• TABLE OF CONTENTS
HIDE
 Title Page
 Introduction
 Materials and methods
 Economic evaluation
 Conclusion
 Appendix I
 Appendix II
 Appendix III






Group Title: Radish response to nitrogen applications : Team III, Dain, Bridgeland, Proenca, Sesto
Title: Radish response to nitrogen applications
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00075665/00001
 Material Information
Title: Radish response to nitrogen applications Team III, Dain, Bridgeland, Proenca, Sesto
Physical Description: 19 leaves : ; 22 cm.
Language: English
Creator: Hildebrand, Peter E
Publication Date: 1988
 Subjects
Subject: Radishes -- Fertilizers   ( lcsh )
Genre: non-fiction   ( marcgt )
 Notes
Statement of Responsibility: P.E. Hildebrand.
General Note: Typescript.
General Note: "April 18, 1988."
 Record Information
Bibliographic ID: UF00075665
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 82967234

Table of Contents
    Title Page
        Page 1
    Introduction
        Page 2
    Materials and methods
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
    Economic evaluation
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
    Conclusion
        Page 20
        Page 21
    Appendix I
        Page 22
        Page 23
        Page 24
        Page 25
    Appendix II
        Page 26
        Page 27
        Page 28
        Page 29
    Appendix III
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
Full Text
















RADISH RESPONSE TO NITROGEN APPLICATIONS


Team III
Dain
Bridgeland
Proenca
Sesto















Dr. Hildebrand

April 18, 1988


























RADISH RESPONSE TO NITROGEN APPLICATIONS


I. INTRODUCTION



The following experiment was conducted to determine the

effects of various levels of nitrogen application on radish bulb

production. The investigation was carried out in conjunction

with Dr. Peter E. Hildebrand's course, AGG 5813, Farming Systems

Research and Extension Methods, during Spring Semester 1988. The

class was divided into six groups, each responsible for two

separate trials of two blocks each. The students were expected

to learn practical applications of statistical analysis tech-

niques, and some basics of experimental design commonly used in

on-farm agronomic trials. On-farm trials constitute the basic

tool that farming systems workers use to predict the success of a

proposed technology from within the constraints of the small-

scale farm.



II. MATERIALS AND METHODS

Site Evaluation

The pre-selected trial site is located in the University of

Florida student research area, on the southwest side of campus.


























RADISH RESPONSE TO NITROGEN APPLICATIONS


I. INTRODUCTION



The following experiment was conducted to determine the

effects of various levels of nitrogen application on radish bulb

production. The investigation was carried out in conjunction

with Dr. Peter E. Hildebrand's course, AGG 5813, Farming Systems

Research and Extension Methods, during Spring Semester 1988. The

class was divided into six groups, each responsible for two

separate trials of two blocks each. The students were expected

to learn practical applications of statistical analysis tech-

niques, and some basics of experimental design commonly used in

on-farm agronomic trials. On-farm trials constitute the basic

tool that farming systems workers use to predict the success of a

proposed technology from within the constraints of the small-

scale farm.



II. MATERIALS AND METHODS

Site Evaluation

The pre-selected trial site is located in the University of

Florida student research area, on the southwest side of campus.














The site was divided into two "villages", with each student team

responsible for two plots or "farms", one in each village. The

land is gently sloping, and the farms and villages are situated

consecutively along a north-south axis. Following is a brief

characterization of the biophysical environment, including cli-

mate, soils, crop history, and weeds.

Climate. The area's climate is humid subtropical, with

brief dry periods corresponding to the early part of spring

(April and May), and fall (October and November). During winter

months rainfall is primarily associated with the convergence of

air masses of unlike temperatures. Summer rains, on the other

hand, are mainly of the convectional type. Mean annual air

temperature is 21 C, and average annual rainfall is 1346mm. The

growing season for the radish experiment coincided with the

period during which the incidence of frontal rains begins to

decrease. In fact, drought stress in Alachua County is most

common during early spring.

Soils. According to the Alachua County Soil Survey Report

(1985), the site's soil is classified as a loamy siliceous hyper-

thermic Arenic Paleudult. Paleudults are highly weathered, and

typically low in fertility. Agricultural value, therefore, is

somewhat limited. The soil moisture regime is udic, meaning that

the soil's control section is dry for not more than 90 days cumu-

lative each year. The soil temperature regime is hyperthermic,

with the mean average soil temperature (MAST) above 22 C, and

with a seasonal difference of less than 5 C.

