CALIFORNIA AGRICULTURAL EXPERIMENT STATION
BULLETIN 722 MARCH 1951
EFFECTS of FERTILIZERS
UPON THE YIELDS, SIZE AND QUALITY
of ORANGE FRUITS
E.R.PARKER • W.W.JONES
THE COLLEGE OF AGRICULTURE • UNIVERSITY OF CALIFORNIA • BERKELEY
Fertilization
is one of the largest items of expense in the production of
citrus crops. It is also one of the most complex problems confronting the
grower. The factors involved are so interdependent, and effects vary so
widely under different conditions, that practical recommendations fre-
quently can be made only after years of carefully controlled experimen-
tation in the field.
Since 1907 the University of California Citrus Experiment Station at
Riverside has been conducting such experiments, attempting to determine
the effects of different fertilizer treatments on orange crops, trees, and
soil.
THIS BULLETIN reports the effects to date of 44 different fertilizer treatments
on the yields, size, and quality of the fruit of navel oranges. Results over
a 22-year period are reported. Resulting changes in the soil and trees
are summarized to afford a better understanding of the fertilizer effects
and thus permit broader application of the results. The studies are being
continued.
Conclusions based on results obtained thus far indicate that:
UNDER THE CONDITIONS PRESENT IN THE EXPERIMENTAL ORCHARD
1. Nitrogen fertilizers and bulky organic sources were the only
materials applied that increased yields. The law of diminishing
returns was found to apply to their use (increased applications
did not bring proportional returns).
2. Annual application of 3 pounds of nitrogen per tree (half from
a concentrated fertilizer, half from manure), with winter cover-
crops, resulted in maximum yields and fruit sizes.
3. Some chemical sources of nitrogen were found to be detrimental
if continuously used as the only fertilizer.
4. Application of bulky organic materials such as manure and cer-
tain agricultural minerals, as well as winter covercrops, tended
to counteract these detrimental effects.
5. Application of potash or of bulky organic materials caused an
increase in fruit sizes.
6. None of the treatments affected the packing-house grade of the
fruit.
A more detailed summary of the results will be found on page 52.
THE AUTHORS: E. R. Parker is Horticulturist in the Experiment Station. W. W. Jones is
Associate Horticulturist in the Experiment Station.
THIS BULLETIN is paper No. 657, University of California Citrus Experiment Station. It was
received for publication May 22, 1950.
Effects of Fertilizers
upon the yields, size and quality of Orange Fruits
E. R. PARKER W. W. JONES
CONTENTS
Page
I. Purpose of the experiments 4
II. How the experiments have been conducted 5
III. Detailed reports on experiments to date 9
A. USE OF CONCENTRATED SOURCES OF NITROGEN 9
Effects on yields, trees, and soil 11
Nitrogen from chemical sources 11
Nitrogen from organic sources 17
Effects on fruit size and grade 18
B. USE OF ORGANIC MATTER 19
Effects of covercrops and of manure 19
On relative yields 20
On fruit size and grade 24
C. USE OF PHOSPHATE AND POTASH 27
Effects when applied with concentrated nitrogen fertilizer 27
On relative yields 28
On fruit size and grade 33
Reliability of observed effects of potash on fruit size 36
D. USE OF BULKY ORGANIC MATERIALS OTHER THAN MANURE 38
Effects on relative yields 39
Effects on fruit size and grade 41
E. USE OF PHOSPHATE, POTASH, AND OF AGRICULTURAL MINERALS WITH
MANURE AND UREA 42
Effects on relative yields, size, and grade of fruit 42
Effects of agricultural minerals on soil reaction and related factors 43
F. USE OF VARYING AMOUNTS OF NITROGEN 47
Effects on yields 47
Effects on fruit size and grade 49
IV. Conclusions from results to date 52
V. Literature cited 55
VI. Acknowledgments 58
I. Purpose of the experiments
THE field experiments which have
been conducted by the University of
California College of Agriculture to de-
termine the effects of fertilizers upon
orange crops may be divided into two
groups:
1. The relatively short trials of 3 to 12
years. Such trials have usually been de-
signed to determine the deficiencies of
available nutrients in the soil and the ef-
fectiveness of various fertilizer practices
in correcting them, as well as the effects
of agricultural minerals and of cultural
practices over short periods of time. They
are, therefore, of considerable practical
value. The yields which have resulted in
11 short-term experiments have been re-
ported. These trials were carried out on
properties of cooperating orchardists
under a wide range of soil types in loca-
tions scattered from Tulare County to
San Diego County. In addition, many
short-term "spot trials" were made in
widely scattered locations to determine
qualitative responses which might be
visible to the eye.
2. The long-term field experiments
dealing with the crop responses of orange
trees. These have the advantages of grad-
ually intensifying the effects of added ma-
terials and of permitting observations
over a longer number of seasons— thus
increasing the reliability of the results.
They also permit studies of (a) the effects
on the trees of trends in the fertility of
the soil under citrus culture and (b) the
cumulative, secondary effects of fertiliza-
tion upon the soil and trees.
The Citrus Experiment Station has
conducted two complex, long-term groups
of fertilizer experiments on its properties
at Riverside. The first of these was started
in 1907 and ended about 1930. The other
experiments, results of which are re-
ported here, have been conducted on trees
planted for the purpose in 1917. The dif-
ferential fertilization of the plots began
in 1927 and is continuing.
Results of the experiments which have
been referred to were published by Vaile
(1922), x Surr and Batchelor (1926) and
by Parker and Batchelor (1942). Other
information on the effects of fertilizers
upon yields is contained in the report of
an extensive survey by Vaile (1924) and
in a more limited survey dealing with
the effects of cultural practices upon
yields and quality of fruit reported by
Parker, Rounds, and Cree (1943) . Addi-
tional information upon fertilization ef-
fects has resulted from experiments and
surveys by the California Agricultural
Extension Service. A great deal of knowl-
edge regarding the nutrition of citrus
trees and the nature of California citrus
soils has also been obtained in labora-
tory and greenhouse investigations. These
studies have provided a better under-
standing of fertility problems and a
broader application of the results of the
field fertilizer experiments.
The reports referred to have not been
concerned with studies of the effects of
application of the "trace" elements— zinc,
manganese, copper, boron, and iron.
Other papers have drawn attention to the
symptoms which are produced when
See "Literature cited," page 55, for citations, referred to in the text by author and date.
[4]
these elements are deficient, and to the
results of treatment.
The purpose of the present paper is to
give a second progress report on the long-
term fertilizer experiments with Wash-
ington Navel oranges being conducted at
Riverside. Results over a period of 22
years (1928-49) are presented. The
effects of the various fertilizers upon
yields during the first 12 years of the
experiment (1928-39) were reported by
Parker and Batchelor (1942) . This paper
gives the yields resulting in the subse-
quent 10-year period (1940-49) and in-
cludes a summary of many of the earlier
yields for reference to permit the ob-
servation of trends over the entire 22-year
period. Information on the effects of the
fertilizers upon the size and commercial
grade of the fruit is given. Observations
are also made as to their effects upon tree
conditions and upon the soil.
II. How the experiments have been conducted
Washington navel orange trees on
sweet orange roots were planted for
this experiment in 1917, in rows 24 feet
apart with trees 20 feet apart in the rows,
on soil classified as Ramona loam. The
site had previously been dry-farmed to
unfertilized grain crops. The land was
fertilized for the first time in 1927, when
the trees were 10 years old. Until then
winter covercrops of yellow bitter clover
(Melilotus indica) or of purple vetch
(Vicia atropurpurea) were grown an-
nually throughout the young orchard.
During the first 6 years of its life, summer
covercrops of black-eye beans or cowpeas
{Vigna sinensis) were also unifoimly
grown.
A record of the cultural practices dur-
ing the preliminary period as well as the
history and outline of the experiment
were published by Batchelor, Parker, and
McBride (1928), and the method of the
distribution of the plots to the various
* treatments was described by Parker and
Batchelor (1932).
Each unit test plot consists of a single
row of 8 Washington Navel orange trees.
The plots are separated by single guard
rows of Valencia orange trees and Marsh
grapefruit trees which alternate in the
row. Four test plots, distributed over the
entire acreage of the experimental area,
are used for each fertilizer treatment.
In addition, one treatment, originally
called the "continuity treatment," is ap-
plied to 25 unit plots in the orchard. The
fertilizer which is applied to the plots of
this treatment is representative of good
commercial practice, and the results are
used for comparative purposes. The con-
tinuity treatment is also used for control
purposes. Its yields, together with the plot
yields which were obtained prior to the
start of fertilization, provide a basis for
adjusting the yields of plots of other treat-
ments for local variations, thereby im-
proving the precision of the results
(Parker, 1942).
No unusual care has been exercised in
the cultural operations performed in the
experimental orchard. Cultivation of the
soil has been similar to that generally
practiced locally. The frequency and
depth of cultivation has been reduced
gradually over a term of years in accord-
ance with prevailing trends. Where the
fertilizers have affected the penetration
of irrigation water into the soil, efforts
[5]
have been made at critical times to pro-
vide adequate soil moisture. During the
most recent 8-year period the interval be-
tween summer irrigations has been re-
duced from the customary 4 weeks to
3 weeks because of increasingly poor
water infiltration in the soil of some of
the treatments.
Fumigation with hydrogen cyanide has
been practiced for the control of scale
insects, and occasional treatments have
been applied for the control of the citrus
red mite. A moderate number of the trees
have developed symptoms of psorosis.
Where the symptoms of this virus disease
have been so severe as to affect the crop
of the trees, the trees have been removed
or their records deleted. A few trees have
been lost as a result of other causes, but
the number of abnormal trees has not
been sufficient to necessitate the elimi-
nation of any single plot.
Although no fertilizers were applied to
the orchard trees during the first 10 years
after planting, the rate of nitrogen appli-
cations used in many of the treatments
during the next 12 years (1927-39) was
smaller than that generally used in com-
mercial orchards. This was considered
especially desirable in facilitating com-
parisons of the relative effectiveness of
different sources of nitrogen. Only one
pound of actual nitrogen was applied per
tree annually in most of the treatments.
It became evident, however, that the
low supply of nitrogen was limiting the
yields of many treatments, and in the win-
ter of 1939-40 the basic rate of applica-
tion of nitrogen was increased to 3
pounds per tree annually. It was hoped
that this change would help to insure the
satisfactory determination of responses to
other elements and also to hasten the ap-
pearance of secondary or indirect effects
of various fertilizers. Since the change in
rate was made the yields have improved
considerably. A small number of other
treatments were also changed in other re-
pects in the season of 1939-40, or subse-
quently, when it was considered that they
had served their purpose and where the
change was expected to make the experi-
ment more valuable.
In most of the treatments, the concen-
trated and chemical fertilizers and agri-
cultural minerals have been applied in
one application in the spring of each year
about 6 weeks before blooming dates.
The bulky organic materials have gener-
ally been applied in the fall prior to the
planting of winter covercrops. In the
treatments for which results are reported
in this paper the fertilizer materials were
applied broadcast as uniformly as pos-
sible in the irrigated areas of the plots.
Symptoms of zinc deficiency (mottle-
leaf) became apparent in the orchard
soon after the fertilizer treatments were
started. They were most severe on plots
which received no organic fertilizers and
on which no covercrops were grown.
Since the fertilizer experiment was not
designed to measure the effects of zinc
deficiency or to test responses to zinc
treatment, all the plots have been uni-
formly treated with zinc sprays since
1934. It was shown in other experiments
that zinc deficiency markedly reduced the
yields of citrus trees and resulted in
smaller fruit of poorer quality (Parker,
1937a and b) .
After the symptoms of mottle-leaf were
eliminated by zinc applications, very
mild and transient symptoms of manga-
nese deficiency have been observed. The
guard rows of the fertilizer experiment
have been used to determine the impor-
tance of these symptoms. Unpublished
data indicate that correction of these
symptoms by the application of manga-
nese materials in foliage sprays has not
resulted in improved yield, size, or grade
of the fruit. The test trees in the fertilizer
plots have therefore not been treated with
manganese. No other extraneous nutri-
tional or toxicity symptoms have been
observed in the orchard, and no response
[6]
to copper sprays has occurred on trees
in other guard rows in the experimental
area.
