Citation
Fertilizer experiments with potatoes on the marl soils of Dade County

Material Information

Title:
Fertilizer experiments with potatoes on the marl soils of Dade County
Series Title:
Bulletin - University of Florida Agricultural Experiment Station ; 352
Creator:
Fifield, W. M.
Wolfe, H. S.
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida Agricultural Experiment Station
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
City of Gainesville ( flego )
Fertilizers ( jstor )
Manganese ( jstor )
Sulfates ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services (UFDC@uflib.ufl.edu) with any additional information they can provide.

Downloads

This item has the following downloads:


Full Text


Bulletn 352December, 1940


UNIVERSITY OF FLORIDA
AGRICULTURAL EXPERIMENT STATION

WILMON NEWELL, Director GAINESVILLE, FLORIDA



Fertilizer Experiments With Potatoes

On The~ Marl SJoilf Dade County

By W. M. FIFIELD and H. S. WOLFE


Fig. 1.-Washing and grading potatoes at the Sub-Tropical Experiment Stationi.

Single copies free to Florida residents upon request to
AGRICULTURAL EXPERIMENT STATION GAINESVILLE, FLORIDA


Bulletin 352




EXECUTIVE STAFF BOARD OF CONTROL I


John J. Tigert, M. A., LL.D., President of the University3 Wilmon Newell, D.Sc., Director3 Harold Mowry, M. S. A., Asst. Dir., Research
J. Francis Cooper, M.S.A., Editor3 Jefferson Thomas, Assistant Editor3 Clyde Beale, A.B.J., Assistant Editor Ida Keeling Cresap, Librarian Ruby Newhall, Administrative Manager K. H. Graham, Business Manager3 Rachel McQuarrie, Accountant3
MAIN STATION, GAINESVILLE
AGRONOMY
W. E. Stokes, M.S., Agronomist' W. A. Leukel, Ph.D., Agronomists Fred. H Hull, Ph.D., Agronomist G. E. Ritchey, M.S., Associate2 W. A. Carver, PH.D., Associate John P. Camp, M.S., Assistant Roy E. Blaser, M.S., Assistant Fred A. Clark, B.S.A., Assistant
ANIMAL INDUSTRY
A. L. Shealy, D.V.M., Animai Industrialist:'
R. B. Becker, Ph.D., Dairy Husbandman' E. L. Fouts, Ph.D. Dairy Technologist8 W. M. Neal, Ph.D. Asso. in An. Nutrition D. A. Sanders, D.V.M., Veterinarian M. W. Emmel, D.V.M., Veterinarian3 N. R. Mehrhof, M.Agr., Poultry Husbandman'
W. G. Kirk, Ph.D., Asso. An. Husbandman-'
D. J. Smith, B.S.A., Asst. An. Hush. P. T. Dix Arnold, M.S.A., Asst. Dairy Husbandman3
L. Rusoff, Ph. D., Asst. in An. Nutrition
0. W. Anderson, M.S., Asst. Poultry Husbandman'
L. E. Mull, M.S., Asst. in Dairy Tech.
SOILS
R. V. Allison, Ph.D., Chemist 3 Gaylord M. Volk, M.S., Chemist F. B. Smith, Ph.D,. Microbiologists C. E. Bell, Ph.D., Associate Chemist H. W. Winsor, B.S.A., Assistant Chemist J. Russell Henderson, M.S.A., Associates L. H. Rogers, M.S., Asso. Biochemist Richard A. Carrigan, B.S., Asst. Chemist
ECONOMICS, AGRICULTURAL
C. V. Noble, Ph.D., Agricultural Economist' 2
Zach Savage, M.S.A., Associate A. H. Spurlock, M.S.A., Associate
ECONOMICS, HOME
Ouida D. Abbott, Ph.D., Home Economist'
Ruth Overstreet, R.N., Assistant R. B. French, Ph.D., Asso. Chemist
ENTOMOLOGY
J. R. Watson, A.M., Entomologist' A. N. Tissot, Ph.D., Associate H. E. Bratley, M.S.A., Assistant
HORTICULTURE
G. H. Blackmon, M.S.A., Horticulturist' A. L. Stahl, Ph.D., Associate F. S. Jamison, Ph. D., Truck Hort.3 R. J. Wilmot, M.S.A., Fumigation Specialist
R. D. Dickey, M.S.A., Asst. Horticulturist J. Carlton Cain, B.S.A., Assistant Horticulturist
Victor F. Nettles, M.S.A., Assistant Horticulturist
F. S. Lagasse, Ph.D., Horticulturist2 H. M. Sell, Ph.D., Asso. Horticulturist2
PLANT PATHOLOGY
W. B. Tisdale, Ph.D,, Plant Pathologist' 3 George F. Weber, Ph.D., Plant Path.3 L. 0. Gratz, Ph.D., Plant Pathologist Erdman West, M.S., Mycologist Lillian E. Arnold, M.S., Asst. Botanist


H. P. Adair, Chairman, Jacksonville W. M. Palmer, Ocala R. H. Gore, Fort Lauderdale N. B. Jordan, Quincy T. T. Scott, Live Oak J. T. Diamond, Secretary, Tallahassee
BRANCH STATIONS
NORTH FLORIDA STATION, QUINCY J. D. Warner, M.S., Agron. Acting in Charge
R. R. Kinkaid, Ph.D., Asso. Plant Path. Elliott Whitehurst, B.S.A., Assistant An. Husbandman
Jesse Reeves, Asst. Agron., Tobacco
CITRUS STATION, LAKE ALFRED A. F. Camp, Ph.D., Horticulturist in Charge.
John H. Jefferies, Asst. in Cit. Breeding Michael Peech, Ph.D., Soils Chemist B. R. Fudge, Ph.D., Associate Chemist W. L. Thompson, B.S., Associate Entomologist
F. F. Cowart, Ph.D., Asso. Horticulturist W. W. Lawless, B. S., Asst. Horticulturist R. K. Voorhees, M.S., Asst. Plant Path.
EVERGLADES STA., BELLE GLADE J. R. Neller, Ph.D., Biochemist in Charge
J. W. Wilson, Sc.D., Entomologist F. D. Stevens, B.S., Sugarcane Agron. Thomas Bregger, Ph.D., Sugarcane Physiologist
Frederick Boyd, Ph.D., Asst. Agronomist G. R. Townsend, Ph.D,, Plant Pathologisi E. W. Kidder, M.S., Asst. An. Husbandman W. T. Forsee, Ph.D., Asso. Chemist B S. Clayton, B.S.C.E., Drainage Engineer2
F. S. Andrews, Ph.D., Asso. Truck Hort.
SUB-TROPICAL STA., HOMESTEAD
W. M. Fifield, M.S., Horticulturist Acting in Charge
S. J. Lynch, B.S.A., Asst. Horticulturist Geo. D. Ruehle, Ph.D., Associate Plant Pathologist
W. CENTRAL FLA. STA.,
BROOKSVILLE
WV F. Ward, M.S. Asst. An. Husbandman in Charge
FIELD STATIONS
Leesburg
M. N. Walker, Ph.D., Plant Pathologist in Charge
K. W. Loucks, M.S., Assistant Plant Pathologist
Plant City
A. N. Brooks, Ph.D., Plant Pathologist
Hastings
A. H. Eddins, Ph.D., Plant Pathologist. E. N. McCubbin, Ph.D., Asso. Truck Horticulturist
Monticello
Samuel 0. Hill, B.S., Asst. Entomologist'
Bradenton
Jos. R. Beckenbach, Ph.D., Truck Horticulturist in Charge David G. Kelbert, Asst, Plant Pathologist
Sanford
R. W. Ruprecht, Ph.D., Chemist in Charge, Celery Investigations W. B. Shippy, Ph.D., Asso. Plant Path,
Lakeland
E. S. Ellison, Meteorologists B. H. Moore, A.B., Asst. Meteorologist2
'Head of Department 21n cooperation with U.S.D.A. 3Cooperative, other divisions, U. of F.







FERTILIZER EXPERIMENTS WITH POTATOES ON THE MARL SOILS OF DADE COUNTY
By W. '\,I. FIFIELD and H. S. WOLFE*

CONTENTS
Page Page
Plan of Experilnents - -- -- - -- - -- ------- --- 4 Sources of Nitrogen 17
Fertilizer Analyses --------------- -------- ------- --- 7 1\1uriate vs. Sulfate of Potash 25
Amounts of Fertilizer per Acre. 11 Manganese 25
Percentage of Nitrogen froin Other Soil Amendments 29
Organic Sources -- --- --- - -------- --- -- ---- 14 Sunnuary 38

INTRODUCTION
The marl soils of southern Dade County which are used for potato production are of a peculiar calcareous formation quite unlike those of any other soils in the United States. They are alkaline, ranging from a pH of about 7.5 to about 8.5. Mechanical analysis would classify this soil as a uniform silt loam, but chemical analysis shows over 90 percent of calcium carbonate. Sand is present only in very small percentage, and organic matter constitutes about 5 percent. This marl, a sedimentary deposit, varies throughout the area from a depth of a few inches to about four feet and is underlaid with a porous limerick. In limited areas muck is found with the marl, particularly in the numerous small sink or pot hole formations and other low places.
The marl is interspersed throughout with remains of tiny shells, evidently similar to those from which it originally was derived. It is relatively low in natural fertility and in organic matter, but soil moisture, rising through capillarity from the water table near the surface during the winter season, ordinarily is sufficient for a good crop.
Some farms in the area are located on the higher, fairly well drained soils which are seldom flooded in summer, and others on lower-lying fields subject to flooding during part of the summer rainy season. Extensive drainage by means of canals and ditches in recent years has reduced the danger from sudden floods but, at the same time, occasionally has contributed to a lack of soil moisture in times of drouth. At present the water table ordinarily is found from two to four feet below the surface during the major part of the growing season, which in this section extends from November to March.
Rotation of crops is not practiced generally in growing po*Formerly Horticulturist in Charge, Sub-Tropical Experiment Station.





















Fig. 2-Summer growth of weeds on a South Dade farm. Note growth in relation to height of tractor.
states in Dade County, where the same field usually is planted to potatoes every year. However, the fields used for potatoes during the winter months produce a heavy green manure crop of weeds (Fig. 2) or of a legume such as velvet beans (Fig. 3) or sesbania during the spring and summer. This is plowed under in the fall.
PLAN OF EXPERIMENTS
The experiments described in this bulletin mostly were carried out on the East Glade farm of the Sub-Tropical Experiment Station, located about six miles east of Homestead on the north side of the Homestead or North Canal. The soil on this farm is typical of the better potato soils of the area, being fairly high, well drained and about three feet deep. Other experiments described were performed in other sections of the adjoining area, known as South Allapattah Gardens, where in the 1938 season about 7,500 acres of potatoes were planted.
Practically all of the potatoes now planted in Dade County are of the Bliss Triumph variety, and all of these experiments were performed with it. In general throughout the tests seed pieces were 11/2 ounces in weight and were spaced 9 inches apart in the rows. The earlier work was with 36-inch rows, and the later with 38-inch rows. This size and spacing required from 28 to 30 bushels of seed per acre.
When the Station was first established in 1930 its equipment, staff and facilities were exceedingly limited and remained so for a number of years. Accordingly, although nearly all of the growers were using machine planters and fertilizer distributors, the early Station experiments had to be fertilized and planted





Fertilizer Experinients with Potatoes


Fig. 3-Summer growth of velvet beans on a potato farm at South Allapattah Gardens. Note high water in the ditch. by hand. Later, as equipment was obtained, a change to machine methods was accomplished.
In planting and fertilizing by hand, furrows of about three inches in depth were opened with a garden tractor, after the field had been plowed and disked. The fertilizer was carefully weighed out and distributed in the furrow uniformly. The furrows were then closed and in about two days reopened for planting, thereby mixing the fertilizer with the soil. Planting was done by hand with the aid of marking chains to insure proper seed piece spacing. Four-row plots, each 1/72 acre, were planted by this method. At harvest time the outside rows of each plot and 5 feet from each end of the two center rows were discarded to eliminate any possible competitive effect between treatments. The comparison of treatments has been based on the yields obtained from the remainder of the plots (1/180 acre) after the discarded portions were eliminated.
Later, when machine planting was adopted, fertilizer was distributed in two continuous narrow bands, approximately at the same level as the seed piece and two inches from it on either side. The distributor was adjusted as accurately as possible for






Florida Agricultural Experiment Station


each mixture. A single-row, assisted-feed planter was used. Plots harvested for data with this method each consisted of a single row, 75 feet long (1/184 acre). In all cases border plots were planted around the field, and except where otherwise indicated buffer rows were planted between treatment plots. Furthermore, about 80 to 85 feet of row were actually planted, the ends being trimmed off at harvest to eliminate errors due to starting and stopping the planter. Variations from these methods occurred in some of the cooperative tests. These exceptions are described under headings of the tests involved.
All of the fertilizer mixtures used on the Station farm were hand mixed. In order to provide a uniform basis of interpretation, all of the test mixtures described in this publication have been converted to a nitrogen basis, according to the 1935 Florida fertilizer law, but early work as reported in the Station's Annual Reports was done with formulas in which the nitrogen was expressed as ammonia.
Replicate plots of all treatments were used, varying in number in different tests, but always arranged in a randomized manner. Time of planting, cultivation and spraying for foliage diseases and insects were uniform throughout each test. The vines were allowed to die before the tubers were dug.
All results are reported in terms of yield of No. 1 tubers (including both A and B sizes) per acre. It was found that the yield of No. 1 tubers, which always ran 85 to 95 percent of the total yield, closely paralleled total yields, and, accordingly, relative results were the same. Grading was done in later years at the Station farm (Fig. 1) with a standard hand grader equipped with official-sized grading chains. Formerly it was done in the packinghouse of the South Florida Potato Growers' Association.
Throughout this bulletin the term "significant" will be used often in discussing differences between yields. Where two average yields differ only slightly, it is conceivable that the differences could be due to other factors than the two fertilizers being tested. Statisticians have worked out methods for use with prescribed plot technique which enable the research worker to determine fairly well whether the difference in average yields obtained is due to the difference in fertilizers being tested, or to other variables. If the differences can be attributed to the fertilizers they are said to be "significant." If the differences apparently were caused by variables other than the fertilizers they would not be significant.





Fertilizer Experiments with Potatoes


In this bulletin statistical significance was determined by a
mean difference in yield being at least 2.2 times as large as its standard error. In a few tests, "Student's" methods were used.
The methods used are all described by Love'. The statistical computations and constants are not shown in the tables except where publication of the odds by "Student's" method is helpful
for interpretation.
FERTILIZER ANALYSES
The most common potato fertilizer analysis used in Dade County at the time the Station was established (1930) was a 48-52, with about 50 percent of the nitrogen derived from organic sources. Experiments soon were begun using the 4-8-5 as a base or control, to determine if a different analysis would give better results.
The first experiments, begun in 1930 on the Station farm, consisted of doubling and halving the percentages of N, P and K, respectively, from the amounts in the common 4-8-5 analysis. All fertilizers had 50 percent of the N derived from organic sources, each contained manganese, and all sources of ingredients were the same for each formula. Each mixture was applied at the rate of 1,500 pounds per acre by hand and was tested in quadruplicate plots, except in 1930-31 when duplicate plots only were used. The tests were conducted on a field in which potatoes had been grown previously only one, the preceding (1929-30), season.
As the work progressed the 4-4-5 treatment was discarded because certain accidental variables occurred which rendered the data valueless.
The average yields for all replications of each treatment (except 4-4-5) are recorded in Table 1.
TABLE 1-AVER-AGE YIELDS oF No. 1 TUBERS (BUSHELS PER AcRF) FROM THE
1930-33 FERTILIZER ANALYSIS TESTS.
Treatment 1930-31 1931-32 1932-33 Average
4-8 -5 140 170 200 170
4-16-5 142 177 218 180
4-8 -V1,, 138 166 195 167
4-8 -10 136 193 209 180
2-8 -5 138 172 201 170
8-8 -5 115 150 186 150
2 41, nitrogen (N), 81 , phosphoric acid (P), 5% potash (K).

1 Love, H. H. Application of statistical methods to agricultural research. The Commercial Press, Ltd. 1938.





8 Florida Agricultural Experimient Station

In each of the three seasons the high nitrogen plot (8-8-5) yielded significantly lower than any of the other treatments, indicating that the additional nitrogen was not only a needless expense but actually harmful. The differences in yield between the other treatments were not very great, but the 3-year average indicated a slightly higher yield for the high phosphate (4-16-5) and high potash (4-8-10) treatments. On the strength of this. and the fact that the low nitrogen treatment (2-8-5) yielded as well as the standard treatment (4-8-5), a new analysis, 3-12-8, was selected as embodying some of the trends indicated by the general results of the test. The cost and source of ingredients of this 3-12-8 were comparable to those of the 4-8-5.
The new 3-12-8 analysis then was tested with the 4-8-5 in succeeding years. The first of these tests was conducted on the Station farm in 1934-35. Each analysis was tested at the rate of 1,000, 1,500 and 2,000 pounds per acre, respectively, applied by hand. The yields from four replications of each treatment were averaged, with the results shown in Table 2, Series 1934-35.


TABLE 2.--AVERAGE YIELDS 01, No. I TUERS (BUSHELS PER NERt) 15GM~ TIHF
1934-36 FERTILIZER ANALYSIS TESTS.

Series Analysis Rate per Acre (lbs.) Yield per Acre
4-8 -5 1,000 167
3-12-8 1,000 171
1934-35 4-8 -5 1,500 199
3-12-8 1,500 186
4-8 -5 2,000 194
3-12-8 2,000 197

Estes 3-12-8 2,000 248
4-8 -5 2,000 237

3-12-8 1,500 279
1935-36 4-8 -5 1,500 278
_1--2-8 2,000 289
4-8 -5 2,000 287


Analysis of the data showed that there was no significant difference in yield between the two analyses, even though at the 1,500-lb. rate the 4-8-5 showed a little higher average yield than the 3-12-8.

Another test was conducted the same year on the commercial farm of J. H1. Estes. About an acre was planted (by machine)






Fcti/in r Experiiments -weit/i Potatoes


with each analysis at the rate of 2,000 pounds per acre, and from each acre 20 representative plots were harvested at random. The yields from the 20 plots of each treatment were averaged with the results shown in Table 2. Series Estes.
In this test a slight but not significant increase in average yield was noted for the 3-12-8 treatment.
Another test was made the following year (1935-36) on the Station farm. The 3-12-8 and 4-8-5 analysis were tested in quadruplicate plots at rates of 1,500 and 2,000 pounds per acre, applied by machine. The average yields are given in Table 2, Series 1935-36. Again, no significant difference in yield was obtained between the two analyses.
Since it was apparent that no appreciable differences in yield were being obtained from analyses varying as widely as a 4-8-5 and a 3-12-8, it was thought wise to test differences in analysis on a more intensive scale. Accordingly, a new series of tests was instituted on the Station farm in 1937-38, in which the percentages of phosphoric acid and potash in a basic 4-8-4 formula were increased progressively, alone and in combination. The nitrogen was not varied, but kept constant throughout at 4 percent, of which half was derived from organic sources.
Data were secured from six replications of each of the 12 treatments, planted by machine. All fertilizer was applied at the rate of 1,500 pounds per acre, and all of it contained manganese. The average yield from each treatment is given in Table 3, for each year.

TABLE S.-AvLRAGE. YIELD O-F No. 1 TU BERS (BUSHELS PLR ACRE) FROM TIE TREATAML\TS Co-MPOSLNG THE FERTILIZER A'\ALN-si1, TESTS, 1937-39.

Analysis 1937-38 1938-39
4-8-4 330 174
4-8-6 331 169
4-8-8 334 176
4-8-10 333 171
4-12-4 342 172
4_12-6 327 179
4-12-8 339 182
4-12-10 331 168
4-16-4 329 169
4-16-6 323 174
4-16-8 336 172
4-16-10 341 179

The yields were extremely high the first year of this test and fairly low the second (a dry) year. Each year, however, a surprising similarity in yield was obtained from all 12 treatments,























'J,
'17




Fig. 4-Portion of the 1937-38 potato fertilizer analysis tests. Note uniform vine growth among all plots.

and the small differences in yield between them were found not to be significant. In other words, no significant increase in yield was obtained from increasing the phosphoric acid beyond 8 percent, or from increasing the potash content beyond 4 percent. No difference was noted in vine growth or quality of tubers between treatments either year (Fig. 4).
It should be noted that this last experiment was conducted on land which had been fertilized and planted in potatoes for five successive preceding years. Since the authors have noted indications that some fertilizer residue tends to carry over from one year to another in these marl soils, there is a possibility that sufficient residual phosphate and potash were present to offset additional quantities added in the higher analyses. Since the lowest analysis tested (4-8-4) yielded as well as the highest (4-16-10), it would seem desirable to test even cheaper formulas, especially in view of the good results from the 2-8-5 and 4-8-21/2 analyses as shown in Table 1, on land with no previous history of potato culture.
These analysis tests, conducted over a period of years, indicate that there seems to be no justification for using analyses,






A r'tilizc r E.~l'p riclts uwit/i Pot atoes


to be applied at rates of about 1,500 to 2,000 pounds per acre, with more than 3 or 4 percent nitrogen, 8 percent phosphoric acid and 4 or 5 percent potash.