Soil samples taken from both the north and south villages














showed some variability in soil type between the two. The

textural class varied from sandy clay loam for the north village,

to loamy sand for the south village. It was thus obvious from

the outset that some variability in water-holding capacity and in

cation exchange capacity would exist between the two villages.

Some of these differences were quantified by the results of the

soil analysis performed by the University of Florida Soil Testing

Laboratory; these will be discussed and analyzed in greater depth

later.

Field historYandweed_ opulation. The field used for the

experiment has been cropped almost continuously for approximately

twenty years. The south village was last in cowpeas and winter

wheat, and the north village had also been planted with a

leguminous crop. The weeds present in the plots were nutsedge,

bermudagrass, wild radish, crawfoot grass, docks, and the

leguminous hairy indigo. Nutsedges constituted the greatest

problem throughout the field. The portion of the field used for

the north farms had been recently treated with chemicals (of type

unknown), but the south farm area had been cropped without the

use of chemical inputs.

Plant materials

Early scarlet globe radishes (Raehanus sativus) were chosen

for field trials due to their short growing season (approximately

five weeks) and high germination rate (85%). The radish is a

member of the mustard family and can be an annual or biennial,

depending on when it is sown and how it is managed. Roots come

in a variety of colors ranging from white to a deep red, and a













variety of flavors ranging from mild to very hot. Radishes are

classified as a cool, hardy crop that flourishes under a wide

range of conditions. Optimum temperature is 50-60 F, and frost

is tolerated to a degree. Radishes will grow on almost any soil,

though sandy and sandy loam soils are preferred. For good root

development, deep, well-worked soils with a pH of 5.5-6.5 and

adequate P and K are required. It is recommended that seeds be

sown 0.5-0.75 inches deep, and at a density of 8-18 per foot.

Rows should be 8-12 inches apart. For class trials, recommen-

dations were slightly different at 8-10 seeds per foot, and 12

inches between rows. With regards to chemical inputs, radishes

generally do not need herbicide, but in sandy, Florida soils it

is recommended that fertilizers be applied at a rate of 200lbs N,

lOOlbs P205, and 100-1501bs K20 per acre. To achieve good root

quality, adequate irrigation is normally important.
Experimental des ign

The experimental design used in these trials was a random-

ized complete block (RCB), with two reps and five treatments, at

each of twelve farms divided into two groups (North and South).

Among the advantages of the RCB design are the ability to

estimate experimental error through ANOVA, ease of application of

Modified Stability Analysis, and simplicity of field layout.

Blocking was parallel to the field gradient on the south locati-

on.

Manaseqent _Practices

Management practices refer to the socio-cultural aspects of

the experiment. By this we mean the human element, the knowledge













and skills possessed by the farmer. Management practices are the

most difficult variables to control, but can have a profound

effect on all aspects of production. Complete information on

trial management practices was not available, so only documented

practices are discussed.

Planting density/deth. Seed was sown at a rate of 1/4

pound per farm to insure adequate seedling emergence. Seeds were

planted at a depth that ranged between .25 and .50 inches. After

thinning, plant density averaged between 10 and 12 plants per

foot.

Fertilizer aeplication. Five ammonium nitrate fertilizer

levels were tested on each farm with each farm receiving two

replications. Fertilizer rates were of 100 pounds/acre, 200

pounds/acre, 300 pounds/acre, 400pounds/acre, and a control

treatment which had no fertilizer applied. Fertilizer applica-

tions consisted of either banding at three inches on both sides

of the seed rows, or broadcasting evenly over the entire farm.

Fertilizer was incorporated into the soil.

Thinning. During the growing season, farmers thinned the

radishes to approximately 10-12 plants per foot. Some farms were

not thinned as well as others.

Tillage. Fields were tilled with both hoes and manual plows

to a depth of approximately four inches. When necessary, some

farmers attempted to reduce field slope by terracing.

rrLiqation. Irrigation was provided twice during the

growing season, both times in the final three weeks before

harvest. Exact irrigation levels are unknown.














Weedinq. Some farmers combined weeding and thinning

operations. Weeding was conducted in a somewhat haphazard

manner.

Methodolo _of Team III

Plantinqdensiit'/deth. Team III used a greater planting

density than any other farm team. Approximately 1/8 pound of

seed was borrowed from other teams and planted in addition to the

1/2 pound originally supplied. Planting depth ranged between .25

and .50 inches.

