A revision of Rigodium (Musci:Rigodiaceae)

MISSING IMAGE

Material Information

Title:
A revision of Rigodium (Musci:Rigodiaceae)
Physical Description:
viii, 258 leaves : ill. ; 29 cm.
Language:
English
Creator:
Zomlefer, Wendy B
Publication Date:

Subjects

Subjects / Keywords:
Mosses -- Classification   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 248-253).
Statement of Responsibility:
by Wendy B. Zomlefer.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001747395
notis - AJG0218
oclc - 26371961
System ID:
AA00003736:00001

Full Text











A REVISION OF RIGODIUM (MUSCI: RIGODIACEAE)


By

WENDY B. ZOMLEFER



















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1991
































Copyright 1991

by

Wendy B. Zomlefer














ACKNOWLEDGMENTS


I express my deepest appreciation and gratitude to Dana

Griffin, III, for his limitless patience and enthusiastic

guidance throughout my graduate program. I also thank

Walter S. Judd, who also generously provided much

encouragement and many helpful criticisms. The other

members of my committee, Bruce J. MacFadden, Jonathan

Reiskind, and James W. Kimbrough have also greatly assisted

in my project and have improved the quality of the

dissertation with their comments. A very special thank-you

goes to Kent D. Perkins for his help in all computerized

aspects of this study, including data-base management, word

processing and processing of loans. I appreciate also the

advice and hospitality of William R. Buck and Richard Harris

(curators at the New York Botanical Garden) during my two

trips to their institution; Bill Buck has been especially

instrumental in my development as a bryologist.

I could not have completed this revision without the

time and patience of many other individuals. Thomas J.

Whitmore, Stephen B. Linda, and Carmine A. Lanciani helped

with aspects of the statistics; W. Mark Whitten, Stephen J.

Walsh, and Douglas D. Turley assisted as computer

consultants. Donna S. Williams spent numerous hours helping

me with the preparation of my specimens for the SEM. Norris


iii









H. Williams, my job supervisor at the Florida Museum of

Natural History, has been especially generous with his

computer facilities, as well as sympathetic to the problems

surrounding my concurrent employment and school activities.

Howard Crum provided excellent instruction during my tenure

as his assistant at the University of Michigan Biological

Station and graciously sent me an unpublished manuscript

related to my dissertation work. Daniel B. Ward and Dan H.

Nicholson advised me on several difficult nomenclatural

problems. Rhoda J. Bryant and Dianna C. Carver helped with

word processing.

I thank also the curators of the herbaria listed in the

materials and methods section for the extended loans of

their specimens. Travel money for three trips to study

bryophytes (one to Michigan and two to Puerto Rico) was

provided by the Department of Botany at the University of

Florida.

By their consistent concern about my graduate program

over the past five years, Graig D. Shaak, Gary S. Morgan,

and Laurie Wilkins (all of the Florida Museum of Natural

History), helped to keep my spirits up. In this regard, I

also thank Susan W. Williams and my sister, Kayla S.

Zomlefer, for their cheerful support.

Finally, I thank my parents, Jack and Dorothy Zomlefer,

who have always encouraged and supported my academic

pursuits, even in my formative years when females were

neither expected nor encouraged to excel in the sciences.















TABLE OF CONTENTS

pace
ACKNOWLEDGMENTS......................................... iii

ABSTRACT................................................. vii

INTRODUCTION............................. ................ 1

TAXONOMIC HISTORY....................................... 3

MATERIALS AND METHODS .................................. 20

MORPHOLOGY .................. ..................... ..... 25

Axes ....... .................... ............ ....... 25
Leaves ............ ........... ... ................. 28
Axillary Hairs.................................... 30
Paraphyllia and Pseudoparaphyllia................. 31
Perichaetia and Perigonia........................... 32
Sporophyte........................................... 33
Spores.............................................. 34
Chromosome Number.................................. 34

PHYLOGENY.. ................... ......................... 57

Characters for Analysis of Representative
Genera of Superfamily Leskeacanae............... 66
Results of Analyses of Genera...................... 69
Characters for Rigodium Species.................... 72
Results of Analyses of Rigodium species............ 74
Taxonomic Philosophy Concerning Species............ 77

PHENETIC STUDIES......................................... 99

Choice of OTUs and Choice of Specimens............. 100
Choice of Characters............................... 103
Results of the Principal Component Analyses......... 105

PHYTOGEOGRAPHY ............. .......................... 117

TAXONOMIC TREATMENT.................................. 122

Key to the Species of Rigodium..................... 125
1. Rigodium toxarion (Schwagr.) Jaeg............... 130
2. Rigodium brachypodium (C. Mull.) Par............ 167
3. Rigodium adpressum Zomlefer.................... 194









4. Rigodium implexum Kunze ex Schwagr.............. 207
5. Rigodium pseudo-thuidium Dus.................... 220

NOMINA AMBIGUA AND NUDA.................................. 234

LITERATURE CITED....................................... 238

APPENDICES

A OTUS (COLLECTIONS) USED IN THE PHENETIC
ANALYSES .............................. ............ 248

B DETERMINATION OF MINIMUM NUMBER OF
MEASUREMENTS PER CHARACTER...................... 254

BIOGRAPHICAL SKETCH..................................... 258








Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

A REVISION OF RIGODIUM (MUSCI: RIGODIACEAE)

By

Wendy B. Zomlefer

May 1991

Chairman: Dana Griffin, III
Major Department: Botany

Rigodium Kunze ex Schwagr., a genus of tropical to

austral-temperate pleurocarpous mosses, is comprised of five

species (plus two varieties). Prior to this study, 26

species were recognized, with five additional species

generally accepted as belonging to other genera. Species of

Rigodium, all dioicous, have wiry stems and branches,

strongly dimorphic stem and branch leaves, extremely

incrassate leaf cells that are prorulose to varying degrees

(depending on the species), and hypnoid peristomes. The

habit varies from arborescent with a well-developed stipe

(B. toxarion, R. adpressum, and R. brachypodium) to more

derived forms with either prolific (R. implexum) or reduced

(R. pseudo-thuidium) branching.

The genus has a disjunct distribution in Central and

South America and southeastern Africa. The plants, usually

epiphytic or terrestrial, generally grow in moist, montane

forests. The center of diversity, where all five species

occur, is in the Andean Cordillera along the Chilean-

Argentinian border in central Chile.


vii









Observations of morphology, including those made with

scanning electron microscopy (SEM), compose the basis of

this revision, which includes computer-assisted analyses of

phenetics (Principal Component Analysis, PCA) and cladistics

(Phylogenetic Analysis Using Parsimony, PAUP). One

previously undescribed species (R. adpressum) and two new

combinations at the varietal level (R. toxarion var.

robustum and R. brachypodium var. tamarix) are presented.

Rigodium is maintained within the monotypic Rigodiaceae,

possibly related to the Thuidiaceae.


viii














INTRODUCTION


As treated here, the genus Ricodium Kunze ex Schwagr.

consists of five species (including two varieties) of

austral-temperate to tropical pleurocarpous mosses from

Hispaniola, Central and South America, eastern Africa, and

Madagascar. The center of diversity of Rigodium, where all

five species occur, is in the Andean Cordillera along the

Chilean-Argentinian border in central Chile. The plants

typically grow as epiphytes (or less often terrestrial,

corticolous, or saxicolous) in moist forests at moderately

high altitudes (up to ca 3800 m).

Although 31 species have been included within Rigodium,

no comprehensive work had been published for the genus until

the present revision. The genus itself is well-defined and

recognizable by the dendroid to frondose habit with wiry

stems and branches and often a well-differentiated stipe

(lower secondary stem). The strongly dimorphic stem and

branch leaves are unusual in the marked tendency toward

ecostate stem leaves (at least those of the "stipe" or lower

stem). Stem leaves are typically strongly spreading to

squarrose, wet or dry. Leaf cells are generally rounded-

quadrate (alar region and margins) to oblong (upper and

midlaminar areas) and incrassate throughout. The plants are









dioicous, and the peristome is hypnoid (Zomlefer and Buck,

1990).

This treatment represents the first attempt to

completely review all pertinent literature and specimens in

order to (1) ascertain the species and their relationships

within the genus and (2) to determine the position of the

genus (in a monotypic family) within the scheme of bryophyte

phylogeny using cladistics. The results presented here are

based on qualitative and quantitative analysis of

morphology, ecology, and anatomy of over 2,400 herbarium

specimens. This study is especially pertinent since many of

the habitats where Rigodium occurs (mainly subtropical to

temperate rain forests) are threatened by human activities

(as for example, logging in the Valdivian forests in Chile).














TAXONOMIC HISTORY


The genus Rigodium was established by Schwagrichen in

1845 with original material of the new species, Rigodium

implexum, coming from southern Chile (Schwagrichen, 1845;

Zomlefer, 1990). The name is taken from the Greek word

"rigos" (= rigid) in reference to the extremely stiff and

wiry stems and branches. Although Muller (1845, 1851)

initially suggested maintaining this new genus rather as a

section within Hypnum, the name became well-accepted in the

literature, with even Muller himself eventually describing

six species using Rigodium as the genus name.

In addition to R. implexum, 19 new species were later

described, plus 11 more transferred to Rigodium from Hypnum

(ten species) and Leskea (one species). Of these 31 valid

species names (Wijk et al., 1967, 1969; Crosby, 1979), five

have been now generally accepted as belonging to other

genera: R. dentatum Dix. = Rhvnchostegiella zevheri (C.

Mull.) Broth., R. reflexum (Stark.) Kindb. = Brachythecium

reflexum (Stark.) B. S. G., R. schlosseri (Sendtn.) Kindb. =

Camptochaete schlosseri (Sendtn.) Par., R. vaqum (C. Mull.)

Reichdt. = Camptochaete vaga (C. Mull.) Broth., and R.

varium (Hedw.) Kindb. = Amblystegium varium (Hedw.) Lindb.

In addition, four more Rigodium types examined in the

initial stages of this study have also been reassessed









(Zomlefer & Buck, 1990): R. toxarioides Broth. & Par. (Fig.

1) and R. crassicostatum Bartr. in Christ. (Fig. 2) both =

Eurhynchium praelonaum (Hedw.) B. S. G.; R.

pteriyvnandroides (Broth.) Broth. (Fig. 3) = Helicodontium

pervirens (C. Mull.) Par.; and R. riparium Sehnem (Fig. 4) =

Helicodontium complanatum Broth. This leaves a total of 22

valid epithets (Table 1) incorporated into the present

study--20 associated with Hispaniola and Central and South

America, and one each with eastern Africa and with

Madagascar.