The original yield data were obtained
annually for each normal tree on a
volume basis. The values obtained in this
way were then converted into pounds of
fruit per tree. The mean yields per tree
of each treatment were then calculated
giving equal weight to each replication.
The data which are presented here are
the mean annual yields in pounds per tree
of the treatments in 5 periods of 4 years
each during 1928-47 and in the 2-year
period, 1948-49. 2
In 14 of the 22 years during which the
effects of fertilizers have been studied, it
has been practical to grade separately in
the packing house the total fruit harvested
from the individual treatments. For this
operation, the fruit produced by all plots
of a treatment was combined to make a
packing-house "lot." The identity of the
fruit of the individual replicates was
therefore lost. The commercial grading
procedures in use at the time were em-
ployed, except that starting with the crop
harvested in 1939 the lower grades of
fruit, which were not usually passed
through the sizers, were specially sized
for these studies. The fruit which would
not pass through the sizing machinery
was usually measured by hand. 3 For this
reason, and also because they describe
the effects which occurred after a con-
siderable period of fertilizer use, the data
since 1939 are used more frequently in
this bulletin.
Data on fruit sizes are presented in
terms of the percentage of the total
volume of the fruit which was of size 220
or larger, i.e., 220 or fewer fruits per
packing box. Such fruits have diameters
of 2.56 or more inches. These fruits are
frequently the most desirable sizes, and
command large price premiums, espe-
cially in years when average sizes are
small. The grades of the fruit produced
by the different treatments are indicated
by the percentage of fruit— by volume—
which was packed in the fancy or choice
grades. 4
2 The data of mean annual yields during the years of each period which are used in this report,
except in the control plots (treatment C) and in plots in the two treatments which received no
fertilizer (treatments 1 and 6) , have been adjusted so as to reduce the effects of chance variations.
This adjustment has not been applied to the yields of individual years. Covariance on the yields
of the control plots has been employed in the adjustment, and variations which are correlated with
the yields of the plots before the start of the differential treatments have been reduced by the
elimination of "yield-block" effects based on yields of 1927 (Parker, 1942). On the average, the
least significant differences (at 5 per cent probability) between the mean annual yields per tree of
any two treatments (each based upon the yields of four plots) in the different 4-year periods are
as follows: for 1928-31, 17 pounds; for 1932-35, 27 pounds; for 1936-39, 27 pounds; for 1940-43,
37 pounds; for 1944-47, 38 pounds; and for 1948-49, 36 pounds. In numerous instances the conclu-
sions given in the paper are influenced by the fact that more than two treatments can be compared
in determining the effects of a given fertilizer. In some instances where materials have been applied
in different quantities the existence of trends in the results also leads to confidence in the con-
clusions.
3 The following grades of fruit were sized in the indicated years of harvest: in 1931, 1932, and
1933, fancy and choice grades; in 1934, fancy, choice, and standards; in 1941, 1945, and 1948,
fancy, choice, standards, and culls. In the other years— 1939, 1940, 1942, 1944, 1946, 1947, and
1949 — all of the grades, and also the packing-house rots, were sized.
Although estimates of effects of chance variations on the size and grade of the fruit cannot be
made by the procedures used in studying the yields, the consistency of the differences between
contrasting treatments and the trends in response, which occur when different quantities of a
material were used, provide a basis for judgment. It is believed that the design of the experiment
has helped to reduce errors in these observations. The significance of certain differences in fruit
size are discussed in a separate section.
4 The estimation of the percentage of fruit in the various grades was made on the basis of the
volume of all the fruit delivered to the packing-house in each year except 1941 and 1945, in which
[7]
Inspection of the packing house data
shows that both size and grade fluctuate
markedly from year to year. For example,
in treatment 42 (3 pounds of nitrogen
per tree annually) the per cent by volume
of fancy and choice fruit had a minimum
value of 69.3 per cent (in 1949) and a
maximum value of 92.4 per cent (in
1940). The per cent by volume of fruits
of size 220 and larger ranged from 35.3
per cent (in 1949) to 90 per cent (in
1941) . These annual fluctuations affected
the fruit of all treatments in about the
same manner.
There are 44 different fertilizer treat-
ments in the experimental orchard. In
this report the data for yield, fruit size,
and fruit quality of small groups of re-
lated or contrasting treatments will be
given.
Comparative yields are presented in
relation to a single base or control treat-
ment, for which the mean annual yields
are given in pounds per tree for the vari-
ous periods.
Data on fruit size and grade are pre-
sented as mean values for terms of several
years.
years the total volume did not include the rots which developed in the packing-house. These rots,
however, represented a very small proportion of the total volume of fruit.
Consideration has been given to alternative methods of reporting fruit size and quality data,
particularly the use of averages giving mean diameters of fruit or mean numbers of fruits per pack-
ing box equivalent. For the purposes of this paper these methods appear less desirable than the
method adopted, namely, the percentage of the larger sizes on a volume basis. The latter method
tends to magnify the differences between treatments, but this has proved to be an advantage. It
is also commonly used by packing-houses, marketing organizations, and others in reports to grow-
ers, and is familiar to them.
These comments also apply to reports of fruit grade data based upon the percentage of the total
volume which is of the various grades, regardless of size. However, a slight bias, for which there
is no measure, is believed to exist in grade statements for years when either large or small fruit is in
demand or in excess supply. Under such conditions these fruits are sold at prices which differ
greatly from the average, and the fruit of such grades is graded more heavily or more leniently
as the case demands. On balance, this probably means that small fruit has been graded more
heavily than large fruit. In the present experiment it seems to explain the observation that the
treatments which have produced the larger fruits also tended to produce larger proportions of
fancy and choice grade fruit, by volume.
8]
III. Detailed reports on the experiments
A. USE OF CONCENTRATED SOURCES OF NITROGEN
THE fertilizer experiments and the
surveys which have been referred to,
as well as the experience of growers, show
that the soils of California lack sufficient
nitrogen for maximum production of
citrus fruits. Yield responses to applica-
tions of nitrogenous fertilizers occur on
an extremely wide range of soil and cli-
matic conditions. The large tonnage of
nitrogenous fertilizers which is applied
annually to citrus orchards is evidence
that this deficiency is generally recog-
nized. Only occasionally, and then usually
for only a few years when planted on very
fertile soil, do newly planted citrus trees
fail to respond to fertilization with such
materials.
The need for liberal applications of
nitrogen to citrus orchards has been ex-
plained, at least partly, by the fact that
this element has a rather transient exist-
ence in the soil. It is much more subject
to leaching than most of the elements,
and a part is also lost to the atmosphere
in gaseous form (Hoagland, 1947; Chap-
man, Liebig, and Rayner, 1949). In the
southern part of California, where the
present experiment is being conducted,
the common practice is to apply sufficient
nitrogenous fertilizers to supply from 2 x /2
to 3 pounds of actual nitrogen per tree
annually. Growers consider larger quan-
tities to be profitable in particular in-
stances. In central California somewhat
smaller amounts seem to suffice.
The use of chemical sources of nitrogen
has increased tremendously in the last 30
years in comparison with the use of con-
centrated organic sources. This has been
the result of greater supply, due in part
to the development of synthetic processes
and lower cost per unit of nitrogen, and
to increased convenience in use. Studies
of the value of chemical fertilizers are
very desirable, both in relation to each
other and in relation to concentrated or-
ganic sources of nitrogen. Such studies
involve two distinct problems. The first
(which is usually the more important)
concerns the relative value of the mate-
rials as sources of nitrogen for the nutri-
tion of the crop plant. The second involves
what may be considered the secondary
effects of the fertilizers on the plants re-
sulting from chemical and physical
changes in the soil which may be caused
by the fertilizers.
Extensive experience with crops other
than citrus under humid conditions over
a period of many years has indicated that
the effects on the soil of certain chemical
sources of nitrogen are cumulative. It is
therefore important to understand the
changes in the soil, trees, and crops which
the continuous use of these fertilizers may
bring about under conditions existing in
irrigated citrus orchards. This informa-
tion is needed in order to recognize the
need for, and method of, preventative or
remedial measures.
Several treatments in the experiment
reported here are of special interest re-
garding the effects of using concentrated
sources of nitrogen. Eight treatments in-
volve the use since 1927 of 4 chemical
sources. These materials are nitrate of
soda, nitrate of lime, sulfate of ammonia,
[9]
1922
-27
1928
-31
1932
-35
1936
-39
1940
-43
1944 1948
-47 -49
FIG. 1. Responses to nitrogen fertilization. Mean annual yields of unfertilized orange trees (treat-
ment 6) and of trees fertilized with nitrate of lime (treatment 21) in 5 periods of four years and
one period of two years. The fertilized trees received 1 pound of nitrogen each year from 1927 to
1939, and 3 pounds of nitrogen per tree from 1940 to 1949. Winter covercrops were grown.
and urea. 5 Each of these materials is used
in separate treatments with winter cover-
crops and also under conditions of clean
cultivation in the winter and summer. An
additional treatment consists of a mixture
of three sources of nitrogen— sulfate of
ammonia, nitrate of soda, and dried
blood— each supplying one third of the
total amount of nitrogen. Dried blood and
cottonseed meal, both sources of organic
nitrogen, have also been used in separate
treatments.
When the experiment was first outlined
(Batchelor, Parker, and McBride, 1928) ,
it was anticipated that any harmful effects
of the continuous use of specific fertiliz-
ers would probably be due to the devel-
opment of excessive soil acidity or to the
accumulation of large amounts of sodium,
both situations resulting in loss of cal-
cium from the soil by base exchange
reactions. The possibilities of simultane-
ous accumulations of toxic concentra-
tions of sodium and of adverse changes
in the physical condition of the soil were
also considered. In efforts to prevent such
possible occurrences, one treatment in
the experiment was established in which
generous applications of gypsum were
made with the usual amount of nitrate
of soda. In another treatment limestone
has been used with sulfate of ammonia.
5 The urea has been used since the spring of 1929. In 1927 and 1928 ammonium nitrate, which
likewise deposits no residue in the soil, was applied to the same plots.
[10]
TABLE 1 — Relative Yields Resulting from the Use of Various Concentrated
Sources of Nitrogen When Used with Winter Covercrops
Treatment*
1928
-31
1932
-35
1936
-39
1928
-39
1940
-43
1944
-47
1948
-49
1940
-49
21. Nitrate of lime
Mean annual yield per tree (pounds)
123
160
118
134
203
185
183
192
21. Nitrate of lime
27. Nitrate of soda
Relative yields (treatment 21 = 100)
100
93
91
96
95
98
87
100
91
89
89
93
96
86
100
102
96
102
97
105
79
100
94
91
95
95
99
84
100
81
59
80
83
97
72
100
78
63
85
84
109
84
100
57 c
63 c
88
76
100
98
100
75
61
84
82
102
81
15. Sulfate of ammonia
18. Urea
12. Mixed sources of nitrogen b
28. Nitrate of soda and gypsum
16. Sulfate of ammonia and
limestone
17. Blood meal
93
87
94
92
102
95
96
91
86
92
90
110
86
102
87
101
43. Cottonseed meal
6. No fertilizer
67
43
17
43
13
12
26
15
a Concentrated nitrogenous materials were applied in the spring of each year at rates to supply one pound
of nitrogen per tree annually during 1927-39 and three pounds of nitrogen per tree annually during 1940-49.
When gypsum was used, it was applied at the rate of one ton per acre each year and limestone was applied
at the rate of 4 tons per acre every fourth year.
b Nitrate of soda, sulfate of ammonia and blood, each supplying one-third of the total amount of nitrogen
applied.
c Mean of only two replicates. Other values are means of four replicates.
In these studies of nitrogen sources the
amount of elemental nitrogen applied in
all treatments has been equal. During the
first 12 crop years (1927-1939) one
pound of nitrogen per tree was applied
annually. It was anticipated that this low
rate of application might permit a more
critical evaluation of the efficiency of the
various materials as sources of nitrogen.
Subsequently 3 pounds of nitrogen per
tree have been applied.
Effects on yields, trees and soil
All of the nitrogen carriers which have
been used in these trials have caused sub-
stantial increases in tree yields as com-
pared to the unfertilized control plots.
The yields of the unfertilized trees grad-
ually declined to a very low level during
the latter half of the experimental period.
