ⫽ 1) with bootstrap values ⬍ 50% (Fig.
Clade six (Dicranella heteromalla) is represented 2), indicating that relationships among the major
by a single exemplar of the Dicranaceae subfamily clades of the subclass are tenuous. Clades one to
Dicranelloideae, thus the clade has no support val- six form a major polytomy, as well there is weak
ues. Clade seven (Pottiaceae) forms a robust clade, support for the relationships of clades 7–11 (DI ⫽
excluding Timmiella crassinervis, which forms a 1,2; BV ⬍ 50%).
basal lineage within the Dicranidae (Fig. 2, Table In order to evaluate our results with respect to
5). Clade eight (Rhachitheciaceae and Erpodiaceae) previous familial classifications, we performed a
has weak BV and strong DI values for the sister constraint analysis on the data. When the trnL-
relationship of the two taxa it represents (Table 5). rps4-rbcL data set is constrained to the ordinal and
Clade nine (Ditrichaceae, excluding Ditrichum flex- familial classification, following Vitt (1984), in-
icaule) forms a robust clade. Clade 10 (Rhabdow- cluding modifications of Vitt et al. (1998), tree
eisiaceae) has moderate support for the six genera length is increased by 242 steps, consistency indi-

TABLE 4. Tree and character statistics from the phylogenetic analyses using parsimony from the different character
sets of the Dicranidae. The analyses column indicates the cpDNA region used in each analysis. The constraint analysis
of trnL-rps4-rbcL provides the results of the data constrained to the orders and families of Vitt (1984) and Vitt et al.
(1998). The following abbreviations refer to the columns. CHARACTERS: total number used–total number of char-
acters used for the different analyses with ambiguous and predominantly missing regions excluded; number informa-
tive–number of informative characters for each analysis, numbers in parentheses indicate the contribution of informative
characters from each respective region; CIa–consistency index with uninformative characters included; CIb–consistency
index with uninformative characters excluded; RI–retention index. TREES: MP–the number of most parsimonious
trees; number of clades forming polytomies–the number of internal polytomous clades.

# Clades
forming
Characters Trees polyto-
Analyses Total # used # Informative CIa CIb RI length MP mies
trnL-rps4 959 334 (115, 219) 0.431 0.367 0.715 1493 24 14
rps4-rbcL 1926 559 (219, 340) 0.411 0.344 0.677 2618 3 4
trnL-rps4-rbcL 2335 674 0.415 0.344 0.678 3191 4 8
Constraint 2335 674 0.385 0.3172 0.636 3433 12 12
264 THE BRYOLOGIST [VOL. 103

FIGURE 1. The three censensus trees from the three different combinations of cpDNA data sets with the number of
informative characters and number of most parsimonious trees given at the top for each respective combination. The
13 major taxonomic clades identified and four basal taxa of the Dicranidae are indicated. The taxonomic composition
of each enumerated clade is identical for each analysis, except clade four (Fissidentaceae) which is polyphyletic in the
rps4-rbcL analysis, and clade seven (Pottiaceae) in the trnL-rps4 analysis with includes Timmiella crassinervis. The
arrangement of the taxa in the resolved consensus tree of the trnL-rps4 data set is distinct from the congruent arrange-
ment of the other two analyses. Decay indices superimposed on the topology indicate the low or lacking support for
the relationships between the major taxonomic clades within the Dicranidae. The heavy black lines indicate the poly-
tomous regions of each analysis to show the unresolved regions for each analysis.
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 265

FIGURE 2. The trnL-rps4-rbcL analysis strict consensus tree of four most parsimonious trees. The exemplar com-
position of the 13 major clades within the Dicranidae is given. Bootstrap (⬎ 50%) and decay indices are superimposed
on the tree for the basal and internal nodes representing the major clades (bootstrap value above and decay index below
the clade). Values within the clades are not shown. The clade defining the Dicranidae is indicated. Clade 12 (Grim-
miales) includes exemplars form the Grimmiaceae and Ptychomitriaceae. Two additional clades are illustrated at the
familial rank, 1) Bryoxiphiaceae, a monotypic family, represented by Bryoxiphium norvegicum, and 2) Scouleriaceae,
also monotypic, represented by Scouleria aquatica. OG–out-group taxa.
266 THE BRYOLOGIST [VOL. 103

TABLE 5. Support indices for the 13 major taxonomic clades found in the three Fitch parsimony analyses: trnL-
rps4-rbcL, rps4-rbcL, trnL-rps4. Additional clades indicated by *, for combined clades or clades excluding single
species. Bootstrap values ⱖ50% and decay indices are given. Additional clades are given to show robust indices with
the exclusion of a species from a major clade or the support for combined clades. Numbers in bold show the highest
BV values and decay indices for each clade. Dashes indicate where support indices do not apply.

Analysis
trnL/rps4/rbcL rps4/rbcL trnL/rps4
Clade Taxon Bootstrap Decay Bootstrap Decay Bootstrap Decay
1 Calymperaceae 100 20 100 18 100 10
2 Octoblepharaceae 100 87 100 62 100 51
*clades 1 ⫹ 2 77 6 72 3 67 5
3 Dicranaceae 62 4 ⬍50 1 78 7
4 Fissidentaceae ⬍50 2 ⬍50 0 63 4
*excluding Fissidens philonotulus 100 48 100 40 100 29
5 Leucobryaceae 59 1 82 4 89 9
*excluding Campylopus eberhardtii 99 16 97 8 50 3
6 Dicranella — — — — — —
7 Pottiaceae 99 15 98 14 90 10
8 Rhachitheciaceae/Erpodiaceae 51 4 52 4 ⬍50 1
9 Ditrichaceae 100 9 100 7 79 5
10 Rhabdoweisiaceae 60 3 52 2 61 4
11 Amphidium 100 13 100 12 93 7
12 Grimmiales 90 8 78 3 99 7
13 Seligeriaceae — — — — — —
*clades 12 ⫹ 13 76 6 53 4 ⬍50 4

