FACTORS CONTRIBUTING TO THE HERBICIDAL
RESISTANCE OF BLACK TITI (Cliftonia monophylla)
WALTER L. BEERS, JR.
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Barbara Merrill Beers
and to the
W. D. "Bill" Smith
former Woodlands Manager, The
Buckeye Cellulose Corporation
The author acknowledges with sincere thanks the advice,
comments, and assistance of Dr. S.H. West, Dr. T.E. Humphreys,
and Dr. M. Wilcox given during this work. Committee service
of Dr. D. McCloud and Dr. R.G. Stanley is likewise appreci-
Acknowledgement is also made to Mrs. Grace B. Colson,
Mr. W.L. Dickerson, Mr. Karl Jurbergs, Mr. J.E. Morris,
Mr. Vasco Murray, Mr. G. Olree, and Dr. C. Schenker, all of
The Buckeye Cellulose Corporation, for their help and encour-
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . .
LIST OF TABLES. . .
LIST OF FIGURES . .
INTRODUCTION . . . . .. . ..
LITERATURE REVIEW . . . . . .
MATERIALS AND METHODS . . . . .
Anatomical study of leaves. . . .
Electron microscopy of leaf surfaces.
Leaf absorption of 2,4-D-1-14C. .
Metabolic degradation of 2,4-D-1-14C.
Preliminary bell jar experiment .
Absorption train experiment . .
EXPERIMENTAL RESULTS . . . . .
Anatomical study of leaves. . . .
Electron microscopy of leaf surfaces.
Leaf absorption of 2,4-D-1-14C. . .
Metabolic degradation of 2,4-D-1-14C.
Preliminary bell jar experiment .
Absorption train experiment .
DISCUSSION . . . . . . .
SUMMARY AND CONCLUSIONS . .. . .
LITERATURE CITED . . . . . .
. . . 1
. . .
* 4 *
* . .
LIST OF TAB3!ES
1. Major anatomical difference between (S) and
(R) titi leaves. . . . . . . 37
2. Analysis of 2,4-D-1--14C absorption by (S) and
(R) titi leaves and significance of the
differences. . . . . . . . 44
3. Results of preliminary bell jar experiment in
which evolved '"CO2 was absorbed in potas--
sium hydroxide solution . . . .. 46
4. Uptake and degradation of 2,4-D--1-4C by de-
tached shoots of (S) and (R) titi during
approximately 140 hours in the dark. . 47
5. Data for computing standard deviation from
the curvilinear regression equation Y
-15.7677 + 28.8199X + (-3.4952) X2 . . 50
6. Data for computing standard deviation from
the linear regression equation Y = 14.2043
+ 4.2602X. . . . . . . . . 51
7. Comparison between mean squares for linear
and curvilinear regression---(R) titi decar-
boxylation data. . . . . . ... 52
8. Analysis of variance for curvilinear regress-
ion--2,4-D-1-14C uptake and breakdown data
for (R) titi trials in the absorption
train . . . . . . . . 53
LIST OF FIGURES
1. Foliage and fruit of black titi and white
titi . . . . . . . . . 2
2. White titi selectively controlled amid re-
sistant black titi, following aerial
treatment with 2,4-D ester applied at the
rate of 6 pounds active ingredient in 5
gallons of spray per acre. . . . .. 4
3. Absorption train used in 2,4-D-1-14C metab-
olism experiments with (S) and (R) titi
shoots . . . . . . . . .27
4. Mass correction curve for barium carbonate
precipitates . . . . . . 29
5. Leaf cross-section of (S) titi (Cyrilla race-
miflora) at mid-rib, showing cuticle (C),
epidermis (E), bundle sheath extension
(BSE), internal phloem (IP), xylem (X), ex-
ternal phloem (EP), palisade mesophyll (PM),
spongy mesophyll (SM), and stoma (S) . 32
6. Leaf cross-section of (R) titi (Cliftonia
monophylla) at mid-rib, showing cuticle (C),
epidermis (E), bundle sheath extension
(BSE), xylem (X), phloem (P), palisade
mesophyll (PM), and spongy mesophyll (SM).
Note the absence of sheath extensions on
the secondary vein to the right of the
main vascular bundle . . .. . . 34
7. Leaf cross-section of (S) titi (Cyrilla race-
miflora), showing secondary vein with
bundle sheath extensions (BSE) leading to
upper and lower epidermis, xylem (X), and
phloem (P) . . . . . . . . 35
8. Electron micrograph showing the smooth adax-
ial surface of an (S) titi leaf . . 39
9. Electron micrograph of the adaxial surface of
an (R) titi leaf, showing a very rough,
ridgy surface. . . . . . . .. 41
10. The dry residue of 10 ul drops of 2,4-D-1-14C
containing 1 per cent Tween 20, applied to
the center of the adaxial surfaces of (S)
and (R) titi leaves . . . . .. 42
11. Cumulative counts per minute evolved as Bal4C03
by (S) and (R) titi during the indicated
dark periods . . ... . . . 48
12. The relationship between 2,4-D (2,4-D-1--14C
uptake and breakdown (decarboxylation) by
(R) titi shoots during approximately 140
hours in the dark . . . . .. .... 54
Black titi cliftonia monophvlla (Lam.) Britt. is a
multiple-stemmed, semi-deciduous shrub that occurs naturally
in the wet lowlands of the Southern Coastal Plain ranging
from West Florida to Texas. The eastward extension of black
titi in Florida appears limited by the Aucilla River in
Jefferson County. A similar plant, white titi (Cyrilla
racemiflora L.) intermingles with black titi over its range,
but extends throughout North Central Florida and beyond,
into Southern Virginia. Both species belong to the family
Cyrillaceae. Where they overlap, floral differences easily
distinguish the two. Black titi bears a panicled raceme,
and white titi, a spike raceme. The persisting fruit
affords another means of identification. Black titi pro-
duces clusters of winged, nut-like drupes, while white titi
develops small globose capsules born on slender spikes
Because of the dense cover it creates, black titi must
be eradicated, or at least severely retarded, prior to re-
forestation of wet lowlands by seeding or planting. Titi
can be top-killed by fire, but vigorous resprouting in-
Fi.. 1. Foliage and fruit of black titi (top) and white
variably occurs from adventitious buds at the root collar.
Although reforestation is possible following a fire, re-
surging titi in the pine understory constitutes a contin-
uing fire hazard to developing plantations. Also, the
brush cover would doubtless increase logging costs in the
In swamps, where poor flotation hampers mechanical
equipment ordinarily used for land clearing, chemical con-
trol is the more practical approach to the problem. How-
ever, repeated attempts to control black titi with 2,4-
dichlorophenoxyacetic acid (2,4-D), 2,4,5-trichlorophenoxy-
acetic acid (2,4,5-T) and mixtures of the two were unsuc-
cessful. White titi scattered through black titi thickets
was selectively controlled by 2,4-D, and severely damaged
by 2,4,5-T alone and when combined with 2,4-D (Fig. 2)
black titi was clearly resistant and white titi susceptible
to these phenoxy herbicides.
The research efforts reported herein were directed
towards elucidation of factors contributing to the herbi-
cidal resistance of black titi. Hereafter, resistant
black titi will be designated "(R) titi," and susceptible
white titi, "(S) titi."
Fig. 2. White titi plant (foreground) selectively con-
trolled amid resistant black titi following aerial
treatment with 2,4-D ester applied at the rate of 6
pounds active ingredient in 5 gallons of spray per acre.
LITERATURE REVI EW
This review will cover some of the problems of wetting
and penetration of leaf surfaces by herbicidal sprays,
followed by a sampling of the literature on metabolism of
2,4-D by plants. Pertinent reviews have been written pre-
viously by Norman, Minarik, and Weintraub (1950); Blackman,
Templeton, and Halliday (1951); Crafts (1953); van Overbeek
(1956); Woodford,. Holly, and McCready (1958); Hilton, Jansen,
and Hull (1963); Franke (1967); and Casida and Lykken (1969).
