Phylogeny and Diversification of Bryophytes

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/51200883
Phylogeny and diversiﬁcation of bryophytes
Article in American Journal of Botany · October 2004

DOI: 10.3732/ajb.91.10.1557 · Source: PubMed
CITATIONS READS
151 1,746
2 authors:
Arthur Jonathan Shaw Karen S Renzaglia

Duke University Southern Illinois University Carbondale
262 PUBLICATIONS 6,149 CITATIONS 139 PUBLICATIONS 3,196 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Induction of stress response and toxicity threshold of Azolla caroliniana in response to AgNP View project
Ultrastructure and pectin composition of guard cell walls View project
All content following this page was uploaded by Karen S Renzaglia on 18 December 2013.
The user has requested enhancement of the downloaded file.

American Journal of Botany 91(10): 1557–1581. 2004.
PHYLOGENY AND DIVERSIFICATION OF BRYOPHYTES1
JONATHAN SHAW2,4 AND KAREN RENZAGLIA3

2
Duke University, Department of Biology, Box 90338, Durham, North Carolina 27708 USA; and 3Department of Plant Biology,
Southern Illinois University, Carbondale, Illinois 62901-6509 USA
The bryophytes comprise three phyla of embryophytes that are well established to occupy the first nodes among extant lineages in
the land-plant tree of life. The three bryophyte groups (hornworts, liverworts, mosses) may not form a monophyletic clade, but they
share life history features including dominant free-living gametophytes and matrotrophic monosporangiate sporophytes. Because of
their unique vegetative and reproductive innovations and their critical position in embryophyte phylogeny, studies of bryophytes are
crucial to understanding the evolution of land plant morphology and genomes. This review focuses on phylogenetic relationships within
each of the three divisions of bryophytes and relates morphological diversity to new insights about those relationships. Most previous
work has been on the mosses, but progress on understanding the phylogeny of hornworts and liverworts is advancing at a rapid pace.
Multilocus multigenome studies have been successful at resolving deep relationships within the mosses and liverworts, whereas single-
gene analyses have advanced understanding of hornwort evolution.
Key words: Anthocerophyta; Bryophyta; bryophyte phylogeny; hornworts; liverworts; Marchantiophyta; mosses; tree of life.
As the only land plants with a dominant gametophyte gen- rophyte: sporangium, seta, and foot. In contrast, an apical mer-
eration, liverworts, mosses, and hornworts exhibit structural istem initial develops in the embryo of tracheophytes and is
and reproductive attributes that are exclusive, unifying, and subsequently responsible for continuous production of repeat-
innovative. Their persistent gametophyte is responsible for ex- ed shoot and root modules in these plants (Bierhorst, 1971;
ploratory growth as well as for proliferation of a new gener- Kato and Imaichi, 1997). Capsules of bryophytes are structur-
ation through either sexual or asexual processes. As a conse- ally elaborate and, in some instances, exhibit complicated
quence, bryophyte gametophytes exhibit a degree of diversity mechanisms for spore production and dispersal. Basal sporo-
and complexity unparalleled in tracheophytes. They are char- phyte elongation with nonsynchronized spore production in
acterized by modular growth (repeated patterns) from a gen- hornworts, elaters in liverworts, and peristomes of mosses pro-
erative apex, range in habit from upright to procumbent, and vide examples of this complexity.
include thalloid to leafy forms (Mishler and DeLuna, 1991). General treatments of bryophyte morphology can be found
Within mosses and liverworts, leafy gametophytes are the in Leitgeb (1874–1881), Campbell (1895), Goebel (1905),
norm, rivaling the leafy sporophytic growth forms of some Smith (1955), Parihar (1965), Watson (1971), Puri (1973),
tracheophytes, especially lycophytes (Renzaglia et al., 2000). Richardson (1981), Schofield (1985), and Crum (2001). The
However, because they depend on water for sexual reproduc- Manual of Bryology, edited by Verdoorn (1932), contains au-
tion, the gametophytes of bryophytes are small relative to most thoritative treatments of selected bryology topics that sum-
vascular plant sporophytes. Sexual reproduction in bryophytes marized the state of our knowledge at that time, and the New
involves release of motile male gametes into the environment Manual of Bryology, edited by Schuster (1984), provided ex-
and requires successful navigation of these naked cells from panded updates more than fifty years later. Both manuals are
the male to the female sex organs via an external water source. still useful. Other edited volumes on various aspects of bryo-
Sporophytes of bryophytes are without exception monospo- phyte biology especially relevant to the tree of life include
rangiate and matrotrophic throughout their life span (Graham Clarke and Duckett (1979), Smith (1982), and Shaw and Gof-
and Wilcox, 2000). Ephemeral and dependent on the game- finet (2000).
tophyte for nutrition and protection, they never exhibit the
modular, indeterminate growth form of the gametophyte gen- The crucial position of bryophytes in embryophyte evolu-
eration. In their greatest structural complexity, bryophyte spo- tion—An unambiguous conclusion from the multitude of con-
rophytes consist of a nutritive foot, elongating pedicel or seta, temporary phylogenetic investigations of streptophytes is that
and a single terminal sporangium or capsule. Formative divi- bryophytes are the first green plants to successfully radiate into
sions in the embryo produce all precursor components of the terrestrial niches. These small, inconspicuous plants have ex-
sporophyte; i.e., distinct embryonic regions are determined to isted for several hundreds of millions of years and have played
develop into the three organographic zones of the mature spo- a prominent role in shaping atmospheric and edaphic change
and the subsequent evolution of all forms of plant life on land.
1
Manuscript received 5 January 2004; revision accepted 15 June 2004.
Explorations of life history phenomena in bryophytes and a
Various aspects of the research presented here were supported by NSF solid understanding of interrelationships among them are nec-
grants from the Systematic Biology and Assembling the Tree-of-Life Pro- essary to reconstruct the early evolution of embryophytes.
grams (DEB-0089131 to AJS and DEB-0235985 and DEB-0228679 to KSR). The concept that the embryo/sporophyte evolved in land
We are grateful to Christine Cargill, Cymon Cox, Barbara Crandall-Stotler, plants through intercalation of mitotic divisions between fer-
Christine Davis, R. Joel Duff, Laura Forrest, Bernard Goffinet, Xiaolan He- tilization and meiosis is widely accepted (Graham, 1993; Gra-
Nygrén, and Juan Carlos Villarreal for providing data and sharing unpublished
manuscripts. We are also indebted to Andrew Blackwell and Scott Schuette
ham and Wilcox, 2000). Based on this axiom, land plant evo-
for contributions of images and expert technical assistance. Molly McMullen lution proceeded in the direction of progressively more elab-
kindly edited an earlier manuscript draft. orate sporophytes. Although generally true, unconditional ac-
4
E-mail: [email protected]. ceptance of this trend leads to conclusions that ignore
1557
1558 AMERICAN JOURNAL OF BOTANY [Vol. 91
processes of reduction and parallel/convergent evolution, phe- development are underway, and, in the context of phylogenetic
nomena that have occurred repeatedly during bryophyte di- analyses of bryophytes, the early evolution of these genes is
versification (Schuster, 1992; Niklas, 1997; Boisselier-Dubayle now a tractable problem for investigation (Himi et al., 2001).
et al., 2002). A defining characteristic of embryophytes is the
meiotic production of spores in tetrads and sporopollenin-im- Relationships among the three lineages—Relationships
pregnated spore walls. Because of their resistance to degra- among the three lineages of bryophytes remain one of the
dation, fossil spores have provided valuable clues to the initial major unresolved questions in plant evolutionary biology
stages of land colonization (Taylor, 1995; Wellman and Gray, (Goffinet, 2000). Virtually every conceivable hypothesis has
2000; Wellman et al., 2003). The earliest confirmed land plant been put forth in regards to primary branching patterns at the
fossils are spores, speculated to be from an ancient liverwort base of embryophytes. Most commonly, bryophytes are
dating to the middle Ordovician, some 475 million years ago viewed as a grade of three monophyletic lineages, with an
(mya) (Wellman et al., 2003). uncertain branching order (Mishler et al., 1994; Qiu et al.,
Gametophytes of bryophytes also provide critical clues 1998). Controversy often focuses on which bryophyte group
about land plant evolution. Thalloid and filamentous growth is sister to all other embryophytes, with two hypotheses most
forms are shared with pteridophytes, but the completely sub- frequently supported: liverworts as sister to other embryo-
terranean and nonphotosynthetic life histories found in many phytes vs. hornworts as the sister group (Mishler et al., 1994;
lycophytes and some ferns show no homology in bryophytes Hedderson et al., 1996, 1998; Malek et al., 1996; Garbary and
(Bierhorst, 1971). The achlorophyllous gametophyte of the liv- Renzaglia, 1998; Qiu et al., 1998; Beckert et al., 1999; Duff
erwort Cryptothallus is a recent acquisition within a strictly and Nickrent, 1999; Nishiyama and Kato, 1999; Soltis et al.,
photosynthetic lineage (Renzaglia, 1982). Unlike pterido- 1999; Nickrent et al., 2000; Renzaglia et al., 2000; Stech et
phytes, bryophyte gametophytes frequently show organ de- al., 2003). A moss-plus-liverwort clade has been recovered in
velopment (leaf, stem, and rhizome) and extensive tissue dif- several of these analyses (Hedderson et al., 1996, 1998; Ni-
ferentiation, including conducting and supportive tissues (Hé- shiyama and Kato, 1999; Nickrent et al., 2000; Renzaglia et
bant, 1977; Ligrone et al., 2000). Production of multicellular al., 2000). Recently, it was postulated that hornworts, not
gametangia was an innovation in embryophytes that was a mosses, are the closest living relative of tracheophytes. This
necessary precursor to embryo development (Graham and Wil- speculation finds support in sequence data as well as in struc-
cox, 2000). Among land plants, only mosses and liverworts tural genomic features (Samigullin et al., 2002; Kelch et al.,
produce superficial gametangia, which are variously protected in press). In contrast, recent analyses of amino acid sequences
by elaborate appendages, including leaves. Hornworts seques- based on entire plastid genomes provided support for a mono-
ter vulnerable organs in internal compartments (Renzaglia et phyletic bryophyte assemblage; however, these results must be
al., 2000; Renzaglia and Vaughn, 2000). viewed with caution because of severe limitations in taxon
A lack of intermediate forms in both life history phases and sampling (Nishiyama et al., in press).
the potential to interpret morphological transitions in opposite The focus of this review is to present the current state of
directions have obscured relationships among bryophytes and knowledge on phylogenetic relationships within, not among,
pteridophytes. Understanding morphological evolution re- hornworts, liverworts, and mosses. Emphasis is placed on syn-
quires unambiguous establishment of phylogenetic relation- thesizing results of recent molecular investigations that have
ships among and within bryophyte lineages. Over the past de- revolutionized interpretations of genetic and morphological di-
cade, great strides have been made toward reaching this goal; versification within each of these groups. Intriguing new per-
however, fundamental questions remain. spectives on character evolution have emerged from these
In addition to elucidating early patterns of morphological studies.
diversification in embryophytes, bryophytes are crucial to un-
derstanding plant genome evolution. Approximately 66% of ANTHOCEROTOPHYTA
genes identified from expressed sequence tag analyses of gene
expression in gametophytes of Physcomitrella patens have ho- Hornwort classification and relationships—For centuries,
mologues in the Arabidopsis genome, consistent with the hy- botanists have marveled at the structural peculiarities of horn-
pothesis that genes expressed in the diploid plant body of an- worts (Hofmeister, 1851; Leitgeb, 1879; Campbell, 1895,
giosperms were expressed in the gametophytes of early land 1917, 1924; Goebel, 1905; Lang, 1907; Bower, 1935). In no
plants and were recruited for sporophytic morphogenesis later other branch of the green tree of life does extension of each
in plant phylogeny (Nishiyama et al., 2003). Phylogenetic and sporophyte involve continuous, presumably indeterminate, ba-
functional analyses of genes expressed in Physcomitrella ga- sipetal growth of a single elongated sporangium. All stages of
metophytes have clarified the phylogenetic history of several spore development, from undifferentiated cells through pre-
important gene families, including MIKC-type MADS-box meiotic/meiotic spore mother cells to sequentially more mature
genes (Krogan and Ashton, 2000; Henschel et al., 2002; Hohe spores, can be found in a single hornwort sporangium. A con-
et al., 2002) and homeobox genes (Champagne and Ashton, stant production of spores therefore ensures dispersal through-
2001). Phylogenetic analyses of the KNOX (homeobox) gene out the growing season for as long as the gametophyte persists.
family across the land plant tree of life have provided insights This mode of sporophyte development has no counterpart in
into the history of gene duplication and functional divergence other plant groups, thus obscuring the phylogenetic position
during embryophyte history (Champagne and Ashton, 2001). of hornworts among green plants.
Because KNOX genes are involved in expression of meriste- Hornworts have remained relatively unexplored at all levels
matic activity in vascular plant sporophytes, functional anal- of phylogenetic inquiry (Renzaglia and Vaughn, 2000; Stech
yses of KNOX genes in mosses, liverworts, and hornworts are et al., 2003; Duff et al., in press). The perception that horn-
central to understanding evolution of plant development in em- worts are invariable, elusive, and difficult to identify has con-
bryophytes. Comparable studies of genes involved in flower tributed to the paucity of systematic studies within the group.
October 2004] SHAW AND RENZAGLIA—PHYLOGENY OF BRYOPHYTES 1559
Even with the advent of molecular systematics and a re-

newed interest in early land plant phylogeny, hornwort sam-
pling has been sparse, with one to three taxa included in most
analyses (Katoh et al., 1983; van de Peer et al., 1990; Mishler
et al., 1994; Bopp and Capesius, 1996, 1998; Hedderson et
al., 1996, 1998; Malek et al., 1996; Qiu et al., 1998; Beckert
et al., 1999; Duff and Nickrent, 1999; Nishiyama and Kato,
1999; Soltis et al., 1999; Nickrent et al., 2000). Among the
dozens of papers on bryophyte phylogeny over the past ten
years, there is only one comprehensive molecular analysis of
within-hornwort relationships, based on rbcL gene sequences
from 20 hornworts (Duff et al., in press). A second study uti-
lizing the plastid trnL intron sampled nine hornworts but fo-
cused on the position of the group among land plants (Stech
et al., 2003). Results of these analyses are congruent and re-
Fig. 1. ‘‘Backbone’’ phylogeny of the one most parsimonious tree based veal novel but intuitive relationships. The rbcL analysis pro-
on rbcL sequences from seven of the 11 named genera of hornworts (Antho- vided much greater resolution of hornwort interrelations be-
cerophyta). Bootstrap percentages are shown below branches. Tree provided
by R. Joel Duff, University of Akron.
cause of more extensive sampling, including additional species
of the five genera included in the trnL study and representa-
tives of three other genera (Folioceros, Leiosporoceros, and
This small, homogeneous assemblage contains 100–150 poor- Nothoceros). The discussion that follows will focus on taxo-
ly delineated species (Schuster, 1992). Concepts of interrela- nomic inferences and morphological character evolution that
tionships among hornworts based on morphology, and the re- emerge from scrutiny of the consensus phylogenetic pattern
sulting classification schemes, show virtually no consensus at supported by these pioneering studies (Fig. 1).
the generic, familial, and ordinal levels (Mishler and Churchill, Diagnostic characters of hornworts are found in both life
1984; Hasegawa, 1988; Hässel de Menéndez, 1988; Schuster, history generations and are variably emphasized by systema-
1992; Hyvönen and Piippo, 1993; Renzaglia and Vaughn, tists (Cargill et al., in press). Growth form (Fig. 2), chloroplast
2000). Twelve genera of hornworts have been named, Antho- structure and number (Fig. 3), antheridial number and jacket
ceros, Dendroceros, Folioceros, Notothylas, Megaceros, cell organization, Nostoc colony organization, and presence of
Phaeoceros, Aspiromitus, Hattorioceros, Leiosporoceros, mucilage canals and thallus outgrowths are taxonomically use-
Nothoceros, Mesoceros, and Sphaerosporoceros, of which ful gametophytic characters. Taxonomically informative fea-
only the first six are widely recognized. tures of the sporophyte include degree of development of his-
Fig. 2. Diversity in growth forms among hornworts. A. Photograph of Anthoceros punctatus L. Small orbicular gametophyte with both immature and almost
ripe sporophytes, growing on soil. Image provided by Christine Cargill. B and C Scanning electron micrographs (SEM) of gametophyte of Dendroceros crispatus
(Hook.) Nees. B. Ventral surface showing monostromatic wings and thickened midrib with bulging Nostoc colonies (arrow). Note the numerous small pores
(mucilage clefts) along either side of the midrib. C. Dorsal surface showing sunken archegonia (arrow) on the midrib and developing sporophytes enclosed
within gametophytic involucre. D. SEM of Notothylas orbicularis (Schwein.) Sull. Small orbicular gametophytes growing on bare soil; note the numerous small,
horizontally oriented sporophytes enclosed in involucres. Bar 5 0.2 mm, except in A, bar 5 3 mm.
internal mucilage canals. Number of antheridia per cavity is

