THE ROOT-PARASITISM OF CASTILLEJA COCCINEA

by
‘William.McLagan Malcolm II

AN ABSTRACT
Submitted to the College of Science and Arts,
Michigan State University of Agriculture and
Applied Science, in partial fulfillment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY
Department of Botany and Plant Pathology

1962

Ft

Castilleaa coccinea (L.) Spreng., commonly called scarlet Indian
paintébrush (Scrophulariaceae), is a species of flowering plants which
extract materials from.the vascular streams of the roots of nearby vas-
cular plants. The attachment to a host plant is made by so-called
haustoria, minute organs on the parasite roots. In each haustorium a
vascular trace from the parasitexxnt penetrates to the xylem of the
host root. Eosin Y, fructose, and sulfate and phosphate ions are known
to cross the haustorial connection, and only in the direction of the
parasite. Presumably most or all of the substances in the vascular
stream of the host can cross into the parasite.

Although the parasite is chlorophyllous and fully photosynthetic,
it does not grow beyond the seedling stage unless it successfully pene-
trates foreign roots, a fact readily demonstrated by growing seedlings
in pots without hosts. This host-requirement is not offset by artifi-
cial feeding of mineral nutrients or the more comon vitamins, phyto-
hormones, or respiratory substrates. Once host-contact is made, the
foliage leaves of the seedling grow rapidly, producing in some four
months a flat, leafy rosette about four centimeters in diameter. How-
ever, seedlings parasitic on even the same host plant have widelyb
varying rates of growth, perhaps a function of the number of their
active haustoria. The rosettes bolt with or without cold treatment,
and form.the showy flowering shoots which give the parasite its common
name of paintAbrush. The parasitic attachment is not necessary for

bolting mature rosettes pulled free of their hosts, washed free

 

of foreign roots and debris, and potted separately will flower normally.

Nonetheless, under natural conditions bolting rosettes enjoy extensive

host contact.

The parasite attaches to a wide variety of host plants, including
plants of species which do not occur in its range. In the greenhouse,
some of these hosts support more vigorous growth of the parasite than
others do. The hosts in a central Michigan community were ranked ac-
cording to their ability to support the parasite. However, the distribu-
tion of the parasite in that community could not be related to the host
ranking. That is, the parasite did not tend to occur more frequently
around the hosts which support it best in the greenhouse.

Seeds of the parasite wifl.germinate at the time of their dissem-
ination, and they retain viability for at least two years when stored
dry. The seeds will germinate at 4°C, but require light and moisture.
Seeds stratified four weeks, then kept in a diurnal temperature range
of 15°-38°C germinate at a level of 87%. The seedling hypocotyls are
clothed abundantly with root hairs. The direction of root growth is
not influenced by the presence of host roots. Induction of haustoria
is not solely a result of physical contact with objects or with living
tissue. Under natural conditions, the hosts suffer no obvious damage,
but greenhouse-grown hosts supporting many parasites suffer twice the

mortality of unattacked hosts.

 

 

 

 

 

 

 

 

 

 

 

’)

THE ROOT-PARASITISM OF CASTILLEJA COCCINEA

by
William McLagan Malcolm II

A DISSERTATION
Submitted to the College of Science and Arts,
Michigan State University of Agriculture and
Applied Sciences, in partial fulfillment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY
Department of Botany and Plant Pathology

1962

page

‘ 15
25
as

69
79

93

10A
133
135

INTRODUCTION
METHODS
RESULTS
THE PARASITE
ontogenj
germination
hypocotyl growth
epicotyl growth

physiology

 

THE HOSTS
THE HOST-PARASITE CONNECTION
THE HOST-PARASITE INTERACTION
the host-requirement
haustorial uptake
damage to the host
parasite micro-distribution
DISCUSSION AND LITERATURE REVIEW
CONCLUSIONS
LITERATURE CITED

p l a t e l
Castilleja coccinea flowering shoot (12X). The flowers and bracts
of the inflorescence are tipped with scarlet, orange, or yellow,

giving the plant its common name, paint-brush.

 

 

 

THE ROOT-PARASITISM OF CASTILLEJA COCCINEA

INTRODUCTION

Castilleja was named by Mutis in 1871 in honor of the Spanish bot-
anist Domingo Castillejo (Fernald, 1950). The name usually is pronounced
kas til LEE ya or kas til LAY a, but sometimes the l'is sounded as an
English 1 rather than a y.

Castilleja is placedin.the subfamily Rhinanthoideae of the figworts
(Scrophulariaceae), a large and cosmopolitan fanily of flowering plants.
The 250 or so species of the genus are mostly North American in distribu-
tion, and are particular1y<xmmmmlin.the western United States (Gleason,
1958). Castilleja is of little economic importance, but has attracted
the attention of taxonomists for sane years because of its complex species
problems.

Castilleja coccinea (L.) Spreng.* (plate 1, frontispiece), commonly
called scarlet Indian paint-brush, ranges from.southern New England and
Canada south to the northern borders of the Gulf states and west to
Oklahoma (Pennell, 1935) (plate 2, page'7). The species in Michigan is
scattered but common close to the shores of both peninsulas and in a few
inland sites (plate 3, Page 9). Plants of the species attack roots of
nearby vascular plants and extract from them materials for'their growth.
The attack on the host plants is made by minute organs on the parasite
roots, so-called haustoria, which adhere to the host roots and lyse or

force access to the vascular elements (Solms-Laubach, 1867/1868) (plates

* the nomenclature in the dissertation is based on Fernald (1950)

 

       

w
.. N

   

     

 

no‘

e‘vi ,Ifenueq

(l

88 J

5Lre Isdduml

r
u
a.

’1

(.Le

 

tasId'A blmo

 

plate 2
Approximate distribution of Castilleja coccinea (based on Pennell, 1935).
(The outline map of North America is based on a Lambert azimuthal equal-

area projection published by Rand McNally in Standard World Atlas, 1951,)

 

 

v
r
\_.
o r
\-'
e
.4} -L.

o .
‘.\.
.‘w
,rl‘

U
2 "
pl
I.
. t,
‘I
”Hi".

H.

~ ";
r
Y
t
..- '1'
l ,

plate 3

Michigan collection sites, by county, of specimens of Castilleja
coccinea entered in the herbaria of Michigan State University and

the University of Michigan (state map from publication A9, Depart-
ment of Conservation, Geological Survey Division, maps of the surface

formations of the northern and southern peninsulas of Michigan).

 

10

A and 5, pagesljiandlt). Although the parasites are chlorophyllous and
fully photosynthetic, to mature they must successfully penetrate foreign

roots. This obligate root-parasitism of Castilleja coccinea is the topic

 

of the dissertation. The approach to the problem is ecological and phys-
iological. Some of the questions the study attempts to answer are 'What
sort of connection does the parasite make with its hosts?', 'What kinds

of compounds cross the haustoria?', 'Where in the life history of the
parasite is the host-parasite interaction important?', 'How wide a host-
range does the parasite have, and do various hosts support the parasite
equally we117', 'Does the host-parasite interaction determine the distribu-
tion and abundance of the parasite in a community?', and 'What damage does

the host suffer?'.

\‘a

,-

ll

p l a t e A
Haustoria of Castilleja coccinea attached to host roots. The drawings

were made from specimens ready for embedding and eventual sectioning (30X).

 

O

 

F

t‘,!'

\o

(I
1 ,K)

Q

'r

.

c2”

‘1'
,0
*-,
'l".
1‘,

, w

 

 

l3

 

 

 

 

I;
3"» '
‘ . x
n
‘ nf‘ 'ff '
_| . . 'Jht‘
.. .._,._,-_, . o. .
' l
A I, ,. u . J
' l
I. t: .' ‘4}‘Il
_ “1.... no . Fix
v t. ‘ -
a. .\ ~~\ . ' " - ‘
-...,.w Muir w r- ‘
fl ,1 .\ ' ‘ O‘-
l) 393 ‘
.wr._v -‘ -"-‘
, A 7
I l
.
‘1
v

 

p l a t e 5
Photomicrograph (270K) of a section through a haustorium of Melampyrum

lineare parasitic on Pinus banksiana, jack-pine. Both the host root

 

and the parasite root which produced the haustorium are in cross-section,
but the vascular trace connecting the two is in longitudinal section.
Melampyrum lineare is a green root-parasite closely related to Castilleja

coccinea, and is common in jack-pine vegetation in Michigan.

. . ..
r I
a. .
. \U . U
V .aflJ-re\
. v:

 

15

M E T H 0 D S
p a r a s i t e f i e 1 d g r O‘W t h d a t a

A heavy pOpulation of Castilleja coccinea was selected early in 1960
for detailed field studies. The population lies south of Houghton Lake
along both sides of old route 0527 3.0 miles south of its junction with
M55, in the NEk of section 33, T22N, RAW, Roscommon County, Michigan.

In April of 1961, 100 rosettes and seedlings of the Houghton pOpula-
tion were marked with wooden or polystyrene pot labels. Their growth and
maturation were recorded at intervals of about a month from April to Sep-
tember. The records included rosette diameter, height of flowering shoot,
and degree of fruit ripeness.

p o t c u 1 t u r e

The studies of Castilleja coccinea nutrition, epicotyl growth, host-

range, and host-parasite interaction required artificial culture of the

parasite and its hosts under laboratory conditions. The culture technique

 

used most often for these studies was simple pot-culture given host
plants to attack, Castilleja coccinea grows readily in flower pots in a
greenhouse.

Preparation of the parasite seed and the hosts for pot-culture is
time-consuming but demands no unusual equipment. Seed is collected by
gathering ripe capsules in a polyethylene bag. The capsules later are
broken Open with forceps over a bowl covered with a piece of window screen.
The screen excludes plant debris but allows the seeds to fall into the bowl.
Potential hosts are dug up and potted in the field, brought into a green-
house, and for about a week allowed to regenerate their damaged roots.

A 3fi-inch pot is plugged at the bottom with a small square of blotter paper

and nearly filled with sand. The sand is tamped lightly into the pot to

16

form a smooth surface. A host is washed free of soil and debris, and its
roots are spread out on the surface of the sand. The roots then are cov-
ered with another quarter-inch of sand, and the pot is watered generously.
Several hundred Castilleja coccinea seeds then are sprinkled evenly over
the surface. The pot is placed in a greenhouse tray in half-inch—deep
water. Germination begins in about four days, and the seedlings show signs
of host-contact within two weeks.

For observation of the haustoria of pot-cultured parasites, the host
and parasite simply are dumped from the pot and their tangled roots care-
fully washed free of sand in a bucket of water. The haustoria are seen
best under a dissecting microscope because of their small size. If kept
moist and handled carefully, a parasite and its host can be replanted suc-
cessfully after being examined for haustoria.

g e r m.i n a t i o n

light reguiremen . (l) 500 Castilleja coccinea seeds were sprinkled into
each of three 3%-inch pots covered with Petri dish bottoms wrapped in
heavy aluminum foil. The foil of one Petri dish was left intact, but both
of the others had a hole which allowed light to enter. One of the holes
'was 7 mm in diameter and the other was AS mm. Both holes were cut close
to the edge of the Petri dishes to accentuate any phototropic response of
the seedlings. The set-ups are shown in plates 6-8 (pages 17-22 ).

(2) 100 Castilleja coccinea seeds were sprinkled into eight sand-filled
3g-inch pots with lids made from aluminum foil-covered plastic Petri dish
tops. The pots were placed in standing water in a greenhouse and the per
cent germination of the seeds recorded at two and four weeks. For controls,

eight pots without lids were set up. Results are in table 22(page 54).

l7

 

 

 

 

\
_
,
I
z
i
.
I. —
.
L,
.
i
k l...
. I.
.i
U . ‘
I
OI
.- r L
vi a
‘H .h
‘.4
. 4v»
.k c
l
. .
~
. x
, ‘ J.

p l a t e 6

Demonstration of light requirement in the germination of the seeds of
Castilleja coccinea. Five hundred seeds were sprinkled on the surface
of the sand in the pot, and a foil-covered Petri dish bottom was placed
over the pot to exclude all light. Only one seed had germinated after

four weeks.

l8

 

l

O

p 1 a t e 7

Effect of low light on the germination of seeds of Castilleja coccinea.
A small hole cut in the foil of the pot-lid afforded a low level of
light. In all, 15 out of 500 seeds germinated after four weeks, a
germination level of 3%. All of the seedlings were markedly etiolated

and showed a strong phototropic growth respcnse.

20

 

21

p 1 a t e 8
Demonstration of normal germination level under a strong-light regime.
The hole in the pot-lid was offset to accentuate the phototropic

response of the seedlings. The germination level at four weeks was 65%.

22

 

23

(3) Five loo-seed lots were kept at 4°C on moist blotter paper in foil-
covered Petri dishes. The per cent germination was recorded at two, four,
and six weeks. Five similar dishes without the foil were set up as con-
trols and illuminated at 4°C with a [+0-th incandescent bulb one foot
from the dishes. Results are tabulated on page 5h (table 2 ).
stratification. 1000 Castilleja coccinea seeds were scattered on a moist
disc of blotter paper in each of three Petri dishes. They then were placed
for one month in a refrigerator held at A°C. With the seeds still on it,
each blotter disc was out in half to make two lots of seeds. One lot was
retained in the cold, but the other was placed in a cool greenhouse (15°-
25°C). Plate $9 is a photograph of both lots of seeds from one dish two
weeks after they were separated.
seed viability. Castilleja coccinea seeds of 0, l, and 2 years age were
sown on the surface of moist sand in 3%-inch pots in‘a greenhouse. The
per cent germination was recorded at two, four, and eight weeks. 100
seeds of each age were tested in each pot, and three replicates were run.
As well, five 100-seed lots were tested of two-year-old'seedstock which
had.been stored at room temperature (20°C). The results of the trials
are in table 2 (page 5h).
h y'p o c o t y 1 g r O‘W t h a n d h a u s t o r i u m o n t o g e n y
Plants of both Melgmpygum lineare and Castilleja coccinea will para-
sitize host plants sandwiched between glass and blotter paper. Such a
set-up is useful for studying the ontogeny of haustoria and the growth of
the parasite hypocotyl. In plate 10 (page 27) is an exploded view of one
of the sandwiches. All the materials in it are marketed in retail stores,

and.its construction is simple. A host which has been potted in the field

21+

p 1 a t e 9-

Demonstration of the effect of stratification on the germination level

of seeds of Castilleja coccinea. A thousand seeds were placed on the
blotter discin the upper Petri dish. They were kept at LOC for a month,
then divided into two lots. The upper lot was retained in the cold but
the lower was warmed to room temperature. Germination of the warmed seeds
reached 87%, about a quarter again as high as the level of germination
typical of seeds which have not been stratified. Germination does occur
even at 4°C if light and moisture are sufficient, but the rate of
germination is low-———————about six weeks are required for the level of

germination to reach 60?.

He's/QM-“ “’61

 

26

. .....‘t.

v

. t
I
a
u
a. ..
.... .. .
.. r
.n.
r
(I.
I .
./ a
I w I
. -.. v
I. -| A
III”;

f
.rce‘wr-‘(fi
J\' 1

t

r. w.

P I“
.-

1 a

r ‘r“ f‘.

x.

 

.s’

e..
,-‘,

. _
lo. .
rr
.. ..
r I. l
v
I l
A7
I I,
..\ a.
v II
.5. a;
r M _
Q l\
n.
C;
: ‘
1L
0»
.
. l
l... r.
,1
l J
or a
_
c, r .
.
a,
.0 0
‘1 If u
A... ..
.41»,
Vv‘. \.
r... a
o\ .I..
‘1 1
L. 4 u
. v t
‘l, «a
3. . w L
. . 7‘
v. r.IL
.
3 a;
E
F. a».
.r..