Spray solutions applied to leaves must first penetrate
the cuticle. Superficially this structure amounts to a skin-
like covering on the leaf that functions to prevent evapor-
ative loss of moisture from underlying tissues. The cuticle
is a complex of polymerized alcohols and long-chain fatty
acids. Cutin makes up the framework of the cuticle. It
consists of polymerized dicarboxylic and hydroxy carboxylic
acids (van Overbeek, 1956) which are cross-linked in a
three-dimensional system featuring intermolecular spaces
(Franke, 1967). Except at the uppermost surface, hemi-
cellulosic and pectic substances are distributed throughout
the cuticle. Wax rods anrd platelets extend upward from the
highly hydrophobic cuticular surface (Leopold, 1964). A
polarity gradient exists from the outer to the inner region
of the cuticle. The low polarity of the wax layer gradually
intergrades to regions of higher polarity internally in the
region of the outer epidermal walls. Working with pear
leaves, Norris and Bukovac (1968) demonstrated the presence
of pectic substances along the cuticle/cell wall interface.
The basal portion of the cuticular layer consists of a
hydrophilic cellulose matrix in which are embedded lipo-
philic platelets. This matrix appears to be in close con-
tact with ectodesmata of the epidermal wall (Jansen, 1964;
The formal and polarity of surfaces waxes car have a very
marked influence on the ease with which a leaf can be wetted.
Such waxes appear to be extruded onto the cuticular surface
at random (Schieferstein and Loomis, 1959) where they harden
into widely varying, characteristic forms (Mueller, Carr,
and Loomis, 1954; Schieferstein and Loomis, 1956). Wax pro-
trusions prevent solid contact between spray droplets and
the cuticular surface (Holly, 1964). The degree of spray
droplet spreading on cuticular surfaces is characterized by
its contact angle (van Overbeek, 1956). Foy and Smith (1965)
presented a method for calculating contact angles from drop
outlines on the leaf. The magnitude of the contact angle
seems dependent upon the following factors: (1) the degree
of polarity of the surface wax, (2) its configuration, or
degree of roughness, and (3) whether an air film occurs
below the drop in the crevices of wax extrusions.
Penetration of the cuticular surface by herbicidal
solutions was reviewed by Currier and Dybing (1959).
Van Overbeek (1956) considered the unbroken cuticle an
improbable point of entry for spray solutions. He proposed
the following alternative routes of entry: (1) through
modified cells which overlie veins, (2) stomates, (3)
cuticular cracks, (4) areas of cuticle stretched thin by
underlying tissue, and (5) a combination of the above.
Recent work indicates that penetration of the cuticular
surface is a limiting step in the passage of growth-regula-
tor compounds through the cuticle, and that this penetration
is a physical process. Norris and Bukovac (1969) studied
the penetration of naphthaleneacetic acid (NAA) through
enzymatically isolated pear leaf cuticle (Pyrus communis L.
cv. Bartlett). The penetration of NAA was found strongly
temperature related, exhibiting a temperature coefficient
(Q]0) of 5.6 between 15 and 25 C. The temperature effects
on permeability were found fully reversible on isolated leaf
discs. This was cited as evidence that cuticular penetration
is a physical process. The authors felt the high Q10 values
observed were probably related to the lipoidal nature of the
cuticular membrane. They did not preclude the possibility
that temperature-induced changes in the cuticle or the
surface waxes may have resulted in higher rates of pene-
tration at higher temperatures. This was not considered
very likely, however, in view of previous work (Norris and
Bukovac, 1968) wherein cuticular waxes were determined to
undergo no visible changes between 5 and 35 C.
The moisture status of leaves may have some role in
spray penetration of astomatous surfaces. When leaves are
exposed to humid conditions the inner hydrophilic region
swells, pulling the hydrophobic wax rodlets apart. This
could create numerous channels for solution movement (Jansen,
1964). Crafts, Currier, and Drever (1958) had previously
shown that leaf absorption of maleic hydrazide maximized in
a humid atmosphere. By contrast, Morton (1966) found
humidity had no effect upon absorption of 2,4,5-T by mes-
quite (Prosopis juliflora var. glandulosa) seedlings. He
attributed this unique result to the xerophilic character
of mesquite and its ability to adapt quickly to humidity
changes. Badiei, Basler, and Santelmann (1966) found that
absorption and translocation of 2,4,5-T decreased with in-
creasing moisture stress in oak seedlings.
Penetration of waxy crevices and stomates is greatly
enhanced by addition of a surface active agent (surfactant)
to the spray solution. The term "surfactant" encompasses
a wide range of chemicals classified as emulsifiers, de-
tergents, and wetting agents (Behrens, 1964). All such
materials exhibit activity at surfaces or interfaces.
Behrens (1964) discussed at length the physical and chemical
properties of surfactants, relating these to effects upon
herbicidal formulations. He visualized a surfactant mole-
cule as one composed of two opposing forces, one compatible
with water and the other with oily substances hydrophilicc
and lipophilic). This two-tailed nature of the molecule
allows it to accumulate at an interface. Although many
examples of surfactant enhancement of herbicides appear in
the literature, in no case has the precise mechanism of
action been elucidated.
Jansen, Gentner, and Shaw (1961) conducted an extensive
study of surfactant activity when they studied the inter-
action between 63 surface active agents and dalapon, 2,4-D,
and amitrol. The surfactants involved were representative
of cationic, anionic, non-ionic, and ampholytic classes.
In general the surfactants increased herbicidal toxicity
with logarithmic increases in concentration. Some brought
about progressive decreases in herbicidal activity, while
others were completely ineffective. Some surfactants tested
independent of the herbicide were phytotoxic in their own
right, while others actually stimulated growth of Zea mays
and Glycine max. The authors found no correlation between
any of the surfactants and the ionogenic class to which they
belonged. This study emphasized the necessity of grooming
a particular surfactant-herbicide mixture to the species
and spray system at hand.
Foy (1962) was able to show a marked surfactant
enhancement of 2,2-dichloropropionic acid (dalapon) uptake
by Zea mays leaves. 14C labelled herbicide was applied with
and without surfactant (0.1% sodium dioctylsulfosuccinate,
Vatsol OT). At intervals of from 15 seconds to 3 hours,
the basal portion of treated leaves was sectioned and
radioactivity determined either by autoradiography or by
extraction and counting in a windowless flow counter.
Treatments with Vatsol OT additive sorbed almost instanta-
neously and moved to the basal portion of the leaf in
approximately five minutes. Those without surfactant showed
delayed absorption and practically no movement from the
treatment locus during the 30-minute evaluation period.
In a later paper, Foy and Smith (1965) studied the
effects of seven different surfactants upon the enhancement
of dalapon on corn. These were: polyoxyethylene sorbitan
monolaurate (Tween 20), polyoxyethylene sorbitan (Tween 80),
x-77 (80% alkylaryl polyoxyethylene glycols, free fatty
acids, and isopropanol), polyoxyethylene polyoxypropylene
polyols (T-1947), 75% sodium dioctylsulfosuccinate (Vatsol
OT), Dynawet, and sodium lauryl sulfate. Materials were
used at concentrations ranging from 0.001 to 5.0%. All
surfactants enhanced herbicidal activity as the rate was
increased. Sodium lauryl sulfate and Dynawet gave the most
pronounced enhancement. In two papers, Jansen (1964, 1965a)
gave several examples of surfactant enhancement of 4,6-
dinitro-6-sec-butylphenol (DNBP), 3-amino-l,2,4-triazole
(amitrole), dalapon, and 2,4-D. In these, the author
attempted to interrelate the physical-chemical properties
of the resulting sprays and the degree of enhancement. He
found, for example, that in one surfactant family, the
alkylbenzenesulfonates (ABS), herbicidal enhancement in-
creased as the benzene ring was moved toward C5 and C6
carbons of the n-dodecyl alkyl group. Also, among the
ethoxylated alkyl amine cationic surfactants certain
variations in the alkyl structure affected herbicidal
It is possible to form a 2,4-D molecule having both
surfactant and herbicidal properties by forming long chain
alkyl-amine salts (Jansen, 1965b).
Crafts and Yamaguchi (1958) studied the uptake and
distribution of labelled 2,4-D by SZbrina and Tradescantia.
They applied 10 pl drops containing 5 pg of carboxyl-
labelled 2,4-D in 50% alcohol with 0.1% Nonic 218 as a
surfactant to either upper or lower leaf surfaces. Direc-
tion and rate of movement were determined by autoradiography.
Material applied to the lower surface moved readily to the
roots; that applied to the upper surface moved predominantly
upward in the plant and outward to the leaf tips. Anatom-
ical study of the two plants involved would probably reveal
the phloem is located on the underside of the vascular
bundles. This would account for downward movement by lower
leaf surface applications. Either phloem is lacking on the
upper side of the vascular bundles, or it is protected by
associated tissue. Autoradiographic analysis seemed to
indicate labelled 2,4-D bypasses mature leaves that are ex-
porting foods. They found that flowers and young fruits are
extremely active sinks, accumulating a high concentration of
Radwan, Stocking, and Currier (1960) studied the trans-
location of labelled 2,4-D and other herbicides by means of
histoautoradiography. Carboxyl-labelled 2,4-D (specific
activity: 6.03 mc/mM) was applied to broad bean (Vicia faba)
in 50% ethanol with either X-77 or Vatsol OT as surfactants.