AMOUNTS OF FERTILIZER PER ACRE
The rate at which fertilizer should be applied per acre is, of course, influenced greatly by the analysis of the fertilizer. At the time these experiments were begun a 4-8-5 analysis was the one most commonly in use. Therefore, it was used extensively for the basic tests. A few other analyses were used from time to time. The tests were distributed over a number of years and over a number of different farms, in order to get as comprehensive data as possible.
The first test was performed on the Station farm in 1933-34. The fertilizer was distributed in the furrows by hand and mixed with the soil two days before the seed pieces were planted. Four treatments were used, 500, 1,000, 2,000 and 3,000 pounds per acre, respectively, of a 4-8-5 analysis, in which two-thirds of the nitrogen was from organic sources. Four replications of each treatment were planted. The average yields from each treatment are shown in Table 4, Series A.
The 2,000-pound rate yielded higher than any of the other treatments. It yielded enough more than the 1,000-pound rate to justify the additional cost. The reduced yield of the 3,000-pound rate was probably due to injury of the potato sprouts, as it was noted that these plots showed slower emergence than did the others. This injury might not have occurred had the fertilizer been distributed by the machine band method, or had it been applied in the furrows earlier before planting.
Another test was conducted on the Station farm the same year in connection with another experiment. A 4-8-5 was also used in this test, but in one series 33 percent of its nitrogen was derived from organic sources, and in the other 66 percent was so derived. Each mixture was applied by hand at both 1,000pound and 2,000-pound per acre rates. The plot arrangement was similar to that of the previous test. The replicate yields were averaged with the results given in Table 4, Series B.
With both types of mixture, the 2,000-pound rate significantly outyielded the 1,000-pound rate. A comparison of the 33 percent and 66 percent mixtures will be presented later (Table 7, Series 1933-34).






Florida A4gricultural Experiment Station


In another test on the Station farm in 1933-34 nitrate of soda was compared with sulfate of ammonia as sources of inorganic nitrogen in a complete 4-8-5 analysis, and muriate of potash with sulfate of potash as potash sources in the same analysis. Each treatment was tested at both 1,000 and 2,000 pounds per acre, in triplicated plots, with the results given in Table 4, Series C.

TABLE 4.-AVERAGE YIELDuS or No. I TUBERS (BUSHELS PER AcR.) FOR VARIOUS
RATES OF APPLICATIION FOE 4-S-5 FERTILIZER, 1933-34.
Series Rate per Acre (lbs.) Fertilizer Variable Yields per Acre
500 Standard mix 176
A 1,000 Standard mix 203
2,000 Standard mix 222
3,000 Standard mix 202
1,000 331/, Organic N 184
B 2,000 33% Organic N 197
1,000 66% Organic N 203
2,000 66% Organic N 222
1,000 Inorg. N as Nitrate of Soda 165
2,000 Inorg. N as Nitrate of Soda 164
1,000 Inorg. N as Sulfate of Ammonia 181
C 2,000 Jnorg. N as Sulfate of Ammonia 197
1,000 Potash all as sulfate 204
2,000 Potash all as sulfate 222
1,000 Potash all as muriate 207
2,000 Potash all as muriate 209

These data are not harmonious. The nitrate of soda plots showed no increase in yield for the higher rate of application. The sulfate of ammonia showed a statistically significant increase. The sulfate of potash plots also showed such an increase for the 2,000 pounds application, but the muriate plots did not. The comparison of these sources of nitrogen and potash as they in themselves affect yields will be discussed later.
The following year (1934-35) the first test was repeated in revised form on the Station farm. The 500-pound rate was eliminated and a 1,500-pound rate was added. Also another analysis, 3-12-8, was added, and this was tested at 1,000, 1,500 and 2,000-pound rates. In both mixtures the nitrogen was derived two-thirds from organic sources. Four replications were used and the fertilizer again was applied by hand. The results from Table 2 are given again for convenience in Table 5, Series 1934-35.
The results indicate that with both analyses the average yields of the 1,500-pound rate were about as good as those of the 2,000-pound rate. The higher yield of the 2,000-pound rate






Fertilizer Experiments with Potatoes


of 3-12-8 was found, upon analysis of the replicate plot data, to be not significant. In the 4-8-5 test the 3,000-pound rate outyielded the lower rates, and with both analyses the 1,000-pound rate was outyielded by the higher rates.
This same year (1934-35) another test was carried out, this time on the commercial farm of J. M. Holferty. A 4-8-5 analysis (50% of nitrogen from organic sources) was used, applied by machine. Approximately one acre was planted at each of two rates, 1,300 and 2,000 pounds per acre. At digging time 10 plots, each 100 feet long, were selected at random for harvest from each block. The No. 1 tubers were graded from each replicate plot and averaged, with results given in Table 5, Series Holferty A. In this instance, the 2,000-pound rate significantly outyielded the 1,300-pound rate.
In another field of the same grower, another similar test was performed two weeks later. The results are found in Table 5, Series Holferty B. This time, although the 2,000-pound rate showed a slight increase in average yield over the 1,300-pound rate, the difference was not statistically significant and thus application of the additional 700 pounds of fertilizer was not justified.
Another test was performed on the Station farm in 1935-36, comparing 1,500 and 2,000-pound rates of both a 4-8-5 and a 3-12-8 analysis (each with nitrogen 66% from organic sources). Six replications of each were planted. This time the fertilizer was applied by machine. The data were averaged for the various replications, as shown in Table 5, Series 1935-36. TABLE 5.-AVERAGE Yrnirs or No. 1 TUBERS (BUSHELS PER ACRE) FOR VARIOUS
RATES OF FERTILIZER APPLICATION, 1934-36.
Series Rate per Acre (lbs.) Analysis Yields per Acre
1,000 4-8-5 167
1,500 4-8-5 199
2,000 4-8-5 194
1934-35 3,000 4-8-5 226
1,000 3-12-8 171
1,500 3-12-8 186
2,000 3-12-8 197
Holferty A 1,300 4-8-5 179
2,000 4-8-5 209
Holferty B 1,300 4-8-5 212
2,000 4-8-5 224
1935-36 1.500 4-8-5 278
2,000 4-8-5 287
1,500 3-12-8 279
2,000 3-12-8 289





Florida A gric cultural ExYperimtent Station


Here, with both analyses, the 2,000-pound rate slightly outyielded the lesser rate in average yield but the differences were not great enough for statistical significance. Therefore, the expense of the extra 500 pounds of fertilizer was not justified.
In 1937-38 still another Station farm test was made of the 1,500 and 2,000-pound rates, in connection with testing 4-8-5 mixtures in which 33, 50 and 66 percent, respectively, of the nitrogen was derived from organic sources. The fertilizer was applied by machine. Four replicate plots were planted, and their yields averaged, as given in Table 6.
TABLE 6.-AVERAGE YIELDS or No. I TUBERS (BUSHELS PLR ACE) FRo VARIOUS
RATES OF APPLICATION OF A 4-8-5 FERTILIZER, 1937-38.
Rate per Acre (lbs.) % of N from Organic Sources Yields per Acre
1,500 33 302
2,000 33 305
1,500 50 304
2,000 50 311
1,500 66 289
2,000 66 296

Again in this test a slightly increased average yield was obtained from the heavier fertilizer application but the increase was too small for statistical significance.
It is apparent that no single rate of application will give the same results in all seasons or under all conditions in the same season. The data of these experiments generally agree, however, that 1,000 pounds of 4-8-5 per acre is probably not sufficient for best results, and that 3,000 pounds is probably too much. There is more definite agreement that the correct amount is from 1,500 to 2,000 pounds per acre. In the great majority of instances no significant yield differences were obtained between these two amounts, and certainly, that being the case, the 1,500-pound rate would be the more profitable for regular use. In a season in which higher prices than usual might be expected, or in an unusually wet season, or on soil more moist than that occurring on the Station's East Glade farm, it is possible that the 2,000-pound rate would be more profitable. It should be pointed out here that the rates of application discussed refer only to the generally used analyses of 4-8-5 and 3-12-8, and not to mixtures containing higher or lower concentrations of plant food.
PERCENTAGE OF NITROGEN FROM ORGANIC SOURCES
Since the percentage of nitrogen derived from organic sources markedly affects the expense of mixed fertilizers, ex-







Fertilizer Experiments within Potatoes


periments were begun in the fall of 1933 to study the effect on yields of varying this percentage in a standard 4-8-5 analysis.
The first year's test was limited to a comparison of 33 and 66 percent of the nitrogen so derived. All sources of ingredients of the two fertilizers were similar, and both included manganese. Nitrogen was derived from cottonseed meal, fish scrap and sulfate of ammonia, phosphate from superphosphate, and potash from sulfate of potash. Each treatment was replicated three times and the fertilizer was applied by hand. Each mixture was applied at both 1,000 and 2,000-pound rates per acre. The replicate plot yields were averaged to obtain the data shown in Table 7, Series 1933-34. In this test, the 66 percent treatment outyielded the 33 percent treatment at both rates of application and significantly enough to offset the additional cost of the 66 percent mixture.

TABLE 7. AVERAGE YI1ELDS oi, No. 1 TU BERS (BUSEL '~~PER ACRE), RESUTLTING
FRO-M 'USE OF A4-~S5 FERTILIZER WITH VARYING PERCENTAGES or NITROGEN
DERIVED FROM ORGAxNIC SQL RLLS.


% of N from Organic Sources


33 66 33 66
33 66


Rate per Acre
(lbs.)


1,000 1.000
IWO0 2,000
1,500 1,500


33 1.500
1935-36 50 1.5M0
66 1,500
33 1.500
1936-37 50 1,500
66 1,500
33 1,500
50 1.5) 0
1937-38 66 1.500
33 2,001)
50 2.000
66 2.000


Yields per Acre
184 203
10c7 222
208 199
276 280 278
194 184 182
302
304 289
3n5 311 296


The following year (1934-35) both percentages were again tested but this time only at 1,500 pounds per acre. Other factors, including application of the fertilizer by hand, were the same as in 1933-34. The average yields are given in Table 7, Series 1934-35. These yields are not significantly different, and since the 33 percent mixture was cheaper, it was therefore more profitable.


Series


1933-34 1934-35






Florida J.ricultw'(d Experbiwiit Sh7tioii


The following year (1935-36) the same tests were repeated with similar procedure, except that the number of replications was increased to 10. In addition, a 50 percent treatment was included. The average yields are given in Table 7. Here again the yields were not significantly different, and the 33 percent treatment again was most profitable.
In the 1936-37 season the same experiment was repeated with technique similar to that of the previous year. The average yields are shown in Table 7. Again this year the 33 percent treatment yielded as well as the others and was therefore most profitable.
All treatments were repeated in the 1937-38 season, this time each mixture being applied at both 1,500 and 2,000-pound rates per acre. Also, this year the fertilizer was applied by machine, with eight replications of each plot. Average yields are given in Table 7. Again this year, and at both rates, the 33 percent treatment yielded as well as either of the others, and therefore again was most profitable.
Thus, in four of the five years of testing, the 33 percent mixture yielded as well as the mixtures containing a higher percentage of N derived from organic sources.
A factor considered most important in selecting the more slowly available sources of nitrogen, and the percentage of them used in the fertilizer, is amount of rainfall. Therefore, for comparative purposes, the rainfall data including these five years, as recorded officially at the Sub-Tropical Experiment Station for the respective crop seasons, are given in Table 8.
There seems to be no definite relation of treatment results to rainfall during the seasons in which these tests were conducted. The first season (1933-34) higher yields were received from the 66 percent mixture. That year an unusually heavy rainfall occurred in October, but the potato plots were not planted until December 8, so that there is not much likelihood of the October rain exerting much influence on the crop. In the 1935-36 season the plots were planted December 2. The season from planting to harvesting (March 18) was even wetter than in 1933-34, and yet the 33 percent mixtures yielded as well as those containing higher percentages of N derived from organic sources.
These tests were conducted over seasons which varied considerably in amount of rainfall, so that it seems quite logical to conclude from them that ordinarily, with other factors equal,






Fcrtili er Experiments with Potatoes


mixtures containing no more than 33 percent of their nitrogen from organic sources should give as satisfactory results as those of higher percentages thus derived. Of course, these conclusions are drawn from experiments in which the organic sources of nitrogen were limited to two typical water-insoluble organics, fish scrap and cottonseed meal.
TABLE S.-MONTHLY ANi) TOTAL RAIN-FALL (IN INCHES) RECORDED DURING THE
CROP SEASONS FB()Nf 1932 TO 1940, HOMESTEAD, FLORIDA.*

Total
October November December January February March (6 months) 1931-32 6.95 1.85 0.26 3.21 0.47 0.92 13.66
1932-33 12.90 4.15 0.81 1.39 0.20 3.05 22.50
1933-34 22.95 1.39 0.33 2.00 0.94 2.13 29.74
1934-35 3.22 0.37 0.41 0.23 0.25 0.53 5.01
1935-36 7.41 5.51 0.69 2.50 4.59 4.83 25.53
1936-37 3.37 3.52 0.89 0.65 2.70 4.41 15.54
1937-38 7.41 0.37 0.55 2.69 0.81 1.72 13.55
1938-39 4.41 2.44 0.71 0.93 0.56 0.09 9.14
1939-40 16.79 0.95 1.80 1.02 2.44 2.93 25.93

None of the potato tests described in this bulletin was planted before
November 10 in any year.
Additional experiments were conducted on sources of nitrogen, in which mixtures containing all water-soluble nitrogen were compared with those containing 50 percent of their N from numerous organic sources. These are described below.
SOURCES OF NITROGEN
These investigations were begun in the fall of 1933, when a comparative study of sulfate of ammonia and nitrate of soda, as inorganic sources of nitrogen in complete fertilizer, was started. The first series of trials was conducted over a period of four years. A 4-8-5 fertilizer was used throughout. The first three years it was applied in the furrow by hand and the last year by machine. The first year it was applied at the rate of 2,000 pounds per acre. In each of the next three years it was applied at 1,500 pounds per acre.
The nitrogen of the fertilizer was derived 25 percent from fish scrap, 25 percent from cottonseed meal, and 50 percent from either sulfate of ammonia or nitrate of soda, according to the mixture tested. In other words, 50 percent of the nitrogen was derived completely from one or the other of these two inorganic sources.
The first year (1933-34) three replications of each treatment were planted. The second (1934-35) year, four, the third








,(1935-36) year, 10, and the fourth (1937-38) year, six replications
-were used. The tests were not planted in 1936-37. The replicate plot yields for each treatment were averaged each year and the average annual yields are given in Table 9.
TABLE 9.-AVERAGE YIELDS OF No. 1 TUBERS (BUSHELS PER ACRE) FROM FERTILIZER
TREATMENTS COMPARING INORGANIC SOURCES OF NITROGEN IN COMPLETE
FERTILIZERS.

Source 1933-34 1934-35 1935-36 1937-38 Average
Nitrate of soda 164 191 284 305 236
Sulfate of ammonia 197 199 278 292 241

There was no appreciable difference between the two treatments except in the first year, when sulfate of ammonia yielded better than nitrate of soda. Just why this difference should have occurred the first year is not explainable on a rainfall basis, or on any other readily apparent basis. At any rate, since the four-year average indicates no significant difference in yield between the two, and since the first year's data were in favor of sulfate of ammonia, it is logical to conclude that this is probably the better source for these soils, especially since its cost per unit of nitrogen was lower than that of nitrate of soda.
A more comprehensive test of nitrogen sources was begun in the fall of 1937. A number of water-soluble and water-insoluble sources were tested in a 4-8-5 complete fertilizer, with manganese added. The source of phosphoric acid (superphosphate) and potash (sulfate of potash) remained the same for all treatments, except for allowances made for the small quantities of these elements carried by some of the organics.
Although most ordinary commercial potato fertilizers contain a number of organic sources of nitrogen, it was decided arbitrarily in this experiment to test the sources singly, deriving half the nitrogen from sulfate of ammonia and half from the organic source being tested. The materials thus tested were cottonseed meal, fish scrap, blood-and-bone tankage, dried blood, milorganite and urea. In addition, a cyanamid test was included, but since 50 percent of cyanamid nitrogen was considered too much, the nitrogen in this treatment was derived 20 percent from cyanamid and 80 percent from sulfate of ammonia. Five other treatments were included. In three of these, all (100 percent) of the nitrogen was derived from sulfate of ammonia, nitrate of soda or ammonium phosphate, respectively. In the


Florida Agricultural Experiment Station





Fertilizzei, Expe;,it)ielits with Potatoes


fourth, the regular Station control 4-8-5 formula was used in which 25 percent of the nitrogen was derived from cottonseed meal, 25 percent from fish scrap and 50 percent from sulfate of ammonia. The nitrogen in the last treatment was derived half from nitrate of soda and half from sulfate of ammonia. Thus in all there were 12 treatments.
Urea and cyanamid are synthetic organics which resemble the inorganic nitrogen sources more closely than they do the natural protein organics, in availability to plants and in solubility in water. They may be distinguished for our purposes as watersoluble organics, in contrast to the older water-insoluble types like cottonseed meal and fish scrap.
The composition of the 12 fertilizer treatments as they were mixed at the Station is given in Table 10.
All of the fertilizers were applied at the rate of 1,500 pounds per acre with a machine distributor. The first year eight replications of each treatment were used. In the second and third years the number was reduced to six. Each plot harvested for data consisted of one row 75 feet long. The rows were spaced 76 inches apart (twice the ordinary distance), but in converting the yield data to an acre basis, a plot was considered arbitrarily as 75 feet x 38 inches or 1/184 acre.
This departure from ordinary plot technique deserves brief discussion. In placing the rows twice the ordinary distance apart the object was simply to eliminate planting of the buffer rows. Satisfactory evidence from previous observations indicated that there woul be no competitive effect between roots and fertilizers in adjacent plots. In converting the data to an acre basis it was recognized that while the plots actually were 75 feet x 76 inches, from a practical standpoint conversion on this basis would yield values far below normal commercial yields, and this would be misleading, since the actual yields per plant were quite normal. As the object of the test was to determine merely the relative effect of the various treatments, and not the absolute value of each, the only sensible procedure was to convert the yields on a normal basis of 38-inch rows. The fertilizer and seed, as well as the yields, were computed on the latter basis. The average yields for each year are given in Table 11,
A primary consideration in the selection of nitrogen sources for potato fertilizers has been their performance in relation to soil moisture. Growers usually consider that the water-insoluble











TABLE 10-.CoTPOSITION, IN POUNDS PER TON, OF TE FERTILIZERS COMPOSING THE SOURCE OF NITROGEN TESTS, 1937-40.


00
E M c
o U2

Treatment, with the Per- E 'd . C Hz
centage N each Source Con- < , c .0 , .
tributesto the Total N in , - S 0 "W 0 C)
the Formula . 5 WO0 9.09 .0
C) C 0 ' CJ Q4 P



Sulfate of ammonia 100% 381 2 800 209 100 510
N itrate of soda 100% ----- 534 ---- ----- ---- ----- ---- ----- ---- ----- 800 209 100 357
Nitrate of soda 50% *190 267 ---- ----- ---- ----- ---- ----- ---- ----- 800 209 100 434
Ammonium phosphate 100% 500 300 209 100 891
Cottonseed meal 50% 190 615 755 190 100 150
Fish scrap 50% * 190 421 - - 740 209 100 340
B & B tankage 50% 190 . . 615 . - 680 209 100 206
Dried blood 50% 190 -------- . . . . . 308 800 209 100 393
Urea 50% * 190 - 95 800 209 100 606
Milorganite 50% 190 - 696 730 209 100 75
Cyanamid 20% * 305 76 800 209 100 510
Control-cottonseed meal 25 ,,
fish scrap 25% * 190 308 210 745 200 100 247
� Balance of N from sulfate of ammonia.





