Bulletin 247
UNIVERSITY OF FLORIDA
AGRICULTURAL EXPERIMENT STATION
Wilmon Newell, Director
Variation in the Tung-Oil Tree
By HAROLD MOWRY
Fig. 1.-The cluster fruiting habit is characteristic of the Florida variety.
This cluster is from a budded tree.
TECHNICAL BULLETIN
Bulletins will be sent free upon application to the
AGRICULTURAL EXPERIMENT STATION,
GAINESVILLE, FLORIDA
LlM ARY
FLORIDA EXPERIMEN-r ,- ...
May, 1932
EXECUTIVE STAFF
John J. Tigert, M.A., LL.D., President of the
University
Wilmon Newell, D.Sc., Director
it. Harold Hume, M.S., Asst. Dir., Research
Sam T. Fleming, A.B., Asst.Dir., Administration
J. Francis Cooper, M.S.A., Editor
R. M. Fulghum, B.S.A., Assistant Editor
Ida Keeling Cresap, Librarian
Ruby Newhall, Secretary
K. H. Graham, Business Manager
fachel McQuarrie, Accountant
MAIN STATION, GAINESVILLE
AGRONOMY
W. E. Stokes, M.S., Agronomist
W. A. Leukel, Ph.D., Associate
G. E. Ritchey, M.S.A., Assistant*
Fred H. Hull, M.S., Assistant
J. D. Warner, M.S., Assistant
John P. Camp, M.S., Assistant
ANIMAL HUSBANDRY
A. L. Shealy, D.V.M., Veterinarian in Charge
E. F. Thomas, D.V.M., Assistant Veterinarian
R. B. Becker, Ph.D., Associate in Dairy Inves-
tigations
W. M. Neal, Ph.D., Asst. in Animal Nutrition
P. T. Dix Arnold, B.S.A., Assistant in Dairy In-
vestigations
CHEMISTRY
R. W. Ruprecht, Ph.D., Chemist
R. M. Barnette, Ph.D., Associate
C. E. Bell, M.S., Assistant
J. M. Coleman, B.S., Assistant
It. W. Winsor, B.S.A., Assistant
H. W. Jones, M.S., Assistant
ECONOMICS, AGRICULTURAL
C. V. Noble, Ph.D., Agricultural Economist
Bruce McKinley, A.B., B.S.A., Associate
M. A. Brooker, Ph.D., Associate
Zach Savage, M.S.A., Assistant
ECONOMICS, HOME
Ouida Davis Abbott, Ph.D., Head
L. W. Gaddum, Ph.D., Biochemist
C. F. Ahmann, Ph.D., Physiologist
ENTOMOLOGY
J. R. Watson, A.M., Entomologist
A. N. Tissot, Ph.D., Assistant
H. E. Bratley, M.S.A., Assistant
E. F. Grossman, M.A., Asso., Cotton Insects
P. W. Calhoun, Assistant, Cotton Insects
HORTICULTURE
A. F. Camp, Ph.D., Horticulturist
Harold Mowry, B.S.A., Associate
M. R. Ensign, M.S., Associate
A. L. Stahl, Ph.D., Assistant
G. H. Blackmon, M.S.A., Pecan Culturist
C. B. Van Cleef, M.S.A., Greenhouse Foreman
PLANT PATHOLOGY
W. B. Tisdale, Ph.D., Plant Pathologist
George F. Weber, Ph.D., Associate
R. K. Voorhees, M.S., Assistant
Erdman West, M.S., Mycologist
*In cooperation with U.S.D.A.
BOARD OF CONTROL
P. K. Yonge, Chairman, Pensacola
A. H. Blanding, Bartow
W. B. Davis, Perry
Raymer F. Maguire, Orlando
Frank J. Wideman, West Palm Beach
J. T. Diamond, Secretary, Tallahassee
BRANCH STATIONS
NORTH FLORIDA STATION, QUINCY
L. O. Gratz, Ph.D., Associate Plant Pathologist
in Charge.
R. R. Kincaid, M.S., Asst. Plant Pathologist
W. A. Carver, Ph.D., Asso. Cotton Specialist
R. M. Crown, B.S.A., Asst. Agronomist, Cotton
Jesse Reeves, Farm Superintendent
CITRUS STATION, LAKE ALFRED
John H. Jefferies, Superintendent
Gee. D. Ruehle, Ph.D., Asst. Plant Pathologist
W. A. Kuntz, A.M., Asst. Plant Pathologist
B. R. Fudge, Ph.D., Assistant Chemist
W. L. Thompson, B.S., Assistant Entomologist
EVERGLADES STATION, BELLE GLADE
R. V. Allison, Ph.D., Soils Specialist in Charge
R. W. Kidder, B.S., Farm Foreman
R. N. Lobdell, M.S., Associate Entomologist
F. D. Stevens, B.S., Sugarcane Agronomist
H. H. Wcdgeworth, M.S., Asso. Plant Path.
B. A. Bourne, M.S., Associate Sugarcane Physi-
ologist
J. R. Neller, Ph.D., Associate Biochemist
A. Daane, Ph.D., Associate Agronomist
M. R. Bedsole, M.S.A., Assistant Chemist
SUB-TROPICAL STATION, HOMESTEAD
H. S. Wolfe. Ph.D., Asso. Horticulturist in Clig
L. R. Toy, B.S.A., Assistant Horticulturist
Stacy 0. Hawkins. M.A., Assistant Plant
Pathologist
FIELD STATIONS
Leesburg
M. N. Walker, Ph.D., Asso. Plant Pathologist
W. B. Shippy, Ph.D., Asst. Plant Pathologist
K. W. Loucks, M.S., Asst. Plant Pathologist
C. C. Goff, M.S., Assistant Entomologist
J. W. Wilson, Ph.D., Assistant Entomologist
Plant City
A. N. Brooks. Ph.D., Asso. Plant Pathologist
R. E. Nolen, M.S.A., Field Asst. in Plant Path.
Cocoa
A. S. Rhoads, Ph.D., Asso. Plant Pathologist
Hastings
A. H. Eddins, Ph.D., Asso. Plant Pathologist
West Palm Beach
D. A. Sanders, D.V.M., Associate Veterinarian
Monticello
Fred W. Walker, Assistant Entomologist
Bradenton
David G. Kelbert, Asst. Plant Pathologist
CONTENTS
PAGE
IN TRODU CTION .......... .. ............................ ....... 5
SCOPE OF INVESTIGATIONS ................ ... ........ ............ 5
FLOWERING HABIT ............. ..... ............................... 7
FRUIT SIZES AND WEIGHTS .......................................... 7
NUMBER OF SEEDS PER FRUIT AND PER POUND ........................ 11
RATIO OF SEED AND HULL IN WHOLE FRUITS ........................... 13
PROPORTION OF KERNEL AND TESTA IN SEED .......................... 13
OIL CONTENT ............ ... ....................................... 14
Y IELDS ............ .............................................. 15
BEHAVIOR OF BUDDED TREES AND SEEDLINGS OF KNOWN PARENTAGE ...... 17
ANALYSIS OF FRUITS FROM RAIFORD STATE FARM ..................... 22
V ARIETIES ........... ...... ............................ ......... 25
SUMMARY .... ................... .... ............................ 31
1 2
3 4
Fig. 2.-Variations in size and external appearance of typical mature
fruits of different seedling trees. The numerals correspond to the tree
numbers of Tables I and II. The internal characters of the same fruits are
shown in Fig. 3.