From this study, it was found that leaf-treated plants
always showed activity in the phloem. But as the distance
from the application point increased, phloem activity de-
creased and xylem activity increased. Since herbicidal
action is usually dependent upon phloem transport, these
results suggest a mechanism of immobilization by plants.
Also, it is reasonable to assume that during period of high
transpiration the tendency for lateral transfer between
phloem and xylem might be accentuated.
Jaworski and Butts (1952) treated potted bean plants
(Phaseolus vulgaris, var. Black Valentine) with alpha-
methylene or carboxyl 14C-labelled 2,4-D. At intervals of
1 to 42 days treated plants were sectioned into leaf and
stem material, homogenized, and extracted with 80 per cent
ethanol. Chromatographic analysis showed the 2,4-D was
rapidly converted to two unknown, radioactive compounds. At
the end of two weeks approximately 65 per cent of the applied
2,4-D was converted to "unknown 1" and 9 per cent to "unknown
2." About 20 per cent of the 2,4-D was recovered, unchanged.
Both unknowns were water-soluble, and ether-insoluble. Un-
known 1 was hydrolizable with 2 N HC1, Takadiastase, and
emulsin. The authors felt it was possibly a glucosidic
detoxification product. The small amount of 2,4-D not
accounted for (6%) was presumed lost to decarboxylation or
transported to the roots.
In looking for factors of selectivity, Fang and Butts
(1954) extended the work of Jaworski and Butts (1952) to two
2,4-D resistant plants, corn (Zea nar~y saccharata var.
Golden Bantam) and wheat (Triticum vulgare var. Hard Winter
White). Using similar treatments and analytical methods the
authors found in addition to unknowns 1 and 2, a third,
which they called "unknown 3." As were unknowns 1 and 2,
unknown 3 was water soluble and ether-insoluble. The authors
concluded that the formation of unknown 3 along with slower
rates of absorption and translocation could account for the
selective resistance of corn and wheat to 2,4-D.
Jaworski, Fang, and Freed (1955) reported later that a
2,4-D complex, presumed to be the "unknown 1" of Jaworski
and Butts (1952), was formed equally by normal and etiolated
bean plants. In addition they found that although 2,4-D was
absorbed by etiolated leaves, its transport was dependent
upon exogenous supplies of sucrose, glucose, and various
other sugars. Since all sugars facilitated the transport of
2,4-D, the authors concluded this action is mediated by food
movement in general.
Using 2,4-D-2-14C and 2,4-D--114C, Williams, Slife, and
Hanson (1960) studied absorption, translocation, and metab-
olism in several annual broadleaved weeds. Absorption was
determined by floating leaf and stem section on radioactive
solutions, washing, drying, and counting. Analysis for
radiocarbon was done by collecting CO2 evolved by Van Slyke
combustion and determining the radic-activity with a
vibrating reed electroscope. Translocation was studied by a
combination of wet combustion analysis and autoradiography.
Side-chain catabolism was studied by sealing the treated
plants in a jar and collecting the evolved CO2 in one ml of
20% KOH. The KOH capsules were collected, dried under a
heat lamp and counted in a G-M counter. The amount of 14CO2
evolved compared to amount absorbed was so slight (less than
0.02%) that this mechanism of 2,4-D detoxification was not
considered to occur generally. According to Williams et al.
(1960), Weintraub (1953) had observed similar results.
Bach and Fellig (1961) showed that growth of excised
bean sections treated with 2,4-D was correlated with dis-
appearance of free 2,4-D from the tissue. They found, how-
ever, that approximately 97 per cent of the herbicide re-
mained in the tissue in a form chromatographically distinct
from pure 2,4-D. It was called merely a "detoxification
product." In a following paper, Bach (1961) attempted to
isolate this "detoxification product." Four-week-old bean
(Phaseolus vulgaris) stems were cut into six-inch lengths
and vacuum infiltrated with a 2 per cent sucrose solution
containing 10-4M 2,4-D-1-14C (2-22 mc/mM). Stems were
incubated at room temperature for four days. After lyoph-
ilization, soxlet extraction, ether soluble and ether in-
soluble fractions were obtained. A host of chemical tests
were run on chromatographed components of both fractions.
Only radioactive isolates were analyzed. Bach found that
approximately 50 per cent of the original radioactivity was
distributed among 10 compounds in the ether extract. The
ether insoluble fraction contained six components. Various
breakdown pathways were postulated. These included ring
fissure, hydroxylation, and decarboxylation.
Luckwill and Lloyd-Jones (1960a) studied the metabolism
of 2,4-D and other phenoxy materials by susceptible black
currant (Ribes nigrum L.) and resistant red currant (Ribes
sativum Syme). 2,4-D, labelled with 14C in either the
methylene or carboxyl position was applied to detached
leaves via the petioles. Evolution of CO2 in darkness was
followed over a period of one week. After this, leaves were
extracted by maceration in sodium bicarbonate or dilute
ammonium hydroxide and analyzed by paper chromatography.
During 139 hours in the metabolism chamber, resistant
currant decarboxylated 48 per cent of the carboxyl and 20
per cent of the methylene carbon from the side chain of
2,4-D. During the same period, susceptible currant degraded
only 2 per cent of the applied material. The authors con-
cluded that differential decarboxylation was the basis of
selective 2,4-D toxicity. Ancillary experiments with
2,4,5-T, 4-chlorophenoxyacetic acid (4-CPA), 2-methyl-4-
chlorophenoxyacetic acid (MCPA), and 2-chlorophenoxyacetic
acid (2-CPA) showed all but the latter compound were
decarboxylated by resistant red currant. 2-CPA was thought
to be rapidly bound or converted to another form not
vulnerable to decarboxylation. Even though extensively
decarboxylated, 2,4,5-T was found decidedly toxic to 2,4-D-
resistant red currant. The authors suggested this species
was unable to cope with the probable decarboxylation pro-
duct, 2,4,5-trichlorophenol, but could selectively inacti-
In a supporting study Luckwill and Lloyd-Jones (1960b)
were able to correlate 2,4-D resistance with decarboxylation
of the molecule by an apple variety (Cox's Orange Pippin)
and its relatives and three strawberry varieties (Royal
Sovereign, Cambridge Favourite, and.Talisman). As with red
currant, Talisman strawberry was susceptible to 2,4,5-T even
though decarboxylation occurred. Only one of sixteen other
test species known to be resistant or susceptible to 2,4-D
carried out any appreciable decarboxylation. Of special
interest was the fact that hawthorne (Crataeus oxycantha)
carried out no side-chain degradation, even though highly
resistant to 2,4-D. As did Williams et al. (1960), the
authors concluded that resistance via the decarboxylation
route does not universally occur and that other groups of
plants must have other resistance mechanisms.
Edgerton and Hoffman (1961) discovered that substitu-
tion of fluorine for chlorine in the 4-position of 2,4-D
(2-chloro, 4-fluorophenoxyacetic acid; 2,4-F) inhibited
decarboxylation of the compound in leaves that otherwise
rapidly metabolized unsubstituted 2,4-D. Detached shoots
of resistant McIntosh apple were treated via the trans-
piration stream with 2,4-D-1-14C and 2,4-F-1-14C then
placed in a darkened metabolism chamber for collection of
evolved 14CO2. In 24 hours only 4 per cent of the 14C was
recovered as 14CO2 from shoots treated with 2,4-F-1-14C.
During the same period, those treated with 2,4-D-1-14C
degraded 33 per cent of the 14C supplied. A parallel study
with 2,4-D-susceptible Staymen apple showed that even in
this species, where 2,4-D is very slowly decarboxylated, a
slight further reduction in breakdown of 2,4-F occurred
(0.14 vs. 0.56%). Basler (1964) showed that 2,4-D, 2,4,5-T
and 2,4-F were not extensively decarboxylated by excised
leaves of blackjack oak (Quercus marilandica Muenchh),
persimmon (Diospyros virqiniana L.), winged elm (Ulmus alata
Michx.), sweet gum (Liquidamvbar styraciflua L.) or green ash
(Fraxinus pennsylvalnica Marsh.).
