Nutritional requirements of high-yield irrigated crop production systems

MISSING IMAGE

Material Information

Title:
Nutritional requirements of high-yield irrigated crop production systems
Physical Description:
xiii, 222 leaves : ill. ; 28 cm.
Language:
English
Creator:
Obreza, Thomas A., 1956-
Publication Date:

Subjects

Subjects / Keywords:
Corn -- Fertilizers   ( lcsh )
Double cropping -- Florida   ( lcsh )
Plant-soil relationships   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1983.
Bibliography:
Includes bibliographical references (leaves 211-221).
Statement of Responsibility:
by Thomas A. Obreza.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000366100
notis - ACA4935
oclc - 10033417
System ID:
AA00003844:00001

Full Text















NUTRITIONAL REQUIREMENTS OF HIGH-YIELD
IRRIGATED CROP PRODUCTION SYSTEMS











By


THOMAS A. OBREZA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1983













ACKNOWLEDGEMENTS


The author wishes to extend his most sincere gratitude

to all who assisted him in the completion of this manuscript.

The most special thanks go to Dr. Fred Rhoads, chairman of

the supervisory committee, for his help, guidance, patience,

and friendship throughout the graduate program. He has

contributed considerably to the author's intellectual growth.

Appreciation is also extended to the other members of the

committee, Dr. W. G. Blue, Dr. P. S. C. Rao, Dr. R. L.

Stanley, Jr., and Dr. F. G. Martin, for their constructive

criticisms which helped put this work into its final form.

The author is very grateful to Mr. Audley Manning for

his friendship and help in the laboratory and in the field.

Thanks go to all the rest of the crew and staff at the AREC,

Quincy, who made the days spent out there pleasurable. The

financial assistance of the Center has also been gratefully

appreciated.

Gratitude is expressed towards fellow graduate students,

faculty, and staff of the Soil Science Department, who are

primarily responsible for making the author's graduate school

experience an enjoyable one.

The author wishes to remember his mother and father,

who have always encouraged him in everything he has








attempted. His only regret is that his mother is no longer

on this earth to see him complete his graduate study and

receive his Ph.D.

Finally, deepest appreciation and love are expressed to

the author's wife, Laurie, who has stood by him and encour-

aged him in good times and bad, and has offered the dis-

tractions necessary for the author to keep his sanity while

preparing this work.


iii












TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS ....................................... ii

LIST OF TABLES............... .......................... vi

LIST OF FIGURES........................................... ix

ABSTRACT ............................................... xii

INTRODUCTION ........................................... 1

CHAPTER 1: LITERATURE REVIEW.......................... 7

Corn Response to Irrigation.......................... 7
The Economics of High Yields....................... 10
Nutrient Management for Corn Production............. 14
Soil Testing................................. ........ 33
Multiple Cropping ................................... 50
Summary ............................................ 53

CHAPTER 2: MATERIALS AND METHODS....................... 57

Field Procedures .................................. 57
Laboratory Procedures ............................. 70
Statistical Analyses and Handling of Data........... 72

CHAPTER 3: RESULTS AND DISCUSSION..................... 77

Yield Response to Treatments........................ 77
Response of Soybean and Grain Sorghum Nutrient
Concentrations to Treatments........................ 82
Response of Other Corn Field Parameters to
Treatments .......................................... 82
Corn Dry-Matter Yield vs. Nutrient Uptake........... 87
Corn Grain Yield vs. Nutrient Uptake................. 93
Stepwise Regression of Grain Yield on Nutrient
Uptake.................................................... 105








Page

Nutrient Use Efficiency............................. 108
N Uptake as Affected by P Uptake.................... 111
Potassium, Mg, and Ca Interrelationships............ 113
Nutrient Uptake vs. Soil-Test Plus Applied Nutrient. 117
Total Corn Dry-Matter Yield vs. Soil-Test Plus
Applied Nutrient.................................... 127
Corn Grain Yield vs. Soil-Test Plus Applied
Nutrient ............................................ 134
Determination of Critical Values of Soil-Test Plus
Fertilizer Nutrient................................. 143
Use of Critical Ranges of Soil-Test Plus Fertilizer
Nutrient for Estimating Fertilizer Requirements of
Irrigated Corn.. .................................... 153
Comparison of Established Soil-Test Rating
Categories and Fertilizer Recommendations to
Critical Ranges of Soil-Test Plus Fertilizer
Nutrient............................................. 156

CHAPTER 4: SUMMARY AND CONCLUSIONS.................... 165

APPENDIX 1: LABORATORY PROCEDURES..................... 175

APPENDIX 2: CATE-NELSON STATISTICAL PROCEDURE.......... 188

APPENDIX 3: FIELD AND LABORATORY DATA................. 191

LITERATURE CITED ........ ... .......................... 211

BIOGRAPHICAL SKETCH.................................... 222













LIST OF TABLES


Table Page

1 Estimated cost of producing 1 hectare of
corn, north Florida, 1976....................... 11

2 Estimated cost of producing 1 hectare of
corn, north Florida, 1979...................... 12

3 Estimated cost of producing 1 hectare of
irrigated corn, north Florida, 1982............. 13

4 Comparison of soil-test rating, fertility
index, relative yield, and recommendations
based on soil tests............................ 46

5 Soil-test rating categories and ranges
for Florida, Georgia, and Alabama.............. 47

6 Fertilizer recommendations for corn from
Florida, Georgia, and Alabama.................. 49

7 Selected properties of the Ruston loamy
fine sand....................................... 58

8 List of fertilizer treatments.................. 62

9 Pesticides and their rates of use............... 65

10 Corn and soybean yields as affected by
fertilizer treatments. 1980................... 78

11 Corn, soybean, and grain sorghum yields as
affected by fertilizer treatments.............. 80

12 Concentrations of P, K, and Mg in soybeans
and grain sorghum as affected by the fer-
tilizer treatments applied to corn. 1981....... 83

13 Response of number of stalks, number of ears,
ear size, and percent barren stalks to
fertilizer treatments. 1980................... 85

14 Response of number of stalks, number of ears,
ear size, percent barren stalks, and percent
lodged plants to fertilizer treatments. 1981.. 86







Table Page

15 Results of the stepwise regression for the
relationship between corn grain yield and
the uptake of the nutrients N, P, K, Mg, Ca,
Zn, and Mn at four sampling dates. 1980
and 1981...................................... 106

16 Mean values of total N uptake at each level
of applied N for 1980, 1981, and combined
data from both years............................ 120

17 Results of the Cate-Nelson procedure calcu-
lations near the critical level of soil-test P
plus fertilizer P for 1980, 1981, and com-
bined data from both years..................... 145

18 Results of the Cate-Nelson procedure calcu-
lations near the critical level of soil-test K
plus fertilizer K for 1980, 1981, and com-
bined data from both years..................... 148

19 Results of the Cate-Nelson procedure calcu-
lations near the critical level of soil-test
Mg pluse fertilizer Mg for 1980, 1981, and
combined data from both years................... 151

20 Soil-test rating categories and ranges for
irrigated corn estimated from calculated
critical ranges of P, K, and Mg................ 158

21 Values of soil-test P obtained from each
experimental plot prior to the 1980 growing
season......................................... 191

22 Values of soil-test P obtained from each
experimental plot prior to the 1981 growing
season....................................... 192

23 Values of soil-test K obtained from each
experimental plot prior to the 1980 growing
season......................................... 193

24 Values of soil-test K obtained from each
experimental plot prior to the 1981 growing
season .... ................................... 194

25 Values of soil-test Mg obtained from each
experimental plot prior to the 1980 growing
season. ........................................ 195


vii







Table Page

26 Values of soil-test Mg obtained from each
experimental plot prior to the 1981 growing
season......................................... 196

27 Observations of corn grain yield. 1980......... 197

28 Observations of corn grain yield. 1981......... 198

29 Observations of total corn dry-matter yield.
1980........................................... 199

30 Observations of total corn dry-matter yield.
1981 ....................... .................... 200

31 Concentrations of N in mature corn plants.
1980 ........................................... 201

32 Concentrations of N in mature corn plants.
1981 ........................................... 202

33 Concentrations of P in mature corn plants.
1980 ............................................. 203

34 Concentrations of P in mature corn plants.
1981 ........................................... 204

35 Concentrations of K in mature corn plants.
1980............................ ................ 205

36 Concentrations of K in mature corn plants.
1981........................................... 206

37 Concentrations of Mg in mature corn plants.
1980 ........................... ................ 207

38 Concentrations of Mg in mature corn plants.
1981..................... ....................... 208

39 Observations of soybean yield. 1980 and 1981.. 209

40 Observations of grain sorghum yield. 1981..... 210


viii













LIST OF FIGURES


Figure Page

1 Soil-water characteristic curve for Ruston
loamy fine sand at two depth intervals........ 59

2 General plot diagram and individual plot
layout........................................ 61

3 Rainfall amounts and patterns at the AREC,
Quincy during the growing seasons of 1980
and 1981..................................... 64

4 The relationship between corn dry-matter
yield and N uptake in 1980 and 1981............ 89

5 The relationship between corn dry-matter
yield and P uptake in 1980 and 1981............ 91

6 The relationship between corn dry-matter
yield and K uptake in 1980 and 1981............ 92

7 The relationship between corn dry-matter
yield and Mg uptake in 1980 and 1981........... 94

8 The relationship between corn grain yield and
N uptake at the silking growth stage.......... 95

9 The relationship between corn grain yield and
N uptake at the mature growth stage............ 96

10 The relationship between corn grain yield and
P uptake at the silking growth stage........... 98

11 The relationship between corn grain yield and
P uptake at the mature growth stage........... 99

12 The relationship between corn grain yield and
K uptake at the silking growth stage........... 100

13 The relationship between corn grain yield and
K uptake at the mature growth stage............ 101







Figure Page

14 The relationship between corn grain yield and
Mg uptake at the silking growth stage......... 103

15 The relationship between corn grain yield and
Mg uptake at the mature growth stage.......... 104

16 Nutrient use efficiency (NUE) calculated for
N as related to corn grain yield and N uptake. 109

17 Nutrient use efficiency (NUE) calculated for
P as related to corn grain yield and P uptake. 110

18 Nutrient use efficiency (NUE) calculated for
K as related to corn grain yield and K uptake. 112

19 The relationship between N uptake and P uptake
by corn at four weeks after emergence......... 114

20 The relationship between N uptake and P uptake
by corn at the silking growth stage (9 weeks
after emergence) .............................. 115

21 The relationships between total Mg and Ca
uptake and total K uptake by corn............. 116

22 The relationship between total N uptake by
corn and the amount of N applied............... 119

23 The relationship between total P uptake by
corn and the amount of soil-test P plus the
amount of fertilizer P applied............... 122

24 The relationship between total K uptake by
corn and the amount of soil-test K plus the
amount of fertilizer K applied................. 124

25 The relationship between total Mg uptake by
corn and the amount of soil-test Mg plus the
amount of fertilizer Mg applied................ 126

26 The relationship between corn dry-matter yield
and the amount of N applied................... 129

27 The relationship between corn dry-matter yield
and the amount of soil-test P plus the amount
of fertilizer P applied...................... 130








Figure Page

28 The relationship between corn dry-matter yield
and the amount of soil-test K plus the amount
of fertilizer K applied....................... 132

29 The relationship between corn dry-matter yield
and the amount of soil-test Mg plus the amount
of fertilizer Mg applied...................... 133

30 The relationship between corn grain yield and
the amount of N applied....................... 136

31 The relationship between corn grain yield and
the amount of soil-test P plus the amount of
fertilizer P applied........................... 138

32 The relationship between corn grain yield and
the amount of soil-test K plus the amount of
fertilizer K applied.......................... 140

33 The relationship between corn grain yield and
the amount of soil-test Mg plus the amount of
fertilizer Mg applied........................ 142

34 Graphical representation of the results of
the Cate-Nelson procedure which determined
critical values of soil-test P plus fertilizer
P for the production of high-yield irrigated
corn.............................. ............ 146

35 Graphical representation of the results of
the Cate-Nelson procedure which determined
critical values of soil-test K plus fertilizer
K for the production of high-yield irrigated
corn........................................... 149

36 Graphical representation of the results of
the Cate-Nelson procedure which determined
critical values of soil-test Mg plus fer-
tilizer Mg for the production of high-yield
irrigated corn .... ............................ 152













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


NUTRITIONAL REQUIREMENTS OF HIGH-YIELD
IRRIGATED CROP PRODUCTION SYSTEMS

By

Thomas A. Obreza

April 1983

Chairman: F. M. Rhoads
Major Department: Soil Science

Increasing costs of energy, water, and fertilizer

demand that fertilizer inputs for irrigated corn (Zea mays

L.) be adequate but not excessive for high yield potential.

Research data relating fertilizer rates and residual plant

nutrients to high corn yields under irrigation in the South-

east are in short supply because most irrigation experiments

have been conducted on soils with high residual levels of

plant nutrients. A knowledge of total nutrient uptake is

necessary to verify that nutrient availability is not

limiting yields.

Irrigated double cropping systems consisting of corn

followed by rye (Secale cereale L.), grain sorghum (Sorghum

bicolor L.), and soybeans (Glycine max L.) were studied in

1980 and 1981 to determine nutrient uptake and yield

responses to applied fertilizers and residual (soil-test)


xii







nutrients. Nitrogen was applied to corn at 168, 336, and

504 kg/ha, P at 0, 29, 59, and 117 kg/ha, K at 0, 209, and

418 kg/ha, and Mg at 0, 67, and 134 kg/ha. Plant samples

were taken at 3, 4, 9, and 13 weeks after emergence for

corn and at maturity for grain sorghum and soybeans for dry-

matter and nutrient concentration measurements. Grain

yields were measured when each crop was mature.

Corn grain yields at the highest level of applied

nutrients averaged about 11 Mg/ha both years. Highest grain

yields were obtained where at least 168 kg N, 29 kg P, and

209 kg K were applied per hectare. The largest soybean

and grain sorghum yields were near 1.7 Mg/ha and 2.8 Mg/ha,

respectively.

Uptake of nutrients by corn increased and did not level

off as the level of applied plus soil-test nutrients

increased. Relationships between grain yield and nutrient

uptake showed that in many cases the amount of nutrients

taken up was much greater than the amount required for

maximum yield. Nitrogen uptake was positively correlated

with P uptake; however, the uptake of large amounts of P with

small amounts of N indicated that N uptake did not influence

P uptake.

Critical ranges of soil-test P, K, and Mg were defined

as ranges below which nutrients limited yields and above

which nutrients did not limit yields. The critical ranges

were 17-35 kg P/ha, 86-147 kg K/ha, and 74-340 kg Mg/ha.


xiii













INTRODUCTION


In the recent past, agricultural researchers and econo-

mists have stressed the need for higher crop yields. This

need has focused on two general areas: expanding world food

demands and the economic returns to the grower. The effect

of rapidly increasing world population on food demand is

obvious: more food will have to be produced each succeeding

year than ever before. When looking at economic returns to

the grower, it can be seen that profits are the incentive

for the grower to produce. Crop production costs are expect-

ed to double in the next 10 years, while crop prices are not

likely to keep pace. In order to increase profits for grow-

ers under these conditions, results of maximum-yield research

need to be focused on developing maximum economic yield pro-

duction systems that growers can use. Such systems should

increase crop yields, lower unit production costs, and in-

crease profits.

