.I

 

‘

 

Ya. riff-2:.
13...... v 3.1;..1 Halal
‘il.if..alfl Iva}...
E):A:Sl
:11
K

i 1’.

29'. iii-.1 If...

. .7. .3117. rtréul.‘
51...

 

SITY LIB BRARI ES

' IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIII III II

3 129300 80077

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

MANAGEMENT STRATEGIES TO HINIHIZE
NITRATE LEACHING
IN SEED CORN PRODUCTION

presented by

EDWARD CHARLES MARTIN

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in AGRICULTURE

941ng

Major professor

Date May 18, 1992

 

MSU is an Affirmative Action/Equal Opportunity Institution 0. 12771

 

 

LIBRARY ‘
I Mlchlgan State
I. University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

law/'5 I I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

._I
IEJI—I
__II__IW

II—II I

MSU to An Affirmative Action/Bond Opportunity lpetitutton
Wm:

 

 

 

 

 

 

 

 

 

 

 

MANAGEMENT STRATEGIES TO MINIMIZE
NITRATE LEACHING
IN SEED CORN PRODUCTION

By
Edward Charles Martin

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Agricultural Engineering

1992

fl 678 7

ABSTRACT
MANAGEMENT STRATEGIES TO MINIMIZE

NI'I'RAT'E LEACHING
IN SEED CORN PRODUCTION

By
Edward Charles Martin

Increasing nitrate concentrations in groundwater in Southwest Michigan has caused the
public to perceive the seed corn industry as a source of the problem. In response, growers are
seeking alternative management strategies to reduce their nitrogen fertilizer applications.
However, as growers apply less nitrogen fertilizer, they risk profit loss if they apply too little. little.

To help evaluate the dimensions of this problem, a simulation model called CERES-1M
was developed to evaluate the impact of nitrogen management strategies on plant growth and
nitrate leaching in seed corn production. Patterned after the CERES-Maize model, CERESoIM
simulated the management operations unique to seed corn production, including the distinction
between male and female plants, and field operations such as the detasseling of the female plants
and the removal of the male plants.

CERES-1M was validated using results from a nitrogen fertilizer study conducted with
hybrid maize. Not using the inbred maize options for these simulations, the yield and biomass
values simulated compared well to the measured values. Also, the drainage and nitrate leaching
values simulated compared well with values measured using drainage iysimeters.

CERES-1M was also compared with two years of data from a nitrogen fertilizer
management study conducted with inbred maize. Using drainage lysimeters, nitrogen strategies
were evaluated to determine their impact on yield and nitrate leaching. CERES-1M showed good
comparisons with the measured biomass data. The mean squared-error of prediction was used to
evaluate the predictive accuracy of the model. All values were less than 12.3% except for the
stover biomass, which was consistently underestimated. The drainage data compared well for all
the iysimeters. The nitrate leaching data also compared well, though there were some

discrepancies during the first year.

Finally, various nitrogen management strategies, including a conventional strategies
presently used by growers and a management strategy called plant response fertilization (PRF),
were simulated using a continuous, multi-year simulation option. The basic premise of PRF is to
apply nitrogen only when the plants experience a nitrogen deficiency. The amount of fertilizer
applied is limited to the amount of nitrogen required to complete the plant’s growth. After
evaluating the strategies based on yield, nitrate leaching, and revenue, a split nitrogen application
was the strategy of choice. Applying minimal nitrogen at planting and one additional application
at cultivation (growth stage V6), the amount of nitrate leaching was minimal and the revenue and
yield were among the highest of the 17 strategies evaluated. For the initial conditions used in
these simulations, a strategy of applying 30 kg N ha’1 at planting with an additional application of
80 kg N liar1 at cultivation time was optimal.

This work demonstrates the value of simulation to evaluate managements strategies to
accomplish the dual goal of minimizing nitrate leaching and maximizing profit. It shows that valid
models can assist growers and policy makers in making informed decisions and help educate the

public on the effects of management schemes on plant growth and the environment.

Dedicated to Henry Miller, the late Dennis Cupp, and all
farmers who participate in on-farm research with their
local universities. Your efforts strengthen agricultural

research and its impact on society.

iv

ACKNOWLEDGEMENTS

I would like to thank of those who have supported me over the past years as I worked
toward this goal. First and foremost is my deep gratitude to my advisor Dr. Joe '1‘. Ritchie. In
1984, we began a professional career together at Michigan State University that has developed into
a friendship of respect and love that will last forever.

Also, I would like to thank the other members of my graduate committee. Dr. Vincent
Bralts, Agricultural Engineering, who gave me the opportunity to work in the Michigan Energy
Conservation Program, an experience I won’t soon forget. Also, Dr. Ted London, Agricultural
Engineering, with whom I have worked since my first day in 1984. My experiences with Dr.
London were the impetus that encouraged me to pursue a career in the Cooperative Extension
Service. Finally, thanks to Fred Henningsen, Cooperative Extension Service, who taught me the
fine art of balancing research with clientele needs. There will not be another Extension agent like
Fred for many years to come.

I would also like to acknowledge the support of Pioneer I-Ii-Bred International, especially
Steve Lupkes, John Lightner, and Richard Peterson. Their assistance, along with many support
personnel, helped to make the seed corn management project a success.

I also want to express my gratitude to my colleagues of the Homer Nowlin Chair. The
number of people is too numerous to name, but I need to mention a special thanks to Brian Baer,
who has helped me with programming and has supported me throughout the years. Also, a very
large thanks to Sharlene Rhines, who not only helped to edit this work, but has supported me
through countless numbers of projects.

Finally, from the bottom of my heart, I thank my family. The support and love I received
from you has been endless. A special thanks to my brother Bob, who has witnessed a
transformation that began my first day on campus in 1978. I cannot thank you enough for your
support over the years. You have been my advisor, my banker, and most importantly, my friend.

Thanks!

TABLE OF CONTENTS

List of Tables ............................................................. viii
List of Figures .............................................................. x
Introduction ................................................................ 1
Management System Components ......................................... 4
Irrigation Considerations .......................................... 4

Crop Considerations ............................................. 5

Soil Considerations .............................................. 7

Objectives ........................................................... 7

Review of Literature ......................................................... 8
Seed Corn Management ................................................ 8
Growing Seed Corn ............................................. 8

Detasseling .................................................... 9

Removal of the Male Plants ...................................... 10

Harvest ...................................................... 10

Nitrogen Management and Scheduling ..................................... 10

Nitrate Movement Through the Rootzone and Plant Uptake ............. 10

Leaching of Nitrates ............................................ 12

Irrigation Management and Its Effect on Nitrogen Management ................. 14

Maize Growth Simulation Models ........................................ 15
Methodology .............................................................. 20
Objective 1: Development of an inbred maize simulation model ................. 20
Commercial Hybrid Maize Study ................................... 20

Development of CERES-1M ...................................... 25

Objective 2: Perform field experiments to validate the simulation model ........... 55

Inbred Calibration/Validation ..................................... 55

vi

Objective 3: Use the simulation model to evaluate the impact of several nitrogen

management strategies on potential leaching in seed corn production ....... 68

CERES-1M Nitrogen Simulations .................................. 68

Evaluation of Nitrogen Management Strategies ........................ 73

Results and Discussion ....................................................... 78

Commercial Hybrid Maize .............................................. 78

Input Data ................................................... 78

Simulated Versus Measured ...................................... 84

Inbred Maize ........................................................ 98

Plant Response Fertilization ...................................... 98

Input Data ................................................... 99

Simulated Versus Measured ..................................... 105

Evaluation of Nitrogen Management Strategies ............................. 135

Summary and Conclusions ................................................... 146
Appendices

Appendix A. Listing of FORTRAN code for new subroutines written ............ 155

Appendix 3. Sample output files of CERES-Maize (CERES-1M) ............... 160

Appendix C. Input data used for hybrid simulations ......................... 168

Appendix D. Input data used for inbred simulations ......................... 192

Bibliography ............................................................. 211

vii

LIST OF TABLES

Table 1. Partial list of published crop growth simulation models taken from Jones and

Ritchie, 1991 ............................................................... 17
Table 2. Average percent biomass loss due to female detasseling for all inbreeds.

Constantine, MI, 1990. ....................................................... 30
Table 3. Linear regression analysis for each individual hybrid for area per leaf for

leaves from the largest leaf to the flag leaf. ....................................... 41
Table 4. Total number of leaves and the leaf number of the largest leaf for each hybrid. ..... 45

Table 5. Measured and predicted values of the leaf number of the largest leaf for
the hybrids. ............................................................... 46

Table 6. Measured and predicted leaf area (cm2) of the largest leaf and the associated
genetic parameter P6 for each hybrid. ............................................ 46

Table 7. Results of regression analysis performed between measured and predicted areas
of leaves for both the Gompertz and linear functions. ............................... 47

Table 8. Nitrogen treatments for 1989 and 1990, inbred nitrogen management study.
Constantine, MI ............................................................. 67

Table 9. List of fertilizer types and their codes used in the CERES-Maize simulation model. . . 71
Table 10. Nitrogen management strategies simulated by CERES-1M for a 60 year period. . . . . 75

Table 11. Averages of the measured soil nitrogen values and the associated standard deviations
for the CMP plot from samples taken on April 15, 1988. Mendon, MI. .................. 81

Table 12. Averages of the measured soil nitrogen values and the associated standard deviations
for the RMP plot from samples taken on April 15, 1988. Mendon, ML .................. 81

Table 13. Averages of the measured soil nitrogen values and the associated standard deviations
for the CMP plot from samples taken on April 11, 1989. Mendon, MI. .................. 82

Table 14. Averages of the measured soil nitrogen values and the associated standard deviations
for the RMP plot from samples taken on April 11, 1989. Mendon, MI .................... 82

Table 15. Yield and stover data measured for the CMP and RMP plots for 1988 and 1989.
Mendon, MI. .............................................................. 85

Table 16. A comparison of several parameters between CERES-1M simulated values and
measured values for the CMP treatment. Mendon, MI, 1988. .......................... 86

Table 17. A comparison of several parameters between CERES-IM simulated values and
measured values for the RMP treatment. Mendon, MI, 1988. ......................... 86

Table 18. A comparison of several parameters between CERES-1M simulated values and
measured values for the CMP treatment. Mendon, MI, 1989 ........................... 87

Table 19. A comparison of several parameters between CERES-1M simulated values and those
measured in the RMP treatment. Mendon, MI, 1989. ............................... 87

Table 20. A comparison of several parameters between CERES-1M simulated (using
300 kg N ha”1 input) and measured values. Mendon, MI, 1988. ........................ 94

Table 21. Averages and the associated standard deviations (SD) for measurements taken for
nitrate-N and ammonium-N concentration in the soil for the four treatments from samples
taken on May 4, 1989. Constantine, ML. ........................................ 102

Table 22. Averages and the associated standard deviations (SD) for measurements taken for
nitrate-N and ammonium-N concentration in the soil (0-75 cm depth) for the four treatments
from samples taken on April 25, 1990. Constantine, MI .............................. 103

Table 23. Average and the associated standard deviation for mineral soil nitrogen
measurements from samples taken on April 25, 1990. Constantine, ML. ................ 104

Table 24. Averages and the associated standard deviations (SD) for yield, biomass, and plant
nitrogen measurements taken at physiological maturity for 1989. Constantine, MI. ........ 106

Table 25. Averages and the associated standard deviations (SD) for yield, biomass, and plant
nitrogen measurements taken at physiological maturity for 1990. Constantine, M1. ........ 107

Table 26. Coefficient of variation expressed as a percentage for yield, biomass, and
plant nitrogen measurements taken at physiological maturity for 1989 and 1990.
Constantine, MI. .......................................................... 108

Table 27. Values of the MSEP for CERES-1M simulated versus measured values for plant
and biomass data collected for the 3 inbred varieties. Constantine, MI, 1989 and 1990. ..... 112

Table 28. MSEP values (kg ha“) of CERES-1M simulated values for above ground biomass
for the 3 inbreeds. for all treatments, for 1989 and 1990. Constantine, MI. .............. 122

Table 29. List of the lysimeters and the nitrogen treatments used for 1989 and 1990.
Constantine, MI ........................................................... 124

Table 30. Coefficient of variation expressed as a percentage for grain yields for a 60 years
CERES-N simulation run. .................................................. 136

Table 31. Coefficient of variation expressed as a percentage for grain yields for a 60 years
CERES-1M simulation run. .................................................. 144

LIST OF FIGURES

Figure l. A map of Michigan showing the areas where high nitrate concentrations have been
measured in groundwater samples. ............................................... 2

Figure 2. A schematic drawing of the top view and end view of one of the lysimeters used in the
nitrogen management study in commercial hybrid maize production. All measurements are in
meters. Mendon, MI. ........................................................ 22

Figure 3. A schematic drawing of the side view of one of the lysimeters used in the nitrogen
management study in commercial hybrid maize production. All measurements are in meters.
Mendon, MI. .............................................................. 23

Figure 4. A schematic drawing of one of the lysimeters used in the nitrogen management
study in commercial hybrid maize production. Mendon, MI. .......................... 24

Figure 5. A schematic drawing of the irrigation system used on the RMP plot. The curved lines
represent the radius of coverage of the four sprinklers. All measurements are in meters.
Mendon, MI. .............................................................. 26

Figure 6. The regression analysis between the linearization of the Gompertz equation using the
area of the largest leaf for A0 and individual leaf number for the seven hybrids tested. ....... 38

Figure 7. The area per leaf versus leaf number for the seven hybrid varieties tested. Also
included is the Gompertz equation using a value of 1000 for A0. ....................... 38

Figure 8. The regression analysis between the linearization of the Gompertz function using a
value of 1200 for A0 and individual leaf numbers for the seven hybrids tested. ............. 40

Figure 9. The area per leaf versus leaf number for the seven hybrid varieties analyzed. Also
included is the Gompertz equation using a value of 1200 for A0. ....................... 41

Figure 10. The regression lines for the area per leaf versus leaf number for leaves after the
largest leaf for varieties A632 and X304. ......................................... 43

Figure 11. The regression lines for the area per leaf versus leaf number for leaves after the
largest leaf for varieties B73 and Z883 ............................................ 43

Figure 12. The regression lines for the area per leaf versus leaf number for leaves after the
largest leaf for varieties M186, M187, and 2882. .................................... 44

Figure 13. Regression line for the leaf number of the largest leaf and the total number of
leaves for six of the hybrid varieties. ............................................. 45

Figure 14. Measured values and predicted values using the Gompertz and linear equations of
area per leaf for hybrids A632 and X304. ......................................... 48

Figure 15. Measured values and predicted values using the Gompertz and linear equations of
area per leaf for hybrids B73 and Z883. .......................................... 48

Figure 16. Measured values and predicted values using the Gompertz and linear equations of
area per leaf for hybrids M186, M187, and Z882. .................................... 49

Figure 17. A schematic drawing of the top view and end view of one of the lysimeters used
in the nitrogen management study in seed corn production. All measurements are in meters.
Constantine, M1. ........................................................... 57

Figure 18. A schematic drawing of the side view of one of the lysimeters used in the
nitrogen management study in seed corn production. All measurements are in meters.
Constantine, Ml. ........................................................... 58

Figure 19. A schematic drawing of one of the lysimeters used in the nitrogen management
study in seed corn production. All measurements are in meters. Constantine, M1. ......... 59

Figure 20. A side view of the sequential sampler used to sample the drainage water from
the lysimeters used in the nitrogen management study for seed corn production.
Constantine, MI. ........................................................... 61

Figure 21. A top view of the sequential sampler used to sample the drainage water from
the lysimeters used in the nitrogen management study for seed corn production.
Constantine, MI. ........................................................... 62

Figure 22. A diagram of the plot layout used in the nitrogen management in seed corn
production study. Constantine, MI. ............................................. 65

Figure 23. A flow diagram of the sequence of nitrogen simulations in the CERES-Maize
model, V2.1. .............................................................. 69

Figure 24. Measured and CERES-1M simulated leaf area index values for the CMP treatment
for 1988. Mendon, MI. ...................................................... 89

Figure 25. Measured and CERES-1M simulated leaf area index values for the RMP treatment
for 1988. Mendon, MI. ...................................................... 89

Figure 26. Measured and CERES-1M simulated leaf area index values for the CMP treatment
for 1989. Mendon, M1. ...................................................... 90

Figure 27. Measured and CERES-1M simulated leaf area index values for the RMP treatment
for 1989. Mendon, M1. ...................................................... 90

Figure 28. Measured and CERES-1M simulated drainage for the CMP lysimeter for 1988-89.
Precipitation is also shown. Mendon, MI. ........................................ 92

Figure 29. Measured and CERES-1M simulated drainage for the RMP lysimeter for 1988-89.
Precipitation is also shown. Mention, MI. ........................................ 92

xi

Figure 30. Measured and CERES-1M simulated (for both 200 and 300 kg N ha'1 pp) nitrate
leaching values for the CMP lysimeter for 1988-89. Mendon, MI. ...................... 93

Figure 31. Measured and CERES-1M simulated nitrate leaching values for the RMP lysimeter
for 1988-89. Mendon, MI. .................................................... 93

Figure 32. Measured and CERES-1M simulated drainage values for the CMP lysimeter for
1989-90. Precipitation is also shown. Mendon, M1. ................................. 96

Figure 33. Measured and CERES-1M simulated drainage values for the RMP lysimeter for
1989-90. Precipitation is also shown. Mendon, M1. ................................. 96

Figure 34. Measured and CERES-1M simulated nitrate leaching values for the CMP lysimeter
for 1989-90. Mendon, MI. .................................................... 97

Figure 35. Measured and CERES-1M simulated nitrate leaching values for the RMP lysimeter
for 1989-90. Mendon, MI. .................................................... 97

Figure 36. CERES-1M simulated versus measured grain yield for the three inbreeds and the
four treatments for 1989 and 1990. Constantine, MI. ............................... 111

Figure 37. CERES-1M simulated and measured kernel weight for the three inbreeds for the four
treatments for 1989 and 1990. Constantine, MI. .................................. 111

Figure 38. CERES-1M versus measured grains per plant for three inbreeds and the four
treatments for 1989 and 1990. Constantine, MI. .................................. 113

Figure 39. CERES-1M versus measured stover yield for the three inbreeds and the four
treatments for 1989 and 1990. Constantine, MI. .................................. 113

Figure 40. CERES-1M versus measured percent nitrogen in the grain for the three inbreeds
and the four treatments for 1989 and 1990. Constantine, MI. ......................... 115

Figure 41. CERES-1M versus measured nitrogen uptake by the grain for the three inbreeds
and the four treatments for 1989 and 1990. Constantine, M1 .......................... 115

Figure 42. CERES~IM versus measured percent nitrogen in the stover for the three inbreeds
and the four treatments for 1989 and 1990. Constantine, MI .......................... 116

Figure 43. CERES-1M versus measured nitrogen uptake by the stover for the three inbreeds
and the four treatments for 1989 and 1990. Constantine, M1 .......................... 116

Figure 44. CERES-1M and measured above ground biomass values for inbred 2, treatment 1
(180pp). Constantine, MI, 1989. .............................................. 118

Figure 45. CERESJM and measured above ground biomass values for inbred 2, treatment 2
(PRF - 80 kg N int-1). Constantine, M1, 1989. .................................... 118

Figure 46. CERES-1M and measured above ground biomass values for inbred 2, treatment 3
(905). Constantine, MI, 1989. ................................................ 119

Figure 47. CERES-1M and measured above ground biomass values for inbred 2, treatment 4
(1805). Constantine, M1, 1989 ................................................. 119

xii

Figure 48. CERES- IM and measured above ground biomass values for inbred 2, treatment 1
(180pp). Constantine, MI, 1990. .............................................. 120

Figure 49. CERES-1M and measured above ground biomass values for inbred 2, treatment 2
(PRF- 0 kg N ha'l). Constantine, MI,1990. ..................................... 120

Figure 50. CERESJM and measured above ground biomass values for inbred 2, treatment 3
(905). Constantine, M1, 1990. ................................................ 121

Figure 51. CERES-1M and measured above ground biomass values for inbred 2, treatment 4
(0 kg N its-1). Constantine, M1,1990. .......................................... 121

Figure 52. CERES-1M and measured LAI values for inbred 1, treatment 1 (180pp).
Constantine, MI, 1989. ...................................................... 123

Figure 53. CERES-1M and measured LAI values for inbred 2, treatment 1 (180pp).
Constantine, MI, 1989. ...................................................... 123

Figure 54. CERES-1M and measured LAI values for inbred 3, treatment 1 (180pp).
Constantine, MI, 1989. ...................................................... 124

Figure 55. CERES~IM and measured drainage values for lysimeter 1, treatment 3 (905).
Precipitation is also shown. Constantine, MI, 1989. ................................ 125

Figure 56. CERES-1M and measured drainage values for lysimeter 2, treatment 2
(PRF- 0 kg N ha‘l). Precipitation is also shown. Constantine, MI,1989. ............... 125

Figure 57. CERES-1M and measured drainage values for lysimeter 4, treatment 4 (1805).
Precipitation is also shown. Constantine, MI, 1989. ................................ 126

Figure 58. CERES-1M and measured drainage values for lysimeter 5, treatment 1 (180pp).
Precipitation is also shown. Constantine, M1, 1989. ................................ 126

Figure 59. CERES-1M and measured nitrate leaching values for lysimeter 1, treatment 3
(9(3). Constantine, M1, 1989. ................................................ 128

Figure 60. CERES-1M and measured nitrate leaching values for lysimeter 2, treatment 2
(PRF - 80 kg N its-1). Constantine, M1, 1989. .................................... 128

Figure 61. CERES-1M and measured nitrate leaching values for lysimeter 4, treatment 4
(1805). Constantine, MI, 1989 ................................................. 129

Figure 62. CERES-1M and measured nitrate leaching values for lysimeter 5, treatment 1
(180pp). Constantine, MI, 1989. .............................................. 129

Figure 63. CERES-1M and measured drainage values for lysimeter 1, treatment 3 (905).
Precipitation is also shown. Constantine, MI, 1990. ................................ 131

Figure 64. CERES-1M and measured drainage values for lysimeter 2, treatment 2
(PRF- 0 kg N ha’l). Precipitation is also shown. Constantine, MI,1990. ............... 131

Figure 65. CERES 1M and measured drainage values for lysimeter 4, treatment 4
(0 kg N its-1). Precipitation is also shown. Constantine, MI, 1990. .................... 132

xiii

Figure 66. CERES-1M and measured drainage values for lysimeter 5, treatment 1 (180pp).
Precipitation is also shown. Constantine, M1, 1990. ................................ 132

Figure 67. CERES-1M and measured nitrate leaching values for lysimeter 1, treatment 3
(905). Constantine, MI, 1989. ................................................ 133

Figure 68. CERES-1M and measured nitrate leaching values for lysimeter 2, treatment 2
(PRF 0 kg N ha’l). Constantine, M1, 1990. ..................................... 133

Figure 69. CERES-1M and measured nitrate leaching values for lysimeter 4, treatment 4
(0 kg N int-1). Constantine, M1, 1990. .......................................... 134

Figure 70. CERES-1M and measured nitrate leaching values for lysimeter 5, treatment 1
(180pp). Constantine, MI, 1990. .............................................. 134

Figure 71. Average grain yield for the 11 strategies evaluated for the 60-year
continuous simulation. ...................................................... 136

Figure 72. Average annual amount of nitrogen fertilizer added for the 11 strategies
evaluated for the 60-year continuous simulation. .................................. 138

Figure 73. Average annual nitrate leaching for the 11 strategies evaluated for
the 60—year continuous simulation. ............................................. 138

Figure 74. Average annual revenue for the 11 strategies evaluated for the 60-year
continuous simulation. ...................................................... 140

Figure 75. Average annual revenue and annual nitrate leaching for the 11 strategies
evaluated for the 60-year continuous simulation. .................................. 140

Figure 76. Average grain yield for the seven sidedressed strategies evaluated
for the 60-year continuous simulation. .......................................... 142

Figure 77. Average annual nitrate leaching for the seven sidedressed strategies
evaluated for the 60-year continuous simulation. .................................. 142

Figure 78. Average annual revenue for the seven sidedressed strategies evaluated
for the 60-year continuous simulation. .......................................... 143

Figure 79. Average annual revenue and nitrate leaching for the seven sidedressed
strategies evaluated for the 60-year continuous simulation. ........................... 143

Figure 80. Cumulative probability of the annual revenue for selected strategies for the
60-year continuous simulation. ................................................ 145

INTRODUCTION

Increasing public awareness of nitrate contamination of groundwater has caused
agliculturalists to focus their attention on nitrogen recommendations and management strategies.
At one time, farmers applied nitrogen fertilizer in sufficient amounts to prevent nitrogen stress
throughout the plant’s life cycle, regardless of plant uptake or leaching potential. Farmers viewed
nitrogen fertilizer over-application as cheap insurance to guard against any nitrogen loss that
might occur during the season. The relatively low cost of nitrogen fertilizer was one factor which
lead to the practice of over-application (Ritchie, personnel communication). However, farmers
are now becoming more keenly aware of the need for better nitrogen management and are seeking
help from research and extension specialists for answers on how to improve nitrogen fertilizer
application rates and timing to reduce the potential for leaching while maintaining a good profit
from their crop yield.

The problem of nitrate contamination of groundwater in Michigan is pervasive.
Exceptionally high nitrate concentrations in groundwater are found in three regions of the state.
As shown in Figure 1, these areas are the Northwest, Central, and Southwest (Kittieson, 1987)
regions. Orchard agriculture is prominent in the Northwest and is one of the largest cherry
production regions in the world. The second area of concern is the Central region where potato is
the major crop. Michigan ranks 10th in potato production in the US (Michigan Agricultural
Statistics, 1989). The third area is the Southwest corner of the state where corn is the major crop
grown. Soybean and dry bean production are also heavy in this region. Seed corn is also grown in
this region with production acres increasing. Seed corn acreage has gone from near zero in 1950

to over 16,000 ha in 1990.

 

 

 

 

Orchard

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Potatoes

Corn
Soybeans

Seed Corn

 

 

 

 

 

Figure 1. A map of Michigan showing the areas where high nitrate concentrations have been
measured in groundwater samples.

3

All three of these regions have at least two common characteristics. First, each area is
dominated by sandy loam and loamy sand soil types. The second is that all three areas have
numerous acres of irrigated land. In the Northwest, the orchards are irrigated mostly by drip or
trickle irrigation while in the Central and Southwest regions, center pivots and traveler or "Big
Gun“ type irrigation systems are mainly used. Because of the soils in these regions, irrigation has
proven to be economical and in many cases, essential.

Though all of these regions are of concern to the general public, farmers in the
Southwest region, seed corn growers in particular, have experienced pressure from the public to
reduce the leaching of nitrates to the groundwater. In one instance, the EPA urged the town of
Constantine, in the center of the seed corn production area, to install a $1.2 million de-ionizer in
their water system to reduce the nitrate-N levels below the 10 ppm government health standard.
Fortunately, a nearby well was located with a low nitrate-N concentration in the water. Mixing
water from this well with water from the well with higher nitrate-N concentrations brought the
water supply within the government standard.

Constantine is located in St. Joseph County, which contains the largest irrigated acreage
of any county in Michigan; approximately 25,000 ha. It also contains the largest area of seed corn
production, with approximately 14,000 ha planted each year.

The St. Joseph County area is dominated by outwash plains that are a result of deposits left by
the glacial melt water in front of the ice. The major parent materials are glacial till, outwash
deposits, alluvium, and organic material. Prairie soils that are considered part of the Great
Prairies of the Midwest are also found in this area. Rivers flow throughout the region and there
are numerous bogs and ponds. Lakes abound along the rivers, especially on the outwash plains.
Groundwater is also plentiful in this area and irrigation wells range from 18 to 70 meters in depth.

Farming takes place primarily on the loamy sands and sandy loams which predominate
this area. The sandy soil conditions, in conjunction with the ample supply of groundwater, make
the area ideal-for crop production while at the same time making it susceptible to groundwater

contamination. As a result, nitrate levels in groundwater have increased during the past few years.

4

In a recent survey conducted in St. Joseph county, water from 2226 wells was tested for nitrate-N
concentration, with 454 or about 20.5 % testing above 10 ppm nitrate-N (King, 1989). With
increased public awareness and concern for environment quality, crop growers must begin to
implement new management strategies that will minimize the impact on the environment.

When developing any management strategy, one must consider the crop type, soil type,
and other components of the management system. The following is a brief description of the

factors influencing the management strategy development.

Management System Components
lm'gation Considerations

Several factors influence irrigation water applications within Southwest Michigan. First there
is the soil, with the potential extractable soil water (PESW) and the extractable soil water (ESW)
being important variables. The PESW is the maximum amount of water the soil can hold that is
available to, or extractable by, the plant. The ESW is the amount of plant extractable water in the
soil at any given time. Farmers normally use the ESW to decide when to start their system,
although this is not always the case and the PESW to determine how much water can be applied
without overfilling the soil profile.

Farmers use a variety of information to help them decide when to irrigate. Some farmers
use tensiometers to estimate the ESW. Tensiometer installation is fairly simple, but they do
require some maintenance and often farmers find them too time consuming. Some experienwd
farmers determine when to irrigate by just feeling the soil or studying the crop. However, with the
recent expansion of computer technology, combined with ever increasing input costs, more farmers
are using computerized irrigation scheduling programs to help them decide when to irrigate and
how much water to apply.

Most irrigation scheduling programs treat the soil like an empty glass. When a glass is
filled it overflows if too much water is put in, thus wasting water. Soil too can hold only a certain

amount of water and overflows if an excess is applied. Some scheduling programs estimate this

5

overflow in terms of drainage, runoff, or surface water retention. Other programs just consider
the excess water lost and unavailable to the plant. As plants begin to use the soil water, the ESW
decreases. Once the ESW drops to a predetermined level, normally 40 - 60 percent of the PESW,
an irrigation application is recommended.

Much work has been done on irrigation scheduling within St. Joseph County. A county-
wide program using computerized irrigation scheduling in conjunction with irrigation system
evaluations has been in operation within the county since 1980 with good success at helping
farmers lower their inputs, conserve energy, and limit leaching. The program was developed by
researchers at Michigan State University and has been refined over the years (Bralts et al., 1983;
Algozin and Bralts, 1986; Shayya et al., 1990, Shayya and Bralts, 1992). Though this program has
served as an excellent resource for scheduling irrigation water, it has never considered the effect
an irrigation management strategy may have on nitrogen management. Many researchers
recognize the importance of managing irrigation and nitrogen together. Watts and Martin (1981)
showed nitrate leaching out of the rootzone was affected by the timing of the irrigation application
and the amount of water applied. Other researchers reached similar conclusions (Watts and
Hanks, 1978; Schepers et al., 1983; Smika, et al., 1977; Mielke et al., 1979). Considering irrigation
and nitrogen management as separate components in the overall management system, each
component affecting the other is imperative. For this reason, irrigation and nitrogen management
must be given equal importance when developing management schemes that will limit potential

leaching.

Crop Considerations

Crops such as soybeans or alfalfa require small amounts of nitrogen as compared to crops
such as corn or potatoes. In this study, we are concerned with seed corn, also called inbred
maize. At the start of a season, inbred maize appears similar to its offspring, hybrid corn. As the

season continues, however, the inbred lines become quite unique.

6
The availability of information on inbred maize growth is limited. Seed companies have

guarded their genetic lines closely, making detailed information unavailable to the general public.
However, there are some basic, well know management strategies in use that require explanation.

F'ust and foremost is the use of male and female plants. Normally, a grower will plant six
or four rows of female (pollen receptor) plants and one or two rows of male (pollen donor) plants
between them. This pattern is repeated throughout the field. The male plants and female plants
are usually of two separate inbred lines and are often planted at different times to assure the male
tassels are shedding pollen at the same time that the female silks are receptive to pollination.
Additionally, young plants may be burned back to delay their development to assure synchronized
pollination. This process is done using propane flamers and driving up and down the rows
burning the tops of the male plants. This bum-back usually delays the plant’s development by
about 28 degree days (10 degrees C, base temperature). A major concern of seed corn growers is
that both plants be at the proper growth stages when pollination takes places. A miscalculation of
a few days can result in poor pollination and thus reduwd yield and profit.

Another operation unique to seed corn management is the detasseling of the female
plants. Before the female plant’s tassels are exposed and shedding pollen, the tops of the plants
are cut or pulled off. This prevents the female plants from shedding their pollen and assures that
only the pollen from the male plants fertilize the silks from the female plants. The result is called
a hybrid. Once pollination is completed, some seed companies require that the male plants be cut
down, while others allow them to continue to grow.

Inbred plants are usually much smaller and yield less than hybrid plants. Once the female
plants have been detasseled and the male plants cut out, one would be hard pressed to identify
what is left in the field as maize. Average yields of irrigated inbred crops in the region range from
as low as 1181 kg ha“1 to 6900 kg ha". Irrigated hybrid crops yield an average of 10,000 kg ha'l.

Farmers have often talked of the low vigor of the inbred plants. The inbreeds are
considered more sensitive to stress from lack of water and/or nitrogen. Wych (1988) concurs with

this, stating that in general, inbreeds have poorer rooting capabilities than hybrids, making them

7

more vulnerable to nutrient and water deficiencies. Consequently, some farmers keep both water
and nitrogen levels high throughout the plant’s growth cycle. Wych, however, goes on to suggest
that due to higher input costs combined with a decrease in commodity prices and an increasing
concern for groundwater contamination, growers should apply nutrients only to maintain necessary

fertility levels.

Soil Considerations

Soil type and its associated characteristics play a major role in any management strategy.
The soils in Southwest Michigan are mostly loamy sands and sandy loams, with variations
throughout the region. These soils have a low PESW and usually require irrigation to meet crop
water needs, especially corn, during short term droughts. The soils also have a low nitrogen
content, thus requiring farmers to apply nitrogen to their crops. Other factors such as soil pH,
bulk density, organic matter content, etc. all impact the soil/nitrogen dynamics and should

therefore be considered when developing a nitrogen management strategy.

Objectives

1. Develop a computer program that adequately simulates inbred maize growth and

development as well as soil-plant interactions for water and nitrogen dynamics.

2. Perform field experiments to validate the simulation model.

3. Utilize the simulation model to evaluate the impact of several nitrogen management

strategies on potential leaching in seed corn production.

LITERATURE REVIEW

Though seed corn is grown in nearly every region of the world, research work done on
inbred growth and development is kept confidential. Competition between seed companies and a
desire to assure that their genetic lines are not duplicated, has kept most of inbred maize research
out of the scientific journals and unavailable to the public. Because of this, related research with
similar crops must be investigated in order to formulate an understanding of seed corn
management.

There has been much work in the area of nitrogen management with commercial hybrid
corn. Researchers have studied the physiological aspects of nitrogen management including plant
uptake of nitrogen, tissue concentrations, and the effects of nitrogen management on yield and
biomass. Furthermore, work, has been done on the environmental impact of nitrogen
management. Research has been conducted studying the movement of nitrogen in the soil and the
potential of nitrate contamination of groundwater resources.

The review of pertinent research will begin with a discussion on the management and
unique field operations that take place in seed corn production. This will be followed by a
discussion of related research performed with hybrid maize. Finally, maize growth simulation

models will be reviewed to assist in the development of the inbred maize growth model.

Seed Corn Management

Growing Seed Corn
Plant densities normally range from 54,000 to 54,000 plants ha'l (22,000 - 26,000 plants
acre“) for inbred female plants, with the male plant population often exceeding that level (Wych,

1988). Planting patterns vary, but are normally 4:1 (four female rows to one male row), 4:2,

9

4:1:4:2, and 6:2. Other patterns such as solid female planting with interplanted males are used,
but are not as common.

Split-date planting is used to assure that the male tassels shed pollen at the proper time
for the female pollination. Planting dates are set by using a combination of days, heat units,
and/or growth stages (Shoultz, 1985). Flaming or 'burn back' is another technique used to retard
plant development to assure good pollination. The plants are physically burned using lighted
propane gas. This burning of the plant causes a short term severe stress that delays plant
development by a day or two. This technique is normally only used on males, since it often leads

to reduwd yields of the burned plants (Fowler, 1967).

Detasseling

After planting, the next major field operation in the seed corn field is the detasseling of
the female plants. This is done prior to silk emergence and pollen shed on the female plants.
The tassels are removed either by hand or by mechanical detasselers. Hand removal is fairly
efficient though often slow. Mechanical detasselers are quicker and more cost effective.
Detasseling must be done with care to prevent unnecessary removal of plant leaf area. Hunter et
a1. (1973) showed that increased leaf removal decreased grain yield. The difference between the
yield when only tassels were pulled as opposed to one, two, or three leaves was 1.5%, 4.9%, and
13.5%. This difference does vary somewhat with inbred and is dependent on the development of
the plant, climatic conditions, operator skill, and other factors. Tests done by Pioneer Hi-Bred
International, have shown little yield differences between hand pulled and mechanically detasseled
fields (Lightner, Personal Communication). Detasseling of a field, whether done mechanically or
by hand, is normally done twice, with 2 days to a week between operations. The second operation
is performed due to uneven stands associated with inbred plants and to assure that all of the

tassels of the female plants are removed.

10
Removal of the Male Plants

Once pollination has occurred, some companies destroy the male plants while others allow
them to mature. Destruction of male plants is done by large "hi-boy“ type machines that have
rotary blades much like the old push type lawn movers. The blades do not simply cut the plants
down, but rather cuts them into small pieces, leaving little possibility for survival. The removal of
male plants serves two purposes. First, it assures that only the grain from the female plants are
harvested. Secondly, it should reduce competition for nutrients and soil water for the female
plants bordering the male plant rows, though there is no scientific evidence to support this theory

(Wych, 1988).

HEM

Harvesting of seed corn occurs just prior to physiological maturity, when grain weight has
reached its maximum. Kernel moisture contents range from 30% to 38% (Knittle and Burris,
1985), with some variations among inbreeds (Carter and Poneleit, 1973). When planning harvest
dates, most companies use the accumulation of heat units and/or black layer formation (Daynard
and Duncan, 1969). Harvesting is planned mainly to avoid damage to the seed by freezing
temperatures, mechanical harvesters (if the seed is too dry), insects, and/or diseases such as car
molds and stalk rots. Whole ears are harvested to minimize seed damage and to allow for
individual ear inspection to assure proper seed quality.

Once the seed is harvested, it is normally taken to a drying depot and proeessed for
storage. Most companies perform tests to assure quality and germination. The seed is

conditioned and put into storage until the next season when it will be sold as hybrid corn seed.

Nitrogen Management and Scheduling
Research into the effect of nitrogen management strategies on the potential for nitrate

leaching has increased as concerns about nitrate contamination of groundwater have grown. As

11

previously mentioned, little work has been done with inbred maize varieties. However, nitrogen
research in commercial hybrid production has been strong.

Much of the work on commercial hybrid corn has been done by predetermining a nitrogen
strategy and then studying its effect on yield, plant growth, leaching, etc. Agricultural research on
nitrates has been separated into two related areas. The first is nitrate within the plant’s rootzone

and plant uptake, and the other is nitrate movement beyond the rootzone and into groundwater.

Nitrate Movement Through the Rootzone and Plant Uptake
To develop a methodology for scheduling nitrogen, the dynamics of nitrogen within the

rootzone must be studied. This includes not only nitrogen movement within the soil, but also
plant uptake.

The driving force in nitrogen fertilizer management should be the plant demand or
uptake. Legg et al. (1979) found that as the nitrogen fertilizer application rate increased, the
amount of fertilizer nitrogen taken up by the plant also increased. However, the percentage of
recovery of the fertilizer nitrogen decreased as the nitrogen fertilizer application rate increased.
This indicates that as nitrogen fertilizer application rates are increased, plant efficienq in using
the nitrogen fertilizer applied in season is decreased. Schepers et al. (1983) concurred with this
conclusion, adding that the N uptake by maize often exceeded the fertilizer N applied. They
concluded that the additional N must have come from some other source, such as mineralized soil
nitrogen, nitrogen in the irrigation water, or residual fertilizer nitrogen in the soil. After further
study, the researchers concluded that the mineralization of soil N was the major contributor of the
additional N source and that from June through August, average daily N mineralized was 2.5 kg
ha'l. They estimated that the maize grown had a maximum daily N demand of 5.0 kg ha'l.
Therefore, even with this relatively high mineralization rate, some addition N would be required to
meet maximum plant demand.

Watts and Martin (1981), found similar results. Using a simulation model they found that

as nitrogen fertilizer amounts were increased, so did the plant’s uptake of fertilizer N. The cost

12

for this increased N uptake was an associated increase in nitrate leaching. This increase in nitrate

leaching is what has caused great concern.

Leaching of Nitrates

The goal of a conservation minded farmer is to use a nitrogen application strategy that
eliminates or at least minimizes nitrate leaching out of the plant rootzone while maintaining
profit. Crop production on certain soils will always result in some nitrate leaching. This is
especially true in coarse texture soils that have a high sand content. Watts and Martin (1981)

used a nitrogen simulation model and predicted that it was impossible to completely eliminate
nitrate leaching and still maintain present day production levels in the central Platte Valley of
Nebraska. In their simulation, Watts and Martin used a fairly droughty soil, a Valentine very fine
sand, which had an available water holding capacity of nine to 10 percent. Using two nitrogen
rates (168 and 253 kg N ha'l), corn production was simulated for this system. In addition,
irrigation applications were simulated ranging from 0.75% to 1.67% replacement of the difference
between evapotranspiration and rainfall every four days. The weather inputs included a below
normal, normal, and above normal rainfall amount. The results of the study indicated that careful
water management did not significantly reduce the loss of water through the soil profile, since
most deep percolation occurred either before any irrigation was required or within days after a
needed irrigation due to unforseen rainfall events. The researchers also stated that early season
percolation could not be avoided on these sandy soils, even though seasonal rainfall was rarely
enough to support crop production. Unfortunately, Watts and Martin simulated only the growing
season (511 - 9/31). Their data show nitrate leaching amounts ranging from 30 to over 100 kg N
ha". Their simulation includes only N uptake data and no yield data. Attempts to limit nitrogen
leaching may result in lower total biomass production, but that does not necessarily mean lower
grain yields. These authors concluded that higher nitrogen rates do not necessarily impact
leaching within the growing season rather, it is the following winter and spring leaching that is

impacted by a season’s nitrogen management.

13
In Minnesota, Timmons and Dylla ( 1979), conducted a similar study with hybrid corn

production comparing irrigation amounts (2.5 and 5.1 cm per application), nitrogen fertilizer types
(granular and liquid), and nitrogen fertilizer applimtion timing (granular, preplant; liquid, four
equal applications throughout the season through the irrigation system). The nitrogen amount
applied was 225 kg N ha“1 and lysimeters where used to collect leachate at a depth of 122 cm.
The results showed the 2.5 cm irrigation application and the 5.1 cm irrigation application
increased nitrate leaching over the non-irrigated control by an average of 18% (70 kg N ha’1 to 82
kg N ha") and 55% (70 kg N her1 to 108 kg N ha“). Again, these are within season leaching
amounts only. The authors also concurred with Watts and Martin (1981) as to when the
maximum percolation occurs. They cite an example in 1975 where 10.2 cm of percolation
occurred before irrigation water was applied to the fertilizer treatments. This percolation
accounted for 70% of the total within season nitrate leaching 1055. Their study also included yield
data which showed no significant yield difference between the irrigation amounts or fertilizer
treatments, with the exception of the non-irrigated and non-fertilized treatments.

One element which causes much confusion in nitrate leaching research is the time span of
the data collection or simulation. Seasonal leaching amounts may be minimal compared to those
that occur between harvest and the next years planting, as previously mentioned by Watts and
Martin (1981). Heavy autumn rain or deep winter snow followed by a spring thaw may cause
exeessive leaching in some regions. Schepers et al. (1983) found that mineralization of N was 312
1; 66 kg ha'1 during the 1980 growing season and 39.8 i 9.0 kg ha'1 between the fall of 1980 and
the spring of 1981. This mineralized N estimation was made from soil‘sampling and takes into
account plant uptake. The study also shows how corn plants can utilize mineralized N for
production and how farmers cm reduce nitrogen fertilization applications. The result was an
average decrease of 94 kg ha'1 of nitrogen fertilizer applied over an area of 3,000 ha. Harvest tests
showed no significant reduction of grain yield. However, the within season mineralized N poses a
threat for off season leaching. This is especially true if plants do not efficiently use mineralized N.

Schepers and Martin ( 1983) simulated the leaching of nitrates in loamy sand soils in the sandhills

14

of Nebraska. They estimated that 20% - 80% of the residual N in the soil on April 1 was leached
out of the rootzone by June 15. This simulation was done over an 11-year period from 1973 to
1984. For fine sand soils, the percent leached increased to 60% - 100% .

The potential for leaching is a factor that often causes farmers to over apply nitrogen.
Schepers (1988) stated that because of the fear of wet spring weather, many farmers consider fall
and early spring applimtions of nitrogen. Even with preplant applications, many farmers over
apply their nitrogen for fear that late spring and early summer rains will cause excessive nitrate
leaching. This is especially true in the subahumid and humid regions. Schepers notes that even
though strong evidence exists that N recovery for corn is highest with sidedress applimtions
(Anderson et al., 1982; Russelle et al., 1983), many farmers are fearful of wet weather causing

delays in the timely applications of nitrogen.

Irrigation Management and Its Effect on Nitrogen Management

Because nitrate is a water soluble form of nitrogen, irrigation can have a signifimnt
impact on nitrogen fertilizer management. Too much irrigation water may muse excessive deep
drainage, which in turn can mrry nitrates out of the rootzone and down toward the groundwater.
On the other hand, too little irrigation water can limit a plants’s uptake of nitrogen musing water
and nitrogen stress.

Much work has been done in evaluating irrigation strategies in respect to nitrogen
management. Russelle et al. (1981) evaluated the effects of nitrogen and water management on
maize yield by varying irrigation amount and applimtion interval. Nitrogen uptake in the grain
lowered with the heavier, less frequent irrigations (10 cm per applimtion in 2 week intervals).
Also, the recovery of fertilizer N from the soil was affected by the irrigation management,
especially on the low sidedress applimtions of nitrogen (112 kg N ha’l). As the irrigation water
amount applied decreased and the applimtion frequency increased, the fertilizer N recovered by

the plant increased. This shows the importance of irrigation management especially at low or

15

marginal nitrogen applimtion rates. The lower and less frequent irrigations left more residual N
in the upper soil profile and increased fertilizer use efficiencies.