ces are lower, 12 MP trees are produced, and there Clade 2 (Octoblepharaceae). Eddy (1990) placed
are 12 unresolved internal branches in the consen- Octoblepharum in a monotypic family, based on
sus tree (Table 4). the specialized expanded costa with several layers
of hyalocysts, diminutive calyptrae, peristome type
DISCUSSION (papillose distally, smooth or reticulate below, ru-
dimentary prostome), and autoicous sexuality. Al-
Monophyly of the Dicranidae.—The results of
len (1994) noted that some neotropical species of
our trnL-rps4-rbcL analysis are used to focus our
Octoblepharum are dioicous, thus autoicous sexu-
discussion of phylogenetic relationships within the
ality can not be used as a distinguishing character
Dicranidae (Figs. 2–3). The consensus tree indi-
cates that the Dicranidae form a well supported of the Octoblepharaceae. Eddy’s distinction of the
monophyletic clade, based on the five outgroup Octoblepharaceae from the Calymperaceae is fol-
taxa. These data are congruent with previous mo- lowed here. In our study, Octoblepharum is repre-
lecular studies that include a broad selection of ar- sented by two exemplar taxa (old world and new
throdontous taxa (e.g., Beckert et al. 1999; Cox & world O. albidum) that share a distinctive long
Hedderson 1999; Goffinet et al. 1998). Within the branch. Additional sampling is clearly needed to
Dicranidae, our results suggest that three of the four determine the affinity of the two families; however,
haplolepideous orders recognized by Vitt (1984) a sister relationship between the Calymperaceae
and Vitt et al. (1998) are polyphyletic (Figs. 3–4; and Octoblepharaceae is well supported (DI ⫽ 6,
Table 1). BV ⫽ 77). Further taxon sampling may break down
Familial relationships within the Dicranidae.— the long branch and improve the support for the
Clade 1 (Calymperaceae). Circumscriptions of the sister relationship of the two families (Graybeal
Calymperaceae have reflected either a restricted 1998). The Calymperaceae and Octoblepharaceae
concept, including Calymperes, Syrrhopodon, and together form a clade that is more closely related
Mitthyridium (e.g., Brotherus 1924; Reese & Stone to other families of the Dicranales, than to the Pot-
1995; Reese et al. 1986) or an expanded concept, tiaceae of the Pottiales.
including Arthrocormus, Calymperopsis, Carina- Clade 3 (Dicranaceae). The current analyses in-
folium, Exodictyon, Hypodontium, Leucophanes, dicate that Dicranaceae (sensu Vitt 1984) are poly-
and Octoblepharum (e.g., Crosby & Magill 1981; phyletic (Figs. 2–3). Our data suggest the following
Vitt 1984). Based on the current data, DNA se- modifications of the familial circumscription 1)
quences from the chloroplast genome support the transfer of genera to the Rhabdoweisiaceae (Arctoa,
expanded concept, regardless of whether Octoble- Oncophorus, Kiaeria, Cynodontium, and Dicra-
pharum is included (BV ⫽ 77, DI ⫽ 6). noweisia); 2) exclusion of the subfamilies Campy-
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 267

FIGURE 3. Phylogram of one of the four most parsimonious (MP) trees from the trnL-rps4-rbcL analysis. The
arrows indicate the clade corresponding to the numbered clade, horizonally opposite. The polyphyletic orders are
indicated–Pottiales (gray-circles); Dicranales (gray-boxed), and Grimmiales (clear-boxed) following Vitt (1984) and
Vitt et al. (1998). Seligeriales (Seligeriaceae) are not shaded. The ordinal classification of Rhachitheciaceae/Erpodiaceae
has not been robustly determined. OG–out-group taxa. A bar (lower left) indicates branch length equivalent to 10
changes.
268 THE BRYOLOGIST [VOL. 103

FIGURE 4. A summary of the trnL-rps4-rbcL strict consensus tree indication the provisional clades that correspond
to three orders within the Dicranidae. Each order is boxed and shaded with a different gray. The Dicranidae are also
boxed and shaded. the numbers in parentheses () after name indicates the number of exemplar taxa representing each
clade. Decay indices are given for the internal nodes. Total number of taxa, parsimony informative characters, most
parsimonious (MP) trees, and tree length (TL) of the MP trees are given at the top. Out-group clades are indicated at
the bottom.
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 269

lopodioideae and Dicranelloideae, and Brothera (Paraleucobryoideae) have been included within the
leana; and 3) synonymy of the Dicnemonaceae Dicranaceae based on the typical bifid, vertically
with the Dicranaceae. Our data support Stech’s pitted-striolate peristome (Frahm 1991a,b; Walther
(1999a) results from trnL data, that Brothera is 1983). Eddy (1990) observed that Brothera has a
more closely related to members of subfamily Cam- similar peristome structure to that of Microdus (Di-
pylopodioideae than the Paraleucobryoideae, al- cranaceae) and its gametophyte structure is similar
though with different inferences for the familial re- to a juvenile Leucobryum (Leucobryaceae). Based
lationship. The current data support Eddy’s sugges- on our molecular data, Brothera is more closely
tion that the Dicnemonaceae should be synony- related to members of the Leucobryaceae, including
mized with the Dicranaceae, and Allen’s concept a paraphyletic Campylopodioideae, than to the Di-
(Allen 1987) that the Dicnemonaceae are closely cranaceae. The data support a monophyletic group
related to the Dicranaceae and represent a ‘‘recently that includes exemplars of the Leucobryaceae, the
derived’’ dicranaceous lineage. The circumscription Campylopodioideae, and Brothera leana. These
of this traditionally large, heterogeneous family is taxa, with typical dicranoid peristomes (reduced in
the focus of a more in depth study (La Farge & Eddy’s Schistomitrium group) and a diversity of
Shaw, unpubl. data). The data here indicate that a costal specialization’s (broadly expanded costae
number of autoicous taxa are purged from the fam- with/without dorsal lamellae, costae consisting of
ily, leaving dioicous or pseudautoicous exemplar leucocysts and hyalocysts) warrant a reevaluation
taxa to represent the family. The family retains taxa of the Leucobryaceae.
with differentiated alar cells and a dicranoid peri- Clade 6 (Dicranella). Dicranella heteromalla is
stome typified by bifid, vertically pitted-striate be- the sole representative of Dicranaceae subfamily
low, transversely striate in mid-section, and distally Dicranelloideae. The position and validity of the
papillose teeth. It should be noted that a broad lineage it represents is tenuous without additional
range of peristome variation occurs within the fam- taxon sampling. At present, it is resolved as an dis-
ily, including smooth and papillose ornamentation, tinct lineage that forms a polytomy with the Calym-
rudimentary prostome development, and entire, peraceae-Octoblepharaceae clade, Leucobryaceae,
perforate, bifid, or trifid teeth. Dicranaceae and Fissidentaceae. The Dicranello-
Clade 4 (Fissidentaceae). The Fissidentaceae ideae have been distinguished by the lack of alar
form a monophyletic group, although weakly sup- cell differentiation, capsules commonly ribbed, and
ported when Fissidens philonotulus is included the typical dicranoid peristome consisting of 16 bi-
(Figs. 2–3; Table 5). The long branch of F. philon- fid, pitted-striate teeth. In contrast, in a more de-
otulus is considered to be a problem of limited sam- tailed study of the Dicranaceae, the Dicranelloideae
pling. Given the unique development of bi-vagi- are resolved sister to the Pottiaceae (La Farge &
nant, distichous leaves, the monophyly of the Fis- Shaw, unpubl. data). Further sampling of the Di-
sidentaceae is presumed. The current data indicate cranelloideae is needed to resolve its placement and
that the Fissidentaceae are well nested within fam- rank.
ilies of the Dicranales (Fig. 3), therefore supporting Clade 7 (Pottiaceae). The Pottiaceae, the largest
its inclusion within the Dicranales (Vitt et al. 1998). family of the Dicranidae with 76 genera (Zander
Peristome variation of the Fissidentaceae is well 1993), are represented by four exemplar species,
within that of the Dicranales (Bruggeman-Nannen- therefore only provisional comments are warranted.
ga & Berendsen 1990). The family is polyphyletic with Timmiella separat-
Clade 5 (Leucobryaceae). Vitt (1984) and Peter- ed from the remaining genera and forming a basal
son (1994) circumscribed the Leucobryaceae to in- clade within the Dicranidae. The Pottiaceae are re-
clude four genera: Leucobryum, Schistomitrium, solved as sister to the Rhachitheciaceae-Erpodi-
Ochrobryum, and Cladopodanthus. In contrast, aceae clade. Zander (1993) described the diversity
Eddy (1990), distinguished a new family, the Schis- of peristome characters in the Pottiaceae. These
tomitriaceae (Schistomitrium, Ochrobryum, and peristomes types are either lacking, rudimentary, or
Cladopodanthus), based on entire, rarely forked, consisting of 16 paired filiform teeth, or 32 com-
papillose teeth, erect, symmetric capsules, the lack pletely divided prongs (twisted or not). The orna-
of dwarf males, and mitrate calyptra. Our molecular mentation ranges from smooth, striated or ridged,
data indicate that Schistomitrium is nested within to papillose or spiculose. Zander (1993) has noted
Leucobryum, which does not support the recogni- peristomial affinities with members of the Ditricha-
tion of Eddy’s distinct family. Clade five forms a ceae (see clade 9). Further sampling is required to
resolved, robust group with the exclusion of Cam- establish a comprehensive assessment of the two
pylous eberhardtii (BV ⫽ 99, DI ⫽ 16). Tradition- families.
ally, the Campylopodioideae (represented here by Clade 8 (Rhachitheciaceae/Erpodiaceae). The
Dicranodontium and Campylopus) and Brothera exemplars of Rhachitheciaceae and Erpodiaceae
270 THE BRYOLOGIST [VOL. 103