In order to identify maximum-yield research concepts,

one must first understand maximum-yield research philosophy.

Maximum-yield research is the study of one or more variables

and their interactions in an integrated system which has the

objective of attaining the highest yield possible for the

soil and climate of the research site (Griffith, 1982).








Economics are of no immediate concern. The final goal is to

maximize yields for any given set of conditions by optimiz-

ing the use of the controllable inputs.

Success in maximum-yield research is required before

any maximum economic yield analysis can be made. Once this

success has been achieved, the following question must be

answered: Are the maximum yields being obtained in research

profitable for the grower? The above definition for maximum-

yield research suggests that maximum economic yields might

be somewhat lower than maximum-yield levels. The maximum

economic yield level can be defined as that yield achieved

from the optimum use of all controllable inputs which gives

the highest possible net return per hectare. This yield

level will vary in its relationship to maximum-yield levels

depending on a number of things, including the type of crop,

the cost of the inputs, and the price that the crop brings.

One controllable input that must be optimized to obtain

high economic yields is fertilizer application. Fertilizers

are added for two main reasons, first to build up the fertil-

ity level of the soil, and second to maintain the soil fertil-

ity level by compensating for nutrients removed by crop uptake

or leaching. The type of nutrient and the characteristics of

the soil will influence the degree to which that nutrient can

be built up in the crop root zone. In the sandy Ultisols of

north Florida, inorganic N in the NO3 form will readily

leach below a crop root zone during periods of excess rain-

fall. Potassium, in the K+ form, has a reduced leaching








potential as compared to NO but significant leaching of K

occurs in north Florida due to the low CEC of the soil.

Phosphorus, Mg, and Ca are relatively immobile in limed

agricultural soils of north Florida due to the formation of

Mg and Ca phosphates. Thus, due to the widely different

leaching potentials of the major mineral nutrients, opti-

mizing their application as fertilizer is not an easy task.

One must attempt to keep an optimum level of each nutrient

in the crop root zone at all times while at the same time

conserving nutrients by trying to prevent their loss through

leaching.

Another controllable input which can affect maximum

economic yield levels dramatically is irrigation. The char-

acteristics of the soil to be irrigated determine how much

and how often water needs to be applied to meet the crop

demand. As with fertilizers, applied water needs to be kept

at optimum levels in the crop root zone while application of

excessive amounts should be avoided. As water is depleted

from the root zone via evapotranspiration, irrigation water

needs to be applied in an amount that will recharge the de-

pleted zone without causing any leaching of mobile nutrients.

A concept of plow-layer soil-water management was

developed at the Agricultural Research and Education Center

in Quincy, Florida, during the 1970s which emerged from the

idea of only recharging the upper part of the crop root zone

as it became depleted of water (Rhoads and Stanley, 1973).

Under this management system, which calls for soil-water








suction to be maintained below 0.02 MPa in the plow-layer

(top 15 cm) throughout the growing season, corn (Zea mays L.)

yields consistently exceeded 12.5 Mg/ha (200 bu/acre) where

not limited by nutrient supply or pests. The highest yield,

about 15.0 Mg/ha, was produced with 71,600 plants/ha in

population studies with plow-layer soil-water management.

In most cases, fertilizer treatments were designed to supply

nutrients in large enough quantities not to limit grain

yields. Nitrogen was generally supplied at the rate of

336 kg/ha, P at 100 kg/ha, and K at 280 kg/ha. Thus, while

much research has been done in the Southeast to improve

irrigation efficiency, research data relating fertilizer

rates and residual soil nutrients to total nutrient uptake

and high (12.5 Mg/ha) corn yields under irrigation in the

Southeast are in short supply. Such data are needed so that

the fertilizer input to high corn yields can be economically

optimized.

Past researchers have used nutrient concentration in

plant tissue as a measure of crop nutritional requirements

(Pierre et al., 1977; Ware et al., 1982; Dow and Roberts,

1982). Percentages of nutrients in the parts of a corn plant

are good indicators of nutrient needs under specified con-

ditions, but may be easily misinterpreted when the crop is

grown at different locations, planting rates, and weather

conditions. It is believed that total nutrient content

(in kg/ha) determined by whole plant analysis and plant

population would be most valuable in determining the amount








of nutrients required for high-population corn irrigated for

high yields. This approach may also be useful in predicting

potential yield if a relation between nutrient uptake at a

specific growth stage and yield can be established.

The long growing season in the Southeast allows for two

crops to be grown in the same season without danger of frost.

Corn planted in March can be harvested for grain in July,

leaving approximately 120 to 140 frost-free days to grow a

second crop such as soybeans (Glycine max L.) or grain

sorghum (Sorghum bicolor L.). This is an excellent way for

Southeastern growers to increase their profits, especially

if land is under irrigation. Double cropping can increase

the efficiency of the fertilizer applied to the first crop,

because residual nutrients such as P and K left over from

fertilizer applied to corn may be utilized by the second

crop rather than be leached. Double cropping also allows

growers to take advantage of their irrigation systems,

because the more these systems are used, the faster they can

pay for themselves.

In 1980, a research program to evaluate the nutritional

requirements of high-yield cropping systems under irrigation

was undertaken at Quincy, Florida. The goal of this program

was to evaluate the effects of applied and residual plant

nutrients on crop uptake and yield under conditions of

established plow-layer soil-water management techniques.

Although the primary focus was placed on the first crop

(corn), the response of two of the second crops (soybeans








or grain sorghum) to residual fertilizer nutrients was also

evaluated.

The objectives of this research were 1) to correlate

double-acid extractable (soil-test) plant nutrients and

soil-test plus fertilizer nutrients with crop uptake of P,

K, and Mg; 2) to establish critical levels of soil-test plus

fertilizer nutrients required for the production of high-

yield irrigated corn; 3) to evaluate nutrient use efficiency

in high-yield cropping systems; 4) to quantitatively monitor

soil-plant-nutrient systems; and 5) to evaluate N, P, K, and

Mg nutritional requirements and interrelationships.













CHAPTER 1
LITERATURE REVIEW

Corn Response to Irrigation


Water, whether from rain or irrigation, is a major

input into the overall corn production management system.

Numerous studies dealing with corn response to soil-water

regime have shown that maximum corn yields cannot be ob-

tained unless the management system used insures that soil

water will not be limiting. Robins and Domingo (1953) and

Robins et al. (1967) stated that corn can tolerate appre-

ciable soil-water stress with only a limited effect on

grain yields except during the growth period from tassel

emergence to completion of pollination. Early-season stress

reduced grain yields only when severe wilting occurred

(Robins et al., 1967). Denmead and Shaw (1960) grew corn

in the field in buried 19 liter crocks and applied soil-

water stress at different growth stages. They found that

the highest grain yields were produced when no soil-water

stress was applied during any growth stage. Haynes (1948)

grew corn in a greenhouse and found that dry-matter pro-

duction was greatest when a permanent water table was main-

tained 15 cm below the soil surface. Rawitz and Hillel

(1969) reported that maximum corn yields were obtained at

soil suctions below 0.1 MPa in the 0-90 cm soil layer.








In the past, grain yields of corn in Florida have been

consistently lower than the national average. The average

corn yield for the state was about 1.0 Mg/ha in 1951 (Lips-

comb and Robertson, 1954) and had risen to only about 3.0

Mg/ha by 1980 (Florida Dept. of Agriculture Field Crops

Summary, 1980) while the national average was above 6.0

Mg/ha (Wright et al., 1980). Water has been the most limit-

ing factor in Florida for corn yields on mineral soils.

This was determined by Rhoads and Stanley (1973), who irri-

gated different varieties of corn when the soil-water suc-

tion in the 15-cm depth plow-layer reached pre-determined

levels (0.03 and 0.06 MPa in 1970 and 0.02, 0.04, and 0.06

MPa in 1971). In both years, corn irrigated at the lowest

value of soil-water suction yielded significantly more than

corn irrigated at the highest soil-water suction. The

highest yielding variety yielded 11.9 Mg/ha when irrigated

at a soil-water suction value of 0.02 MPa and 7.2 Mg/ha with

no irrigation. Each irrigation treatment had fertilizers

applied such that nutrients would not limit yields. In an

associated experiment, Stanley and Rhoads (1971) obtained

yields of larger than 12.5 Mg/ha using plow-layer soil-water

management procedures where irrigation was applied at 0.03

MPa of soil-water suction. Rhoads and Stanley (1975) tested

this same type of management on three soils of different

textural classes and found that for all three soils, irri-

gated corn yields were significantly higher than unirrigated

corn yields, even during a year where rainfall was relatively







plentiful during periods of high crop demand for water and

nutrients. The maximum yield averaged over all three soils

was 9.5 Mg/ha while the minimum (unirrigated) yield aver-

aged 6.7 Mg/ha.

Dry weather usually reduces unirrigated corn yields

significantly each year in the Southeast. In addition, the

rainfall that does occur falls in patterns such that fer-

tilizer losses due to leaching can be a problem in most

years. As previously seen, irrigation is needed to obtain

maximum corn yields, but should only be applied to recharge

the top 15-20 cm of soil to minimize fertilizer movement

(Rhoads and Stanley, 1975). Thus, the concept of plow-layer

soil-water management is a logical solution to the problem

of irrigation scheduling in the southeastern United States.

Research results from Texas (Shipley et al., 1973) support

this concept.

Rhoads (1981b) pointed out several practical advantages

to plow-layer soil-water management. First, less water is

required per irrigation. This allows large center pivot

irrigation systems to cover the area to be irrigated before

plants become severely water-stressed. Also, the loss of

nutrients due to heavy summer rains that occur unpredictably

in the Southeast would not be as great because light, fre-

quent irrigation would make a minimum contribution to

nutrient movement. It has been pointed out by Rawlins and

Raats (1975) that uniform, frequent irrigation optimizes the

root environment while drastically reducing water use.








The Economics of High Yields


Adding irrigation to any corn production management

scheme, while increasing yields, will also increase produc-

tion costs per hectare. The question to be answered then

is: Is the increased cost of production of irrigated corn

more than offset by the increased yields caused by irriga-

tion? Data presented in Tables 1 and 2 help answer this

question. In 1976, the irrigated corn grower in north

Florida spent 1.6 times as much money per hectare as did the

growers of unirrigated corn, but received slightly over 11

times as much revenue (Table 1). In 1979, irrigated corn

growers spent 1.7 times as much money per hectare as did

unirrigated corn growers and received nearly five times as

much revenue (Table 2). The returns obtained from irrigated

corn management appear to far outweigh the additional pro-

duction costs.

Data from 1982 (Table 3) show that in 6 years the cost

of producing a hectare of irrigated corn had doubled while

the projected yield and the price received per megagram of

corn did not nearly increase in the same proportion. Farm-

ing, like any other business, faces the impact of inflation-

ary pressures. According to Dibb (1982), one way to ease

these pressures is better marketing of crops by growers.

However, he stated that the key to survival in the business

of farming is in improved productivity and that higher

yields will be the primary path to economic survival for







Table 1. Estimated cost of producing 1 hectare of corn,
north Florida, 1976.


Irrigated Unirrigated
Item Unit Price Quantity Value Quantity Value


kg $0.102


9408 $959.62


3763 $383.83


Variable costs:
Seed
Fertilizer
Nitrogen
Lime t
Insecticide
Herbicide
Tractor
Truck, pickup
Truck, 2-ton
Machinery
Combine
Labor
Irrig. costs
Interest on
above expenses


Total variable costs

Fixed costs:


Tractor
Truck, pickup
Truck, 2-ton
Combine
Machinery
Irrigation


kg 1.54
kg 0.117
kg 0.53
on 15.43
kg 1.30
kg 4.93
hr 3.27
km 0.031
km 0.069
hr 1.04
hr 6.57
hr 2.50
ha 30.37


16.8
896
168
0.74
16.8
4.5
5.46
79
79
5.46
1.0
7.4
1.0


0.05 362.06


hr 3.11
km 0.063
km 0.078
hr 16.35
hr 2.15
ha 126.09


5.46
79
79
1.0
5.46
1.0


Total fixed costs
Total costs
Returns to land and management


25.87
104.83
89.04
11.42
21.84
22.19
17.85
2.45
5.45
5.68
6.57
18.50
30.37


13.4
616
140
0.74
16.8
4.5
5.46
79
79
5.46
1.0
7.4
-


18.10 278.86


380.16


16.98
4.98
6.16
16.35
11.74
126.09
182.30
562.46
397.16


5.46
79
79
1.0
5.46


Source: Rhoads and Russell (1977).


Revenue


20.64
72.07
74.20
11.42
21.84
22.19
17.85
2.45
5.45
5.68
6.57
18.50

13.94

292.80


16.98
4.98
6.16
16.35
11.74

56.21
349.01
34.82








Table 2. Estimated cost of producing 1 hectare of corn,
north Florida, 1979.


Irrigated Unirrigated
Item Unit Price Quantity Value Quantity Value


kg $0.108


10349 $1117.69


4704 $508.03


Variable costs:
Seed
Fertilizer
Nitrogen
Lime t
Insecticide
Herbicide
Tractor
Truck, pickup
Truck, 2-ton
Machinery
Combine
Labor
Irrig. costs
Interest on
above expenses


Total variable costs

Fixed costs:


Tractor
Truck, pickup
Truck, 2-ton
Machinery
Combine
Irrigation


kg 1.87
kg 0.123
kg 0.53
on 18.74
kg 1.72
kg 7.37
hr 6.19
km 0.063
km 0.088
hr 1.48
hr 11.50
hr 3.50
ha 73.13


0.06 457.15


hr 7.79
km 0.063
km 0.078
hr 2.67
hr 36.83
ha 135.77


Total fixed costs
Total costs
Returns to land and management


Source: Wright et al. (1980).


Revenue


20.2
896
168
0.74
16.8
4.5
3.09
79
79
3.09
1.48
5.26
1.0


13.4
616
140
0.74
16.8
4.5
3.09
79
79
3.09
1.00
5.26

316.51


37.77
110.21
89.04
13.87
28.90
33.17
19.13
4.98
6.95
4.57
17.02
18.41
73.13
27.43

484.58


24.07
4.98
6.16
8.25
54.51
135.77
233.74
718.32
399.37


3.09
79
79
3.09
1.00


25.06
75.77
74.20
13.87
28.90
33.17
19.13
4.98
6.95
4.57
11.50
18.41
-

18.99

335.50


24.07
4.98
6.16
8.25
36.83

80.29
415.79
92.24


3.09
79
79
3.09
1.48
1.0








Table 3. Estimated cost of producing 1 hectare of irrigated
corn, north Florida, 1982.