This is illustrated in figure 1 and table 1.
NITROGEN FROM CHEMICAL SOURCES
Attention will be called first to the ef-
fects of nitrogen from chemical sources
upon the yields of the trees. Since the rela-
tive effects of these materials were com-
parable when they were used either with
or without winter covercrops, only the
yields obtained in the covercropped plots
are presented here. Table 1 gives relative
yields by periods of 4 years as well as
mean relative yields for the 12- and the
10-year periods, when the different rates
of nitrogen application were used. The
yields resulting from the use of nitrate of
lime with covercrop (tr. 21) are taken as
100 per cent.
[ii J
Different results were obtained in the
two periods. The relative yields for the
first 12 years of the experiment, during
which the applications were made at the
rate of one pound of nitrogen per tree
each year, showed (Parker and Batche-
lor, 1942) that the yield responses to the
various chemical sources of nitrogen were
very similar. The mean yields for this
period, 1928-39, resulting from the use
of nitrate of soda (tr. 27) , sulfate of am-
monia (tr. 15), and urea (tr. 18) were
within 7 per cent of those trees fertilized
with nitrate of lime (tr. 21 ) .
These differences are well within the
realm of experimental error. Apparently
any secondary effects on the soil were of
minor consequence at that time. More-
over, the use of the agricultural minerals
had no effect on yields. Thus gypsum
when applied with nitrate of soda (tr. 28)
or limestone with sulfate of ammonia (tr.
16) did not give significantly different
yields from those which were obtained
by the use of the respective nitrogen car-
riers alone (trs. 27 and 15) .
The increase in the rate of fertilizer ap-
plication which was made in the spring of
1940 in these plots offered an opportunity
to determine responses to the increased
supply of nitrogen. It also hastened the
development of harmful secondary effects
of certain of the chemical nitrogen fer-
tilizers. The secondary effects are de-
scribed in some detail since they are of
considerable magnitude and importance.
The effects of various amounts of nitro-
gen under more "normal" soil conditions
will be discussed elsewhere in this paper.
Differences in yields resulted within a
few years after the increase in rate of
application in 1940. These differences
cannot be attributed to the effectiveness
of the materials as sources of nitrogen.
In fact, the absorption of nitrogen by the
trees in all these nitrogen source treat-
ments was relatively high and nearly
equal as indicated by leaf analysis (2.37
to 2.47 per cent N on the dry weight basis
in February, 1947). Moreover, nitrogen
can be hardly a limiting factor in the
lowest-yielding treatments, since the
quantity of nitrogen in the soil of these
treatments has increased.
The differences in yields were due to
adverse soil conditions brought about by
certain fertilizers as a result of the accu-
mulation of residues or by cumulative
chemical effects. Nitrate of lime appar-
ently had the least effect of this nature.
The yields of the trees which received
this fertilizer have been maintained at a
relatively high level throughout the entire
experimental period. (See table 1.)
Relative to the yields of the nitrate of
lime treatment (No. 21) , markedly lower
average yields resulted in 1940-49 from
the use of nitrate of soda (tr. 27) and also
of sulfate of ammonia (tr. 15) , while urea
(tr. 18) and the mixed nitrogen sources
(tr. 12) gave somewhat lower yields than
nitrate of lime. The use of gypsum pre-
vented a depression of yields when used
with nitrate of soda (tr. 28). The use of
limestone with sulfate of ammonia (tr.
16) was also very beneficial. 6
Similar abnormal symptoms first de-
veloped in 1942 on the trees in plots
which received only nitrate of soda and
on those which received only sulfate of
ammonia. A very limited amount of new
growth developed that year and the older
leaves assumed a dull, grayish color. A
heavy drop of old leaves occurred in the
winter of 1942-43. Many of the old leaves
wilted and burned at the tip before drop-
ping. A large part of the young growth
which appeared in the spring of 1943 was
also affected with tip burn and dropped
(plate 1, page 29). As a result, the trees
were very thinly foliated in the early sum-
mer of that year, and the dullness of color
The effect of limestone was apparent on the structure of the soil, on tree condition, and upon
yields, although its effect on yields was not as great as was expected from the other observations.
The lower yield, just bordering on statistical significance, is perhaps due to chance. The yields
of this treatment (No. 16) have been somewhat low since fertilization began in 1927.
[12]
120-
> 60
<
J
Ul
QC
40
-, ^
*/ NITRATE OF Lime
ST. HITRATE OF SODA
v w.
/j-» •
/5. SULFATE OF
HMOMA
It. HIKED SOURCES
w
\\
1922
-27
1928
-39
FIG. 2. Effects of several concentrated nitrogen fertilizers upon relative yields of oranges during
the 22-year period 1928-49. The yield of treatment No. 21 (nitrate of lime) is taken as 100 per
cent in each year. No fertilizer was applied prior to 1927. From 1927 to 1939 each material sup-
plied one pound of nitrogen per tree each year. Starting in 1940, 3 pounds of nitrogen were ap-
plied annually. The values for annual yields have not been adjusted (see footnote 2, page 7).
of the remaining foliage increased. In the
treatments receiving urea or mixed nitro-
gen sources the abnormal leaf drop was
not observed but dull-colored foliage and
a reduction of amount of new growth
were noticeable. These symptoms were
not found on trees in plots receiving gyp-
sum and nitrate of soda or in plots treated
with limestone and sulfate of ammonia.
The reduction in tree yields accompa-
nied or slightly preceded the appearance
of foliage symptoms. Figure 2 shows tne
mean yields of these nitrogen source
treatments, relative to the yield of the
nitrate of lime treatment, by years during
1940-49, when the increased rate of ap-
plication was used, as well as the annual
average relative yields for earlier periods.
Note that the relative yields fell rapidly
after 1940 and reached their lowest values
in the crop harvested in the spring of
1943. The improvement in relative yields
since then is believed to be due to a reduc-
tion of the interval between irrigations
(from 4 to 3 weeks during the periods of
greatest water usage) which was begun in
the summer of 1943. It is evident from
figure 2, however, that this expedient has
not resulted in normal yields.
Plates 2 and 3 (pages 30 and 31) show,
also, that the growth and foliage of the
trees have not become normal. It seems
probable that further tree decline may be
expected as the soil conditions further
deteriorate as a result of additional appli-
cations.
The harmful effects of the continuous,
sole use of large quantities of certain
chemical fertilizers have been associated
with the development of poor soil struc-
ture, as indicated by a decreased per-
meability of the soil to water. Many
systematic observations of depth of pene-
tration of irrigation water have been
made in these plots. Even during the later
part of the 1928-39 period, it was appar-
ent that nitrate of soda and sulfate of am-
monia were causing decreased water
[13]
infiltration during irrigation, and that
this depression was prevented (1) when
gypsum was used with nitrate of soda and
(2) when limestone was used with am-
monium sulfate. Urea and the mixed
sources of nitrogen had a less marked ef-
fect on water movement.
These effects on infiltration rate were
intensified following the increase in the
rate of fertilizer application in 1940, and
soon thereafter it became evident that the
tree yields also were depressed in treat-
ments which most reduced the quantity of
irrigation water absorbed. This relation-
ship is clearly shown in figure 3 which il-
lustrates for these chemical treatments the
very high correlation between the mean
depth of penetration of irrigation water
during the summer and fall of the years
1942 and 1943, and mean yields har-
vested in the following spring (1943 and
1944). Similar correlations between
yields of other crops and other indices of
soil structure have been reported else-
where (Richards, Neal, and Brill, 1949).
Extensive studies of the more seriously
affected treatments indicated that it was
not possible to introduce sufficient water
into the root zone during irrigation to
carry the trees through the normal inter-
val between irrigations. Many expedients
in the management of water were tried,
but none was successful, hence the inter-
val between irrigations was shortened in
the summer of 1943. These observations,
and the improvement in the yields and
conditions of the trees in the most severely
affected plots following the change in the
irrigation interval indicate that moisture
stress was an important factor in causing
lower yields and poor tree condition.
Considerable quantities of soluble salts
have accumulated in the root zone of the
soil of certain treatments as a result of
decreased leaching losses. It is certain
that these aggravated the moisture stress,
and may have exerted some direct toxic
effect. It is of interest that the most se-
verely affected trees have not shown typi-
cal wilting during hot periods when
moisture stress has been general.
Laboratory studies of the surface soil
from these treatments by Aldrich, Parker
and Chapman (1945) have shown that
the various chemical fertilizers had
marked effects on soil structure and the
chemistry of the soil. In the cases of the
sulfate of ammonia and the nitrate of
soda treatments it was found that the de-
terioration of soil structure was the result
of the replacement of the exchangeable
calcium of the soil. Where nitrate of soda
was applied, sodium accumulated on the
base exchange complex of the soil while
calcium was lost. This resulted in a typi-
cal high-sodium soil having poor struc-
ture characterized by small pore spaces,
poor aeration, reduced permeability, and
an accumulation of salts. These harmful
effects of nitrate of soda were not ob-
served in the soil samples from plots
treated with gypsum (calcium sulfate)
and nitrate of soda, for in this case an
abundance of soluble calcium (in relation
to sodium) was available to maintain
good soil structure.
In the case of the soil of the sulfate of
ammonia treatment the situation was
found to be more complex. This material
is strongly acid-forming when applied to
the soil and in these plots has resulted in
very low pH values (see table 9, page 44) .
Acid soils do not necessarily have poor
structure. However, it was shown (Al-
drich et al., 1945) that in these plots the
acidity which has been developed in the
soil not only caused replacement of cal-
cium but also was so intense that nitrifi-
cation of the ammonium ion supplied by
ammonium sulfate was prevented, and
that the accumulation of this ion on the
base exchange complex was responsible
for the deterioration of the physical con-
dition of the soil. In these laboratory
studies the soil of the limestone-sulfate of
ammonia treatment exhibited good struc-
ture and was well supplied with calcium.
It was also normal in regard to the con-
[14]
a 120
O 100
o.
I
+
- 80
a
z
<
A
* 60
a
j
u
>
40 —
20
-
1 1
1
•
• tl.
\
NITRATE OF S00A,
6YPSVM, C C
NITRATE OF. LIME, C.C
-
—
-
• to.
NITRATE
OF
LIME
-
• ib urea, c c
• IS
SULFATE OF AMMONIA,
LIMESTONE, C C
-
-
• / URCA
• tT NITRATE OF SODA, C C
-
—
-
•
• IS SULFATE OF AMMONIA, C C.
% 16 NITRATE OF S00A
SULFATE OF AMMONIA
1 1
1
1
-
2.0 2.5
MEAN PENETRATION, 1942 AND 1943 — IN FEET
FIG. 3. Mean annual yields (unadjusted) harvested in the spring of 1943 and 1944 from trees
receiving various concentrated sources of nitrogen in relation to mean depth of penetration of
irrigation water during 5 irrigations in 1942 and 1943. The fertilizers had been used since 1927.
Water penetration was measured to a depth of only 3 feet. Treatments which were winter cover-
cropped are indicated by "C. C."
centration of ammonium ions. Similar
effects of sulfate of ammonia on nitrifica-
tion in acid citrus soils have been re-
ported from Australia (Parbery, 1945).
It has been shown that the use of gyp-
sum has prevented the harmful effects of
nitrate of soda in the soil or trees in these
trials when the two were applied together.
Likewise the use of limestone has pre-
vented injury from the excessive use of
sulfate of ammonia so far as the soil is
concerned and probably insofar as the
trees and their yields can be determined.
Limestone has been used for many years
in the treatment of excessively acid soils.
In the process of neutralizing the acid of
the soil, gypsum— a slowly soluble calcium
compound— is formed by reaction in the
soil. Dolomitic limestone, which contains
appreciable quantities of magnesium, is
sometimes preferred in Florida, especially
if magnesium is deficient as a plant nu-
trient in the soil. In the treatment of alkali
soil which contains an excess of sodium
the needed calcium must be supplied in a
soluble form. Gypsum is the standard
method of supplying calcium under the
latter conditions.
That gypsum and limestone can also
be used in reclaiming sodium or acid cit-
rus soils as well in preventing these con-
ditions is indicated by their application
to two of the most seriously affected plots
in the sulfate of ammonia treatment (No.