\n
a . .VK
’ x
. V I

IcIJ
er ,
\ no . ..
e . c ’7
.1 (a
c .
or...
\.J
. . n
r7.”
.2.
.P

 

,-
t
I
v. ~h
.
4.
l 1 .o l l
a
o
h u
.
1 1‘?!
cue...‘ .
" |.
x
.t a
.9‘.
‘10:}...
.‘. I111 v

pi“

\ arr‘

 

p 1 a t e 1 0

Exploded view of glass-blotter-glass 'sandwich'. The layers of the
sandwich are, in order, glass, blotter, sand (Optional), blotter, and
glass. The assembled layers are held together on three sides by

aluminum channeling. If no sand is used in the 'stuffing' of the sandwich,
aluminum channeling need be placed only on two sides. The sandwich is
used in the study of haustorium formation by root-parasites. The root
systems of both the root-parasites and their hosts can be observed

easily with a dissecting microscope, and the locations of haustoria

marked with a grease pencil right on the glass of the sandwich.

 

 

 

 

 

 

 

 

 

28

and allowed to regenerate its roots for a week is washed free of soil and
debris, and its roots are flattened out on a four-inch square of blotter
paper. Several of the roots are positioned along the t0p inch of the
blotter. A four-inch square of double-weight glass then is placed over
the roots to flatten them and hold them in position against the blotter.

A second piece of glass is placed behind the blotter to form a sandwich
‘with 'bread' of glass and 'stuffing' of the blotter and host. The sand-
wich is held together along its aides and bottom by half-inchdwide
aluminum U-channeling (plate 11, page 30). The assembled sandwich must
be about as thick as the interior width of the channeling. This width

is made up by placing sand or more squares of blotter paper between the
two pieces of glass. Large hosts are best accommodated in larger sand-
wiches, 6 or 8 inches on a side. The ccmpleted sandwich is placed upright
in a tray of water in a greenhouse. After a few days, when the host roots
are 'trained' to stay in positflxh.the front piece of glass is removed and
seeds of Castilleja coccinea sown as close as possible to the tOp of the
blotter. The parasite seedlings must emerge from the top of the sandwich,
which they can not do if they are placed far down on the blotter or if the
sandwich is too tightly pressed together. After the seedlings are estab-
lished, any disassembling of the sandwich must be done carefully, since
the haustoria are minute and fragile. Also, debris-eating maggots often
invade the sandwiches, and, in times of food stress, destroy the parasite
seedlings. Such pests can be eliminated by 25% Malathion or some other
insecticide.

e p i c o t y l g r O‘W t h a n d h o s t - r a n g e

The growth of individual rosettes was followed for several months by

.-

I.

.élw inlwoe 103 Ybseq

       

bns boIdm

rat!
4 ‘ I

29

II ”BI”:

old“asla A

“.0 to 35091

IMO. .113 .831 .‘l
t —

 

p l a t e 1 1
A glass-blotter—glass sandwich assembled and ready for sowing with

seeds of Castilleja coccinea. The host shown is a rosette of Rudbeckia

hirta, the common black-eyed Susan.

 

 

 

31

labelling selected rosettes grown in pot-culture with various hosts. The
rosettes were labelled with paper discs marked with a number and glued to
the heads of insect pins pushed into the sand of the host-pot (plate 12,
page 33). In all, 350 rosettes were studied in this way. Their growth
was measured as centimeters of rosette diameter and number of foliage
leaves. Later, height of the flowering shoot was recorded in centimeters.
p hey si 0 1 o g y

starch test for photosygthesis. Leaves of Castilleja coccinea to be tested
for their starch content were killed in boiling water and leached of their
chlorophyll in boiling ethanol. They were then immersed in boiling water
again, and finally soaked in iodine-potassium iodide solution for five
minutes. The test-solution was made up of 15 g of potassium iodide in a
liter of distilled water, to which 5 g of crystalline iodine was added.

A positive reaction was the formation of an intense blue complex of starch
and iodine.

starch sygthesis assay. All the leaves used in the tests for starch syn-
thesis were excised and floated on distilled water in darkness for 24 hours
to deplete their food reserves. At the end of this quarantine period, the
leaves were divided into four lots of ten leaves each. The first lot was
tested for starch immediately. The second lot was placed in sunlight and
tested for starch after six hours. The third was placed in a small beaker
inside a tightly-closed jar containing a half-inch of 10% potassium hydrox-
ide which removed the carbon dioxide from the atmosphere inside the jar.
The jar was kept in total darkness for six hours, then exposed for another
six hours to sunlight and the leaves tested for.starch. The leaves of the

fourth lot were smeared with petroleum jelly, refloated on distilled water

32

 

p l a t e l 2

A rosette of Rudbeckia hirta, the common black-eyed Susan, under

 

attack by many month-old rosettes of Castillejg coccinea. Some of
the parasite rosettes are labelled with paper discs glued to the
heads of insect pins, a technique used in following the growth of

individual rosettes over a long period of time (1.6X).

 

33

1'7
( .

..

 

34

in a Petri dish, left in sunlight for six hours, and then tested for
their starch content.

carbon fixation assay. One set of experiments was designed to test the
ability of host-free Castilleja coccinea rosettes to fix carbon dioxide
when illuminated. Castilleja coccinea rosettes of various ages were
pulled free of their hosts, separated into four groups, and placed on
moist blotter in four plastic Petri dishes. They were kept in darkness
for twelve hours. All four groups than were given 01902 for thirty minutes,
but two groups were lighted and two retained in the dark. They then were
weighed, killed, macerated, each pipetted as three aliquots into planchets,
dried, and their activity estimated with a 1000-count.

Also an estimation was made of the level of photosynthesis of rosettes
still parasitic on host plants. The several rosettes selected were para-
sitic on Lactuca canadensis in two pots. The bottom from a clear plastic
Petri dish was placed over each pot, isolating the rosettes and their host
inside a small volume of air. In the pot with the plants was a vial of
NaHClAOB. An excess of acid was added to the bicarbonate with a medicine
dropper stuck through a hole in the Petri dish bottom. Cellophane tape
was placed quickly over the hole after the acid was added, and the whole
assembly was placed in a hood under strong illumination behind a water
heat-barrier. After 30 minutes, the host and the parasites were weighed
in an automatic balance, killed in boiling distilled water, macerated in
95% ethanol, and their activity sampled with dried aliquots in planchets.

The set-up is shown in plate 13 (page 36).

 

p l a t e l 3
Diagram of experimental set-up for assay of photosynthesis by Castilleja

coccinea and its host, Lactuca canadensis.

 

 

acid drOpner

host plant

————— plastic lid

. . v ,‘ / ,
__ vial Oi, Lambs,

parasite rosette

 

 

n1 .
;5-1nch pot

 

37

nutrition supplements. The two techniques used for growing Castilleja
coccinea in nutrition supplement experiments were aseptic agar culture
and sand culture.

Four types of media were used in the agar culture-—————-—-Bacto—Agar
2%, Bacto-Agar 2% made up with Shive's nutrient solution and a micro-
nutrient supplement, Bacto tryptone glucose extract agar 2%, and Bacto
tryptone glucose extract agar 2% made up with Shive's solution and a
micro-nutrient supplement. The agar was poured into inch-byhthree-inch
capless vials, which were stOppered with non-absorbent cotton protected
from.moisture by aluminum foil. The bottom inch of each vial was dipped
into patent black masking ink and wrapped with aluminum foil to exclude
light from the root zones of the seedlings. The vials and agar then were
sterilized for fifteen minutes at fifteen pounds pressure. Castilleja
coccinea seeds were surface-sterilized for thirty minutes in a solution
of calcium hypochlorite (50 g / 700 ml distilled water), and placed with
flamed forceps Onto sterile agar in Petri dishes. After four days, those
seeds which showed no contamination were transferred to the vials.

In sand culture, coarse washed sand was spread into either Petri
dishes or small capped vials, moistened to capacity with the proper nu-
trition supplement, seeded with about fifty Castilleja coccinea seeds, and
placed in a cool greenhouse. Contamination was rarely a problem, simply
because the supplements were not respiratory substrates. In general, the
Petri dishes were more satisfactory for this sort of study, because many
seedlings could be maintained, and lighting them was simplified. However,
in sterile culture, Petri dishes were found to be useless, since contamin-

ants invariably got in, apparently through the film of water which condensed

38

on the top of the dish during the daytime. In the case of sand culture
with distilled water, the sand used was acid-leached marine white sand.
h o s t - r a n g e

Castilleja coccinea was grown from seed in pots with single hosts,
and its success noted for each of the hosts. All the hosts were collect-
ed in the Houghton Lake study area, transported in 3%-inch pots to East
Lansing, and placed in standing water in a greenhouse at Michigan State
University. After a week, the host plants which did not survive potting
were discarded, and the remaining hosts repotted in strained, washed sand
in new'3é-inch pots plugged with blotter paper. Ten pots of each host
were made up in this way, except in the cases of Lactuca canadensis,
Fragaria vir 'niana, and Chrysanthemum leucanthemum, of which respective-
ly 50, 20, and 20 pots were made up. In each of the host pots were sown
over 500 seeds.
h a u s t o r i u m a n a t o m y a n d o n t o g e n y

The anatomy and ontogeny of haustoria, either field-collected or har-
vested from,glasseblotter sandwiches, were studied by sectioning. Each
haustorium was killed and fixed still attached to a portion of the host
root it had attacked. The fixing solution used was FAA, and the specimens
were readied for embedding by immersion in the standard tertiary butanol
histologic solutions. The specimens were sectioned in rubber paraffin,
stained with safranin and fast green, and mounted serially in balsam.

Haustoria of known age were harvested by growing the parasite and its
host in glasséblotter sandwiches (see page 27), locating the incipient
haustoria with a dissecting microsc0pe, noting the location and date right

on the glass of the sandwich with a grease pencil, and later disassembling

39

the sandwich to collect haustoria of various ages.
h o s t - r e q u i r e m e n t

The host-requirement in epicotfl.growth of Castilleja coccinea was
established in two ways-———————-by growing hundreds of seedlings in pots
with and without host plants, and by divorcing already-establiShed ro-
settes from their hosts.

In the first technique, all the pots used in the experiments were
new, and the sand in them was strained and washed to exclude all plant
debris.

In the second technique, both greenhouse-grown and field-collected
rosettes were used. In both cases, the rosettes were teased away from
their hosts gently, and repotted alone in strained, washed sand in new
pots.

h a u s t o r i a l u p t a k e

Transfer of eosin Y and labelled phosphate, sulfate, and fructose
across the haustoria was done with potted specimens grown from seed in
the greenhouse. The hosts were either Fragaria virginiana or Antennaria
neglecta. In all the transfer experiments, the substance to be tested
was introduced into the host or parasite simply by dipping a cut leaf
blade or petiole into a solution of the substance for about a half hour.
Transfer was detected by a color change in the case of eosin Y, and by
beta emission in the case of the labelled compounds. In transfer of the
labelled compounds, because the source of nuclide was in the same pot
with the plant being monitored, a particle shield was set up around the
plant and the Geiger-Muller tube. The experimental arrangement is dia-

grammed on page L1, (plate It).

[+0

p l a t e l A
Drawing of experimental set-up for the transfer of labelled compounds

between Castilleja coccinea and its hosts.

"T7

a“...

"-..~

Jr
H

s--.’_‘ 4"
.g g W:’:\[~0, >

Is g,‘

a

 

 

 

 

 

 

 

 

 

 

to sealer

monitor

monitor support

ring stand

host plant

beta shield

parasite rosette
host petiole
nuclide reservoir

#2

p a r a s i t e m i c r o - d i s t r i b u t i o n

The correlation between the various hosts' ability to support Qééfill?
lgja coccinea and the field distribution of Castilleja coccinea was tested
with a statistical analysis of field sampling data. It was assumed that
if the parasite enjoys better growth when parasitic on particular hosts,
then in the field it would tend to occur more frequently close to those
hosts than close to any others. Conversely, one could expect to find the
density of those preferred hosts higher close to the paraSites than away
from them. This assumption was the base of the sampling procedure. In a
sense, the assumption is that the act of locating the parasite in the field
is more often than not an act of locating the preferred hosts, since the
parasite survives best when parasitic on those hosts.

A 6- by 3-meter community of Castilleaa coccinea and its hosts in the
Houghton Lake study area was divided into fifty quadrats 60 cm on a side.
Each host species in the community was placed in one of three categories
of root-system-diameter on the basis of excavations of the root systems of
representative plants of each species. These diameters were 30, AO, and
50 cm. A sampling device made of concentric wire circles with these three
diameters then was used in the field to estimate how many plants of each
host species logiéally could be suspected of contributing to the support
of a given parasite rosette. If a quadrat contained a rosette of Castilleja
coccinea, the concentric circles were placed so that the parasite was at the
center. In this way the density of particular hosts around the parasite was
calculated. The density of the same hosts occurring some distance away from
any parasites also was calculated. For this task, the concentric circles

were placed in the middle of quadrats in which no Castilleja coccinea ro-

1+3

sette occurred. The statistical analysis involved testing for a signifi-
cant difference between the density of each host near a rosette and its
density away from a rosette. All of the fifty quadrats were used. 22 had
no Castille a coccinea rosettes, while 28 did. The test used was based

on the t distribution. A sample calculation is given on pageldi.

Ah

STATISTICAL COMPARISON OF DENSITY OF LOBELIA SPICATA CLOSE TO AND DISTANT

FROM CASTILLEJA COCCINEA.

Number of Lobelia plants present in each of 28 quadrats in which Castilleja

 

12, 5: 3, 3a 6: 4: 5: 7: 8a 3: 2: A: 3:
3, l, l, 6, 3, 6, O, O, O, O, O, O, O, O, 0. Number of Lobelia plants

coccinea also was present

present in each of 22 quadrats in which Castilleja coccinea was absent

 

1, Ll" 8, l, l! 5, 3, l, 7’ 1’ 6’ 5, 1+, 1’ 0’ 0, 0’ O, O! 0’ 0, 0'

X1 Castilleja present X2 Castilleja absent
2x = 85 2x = 68
2x2 = 507 2x2 = 21.6
(2102 / :1: 258.0357 (2102 / n= 104.7272
2 = 3.0357 1? = 2.1818

s: = (507 - 258.0357 + 22.6 - 104.7272) / (1.8) = 8.1299

332.21 _ 22 = 8.1299 (1/28 + 1/22) = 0.6599

ail _ x2 = V 0.6599 = 0.8123

t = (3.0357 - 2.1818) / (0.8123) = 1.0512

t97.5(d.f.48) = 2.011 (Hald, 1952)

2.011 :>' 1.0512, therefore the density of Lobelia close to Castilleja is

not significantly different from what it is distant from Castillej .

AS

R E S U L T S
T H E P A R A S I T E
o n t o g e n y

An ovary of a Castilleja coccinea plant produces up to 300 seeds,
each less than 1 mm in length (plates 15 and 16, pages [.7 and 49). The
fruits of the adult plant are two-loculed capsules, from which the seeds
are shaken by wind or passing animals. The seeds are light, averaging
0.056 mg in weight, and they scatter readily.

Castilleja coccinea in the Heughton Lake area typically is biennial
in its development. In the laboratory the seeds will germinate at the
time of capsule dehiscence, but in the field the bulk of germination
occurs in the spring. Maturation of the rosettes umually is complete
by the end of the summer. The rosettes bolt the following spring, often
very early. In greenhouse culture, they do not require a period of cold
to bolt, which suggests that in the field they may behave occasionally
as annuals and bloom in their first season of growth. Tabulated on pages
50-52 (table 1 ) are the results of a field study of 100 rosettes and
seedlings in the Houghton Lake area. The data include rosette diameter,
height of flowering shoot, and degree of fruit ripeness.

g e r m.i n a t i o n

Germination.of the seeds requires moisture and light. Seeds placed
on the well-lighted surface of a water-retentive material will germinate
in four to seven days at room temperature. Thus for adequate storage,
the seeds must be kept dry.

Germination is precluded in darkness, but the light requirement is

met with low light intensity. However, seedlings kept in low light are

 

 

p l a t e l 5

Seeds of Castilleja coccinea (90X).

 

.....

a

'30 a s 3-1211'1193 a? j,

" iii” 913 euqug ,”
sjéw od WIfiT¥I 0;: eeaow. a“ 2

A8

9&1me

QM‘ Ens“!

. I’M. “\[1 O? M. “311
v, , g

3;" "'fi "‘1 'l‘”:

9’
v

JFIQEIf1"-”.h

v”.

59

W194

     
 
 
  

;3tellid853
AR .beomos-v
21

3b0" bees

‘1 .
figseaessn
.’. |

M , V

(.3 '
s A

r‘.—. .