Item Unit Price Quantity Value

Revenue kg 0.112 11436 $1280.83
Variable costs:
Seed kg 2.64 20.16 53.22
Fertilizer kg 0.22 1680 369.60
Lime ton 19.84 0.74 14.68
Insecticide kg 2.09 16.8 35.11
Herbicide kg 8.80 6.16 54.21
Tractor hr 7.96 3.56 28.34
Truck, pickup km 0.063 79 4.98
Truck, 2-ton km 0.094 79 7.43
Machinery hr 1.83 3.56 6.51
Combine hr 12.62 1.48 18.68
Labor hr 3.75 9.88 37.05
Irrigation costs ha 90.37 1.0 90.37
Land rent ha 86.42 1.0 86.42
Interest on 0.075 806.60 60.50
cash expenses
Total variable costs 867.10

Fixed costs:
Tractor hr 11.06 3.56 39.37
Truck, pickup km 0.081 79 6.40
Truck, 2-ton km 0.10 79 7.90
Combine hr 38.52 1.48 57.01
Machinery hr 3.20 3.56 11.39
Irrigation ha 160.98 1.0 160.98
Total fixed costs 283.05
Total costs 1150.15
Returns to land and management 130.68

Source: Eason and Rhoads (1982).








growers in the future. In Florida, efficient water manage-

ment using optimum irrigation scheduling techniques is one

of the main ways that corn growers will be able to increase

productivity.


Nutrient Management for Corn Production


When an improvement in management such as irrigation is

introduced which will increase the potential yield, the

nutrient requirement of the crop can be expected to increase

proportionally (Kurtz and Smith, 1966). Much research relat-

ing applied nutrients to corn plant uptake and nutrient up-

take to corn yields has been obtained under unirrigated

conditions. However, previous research can offer useful

guidelines when studying these relationships for irrigated

corn.

Nitrogen

Nitrogen is needed in large amounts for high corn

yields and often is a major limiting factor to corn pro-

duction especially in the sandy Ultisols of north Florida.

Most N in the soil exists in combination with the soil

organic matter. Rarely is more than 5% of the total N

content of the soil in an inorganic (available) form (Kurtz

and Smith, 1966). The most important forms of available N

are the ammonium ion (NH ) and the nitrate ion (NO3); how-

ever, the nitrate form dominates in nearly all agricultural

soils due to rapid biological nitrification of any ammonium

that may become available through mineralization or







fertilization (Alexander, 1965). Nitrate has been shown to

be highly mobile in practically all soils of the southeast-

ern United States (Rhoads, 1970). Excess water from rains

or irrigation may quickly leach nitrate in highly permeable,

coarse-textured soils because it is not adsorbed by soil

colloids. Thus, since the environmental conditions strongly

dictate the fate of N, very intensive management is required

in order to keep it in the root zone of most Florida soils.

Since available N is very mobile in sandy soils, annual

applications of N fertilizers are needed to obtain high corn

yields. Important sources of fertilizer N include anhydrous

ammonia, urea, ammonium nitrate, N solutions, ammonium sul-

fate, and ammonium phosphates. Under most conditions, all

forms of N fertilizer are equally effective per unit of N

in supplying the needs of a corn crop (Larson and Hanway,

1977).

The environmental conditions, management practices, and

the chemical and physical properties of the soil determine

the type of application schedule that can be used to keep

the applied N in the root zone of the crop. The choices

usually include fall application, preplant application,

single or multiple postemergence sidedress applications, or

application with irrigation water. Stevenson and Baldwin

(1969) applied N to fine-textured soils in the fall and

spring. The spring-applied N was either a preplant or post-

emergence sidedress application. They found that more corn

grain was produced with spring-applied N, with no difference








between preplant or sidedress application. Miller et al.

(1975) did a similar study on a medium-textured soil and

found that postemergence sidedress applications gave the

highest corn yields. Both of these studies were done on

unirrigated corn. Reasons for poor response to fall-applied

N were given as denitrification and leaching.

As soils become more coarse-textured and increase in

permeability, it becomes more likely that corn yield will

respond to postemergence sidedressings of N or N applied in

irrigation water. Anderson et al. (1982) found that corn

yields were higher when N was all applied preplant rather

than during the growing season with limited irrigation on

medium- to fine-textured soils. However, Russelle et al.

(1981) found that when irrigation was light and frequent on

soils of relatively low permeability, sidedressed applica-

tions of relatively low rates of N gave the highest ferti-

lizer use efficiency for corn grain production. Rehm and

Wiese (1975), working on a highly-permeable, coarse-textured

soil, compared conventional preplant and sidedress applica-

tions of N to these same treatments plus supplementary N

added through irrigation water. They found that the appli-

cation of N with irrigation water increased corn grain yield,

and that a higher proportion of the applied N was recovered

with this management practice.

The timing of the sidedressing or irrigation applica-

tions of N becomes important when extremely permeable sandy

soils are involved. Jung et al. (1972) experimented with







N fertilization schedules for irrigated corn and found that

the N applied during either the 5th, 6th, 7th, or 8th week

after planting was most effective as shown by increased

grain and tissue yields. Nitrogen applied after the 8th

week was associated with a distinct reduction in N uptake

and grain and tissue yields. Stanley and Rhoads (1977) also

investigated N application to irrigated corn and found that

grain yield was not affected by application schedule pro-

vided that nutrients were applied before 6 weeks after

planting. Delaying 6 weeks or more caused a significant

yield reduction.

Rhoads (1981b) suggests that the best way to time N

application to minimize nitrate leaching is through program

fertilization, which can be simply defined as making nutri-

ents available to plants as needed for maximum growth rate.

The most critical period for programming N application for

corn is between 4 and 10 weeks after planting. The frequen-

cy of application is dependent on the occurrence of leaching

rains, which are defined as rains that fall in amounts which

have the potential to leach nitrate below the crop root zone.

A specific schedule that has worked well on soils with a

sandy clay loam subsoil (a characteristic of many South-

eastern soils) is to apply one fourth of the N at emergence,

one fourth at 45-60 cm plant height, one fourth at 100-120

cm, and one fourth at 180 cm (Rhoads, 1981b).

The amount of N needed by a corn crop will depend,

among other things, on the variety, the plant population,








the yield goal, the supply of other nutrients, and the

environmental conditions. When estimating the N fertilizer

requirements of a corn crop, one must take into account the

N supplying power of the soil, or potential for N mineral-

ization. The mineral soils of the Southeast do not have

N-supplying powers sufficient to give satisfactory yields,

and considerable amounts of N fertilizers must be added to

obtain high yields. One must also take into account the use

efficiency of the applied fertilizer. Although split appli-

cations of N help to increase use efficiency, recovery of

added inorganic N can be sometimes as low as 25-50% (Allison,

1955). Under the intensive management for high-yield corn

used at the AREC, Quincy, Florida, yields in excess of 15

Mg/ha have been obtained when a total of 336 kg/ha of N was

applied (Rhoads, 1981b).

It is very difficult to compare corn grain yields

versus amount of N applied for different locations and times

because the management schemes, varieties, soil types, and

environmental conditions are so variable. One parameter

that can be used to compare between locations is nutrient

use efficiency (NUE). Nutrient use efficiency is defined as

the grain yield to total nutrient uptake ratio (abbreviated

GNR for N after Rhoads and Stanley, 1981). The concept of

NUE helps identify production practices that maximize yield

response to fertilization. Cropping systems with high NUE

would be expected to be the most desirable due to the high

return (yield) obtained from the investment (fertilizers







and irrigation), and should be more energy efficient (Rhoads

and Stanley, 1981).

The values of GNR for corn calculated from the litera-

ture have a very wide range and have been found to depend on

the yield level and the general availability of soil-water.

According to Kurtz and Smith (1966), GNR values should range

between 30 and 60. Benne et al. (1964) measured total N

uptake and grain yield of unirrigated corn and found that

a yield of 6.7 Mg/ha had an associated GNR of 32, while

Chandler (1960) did a similar study and found that a 7.0

Mg/ha yield had a GNR of 46. Unirrigated corn was also

grown by Hanway (1962a), who found that a GNR of 42 was

associated with a yield of 6.8 Mg/ha, while a 1.2 Mg/ha

yield had a GNR of 33.

Chancy and Kamprath (1982) did 2 years of work with

unirrigated corn, but their results were markedly different

between years due to differing amounts of rainfall. The

first year, rainfall was adequate to prevent severe plant

water stress, and yields ranged from 2.4 to 11.3 Mg/ha. The

corresponding GNR values were 63 and 75. The second year,

rainfall was limited, especially during grain filling. The

yields ranged from 1.4 to 5.3 Mg/ha, and the corresponding

GNR values were 46 and 62.

Several workers have studied grain yield-N uptake

relationships with irrigated corn and have generally found

GNR values to be higher than those associated with un-

irrigated corn. Rhoads and Stanley (1981) used plow-layer








soil-water management and program fertilization with corn

and measured a GNR of 59 from a yield of 11.5 Mg/ha. Rehm

and Wiese (1975) used split applications of N applied as

sidedressed anhydrous ammonia and with irrigation water and

found that a grain yield of 9.4 Mg/ha gave a GNR of 54.

Jung et al.(1972) studied irrigated corn response to time,

rate, and source of applied N and found that yields above

7.0 Mg/ha had an average GNR of 52, while yields below 4.0

Mg/ha had an average GNR of 41. Olson (1980) obtained an

average GNR of 54 from an average irrigated corn yield of

8.8 Mg/ha. Bigeriego et al. (1979) measured an average

GNR of 51 from an average grain yield of 10.9 Mg/ha. Thus,

it appears that increasing yield levels by improving manage-

ment, including the addition of irrigation, will increase

the amount of grain produced per unit of N taken up.

Phosphorus

Phosphorus is considered to be a major nutrient re-

quired for high corn yields, although it is taken up in much

smaller quantities than either N or K. It is generally con-

sidered that plants absorb most of their P as the primary

orthophosphate ion (H2PO4), although smaller quantities of
2-
the secondary orthophosphate ion (HPO4 ) can be absorbed.

The relative amounts of these two ions which will be taken

up by corn plants are affected by the pH of the soil. The
2-
H2P04 ion is favored in acid soils while the HPO ion

dominates in soils with a pH greater than 7.0 (Buehrer,

1932).








Phosphorus in soil can be classified generally as

organic or inorganic, depending on the nature of the com-

pounds in which it occurs. The organic fraction is found

in humus or other organic materials, while the inorganic

fraction occurs in numerous combinations with Fe, Al, Ca,

Mg, F, and other elements. These inorganic forms of P are

usually only slightly soluble in water. Phosphates can also

combine with clays to form generally insoluble clay-phos-

phate complexes.

Virgin Ultisols in north Florida are normally very

deficient in P, and additions of this element are usually

needed most years to obtain high corn yields. Availability

of added P is limited due to rapid P fixation by soils. In

alkaline soils, P can precipitate as dicalcium phosphate or

precipitate on solid phase calcium carbonate. In acid soils,

P can precipitate as Fe or Al phosphates or react with sili-

cate clays and become fixed (Tisdale and Nelson, 1975).

Phosphorus fertilizer sources for corn are primarily

superphosphates and ammonium phosphates. Superphosphates

are most widely used as a single source of P (Larson and

Hanway, 1977). Ordinary superphosphate contains about 8.8%

P and concentrated superphosphates contain 20-22% P. Ammo-

nium phosphates include a wide variety of materials pro-

duced by ammoniation of phosphoric acid.

Liming of acid soils can often help make added P more

available to plants. Woodruff and Kamprath (1965) found

that P adsorption by soils decreased as the soil was limed








due to neutralization of exchangeable Al. Fox et al. (1962)

also found a relationship between lime and P solubility and

availability. The optimum level of liming for improved P

solubility was between pH 5.0 and 6.0. Even though liming

improves P solubility, P still does not move very much in

the coarse-textured soils of the Southeast (Blue, 1970;

Neller et al., 1951). Therefore, leaching of applied P in

limed north Florida Ultisols is not a problem.

Timing of P application is not nearly as important as

timing of N application due to the immobility of applied P

in most agricultural soils. Experiments in Iowa (TVA Pro-

gress Rept., 1960-1963) showed that application of water-

soluble P to corn at any time prior to tasseling was about

equally effective for producing total dry-matter and corn

grain. Significantly less yield response was obtained from

applications more than 40 days after planting. The percen-

tage of P in the plants decreased with time and varied

widely in different parts of the plant but was not influ-

enced significantly by time of application. The primary

advantage of early-applied P seems to be promotion of early

root and top growth, which provides greater exploitation of

soil nutrients and greater utilization of available light

energy in photosynthesis (Caldwell and Ohlrogge, 1966).

Placement of P fertilizers can have an effect on the

availability of P to plants. Traditionally, P has been

applied either by broadcasting and mixing with the surface

plow-layer by tillage operations, or by banding near the row








when the crop is planted. Broadcast applications can be

advantageous because large amounts of fertilizer can be

applied without the danger of injuring the plants, and dis-

tribution of nutrients throughout the plow-layer encourages

deeper rooting (Tisdale and Nelson, 1975). Banded appli-

cations, while encouraging root growth only in and around

the fertilizer band, are more efficient than massive broad-

cast applications in certain high P-fixing soils (Sanchez

and Uehara, 1980).

Kamprath (1967) found that after 7 years, similar corn

yields were obtained where annual banded applications of

22 kg P/ha were made (providing a total of 154 kg P/ha) as

were obtained from an initial application of 350 kg P/ha.

Banding, therefore, saved about half the P requirement. In

soils with extremely high P-fixation capacity and very low

levels of available P, the results are completely different.

Studies by Yost et al. (1979) indicated that banded applica-

tions are inferior to broadcast applications for the first

corn crop planted in soils with the properties listed above.

In Illinois, Welch et al. (1966) found that the relative

efficiency of broadcast P as compared to banded P for corn

on three soils ranged from 0.49 to 1.23. On two of the

soils, higher yields were obtained by a combination of banded

and broadcast P. It was emphasized that as the fertility

level of the soil is increased, the advantage for band

application would be expected to decrease.








The amount of P needed by a corn crop will depend on

the same factors that affect the N needs of the crop (vari-

ety, plant population, yield goal, etc.). When estimating

the P fertilizer requirements of a corn crop, one must have

some idea of the relative availability of the P that exists

in the soil. Mineral soils in the Southeast that have a

long history of P fertilization will tend to have a larger

P-supplying capacity than those soils which have only

recently come into cultivation, simply because the large

amount of P that has been added to them over time has

decreased the amount of P-fixing sites in the soil. The

efficiency of applied P even under low P-fixing conditions

is usually less than 25%, however (Tisdale and Nelson, 1975).

Under the intensive management for high-yield corn used at

the AREC, Quincy, Florida, yields in excess of 15 Mg/ha have

been obtained when a total of 100 kg/ha of P was applied to

supplement the P existing in the soil (Rhoads, 1981b).

In contrast to the number of studies on N use,

research on the utilization efficiency of P has been limited.

The ratio of corn grain yield to total P uptake (GPR), as

with N, varies with yield level and availability of soil-

water. Hanway (1962a) found that a 1.2 Mg/ha grain yield

had a GPR of 119, while a 6.8 Mg/ha yield had a GPR of

228. Studies in Ohio (Anonymous, 1963) showed that about

217 kg of grain was produced for each kg of P contained

in the above-ground portion of a 7.5 Mg/ha corn crop.

