5:: g...“

 

_‘_____:__::_ 1:,

4‘9

«Kw.

#3!‘

n5 .

Ii! 4‘

3.374%

IIIIIIII IIIIIII ”III III]

312|93 10747 0704

 

 
   
     
      

amfima av
HUM; & SflNS’
800K BINDERY INC.

IIRARY BINDERS

.IIOIIID' “RUBI-

 

ABSTRACT

THE BIOLOGY OF DECODON VERTICILLATUS (LYTHRACEAE)
IN MICHIGAN
By
Sue Ellen Blaisdell

The morphology, systematics, phytogeography, ecology, and repro-
ductive biology of Decodon verticillatus (L.) Ell. are considered.
Nearly 90% of the stalks of this plant were found to have three leaves
per node with the remainder having one, two, or four leaves per node.
Nearly 85% of the flowers were found to be five-merous with the remain-
ing ones being three-, four», or six-merous. Intraspecific variation
in the species with respect to the pubescence character of the leaves,
stems, and inflorescences in considered not significant enough to war-
rant taxonomic recognition. The past and present distribution of
Decodon and its ancestors is discussed. Vegetative and sexual repro-
duction are compared. Other species of the Decodon mat are noted.
Reproductive biology is described in detail. The tristylous flowers
were found to have lengths of parts of each level as expected except
the short-styled flowers in which the styles and long stamens were
shorter than expected. Pollen grains did not show a significant
trimorphism correlated with stamen length. Legitimate cross-pollina-
tions, in the Darwinian sense, were no more productive than illegiti-
mate crosses. Thus, Decodon does not have a typical incompatibility
system complementing the trimorphism. Self-pollinations produced more

seeds but fewer capsules than cross-pollinations of parts of the same

Sue Ellen Blaisdell
lengths as the self-pollinations. The structure of Decodon populations
with respect to style type does not differ significantly from the 1:181

ratio predicted by Darwin for trimorphic populations.

 

THE BIOLGBY OF DECODON VERTICILLATUS (LYTHRACEAE)

IN MICHIGAN
By

Sue Ellen Blaisdell

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
fer the degree of
MASTER OF SCIENCE

Department of'Botany and Plant Pathology

197h

L?“ Q 7C“,\/1

TABLE OF

LIST OF TABLES . . . . . . .
LIST OF FIGURES . . . . . .
INTRODUCTION . . . . . . . .
MORPHOLOGY . . . . . . . . .
SYSTEMATICS . . . . . . . .
PHYTOGEOGRAPHY . . . . . . .
ECOLOGY . . . . . . . . . .
REPRODUCTIVE BIOLOGY . . . .

‘Elggal Morphology . . .

Incompatibilityggystem

Population Structure .
SUMMARY AND CONCLUSIONS . .

BIB LI WRAPHY O O O O O O O 0

ii

CONTENTS

Page

0 O O C O O O O I O O O O O O O 0 iii

. . . . . . . . . . . . . . . . . iv
. . . . . . . . . . . . . . . . 1
. . . . . . . . . . . . . . . . . 3
. . . . . . . . . . . . . . . . . 7
. . . . . . . . . . . . . . . . . 19
. . . . . . . . . . . . . . . . 27
. . . . . . . . . . . . . . . . . 27

o o o o o o o o o o o o o o o o o 36
o o o o o o o o o o e o o o o o O 39
o o o o o o o o o o o o o o o o 41

Table

annotat-

LIST OF TABLES

Leaf measurements of specimens from personal collection .

Leaf measurements of specimens in the Michigan State Uni-
verSity Herbarium O O O O O O O O O O O O O O O 0 O O O 0

Species of the Decodon mats of nine localities in Michi-

gan . . . . . . . . . . . . . . . . . . . . . . . . . . .
Measurements of flowers from herbarium Specimens . . . .
Measurements of preserved flowers from Mud Lake . . . . .
Measurements of pollen from the six types of stamens . .

Results of legitimate and illegitimate pollinations . . .

Page

12

13

23
29
29
32
34

Comparison of selfed and crossed illegitimate pollinations 37

iii

LIST OF FIGURES

Page

Distribution of pubescent, glabrous, and intermediate plants
OfDeCOdonver‘tlclllatUS....... 0 0000000000

Distribution of pubescent, glabrous, and intermediate plants
of Decodon verticillatus in Michigan . . . . . . . . . . . .

Comparison of petiole lengths of glabrous plants, pubescent
plants, and both combined . . . . . . . . . . . . . . . . .

Measurements of styles and stamens of the three flower forms

iv

10

14
30

INTRODUCTION

Studies of the biology of individual species provide useful data,
especially to taxonomists concerned with more inclusive taxa and to
ecologists studying interactions of different kinds of plants (Pen—
found, 1952; Tessene, 1969). Nevertheless, research of this type is
rare, even for common members of the local flora. This study of
Decodon verticillatus (L.) Ell., a monotypic genus of the Lythraceae,
is one such example. Decodon is an emergent aquatic ferming mats at
the margins of lakes and slow flowing rivers and in very wet marshes in
eastern North America.

This study considers the morphology, systematics, phytogeography,
ecology, and reproductive biology of Decodon. In addition, it aims at
answering two specific questions. First, what is the nature of the
breeding system? Decodon is a heterostylous, trimorphic plant. As in
other plants with trimorphic flowers, the stamens are at three levels.
Short-styled flowers have mid and long stamens; mid-styled flowers have
short and long stamens; and long-styled flowers have short and mid sta-
mens. Since the style type is genetically controlled, only one flower
form is found on each plant. Studies in Lythrum salicaria (Darwin,
1889) and Pontederia cordata (Ornduff, 1966) revealed that in each case

an incompatibility system was linked with the trimorphism. Crosses
using styles and stamens of the same levels (eg. long style x pollen
from long stamens) were highly fertile, whereas crosses between

1

different levels showed a low percentage of fertility. We may ask,
therefore, does Decodon also have an incompatibility system linked to
its trimorphism?

Secondly, is there sufficient evidence to warrant taxonomic recog-
nition of the two varieties which were described by Torrey and Gray
(18h0) on the basis of presence or absence of pubescence?

Two summers were spent making field observations in the Lower
Peninsula of Michigan. In addition to field observations, herbarium
specimens were examined from the Smithsonian Institution, University
of Michigan, Western Michigan University, and Michigan State Univer-
sity. Collections were made of Decodon and the vegetation associated
with it. Voucher specimens are deposited in the Michigan State Univer-

sity Herbarium.

MORPHOLOGY

The underground or, as the case may be, underwater parts of 239g
gggg consist of woody rhizomes which are often an inch thick at the
base of the upright stem, and various kinds of true roots including
hair-like red roots and thicker (1 mm wide) white roots.

Spongy aerenchyma surrounds the woody tissue of the rhizomes and
stem-bases. It also forms on any part of the stem which comes in con-
tact with the water. This tissue and its development were discussed
in detail by Schrenk (1889). Generally, stalks which have grown long
will arch and touch the water, form spongy tissue, and then fern white,
red-tipped roots. They soon become attached to the substrate and form
new plants. In the fall, when all the stalks die, the new plants are
separated.