1,.

.1

p l a t e l 6
Seed of Castilleja coccinea (lSOX). The seed coat is cellular-favose,
or honey—combed. A transparent, membranous tissue covers the 'windows'

of the seed coat, but imposes no barrier to water or light, two of the

factors necessary for germination.

 

 

 

50
t a b l e 1

GROWTH DATA OF A CASTILLEJA COCCINEA POPULATION IN THE HOUGHTON LAKE AREA

8

 

seedling (epicotyl spread less than 0.5 cm)

rosette (0.5 cm or more in diameter)

 

r

 

bolting rosette

f —— flowering or fruiting

The numbers after glare rosette diameters in centimeters, and after g_are
heights of flowering shoots in centimeters. The numbers 1-3 after §,indi-
cate (1) flowers largely immature, but at least one Open, (2) about half

the inflorescence has ripening capsules, but some corollas are still intact,

(3) all the corollas are withering, and most of the capsules are ripe.

.-1 H .—u
d a t e \o \0 r4 ~o rs r4
6 s‘ ‘9 00 “2 ‘9
N H m .--I m \0
tag 1,: .5 (x. L: 00 o:
1 e3 e20f2 e21f3 f3 f3 f3
2 e8 e20f3 el9f3 f3 f3 f3
3 el e25f3 e25f3 f3 f3 f3
4 r2 r2 r2 r2 r2 r2
5 e2 missing dead r2 dead r2 dead r2 dead r2
6 e7 e2lf3 e20f3 f3 f3 f3
7 eh e16f3 el6f3 f3 f3 f3
8 e2 e10fl e12f2 f3 f3 f3
9 e3 elhfl f3 f3 f3 f3
rl e5 ellfl e12f2 f3 f3
r3 r3 rh r5 r5 r6
r2 r3 rA rh r4 rh
s r% rl missing missing missing
e2 e9fl e9f3 f3 f3 f3
e7 e18f3 e18f3 f3 f3 f3
e6 e22f3 e22f3 f3 f3 f3
ea el9f3 el9f3 f3 f3 f3
3 rl rl missing missing missing
e9 e25f2 e25f3 f3 f3 f3
s r% rl r1 r1 r1
8 s r% dead missing missing
5 r1 r2 r2 r2 r2
3 missing missing missing missing missing
e6 e8f3 f3 f3 f3 f3
3 s r% r% r% missing

e10

e10
r1
eh
elO
cl.
e2
95

6.17.61

el7fl
e12fl
e12f2
r1

ehfl

’1
MW!“

1'3

e12f1
e23fl
ellfl
elhfl
el9fl
e28f2
e9

e18f2
elhf2
el6fl
el6f3

e26fl
e28fl
e25f3
e15
e24f3
missing
e13f2
e22fl
e21fl
ellfl
e22f2

51

( c o n t i n u e d )

703061

el7f3
e12f3
e15f3
r2
ehfl
rl
e5f2
elOfl
r6
r2

r2

r2

r1

r1

el

r3
e13f3
e25f3
e10f3
e20f2
e25f2
e28f3
e17
e18f3
e13f3
el6f3
el6f3
el
e27f2
e35f2
e28f2
e27fl
02MB
missing
e12f3
e22fl
e22f3
ellf2
e22f3

missing
r2

r2

r1

r1

r1
miSsing
e1

r3

f3

f3

f3
e20f3
f3
e27f3
broken
f3

f3

f3

f3

necrotic

e33fl
e32f2
e32f2
e30f2
f3
missing
f3
e24f2
f3

f3

f3

f3
missing
r2
r2
r1
dead r1
r1
missing
dead e2
r3
f3
f3
f3
f3
f3
f3
f3
f3
f3
f3
f3
missing
f2
f3
f3
f3
f3
missing
f3
f3
f3
f3
f3

e10

6.17.61

e2hf2
e26f2
el7f1
e17f3
el6f2
el7f2

e8fl
e8fl
e8fl
e12f3
e13fl
e12f2
el7f3

CI

52

( c o n c l u d e d )

e10f2
ellf2
e20fl
e15f3
e13f3
el7f3

r3
missing
missing
missing
missing
missing
r3
e22f2
missing
dying
r2

r1

r2
missing
f3

f3

f3

f3

f3

f3

f3

r3
missing
missing
f3
f3
f3
f3
f3
f3

r2

dead
missing
missing
missing
missing
missing
r3

f3
missing
dead e2
r2

r2

r3
missing
missing
f3

f3

f3

f3

f3

f3

53

markedly etiolated. In table 2 (page 54) are compared the germination
trials of illuminated and darkened seeds. The set-ups are shown in plates
6 -8 (pages 17-22).

Lighted and moistened seeds will germinate at 4°C, but only at a low
rate. However, seeds chilled moist for one month and then warmed and
lighted germinate both more quickly and at a higher per cent than do seeds
which have not been so stratified. However, it must be emphasized that
seeds collected in.mid-summer genninate at a 70% level without cold treat-
ment, and therefore it can not be said that stratification is a germination
requirement, even though it does increase both the rate and the level of
germination.

Tabulated on page 56.(table 22) are the results of genuination trials
in the greenhouse of seeds of 0-2 years age stored dry at A°C and of two-
year-old seeds stored dry at room temperature. Seeds kept dry at either

temperature suffer little decline in viability over a period of two years.

SA

on

s.o

Ne
Hm
mm
sm

axon: m exam: 0

:oapmsaeaow peoo pod omwao>w

Ame asonroa m mHopmsaxopddm .emcma ondpmpooeop HmQLSHe oommlomH **

mosap econommae us may one: mopmoHHdoa one *

UOON on
seats: ens eoeemaa
cone .mxoos_4 new

0 o m oooa use .eeaos .eoaeeaee

ma m.o *s ooa

o o *s ooa
no mm m OOH
m.H O m OOH
Ob Om H OOH
m.O O H OOH
mm N@ m OOH
me Oh m OOH
we Nm m OOH
m0 HO m OOH
t
9303 4 9303 N w m.
.m //

C
“n .u
D. e
n a

use .emaos .eseemaa
Doe .pmHoE .eeCOxnme
omSOQCooam

emceetme

omzozcoeam

eomoxame

omsozceeam
ow50£Cooam
omsoncooum

*kemooncoeam

ecoEpmopp

Doom
eoom
eoom
eoom
eoom
eoom
eoom
Doom
004

use

004

ommaoum

mesa sass ma
mesa sass ma
mesa sass ma
mesa sane ma
mesa sass ma
mesa sass ma
mesa sass ma
oesa sass m
oesa sane m
aesa sneeze m

mesa sane ma

oopoeHHoo opmo

when Hm
*msee as .m
*mmme an .m

ease em

meme em

whee N
when N
ease ems
meme ems

ease mes

mbme N

0mm econ

A<>H>mDMIOZHHOmmm Qz< ZOHH¢ZHzmMOIOMMm mo mq<HmB mmbomzmmmo mo menbmmm

N o H n m p

55

h y p o c o t y l g r o w‘t h

Three or four days after Castilleja coccinea seeds are moistened,
lighted, and warmed, the cotyledons turn green and can be seen easily
through the seed coat. On‘the fourth or fifth day, the elongating radi-
cle of the embryo pushes through the seed coat, usually bending in a
strong negative geotropic response (plate 17, page 57).

The seedling hypocotyl continues to elongate steadily after germina-
tion. Roots of greenhouse seedlings grown under glass often reach a length
of 5 cm. The root system of these seedlings consists of a very few slender
roots sparsely 'branched and growing almost vertically downward. Under
greenhouse conditions, growth continues for up to four months without bene-
fit of host-contact.

The apex of the radicle of the germinating embryo is conical and sub-
tended by'a prominent collar of root hairs up to 0.5 mm long (plate 17,
page 57). As the radicle elongates, root hairs are produced behind its
apex (plate 18, page 59), but not so 00piously as on the roots of many
autotrophic plants. These subapical root hairs are seen readily on seed-
lings grown behind glass. After extensive contact with a host has been
made, the root spices die back to the most distal haustorium, and many root

hairs are lost.

"‘43

6

:

i

’. s
‘1. -‘
‘ !

'fi’t.

Q»
,3

'Z 9 "~’_"

me“-

.‘,

«am

an
1 a
1,.
r .
s
. x
c
u
.
b
a
D

‘13 r::3f

.‘5-

e
VI.

3 '1
. 57'} 1817631100551

i

56

10 981105

 

+7.. A ,.

33.. £0998 seniooz_ i;”rii~“

“,_a_g--.-c.o- — Dun-a

fiuenism-w a T ..

u'".

. .
I. . .
‘0 my,"

p l a t e l 7

Young Castilleja coccinea seedling (lSOX). The conical radicle is
subtended by a prominent collar of root hairs. The honey-comb seed
coat ruptures easily, and is lost soon after expansion of the coty-

ledons.

 

3).,

“T "VKK!

A
1x

(\

p l a t e l 8

Qgstilleja coccinea seedling with expanded cotyledons (150X). The

 

glandular multicellular trichomes abundantly clothe the cotyledons,
leaves, stem, and flowers of the plants. Root hairs are produced
initially in a collar on the hypocotyl at the 'soil-line', but later

are produced just behind the apex of each root tip.

 

60

e p i c o t y l g r 0‘w t3h

When a Castilleja coccinea seedling contacts a host root and success-
fully penetrates to the vasoular stream, the foliage leaves of the seedling
expand dramatically, often doubling in length in 24 hours. Graphed in
plate 19 (page 62) are the growth curves of three rosettes grown behind
glass with host plants.

Same-age rosettes, even those parasitic on the same host plant, have
widely-varying growth rates. The growth curves of three rosettes parasitic
on different hosts are drawn in plate 20 (page 63), with time graphed against
the number and the length of the foliage leaves. As well, drawn in plate 21
(page 64) are the growth curves of three groups of rosettes, each group growb
ing together in one pot and parasitic on only one host plant (see plate 12,
page 33)-

Host-contact is essential for maturation of the rosettes of Castillejg
coccinea. This assertion is based on two lines of evidence. (1) Immature
rosettes, when divorced from their hosts, fail to mature further, and die
after 2-3 weeks of apparent inactivity. (2) Some four thousand seedlings
in 12 lots have been kept without hosts on a variety of substrates, and
none of them.deve10ped more than five foliage leaves. About 500 of the
seedlings were kept host-free for as long as four months. Given host-
contact, a plant in those four months nonmally would have formed a rosette
2-3 cm.in diameter and.made up of 10-15 leaves about 15 mm in length (plate
29 , page 95)-

The first indication of bolting is the production of mitten leaves.

The rosette leaves are always entire, while those of the flowering shoot

are single or double mittens, often deeply lobed (plate 22, page 66).

61

Greenhouse-grown rosettes readily flower without cold-treatment,
usually by the sixth.month after sowing. Thus cold-treatment is not an ‘
important factor in triggering bolting of mature rosettes.

The bolting Shoot soon produces flowers and the showy colored bracts
of the inflorescence (plates 23 and ‘1 , page 68 and fronti spiece). The in-
florescence is indeterminate in its growth pattern, and 5-20 flowers are
produced on a typical shoot. Flowering in the Houghton Lake area begins
in early May, and continues into September, with the peak in late June.
The capsules ripen from June through September, with the peak in.mid—July.
However, dehiscence of the abaxial locule of the capsule is delayed as
much as a.month, a mechanism which disseminates seeds throughout the grow-
ing season.

The parasitic attachmentis not required for bolting and flowering
once the rosette has reached maturity; Tabulated on page 69(tab1e 3)
are the heights of flowering shoots of 30 mature rosettes, half of them
teased free of their hosts and repotted, and half of them left undisturbed
in the small blocks of vegetation they were collected in. The difference
in.mean height between the two lots of rosettes is no greater than that
expected by chance, as tested with the t distribution.

Although the rosettes can flower detached from their hosts once they
reach maturity, there is no evidence that they are detached or otherwise
isolated from their hosts during their normal flowering period. Eosin
transfer experiments in the greenhouse are as successful with bolting
rosettes as with young rosettes, and in histologic section the haustoria

of adult plants in the field appear functional.

11 mm

1

d i a m e t e r

r o s e t t e

 

p l a t e l 9

Growth of Castillejn coccire< rosettes in response tn hast—corzac

 

 

p l a t e 2 0
Growth curves of three greenhouse-grown Castilleja coccinea rosettes

parasitic on different hosts.

m i l l i m e t e r s

a m e t e r i n

d i

 

20 30 to 50

r o s e t t e

t e i n d a y s

3.1.
B

l parasitic on Hieracium aurantiacum

 

2 parasitic on Lactuca canadensis

3 parasitic on Antennaria neglects

 

 

 

 

 

 

 

m c w o E H p
em ea ce as om oa oe om om oa
P: b u - u n n n n . a
was. emw 2:. “so manganese moses pee: mamuoemsmo 33qu neon

 

 

ram

.mpmoz mHmch :o owpflmmhma

TmH

FOA-

ON
Pom

 

 

mN

 

.os

Om
moppemon mo museum 09:3 eao wearaoo piano

HN eueHd

p l a t e 2 2

Five Castilleja coccinea rosettes parasitic on Krigia biflora, which has
died back to only two leaves in the center of the pot. The largest
rosette is bolting and producing the lobed 'mitten' leaves of the
flowering shoot. All the rosettes are the same age, about six menths,
but are of markedly different size. Initially 600 seeds were sown in
the host-pot, and over 100 seedlings survived their first month. Those

which failed to attack the host eventually died (1.8X).

 

67

p l a t e 2 3
Bolting rosette of Castilleja coccinea. The same rosette two weeks

younger is pictured in plate 22 (page 66).

 

 

69

t a b l e 3
HEIGHTS OF FLOWERING SHOOTS OF NORMAL AND HOST-DIVORCED ROSETTES IN CM

normal host-free

20 12
20 12 normal host-free
25 10 n = 15 n = 15
21 23 2x = 256 Ex = 255
16 11 2x2 = 4390 2x2 = 1.397
10 11. (2x)2/n=z.369.0667 (2x)2/n 4335.0000
11. 19 i = 17.0667 1'; = 17.0000
9 28 2
s = 0.1152
18 9 ’2‘
- _ = 0.01936
a m H'&
s- _. = 0.1391
19 11+ x1 x2
1: = 0.117%
25 16
8 16 t97.5(d.f.28) — 2.01.8 (Hald, 1952)
0.1.794 < 2-0h8
17 26

difference in means not significant at 5% level
12 27

7O

pih yrs i o l o g y
p h o t o s y n t h e s i s

Castilleja coccinea is a species of wet meadow and savannah vegetation.
It rarely occurs in the shade of other plants. Its leaf mesophyll is undif-
ferentiated and highly lacunar (plate 24, page 72). The epidermal cells of‘
the leaf lie in a single layer, are conical, and lack a cuticle. Glandular,
multicellular trichomes are abundant on the stem and on both leaf surfaces.
Stomata also are plentiful~———————:their density averages about 29,700 per
square centimeter (10 counts) on the lower (abaxial) leaf surface, and A950
per square centimeter (10 counts) on the upper (adaxial) leaf surface.

On the basis of experimental evidence, Castillejg coccinea is photo—
synthetic. In one set of experiments, presence of starch in leaf tissue was
considered a demonstration of photosynthetic activity. The evidence is that
(l) starch is formed by seedlings and excised rosette leaves under strong
illumination, but not in darkness, (2) seedlings and excised leaves illum-
inated in an atmosphere free of carbon dioxide synthesize no starch, and
(3) illuminated seedlings and excised leaves synthesize no starch when their
stomata are plugged with petroleum jelly (table A, page 73).

Another set of experiments was designed to test the ability of host-free
Castilleja coccinea rosettes to fix carbon dioxide when lighted. Castilleja
coccinea rosettes of various ages were pulled from their host-pots, separated
into four groups, and placed on moist blotter discs in plastic Petri dishes.
They were kept in darkness for twelve hours. All four groups then were given
C1402 for 30 minutes during which two of the groups were lighted and two were
retained in the dark. They then were killed and prepared for radioactive
assay. The results are in table 5 (page 7A). The darkened rosettes showed

almost no increase over background, while the lighted rosettes showed

substantial levels of radioactivity.