Benne et al. (1964) found a GPR of 217 associated with a








grain yield of 7.5 Mg/ha, while Chandler (1960) obtained a

GPR of 281 from a corn crop which yielded 7.9 Mg/ha. All of

the above studies were with unirrigated corn.

Rhoads and Stanley (1981) calculated the 2-year

average GPR value for irrigated corn to be 312. The average

grain yield associated with this value was 11.5 Mg/ha.

Rhoads et al. (1978) obtained a GPR of 330 from a yield of

6.6 Mg/ha. The corn in this study was grown on a deep sand

with plow-layer soil-water management and program fertiliza-

tion. Thus, it appears that, as with N, the improvement of

management practices which improve yields of corn will

increase the amount of grain produced per unit of P taken up.

Potassium

Potassium is the third major soil-supplied nutrient

required for corn growth. As with N, large amounts of K are

required for high corn yields. Potassium is absorbed by

plants as the K+ ion and is found in soils in varying

amounts. Native K in soils originates from the decompo-

sition of rocks containing K-bearing minerals. Soils which

contain large amounts of native K are generally fine-tex-

tured and moderately weathered, while coarse-textured, more

highly weathered soils usually contain extremely small

amounts of K due to excessive leaching.

Of the total amount of K in the soil, only a fraction

is readily available to plants. Soil K can be classified

into three forms termed 1) relatively unavailable, 2) slowly

available, and 3) readily available. Unavailable K exists








as part of the crystal structure of unweathered or only

slightly weathered K-bearing minerals. Slowly available K

is associated with minerals such as illite which appear to

alternately release or fix it, depending on several factors.

Readily available K, the combination of water-soluble K and

exchangeable K, is present either in the soil solution or is

held on the exchange complex of the soil. The amount of K in

this form is usually so small that on many soils containing

large amounts of total K, crops may respond to additions of

K fertilizer (Tisdale and Nelson, 1975). The sandy, highly

leached soils of north Florida frequently are deficient in

total and available K for intensive corn production because

the K+ ion moves quite rapidly in sands (Rhoads, 1981b).

Therefore, K fertilizer must be applied to north Florida

soils to supply most of the crop requirement (Khomvilai and

Blue, 1976). Potassium chloride is the main form of K

applied to corn (Larson and Hanway, 1977).

The retention of native and applied soil K has been

found to depend on soil pH and degree of Ca saturation.

Nolan and Pritchett (1960) reported that the leaching of K

from Lakeland fine sand at pH 4.2 was 1.75 times greater

than at pH 5.3 and 2.75 times greater than at pH 6.3.

Mehlich and Reed (1954) obtained increased K retention by

increasing soil Ca saturation. This was explained on the

basis that K can replace Ca from the exchange complex much

more readily than it can replace H or Al. Khomvilai and

Blue (1976) found that increased amounts of K were held in








three Florida soils when a relatively small amount of lime

was added, but additional increments of added lime decreased

K retention due to competition for adsorption sites by Ca

and to a lesser extent Mg.

Soils of north Florida have cation exchange properties

dominated by organic matter and hydroxy Al and Fe. Native

CEC is usually low, but liming these soils can increase the

CEC by neutralizing H ions from the hydrolysis of Al or Fe

and by displacing hydroxy Al and Fe from the colloidal sur-

face, thus freeing exchange sites (Kamprath, 1970). As soil

pH increases, carboxyl and phenolic radicals on organic

matter lose H+ and become negatively charged, also increasing

the CEC of the soil (Coleman and Thomas, 1967). However,

these sites have been found to have a greater affinity for

polyvalent cations (Broadbent and Bradford, 1952). There-

fore, activated organic matter exchange sites are not highly

effective in retaining K.

In medium- to fine-textured soils with high CEC, K is

commonly applied in the fall of the preceding year, before

planting in the spring, or at planting. These methods are

usually more efficient on the soil types mentioned above

because K does not move down through them very rapidly and

the opportunity exists for the K to be incorporated into the

soil (Tisdale and Nelson, 1975). However, Florida soils are

different due to their sandy nature and low CEC values.

Fertilizer K applied in its entirety before planting has an

excellent chance of leaching below the crop root zone with








heavy rains. Therefore, split applications and timing

applications to coincide with maximum uptake by the crop

have been recommended in Florida (Blue, 1973). Rhoads

(1981b) has found that two applications of K appear to be

adequate if one third of the K is applied at planting and

the remaining two thirds 6 weeks after planting. Since corn

takes up almost 90% of the K that it needs by the silking

stage (about 9 weeks after emergence), it is important that

all K applications be made by several weeks before this

stage to insure that the crop will be able to meet its

needs (Rhoads, 1981a).

Potassium fertilizers, like P fertilizer, can be ap-

plied by broadcasting and incorporating or by banding next

to the row. The advantages of each method for K are the

same as those for P. Placement of K is important from the

standpoint of corn germination, because reduced germination

may result if comparatively large amounts of K are placed

too close to the seed (Cummins and Parks, 1961).

Barber (1959) found no significant differences in corn-

leaf K concentration between banded or broadcast applica-

tions of K. Similarly, Wittels and Seatz (1953) found that

yield, stalk breakage, and K content of the corn plant were

not affected by K placement. However, in both of these

studies the response to K was not large, and it would be

difficult to measure any yield differences due to placement

under these conditions.








Parks and Walker (1969) found that banding K gave

higher corn yields than broadcasting, but that the effect

of placement on yield decreased as the level of soil-test K

and the total K available for the plant increased. On soils

testing low in K, Parks et al. (1965) reported that higher

corn yields were obtained with row-applied K than with

broadcast K. The yield differences due to method of appli-

cation were greater at low rates of K than at high rates.

On soils testing high in K, there was neither a response to

method of application nor amount of applied K. In a study

in Illinois, Welch et al. (1966) found that the relative

efficiency of broadcast K as compared to banded K ranged

from 0.33 to 0.88.

The amount of K needed by a corn crop, as with N and P,

will depend on such things as the variety, the plant popula-

tion, and the yield goal. The amount of K fertilizer that

needs to be applied will not necessarily depend on the total

amount of K present in the soil, but will depend on the

amount of the total K that is in an available form to the

plants. In Florida, the available K contents of the highly

leached, sandy soils are very low, and even under cropping

systems where K removal is small or recycling occurs, applied

K usually does not accumulate (Blue, 1974). However, the

efficiency of applied K is high, usually between 50 and 75%

(Tisdale and Nelson, 1975). Under intensive management of

irrigated corn in north Florida, grain yields above 15 Mg/ha








have been obtained when 280 kg/ha of K was applied (Rhoads,

1981b).

Research studying the utilization efficiency of K as

related to corn grain yield has given much more variable

results than the same type of research involving N or P.

The ratio of corn grain yield to total K uptake (GKR) has

shown no dependence on irrigation or yield level. Barber

and Mederski (1966) implied that corn should produce about

50 kg of grain for each kg of K taken up. Working with

irrigated corn, Rhoads and Stanley (1981) obtained a GKR

of 50 from a 2-year average grain yield of 11.5 Mg/ha.

Rhoads et al. (1978) found a GKR of 61 associated with a

corn yield of 6.6 Mg/ha using plow-layer soil-water manage-

ment and program fertilization on a deep sand. Sparks et

al. (1980) also grew irrigated corn and obtained a 2-year

average GKR of 92 from an average yield of 6.7 Mg/ha. In

unirrigated corn studies, Hanway (1962a) found GKR values of

37 and 105 associated with yields of 1.2 and 6.8 Mg/ha,

respectively, while Benne et al. (1964) obtained a GKR of

33 from a grain yield of 7.5 Mg/ha and Chandler (1960) found

a GKR of 56 associated with a grain yield of 7.9 Mg/ha.

The translocation pattern of K from other plant parts

to the grain during the grain-filling period may have caused

the extremely variable results illustrated above. According

to Hanway (1962a), translocation of N and P to the grain as

it matures occurs in large amounts, while translocation of K

is much less. At maturity the corn grain contains about








62-72% of the total N and 72-82% of the total P in the

plant, but only about 31-44% of the total K. Thus, there

is a potential for more variability in the amount of trans-

location of K between other plant parts and the corn grain

which could cause the variability in GKR values.

Magnesium
2+
Magnesium is absorbed by corn plants as the Mg ion

and is considered a secondary nutrient. Magnesium is needed

in relatively small amounts by corn. Olson and Lucas (1966)

estimate that a 6.3 Mg/ha corn crop will take up a total of

25 kg/ha of Mg. Magnesium is normally in low supply in the

coarse-textured soils of north Florida. These soils usually

contain only small amounts of exchangeable Mg, a condition

that can be aggravated by the addition of large quantities

of fertilizer salts which contain little or none of this

element. Any exchangeable Mg in these soils is released by

ion exchange when these fertilizers are added, and large

quantities of chlorides or sulfates accelerate Mg removal by

leaching (Tisdale and Nelson, 1975).

Magnesium deficiency can be prevented or eliminated by

use of dolomitic limestone or soluble Mg fertilizers (Blue,

1974). Levels of Mg for good corn growth are usually avail-

able when pH is corrected to 6.0 or higher with dolomitic

lime (Wright et al., 1980). The lime is normally applied

well in advance of the time of planting of the corn crop

and is incorporated into the soil.








Calcium

Calcium is absorbed by corn plants as the Ca2+ ion.

Like Mg, Ca is considered to be a secondary nutrient. In

limed agricultural soils of north Florida, sufficient Ca is

usually available and is not a limiting factor to high corn

yields. Levels of Ca for good corn growth can be found

where calcitic or dolomitic lime has been applied to raise

the pH of the soil to 6.0 or higher (Wright et al., 1980).
2+
The reactions in which Ca2+ can be involved in the soil are

very similar to those of Mg2+ mentioned previously.

Sulfur

Sulfur, another secondary nutrient, is absorbed by corn
2-
plants as the SO ion. Sulfur is a constituent of organic

matter and since many highly weathered soils in north Florida

are low in organic matter, S deficiencies should be expected

(Blue, 1974). Since the available form of S is an anion, S

is easily leached from sandy soils similar to nitrate.

Wright et al. (1980) recommended that to eliminate any

possibility of a S deficiency, about 40 kg/ha of S should be

applied to north Florida irrigated corn annually in split

applications as with N. Generally, half of the S may be

applied preplant with the remainder applied as a sidedressing

or through the irrigation system with N.

Micronutrients

In north Florida, the micronutrients Fe, Mn, and Mo in

most years exist in sufficient quantities in the soil such

that they do not limit irrigated corn yields. It has been








recommended that a micronutrient mixture should be applied

to irrigated corn in north Florida at planting time to

satisfy Zn and Cu needs. A total application of about

6 kg/ha of Zn is needed for high corn yields. Increases

in yield have been noted for corn when B was applied. This

element should be applied as a sidedressing or through the

irrigation system in three to four applications at a total

rate of about 2 kg/ha (Wright et al., 1980).


Soil Testing


Soil testing can be considered to be any chemical or

physical measurement made on a soil, but has been given both

restricted and broad meanings. The term is restricted in

the sense that it has come to mean rapid chemical analysis

to assess the available nutrient status of the soil, and

broadened to include interpretations, evaluations, and

fertilizer recommendations based on results of chemical

analyses and several other considerations (Melsted and Peck,

1973).

Of the elements required for corn production, N, P,

and K are the most deficient in Florida mineral soils. Soil

pH is also a common limitation to plant growth, but if the

soil has been limed, acidity will not be a problem and Ca

and Mg levels will usually be high enough for good corn

growth. Sulfur, Zn, and B can also be deficient in Florida

soils, but yearly requirements of these elements are small

and annual maintenance applications are usually made to








irrigated corn without the benefit of a soil test. The

complex nature of soil N precludes the use of rapid, inex-

pensive, and reliable tests to measure this element (Whitty

et al., 1977). Because Florida mineral soils do not retain

available N for long periods of time, the normal recommenda-

tion is to apply N yearly to field crops without first

making a soil test. Soil testing, therefore, predominantly

involves P, K, Mg and pH in Florida (Kidder, 1981a), with

secondary and micronutrient analyses varying widely depend-

ing on location.

The nutrient status of a soil may be evaluated by

several means, among which include field plot fertilizer

trials, greenhouse pot experiments, plant tissue testing,

and rapid chemical analysis of the soil (Melsted and Peck,

1973). There are some limitations to each of these methods.

Fertilizer field trials cannot be conducted on every grow-

er's field, and the results may not be applicable to other

fields where management practices are different. Greenhouse

fertilizer experiments often yield results that cannot be

quantitatively extrapolated to the field. Plant tissue

testing can help explain what was wrong at a certain time

on a specific soil but does not quantitatively predict

fertilizer needs. In contrast, rapid chemical analysis, or

soil testing can be fast, inexpensive and accurate, and can

help transform field, greenhouse, and laboratory research

into reliable predictions of lime and fertilizer needs

before crops are planted (Melsted and Peck, 1973). However,








a soil-testing program requires a very large amount of back-

ground research before it can become dependable.

The types of information needed to support an effective

soil-testing program include identification of the signifi-

cant chemical forms of the plant nutrients in the soils of

the area, the extractants most suitable for accurately

measuring the available nutrient forms, field sampling tech-

niques, soil-testing procedures, the relative productive

capacity of soils for various crops grown in the area, and

the differential response of various rates and methods of

fertilizer application for different crops (Melsted and Peck,

1973). Of these six types of information, the first four

would probably not change significantly with changing crop

management practices over time. However, the last two would

probably be altered as management practices designed to give

high yields (e.g. program fertilization, irrigation, in-

creased planting density) are introduced into the overall

production system. Thus, soil-testing programs need to

remain current with respect to management practices, and

research relating yields of crops under new production

schemes to soil-test levels is necessary to determine if

new calibrations are required.

Most experiments that have investigated relationships

between crop yield or yield responses to applied fertilizers

and soil-test results have been done under unirrigated

conditions (Bray, 1944; Dumenil, 1952; Hanway, 1962c; Grunes

et al., 1963; Rouse, 1968), but a few have been done with








irrigation (Grunes et al., 1963; Liebhardt et al., 1976).

However, even the most modern of fertilizer recommendations

based on soil tests have originated from data that were

obtained from field trials that were not under irrigation

(Plank, 1978; Whitty et al., 1977; Cope et al., 1980).

Current literature in this area is void of information

relating high irrigated corn yields to soil tests and

applied fertilizers on soils of the southeastern United

States. It has been stated that when an improvement in

management (such as irrigation) is introduced which will

increase the potential yield, the nutrient requirement of

the crop can be expected to increase proportionally (Kurtz

and Smith, 1966). Thus, as the amount of agricultural land

area under irrigation increases, soil-testing research under

irrigated conditions will become more necessary.