Fertilizer_a2elication. Fertilizer was incorporated in

bands no closer than two inches to the seed row on either side.

Care was taken to assure equal distribution of fertilizer.

Thinning. Due to extremely high germination, thinning was

necessary on both farms. On the south farm, rows were thinned to

approximately 12-15 plants per foot. On the north farm, thinning

left a slightly higher plant density that ranged from 15-18

plants per foot.

Tillage. Fields were tilled to a depth of four inches using

a manual plow. After tilling, a rake was used to slightly level

plot slope and break up large soil clumps.

Irrigation. Irrigation was the same as on other farms, but

our south farm suffered from drought stress.

Weeding. Weeding was carried out during thinning operations

and consisted mostly of removing grasses and sedges by hand.














III. ECONOMIC EVALUATION


An evaluation was done to determine the net income, per

acre, of farms using each of the five technologies. The logic

behind this analysis was as follows:



1 Radish Market Price = $450.00

2 Price of Ammonium Nitrate (33.5% N), per ton = $240 (contains

6701bs of N)= $0.36/lb of N.

3a. With fertilizer; Fixed Costs=Seed+Tillage + Fert. Ap.+Labor=

$100

b. Without fertilizer; Fixed Costs= Seed + Tillage + Labor = $50

4 Variable Costs = Fertilizer + Application

5 Net Income = Gross Income (Variable + Fixed Costs)

or,

Net Income =(Radish Wt. in Tons $450.00) -[(Fert Rate*

$0.36)+Fixed Costs))

Net Income was estimated for all yield values on the dataset, and

then regressed with the environmental index as the independent

variable; a regression was done for each N-treatment.



IV. RESULTS AND DISCUSSION

Response to N-treatments

All data is presented in Appendix 1. Treatments varied

considerably from the north to the south location. The overall

mean yields were 1.24 and 0.278 T/A, for the north and south

locations, respectively. The overall treatment means corre-














sponded to, in ascending order, 200, 400, 300, 100, and 0 Ibs

N/A, or, 0.988, 0,979, 0,903, 0,761, and 0.174 Tons of radishes/

A. Two MSA analysis were done to determine yield response to N

treatment, the first with 120 and the second with 115 observa

tions. The analysis with 115 observations was carried out as we

considered that the results of five observations were essentially

a result of management practices (i.e. washed away plots, lost

data, etc.), and not the result of different N rates. The

ultimate goal of this analysis was to compare it to the 120

observation MSA, in order to determine if a significant shift in

any treatment occurred. The eliminated observations were

S.1.1.100, S.1.2.400, S4.2.300, S.5.1.400 and S.5.2.400 (Loca-

tion. Farm #. Rep #. N-Trt).

Analysis of N Treatments With 120 Observations

The average of all treatments at one location is the

Environmental Index value (E); for the north farms E ranged

between 1.732 and 0.932, and for the south it ranged between

0.923 and 0.045. The fact that there is no overlap of ranges

indicates that the south farms were less productive than the

north ones for radish growth. Having calculated the E value for

each farm, we carried out the Modified Stability Analysis (MSA),

with the graph on the following page:








Effect of N on Raidish fieldd (1.23 o1~selv,)


N::1
N:2
H::3
N::,4


e'nvri~rolnrint~a1 i ndei:v


2: 3.


1, 8I


1,36;


-

,8611














The equations for the N-treatments with 120 observations are

as follows:

Olbs N/A Yield = 0.0619 + 0.147*E

1001bs N/A Yield = 0.1457 + 0.809*E

2001bs N/A Yield = 0.0536 + 1.228*E

3001bs N/A Yield = -0.2041 + 1.455*E

4001bs N/A Yield = -0.0572 + 1.361*E

At the outset we thought that there should be four recommen-

dation domains even though yield differences between some treat-

ments did not appear to be very different. The narrow range of

environments in this experiment does not allow for a clear delin-

eation of recommendation domains. We divided our E range into

three recommendation domains (RD), these being:

RD 1; 0.044 < E < 0.200

RD 2; 0.200 < E < 1.300

RD 3; 1.300 < E < 1.730

In RD 1, the lesser productive environment, 100 Ibs N

produced the highest yields. For RD 2 the highest yielding

treatment was 200 lbs N. The yield differences between 200 and

300 Ibs N in RD 3 are not very discernible; we therefore consult-

ed our net income analysis, and decided that although the higher

N treatment out-yielded the lower one, in economic terms the

profit would be less with 300 lbs N in this range. In RP3 3001bs

of N resulted in higher yields.