Martin et al. (1982) found that irrigation management signifimntly impacted the nitrogen
uptake by maize. The researchers used various nitrate-N concentrations in the irrigation water
and a field mlibrated simulation model to study the interaction of nitrogen and irrigation water in
the Platte Valley region of Nebraska. They concluded that N uptake by maize was signifimntly
influenced by the amount of irrigation water applied, and to a lesser extent by the nitrate-N
concentration of that water. In addition, the plant’s uptake efficiency of the fertilizer was also
sensitive to irrigation water amounts applied.

The authors simulated for 4 nitrogen fertilizer applimtion amounts, 0, 45, 135, and 225 kg
ha'l. The nitrogen applimtions were simulated with 45 kg N ha‘1 applied at planting and the
remainder of the nitrogen applied in a single side-dress applimtion one month later. The
irrigation amounts were simulated for "Irrigation Replacement Fractions (IRF')" ranging from 1 to
4. An IRF value of 1 represented an irrigation water applimtion equal to the difference between
evapotranspiration and the effective rainfall since the last irrigation (effectively refilling the soil
profile). All irrigation applimtions were simulated on a 4 to 5 day frequency. The results of the
simulation showed that for irrigation water with a nitrate-N concentration of 10 and 25 ppm, the
fertilizer N uptake efficiency for IRF equal to 1 was always greater than 0.5, regardless of the
nitrate-N concentration in the irrigation water or the nitrogen fertilizer applied. However, for
IRF equal to 3 or 4, the fertilizer N uptake efficiency was never greater than 0.25, again without

regard to nitrate-N concentrations or nitrogen fertilizer applied.

Maize Simulation Growth Models

Crop growth simulation models have been gaining acceptance as a viable research tool.
These models usually integrate weather, soil, and management variables to simulate the growth
and development of a particular crop. Crops that have been simulated include maize, soybeans,

alfalfa, and cotton, to name a few. A partial list of published crop simulation models is in Table l

16

(Jones and Ritchie, 1991). A simulation model for inbred maize does not exist at this time.
However, hybrid maize simulation models that have gained acceptance could be adapted to inbred
maize.

Though inbred maize does differ from hybrid maize, the basic structure and development
of both crop types are the same. In 1980, Stapper and Arkin published a maize simulation model
mlled CORNF (Stapper and Arkin, 1980). This model used a daily time step and simulated maize
growth and development. The model also included a soil water component that allowed for
estimations of ET, soil water content, drainage and runoff. This model does not include nitrogen
subroutines that would allow for soil nitrogen content estimation and/or nitrate leaching. Also,
the CORNF model does not simulate soil water below the wilting point or above field mpacity.
This could muse some problems during water stress periods or periods of excessive rainfall.

The CERES-Maize model is a daily time step model that simulates plant growth and
development as well as soil water and soil nitrogen interactions. The first version was published
in 1986 (Jones and Kiniry, 1986) and version 2.10 was released in 1989 (Ritchie et al., 1989).
CERES-Maize is just one of a family of growth simulations models that are being developed
through the USAID sponsored International Benchmark Sites Network for Agrotechnology
Transfer (IBSNAT) program (Uehara, 1985). Other IBSNAT models include soybeans -
SOYGRO (Wilkerson, 1983), peanuts - PNUTGRO (Boote et al., 1989), and wheat - CERES-
Wheat (Ritchie and Otter, 1985).

Aside from the usual validation needed for model development, CERES-Maize has
received much testing and use by researchers outside of the original development group. Hodges et
al. (1987) illustrated the use of CERES-Maize in predicting regional yield estimations for the US
Cornbelt. In this study, for the years 1982 (mlibration year), 1983, 1984, and 1985, CERES-Maize
production estimates were 92%, 97%, 98%, and 101%, of the figures reported by USDA in
January, 1987. Piper and Weiss (1990) evaluated CERES-Maize for reduction in plant population
and leaf area during the growing season. Their results showed that CERES-Maize gave good

results except that it had difficulty predicting yields at high and low plant populations, though it

17

Table 1. Partial list of published crop growth simulation models taken from Jones and Ritchie,
1991.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

  

Crop Model Name Reference
Alfalfa ALSIM 1 (Level 2) Fick (1981)
Barley CERES-Barley Ritchie et al (1989)
Cotton GOSSYM Baker et al. (1983)
Dry Bean BEANGRO Hoogenboom et al. (1989)
Maize CERESoMaize V1.0 Jones and Kiniry (1986)
CERES-Maize V2.1 Ritchie et. al (1989)
CORNF Stapper and Arkin (1985)
VT-MAIZE Newkirk et al. (1989)
Peanut PNUTGRO Boote et al. (1989)
Pearl Millet CERES-Millet Ritchie and Alagarswamy (1989)
Potato SUBSTOR Hodges et al. (1989)
Rice CERES-Rice Godwin et al. (1990) H
Sorghum SORGF Arkin et al. (1976) ll
CERES-Sorghum Ritchie and Alagarswamy (1989) l
Soybean ' SOYGRO V5.00 Wilkerson et al. (1983) I
SOYGRO V5.42 Jones et al. (1989)
Wheat CERES-Wheat V1.0 Ritchie and Otter (1985)
CERES-Wheat V1.0 (Nitrogen) Godwin and Vlek (1985)
CERES-Wheat v2.1 Godwin et al. (1989)
TAMW Maas and Arkin (1980)

 

 

 

  

18
did follow yield trends at the lower plant densities. CEREscMaize has also been used in the semi-

arid tropim (Carberry et al., 1989) and as part of a newly developed light interception model
(Hodges and Evans, 1990).

Work has also been done using CERES-Maize to aid in the economic analysis of
irrigation strategies. Algozin et al. (1988) used the model to generate yield and irrigation water
use information to perform an irrigation budget analysis for several irrigation strategies used in
Michigan. Boggess and Ritchie (1988) used a similar approach for analyzing different irrigation
regimes with respect to economies and risk aversion in humid regions. Their conclusion was that
maximum profit is obtained when less irrigation water is used than needed to produce maximum
yield. In other words, the extra cost associated with irrigating the crop to obtain maximum yield
was greater than the increased revenue from the extra yield. Therefore, to obtain the highest
amount of profit, maximum yields were not always obtained. Other works involving CERES-
Maize with irrigation analysis include Martin et al. (1985) and Worman et al. (1988).

In one performance test, de Vos and Mallett (1987) compared CORNF and CERES-
Maize. Though the authors did not find one model superior over the other, they did indimte that
CERESoMaize predicted soil water patterns better than CORNF, although the inputs for CORNF
were simpler than for CERES-Maize. In the final analysis, the authors concluded that, 'CEREso
Maize provided more realistic simulations, although its overall performance was not substantially
better than that of CORNF.“

A third maize model, VT-Maize (Newkirk et al., 1989) is a combination of separate
published models. The developers took what they considered to be the best parts of several
different simulations models and combined them into one maize model. The phenologic and
physiologic development used was based on the CERES-Maize nitrogen version (Jones and Kiniry,
1986). The soil subsystem processes and interactions were based on the RHIZOS portion of the
GOSSYM model (Baker, et al., 1983), while soil temperature is updated using the approach found
in the EPIC model (Williams and Renard, 1985). Inputs for VT-Maize are similar to CERES-

Maize for the phenologiml inputs. However, the soil inputs are much more detailed and difficult

19

to obtain. Soil water and nitrogen distribution and interaction are based on a two-dimensional
model, that is defined by a grid or cell matrix.

All three models have strong and weak points. The CORNF model does not contain the
necessary elements for evaluating nitrogen management on yield and leaching. VT-Maize uses the
same basic growth subroutines as CERES-Maize, but the soil water and soil nitrogen data are
more detailed and whether this increased complexity results in a more precise prediction and/or
simulation is not yet clear bemuse the developers did not provide evidence though independent

validation.

METHODOLOGY

OBJECTIVE 1: Development of an Inbred Maize Simulation Model

Though the production of inbred maize is different from that of hybrid maize, the growth
and phenologiml development of both plants are similar. For this reason, it was decided that a
presently used hybrid simulation model would be adapted for inbred growth. The model chosen
was CERES-Maize bemuse of its relatively wide use within the scientific community (Algozin et
al., 1988; Boggess and Ritchie, 1988; Carberry et al. 1989, Hodges and Evans, 1990; Hodges, et al.,
1987; Martin et al., 1985; Worman et al., 1988; and de Vos and Mallett, 1987) and the ease with
which inputs are obtained. In addition, the CERES-Maize model has nitrogen components that
will aid in developing of nitrogen fertilizer management schemes. The first step to reprogramming
CERES-Maize for inbred maize is to mlibrate/validate the model for hybrid maize grown within

the region of interest.

Commercial Hflrid Maize Study
Though a validation of the CERES-Maize program was presented in Jones and Kiniry

(1986), there was no validation on the program’s ability to simulate soil nitrogen leaching. Since
the leaching of nitrates is integral to the formulation of any nitrogen management scheme, it
follows that this portion of the model requires further validation than that done by Jones and

In the summer of 1986, two drainage lysimeters were placed in a farmer’s field near
Mendon, Michigan. Both lysimeters were 'disturbed' profile lysimeters. Installation began with an

analysis of the soil and defining specific soil layers. Next, the soil was removed, layer by layer, and

20

21

placed into individual piles. Then, once the lysimeter was lowered into the ground, the soil was
placed back into the lysimeter, again layer by layer, and packed so as to best represent the soil in
its original condition. The first lysimeter was placed under an existing center pivot irrigation
system that the farmer used to irrigate hybrid maize. The other lysimeter was plamd outside the
arm of the center pivot but within the same field as the first. The tamer controlled the water and
nitrogen management of the lysimeter under the existing center pivot irrigation system (the
conventional management practice - CMP) while researchers at MSU controlled the water and
I nitrogen management of the other (the research management practice - RMP). The data collected
included soil water and soil nitrogen data as well as information on plant growth, phasic
development, and yield. Also, the drainage volume was measured and samples were collected to
analyze the drainage for nitrate concentration.

The lysimeters used in this experiment were 1.52 m wide, 1.22 m long, and 1.83 m deep
(see Figures 2, 3, and 4). The corn was planted parallel with the 1.52 m width to assure that two
rows of maize (at 76.2 cm spacings) would be planted over the lysimeters. The lysimeters were
placed approximately 45.7 cm below the ground surface so that normal tillage operations could
take place. The farmer was responsible for all tillage, planting, and pesticide applimtions for both
plots. Both plots were treated exactly the same except for the irrigation water and nitrogen
fertilizer applimtions (excluding the nitrogen in the starter fertilizer that was the same for both

plots).

Drainage Sampling

Leachate sampling began in 1988. The sampling took place approximately every two
weeks, or when the drainage volume warranted. The drainage was collected in polyethylene
containers lomted under the lysimeter drainage pipe in the collection area. The drainage samples
were analyzed for nitrate concentration. In 1989, a more sophistimted collection device was

installed that allowed for the automatic collection of drainage samples. Using a dataloggel’. the

22

42 £852 £58.: 5 2a neon—238E =< £26305 use...
HEP? Ethan—E8 E 33... EoEowmeaE eowebm: 05 5 new: £20.55 05 no 25 Co 32> one 23 32> no. 2: Co 9:38.. c.3805... < .N BEE

3WH> mom

3:33 .32
.5 sea
Eases...
.I.._ I._ A

 

3mH> 22m

  

 

 

 

 

 

 

 

 

 

 

 

 

 

_._HL\ .
HlllllllUl'atlillll xxxx x x x
gulf—ll. x x xxx xx
\\\\\\\\\\\\\\x
xxxxxxxxxxxxxxxlr
xxxxxxxxxxxxxx .
xxxxxxxxx\\\x\ua
x xxxxx \
x xxx xxxxx
\\\\\\\\\\\
xxx x xxx
9:3 I. xxxxxxxxxxxxx.
. \xxxxxxxxx xxxxxxxxxu
\ \ \ \
32' \‘ a \ xxxxxxxxxxxx xxx xxx xxxux xxx.
I. x x xxxx x x
Eva? — \ xxxhhxkxxxxkxxxtxkx
F—D. r|lrmegl|L

 

 

 

 

I
i.
I
L

 

 

 

42 £852 £22: 5 o..." 8522:...on =< dozen—6.5
SEE ESE 365:.an 5 3:3. EquwmemE smog: 2: 5 com: cos—=3». on. Co 28 Co 33> 02m 95 Co 3.36:. 2.25:3. < .m 033..."

 

season. as:

_i| N3 Ili— EE .5

\l Baa see 555
mcfifinaum I .I4IIII

 

 

7

A.

 

 

 

 

 

 

usage .2 gas
name
1 gas... usage
NS
8.8 \

533:8 as: m. 2 sea. :8
l/ a: €535
. as as? a...

_

 

 

 

 

 

 

I4“
I
I

888 8.3
533:3 as;

 

 

 

24

£2.52 deco—.8... 83:. 2.5.3 .wUCoEEoo E 3.5... EoEowEmE Swazi o... E 99.: EoBEEb 2: Co 25 Co 332—. 33805... < .v oSuE

42

 

amass as“;

egg .5
33.. as... sea

       

x x
x

  

    

 

\
\ x
\ \

III.

I.

.I I'll"
I

    

_ x
*
.

8.3

538:8 cases

 

      

—
.
_
r
.
—
_
—
—
_
—
_
.

.
‘VN‘N‘N‘
xxxxxxxxxxxxx
xxxxxxxxxxvxx

$888.!

grass .95;

 

 

:8 5.. 3:: has:

 

 

25

volume of drainage was measured and a sample was taken whenever a 6 mm depth of water had

drained.

Research Plot Irrigation System

The research plot, lomted at the end of the field, required its own irrigation system.
Figure 5 shows a diagram of the irrigation system layout. A system was setup with four sprinklers
at 9.14 m spacings, with the lysimeter in the center. The system employed the use of a Rainbird
irrigation timer-controller that was connected to a solenoid valve lomted in the main line of the
irrigation system. The controller was a 14-day timer, designed so that the user could chose the
time to irrigate, the length of time the system was on and the day the system would turn on. The
system was also equipped with an automatic device that shut the system down once a
predetermined amount of rain had fallen. The rainfall was collected in a small mlibrated bucket
that was lomted on top of a spring activated switch connected to the controller. System
reactivation occurred when the water in the bucket evaporated. To give more flexibility for
controlling the irrigation events, the timer was rewired. First, both a 'start“ and a 'stop' button
were installed to turn the system on and off manually. Also, the system was altered so that if a
sufficient rainfall amount had occurred to shut down the controller operation, the shut down
would be irreversible, until the start button was pressed again. This avoided any over watering
that might occur since the evaporation out of the bucket may not represent actual water loss from
the research plot and could muse an irrigation event to occur before it was required.

Using the data gathered in 1988 and 1989, the CERES-Maize program was compared
against field data. Comparisons are made with biomass and grain measurements in addition to

nitrate leaching data recorded.

Qe_velopment of CERES-1M
The changes required to simulate inbred growth dealt mainly with production events such

as the detasseling of the female plants and removal of the male plants, and how these operations

26

 

 

 

9.14

 

9. l4 So‘ilitlars

 

 

 

e a

Spriritler pattern

 

 

 

Figure 5. A schematic drawing of the irrigation system used on the RMP plot. The curved lines
represent the radius of coverage of the four sprinklers. All measurements are in meters. Mendon,
MI.

27
effect plant growth and development. The new model mlled CERES-1M (Inbred Maize) simulates

inbred maize growth and development and allows for the necessary inputs relative to inbred
production (male and female plants, detasseling, etc). Additionally, changes were made in the
model in areas where shortcomings have been observed in past uses of CERES-Maize. Before

discussing the changes made, a short summary of the present CERES-Maize structure is required.

Present CERES-Maize Input Structure
CERES-Maize presently has 9 input files (1 unused at the present time) in which all data
required for a simulation can be entered. These files include information on soil water/nitrogen
data, daily weather data, fertilization and irrigation data, management information (i.e., plant
population, planting date, variety sown, etc.) and genetic parameter inputs. A description of these
files can be found IBSNAT Techniml Report 5, "Documentation for IBSNAT Crop Model Input
& Output Files, Version 1.1: for the Decision Support System for Agrotechnology Transfer
(DSSAT V.2.l),' (1990). The files are defined as follows:
File 1. Weather data - Daily solar radiation, max/min temperature, and rainfall.
File 2. Soil Data - Drained upper limit, lower limit, soil albedo, etc.
File 3. Unused at the present time
File 4. Soil Nitrogen Balance Parameters - Amount of the organic residue of previous crop,
depth of incorporation, C:N ratio of residue, and dry weight of root residue of previous
crop.

File 5. Soil Profile Initial Conditions ~ Water content, soil ammonium, soil nitrate, and soil pH,
all by layer.

File 6. Irrigation Management Data - Date of irrigation and amount of water applied.

File 7. Fertilizer Management Data - Date of applimtion, amount applied (actual N), depth of
applimtion, and type of fertilizer used.

File 8. Treatment Management Data - Sowing date, sowing depth, irrigation management (no
irrigation, according to input (File 6), automatic irrigation, or water assumed non-
limiting), nitrogen management (according to input (File 7) or nitrogen assumed non-
limiting), irrigation system efficiency, depth of soil considered if irrigation is automatic,
available soil moisture trigger used to automatimlly irrigate, phyllochron interval, and
multi-year switch (indimtes number of years of simulation).

28
File 9. Genetic Coefficient Data - Degree days from emergence to end of juvenile stage and from
emergence to physiologiml maturity, photoperiod sensitivity, potential kernels per car, and
potential kernel growth rate.

In addition to these 9 input files, CERES-Maize also requires two other input files. One
is named I'WT'I-1.DIR" and contains the names of the weather files that can be used in the
simulation run. Any number of weather files can be listed. The other file is named
'MZEXPDIR' and contains the name of the eight input files previously listed plus the names of
five output files (a description of the output files will be given later). The files are listed by
experiment. Since an experiment can have several treatments, the input files themselves can be
separated by treatment. Thus, the "MZEXPDIR' file may contain 4 experiments, each having
their own individual input and output file names. Within each of the 8 input files, experiments

can be further separated into treatments. A more detailed description of the individual inputs

within these files can be found in Jones and Kiniry (1986).

Restructuring of Input to Accommodate Male and Female Plants

Since male and female plants play different roles in seed corn production, there needed to
be a method for identifying each within the input structure of CERES-1M. To keep changes to
the input structure within the presently used CERES-Maize model, additional variable inputs were
added to misting input files rather than completely restructuring the input format. Since the role
of male or female is not necessarily dependent on genetic type, each treatment must be allowed to
be either designated male or female. Therefore, the new parameters were placed within the
treatment management input file (File 8).

The first additional input is ISEX. ISEX represents the role the treatment plays in the
production of the seed corn. If ISEX equals 1, it is assumed to be a male plant. If ISEX equals 2,
it is assumed to be a female plant. If there is no input for ISEX, the simulation is assumed to be
for hybrid maize, and all inbred maize related changes are ignored.

The other two new inputs are IDET and IMCUT. IDET is the day-of-year (DOY) that

detasseling occurred. IDET is used only if ISEX is equal to 2 (female). Even if a value exists for

29
IDET, detasseling is not simulated unless ISEX equals 2. If IDET equals 0, then detasseling is not

simulated. Finally, if IDET is set equal to 999, then detasseling is automatimlly simulated within
the program structure. A new subroutine, DETASS, was written to simulate the detasseling
operation. A listing of the source code of the DETASS subroutine is in Appendix one.

Once detasseling has been simulated, a reduction in both leaf area and leaf weight is
required. According to tassel samples taken in 1990, an average of 3 leaves per plant were
removed during the detasseling operation. The average leaf weight loss compared to total plant
biomass was 13.1 % (Table 2). However, the biomass samples taken for inbred 1 were taken 1
week after detasseling while the biomass sample taken for inbreeds 2 and 3 were taken on the
same day of detasseling. This is reflected in the lower percentage of percent biomass loss for
inbred 1 shown in Table 2. Therefore, averaging inbreeds 2 and 3, the average percent biomass
loss due to detasseling is 14.05%. Therefore, once detasseling is simulated, leaf biomass is reduced
by 14 % of the total plant biomass and leaf number is reduwd by 3.

For automatic detasseling, tassel removal is assumed to take place just prior to silking,
which is the norm for field situations. Early detasseling can muse excessive loss of leaves and lead
to yield reduction while late detasseling can muse contamination of the cross breeding (Wych,
1988). Therefore, automatic detasseling is assumed to occur when 90% of its thermal time ('1'?)
requirement to complete CERES-Maize phenologiml growth stage 3 (tassel initiation to end of
leaf growth and silking). Though this may not always be the mse, it does provide for the
detasseling operation to take place before any pollination contamination would take place but not
too early to muse excessive leaf loss.

The subroutine DETASS is mlled whenever the DOY is equal to the detasseling date
(IDET) that was entered in File 8 or, if IDET is equal to 999, when the TT‘ reaches 90% of the
total TI‘ required to complete stage 3. Once detasseling is simulated, a message is printed on the
summary output file and to the screen, informing the user that on that day female detasseling was

simulated.

30

Table 2. Average percent biomass loss due to female detasseling for all inbreeds. Constantine,
MI, 1990.

 

 

 

 

 

 

Variety Detasseled Total Biomass' % Biomass I
Biomass (kg ha") Loss I

I (k8 113-1) I
Inbred 1 413.3 3690.2 11.2 ;

l

, Inbred 2 443.9 33127 13.4 ;
Inbred 3 515.3 3505.4 14.7 f

" Biomass samples for inbred 1 were taken 1 week after detasseling. Biomass samples for
inbreeds 2 and 3 were taken on the day of detasseling. The total biomass samples include the
weight of the detasseled biomass.

The other new input in the treatment management file is IMCUT, the DOY the male
plants are cut down. As with the females plants and the input IDET, IMCUT is ignored,
regardless of its value if ISEX is not equal to 1 (male). However, if ISEX is equal to 1, then
IMCUT is used to determine when the male plants are to be cut down. As with the female
detasseling operation, if IMCUT is set equal to 999, automatic male plant removal is simulated. If
IMCUT is equal to 0, the male plants are not removed. However, a value of 0 for IMCUT is
invalid when simulating an entire seed corn field with male and female interactions. A new
subroutine mlled MCUT was written to handle the simulation of the cutting out of the male
plants. A listing of the FORTRAN code for the subroutine IMCUT is in Appendix one.

If ISEX is equal to 1 (indimting male plants), CERES-1M then determines if the removal
of the male plants is to be simulated. If IMCUT equals 0, then the removal is not simulated and
plant growth is simulated to physiologiml maturity. If IMCUT does not equal 0 (indimting that
removal is to be simulated), then CERES-1M determines if the removal date is to mlculated
internally (IMCUT equal to 999) or if the removal is to simulated on the DOY entered (IMCUT
equal to 1366). If the removal is to be simulated on the day entered in the input file, CERES-1M
checks to determine whether the DOY equals IMCUT. If it does, a message is sent to the

summary output file and the screen informing the user that the removal of the males plants has

31

been simulated and that the growth routines have been terminated. This termination of growth is
accomplished by setting the plant growth stage to 6, physiologiml maturity. This terminates the
growth portions of the model but allows the soil water and soil nitrogen routines to continue
processing.

If IMCUT equals 999, the operation is simulated automatimlly. Male plants are removed
mainly to protect the germplasm. Many companies feel that if the male plants are allowed to
begin grain filling, their germplasm could be duplimted by other rival companies. Therefore,
though the early removal of male plants may muse reduced yield due to poor pollination, late
removal may muse anxiety for the seed companies. In the automatic removal of the male plants,
it is assumed that the male plants are cut out when 35% of the required TT to complete grain
filling has been reached. This may leave the males in the field too long for some growers, but is a
good approximation and guarantees no loss of yield.

Another area of concern for inbred growth is the burning back of plants to delay
development in order to extend the pollination period. This is done only on male plants and is
rarely done on the entire male population. At most, only 50% of the male plants are burned
back. Bemuse of this, no reprogramming was done to account for this operation.

The male and female plants may vary in genetic type, plant population, and developmental
stage as well as having separate field operations performed on them (ie, detasseling and removal).
The interactions of these plants and how these interactions effect growth and soil water and soil
nitrogen conditions must be accounted for. One consideration is the planting pattern used.
Though most seed corn fields are planted in a pattern of four female rows per each male row
(4:1), other patterns such as 4:2, 4:1:4:2, and 6:2 are used. In the new CERES-1M model, the ratio
of female rows to male rows is a management input and is placed in the treatment management
file, File 8. The input, mlled PRAT‘IO (Planting RATIO) is a single real number that relates the
ratio of female rows to male rows. Patterns such as 4:1 and 4:2 are entered as 4 or 2. Other

patterns, such as the 4:1:4:2 are not allowed.

32

During the season, there is no interaction between the male and female plants. To run a
simulation for a seed corn field, the user enters the treatment used for the female and male plants,
making sure that all of the appropriate inputs relative to male and female plants are entered (i.e.,
ISEX, IDET, IMCUT, etc.). The model then makes two separate runs; one run for the female
plants and one for the male plants. The male plant’s simulation is run first. Using the inputs
given, the simulation runs until the male plants are cut down. Once the removal of the male
plants has been simulated, the model continues to run the soil water and soil nitrogen simulations
for an additional 60 days. During this time, the daily soil water and soil nitrogen values, all by soil
layer, are stored in a temporary file, to be used later at the end of the female plant’s simulation.

After the male plant’s simulation is completed, CERES-1M begins to simulate the female
plants. Once the female plants have reached full maturity, the temporary file created by the male
plant’s run is read. When the DOY of the female plant’s maturity equals the DOY in the male
output file, the soil water and soil nitrogen contents of the two runs are combined. The
combining of these data is accomplished by using PRATTO and weighing the appropriate values
for soil water and soil nitrogen to produce combined values. These new soil water and nitrogen

values are used for the remainder of the female simulation.

Present Structure of CERES-Maize Output Files

There are five output files in CERES-Maize. These files contain data on the growth and
phenologiml development of the plant, including biomass accumulation, grain filling, and nitrogen
uptake and concentration. Soil water and soil nitrogen data are also given. There is one output
file that gives a year-end summary of the simulation, three component files that give data
throughout the simulation, and a fifth file that is a short summary file used only during multi-year
runs. All of these files are discussed in Techniml Report 5, 'Documentation for IBSNAT Crop
Model Input and Output Files, Version 1.1", IBSNAT (1990). The following is a brief discussion

of each file. An example of each of these output files is in Appendix two.

33

The first output file is the summary output. This file contains a summary of the input
data and output data. Input data such as soil parameters, genetic parameters, experiment and
treatment identifiers, and fertilizer and irrigation applimtions are all echoed in this file.
Additionally, phenologiml data, such as end of juvenile stage and beginning of grain filling are
given, with dates and TT, precipitation, and nitrogen uptake. Finally, the summary output file
gives a comparison between observed and predicted data. This includes silking date and maturity
date, kernel and yield data, biomass data, and nitrogen uptake and concentration data.

The component output data are separated into three files; the biomass, the soil water, and
the nitrogen output files. Each of these files have the option to print output on a daily basis or at
any daily frequency up to 99 days.

The biomass output file contains information on biomass accumulation, leaf number, LAI,
root-stem-grain-leaf weights, rooting depth and root length volume for soil layers 1, 3 and 5. The
default for writing to this output file is once every 7 days. However, the output interval can be
changed.

The soil water output file contains information on the status of the soil water and the
parameters that influence it. The data given include potential and actual evapotranspiration, plant
evaporation, the soil water contents of soil layers 1 through 5, the profile soil water content, and
maximum and minimum temperature, solar radiation, and precipitation.

The third component output file contains information on plant and soil nitrogen status.
The data given include vegetative and grain nitrogen uptake, percent nitrogen in the above ground
biomass, and nitrate concentrations for soil layers 1 through 5 and ammonium concentrations for
soil layers 1 through 3.

The fifth output file is used only during multi-year runs. This output file contains
summary data on nitrogen and water usage, biomass, yield, cumulative evapotranspiration, etc.
This file is important for multi-ymr runs, since the other output files can become quite

cumbersome and difficult to decipher when there are several years of data.

34

Together, the output files given by CEREscMaize provide users with opportunities to
check predicted versus observed values for a host of plant and soil parameters. By changing the
timing of the output, daily values for soil water, soil nitrogen, and plant biomass accumulation can

be given, allowing for any number of comparisons throughout the simulation.

Restructuring of Output Files

Unlike the input file structure, the output file structure requires very few changes to
accommodate inbred maize simulation. Some minor changes were made to the summary output
file to allow for the new inputs such as ISEX and PRAT‘IO to be printed out. Messages of when
the female plants were detasseled and when the male plants were removed were also added.
Additionally, when the male and female soil water and soil nitrogen parameters are combined at
the end of the female growing season, a message to alert the user that this has taken place is also
written to the summary output file. However, water drainage and nitrate leaching are not given
within the present output file structure and are required to effectively evaluate nitrogen
management strategies on potential groundwater contamination.

Though both water drainage and nitrate leaching are simulated within the model, these
data are not written to any output file. A new output file, 'OUT'LCI-LDAT‘ was created to give
data on water drainage and nitrate leaching. This file uses the same output step used by the soil
water output file. Therefore, if the user defines the frequency of output for the soil water file as
once every 10 days, the drainage and leaching data are also written once every 10 days. These data

can then be compared against observed data obtained through the use of the lysimeters.

New Leaf Area Relationship

In addition to the changes required to simulate inbred growth, changes were also made to
some of the basic relationships in CERES-Maize. An area that has yielded inconsistencies
between predicted and observed data is leaf area. This is especially true for the smaller leaves at

the beginning of the season when CERES-Maize over-predicts leaf area (Ritchie, personal

35

communimtion). This leaf area over prediction mn muse errors in biomass accumulation and in
evaporation from the plants and soil surface. A new aspect of the CERES-1M model is a new set
of functions to describe leaf arm growth.

Currently, CERES-Maize uses a series of discontinuous functions to simulate plant leaf
area. These functions are separated based on leaf number and were fitted for the plant leaf area
for leaves 1-3, 4-11, 12 to total leaf number (TLNO) minus three, and from TLNO minus three to
the flag leaf, with leaf number as the independent variable. Muchow and Carberry (1989)
described the use of functions based on leaf number that would describe the area of individual
leaves, not total plant leaf area as in CERES-Maize. Using the leaf number break points of
CEREsaMaize, they developed new functions to describe the area of a leaf based on the leaf
number. They also discussed the use of a single exponential function. Using the leaf number,
area of the largest leaf, and leaf number of the largest leaf, they found the following function to

best fit the data, with an R2 value of 0.98:

Y - YOEXP(—0.0344(X — X0)2 + 0.000731(X — 2(0)?) [El]

where Y is the area of the leaf (cmz), Yo is the area of the largest leaf (cmz), X is the leaf
number, and X0 is the number of the largest leaf. Though this function yielded a high R2 value, it
still had difficulty accurately describing the young, small leaves.

A function that does provide for small incremental increases at low values is the
Gompertz function. Described by Richards (1959, 1969), the Gompertz function produces an

asymmetric sigmoid curve that gives the small leaf area values when the plants are young. The

Gompertz function is described by:

A - A e'b°"" [E2]

where A represents the area of leaft (cmz), A0 is the area of the largest leaf (cmz), and b and k
are constants that regulate the spread of the curve and its position along the leaf number axis.

Though the Gompertz function has been used primarily in animal and population studies
(Richards, 1969), it has begun to be used more in growth relationships of higher plants. Baker et
al. (1975) applied the Gompertz equation to maize leaf area development and evaluated the
influence of environmental conditions on the parameters of the function. More recently, Ritchie
and Johnson (1990) proposed the use of the Gompertz function to predict leaf arm index (LAI)
for a variety of plants. Here, the Gompertz function is used to describe the leaf area of individual
leaves, from the first leaf to the largest leaf.

To define the constants b and k, the equation must first be linearized. Substituting leaf

number (LN) for the parameter t (equation [132]), we get:

logc (logc(A/Ao) - logcb - k(LN) [E3]

When logc(logc(A/Ao)) is plotted against leaf number (LN), the resulting linear regression gives a
line with a slope k and a Y-intercept 10ng that gives the constant b when exponentiated.

To determine these values, leaf area data gathered in Michigan was used. The data
consisted of five hybrids grown on the MSU farms (Muchena and Ritchie, 1989) and data from
two hybrids grown in St. Joseph County, Michigan, in 1986 and 1987. The hybrids used from the
Muchena and Ritchie study included A632 x W117 (A632), B73 x M017 (B73), 883 z 045 (2883),
882 z 105 (Z883) and X304C (X304). The two hybrids from St. Joseph County will be referred to
as M186 and M187. The first step was to transform the leaf area data into its linear form by first

taking the inverse of the areas, multiplying them by the area of the largest leaf, and then taking

37

the loge twice. Once this was accomplished, leaves greater than or equal to the largest leaf were
deleted from the data set. Any extremely small values (less then 0.1) were also removed. This
normally deleted the leaf just before the largest leaf. This was done to assure that these extremely
small numbers did not dominate the regression fit, which is often the tendency when either inverse
or loge transformations are involved. The result is the regression analyses found in Figure 6.
Though an R2 of 0.936 was achieved with the regression, Figure 6 does show some smttering of
the data. The next step was to formulate the actual Gompertz function.

Since the Gompertz is an asymmetric function, using individual maximum leaf area would
muse under estimation of the largest leaf area. For example, the hybrid 882 z 105 had 27 leaves.
Using the Gompertz with an A0 value of the largest leaf would muse the function to
asymptotimlly approach A0 at leaf 27. Therefore, a value larger than the area of the largest leaf
must be used. Since the largest area of all of the hybrid leaves is 896 cm2, from hybrid 882 z 105,
a value of 1000 was used for the variable A0. The value of b is the exponential of the Y-intercept

and the value of k is the slope. The regression analysis yields a Gompertz equation as follows:

A _ 10006675 c-0.211LN [54}

Using equation [E4], the predicted leaf area can be plotted against leaf number along with the
measured values for all of the hybrids. This is shown in Figure 7.

As seen in Figure 7, the Gompertz does a good job of fitting the data. However, once
again the smaller leaves at the beginning of the season are all over estimated, due to the high
value of b: -6.75. Additionally, one can see that the function does not adequately fit all of the
data sets up to the maximum leaf area. This is due to the value of AD: 1000.

To get a better fit, the value of b must be decreased. However, this can only be done by
increasing A0. Using the method of trial and error, a new AD value was determined: 1200. Using

1200 instead of the maximum area of the largest leaf, equation [E3] was remlculated. The new

 

 

 

 

 

2.0

1g: Y = 1.91 + (—.211 t x) L
. '5 2 _ .
1.6- ‘\ R — 0.936 _
A . \ . A0 = Area of Largest Leof .
2 1.4~ :\ I r
\ .1 \ a
O 1.2- I l -
3 . \i g 1
0°: 1.0-1 -
o I , \ .
d 0.8-1 k " .
on: . i \\ I .
3 0.6: ‘\\ '1
0.4‘ '\\{ 1
0.2“ I \ It -'

0.0 r I T T r I r I TY I I r I 1 I '\I y I T
O l 2 3 4 5 6 7 8 9 10 11

Leaf Number

 

 

 

Figure 6. The regression analysis between the linearization of the Gompertz
equation using the area of the largest leaf for A0 and individual leaf number for
the seven hybrids tested.

 

 

 

 
   

 

 

 

1200
1100‘ Y = AotEXP(—6.75tEXP(—O.211¢X)) .
. A0 = 1000 1
1000-
A ‘ .
“ 900- .
E . ‘
¢; 800— _
E 7001 _
—i 6001 1
L _‘ _I
8' 500‘ 3k Ml86 .
0 400‘ x MI87 -
0 l + x304 ‘
L - .1
< 300‘ o 2883 .
200‘ A 2882 a
2 El 873 ‘
iOOJ O A632
O-l I'"'r7'"l'l'I‘l'Tvr'lYI'r'l'iv
0 2 4 5 81012141618202224262830323436
Leof Number

 

 

 

Figure 7. The area per leaf versus leaf number for the seven hybrid varieties
tested. Also included is the Gompertz equation using a value of 1000 for A0.

39

regression is shown in Figure 8. The resultant regression yields an R2 value of 0.981 and the data
are less smttered than that found in Figure 6 when A0 was equal to the area of the largest leaf.
The value of the Y-intercept was increase to 2.01 while the slope remained fairly constant,
decreasing only slightly.

Once again, taking the exponential of the Y-intercept, we get a new value for b: -8.08.

The slope of the line, k, is -0.193. Using these values along with an A0 of 1200, we can rewrite
equation [E4] as:

A - 120053-08 {om L" [135]

This new Gompertz function can then be plotted against leaf number along with the measured
data for all of the hybrids as shown in Figure 9.

Figure 9 shows the new Gompertz more closely describes the leaf area for the smaller
leaves. However, there is still some "noise” in the upper portion of the curve. This noise is most
likely due to the larger leaf sizes. Large leaves appear on maize plants due to one of two
conditions. The first is that the plant grows many leaves and therefore has more time to build up
leaf area. The other is genetic. Some varieties simply grow larger leaves than others. Muchow
and Carberry (1989), mentioned in their conclusions that there was clearly 'A genotypic difference
in the coefficients of the functions describing leaf growth” (referring to their coefficients Yo, the
area of the largest leaf and X0, the leaf number of the largest leaf). To account for this genetic
difference a new genetic parameter is required. To keep within the input format already present
In CERES-Maize, a new genetic parameter was added, P6. P6 is a parameter that describes the

leaf type of the variety. A value of 1.0 would indimte that the leaves are of average size. Values

 

 

 

 

 

2.0
(8.“ K Y = 2.01 + (-.195 t x) L
. " \ R’ = 0.981 .
1.6- l\ . _
A . \\ Ao =1200 .
A 1.4- _
<0 . ' \i ,
3 1.2-.1 ‘ \ i j
G)
0‘ 1°01 E\ I ‘
O .. \ 'l
_J
v 0.8‘ ‘\i “
a: « s ‘
O 0.5-1 I\ I q
—l . I \I .
0.4- .\ , -
-t ' a
0.2- \:\ _
. \'
O-O ' I . I ' l ' I T r ' f1 l ' I ' l ' 1 r
o 1 2 s 4 5 6 7 8 9 10 11
Leaf Number

 

 

 

Figure 8. The regression analysis between the linearization of the Gompertz
function using a value of 1200 for A0 and individual leaf numbers for the seven
hybrids tested.

of 1.2 or 0.8 would indimte either a large leaf type or a small leaf type. The parameter P6 was

used as a multiplier and the new Gompertz reads:

A - 1200P6e‘8'08 “0'1” LN [E5]

Though the Gompertz function adequately describes the leaf area of the first leaf to the
largest leaf, another equation is needed to describe the area of the remaining lmves. Figures 7
and 8 indimte that the remaining leaves show an almost linear decrease in size. Muchow and
Carberry (1990), used a linear relationship to describe the area per leaf of the last four leaves of a

grain sorghum plant. Therefore, in CERES-1M, a linear function was used to describe the area

41

 

 

 

 

 

 

 

1200 P 8 E P 3
‘ Y = P6 EX — .08 X -0.19 X
11004 Act e ( a ( e _
10004 _
E ~ .
3 800: "
E 700: e
_l 600- 1
L.
3 500 its Ml86 '.
o 400— x M|87 -
‘D ‘ + x304 ‘
L u .1
< 500. o 2883 1
200— A 2882 “
j 1:1 87:5 _‘
100 I O A632 .1
0“ ' 1 T r v r ' ff

1'1 I'l'r'l'l'l'rvl'l'l'
0 2 4 6 81012141618202224262830323436

Leof Number

 

 

 

Figure 9. The area per leaf versus leaf number for the seven hybrid varieties
analyzed. Also included is the Gompertz equation using a value of 1200 for A0.

per leaf of all leaves after the largest leaf. Using the standard linear equation:

Y-aX+b WW

where Y is the area per leaf (cmz), a is the slope of the line, b is the Y-intercept and X is the leaf
number. Taking data from the largest leaf to the last leaf, linear equations were fitted for each
individual hybrid. The regression analyses for the hybrids are graphimlly shown in three figures:
Figure 10 contains data on the hybrids A632 and X304; Figure 11 contains data on hybrids B73
and 2883; and Figure 12 contains data on hybrids M186, M187, and 2882. The data were separate
into the three figures based solely on appearance and ease of reading. Regression data for all 7
hybrids can also be found in Table 3.

As Table 3 shows, the slope of the linear equation is approximately similar for all of the

hybrids, mcept 883 z 045. The reason for the difference is the hybrid’s flattened curvature around

42

Table 3. Linear regression analysis for each individual hybrid for area per leaf for leaves from the
largest leaf to the flag leaf.

 

 

 

 

 

 

 

 

‘ A632 x mm 75.57 0.995 I
I B73 1: M017 77.71 0.962
I 882 z 105 79.56 0.998
I 883 z 105 43.50 0.963
I x304c 67.51 0.984 I
M187 74.20 0.995
M186 . 62.21 0.989

 

 

 

the maximum leaf area. However, the other data shows fairly good agreement that would indicate
a common slope. Therefore, using the slopes from the remaining six hybrids, a single slope for all
hybrids was determined. This was done by averaging all of the slopes. The resulting linear

equations is

Y - -72.80X + b [53}

where Y is the area of the leaf X (cmz), b is the area of the largest leaf (obtained by the
Gompertz equation), and 4280 is the slope. This equation is used to predict the area of each leaf
from the largest leaf plus 1 to the last leaf.

To obtain the parameter b, the area of the largest leaf is needed. To avoid additional
inputs, the leaf number of the largest leaf is estimated and used in the Gompertz function to
estimate the area of the largest leaf.

To obtain the leaf number of the largest leaf, 8 linear regression was performed on the
data with total leaf number (TLNO) as the independent variable. Table 4 gives the TLNO and
the leaf number of the largest leaf for each hybrid. As the Table 4 shows, the difference between
the TLNO and the number of the largest leaf ranges from 5 (A632 7: W117) to 10 (X304C).

Because of its unusually low number, A632 was not used in the regression analysis. Again, using

43

 

 

 

 

 

1200
1 o x504: Y - 2246.16 — 67.51ox : R’ - 0.984
“001 1:1 A652: Y = 1552.86 - 75.57ax : R’ = 0.995 i
1000- ‘ -
(P 9004 6‘0. i
E 1 03-0. “
ii 800: :3. j
,0
E 700: j
.1 600- q 0f -
53 500- a "'9. -
o. «1 lab «1
200‘ ab ..
100~ -
4 ..
O ' I I r f I %r I T I T ' I ' I ' I ‘ I '
8 1O 12 14 16 18 20 22 24 26 28 30
Leaf Number

 

 

 

Figure 10. The regression lines for the area per leaf versus leaf number for leaves
after the largest leaf for varieties A632 and X304.

 

 

 

 

 

1200
1 o 875 : Y = 1732.79 — 77.71.x : R’ - 0.962 1
“001 U 2885: Y = 1324.51 - 43.501x : R’ = 0.963 1
1000« -
A ‘ 4
N - q
E 900 ‘ .1
3, 800~ j
g 70°: 1
-. U
_l 600‘ “q a "a... a “
o 400~ 0 °‘-.. -
q) 1 a...“ .
2 30°“. ° "a. '1
200- ° j
100- ° 8
O f I ' T ‘ I f T r I ' l ‘ T ' I ' I ' I '
8 10 12 14 16 18 20 22 24 26 28 30
Leaf Number

 

 

 

Figure 11. The regression lines for the area per leaf versus leaf number for leaves
after the largest leaf for varieties B73 and 2883.

 

 

 

 

 

1200
o 2882 : Y - 2417.45 - 79.569X : R' - 0.998 4
“001 1:1 M186 : Y = 1586.51 — 62.219X : R’ = 0.989 j
1000- A MI87 ; Y - 1700.62 - 74.20-x : R’ - 0.995 -
r? 900-1 0 .j
E 4 o .
3 800‘ -
“- 1 A. '0... j
g 7007 0 °. '9._ 1
.1 600-1 o "-6 a
L '1 0.3.9 ‘1
Q) 500« “a. 6‘ 2
O. 1 0.. ”A o q
8 400-4 0 3", n'o :
2 300: “'0‘ 2'62 -.
200~ 6.6- A "6 7
1004 6

0 # I I I I I I I I I T I ' I ' I ‘ I I I '
8 1o 12 14 15 18 20 22 24 26 28 :50
Leaf Number

 

 

 

Figure 12. The regression lines for the area per leaf versus leaf number for leaves
after the largest leaf for varieties M186, M187, and Z882.

the standard linear equation (equation [136]), a regression analysis was performed (Figure 13).

The resulting equation is:

Y - 0.796X - 3.53 [E91

with Y equal to the leaf number of the largest leaf and X equal to the T'LNO. From this
equation, the leaf number of the largest leaf for each hybrid was determined. Table 5 gives a
comparison between the actual leaf number of the largest leaf and the predicted one. As expected,
the hybrid A632 1: W117 shows the largest difference. This is due to its unusually high ratio of the
leaf number of that largest leaf to the total leaf number.

Using the number of the largest leaf from equation [E8], in conjunction with the
Gompertz Function (equation [56]), the area of the largest leaf can be determined. Values of the

new genetic parameter P6 were used in this mlculation. Table 6 gives the actual and predicted

45

 

 

 

 

 

20

,_ . Y = -5.55 + .7963X o _.--'o
0 R’ = 0.976
8 18~ ‘
‘63 . .-
cv
8‘ 16-1 7 ‘
o J
_l I . o
“6 144 J
6 o ~
.0 J
g 12- °
3
Z -1
“5 104 -
Q)
_J J

8 I I I I I I I I I I I I I 1 I

14 16 18 20 22 24 26 28 50

Total Leof Number

 

 

 

Figure 13. Regression line for the leaf number of the largest leaf and the total
number of leaves for six of the hybrid varieties.

Table 4. Total number of leaves and the leaf number of the largest leaf for each hybrid.

 

 

 

 

 

 

 

 

 

5 fl
Hybrid Total Leaf Number Number of largest Leaf

A632 x W117 18 13
B73 1: M017 20 13
883 z 045 24 15

882 z 105 27 19 !

X304C 29 19 I
M187 21 13
M186 20 12

 

 

 

 

46
Table 5. Measured and predicted values of the leaf number of the largest leaf for the hybrids.

 

Hybrid Actual Number Predicted Number Difference

    
   
   
   
   
   
    

 

 

 

 

  
  
  
  
 

A632 x W117 13 10.8 - 2.2
B73 x M017 13 12.4 - 0.6
883 z 045 15 15.6 + 0.6

 

882 z 105 19 18.7 - 03

X304C 19 18.7 - 03
M187 13 13.2 + 0.2
M186 12 12.4 + 0.4 |

 

 

 

   

 

      

 

 

 

Table 6. Measured and predicted leaf area (cm2) of the largest leaf and the associated genetic
parameter P6 for each hybrid.