greater than 20, a feature shared only with Anthoceros and
Folioceros (Cargill et al., in press). However, unlike other
hornworts in which Nostoc is in discrete spherical colonies
within the ventral thallus (Fig. 2B), those in Leiosporoceros
occur in branching strands that run longitudinally and are se-
questered in the thallus midregion. Ventral mucilage clefts that
enable Nostoc to enter and establish colonies in other taxa
(Fig. 6B) are lacking in Leiosporoceros. Chloroplasts of Leio-
sporoceros are also readily differentiated from those in other
hornworts. Starch is neatly aggregated around the periphery
of the organelle, with a central elaboration of photosynthetic
membranes; the chloroplasts have no pyrenoids (Fig. 3A).
The sporophyte of Leiosporoceros is elongated and robust,
and its anatomy departs significantly from that of other horn-
worts (Fig. 4A). The suture is highly differentiated and visible
as a deep longitudinal groove. The assimilative and sporoge-
nous regions are massive when compared with other horn-
worts. Several layers of small spore tetrads are surrounded by
mucilage and interspersed with groups of large, elongated
pseudoelaters (Fig. 5A). Stomata are abundant and appear sim-
ilar to those in more derived taxa (Fig. 6A). Clearly, the shared
traits between Leiosporoceros and other hornworts provide in-
sight about plesiomorphies within the group. For example, sto-
mata and large numbers of antheridia are best interpreted as
ancestral hornwort traits. On the other hand, unique morpho-
logical traits that characterize Leiosporoceros are presumed
autapomorphies and likely reflect the deep evolutionary sep-
aration of this genus from other hornworts. Further molecular
and morphological studies are required to evaluate these hy-
potheses.
After Leiosporoceros, Anthoceros plus Folioceros form a
Fig. 3. Transmission electron micrographs of chloroplasts in hornworts. clade sister to other hornworts (Fig. 1). Taxonomic treatments
A. Leiosporoceros dussii (Steph.) Hässel. Chloroplast in the assimilative layer have generally recognized a sister relationship between Antho-
of the sporophytes showing peripheral starch and centralized grana. B. Foli-
oceros fuciformis Baradw. Central pyrenoid with lens-shaped subunits sepa-
ceros (including Folioceros) and Phaeoceros, placing them in
rated by narrow grana and surrounded by starch grains. Bar 5 0.5 mm. the same family or subfamily. Thus genetic divergence be-
tween Anthoceros plus Folioceros and the remaining taxa ap-
pears problematic at first glance. Similarities between Antho-
togenic regions (Fig. 4), spore and pseudoelater architecture ceros, Folioceros, and Phaeoceros include rosette-like habits
and ultrastructure (Fig. 5), and the presence of columella and (Fig. 2A), large solitary chloroplasts with well-developed py-
stomata (Fig. 6). renoids (Fig. 3B), and comparable sporophyte anatomy (Fig.
4B). However, clearly defined features distinguish Anthoceros
The backbone of hornwort phylogenetic relationships— and Folioceros from other hornworts; these include dorsal la-
Molecular evidence corroborates morphological inferences mellae, schizogenous mucilage cavities, and antheridia in large
that the hornworts are monophyletic (Fig. 1). The genus Leios- groups of up to 50 per cavity, as compared to 1–4 (–6) an-
poroceros, which was named for its unusually small, mono- theridia per cavity in other hornworts. Darkly pigmented
lete, smooth spores produced in isobilateral tetrads (Figs. 4A, spores with well-defined trilete marks also serve to differen-
5A), is sister to the remaining hornworts. The position of tiate these taxa from other hornworts (Fig. 5B). Molecular ev-
Leiosporoceros among hornworts has been controversial. Häs- idence that Anthoceros plus Folioceros form a clade sister to
sel de Menéndez (1986, 1988) segregated it into an autono- all other hornworts (except Leiosporoceros) has reinforced the
mous family and order, whereas Hasegawa (1988) and Schus- taxonomic value of diagnostic morphological features that are
ter (1992) recognized it as a subgenus of Phaeoceros. In con- restricted to these two genera.
trast, Hyvönen and Piippo (1993) supported a sister relation- Close affinity between Folioceros and Anthoceros is well
ship between Leiosporoceros and Folioceros, based primarily supported by rbcL data and reflected in most current classifi-
on morphologically similar pseudoleaters (Fig. 5A). More de- cations (Hasegawa, 1988, 1994; Schuster, 1992; Hyvönen and
tailed examination of Leiosporoceros dussii reveals morpho- Piippo, 1993). In contrast, Hässel de Menéndez (1988) seg-
logical and molecular features heretofore undescribed in any regated Folioceros into a monotypic family and order based
hornwort. Moreover, Leiosporoceros gene sequences have ex- on spore ornamentation and pseudoelaters. Folioceros has
tremely low levels of RNA editing (J. Duff, University of Ak- thick-walled, reddish brown, highly elongated pseudoelaters,
ron, unpublished data) and thus differ from other hornworts whereas Anthoceros has short, thin-walled multicellular pseu-
that have been shown to have extensive editing (Yoshinga et doelaters similar to those of Phaeoceros (Fig. 5D). Additional
al., 1996; Duff et al., in press). The gametophyte resembles differences are found in the placenta and chloroplasts of these
that of Phaeoceros and Megaceros in that it is fleshy and lacks two taxa (Vaughn et al., 1992; Vaughn and Hasegawa, 1993).
Fig. 4. Cross sections of hornwort sporophytes. A. Light micrograph of Leiosporoceros dussii (Steph.) Hässel. Tissue is differentiated from outside to inside
as follows: single-layered epidermis, 9–10 layers of assimilative cells, abundant sporogenous tissue with several layers of tetrads intermixed with elaters, and
an indistinct columella. The suture is clearly defined as a longitudinal groove that extends nearly to the sporogenous tissue. Bar 5 100 mm. B. Scanning electron
micrograph of Phaeoceros carolinianus (Michx.) Prosk. In contrast to Fig. 4A, this sporophyte contains an assimilative zone of four cell layers, sporogenous
tissue with one layer of large tetrads intermixed with small elaters, and a columella of 16 cells. Bar 5 1 mm.
With only a single species included in the rbcL sequence anal- (Lange, 1907; Proskauer, 1960). An evaluation of this hypoth-
ysis, it is not possible to evaluate monophyly of Folioceros. esis requires increased taxon sampling across the hornworts.
The remaining hornworts form a monophyletic group that Diversity within Phaeoceros is particularly evident in spore
includes two well-supported assemblages: Phaeoceros laevis morphology (Schuster, 1992). Phaeoceros laevis s. l., includes
sensu lato (represented in Fig. 1 by P. carolinianus) plus No- species with spiny papillate spores, whereas ornamentation in
tothylas and Megaceros plus Dendroceros. A close affinity the remaining species varies from vermiculate to blunt, wart-
between Phaeoceros and Notothylas was suggested by Hässel like projections. As described later, the three representatives
de Menéndez (1988), who placed these two genera in the fam- of Phaeoceros with vermiculate spores included in molecular
ily Notothyladaceae. Both genera have chloroplasts with analyses are more closely related to Megaceros than to P. lae-
prominent pyrenoids, spores with an equatorial girdle (Fig. vis s. l.
5C), and 2–4 (–6) antheridia per chamber. However, because A close relationship between Megaceros and Dendroceros
of the distinctive sporophyte of Notothylas (Fig. 2D), most is evident in morphological characters such as spiraled pseu-
systematists have segregated this genus into a monotypic sub- doelaters (Fig. 5E, F), absence of stomata, and solitary an-
family, family, or order (Singh, 2002). Notothylas is the only theridia. The only epiphytic hornwort, Dendroceros, has a
hornwort taxon in which growth of the sporophyte is abbre- thickened, central midrib with perforated wings (Fig. 2B, C);
viated, spore production appears synchronized, stomata are ab- large, central pyrenoids in each plastid; and multicellular
sent, and the columella is normally absent to poorly developed, spores (Fig. 5E) (Hasegawa, 1980; Renzaglia and Vaughn,
a combination of characters that indicate affinities with liver- 2000). Diagnostic features of Megaceros include unicellular
worts. Consequently, it has been suggested that the Notothylas green spores with distal mammilla (Fig. 5F), the absence of a
sporophyte is plesiomorphic, representing a structural ‘‘link’’ pyrenoid, and multiple plastids per cell (Hasegawa, 1983; Val-
with other bryophytes. Under this interpretation, hornwort ra- entine et al., 1986; Vaughn et al., 1992). However, as discussed
diation involved an elaboration of sporophytes in more derived next, the demarcation between Megaceros and Dendroceros is
taxa (Campbell, 1895; Mishler and Churchill, 1984; Graham, not always well defined, especially with regard to growth form
1993; Hyvönen and Piippo, 1993; Hasegawa, 1994). An al- (Proskauer, 1953; Hässel de Menéndez, 1962).
ternative hypothesis, supported by molecular data, is that spo- A clade containing two species of Dendroceros is sister to
rophytes in Notothylas are not representative of the ancestral a monophyletic assemblage that includes species previously
condition in hornworts but are highly reduced and specialized placed in Megaceros, Phaeoceros, Nothoceros, and Dendro-
(Lang, 1907; Bartlett, 1928; Proskauer, 1960; Renzaglia, 1978; ceros (Fig. 1). This taxonomically heterogeneous group in turn
Schofield, 1985; Schuster, 1992). Features such as the exis- consists of two clades: the first includes two Old World species
tence of a relictual and largely nonfunctional suture in some of Megaceros, the Austral-Asian M. flagellaris and M. denti-
species support the derived nature of the Notothylas sporo- culatus (Hasegawa, 1983; Glenny, 1998), and the second is an
phyte. If parallel reduction in sporophye complexity occurred assemblage of species from four generic segregates. Three
among hornwort genera, Notothylas may be polyphyletic Phaeoceros species, P. coriaceus (Steph.) Campbell, P. hirti-
Fig. 6. Scanning electron micrographs of pores in hornworts. A. Stoma

in the sporophyte epidermis of Phaeoceros carolinianus (Michx.) Prosk. B.
Ventral mucilage cleft in the gametophyte epidermis of Dendroceros crispatus
(Hook.) Nees. Bars 5 20 mm.
ently inappropriate, placements in Dendroceros or the newly

delineated Nothoceros. The existence of unicellular, mamillate
spores, and plastids devoid of pyrenoids clearly place these
species in Megaceros. Thus, scrutiny of the morphology of
these seemingly disparate hornwort species reveals features
that solidify their inclusion in the Megaceros clade. The well-
developed costa and monostromatic wings in these taxa were
likely a result of parallel evolution with Dendroceros.
Fig. 5. Scanning electrom micrographs (SEM) of spores and pseudoelaters Inferences about morphological evolution in hornworts
of hornworts. A. Leiosporoceros dussii (Steph.) Hässel. Abundant elongated, from molecular analyses—One intriguing feature of horn-
slightly spiraled–thickened pseudoelaters with scattered small, smooth, iso-
bilateral spore tetrads (Sp). B. Anthoceros spores showing spinelike echinae
worts is the large, solitary chloroplast with a prominent py-
on the distal surface (left) and a well-developed triradiate ridge on the prox- renoid, which is shared with green algae but has no parallel
imal face. Image provided by Christine Cargill. C. Phaeoceros spores showing in any other embryophyte group. Within hornworts, pyrenoids
papillose distal surface (right) and proximal surface with an indistinct trilete appear to have been lost multiple times. Similar pyrenoid loss-
mark similar to the equatorial girdle (not visible). D. Short multicellular pseu- es (and gains) have been described in several algal lineages
doelaters of Phaeoceros. E. Dendroceros crispatus (Hook.) Nees. Multicel- (Hoham et al., 2002; Nozaki et al., 2002). In hornworts, chlo-
lular spores and spiraled pseudoelaters. F. Megaceros gracilis (Reich.) Steph.
Spiraled pseudoelater and spore with proximal surface facing downward and
roplast compartmentalization characterizes several taxa, in-
distal surface with mamilla upward. Bar 5 10 mm, except in A, bar 5 50 cluding Leiosporoceros and certain species of Phaeoceros, An-
mm. thoceros, and Megaceros (Burr, 1970; Valentine et al., 1986;
Vaughn et al., 1992; Duff et al., in press). This arrangement
is consistent with a carbon-concentrating mechanism typical
calyx Steph., and Phaeoceros chiloensis (Steph.) Hässel de of organisms with pyrenoids, including other hornworts (Smith
Menéndez, are included within the second Megaceros clade and Griffiths, 1996, 2000; Hansen et al., 2002). It has been
(Stech et al., 2003; Duff et al., in press). All these species speculated that the ‘‘pyrenoid-like’’ area evident in certain
have spores with markedly different architecture from those of hornworts represents a transitional state from presence to com-
other Phaeoceros (Fig. 5C). Moreover, these species have plete absence of the pyrenoid (Burr, 1970). An evolutionary
monoandrous androecia and multiple chloroplasts that lack py- inference supported by this interpretation, in addition to the
renoids, two diagnostic characters of Megaceros (Duff et al., phylogenetic topology presented in Fig. 1, is that a solitary
in press). However, lack of spiraled pseudoelaters, existence plastid with a pyrenoid is plesiomorphic in hornworts. In
of stomata, and yellow, not green, spores are defining features Leiosporoceros, the plastid is solitary but without a pyrenoid,
of Phaeoceros. Bartlett (1928) and Proskauer (1951) noted that the remnant of which is a compartmentalized organelle with
morphological boundaries between Megaceros and Phaeocer- peripherally aggregated starch and centralized grana and plas-
os are blurred and that similarities in growth form and chlo- toglobuli (Fig. 3A). Independent losses of the pyrenoid with
roplast structure suggest a close relationship between the two. or without organellar compartmentalization occurred at least
Molecular analyses have indicated that Phaeoceros with ver- once each in Megaceros, Notothylas, and Anthoceros (Vaughn
miculate spores and mamillae on the distal faces are more et al., 1992; Singh, 2002; Duff et al., in press). In Phaeoceros
closely related to Megaceros than to Phaeoceros with papillate hirticalyx, P. coriaceus, and P. chilioensis, species that are
spores (P. laevis s. l.) (Fig. 5B). probably better placed in Megaceros, loss of the pyrenoid may
The crown group of Megaceros consists of M. aenigmati- be interpreted as preceding the evolution of spiraled pseudoe-
cus, the only North American representative of the genus; M. laters and stomatal loss.
vincentianus, the only species from the Neotropics; Dendro- As structures that facilitate gas exchange, stomata are im-
ceros canaliculatus (5 M. canaliculatus), and Nothoceros gi- portant innovations in the diversification of land plants. Their
ganteus (5 M. giganteus). The last two species have thickened presence in hornworts has been viewed either as a synapo-
midribs and wings, which accounts for previous, and appar- morphy with mosses and tracheophytes or as a homoplastic
acquisition within hornworts (Mishler and Churchill, 1984;