form a weak monophyletic clade (BV ⫽ 51, DI ⫽ aceae form a monophyletic clade, albeit with low
4) nested within the Dicranidae. Edwards (1979) support indices (BV ⫽ 60, DI ⫽ 3). Our molecular
described a seligeriaceous peristome as having a data support an expanded concept of the family to
well developed endostome and reduced exostomial include Arctoa, Cynodontium, Dicranoweisia, Kia-
membrane that adheres to the base of the endosto- eria, and Oncophorus. The data suggest that the
me. He observed this type of prostomial develop- family includes autoicous genera characterized by
ment in other haplolepideous taxa and suggested entire or bifid, smooth, striate or papillose peri-
that Venturiella (Erpodiaceae) has a similar type of stome teeth, and taxa with or without developed
peristome. The Rhachitheciaceae and Erpodiaceae alar cells. Further taxon sampling of additional
have recently been transferred from the diplolepi- families within the Pottiales and Dicranales is re-
deous Orthotrichales to the Dicranidae based on quired to test the monophyly of this clade. The fam-
rbcL sequences (Goffinet et al. 1998). Their place- ily is sister to a clade including the Pottiaceae,
ment is tested here within a broad sampling of the Rhachitheciaceae/Erpodiaceae, and Ditrichaceae,
subclass. Their apparent relationship may be prod- but this relationship is also with weak support in-
uct of long branch attraction rather than common dices (BV ⬍ 50%, DI ⫽ 1).
ancestry (Felsenstein 1978; Hendy & Penny 1989), Clade 11 (Amphidium). The systematic position
therefore their relationship and placement are con- of Amphidium has been problematic, given the lack
sidered provisional. Our data indicate that the two of a peristome. Some authors have considered Am-
families are more closely related to pottiaceous taxa phidium a member of the Dicranidae, with morpho-
than to the Seligeriaceae. Prostomial development logical affinities to Rhabdoweisia (e.g., Loeske
found in members of the Rhachitheciaceae and Se- 1910; Malta 1926; Vitt 1973), whereas others have
ligeriaceae has occurred frequently in scattered taxa considered it a member of the Orthotrichales (Crum
within all haplolepideous orders–Pottiales, Grim- & Anderson 1981; Lewinsky 1976; Smith 1978).
miales, and Dicranales (e.g., Edwards 1979; Gof- Goffinet et al. (1998) excluded Amphidium from the
finet 1998; La Farge-England 1998). Orthotrichales based on rbcL sequence data and in-
Clade 9 (Ditrichaceae). The Ditrichaceae are dicated its affiliation with members of the Dicran-
polyphyletic, with Ditrichum flexicaule forming a idae. Subsequent molecular data have supported
distinct, isolated basal lineage. Increased taxon Goffinet et al. (1998) exclusion of Amphidium from
sampling within the Ditrichaceae should improve the Orthotrichales and its transfer to the Dicranidae
the phylogenetic assessments of the family. The Di- (Stech 1999b). Our analyses indicate that it is clear-
trichaceae are distinguished from the Dicranaceae ly a member of the Dicranidae, as well as an in-
by the terete, filiform peristomes that are divided to dependent lineage sister to the ‘pottiaceous’ clades
the base, forming 32 papillose prongs (Crum & An- (7–10). Based on the poor support values (BV ⬍
derson 1981; Eddy 1988; Matsui & Iwatsuki 1990). 50%, DI ⫽ 1) for the sister relationship of Amphi-
The Ditrichaceae are resolved as sister to a clade dium to clades 7–10, the data are considered insuf-
including the Pottiaceae and Rhachitheciaceae/Er- ficient to determine a new familial status of Am-
podiaceae, although this relationship is weak (BV phidium without further taxon sampling (e.g.,
⬍ 50%, DI ⫽ 2) it does suggest that the Dicrana- Drummondia).
ceae are more distantly related. Previous classifi- Clades 12–13. The Seligeriaceae and Grimmi-
cations have inferred a close relationship among the ales, excluding Scouleria, form a monophyletic
Dicranaceae and Ditrichaceae (Table 1). Both the clade. The Grimmiales are represented by exem-
Pottiaceae and Ditrichaceae share the papillose, di- plars from the Grimmiaceae and the Ptychomitri-
vided peristome structure. Zander (1993) discussed aceae that have a robust sister relationship. As well,
the relationship of several Ditrichum species to the sister relationship of the Seligeriaceae and
Barbula and suggested that these species should be Grimmiales has strong support. The data suggest
transferred to Barbula, based on costal anatomy the synonymy of the Seligeriales with the Grim-
and the twisted peristomes. His observations are miales, yet further taxon sampling is required to test
consistent with the current analysis, that suggests a this result.
close phylogenetic relationship between the Ditri- Bryoxiphiaceae. Bryoxiphium norvegicum is the
chaceae and Pottiaceae. representative taxon of the monotypic Bryoxiphi-
Clade 10 (Rhabdoweisiaceae). The Rhabdowei- aceae. The family has been placed in its own order
siaceae were initially described by Limpricht (Crum 1994; Crum & Anderson 1981), considered
(1886) to include Cynodontium, Dichodontium, Or- a member of the Dicranales (Brotherus 1924; Vitt
eas, Oreoweisia, and Rhabdoweisia. Anderson and 1984) or included as a member of the Fissidentales
Crum (1959) and Vitt (1984) have used a restricted (Walther 1983). The eperistomate condition has
concept of the family to include Amphidium, Rhab- made it difficult to place Bryoxiphium, for peri-
doweisia, and Rhabdoweisiella. The Rhabdoweisi- stome structure traditionally provides the criteria
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 271