The main stem is generally hexagonal in cross-section with three
leaves at each node. However, secondary shoots and submerged shoots
usually have square stems and two leaves at each node. Occasionally,
plants are feund with alternate leaves and 6-ang1ed stems or with four
leaves per node and B—angled stems.

Main stalks were checked at random along the water edge of a
Decodon mat at the Mud Lake locality (see page 25). The number of
leaves at each node was recorded. These, without exception, corre-
sponded to the angles of the node as previously indicated. The results

were as follows:

a

 

 

leavesper nogg number of stalks ‘percent of total
1 6 2.4
2 6 2.4
3 221 89.1
a 15 6.0

The leaves are generally lanceolate or linear lanceolate, 6-10 cm
long and 1.5-2.5 cm wide on the main stem. The greatest width of the
leaf is from one-half to one-third the length of the leaf blade. The
petiole is generally 2-9 mm long. The lower leaves are usually longer
and wider, often about 15 cm long and h cm wide. The leaves of second-
ary branches and those (undifferentiated bracts) subtending inflores-
cences become gradually shorter and narrower. All leaves are a rela-
tively light, bright green color. They turn a brilliant, deep red in
September, and fall soon after.

Since an inflorescence is found at the base of each leaf of the
flowering stalk, a typical stalk has three inflorescences in a vertical
at each node, each inflorescence having eight flowers. An individual

inflorescence is made up of two parts, one superimposed on the other
(Koehne, 187w). The top part has five flowers and two bracteoles,

while the bottom has three flowers and two bracteoles (see diagram).

top part bottom part

<_——bracteole

Kpoint of attachment <—— flower

___..‘.__~.._—»__,.___ _._

The bracteoles are 10-12 mm long and 6-8 mm wide. They fall off before
the fruits mature. The center flowers open first and these are more

likely than the others to be 6-merous instead of the normal 5-merous.

5

The outer flowers are more likely to be h—merous.

Pubescence in varying amounts may be present on the stems, under-
surface of leaves, and inflorescences. The hairs are uniseriate and
usually multicellular and branched. They may be entirely lacking or
abundant on all organs, producing a brownish, fuzzy appearance: or
there may be intermediate degrees of pubescence. Hairs may be abundant
on the leaves but absent from the stems and inflorescences. or vice
versa.

The flowers are generally 5-merous, having five calyx lobes, five
appendages, five petals, and five stamens. However, three-, four-, and
six-morons flowers are occasionally found. In a collection of inflo-
rescences from the Marsh near the German Lutheran Church in Jackson
County (see page 22), the buds, flowers, and fruits were feund to have
multiples of floral parts as follows!

number of buds,

 

 

flowers & fruits percent of total
3-merous 7 .6
h-merous 134 11.0
5-merous 1035 84.9
6-merous Q3 3.5

The flower has a cup- or urn-shaped hypanthium surrounding the
ovary. This measures 3.5-5.5 mm wide and 3.5-5.0 mm long at anthesis.
The top edge of the hypanthium has ten triangular protruberances from
which it gets the name Decodon meaning "ten teeth”. Five of these,
the sepals, are obtuse triangles 1-2 mm long. The alternating five are
calyx appendages which are narrow, subulate, often twisting, and 1.5-
2.5 mm long. These two types of teeth fit snugly together at their
edges to form the valvate bud.

The corolla consists of pink, crinkled petals, 6-10 mm long and

6

Z-h mm wide. They are attached at the base of the calyx appendages.
The androecium consists of ten stamens attached to the hypanthium
midway between the base of the ovary and the point of attachment of the
petals. Five are shorter in length and are attached directly opposite
the petals, while the five longer stamens are borne opposite the
sepals. Generally, long stamens protrude 15-20 mm from the base of the
ovary: mid stamens, 9-1“ mm; and short stamens, 4-7 mm, Just reaching
the edge of the cup. The largest anthers are just over one mm long
and these are found only on the long filaments. Medium-sized anthers
are found on mid-lengthed filaments, and small anthers, just slightly
over half the size of the large ones, are found on the short filaments.
The pollen grains are 24-29 microns in diameter and are tricolporate.
The gynoecium consists of one ovary with three or four carpels.
The base of the ovary is modified into a nectar-secreting gland. The
single style may be long (15-20 mm), mid (9-1“ mm), or short (4-6 mm).
The stigma, which is more or less globular when receptive, is about
twice the diameter of the slender style and is papillate. The long and
mid styles form an abscission layer and break off before the fruit is
nature, but the short styles persist longer and are destroyed when the
loculicidal capsule dehisces. The seeds are tetrahedral-shaped, brown

when mature, and about 1-1.5 mm in their greatest dimension.

SYSTEMATI CS

Decodon verticillatus (L.) Ell. was first described as Lythrum
lggggpillatum by Linnaeus in Species Plantarum (1753). In Elgrg
Caroliniana (1788), Halter called the plant Anonymous aquaticus. How-
ever, since Anonymous is not considered a proper name according to the
International Code of Botanical_Nomenclature (Stafleu, et al.. 1972),
this name is invalid. In Systema Naturae (1791). Gmelin first used the
currently accepted generic name along with the specific epithet which
Halter used, thus calling the plant Decodon aguaticus. Later, the
plant was named Decodon verticillatum by Elliott in A Sketch of the
Botany of South Carolina and Georgia (1821). Humboldt, Bonpland, and
Kunth, in 182“, transferred Decodon to the genus Nggggg and retained
Linnaeus' specific epithet (m verticillata mm).

Thus, the species has been treated in three different genera.
Decodon verticillatus is not, however, closely related to the genus
thhrum. In his study of the Lythraceae, Koehne (1903) placed Lythrum
in a different subfamily on the basis of the ovary septa being inter-
rupted above the placenta as opposed to the complete septa of the sub-
family in which he placed Decodon. Koehne, however, considered.§ggggg
to be more closely related to Decodon, but he retained it as as separate
genus on the basis of the capsule dehiscence which is by'a small oper-
culum.in 323523 and loculicidal in Decodon. The temperate distribution

of Decodon versus the pantropical distribution of’Nesaea also supports

7

this separation.

Two varieties of Decodon verticillatus were designated by Torrey
and Gray (1890). Decodon verticillatgg var. bescens, with stems and
lower surface of leaves more or less tomentose-pubescent, was appar-

ently based on the type of Linnaeus and should, therefore, be called

 

Decodon verticillatus var. vertipillatus. The other variety, Decodon
'zggticillgtus var. laevigatus, is glabrous with bright green leaves.
Torrey and Gray (18h0) noted that the pubescent variety is found in the
southern states and the glabrous variety in the North. Fernald (1917)
looked more closely at the distribution of the varieties. He found
that the first variety ”shows a strong inclination to fellow the coast-
al plain and related areas,” whereas var. laeviggtus is ”rare or local
in the coastal plain region but more general inland.“

Until more populations can be studied in detail, I am inclined not
to recognize infraspecific taxa. Nevertheless, it should be noted that
there is a slight geographic segregation in the distribution of the
pubescence character. Figure 1 shows the distribution of pubescent,
glabrous, and intermediate plants throughout the range. The southern
half of the Atlantic coastal plain is inhabited only by pubescent
plants. Pubescence may predominate here as an adaptation to reduce
water loss. The high transpiration rate of these plants became obvious
to me when they were being grown in cans which, during the hottest and
sunniest times in the summer, needed to be filled with water twice a
day. At the northern end of the range, glabrous plants predominate,
possibly due to their ability to better withstand the winters. In His-
consin, Ugent (1962) found glabrous plants to be distributed further

north than the pubescent form, but in Michigan (see Figure 2), both

5—".-- -L___-

 

 

 

 

 

 

mspfiflaoafig sovoomn
mo sundae sewage—Hound use .295us .uceomepan mo soavspahvna

 

 

. H NEG:

10

m 2. Dietrihtiu of pubescent. glam, and inc-saute plants of
new 1- Mom-I-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11

forms are widely distributed in the Lower Peninsula, although glabrous
plants predominate to the north.