Kenrick and Crane, 1997; Renzaglia et al., 2000). The pres-
ence of stomata in Leiosporoceros, Anthoceros, and Foliocer-
os supports the contention that these structures are plesiomor-
phic in hornworts and may be homologous to those in mosses
and/or tracheophytes. A clear case of homoplasy is the loss of
stomata in at least three, possibly four, hornwort lineages: No-
tothylas, Dendroceros, and Megaceros. Stomatal loss may
have accompanied modifications in sporophyte development,
e.g., maturation of the sporophyte within the protective ga-
metophytic involucre where gas exchange is limited (Notothy-
las, Fig. 2D and Dendroceros, Fig. 2C). Stomatal loss in Me-
gaceros is associated with occurrence of these species in pe- Fig. 7. ‘‘Backbone’’ tree showing phylogenetic relationships among the
riodically inundated habitats. The existence of P. coriaceus, major clades of liverworts, redrawn from Davis (in press). The topology is
P. hirticalyx, and P. chiloensis, three terrestrial species with from a maximum likelihood analysis of 12 nuclear, plastid, and mitochondrial
stomata, supports this hypothesis. The topology presented in genes. Broken-bold branch indicates uncertainty in the placement of Haplomi-
Fig. 1 necessitates at least two losses of stomata in the Me- trium.
gaceros clade.
The interpretation set forth by Proskauer (1951) and Schus-
ter (1992) that mucilage clefts on the ventral side of the ga- leafy forms are viewed as ancestral (see literature review in
metophyte in hornworts (Fig. 6B) are homologous to sporo- Crandall-Stotler and Stotler, 2000 and Davis, in press [Figs. 7,
phytic stomata (Fig. 6A) is not supported by molecular anal- 8]). Morphological studies supported the concept that simple
ysis. Absence of mucilage clefts in Leiosporoceros and the thalloid liverworts are more closely related to leafy types than
specialized function of these structures in all other hornworts to complex thalloids. Classification schemes reflect this inter-
indicate that gametophytic ‘‘stomata’’ evolved after hornwort pretation with hepatics typically divided into two groups: mar-
diversification simply as an entryway for the cyanobacterium, chantioid or complex thalloid liverworts (Marchantiopsida,
Nostoc. Marchantiidae) and jungermannioid liverworts, including the
It is reasonable to hypothesize that the habit of extant mem- leafy (Jungermanniopsida, Jungermanniidae) and simple thal-
bers of Anthoceros, Leiosporoceros, and Phaeoceros represent loid taxa (Jungermanniopsida, Metzgeriidae). Complex thal-
the ancestral condition in hornworts and may be related to loid types usually have air chambers with dorsal pores and
their common occurrence on exposed soil. Morphological di- differentiated internal tissues (Fig. 9A). Less commonly, the
versity in other taxa likely results from radiation into and con- thallus resembles the simple thalloid type in the lack of inter-
sequent adaptations to specialized habitats; e.g., Dendroceros nal or epidermal differentiation (e.g., Sphaerocarpos, Mono-
is an epiphyte, and Megaceros is restricted to tropical or tem- clea, and Dumortiera). Gametophytes of leafy liverworts
perate sites where it often occurs submerged in streams. Di- range from radially symmetrical with three rows of morpho-
versification of Dendroceros may be correlated with the evo- logically similar leaves (isophyllous) to dorsiventral with two
lution of angiosperms, which provided abundant new bark and rows of lateral leaves and an additional row of reduced (to
leaf habitats (Ahonen et al., 2003). Notothylas is an ephemeral absent) ventral underleaves or amphigastria (anisophyllous;
hornwort that grows as a pioneer on soil. Unlike other genera Fig. 9C). Simple thalloid (metzgerialean) organizations show
in which spores are wind dispersed, Notothylas spores are dis- less variability, from fleshy undifferentiated thalli to those with
persed by water or facultatively by insects or other animals, prominent midribs and monostromatic wings (Fig. 9B). Leaf-
thus eliminating the ‘‘need’’ for vertical elongation of the spo- like lobes or lobules in some taxa blur the distinction between
rophyte. leafy and simple thalloid forms. Internal differentiation of wa-
ter-conducting tissue is restricted to Haplomitrium and certain
MARCHANTIOPHYTA simple thalloid taxa, whereas conducting parenchyma is wide-
spread among both complex and simple thalloid forms, but not
Liverwort classification and relationships—The immense leafy taxa (Hébant, 1977; Kobiyama and Crandall-Stotler,
morphological diversity among the 377 genera and 6000–8000 1999; Ligrone et al., 2000).
species of liverworts has presented significant challenges to Liverworts are distinguished from hornworts and mosses by
systematists (Schljakov, 1972; Schuster, 1984; Crandall-Stotler the possession of oil bodies, unique organelles in which ter-
and Stotler, 2000). Within this monophyletic assemblage are penoids accumulate (Fig. 9D). All other embryophytes, in-
several morphologically isolated elements that represent prod- cluding mosses and hornworts, produce cytoplasmic oil drop-
ucts of deep divergences (Garbary and Renzaglia, 1998; Ren- lets (usually triglycerides), but they are not sequestered in spe-
zaglia et al., 2000). Morphological heterogeneity in the group cialized organelles. Although the function of the oil body is
is particularly evident in growth form of the gametophyte, controversial, these single-membrane-bound organelles are re-
which shows the greatest range of variability among bryo- stricted to hepatics and occur in approximately 90% of taxa.
phytes. Since the starting point of liverwort nomenclature Derived from endoplasmic reticulum in meristematic cells
(Linnaeus, 1753) and the beginning of their systematic treat- (Duckett and Ligrone, 1995), oil bodies provide valuable tax-
ment (Endlicher, 1841), hepatics have been organized into onomic information because their size, shape, number, and col-
three groups based on growth form: (1) complex thalloids, (2) or are taxon specific (Crandall-Stotler and Stotler, 2000).
simple thalloids, and (3) leafy liverworts. Conflicting concepts Unlike hornworts, but comparable to mosses, is the produc-
of diversification have led to opposing views on the direction- tion of a variety of organized external appendages, most of
ality of change within liverworts; that is, whether thalloid or which function in protecting fragile tissues. For example, var-
ious mucilage papillae, hairs, scales, bracts, cups, or flask- taxa. Davis (in press) provided the most extensive analysis of
shaped structures protect the meristem, gemmae, and other relationships among leafy liverwort genera available to date.
vegetative organs. Especially vulnerable are the superficial sex Davis (in press) reconstructed ‘‘backbone’’ relationships
organs that often occur in clusters protected by flaps of tissue, among liverworts based on a combined data set including two
leaf lobes, young leaves, or modified branches (Fig. 9B). nuclear, three mitochondrial, and eight loci sequenced from 20
The uniformity and uniqueness of liverwort sporophytes liverworts and three outgroup mosses (Fig. 7). The data were
provide compelling evidence for monophylly of hepatics. Un- analyzed using maximum parsimony, maximum likelihood,
like mosses and hornworts, sporophytes of liverworts reach and Bayesian inference, and most of the results were robust
maturity within the confines of protective gametophytic tissue to these alternative methods. The liverworts are resolved as
that develops from the shoot/thallus (5 perigynium or coelo- monophyletic, as are class Marchantiopsida (complex thal-
caule) and/or archegonium (5 calyptra). Additional gameto- loids) and Jungermanniidae (leafies). Metzgeriidae are re-
phytic structures such as perianths, pseudoperianths, bracts, solved as a grade paraphyletic to Jungermanniidae, in agree-
scales, and involucral flaps may further surround the sporo- ment with earlier studies. Although Forrest and Crandall-Sto-
phyte and associated protective tissue. In such a milieu, pho- tler (in press) sampled different species, results of their anal-
tosynthesis is limited, and the sporophyte derives nourishment ysis of five plastid loci are congruent with those of Davis (in
from the gametophyte through a placenta. The seta is pale to press).
hyaline, and the capsule is devoid of stomata. The majority of Although Haplomitrium has generally been regarded as an
liverwort sporophytes are differentiated into foot, seta, and early-diverging lineage within the liverworts (Smith, 1955;
capsule; in the occasional marchantioid taxon (e.g., Riccia, Schuster, 1984; Renzaglia et al., 1994), the precise placement
Corsinia), the seta and/or foot is vestigial or absent. At com- of this genus remains problematic. The gametophyte of Hap-
pletion of meiosis and spore development, cells of the seta lomitrium is erect and radially symmetrical and therefore rem-
typically undergo rapid elongation through water imbibition iniscent of both jungermannialean liverworts and mosses. Pri-
and thus elevate the capsule away from the substrate. Sterile, or to the discovery of antheridia and sporophytes in Takakia
elongated elaters have hygroscopic, spiraled, inner-wall thick- (Smith and Davison, 1993; Renzaglia et al., 1997), Haplomi-
enings, that are strategically interspersed among spores to fa- trium was considered closely related to Takakia because of
cilitate their separation and dispersal (Fig. 9E). Capsule de- gametophytic similarities (Schuster, 1972, 1984). More recent
hiscence normally entails a patterned separation into four lon- molecular and morphological data have come together to so-
gitudinal valves, but variations range from two valves through lidify the placement of Takakia among mosses (see later). Di-
irregular fragments or plates to cleistocarpous capsules. vergent opinions have been expressed with regard to the re-
lationship of Haplomitrium to other hepatics. A conclusion
The backbone of liverwort phylogenetic relationships— from Bartholomew-Began’s (1990, 1991) extensive morpho-
Crandall-Stotler and Stotler (2000) used morphological char- genetic reevaluation of Haplomitrium was that the genus is a
acters in a cladistic analysis of liverworts. Their analyses in- member of the simple thalloid lineage. In their analysis of land
cluded 34 taxa and 61 characters, and they resolved two main plant relationships based on rbcL sequences, Lewis et al.
lineages: complex thalloids (Marchantiopsida) and simple thal- (1997) noted that the precise position of the genus depended
loids plus leafies (Jungermanniopsida: Metzgeriidae, Junger- on the data set analyzed (1st and 2nd vs. 3rd positions, all po-
manniidae, respectively). However, their sampling was not ex- sitions, ‘‘ts/tv’’ weighting); Haplomitrium fell out sister to all
tensive enough to address phylogenetic issues within any of other embryophytes, sister to all other liverworts, or nested
the major clades. There are a few taxa for which placement within the liverworts and sister to the leafy taxa. Nuclear 18S
relative to the three large groups is ambiguous on the basis of rDNA sequences resolved Haplomitrium (without bootstrap
morphological, ultrastructural, and chemical features. These support) as sister to the class Jungermanniopsida (i.e., leafies
include Treubia and Apotreubia (Treubiales), Monoclea (Mon- plus simple thalloids; Hedderson et al., 1996).
ocleales), Sphaerocarpos, Geothallus and Riella (Sphaerocar- Recent multigene analyses have focused on two hypotheses:
pales), Blasia and Cavicularia (Blasiales), and Haplomitrium Haplomitrium is either sister to Jungermanniopsida or sister to
(Haplomitriales [Calobryales]). Early molecular analyses of all other liverworts. In contrast to almost all other nodes on
the liverworts were limited to single genomic regions with her tree, Davis (in press) reported that the placement of Hap-
limited taxon sampling (e.g., Lewis et al., 1997; Bopp and lomitrium varied among analyses. Under parsimony, likeli-
Capesius, 1998; Beckert et al., 1999; Stech and Frey, 2001) hood, and Bayesian methods, Haplomitrium is resolved with
but recent multigene analyses with increased sampling have strong support as sister to Jungermanniopsida (simple thalloids
begun to clarify phylogenetic relations among (and within) the plus leafies), and this inclusive clade is in turn sister to Mar-
major groups of liverworts. Phylogenetic relationships within chantiopsida (complex thalloids; Fig. 7). However, the most
Marchantiopsida (complex thalloids) from DNA sequence data complex heterogeneous Bayesian substitution model, with 21
were analyzed by Bischler (1998), Wheeler (2000), and Bois- partitions, yielded Haplomitirum as the sister group to all other
selier-Dubayle et al. (2002). Forrest and Crandall-Stotler (in liverworts. Forrest and Crandall-Stotler (in press) and Qiu
press) focused on Metzgeriidae (simple thalloids), whereas He- (2003) reported that Haplomitrium plus Treubia form a clade
Nygrén et al. (in press) sampled a wide diversity of liverwort sister to all other hepatics. However, the sister-group relation-
←
Fig. 8. Phylogenetic relationships among liverworts, especially the Jungermanniidae (leafies). Homogeneous Bayesian 95% majority rule tree from a four-
gene data matrix (Davis, in press). Bold branches indicate significant support for the clade in all Bayesian analyses (homogeneous and heterogeneous posterior
probabilites $95). Parsimony bootstrap values $50 are shown on the tree.
Fig. 9. Morphological diversity in liverworts. A. Photograph of complex thalloid gametophyte of Conocephalum conicum (L.) Lindb. Note polygonal air
chambers on the dorsal surface. Bar 5 1.0 cm. B. SEM of simple thalloid gametophyte of Pallavicinia lyellii (Hook.) Gray showing monostromatic wings and
thickened midrib. Flaps of tissue on either side of the midrib cover protect antheridia in this male plant. Bar 5 1.0 mm. C. Photograph of leafy gametophyte
of Bazzania trilobata (L.) S. Gray showing incubous leafy insertion from dorsal aspect. The shoot on the right is seen from the ventral side revealing the small
row of underleaves. Bar 5 1.0 mm. D. Light micrograph of large oil bodies in the leaves of Calypogeia muelleriana (Schiffn.) K. Muell. Bar 5 50 mm. E.
Light micrograph of spores and elongated spiraled elaters in Pallavicinia lyellii (Hook.) Gray. Image provided by Scott Schuette. Bar 5 50 mm.
ship was unsupported. When Treubia was excluded from the complex thalloid liverworts include relatively drought-resis-
analysis by Forrest and Crandall-Stotler (in press), the position tant species. Many morphological features of Marchantiopsida
of Haplomitrium was unresolved. Thus, the affinities of Hap- indicate xeromorphic adaptations (Schuster, 1992; Wheeler,
lomitrium are not yet satisfactorily resolved; Davis (in press) 2000). In addition to air chambers in the dorsal part of the
felt that the weight of the current evidence supports a position thallus (Fig. 9A), marchantioid liverworts are characterized by
for the genus as sister to the class Jungermanniopsida, whereas two types of rhizoids (smooth and pegged), archegonial in-
Qiu (2003) and Forrest and Crandall-Stotler (in press) favor a volucres, unlobed spore mother cells, four primary androgones
position as sister to all other hepatics. in the antheridium, six rows of neck cells in the archegonium,
Although unexpected, the affinity between Treubia and idioblastic oil body cells, ventral thallus scales, unistratose
Haplomitrium finds support in morphology. Both are ‘‘leafy’’ capsule walls, and a simple locomotory apparatus in the small
taxa with gametangia situated in leaf axils or lobules. Treubia biflagellated sperm cell (Schuster, 1966, 1992; Renzaglia et
is decisively more dorsiventral, with an oblique to transverse al., 2000; Renzaglia and Garbary, 2001). Of these, only fea-
leaf insertion (succubous) and small dorsal lobules (Renzaglia, tures of the sperm appear to be universal in all species.
1982), whereas some species of Haplomitrium tend toward Although Marchantiopsida are resolved as monophyletic,
anisophylly and succubous insertion (Bartholomew-Begin, traditional relationships among taxa generally are not sup-
1991). In both genera, a tetrahedral apical cell is responsible ported by molecular data. The classical morphological sepa-
for shoot growth. Perhaps the most compelling evidence for a ration of this liverwort class into three orders; i.e., Monocle-
close relationship between these two genera icomes from the ales, Sphaerocarpales, and Marchantiales, is challenged by nu-
peculiar yet similar sperm cells that they produce. Cladistic cleotide sequence data (Wheeler, 2000; Boisselier-Dubayle et
analyses based on spermatogenesis consistently recovered a al., 2002). Incongruence between morphological and molecu-
Treubia plus Haplomitrium clade that is sister to the remaining lar patterns may be attributed to parallel changes in multiple
liverworts (Garbary et al., 1993; Renzaglia and Garbary, lineages (Boisselier-Dubayle et al., 2002).
2001). Stech et al. (2000) elevated Treubia to class Treubiop- The multigene analyses of Davis (in press, Fig. 8) and For-
sida based on trnL intron sequence divergences between it and rest and Crandall-Stotler (in press) provided strong support for
other liverworts. the placement of Blasia as a member of the complex thalloids,
a result that conflicts with the traditional placement of this
Systematics and phylogeny of the Marchantiopsida (com- liverwort within the simple thalloids (Renzaglia, 1982). Sperm
plex thalloid liverworts)—Unlike other hepatic groups, the cell features, persistent ventral scales, a small wedge-shaped
Fig. 10. Diagrammatic representation of apical cell shapes in liverworts