15) and also in the nitrate of soda treat-
ment (No. 27) in the spring of 1946. In
the former, limestone has been applied
annually at the rate of one ton per acre
each year. In the latter, one ton of gypsum
(later increased to 2 tons) per acre has
been applied annually. The application of
the nitrogen fertilizers in these plots has
[15]
been continued. An improvement in tree
condition was noticeable in the spring of
1947 and by the fall of 1948 the trees
were in very much better condition than
the trees of the two plots of each treat-
ment which did not receive gypsum or
limestone. Tree yields also increased in
the two crops set after the calcium treat-
ments had become effective.
The effects of large quantities of urea
on relative yields may possibly be due to
harmful products which may form tem-
porarily in the soil (Foote, 1945). How-
ever, it seems more probable that they
result from a moderate deterioration of
soil structure due to the replacement of
calcium. The latter theory is suggested by
the reduced infiltration capacity of the
soil of the urea-treated plots and by the
observations of Pierre ( 1931 ) , who found
urea to have about one-half the acid-
forming capacity of ammonium sulfate.
It also seems likely that the depression of
relative yields of the mixed sources of
nitrogen treatment (No. 13) has been due
to the loss of active calcium. All of the
ingredi°nts in this treatment (sulfate of
ammonia, nitrate of soda, and dried
blood) tend to replace calcium in the soil
by base exchange either directly or in-
directly through the production of acid-
ity. If lack of adequate exchangeable
calcium is responsible for the poorer soil
structure and lower relative yields in the
urea and the mixed sources of nitrogen
treatments, applications of gypsum or of
limestone should be helpful.
The deterioration of soil structure and
the decrease in relative yields which have
resulted from the continuous use of large
amounts of nitrate of soda, of sulfate of
ammonia, or of urea, have been more
severe in plots where no covercrops have
been grown than in the covercropped
plots which have received the same fer-
tilizers and for which data are presented
here (see plates 2 and 3, pages 30 and
31). Apparently the effect of the cover-
crop is expressed through its ability to
favor good soil structure and water infil-
tration, as indicated in figure 3 and as
suggested by the effects of added organic
matter in the text table on page 21. The
application of bulky organic materials to
the soil is also very helpful in the preven-
tion of soil deterioration and in the recla-
mation of structurally deteriorated soil.
This will be discussed in another section.
The manner in which the chemical ni-
trogenous fertilizers have been used in
this experiment has been extreme and
not in accord with the best commercial
fertilizer practice. The chemicals have
be~n used without variation to the exclu-
sion of all other fertilizers, added organic
materials or agricultural minerals. Under
common farming conditions, when fer-
tilizers are chosen on the basis of their
unit cost, it is expected that a variety of
materials be used. It is probable that
some would have little effect on the ex-
changeable calcium in the soil and that
some, such as organic matter, might tend
to improve the structure of the soil. More-
over, the use of applications as heavy as
those used since 1939 has been shown to
be more harmful to soil structure than
the use of the same total quantity of ma-
terial when applied in several small doses
(Fireman, Magistad, and Wilcox, 1945) .
Furthermore, the soil on which the ex-
periment is located is relatively low in
exchangeable and total calcium when
compared to many irrigated soils on
which citrus is grown in California. The
irrigation water is also relatively low in
calcium. In a great many locations harm-
ful effects would not be expected to oc-
cur within the same length of time as they
did in this particular orchard even if the
manner of use were the same (Aldrich,
Chapman, and Parker, 1944). However,
if the period of use were long enough,
similar results might be expected on
many soils.
The results reported here may be help-
ful in the choice of fertilizers for loca-
tions where the soil is naturally quite
[16]
acid (pH less than 5.5) or where it has
been made so by application of sulfur or
of acid-forming fertilizers. It is obvious
that the liberal use of sulfate of ammonia,
or of other materials which are highly
acid forming, on such soil would not be
advisable. Likewise, nitrate of soda
should not be applied (1) to soils which
already contain an excess of sodium or
(2) under conditions where a high so-
dium irrigation water is used. Low buffer
capacity and low base exchange capacity
of soils render them more susceptible to
injury by excessive applications of ni-
trate of soda or sulfate of ammonia. In
some citrus regions, such as South Africa,
citrus soils vary so greatly in acidity and
exchange capacity that great care is ad-
vised in the choice of the nitrogenous
fertilizers to be used (Naude, 1949).
Although the use of large annual quan-
tities of nitrate of soda or ammonium sul-
fate is less harmful if the total amount
is divided into several small applications,
there was evidence in the early years of
the experiment that, at a low rate of ni-
trogen supply, split applications were no
more efficient in providing nitrogen for
the trees than single annual applications.
During the first 12 years of the experi-
ment, when one pound of nitrogen was
applied per tree each year, there was a
series of treatments involving the use of
equal quantities of nitrate of lime applied
at different seasons of the year, and one
treatment involving 3 split applications
during the year. Under the conditions of
partial nitrogen deficiency which existed
then, the yield responses of the trees were
essentially equal whether the fertilizer
was applied in one treatment in Febru-
ary, in June, and in October, or in three
split applications made in those months.
Packing-house data obtained in 6 years
of the period during which these particu-
lar trials were in operation show that the
fruit size and the commercial grade of the
crop were not affected by the method in
which nitrate of lime was applied.
NITROGEN FROM CONCENTRATED
ORGANIC SOURCES
The yields which have been obtained
from the use of concentrated organic
sources of nitrogen are also given in
table 1 relative to the yields resulting
from the use of nitrate of lime. In the first
12 years of the experiment, 1928-39,
when one pound of nitrogen was applied
annually per tree, the use of dried blood
meal (tr. 17) or of cottonseed meal (tr.
43) in covercropped plots resulted in
yields which did not differ significantly
from those obtained with nitrate of lime
(tr. 21) or with the other chemical fer-
tilizers. However, in the following period,
1940-49, when 3 pounds of nitrogen per
tree were applied, the trees of the cotton-
seed meal treatment (No. 43) produced
yields which, though on the average equal
to those of the nitrate of lime treatment,
were superior to those resulting from the
use of other concentrated materials. The
yields of the dried blood treatment were
slightly lower than those obtained with
nitrate of lime, but the difference is of
doubtful statistical significance. Dried
blood and cottonseed meal are ordinarily
more expensive per unit of nitrogen than
the chemical fertilizers, as they are used
extensively as feed for livestock.
The use of bulky organic materials as
a part of fertilizer programs, including
the use of calcium-replacing fertilizers,
is expected to be particularly helpful. In
several instances the use of sufficient
manure to provide one-half of the applied
nitrogen (the other half being derived
from a concentrated source) has given
satisfactory results over a period of many
years. In one case, nitrate of soda was
used as the source of nitrogen (tr. 29).
Sulfate of ammonia has also been used
with manure in parts of the orchard
which are adjacent to the experimental
plots. In addition, the application of urea
with manure has given good results in
treatment C, as will appear later (see
table 3, page 21).
[17]
Effects on fruit size
and grade
Observations upon the size and com-
mercial grade of the fruit harvested from
unfertilized trees (tr. 6) and from trees
which were fertilized with the various
concentrated sources of nitrogen are sum-
marized in table 2. Data were obtained
on 6 crops during 1928-40 which were
affected by applications of the various
fertilizers at the rate of one pound of
nitrogen per tree each year. During
1941-49 grades and sizes were deter-
mined on the crops in 8 years. It was
during this period, when applications
were made annually at the rate of 3
pounds of nitrogen per tree, that the
harmful effects of certain fertilizers ap-
peared.
It will be noted that in each of the
periods the fruit of the unfertilized
treatment (No. 6) was larger, i.e., there
was a larger proportion by volume of
fruit of size 220 and larger, than the fruit
of any of the fertilized treatments. This
is believed to be the result of the small
crop produced by the unfertilized trees.
The use of the various chemical sources
of nitrogen (trs. 21, 27, 15, 18, 12) ap-
pears to have resulted in nearly equal
sizes in both of the periods of the experi-
ment when different rates of application
were used. Although the two nitrate
sources of nitrogen seem to have pro-
duced slightly larger fruit, this was not the
case when the fertilizers were used with-
out covercrops. Supplements of gypsum
to the nitrate of soda treatment (No. 28)
or the use of limestone with sulfate of am-
monia (tr. 16) had little influence upon
the size of the fruit. In view of the marked
effects upon the soil and the trees of the
continuous use of nitrate of soda and of
sulfate of ammonia, the failure of these
materials to significantly affect fruit sizes
is of considerable interest.
TABLE 2 — Effects of Various Concentrated Sources of Nitrogen upon Percentage
of Fruit by Volume of Desirable Sizes and Grades in Two Periods of
Time in Which Different Rates of Application Were Used
Fertilizer treatment 8
220' s and larger,
per cent
Fancy and choice,
per cent
1928-40
1941-49
1928-40
1941-49
21. Nitrate of lime
55.2
52.2
50.5
51.0
50.1
53.0
53.8
58.1
57.8
52.1
54.5
52.1
53.9
54.0
87.4
81.6
86.5
85.2
86.2
84.5
86.1
78.8
73.3
70.0
75.3
71.7
75.0
74.3
27. Nitrate of soda
15. Sulfate of ammonia
18. Urea
12. Mixed sources of nitrogen b
28. Nitrate of soda and gypsum
16. Ammonium sulfate and limestone
17. Dried blood meal
52.6
61.8
55.9
59.8
86.8
86.1
75.0
77.6
43. Cottonseed meal
6. No fertilizer
64.2
61.1
87.2
79.2
a All with winter covercrops. One pound of nitrogen was applied annually per tree 1927-39, and three
pounds per tree annually 1940-49 from the sources indicated. Gypsum was used at the rate of one ton per
acre annually, and four tons of limestone per acre were applied every fourth year in treatments 28 and 16,
respectively.
b Nitrate of soda, sulfate of ammonia, and blood meal, each supplying one-third of the total amount of
nitrogen applied each year.
[18]
The effects of the organic concentrates,
dried blood, and cottonseed meal appear
to differ slightly, as may be seen in table
2. The use of dried blood (tr. 17) resulted
in fruit of about the same size as that
produced by the plots receiving the chem-
ical sources of nitrogen. However, cotton-
seed meal (tr. 43) caused a slightly
larger proportion of 220's and larger, by
volume, than any of the other concen-
trates. In both periods the size of the fruit
was nearly as large as that produced by
the unfertilized trees (tr. 6) although the
yield was much greater (see table 1, page
11), The average increase in size of the
fruit of the cottonseed meal plots is not
thought to be due to chance. This meal
contains appreciable concentrations of
mineral elements and of organic matter.
In view of results to be reported later in
this paper it appears probable that or-
ganic matter and the nutrient elements
exert an effect on fruit size in this or-
chard. If so, the observations on size of
the fruit of the cottonseed meal treatment
are in accord with other data reported
later. Dried blood, in comparison, was
applied in smaller quantities because of
its higher nitrogen content (usually 12-
14%). It also contains small quantities
of mineral elements. Moreover, its or-
ganic compounds decompose very rapidly
and completely in the soil. For these rea-
sons dried blood appears to react in a
manner similar to that of the chemical
sources of nitrogen.
Table 2 also shows the percentage of
the total volume of fruit which was of
fancy and choice grades. It is evident that
the fertilizers have had very little effect
in either period of this experiment on
the grade of the fruit as determined by
commercial grading procedures. This is
especially interesting because of the strik-
ing effects of the sole use of nitrate of
soda (tr. 27) and sulfate of ammonia (tr.
15) upon tree condition and yields in the
latter half of the experimental period.
B. USE OF ORGANIC MATTER
Effects of covercrops and
of manure
Organic matter is supplied to citrus
soils in California by the growing of
covercrops during the winter rainy sea-
son and by the application of bulky or-
ganic materials such as various kinds of
manures, straws, hays, cottonseed hulls,
etc. The value and management of winter
covercrops have been discussed in detail
by Batchelor (19486). They frequently
increase the productivity of the soil and
help control soil erosion. Their use for
the improvement of soil structure in citrus
orchards frequently results in increased
rate of water infiltration during irriga-
tion. In California this latter effect is
especially prominent on the older granitic
soils, such as the Ramona loam soil upon
which the present experiment is con-
ducted.