The authors of the original species descriptions have

generally emphasized subtle differences in characters, such

as the relative size and fragility of the plants and the

coarseness of the serrations on the leaves. By apparently

comparing type specimens only, the authors did not

sufficiently evaluate the intergradation of these characters

throughout the genus and thus may have overlooked numerous

specimens which bridged the apparent morphological gaps

between the type specimens. The most obvious "species"

based on the old literature are R. pseudothuidium/R.

hvlocomioides (both with distinct habit and a reduced costa

in the stem leaves) and R. implexum (very stiff plants with

unusual right-angle branching). Certain names tended to be

associated only with collections from the particular country

of the type specimen, such as "R. solutum" with Ecuador and

"B. leptodendron" with Bolivia (see Table 1). Adding to the

confusion, authors of floristic treatments tended to lump









indiscriminately several names in synonymy, such as Jaeger

[1878; R. toxarion = R. implexum, B. brachypodium, and

Hypnum (Riqodium) solutum], Muller [1882, Hvpnum (Riqodium)

argentinicum = E. implexum and R. brachypelma

(brachypodium)], and Mitten [1869; Hypnum (Rigodium)

toxarion = H. implexum, H. brachypodium, H. solutum, and R.

lechleri (arborescens)]. Brotherus' (1925) key to 17

species separates out only the few obvious forms (i.e., R.

implexum, R. hvlocomioides, and several species no longer

included in Rigodium) and lumps the remaining 12 species (by

country) under the first general dichotomy of the key (habit

type).

The great need for a full-scale revision of Rigodium

has been noted by authors from Reimers (1926) to Robinson

(1975). In reviewing Chilean collections of bryophytes and

examining numerous specimens of Rigodium in particular,

Reimers (1926) and Herzog (1938, 1954) both concluded that

making any decisions about the taxonomic validity of

Rigodium species was virtually unfeasible without

monographic studies. In fact, Reimers promised a

forthcoming revision (which never appeared). Herzog (1939)

and Bartram (1960) each described a new species of Rigodium

(R. pendulum and R. crassicostatum, respectively), while at

the same time admitting the near impossibility of applying

any specific names without a critical revision of the entire

genus.









Other than the keys of Brotherus (1909, 1925), no

attempt has been made to distinguish all the species of the

genus, although several workers have treated a few species

of Rigodium with a regional, floristic perspective. Most

floras do not include descriptions or keys but only list

those species names (and sometimes the specimens) associated

with the area (e.g., Seki, 1976). Notable exceptions are

the floras of Sehnem (1976) and Robinson (1975), which

include keys and some discussion comparing several species.

These two floras, however, are restricted in scope and do

not address Rigodium as a whole. Sehnem's keys for seven

species of Rigodium in his treatment of the mosses of

southern Brazil use vague characters such as a "delicate

habit, similar in aspect to Pterqynandrum or Thuidium." He

described two new species, R. riparium and R. pallidum.

Rigodium riparium Sehnem is actually Helicodontium

complanatum Broth. (Zomlefer & Buck, 1990), and R. pallidum

Sehnem has been reduced to synonymy with R. toxarion

(Schwagr.) Jaeg. in the present study. Robinson's treatment

of the mosses of the well-collected Juan Fernandez Islands

includes detailed discussions of four species (B.

arborescens, R. hvlocomioides, R. toxarion and R. robustum)

with comments on his opinions of synonomy and relationships

to several other species on mainland Chile. By

concentrating on a small, well-collected area, he apparently

made some assumptions about Riqodium species that reflect

better the local variation rather than that found throughout









whole range of the genus. His conclusions are commented on

under the appropriate species in this study.

Coupled with problems in delimitation of the species is

a lack of agreement on the affinities of Rigodium to other

groups. As with many pleurocarpous taxa, Rigodium was

initially allied with Hypnum/Hypnaceae (e.g., Muller, 1851).

By the early twentieth century, more pleurocarpous families

had been delimited, and Rigodium was moved to the

Brachytheciaceae by Brotherus (1909). Later, he (1925)

transferred it to the Lembophyllaceae. Associating the

genus with these two families has persisted until recently.

For example, Robinson (1975) placed Riqodium in the

Brachytheciaceae due to its similarity stipess and strong

heterophylly) with Stokesiella [hom. illeq.; = Eurhynchium

fide Buck (1988)]. More recent publications (Crosby &

Magill, 1981; Vitt, 1982; Walther, 1983) maintain Ricodium

under the Lembophyllaceae, although the validity of this

family, as first conceived by Brotherus, has continued to be

questioned (Crum, 1991). Buck (1980a) reduced the

Lembophyllaceae to two genera (Camptochaete and

Lembophyllum) and suggested placement of Rigodium in the

Thuidiaceae, a move supported by Vitt (1984). According to

Buck, pertinent characters linking Rigodium to the

Thuidiaceae include the pinnate branching, similar leaf

shapes, short leaf cells, and strong single costae.

Bartram (1949) was the first to suggest placement of

Rigodium in a separate family, if not the Brachytheciaceae









or Leskeaceae (in which he included the Thuidiaceae). Crum

(1981, 1984) recently erected the monotypic family

Rigodiaceae, which has been accepted by some authors (Buck &

Vitt, 1986), and tentatively supported by others (Walther,

1983). Although he later (1991) agrees with Buck (1980a)

and acknowledges a close resemblance of Rigodium to the

Thuidiaceae, Crum maintains that the "stipitate growth,

smooth leaf cells, and absence of paraphyllia" merit

separate familial status for the genus.

The present study recognizes five species (including

two varieties) of Rigodium, including one species described

here for the first time: R. adpressum Zomlefer, R.

brachypodium (C. Mull.) Par. var. brachypodium, E.

brachypodium (C. Mull.) Par. var. tamarix (C. Mull.)

Zomlefer, R. implexum Kunze ex Schwagr., R. pseudo-thuidium

Dus., B. toxarion (Schwagr.) Jaeg. var. toxarion, and R.

toxarion (Schwagr.) Jaeg. var. robustum (Broth. in Skottsb.)

Zomlefer. Considering the initial 22 epithets, the

recognition of only five species represents a 77% rate of

over-description. This figure is consistent with moss

revisions in general: according to Touw's (1974) survey, 73

to 79% of published specific and infraspecific names have

been reduced in synonymy in moss revisions from 1901 to

1974.

The results of the preliminary cladistic study and

character analysis in this revision indicate that Rigodium

is phylogenetically intermediate between Echinodium







9

(Echinodiaceae) and Thuidium (Thuidiaceae), and that all

three taxa are basal in Buck and Vitt's (1986) superfamily

Leskeacanae of the Hypnales (see detailed discussion in

section on phylogeny). With Echinodium as an outgroup for

Rigodium, cladistic analyses identify E. toxarion as the

most ancestral species, R. implexum and R. pseudo-thuidium

as the most derived sister-species in the genus, and R.

brachypodium and R. adpressum as phylogenetic intermediates.




















FIG. 1. Type of Rigodium toxarioides Broth. & Par. =
Eurhynchium praelonqum (Hedw.) B. S. G. A. Habit. B.
Detail of stem. C. Stem leaf. D. Branch leaf. E-H.
Areolation of leaf. E-F. Basal cells. G. Median cells. H.
Detail of upper costa showing dorsal spine. All from
Apollinaire s.n., 1905 (H-BR). Copyright 1990 by the
American Bryological and Lichenological Society. Reprinted
with permission from The Bryologist 93 (3): 305.

















D

if


I30
30Orm


50,um





















FIG. 2. Type of Rigodium crassicostatum Bartr. in
Christ. = Eurhynchium praelonqum (Hedw.) B. S. G. A. Habit.
B. Detail of stem. C. Stem leaf. D. Branch leaf. E-H.
Areolation of leaf. E-F. Basal cells. G. Median cells. H.
Detail of upper costa showing dorsal spine. All from
Christophersen & Meiland 907 (BM). A and B copyright 1990
by the American Bryological and Lichenological Society.
Reprinted with permission from The Bryologist 93 (3): 305.


















D


30,um


0.25 mm





















FIG. 3. Type of Rigodium pteriqynandroides (Broth.)
Broth. = Helicodontium pervirens (C. Mull.) Par. A. Habit.
B. Stem leaf. C. Branch leaf. D-G. Areolation of leaf.
D. Basal cells. E. Median cells. F. Acumen. G. Detail of
upper costa showing dorsal spine. H. Capsule. All from
Lindman 52 (H-BR). Copyright 1990 by the American
Bryological and Lichenological Society. Reprinted with
permission from The Bryologist 93 (3): 306.













B C
0.25mm G





A 0.5mm


A E


30,m


30,m





















FIG. 4. Type of Riqodium riparium Sehnem =
Helicodontium complanatum Broth. A. Habit. B. Portion of
plant with sporophyte and perigonia. C. Stem leaf. D.
Branch leaves. E-G. Areolation of leaf. E. Basal cells.
F. Median cells. G. Acumen. H. Capsule. All from Sehnem
21515 pacaA). Copyright 1990 by the American Bryological
and Lichenological Society. Reprinted with permission from
The BryoloQist 93 (3): 307.













r


0.5 mm

F


0.25mm















*-
'0
CO
COO C
0 0
0 0 0
*4-l A *A
X 0 X
O 0


KI KI al


4-4
0
U,
0 t
>i
C





-,4
o w
0 0



MC





CO
0)


'0





O w
0 0
U U

Q)c 4
o0 a
4P 0


,) --4 41

G-M .P(


14 -4





W E-4

*^ C
* *09






Ot
0 (0
:0





w 9
-4 CO



X -H
.0 w
o 0



S-H a
* C


-'0 0-






aU
1-1


c c
0 0
*H *H
4 4
0 0
X X
o0


.I aI


U)
r-I
(0 3
-H 0
0 o
0 0 -1


'0
*4



'0

C










(1
C Q
0


Ua -4

- *H







H C
U 0































0
CM





r-H





-4

0


z 0 0
X -4 -A
W) -4 -4
r.1 (0 (0
6 0 0
4.3 4.

*I -PI I


.4-H














cI







0
p4
0 3







0






-.