 

 

 

 

 

 

 

 

 

P6 AZual Area Predicted Area Difference

(cm’) (662) (cm’) I

A632 x W117 1.0 579 439.9 - 139.1 I
B73 x M017 1.1 654 631.0 - 23.0
883 z 045 1.1 627 921.9 + 294.9
882 z 105 0.9 896 869.9 - 26.1
X304C 1.0 925 966.6 + 41.6
M187 1.2 737 764.2 + 27.2
M186 1.2 633 688.4 A + 55.9

 

 

 

 

 

area of the largest leaf for each hybrid. Most hybrids show good correlation except for A632 x
W117 and hybrid 883 z 045.

Using the values of the area of the largest leaf obtained by the Gompertz function, the
linear function can then be used to determine the area of the remaining leaves. Therefore, there
now exists two equations which describe the area of a leaf, both based on leaf number. A
regression was performed on all of the hybrids using both the Gompertz and the linear equation.
The actual and simulated data for the area per leaf for all seven hybrids are shown in three

figures: Figure 14 contains data on hybrids A632 and X304; Figure 15 contains data on hybrids

47
B73 and 2883; and Figure 16 contains data on hybrids M186, M187, and 2882. Also, Table 7 gives

the R2 values for each hybrid and for the Gompertz and linear portions of the curves. For the
Gompertz equation, the R2 values range from 0.984 for the 883 z 045 to 0.999 for hybrids 882 z
105, X304C and M186. The low R2 value for hybrid 883 z 045 is due mainly to the
uncharacteristic 'flat' top around the largest leaf. For the linear fit, R2 values range from a low of
0.883 for hybrid A632 x W117 to a high a 0.999 for hybrid 883 z 045. The low R2 value for hybrid
A632 x W117 is clearly due to the large difference between actual and predicted largest leaf
number. The measured value was leaf 13 while the predicted largest leaf number was 10.8. The

Table 7. Results of regression analysis performed between measured and predicted areas of leaves
for both the Gompertz and linear functions.

 

 

 

 

 

 

 

 

I Hybrid Gompertz R2
l A632 x wu7 .998
I B73 x M017 .998
I 883 z 045 .984
882 z 105 .999
.999
M187 .996
I M186 .999

 

 

 

 

others all show good agreement for both the Gompertz and the linear equations used, with R2
values all above 0.99.

No new subroutines were written for the new leaf area relationship. The first is
LEAFAR (LEAF ARea). A listing of the subroutine is in Appendix one. LEAFAR is used twice
in a single season run. It is used in the beginning of the simulation to fill an array with leaf
number and plant leaf area, using the Gompertz and linear functions and summing the individual
leaf areas. Since the program does not know what the final leaf number will actually be, a value

of 22 leaves is used. Thus, before the season begins, the program first calculates the leaf number

 

 

 

 

 

 

1400
,Ax304 .
¥A632
1200- _
p 1
E1000- .
0
V
“6 800~ _
Q)
-J 1
a; 600- ~
0. 1 .
8 400— -
I.
“ 1
200- 5
.1
O '1' IITI'I'I'I‘I'I‘I'I'Ijljl'
0 2 4 6 81012141618202224262850
LeofNumber

 

 

 

Figure 14. Measured values and predicted values using the Gompertz and linear
equations of area per leaf for hybrids A632 and X304.

 

 

Arec per Leof (cm?)

 

 

 

 

I'I'I'I'I'I'I'I'I'I'I'I'II
0 2 4 5 81012141618202224262830

Lecf Number

 

 

 

Figure 15. Measured values and predicted values using the Gompertz and linear
equations of area per leaf for hybrids B73 and Z883.

49

 

 

 

 

 

 

1400
o 2882 .
1200- _
g" .
E 1000- _
0
V 4
“6 800- -
cu
_’ 4
a; 600- -
Q- 1
8 400- -
L.
< . 1
200- -
01 'IfiIIIIIIIIIII'I'I'I'IfiIII'
024681012141618202224262830
Leof Number

 

 

 

Figure 16. Measured values and predicted values using the Gompertz and linear
equations of area per leaf for hybrids M186, M187, and 2882.

of the largest leaf, using equation [E9]. Then, using the Gompertz equation, equation [E6], the
area of the first leaf to the largest leaf is calculated. Finally, using the linear equation, equation
[ES], the remaining leaf areas are determined. All of the leaf areas are placed into an array and
the values accumulated.

As the season progresses, CERES-Maize determines the total leaf number based on the
genetic parameters entered and the accumulated TT. This occurs at tassel initiation. At this
point, the leaf area array is reealculated as it was at the beginning of the season but with one
exception. Instead of using the default value of 22 leaves for maximum leaf number, the program
uses the calculated total leaf number. The result is a new array, with plant leaf area values based
on the Gompertz function for leaves 1 through the largest leaf and the linear function for the
remaining leaves. Though individual leaf areas have been determined, the next step is to calculate

total plant leaf area.

50
When simulating leaf growth, CERES-Maize calculates leaf tip appearance. The variable

XN represents the number of leaf tips and XN does not always come in whole numbers. While it
is impossible to have 3.4 tips appearing, the variable XN is used only as a counter, based on
degree days. When a sufficient amount of TI‘ has occurred, the a new leaf appears. For example,
if enough TT has accumulated for three leaves plus half the TI‘ required for a fourth leaf, then
XN would have the value of 3.5. To deal with these decimal numbers, an interpolation of the
array is required.

A new subroutine was created to handle the interpolation of the leaf area array,
LFINTER (LeaF INTERpolation). LFINTER uses a linear interpolation to approximate the
potential plant leaf area growth. A listing of this subroutine is in Appendix one. Using the leaf
tip number, XN, provided by CERES-Maize, LFINTER interpolates the plant leaf area growth by
using the array set up in LEAFAR. The plant leaf area growth determined by the interpolation is
actually a potential growth. This potential must be subjected to reduction due to nitrogen or
water stress. In the present CERES-Maize program, this reduction is made by the following
statement:

PLAG = PLAG " AMIN1(NDEF2,SWDF2)
where PLAG is the plant leaf area growth, AMINl is a FORTRAN function that returns a value
equal to the minimum value of the variable list within the parentheses, in this case a nitrogen
(NDEFZ) and water (SWDFZ) stress factor. These stress factors have values that range from 0.0
to 1.0, with 1 representing no stress and 0 representing maximum stress. This same procedure is
used in the new leaf area model. Once the subroutine LFINTER has determined the potential
leaf area growth, the actual leaf area growth is reduwd if appropriate. Although the functions for
relating individual leaf area show good correlation with measured leaf area, the area expansion
rate must be determined with another function.

Muchow and Carberry (1989) used the number of fully expanded leaves in their equations
as the independent variable when predicting total plant leaf area. Their function worked well

when describing individual leaf areas, however it did not provide good agreement when predicting

51

total plant leaf area. The reason for this was that their function did not account for expanding
leaf area. Consider for example, when there are five fully expanded leaves, there will likely be an
additional two or three leaves above leaf five that are expanding. They assumed that the leaf area
at any given time in the development of leaves is equal to the area of the leaves fully expanded
plus the area of the next two sequential leaves. This methodology was also applied to grain
sorghum (Muchow and Carberry, 1990) with equally good results except that the fully expanded
area of the next 1.6 leaves was assumed.

In the new CERES-1M, the independent variable is leaf tips, XN, not fully expanded
leaves. When XN equals five, that does not represent five fully expanded leaves, and to use the
leaf area data in the array that corresponds to five leaves would cause an overestimation of plant
leaf area. In actuality, there are usually three to four leaves expanding at a time. For CERES-1M,
it was determined that the expanding leaf area could be adequately approximated by assuming that
the total leaf area was equal to the area of the fully expanded leaves for two leaves less than the

leaf tip number.

Other Modifieations

The ability to predict nitrogen stress and its effect on plant growth and yield is one area
that is vital to this work. Though the present CERES-Maize model does contain a subroutine to
determine nitrogen deficit factors, the approach used often gives erroneous nitrogen stress at
relatively high nitrogen levels while the stress induwd at low nitrogen levels is often not severe
enough to adequately simulate field observations.

There are three nitrogen deficit factors used in the CERES-Maize model. All three of the
factors are a function of NFAC, a variable that relates the actual plant nitrogen content with the
minimum required for adequate growth. The variable NFAC is a 0 to 1 variable, with 1 indicating
adequate nitrogen content and 0 representing maximum deficienqv. The three nitrogen deficit

factors follow the same logic, with a 0 to 1 range and with 1 being no nitrogen stress and a value

52

of 0 indicating maximum nitrogen stress. The three are used to provide varying sensitivity values
to different physiological process.

The first nitrogen factor is defined as NDEFl, which affects the photosynthesis of the
plant and is used when calculating the plant’s carbon assimilation. In its original form, this factor
slowly decreases as the variable NFAC decreases in the form of:

IF(NFAC .GT. 0.5) THEN
NDEFl = NFAC ' 0.4 + 0.6
ELSE
NDEFl = NFAC ‘ 1.2 + 0.2
ENDIF
However, the decrease when NFAC is only slightly reduced has been found to be too severe and is
also too slight when NFAC is greatly reduced. Following similar adjustments made in the
CERES-Wheat model (Godwin et al., 1989) NDEFI was redefined. The new code is:
IF(NFAC .LE. 0.45) THEN
NDEFl = NFAC ‘ 2.0 + 0.1
ELSE
NDEFI = 1.0
ENDIF

This new formulation does not allow photosynthesis to be affected by nitrogen stress until
NFAC is reduced to 0.45. At that point, the reduction is slightly more than the old code.

The second nitrogen deficit factor is NDEFZ. This factor affects plant growth, namely leaf
growth. In its original form, the relationship reads:

NDEH :- NFAC ‘ 0.95
In this relationship, the highest value NDEFZ can obtain is 0.95, indicating that a nitrogen stress
was occurring regardless of the plant’s nitrogen content. Whether this was an error in coding or a
functional error, this relationship also required adjustment. Using the CERES-Wheat model as a
basis, NDEFZ is defined as:

NDEIQ - NFAC ‘ NFAC

This new formulation does not vary much from the old, but rather corrects the problem of

always having a small nitrogen stress when there was adequate nitrogen.

53
The final nitrogen factor is NDEF3, which is used in the grain filling functions. In its
present form, it reads:
IF(NFAC .LT. 0.8) THEN
NDEF3 = 0.2 + NFAC
ELSE
NDEF3 = 1.0
ENDIF
This relationship allows for the nitrogen deficit factor to remain at 1 until the plant nitrogen
content becomes reduced significantly. Since this relationship follows logical deduction, it was not
altered.

Another change that was made to the CERES-Maize model is the relationship between
potential carbon assimilation (PCARB) and photosynthetically active radiation (PAR). The
present formulation reads:

PCARB = 5.0 ‘ PAR / PLANTS ‘ (1. - AMAX1(Y1,Y2))

where PCARB is the potential carbon assimilation (per plant), PAR is the photosynthetically
active radiation (M1 in:2 d'l), PLANTS is the plant population (plants n14), and Y1 and Y2 are
LAI factors related to row spacing that range from 0 to 1. AMAXI is a FORTRAN function that
returns a value equal to the maximum value of the variables listed within the parentheses. In the
past, CERES-Maize has often over-estimated biomass accumulation, a result of the radiation use
efficiency coefficient 5.0 used in this formulation (Ritchie, personal communication). Work done
by Kiniry et a1. (1989) showed that the mean value for this coefficient was closer to 3.5. However,
their study only included above ground biomass accumulation and their coefficients ranged from a
high of 4.5 to a low of 2.1, with the lower values representing extremely low plant populations.
Therefore, taking into account that the Kiniry et al. study only observed above ground biomass
and that the mean was 3.5, it was decided that the new coefficient should be more reasonably
approximated by 4.5.

The final major change made to the CERES-Maize program was the method used for

determining grain number. In its present form, the model determines grain number based on the

54
average carbon assimilation rate from 50% silking to the beginning of grain filling. The formula

reads:

PSKER = SUMP '100/IDURP ‘ 3.4/5.0

6PPC = PSKER/TZOO

GPP a: 62 ‘ 6PPC +50
where SUMP is the cumulative carbon assimilated during CERES-Maize phenological growth
stage 4 (hour silking to effective grain filling, g plant-1), IDURP is the duration of stage 4 (d),
PSKER is the average rate of earbon assimilation during stage 4 (g plant'1 d“), 62 is the potential
number of grains per car (entered as an input in File 9, the geneties file) and 6PP is the predicted
grains per plant. Additionally, the effect of nitrogen within the plant is taken into consideration
and 6PP is reduced if there is a nitrogen deficiency. The 6PP formula presently used is based on
work done by Edmeades and Daynard (1979). In their paper, the authors suggest a correlation
between biomass accumulation rate just prior to anthesis and kernel number. This theory is well
supported through the literature for a variety of cereal crops (maize - Hawkins and Cooper, 1981;
Kiniry and Ritchie, 1985: wheat - Fischer and Maurer, 1976; Rawson and Bassa, 1979: millet - Ong
and Squire, 1984: rice - Evans and DeDatta).

Problems have occurred with this relationship as the model has been used at many
locations in the world. Using the rate of earbon assimilation rather than accumulated assimilation
may cause some problems. Edmeades and Daynard (1979) found that carbon assimilation rate had
a hyperbolic relationship with grain number. However, Hawkins and Cooper (1981) found the
relationship to be linear in maize. To simplify this relationship, it was decided that the total
mrbon assimilated during stage 4 would be related to grain number. The new relationship reads:

6PP = ACCTOP ” 62
where ACCI‘OP is the accumulated above ground biomass during stage 4 and 62 is a new genetic
parameter describing potential grain number per plant. This new 62 was determined by

calibrating the hybrids and inbreeds used in the simulations.

55
Another change needed in the model for this work was the duration of the simulation.

Normally, CERES-Maize simulates soil and plant dynamics from the first day of simulation (given
in input File 8) until physiological maturity. However, with the use of male plants and the need
to investigate drainage and leaching both before and after the growing season, an alteration was
made. The model still begins simulation on the day entered in File 8, however the model
continues to simulate soil water and soil nitrogen dynamics until the end of the weather data. In
doing this, weather files can be arranged such that the simulation takes place from the beginning
of one growing season until the beginning of the next growing season. This allows drainage and
leaching dynamies to be simulated throughout the entire year.

Once the change was made to the model to allow for year round simulation, changes
needed to be made to allow for snowfall. When snowfall occurs, the amount of precipitation, in
rainfall equivalent, is entered in the weather file. In its present form, CERES-Maize reads the
rainfall data and begins to simulate the distribution of the rainfall through the soil profile.
However, since the rain is in reality snow, it is not available for distribution to the soil profile
until it melts. Though CERES-Maize had no routines available to handle snowfall, CERES-
Wheat did. The subroutine SNOW was added to CERESJM and allows for the accumulation of
snow and, if temperatures warrant, the melting of snow. Determination of both snowfall and snow
melt are based on maximum temperature. A listing of the subroutine is in Appendix one.

Other changes to the model were made as deemed necessary, though no basic functional
changes were made except for those mentioned above. Error traps were placed where appropriate

and input/output changes were made where required.

OBJECTIVE 2. Perform field experiments to validate the simulation model.

bred libratio alidati

In 1988, researchers at Michigan State University, in cooperation with Pioneer Hi-Bred

International, he, began a study to investigate alternative nitrogen management strategies for

56
seed corn production. Conducted in Constantine, Michigan, the study was designed to aid local

seed corn growers in managing their nitrogen applications. The effect of various nitrogen
management strategies on nitrate leaching and the growth and yield of inbred maize were subjects
of study. Four nitrogen treatments were used with three inbred varieties. Drainage lysimeters
were installed to determine the impact of nitrogen applications on the leaching of nitrates from

below the rootzone.

Lysirneter Installation and Drainage Sampling Systems

During fall of 1988, five drainage lysimeters were installed: four were 'disturbed' while the
fifth was an undisturbed monolith. The dimensions of the lysimeters were 0.91 m by 3.81 m by
1.83 m deep (see Figures 17, 18, and 19), and they were placed approximately 45.7 cm below the
soil surface to allow normal field operations to take place. The disturbed lysimeters were installed
using the same method described earlier in the hybrid drainage study. The one undisturbed
lysimeter was installed by first cutting a soil monolith out of a barrow site in the field, placing the
lysimeter walls around it, sliding a plate of steel under it, and then lifting the containerized soil
with a crane. The lysimeter, with the monolith inside, was then turned over and the bottom of the
lysimeter was welded on. Finally, the soil was excavated where the lysimeter would finally rest and
the entire lysimeter, with the monolith still intact, was again inverted and placed in the ground.

A sampling system was installed in June 1989. Modified tipping buckets from rain gauge
sensors were chosen to measure the flow from the lysimeters because they could easily be
interfaced with a datalogger. Tipping bucket assemblies (Sierra Misco) were attached to an
aluminum mounting plate. The mounting plate was bolted to the top of a 38 liter polyethylene
reservoir. The lysimeter soil drainage water passed through a 1.27 cm diameter pipe in the
lysimeter wall, located at the lowest point. The drainage water was then funneled to the tipping
bucket assembly. The water dumped by the tipping buckets was then funneled to the reservoir. A
PVC pipe was fed through the aluminum mounting plate to the bottom of the reservoir and

connected to the suction side of a 12-volt diaphragm pump (Flojet) using a flexible hose. The

57

42 6525200 .822: E 26 35823on =< doses—65
ES 39. 5 33¢. EoEomnauE Smog: 65 E 8...: £688?— ofi Lo 25 .8 32> 2.0 25 Be? e9 2: Lo marine 6:25:96. 4. .2 2:3."—

3; QB

 

”88.. 3...
838:3 823

     

3m; em
as 5.88:8 NS 1 sis

as“;
as... 1631\4 \| 3:63
a ease , 863 Ea: E. w

Il/fl E. ban—Se «Sago—ms; ......... 1* ,
. 1%; DWSWI i...
I.
.6.» , .1115
1g? a.” ”Magda.” |\.|L

 

 
    

 

 

 

 

 

 

 

 

 

 

 

638:3 8.55

 

 

 

58

.52 6558860 .522: 5 0.5 8:652:80:—

=< 4.5.2.85 58 Bow 5 .32... 56569655 comes? 2: 5 8...: E8655 2: Ho 25 _o 35> 65.6. 2: .8 M5356 53528 < .3 653m

 

 

_lllm..llr_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a
P

 

 

 

_lla..L

U\l

553:8 usfimfimza

8.5 833:8 _ . .
6956.5 6...... > W mm
. 8...... i J. 82...... .2
mm L 95%: w [I11 .358 B as“
g.
a .5 8 l .53 6958“.
335 um 8. 5:58 1/ 1 632.33
8;
6595855 B _.
.
m1 2...... :6
/ 2» @5538“.

 

MT 16.... 6.2.6.... 5....

 

 

59

.=2 65552.8 .232: 5

2a anon—2385 =< £263.05 58 39. 5 33m EoEommumE cowobE on. 5 com: E2259: 2: .o 25 .8 9522—. nun—=28 < .2 oEmE

 

:3 3% E.
.235 33%:

53.... .5 Mann”,
3.3 a a5 |\ ,

 

 

 

     

.2:
\I 3&3. 3 =35

/ 2 5%

 

 

him
as 53.3 s...

 

 

cause

/I 533:3 5.55

 

 

 

60

pressure side of the pump was connected to a 1.27 cm diameter PVC pipe leading to a sequential
sampler loeated outside the plot, approximately 12 m from the lysimeter. The PVC pipe and
cabling between the collection area and the sampler box ran through a 7.62 cm diameter conduit
buried underground at the time of the lysimeter installation.

The sequential sampler was contained in a waterproof wood box (see Figures 20 and 21).
The wood box was beneath the soil surface so the box lid was just above the soil surface. A
solenoid valve (Spraying Systems) was used to divert a portion of the accumulated drainage water
contained in the lysimeter reservoir during the pumping stage to obtain a sample for chemical
analysis. The solenoid valve was plawd above the sequential sampler and the diverted water
funneled through a small PVC pipe (1.27 cm diam.) connected to a rotating pipe. The diverted
water flowed through the pipe by gravity into a 500 ml wide mouth polyethylene sample bottle.
The pipe was rotated by a lZ-volt gearmotor (Grainger) when the sampling qcle was completed.
The major components of the sequential sampler included a 12-volt gearmotor, a roller micro-
switch, and a l4-tooth sprocket. The 14-tooth sprocket allowed the sequential sampler to collect
14 samples before the sample bottles required emptying. The sprocket was directly connected to
the gearrnotor shaft. The micro-switch and sprocket combination stopped the gearmotor so the
rotating pipe was exactly over the next empty sample bottle.

The electronic control system consisted of a Campbell Scientific CRIO micro—datalogger
interfaced with a Campbell Scientific SDM-SW8A Switch Closure Module and SDM-CD16
Control Port Expansion Module, placed in a separate waterproof box outside of the plot area and
next to a weather station. The SDM-SW8A increased the number of pulse channels available on
the CRIO for measuring multiple tipping bucket sampler pulses. The SDM-CD16 allowed the
datalogger control of remotely powered electrical devices through the use of 12-volt relays
controlling the current going to the 12-volt components (the pumps, the valves, and the
gearmotors). Power was supplied to the components via a 12-volt battery placed in the datalogger

box and in each sequential sampler box. The batteries were connected in parallel and were

61

42 6558200 £2832; 58 33 e8
has: Eva—ounce:— =ouo£= 2: 5 new: flange: 2: Soc 833 own—:8“. 05 29:9. 9 3m: SEES 3.533. 2: .8 32> on: < .8 Baum

 

mmfieyon mfiaemm

mafia mcflgapom

 

Hmcczm

 

x05
causewmycau amccfi

   

Hfim: mac ch
m>Hm> mewpam>fla

 

 

 

62

42 6523200 dozen—co... ES 38 .8
b5: Eoeomaaae .5ng 05 5 3...: 8225.5 2: fig .25 own—5:. 2: 0383. 9 v8: SEES 325:3» 2: Co 303 no. < Am 2:3."—

 

 

x8
225589 ES l/I

3E8 £95m |\

 

 

 

 

33 958.9: |\

 

 

 

63

charged by a battery charger connected when the voltage became low. The use of the 12-volt
system ensured continuous sampler and datalogger operation in case of AC failure.

The objectives for the water sampling system were to record the hourly flow of water
draining from the lysimeters and to sample a portion of that water after some predetermined
volume has been obtained. The sampling cycle began when a certain volume had accumulated in
the lysimeter reservoir as determined by the number of tipping bucket pulses. The volume to take
a sample can easily be changed by a simple entry in the datalogger program. Once a
predetermined number of tips have occurred, the datalogger program entered into a programming
loop. Once in this loop, the program directed the SDM-CD16 to ground the pump relay control
circuit that operated the pump and began pumping out the sample. The sample was then pumped
through the PVC pipe buried under the plot, through the diverting valve and into a dry well next
to the sequential sampler box. After 1 minute (used to expel any water left in the line from the
previous sample) the SDM-CD16 grounded the relay controlling the diverting valve. The valve
was opened for eight seconds to allow a sample to be taken. The pump continued to run for an
additional six minutes to pump any remaining sample into the dry well. Once the pump turned
off, the relay controlling the gearmotor was grounded for four seconds. That began to rotate the
diverting pipe. The gearmotor relay remained closed until a shaft connected to the diverting pipe
rotated a sufficient amount to allow the roller micro-switch to be closed by a sprocket tooth. The
gearmotor operation was then controlled by the micro-switch. When the roller was in the valley
of the sprocket, the micro-switch was opened, stopping the gearmotor and leaving the diverting
pipe aligned with the next sample bottle. This cycle was repeated for each lysimeter, but only one
sampling cycle could occur at any one time. After the sampling cycle, the time the sample was
taken, the sample number, and the total volume the sample represented was written to the final

output storage in the datalogger.

Experimental Design and Plot Layout

The plots were planted parallel to the .91 m dimension of the lysimeters so that an entire
block (four female rows with 1 male row, 76.2 cm spacing) could be planted within the lysimeter
area. The male rows were planted in the center of the lysimeter area with two rows of females on
either side. The entire plot dimensions were 5.24 m by 7.62 m, with eight rows of females and two
rows of males. Figure 22 shows a detailed plot plan. All normal seed corn production field
operations were performed on the plants, i.e., detasseling and male removal.

Three inbreeds were used in this experiment, with four nitrogen treatments. The inbreeds
used were an early maturity (inbred 1), a mid-season maturity (inbred 2), and a late season
maturity (inbred 3) variety. Each treatment was replicated four times. The lysimeters were all

planted with inbred 2.

Plant Growth, Development and Yield Measurements

The plant growth measurements included total above ground plant biomass and were
taken four times throughout the season. Each sample was analyzed for total nitrogen, including
the grain and cob. Developmental factors such as shedding dates for the males and silking dates
for the females were also noted. Harvest data included ear counts, stalk counts, barren plants,

plants with double cars, car length and grain yield.

Soil - Nitrogen Data

Soil samples were taken to evaluate the soil’s nitrate and ammonium contents. Samples
were taken at the beginning of each season to establish initial conditions. Also, samples were
taken just prior to detasseling and then again towards the end of the season. Sampling depths
were determined by examining the nitrogen analysis of the previous sample, and were changed
accordingly. Soil samples were also taken before the lysimeters were installed and analyzed for

percent organic matter, sand, silt, and clay.

d2 .oEESmSO $2.3. cone—62m E8 38 E EoEomS—aa couch:— ofi 5 com: 39?. SE 05 Co Ease < .NN oEwE

 

65

 

LmeEume mHfiL—OLQ 10m “UmDLJJmaUCD V
S

LmeEamml. mHCOLQ 20m UmDLJDwHD @

Lm—ULom

LmULUm

 

Haze—4
Heme—9|

TV fill $le “EITEI

F mg L

_/ \_

 

 

Irrigation and Nitrogen Application

The irrigation water applications for this study were scheduled using the 'Micro-
Scheduler“ computer program as described by Shayya et al. (1990). The program uses crop curves
developed for specific areas in Michigan. Though no crop curve exists for inbred maize, the curve
for hybrid maize has been used for irrigation scheduling on inbred fields with good results. This
curve was used for scheduling the irrigation of this study.

Nitrogen applimtions were separated into four treatments. These treatments are listed in
Table 8. Treatment 1, 180 kg N ha'1 preplant, was considered to be a typical treatment used by
many growers in the area.

Treatment 2, the research management strategy (RMP), used a new concept of Plant
Response Fertilization (PRF). The basic premise of PRF is to allow the plants to indicate the
need for nitrogen by showing signs of nitrogen deficiency. In 1989, the first year of the study, 30
kg N ha‘1 was applied preplant to assure adequate early plant growth. Then, an observed color
difference was used as a factor to determine if any additional nitrogen was required. In the second
season, 1990, the 30 kg N ha'1 preplant was not applied and more scientific approaches were used
to detect a nitrogen deficiency. One method employed the use of a multispectral radiometer
(MSR). The MSR allows for near simultaneous inputs of signals representing incident as well as
reflected irradiation. This enabled measurements to be taken from the crop canopies when sun
angles or sunlight conditions were less than ideal. Positioning the radiometer over the crop
canopies, measurements were taken at three wavelengths, 450 nm, 650 nm and 800 nm. The 450
nm is'the red region and the 650 nm is the blue region. These two regions comprise the
reflectance of the green color within the plant. The 800 nm represents the infrared and gives an
indication of plant biomass. Readings obtained from the MSR were in percent reflectance. The
more reflectance that occurred, the greener the plant leaves (in the ease of the 450 and 650 nm
wavelengths) or the more plant biomass there was (in the case of the 800 nm wavelength). The

radiometer was positioned approximately 1 meter above the crop canopy centered on the row.

67

Table 8. Nitrogen treatments for 1989 and 1990, inbred nitrogen management study, Constantine,
Michigan.

   
    
  

 

 

 

 

 

Treatment Nitrogen Treatment. Nitrogen Treatment.
(kg ha") 1989 (kg ha'l) 1990
l 1 180 - preplant 180 - preplant

2 30 - preplant 40 - PRF (1)8
50 - PRF o - PRF (2&3):

3 30 - preplant 30 - preplant
60 - sidedress 60 - sidedress

4 30 - preplant 0 treatment
150 - sidedress L

 

 

 

' Preplant applications were applied at planting and sidedress applications were applied between
the 6—8 collared leaf stage.

5 PRF treatment for inbred l was 40 kg N ha'1 and the PRF treatment for inbreeds 2 and 3 was 0
nitrogen added.

Readings were taken on one replication of PRF plots and on one replication of the 180 kg N ha’1
preplant plots (considered to be nitrogen non-limiting) for each inbred. Eight readings were taken
for each inbred and treatment, giving a total of 48 readings taken at a time. Plant height
measurements were also taken to detect any reduction in plant growth due to a nitrogen
deficiency.

Treatment 3 was a split application of 90 kg N lba‘1 with 30 kg N ha'1 applied preplant
and an additional 60 kg N ha‘1 applied sidedress. Treatment 4 was a split of 180 kg N ha'l, with
30 kg N ha‘1 applied at planting and an additional 150 kg N ha‘1 applied sidedress. However, due
to the small differences measured in yield between all of the treatments in 1989, it was decided
that this treatment would be changed for 1990. The new treatment 4 was a 0 kg N ha'1 to provide
a baseline from which to analyze the nitrogen effect of the other three treatments. All preplant

applications were made at planting and sidedress applications were applied between the 6—8 leaf

collared stage.

68
OBJECTIVE 3. Use the simulation model to evaluate the impact of several nitrogen

managements strategies on potential leaching in seed corn production.

With the completion of CERES-1M, attention then was focused on evaluating the impact
of nitrogen management strategies on the leaching of nitrates and crop yield. To achieve this goal,
an automatic fertilization subroutine was written into the CERES-1M programming. However,

before discussing nitrogen scheduling, a brief overview of how CERES-1M simulates soil nitrogen

dynamics is required.

CERES-1M Nitrogen Simulations
The simulation of nitrogen dynamics performed by the CERES-1M model is the same as

simulated in the CERES-Maize model, with the exception of the change mentioned earlier in this
section (i.e. adjustment of the nitrogen deficit factors) and the new autoofertilization routine.
Godwin and Jones (1991) give a detailed description of the nitrogen simulations used in the
CERES-N model, the same routines used in the CERES-Maize and CERES-Wheat models. The
following is a brief description of the nitrogen simulations found in CERES-1M.

CERES-1M simulates soil nitrogen dynamics on a daily basis if the nitrogen switch is
turned on. If the nitrogen switch is turned off, nitrogen is assumed non-limiting. The switch is an
input in the treatment management file, File 8. All nitrogen transformations are simulated within
the subroutine NTRANS. Within NTRANS, the subroutine is divided into four parts: 1) fertilizer
incorporation (if nitrogen was added to the soil on that day); 2) mineralization; 3) nitrification; 4)
denitrification. A flow chart of the soil nitrogen dynamics is shown in Figure 23.

If the nitrogen switch is turned on, the model simulates soil nitrogen dynamics for the
entire length of the weather data provided. Before the crop is planted, only the soil dynamics are
simulated but once the crop begins to grow, plant and soil nitrogen dynamics are simulated.

When the crop reaches physiological maturity, only soil dynamics are simulated. To better

69

.~.~> .68.: ofimiémmmo 2: E meets—=87. 535: we 8:033. 2: Co 829% >6: 4. .mn 05E...—

 

 

 

 

 

 

 

Z—(N‘U )0 ”MID—.5
Z UEU

 

 

 

 

 

70

understand how nitrogen dynamics are simulated, a short step-by-step procedural explanation
follows.

The first step in simulating the nitrogen dynamics is to read in the soil nitrogen input
data. This includes data from File 4, soil nitrogen balance parameters. These include any above
ground organic residue left from the previous crop and the depth of incorporation of that residue,
the carbon:nitrogen (C:N) ratio of the residue, and the dry weight of any root residue from the
previous crop. Initial soil nitrogen data are read from File 5, soil profile initial conditions. From
this file the program reads the initial amount of water, nitrate (NO3‘), ammonium (NHf), and
the pH of each soil layer. Finally, the program reads in the fertilizer data stored in File 7. This
data includes that day of application, the amount of nitrogen applied, the depth of incorporation,
and the type of fertilizer used. Once the data have been read, the model initializes the soil
nitrogen parameters.

Not all of the required soil nitrogen parameters are entered as inputs, so CERESJM
performs a few initial calculations to obtain values for the various nitrogen pools and sinks within
the soil. For example, values of the amount of fresh organic matter and the associated nitrogen
content must be determined. Once all of the initial conditions have been calculated, the model
can begin its simulation.

The first step in the simulation is to check whether fertilizer was added on the day of
simulation. If no fertilizer was added, normal soil nitrogen dynamics are simulated. However, if
nitrogen was added, it must be incorporated into the soil according to the inputs given.

In adding nitrogen from a fertilizer application to the soil, CERES-1M first determines to
which soil layer that the nitrogen is to be added to and the amount and type of nitrogen added.
CERES-1M allows for the use of 12 types of fertilizer. A list of these can be found in Table 9.
Within this list, four forms of nitrogen are defined; ammonium, ammonia, nitrate, or urea.
Depending on the composition of the fertilizer, varying amounts of each form is added to the soil.

For fertilizer types 2, 6, 7, and 11, the model simply adds the correct amount of ammonium to the

71
Table 9. List of fertilizer types and their codes used in the CERES-Maize simulation model.

CERES-Maize Fertilizer Type
Input Code

 

Ammonium Nitrate

 

Ammonium Sulphate

 

Ammonium Nitrate-Sulphate
Anhydrous Ammonia

 

 

Urea

Diammonium Phosphate

 

Monoammonium Phosphate

 

Calcium Nitrate

Aqua Ammonia

 

——__—_-——
\O on ‘1 0‘ VI «5 b) N H

s—a
C

Urea Ammonium Nitrate

 

Calcium Ammonium Nitrate

H
H

 

Potassium Nitrate

I

 

I:
L

appropriate soil layer. For types 4 and 9, which are ammonia based, the conversion to ammonium
is assumed to be instantaneous. All of the nitrogen is assumed to be in the ionic form, and is
added to the appropriate soil layer as ammonium. For fertilizer types 1 and 3, both ammonium
and nitrate are added to the appropriate soil layer. The composition of the fertilizer nitrogen is
assumed to be 50% ammonium and 50% nitrate. For fertilizer type 10, ammonium, nitrate and
urea is added. The composition of this fertilizer nitrogen is assumed to be 25% ammonium, 25%
nitrate, and 50% urea. Finally, for fertilizer type 5, only urea is added.

The next step is to perform soil nitrogen dynamics. First, if any urea was added, the
model begins to simulate urea hydrolysis. Hydrolysis of urea is a chemical reaction by which the
urea is converted to ammonia, NH3 and then to ammonium, NH4“.

Then the model simulates mineralization. Mineralization is the process by which organic

nitrogen is converted to mineral nitrogen. Also, the immobilization of nitrogen, a process by

72

which mineral nitrogen is converted to organic nitrogen, is simulated. This makes the nitrogen
unavailable for plant uptake, thus the term immobilization.

The next step is to simulate the conversion of ammonium to nitrate. Referred to as
nitrification, this process involves two steps. First, ammonium is converted to nitrite, NOz', and
then the nitrite is converted to nitrate. CERES-1M assumes a constant rate for the reaction. The
actual amount of ammonium that is converted into nitrate is based on this rate along with soil
temperature and soil water content. Finally, denitrification is simulated. This process is a
chemical reduction of nitrate to dinitrogen, N2, or nitrogen gas. The gas then escapes through the
soil and is lost to the atmosphere.

Once all of the soil dynamics have been simulated, the model simulates nitrogen
movement due to water movement. The model simulates the redistribution of soil nitrogen
attributed to the redistribution of soil water. Soil water concentrations are calculated and the
appropriate amounts of nitrogen are distributed among the soil layers. Any soil water that drains
from the last layer is considered to be drainage and the amount of both water and nitrogen loss is
recorded.

After the soil nitrogen has been simulated, nitrogen uptake by the plants must be
simulated. In CERES-1M, this is a three step process. First, nitrogen deficit factors must be
determined to simulate any effect of nitrogen stress on plant grth for the day of simulation.
Then, plant biomass accumulation is simulated. Finally, based on the new plant biomass, soil
nitrogen is removed from the soil and translocated in the plant. This process is repeated for every
day of the growth cycle.

The nitrogen cycle is a series of chemical reactions and interactions within the soil
environment. Many factors, including soil water content, pH, C:N ratios, temperature, and organic
carbon content play a role in how fast certain reactions take place and how much and what form
the soil nitrogen is in. Additionally, the translocation of nitrogen from the soil to the plants, and
within the plants themselves, create problems in modelling. CERES-1M uses empiriml formulas

to simulate these nitrogen interactions.

73

Evaluation of Nitrogen Management Strategies

Most present day nitrogen management strategies rely on yield goals to determine the amount
of nitrogen fertilizer to be applied. Some work has been done in testing soils for nitrogen content
and relating the results to the expected amount of nitrogen required for good crop growth.
Blackmer, et al. (1989), showed fair correlation between late spring soil tests and corn yields in
Iowa. However, these tests cannot fully account for the soil’s ability to mineralize organic
nitrogen. This mineralized pool of nitrogen is what must be utilized to its fullest to limit nitrate
leaching. One alternative to traditional methods of estimating plant nitrogen requirements is the
method mentioned earlier, Plant Response Fertilization. By watching the plants for signs to
determine when a nitrogen deficiency exists, nitrogen can be applied only when needed, thus
taking full advantage of the mineral nitrogen pool within the soil. To test the feasibility of such a
strategy, new code was written for CERES-1M that evaluates the plant’s nitrogen content in
relationship to the minimum required for adequate growth Once the plant’s nitrogen content
falls below this critical level, nitrogen must be added to the soil.

The auto-fertilizer routine uses the nitrogen deficit factor NFAC. As mentioned earlier,
NFAC is a 0 to 1 variable that relates the actual nitrogen concentration of the plant to the
minimum required for adequate growth. A value of 1.0 for NFAC indimtes no nitrogen stress and
a value of 0.2 is maximum nitrogen stress. The threshold value of NFAC chosen to trigger a
nitrogen fertilizer application is 0.90. This value was chosen because preliminary simulations
showed no significant increase in grain yield when nitrogen fertilizer was applied prior to this
value. Once NFAC is less than or equal to 0.90, the model simulates the addition of nitrogen
fertilizer to the soil.

Using CERES-1M in conjunction with simulated weather, nitrogen strategies in seed corn
production were evaluated for their impact on yield, leaching, and revenue. The simulated
weather used was compiled using real weather data and a weather estimation program called
WGEN (Richardson, 1984). Sixty years of weather data were simulated for this exercise. The

weather data generated included the 4 inputs required by CERES-1M, i.e., daily solar radiation,

74

maximum temperature, minimum temperature, and precipitation. This site chosen for the
simulation was the center of St. Joseph county, Michigan.

The soil inputs used were taken from the hybrid study used for the hybrid calibration done
earlier. These data were chosen because they represented a soil in production. The soil data from
the inbred study were from a newly cultivated field, and did not necessarily represent normal field
conditions. The variety chosen for these simulation was inbred 1 from the inbred analysis

performed earlier. Inbred one was chosen because it was best simulated by the CERES-1M

program.

Reprogramming of CERES-1M to Accommodate Multi-year Simulations

In addition to the auto-fertilization routine that was written, CERES-1M required other
changes in the code in order to simulate 60 years of production. The simulations for the hybrid
and inbred calibrations performed earlier were on a yearly basis. The inputs required for initial
conditions were entered for each year of simulation. However, in the 60 year simulation, it was
decided that the model should be continuous, each year depending on the previous year’s biomass
production and nitrogen uptake. In order to accommodate this, a few changes were required.

The obvious change was to reprogram the model to run continuously without reading
input files at the beginning of each simulation. Additionally, the residue inputs for each year,
except for year one, were the residue left by the previous years’s crop. The inputs for year one
were read in through the traditional input files. However, for the remaining 59 years, the residue
simulated by CERES-1M was entered as the inputs for the fresh organic matter for the following

year.

Nitrogen Strategies Evaluated
The possibility of nitrogen management strategies that can be evaluated are endless. To
keep the amount of data within a quantifiable amount, only 14 strategies were chosen. These

strategies are listed in Table 10.

75

Table 10. Nitrogen management strategies simulated by CERES-1M for a 60 year period.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Strategy Number Description
1 Zero nitrogen added
2 180 kg N ha‘1 applied at planting
3 180 kg N ha'1 sidedressed (30 kg N at planting and i
an additional 150 kg N at V6) ‘
4 90 kg N ha‘1 sidedressed (30 kg N at planting and
an additional 60 kg N at V6)
5 30/30 PRF - If less than 30 kg N are required then
30 kg N will be applied (time < silking)
6 30/30 PRF - If less than 30 kg N are required then
30 kg N will be applied (time < grain fill)
7 30,0 PRF - If less than 30 kg N are required then
no nitrogen will be applied (time < silking)
! 8 30/0 PRF - If less than 30 kg N are required then
no nitrogen will be applied (time < grain fill)
I
9 15/15 PRF - If less than 15 kg N are required then
30 kg N will be applied (time < silking)
10 15/15 PRF - If less than 15 kg N are required then
30 kg N will be applied (time < grain fill)
11 15/0 PRF - If less than 15 kg N are required then
no nitrogen will be applied (time < silking)
12 15/0 PRF - If less than 15 kg N are required then
no nitrogen will be applied (time < grain fill)
13 PRF - The exact amount of nitrogen required is
applied (time < silking)
14 PRF - The exact amount of nitrogen required is
applied (time < grain fill)

 

 

 

76
The nitrogen strategies used in the inbred study were all evaluated. This included a zero

nitrogen fertilizer, a 180 kg N its-1 applied at planting (180 pp), 180 kg N ha-l sidedressed (180 sd)
with 30 kg N ha'1 at applied planting and an additional 150 kg N ha'1 applied at growth stage V6,
a 90 kg N ha'1 sidedressed (90 sd) with 30 kg N ha’1 at applied planting and an additional 60 kg N
ha1 applied at growth stage V6, and a PRF (Auto N) treatment. The PRF treatment was
accomplished by using the nitrogen factor NFAC as an indicator of nitrogen stress.

The PRF strategies evaluated used the auto-fertilimtion routine mentioned earlier. Once
nitrogen stress was detected, CERES-1M used a nitrogen uptake curve developed from the inbred
study to make a determination of the required amount of nitrogen needed by the plant to
complete its growth without nitrogen stress. A nitrogen application of this required amount was
then simulated. An irrigation water application of 7 mm was also simulated on the day of a
nitrogen fertilizer application. The irrigation application was made to assure that the nitrogen
penetrated the soil surface, since all nitrogen applications were assumed to be sprayed on in the
form of urea-ammonium nitrate (28% N). However, if sufficient rainfall occurred the day of a
nitrogen fertilizer application, then no irrigation was applied.

It was decided that if a nitrogen stress occurred earlier than the end of the juvenile stage,
that a base amount of 50 kg N ha'1 would be applied. This was done based on previous
experience that any nitrogen stress that occurred early in the plants growth would require a
substantial amount of nitrogen. This would also indicate that the of mineralization would not be
very high. The intent of PRF is to make use of the mineralized nitrogen to its fullest extent.
Also, the fertilizer amounts needed late in the PRF strategy may be small, too small for practical
consideration. Therefore, four additional strategies were added.

Using the PRF methodology just mentioned, four more strategies were evaluated that
contained the same assumption that early fertilizer applications would be set to 50 kg N ha".
However, instead of adding what was required by the plant in the later stages, it was decided that
two threshold values would be used. One threshold value was set at 15 kg N ha'l. If, after the

end of the juvenile stage, the amount of nitrogen required was less than 15 kg N ha‘l,

77
then 15 kg N ha'1 would be added. The 15 kg was considered to be the lowest practical amount of

nitrogen fertilizer to apply. Additionally, a strategy was setup that if less than 15 kg ha'1 of
nitrogen was required, then no nitrogen fertilizer would be added. This allowed for a comparison
of whether the 15 kg threshold had any real impact Also, a threshold of 30 kg N ha-1 was also
chosen for evaluation. Both strategies that were used with the 15 kg threshold were used with the
30 kg threshold.

Finally, there was the timing of the late application with respect to plant growth stage.
Late applications of nitrogen fertilizer often have little effect on grain yield because the nitrogen
does not have time to be available for plant uptake before grain filling is started. To evaluate this,
each of the PRF strategies, including the 15 and 30 kg threshold strategies and the PRF-Auto N
strategy, were simulated with two different growth stage related restrictions. One restriction was
that if a nitrogen stress was detected after silking, no nitrogen fertilizer would be applied. The
second restriction was that if nitrogen stress was detected after the beginning of grain filling, no
nitrogen would be added. Growers would generally agree with the second limit of not apply
nitrogen after the beginning of grain filling. The addition of nitrogen after silking is done, though
this strategy is not used widely. With the addition of these conditional strategies, the total number
of strategies simulated was 14 (see Table 10).

All of these strategies were evaluated for their impact on grain yield and nitrate leaching.
Yearly data on irrigation and nitrogen fertilizer applications were also written to an output file to

allow for further mment of these strategies.

RESULTS AND DISCUSSION

Commercial Hybrid Maize

The data used for the hybrid validation/calibration in this research was part of a five year
project to study the effects of irrigation and nitrogen fertilizer application management on hybrid
corn production and nitrate leaching below the rootzone. Drainage lysimeters were installed in
1986 and were used to measure the leaching at a depth of 2.29 m. The years used for this
validation were 1988 and 1989.

The two drainage lysimeters were installed in a farmer’s field in St. Joseph County, MI.
One lysimeter was located under an existing center pivot irrigation system and the farmer managed
the nitrogen fertilizer and irrigation water applications to the plot. This treatment is referred to
as the conventional management practice (CMP). The other lysimeter was located in the same
field but outside of the area irrigated by the center pivot. This lysimeter was under the
management of the researchers at MSU and the treatment used is referred to as the research
management practice (RMP). Both lysimeters received the same management in terms of tillage,
planting, and pesticide control. However, the irrigation and nitrogen management schemes

differed greatly.