Fertilizer Experients wzit/i Potatoes


TABLE 11. AVERAGE YIELDS Or -NO. 1 TUBERS (BUS3HELS PER ACRE) FROM THE
VARIOUS SOURCE OF-NITROGE-, TREAT-MENTS, U SING A 4-S-5 ANALYSIS Al 1,500
POUNDS PER AC~RE.
Fertilizer
Nitrogen Source Cost 1937-38 1938-39 1939-40 Average
1,500 Pounds
100% from sulfate of ammonia $17.96 237 144 249 210
100% from nitrate of soda 19.35 240 144 246 210
50% from nitrate of soda* 18.65 242 149 239~ 210
100% from ammonium
phosphate 22.38 242 151 247 213
50% from cottonseed meal* 21.84 256 145 241 214
50% from fish scrap: 22.64 246 150 239 212
50% from blood-and-hone
tankage* 22.55 263 150 257 223
50% from dried blood* 23.16 256 148 255 220
50% from urea: 18.37 249 155 238 214
50% from milorganito" 22.33 246 170 255 224
20% from cyanamid" 18.35 252 145 237 211
25% from fish scrap, 25% cottonseed meal 22.24 245 154 236 212
Balance of N from sulfate of ammonia. sources are best in a wet year. Referring back to Table 4 and to the seasons 1937-38 through 1939-40, it will be noted that these seasons were quite variable with regard to rainfall. The 1937-38 season may be considered as intermediate, with a total of 13.55 inches. The 1938-39 year was a dry one, with only 9.14 inches, as was reflected in the lower yields. The 1939-40 season was an unusually wet one, with rainfall well distributed throughout the growing season. Thus it may be considered that the test was run under conditions which varied considerably from dry to wet. With all of this variation, the yields did not vary much among the treatments, arnd therefore tile expert iment was termninated with the 1939-40 test.
The results first will be discussed by individual years. rt was found that a minimum difference in yield of 7 bushels per acre was necessary for significance in the 1937-38 data. Therefore, the treatments sulfate of ammonia, nitrate of soda, the combination of these two, ammonium phosphate, fish scrap, milorganite and the fish scrap-cottonseed meal combination as a group may be considered as all having yielded about alike. Slightly higher yielding were urea and cyanamid and a little higher yielding than these were tankage, dried blood and cottonseed meal. The yields of all 12 treatments were so nearly alike, however, that fine distinctions in yield differences are hardly justified.
It was found that at least a 6-bushel difference was required for significance in the 1938-39 data. On this basis, all of the






Florida A1gricultural Experiment Station


treatments except urea, milorganite and the fish scrap-cottonseed meal combination yielded about alike. The latter three may be considered as higher yielding than some of the others, but only the milorganite significantly outylelded all of the others.
In 1939-40 a 7-bushel difference was found necessary for significance. This year milorganite, dried blood and tankage stood out as being significantly highest yielding, although minor significant differences occurred between some of the other treatments, and a few-sulfate of ammonia, nitrate of soda and ammonium phosphate-yielded almost the same as these three water-insoluble organics.
From this brief consideration of the yearly data it will be noted that while certain trends in yield differences appear, the differences in most cases are so small as to leave some doubt as to their practical application. Conclusions can best be drawn from the data secured by averaging the treatment yields over the three-year period.
For these data (last column, Table 11) it was found that a difference of more than 4 bushels per acre was necessary to establish significance. On this basis the treatments sulfate of ammonia, nitrate of soda, the sulfate of ammonia-nitrate of soda combination, ammonium phosphate, cottonseed meal, fish scrap, urea, cyanamid and the fish scrap-cottonseed meal combination all yielded about alike. Yielding slightly but significantly higher were milorganite, tankage and dried blood.
In order to help determine the relative profitableness of the treatments, the fertilizer cost per bushel of potatoes produced was determined from these data for each season. The costs are shown in Table 12.
TABLE 12.-FiETIT IZER COST PER BUsuviL Or POTATOES FROM% THlE VARIOUS SOURCEup NITRO-uN TREATIENTS. (ARRANGED IN ORDER OF AVERAGE COST BASED ON
Ax ERA ; Y1 IELD FO TILEE YEARS.)


50% fror 100% fror 20% fror 50% fror 100% fror 50% fro' 50%7 frost 50% fror 50% fror


Nitrogen Source 1937-38 1938-39 1939-40
n urea1 $ .073 $ .119 $ .077
n sulfate of ammonia .076 .125 .072
n cyanamid* .073 .127 .077
n nitrate of soda1' .077 .125 .078
n nitrate of soda .081 .133 .079
mn milorganite* .091 .131 .088
* blood-and-bone tankage* .086 .150 .088
n cottonseed meal* .085 .151 .091
n dried blood* .090 .156 .091


25% from fish scrap, 25% cottonseed
meal'
100% from ammonium phosphate 50% from fish scrap*


Average
$ .086
.086 .087 .089 .092 .100 .101
.102 .105


.091 .153 .094 .105 .092 .144 .091 .105 .092 .151 .095 .107


*Balance of N from sulfate of ammonia.






Fertilizer Experimients -with Potatoes


Each year, and for the three-year average, the fertilizer cost of producing a bushel of No. 1 potatoes was less from the water-soluble sources-urea, sulfate of ammonia, cyanamid and nitrate of soda-than from the other sources.
Since it has already been shown that cottonseed meal, ammonium phosphate, fish scrap and the fish scrap-cottonseed meal combination were significantly outyielded by milorganite, dried blood and tankage, there is no question but that the latter three were more profitable than these others, which cost about the same.
In comparing milorganite, dried blood and tankage with the cheaper sources-urea, sulfate of ammonia, cyanamid and nitrate of soda-however, a little more difficulty occurs.
Considering the first three treatments as one group and the latter four treatments as another group, by averaging the data for each it is calculated that the first, or insoluble materials, group required $4.17 per acre more of fertilizer cost to produce an additional 11 bushels of No. 1 potatoes. These potatoes had an undug field value of about 82 cents per bushel according to Howard and Steffani3, or $9.02 for the 11 bushels. Thus the extra $4.17 cost for fertilizer of the first group returned an increased yield worth $9.02, as compared with the second group, or a difference of $4.85 per acre in favor of the more insoluble organic nitrogen group.
The above calculations are presented for the benefit of those who are interested in knowing just what the profit and loss of the various treatment groups were under the conditions in which the experiments were performed. From a broader and more practical viewpoint, however, the authors believe that definite conclusions on the basis of these particular "cost and returns" data alone are not warranted, because conditions are almost never the same in any two years. Fertilizer costs, yields and returns vary considerably. Furthermore the calculations are based on average yields, which in themselves are only significant within certain limits, as already pointed out. Also it must be recognized that as yields increase on an acre of land, the costs per bushel for land rent, machiner 'y use, spray materials, seed, labor and supervision decrease, which probably would more than offset small increases in fertilizer cost.
It would seem more sensible, therefore, to consider the re3 Howard, R. H. and C. HI. Steffani. A studv of potato farming in Dade Couinty. Florida, seasons 1934-35 to 1938-39. Fla. Agr. Ext. Ser. Mimeo. Circ. Dept. Agr. Econ. Potatoes AE4. 1939.






24 Floi-ida Agi-icidtinal Experinicnt Station
sults in more general terms. From this standpoint it is concluded that in general as used in this experiment, the fertilizer sources milorganite, blood-and-bone tankage (medium grade) and dried blood slightly outyielded the other sources. The differences in yield, although small, were profitable for the three-year period based on cost and returns data cited but in certain of the years some of the water-soluble sources were equally as profitable. It would seem wise to continue the use of either milorganite, dried blood, or blood-and-bone tankage in fertilizers for potatoes in Dade County for the present. There is no evidence to indicate that any one of these three is better than the other. Since prices fluctuate from time to time, whichever one is cheaper per unit of nitrogen at the time of purchase is the one to choose.
Sulfate of ammonia and nitrate of soda yielded alike, as sole sources of nitrogen, and no advantage was demonstrated from combining the two on a half and half basis, Price differential per unit of nitrogen would accordingly determine the preference for use in potato fertilizers in this area.
Ammonium phosphate was the least economical of the watersoluble nitrogen sources tested. Although its performance was quite satisfactory, its use cannot be recommended because of its higher cost. It is possible, however, that it might be a desirable source for use in high analysis fertilizers, which were not included in the scope of this test.
The use of urea and cyanamid is worthy of consideration. They both produced yields in the upper range of the water-soluble sources, and both sources were among the treatments showing least fertilizer cost per bushel. Since many growers are of the impression that urea leaches less readily from the soil than do some of the other water-soluble sources, this would give it an apparent advantage in wet seasons, especially if an entirely water-soluble fertilizer is desired. No data are available for Dade County soils on the leaching properties of urea.
The relatively good showing of the 20 percent cyanamid treatment is noteworthy. Some fertilizer manufacturers would like to use a small amount of this ingredient in their mixtures because of its conditioning effect. The data of this experiment indicate that there should be no objection to this practice so far as potato fertilizers for these soils are concerned.
The performances of cottonseed meal and fish scrap indicate that they are less desirable than the other water-insoluble sources tested, and, because of their present higher cost, show no economical advantage over the water-soluble sources with which theY were also compared.






Fcrtilizcr Expcirimiciits -wit/ Pot atocs


MURIATE VS. SULFATE OF POTASH
There has been a great deal of debate among growers as to the relative merits of muriate and sulfate of potash for potato fertilizers. A series of experiments was begun in the fall of 1933, testing these two ingredients. Each was used as the sole source of potash in a regular 4-8-5 fertilizer, with manganese. The rest of the mixture in each case was composed of sulfate of ammonia, fish scrap, cottonseed meal and superphosphate.
The first two years, four replications of each treatment were planted. The third year 10 replications and the last year six replications were used. The first three years the fertilizer was applied by hand. The last year it was applied by machine. The tests were not conducted in 1936-37. The first year a ton per acre was used, but in each of the other three years the fertilizer was applied at the 1.500 pound rate. The plots were on different ground each year.
The replicate plot yields each year were averaged and the annual average yields from each treatment are given in Table 13.
TABLE 13.-AN-ERAGF YIELDS 0F NO. 1 TUBERS (BUSHEILS PER ACRE) IROM- PLOTS
COAIPARTIG THE Mi RIATE AD SU LFATE FOR[M ASo SORCE Or POTASH I,, A
4-8-5 FERTILIZER.
Source 1933-34 1934-35 1935-36 1937-38 Average
Sulfate of potash 222 199 278 298 249
Muriate of potash 209 189 281 301 245

The data of Table 13 indicate only slight differences in yield of tubers between the two sources of potash, and these differences were found not to be statistically significant. The findings of this experiment indicate that either source may be used with equal results.
It should be noted that each year these tests were carried out in a different field. Just what effect would be had by using either source continually on the same field remains to be determined.

MANGANESE
As a result of the pioneering work of Skinner and Ruprecht4 and of leading growers, most of the potato growers in the Homestead area were using manganese sulfate in their fertilizers by the time the Sub-Tropical Experiment Station was established

4 Skinner, J. J., and R. W. Ruprecht. Fertilizer experim-ents with truck crops. Fla. Agr. Exp. Sta. Bul. 218. 1930.






26 Florida Agricultural Experiment Station

in 1930. The usual amount of manganese sulfate added was 200 pounds per ton but some growers were using as much as 400 pounds per ton. Experiments were started by the Station in the fall of 1931 to determine the effect on yields of lesser amounts of manganese, and of the residual effect of manganese.
RATE OF APPLICATION
A series of plots was laid out on the East Glade farm to be continued over a period of years. Four manganese treatments (65 percent manganese sulfate) were included: 0 lbs., 50 lbs., 100 lbs. and 200 lbs. per ton of 4-8-5 fertilizer, respectively. The fertilizer was applied at the rate of 1,500 pounds per acre and was distributed in the furrow by hand. The test included three replicate plots of each of the four treatments. Average yields from each treatment are given in Table 14.
TABLE 14. YrIS OF No. I TUBERS (BUSHTELS PER ACRE) RESULTING FROM VARIOUS
RATES OF APPLICATION OF MANGANESE SULFATE PER TON OF 4-S-5 FERTILIZER. Season 0 lbs. 50 lbs. 100 lbs. 200 lbs.
1931-32 128 126 155 147
1932-33 195 237 228
1933-34 174 196 216 199
1934-35 137 139 146 133
Average 159 174 186 159

The data are quite consistent in showing that 100 pniincts of manganese sulfate per ton (75 pounds per acre) was equally as good as, if not better than, the greater and lesser amounts tested.
In order to check these results on a commercial scale, seye~ral cooperative tests were carried out with growers.
The first of these was conducted on the farm of W. J. 'Vick irn 1934-35. Eleven rows, each one-fourth mile long, were planted n the regular commercial manner for each treatment, aind eachj treatment was teplicated once. The regular basic fertilizer was a 4-8-5 applied at 1,500 pounds per acre. The three treatments consisted of 100, 150 and 200 pounds per ton, respectivelIy, of 65 percent manganese sulfate mixed with the fertilizer. Ten strips each 50 feet long were harvested at random from each plot and averaged for each treatment to obtain the data shown in Table 15 (Vick). The data show clearly that the 100-pound rate yielded slightly better than the higher rates of manganese application.
Another test was conducted in 1935-36 on a commercial scale on one of the farms of F. C. Peters, Inc. This farm, by the way, had never before had applications of manganese, so far as the owner knew. Approximately one-quarter acre was planted with





Fcrtili--cr E.vperiiiicids -,,ith Potatoes


TABLE 15-AVERAGE YIELD; OF -NO. I TUBERS (BUSHELS PER ACRE) REsULIING FROM
VARIOUS RATES OF APPLICATION OF MANGANESE SULFATE PER TON OF 4-8-5
FERTILIZER IN COOPERATIVE TESTS.

Lbs. of Manganese Sulfate per Ton Cooperator 100 150 200
Vick 235 219 216
Peters 244 245 244

each of the three manganese (65 percent) treatments, 100, 150 and 200 pounds per ton. The 4-8-5 fertilizer to which the manganese had been added was applied by machine at the rate of one ton per acre. Ten sub-plots, each 2 rows 50 feet long, were harvested at random from each treatment plot. The average yields are also given in Table 15 (Peters). The data show no significant differences in yield among the three treatments, again indicating the 100-pound rate to be as effective as the higher rates.
This particular test has further significance in view of the idea among some growers that more than 100 pounds of manganese sulfate per acre is needed for potatoes on "new" land. Apparently this idea is erroneous,
RECENCY OF APPLICATION
Experimental work was also started on the Station's East Glade farm in the fall of 1931 to study the residual effect of manganese over a four-year period. The plots, each 1/72 acre in size and each with three replications, were established with permanent boundaries for the four-year period. A 4-8-5 fertilizer, applied by hand at the rate of 1,500 pounds per acre, was used each year. When manganese applications were made, 65 percent manganese sulfate was mixed directly with the fertilizer, at the rate of 100 pounds per ton.
Five basic treatments were included. Manganese was applied to four of the plots the first year and omitted from the fifth. The second year manganese was applied to Plots 1, 2 and 3 and omitted from 4 and 5. The third year only Plots 1 and 2 received manganese, and the fourth year only Plot 1. Thus by the end of the fourth year data had been secured from plots receiving no manganese for four5, three, two and one years, respectively, and at the same time from plots receiving manganese for no, one, two, three and four years in succession, respectively. The replicate yields of No. 1 tubers of the various treatments

'As a matter of fact these plots, to the writers' knowledge, had received no manganese for at least eight years.






Florida Agricultural Experimcnt Station


have been averaged and are given in Table 16. Figures in italics indicate where manganese had been applied. TABLE 16.-YIELDS OF No. 1 TUBERS (BUSHELS PER ACRE) As RELATED To RECENCY
or APPLICATION or MANGANESE SULFATE.
Treatment 1931-32 1932-33 1933-34 1934-35
A 134 161 157 142
B 155 183 186 146
C 148 203 179 145
D 145 21; 196 148
E 149 217 216 160

The yields fluctuated from year to year as is expected of different seasons. The data show consistently that each year the yields were higher in the plots receiving manganese that year. Because of the normal yield fluctuations from year to year, the data are difficult to interpret from the standpoint of residual effects. Only a generalization can be made-that there is a tendency for the effect of manganese to continue somewhat into the second and possibly into the third year after its application, but that for all practical purposes, however, manganese should be applied each year, through a period of at least four years.
Two years later (1937-38) another experiment was conducted on another section of the farm, to which manganese had been applied for at least five years in succession. Replicate plots (six each) with and without manganese (75 pounds per acre) were compared for yield. This year the fertilizer was applied by machine. The average yields, from Table 20, are given in Table 17 (1937-38). The addition of manganese did not increase the yield.
TABLE 17.-AVERAGL YIELDS OF No. I TUBERS (BUSHELS PER ACRE) OBTAINED IVITH
AND WITHOUT MANGANESE IN A STANDARD 4-8-5 FERTILIZER ON LAND WHERE
MANGANESE HAD BEEN APPLIED FOR SEVERAL YEARS.
Season Control Manganese
1937-38 292 292
1938-39 142 139

Again in 1938-39 the test was repeated in another field which had received manganese annually for at least seven years. Plots with and without manganese were planted as before, with the results (from Table 21) given also in Table 17 (1938-39). These data also show no increase in yield resulting from the application of manganese.
The results of these experiments indicate that for the first four years, at least, manganese sulfate should be applied annual-





Fertilizer Experiments with Potatoes


ly, but that after a period of five years or more of successive applications, the manganese content probably may be omitted entirely from the fertilizer one year without decreasing the yields. The data here presented indicate not only that manganese exhibits some residual effects, but also that there may result an accumulation of it in the soil from year to year in a form available to the potato plant. It is possible, of course, that the residual effect of manganese becomes more pronounced when the fertilizer is applied by machine in the band method than by hand where it becomes more thoroughly mixed with the soil.
It is worthy of note that fairly good yields of potatoes were obtained during several years of growing them without adding any manganese on land with no previous history of manganese applications (treatment A, Table 16). Although the addition of manganese with the fertilizer increased the yields profitably, its omission resulted only in lessened yields and slightly smaller vine growth (Fig. 5). No other characteristic foliage symptoms of manganese deficiency were noted.

OTHER SOIL AMENDMENTS
The results obtained with manganese naturally promoted a parallel interest in other soil amendments. Accordingly, some experiments with these were started on the Station farm in 1934, to determine if the addition of any of these other elements might exhibit a beneficial effect on potato yields. The first trials were of a more or less preliminary nature and included tests of zinc, copper, iron and magnesium, as well as manganese in various

Fig. 5-The 1932-33 potato manganese test. The plot in the background received 100 pounds of manganese sulfate: that in the foreground no manganese for two years.






F lorida .16-ricultunda Expcrinwnt Station


combinations. The materials used as sources of these elements and amounts applied per ton of fertilizer were as follows:
1. Zinc-from 89 percent zinc sulfate @ 75 pounds per ton.
2. Copper-from bluestone crystals (copper sulfate) @ 100
pounds per ton.
3. Iron-from copperas (ferrous sulfate) @ 125 pounds per ton. 4. Magnesium-from commercial epsom salts (magnesium sulfate) @ 300 pounds per ton.
5. Manganese-from 65 percent manganese sulfate @ 100 pounds
per ton.
6. Calcium-from gypsum (calcium sulfate) @ 1000 pounds per
ton.
All of these ingredients, in whatever combination they were used, were mixed with a regular 4-8-5 fertilizer (50 percent of N from organic sources), which was applied by hand at the rate of 1,500 pounds per acre. Thus the elements were applied per acre at three-fourths the amounts listed above.
The test the first year was composed of 14 treatments, applied to land which had had manganese applications for several years previously. The treatments are listed in Table 18 in their order of yield of No. 1 tubers. The plots consisted of single rows, 50 feet long and 38 inches apart, and each treatment replicated 16 times, in two series of 8 plots each. The tubers from each replicate plot were graded and the data converted to an acre basis and averaged for each treatment, with results shown in Table 18.
TABLE is. ANERAGE YIELDoS or No. 1 TUBERS (R sin 'LS PER ACRE) FROMf THE
TREATMENTS COMPOSING THlE 1934-3s MICRo NUTRIFNNT FuES rTFSTS.


Elements Added to 4-8-5 Fertilizer

Manganese-zinc -copper Manganese -zinc -copper-iron Manganese-zinc Manganese-magnesium Iron-zinc
Zinc
Manganese-imagnes ium -zinc- iron -copper Iron -zinc-copper Manganese-copper Manganese-calcium-zinc-iron -copper Iron
Iron-copper Copper
Manganese (control)


Increase
Yield Over Odds
Control
237 38 4298:1
227 28 71:1
223 24 121:1
223 24 121:1
217 18 27:1
216 17 43:1
216 17 17:1
215 16 14:1
205 6 2:1
201 2
199 0
196 - 3
192 - 7
199


According to the data of Table 18, a number of the treatments showed average yields somewhat higher than the control or manganese treatment.