0




0
'4



-4


r-4 (0
(0 U
P ^
CO
04:











co
-4
















0



.)
a



O


(0


5 5 O





co .0 N IQ H1
3 30 3
C C C
0 0 0
> > >

COJ3 N3 H 4
eo ml a ui c
CO1 W- r l -













3 0

3 C 'OX
C 4 C C C C C C1 C 0-H C
00 0 0 0 0 0 I OE0 0 4 0
.--4 *r -H -4 *HH 0 3 *- > (a -H
4O 4 4 4 W Wc '0 kW.Q w 1c 54
(a0 al ( (c R% 4) (U : (0 0 (o 0( t (0
Xw X X X X 4) X 4) X54 X (04J x
0 O O O 0 0 n 0 0 1 0

* (U 0 (
MI> rI I Xi I c MI 1I PI > 1i 9I > 1I


C0 iS
U UC
0) w0 U 4- -Hr
C Q) & *M Q) *^ *H c C
4 H N H N N U )
,4H .- 0 r -- 4
M r i 0 (0 0 (
H U z co U O M m 4


o %o %0 ON i-I H- LI
M, N r- M 0 0 0O


*4
C
a) w 4
Z4 0 0
C '0

30 0
bH M


Z 0


,e ) -a, 4r
X 4 *E 00
u < aI- ) co


c I


-o H o i
a a


H
4 r-H
') :
E4

0 U
) *
C N r

-49








Oipl Ol K


4
51





0 )
H
(U





H
0
o 4)

IU X



I PI

MI KI


*'
a,
-4

0



(U
C


*5

E-4




0
H
MI


+1
0
o


-4







o 0
0 0
0 0
w














MATERIALS AND METHODS


In the course of this study over 2,400 specimens were

examined from the collections of 43 herbaria: ALTA, B, BA,

BM, BR, C, CANM, CHR, DUIS, E, F, FH, FLAS, G, GB, GOET, H,

IA, JE, KRAM, L, LPB, M, MANCH, MEXU, MICH, MO, MPU, NAM,

NY, O, OXF, P, PACA, RB, RO, S, TENN, U, UPS, UWM, VBI, and

W. The acronyms are standardized according to Holmgren et

al. (1990); specimens indicated as from "Herb. Matteri" are

from the personal herbarium of C. Matteri. The label data

from each specimen were entered into a data base on an IBM-

compatible PC. The user interface for the program, similar

in structure to dBase III PLUS (Ashton-Tate, registered

trademark), was compiled by Kent D. Perkins specifically for

this project using CLIPPER (Nantucket Corp., registered

trademark) language. As exsiccatae are common in bryology,

cataloging the numerous specimens with a data base greatly

facilitated the identification of duplicates and fragmentary

specimens that had incomplete or missing data on the labels.

As computer "records," the specimens could also then be

quickly sorted into any appropriate category ("field")--such

as by collector, country, elevation, type designation, or

reproductive state--for use in specimen citations, ranges,

and exsiccatae lists. The modern spellings of geographical

names are according to the United States Office of









Geography's Official Standard Names gazetteer volumes for

Latin America (1955, 1956, 1957, 1963, 1967, and 1968).

Hydrated specimens and permanent slides were examined

for the detailed measurements and qualitative characters

used in the descriptions, phenetic studies, and cladistic

analyses. (The specific methodologies for the cladistic and

phenetic analyses are discussed in their particular

chapters.) Prior to examining them, appropriate portions of

specimens were steeped for several minutes in boiling water.

Occasionally, the entire contents of a packet were soaked if

the plants were inextricably intertwined. The strongly

clasping stem and stipe leaves (only ca 0.45 1.70 mm long)

were individually teased off the stems with fine-tipped

forceps. At least ten to twenty leaves were mounted for

each stipe and stem leaf slide. Leaves, stems, and capsules

were sectioned by hand using single-edged razor blades.

Capsules were cut vertically and the columnae and spores

flushed out. The two capsule sections were then cut into

fourths for the slide.

Permanent slides of whole leaves and sectioned parts

were prepared with Hoyer's solution (Anderson, 1954). The

slides were then heated following application of the cover

slips to clear the air bubbles and flatten the specimens.

Temporary water mounts were prepared of capsules from

specimens with scanty sporophyte material, which was

returned to its packet after examination. At least five

slides were made of each specimen examined, including whole









mounts of the leaves (stipe, stem, branch, perigonial, and

perichaetial) and sections of the stem, stem leaves, and

capsules.

Measurements of the stems, branches, and capsules, as

well as qualitative observations on habit features (e.g.,

leaf position), were made with the aid of a Wild M5A

Stereomicroscope. Specimens on slides (leaves, cells,

peristome features, and spores) were measured with a Wild

M20 compound microscope equipped with drawing tube.

Measurements were made with an appropriately scaled gauge on

outlines sketched with the drawing tube or directly on the

projected image from the drawing tube. Detailed data

sheets, which included xerox copies of the actual (hydrated)

specimen as a record of the habit, were compiled to organize

the measurements.

The surface details of stem and perichaetial leaves

papillaee, prorosities), peristomes, and spores were

examined and photographed using a Hitachi S-450 scanning

electron microscope (SEM). Specimens were attached to studs

with double-sided tape and then plated with gold in an Eiko

IB-2 ion coater for five minutes at 8 mA prior to SEM

examination. Capsules peristomess) did not require

rehydration and were attached to the stubs in the dry

condition directly from the herbarium specimen.

Spores and leaves, which tended to be collapsed in the

dried state, required special preparation. Stem leaves were

left attached to ca 0.5 1.0 cm stem segments, perichaetia









were removed as a whole unit, and spores were left in their

capsules during rehydration. Due to their small size,

perichaetia and capsules with spores were sandwiched between

1 cm Nucleopore polycarbonate filter papers placed in

modified Beem capsules. All leaves and spores were first

hydrated for 15 minutes in warm water. The leaves of

Rigodium toxarion and R. brachypodium were next placed

through a series of successive ten-minute washes in 25, 50,

75, and 95% ethanol, respectively, followed by two

successive ten-minute rinses in 100% ethanol. The specimens

were subsequently subjected to a critical point drying

sequence composed of five fifteen-minute wash cycles of

liquid CO2 in a Balzers CPD-101 Critical Point Dryer. These

leaves were then coated with gold as described above.

The perichaetia and stem leaves of R. adpressum and all

spores required fixation prior to ethanol rinses and

critical point drying. Otherwise, entire spores and the low

papillae at the base of the perichaetial leaves tended to

collapse, and the decurrent portion of the stem leaves of R.

adpressum curled over. After rehydration in water, these

specimens were soaked in 5% gluteraldehyde in .2M

NaCacodylate buffer for two hours. They were then rinsed

three times in .05M buffer for five minutes and placed in a

1% solution of osmium tetroxide (Os04) in buffer overnight.

After four rinses with distilled water, these specimens were

taken through the alcohol washes, critical point drying








24

sequence, and coating preparation as described above for the

capsules and other leaves.
















GENERAL MORPHOLOGY


Salient morphological features of the genus Rigodium

are included in this chapter, along with a discussion on the

terminology (particularly that of the axes and corresponding

leaves) incorporated in this monograph.

Axes

Traditionally, bryologists have used the terms

"primary" and "secondary" to distinguish the stoloniferous

vs. the erect/pendent axes, respectively, of a pleurocarpous

habit (e.g., Arzeni, 1954). Argent (1973) describes in

detail the difficulty in applying these designations to the

different types of shoots in the Pterobryaceae, a

pleurocarpous group. His simplified terminology (stolon,

stems, branches, and flagella) has been adopted by others

(e.g., Allen & Crosby, 1986). Some of Argent's ideas have

been incorporated into this monograph, with the addition of

a few terms peculiar to the habit of Rigodium (stipe vs.

stem, primary, secondary, and tertiary branches), and are

defined below:

1. "Stolons" (or rhizomes) are often referred to as

"primary stems" by other authors (e.g., Crum, 1981, 1984).

These axes are usually creeping, deeply colored (brown-red

to dark brown) and bear rhizoids. Leaves, when present, are










often achlorophyllous and less well-developed than those on

other axes.

2. "Stems" and stipess" are general terms for the two

parts of the main axis (or secondary stem of some authors)

of plants which are erect (or pendulous) and are derived

from the stolons. The stipe is the lower, unbranched

portion of the main shoot which is often darkly pigmented.

In Rigodium, stipe leaves are ecostate but are otherwise

well-developed. The term "stem" refers here to the upper,

branched portion of Rigodium plants where the leaves are

generally costate.

3. Branches are ranked in this study as primary,

secondary, and tertiary. Primary branches arise from stems

and are characterized by leaves intermediate in shape and

size of those of the stem/stipe and secondary/tertiary

branches. Secondary branches, which arise from the primary

branches, have narrowly ovate leaves without a well-

developed acumen. Tertiary branches are borne on secondary

branches. The leaves are similar in shape but smaller than

those of the secondary branches.

The delimitation of these types of axes outlined above

is not always clear-cut in Rigodium since they may

intergrade with one another. Allen and Crosby (1986)

suggest that this intergradation in Squamidium is due to

probably very weak genetic control coupled with

environmental influences, which may also be the case in

Rigodium. For example, in addition to the secondary










branches, the upper main axis (stem) may give rise directly

to new main axes (stems), resulting in "stem-like primary

branches" (Figs. 7H-J, 8H-J) that are similar to the

innovations in genera such as Hylocomium (Crum & Anderson,

1981). In Rigodium, though, these "innovations" are not

necessarily restricted to the area just below the

inflorescences, but may arise anywhere along the stem.

Although the term is not used here, Rigodium species also

have "flagellae" as defined by Allen and Crosby (1986): axes

stolonss) which develop from the tips of branches. However,

in Rigodium both branches and stems may root at the tips and

produce new plants (produce either stolons or stems; Fig.

5). Rhizoids, then, may be found at the tips of rooting

stems and branches as well as on the substrate sides of

stolons and stems.

The branching in Rigodium is monopodial (as defined by

Crum & Anderson, 1981) with irregularly pinnate branching.

Three of the species (R. brachypodium, R. adpressum and R.

toxarion) are dendroid with distinct creeping vs. pendulous

(or erect) portions. In R. pseudo-thuidium and R. implexum

the stipe is less distinct and short. R. implexum is

profusely branched, forming well-developed cushions (Fig.

43A). Plants of R. pseudo-thuidium have very little

secondary branching (Fig. 46A).

The various types of axes are similar in anatomy

although branches are much smaller in diameter and with a

less-developed central strand. In all species, several









layers (three to six rows) of small and incrassate epidermal

cells (Fig. 6A and B) form a well-developed rind which

probably contributes to the characteristic stiffness of the

plants. The epidermis has a furrowed appearance in surface

view (Fig. 11C). The cortex (Fig. 6B) consists of large,

hyaline thin-walled cells which usually become smaller

toward the center of the axis. Sometimes the smaller cells

of the cortex form a loosely organized ring around the

central strand as in Fig. 6B. Small, reddish, incrassate

cells comprise the central strand which varies in degree of

development (Fig. 6C and D) but generally consists of ten to

twenty (or sometimes more) cells.

Leaves

Riaodium is unusual in having ecostate lower stem

(stipe) leaves and costate upper stem leaves, in addition to

the dimorphic stem and branch leaves. Due to intergradation

of axes and corresponding leaf forms, care must be taken in

assigning and comparing "leaf types." Figures 7 and 8 show

the relative changes in size and morphology of leaves along

particular axes in R. toxarion (long acumen) and R.

brachypodium (short acumen). In this study (phenetic

analysis and all descriptions), "stipe leaves" are defined

as those ecostate leaves from the mid-stipe, "stem leaves,"

as those from the mid- to upper stem, and "branch leaves,"

as those from the secondary branches (Fig. 23A).

In all species, stipe and stem leaves are broadly ovate

and taper abruptly to a long or short acumen. The strongly









clasping and concave leaf bases tend to be decurrent (in R.

adpressum, strongly so). All species show a tendency to

form auriculae (Fig. 9A-E), although this character is most

consistent in R. pseudo-thuidium. Stipe and stem leaves are

typically wide-spreading to squarrose, wet or dry, but the

leaves of R. adpressum are appressed with spreading tips.

Branch leaves are consistently ovate (R. pseudo-thuidium) to

usually narrowly ovate (all other species) and are typically

erect to spreading.

Leaf cells are incrassate throughout and the alars are

not differentiated. Cell walls vary from even (R. toxarion)

to uneven (E. brachypodium). Due to the thick walls, the

cells are consistently prorulose, most weakly so in R.

toxarion (Fig. 10A). In the most extreme case (R.

brachypodium), the midlaminar cells project as well-

developed papillae (prorulae) on the dorsal surface (Fig.

10C and D). Although leaf cells in Rigodium are reported as

smooth (Crum, 1981, 1984, 1991), central papillae occur on

the dorsal side of some cells in the decurrent portion of

the stem and stipe leaves of R. adpressum (Figs. 11D, 40F

and G) and at the base of perichaetial leaves of some

specimens of all species (discussed below under perigonia

and perichaetia).

An outstanding feature of the areolation of Rigodium

stem, stipe, and branch leaves is an apparent intramarginal

border very visible at low power (Fig. 12A and B) and also

distinguishable in cross sectional view of the leaves (Fig.









12C), as well as in surface view with the SEM (Fig. 10B).

The border is formed by four to nine rows of cells that have

much thicker walls than those of the midlaminar cells. The

cells comprising this border are often a different shape

than those of the midlaminar portion of the leaf, as well.

For example, the seven- to ten-cell border in R. pseudo-

thuidium typically is composed of longer and wider cells

than the rest of the leaf, while the narrower border (four

to six cells) of R. adpressum consists of much shorter

cells. The intramarginal border is always present to some

degree but its appearance varies from weak (R. toxarion) to

very strong (R. adpressum), depending on the relative

thickness of the cell walls in comparison to the midlaminar

cells.

The costa is absent from the stipe leaves and is

usually present in the stem leaves. The stem leaves of all

species, however, show a tendency for a reduced costa (e.g.,

Fig. 9F-I), which is most consistent and characteristic of

R. pseudo-thuidium. The costa is strongest in R. implexum

(five-stratose in the thickest part) and to a lesser extent,

R. brachypodium (47-96 um wide at base). Branch leaves are

always costate in Riaodium, although the costa may be weak

in R. pseudo-thuidium (Fig. 9H and I).

Axillary Hairs

Recently (Griffin & Buck, 1989), axillary hairs have

been emphasized as useful characters for characterizing

taxonomic groups, especially genera (Saito, 1975) or









groupings of genera, such as subfamilies (Griffin, 1990).

Although not useful for distinguishing between species of

Rigodium, the structure of the axillary hairs is consistent

within the genus: two pigmented basal cells and an elongate,

hyaline terminal cell (Fig. 6E-I). The lower basal cell is

sometimes very lightly pigmented and difficult to see. The

axillary hairs are profuse at the tips of stems and branches

and are easiest to locate in E. adpressum appressedd leaves

which prevent loss of the delicate hairs) and R. pseudo-

thuidium (relatively large leaves with broad bases).

Paraphvllia and Pseudoparaphyllia

Paraphyllia and pseudoparaphyllia are small,

chlorophyllous, filamentous or foliose structures arising on

the stems of pleurocarpous mosses (Rohrer, 1985).

Historically, distinctions between the two structures have

been blurred until recently with the investigations of

Iwatsuki (1963) and Ireland (1971). Paraphyllia, usually

abundant when present, occur everywhere on stems and

branches (Ireland, 1971). Rigodium lacks these structures.

Pseudoparaphyllia, not particularly abundant when found

on moss plants, are restricted to areas around the branch

primordia and mature branches. The structures, distinct for

many species, have recently been emphasized in moss taxonomy

(Buck, 1984). However, foliose pseudoparaphyllia and

rudimentary branch leaves are sometimes confused (Allen,

1987). Generally, pseudoparaphyllia arise strictly on the

stem, while immature branch leaves occur at the base of









branches (Buck, 1984). Ovate, foliose, ecostate structures

occur at the base of branches of Rigodium (e.g., Figs. 7K,

8K). These intergrade with the branch leaves above (Figs.

7L-O, 8L-O) and are interpreted here as rudimentary branch

leaves and not as pseudoparaphyllia. Pseudoparaphyllia do

not occur in Rigodium.

Perichaetia and Perigonia

Rigodium is dioicous, and male and female plants are

indistinguishable. Reports of monoicism in the literature

(e.g., Sehnem, 1976) were due to the inclusion of discordant

elements within the genus, such as species of Helicodontium

(Zomlefer & Buck, 1990). Specimens with archegonial plants

are much more common than those with antheridia. The

infrequency of male populations compared to female in

dioicous mosses has been noted often (Longton & Greene,

1979; Stark, 1987) but not explained.

Both perichaetia and perigonia are bud-like structures

arising laterally along the axes of the plant. Perichaetia

are most common along the stems (but also occur on the

primary branches), whereas perigonia are most frequent along

primary branches, occurring much less often on the stems

(and secondary branches as well). Perigonial leaves (e.g.,

Fig. 28F) are similar to branch leaves in shape and

areolation but are ecostate and also tend to be more broadly

ovate. The inner perigonial leaves may also taper to a

somewhat longer acumen. Perichaetial leaves, weakly costate

to ecostate, are consistently subulate with spreading to









squarrose acumens in all species. The areolation, oblong to

rhombic cells at the base and shorter flexuous cells above,

thus differs from that of the stem leaves (shorter cells

below, cells above not flexuous). In addition, dorsal

central papillae may be found within the some cells near the

base of the outer perichaetial leaves (Figs. 11A and B, 41B

and C). These papillae are consistent on all available

collections of R. adpressum with archegonial plants and are

occasionally to frequently present in the other four species

as well (e.g., Fig. 35B).

Representative examples of gametangia and paraphyses of

Rigodium are illustrated in Figs. 13A-D. Of note is that

the structure of the paraphyses is somewhat similar to that

of the axillary hairs: usually two (but sometimes more)

pigmented cells at the base plus elongate hyaline terminal

cells (several as opposed to one in the axillary hairs).

The paraphyses subtending the antheridia (Fig. 13C) tend to

be stouter than those found within perichaetia (Fig. 13A).

Sporophvte

Mature sporophytes are not commonly found on Rigodium

specimens. The morphology of the capsule is consistent

among the species, although the urns of R. pseudo-thuidium

(Fig. 47H) are somewhat larger than those of the other

species. The asymmetric, horizontal to inclined capsule of

Rigodium has a well-developed rostrum (Fig. 13F) on the

operculum. The annulus, consisting of two rows of

incrassate cells, is illustrated in Fig. 13H. The









phaneroporous stomates (Fig. 13G) occur at the base of the

urn.

Rigodium has a typical hypnaceous peristome, which does

not vary in structure between species (Figs. 131, 14). The

exostome teeth, cross striolate below (Fig. 14C and D) and

papillose above (Fig. 14B), have well-developed lamellae on

the ventral side (Fig. 14E and F). The endostome consists

of a high basal membrane, well-developed segments (Fig. 14A)

and usually two to three cilia.

Spores

The typical spherical and papillose spores are

illustrated in Figs. 13J and 15A-D.

Chromosome Number

A chromosome number of n = 11 has recently been

reported by Deguchi and Oginuma (1990) for "Rigodium

implexum" (Dequchi 31548, KOCH; specimen not seen).





















FIG. 5. General Morphological Features of Rigodium:
Rooting at Tips. From T. Muller 2635 (L).
















































5mm





















FIG. 6. General Morphological Features of Rigodium:
Stem Anatomy and Axillary Hairs. A-D. Stem cross sections.
A. Whole stem section indicating position of wedge featured
in B. B. Detail of section showing rind, cortex and well-
developed central strand. C-D. Variation in the development
of the central strand. C. Weak central strand. D. Typical
development of the central strand. A-B from Sehnem 4686
(PACA, E. toxarion); C from Crosby 12984 (MO, R. toxarion);
D from Skottsberq & Skottsberg 418 (S, R. toxarion). E-I.
Representative axillary hairs from each species. E. R.
adpressum (Roivainen 1597, H). F. R. toxarion (Wasum et al.
3964, FLAS). G. R. brachvpodium (Matteri 2274, BA). H. R.
implexum (Crosby 12504, MO). I. R. pseudo-thuidium
(Schiavone 2451, BA).
























A


r------a -
S----------


0.2mm







C


50,jm


E








25um





















FIG 7. General Morphological Features of Rigodium:
Changes in Leaf Morphology (Leaf with Long Acumen) along
Various Axes. A. Habit. B. Leaf from stolon. C-E. Stipe
leaves. C. Leaf from stipe base. D. Leaf from mid-stipe.
E. Leaf from stipe apex. F-G. Stem leaves. F. Leaf from
mid-stem. G. Leaf from stem apex. H-J. Leaves from "stem-
like" branch (innovation). H. Leaves from innovation base.
I. Leaf from mid-innovation. J. Leaf from innovation apex.
K-M. Primary branch leaves. K. Leaves from branch base.
L. Leaf from mid-branch. M. Leaf from branch apex. N.
Leaves from secondary branch base (left) to apex (right).
O. Leaves from tertiary branch base (left) to apex (right).
All from Tonduz rPittier No:1 5667 (BR, E. toxarion).



























































0.3mm





















FIG. 8. General Morphological Features of Rigodium:
Changes in Leaf Morphology (Leaf with Short Acumen) along
Various Axes. A. Habit. B. Leaf from stolon. C-E. Stipe
leaves. C. Leaf from stipe base. D. Leaf from mid-stipe.
E. Leaf from stipe apex. F-G. Stem leaves. F. Leaf from
mid-stem. G. Leaf from stem apex. H-J. Leaves from "stem-
like" branch (innovation). H. Leaves from innovation base.
I. Leaf from mid-innovation. J. Leaf from innovation apex.
K-M. Primary branch leaves. K. Leaves from branch base.
L. Leaf from mid-branch. M. Leaf from branch apex. N.
Leaves from secondary branch base (left) to apex (right).
O. Leaves from tertiary branch base (left) to apex (right).
All from Elliott 139 (H, R. brachypodium).
























































0.3mm





















FIG. 9. General Morphological Features of Rigodium:
Aspects in Leaf Morphology. A-E. Examples of auriculate
leaves in the various species (rare in all but R. pseudo-
thuidium). A. R. pseudo-thuidium (Halle & Skottsberq 842,
S). B. R. brachvyodium (Dusen 357, W). C. R. adpressum
(Kuhnemann 5260, ALTA). D. R. implexum (Poeppia s.n., BM).
E. R. toxarion (Landrum 240, MO). F-I. Stem leaves from a
collection of R. toxarion, showing unusual variation in
costa development (Tonduz rPittier No:1 5667, BR).







































0.3mm


0.3mm





















FIG. 10. General Morphological Features of Rigodium:
Scanning Electron Micrographs of Leaf Surfaces. A. Dorsal
side of R. toxarion stem leaf showing weakly prorulose
midlaminar cells. B. Ventral side of R. brachypodium stem
leaf (arrow indicates intramarginal band of cells). C.
Dorsal side of R. brachypodium stem leaf showing projecting
end walls (prorulae) of midlaminar cells. D. Detail of C.
A from Tonduz [Pittier No:l 5667 (BR); B-D from Crosby 12992
(MO). A: line = 25 um; B, C: = 50 um; D: = 10 um.







46





















FIG. 11. General Morphological Features of Rigodium:
Scanning Electron Micrographs of Leaf Surfaces in R.
adpressum. A. Whole perichaetium (arrow indicates position
of detail in B). B. Detail of dorsal surface of
perichaetial leaf base showing papillae. C. Stem leaves
(arrow indicates decurrent portion). D. Detail of
decurrency showing papillae. A-B from Dusen 397 (S), C from
Matter & Schiavone 4985 (Herb. Matteri, D from Cantino M-
11 (MO). A: line = 250 um; B: = 15 um; C: = 100 um; D: = 25
um.







48





















FIG 12. General Morphological Features of Rigodium:
Areolation of Stem Leaf. A. Stem leaf showing location of
view in B. B. Portion of stem leaf with various zones
labelled. C. Cross section of stem leaf with zones
corresponding to B. All from Claude-Joseph 5775 (FH, R.
implexum).


























/ '0.2mm



intramarginal
-*-costa -midlaminar cells -cells marginal cells







5030 frn ,q
80 0 \




0 00 SS225S22235^
0^ 50AIm- 41





















FIG. 13. General Morphological Features of Rigodium:
Sporophytic and Associated Gametophytic Structures. A.
Paraphysis from perichaetium. B. Archegonium. C.
Paraphysis from perigonium. D. Antheridium. E. Calyptra
from developing sporophyte. F. Capsule. G. Stomate at base
of capsule urn. H. Annulus (still attached to urn and
operculum). I. Portion of peristome. J. Spore. A-B from
Kuhnemann 5260 (ALTA, R. adpressum); C-D from Ruthsatz s.n.
(H, R. adpressum); E from Matteri 432 (BA, R. adpressum); F
from Lechler 629 (MICH, R. brachypodium); G from Stevermark
32769 (F, R. toxarion); H from Ventura 18869a (MEXU, R.
toxarion), I from Hammen et al. 2676 (FLAS, B. toxarion); J
from Matteri & Schiavone 4985 (Herb. Matteri, R. adpressum).




















3q


E











0.5mm


A


H 25jm


I I
100jum


" ', /


10Jum


operculum


annulus



urn


50 um


100um





















FIG. 14. General Morphological Features of Rigodium:
Scanning Electron Micrographs of the Peristome. A. Overview
of peristome. B. Apex of exostome tooth showing papillose
outer surface. C. Base of peristome tooth showing cross-
striolations on the outer surface. D. Detail of C. E.
Inner surface of peristome tooth showing lamellae. F.
Detail of E. A from Kunkel 2029 (H, R. brachypodium); B-D
from Barrandequv 345 (MO, R. pseudo-thuidium); E-F from
Neaer 10 (L, R. brachypodium). A: line = 100 um; B, C: = 25
um; D, F: = 10 um; E: = 50 um.







54





















FIG. 15. General Morphological Features of Rigodium:
Scanning Electron Micrographs of Spores. A-D.
Representative spores. A-B from Hollermayer 281 Ep.. (S, R.
brachypodium); C-D from P6cs 6782/B (VBI, R. toxarion). All
lines = 5 um.







56















PHYLOGENY


Cladistic methodology has been developing in zoology

and botany over the past three decades, but has been

infrequently applied to systematic studies in the

bryophytes. The disadvantages of using bryophytes as the

subjects for cladistic study are discussed in detail by

several bryologists, such as Robinson (1985, 1986) and Buck

(1986). However, these problems with cladistics (e.g.,

difficulty of determining homology, choosing outgroups, and

dealing with character loss) exist in "traditional"

taxonomic work as well (Mishler, 1986). The primary

benefits of cladistic analysis are: (1) the care in which

characters is presented: the characters, their polarity, and

their level of universality must be stated explicitly in

order to support a cladogram, and (2) the ability to

construct an explicit hypothesis of genealogical

relationships.

Few studies have incorporated pleurocarpous groups

(including Rigodium) in a phylogenetic framework.

Historically, most bryologists have concentrated on

describing new species and genera; few have dealt with

delimitation of groups based on phylogeny (Buck & Crum,

1990). Family definitions, established at the turn of the









century and generally based on European taxa, became more

and more heterogeneous as newly described genera were added.

Recently, the circumscriptions of moss families and genera,

particularly those of the pleurocarps, are being radically

revised, but no scheme using strict Hennigian principles for

the pleurocarps has yet been published.

Trees showing relationships among pleurocarpous

families have been presented by Vitt (1984) and Buck and

Vitt (1986). Vitt (1984) applies character analysis at the

subordinal level (= orders in other classifications, i.e.,

Bryineae, Hypnineae, Leucodontineae, Hookeriineae) for the

Bryales, but his dendrogram of the families groups them

"according to similarity." In this scheme, Rigodium is

included within Thuidiaceae, shown as a sister group to the

Echinodiaceae. The survey of pleurocarpous families

presented by Buck and Vitt (1986) results in some

revolutionary alignments of families, some of which are

shifted between orders. The authors admittedly use both

symplesiomorphies and synapomorphies in the construction of

their "pseudocladogram." Explanations of the polarities of

characters, as well as their level of universality, are not

often offered but are convincing when given (e.g., for

peristomal development characters and leaf cell shape).

Figure 16 is a redrawn version of their cladogram of the

Hypnales with the cited plesiomorphic characters deleted,

resulting in unresolved branching points in the

superfamilies Hypnacanae and Brachytheciacanae. For









simplicity, autapomorphies are not included here as the

assumption is being made that the families are monophyletic.

The numbers of the character sets, listed in Table 2, follow

the original list of Buck and Vitt (1986). Numbers for the

plesiomorphic and autapomorphic characters are deleted here.

According to Buck and Vitt, the members of the Hypnales

lineage, all pleurocarpous, are characterized additionally

by hypnoid peristomes, elongate leaf cells, horizontal and

asymmetric capsules, and distinctive exostomal ornamentation

(outer surface cross-striolate below and coarsely papillose

above; inner surface with projecting trabeculae). The

Hypnales are further divided into three suborders: the

Hypnodendrineae, Fontinalineae, and Hypnineae. The

Rigodiaceae are classified within the Hypnineae, the most

advanced hypnalean suborder, characterized by well-developed

peristomal development (character 11, shouldered teeth).

The division of this suborder into three superfamilies is

based on gametophytic characters: the families of the basal

superfamily Leskeacanae (in which the Rigodiaceae are

included) are linked by the strong synapomorphy of

secondarily shortened leaf cells, plus a weaker character

cited as a tendency for papillosity (of various types) of

leaf cells. (That the latter character is found in other

groups is not directly addressed by the authors, and

additionally, the various kinds of papillae found in these

and other families are not distinguished.) The authors









postulate that these characters may be adaptations for water

retention in exposed habitats.

The five families composing the Leskeacanae

(Pterigynandraceae, Leskeaceae, Thuidiaceae, Echinodiaceae

and Rigodiaceae) have undergone numerous and ongoing

revisions. The family Pterigynandraceae was erected by Buck

(1980a) as a monotypic family, which has been recently

expanded to include four more genera (Buck & Crum, 1990).

The family is characterized by reduced costae, prorulose

leaf cells, and reduced peristome (smooth exostome;

endostome reduced to slender segments).

The Leskeaceae have historically been considered part

of, or close to, the Thuidiaceae (e.g., Grout, 1932) and due

to the reduced peristome of some genera, have even been

included within the Isobryales (Leucodontales; Buck, 1980a;

Vitt 1982, 1984). Buck and Vitt (1986) recognize a more

restricted Leskeaceae, as first suggested by Crum (1973),

with Anomodon and allies excluded. According to their

definition, the family is reduced to Leskea and Pseudoleskea

plus their associated relatives and are characterized by

strong tendencies of cells to be papillose over the lumina,

limited presence of paraphyllia, and orange exostome with

reduced striations (usually papillose). Recently, Buck and

Crum (1990) transferred several genera (e.g.,

Brvohaplocladium and Claopodium, discussed below) to the

Leskeaceae from the Thuidiaceae, primarily recognizing the

family on the basis of gametophytic characters and









perceiving "a habitat-driven, sporophytic reduction series

with Leskea as the reductionary extreme."

The Thuidiaceae s.l., sometimes combined with the

Leskeaceae, historically included several subfamilies (see

Watanabe, 1972). Buck and Vitt (1986) excluded Anomodon

from the family and simply defined it on the basis of

papillose leaf cells and abundant paraphyllia. Buck and

Crum (1990) further reduced the family to only the subfamily

Thuidioideae, with members of the other subfamilies

transferred to other families (discussed below).

The Echinodiaceae, a small monotypic family, have

sometimes been included in the Isobryales (e.g., Walther,

1983), but according to Buck and Vitt (1986) and Churchill

(1986), the plants have a standard hypnalean peristome. The

coarse, poorly branched plants have an unusual (and possibly

relict) Macaronesian-Australian-New Zealand distribution.

The monophyly of the superfamily Leskeacanae is based

on Buck's and Vitt's suite of characters represented by "12"

in Fig. 16; of these, the secondarily derived short leaf

cells is their strongest and best defined synapomorphy. For

the purposes of this study, the assumption is being made

that the superfamily is a natural and monophyletic group,

until such a time when a higher level phylogenetic analysis

(based on shared derived characters) is completed on all the

families of the Hypnales (as well as for the Leucodontales

and Hookeriales). These families continue to undergo much

redefinition, including transfer of genera between them









(even since the cladogram of Buck and Vitt was published).

The following representative genera of the superfamily have

been chosen for a preliminary cladistic analysis in an

attempt to better ascertain the phylogenetic position of

Riqodium. The affinities of several of the genera

themselves, as outlined below, have also not been agreed

upon as they are shifted around between families of the

Leskeacanae.

Pteriqynandrum, a monotypic genus (Buck, 1980b), was

included in the Entodontaceae (Crosby & Magill, 1981; Vitt,

1982) and tentatively in the Leskeaceae in the Isobryales

(Vitt, 1984). The genus initially comprised the

Pterigynandraceae (Buck, 1980a) until Buck and Crum (1990)

expanded the family. Buck (1980b) stated that the genus was

clearly an isolated group with no obvious connections to

other genera, but Buck and Crum (1990) acknowledge a

relationship with Leskeaceae and Thuidiaceae.

Leskea and Pseudoleskea represent the major groups of

Leskeaceae s.s., with Pseudoleskea characterized by a normal

hypnalean peristome, and Leskea, having the reductionary

extreme of the sporophyte.

Thuidium, Brvohaplocladium, Claopodium, Heterocladium,

and Helodium represent the diverse Thuidiaceae s.1.

Thuidium, Bryohaplocladium [= Haplocladium nom. ille fide

Watanabe & Iwatsuki (1981)], and Claopodium have been

included in the subfamily Thuidioideae s.l. The close

relationship between Claopodium and Brvohaplocladium has









long been recognized (Noguchi, 1964; Watanabe, 1972), and

the genera have been sometimes included within the

Anomodontaceae, as well as the Thuidiaceae. Buck and Crum

(1990) recently transferred them to the Leskeaceae as part

of the reduction series within the family. Of note here is

that Noguchi (1964) in his monograph of Claopodium suggested

that it may possibly not be monophyletic. Heterocladium,

previously within the subfamily Heterocladioideae of the

Thuidiaceae s.l., has been added to the expanded

Pterigynandraceae (Buck & Crum, 1990) as part of a reduction

series within that family. Helodium, a small genus of three

species (Abramova & Abramov, 1972), was formerly included

within the subfamily Helodioideae. Buck and Crum (1990),

somewhat unsure of its placement, erected the new monotypic

family, Helodiaceae, to accommodate it.

Echinodium is the small genus of six species comprising

the Echinodiaceae. According to Churchill (1986), the

monophyly of the genus is questionable since it is not well-

defined on the basis of shared derived characters.

The problems in placement and previous inclusion of

Rigodium within various families (Brachytheciaceae,

Lembophyllaceae, Thuidiaceae) are discussed in detail in the

section on "Taxonomic History." Crum (1981, 1984) recently

circumscribed the monotypic Rigodiaceae.

Phylogenetic reconstruction and character polarization

depend on the proper selection of an outgroup. Often a

functional or provisional outgroup is the best available









choice (see Watrous & Wheeler, 1981). According to Buck and

Vitt (1986; Fig. 16), the suborders Fontinalineae

(Fontinalaceae) and Hypnodendrineae (Pterobryellaceae,

Thamnobryaceae, Pleuroziopsidaceae, and Hypnodendraceae) are

the outgroups for the suborder Hypnineae (including

superfamily Leskeacanae). The choice of using these groups

is problematical for the following reasons. Most of these

families had been scattered among several orders and are

tentatively allied by Buck and Vitt in the Hypnales for the

first time. The use of characters of the Fontinalaceae

(traditionally in the Isobryales) is difficult due to the

extreme modifications in both the sporophyte and gametophyte

for an aquatic habitat. The synapomorphies (character set

2: exostome teeth long, gametophores dendroid, endostome

segments gaping, leaves coarsely serrate) for the

Hypnodendrineae may be weak, as arguments concerning the use

and polarities of these particular characters are not

presented. Two of these families, the Pterobryellaceae and

Thamnobryaceae, were previously subfamilies of the

Neckeraceae (subfamily Pterobyelloideae and Thamnoideae,

respectively) in the order Isobryales (Vitt, 1982, 1984).

(Buck and Vitt maintain the rest of the Neckeraceae in the

Isobryales.) The monotypic Pleuroziopsis comprises the

Pleuroziopsidaceae (Ireland, 1968), previously allied with

the Climaciaceae (Isobryales), but now segregated into its

own family due to several unusual characters. The

Hypnodendraceae have traditionally been considered as an









acrocarpous family in the Bryales (Vitt, 1982, 1984), but

Buck and Vitt (1986) follow Touw (1971) in interpreting the

habit as having compressed pinnate branching.

For the preliminary investigation of the representative

genera of superfamily Leskeacanae, Pleuroziopsis

(Pleuroziopsidaceae) and Thamnobryum (Thamnobryaceae) were

chosen as provisional outgroups for rooting the trees

presented here (Figs. 17, 18). As discussed below, the

results of these analyses suggest Echinodium as the possible

outgroup for the cladistic analysis of RiQodium species,

enabling the polarization of characters within Riqodium in

an effort to determine the possible phylogenetic

relationships of the species. These cladistic analyses were

generated using the software program Phylogenetic Analysis

Using Parsimony (PAUP), Version 3.0 (Swofford, 1990) on a

Macintosh computer. PAUP and similar software are

summarized and compared by Platnick (1987). The cladograms

were rooted using the particular outgroups as indicated in

the discussions below under the appropriate analyses. The

searches for the shortest trees included the heuristic

option "MULPARS" (multiple parsimony) along with branch-

swapping, as well as branch-and-bound. Equivalent results

were obtained with both methods. When more than one minimal

tree resulted from an analysis, a strict consensus tree also

was calculated. Characters for multistate taxa, as coded in

Tables 3 and 5, were included and interpreted as

"polymorphism" (as opposed to "uncertainty").










Autapomorphies were excluded from the computer analyses (so

as to not influence the consistency index); those of the

Rigodium species are listed in the discussion of that

analysis. For autapomorphies of the families for

superfamily Leskeacanae, see Buck and Vitt (1986).

Characters for Analysis of Representative Genera of
Superfamily Leskeacanae

Out of numerous possible features examined, only those

with the potential for elucidating phylogenetic

relationships (i.e., those potential synapomorphies) were

included in the PAUP analyses. Table 3 represents the basic

data set for the three analyses of the ten representative

genera of superfamily Leskeacanae (plus two outgroups). The

18 characters are listed numerically and discussed briefly

below.

1. Development of the stipe (0 = well-developed). As

discussed in the section on morphology, the stipe is the

unbranched part of the lower stem and is an indication of

the development of an arborescent habit type.

2. Differentiation of secondary orders of branching (0

= differentiated branching). Some genera, such as Thuidium

and Thamnobryum are characterized by secondary (and even

tertiary) branching, while in others, such as Leskea, the

primary branches do not ramify.

3. Organization of the axes into stems vs. branches (0

= well-organized stems and branches). In Leskea and

Pseudoleskea, for example, the few branches that do occur










are not very well-developed and are not distinguishable from

the main stems. In Thuidium, or even Bryohaplocladium (with

essentially no secondary branching), the branches are

numerous and well-developed.

4. Presence and types of paraphyllia (0 = no

paraphyllia). Riqodium lacks these structures, which are

discussed in detail under morphology. The three forms of

paraphyllia filamentouss, branched and foliose) are numbered

(but not ordered) differently in the matrix (see Table 3).

Note that a genus may be characterized by more than one

morphology.

5. Presence of pseudoparaphyllia (= 0).

Pseudoparaphyllia, also absent in Ricodium, are discussed

with paraphyllia in the section on morphology.

6. Dimorphic stem and branch leaves (0 = no

dimorphism). In certain genera (e.g., Rigodium and

Thuidium), the stem leaves are broadly ovate at the base and

taper to an acumen, but the branch leaves are narrowly ovate

and acute. Pleuroziopsis, with very broadly ovate and blunt

stem leaves and ovate, deeply toothed branch leaves, here is

considered to have a different type of leaf dimorphism.

7. + Subulate stem leaves (0 = non-acuminate leaves).

These stem leaves are ovate at the base and more or less

abruptly taper to an acumen (e.g., Claopodium and Rigodium).

This feature is related to character 7, but not all genera

with this stem leaf shape have dimorphic stem and branch

leaves (e.g., Echinodium).









8. Reduced costa (0 = costa present). A very weak to

absent costa is present in genera such as Heterocladium (as

well as certain species of Rigodium).

9. Quadrate alar cells in rows (0 = alars not

differentiated). Pterigynandrum, Leskea and Pseudoleskea

share this feature, but in the other genera, the alar region

is either undifferentiated or not differentiated in this

manner.

10. Prorulose midlaminar cells (0 = cells non-

prorulose). A distinction is made here between papillose

("11" below) and prorulose cells, in which the end walls

project dorsally, sometimes even forming well-developed

papillosities (prorulae) as in Pteriqynandrum and Rigodium.

11. Papillose leaf cells (0 = smooth leaf cells). A

loose interpretation of "papillose" is applied here,

including mammillose (low papillose) and multiple papillate

forms. Mammillose cells bulge in the lumen as well as in

the wall. Basically, all of these wall projection types

occur centrally in the cell.

12. Autoicous sexual condition (0 = dioicous). This

indication of sexuality was included in the analysis as a

few taxa (e.g., Bryohaplocladium and some species of

Helodium) differed from the rest of the group (dioicous).

13. Lack of cilia (0 = cilia present). This character

is an indication of a reduced hypnalean endostome, as found

in Leskea, Pseudoleskea, and Pteriqynandrum.










14. Reduced ornamentation of exostome (0 = standard

exostome ornamentation). As discussed by Buck and Vitt

(1986), the standard hypnalean exostome tooth is cross-

striolate below and papillose above; teeth of Leskea and

Pseudoleskea demonstrate irregularly striolate to entirely

papillose ornamentations.

15. Erect capsules (0 = inclined to horizontal

capsules). (See below).

16. Symmetric capsules (0 = asymmetric capsules).

Erect and symmetric capsules may be correlated characters.

A typical hypnalean capsule is horizontal to inclined and

asymmetric in shape.

17. Reduction of the annulus (0 = well-developed

annulus). In Leskea and Pseudoleskea, the annulus is

reduced to absent.

18. Conic operculum (0 = rostrate operculum).

Opercula are either rostrate (e.g., Rigodium) or lack the

beak (conic), as in Leskea.

Results of Analyses of Genera

Three analyses were evaluated using the basic matrix in

Table 3. In one analysis, Thamnobryum functioned as the

outgroup (Pleuroziopsis deleted), and in the second,

Pleuroziopsis was the designated outgroup (Thamnobryum

omitted). In the third analysis, both genera were included

as outgroups. Thus, certain characters are polarized

differently according to the particular outgroup(s) chosen.










Only one most parsimonious tree (50 steps, CI=0.760;

Fig. 17) was generated when Thamnobryum was used as the

outgroup. Echinodium is basal (as opposed to terminal in

Buck and Vitt's cladogram, Fig. 16), with Riqodium

positioned between it and Thuidium. In addition, Leskea-

Pseudoleskea-Pteriqynandrum, basal in Fig. 16, are the most

derived groups here. The cladogram also does not support

Buck and Crum's (1990) transfer of Bryohaplocladium and

Claopodium from Thuidiaceae to Leskeaceae, as inclusion of

these two genera with Leskea and Pseudoleskea in one family

would result in a polyphyletic group. Bryohaplocladium and

Claopodium appear phylogenetically closer to Leskea than to

Thuidium, however. Likewise, Pterigvnandrum and

Heterocladium (Pterigynandraceae; Buck and Crum, 1990) are

not shown as sister groups. The several homoplasies on the

lineage leading to Pteriqynandrum, however, may indicate

problems with its placement in this cladogram.

Thirteen characters (1, 2, 5, 6, 7, 9, 11, 13, 14, 15,

16, 17, 18) do not exhibit parallelism, although reversals

(in 5, 6, 7, 14, 11) and polymorphism (5, 11, 15, 16, 18; in

multistate taxa) occur. Also, six of these characters (9,

13, 15, 16, 17, 18) are associated with the Ptericvnandrum-

Leskea-Pseudoleskea group.

Rigodium and Echinodium (and other taxa; see Fig. 17)

share characters 7 acuminatee leaves) and 11 (papillose

cells); the synapomorphy of lack of pseudoparaphyllia (5)

and dimorphic leaf shape (6) link Rigodium to Thuidium (and









other taxa). The Rigodium lineage is characterized by

prorulose cells (10, a feature showing homoplasy), and the

polymorphic features of the weak costa (8) and papillose

cells (11), change from 0->1,0 and 1->1,0, respectively.

In the second analysis (Pleuroziopsis as outgroup), the

results are much less resolved; 27 equally parsimonious

trees (53 steps, CI=0.717) were discovered. The consensus

tree (Fig. 18B) is basically polytomous, although

Pteriqynandrum, Leskea, and Pseudoleskea are established as

a monophyletic group. However, one of the 27 trees is

identical in topology to the cladogram derived with

Thamnobryum as the outgroup (Fig. 17), and in many trees,

(e.g., Fig. 18A), Echinodium is basal within the group, with

Ricodium phylogenetically adjacent.

With Thamnobrvum and Pleuroziopsis functioning as a

combined outgroup, 33 most parsimonious trees were generated

(54 steps, CI=0.704). The use of both taxa simultaneously

to root cladograms was problematical as one or the other of

these genera appears within the ingroup in several of the

trees (i.e., the tree could not be rooted so that the

ingroup was monophyletic). This issue weakens the validity

of using both these taxa as outgroups. However, several of

the trees (where the combined Pteriqynandrum-Thamnobryum

does appear as the outgroup) are similar in topology to

those discussed above (e.g., Fig. 17) and have Echinodium in

a cladistically basal position, followed by Riqodium and

Thuidium.









The consensus tree of the previous analysis (Fig. 18B)

indicates that more detailed phylogenetic investigations

(with more genera) are needed to better define the

relationships of the families within the Leskeacanae. The

present analyses do support maintaining Rigodiaceae as a

monotypic family, since inclusion with Thuidiaceae

(Thuidium) would result in a paraphyletic family.

Characters for Rigodium Species

The matrices of Tables 4 and 5 are the data sets used

in the analyses of Rigodium species. Eight characters were

polarized using Echinodium as the outgroup. Although the

number of OTUs differ (see discussion following), the number

of characters remained constant for both analyses. Six

characters are qualitative, and two are quantitative (one

continuous and one discontinuous). The problems in forming

taxonomic characters from measurements are summarized in

Almeida and Busby (1984). The approach in this revision

(see Fig. 19A and B) involves the traditional method of

scanning the full range of the material and partitioning the

scale into classes of varied range. The character states

are, thus, somewhat arbitrarily delimited. Only the

characters with an obvious gap (number of cells in the

intramarginal band, Fig. 19A; stem leaf decurrency/total

leaf length, Fig. 19B) were used in the cladistic analysis,

although three other characters (intramarginal vs.

midlaminar cell length, Fig. 19C; stem leaf length/width,

Fig. 19D; and stem leaf acumen length/length from point of









insertion, Fig. 19E) show basic trends. These three latter

features are discussed under the appropriate species

discussions following the diagnoses in this revision.

The following lists the characters as their derived

state ("1" in Tables 4 and 5; character 6 has two derived

states, "1" and "2.")

1. Rooting at tips. As discussed in the morphology

section, several species of Rigodium (e.g., R. toxarion)

develop rhizoid clusters at the tips of stem and branches

and thereby form new plants.

2. Loss of well-differentiated stipe. Two species (R.

implexum and R. pseudo-thuidium) lack the arborescent habit

and the associated well-developed stipe (lower unbranched

portion of the stem).

3. Right angle branching. In R. brachypodium var.

tamarix (Fig. 38A), B. implexum (Fig. 44G), and R. pseudo-

thuidium (Fig. 46A), the primary branches are spreading,

evenly spaced and arise at right angles to the stem.

4. Decurrent stem leaves. Stem leaves were ranked as

"decurrent" when the stem leaf decurrency/total leaf length

ratio exceeded 8% (Fig. 19B). Note that a percentage

greater than 14 is an autapomorphy for R. adpressum.

5. Uneven midlaminar cell walls. In R. toxarion, (as

in Echinodium), these cell walls are not very porose (Fig.

27J), compared to the other species with conspicuously

uneven walls (e.g., R. pseudo-thuidium, Fig. 46M).









6. Visibly prorulose midlaminar cells. In R.

toxarion, the cells are prorulose, but not easily

discernible (Fig. 27J), while in R. adpressum, R. implexum

and R. pseudo-thuidium these cells are noticeably thickened

at the end walls (e.g., Fig. 46M). In R. brachypodium (both

varieties) the end walls project as dorsal papillae

(prorulae; Fig. 10C and D). In the matrices, the thickened

but not papillose condition is indicated as "1," and the

papillose extreme, as "2." This does not, however, mean

that condition "2" is necessarily more derived that "1;"

other than the fact that "0" is basal (outgroup comparison),

the actual sequence of events (i.e., 0->1->2, 0->2->1, or

1<-0->2) is unknown here (see discussion of results).

7. Seven or greater cell rows in intramarginal band.

A natural division in the number of cell rows is apparent at

seven cells (Fig. 19A), with R. implexum and R. pseudo-

thuidium both characterized by this derived condition..

8. Extremely thickened cell walls of intramarginal

band. In E. toxarion (Fig. 27K) the intramarginal band is

barely visible, even in cross section (Fig. 271). In the

other four species, the walls are extremely incrassate and

easily distinguished from the midlaminar cells, even in

surface view (e.g., Fig. 46L).

Results of Analyses of Rigodium Species

The data sets (Tables 4 and 5) for the two analyses

were identical except for the configuration of R.

brachypodium. In the first analysis, this species was split









into its two varieties, R. brachypodium var. brachypodium

and R. brachypodium var. tamarix, in order to test the

monophyly of the species (Table 4). In the second

examination (Table 5), the species was assumed to be

monophyletic with two polymorphous character states (3,

right angle branching; 4, decurrent stem leaves).

Figure 20 is the single most parsimonious tree (11

steps, CI=0.818) for Riqodium species with the two varieties

of R. brachypodium designated as separate OTUs. Ricodium

toxarion is basal, followed by R. brachypodium and R.

adpressum, and R. pseudo-thuidium and R. implexum are

derived sister-species. Of interest is that R. brachypodium

is not monophyletic here, but paraphyletic.

The only parallelism involves character 3 (right angle

branching) for R. brachypodium var. tamarix and the R.

implexum-R. pseudo-thuidium clade, which may reflect that

the ramification is actually different in these two

instances. In R. brachypodium var. tamarix the branches are

regularly pinnate (Fig. 38A), while the branching of the

other two species is irregularly pinnate. In fact, the

habit type of R. implexum (Fig. 22A) may have been derived

from a form like R. pseudo-thuidium with several innovations

arising from the main stem (Fig. 22B); see also discussion

under R. pseudo-thuidium).

The only reversal in the cladogram is where "rooting at

the tips" (1) was lost for the R. implexum-R. pseudo-

thuidium clade, possibly because this feature is correlated









with the loss of the aborescent or frondose habit. This

analysis also suggests that the direction within character 6

(degree of prorulosity) is 0->2->1, i.e., the less prorulose

state (1) is secondarily derived from the extreme papillose

type (2).

In the second analysis, R. brachypodium was considered

to be a monophyletic multistate taxon (polymorphic

characters 3 and 4; Table 5). Two equally parsimonious

cladograms were generated (12 steps, CI=0.917; Fig. 21A and

B), with one tree (Fig. 21B) equivalent to the strict

consensus tree. In contrast to the previous analysis, where

R. adpressum is more derived than R. brachypodium (which is

paraphyletic), the position of the former species here is

either phylogenetically intermediate between R. toxarion and

E. brachypodium (Fig. 21A) or unresolved (Fig. 21B). The

other portions of the tree (R. toxarion basal; R. implexum-

R. pseudo-thuidium terminal sister-taxa) are supported as in

the previous analysis.

Within the multistate E. brachypodium, homoplasy is

indicated for characters 3 (right angle branching) and 4

(decurrent stem leaves). Also, the direction of evolution

for character 6 is towards increasing degree of prorulosity

(0->1->2), instead of the secondarily derived reduction in

the previous analysis.

For both analyses, character 1 (rooting at tips--which

reverses within the terminal portion) is a defining feature

of Rigodium. In addition to this feature, autapomorphies of









the genus (i.e., synapomorphies of its species) include the

presence of a differentiated intramarginal band of cells and

extremely incrassate cells. Thus, the genus is clearly

monophyletic.

Taxonomic Philosophy Concerning Species

As suggested in Budd and Mishler (1990), the working

definitions of species used by taxonomists are, in general,

either phenetic or basically cladistic. In this revision, a

species was defined primarily as a phenetically cohesive

unit that is separated from other similar entities by a

consistent morphological gap. Monophyly cladisticc

approach) was examined secondarily in the analyses discussed

above. Rigodium toxarion (the basal species) is

metaphyletic (Donoghue, 1985; Mishler, 1985), as no

autapomorphies define it. Rigodium adpressum, R. implexum

and R. pseudo-thuidium are monophyletic, each distinguished

by the following autapomorphies: (1) R. adpressum: appressed

leaves, strongly revolute leaf margins, decurrency of stem

leaf greater than 14% of total leaf length (Fig. 19B), low

central papillae in some cells of the stem leaf decurrency,

and intramarginal cells shorter than the midlaminar cells

(Fig. 19C); (2) R. pseudo-thuidium: reduction of secondary

and loss of tertiary branching and loss/reduction of stem

leaf costae; and (3) R. implexum: free-living, terrestrial

"tumbleweed" habit with extremely wiry axes and profuse

branching. As discussed above, R. brachypodium may be

monophyletic or paraphyletic. Since the strong character 6









(end walls with prorulae) and intermediate habit forms

(detailed in the discussion of R. brachypodium var.

brachypodium) link the two varieties, from a practical

standpoint this taxon is considered as a single species in

this revision, and it is considered probably monophyletic.

A variety here is defined as a monophyletic (or

metaphyletic) group within a species that is separated from

other such entities by a small morphological gap (Wilken &

Hartman, 1991). Thus, the variety in the traditional sense

of an allopatric race (Mayr, 1969), i.e., having one to few

conspicuous differences associated with a certain

geographical range, is not applied in this revision (see

Stuessy, 1990, for discussion on infraspecific categories).

The category of variety is applied within two species, R.

toxarion and R. brachvyodium. Although both varieties of R.

toxarion are sympatric, the traditional varietal concept

somewhat applies since R. toxarion var. robustum is

restricted to the Juan Fernandez Islands. This variety is

additionally defined by the autapomorphy of narrow branch

leaves (see details under discussion of R. toxarion var.

robustum), while R. toxarion var. toxarion is metaphyletic.

As discussed above, the two varieties of R. brachypodium

here are considered to comprise one species due to the

presumed synapomorphy of prorulae, as well as intermediates

in habit type. Rigodium brachypodium var. tamarix is

defined by the autapomorphy of the habit type (elongated







79

stems with regularly pinnate branching), while R.

brachypodium var. brachypodium is considered metaphyletic.

















0)
0 (0




w1




03-i
to

Cr O
) a
Q3i











4.)


00)
04
0*4



















oE-1
C
aO

















- 0)
0


a) o
cu




04



4 4
0)







~-P
>* 0
*C


ax




CO *










Xi4-
u w
OQi-



















w
w
<
w z



.wZa.


cDo#
^2








Table 2. A list of the synapomorphic characters for the
order Hypnales as featured in the cladogram (Fig. 16)
modified from Buck and Vitt (1986). The characters
indicated by the authors as symplesiomorphic and
autapomorphic are deleted here, hence the
discontinuities in the numbering system on the
cladogram and for their characters listed below.


1. Peristome hypnoid; capsules horizontal, asymmetric; leaf

cells elongate

2. Exostome teeth long; gametophores dendroid; endostome

segments gaping; leaves coarsely serrate

3. Leaf cells thin-walled; alar cells thin-walled, quadrate

5. Leaf cells long

7. Opercula + rostrate; leaves with thickened margins

11. Exostomial shoulder well-developed

12. Leaf cells + papillose; alar cells reduced; leaf cells

secondarily short; plants xerophytic

13. Leaves subulate

14. Capsules erect; costa weak; paraphyllia reduced;

peristome reduced

16. Stem leaves differentiated; branching well-developed

21. Leaf cells long

22. Leaves varying around lanceolate

25. Peristome yellow-brown; costa poorly developed; leaf

cells shortened

26. Alar cells small; capsules short; substrate of wood and

rocks









Table 2--continued


29. Alar cells covering costal base; pseudoparaphyllia

filamentous

30. Costal spine present

34. Leaf cells short; peristome reduced; plants epiphytic

37. Costa lost

40. Exostome shoulder amplified; opercula long-rostrate;

alar cells amplified (shape)

45. Plants flattened








84
Table 3. Data matrix for the cladistic analysis of selected
genera of the superfamily Leskeacanae. "?" = missing
data. Polymorphous characters are indicated as: a =
1,2,3, b = 1,2; c = 0,2; d = 0,1. Thamnobryum and
Pleuroziopsis functioned as outgroups in the analyses.


Genus Characters



1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8
111111111
123456789012345678

Thamnobryum 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Pleuroziopsis 0 0 0 0 1 0 0 1 0 1 0 0 1 0 0 0 ? ?

Bryohaplocladium 1 1 0 a 1 1 1 0 0 1 1 0 O O O 0 d

Claopodium 1 1 0 c 1 d 1 0 O 0 1 0 O O O O O O

Echinodium O0 1 0 O 0 1 0 O 0 d 0 O O O O O 0

Helodium 1 1 0 a 1 O 1 0 0 1 1 d O O O O O O

Heterocladium 1 1 0 b d 1 1 1 0 1 1 0 O O O 0 d

Leskea 1 1 1 b 1 0 0 0 1 0 1 1 1 1 1 1 1 1

Pseudoleskea 1 1 1 b 1 0 1 0 1 0 1 0 1 0 dd 1 1

Pterivqnandrum 1 1 1 0 0 0 0 1 1 1 0 0 1 1 1 1 0 0

Riqodium O O O0 1 1 1 d 0 1 d O O O O O O0

Thuidium 1 0 0 a 1 1 1 0 O 0 1 0 O O O O O O





















FIG. 17. Most parsimonious tree for representative
genera of the superfamily Leskeacanae using Thamnobryum as
the outgroup. Characters are listed in the text; for the
character states see data matrix (Table 3).

























o 0 0 0 -


' parallelism
=reversal


polymorphic characters
a =1,2,3 c = 0,2
b=1,2 d =0,1


tham





















FIG. 18. Representative trees for the genera of the
superfamily Leskeacanae using Pleuroziopsis as the outgroup.
A. Sample tree (out of 27 equally parsimonious cladograms).
Characters are listed in the text; for the character states
see data matrix (Table 3). B. Consensus tree.







88




. o 0 o 0 0 L- .
S C) r) D
0 "-C r- 0 -C C. _. Q.

1a2 6d' c8,11 4b1' 14
18d \ 4 12 b' .15d',16d'
a4a' (2 1718
8 1 \ \12d' 10'
4b'1
4a' 5d' 7,13
8d' 10' 18d' 3'15',16'
3' 1 1d' 10'
5
10' 6
11d'


2y parallelism
= reversal
1
4(0-2) polymorphic characters
a = 1,2,3 c = 0,2
b=1,2 d=0,1
7,11,13
pleu a'


o o -oo r- o o- o "
S 0 _? M.





















FIG. 19. Morphological Trends in Rigodium species. A.
Number of cell rows in the intramarginal band of stem
leaves, showing division at 8 cells. B. Range of variation
in stem leaf decurrency length/total leaf length, showing
gaps at 8 and 13. C. Range in intramarginal and marginal
cell lengths. D. Range in stem leaf length/width. E.
Range in stem leaf acumen length/leaf length from point of
insertion. TOX = R. toxarion; BRAB = R. brachypodium var.
brachypodium; BRAT = R. brachypodium var. tamarix; ADP = R.
adpressum; PSE = E. pseudo-thuidium; IMP = R. implexum.
Asterisks (*) in B and C indicate autapomorphies (for R.
adpressum). Character groupings depicted in A and B were
used in cladistic analyses.




















t--- m-- r

v -$
..--o



-NrC
------
m---<


4 6 8 10
Number of Cells


B
IMP

PSE

ADP

BRAT

BRAB

TOX


V V

-I
*


I I


5 10 15 20
Decurrency L/Total Leaf L


I------ --

I --------------------l
II
-.-- .-- -

r ^---^

I ---------
S--------------
-
.


10.0 20.0
Cell Length (um)


I III


0.75 1.25 1.75
Stem Leaf L/W


30.0 40.0
intramarginal
--- -imidlaminar


IMP

PSE

ADP

BRAT

BRAB

TOX


I II


20 o 0 60
Acumen L/Leaf L


A
IMP

PSE

ADP

BRAT

BRAB

TOX


C IMP

PSE

ADP

BRAT

BRAB

TOX







D
IMP

PSE

ADP

BRAT

BRAB

TOX


- -' -- --'- --- -- ---- ----- ---- --- -- ----










Table 4. Data matrix for the cladistic analysis of
Rigodium species where the varieties of R.
brachypodium are treated as separate OTUs.
Echinodium functioned as the outgroup in the
analysis.


Species


Echinodium

R. toxarion

R. brachypodium var. brachypodium

R. brachypodium var. tamarix

R. adpressum

R. implexum

R. pseudo-thuidium


Characters

1 2 3 4 5 6 7

0 0 0 0 0 0 0

1 0 0 0 0 0 0


1 0

1 0

1 0

0 1

0 1


2 0

2 0

1 0

1 1

1 1






















FIG. 20. Most parsimonious tree in analysis of
Rigodium species with the two varieties of R. brachypodium
("brat" and "brab") treated as separate OTUs. For
characters see the text; for the character states see Table
4.