as oriented in plants growing horizontally. The shoot tip is directed toward
the left. A. Tetrahedral cell with three cutting faces. B. Wedge-shaped or Fig. 11. Formative divisions in the lateral derivative from apical cells of
cuneate cell with two lateral, one dorsal and one ventral cutting face. C. liverworts. Apices are vertically oriented so that the shoot tip is facing up.
Lenticular or lens-shaped cell with two lateral cutting faces. D. Hemidiscoid A. Lateral thallus, wing and ‘‘leaf’’ development from a central wedge-shaped
cell with two lateral and one posterior cutting face. This cell is rare in liv- cell (single initial) (L) in a three-celled derivative. This type of development
erworts and was developmentally and evolutionarily derived from a wedge- characterizes all simple thalloid and complex thalloid taxa and occurs in de-
shaped cell. rivatives from all four types of apical cells. B. ‘‘True’’ leaf development in
leafy liverworts from two initials. In this five-celled derivative, two leaf ini-
tials (L) are determined to develop into the bifid and conplicate–bilobed
leaves.
apical cell, and a Monoclea-like female involucre provide mor-
phological evidence for the inclusion of Blasia in the complex
thalloid lineage (Renzaglia and Duckett, 1987; Pass and Ren- in Marchantiopsida. However, congruence among the multi-
zaglia, 1995; Renzaglia and Garbary, 2001). Previous molec- gene analyses provided support for Monoclea close to Du-
ular analyses based on one or two gene sequences do not agree mortiera.
in the placement of Blasia. Stech and Frey (2001) resolved
Blasia as sister to Jungermanniopsida (simple thalloids plus Systematics and phylogeny of Metzgeriidae (simple thal-
leafies) and described the new class, Blasiopsida. Their study loid liverworts)—Clearly not a monophyletic group, Mezger-
was based solely on trnL intron sequences (ca. 500 bp), and iidae traditionally include some 30 highly diverse genera of
the relationship was without bootstrap support. Wheeler ‘‘leafy’’ and thalloid forms. Although four apical cell types are
(2000) found that Blasia grouped with the simple thalloids found in the group (Fig. 10), a unifying feature of apical
(Metzgeriidae) based on 26S nrDNA (also without bootstrap growth in these plants is development of wings and leaves
support), and He-Nygrén et al. (in press) resolved Blasia as from a central wedge cell (single initial) that forms in the
sister to the remaining liverworts. newly produced apical derivative (Fig. 11A) (Renzaglia,
After Blasia, Sphaerocarpos is the next divergent taxon 1982). Simple thalloid genera are distinguished from leafy liv-
(Fig. 8). A position for Sphaerocarpales (Sphaerocarpos, Riel- erworts (Jungermanniidae) in that they are anacrogynous: ar-
la, Geothallus) among complex thalloids is generally support- chegonia are produced along the mid-thallus of either the
ed by morphology (Smith, 1955; Bishler, 1998; Crandall-Sto- main, lateral, or ventral shoots. Consequently, the apical cell
tler and Stotler, 2000; Boisselier-Dubayle et al., 2002). How- is not transformed into permanent tissue after archegonial de-
ever, with additional taxon sampling, the sister relationship velopment, and sporophytes do not terminate the shoot as in
between Sphaerocarpales and the remaining Marchantiopsida acrogynous Jungermanniidae. Additional features that unify
is called into question. Based on LSU rDNA sequences, the simple thalloid taxa, but are also found in leafy liverworts,
Wheeler (2000) and Boisselier-Dubayle et al. (2002) reported are the development of antheridia from two primary andro-
that Sphaerocarpales were placed within Marchantiaceae. Sim- gones, oil bodies in all cells, lobed sporocytes, smooth rhi-
ilarly, Sphaerocarpos nested between Neohodgsonia and Mar- zoids, and five rows of neck cells per archegonium.
chantia in the five-gene analysis of Forrest and Crandall-Sto- Two assemblages of simple thalloid taxa are paraphyletic
tler (in press). The implication from these results is that the (simple thalloid I and II in Fig. 8) within Jungermanniopsida.
relatively simple morphology of both generations in Spharo- The first group (simple thalloid I) is the most diverse and
carpales may not be plesiomorphic but rather the product of includes Phyllothallia, generally placed in Treubiales, most
extreme simplification in ephemeral or aquatic habitats. members of Fossombroniales, and suborder Pallaviciniineae of
Air chambers are found in the crown group taxa (Marchan- Metzgeriales (classification according to Crandall-Stotler and
tia, Preissia, Targionia, Riccia) (Wheeler, 2000). One lineage, Stotler, 2000). Placement of Phyllothallia, Pellia, Calycularia,
Monoclea plus Dumortiera, has secondarily reverted to a mor- and Noteroclada is not resolved in the five-gene analysis of
phologically simple thallus devoid of chambers, perhaps ad- Forrest and Crandall-Stotler (in press); all represent genetically
aptations to the semi-aquatic habit of these plants (Wheeler, and morphologically divergent taxa. Phyllothallia and Noter-
2000). The production of archegoniophores (carpocephala) oclada are distinctly ‘‘leafy’’ in habit, but development is from
that elevate sporophytes above the gametophyte also evolved a wedge-shaped cell (Fig. 10B) in the former and a tetrahedral
within the crown Marchantiopsida group. Independent losses apical cell (Fig. 10A) in the latter (Renzaglia, 1982). Pellia
of these structures occurred in riccioid taxa (Riccia, Riccio- and Calycularia are fleshy thalloid types, both with wedge
carpos, Oxymitra) and Monoclea (Wheeler, 2000). Reduction (Fig. 10B) and hemidiscoid apical cells (Fig. 10D). Although
in sporophyte complexity is likewise a derived feature of ric- support for an assemblage that includes Fossombronia, Aus-
cioid liverworts (Renzaglia et al., 2000; Boisselier-Dubayle et trofossombronia, Petalophyllum, and Allisonia is weak, this
al., 2002). With a jungermannioid-like sporophyte elevated on group includes most of the genera traditionally placed in Fos-
a fragile and highly elongated seta, Monoclea seems inappro- sombroniales (Crandall-Stotler and Stotler, 2000). All have
priately placed within this crown group. Additional characters spheroidal capsules, which are typically irregular or nonval-
of the genus, including a free nuclear embryo and monoplas- vate in dehiscence. Most exhibit a ‘‘leafy’’ growth form with
tidic meiosis in some species (Schofield, 1985; Renzaglia et either lens-shaped (Fig. 10C) or tetrahedral (Fig. 10A) apical
al., 1994), support a more traditional placement of Monoclea cells. Suborder Pallaviciniinae of Metzgeriales are recovered
as monophyletic and include Hymenophyton, Moerckia, Hat- a single leaf (Zartman, 2003). These organisms are important
torianthus, Podomitrium, Pallavicinia, Jensenia, Xenothallus, components of tropical forest diversity, and the diversity of
and Symphyogyna (Crandall-Stotler and Stotler, 2000; Forrest epiphylls (almost exclusively Lejeuneaceae) is a sensitive in-
and Crandall-Stotler, in press). This morphologically uniform dicator of habitat change associated with forest fragmentation
group contains upright or procumbent taxa, most with prom- (Zartman, 2003).
inent midribs and monostromatic wings (e.g., Pallavicinia, Jungermanniidae are distinguished from Metzgeriidae in
Fig. 9B). Lens-shaped apical cells (Fig. 10C) are responsible having tetrahedral apical cells, gametophytic shoots with (usu-
for vegetative growth. An autapomorphy of this group is the ally) well-differentiated stems and leaves, leaves formed from
production of specialized strands of dead, water-conducting two initial cells, acrogynous perichaetia (terminating the main
cells that predominate in most taxa (Ligrone et al., 2000). Ex- stem or branch), bracts and perianths (modified, fused leaves)
tensive variability is seen in position and type of protective associated with the perichaetium, and capsules that regularly
structure associated with gametangia and sporophytes (Ren- dehisce into four valves (Crandall-Stotler and Stotler, 2000).
zaglia, 1982; Fig. 9B). The perianths of leafy liverworts are diverse and provide im-
The apparent affinity between suborder Metzgeriinae (sim- portant taxonomic characters in many genera and families.
ple thalloid II) and Jungermanniidae (‘‘true’’ leafy liverworts) Leaves of leafy liverworts may be entire or more often have
in the multigene analyses of both Davis (Fig. 8) and Forrest two large lobes or teeth. They are most commonly differen-
and Crandall-Stotler (in press) was unexpected. Members of tiated as two rows of lateral leaves and a single row of ventral
Metzgeriinae epitomize the simple thalloid condition, with underleaves (amphigastria; Fig. 9C). Underleaves are frequent-
fleshy (Aneuraceae) and midrib-plus-wing (Metzgeriaceae) or- ly small or lacking. Insertion of the lateral leaves may be trans-
ganizations. All of these thalli develop from a lens-shaped api- verse, or, more commonly, they are oblique and the plants are
cal cell (Fig. 10C), and no ‘‘leafy’’ forms exist (except perhaps more or less flattened because the leaves overlap. In plants
Pleurozia, discussed next). Endogenous branches in Metzger- with incubous leaf orientation, the forward leaf margin over-
iaceae are reminiscent of those in leafy liverworts (Renzaglia, laps the trailing margin of the next younger leaf, resembling
1982). the arrangement of roof shingles (Fig. 9C). In sucubous ori-
One of the most surprising results from the Davis (in press) entation, forward margins of older leaves are covered by over-
analyses was the placement of Pleurozia in Metzgeriinae (sim- lapping trailing margins of the younger leaves. In species with
ple thalloid II) rather than among the ‘‘true’’ leafy liverworts complicate–bilobed leaves, lateral leaves are each folded to
(Figs. 7, 8). Pleurozia is composed of about 11 species dis- form ventral and dorsal lobes. The dorsal lobe is larger in most
tributed primarily in the tropics. Leaves are complicate-bi- taxa, and the ventral lobe may be highly modified into the
lobed, and for that reason, Pleurozia has traditionally been form of a pouch or helmet-shaped lobule that holds water.
included in or near Porellales within the leafy liverworts Schuster (1966, 1984) assumed that the most primitive liv-
(Schuster, 1984; Crandall-Stotler and Stotler, 2000). However, erworts would be the most mosslike, with leafy, radially sym-
leaf morphology in Pleurozia is unique in that the leaf lobule metric gametophytes and therefore placed leafy taxa at the
is dorsal in orientation, not ventral (Thiers, 1993), and the base of his subjectively derived ‘‘phylogenetic trees’’ (e.g.,
plants grow from a lenticular apical cell (Crandall-Stotler, Schuster, 1966, pp. 406, 696). He considered leafy taxa with
1976) rather than a tetrahedral cell as in all ‘‘true’’ leafy liv- radial symmetry and three rows of transversely (or nearly so)
erworts. The placement of Pleurozia in the metzgerioid liv- inserted leaves (e.g., Herbertineae) to be early diverging
erworts indicates that the ‘‘leafy’’ gametophytes of Pleurozia, groups, and from these he showed the branching of lineages
with their complicate–bilobed leaves, may have evolved con- or clusters of lineages with increased anisophylly and more
vergently in a group otherwise characterized by thalloid ga- obliquely inserted leaves (Schuster, 1966, 1972, 1984). One
metophytes. In contrast to the single leaf initial in simple thal- group includes Schistochilaceae, Cephaloziaceae, Lepidozia-
loids that have ‘‘leafy’’ gametophytes (Fig. 11A), leaves of ceae, and Pleuroziaceae, whereas the other progresses through
Pleurozia develop from two initial cells, as is typical of leafy Ptidiaceae to Jungermanniaceae, Frullaniaceae and Lejeune-
liverworts (Fig. 11B; Crandall-Stotler, 1976). The phyloge- aceae. (His diagram shows extant families ancestral to other
netic position of Pleurozia should be further investigated, al- families.) The classification of Crandall-Stotler and Stotler
though its placement within subclass Metzgeriidae is strongly (2000) has a sequence of families in five orders, Lepicoleales
supported by both the 12- and four-locus analyses of Davis (in (including Ptilidiaceae, Lepicoleaceae, Schistochilaceae, and
press). Lepidolaenaceae), Jungermanniales (including Herbertaceae,
Balantiopsidaceae, Geocalycaceae, Lepidoziaceae, Cephalo-
Systematics and phylogeny of the Jungermanniidae (leafy ziaceae, Jungermanniaceae, and Gymnomitriaceae), Porellales
liverworts)—The leafy liverworts, with some 4000–6000 spe- (including Porellaceae, Jubulaceae, and Lejeuneaceae), and the
cies, are by far the largest of the liverwort groups. They occur monotypic Radulales and Pleuroziales. Their classification im-
in most terrestrial and aquatic habitats but are especially di- plies similar concepts of evolution in leafy liverworts to those
verse in high-moisture environments. Many species are epi- of Schuster.
phytic on bark, and in the tropics, epiphyllous liverworts may In a liverwort backbone tree based on 12 loci, Davis (in
cover the leaves of angiosperm trees and shrubs in shaded, press) resolved two major clades within subclass Jungerman-
high-humidity forests. More than 75% of the liverworts of niidae (Fig. 7). One clade contains most of the taxa with com-
tropical lowland forests and almost all the epiphylls belong to plicate–bilobed, incubous (or transverse) leaves (mainly Po-
Lejeuneaceae (Gradstein, 1994, 1997). Lejeuneaceae comprise rellaceae, Jubulaceae, Radulaceae, and Lejeuneaceae), whereas
approximately 93 of the 307 genera (30%) of leafy liverworts, the other contains the remaining families of leafy liverworts.
and well over 1000 species (Gradstein, 1979, 1994, 1997; In a more taxon-extensive analysis that included 81 liverworts,
Crandall-Stotler and Stotler, 2000). In lowland equatorial for- two mosses, and a hornwort, based on sequences from 26S
ests, as many as 20 species of Lejeuneaceae may co-occur on nrDNA, two plastid loci ( psbA and rps4), and mitochondrial
nad5, the same two leafy liverwort clades were resolved (Fig. occurs during development, and this plasticity may have pro-
8). The noncomplicate-bilobed group consists of three sub- vided fuel for evolutionary change (Renzaglia et al., 2000).
clades for which sister group relationships are ambiguous (A, Depending on the position of Haplomitrium in the trees, either
B, and C in Fig. 8). Species in clade A have incubous or a tetrahedral or wedge-shaped cell is plesiomorphic. Similarly,
transverse leaf insertion, well-developed underleaves, and mul- either an upright ‘‘leafy’’ habit or a flattened thallus is ances-
tilobed lateral leaves (Davis, in press). Herbertus, assumed by tral in hepatics; both hypotheses have garnered support
Schuster (1984) to be primitive among leafy liverworts, is re- (Schuster, 1992; Mishler and Churchill, 1984). Outgroup com-
solved in a derived position within clade A (Fig. 8). Moreover, parisons provide no further resolution of this issue as hornwort
other isophyllous taxa (e.g., Anthelia, Triandrophyllum) are and pteridophyte gametophytes are thalloid with wedge-
also resolved in relatively derived phylogenetic positions. Taxa shaped apical cells, whereas mosses are leafy with tetrahedral
in clade B have sucubous or transverse leaves and generally cells.
lack underleaves; however, lateral leaf shape is variable. Leaf Within liverworts, significant evolutionary changes can be
shape, insertion, and underleaf development are highly vari- inferred at the cellular level based on the consensus topology
able in clade C, but many of the species are characterized by of recent molecular analyses. Monoplastidic meiosis occurs in
having perichaetia formed in fleshy perigynia or marsupia, all mosses and hornworts. However, it is restricted in liver-
which do not occur elsewhere in the leafy liverworts. worts to Haplomitrium, Blasia, and Monoclea (Renzaglia et
Among suborders of leafy liverworts recognized by Schus- al., 1994) and is best interpreted as plesiomorphic. Monoplas-
ter (1984), only Radulineae and Balantiopsidineae are mono- tidic meiosis involves precise control of plastid division and
phyletic based on the four-locus analysis of Davis (in press). migration prior to chromosomal separation (Brown and Lem-
The classification of Crandall-Stotler and Stotler (2000) is also mon, 1990). Polyplastidic meiosis predominates in liverworts
in conflict with many of the phylogenetic inferences from Da- and is a derived state. Similarly, lobed spore mother cells that
vis’s analysis (Fig. 8). Notably, Lepicoleales are extensively occur in liverworts such as Haplomitrium, Treubia, and Blasia
polyphyletic, and Radulales are nested within Porellales. Her- are shared with other bryophytes and represent a plesio-
bertaceae, Lepidoziaceae, Balantiopsidaceae, Cephaloziaceae, morphic condition (Brown and Lemmon, 1988). Sporocyte
Porellaceae, and Radulaceae are supported as monophyletic. lobing was lost within Marchantiopsida, whereas spores united
Lejeuniaceae are monophyletic only if Bryopteris is included in permanent tetrads are viewed as derived within Sphaero-
within them (Bryopteridaceae, fide Crandall-Stotler and Sto- carpales. Among bryophytes, pre-meiotic patterning of spore
tler, 2000). Jungermanniaceae, Gymnomitriaceae, Geocalyca- wall ornamentation occurs in Apotreubia and Haplomitrium
ceae, Cephaloziaceae, Lepidolaenaceae are paraphyletic (Da- and presumably has been lost in more derived liverwort line-
vis, in press). ages (Brown et al., 1986). Further ultrastructural studies across
a range of hepatic groups are likely to provide new insights
Inferences about morphological evolution in liverworts into the nature and direction of changes in cellular processes
from molecular analyses—Leaves or leaflike lobes have during early land plant evolution.
evolved in every major group of hepatics. Haplomitrium and
Treubia have leafy appendages. Blasia and Sphaerocarpos, BRYOPHYTA
taxa within the marchantioid line, have leafy habits. Phylloth-
allia, Noteroclada, and Pleurozia, with leafy gametophytes, Moss classification and relationships—Division Bryophy-
are scattered among simple thalloid taxa. The leaves of these ta, or mosses, include about 10 000 species (Crosby et al.,
plants are typically succubous to transversely inserted and may 2000). Systematic knowledge about the mosses has grown
be formed from any one of three apical cell types: wedge- steadily since Hedwig (1801), the starting point for moss no-
shaped, lenticular, or tetrahedral. In Phyllothallia and Noter- menclature (excluding Sphagnum), recognized 32 genera.
oclada (Fig. 10A), each leaf develops from a single initial, Most classifications of the 19th century depended on gameto-
whereas in Pleurozia there are two initials. A single initial is phyte characters for defining the major groups of mosses (e.g.,
also responsible for development of wings and lateral thallus Bruch et al., 1851–1855; Kindberg, 1897). Mosses (excluding
in all simple thalloids, complex thalloids, hornworts, and many Sphagnum and Andreaea) were divided into acrocarpous and
pteridophyte gametophytes and is thus best viewed as plesio- pleurocarpous taxa (Mitten, 1859). Acrocarpous mosses have
morphic. archegonia terminating the main stems, which tend to be
An autapomorphy of the Jungermanniidae is the production sparsely if at all branched. Pleurocarpous mosses, in contrast,
of bifid leaves from two leaf initials in a derivative from a have archegonia borne laterally along relatively highly
tetrahedral apical cell (Fig. 10B). Once ‘‘locked’’ into this pat- branched, generally procumbent or pendent, extensively inter-
tern of cell divisions, a number of variations on the ‘‘typical’’ woven stems. The two forms of gametophyte architecture are
bifid leaf of hepatics evolved, including complicate–bilobed often obvious, but some taxa are confusingly intermediate
leaves. A narrower ventral cutting face in the apical cell is (e.g., Rhizogoniaeae, Orthotrichaceae, Hedwigiaceae) because
associated with a smaller size or absence of underleaves that they have moderately branched stems with archegonia termi-
originate from it. Incubous leaf insertion results from a ventral nating short to long lateral branches. La Farge-England (1996)
(downward) tilt of the apical cell (Crandall-Stotler and Stotler, clarified the definitions of these forms of gametophytic archi-
2000), a feature that is often correlated with taxa that grow on tecture and discussed possible phylogenetic relations between
vertical substrates such as tree bark (e.g., Leujeuniaceae). Con- taxa characterized by acrocarpous, pleurocarpous, and clado-
versely, succubous leaf arrangements are correlated with a dor- carpous gametophyte architecture (the latter including the
sal (upward) tilt of the growing tip. seemingly intermediate forms).
Few conclusions can be drawn at present about the evolu- Philibert (1884–1902) published a series of seminal papers
tion of apical cell shapes and growth forms in liverworts. describing variation in the structure of the moss peristome
Transformation from one apical cell type to another readily (sporophytic) and distinguished several basic types character-
plates of cell wall material; most of the anticlinal walls are