1 2 4 5
.4.;~~~
Fig. 3.-Variations in size and internal appearance of typical mature
fruits of different seedling trees. The numerals correspond to the tree
numbers of Tables I and II. The external characters of the same fruits
are shown in Fig. 2.
VARIATION IN THE TUNG-OIL TREE
By HAROLD MOWRY
The progress made in the commercial development of tung-oil
growing is indicative of the active interest manifest in the possi-
bilities of this new crop. Since the initial plantings of 1923, the
Florida acreage of tung-oil trees has increased to a present total
of about 10,000 acres. This acreage quite probably is but a fore-
runner of much heavier planting, if present groves show anything
like anticipated results in approaching maturity.
Like nearly all first commercial ventures in the growing of the
various fruit and nut trees, the whole of the tung-oil acreage
now planted is of seedling stock. A large part is from seeds
obtained from any available source, little or no attention having
been given to the desirability or productiveness of the parent
trees. The condition is not due to a lack of diligence but to a
scarcity of information at the time of planting as to the degree
of variation existent in seedling trees. So few bearing trees have
been available for study that it is only recently that any com-
parisons were possible between parent and seedling or parent and
scion or of differences in the fruits and tree behavior of various
seedlings.
Heavy yield has proved to be the most desired characteristic in
the tung-oil tree. There are other factors, however, influencing
oil production that should be taken into consideration, since there
is considerable variation in size of fruit and seed, relative weights
of seed and hull, ratios of weight of kernel to testa (seed shell or
coat), and percentages of oil in the seed. The primary object is
the production of oil in maximum amounts and to attain it the
planting must necessarily consist of stock that will produce the
greatest amount of oil per tree.
As more groves have reached a bearing age it has become
apparent, even from casual observation, that wide variation exists
in seedling trees. The visual variation is manifested chiefly in
the number, size and shape of the fruits, and the single or cluster
habits of fruiting. Some minor differences in foliage and in
blossom color also have been noted.
SCOPE OF INVESTIGATIONS
The purpose of the studies outlined herein has been to deter-
mine and evaluate the extent of any differences occurring in tung-
oil (Aleurites fordi Hemsl.) trees and fruit that the most desir-
VARIATION IN THE TUNG-OIL TREE
By HAROLD MOWRY
The progress made in the commercial development of tung-oil
growing is indicative of the active interest manifest in the possi-
bilities of this new crop. Since the initial plantings of 1923, the
Florida acreage of tung-oil trees has increased to a present total
of about 10,000 acres. This acreage quite probably is but a fore-
runner of much heavier planting, if present groves show anything
like anticipated results in approaching maturity.
Like nearly all first commercial ventures in the growing of the
various fruit and nut trees, the whole of the tung-oil acreage
now planted is of seedling stock. A large part is from seeds
obtained from any available source, little or no attention having
been given to the desirability or productiveness of the parent
trees. The condition is not due to a lack of diligence but to a
scarcity of information at the time of planting as to the degree
of variation existent in seedling trees. So few bearing trees have
been available for study that it is only recently that any com-
parisons were possible between parent and seedling or parent and
scion or of differences in the fruits and tree behavior of various
seedlings.
Heavy yield has proved to be the most desired characteristic in
the tung-oil tree. There are other factors, however, influencing
oil production that should be taken into consideration, since there
is considerable variation in size of fruit and seed, relative weights
of seed and hull, ratios of weight of kernel to testa (seed shell or
coat), and percentages of oil in the seed. The primary object is
the production of oil in maximum amounts and to attain it the
planting must necessarily consist of stock that will produce the
greatest amount of oil per tree.
As more groves have reached a bearing age it has become
apparent, even from casual observation, that wide variation exists
in seedling trees. The visual variation is manifested chiefly in
the number, size and shape of the fruits, and the single or cluster
habits of fruiting. Some minor differences in foliage and in
blossom color also have been noted.
SCOPE OF INVESTIGATIONS
The purpose of the studies outlined herein has been to deter-
mine and evaluate the extent of any differences occurring in tung-
oil (Aleurites fordi Hemsl.) trees and fruit that the most desir-
Florida Agricultural Experiment Station
able types of trees might be more accurately selected; to deter-
mine whether certain variations noted in different trees were
inherent or due to environmental influences; and to ascertain
whether desired characters might be perpetuated through seeds
or by vegetative methods of propagation.
Twenty-three trees of the Station plantings were chosen for
study, 10 of them being about 18 years of age and the others in
their seventh, eighth, and ninth years from the time of planting.
All trees of the Station plantings had been regularly fertilized
and with the younger trees a system of summer cover-cropping
with legumes had been followed. Analyses of fruits from nine
seven-year-old trees growing at the Florida State Farm at Rai-
ford also are included. The soils of both plantings are of the
Norfolk series-that of the Station planting a deep, medium to
coarse sand and that of the Raiford planting a deep, medium to
fine sand. The older trees (1-10 inclusive) are spaced 10 feet
in the row; all the others 25 feet.
All analyses, other than of records of yield over several years,
were made of the 1931 crop. Although a prolonged drought
occurred during the fall months of that year, it is believed it had
but little effect on fruit size and composition, since all fruits were
nearly, if not wholly, mature before any of the trees gave evidence
of insufficient soil moisture. Any differences from normal in the
fruit would tend to be equalized, as all trees of the Station plant-
ings underwent like weather conditions. That would also hold
true for the Raiford trees as the precipitation for the year of
the Gainesville and Raiford areas did not materially differ. Sea-
sonal conditions prior to the drought were conducive to a thrifty
growth in the trees and the crop produced should be represen-
tative of both yield and quality under favorable conditions. It is
recognized that the data collected cover but a single season and
that they would be materially strengthened were the analyses to
be extended over two or more years. However, it is not considered
likely that fluctuations from year to year are great enough to
justify an indefinite withholding of the results thus far secured
until future collateral investigations may be completed.
The factors considered and on which studies were made include
yield, fruit size and weight, number of seeds per fruit, number
of seeds per pound, percentages of hull and of seed in the whole
fruits, and percentages of kernel, oil, and testa (seed coat) in
the seed. For each tree, unless exception is noted, 200 fruits
taken from all parts of the tree were used for analysis. Measure-
ments were made on matured but not dried nor shrunken fruits,
Bulletin 247, Variation in the Tung-Oil Tree
and the weights, other than oil content of kernels, on air-dried
materials. Percentages of oil in the kernels were calculated on
an oven-dry basis.
FLOWERING HABIT
The flowers of the tung-oil tree are not perfect but are unisexual.
Stamens and pistils are produced in separate flowers but both
flower types are borne on the same tree monoeciouss). They are
borne in large, loose cymes at the terminals of twigs of the pre-
vious season's growth. The pistillate flower, if present, is borne
as the central or apical flower of the cyme and all lateral flowers
are staminate, except in trees of the so-called cluster type wherein
one to several pistillate flowers are produced laterally in a single
flower cluster.
Apparently pollination is amply provided, since the pistillate
flowers are surrounded in the cluster by staminate ones at an
approximate ratio of 60 to 1. Because of the preponderant pro-
portion of staminate blossoms and the cymose type of flower
arrangement, it is possible that there is much more self- or close-
pollination in trees than cross-pollination among them. This in
part may account for a fairly large proportion of seedlings closely
resembling the parent.
In three instances, one on each of three trees, individual flowers
have been found having both stamens and pistil (perfect or her-
maphrodite flowers). No special significance can be attached to
such flowers because they are so rare, and besides, no pollination
troubles have ever been reported. Were there isolated plantings
of a single budded variety, it is possible that effective pollination
would be hindered but such a condition is not believed probable
since no trouble has been experienced in the setting of fruit from
blossoms that were hand-pollinated with pollen from the same
tree.
Another floral difference noted and found to be transmitted to
the seedlings was the absence of the internal pink or reddish
coloration on the petals in the throat of the flower. The trees
with the wholly white or faintly yellow blossoms seemingly do not
differ in other respects from the ordinary seedling.
FRUIT SIZES AND WEIGHTS
Tung-oil fruits vary considerably in shape, the range for most
being from oblate to oblong or ovate, but some are nearly spherical
while others are inverted pyriform in outline (Fig. 4). Most are
more or less necked at the stem end and nippled at the apex. The
Bulletin 247, Variation in the Tung-Oil Tree
and the weights, other than oil content of kernels, on air-dried
materials. Percentages of oil in the kernels were calculated on
an oven-dry basis.
FLOWERING HABIT
The flowers of the tung-oil tree are not perfect but are unisexual.
Stamens and pistils are produced in separate flowers but both
flower types are borne on the same tree monoeciouss). They are
borne in large, loose cymes at the terminals of twigs of the pre-
vious season's growth. The pistillate flower, if present, is borne
as the central or apical flower of the cyme and all lateral flowers
are staminate, except in trees of the so-called cluster type wherein
one to several pistillate flowers are produced laterally in a single
flower cluster.
Apparently pollination is amply provided, since the pistillate
flowers are surrounded in the cluster by staminate ones at an
approximate ratio of 60 to 1. Because of the preponderant pro-
portion of staminate blossoms and the cymose type of flower
arrangement, it is possible that there is much more self- or close-
pollination in trees than cross-pollination among them. This in
part may account for a fairly large proportion of seedlings closely
resembling the parent.
In three instances, one on each of three trees, individual flowers
have been found having both stamens and pistil (perfect or her-
maphrodite flowers). No special significance can be attached to
such flowers because they are so rare, and besides, no pollination
troubles have ever been reported. Were there isolated plantings
of a single budded variety, it is possible that effective pollination
would be hindered but such a condition is not believed probable
since no trouble has been experienced in the setting of fruit from
blossoms that were hand-pollinated with pollen from the same
tree.
Another floral difference noted and found to be transmitted to
the seedlings was the absence of the internal pink or reddish
coloration on the petals in the throat of the flower. The trees
with the wholly white or faintly yellow blossoms seemingly do not
differ in other respects from the ordinary seedling.
FRUIT SIZES AND WEIGHTS
Tung-oil fruits vary considerably in shape, the range for most
being from oblate to oblong or ovate, but some are nearly spherical
while others are inverted pyriform in outline (Fig. 4). Most are
more or less necked at the stem end and nippled at the apex. The
Florida Agricultural Experiment Station
external surface may be quite smooth or may be somewhat irregu-
larly roughened near the apex; on some, distinct longitudinal
ridges are present. Usually the variation on an individual tree
is slight but an occasional seedling bears fruits of widely differing
shapes.
Fig. 4.-Extremes in size and shape of fruits from different seedling trees.
In Table I the trans-
verse measurements of
the fruits are of the ex-
i treme diameter--the lon-
lgitudinal measurements
include the length of the
neck, if present, at the
base of the fruit but not
of the nipple of the apical
end. All linear measure-
ments are in centimeters
(2.54 centimeters=l
inch).
It will be noted in most
instances that the fruits
were of greater breadth
than length, though the
7J I X 7 /a AC $D4 ANr I/ opposite was true of the
Fig. 5.-Transverse diameter of fruits. fruits of a few trees.
TABLE I.-FRUIT ANALYSES OF SEEDLING TUNG-OIL TREES.
Av. Av. Av. Fruit Fruit Seed
Fruit- Weight No. No. Mature-Not dried Air-dry Air-dry Kernel
Tree ing Indv. Seed Seed
Type* Fruits Per Per lb. Mean Mean Percent Percent Percent Percent Percent Percent
(Air-dry Fruit (Air-dry) Breadth Length Seed Iull Kernel Testa Mois- Oil
(oz.) (cm.) (c.) ture Oven-dry)
1 S .67 4.38 167 4.82 .010 4.71 + .010 62.6 37.4 59.6 40.4 5.64 60.92
2 Cl .74 4.44 143 5.45 + .014 5.01 + .012 63.0 37 0 59 9 40.1 8.17 58.50
3 1.44 4.11 115 5.92 + .018 5.47 + .014 55.0 45 0 65 3 34.7 6.20 60.89
4 S .44 4.15 227 3.48 + .010 3.76 + .009 63.6 36 4 58.9 41.1 5.07 62.08
5 -S 79 5.09 176 4.57 + .014 4.31 .012 5.6 41.4 57 4 42.6 5.68 59.71
6 S** .87 4.41 129 5.25 + 014 5.06 .012 63.0 30 37.4 5.80 59.09
7 S .86 4.48 133 5.36 018 4.99 .012 63.1 36.9 63.3 36.7 5.75 59.34
8 S 1.36 4.56 97 6.65 + .015 5.96 .013 54 5 45.5 65.1 34.9 6.00 60.65
9 S .78 4.54 171 5.51 + .011 5.63 .011 60.7 39.3 56 0 4.0 53.97
10 S 1.34 4.36 6.29 + .015 5.63 .013 60 9 1 34 10 72 60.54
A Cl .90 4.56 128 5.26 + 014 4.92 + .011 63.3 36 7 61.5 38.5 9.05 60.14
B S 1.05 4.28 130 5.92 + .022 5.68 .017 50 2 49.8 62.5 375 6.36 55.51 m
C Cl .75 -4.38 154 5.47 + .016 5 16 .018 60.6 39 4 54.5 45.5 650 56.52 Z
D Cl .95 4.38 111 5.47 + .022 4.96 .011 66.4 33.6 63.6 36.4 10.39 58.43
~-~-----I57--t- -
E' 7S 1.11 4.54 108 6.31 + .012 5.80 + .010 61.0 390 62 8 37.2 6.76 58.97
F S 1.10 4.19 107 6.00 .014 5.75 + .010 57.5 42.5 60.9 39.1 5.15 58.27
G Cl .77 4.39 147 5.52 + .016 5.28 + .014 62.2 37.8 59.9 40.1 6.63 60.15
H Craig 2.08 7.28 105 8.03 + .026 6.07 + .017 53.5 46.5 61.1 38.9 4.83 56.63 w
Trees 1-10, inclusive, 18 years old; A-E, inclusive, 9 years; F, G, and H, 8 years.
*Cluster or single.
**Has twice borne a small part of its fruit in clusters.
Florida Agricultural Experiment Station
Among fruits of the dif-
ferent trees the mean
transverse and longitudi-
nal diameters were found
to vary widely (Figs. 2, 3,
5 and 6). The mean
transverse diameter va-
ried from 3.48 cm. (Tree
4) to 6.65 cm. (Tree 8)
Sand the mean longitudinal
diameter from 3.76 cm.
(Tree 4) to 5.96 cm. (Tree
8). The Craig variety
r4 a F Yr(7 o A cO~ ~ had a mean transverse
diameter of 8.03 cm. and
Tig. 6.-Longitudinal diameter of fruits. a mean longitudinal di-
ameter of 6.07 cm.
The average weight in
the several trees of single,
air-dried fruits varied
from .44 ounces (Tree 4)
to 1.44 ounces (Tree 3).
Like fruits of the Craig
variety had an average
weight of 2.08 ounces
(Fig. 7).
The thickness of the
undried hulls of fruit
from different trees also
varied materially but on
thorough drying the dif-
ference was considerably
1~a 3 ~' 7*s a Go a _c lessened. The thickness
Fig. 7.-Weight of individual air-dried of the dried hulls, taken
fruits. at the point of minimum
thickness, was from .15
cm. to .35 cm. It follows that in most instances the percentage
weight of hull in the fruit increases in direct ratio to its thickness
and the percentage weight of seed is proportionately lowered.
Bulletin 247, Variation in the Tung-Oil Tree
NUMBER OF SEEDS PER
FRUIT AND PER POUND
In tung-oil blossoms
the ovaries are usually
o.. four- or five-celled and, if
fertilized, result in either
four or five seeds to the
fruit. Partial fertiliza-
tion, wherein all ovules
are not fertilized, may be
the cause of fewer seeds
oo than cells in the ovary.
The number of seeds
in the individual fruits
seemed one of the most
constant characters, al-
though the number va-
ried among different
-rEL./ a r MO- 7 r ~w so Cu r trees (Tables IandII and
Nub.er MA. r-Dried Seed.. Pr PoFigs. 21 and 22). Infre-
l ig. 8.-Number of air-dried seed per pound. quently, trees have been
found having six or seven
seeds in some of their
fruits and, in single fruits
from fasciated twigs of
the Craig variety, 20 or
more seeds have been
noted. Exclusive of the
Craig, the greatest aver-
age number of seeds per
fruit for any one of the
trees included in the
study was 5.09 and the
.1 least 4.11. The Craig
averaged 7.28 seeds per
fruit.
There was a direct cor-
-aL ) 7 I7 I. o4 A o Lf relation between size of
well-filled seeds and the
Fig. 9.-Percent seed in air-dry fruit. number of seeds per
pound; the number in in-
verse ratio to the size. The extreme variation in number of seeds
per pound was from 85 (Tree 10) to 227 (Tree 4). (Fig. 8.)
TABLE II.-OIL CONTENT AND YIELD OF TUNG-OIL FRUITS AND SEEDS.
Oil percentages Fruit Yield** Percent
________ !- __________----------------------------------P percent
of Kernel of Seed of Fruit 19271 1928 1929 1931 Total 4 years of 4-year L
Tree* -- total yield
(oven-dry (air-dry (air-dry lb b. b. o. b. o. b. of the
basis) basis) basis) 10 trees
1 60.92 34.26 21.45 17 8 14 5 37 6 38 1 107 4 7.35
2 58.50 32.18 20.27 55 0 25 8 89 12 72 8 242 12 16.64
__ ______________________________ 0
3 60.89 37.31 20.52 13 0 2 7 4 14 14 0 34 5 2.34 2
4 62.08 34.74 22.09 20 0 22 11 31 0 30 12 104 7 7.16
5 58.48 31.68 18.56 5 0 4 8 4 6 7 10 21 8 1.47
6 59.11 34.52 21.75 50 0 11 5 66 14 64 1 192 4 13.17
7 59.34 35.42 22.35 15 0 5 4 35 12 19 3 75 3 5.15
8 60.65 36.32 19.79 48 0 23 5 39 5 38 9 149 3 10.22
9 53.97 28.07 17.04 90 1 0 84 12 141 11 164 13 481 4 32.97
10 60.54 35.67 21.72 14 0 9 2 5 0 23 6 51 8 3.53
A 60.14 33.65 21.30 1 12 11 9 27 6 40 12 81 7
B 55.51 32.49 16.31 0 3 4 3 16 14 7 13 29 1
C 56.52 28.80 17.45 2 115 5 12 16 14 21 12 47 5
D 58.43 33.30 22.11 4 1 9 8 20 0 44 5 75 14 o
E 58.97 34.53 21.06 7 4 12 1 16 3 26 6 61 14
F 58.27 33.96 19.53 5 14 8 9 24 11 55 9 94 11
G 60.15 33.66 20.94 6 6 5 0 34 15 58 2 104 7
H 56.63 32.95 17.63 3 4 2 15 8 6 15 9 30 2
*Trees of this Table are the same as those of Table I.
**Weights are of hulled, air-dried seed.
Bulletin 247, Variation in the Tung-Oil Tree
RATIO OF SEED AND HULL IN WHOLE FRUITS
With other factors equal, the fruit having the greatest pro-
portion of seed to hull, by weight, would be the most desirable.
No apparent advantage was found in the trees producing thick-
hulled fruit. (Figs. 4 and 20.)
In the different trees of the Station plantings the highest per-
centage of seed found in the whole, air-dried fruit was 66.4 per-
cent and the lowest 50.2 percent, an extreme difference of 16.2
percent (Table I and Fig. 9). If comparison is made of the seed
weight percentages of trees given in Tables I and IV it becomes
evident that to maintain a high proportional weight of seed to
hull the trees must have suitable cultural attention in the form
of cultivation and fertilization, as well as sufficient soil moisture.
PROPORTION OF KERNEL AND TESTA IN SEED
Although the fruits of the tung-oil tree cannot be correctly
termed nuts, the seeds might popularly fall into that classification
since they have a hard bony covering comparable to the shell of a
nut. The seed-coat, or testa, has been designated by some as the
pellicle but, since pellicle is defined as a papery skin or delicate
superficial membrane, the word certainly is neither descriptive
nor appropriate. A clear distinction should be drawn between the
outer hull or husk which envelops the seeds and the testa, or
seed-coat, containing the kernel.
With a variation in
70 weight of 54.5 to 66.0 per-
cent kernel in the seed,
4o or 34.0 to 45.5 percent
testa, the widest differ-
0o ~ence among the several
trees was 11.5 percent
between the highest and
lowest (Fig. 10). Gen-
30 erally, although not by
any means invariably,
there was a fairly close
correlation between size
of seed and percent ker-
nel; that is, the larger
the seed the greater per-
fL / > .J v J 7 @ 4o a D E o F.* a cent of kernel contained.
Fig. 10.-Percent kernel in air-dry seed. No apparent correlation
Bulletin 247, Variation in the Tung-Oil Tree
RATIO OF SEED AND HULL IN WHOLE FRUITS
With other factors equal, the fruit having the greatest pro-
portion of seed to hull, by weight, would be the most desirable.
No apparent advantage was found in the trees producing thick-
hulled fruit. (Figs. 4 and 20.)
In the different trees of the Station plantings the highest per-
centage of seed found in the whole, air-dried fruit was 66.4 per-
cent and the lowest 50.2 percent, an extreme difference of 16.2
percent (Table I and Fig. 9). If comparison is made of the seed
weight percentages of trees given in Tables I and IV it becomes
evident that to maintain a high proportional weight of seed to
hull the trees must have suitable cultural attention in the form
of cultivation and fertilization, as well as sufficient soil moisture.
PROPORTION OF KERNEL AND TESTA IN SEED
Although the fruits of the tung-oil tree cannot be correctly
termed nuts, the seeds might popularly fall into that classification
since they have a hard bony covering comparable to the shell of a
nut. The seed-coat, or testa, has been designated by some as the
pellicle but, since pellicle is defined as a papery skin or delicate
superficial membrane, the word certainly is neither descriptive
nor appropriate. A clear distinction should be drawn between the
outer hull or husk which envelops the seeds and the testa, or
seed-coat, containing the kernel.
With a variation in
70 weight of 54.5 to 66.0 per-
cent kernel in the seed,
4o or 34.0 to 45.5 percent
testa, the widest differ-
0o ~ence among the several
trees was 11.5 percent
between the highest and
lowest (Fig. 10). Gen-
30 erally, although not by
any means invariably,
there was a fairly close
correlation between size
of seed and percent ker-
nel; that is, the larger
the seed the greater per-
fL / > .J v J 7 @ 4o a D E o F.* a cent of kernel contained.
Fig. 10.-Percent kernel in air-dry seed. No apparent correlation
Florida Agricultural Experiment Station
existed, however, between kernel weight and the percentage
weight of seed or of hull in the fruit.
It has been noted that in some years there was a small propor-
tion of seeds that contained poorly filled, shriveled, or, in some
instances, no kernels. Again, under-nourishment, drought, poor
drainage, or other adverse conditions may result in light-weight
kernels instead of the solid, plump ones of seeds from vigorous
trees. A comparison of kernel percentage data in Tables I and
IV will show existent variation due to some of such causes.
OIL CONTENT
The oil content of the seed kernels from the different Station
trees analyzed was found to vary from 53.97 percent to 66.31
percent (oven-dry basis), a maximum difference of 12.34 percent.
When calculated on an air-dry basis the oil content of the whole
seed varied from 28.07 percent to 40.62 percent, and of the whole
fruit from 16.23 percent to 24.53 percent. It is significant that
the maximum differences found in oil content of the whole air-
dried fruits of any trees of Station plantings were between those
of a seedling (K1) and a budded branch (K2) of the same tree;
the former had the lowest and the latter the highest percentage
of oil. Data presenting oil
7.. percentages of Station
S: plantings are given in
Tables II and III and
Figures 11 and 12, and of
the Raiford plantings in
Table IV. There was a
VI marked difference in the
oil content of fruits of the
two plantings in most in-
3o stances, the differences
probably being due in
part to soil variations but
mainly to unlike cultural
practices.
so Because of variation
always present in plant
materials, it is probable
f4 10 2Ii# 7 e 8'o 4 OoE ACr that a difference in oil
content of less than 3 or
Fig. 11.-Percentage of oil in kernel, oven- 4 percent should not be
dry basis.
Bulletin 247, Variation in the Tung-Oil Tree
considered as significant. Samples of kernels from most trees
listed in Table II, it will be noted, were within the comparatively
narrow range of 58 to 61 percent.
The method followed in determining the percentage of oil was
as follows: Mature fruits, 200 in number, were collected from all
parts of the tree, the
seeds were removed from
the hull and allowed to
become thoroughly air-
dried. The seeds were
then well mixed and 50
were taken as a sample
for analysis. After re-
moval of the test (the
hard, bony seed-coat) the
o kernels were ground to a
meal in a Wiley mill and
the meal of each lot was
separately stored, until
extraction, under a 24-
inch vacuum in glass.
Eight to 10 gram samples
were taken, the moisture
s removed in a vacuum
oven at 70'C., and the oil
then extracted with ether.
/rqo t J r 6 7 1 / 'oA 0 C o 0 C
YIELDS
Fig. 12.-Percentage of oil in seed, air-dry In the culture of the
basis.
tung-oil tree, as with
other tree crops, heavy yields are vital to profitable production.
Soil type, soil moisture, and cultural practices may exert a strong
influence on the size of the crop but the trees' fruiting habits also
must be included among the major contributing factors affecting
yield. An inherently heavy-fruiting type of tree is required to
secure maximum yields, since, irrespective of soil fertility and
suitability of cultural practices, a tree of low-yielding capacity
will continue year after year as such.
Seedling trees vary widely in fruitfulness, as is strongly empha-
sized by tree yield records over a period of years. Production
data of the several Station trees for the years 1927, 1928, 1929,
and 1931 are presented in Table II. The yields for 1930 are not
Florida Agricultural Experiment Station
included as they were severely curtailed by a late frost occurring
when the trees were in bloom. That frost was not general, as the
bloom of trees less than three miles distant was but slightly
affected.
The yields of Trees 1 to 10 inclusive1, with an average age of
18 years, are of interest in that for the above four-year period
two trees-2 and 9-produced as much as the combined yields of
1eEiE t 1 .3Y 4 i f 7/
Fig. 13-Comparative yields in pounds
of air-dried seeds of 18-year-old
trees; black, 1931 yield; light, total,
yield of 1927, 1928, 1929, and 1931.
the other eight trees of
the planting. By follow-
ing the size of the crop
borne by each tree year
after year, the scanty
bearing habit of some and
the fruitfulness of oth-
ers is strikingly demon-
strated (Fig. 13). It is
recognized that in the 10-
foot spacing of the 10
trees in question there
has not been adequate
room for the miximum
development of each.
However, for all other
than the two end trees
the crowding in the row
has been alike and the
end trees did not yield
proportionate increases
because of their having
greater proportionate
room.
Tree production records
of all Station plantings
have shown that the
trend of seedling trees
toward consistently vari-
able yields has been simi-
lar to that of the above 10
trees. Even though it is
quite probable that the
same situation holds true
1Record of previous yields of those trees is available in Florida Experiment
Station Bulletin 221.
Bulletin 247, Variation in the Tung-Oil Tree
in most commercial plantings, the condition should not be con-
strued as meaning that those plantings will not bear profitably on
reaching maturity. Rather, the situation appears to be that pro.
duction in future plantings may be materially increased by a care-
ful and intelligent selection of seed for planting or by budding
seedling nursery stock with buds from mature trees of proven
merit. There is no longer a dearth of seed for planting so that
the further planting of seed from any or all sources, with the
consequent high degree of seedling variation, is not necessary.
Yield data here presented have possibly been influenced to some
extent by environmental factors, but the differences must be
largely of hereditary nature. There is little doubt that the acre
yield can be increased over that of miscellaneous seedlings by the
relatively simple expedients of seed selection or use of budded
stock.
Of the several variations apparent in the yield and in the char-
acteristics of the fruit, that of yield is of first importance. Heavy
yield in any tree is of far greater value than the presence of minor
desirable characteristics in a less prolific tree since, in most in-
stances, only a few extra pounds of fruit are required to offset
gains accruing from the minor varietal difference.
Yield, however, should not be the sole basis of future selection.
All additional characteristics that will make for a greater per-
centage of oil in the fruit are of value and when combined with
high yield will eventually assure highest returns.
BEHAVIOR OF BUDDED TREES AND SEEDLINGS OF KNOWN
PARENTAGE
Apparently the Chinese have made no effort to propagate the
tung-oil tree by budding or grafting. No information, therefore,
has been available on this phase of propagation other than the
knowledge that budding could be accomplished were it to be found
of practical worth.
Only a few budded trees were planted in field formation on the
test grounds of the Station at the time the first budding was
done there because of a lack of available planting space. The data
of Table III, however, definitely show that characteristics of the
tree from which the scions were taken were transferred to the
budded trees. It is nothing new in horticultural practice that in-
herent characters in woody plants amenable to budding or graft-
ing may be perpetuated by those means. The tests herein re-
ported merely have indicated that the tung-oil tree falls into the
TABLE III.-FRUIT ANALYSES OF PARENT TUNG-OIL TREE, ITS SEEDLINGS AND SCIONS.
Av. Av. Av. Fruit Fruit Seed
Fruit-Weight No. No. Mature-Not dried Air-dried Air-dried Kernel Seed Fruit Yield-Air-dried hulled seed
ing Indv. Seed Seed
Type Fruits Per Per Mean Mean Per- Per- Per- Per- Per- Per- Per- Per- Total
Tree Cluster (Air- Fruit Pound Breadth Length cent cent cent cent cent cent cent centotal
or dry) (cm.) (cm.) Seed Hull Ker- Testa Mois- Oil Oil Oil 1927 1928 1929 1931 4-year
Single (oz.) nel ture '(Oven- (Air- (Air- -----
Sdry) dry) dry) lb. oz. lb. oz. lb. oz. lb. oz. b. oz.
2 Cl .74 4.44 143 5.45 + .014 5.01 .012 63.0 37.0 59.9 40.1 8.17 58.50 32.18 20.27 55 0 25 8 89 12 72 8 242 12
S9
Sdlg.2 Cl .79 4.67 145 5.47 + .013 4.63 + .010 63.5 36.5 60.1 39.9 5.35 60.77 4.44 21.87 0 4 4 6 7 1 40 4 51 15
S10
Sdlg.2 Cl ... .... ....... .... 0 3 2 2 6 12 2 17 10
Bll
Bud2 Cl .74 4.58 168 5.32 + .011 4.75 + .008 61.5 38.5 57.5 42.5 6.15 58.51 32.11 19.75 0 5 2 5 4 6 28 4 35 4
B12 i
Bud2 Cl .77 4.63 159 5.29 + .012 4.70 + .010 60.5 39.6 57.4 42.6 5.98 59.712.28 19.53 0 8 2 13 5 6 29 2 37 13
d2 Cl .96 4.64 128 5.73 .027 604 396 645 35.5 500 6631 40.62 24.53
B CI .96 4.64 128 5.73 + .027 5.29 + .024 60.4 39.6 64.5 35.5 500 6631
K1 Too irregular for
Sdlg. S ... 2.4 126 systematic measurement. 44.4 55.6 61.9 38.1 5.43 '62.44 i36.55 16.23 .. ..
Tree 2 is the parent; S9 and S10, its seedlings; B11 and B12, buds frcm Tree 2.
K2-a bud from Tree 2 inserted in the older tree K, the latter a low yielder of mis happen fruits.
*Analysis on 50 fruits, 1931.
Trees S9, S10, B11, and B12 planted 1925; K1, about 1917.
Bulletin 247, Variation in the Tung-Oil Tree
same classification as other fruit or nut trees with reference to
varietal continuity by practice of asexual propagation methods.
Tree 2 (Table I) of the older Station trees was selected as a
source of budwood and seed because of its relative high yields
at the time and its cluster habit of fruiting.
Fig. 11.-A seedling of Tree 2, Table II./ Of same age as budded tree in
Fig. A,
In conjunction with the planting of budded trees, two seedlings
of the same parent tree from which buds were taken were included
in the same row (Figs. 14 and 15). There is no variation of ap-
preciable magnitude between the fruit of these seedlings and that
of the parent. The seedling trees are of the same age as the
budded ones but are approximately one-third larger in size. Al-
though the number of seedlings in the test is not great enough
to justify final conclusions, the result is in line with observations
made in commercial plantings. It appears that some seedling
trees will in fairly large measure and percentage bear fruit closely
resembling that of the parent. Plantings have been made on a
much larger scale that will give more conclusive evidence in the
future. In the meantime, advantage could well be taken of the
knowledge in hand and only seed from carefully selected parent
trees planted.
An analysis of fruits is given in Table III of the parent Tree 2,
Florida Agricultural Experiment Station
two of its seedlings-S9 and S10-and three budded trees- B1l,
B12, and K2, the budwood from Tree 2. The similarity in external
appearance of typical fruits from the parent, seedling, and buds
is shown in Figure 16. Variations among these trees in fruit size,
percentage of hull, seed, kernel, and oil are relatively small and
in contrast to the variation in mixed seedlings from various
sources as shown in Table I. The 1931 yield, as well as the total
yield (Table III), of the two seedlings (S9 and S10) and of the
two budded trees (B11 and B12) shows a wide difference in indi-
vidual yield of the seedlings and a striking similarity in produc-
tion of the budded trees. Further reference to Table III shows
that high production for one seedling and low yield for the other
have been characteristic of the trees for the full time they have
been in production. Yields of the budded trees B11 and B12 differ
materially from those of the seedlings in that they have been al-
most identical in amount each year. Neither of the budded trees
bore the amount of fruit as did the heavier fruiting seedling, but
the uniformity of the buds gave them a greater total yield than
that of the seedlings.
Fig. 15.-A budded tree, the bud from Tree 2, Table III. Of same age as
seedling in Fig.. Note the heavy crop of fruit.
Bulletin 247, Variation in the Tung-Oil Tree
A summary of results from the seedling and budded trees of
the same parentage shows the fruits of both to be nearly identical
in all characteristics; the seedling trees being about one-third
greater in size than the buds of the same age; the buds yielding
in nearly identical amount and in total exceeding that of the
*D
Fig. 16.-Upper, fruit of type tree (Tree 2) of Florida variety; center,
fruits from two seedlings (Trees S9 and S10) of Tree 2; lower, fruits from
buds (Trees 11B and 12B) from Tree 2.
seedlings; and one seedling yielding heavily and the other being
a low producer.
Tree K1 (Table III) is of mature age, a low yielder, and bears
its fruits singly. Its fruit is of very irregular size and shape. A
bud (K2) of Tree 2 (Table I) inserted in one of the larger limbs
of this tree resulted in a single, large, upright branch bearing
typical Tree 2 fruits in clusters. The variation in the shape and
size of fruits from the original tree K1 and of those from the
budded branch (K2) are shown in Figures 17. The vigor of the
tree K1 and its low yield may account for the somewhat larger
size and higher oil content of the fruits of the budded branch
over those of other buds of the same variety recorded in Table III.
The test of two types or varieties growing on the same root-
stock and fruiting at the same time has demonstrated that the
widely differing fruit characteristics of different seedlings are
inherent and are due only in small part to environmental in-
Florida Agricultural Experiment Station
fluences. It is significant that the yield of tree K1 has been al-
most negligible while that of the budded branch (the scion from
a prolific tree) has been heavy and that, based on analytical data
from representative fruits, the yield of oil from 100 pounds of
Fig. 17.-Upper, misshapen fruits of seedling Tree Kl; lower, fruits of the
Florida variety from budded branch (K2) of Tree K1.
air-dried fruits from the original seedling K1 would be 16.23
pounds, while a like amount of fruit from the branch of K2 budded
in its top would yield 24.53 pounds.
ANALYSES OF FRUITS FROM RAIFORD STATE FARM
In Table IV are data on fruits of the 1931 crop from nine trees
of a seven-year-old planting of the State Farm at Raiford. The
trees were grown as nursery stock at the Experiment Station at
Gainesville and are of the same mixed seedling parentage as the
early Station plantings. The soils on which the trees are planted
are deep, medium to fine sands of the Norfolk series with an ex-
tremely low content of organic matter.
According to the Farm horticulturist2, the trees were planted
on soils of low native fertility and have received but scant cultural
attention. He states, "The soil on which the trees are planted is
extremely poor white sand, very well drained, and the trees have
never been properly cultivated. The total amount of fertilizer
given them has, to date, amounted to about two pounds per tree.
2Frink, Aubrey J. Correspondence with the author, January, 1932.
TABLE IV.-ANALYSES OF TUNG-OIL FRUITS GROWN AT THE RAIFORD STATE FARM.
Av. Weight
Indv.
Tree Fruits (air-
dry basis)
(oz.)
F-1 .93
F-2 .69
F-3 1.33
F-4 .59
F-5 1.11
F-6 .56
F-8 .91
F-9 .67
F-10 1.44
Av.
No.
Seeds
per
Fruit
4.33
4.21
5.4
4.4
4.48
4.44
4.46
4.64
6.21
Av.
No.
Seeds
per
lb.
149
167
129
222
113
253
153
214
168
Fruit (Air-dry)
Percent Percent
Seed Hull
50.30 49.70
58.66 41.34
50.40 49.60
53.38 44.62
56.81 43.19
50.48 49.52
51.11 48.89
51.36 48.64
40.91 59.09
Seed (Air-dry) Kernel Seed
Percent Percent Percent Percent Percent
Kernel Testa Mois- Oil (oven- Oil (air-
ture dry basis) dry basis)
54.89
53.73
54.58
46.01
60.09
40.38
52.77
47.82
43.93
45.11
46.27
45.42
53.99
39.91
59,62
47.23
52.18
56.07
4.23
4.50
4 31
9 87
3.95
9.34
4.69
13.82
11.02
59 19 31.11
60 18 30 88
57. 67 30.12
52.53 21.77
61.70 35 60
51 21 18.75
58.46 29.40
50.92 20.98
39.58 14.30
Fruit
Percent
Oil (air-
dry basis)
15.65
18.11
15.18
11.62
20.22
9.47
15.03
10.76
5.85
1931
Yield lbs.
Air-dry
Whole
Fruits 4,
28
39
25 J
28
31
24
21
39 &
19
_____ ~
Based on samples of 25 fruits from each tree.
Florida Agricultural Experiment Station
Our tung trees apparently have not suffered at all from the
drought of the past fall and winter."
In appearance and size the fruits of this planting differed little
or none from those of any other mixed seedling plantings. The
effect of thin soils coupled with a lack of cultural attention, in-
cluding fertilization and cover-cropping, was clearly evidenced in
several of the samples examined by the lowered proportion of
^ if1
71,fc i. t. S j Y 1/ 8 t 9 A 3 c 0 e_ c 6- H
-re s 2. 3 y s 6 I 7o
Fig. 18.-Percentage of oil in air-dry fruit; left, Station planting; right,
Raiford planting.
seed weight to hull weight, light kernel weight, soft, flabby ker-
nels in some, and, with many, unfilled seed. The oil percentage
of the kernels was lower, on the whole, than of those from fer-
tilized Station trees. The chief and most significant difference
lay in the low percentage weight of oil in the whole seed and in
the whole fruit.
It is not possible to project accurate averages because of the
variation in moisture content of the fruits. However, a compar-
ison of the oil content of the fruits (Tables II and IV) shows that
a ton of mixed, air-dried fruits of the former contained at least
100 pounds more oil than a like weight of fruits of the latter (Fig.
18). Differences of such magnitude are worthy of serious con-
sideration in determining a policy of cultural practices (cultiva-
Bulletin 247, Variation in the Tung-Oil Tree
tion, fertilization, and cover-cropping) to be followed, if greatest
returns are to be had.
Varietal differences, other than number or amount of fruits
produced, for the most part lose their significance in undernour-
ished trees. The vigor of growth in seedling trees and soil fer-
tility are major factors to be considered in determining varietal
characters based on fruit composition and in the choosing of
parent trees for either seed or vegetative propagation.
VARIETIES
McClure" states that in China varietal names are given to tung-
oil trees, but the names applied differ in application in the differ-
ent provinces. Those varieties are: the Ten-Thousand Year Tung,
the trees of which are said to come into bearing at a normal age
of three years, to be long-lived, and to bear fruits with large seeds;
the Early Year Tung-trees are precocious, i. e., bearing at an
early age; and the Five Clawed Tung-applied to trees having
lobed leaves. The last character probably is not permanent but
found only on young trees. He advises that the use of varietal
names is not general and no apparently sound basis could be found
for their application.
Inasmuch as the above varietal differences are evidently of
little significance and have not been apparent in American plant-
ings, there is no reason for any attempt to follow the Oriental
classification of variation.
The differences appearing in Florida plantings that have led
to a naming of varieties have been: the large sized and often mis-
shapen fruits of the Craig; the willowy growth and peculiarly dis-
torted foliage of the Moore; and the production of fruits in clus-
ters of the so-called "cluster" type. The characteristics of the
Craig and Moore are given below, together with those of prolific
trees of a cluster bearing habit under the new varietal name-the
Florida. The varietal name Florida seems peculiarly fitting in
that the cluster bearing habit in tung-oil trees apparently has
not been reported outside of Florida plantings and was first
brought to attention in plantings of the Florida Experiment Sta-
tion.
The legitimate use of definite varietal names will do much in
stabilizing types of trees for future plantings. It is probable that
other varieties will be brought forth in future as and when differ-
3McClure, F. A.: Tung-Oil in the Yangtze Valley. Lingnan Sci. Jour.,
Vol. 9, No. 3., Oct., 1930.
Florida Agricultural Experiment Station
ences in trees or their fruit show them to be superior to ordinary
seedlings or existing varieties. Indiscriminate naming of new
varieties is to be discouraged and no new variety should be brought
forth until its worth has been definitely proved.
Fig. 19.-An abnormal and mal-formed fluit of the Craig variety re-
sulting from the fasciation of the fruiting twig. Such monstrosities are
not uncommon in this variety.
CRAIG
The Craig variety derived its name from that of the Florida
grower in whose planting a tree with the typical malformed
fruits was first observed. Its chief distinction lies in the large
size of its fruits, many of which, as a result of fasciation in the
twigs, are very large and distorted or huge, flatly oval specimens
(Figs. 19 and 20). In a single large misshapen fruit there may
be up to 22 seeds, many of which are quite flat rather than plump-
ly rounded. Tree H of Tables I and II was selected as the type
tree of the variety.
Technical Description:-The tree in shape, size, and foliage has
no characteristics definitely distinguishing it from vigorous seed-
lings. Fasciation of twigs is of common occurrence. Fruits dis-
tinctive due to large size and abnormal shape of many; shape
Bulletin 247, Variation in the Tung-Oil Tree
normally depressed globose or oblate; transverse diameter 5.7
to 9.1 cm. with a mean of 8.03 .026 cm.; longitudinal diameter
from 5.0 to 7.0 cm. with a mean of 6.07.017 cm.; average weight
air-dry fruits, 2.08 ounces; seeds, 7 to 8 in each fruit, averaging
7.28 per fruit and 105 (air-dry) per pound; whole fruits, 53.5 per-
cent seed and 46.5 percent hull; seed (air-dry), 61.1 percent ker-
percent oil.
The yield of this variety in the tree observed has not been above
average and it at present shows no superiority over ordinary seed-
lings.
MOORE
The Moore variety was given its name by Rolfe Buckley, who
first noted the variation in plantings of the Alachua Tung-Oil
Corporation near Gainesville.
The foliage of the variety differs distinctly from that of other
seedlings. Instead of having the usual acute or acuminate apex,
the leaf is distorted as if the midrib had failed to grow to its full
length. The leaf curls inwardly and appears in many specimens
Florida Agricultural Experiment Station
to consist of two lobes. The leaf is greatly reduced in size and
there is generally but one gland on the petiole near the leaf base,
instead of two. Figure 21 shows the characteristically distorted
foliage.
Fig. 21.-Typical fruit and foliage of the Moore variety.
The fruits are much smaller in size than the average and have
an exceptionally thin, almost paper-like hull. The tree is willowy
in growth habit, well branched, and very prolific but lacks vigor.
If a more vigorous strain were to be developed, it would be valuable
because of its fruitfulness and thinness of hull but, in its present
status, it apparently is of little commercial worth.
According to Buckley, only a very small percentage of its seed-
lings show the typical characteristics of the parent, the others
reverting to the ordinary type.
FLORIDA
Both Station and commercial plantings have demonstrated the
fruitfulness of trees of the cluster bearing habit, tree records
showing them to have produced the greater average yield. As an
instance of this heavier bearing of cluster trees, in a block of
90 nine-year-old trees in which 49 bore their fruit in clusters
and 41 singly, the average of the 1931 yield of the "clusters" was
Bulletin 247, Variation in the Tung-Oil Tree
slightly over 27 pounds of air-dried hulled seed, while that of the
"singles" was but slightly over 181/2 pounds.
The variety name Florida has been given to the parent tree of
which the above cluster bearing trees are seedlings. Tree 2 of
Tables I, II, and III is the type tree of the variety. The behavior
of its seedlings and scions is given in Table III and in the para-
graph on budded trees.
i l^ r. p i j'i'i' i i !)
Fig. 22.-Typical fruits of the Florida variety.
No measurable differences were apparent between the foliage
and individual flowers of trees of this variety and those of miscel-
laneous seedling trees. At the time of blossoming the cyme has
not only the apical pistillate flower but, generally, also one or
more pistillate blossoms borne laterally. This type of flowering
leads to the production of several fruits at the twig terminals-
now commonly referred to as "cluster" fruiting. The fruits, as
shown in Figures 1, 16, 17, and 22, are somewhat oblate but nearly
spherical in outline.
Florida Agricultural Experiment Station
Technical Description:-Tree spreading, prolific; bearing its
fruit in clusters. Flower cymes usually with one or more lateral
pistillate blossoms in addition to an apical one. Fruit oblate-
spherical, somewhat flattened at apical end; base with short neck
and apex with a rather sharply pointed nipple; hull thin; surface
smooth, distinctly marked with deep green longitudinal lines co-
incident with the inner segment or cell walls containing the indi-
vidual seeds; color usually green at maturity but a deep red where
exposed to full sun; size medium, the transverse diameter from
4.6 cm. to 6.1 cm. with a mean of 5.45 .012 cm. (standard devia-
tion .304, and coefficient of variability 5.57 percent), the longi-
tudinal diameter, including neck, from 4.3 to 5.6 cm. with a mean
of 5.01 .012 cm. (standard deviation .255, and coefficient of
variability 5.08 percent) ; average weight of air-dried fruits .74
ounces; seeds 4 or 5 to each fruit, average 4.44 per fruit and 143
(air-dry) per pound; whole fruits (air-dry) 63.0 percent seed
and 37.0 percent hull; seed (air-dry) 59.9 percent kernel and 40.1
percent testa; kernel (oven-dry) 58.50 percent oil; seed (air-dry)
32.18 percent oil; and whole fruit (air-dry) 20.27 percent oil.
The Florida variety is principally characterized by its cluster
fruiting habit and prolific yields. Not all trees bearing fruit in
clusters can be said to be of this variety but, rather, seedling
strains of the variety. Most seedling trees, although not all, bear-
Fig. 23.-A row of budded tung-oil trees. Seed planted February, 1929;
transplanted to field, January, 1930; budded in field, April, 1930; photo-
graphed, October, 1931.
Bulletin 247, Variation in the Tm,,-Oil Tree 31
ing their fruit in clusters produce fruits that are quite similar
in all respects, the greatest variation seeming to be one of yield.
With such a variation, however, seedlings of that variety, as has
been shown, average larger yields than do mixed seedlings bearing
their fruit singly. With a single exception, no recorded tree yield
of single fruiting trees has exceeded that of cluster trees. That
exception is Tree 9 of Table I, and its seedlings as well as scions
are undergoing tests to determine its comparative value with the
Florida variety in larger field plantings (Fig. 23).
There is no reason, without data from much further study and
breeding to prove the conjecture, to suspect the tree chosen as
the type of the Florida variety to be homozygous in genetic com-
position. Because of this and the probability of cross-pollination
there is little likelihood that the type tree is a pure-line strain and
that it, or other cluster fruiting trees, will in all instances produce
seedlings identical in characteristics. Some variation, to a degree
as yet undetermined, is to be expected in seedling trees from
whatever source. Variation due to such causes may be eliminated
by asexual methods of propagation, of which budding appears to
be the most feasible.
SUMMARY
Studies were made of the variations in the fruits and yields of
tung-oil trees.
A comparison of variation of the average of the individual trees
analyzed showed a range in:
Total yield (air-dry seed) 1931 crop of:
(a) 18-year-old trees: 7-', lbs. to 1643/4 lbs. per tree.
(b) 9-year-old trees: 7%3 lbs. to 581/g lbs. per tree.
Total four-year yield (1927, 1928, 1929, 1931) air-dry seed of:
(a) 18-year-old trees: 211/ lbs. to 48114 lbs. per tree.
(b) 9-year-old trees: 29 lbs. to 104/2 lbs. per tree.
Fruit size (mature but not dried fruit) of:
3.48 .0096 cm. to 8.03 .0262 cm. in transverse diameter.
3.761 .0094 cm. to 6.07 .0168 cm. in longitudinal diameter.
Individual fruit weights (air-dried) of .44 to 2.08 ounces.
Number of seeds per fruit of 4.11 to 7.28.
Number of seeds (air-dried) per pound of 85 to 227.
Percentage weight of seed in air-dried fruit of 53.5 to 66.4
percent and hull proportionately 46.5 to 33.6 percent.
Kernel weight of air-dried seed of 54.5 to 66.7 percent.
Oil content of kernels (moisture-free basis) of 53.97 to 66.31
percent.
32 Florida Agricultural Experiment Station
Combining the variable factors in different trees, it was found
that the oil content per hundred pounds of their air-dried fruits
varied from 17.04 lbs. to 24.53 lbs.
Undernourishment in the tree adversely affects the filling of the
seed kernel with a consequent reduction of oil percentage in the
whole fruits or seed.
Desirable characteristics can be perpetuated by means of asex-
ual methods of propagation.
On seedlings of known parentage, fruit characters have been
closely identical with those of the parent but a wide yield varia-
tion was apparent in the trees under observation.
Three varieties, the Craig, Moore, and Florida, are described.
|