In this experiment the 4 chemical ni-
trogen sources already discussed (nitrate
of lime, nitrate of soda, urea, and sulfate
of ammonia) have been used with and
without winter covercrops. Manure has
also been used with and without winter
covercrops, and unfertilized treatments
have been maintained under both systems
of culture.
To prevent weed growth the land in the
clean culture series of treatments has been
cultivated one to three times during the
winter while the covercrops were being
grown on the other plots. In the cover-
cropped plots Trieste mustard {Brassica
juncea, cult.) or bitter clover (Melilotus
indica) were usually planted, but in the
last few years mixtures of purple vetch
(Vicia atropurpurea) with either mus-
tard, oats (Avena sativa) or barley {Hor-
deum sativum) were used. The changes
in kind of covercrop were due to in-
creased shading as the citrus trees became
larger, to weed competition, or to other
[19]
UJ
o
oc
UJ
130
a.
1
o
-J
120
UJ
>-
110
UJ
>
1922 1928 1932 1936 1940 1944 1948
-27 -31 -35 -39 -43 -47 -49
FIG. 4. Relative mean increase in total orange yields, resulting from the use of covercrops with
4 concentrated chemical sources of nitrogen over periods of 4 or 2 years from 1927 until 1949.
The relative yields of plots which were without covercrops equals 100. An equal number of trees
were fertilized with nitrate of lime, nitrate of soda, urea, and sulfate of ammonia under each
system of culture. One pound of nitrogen was applied annually per tree from 1927 to 1939, and
three pounds of nitrogen from 1940 to 1949. The lower relative yield of the covercropped plots
in 1948 and 1949 appears to be due to chance variations in the yields of 1949.
factors which affected the stand of par-
ticular covercrops.
On the unfertilized plots the growth of
covercrops has been very poor for many
years, and in years when mixtures of
seeds have been sown the legumes have
predominated. The application of nitrog-
enous fertilizers has greatly stimulated
covercrop growth. This improved growth
was noted following the general increase
in rate of nitrogen fertilization in 1939-
40. The growth of covercrops of all kinds
has been further increased by applica-
tions of organic matter at about the same
time that the covercrops were planted.
This may be due, at least in part, to water
relations in the soil which were more fa-
vorable to the young covercrop plants.
ON RELATIVE YIELDS
The average differences in yields result-
ing from the use of winter covercrops
with the 4 chemical sources of nitrogen
mentioned above are given in figure 4 for
the various experimental periods, as well
as for the preliminary period 1922-27
when no fertilizers were applied. In the
graph the mean yields of the clean culture
plots are taken as equal to 100 per cent
for each period. The relative yields of the
covercropped plots increased gradually
during the experimental period until a
difference of about 30 per cent was
reached in 1940-47. During 1948-49 the
difference was only 19 per cent, but this
lower value appears to be due to chance
seasonal variations which affected the
crop of fruit picked in 1949.
Although the growing of covercrops
has increased yields in this experimental
orchard, the amount of the increase has
been influenced by the source of nitrogen
which was used. This has been particu-
larly noticeable in the later years when
heavy applications of fertilizer were
made. Thus the percentage increase in
[20]
yield during 1940-47 due to the growing
of covercrops from treatments involving
the chemical fertilizers as well as from
treatments involving manure as the
source of equal amounts of nitrogen was
as follows:
Nitrate of soda 92 per cent
Sulfate of ammonia .... 34 per cent
Urea 14 per cent
Nitrate of lime 14 per cent
Manure 2 per cent
It will be noted that the growing of cover-
crops caused a greater percentage in-
crease in orange yields when nitrate of
soda or sulfate of ammonia was used as
a nitrogen source. These were the fertiliz-
ers which, in the same period of time, had
the more harmful effects on soil structure
and relative yields. Where nitrate of lime
or urea was used, the effect of the cover-
crop on yields was less, and about equal.
No tangible effect of covercrops on yields
resulted when manure was used. The lat-
ter point is of especial interest, since it
indicates that the application of bulky or-
ganic materials may supplant covercrops
where soil erosion is not a hazard. It is
in agreement with the observations of
Vaile (1924) and of Batchelor (19486).
The effects on yields of the use of cov-
ercrops and of varying amounts of or-
ganic matter derived from manure are
given in table 3 for each period of the
experiment. Equal amounts of nitrogen
were applied in all treatments in each pe-
riod of the experiment, but only one
pound of nitrogen was applied per tree
TABLE 3 — Relative Yields as Influenced by Organic Matter from Covercrops
and from Manure When Used in Various Ways, When Equal Amounts
of Nitrogen Are Used in the Fertilizer Program
Treatment'
1928
-31
1932
-35
1936
-39
1928
-39
1940
-43
1944
-47
1948
-49
1940
-49
Mean annual yield per tree (pounds)
21. Nitrate of lime, with
covercrop
123
160
118
134
203
185 183 192
Relative yields (treatment 21 = 100)
21. Nitrate of lime, with
covercrop
20. Nitrate of lime, without
covercrop
C. Urea and manure with
covercrop
30. Manure in fall, without
covercrop
32. Manure in spring, without
covercrop
31. Manure in fall, with
covercrop
100
97
93
97
85
92
100
100
100
100
100
100
76
86
83
82
94
101
100
103
98
101
118
119
86
103
94
83
106
109
79
80
81
68
98
102
83
70
81
80
113
111
89
102
95
80
85
88
100
90
111
97
86
98
18. Urea, with covercrop
96
84
a A total of one pound of nitrogen from all sources was applied annually per tree in each treatment during
1928-39, and 3 pounds of nitrogen were similarly applied during 1940-49. Chemical fertilizers were applied
in early spring and manure was applied at the time of planting covercrops in> the fall except as indicated.
[21]
annually during 1928-39, and 3 pounds
per tree each year during 1940-49. In
this table the yields of the covercropped
nitrate of lime treatment (No. 21) are
taken as 100 per cent.
It is evident that where nitrate of lime
was used under conditions of clean culti-
vation (tr. 20) that the yields were con-
sistently lower than the yields of the cov-
ercropped nitrate of lime plots (tr. 21).
In the last 10 years the average yields of
the covercropped plots have been about
11 per cent greater. Where covercrops
were grown and the quantity of organic
matter further increased by the addition
of manure in such an amount that it sup-
plied one-half of the additional nitrogen
applied (the balance of the nitrogen be-
ing supplied by urea [tr. C] ) , no demon-
strable increase in yield over that of the
covercropped nitrate of lime treatment
(No. 21) occurred during the 12 years,
1928-39 (when one pound of nitrogen
was applied annually to each tree). Nor
was there any increase during the first
subsequent period of 4 years (1940-43)
when nitrogen applications were made
at a higher rate.
Thus during the first 16 years of the
experiment, it is evident that the use of
manure with urea, each supplying one-
half of the nitrogen applied, has resulted
in about the same yields as those obtained
with nitrate of lime when covercrops were
grown in each case. However, in the
1944-47 and the 1948-49 periods, ma-
nure and urea (tr. C) gave an apparent
increase in yield amounting to 18 per
cent over the yield of the nitrate of lime-
covercrop treatment (No. 21). This dif-
ference is on the borderline of statistical
significance. 7
It will be recalled (see p. 12 and table
1) that in the 1940^9 periods the use
of urea did not produce as high yields as
did the use of nitrate of lime when both
were used with winter covercrops. It is
therefore of particular interest to com-
pare, in table 3, the relative yields result-
ing from the use of manure and urea (tr.
C) with those of the urea treatment (No.
18). It may be seen that the yields were
about equal during 1928-39, when one
pound of nitrogen was applied per tree
annually, but that the yields of the urea
treatment were considerably poorer than
those of the manure-urea treatment dur-
ing 1940-49 when 3 pounds of nitrogen
were applied annually to each tree. In this
case, a moderate quantity of manure— ap-
plied at a rate supplying one-half the total
nitrogen applied— has been very bene-
ficial and has prevented the appearance
of secondary harmful effects of urea
which have been referred to previously.
That manure has also improved the soil
structure is indicated by increased rates
of infiltration of water.
Manure has also been used experimen-
tally to supply all of the applied nitrogen.
When used under conditions of clean cul-
tivation, fall applications of manure alone
(tr. 30) produced yields during 1928-39
(when one pound of nitrogen was ap-
plied) which were about equal to those
of the covercropped treatment in which
manure supplied only one-half of the ap-
plied nitrogen (No. C). However, in the
1940-49 period (when 3 pounds of ni-
trogen were applied) the yields of the
manure-alone treatment tended to be
lower than those of the manure-urea
treatment.
When the supply of organic matter was
still further increased in a manure treat-
ment by the growing of winter covercrops
(tr. 31), the yields in 1928-39 (when
one pound of nitrogen was applied per
tree annually) were markedly below
those of the covercropped treatment in
which manure supplied only one-half the
nitrogen (tr. C) . They were also less than
those of the clean-cultivated manure treat-
ment (No. 30) in the same period. In
the 1940-49 period (when 3 pounds of
nitrogen were applied), the growing of
7 Where the probability of chance occurrences is 5 per cent.
[22]
winter covercrops had little, if any, effect
on the yields of the manure-alone treat-
ments (Nos. 30 and 31), but the yields
of both of these were slightly less than
the yields of the manure-urea-covercrop
treatment (No. C).
It may be seen from table 3 that spring
applications of manure as the sole fer-
tilizer under conditions of clean cultiva-
tion (tr. 32) resulted in low average
relative yields in both periods when dif-
ferent rates of fertilization were used
(compare with treatment 30) .
The most effective use of manure re-
sulted when it was used in the fall to
supply one-half of the total amounts of
nitrogen applied and when winter cover-
crops were grown (tr. C). The use of
manure as the sole source of nitrogen did
not at any time result in greater yields
and under some conditions it resulted in
lower yields than those obtained in treat-
ment C. When manure was used alone
the time and rate of the application— as
well as the winter covercropping prac-
tices—affected the yields.
The effects of these factors may be
explained by their influence upon the
availability of nitrogen to the trees, par-
ticularly in the blooming and fruit setting
period. Covercrops are competitive for
nitrogen during the period of their
growth. Moreover, the microorganisms
which cause the decay of covercrops and
manure absorb nitrogen from the soil
during the period of decomposition, since
such materials do not contain enough ni-
trogen to satisfy the needs of the organ-
isms which act upon the organic matter
applied. The period of decomposition
also may be prolonged under such condi-
tions (Broadbent and Bartholomew,
1949) . The amount of nitrogen available
to the trees is thus reduced. This competi-
tion is temporary, since the nitrogen— or
most of it— is later made available. How-
ever, smaller crops may be expected if
nitrogen availability is reduced at sea-
sons which are critical for fruit set-
ting, especially if the reserve supply of
nitrogen is small. Under conditions of
limited nitrogen supply such as existed
during 1928-39, the depressing effects
of the sole use of manure with covercrops
was more severe than in the last 6 years
of 1940-49 when heavier applications
were made. This difference appears to
be due to the accumulation of greater
quantities of available nitrogen in the
soil and, perhaps, in the trees.
It may be seen that the relations be-
tween the total quantities of organic mat-
ter and nitrogen which are applied in
the fertilizer program may be of con-
siderable practical importance. These are
again referred to in a subsequent section
(pp. 38-39).
Although other influences are not pre-
cluded, it is probable that the beneficial
effects upon orange yields resulting from
the growing of covercrops and the addi-
tion of moderate amounts of organic mat-
ter are primarily due to an improvement
in the physical structure of the soil. The
importance of good soil structure in this
orchard has already been illustrated in
connection with the effects of chemical
fertilizer. Parker and Jenny (1945) have
presented data which show that the
soil structure, as indicated by the rate
of water infiltration and by other meas-
ures, has been reduced by cultivation and
traffic in this orchard and that additions
of organic matter in manure counteract
this effect. Although the beneficial effect
of the organic matter upon the soil struc-
ture was related to the quantity applied,
it is evident that increases in yield due
to applications of organic matter may be
limited by other factors.
The effects of covercrops and organic
matter applications which are discussed
here would probably have been much
greater if the plots in the experiment had
not been periodically sprayed with zinc
compounds for the treatment and pre-
vention of zinc deficiency (mottle-leaf) .
Before zinc treatment was begun in 1934.