Input Data

The input data for the simulation runs includes weather data, soils data, genetic data, and
management data. These data make up the eight input files required by CERES-1M. The

following is a brief discussion of the input files.

78

79
Weather Data

Weather data were collected on site using a LICOR weather station. The station
consisted of a LICOR LI-1200 minimum data set recorder, a LICOR LI-ZOOSA pyranometer
sensor, a LICOR 1000-07 instrument enclosure with an air temperature sensor, and a Sierra Misco
tipping bucket rain gauge. The weather data were recorded on a daily basis and included solar
radiation, maximum and minimum temperature, and rainfall. During the winter months, the solar
and temperature data were used from the weather station on site, but the precipitation data were
taken from a nearby weather station that recorded snowfall. A listing of the weather data used for

the simulation runs is in Appendix three.

Soil Nitrogen Balance Parameters

To simulate nitrogen dynamics, CERES-1M requires information on the addition of fresh
organic matter such as the amount of residue from the previous crop and the depth of
incorporation. Also, the C:N ratio of the residue is required. Finally, the root residue from the
previous crop is needed. The plots used in this study were both in commercial hybrid maize
production for several years before 1988. For the 1988 simulation, the C:N ratio and amount of
residue was estimated based on the grain yield from the previous year. The depth of incorporation
was based on the tillage depth in the spring of 1988. In 1989, the data from the previous year’s
harvest were used and root residue was again estimated. Input File 4 (soil nitrogen balance

parameters) was created with these data. All of the residue data used for the simulations are in

Appendix C.

Soils Data

The soils data were collected at the beginning of each season before fertilizer was applied.
In 1987, soil samples were taken and the bulk density of each plot was determined. These
numbers were used in the 1988 and 1989 simulation runs. On April 15, 1988, three soil samples

were taken from each plot in increments of 15, 15, 30, 60, and 30 cm to a depth of 150 cm. The

80

samples were analyzed for nitrate and ammonium content. The measured values from one sample,
from the RMP plot, were discarded because of its unexplainably high value.

In 1989, on April 11, four samples from each plot were taken and analyzed for nitrate-N
and ammonium-N concentrations. The soil layers were separated as in 1988. Additionally, the
organic carbon percentage of the soil and the soil pH were measured for each soil layer. The
organic carbon and pH data were used in the 1988 simulations as well as in 1989 because they do
not vary greatly from year to year and no measurements were taken in 1988.

There was large variation in the measured NO3' and NH4+ contents among the soil
samples taken. Tables 11 and 12 gives the average values used and the standard deviation (SD)
associated with those measured values for the soil nitrogen components for the CMP and the
RMP plots for 1988. The 1989 data for the CMP and RMP plots are found in Tables 13 and 14.
All four tables show the difficulty in establishing good initial conditions for soil nitrogen. The soil
water inputs, i.e., drained upper limit, lower limit, saturation, were estimated based on soil survey
data and the sand, silt, and clay content of the soil. All of these data were used to create input

File 5 (soil profile initial conditions).

Irrigation and Nitrogen Applications

The irrigation input file was created by using a combination of measured values and
farmer records. The farmer supplied a schedule of when an irrigation took place and the amount
of water applied on the CMP plot. Researchers also kept a record of when irrigation water was
applied and for how long the system ran on the RMP plot. These data were compared against
data recorded by wedge type and tipping bucket rain gauges located in the fields near the
lysimeters. Using the three pieces of data, the amount of irrigation water applied and when that
water was applied was determined.

The nitrogen fertilizer application dates and amounts for the CMP plot were supplied by
the farmer. The preplant applications were made by knifing anhydrous ammonium into the soil at

a depth of 25 cm. The nitrogen applied at planting was in the form of diammonium phosphate

Table 11. Averages of the measured soil nitrogen values and the associated standard deviations

81

for the CMP plot from samples taken on April 15, 1988. Mendon, MI.

Nitrate-N (ppm)

Ammonium-N (ppm)

 

 

 

 

 

 

 

 

DEPTH (cm) AVE so AVE so
015 .905 .429 3.99 1.93
15-30 .743 .323 0.331 0393
30-60 .516 .205 0.305 0.317
.012 0.730 0.425

.060 0.600 0.201

 

Table 12. Averages of the measured soil nitrogen values and the associated standard deviations

 

for the RMP plot from samples taken on April 15, 1988. Mendon, MI.

Nitrate-N (ppm)

   

 

Ammonium-N (ppm)

 

 

 

 

 

 

 

 

 

DEPTH (cm) AVE so AVE so

I 015 .905 .005 2.29 .345

I 15-30 .57 .11 1.12 .690

3060 .545 .175 1.94 .105

60-120 .305 .105 0.67 .115
120150 .535

 

 

 

 

82

Table 13. Averages of the measured soil nitrogen values and the associated standard deviations
for the CMP plot from samples taken on April 11, 1989. Mendon, MI.

 

   

Organic Carbon

 

 

 

 

 

 

 

 

 

 

 

Nitrate-N Ammonium-N
DEPTH (PM) (PM!) (%)

(cm) AVE so AVE so AVE so AVE so
0-15 1.19 .895 2.31 .120 .780 .080 7.45 .250

I 15-30 .810 .180 1.92 .605 .725 .185 6.70 .100
30-60 1.11 .530 2.61 .720 .370 .090 5.70 .2“)

I 60120 1.11 .078 1.39 .890 .375 .115 6.45 .050
I 120-150 1.81 .125 1.57 .225 .320 .020 7.45 .950

__—

 

 

 

 

 

Table 14. Averages of the measured soil nitrogen values and the associated standard deviations
for the RMP plot from samples taken on April 11, 1989. Mention, MI.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Nitrate-N Ammonium-N Organic Carbon-
DEPTH (ppm) (ppm) (%)
(cm) AVE so AVE so AVE so
015 .225 .225 1.99 .330 .600 .001
15-30 .235 .235 1.56 .410 .660 .050
I 30-60 .570 .120 1.43 .335 .365 .045 .
60120 .495 .025 0.47 .465 .190 .010 6.10 .500 I
{-13.0150 .390 .090 0.22 .215 .205 .035 7.25 .145

 

 

 

 

83
(DAP) and was sidedressed with the other fertilizers added at the time. The RMP also received

the DAP application at planting. All additional nitrogen added to the RMP plot was in the form
of ammonium nitrate and applied by hand.

In 1988, the CMP plot received 210 kg N ha'li 200 kg N ha'1 preplant as anhydrous
ammonium and 10 kg N ha'1 at planting as DAP. The RMP plot received a total of 130 kg N ha’
1: 10 kg N ha'1 at planting as DAP and then 2 additional applications of 40 kg N ha'1 and 80 kg N
ha'1 as ammonium nitrate. Additionally, the CMP plot received 347 mm of irrigation water while
the RMP plot received only 267 mm.

In 1989, the nitrogen applied to the CMP plot was the same as in previous year, with 200
kg N ha'1 applied preplant as anhydrous ammonium and 10 kg N ha'1 applied at planting as DAP.
The RMP plot received the 10 kg N ha-1 at planting and then an additional 113.1 kg N M1 as
ammonium nitrate split into 3 applications of 30, 35.6 and 47.5 kg N ha‘l. The total irrigation
water applied to the CMP plot was 131 mm and the RMP plot received 49 mm of irrigation water.
The input data for the irrigation water applications (File 6) and the nitrogen fertilizer applications

(File 7) are in Appendix three.

Management Input Data

The management data required to run CERES-1M are easily obtainable if good field
records are kept. In this study, the grower was given a form to fill out after each field operation.
Phone calls were made to him to verify the data. The plant population data used for the
simulation runs were measured after emergence bemuse CERBSoIM requires the plant population,
not the planting population. Often the planting population is higher than the actual field
population because 100 percent germination does not occur. In 1988, both plots had a plant
population of 6.8 m'z. However, in 1989, wet weather and mechanical problems caused a late
planting for the study and the plant population was not as homogenous. Planting normally takes

place by the first week in May, but this year planting was delayed until May 20. The plant

84
populations for 1989 were 8.3 m'2 for the RMP plot and 7.3 n1'2 for the CMP plot. The difference

in plant populations was due to poor plant germination on the CMP plot.

Genetic Coefficient Information

The genetics coefficients for the hybrid used in this study were created for CERES-1M by
using data from previous years and data from other experiments within the state using the same or
a similar hybrid type. In CERES-1M however, the potential grain number has been changed to a
grain number factor. This parameter was developed by using the 1988 data and adjusting the
value to match the CMP plot harvest data. The same value was then used for the RMP treatment
in 1988 and both the CMP and RMP treatments in 1989. The new leaf area parameter was

determined in the same way, again using the 1988 CMP plot for calibration.

Simulated Versus Measured

The comparisons made between CERES-1M simulated values and those measured in the
field for this study include year end plant biomass and yield data, soil water drainage and nitrate
loss below the plant rootzone. In the CMP and RMP treatments, for 1988 and 1989, no
replications were made. However, some measurements were made on several plants within the
plot areas.

Both plots were approximately 8 m X 10 m. In both years, harvest data were collected by
harvesting approximately 2.1 1112 (2.63 m of a row with 0.80 m spacing). Over an area of
approximately 80 m2, each treatment was split into 36 sub-plots, with grain data collected on all.
Additionally, four sub-plots were chosen at random and total above ground biomass samples
taken. The plant nitrogen data were obtained by using biomass and grain samples from two of the
four sub-plots to determine total above ground biomass. As with the soil analysis, some variation
existed in the measured data. Table 15 gives the averages and the SDs for the yield and stover for

the CMP and RMP plots for 1988 and 1989.

85

Table 15. Yield and stover data measured for the CMP and RMP plots for 1988 and 1989.
Mendon, MI.

 

Yield (kg ha‘l) Stover (kg ha-l)

 

P10, 1933 1939 1933 1939

 

AVE SD AVE SD AVE SD AVE SD

 

l CMP 10848 1170 11095 1667 9370 620.45 10546 834.24

 

 

 

 

 

 

 

 

 

fl RMP 9396 1410 10882 1579 8920 595.32 7853 811.82

 

Plant Biomass and Yield Data

The weather during the 1988 growing season was not typical. There was widespread
drought throughout much of the US, causing crop damage and yield losses. Fortunately, both
plots in this study were irrigated.

As seen in Table 16, there was good agreement between the measured values and those
simulated by CERES-1M for the CMP plot for 1988. The grain yield and kernel data showed
excellent agreement. There was a discrepancy in the amount of stover produwd and the grain
nitrogen content. However, the overall nitrogen uptake data compared well. For the RMP plot,
Table 17, the comparisons were good for all of the data. Yield, kernel, and nitrogen data showed
good agreement between the measured and simulated values.

The 1989 season saw a return to more traditional weather patterns. Again, using the data
collected, the input files for CERES-1M were created. A comparison of the year end yield and
biomass data are given in Tables 18 and 19, for the CMP plot and the RMP plot.

Results of the simulation for the CMP plot for 1989 again showed good agreement with
the measured values (Table 18). The yield and kernel data comparisons showed well. As in 1988,
CERES-1M simulated the percent nitrogen content higher than that measured resulting in the
grain nitrogen uptake being overestimated as well. However, the simulated total nitrogen uptake

again compared well with the measured values.

86

Table 16. A comparison of several parameters between CERES-1M simulated values and
measured values for the CMP treatment. Mendon, MI, 1988.

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter CERES-1M Measured Difference
Silking Date 201 200 + 1
Grain Yield (kg ha-l) 10112 10343 . 6.3 96
Kernel Weight (g) 0.279 0.236 - 2.4 96
Grains nr2 3063 3279 - 6.6 %
Grains car1 452 434 - 6.6 %
Maximum LAI 4.63 5.06 - 35 %
Biomass (kg hat-1) 19700 13762 + 4.5 96
Stover (kg ha-l) 11155 9370 +190 96
Grain Nitrogen % 1.71 1.39 +23.0 %
Tot N Uptake (kg ha-l) 219.7 2037 + 5.3 96
Stover N Uptk (kg ha-l) 73.2 73.2 - 6.4 96
Grain N Uptk (kg ha“ 146.5 130.5 +123 96

 

 

 

 

 

 

Table 17. A comparison of several parameters between CERES-[M simulated values and
measured values for the RMP treatment. Mendon, MI, 1988.

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter CERES-1M Measured Difference
Silking Date 201 202 -1
Grain Yield (kg ha“) 10014 9396 + 6.5 96
Kernel Weight (g) 0.279 0.259 - 7.7 %
Grains m-2 3033 3139 - 3.3 %
Grains ear1 447 463 - 3.5 96
Maximum LAI 4.26 3.81 +ll.8 %
Biomass (kg ha") 15334 17055 - 6.9 96
Stover (kg ha-l) 7423 3920 -133 96
Grain Nitrogen % 1.54 1.39 + 9.7 %
Tot N Uptake (kg ha") 163.5 1433 + 9.9 96
Stover N Uptk (kg hit-1) 33.4 35.7 + 6.4 96
Grain N Uptk (kg ha'” 130.1 113.1 +150 96

 

 

 

 

 

 

Table 18. A comparison of several parameters between CERES-1M simulated values and

87

measured values for the CMP treatment. Mendon, MI, 1989.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter CERES-IM Measured Difference
Silking Date 215 216 - 1
Grain Yield (kg ha-l) 10714 11095 . 3.4 %
Kernel Weight (g) 0.251 0.262 - 4.2 %
Grains nr2 3605 3535 + 0.6 %
Grains ear-l 495 491 + 0.3 %
Maximum LAI 4.34 4.33 +11.3 %
Biomass (kg ha-l) 13969 19993 - 5.1 %
Stover (kg ha!) 9916 10546 - 6.0 %
Grain Nitrogen % 1.65 1.40 +17.9 %
Tot N Uptake (kg ha-l) 2127 194.0 + 9.6 %
Stover N Uptk (kg ha” 63.0 63.4 - 0.6 %
Grain N Uptk (kg ha” 149.7 131.3 + 14.0 96

 

 

 

 

Table 19.
measured

A comparison of several parameters between CERES-IM simulated values and those

in the RMP treatment. Mendon, MI, 1989.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter CERES-[M Measured Difference

Silking Date 214 214 0

Grain Yield (kg ha-l) 10073 10332 - 7.4 %
Kernel Weight (g) 0.263 0.239 +10.0 %
Grains m2 3233 3343 -15.7 96
Grains car1 391 463 ~15.6 %
Maximum LAI 5.03 4.13 +215 96
Biomass (kg ha“) 16500 17043 - 3.2 96
Stover (kg ha") 7939 7353 + 1.7 96
Grain Nitrogen % 1.36 1.26 + 7.9 %
Tot N Uptake (kg ha-l) 143.9 155.2 - 4.1 96
Stover N Uptk (kg ha” 33.0 39.3 -16.0 96
Grain N Uptk (kg ha“) 115.9 115.9 - 0.0 96

 

 

 

 

 

 

88
The simulation for the RMP plot for 1989 also showed good agreement with the

measured yield and nitrogen components (Table 19). However, individual kernel data was under
predicted by the simulation.

Another area of interest in these simulations was the leaf area index (LAI). With the
introduction of a new leaf area formulation, analyzing how well CERESoIM descn’bed the leaf area
of the plants in the treatments is appropriate. Throughout the growing season 15 plants were
measured four times in 1988 and five times during the 1989 season.

In 1988, CERES-1M showed good agreement in simulating LA] for both the CMP and the
RMP treatments, Figures 24 and 25. Figure 24 shows the LAI simulated by CERES-1M and the
measured values and the associated SDs for the plants in the CMP treatment. CERES-1M
simulated LAI well for the first two measurements but underestimated LAI for the third and
fourth measurement. This resulted in an underestimation of the maximum LAI, as shown
previously in Table 16. In the RMP treatment, Figure 25, CERES-1M again did a good job of
simulating LAI. However, CERES-1M under-predicted LAI during the early vegetative phase.
This may be due to the water stress that was imposed on the plants in the RMP plot. This stress
may have caused some variations in individual plant growth.

In 1989, CERES-1M showed good agreement with the measured LAI, Figures 26 and 27.
In the CMP treatment, Figure 26, CERES-1M showed close agreement with the LAI measured. In
the RMP treatment, Figure 27, CERES-[M showed good agreement with a slight underestimation
during the middle of the vegetative growth stage similar to that seen in 1988, Figure 25.

The results shown here illustrate the ability of the model to simulate the biomass and
nitrogen components of hybrid maize. Though some deviations exist, the overall results compare
well giving validity to the model’s phasic and developmental capabilities. However, since this

study involves nitrate leaching, a comparison of those values is also warranted.

 

 

 

 

 

6.0
- — CERES-m .

“1‘ j a MEASURED 1
I 5.0- -
E . .
o d
N 4.0; 3
E . .
o . .
V ‘ d
a" 3.01 -:
U
E . .
o 2.0— -t
q; . 4
L
< j I
“5 1.0- -1
Q) ‘ a
_J 4

0.0 . . 1 3 r . , . .

4/29/33 5/19 6/8 6/23 7/13 3/7 8/23 9/16

Date.

 

 

 

Figure 24. Measured and CERES-1M simulated leaf area index values for the
CMP treatment for 1988. Mendon, MI.

 

 

 

 

 

5.0
« — CERES—m .
g," . a MEASURED .
I ~ .
E 4.0- ~
0 1 1
N ‘ '
1 I

8 3.0- .
V d 4
a: ‘ 1
'o . s
s 2.0— I _
o ‘ ‘
m ‘
a . -
,_ 1.0— -
O '4 1
Q)
A 1

0.0 . . a - - 4 1 . .

4/29/33 5719 6/3 6/23 7/13 3/7 3/28 9/16

Date

 

 

 

Figure 25. Measured and CERES-1M simulated LAI values for the RMP
treatment for 1988. Mendon, MI.

 

 

‘ — CERES—IN
at MEASURED

A
A l A l J J4 l A I A

J A I

Leaf Area Index (cm2 cm—2)

d
O
1.
11.4;

 

 

 

0.0 . . . a - . .
5/19/88 (5)3 6/28 7/13 877 3/27 9/16 10/6

Date

 

 

 

Figure 26. Measured and CERES-1M simulated leaf area index values for the
CMP treatment for 1989. Mendon, MI.

 

 

. — CERES-1M i
a MEASURED 1

Leaf Area Index (cm2 cm‘z)
1 a 1 L4 LLLL a

 

0.0— . . T . . 3
5/19/33 6/8 6/23 7/18 ab 8/27 9/16 10/6

Date

 

 

 

 

 

 

Figure 27. Measured and CERES-1M simulated leaf area index values for RMP
treatment for 1989. Mendon, MI.

91

Drainage and Nitrate Leaching

Nitrate leaching was measured in both plots using non-weighing lysimeters. Samples were
taken approximately every two weeks, or as warranted by drainage events. Drainage amounts were
measured and samples were analyzed for nitrate-N concentrations. Although the lysimeters were
installed in 1986, the drainage data were not collected and studied intensively until 1988.

Collection of the drainage data for the 1988 season began on April 10th. With the onset
of the summer drought, the drainage from the RMP lysimeter was low compared to the CMP
lysimeter. At summer’s end, heavy autumn rain caused excessive drainage. Relatively high
drainage amounts were recorded over most of the winter, except for periods of extreme cold when
the soil would freeze. The CERES-1M model was run for the entire length of the drainage record,
from April 10, 1988 to April 10, 1989. Figures 28 and 29 show the drainage amount for the CMP
and RMP lysimeters along with the CERES-[M simulated values and the precipitation. As seen in
Figures 28 and 29, CERES-1M performed well in simulating soil water drainage. Though the
timing of the drainage events do not always coincide, CERES-1M did a good job of mimicking the
general curve described by the measured data.

The measured nitrate accumulated for the drainage water sampled from the CMP and the
RMP lysimeters for 1988 is shown in Figures 30 and 31, along with the CERES-IM simulated
values and the precipitation. Figure 30, the CMP lysimeter, shows one set of measured data and
two CERES-[M runs. The first CERES-1M run is that with the original fertilizer input of 200 kg
N ha'1 applied preplant in the form of anhydrous ammonium. This curve underestimates nitrate
loss by more than 50%. After discussions with the farmer, it was concluded that an over-
application of nitrogen was possible and highly likely. Also, because the drainage component
compared so well, it was decided that a new fertilizer input would be used and CERES-1M was
rerun to test the hypothesis that in 1988, the amount of fertilizer applied as anhydrous ammonium
was greater than 200 kg N ha’1 reported by the farmer. A value of 300 kg N ha’1 was used and

the result is the second CERES-1M line shown in Figure 30. At 300 kg N ha-l, CERES-1M

 

‘l 400

 

, — CERES-IM

H Measured r
1200‘ -- Precip '- "

1000-3
800‘

600 -

Drainage (mm)

 

 

 

0 fl ' I Y I
4/10/33 7R1 10/27 2/3/39 5/14

Date

 

Y

 

 

 

Figure 28. Measured and CERES-1M simulated drainage for the CMP
lysimeter for 1988-89. Precipitation is also shown. Mendon, MI.

 

 

 

 

 

 

 

1400
, — CERES-1M
H Measured
1200' -- Precip '7 "‘
A 1000- -
v 800- _
o . .
8
C 600- -
'6 J
L
O 400- .
200- -
0 A”d{:‘ "— _- . , . I ,
4/10/33 7/19 10/27 2/5/39 5/14
Date

 

 

 

Figure 29. Measured and CERES-M simulated drainage values for the RMP
lysimeter for 1988-89. Precipitation is also shown. Mendon, MI.

 

 

180
4 — CERES-IM (200) . .
150.. H Measured

T -- CERES-IM (300)

NOT

 

120‘

q

lOOn
80-
60-

.4

40-1

Nitrate Leached (kg N ha")

20--
J

 

 

 

Y

j
2/3/89 5 14

fi'

0 " fl 1
4/10/33 7/19 10/27
Date

 

 

 

Flgure 30. Measured and CERES-1M simulated (for both 200 and 300 kg N
ha'1 pp) nitrate leaching values for the CMP lysimeter for 1988-89. Mention,
MI.

 

 

 

 

 

80
'i —' CERES-1M

71‘ 70‘ H Measured ..
2 N 1

50-1
z . T
S 50" .4
V 1 q
8 40- -
3 . .
0 30-1 ..
Q)
_l i .
3 20" q
C 4
L
32: 10: q

0 -. . l v , .
4/ 1 0/ 88 7/19 10/27 2/5/39 5/14
Date

 

 

 

Figure 31. Measured and CERES-[M simulated nitrate leaching values for the
RMP lysimeter for 1988-89. Mendon, MI.

94
showed excellent agreement with the measured values. Additionally, the phasic and developmental
outputs that were shown in Table 16 changed little. Table 20 shows the comparison between the
measured values and CERES-1M with a fertilizer input of 300 kg N ha‘1 instead of the 200 kg N
ha'1 as reported by the farmer. As seen in Table 20, the stover nitrogen uptake is increased
slightly by 6.7 kg N ha'1 and the grain nitrogen content increased by 0.02% causing an increase in
the grain nitrogen uptake of 0.3 kg N ha'l. All of the other values remained the same. Figure 30,
however, shows excellent agreement between the measured values and those simulated by CERES-
IM when the 300 kg ha'1 input is used.

The RMP nitrate results are shown in Figure 31. Here, CERES-1M showed a bias
throughout most of the simulation period. This may be due to the difficulty in determining the
day to begin accumulation of the drainage values simulated by CERES-1M should begin. In the
measured data, there seemed to be more lag time between rainfall and drainage than is simulated

Table 20. A comparison of several parameters between CERES-[M simulated (using
300 kg N ha'1 input) and measured values. Mendon, MI, 1988.

 

 

 

 

 

 

 

 

 

 

 

 

 

Parameter CERES-1M Measured Difference
Silking Date 201 200 + 1
Grain Yield (kg ha") 10112 10343 - 6.3 96
Kernel Weight (g) 0.279 0.286 - 2.4 %
Grains in2 3063 3279 - 6.6 %
Grains car1 452 484 - 6.6 %
Maximum LAI 4.63 5.06 - 8.5 %
Biomass (kg ha-l) 19723 18762 + 5.1 96
Stover (kg ha") 11133 9370 +193 96
Grain Nitrogen % 1.73 1.39 +245 %
Tot N Uptake (kg ha") 227.4 203.7 + 39 96
Stover N Uptk (kg ha‘l) 79.9 73.2 + 22 96
Grain N Uptk (kg ha” 147.5 130.5 +130 96

 

 

 

 

 

 

95
by CERES-1M. However, even with this bias, CERES-1M simulated the measured accumulated

nitrate leaching loss well showing the pulses and plateaus measured in the field.

In 1989, the drainage comparisons between the measured and the CERES-TM simulated
values (Figures 32 and 34) again showed good agreement for the CMP and RMP lysimeters. As in
1988, there were discrepancies between the measured and simulated values as to when drainage
occurred. This is especially true for the RMP lysimeter, Figure 33, during the winter months.
However, the total drainage amounts for both lysimeters compared well.

The nitrate leaching data measured in the CMP and RMP Mimeters and those values
simulated by CERES-1M for 1989 are shown in Figures 34 and 35. In both cases, CERES-1M
slightly overestimated the total amount of nitrate leached by approximately 10 kg N ha-l. Figure
34, the CMP lysimeter, shows that CERES-1M overestimated the nitrate loss throughout the
simulation period. CERES-1M showed more nitrate loss occurring during the summer and less
during the fall than that measured. However, by the end of the simulation period, the total nitrate
loss compared well.

Figure 35 shows the simulated and the measured nitrate leached for the RMP lysimeter.
CERES-1M slightly underestimated nitrate loss through most of the growing season and into
autumn. However, both the measured and the simulated curves show a sharp increase during the
winter months and the values of the measured and predicted total nitrate loss again compared
well.

These results have shown that CERES-1M can reasonably simulate drainage and nitrate
leaching in hybrid maize production. Discrepancies between the measured and simulated values of
when drainage actually occurs does not seem to have a large impact on the models ability to
simulate nitrate leaching. The next step is to compare CERES-1M with inbred maize in seed corn

production.

 

Drainage (mm)

 

1200

 

1100:
1000:
900i
3003

7004

600%
500«
4003
300%
2003
1001

0:

 

4/10/89

— CERES-IM
H Measured
—- Precip

 

lalalnlhlaiJ

LLILIA

AIJLA

 

7/19 Y

 

V

10}27 2/5790 5/14

Date

 

Figure 32. Measured and CERES-1M simulated drainage values for the CMP
lysimeter for 1989-90. Precipitation is also shown. Mendon, MI.

 

Drainage (mm)

1000

 

900~
800;
700:
600;
sool
4669
300;
zool

1
100-

4

< — CERES-IN

H Measured
-- Precip

 

 

/

- --- A — -

I

 

Ll‘lllLlALh

 

 

V

7/19 10}27 2/5790

Date

0
4/10/89 5/14

 

 

 

Figure 33. Measured and CERES-1M simulated drainage values for the RMP
lysimeter for 1989-90. Precipitation is also shown. Mendon, MI.

 

 

 

 

 

140q — CERES-IM _

9 ‘ H Measured ‘
l

o 120- _
z 1
Z 100‘ .4

U1

x - .
v

.0 80- ..

°’ . i
'6

O 60- -

to . ,
_J

B 40‘1 —i

O ‘ ..

\—

f: 20~ ..
Z d
O . I r , . 1 .
4/10/89 7/19 10/27 2/3/90 5/14

Date

 

 

 

Figure 34. Measured and CERES-IM simulated nitrate leaching values for the
CMP lysimeter for 1989-90. Mendon, MI.

 

 

60
— CERES-IM
55 - -

H Measured

50-
45-
40-
CBS-i
30-
25-
20-
15-
10-
5..

 

Nitrate Leached (kg N lea-1)

 

 

 

 

r“

10}27 ' 2/5790 ' 5/14
Date

0 == .
4/10/89 7/19

 

 

 

Figure 35. Measured and CERES-1M simulated nitrate leaching values for the
RMP lysimeter for 1989-90. Mendon, MI.

Inbred Maize

The inbred phase of this research was part of a joint research project between Michigan
State University and Pioneer Hi-Bred, Int., Inc. Begun in 1988, the project’s focus was to evaluate
the impact of nitrogen fertilizer application management on grain yield and the leaching of nitrate
below the rootzone in seed corn production. Using drainage lysimeters, soil-water drainage was
measured at a depth of 2.29 m below the surface and analyzed for nitrate content. Using three
inbred varieties, the nitrogen fertilizer treatments evaluated ranged from a conventional strategy of
180 kg N ha'1 applied at planting to 90 kg N ha'1 sidedressed to no nitrogen fertilizer added.
Using the data collected, input files for the CERES-1M simulation model were created and

simulation runs were made for each inbred and treatment for both 1989 and 1990.

Plant Resmnse Fertilization

As previously mentioned, one treatment for the inbred study was a research nitrogen
management treatment mlled plant response fertilization, or PRF. The basic premise of PRF is to
develop a methodology that once a nitrogen stress is detected, only the required amount of
nitrogen needed to sustain the plants until physiological maturity is applied. During the first
season, 1989, 30 kg N ha'1 were applied at planting for this treatment. This was done to assure
adequate plant growth, since there were no records of yield or growth for this field. Additionally,
50 kg N ha‘1 were applied later that season because of a color change observed by the researchers.
The plants in the PRF treatment appeared lighter green in their leaf color than the plants in the
other treatments. In the second year, a more scientific approach was used.

In 1990, two methods were used to determine nitrogen stress in the plant. One was
comparing the height of well fertilized plants with the height of the plants in the PRF treatment.
Also, a multi-spectrum radiometer (MSR) was used to measure the reflectance of the plants.
Again, the reflectance of the plants in the PRF treatment was compared with well fertilized plants.
Using multiple light ranges, the MSR can measure the reflectance in the blue and red ranges of

light, allowing for a determination of the 'greenness' of the plant. The theory was that plants

99

experiencing nitrogen stress would be less green than non-stressed plants. Unfortunately, it was
discovered that this methodology did not work well with the inbred plants. There was a problem
with incomplete ground cover. Due to the relatively low plant populations (5.5 pl m'z), the
reflectance of the soil surface was often measured and caused erroneous readings. The onset of
nitrogen stress often begins with a yellowing of the bottom leaves, which are often shielded from
the MRS by the upper leaves of the plants. In the end, it was the plant height measurements that
best determined nitrogen stress. In only inbred 1 was a stress detected. Using nitrogen uptake
data from the previous year, a determination was made on the amount of nitrogen the plants
required to complete their growth. On July 23, an additional 40 kg N ha'1 was applied to the PRF
treatment plots with inbred 1. The other inbreeds never showed any significant difference in plant

height comparisons with the 180 kg ha‘1 treatment.

Input Data

The input data required to simulate inbred maize are similar to that for hybrid maize.
The same 8 input files are used and the information within the files is the same except for the few
additional parameters required for inbred simulation, i.e., male/female plant designation,
detasseling date, etc. The following is a brief discussion on how the information required for

these inputs files was obtained.

Weather Data

The weather data for this study were collected on site using 3 Campbell Scientific 1200
weather station. The station included a CR10 datalogger that was programmed to record hourly
data. The instruments used to collect the weather data included 3 1.12008 LICOR pyranometer
(solar radiation), 3 LI-207 temperature and relative humidity probe (maximum and minimum air
temperature), and a Weathertronies 0.01 in. tipping bucket rain gage (rainfall). Other
meteorological measurements taken included relative humidity and wind speed and direction.

Also, the CR10 was used in conjunction with other peripherals to control the drainage sampling

100

system for the five lysimeters. Weather data were collected year round. During the winter
months, precipitation data were obtained from a nearby weather station that recorded snowfall. A

listing of the weather data is in Appendix D.

Soil Nitrogen Balance Parameters

Information on the incorporation of fresh organic matter into the soil profile is required
by the CERES-1M program. These data include organic residue of previous crop, depth of
incorporation, C:N ratio of organic residue, and root residue of previous crop. In 1989, estimates
for these values were made based on observations made at the time of the installation of the
lysimeters. Also, a new input was added in to allow for the user to enter the C:N ratio of the root
residue. This was done because CERES-1M assumes the C:N ratio for the roots to be 40, a
common value used for maize roots. Because the plot area was an old alfalfa field, with some
alfalfa and pasture grasses growing, the C:N ratio of the roots was probably lower. The new input
allowed for the C:N ratio of the root residue to be entered.

In 1990, data from the previous years crop were used as the inputs for these parameters in
conjunction with the inputs used for 1989. The crop residue entered was the stover biomass from
1989. For the root residue, it was assumed that one-third of the root residue fi'om the crop in
1988 was still present in 1990. All of the residue data used in the 1989 and 1990 simulations are

listed in Appendix o.

Soils Data

To obtained the necessary inputs for soil nitrate and ammonium initial conditions, soil
samples were collected at the beginning of each season. Data for organic carbon, bulk density,
and pH were taken when the lysimeters were installed and used for both the 1989 and 1990
simulations.

In 1989, soil samples were taken from each lysimeter area. The sampling consisted of 2

samples for every 15 centimeters of depth, down to a depth of 120 cm. Each sample was

101

comprised of 4 sub samples that were mixed together. The measured ppm values for nitrate-N
and ammonium-N and the related standard deviations (SD) are found in Table 21. The data
shows considerable variation.

In 1990, the soil sampling was simplified and the soil was analyzed every 25 centimeters
down to a depth of 200 cm. Single samples, comprised of 10 subsamples were analyzed for the
top 75 cm on two treatments (180pp and 905). The other two treatments (PRF and 0) had four
samples analyzed, each comprised of 4 sub-samples, for the top 75 cm. Additionally, only 2
treatments (PRF and 0) were sampled beyond the 75 cm depth. Two samples were taken down to
a depth of 200 cm, each comprised of four sub-samples. This new soil sampling procedure was
done due to the time and cost of analysis for soil samples and because deep soil analysis yielded
very little difference between treatments in 1989. The averages and SDs of the measured
concentration of nitrate-N and ammonium-N for the 1990 soil analysis are shown in Tables 22 and
23. The data from 0-75 cm for each treatment is in Table 22. Table 23 contains the soil nitrogen
values from 75 - 200 cm that were used for all treatments in 1990.

The soil water parameters used for the simulation were estimated based on soil survey
data and mechanical analysis of the soil. All of the soil input data used for the simulations are
listed in Appendix D. This information was used to create the soil profile initial conditions (soil-

water and bulk density - input File 2; soil nitrogen and soil pH - input File 5).

Irrigation and Nitrogen Applications

The irrigation applications were measured using wedge type rain gages. Four gages were
placed around each lysimeter and measurements were taken after each irrigation event. All
nitrogen applications were done by hand except for the nitrogen applied at planting, which was
applied mechanically. The amount of irrigation water applied (File 6) and the nitrogen fertilizer

applied (File 7) for each treatment are listed in Appendix four.

102

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

83 w: 8.8 3.8 8.9 NS 83 3.8 83 8.3 83 8.8 8.3 83 83 88.8 as __
8.83 8.8 83 wed 83 8.8 83 8.8 83 a: 83 8.8 8.9 88.8 :3 2.8 we _—
83 eed 83 etc 83 82 83 8.8 83 _E 83 :._ 8.9 :8 :3 2: 8e
83 e3. :3 88 8.83 83. :3 end 83 :8 :3 wed 8.3 38 :3 we; we
:3 9.8 83 8.3 83 8.8 83 e3 83 8.8 :3 83 8.3 8.8 83 3.8 8
:3 $8 83 28 8.: 8.8 83 :8 83 88.8 83 $8 83 8.8 83 8.8 a.
8.8 83 83 2.8 83 68.8 83 8.8 :3 82 83 Nam 83 83 83 m: on
83 3.8 83 8.: 83 a: 83 3.2 8.3 $8 83 ee.e :3 8.8 83 :8— 2
one“: 9.53 855V 958 8:83 8:9: 9&3 8:9:
nmz .602 +st .802 2:2 .602 .82 .802
8.3 68362 .883 68362 83 68362 883 68262. 83
.3 cm: Emma

mow“

 

moo

 

my:

 

 

 

86::qu anew 5.. 82.3 cowaEZ =om

 

 

42 853288 .32 .v .82 ea cos—S 835$ Eat 8:253: .38 2: 8..
=9. 05 5 5282850 2-53583.“ 23 2-28:: .5. coo—S 3380382: .8 53 82333 282.86. 8886.88 05 was 8w80>< 4N ozfl.

103

 

 

 

 

 

 

 

 

 

83 8.. 83 8.. 8.. a. 83 8.. 83 8.. 3 8.8 8
83 2 83 8.8 8.. 8.8 83 8.. 83 .8 8.8 3. on
83 .8 83 8.8 .8 3. 83 8.8 83 3. 8.8 e... 88
838 8:83 838 8:33 833 8:83 838 838
+882 .802 +52 .802 4.32 .802 48:2 .802 83
63 owfio>< $8.92 53 owes}. 38.634. IFme
6 8 my: .8 8.

 

 

 

 

 

3258.8. scam .8 82.3 :omoEZ =8

 

 

42 682.828 dee— .nm :8? _.o :83. 8358 Ea.“ 35.58.. .58 on. .8 £38. Eu n88 :8
on. E 8:838:00 2-83.5858 88 2-88:: .8 :38 £556.38... .8 53 88.85.. 6.88.88 8.8688 2.. 8:. 88.638 .NN 28....

 

104

Table 23. Average and the associated standard deviation for mineral soil nitrogen measurements
from samples taken on April 25, 1990. Constantine, MI.

 

 

 

 

 

 

 

 

 

 

Soil Nitrogen Values
for All Treatments
DEPTH Average (SD)
(cm) NO - NH +
3 4
(Win) (Win)
100 0.86 (.30) 0.59 (.10)
125 0.82 (.11) 0.53 (.11)
150 1.12 (.34) 0.50 (.07)
175 1.41 (.46) 0.66 (.29)
200 1.67 (.60) 0.65 (.29)

 

 

Management Input Data

Management information and field operation records were well kept by personnel at the
site. Dates of planting. detasseling, and the removal of the male plants were all recorded.
Additionally, field notes were taken as to the condition of the plots throughout the season. Plant
populations were obtained by planting at a high rate and then thinning the plots to the desired
population of 5.5 pl m‘z. All of the management information used to create the management

input files (File 8), is listed in Appendix four.

Genetic Coefficients

Since no previous data existed concerning the phasic development of the inbred lines, the
genetic coefficients needed to be established. Using data gathered from treatment 1 in 1989 (180
kg N ha‘1 at planting), the genetic coefficients were adjusted to give reasonable results. This was
done for each of the inbreeds. The same genetic coefficients were then used for the rest of the
treatments in 1989 and all the treatments in 1990. The listing of the genetic parameters used to

create the genetic input file (File 9) is in Appendix four.

105

Simulated versus Measured

There was variation in the biomass and yield measurements taken in the inbred study.
The year end data used for comparison with the values simulated by CERESJM are a result of
two separate sampling procedures. The yield data given are part of a harvest sampling taken at
physiological maturity. The data values used for comparison are averages of eight samples, two
samples from each replication within a treatment. The samples were about 1/1235 of a hectare
(1/500 of an acre) in size. From this data set, kernel weight, kernel number, and grain yield were
determined.

The stover and nitrogen data were obtained from a biomass sampling that took place four
times throughout the season. Six plants from each replication were removed and dry weight and
nitrogen contents were determined. The biomass samples were divided in leaf and stem, cob, and
grain. These samples were taken at growth stage V6, silking, 50 % grain filling, and at
physiological maturity. These samples were not part of the harvest sample previously mentioned.
The stover and cob weights measured in these biomass samples were added to the grain yield data
from the harvest sample to obtain above ground biomass. Additionally, the nitrogen
concentrations obtained for the grain in the last biomass sample was used for grain N (%)
comparisons. In turn, this grain N (%) was multiplied by the grain data from the harvest sample
to obtain grain N uptake. Finally, the percent nitrogen in the cob and in the stover (leaves and
stems) was averaged (weighted) to obtain the stover N percentage. Again, this in turn was
multiplied by the total stover and cob weight to obtain a value for the nitrogen uptake by the
stover.

The averages and the associated SDs for the key parameters of the simulation (i.e., grain
yield, kernel number, stover, % N grain, % N stover) for 1989 and 1990, taken at physiological
maturity, are shown in Tables 24 and 25. These average values are used later to compare against
CERES-1M simulated values.

Taking the data from Tables 24 and 25, a third table, Table 26 was created that gives the

coefficient of variation (CV) for the various parameters analyzed. The CV is the standard

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.9 8.9 8.9 8.9 889 889 889 8.9 8.9 8.9 889 8.9

8... 88.8 8.. 8.8 88.. 88.. 8... 8.8 8.8 88.8 88.8 8.8 5.8%..“ 88.
889 8.9 889 8.9 889 8.9 889 8.9 889 8.9 8.9 8.9

88.. 8.... a... 8.... 88.. 88.. 88.. .8.. 88.. 8.... 88.. 88.. 58%....” 88
8.888. 888. 8.888. 8.888. 28.. 888.. 8.888. 8.888. 8.88... :88. 8.88. 8.88... A 8
88.8 89.8 8888 88.8 8888 3.8 :88 88.88 888.. 888.. 88.. 88.. ”was“.
88.. 88.. 88.. 88.. 88.. 8.8.. 88.. 88.. 88.. 88.. c 8.. 88.. 8

888. 88. 888. 8.8. 888. .88. 888. 888. 888. 888. 888. 888. .3 ”8....3.
8.888. 8.8... 8.88... 8.888. 8.888. 8.8. 8. 8.88. 8.888. 8.88. 8.888. 8.8. .. 8.8 _ 8. 8
8.88 888 888 888 888 888 8888 838 8888 888 $88 .888 fly»...
8. 8 my... 88. 88. 8 my... 88. 8. 8 ".8... 88. .2822...

Em. - owa.o>< am. - owflo>< Em. . om8.o>< 30:23....
8 828... 8 828... . 828:.

 

 

 

.2 .o=_.=88=00 .33 .8

8.8.8:. .8385... .8 :38 9:082:88... comes... 2...... 8.8 88888.. .88.. .8 Em. 2.8.8.8.. 8.88.8.8 3.88888 2.. 8:8 898.03. .vw 038...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

889 889 889 88.8. 889 889 889 889 889 889 889 8.9

88.. .8.. 88.8 8.8 8... 88.. 88.. 88.. 8... 88.. 88.. 88.. 8N3.“ 8..
8.9 :89 889 889 8.9 889 8.9 8.8. 2.9 8... 889 8.8.

88.. 8.... 8.... 88.. 8.... 88.. 88.. .8. 88.. 8.. 88.. 88.. 58%..w 88
8.88... 8.8.8. 8.8... 888.. 888.. 8.88... 8.88. 8.888. 8.8... 8.88. 8.888. 2.889 A 8 .
8888 8.88 :88 8888 888.. 8888 888 88.8 88.8 8888 8.88 888 fine“
88.. 88.. 88.. 88.. 88.. .88.. 8.8.. 8.8.. 8.8.. 8.8.. 288.. 88.. 8

3.8. 888. 888. 3.8. 888. 888. 888. 888. .88. .88. 88.. .88. 83 ”are.
8.888. 8.888. 8.888. 8.888. 2.88. 8.888. 8.88... 8.889 8.888. 2.888. 8.88. 8.8.8. 8
88... .88.. 888 8888 8888 88.8 888 88.8 8888 8.8 88.. 838 buy»...

8 88 "8. 888. 8 .8 "8. 8888. 8 m8 "8.. 8888. .2852...
am. - ow~.o>< am. . 088.34. Em. - 088.03.. 808.3...
8 828... 8 828... . 828...

 

 

 

 

AZ 65.—.8850 .83 .8

8.38... .8888883 .8 88.8 $580.33... 8080...: Eu... 8... .8880... .80.» .8 6m. 888.882. 8.8888... 3.80888 8.. 8... 888.03.. 8.8 038...

 

108

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88.88 88.8. 88.8. 88.8. 88.8 .88. 88.88 88.8. 33.8 5888.2 88
88.8 88.8 88.8 8.8 88.8 88.8 88.8 8.8 520 5888.2 88
88.8 88.8. 88.8 8.8. 88.8 88.8. 8.... 88.8. 8.-.... 8... 896.8
88.8 88.8 8.8 88.8 88.8 88.8 88.8 8.8 .8. .85 .858.
88.8 88.8 88.8. 88.8 88.8 88.8 88.8 .88 .78.. 8... 22>
888. 888. 888. 888. 888. 888. 888. 888.
8225:88—
>u >0 >0
8883.8 8 828... 8 828... . 8288. L

 

 

 

.2 88888.88 .888. 888 888.

8.. 5.8388 383228.38 .8 88.8. 5:052:88... coach... 2...... .858 88882.. 6.0.5 8.. ems—.882. 8 88 388.98 82.8.8.3 «o 20.8an .3 058.8.

109
deviation divided by the average and multiplied by 100. The CVs of the measured data for each

inbred and treatment for 1989 and 1990 are shown in table 26.

The data from 1989 in Table 26, show that the variation in grain yield is fairly small, with
the CVs averaging 6.70 %. The lowest CV was inbred l, PRF treatment, with a CV of 1.7 %.
The highest was inbred 2, 1805 treatment, with a CV of slightly over 14.2 %. The kernel weight
data all looked good and showed little variation. The stover data, however, showed quite of
variation within replications with an average CV of 13.82 % (Table 26).

Tables 24, 25, and 26 also contain data for the percent nitrogen in the grain and in the
stover. This follows a similar patter to the yield and stover data. The average CV of the percent
nitrogen in the grain was 6.70 %. However, the CV increased in the percent nitrogen in the
stover to 13.9%.

The CVs of the measured values for the 1990 season followed the same trends as seen in
1989 (Table 26). However, the CVs increased in 1990 for each parameter except stover. The
reason for the increase in CVs is not known, except that the low nitrogen treatments may have
caused some sporadic nitrogen stress which caused some increases in the variability of the inbreeds
growth. However, there is no statistieal data to support this theory. The decrease in the CVs of
the measured stover from 1989 to 1990 may have been due to late biomass sampling in 1989. In
1989, the last biomass samples was taken until a few weeks after physiological maturity. This late
sampling may have caused non-uniformity in stover amount due to loss of leaves from wind or

rain.

Plant and Biomass Data

Overall, the comparisons between CERES-1M simulated values and those measured in the
field showed good agreement. The simulated values versus the measured values for the grain
yield, kernel weight, grains per plant, and stover yield (at physiological maturity) are shown in

Figures 36 through 39. The nitrogen related biomass data, which includes information on percent

110

nitrogen in the grain, nitrogen uptake by the grain, percent nitrogen in the stover, and nitrogen
uptake by the stover (at physiological maturity) are shown in Figures 40 through 43.