for higher level groupings. Our molecular data in- the Orthotrichales and Hypnales of the Diplolepi-
dicate that it forms a basal lineage within the Di- deae. Dioicous taxa are the presumed plesiomorph-
cranidae however its ordinal status or familial af- ic with autoicous or specialized pseudautoicous
filiations have yet to be determined. taxa apomorphic.
Scouleriaceae. Scouleria aquatica has been not- Prostomial development is found in all orders of
ed for its unique haplolepideous peristome with 32 the Dicranidae in scattered taxa within various fam-
irregular teeth and two rudimentary prostomial lay- ilies (e.g., Leucoloma, Octoblepharum, Seligeria).
ers (Edwards 1979). It was recently removed from The distribution among the genera suggests that
the Grimmiaceae and placed in its own family, prostomes have evolved independently in multiple
Scouleriaceae (Churchill 1981). Edwards (1979) lineages or are residual features. Edwards (1979)
suggested that it had peristome similarities to the found that for some taxa the additional peristome
Seligeriaceae. Our current molecular data supports layers are shed with the operculum, therefore rarely
the isolation of this taxon as one of the basal lin- observed in the specimens. The distribution and
eages in the Dicranidae. evolutionary significance of this character needs
Ordinal relationships within the Dicranidae.— further study in the Dicranidae.
Our results suggest that within the Dicranidae, 13 Taxa of the Grimmiales clade have 16 entire, per-
major clades form three monophyletic ordinal forate, or divided teeth (smooth or papillose) that
groups: Dicranales, Pottiales, and Grimmiales (Fig. are paired (eight teeth) or form 32 filiform prongs.
4). Identification of these clades implies the need In the Pottiales, peristomes typically consist of 16
for revised familial circumscriptions within the or- entire, bifid, perforate teeth or 32 filiform prongs.
ders (Vitt 1984). This re-arrangement of the famil- The teeth are either flat or terete, typically papillose
ial relationships is considered provisional, based on on the outer surface, sometimes striate to smooth
a significant lack of representative samples for all at the base. Members of the Pottiaceae develop spe-
families and low levels of support for some of the cialized high basal membranes in selected taxa. The
relationships. Despite these caveats, our results do Dicranales are characterized by 16 bifid teeth that
permit some preliminary considerations of peristo- are striate at the base and often papillose distally
mial and ecological evolution (Table 6). (although a number of taxa in the Leucobryaceae,
Morphological and ecological implications.— Dicranaceae, and all members of the Calympera-
Given our data set, and the topology of the sum- ceae have entire, papillose teeth). Peristome varia-
mary tree (Fig. 4) the distribution of selected mor- tion within the orders of the Dicranidae indicates
phological and ecological characters (Table 6) are that papillose teeth occur in all lineages, but that
used to suggest the following trends within the Di- striate teeth are not found in the Grimmiales. With-
cranidae. These data are not exhaustive. The basal in lineages with well-developed, vertically striate
taxa are predominantly saxicolous or saxicolous- teeth, recurrent trends lead to reduced papillose or
terricolous in their substrate preferences. This in- smooth teeth (e.g., La Farge-England 1998). Sev-
cludes the Bryoxiphiaceae, Scouleriaceae, and four eral lineages that show a trend from well-developed
of the six taxa in Pottiales clade. The Dicranales papillose teeth as the ancestral stock to smooth, re-
clade includes families from a broad spectrum of duced, paired peristome teeth as in Octoblephara-
habitats; all of which have taxa that specialize in ceae (Eddy 1990) and Rhachitheciaceae (this
epiphytic niches. The Pottiales clade has two fam- study), and Pottiaceae (Zander 1993). Lacking ro-
ilies that are specialized and restricted to epiphytic bust ordinal relationships, interpreting peristome
habitats. From the topology of the summary tree, a evolution within the Dicranidae is considered pre-
generalized ecological trend is from basal saxico- mature.
lous-terricolous habitats to more distal humicolous- The variation of stereid band development is var-
lignicolous-epiphytic habitats. ied within the orders. Within the Grimmiales clade,
Sexuality within the Dicranidae includes dioi- the Seligeriaceae lack stereid band development,
cous basal taxa; dioicous or autoicous taxa in the the Grimmiaceae develop a single stereid band,
Grimmiales, Pottiales, and Dicranales; and dioicous while the Ptychomitriaceae develop two stereid
or pseudautoicous taxa within the Dicranales. Pseu- bands. Within the Pottiales clade, the costal anato-
dautoicous taxa develop specialized dwarf males my shows a broad range, 0–2 stereid bands, with
that increase the probability of fertilization in an some taxa of the Ditrichaceae having a broadly ex-
otherwise dioicous taxon by having the male spores panded costa. Reduced costae occur in the Rhach-
germinate on the mature female plants. The dwarf itheciaceae and Erpodiaceae that are restricted to
males commonly develop on the stem tomentum or epiphytic habitats. The Dicranales include taxa with
in the leaf axils. This type of sexuality is common specialized costal development. Broadly expanded
among members of the Dicranaceae. Pseudauto- costae showing a large number of guide cells (⬎ 8,
icous taxa have also been independently derived in e.g., Campylopus, and some species of Dicranolo-
 272
TABLE 6. Selected characters of the 13 major clades represented in the current analyses. The categories are as follows: Taxon; substrate; sex; carpy–the development of
cladocarpy indicated in otherwise acrocarpous lineages; peristome characters–presence of prostomial development in the OPL (outer peristome layer), structure of the peristome
tooth, the number (#) of teeth, and ornamentation on the tooth; alar cells–presence of alar cell differentiation; and costal anatomy–the number of stereid bands developed and
the presence of broadly expanded costae or specialized development of chlorocysts (chloro) and leucocysts (leuco). The character data set was compiled for the representative
families from Allen (1987), Crum and Anderson (1981), Bruggeman-Nannenga and Berendsen (1990), Eddy (1988, 1990), Frahm 1991a,b, La Farge-England (1996, 1998),
Lawton (1971), Magill (1981), Sharp et al. (1994), Smith (1978), Zander (1993), and personal observations.