Wilcox, Moreland, and KlingTnan (1963) discovered that
excised oat, barley, and corn roots produced a ring-
hydroxylated metabolite from phenoxy-n-aliphatic acids,
acetic through hexanoic. The metabolite, identified as
4-hydroxy-phenoxyacetic acid, was produced from compounds
bearing even-numbered side chains. Since there was no
evidence of hydroxylated products of chain lengths longer
than acetic, the authors concluded that beta oxidation of
the side chain may have preceded the hydroxylation. The
greater hydroxylation activity observed in oat, barley, and
corn compared to that in peanut, soybean, and alfalfa roots
suggested ring hydroxylation may be a detoxication mechanism
in the former. Thomas, Loughman, and Powell (1963) isolated
the same metabolite from mesocotyl segments of Avena sativa.
In a later paper, the same authors (Thomas, et al., 1964a)
reported that phenoxyacetic acids with an unsubstituted four
position were hydroxylated there, and the resulting product
was accumulated as a glucoside. Phenoxyacetic acids bearing
a chlorine atom in the four position were not hydroxylated
to the same extent.
Faulkner and Woodcock (1964) isolated a 2,4-D metabolite
from Aspergillus niger, which they identified as 2,4-dichloro-
5-hydroxy-phenoxyacetic acid. A minor metabolite proved to
be 2,5-dichloro-4-hydroxy-phenoxyacetic acid. The formation
of this metabolite was unusual in that it not only involved
a hydroxyl-chlorine replacement, but also a chlorine shift
from the four to five position. Thomas et al. (1964b) dis-
covered that a similar chlorine shift occurred in seedlings
of Phaseolus vulgaris var. Canadian Wonder. A phenolic
glucoside was isolated. After treatment with beta-glucosidase,
the ether-soluble aglycone was found to consist of a mixture
of 2,5-dichloro-4-hydroxy-phenoxyacetic acid and 2,3-dichloro-
Crosby (1964) studied the metabolism of 2,4-D by red
kidney bean (Phaseolus vulgaris L.). The major chlorine-
containing ether-soluble plant fraction proved to be free
2,4-D. The ether-insoluble fraction yielded 2,4-D and another
chlorinated compound upon acid hydrolysis. A small amount of
2,4-dichloroanisole was isolated, but no 2,4-dichlorophenol
An interesting detoxication mechanism for jimsonweed
(Datura stromonium) was discovered by Fites, Slife, and Hanson
(1964). They used carboxyl-labelled 2,4-D-1-14C (specific
activity: 1 mc/mM). A 10 microliter drop of solution con-
taining 22.1 mg of 2,4-D in 2 ml of 50 per cent ethanol was
applied to the first nearly mature leaf of experimental plants.
At different intervals, treated plants were harvested and
divided into stems, roots and leaves. Following ether ex-
traction the changes in relative activity of various plant
portions were evaluated. After 21 days most of the activity
resided in the lower stems and in the roots. The nutrient
solutions used were found to be radioactive, suggesting con-
siderable 2,4-D leakage occurred from the roots. A follow-up
experiment confirmed that approximately 25 per cent of the
original 2,4-D applied had been expelled intact from the
roots. The net result was a transfer of material from the
leaf surface to the medium. As 2,4-D in the apical region
of the plant reached a certain tolerable concentration,
MATERIALS AND M1ETIHODS
Anatomical Stud oof Leaves
Mature leaves of (S) and (R) titi were collected from
the natural habitats of the two species in Taylor and
Franklin counties, respectively. The base and apex of
sample leaves were snipped free leaving a central segment
approximately 1 cm long. After dehydration in a graded
series of ethanol solutions leaf segments were embedded in
Fisher Tissuemat (melting point, 52-54.5 C). Leaf cross-
sections 12 microns thick were prepared with a Spencer
Model 815 rotary microtome. Resulting sections were
mounted on glass microscope slides and stained with
safranin-fast green according to the procedure of Johansen
(1940). Photographs of the stained sections were made
with a Spencer Model 675 photomicrographic camera.
Electron Microscopy of Leaf Surfaces
Preparation and microscopy of titi leaves were done at
The Buckeye Cellulose Corporation, Cellulose and Specialties
Division laboratory, Memphis, Tennessee, by Mr. K.A. Jurbergs
and his staff.
Mature leaves from potted two-year-old (S) and (R)
plants grown from cuttings were prepared for electron
microscopy by means of a two-step replication technique.
Adaxial leaf surfaces were first impressed into a poly-
styrene film softened by heating,. After the film hardened
the leaves were stripped off and discarded. A chromium
layer approximately 150A thick was deposited on the leaf
impressions by vacuum evaporation of metallic chromium.
Another layer of about the same thickness was deposited
on the chromium by vacuum evaporation of carbon. Desired
areas were cut from the layered replicas and placed on
copper specimen grids with the polystyrene film in contact
with the grid. The polystyrene was then dissolved away in
a solution consisting of 75 per cent ethylene dichloride
and 25 per cent xylene. Final specimens consisted of a
chromium replica supported by carbon. Micrographs were
made with a Phillips Model EM-75B electron microscope.
Leaf Absorption of 2,4-D-- 14C
Mature (S) and (R) titi leaves were harvested from
two-year-old greenhouse-grown plants and inserted into
one-inch cubes of florist's foam plastic hydrated with
0.1 M glucose solution. Four petri dishes were prepared,
each holding four leaves each of (S) and (R) titi. Just
prior to insertion, a fresh face was cut on the petioles
with a sharp scalpel to ensure good continuity between con-
ducting tissues and the hydrating solution. After a five-
minute equilibration period 10 yil drops of 0.1 M radio-
labelled 2,4-D (2,4-D---I-14C; S.A. 0.014 pc/)jl) solution,
as the sodium salt, were applied to the center of the
adaxial leaf surfaces. The uncovered dishes were illumi-
nated with approximately 200 f.c. of incandescent light.
After 2 hours, the dry herbicide residue was washed from the
leaves into stainless steel planchets with 5 per cent ammo-
nium hydroxide solution. Pianchets were dried under a heat
lamp and counted in a Nuclear-Chicago Model 8775 scaler.
The scaler was attached to a Nuclear-Chicago Model 3053
sample -changer, which housed a thin-window Geiger tube.
2,4-D-1-14C uptake was estimated from the difference
between radio-activity applied and that recovered from leaf
washings. Count data were converted to micrograms of
2,4-D-1-14C taken up. The experiment was analyzed as a
paired "t" test as outlined by Snedecor and Cochran (1956).
Metabolic Degradation of 2,4-D-l-14C
Preliminary Bell Jar Experiment
A preliminary experiment similar to that of Williams
et al. (1960) was devised to check for metabolic degradation
of 2,4-D-1-14C by (S) and (R) titi via the decarboxylation
An (R) titi shoot 10.cm in length was stripped of all
but five terminal leaves and placed in a 125 ml Erlenmeyer
flask containing distilled water. The shoot was steadied
in a vertical position with a cotton plug placed in the
neck of the flask. The leaves to be treated were held in
a horizontal plane with cellophane tape. A 10 ul drop of
0.1 M 2,4-D-1-14C (S.A. = 0.014 uc/10 ul) solution con-
taining 1 per cent v/v polyoxyethylene sorbitan monolaurate
(Tween 20) was applied to the center of the adaxial sur-
faces of four leaves, with the fifth being left untreated.
The treated plant and flask were sealed with a two-liter
bell jar along with a planchet containing 1 ml of 20 per
cent potassium hydroxide solution. The assembly was placed
in a controlled temperature room (72 + 30F) and exposed to
the following light and dark periods: 9.5 hours' light, 8
hours' dark, 16 hours' light, and 8 hours' dark. After
this period, 41.5 hours in all, the planchet containing
potassium hydroxide was removed from the bell jar, dried
under a heat lamp, and counted in the scaler. Treated
leaves were washed individually into planchets with 5 per
cent ammonium hydroxide for determination of residual radio-
Another replicate of the same experiment was carried
out with a comparable shoot collected from an (S) plant.
The light and dark periods; were essentially the same as
those for (R) titi.
Absorption Train Ex2p riLme nt
An absorption train (Fig. 3) similar to that of Morgan
and Hall (1962) was assembled for measuring the incremental
evolution of 14C02 from (S) and (R) titi shoots treated with
a 0.01 M solution of 2,4-D--1-4C (S.A. = 1 mc/mM).