Fertilizer Experii7zeizts with Potaioes


The data were analyzed statistically by "Student's" t test. Significance of differences in yield over that of the control plot is indicated in the column "odds". Odds of 30:1 or better are considered satisfactory for significance. On this basis the first six treatments listed in Table 18 may be considered as having significantly outyielded the control treatment, since the one treatment with odds of 27:1 closely approaches the arbitrary 30:1 standard.
A striking fact is that most of the combinations including zinc are represented in these six treatments, indicating that applications of zinc sulfate may be beneficial on these soils. This is especially indicated by the fact that zinc alone and the zincmanganese combination outyielded manganese alone.
The addition of copper and iron in combination with manganese and zinc did not significantly increase the yield over that obtained by adding the latter two alone, and since iron and copper alone or together did not increase the yield, the data of this test indicate they were not beneficial.
Evidence is provided that the addition of magnesium sulfate with manganese was beneficial, although when magnesium was added to the larger combination of manganese- zinc-iron- copper, the yield was lower than that of a similar mixture without magnesium.
The only mixture in which calcium sulfate was added failed to show an increase over a similar mixture without it.
The following year (1935-36) an experiment was conducted on different ground comparing the effects on yield of adding 89 percent zinc sulfate to the fertilizer at the rate of 50 pounds per ton, with and without manganese. The regular 4-8-5 fertilizer was applied by hand at the rate of 1,500 pounds per acre. Ten replications of each treatment were used. The average yields from each treatment are given in Table 19 (Series 1935-36). No increase in yield was obtained from the zinc applied in this experiment.
The next year (1936-37) similar tests were conducted with zinc sulfate (89 -oercent) at the rates of 5, 25 and 50 pounds per ton, and with magnesium sulfate (commercial epsom salts) at rates of 25, 100 and 400 pounds per ton. A control treatment of 100 pounds of manganese sulfate was included also. The fertilizer to which they all were added was the regular 4-8-5 mixture containing 100 pounds of manganese sulfate per ton, and it was applied by hand at the rate of 1,500 pounds per acre. Each of the seven treatments was replicated 10 times. The average yields





Florida Agricultural Experiment Station


TABLE 19.-AVERAGE YIELDS OF No. I TUBERS (BUSIIELS PER ACRE) OBTAINED 1_';
TESTS OF VARIOUS MicRo-NuTRIE.NT ELEMENTS ADDED TO STANDARD 4-8-5
FERTILIZER, 1935-37.
Series Elements Added Yields
Manganese (100)* (Control) 278
1935-36 Manganese (100) and zinc (50) 277
Zinc (50) 270
Manganese (100) (Control) 182
Zinc (5) 187
Zinc (25) 189
1936-37 Zinc (50) 184
Magnesium (25) 191
Magnesium (100) 185
Magnesium (400) 187
Figures in parentheses refer to pounds of the compound added per ton of fertilizer.
are given for each treatment in Table 19, Series 1936-37. Here again the zinc failed to increase the yield over the manganese treatment and magnesium also had no beneficial effect. The small differences in average yields are not significant,
The following year (1937-38) the micro-nutrient element tests were expanded to include more treatments and the fertilizer was applied by machine. This year five replications were planted of each treatment in a randomized block arrangement, with each treatment replicate paired with a control plot. The buffer rows, planted between treatment and control plot rows, were fertilized with the control mixture. Zinc sulfate (89 percent), copper sulfate (crystal bluestone), iron (commercial ferrous) sulfate, iron (technical ferric) citrate, and magnesium sulfate (commercial epsorn salts) were used this season in various combinations with and without manganese. The field in which the test was conducted had been planted annually to potatoes for at least the five preceding years, with manganese applications made each year.
All ingredients were added to the regular 4-8-5 mixture, which was applied at the rate of 1,500 pounds per acre. The control mixture was this same 4-8-5 with no manganese added. The treatments tested and the average yields from them and from their respective control plots are given in Table 20.
Statistical analysis of the data from which this table was compiled involved the use of "Student's" Z test, as applied to a paired plot arrangement. Odds of less than 30:1 are not considered significant.
Zinc at the 25-pound rate, both alone and in combination





Fertilizer Experi;neWs u4th Potatoes


TABLE 20.-AVERAGE YIELDS OF NO. I TUBERS (BUSHELS PER ACRE) O.U THE TREATMETNTS AND OF THEIR RESPECTIVE CONTROL PLOTS, 193,-3S MICRO-NUTRIENT
ELEMENT TESTS.
Ave. Increase
Elements Added to 4-8-5 Fertilizer Treatment Control of Treatment Odds Yield Yield Over Control
Zinc (75) manganese (100)" 310 285 25 23:1
Zinc (75) manganese (100) iron
sulfate (125) 308 280 28 36:1
Iron (ferric) (134) 304 290 14 73:1
Zinc (25) manganese (100) 304 278 26 60:1
Zinc (25) 303 285 18 49:1
Zinc (75) 300 290 10 4:1
Magnesium (400) manganese (100) 299 290 9 3:1
Iron (ferrous) (125) 299 276 23 44:1
Iron (ferrous) (125) manganese (100) 294 293 1 1:1
Zinc (75) copper (100) manganese
(100) 292 281 11 21.1
Manganese (100) 292 292 0 0
Zinc (75) iron (ferrous) (125) copper
(100) manganese (100) 285 279 6 6:1
Figures in parentheses refer to pounds of the compound added per ton of fertilizer.
with manganese, increased yields markedly. No explanation can be offered for the poor results with zinc at 75 pounds, where the increase was not significant.
Manganese failed to increase yields when applied alone. When applied with zinc the increase was not appreciable over the yield of zinc (25) alone. This failure might be expected, in view of the continual manganese applications made in previous years on the field where these tests were made.
Iron salts, both ferric and ferrous, increased yields significantly when applied alone, but not in combination with other elements. Iron added to zinc and manganese did not increase the yield over that of the latter two alone, nor did its addition with manganese increase the yield. The lack of increase from iron in combination with other elements is difficult to understand and no explanation is offered.
Neither magnesium nor copper caused increase in yields.
Another series of experiments along these same general lines was conducted in 1938-39. The 4-8-5 fertilizer, to which the various amendments were added, was applied by machine as before, at 1,500 pounds per acre. Five replications were planted of each treatment and its respective paired control plot on land which had received manganese for the preceding seven years. Zinc sulfate was added at 5, 25 and 75 pounds per ton, iron (ferrous) sulfate at 125 pounds and magnesium sulfate at 25





Florida Agricultural Experiment Station


and 400 pounds per ton. Copper sulfate and iron citrate were not used this year but boron, in the form of borax, was added at the rates of 12 and 24 pounds per ton. The treatments and average yields from them and from their respective control plots are
given in Table 21.
It was found that none of the small differences in average
yields this year were significant. Thus this year manganese, magnesium and zinc were of no effect in increasing yields. Likewise no increase was obtained from either iron or boron.

TABLE 21-AVERAGE FIELDS OF No. 1 TUBERS (BUSHELS PER ACRE) Or THE TREATAIENTS AND OF THEIR RESPECTIVE CONTROL PLOTS, 1938-1939 MICRO-NUTRIENT
ELEMENTS TESTS.
Average Increase
Ingredients Added to 4-8-5 Fertilizer Treatment Control of Treatment Yield Yield Over Control
Zinc (75) manganese (100)* 156 144 12
Iron (125) manganese (100) 155 142 13
Zinc (25) manganese (100) 149 140 9
Zinc (5) manganese (100) 148 142 6
Boron (12) manganese (100) 147 142 5
Magnesium (25) manganese (100) 144 137 7
Zinc (75) iron (125) manganese (100) 143 143 0
Iron (125) 141 142 - 1
Manganese (100) 139 142 - 3
Magnesium (400) manganese (100) 136 143 - 7
Boron (24) manganese (100) 132 139 - 7

Figures in parentheses refer to pounds of the compound per ton of
fertilizer.
SUMMARY OF MICRO -NUTRIENT ELEMENT STUDIES
Considering the experimental results over a period of years, the data are rather convincing in establishing that the addition of 14ron sulfate, copper sulfate and magnesium sulfate to the soil in fertilizer form is not to be recommended as a general practice in growing potatoes on these marl soils. In occasional instances their use resulted in slight increases in yield but the evidence is strong enough otherwise to indicate that their use in the amounts tested, even as yield "insurance," is unwarranted.
The use of calcium sulfate, or gypsum, will be discussed later. It did not increase the yield in the 1934-35 test.
Boron, used only one year at two different rates of application, failed to improve the yields,
The evidence is not so convincing concerning the use of zinc sulfate. In two of the five seasons zinc increased the yields significantly and the increases, although not very large, were profitable. This success is not to be ignored, although the fact






Fertilizer Experi)iirnts -,,ith Potatiws


that its addition to the fertilizer in the majority of years did not result in increased yields scarcely permits its recommendation for general use. The nature of zinc deficiency in these soils is not yet understood and it is not known if deficiencies occur in some years and not in others. Certainly the method of adding zinc needs further study. As in the case of manganese, no obvious symptoms of deficiency which could be attributed to zinc have been observed. It is possible that this element, as well as some of the others, may better be applied as a foliage spray, and experiments are now under way at this Station to determine this point.
Since this experiment was conducted with an ordinary 4-8-5 fertilizer with 50 percent of its nitrogen derived from organic sources, the results might not be applicable in instances where a radically different fertilizer program is followed. Also the fact that benefits from adding these various micro-nutrient elements have not been pronounced up to the present time does not exclude the possibility that their use may be required later on after intensive potato cropping has progressed year after year on the same land.
In addition to the more complete experiments thus far described a number of minor tests were carried out with sulfur, gypsum and manure.
SULFUR
Since the soil of this area is alkaline it seemed logical to determine what effect applications of sulfur might have on the resultant yields of potatoes. Accordingly, a test was conducted in 1932-33, in which a control 4-8-5 fertilizer treatment applied at 2,000 pounds per acre was compared with a similar treatment to which had been added 1,000 pounds of dusting sulfur per acre, applied in the furrows by hand with the fertilizer two days before planting. The resultant average yields from three replications of each treatment are given in Table 22, Series 1932-33.
No change in yield resulted from the addition of sulfur. The effect of the sulfur on the pH of the soil was not determined this season, but in 1938 Ruehle', in studies of potato scab, applied 1,000 pounds of sulfur per acre in a similar manner and found that the pH was changed slightly but the change was of short duration.

6Ruehle, G. D. Control of potato diseases in Dade County. Fla. Agr. Exp. Sta. Ann. Rept. 1938, p. 191.






Florida z1gricultural Experiment Station


TABLE 22.-AVERAGE YIELDS or No. I TUBERS (BUSHELS PER ACRE) RESULTING FROA'1
USE OF VARIOUS AMOUNTS Or SULFUR WITH A STANDARD 4-8-5 FERTILIZER.
Series Treatment Rate per Acre Yields
1932-33 Sulfur 1,000 pounds 195
Control -------------------- 197
1938-39 Sulfur 4,000 pounds 125
Control -------------------- 149
Filler Sulfur 123 pounds 140
Phos-rock 123 pounds 139

In 1938-39 a cooperative test was carried out on a field which had produced poor crops of potatoes due to its sea salt content. Analysis of the soil showed a chlorine concentration of about 4,000 parts per million in the surface and 1,100 p.p.m. at the 2 to 4-inch depth. The analysis was made by the Department of Chemistry and Soils of the Florida Experiment Station at Gainesville. Sulfur was among the materials tested on this soil. It was broadcast over the surface at the rate of 2 tons per acre and disked in the day before planting. A 4-8-5 fertilizer was applied at planting time at the rate of one ton per acre. Resultant average yields per acre from this test are given in Table 22, Series 193839. In this case sulfur actually reduced the yield.
This same year a test was conducted on the Station farm in which two 4-8-5 fertilizer mixtures, identical except for their filler material, were compared. One was made up in the ordinary manner, using raw rock phosphate as filler, and the other with flowers of sulfur as filler. Each filler was added in the amount of 164 pounds per ton, and the complete fertilizers were applied at the rate of 1,500 pounds per acre by machine. Each plot was replicated five times. The replicate yields were averaged for each treatment and are also found in Table 22, Series Filler. In this test the yields were the same for both treatments.
These three experiments, then, indicate that no increase in yield of potatoes may be expected from the addition of sulfur in the amounts and by the methods tested.
GYPSUM
The effect on yields of applying commercial calcium sulfate (gypsum or landplaster) was also studied over a period of years, as the result of preliminary tests in pot cultures which showed a stimulating effect of calcium on growth of tomato plants in marl soil.
In 1932-33 gypsum was applied to the furrows by hand at the rates of 1,000 pounds per acre, along with a regular 4-8-5





Fertilizer Experiments with Potatoes


fertilizer applied at 2,000 pounds per acre. This treatment was compared with a similar one without gypsum added. Three replications of each were planted and the resultant average yields are found in Table 23 (1932-33). The difference between these two yields was too small to be of significance.
The following year gypsum was again applied with a 4-8-5 mixture by hand at two different rates. The treatments, each replicated four times, and their average yields are given in Table 23 (1933-34) also. This time small but significant increases in yield could be attributed to the gypsum, but it is extremely doubtful if the additional cost and labor of applying the gypsum were justified.

TABLE 23.-AN7ERAGE YIELDS OF NO. I TUBERS (BUSHELS PER ACRE) OBTAINED L,
TRIALS oF Gypsu- ,i ADDED To REGULAR 4-8-5 FERTILIZER.
Series Treatment Rate per Acre (lbs.) Yields
1932-33 Control 2000 196
Control + gypsum 2000 + 1000 202
Control 2000 198
1933-34 Control + gypsum 2000 + 1000 217
Control 1000 172
Control + gypsum 1000 + 500 187
1934-35 Mixture 1500 227
Mixture -r gypsum 1500 + 750 201
1938-39 Control 2000 149
Control + gypsum 2000 + 4000 132

Another test in 1934-35 has already been described (Table 18). Gypsum at 750 pounds per acre was added with a 4-8-5 mixture (at 1,500 pounds) containing manganese, zinc, copper, and iron, and compared with a similar treatment without gypsum. The results are given again in Table 23 (1934-35) for comparison. In this test gypsum decreased the yield.
The cooperative experiment of 1938-39 in the field with a high chlorine content, which contained the sulfur experiment, also contained a treatment in which gypsum was broadcast at the rate of 2 tons per acre and disked in prior to planting. Average yields per acre of the 10 replications of the gypsum and control plots are included in Table 23 (1938-39). Again in this experiment gypsum failed to increase the yield.
Thus, it is concluded that the addition of gypsum in the amounts tested to these marl soils is not justified from the standpoint of increasing potato yields.





Florida Agricultural Experiment Station


MANURE
A series of replicated plots was planted in 1932-33 to test the effect on yield of adding stable manure to the soil. Stable manure was broadcast over the plots at the rate of 6 tons per acre and disked in prior to planting. The potatoes in the manure and control plots both were fertilized with a ton per acre of 4-8-5 containing 100 pounds of manganese sulfate. The average yields are given in Table 24 (1932-33). The manure significantly increased the yield, but at local manure prices (about $5 per ton) the treatment was not profitable. TABLE 24.-AVERAGE YIELDS OF No. I TUBERS (BUSHELS PER ACRE) OBTAINED 1N
TRIALS 01, MANURE APPLIED IN ADDITION TO STANDARD 4-8-5 FERTILIZER CONTAINING MANGANESE.
Series Treatment Rate per Acre (tons) Yields
1932-33 Control 1 196
Control + manure I + 6 220
1938-39 Control 1 149
Control + manure I + 8 162

Again in the 1938-39 season manure was added as one of the treatments on the salty land to which the sulfur and gypsurn were applied. In this experiment stable manure was broadcast uniformly at the rate of 8 tons -per acre and disked in just prior to planting. The potatoes were fertilized in the regular cornmercial manner with a ton of 4-8-5 containing 100 pounds of manganese sulfate. Average resultant yields of tubers are given in Table 24. Again the manure increased the yield but not enough to warrant the additional cost.
These two experiments indicate that stable manure has a beneficial effect on yields, but as applied in these tests its additional cost was -Drohibitive. Incidentally, in neither test did the percentage of potato scab (Actinomyces scabies (Thax.) GUssow) increase. There was practically no scab in any of the plots.
SUMMARY
Fertilizer experiments conducted at the Sub-Tropical Experiment Station over a period of 10 years are reported in which analyses, amounts, sources of nitrogen and potash, and applications of manganese and other soil amendments were studied as affecting yields of Bliss Triumph potatoes grown on the marl soils of Dade County, Florida.
Analyses varying from a 2-8-5 and 3-12-8 to 8-16-10 were tested. The results indicate that for mixtures to be applied at rates of about 1,500 to 2,000 pounds per acre there is no justifica-





Fertilizer Expcrinients with Potatoes


tion for increasing the analysis beyond 3 or 4 percent nitrogen,
8 percent phosphoric acid and 4 or 5 percent potash.
Experiments with ordinary 4-8-5 and 3-12-8 analyses showed that the correct amounts of these to apply was from about 1,500 to 2,000 pounds per acre, and in most instances the most profitable amount was 1,500 pounds.
In four of the five years of testing a 4-8-5 mixture in which 33 percent of the nitrogen was derived from organic sources yielded as well as mixtures containing a higher percentage of nitrogen so derived. No definite relation of treatment yields to monthly rainfall during the crop season was observed during the five years over which these tests were conducted.
The source-of -nitrogen tests indicated that the organic materials milorganite, blood-and-bone tankage (medium grade) and dried blood slightly outyielded the other sources, and profitably so. No significant differences in yield were obtained among these three. Fish scrap, cottonseed meal, urea and cyanamid yielded slightly less, and about the same as mixtures containing all their nitrogen from ammonium phosphate, sulfate of ammonia or nitrate of soda. No advantage was demonstrated from combining the latter two materials. Likewise a fish scrap and cottonseed meal combination fielded no better than mixtures in which these materials were used separately.
The urea and cyanamid treatments both produced yields in the upper range of the water-soluble sources, and both were among the treatments showing the lowest fertilizer cost per bushel.
No significant differences in yield were obtained when sulfate of ammonia and nitrate of soda were compared as inorganic sources of nitrogen in a 4-8-5 formula containing 50 percent of its nitrogen from organic sources.
Likewise no significant differences in yield were obtained between the sulfate and muriate forms of potash as sources of potash in a similar 4-8-5 fertilizer.
Applications of 100 pounds of 65 percent manganese sulfate per ton of fertilizer gave as good yields as greater amounts. It was found that on "new" land applications of manganese sulfate should be made annually for at least the first four years. After a period of five or more years of successive applications, however, results indicated that the manganese content of the fertilizer could be entirely omitted at least one year without decreasing the yield.





Florida Agricultural Experiment Station


Applications of magnesium sulfate (epsom salts), copper sulfate (bluestone), iron sulfate (copperas), borax, iron citrate, sulfur and calcium sulfate (gypsum) failed to increase yields profitably. There were some seasons when zinc sulfate increased yields but results were too inconsistent to warrant its general recommendation as a fertilizer ingredient.

Stable manure increased the yields slightly when applied at 6 to 8 tons per acre with commercial fertilizer, but not sufficiently to justify the cost.

ACKNOWLEDGMENTS
The Board of County Commissioners of Dade County assisted by appropriating part of the funds for these experiments each year since the 1935 season. Dr. G. D. Ruehle supervised much of the spraying of the plots in later years, and Messrs. L. R. Toy, Ivan Moser, Lloyd Field and Roy Nelson assisted with much of the field work involved in conducting the tests.
Messrs D. P. Blake, Jr., Luther Chandler, F. M. Dolan, J. H. Estes, J. M. Holferty, F. C. Peters, Frank Rue and W. J. Vick, growers, kindly cooperated in some of the tests reported. County Agent Charles Steffani also assisted with the cooperative work.
The Redland District Chamber of Commerce provided six acres of test plots for some of the experiments conducted in 1932-33.