Because of the large quantity of data involved in the inbred validation, a simple
comparison of simulated and measured values like that used in the hybrid demonstration shown
early, would lead to cumbersome tables with results difficult to quantify. Traditionally, model
evaluation has been done by testing the hypothesis that the regression line of simulated versus
observed values has a slope of l and passes through the origin (Carter, 1986). However, this
approach does not directly relate the predictive accuracy of the model. One commonly used term
to descn’be the measure of predictive accuracy is the mean squared error of prediction (MSEP)
(Wallach and Goffinet, 1989). The MSEP is a statistical term that represents an average weighted
difference between observed and simulated data. Muchow and Carberry, 1990, used the MSEP to
evaluate their mathematical models which described phenology and leaf-area development in

tropical grain sorghum. The MSEP, also called the root mean squared deviation, is defined as:

MSEP-[ ()3 (O-P)2) /n]°-5

Where 0 is the observed value, P is the predicted, or simulated value, and n is the number of
paired values. In Figures 36 through 43 the data points are shown with the 1:1 line (solid) and
the traditional regression line (dashed). Also shown is the regression line formula and the
associated R2 value. Additionally, the MSEP values for the components compared are given in
Table 27.

The grain and biomass data are shown in Figures 36 through 39. The grain yield
comparisons (Figure 36) showed good agreement between the simulated and measured values.
The regression analysis yielded a regression line with a slope of 1.08 and an R2 value of .656.
Additionally, the y-intercept was -549, indicating that CERES-1M has a bias to underestimate

yield. However, further study of figure 36 shows that it is the data from inbred 2 and 3 for 1990

111

 

 

 
 
 
 
    

 

 

 

O Inbred I - 89 a ,

A 5500- El Inbred 2 - 89 m «:0" -
'— A Inbred 3 - 89
'0 e Inbred l - 90 . o -
.c E Inbred 2 - 90
C» 5500- A Inbred 3 90 _
:5
U '1
.13
>- 4500- -I
U a
.93 i 1
2
3 _ 1:1 line -
,E 3500 I ,

A
m 4 Y = -549 + 1.080-X (--—) -

A 2
R = .656
2500 . 1 v r . I ' I
2500 3500 4500 5500 6500
Measured Yield (kg h0'1)

 

 

 

Figure 36. CERES-1M simulated versus measured grain yield for the three
inbreeds and the four treatments for 1989 and 1990. Constantine, MI.

 

 

  
  
 
  
  
  

 

 

 

0.27
1 O Inbred I — 89
3 y El Inbred 2 - 89
V A Inbred 3 - 89
.8: 0 Inbred I — 90
'5, 0‘25- 8 Inbred 2 — 90 d
'5 A Inbred 3 - 90
3
T)
E 0.238 ‘
0
x 1
'U
3 '1
o _ ’ -
.5 0.21 .
E * ‘
27> . 1.1 I. Y = .096 + .575rx (---> 1
. . "‘0 R°= 0.368 ‘
0.19 - T ' r y I ' I ' I r j ' ' 1
0.19 0.20 0.21 0.22 0.23 0.24 0.25 0.26 0.27
Measured Kernel Weight (g)

 

 

 

Figure 37. CERES-[M simulated and measured kernel weight for the three
inbreeds for the four treatments for 1989 and 1990. Constantine, MI.

112

Table 27. Values of the MSEP for CERES-1M simulated versus measured values for plant and
biomass data collected for the 3 inbred varieties. Constantine, MI, 1989 and 1990.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TOTAL _
FWR 1939 1990 1989 1990 1989 1990 (% of Measured)
Yield (kg Ira-1) 237 715 493 1059 118 880 673 (123%)
Kernel WI. (g) .002 .0086 .008 .035 .008 .019 0.017 (7.2%)
Grains Pram-1 19.10 42.36 20.27 28.76 14.09 64.13 35.99 (9.4%)
Stover (kg ha'l) 275 441 683 1924 678 915 978 (16.7%)
N% Grain .128 .156 .110 .071 .203 .170 0.146 (93%)
N Grn (kg ha'l) 4.40 13.65 12.11 14.35 1552 15.21 13.20 (18.2%) I
N% Stover .268 .607 .470 .505 398 .296 0.440 (453%) I
N Stvr (kg ha'l) 12.69 36.04 35.94 52.65 27.54 23.36 33.73 (57.6%) I

 

' Numbers in parenthesis are the MSEP value divided by the measured value.

that are the main muses of the low y-intercept and R2 value. The MSEP values given in Table 27

confirm this. However, the overall MSEP shows a good predictive accuracy, with a value of 673 kg

ha'l, or 12.3%.

In comparing simulated kernel weight with measured values (Figure 37), CERES-1M often

simulated the same number of kernels regardless of the treatment. This follows the logic of the

program since the simulation runs did not show any evidence of signifimnt water or nitrogen

stress, regardless of the treatment. The data points show a horizontal spread for all of the

inbreeds for 1989 and 1990. Also, inbred 2 and again to a lesser degree, inbred 3 in 1990 show

the poorest comparisons. Table 27 shows good predictive accuracy for kernel weight simulation.

The MSEP value for all inbred and treatment combinations for 1989 and 1990 was

0.017 g, or 7.2%. This is a good example of how the traditional statistiml analysis does not

provide a clear indimtion of predictive accuracy of the model.

When grains per plant were compared between simulated and measured values (Figure 38)

the highest R2 value was obtained: 0.797. This coincides with the low MSEP values found in

113

 

 

 
 
  
  
   

 

 

 

500
1 Y = .869 + .976ex (---) e e 1
“ R== 0 797 ° ’
5 450~ ' 8 -
E
L i
3 4oo~ .
m -I
C
'5 350-3 -
L.
0 4 J
U A
3 300- o Inbred 1 - 89 -
2 g D Inbred 2 - 89 g
g . , A Inbred 3 - 89
._ 250-< 1-1 ""8 e Inbred 1 - 90 1
m ‘ Ia Inbred 2 — 9o 4
A Inbred 3 - 90
200 I I I ‘l I I I I I I '
200 250 300 350 400 450 500
Measured Grains per Plant

 

 

 

Figure 38. CERES-1M versus measured grains per plant for the three
inbreeds and the four treatments for 1989 and 1990. Constantine, ML

 

 

       

 

 

 

9000q Y = 2815.55 4' .488‘)‘ (---) q
A
T fl R2: .546 e
g 8000- 4
0"
.x a
v 7000- t ' -
L
Q) T -I
> I
2 6000— -
m
'0
Q) _ O Inbred 1 - 89 _
*5 5000 1:1 Inbred 2 — 89 J
3 ‘ A Inbred 3 - 89
"g 4000.. 1:1 line a Inbred I - 90 .I
(f) 4 B Inbred 2 - 90
A Inbred 3 - 90
3000 v r . I f I I l ' I ' I
3000 4000 5000 6000 7000 8000 9000
Measured Stover (kg ha-1)

 

 

 

Figure 39. CERES-1M versus measured stover yield for the three inbreeds
and the four treatments for 1989 and 1990. Constantine, MI.

114

Table 27. There is some discrepancy with inbred 3 in 1990, but the remainder of the data
compared well.

The last of the biomass related comparisons made was the stover (Figure 39). Though a
relatively low R2 value was obtained from the regression analysis (0.546), Figure 39 does show a
fair correlation. As in previous comparisons, data from inbred 2 in 1990 showed the poorest
comparison. This is confirmed by the high MSEP value in Table 27. If the data for inbred 2,
1990, was removed from the analysis, the new R2 value would be only .608. The MSEP value
would be reduced from 978 kg ha'1 (16.7%) to 634 kg ha" (10.8%).

The nitrogen related plant data are graphically shown in Figures 40 through 43. The
MSEP values are also given in Table 27. The percent nitrogen in the grain and related nitrogen
uptake by the grain are shown in Figures 40 and 41. A comparison of the percent nitrogen in the
grain showed a seemingly poor correlation with a R2 value of only 0.135 because of the low range
of values. However, the MSEP data in Table 27 indicates good predictive accuraq'. In the
simulation of nitrogen dynamics in CERES-TM, the maximum percent nitrogen that can be
obtained in the grain is 1.70 percent. Given that no stress occurs and there is ample nitrogen for
grain filling, the model will normally simulate a grain nitrogen content approximating 1.70%. The
almost horizontal regression line in Figure 40 illustrates this. However, the MSEP given in
Table 27 for percent nitrogen in the grain shows an error of less than 10% for the 3 inbreeds over
the 2 year period. This is another example of how R2 values can be misleading in evaluating
model performance.

In the comparison of nitrogen uptake by the grain (Figure 41), the data compared well.
The 1?.2 and MSEP values are reasonable and the regression line formula indicates a good
comparison.

figures 42 and 43 contain information on the nitrogen in the stover. For both the
percent nitrogen in the stover (Figure 42) and the nitrogen uptake by the stover (Figure 43)
CERES-1M underestimated as compared to the measured values. The values for the measured

percent nitrogen range from 0.75 to 1.3 percent. However, CERES-1M simulated values range

115

 

 

. Y = 1.27 + .2528x (—--) .
1,9. R’= 0.135 -

   

 

 

A

N

v

C

Q)

as

8

:2: 1.7- ..

z . .

,E 16‘ ..

O

(3 -I -I
1.5“ 4

B ‘ o InbredI-89 1

'5 ' . 1:1 lme A Inbred3-89 ‘

E 4 e Inbred I -90

“(7) L3 81 Inbredz-so‘

I A Inbred3-9o ‘

 

T Y T V I V

I I I I
1.5 1.6 1.7 1.8 1.9 2.0

Measured Grain Nitrogen (%)

I I I I
1.2 1.3 1.4

 

 

 

Figure 40. CERES-1M versus measured percent nitrogen in the grain for
three inbreeds and the four treatments for 1989 and 1990. Constantine,
MI.

 

 

   

Inbred l - 89 -

 

 

 

 

_‘ O
55 U Inbred 2 - 89

1 . A Inbred 3 - 89 1
45 1:1 ""6 . e Inbred I - 90 -

‘ B Inbred 2 - 9O

‘ A Inbred 3 - 90

35 ' I T I r I ' I i I fl l

35 45 55 65 75 85 95

Simulated N Uptake by Grain (kg ha")
03
(1'
1

Measured N Uptake by Grain (kg ha")

 

 

 

Figure 41. CERES-1M versus measured nitrogen uptake by the grain for
the three inbreeds and the four treatments for '1989 and 1990.

Constantine, MI.

116

 

 

 

 

 

 

1.4
A 1 31 O Inbred 1 - 89 Y = .282 + .321tX (---) -
N ' D Inbred 2 - 89 R3- 0135 ..
V 1 2: A Inbred 3 - 89 _
S . J 9 Inbred 1 - 90 In line 1
8" 1.14 Q Inbred 2 - 90 -I
L. ‘ A Inbred 3 - 90 1
.t’ 1.0‘ -I
Z I I
L 0.9‘ ..I
3 J ‘ ‘
3 0.8- . ..
m 0.7" ’ -I
O < 4
3 0.6- -
_0 - .
3 0.5- .4
E 1 1
(7) 0.4-1 -I
0.3 ' I ' I ' T ‘ I ' I V I Y 1 ‘ I ' I ' I r
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Measured Stover Nitrogen (%)

 

 

 

Figure 42. CERES-1M versus measured percent nitrogen in the stover for
the three inbreeds and the four treatments for 1989 and 1990.
Constantine, MI.

 

 

  
    

 

 

 

A
'0 I20
_c . o Inbred I - 89 Y = 1289 + .5561.x (---)*
o, “0': c1 Inbred 2 - 89 R’- 0.306 1
x A Inbred 3 - 89 ..
v 100-
h 4 Q Inbred 1 - 90 1.1 line
'5’ 90-1 B Inbred 2 - 90 '
8 1 A Inbred 3 — 90 "
(I) 801 -
5‘ 70': ‘1
g 60': '2‘
‘a 50‘: , - ‘ ‘.
D 40‘ _(
z 4

n
g .0- I
3 O 1 1
2 2 1 j
g 10 fo I I I I I I I . I I I I I . I I I I
I7) 10 20 30 4O 50 60 70 80 90 100 110 120

Measured N Uptake by Stover (kg 110-1)

 

 

 

Figure 43. CERES-1M versus measured nitrogen uptake by the stover for
the three inbreeds and the four treatments for 1989 and 1990.
Constantine, MI.

117
from 0.35 to 0.85 percent. This seems to indicate that CERES-1M allowed the nitrogen content of

the stover to fall too low and perhaps did not allow for enough nitrogen uptake to be simulated.
There was also some question as to the cob nitrogen content. Though CERES-1M simulates the
growth of the car, it does not simulate the nitrogen in the ear independently of the other plant
components. The measured data used for comparison of the stover nitrogen includes the nitrogen
in the cob. This may explain the underestimation by CERES-1M. The discrepancy in the nitrogen
uptake by the stover followed a similar pattern.

Another comparison made was above ground biomass accumulation during the growing
season. As mentioned earlier, biomass samples were taken four times during the growing season.
Bemuse of the large number of comparisons, only inbred 2 will be discussed in depth. Inbred 2
was chosen because it showed the best and worst comparisons for above ground biomass and is the
inbred that was grown on the lysimeter plots. However, all of the MSEP values for all of the
inbreeds are in Table 28.

The MSEP values for the biomass were larger in 1990 than in 1989, except for inbred 3,
sample 2 (Table 28). The largest increases came from inbred 2 and inbred 3. It is difficult to
determine the exact causes of these increases. One definite factor is the zero nitrogen treatments
in which CERESJM underestimated grain and stover yields. Other factors such as amount of
biomass removed by detasseling may have contributed. CERES-[M simulated values and those
measured for above ground biomass for inbred 2, treatments 1 through 4, for 1989, are shown in
Figures 44 through 47. For all four treatments, there was excellent agreement between CERES-
IM and the measured values.

The comparison between modelled and measured above ground biomass for 1990 are
shown in Figures 48 through 51. All four figures show a similar pattern: CERES-1M either
simulated well or slightly overestimated above ground biomass accumulation through to the third
sampling. Biomass accumulation for the fourth sample is underestimated by CERES-1M for all

treatments.

118

 

 

 

 

 

A
T 14000.- .
O . e
‘c 12000 - I e
o .
x - .
v
.0 10000.- -
m
C ‘ .
E 8000- _
.9
m . .
U 6000.- -
C . .
3
9 4000.- -
o 1
2 2000.4
8 I Measured 4
< 1 — CERES—III
. . , . . . , . . .
5/20/89 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 44. CERES-1M and measured above ground biomass values for inbred 2,
treatment 1 (180pp). Constantine, MI, 1989.

 

 

 

 

 

T 14000.- —
O - .
‘c 12000— I —
o .
x . .
V
$1 10000.- -
o . I .
E 8000— -
.9
m . .
U 6000.- —
C . .
3
e 4000— 1
o . .
2 2000- —
.8 II. Measured
< — crass—m
0. fl , . , e - , . . a
5/20/89 6/9 6/29 7/19 8/8 8/28 9217 10/7
Date

 

 

 

Figure 45. CERES-1M and measured above ground biomass values for inbred 2,
treatment 2 (PRF - 80 kg N 11:14). Constantine, MI, 1989.

119

 

 

 

 

 

A
'1' 14000.1 e
o . .
‘C 12000 - -
o .
x < .
v
0, 10000- -
m I
o .
E 8000.- 3
.9 J
m .
1, 6000.- -
C . .
3
9 4000— -
O . .
9 2000.~ -
8 ‘ 1 Measured
< 0 — CERES—IM

. . . . . . . . . .

5/20/89 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 46. CERES-IN! and measured above ground biomass values for inbred 2,
treatment 3 (90s). Constantine, M1, 1989.

 

 

 

 

 

A
'I' 14000.- 1
C e 1
‘C 12000 —
OI ' '1
X e 4
V
a, 10000.- _
(D
O ‘ .
E 8000- _
.9
m . .
U 6000.- -
C 4 .
3
g 4000.— -
o . .
"g 2000- —
8 . 1 Measured
< o — CERES—IN

- I I v *v’ I I r r fir

5/20/89 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 47. CERES-1M and measured above ground biomass values for inbred 2,
treatment 4 (1805). Constantine, MI, 1989.

 

 

 

 

 

A
'1— 14000.- -
C < .
S 12000 e -
o .
x 4 1
v
a, 10000.9 _
m
D ‘ 4
E 8000- -
.9
m . .
U 6000.- -
C . .
3
e 4000— —
O . I .
2 2000- -
_8 . I Measured
<1: 0 — CERES—IM
. . , . . . , . .
5/20/90 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 48. CERES-1M and measured above ground biomass values for inbred 2,
treatment 1 (180pp). Constantine, MI, 1990.

 

 

 

 

 

T 140001 I -
o . .
.C

12000- —
U’
x . <

v
a, 10000- _
(I)

O ‘ .
E 8000.~ —

.9
m . .
t, 6000.9 -
C . .
3
g 4000. - —
o . I
"g 2000— -
_8 . I Measured
< c — CERES—IM

- . I . T I I I I . I I
5/20/90 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 49. CERES-IN! and measured above ground biomass values for inbred 2,
treatment 2 (PRF - 0 kg N ha'l). Constantine, MI, 1990.

121

 

 

 

 

 

 

A

I

O

c 12000—

0, . "1

ex ‘ 4

v

v, 10000— -

m

U .

E 8000— —

.9

m .

U 6000— —

C 1 1

3

e 4000— —

o .

2 20004 -

O I I Measured ,

.0

< — canes-m
o. . , . . r . T . . .
5/20/90 6/9 6/29 7/19 8/8 8/28 917 10/7

Date

 

 

 

Figure 50. CERES-1M and measured above ground biomass values for inbred 2,
treatment 3 (908). Constantine, MI, 1990.

 

 

 

 

 

A
T 14000— -
O - .
I 12000— I —
o .
x — .
\J
,,, 10000— -
0)
C ' I
E 8000— 4
.9
m . .
U 6000— —
C .
8 4000
5 '. I .
g’ 2000— —
_8 4 I Measured
< o — CERES—IM

. . , . , . , . , . . , .

5/20/90 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 51. CERES-1M and measured above ground biomass values for inbred 2,
treatment 4 (0 kg N ha"). Constantine, MI, 1990.

122

Table 28. MSEP values (kg ha'l) of CERES-1M simulated values for above ground biomass for
the 3 inbreeds, for all treatments, for 1989 and 1990. Constantine, MI.

 

 

 

 

 

 

 

MSEP FOR ABOVE GROUND BIOMASS (kg ha'l)
Sampling- INBRED 1 INBRED 2 INBRED 3 AVERAGE
Dam 1989 1990 1989 1990 1989 1990
1 144 376 125 358 50 254 249
2 943 1278 428 1631 1064 1008 1119
3 794 878 712 1039 957 1555 986
4 350 425 479 2989 828 1643 1463

 

 

 

 

 

 

 

 

 

 

° Sampling dates for all inbreeds for 1989 were 1) June 28; 2) July 27; 3) August 18;
and 4) September 27. Sampling dates for 1990 for inbred 1 were 1) June 24; 2) July 27; 3) August
16; and 4) September 12. Sampling dates for inbrwds 2 and 3 for 1990 were 1) June 24; 2) July
27; 3) August 23; and 4) September 25.

The final comparison made between simulated and measured values was leaf area index
(1A1). LAI measurements were taken in 1989 only. All three inbreeds were measured using 15
plants from the 180 kg pp treatment (treatment 1). The comparisons of CERES-1M simulated
and measured LAI values are shown in Figures 52 through 54. There was excellent agreement
between CERES-1M simulated and measured LAI values. As seen in Figure 52, CERES-IM
adequately simulated LA] for most of the season, though there is some discrepanq just after

detasseling, The other two inbred, Figures 53 and 54, show excellent agreement throughout the

season.

Drainage and Leaching Data

The lysimeter plots in this study were all planted with inbred 2. There were 5 drainage
lysimeters installed: four with disturbed soil profiles and one with an undisturbed soil profile.
There were four treatments and the undisturbed lysimeter was used as a replicate for one of the
treatments. To keep the data consist with lysimeter type, only the undisturbed lysimeters were
evaluated with the CERES-1M simulated data. Table 29 shows the lysimeter number and the

nitrogen treatment used for 1989 and 1990. The undisturbed lysimeter was lysimeter number 3.

123

 

 

Leaf Area Index (cm2 cm’z)
I}.
l

 

4 It Measured
— CERES-m

0.0 t '
8/8 8/28 9/17

5/20/89' 6)9 f 6/29 ' 7/'19
Date

 

Y

 

 

 

 

 

Figure 52. CERES-1M and measured LAI values for inbred 1, treatment 1
(180pp). Constantine, MI, 1989.

 

 

 

 

 

3.0

A 4

N

l 2.5.. .J

E

o 4

N _ -I

E 2.0

U

V

a; 1.5- -

.0

E

o 1.0a ..

Q)

L

< 1

“5 o.5~ .1

Q)

_, :1: Measured 1

— canes-m
0.0 4 , . . . , . - .
5/20/89 6/9 6/29 7/19 8/8 8/28 9/17 10/7
Date

 

 

 

Figure 53. CERES-1M and measured LAI values for inbred 2, treatment 1
(180pp). Constantine, MI, 1989.

124

 

 

 

 

 

 

313

,-\ J - ,

N T

| 2.5~ -

E

0

c9 _ -

E 200

U

V

5; 1.5-‘ ~

'0

E

o 1.0— ..

0)

L

< 1

“5 0.5— .

Q)

_, 1 as Measured

— canes-m
cxo . , . . , . , . s. .
5/20/89 6/9 6/29 7/19 8/8 8/28 9717 10/7
Date

 

 

 

Figure 54. CERES-1M and measured LAI values for inbred 3, treatment 1
(180pp). Constantine, MI, 1989.

Table 29. List of the lysimeters and the nitrogen treatments used for 1989 and 1990.
Constantine, MI

TREATMENT (kg N ha‘l)
1989 1990

 
  

    
   

 

LYSIMETER ('l‘rt n#)

      

 

 

l (3) 90 split 90 split
2 (2) 80 PRF 0 PRF
4 (4) 180 split 0

5 (1)

   

 

        
  

 

     

 

   

 

 

     

180 preplant 180 preplant

Comparisons between the measured and simulated drainage data showed excellent results
for both years of the study. The drainage data for 1989 is shown in Figures 55 through 58. For
all four lysimeters, CERES-1M did a good job of following the flow of drainage water throughout
the year, though there was some discrepancy at the beginning of the season. One explanation for

this is the lag time for drainage to occur. Measurements taken have shown a lag time between

125

 

 

300- — CERES-IM ..
4 H Measured ,‘ q
700_‘ -- Precip [J q

d

600 ..

d

500 -

400 ..

4

300 -

Drainage (mm)

200 -1

d

100-1

1

 

 

 

 

 

o - . . , - , . r . , . .
5/50/89 7/19 9/7 10/27 12/16 2/4/90 :5/26 5/15
Date

 

 

 

Figure 55. CERES-1M and measured drainage values for lysimeter 1,
treatment 3 (905). Precipitation is also shown. Constantine, MI, 1989.

 

 

 

 

8001 -— CERES—IN ,r T
, H Measured /’ 4
700-1 "- Precip r, q
n '1} q
fg 600- I -
4 ——-"' 1

E g /___,x
v 500 J 4
Q) 4 'J 4
g 400'1 / "‘
C d / <
8 300-1 I, -
O < J,’ ll
200~ J ..
d I 1

I!

100- -’ -

 

 

 

10}27 Y 12}1612/4'/901 3/26f5/15
Date

0 - . . ,
5/30/89 7/19 9/7

 

 

 

Figure 56. CERESJM and measured drainage values for lysimeter 2,
treatment 2 (PRF . 0 kg N ha'l). Precipitation is also shown. Constantine,
MI, 1989.

126

 

 

 

 

 

 

 

800- — CERES-IM J -
. H Measured ’1 .
700_ —- Precip Jr _'
. r/I .
E 600‘ ' ~
1 /
,._._
f, 500- r,_-_« -
o ‘ r." i
O 400‘ I "
O
.E . / i
g 300~ /,---— -
o . ,e .
200~ 1" -
1 ll 1
100.. ’1 ..
J A _ .4 1
0 7 ' ' T ' I ' 1 fl 1 ' '
5/30/89 7/19 9/7 10/27 12/16 2/4/90 3/26 5/15
Date

 

 

 

Figure 57. CERES-[M and measured drainage values for lysimeter 4,
treatment 4 (1805). Precipitation is also shown. Constantine, MI, 1989.

 

 

 

 

 

 

800- — CERES-4M I)" -

, H Measured P/ .

700‘ —- Precip /’ _

4 l J

l d

E 600i /-——fi,__.l ‘

f, 500- 1,! -

Q) '4 I T

0’ 400- / -‘

O 2

.E ‘ ”" ‘

g 300- ,J _

Q ~ I, 4

2004 I e

/
1 1 ‘
100- /' -
o f . , . , . , . , - .
5/30/89 7719 9/7 10/27 12/16 2/4/90 3/26 5/15
Date

 

 

 

Figure 58. CERES-1M and measured drainage values for lysimeter 5,
treatment 1 (180pp). Precipitation is also shown. Constantine, MI, 1989.

127

precipitation events and the drainage that these events cause. CERESolM cannot adequately
simulate this lag time because the bottom of the lysimeter causes a time delay in the outflow.
However, most of the data showed good agreement.

The nitrate leaching data for 1989 did not compare as well as the drainage data. The
comparison between CERES-1M and the measured values for nitrate leaching are shown in
Figures 59 through 62. CERES-IM did a good job of simulating the nitrate leached for lysimeters
1 (905) and 2 (PRF-80 kg N ha"), though the timing of when the nitrate leached does differ
between the measured and simulated values (Figures 59 and 60). However, for lysimeters 4 (1805)
and 5 (180pp), CERES-1M overestimated the nitrate leached by a factor of almost two (Figures 61
and 62). As seen in all four figures, there was a large increase in the nitrate leaching at the end of
October. The reason for this was that a large irrigation was added to the lysimeters.

Following the summer of 1989, in which almost no drainage occurred, a large amount of
irrigation water was applied to the plots to test the lysimeters and to assure there were no leaks in
the system. The amount of irrigation water applied ranged from 250 mm to 350 mm. The
irrigations took place over a 2-day period and applications were split between the two days.
Fortunately, the application of the irrigation water showed no leaks in the systems. However, the
large amount of drainage that occurred may have caused difficulties in simulating the nitrate
leached. The large discrepancy between the simulated and measured values of nitrate leaching
may have been caused by macropore flow. Macropore flow is water that flows through the
macropores of the soil, coming in contact with little of the soil particles and leaching out minimal
amounts of chemicals. If macropores were present in the lysimeters 4 (1805) and 5 (180pp), this
could account for the discrepancies. Also, it should be noted that the measured amounts of
nitrate leached from the lysimeters was not proportionate to the nitrogen fertilizer applications
made.

The lysimeter with the largest amount of nitrate leaching was lysimeter l, which received a
nitrogen fertilizer application of 90 kg N ha'1 and leached a total of 90 kg N ha’l. Lysimeter 5

received 180 kg N ha‘1 of nitrogen fertilizer and leached about 80 kg N ha“. Lysimeters 2 and 4

128

 

 

 

 

 

 

 

 

140
g — CERES-IM '

,1“ H Measured 1
| 120‘ "‘
O 4 .
.c
2 ‘00“ "‘
U5 " .4
:3 80~ -
U 4 4
2
o 60- -
O 4
3

40“ d
3
8 .
:0: 20- -
Z i 1

o e e - ‘, . ,‘ . , . , . .
5/30/89 7/19 10/7 11/27 12/16 2/4/90 3726 5/15
Date

 

 

 

Figure 59. CERES-[M and measured nitrate leaching values for lysimeter 1,
treatment 3 (908). Constantine, MI, 1989.

 

 

 

 

 

 

 

14o
. — CERES-UM q
9 HMeosured
I ‘20“ 'l
0 j 1
.C
2 100‘ "‘
3 .
v 80~ —
'0 .
2
o 60- -
O .
Q)
-’ 4o~ -
3
o .
,4: 20- .
z .
o—é=.,.€.,.,..
5/30/89 7/19 9/7 10/27 12/16 2/4/90 3/26 5/15
Date

 

 

 

Figure 60. CERES-IM and measured nitrate leaching values for lysimeter 2,
treatment 2 (PRF - 80 kg N ha-l). Constantine, MI, 1989.

 

 

 

 

 

 

240
- — CERES—IM 1
A j H Measured '
_g ZOO-'1 1
\ .
z j .
or 160‘ -
.x . i
V ‘ I
8 . 1
_s: 120: ‘4
U
C ‘ .1
Q) 4 .
—’ 80-1 .-
0 1
*4 -1
O 1
E 40~ -
Z 1
1 i
0 :4 r I . I s I . I a f
5/30/89 7/19 9/7 10/27 12/15 2/4/90 3/26 5 15
Date

 

 

 

Figure 61. CERES-1M and measured nitrate leaching values for lysimeter 4,
treatment 4 (1805). Constantine, MI, 1989.

 

 

 

 

 

 

 

280
j — CERES—IM 3
1 H Measured ‘
A 240- -
O ‘ 4
.C 1 4
E 200. _
x “ 1
v ‘ j
.
8 160': j
.C a
o . 1
8 120— ~
_I ‘ :
a) 1 1
‘6 8°“. 1
b .
Z 403 1
- - - _3'
0 f" ' r' 1 Y I 7 l f r ' '
5/50/89 7/19 9/7 10/27 12/16 2/4/90 3/26 5 15
Date

 

 

 

Figure 62. CERES-IM and measured nitrate leaching values for lysimeter 5,
treatment 1 (180pp). Constantine, MI, 1989.

130
both leached about 55 kg N ha‘l. However, lysimeter 2 received only 80 kg N ha‘1 while

lysimeter 4 received 180 kg N ha-l. The nitrate leaching simulated by CERES-1M follows the
expected pattern with lysimeter 4 and 5, the 180 kg N ha‘1 treatments, leaching about the same
nitrate (160 kg N ha-l) and lysimeters 1 and 2, which received 90 kg N ha-1 and 80 kg N hit-1,
leaching about same amount of nitrate (80 kg N ha'l). Since the simulated data of CERES-1M
follows logic concerning nitrogen added and nitrate leached, it seems to indicate that there were
interactions which took place that CERES-1M could not account for or that the initial state was
not properly described.

In the second year of the study, CERES-1M continued to show good ability to simulate the
soil-water drainage from the lysimeters. The measured and simulated drainage values for the four
lysimeters for 1990 are shown in Figures 63 through 66. Once again, some discrepancies occurred
between the timing of the drainage, but overall, CERES-[M followed the drainage patterns
measured.

The nitrate leaching data for 1990 is shown in Figures 67 through 70. All four lysimeters
showed good agreement between CERES-1M simulated and measured values. All of the
lysimeters showed a slight increase in loss during the early summer and then a leveling off of
nitrate loss during the growing season. This is partly due to the plants using soil nitrogen and the
low drainage that occurred during the season. Likewise, the graphs show the increase in nitrate
leaching that occurred after the growing season, during the fall and early winter months. One
graph of interest is that for lysimeter 5, the 180pp treatment (Figure 66). The nitrate loss for the
lysimeter followed a similar pattern to that shown by the other lysimeters for most of the year.
However, toward the end of the year, the nitrate loss sharply increased. This is probably due to
the nitrogen fertilizer treatments over the last two years (180 kg N ha'1 applied each year).
CERES-1M simulated this loss well, though the total nitrate loss was slightly overestimated.

The data presented here have shown that CERES-1M can adequately simulate inbred
maize yield and development. Additionally, CERES-IM can simulate drainage through the soil

profile and the nitrate loss associated with the drainage. The next step was to use CERES-1M

131

 

 

 

 

 

800- — CERES-IM I, -
. H Measured (P -
700 _ — - Precip - Pi -
1 ,,; 1
A 600a ' -
E . ’1’ .
l

E 500- / ~
g d r_J d
o 400‘ [.4 "
.E ‘ ,/ -
E 300‘ f -
o . 1’ .
200- )4 4

4 ’J"
100'1 ,x", '1
«I d

O ' l ' 1 V l ' t ' l
5/10/90 6/29 8/18 10/7 “/26 1/15/91
Date

 

 

 

Figure 63. CERES-1M and measured drainage values for lysimeter 1,
treatment 3 (908). Also shown is the precipitation. Constantine, MI, 1990.

 

 

 

 

 

800~ — CERES-IM ./ -
4 H Measured (F 4
_- ' l
700_ Precrp r--" _
. I”) .
A 600~ ’ -
E . J, .
r

E, 500~ /, _
g '1 r—J I
o 400‘ [.4 "
.E ‘ ,/ 1
9 300- r -
O 4 j, q
200-1 l4 .t
1 ’J-' ,
100-1 ,/"’ -

1i

0 ' l ' ' 1 ' f ' I
5/10/90 6/29 8/ 18 10/7 11 /26 1 /15/91
Date

 

 

 

Figure 64. CERES-1M and measured drainage values for lysimeter 2,
treatment 2 (PRF - 0 kg N ha'l). Precipitation is also shown. Constantine,
MI, 1990.

132

 

 

 

 

 

300- — CERES—IM ./ _
. H Measured (P 4
700 _ - - Precip _- J _
‘1 [_.Jr -1
A 600- ' -
E . r) .
r"
E, 500- / -
a) ‘ r.._J J
O, 400‘ 4.4
o l
.E ‘ ,/ .
g 3004 r -
o 4 J’ 4
200' )4 -1
1 ’Jd
100‘ ,x", "1
-I
0 ' T ' 1 7 l ' r ' 1
5/10/90 6/29 8/18 10/7 11/26 1/15/91
Date

 

 

 

Figure 65. CERES-1M and measured drainage values for lysimeter 4,
treatment 4 (0 kg N ha'l). Precipitation is also shown. Constantine, MI,
1990.

 

 

 

 

 

goo- — CERES-1M ./ ..,
4 H Measured (‘- -
_- - 1

700‘ Precup r—" _

4 r/,4 1

A GOO-1 .4
E 4 I, .

I"

f, 500- / -
a ‘ r—J ‘
o 400‘ [.1 -1
.E ‘ ,/ 1
E 300- If .4
O -1 j 1
200- )4 .I

4 ’JH q
loo-,x" -1

4’ c1

0 . I 4 I . I . I . I
5/10/90 5/29 8/18 10/7 11/26 1/15/91
Date

 

 

 

Figure 66. CERES-1M and measured drainage values for lysimeter 5,
treatment 1 (180pp). Precipitation is also shown. Constantine, MI, 1990.

133

 

 

160

, — CERES-IM .
140- H Measured _
120- ~

1

_a
l
J

L

a1

0
l
1

Ci
0
l l
l

L

Nitrate Leached (kg N ltd—1)
I5

20- -

 

 

 

T

0 r' r v l 1 ' 1 4
4/10/90 5/50 7/19 9/7 10/27 12/16 2/4/91

Date

 

 

 

Figure 67. CERES-1M and measured nitrate leaching values for lysimeter 1,
treatment 3 (905). Constantine, MI, 1989.

 

 

 

 

 

160

,T.‘ 140- H Measured _
o . .
I 120- J
z . .
3‘ 1004 e
v 4 4
3 804 4
6 l ‘
o 60— -
Q)

_l <

3 404 -1
O 4 .1
L

r: 20- —
z 1

o 44 . 4 I . I . I .
4/10/90 5/30 7/19 9/7 10/27 12/16 2/4/91
Date

 

 

 

Figure 68. CERES-1M and measured nitrate leaching values for lysimeter 2,
treatment 2 (PRF - 0 kg N ha“). Constantine, MI, 1990.

134

 

 

160
. — CERES—1M 1
.f‘ 140_ H Measured _
lo 1 q
5 120.4 4
Z 4 4
3‘ 100- 1
V a -1
8 80- -1
'5 4 -
0 60a -
Q)
_l 1 4
Q) 40"1 -1
«OJ
C 1 1
b 20
Z 1 ‘

 

 

 

o .- . . I . 4 I .
4/10/90 5/30 7/19 9/7 10/27 12/16 2/4/91

Date

 

 

 

Figure 69. CERES-1M and measured nitrate leaching values for lysimeter 4,
treatment 4 (0 kg N ha'l). Constantine, MI, 1990.

 

 

160
. — CERES-1M 4
,1" 140‘ H Measured 4
I 4 .
o
-C 120- -
z . 1
3‘ 1001 -
v < «1
U —1 d
111 80
s . 1
o 60« -
a)
_J A
a) 40- 4
H
o 4 .
b 20
z . .

 

 

 

0 ' V f I I I '
4/10/90 5/30 7/19 9/7 10/27 12/16 2/4/91

Date

 

 

 

Figure 70. CERES-1M and measured nitrate leaching values for lysimeter 5,
treatment 1 (180pp). Constantine, MI, 1990.

135

investigate the possible effects of various nitrogen management schemes on grain yield and
nitrate leaching. To do this, several years of simulated weather data were used and some

additional reprogramming of the model was performed.

Evaluation of Nitrogen Management Strategies

The nitrogen management strategies were evaluated over a 60 year time period. The data
were “mined in 30 year intervals to determine if the first 30 years was significantly different than
the last 30 years. However, the data showed no significant differences for any of the strategies.
Additionally, differences in the PRF strategies with regards to the plant stage limitation set were
also minimal, except for the 30 kg N ha'1 threshold value. Therefore, the strategy that limited
nitrogen fertilizer application after grain filling were deleted from the analysis for the PRF-Auto
N , the PRF-15/15, and the PRF-15/0 strategies. The analyses for the PRF-30m and the PRF-
30/0 included the growth stage limitations as an example.

The average yield for the 60 year period for the strategies evaluated are presented in
Figure 71. The zero nitrogen strategy had the lowest average yield, as expected. The two 180 kg
N ha‘1 strategies yielded the highest, with an average of 5758 kg ha'l. The next highest yielding
strategies were the two PRF-30m strategies and followed by the PRF-15/15 and the PRF-Auto N.
The lowest average yields, besides the zero nitrogen strategy, were produced by the two PRF-30M
strategies. This would indicate that if 30 kg of nitrogen is required, some nitrogen should be
added. The data from the PRF-15/0 suggest that the threshold value should be somewhere
between 15 and 30 kg N.

In conjunction with average grain yield, it is also important to know the variation of the
yields from year to year. Often, the strategy that yields the highest average grain yield is not the
most stable strategy. However, in this case it was. Table 30 contains data on the CVs of the grain
yields for the 11 strategies presented in Flgure 71. The strategy with the lowest CV was the 180
pp strategy. The highest was, as expected, the zero nitrogen treatment. The two PRF-30K)

strategies also showed relatively high variation. The remainder of the strategies had

 

136

 

0
fl 0
.m U... m
e “mull:
0.. sgflm
e (<89
L p ((<<
t d I.\I.\ SM
n Pswww0001o
e Nww ////wwt
m 41mmmmmiin
W :-__:--=-
W 125456789WH

 

n
.m
m
V
f
o
m
m
m

Treatment Number

 

 

W m m m m m ..
A—Ioc 9;
22> E80 ouono><

 

 

Figure 71. Average grain yield for the 11 strategies evaluated for the 60-year continuous

simulation.

 

Table 30. Coefficient of variation expressed as a percentage for grain yields for a 60 years

CERES—1M simulation run.

 

137

relatively the same variation, with CVs ranging from 16.29 to 19.71. The highest of these was the
90 sd strategy.

All of the strategies evaluated had the same inputs and initial conditions except for the
nitrogen management. For some strategies, the amount of nitrogen fertilizer applied was fixed
(i.e., 180 pp, 180 sd, 90 sd). For the PRF strategies, the amount of nitrogen fertilizer applied
varied depending on the threshold value used. The average amount of nitrogen fertilizer added
for each strategy is shown in Figure 72. The strategies of most interest are the PRF strategies.

The PRF-30m strategies show a slight increase between the silking day restriction and
the grain fill restriction. This increase was also seen in the PRF-15/15 (only silking day restriction
shown). The two PRF-30/0 strategies applied the same amount of nitrogen, but less than either
PRF-30m strategy. This trend was also seen in the PRF-1510 strategies (only silking day
restriction shown).

Leaching amounts are presented in Figure 73. These data follow a somewhat expected
trend with the two 180 kg N ha'1 strategies leaching the most nitrate with an average of
approximately 111 kg N ha’1 yr“. Also, the zero nitrogen strategy yielded the lowest leaching
average, with about 50 kg N ha'1 leaching annually. One strategy that also produced relatively low
leaching amounts was the 90 kg N ha'1 sidedress. The leaching amount from this strategy was
slightly lower than the two PRF-30,0 strategies while the yield was significantly higher. However,
the yields from the other PRF treatments were slightly higher than the 90 ad. This may indicate
that the 90 sd strategy, which is fairly straight forward and easy to manage, is comparable to the
more manage intensive PRF strategies.

A final evaluation made of these nitrogen strategies was a simple economic analysis.
Since only the nitrogen fertilizer applications and irrigation water applications were the only

inputs that varied among strategies, only the costs associated with these two inputs were

 

 

I )
)

 

.n. > m
'
e Mm»,
9 seam
e
L D. I<\I,.\\<<
t d (( 5"
n DswwwOOOIO
e ammo///wwwt
m orrammmsirn
W ----:--
W 123456789wfl
1‘14-q - anus-‘4‘:

 

1 u 1 d 4 q 4
m m m
ATS. z 9:
oovu< codzto... z we #5954 00803

W

L

 

aM

 

Treatment Number

 

 

Figure 72. Average annual amount of nitrogen fertilizer added for the 11 strategies evaluated

for the 60-year continuous simulation.

 

 

.4. > m,

6 «mm;

9 serum.

8

L D. «k\<<
((

sL Fwdww ”N

n a 000 o

6 N00 //////t

m BBOOOOMSGM

LL 0119333 4|

w .=.=.=-=.g.

r 12345678901

T 11

4 _ 11. 4 a 4 4

 

 

 

:12. z 9:
vogue».— oaeezz omEo><

 

Treatment Number

 

 

Figure 73. Average annual nitrate leaching for the 11 strategies evaluated for the 60-year

continuous simulation.

139

considered. Thus, the fixed production costs are not subtracted. The formula used to compute

the average revenue is:
REV = YLD " 0.108 -((NIT‘ ‘ 0.40) + (NFERT " 12.00) + (AMIRR ‘ 0.14))

where REV is the revenue (3 ha“), YLD is the grain yield (kg ha‘l), NIT is the amount of
nitrogen fertilizer added (kg N ha'l), NFERT is the number of nitrogen fertilizer applications
made (applications yr’l), and AMIRR is the amount of irrigation water applied (mm). The
constants used are based on cost incurred in the inbred study discussed earlier.

The revenue values from the 11 strategies evaluated are shown in Figure 74. Once again,
the zero nitrogen strategy had the lowest amount, with a revenue of only $14.12 ha". The 180 pp
strategy yielded the highest average revenue, $529.30 ha'1. The lowest revenue, besides the zero N
strategy, was obtained with the two PRF-3010 strategies. The remainder Of the strategies varied
from about 5498 ha'1 to $518 ha'l. Also, it should be noted that the 90 sd strategy again ranked
high as it did in the yield data. Only the 180 pp, the 180 sd, and one of the PRF-30m strategies
yielded more revenue. With the low leaching amount shown earlier, the 90 sd strategy must be
considered as the viable strategy of this evaluation in terms of revenue, minimal leaching, and
management required.

Another alterative method to evaluating these strategies is to plot the revenue data
against the leaching data, as shown in Figure 75. With revenue on the y-axis and leaching on the
x-axis, the most optimal strategy in terms of revenue would be the data point closest to the top of
the graph. However, the most optimal strategy in terms of nitrate leaching is the data point
furthest to the left of the graph. Unfortunately, the most optimal point in terms of revenue is also
the least optimal point in terms of leaching. The 180 pp strategy yielded the largest revenue and
the largest annual nitrate leaching. The converse is true for the most optimal strategy in terms of
leaching. The zero nitrogen strategy yielded the lowest nitrate leaching but also yielded the lowest
revenue. Therefore, the point of compromise must lie between these two strategies. In Figure 75,

this is the point that lies furthest to the top and furthest to the left, which is the 90 sd strategy.

140

 

 

 

 

 

£1!“
WV
A
'- 500. 5 3 r p r r -
| fl d fl d v ' f d
o 5 5 5 5 5 5 5 5
1: ‘ f v t a d r r f f d
v v t t f v f f f d
f f v f d l d d d d
0 400-1 v a r a a a a a a a -
v 5 5 5 5 5 5 5 5 5 5
1n 4 a r r a a a a r a a Treatment Legend
d d f d d f I d d f
3 f fl f fl f f d f d f 1 I o N
C 3004 5 5 5 5 5 5 5 5 5 5 -
0 a a a a a a a a a a 2 - 180 pp
> a a a r a a a a a a
a 1 5 5 5 5 5 5 5 5 5 5 3 ' ‘80 =11
I! 5 5 5 5 5 5 5 5 5 5 4 9° -
1» 2°°1 5 5 5 5 5 5 5 5 5 5 ‘ 5-30/50 <
0' a a a a a a a a a v 6 - 30/30 < 9m fill)
a T v d d d d d f d d f ,
L a f p f f v d f f v 7 I 30/0 < Silk)
o a a r v a r a a a a
> 1001 a a a a a a f v I v 4 8 - 30/0 < 9m till)
d d d I d f 4 v d v
< a a d a a v t d d a 9 a 15/0
f f d d p p fl ' t d
5 5 5 5 5 5 5 5 5 5 10 - ‘5/15
E a E 5 1 fl 5 1 fl 5 11 - Auto N
3 9
Treatment Number

 

 

 

Figure 74. Average annual revenue for the 11 strategies evaluated for the 60-year
continuous simulation.

 

 

 

 

 

600

/'\ - E ‘
‘I‘ 500— ° a‘ x at

g . . 1
#9 400- _
V o o N

g ‘ 1:1 180 pp '
ac) 3004 A 180 sd _
> 0 90 Sd

cg ‘ x 30/50 (< silk)

d) 200- 3K 350/30 (< .grn fill) _.
0" V 430/0 (< Silk)

E ‘ 19 30/0 (< grn fill)

3)) 100- El 15/0 _
<1: A 15/15

‘ 0 Auto N
O
C l I I l l l I
40 50 60 70 80 90 100 1 10 120
Average Nitrate Leached
(kg N ho—1)

 

 

 

Figure 75. Average annual revenue and annual nitrate leaching for the 11 strategies
evaluated for the 60-year continuous simulation.