# Alar Costal
Taxon Substrate Sex Carpy OPL Peristome Teeth Ornamentation cells anatomy
Incertus sedis:
Bryoxiphiaceae saxicolous dioicous none 1
Scouleriaceae saxicolous dioicous 2 prostomes divided 16 (32) smooth 2
GRIMMIALES
Seligeriaceae saxicolous autoicous/dioicous prostome entire 16 smooth alars 0
Ptychomitriaceae saxicolous autoicous prostome entire/cleft/perforate 8, 16 smooth/papillose 2
Grimmiaceae saxicolous autoicous/dioicous cladocarpy prostome entire/cleft/perforate/di- 16 (32) papillose 0, 1
vided
POTTIALES
Amphidium saxicolous autoicous/dioicous none 1
Rhabdoweisiaceae saxicolous/terricolous autoicous entire/bifid/perforate 16 smooth/striate/pa- alars 2
pillose
Ditrichaceae saxicolous/terricolous autoicous/dioicous divided 16 (32) papillose 1, 2, expanded
THE BRYOLOGIST

Pottiaceae saxicolous/terricolous/epi- autoicous/dioicous cladocarpy prostome entire/divided 16 (32) smooth/striate/pa- 1,2

phytic pillose
Rhachitheciaceae epiphytic autoicous/dioicous entire 8 smooth 0,1
Erpodiaceae epiphytic autoicous cladocarpy entire 16 papillose 0
DICRANALES
Dicranella saxicolous/terricolous autoicous/dioicous bifid 16 striate 2
Fissidentaceae saxicolous/terricolous/ligni- autoicous/dioicous cladocarpy entire, bifid 16 smooth/striate/pa- 2
colous/epiphytic pillose
Leucobryaceae terricolous/lignicolous/epi- dioicous/pseudauto- cladocarpy prostome entire, bifid 16 striate/papillose 2, expanded
phytic icous chloro/leuco
Dicranaceae terricolous/lignicolous/epi- dioicous/pseudauto- cladocarpy prostome entire, bifid 16 smooth/striate/pa- alars 2, expanded
phytic/humicolous icous pillose chloro/leuco
Octoblepharaceae saxicolous/humicolous/epi- autoicous/dioicous prostome entire 8, 16 smooth expanded chlo-
phytic ro/leuco
Calymperaceae epiphytic dioicous cladocarpy prostome entire 16 papillose 2, expanded
chloro/leuco
[VOL. 103
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 273

ma and Dicranum), occur in two of the six clades characters to the chloroplast genome alone, one
(excluding Dicranella and Fissidens), or the ex- cannot expect the results presented here to provide
panded costae composed of specialized leucocysts a fully accurate phylogeny. Further sampling, es-
and chlorocysts (e.g., Octoblepharum, Leucobryum, pecially from the nuclear genome is needed.
Paraleucobryum) occur in four of the six clades. Each of the three separate analyses produced
The Ditrichaceae, placed here in the Pottiales clade, similar results, identifying the same 13 major
also develops a broad costa in selected taxa. Given clades within the Dicranidae (Fig. 1). The least re-
the polytomy of the familial relationships within the solved phylogeny (trnL-rps4) showed ambiguity of
Dicranales, the ancestral history of these costal fea- generic and species level relationships, with deeper
tures cannot be established without a refined phy- level relationships resolved. The data demonstrate
logeny of the taxa. It can be stated with certainty that an increase of informative characters (334 to
that the expanded costa (with a large number of 559 or 674) provides greater phylogenetic resolu-
guide cells) has occurred more than once and the tion (24 to three or four MP trees). There was little
development of leucocysts and chlorocysts has oc- difference between the trnL-rps4-rbcL and rps4-
curred multiple times within the Dicranidae. rbcL analyses, although the trnL-rps4-rbcL data set
Our analyses indicate that all three orders of the had less resolved relationships at deeper levels on
Dicranidae have taxa with differentiated alar cells. the tree (ordinal and familial), despite an increase
All members of the Dicranaceae and scattered taxa of 115 informative characters. The use of all three
in the Rhabdoweisiaceae (as defined here) and Blin- regions improves support for nine of the major
dia of the Seligeriaceae (Vitt 1976) have differen- clades (1–13), and three additional clades (Table 5).
tiated alar cells. As well, our phylogenetic recon- Although the maximum use of available characters
struction indicates that cladocarpy has developed in is advocated here, the different sets of characters
seven distinct clades, nested within predominantly show their respective utility for resolving different
acrocarpous lineages. This supports the previous taxa within the Dicranidae. The data sets are con-
suggestion by La Farge-England (1996) that cla- gruent in their identification and support for the 13
docarpy has evolved independently among multiple major clades, despite variation of informative or to-
isolated terminal groups within otherwise acrocar- tal number of characters. All three analyses dem-
pous lineages. Our data provide a working hypoth- onstrate the need for additional data to establish
esis for more detailed phylogenetic analyses to rig- robust familial and ordinal relationships within the
orously evaluate morphological and ecological evo- Dicranidae, because of poor support indices at
lution within the Dicranidae. those levels (Fig. 1). The trnL-rps4-rbcL analysis
Utility of chloroplast regions for the Dicrani- (maximum characters) provides the most robust
dae.—Although all of our DNA sequences have phylogeny of the 76 taxa.
been generated from the single chloroplast genome, For comparison, the study of the pleurocarps by
different regions have been phylogenetically infor- Buck et al. (2000) also used the rps4 and trnL re-
mative at various hierarchical levels. In the litera- gions for a similar number of taxa (86) with a sim-
ture, rbcL has been used to address higher level ilar percentage of informative characters (⫾32%).
relationships (class and orders) of bryophytes (e.g., In contrast, their results revealed topologies with
Goffinet et al. 1998; Lewis et al. 1997). Whereas, short internal branch lengths that lacked strong sta-
rps4 and/or trnL have been used primarily to ex- tistical support for familial or ordinal clades. A
amine the relationships of orders, families, and sub- comparison of our results with those of Buck et al.
families within arthrodontous mosses (e.g., Buck et (2000) emphasizes the variation in utility of partic-
al. 2000; Cox & Hedderson 1999; Stech 1999a). ular genomic regions for phylogeny reconstruction
Single gene phylogenies were initially the norm, in different groups (Soltis & Soltis 1998). Evident-
whereas multiple regions are the norm currently ly, families and orders of pleurocarps are very dif-
(see various papers of this symposium series). ferent in evolutionary age, or evolutionary rate (or
Our study explored the utility of three different both) compared to those traditionally recognized in
combinations of genomic regions 1) a variable, haplolepideous mosses. Therefore our data indicate
non-coding region with a coding region (trnL- the utility of selected chloroplast regions for a
rps4); 2) two coding regions (rbcL-rps4); and 3) all broad selection of taxa in the Dicranidae.
three regions combined. The individual regions in-
dicated that for our data set, rps4 evolves 1.4 times CONCLUSIONS
faster than trnL region and 1.6 times faster than the
rbcL region. However, it should be noted that most We identify 13 major clades within the Dicrani-
of the potential variation was excluded from the dae, with four species forming distinct basal line-
trnL region due to ambiguous alignment over the ages. The strongest estimate of phylogenetic rela-
broad taxon sampling. With the restriction of the tionships for the taxa sampled was from the use of
274 THE BRYOLOGIST [VOL. 103