Full Text

PAGE 1

Bulletin 352 D e c embe r , 1940 UNIVERSITY OF FLORIDA AGRICULTURAL EXPERIMENT STATION WrLMON NEWELL, Dir ec tor GAINESVILLE , FLORIDA fertilizer Experiments With Potatoes On Th e Marl SoJs of Dad e Count y By W. M. FIFIELD and H. S. WoLFE Fi g. 1. W as hin g a nd grading potatoes at the S u b-Tropical E xperime nt St ation. Single copies . free to Florida residents upon request to AGRICULTURAL EXPERIMENT STATION GAINESVILLE, FLORIDA

PAGE 2

EXECUTIVE STAFF John J. Tigert, M. A., LL.D., President of th e University• Wilmon Newell, D.Sc., Director• Harold Mowry, M. S. A., Asst. Dir., Research J. Franc is Cooper, M.S.A., Editor• Jefferson Thom as, Assistant Editor• Clyde Beale, A.B.J., Assistant Editor• I da K ee ling Cresap, Librarian Ruby Newhall, Administrative Manager• K. H. Graham, Bus ine ss Manager• Rachel McQuarrie, Accountant" MAIN STATION, GAINESVILLE AGRONOMY W. E. Stokes, M .S ., Agronomist 1 W. A. Leukel, Ph.D., Agronomist Freel. H Hu ll, Ph.D., A grono mist G . E. Ritchey, M.S., Associate• W. A. C a rver, PH.D ., Associate John P. Camp, M.S., Assistant Roy E. B laser, M.S., Assistant Fred A. Clark , B.S .A., A ss istant ANIMAL INDUSTRY A. L. Shealy, D.V.M., Anima1 Industriali st" R . B. Becker, Ph . D., Dairy Husbandman• E. L. Fouts , Ph.D .. Dairy Technologist• W. M. Nea l, Ph .D. Assa. in An. Nutrition D. A. Sanders, D.V.M., Veterinarian M. W. Emmel, D . V.M., Veterinarian• N. R. Mehrhof, M.Agr., Poultry Hus bandman• W . G. Kirk, Ph.D., Asso . An . Husband man 3 D. J. Smith, B.S.A., As st. An. Husb . P. T. Dix Arnold, M.S.A., Asst. Dairy Husbandman• . L. Rusoff, Ph. D., Asst. in An. Nutrition• 0. W. Anderson, M.S., Asst. Poultry Husbandman• L. E. Mull, M.S., Asst. in Dair y Tech. SOILS R. V. Allison, Ph . D., Chemist' Gaylord M. Vol k. M.S., Chemist F . B. Smith, Ph . D .. Microbiologist C . E. Bell, Ph . D . , Associate Chemist H. W. Winsor, B.S.A., As s istant Chemist J. Russ e ll Hend erso n, M.S.A., Associate• L. H. Rogers, M.S., Assa. Biochemist Richard A. Carri gan, B .S., Asst. Chemist ECONOMICS, AGRICULTURAL C. V. Noble , Ph.D., Agricultural Economist 1 3 Zach Savage, M.S.A .. Associate A . H. Spurlock , M.S.A., Associate ECONOMICS, HOME Ouida D. Abbott, Ph.D., Home E co omist1 Ruth O verst reet, R.N., Assistant R . B. French, Ph .D., Asso. Chemist ENTOMOLOGY J . R. W a tson, A.M., Entomologist1 A . N. Tissot, Ph.D., As sociat e H . E. Bratley, M . S.A., Assistant HORTICULTURE G. H. Blackmon, M.S.A., Horticulturist> A. L. Stahl, Ph . D ., Associate F. S. Jamison , Ph . D., Truck Hort . R . J . Wilmot, M.S.A., Fumigation Specialist R. D. Dickey, M.S.A., Asst. Horticulturist J. Carlton Cain, B.S.A., Assistant Horticulturist Victor F . Nett l es, M.S.A . , Assistant Horticulturist F. S. La gas se, Ph.D., Horticulturist' H. M. Sell, Ph.D. , Assa. Horticulturist 2 PLANT PATHOLOGY W. B. Tisdale, Ph . D., Plant Pathologist• George F. Weber, Ph.D., Plant Path.3 L. 0. Gr atz , Ph . D., Plant Pathologist Erdman West, M.S., Mycologist Lillian E. Arnold , M.S. , Asst. Botanist BOARD OF CONTROL H. P. Adair, Chairman, Jacksonville W. M. Palmer, Ocala R. H. Gore, Fort Lauderdale N. B. Jordan. Quincy T. T. Scott, Live Oak J . T. Diam ond, Secretary, T a llahassee BRANCH STATIONS NORTH FLORIDA STATION, QUINCY J. D. Warne r, M.S., Agron. Acting in Charge R. R. Kinkaid, Ph.D., Assa. Plant Path. Ellio tt Whitehurst, B.S.A., Assistant An. Husbandman Jesse Reeves, Asst. Agron., Tobacco CITRUS STATION, LAKE ALFRED A. F. Camp, Ph.D., Horticulturist in Charge. John H. Jefferies, Asst. in Cit. Breeding Michael PeeciJ, Ph.D., Soils Chemist B. R. Fudge, Ph .D., Associate Chemist W. L. Thompson, B.S., Ass ocia te Entomologi st F. F. Cowart. Ph.D., Asso. Hort icu lturist W. W. Lawle ss, B. S., Asst. Horticulturist R. K. Voorhees, M.S., Asst. Plant Path. EV ERG LADES STA ., BELLE GLADE J. R. Neller, Ph.D., Biochem is t in Charge J. W. Wilson, Sc.D., Entomologist F. D. Stevens. B.S., Sugarcane Agron. Thomas Bregger, Ph . D., Sugarcane Physiologist Fred erick Boyd, Ph.D ., Asst. Agronomist G. R. Townsend, Ph.D., Plant PathologiS: R. W. Kid der, M.S., A sst. An. Husbandman W. T. Forsee, Ph.D., Assa. Chemist B S. Clayton, B.S.C.E ., Drainage En gine er• F. S. Andrews, Ph.D., Assa. Truck Hort. SUB-TROPICAL STA., HOMESTEAD W. M. Fifield, M. S., Horticulturist Act ini; in Charge s. J. Lynch, B.S . A., Asst . Horticulturist Geo. D. Ruehle, Ph.D., Associ ate Plant Pathologist W. CENTRAL FLA. STA., BROOKSVILLE W F. Ward, M.S .. As st . An. Husband man in Charge 2 FIELD STATIONS Leesburi; M. N. Walker, Ph.D., Plant Pathologist in Charge K. W. Loucks, M.S., Assistant Plant Pat ho l ogist Plant City A. N. Brooks, Ph.D., Plant Pathologist Hastings A. H. Eddins, Ph.D., P lant Pathologist . E. N. Mccubbin, Ph.D. , Assa. Truck Horticultu rist Monticello Samuel 0. Hill , B.S., As s t. Entomologist• Bradenton Jos. R. Beckenbach, Ph . D., Truck Horti culturist in Charge David G. Kelbert, Asst. Plant Pathologist Sanford R. W. Ruprecht, Ph . D . , Chemist in Chante, Celery Inve stiga tion s W. B. Shippy, Ph.D., Asso. Plant Path. Lakeland E. S . Ellison , Meteorologist2 B. H . Moore, A.B., Asst. Meteorologist' 'Head of Department 2 1n cooperation with U.S.D.A. •Cooperative, other divisions, U. of F .

PAGE 3

FERTILIZER EXPERIMENTS WITH POTATOES ON THE MARL SOILS OF DADE COUNTY By \V. M. FIFIELD and H. S. WOLFE* CONTENTS Page Page Plan o f Experim e nts_ _ ______ 4 Sources of Nitrog e n .... 1 7 Fertili ze r An a l y se s 7 Muii at e vs. Sulfat e of P o tash ____ __ ..... 25 Amounts of Fertilizer per Acre ___ ____ ... . 11 l\Iangan e se ----_ _ ...... 25 P e rcent a ge of Nitrogen from Other Soil Amendments 29 Or ga nic Sourc es 14 Sunnnar y ----38 INTRODUCTION The marl soils of southern Dade C ou nt y which are used for potato production are of a peculiar calcar eo us formation quite unlike those of any other soils in the United States. The y are alkaline , ranging from a pH of about 7.5 to about 8.5. Mechanical analysis would classify this soil as a uniform s ilt loam, but chem ical analysis shows over 90 percent of calcium carbonate. Sand is pr ese nt onl y in ver y small percentage, and organic matter constitut es about 5 percent. Thi s marl, a sed imen tary deposit, varies throughout the area from a depth of a few inches to about four feet and is underlaid with a porous limerock. In limited areas muck i s found with the marl, particularly in the numerous small sink or pot hole formations and other low places. The marl is interspersed throughout wit h remains of tiny shells, evidently similar to those from which it originally was deriv ed . It is relatively low in natural fertilit y and in organic matt er, but soil moisture , rising through capillarity from the water table near the surface during the winter season, ordin arily is sufficient for a good crop. Some farms in the area are locat ed on th e higher, fairly well drained soils which are seldom flooded in summer, and others on lower-lying fields subject to flooding during part of th e summer rainy season . Extensive drainage by means of canals and ditches in recent years has r ed uced the danger fr o m sudden floods but, at th e same time , occasionally has contributed to a l ac k of soil moisture in times of drouth. At present the water table ordinarily is found from two to four fe e t below the surface during the major part of the growing season. which in this section extends from November to March. Rotation of crops is not practiced generally in growing po*Formerly Horticulturist in Charge, Sub-Tropical Experiment Station.

PAGE 4

Fig . 2.-Summer growth of weeds on a South Dad e farm. Note growth in relation to height of tractor. tatoes in Dade County, where the same field usually is planted to potatoes every year. However , the fields used for potatoes dur ing the winter months produce a heavy green manure crop of weeds (Fig. 2) or of a legume such as velvet beans (Fig. 3) or . sesbania during the spring and summer. This is plowed under in the fall. PLAN OF EXPERIMENTS The exper iment s described in this bulletin mostly were carried out on the East Glade farm of the Sub-Tropical Experi ment Station, locat ed about six miles east of Homestead on the north side of the Homestead or North Canal. The soil on this farm is typical of the better potato soi ls of the area, bein g fairly high, well drained and about . three feet deep. Other experiments described were performed in other sections of the adjoining area , known as South Allapattah Gardens, where in the 1938 season about 7 , 500 acres of potatoes were planted . Practicall y all of the potatoes now planted in Dade County are of the Bliss Triumph variety, and all of these experiments were performed with it. In general throughout the tests seed pieces were 1 ounces in weight and were spaced 9 inches apart in the rows. The ear lier work was with 36-inch rows, and the later with 38-inch rows. This size and spacing required from 28 to 30 bushels of seed per acre. When the Station was first established in 1930 its equipment, staff and facilities were exceedingly limited and remained so for a number of years. Accordin gly, although nearly all of the growers were using machine planters and fertilizer distributors , the early Station experiments had to be fertilized and planted

PAGE 5

F e rtili ze r Exp e r iments wit h Potato es 5 Fig . 3.-Summ e r gro w th o f ve l ve t b e ans on a potato farm at S o uth Allapattah Gard e ns . Not e high wat e r in th e ditch . b y hand. L ater , as eq uipm ent was obtained, a change to ma chin e methods was a cc om plished . In plan ting and fertilizing by hand, furro ws of about thr ee inche s in dept h were opened wit h a garden tractor , after th e field had been p l owed and d isk ed. The fertilizer was carefull y weighed out and distributed in t h e furrow uniforml y . The fur rows were then closed and in about tw o da ys reopened for plant ing, thereby mixing the fertilizer with the soil. Planting was done b y hand with the aid of marking chains to insure proper seed piece spacing. Four-row plots, each 1/72 acre, were planted b y this method . At harvest time the outside rows of each plot and 5 feet from each end of the two c e nter rows were discarded to eliminate an y possible competitive e ffect between treatments . Th e comparison of treatments has been based on the yields o bt ained from the remainder of the plots (1/180 ac re) after th e d i scarded portions were eliminated. Later , when mach ine planting was adopted, fertilizer was distributed in two continuous narrow bands , approximately at the same level as the seed piece and two inches from it on either side. Th e distributor was adjusted as accurately as possible for

PAGE 6

G Florida Agricultural Experiment Station e ach mixture. A single-row, assisted-feed planter was used. Plots harvested for data with this method each consisted of a single row, 75 feet long (1/184 acre). In all cases border plots were planted around the field, and except where otherwise in dicated buffer rows were planted between treatment plots . Furth ermore, about 80 to 85 feet of row were actually planted, the ends being trimmed off at harvest to eliminate errors due to s tarting and stopping the planter. Variations from these meth ods occurred in some of the cooperative tests. These exceptions are described under headings of the tests involved . All of the fertilizer mixtures used on the Station farm were hand mixed. In order to provide a uniform basis of interpreta tion, all of the test mixtures described in this publication have been converted to a nitrogen basis, according to the 1935 Florida fertilizer law, but early work as reported in the Station's Annual Reports was done with formulas in which the nitrogen was ex pressed as ammonia. Replicate plots of all treatments were used, varying in num ber in different tests, but always arranged in a randomized manner. Time of planting, cultivation and spraying for foliage diseases and insects were uniform throughout each test. The vines were allowed to die before the tubers were dug. All results are reported in terms of yield of No. 1 tubers (in cluding both A and B sizes) per acre. It was found that the yield of No. 1 tubers, which always ran 85 to 95 percent of the total yield, closely paralleled total yields, and, accordingly, rela tive results were the same. Grading was done in later years at the Station farm (Fig. 1) with a standard hand grader equipped with official-sized grading chains. Formerly it was done in the packinghouse of the South Florida Potato Growers' Association. Throughout this bulletin the term "significant" will be used o ften in discussing differences between yields. Where two aver age yields differ only slightly, it is conceivable that the differ ences could be due to other factors than the two fertilizers being tested. Statisticians have worked out methods for use with prescribed plot technique which enable the research worker to determine fairly well whether the difference in average yields obtained is due to the difference in fertilizers being tested, or to other variables. If the differences can be attributed to the fer tilizers they are said to be "significant." If the differences ap parently were caused by variables other than the fertilizers they would not be significant.

PAGE 7

Fertiliz e r Experiments with Potatoes In this bulletin statistical significance was determined by a mean difference in yield being at least 2.2 times as large as its standard error. In a few tests, "Student's" methods were used. The methods used are all described by Love 1 . The statistical computations and constants are not shown in the tables except where publication of the odds by "Student's" method is helpful for interpretation. FERTILIZER ANALYSES The most common potato fertilizer analysis used in Dade County at the time the Station was established (1930) was a 48-52, with about 50 percent of the nitrogen derived from organic sources. Experiments soon were begun using the 4-8-5 as a base or control , to determine if a different analysis would give better results. The first experiments , begun in 1930 on the Station farm , consisted of doubling and halving the percentages of N, P and K, respectivel y, from the amounts in the common 4-8-5 analysis. All fertilizers had 50 percent of the N derived from organic source s, each contained manganese, and all sources of ingredients were the same for each formula . Each mixture was applied at the rate of 1 , 500 pounds per acre by hand and was tested in quadruplicate plots, except in 1930-31 when duplicate plots only were used. The tests were conducted on a field in which pota toes had been grown previously only one, the preceding (1929-30), season . As the work progressed the 4-4-5 treatment was discarded becaus e certain accid e ntal variables occurred which rendered the data valueless. Th e average yields for all r e plications of each treatment ( ex c e pt 4-4-5) are recorded in Table 1 . TABLE 1. A\ 'ER A GE YIELDS OF Ko. 1 TFB E RS (BUSHELS PER AcRE) FR OM THE 19 3 033 FERTILI ZE R ANALY S IS TE S TS. Tr eat ment 1930-31 4-8 -5 140 4-16-5 142 4-8 -2 13 8 4-8 -10 13 6 2-8 -5 138 8-8 -5 115 1931-32 170 177 166 193 172 150 1932-33 200 218 195 209 201 186 Average 170 180 167 180 170 150 2 4 % nitr oge n (N) , 8 % phosphoric a cid (P), 5 % potash (K). 1 Love, H. H. Application of statistical methods to agricultural re search. The Commercial Press, Ltd. 1938.

PAGE 8

8 Florida Agricultural Experiment Station In each of the three seasons the high nitrogen plot (8-8-5) yielded significantly lower than any of the other treatments, indicating that the additional nitrogen was not only a needless expense but actually harmful. The differences in yield between the other treatments were not very great, but the 3-year average indicated a slightly higher yield for the high phosphate ( 4-16-5) and high potash (4-8-10) treatments. On the strength of this, and the fact that the low nitrogen treatment (2-8-5) yielded as well as the standard treatment ( 4-8-5), a new analysis, 3-12-8, was selected as embodying some of the trends indicated by the general results of the test. The cost and source of ingredients of this 3-12-8 were comparable to those of the 4-8-5. The new 3-12-8 analysis then was tested with the 4-8-5 in succeeding years. The first of these tests was conducted on the Station farm in 1934-35. Each analysis was tested at the rate of 1,000, 1,500 and 2,000 pounds per acre, respectively, applied by hand. The yields from four replications of each treatment were averaged, with the results shown in Table 2 , Seri es 1934-35. TABLE 2. -AVERAGE YIELD S 01, No. I Tt l BEJ{ S ( Bl 1 SHEI .S PER .\cRE ) 1-'R 0M THE 193436 FERTILIZER ANA LYSIS TEST S . Series Analy s is Rate p e r Acre (lbs.) Yield p e r Acre 4-8 -5 1 , 000 167 3-12-8 1,000 171 1934-35 4-8 -5 1,500 199 3-12-8 1,500 186 4-8 -5 2,000 194 3-12-8 2 , 000 197 Estes 3-12-8 2 , 000 248 4-8 -5 2 , 000 237 3-12-8 1,500 279 1935-36 4-8 -5 1,500 278 3.=.12-8 2 , 000 289 4-8 -5 2,000 287 Analysis of the data showed that there was no significant difference in yield between the two analyses, even though at the 1,500-lb. rate the 4-8-5 showed a little higher average yield than the 3-12-8 . Another test was conducted the same year on the commer cial farm of J. l-I. Estes. About an acre was planted (by machine)

PAGE 9

F crt ilizcr Exp e riments with Potatoes 9 with each analysis at the rate of 2,000 pounds per acre, and from each acre 20 representative plots were harvested at random. The y ields from the 20 plots of each treatment were averaged with the results shown in Table 2 , Series Estes. In this te st a slight but not significant increase in average yie ld was noted for the 3-12-8 treatment. Another test was made the following year (1935-36) on the Station farm. The 3-12-8 and 4-8-5 analysis were tested in quad ruplicate plots at rates of 1 , 500 and 2 , 000 pounds per acre , applied by machine. Th e average yie lds are given in Table 2, Series 1935-36. Again, no significant difference in yield was obtained between the two analyses. Since it was apparent that no appreciable differences in yield were being obtained from analyses varying as widely as a 4-8-5 and a 3-12-8, it was thought wise to test differences in analysis on a more intensive scale. Accordingly, a new series of tests was instituted on the Station farm in 1937-38, in which the percent ages of phosphoric acid and potash in a basic 4-8-4 formula were increased progressively , alone and in combination. The nitrogen was not varied, but kept constant throughout at 4 percent, of which half was derived from organic sources. Data were secured from six replications of each of the 12 treatments, planted by machine. All fertilizer was applied at the rate of 1.500 pounds per acre, and all of it contained man ganese. The average yield from each treatment is given in Table 3, for each year. TABLE 3.-AvrnA Gr. Y11::rn OF :--..o. 1 Tl'BERS (BusHu s PER ACRE) FRO:I.I THI-: TREAT ::rExTs Co::,rPOSTXG THI-: FER TILIZER ANAL\' SIS T ESTS, 193739 . Analysi s 1937-38 1938-39 4-8-4 330 174 4-8-6 331 169 4-8-8 334 176 4-8-10 333 171 4-12-4 342 172 4-12-6 327 179 4-12-8 339 182 4-12-10 331 168 4-16-4 329 169 4-16-6 323 174 4-16-8 336 172 4-16-10 341 179 The y ield s were extremely high the first year of this test and fairly low the second (a dry) year. Each year, however, a sur prising similarity in yield was obtain ed from all 12 treatments,

PAGE 10

Fig. 4.-Portion of th e 1937-38 potato fertilizer analysis tests. Not e uniform vine growth among all plot s. and the small differences in yield between them were found not to be significant. In other words, no significant increase in yield was obtained from increasing the phosphoric acid beyond 8 percent, or from increasing the potash content beyond 4 percent . No differ ence was noted in vine growth or quality of tubers between treatments either year (Fig. 4). It should be noted that this last experiment was conducted on land which had been fertilized and planted in potatoes for five successive preceding yea r s. Since the authors have noted indications that some fertilizer residue tends to carry over from one year to another in these marl soils, there is a possibility that sufficient residual phosphate and potash were present to offset additional quantities added in the higher analyses. Sinc e the lowest analysis tested ( 4-8-4) yie lded as well as the highest ( 4-16-10), it would seem desirable to test even cheaper formulas , especially in view of the good results from the 2-8-5 and 4-8-2 analyses as shown in Table 1, on land with no previous histor y of potato culture. These analysis tests , conducted over a period of years, in dicate that there seems to be no justification for using analyses ,

PAGE 11

F er til iz er Erp c rim c nts itlt Potato es 11 to be applied at rates of about 1,500 to 2,000 pounds per acre, with more than 3 or 4 percent nitrogen, 8 p e rcent phosphoric acid a nd 4 or 5 perc e nt potash. AMOUNTS OF FERTILIZER PER ACRE The rate at which fertilizer should be applied per acre is, of course, influ e nced greatl y by the analysis of the fertilizer. At the time the se experimen t s were begun a 4-8-5 analysis was the one most c o mmonl y in u s e . Therefore, it was us e d exten s ively for the b a sic test s . A few other analyses were used from t im e to tim e. The test s were distributed over a number of ye ars and ov e r a numb er of different farms, in order to get as c omprehensive data as p os sible . The first t e st was p e rform e d on the Station farm in 1933-34 . Th e fertilizer w a s distribut e d in the furrows by hand and mixed with the soil two days b e fore the seed pieces were planted . Four treatments were used , 500, 1,000 , 2,000 and 3,000 pounds per acre , re s pectively , of a 4-8-5 anal y si s, in which two-thirds o f the nitrog e n w a s from o rganic sources . Four replications o f each tre a tment were plan te d. The av e rage y ield s from each t reatment are shown in Tabl e 4, Seri e s A. The 2,000-pound rat e y i e lded high e r than an y o f the other treatments . It y ielded enou g h more than the 1 , 000-pound rate to justif y the additional c o st . Th e reduced y ield of the 3,000-pound rate was probabl y due to in j ury of the potato sprouts, as it was not e d that th e s e plots sh o w e d slow e r e mergence than did the o th e rs. This injur y might n o t have occurred had the fertilizer been distribut ed b y the m a chine band method , o r had it been a pplied in th e furro w s e arli e r befor e planting . Another t e st was conducted on th e Station farm the same year in conn e ction with another experiment. A 4-8-5 was also us e d in this t e st , but in on e s e ries 33 p e rc e nt of its nitrogen was d e rived from o rganic s o urc e s , and in the other 66 percent was s o derived. Each mixtur e w as applied b y hand at both 1,000 pound and 2 , 000-pound p e r acre rat e s . The plot arrangement was similar to that of th e previous test. The replicate yields we re averag e d with th e r e sults given in Table 4, Series B. With both t y pes of mi x ture , the 2 , 000-pound rate signifi c antly outyielded the 1,000-p o und rat e . A comparison of the 33 percent and 66 percent mixtures will be presented later (Table 7 , Series 1933-34).