AnalYtsis of N Treaments With 115 Observations

This analysis is graphed on the following page. There are

no obvious differences in








Effect of ~1 on Radish Yiield (115 observ,)


N::1

N::3

=1:4


I, BIT!

1, 37-








V


'I

Sf-


e'rv:ironmental index














graphs and RDs to the previous analysis with 120 observations.

All following analyses will therefore use 120 observations.

Soil Phoshorus Anal ysis

Soil P content gives an indication of plant available P. In

these trials the soil P content of north locations (range 1480 to

1630ppm) was much higher than that of south locations (175 to

297ppm). This is related to the higher clay content of the north

location. Regression of soil P content with yield showed that

there is a positive correlation between the two. Lower soil P

content was corresponded to lower yields. This is graphed on the

next page and a scatterplot graph is presented in Appendix 2.

Soil Potassium Analysis

The soil K content range was 71.4 113 ppm and 36.7 58.9

ppm in the north and south location, respectively. Noteworthy is

the fact that the lowest average farm yield in the north farms

(farm 4) corresponded to the lowest soil K content, and that the

highest south farm average yield had the highest K content in

that area. In our opinion the north farms had near adequate

plant available K, while in the south this quantity was very low

(even more so if we consider that soil K is not all plant avail-

able). A positive correlation exists between soil K and yield

and this is graphed on the page after the soil P graph. A

scatterplot graph is included in Appendix 2.

Water Stress

Although no quantified data was available, observations

indicate that water stress occurred to a greater extent in south

farms.







E Ef'Ict oif Soil Pi:OsphfOi on Radish Yield







_Effect of S:oil Potassium oln Radishl Yield



L6,65-- -----
1,1.6- /







7 .








Soiil Potass!iu# ;s v, Soil Phosphokln s

17636-6



V /
-,-
HIM-




5 ,66-
,e







5 6 6 "



1.( 5--------
56,,7 76,7 96,7 11
0 t ass:i, u (pp~M













Soil Nitrogen, Phosohorus, and Potassium

A linear regression was performed using K as the independent

variable and P as the dependent one. A strong positive relation-

ship exists between the two. This relation is graphed on the

preceding page. To further our study of the effects of N, P, and

K on radish growth an analysis of the effect of the three

variables on yield was conducted using the Stepwise procedure in

SAS, with the variables pH, N, P, K, Al, and O.M. This analysis

showed P as constituting the best one-variable model, P and N as

constituting the best two-variable model, and N, P, and K gave

the best three-variable model. The other variables did not yield

a noteworthy model. This procedure is included in Appendix 3.

From this analysis we concluded that part of the interaction of N

treatments was due to very diverse P and K contents of the soil

at each location.

Economic Analysis Results

The method that was used to estimate Net Income (NI), in

US$, was described in the 'Materials and Methods' section. From

regression with E we obtained the following equations:

0 Ibs N; NI= -22.2 + 66.357-E

100 lbs N; NI = -70.2 + 364.054*E

200 lbs N; NI = -147.5 + 552.525*E

300 Ibs N; NI = -299.3 + 654.64*E

400 lbs N; NI = -269.0 + 612.43*E

A graph of these equations is presented on the following


page.









Effect of Niitogen on Nett Incomie for nadlishes
~i t18 i tl=,t ----------------
W::i.
A H-M
a it: a
638' N::3



-'' ,


34 ..N~"- -


A J --'
S~h j / *














From the graph on the previous page we divided the E range

into three Net Income Recommendation Domains (NIRDs), and these

were:

NIRD 1; 0.0447< E < 0.400

NIRD 2; 0.400 < E < 1.400

NIRD 3; 1.400 < E < 1.732

A farmer operating in NIRD 1 will increase net income using

100 Ibs N/A. For NIRD 2 the 200 lbs N/A results in higher net

incomes. For NIRD 3 the 300 Ibs of nitrogen per acre correspond

to the higher net incomes.

In our Net Income analysis we have not considered the

effects of radish quality on income. In the trials, radishes

from the heavier texture north farms were irregularly shaped,

while those of the lighter textured south farms tended to be

rounder. As a result, the radishes from the north farms will

fetch a lower market price than their southern counterparts.