the maximum number of informative characters. phyly of groups that had limited sample size and
The data support the following relationships and an increase of parsimony-informative characters
inferences 1) the Dicranidae (i.e., haplolepideous should include data from other genomic regions. It
mosses as defined by peristome structure) form a is especially important to expand phylogenetic
monophyletic taxon; 2) the Dicranales, as tradition- analyses to include sequences from the nuclear ge-
ally circumscribed, are polyphyletic; 3) the Calym- nome, since chloroplast genealogy may differ from
peraceae and Octoblepharaceae form a monophy- the organismal phylogeny because of reticulation
letic group and are more closely related to members and lineage sorting. It will be interesting to com-
of the Dicranales than Pottiales; 4) the Dicranaceae pare the chloroplast phylogeny with patterns of re-
are polyphyletic; 5) the Ditrichaceae and the Rhab- lationships based on nuclear loci. As in numerous
doweisiaceae share a more recent common ancestor phylogenetic studies, our most ambiguous results
with members of the Pottiaceae than with other are concerned with relatively deep relationships
genera of the Dicranales; 6) the Fissidentaceae are among the major clades. Adding characters is prob-
nested within the Dicranales and should not be rec- ably the only way to resolve deep relationships
ognized at the ordinal level; 7) the Leucobryaceae (e.g., Graham & Olmstead 1999; Newton et al.
should be expanded to include Brothera (Paraleu- 2000).
cobryoideae, Dicranaceae), Dicranodontium, and In addition to such taxonomically broad ap-
Campylopus (Campylopodioideae, Dicranaceae); 8) proaches, ongoing research in our respective labo-
Rhachitheciaceae and Erpodiaceae form a clade ratories, focuses on genera and species of Dicran-
that is nested within the Dicranidae, probably sister aceae, Rhabdoweisiaceae, and Calymperaceae. The
to the Pottiaceae; 9) Amphidium forms an distinct position of the Rhachitheciaceae and Erpodiaceae
clade sister to the pottiaceous clade; 10) the Seli- also warrants further study. The hypothesis that the
geriaceae, represented by Blindia acuta, appears to Grimmiales include the Seligeriaceae requires in-
have a sister relationship with the Grimmiales; and creased taxon sampling to test its robustness. Phy-
11) four taxa form the basal lineages within the logenetic hypotheses provided by the current study
Dicranidae: Bryoxiphium norvegicum, Scoularia will facilitate the choice of out-group taxa for more
aquatica, Ditrichum flexicaule, and Timmiella cras- intensive studies of particular families and genera.
sinervis. Furthermore, the Dicranidae are morphologically
These data imply that traditional generic circum- and ecologically diverse, and future studies should
scriptions of haplolepideous families warrant the address ecological and evolutionary questions in a
purging or adding of selected taxa (cf. Vitt 1984). phylogenetic framework. Our study provides a
Ordinal circumscription and relationships among working hypothesis for such research.
the families within the subclass are considered pro-
visional, based on the lack of robust support for the ACKNOWLEDGMENTS
internal nodes. A summary of the inferred provi-
We acknowledge the support from the Mellon Foun-
sional classification of the Dicranidae is shown in dation Fellowship Award at Duke University; the Green
Table 1. Bryoxiphium and Scouleria are retained in Plant Phylogeny Coordination Research Group, USDA
distinct families. The Grimmiales include the Seli- grant 94-37105-0713 (DOE/NSF/USDA Panel on Collab-
geriaceae, Grimmiaceae, and Ptychomitriaceae. The orative Research in Plant Biology); Co-PIs Mark A. Buch-
heim, Brent D. Mishler, Russell L. Chapman; National
Dicranales are modified to consist of 11 families. Science Foundation grant DEB-9712347 (PEET: Partner-
The Pottiales include Amphidum, the Rhabdowei- ships for Enhancing Expertise in Taxonomy; Monographic
siaceae, Ditrichaceae, Rhachitheciaceae, and Erpo- research in the moss family Calymperaceae), PI: Brent D.
diaceae. The Calymperaceae are moved to the Di- Mishler; and Natural Sciences and Engineering Research
Council of Canada grant A-6390 (to Dale H. Vitt), Uni-
cranales, which also include the Octoblepharaceae
versity of Alberta. We thank Sandra Boles (Duke Univer-
and Fissidentaceae. The Dicnemonaceae are syn- sity) for generating a set of rps4 sequences, Patricia San-
onymized with the Dicranaceae. Seven families, the chez (University of California, Berkeley) for trnL se-
Archidiaceae, Drummondiaceae, Schistostegiaceae, quences, and Terry O’Brien (University of California,
Serpotortellaceae, Ephemeraceae, Splachnobry- Berkeley) for verification of voucher identifications in the
Mishler Lab. We thank Sean W. Graham and an anony-
aceae, and Wardiaceae were not included in this mous reviewer for useful critiques of the manuscript.
study and are thus given an uncertain placement
within the Dicranidae. The familial position of Am- LITERATURE CITED
phidium and Dicranella is also not determined from
this data set. ALLEN, B. H. 1987. A revision of the Dicnemonaceae.
In order to address the ambiguity of ordinal and Journal of the Hattori Botanical Laboratory 62: 1–100.
. 1994. Moss Flora of Central America. Part 1.
familial relationships within the Dicranidae, future Sphagnaceae-Calymperaceae. Monographs in System-
research should concentrate on two approaches. In- atic Botany, Missouri Botanical Garden 49.
creased taxon sampling is needed to assess mono- ANDERSON, L. E. & H. CRUM. 1859. Cytotaxonomic stud-
2000] LA FARGE ET AL.: HAPLOLEPIDEOUS MOSSES 275