Detached (S) and (R) titi shoots 8 to 10 cm long,
bearing seven terminal leaves were used in this experiment.
Herbicidal treatment was made by placing the shoots in a
2 ml volumetric flask containing a weighed amount of
2,4-D-l-14C solution, and allowing them to absorb the
labelled material via the transpiration stream. The neck
of the flask was tightly plugged with cotton to prevent
loss to evaporation. The assembly was placed 10 inches
from a slowly moving six-inch fan, and illuminated with
approximately 200 f.c. of incandescent light during the
uptake period. Periods were varied from 10 minutes to 3
hours to give a range of herbicide dosages in the experi-
mental material. Uptake was estimated from the loss in
weight of the flask and solution. After treatment, the
shoot was transferred to a 5 ml volumetric flask containing
distilled water. The neck of the flask was plugged with
cotton as before. The assembly was placed in the metab-
Fig. 3. Absorption train used in 2,4-D-1-14C metabolism
experiments with (S) and (R) titi shoots.
olism chamber, which then was covered with a light-tight
box for the duration of the test. At the start of the
metabolism period, the chamber w:s purged of atmospheric
CO2 by pulling in a supply of air through scrubbers con-
taining 1 N solutions of sodium hydroxide. The chamber was
then sealed off with screw clamps. Each treated shoot
remained in the chamber for approximately 140 hours. Per-
iodically during this time the chamber air was absorbed into
200 ml of 0.1 N sodium hydroxide. The absorber solution was
transferred to a 500 ml Erlenmeyer flask. To this was added
about 15 ml of 2 N barium chloride to recover any evolved
14CO2 as precipitated Ba14CO3. The precipitate was collect-
ed by vacuum filtration onto a sintered stainless steel disc
covered with a thin layer of Celite filter aid. After
drying under the heat lamp, the precipitate and disc were
placed in a copper planchet and counted in the scaler.
Counts were corrected for precipitate mass by means of the
curve shown in Fig. 4.
Seven metabolism experiments were run with (R) and
three with (S) titi. A slight modification was made in the
fourth metabolism experiment, designated R-4. Just prior
to treatment with 2,4-D-1-14C solution, the shoot was
allowed to take up, via the transpiration stream, 2 ug of
fluorine from a solution 100 p.p.m. in sodium fluoride.
0 10 20 30 40 50 60 70 80
Sample Density- -g/cm2
Fig. 4. Mass correction curve for barium carbonate
Anatomical Study of Leaves
Microscopic examination of leaf cross-sections of (S)
and (R) titi (Figs. 5 and 6, respectively) revealed many
(S) titi has a thin cuticle (2.4 microns) overlying an
epidermis consisting of large, thin-walled cells loosely
arranged above the main vascular bundle. The main and
secondary vascular bundles are collateral, having both in-
ternal and external phloems. A well'-defined bundle sheath
extension pyramids upward from the main vascular bundle to
the loosely arranged region of the epidermis. Bundle sheath
extensions lead from secondary veins to both upper and lower
leaf surfaces (Fig. 7). The mesophyll is bifacial. Two
layers of palisade parenchyma are evident, with a third
intergrading into spongy parenchyma below. Stomates are
lacking on the upper leaf surface but occur frequently on
(R) titi has a thick cuticle (13 microns) overlying an
epidermis consisting of small, thick-walled cells compactly
arranged above the main vascular bundle. Main and secondary
Fig. 5. Leaf cross-section of (S) titi (Cyrilla racemiflora) at mid-rib, x 330, showing
cuticle (C), epidermis (E), bundle sheath extension (BSE), internal phloem (IP), xylem
(X), external phloem (EP), palisade mesophyll (PM), spongy mesophyll (SM), and stoma (S).
Fig. 6. Leaf cross-section of (R) titi (Cliftonia monophylla) at mid-rib, x 330, showing
cuticle (C), epidermis (E), bundle sheath extension (BSE), xylem (X), phloem (P), pali-
sade mesophyll (PM), and spongy mesophyll (SM). Note the absence of sheath extensions
on the secondary vein to the right of the main vascular bundle.
Fig. 7. Leaf cross section of (S) titi (Cyrilla race-
miflora), x 330, showing secondary vein with bundle
sheath extensions (BSE) leading to upper and lower
epidermis, xylem (X), and phloem (P).
vascular bundles are bicollateral, with internal phloem
lacking. A poorly defined sheath extension leads upward
from the main vascular bundle to the epidermis. Secondary
veins are deeply embedded in the leaf, without sheath
extensions. As with (S) titi, the mesophyll of (R) titi
is bifacial. Stomates are absent on the upper leaf surface
and only sparsely present on the lower.
The major anatomical differences are summarized in
Electron Microscopy of Leaf Surfaces
Examination of electron micrographs of (S) and (R) titi
leaf surfaces show marked differences in cuticular texture.
(S) titi (Fig. 8) has a relatively smooth, homogeneous
surface through which the dim outlines of epidermal cell
walls can be seen. (R) titi (Fig. 9), by contrast, exhibits
a rough, ridgy surface, probably waxy and very hydrophobic.
This would suggest (R) titi leaf surfaces are much more
difficult to wet by herbicidal sprays, as considerable re-
duction in spray droplet surface tension would be required
to cause them to relax into the waxy crevices and other
irregularities. On the other hand, (S) titi surfaces could
be wetted with comparative ease. This difference in wett-
ability is illustrated in Fig. 10, which shows the extent
of spread of 10 ul drops (0.01 M 2,4-D-1-14C, 1 per cent v/v
Table 1. Major anatomical differences between (S) and (R) titi leaves.
Feature (S) Titi (R) Titi
cuticle 2.4 microns thick 13 microns thick
epidermis large, thin-walled cells small, thick-walled cells
loosely arranged above compactly arranged above
main vascular bundle main vascular bundle
vascular bundles collateral bicollateral
bundle sheath present on all veins weakly defined on main vas-
extensions cular bundle but lacking on
stomates plentiful on lower epidermis sparsely distributed on lower
Fig. 8. Electron micrograph showing the smooth adaxial surface of an (S) titi leaf, x
^ ( 4 0.*ev
Fig. 10. The dry residue of 10 ul drops of 2,4-D-1-14C
containing 1 per cent Tween 20, applied to the center
of the adaxial surfaces of (S) and (R) titi leaves.
in Tween 20) applied to the center of the leaves of the two
species. On the (S) titi leaf, the droplet spread quickly
in both directions along the mid-vein and migrated toward
the lateral margins before drying. On the (R) titi surface,
the droplet dried essentially in place, with little or no
spreading from the point of application.
Leaf Absorption of 2,4-D-1-14C
Results and analysis of the 2,4-D-1- 4C absorption
trials with excised leaves of (S) and (R) titi are given
in Table 2.
(S) titi leaves absorbed on the average, 18.61 + 7.32
ug of the 221 ug of 2,4-D-1-- C applied in 10 ul drops.
This amounted to an absorption of 8.4 per cent. Under the
same experimental conditions, (R) titi absorbed 14.54 +
2.32 ug, for an uptake of 6.6 per cent. A comparison of
the two averages shows that (S) titi took up 1.28 times
more 2,4-D-1-14C than (R) leaves. This difference was
significant at the 5 per cent level of t.
Metabolic Degradation of 2,4-D-- 14
Preliminary Bell Jar Experiment
As mentioned earlier the bell jar experiment was
instituted to check for decarboxylation of 2,4-D-1-14C by
the titi species as a prelude to a more extensive absorption
train experiment. Results of the bell jar trials are
Table 2. Analysis of 2,4-D-1-14C absorption by (S) and (R)
titi leaves and significance test of the differences.
Leaf 2,4-D-1-14C Absorption Difference
Pair No. (S) Titi (R) Titi D = X1 X2
ug ug ug
1 10.5210 4.9098 5.6112
2 50.5008 4.9098 45.5910
3 17.5350 11.9238 5.6112
4 25.6011 17.5350 8.0661
5 10.5210 34.3686 -23.8476
6 19.2885 21.3927 2.1042
7 11.5731 15.0801 3.5070
8 46.6431 4.9098 41.7333
9 83.1159 20.6913 62.4246
10 17.1843 12.6252 4.5591
11 88.0257 9.8196 78.2061
12 26.3025 19.9899 6.3126
13 24.1983 10.8717 13.3266
Total 431.0103 189.0273 241.9830
Mean 18.6141 14.5405 18.6140
Calculation of "t" and Test of Siqnificance
S2 = D2 n
= 14784.9320 -
2 = S2 = 856.7201 = 65.9015 S- = 8.1182
d d d
t = d uD = 18.6140 0 = 18.6140 0 = 2.2928*
S- 8.1182 8.1182
d.f. = n 1 = 12; t.05 = 2.160
summarized in Table 3, below. Both (S) and (R) titi carried
out some decarboxylation (2.4 and 5.4%, respectively) and
(S) titi absorbed more 2,4-D--1-14C than (R) titi (5.43 vs
3.10%). Because of the lack of replication, this experiment
did not lend itself to statistical analysis.
Absorption Train Experiment
2,4-D-1-14C breakdown data from the absorption train
trials are given in Table 4. Breakdown was expressed as a
per cent of 2,4-D-1-14C taken up via the transpiration
stream. Depending on dosage, (R) titi degraded from 5 to
13.9 per cent of the 2,4-D-1-14C during dark periods of
approximately 140 hours duration. Two of the three (S) titi
trials resulted in no degradation. A third, S-2, showed a
breakdown of 0.8 per cent. A marked difference in
2,4-D-1-14C metabolism exists between (S) and (R) titi
(Table 4). This difference is shown more strikingly in
Fig. 11, where cumulative counts per minute evolved as
Bal4CO3 are plotted over time for the trials listed in Table
4. Trial R-4 was omitted from the plot because of the pre-
treatment it received with sodium fluoride solution.
In order to evaluate the relationship between
2,4-D-1-14C uptake and breakdown, the data in Table 4 were
analyzed by both linear and curvilinear regression analysis.
The linear regression, Y = 14.2043 + 4.2602X, was rejected
Table 3. Results of preliminary bell jar experiment in
which evolved 14CO2 was absorbed in potassium hydroxide
amounts per 10 minutes.
bExpressed as a per cent of total activity applied.
CExpressed as a per cent of activity absorbed.
Table 4. Uptake and degradation of 2,4-D-1-14C by detached
shoots of (S) and (R) titi during approximately 140 hours
in the dark.
Trial and 2,4-D---14C 2,4-D-1-14C
Plant Uptake Breakdown
apretreated with 2 ug flourine as sodium fluoride.
0 10 20 30 40 ISO 0 70 O 90 100 110 120 130 1l0
Hours in Dark
Fi. 11. Cumulative counts per minute evolved as Ba 4C03
by (S) and (R) titi during the indicated dark periods.
by means of the calculations given in Tables 5, 6, and 7.
A significant F value was determined (Table 7) when the
mean square deviations from the linear and curvilinear
regressions were compared. The second degree curve,
Y 15.7677 + 28.8199X + (-3.4952) X2, was considered to
fit the data best, due to this significant reduction in the
sums of squares.
The analysis of variance for curvilinear regression
(Table 8) showed that a significant amount of the vari-
ability was accounted for by the second degree curve.
Further, the coefficient of determination (R2) of 0.93
(Table 8) indicated that approximately 96 per cent of the
variability in breakdown was accounted for by differences
in 2,4-D-1-14C uptake by the plants.
Herbicide breakdown (Fig. 12) in the (R) titi shoot
maximized at a calculated level of 400 pg total, or 57 pg
per leaf (all shoots bore seven leaves). This result
suggested that toxic levels were reached at a 2,4-D con-
centration of 57 pg per leaf. By extrapolation of the
curve it can be estimated that little or no breakdown would
occur beyond a concentration of about 800 ug per shoot, or
114 pg per leaf.
Confidence intervals were calculated for the degrada-
tion curve (Fig. 12) to determine whether 2,4-D-1-14C
Table 5. Data for computing standard deviation from the curvilinear regression
equation Y = -15.7677 + 28.8199X + (-3.4952) X2.
Plant and Uptake Breakdown
Trial X Y Y dyx dyx2
ug x 102 ug ug ug ug
R-l 3.562 46.6 42.5424 +4.0576 16.4641
R-2 2.372 29.6 32.9281 -3.3281 11.0762
R-3 1.039 14.4 10.4031 +3.9969 15.9752
R-5 6.020 30.3 31.0606 .-0.7606 0.5785
R-6 1.582 17.5 21.0779 -3.5779 12.8013
R-7 1.240 14.2 14.5947 -0.3947 0.1557
S = Ed 2 = 57.0510 = 19.0170
Syx = + 4.3608
Table 6. Data for computing standard deviation from the linear regression
equation Y = 14.2043 + 4.2602X.
Plant and Uptake Breakdown
Trial X Y Y dyx d 2
ug x 10 ug ug ug ug
R-1 3.562 46.6 29.3791 17.2209 296.5593
R-2 2.372 29.6 24.3094 5.2906 27.9904
R-3 1.039 14.4 18.6306 -4.2306 17.8979
R-5 6.020 30.3 39.8507 -9.5507 91.2158
R-6 1.582 17.5 20.9439 -3.4439 11.8604
R-7 1.240 14.2 19.4869 -5.2869 27.9513
Syx = dyx2 = 473.4751 = 118.3687
S = + 10.8797
Table 7. Comparison between mean squares for linear and curvilinear
regression--(R) titi decarboxylation data.
Analysis of Variance
Source of Variation d.f. SS MS
Deviations from Linear Regression 4 473.4751
Deviations from Curvilinear Regression 3 57.0510 19.0170
Curvilinearity of Regression
F = 416.4241 = 21.8974*
F .05; n = 4,3 = 9.12
Table 8. Analysis of variance for curvilinear regression -
2,4-D-1-14C uptake and breakdown data for (R) titi trials
in the absorption train.
Analysis of Variance
Variation d.f. SS MS F
Total 5 799.9333 159.9867
Regression 2 742.8917 371.4458 19.535*
Error 3 57.0416 19.0139
R2 = 742.8917
F = 371.4458 = 19.535*
F 05; n = 2,3 = 9.55
SBLACK TITI (Cliftonia onoPhvlla)
S= -15.7677 + 28.8199X + (-3.4952)X2
0 1 3 4 7
2,4-D Uptake x 10-2 -
Fiq. 12. The relationship between 2,4-D (2,4-D-1-14C)
uptake and breakdown (decarboxylation) by (R) titi
shoots during approximately 140 hours in the dark.
breakdown by the R-4 shoot, pretreated with 2 ug of fluoride
ion, departed significantly from predicted values based on
the amount of herbicide taken up. The standard deviation
from curvilinear regression (Syx), calculated in Table 5,
was determined to be + 4.3608 ug. Shoot R-4 took up
254.4 ug and degraded 23.6 ug of the applied material
(Table 4). Based on the regression in Fig. 12, the pre-
dicted breakdown for an uptake of 254.4 ug was found to be
34.9 ug. The difference between actual and predicted
values was 1.1.3 ug, or 2.60 times greater than one standard
deviation (11.3290 e 4.3608 = 2.60). Since the value of
t.05 for five degrees of freedom is 2.57, we can say the
quantity of 2,4-D-1-14C degraded by shoot R-4 departed
significantly from the other six shoots evaluated in the
absorption train. Presumably this departure, amounting to
a 32 per cent reduction in breakdown, was fluorine-related,
as experimental conditions were otherwise held constant.
Unquestionably, leaf anatomy plays no small role in
determining the degree of herbicidal susceptibility of any
given plant. Before a spray deposit bearing systemic
materials can produce the desired effect, it must first
penetrate into the leaf. Foliar absorption must occur
directly through the leaf cuticle and through stomates if
present. However, once into stomatal chambers, the spray
solution is still confronted by a cuticular lining which
must be crossed (Norris and Bukovac, 1968). While being
absorbed, the herbicide must be kept in a fluid state. The
quantity of active material absorbed is a function of its
concentration in the spray droplet, leaf area covered, and
the rate of evaporation of the solution.
After crossing the cuticle, the herbicide passes through
and between the cells surrounding the main and secondary
veins gaining access to them through bundle sheath extensions
when present (Esau, 1953; van Overbeek, 1956). Upon entering
the veins, solution movement is bi-directional, part of it
moving upward in the xylem via the transpiration stream, and
part toward the root in the phloem, in the food transport
system. Thereafter, the herbicide must reach a sensitive
site and accumulate to toxic concentration in some active
From the standpoint of herbicidal susceptibility, (S)
titi (Fig. 5) has an ideal leaf anatomy. A spray droplet
landing above the main vein is confronted by a thin, smooth
cuticle overlying loosely arranged epidermal cells. Droplet
spreading (Fig. 10) and solid contact are favored by the
lack of wax protrusions on the leaf surface (Fig. 8). Due
to the large voids between epidermal cells, the herbicide
probably reaches the upper bundle sheath extension via an
apoplastic route. Once in the bundle sheath extension, the
herbicide solution is funnelled directly to the internal
phloem. Sheath extensions on secondary veins (Fig. 7) like-
wise channel spray materials to the phloem, although some are
doubtless intercepted by xylem tissue overlying the external
phloem (internal phloem was found lacking in secondary veins
of (S) titi).