141

Once again, the 90 sd strategy must be considered as a viable strategy to minimize nitrate leaching
while maintaining a profitable yield.

To investigate a possible improvement on the 90 sd strategy, additional sidedress strategies
were simulated. These strategies include a 100, 110, 120, 130, 140, and 150 kg N ha". In each
case, 30 kg N ha'1 was applied at planting and the remainder of the nitrogen was applied at V6,
the same time the second application of the 90 sd was made. This was done to determine if there
existed an amount between the 90 sd and the 180 pp strategy that yielded the same or greater
revenue than the 180 pp strategy but still minimized the nitrate leached.

The yield data for the different sidedress strategies is shown in Figure 76. The yields
steadily increased as the amount of nitrogen fertilizer added increased until about 130 kg N ha'l.
At this point, the yield differences between the strategies became extremely small. The CVs for
the yields are shown given in Table 31. The nitrate leaching data are shown in Figure 77. The
amount of nitrate leached increased as the amount of nitrogen fertilizer applied increased.

The data on the revenue for the strategies are shown in Figure 78. The revenue graph
followed more closely the yield data shown in Figure 76. However, as the amount of nitrogen
applied increased past 130 kg N ha'l, the amount of revenue actually decreased. This is because
the extra profit from the increase in yield cannot pay for the extra cost in nitrogen fertilizer.

Finally, the revenue data versus the leaching data are shown in Figure 79. The data
points formed an arc within the graph, with the 130 kg N ha“1 at the top. Though the 130 sd
strategy does provide for about 327 ha‘1 more in revenue over the 90 sd strategy, it also leaches an
additional 22 kg N ha'1 of nitrate. A possible optimal strategy taken from a path along the curve
where the curvature is the greatest at about the 110 sd treatment. Note that compared to the 90
sd treatment, the 110 sd treatment gives 263 kg ha'1 more yield and $9.65 ha'1 more revenue, with
only about 4 more kg N ha'1 leaching.

As previously mentioned, the PRF-30m strategy yielded a slightly higher revenue than
the 90 sd strategy, but with some additional nitrate leaching. The 110 sd strategy also yielded a

slightly higher average revenue than the 90 sd strategy, but with only a minimal increase in

 

 

 

142

 

 

 

 

 

 

 

 

 

 

 

r
d m. d
n w n
m. w.
u m m
m «magnum m m umwmwmu
.4. ”meme m .m. amass
Mun-u...- os w.....-....-
A” 1234567 m n 1234567
q 1 - - 4 u a 1 d
we
555555555555555555555 m
.s...
.55555555555555555555 m
4. w r
e
55555555555555555555: m m 2.5 m
U s
N m M
ra:ataav55t:dardrtartsastagagfl cm W @144 it
e m.
m m m
5555555555555555555 m r g1... m
.n m m
T
d
.55aaaa=5aararvv‘a‘r‘v’r"! ”W. Egg: 2
.n
45555555555’55555455 m n. 1
e .m
mm
1 J‘ 1 - 4 - _ o AVV m . q 4 — 4 - 4 < T #
Mam: 4.8:...
. m
:12. 9; m m “To; 2 9;
0
Bo; £80 3803. New 32003 29.22 00803.
H w

 

 

Figure 77. Average annual nitrate leaching for the seven sidedressed strategies evaluated for

the 60-year continuous simulation.

 

143

 

 

Treatment Legend

 

 

can

 

m.w.m.m m

Clo: 3 03:93. ooo..o><

 

Treatment Number

 

 

Figure 78. Average annual revenue for the seven sidedressed strategies evaluated for the

60-year continuous simulation.

 

 

 

 

 

 

...._...._..4._...4.........._....4444..
dddddd.

v dSSSSSSl
5000000.

00112345.

914111111.

ODAOXXVI

.- .

x H

o .1

A 4

u .

O 4
414444421....4....4444....4_4444_4444
C 5 O 5 0 5 0 5 C
4 3 3 2 2 1 4| 0 0
5 5 5 5 5 5 5 5 5

flag 3 osco>om moono><

90

85

80

75

7O

65

60

55

50

Average Nitrate Leached

_1)

(kg N ho

 

 

Figure 79. Average annual revenue and nitrate leaching for the seven sidedressed

strategies evaluated for the 60—year continuous simulation.

144

nitrate leaching. To help compare these strategies, the annual revenues for the 60 year period
were sorted and plotted as cumulative probabilities in Figure 80. As seen in Figure 80, all of the
strategies obtained approximately the same revenue at 100% cumulative probability. As the
cumulative probability decreases, the curve of 90 sd strategy begins to diverge from the other two
curves. This shows the increase in variability in the 90 sd strategy. It also shows that the 110 sd
and the PRF-3030 strategies have virtually the same probability of revenue. The strategy that
should be chosen in Figure 80 is the strategy that lies furthest to the 5 right. However, since the
110 sd and the PRF-30m strategy cannot be distinguished from each other, either of these
strategies can be chosen based on cumulative probability. However, based on revenue, leaching,
and management required, the 110 sd strategy would be the strategy of choice.

Table 31. Coefficient of variation expressed as a percentage for grain yields for a 60 years
CERES-1M simulation run.

Coeflicient of Variation

 

   
  
  
  
  
  
 

 

 

 

 

 

 

 

 

  

 

Strategy

90 sd 19.71
I 100 sd 18.88

110 sd 18.27

120 sd 17.70

130 sd 16.71

140 sd 16.29

150 sd 16.07
_ J

 

 

145

 

 

     

 

 

 

120
g 100~ _
3. 4
~35 80— -
o
.0 '1
9
0. 60+ "‘
0
.Z
:5 40- -
3
E
3 20... H110 sd ..
0 H90 sd
1 9'8 30/30 (< silk)
o-,.,.,... f,.
200 300 400 500 600 700 800 900
Revenue ($ fro-'1)

 

 

 

Figure 80. Cumulative probability of the annual revenue for selected strategies for the 60-year
continuous simulation.

SUMMARY AND CONCLUSIONS
Objective 1: Develop a computer program that adequately simulates inbred maize growth and
development as well as soil-plant interactions for water and nitrogen dynamics.

The CERES-1M model developed for this study has shown a good ability to simulate
inbred growth and development. Using the CERES-Maize model as a basis has allowed for the
development of a model that can actually simulate both inbred and hybrid maize.

In order to simulate inbred maize growth and development, CERES-Maize was first
reprogrammed to simulate the unique field operations of seed corn production such as detasseling,
the removal of the male plants, and the designation and interaction of male and female plants.
The removal of the male plants was easily handled by changing code to end plant growth on the
date of removal. The detasseling routine required more integration with other components of the
model but the change was not drastic. The interaction of the male and female plants did,
however, pose several questions.

In seed corn production, the male plants are grown only to provide the necessary pollen
needed to pollinate the female plants. The actual yield of the male plants is of no concern to
most growers, though some companies allow the growers to harvest the male rows and sell the
grain. A decision needed to be made whether or not to simulate the growth and development of
the male plants. After much discussion, it was decided that growth of the male plants be
simulated to take into account any soil nitrogen effect the male plants might have. Further study
of the interaction between the male and female plants should be made. The methodology used in
CERES-1M is simplistic but may be all that is necessary. However, if the male inbred is drastically

different from the female inbred, the interaction between the plants may be of more importance.

146

147

One of the major changes to the CERES-1M model was the incorporation of a new leaf
area relationship. Integrating individual leaf growth with plant leaf area gives the model more
flexibility. For instance, the detasseling routine could be rewritten to allow for the specific loss of
the biomass of the last three leaves at the time of detasseling. In its present form, CERES-1M
reduces biomass based on a percentage of total above ground biomass. A more precise reduction
may help to avoid the problems of biomass simulation as seen in the inbred comparisons shown
earlier.

The refining of the nitrogen deficit factors should continue as more is learned about the
effects of stress on inbred maize growth and development. Most growers of seed corn agree that
the inbred varieties are less tolerant of water and nitrogen stress, though little has been done to
prove this. Work done is this area would help to clarify this suspicion and also help to refine the
model’s ability to simulate the effects of water and nitrogen stress on inbred maize, if appropriate.

All of the other changes made to either accommodate seed corn production or to improve
the empirical formulas of CERES-Maize to work well. The new CERES-1M model should be
considered a first version from which to make improvements. It can be use to simulate hybrid or
inbred maize, single season or continuous, multiple years. These new options give the model
flexibility in its usefulness. In the future, work should continue to refine and simplify the code.

Also, user interfaces need to be improved to keep the model user friendly.

Objective 2: To perform field experiments to validate the simulation model.

The validation of CERES-[M was presented in two phases. First, data from a hybrid
study were used to help validate the models ability to simulate nitrate leaching. This hybrid study
was done in a field that was under cultivation for many years. This helped to provide historical
data on crop yield and management. Also, the lysimeters had been in placed for a few years
before intensive data collecting began. Waiting a few years after the lysimeters were installed

before collecting the data minimized the impact of any soil disturbances caused by the installation.

148
For these two reasons, data from the hybrid study compared fairly well with values simulated by

CERES-1M.

Though 'on-farm' research is normally more applicable than research conducted in
greenhouses or on small research plots, problems can occur when working with growers. The
problem experienced with the nitrogen application in the hybrid study in 1988 are not uncommon.
However, good communication and trust between the researcher and the farmer can usually
minimize these errors. It is imperative that the participating farmer understand that though it is
important to apply the correct amount, it is just as important to know how much was applied,
even if the amount was incorrect.

The inbred validation phase of this study was key to the development of CERES-1M. The
data collected over the two year period was designed to aid in the development of the model as
well as to determine the impact of nitrogen management in seed corn production. Of all the data
collected, the soils data proved to be the most difficult to quantify.

The soil nitrate and ammonium data collected throughout the inbred study have shown
inconsistencies with large variations between samples. At the time of the installation of the
lysimeters, measurements were taken to determine soil organic matter content and soil pH. Due
to the d'mturbance caused by the installation of the lysimeters, these measurements probably
should have been taken yearly, at least for the first few years.

The amount of nitrate in the soil is dependent on a number of factors including microbial
activity, mineralization, and denitrification. Sampling the soil within the lysimeters would cause
sink holes to develop which would cause errors in the amount of drainage water and its nitrate
content. However, sampling around the lysimeters must also be limited because the plot area is
not infinitely large and after several years of research, the soil samples collected would not
represent the actual soil conditions in the plot area. With these two limitation, soil sampling
procedures have been chosen to limit sampling size and number while still trying to maintain some

sense of accuracy. Unfortunately, Michigan soils are not homogenous and limiting sample size or

149

number can introduce large errors. However, the sampling procedure used is probably the most
optimal given the limitations of this study. In future years, as in the past, the soil sampling
procedures should be reviewed yearly and changes should be made when appropriate. The
investments of time, money, and people in this project are large and this study will undoubtedly
continue for several more years.

The comparisons between CERES-1M and the inbred data showed a good correlation for
most of the parameters studied. The yield and kernel data compared well as did the grain
nitrogen. The stover data, however, did not compare as well. One problem is that there is no
mechanism in CERES-1M to take into account the biomass production and nitrogen uptake by the
cob. By most measures, cob data is inconsequential when compared to the stover or grain.
However, this seemingly small amount becomes a more integral component when simulating
inbred maize. The overall plant size is smaller than in hybrid maize, making cob biomass a larger
contributor to the overall above ground biomass. Changes should be made to separate cob
biomass accumulation from the stover in CEREScIM.

The stover nitrogen data showed the poorest comparison of all the components. CERES-
IM consistently underestimated the percent nitrogen in the stover. This in turn gave a poor
comparison for the nitrogen uptake by the stover. However, there is large variation in the
measured values as seen in Table 23. According to researchers at Pioneer Hi-Bred Int., this is not
uncommon. Inbred maize normally shows large variations in both yield and biomass. Often, the
detasseling operation is performed two or three times on a single field because plant height is
irregular and some tassels are not removed with just one pass. This is especially true for when the
tassels are removed mechanically. With this inherent variation, it is difficult to determine the
sources of error in the simulation. At the present, stover samples are taken with leaves and stems
mixed together. It may be useful if these were separated since CERES-1M simulates the growth of
these two components separately. Also, it may be useful for CERES-1M to be reprogrammed to
simulate nitrogen uptake by the leaves and stems separately. In its present form, CERES-1M only

simulates stover nitrogen uptake.

150

Another area that requires more investigation is the detasseling Operation and its effect
on yield and biomass production. Presently, CERES-1M uses a percent reduction in biomass loss
to simulate the loss of leaves due to detasseling. A more precise method may be to use the new
leaf area model to follow the biomass accumulation by individual leaves. Also, more work needs
to be done in the impact of detasseling on carbon assimilation, if any. Undoubtedly, the
assimilation of carbon will be reduced by the loss of leaf area. However, detasseling may cause a
shock to the entire plant system causing a temporary reduction the assimilation rate. Work should
also continue on the effects of timeliness of the detasseling operation on grain yield. Late
detasseling is rarely practiced since one of the companies major goals is to assure the quality of
their seed. However, if the female plants are detasseled too early how and to what degree will this
impact yield?

Future research should also focus on the impact of the male plants on nitrate leaching.
Work done within St. Joseph county has shown that minimal amounts of nitrogen fertilizer can be
applied to the male plants without causing a reduction in female yield. Data from Pioneer Hi-
Bred Intl. confirm this, showing that a moderate nitrogen stress does not effect the quantity or
quality of the male pollen. However, steps should be taken to assure that the nitrogen stress is
not too severe to cause a delay in phasic development which can cause poor pollination. If male
plants can be given limited amounts of nitrogen fertilizer, this should reduce the amount of nitrate
leaching. Work can be done both in the field and using CERES-1M to try to study the impact of
this strategy. With male plants accounting for 20% - 25% of the entire field, the potential impact
on reducing nitrate leaching may be great.

As work continues at the lysimeter site and at other research sites, CERES-M should be
reprogrammed and refined. The model presented here will act as a base for others to work from.
Like the original CERES-Maize V1.0, CERES-1M will continually be challenged and, if
appropriate, changed. With the concern for environmental integrity, companies like Pioneer Hi-
Bred Intl. are becoming more keenly aware of responsibility to help provide their growers with the

most up-to-date information possible. Models such as CERES-1M can help demonstrate the

151

effects of management strategies on grain yield and nitrate leaching. Combined with other
information, growers will be able to make more informed decisions about which management

strategy best fits their overall farm management scheme.

Objective 3: Utilize the simulation model to evaluate the impact of several nitrogen
management strategies on potential leaching in seed corn production.

The evaluation of the management strategies in this section provided some interesting
information but provided even more stimulating questions. The PRF threshold values of 15 kg
and 30 kg were more or less arbitrarily selected. Results showed that the 30 kg value was too high
since the PRF-30/0 strategies showed a significant decrease in yield as compared to the PRF-3030
strategies. Also, the 15 kg value may be too low, though it is difficult to determine because no
yield loss was observed between the PRF-15M and the PRF-15/15 strategies. Additional
simulations could be made to determine a proper threshold value.

The simulations evaluated in this study seemed to indicate that the sidedress strategy was
as good if not better than any PRF strategy. In terms of leaching, the 90 sd strategy did leach less
than any PRF strategy. However, all of the PRF strategies yielded higher, except for the PRF-30M
strategy. Only one PRF strategy, PRF-3030 (with the silking day limitation) produced a higher
revenue.

Continued analysis of the sidedress strategy lead to the discovery that the 90 sd strategy
was still not the most optimal, but that perhaps a strategy between 110 - 120 sd will probably yield
the highest revenue while keeping nitrate leaching to a relatively low amount. However, other
factors must be considered. .

First and most importantly is the initial conditions of the field in question. Given the
initial conditions used in this study the 90 sd seems to produce the most reasonable revenue with
the minimal amount of leaching. However, the 110 sd strategy could produce an extra $20.00 ha'1
over the 90 sd strategy while leaching only about 4 kg N ha'1 more. Additionally, the CV for the

yield of the 90 sd strategy was 19.71 and only 18.27 for the 110 sd strategy. Furthermore, the CV

152
for the yield of the PRF-3030 (<silk) was only 16.54. If a grower was concerned about yield

stability, then the PRF strategy should be chosen. However, the success of the sidedress strategy
presented here has also been recorded by a recent PRF study with a seed corn grower in St.
Joseph county.

In 1990, a field study was conducted to test the applicability of a PRF strategy on a field
scale. Using the MSR and the plant height measurements described earlier, an eight hectare field
was fertilized according to a PRF strategy similar to the PRF- Auto N strategy evaluated earlier.
The farmer asked to participate in the study had been a seed corn grower for Pioneer Hi-Bred
Intl. for several years and had always produwd high yields. Additionally, this particular farmer
had done his own nitrogen fertilizer tests and his annual nitrogen fertilizer application was one of
the lowest in the county.

The farmer’s basic nitrogen management scheme was to apply a nominal amount of
nitrogen fertilizer at planting time. Then, a second applimtion would be applied at cultivation
time, approximately growth stage V6-V8. A few years ago, he was applying a total of 180 kg N ha'
1 and wanted to find out if he could lower the amount of nitrogen fertilizer without yield loss.
Each year he would set aside a small portion of his field and apply 10 kg of nitrogen less and then
compare the yields. It was not until he was down to 110 kg N ha'1 that a yield decrease was
found. Therefore, his new strategy is to apply about 10-30 kg N ha'1 at planting and the
remainder at cultivation for a total of 120 kg N ha'l. Yield records at Pioneer hi-Bred Intl. have
shown his average yields to be always above the county average and among some of the top in the
county.

In the PRF study, the farmer applied 24 kg N be11 at planting and then an additional 90
kg N hr1 at cultivation to all of his fields except the eight hectares under the PRF strategy. Using
plant height as an indicator of stress, a nitrogen deficiency was detected in the plants in the PRF
field. Unfortunately, heavy rains over the next few days caused a delay in the nitrogen application.
When the yield results came in, the farmers strategy had yielded about 0.064 kg ha’1 (4 bu ac‘l)

more than the PRF strategy. Though this amount is small, the formulas used by the seed

153

companies to compensate the farmers for the expected lower yields of the inbreeds make this a
significant difference. In this study, using a simple economic analysis similar to that used with the
60 years analyses, the farmer made an additional $60 ha'l. If the PRF fertilizer application could
have applied in a more timely manner, this difference may not have existed. However, it is
interesting to see that the evaluations of the various nitrogen management strategies performed
earlier confirm the results of the farmers own 'on-farm' nitrogen management tests.

Data from the literature as well as that recorded in Michigan suggest that a zero nitrate
leaching strategy is non-obtainable in most inbred maize fields using present management
strategies. The zero nitrogen fertilizer strategy simulated here indicates that at best, this soil
under inbred maize production will leach and average of 50 kg N ha‘I. Perhaps if the initial
conditions were changed or if the biomass at the end of a season were removed instead of plowed
into the soil, then the total organic N content of the soil could be reduwd to a level which might
approach a zero nitrate leaching level. Models such as CERES-Maize and CERES-1M could be
reprogrammed to look at the posSlhle strategies that might accomplish this goal. However, it
seems imposslble that maize can be grown without some minimal amount of nitrate leaching.
There was likely some nitrate leaching before the land was cultivated.

The question then becomes, are seed corn growers applying too much nitrogen fertilizer?
A 1990 survey of seed corn growers in St. Joseph county indicated the answer is yes. According to
the survey, growers applied an average of 180 kg N ha'1 to their md corn fields. Data from the
study presented here suggests that this is too much. However, the variability data also presented
here show that the growers are following are strategy of low variability, not necessarily high
revenue. Since farmers are traditionally a risk adverse group, this over application of nitrogen
fertilizer is not unexpected. If farmers were to lower their nitrogen fertilizer inputs to help limit
nitrate leaching losses, they would be accepting a greater risks. The question now becomes who

should bear the burden of this risk?

154

The use of models such as CERES-1M help to show gaps in our knowledge base as well as
help to answer some 'what it" questions. With the power of personal computers increasing almost
daily, growth simulation models will gain more common use within the agricultural community.
Models such as CERES-1M can help to educate and inform agricultural and non-agricultural
clientele regarding the effects of management schemes on plant growth and the environment.
Hopefully, the data provided by models such as these can help farmers and policy makers make

more informed decisions.

APPENDICES

APPENDIX A

APPENDIX A

A listing of the subroutine DETASS, that simulates the detasseling of
the female plants.

0000000000

0000000

000000

SUBROUTINE DETASS

*ttttttttt DETASSELING 30330011“; *twtttttitt*ttttttttwttt********

Version 2.18:

By: E. Martin
January, 1991
Called by: GROSUB Program

include 'maizl.blk'
include 'mai22.blk'
include 'maiz4.blk'
include 'comibs.blk'

REDUCE LEAF NUMBER AND PLA DUE TO LOSS OF PLA

I? ISTAGE LE TO 3, THEN LOSS OF BIOMASS DUE TO DETASSELING IS
APPROXIMATELY 14.0% OF THE TOTAL PLANT BIOMASS

REDUCE LEAF NUMBER BY 3

IF (ISTAGE .LE. 3) THEN

LN - LN-3

LPWT- LFWT -((Lrwr+sruwr)*.14)

pLa- PLA-((LFWT*.14)**.8*267)
CALL CALDAT

IF (IPHOUT) THEN
write (*,100)ND,Month
write (NOUTl,100)ND,Month

ENDIF

IDET s DOY

idett = l

RETURN
ELSE

I! ISTAOE IS GREATER THAN 3, THEN LOSS OF BIOMASS DUE TO
DETASSELING IS APPROXIMATELY 10.5‘ OF THE TOTAL BIOMASS

LEAF LOSS REMAINS AT 3

155

APPENDIX A

156

A listing of the subroutine DETASS (cont.)

C

LN - LN-3
LEW? - LFWT -((LFWT+STHWT)*.105)
PLA - PLA-((LFWT*.105)**.8*267)
CALL CALDAT

I? (IPHOUT) THEN

write (*,100)ND,Month

write (NOUTl,100)ND,Honth
ENDIP

IDET - DOY

idett - l

RETURN
ENDIF

100 tonnar(1x,r2,1x,n3,'

END

FEMALE PLANTS DETASSELED',)

A listing of the subroutine MCUT that simulates the removal of the male
plants

000000000000

100 EORMAT(1X,IZ,1X,A3,'

+

SUBROUTINE MCUT

********* SUBROUTINE TO DETERMINE MALE PLANT REMOVAL

Version 2.018

Added

for CERES-1M

Written by: E. Martin

February 12, 1991

Called by Phenol

include 'maizl.blk'
include 'mai22.blk'
include 'comibs.blk'

CALL CALDAT

Ir (tracer) THEN
warrs(*,100)uc,nonra
warrn(uour1,100)nn,nonrn

ENDIF
ISTAGE

8 6

IMCUT-DOY

RETURN

REMOVAL 0? MALE PLANTS.')

END

*************

GROWTH rsaurnarsn DUE TO'./.'

157

APPENDIX A

A listing of the subroutine LEAPAR that simulates the growth of plant
leaf area. This subroutine fills the array used to predict leaf area.

f3CIO¢3CIO<3CIOf3C§0

0 0(1‘) 0f)() 0

0()0

100

SURROUTINE LEAFAR

********* SUBROUTINE TO DETERMINE LEAP AREA GROWTH *************
Version 2.018

Added for CERES-1M

Written by: E. Martin and J. T. Ritchie
May 1991

Called by STAGEI STAGEZ STAGE3 STAGE4 GROSUB

include 'mai22.b1k'
include 'mai23.blk'
include ‘maiz4.blk'
include 'ed.blk'

TAREA-0.0

**** Calculate Area of Largest Leaf

armax - 1200 * P6 * (EXP(-8.08*EXP(-O.193*xmaxlf)))
Calculate the Area of Individual Leaves

DO 100 I-1,TLNO
if (i .le. xmaxlf) then
ease coupggrz ease
PLAl - 1200 * P6 * (ExP(-8.08*EXP(-0.l93*I)))
else
seen LINEAR sens
PLAl - armax - 72.80*(i-xmaxlf)
ENDIF

Total Up the Areas and Store them in an Array

TAREA-TAREA+PLA1
menu I) - TAREA

CONTINUE

RETURN

END

158

APPENDIX A

A listing of the subroutine LFINTER that simulates the growth of leaf

area.

This subroutine is used to read the array already filled by the

subroutine LEAPAR.

00000000000

000

000

000

90
100
200
300

SUBROUTINE LFINTER

tittttttfi SUBROUTINE To DETERMINE W m GROWTH *************
Version 2.018

Added for CERES-1M

Written by: E. Martin and J. T. Ritchie May 1991

Called by STAGEI STAGEZ STAGE3 STAGE4 GROSUB

include 'maizl.blk'
include 'mai22.blk'
include 'mai23.blk'
include 'maiz4.blk'
include 'ntrc1.blk'

Reduced the Leaf Number to Obtain the Correct Area

IP(XN .LE. 6) THEN
XXN I XN-xn/3
ELSE
XXN I XN-Z
ENDIF
I .. AINT(XXN)

Perform a Linear Interpolation on the Array 'AREALP"

PLANE“ I AREALF(I)+((xxn-i)*(arealf(i+1)-arealf(i)))
PLAG I PLANEW - PLA
Ir (PLAG .LE. 0) PLAG - 0.0

Reduce the Plant Leaf Growth if Appropriate

IP(ISTAGE .EQ. 1) THEN
PLAG I PLAG * SWDFZ
ELSE
PLAG I PLAG *AMIN1(NDEP2,SWDP2)
ENDIF
PLA-PLA+PLAG
Call Detasseling Subroutine if Appropriate
I? (ISEX .EQ. 2) THEN
I! (DO! .EQ. IDET) THEN
CALL DETASS
ELSE I! (IDET .89. 999) THEN
IP(SUMDTT .GE. O.90*P3) THEN
CALL DETASS
ENDIF
ENDIF
ENDIF
RETURN
format(/,2x,i2,2(2x,f7.2),2x,i2)
format(33,3(3x,f10.2))
format(2x,4(f10.2,3x))
END

159

APPENDIX A

A listing of the subroutine SNOW that simulates the melting of snowfall.

SUBROUTINE SNOWEALL (TEMPMX, PRECIP, RAIN, SNOW)

Modified by A. Gerakis 8-20-91

0000

REAL TEMPMX, PRECIP, RAIN, SNOMLT, SNOW

I! (TEMPMX.GT.1.) THEN
SNOMLTITEMPMX+RAIN*O.4
I! (SNOMLT.GT.8NOW) THEN

SNOMLT I SNOW
ENDIF
SNOWISNOW-SNOMLT
PRECIPIPRECIP+snomlt

ELSE
SNOWISNOW+RAIN
PRECIP I PRECIP - RAIN
RAINIO.

ENDI!

RETURN

END

APPENDD( B

APPENDIX 3

SUMMARY OUTPUT

RUN

INST_ID

EXP.
TRT.

Output files of CERES-1M (CERES-MAIZE)

1 OUTPUT SUMMARY

WEATHER
SOIL
VARIETY
IRRIG.

:PO2

:MS SITE ID: CN

- 180PP

_ EXPT_NO: 01 YEAR : 1989 TRT_NO:
:1989 LYSIMETER EXP INBRED 1 (CONSTANT)
:In 1 PE

:1989 Constant, MI
:ELSTON SANDY LOAM,

LATITUDE I 41.50, SOWING DEPTH 8

PLANT POPULATION I
GENETIC SPECIFIC CONSTANTS P1 I3200.00 P2 3
I1.15 G3 I 7.450 ISEX ' 2 EzM RATIO 34.0

SOIL PROFILE DATA [ LOCATION:
.13 U8

SOIL ALBEDO'

G2

.035
.035
.037
.037
.037
.028
.020
.019
.019
.019

5.4

Units are in

xe/nn

DEPTH-CM

Oe- lSe
15s- 30c
30c- ‘Se
45s- 60e
60e- 75c
7Se- 90s
90.- 120.

120.- 150s
150.- 180.
18°e- 210s
T 0.- 210.

* NOTE:
FERTILIZER INPUTS
DA! 0’ YEAR

136

180.00

5.

:ACCORDING TO THE FIELD SCHEDULE.

CM:

5.2 PLANTS/SQ METER

CONSTANT

8.0 SWCON8

.177
.177
.176
.176
.176
.134
.110
.105
.105
.105

28.0

.385
.385
.385
.385
.385
.375
.353
O 353
.353
.353

76.9

kg / hectare.

DEPTH

1.00

.40

L0 LIM UP LIM SAT SW EXT SW

.142
.142
.139
.139
.139
.106
.090
.086
.086
.086

22.5

SOURCE

.30

IN SW

.177
.177
.176
.176
.176
.134
.110
.105
.105
.105

28.0

NR

1.000

.900
.700
.700
.500
.080
.040
.005
.002
.002

UREA AMMONIUM NITRATE

160

P5 I650.00 P6 8

RUNOPF CURVE NO.-

N03

1

QHQHHOO

1

.85

78.0

NH4

-m9/k9---
2.3

161

APPENDIX B

SUMMAR! OUTPUT (cont.)

THE PROGRAM STARTED ON, 121 DAY OF YEAR

DATE CDTT PHENOLOGICAL STAGE BIOM LAI NUPTK Ni CET RAIN PESW

16 May 0. SOWING g/m‘2 kg/ha ---mm---- cm
17 May 5. GERMINATION 33. 20. 21.
23 May 47. EMERGENCE 13. 17. 21.
10 Jun 246. END JUVENILE 3. .05 1.1 3.92 63. 205. 26.
16 Jun 300. TASSEL INITIATION 5. .12 2.3 4.18 76. 226. 25.
22 Jul FEMALE PLANTS DETASSELED

24 Jul 840. 75% SILKING 371. 1.55 87.6 2.36 218. 317. 18.
4 Aug 1007. BEGIN GRAIN FILL 571. 1.41 86.7 2.35 269. 421. 24.
9 Sep 1463. END GRAIN PILL 1016. .61 27.1 .99 395. 510. 19.
1 Sep 1491. PHYS. MATURITY 1016. .61 27.1 .99 399. 510. 19.

YIELD (xG/RA)- 6390. (BU/AC)I101.7 FINAL Gpsn- 2318.
RERNEL WT.(mg)I232.9 .

ISTAGE CSDl CSD2 CNSDl CNSD2 S T A G E OF G R O W T H
1 .00 .00 .00 .00 EMERG to END JUVENILE PHASE
2 .00 .00 .00 .00 END JUV to TASSEL INITIAT
3 .00 .00 .00 .00 TASSEL INITIATION to SILK
4 .00 .00 .00 .00 SILRING to BEGIN GRAIN FILL
5 .00 .00 .00 .00 GRAIN FILLING PHASE

* NOTE: In the above table, 0.0 represents minimum
stress and 1.0 represents maximum stress for water (CSD)
and nitrogen (CNSD), respectively.

5 IRRIGATION APPLICATIONS AT 1.00 EFFICIENCY

DAY OF YR 200 203 207 208 214
AMOUNT mm 16. 32. 37. 1. 51.

IRRIGATION TNIs SEASON 2 137. mm
PREDICTED OBSERVED

SILKING DATE 205 206
MATURITY DATE 254 270
GRAIN YIELD (xG/NA) 6390. 6551.
KERNEL WEIGHT (G) .233 .233
GRAINS PER sq METRE 2318. 2376.
GRAINS PER EAR 450.07 461.00
MAX. LAI 1.55 1.69
BIOMASS (xG/RA) 10164. 9545.
sTRAw (KG/HA) 4765. 4009.
GRAIN Ns 1.70 1.69
TOT N UPTAKE (xG N/RA) 119.0 129.2
sTRAw N UPTAKE 27.1 35.7
GRAIN N UPTAKE 91.8 93.5

162

APPENDIX B

BIOMASS OUTPUT FILE

RUN

1

IN8T_ID

EXP.
TRT.

WEATHER

SOIL

VARIETY
IRRIG.

DAY
OYR
149
156
163
170
177
184
191
198
205
212
219
226
233
240
247

SDTT

59.
143.
217.

35.
147.
240.
362.
453.

10.
113.
211.
294.
380.
475.
555.

In 1 FE - 180PP
:MS SITE_ID: CN EXPT_NO: 01 YEAR : 1989 TRT_NO:
:1989 LYSIMETER EXP INBRED l (CONSTANT)
:In 1 FE - 180PP
:1989 Constant, MI
:ELSTON SANDY LOAM,
:PO2
:ACCORDING To THE FIELD SCHEDULE.

BIO LN LAI ROOT STEM GRAIN LEAF RTD PTF
g/m2 I ----- weight in g ----- i (cm)
2. 3 .00 .23 .20 .00 .20 23. .00
2. 5 .02 .38 .20 .00 .20 41. .00
3. 7 .07 .94 .20 .00 .40 57. .75
8. 9 .17 1.89 .20 .00 1.36 72. .87
32. 12 .53 5.44 .71 .00 5.46 95. .48
89. 14 1.08 14.46 4.20 .00 13.13 117. .47
203. 17 1.59 24.64 17.95 .00 21.37 142. .72
319. 20 1.77 33.22 37.53 .00 24.46 164. .38
371. 17 1.55 35.48 42.70 .00 20.71 185. 1.00
490. 17 1.44 38.01 50.00 .00 18.43 208. .49
612. 17 1.37 40.65 54.96 7.90 16.69 210. .88
713. 17 1.32 39.87 54.96 27.72 16.58 210. 1.00
813. 17 1.23 39.06 54.29 47.90 16.46 210. .94
897. 17 1.06 37.90 49.30 69.32 16.35 210. 1.00
979. 17 .85 36.73 45.72 88.95 16.23 210. .99

1
L1 L3 L5
I" RLV-'l
.1 .0 .0
.1 .0 .0
.2 .1 Co
.3 .2 .0
.8 .5 .3
2.0 1.3 .8
3.4 2.3 1.6
4.3 3.0 2.1
4.7 3.5 2.5
4.7 3.7 2.6
4.8 4.2 3.0
4.7 4.2 3.0
4.7 4.1 3.0
4.6 4.0 3.0
4.5 4.0 2.9

163

APPENDIX B

SOIL WATER OUTPUT FILE

RUN

INST_ID :MS SITE ID: CN

- 180PP
EXPT NO: 01 YEAR : 1989 TRT_NO:

In 1 FE

 

EXP. :1989 LYSIMETER ExP INERED 1 (CONSTANT)

TRT. :In 1 FE - 180PP

WEATHER :1989 Constant, MI

SOIL :ELSTON SANDY LOAM,

VARIETY :P02

IRRIG. :ACCORDING TO THE FIELD SCHEDULE.

* Units are in MJ/square meter.
AVERAGE --------- PERIOD SW CONTENT W/DEPTH

DAY EP ET E0 SR* MAX MIN PREC SW1 SW2 SW3 SW4
OYR (mm) (mm) (mm) c 0 (mm)
127 .0 2.8 3.5 19. 13.2 2.0 8.40 .13 .15 .16 .16
134 .0 1.3 3.2 16. 16.4 5.0 10.50 .15 .15 .16 .16
141 .0 2.0 4.0 18. 22.5 9.5 17.30 .15 .16 .17 .16
148 .0 2.7 4.3 19. 22.9 13.1 16.00 .13 .15 .16 .16
155 .0 3.5 3.9 16. 24.5 14.9 172.30 .21 .22 .25 .24
162 .2 1.5 5.6 24. 25.3 11.9 .00 .17 .17 .18 .18
169 .2 2.6 3.3 15. 21.7 13.5 20.80 .15 .16 .17 .18
176 1.2 2.0 5.1 20. 29.0 17.5 .00 .12 .14 .15 .15
183 2.7 3.9 5.7 24. 27.8 15.7 6.90 .09 .11 .12 .13
190 4.1 4.9 6.1 25. 31.3 17.8 24.90 .17 .11 .08 .09
197 3.0 4.6 4.6 20. 27.0 17.3 .00 .06 .07 .07 .07
204 2.3 3.5 3.6 15. 26.9 16.5 59.70 .21 .17 .14 .07
211 1.9 3.8 3.8 16. 28.7 17.9 49.20 .20 .19 .18 .16
218 2.6 5.3 5.3 23. 29.5 16.7 79.40 .22 .22 .22 .21
225 2.4 4.3 4.3 22. 25.9 11.7 6.90 .16 .16 .17 .16
232 2.6 3.7 3.9 19. 26.5 13.7 20.10 .18 .17 .14 .14
239 1.9 3.9 3.9 19. 27.3 16.2 2.00 .11 .13 .14 .13
246 1.8 2.6 3.1 16. 25.5 14.7 .00 .08 .10 .11 .12
253 1.1 2.5 3.0 15. 26.7 16.8 35.10 .17 .16 .12 .11
260 .0 1.8 2.5 12. 19.3 10.1 29.60 .17 .17 .17 .16
267 .0 1.4 3.5 17. 22.1 6.3 .00 .13 .14 .15 .15
274 .0 .6 3.7 18. 21.5 4.2 .00 .12 .14 .14 .14
281 .0 .5 2.0 11. 15.2 1.9 1.00 .11 .13 .14 .14
288 .0 .5 2.9 14. 22.5 4.2 2.80 .12 .13 .14 .14
295 .0 .7 1.1 7. 9.8 1.7 31.80 .19 .18 .17 .16
302 .0 1.6 2.4 11. 23.4 5.8 .00 .14 .15 .16 .16
309 .0 .6 .8 5. 12.6 2.9 5.90 .16 .15 .15 .16
316 .0 .6 .8 5. 9.1 .9 15.20 .17 .17 .17 .17
323 .0 .4 .6 5. 8.3 -2.3 33.80 .20 .20 .20 .19
330 .0 .8 1.1 9. 6.5 -4.2 .00 .17 .17 .18 .17
337 .0 .4 .5 5. 5.7 -4.6 10.70 .18 .18 .18 .17
344 .0 .1 .2 4. .3 -7.5 .30 .17 .17 .17 .17
351 .0 .1 .1 7. -7.0 -17.7 .00 .17 .17 .17 .16
358 .0 .0 .0 6. -9.9 -20.8 .00 .17 .17 .17 .17
365 .0 .1 .2 4. -.5 -8.6 .00 .16 .16 .17 .17

7 .0 .3 .4 5. 2.1 -5.0 13.00 .18 .18 .18 .18

TOTAL
SW5

.16
.15
.16
.16
.23
.17
.17
.15
.13
.10
.08
.07
.10
.18
.16
.14
.13
.12
.12
.13
.14

.14
01‘
.14
.16
.16
.16
.18
.17
.17
.16
.16
.17
.17
.17

PESW
(cm)

21.4
21.2
21.5
21.2
29.9
25.7
23.5
20.7
18.3
17.3
14.4
18.6
20.8
25.2
22.4
21.1
19.0
17.2
19.1
20.8
19.8
19.4
19.2
19.1
21.8
20.7
20.9
22.0
25.0
23.4
23.3
22.5
22.0
22.0
21.9
23.0

164

APPENDIX B

SOIL WATER OUTPUT FILE (cont.)

* Units are in MJ/square meter.

 

AVERAGE --------- PERIOD SW CONTENT W/DEPTH TOTAL
DAY EP ET EO SR* MAX MIN PREC SW1 SW2 SW3 SW4 SW5 PESW
0“ (an) (an) (m) C C (nun) (cm)
14 .0 .3 .5 5. 3.4 -2.9 13.10 .20 .19 .19 .18 .17 23.8
21 .0 .4 .6 6. 5.1 -6.5 14.30 .18 .18 .19 .18 .17 24.4
28 .0 .3 .4 9. -1.8 -13.9 .00 .17 .17 .18 .17 .16 23.0
35 .0 .8 1.2 10. 7.9 -6.0 .80 .15 .16 .16 .16 .16 21.7
42 .0 .7 1.0 7. 8.2 -.7 .80 .14 .15 .16 .16 .16 21.0
49 e0 e5 e7 7e ‘e3 -5e7 4.00 e15 e15 e16 e16 e16 21c].
56 .0 .6 .8 12. 2.0 -9.0 88.60 .22 .22 .23 .22 .21 27.5
63 .0 .9 1.3 15. -8.8 3.00 .17 .18 .18 .18 .18 25.4
70 .0 .6 .8 9. -3.4 64.90 .27 .26 .25 .22 .20 29.1
77 .0 1.5 2.1 10. 9.9 4.90 .18 .19 .19 .19 .18 26.5
84 .0 1.2 1.7 14. -2.7 30.20 .19 .20 .21 .20 .19 26.3

165

 

l

2

 

APPENDIX E
NITROGEN OUTPUT FILE

RUN 1 In 1 FE - 180PP

INST_ID :MS SITE_ID: CN EXPT_NO: 01 YEAR : 1989 TRT_NO:
EXP. :1989 LYSIMETER EXP INBRED 1 (CONSTANT)

TRT. :In 1 FE - 180PP

WEATHER :1989 Constant, MI

SOIL :ELSTON SANDY LOAM,

VARIETY : P02

IRRIG. :ACCORDING To THE FIELD SCHEDULE.

TOPS NFAC VEG N GRAIN N03 N03 N03 N03 N03

DAY N t UPTR UPTR 1 2 3 4 5 1
OYR - kg N/ha -: ug N/g soil
127 4.40 1.00 .0 .0 14.4 5.6 3.1 3.9 2.6 1.9
134 4.40 1.00 .0 .0 15.0 6.0 3.1 4.4 3.1 1.8
141 4.40 1.00 .0 .0 74.6 17.2 4.0 4.8 3.4 5.8
148 4.47 1.00 .9 .0 72.0 25.4 4.9 5.1 3.7 3.5
155 4.50 1.00 .9 .0 16.5 23.6 23.8 20.4 13.8 2.6
162 4.26 1.00 1.1 .0 18.4 19.8 17.7 17.6 14.9 2.4
169 4.00 1.00 2.7 .0 15.0 17.7 17.1 18.2 17.0 2.0
176 4.00 1.00 10.2 .0 16.2 16.5 15.7 16.8 16.1 1.8
183 3.52 1.00 26.8 .0 15.4 14.9 13.9 14.8 14.8 1.7
190 3.13 1.00 54.2 .0 10.7 14.4 12.1 12.7 12.4 1.5
197 2.73 1.00 76.6 .0 7.4 11.8 11.3 11.6 11.1 1.5
204 2.46 1.00 87.7 .0 2.9 10.0 13.4 11.3 10.9 1.5
211 2.53 1.00 87.1 .0 2.0 6.2 12.2 16.6 12.1 1.6
218 2.36 1.00 81.3 5.2 1.2 3.1 7.6 15.2 18.2 1.7
225 1.95 1.00 64.1 21.6 1.8 3.0 6.0 12.3 17.3 1.6
232 1.53 1.00 51.5 39.2 2.0 3.4 5.6 11.4 16.5 1.5
239 1.34 1.00 44.3 57.9 2.0 2.8 4.8 10.3 15.7 1.5
246 1.17 1.00 35.7 75.7 2.4 2.8 4.3 9.3 14.6 1.4
253 1.04 1.00 27.1 91.8 1.8 3.1 4.2 8.8 13.7 1.4
260 .99 1.00 27.1 91.8 1.6 2.8 4.5 9.5 14.1 1.4
267 .99 1.00 27.1 91.8 2.1 2.9 4.3 9.1 14.4 1.3
274 .99 1.00 27.1 91.8 2.6 3.1 4.3 9.0 14.4 1.2
281 .99 1.00 27.1 91.8 2.9 3.3 4.3 9.0 14.4 1.1
288 .99 1.00 27.1 91.8 3.2 3.5 4.3 9.0 14.5 1.1
295 .99 1.00 27.1 91.8 2.4 3.9 4.8 9.3 14.5 1.0
302 .99 1.00 27.1 91.8 2.8 3.7 4.6 9.0 15.0 1.0
309 .99 1.00 27.1 91.8 3.1 3.9 4.6 9.0 15.1 1.0
316 .99 1.00 27.1 91.8 2.6 4.1 4.8 9.0 15.1 .9
323 .99 1.00 27.1 91.8 1.6 3.2 4.4 8.3 14.7 .8
330 .99 1.00 27.1 91.8 1.6 2.8 3.8 7.1 13.3 .8
337 .99 1.00 27.1 91.8 1.4 2.6 3.6 6.7 12.8 .8
344 .99 1.00 27.1 91.8 1.4 2.6 3.5 6.5 12.2 .8
351 .99 1.00 27.1 91.8 1.4 2.5 3.4 6.3 11.8 .8
358 .99 1.00 27.1 91.8 1.4 2.5 3.4 6.4 12.0 .8
365 .99 1.00 27.1 91.8 1.4 2.5 3.4 6.4 12.0 .8

7 .99 1.00 27.1 91.8 1.1 2.3 3.4 6.4 11.6 .8

HHHHHHHHHHHHHHHHHHHHHHHMHHH
O O O I O O O O 0 0 O O O O O O O O 0 O O O O O O O O

OOOOOHHNNNUbhfihUNNWDMGQDOOUh

NH4 NH4 NH4

3

166

APPENDIX B

NITROGEN OUTPUT FILE (cont.)