ies on mosses of the Canadian Rocky Mountains. Bul- patibility methods will be positively misleading. Sys-
letin of the National Museum of Canada 160: 1–89. tematic Zoology 27: 401–410.
BECKERT, S., H. MUHLE, D. PRUCHNER & V. KNOOP. 1999. . 1985. Confidence limits on phylogenies: an ap-
Phylogenies from mitochondrial DNA in bryophytes proach using the bootstrap. Evolution 39: 783–791.
and cryptogams, poster 1073. XVI International Bo- FRAHM, J. P. 1991a. Dicranaceae: Campylopodioideae,
tanical Congress, St. Louis, U.S.A. August 1–7. Paraleucobryoideae. Flora Neotropica Monograph 54:
BLOMQUIST, H. L. & L. L. ROBERTSON. 1941. The devel- 1–238.
opment of the peristome in Aulacomnium heterosti- . 1991b. A phenetic and cladistic study of the
chum. Bulletin of the Torrey Botanical Club 68: 569– Campylopodioideae. Journal Hattori Botanical Labo-
584. ratory 69: 65–78.
BREMER, K. 1994. Branch support and tree stability. Cla- GOFFINET, B. 1998. The Rhachitheciaceae: revised generic
distics 10: 295–304. circumscription and ordinal affinities. THE BRYOLOGIST
BROTHERUS, V. F. 1924. Musci (Laubmoose) III. Unterk- 100: 425–439.
lasse Bryales. Spezieller Teil.10 Band, 1 Halfte, pp. , R. J. BAYER & D. H. VITT. 1998. Circumscription
143–314. In A. Engler & K. Prantl (eds.), Die Natür- of the Orthotrichales (Bryopsida) inferred from rbcL
lichen Pflanzenfamilien, ed. 2. Engelmann, Leipzig. sequence analyses. American Journal of Botany 85:
BRUGGEMAN-NANNENGA, M. A. & W. BERENDSEN. 1990. 1324–1337.
On the peristome types found in the Fissidentaceae & C. J. COX. 2000. Phylogenetic relationships
and their importance for the classifications. Journal of among basal-most arthrodontous mosses with special
the Hattori Botanical Laboratory 68: 193–234. emphasis on evolutionary significance of the Funari-
BUCK, W. R., B. GOFFINET, & A. J. SHAW. 2000. Testing neqe. THE BRYOLOGIST 103: 212–223.
morphological concepts of orders of pleurocarpous GRAHAM, S. W. & R. G. OLMSTEAD. 1999. Abstract 7.4.5.
mosses (Bryophyta) using phylogenetic reconstruc- A phylogeny of basal angiosperms inferred from 17
tions based on trnL-trnF and rps4 sequences. Molec- chloroplast genes. XVI International Botanical Con-
ular Phylogenetics and Evolution 10: (in press). gress, St. Louis, U.S.A. August 1–7.
CHURCHILL, S. 1981. A phylogenetic analysis, classifica- GRAYBEAL, A. 1998. Is it better to add taxa or characters
tion and synopsis of the genera of the Grimmiaceae to a difficult phylogenetic problem? Systematic Biol-
(Musci), pp. 127–144. In V. A. Funk & D. R. Brooks ogy 47: 9–17.
(eds.), Advances in Cladistics. Proceedings of the First HEDDERSON, T., C. J. COX & J. G. GIBBINGS. 1999. Phy-
Meeting of the Willi Hennig Society. The New York logenetic relationships of the Wardiaceae (Musci); ev-
Botanical Garden. idence from 18S rRNA and rps4 gene sequences. THE
COX, C. J. & T. J. HEDDERSON. 1999. Phylogenetic rela- BRYOLOGIST 102: 26–31.
HENDY, M. D. & D. PENNY. 1989. A framework for the
tionships among the ciliate arthrodontous mosses: ev-
quantitative study of evolutionary trees. Systematic
idence from chloroplast and nuclear DNA sequences.
Zoology 38: 297–309.
Plant Systematics and Evolution 215: 119–139.
LA FARGE-ENGLAND, C. 1996. Growth form, branching
, B. GOFFINET, A. E. NEWTON, A. J. SHAW & T. A.
pattern, and perichaetial position in mosses: Cladocar-
HEDDERSON. 2000. Phylogenetic relationships among
py and pleurocarpy redefined. THE BRYOLOGIST 99:
the diplolepideous-alternate mosses (Bryidae) inferred
170–186.
from nuclear and chloroplast DNA sequences. THE
. 1998. The infrageneric phylogeny, classification
BRYOLOGIST 103: 224–241.
and phytogeography of Leucoloma Brid. (Dicranaceae,
CROSBY, M. R. & R. E. MAGILL. 1981. A Dictionary of
Bryopsida). THE BRYOLOGIST 10: 181–220.
Mosses. Monographs in Systematic Botany from the
LAWTON, E. 1971. Moss Flora of the Pacific Northwest.
Missouri Botanical Garden 3: 1–43.
Hattori Botanical Laboratory, Nichinan, Japan.
CRUM, H. 1994. Bryoxiphiales, p. 82. In A. J. Sharp, H. LEWINSKY, J. 1976. On the systematic position of Amphi-
A. Crum & P. M. Eckel (eds.), The Moss Flora of dium Schimp. Lindbergia 3: 227–231.
Mexico. Memoirs of the New York Botanical Garden LEWIS, L. A., B. D. MISHLER & R. VILGALYS. 1997. Phy-
69. logenetic relationships of the liverworts (Hepaticae), a
& L. E. ANDERSON. 1981. Mosses of Eastern basal embryophyte lineage, inferred from nucleotide
North America, Vol. 1. Columbia University Press, sequence data of the chloroplast gene rbcL. Molecular
NY. Phylogenetics and Evolution 7: 377–393.
DIXON, H. N. 1932. Classification of Mosses, pp. 397– LIMPRICHT, K. G. 1885–1895. Die laubmoose Deutsch-
412. In F. Verdoorn (ed.), Manual of Bryology. Mar- lands, Oesterreichs und der Schweiz. In Dr. l. Raben-
tius Nijhoff, The Hague. horst’s Kryptogamen Floravon Deutschland, Oester-
DOYLE, J. J. & J. L. DOYLE. 1990. Isolation of plant DNA reich und der Schweiz, 2nd ed. Eduard Kummer, Leip-
from fresh tissue. Focus 12: 13–15. zig.
EDDY, A. 1988. Sphagnales to Dicranales. A Handbook of LOESKE, L. 1910. Studien zur vergleichenden Morphologie
Malesian Mosses, Vol. 1. Natural History Museum und phylogenetischen Systematik der Laubmoose. M.
Publications, London. Lande, Berlin.
. 1990. Leucobryaceae to Buxbaumiaceae. A MADDISON,W. P. & D. R. MADDISON. 1992. MacClade ver-
Handbook of Malesian Mosses, Vol. 2. Natural His- sion 3. Analysis of phylogeny and character evolution.
tory Museum Publications, London. Sinauer Associates Inc, Sunderland, MA.
EDWARDS, S. 1979. Taxonomic implications of cell pat- MALTA, M. 1926. Die Gattung Zygodon Hook. et Tayl.
terns in haplolepidous moss peristomes, pp. 317–346. Eine monographische Studie. Acta Hatori Botanici
In G. C. S. Clarke & J. G. Duckett (eds.), Bryophyte Universitatis Latviensis 1: 1–184.
Systematics, Academic Press, London. MAGILL, R. E. 1981. Bryophyta, Part 1. Mosses. Fasc. 1
ERIKSSON, T. 1998. AutoDecay ver.4.0. Department of Sphagnaceae-Grimmiaceae, pp. 1–291. In O. A. Leist-
Botany Stockholm University, Stockholm. ner (ed.), Flora of Southern Africa, Department of Ag-
FELSENSTEIN, J. 1978. Cases in which parsimony and com- riculture and Fisheries, Pretoria.
276 THE BRYOLOGIST [VOL. 103