Testing Soils for Phosphorus

As previously mentioned, P compounds in soils are only

slightly soluble. Because of this, the amount of P in the

soil solution at any given time is small. If growing plants

absorb P only from the soil solution, in order for normal

growth and P uptake to occur the soil solution must be

renewed several times each day during the growing season

(Barber, 1962). The factors involved in the renewal of the

soil solution are the amount of P which can be solubilized,

its degree of solubility, and the rate of its diffusion from

the solid surface to the plant root surface (Thomas and

Peaslee, 1973).







Extractants for soil P are available from which both

P supply and solubility can be estimated. The kinds of P

compounds present and other chemical properties of the soil

will strongly influence the characteristics of the extract-

ing agent. Soils of different climates (e.g. arid or humid)

would be expected to have a different distribution of P

compounds. North Florida Ultisols are acid, highly weathered

soils which are dominated by Al- and Fe-P and contain very

little Ca-P.

When P is added to soils, it is usually made less

soluble by adsorption or precipitation reactions. Phosphorus

can become adsorbed on calcium carbonate (Olsen, 1953), Fe

and Al oxides (Dean and Rubins, 1947), or clay minerals.

With time, it has been shown that the adsorption of P can

change to preciptiation (Low and Black, 1947). Haseman

et al. (1950) showed that a number of crystalline phosphates

could be identified when soils were reacted with strong

phosphate solutions.

Extractants are available which can selectively remove

various fractions of the total P in soils while leaving

other fractions untouched. It is necessary that the extrac-

tant removes all or a proportionate part of the available

form or forms of P from soils with variable properties and

extracts them in an amount that is correlated with the

growth and response of each crop to P under various condi-

tions (Bray, 1948). Soil-test extractants for P can be

classified into one of several categories based on the








chemical nature of the extracting solution. The four main

categories of P extractants are 1) dilute concentrations of

strong acids such as HC1, HNO3, and H2SO4; 2) dilute concen-

trations of strong acids plus a completing agent such as F ;

3) dilute concentrations of weak acids such as citric,

lactic, or acetic acids; and 4) buffered alkaline solutions

such as NaHCO3.

According to Kamprath and Watson (1980), there are four

basic reactions by which P in the solid phase in soils is

removed. The first reaction is the solvent action of acids.

The acid solutions used provide sufficient H ion activity

to dissolve Ca phosphates and will also solubilize some Al

and Fe phosphates. The order of greatest solubility in acid

solutions is Ca-P >Al-P >Fe-P (Thomas and Peaslee, 1973).

The second reaction is anion replacement. Phosphorus

adsorbed on surfaces of calcium carbonate and hydrated

oxides of Fe and Al can be replaced by such ions as acetate,

lactate, citrate, sulfate, and bicarbonate (Dean and Rubins,

1947; Olsen et al., 1954). When the organic anions and

sulfate are present in acid solutions they reduce readsorp-

tion of P. The third reaction is completing cations binding

P. Fluoride ions are very effective in completing Al ions

(Chang and Jackson, 1957) and will also precipitate Ca

(Thomas and Peaslee, 1973). Thus, P in soils containing

Al-P or CaHPO4 will be extracted by solutions containing F

ions. The fourth reaction is hydrolysis of cations binding

P. Extracting solutions containing OH ions extract P from







Al-P and Fe-P due to the hydrolysis of Al and Fe. Thus

NaHCO3, which is buffered at pH 8.5, is effective in

extracting P from Al and Fe phosphates.

The extractant used to test soils for P in most of the

southeastern states including Florida is the double-acid, or

North Carolina extracting solution (0.05N HC1 in 0.025N H SO4)

(Sabbe and Breland, 1974). This extractant appears to work

best on soils with low CEC values that are relatively highly

weathered and contain Al-P and/or Ca-P (Thomas and Peaslee,

1973). In soils containing appreciable amounts of Ca-P, the

majority of P extracted comes from the dissolution of this

form. However, in soils containing only small amounts of

Ca-P, Al-P is the principal form of soil P removed (Kamprath

and Watson, 1980). For acid to neutral soils, Al-P is the

primary source of plant P along with any CaHPO4 that may be

present (Kamprath and Watson, 1980), so the type of P

extracted by the double-acid solution and the source of P

for plant growth in acid soils correspond well. In addition,

it has been reported that the double-acid gave much better

correlations with estimates of labile P and percent yield on

soils which had predominantly kaolinitic clay minerals as

compared to soils which had 2:1-type clay minerals (Fitts,

1956). Thus, the double-acid extracting solution is well

suited for use with acid, kaolinitic, highly weathered

Ultisols of the southeastern United States.








Testing Soils for Potassium

As previously mentioned, only a small fraction of the

total K in the soil is readily available to plants. The

remaining K can be classified as slowly available or rela-

tively unavailable. Relatively unavailable K is a component

of some of the primary minerals in the soil. The K in these

minerals is released slowly by weathering and the rate

usually is not rapid enough to be of significance for an

immediate crop (Doll and Lucas, 1973). Slowly available K

is found in secondary minerals such as illite, vermiculite,

and chlorite. These forms of K tend to be released to a

readily available form when the levels of this form in the

soil decrease (Doll and Lucas, 1973). Readily available K

includes both water soluble K and exchangeable K.

The readily and slowly available forms of K together

may compromise the majority of the K that is available for

plant uptake during the growing season. It was shown by

Tabatabai and Hanway (1969) that the rate and amount of non-

exchangeable K released during intensive greenhouse cropping

was directly related to the clay content of several Iowa

soils. Most of the nonexchangeable K that becomes readily

available originates in the clay fraction, but in some soils

the silt fraction can contribute significantly (Doll et al.,

1965). Liebhardt et al. (1976) and McLean and Simon (1958)

have shown that when exchangeable K is removed from the soil,

the release of nonexchangeable K can be quite rapid. On two

of three soils tested by McLean and Simon, the amount of K








released over a 49-day period after the original exchangeable

K was removed was greater than the amount initially present

in the exchangeable form.

Under conditions of intensive cropping, uptake of K

from different soils has been correlated with the initial

level of exchangeable soil K (Conyers and McLean, 1969).

Additional field studies showed excellent correlation be-

tween exchangeable soil K and crop yield and crop content of

K (Bray, 1944; Hanway, 1962c). However, because of differ-

ences between soils, response curves for different crops on

specific soils of a given area need to be determined if

reliable fertilizer recommendations are to be developed from

soil tests. Differences in soil texture, especially clay

content, and types of minerals in the soil have marked

effects on the rate of K fertilizer needed for different

soils (Doll and Lucas, 1973).

The most common methods used for determining exchange-

able K usually include water-soluble (soil solution) K, but

the amount of K in this form is so small in relation to

exchangeable K that the two forms are normally determined

together and referred to simply as "available" K (Doll and

Lucas, 1973). Available K can be extracted with neutral

normal ammonium acetate (1.ON NH4OAc buffered at pH 7.0),

but extractions with cold dilute acids such as HC1 or H2SO4

will remove about the same amount of K as NH4OAc (Doll and

Lucas, 1973). Thus, the double-acid extracting solution









used in Florida for P determination is also a satisfactory

extractant for K determination.

Testing Soils for Magnesium

It has been shown for some New Jersey soils that the

total content of soil Mg is about the same as that of K, but

a greater proportion of the soil Mg is usually in an ex-

changeable form (Bear et al., 1945). In these soils, the

main sources of Mg and K are primary and secondary minerals.

In limed north Florida Ultisols, the total content of Mg

would be expected to be significantly higher than the total

K content due to differences in the solubility of the major

source of each element. The main source of Mg is dolomitic

lime which is only slightly water-soluble, while the main

source of K is K fertilizer which is very water-soluble and

can easily leach from sandy soils (Rhoads, 1981b). As

stated earlier, levels of Mg for good plant growth are

usually present when the soil pH is corrected to 6.0 or

higher with dolomitic lime (Wright et al., 1980).

The immediate source of available Mg is the Mg that

exists on the exchange complex of the soil (Doll and Lucas,

1973). The standard method of extracting the available Mg

is with neutral normal ammonium acetate (Doll and Lucas,

1973). However, Sartain et al. (1976) have shown that for

Florida soils the double-acid extracting solution removes

amounts of Mg that are very similar to those amounts ex-

tracted by NH4OAc at either pH 7.0 or 4.8. Thus, the three

main plant nutrients that Florida soils are tested for







(P, K, and Mg) may be extracted and measured using one

extraction with the double-acid solution.

Information Necessary for Calibrating Soil-Test Results

If soil tests are to be used as a guide for crop pro-

duction, then the relationship between soil-test values and

crop response needs to be investigated for specific crops on

specific soils under the environmental conditions where the

crop is to be grown. Evaluation of soil-test values based

on a soil-chemical factor such as cation-saturation ratios

do not take into account plant differences, and greenhouse

experiments with seedling plants under artificial conditions

such as soil in pots do not represent what happens in the

field (Cope and Rouse, 1973). Field experiments are neces-

sary because they are the only way that all interactions

between the plant, the soil, and the environment can be

integrated into the soil-test calibration process.

Soil tests can be calibrated using either short-term or

long-term field experiments. Short-term experiments, con-

ducted for one year, are useful in interpreting the rela-

tionship between soil-test values and sufficiency for maxi-

mum yield. According to Cope and Rouse (1973), a primary

objective of these types of experiments is to determine the

crop yield from an adequate application of a nutrient and

the relative yield from plots receiving none. The soil-test

values may then be plotted versus relative yields produced

and a statistical procedure used to determine critical

values for each nutrient (Cate and Nelson, 1971) or define








ranges for classifying levels of soil-test nutrients (Nelson

and Anderson, 1977). Short-term experiments have the limi-

tation that they provide no measure of the cumulative

effects of fertilizer treatments on yield or soil buildup

or depletion. However, long-term experiments can supply

this information. When experiments are done over many years

at the same location, yield responses during favorable grow-

ing seasons where the nutrient being studied is the major

controlling factor in determining yields are almost assured.

As soils are fertilized year after year, their management

has an increasing effect on their fertility (Cope and Rouse,

1973). Long-term experiments can therefore be more control-

lable and can give suitable data to relate soil-test levels

to actual yield instead of relative yield.

Rating Soil-Test Values

Most soil-testing laboratories use some kind of rating

system for evaluating soil-test values. Many use the de-

scriptive terms "low," "medium," and "high" and assign a

range of soil-test values to each term in order to classify

the nutrient status of a soil and to aid in fertilizer

recommendations. The assignment of specific values to each

descriptive term depends on the soil texture or type and the

amount of fertilizer recommended depends on the type of crop

to be grown and the management practices to be used (Whitty

et al., 1977).

Another type of rating system was developed in Alabama

by Rouse (1968). He proposed the use of a combination of








ratings defined in terms of relative yield and a fertility

index expressed as percent sufficiency. According to Cope

and Rouse (1973), this system is desirable because it indi-

cates the expected relative yield that would be obtained

without the addition of any of the nutrient in question.

The index values below 100 indicate the percent sufficiency

of an element in the soil without the addition of the ele-

ment as fertilizer. Index values above 100 indicate that

the soil supply of an element is high enough not to limit

yields and indicates the nearness of the supply to an

excessive level. The values are referred to as the "Soil

Fertility Index" and are reported to the nearest multiple of

10 from 0 to 990. Relationships among soil-test ratings and

their definitions, fertility index values, and relative

yields are shown in Table 4.

Soil-test ratings for three Southeastern states, each

of which use the double-acid extracting solution in their

soil-testing laboratories, are shown in Table 5. The ratings

are those which were derived for sandy coastal plain soils

with low CEC in each of the three states. The ratings for

P are very similar between the three states, but the ratings

for K and Mg are much lower in Alabama than in Florida or

Georgia. This variation could reflect differences in soil-

test calibration philosophy, management practices, climate,

forms of nutrients in the soil, or soil properties between

the states. In the discussion of K by Whitty et al. (1977)

in Florida, there seemed to be quite a bit of concern about
















0
W

5i

'o




ra

,-




(Z
-1
4-)

r4
0)











-,-

-O
r4

















U,

4 E



LIo
-
.-4
-l

O-I
0


















0,
4-1


tn
C









CO 4
-rIW


>U)
4-) -H
cn





m 0
*rf









0 C
C




(O








E-4
om


0
-rH
-
Cd




0














O4-



>i



0
S4-)
-H

(0 u


-rl








4J -
CH C





*H3
X-





a) 0
h -


4-a
O













0 4
U)rc


04






04 0


o0 0



4J (U (
Q4 a 0 0
C o d

) -H 5.- 0
N 90 a) 4-4
-H r-l N
iq- -H

4-1 *> -H -q

W1) 5--l I 3
44 -H )
O 44 a)


(0 0 0
i-l 4- -H
c 0 *Hl t


N 0

-IH r *

4- >i c
54 WJ Q-e
W) -H ,! 0


V 0


0 0U
-H r-H -i


OWU N
-H -H r-
-l i-I4 0-
u U) -H -H

Cd Q) 5-iV '
v- 0 4-
(40 W
3 -H W) --
c E
S-i-)CO -H
4 ct m *H


1-I
-l 0)



0


,co
V -H


0


0


N



-H
Q) 40
N


-H ,




0 0 -,
Z ,.C


04

H 0
EU

C(
0 W


-4rl 40
4-1


M-I-)-H
U 4-)


'1 C
4 rO
4 0 Q)
(0 41



-H >



-3
NC- r
I-a0




> <6


0
0










r--





0



-I
o\













o
-l




M

















5)
-3












-1
4









-vI













CJ




















(10)
.I1
,<

4J



























0
4U)
14




























0 0
10)






a-



















4-)
3
i-(









r 6
<~a





>i


t
E -4





>


r- A



0
CO 0
C0 CO





ON
r-'



I N
o\

I r-'




N A
r.4










I


N


I r






in
Co
LO A


0
o
00



.--. --


,I O

-< O)
0 0 0





000
k0 4Q
'0 ( Cl








rapid losses through leaching and crop uptake. Cope et al.

(1980) in Alabama seemed much less concerned about K losses,

and even stated that K may accumulate in soils where recom-

mended rates of K are applied.

Fertilizer Recommendations Based on Soil Tests

When fertilizer recommendations are to be made from

soil-test values, it is a good idea if the recommended

amounts are larger than what response data indicate they

should be. This is known as adding a "safety factor," and

the reason for it lies in the fact that losses to a grower

from the use of too little fertilizer are normally much

greater than from the use of more than is needed (Cope and

Rouse, 1973). The general procedure for determining the

amount of fertilizer to be added to soils in known soil-test

rating classes is to examine fertilizer response data from

soils varying widely in fertility and estimate the average

amount of fertilizer to which crops grown on soils in each

rating class will respond. These amounts are then normally

increased by the safety factor before being put into a

recommendation table.

Fertilizer recommendations for corn from Florida, Geor-

gia, and Alabama are shown in Table 6. Although the recom-

mendations for all three states have been based on field

data obtained under unirrigated conditions, an attempt has

been made to supply recommendations for irrigated corn grow-

ers by increasing the values for unirrigated corn. Florida

attempts to adjust their recommendations for plant population.