No apparent ecological differences are found between the glabrous
and pubescent plants. In two lakes in southern Michigan with extensive
Decodon mats (Park Lake and Mud Lake: see page 25), both glabrous and
pubescent plants are abundant along with many different intermediates.

Torrey and Gray (1840) described var. laevigatus as being glabrous
with bright green leaves. This does not constitute two characters
separating the varieties, however, since the bright green color is due
to the lack of pubescence. In the dull, often brownish variant, color
is due to the presence of pubescence.

Using a sample of specimens from Michigan, various leaf parameters
were measured from one leaf of the first whorl below the lowest inflo-
rescence. The resulting data are given in Table 1, comparing the gla-
brous and pubescent variants. Student's ”t” tests were made to deter-
mine significant differences between the two. The results show that
only petiole length seems to show correlation with pubescence, the
pubescent plants having longer petioles than the glabrous plants.

Another similar sample was measured using specimens from Michi-
gan State University. This time three random leaf measurements were
made from each specimen: these were averaged to give the data in Table
2. Again, 't' tests were made, and only petiole lengths showed a sig-
nificant difference. A histogram was made (Figure 3) showing the dis-
tribution of petiole lengths using the same specimens used in Table 2.
This shows an extensive amount of overlapping, and the character would
not appear to be satisfactory fer separating varieties.

To determine pubescence types for Figures 1, 2, and 3 and.Tables

12

 

.ecenoeeas mavenoaeaemamm
.ssmanoaz :« umpoeaaoo who: mcosaoomm HH<

 

 

 

 

 

 

 

 

 

 

 

 

 

N
.ooswommnoamca use :oaop anon: pmawm m2» mo mesa use song smxmo mm: aqueousmsos commH
mwom.w a v mus. I p :0:.« I v :50.“ I p
we.m om.“ om.m~ om.es oe.mH mm.ea om.sm _ ms.ee x
me.s mm.~ um.nm mm.mme mm.nm as.mm so.mam ms.oeo Nx
me. mm. we.m ma.e ~m.~ mm.a ma.m ma.a x
o. 0.: m:.~ w.m
a. m.m m.N m.a
.3. Oe+N meH 000
we mom “we.“ New
a. a. s.m m.s H.N mm.a N.a a.@
God .3. 3.“ men Wham mHeN 3eHfi How
u. m. 0.: m.m «.N :.N m.o« m.m
mm. m. m.: m.s ma.~ mm.a n.0e ma.e
m. o. m.m m.d mm.m o.m m.m 0.0
m. o. o.m o.m mm.H n.a “.5 m.w
useomonsm msompsao psmomonsm mmmmmmwm mmmmmmmmm. mdohpsau easemenzm msompmau
Asov cameos oHoneoa omens neeax\neweoa Asov ease: unog Aeov apnoea ewes

N.H:oapooaaoo Hmcowmem scam msmsfioomm mo mvsosensmsms when .H mqm<e

dEZuLELCI >Odm.»c.>«r_3 Stored“. CTNdCOHZ 029 Cd DCTEfiOTCT .MC TGCQETLSMOTE .HTOQ .N qu‘fi

 

.smsaoemm ssammpmmz mco scam mm>soa amaze mo mmspm>s on ma vcmseusmmes momma

13

 

 

 

 

 

 

 

 

sea.s . e aeo.~ . e oas.a . 9 now. - e

amn.m mmo.m ea.ms -.~m ms.e~ aw.m~ em.me aN.N0a x

moo.m smm.m mo.aea mm.~a~ m~.om aa.mm Hm.swa as.~mm mx

one. Lem. me.m No.3 a~.~ oe.e ae.a no.5 x
mma. me.m ms.~ oa.m

use. new. em.~ mm.s ms.m mm.~ ea.a em.e

mmm. mmm. we.m me.m ~a.~ ms.m om.s na.e

use. mmm. ma.m 35.: am.m ma.a am.m ce.w

00:. com. aH.: so.m ca.“ mm.m ao.a om.e

mmo. use. ma.: es.m om.~ om.a aa.o m~.o

new. con. mm.s mw.m aa.a aa.m o«.m mm.m

mme. mmm. me.m mm.m ms.m oH.~ oo.m mm.m

use. oom. e~.m mm.m mm.~ oo.a mm.w No.m

coo. new. om.m me.a Na.~ ma.“ mm.m ee.a

men. 00:. mm.m #3.: ma.~ as.“ m~.e aH.e

com. mmm. aa.m no.3 mm.a mo.~ om.s om.m

oos. awn. on.m no.3 am.~ oe.a as.m mm.a

 

 

 

 

 

 

pcmowepzm msomnmau
Aeov sesame oaoaeoa

ascomepzm msoppsau
Asov npmceq mesa

vsoowensm msoupmao
oases aesaz\semeoq

pcmomopsm machnsao
Aeov sees: anoa

«snaumnumx hvdmmo>ass spasm sewage“: one an maesdoomm mo memesehsmmes mesa .N mqm<s

FIGURE 3.

GLKBROUS

PUBESCENT

BOTH COMBINED

1h

Comparison of petiole lengths of glabrous plants,
pubescent plants, and both combined

_ .m._--. m4...--r__...._.__..q.

 

 

 

 

 

 

1.5 2.5 3.5 “.5 5.5 6.5 7.5 8.5 9.5
petiole length in millimeters

 

  

 

 

 

’1.5 2.5 3.5 “.5 5.5 6.5 7.5 8.5 9.5
petiole length in millimeters

 

 

 

 

 

 

1.5 2.5 3.5 “.5 5.5 6.5 7.5 8.5 9.5
petiole length in millimeters

25$
20!

1’

«2m

195

105

' 305
I 25‘
I 205

10‘

15

1 and 2, leaves, stems, and inflorescences were each rated on a basis
of zero to three, zero being no pubescence whatsoever and three being
very pubescent. If the specimen received a total rating of five or
more. it was considered pubescent. If it received zero, it was gla-
brous. If it had a total rating from one to four, it was considered
intermediate.

The frequent occurence of both forms together and the numerous
intermediates are the major reasons I do not at this time feel that

two varieties should be recognized.

PHYTOGEOGRAPHY

The Lythraceae apparently originated in the Indo-Malayan region of
the Old World during the lowermost Eocene (Graham and Graham, 1971).
From there, the family spread and radiated into many new genera and
species. Fossil genera very similar to Decodon flowers and fruits,
named Sahnianthus and Enigmocarpon respectively, have been found in the
Deccan Intertrappean Beds of India. These beds are dated Uppermost
Cretaceous to Eocene. The flowers are trimorphic, but they differ from
Decodon in that they are apparently h-merous and 6-8-loculed. Fossil
seeds of three species of the extinct genus Diclidocarya were described
by E. M. Reid (Nitikin, 1929) from Tertiary deposits in Europe and cen-
tral Russia. Two of these were later transferred to the genus Decodon.
One remaining species, Diclidocarya menzelii. is found only in the
Oligocene deposits of Senftenberg, Germany. Decodon gibbosa, which has
much more narrowly triangular seeds than the extant Decodon, is found
only in the Pliocene of Pont-de-Gail, France. Decodon globosus, which
is very similar to Decodon verticillatus, is found in several deposits
of the Oligocene and Pliocene from France, Germany, Belgium, the Nether-
lands, and the Voronezh District of Russia. It appears that the fore-
runners of the extant Decodon migrated from India across Asia to Europe.
More the genus diversified. The ancestors of Decodon verticillatus
were apparently carried to North America from Europe by some unknown
means just prior to the Pleistocene glaciations. The glaciers then

16

 

17

exterminated those remaining in Europe. Similarly, many other plants
have been exterminated in Europe because the east-west mountain ranges
prevent migration southward away from advancing glaciers.

The present-day distribution of‘Q. verticillatus appears to have
been effected by such factors as the Pleistocene glaciation, migration
patterns of water-birds which disperse Decodon seeds. presence of lakes
and marshes, and the climate. For example, in Uisconsin, Decodon is
found only north and east of the Cary End Moraine of the most recent
glaciation where it is relatively abundant (Ugent. 1962). It is entire-
ly lacking from the nearby unglaciated "Driftless Area". It was appar-
ently able to compete only in the new unvegetsted lakes left by the
glaciers.

waterfowl flyways in North America generally, but with many excep-
tions, follow the seacoast northward. then fan out to cover the area
from British Columbia to Nova Scotia (Lincoln, 1950). It is not sur-
prising. therefore, that Decodon which. according to Ridley (1930), is
duck-dispersed is most abundant along the Atlantic coast and in the
northeastern states.

The distribution of Decodon is also dependent on the distribution
of lakes. in particular those with muddy bottoms. and marshes which
have standing water the year around. These habitats are common along
the Coastal Plain and in glaciated areas, but they are almost non-exis-
tent on the Great Plains. The species is therefore restricted to the
eastern portion of North America. It is interesting to note that in
southern Michigan the few counties in which no records of Decodon have
been feund are flat with extensive farm-land and few lakes.

In the northern part of the range of Decodon, the cold climate is

18

a limiting factor. There, seed germination is low, and the seedlings
do not grow fast enough to become established. The distribution of
Decodon is thus bounded on two sides by ocean. on one by the Great

Plains, and on the other by a cold morthern climate.

ECOLOGY

Although Decodon verticillatus produces an abundance of seeds,
vegetative reproduction is more important than sexual reproduction in
the species. Seedlings have not been observed in Michigan lakes appar-
ently because the conditions for germination and establishment are sel-
dom met. These conditions are high temperature and bright sunlight.

In the laboratory, the seeds placed on moistened filter paper in petri
dishes germinate best at temperatures above 72°F and in the light.
Germination begins in ten days to twowweeks or in about four days for
well-scarified seeds. Even then, only a relatively small percentage of
seeds germinate. By the time (usually July in southern Michigan) that
the conditions in the field have been appropriate long enough to pro-
duce seedlings, the older plants have already grown and shaded the
seedlings growing on the already-formed mat. Thus. the seedlings do not
compete well with the mature plants and survival is greatest where there
is no overhead (for them) growth. They must also be buoyed up by plant
material such as sedges at the edges of bogs or floating mud-jams
caused by water-lily uprootings. Such places provide the high tempera-
tures conducive to germination and seedling development. The seedlings
must be well protected from buffeting water until they are large enough
so as not to be buried in the mud or submerged in the water. Thus. a
specific microhabitat. resulting in the correct nicroclimate, is neces-
sary for establishment. Since these microhabitats are not common.

19

#17:

 

20

neither are Decodon seedlings common.

However. vegetative reproduction is a continuous and apparently
successful process. Near the end of July (in Michigan). the stalks
have grown long enough that they are unable to support themselves up-
right. They soon bend over and the tips gradually touch the water.
Then the tips grow upward again. and spongy tissue and roots are formed
on the part touching the water. Eventually. a segment of the stem,
usually 25 centimeters or more in length, will produce roots about 10
centimeters long. The rooted portion may become attached to the sub- ‘wu
strata or it may float free in open water. In the fall. the stems die
but often remain attached to the parent plant until spring. At that
time newly formed plants that are not securely attached to the sub-
strate are easily dislodged and buffeted about the lake. They may
either be caught by snags of vegetation or end up on the leeward side
of the lake. In either case. they may be dispersed a considerable dis-
tance. depending on the size of the lake. and start new clones. If.
however, the new plant remains attached, it may contribute to the out-
ward expansion of the mat or to the development of a new mat, if the
parent plant is relatively young. Older mats may also be reinforced by
this type of growth. Decodon also reproduced by underground, woody,
rhizomes which are generally short (5-30 centimeters between vertical
stalks). It is this underwater lattice which results in the formation
of a sturdy. but holey, mat.

Th¢.spongy aerenchyma formed by Decodon is very important to the
vegetative reproduction of the plant. This tissue provides the buoyancy

critical to the building of the mat and survival of new plants formed

by vegetative means. If, by chance, this tissue is destroyed (it is

21

easily stripped off the stem), it will readily grow back. Three days
after removal of aerenchyma, the stem appears to split. and white,
spongy tissue begins to emerge. within two weeks, the aerenchyma is
fully replenished.

Since Decodon is an important mat-forming species, it establishes
its own habitat. The flora associated with this mat habitat can be
separated into three groups: 1) submerged and floating aquatic plants
living in the puddles between Decodon plants. 2) plants developing early

and flowering before Decodon is leafed out. and 3) plants growing at [ «

 

the bases of the stalks and forming hummocks when the old stalks die.

Several species of Utricularia are common in shallow water between
Decodon plants associated with rich organic material. They need very
little light and can thus survive there even in mid-summer. Submerged
aquatics such as Potamogeton, Ceratgphyllum, Myriophyllum. and.§ajg§
gagigg may also be found beneath the plants. particularly if the mat is
more open. Duckweeds (Wolffia. Spirodela, and £2223) are often scat-
tered beneath the plants of the mat where they may be blown. Here they
are protected from rough water, but they do not receive much sunlight.
and their growth is limited. Often, however. no other species are
found in this shallow. muddy area especially in newer mats.

Several species of.§agg§ and Eleocharis are well-adapted to life
on the Decodon mat. They, somewhat like the spring flowers in a decid-
uous forest. grow early and flower before the leaves come out. Eggr
ggmine palustris and_Lysimachiapthyrsiflora also flower early in the
yearly cycle of the mat.

Hammocks are formed by the Decodon mat. These usually occur far-

thest from the water where the plants are smaller and the areas between

22

and below the plants are usually filled in. First mosses then
Thelyptris palustris and a number of angiosperms grow on the base of
the Decodon plant. Eventually the Decodon plants fail to emerge in
the summer, and a well-formed hummock is left.

Each Decodon locality has its own unique characteristics and flora.
In the course of this study many of these were observed. In the fol- Ewfi
lowing paragraphs, brief accounts are given of the habitats at nine

localities which were studied in considerable detail. Table 3 docu- %

 

ments the flora of the Decodon mat at each locality. r :1
1. Marsh near German Lutheran Church, Waterloo Recreation Area,

Jackson County, Waterloo Township,_Section 27. This is a very wide

belt of wet, marshy land following the Portage River. Reithmiller

Road crosses the marsh and the river near the church. East of the road

are large clumps of Decodon scattered widely throughout the marsh.

Between the Decodon patches is Open water, occasional hummocks with

sedges and grasses on them, and patches of Nuphar and Pontederia.

 

Sparganium, Typha, and Scirpus also are common. This marsh is almost
impenetrable by humans since it is too deep for waders and too thick
fer boats. This is one of the few marshes which is wet enough year
around for Decodon. The only other marshy habitats I have observed in
which it grows are along creeks and rivers such as the Huron River
(Hashtenaw County) and Wilder Creek, a tributary of the Kalamazoo River
(Calhoun County).

2. ‘ngg:Lake, Yankee Springs Recreation Area,_Barry County,
Yankee '32 ings Township, Sections 20-21 and 28-23. Most of the Decodon
grows at the rear of coves on the north and south ends of the lake.

There it forms a relatively narrow (3 meters wide) and uniform band

23

TABLE 3. Species of the Decodon mats at nine localities in Michigan1

Michigan Center

Lake

German Luthern

Church
yoLong Lake
South Lake
Hankard Lake
h—mile Lake

Marsh near

Mud Lake
Park Lake

Bear Lake

Chara sp.
Fontinalis sp.
Onoclea sensibilis
Thelyptris palustris x
Typha SPD- x x x
Sparganium spp.
Potamogeton spp.
Najas marina x x
Sagittaria spp. x x
Elodea sp.

Calamagrostis canadensis x x
Leersia spp. x
Phalaris arundinacaea

Zizania aquatica

Carex aquatilis x
C. hystericina

C. lasiocar

Carex spp. Eithers) x
Eleocharis spp. x x
Cladium mariscoides

Scirpus acutus x
S. subterminalis

S. validus x x
Peltandra virginica x
Wolffia columbiana

Spirodela polyrhiza x x
Lemna minor x x x x x

L. trisulca x

Pontederia cordata x x
Juncus sp. x

Liparis loeselii x

Ceratophyllum demersum x x

Ulmus sp. x

Boehmeria cylindrica

Pilea fontana x x
Betula pumila x
Polygonum Spp, x x x x x
Rumex orbiculatus x x
Triadenum sp.

Viola blanda

Salix spp.
Cardamine palustris

X

XX
X
X

XX

XXXXX X X
XXXXX XXXX X X
X

XX
X

XX
XX

X
XX

XXXX

24

Table 3 (cont'd.)

Michigan Center

Lake

Marsh near
German Lutheran
Church

Park Lake
Hankard Lake
4-mile Lake

Mud Lake
Bear Lake

Long Lake

Lysimachia thyrsiflora
Rosa sp.

Spirea alba

Lythrum salicaria x x
Myriophyllum spp.
Proserpinaca palustris
Cornus stolonifera x x
Impatiens sp. x x
Hydrocotyl sp. x

Cicuta bulbifera x x

Cassandra calyculata x x
Vaccinium sp. x
Asclepias incarnata x
Cuscuta gronovii x x x
Solanum dulcamara
Lycopus spp. x
Scutellaria spp.
Gerardia purpurea x
Utricularia gibba

U. intermedia

Us minor

U. purpurea

U. vulgaris x
Campanula aparinoides

Cephalanthus occidentalis

Galium trifidum

Aster junciformis

Bidens spp.

Eupatorium perfoliatum

x x South Lake

X

XX

 

X
X
X

X

XXXXXX X
XXXXXXXXXXX
XXXX XX

1Each x denotes the presence of one taxon at one locality.

25

across the edge of the cove. The habitat seems well-suited to the
brackish-water species Najas marina, which grows in the submerged mud
in and around the Decodon plants. This is an interesting lake because
bubbles of swamp gas emerge behind a row boat as the mud is disturbed
by the ears.

3. South Lake, Jackson County, Sandstone Towflship, Section 2. fl 4
An old Decodon mat occurs mostly on the west side of this small lake. Hmm‘
Diversity of the mat flora (see Table 3) may appear greater for this

r "‘ 'f-

locality since more extensive collections have been made there through-

 

out the growing season than at the other localities.
b. Mud Lake,_Haterloo Recreation Area, Hashtenaw County, Lyndon

Township, Section 31. This lake was formed by the construction of a

 

dam across a small creek flowing through a marsh. Extensive Decodon
mats have formed on all sides except the east, and several Decodon
islands have deve10ped in the lake. Broad areas of shallow water
apparently are conducive to the growth of large mats of very robust
plants.

5. Park Lake,gC11nton County, Bath Township, Section 28-29. Park
Lake is very similar to Mud Lake in that it has large, shallow areas
with large mats. It is polluted from sewage from the houses along the
east and north shores. Along the west shore massive areas of Decodon
mat have formed, probably within the last fifty years. Some Decodon
islands have also formed, and occasionally a single plant has been able
to establish itself in the shallow, open water. This lake has been
treated for excessive weed growth with no apparent harm to the Decodon.
Like Long Lake, Park Lake has Najas marina growing in and around the

Decodon.

26

6. Bear Lake,_Michigan State University,p1ngham County,Lansigg
Township, Section_3§. A bog with thick peat beds surrounds the open
water of Bear Lake. There is a relatively small floating bog mat. The
Decodon is often mixed with Cassandra calyculata on the bog margin.
Elsewhere, it is the principle species forming the mat encroaching upon
open water.

7. Rankard Lake, Waterloo Recreation Area, Uashtenaw County, Lyn-
don Township, Section 20. This small lake has a homogeneous margin
dominated by Decodon, Typha, and Cassandra.

8. Michigan Center Lake, Jackson County, Napoleon and Leoni Town-
'ghipg. The Grand River provides the drainage for this large lake. The
only area where Decodon was found was the northeast part near the fish-
ing access. Extensive mats are formed which border the marshes to the
east, but these have a greater admixture of other species than does,
for example, the pure Decodon mat at Park Lake.

9. Four-mile Lake,_flashtenaw County, Lima Township, Section #.
This is an unusual place to find Decodon since it is a calcareous lake
with lime pits nearby. Lyphrum salicaria, which is more often found in
calcareous habitats, grows abundantly in the marshes at the edge of the
lake. In shallow water, Decodon has rooted and become established in
several places. Near the fishing access at the south end it has formed

a mat over accumulated organic debris.

REPRODUCTIVE BIOLOGY

Tristylous breeding systems were first studied in considerable
detail by Charles Darwin (1889). Ip_Different Forms of Flowers on
Plants of the Same Species, Darwin presented data on tristylous and
other species of Lythrum, Oxalis, and Pontederia. He also mentioned
Decodon verticillatus and presented measurements of pollen grains.
Since Darwin, many studies on the genetics of trimorphism have been
conducted (Barlwo, 1923; East, 1927, 1932; Fisher and Mather, 191.3).
but, until lately very little has been done to study the incompatibility
system thought to be associated with trimorphism. Ornduff (196b, 1966)
recently investigated the breeding systems of Oxalis suksdorfii and
Pontederia cordata. Darwin's work on Lyghrum and the studies by Orn-

duff have served as guides for this research.

‘Elgggl Morphology

Theoretically, in a tristylous flower, the length of the reproduc-
tive parts at each level should be the same as those on all other plants
regardless of the style type. For example, the mid-length style and
the mid-length stamens of both long» and short-styled flowers should
have similar lengths. Also, the mid-length parts should be mid way
between the long and short parts. However, this is seldom the case in
nature. For example, Ornduff (196“) found that in Oxalis suksdorfii

the mid stamens of the long-styled flowers were shorter and the mid

2?

28

stamens of the short-styled flowers were longer than expected. From
these results, he concluded that the species is tending toward distyly.
His incompatibility data also supported this idea.

Flowers of Decodon were measured using herbarium specimens from
many portions of the general range and specimens from my personal col-
lection. One flower per specimen was measured. The results are pre-
sented in Table h. It is noteworthy that the short-styled flowers have
two measurements which are significantly different from measurements of
corresponding levels of other forms. The style of the short-styled
flower is shorter than expected and the mid and long anthers are nearer
in length to each other than expected. Thus, the short-styled flower
appears to be tending toward distyly. However, no corresponding tenden-
cy occurs in the long-styled flowers. To show more dramatically these
tendencies, histograms of measurements were made for each level of each
flower form (see Figure 4). The amount of overlap between mid and long
stamens of short-styled flowers should be noted.

Measurements were taken of another sample of specimens preserved
in alcohol all of which were collected at the Mud Lake locality. These
results are given in Table 5. The average values for levels of the long
and short-styled flowers were similar. However, the standard deviations
were much greater for the short—styled flowers. The flowers designated
as mid~styled from this locality were distinguished on the basis of
style-length in the flowers measured. Most of these flowers were in-
fected with Cylindrocephalum, one of the imperfect fungi. The presence
of this organism causes a diminution of differences in lengths of dif-
ferent levels of parts of the flower. Not all flowers on a plant may

be equally affected. In fact, entirely normal flowers may be found on

 

29

.mmmonpsmhmm ca mcodpsa>eu unsucdem new: as :a mane:fl

----- Ao.~v o.ms Aa.mv o.ma Am.ov o.m em sperm
Am.sv m.a --..- Ao.~v H.Hs As.mv a.e ms ea:
Ae.ov m.e Ae.ov a.Ms ----- A~.sv a.ma em wees
mesons ensues cameos mesons oflepm osoenn assets osaem

seesaw pmonm seesaw was sosmpm msoq ca mpesoam
ho Honssz

flexed cs: scum mmmzoam Um>mmmema mo musesmmSmmm: .m mqm<9

.mmmenvcmmna ca msoapna>ep pumpsmpm sad: as s“ meme:

a
.mm.--se>es euoem .Om.--Ho>eH es: .mo.ttso>ea meoa .noonenoamHe senoaesewsn ennoa
ta--- Am.sv H.~H Ae.~V m.sa Ao.av N.m we euonm
Ao.sv m.n st--- Am.~v m.aa AH.~v a.aa an em:
Ae.ov me.m Aa.sv s.as ----- Aa.NV e.as me weoq
nemees cameos messes assess ossem osaeen gasses oseem
seesaw puocm seesaw pd: seesaw msoq 2H whoson
mo Honssz

HWGQEHOOQW ESHHBHQS EON,“ wH®30HM .HO mnvEQEmHSmde e3 am<9

30

FIGURE 4‘. Hanumts of styles and stamens of the three newer {ones
Short-styled flowers ‘ of total

not
35$
30M
25%
20!
z 51
1

   

OS
51
o

In 3 5 0 - 9 10 1112 1 1 151 171

l 21
Style Mid stamens Long stamens

Mid-styled flowers 5 of total

 

_ u —-———e ——«+ tow

» - - _. as:

- ---— — 30!

.. Hm 2;

t__ . ‘ ___‘|I“ - ___4 19‘

-7 In 1| "1|“ - , 1“

m m tun ... , ,
N'- 1" H I "1’ ‘l 12> H " ’ ,

‘ 0
mm 8 010 12131 151-1-192021 22
‘ stamens Style Lang stems

Long-styled flowers C of total

W
31
)6
2’
2G
1!
1“

fl
0

 

 

 

 

 

 

 

     

   

 

      
       

h a 7 9 1O 12 1 1 15 1 1 1
Short stamens Mid stamens Style

31

the same flowering stalk with extremely altered flowers. Neither is
the disease limited to any style type, although diseased flowers usually
appear to be mid-styled. At this locality most of the infected plants
were short-styled. True mid-styled plants were actually rare here.

Pollen grains in tristylous species may also show trimorphism.
For example, Hazen (1918) found that the pollen of short anthers of
Pontederia cordata was very small, that of the mid anthers was medium-
sised, and that of the long anthers was very large in comparison. The
difference between pollen of anthers of the same level but from flowers
of different style types was not significant. Pollen of Lyghrum (Dar-
win, 1889) and m (Ornduff, 1964) is also trimorphic, but the
differences are not as profound.

Measurements of Decodon pollen were made using dried material.
The anthers were removed from the flowers with forceps, placed in a
drop of Pohl's solution, and mashed. The debris was removed and the
slide allowed to try. A drop of Hoyer's solution was placed over the
pollen and a cover-slip added. Ten pollen grains were then measured
on each slide since the grains were relatively uniform in size. Seven
to eleven slides of each pollen type were examined. More variation was
found between different flowers even though the anthers were of the
same level and from flowers of the same style type. The results are
given in Table 6. Although the pollen from the short stamens is smal-
ler than that of the mid stamens and the pollen from the long stamens
is larger, these differences are not statistically significant.

My results are considerably different from the measurements given
by Darwin, who apparently measured relatively few grains. He did not

give separate measurements for each type of anther but instead lumped

32

om

OOH

on

OAH

ow

om

umnsmmms
undenm
no nepesz

fir

m.fi

m.a

o.H

o.a

m.H

u.H

coasea>on
8358

“.mm

0.0N

3.0m

3.0N

N.m~

e.m~

use:

H

n.0mnw.NN

m.Hmnm.mm

u.mmuw.mfi

m.mmum.mm

m.mmnm.NN

h.mmcm.«m

omcmm

o mCOHOHE EH m PCmeoH—Jmmmz

AHV navemscommmmz
Aav sameness

H ago» :02
a sewage“:

“av eszewua>
Adv mppomscosmnez
Amy sameness

Aav oeuepco
Ms snow :m2
0 sameness

2 2535
An sameness

adv mpamwsnosmmmz
Amy sewage“:

moHpHHeooH
cospomHHoo

a

mnexoam voahpmucas
mo usesmvm wcoq

mueson veahvmuphonn
mo msoswum mcoq

mumaoam veazpmuwsoa
ho newsman we:

mhmxoam umahvwcvhonm
mo newsman UH:

mnozoam ceahpmuwcoa
mo ecosepm phonm

muozoam ceahvmuufis
mo maesmpm phonm

mam» seesaw

usesspm mo mommy xfim on» sohm soHHom mo nauseousmso: .0 mqm<a

33

them together into the three levels. His results were:

Long stamens: 13/7000 inch (47 microns)

Mid stamens: 9-10/7000 inch (33-36 microns)

Short stamens: 8-9/7000 inch (29-33 microns)
These measurements are apparently uncharacteristic of the species as a
whole, but they may be representative of the plants Darwin grew from
the seeds which Asa Gray sent him. It is noteworthy that three groups
of measurements made using two specimens from Massachusetts (Cambridge
and floods Hole) were unusually small. Since the material Darwin used

may have come from one of these abnormal populations, his measurements

may be questionable.

Incompatibility §ystem

Materials, Methods, and Results. The plants used for the incon-
patibility study of Decodon were grown in #10 cans or in metal tube
which were placed out-ofzdoors between the wings of‘a greenhouse.
These plants were collected from various localities in southern Michi-
gan, mostly from Park Lake. In August, 1973. the flowers were emas-
culated, and 3-30 flowers on each stalk were pollinated with the same
type of pollen, bagged with cheesecloth, and labeled as to the type of
cross. The seeds were collected in October after the first frost. The
results are given in Table 7. In only two crosses, one legitimate and
one illegitimate, did fewer than 50% of the flowers form capsules.
These two crosses also had relatively low seed set per capsule. All
other crosses displayed a considerable degree of compatibility and pro-
duced abundant seeds.

The results in Table 7 may have been subject to various effors.

For example, sufficient pollen may not have been applied to all stigmas

.omam nman xae>fipmaem ohm mmono menu mom mesam> nonpo on» pen» cepoaneha on see

pa .nonss: man» mo manna one no .mmmn npoaoemoono one canvas weanamseh muoom emonp no women
ma mmouo mom memem no sense: ewmum>e one .uevoeaaoo one: zone seamen coomaneu measmomo emenaw
.nmezoam deamvmucas

mo mnmnpcm phone enp scum ceaaoa npdz newscaaaoa one: mweaoam noaaumnvuonm pen» memos 2m x m

.mamsmxo nob .npmcoa seamen 0» seems mnevpma aamsm use newnea cahpm 0» Homes mwoupma Heuammou

0.00 0.0m $.mm mm mm me x A
o.om m.om 0.0a om mN as x A
m.mm m.mm m.ms as em 2» x q
e.~m ~.oe s.s~ as am An x a
0.0m 0.m~ 0.0m mm on :n x :
o.ow m.em m.a~ mm mm as x z
m.Hm 0.0m 0.0N 0N mm EH x :
~.m~ 0.~e e.~s 5 am me x :
o.ooH 0.00 0.00 0a 0H 2H x m
0.30 0.om m.mm «a Ma ma x m
o.oo e.am 0.- m m as x m
N.o0 0.0m 5.5a o M“ ms w m
AssesspamoHHHv
0.ma e.e~ m.om em me :H x a
0.0m e.ma m.da an em ma x a
“.mu o.mm m.mm mm mm A: x :
m.mm 3.0m m.ms am as me x z
2...- :.... New .. on New x m
0.mm a.o: o.em we mm :m x m
amassepammqv
moasmmmo wnaosuoma noosponm oasmamo uommouo Hmsoam noosnoum someone mmoho
muesoam mo pceosem Hem meoem mo Hem mpeem wo measmamo whezoam a
Hopes: owmne>< Hopes: emmum>< mo Hepssz mo nopssz

mcospecsasoe msezdpsmeaas use mamaspsmmfi no assumes .a mamas

35

or the pollen may have been contaminated. Nevertheless, a lower seed
set would not be expected in either case. Also, contamination in the
bags (only by small, non-pollinating insects and spiders), honeybees
attacking bagged stalks (apparently the smell of the bountiful nectar
inside was too much for theml). and long-styles protruding through the
bags may have caused an increase in seed set. Nevertheless, I feel
that these sources of error can, in no way, account for the differences
between the expected results and the results actually obtained.
Discussion. Darwin (1889) was the first to demonstrate a connec-
tion between trimorphism and an incompatibility system. He used
Lythrum salicaria to make the 18 possible crosses of stamens and styles.
His results showed that crosses of organs of the same level were highly
fertile. He called such crosses legitimate. Those in which pollen
from one level pollinated a style of another level, he called illegiti-
mate crosses, and these were generally incompatible. Illegitimate
crosses with the mid-style, however, were not all incompatible. One
was completely compatible, and two were partially compatible. Ornduff
(1966) obtained very similar results with Pontederia cordata. Darwin's
data are based on a relatively small sample, and his conclusions are
based on the percent of flowers producing capsules irrespective of the
number of seeds set per capsule. However, Ornduff's data are based on
a very large sample, and, since Pontederia has a single-seeded fruit,
the conclusions are the same regardless of whether the number of seeds
per capsule or the percent of flowers producing capsules is used. The
data presented for Decodon, however, do not show this type of incom-
patibility system regardless of the method of determining seed set.

The results, along with those which follow, suggest that Decodon

36

evidently has no incompatibility system or at least no simple one.

In Table 7, self— and cross-pollinations were lumped together for
six of the illegitimate crosses. This was considered appropriate since
Ornduff (1966) found the results of such self- and cross-pollinations
to be nearly identical. However, in Decodon, there is an important
difference between the two. In this study, a few of these self; and
cross-pollinations were labeled as such, and the results of the indi-
vidual crosses are given in Table 8. Although the cross-pollinated
flowers produced a greater percentage of capsules, the self-pollinated
flowers produced significantly more seed per capsule than the similar
cross-pollination. Possibly the explanation for this lies in the
genetics of the organism. It would be interesting to know if the in-
hibition is occurring at the stigma, style, or ovary and when the stig-
mas are most receptive. Tests should have been made with single plants
comparing the effect of pollen from the same flower, pollen from dif-
ferent flowers of the same stalk, and pollen from different plants of
the same style type. It is noteworthy that in the crosses of long-
styled flowers a much smaller percentage formed capsules. This may be

the result of a slow pollen-tube growth rate.