This can be factored into our model by considering that the Net

Income of farms with E < 0.925 will be somewhat above the

regression lines, while that of the farms above this value will

be below the regression lines.














V. CONCLUSION


This experiment showed us that different nitrogen levels

have an effect on radish growth. By using a greater planting

density than all the other teams, Team III had more leaf produc-

tion than any other team, but less radish weight than some of the

other teams, demonstrating that planting density has an effect on

radish production.

Through the MSA analysis we were able to establish three

recommendation domains for Yield/A and Net Income; these domains

are also the constraints considered. We would recommend nitrogen

treatments of 100, 200 and 300 Ibs/acre for low, medium and high

productivity environments (respectively) in order to maximize

yields and net income. Radishes grown in soils containing more

clay had irregular shapes and would therefore receive smaller

returns at market.

The analysis of soil Potassium and Phosphorus showed a

correlation between these two elements and radish production.

Lack of potassium occurred in the south farms, although no

visible symptoms were present (hidden hunger), as some of the

major contributions of this nutrient occur in the root zone. The

variability of K and P soil contents were probably the greatest

contributors to the observed interaction between nitrogen

treatments. We would recommend that K and P be added to all of

the south farms at a rate of 50 and 100 lbs/A, respectively.

Some of the north farms would also benefit from the addition of

potassium.














Water stress was a problem, especially in the south farms

where soil clay content was lower. This could have been compen-

sated for by planting deeper and/or through more frequent

irrigation.












APPENDIX I











DATA RADISH;
FARM PH BPH
N1 5.8 7.37
N2 6.0 0.00
N3 6.3 0.00
N4 6.0 0.00
N5 6.3 0.00
N6 6.1 0.00
$1 6.1 0.00
82 5.9 7.66
S3 6.3 0.00
S4 6.1 0.00
S5 5.6 7.64
S6 5.9 7.70
YEAR FARM
87 N1
87 N1
87 N1
87 N1
87 N1
87 N1
87 N1
87 N1
87 N1
87 N1
87 N2
87 N2
87 N2
87 N2
87 N2
87 N2
87 N2
87 N2
87 N2
87 N2
87 N3
87 N3
87 N3
87 N3
87 N3
87 N3
87 N3
87 N3
87 N3
87 N3
87 N4
87 N4
87 N4
87 N4
87 N4
87 N4
87 N4


P K
1480 94.9
1490 97.1
1630 113
1590 71.1
1560 102
1560 77.2
261 41.f
428 58.9
175 38.E
189 36.'
231 55.E
297 46.'
REP NTRT
1 0
1 100
1 200
1 300
1 400
2 0
2 100
2 200
2 300
2 400
1 0
1 100
1 200
1 300
1 400
2 0
2 100
2 200
2 300
2 400
1 0
1 100
1 200
1 300
1 400
2 0
2 100
2 200
2 300
2 400
1 0
1 100
1 200
1 300
1 400
2 0
2 100


AL OM
? 674 0.71
715 0.81
678 0.91
S662 0.71
704 0.841
670 0.74
5 384 0.65
359 0.52
3 394 0.58
S377 0.45
2 520 0.55
S398 0.94
YLD
12.5
51
212.5
90
264
27
78
330
260
240
0
309.7
479.5
371.2
486.4
32.09
192.06
341.9
370.3
301.3
22
86
72
207
86
231
227
187
234
296
0
217
83
48
105
150
167


E
0.9394695
1.73153533
0.98929440
0.93226590
1.5271632
1.34347140
0.055903125
0.92272113
0.2095047
0.268134
0.14064714
0.1530765











87 N4 2 200 334
87 N4 2 300 125
87 N4 2 400 324
87 N5 1 0 60
87 N5 1 100 147
87 N5 1 200 151
87 N5 1 300 465
87 N5 1 400 289
87 N5 2 0 36
87 N5 2 100 317
87 N5 2 200 373
87 N5 2 300 349
87 N5 2 400 357
87 N6 1 0 0
87 N6 1 100 312
87 N6 1 200 177
87 N6 1 300 353
87 N6 1 400 243
87 N6 2 0 26
87 N6 2 100 186
87 N6 2 200 348
87 N6 2 300 356
87 N6 2 400 237
87 51 1 0 0
87 S1 1 200 10
87 81 1 300 15
87 S1 1 400 13
87 S1 2 0 12.5
87 S1 2 100 10
87 S1 2 200 9
87 Si 2 300 5
87 S2 1 0 33.3
87 S2 1 100 103.7
87 S2 1 200 341.7
87 S2 1 300 181.7
87 S2 1 400 368.1
87 S2 2 0 20.7
87 S2 2 100 152.8
87 S2 2 200 149.9
87 S2 2 300 46.5
87 S2 2 400 138.7
87 S3 1 0 0
87 S3 1 100 34
87 S3 1 200 124
87 S3 1 300 15
87 S3 1 400 33
87 S3 2 0 0
87 S3 2 100 45
87 S3 2 200 16
87 S3 2 300 37
87 S3 2 400 45
87 S4 1 0 23