49


000 000 0 0 0 0


m fn 0 00 m r- m m
-I CN N N mm 00


r- o io Ln r-
m Irn in Ln %.0 f- m


CN (NODa -W N
ri r -l 1
rH f-i r-


Sm 0 c
r-im m %. co
1-1 r-I 1-1 t-q r-q


I I H,
N


000 000 0 0 0 0


oo c o oo 0 0o
0N L L 0O 0 O 0 L

ON rI rN rl C4 o

0 Ro\ m 0)10 0 0 0 a)
cM(N(N r Ncrm c N m CN (N


m om m a) o a) r- 0m om
m 1- Ln M No r- m L (N LA


Soo 010


I I


cN -*, r- r- m N 0 (N -iM Ln Lr-Ao 0 D Ln m 0
--l r-l -l N (V r i C(N r N





(N '3L Ln D r-
NM I I 3n 1o


S-4 4J


0 fl C)


000
000
000


Nmm


4- 41 4-1

0 (0 C




000

0 D
m\~f
1-:31


0

O
U


4-r


--4


I

0
Cz


.0
Z


ro
-,l

04
0

u3


0
N


m in



o 0


I >

0 -P


me








Georgia attempts to adjust their recommendations for yield

goal. Their recommendations presented are for an expected

yield of 6.3 Mg/ha. For expected yields less than 6.3 Mg/ha,

the recommendations should be reduced by 25%. For expected

yields greater than 6.3 Mg/ha, the N rate should be increased

by 68 kg/ha, the P rate by 10 kg/ha, and the K rate by

37 kg/ha for every 3.1 Mg/ha increment over 6.3 Mg/ha. For

irrigated corn, it is recommended that the rates of P and K

be increased by 25% and this has been shown in the table

(Plank, 1978). The lower recommendations for Alabama corn

as compared to those for Florida and Georgia corn may be due

to differences in expected plant populations, soil nutrient-

holding capacities, expected moisture during the growing

season, or yield goals.


Multiple Cropping


In the southeastern United States, the form of multiple

cropping used most often is double cropping, where total

production from a given area of land during one farming year

is maximized by growing two crops in sequence. In a sense,

double cropping helps growers increase the amount of land on

which crops are grown, thereby helping to meet expanding

world food demands. At the same time, the systems require

careful management if they are to be successful and profit-

able. Land use becomes more intensive, and this encourages

more efficient utilization of machinery, labor, and capital

investment (Lewis and Phillips, 1976).








Double cropping in the Southeast is possible due to the

long growing season. If the first crop, such as corn, is

planted in late February or early March it can be removed by

early to mid-July, leaving approximately 120 to 140 frost-

free days for a second crop (Wright, 1982). No-tillage

planting also contributes to the success of double cropping

by allowing the second crop to be established quickly.

Additional advantages of minimun tillage include reduced

labor and fuel requirements, increased flexibility in timing

of farm operations, conservation of soil moisture, and

control of erosion (Lewis and Phillips, 1976).

In the past, southeastern double-cropping systems have

consisted of small grains such as rye (Secale cereal L.),

barley (Hordeum vulgare L.), wheat (Triticum aestivum L.),

or oats (Avena sativa L.) grown as winter forage or grain

crops followed by corn, grain sorghum, or soybeans. If the

first crop is utilized as forage, it can be removed early

enough so that conventional tillage practices and recom-

mended planting dates for summer crops can be followed. If

the first crop is harvested for grain, the delayed planting

of summer crops may result in reduced yields (Nelson et al.,

1977). Thus, minimum tillage practices may help in this

type of situation.

Nelson et al. (1977) conducted a study of the effects

of conventional tillage and no tillage on double-cropping

systems involving wheat or barley followed by either corn

or grain sorghum at two locations in Georgia. Corn and








grain sorghum yields did not significantly differ between

tillage treatments which were planted on the same date.

No-till corn and grain sorghum produced higher yields when

early-planted (after small grain for forage) than when late-

planted (after small grain for grain). Irrigation was found

to increase conventionally tilled corn yields by 31% and

grain sorghum yields by more than 20%. Irrigation also

increased yields of no-till corn and grain sorghum, but the

increases were of smaller magnitude than for conventional

tillage. This was attributed to the fact that the soil-

water content in the unirrigated no-till treatment remained

higher relative to the unirrigated conventionally tilled

treatment.

Double-cropping systems consisting of irrigated corn

for grain as the first crop followed by soybeans or grain

sorghum are now being used in north Florida. According to

Wright (1982), irrigation water is generally the key factor

in making two grain crops in one season. Studies in Quincy,

Florida (Wright, 1982) have shown that grain sorghum and

soybean yields can be high (above 7.0 Mg/ha grain sorghum

and 2.5 Mg/ha soybeans) as long as they are planted in July.

With good hybrids and good management, grain yield of one

variety of corn planted 27 February was 17.7 Mg/ha followed

by 7.6 Mg/ha of grain sorghum planted on 28 July. A com-

bination of these two crops resulted in a combined grain

yield of over 25 Mg/ha in one season. Only 112 kg/ha of N

was applied to the grain sorghum with other nutrients coming







from the residual fertilizer applied to the corn. Little

difference was noted in grain sorghum or soybean yields when

either no-till or conventionally planted.


Summary


Much of the research that has dealt with water and

nutrient management practices for corn has had a common aim:

to make sufficient amounts of water and nutrients available

for uptake at the times during the growing season when they

are needed, while simultaneously increasing the efficiency

of their use. Practices which have been developed with this

objective in mind include plow-layer soil-water management,

multiple applications of nutrients, fertilization with irri-

gation water, liming, and banded or broadcast application of

fertilizers. Using a combination of these management prac-

tices, irrigated corn can be produced with an economic advan-

tage to growers.

Intensive soil-water management is necessary to obtain

high corn yields in north Florida due to low soil-water

holding capacities of surface soils and periods of infre-

quent rainfall. Corn cannot undergo extended periods of

water stress during any part of the growing season without a

reduction in grain yield. The use of plow-layer soil-water

management, which consists of recharging the plow layer to

field capacity when tensiometers placed at 15-cm depth reach

a value of 0.02 MPa of soil suction, can assure that water

will be available to the crop while helping to prevent the

leaching of plant nutrients.







Since most of the available N in Florida mineral soils

exists as NO3, which will not be adsorbed by negatively-

charged soil colloids, a comprehensive effort must be made

to keep a sufficient supply of N in the crop root zone for

high yields. Availability of N can be increased with the

use of multiple applications of N fertilizer timed with the

needs of the crop or with fertilization with irrigation

water. When multiple applications of N are used in combina-

tion with plow-layer soil-water management, losses of N

through leaching are minimized and N fertilizer use effi-

ciency increases.

Phosphorus availability can become limited because of

the insolubility of P compounds in the soil. Applied P does

not move very much in the soil, thus the timing of applica-

tion of P is not as important as with N. Management prac-

tices that have been developed to increase P availability

include liming and banded or broadcast application of P

fertilizer. Liming to a pH between 5.0 and 6.0 can increase

the solubility of some P compounds in the soil. Broadcast

applications of P fertilizer can generally supply sufficient

available P in low P-fixing soils, but in high P-fixing

soils P availability and fertilizer P efficiency may be

increased using banded applications.

Since the available form of K is a cation which can be

adsorbed by negatively-charged soil colloids, movement of K

in the soil is not as great as N in the NO3 form. However,

K can leach in Florida mineral soils due to their low value








of CEC. Fertilization schedules where the K application is

split one or more times have been developed to increase K

availability in the soil. It is extremely important to make

K available from the time of planting through the silking

growth stage for corn, because about 90% of the total K in

a mature corn plant is taken up during this time period.

In order to determine quantities of nutrients that

should be applied to soils such that high corn yields can

be obtained, the nutrient status of the soil should be

evaluated. This can be accomplished in a fast and inexpen-

sive way using rapid chemical analysis, or soil testing.

Soil-test calibration is necessary before results of soil

tests can be helpful in making fertilizer recommendations.

Calibration involves the establishment of relationships

between crop response and levels of soil-test nutrients and

applied fertilizers. After this has been accomplished,

soil-test rating systems may be created and used in making

fertilizer recommendations.

While much information has been obtained which has

improved irrigation efficiency and availability of applied

water and nutrients for corn, this alone is not enough to

economically optimize the fertilizer input to high corn

yields. Research relating fertilizer rates and residual

(soil-test) nutrients to high irrigated corn yields is

necessary so that optimum quantities of nutrients required

to produce maximum yields can be determined. As this infor-

mation becomes more developed and refined and is combined





56


with management practices designed to increase availability,

maximum irrigated corn yields should be produced more

efficiently at greater profits to growers.














CHAPTER 2
MATERIALS AND METHODS

Field Procedures


A double-cropping system consisting of corn followed

by either rye or soybeans was established at the University

of Florida Agricultural Research and Education Center at

Quincy, Florida,in 1980. The following year the same system

was repeated, except that grain sorghum was grown in place

of rye. The land selected for the system was a tract of

approximately 0.6 ha which had not been previously cropped.

The soil was a Ruston loamy fine sand (fine-loamy, siliceous,

thermic Typic Paleudult), and the entire area was under a

center pivot irrigation system. Selected soil properties

are shown in Table 7, and soil-water characteristic curves

are shown in Fig. 1.

In the summer and fall of 1979, the land underwent

preparation for the following spring planting. On 29 August,

the soil was turned with a bottom plow. Soil samples were

taken from each future experimental plot on 4 September.

A total of fifteen 2.5-cm cores taken to a depth of 15 cm

were removed from each plot and the soil was analyzed for

extractable Ca. Depending on the level of Ca already in the

soil, fine calcitic limestone was applied to each plot on













LA)
LA

-S


S5,












0









0












4 C











rd







-I)
c"














































ru
U
ra
+J




U-














4J
&1


Q,


I I I










I I I


0-
tr




Cm CN CN N (N C r)










C C C C r.
S (c co (Cd0 0 '-
w U) En En r-i u E
(d

E E E e !3 T
0 0 0 0 Cd C
0 0 0 0 (N (N








I 0 0 0 0





S\D(N 0 (N 0







-1 mi IV (N m
I a *
0ILA 0 LA 0 0





I-3 rO
I O~~d~ c
I OX 03 3 I




Incn0

I ~ OV 0
I-l ~D3


d O 5

(N \ 0











- 5 0
i- I

*



00
o





00 0
/ O




*2






0
o"








O r

/ / 0 0
/u 4L .NO I
":H
4)








0 6 0
W/O 'N3 NO VMm
iios oiui
y9 >^ u >
x~~ mr r







17 October such that the applied plus extractable Ca totalled

400 kg/ha. Following incorporation of the limestone, rye

was planted to serve as a cover crop for the area over the

winter.

Soil samples were again taken from each plot on 20 Feb.

1980 in the same manner as before, and the soil was analyzed

for the extractable plant nutrients P, K and Mg. Following

disc harrowing of the rye cover crop, corn.hybrid Pioneer

Brand 3369A was planted on 17 March. The seed was over-

planted and the plants were thinned to the desired popula-

tion of 86,000/ha shortly after emergence. Row spacing was

alternate 30.5 cm and 61 cm widths with 25.4 cm between

plants within rows, which were planted in a north-south

direction.

The experiment was set up in a randomized complete

block design, with blocks consisting of the two different

crops following the corn (rye and soybeans in 1980). There

were eleven fertilizer treatments and four replications of

each treatment-block combination for a total of 88 plots.

Plot size was 7.3 by 6.1 m, and each contained 16 rows of

corn plants. A plot diagram is shown in Fig. 2.

The fertilizer treatments are listed in Table 8.

Nitrogen was applied as NH4NO3 solution banded next to the

rows in four applications, each 25% of the total amount

applied, at 3, 6, 8, and 10.5 weeks after planting. The

dates of application were 7 April, 28 April, 13 May, and

28 May, and the N was irrigated into the soil each time it









GENERAL PLOT DIAGRAM


T
I I


62mg


FT


W


]
I
]
]
]
I
1


4-

4-

4-
4-

4-
4-


First crop
Corn
Corn
Corn
Corn
Corn
Corn
Corn
Corn


Second crop
Rye, Sorghum
Soybeans
Rye, Sorghum
Soybeans
Rye, Sorghum
Soybeans
Rye, Sorghum
Soybeans


INDIVIDUAL PLOT LAYOUT


S00 Harvested
SArea





213141516 Row number

213 1415 16 *- Row number


< 7.3m


Figure 2. General plot diagram and individual plot layout.


6.1






1


I I I I I I I I 1
I IIIII I I I
I I I I I I I [ I
I I I I I I I I I
111111 1 I I


111
I 1 11111111 1 I





62



Table 8. List of fertilizer treatments.


Treatment
number N P K Mg

-----------------kg/ha------------------

1 336 0 0 67

2 336 0 418 67

3 336 29 418 67

4 336 59 418 67

5 336 117 418 67

6 336 59 0 67

7 336 59 209 67

8 168 59 418 67

9 504 59 418 67

10 336 59 418 0

11 336 59 418 134







was applied. Phosphorus and K were applied preplant broad-

cast as triple superphosphate and KC1, respectively, on 26

February in accordance with the treatments. Magnesium was

applied broadcast as MgSO4 on 1 April. Other nutrients

applied at a uniform rate to all plots included B, S, and

Zn. Boron was applied with the NH4NO3 solution at a total

rate of 1 kg/ha, while S and Zn were applied as ZnSO4 at a

rate of 4.5 kg S/ha and 9.0 kg Zn/ha on 1 April.

The practice of plow-layer soil-water management was

employed throughout the growing season. Irrigation water

was applied from the center pivot system to recharge the

plow layer (top 15 cm) to field capacity when the soil-water

tension measured by tensiometers placed at 15-cm depth

reached 0.02 MPa. Approximately 2.5 cm of water was applied

per irrigation. Rainfall data were collected from National

Oceanic and Atmospheric Administration instruments located

on the grounds of the Quincy Experiment Station and are

shown in Fig. 3.

A comprehensive effort was made to minimize the pos-

sibility that corn grain yields would be reduced due to

disease, weed, or insect problems. The crop was scouted

frequently throughout the growing season for injury and

appropriate pesticides or cultivation were used to combat

any pests that appeared. Chemicals applied and the purpose

for their use are shown in Table 9. Those materials that

were used on the second crops in 1980 and 1981 (soybeans

and grain sorghum) are also included in this table. All











N
*


1981 .L.. ,..L ,. L L
1981 I I


N N


N


N


A A -. & -- a a M ~* I.


25
DAYS


50
AFTER


75 100
PLANTING


Figure 3.


Rainfall amounts and patterns at the AREC,
Quincy during the growing seasons of 1980 and
1981. Nitrogen applications to corn are
indicated by the letter "N".