MATSUI, T. & Z. IWATSUKI. 1990. A taxonomic revision of the plastid gene rps4. Plant Systematics and Evolution
the family Ditrichaceae (Musci) of Japan, Korea, and 204: 109–123.
Taiwan. Journal of the Hattori Botanical Laboratory STECH, M. 1999a. A reclassification of the Dicranaceae
68: 317–366. (Bryopsida) based on non-coding cpDNA sequence
NEWTON, A. E., C. J. COX, J. G. DUCKETT, J. A. WHEELER, data. Journal of the Hattori Botanical Laboratory 86:
B. GOFFINET & B. D. MISHLER. 2000. Evolution of the 137–159.
major moss lineages: phylogenetic analyses based on . 1999b. A molecular systematic contribution to
multiple gene sequences and morphology. THE BRY- the position of Amphidium Schimp. (Rhabdoweisi-
aceae, Bryopsida). Nova Hedwigia 68: 291–300.
OLOGIST 103: 187–211.
SWOFFORD, D. L. 1998. PAUP* 4.0b2–Phylogenetic Anal-
PETERSON, W. 1994. Leucobryaceae, pp. 169–171. In A.
ysis Using Parsimony (* and other methods). Version
J. Sharp, H. A. Crum & P. M. Eckel (eds.), The Moss 4. Sinauer Associates, Sunderland, MA.
Flora of Mexico. Memoirs of the New York Botanical TABERLET, P. L. GIELLY, G. PAUTOU & J. BOUVET. 1991.
Garden 69. Universal primers for amplification of three non-cod-
RAMBAUT, A. 1995. Sequence Alignment Program (Se–Al) ing regions of chloroplast DNA. Plant Molecular Bi-
v1.d1. Department of Zoology, University of Oxford, ology 17: 1105–1109.
Oxford. TAYLOR, E. C. 1962. The Philibert peristome articles. An
& I. G. STONE. 1995. The Calymperaceae of Aus- abridged translation. THE BRYOLOGIST 65: 175–212.
tralia. Journal of the Hattori Botanical Laboratory 78: VITT, D. H. 1973. A revision of the genus Orthotrichum
1–40 in North America, north of Mexico. Bryophytorum
& R. H. ZANDER. 1987. Serpotortellaceae, a new Bibliotheca 1: 1–208.
family of mosses (Musci) from Madagascar and Ré- . 1976. The genus Seligeria in North America.
union. THE BRYOLOGIST 90: 230–236. Lindbergia 3: 241–275.
REESE, W. D., T. KOPONEN, & D. H. NORRIS. 1986. Bryo- . 1984. The classification of the Bryopsida, pp.
phyte flora of the Huon Peninsula, Papua New Guinea. 676–759. In R. M. Schuster (ed.), New Manual of
XIX. Calymperes, Syrrhopodon, and Mytthyridium Bryology. The Hattori Botanical Laboratory, Vol. 2,
Nichinan, Japan.
(Calymperaceae, Musci). Acta Botanica Fennica 133:
, B. GOFFINET & T. HEDDERSON. 1998. The ordinal
151–202. classification of the mosses. Questions and answers for
SHARP, A. J., H. CRUM & P. M. ECKEL (EDS.). 1994. the 1990s, pp. 113–123. In J. W. Bates, N. W. Aston
Sphagnales to Bryales, pp. 31–417. The Moss Flora & J. G. Duckett (eds.), Bryology for the Twenty-first
of Mexico. Memoirs of the New York Botanical Gar- Century. Maney and the British Bryological Society,
den 69. Leeds, U.K.
SHAW, A. J., B. D. MISHLER & L. E. ANDERSON. 1989. WALTHER, K. 1983. Bryophytina. Laubmoose. In J. Ger-
Peristome development in mosses in relation to sys- loff & J. Poelt (eds.), A. Engler’s Syllabus der Pflan-
tematics and evolution. IV Haplolepideae: Ditricha- zenfamilien. Aufl.13.V 1–108. Gebrüder Bornträger,
ceae and Dicranaceae. THE BRYOLOGIST 92: 314–325. Berlin.
SMITH, A. J. E. 1978. Moss Flora of Britain and Ireland, WHEELER, J. A., D. P. WALL, K. JOHANNES & B. D. MISH-
LER. 2000. Congruence and convergence in the moss
Cambridge University Press, Cambridge.
family Calymperaceae: phylogenetic analysis of two
SOLTIS, D. E. & P. E. SOLTIS. 1998. Choosing an approach
chloroplast genes (rbcL and rps4) and morphology.
and an appropriate gene for phylogenetic analysis, Systematic Biology (in press).
pp.1–42. In D. E. Soltis, P. E. Soltis, & J. J. Doyle ZANDER, R. H. 1993. Genera of the Pottiaceae: mosses of
(eds.), Molecular Systematics of Plants II DNA Se- harsh environments. Bulletin of the Buffalo Society of
quencing. Kluwer, Boston. Natural Sciences 32: 1–378.
SOUZA-CHIES, T. T., G. BITTAR, S. NADOT, L. CARTER, E.
BESIN & B. LEJEUNE. 1997. Phylogenetic analysis of
Iridaceae with parsimony and distance method using ms. received Dec. 8, 1999; accepted Feb. 10, 2000.