87 S4 1 100 132
87 S4 1 200 37
87 S4 1 300 21
87 S4 1 400 31
87 S4 2 0 0
87 S4 2 100 97
87 S4 2 200 55
87 S4 2 400 6
87 S5 1 0 5
87 S5 1 100 33
87 S5 1 200 11
87 S5 1 300 50
87 85 2 0 0
87 S5 2 100 17
87 S5 2 200 48
87 S5 2 300 0
87 S6 1 0 0
87 56 1 100 51
87 S6 1 200 20
87 S6 1 300 5
87 S6 1 400 14
87 86 2 0 5
87 Sb 2 100 79
87 S6 2 200 40
87 S6 2 300 6













APPENDIX II












17:04 Saturday, April 16, 1988 10E


Plot of YIELD*P


Legend: A = 1 obs, B = 2 obs, etc.


B A
B
A C
A A


B A


A A B
AA
A A


AA A
CAA B
BCA
B BFB
BCEDD


B AB
A
A A
AA
B AA
A
A AA


800 1000 1200 1400 1600 1800


YIELD 1

3.0 +


2.5 +


2.0 +


1.5 +


1.0 +


0.5 +





0.0 +


SAS


0 200 400


600













17:04 Saturday, April 16, 1988 100


Plot of YIELD*K


Legend: A = 1 obs, B = 2 obs, etc.


B A
A
A A


YIELD 1

3.0 +





2.5 +






2.0 +





1.5 +





1.0 +





0.5 +





0.0 +


----------+------+-----------+------+----+-------+---- ------+--


70 80


SAS


A A
A
A A
A


110


120


30 40


100













SAS 17:04 Saturday, April 16, 1988 104


Plot of P*K Legend: A = 1 obs, B = 2 obs, etc.

P



1800 +


J
1600 + J
J J
*J J

1400 +



1200 +



1000 +



800 +



600 +



400 +


J J
200 + JJ



0 +
----------+------+------+-----+------+---- ---------- +---------

30 40 50 60 70 80 90 100 110 120

K
I













APPENDIX III







SAS 17:04 Saturday, April 16, 1988 113

Maximum R-square Improvement for Dependent Variable YIELD

Step 1 Variable P Entered R-square = 0.37282837 C(p) = 22.02179879

DF Sum of Squares Mean Square F Prob>F

Regression 1 28.84018890 28.84018890 70.15 0.0001
Error 118 48.51494694 0.41114362
Total 119 77.35513584

Parameter Standard Type II
Variable Estimate Error Sum of Squares F Prob>F

INTERCEP 0.07436077 0.10074279 0.22400277 0.54 0.4619
P 0.00075665 0.00009034 28.84018890 70.15 0.0001

Bounds on condition number: 1.0000, 1.0000


The above model is the best 1 variables model found.

A:\REGSTEP.WP Doc 1 Pg 87 Ln 13 POS 2

Step 2 Variable NTRT Entered R-square = 0.46791961 C(p) = 3.09504714

DF Sum of Squares Mean Square F Prob>F

Regression 2 36.19598527 18.09799263 51.45 0.0001
Error 117 41.15915057 0.35178761
Total 119 77.35513584

Parameter Standard Type II
Variable Estimate Error Sum of Squares F Prob>F

INTERCEP -0.27577721 0.12061107 1.83917467 5.23 0.0240
NTRT 0.00175069 0.00038286 7.35579637 20.91 0.0001
P 0.00075665 0.00008357 28.84018890 81.98 0.0001

Bounds on condition number: 1.0000, 4.0000
----------------------------------------------------------------------------------


The above model is the best 2 variables model found.