resorbed prior to maturity. Arthrodontous peristomes may have
one or two rings of teeth. Diplolepideous peristomes typically
consist of an outer ring of 16 teeth, the exostome, and an inner
more delicate ring, the endostome (Fig. 13F, K). Endostome
teeth, when present, are referred to as segments and may be
united below as a basal membrane. Diplolepideous peristomes
take their name from the fact that a vertical line extends down
the outer surface of each exostome tooth because each consists
of two vertical rows of cell wall plates on the dorsal surface.
These plates are derived from two columns of periclinal cell
walls (Fig. 13F). Haplolepideous peristomes, in contrast, (typ-
ically) consist of a single ring of 16 teeth, and each tooth lacks
a median vertical line extending down its dorsal surface (Fig.
Fig. 12. ‘‘Backbone’’ phylogenetic tree of mosses showing relationships 13I). The ‘‘haplo’’ of haplolepideous refers to the single col-
among major clades characterized by different peristome types. This is the
single optimal tree under maximum likelihood (Cox et al., in press).
umn of cell wall plates on the outer (dorsal) surface, not the
fact that most haplolepideous peristomes consist of a single
ring of teeth. In fact, some haplolepideous peristomes have
izing large groups of taxa. Fleischer (1923) developed a rad- additional irregular teeth formed external to the main peri-
ically new classification of moss diversity based on Philibert’s stome. Arthrodontous peristomes are derived from the inner-
peristome observations for his flora of Java and adjacent re- most three layers of the amphithecium: inner peristomial layer
gions, and variations on this classification were utilized (IPL), primary peristomial layer (PPL), and outer peristomial
through almost all of the 20th century. The most influential layer (OPL). Whereas haplolepideous peristomes are formed
classification utilizing Philibert’s observations and Fleischer’s from the IPL and PPL, diplolepideous peristomes are derived
taxonomic concepts was Brotherus’ (1924–1925) worldwide from all three peristomial layers. Exostomes form from adja-
synopsis of mosses for Engler and Prantl’s Die natürlichen cent OPL and PPL cells, and endostomes are formed from
Pflanzenfamilien. With minor modifications, the Brotherus sys- adjacent PPL and IPL cells.
tem formed the basis for moss classification (e.g., Vitt, 1984) Diplolepideous peristomes may have endostome segments
until the last five years, during which insights from molecular that lie opposite the exostome teeth (the Funaria- or diplole-
analyses have accumulated (Buck and Goffinet, 2000) (Fig. pideous-opposite type; Fig. 13K) or alternate with them
12). (Bryum or diplolepideous-alternate type; Fig. 13F). Endo-
Toward the end of the 20th century, Edwards (1979) and stomes in the Funaria-type consist of relatively massive teeth
Vitt (1984) provided refined insights into differences between (reduced or absent in some taxa) without a basal membrane.
the basic peristome types in mosses. Evans and Hooker (1913), The endostomes in (well-developed) Bryum-types are more
Blomquist and Robertson (1941), Shaw and Anderson (1988), membranous and consist of a basal membrane and 16-keeled
Shaw et al. (1987, 1989a, b), and Goffinet et al. (1999) doc- segments. Narrow cilia may occur between the endostome seg-
umented developmental characteristics of the peristome types ments. Vitt (1981) argued that peristomes found in Orthotri-
and differences between them. chaceae constitute another basic type, but this interpretation
Two basic types of peristome, nematodontous and arthro- was not supported by morphological studies of Shaw (1986)
dontous, are distinguished by whether the teeth are formed or Goffinet et al. (1999).
from whole dead cells or just remnants of cell walls, respec- Peristomes characterizing Buxbaumiaceae and Diphysiaceae
tively. Nematodontous peristomes are heterogeneous in both have been interpreted as intermediate between nematodontous
development and mature structure. The so-called Polytrichum and arthrodontous types (Edwards, 1979, 1984; Vitt, 1984). In
type (Shaw and Robinson, 1984) consists of 32 or 64 teeth Buxbaumia, outer teeth are derived from whole cells, whereas
united at their tips by a membranous epiphragm (Fig. 13A– inner teeth are arthrodontous (Fig. 13E). More than three am-
E). The teeth are formed from whole cells derived from the phithecial layers contribute to peristomes, as in nematodontous
innermost four to eight layers of the amphithecium. (Endothe- types. The Diphyscium peristome is entirely arthrodontous but
cium and amphithecium are embryonic tissues that differen- is a pleated cone unlike any other arthodontous peristome (Fig.
tiate early in the ontogeny of bryophyte capsules.). Polytri- 13H).
chum-type peristomes are uniquely characterized by a series Peristomial formulae describe the numbers of cells in the
of early anticlinal divisions in the amphithecium, and, because three peristomial layers as revealed by patterns of vertical and
peristome development involves remarkably regular alternat- horizontal lines visible on mature peristome teeth. These lines
ing anticlinal and periclinal divisions, the amphithecium ends represent remnants of anticlinal walls from cells in peristomial
up having double the number of cells compared to arthrodon- layers and thus, numbers of cells in the layers. Formulae spec-
tous and other nematodontous types (Fig. 13B). The Polytri- ify numbers of cells in the IPL, PPL, and OPL in 1/8 of the
chum-type peristome is further characterized by complex pat- capsule’s circumference (Edwards, 1984). Haplolepideous
terns of cell deformation during development (Fig. 13C). The peristomes are characterized by a 4 : 2 : 3 (OPL : PPL : IPL)
other form of nematodontous peristome, the Tetraphis-type, is formula (rarely 4 : 2 : 1), and diplolepideous peristomes by
simpler in development, including the absence of the addi- formulae of 4 : 2 : 4–12 (Edwards, 1984; Shaw and Rohrer,
tional anticlinal divisions found in the Polytrichum-type (Shaw 1984). Another significant feature distinguishing peristome
and Anderson, 1988), and consists at maturity of four massive types is whether anticlinal walls in IPL are laterally offset with
teeth derived from the entire amphithecium (Fig. 13D). regard to anticlinal walls in PPL (Fig. 13G, J). Offset walls
Arthrodontous peristomes consist of (mainly) periclinal characterize haplolepideous and diplolepideous-alternate peri-
stomes but not the diplolepideous-opposite (Funaria-) type. but present in Andreaeobryum), mode of capsule dehiscence,
Anticlinal walls in the Polytrichum-type nematodontous peri- and timing of perichaetium differentiation relative to sporo-
stomes are not offset (Wenderoth, 1931), and there is little if phyte development. Newton et al. (2000), Goffinet et al.
any offsetting of the walls in Tetraphis-type nematodontous (2001), and Cox et al. (2004) resolved Andreaea and An-
peristomes (Shaw and Anderson, 1988). The peristome of Di- dreaeobryum in a single clade, but without impressive support
physcium has a developmental pattern that conforms in all de- from the bootstrap, Bremer support indices, and Bayesian pos-
tails to the haplolepideous peristomial type and has a 4 : 2 : terior probabilities. It remains possible, though not likely, that
3 formula at maturity despite the unique structure of the ma- Andreaea and Andreaeobryum form a paraphyletic grade lead-
ture peristome (Shaw et al., 1987). ing to the ‘‘true’’ (peristomate) mosses. Murray (1988) noted
Developmental studies have succeeded (even if based on morphological similarities that might link Andreaeobryum
few taxa) in defining when and how the basic peristome types with Takakia.
differ from one another. These studies do not, however, claritfy One of the most exciting insights from phylogenetic anal-
phylogenetic relationships among the types. One of the central yses of mosses is that the monospecific genus, Oedopodium,
goals of higher-level phylogenetic analyses for mosses has appears to be sister to all remaining peristomate taxa (Fig. 15).
been to resolve these relationships. Much progress has been A critical position for Oedopodium corroborates the results of
made, but a full resolution is still forthcoming. Newton et al. (2000) from combined analyses of morphology
and four plastid DNA regions and of Goffinet et al. (2001)
The backbone of moss phylogenetic relationships—Vari- based on a taxon-extensive analysis of plastid rps4 sequences.
ous approaches to resolving relationships among mosses have Goffinet et al. (2001) resolved Oedopodium as sister to a clade
included taxon-extensive analyses using a single plastid gene containing Tetraphidaceae and Polytrichaceae at or near the
(rps4: Goffinet et al., 2001; rbcL: Tsubota et al., 2002), and base of peristomate mosses, but relationships were not fully
analyses of multigene, multigenomic data sets with more syn- resolved and without bootstrap support. Newton et al. (2000)
optic taxon sampling (Cox et al., 2004). Both approaches have and Magombo (2003), based on four plastid DNA regions,
their merits, but it is clear that resolution of well-supported resolved Oedopodium, both with moderate to strong bootstrap
relationships among the major groups of mosses will not be support, as sister to all peristomate mosses.
accomplished using one or a few genes, even if such analyses Oedopodium griffitheanum (Dicks.) Schwaegr. is a small ac-
succeed in placing more genera into monophyletic groups. The rocarpous moss with soft obovate to spathulate leaves, thin-
best-supported ‘‘backbone’’ for mosses was derived from an walled hexagonal leaf cells, erect capsules with a well-devel-
analysis of eight genes representing the mitochondrial, plastid, oped, long-tapered sterile neck region, a well-developed oper-
and nuclear genomes of 30 exemplars that represent major culum but no peristome (Fig. 15). Stalked multicellular gem-
lineages based on previous studies (Cox et al., 2004). The mae are sometimes formed in the leaf axils (Smith, 1978). The
following synopsis is based on that analysis, with discussion species is uncommon, but reported from Alaska, Greenland,
of supportive and/or contradictory evidence when appropriate. Britain, Scandinavia, and Japan, and it is disjunct in the South-
With sequences from four species of liverworts as the out- ern Hemisphere on the Falkland Islands, where it grows on
group, the Bayesian reconstruction presented by Cox et al. peaty soil, typically in rock crevices (Smith, 1978; Mahú,
(2004) indicated that Sphagnum and Takakia form a clade sis- 1979; Noguchi, 1988). Oedopodium has previously been clas-
ter to all remaining mosses (Fig. 12). A close relationship be- sified near Funariaceae, mainly because of similarities in ga-
tween Sphagnum and Takakia was also resolved by Hedderson metophyte morphology (especially the broad, soft-textured
et al. (1998), Newton et al. (2000), and Yatsentyuk (2001) leaves with large, thin-walled cells). The absence of a peri-
from nucleotide sequences, although Newton et al. (2000) stome in Oedopodium may well be plesiotypic, although the
were not able to identify any morphological synapomorphies possibility of secondary loss cannot be eliminated (Cox et al.,
uniting the two genera. Gametophytes of Sphagnum and Tak- 2004).
akia could not be more divergent: those of Takakia are tiny, Mosses characterized by nematodontous peristomes form a
simple in structure, and reminiscent of liverworts, whereas grade paraphyletic to the arthrodontous taxa (Fig. 12). Poly-
those of Sphagnum are large and characterized by a number trichales form a monophyletic group sister to the rest of the
of autapomorphies. The sporophyte of Takakia is mosslike in peristomate mosses; Tetraphis (representing the small family,
development (Renzaglia et al., 1997) with a well-developed Tetraphidaceae) is next diverging, then Buxbaumia, and Di-
seta, a cylindrical capsule, and spiraled dehiscence (Fig. 14), physcium. Monophylly of Polytrichales based on the eight-
whereas capsules of Sphagnum are ovoid in shape, open by gene data set corroborated earlier results of Hyvönen et al.
an apical operculum, and are elevated on gametophytic pseu- (1998), Newton et al. (2000), and Magombo (2003). Tetra-
dopodia. In the analyses of Cox et al. (2004), support for the phidaceae, characterized by four massive, nematodontous peri-
clade containing Sphagnum and Takakia was lower when sub- stome teeth, are not part of the monophyletic Polytrichaceae.
stitution patterns were modeled separately for each of the eight This result makes sense in terms of peristome structure and
genomic regions than when a single model was applied, thus development; Tetraphis does not have the ‘‘extra’’ anticlinal
raising the possibility that resolution of the Sphagnum-Takakia division that characterizes the amphithecial layers of Polytri-
clade may be an artifact. Phylogenetic relationships among chaceae (Shaw and Anderson, 1988) nor the complex pattern
species within Sphagnopsida (Sphagnum and Ambuchanania) of cell malformation that occurs during peristome development
have been described by Shaw (2000) and Shaw et al. (2003a). in Polytrichaceae. Aside from their nematodontous structure,
After Sphagnum and Takakia, the next diverging clade of peristomes of Tetraphidaceae and Polytrichaceae have little in
mosses contains the two genera, Andreaea and Andreaeo- common.
bryum. These two mosses, although similar in gross morphol- Molecular phylogenetic analyses, as well as peristome struc-
ogy, differ in several seemingly fundamental morphological ture and development, support the interpretation that Buxbau-
features including development of a seta (absent in Andreaea mia and Diphyscium are intermediate between nematodontous
and arthodontous mosses. The two families are paraphyletic to