In contrast to (S) titi, (R) titi (Fig. 6) has many of
the anatomical attributes of a resistant species: thick,
ridgy cuticle, compact epidermis, scanty phloem, and deeply
embedded veins lacking sheath extensions to the epidermis.
Since internal phloem is lacking in (R) titi, the upper
bundle sheath extension leads directly from the epidermis
to the xylem of the main vascular bundle. Herbicide solution
entering the leaf in the vicinity of mid-rib is probably
carried to the xylem elements where it is sidetracked into
the transpiration stream. Solutions entering the leaf at
points other than near the mid-rib probably move through the
mesophyll with difficulty, as bundle sheaths are lacking on
secondary veins in this species.
Results given in the present study showed that (S) titi
absorbed 28 per cent more 2,4-D-1-14C (sodium salt) than (R)
titi under the same experimental conditions. This difference
is surprisingly small considering the disparity in cuticular
thickness observed (2.4 vs. 13.0 microns). However, as in-
dicated earlier (Norris and Bukovac, 1969), the limiting step
in leaf absorption is apparently the initial penetration of
the cuticular surface, not the thickness of the cuticle pae
Since the uptake trials involved mid-rib placement of
droplets, the results could have been slightly biased in fa-
vor of (R) titi. Close scrutiny of Fig. 6 will show light
streaks bisecting the cuticle along the concave surface of
the leaf above the main vein. The streaks, which terminate
at the anticlinal walls of epidermal cells, possibly function
as tension zones to accommodate changes in leaf size due to
variable turgidity. One might expect that solutions would
preferentially enter such tension zones in the cuticle.
Despite this possible bias, the difference in uptake was
significant, and it was correlated with susceptibility.
This result was at variance with those of Fang (1958),
Luckwill and Lloyd-Jones (1960a, 1960b), Foy (1962), and
others, who observed no such correlation between susceptible
and resistant species.
The amount of decarboxylation determined for (R) titi
was low compared to that of resistant red currant, straw-
berry, and apple, reported by Luckwill and Lloyd-Jones
(1960a, 1960b), and Edgerton and Hoffman (1961). It is pro-
posed here that (R) titi is herbicidally resistant by virtue
of the combined effects of poor leaf absorption and the
ability to carry out some decarboxylation of 2,4-D. As pre-
viously stated, (R) titi was able to withstand aerial
applications of 2,4-D ester applied at the rate of 6 pounds
active ingredient in 5 gallons of spray per acre. The
application equipment used produced spray droplets averaging
about 400 microns in diameter. With the above spray volume,
and allowing 20 per cent spray loss to the atmosphere, this
would yield about 75 drops per square inch. The concentra-
tion used would supply 48 x 10-2 mg 2,4-D per square inch.
(R) titi leaves individually cover an average area of 0.875
square inches. Allowing for inclination and overlap, the
effective leaf area is probably reduced to approximately 0.5
square inches. Thus it is estimated that each (R) titi would
intercept (0.5)(48 x 10-2) 24 x 10.2 mg or 240 ig 2,4-D
via the commercial application. To reach the toxic threshold
of 57 pg per leaf, the commercially applied material would
need to penetrate the (R) titi leaf at a level in excess of
23 per cent. This is a reasonable level of penetration which
should be attainable through surfactant enhancement.
Introduction of fluoride ion into the (R) titi leaf just
prior to treatment with 2,4-D-l-14C had a sparing effect on
decarboxylation. This result,though similar, is not fully
equatable to that of Edgerton and Hoffman (1961) since they
utilized a fluorine substitution in the 2,4-D molecule rather
than the free ion for inhibition of the systems) involved.
Addition of fluoride ion to spray mixtures may provide a use-
ful means of potentiating 2,4-D in (R) titi. Also, it would
be interesting to check for the selective toxicity of 2,4,5-T
as observed by Luckwill and Lloyd-Jones (1960a, 1960b) in
other decarboxylating species.
SUMMARY AND CONCLUSIONS
The failure to control black titi (Cliftonia mono-
phvlla) with aerial applications of 2,4-D led to studies
of the factors of resistance, using closely related
2,4-D-susceptible (S) white titi (Cyrilla racemiflora) for
Anatomical study of leaves showed resistant (R) titi
to have a thick cuticle, compactly arranged epidermal cells,
no internal phloem in the main vascular bundle, and no
bundle sheath extensions. (S) titi had a thin cuticle,
loosely arranged epidermal cells, internal phloem just under
the epidermis, and well-defined bundle sheath extensions on
all veins. Ultra-structure studies of adaxial leaf sur-
faces revealed those of (R) titi to be rough and ridgy, and
those of (S) titi, relatively smooth and homogeneous.
Ten il drops of 2,4-D-1- C solution containing Tween
20 surfactant were applied to excised leaves of (S) and (R)
titi. (R) titi absorbed 6.6 per cent and (S) titi, 8.4 per
cent of the applied material. The difference was signifi-
cant at the 5 per cent point of t.
Detached shoots of (S) and (R) titi were fed 2,4-D-1-14C
solution via the transpiration st-ream then sealed in a dark-
ened metabolism chamber for measur:iement of evolved 14CO2.
Depending on dosage, (R) titi shoots degraded from 5 to 13.9
per cent and (S) titi, 0 to 0.8 per cent of the applied her-
bicide during approximately 140 hours. Curvilinear regress-
ion analysis of (R) titi degradation data suggested 2,4-D
breakdown maximized at a leaf concentration of 57 pg, and
Pretreatment of an (R) titi shoot with 2 y.g of fluorine
as sodium fluoride resulted in a significant reduction in
decarboxylation of 2,4-D-1-14C.
From the foregoing studies it can be concluded that:
1. Major factors contributing to the herbicidal re-
sistance of black titi include its thick, ridgy
cuticle, peculiar arrangement of internal leaf
tissues, and comparatively low and high rates of
2,4-D penetration and decarboxylation, respectively.
2. Enhancement of herbicide penetration is the pro-
bable key to control of black titi, as economical
levels of 2,4-D were apparently leaf-toxic when
applied via the transpiration stream.
3. As a practical measure, fluoride ion supplied sep-
arately to black titi may lead to a significant
reduction in the toxic concentration indicated for
control of this species.
Bach, M.K. 1961. Metabolites of 2,4-dichlorophenoxyacetic
acid from bean stems. Plant Physiol. 36:558-565.
Bach, M.K. and J. Fellig. 1961. Correlation between in-
activation of 2,4-dichlorophenoxyacetic acid and cessation
of callus growth in bean stem sections. Plant Physiol.
Badiei, A.A., E. Basler, and P.W. Santelmann. 1966. As-
pects of movement of 2,4,5-T in blackjack oak. Weeds,
Basler, E. 1964. The decarboxylation of phenoxyacetic
acid herbicides by excised leaves of woody plants. Weeds,
Behrers, R.W. 1964. The physical and chemical properties
of surfactants and their effects on formulated herbicides.
Blackman, G.E., N.G. Templeton, and D.J. Halliday. 1951.
Herbicides and selective toxicity. Ann. Rev. Plant
Casida, J.E. and L. Lykken. 1969. Metabolism of organic
pesticide chemicals in higher plants. Ann. Rev. Plant
Crafts, A.S. 1953. Herbicides. Ann. Rev. Plant Physiol.
Crafts, A.S. and S. Yamaguchi. 1958. Comparative tests on
the uptake and distribution of labelled herbicides by
Zebrina pendula and Tradescantia fluminensis. Hilgardia,
Crafts, A.S., H.B. Currier, and H.R. Drever. 1958. Some
studies on the herbicidal properties of maleic hydrazide.
Crosby, D.G. 1964. Metabolites of 2,4-dichlorophenoxyacetic
acid (2,4-D) in bean plants. J. Agr. and Food Chem. 12:3-6.