K‘-

:

.

.

.

J

'.

v:

.

l
:‘7‘...-," "

I
- . ._ . V l._ q ,
5.. 1.. - g“. Q 1
_ ‘ ‘ ‘ —— ‘ J I

Q . .'
I
V
\
I ‘ ’\
. ‘A
.'~ 1
5'
u‘
‘(
”I
. ,
v
r
\
.. ‘
l ““u.
‘7 .
\
C
- v
I»
c"‘
I i
,4
\
. T
I
f . e
" \
t
'. 1 -
‘ ‘1 - '
f “"w -‘ ‘.. I 5.
‘u
. '2.
o
1"
fl
“
, . . . . 1‘ ;-
,t'
5
‘.
t} - .
l.

1
o . .
~ 'f r C‘ j. '3 ‘ \. {In _- . 5
k. a . ' ‘ ll: (' . ~_4 v ) .. ‘5’
e O l . r ' J a
7 ‘f I‘ I (2" j \ L“ LEI. : .
‘7‘ »\~ll_ '1 \—

O o
1; I - new {71113

b .‘0’

r t ,1 r

2 u. _' \ .4}
, .

"5, o' 4‘ ' (I.
. p I J. C'L“,

r
,5 r .. .. l.
L _ .A \ ‘." -- ~-
.

(

p l a t e 2 A
Photomicrograph of a cross-section of a leaf of Castilleja coccinea (SOOX).
- The mesophyll of the leaf is undifferentitated and highly lacunar. The

epidermal cells are conical in vertical section.

a m c H o #
wmmcfiem ow ammam mom ma>mom Hz MNOHmmc Hm><mm >2u momalbwmm mmwchzom ow o>maHrrmu> oooonwp

IT mdwwo: peomoSd

II mdmdo: humane

 

 

 

 

mere Homwae deemesasa «Newman Powwow moonwusmm
QmQHHmsd 5050* n+. 1+1
@3533 5050 II II
ameHmsd now oNoHsaoa 1|. 1|.
amqumrd «weave» chmmaQ .II II.

 

73

*mHH How<eu teen «Newman can Hwomdoa o: nemewwwoa zwaoe H5 mafia» numrom H: are gnaw woe

Np vogue camouo «Homeaoav

.mpcsoo OOOH Gama» op 05mmfip mo mpnwfloz Hedda pom meEHp mam GEsaoo pmpooppoo one QH mepsmflm
esp .msne .UeHQEdm we>mmH opflmwpmm esp mcosm pgmflmz cw meoGoLeMMHe esp pom msoaam coapoepnoo mflne **

.mpcsoo OOOH Upoomp op popesvep mmpscws cw asap exp mfl mpsmfim mEHp Seem *

ao.H mm.o mm.o em.o mm.o mammo.o samea e

.4 oo.H oo.H mo.a no.0 ma.o eaaao.o samaa m
7. mm.©m Nd.©m mm.om N4.wm 00.0m Omamo.o mmecxpmp N
cm.am oq.am ma.am mH.em mm.mm m aamao.o mmecxtme H

**Uepooppoo ommem>m Amposzam oopcpv *oEflp pnmflez eeflwwp psmwa QSOLm

 

.mmppemoh mmcHoOOo wheaawpmwo mo mo>mmH powwoxo pocoxpmp paw Uopmcflesaaw hp cowpmxwm JHIGODLwo mo somwpwdEoo

m c a p m a

75

An estimation also was made of the level of photosynthesis of rosettes
still parasitic on host plants. C1402 was given to Caatilleja coccinea
rosettes and their hosts for 30 minutes. The host and the parasites then
were weighed, killed, macerated, and their activity sampled. The results
are in table 6 (page 76). The level of 002 fixation by the parasites was
only slightly higher than that of excised leaves. However, the host plant

fixed about five times as much carbon per gram of tissue as the parasite did.

76

.opcsoo 000a paowm on odomwp
opfimonoa new pmon mo onnmfioz Hmsvo pom moEHp one cesaoo Uopoonpoo one Ca mopsmflm one .msnB .oHQEoo
mood pmon onp ego oHQEoo mood opfimoaod one noozpon pnmwos cw oonopoMMflp one now ozoaao coapoopnoo mane **

.mpasoo 000H Upooop op noneswon Amaoflhp 0H mov oopdcfis CH oEHp omopo>o onp mfl opdmwm oEflp noom *

 

 

 

 

0m.0 00.0 Hm.0 00.0 40.0 ddm00.0 oonwOOOO ofioaawpmoo
NH.0 NH.0 HH.0 0H.0 mH.0 Hm00H.0 mflmnoeocoo mospooq
00.0 m0.0 00.0 0H.H 00.0 mmm00.0 oonfloooo mfiwaawpmmu
4H.0 4H.0 0H.0 mH.0 HH.0 m 0HNNH.0 mfimcopocoo weepowq
* * 0 o p o o p p o o o m o n o > o Amposzao oopnpv * o E w p p n m a o 3 p H o a

 

 

.oaasoo owpfimopoa o CH oocwoooo ofioaawpmoo 02o mwmnovonoo manpooq An :Oflpoxflm Jalconpoo mo acownoaaoo

0 o H n o p

77

p n a t u r e o f t h e p a r a s i t i s m

In view of the facts that Castilleja coccinea is photosynthetic and
yet requires host-contact for'maturation, and responds quickly to host-
contact, it seems likely that some micro-metabolite is one of the sub-
stances the parasite uses from.its hosts. For this reason, host-less
parasite seedlings grown in sand or agar were irrigated with a few of the
best-known plant vitamins and hormones. However, none of the treatments
simulated the effects of host-contact. A list and description of the
treatments is in table '7(page 78). In general, the concentrations of
the compounds were chosen to include the range of concentrations reported

effective for each compound.

gag m use .o .04 .00m .OOOH

and N 0co .m .m .0H

0 A A q A
2 MHIOH Ucm HHIOH 0IOH 0IOH mIOH MIOH

q A q q q
2 manoa 6am Hauoa 0-0a anoa muoa

ea
mm a ooa\a use
and 00H\H pno
sea ooa\a ace
Sea ooa\a use
gag ooa\a use
sea ooa\a use
a o a s
.oASpHSO

.oH\H .H .oa
.oa\a .H .oa
.OH\H .H .oa
.oH\H .H .oa
.OH\H .H .oa
.OH\H .H .oa

mIOH
.ooa
.ooa
.ooa
.ooa
.ooa

.OOH

o p p c o o c o o

oLSpHSO

opSpHSo

econ

pnoo

opszSo homo OHpaooo

oQSpHSo homo oflpqomo

oLSQHSo neon
endeSo pcom
oaSPHSo Unom
oadeSo Unom
oaSpHSO Unom
oLSQHSo econ
opszSo econ
opSpHSo 0Com
opspHSo econ
oadeSO econ
a o a p o e

 

pope: emaaflsmaa

 

mwmcoponoo weapooq mo pooppxo nope:
pogo pooppxo oooosam oCOpahop Opoom
Amaaa .osaamv coassHOm sameness
606m oaaaoemnnaw

capoCflx

peso oflpoooocoaonpndon

ewoo ofloooooaopcfl

oefinoanooapmn ocfixopflpmm

pfloo ownpoomo

HopfloOCH

:wpown

Anfioo owcflpooficv swoofin

ocaeaaas

U n s o Q E o o

ammo one econ :fi oonwooOU Mfimaawpmoo no>flm mucospoopp opflaonoposlonofla mo pqu

a o H a a p

79

T H E H O S T S

The host-lists for various root-parasites have been compiled largely
(by excavating the parasite roots and simply noting the plants the haustoria
are attached to. Unfortunately, some if not all root-parasites also form
haustoria on materialswflfibh.are nutritionally inert. Castillega coccinea,
for example, will form.haustoria on pebbles, grains of sand, aluminum foil,
and leached pith. The suspicion arises, therefore, that not all the hosts
a parasite makes root connections with are functional hosts. This diffi-
culty can be avoided by growing the parasite in pots with single hosts
and noting its success on each one (plate lih page 33).

In this way Castilleja coccinea was tested for its ability to para-
sitize successfully l7 vascular species common in an area of central Michi-
gan where Castilleja coccinea itself is abundant. The host species used

in the study were mtenngzgg neglecta (plate 26, page 83), Fragaria virgini-

ggg, Chgysanthemum.leucanthemum, Achillea.millefolium, Populus deltoides,
Lobelia a icata, Alg__m sa, Hieggcium ur tiacum, Lactucg canadensis
(plate 29, page 95), Egigig biflora (plate 22, page 66), Dggthonia s icata,
Rudbecgia hirtg (plate 12, page 33), Egbg§,his idus, Panicum sphaeroides,
and three Solidggo species, S. juncea (plate 25, page 81), S. raminifoli ,
and S. ruggsa. Specimens of each species are filed in the Michigan State
University herbarium.

All of the 17 hosts supported the parasite's growth, but the rosettes
parasitic on Popglus deltoides, Alggg osa, and §g§g§,hispidgs never
matured (table 8, page 81.).

Under natural conditions a Castilleja coccinea rosette usually para-

sitizes several hosts around it, and these rarely are of the same species

80

 

p 1 a t e 2 5
A mature rosette of Castilleja coccinea parasitic on a rosette of

Solidago juncea.

81

 

 

 

[‘42

p l a t e 2 6
Two rosettes of Castilleja coccinea parasitic on a vegetative shoot of

Antennaria neglecta.

83

136

 

8h
t a b l e 8

LIST OF HOST SPECIES IN GREENHOUSE CULTURE OF CASTILLEJA COCCINEA

 

 

 

 

 

 

Achille§.millefolium_ Lactuca canadensis
* Alggg £25255 Lobelia §picata
Antennaria neglecta Panicum.§phaeroidg§
Chgzsanthemum.leucanthemum. * Papulus deltoides
fgaleua Llama; * Lubes _a_his idus
Danthonia spicata Rudbeckia higgg
Fragaria vir niana Solidago ggaminifolia
+Hieracium aurantiacum S. uncea
+bKalanch03 verticillata S. dsa
Krigia biflora +Trag0pogpn pratensis

* supported only early growth of the parasite

+ species not native to the range of Castilleja coccinea
total genera represented 18

families represented 7

monocotyledons 2

85

unless the host daninates the area. This multiple host-contact is demon-
strated readily by excavating and tracing the roots of the parasite. Howb
ever, it is difficult to trace more than a few of the connections, since
in the ta3k of tracing each one, many haustoria, host roots, and parasite
roots are destroyed.

Castilleja coccinea successfully parasitizes hosts of a variety of
Species, but usually in greenhouse culture a few host species appear to
support its growth better than others do. In table ‘9 (page«86), several
hosts are ranked according to the number of parasites they support in
greenhouse culture. This ranking may reflect nothing more than the density
of their root systems, but nonetheless in greenhouse culture there is a
striking difference in the vigor and numbers of the parasite on different
hosts. The ranking of each host species is based on the average number

of parasite rosettes supported by that host in 10 different pots.

86
table 9

RANKING OF SELECTED HCBTS ACCORDING TO THE AVERAGE NUMBER OF CASTILLEJA

COCCINEA ROSETTES THEY SUPPORT IN GREENHOUSE CULTURE *

0‘) U)
n n
" 8 to
g 3 c:
<r co ‘3
H
a if: .8
Lgctugg can ns 28.6 12.1 1.3
Frggagia m 21.5 10.1. 0.9
W legcggthemum 18.0 10.5 0.8
was 1111:1023 22.9 9.2 0.6
Rudbecng mm 19.7 8.2 0.6
Mg 321.0th 10.7 9.2 0.5
Antennaria 29319.20: 4.3 9.5 0.5
22119332 32am 8.3 7.9 0.5
W agratiacum 1.1.7 7.7 0.3
Sogdgo m 1.7 1.1+ 0.3
mm 2-1 0.9 0-2
mm gpicatg 7.0 1.2 0.1
m sphaggoides 1.0 0.7 0.1
Resales 121mm 0-3 0-h 0
Rubus W 0.5 0.3 0
flag Egosg 0.2 0.2 0
c o n t r o l O 0 0

* averages based on a total of 10 pots per host species

87

T H E H 0 S T - P A R A S I T E C O N N E C T I 0 N
The mature haustorium of Castilleja coccinea consists of four distinct

(1) a central vascular trace which con-

 

regions (plate 27, page 89)
nects the stele of the host root with the stele of the parasite root, (2)

a mass of denselybstaining parenchyma surrounding the vascular trace, (3)

an outer rind or cortex of vacuolate parenchyma, and (h) a pad of columnar
cells next to the host root.

The rate and course of haustorium development in Castilleja coccinea
and Melampxrum lineare were studied by harvesting and sectioning haustoria
of known age. Several typical haustoria studied in this way are diagrammed
in plate 28'Cpage 91) in successive stages of development. These stages
include (1) the formation of papillae on the parasite root at the point of
eventual haustorium formation, (2) the lySing or softening of the host root
apparently by secretions of the columnar cells, (A) the penetration of the
host root by an elongating wedge of parenchyma tissue originating in the
interior of the haustorium, (5) the differentiation of vascular elements
in the center of the parenchymatous wedge, and (6) the differentiation of
a mass of densely-staining parenchyma around the central vascular trace.
The internal cells of the host root rarely are distorted next to the haus-
torium, suggesting that they are not forced aside or compressed by the pene-
trating haustorial tissue.

The formation of haustoria apparently is not induced by living tissue
alone. To be sure, most of the haustoria of Melampyrum lineare and Qéélllf
lglé coccinea plants are found attached to host roots, but plants of both
species also form.haustoria on sand grains, pebbles, and organic debris.

In the laboratory Mglgmpyggm,lineare attacks its own discarded seed coat,

and both it and Castilleja coccinea produce haustoria on the cut surface of

p l a t e 2 7

Photomicrograph of haustroium of Castilleja coccinea, about AOOX.
Only part of the vascular trace connécting the stele of the host with
that of the parasite is visible, since the trace meanders somewhat in

the body of the haustorium.

89

 

 

.- _ - h
J
-1 ., '
.-
-.a
e
\r ,—r v\,-(“ "V
. s .
' '. 7 “ "" -.-_. -
x
' ‘ I. -
‘l. _ - h.
l
.
.
.o
.
-
.v
0.. A
. . ' .~
.
.
\
I §
0
a
it ' ‘
.
u
.
_.r
‘r

l
u...

 

 

.
I
D
' l
.
.
.
,V
.
I
Q
.' '
\

Q

1
t
.
.
.
f
.
A-
- I

 

 

p l a t e 2 8

Stages in the develOpment of haustoria by Castilleja coccinea.

92

moist, leached elder pith. Moreover, Castilleja coccinea and Melampygum
lineare plants grown behind the glass of glass-blotter sandwiches form
haustoria at only a few of the numerous sites where parasite and host
roots touch. Thus, physical contact alone does not induce haustorium
formation. It also appears that the roots of these two root-parasites

do not seek out host roots chemotactically. The roots of over 100 Mglgm—
21; m lineare seedlings and 200 Castilleja coccinea seedlings grown in

38 separate glass-blotter sandwiches were not visibly attracted to host

roots.

93

T H E H 0 S T - P A R A S I T E I N T E R A C T I 0 N

Four aspects of the interaction of Castilleja coccinea and its hosts
were studied---- the host-requirement of the parasite, transfer of
materials across the haustoria, damage to the host in the interaction, and
the effect of the interaction on the distribution of the parasite in a
community.
host-rgguirement. Contact with a host is essential for maturation of the
rosettes of Castilleja coccinea. This assertion is based on two lines of
evidence. (1) About four thousand seedlings have been kept without hosts
on a variery of substrates, including sand, loam, blotter paper, elder
pith, agar, agar with mineral nutrient supplements, and agar with full
nutrient supplement. None of the seedlings developed more than five foli-
age leaves, and none of the leaves was longer than six millimeters. About
500 seedlings were kept for as long as four months. Given host-contact, a
plant in those four months normally would have formed a rosette 2-3 cm in
diameter made up of 10-15 leayes about 15 mm in length (plate 29, page 95).
(2) Immature rosettes, when divorced from their hosts, fail to mature fur-
ther, and die after 2-3 weeks of apparent inactivity. Death usually involves
necrosis of the apical region. Since mature rosettes can be pulled free
of their hosts without ill effect, it is unlikely that physical injury alone
causes the death of the young rosettes.
haustorial uptake. Haustorial transfer of materials was tested with four
compounds, two of them common ions (sulfate and phosphate), one the hexose
fructose, and one eosin, a mildly toxic dye. All four compounds will tra-
verse the haustorial connection in the direction of the parasite but appar-

ently not in the direction of the host. The eosin Y was introduced into the

’0
,
>0
'\
.
.
'\
‘0
V
‘\
W .
a‘ v'.
39’".
A.
.r.
[5. \

.’ '

.’ ..

.a
a} 'a‘
.v ,I‘.
*.
'. l .
‘ N I 9 - ‘ v. *3
_ m -

91+

‘
J
J
‘9
i
‘|
4
J
.
‘
a
.
’
\
.
. I
, 4 )
§
' ,- .
_ . .
'1' A7" " no .- r..

 

.o’v

)Fifl'.

'- a.

e
'2 »
-;-_: ""'-'-.—

-<‘—- -

p 1 a t e 2 9
A Castilleja coccinea rosette parasitic on a young rosette of Lactuca

canadensis. The parasite is four months old and nearly mature (1.7X).

95

 

 

\I)
l‘)‘ I!!!

\I’

I

.I

09‘s,).
-.oo.\. I..fal!7.l.v

96

host through a cut petiole, and was detected in the parasite by its color.
After the transfer, the parasite and its host were dumped from their pot,
their roots washed, and the path of the eosin traced through the two root
systems. Only some of the host roots had carried the dye, but earlier
experiments with eosin transfer had suggested this would be the case. In
these earlier transfers, pots were used which contained several parasite
rosettes. Introduction of the dye into only one of the several host peti-
oles resulted in accumulation of the dye in only one or a few of the para-
sites. Transfer of the fructose and the two anions was studied by using
radio-isotopes and monitoring their accumulation in the parasite tissue
with a Geiger-Maller tube connected to a sealer and timer. In plate 30
(page 97) are the data from a transfer of Clh-fructose to the parasite

from Frggaria yiggigiggg. Similar graphs for transfer of S3504 and P3204
are on pages 31 and 32 (plates 98 and 99).

daggge to the host. Only a crude analysis has been made of the perfonmance
of a plant under attack by Castilleja coccinea. The rates of attrition and
loss of heayily-attacked Lgctuc; ca ens s are noticeably different fnan
those of parasite-free hosts. Of 22 potted Lactuca canadensis hosts attacked
by twenty of more rosettes for h—8 weeks, only 5 survived. In contrast, the
rate of die-off of unattacked controls was 7 out of AO.

pagasite micro-distribution. Castilleja coccinea in the greenhouse is sup-
ported better by some hosts than by others. Thus, the hosts in a given
community can be ranked according to their ability to support the parasite.
One such community was studied to determine if the distribution of the
parasite in it reflects the ranking of hosts. It was presumed that the

parasite is more abundant around the hosts which support its growth best.

97

 

 

 

on 0N 0H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.oocpoooo owoaappmoo op ocop2pmnp> optomonm Eonm nomoconp Aoomamn&ov owopoznm mo nampw

on opapa

0m

on

on

(eqnutw-Jteq Jed squnoo) eqqesoa eatseaed JO fixatioeotpea

98

on on om op

 

 

.oonoooo onoaaflpmmo op ocoflnpmnw> oppomonm EOAM homeconp nomfim o no neono

H m o p o H Q

00H

00m

00m

edqesoa eatseaed Jo fixatioeotpea

(eqnutw-Jteq Jed squnoo)

99

om

on on om op

om

00H

omH

 

 

.oocpoooo omoHprmmo op ocoficwmnp> opuowoam Eoum homeconp nommm o no naoho

mm wasps

(eqnqu-Jteq Jed squnoo) eqqasoa eqtseaed JO KQIAIQOBOIPBJ

100

Thus, the study was set up to compare the density of each host away from
the parasite and close to it. However, no significant difference was
found between the densities of hosts close to the parasite and the densi-
ties of the same hosts some distance away from the parasite. The sampling
data are tabulated on pages]Ifl.anle2.(tableEfi. The numbers indicate how
many individuals of the host species occurred in a quadrat, but where the
individuals were too numerous to count or were clonal, as many grasses are,
presence is noted simply by the letter 2, and no attempt is made to record
the number of individuals present. Results of the statistical analysis of
the data are on page 103(table 9). Only some of the host Species were
tested statistically. Those that were not tested either turned up in too
few quadrats (fewer than 5), or else they were clonal and individuals could

not be counted accurately.

a coccinea present in quadrat

.i

f

 

o Castille

8

t a b l e

 

x Castilleja coccinea absent from quadrat

Castilleja coccinea micro-distribution sampling data

 

m. b e r

n 11

q u a d r a t

h o s t

om
N—d’
OM
ON

OH

0 0 0 0 0 O 0 0 0 O O 0 0 0 0 0 0 0 0 0 0 0 O 0 0

Achilles millefolium

Alnus rugosa

0 0 0 O 0 0 0 0 0 0 0 0 l l l 0 0 0 0 0 0 l 0 0 O

 

0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000
0000

0000

 

:53

F:

o

J:

.p

n
“5 ‘3
*5 :3
o m o
r1 ¢4r+
o c‘S-e
c.ea
o o c o
75235
S m min
E o
600‘”
.p p L m
c o o.n
<cc3koc3

 

 

 

101

0000
0000
0000
oomo
001-10
0000
0000
0000
0000
0000
$000
0000
0000
00:40
0000
00:40
0000
0000
00m0
0000
$000
00:40
0000
@000

0000

Virginiana

 

 

 

Eleocharis compressa
Equisetum arvense

 

Danthonia §picata

Fragaria

0000
0000

OOHO

O

000
000
000
000
000
000
000
ONO
000
0H0
01-10
0N0
000
000
000
000
000
00
ON

C>c>

0 O O 0 O O 0 O 0 0 0 0 0 0 0 0 0 O 0 0 O

00

 

 

o
a:
-H
04-3
an
0001‘s
0:45;.
:0ng
H0514
CHCH
cogs-Ma
~r-i 0,0
Saar-l
(00030
$210311
0:40
.O0Cor—1
cue-{SS4
:1:th

 

 

O O 0 O O O O 0 0 0 0 0 O 0 O O O 0 O 0 O 0 0 O O

Lactuca canadensis

 

0r—I
0C0
00
OH
00
00
00
00
00
00
0H
06'\
0—3
on
0.3
0H
00
00
00
ON
00
00
00
00

00

 

 

(0
(0+3
£403
00
8*“
“5%
(060
‘H0H
S-«r—i
+30
‘09
"-40
I—lr—‘l

l 2 0 l O O 0 1 0 0 0 l O O 1 O O 0 2 3 l 3 1. l l

 

Oxalis violacea.

00 0:40:00
00 0r-h-{00
00 0HOOO
00 0m000
00 0c-lr-100
0000000
00 004000
0000000
0<3 0<DCDCDCD
o<3 o<3u4<3<3
0000000
00 00H00
oco oracococp
0O 0HOOO
00 00000
00 000100
0000000
0000r-l00
o<3 oto<3<o<3
o<3 o.a<o<3<o
00 0.4000
0000.400
00 00000
00 000100
00 0mr—l00
C.
o
21 .5
<0 0
8 Canola)
L 7§yp:j.g
33323 a in $53
£1.50 005.2 0::
0480)“) +3?"
0 a o ova
+3 Dita-E0
g m Era o m
p 0030 5
0 030 020 >46
0H S th 0
n C 0.4'0 50H
0-H o o 5 o
(LO-«04040110103

 

 

 

 

 

 

 

Sisvrinchium montanum

C) 0 0 O 0 O O 0 O O 0 0 0 O O O 0 O 0 0 0

A4

 

-:

l O l O 0 O 0 O 0 0 O 0 0 O 0 O O 0 O 0 2

O O 0

 

Spiraea alba

000
H30
NHO
OmO
0\00
qmo
0—30
HHO
0‘00
0—30
0610
000
000

omo

.5100

0H0
r-{NO
r-I—EIO
0N0
0.20
000
O\O
”0
H0

H0

Solida o ru osa

 

Solidaggggraminifolia

Solidago juncea

 

(concluded)

8

t a b 1 e

o-distribution sampling data

a coccinea micr

gj

 

Castill

I‘

n u m b e

q u a d r a t

°€R
>¢O\
aco
gt»
>00
OLA

4:
”3
”53
°§
0:4
cc:

4:
cc»
NCO

cw
oc~
oxo
oxn

rm
0.:
Mr“

cw
New
N:4
Kc:

cw
ocx
0:0
or»

xxo
o:

Achillea millefolium

Alnus ru osa

O 0 O O O O .O O O O_ O O 0 O 0 O O O :0 O O O O O O

 

00
00
00
00
00
00
0O
00
00
00
0:4
00
00
00
00
00

00

0 0 O O O O O O O l O 1 O O O O O O O O 0 O 0 O O

00

 

Antennaria neglecta

Carex aurea

 

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Carex granularis

0000
0000
0000
0000
0000
0:40?!

0000

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

leucanthemum

emum

r

 

Chr

Vsanth

A

 

0

0

8 8 6

68866886688

8

6

8

8

A

 

Danthonia snicata

102

0:400
0000
0000
0.:1':40
00100
0000
0000
0000
<3<D<D<3
0:400
0000
00010
0000
0000
C>C>c>C>
omoo
0:100
0N00
0000
0000
0000
0000
0000
0000
0000
:13 Cd
0 C
(D 0
00-H
£46) CCU
04g. :4
E 0 ' 0
DES—:0
o -H<U
c6:>:4
U) (U
-r—:
h S w-H
(Up-Eh
.130 cu
O-H t1 (1)
m 3 fits
HUS-.m-
LTJLIJMZLZ

 

 

 

 

O O O 1 0 O O 0 O O O O 1 O O O O O O O O 0 O O 1
0 0 O 1 O 0 l 0 O O O O O 1 0 1 0 0 0 0 O 0 O O 1

iacum

t

ler'ac1um auran

H

 

Juncus effusus
Krigia biflora

 

O O 0 O O 0 O O O 0 O 0 O O O O O O 0 O O O O 0 O

0 0 O O 0 O O 0 O O O O O 0 O 0 O O O O O 0 0 O

 

0000
C343r4
0:40
010:4
00:4
0600
0H:4
ONH
0C00
00-0
0100
0‘30
0:4:4
000
C)"\"\
OWN
01-00)
0001
000:4
000
0100
020
000
0M0!

0:4:4

 

U)
.r—1
(0
C000
0544-30)
U000
CUQ’ECU
% m-Hr4
(UCLO
O m-H
>
(150360
O'H0r4m
'JS-«r4-H
-p-p 0:4
00,0613
(IS-HON
r-Jr—Jr—JO

 

 

 

00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
:90
0:4
00
00
00
00

00

 

Panicum aghaerocarcon

Pinus strobus
Poa compressa

 

00N00000M0

001:4000

0010000

0010000

0100000

00:40
01:40
:400
:400
N00
:400
01:40
MHO

0100

666666666

:4:40
0000
000010
00000
00:40
0:400
0—‘100
0N00
0:400
0:400

001010

 

Rumex acetosella
Scirpus lineatus

 

Polygala sanguinea
Rndbeckia hirta

 

 

0

0

0

0

00

0

0

00

00

00

o1c>r4c>
.4<3.4<3
C)C3r4<3
f4C>c>c>
r4CDr4<D
C,c>c>c>
(3:4(VCD
c>c>uxc>
C>F4"\C)
p1c>uxc>

00:70

000000

000000

00:40.30

000000

00:40l‘0

000:4-30

00:40010

0000100

C>C>C>C>c\c>

0000:40

0000:4:4

0000:4:4

000:400

 

hium montanum

isyrlnc

 

S

 

 

w
o
(H
.H
C2322
'E o o
“Eff;
,0
r3 m~d E
0 8J0
m c -
m film m
m'U-U'U
L-H-H-H
-Hr4r4:4
9+0 0 o
cncncnco

 

t a b 1 e 9

COMPARISON OF DENSITY OF HOSTS CLOSE TO AND DISTANT FROM CASTILLEJA COCCINEA

values of t computed for selected hosts of Castilleja coccinea

my. msaaa 1438A
Egui set-g3 we 0.1819
QEBEEE 2§£E§E§. 1°09O7
Lobelia spiggtg 1.0512
Polygalg sggguinea 1.9137
Rudbeckia flirt; 0.3265
Spiraea sly; 0.0616
Solidago ggggigigglig’ 0.1070
S. unce 0-7595

theoretical t 2.011 (Hald, 1952)

97.5 (d.f.l.8) =
All the t values computed for the hosts are less than 2.011. Thus, for any one
of the hosts tested, there is no reason to believe that its density close to
the parasite is different from its density some distance away from.the para-
site. Stated another way, the ability of each host to support the parasite

in the greenhouse is not reflected in the distribution of the parasite in

the community. A similar analysis was made for Lobelia s icata, Rudbecgia
gigtg, and Solidago jggggg taken as a group. or the hosts which occurred

often in the sampling data, these three supported the parasite best in green-
house culture. As separate species, they were not significantly denser around
the parasite than away from it, but they were tested as a group to increase

the sensitivity of the analysis. However, even as a group they showed no

significant tendency to occur more frequently close to the parasite

t = 0.9797, which is less than t(d.f.l£.8) 97.5 = 1.978 (Hald, 1952).

1.01:,

DISCUSSION LND LITERATURE REVIEW

v a s c u 1 a r p a r a s i t e c l a s s i f i c a t i o n

vascular plant parasites are classified according to various schemes,
but most often they are divided into two large groups, those with chloro-
phyll and those without. The chlorophyllous parasites are variously termed
hemi-parasites, half-parasites, or hydro-parasites, since they are presumed
photosynthetic and dependent on their hosts for only water and mineral nu-
trients. Unfortunately, the classification of parasites by their color is
artificial--———-—- some genera contain both green and nonrgreen plants,
and in fact some parasites produce chlorOphyll as adults but not at all in
their first year or two or growth.

Two additional characters used in classifying vascular parasites are
the production of haustoria and the location of the haustoria on the host.

Host of the vascular parasites produce haustoria, and most haustoria are

 

formed on the roots of hosts. However, the exceptions are important ones
the mistletoes and dodders produce their haustoria above-ground on the
leaves and stems of their hosts. Such exceptions complicate attempts at
classifying vascular plant parasites. The difficulty can be resolved fully
only with an elaborate, and cumbersome, classification system. One scheme
is outlined on page 105. It is artificial but compatible with what is known
of heterotrophy in higher plants. Qggtillgjg_coccine§ and other species of

green root-parasites belong to the category marked with an asterisk.

105

V A S C U L A R P L A N T S
c h 1 o r o p h y l 1 o u s
antotrophic
root-grafting
haustoriate
Qt haustoria below-ground (hypogeal)
haustoria above-ground (epigeal)
a c h 1 o r o p h y l l o u s
saprOphytic
parasitic
not haustoriate
haustoriate
haustoria hypogeal
haustoria epigeal
A root-paggsite is any vascular plant which parasitizes the roots of
other vascular plants by means of haustoria, or, more simply, a root-
parasite is any vascular plant forming root haustoria. The definition is
a loose one, but must remain so until root-parasitism is better under-
stood-—-———-as yet little is known of what compounds the parasites obtain
from their hosts, and what they do with those compounds. The phrase green
root-parasite refers to vascular plants which are green and which produce
root haustoria-—-————-Castilleja falls in this group. F l arasite or
holopargsite applies to (1) any non-green vascular plant, (2) the young
unpigmented stages of Tozzia and Stri a, which are chlorophyllous as adults,
and (3) root-grafting or haustoriate vascular plants which produce non-
functional chlorophyll or which produce chlorophyll only during food stress.
Vgscular pgggsite is taken to mean a vascular plant which requires interaction
with a host at some time during its life cycle. This definition, however,
can be applied rigorously only to individual plants, since plants of even
the same species vary in their host requirement as a result of either their
genetic make-up or the variables in their environment. For example, thick

stands of the green root-parasite Euphrasia migima grown without hosts often

include a few flowering individuals, but plants grown singly rarely flower

106

(Heinricher, 1917). Apparently the parasites attack one another, and some
get more than they give up, thereby gaining an edge over their neighbors.
Chance, or else genetic variability among the parasites, determines which
individuals prosper. Nonetheless, regardless that individual root-parasites
can be induced under artificial conditions to mature without host-contact,
probably root-parasites as populations do require hosts. Under natural con-
ditions and taken as populations, it is likely that no species of root-
parasites can survive without hosts.

t h e g r e e n r o o t - p a r a s i t e s

The green root-parasites belong for the most part to the figwort and sand-
alwood families, reapectively the Scrophulariaceae and Santalaceae. To the
Scrophulariaceae alone belong almost 500 green root-parasitic species (Can-
non, 1909). The eastern North American genera of parasitic Scrophulariaceae
are Pedicularis, Gerardi , Melam rum, Eu hrasia, Odontites, Rhinanthus,
Orthocar us, and Castilleja. only Comandra of the Santalaceae ranges into
our area.

As a group, the green root-parasites include herbs, shrubs, and trees,
but most of the species native to North America are herbaceous. They rarely
look any different from.autotrophic plants, and the parasitism of some of
them went unsuspected for as long as 60 years after publication of the first
account of root-parasitism.

As might be suspected, root-parasites often are noxious agricultural
and silvicultural pests, and surprisingly, this is true of both green and
non-green root-parasites. In this country, a green root-parasite introduced
from Africa (witchweed, §t§ig§ asiaticg) seriously reduces corn production

in the Carolinas. In turn-about, however, plants of at least one genus of

107

green root-parasites are exploited commercially-——-—-—- the aromatic wood
of Sgntalum.albgm, the Indian sandal of commerce, is prized in the cabinet
and perfume trades (Bailey, 1951).

T H E P A R A S I T E

o n t o g e n y

The tolerance of Castilleja coccineg seedstock to two-year desiccation
is not shared by the seeds of Melam , a genus of root-parasites in the
same subfamily as Castilleja. Gislen in 19h? and Gautier many years earlier
(1908) noted that Melampyzgm seeds quickly lose viability when allowed to
dry out for even a few days. Seedstock of Melampyrum arvense dried for ten
days germinates at only a low per cent, and seeds kept dry for twenty days
die (Gislen, l9h9, and Heinricher, 1909). Using this fact, Gislen was
able to assert that figlgmpyggghdid not reach Sweden as a contaminant of
wheat in the early days of oceanic shipping, since the seeds surely would
have died in the weeks required for ocean transport.

The high per cent germination of unchilled Castilleda coccinea seeds
is in marked contrast to some other root-parasites, both native and Old
world. The seeds of Melamgyggm lineare, a green root-parasite common in
Michigan jack-pine vegetation, require 80-100 days of moist storage at AOC
to break what is really a double dormancy-—-———-about #0 days for radicle
dormancy and an additional 40 days for epicotyl dormancy (Cantlon gt_al;,
1962). Mbr60ver, only a third of each season's seed crop germinates after
the four months of cold-treatment. Under natural conditions the remaining
two-thirds of the seed crop presumably germinates in the second or third
years. In the laboratory these ungerminated seeds can be stimulated to

germinate by treatment with gibberellic acid, but not by leaching,

108

scarification, or treatment with various growth supplements (Curtis and
Cantlon, 1962).

Presence of a host is not required for germination of the seeds of
Castilleja coccine , but seeds of several other root-parasites do have
such a requirement. Absence of a host precludes germination of §triga

luteg (Brown and Edwards, l9Ah), S. hermonthica (Brown gt 51;, 1949),

 

Orobggphe mine; (Brown gt 91:, 1951), 0. speciosg (Chabrolin, 193A),

0. cgméga (Bartcinskii, 193A), Alectra vogelii (Botha, 1948), and several
other species. Striga asiatica, a native of Africa and since 1956 a serious
threat to corn yield in the Carolina states, also is a member of this group.
Using §§rig§_lutg§ and §tgigalhermonthica, Brown and his associates in En-
gland worked for several years first to substantiate the requirement and
then to discover what compound is involved in the germination stimulation.
Unfortunately, the stimulant has never been identified, but its character-
istics are well-established (Brown, 19A6). (1) It is active in hormonal

6 grams / liter. (2) The parasite seeds react

concentrations of 10"3 to 10-
to the stimulant with as little as 30 seconds exposure-———————-60 seconds
exposure induces a 70% germination level in Ordb che, compared with 0.1%
germination with no exposure. (3) The substances are not simple inorganic
compounds, nor are trace elements involved. (A) Probably a variety of
activators in nature stimulate the genuination of the various species of
parasites. There is no single universal stimulant. (5) No known vitamin
or phytohormone is involved. (6) The source of the activator is not
unique-——--———different tissues from various species produce it.

Although the natural stimulant still is not isolated and identified,

some compounds have been found which have the same effects. These include

109

D-xyloketose, thiourea, kinetin (6-(2—furfuryl) amino purine), and certain
other 6-substituted aminopurines (Brown gt él;, 1949, and Wbrsham gt,§l;,
1959). The last of these, the purine derivatives, have an effect beyond
germination stimulation. They substitute at least in part for host-
contact as well. §t§ig§ seedlings left in the natural stimulant genminate,
but never produce cotyledons and their shoot apices do not elongate, but
those kept in purine solutions do (Wbrsham.gt_al;, 1959).

Brown gt ah (19A9b) studied the physiological effects of sane of the
substances which induce §tgiga,gennination. Both the natural stimulant and
D—xyloketose produce up to 60% increases in the volume of root segments of
pea and corn. The volume increase in a result of cell extension, not divi-
sion. The concentration of D-xyloketose used was less than 10 ml / l, and
according to Brown is '...the first record...of a stimulating effect on the
growth of plant tissues...[by] a simple sugar in...[so low a concentration1'.

Root-parasites which require a host-secretion for germination are
thought simply to have lost the ability to produce by themselves enough of
some compound vital in germination. The evidence for this comes from work on
§§gig§ and grobggche. Brown and his colleagues (1946, 1951) noted that
moistening the seeds of Orobanche and §tgig§‘before treating them with the
host stimulant enhances their germination. The per cent germination in-
creases steadily the longer the seeds are given this pre-treatment, until,
after 1h days for Orobanche and 21 for Stri , there is no appreciable in-
crease in the per cent. After this Optimum length of pre-treatment, the
germination percentage of Orobanche seeds remains almost constant for at
least a year, but that of §tgig§ falls off noticeably. Brown and his co-
workers suggested that the seeds, once moistened, begin to produce by them?

selves a stimulant which is or is similar to the stimulant secreted from

110

host roots. As this natural stimulant builds up, less and less host stimu-
lant is needed to trigger germination. But eventually, for some reason,
the parasite seed stops accumulating its own stimulant. In Orobanche,
after 1h days the synthesis of the stimulant equals its breakdown, or else
synthesis stops and the stimulant already produced is protected from des-
truction. In Stri a, on the other hand, after 21 days the breakdown of
the stimulant occurs faster than its syntheSis.

Brown's idea suggests a manner in which root-parasitism.may have
evolved. A plant which produces haustoria has available for its use every-
thing its host translocates in conductive tissue. It is thus no longer
under selective pressure to maintain a working autotrophic metabolic system.
By genetic accident it can lose some functions of a normal autotroph and
suffer not at all. Even mutations that are ordinarily lethal might be per-
petuated. In this process, however, the root-parasite becomes more or less
dependent on its host. Orobanche and Stgigg apparently have lost only partly
their ability to produce a compound essential in genmination. This loss
makes them.host-dependent but not strictly parasitic, since what they re-
quire is a root secretion which the host normally 'discards'. They have
suffered other metabolic losses BS‘Wfill, however, losses which do make them
parasitic. Quite aside from.their germination stimulation, haustorial con-
tact with a host is necessary for their growth and maturation after germina-
tion. Castillejg coccinea requires no germination stimulant. Evidently in

 

this area of metabolism it is autonomous, and only in its maturation pro-
cesses is it host-dependent.
Root-parasites tend to be similar in their morphology. Systematists

recognize this, and have grouped the bulk of them into only a dozen families.

111

This morphological kinship of root-parasites makes sense if root-parasitism
does arise by genetic accident in root-grafting or haustoriumpforming plants.
Genetic systems which are alike enough to produce plants of similar morphol-
ogy doubtless are alike enough to suffer similar genetic accidents.

For that matter, it may be that the genetic system of any plant would
lose what Striga, Castilleja, or other root-parasites have lost, given the
chance-—-—-—- that is, given the ability to produce haustoria. To be sure,
root-parasites are characterized by more than the ability to produce haus-

toria their transpiration and guttation volume is large (Seeger, 1910),

 

the osmotic concentration of their tiesues is high (Solomon, 1952), they often
lack root hairs (Heinricher, 1917), and many of them germinate only when
stimulated by a host secretion (Brown, 1946). However, when root-parasites
first evolved, they may not have enjoyed these mechanisms, most of which
make them. more effective root-parasites. Instead, they may have acquired
such features after their root-parasitism.was firmly established genetically.
This seems very likely in the case of host stimulation of germination———————
without hosts the parasites can not survive beyond germination, and therefore
selection doubtless favors mechanisms for ensuring host-contact. §t§ig§ and
Orobanche have one such mechanism, the stimulation of germination by substances
diffusing from host roots. Host-specificity is the next step in this selec-
tion process. Apparently some hosts afford better parasite growth than do
other hosts, and therefore selection favors mechanisms for 'recognizing'
these hosts, such as a response to their characteristic root-secretions.

This, incidentally, illustrates the axiom that a mutation or new gene
combination varies in its selective value with the particular needs of the

plant that produced it. Probably seeds of many plants, not just root-parasites,

112

by mutation have become dependent on root-secretions for their germination.
But, whereas such a dependence is of definite survival value to a root-
parasite, it has no survival value or even is detrimental to autotrOphic
plants. Accordingly, selection would increase rapidly the frequency of
such a mechanism in populations of root-parasites but not at all in popu-
lations of autotrOphs.

If root-parasites have indeed lost some of the functions of autotrophs,
there is the possibility that a portion of their genetic code is not used.
That is, since a root-parasite has an external supply of vital compounds,
then its gene loci involved in the synthesis of such compounds can be
altered by mutations. Eventually, such loci might serve useful functions,
although to be sure it is naive to think that many of such freely-mutating
loci would be anything but detrimental. At any rate, the more a root-parasite
becomes dependent upon its host, the more loci are freed for coding new gen-
etic information, at least some of which could be useful to the parasite.
This might also explain why Castilleja varies widely in its morphology.

The ability to produce haustoria is the most striking feature of root-
parasites. The first step in the evolution of haustoria likely was a ten-
dency to fonn root-grafts. This speculation is supported by two observations.
First, a tendency to root-graft is common among plants. The roots of some
pines growing in stands graft freely, and, in fact, the living root systems
of otherwise dead trees are kept alive by photosynthate contributed by other
trees in the stand in a sort of host-parasite exchange (Bormann, 1957).
Presumably the advantage to the contributor-trees is the supply of nutrients
and water absorbed by the parasitic root systems. Moreover, some trees in
such stands dominate the bulk of their root-grafts with their neighbors,

and thus receive an inordinate supply of nutrients and water. As a result,

113

they enjoy better growth. This helps explain how’s few individuals in a
stand of geneticallybidentical trees can mature faster than their neighbors.
Second, man's extensive success with grafting suggests that there are
few barriers to the intimate union of unrelated plants. In this regard,
it is significant that all of the host-specific root-parasites exploit
hosts unrelated to them, and the root-parasites with a wide host range
attack members of many different families. For that matter, the inter-
action of scion and stock of grafted plants often is a sort of parasitism.
A flower bud of one plant grafted into a twig of another contributes noth-
ing but its genetic information, while both the support and the raw mater-
ials for its flowering and fruiting are contributed by the stock plant.
Granted that there is a widespread grafting capacity in plants, there

are three mechanisms for _effecting the grafts (MacDougal and Cannon, 1910)

 

(l) grafting of two roots growing close together, (2) grafting of one shoot,
or its adventitious roots, with another shoot, and (3) grafting of a stem
with the roots of a seedling lodged in its bark or in a wound. Root-grafting
doubtless is the most likely of the three mechanisms. Roots grow entangled
in the soil, often touching one another, and rarely are disturbed. In con-
trast, aerial contact is infrequent, and is subject to disturbance by
environmental forces such as wind. Successful aerial parasites either
twine about their hosts as dodder does, or produce adhesives for ensuring
firm host contact, as do the seeds of the Arceuthobium.mistletoes (Peirce,
1905). This suggests that most higher plant parasites are root-parasites
simply' because root-grafting is both more common and more stable than

other natural grafting mechanisms.

110

The selective advantage of vertical growth of the roots of Castilleja
coccinea seedlings lies in early contact with host roots. The sandy soil
that Castilleja coccinea typically grows in dries out to a depth of 1-2 cm
in the summer months. As a result, few roots grow close to the surface,
and, for a parasite seedling to attack a host, it must first reach the

root zone below the dry surface soil. In contrast to Castilleja coccinea,

Melampxn_1m_ lineare is not strongly geotropic in its early growth. And,
significantly, the site of germination of the seeds of Melampygum lineare

is moist jack-pine litter. The fine rootlets of jack-pine and other hosts
are protected from desiccation by the forest litter, and thus grow near
the surface of the soil well within reach of the elongating Melampyrgm
radicles.

The production of root hairs by Castilleja coccinea apparently is in
contrast to some European and Asian root-parasites (Heinricher, 1917, and
Stephens, 1912), which lack root hairs altogether. Melampygum produces
minute, cylindric appendages somewhat like root hairs (Sablon, 1887) but
called papillae or trichomes (plate 33, page 116). Similar structures have
been described for Lathraea (Heinricher, 1895), Cuscuta (Peirce, 1893),
Santalgm (Barber, 1906), and several other genera. Their function is not
known, although Kusano went so far as to call them hair-tendrils in A2312?
gtig (1908/1909) because '...[They]seem to attach...firm1y to the...[host
root] and then to coil or contract throughout...[their] whole length,
whereby the seedling is drawn closer to the host...'. Kusano noted the
tendril action only in Aeginetia, however, and speculated that the papillae
or trichomes of Melampyrum, Santalum, and Lathraea serve simply to attach

the parasite root to the host root. It does seem that on Melampygum lineare

_ f (
C.
N

’ O
[o-" '- ’
" -..‘ r.
‘1,” ,4) J
\u
I .2.

 

’3

p l a t e 3 3

Germinating seeds of Melamoyrum lineare. The root hair-like trichomes

 

are thought to secrete a substance which attacks the roots of nearby

vascular plants in the early'sfiages<1f haustorium penetration (heinribher,

1909).

116

 

a.

w...

 

 

 

 

 

 

117

seedlings they are most abundant at points on the root where haustoria are
developing. Melampygum lineare seedlings placed on moist elder pith pro-
duce a number of haustoria on the pith, and the papillae are observed eas-
ily with a dissecting microscope or hand lens. As well, in histologic
sections of Melampyrum lineare haustoria, the papillae show up as long
vacuolate cells. Their walls are continuous with the walls of the host
cells they touch. It thus appears that they secrete either a cementing
substance which effects firm contact with the host root, or a lysing sub-
stance which acts on the host root during penetration. Castilleja coccinea
haustoria in section also show papillae-like cells in addition to root
hairs, but the papillae are not obvious in living material. However, this
might be a result of the small size of both the papillae and the haustoria
of Castilleja coccinea.

When root-parasites first attracted attention at about the turn of
the century, lack of root hairs was considered good evidence that the hosts
afforded the parasites only water and mineral salts. Most of the root-
parasites are chlorOphyllous, and the only gross morphological differences
between them and autotrophs are the absence of root hairs and the presence
of haustoria. Thus, reasonably enough, they were presumed to be autotrophs
which had lost root hairs, the usual organs of absorption, and.instead pro-
duced haustoria to rOb other plants of water and mineral nutrients (Koch,
1889, Kostytschew, 192A, and Heinricher, 1917). Recent evidence (Hartel,
1959) suggests that the parasites do extract water and ions from their
hosts, but they do not stOp there, and take food substances and phytohor-

mones as well (Nelson and Rogers, 1959).

118

p h z s i o l o g y

The demonstration of starch in illuminated leaves and seedlings of
Castilleja coccinea suggests that the paraSite is photosynthetic. The
starch test for photosynthetic activity doubtless is a valid one, but
results of such tests on root-parasites must be interpreted with caution,
since the glucose or glucose-l-phosphate utilized by the parasite in
starch synthesis may be provided by host plants. To avoid this possibility,
every leaf used in the experiments was excised and floated on distilled
water in darkness for 2h hours to deplete its food reserves. Nonetheless,
the starch detected in the parasite leaves perhaps was derived from host-
provided reserves of sugar or its phosphorylated derivatives.

The conclusion from the photosynthesis experiments is that Castilleja
coccinea has its own photosynthetic machinery and uses it. However, since
haustorial transfer of fructose has been demonstrated, there is good reason
to suspect that the parasite, by plugging into the vascular streams of
other plants, obtains food materials it does not itself synthesize. That
is, Castilleja coccinea is photosynthetic, but it doubtless receives host
photosynthate as well. Whether or not the parasite ever uses the metabo-
lites it pirates is not yet established. There are three possibilities in
this regard'----(l) the parasite does not metabolize the food materials
it receives from its host, (2) the parasite's nutrition is supported by
photosynthate from both itself and its host, and (3) the parasite does not
metabolize its own photosynthate.

Heinricher (1917) thought that green root-parasites do not use the
food substances pirated from their hosts, but he did not support this pre-

sumption. When root-parasites first came under serious scrutiny, it was

119

possible to demonstrate transfer of only dyes across the haustoria. As a
result, it was speculated on purely anatomical grounds what the parasite
normally obtains from.its hosts. Those parasites that have both phloem
and xylem connections in their haustoria were thought to be dependent on
their hosts for food, water, and mineral nutrients, but those with only
xylem.connections were thought dependent only for water and mineral nu-
trients. For example, Peirce (1893) presumed that the mistletoe species
Eigggm_§l§um, because it has no sieve tubes in its haustoria, is a 'water-
parasite', and that '...its host performs for it only the functions of a
root--———-absorption, conduction, and mechanical support'. On this
basis, it would be concluded that Stgiga and Castilleja are not dependent
on their hosts for food materials, since neither has phloem elements in

its haustoria or phloem connections with its hosts (Stephens, 1912). Howe
ever, fructose readily crosses the haustoria of Castilleja, and according
to Rogers and Nelson (1959) 'sugar' crosses the haustoria of §tgig_. None-
theless, these experimental sugar transfers do not demonstrate that the
parasites normally obtain sugar from.their hosts. The details of the §§gig§
transfer have not yet been published, but in at least the Castilleja trans-
fer the labelled sugar was introduced into the host through a cut petiole.
As a result, the fructose surely had access to xylem tissue as well as to
phloem, and could have entered into the parasite across the wholly xylem
connections of the haustoria. Thus, it still is not known if Castilleja
coccinea takes food substrates from its hosts. The problem could be solved
by (l) demonstrating the presence of photosynthesis products in the xylem
of normal hosts, or by (2) introducing labelled sugar into only the phloem

of the host and noting whether transfer occurs. Of course, even if host

120

photosynthate is found to enter the parasite, there remains the question
of whether the parasite uses that photosynthate.
T H E H 0 S T 5

All the hosts of Castilleja coccinea tested are perennial or biennial.
However, since the parasite will attack the common radish, Raphanus sativus,
and corn, gg§,m§y§, it is at least likely that annuals serve as suitable
hosts. Also, the hosts need not be matures----seedlings of Tragopogon

ratensis, Lactuca canad nsis, Danthonia s icata, and Kalanchog verticillata
are attacked readily.

Melampygum in Europe (Heinricher, 1909) apparently only rarely attacks
annuals in artificial culture. In contrast, however, Orthocarpus purpuras-
Eggs, a root-parasite indigenous to the American deserts, parasitizes mostly
annuals (Cannon, 1909). The parasite itself is a desert annual, and germin-
ates and matures only when rain is sufficient. The many other desert annuals
with the same rain-dependence are the most abundant hosts in the area, and
thus are the most often attacked.

In the greenhouse Castilleja coccinea successfully parasitizes a variety
of hosts, but a few hosts appear to support its growth better than others do.
The reason for this is not yet clear, although it is suspected that the best
hosts are simply those with the densest root systems. For example, Lactuca

canadensis doubtless is the best host of the twenty of so tested, and it

 

produces the most massive and branched root system. Whatever the reason for
the parasite's varying growth on different hosts, almost all the plants
tested supported the parasite to maturity. Thus, it seems reasonable to

assert that Castilleja coccinea can parasitize almost any higher plant it

 

normally occurs with. Moreover, the parasite attacks plants which never

occur in the same area with it. For example, Castilleja coccinea will

121

parasitize Kalanchog verticillata, a succulent greenhouse herb native to
South Africa, and legg§_blum§i, a native of Java.
THE HOST-PARASITE CONNECTION

The mechanism by which root-parasites extract host materials is not

host root

 

fully understood. There are at least three possibilities
pressure, osmotic absorption by haustoria, and transpiration pull (Hartel,
1959). Each has been championed as the sole mechanism for the extraction
process, but seeming exceptions inevitably have been cited. Nonetheless,
(1) most plants maintain positive root pressure, (2) the tissues of a root-
parasite typically have a higher osmotic concentration than that of its
hosts, and (3) the transpiration volume of many root-parasites is unusually
high.

osmotic absorption. MacDougal and Cannon (1910) studied in an unusual way
the osmotics of host-parasite pairs. They pierced the succulent stems of
cacti and slipped in cuttings of various autotrophs to serve as 'parasites'.
None of the 'parasitic' plants maintained itself against a host of higher
osmotic concentration. In contrast, those with a favorable osmotic differ-
ential usually were successful parasites, and some of them survived for two
years.

Solomon (1952) studied the osmotics of an actual root-parasite and its
hosts. He used Strig§,parasitic on sorghum, which was grown in sand and
watered with various concentrations of nutrient solutions. As expected,
he found that the osmotic value of the sorghum increases as the osmotic
value of its culture solution increases. But, in contrast, that of the
parasite is high to start with, and increases only slightly at very high

nutrient concentrations. At the highest concentration the osmotic values

122

of both parasite and host are about equal at 9 atmospheres. But at the
lowest concentrations, the osmotic value of the parasite exceeds that of
the host by nearly 8 atmospheres. And, interestingly, the yield of the
sorghum at the highest concentration is about the same attacked and unat-
tacked by the parasite, but at lower concentrations, the yield drops off
in the attacked plants. Solomon reasoned that, at the low nutrient con-
centrations, because there is an 8-atmosphere difference in osmotic pres-
sure between the host and the parasite, the parasite is able to deprive
the host of large quantities of water and thus reduce its yield. In con-
trast, at the high concentrations of nutrient solution, when the osmotic
pressures of host and parasite are practically the same, the parasite can
extract almost nothing from the host, and in fact leads a precarious exis-
tence. The sorghum plants are not always attacked under such conditions,
and the Sppigg plants which do attack die early, often without flowering.
transpiration pull. ngtel (19Al) has studied what he calls the stress
on the water economy of the host caused by the parasite's transpiration
pull. He measured in the field the daily transpiration patterns of para-
sites (Pedicularis foliosa) and of both attacked and unattacked hosts
(Aygpg caucasica). In graphing the results, he noted no marked difference
in the transpiration patterns of the parasites and the unattacked hosts,
but he did note a difference between those of the attacked and the unat-
tacked hosts. The daily transpiration level of the parasitized host drops
markedly in the afternoons. This drop in transpiration Hartel attributed
to the effect of the parasite's transpiration pull. He asserted that the
parasite pumps enough water out of the host to cause a passive closure of

the host stomates, drastically reducing the transpiration volume of the

123

host. As a result of the stomate closure the host suffers a decrease in
gas exchange and a corresponding decrease in photosynthesis. But, suppos-
edly, the host compensates for this loss of photosynthate by an increase
in photosynthesis during the early-morning hours of each day, before the
parasite can draw off enough water to close the host's stomates.

Seeger (1910) compared the transpiration levels of two root-parasites
and six.autotrophs. According to his data, both root-parasites transpire

more heavily than any of the autotrOphs tested

'...a piece of leaf of Epphpapia ppstkoviana or Odontites verna
transpires five-fold that of Nuphar luteum (emergent leaf), six

to seven times more than Gentiana, Callisia, Lamium, and other
mesophytes, forty times more than Rhododendron (xerophyte). And,
Veronica chamaegpys, also a member of the Scrophulariaceae, loses
only one-third as much water in transpiration'. [German translation]
The functions of the four portions of the haustorium of Castilleja
coccinea and other root-parasites are only partly understood. The central
vascular trace doubtless conducts host materials into the parasite, but
the function of the densely-staining parenchyma around the vascular trace

is not known. The cortical rind presumably protects the haustorial con-

nection. The pad of columnar cells, at least in the haustoria of Cuscuta,

 

'was thought by Peirce (1893) to '...exude...a solvent which attacks and
dissolves the walls and contents...of the cortical and epidermal cells...
[of the host]'. Similar structures occur in the haustoria of most root-
parasites. In Castilleja coccinea and Melampyzpm lineare haustoria the
walls of the pad cells appear bonded to the host (plate 3', page Vi), which
does suggest lysing or softening of the host cellso The cells of the host
root rarely are distorted next to the haustorium, suggesting that they are
not forced aside or compressed by the penetrating haustorial tissueo This

is the best evidence that the haustoria lyse rather than force access to

124

the stele of the host. Apparently the elongating wedge of haustorium tis-
sue merely fills a cavity in the host root formed by the breakdown of the
tissue lying in its path.

The factors inducing the formation of haustoria by root-parasites are
largely unknown. Heinricher (1897, 1931) and Sablon (1887) asserted that
the parasites produce haustoria only in the presence of living roots. Other
workers have suggested that physical contact alone induces haustorium for-
mation. Barber (1907) found a sandal tree (Santalum alppm) root which
'...had made five attempts to penetrate a refractory particle of quartz'.
Dodder (Cuscuta) haustoria can be induced experimentally by pressing the
dodder shoots in tinfoil (Knapp, 1954). Peirce as early as 189A induced
dodder haustoria simply by locally irritating the shoots. But, he noted
that a haustorium formed in this way does not mature unless it soon obtains
nutrients from whatever it penetrates. The developing haustorium is not
sustained by food transported from some other portion of the dodder plant.
Such a mechanism has obvious selective advantage. Only those haustoria
which contribute to the parasite's nutrition ever mature, '...an economy
of materials and of energy' (Peirce, 189A).

The roots of Castilleja coccinea plants grown behind glass in glass-
blotter sandwiches do not visibly grow toward nearby host roots. The lack
of a chemotactic mechanism for locating host roots would seem disadvantag-
eous to root-parasites, but the density of roots is high in many soilso
No chemotactic mechanism has been demonstrated for even the host-specific
root-parasites. In fact, Orobanche is stimulated to germinate by a host
substance diffusing as far as a centimeter from its host, and yet the

seedling radicle can elongate only a scant two millimeters (Sunderland,

125

1960). As a result, many seedlings die because they are unable to elongate
enough to attack the very root which stimulated their germination. But even
though root-parasites do not locate host roots chemotactically, it is sig-
nificant that most parasites with a wide host-range genninate independently,
while host-specific parasites usually require a host secretion for germin-
ation.

T H E H 0 S T - P A R A S I T E I N T E R A C T I O N

h o s t - r s q u i r e m e n t

The inability of Castilleja coccinea to grow beyond the seedling stage
without host-contact is shared by only some root-parasites. Others, includ-
ing a feW'Melampypum species, are able to mature in the absence of a host
(Heinricher, 1909). Usually such plants are stunted, discolored, and sensi-
tive to drought, but they nonetheless flower and set seed. Thus, it is ob-
vious that some green root-parasites are less dependent on their hosts than
Castilleja coccinea is.

The results of divorcing Castilleja coccinea from its hosts demonstrate
that the parasitic attachment is not required for bolting once the rosette
has reached maturity. However, flowering in rosette-forming plants such as
Castilleja coccinea involves two processes, the bolting of the flowering
shoot and the formation of floral primordia. It is not yet known if the
host is required for initiation of floral primordia, but it is likely since
rosettes which are divorced from their hosts before they are mature typically
cease apical growth, and eventually suffer apical necrosis.

Although the rosettes can flower detached from their hosts once they
are mature, there is no evidence that they are detached in the wild during

their flowering period. To be sure, some of the haustoria of a rosette

126

likely plug up during the winter months, and some host roots or even whole
hosts die, and as well some parasite roots die. Thus, unless new haustoria
are formed in the second season of growth, a year-old rosette has less host-
contact than does a rosette about four months old. Nonetheless, bolting
rosettes doubtless enjoy extensive host-contact.

The selective advantage of the ability to bolt without a host is not
clear. Of course, if the hosts were all annuals, such a mechanism would
be essential for the survival of a biennial parasite such as Cgstilleja
coccinea. But, the fact is, few of the hosts of Castilleja coccinea, at
least in the Houghton Lake area, are annuals. Polygala sanguinea and
Trifolium agrarium are the only abundant annuals in the vegetation. None
of the four or five dominants of the vegetation is an annual. However, it
is likely that some of the roots of the perennial hosts die back during
the winter months. Any advantage the parasite gains in being independent
during its second year may derive from this die-back of host roots and the
resulting loss of functional haustoria.

It would be interesting in this regard to study the perennial Castilleja
species so common in boreal regions and western North America. For three
reasons it is unlikely that these perennial Castilleja are independent of
their hosts even after they reach maturity (assuming that they are parasitic).
(1) The adult plants of Pedicularis and Gerardia, two genera of root-parasites
related to Castilleja, maintain many haustoria which in microsc0pic section
appear functional. (2) The haustoria of Comandra and Gerardia attain a
diameter of l-2 cm, suggesting that they grow and function for several years.
(3) Mature perennial Castilleja is notoriously hard to transplant, even into

pots of its native soil. If perennial Castilleja are indeed host-dependent,

127

they likely maintain host-contact either by a spring burst of root-growth
and haustorium production or by the yearly enlargement of existing haus-
toria in pace with the growth of host roots.
i o n u p t a k e

Beath gt_§l;,(l9hl) asserted that at least one species of Castille‘a,
Castilleja chromosa, accumulates minerals independently of its host plants.
Beath and his colleagues were interested in Castilleja as an indicator
plant in seleniferous areas in the western states. They assumed that all
Castilleja plants are root-parasitic, and could contain only as much
selenium as their hosts. They found, however, that their samples of gas:
tilleja chromosa parasitic on Chrysothamnus pumilis contained 258 ppm of

selenium but the Chrysothamnus only 1 ppm. Moreover, the converse was

 

true they got a low selenium count for Castilleja chromosa parasitic
on Astragalus bisulcatus, normally a selenium-accumulating species. Their
conclusion was that '... astilleja chromosa can absorb selenium from a raw
seleniferous shale regardless of the kind of host plant with which it is
associated'.

Beath and his co-workers assumed that Castilleja can not selectively
‘accumulate selenium.from its hosts. However, this is only an assumption.
Probably the haustoria impose no permeability barrier to most substances
in the vascular system of the host, since the connection between host and
parasite in histologic section appears to be a xylem-to-xylem contact.
Nonetheless, the parasite tissue may be selective in its use of the host
materials which reach it. Thus the lack of correspondence in selenium con-

tent of host and parasite may indicate (1) as Beath suggests, independent

absorption by Castilleja or (2) selective use by the paraSite of whatever

128

host substances reach it through its haustoria. At any rate, Beath has
shown that the selenium content (not necessarily the seleniumrabsorbing
capacity) of Castilleja chromosa plants is independent of whatever hosts
they attack.

d a m a g e t o t h e h o s t

In any discussion of parasitism, there arises the question of damage
to the host. It seems only logiCal that a plant suffers a lower rate of
growth when supporting a paraSite. In theory, the parasitism does no
damage only if the host has more than it needs of whatever the parasite
takes. Certainly it would be to the advantage of the parasite not to kill
its hosts or even to impair their vigor, especially if the parasite is a
perennial.

Some plants enjoy 'luxury uptake' or mineral nutrients, and often
have apparent 'excesses' of compounds such as free amino acids. A root-
parasite likely would not damage its hosts by extracting such materials.

If root-parasites do pirate only 'excess' host compounds, they are exploit-
ing a source of energy and materials not available to other higher plants
except as root secretions. However, it could be argued that such excess
host compounds are available only at certain times in the host's life
cycle. Unless the parasite restricts its extraction to those periods,

it must extract more than just 'excess' host compounds.

The question of host damage will be answered satisfactorily only when
it is discovered what the parasite takes from its hosts, but the prOblem
can be studied in other ways until that information is obtained. ‘What is
needed is a statistical analysis of the performance of pairs of host plants,
one plant of each pair attacked and the other not. A significant difference

in, say, plant height or number of leaves between the attacked and unattacked

129

hosts would indicate that the parasite does lower the vigor of its hosts.

It would be ideal, of course, to use genetically-identical hosts for such

a study, and, conveniently, this is possible---- Castilleja coccinea
‘will parasitize Kalanchog verticillata, a greenhouse herb which asexually
produces hundreds of geneticallyeidentical plantlets along the margins of
its leaves. This host has another advantage in that Castilleja coccinea
‘will parasitize the plantlets when they are not much larger than the para-
site itself, and thus any drain the parasite imposes on them will be obvious.
An analysis of host-damage using Kalanchog and Lactuca canadensis as hosts
will be made in growth-control chambers during the winter of 1962-1963.

A crude analysis has been made of the attrition and loss of Lactuca
canadensis under heavy attack by Castilleja coccinea. The results suggest
that the host suffers damage only when the parasites are abundant. However,
such a heavy attack by Castilleja coccinea rarely occurs naturally. Not
only are the rosettes scattered in the field, but also each rosette typically
attacks several plants around it‘-—-—-—-this is demonstrated readily by ex-
cavating the rosette's root system. Thus, the attack of a rosette is divided
among several hosts, and presumably no one host suffers unduly. The results
of artificial host-damage experiments must be evaluated with this in mind.

It is likely that given a very heavy parasite attack, any host plant would
suffer a lower rate of growth.

Although Castilleja coccinea typically does not subject its hosts to
heavy attack, some root-parasites do, and inflict major losses in the yield
of crops. An alarmed California farmer wrote to J. Burtt Davy in 1898

'You will find enclosed a branch of a weed [Orthocagpus pusillus

Benth.:l which has lately made its appearance in our valley lands,
and it appears to take possession, and wherever it comes the grass

130

disappears, but still it does not seem to grow thick enough to choke
out the grass... Many of our fields that have always been very pro-
ductive of clover and also rye-grass are being covered with this,
and at a distance it looks like moss...'.

 

Alectra vogglii causes 'serious damage to leguminous cr0ps...of the
Transvaal and Southern Rodesia' (Botha, 1948). §§§ig§,wherever it becomes
established reduces considerably the yield of its corn and sorghum hosts.
ngtel (1941) studied the carbohydrate content of alpine grasses attacked
and unattacked by Pedicularis. He found no difference in carbohydrate con-
tent of attacked and unattacked plants. The raw fiber content of the hosts
similarly does not change during parasitic attack. Hartel did note, however,
that the attacked plants wither earlier in the season than unattacked plants
do.

In short, the evidence indicates that hosts of root-parasites, inclu-
ding Castilleja coccinea, accrue no unusual damage unless the parasites are
abundant. But, unfortunately, quite a few species of root-paraSites do
build up heavy populations and become pests, especially on agricultural land.
p a r a s i t e m.i c r o - d i s t r i b u t i o n

The environment of any plant is at least partly biotic, but the environ-

ment of a root-parasite is largely biotic the parasite's vascular

 

stream is continuous with the vascular streams of best plants. To be sure,
any environment can be defined in physical and chemical tenns, and the dis-
tinction erased between biotic and abiotic. Thus it is immaterial to the
plant whether a given environmental factor is biotic or abiotic. But, taken
as a whole, the environment of a root-parasite, because it is largely biotic,
surely is different from that of an autotnoph. Thus, the variables of a

root-parasite's environment are not all the same as those of an autotroph's

131

environment. For example, water uptake by a root-parasite is affected not
only by soil moisture, but also by the uptake capacity of host plants. In
short, the host is a major factor in the environment of any root-parasite.

The distribution and abundance of a plant species are results of the
interactions of the genetic make-up of the individuals of that Species and
the many variables of the environment. Where the rigor of a few or even
one environmental variable exceeds the species' geneticallybfixed capaci-
ties, the species is not successful. This is as true for micro-distribution
and abundance as it is for, say, continental distribution and abundance,
although environmental variables tend to vary less over small surfaces.
As yet, no plant is known so well that its success can be predicted under
simulated environments. In fact, few are well known to predict their per-
formance under varying regimes of the most important environmental factors
of light, moisture, and temperature. Moreover, the detailed pattern of
these factors in the environment through time and in space is poorly known.

Because of this ignorance of what factors in the environment really
control a species' field performance, and how those factors are distributed,
it is wOrthwhile to study the micro-distribution of a root-parasite such as
Castilleja coccinea. Castilleja coccinea is known to require in its environ-
ment a host or hosts, and the location of those hosts can be plotted accur-
ately. Thus here is a plant, one of whose environmental requirements is
clear-cut and also has an obvious pattern in the environment.

In greenhouse culture the plants of some host species support the
growth of Castilleja coccinea better than others do. Thus, the hosts in
a given community can be ranked according to their ability to suppcrt the

parasite. Then, on this basis, one can try to explain the field distribu-

tion of Castilleja coccinea in that community. However, the micro-distribution

of Castilleja coccinea in one such community in the Houghton Lake area did

 

not reflect this host ranking. The reasonsfor the failure might be several
(1) compared with other factors which influence the distribution of Qééiilf
lgja_coccinea, host-suitability might be insignificant, (2) the sample size
may have been too small, (3) the estimations of root-spread of the hosts, an
assumption made in the study, may have been in error, and (A) at least some
flowering (secondeyear) parasites were included in the study, and the hosts
recorded for them in reality may not have been hosts, since bolting rosettes
do not require host-contact. In this last case, the hosts which nourished
the parasite might have died previous to the study. If other, nonphost plants
took their place, spurious data were collected in the study, and even if noth-
ing took the place of the real host, that host nonetheless went undetected.
Further studies will be made of this question of the effect of hosts on para-
site micro-distribution. One approach will be experimental-—-——-—-the para-
site will be grown in controlled environments, and its survival noted for

various combinations of hosts and various host densities.

133

C 0 N C L U S I 0 N S

l. Castilleja coccinea is biennial in its develOpment, at least in the
Houghton Lake area of Michigan. Seed germination and maturation of the
rosettes occur during the first season, and bolting of the rosettes,
flowering, and fruiting occur in the second season.

2. Seed germination requires light, moisture, and moderate temperatures.
Presence of a host is not a germination requisite, nor do host roots
influence the direction of growth of the parasite roots.

3. Production of haustoria by parasite roots is not solely a result of
physical contact with an object or with living tissues.

A. The xylem of host and parasite are connected through the vascular
trace in each haustorium. Eosin Y, Gin-fructose, S3504, and PBZOA will
cross the haustorial connection into the parasite, and presumably'most
or all the substances transported in the host xylem cross as well.

5. The parasite apparently has its own photosynthetic machinery and
uses it.

6. Nonetheless, successful penetration of host roots is necessary for
growth of the parasite beyond the seedling stage.

7. Thus, it is likely that the host compound (or compounds) most vital
to the parasite is a phytohormone or enzyme prosthetic group rather than
a respiratory substrate. This compound is not one of the more common
vitamins or auxins.

8. The parasitic attachment is not necessary for bolting of the rosette,
but under natural conditions bolting rosettes do enjoy extensive host-
contact.

9. The parasite has a wide host-range, and will attack plants which do

134

not occur in its natural range.

10. However, the hosts do not support the parasite equally well, and on
the basis of greenhouse studies the hosts in a given community can be
ranked according to their ability to support the parasite.

11. This ranking of hosts is not reflected in the field micro-distribution
of the parasite in the community.

12. The hosts accrue little damage in the host-parasite interaction except

when under heavy parasite attack.

135

L I T E R A T U'R E C I T E D

 

Bailey, L. H., 1951, Manual of Cultivated Plants, The MacMillan Company,
New'York.
Barber, C. A., 1906, Studies in root-parasitism, the haustorium of

Santalum album, part 1. early stages, up to penetration, Mem.

Dept. Ag. India, Bot. Ser. 1(1)I:1-3o.

 

, 1906, Studies in root-parasitism, the haustorium of
Santalum album, part 2. the structuie of the mature haustorium
and the inter-relations between host and parasite, Mem. Dept. Ag.

India, Bot. Ser. l(l)II:l-58.

 

, 1907, Parasitic trees in southern India, Proc. Cam.
Phil. Soc. 14:246-256.

Bartcinskii, 3., 1934, c. R. Acad. Sci. U.R.S.S. 1:343 (cited in
Knapp, R., 1954).

Beath, C. A., C. S. Gilbert, and H. F. Eppson, 1941, The use of indicator
plants in locating seleniferous areas in western U.S.A. IV. progress
report, Amer. J. Bot. 28(10):887e900.

Bormann, F. H., 1957, Moisture transfer between plants through intertwined
root systems, Plant Phys. 32:48-55.

Botha, P. J., 1948, The parasitism of Alectra vogelii Benth. with special
reference to the germination of its seeds, South Afr. J. Sci. 44:119.

Brown, H., and M. Edwards, 1944, The germination of the seed of §tgig§
lutgg, I. host influence and the progress of germination, Ann. Boto
N.s. 8:131-148.

Brown, H., 1946, Biological stimulation in germination, Nature (London)

157(3977):64-68.

(J?

136

Brown, R., A. W. Johnson, E. Robinson, and A. R. Todd, 1949, The stimulant
involved in the germination of Stgigg hermonthica, Proc. Roy. Soc.
London B 136:l~12.

Brown, R., E. Robinson, and A. W. Johnson, 1949b, The effects of D-xyloketose
and certain root exudates in extension growth, Proc. Roy. Soc. London
B 136:577-591.

Brown, R., A. D. Greenwood, A. W. Long, and G. J. Tyler, 1951, The stimulant
involved in the germination of Orobanche ming;_Sm. 2. Chromatographic
purification of crude concentrates, Biochem. Jour. 48(5):564-568.

Cantlon, J. E., E. J. C. Curtis, and W. M. Malcolm, 1962, manuscript submitted
to Ecology.

Chabrolin, C., 1934, La germination des graines d'Orobanche, Compt. Rend.
Acad. Sci., Paris 198:2275-2277.

Curtis, E. J. C., and J. E. Cantlon, 1962, manuscript.

Davy, J. B., 1898, Parasitism of Orthocarpus usillus, Erythea 6:93.

Fernald, M. L., 1950, Gra '8 Manual of Botan , Eighth Edition, American Book
Company, New York.

Gautier, L., 1908, Sur la parasitisme du Melampygum ratense, Rev. gen. Bot.
20:67-84.

Gislen, Torsten, 1949, Problems concerning the occurrence of Melampygum
arvense in Sweden, Oikos 1:208-234.

Gleason, H. A., 1958 (second printing), The New Britton and_§rown Illustratgd

Flora of the Northeastern United States and Adjacent Canada, New York

Botanical Garden.

 

Hald, A., 1958 (second printing), Statistical Tables and Formulas, Johaniley

and Sons, New York.

137

ngtel, 0., 1941, fiber die 0kologie einiger Halbparasiten und ihrer Wirts-

pflanzen, Ber. dtsch. bot. Ges. 59:136-148.

 

, 1956, Der Wasserhaushalt der Parasiten, Handb. d. Pflanzen-

phys. 3 (Pflanze und Nasser):951-960.

 

, 1959, Der Erwerb von Nasser und Mineralstoffen bei Hemi-
parasiten, Handb. d. Pflanzenphys. ll (Heterotrophie):33-45.
Heinricher, E., 1897, Die grfinen Halbschmarotzer I. Odontites, Eu hrasia,

und Orthantha, Jahrb. f. wiss. Bot. 31:77-124.

 

—~ , 1909, Die grfinen Halbschmarotzer V. Melam um, Jahrb. f.

wiss. Bot. 46:273-376.

 

, 1917, Zur Physiologie der Schmarotzenden Rhinantheen,

besonders der Halbparasitischen, Naturwissenschaften 5:113-119.

 

( , 1931, Mono ra hie der Gattun Lathraea, Jena.

Knapp, R., 1954, Experimentelle Soziologie der hgheren Pflanzen, Ulmer,
Stuttgart.

Koch, L. von, 1889, Zur Entwicklungsgeschichte der Rhinanthaceen (Rhinanthus
m_i_no_r Ehrh.), Jahrb. r. wiss. Bot. 20:1-37.

Kostytschew, S., 1924, Untersuchungen fiber die Ernghrung der grgnen Halb-
schmarotzer, Beih. bot. Abl., Abt. 1, 40:351-373.

Kusano, 8., 1908/1909, Further studies on Aeginetia indigg, Bull. Coll. Ag.,
Tokyo Imp. Univ. Japan 8:59-77.

MacDougal, D. T., and W. A. Cannon, 1910, The conditions of parasitism in
plants, Car. Inst. wash. Pub. 129.

Peirce, G. J., 1893, On the structure of the haustoria of some phanerogamic

parasites, Ann. Bot. 7:291-328.

*‘7
V’

138

Peirce, G. J., 1894, A contribution to the physiology of the genus Cuscuta,

Ann. Bot. 8:53-118.

 

, 1905, Dissemination and germination of Arceuthobium
occidentale, Eng., Ann. Bot. 19:99-114.

Pennell, F. W., 1935, The ScrOphulariaceae of eastern temperate North
America, Monog. Acad. Nat. Sci. Phila. 121-650.

Rogers, W. E., and R. R. Nelson, 1959, Studies on the translocation of
minerals and carbon compounds between corn and §tgig§ asiatica, Jour.
Elisha Mitch. Sci. Soc. 75(2):67 (abstract).

Sablon, Leclerc du, 1887, Recherches sur 1es Organes diAbsorption des
Plantes Parasites (Rhinanthacees et Santalaceae), Ann. Sci. Nat.

Bot. Ser. 7(6):90—117.

Shive, J., 1915, Athree-salt nutrient solution for plants, Amer. Jour.
Bot. 2:157-160.

Seeger, R., 1910, Versuche fiber die Assimilation von Euphrasia (5.1.) und
fiber die Transpiration der Rhinantheen, Sitzber. Kais.
Akad. Wiss., Math.-Nat. Klasse 119:987-1004.

Solms-Laubach, H. Grafen zu, 1867-1868, Ueber den Bau und die Entwicklung
der Ernghrungsorgane parasitischer Phanerogamen, Jahrb. f. wiss. Bot.
6:509-638.

Solomon, 8., 1952, Studies in the physiology of phanerogamic parasitisnxwith
special reference to §£§i£§.i§£§é Lour. and S. densiflora Benth. on
Andropogon sorghum Hack. l. the osmotic pressure of the host and para»
site in relation to the nutrition of the host, Proc. Indian Acad. Sci.,

seCto B 35:122‘1316

139

Stephens, E. L., 1912, The structure and deve10pment of the haustorium of
Stgiga lutea, Ann. Bot. 26:1067-1076.

Sunderland, N., 1960, Gennination of the seeds of angiospermous root
parasites, The Biology of weeds, a symposium of the British
Ecological Society, Blackwell Scientific Publications, Oxford.

Vallance, K. B., 1951, Germination of seeds of Bartsia Odontites, Nature
167:732.

‘Worsham, A. D., D. E. Moreland, and G. C. Klingman, 1959, Stimulation of
§trig§_asiatica (witchweed) seed germination by 6-substituted purines,

Science 130(3389):1654-1656.

     
       

 
 

NNANu

3

ml