the rest of the arthrodontous clade, forming a bridge from taxa
with nematodontous peristomes (Fig. 12). Newton et al. (2000)
resolved Buxbaumia and Tetraphis as a monophyletic group
between Polytrichaceae and Diphyscium, but their topology
was otherwise similar in the intermediate placement of these
taxa between nematodonts and arthrodonts. Goffinet et al.
(2001) resolved Buxbaumia in an unsupported clade with Te-
traphidaceae and Oedopodium (in contrast to Fig. 12), but their
analysis resolved Diphyscium as sister to arthrodontous taxa
(as in Fig. 12). With more extensive taxon sampling within
Diphysiaceae, Magombo (2003) confirmed that the family is
monophyletic and corroborated its phylogenetic position be-
tween Buxbaumiacae and arthrodontous mosses. A placement
of Buxbaumia near the basal node of the Diphyscium-arthro-
dontous clade is supported by a shared deletion of approxi-
mately 200 nucleotides in the rps4 gene (Goffinet et al., 2001;
Cox et al., 2004). The deletion is absent in nematodonts (Po-
lytrichaceae, Tetraphidaceae) as well as in Oedipodium,
Sphagnum, Takakia, Andreaea, and Andreaeobryum.
Relationships within arthrodontous mosses are less well es-
tablished, and internal branches down the backbone of the ar-
throdonts in the eight-gene tree are notably short (Fig. 12). If
these short branches reflect time rather than a shift in substi-
tution rate, the shape of the tree indicates a rapid radiation of
arthrodontous mosses. Arthrodontous taxa are resolved in two
lineages. One includes the genus Timmia plus the Encalypta-
ceae and Funariaceae. Timmiaceae are a small family of
Northern Hemisphere mosses with one genus and fewer than
10 species (Brassard, 1979). The peristome of Timmia is
unique, although unambiguously arthrodontous. It has typical
diplolepideous exostome teeth, but the endostome consists of
a membranous basal membrane from which approximately 64
cilia arise (Fig. 13L); normal endostome segments are not Fig. 14. Sporophyte of Takakia ceratophylla (Mitt.) Grolle. Scanning
electron micrograph showing cylindrical capsule with a spiraled suture (S)
formed. There has been much speculation about homology of and persistent terminal calyptra (C). At the base of the sporophyte is the
cilia in the Timmia endostome (Vitt, 1984) but little consensus. remnant of the archegonium and spirally inserted terete phyllids (leaves) of
In the eight-gene tree in Fig. 12, Timmia is sister to a clade the gametophyte. Inset (light micrograph) shows dehisced capsule with spores
containing Encalyptaceae and Funariaceae, which is consistent adhering along the spiraled opening. Bar 5 0.2 mm.
with the topology recovered from analyses of plastid and mi-
tochondrial sequences (Beckert et al., 1999, 2001; Goffinet and
Cox, 2000; Magombo, 2003). It is possible that sequences calypta have not been investigated but might offer insights into
from the nuclear genome produce a different position for Tim- peristome evolution. Funariaceae are also diverse in peristome
mia as sister to all arthrodontous mosses (Cox and Hedderson, structure, ranging from well developed to absent. When pre-
1999; Newton et al., 2000), but support for this potential con- sent, however, they consistently have a Funaria-type diplole-
flict is weak at present. pideous-opposite morphology. The absence of peristomes in
Interpretations of peristome structure in Encalypta (Enca- some Funariaceae is generally interpreted as secondary reduc-
lyptaceae) have been controversial. Vitt (1984) suggested that tion, and phylogenetic analyses do not contradict this conclu-
the genus encompasses characteristics of nematodontous and sion. Some Funariaceae, including the ‘‘model organism’’
arthrodontous peristomes, including species with diplolepi- Physcomitrella patens (Hedw.) Bruch & Schimp., have cleis-
deous-alternate and diplolepideous-opposite types (Vitt, 1984). tocarpous capsules (i.e., no differentiated operculum or peri-
Developmental patterns in the sporophytes of and phylogenetic stome).
relationships among the peristomially diverse species of En- The eight-gene tree in Fig. 12 indicates a sister-group re-
←
Fig. 13. Anatomy and development of moss peristomes. A. Polytrichum juniperinum Hedw. Bar 5 45 mm. B. Transverse section of a young Polytrichum
capsule showing two innermost amphithecial layers of 32 cells (arrows) resulting from ‘‘extra’’ periclinal divisions early in development. Bar 5 100 mm. C.
Transverse section of an older Polytrichum capsule showing cell deformation. Bar 5 30 mm. D. Tetraphis peristome. Bar 5 350 mm. E. Buxbaumia aphylla
Hedw. peristome. Bar 5 425 mm. F. Bryum pseudotriquetrum (Hedw.) Gaertn. et al. (diplolepideous-alternate) peristome showing 16 outer exostome teeth
surrounding a keeled endostome. Bar 5 200 mm. G. Transverse section of a young Funaria hygrometrica Hedw. capsule showing symmetric anticlinal divisions
(arrows) in the inner peristomial layer (IPL). Bar 5 100 mm. H. Diphyscium peristome. Bar 5 450 mm. I. Dicranum scoparium Hedw. (haplolepideous)
peristome. Bar 5 75 mm. J. Transverse section of a young Bryum peristome showing asymmetric anticlinal divisions (arrows) in the IPL. Bar 5 100 mm. K.
Funaria peristome (diplolepideous-opposite). Bar 5 125 mm. L. Timmia peristome, viewed from the inside, showing the endostome consisting of cilia but no
segments. Bar 5 55 mm.
lationship between taxa characterized by haplolepideous and rocarps from which pleurocarps evolved (Cox and Hedderson,
diplolepideous-alternate peristomes but with weak bootstrap 1999; Cox et al., 2000; Tsubota et al., 2002). Goffinet et al.
and/or Bayesian support. This is an important relationship in (2001) and Cox et al. (2004) resolved Orthotrichacae as sister
the context of understanding peristome evolution. As noted to Splachnaceae (the dung mosses). This result has significant
before, Newton et al. (2000) found evidence of a close rela- support (i.e., .95% posterior probability) in the analyses of
tionship between Encalyptaceae and haplolepideous taxa, but Cox et al. (2004).
they nevertheless resolved the haplolepideous plus Encalyp- The traditional family Rhizogoniaceae (e.g., Brotherus,
taceae clade as sister to the diplolepideous-alternate mosses. 1924) are consistently resolved as nonmonophyletic by mo-
Within the diplolepideous-alternate clade, acrocarpous taxa lecular data (Goffinet et al., 2001). Members of Rhizogon-
form a paraphyletic grade leading to pleurocarps (Fig. 12). iaeae, however, along with the genera Orthodontium (tradi-
Pleurocarpous mosses form a strongly supported monophyletic tionally placed in Bryaceae; Brotherus, 1924; Vitt, 1984) and
group derived from an acrocarpous grade in all analyses with Aulacomnium, appear to be close to the ancestral acrocarps
sufficient sampling conducted to date (Buck et al., 2000; De from which pleurocarpous mosses originated (Cox et al., 2000;
Luna et al., 2000; Tsubota et al., 2002). De Luna et al., 2000; Goffinet et al., 2001; Tsubota et al.,
2002). These taxa are critical to questions about the origins of
Phylogenetic relationships within acrocarpous and clado- pleurocarpy, and progress is being made resolving relation-
carpous mosses—Cox and Hedderson (1999) reconstructed re- ships among taxa traditionally classified in Rhizogoniaceae (A.
lationships among acrocarpous mosses with diplolepideous-al- Newton, British Museum, Natural History, personal commu-
ternate peristomes based on nuclear 26s rDNA and plastid nication).
rps4, and trnL-trnF sequences. Their study upset many long-
established taxonomic concepts. In particular, they showed that Phylogenetic relationships among pleurocarpous moss-
the large family, Bryaceae, is phylogenetically heterogeneous. es—The pleurocarpous mosses include some 5000 species, ap-
Leptobryum, always previously classified in Bryaceae, turned proximately 50% of all mosses. Pleurocarps are diverse in
out to be in Meesiaceae, a conclusion corroborated by subse- tropical forests, although they are also well represented in
quent studies (Cox et al., 2000; Goffinet and Cox, 2000; Gof- Northern and Southern Hemisphere temperate regions. It is
finet et al., 2001). Orthodontium was removed from Bryaceae well established that the pleurocarps are monophyletic and
in favor of a placement among (largely unresolved) acrocar- evolved from acrocarpous ancestors (De Luna et al., 2000;
pous genera near the base of the pleurocarps. Most striking, Newton et al., 2000; Goffinet et al., 2001; Cox et al., 2004).
however, was their finding that even the core bryaceous gen- Pleurocarpous mosses have traditionally been classified in
era, Bryum, Brachymenium, Pohlia, Mielichhoferia, do not three orders: Hookeriales, Hypnales, and Leucodontales. There
form a monophyletic group. Pohlia and related genera (e.g., is now little question that Leucodontales, defined primarily on
Mielichhoferia, Mniobryum) are part of a clade including taxa the basis of reduced peristomes (Brotherus, 1924–1925; Vitt,
traditionally classified in Mniaceae, leaving only Bryum and 1984; Buck, 1991), are nonmonophyletic (Buck et al., 2000;
related genera to form a more restricted Bryaceae. Pohlia has De Luna et al., 2000; Tsubota et al., 2002). Relationships with-
relatively narrow, nonbordered leaves and long leaf cells; in Hypnales have been recalcitrant to phylogenetic resolution
Bryum and relatives have broader, frequently bordered leaves because of short branch lengths at the base of the hypnalean
and shorter hexagonal or rhombic leaf cells; and Mniaceae are clade (Buck et al., 2000). Shaw et al. (2003a) provided mo-
characterized by broader still, sometimes elliptical leaves gen- lecular evidence that Hypnales underwent a rapid radiation
erally with a strong border and isodiametric cells. The un- early in their history. Consequently, resolution of family re-
equivocal placement of Pohlia in Mniaceae could never have lationships within Hypnales is likely to require a tremendous
been predicted from morphological observations and showed amounts of nucleotide sequence data and/or comparative in-
clearly how misleading morphological patterns can be about formation about genome structure. Although relationships
phylogenetic relationships (notwithstanding many congruent among hypnalean families are largely unresolved at present,
patterns of relationship inferred from morphology and molec- some apparently monophyletic groups have been identified
ular data in the mosses). Moreover, phylogenetic insights and generic relationships within them investigated (Chiang and
gained from molecular analyses raise questions about the na- Schaal, 2000; Quandt et al., 2000; Tsubota et al., 2001a, b;
ture of large morphological transitions within monophyletic Blöcher and Capesius, 2002; Pedersen and Hedenäs, 2002;
groups such as Mniaceae. Stech et al., 2002; Vanderpoorten et al., 2002a, b). A difficult
Cladocarpous taxa have archegonia borne on lateral branch- but critical issue confronting phylogenetic analyses of generic
es, seemingly intermediate between acrocarpous and pleuro- relationships within families of hypnalean pleurocarps has
carpous architectures (La Farge-England, 1996). Diverse been the identification of well-supported monophyletic groups
groups of cladocarps include Hedwigiaceae, Orthotrichaceae, appropriate for detailed investigations.
and Rhizogoniaceae. Placement of Orthotrichaceae is also im- Buck et al. (in press) resolved ordinal relationships in the
portant in the context of interpreting basic peristome types in pleurocarps based on four genes (nuclear 26S rDNA, plastid
mosses (discussed earlier). Unfortunately, relationships of Or- rps4, trnL-trnF, and mitochondrial nad4). They found (with
thotrichaceae are still unresolved, although all studies to date strong support) that a clade including traditional Garovagli-
have indicated that the family is nested within groups char- aceae and Ptychomniaceae is sister to Hookeriales plus Hyp-
acterized by diplolepideous-alternate peristomes (Goffinet et nales. These orders are also supported as monophyletic. On
al., 2001), possibly among a group of relatively derived ac- this basis, Buck et al. (in press) reclassified the pleurocarps in
←
Fig. 15. Oedipodium griffitheanum (Dicks.) Schwaegr. habit. Drawing by C. Zartman.
two superorders, the Pychomnianae and Hypnanae. Ptychom- The teeth of haplolepideous peristomes develop in positions
nianae include the single order, Ptychomniales (with one fam- that would be opposite exostome teeth were the latter formed
ily: Ptychomniaceae), whereas Hypnanae encompass Hooker- (clearly shown in Vitt, 1981). Thus, in terms of development,
iales and Hypnales. They also resolved familial and generic the single row of teeth in haplolepideous peristomes are ho-
relationships within the Hookeriales, recognizing seven fami- mologous with the opposite endostome segments of the Fu-
lies, and reconstructed the evolution of morphological char- naria-type. Like Funaria-type endostomes, haplolepideous
acters on the basis of their results. Obtaining phylogenetic res- peristomes also lack a basal membrane and are relatively mas-
olution within Hookeriales proved less problematic than in sive. Groups characterized by haplolepideous and diplolepi-
Hypnales because Hookeriales do not appear to have under- deous-alternate peristomes appear to be sister groups, implying
gone the sort of rapid radiation that characterizes Hypnales that asymmetric anticlinal cell divisions in the IPL during peri-
(Shaw et al., 2003b). Branch lengths along Hookerialean back- stome development may be a synapomorphy for that clade.
bone are substantially longer than in Hypnales. Diplolepideous-alternate peristomes have additional anticlinal
IPL walls that become offset during development relative to
Inferences about morphological evolution in mosses from those in the PPL. This pattern is likely a synapomorphy for
molecular phylogenies—Cox et al. (2004) conservatively stat- the clade characterized by such peristomes.
ed the morphological implications of their phylogenetic results
for the mosses. These inferences are briefly summarized here. CONSPECTUS
Taxa near the base of the moss tree have the capsule ele-
vated on a gametophytic pseudopodium rather than on a spo- We are currently in a period of exponential change in our
rophytic seta (i.e., Sphagnum and Andreaea), but the ancestral understanding of bryophyte phylogeny. Relationships among
condition in mosses is ambiguous because Takakia and An- the major moss clades are relatively well resolved in compar-
dreaeobryum have a seta. Cox et al. (2004) concluded that the ison to the liverworts and hornworts. However, current work
pseudopodium evolved independently in Sphagnum and An- on all three groups is progressing at such a fast pace that even
dreaea. Because stomata are absent in Takakia, Andreaea, and by the time this review is in print, new discoveries and insights
Andreaeobryum and those of Sphagnum are nonfunctional, are likely.
Cox et al. also concluded that stomata-like structures in Sphag- The molecular hypothesis presented here on hornworts (Fig.
num may not be homologous with stomata of more derived 1) is a critical first step toward a modern phylogenetic under-
mosses or to those of tracheophytes and hornworts. Their pres- standing for the group. Comprehensive analyses using genes
ence in some hornworts indicates, to the contrary, that stomata from all three genomes in combination with morphological
may be homologous in mosses and hornworts, which implies data, with sampling from all 12 genera of hornworts, are re-
multiple losses in mosses. Alternatively, stomata could have quired to verify the novel relationships described earlier. Sys-
been lost once in the early evolution of mosses and regained tematic studies of hornworts have lagged so far behind those
in class Bryopsida. of other land plants that any molecular analysis must first be-
The acrocarpous habit is clearly pleisiotypic in peristomate gin with a clear delineation of morphological characteristics in
mosses, but acrocarps are a paraphyletic group within which the specimens/species that are examined. Worldwide collecting
pleurocarps are nested. Although resolution among cladocar- and basic taxonomic evaluations are essential. With their
pous taxa is poor, it appears that cladocarpy evolved several unique adaptations to land, including basal elongation of the
times. It may be that hypnalean pleurocarps evolved from a sporophyte and internalization of vulnerable tissues, hornworts
cladocarpous ancestor, but additional resolution among acro- will continue to provide essential information about early land
carps near the origin of pleurocarps is needed. plant evolution.
Absence of a peristome in Oedopodium makes the phylo- A general understanding of liverwort relationships has
genetic node at which peristomes originated ambiguous, and emerged from recent molecular studies, and this has led to
it is not clear whether nematodontous peristomes of the Po- major reinterpretations of evolutionary changes in the group.
lytrichum-type evolved independently of arthrodontous peri- For example, isophylly in leafy liverworts and conducting tis-
stomes. The unique anticlinal divisions in the IPL of Polytri- sue in simple thalloid taxa are now clearly seen as derived,
chaceae, leading to twice the ‘‘normal’’ number of cells in this not ancestral. However, critical unanswered questions in liv-
layer, may be an apomorphy for that clade. These divisions erwort phylogeny remain, and these include the position of
appear to be characteristic not only of Polytrichum, but also Haplomitrium (and Treubia), the placement of Pleurozia
Atrichum and Pogonatum, also in Polytrichaceae (Shaw, L. among the simple thalloid vs. leafy clades, and the precise
Anderson, Duke University, and B. Mishler, University of Cal- positions of Sphaerocarpos and Lunularia within the complex
ifornia, Berkeley, unpublished data). These divisions do not thalloid lineage. Relationships among simple thalloid taxa, es-
occur in the developing sporophyte of Sphagnum or Andreaea pecially Pellia, Phyllothallia, Calycularia, and Cavicularia
(Shaw, unpublished data). Phylogenetic considerations lend (the last sister to Blasia, Renzaglia, 1982) have not been re-
support to the hypothesis of Vitt (1984) that the Funaria-type solved using multilocus studies to date and will require addi-
arthrodontous peristome with opposite exostome and endo- tional sequences and more taxon sampling. Most striking is