Step 3 Variable K Entered R-square = 0.47449752 C(p) = 3.64744568

DF Sum of Squares Mean Square F Prob>F

A:\REGSTEP.WP Doc 1 Pg 87 Ln 48 POS 2


SAS


17:04 Saturday, April 16, 1988 11h


Maximum R-square Improvement for Dependent Variable YIELD
r. th- - -- ~_C1~~-- JC) L


.. .~.~I

















Step 3 Variable K Entered


Sum of Squares


Regression
Error
Total


3
116
119


36.70482017
40.65031567
77.35513584


Mean Square

12.23494006
0.35043376


F Prob>F


34.91 0.0001


17:04 Saturday, April 16, 1988 114


Variable

INTERCEP
NTRT
P
K


Parameter
Estimate

-0.49233750
0.00175069
0.00054231
0.00591920


Standard
Error


0.21630966
0.00038212
0.00019646
0.00491222


Type II
Sum of Squares


1.81543185
7.35579637
2.67036091
0.50883490


F Prob>F


5.18
20.99
7.62
1.45


0.0247
0.0001
0.0067
0.2307


Bounds on condition number: 5.5480, 36.2882


The above model is the best 3 variables model found.
A:\REGSTEP.WP Doc 1 Pg 87 Ln 45 POS 4


SAS


R-square = 0.47449752


C(p) = 3.64744568







Step 4 Variable OM Entered


Sum of Squares


Regression
Error
Total


Variable

INTERCEP
NTRT
P
K
OM


4
115
119


37.31653063
40.03860520
77.35513584


Parameter
Estimate

-0.20540245
0.00175069
0.00056019
0.00773333
-0.61227572


Standard
Error


0.30552663
0.00038088
0.00019628
0.00508395
0.46191773


Mean Square

9.32913266
0.34816178


Type II
Sum of Squares


0.15735994
7.35579637
2.83588976
0.80558514
0.61171047


F Prob>F


26.80 0.0001


F Prob>F


0.45
21.13
8.15
2.31
1.76


0.5027
0.0001
0.0051
0.1310
0.1876


Bounds on condition number: 5.9815, 56.8366


The above model is the best 4 variables model found.


Step 5 Variable AL Entered
A:\REGSTEP.WP


R-square = 0.48579953 C(p) = 5.16021035
Doc 1 Pg 88 Ln 39 POS 4


C(p) = 3.90717006


R-square = 0.48240534










Step 5 Variable AL Entered


Regression
Error
Total


Variable


5
114
119


Sum of Squares

37.57908851
39.77604732
77.35513584


Parameter
Estimate


Standard
Error


Mean Square

7.51581770
0.34891270


Type II
Sum of Squares


F Prob>F


21.54 0.0001


F Prob>F


INTERCEP 0.12882724 0.49193332 0.02392873 0.07 0.7939
NTRT 0.00175069 0.00038129 7.35579637 21.08 0.0001


0.00073443
0.00916365
-0.00105086
-0.63998189


0.00028099
0.00534986
0.00121141
0.46351729


2.38365483
1.02369151
0.26255788
0.66515101


Bounds on condition number: 11.3990, 156.3


The above model is the best 5 variables model found.


A:\REGSTEP.WP


Doc 1 Pg 89 Ln 2


Step 6 Variable PH Entered


Regression
Error
Total


Variable


INTERCEP
NTRT
PH
P
K
AL
OM


6
113
119


R-square = 0.48652752


Sum of Squares

37.63540278
39.71973306
77.35513584


Parameter
Estimate

-0.63104943
0.00175069
0.11892670
0.00068909
0.00914262
-0.00085709
-0.66931739


Standard
Error


1.96160258
0.00038270
0.29712150
0.00030393
0.00536993
0.00130873
0.47097158


Mean Square

6.27256713
0.35150206


Type II
Sum of Squares


0.03637748
7.35579637
0.05631427
1.80689073
1.01890162
0.15075644
0.70990900


C(p) = 7.00000000

F Prob>F


17.85 0.0001


F Prob>F


0.10
20.93
0.16
5.14
2.90
0.43
2.02


0.7483
0.0001
0.6897
0.0253
0.0914
0.5139
0.1580


Bounds on condition number: 13.2381, 216.5

The above model is the best 6 variables model found.
No further improvement in R-square is possible.


Doc 1 Pg 89 Ln 11 POS 3


6.83
2.93
0.75
1.91


0.0102
0.0895
0.3875
0.1701


POS 2


C(p) = 5.16021035


R-square = 0.48579953


A:\REGSTEP.WP




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs