.avl4'!

 

 

 

This is to certify that the

thesis entitled

A Life History Study of Diarrhena
americana Beaum: the Rhizome System

 

 

presented by
Donna Elizabeth Herendeen

has been accepted towards fulfillment
of the requirements for

Masters Botany

degree in

/ é. . Siephenson

Major professor

 

 

 

DMe February 20, 1987

 

0.7639 MS U i: an Affirmative Action/Equal Opportunity Institution

 

MSU

 

LIBRARIES
._:_‘—n_

RETURNING MATERIALS:

 

 

 

Place in book drop to
remove this checkout from
your record. FINES wil]
be charged if book is
returned after the date
stamped beiow.

 

 

AUG 3&7; 293%

Uu'

 

 

A LIFE HISTORY STUDY OF DIARRHENA AMERICANA BEAUV.:
THE RHIZOME SYSTEM

 

BY

Donna Elizabeth Herendeen

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Botany and Plant Pathology

1986

ABSTRACT

A LIFE HISTORY STUDY OF DIARRHENA AMERICANA BEAUV.:
THE RHIZOME SYSTEM

by
Donna Elizabeth Herendeen

The relationship between the rhizomatous growth form of
Diarrhena americana and its tendency to form populations
consisting of varibly sized patches was investigated.

Intact rhizome systems were removed form the field, prepared
and mapped for examination of the growth form and for the
interpretation of patch formation. Rhizome development was
found to be sympoidial and rhizome branches or segments were
found to persist for several years. Small patches (< 40 cm
diameter) were interpreted as single clones: larger patches
( > 150 cm diameter) were either single clones or two or
more merging clones.

Clonal structure was examined through culm densities
and above ground culm dry weights of complete small patches
and from transects through large patches. The results of
these investigations reveal zones of differential culm vigor
and reproduction.

Observations made in the field and on transplanted
clones indicate that phenology, flowering and resource

allocation may be influenced by light intensity.

To my parents

ACKNOWLEDGEMENTS

I would like to acknowledge the assistance and guidance
of my major professor Dr. S. N. Stephenson. I would like to
thank my committee members Dr. F. Ewers and Dr. J. Hancock
for their valuable advice and comments. I would like to
thank Dr. P. Werner for helpful suggestions during the
initial planning stages. I would like to thank Mr. P.
Fields for assistance with field work and data collection.
Finally, I would like to acknowledge the assistance, support

and considerable patience of Mr. Patrick Herendeen.

iii

TABLE OF CONTENTS

List of Tables............................................vi
List of Figures..........................................vii
Introduction...............................................l
Literature review..................... ..... ................3

Methods.
Establishment of permanent field transects and
quadrats.............................................11
Data collection on field transects....................12
Whole clones..........................................14
Data collection on whole clones.......................l6
Maps..................................................l7
Fragmented patches....................................18
Seeds and seedlings...................................19

Field site description.
Location..............................................20
Soil..................................................20
Vegetation............................................21
Light.................................................22

Climate...0....O...0.00.00.00.0000000000000000000I0.0.22

Results and Discussion.
Initial ovservations..................................28
Field transects.
Results...........................................3O
Discussion........................................34
Clone structure and organization.
Results...........................................38
Above ground patterns.
Whole clone transects.....................38
Whole clone transplant study..............40
Below ground patterns.
Whole clone rhizome study.................44
Growth pattern of culms and tillers.......44
The age of clones.........................61
Whole clone fragments.....................62
Discussion........................................62
Seedlings.............................................69

seeds. 0 O I O O O O O O O I O O O O O O O O O O O O O O O O O O O I O O O O O O O I O O O O O O O O .71
General Discussion........................................74
conCIUSions. O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O .78

Literature Cited.0.0...O...OOOOOOOOOOOOOOOOOOOOO0.00.00.0080

iv

Appendix A Test of goodness of fit to a uniform
distribution of field transect culms.....................84
Appendix B Test of goodness of fit to a uniform
distribution of field transect fall tillers..............85
Appendix C Correlation of field transect dry weight
and RE...................................................86
Appendix D Test of goodness of fit to a uniform
distribution of whole clone transect culms...............87

Appendix E
dry weight
Appendix F
Appendix G
Appendix H
Appendix I
Appendix J

Correlation of whole clone transect quadrats

and RE........................................88
Correlation of RE and mean culm weight........89
Summary of culm distribution..................9O
Summary of tiller-culm relationships..........92
Seed weights..................................93
Data tables...................................94

LIST OF TABLES

1. Frequency of tiller types per culm base. 59
2. Percentage of culm bases with one, two, three or four

tillers. 6O
3. Culm and tiller summary for rhizome fragments. 63

4. Harvested culms from field transects: test of goodness
of fit to a uniform distribution. 84

5. Field transect fall emerging tillers: test of goodness
of fit to a uniform distribution. 85

6. Correlation of mean culm dry weight and RB values for
each quadrat along field transect clones. 86

7. Number of harvested culms from whole clone transects:
test of goodness of fit to a uniform distribution. 87

8. Correlation of mean culm dry weight and RE values for
each quadrat along whole clone transects. 88

9. Data table and correlations of RE, percent reproductive
culms and mean culm weight for whole clones and field
transects. Values correlated are based on total above
ground biomass harvested. 89

10. Summary of culm distribution illustrated by the whole
clone maps. 90

11. Summary of tiller-culm relationships illustrated by
whole clone maps. 92

12. Individual seed weights. 93

13. Field transect dry weights. Above ground dry weight
only. Reproductive effort for each quadrat and the entire

transect are included. 94
14. Culm tagging results. 95
15. Field transect harvest and fall tiller data. 96
16. Whole clone transect data. 97
17. Whole clone data. 98

18. A sample of apical spiklet, basal floret seed weights
from 23 inflorescences from clone 8. 99

vi

LIST OF FIGURES

1. Light levels at plant height at the Van atta Road

field site. 23
2. Summary of temperatures for the East Lansing area. 24
3. Precipitation for the East Lansing area. 26

4. Field transects compared. Total number of culms
compared for field transects 27, 33 and 35 29

5. Histogram of culm number for field transects 35, 27
and 33. Shaded area represents number of reproductive
culms. 31

6. Histogram of fall emerging tillers on field transects
35, 27 and 33. Each bar represents the total number of
fall tillers for each quadrat along the transect. 32

7. Field transects 35, 27 and 33. Each bar represents
total standing crop for each quadrat and is divided
into reproductive (shaded) and vegetative portions. 33

8. Summary figures for transects 27, 35 and 33 comparing
total number of tillers and number of reproductive culms
along each transect. 35

9. The total number of culms in quadrats along the whole
clone (patch) transects. 39

10. Histograms of total above ground dry weight from whole
clone (patch) transect quadrats. Dry weights are parti-
tioned into reproductive (shaded) and vegetative portions.
RE for each quadrat is above each bar. 41

11. Comparison of area and total number of culms for whole
clones (patches). Total culm number is partitioned into
reproductive (shaded) and vegetative portions. 42

12. Histogram of whole clone (patches) dry weights. Dry
weight is partitioned into reproductive (shaded) and

vegetative portions. Numbers above bars represent RE. 43
13. Map of clone A 46
14. Map of clone 8 48
15. Map of clone C 50

vii

16. Map of clone D
17. Map of clone E

18. Diagramatic explanation of the terms "position 1",
"position 2", "proximal tiller” and "distal tiller”.

l9. Seedling diagrams, not to scale, of the underground
development of six laboratory seedlings.

20. Photographs of site, seedling and rhizomes.

viii

52

54

55

7O

73

INTRODUCTION

The grass Diarrhena americana Beauv. is a deciduous
forest taxon of uncertain affinities, it appears to occur
only in flood plain forests and has only been reported from
a few isolated locations in Michigan where it is at the
northern edge of its range. The species occurs more
frequently in flood plain forests of states to the south and
west. The Michigan population used in this study is
composed of scattered patches of various sizes, located on
the Red Cedar River floodplain in northern Ingham County,
Michigan.

Diarrhena americana is commonly described as a
rhizomatous grass. However, the extent and persistance of
its rhizome connections, as well as several other growth
characteristics are undocumented. Populations of Q;
americana occur as aggregates of variable size patches. The
purpose of this study was to investigate the growth habit of
the species, and specifically, to examine the relationship
between the growth form of 2; americana and the development
of populations composed of patches.

Information collected from an investigation of the
below ground structure of small patches will be used to
interpret above ground patterns and the morphological basis
for the formation of patches within a population.

This investigation of clonal organization asked two

1

major questions: 1. "Are the patches a single clone or
several?", and 2. ”What is the pattern of rhizome growth in

this grass?"

LITERATURE REVIEW

The many published studies that concentrate on
underground organization in plants vary in approach and
emphasis, but have in common that they investigate some
aspect of the clonal growth habit. Some studies have
focused on the non-random distribution pattern in vegetation
that has clonal dominants. These examined the relationship
between rhizome branching and patterns of above-ground shoot
or tiller distribution. K. A. Kershaw's studies on patterns

produced by Agrostis tenuis (Kershaw 1958), Carex bigelowii

 

and Calamagrostis neglecta (Kershaw 1962), Dactylis

 

glomerata, Lolium perenne and Trifolium repens (Kershaw
1959) and others (in Kershaw 1963) are some of the best
examples.

Other studies, also investigating pattern in
vegetation, place an emphasis on life history. Watt's work
is an example of the analysis and characterization of the
dynamic growth and change within a clonal community; Watt
(1947 a, b) working with bracken, Pteridium aquilinum,
demonstrated that a clonal population goes through various
stages of development as the clone spreads vegetatively.
The leading edge he called a pioneer phase followed by the
building and mature phases. The oldest, nonproductive

portion of the clone was the degenerate phase.

3

More recent similar studies have emphasized the
demography of ramets within the clonal population. In a
study of £3525 arenaria on sand dunes, Nobel _E _l, (1979)
followed cohorts of ramets on a clone displaying
developmental phases similar to those Watt (1947) found in
bracken populations. Dickerman and Wetzel (1985), with less

emphasis on the below-ground morphology, did a demographic

study of the individual shoots of Typha latifolia over two

 

growing seasons.

In a separate group of studies, investigators have
attempted to assign some adaptive significance to a
particular clonal architecture or the clonal life form in
general. Lovett Doust (1981) has introduced terms to
characterize the two extremes in clonal life form and their
adaptive significance. One extreme is the 'guerilla”
strategy, which is characteristic of those clones which have
widely spaced ramets that "infiltrate the surrounding
vegetation”. The other extreme being "clonal species which
present a tight-packed advancing front of ramets, a
'phalanx"which excludes other plants from the clonal
territoryu" (1981, gu754). This terminology suggests
tactical perception on the part of the clone and gives the
impression of strategic planning. Strategy on the part of a
clone is not possible, but genetic selection for optimal
patterns of ramet production and placement are. In their
study of Solidago canadensis, Smith and Palmer (1976) found
that the branching pattern was not significantly different

from an optimal branching pattern model. They suggest that

there is selection for precise and efficient geometric
patterns in clonal plants.

Not only is the pattern by which ramets are deployed
important but perhaps the longevity of the connections
between ramets may be important. Some research suggests
that maintaining costly ramet connections may be
advantagous. Barnett and Bazzaz (1985) found that if only a
portion of an interconnected clone is under stress due to
competiton of neighbors, the impact on the ramets under
stress will be less than if the entire clone was under
stress. From their work it appears that clonal connections
buffer the ramets from local variability in the environment.
This may or may not explain the findings of another study of
Clintonia borealis (Ashumn and Pitelka 1985) which concluded
that there were no adaptive genetic differences between
plants that were collected from sites with widely varying
light levels. They attributed this to a broad ecological
tolerance, but the source of this tolerance was not

mentioned: perhaps it could be due to Clintoniais persistant

 

rhizome network which was dealt with in a separate paper
(Pitelka 23 El, 1985).

Basic life history studies of rhizomatous plants have
focused on resource allocation (Armstrong, 1984: Ashmun st
31, 1985), on the pattern of shoot development (Bernard,
1975), and on the branching pattern itself and how the
pattern fits into the life history of a species (Bell, 1974,

1979). Several studies have addressed many different

6

aspects of a life history at once (Pitelka _£._l' 1985:
Phillips, 1953).

Several studies (Bell and Tomlinson, 1980: Bell _£Hgl,
1979) have taken the few morphologically detailed life
history studies available and have developed computer
simulations which, using life history data on a particular
rhizomatous growth pattern, can devise models and generate
patterns which can represent present and future growth of a
clone through the addition of ramet modules.

Studies relevant to the present work are those which
focus primarily on the growth form and ecology of a
particular clonal grass species. Caldwell's (1975) study of
Spartina townsendii colonies on mud flats, Perkin%3(l968)

work on the ecology of Nardus stricta in British grasslands,

and Maun's (1985) papulation biology study of Ammophila

 

breviligulata.and Calamovilfa longifolia on the sand dunes

 

 

of Lake Huron are a few examples that concentrated on the
nature and construction of the clones.

None of the grasses studied have a rhizome system that
resembles that of Diarrhena. However, none of the clonal
studies deal with a grass in a woodland environment. 2;
americana is unusual in that it is a temperate woodland
grass. Pohl, in "The Grasses of Iowa" (1966, p. 343)
provides a list of characteristic grass species for the

major habitat types in Iowa. 2; americana is included in

 

the listing under "lowland forest”. The lowland forest list
contains only seven species of grasses. The list for upland

forest contains 20. The remaining habitat types have the

following numbers of characteristic grass species listed:
lowland prairie (24), dry prairie (23), marsh, fen, bog, and
wet shores (22), meadows (25), weeds of cultivated fields
and waste ground (33). Pohl does not note whether the lists
are complete. However, even if only the most common species
are listed, only 18% of these species are characteristic of
woodland communities. Further, only 5% of the grass species
are characteristic of lowland forest, the habitat of 2;
americana. Clearly grasses are poorly represented in the
forest communities.

It has been shown that woodland species differ in
certain respects from those found in more open environments.
Woodland perennial species tend to allocate a significantly
greater portion of their resources to leaves and underground
organs: seeds tend to be larger and fewer, and the total
biomass allocated to reproduction is smaller in woodland
species as compared to field species (Abrahamson, 1979).
Such differences in life history and growth habit make the
comparison between woodland and non-woodland grass species
difficult.

The bamboos are the only group of woodland grasses for
which any information on underground morphology is
available. McClure (1960), reviewed all aspects of these
plants including the variety of rhizome branching patterns
and systems that he observed. Some of the bamboo rhizome
systems McClure illustrated strongly resemble that of

Diarrhena. Recent studies in grass systematics suggest that

8

the genus Diarrhena may be more closely related to the

 

bambusoid grasses than to other grasses. These studies will
be discussed below.

Although there is very little known about the life
history of Diarrhena americana, there has been work on its
anatomy and systematic placement. Tateoka (1957, 1960)
examined the characteristics of leaf structure and

chromosomes of the Asian species of Diarrhena, and Anderson

 

(1958) and Renvoize (1985) have investigated the same

features in Diarrhena americana. The most extensive

 

anatomical study of Diarrhena dealt with embryo structure.

 

Schwab (1971) investigated the embryo structure of Diarrhena

 

and demonstrated that the embryo was of a bambusoid type and
not festucoid, as it had been previously considered. This
change in the designation of the embryo type eventually lead
to the genus' assignment to a different subfamily.

Systematic studies of Diarrhena have traditionally put

 

the genus in the tribe Poeae (Festuceae), subfamily

Pooideae. There has been a trend towards removing Diarrhena

 

from the Pooideae, but there has been little agreement as to
where it should be placed. Tateoka (1957, 1960) working
with the Asian species of Diarrhena, first suggested that
because of its distinctive morphology, the genus should be
moved from the Poeae to a new tribe Diarrheneae. Anderson
(1958) reviewed the morphlolgical variability of the genus
and suggested that it should also be removed from the
subfamily Pooideae and (for the lack of a better place, in

his opinion) could be placed in the Eragrostoideae (called

the Chloridoideae by Mac Farlane and Watson, 1980). In
dealing with the subfamily Pooideae, Mac Farlane and Watson
(1980) discussed the genus Diarrhena and concluded that
because Diarrhena has many non-pooid characters, including a
bambusoid embryo type, it should be removed from the

subfamily Pooideae. They dismissed the idea of Diarrhena (a

 

C3 grass) being placed in the predominantly C4 Chloridoideae
(Eragrostoideae)). They also could not place the genus to
their satisfaction, but suggested a relationship close to
the oryzoid and bambusoid grasses. Hilu and Wright (1982),
in a cluster analysis of the Poaceae, suggested the creation
of a new subfamily, Nardoideae, which would contain
Diarrhena and two other problem genera, Lygeum and Nardus.
The new subfamily would be located near the clusters formed
by the oryizoid and pooid grasses. Watson gt _1’ (1985),

working at the subfamily and supertribe level, placed

Diarrhena in the subfamily Bambusoideae, supertribe

 

Oryzanae. This arrangement grouped Diarrhena with the

 

herbaceous bamboos and the oryzoid grasses and regarded the
tribe Diarrheneae as an outlier to the bambusoid group.

Renvoize (1985) placed Diarrhena in a group he called the

 

bamboo allies which was much the same arrangement as Watson
33 _l (1985). Campbell (1985) opted not to suggest any
relationships and has placed the Diarrheneae in a catch all
list of "unplaced tribes".

The placement of Diarrhena will probably always be a

 

problem and the distribution of this genus may suggest why.

10
Two phytogeographic papers (Li, 1952: Koyama and Kwano,
1964) label Diarrhena an east Asia-eastern North American
disjuct genus. This type of disjunct distribution,
according to Li (1952), indicates that the genus was
probably a component of the widespread forests of the
Tertiary, remnants of which are today apparently only found
in locations such as eastern Asia and eastern North America.
Anderson (1958, p. 18-19), after attempting to place
Diarrhena systematicly and using Li's hypothesis, had
concluded the following:

“If Diarrhena were a remnant of the flora that existed in
this area at one time, as its present day distribution
suggests, then it is indeed a very ancient genus and could
conceivably have remained relatively unchanged and possess
characters of groups which later became separated into major

trends that we recognize in the grass family today.”

If Anderson's consclusion is valid then Diarrhena will

 

never be placed adequately in any of the existing taxonomic
schemes and problems with the systematic placement of
Diarrhena will persist. As more information about this genus
is gathered, perhaps a clearer picture of the relationship

of Diarrhena to the remainder of the grass family will

 

develop.

METHODS

Establishment of permanent field transects and quadrats

 

 

Three Diarrhena patches were selected randomly from a
pool of 35 numbered patches. The 35 patches were of similar
size and shape and were located within the main population.
Patches at the extreme edges of the population were avoided,
as were extremely large or irregularly shaped and small
patches. The objective in the selection process was to
obtain representatives typical of the pOpulation.

Transects through the three Diarrhena patches were

 

established (4/11/85) to examine variations in tiller
production, culm weight and the proportion of reproductive
culms. These transects through large patches were called
"field transects” and the patches "field transect clones”.
The transects were positioned through the centers and
extended beyond the margins of each patch.

The transects were established by first measuring the
diameter and finding the approximate center of each patch,
randomly selecting the direction for the transect to run
through the center, and then laying a cord which was knotted
at 10 cm intervals along the selected transect direction.
The transect extended to the edge of the patch at both ends.

Contiguous 10 cm x 10 cm quadrats were positioned
along the transect, with the first quadrat being directly
over the center. The remaining quadrats were then positioned

‘11

12

such that they bisected the patch and extended outward until
the margin of the patch was reached. The corners of each
quadrat were marked with white plastic stakes. These
permanent quadrats were used to monitor differences across
the patch at 10 cm intervals. Every culm (defined here as
an erect leafy shoot) present within the transect quadrats
was marked with an aluminum tag. The tag consisted of a one
inch strip of aluminum with a quarter inch whole punched in
one end. The tags were secure on the culm after the first
leaf was completely unrolled and could not be removed
without severly damaging the culm. Wooden stakes were
positioned at the last quadrats on either end of the

transect to facilitate finding the ends of the transect.

Data collection 23 the field transects

Each quadrat along the three transects was monitored.
The intial shoots were tagged, and then mapped onto a 10 cm
x 10 cm representation of the quadrat on a data sheet (the
map to be used if, for any reason, the tags were removed or
disturbed). In the fall (9/14/85) the number of shoots
initally tagged was recorded for each quadrat and any
additions to the quadrat were recorded as well as any

deaths.

Harvest for dry weight measurements.
Before clipping all above ground material in the
transects, the following were first noted on the field maps:

1) any new culms emerged since the last tagging, 2) which

l3

culms were reproductive, 3) which culms were vegetative and
4) which culms had died. In addition, each quadrat was
checked for seedlings. After recording this information
(9/14/85), all above ground material in each quadrat was
harvested separately. All vegetative culms for each quadrat
were bagged together. Reproductive culms were separated into
a vegetative portion and a reproductive portion. The
division between the two portions was considered to be at
the point where the culm emerges from the upper most leaf
sheath. Reproductive effort (RE) was calculated on a per
quadrat basis using above ground biomass. The reason for
using the quadrat as the unit of data collection was that
there is the possibility that vegetative culms may
contribute to the reproductive effort of adjacent
reproductive culms through rhizome (lateral belowgrownd
shoots) connections. Therefore, the 10 x 10 cm quadrat was
used as the unit of data collection, rather than each culm.
The vegetative portions of all culms in a quadrat were
combined, as were all the reproductive portions, weighed,
and the RE was then calculated from these combined totals.
By using the quadrat to calculate RE the problem of
transport is reduced and the RE value calculated represents
the performance of the patch in that 10 x 10 cm area along
the transect. Changes in the RE value along the transect
would indicate zones of reproductively active vs. inactive
culms in a patch.

The inflorescences were individually bagged and

14

numbered according to transect and quadrat number. After
harvesting all culms, the quadrat markers were left in
place to monitor any future shoot emergence (in the form of
upwardly turning rhizomes, here called tillers). (After the
culms were clipped the metal tags were no longer secure, but
the stubs from the clipped culms (culms that had been
present during harvest) were obviously different from new
emerging shoots (tillers).

All material harvested was dried for 24 hrs at 100°C.
The number and average weight of seeds per inflorescence was
recorded. The dry weight of all reproductive and vegetative

material was recorded for each quadrat.

Whole Clones

To investigate the organization and growth of a patch
it was necessary to monitor all the culms of an entire patch
and dig it up at the end of the growing season to determine
the underground connections. It would be important to know
how variable patches were and what the effects of
disturbance due to transplanting were on the clonal growth
pattern. Five patches were selected for this investigation.
The criteria for selection were different from those for the
transects. The patches had to be roughly of equivalent
size, but smaller than those used for the transects. The
patches used here had to be approximately one half meter in
diameter, or less, in order to: 1) make marking and
monitoring all culms and culm additions feasable 2) to be

able to transplant some clones into bushel size tubs, and 3)

15

to be of a workable size to make entire patch maps in a
reasonable amount of time.

One half meter patches were not as common as the larger
patches used for the transects and therefore more difficult
to find. Five small patches were found. One was marked
with a flag, all of its culms were tagged and the patch was
left in place undisturbed. A second patch was completely
excavated, placed in a large metal tray, and replanted, tray
and all, in the exact position from which it had been
removed. All of the culms in this patch were also marked
with metal tags. The remaining three patches were
completely excavated, placed in bushel tubs and transported
to the M.SJL greenhouses. All culms were tagged with metal
tags.

The greenhouse replicates were grown under uniform
conditions in order to eliminate possible growth pattern
variations due to environmental variations on the flood
plain, ecg. light, moisture and nutrients. The patches that
remained on the flood plain, one undisturbed and one
transplanted, were used to detect any unusual patterns that
may have occured due to the shock of transplanting.

The field patches were checked throughout the summer for
new shoots. The green house patches were removed from the
greehouse on 2 May 1985 and placed on a snowfence shaded
outdoor bench for the remainder of the summer. This move,
not origionally planned, was needed due to uncontrollable

hot, dry conditions in the greenhouse. Potted patches were

16

watered when dry and fed with ”Schultz-Instant" liquid plant
food once a month. The smaller patches used for the whole
clone studies will, for discussion, be termed "whole clones"
and transects through these patches are called "whole clone

transects".

Qata collection aa'gtata clones.

Three groups of data were taken (9/3/85) from the whole
clones: transect dry weights, whole clone dry weights, and
maps of the rhizome network with culm locations and
connectionsm To make culm identification possible for
mapping after harvest, all culms were first tagged at the
base with color coded tape indicating that they were
reproductive or vegetative. A small scale (in comparison to
the field transects) transect was then placed through each
whole clone. Culm density, dry weight, and the frequency of
reproductive culms were recorded for 2 cm x 2 cm quadrats
along this transect. The transect passed through the center
of the patch and was used to detect differences in above
ground culm pattern across the patch. The culms within the
transect were first numbered and color coded with tape and
then harvested, leaving the color coded culm bases attached
to the rhizome system. The culms were then dried at 100° C
for 24 hrs and weighed. Before weighing, the reproductive
and vegetative portions were separated in the same manner as
for the harvested material in the field transects.

After the harvesting of all material within the

transect, all the remaining culms were harvested to obtain

17

the total above ground biomass for each of the whole clones.
Reproductive and vegetative material was again weighed
separately; The total biomass allocated to reproductive and
vegetative portions of the clone was calculated by adding
this total to that of the transects.

The remaining portion of the clone including the above
ground color coded (reproductive or vegetative) stubs of the
harvested culms, was removed from the tub in which it had
been growing, placed on a metal screen and allowed to soak
in a basin until completely saturated. The saturated mass
of soil, roots and rhizomes was then gently washed until
the root and rhizome mat was completely exposed. It was

impossible to distinguish Diarrhena roots from roots of

 

other plants which were completely entangled in the root and
rhizome mat. All roots were therefore completely removed and
were not used in any biomass calculations. It also should
be mentioned that the rhizome network was obscured by the
dense root mat and that it was necessary to remove the roots
to facilitate accurate observations of the rhizome
connections. After washing and root removal, that which
remained was the rhizome network with the attached color
coded culm bases. A11 aluminum tags which had been placed
on the clone in the beginning were held in place by the

colored tape.

Mags.

After cleaning the rhizome network, it was allowed to

18

air dry. The rhizome systems were then placed on a paper
covered light table and all rhizome connections and culm
locations were traced and marked onto the paper. The
resulting map was a two dimentional representation of the
rhizome system and the attached color coded culms. 'This map
was then traced and translated into the black and white
format included here. Erect leafy shoots were called

culms. Lateral belowground shoots were called rhizomes.
Upward turning rhizomes which may produce new culms were
called tillers. The term culm base was used to refer to the
junction between culms and rhizomes, as they have been
defined here. Culm bases remain after the senescence of the

culm and give rise to new tillers.

Fragmented Patches

To investigate the possible reaction of rhizomes when
they are fragmented, three patches were excavated, taken to
the greenhouse and cut into pieces. They were each cut into
four triangular "pie” pieces, each piece with a center and
edge portion. The culms were tagged and these clone
fragments were planted into metal trays and treated the same
as the whole clones A-C for the entire summer. The
below ground portion was rinsed, the roots removed, and
notes were taken on the patterns observed. Schematic
diagrams were made for each fragment rhizome system. No
traced maps were made. The purpose of this treatment was to
determine if there was any regeneration in older portions of

the clone or additional vegetative growth triggered by

l9

fragmentation.

Seeds and Seedlings

 

Seed from the fall of 1984, some which were
refrigerated from October 1984 to March 1985 and some which
had been stored at room temperature, was stratified in sand
from March 15 to May 1. The only seeds to germinate were
those which had been refrigerated. These seedlings were then
placed in a flat inside an aquarium in a north facing window
and watered regularly. These seedlings were observed and
schematic diagrams of the underground development were made

for several of them.

FIELD SITE DESCRIPTION

 

Location

The geographic coordinates of the study site are 84'
26" W, 42' 44" N. The site is in south central Michigan,
Ingham Co., five miles east of East Lansing, Michigan (SWl/4

Sec 25, SEl/4 Sec 26 T4N R1W). The Diarrhena population

 

used for this study is at the intersection of the Red Cedar
River and Van Atta Road. The population is north of the

river and west of the road.

Soil

 

The soil at the field site, according to the soil
survey of Ingham County, Michigan (1979), is Ceresco fine
sandy loam, typically, "umsomewhat poorly drained, nearly
level soil .u on flood plains along streams and rivers. It
is subject to frequent flooding."

The field site was flooded twice, 15 March and 4 April
1985, during the time period covered by this study. Both
incidents of flooding lasted several days and completely
covered the Diarrhena population used in this study. The
site was also flooded in March 1986 (Stephenson, personal
comm.), the spring after the study was completed, suggesting
that the flooding, as mentioned in the soil survey, is

frequent.

20

21

Vegetation

 

The vegetation at the field site can be compared to
that which Curtis (1959, pp.156-168) described as southern
lowland forest in Wisconsin: a forest dominated by silver
maple, found along rivers and frequently flooded. Kron and
Walter (1986 a,b) studied a community that is similar, but
not identical, to the one on Van Atta Road. Their study
site, also along the Red Cedar River in Ingham County,
Michigan, included many habitat types which are not
represented at the site used in this study. Those species
listed by Kron and Walters (1986a) which are also common at
the Van Atta Rd. site are listed below. This list is not
complete and serves only to characterize the vegetation at
the site. The taxa are listed in the order of importance as

determined by Kron and Walters.

CHARACTERISTIC SPECIES 95 FIELD SITE

Woody

Acer nigrum Michx. f.
Acer saccharum Marsh.
Populus deltoides Marsh.
Platanus occidentalis L.
Fraxinus americana L.
Ulmus americana L.

 

 

 

Spring Herbaceous

Dentaria laciniata Muhl.
Geranium maculatum L.
Erythronium americanum Ker.
Asarum canadensis L.

Viola canadensis L.

Allium tricoccum Ait.

 

22

Dicentra cucullaria (L.) Desf.
Sanguinaria canadensis L.

Dicentra canadensis (Goldie) Walp.
Claytonia virginica L.

Enonymus obovatus Nutt.

Geum canadense Jacq.

Cardamine douglassii (Torr.) Britt.

 

 

Fall Herbaceous

Asarum canadense L.

Viola canadensis L.

Hydrophyllum spp.

Enonymus obovatus Nutt.

Geranium maculatum L.

Viola sororia Willd.

Boehmeria cylindrica (L.) SW.

Geum canadense Jacq.

Laportea canadense Jacq.

Circaea quadrisulcata (Maxim.) Franch and Sav.

 

 

 

 

 

 

 

9.922

The canopy of the forest closes in late May or early
June: leaf drop is usually complete by late October. Light
levels are low with the exception of sun flecks and small
canopy openings. An estimate of the light level at plant
height on the forest floor is represented by figure 1. The
light measurements (provided by Dr. S. Stephenson) suggest
that a very small percentage of full daylight is reaching

the forest floor.

Climate

Figures 2 and 3 summarize temperature and
precipitation (respectively) for the East Lansing area. The
data includes long term averages (for the period 1939-1969),
departures from normal and summaries for 1985. From these

figures it should be noted that the mean maximum temperature

PCT. OF READINGS

23

LIGHT INTENSITY FREQUENCY HISTOGRAM

50

 

30-
20s

IO-q xL—ii_£*_a—4}-___—_£r —_£r_— .if/n

o r ‘ l

5 1O 15 20
PERCENT OF OPEN LIGHT INTENSITY

 

 

 

 

25

Figure 1 Light levels at plant height at the Van atta Road field
site.

DEG. C.

Figure 2.

24

1 985 SUMMARY

 

 

 

 

40
30d
20‘ .I * IT :1
. j i I
101 :I :i
.i ........... I ................................ 1 ........
"I I I i
-20,
I
-301 f I"— r T 1— 1 1 r I t fir r
an: Ft! IN“! NH! MAY JUN .nn.‘nu3 an! 06! NOV DEC
Moan monthly temp (x). + moan max. - moan mln
Summary of temperatures for the East Lansing area.

25

 

 

 

 

 

SUMMARY
Long-form normal
30
29.
d
3; IO-'
O
“W; ................................ I “1
"° finrfiuhfi—M'Avahiaafiosfioéruévfic
/
DEPARTURES
1985 SUMMARY
I0 '4-fuflfl'
-4»-:uuur

 

 

 

 

Figure 2 continued

DEPARTURES
1985 SUMMARY ,

26

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m ---..I..- , _
m. .. .. .. s. ........... ... $.T
s T I I- .m
g ,m §§ .w.
NN T N e
.m Y m
N 1m m m
T T
N T m
m .m i
\NNN a
NNN. a NNN§
N m. m l I .w
TIER; ........+.--_...--IT m m a .

$3 5..

Figure 3.Precipitation for the East Lansing area.

27

in April was considerably higher than normal. However, the
mean maximum and minimum temperatures were well below normal
in June, July and August. The early summer (May-June) was
drier than average while the later portion of the growing
season (August—November) was wetter. It is not known how
much these deviations from normal precipitation and
temperature have influenced the growth of Diarrhena

americana in 1985. However, it might be expected that the

 

flood plain habitat would moderate these deviations.

RESULTS AND DISCUSSION

Initial observations.

The Diarrhena americana population at the Van Atta Rd.
site occurs as dense patches of variable size. These occur
on the floodplain as single isolated units, or more commonly
in localized groups. Culm densities vary within the
patches, providing a conspicuous secondary scale of non-
uniform distribution (Fig. 4).

During the 1985 growing season the Vanatta Road

Diarrhena population produced one cohort of culms. All of

 

these culms appeared to have developed from tillers that had
emerged the previous fall. Most tillers produced during the
late fall of 1985 were observed to produce one or two
leaves: others remained in the bud stage.

In the course of the study it was observed that in the

lab the fruits of kamericana will float in water for up to

 

one week. The potential for 2; americana to disperse via

 

water along flood plain systems must be great. No fruits
were observed floating during the two floods that occurred

during this study.

28

29

14‘
FIELD TRANSECTS

13% D
0-'33
12~ 0- 27
‘A- 35

11-

‘10--

ad

7-:

number of culms

 

 

”-
b-

Figure 4. Field transects compared. Total number of culms compared for
field transects 27, 33 and 35.

30

Field Transects.

 

Results
Patch Structure and Organization

Variations in culm densities along the field transects
is shown in Fig. 5. An uneven distribution of culms is
demonstrated in all three patches: in each case there is an
‘empty central zone surrounded by a high density zone. A
ChiZ goodness of fit test (Appendix A) indicated significant
differences from an expected uniform distribution within the
patches (P = <.05).

The number of tillers emerging in the fall after each
quadrat was harvested is shown in Fig. 6. The distribution
pattern of the fall emerging tillers was similar to that of
the current seasons culms, but due to an overall lower
density the distribution did not differ significantly from a
uniform distribution (Appendix B).

The number and distribution of reproductive culms along
two of the field transects was similar to that of total
culms (Fig. 5). The small sample size precluded tests of
significance for distribution within patches, however, there
was a significant association between flowering culm
occurrence and greater than average culm density (Chi2 =
4.86, P = .027).

No flowering culms were produced along field transect

33. \

Standing Crop and Reproductive Effort

Figure 7 shows the above-ground standing crOps for the

12‘

104

6.1

4.

number of culms

2-I

 

31

 

"'15: # 3s

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

14-

12'I

10.1

6-4

.4.

number of culms

 

.b

qp
'
w-

quadrat

# 27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10-

8-1

number of culms

 

# 33

 

 

 

 

 

 

 

__T__k

 

 

 

 

 

 

 

 

 

 

 

 

 

I.

 

Figure

5.

2 a 4 s 5 7 a s 10 11 12 1a 14 1s
quadrat

Histogram of culm number for field transects 35, 27 and 33.
Shaded area represents number of reproductive culms.

32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8-
# 35
6-
H “' I—__
m
1%
a 2-
a
1 ' 2 3 4 s a 7 a 9 1o 11 12 1a T144r1s'
quadrat
d.-
6- # 27
‘u)
H
o
‘2
s 2'
‘ 1
1 2 8 4 5 6 7 8 9 1O 11 12 13 14 15
quadrat
8 I
T 1 33
6-1
In
<11 ‘1
a l—‘
E
a 2d—‘ J—
1 2 3 ‘ 5 5 7 8 9 1o 11 12 13 14 15
quadrat

Figure 6. Histogram of fall emerging tillers on field transects 35, 27
and 33. Each bar represents the total number of fall tillers
for each quadrat along the transect.

 

 

 

 

 

 

 

33

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

51
F .11
I 0"... # 35
3.....4.
A {of-'25.:
m Lbs—A
e 4»
(U
H
60
V
1.:
S
2?
0 3w
3
>4
«‘5
.2 1 «03
g 24-
H
on
2 .06
0 TWO—*
.0 hl-I—I-hm
Q 1 ‘-
1—1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
quadrat
# 27
3‘.
.04 .02
m .04 1 .
5 z“ . 71111 .:2223
La
00
.03
.14
.09 '—"
11. m’ 1L.-'_1a_.
1_1
1 2 3 4 5 5 7 8 9 10 11 12 13 14 15
quadrat
2T
m # 33
a ‘1!»
a .
60
—I
v—I
1_‘ 1‘1
1 2 3 4 5 B 7 8 9 10 1 1 12 13 14 15

Figure 7. Field transects 35, 27 and 33.
standing crOp for each quadrat and is divided into reproduc-
tive (shaded) and vegetative portions.

Each bar represents total

34

three field transects and reproductive effort for field
transects 27 and 35. There is a significant positive
correlation between the RE for each quadrat and the average
culm weight within each quadrat along the transects 27 and
35. The average culm weight calculated over all quadrats
for transects 27, 33 and 35 showed that the non—reproductive
transect had the lowest average culm weight (Appendix C ).

Figure 8 presents the total number of culms harvested,
the number of reproductive culms, and the number of fall
emerging tillers for each transect. The shaded areas on
the figure represent zones where the number of fall emerging
tillers is greater than the number of culms of that season.
In figure 8, peaks in reproduction are accompanied by peaks

in total culms and fall tillers.

Discussion
The significantly uneven distribution of culms along

all field transects supports the idea that there is a

pattern in the distribution of culms within a Diarrhena

 

patch. A bimodal pattern of distribution is shown in Figure
4.

The nonsignificance of the fall emerging tillers'
departure from a uniform distribution does not necessarily
mean that the bimodal non uniform pattern will not occur the
following year. Later sections will show that the Diarrhena
clones studied maintain, at least, the same number of culms

from year to year, and that for most whole clones (A —E) the

number of culms

13‘

121

10"

8-1

51

5-1

4..

3‘

   

 

35

0- total I
A’ fall emerging tillers I
0- reproductive culms '

FIELD mnsscr # 27 %1

'"TD

$ ,-
\\D\
, --

‘

’1‘--
\~
\
S
v ‘

RD
\ \.
/
/
_”,a’
‘
“a
I
a":
9\\
Wk. 0\\
\“
~
’ v ’ '
>\
\
hon—”‘v—

/
5“
.‘zzfijbo.
\.
\
-OO'T"-
k H

I". s
.‘
I
I
I ,.
- §
.' -
o
‘

~ ~
., l
\
J.. ‘
A
5
I "4’
.00
u . ‘
.a. I“... . .
- ' .31.; ., _ _.
' .-
’ ’ ’

F.
1.
:1
1
o
0‘ I
0"JD
33.3325} #14
oo ..‘.:;o_ I
./
J "

 

 

quadrat

Figure 8. Summary figures for Transects 27, 35 and 33 comparing

total number of tillers and number of reproductive culms
along each transect.

 

36

"‘ ,3 FIELD mnssc'r # 35

1o- \ . 0- total
. A- fall emerging tillers
1 .- reproductive culms

'\ II

1’
9- I \
I

7-1

number of culms
(n
l

 

 

FIELD TRANSEC'I # 33

" o- total N
A' fall emerging tillers 1‘

8- 0- reproductive culms /I \

7- o

I \

 

 

 

 

 

Q
.5 .
a 5‘ ' ' A
“" I ‘\ /.:‘
o .1 ,
1. ‘ A ' . [R
.8 A." I / ‘ [:1’
E ‘1 {21.1% ‘ \ ..i
s 3 r .. .. \
E [5:13;] \\ /' ‘J‘t \ \
’ o 3:.
1 y 113
2~ 4 1 1 , 1 a
1'] \1 ’l ‘1 “
1 .1 C1 1 ’p—ua’ \ \
\ I A..—— 'An— cut-b
o g—4——-?—#o_fi E H 044:; —o
1 '2 3 4I 5 o 7 a 1'1 1T 0 1'1 1'2 1'3 1'4 15

Figure 8 continued.

37

number of tillers produced for the next season is greater
than the number of culms present in the current season. The
number of fall emerging tillers for each of the field
transects is less than the number of culms for the current
season, and it is possible that all tillers which may
contribute to next years culm distribution pattern may not
have emerged.

The significant positive association between the number
of flowering culms and a greater culm density indicates that
it is only the most active area of the clone which
reproduces. Later sections show that, after studying the
rhizome system of whole clones, there is only one area of
the clone which is active and therefore all culms, flowering
or not, occur in the same area of the clone. The appearance
of zones with different levels of activity may be due to
some of the new rhizomes from the active portion of the
clone circling back into inactive areas.

The significant positive correlation of RE values with
average culm weight (culm weight being the vegetative
portion only) also indicates that the more vigorous areas of
the clone (areas greater average culm weights) are usually
the areas of reproduction, and again this may be explained
by the single active zone of the rhizome system.

The positive correlation of high RE values with high
average culm weight occurs again in a comparison of the
average culm weight over all quadrats for each of the field
transects 27, 33 and 35. In this case the average culm

weight may an indicator of clone vigor; The transect with

38
the lowest average culm weight had no flowering culms.

Transect 33, with the lowest average culm weight and no
reproduction, was the only clone used in this study that was
completely overtopped by nettles for most of the growing
season. If mean culm weight is an indicator of clone
vigor, there may be a connection between plant vigor, the
lack of reproduction and the reduced light level that would
be produced by a canopy of nettles.

In general, the reproductive effort displayed by
Diarrhena is not unusual for a forest herb. The
reproductive effort values of the field transects fall
within the range of values for a large number of shade-

tolerant temperate forest herbs (Bierzychudek, 1982).

Clone Structure and Organization

 

Results
Above Ground Patterns

Whole clone transects.

Figure 9 summarizes the total number of culms along
transects for the five whole clones that were studied.
Although a Chi2 goodness of fit test failed to find any
significant difference from a uniform distribution of culms
for all whole clones except D (Appendix D), all five
transects show an uneven distribution of culms, and clones C

and E show a bimodal pattern similar to that observed in the

number of culms

 

39

PATCHES A,B,C

 

61»

.number of culms
?

 

PATCHES D & E

 

 

Figure

quadrat

9. The total number of culms in quadrats along the whole clone

(patch) transects.

 

A=
13=C

.2
03E

40

field transects. Figure 10 presents values for standing
crop and reproductive effort along transects through clones
A-E. There is a positive correlation for clones A, B, C and
D between vegetative biomass (mean culm th and RE

(Appendix E).

Whole clone transplant study.

The number of vegetative and reproductive culms in each
clone A-E is illustrated in Figure 11. Clones A and C had
the greatest proportion of reproductive culms, followed in
order by clones B, E and D with the lowest number of culms
flowering.

Figure 12 presents total standing crop and reproductive
effort for clones A-E. The reproductive effort of clones D
and E was much lower than in clones A, B and C: the clones
appear to differ according to their growing environment.
Those clones grown at Michigan State University, A, B and C,
had a higher RE than those, D and E, which remained on the
floodplain.

When whole clone data is used (vs. using whole clone
transect data) to test for a correlation between RE values
and the mean culm weight, there is no significant
correlation (Appendix F). There is, however, a very strong
positive correlation between mean culm weight and the
percentage of the mean culm weight contributed by the
vegetative portion of reproductive culms. Those clones with
a greater proportion of reproductive culms had a higher

clonal average culm weight.

41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e.
PATCH A PATCH D
50!
4.
03 m
E
a: S 3.
:10 1..
34 m
2‘ 00.0. 2‘
ans: , .08
1- . Tu-~- .25
'2? ~04 ' It:
I=r__J—‘—-1
‘ 2 3 ‘ '5 5 1 2 3 4 '5 o
quadrat quadrat
5.
PATCH 8 PATCH E
5-:
4.
U!
E
(B
a 3‘0 3“
m
3‘.
2d
.18
1 .04
"".’.1-- -
1. :04 E: F1222:
1 2 3 4 5 6 1 2 3 4 5 5
6 quadrat quadrat
PATCH C
2
n
H
on

 

quadrat

Figure UJHistograms of total above ground dry weight from whole clone
(patch) transect quadrats. Dry weights are partitioned into
reproductive (shaded) and vegetative portions. RE for each
quadrat is above each bar.

number of culms

area C111

90“

80--

70...

501* . :13: 2.“: '.'-':'-’ 1:: l
. o. ' : 2: ‘ '11... . .
o ‘. . :1 :...:....oo

so»
46...
inc-I
204.

10'1-

 

 

 

 

 

 

 

42

 

 

 

 

 

 

1500-

I

1200‘

600-1-

3001'

 

 

 

 

 

B

C

whole patches

 

 

 

 

 

 

B

 

 

 

 

C D E
whole patches
Figure 1J.Comparison of area and total number of culms for whole clones

(patches). Total culm number is partitioned into reproductive
(shaded) and vegetative portions.

43

600

 

 

Z... .:..:
55 ‘iii '29
so" 1 1.33,,- m
:o::.o.o.o ::.:'.'. 0‘0: :0 .2 . .
: :1 ..4 it: '1': .' ..'. a". .. 'I
Tiiff: 4' - 1 5': r
. . . o .o : o z... .4
"1'5: .’.'-'-o 's- .'
40"- :‘:.o.. o.‘ * P:::o.. .'
..o..o..:. Io... .....
. .0. .0... ‘ 0.. : .
I. ..o. ' o Id

301’

20--

above ground dry weight (grams)

 

10" E

 

 

 

 

 

 

 

 

 

 

 

 

A ' B C D E

whole patches

Figure 12 Histogram of whole clone (whole patches) dry weights. Dry
weight is partitioned into reproductive (shaded) and vege—
tative portions. Numbers above bars represent RE.

44

Below Ground Patterns

 

WhOle clone rhizome study.

After the rhizome connections for whole clones A-E were
mapped (Figures 13-17), it was apparent that all five whole
clones are each a single rhizome system. In several of the
clones there was fragmentation in the older sections, and
this made it impossible to eliminate the possibility that
the clones may be two individuals of equal age. This is
unlikely, however, because there appears to be only one

center in each clone.

The growth pattern of culms and tillers.

The general growth pattern of the clone is easily
discovered by following along each rhizome arm on the whole
'clone maps. There are only two points on each rhizome
branch occupied by living culms. These two locations on the
rhizome branch will be referred to here as position 1 and
position 2 (Figure 18). Culms at position 1 are defined as
those which either arise from distal tillering and promote
clonal expansion, or are the last culms at the end of a
rhizome branch. Position 2 culms are those emerging at the
base of a dead culm, most often the previous year's position
1 or distal tiller, and are not the distal culms on a

rhizome branch.

Figure 13.

Map of clone A.

45

 

46

 

 

 

 

 

Figure 14.

Map of clone B.

47

LR?“ No \

\/ .///" ' T9941 1
‘1' I 5‘ 9/
”K \‘fi/ //:\“'\/fi >C%\J*

/ . k“ ‘3 , _1 ‘// I
o\‘"\ .; o %§\1r /\I 1
“it” .: roux "- >1 *
.7 I! \I . I\
411, ' “k 1 - ’ I 1: KW)
1‘7”" 4 ' x l H 4 / ‘

4.223 1 ' 4 ‘ ‘ \n
\Lj‘ NI .>\> 1 I . l‘ I EVXV

.51. /% \ F’Kf’T—i
.‘j { fi/‘Oi’ 'Aflm f. \/ 7 1
1 I N ” , es My

 

 

Legend

0 tiller
0 reproductive culm

a vegetative culm
I) I dead culm base

*ccnter of clone
rhizome connections

 

 

 

 

 

 

 

centimeter scale

 

O
.-
N
u
3‘
u-

Figure 15.

Map of clone C.

49

 

 

50

f———‘—

n “1:“,

O 7"“I‘F’J'JL‘C‘L'JIE ("l-LI“
a vvt'cmtiw: culm
I Culld :Lxlm imsc

*m‘ntL-r N‘

 

clone
i ———r!:izvn~c ”June-tins”-
.____—

\y
rontimetnr scale ]

 

 

 

Figure 16.

Map of clone D.

51

Figure 17.

Map of clone E.

53

 

 

Logan!

0 tiller

0 reproductive cul-
o vegetattvc cul-
I dud cu].- bun

*center of clone
—-— rhizou conncctlons

g

 

 

 

centtmtor scale

 

 

54

55

Legend

0 reproductive culm

l
o tiller }
|
A vegetative culm t

 

 

 

  
 

I dead culm base
i a
*center of clone
-———-rhizome connections |
- position
1 <3
\fi’.
position
I. 2
' IA
distal
tiller
. proximal
tiller

  

 

 

 

 

 

 

I 0 o
I
I
l
l
l
n
(
I J z
/ I
' l o
I
l
I . I ' proximal
l u “ 0 } tiller ‘.
5", K I ‘f ..
. )g . ‘ I,
I ‘ I I
\“ \ I J 7 I o /
(u‘ l ‘.“\ ’4.§-:.’;% //
I ‘ - ‘\ [’1’ ‘ ‘ ’
::’ A l ‘ :\\‘_ ."( ‘:\-Jd~'
V distal
. tiller
p051tion position
2 1

Figure 18 Diagramatic explanation of the terms "position 1", "position
2", "proximal tiller" and "distal tiller".

56
Culm Patterns

Appendix G lists the frequency of vegetative and
reproductive culms at position 1 and position 2 on each of
the five maps. Few culms located at position 1 are
associated with a dead culm base from the previous year,
though there are cases where this occurs. This situation
would be created when a culm from the previous year produced
only a proximal tiller. All culms at position 2 are
accompanied by a dead culm base of the previous year. There
is usually only one culm at any point or position. Early
tiller emergence occurred on clones A - C due to more
favorable growing conditions. Clone 8 had the greatest
number of precocious or early emerging tillers. Many of
these tillers had developed a leaf and were included in dry
weight measurements as vegetative dry weight. These same
precocious tillers appear as vegetative culms on maps but
were not consisdered vegetative culms for the purpose of
describing the culm pattern because they were not developed
and would not develop into culms until the next season. The
early tillers which contributed vegetative dry weight give
the appearance of two culms at one position on the rhizome
system maps. The two are not of the same cohort and the
precocious tiller was ggly considered a culm for dry
weight measurements.

The similarities and differences in the occurrence of
reproductive and vegetative culms in the five mapped clones

were noted and are summarized in outline form below.

57

I. Frequency of culm patterns
A. For clones B, C and D the most common order of
occurrence for culms at position 1 and 2 was to have:
- a vegetative 25 reproductive culm at position 1
- no culm at position 2

B. For clones A and B it was as common to have both
positions occupied as it was to have position 2
unoccupied.

percent of rhizome arms with
both pos. only pos. 1

clone occupied occupied
A 58 43
B 24 78
C 36 64
D 27 73
E 53 47

II. When both positions 1 and 2 were occupied:
A. Clones A, B and C frequently had reproductive culms at
position 2.

B. Clones D and E never had reproductive culms at position
2.

Tiller'Patterns

Appendix H lists the the observed relationships of
tillers to vegetative and reproductive culm bases at
position 1 and 2. The trends found in tiller patterns for
clones A-E are summarized below:

I. Tillers per culm
A. The average number of tillers per culm base, excluding
those culm bases without tillers.

 

mean number

clone tillers/culm base
A 2.2
B 2.0
C 2.2
D 2.2
E 2.6

58

B. The average number of tillers per.culm base including
those bases that produced no tillers. All clones
except B have the potential to increase in size if all
tillers develop into culms in the next growing season.

 

mean number

clone tillers/culm base
A 1.7
B 1.0
C 1.8
D 1.9
E 2.3

C. The maximum number of tillers per culm base is higher
for reproductive culms, particularly those reproductive
culms at position 1.

D. The maximum number of tillers per culm base is four.

II. Frequency of tiller patterns
A. For clones A and E the most frequent patterns had a
culm base producing both proximal and distal culms.

B. For clones B, C and D there is a nearly even split
between tiller patterns with both proximal and distal
tillers and those with either proximal or distal
tillers.

percent culm bases with:

clone either both
A 29.7 70.9
B 51.2 48.8
C 49.1 50.9
D 54.9 45.1
B 27.8 72.2

C. Of those patterns in which a culm base produces both
proximal and distal tillers the two most common are:
1. one proximal and one distal (1/1)
2. one proximal and two distal (1/2)

D. There is no pattern with only distal or proximal
tillers that is consistently more common than any

other, although the pattern of two distal tillers (0/2)
was very common for clones C and D.

Table 1 lists the number of culm bases which have proximal
tillers, distal tillers, or both. Table 2 lists the

percentage of culm bases in each clone which produced 1, 2,

59

Table 1. Frequency of tiller types per culm base.

 

Types of tillers
Map Proximal only Distal only Both
6 8

A 33
B 10 12 21
C 4 23 28
D 17 17 23
E 0 5 l3

 

60

Table 2. Percentage of culm bases with 1, 2, 3, or 4
tillers.

 

Tillers per culm
Map l 2 3 4

A 20 4O 40 O
B 28 46 26 O
C 15 51 33 1
D 14 57 27 2
E 22 11 56 11

 

61

3 and 4 tillers.

The age of whole clones.

The age of whole clones (A-E) can be estimated if a
preliminary assumption is made that the clone never produces
more than one cohort of culms per year. Ageing the clone is
calculated by finding the terminal end of any rhizome arm
and counting backward into the dead center of the clone.
Each point or "node“ along the rhizome that produced a culm
would represent the termination of one year‘s growth (Figure
18). Finding the center of a clone can be made difficult by
fragmentation and deterioration in the oldest portions of
the clone. The center can still be found by counting several
rhizome arms at different points on the clone to narrow down
the area in which the center could occur. Consulting the
actual rhizome system, not the map, at this point is helpful
in determining the most likely point of origin. While the
maps provide an accurate representation of the rhizome
system as viewed in two dimensions, it is the third
dimension, out of the plane of the paper, that gives the
best clue as to the seedling origin of the clone. The best
clue is the growth angle of the rhizome and culm bases.

Culm bases slant outward from their point of connection with
the rhizome network. A region that has all culm bases
slanting away from it represents the center of the clone.
Centers are marked on the clone maps. On some maps the
centers are in areas which have fragmented.

Using this method to age the clones, the following

62

results were obtained: A, seven years: B, seven to eight

years: C, six years: D, ten years and E, five years.

Whole clone fragments.

Data collected from the fragmented whole clone study
are listed in table 3. The fragmented whole clones had no
regrowth in the older portions of the clone. Tiller
production occurred only from the two types of tillering
culm bases,positions l and 2, as was found for the whole
clones. The whole clone fragments produced reproductive
culms and had a greater number of tillers per culm compared
to the whole clones. The older portions of the whole clone
fragments became extremely brittle, unlike the persistent
connections of an unfragmented system. The clone fragments,
for the most part, degenerated into small disconnected

segments.

Discussion

It was found through the field transects that there is
an empty zone (or zones) in a Diarrhena patch. Each zone is
sometimes surrounded by an actively expanding zone. The
whole clone transect data do not show a pattern identical to
that found with the field transects. The field transects
were placed through considerably larger patches than the

half meter patches that were used for the whole clone

63

Table 3.

x Tillers

Culm and tiller summary for rhizome fragments.

 

per culm

Tillers per culm

Veg
Culms

Repro
Culms

Fragments

3850

2222

(2,213,113,3131211

,3,3,2,3,3

8222

5272

1234.
FFFF

4.055

2312

2500

54.66

1234.
GGGG

2230

84.26

1234
HHHH

 

64
measurements. The use of smaller patches for the whole

clone transects may be the reason for the difference between
the whole clone transect and the field transect data.

The data from the field transects show a pattern of
growth that could be produced by either a large clone, or by
several to many small clones located together with an
irregular density pattern. The inactive zones could be
dead old growth centers or they could unoccupied space. If
the larger field transect patches consist primarily of a
single individual, as the small whole clone patches do,
then it is reasonable to expect some similarities in the
data collected for both patch sizes. But if the patches are
clones expanding outward in a ring like fashion, a small
clone would have a less strongly developed division between
active and inactive areas (as in Spartina in Caldwell,
1957), while a larger clone which would have spread further
from its inactive center would show a clearer ring pattern.
By referring to the whole clone rhizome maps it is not
difficult to see the growth pattern that could be behind the
bimodal pattern observed in the field transects and to
explain the lack of pattern development in the smaller whole
patches. The maps clearly show that the whole clones have a
large, though not always centrally located, region that
produces no culms and a peripheral region that produces a
number of culms, followed by outwardly expanding rhizome
segments which turn upward to produce new culms. The ring-
like pattern was detected by the whole clone transects but

was not developed to the point to where it would test

65

signifcantly different from a uniform distribution. The
poor pattern development is due to rhizomes circling back
into the older inactive areas of the clone, thus blurring
the ring shaped pattern of growth. In larger clones there
would be a greater empty region and circling back would be
less likely to fill in this space. This "circling back" of
rhizomes is characteristic of rhizome networks which branch
at predictable or regular angles (Bell and Tomlinson 1980).

The distribution of culms along a transect through
whole clone D tested significantly different from a uniform
distribution. It would be the most likely whole clone to
show the development of pattern because it is the largest
and the oldest of the small patches used for the whole clone
studies. In an older clone the ring pattern would be
strongly deve10ped and more likely to be significantly
different from a uniform distribution. Whole clone D would
have fewer rhizomes filling the inactive areas the older
portions of the clone.

The most common Q13 pattern for clones B, C and D is
to have a culm Bill in position one. For the same three
clones the common tiller pattern is to have half of the culm
bases with either proximal or distal tillers and half with
Eggh proximal and distal tillers. Clones A and B have culm
patterns where half of their rhizome arms have bgth position
1 and 2 occupied and half with either position occupied.

For the same two clones the common tiller pattern is to have

both proximal and distal tillers. When a culm base produces

66

both a proximal and a distal tiller there is a potential for
two new culms the following year on that rhizome arm. The
distal tiller would developlinto a position one culm, the
proximal tiller into a position two culm.

Projecting deve10pmenta1 patterns, clones B, C and D
would have had 50% of their rhizome arms with both positions
1 and 2 occupied by culms next year, if all tillers became
culms. Only 30% of the rhizome arms had both positions
occupied during the current year. Clones A and B would have
had 70% of their rhizome arms with both position 1 and 2
occupied next year, but this year had only 50% with both
filled. The most common tiller pattern should produce the
most common culm pattern if the same culm pattern is
maintained from year to year. Given the distribution of fall
tillers on the whole clones, the culm pattern observed on
these clones this year would likely not be observed again
next year. Either there are losses of proximal tillers over
winter or there is year to year variability in the patterns
of tiller production and culm pattern. Only a study over
several growing seasons could answer this question.

In summary, by consulting the maps, and tables derived
from the maps, it can be concluded that the whole clone
patches are single clones and that the centers are inactive.
The larger patches used for the field transects are probably
single clones (or at most two or three merging clones) with
growth patterns like the mapped whole clones. The age of a
clone can be approximated, given a map of the rhizome

system. There are two functional types of tillers, distal

67
(away from the existing culm) and proximal (near the
existing culnn. The culms at position 1 (formerly distal
tillers from the previous year) produce more tillers, on
average, than culms at position 2. From the calculated
average number of tillers per culm, it also appears that
all five clones were replacing their culms every year and
that some would have produced twice as many culms next year
as they had this year, provided all tillers emerged as
culms.

The whole clone comparisons produced interesting
results. The cones were grown in two separate locations and
there were some differences between clones that correlate
with the location at which they were grown. Clones A, B and
C, were grown at Michigan State University, had higher RE's
than those remaining on the floodplain (D and E), and the RE
values calculated for the field transect clones 35 and 27.
The clones A, B and C also had reproductive culms in
position 2 on the rhizome network, where as clones D and E
did not. These differences are attributed to some
environmental factor which differed between the two growing
sites.

One of the more obvious differences between the
floodplain and the transplant environments was light
intensity. The transplanted clones growing under the higher
light intensity (A, B, and C) had the highest RB values and
the greatest number of points on the clone (both positions 1

and 2) filled by reproducing culms. In contrast, those

68

clones growing under the lower light intensities of the
flood plain had fewer reproductive culms, lower RE values.
Field transect 33, which was not only in the low light of
the flood plain forest but in addition was over topped by a
canopy of nettles had no reproductive culms.

The clones grown at Michigan State flowered in June,
whereas those on the flood plain flowered in late July.
This result could also be attributed to differences in light
levels between the slightly shaded location at MJLU. as
compared to the heavily shaded floodplain.

In order to separate the effects of location from
those of transplanting, clone D was transplanted but not
relocated. There appeared to be adverse effects associated
with transplanting. Clone D had the lowest RE value and the
fewest number of reproductive culms of the whole clones
studied. A larger sample size would be necessary to provide
conclusive results. There could be other reasons for the
poor performance of clone D. This clone was the-oldest and
the largest whole clone studied, therefore age and size can
not be eliminated as factors contributing to the lower RE
and reproductive culm numbers. The RB value for clone D is
closer to that of the field transect clones than it is to
the whole clones. The field transect clones are larger and
are assumed to be older than all of the clones used for the
whole clone studies. These facts suggest the possibility
that the lower RE for clone D could also be attributed to
the age of the clone.

Clones A, B, and C were also transplanted but they did

69

not have low RE values or lower numbers of reproductive
culms. This could either be attributed to a balancing of
any negative effects of transplanting on flowering by
positive effects of increased light, or to there being no
negative effects of transplanting.

Although there are differences in RE values and
flowering among the clones used for this rhizome study,
there is insufficient data provided by this study to allow a
conclusion as to the effects of transplanting and light
levels on the clones. The effects of light on reproduction
appear important and require further study.

Expected changes in rhizome growth pattern such as an
increase in the number of culm cohorts and the regeneration
of older sections of the clone did not occur in clones which
were transplanted and/or grown in higher light levels. It
would appear that the pattern of rhizome growth is fairly
fixed. The results of the fragmented clones showed an
increase in tiller production, but only in the usual areas

of growth.

Seedlings

Seedlings were raised from fruits that were germinated
in the lab. Seedlings were harvested and the developing
rhizome system was sketched (Figure 19). Six seedlings were
examined and the resulting figure is a schematic
representation of the rhizome system. This diagram is not

drawn to scale. Five of the six seedlings produced a bud on

70

   

 

0,.---

 

..-veg. culm O -tiller fi-initial culm

 

..... horizontal culm rhizome

Figure 19. Seedling diagrams, not to scale, of the underground
' development of six laboratory raised seedlings..

71
the stoloniferous seedling culm. 'This bud developed into
shoots and culms. The horizontal culm was rooted to the
soil at a node and was the only addition to the develOping
clone which was not produced by tillering culm bases. This
production of a bud on an above ground portion of a culm
only occurred in the seedlings and was produced only on the
first culm of the seedling.

Seedling establishment is relatively rare at the Red
Cedar flood plain locality. Initial searching in the field
produced a single seedling (located in July) and it
consisted.of a single very small culm, not at all like that
of the mature plant in size. In the late fall more
seedlings were observed. They may or may not have been
present during July. It is not known whether any of these
seedlings were present in June when no seedlings could be
found. All seedlings were found outside a large clone,
approximately 30 cm from the edge of the clone.‘The
seedlings observed in the field in the fall had tillers
emerging from the culm base, as illustrated in Figure 20.
The seedlings in the lab flowered, though none of the
seedlings in the field reproduced nor did they reach the
level of shoot and tiller development that was observed in

the lab seedlings.

Seeds

 

A subsample of seeds was weighed for all clones used in

this study. The data have been included in Appendix I.

72

Figure 20. Photographs of site, seedlings and rhizomes. Field
site 10/25/1985 (1), field site seedling (2), rhizome
of clone E (3-5).

 

GENERAL DISCUSSION

Given the information from the whole patch maps, it is
possible to interpret the results of the field transects
(Figure 8) in several ways. Transect 27 (Figure 8) can be
interpreted as representing one clone with new tillers
appearing at the outer edges, and with rhizome tillers that
are circling back and filling in the older central sections.
An alternate interpretation is that transect 27 is passing
through a patch that consists of two clones that are in the
process of merging, with the right hand clone (as seen on
Figure 8) being older and having a nonreproductive center.
Transect 35 (figure 8) is probably a single clone that is
simply expanding outward with no recolonization directed
inward. Transect 33 (Figure 8) may consist of two
partially overlapping clones. The peaks on the right side
of the transect show very little in the way of the familiar
growth pattern (peaks of activity bordered by areas of
increased tiller activity). The left portion of the
transect may represent a separate, expanding clone. This
transect, even after investigating the underground
construction on other clones, could not be easily
interpreted. Actual excavation of parts of the transect
would be the only way to investigate this further and this
was not attempted.

Alternatively, all three transects could represent

single clones and all of the internal growth could be simply

74

75

a product of the old rhizome system that is circling back
into the center. This type of behavior has been documented
for Spartina townsedii that is growing without competition
on mud flats (Caldwell, 1957).

It can be concluded here that the smaller patches are
single clones and that probably the larger patches also
consist of a single or small number of clones. The
organization of the large patch rhizome systems is probably
the same as that observed for the small patches (the mapped
whole clone patches): the large patches are simply older
clones (or groups of older clones). It is not possible to
determine whether the field transects through the large
patches are sampling one or several clones without extensive
excavation of their rhizome system.

Judging from published studies, the clonal structure
and growth habit of Diarrhena americana is not of a common
type. Published studies report that other grasses form
large ring shaped clones, but none have a rhizome system or
a growth rate like that of Diarrhena. Spartina townsedii
(Caldwell, 1957) forms rings but they are rings of active or
inactive tussocks. The ring pattern of Eggga erecta (Austin,
1968) in turf is only detected by quadrat sampling.

Agrostis tenuis forms rings but the rhizomes branch and

produce tillers more profusely than Diarrhena. Agropyron

 

repens (Palmer, 1958) has been reported to have a cohort of
rhizomes which elongate during the summer, emerge in the
fall and develop into culms the following spring but the

tiller and rhizome development after the spring emergence of

76
the new culms is much different. Up to 50 tillers can be
formed from the inital spring emerging culm (as apposed to

Diarrhenafs 4). Tillers are produced (by Agropyron) at the

 

base of the culm, along rhizomes, and along new tillers and
rhizomes produced in the same season.

Holttum (1955) noted that the rhizome systems of most
grasses are monopodial: some bamboos, which are sympodial,

are the exception. Diarrhena has a sympodial pattern of

 

growth, and in that respect is more similar to the bamboos
than to other groups of grasses. The only published reports

of a rhizome branching pattern resembling that of Diarrhena

 

is that found in some bamboos, as illustrated by McClure
(1966, p. 26). McClure does not give any information on the
growth rate of or the shape of the clones produced by the
different types of rhizomes he illustrates. Any further

comparison with Diarrhena would require that information.

 

A greater number of rhizome systems have been studied
among the nongraminaceous herbs. The forest herbs reviewed
by Antos and Zobel (1984) have rhizome systems that are
characterized by having greatly elongating rhizomes.
Clintonia borealis, studied by Pitelka g£_al;_(l985) and

Aralia nudicaulis, by Edwards (1984) are also unlike that of

 

Q; americana. Two studies on Carex, one on g; lacustris

 

(Bernard 1975) and the other on g; arenaria (Nobel 35 al.,
1979) both address rhizome growth rate and structure.
Neither Carex species has a rhizome branching pattern

similar to that of Diarrhena. Tiller production by g;

77

arenaria is greater than Diarrhena, producing several

 

cohorts of culms per year. C. lacustris is similar to

 

Diarrhena in that it produces one culm cohort per year.
This species differs in that it appears to form dense clumps

rather than forming loose patches. Alpina speciosa (Bell,

 

1979) has some similarities with Diarrhena. The similarity

 

is due to both systems being based on a sympodial branching
system with modules added in a regular pattern, a branching
angle which allows a rhizome to circle back into the clone,
and the restriction of growth to the perimeter of the clone.
The sympodial bamboo illustated by McClure (1966” p. 26)
appears to be similar in branching angle and regularity of
growth to Alpina speciosa and may result in a similar
rhizome network. The similarity of Diarrhena to other
sympodial systems reinforces the idea that the rhizome
system of Diarrehena most closely resembles that of a
sympodially branching bamboo. Although many similarities in
vegetative characters are due to convergent evolution and
habitat similarities, recall that the current thought is to
place Diarrhena with the bambusoid grasses. This study

provides some support for the placement of Diarrhena into

 

the bambusoid grasses.

CONCLUSIONS

Whole clones

l. The small patches of Diarrhena americana consist of

 

single clones.

2. The rhizome branching pattern is sympodial and produces,
in the simplest cases, a clone in the form of an expanding
ring.

3. This ring pattern is complicated by circling back of
rhizome branches from the expanding clone and by asymmetric
growth.

4. Rhizomes occur in two lengths, one that remains proximal
to the existing culm, and one that elongates laterally. The
elongating rhizomes may circle back toward the center of the
clone.

5. Connections between ramets are persistent. This results
in an extensive rhizome network.

6. The rhizome network of Diarrhena most closely resembles

 

that of a sympodially branching bamboo.

7. No regrowth on older rhizomes occurs, even when the clone
is fragmented.

8. The clone produces one cohort of tillers per year. The

tillers emerge in late fall and early spring.

Field transects
9. The larger patches (such as those that were used for the

field transects) may be either single clones or several

78

 

79
merged clones. If the large patches consist of more than

one clone, the number of clones involved is small, probably

less than three separate rhizome networks.

Seedlings

 

10. Seedling establishment does occur on the floodplain.
However it is not common. Several seedlings were located

near existing patches.
11. Seedlings were not observed to recolonize the center of

a clone.

Light levels

 

12. There are indications that light intesity may effect

flowering, reproductive effort and phenology.

LITERATURE CITED

 

LITERATURE CITED

Abrahamson, WJL 1979. Patterns of resource allocation in
wildflower populations of field and woods. Amer. J.
Bot. 66: 71-79.

Anderson, 0.8. 1958. Taxonomy and distribution of the genus
Diarrhena. Iowa State College Masters thesis.
unpublished.

 

Antos, J;A. and DuB. Zobel. 1984. Ecological implications
of belowground morphology of nine coniferous forest
herbs. Bot. Gaz. 145: 508-517.

Armstrong, R.A. 1984. On the quantitative theory of
reproductive effort in clonal plants: refinements of
theory, with evidence from goldenrods and mayapples.
Oecologia (Berlin) 63: 410-417.

Ashmun, J.W., R.L. Brown and L.F. Pitelka. 1985. Biomass
allocation in Aster acuminatus: variation within and
among papulations over five years. Can. J. Bot. 63:
2035-2043.

and L.F. Pitelka. 1985. Pepulation biology of
Clintonia borealis. II. survival and growth of
transplanted ramets in different environments. Journal
of Ecology 73: 185-198.

 

Austin, M.P. 1968. Pattern in a Zerna erecta dominated
community. Journal of Ecology 56: 197-218.

Bell, AHD. 1974. Rhizome organization in relation to
vegetative spread in Medeola virginiana L. Journal of
the Arnold Arboretum 55: 458-468.

 

. 1979. The hexagonal branching pattern of rhizomes of
Alpinia speciosa L. (Zingiberaceae). Ann. Bot. 43:
'209-223.

, D. Roberts and A. Smith. 1979. Branching patterns:
the simulation of plant architecture. J. Theor. Biol.
81: 351-375.

, P.B. Tomlinson. 1980. Adaptive architecture in
rhizomatous plants. Botanical Journal of the Linnean
Society 80: 125-160.

Bar:

.3.

p
.3:

81

Bernard, JQM. 1975. The life history of shoots of Carex
lacustris. Canadian Journal of Botany 53: 256-260.

Bierzychudek,P. 1982. Life histories and demography of
shade-tolerant temperate forest herbs. A review. The
New Phytologist 90: 777-783.

Caldwell, PJL 1957. The spatial development of Spartina
colonies growing without competition. Annals of Botany
21: 203-215.

 

Campbell, C. 1985. The subfamilies and tribes of
Gramineae(Poaceae) in the southeastern U.S. Journal of
the Arnold Arboretum 66: 123-199.

Dickerman, J.A. and R.G. Wetzel. 1985. Clonal growth in
Typha latifolia: pOpulation dynamics and demography of
the ramets. Journal of Ecology 73: 535-552.

Edwards, J. 1984. Spatial pattern and clone structure of
the perennial herb Aralia nudicaulis L. (Araliaceae).
The Bulletin of the Torrey Botanical Club 111: 28-33.

 

Hartnett, D.C. and F.A. Bazzaz. 1985. The integration of
neighbourhood effects by clonal genets in Solidago
canadensis. Journal of Ecology 73: 415-427.

 

Hilu, K.W. and K. Wright. 1982. Systematics of Gramineae:
A cluster analysis study. Taxon 31: 9-36.

Holttum, R.E. 1955. Growth habits in monocotyledons.
Variation on a theme. Phytomorphology 5: 399-413.

Kershaw, KAA. 1958. An investigation of a grassland
community. I. The pattern of Agrostis tenuis. Journal
of Ecology 46: 571-592.

 

. 1959. An investigation of the structure of a
grassland community. II. The pattern of Dactylis
glomerata, Lolium perenne and Trifolium repens. III.
Discussion and conclusions. Journal of Ecology 47: 31-
53.

 

 

. 1962. Quantitative ecological studies from the
Landmannahellir, Iceland. II.'The rhizome behavior of
Carex bigelowii and Calamagrostis neglecta. Journal of
Ecology 50: 171-179.

 

. 1963. Pattern in vegetation and its causality.
Ecology 44: 377-388.

Koyama, T and S. Kawano. 1964. Critical taxa of grasses
with North American and Eastern Asiatic distribution.
Canadian Journal of Botany 42: 859-884.

82

Li, HJH 1952. Floristic relationships between eastern
Asia and eastern North America. Trans. Am. Phil. Soc.
42: 371-429.

Lovett Doust, L. 1981. Population dynamics and local
specialization in a clonal perennial (Ranunculus
repens). I. The dynamics of ramets in contrasting
habitats. Journal of Ecology 69: 743-755.

 

Maun, M.A. 1985. Population biology of AmmOphila
breviligulata.and Calamovilfa longifolia_on Lake Huron
sand dunes. I. Habitat, growth form, reproduction and
establishment. 63: 113-123.

 

 

Macfarlane, ILD. and L. Watson. 1980. The circumscription
of Poaceae subfamily pooideae with notes on some
controversial genera. Taxon 29: 645-666.

McClure, F.A. 1966. The bamboos: A fresh perspective.
Harvard University Press, Cambridge, Mass.

Noble, J.C., A.D. Bell and J.L. Harper. 1979. The
population biology of plants with clonal growth. I. The
morphology and structural demography of Carex arenaria
L. Journal of Ecology 67: 983-1008.

Palmer, J.H. 1958. Studies in the behavior of the rhizome
of Agropyron repens (L.) Beauv. I. The seasonal
development and growth of the parent plant and rhizome.
New Phytologist 57: 145-159.

Perkins, D.F. 1968. Ecology of Nardus stricta L. II. annual
growth in relation to tiller phenology. Journal of
Ecology 56: 633-645.

Phillips, M.E. 1953. Studies in the quantitative morphology
and ecology of Eriophorum angustifolium Roth. The
rhizome system. Journal of Ecology 41: 295-318.

Pitelka, L.F., S.B. Hansen and J.W. Ashmun. 1985.
Population biology of Clintonia borialis.]; Ramet and
patch dynamics. Journal of Ecology 73: 169-183.

 

Pohl, RJL 1966. The grasses of Iowa. Iowa State'Journal
of Science 40: 341-566.

Renvoize, SJL 1985. A survey of leaf-blade anatomy in
grasses V. the bamboo allies. Kew Bulletin 40: 509-
535.

Schwab, CLA. 1971. Floral structure and embryology of
Diarrhena (Gramineae). Iowa State University Doctoral
Disseration. Ames Iowa. unpublished.

 

83

Smith, AHP. and J.O. Palmer. 1976. Vegetative reproduction
and close packing in successional plant species.
Nature,London 261: 232-233.

Tateoka, T. 1957. Notes on some grasses III. Bot. Mag.
Tokyo 70: 9-12.

. 1960. Notes on some grasses. X. Some thought on
Festuceae, Festucinae with special reference to their
morphology. Canadian Journal of Botany 38: 951-967.

Watson, L., H.T. Clifford and M.J. Dallwitz. 1985. The
Classification of Poaceae. Subfamilies and Supertribes.
Australian Journal of Botany 33: 433-484.

Watt, A.S. 1947a. Contributions to the ecology of bracken
(Pteridium aquilinum). III. Frond types and the make-up
of the population. New Phytologist 44: 156-178.

. 1947b. Pattern and process in the plant community.
Journal of Ecology 35: 1-22.

 

APPENDICES

84

 

APPENDIX A
Table 4. Harvested culms from field transects: test of
goodness of fit to a uniform distribution.
Field
transect X2 d.f. prob.
35 21.696* 7 .00287
27 41.316* 10 9.92E-6
33 24.407* 12 .0179

 

 

Data is from table 15.
*= Statistically significant departures from a uniform

distribution.

85

APPENDIX B

Table 5. Field transect fall emerging tillers: test of
goodness of fit to a uniform distribution.

 

Field

transect X2 d.f. prob.
35 9.323 8 .3185
27 17.383 11 .0970
33 13.902 10 .1775

 

 

Data is from table 15.
.All values not statistically different from a uniform
distribution.

86

APPENDIX C

Table 6. Correlation of mean culm dry weight and RE values
for each quadrat along field transect clones.

 

mean culm

quad weight RE
4 .38 .11
5 .15 .00
6 .30 .06
Field 7 .16 .00 Corr. coef= .7635*
trans 8 .00 .00 c.v. (prob=.05)= :_.6642
#35 9 .28 .00
10 .07 .00
11 .17 .00
12 .38 .05
2 .11 .00
3 .27 .04
4 .21 .09
5 .05 .00
Field 6 .00 .00 Corr. coef= .6876*
trans 7 .00 .00 c.v. (prob=.05)= 1 .5740
#27 8 .23 .04
9 .25 .00
10 .21 .00
ll .18 .03
12 .16 .02
13 .36 .14

Field transect #33- no reproduction, no RE to correlate.

 

 

87

APPENDIX D

Table 7. Number of harvested culms from whole clone
transects: test of goodness of fit to a uniform distribution

 

Whole

clone x2 d.f. prob.
A 4.992 4 .2881
8 8.600 4 .0719
C 7.443 4 .1143
D 12.200* 4 .0159
E 1.000 2 .6065

 

Data from table 16.
* Statistically significant departure from uniform
distribution.

 

88

APPENDIX E

Table 8. Correlation of mean culm dry weight and RE values
for each quadrat along whole clone transects.

 

whole mean culm
clone quad weight RE

1 .77 .30
2 .46 .29 *

A 3 .16 .00 Corr. coef= .8519
4 .43 .04 c.v. (prob=.05)= + .8116
5 .65 .30 ‘-
6 .70 .34
1 .38 .13
2 .52 .22 *

B 3 .48 .26 Corr. coef= .8355
4 1.19 .28 c.v. (prob=.05)= : .8116
5 .00 .00
6 .83 .23
l .36 .29
2 .48 .32 *

C 3 .61 .25 Corr. coef= .8147
4 .00 .00 c.v. (prob=.05)= i .8116
5 .64 .25
6 .54 .28
1 .03 .00
2 .22 .00 *

D 3 .00 .00 Corr. coef= .9493
4 .11 .00 c.v. (prob=.05)= + .8116
5 .28 .08 -
6 .66 .25
2 .27 .04

E 3 .59 .18 Corr. coef= .3865 n.s.
4 .47 .00 c.v. (prob=.05)= i.-9611
5 .63 .04

 

 

89

APPENDIX E

Table 9. Data table and correlations of RE, percent
reproductive culms and mean culm weight for whole
clones and field transects. Values correlated are
based on total above ground biomass harvested.

 

mean percent

culm repro1

weight RE culms
whole A .57 .29 68
clones B .49 .28 50
C .52 .29 78
D .19 .08 18
E .57 .16 50
field 27 .21 .50 11
trans 33 .15 .00 0
clones 35 .26 .06 11

 

correlation matrix:
mean
culm
weight RE
RE .3144
pcnt repro. culms .9299* .3737

 

Critical value (Z-tail, prob=.05) =‘i .7048

 

1 percent of total number of culms that are reproductive.

 

90
APPENDIX G

Table 10. Summary of culm distribution illustrated by the whole clone
maps.

Culm Organization

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency of position

Culm Reproductive Vegetative Order of occurrenceZ
position1 1 1 2 2 1 1 2 2 V/R R/V V/V R/R V/D R/D Prec3

w/d w/d w/d w/d Tlrs f
Map A 32 0 0 10 ll 0 0 8 3 6 4 10 2 15 6
Map B 31 l 0 7 7 0 0 6 0 3 1 4 3 22 26
Map C 41 l 0 6 7 2 0 6 0 8 0 8 3 25 2
Map D 11 0 0 0 38 0 0 ll 0 5 8 0 29 6 0
Map E 10 0 0 0 5 0 0 5 0 7 l 0 4 3 0
1 position, see Figure 18 for diagramatic representation E

1- most distal culm
2- one node back from most distal culm
w/da culm occurring at a node with a dead culm base.

”Order of occurrence“ refers to the sequence in which
reproductive and vegetative culms appear in position 1 and 2
notation is in the form of: x/y, x- culm type in position 1
and ya culm type in position 2. Types of culms are:

R3 reproductive culm, Va vegetative culm and 08 dead culm.

3precocioustillers. Earlyemergenceoftillers,probably
induced by artificial growing conditions. See page 56 in text.

91

Table 11 Summary of tiller-culm relationships illustrated by whole
clone maps.

92

.mu0HHHu 0>mz
umzu n0mmn mnocu uOu >Hco ©0umHsuHmu nH 0009 EH30 900 mu0HHHu uo u0nssc 00mu0>0 0:9 .005 ouHuc0 0zu
ecu 0nmn EH50 900 mu0HHHu uo D0959: 00mu0>m 0;» 0:0 ncesHou HHm ecu anuoD m0bH>0uQ coHuu0m nHmuoa 0:9

 

 

 

 

 

 

v

.m0Hu000umu nsoHum> 0;» saw 00Hoc0s00uu 0mm sumo

.m0coHo 0gb :0 99000 coch muwHHHu HneHuOLQ 0:0 Hnuch uo ncoHumcheou nsoHum> 0zu 0am ncumuumm L0HHHB

nu0HHHu HmumHo no 0095:: "N

nu0HHHu HmeonuQ uo u0nesc n»

0009 EH30 L00 nu0HHHu 30c uo 9095:: Hmuo» nx
0n0c: .HN\>V x “No auOu 0gb cH coHumuoz .ncu0uumd umHHHa m

.cu0uumd umcu uo numHHHu cuHs 00009 eHso uo u0nesc HmuoU 0cu
0H Hey cesHou HmHOH 0:9 .n0umcH0Huo L0HHHu m zuHcs eouu 0009 eHsu no 0Q>u 0zu >uHuc0cH n0Hu000umu 00029 N

0nmn eHso Houan unoe sauu xumn 000: 0:0 nN

E0un>n 0souch a co 0MMn eHsu Hmuan unoe nH
coHumuc0n0uQ0u uHuoemu00Hc ecu 0us0Hu 000 .coHuHmom H

EH50
o.N o o.H N.m m.H N.N o 0.H n.N H.N N.N h.H m.H v.N o o.N h.H o.H H.N m.H N.N o.H o.H m.N m.H \nuHa x
0v 0 m Nm 0 HHH o b on 05 NNH 0H m on 0 mm NH N om HH «OH 0 m 05 0 nu0HHH9
0H 0 m 0H m Hm o 0 HH mm mm m 0 m0 0 me n N mN o be m m Nm m neHsu
vnH0u09
o o o o o o o o o o H o o H o o o o o o o o o o o HM\HV v
H o o H o o o o o o o o o o o o o o o o o o o o o Hmev -
H o o- H o H o o H o o o o o c o o o o o o o o o o HN\NV v
m o H H H H o o o H H o o H o m H o v o H o o o H HH\NV m
n o o h 0 0H 0 o m m mH o o mH o m o o m 0 0H 0 0 0H 0 HN\HV m
o o o o o m o H H H N o o N o H o o H o o o o o o HN\oV m
o o o o o o o o o w m o H N o m H o o N H o H o o H0\Nv N
H o o o H HH 0 o N m HH H H a 6 HH N o m m «H N N m N HH\HV N
H o o o H NH o H N a «H m o HH o o o o o o v H o n o HNxov N
o o o o o m o N o m H o H o o h N H m H m H N H H H0\Hv H
v o N o N N o o o N h N m N o m H H m o v H o N H ”Hmoy H
mucu0uum

90HHHH
a Q00 00> Q0u 00> a Q0u 00> Q0u 00> a Q0u 00> Q09 00> a a0u 00> m0u 00> B m0u 00> d0u 00> NeHso
N H N H N H N H N H HcoHuHmom
m o o . m 4 mm:

 

 

 

 

 

.N can H mcoHanoa um n0mmn EH50 0>Huoscoum0u 0cm 0>Hu0u000> Ou nu0HHHu 30: yo dencoHu0H0u 0zu uo >umessm

.m va=0Qa<

 

 

 

 

93
APPENDIX I
Table 12. Individual seed weights (9)1
Clone A B C D E
.0079 .0057 .0090 .0081 .0077
.0100 .0002 .0075 .0093 .0071
.0094 .0054 .0089 .0089 .0063
.0092 .0056 .0072 .0073 .0060
.0095 .0058 .0083 .0073 .0068
.0100 .0043 .0081 .0034 .0078
.0100 .0061 .0094 .0094 .0070
.0094 .0058 .0098 .0051 .0053
.0100 .0002 .0081 .0094 .0075
.0093 .0066 .0082 .0093 .0080
.0100 .0058 .0092 .0086 .0055
.0092 .0010 .0078 .0085 .0041
.0095 .0071 .0082 .0084 .0057
.0099 .0058 .0092 .0053
.0101 .0006 .0090 .0055
.0096 .0072 .0102
.0092 .0059 .0096
.0102 .0054 .0087
.0084 .0002 .0089
.0094 .0070 .0098
.0091 .0054 .0088
.0085 .0002 .0097
.0095 .0067 .0095
.0066 .0059
.0082 .0074
.0102 .0011
.0100 .0065
.0100
.0091
.0090
.0090
.0098
.0098
.0090
x .0094 .0046 .0088 .0079 .0063
s .0008 .0026 .0008 .0018 .0011
S .0001 .0005 .0002 .0005 .0003
Max. .0102 .0074 .0102 .0094 .0080
Min. .0066 .0002 .0074 .0034 .0041
x number of seeds/infl.
16.9 20.0 12.5 4.2 18.5

 

1

For clones A,B and C seeds are from one inflorescence.
For clones D and E seeds are from three inflorescences.

.'

94

APPENDIX J Data Tables.

Tablel3 Field transect dry weights. Above ground dry weight only.
Reproductive effort for each quadrat and the entire transect
are included.

Field Transect Dry Weights

 

 

 

Field

Transect #35 #27 #33

Plot Total Veg. Repr REP1 Total Veg. Repr REP Total Veg. Repr REP
1 0 0 0 0 0 0 0 0 0.31 0.31 0 0
2 0 0 0 0 0.11 0.11 0 0 0.56 0.56 0 0
3 0 0 0 0 1.99 1.92 0.07 0.04 0.44 0.44 0 0
4 4.66 4.13 0.53 0.11 1.16 1.05 0.11 0.09 0.33 0.33 0 0
5 0.60 0.60 0 0 0.05 0.05 0 0 0.86 0.86 0 0
6 1.27 1.19 0.08 0.06 0 0 0 0 0.68 0.68 0 0
7 0.47 0.47 0 0 0 0 0 0 0.43 0.43 0 0
8 0 0 0 0 2.12 2.03 0.09 0.04 0 0 0 0
9 0.28 0.28 0 0 0.75 0.75 0 0 0.72 0.72 O 0
10 0.83 0.83 0 0 1.26 1.26 0 0 1.18 1.18 0 0
11 1.33 1.33 0 0 1.46 1.41 0.05 0.03 0.31 0.31 0 0
12 1.19 1.13 0.06 0.05 2.15 2.10 0.05 0.02 0.48 0.48 0 0
13 0 0 0 0 1.26 1.08 0.18 0.14 0.17 0.17 0 0
l4 0 0 0 0 0 0 0 0 0.15 0.15 0 0
15 0 0 0 0 0 0 0 0 0 0 O 0

 

Total dry weights and reproductive effort for transects:

#35 #27 #33
total repro. total repro. total repro
10.63 0.67 12.31 0.55 6.62 0.00
RBt2= 0.06 REt= 0.05 ast= 0.00

 

l
REp= reproductive effort for plot= repro dry wt/total dry wt

2 REt= total reproductive effort for transect

95

APPENDIX J

Table 14. Culm tagging results.

 

Transect 35 Transect 27 Transect 33
number of culms number of culms number of culms
tagged new died harv tag new died harv tag new died harv
quad
1 0 0 0 0 0 0 0 0 3 0 2 1
2 0 0 0 0 2 0 1 l 3 0 0 3
3 0 0 0 0 12 O 5 7 4 0 0 4
4 11 0 0 ll 5 0 0 5 3 0 0 3
5 5 0 1 4 l 0 0 1 3 2 0 5
6 8 0 4 4 0 0 0 0 3 0 l 2
7 3 0 0 3 l 0 1 0 2 0 0 2
(B 2 0 2 0 ll 1 3 9 0 0 0 0
9 1 0 0 1 3 0 0 3 7 0 l 6
10 6 0 2 4 6 0 0 6 8 1 0 9
ll 7 1 0 8 7 1 0 8 2 0 0 2
12 3 0 0 3 13 1 1 13 4 0 0 4
l3 0 0 0 O 3 0 0 3 3 0 1 2
14 0 0 0 0 0 0 0 0 3 0 1 2
15 0 0 0 0 0 0 0 0 0 0 0 0
Total 38 56 45

 

96

APPENDIX J

Table 15. Field transect harvest and fall tiller data.

 

Number fall Number of culms at harvest:

Sprouting tillers Trans 35 Trans 27 Trans 33
Trans. 35 27 33 Total Rep. Total Rep. Total Rep.

quad. 1 0 0 2 0 0 0 0 l 0

2 0 5 4 0 0 1 0 3 0

3 4 3 0 0 0 7 1 4 0

4 5 3 2 ll 2 5 1 3 0

5 4 2 3 4 0 1 0 5 0

6 3 0 4 4 1 0 0 2 0

7 1 3 l 3 0 0 0 2 0

8 0 8 0 0 0 9 1 0 0

9 1 4 1 l 0 3 0 6 0

10 4 5 1 4 0 6 0 9 0

11 4 8 2 8 0 8 1 2 0

12 5 5 5 3 1 13 l 4 0

l3 0 6 0 0 0 3 l 2 0

14 0 l 0 0 0 0 0 2 0

15 0 0 0 0 0 0 0 0 0

total 38 4 56 6 45 0

 

 

97

APPENDIX J

Table 16. Whole clone transect data.

 

Total # Veg # Rep Total Veg Repro

Clone Plot Culms Culms Culms Dry Wt. Dry Wt. Dry Wt. RE
5 0 . 5 5.50 3.85 1.65 .30
1 0 1 .65 .46 .19 .29
2 2 0 .32 .32 0.00 0.00
l 0 l .45 .43 .02 .04
2 1 1 1.86 1.31 .55 .30
2 0 2 2.11 1.40 .71 .34
2.17 0.50 1.67 1.82 1.30 .52 .21
4 2 2 1.73 1.51 .22 .13
5 2 3 3.31 2.60 .71 .23
4 2 2 2.60 1.92 .68 .26
1 0 1 1.66 1.19 .47 .28
0 0 0 0.00 0.00 0.00 0.00
1 0 1 . 1.08 .83 .25 .23
2.50 1.00 1.50 1.73 1.34 .39 .19
4 3 1 2.03 1.45 .58 .29
6 0 6 4.29 2.90 1.39 .32
2 0 2 1.61 1.21 .40 .25
0 0 0 0.00 0.00 0.00 0.00
3 0 3 2.58 1.93 .65 .25
2 0 2 1.51 1.08 .43 .28
2.80 0.50 2.30 2.00 1.43 .58 .23
1 1 0 .03 .03 0.00 0.00
1 1 0 .22 .22 0.00 0.00
0 0 0 0.00 0.00 0.00 0.00
6 6 0 .65 .65 0.00 0.00
4 2 2 1.19 1.10 .09 .08
1 0 l .88 .66 .22 .25
2.20 1.70 0.50 0.50 .44 .05 .06
0 0 0 0.00 0.00 0.00 0.00
3 2 1 .85 .82 .03 .04
2 l 1 1.44 1.18 .26 .18
1 1 0 .47 .47 0.00 0.00
2 1 l 1.31 1.26 .05 .04
0 0 0 0.00 0.00 0.00 0.00
2.00 1.00 0.75 1.00 .93 .09 .07

.03 >06 Hmuou\.03 >06 0>Huusno0m00 "000000 0>HuusUO0m0m Imm

 

98

H
0H. hm.HH 0H.N m0.MH m.mH 00m ON OH OH a
mo. Hv.HH mo.H m¢.NH N.v 00mH 00 HH 00 Q
0N. ov.mm 0H.¢H 0m.0v m.NH thH mm mv mH 0
mm. mv.0m 0¢.mH 00.0w 0.0N 000 om 00 00 0
mm. 0>.vm mm.mH 00.00 0.0H mum Ho N0 0H <

4mm m0> 0000M H0008 .HucH\c00m m00< H0009 mEHsu mEHsO 0coHo
.03 >00 0:5000 0>on< a .0>< 00m0m .00>

 

 

 

.numc mcoHo mHoez .AH mHnma

H. xHszmg

APPENDIX J

99

Table 18. A sample of apical spiklet, basal floret seed
weights from 23 inflorescences from clone B.

 

seed weights

.0103 .0072
.0083 .0094
.0099 .0097
.0100 .0104
.0025 .0091
.0115 .0083
.0107 .0102
.0102 .0102
.0108 .0103
.0063 .0037
.0100 .0101

.0100

summary

x= .0091
= .0022

 

 
 

WWII

763

      

1111111111111