stome teeth is primitive in arthrodonts. Cox et al. (2004) noted the lack of representation of critical genera and families in any
that although Timmia lacks endostome segments, the most par- one study and the need for a concerted, collaborative effort to
simonious interpretation of its endostome (which consists of a obtain and share specimens of poorly known taxa.
basal membrane and cilia), given its phylogenetic position, is Still outstanding questions with regard to moss phylogeny
that it conforms to the opposite type. Developmental studies include the relationship among Takakia, Sphagnum, and An-
of peristomial layers in Timmia are sorely needed. Of critical dreaea (do they form a monophyletic group sister to all other
importance is whether or not anticlinal walls in the PPL and mosses?), the origin and evolution of the major peristome
IPL are offset. types, and the nature of the acrocarpous ancestors of pleuro-
carpous taxa. Fundamental morphological data on Takakia, in- BOISSELIER-DUBAYLE, M.-C., J. LAMBOURDIERE, AND H. BISHLER. 2002.
cluding embryology, sporophyte development, and apical Molecular phylogenies support multiple morphological reductions in the
liverwort subclass Marchantiidae (Bryophyta). Molecular Phylogenetics
growth are necessary to identify structural changes within and Evolution 24: 66–77.
mosses. Additional work on developmental anatomy of peri- BOPP, M., AND I. CAPESIUS. 1996. New aspects of bryophyte taxonomy pro-
stomes is also needed in conjunction with ongoing phyloge- vided by a molecular approach. Botanica Acta 106: 368–372.
netic work. Development of peristomial cell layers in Oedi- BOPP, M., AND I. CAPESIUS. 1998. A molecular approach to bryophyte sys-
podium (which lacks a peristome) is critical, as is the devel- tematics. In J. W. Bates, N. W. Ashton, and J. G. Duckett [eds.], Bryology
opment of the unique Timmia peristome. Resolution of family- for the twenty-first century, 79–88. Maney and the British Bryological
Society, Leeds, UK.
level relationships in mosses, especially in closely related BOWER, F. O. 1935. Primitive land plants. Macmillan and Company, London,
pleurocarps, will require expanded data sets based on not just UK.
one or two genes, nor even five, but probably 15 to 20. Avail- BRASSARD, G. R. 1979. The moss genus Timmia. I. Introduction, and revision
ability of primers for such multilocus analyses are probably of T. norvegica and allied taxa. Lindbergia 5: 39–53.
not far off, given the intensive genomic work currently un- BROTHERUS, V. A. 1924–1925. Musci (Laubmoose). III. Unterklasse Bryales:
derway on a wide diversity of land plants, including mosses. II. Spezieller Teil. In A. Engler [ed.], Die Natürlichen Pflanzenfamilien,
277–700. Engelmann, Leipzig, Germany.
Molecular technologies are improving at a rate that is un- BROWN, R. C., AND B. L. LEMMON. 1988. Sporogenesis in bryophytes. Ad-
predictable and incomprehensible; we speculate that within the vances in Bryology 3: 159–223.
next decade most of the phylogenetic questions raised in this BROWN, R. C., AND B. L. LEMMON. 1990. Monoplastidic cell division in
review will be resolved. In addition to sequences from mul- lower land plants. American Journal of Botany 77: 559–571.
tiple genes, complete organellar genomes will soon be avail- BROWN, R. C., B. L. LEMMON, AND K. S. RENZAGLIA. 1986. Sporocytic
able for representatives of each major lineage within bryo- control of spore wall patterns in liverworts. American Journal of Botany
73: 593–596.
phytes. Structural features of genomes, including intron pres- BRUCH, P., W. P. SCHIMPER, AND T. GÜMBEL. 1851–1855. Bryologia Europaea
ence, and gene order and deletions, likewise will continue to seu Generum Muscorum Europaeorum Monographice Illustrata.
provide informative phylogenetic evidence. Further under- Schweizerbart, Stuttgart, Germany.
standing of the existence and expression of developmental BUCK, W. R. 1991. The basis for familial classification of pleurocarpous
genes, especially homeobox and MADS-box genes, in all three mosses. Advances in Bryology 4: 169–185.
groups of bryophytes will provide clues to the evolution of BUCK, W. R., C. J. COX, A. J. SHAW, AND B. GOFFINET. In press. Ordinal
relationships of pleurocarpous mosses, with special emphasis on the
structural complexity within these plants and evolutionary re- Hookeriales. Systematics and Biodiversity.
lationships to more complicated organ systems of tracheo- BUCK, W. R., AND B. GOFFINET. 2000. Morphology and classification of
phytes. Comparative morphological/ultrastructural studies of mosses. In A. J. Shaw and B. Goffinet [eds.], Bryophyte biology, 71–
living and fossil taxa are required to fill in the gaps in knowl- 123. Cambridge University Press, New York, New York, USA.
edge as well as to fully comprehend structural changes. In BUCK, W. R., B. GOFFINET, AND A. J. SHAW. 2000. Testing morphological
addition to perfecting data collection methodologies, the pri- concepts of orders of pleurocarpous mosses (Bryophyta) using phylo-
mary challenge that lies ahead is in developing methods of genetic reconstructions based on trnL-trnF and rps4 sequences. Molec-
ular Phylogenetics and Evolution 16: 180–198.
analyzing and combining the large and diverse data sets that BURR, F. A. 1970. Phylogenetic transitions in the chloroplast numbers of the
are rapidly materializing. Only then will the intricate details Anthocerotales. I. The number and ultrastructure of the mature plastids.
of early land plant interrelationships be clearly illuminated. American Journal of Botany 57: 97–110.
CAMPBELL, D. H. 1895. The structure and development of mosses and ferns
LITERATURE CITED (Archegoniatae). Macmillan and Co., London, UK.
CAMPBELL, D. H. 1917. Growth of isolated sporophytes of Anthoceros. Pro-
AHONEN, I., J. MUONA, AND S. PIIPPO. 2003. Inferring the phylogeny of the ceedings of the National Academy of Science, USA 3: 494–496.
Lejeuneaceae (Jungermanniopsida): a first appraisal of molecular data. CAMPBELL, D. H. 1924. A remarkable development of the sporophyte of
Bryologist 106: 297–308. Anthoceros fusiformis Aust. Annals of Botany 38: 473–483.
BARTHOLOMEW-BEGAN, S. E. 1990. Classification of the Haplomitriales and CARGILL, D. C., K. S. RENZAGLIA, J. C. VILLAREAL, AND R. J. DUFF. In
Metzgeriales into the subclass Metzgeriidae, subclass nov. (Hepatophyta, press. Generic concepts within hornworts (Phylum Anthocerotophyta).
Jungermanniopsida). Phytologia 69: 464–466. Australian Journal of Botany.
BARTHOLOMEW-BEGAN, S. E. 1991. A morphogenetic re-evaluation of Hap- CHAMPAGNE, C. E. M., AND N. W. ASHTON. 2001. Ancestry of KNOX genes
lomitrium Nees (Hepatophyta). Bryophytorum Bibliotheca 41: 1–297. revealed by bryophyte (Physcomitrella patens) homologs. New Phytol-
BARTLETT, E. M. 1928. A comparative study of the development of the spo- ogist 150: 23–36.
rophyte in the Anthocerotae, with special reference to the genus Antho- CHIANG, T.-Y., AND B. A. SCHAAL. 2000. The internal transcribed spacer 2
ceros. Annals of Botany 42: 409–430. region of the nuclear ribosomal DNA and the phylogeny of the moss
BECKERT, S., H. MUHLE, D. PRUCHNER, AND V. KNOOP. 2001. The mito- family Hylocomiaceae. Plant Systematics and Evolution 224: 127–137.
chondrial nad2 gene as a novel marker locus for phylogenetic analysis CLARKE, G. C. S., AND J. G. DUCKETT [EDS.]. 1979. Bryophyte systematics.
of early land plants: a comparative analysis in mosses. Molecular Phy- The Systematics Association Special Volume 14. Academic Press, New
logenetics and Evolution 18: 117–126. York, New York, USA.
BECKERT, S., S. STEINHAUSER, H. MUHLE, AND V. KNOOP. 1999. A molecular COX, C. J., B. GOFFINET, A. E. NEWTON, A. J. SHAW, AND T. A. J. HEDDER-
phylogeny of the bryophytes based on nucleotide sequences of the mi- SON. 2000. Phylogenetic relationships among the diplolepideous-alter-
tochondrial nad5 gene. Plant Systematics and Evolution 218: 179–192. nate mosses (Bryidae) inferred from nuclear and chloroplast DNA se-
BIERHORST, D. W. 1971. Morphology of vascular plants. Macmillan Co., New quences. Bryologist 103: 224–240.
York, New York, USA. COX, C. J., B. GOFFINET, A. J. SHAW, AND S. BOLES. 2004. Phylogenetic
BISCHLER, H. 1998. Systematics and evolution of the genera of the Mar- relationships among the mosses based on heterogeneous Bayesian anal-
chantiales. Bryophytorum Bibliotheca 51: 1–201. ysis of multiple genes from multiple genomic compartments. Systematic
BLÖCHER, R., AND I. CAPESIUS. 2002. The systematic position of the Hy- Botany 29: 234–250.
popterygiaceae (Bryopsida) inferred from rps4 gene sequences. Cryp- COX, C. J., AND T. A. J. HEDDERSON. 1999. Phylogenetic relationships among
togamie, Bryologie 23: 191–207. the ciliate arthrodontous mosses: evidence from chloroplast and nuclear
BLOMQUIST, H. L., AND L. L. ROBERTSON. 1941. The development of the DNA sequences. Plant Systematics and Evolution 215: 119–139.
peristome in Aulacomnium heterostichum. Bulletin of the Torrey Botan- CRANDALL-STOTLER, B. 1976. The apical cell and early development of Pleu-
ical Club 68: 569–584. rozia purpurea Lindb. Lindbergia 32: 197–208.
CRANDALL-STOTLER, B. 1981. Morphology/anatomy of hepatics and antho- GOFFINET, B., C. J. COX, L. E. ANDERSON, AND B. D. MISHLER. 1999. Peri-
cerotes. Advances in Bryology 1: 315–398. stome development in the Orthotrichaceae. Bryologist 102: 581–594.
CRANDALL-STOTLER, B., AND R. E. STOTLER. 2000. Morphology and clas- GOFFINET, B., C. J. COX, A. J. SHAW, AND T. A. J. HEDDERSON. 2001. The
sification of the Marchantiophyta. In A. J. Shaw and B. Goffinet [eds.], bryophyta (mosses): systematic and evolutionary inferences from an rps4
Bryophyte biology, 21–70. Cambridge University Press, Cambridge, UK. gene (cpDNA) phylogeny. Annals of Botany 87: 191–208.
CROSBY, M. R., R. E. MAGILL, B. ALLEN, AND S. HE. 2000. A checklist of GRADSTEIN, S. R. 1979. The genera of the Lejeuneaceae: past and present.
the mosses. Missouri Botanical Garden, St. Louis, Missouri, USA, web- In G. C. S. Clarke and J. G. Duckett [eds.], Bryophyte systematics, 83–
site: www.mobot.org/MOBOT/tropicos/most/checklist.shtml. 107. Academic Press, New York, New York, USA.
CRUM, H. A. 2001. Structural diversity of bryophytes. University of Michigan GRADSTEIN, S. R. 1994. Lejeuneaceae: Ptychantheae, Brachiolejeuneae. Flo-
Herbarium, Ann Arbor, Michigan, USA. ra Neotropica Monograph 62: 1–225.
DAVIS, E. C. In press. A molecular phylogeny of leafy liverworts (Junger- GRADSTEIN, S. R. 1997. The taxonomic diversity of epiphyllous bryophytes.
manniidae: Marchantiophyta). In B. Goffinet, V. C. Hollowell, and R. E. Abstracta Botanica 21: 15–19.
Magill [eds.], Molecular systematics of bryophytes: progress, problems GRAHAM, L. E. 1993. Origin of land plants. John Wiley and Sons, Inc., New
& perspectives. Monographs in systematic botany from the Missouri Bo- York, USA.
tanical Garden, St. Louis, Missouri, USA. GRAHAM, L. E., AND L. W. WILCOX. 2000. The origin of alternation of gen-
DE LUNA, E., W. R. BUCK, H. AKIYAMA, T. ARIKAWA, H. TSUBOTA, D. erations in land plants: a focus on matrotrophy and hexose transport.
GONZÁLEZ, A. E. NEWTON, AND A. J. SHAW. 2000. Ordinal phylogeny Philosophical Transactions of the Royal Society, London, B, Biological
within the hypnobryalean pleurocarpous mosses inferred from cladistic Sciences 355: 757–767.
analyses of three chloroplast DNA sequence data sets: trnL-F, rps4, and HANSEN, D., T. J. ANDREWS, AND M. R. BADGER. 2002. Variability of the
rbcL. Bryologist 103: 242–256. pyrenoid-based CO2 concentrating mechanism in hornworts. Functional
DUCKETT, J. G., AND R. LIGRONE. 1995. The formation of catenate foliar Plant Biology 29: 407–416.
gemmae and the origin of the oil bodies in the liverwort Odontoschisma HASEGAWA, J. 1980. Taxonomical studies on Asian Anthocerotae. III. Some
denudatum (Mart.) Dum. (Jungermanniales): a light and electron micro- Asian species of Dendroceros. Journal of the Hattori Botanical Labo-
scope study. Annals of Botany 76: 406–419. ratory 47: 287–309.
HASEGAWA, J. 1983. Taxonomical studies on Asian Anthocerotae. III. Asian
DUFF, R. J., AND D. L. NICKRENT. 1999. Phylogenetic relationships among
species of Megaceros. Journal of the Hattori Botanical Laboratory 54:
land plants using mitochondrial small-subunit rDNA sequences. Ameri-
227–240.
can Journal of Botany 86: 372–386.
HASEGAWA, J. 1988. A proposal for a new system of the Anthocerotae, with
DUFF, R. J., D. C. CARGILL, J. C. VILLARREAL, AND K. S. RENZAGLIA. In
a revision of the genera. Journal of the Hattori Botanical Laboratory 74:
press. Phylogenetic relationships of the hornworts based on rbcL se-
105–119.
quence data: novel relationships and new insights. In B. Goffinet, V. C.
HASEGAWA, J. 1994. New classification of Anthocerotae. Journal of the Hat-
Hollowell, and R. E. Magill [eds.], Molecular systematics of bryophytes:
tori Botanical Laboratory 76: 21–34.
progress, problems & perspectives. Monographs in systematic botany
HÄSSEL DE MENÉNDEZ, G. G. 1962. Estudio de los Anthocerotales y Mar-
from the Missouri Botanical Garden, St. Louis, Missouri, USA. chantiales de Argentina. Opera Lilloana 7: 1–297.
EDWARDS, S. R. 1979. Taxonomic implications of cell patterns in haplole- HÄSSEL DE MENÉNDEZ, G. G. 1986. Leiosporoceros Hässel n. gen. and Leios-
pidous moss peristomes. In G. C. S. Clarke and J. G. Duckett [eds.], porocerotaceae Hässel n. fam. of Anthocerotopsida. Journal of Bryology
Bryophyte systematics, 317–346. Systematics Association Special Vol- 14: 255–259.
ume 14. Academic Press, New York, New York, USA. HÄSSEL DE MENÉNDEZ, G. G. 1988. A proposal for a new classification of
EDWARDS, S. R. 1984. Homologies and inter-relationships of moss peri- the genera within the Anthocerotophyta. Journal of the Hattori Botanical
stomes. In R. M. Schuster [ed.], New manual of bryology, vol. 2, 658– Laboratory 64: 71–86.
695. Hattori Botanical Laboratory, Nichinan, Japan. HÉBANT, C. 1977. The conducting tissues of bryophytes. Bryophytorum Bib-
ENDLICHER, S. 1841. Enchiridion botanicum. W. Engelmann, Leipzig, Ger- liotheca 10: 1–157, 80 plates, J. Cramer, Vaduz, Lichtenstein.
many. HEDDERSON, T. A., R. L. CHAPMAN, AND C. J. COX. 1998. Bryophytes and
EVANS, A. W., AND H. D. HOOKER, JR. 1913. Development of the peristome the origins and diversification of land plants: new evidence from mole-
in Ceratodon purpureus. Bulletin of the Torrey Botanical Club 40: 97– cules. In J. W. Bates, N. W. Ashton, and J. G. Duckett [eds.], Bryology
109. for the twenty-first century, 65–77. Maney Publishing, Leeds, UK.
FLEISCHER, M. 1923. Die Moose der Flora von Buitenzorg 4: XI–XXIII. E. HEDDERSON, T. A., R. L. CHAPMAN, AND W. L. ROOTES. 1996. Phylogenetic
J. Brill, Leiden, Netherlands. relationships of bryophytes inferred from nuclear encoded rRNA gene
FORREST, L. L., AND B. J. CRANDALL-STOTLER. In press. A phylogeny of sequences. Plant Systematics and Evolution 200: 213–224.
the simple thalloid liverworts (Jungermanniopsida, subclass Metzgeri- HEDWIG, J. 1801. Species Muscorum frondosorum Descriptae et Tabulis
idae) as inferred from five chloroplast genes. In B. Goffinet, V. C. Hol- Aeneid 1xxvii Colratis Illustratae. Barthii, Leipzig, Germany.
lowell, and R. E. Magill [eds.], Molecular systematics of bryophytes: HENSCHEL, K., R. KUFUJI, M. HASEBE, H. SAEDLER, T. MÜNSTER, AND G.
progress, problems & perspectives. Monographs in systematic botany THEISSEN. 2002. Two ancient classes of MIKC-type MADS-box genes
from the Missouri Botanical Garden, St. Louis, Missouri, USA. are present in the moss Physcomitrella patens. Molecular Biology and
GARBARY, D. J., AND K. S. RENZAGLIA. 1998. Bryophyte phylogeny and the Evolution 19: 801–814.
evolution of land plants: evidence from development and ultrastructure. HE-NYGRÉN, X., I. AHONEN, A. JUSLEN, D. GLENNY, AND S. PIIPPO. In press.
In J. W. Bates, N. W. Ashton, and J. G. Duckett [eds.], Bryology for the Dichotomous evolution in leafy liverworts. In B. Goffinet, V. C. Hollow-
twenty-first century, 45–63. Maney and British Bryological Society, ell, and R. E. Magill [eds.], Molecular systematics of bryophytes: pro-
Leeds, UK. gress, problems & perspectives. Monographs in systematic botany from
GARBARY, D. J., K. S. RENZAGLIA, AND J. G. DUCKETT. 1993. The phylogeny the Missouri Botanical Garden, St. Louis, Missouri, USA.
of land plants: a cladistic analysis based on male gametogenesis. Plant HIMI, S., R. SANO, T. NISHIYAMA, T. TANAHASHI, M. KATO, K. UEDA, AND
Systematics and Evolution 188: 237–269. M. HASEBE. 2001. Evolution of MADS-box gene induction by FLO/
GLENNY, D. 1998. A revised checklist of New Zealand liverworts and horn- LFY genes. Journal of Molecular Evolution 53: 387–393.
worts. Tuhinga 10: 11–149. HOFMEISTER, W. 1851. Vergleichende Untersuchungen der Keimung, Entfal-
GOEBEL, K. 1905. Organography of plants, especially of the archegoniatae tung und Fruchtbildung hoherer Kryptogamen. F. Hofmeister, Liepzig,
and spermatophyta. II. Special organography. Clarendon Press, Oxford, Germany.
UK. HOHAM, R. W., T. A. BONOME, D. W. MARTIN, AND J. H. LEEBENS-MACK.
GOFFINET, B. 2000. Origin and phylogenetic relationships of bryophytes. In 2002. A combined 18S rDNA and rbcL phylogenetic analysis of Chlo-
J. Shaw and B. Goffinet [eds.], The biology of bryophytes, 124–149. romonas and Chlamydomonas (Chlorophyceae, Volvocales) emphasizing
Cambridge University Press, Cambridge, UK. snow and other cold-temperature habitats. Journal of Phycology 38:
GOFFINET, B., AND C. J. COX. 2000. Phylogenetic relationships among basal- 1051–1064.
most arthrodontous mosses with special emphasis on the evolutionary HOHE, A., S. A. RENSING, M. MILDNER, D. LANG, AND R. RESKI. 2002. Day
significance of the Funariineae. Bryologist 103: 212–223. length and temperature strongly influence sexual reproduction and ex-
pression of a novel MADS-box gene in the moss Physcomitrella patens. NEWTON, A. E., C. J. COX, J. G. DUCKETT, J. WHEELER, B. GOFFINET, B.
Plant Biology 4: 595–202. D. MISHLER, AND T. A. J. HEDDERSON. 2000. Evolution of the major
HYVÖNEN, J., T. A. HEDDERSON, G. L. SMITH MERRILL, J. G. GIBBINS, AND moss lineages. Bryologist 103: 187–211.
S. KOSKINEN. 1998. On phylogeny of the Polytrichales. Bryologist 101: NICKRENT, D. L., C. L. PARKINSON, J. D. PALMER, AND R. J. DUFF. 2000.
489–504. Multigene phylogeny of land plants with special reference to bryophytes
HYVÖNEN, J., AND S. PIIPPO. 1993. Cladistic analysis of the hornworts (An- and the earliest land plants. Molecular Biology and Evolution 17: 1885–
thocerotophyta). Journal of the Hattori Botanical Laboratory 74: 105– 1895.
119. NIKLAS, K. J. 1997. The evolutionary biology of plants. University of Chi-
KATO, M., AND R. IMAICHI. 1997. Morphological diversity and evolution of cago Press, Chicago, Illinois, USA.
vegetative organs in pteridophytes. In K. Iwatsuki and P. H. Raven [eds.], NISHIYAMA, T., AND M. KATO. 1999. Molecular phylogenetic analysis among
Evolution and diversification of plants, 27–43. Springer-Verlag, New bryophytes and tracheophytes based on combined data of plastid coded
York, New York, USA. genes and the 18S rRNA gene. Molecular Biology and Evolution 16:
KATOH, K., H. HORI, AND S. OSAWA. 1983. The nucleotide sequences of 5S 1027–1036.
ribosomal RNAs from four Bryophyta-species. Nucleic Acids Research NISHIYAMA, T., T. FUJITA, T. SHIN-I, M. SEKI, H. NISHIDE, I. UCHIYAMA, A.
11: 5671–5674. KAMIYA, P. CARNINCI, Y. HAYASHIZAKI, K. SHINOZAKI, Y. KOHARA,
KELCH, D. G., B. D. MISHLER, AND A. DRISKELL. In press. Inferring phy- AND H. HASEBE. 2003. Comparative genomics of Physcomitrella patens
logeny using genomic characters: a case study using land plant plas- gametophytic transcriptome and Arabidopsis thaliana: implication for
tomes. In B. Goffinet, V. C. Hollowell, and R. E. Magill [eds.], Molecular land plant evolution. Proceedings of the National Academy of Science,
systematics of bryophytes: progress, problems & perspectives. Mono- USA 100: 8007–8012.
graphs in systematic botany from the Missouri Botanical Garden, St. NISHIYAMA, T., P. G. WOLF, M. KUGHITA, R. B. SINCLAIR, M. SUGITA, C.
Louis, Missouri, USA. SUGIURA, T. WAKASUGI, K. YAMADA, K. YOSHINAGA, K. YAMAGUCHI,
KENRICK, P., AND P. R. CRANE. 1997. The origin and early diversification of K. EUDA, AND M. HASEBE. In press. Chloroplast phylogeny indicates
land plants: a cladistic study. Smithsonian Institution Press, Washington that bryophytes are monophyletic. Molecular Biology and Evolution.
D.C., USA. NOGUCHI, A. 1988. Illustrated moss flora of Japan, part 2. Hattori Botanical
KINDBERG, N. C. 1897. Genera of European and North American Bryineae Laboratory, Nichinan, Japan.
(Mosses) synoptically disposed. Bonniers Boktryckeri Aktiebolag, Gö- NOZAKI, N., K. ONISHI, AND E. MORITA. 2002. Differences in pyrenoid mor-
teborg, Sweden. phology are correlated with differences in the rbcL genes of members
KOBIYAMA, Y., AND B. J. CRANDALL-STOTLER. 1999. Studies of specialized of the chloromonas lineage (Volvocales, Chlorphyceae). Journal of Mo-
pitted parenchyma cells of the liverwort Conocephalum Hill and their lecular Evolution 55: 414–430.
phylogenetic implications. International Journal of Plant Sciences 160: PARIHAR, N. S. 1965. An introduction to Embryophyta, vol. 1, Bryophyta,
351–370. 5th ed. Central Book Dept., Allahabad, India.
KROGAN, N. T., AND N. W. ASHTON. 2000. Ancestry of plant MADS-box PASS, J. M., AND K. S. RENZAGLIA. 1995. Comparative microanatomy of the
genes revealed by bryophyte (Physcomitrella patens) homologues. New locomotory apparatus of Conocephalum conicum (Hepaticae, Conoce-
Phytologist 147: 505–517. phalaceae). Fragmenta Floristica et Geobotanica 40: 365–377.
LA FARGE-ENGLAND, C. 1996. Growth form, branching pattern, and peri- PEDERSEN, N., AND L. HEDENÄS. 2002. Phylogeny of the Plagiotheciaceae
chaetial position in mosses: cladocarpy and pleurocarpy redefined. Bry- based on molecular and morphological evidence. Bryologist 105: 310–
ologist 99: 170–186. 324.
LANG, W. H. 1907. On the sporangium of Notothylas. Annals of Botany 21: PHILIBERT, H. 1884–1902. De l’importance du péristome pour les affinities
201–210. naturelles des mousses. Revue Bryologique 11: 49–52, 65–72 (1884).
LEITGEB, H. 1874–1881. Untersuchungen über die Lebermoose, I–V. I. Blasia Études sur le péristome. Revue Bryologique 11: 80–87 (1884); 12: 67–
pusilla, 1874; II. Die Foliosen Jungermannieen, 1877; IV. Die Riccieen, 77, 81–85 (1885); 13: 17–26, 81–86 (1886); 14: 9–11, 81–90 (1887);
1879; V. Die Anthoceroteen, 1879; VI. Die Marchanteen, 1881, vols. I– 15: 6–12, 24–28, 37–44, 50–60, 65–69, 90–93 (1888); 16: 1–9, 39–44,
III, Jena: O. Deistungs Buchhandlungs, vols. IV–VI. Leuschner and Lu- 67–77 (1889); 17: 8–12, 25–29, 39–42 (1890); 23: 36–38, 41–56 (1896);
bensky, Graz, Austria. 28: 56–59, 127–130 (1901); 29: 10–13 (1902).
LEWIS, L., B. D. MISHLER, AND R. VILGALYS. 1997. Phylogenetic relation- PROSKAUER, J. 1951. Studies on Anthocerotales. III. The genera Anthoceros
ships of the liverworts (Hepaticae), a basal embryophyte lineage, inferred and Phaeoceros. Bulletin of the Torrey Botanical Club 78: 331–349.
from nucleotide sequence data of the chloroplast gene rbcL. Molecular PROSKAUER, J. 1953. Studies on Anthocerotales. IV. Bulletin of the Torrey
Phylogenetics and Evolution 7: 377–393. Botanical Club 80: 65–75.
LIGRONE, R., J. G. DUCKETT, AND K. S. RENZAGLIA. 2000. Conducting tis- PROSKAUER, J. 1960. Studies on Anthocerotales. VI. On spiral thickenings in
sues and phyletic relationships of bryophytes. Philosophical Transac- the columella and its bearing on phylogeny. Phytomorphology 10: 1–19.
tions of the Royal Society 355: 795–813. PURI, P. 1973. Bryophytes. A broad perspective. Atma Ram & Sons, Dehli,
LINNAEUS, C. 1753. Species plantarum. Laurentii Salvii, Stockholm, Sweden. India.
MAGOMBO, Z. L. K. 2003. The phylogeny of basal peristomate mosses: ev- QIU, Y.-L., Y. CHO, J. C. COX, AND J. D. PALMER. 1998. The gain of three
idence from cpDNA, and implications for peristome evolution. System- mitochondrial introns identifies liverworts as the earliest land plants. Na-
atic Botany 28: 24–38. ture 394: 671–674.
MAHÚ, M. 1979. Familias y géneros de musgos Chilenos. Bryologist 82: 513– QUANDT, D., R. S. TANGNEY, J.-P. FRAHM, AND W. FREY. 2000. A molecular
524. contribution for understanding the Lembophyllaceae (Bryopsida) based
MALEK, O., K. LÄTTIG, R. HIESEL, A. BRENNICKE, AND V. KNOOP. 1996. on noncoding chloroplast regions (cpDNA) and ITS2 (nrDNA) sequence
RNA editing in bryophytes and a molecular phylogeny of land plants. data. Journal of the Hattori Botanical Laboratory 89: 71–92.
European Molecular Biology Organization 15: 1403–1411. RENZAGLIA, K. S. 1978. A comparative morphology and developmental anat-
MISHLER, B. D., AND S. P. CHURCHILL. 1984. A cladistic approach to the omy of the Anthocerotophyta. Journal of the Hattori Botanical Labo-
phylogeny of the ‘‘bryophytes.’’ Brittonia 36: 406–424. ratory 44: 31–90.
MISHLER, B. D., AND E. DE LUNA. 1991. The use of ontogenetic data in RENZAGLIA, K. S. 1982. A comparative developmental investigation of the
phylogenetic analyses of mosses. Advances in Bryology 4: 121–167. gametophyte generation in the Metzgeriales (Hepatophyta). Bryophythe-
MISHLER, B. D., L. A. LEWIS, M. A. BUCHHEIM, K. S. RENZAGLIA, D. J. ca Bibliotheca 24: 1–253.
GARBARY, C. F. DELWICHE, F. W. ZECHMAN, T. S. KANTZ, AND R. L. RENZAGLIA, K. S., R. C. BROWN, B. E. LEMMON, J. G. DUCKETT, AND R.
CHAPMAN. 1994. Phylogenetic relationships of the ‘‘green algae’’ and LIGRONE. 1994. The occurrence and phylogenetic significance of mon-
‘‘bryophytes.’’ Annals of the Missouri Botanical Garden 81: 451–483. oplastidic meiosis in liverworts. Canadian Journal of Botany 72: 65–72.
MITTEN, W. 1859. Musci Indiae Orientalis. An enumeration of the mosses of RENZAGLIA, K. S., AND J. G. DUCKETT. 1987. Spermatogenesis of Blasia
the East Indies. Journal of the Proceedings of the Linneaen Society (Lon- pusilla, from antheridial initial through mature spermatozoid. Bryologist
don), Supplement Botany 1: 1–171. 90: 419–449.
MURRAY, B. M. 1988. Systematics of the Andreaopsida (Bryophyta): two RENZAGLIA, K. S., R. J. DUFF, D. L. NICKRENT, AND D. GARBARY. 2000.
orders with links to Takakia. Beiheft zur Nova Hedwigia 90: 289–336. Vegetative and reproductive innovations of early land plants: implica-
tions for a unified phylogeny. Transactions of the Royal Society, London SMITH, E. C., AND H. GRIFFITHS. 2000. The role of carbonic anhydrase in
355: 769–793. photosynthesis and the activity of the carbon-concentrating-mechanism
RENZAGLIA, K. S., AND D. J. GARBARY. 2001. Motile male gametes of land in bryophytes of the class Anthocerotae. New Phytologist 145: 29–37.
plants: diversity, development and evolution. Critical Reviews in Plant SMITH, G. M. 1955. Cryptogamic botany, vol. II, Bryophytes and pterido-
Sciences 20: 107–213. phytes. McGraw-Hill Book Co., New York, New York, USA.
RENZAGLIA, K. S., K. D. MCFARLAND, AND D. K. SMITH. 1997. Anatomy SOLTIS, P. S., D. E. SOLTIS, P. G. WOLF, D. L. NICKRENT, S. M. CHAW, AND
and ultrastructure of the sporophyte of Takakia ceratophylla (Bryophyta). R. L. CHAPMAN. 1999. The phylogeny of land plants inferred from 18S
American Journal of Botany 84: 1337–1350. rDNA sequences: pushing the limits of rDNA signal? Molecular Biology
RENZAGLIA, K. S., AND K. C. VAUGHN. 2000. Anatomy, development and and Evolution 16: 1774–1784.
classification of hornworts. In J. Shaw and B. Goffinet [eds.], The biol- STECH, M., AND W. FREY. 2001. CpDNA—relationship and classification of
ogy of bryophytes, 1–35. Cambridge University Press, Cambridge, UK. the Jungermanniopsida (Hepaticophytina, Bryophyta). Nova Hedwigia
RICHARDSON, D. H. S. 1981. The biology of bryophytes. Blackwell Scientific 72: 45–58.
Publications, Oxford, UK. STECH, M., J.-P. FRAHM, H. H. HILGER, AND W. FREY. 2000. Molecular
SAMIGULLIN, T. K., S. P. YACENTYUK, G. V. DEGTYARYEVA, K. M. VALIEHO- relationships of Treubia Goebel. (Treubiaceae, Treubiopsida) and high
ROMAN, V. K. BOBROVA, I. CAPESIUS, W. F. MARTIN, A. V. TROITSKY, taxonomic level classification of the Hepaticophytina. Nova Hedwigia
V. R. FILIN, AND A. S. ANTONOV. 2002. Paraphyly of bryophytes and 71: 195–208.
close relationship of hornworts and vascular plants inferred from analysis STECH, M., T. PFEIFFER, AND W. FREY. 2002. Molecular generic classification
of chloroplast rDNA ITS (cpITS) spacer sequences. Arctoa 11: 31–43. of the Hypopterygiaceae (Bryopsida), with the description of a new ge-
SCHLJAKOV, R. N. 1972. On the higher taxa of liverworts—class Hepaticae nus, Arbusculohypopterygium gen. nov. Studies of austral temperate rain
s. str. Botanicheskii Zhurnal 57: 496–508. forest bryophytes 10. New Zealand Journal of Botany 40: 207–221.
SCHOFIELD, W. B. 1985. Introduction to bryology. Macmillan, New York, STECH, M., D. QUANDT, AND W. FREY. 2003. Molecular circumscription of
New York, USA. the hornworts (Anthocerotophyta) based on the chloroplast DNA trnL-
SCHUSTER, R. M. 1966. The Hepaticae and Anthocerotae of North America, trnF region. Journal of Plant Research 116: 389–398.
vol. I. Columbia University Press, New York, New York, USA. TAYLOR, W. A. 1995. Spores in earliest land plants. Nature 373: 391–392.
SCHUSTER, R. M. 1972. Phylogenetic and taxonomic studies on Jungerman-
THIERS, B. M. 1993. A monograph of Pleurozia (Hepaticae; Pleuroziaceae).
nidae. Journal of the Hattori Botanical Laboratory 36: 321–405.
Bryologist 96: 517–554.
SCHUSTER, R. M. 1984. New manual of bryology, vols. 1–2. The Hattori
TSUBOTA, H., H. AKIYAMA, T. YAMAGUCHI, AND H. DEGUCHI. 2001a. Mo-
Botanical Laboratory, Nichinan, Miyazaki, Japan.
lecular phylogeny of the genus Trismegistia and related genera (Sema-
SCHUSTER, R. M. 1992. The Hepaticae and Anthocerotae of North America,
tophyllaceae, Musci) based on chloroplast rbcL sequences. Hikobia 13:
vol. VI. Field Museum of Natural History, Chicago, Illinois, USA.
529–549.
SHAW, J. 1986. Peristome structure in the Orthotrichaceae. Journal of the
TSUBOTA, H., H. AKIYAMA, T. YAMAGUCHI, AND H. DEGUCHI. 2001b. Mo-
Hattori Botanical Laboratory 60: 119–136.
lecular phylogeny of Sematophyllaceae (Hypnales, Musci) based on
SHAW, J. 2000. Phylogeny of the Sphagnopsida based on nuclear and chlo-
roplast DNA sequences. Bryologist 103: 277–306. chloroplast rbcL sequences. Journal of the Hattori Botanical Laboratory
SHAW, J., AND L. E. ANDERSON. 1988. Peristome development in mosses in 90: 221–240.
relation to systematics and evolution. II. Tetraphis pellucida (Tetraphi- TSUBOTA, H., T. ARIKAWA, H. AKIYAMA, E. DE LUNA, D. GONZALEZ, M.
daceae). American Journal of Botany 75: 1019–1032. HIGUCHI, AND H. DEGUCHI. 2002. Molecular phylogeny of hypnobry-
SHAW, J., L. E. ANDERSON, AND B. D. MISHLER. 1987. Peristome develop- alian mosses as inferred from a large-scale dataset of chloroplast rbcL,
ment in mosses in relation to systematics and evolution. I. Diphyscium with special reference to the Hypnaceae and possibly related families.
foliosum (Buxbaumiaceae). Memoirs of the New York Botanical Garden Hikobia 13: 645–665.
45: 55–70. VALENTINE, L. J., E. O. CAMPBELL, AND D. H. HOPCROFT. 1986. Study of
SHAW, J., L. E. ANDERSON, AND B. D. MISHLER. 1989a. Peristome devel- chloroplast structure in 3 Megaceros species and 3 Dendroceros species
opment in mosses in relation to systematics and evolution. III. Funaria (Anthocerotae) indigenous to New Zealand. New Zealand Journal of Bot-
hygrometrica, Bryum bicolor, and B. pseudocapillare. Systematic Botany any 24: 1–8.
14: 24–36. VAN DE PEER, Y., R. DE BAER, J. CAUWENBERGHS, AND R. DE WACHTER.
SHAW, J., L. E. ANDERSON, AND B. D. MISHLER. 1989b. Peristome devel- 1990. Evolution of green plants and their relationship with other pho-
opment in mosses in relation to systematics and evolution. IV. Haplole- tosynthetic eukaryotes as deduced from 5S ribosomal RNA sequences.
pideae: Ditrichaceae and Dicranaceae. Bryologist 92: 314–325. Plant Systematics and Evolution 170: 85–96.
SHAW, J., C. J. COX, AND S. B. BOLES. 2003a. Polarity of peatmoss (Sphag- VANDERPOORTEN, A., L. HEDENÄS, C. J. COX, AND A. J. SHAW. 2002a. Phy-
num) evolution: who says mosses have no roots? American Journal of logeny and morphological evolution of the Amblystegiaceae (Bryopsida).
Botany 90: 1777–1787. Molecular Phylogenetics and Evolution 23: 1–21.
SHAW, J., C. J. COX, S. B. BOLES, AND B. GOFFINET. 2003b. Phylogenetic VANDERPOORTEN, A., L. HEDENÄS, C. J. COX, AND A. J. SHAW. 2002b. Cir-
evidence for a rapid radiation of pleurocarpous mosses (Bryopsida). Evo- cumscription, classification, and taxonomy of Amblystegiaceae (Bryop-
lution 57: 2226–2241. sida) inferred from nuclear and chloroplast DNA sequence data and mor-
SHAW, J., AND B. GOFFINET. 2000. Bryophyte biology. Cambridge University phology. Taxon 2002: 115–122.
Press, Cambridge, UK. VAUGHN, K. C., AND J. HASEGAWA. 1993. Ultrastructural characteristics of
SHAW, J., AND H. ROBINSON. 1984. The development, evolution, and function the placental region of Folioceros and their taxonomic significance. Bry-
of peristomes in mosses. Journal of the Hattori Botanical Laboratory ologist 96: 112–121.
57: 319–335. VAUGHN, K. C., R. LIGRONE, H. A. OWENS, J. HASEGAWA, E. O. CAMPBELL,
SHAW, J., AND J. ROHRER. 1984. Endostomial architecture in diplolepideous K. S. RENZAGLIA, AND J. MONGE-NAJERA. 1992. The anthocerote chlo-
mosses. Journal of the Hattori Botanical Laboratory 57: 41–61. roplast: a review. New Phytologist 120: 169–190.
SINGH, D. K. 2002. Notothylaceae of India and Nepal (a morpho-taxonomic VERDOORN, F. [ED.]. 1932. Manual of bryology. Martinus Nijhoff, The
revision). Bishen Singh Mahendra Pal Singh, Dehra Dun, India. Hague, Netherlands.
SMITH, A. J. E. 1978. Moss flora of Britain and Ireland. Cambridge University VITT, D. H. 1981. Adaptive modes of the moss sporophyte. Bryologist 84:
Press, Cambridge, UK. 166–186.
SMITH, A. J. E. [ED.]. 1982. Bryophyte ecology. Chapman and Hall, New VITT, D. H. 1984. Classification of the Bryopsida. In R. M. Schuster [ed.],
York, New York, USA. New Manual of Bryology, vol. 2, 676–759. Hattori Botanical Laboratory,
SMITH, D. K., AND P. G. DAVISON. 1993. Antheridia and sporophytes in Nichinan, Japan.
Takakia ceratophylla (Mitt.) Grolle.: evidence for reclassification among WATSON, E. V. 1971. The structure and life of bryophytes, 3rd ed. Hutchinson
the mosses. Journal of the Hattori Botanical Laboratory 73: 263–271. & Co., Ltd., London, UK.
SMITH, E. C., AND H. GRIFFITHS. 1996. A pyrenoid-based carbonic-concen- WELLMAN, C. H., AND J. GRAY. 2000. The microfossil record of early land
trating mechanism is present in terrestrial bryophytes of the class An- plants. Philosophical Transactions of the Royal Society, London, B 355:
thocerotae. Planta 200: 203–212. 717–732.
WELLMAN, C. H., P. OSTERLOFF, AND U. MOHLUDDIN. 2003. Fragments of DNA. Ph.D. dissertation, Moscow State University, Moscow, Idaho,
the earliest land plants. Nature 425: 282–285. USA.
WENDEROTH, H. 1931. Beiträge zur Kenntniss des Sporophyten von Polytri- YOSHINGA, K., H. IINUMA, T. MASUZAWA, AND K. UEDAL. 1996. Extensive
chum juniperinum Willdenow. Planta 14: 244–385. RNA editing of U to C in addition to C to U substitution in the rbcL
WHEELER, J. A. 2000. Molecular phylogenetic reconstruction of the mar- transcripts of hornwort chloroplasts and the origin of RNA editing in
chantioid liverwort radiation. Bryologist 103: 314–333. green plants. Nucleic Acids Research 24: 1008–1014.
YATSENTYUK, S. P. 2001. Molecular phylogeny of the Bryophyta and Ly- ZARTMAN, C. 2003. Habitat fragmentation impacts on epiphyllous bryophyte
copodophyta, according to results from some sequences of chloroplast communities in central Amazonia. Ecology 84: 948–954.
View publication stats

Phylogeny and Diversification of Bryophytes

Uploaded by

Copyright:

Available Formats

Phylogeny and Diversification of Bryophytes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Phylogeny and Diversification of Bryophytes

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Phylogeny and diversiﬁcation of bryophytes

Article in American Journal of Botany · October 2004

Arthur Jonathan Shaw Karen S Renzaglia

SEE PROFILE SEE PROFILE

Ultrastructure and pectin composition of guard cell walls View project

The user has requested enhancement of the downloaded file.

PHYLOGENY AND DIVERSIFICATION OF BRYOPHYTES1

JONATHAN SHAW2,4 AND KAREN RENZAGLIA3

Even with the advent of molecular systematics and a re-

internal mucilage canals. Number of antheridia per cavity is

Fig. 6. Scanning electron micrographs of pores in hornworts. A. Stoma

ently inappropriate, placements in Dendroceros or the newly

acquisition within hornworts (Mishler and Churchill, 1984;

Fig. 10. Diagrammatic representation of apical cell shapes in liverworts

plates of cell wall material; most of the anticlinal walls are

and arthodontous mosses. The two families are paraphyletic to

View publication stats

You might also like