TOPS NFAC VEG N GRAIN N03 N03 N03 N03 N03 NH4 NH4 NH4

 

 

.5

DAY N 8 UPTR UPTX 1 2 3 4 5 1 2 3

OYR - kg N/ha -: ug N/g soil :
14 .99 1.00 27.1 91.8 .9 2.1 3.1 6.1 11.2 .8 .8 .5
21 .99 1.00 27.1 91.8 .7 1.5 2.5 5.1 10.2 .8 .8 .5
28 .99 1.00 27.1 91.8 .7 1.5 2.3 4.7 9.5 .8 .8 .5
35 .99 1.00 27.1 91.8 .8 1.4 2.3 4.6 9.1 .8 .8 .5
42 .99 1.00 27.1 91.8 1.0 1.5 2.2 4.5 9.1 .7 .7 .5
49 .99 1.00 27.1 91.8 1.0 1.6 2.3 4.7 9.1 .7 .7 .5
56 .99 1.00 27.1 91.8 .5 1.0 1.7 3.8 8.0 .7 .7
63 .99 1.00 27.1 91.8 .5 .8 1.3 2.5 5.4 .7 .7 .5
70 .99 1.00 27.1 91.8 .4 .6 1.1 2.2 4.8 .7 .7 .5
77 .99 1.00 27.1 91.8 .4 .6 .7 1.4 3.0 .8 .8 .5
84 .99 1.00 27.1 91.8 .4 .5 .7 1.1 2.6 .8 .8 .5

APPENDIX C

167

.oma
.omH
.omH
.omH
.omH
.omd
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH
.omH

NNNNNNNNNNNNNNNNNNNNNNNNNN

madmz th

mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m
mH.m

mfiqm

mm.m
mm.0
om.m
NH.h
Vm.m
mfl.m
Hh.m
Hm.m
hN.h
Nh.w
no.5
bfi.m
Hm.m
b¢.m
mh.m
Hm.o
mm.m
5H.o
mfl.o
om.o
mo.m
mo.m
v0.0
mo.m
vm.m
mm.h

mm.HH
ov.HH
hm.oH
Nm.NH
mm.oH
mo.oH
mn.HH
mv.Ha
mm.MH
hh.HH
vN.NH
OO.HH
om.oH
mm.HH
mm.fiH
mm.HH
mm.oH
cm.HH
mN.HH
mN.NH
mm.oH
mm.m

mm.ma
00.HH
hm.oH
Hm.MH

--- .e\e --

DAM

WZOHm

.Hmv .emv
.omm .mec
.oom .hee
.hme .oNv
.mNm .ooe
.mhm .m¢¢
.NNm .mnm
.mom .NQ?
.Nom .bwv
.mem .va
.mmm .Om¢
.va .hmv
.mmm .vhv
.Hmm .hov
.mom .m¢v
.Nmm .moe
.vmm .bwv
.vom .vmc
.oom .bhv
.ONv .Oov
.Nme .mo¢
.mmm .mmq
.mov .MOm
.Omm .h¢¢
.vom .mwv
.mmo .mmm
Ill ES Ill
2H¢m END

00.
oo.
oo.
oo.
oo.
oo.
co.
co.
oo.
00.
Ho.
00.
Ho.
00.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
00.
Ho.
00.
oo.
oo.

mmam

oo.
oo.
oo.
00.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
00.
00.

at:

.mn

.mm
.hm
.vHH
.mm
.hm

mmH
HOB

OmmOMNMNOHMOMOMI‘VHmf‘OMr-‘lhfl'm

b.0NH
0.0HH
m.mHH
v.m¢H
m.mHH
M.NHH
m.HNH
h.mNH
m.vma
m.~n~
m.NNH
m.MHH
h.¢oH
H.HNH
m.hHa
¢.mmd
m.NHH
w.mNH
m.NNH
H.HMH
m.voH
m.moH
¢.hNH
m.vNH
h.¢HH
«.mmH

e\mx

.mmHz $8032

00.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
00.
Ho.
00.
oo.
00.

m
mmfim

mAHh HDQBDO

oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
00.
Ho.
00.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
oo.
00.

m
8H2

.¢H
.ON
.vH
.HH
.OH
.nH
.mm
.OH
.vN
.mH
.HN
.ma
.m

.Nm
.hH
.m

.0H
.mm
.mm
.HH
.mm
.hN
.mH
.hH
.HH
.hH

e\zox
mquz

mOH Nb
NOH Oh
Nam Ho
do“ Oh
OHH eh
00H 05
VHH on
mHH Hm
mHH mm
mad we
wad ch
HHH mh
NHH Nb
MOH mo
MHH ch
MOH mm
mad he
ONH we
wHH be
MNH Nb
50H 05
med «5
oNH mm
wan me
Can Mb
mHH mm

lmwfioll
210 214

Mmdzzbm mdeIHBADZ

m ”Hazmmmd

APPENDIX C

Inputs used in the hybrid simulations

ID I

WEATHER DATA (1988/1989)

MSCN

Latitude I 42.00
Longitude I 85.50

YR-

X03!

SOL I Solar Radiation (Mj m'z)

TMAX
TMIN
RAIN

ID

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

CVCP

I Maximum Daily Temperature (°C)
I Minimum Daily Temperature (°C)
I Rainfall (mm)

YR DOY SOL TMAX TMIN RAIN DATE
88 1 5.6 - 6.7 -17.1 0.0 01 JAN
88 2 8.7 - 3.4 -18. 0.0 02 JAN
88 3 6.7 - .1 -10.8 0.0 03 JAN
88 4 6.1 - 3.4 -21.9 0.0 04 JAN
88 5 8.5 -14.9 -26.4 0.0 05 JAN
88 6 8.7 -10.2 -26.8 0.0 06 JAN
88 7 5.7 - 8.7 -28.9 0.0 07 JAN
88 8 8.3 - 3.7 -27.8 0.0 08 JAN
88 9 6.2 - 4.4 -19.1 0.0 09 JAN
88 10 8.4 - 7.0 -24.8 0.0 10 JAN
88 11 8.2 .1 -15.9 0.0 11 JAN
88 12 2.3 4.1 - 2.8 0.0 12 JAN
88 13 5.8 - 2.9 -14.2 0.0 13 JAN
88 14 8.0 - 5.5 -20.8 0.0 14 JAN
88 15 8.8 1.5 - 8.9 0.0 15 JAN
88 16 9.3 8.4 - 2.0 0.0 16 JAN
88 17 .9 5.0 1.5 16.2 17 JAN
88 18 8.6 7.4 - 3.4 1.8 18 JAN
88 19 1.3 2.1 - 4.1 3.6 19 JAN
88 20 .8 6.9 - .7 1.3 20 JAN
88 21 6.6 .9 - 5.7 0.3 21 JAN
88 22 3.0 - .1 - 5.8 0.0 22 JAN
88 23 5.9 - 1.3 - 4.4 0.0 23 JAN
88 24 3.5 1.1 - 8.5 0.5 24 JAN
88 25 4.7 - 1.0 -16.5 0.0 25 JAN
88 26 8.4 .6 -16.8 0.3 26 JAN
88 27 10.5 - 6.1 -20.4 0.0 27 JAN
88 28 8.0 1.1 -10.8 1.5 28 JAN
88 29 9.6 8.6 - 5.9 0.5 29 JAN
88 30 2.3 10.2 5.3 '0.0 30 JAN
88 31 .9 10.4 2.1 12.1 31 JAN
88 32 1.0 2.1 - 1.9 2.8 01 FEB
88 33 5.4 - 1.9 - 7.3 0.0 02 FEB

168

IWHMHXC

ID

GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP
GNP

HHH
Hmmhumumnnmqmmmupmmummomepn

H H HP HP
upmmqnommnunqemuommoomooou

HHH

14.9

13.0
17.0
13.3
2m

17.4
13.9

00000000 Hemmwo HMO-‘00 HQGM§DHHMUHH5GQ

H

H H

HHH HHH

uuo

OO.

bedmoqmoemmqmowmehbwuquqmomommmMHHuOM
Illlll

L0

cecccccc
HMUIUQHON

wownmqmn

N

I
H

GOU‘IUIfiNU‘DGUUUmUOU‘GOO‘DQOQ‘DQU’NMNHGHOHUHNIJUGGUIUGHQ

'0 MPH OGUGQ‘GDGGOGG NGHNHISHmGthGGOOOGNOIIUIIDUIQ§

-1
-1

me

E

OI!UH-‘OOGOOOOOOHOOOOO‘CONOOOOOOOOOOOOOOOOOONOOO‘OU‘OOOOOOOOOMO
eeeeeeecceeeeeeeeeeeeeeecceeceeeeceeceeeeeeceeeeeeeeeeeeee

ONHUOOISUOOUOOOOO’O“OUOOUOOOOOOOOOOOOOOOOOUwOOQOdOOOOOOOOOUO

EEEEEESEEEEEEEEEEEEEEEEEEEEEEEE

APPENDIX C

ID

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

CVCP
CVCP
CVCP
CVCP

CVCP

SOL

8.7

4.8
12.3
13.9
21.4

1.5
24.9
24.6
24.1
21.6
21.0
24.5
22.9
21.3
11.7
25.1
22.4
25.1
26.1

19.4
14.5
14.1
19.5
25.8
16.5
4.4
22.8
27.6
27.7
28.1
28.3
20.4
21.1
22.6
28.8
24.1
19.9
13.2
10.2
29.5
14.0
26.5
29.0
14.0
23.5
25.9
29.1
19.0
17.
20.6
25.

28.7
30.4
30.3
29.2
27.8

14.1
15.4
22.9
18.6
29.0
20.1
17.1
17.3
19.3
22.0
19.7
20.1
22.8
14.0

15.5
23.

10.0
12.9
12.3
14.4
21.3
14.6
20.0
19.3
10.7
15.3
18.0
23.2
24.1
24.5
20.8
20.0
23.4
27.4
28.4
29.0
21.7
17.7
23.1
26.8
22.0
24.7
31.6
19.4
21.8
24.6
24.5
26.4
29.4
32.3
22.6
21.8
20.4
25.8
28.9
30.8

U‘1GWOOGGONQHNGUNOCONHHONNOUONHUHUINGMQNONGHHOUU‘WGGQH

TMIN RAIN

O 0 0 0 0 0 O 0 0 0 0 O O 0 0 O 0 0 O 0 0 O O O O O O 0 0 0 O 0 0

use

p

p
GHOHUGGHh‘DfifiUGOO—‘QdbUUGOOUQGhU‘UUOGHO‘DONUhhMUUOOUloml-lm

p
N
O

13.4

170

p
O H
O 0

MN

OOOO‘OOOOOOOOOOOOOO‘DGOOOOOOOOOQOOOOHOOHOOOOOHOOOOOOOUOOM

O

OOOOGwOOOOOOOOOOHOObHOOOOOOOOOOOOOONOOMOOUOONOOOOOOOUOO

APPENDIX C

ID
CVCP

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

CVCP

207

SOL

28.9
28.7
29.2
29.1
8.6
30.
30.1
29.6
27.7
30.4
23.5
31.2
30.8
29.4
30.0
29.7
28.5
24.1
27.1
23.1
30.3
26.0
22.3
27.7
16.8
31.1
24.7
27.7
24.7
30.6
12.2
21.0
27.4
28.5
28.0
30.9
30.4
28.7
26.8
24.7
27.0
25.0
8.1
24.2
25.5
23.4
28.1
22.7
20.1
24.0
4.5
22.7
8.0
22.9
21.7
25.1
27.6
15.7

32.5
34.3
34.5
34.4
20.5
26.4
26.4
30.5
33.5
33.8
27.5
19.9
24.1
29.4
33.0
34.7
35.4
34.4
29.7
31.0
34.4
33.8
32.9
37.4
35.4
28.9
31.7
39.4
25.6
29.1
27.8
26.8
25.3
27.7
30.8
33.2
36.6
38.7
39.5
39.8
40.0
37.1
28.0
31.4
31.8
31.9
34.4
37.1
36.2
32.8
25.7
32.7
23.4
30.5
28.1
31.4
31.0
30.0

TMIN

171

z

GQOOOUIOUOIIHhOOUwOOOOOOOOOOOGOOOOOOOU‘OOOOGOOOOOOUOOOOODOOOO

H

H

p
OOOOOOOUOMUMOOOOOOOOOOOOOOOHOOOOOOO‘OOOOOOOOOOOOOOOOOOMOOOO

3
’3

Eééii

§§§§§§§§§§§§§§§§EE§§§§§§§§§§§Eaééééiééiéééaéaéiéfiiiii

APPENDIX C

ID

CVCP
CVCP
CVCP

CVCP
CVCP

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

208

SOL

24.6
27.1
26.9
26.0
16.8
26.4
24.6
24.9
21.6
21.7
13.2
27.2
27.0
23.8
14.0
14.2
20.7
23.8
19.6
20.4
21.0
23.1
24.4
11.
30‘
23.8
23.3
18.4
14.2
23.6
23.2
20.6

18.7
19.8
17.7
22.4
21.6
16.1

29.5
31.8
35.6
34.0
32.3
32.3
38.7
37.0
38.1
36.5
34.0
31.5
34.5
34.6
31.1
33.3
33.1
36.2
34.5
34.3
31.8
34.7
35.6
28.6
19.

27.9
26.8
26.7
26.2
26.8
25.7
26.0
21.2
24.5
23.7
25.0
27.7
27.8
29.5
23.1
22.8
16.4
19.3
22.8
24.8
26.7
28.3
28.0
27.3
21.6
25.6
22.9
25.6
29.2
29.3
21.5
14.3
17.7

TMIN

15.5
11.5
18.8
18.1
22.2
15.9
20.6
22.2
21.7

16.9

HsaH H
HuummmmqmmmmmmnuqemumHmemmo

j...
GONOGQOhOOGMOQQOHGOGHGHGUJHH

HHHHH

172

E
z

p

w

p

H

OHO‘DOOOOONOOOOOOUQIIUIOOOOU‘OOOOGOOOUIwOONOOOQGGOOUGOOOOOUOOOO

p

but H
OO‘OWOOOHOQOOOOOOOOUIGOOOOONOOOHOOOOHOOGOOOO‘NOOONOOOOOMOOOO

3

SEP

SEP _

SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP
SEP

APPENDIX C

ID

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

H Hwah- ta
0 O 0 0 O 0 0 0 O 0

HI IHF‘H
O

0 0 O 0 0 O 0 0 0 0 O O 0 0 0 O

thh‘

H
dONQIIIJIOUOUNéMOI-‘IIUINNNOQ\INMNHOHGOONMH‘MGOOOGNNhGGQ§

hNOGOhNUUGUGHONUNbHUIJI-‘UIOGUINOMGMOOOUM’QGMHNGHMGODU‘I

23.0

Q
s

H
nmmoomemqumeammeonHwHooomme

s4 IHF‘H

Hldh‘

NdGQhOGhHOGGMI—‘hfiddWGQOU‘UUQO

H

TMIN RAIN

11.9

F‘H Hsa

I
G401IDIOU'DUUGI-‘Ul\DQQDOQGUUQUNNHOHNUIHQOIDH

NHOOHHfiI—‘HNOOHHNMMGHIGGNI-‘UfidGNOOIDUIUIAO

IJF‘H
O

I
HHOOOUOHMOHOON
MUWUHDGOQQHGQQ

173

15.7

H
OOHOOOMUNQUIHHOGOOOOOOUIOOOUIGUIGUQOHOOOOOOGGOOOOQOQ’MNOOOOOOO

(”H
0

H
O

ION
OOGOOHIIOUIfiHfibNbOOOOOOOOOOIJHOhONHOIIOOOOI-‘OOOOOOGOOIJGOOOOOON

O O O 0 O O O O O O O 0 O 0 O O O 0

APPENDIX C

ID

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

:6;
E

MHOGOGQU‘NHGGMN‘IdQHHI—‘HNMOU‘NUQUU‘HUIUUQUUIDQUQGQOQQHUINHHOIDQNMOO

OMfiGbQGQQQQO‘GOGHHHHQU‘GMGQQUNHIONIDUIQIDU'IUIDOwNGWUI-‘OUIQNNNONIOHID‘D

HHHH

H

IIII
UOHNUOUIfi‘HHUQNMW‘OfiU‘DOQNI—‘G‘NHGOU‘bmfipmmHHNhua‘uwmm
ccccccccececcccccccccceecccecceecceeeeeeeceecee

HWGMHOUUUQOO‘GMUIG‘OHHbNGI-‘QUHOQG‘DQIDOOHQNNQNNUQGUIUI

H

U
h

1.2

174

TMIN RAIN DATE
3.8 9.3 19 NOV
0.2 17.0 20 NOV
1.3 0.0 21 NOV
-1.2 0.0 22 NOV
-O.9 0.0 24 NOV
2.8 0.0 25 NOV
10.5 5.1 26 NOV
0.5 3.9 27 NOV
-3.0 0.8 28 NOV
'3.9 0.0 29 NOV
-1.3 0.0 30 NOV
-7.2 1.0 1 DEC
-6.9 0.3 2 DEC
-1.3 0.0 3 DEC
-6.7 0.0 4 DEC
'2.1 0.0 5 DEC
-O.2 0.0 6 DEC
-1.0 0.0 7 DEC
'6.7 0.0 8 DEC
-7.9 0.0 9 DEC
-12.6 0.0 10 DEC
-15.0 0.0 11 DEC
-10.6 0.0 12 DEC
'6.3 0.5 13 DEC
-0.5 0.0 14 DEC
-10.2 0.0 15 DEC
-13.6 0.0 16 DEC
-IO.6 0.0 17 DEC
-10.4 0.0 18 DEC
-8.0 0.0 19 DEC
1.1 58.1 20 DEC
-3.5 0.0 21 DEC
”3.9 17.5 22 DEC
1.8 1.8 23 DEC
-O.4 0.0 24 DEC
-O.6 0.0 25 DEC
'7.1 0.0 26 DEC
-1.7 20.8 27 DEC
-l.7 0.0 28 DEC
-1.7 0.0 29 DEC
'11.7 0.3 30 DEC
-7.6 0.5 31 DEC
-7.1 0.0 01 JAN
'10.2 0.3 02 JAN
-10.8 0.0 03 JAN
-16.8 0.0 04 JAN
'9.2 3.1 05 JAN
-O.9 3.6 06 JAN
-O.9 15.7 07 JAN
-lO.3 0.0 08 JAN
-12.6 0.0 09 JAN
'0.6 0.0 10 JAN
-6.0 0.8 11 JAN
'3.7 1.8 12 JAN
-6.1 0.0 13 JAN
-7.0 0.0 14 JAN
'2.6 2.6 15 JAN

IWHMHXC

ID

GNP
GNP
GNP
GNP
GNP
GNP
GNP

GNP
GNP
GNP
GNP
GNP

E’

O O O O O
N‘DfiObD-‘OHO‘D‘Dthd'DVDOhIUIHGUQwOOO‘QMhUIUI‘OUNI-‘OQNM‘D

HHH
O O O O O O

HH

HHH H

H H

HHH
mmONUMNNwNNHMOOHHNhN¢HHHHwHOOHHNOOOdOOQO

HHH
O

OOGQMQGQNIUIUUIQ
O O O O O O O O O O O O

HH HHHHH
UbOOMGdOdQfiNG

thGGUG‘DGOUOQGH

OGDHU§§OQU~JQOOVD

H

H
0!
O

h

I I I I IH
Q‘GUUO
O O O
hQGUOU

llll
Hwhb
O

I
#mHUmmfii-‘NOGOHHU‘NQOU‘OHHOHOHHUINU‘U
ooooooooooooooooooooooo

H
GMW‘GGOOQUN‘HMHU‘O‘D‘DG‘DQOGb‘DUMQhOHO-fihm

H

HhGOUO‘ONOHD—‘ON

I
OmeONOh’NOOO

U5

2
2

N

UlOOOOOOOUICMOOOOOOOUOOOUHOOOOOOMOOOOOOOOOOOOOQHOOU‘IOOOOOOOOOO

H H
OOOOOOOOOOQOOOOOOHOOOOMOOOOHOGCOOOOOOOOOOOOOUIOOONOOOOOOOOO

SEEEEEEEEEEEEE

APP]

APPENDIX C

16.7

80L

"E

MHH
omononuoomo
O O O O O O O
oOHmhmonqwqau

21.

Hr»
NO
O O
dfi

10.3

Hra

O O O O O O O O
HHHOHHQQU‘OH‘DO

H H
oounoquqooqw

15.2
16.
21.1
24.1
24.4
24.5
12.4
16.9
17.6
17.4
9.7
17.
19.6
19.8
8.2
11.2
19.1
16.5
18.6
17.5

UHQOQOMMOHUUQGOHOOOMQOHO‘AOON
O
NO‘N‘O‘D‘OQwaMde‘DbOO-‘UIUUH‘DUO

Ht‘h‘
O O

176

0.0

F‘H

OOOOOOOOOQOOOHOOQOOOOOOOUOOO‘OOOOOOOOOUIUINOOOO'OOOOOOO
OOMUOOUU‘IOQOOOQOONOOOUOOONMQQOOOOOOOUOQQUOOOOHOOOOOO

O O O O O O O

3
a

EEEEEEEEEEEEEEEEE

APPENDIX C

132

80L

11.8

6.6
13.3
17.8
25.6
2703
16.5

4.3
13.9
28.6
21.7
29.5
17.4
17.0
29.7
29.0
27.5
11.6
13.1
12.2

7.7
30.3

28.0
18.8
29.3
27.0
21.6
21.3
23.5
29.2

4.0
18.2
16.7
15.6

8.4
26.9
18.3
13.6
19.9
27.9
18.4
24.6
28.4
28.5
18.7

29.9
31.4
31.3
29.8
23.9
18.0
19.9
27.4
26.1
27.7
23.9

17.8
13.6
17.7
19.0
23.6
27.0
26.1
20.1
23.6
26.3
26.9
26.7
26.1
25.2
24.6
18.7
22.2
21.0
29.3
28.7
22.3
26.2
21.9
23.6
24.2
28.9
30.6
29.1
24.8
21.1
25.6
17.6
24.8
24.9
20.9
15.9
25.4
27.0
26.9
25.9
29.3
31.0
33.2
33.1
33.9
34.9
25.9
27.5
24.8
28.7
31.9
32.6
28.4
31.8
32.5
34.7
33.2
32.9

Hra nu
qmmnmmenmm

F‘H

177

E

H

H

NU
O

H

OOOOHOOOOOOUIOOOOOOOOOOUIONHNOOOOOOOOO‘OUONOOOOQOOOON‘IOOOUOUQ
O
OOOOUIOOOOOUQUOOOOOOOOO#OUUIUIOOOOOOUOuOOUI-‘OOOOQOOOOHQOOOHOUN

§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

APPEI

2

__A-nnnnn00009999

APPENDIX C

ID
CVCP

CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP
CVCP

CVCP

190
191
192
193
194

196
197

199
200
201
202

204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247

22.6
26.8
18.0

15.9
28.7
23.4
28.9
26.2
20.6
6.8
4.6
16.3
22.9
24.7
21.3
21.2
17.6
18.3
22.0
22.9
4.5
24.7
26.1
25.5
22.3
17.6
23.8
19.4
25.2
24.0
22.2
19.4
22.4
17.8
23.8
19.4
15.0
18.1
25.2
24.3
22.0
13.4
24.6
11.4
11.7
20.6
23.9
21.9
17.5
6.1
7.1
23.
16.6
14.0
22.5
18.7
19.9

TMAX

31.6
33.4
32.9
20.6
25.1
27.7
27.2
30.0
31.1
31.2
22.3
19.0
27.6
30.1
31.1
31.7
32.0
32.0
30.6
25.9
27.0
19.6
28.4
29.6
29.5
31.8
30.2
30.5
25.7
20.3
25.4
26.7
28.8
29.4
29.8
28.4
28.8
27.4
24.6
27.8
26.9
29.2
27.5
29.5
26.8
25.8
26.0
27.3
28.2
31.1
26.2
26.9
26.7
28.6
26.3
24.0
25.6
25.6

TMIN RAIN

18.4
23.2
19.6
15.3
14.8

9.3
11.3
11.4
12.2
12.1
17.1
17.2
16.5
16.7
16.0
18.5
20.2

178

50.6

Hra

H

F‘H

N
OOOOOOHWOOOOONOHOOOOHOOHOOOOOONWOOOO‘DOO‘OHOOOU‘IQUOOOOOOOOO

O O O O O O O O O O O O O O O
OOOUOOQOOOOOOOOQOOOUOOOMOOOOOO‘DO‘OOOOMOOHUIUOOOHOOUOOOOOOOO

0‘

U

§§§§§§§§§§§§§§§§§§§§§§§ E

APP!

CV!

APPENDIX C

298
299
300
301
302
303
304
305

18.1

HHHMH N
qmqooomHNH
O O O O O O O O O

H
0|
Q‘DQNOUNNUOOG

H
0

19.1
19.4
19.1
18.9
17.9
18.1

H Fl
unown
O O O
how

17.6

HHHHH HH

H

H
NNNHOWNIUNOHQMUWGMMQOO‘OQ

OGOQ‘GONOQQOMQMMQQObUUG

H Hid

28.5
25.7
29.3
30.
27.9
23.4
25.2
23.8
11.4
12.2
21.2
16.1
23.8
24.7
25.1
25.2
26.1
23.2
13.8
16.8
20.4
16.9
18.2
22.7
25.1
24.2
25.9
20.3
13.5
16.4
18.8
17.2
11.8
12.4
12.9
14.4
21.7
23.5
26.5
29.6
27.1
25.2
8.2
7.4
1.1
3.3
9.9
17.0
21.2
24.6
24.8
3.0
24.0
24.1
22.9
23.0
14.0
13.9

TMIN RAIN

12.7
19.3
18.1
19.5

UGbQO—‘QU‘IQGGUU’UGHONQwaNI-‘fiQO‘OO‘DO-‘NIQUGfiQHhQGOUGWUbUO

I
IUGUU‘OQONMOOONUG‘DQONOQQUHU‘HNMO‘QhQOUUI—‘NOOOQQU‘IMG‘

H

179

OOOOOOWO‘DOHUOOOOOHOUOOHUOUOOOOOOOOOUJUOOOO‘OOGbOOOO-‘D‘DOQ

OOOOONQU‘0&0OOOOOUIOOOONOOOOOOOOOOOOOOOOOO‘ONOOOOU‘IHfiOO

000

H
OUI

DATE

SEP

§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

180

APPENDIX C

ID YR DOY SOL TMAX THIN RAIN DATE
CVCP 89 306 3.4 7.9 -1.3 .3 02 NOV
CVCP 89 307 6.8 5.3 -4.9 .3 03 NOV
CVCP 89 308 3.6 7.6 '1.2 .0 04 NOV
CVCP 89 309 4.3 14.4 3.9 2.1 05 NOV
CVCP 89 310 3.9 11.9 -1.8 .0 06 NOV
CVCP 89 312 2.4 8.8 3.3 .5 08 NOV
CVCP 89 313 7.2 8.2 1.6 1.0 09 NOV
CVCP 89 314 4.2 7.7 -1.0 2.1 10 NOV
CVCP 89 315 3.6 10.2 .5 .0 11 NOV
CVCP 89 316 9.7 9.7 -5.6 .0 12 NOV
CVCP 89 317 9.6 21.8 2.3 .0 13 NOV
CVCP 89 318 1.5 15.6 10.8 13.1 14 NOV
CVCP 89 319 .7 14.1 -.1 6.7 15 NOV
CVCP 89 320 4.7 .3 -4.0 2.5 16 NOV
CVCP 89 321 6.1 -1.9 -11.1 4.5 17 NOV
CVCP 89 322 8.1 -.3 -9.5 1.3 18 NOV
CVCP 89 323 8.3 6.5 -7.3 6.0 19 NOV
CVCP 89 324 9.3 11.7 .9 .0 20 NOV
CVCP 89 325 7.5 1.0 -4.1 .0 21 NOV
CVCP 89 326 3.7 -1.0 -7.9 .0 22 NOV
CVCP 89 327 5.3 -.8 -11.2 1.3 23 NOV
CVCP 89 328 9.2 4.8 -9.3 1.3 24 NOV
CVCP 89 329 5.6 9.8 1.6 .0 25 NOV
CVCP 89 330 9.1 11.4 -1.3 .0 26 NOV
CVCP 89 331 .8 15.2 1.8 0.5 27 NOV
CVCP 89 332 2.1 10.6 -5.7 1.3 28 NOV
CVCP 89 333 6.2 -1.8 -8.1 1.3 29 NOV
CVCP 89 334 8.9 5.4 -5.1 .0 30 NOV
CVCP 89 335 8.6 7.8 -4.2 .0 01 DEC
CVCP 89 336 3.4 3.1 -4.9 2.5 02 DEC
CVCP 89 337 6.2 -2.4 -9.2 1.3 03 DEC
CVCP 89 338 4.9 1.1 -7.8 .0 04 DEC
CVCP 89 339 2.0 1.4 -6.2 .0 05 DEC
CVCP 89 340 2.3 3.3 -6.6 .0 06 DEC
CVCP 89 341 4.7 -3.1 -8.5 .0 07 DEC
CVCP 89 342 8.4 -2.2 -10.9 .0 08 DEC
CVCP 89 343 4.7 -.4 -12.3 .0 09 DEC
CVCP 89 344 1.7 1.1 -2.7 3.8 10 DEC
CVCP 89 345 8.1 -.8 -7.0 .0 11 DEC
CVCP 89 346 6.6 -3.4 -15.1 .0 12 DEC
CVCP 89 347 4.9 -4.7 -18.4 2.5 13 DEC
CVCP 89 348 6.7 -7.4 -20.7 1.3 14 DEC7
CVCP 89 349 4.8 -8.6 -22.3 1.3 15 DEC
CVCP 89 350 7.2 -6.9 '24.8 .0 16 DEC
CVCP 89 351 6.6 -9.5 -24.7 .0 17 DEC
CVCP 89 352 7.9 -6.1 -18.2 3.2 18 DEC
CVCP 89 353 1.9 -3.7 -12.9 1.3 19 DEC
CVCP 89 354 2.6 -8.8 -16.3 2.5 20 DEC
CVCP 89 355 6.3 -11.3 -29.0 .0 21 DEC
CVCP 89 356 7.1 -8.3 -33.8 .0 22 DEC
CVCP 89 357 7.8 -6.4 -25.6 .0 23 DEC
CVCP 89 358 5.6 -10.7 -26.6 .0 24 DEC
CVCP 89 359 2.2 -1.0 -12.2 5.0 25 DEC
CVCP 89 361 2.8 -4.2 -14.8 1.3 27 DEC
CVCP 89 362 5.0 .0 -6.3 .0 28 DEC
CVCP 89 363 2.6 1.7 -2.2 0.5 29 DEC

APP]

_ _--nnnnnnn00009999922

181

APPENDIX C

ID YR DOY SOL TMAX THIN RAIN DATE
CVCP 89 364 1.6 -.1 -3.7 16.0 30 DEC
CVCP 89 365 1.1 .5 -2.3 .0 31 DEC
CVCP 90 1 2.25 -.3 '3.9 .0 01 JAN
CVCP 90 2 8.26 3.6 -5.5 .3 02 JAN
CVCP 90 3 7.13 5.6 ’2.2 .0 03 JAN
CVCP 90 4 1.52 5.3 '3.4 23.6 04 JAN
CVCP 90 6 4.69 2.5 '4.6 .0 06 JAN
CVCP 90 7 8.06 6.3 -3.8 .0 07 JAN
CVCP 90 8 7.68 7.7 '3.1 .0 08 JAN
CVCP 90 9 1.31 3.7 1.3 15.1 09 JAN
CVCP 90 10 1.44 1.4 '1.8 .0 10 JAN
CVCP 90 11 3.31 3.4 -2.1 .0 11 JAN
CVCP 90 12 6.04 -2.1 -5.7 .0 12 JAN
CVCP 90 13 6.88 .3 '10.9 .0 13 JAN
CVCP 90 14 5.13 3.9 -6.2 .0 14 JAN
CVCP 90 15 7.19 7.7 -.2 .0 15 JAN
CVCP 90 16 1.38 10.3 1.6 .8 16 JAN
CVCP 90 17 .94 11.3 3.4 18.2 17 JAN
CVCP 90 18 1.35 3.3 -4.7 .0 18 JAN
CVCP 90 19 6.32 1.9 -6.6 .0 19 JAN
CVCP 90 20 1.39 .1 '2.0 .3 20 JAN
CVCP 90 21 2.17 .5 '1.3 .0 21 JAN
CVCP 90 22 2.30 1.6 -2.1 .0 22 JAN
CVCP 90 23 2.12 7.0 '3.4 .3 23 JAN
CVCP 90 24 6.07 7.5 1.9 .0 24 JAN
CVCP 90 25 1.70 9.1 -3.0 25.9 25 JAN
CVCP 90 26 7.37 1.1 '4.6 .0 26 JAN
CVCP 90 27 6.87 9.4 -1.3 .0 27 JAN
CVCP 90 28 9.36 4.2 '6.7 .0 28 JAN
CVCP 90 29 7.84 4.3 '5.7 .0 29 JAN
CVCP 90 30 7.99 6.5 '4.3 .0 30 JAN
CVCP 90 31 10.95 6.5 '6.7 .0 31 JAN
CVCP 90 32 1.13 6.8 .6 18.7 01 FEB
CVCP 90 33 2.20 3.1 '4.5 14.9 02 FEB
CVCP 90 34 1.14 -.4 '4.0 .8 03 FEB
CVCP 90 35 2.24 -1.0 -4.9 .3 04 FEB
CVCP 90 36 8.75 6.6 -4.6 .3 05 FEB
CVCP 90 37 6.07 6.3 '1.1 .0 06 FEB
CVCP 90 38 2.82 5.3 .3 .0 07 FEB
CVCP 90 39 11.19 14.4 -.2 .0 08 FEB
CVCP 90 40 6.63 12.7 .2 1.3 09 FEB
CVCP 90 41 12.03 5.2 '5.8 .0 10 FEB
CVCP 90 42 9.32 3.7 -5.7 .0 11 FEB
CVCP 90 43 9.31 7.8 '9.0 .5 12 FEB
CVCP 90 44 4.37 15.3 '1.5 .5 13 FEB
CVCP 90 45 2.41 -1.5 '5.7 .0 14 FEB
CVCP 90 46 1.96 -1.2 -5.6 .0 15 FEB
CVCP 90 47 3.89 2.6 -4.9 5.9 16 FEB
CVCP 90 48 13.54 4.8 -8.9 8.2 17 FEB
CVCP 90 49 14.53 4.5 -7.8 1.5 18 FEB
CVCP 90 50 15.03 1.2 '9.7 .0 19 FEB
CVCP 90 51 15.30 3.8 '14.1 .0 20 FEB
CVCP 90 52 13.66 8.8 “5.7 .0 21 FEB
CVCP 90 53 1.52 5.3 -.0 43.2 22 FEB
CVCP 90 54 13.56 4.1 '6.3 2.1 23 FEB
CVCP 90 55 10.75 -1.0 -11.5 .0 24 FEB
CVCP 90 56 15.77 1.6 '19.9 .0 25 FEB

182

APPENDIX C

10 YR 002 SOL TMAX THIN RAIN DATE
ever 90 57 10.78 -.2 —16.0 .0 26 FEB
ever 90 58 11.77 7.3 —4.7 .0 27 FEB
ever 90 59 16.93 2.7 -10.9 .0 28 FEB
ever 90 60 16.80 5.8 -9.3 .0 29 FEB
cvcp 90 61 16.82 8.7 -2.8 .0 01 MAR
ever 90 62 16.42 2.4 -9.2 .0 02 MAR
cvcp 90 63 11.44 2.9 -11.5 .0 03 MAR
over 90 64 3.09 -1.1 -4.7 .0 04 MAR
ever 90 65 18.09 .6 -11.6 1.0 05 MAR
ever 90 66 17.28 4.4 -15.3 .3 06 MAR
cvcp 90 67 5.34 6.0 -7.9 12.6 07 MAR
cvcp 90 68 4.61 6.6 3.7 1.3 08 MAR
ever 90 69 5.78 14.2 1.1 26.0 09 MAR
ever 90 70 7.86 23.1 5.5 4.6 10 MAR
ever 90 71 14.53 25.5 15.8 .3 11 MAR
ever 90 72 5.75 23.1 14.7 .0 12 MAR
cvcp 90 73 13.57 25.6 14.9 .0 13 MAR
ever 90 74 6.05 22.1 8.8 9.3 14 MAR
ever 90 75 13.93 16.0 6.8 .0 15 MAR
ever 90 76 11.00 11.3 1.4 .3 16 MAR
ever 90 77 10.75 6.0 -.8 2.1 17 MAR
ever 90 78 12.06 .6 -4.8 .0 18 MAR
over 90 79 19.40 7.6 -8.6 .0 19 MAR
ever 90 80 1.79 7.9 3.1 11.7 20 MAR
ever 90 81 1.79 7.9 3.1 11.7 21 MAR
ever 90 82 1.79 7.9 3.1 11.7 22 MAR
ever 90 83 16.28 3.2 -5.7 .0 23 MAR
ever 90 84 18.96 5.3 -6.6 .0 24 MAR
cvcp 90 85 22.25 8.0 -6.4 .0 25 MAR
ever 90 86 20.28 6.7 -6.2 .0 26 MAR
cvcp 90 87 22.75 9.7 -9.5 .0 27 MAR
ever 90 88 18.45 13.1 -6.6 .0 28 MAR
cvcp 90 89 4.16 6.0 2.4 12.9 29 MAR
ever 90 90 6.86 8.6 3.8 .8 30 MAR
cvcp 90 91 5.30 9.6 4.2 4.3 31 MAR
ever 90 92 6.46 12.6 6.0 2.3 01 APR
ever 90 93 3.33 9.4 .8 4.8 02 APR
ever 90 94 11.09 7.7 -1.0 .0 03 APR
ever 90 95 15.09 11.5 -3.6 .0 04 APR
cvcp 90 96 9.64 6.8 -2.6 .0 05 APR
cvcp 90 97 14.92 7.0 -6.6 .0 06 APR
ever 90 98 15.77 6.2 -6.8 .0 07 ApR_
ever 90 99 12.14 12.8 -3.7 .0 08 APR
ever 90 100 9.41 14.7 6.0 .0 09 APR

183

APPENDIX C
SOIL NITROGEN BALANCE PARAMETERS (1988/1989)
TR! - Treatment number
STRAW - Weight of organic residue of previous crop (kg ha”)
8DEP - Depth of incorporation of residue (cm)
SCN - C:N ratio of the residue of previous crop
ROOI - Root weight of previous crop
RCN - C:N ratio of roots of previous crop

CVCPBBOI - MSCN (ID) 89 (Year) 01 (01 . REP; 02 - CMP)

ID TRT STRAW SDEP SCN ROOT RCN
CVCP8801 1 9000. 5. 60. 2500. 40.
CVCP8802 1 9000. S. 60. 2500. 40.

CVCP8901 1 7220. 5. 60. 4000. 40.
CVCP8902 1 7800. S. 60. 4000. 40.

184

APPENDIX C

son. PROFILE morn-Runs (1988/1989)

NOTE

OO

These parameters are used for both the RMP and CMP
simulations for both 1989 and 1990.

Soil Name I Oshtemo Sandy Loam

Soil Albedo I 0.11

Upper limit of stage 1 soil evaporation I 7.00 mm

Soil water drainage constant, fraction drained per day I 0.67
SCS runoff curve number I 62.00

Annual average ambient temperature I 9.4 (°C)

Annual amplitude in mean monthly temperature I 15.5 (°C)
Mineralization factor (DMOD) I 1.0

DLAYR I Thickness of soil layer (cm)
LL I Lower Limit of plant extractable soil water (cm? cmf3)
DUL I Drained Upper Limit of plan; exSractable soil water (cm3 cm'3)
SAT I Saturated water content (cm ')
SW I Default soil water content ( cm'3)
NR I Root weighing factdf to determine new root growth distribution
80 I Bulk Density (g cm')
00 I Organic carbon (4)
DLAYR LL DUL SAT SW NR 80 0C
15. 0.080 0.175 0.322 0.175 1.000 1.70 0.78
15. 0.085 0.155 0.315 0.155 .640 1.64 0.73
15. 0.083 0.139 0.311 0.139 .470 1.77 0.37
15. 0.085 0.133 0.302 0.133 .350 1.73 0.37
15. 0.071 0.128 0.299 0.128 .260 1.61 0.38
15. 0.073 0.126 0.297 0.126 .190 1.53 0.38
15. 0.073 0.133 0.296 0.133 .140 1.60 0.38
15. 0.072 0.133 0.296 0.133 .090 1.55 0.35
15. 0.069 0.135 0.320 0.135 .050 1.62 0.32
30. 0.069 0.135 0.320 0.135 .030 1.61 0.22
30 0.072 0.135 0.320 0.135 .010 1.61 0.12

APPENDIX C

TRT I
DLAYR I
S" 3

N03 -
P8 I

cvcpaao1 - Mscu (10)

01 CVCP
DLAYR

15.
15.
15.
15.
15.
15.
15.
15.
15.
30.
30.

01 CVCP
15.
15.
ISO
15.
15.
15.
15.
15.
15.
30.
30.

Treatment number
Thickness of soil laye

185

SOIL PROFILE INITIAL CONDITIONS (1988)

Soil water content (
NH4 I Soil ammonium content (mg elemental N per kg of soil)
Soil ammonium content (mg elemental N per kg of soil)
p8 of soil in a 1:1 soilxwater slurry

8801
SW

00.175
00.155
00.139
00.133
00.128
00.126
00.133
00.133
00.135
00.119
00.128

8802

00.175
00.155
00.139
00.133
00.128
00.126
00.133
00.133
00.135
00.135
00.135

OOOOOOOHHHN §
0 s s s s s
HHUMWQQ‘D‘DHW §

”Rommmmmmmo

OOOOOOOOOOb

88 (Year)
N03 PH
0.9 5.8
0.6 5.6
0.5 5.7
0.5 5.7
0.3 6.1
0.3 6.1
0.1 6.1
0.1 6.1
0.1 6.1
0.1 6.1
0.1 6.1

OOOOOOOOOOO
HUMO‘GO‘GU‘U‘Q‘D

GOGGGGO‘U‘U‘GQ
O
GGMMUIUIUIQIJQUI

cmf3)

01 (01 - RMP: 02 - CMP)

186

APPENDIX C

SOIL PROFILE INITIAL CONDITIONS (1989)

TRT I Treatment number

DLAYR I Thickness of soilcigyer

SW I Soil water content ( cm")

NH4 I Soil ammonium content (mg elemental N per kg of soil)
N03 I Soil ammonium content (mg elemental N per kg of soil)
PH I p8 of soil in a 1:1 soilxwater slurry

cvcpa901 - MSCN (ID) 89 (Year) 01 (01 - RMP; 02 - CMP)

01 CVCP8901

DLAYR SW NH4 N03 PH
15. 00.175 2.0 0.4 5.8
15. 00.155 1.6 0.6 5.6
15. 00.139 1.5 0.6 5.7
15. 00.133 1.5 0.6 5.7
15. 00.128 0.9 0.5 6.1
15. 00.126 0.9 0.5 6.1
15. 00.133 0.9 0.5 6.1
15. 00.133 0.9 0.5 6.1
15. 00.135 0.4 0.4 7.3
30. 00.135 0.2 0.2 7.3
30. 00.135 0.1 0.1 7.3

01 CVCP8902
15. 00.175 2.3 1.2 7.5
15. 00.155 1.9 0.8 6.7
15. 00.139 2.6 1.1 5.7
15. 00.133 2.6 1.1 5.7
15. 00.128 1.4 1.1 6.5
15. 00.126 1.4 1.1 6.5
15. 00.133 1.4 1.1 6.5
15. 00.133 1.4 1.1 6.5
15. 00.135 1.2 0.9 5.5
30. 00.135 1.0 0.7 5.5
30. 00.135 0.8 0.5 4.5

187

APPENDIX C

IRRIGATION WATER APPLICATION DATA (1988)

TRT I Treatment
CVCP8801 I CVCP (ID) 88 (Year) 01 (01 I RMP; 02 I CMP)

DOY I Day Of the Year
ANIRR I Amount of irrigation water applied (mm)

CVCP8801 CVCP8802
DOY ANIRR DOY ANIRR
168 15. 165 35.
169 8. 167 16.
176 13. 171 23.
178 13. 174 16.
180 13. 175 33.
182 13. 180 37.
184 13. 183 18.
186 13. 191 12.
188 13. 194 19.
190 16. 196 28.
192 13. 200 6.
197 13. 202 18.
210 13. 203 7.
211 13. 210 6.
213 13. 214 15.
215 13. 215 9.
217 13. 216 13.
225 13. 217 10.
227 13. 221 10.
231 7. 222 16.

239 13.

188

APPENDIX C

IRRIGATION WATER APPLICATION DATA (1989)

TRT I Treatment
CVCP8801 I CVCP (ID) 88 (Year) 01 (01 I RMP: 02 I CMP)

DOY I Day Of the Year
ANIRR I Amount of irrigation water applied (mm)

CVCP8901 CVCP8902
DOY AHIRR DOY AHIRR
176 1. 184 13.
185 4. 185 13.
198 15. 187 15.
202 3. 188 15.
227 13. 189 13.
228 13. 198 15.

199 15.
200 13.
201 13.
226 15.

228 13.

189

APPENDIX C

NITROGEN FERTILIZER APPLICATION DATA (1988/1989)

CVCP8801 I CVCP (ID) 88 (Year) 01 (01 I RMP; 02 I CMP)
DOY I Day Of the Year
ANPERT I Amount of fertilizer nitrogen added (kg ha“)
DPERT I Depth of incorporation (cm)

IFTYPE I Type of fertilizer used (see Table 9)

CVCP8801

DOY AMFERT DFERT IFTYPE

126 10.0 5.0 6

169 40.0 3.0 1

190 80.0 3.0 1

CVCP8802
DOY AHFERT DEERT IFTYPE

111 200.0 25.0 4
126 10.0 5.0 6

CVCP8901
DOY AHFERT DPERT IPTYPE

140 10.0 5
176 30.0 3
185 35.6 3
202 47.5 3

HHHm

CVCP8902
DOY AMFERT DPERT IFTYPE

125 200.0 25.0 4
141 10.0 5.0 6

190

APPENDIX C

MANAGEMENT DATA (1988/1989)
Input management data for RMP and GNP simulations for 1988.

Day simulation began I 74

Day of sowing I 126

Plant population I 6.78 pl m'z
Row spacing I 0.773 m

Planting depth I 5.0 cm

Input management data for RMP and CMP simulations for 1989.

Day simulation began I 101

Day of sowing I 140

Plant population . 8.29 pl 8'2 (RMP) ; 7.29 p1 m“? (0MP)
Row spacing I 0.773 m

Planting depth I 5.0 cm

191

APPENDIX C

GENETIC DATA (1988/1989)

P1 I Growing degree days (base 8 °C) from seedling emergence to the end
of the juvenile stage (d °C).

P2 I Photoperiod sensitivity coefficient (1 hr")

PS I Growing degree days (base °C) from silking to physiological
maturity (d.°C).

P6 I New leaf area coefficient
62 I New grains per plant coefficient

G3 I Potential kernel growth rate (mg kernel'1 day").
This genetic variety was used for all treatments for both 1988 and 1989.