Population Structure

Darwin suggested that an equilibrium population of a tristylous
plant would have a 18181 ratio of style lengths. But, due to extensive
vegetative reproduction, local populations of Decodon may seldom reach
an equilibrium. However, this may be achieved in the species as a
whole. From the results of the flower measurements (see Table 4), I
found in a sampling of herbarium specimens that there were 65 long-

styled plants, 57 mid-styled plants, and 68 short-styled plants. This

37

.emmwowo nmonm eno soyw mnHaHsmmh mpemm mo Mensa: one mpnemmhmeu panes: nomm

 

“.me ~.es a.sa m.sw m.mm o.ma
as new 0.3 a? 0.3 0.3
e.ms e.m~ m.o~ N.am s.o~ n.mm
an
an
an
on
an
mm
«a mm
we Nm
an mm mm
m: Hm on
e: me on em
mm N: as on an
mm o m on an
mN o m as aw
em 0 a as aw
a o a o: es en
0 o 0 am as em
0 o H o m on
o o o o o o
60mmOHo UGMHmw Ummmonno UGMHmm Ummmonno UwMHmm
an x a 2m x z ma x m

 

 

 

H

 

H

measmmmo wanmom

 

0.05 0.00 mnemon mo uncouem

. . coswom eHsmamo you neosnowm
0 Hm 0 mm mneem mo Mensa: emmwe><

. . nemmono Hosoam Hem couscous
3 0H 0 0H women no Mensa: mwmwo><

0N

mm

mm

mm mm

as an

NH 0m

0 0

0 0

commono coMHom

me x m

mnoHpmsHHHom oumEHvaeHHH someone was neMHem mo nomHsmasoo .0 mqm<e

38

ratio is not significantly different from Darwin's prediction. Although
there is statistically no difference between these, it is my subjective
observation that the short-styled plants are most common and the mid-
styled plants least common in Michigan. Interestingly, short-styled
herbarium specimens are also greatest in number, and mid-styled ones

are the least frequent. One point which may affect the apparent com-
position of a Decodon population is that the short-styled plants are

the first to flower and the long-styled plants flower later and may
still be blooming after several frosts. Thus, anyone visiting a popu-
lation may get a biased view of the frequency of style lengths unless

he visits it several times throughout the flowering season.

SUMMARY AND CONCLUSIONS

This research is regarded as a preliminary study and serves to
ellucidate several problems in Decodon. The following paragraphs sum-
marize the results of each section of the study and point out areas for
further investigation.

The general morph010gy of Decodon is discussed with special refer-
ence to the number of leaves per node and the number of floral parts
per flower. Nearly 90 percent of the Decodon stalks were found to have
three leaves per node; the remainder had one, two, or four leaves per
node. Nearly 85 percent of the flowers were found to be 5-merous; the
remainder were three-, four-, or six-merous.

The nomenclature and taxonomic relationships of Decodon are dis-
cussed. Intraspecific variation is dealt with in respect to the pubes-
cence character. It appears that discontinuity of morphological forms
and geographical segregation may not be significant enough to warrant
taxonomic recognition of two varieties. A larger number of Specimens
should be studied to better determine the distribution of pubescence
types. A study of the physiology of glabrous and pubescent plants
might provide an explanation as to geographical segregation of pubes-
cence types on the extremes of the ranges.

The geological history of Decodon and its close relatives is des-
cribed. The modern distribution of Decodon reflects events of glacia—

tion, migration patterns of waterfowl, distribution of suitable

39

habitats, and the climate.

Vegetative and sexual reproduction are compared. The flora of the
Decodon mat is categorized in three groups: 1) plants living in the
shallow water between Decodon plants, 2) those developing early, and
3) those growing at the bases of the stalks eventually forming hummocks.
The plants of the Decodon mats of nine localities are listed and the
localities described. The exact conditions under which Decodon plants
become established are not known.

Reproductive biology is discussed in detail. The tristylous flow-
ers were found to have lengths of parts of each level as expected,
except for the short-styled flowers in which the styles and long sta-
mens were shorter than expected. Pollen grains did not show a signifi-
cant trimorphism correlated with stamen length. Since legitimate
cross-pollinations were no more productive than illegitimate crosses in
most cases, Decodon does not have the typical incompatibility system
complementing the trimorphism. Self-pollinations produced more seeds
but fewer capsules than cross-pollinations of parts of the same lengths.
The structure of Decodon populations is not significantly different
from the 1:1:1 ratio predicted by Darwin for trimorphic populations.
More similar studies of the floral morphology and incompatibility of
Decodon should be done using plants from other parts of the range in
order to determine if a typical tristylous incompatibility system is

functional elsewhere.

BIBLIOGRAPHY

Barlow, N. 1923. Inheritance of the three forms in trimorphic species.
J. Genet. 13: 133-146.

Darwin, C. 1889. The different forms of flowers on plants of the same
species. Appleton and Co., New York.

East, E. M. 1927. The inheritance of heterostyly in Lythrum salicaria.
Genetics 12: 393-414.

1932. Further observations in Lythrum salicaria. Genetics 17:
327-3340

Elliott, S. 1824. A sketch of the botany of South-Carolina and Geor-
gia. J. R. Schenck, Charleston, S. C.

Fernald, M. L. 1917. Decodon verticillatus var. laevigatus. Rhodora
19: 154-155.

Fisher, R. A. and K. Mather. 1943. The inheritance of style length
in lythrum salicaria. Ann. Eugenics 12: 1—23.

Gmelin, J. F. 1791. Systema naturae 2: 656, 677.

Graham, A. and S. A. Graham. 1971. The geological history of the
Lythraceae. Brittonia 23: 335-346.

Hazen, T. E. 1918. The trimorphism and insect visitors of Pontederia.
Mem. Torrey Bot. Club 17: 459-484.

Koehne, E. 1874. Ueber inflorescenz und trimorphismus von Decodon
(Nesaea) verticillatus einer nordamerikanischen Lythracee. Verh.
Bot. Ver. Prov. Brandenburg 16: 42-43.

Lincoln, F. C. 1950. Migration of birds. U. S. Dept. of the Interior,
Fish and Wildlife Service, Circular 16.

Nikitin, P. A. 1929. The systematic position of the fossil genus
Diclidocarya E. M. Reid (with a note by Mrs. E. M. Reid). J. Bot.
678 33‘380

Ornduff, R. 1964. The breeding system of Oxalis suksdorfii. Amer. J.

 

41

42

1966. The breeding system of Pontederia cordata L. Bull. Torrey
Bot. Club 93: 407-416.

 

Penfound, N. T. 1952. An outline for ecological life histories of
herbaceous vascular hydrophytes. Ecology 33: 123-128.

Ridley, H. N. 1930. The dispersal of plants throughout the world.
L. Reeve and Co., Ashford, Great Britain.

Schrenk, J. 1889. On the floating tissue of Nesaea verticillata (L.)
HBK. Bull. Torrey Bot. Club 16: 315-323.

 

Stafleu, F. A. ethal. 1972. International Code of Botanical Nomencla-
ture. Regnum Vegetabile 82: 1-426.

Tessene, M. F. 1969. Systematic and ecological studies on Plantago
cordata. MiChe BOte 8: 72-104.

Torrey, J. and A. Gray. 1840. Flora of North America. (Hefner Pub-
lishing Co., New York, 1969).

Ugent, D. 1962. Preliminary reports on the flora of Wisconsin. No.
47. The orders Thymelaeales, Myrtales, and Cactales. Trans. Wis-
consin Acad. Sci. 51: 83-134.

Walter, T. 1788. Flora caroliniana. London. (Facsimile edition,
Arnold Arboretum, Jamaica Plains, 1946).

Wilson, E. H. 1925. A neglected aquatic. Garden (London) 89: 673.