[23]
the trees in the clean-culture plots which
received no bulky organic material in the
fertilizer treatment were seriously af-
fected with zinc-deficiency symptoms and
the yields of the most severely affected
trees were depressed considerably. By
1935 the application of foliage sprays
containing zinc had removed the symp-
toms of zinc deficiency and its depressing
effects upon yields (Parker, 1937a).
The effect of zinc treatment on a tree in
a clean-culture plot is shown in figure 5.
The results of the experimental treat-
ments which are reported in this section
show that covercrops have generally in-
creased orange yields in this cultivated
orchard. The growing of covercrops may
supply all the organic matter which is
needed by soil which has good structure.
When unwise use of chemical fertilizers
or compaction has resulted in structural
deterioration, the further addition of or-
ganic matter in the form of manure— at
rates such that the manure supplies one-
half of the total quantity of nitrogen— has
further increased yields. However, the use
of manure to supply all of the applied
nitrogen has depressed yields, especially
when the manure was used in conjunc-
tion with the growing of covercrops or
applied in the spring rather than in the
fall. This depression was more pro-
nounced when the rate of nitrogen appli-
cation was low.
When manure was used in the fall to
supply one half the total amount of nitro-
gen (the balance being supplied by a
chemical fertilizer) it was an efficient
source of nitrogen.
EFFECTS OF ORGANIC MATTER FROM
COVERCROPS AND FROM MANURE
ON FRUIT SIZE AND GRADE
Average effects upon fruit sizes and
commercial grade of variations in cover-
cropping practices and in the use of ma-
nure are shown in table 4 for the 1941-
49 period. In all cases equal amounts of
nitrogen were applied.
It is apparent, when nitrate of lime was
used as the sole fertilizer, that the grow-
ing of covercrops (tr. 21) resulted in a
slightly greater percentage of large-sized
fruit than occurred in the clean-cultured
treatment (No. 20) . In other comparisons
of the effects of covercrops when used
with chemical fertilizers in this experi-
ment, a similar increase in fruit size has
generally been observed.
When covercrops were grown and the
quantity of organic matter applied to the
soil was further increased by the use of
manure in sufficient quantity to supply
one-half of the nitrogen (tr. C), a slight
additional gain in the proportion of
large-sized fruit occurred. It may be seen
from the table that the total gain of this
treatment (No. C) over the treatment
TABLE 4 — Effects of Covercrops and of Manure on the Size and Commercial
Grade of the Fruit — Harvests of 1941-49
Treatment a
220's and larger,
per cent
by volume
Fancy and choice,
per cent
by volume
20. Nitrate of lime, without covercrop
52.7
58.1
63.4
67.4
69.6
70.3
75.5
78.8
79.5
78.9
79.6
80.7
21. Nitrate of lime, with covercrop
C. Urea-manure, with covercrop
30. Manure in fall without covercrop
32. Manure in spring, without covercrop
31. Manure in fall, with covercrop
a In all cases, one pound of nitrogen was applied per tree in each year 1928-39, and 3 pounds of nitrogen
per tree were applied annually in 1940-49.
[24]
FIG. 5. Effects of zinc treatment on an orange tree which was fertilized
with inorganic nitrogenous materials. (Top) April, 1934, before treatment.
(Bottom) February, 1935, after one season's response to treatment with
zinc foliage spray. The spray was applied in the spring of 1934.
[25]
which received nitrate of lime but no or-
ganic matter (No. 20) was, however, ap-
preciable (63.4% of 220's and larger as
compared to 52.7%).
The table also shows that the use of
manure as the sole source of nitrogen
in the absence of covercrops resulted in
still larger fruit sizes, whether the ma-
nure was applied in the spring (tr. 32)
or in the fall (tr. 30) , and that the largest
fruit (70.3% of 220's and larger) was
produced when manure alone was used
with covercrops (tr. 31). The largest
quantity of organic matter was applied
to the soil in the last treatment mentioned.
These results indicate an increase in
size of oranges resulting from the grow-
ing of covercrops and also from the ap-
plication of organic matter in manure in
this orchard. The magnitude of the fruit
size response was positively correlated
with the quantity of organic matter which
was applied to the soil. This positive
relationship between size of fruit and
quantity of organic matter applied is par-
ticularly interesting when comparisons
are confined to treatments 20, 21, and C,
in which the range of organic matter ap-
plication is from zero to that supplied
by covercrops plus a moderate quantity
of manure. In these treatments the aver-
age yields, as well as the average fruit
sizes, increased with increasing applica-
tions of organic matter. Although the use
of still larger quantities of organic mat-
ter in the manure-alone treatments (Nos.
30, 31, and 32) resulted in the largest
fruits, the yields of fruit from these treat-
ments tended to be slightly less than those
of the manure-urea treatment (No. C) in
the period under discussion. In view of
the general inverse relationship between
size of crops and fruit size (Parker, 1934)
it is possible that the increases in the
proportion of large size fruit of the
manure-alone treatments over the ma-
nure-urea treatment (No. C) were influ-
enced by the smaller crops of the former
treatments.
In this orchard, the effects of adding
organic matter upon fruit size appear to
be due to two factors:
( 1 ) The result of their effects upon the
structure of the soil, improved water
penetration, and related factors (Parker
and Jenny, 1945). In various orange or-
chards, prolonged water shortage has
frequently been observed to result in
small orange sizes. Similar effects have
also been observed with lemons (Furr
and Taylor, 1939) and with deciduous
fruits (Hendrickson and Veihmeyer,
1948). Smaller fruit at the lower end
of irrigation runs under the furrow
system have been commonly observed
and reported for numerous orange or-
chards (Taylor, 1941; Teague, 1949;
Puffer, 1949; and Yarick, 1949). Al-
though severe water shortages have not
been observed in the organic matter treat-
ments under discussion, variations in
water supply have undoubtedly occurred.
(2) The presence of chemicals other
than nitrogen. Data to be presented in
the next section show that fruit sizes have
been increased by the application of
potash. Manure and other bulky organic
materials contain appreciable quantities
of potash, phosphate, and other nutrients
as well as nitrogen and organic matter.
For example, over a period of 22 years
the manure applied in this experiment
has contained nitrogen (N), phosphoric
acid (P 2 5 ), and potash (K 2 0), and or-
ganic matter in the proportions 1.0 : 0.75 :
2.10:36. Thus, when one pound of nitro-
gen was applied in manure, about 2
pounds of potash were applied. Leaf
analysis data indicate that the potash
which is supplied by manure is absorbed
by the trees. Moreover, Jones and Parker
(1949) have shown that the concentra-
tion of potassium in the juice of the
oranges from these plots has also been in-
creased by the manure applications. The
effects of covercrops upon potash absorp-
tion are not so obvious, and their effects
on size of fruit may be primarily related
[26]
to an improvement in soil structure and
water relations.
Effects on the commercial grades of the
fruit of growing covercrops and adding
varying quantities of organic matter in
the form of manure are also shown in
table 4 for observations made during
1941-49. In all comparisons, the differ-
ences are small. Very slightly lower per-
centages in the proportion of fruit by
volume which was packed in the fancy
and choice grades occurred in the clean-
cultivated plots receiving nitrate of lime
(tr. 20). The use of covercrops and/or
manure appears to have improved the
grades slightly, but the grades in all treat-
ments receiving them were about the
same. It is of interest that the heaviest
rates of application of organic matter did
not lower packing-house grades.
C. USE OF PHOSPHATE AND POTASH
Effects when applied with
concentrated nitrogen
fertilizer
In numerous areas of California, vege-
table, forage or cereal crops have shown
responses in yields to fertilization with
materials containing phosphate. On a few
soil types in this state, olives and decidu-
ous fruit trees have responded to potash
fertilization. The results of field experi-
ments with citrus in California heretofore
reported have not indicated increases in
yield or changes in grade or size of the
fruit as a result of applications of these
materials. In comparison with those of
other regions, California citrus soils con-
tain fairly high contents of these elements.
Moreover, Chapman (1934 and 1941)
found that substantial accumulations of
phosphate and potash have occurred in
many citrus orchards as a result of the
application of various fertilizers, espe-
cially of bulky organic materials. Re-
cently Aldrich and Coony (1951) have
obtained responses by lemon trees to
phosphate applications on certain acid
soils where bulky organic materials were
not used.
The use of manure and other bulky or-
ganics to supply about half of the nitrogen
applied has been generally recommended
and practiced. This procedure became so
well established that several fertilizer
trials were conducted on land which had
previously received manures, and others
were carried out to determine the effects
of various supplements to the nitrogen-
manure program. In recent years, how-
ever, bulky organic materials have at
times been very expensive to buy or to
apply, so that many growers discontinued
their use and increased their applications
of chemical sources of nitrogen. This sub-
stitution has also been made in many
cases where weed spraying and non-
cultivation systems of soil management
have been adopted. Since bulky organic
materials supply large amounts of phos-
phate and potash, the long-term effects
of the use of the chemicals in the absence
of applications of bulky organic mate-
rials is especially interesting at this time.
The conditions under which the present
experiment was conducted permit an un-
usual opportunity to evaluate yield and
fruit responses of orange trees to applica-
tions of phosphate and potash without
complications from the residual effects
of previous fertilization.
Phosphate and potash have been used
singly and together with urea as a source
of nitrogen. The studies have been con-
ducted under conditions of clean culture,
as well as when winter covercropping has
been employed. The usual rate of appli-
cation of potash was one pound of K 2
per tree annually. However, in one cover-
cropped trial (No. 10) it has been used
at the rate of 3 pounds of K 2 per tree
since 1940. Phosphate was applied at the
rate of one pound of P 2 5 per tree an-
nually. The responses obtained without
covercrops were generally similar to
[27]
those with covercrops, hence only the re-
sults obtained with winter covercrops are
reported here.
In addition, one treatment (No. 13)
involved the use of phosphate and potash
in a factory-mixed fertilizer (8-8-8). The
nitrogen in this fertilizer was supplied
by 3 concentrated sources of nitrogen.
The quantities of phosphate (P 2 5 ) and
of potash (K 2 0) applied in this treatment
were equal to the quantity of nitrogen
applied. This treatment is contrasted with
one (No. 12) in which equal amounts of
the nitrogenous materials are used with-
out phosphate or potash. Covercrops were
grown in the plots of both treatments.
ON RELATIVE YIELDS
Relative yields of these groups of treat-
ments over a period of 22 years are given
in table 5. The nitrogen treatment (No.
18) which received only urea is used as
a basis of comparison.
In the series in which urea was used
as a nitrogen source, potash used alone or
with phosphate has not caused significant
increases in yields, as judged by analysis
of variance (Snedecor, 1946). However,
the differences between comparable treat-
ments are generally very slightly in favor
of those which have received potash. The
yield differences are greatest in the last
2 years for which data are available
(1948-49). Additional time will be
needed to determine if a significant trend
is developing in this orchard after 20
years of differential fertilization. The
data of earlier years indicate that the
slightly larger yields of the potash treat-
ments are due to the somewhat larger
fruits which they produced (see p. 36),
and it has been found that the fertilizer
had no effect on the number of fruits
harvested per tree (Parker and Jones,
19506).
Table 5 also shows that in the urea
series of treatments the addition of phos-
phate failed to cause a yield response.
In fact there was a suggestion in 1936-39
that the use of phosphate with covercrops
[28
(tr. 8) slightly depressed the yields of
the trees (Parker and Batchelor, 1942).
This depression occurred during the pe-
riod of greatest nitrogen deficiency in
these plots and persisted during the 1940-
43 period. It is not particularly surpris-
ing in view of the frequently observed
antagonism between the absorption of
phosphates and nitrates by plants.
It may also be seen from the table that
the use of potash and phosphate in a
factory-mixed fertilizer with several nitro-
gen sources (tr. 13) failed to result in
significantly better yields than treatments
consisting of those same 3 nitrogen
sources alone (tr. 12) or of urea (tr. 18) .
It was previously pointed out in this
bulletin that the continuous use of urea
and mixed nitrogen sources has caused a
slight deterioration of the soil structure
in recent years, and has not produced as
high yields as nitrate of lime alone. Yield
data for the nitrate of lime treatment (No.
21) are repeated in table 5 to show that
the addition of phosphate or potash did
not overcome the depressing effects of
these nitrogen sources on yields. How-
ever, when manure was used with urea,
each supplying one-half of the nitrogen
applied (tr. C, bottom line of table 5) , the
yields in recent years (1940^19) have
been equal to or superior to those result-
ing from the use of phosphate or potash
with nitrogen from either mixed nitrogen
sources or urea. This indicates that phos-
phate, potash, or other chemicals in the
manure have not been responsible for the
increased yields of the urea-manure treat-
ment.
The failure of phosphate and potash to
significantly increase yields during a
period of at least 20 years in these experi-
ments has not been due to lack of avail-
ability of these fertilizers to the trees.
Chapman (1934, 1941) has shown that
phosphate has penetrated the soil as far
as the root-zone of the trees, and that pot-
ash accumulation has been increased in
the leaves of trees fertilized with this ele-
Wkm
PLATE 1. Tip-burn of old and young leaves which dropped in the spring of 1943 from trees which
had been fertilized solely with sulfate of ammonia for 16 years. Leaves having similar symptoms
were shed by trees which had received only nitrate of soda.
[29]
y&z-te
PLATE 2. Defoliation and lack of growth resulting from the continuous
use of sulfate of ammonia, both without (top) and with (bottom) winter
covercrops. Photographed May, 1948. Compare with plate 4 (bottom).
These harmful effects were prevented by the use of limestone.
[30]
PLATE 3. Defoliation and lack of growth resulting from the continuous
use of nitrate of soda, both without (top) and with (bottom) winter cover-
crops. Photographed May, 1948. Compare with plate 4 (bottom). These
harmful effects were prevented by the use of gypsum.
[31]
PLATE 4. Response to nitrogen. Unfertilized orange tree 31 years old
(top) and comparable tree (bottom) fertilized annually since 1927. The
fertilizer supplied nitrogen from urea. Photographed May, 1948.
[32]
ment. Haas (1949) also showed that the
potash content of the flowers was in-
creased by fertilization. Evidence of pot-
ash absorption by the leaves is presented
later in this publication. Jones and Parker
(1949) have also found that the applica-
tion of phosphate as well as potash has
resulted in increased concentrations of
these elements in the juice of the fruits.
Availability of the potash and phosphate
which have been applied to the soil is also
shown by the fact that certain minor
changes in the characteristics of the in-
ternal quality of the fruits have resulted
(Jones and Parker, 1949) . The failure of
fertilization with these elements to cause
significant increases in yields is therefore
of particular interest.
The growth of covercrops has not been
affected by the application of either phos-
phate or potash in these experiments. An-
nual crops are usually more sensitive to a
restricted phosphate supply than are tree
crops. In view of the fact that the control
trees in this series of trials have been
growing in the field for 32 years with the
application of only concentrated nitrogen
fertilizers, and that the land had been
previously dry-farmed to cereal crops
without fertilization for at least 38 years,
it is evident that the soil in this orchard
(Ramona loam) possesses a relatively
good supply of phosphate and potash.
ON FRUIT SIZE AND GRADE
The effects of phosphate and potash
when used with concentrated sources of
nitrogen upon the commercial grade and
TABLE 5 — Effects upon Relative Orange Yields of Phosphate and of Potash
When Used with Concentrated Nitrogen Fertilizers
Treatments a
1928-39
1940-43
1944-47
1948-49
1940-49
18. Urea
Mean annual yield per tree (pounds)
127
163
158
161
161
18. Urea
Relative yield (treatment 18 = 100)
100
90
103
103
97
100
85
101
94
88
100
96
105
100
96
100
94
120
107
111
100
91
106
99
96
8. Urea and phosphate
11. Urea and potash
9. Urea, phosphate and potash
10. Urea, phosphate and potash b
12. Nitrogen from three sources c
13. Nitrogen from three sources c with phosphate
and potash in mixed fertilizer (8-8-8) d
100
106
103
101
98
105
87
102
98
102
21. Nitrate of lime
106
104
122
126
117
139
114
135
119
123
C. Urea and Manure
a All with winter covercrops. The rate of application of nitrogen for each treatment was one pound per tree
annually during 1927-39 and 3 pounds per tree annually during 1940-49. Phosphate was applied at the rate
of one pound of P^O, per tree each year in the form of treble superphosphate except as noted for treatment 13.
The annual application of potash was one pound of K 2 per tree from sulfate of potash except as noted for
treatments 10 and 13.
b Potash from muriate of potash at the rate of one pound of K„0 per tree annually during 1927-39. During
1940-49, the annual application was three pounds of K per tree from sulfate of potash.
c Nitrate of soda, sulfate of ammonia, and blood, each supplying Vz of the annual amount of nitrogen
applied.
d The quantities of P 5 and IC.O were equal to the nitrogen applied.
[33]
the size of the fruit harvested from the
experimental trees are indicated in table
6. Averages are given for the observations
made in the crops picked in 1941-49
when 3 pounds of nitrogen were applied
per tree and the supply of nitrogen was
not limiting yields to a practical degree.
The table shows that when used with
urea the additions of potash or of phos-
phate, singly or together, had little effect
on the proportion of the fruit which was
of fancy and choice grades. This is par-
ticularly evident when the results for the
individual years are studied. Similar
comments may be made in regard to the
effects of the use of a mixed fertilizer con-
taining phosphate and potash (tr. 13) as
compared to the results of applications of
identical sources of nitrogen without the
phosphate and potash (tr. 12). The dif-
ferences in fruit grade between these two
treatments appear to be due to chance
variations.
Comparisons of the grades of the treat-
ment receiving nitrate of lime without
phosphate or potash (tr. 21) with the
urea-phosphate-potash treatments and
with the mixed nitrogen treatments which
are given in table 6 indicate that the for-
mer, although producing larger crops,
did not have fruit of lower grade. Like-
wise, when the phosphate and potash
treatments are compared with the high-
yielding manure-urea treatment (No. C)
it may be seen that the fruit of the latter
was, on the average, at least as good in
grade.
The effects of phosphate and potash
fertilizers upon the size of fruit during
1941-49 are also shown in table 6. In
none of the comparisons do the averages
suggest that the application of phosphate
has increased fruit sizes. The comparisons
which may logically be made are between
the urea-phosphate treatment (No. 8) and
the urea treatment (No. 18) , and also be-
TABLE 6 — Effects of Phosphate and Potash When Used with a Chemical Nitrogen
Fertilizer upon Percentage of Fruit of the Larger Sizes and Better Grades
Harvests of 1941-49
Treatments a
220' s and larger
per cent
Fancy and choice
per cent
18. Urea
54.5
55.0
62.9
60.0
66.7
75.3
74.6
75.2
78.0
76.8
8. Urea and phosphate
11. Urea and potash
9. Urea, phosphate and potash
10. Urea, phosphate and potash b
12. Nitrogen from three sources °
52.1
56.7
71.7
75.0
13. Nitrogen from three sources c with phosphate and
potash in factory-mixed fertilizer (8-8-8) d
21. Nitrate of lime
58.1
63.4
78.8
79.5
C. Urea and manure
a All with winter covercrops. The rate of application of nitrogen for each treatment was one pound per tree
annually during 1927-39 and 3 pounds per tree annually during 1940-49. Where phosphate was applied, one
pound of P.,0 5 was applied per tree each year in the form of treble superphosphate except as noted for treat-
ment 13. The annual application of potash was one pound of IC,0 per tree from sulfate of potash except as
noted for treatments 10 and 13.
b Potash from muriate of potash at the rate of one pound of K 2 per tree annually during 1927-39. During
1940-49, the annual application was 3 pounds of K 2 per tree from sulfate of potash.
c Nitrate of soda, ammonium sulfate and blood, each supplying Vb of the annual amount of nitrogen applied.
d The quantities of P 2 5 and K.O were equal to the nitrogen applied.
[34]
tween the urea-phosphate-potash treat-
ment (No. 9) and the treatment which
involves the use of equal quantities of
potash and urea (No. 11). From these
data it is not likely that the effects of bulky
organic materials on fruit size, mentioned
earlier, are due to the phosphate con-
tained in them.
However, the data indicate that the
application of potash, when used with
urea, has caused an increase in the pro-
portion of large-sized fruit. In the early
years of the differential fertilizer treat-
ments (1928-39) during which one
pound of nitrogen was applied, the effect
was small and variable. It appeared to be
either within the limits of experimental
error or to be influenced by nitrogen de-
ficiency. In later years it has been more
consistent. From table 6 it will be seen
that, during 1941-49, 62.9% of the total
volume of fruit picked from the urea-
potash treatment (No. 11) was of size
220's or larger, while 54.5% of the fruit
of the urea treatment (No. 18) was of the
large size. Likewise, the urea-phosphate-
potash treatments (No. 9 and No. 10)
have resulted in a greater proportion of
large fruits than the urea-phosphate treat-
ment (No. 8). An increase is especially
noticeable in the case of the treatment
(No. 10) which has received 3 pounds of
potash (K 2 0) per tree annually since
1939. Of the total numbei of boxes pro-
duced by this treatment 66.7% were of
large sizes, as compared with 55.0% for
the urea-phosphate treatment (No. 8).
The reliability of differences due to fer-
tilization with potash are discussed below.
From the data given in table 6 it will
be noted that the inclusion of phosphate
with potash has not improved the re-
sponse in fruit size resulting from the use
of potash alone. In fact, the averages are
a little lower for the nitrogen, phosphate,
and potash treatment (No. 9) than for the
treatment in which equal amounts of only
nitrogen and potash were used (No. 11).
In the case of the treatment involving the
use of factory-mixed fertilizer containing
phosphate and potash with three sources
of nitrogen (tr. 13), the apparent slight
increase in size of the fruit over that re-
sulting from the use of the same nitrogen
sources alone (tr. 12) appears, therefore,
to be due to the potash in the factory-
mixed fertilizer.
It may also be seen from table 6 that
the use of potash with urea (tr. 11) or
with urea and phosphate (trs. 9 and 10)
has in recent years (1941-49) resulted
in larger fruit than that produced by the
nitrate of lime treatment (No. 21). It
does not seem probable that the difference
in fruit size is due to the larger yields of
the nitrate of lime treatment since, if this
were so, the fruit of the urea-alone treat-
ment should then be affected in the same
way as that in the urea-potash treatment.
On the other hand, relatively large fruits
and larger yields were produced by the
manure-urea treatment (No. C) , presum-
ably because of the effect of the manure
upon potash supply, soil structure, and
moisture relations.
Although no significant effects of phos-
phate and potash on the commercial grade
of the fruit have been observed, these fer-
tilizers have been found to affect meas-
urably the internal quality of the fruits
of this experiment. Jones and Parker
(1949) found that the application of pot-
ash fertilizers has slightly increased the
total acidity and the vitamin C content of
the juice, whereas the use of phosphates
has decreased the concentration of these
two components of the juice. The soluble
solids of the juice were not significantly
affected by either phosphate or potash.
Phosphates, however, tended to produce
fruit having slightly greater specific grav-
ity and greater juice percentage by vol-
ume. This last effect is due to somewhat
thinner rind.
These effects on fruit quality are so
small in our trials that they are probably
of no commercial importance. All of them
have been observed elsewhere, either in
[35]
water culture studies where the supply of
the elements was drastically curtailed or
in field trials which were conducted in
other citrus areas under conditions where
applications of phosphate or potash have
not resulted in yield responses (see Jones
and Parker, 1949, for literature review).
Reliability of the observed
effects of potash on fruit size
Although it is impossible to determine
the statistical probabilities of significant
differences between the percentage of
220's and larger produced by two treat-
ments differing in respect to potash fer-
tilization (see footnote 3, page 7), it is
possible to find the probability of chance
occurrence of the mean of a series of such
differences. Five treatments may be
paired with 5 other treatments which are
identical except for the application of
sulfate of potash. All of these treatments
received equal amounts of nitrogen from
urea, while some were clean cultivated
and some winter covercropped. 8 Jones
and Parker (19506) have determined
the mean differences between these pairs
of treatments for each year in which fruit
sizes were determined. During the early
years of the experiment, 1931-34, the
mean differences were small, variable,
and generally not significant. In the pe-
riod of 1939-49, however, the increases
in fruit size due to potash fertilization
were usually consistent, but varied con-
siderably in magnitude from year to year.