PAGE 12

12 Florida Agricultural Exp e riment Station In another test on the Station farm in 1933-34 nitrate of soda was compared with sulfate of ammonia as sources of inorganic nitrogen in a complete 4-8-5 analysis, and muriate of potash with sulfate of potash as potash sources in the same analysis. Each treatment was tested at both 1,000 and 2,000 pounds per acre, in triplicated plots, with the results given in Table 4, Series C. TABLE -+.-A v ER . \GE Yn:w s or No. 1 TunERS (BusRELS PER AcRE) FOR VARIO US RATES OF APPLIC\T IO X F OR 4-8-5 FERTILIZER, 1933-34. Series Rate per Acre (lbs.) A B C 500 1,000 2,000 3,000 1,000 2,000 1,000 2,000 1,000 2,000 1,000 2,000 1,000 2,000 1,000 2,000 F e rtilizer Variable Standard mix Standard mix Standard mix Standard mix 33 % Organic N 33 % Organic N 66 % Organic N 66 % Organic N Yields per Acre 176 203 222 202 184 197 203 222 Inorg . N as Nitrate of Soda 165 Inorg. N as Nitrate of Soda 164 Inorg. N as Sulfate of Ammonia 181 Inorg. N as Sulfate of Ammonia 197 Potash all as sulfate 204 Potash all as sulfate 222 Potash all as muriate 207 Potash all as muriate 209 These data are not harmonious. The nitrate of soda plots showed no increase in yield for the higher rate of application . The sulfate of ammonia showed a statistically significant in crease. The sulfate of potash plots also showed such an increase for the 2,000 pounds application, but the muriate plots did not. The comparison of these sources of nitrogen and potash as they in themselves affect yields will be discussed later. The following year (1934-35) the first test was repeated in revised form on the Station farm. The 500-pound rate was eliminated and a 1,500-pound rate was added. Also another analysis, 3-12-8, was added, and this was tested at 1,000, 1,500 and 2 , 000-pound rates . In both mixtures the nitrogen was de rived two-thirds from organic sources. Four replications were used and the fertilizer again was applied by hand. The results from Table 2 are given again for convenience in Table 5, Series 1934-35. The results indicate that with both analyses the average yields of the 1,500-pound rate were about as good as those of the 2 , 000-pound rate. The higher yield of the 2 , 000-pound rate

PAGE 13

Fertilizer Exp e rim e nt s , i •itlt Potatoes 13 of 3-12-8 was found, upon analysis of the replicate plot data , to be not significant. In the 4-8-5 test the 3,000-pound rate out yielded the lower rates, and with both analyses the 1,000-pound rate was outyielded by the higher rates. This same year (1934-35) another test was carried out, this time on the commercial farm of J . M. Holferty. A 4-8-5 analysis (50% of nitrogen from organic sources) was used, applied by machine . Approximately one acre was planted at each of two rates, 1 ,300 and 2,000 pounds per acre. At digging time 10 plots, each 100 feet long, were selected at random for harvest from each block. The No. 1 tubers were graded from each replicate plot and averaged , with results given in Table 5 , Series Holferty A. In this instance, the 2 , 000-pound rate significantly outyielded the 1,300-pound rate. In another field of the same grower, another similar test was performed two weeks later. The results are found in Table 5, Series Halferty B. This time , although the 2,000-pound rate showed a slight increase in average yield over the 1,300-pound rate, the difference was not statistically significant and thus ap plication of the additional 700 pounds of fertilizer was not justified . Another test was performed on the Station farm in 1935-36 , comparing 1 , 500 and 2,000-pound rates of both a 4-8-5 and a 3-12 -8 analysis (each with nitrog en 66 ~;-, from organic sources). Six replications of each were planted. This time the fertilizer was applied by machine. The data were averaged for the various replications, as shown in Table 5, Series 1935-36. T.-\BLE 5.-.-\n:RAG Ymr.Ds OF No. 1 TUBERS (BUSHELS PER ACR E) FOR V . .\Rl01;S RATE S OF FERTILIZER APPLICATIO N , 19 3 436 . Seri es Rate per Acre (lbs . ) Analysis Yields per Acr e 1,000 4-8-5 167 1 , 500 4-8-5 199 2 , 000 4-8-5 194 19343 5 3,000 4-8-5 226 1,000 3-12-8 171 1,500 3 -12-8 186 2,000 3-12-8 197 Half e rty A 1,300 4 8-5 179 2,000 4 -8-5 209 Holferty B 1,300 4-8-5 212 2,000 4-8-5 224 19353 6 1.500 4-8-5 278 2 . 000 4-8-5 287 1.500 3-12-8 279 2 , 000 3-12-8 289

PAGE 14

14 Florid a Agricultural Experim e nt Station Here, with both analyses, the 2,000-pound rate slightly out yielded the lesser rate in average yield but the differences were not great enough for statistical significance. Therefore, the ex pense of the extra 500 pounds of fertilizer was not justified. In 1937-38 still another Station farm test was made of the 1 , 500 and 2 , 000-pound rates, in connection with testing 4-8-5 mixtures in which 33, 50 and 66 percent, r e spectively, of the nitrogen was derived from organic sources . The fertilizer was applied by machine. Four replicate plots were planted, and their yields averaged , as given in Table 6. TABLE 6.-A VE R AGE YI ELD S OF No . 1 T UBE RS ( B US HE LS PER A CRE) F OR VARIO US R ATE S OF A PP LICATI ON OF A 4 8-5 F ER TILIZ ER , 193738 . R a te p e r Acre (lbs.) 1,500 2 , 000 1 , 500 2,000 1,500 2,000 % of N from Organic Sourc e s 33 33 50 50 66 66 Yi e lds p e r Acr e 302 305 304 311 289 296 Again in this t e st a slightly increas e d average yield was ob tained from the heavier fertilizer application but the increase was too small for statistical significanc e . It is apparent that no single rate of application will give the same results in all s e asons or under all conditions in th e sam e season. The data of these e xperiments generally agr e e, h o wever, that 1,000 pounds of 4-8-5 per acr e is probabl y not sufficient for best results , and that 3,000 pounds is probably too much. There is more definite agreement that th e correct amount is from 1 , 500 to 2 , 000 pounds per acre. In the great majorit y of instance s no significant yield d iffer e nces were obtained betw ee n thes e two amounts , and certainly , that being th e cas e, the 1,500-pound rate would b e the more profitable for r e gular us e. In a season in which higher prices than usual might be expected, or in an unusuall y wet s e ason , or on so il m o re moist th a n that occur ring on the Station ' s E a st Glad e farm, it i s possible that the 2,000-pound rate would b e more profitable. It should be pointed o ut h e re th a t the rat e s of appli ca tion discuss e d ref e r onl y to th e generally us e d anal y ses o f 4-8-5 a nd 3-12-8, and not to mixtur es containing higher or lower conc e ntrations of plant food . PERCENTAGE OF NITROGEN FROM ORGANIC SOURCES Since th e percentage of nitrog e n d e rived from organic sources mark e dly affects the expense of mixed fertilizers, ex

PAGE 15

F e rtili ze r Exp e riments with Potatoes 15 periments were begun in the fall of 1933 to study the effect on yields of varying this percentage in a standard 4-8-5 analysis. The first year's test was limited to a comparison of 33 and 66 percent of the nitrogen so derived. All sources of ingredients of the two fertili z ers were similar , and both included manganese . Nitrogen was derived from cottonseed meal, fish scrap and sulfate of ammonia, phosphate from superphosphate, and potash from sulfate of potash. Each treatment was replicated three times and the fertilizer was applied by har,rl. Each mixture was applied at both 1 , 000 and 2 , 000-pound rates per acre. The re plicate plot yields were averaged to obtain the data shown in Table 7, Series 1933-34 . In this test, the 66 percent treatment out y ielded the 33 percent treatment at both rates of application and significantly enough to offset the additional cost of the 66 percent mixture. TABLE 7. -. -\YER.,G E YILLDS O F ;,.:;o. 1 TcB E R S (B US HELS PE R AcRE), RESULTING FRO~ I l 'S E O F A .\S 5 FERTI LIZE R WIT H VAR Y I NG PER CE N"TAGES OF NITR O GEN DERIV E D FRO :\ ( 0RGA X I C So tR CES. % of N from Rate per Acre S e ri es Organic Sourc e s (lbs . ) Yi e lds per Acre 33 1,000 184 19 3 3-34 66 1.000 201 33 2 . noo l q7 66 2,000 222 -1934-35 3 3 1.500 208 66 1,500 199 -----33 1.500 276 1935-36 50 1.5'10 280 66 1,500 278 3 3 l.fiOO Hl4 1936-37 50 1.500 184 66 1.500 182 ----33 1.500 302 50 lfi"l() ::104 19 3 7-38 66 1.500 289 33 200') ~05 50 2.000 311 66 2.000 296 The following year (1934-35) both percentages were again tested but this time onl y at 1 , 500 pounds per acre. Other factors, including application of the fertilizer by hand, were the same as in 1933-34. The average yields are given in Table 7, Series 1934-35. These y ields are not significantly different, and since the 33 percent mixture was cheaper, it was therefore more profitable .

PAGE 16

16 Florida .-lgri c ultural E:rpcri111 c nt Station The following year (1935-36) the same tests were repeated with similar procedure, except that the number of replications was increased to 10. In addition, a 50 percent treatment was included. The average yields are given in Table 7. Here again the yields were not significantly different, and the 33 percent treatment again was most profitable. In the 1936-37 season the same experiment was repeated with technique similar to that of the previous year. The average yields are shown in Table 7. Again this year the 33 percent treatment yielded as well as the others and was therefore most profitable. All treatments were repeated in the 1937-38 season, this time each mixture being applied at both 1,500 and 2,000-pound rates per acre. Also, this year the fertilizer was applied by machine, with eight replications of each plot. Average yields are given in Table 7. Again this year , and at both rates, the 33 percent treatment yielded as well as either of the others, and therefore again was most profitable. Thus, in four of the five years of testing, the 33 percent mixture yielded as well as the mixtures containing a higher percentage of N derived from organic sources. A factor considered most important in selecting the more slowly available sources of nitrogen, and the percentage of them used in the fertilizer, is amount of rainfall. Therefore, for com parative purposes, the rainfall data including these five years , as recorded officially at the Sub-Tropical Experiment Station for the respective crop seasons, are given in Table 8. There seems to be no definite relation of treatment results to rainfall during the seasons in which these tests were con ducted. The first season (1933-34) higher yields were received from the 66 percent mixture. That year an unusually heavy rainfall occurred in October, but the potato plots were not planted until December 8, so that there is not much likelihood of the October rain exerting much influence on the crop. In the 1935-36 season the plots were planted December 2. The season from planting to harvesting (March 18) was even wetter than in 1933-34, and yet the 33 percent mixtures yielded as well as those containing higher percentages of N derived from organic sources. These tests were conducted over seasons which varied con siderably in amount of rainfall, so that it seems quite logical to conclude from them that ordinarily, with other factors equal,

PAGE 17

F e rtili ze r Experiments with Potatoes 17 mixtures containing no more than 33 percent of their nitrogen from organic sources should give as satisfactory results as those of higher percentages thus derived. Of course, these conclusions are drawn from experiments in which the organic sources of nitrogen were limited to two typical water-insoluble organics, fish scrap and cottonseed meal. TABLE 8. '.\foXTIILY .~XD TOTAL R ,\IXHLL ( IX INCHE S ) RECORDED DURl:s!G THE CR OP SE. ~S OXS F R(J:\ 1 )0 3 2 T O 19..\0, Ho :-.rES TE .-\D, FLORIDA.* Total October November Dec ember January February March (6 months) 1931-32 193233 1933-34 193435 1935-3 6 19 3 6-37 1937-3 8 1938-3 9 1939-40 6.95 12.90 22 .9 5 3 . 22 7.41 3.3 7 7.41 4.41 16 . 79 1.85 4.15 1.39 0.37 5.51 3.5 2 0 .3 7 2 . 44 0 . 95 0.26 0 . 81 0 . 33 0.41 0 . 69 0.89 0.55 0.71 1.80 3.21 1.39 2.00 0.23 2.50 0.65 2 . 69 0.93 1.02 0.47 0.20 0.94 0.25 4.59 2 . 70 0.81 0.56 2.44 0.92 3.05 2.13 0 . 5 3 4.83 4.41 1.72 0.09 2.93 13.66 22 . 50 29.74 5.01 25.53 15.54 13 . 55 9.14 25.93 Non e of the potato tests describ e d in this bulletin was plant e d b e for e No vem b e r 10 in any y ea r. Additional experiments ,Yer e conducted on sources of nitro gen, in which mixtures containing all water-soluble nitrogen were compared with those containing 50 percent of their N from numerous organic sources. These are described below. SOURCES OF NITROGEN Thes e investigations were begun in the fall of 1933, when a comparative study of sulfate of ammonia and nitrate of soda , as inorganic sources of nitrogen in complete fertilizer, was started . The first series of trials was conducted over a period of four years. A 4-8-5 fertilizer was used throughout. The first three y ears it was applied in the furrow by hand and the last year b y machine. The first yea r it was applied at the rate of 2 ,0 00 pounds per acre. In e ach of the next three years it was applied at 1 , 500 pounds per acre . The nitrogen of the fertilizer was derived 25 percent from fish scrap, 25 percent from cottonseed meal, and 50 percent from either sulfate of ammonia or nitrate of soda, according to the mixture tested . In other words, 50 percent of the nitrogen was derived completely from one or the other of these two inorganic sources. The first ye ar (1933-34) three replications of each treatment were planted. Th e second (1934-35) year, four, the third

PAGE 18

]8 Florida Agricultural Experiment Station (1935-36) year, 10, and the fourth (1937-38) year, six replications were used. The tests were not planted in 1936-37. The repli cate plot yields for each treatment were averaged each year and the average annual yields are given in Table 9. TABLE 9.-AVERAGE YIELDS OF No. l TUBERS (BUSHELS PER ACRE) FROM FERTILI ZE R TREATMENTS COMPARING INORGANIC SO URCE S OF NITROGEN IN COMPLETE FERTILIZERS. Source 1933-34 Nitrate of soda 164 Sulfate of ammonia 197 1934-35 191 199 1935-36 284 278 1937-38 305 292 Average 236 241 There was no appreciable difference between the two treat ments except in the first year, when sulfate of ammonia yielded better than nitrate of soda. Just why this difference should have occurred the first year is not explainable on a rainfall basis, or on any other readily apparent basis. At any rate, since the four-year average indicates no significant difference in yield between the two, and since the first year's data were in favor of sulfate of ammonia, it is logical to conclude that this is probably the better source for these soils, especially since its cost per unit of nitrogen was lower than that of nitrate of soda. A more comprehensive test of nitrogen sources was begun in the fall of 1937. A number of water-soluble and water-insolu ble sources were tested in a 4-8-5 complete fertilizer, with man ganese added. The source of phosphoric acid (superphosphate) and potash (sulfate of potash) remained the same for all treat ments, except for allowances made for the small quantities of these elements carried by some of the organics. Although most ordinary commercial potato fertilizers contain a number of organic sources of nitrogen , it was decided arbi trarily in this experiment to test the sources singly, deriving half the nitrogen from sulfate of ammonia and half from the organic source being tested. The materials thus tested were cottonseed meal, fish scrap, blood-and-bone tankage, dried blood, milorganite and urea. In addition, a cyanamid test was included, but since 50 percent of cyanamid nitrogen was considered too much, the nitrogen in this treatment was derived 20 percent from cyanamid and 80 percent from sulfate of ammonia. Five other treatments were included. In three of these, all (100 per cent) of the nitrogen was derived from sulfate of ammonia, nitrate of soda or ammonium phosphate, respectively. In the

PAGE 19

Fertilizer Experiments with Potatoes 19 fourth, the regular Station control 4-8-5 formula was used in which 25 percent of the nitrogen was derived from cottonseed meal, 25 percent from fish scrap and 50 percent from sulfate of ammonia. The nitrogen in the last treatment was derived half from nitrate of soda and half from sulfate of ammonia. Thus in all there were 12 treatments. Urea and cyanamid are synthetic organics which resemble the inorganic nitrogen sources more closely than they do the natural protein organics, in availability to plants and in solubility in water. They may be distinguished for our purposes as water soluble organics, in contrast to the older water-insoluble types like cottonseed meal and fish scrap. The composition of the 12 fertilizer treatments as they were mixed at the Station is given in Table 10. All of the fertilizers were applied at the rate of 1,500 pounds per acre with a machine distributor. The first year eight repli cations of each treatment were used. In the second and third years the number was reduced to six . Each plot harvested for data consisted of one row 75 feet long. The rows were spaced 76 inches apart (twice the ordinary distance), but in converting the yield data to an acre basis , a plot was considered arbitrarily as 75 feet x 38 inches or 1/184 acre. This departure from ordinary plot technique deserves brief discussion . In placing the rows twice the ordinary distance apart the object was simply to eliminat e planting of the buffer rows. Satisfactory evidence from previous observations indicated that there would be no competitive effect between roots and fertilizers in adjacent plots . In converting the data to an acre basis it was recognized that while the plots actually were 75 feet x 76 inches, from a practical standpoint conversi o n on this basis would yield values far below normal commercial yields. and this would be misleading , since the actual yields per plant w e re quite normal. As the object of the test was to determine m e rel y the relative effect of the various tr ea tments, and not the absolute value of each, the only sensible procedure was to con v er t the y ields on a normal basis of 38-inch rows. The fertilizer and seed . as \\ ell as the y ields . were computed on the latter basis. The average yields for each y ear are given in Table 11. A primary consideration in the selection of nitrogen sources for potato fertilizers has been their performance in relation to soil moisture. Growers usually consider that the water-insoluble

PAGE 20

N 0 TABLE 10.-CO M P OS ITI ON , I X P Ol:N DS PE R T ON, OF THE F ERT ILI ZERS CoM P O SDIG TH E S OURCE OF N IT RO G EN TEST S , 1 93 7-40 . Q) ctS ctS a .i:: a., 0. .s:: ..., 0 "' "' ctS s ctS 0 ctS a., ctS .... a., a., 'O .i:: a., ;:I Treatment, with the Pers ctS 0 ctS 0 0.. ctS r./). .i:: centage N each Sourc e Con ctS 0. a., 0.. i:1 I r.110 O'.l9 0JJO roo ctS :i... ....,0 -o o , 0 ~ 'Clo 0 .. ' .. 0 -c:o 0J) ctS ' ctS I So I ctS' = a., ' ctS-.:t< "" ,i:: M ~"'!' Q) I OIN ~ "" <+-... -, ..., ' I Ul I .. I .. I ' :>, ' ;:::1"'1 -o o:;:: <', ;:I,... -~It:) 0,e.) -It:) It:) ~r:;:I' r./).oN z ......
PAGE 21