3ha


11980


71


E


_J
11.

z


0n
0-


I I


I l I


125














m

CD C; *
. .. O O .- O O t -


4J





S-I
4J

C

OC

a)
U)


O0
02
a)











0
r-
02
07







U
a:








0








0
o











0a








m


4 4
c~
0ie


0n


a 0) 0

U a) a)
c U

a)
54 0) (
a) (4 54


O E
L P4 P4


0n
0a





a)
t,-i aC)
a O 4. .
0 4-J O U U U
a ) m a a a


En ElC

a) a,

0 m 0 0 0 0 0




O O O O O O O O O O O O O
U 3 r U z3 E 0
0 0 0 0 0 0 0 0 0 0 0 0 0
u u u u u m) U) U) u u U) LO U


-4
>1 -
S Q 0 z


, Ol 0
o a) -4 -0


m- -4 4 0
C C 0
~ z c (


0
.a

O -



N qC 4 3
C a) CN -. N -

-3 i ( o Ic a H aE
*d N H l ra >C
4 -) N i a -i -
0 -i ^ Cd Cd U -4 a)




v O 0-l 4 O ) U r
ni U 4-4 aj 4 -4d -H
4 c d 4 -) > E c
Cd U '4 a) Cd -H -H <
m ) : OS 0a) -H c0
pl M M 4 4 0 U) 0


U)
0
a)







a) a)
0

0 rt

- .4








chemicals were applied at their recommended rates for the

soil properties of the Ruston series.

Above-ground corn plant samples were taken from each

plot at 38, 49, 79, and 108 days after planting. Each

entire plant was dried, weighed, ground, and analyzed for

N, P, K, Mg, Ca, and Zn concentrations. Total nutrient

uptake was then calculated by multiplying the individual

nutrient concentrations by the dry weight of the plant

samples.

Corn was harvested from each plot on 22 July. The

harvested area included 2.2 m sections of row numbers 3

through 6 and 11 through 14 from each of the 16-row plots.

The harvested segments of each row were taken near the

center of the row to eliminate border effects. Counts of

stalks and ears within each harvested area were also made.

Corn grain yield was calculated as follows: First, the

entire sample of ears from each harvested area was weighed.

Then 10 "average" sized ears were selected from the sample

and were weighed together. The 10-ear subsample was then

shelled, the grain was weighed, and the grain moisture con-

tent was determined. The weight of the grain from each

entire harvested area was then calculated by multiplying the

subsample grain weight by a factor which depended on the

number of ears which came from each particular harvested

area. The final weights were then extrapolated to hectare

area yields. All final yields were corrected to a grain

moisture content of 15.5%.







The remainder of the corn in the field was harvested

with a mechanical corn picker. The stalks were chopped with

a rotary mower and were left on the field. Paraquat was

applied to kill any remaining vegetation and the corn stub-

ble and remaining plant material were not turned under the

soil in accordance with no-tillage practices. On 30 July,

soybeans ('Cobb' variety) were planted in their designated

blocks in 76-cm rows, which were planted in a north-south

direction. Plants were 5 cm apart within rows, giving a

plant population of 263,000/ha. Seeds were inoculated with

Rhizobium spp. at planting. On 10 October, rye ('Wrens

Abruzzi' variety) was planted broadcast in its designated

blocks at a rate of 133 kg seed/ha.

Neither the soybean nor the rye plots had any fertili-

zer applied to them; all soil fertility was due to residual

nutrients from the fertilizer applied to the corn plus the

native fertility. Irrigation scheduling was not as stringent

for the second crops as it was for the corn. Irrigation

water was applied in combination with rain water such that

the total application amounted to about 2.5 cm per week.

The soybeans were harvested on 2 December. The har-

vested area included 4.6 m sections of the four middle rows

of each plot. The row segments harvested were taken near

the center of the row to eliminate border effects. Soybean

yield was calculated as follows: Each harvested plant was

put through a machine which separated the pods from the rest

of the plant. All of the mature seeds from the pods were








colle ted, weighed, and measured for their moisture content.

The w ights were then extrapolated to hectare area yields.

All f nal yields were corrected to a grain moisture content

of 12

e rye was left as a winter cover crop until 20 Feb.

1981, when it was turned under the soil with a disc harrow.

On 24 February, soil samples were taken from each plot in a

simile r manner as in 1980 and were analyzed for P, K, Mg,

and C Corn was planted on 10 March. The variety, plant

popul; ion, and row spacing were the same as in 1980.

1 artilizer treatments were identical to those of the

previc -s year (Table 8). Nitrogen was applied 3, 5, 7, and

9 weel 3 after planting. The dates of application were 31

March, 13 April, 29 April, and 12 May. Phosphorus and K

were c :plied preplant broadcast on 4 March. Boron was again

appliE I along with the N. Zinc and S were applied on 31

March, and Mg was applied on 2 April. Soil-water management

and pE ;t management schemes were the same as those used in

1980.

I )ove-ground plant samples were taken on 14 April, 22

April, 21 May, and 22 June. Corn was harvested on 12 July.

The he -vest and yield calculation procedures were identical

to the ;e used the previous year. Following the harvest, the

corn E -ubble was chopped with a rotary mower and left on the

soil s irface.

I1 1981, grain sorghum replaced rye as a second crop.

Sorghu i and soybeans were planted in their designated blocks








in east-west rows on 29 July. The sorghum was planted in

76-cm rows at a rate of 16.8 kg seed/ha. The soybean row

spacing was narrowed to 25 cm, and the plant population was

increased slightly over that of 1980. Seeds were inoculated

with Rhizobium spp. at planting.

Nitrogen as NH4NO3 was applied to the grain sorghum at

a total rate of 118 kg/ha in a split application. One third

of the N was applied on 17 August and the remaining two

thirds was applied on 9 September. No other fertilizers

were applied to either the sorghum or the soybeans. Irri-

gation was again used such that applied water plus rain

water equalled approximately 2.5 cm per week.

Plant samples for analysis of N, P, K, Mg, Ca, Mg, Mn,

and Zn concentrations were taken on 21 October for the soy-

beans and 4 November for the grain sorghum. Soybeans were

harvested on 18 November using the same harvest and yield

procedures as used in 1980. Grain sorghum was harvested in
2
late November using the following harvest procedure: A 4 m

(0.0004 ha) area was marked off in the center of each plot

and the grain sorghum was harvested from within this area.

The grain was collected, weighed, and measured for moisture

content. The weights were then extrapolated to hectare area

yields after correcting them to a grain moisture content of

12%.








Laboratory Procedures


The soil samples taken from each plot were dried in a

chamber under heat supplied by incandescent light bulbs,

then were run through a grinder to break up the aggregates

and mix the soil. Five grams of soil from each plot were

extracted with 25 ml of double-acid (0.05N HC1 in 0.025N

H2SO4) extracting solution. The extraction procedure con-

sisted of continuous shaking of the soil-solution mixture

for 30 min. followed by filtering. The soil extract was

analyzed for P by the colorimetric molybdenum blue method

(Olsen and Dean, 1965), and K, Mg, and Ca by flame emission

(Rich, 1965). Flame emission analyses were performed using

a Beckman model DU emission spectrophotometer. Soil pH was

measured using 1:1 soil-water pastes and a glass electrode

pH meter (Peech, 1965).

In order to characterize the soil somewhat further,

CEC, organic C content, and soil texture were measured using

soil samples taken at the beginning of this experiment.

Soil samples from six depths (15, 30, 45, 60, 90, and 120 cm)

were obtained from 12 widely separated areas in the field.

Cation exchange capacity was determined using the calcium

saturation method (Jackson, 1958) and soil texture by the

hydrometer method (Day, 1965). Organic C was determined for

the top three depth increments only by the Walkley-Black

method (Jackson, 1958). The results of these procedures are

summarized in Table 7.








The whole above-ground plant samples taken from each

plot were dried in the same manner as the soil samples,

weighed,and ground up into a fine consistency. One gram of

plant material from each plot was placed in a beaker and put

into a muffle furnace at 5500C until it was completely

ashed. The residue was then dissolved in HC1 according to

the procedure described by Chapman and Pratt (1961). The

resulting solution was analyzed for total P, K, Mg, Ca, and

Zn in 1980. The following year, the solution was also

analyzed for Mn. The elements P, K, Mg, and Ca were measured

using the same procedures referenced previously. The ele-

ments Zn and Mn were measured by atomic absorption (Isaac

and Kerber, 1971) using a Perkin Elmer model 2380 atomic

absorption spectrophotometer. In a separate procedure, 0.2 g

of plant material from each plot was used in a micro-

Kjeldahl procedure (Chapman and Pratt, 1961) to measure

total N in the tissue. The amount of each nutrient measured

was converted to a percentage form, and nutrient uptake by

the plant was then calculated by multiplying the percent

nutrient in the plant by the total dry weight of the plant.

This calculation was made for plants from all four sampling

dates each year.

The procedure for each laboratory analysis referenced

in this section can be found in Appendix 1.








Statistical Analyses and Handling of Data


The experiment was a randomized complete block design.

The factors N applied, P applied, K applied, and Mg applied

were converted to a single factor labelled as treatment,

with 11 levels. Analysis of variance was performed with

respect to the treatments and blocks on corn grain yield,

corn dry-matter yield, number of stalks/ha, number of ears/ha,

ear size, percent barren stalks/ha, percent lodged plants/ha,

soybean yield, and grain sorghum yield. The procedure ANOVA

of the Statistical Analysis System (Goodnight, 1982a) was

employed in performing the analyses. In those cases where

it was shown that all treatment means of a dependent vari-

able were not equal (using the 5% level of significance),

Duncan's Multiple Range Test (Steel and Torrie, 1960) was

used to determine the means that were different from one

another.

After the effects of the overall treatments were inves-

tigated, response of dependent variables such as corn dry-

matter yield and corn grain yield to levels of independent

variables such as nutrient uptake or level of a nutrient in

the soil was examined. When investigating the effect that a

single nutrient had on any of the dependent variables, it

was necessary to attempt to eliminate any effects of other

nutrients. In other words, the nutrient in question needed

to be the only factor which caused the dependent variable to

respond. Therefore, when the response of a dependent







variable (such as yield) to an individual nutrient was

investigated, data used and presented in the investigation

were obtained from that set of treatments in which only the

nutrient in question varied in application rate while the

remaining nutrients were applied at constant levels. It was

assumed that non-yield-limiting amounts of the remaining

nutrients were available for uptake from the relatively high

constant levels that were applied, and that any limited

yields were due to less than optimum uptake of the nutrient

being varied. The treatment combinations used for each

nutrient were treatments 4, 8, and 9 for N, treatments 2, 3,

4, and 5 for P, treatments 4, 6, and 7 for K, and treatments

4, 10, and 11 for Mg (see Table 8).

A multiple regression model was used to test for the

importance of the uptake of the nutrients N, P, K, Mg, Ca,

Zn, and Mn in influencing corn grain yield. In the model,

uptake amounts of the individual nutrients were taken as

independent variables and grain yield was taken as the

dependent variable. A stepwise regression technique was

used. The procedure began by finding the one-variable

model that produced the highest R2, while at the same time

satisfying the condition that all introduced variables have

a significance level of 15% for entry into the model. Then

other variables were added to the model until there were

none left which met the entrance requirement. Thus, only

variables which were important in influencing grain yield

were included in the models. Coefficients of single and








multiple correlation were calculated by taking the square

root of the coefficient of determination. This procedure

was performed separately on data from each of the four

sampling dates within each year. The STEPWISE procedure of

the Statistical Analysis System (Goodnight, 1982b) was used

in performing the regressions.

When fitting equations to fertilizer nutrient and soil-

test nutrient response data where corn grain yield was the

dependent variable, the Mitscherlich equation was used (Mel-

sted and Peck, 1977). The form of this equation that was

fitted was

Y = Yo(1-exp(-cX)),

where Y is yield, Yo is maximum yield when all nutrients are

at optimum but not excessive levels, X is the level of the

nutrient or growth factor being considered, and c is a

constant related to the type of nutrient or growth factor.

It was assumed that the corn hybrid used in this experiment

had a finite grain yield potential, Y which was affected

by the level of nutrient X in the proportion (l-exp(-cX))

provided that all other nutrients were non-limiting to grain

yield. The exponential part of the Mitscherlich equation

indicates that if extremely low values of the limiting

nutrient are present, grain yield will increase as more of

this element is added, but not in direct proportion to the

amount added. The increase in yield with each successive

addition of the element is progressively smaller. The con-

stant c determines how quickly or slowly grain yield will







reach its plateau as increments of nutrient are added. The

NLIN (non-linear) procedure of the Statistical Analysis

System (Goodnight and Sail, 1982) was used to determine the

parameters Yo and c in the Mitscherlich equation for grain

yield vs. nutrient uptake and grain yield vs. soil nutrient

level relationships.

A commonly used measure of the goodness of fit for
2
linear regression models is r the amount of the total pro-

portion of variability explained by the regression model.

It is defined as the ratio of the regression sum of squares

to the corrected total sum of squares. However, for non-

linear models this ratio may exceed 1.0 since non-linear

models usually do not contain a parameter for the population

mean. In an attempt to report a measure of the goodness of

fit for the Mitscherlich models, r2 will be defined as 1.0

minus the ratio of the residual sum of squares to the cor-

rected total sum of squares (Ware et al., 1982). Although

this ratio underestimates the degree of fit for non-linear

models containing asymptotic parameters such as Y in the

Mitscherlich model, it can be used as a basis for comparing

the degree of fit for different non-linear models (Ware et

al., 1982).

One of the main objectives of this research was to

establish critical levels of soil-test nutrients or soil-

test plus fertilizer nutrients required for the production

of high-yield irrigated corn. This was done using the Cate-

Nelson procedure for partitioning soil-test correlation data








into two classes (Cate and Nelson, 1971; Nelson and Ander-

son, 1977). In this procedure, soil-test correlation data

are divided into two categories: one which indicates that

there is a high probability of a response to fertilizer and

one which indicates a low probability of a response. Origi-

nally, this technique was graphical, where vertical and

horizontal lines were superimposed on a plot of yield vs.

soil-test nutrient so as to maximize the number of points in

the first and third quadrants. The point where the vertical

line intersected the abscissa was used to divide the data

into two classes, and was termed the critical level. An

analysis of variance method has since been developed to

determine the critical level mathematically. The steps

involved in this procedure are found in Appendix 2.

All analyses using the Statistical Analysis System were

done utilizing the computing facilities of the Northeast

Regional Data Center of the State University System of

Florida, located on the campus of the University of Florida

in Gainesville.













CHAPTER 3
RESULTS AND DISCUSSION

Yield Response to Treatments


The response of corn dry-matter yield, corn grain

yield, soybean yield, and grain sorghum yield to fertilizer

treatments is shown for 1980 and 1981 in Tables 10 and 11,

respectively. Analysis of variance indicated that the

second crops in the rotations (rye and soybeans in 1980 and

grain sorghum and soybeans in 1981) did not influence corn

dry-matter or grain yields; hence, the means for these

variables presented in the tables were calculated from

combined data from both rotations.

Corn grain yield in 1980 ranged from about 7 Mg/ha (110

bu/acre) where neither P nor K was applied (treatment 1) to

about 8.6 Mg/ha (137 bu/acre) where either P or K was not

applied (treatments 2 and 6). Where all nutrients (N, P, K,

and Mg), regardless of quantity, were applied, yields were

about 11 Mg/ha (175 bu/acre). Total dry-matter yield of the

above-ground portion of the corn plants ranged from about

15 Mg/ha at low levels of applied fertilizers to about 20

Mg/ha at higher levels in 1980.