During these years the minimum increase
in the percentage of 220's and larger, by
volume, was 2.3 and the maximum was
12.4. In general, the largest increase in
the percentage of 220's and larger was
greatest in the years when the size of the
8 The treatments paired are: 4-3, 5-2, 11-18, 9-8, and 10-8. Treatments 4, 5, 9, and 11 received
one pound of K 2 per tree annually, while treatment 10 received one pound of K-0 each year dur-
ing 1928-39 and 3 pounds annually in subsequent years during 1940-49. Treatments 2, 3, 4, and 5
are without covercrops.
9 Since the absorption of one element was affected by that of another element, simple correla-
tions between fruit size and the concentration of other elements in the leaves were also found.
However, the application of partial regression statistics showed that the potassium correlation is
the only one which is firmly related to fruit size.
fruit of the trees which did not receive
potash fertilization was small, and least
when the fruit of those trees was large.
It thus appears that the responses to pot-
ash were influenced by seasonal effects,
probably of a climatic nature, as well as
by a gradual depletion of the potash sup-
ply in the soil.
It was also found that potash fertiliza-
tion did not affect the number of fruits
which were harvested from these plots.
It is, therefore, evident that the small in-
crease in the yield of the treated trees
which is indicated in table 5 is due to the
larger fruits which they produced. In this
respect the results of the present experi-
ment are similar to those of Benton and
Stokes ( 1931 ) with oranges in New South
Wales.
The relationship between potash and
fruit size in these experiments is also
indicated by correlations which have been
obtained between the percentage of 220's
and larger and the potassium content of
the leaves (Parker and Jones, 1950a, b) .
Spring cycle leaves were harvested in De-
cember, 1948, from trees in all replicates
of 30 treatments. These treatments varied
in respect to the use of several inorganic
and organic fertilizers, soil amendments,
and covercrops. Nitrogen had been ap-
plied in all cases, and trees were reason-
ably normal since treatments with badly
deteriorated soil were avoided. A sizable
positive correlation was observed between
the concentration of potassium in the
leaves and the size of the fruit.
Figure 6 shows the relationship between
the percentages of size 220 and larger for
the crops of 1941-49 and leaf potassium.
These two variables may be seen to have
increased simultaneously until the potas-
[36]
A
i — r
80
70
60 —
z
u
o
DC
UJ
a.
z 5C
<
UJ
2
10
i r
i r
R = + 8801
O INORGANIC FERTILIZERS
• BULKY ORGANIC FERTILIZERS
OL/
0.7 0.8 0.9 1.0 I.I 1.2 1.3 1.4 1.5 1.6
PER CENT POTASSIUM IN DRY WEIGHT OF LEAVES
FIG. 6. Relation between the proportion of large-sized oranges in the crops harvested during
1941-49 and the concentration of potassium in the leaves in December, 1948, for 30 treatments
in the experimental orchard. The quantities of potassium applied were not equal in all treatments.
sium in the leaves reached approximately
1.3 per cent of their dry weight, and at
about that point the curve levels out. Al-
though similar curves for the crops of in-
dividual years vary in slope and region of
leveling, it appears that this leaf content
may represent an average concentration—
in December samples of spring-cycle
leaves— which may be critical for fruit
sizes in navel oranges. Extensive studies
of the seasonal trends of leaf composition
in this orchard (Jones and Parker, 1950)
indicate that it would be about 1.5 per
cent of the dry weight if leaves of the same
cycle were taken in August or September.
Although the nature of the rootstocks
(Haas, 1948a; Smith, Reuther, and
Specht, 1949; Chapman and Brown,
1950) , the age, and period of sampling of
the leaves complicate comparisons of leaf
analysis data of various investigators, it
is interesting to note that the potassium
content of the leaves of the trees in this
experiment, as given in figure 5, is greater
[37
than that which has been found to accom-
pany reduced yields or visible symptoms
of potassium deficiency (Arnot, 1946;
Chapman, Brown, and Rayner, 1947;
Chapman and Brown, 1950; Haas,
19486). However, two investigators with
oranges (Bahrt and Roy, 1940) and an-
other with grapefruit (Innes, 1946) re-
port that potassium applications resulted
in increases in both yield and fruit sizes
under conditions such that the potassium
content of the leaves (age not specified)
of untreated trees was within the range
shown in figure 5. However, these inves-
tigators did not indicate the effects of
increased size of fruits upon the yields.
No visible symptoms of potassium defi-
ciency (Chapman, Brown, and Rayner,
1947; Camp, Chapman, and Parker,
1949) have been recognized in the foliage
and twigs of the trees of the present ex-
periment and, as has been stated, potash
fertilization did not increase the numbers
of harvested fruits. The effects of potash
fertilization upon the size, and internal
quality, of the fruit are apparently cor-
related with the absorption of potassium
in amounts greater than that which limits
fruit setting, premature dropping, or
readily recognized visible symptoms of
deficiency (other than fruit of small size) .
Small sizes of oranges and grapefruit
have been recognized as a symptom of
potash deficiency in sand and solution
cultures (Chapman, Brown, and Rayner,
1947; Haas, 19486). This has also been
the case in field trials with oranges in
Florida (Bahrt and Roy, 1940; Sites,
1948), in Japan (Takahashi, 1931), and
in New South Wales (Arnot, 1946), as
well as with grapefruit in Jamaica (Innes,
1946) and in Florida (Sites and Bowers,
1948). The present trials indicate that
they are a very sensitive index of the
deficiency of this element.
Only a part of the variation in the size
of fruit in this orchard can, however, be
accounted for by correlations with the
concentrations of potassium in the leaves
of the December, 1948, samples. This is
indicated by the annual variations in the
size of fruit of all treatments, and, in any
one year or periods of years, by the devia-
tions from the correlation line of the
mean size of fruit of individual fertilizer
treatments (see figure 6, page 37) .
It is notable that the use of bulky or-
ganic materials has resulted in large in-
creases in the potassium in the leaves.
No estimate of the relative efficiency of
such materials as sources of potassium,
in comparison with sulfate of potash or
other salts, should be drawn directly from
the data of figure 6, since unequal quan-
tities of this element were applied. How-
ever, it may be stated that the leaves of
the treatment (No. 42), which has re-
ceived manure supplying 1.5 pounds of
nitrogen per tree each year since the start
of the experiment, contained 1.22 per cent
of potassium in the dry weight of Decem-
ber samples. From information presented
in table 8 (page 43), it is evident that
supplementing this treatment with sulfate
of potash (tr. 14) did not increase fruit
size, although leaf potassium was in-
creased to 1.59 per cent.
D. USE OF BULKY ORGANIC MATERIALS OTHER THAN MANURE
Various bulky organic materials other
than manure are frequently applied in
citrus orchards in California. Lima bean
straw, alfalfa hay, and cereal straw are
among those commonly used. The use of
such materials is influenced by their costs,
the presence or absence of weed seeds,
convenience of application, and their
chemical analyses. The hays and straws
contain a higher percentage of organic
matter than manures, but their nitrogen
content varies considerably.
In the present trials, the sources of
organic matter named above have been
applied to certain treatments on the basis
of their content of either nitrogen or or-
[38]
ganic matter. In all cases they have been
used in conjunction with urea as a sup-
plemental source of nitrogen in such a
way that the total amount of nitrogen ap-
plied in the fertilizer programs was equal
in contrasting treatments in the same pe-
riods of time. Data concerning these
treatments are presented only for the
1928-1939 crops since changes in this
series of treatments were made at the
conclusion of this 12-year period, the
results of which have not yet materialized.
Effects on relative yields
During 1928-39, the total nitrogen ap-
plication in this series of treatments was
only one pound per tree per year, and
partial nitrogen starvation existed in the
later years of the period. One-half of the
nitrogen was supplied by the bulky or-
ganic materials in the following treat-
ments: manure (No. C) , alfalfa hay (No.
36), and lima bean straw (No. 37). The
other half of the nitrogen was supplied
by urea. The quantity of actual organic
matter applied in these 3 treatments there-
fore varied in accordance with the analy-
sis of the organic material. In one cereal
straw treatment (No. 39) the straw was
applied to supply the same amount of
organic matter as in the manure-urea
treatment (No. C), while in a second
cereal straw treatment the quantity of or-
ganic matter was three times that amount.
The actual quantities of organic matter,
of nitrogen, and of other nutrients ap-
plied from all materials in the fertilizer
program are shown in table 7.
Comparisons of the yields of these and
other treatments during 1928-39 sug-
gested (Parker and Batchelor, 1942) that
all of these bulky organic materials are
of equal value as sources of organic mat-
ter and nitrogen. This is provided they
are applied with regard to their analysis
of organic carbon and nitrogen, and are
19 Note that the nitrogen in both the organic matter and in the supplementary concentrate is in-
cluded in this computation. The content of organic carbon is computed by multiplying the per-
centage of organic matter of fertilizer materials by 0.58.
supplemented by additions of concen-
trated nitrogen fertilizer in such a way
that the ratio of organic carbon to nitro-
gen in the fertilizer program is not exces-
sive. The evidence indicated that fertilizer
programs in which the organic carbon:
nitrogen ratios 10 were greatly in excess
of 10:1 frequently resulted in lower yields
caused by decreased availability of ni-
trogen to the trees. This decrease in avail-
ability is a result of the failure of the
added organic matter to supply sufficient
nitrogen for the microorganisms which
decompose it (Batchelor, 1933, 1948a).
The absorption from the soil of the nitro-
gen the organisms require is believed to
explain the depression of yields resulting
from the use of manure alone as shown
in table 3 (page 21).
The yields resulting from the use of
the various hays and straws with urea
are given in table 7 relative to the yields
of the manure-urea treatment (No. C).
It will be noted that most of the programs
resulted in about equal yields; only the
yields of the treatment consisting of
heavy applications of cereal straw (No.
38) were relatively low. The carbon:
nitrogen ratio of the fertilizer applied
in this treatment was 34: 1, an excessively
high ratio. The C:N ratios of the other
treatments are nearly 10:1, but that of
the lima bean straw treatment is 19:1.
The latter treatment resulted in normal
yield, and illustrates the approximate
nature of this ratio, especially where dif-
ferent classes of bulky organic materials
are concerned. Emphasis should also be
placed upon the fact that this ratio is
most critical when fertilizer applications
are made to soil having a low nitrogen
supply. Organics with a wide carbon:
nitrogen ratio may be more safely ap-
plied in large quantities to soils with large
reserves of readily available nitrogen
than to those with only small reserves.
[39]
3 O
U. Q.
5 N
>■ r
c £
o
o .>
St
o a.
< o
*?
*.
.a o
3£
a c
I s
Si
c E
•■ o
1 "5
£
«n +■
***
>
T3
C w
CO CD
>>2«
2*1
00 ■<* t-; 00
"^1
§*>g
■ oo id oi in
t>
co 2 ft
CO 00 00 00
oo
i-s
w .
2 »~
S£?g
^ 00 W N
C5
■
, 00 CO 00
o
e3
bo
r-4 rtf CO Ifi
tH
o
i-i d d d
tH
c ®
. rt CD
03 is
Is
*E &
«a .
o CO
CD
0.3
n n t- h
rH
CO iH © iH
(N
oPh
J3 W
d d d d
d
Pi
Ph
G
3
CD
a
d
CD
in
-^
O
m m th io
-^
ft
d d d d
d
s
2
*« CD
co
T3
1 CD
>
d J-+3
C > S d
|
1 Sg'3'2
. 3 a © .§
1 §
"° CO t>
1 CO CO
o -p -a
°-g §
■^ E «
o S
-D «J «
I*?
a £
2-Sio
--= £
.5 *-
lu- E
o -o
£ c 5
Ol tt) "o
-C o w
0^
CO ■>- —
u ||
= -ll
S o Q-
B 0> CD
g o I
*= *o ?
o c a
10 C u
u 3
a. o- >
c D 5
to c
* £
a c
2 "S 5
-^ a o
"5 2-m,
a 2 5
< E
a
II
c
o
A
o
J a) cf- £
E *- O CD I) o
1 2« p i g°
2 — o *■ a p c
a c g-o c £ g
J) — .5 CD O O U
C _• X H CD