F e rtili ze r Exp e riments with P o tato es 21 TABL E 11.: hERAGt YIE L IJ S OF ::--u. 1 Tt " B ER S ( Bi; S HELS PER ..\CRE ) l ' R011 TH E VARI OG S S O URC E O F-:\ITRO GE X TREADI EX T S , l'SIX G A +8 -5 ANALY SI S AT 1,500 P ou x os P ER A CRL Fertilizer Nitrogen S o ur ce Co s t 19 3 7-38 1938-39 1939-40 Av i:: rage 1 , 500 Pound s 100 % from sulfate of ammonia $17 . 96 2 3 7 144 249 210 100 % from nitrate of sod a 1 9 .35 240 144 246 210 50 % from nitrate of sod a ''' 18.65 242 149 239 210 100 % from ammonium phosphate 22.38 242 151 247 213 50 % from cottons ee d m e al"' 21.84 256 145 241 214 50 % from fish scr a p ' ' 22.64 246 150 239 212 50 % from blood-and-bone tankag e' ' 22 . 55 263 150 257 223 50 % from dried blood ' ' ' 23.16 256 148 255 220 50 % from urea" 18 . 37 249 155 238 214 50 % from milorganite '' ' 22.33 246 170 255 224 20 % from cy a namid''' 18.35 252 145 237 211 25 % from fish scrap , 25 % c o ttons e ed meal ' ' 22.24 245 154 236 212 ,:, Balanc e of N from sulfate of ammonia. sources are best in a wet year. Referring back to Table 4 and to the seasons 1937-38 through 1939-40, it will be noted that these seasons were quite variable with r e gard to rainfall. The 1937-38 season may be considered as intermediate, with a total of 13.55 inches. The 1938-39 year was a dry one, with only 9.14 inches , as was reflected in the lower yields. The 1939-40 season was an unusually wet one, with rainfall well distributed throughout the growing season . Thus it ma y b e considered that the test was run under conditions which varied considerably from dry to wet. With all of this variation , the yields did not vary much among the treatments, and therefore the experiment was termi nated with the 1939-40 test. The results first will be discussed b y indi v idual years . rt was found that a minimum difference in yield of 7 bushels per acre was necessary for significance in the 1937-38 data. There fore, the treatments sulfate of ammonia, nitrate of soda, the combination of these two, ammonium phosphate, fish scrap, milorganite and the fish scrap-cottonseed meal combination as a group may be considered as all having yielded about alike. Slightly higher y ielding were urea and cyanamid and a little higher yielding than these were tankage, dried blood and cotton seed meal. The yields of all 12 treatments were so nearly alike , however , that fine distinctions in yield differences are hardly justified. It was found that at least a 6-bushel difference was required for significance in the 1938-39 data. On this basis, all of the

PAGE 22

22 Florida Agricultural Experiment Station treatments except urea, milorganite and the fish scrap-cottonseed meal combination yielded about alike. The latter three may be considered as higher yielding than some of the others, but only the milorganite significantly outyielded all of the others. In 1939-40 a 7-bushel difference was found necessary for significance. This year milorganite, dried blood and tankage stood out as being significantly highest yielding, although minor significant differences occurred between some of the other treatments, and a few-sulfate of ammonia, nitrate of soda and ammonium phosphate-yielded almost the same as these three water-insoluble organics. From this brief consideration of the yearly data it will be noted that while certain trends in yield differences appear, the differences in most cases are so small as to leave some doubt as to their practical application. Conclusions can best be drawn from the data secured by averaging the treatment yields over the three-year period. For these data (last column, Table 11) it was found that a difference of more than 4 bushels per acre was necessary to establish significance. On this basis the treatments sulfate of ammonia, nitrate of soda, the sulfate of ammonia-nitrate of soda combination, ammonium phosphate, cottonseed meal, fish scrap, urea, cyanamid and the fish scrap-cottonseed meal combination all yielded about alike. Yielding slightly but significantly higher were milorganite, tankage and dried blood. In order to help determine the relative profitableness of the treatments, the fertilizer cost per bushel of potatoes produced was determined from these data for each season. The costs are shown in Table 12. TABLE 12.-FERTILIZER COST PER BUSHEL OF POTATOES FROM THE VARIOUS SOURCE UF-c'JJTRO'.'EN TREATMENTS. (ARRANGED IN ORDER OF AvERAr:E CusT BASED o" .'..\"ER.\'.:E YIELD FOR THREE YEARS.) __ N_1_ t_ro_g~e_n_S_o_u_r_ce _____ ~l 937-38 1938-39 1939-40 Average 50% from urea* $ .073 $ .119 $ .077 $ .086 100% from sulfate of ammonia .076 .125 .072 .086 20% from Cyanamid* .073 .127 .077 .087 50% from nitrate of soda* .077 .125 .078 .089 100% from nitrate of soda .081 .133 .079 .092 50% from milorganite''' .091 .131 .088 .100 50% from blood-and-bone tankage* .086 .150 .088 .101 50% from cottonseed meal* .085 .151 .091 .102 50% from dried blood* .090 .156 .091 .105 25% from fish scrap, 25% cottonseed meal* 100% from ammonium phosphate 50% from fish scrap* .091 .092 .092 * Balance of N from sulfate of ammonia. .153 .144 .151 .094 .091 .095 .105 .105 .107

PAGE 23

Fertilizer Exp e rime11ts with Potato e s 23 Each year, and for the three-year average, the fertilizer cost of producing a bushel of No. 1 potatoes was less from the water-soluble sources-urea, sulfate of ammonia, cyanamid and nitrate of soda-than from the other sources. Since it has already been shown that cottonseed meal , am monium phosphate, fish scrap and the fish scrap-cottonseed meal combination were significantl y out y ielded by milorganite , dried blood and tankag e, there is no question but that the latter thr e e were mor e profitable than these others , which cost about the sam e . In comparing milorganite , dried blood and tankage with the cheaper sources-urea , sulfat e of ammonia, c y anamid and nitrate of soda-how e v e r, a little more difficulty occurs . Considering the first three treatments as one group and the latter four treatments as another group , by averaging the data for each it is calculated that the first , or insoluble materials , group required $4 . 17 per acre more of fertilizer cost to produce an additional 11 bushel s of No . 1 potatoes. These potatoes had an undug field value of about 82 cents p e r bushel according to Howard and Steffani 3 , or $9 . 02 for the 11 bushels. Thus the extra $4.17 cost for fertilizer of the first group return e d an incr e ased yield worth $ 9.02, as compared with the second group , or a dif ference of $4 . 85 per acre in favor of the more insoluble organic nitrogen group. The above calculations are pr e s e nted for the benefit of those who are interest e d in knowing just what the profit and loss of the various treatment group s were under the conditions in which the experiments were performed . From a broader and more practical viewpoint , however , the authors believe that d e finite conclusions o n the basis of these particular "cost and returns" data alone are not w a rranted, because cond i tions are almost never the same in any two years. Fertilizer costs, yields and r e turns vary considerably. Furthermore the calculations are based on average yields, which in themselves are only sig nificant within certain limit s, as already pointed out. Also it must be r e cognized that as y ields incr e ase on an acr e of land, the costs per bushel for land r e nt , machinery use, spr a y mater ials , s e ed , l a bor and sup e rvision d e crease, which probabl y would more than off s et small i ncreases in fertili z er c o st. It would seem more sensible, therefore , to consider the re3 Howard , R. H . and C . H. St e ffani . A studv of potato farming in Dade County, Florida, seasons 1934-35 to 1938-~9. Fla. Agr. Ext. Ser. Mimeo. Circ. D e pt . Agr. Econ. P o tat o es AE4 . 1939 .

PAGE 24

24 Florida Agricultural Exp e riment Station sults in more general terms. From this standpoint it is concluded that in general as used in this experiment, the fertilizer sources milorganite, blood-and-bone tankage (medium grade) and dried blood slightly outyielded the other sources. The differences in yield, although small, were profitable for the three-year period based on cost and returns data cited but in certain of the years some of the water-soluble sources were equally as profitable. It would seem wise to continue the use of either milorganite, dried blood , or blood-and-bone tankage in fertilizers for potatoes in Dade County for the present. There is no evidence to indi cate that any one of these three is better than the other. Since prices fluctuate from time to time, whichever one is cheaper per unit of nitrogen at the time of purchase is the one to choose. Sulfate of ammonia and nitrate of soda yielded alike, as sole sources of nitrogen, and no advantage was demonstrated from combining the two on a half and half basis. Price differential per unit of nitrogen would accordingly determine the preference for use in potato fertilizers in this area. Ammonium phosphate was the least economical of the water soluble nitrogen sources tested . Although its performance was quite satisfactory, its use cannot be recommended because of its higher cost . It is possible , however, that it might be a desirable source for use in high analysis fertilizers, which were not in cluded in the scope of this test. The use of urea and c y anamid is worthy of consideration. They both produced yields in the upper range of the water-solu ble sources, and both sources were among the treatments show ing least fertilizer cost per bushel. Since many growers are of the impression that ur ea leaches l ess readily from the soil than do some of the other water-soluble sources, this wo uld give it an apparent advantage in wet seasons , especially if an entirel y water-solubl e fertilizer is desired. No data are available for Dade Count y soils on th e leaching properties of urea. Th e relatively good showing of the 20 percent cyanamid treatment is noteworthy . Some fertilizer manufacturers would like to use a sm all amount of this ingredient in their mixtur e s because of its conditioning effect. The data of this experiment indicate th a t there should be no objection to this practice so far as potato f e rtilizers for these soils are concerned. The p e rformanc es of cottonseed meal and fish scrap indicate that the y are l ess desirabl e than the other water-insoluble sources tested. and, because of t heir present higher cost, show no econ omical advantage over the water-soluble sources with which thev were also compar ed.

PAGE 25

F e rtilizer Expcrimrnts ; ci th Pot a toes 2 5 MURIATE VS. SULFATE OF POTASH There has been a great deal of debate among growers as to the relative merits of muriate and sulfate of potash for potato fertilizers. A series of experiments was begun in the fall of 1933, testing these two ingredients. Each was used as the sole sourc e of potash in a regular 4-8-5 fertilizer, with manganese. The rest of the mixture in each case was composed of sulfate of ammonia, fish scrap , cottonseed meal and superphosphate. The first two years, four replications of each treatment wer e planted. The third year 10 replications and the last year six replications were used. The first three years the fertilizer was applied by hand. The last year it was applied by machine. The tests were not conducted in 1936-37. The first y ear a ton per acre was used, but in each of the other three years the fertilizer was applied at the 1 , 500 pound rate. The plots were on different ground each year. The replicate plot y ields each yea r wer e av e raged and the annual average y ields from each treatment are giv e n in Table 13. TABL E 13.-A YER AGE YIELDS OF ::--:o. 1 Tl-BLR S I BrSII EL S PER ACRE) FRO.'.I I P LO T S COIII PA R I XG TH E :'lfl Hr..\T E . \ X D Sl ' L L .\TE FOR:\[S A S SO l' R C E O F P o nsrr I X A 4 8 -5 F ERTILI Z ER. Sourc e Sulfat e of p otas h Muri a t e of p ot ash 193334 19343 5 19 3 5 -3 6 1 93 7-38 Aver age 222 209 19 9 189 278 2 8 1 298 301 249 245 The data of Table 13 indicate onl y slight differ e nces in yield of tubers between the two sources of potash, and these differ ences were found not to be statisticall y significant. The findings of this experiment indicate that either source may be used with equal results. It should be noted that ea ch yea r thes e tests were carried out in a different field . Just what effect would be had by using eith e r source continuall y on th e same field remains to be de termined. MANGANESE As a result of the pioneering work of Skinner and Ruprecht 4 and of leading growers, most of the potato growers in the Homes stead area were using manganese sulfate in their fertilizers by the time th e Sub-Tropical Experiment Station was established 4 Skinn e r, J. J .. and R. W. Ruprecht. Fertilizer ex perim e nts with truck crops. Fla. Agr. Exp. Sta. Bu!. 218 , 1930.

PAGE 26

26 Florida Agricultural Exp erime nt Station in 1930. The usual amount of manganese sulfate added was 200 pounds per ton but some growers were using as much as 400 pounds p er ton. Experiments were started by the Station in the fall of 1931 to determine the effect on yields of lesser amounts of manganese, and of the residual effect of manganese. RATE OF APPLICATION A seri es of plots was laid out on the East Glade farm to be continued over a period of years. Four manganese treatments (65 percent manganese sulfate) were included: 0 lbs ., 50 lbs., 100 lbs. and 200 lbs. per ton of 4-8-5 fertilizer, respectively. The fertilizer was applied at the rate of 1,500 pounds per acre and was distributed in the furrow by hand. The test included three repli cate plots of each of the four treatments. Average yields from each treatment are given in Table 14. TABLE 14. YIELDS OF No. l TUB ERS (BUSHELS PER A CRE) RESULTING FROM VARIO US RATE S O F APPLIC'ATI ON OF MA NG ANESE SULFATE PER TON OF 48 -5 FERTILIZ ER . S easo n 0 lbs. 50 lbs . 100 lbs. 200 lbs. 19 3 1-32 128 126 155 147 1932-33 195 237 228 19 33 -34 174 196 216 19') 19 3 4-35 137 139 146 133 Av e rage 15 9 174 186 159 The data are quite consistent in showing that 100 porn~d~; of manganese sulfate per ton (75 pounds per acre) was equally as good as, if not better than, the greater and less er amounts testP.d. In order to check these results on a commercial scale, sev er al cooperative t e sts were carried out with growers. The first of these was conducted on the farm of v;;. J. Vic~ ir , 1934-35 . Eleven rows, each one-fourth mile long, were planted •n the regul ar comm er cial manner for each treatment, ancl each treatment was replicated once. The regular basic fertilizer was a 4 8-5 applied at 1 ,500 pounds per acre. The three treatments consisted of 100, 150 and 200 pounds per ton , respectively, of 135 percent manganese sulfate mixed with the fertilizer. Ter. strips e<1ch 50 feet long were harvested at random from each plct and averaged for each treatment to obtain the data shown in Table 15 (Vick). The data show clearly that the 100-pound rate yielded slight ly better than the higher rates of manganese applicaticm. Another test was conducted in 19 35 -36 on a commercial scale on one of the farms of F. C. Peters, Inc. This farm, by the way, had never before had applications of manganese, so far as the owner knew. Approximately one-quarter acre was planted with

PAGE 27

F c rtili: xr Erp c ri111 c 11ts , i' ith Potatoe s 2i TABLE 15 . AV ER AGE YIELDS OF ::'\o. 1 T1.. :urns (B cS HEL S P ER A CR E) R tS l.'LTI XG FROM VARIOvS RATE S OF .-\PPLIC.HJOX OF MAXGA:
PAGE 28

28 Florida Agricultural Experiment Station have been averaged and are given in Table 16. Figures in italics indicate where manganese had been applied. TABLE 1 6 .-YI E L DS OF No. 1 T U BE RS (BUS HEL S PER A CRE) AS RELATED TO RECE:-i CY OF APPLI CATI ON OF MANGAN ES E SULFATE. Treatment 1931-32 1932-33 1933-34 1934-35 A 134 161 157 142 B 155 183 186 146 C 1 48 203 179 145 D 145 215 19 6 148 E 149 217 216 160 The yields fluctuated from year to year as is expected of different seasons. The data show consistently that each year the y ields were higher in the plots receiving manganese that year. Because of the normal yield fluctuations from year to year, the data are difficult to interpret from the standpoint of residual effects. Only a generalization can be made-that -there is a ten dency for the effect of manganese to continue somewhat into the second and possibly into the third year after its application , but that for all practical purposes, however, manganese should be applied each year, through a period of at least four years. Two years later (1937-38) another experiment was conducted on another section of the farm, to which manganese had been applied for at least five years in succession. Replicate plots (six each) with and without manganese (75 pounds per acre) were compared for yield. This year the fertilizer was applied by machine. The average yields, from Table 20, are given in Table 17 (1937-38). The addition of manganese did not increase the yield. TABLE 1 i.-AVERAGE YlELDS m l\o . 1 T UBE RS (BUSH EL S PER A CRE) OBT.U XED WI TH AND WIT HOU T MA NG ANES E IN A S TAN DARD 4-8-5 FE RTI LIZE R O N LA ND WH ERE MANGANE SE HAD BEEN APPLIED FOR SEVERAL YEAR S . Season 1937-38 1938-39 Control 292 142 Mangan ese 292 139 Again in 1938-39 the test was repeated in another field which had received manganese annually for at least seven years. Plots with and without mangan es e were planted as before , with the results (from Table 21) given also in Table 17 (1938-39). These data also show no incr e ase in yield resulting from the application of manganese . The results of these experiments indicate that for the first four years, at least, manganese sulfate should be applied annual

PAGE 29

F e rtili ze r Exp e rim e nts with Potato e s 29 ly, but that after a period of five years or more of successive applications , the manganese content probably may be omitted entirely from the fertilizer one year without decreasing the yields. The data here presented indicate not only that manganese ex hibi ts some residual effects , but also that there ma y result an accumulation of it in the soil from year to year in a form avail ab l e to the potato plant. It is possible , of course , that the residual effect of manganese becomes more pronounced when the fertilizer is applied by machine in the band method than b y hand where it becomes more thoroughly mixed with the soil. It is worthy of note that fairl y good yie lds of potatoes wer e obtained during several years of growing them without adding any manganese on land with no previous history of manganese applications (treatment A , Table 16). Although the addition of manganese with the fertilizer increased the yie lds profitably, its omission resulted onl y in lessened y ields and slightly smaller vine growth (Fig. 5). No other characteristic foliage s y mptoms of manganese deficienc y were noted. OTHER SOIL AMENDMENTS The results obtained with manganese naturally promoted a parallel interest in other soil amendments. Accordingl y, some ex periments with these wer e started on the Station farm in 1934 , to determine if the addition of any of these other elements might exhibit a beneficial effect on potato y ields . The first trials were of a more or less preliminar y nature and included tests of zinc , copper , iron and magnesium , as we ll as manganese in various Fig. 5.-Th e 193233 pot ato mangan ese t est. Th e pl ot in the background r eceived 100 pound s of manganese s ulfat e; that in th e for e ground no mangan ese for two yea r s .

PAGE 30

30 Florida .lgricultural Hxp e rimcnt Station combinations. The materials used as sources of these elements and amounts applied per ton of fertilizer were as follows: 1. Zinc-from 89 percent zinc sulfate @ 75 pounds per ton. 2. Copper-from bluestone crystals (copper sulfate) @ 100 pounds per ton . 3. Iron-from copperas (ferrous sulfate) @ 125 pounds per ton. 4. Magnesium-from commercial epsom salts (magnesium sul fate) @ 300 pounds per ton . 5. Manganese-from 65 percent manganese sulfate@ 100 pounds per ton. 6. Calcium-from gypsum (calcium sulfate) @ 1000 pounds per ton. All of these ingredients, in whatever combination they were used, were mixed with a regular 4-8-5 fertilizer (50 percent of N from organic sources) , which was applied by hand at the rate of 1,500 pounds per acre. Thus the elements were applied per acre at three-fourths the amounts listed abov e . The test the first year was composed of 14 treatments , applied to land which had had manganese applications for several years previously. The treatments are listed in Table 18 in their order of yield of No. 1 tubers. The plots consisted of single rows, 50 feet long and 38 inches apart, and each treatment replicated 16 times, in two series of 8 plots each . The tubers from each replicate plot were graded and the data converted to an acre basis and averaged for each treatment, with results shown in Table 18. TABLE l S .-AV ER AGE Y IEL D S OF No. 1 TU BE R S (Bu s HJ , LS PER A C R E) FRO:\r TH E TREAT MEN TS COMP OS ING THE 1 93 4-35 MI C RO-NUTRI ENT ErnME N T TESTS. El e m e nts Added to 4-8-5 Fertiliz e r Manganese-zinc-copp e r Manganes e z inc-copper-iron Manganes e z inc Manganes e -magnesium Iron-zinc Zinc Manganese m a gnesiumz inc-ir o n-copper Iron-zinc-copp e r Manganese-copper M a nganese-calcium-zinc-iron-copp e r Iron Iron-copper Copp e r Man g anese ( control) Yield 237 227 223 223 217 216 216 215 205 201 199 Hlfi 192 199 Increas e Ov e r Control 38 28 24 24 18 17 17 16 6 2 0 -3 -7 Odds 4298:1 71 :1 121:1 121 :1 27:1 43:1 17:1 14 : 1 2 : 1 According to the data of Table 18, a number of the treat ments showed average yields somewhat higher than the control or manganese treatment.