Total corn dry-matter yield and grain yield showed no

significant response to applied N above the 168 kg N/ha rate











r-


















r)



a)
4-)











4-)

C
O
















rQ
C









0
U


m C>

N0
o r-

N cO

-c CO
1-1 1-


00




qr co o
( LA N















00 qr' CM
( rN rn
*H 'H
'H 0
1-1 I-


mn
0r LA

a; cy
'- 1-1


0
.Q CU


m q














0 (0

n (
oo












0 o
,-I 1-1









.C)




oc 0

nH N
C-- Cri

















co
-H CH
m0 C


N N N N N CN N N


O 00


00 00 0
H- MH
w. w' P*


00 O\ O\
AJ rLA'
'H


C', ~D C',
C', Cm C', C'


,Q
0

Crd


Ln
*H

'H 'H


4 4

CN e
tZ LA







o '
(0 ()











*


o Or 00 0o 0c0o





inL m Lon mL



OoD m o o o
Mm o m
mm H m


-. m V' t A ,o N co m om 0 r-i
r-1 r-


m cr
0) 0
o o
0 o l
1-1i 1-I


.0

4-4
(Cd




a,)

4-4
.1-1
ro






oH







CU
0
-H

4-4











Hr-
4-
r-


U)
C
4n








4-d


















H
r-4












0
-r-4


















-H




0 +o










do
da)





























m



4,1
EU












rt (U
(U







(treatments 8, 4, and 9 in Table 10). These variables also

showed no response to applied Mg above the zero rate (treat-

ments 10, 4, and 11). There was a significant response of

dry-matter and grain yield to applied P above the zero

rate (treatments 2, 3, 4, and 5), but no significant dif-

ference between yields at each of the three levels of

applied P even though there was an upward trend in yield

level as the rate of applied P increased. There was no

significant response of corn dry-matter yield to applied K

above the zero rate (treatments 6, 7, and 4), but there was

a significant increase in corn grain yield between applied K

rates of 0 and 209 kg K/ha. There was no additional sig-

nificant response of grain yield when applied K was increased

to 418 kg/ha.

Soybean yield in 1980 ranged from about 0.77 Mg/ha

(11 bu/acre) where neither P nor K was applied to about

1.29 Mg/ha (19 bu/acre) where all nutrients, regardless of

quantity, had been applied. Yield responded to residual P

left from the fertilizer applied to the corn, but not to

residual K or Mg. Yield was significantly higher where P

had been applied at a rate of 117 kg/ha than where no P

was applied.

Corn grain yield in 1981 ranged from about 7 Mg/ha

(110 bu/acre) where neither P nor K was applied to about

10.9 Mg/ha (174 bu/acre) where all nutrients were applied,

regardless of quantity (Table 11). Unlike 1980, grain

yields in 1981 in the absence of applied P or K were not












*4






54




N

-H

4-







O
4,




4-4





co








O0
U)
















*r
1-1




r-o
4-)













0











U0
-41
,4.















Cr
Crl
00


















'-1-







*-r4
rd
&'
TI















C a
r- 0


0 -





a
-




a>1
*d











ro


-H




r)
>u
0






U-l
0
C(











u





-P



>1
c4


Sd





U
4u


Q rCd

Y-I s
LA l
'-0 '.

C) '9


00 k.0 *q
Co- Ln re)
m- wt -

rlrlr


0 00
M r-


.Io 00
v-Iu


LA) 00 t.
C 00 om
m

nOO
* *


rd od



* C)

r- v-


Cd rd

(N 00
LA


0 -4









it M




* *
r-i i


r- r- r. r0 r. r-0 r- r-' '0O


C0 C c00 00 0 0




o o) 0' 0r N 01
CN LA r-I Ln
r-1I


k. '.0 '0 %D 't 0
mmmmmm


o01 co 0o co0 0
0 r-I r-i r- V-4
CN -4 *. -qT IT


01 01 01 o0 o0




m LA LA LM L


r-I N m -v' L k.0 r Co o0 r- i
~N1d1 r-iO


m r N m oo co com r- N














N l q3 i-I Or m (N LA (Y) LAn 00
m N- 'k.0 00 m LA 00 V m C>
c -d L Ln .n t :1 C) LA CNT

O r-l -1 i- H 1 r-- i-1 r4


a),


cq
cdC
mE


CO



i-4
c) C
4-)4






Z 4
r-1 .- O

M O
0 Q)


r- aQ) i



S*H -C
C 0





0 a)




0 -4-) -H-
uarm

E-< >1
4-) .,
4) 4-
U tn C
r-C
(0 U

E 4-4

r-A
4 o --
a) -H -H
4J U)

U!
-4-)4H






0 m a

0 r)



CO 4.) (O

M C4

0 r 0
U 4 r.
0



4-) 4-)
C r-I




r-) *n
> 0


SdP -r-

-Kt


Q Cd

m (N
N N
CM N
N C)
rr
i-i
O;
r-







similar. The grain yield where no P was applied (7.3 Mg/ha

or 116 bu/acre) was significantly lower than the yield

where no K was applied (10.3 Mg/ha or 164 bu/acre). Corn

dry-matter yield in 1981 was lower than in 1980, ranging

from about 10.0 Mg/ha at low levels of applied fertilizers

to about 18.3 Mg/ha at higher levels.

As in 1980, corn dry-matter and grain yield showed no

significant response to applied N above the lowest rate of

168 kg N/ha and no significant response to Mg application.

There was a significant yield increase for both dry matter

and grain as the amount of P applied increased from 0 to 29

kg/ha of P, but no further significant yield response was

seen as more P was applied even though there was an upward

trend in yield level. Unlike 1980, there was no significant

dry-matter or grain yield response to applied K.

Soybean yield in 1981 ranged from about 0.84 Mg/ha

(12.5 bu/acre) where neither P nor K was applied to corn

to about 1.69 Mg/ha (25 bu/acre) where all nutrients were

applied, regardless of quantity. Grain sorghum yield ranged

from about 1.36 Mg/ha (21.5 bu/acre) where no P was applied

to about 2.81 Mg/ha (44 bu/acre) where P was added. Neither

soybean nor grain sorghum yield significantly responded to

any residual nutrients from fertilizer applied to corn.








Response of Soybean and Grain Sorghum
Nutrient Concentrations to Treatments


Table 12 shows the response of nutrient concentration

in the 1981 soybean and grain sorghum plants to the ferti-

lizer treatments. There was no significant difference

between mean K or Mg concentrations for either the soybeans

or grain sorghum, but there was a difference in P concen-

tration. The P concentration in both the soybeans and grain

sorghum was significantly higher where 117 kg/ha of P was

applied to the corn than where either 0 or 29 kg/ha of P had

been applied. However, according to data in Table 11 this

increase had no effect on grain yield. The fact that there

was a concentration response to residual P but no response to

residual K indicated that applied P remained available in the

crop root zone longer than applied K, which probably leached.

It is well established that K is a more mobile nutrient than

P in Florida mineral soils.


Response of Other Corn Field Parameters to Treatments


The treatment means of additional data collected from

each corn plot in 1980 and 1981 are presented in Tables 13

and 14, respectively. Variables measured included number of

stalks and number of ears per hectare, ear size, percent

barren stalks, and percent lodged plants. Analysis of vari-

ance indicated that the second crops in each rotation did

not influence any of the variables listed, thus the means












NL 0 L

coo
CN CN CN
*; C; *


.Q


a








0






C'
4-















od









0
C
rn
O





























l 0
0


x-)


a. )
ro -H


04



r-
co

























04-)
C N
0a)

U 4



C4



-H4
0)





Cr-l
(0 r
(^r















Q) i-


0








0
O







cO


0
.H
a






c
0
4-)
(0









-H
4-)

C
0)


U





0


















-H
C






a)
0


0








-H

a







r

4-
-1


0 0

* O *
0 0






.n




No











0
0 C
m *
r-l



U






0 0



. Q
.Q Cd
Co o
CO M
r- N

0


D \D' .D0 0 O'.O' D .0 '0W


0 00
1-1l


CO CO CO
r--l r-
v d


0 0 Ov 01 N
CM n r-L i
1-


0 mv CO0
0 r
N l'q


CO 00 00
v- I qr
<<. .


LA LA LA LA LA LA


m'D Y) CO

CA CA


*z %D t0
0 'n m
LA n CA


N-i m L 'o co 0 o -0 -I
~Nm~v-I v-I


L I)
iN o
c'3 o,






C* 0
LO





0 0
00



















N N
fo (d
0 O

,- 0






0 0
o o
N N

0 0



-.Q



N N

00


LO I) o
n CN 0
N No

000






000
LA LAn o


000





0 0 0
0 00
0 r C








ON N
CN CN CN

r- -l r-I



U 0







SLA LU

Sco






000


) I
C



aod

3r0




4.4-
4 r-4
CO



con




r -i ()4J
au





OcOH
C- c
0) -0




4-4
* -H 0
o-r o E
z 0
m a)p
-H 4-4

S4-)


-H C
0 Q a)

a)
a 4-
1 -I 4-


a m




4) 4-
4-) a) 4-1
a) z

0rd U


d a) -H


a) -H -H

4-) r-I

aU))




00
4- (z



0 4-)
U4-) C4


c a)



0 Qr 0

0



4-) 4-)



() a)4
r-l cr
it U4





*L) ra



-K


0L

I
U


r^








presented in the tables were calculated from combined data

from both rotations.

The desired corn plant population at planting was

86,000/ha. In 1980, the average plant population of only

three of 11 treatments exceeded this value. However, the

greatest stalk reduction for any treatment was less than

3,000 plants/ha, or less than 3.5% below the desired popu-

lation. In 1981, eight of 11 treatment means exceeded the

desired level and the lowest value for any treatment was

within about 500 plants/ha of it (Table 14). Neither the

number of stalks nor the number of ears was affected by the

nutrient treatments in each year. The percent barren stalks

in both 1980 and 1981 was not significantly affected by the

treatments, although the highest values for this variable

tended to be associated with treatments which supplied low

levels of P, K, and Mg both years. The treatment where all

nutrients were applied except K (treatment 6) had the

highest percentage of lodged plants in 1981. This was

expected since K is a major factor in determining the struc-

tural stability of corn plants (Barber and Mederski, 1966).

Lodging reduces the harvestable yield where mechanical

pickers or combines are used. Treatments with the lower

percentages of lodged plants were those where low amounts of

N and P were applied. These treatments tended to result in

smaller, lighter ears.

Ear size in 1980 ranged from about 80 g of grain/ear

where neither P nor K was applied to about 128 g/ear where

















-I CO O 0 0 0 0 o C C44 0
-4 r 00 0000 0 00 0


C)
4-
Cd
-10
a,

4-J



0







0r)
Cd









Cd



0





00


Q-
(0









r. *




-O





4 -
00


4
C I




am
C m















0-t
.El
4J4




IC
E-1
02




C)





a-Q
CO(d


0 oc
r-4
I;T


(N 10 c -
CM4 Ln r-
-4


^0 0O ^0 D
mmm rm
rQ r ro) ro r


0 O 00 0
0 r-1 r--
o M m ,,
OrlH
N < < S


LO Ln Ln Lt Ln


t0 .0 00
m C O
re Cf 1-1


0 0 '.0
l0 o C
Ln (


mN- (1 .o N mC m 0 -
r- i-


r- r- r- [- r- r- r- c- r- 0 It3
-D s 1%0D 'D 'D tD D %D tD re


, I






CO
4-) r-I




r 0







CdOC)
H C


- 0 -
rO .Q









2 0
4-)i0 i
















m 0
zH 0











O 4
4-i r-






c)4
4 -4-4










Cd 0
)- 02z
mC













4 4




0 2
dC Cd
CO -) -H
4-1 4- r-i













> 1
r0 fa




















0 0
4 -4






0 4-i
H -4 M
0 o' -0





rc-
0l)
,c U4
*il U (
43 d

*rc1l-

> Q):








(0 U
(UcfP *
S n )
E bK











rro
N u


L- tO


0 .0 0 0 0
.Q C 0 Q0 r Q U

r- m c r0 .o

N r4 a t L LA 'l L


(0






4
4-)
co



0)










N
4-)
0
r:
CO

















N
S*























m
^1 4-4
tC

























0u 0



E ro
eO

n-4 4
Cd



N
0 -r
00-
4 -H




e a





















0 0
4-)











En o







f
r-4




3 )


13 03

() 4-1
0C
o0 (0










OH
C4










m

4





14










U)



4-4








rCd








-H a
-4
-l













4-4
a,



4,-
I..
O)z


-4 -
.c~jv ~c.


0 0 0v O N
(N LA. r*i
r--4


0 m
0
(N


om
Ln ql


N LO







o O"
m
1-


--0-4 -r
v .T .*d 4 ~


m0 0m 0m1 m v 0m m
LA Ln LA LA L LA


to kiD t.o tZ k.0 %D tD wO T t.0 11D
m mm mmm m om m
m mm mm mm n m m






(- N m LA k N co w 0 r-i0
1-4 -i


(N CNO v-i 0 (Nr r-4 I 0


0) 0


Cd
UP)C




0 E



O
4J 4-) 0
r, 4-)
C O



4-4 o
-H0 O
0 0

-0 0J 4-


Ur-I 4 C4-4
4


-4 .4-4
t I4-) -H-
rC



4- 0) 4-

U C
kQj
4J-l -1






0 I
U) r -H)
0 4u
d0 U -Hl








U-4 *E






0 4
S-tP H-
r-p -1H


O C


0C








4-),









> (
3 i 1o





r-1 -H






0 o 0
0





0 U)



*LaX


tD %o %.c o tZ t D D t


I


d4
U)U
H,


0 0C
1-4








all nutrients were applied, regardless of quantity. Ear

size where either P was not applied or K was not applied

was similar and averaged 98 g/ear. Ear size showed a

significant response to the first level of applied P and K

above the zero level, but did not respond to any increased

amounts of either element. There was no significant

increase in ear size due to increased application of N above

the 168 kg/ha level or to Mg above the zero level.

Ear size in 1981 ranged from about 83 g/ear where P was

not applied to about 130 g/ear where all elements were

applied, regardless of quantity. Unlike 1980, ear sizes

where either P or K was withheld were not similar. Ear size

significantly responded to the first level of applied P

above the zero level, but did not significantly respond to

any addition of K. However, the average ear size where K

was applied was considerably larger than the average ear

size at the zero rate of K. As in 1980, there was no sig-

nificant increase in ear size due to application of N or Mg

above their lowest rates.


Cdrn Dry-Matter Yield vs. Nutrient Uptake


The relationship between corn plant dry-matter yield

and nutrient (N, P, K, and Mg) uptake is shown in Figs. 4-7.

In each plot, the points appear in four groupings of either

three or four. These groupings correspond to the four dif-

ferent times during the growing season that plant samples

were taken, and each point within a group represents the