Currier, H.B. and C.D. Dybing. 3959. Foliar penetration of
herbicides--review and present status. Weeds, 7:195-213.
Edgerton, L.J. and M.E. Hoffman. 1961. Fluorine substitu-
tion affects decarboxylation of 2,4-dichlorophenoxyacetic
acid in apple. Science, 134:341-342.
Esau, K. 1953. Plant Anatomy. Wiley and Sons, New York.
Fang, S.C. and J.S. Butts. 1954. Studies in plant metab-
olism. III. Absorption, translocation and metabolism
of radioactive 2,4-D in corn and wheat plants. Plant
Fang, S.C. 1958. Absorption, translocation, and metabolism
of 2,4-D-l-14C in pea and tomato plants. Weeds, 6:179-186.
Faulkner, J.K. and D. Woodcock. 1964. Metabolism of 2,4-
dichlorophenoxyacetic acid (2,4-D) by Aspergillus niger
van Tiegh. Nature, 203:865.
Fites, R.C., F.W. Slife, and J.B. Hanson. 1964. Trans-
location and metabolism of radioactive 2,4-D in jimsonweed.
Foy, C.L. 1962. Penetration and initial translocation of
2,2-dichloropropionic acid (Dalapon) in individual leaves
of Zea Mays L. Weeds, 10:35-39.
Foy, C.L. and L.W. Smith. 1965. Surface tension lowering
wettability of paraffin and corn leaf surfaces, and herbi-
cidal enhancement of dalapon by seven surfactants. Weeds,
Franke, W. 1967. Mechanisms of foliar penetration of solu-
tions. Ann. Rev. Plant Physiol. 18:281-300.
Hilton, J.L., L.L. Jansen, and H.M. Hull. 1963. Mechanisms
of herbicide action. Ann. Rev. Plant Physiol. 14:353-384.
Holly, K. 1964. Herbicide selectivity in relation to form-
ulation and application methods, p. 423-464. In L.J. Audus
(ed.), The Physiology and Biochemistry of Herbicides.
Academic Press, London and New York. 555 p.
Jansen, L.L., WA. Gentner, and W.C. Shaw. 1961. Effects
of surfactants on the herbicidal activity of several
herbicides in aqueous spray systems. Weeds, 9:381-405.
Jansen, L.L. 1964. Surfactant enhancement of herbicide
entry. Weeds, 12:251-255.
Jansen, L.L. 1965a. Effects of structural variations in
ionic surfactants on phytotoxicity and physical-chemical
properties of aqueous sprays of several herbicides.
Jansen, L.L. 1965b. Herbicidal and surfactant properties
of long-chain alkylamine salts of 2,4-D in water and oil
sprays. Weeds, 13:123-130.
Jaworski, E.G. and J.S. Butts. 1952. Studies in plant
metabolism. II. The metabolism of 14C labelled 2,4-
dichlorophenoxyacetic acid in bean plants. Arch. Biochem.
Jaworski, E.G., Fang, S.C., and V.H. Freed. 1955. Studies
in plant metabolism. V. The metabolism of radioactive
2,4-D in etiolated bean plants. Plant Physiol. 30:272-
Johansen, D.A. 1940. Plant Microtechnique. McGraw-Hill,
New York and London. 523 p.
Leopold, A.C. 1964. Plant Growth and Development. McGraw-
Hill, New York. 466 p.
Luckwill, E.C., and C.P. Lloyd-Jones. 1960a. Metabolism of
plant growth regulators. I. 2,4-dichlorophenoxyacetic
acid in leaves of red and black currant. Ann. Appl. Biol.
Luckwill, E.C., and C.P. Lloyd-Jones. 1960b. Metabolism of
plant growth regulators. II. Decarboxylation of 2,4-
dichlorophenoxyacetic acid in leaves of apple and straw-
berry. Ann. Appl. Biol. 48:626-636.
Morgan, P.W. and W.C. Hall. 1962. Effect of 2,4-dichloro-
phenoxyacetic acid on the production of ethylene by cotton
and grain sorghum. Physiol. Plant. 15:420-427.
Morton, H.L. 1966. Influence of temperature and humidity
on foliar absorption, translocation, and metabolism of
2,4,5-T by mesquite seedlings. Weeds, 14:136-141.
Mueller, L.E., P.H. Carr, and W.E. Loomis. 1954. The sub-
microscopic structure of plant surfaces. Am. J. Botany,
Nash, A.J. 1965. Statistical Techniques in Forestry.
Lucas Bro., Columbia, Mo. 146 p.
Norman, A.G., C.E. Minarik, and R.L. Weintraub. 1950.
Herbicides. Ann. Rev. Plant Physiol. 1:141-168.
Norris, R.F. and M.J. Bukovac. 1968. Structure of the pear
leaf cuticle with reference to cuticular penetration. Ann.
J. Botany, 55:975-983.
Norris, R.F. and M.J. Bukovac. 1969. Some physical-kinetic
considerations in penetration of naphthaleneacetic acid
through isolated pear leaf cuticle. Physiol. Plant. 22
Overbeek, van, J. 1956. Absorption and translocation of
plant regulators. Ann. Rev. Plant Physiol. 7:355-372.
Radwan, M.A., C.R. Stocking, and H.B. Currier. 1960. Histo-
autoradiographic studies of herbicidal translocation.
Schieferstein, R.H. and W.E. Loomis. 1956. Wax deposits on
leaf surfaces. Plant Physiol. 31:240-247.
Schieferstein, R.H. and W.E. Loomis. 1959. Development of
the cuticular layers in angiosperm leaves. Am. J. Botany,
Snedecor, G.W. and W.G. Cochran. 1956. Statistical Methods.
Iowa State Univ. Press. 534 p.
Thomas, E.W., Loughman, B.C., and R.G. Powell. 1963. Hydrox-
ylation of phenoxyacetic acid by stem tissue of Avena
sativa. Nature, 199:73-74.
Thomas, E.W., Loughman, B.C., and R.G. Powell. 1964a. Meta-
bolic fate of some chlorinated phenoxyacetic acids in the
stem tissue of Avena sativa. Nature, 204:286.
Thomas, E.W., Loughman, B.C., and R.G. Powell. 1964b. Meta-
bolic fate of 2,4-dichlcrophenoxyacetic acid in the stem
tissue of Phaseolus vulqaris. Nature, 204:884-885.
Weintraub, R.L. 1953. Metabolism of 2,4-D by microorgan-
isms and higher plants. Proc. NCWCC, 10:6-.
Wilcox, M., Moreland, D.E., and G.C. Klingman. 1963. Aryl
hydroxylation of phenoxyaliphatic acids by excised roots.
Physiol. Plant. 16:565-571.
Williams, M.C., F.W. Slife, and J.B. Hanson. 1960. Absorp-
tion and translocation of 2,4-D in several annual broad-
leaved weeds. Weeds, 8:657-665.
Woodford, E.K., K. Holly, and C.C. McCready. 1958. Herbi-
cides. Ann. Rev. Plant Physiol., 9:311-358.
Walter Leon Beers, Jr. was born June 28, 1927, in
Greensburg, Westmoreland County, Pennsylvania, the son of
Margaret E. and Walter L. Beers, Sr. He attended the
Greensburg public schools and was graduated from the
Greensburg High School in June, 1945. Immediately there-
after he enlisted in the United States Marine Corps for a
period of four years.
In July, 1949, he married Barbara Jean Merrill. They
have three children, Walter Bradley, Barry Leon, and
In September, 1949, he entered the Pennsylvania State
University and in February, 1953, he was conferred the
degree of Bachelor of Science with a major in Forestry.
He then entered the Graduate School of the same institution
for one term, majoring in Forest Management.
In June, 1953, he accepted employment with The Buckeye
Cellulose Corporation, Perry, Florida, and until now has
worked as a Research Forester in the Woodlands Department.
He entered the University of Florida Graduate School
in February, 1962, majoring first in Botany and later in
Agronomy. The degree of Doctor of Philosophy was conferred
on him in December, 1969.
He is a member of the Society of American Foresters,
Kiwanis International, and Xi Sigma Pi, Gamma Sigma Delta,
Phi Sigma, and Alpha Zeta honorary organizations.
This dissertation was prepared under the direction of
the chairman of the candidate's supervisory committee and
has been approved by all members of that committee. It was
submitted to the Dean of the College of Agriculture and to
the Graduate Council, and was approved as partial fulfill-
ment of the requirements for the degree of Doctor of
CfDean, College of Agriculture
Dean, Graduate School