PAGE 31

Fertiliz e r Experiments with Potatoes 31 The data were analyzed statistically by "Student's" t test. Significance of differences in yield over that of the control plot is indicated in the column "odds". Odds of 30:1 or better are con sidered satisfactory for significance. On this basis the first six treatments listed in Table 18 may be considered as having signi ficantly outyielded the control treatment, since the one treat ment with odds of 27:1 closely approaches the arbitrary 30:1 standard. A striking fact is that most of the combinations including zinc are represented in these six treatments, indicating that ap plications of zinc sulfate may be beneficial on these soils. This is especially indicated by the fact that zinc alone and the zinc manganese combination outyielded manganese alone. The addition of copper and iron in combination with man ganese and zinc did not significantly increase the yield over that obtained by adding the latter two alone, and since iron and copper alone or together did not increase the yield, the data of this test indicate they were not beneficial. Evidence is provided that the addition of magnesium sulfate with manganese was beneficial, although when magnesium was added to the larger combination of manganese-zinc-iron-copper, the yield was lower than that of a similar mixture without mag nesium. The only mixture in which calcium sulfate was added failed to show an increase over a similar mixture without it. The following y ear (1935-36) an experiment was conducted on different ground comparing the eff e cts on yield of adding 89 percent zinc sulfate to the fertilizer at the rate of 50 pounds per ton, with and without manganese. The regular 4-8-5 fertilizer was applied by hand at the rate of 1 , 500 pounds per acre. Ten replica tions of each treatment were used . The average yields from each treatment are given in Table 19 (Series 1935-36). No increase in yield was obtained from the zinc applied in this experiment. The next year (1936-37) similar tests were conducted with zinc sulfate (89 percent) at the rates of 5, 25 and 50 pounds per ton, and with magnesium sulfate (commercial epsom salts) at rates of 25, 100 and 400 pounds per ton. A control treatment of 100 pounds of manganese sulfate was included also. The fertilizer to which they all were added was the regular 4-8-5 mixture con taining 100 pounds of manganese sulfate per ton, and it was applied by hand at the rate of 1,500 pounds per acre. Each of the seven treatments was replicated 10 times. The average yields

PAGE 32

32 Florida Agricultural Experiment Station TABLE 19.-AVERAGE YIELDS OF No. 1 TUBERS (BusnLs PER ACRE) OBTAINED I:S TEsTs OF VARIOUS MICRO-NUTRIENT ELEMENTS ADDED TO STANDARD 4-S-S FERTILIZER, 1935 3 7. Series 1935-36 1936-37 Elements Added Manganese (100)* (Control) Manganese (100) and zinc (50) Zinc (50) Manganese (100) (Control) Zinc (5) Zinc (25) Zinc (50) Magnesium (25) Magnesium (100) Magnesium (400) Yields 278 277 270 182 187 189 184 191 185 187 ,:, Figures in parentheses ref e r to pounds of the compound added per ton of fertilizer . are given for each treatment in Table 19, Series 1936-37. Here again the zinc failed to increase the yield over the manganese treatment and magnesium also had no beneficial effect. The small differences in average yields are not significant. The following year (1937-38) the micro-nutrient element tests were expanded to include more treatments and the fertili zer was applied by machine. This year five replications were planted of each treatment in a randomized block arrangement, with each treatment replicate paired with a control plot. The buffer rows, planted between treatment and control plot rows, were fertilized with the control mixture. Zinc sulfate (89 per cent), copper sulfate (crystal bluestone), iron (commercial fer rous) sulfate, iron (technical ferric) citrate, and magnesium sulfate (commercial epsom salts) were used this season in various combinations with and without manganese. The field in which the test was conducted had been planted annually to potatoes for at least the five preceding years , with manganese applica tions made each year. All ingredients were added to the regular 4-8-5 mixture, which was applied at the rate of 1,500 pounds per acre. The control mixture was this same 4-8-5 with no manganese added. The treatments tested and the average yields from them and from their respective control plots are given in Table 20. Statistical analysis of the data from which this table was compiled involved the use of "Student's" Z test, as applied to a paired plot arrangement. Odds of less than 30:1 are not consid ered significant. Zinc at the 25-pound rate, both alone and in combination

PAGE 33

Fertiliz e r Exp e rime11ts with Po tatoes 33 TABLE 20.. .\n : RAGE YIELDS OF :\'o. 1 Tr BERS (Bl ' SHELS PER .-\CREI OF THE TREAT MENTS AXD OF T HEIR R E SPECil\'E Co:nRO L PLOTS , 19 3i 3S ;\fr C RO :\'t:TRIEX T ELEl\lE:-iT T ESTS. Ave. Increas e El ements Added to 4 8-5 F er tili ze r Tr eat m ent Contro l of Tr eat ment Odds Yield Yield Over Control Zinc ( 75) manganes e (100) ,:, Zin c (75) manganes e (100) iron 3 10 285 25 23: 1 su l fate (125) 308 280 28 36:1 I ron (ferric) (134) 304 290 14 73:1 Zi nc (25) manganes e (100 ) 304 278 26 60:1 Zin c (25) 3 03 285 18 49:1 Zinc (75) 300 290 10 4: 1 Magnesium (400) manganese (100) 299 290 9 3:1 I ron (ferrous) (125) 299 276 2 3 44:1 Ir on (ferro us ) (125) manganese (100) 294 293 1 1 : 1 Zin c (75) copper (100) mangan ese (100) 292 281 11 21.1 Mangan ese (100) 292 292 0 0 Zi nc (75) iron (ferrous) (125) copper (100) manganes e (100) 28 5 279 6 6:1 ,:, Figures in p are nth ese s refer to pounds of the compound added per to n of fe rtili zer. with manganese, increased yields markedly. No exp l anation c an be offered for the poor results with zinc at 75 pounds , where th e increase was not significant. Manganese failed to increase y i elds when applied alone . When applied w ith z in c the incre ase was not appreciable over the yie ld of zinc (25) alone. This failure might be expected, in view of the continual manganese applications made in pre vio u s years on the fi e ld where these tests were made. Iron salts, bo th ferric and ferrous, increased y i e lds signifi cantly whe n applied alone , but not in combination with other elements . Ir on added to zinc and m a nganese did not increas e the y ield over that of the l atter two alone, nor did its addition with manganese increase the yield. The lack of incr ease from iron in combination with other elements is difficult to understand and no exp l anation is offered. Neither magnesium nor copper caused in crease in y iel ds. Another ser ie s of experiments al on g these same g ene ral lines was conducted in 1 938 39. Th e 4-8-5 fertilizer , to wh ich th e various amendments were ad ded, was applied by machine as before, at 1 , 500 pounds per acre . Five replication s were planted of ea ch treatment and its respective paired control pl o t on land which had received mangan ese for the preceding seven yea rs. Zi nc su lfate was added at 5, 25 and 75 pounds per ton, iron (ferrous) sulfate at 125 pounds and magnesium sulfate at 25

PAGE 34

34 Florida Agricultural Experiment Station and 400 pounds per ton. Copper sulfate and iron citrate were not used this year but boron, in the form of borax, was added at the rates of 12 and 24 pounds per ton. The treatments and average yields from them and from their respective control plots are given in Table 21. It was found that none of the small differences in average yields this year were significant. Thus this year manganese, magnesium and zinc were of no effect in increasing yields. Like wise no increase was obtained from either iron or boron. TABLE 21. AVERAGE YIELDS OF No. 1 TUBERS ( BUSHELS P E R A C R E ) OF THE TREAT1 1-IE N T S A N D OF TH E IR RESP EC TIV E CO N TR O L PLOTS , 19 38 -1 9 39 MI C R O -NUTRIENT EL EM E NTS TESTS. Average Increase Ingr e di e nts Added to 4-8-5 Fertiliz e r Treatment Control of Treatment Yield Yield Ov e r Control Zinc (75) manganese (100)' '' 156 Iron (125) manganese (100) 155 Zinc (25) manganese (100) 149 Zinc (5) manganese (100) 148 Boron (12) manganes e (100) 147 Magnesium (25) manganese (100) 144 Zinc (75) iron (125) manganese (100) 143 Iron (125) 141 Mang a nese (100) 139 Magn e sium (400) manganes e (100) 136 Boron (24) manganes e (100) 132 144 142 140 142 142 137 143 142 142 143 139 12 13 9 6 5 7 0 1 -3 -7 -7 * Figur e s in p a renth e s e s r e f e r to pounds of the compound per ton of fertilizer. SUMMARY OF M:ICRO-NUTRIENT ELEMENT STUDIES Considering the experimental results over a period of years , the data are rather convincing in establishing that the addition of iron sulfate, copper sulfate and magnesium sulfate to the soil in fertilizer form is not to be recommended as a general practic e in growing potatoes on these marl soils. In occasional instances their use resulted in slight increases in yield but the evidence is stron g enough otherwise to indicate that their use in the amounts tested, even as yield "insurance , " is unwarranted. The use of calcium sulfate , or gypsum, will be discussed later. It did not increase the yield in the 1934-35 test. Boron, used only one year at two different rates of applica tion, failed to improve the yields. The evidence is not so convincing concerning the use of zinc sulfate. In two of the five seasons zinc increased the yields significantly and the increases , although not very large, were profitable. This success is not to be ignored, although the fact

PAGE 35

Fertili ze r t:xpcri111c11ts ,, it/1 Potaf{}!'S that its addition to the fertilizer in the majority of years did not result in increased yields scarcely permits its recommendation for general use. The nature of zinc deficiency in these soils is not yet understood and it is not known if deficiencies occur in some years and not in others. Certainly the method of adding zinc needs furth e r study. As in the case of mangan es e , no obvious symptoms of deficiency which could be attributed to zinc have been observed. It is possible that this element, as well as some of the others , may better b e applied as a foliage spray , and ex periments are now under way at this Station to determine this point. Since this experiment was conducted with an ordin a ry 4-8-5 fertilizer with 50 percent of its nitrogen derived from organic sources, the results might not be applicable in instances where a radically different fertili z er program is followed. Also th e fact that benefits from adding thes e various micro-nutrient e lements have not been pronounced up to the present time does not exclude the possibility that their use may be required later on after intensive potato cropping has progressed year after year on the same land. In addition to the more complet e experiments thus far described a number of minor t e sts were carried out with sulfur , gypsum and manure. SULFUR Since the soil of this area is alkaline it seemed logical to determine what e ffect applications of sulfur might have on the resultant yields of potatoes. Accordingly, a test was conducted in 1932-33, in which a control 4-8-5 fertilizer tr e atm e nt applied at 2,000 pounds per acre was compared with a similar treatment to which had been add e d 1,000 pounds of dusting sulfur per acre, applied in the furrows by hand with the fertili z er two days before planting. The resultant average yields from three replica tions of each treatment are given in Table 22, Series 1932-33. No change in yield resulted from the addition of sulfur. The effect of the sulfur on the pH of the soil was not determined this season, but in 1938 Ruehle 0 , in studies of potato scab, applied 1,000 pounds of sulfur per acre in a similar manner and found that the pH was changed slightly but the change was of short duration. 6 Ru e hl e, G . D. Control o f potato dis e as e s in Dad e County. Fla. Agr. Exp. Sta. Ann. R e pt. 1938. p. 191.

PAGE 36

36 Florida Agricultural Experiment Station TABLE 22.-AVERAGE YIELD S or No. 1 TUBERS (BUSHELS PER ACRE) RESULTL'l'G F R OM USE OF VARIOUS AMOl:NTS OF SULFUR WITH A STANDARD 4-8-5 FERTILIZER. Series Treatm e nt Rat e per Acre Yi e lds 1932-33 Sulfur 1,000 pounds 195 Control -----------------197 1938-39 Sulfur 4,000 pounds 125 Control -----------------149 Filler Sulfur 123 pounds 140 Phos-rock 123 pounds 139 In 1938-39 a cooperative test was carried out on a field which had produced poor crops of potatoes due to its sea salt content. Analysis of the soil showed a chlorine concentration of about 4,000 parts per million in the surface and 1 , 100 p.p.m. at the 2 to 4-inch depth. The analysis was made by the Department of Chemistr y and Soils of the Florida Experiment Station at Gaines ville. Sulfur was among the materials tested on this soil. It was broadcast over the surface at th e rate of 2 tons per acre and disked in the day befor e planting. A 4-8-5 fertilizer was applied at planting time at the rate of one ton per acre. Resultant average y ields per acre from this test are given in Table 22 , Series 193839. In this case sulfur actually reduced the yield. This same year a t es t was conducted on th e Station farm in which two 4-8-5 fertilizer mixtures, identical except for their filler material , were compared. One was made up in the ordinary manner, using raw rock phosphate as filler, and the other with flowers of sulfur as filler. Each filler was added in the amount of 164 pounds per ton, and the complete fertilizers were applied at the rate of 1,500 pounds per acre by machine . Each plot was replicated five times. The replicate yields were averaged for each treatment and are also found in Table 22 , Series Filler. In this test the yields were the same for both treatments. These three experiments, then, indicate that no increase in yield of potatoes may be expected from the addition of sulfur in the amounts and b y the methods tested . GYPSUM The effect on yields of applying commercial calcium sulfate (gypsum or land plaster) was also studied over a period of years, as the result of preliminary tests in pot cultur e s which showed a stimulating effect of calcium on g rowth of tomato plants in marl soil. In 1932-33 gypsum was applied to the furrows by hand at the rat es of 1,000 pounds per acre, along with a regular 4-8-5

PAGE 37

Fertilizer Experiments with Potatoes 37 fertilizer applied at 2,000 pounds per acre . This treatment was compared with a similar on e without gypsum added . Three repli cations of each were planted and the resultant average yields are found in Table 23 (1932-33). The difference between these two yields was too small to be of significance. The following year gypsum was again applied with a 4-8-5 mixture by hand at two different rates. The treatments, each replicated four times, and their average yields are given in Table 23 (1933-34) also. This time small but significant increases in yield could b e attributed to the gypsum, but it is extr e mely doubtful if the additional cost and labor of applying the gypsum were justified. TABLE 2 3 .-AVERAG E YIELDS o r ); o , 1 T i:BE RS (Bl 'SII E L S PER A C R E ) OBTA IBE D 11"' T RI AL S O F G YPS l 'l-1 ADLJE D TO REG l'LAR 4 -S -5 F ER TI L IZER. S e ries Tr ea tment Rate per Acr e (lbs.) Yields 1932-33 Control 2000 196 Control + gypsum 2000 + 1000 202 Control 2000 193 1933-34 Control + gypsum 2000 + 1000 217 Control 1000 172 Control + gypsum 1000 + 500 187 1934-35 Mixtur e 1500 227 Mixture + gypsum 1500 + 750 201 1938-39 Cont r ol 2000 149 Control + gypsum 2000 + 4000 132 Another test in 1934-35 has already been described (Table 18). Gypsum at 750 pounds per acre was added with a 4-8-5 mixture (at 1 , 500 pounds) containing manganese , zinc, copper, and iron, and compared with a similar treatment without gypsum. The results are given again in Table 23 (1934-35) for comparison. In this test gypsum decreased the y ield. The cooperative experiment of 1938-39 in the field with a high chlorine content, which contained the sulfur experiment, also contained a treatment in which gypsum was broadcast at the rate of 2 tons per acre and disked in prior to planting. Average yields per acre of the 10 replications of the gypsum and control plots are included in Table 23 (1938-39). Again in this experiment gypsum failed to increase the yield. Thus, it is concluded that the addition of gypsum in the amounts tested to these marl soils is not justified from the stand point of increasing potato y ields .

PAGE 38

38 Florida Agr icu ltural Experiment Station MANURE A series of replicated plots was planted in 1932-33 to test the effect on yield of adding stable manure to the soil. Stable manure was broadcast over the plots at the rate of 6 tons per acre and disked in prior to planting. The potatoes in the manure and control plots both were fertilized with a ton per acre of 4:-8-5 containing 100 pounds of manganese sulfate. The average yields are given in Table 24 (1932-33) . The manure significantly in creased the yield, but at local manure prices (about $5 per ton) the treatm e nt was not profitable. TABLE 24 .AVERAGE Y I E LD S OF No . l TU!lEI!S (BUS H ELS PER A CR E ) O13TAL'H:D IN TRIALS OF MAN 1.,RE APP LIED IN A DDlT lON TO STAN D A RD 4-85 FERTI LIZE R CO TAINI NG MANGAN ESI•; . S e ri es Tr ea tm e nt Rate p e r Acr e ( tons) Yi e lds 1932-3 3 Control 1 19 6 Control + m a nure 1 + 6 2 20 1938-3 9 Control 1 1 49 Contr o l + m an ure 1 + 8 162 Ago.in in the 1938-39 season manure was added as one of the tr ea tments on the salty land to which th e sulfur and gypsum were applied. In this exper im e nt stable manure was br o adca st uniformly at the ra te of 8 tons per acre and disked in just pri o r to pl a nting. The potatoes were fertilized in the regular commer cial manner with a ton of 4-8-5 containing 100 pounds of man ganese sulfat e . Average resultant yields of tubers are given iD Table 24. A ga in the manure increased the yield but not enough to warrant the additional cost. These two experiments indicate that stable manure has a beneficial effect on yi e lds, but as applied in these tests its ad diti ona l cost was prohibitive. Incidentally, in neither test did the p e rcentage of potato scab ( Ac tinom yces sca b ies (Thax.) Gussow) increase. There was practically no scab in any of the plots. SUMMARY Fertilizer experiments conducted at the Sub-Tropical Ex periment Station over a period of 10 ye ars are reported in which analy ses, amounts, sources of nitrogen and potash , and applica tions of manganes e and other soil am e ndments w e re studied as affecting yields of Bliss Triumph potatoes grown on the marl soils of Dade County, Florida. Analyses varying from a 2-8-5 and 3-12-8 to 8-16-10 were tested . The results indicate that for mixtures to be applied at rates of about 1,500 to 2,000 pounds per acre there is no justifica

PAGE 39

F e rtilizer Experiments with Potatoes 39 tion for increasing the analysis beyond 3 or 4 percent nitrogen, 8 percent phosphoric acid and 4 or 5 percent potash. Experiments with ordinary 4-8-5 and 3-12-8 analyses showed that the correct amounts of these to apply was from about 1,500 to 2,000 pounds per acre _ , and in most instances the most profit able amount was 1,500 pounds. In four of the five years of testing a 4-8-5 mixture in which 33 percent of the nitrogen was derived from organic sources yielded as well as mixtures containing a higher percentage of nitrogen so derived. No definite relation of treatment yields to monthly rainfall during the crop season was observed during the five years over which these tests were conducted. The source-of-nitrogen tests indicat e d that the organic materials milorganite, blood-and-bone tankage (medium grade) and dried blood slightly outyielded the other sources, and profit ably so. No significant differences in yield were obtained among these three. Fish scrap, cottonseed meal, urea and cyanamid yielded slightly less , and about the same as mixtures containing all their nitrogen from ammonium phosphate, sulfate of am monia or nitrate of soda. No advantage was demonstrated from combining the latter two materials. Lik ew ise a fish scrap and cottonseed meal combination yielded no better than mixtures in which these materials were used separately. The ur e a and cyanamid treatments both produced yields in the upper range of the water-soluble sources, and both were among the treatments showing the lowest fertilizer cost per bush e l. No significant diff e rences in yie ld were obtained when sul fat e of ammonia and nitrate of soda were compared as inorganic sources of nitrogen in a 4-8-5 formula containing 50 percent of its nitrogen from organic sources. Likewise no significant differences in yield were obtained between the sulfate and muriate forms of potash as sources of potash in a similar 4-8-5 fertilizer. Applications of 100 pounds of 65 percent manganese sulfate per ton of fertilizer gave as good yields as greater amounts. It was found that on "new" land applications of manganese sul fate should be made annuall y for at least the first four years. After a period of five or more years of successive applications, however, results indicated that th e manganese content of the fertilizer could be entirely omitted at least one year without decreasing the yield.

PAGE 40

40 Florida Agricultural Experiment Station Applications of magnesium sulfate (epsom salts), copper sulfate (bluestone), iron sulfate (copperas), borax, iron citrate, sulfur and calcium sulfate (gypsum) failed to increase yields profitably. There were some seasons when zinc sulfate increased yields but results were too inconsistent to warrant its general recommendation as a f e rtilizer ingredient. Stable manure increased the yields slightly when applied at 6 to 8 tons per acre with commercial fertilizer, but not sufficient ly to justify the cost . ACKNOWLEDGMENTS Th e Board of County Commission e rs of Dad e County assist e d by appropriating part of th e funds for these experim e nts each year since the 1935 season. Dr. G. D . Ruehl e sup e rvised much of the spraying of the plots in later years, and Messrs. L . R. Toy , Ivan Moser , Lloyd Fi e ld and Roy Nelson assisted with much of the field work involved in conduct ing the tests. Messrs D. P. Blake, Jr., Luther Chandler, F. M. Dolan, J. H. Estes, J. M. Halferty, F. C. Peters, Frank Ru e and W. J. Vick, growers, kindly cooperated in some of the tests reported. County Agent Charles Steffani also assisted with the cooperative work. The Redland District Chamber of Commerce provided six acres of test plots for some of the experiments conducted in 1932-33.