Variety ID I P103475

P1 I 220.00
P2 I 0.30
PS I 740.0
P6 I 1.0

G2 I 1.10
G3 I 8.90

APPENDIX D

.- ' .
f .u.‘ '

 

 

APPENDIX D

Inputs used in the inbred simulations

WEATHER DATA (1989/1990)

ID I MSCN

Latitude I 41.50

Longitude I
YR I Year

85.40

801. I Solar Radiation (Mj 0'2)

TMAX I Maximum Daily Temperature (°C)
TMIN I Minimum Daily Temperature (°C)

RAIN I Rainfall (mm)

ID YR DOY SOL TMAX TMIN RAIN DATE
MSCN 89 121 23.91 15.9 3.3 2.8 01 MAY 89
MSCN 89 122 6.71 9.3 2.9 3.8 02 MAY 89
MSCN 89 123 26.88 16.3 1.3 0.0 03 MAY 89
MSCN 89 124 21.34 19.0 6.0 1.8 04 MAY 89
MSCN 89 125 21.05 16.5 5.4 0.0 05 MAY 89
MSCN 89 126 12.87 6.8 -1.5 0.0 06 MAY 89
MSCN 89 127 23.31 8.6 -3.3 0.0 07 MAY 89
MSCN 89 128 16.99 16.3 -0.6 1.8 08 MAY 89
MSCN 89 129 11.49 16.8 7.5 1.3 09 MAY 89
MSCN 89 130 28.14 18.0 2.6 0.0 10 MAY 89
MSCN 89 131 29.11 16.9 3.4 0.0 11 MAY 89
MSCN 89 132 10.10 16.1 7.3 6.6 12 MAY 89
MSCN 89 133 7.22 13.9 7.7 0.8 13 MAY 89
MSCN 89 134 12.37 16.8 6.8 0.0 14 MAY 89
NSCN 89 135 13.35 17.8 8.7 0.8 15 MAY 89
MSCN 89 136 25.71 22.6 4.9 0.0 16 MAY 89
MSCN 89 137 28.14 25.8 6.7 0.0 17 MAY 89
MSCN 89 138 14.78 24.7 11.9 2.0 18 MAY 89
MSCN 89 139 5.15 21.1 16.1 14.0 19 MAY 89
MSCN 89 140 11.95 21.7 10.5 0.5 20 MAY 89
MSCN 89 141 28.53 24.0 7.9 0.0 21 MAY 89
MSCN 89 142 1.29 24.1 17.1 0.0 22 MAY 89
MSCN 89 143 29.34 25.9 11.5 0.0 23 MAY 89
MSCN 89 144 14.16 26.0 20.6 12.7 24 MAY 89
MSCN 89 145 7.36 24.4 18.8 3.3 25 MAY 89
MSCN 89 146 28.95 22.1 12.1 0.0 26 MAY 89
MSCN 89 147 27.67 16.8 7.9 0.0 27 MAY 89
MSCN 89 148 27.01 20.8 3.4 0.0 28 MAY 89
MSCN 89 149 12.42 21.3 9.8 0.0 29 MAY 89
MSCN 89 150 16.07 27.5 17.7 46.5 30 MAY 89
MSCN 89 151 9.16 28.3 18.1 99.6 31 MAY 89
MSCN 89 152 10.87 24. 18.0 9.7 01 JUN 89
MSCN 89 153 29.76 25.3 15.1 0.0 02 JUN 89
MSCN 89 154 8.41 21. 14.5 16.5 03 JUN 89

192

“PERM!

APPENDIX D

155

212

SOL

28.52
20.05
28.57
25.23
22.04
20. 41
21.85
29.51
4.82
11.98
14.11
13.98
7.07
26.22
23.39
8.16
19.18
24.32
15.89
21.05
26.27
28.41
19.95
8.69
30.35
30.85
30.26
28.06
22.67
24.54
24.11
26.31
24.47
27.12
24.46
22.63
26.02
16.91
8.70
14.76
26.19
23.06
24.81
28.04
19. 18

5.20
4.11
11.87
19.51
19.83
15.68
15.59
15.46
16.94
22.35
23.53
3.94
23.98

23.0
24.3
27.6
29. 2
28. 1
23. 8
19. 8
24. 5
17. 7
24.1
23.9
20.0
15.1
24.3
26.6
23.3
25.4
28.0

31.9

24.8
26.0
23.5
26.8
30.0
30.7
27.4
30.9
31.8
32.7
32.8
31.1
32.2
34.1
32.1
18. 4
24. 6
26.4
25.9
27.5
28.6
29.9
21.9
20.6
27.1
29.9
30.6
31.2
32.2
31.0
32.2
26.2
27.1
20. 7
27. 9

THIN

11.2
11.8
11.7
15.6
17.3
12.0

9.9

5.3
15.8
16.5
15.2
10.7
10.4
10.6
15.0
17.4
16.1
15.2
17.7
19.1
18.9
17.9
20.0
19.2
13.8

8.9

20.5
18.2
19.3
17.4
17.1
18.1
18.5
15.2
18.7
24.2
21. 2
16. 7
16. 3
12. 1
17.0
13.4
13.4
14.0
17.7
18.2
18.0
17.5
16.6
19. 4
21.4
21.1
21.7
15.3
10.4
15.7
15.1

193

E
2

OGOOONOOOOOOJIIOOOOOOOOO#00OOOOOOOOOOO‘OOOOOOOOOOO‘OOOOOOOOOO

0‘OOOU‘UUOOOQUGOOOOOOOOOOOOOOOOOOOOOO‘OOOOOOOOOOOOQOOOOOOOOOO

U

E§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§ E

APPEb

194

APPENDIX D

ID YR DOY SOL TMAX TMIN RAIN DATE
MSCN 89 213 25.06 28.8 14.5 0.0 01 AUG
MSCN 89 214 25.97 30.1 15.7 0.0 02 AUG
MSCN 89 215 22.96 32.8 18.2 0.0 03 AUG
MSCN 89 216 18.24 31.3 22.7 3.8 04 AUG
NSCN 89 217 23.76 30.8 19.4 24.6 05 AUG
MSCN 89 218 21.50 24.6 11.6 0.0 06 AUG
MSCN 89 219 25.24 19.4 8.1 0.0 07 AUG
MSCN 89 220 21.07 23.8 8.7 0.0 08 AUG
MSCN 89 221 20.24 26.0 11.9 0.0 09 AUG
MSCN 89 222 21.93 27.9 12.4 0.0 10 AUG
MSCN 89 223 21.87 28.3 13.4 2.3 11 AUG
MSCN 89 224 17.56 27.9 14. 0 4.3 12 AUG
MSCN 89 225 24.88 27.9 13. 3 0.3 13 AUG
MSCN 89 226 18.25 28.6 15. 5 0.0 14 AUG
MSCN 89 227 15.73 26.6 16.4 0.3 15 AUG
MSCN 89 228 19.16 23.5 12.5 0.0 16 AUG
MSCN 89 229 23.10 26.2 10.4 0.0 17 AUG
MSCN 89 230 23.86 25.7 11.6 0.0 18 AUG
MSCN 89 231 21.26 27.2 11.2 0.0 19 AUG
MSCN 89 232 12.50 27.9 18.1 19.8 20 AUG
MSCN 89 233 24.98 28.9 16.2 0.0 21 AUG
MSCN 89 234 11.80 27.1 19.2 2.0 22 AUG
MSCN 89 235 9. 88 26.1 19.4 0.0 23 AUG
MSCN 89 236 19.70 25.4 16.3 0.0 24 AUG
MSCN 89 237 24.25 26.1 11.6 0.0 25 AUG
MSCN 89 238 21.42 27.3 12.2 0.0 26 AUG
MSCN 89 239 18.40 30.4 18.4 0.0 27 AUG
MSCN 89 240 5.60 25.7 18.1 0.0 28 AUG
MSCN 89 241 7.16 26.4 20.6 0.0 29 AUG
MSCN 89 242 23.66 25.9 13.9 0.0 30 AUG
MSCN 89 243 16.43 27. 8 11.5 0.0 31 AUG
MSCN 89 244 12.66 25. 7 18. 0 0.0 01 SEP
MSCN 89 245 23.43 22. 9 12. 2 0.0 02 SEP
MSCN 89 246 19.86 24. 0 8. 9 0.0 03 SEP
MSCN 89 247 20.09 24.5 9.5 0.0 04 SEP
MSCN 89 248 17.29 26.9 14.0 0.0 05 SEP
MSCN 89 249 5.53 24.6 19.4 11.2 06 SEP
MSCN 89 250 14.03 29.2 19.6 8.6 07 SEP
MSCN 89 251 16.12 30.0 20.2 8.4 08 SEP
MSCN 89 252 11.14 27.7 19.6 6.9 09 SEP
MSCN 89 253 17.77 24.2 15.1 0.0 10 SEP
MSCN 89 254 11.64 24.0 14.5 0.0 11 SEP
MSCN 89 255 18. 98 22.5 12.8 10.2 12 SEP
MSCN 89 256 1.74 12.8 10.9 9.4 13 SEP
MSCN 89 257 1.51 13.1 11.1 9.4 14 SEP
MSCN 89 258 22.65 20.8 5.5 0.3 15 SEP
MSCN 89 259 7.43 18.4 10.3 0.3 16 SEP
MSCN 89 260 18.20 23.6 5.7 0.0 17 SEP
MSCN 89 261 19.89 24.3 7.0 0.0 18 SEP
MSCN 89 262 20. 98 25.6 7.8 0.0 19 SEP
MSCN 89 263 16.63 25.6 9.0 0.0 20 SEP
MSCN 89 264 14.67 26.1 10.2 0.0 21 SEP
MSCN 89 265 7. 90 24.3 10.6 0.0 22 SEP
MSCN 89 266 16.87 12.1 1.7 0.0 23 SEP
MSCN 89 267 19.78 16.4 -1.9 0.0 24 SEP
MSCN 89 268 18.94 19.3 0.4 0.0 25 SEP
MSCN 89 269 19.15 16.2 5.8 0.0 26 SEP
MSCN 89 270 18.98 17.5 -0.7 0.0 27 SEP

APPENDIX D

DOY

271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286

325
326
327
328

SOL

18.78
17.07
17.74
14.98
3.98
16.03
17.93
8.37
11.20
9.15
9.56
15.11
7.59
15.93
16.22
15.57
14.06
12.16
8.39
1.75
10.62
2.92
4.19
6.91
13.99
10.00
12.41
10.62
11.78
11.95
11.51
11.29
1.28
1.34
10.40
3.23
8.54
3.36
4.57
4.54
2.43
1.96
7.93
4.46
3.69
9.43
9.74
1.31
0.58
2.64
6.21
804s
8.58
9.66
5.94
3.51
14.17
14.17

22.6
25.4
23.9

12.8

F‘H
O O O O O O O O

HUJNJ 4‘

H

beNNQNHHMGHWOGQOG‘Ifivth

OOUOOthi-‘GH‘OUI‘OGLHUH‘OHQQ

TMIN

UHDH
s s s
‘0‘“

HH I

H

NMQMOOOHOI’OGGQHOMGGOU‘OUOOOdfifiomO‘OGGUU‘UOHGHQNQOOQhHO-‘GOOUI

H

I
'0‘
ssssssss

I
HH
ooqwnmmeouuNOOHbHOAOHoowamhmmqqMHOOOOhooonmomeHHOOHomq

195

E
2

H

OOOOOGOOUIUhOOOOOU00000OOUOOOOOOOOUGUUOHOOOOOOOOOOOOOOOOOOO

H

N
OOOOOUOOOHDOOOHPONONOOOONOOOOOOOOOfi‘DNO-fifi-‘OOOOONOOOOHOOOOOOO

D

5

SEP

SEP

§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

H

_555‘551‘n‘fifiafififififififififi

APPENDIX D

329

5.43
9.24
0O76
1.81
7.88
9.03
8.42
3.76
6.02
5.30
2.46
2.04
5.49
8.55
4.76
1.96
8.34
7.80
5.75
7.12
5.84
7.88
7.04
8.26
1.81
3.94
5.85
8.09
8.01
5.72
2.65
7.22
3.07
4.87
2.93
1.56
6.72
2.51
8.48
7.39
1.00
4.26
5.59
8.77
7.81
1.10
1.54
5.23
5.33
7.02
5.36
5.26
1.43
2.43
9.57
9.69
3.23
8.24

E

I HHH

QNOUOOMNHGNIO

NMOMHON‘OGGOG

I
N
0
.

-3e1

OGGHUGOQHHbNfiGMNOGU‘IflONOI—‘O

OHQGOQ‘QOGMGUHOUIGIFUIOUNGOJUI

THIN RAIN

106
-2.1
1.5
-4.9
-7.0
-4.3
-4.1
-5.1
-8.0
-7.4
-5.4
-6.2
-8.2
-10.5
-11.2
-3.3
-7.2
-15.4
-16.6
-19.3
-20.9
-20.7

-8.6
-3.3
-5.6
-1.0
-6.1
-7.9
-1.3
-9.5
-13.8

196

0.0
0.0
10.4

OOOOOIIGI)OOOOOOOOOOUOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
sssssssssssssssssssssssssssssssssssssssssssssssssssssss

OUOO’OHGWOOUOOwOOOmeOOOOOOOOOOOOOOOOOOOOOOOOOOOUOOIADOOOO

DATE

H
O‘DOQGM‘UNH

HHHHH
mouMH

NOV

APPENDIX D

10.35
6.55
8.85

10.81
9.70
6.07

11.79

11.79

12.00

12.00

11.15

11.65
8.51
4.16
2.64
5.87
0.25

11.77
7.24

12.20
8.23

10.21
4.69
2.13
2.35
4.50

13.65

14.89

15.53

15.61

13.76
1.47

13.31

10.40

16.56

10.36

14.14

16.94

17.16

17.05

15.47

13.57
3.41

18.12

17.47
4.78
4.54
5.37
8.46

13.53
5.32

12.43
4.43

13.98

11.85

10.82

11.47

19.09

E

I
#GslhHNONHQUHMNMOhGGI—IObHI-‘OOUI0U§bb§d0UhUWUhfiOHMGMOU

HIIhI
whommH«OhmOOOHbMHthmHOHommwmomumwabQObOHmmHmMGOH

h‘H

H

NH

25.
22.9

THIN

I
0

O O O O

outho

IIIII

HHHH H

mum. m
O

I
H
H
O
b

-11s4

II

IHH
ooam
Iowa

I
bbU‘OQUfiHHUHI—‘OH

NOOHGOGOQHHOMMU

Hrahl
UNHUHOOQHHUO

II
QGONUWGU‘QQHU
s s s

197

2

MN
OONOONOOOHIDONOOOOOOOOUOOONQQOOHHHOOOOOOOOOOOOOOOOOOOOOOOOO

oouuouooomomqoooooooooooomoooommooooooumoooowooomooooooooo

HHHN

AP!

APPENDIX D

SOL

14.48
1.76
16.61
17.05
17.62
20.38
22.10
17.39
2.32
5.13
4.48
6.03
2.64
13.31
12.75
9.13
14.66
17.27
11.69
8.68
2.21
13.81
18.67
16.81
4.82
12.44
17.17
22.69
26.33
14.33
2.04
26.06
26.42
24.83
21.72
23.91
22.71
23.23
14.10
21.31
12.38
24.60
25.94
8.48
1.47
25.87
26.20
27.52
23.61
20.67
9.58
29.78
2.98
22.87
7.06
6.45
8.15
15.59

.333

HH H HHH H

H
NGHG‘Q‘QUbGOQONOGGOOGGMb

F‘H
m

F‘H
Ink)&
O O

UNI-J
NOU‘I
O O

to
UI
UI‘DIIOO‘DNOHIINwfib-IMIOGOQOwaNbUIOIOfiUlI-‘OII‘DGOG

”IAIN
000

29.6
25.0
22.0
22.7
15.8
18.9
16.

17.1
17.2
24.5
26.4
26.1
18.4
15.2
10.7
19.1
17.6
20.3
23.0
18.8

THIN RAIN

'1.7

3.5
-4e8
-5e7

1e2
-4e9
-8.0
'6.0

h”
O O
III.

I
NMHNH‘NMGU‘GU‘HNOHGM

O
“GONGUQHOGQIOm‘dwUUIN

I
5

Hrs
O O O O O O O O O

HQOQOQNUO‘GMGGU‘ONG

O
HWD‘OONQHHONGGOGWG

F‘HFI

198

U

04NOHNOU‘OOOOOQHOOOOOOOOOOOOGOOOHOdHOI—IddOOOOOOhI-‘NONOOOOOOOO

H

H

H

OhuwuhomooooouooooomoooooowuoouuoamomooooowOOHmmmqoooooouo

H

UN”

O O O O O O O O O O O O O O O O O O O O O

APPENDIX D

28.59
10.49
20.39
10.98
27.27
20.15
17.16

6.07
10.46
25.21
22.59
29.24
30.87
31.04
21.22
12.24
19.55
26.19
13.68
18.14
25.85
15.56
27.50
26.08
21.59
26.14
28.72
19.50
30.35
25.79
20.32
29.74
25.25

9.43
28.14

8.18

9.29
29.33
29.56
13.39
23.65
14.44
15.60
24.62
28.88
30.31
28.57
28.39
20.06
21.34
21.76
25.79
24.12

8.86

4.51
18.93

9.94

7.94

19.3
21.7
19.8
15.7
20.1
20.4
22.1
16.9
21.4
25.0
24.4
20.9
21.7
24.0
28.8
26.6
25.0
17.9
17.2
25.

27.2
27.3
25.0
24.1
26.7
29.1
34.0
30.9
30.6
31.4
35.6
29.3
26.6
24.2
29.7
22.8
18.9
23.1
28.8
25.5
31.5
29.6
27.6
31.3
26.9
28.5
31.2
37.7
29.8
24.0
25.

35.5
32.9
26.

20.9
24.7
20.2
22.1

THIN

HHHH H

N
UUGOOOGUQHOUbOGQOOOO

F‘H
O
NdOdOGOGUhMQU‘I—‘QOOHNGUN‘O

HH
GI]
O

14O
11.7
18.0
19.3
20.5
17.9
15.6
23.3
16.1
10.6
17.0
16.5
14.7
1302
11.1

7.4
14.8
14.4
19.1
21.4
20.6
17.0
14.4
14.1
24.1
18.9
1‘O2
11.4
17.0

N
N
O

O

20.9
16.1

HII
hue
O O
Nab

15.1

199

E

H

H

HOOOOOOOOOOOOOOQMOOOONUIOIIOOOOOU0OOOOOOOOOHNOOOOOOOHOOOOOOO
O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
GUI00)U0OOOOONOOOOI—IUGOOU'lwowOOOOOwOOOOOOOOOOWGOOOOOOObOOOwOOO

H

l0

§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

APPENDIX D

253

SOL

TMAX

23.7
27.4
29.7
31.9
31.6
26.0
29.8
21.3
25.5
27.8
29.8
28.7
28.9
31.1
30.7
26.4
23.8
26.8
27.8
29.4
24.9
25.8
20.7
23.0
28.7
28.6
28.7
27.4
25.5
23.1
25.8
28.3
29.4
27.0
31.6
29.1
23.6
21.5
22.9
24.6
28.4
29.5
32.3
33.6
34.7
30.3
29.2
30.1
31.3
31.2
27.2
31.8
34.3
36.5
280s
26.3
28.3
29.3

TMIN RAIN

16.3
1407
18.8
18.3
19.0
19.9
19.1
1502
12.1
12.4
12.9
15.8
16.4
18.3
18.3
18.
13.0
9.1
11.2
12.7
19.7
15.9
11.4
10.9
7.2
12.
12.3
14.9
16.4
15.3
11.7
14.0
17.1
17.1
21.2
20.1
17.3
19.9
19.1
18.3
15.8
16.2
17.7
20.1
22.5
18.8
14.3
14.4
15.5
19.3
14.4
16.9
22.0
21.0
17.9
14.2
16.8
17.5

200

N

UM

H

O O O O O O O O O O O O O O O O

N
OOODUOOOOOOOOOOOOOOOW\DONOOOOIDHIIOOOOOOIOOOOOOOOOOOOOUIOUIUIUOOO
OOOHOOUIOOOOOOUIWOOOOObeUOOOOHQOOOOOOOGOOOOOOOOOOOOOMO’HUOQU

IO

W

O O O O O O O O O O O O O O

D

E

§§§§§§§§§§§§§§§§§

SEP

APPENDIX D

254

25.6
29.9
29.1
25.6
20.8
1802
18.1
17.7
17.8
22.3
18.8
1803
12.9
18.9
24.0
25.6
27.9
23.5
19.8
17.0
19.8
21.7
27.4
14.9
28.8
29.0
19.7
11.5

8.1
11.6
15.8
17.8
18.8
21.9

26.2

5.4

TMIN

HHHH

HII HII

OUIGOOIIOUOHOUINIIéGHGGI—‘UHGQNOOGMU‘GUIOGUGNhOQGNNmGI-‘Q

H

ONNwaIIUIUI0|HHDO‘NHQMhthMOOOHUIGIIHOIOONUINMIOOIOOIDOGO

HI»

201

2

PM

N

U

0)

IbN

OOUIQOOOOOOOOOOOOCOCO-“#0001000OHGOODOOOONOHOOOOHHUO‘HOhONOOO
OON‘DOOOOOOOOOOOOOOOUO‘O‘OO’GOOOUNHU‘OOGUHHOOOIBOOOOOUIQOHIfluNOOOOO

§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§

APPENDIX D

365

9.65
2.25
8.74
3.36
10.72
9.48
10.11
9095
2.13
9.28
5.08
5.03
8.83
1.72
4.14
6.31
3.54
6.47
2.12
0.79
0.88
5.72
8.66
4.20
5.75
1.09
5.52
7.17
2.66
8.30
6.51
8.42
7.73
7.85
4.35
1.76
7.91
1.90
0.93
1.48
0.75
4.98
5.63
1.49
7.06
7.66
4.08
3.62
3.62
3.62
3.62
0.90
1.23
8.01

E

H

IHNJH
momHmmqmomm

O
UNOOHQGMUQOU‘Q

H

FIH

H

OUOfibbfiObMGGMUGhII-‘UmmNWMQéhMOGbOQN

H
oqommmauoouuoqoboqwomoumuommemAJ-qm

THIN RAIN

-4e7

MIDNNHUGGNHNN

H

I-‘NNUIOOUIOONHH

-2.9

202

O

HON
OOGOOOOOOOMOOONOO‘OOOOOOOOOOOO’IOOOOI}GOOOHMOOOOOOOOOOOOOO

H
O

OUNOOOOOOOQOOQQOO’OOU‘I00000000500000!\DQOOOOOOOOOOOOOOOUO

d

O O O O O O O O O O O O O O O O O O O O O O O O O O

DATE

NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
NOV
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC
DEC

TR?
ST]
SD]

RO‘
RC!

HS(

HSC
HSC
HSC
HSC

203

APPENDIX D

son. NITROGEN BALANCE PARAMETERS (1989/1990)

TRT Treatment number

STRAW I Weight of organic residue of previous crop (kg haq)
SDEP I Depth of incorporation of residue (cm)

SCN I C:N ratio of the residue of previous crop

ROOT I Root weight of previous crop

RCN I C:N ratio of roots of previous crop

MSCN8901 I MSCN (ID) 89 (Year) 01 (Inbred number)
ID TRT STRAW SDEP SCN ROOT RCN
MSCN8901 1 500. 10. 20. 1500. 20.
MSCN8901 2 500. 10. 20. 1500. 20.
MSCN8901 3 500. 10. 20. 1500. 20.
MSCN8901 4 500. 10. 20. 1500. 20.
MSCN8902 1 500. 10. 20. 1500. 20.
MSCN8902 2 500. 10. 20. 1500. 20.
MSCN8902 3 500. 10. 20. 1500. 20.
MSCN8902 4 500. 10. 20. 1500. 20.
MSCN8903 1 500. 10. 20. 1500. 20.
MSCN8903 2 500. 10. 20. 1500. 20.
MSCN8903 3 500. 10. 20. 1500. 20.
MSCN8903 4 500. 10. 20. 1500. 20.
MSCN8901 I MSCN (ID) 90 (Year) 01 (Inbred number)
ID TRT STRAN SDEP SCN ROOT RCN
MSCN9001 1 4509. 10. 30. 2750. 30.
MSCN9001 2 4629. 10. 30. 2750. 30.
MSCN9001 3 4599. 10. 30. 2750. 30.
MSCN9001 4 4727. 10. 30. 2750. 30.
MSCN9002 1 7298. 10. 30. 2750. 30.
MSCN9002 2 6933. 10. 30. 2750. 30.
MSCN9002 3 6795. 10. 30. 2750. 30.
MSCN9002 4 6688. 10. 30. 2750. 30.
MSCN9003 1 5957. 10. 30. 2750. 30.
MSCN9003 2 6142. 10. 30. 2750. 30.
MSCN9003 3 6242. 10. 30. 2750. 30.
MSCN9003 4 5969. 10. 30. 2750. 30.

204

APPENDIX D

SOIL PROFILE PROPERTIES (1989/1990)

NOTE: These parameters are used for all treatments in both 1989 and
1990.

Soil Name I Elston Sandy Loam

Soil Albedo I 0.13

Upper limit of stage 1 soil evaporation I 8.00 mm

Soil water drainage constant, fraction drained per day I 0.40
SCS runoff curve number I 78.00

Annual average ambient temperature I 9.4 (°C)

Annual amplitude in mean monthly temperature I 15.5 (°C)
Mineralization factor (DMOD) I 1.0

DLAYR I Thickness of soil layer (cm)

LL I Lower Limit of plant extractable soil water (cm3 cm'3)

DUL I Drained Upper Limit of plan; exgractable soil water (cm3 cm'3)
SAT I Saturated water content (cm 2:3-)

SW I Default soil water content ( cm'3)

NR I Root weighing factor to determine new root growth distribution
8D I Bulk Density (9 cm')

00 I Organic carbon (5)

DLAYR LL DUL SAT SW NR 80 OC

15. 0.035 0.177 0.385 0.177 1.000 1.52 1.04
15. 0.035 0.177 0.385 0.177 .900 1.55 1.04
15. 0.037 0.176 0.385 0.176 .700 1.72 0.71
15. 0.037 0.176 0.385 0.176 .700 1.62 0.71
15. 0.037 0.176 0.385 0.176 .500 1.63 0.47
15. 0.028 0.134 0.375 0.134 .080 1.61 0.47
30. 0.020 0.110 0.353 0.110 .040 1.56 0.42
30. 0.019 0.105 0.353 0.105 .005 1.56 0.17
30. 0.019 0.105 0.353 0.105 .002 1.58 0.00
30. 0.019 0.105 0.353 0.105 .002 1.58 0.00

205

APPENDIX D
SOIL PROFILE INITIAL CONDITIONS (1989)
TRT I Treatment number
DLAYR I Thickness of soil laye
SW I Soil water content ( cm'3)
NH4 I Soil ammonium content (mg elemental N per kg of soil)
No3 I Soil ammonium content (mg elemental N per kg of soil)
PH I pH of soil in a 1:1 soilxwater slurry

MSCN8901 I MSCN (ID) 89 (Year) 01 (Inbred)

NOTE: The same inputs were used for all three inbred for 1989.

TRT

01 MSCN8901 03 MSCN8901

DLAYR SW N84 N03 PH
15. .177 2.3 13.8 6.5 15. .177 1.9 10.6 6.5
15. .177 1.7 4.8 6.6 15. .177 2.4 5.1 6.6
15. .176 .4 3.1 6.8 15. .176 3.2 2.7 6.8
15. .176 4.0 2.1 7.0 15. .176 2.3 1.6 7.0
15. .176 2.7 1.7 7.0 15. .176 4.2 0.9 7.0
15. .134 3.2 1.1 7.0 15. .134 1.4 0.8 7.0
30. .110 3.8 0.7 7.0 30. .110 3.3 0.9 7.0
30. .105 6.2 0.5 7.0 30. .105 1.7 0.5 7.0
30. .105 1.0 0.5 7.0 30. .105 1.0 0.3 7.0
30. .105 1.0 0.5 7.0 30. .105 1.0 0.1 7.0

02 MSCN8901 04 MSCN8901
15. .177 8.9 10.0 6.5 15. .177 3.6 11.0 6.5
15. .177 2.7 3.5 6.6 15. .177 1.1 3.5 6.6
15. .176 2.5 2.5 6.8 15. .176 2.7 2.1 6.8
15. .176 2.2 1.3 7.0 15. .176 2.5 1.8 7.0
15. .176 4.1 1.0 7.0 15. .176 4.3 0.8 7.0
15. .134 1.7 1.1 7.0 15. .134 1.0 0.5 7.0
30. .110 1.5 0.8 7.0 30. .110 5.0 1.0 7.0
30. .105 1.7 0.6 7.0 30. .105 1.8 0.4 7.0
30. .105 1.0 0.4 7.0 30. .105 1.0 0.2 7.0
30. .105 1.0 0.2 7.0 30. .105 1.0 0.1 7.0

206

APPENDIX D

SOIL PROFILE INITIAL CONDITIONS (1990)

TRT I Treatment number

DLAYR I Thickness of soil laye

SW I Soil water content ( cm'3)
NH4 I Soil ammonium content (mg elemental N per kg of soil)
N03 I Soil ammonium content (mg elemental N per kg of soil)
PM I pH of soil in a 1:1 soi1:water slurry

MSCN9001 - MSCN (10) 90 (Year) 01 (Inbred)

NOTE: The same inputs were used for all three inbred for 1990.

TRT

01 NSCN9001 03 NSCN9001

DLAYR SW NH4 N03 PH
15. .177 2.5 4.9 6.5 15. .177 2.1 4.1 6.5
15. .177 2.3 4.7 6.6 15. .177 1.8 3.5 6.6
15. .176 2.0 4.3 6.8 15. .176 1.3 2.5 6.8
15. .176 1.5 3.4 7.0 15. .176 1.6 2.1 7.0
15. .176 1.3 3.0 7.0 15. .176 1.8 1.9 7.0
15. .134 0.6 0.9 7.0 15. .134 0.6 0.9 7.0
30. .110 0.6 0.8 7.0 30. .110 0.6 0.8 7.0
30. .105 0.5 1.1 7.0 30. .105 0.5 1.1 7.0
30. .105 0.6 1.4 7.0 30. .105 0.1 1.4 7.0
30. .105 0.6 1.6 7.0 30. .105 0.1 1.6 7.0

02 NSCN9001 04 NSCN9001
15. .177 2.4 4.1 6.5 15. .177 2.1 3.5 6.5
15. .177 2.2 2.8 6.6 15. .177 1.8 3.2 6.6
15. .176 1.8 2.1 6.8 15. .176 1.1 2.6 6.8
15. .176 1.4 1.6 7.0 15. .176 1.0 1.7 7.0
15. .176 1.2 1.4 7.0 15. .176 1.0 1.3 7.0
15. .134 0.6 0.9 7.0 15. .134 0.6 0.9 7.0
30. .110 0.6 0.8 7.0 30. .110 0.6 0.8 7.0
30. .105 0.5 1.1 7.0 30. .105 0.5 1.1 7.0
30. .105 0.6 1.4 7.0 30. .105 0.1 1.4 7.0
30. .105 0.6 1.6 7.0 0.1 1.6 7.0

30. .105

APPENDIX D

TRT

207

IRRIGATION WATER APPLICATION DATA (1989/1990)

3 Treatment

MSCN8902 I MSCN (ID) 89 (Year) 02 (Inbred)

DOY
AHIRR

NOTE:

1 MSCN8902

DOY

200
203
207
208
214
303
304

ANIRR

16.
32.
37.

1.
51.
58.
57.

2 MSCN8902

16.
20.
33.

1.
34.
60.
60.

I Day Of the Year
I Amount of irrigation water applied (mm)

The inputs were used for all the simulation runs for all the
inbreds in 1989. However, the last two irrigations listed for
each treatment were applied only to the lysimeter plots and were
not used for the other runs.

3 MSCN8902

200 12.
203 8.
207 11.
208 1.
214 27.
303 100.
304 100.

4 MSCN8902

200 16.
203 20.
207 27.
208 1.
214 30.
303 73.
304 73.

NOTE: These inputs were used for all the simulation runs for all the

1 MSCN9001

190
202
220

18.
22.
23.

2 MSCN9001

190
202
220

18.
22.
23.

inbreds in 1990.
3 MSCN9001
190 18.
202 22.
220 23.
4 MSCN9001
190 18.

202 22.
220 23.

208

APPENDIX D

NITROGEN FERTILIZER APPLICATION DATA (1989/1990)

TRT I Treatment
MSCN8901 I MSCN (ID) 89 (Year) 01 (Inbred)

DOY I Day Of the Year

AMPERT I Amount of fertilizer nitrogen added (kg ha”)
DFERT I Depth of incorporation (cm)

IPTYPE I Type of fertilizer used (see Table 9)

NOTE: These inputs were used for all the inbred simulations for 1989.

1 MSCN8901 3 MSCN8901

DOY AMFERT DFERT IFTYPE) 136 30.0 1.0 10
179 60.0 1.0 1

136 180.0 1.0 10

2 MSCN8901 4 MSCN8901

136 30.0 1.0 10 136 30.0 1.0 10

200 50.0 1.0 1 179 150.0 1.0 1

NOTE: These inputs were used for all t e inbred simulations for 1989
except for treatment 2. The 40 kg ha' shown here was only applied to
inbred 1, not to inbred 2 or 3.8.

3 MSCN9001
1 MSCN9001
122 30.0 1.0 1
122 180.0 1.0 1 179 60.0 1.0 1
4 MSCN9001
2 MSCN9001

122 00.0 1.0 1
204 40.0 1.0 1

209

APPENDIX D

MANAGEMENT DATA (1989/1990)

NOTE: This data was used for all treatments and all inbreds for 1989
except for detasseling day and the day the male plants were
removed.

Day simulation began I 121

Day of sowing I 136

Plant population I 5.15 pl m'z

Row spacing I 0.762 m

Planting depth I 5.0 cm

Inbred l Inbred 2 Inbred 3

Detasseling day 203 208 208

Removal of male plants 213 233 233

NOTE: This data was used for all treatments and all inbreds for 1990
except for the detasseling day and the day the male plants were
removed.

Day simulation began I 115

Day of sowing I 122

Plant population I 5.15‘pl III'2

Row spacing I 0.762 m

Planting depth I 5.0 cm

Inbred 1 Inbred 2 Inbred 3

Detasseling day 201 208 208

Removal of male plants 226 226 226

210

APPENDIX D

GENETIC DATA (1989/1990)

Pl I Growing degree days (base 81°C) from seedling emergence to the end
of the juvenile stage (d °C).

P2 I Photoperiod sensitivity coefficient (1 hr”)

P5 I Growing degree days (base 0C) from silking to physiological
maturity (d °C).

P6 I New leaf area coefficient
G2 I New grains per plant coefficient

G3 I Potential kernel growth rate (mg kernel'1 dayIU.

These genetics varieties were used for all treatments for both 1988 and
1989.

Variety ID I Inbred 1

P1 I 200.00
P2 I 0.30
PS I 650.0
P6 I 0.85
G2 I 1.15
G3 I 7.50

Variety ID I Inbred 2

P1 I 220.00
P2 I 0.30
PS I 700.0
P6 I 0.90
G2 I 1.14
G3 I 7.35

Variety ID I Inbred 3

P1 I 240.00
P2 I 0.30
PS I 700.0
P6 I 0.75
G2 I 1.09
G3 I 7.10

BIBLIOGRAPHY

BIBLIOGRAPHY

Algozin, K.A. and Bralts, V.F. 1986. Report of results for irrigation system evaluation and
scheduling demonstration project. Cooperative Agreement 68-5D21-5—008, Soil
Conservation Service, USDA.

Algozin, K.A., V.F. Bralts, and J.T.Ritchie. 1988. Irrigation strategy selection based on yield,
water and energy use relationships: A Michigan example. J. of Soil and Water Cons.
43(5):428-43l.

Anderson, C.K., LR. Stone, and LS. Murphy. 1982. Corn yield as influenwd by in-season application
of nitrogen with limited irrigation. Agron. J. 74:396-401.

Anonymous, "Documentation for IBSNAT Crop Model Crop Model Input & Output Files, Version
1.1: for the Decision Support System for Agrotechnology Transfer (DSSAT V2.1)". Technical
Report 5. USAID/University of Hawaii.

Arkin, G.F., R.L. Vanderlip and J.T. Ritchie. 1976. A dynamic grain sorghum growth model.
Transactions of the ASAE 19(4):622-626, 630.

Baker, C.l-I., R.D. Horrocks, and CE. Goering. 1975. Use of the Gompertz function for predicting
corn leaf area. ASAE 18323826, 330.

Baker, D.M., J .R. Lambert and J .M McKinion. 1983. GOSSYM: A simulation of cotton crop
growth and yield. 134 p. South Carolina Agricultural Experiment Station, Clemson,
SC Bulletin 1089.

Boggess, W.G. and Ritchie, J.T. 1988. Economic and risk analysis of irrigation decisions in
humid regions. J. Prod. Agrie, 1:116-122.

Boote, Kl, J .W. Jones, G. Hoogenboom, G.G. Wilkerson, and SS. Jagtap. 1989. PNUTGRO
V1.02: Peanut crop growth simulation model. User’s Guide. Florida Agric. Exp. Stn.
Journal number 8420. University of Florida, Gainsville, FL.

Bralts, V.F., T.L. London, RC. Brown, and J.A. Westendorp. 1983. Report of results for
irrigation system evaluation and scheduling demonstration project. Cooperative
Agreement 58-5D21-3-16, Soil Conservation Service, USDA.

Carberry, P.S., R.C. Muchow and R1. McCown. 1989. Testing the CERES-Maize simulation
model in a semi-arid tropical environment. Field Crops Res. 20:297-315.

Carter, MW. and Poneleit, CG. 1973. Black layer maturity and filling period variations
among inbred lines of corn (Zea mays 1.). Crop Sci, 13:436-439.

211

212

Carter, N. 1986. Simulation modelling. 19, Plant Virus Epidemits. G.D. McLean, R.G.Garret and
W.G. Ruesink (eds.). Academic Press, Sydney, N.S.W., pp. 193-215.

Daynard, TB. and Duncan, J.C. 1969. The black layer and grain maturity in corn. Crop Sci.
9:473-476.

de Vos, RN. and Mallett, J.B. 1987. Preliminary evaluation of two maize (Zea mays L)
growth simulation models. S. Afr. J. Plant and Soil. 4(3):l31-136.

Edmeades, 6.0. and Daynard, T.B. 1979. The relationship between final yield and photosynthesis at
flowering in individual maize plants. Can. J. Plant Sci. 59:585-601.

Evans, LT. and De Datta, SK. 1979. The relationship between irradiance and grain yield of irrigated
rice in the tropics, as influenced by cultivar, nitrogen fertilizer applications and month of
planting. Field Crops Res. 2:1-17.

Fick, G.W. 1981. ALSIM 1 (Level 2) User’s Manual. Agron. Mimeo 81-35, Dept. of Agron.,
Cornell Univ., Ithaca, NY.

Fischer, RA and Maurer, OR. 1976. Crop temperature modification and yield potential in dwarf
spring wheat. Crop Sci. 2:1-17.

Fowler, W. 1967. Cultural practices for today’s seed field. pp. 53-58. I_I; Proc. 32nd Annual
Hybrid Corn Ind. Res. Conf. ASTA, Washington DC.

Godwin, D.C., J.T. Ritchie, U. Singh and L Hunt. 1989. A user’s guide to CERES-Wheat -
V2.10. Inter. Fert. Dev. Center, Muscles Shoals, AL 86 p.

Godwin, DC. and Jones, CA. 1991. Nitrogen Dynamics in Soil-Plant Systems. pp. 287-321. Ln:
Ritchie, J.T. and RJ. Hanks (eds.) Modeling Plant and Soil Systems. Agronomy Monograph
no. 31. ASA, Madison, WI.

Godwin, DO, U. Singh and L Hunt. 1990. A user’s guide to CERES-Rice. Inter. Fert. Dev.
Center, Muscles Shoals, AL (In press).

Godwin, DC and Vlek, P.LG. 1985. Simulation of nitrogen dynamics in wheat cropping
systems. 19: W. Day and R.K. Arkin (eds.) Wheat growth and modeling. Plenum
Press, New York, NY.

Hawkins, RC and C00per, PJ.M. 1981. Growth, development and grain yield of maize. Experimental
Agric. 17:203-207. '

Hodges, T., 8.8. Johnson and LA. Manrique. 1989. SUBSTOR: A model of potato growth and
development. l_n Agronomy Abstracts: 16, ASA. Madison, WI.

Hodges, T., D. Botner, C. Sakamoto, and J.H. Haug. 1987. Using the CERES-Maize model
to estimate production for the US cornbelt. Agric. and For. Meteorol. 40:293-303.

Hodges, T. and Evans, D.W. 1990. Light interception model for estimating the effects of row
spacing on plant competition in maize. J. of Prod. Agric. 3:190-195.

213

Hoogenboom, G., J.W. White and J.W. Jones. 1989. A computer model for the simulation
of bean growth and development. La; Advances in bean (Phaseolus vulgaris L.)
research production. Publication CIAT no. 23, Cali, Columbia.

Hunter, R.B, CG. Motimore, and LW. Kannenberg. 1973. Inbred maize performance
following tassel and leaf removal. Agron. J. 65:471-472.

Jones, CA. and Kiniry, J.R.(eds.) 1986. CERES-Maize: A simulation model of maize growth and
development. Texas A&M University Press, College Station, Texas.

Jones, J.W., K.J. Boote, S.S. Jagtap, G. Hoogenboom and 6.6. Wilkerson. 1989 SOYGRO
V5.42. Soybean crop growth simulation model. User’s guide. Florida Agric. Exp. Stn.
Journal No. 8304. University of Florida, Gainsville, FL.

Kiniry, J.R., C.A. Jones, J .C. Toole, R. Blanchet, M. Cabelguenne, and DA. Spanel. 1989. Radiation-
use efficiency in biomass accumulation prior to grain-filling for five grain-crop species. Field
Crops Res. 20:51-64.

Kiniry, J .R. and Ritchie, J.T. 1985. Shade-sensitive interval of kernel number of maize.
Agron. J. 77:711-715.

King, R. Cooperative Extension Director, St. Joseph County. Centreville, Ml, 49032

Kittieson, KM. 1987. The groundwater problem in Michigan: An overview. pp. 69-84. l_n:
F.M. D’Itri and LG. Wolfson (eds.) Rural Groundwater Contamination. Lewis
Publishers, Chelsea, MI.

Knittle, KH. and Burris, J.S. 1985. Effect of kernel maturation on subsequent seedling vigor
in maize. Crop Sci. 16:851-855.

Legg, J .0, G. Stanford, and 0.1.. Bennett. 1979. Utilization of labeled-N fertilizer by silage corn under
conventional and no-till culture. Agron. J. 71:1009-1015.

Lightner, J. Research Agronomist. Pioneer Hi-Breds Int., Inc. Johnston, IO.

Maas, SJ. and Arkin, GP. 1980. TAMW: A wheat growth and development simulation
model. Research Center Program and Model Documentation No. 803, Texas Agric.
Exp. Stn., Blackland Research Center, Temple, TX.

Martin, D.L, D.G. Watts, LN. Mielke, K.D. Franks, and DE Eisenhauer. 1982. Evaluation of
nitrogen management for corn production using water high nitrate. Agron. J. 46:1056-1062.

Martin, E.C., J.T. Ritchie, and TL London. 1985. Use of the CERES-Maize model to
evaluate irrigation strategies. pp. 342-350. _l_r_;: Advances in Evapotranspiration.
Proceedings of the National Conference on Advances in Evapotranspiration.
December, 1985. Chicago, IL. Am. Soc. Ag. Egr., St. Joseph, MI.

Michigan Agricultural Statistics. 1989. Michigan Department of Agriculture, USDA. Bulletin
number 8956. 22 p.

Mielke LN., J.S. Schepers, and ILA. Richards. 1979. Nitrogen leaching Under Center-Pivot Irrigation
on Sandy Loam Soil. pp. 85-94. 13;: Proc. Tech. Conf, Irrig. Assn. Feb 18-21, 1979. San
Francisco, CA.

214

Muchena, P.K and Ritchie, J.T. 1989. Growth characterization for several maize genotypes. Field
Crops Res. In Review.

Muchow, RC. and Carberry, RS. 1989. Environment control of phenology and leaf growth in a
tropically adapted maize. Field Crops Res. 20:221-236.

Muchow, RC. and Carberry, PS. 1990. Phenology and leaf-area development in a tropical grain
sorghum. Field Crops Res. 23:221-237.

Newkirk, KM., J.C. Parker, J.C. Baker, EW. Carson, T.B. Brumback, Jr., and O. Balci. 1989.
User’s guide to VT-Maize Version 1.0 (R). Virginia Water Resources Res. Ctr.,
Virginia Polytechnic Inst. and State Univ., Blacksburg, VA.

Ong, GK. and Squire, GR. 1984. Response to temperature in a stand of pearl millet. J. of Exp.
Botany. 35(157):1233-1240.

Pelchat, A. 1986. Effects of tillage and winter cover crops on nitrogen requirements of corn in
Indiana. Ph.D. diss., Purdue University, West Lafayette, IN.

Piper, EL and Weiss, A. 1990. Evaluating CERES-Maize for reduction in plant population
or leaf area during the growing season. Agric. Sys. 33(3):199-213.

Rawson, HM. and Bagga, AK. 1979. Influence of temperature between floral initiation and flag leaf
emergence on grain number in wheat. Australian J. Plant Phys. 6:391-400.

Richards, FJ. 1959. A flexible growth function for empirical use. J. Exp. Bot. 10:290-300.

Richards, FJ. 1969. The quantitative analysis of growth. pp. 3-76. l_n: F.C. Stewart (ed.) Plant
physiology. Vol. VA. Analysis of growth: Behavior of plants and their organs. Academic Press,
New York.

Richardson, CW. and Wright, DA 1984. WGEN: A Model for Generating Daily Weather Variables.
USDA-ARS, ARS-8, 83 p.

Ritchie, J. T. Professor, Homer Nowlin Chair, Department of Crop and Soil Sciences, College of Ag.
and Nat. Res. Michigan State University, East Lansing, MI.

Ritchie, J.T. and Alagarswamy, G. 1989. Simulation of sorghum and pearl millet phenology.
pp. 24-26. 13: SM. Virmani, H.L.S. Tandon and G. Alagarswamy (eds.) Modeling the
growth and development of sorghum and pearl millet, Research Bulletin No. 12
ICRISAT. Andhra Pradesh, India.

Ritchie, J.T. and Johnson, BS. 1990. Soil and plant factors affecting evaporation. pp. 363-390. Ln;
Stewart, RA. and Nielsen, D.R. (eds.) Irrigation of Agricultural Crops. ASA, CSSA, and
SSSA, Madison, WI.

Ritchie, J .T., 8.8. Johnson, 8. Otter-Nacke and DC. Godwin. 1989. Development of a barley
simulation model. Final progress report. USDA no. 86-CRSR-2-2867. Michigan State
Univ., East Lansing, MI.

Ritchie, J.T. and Otter, S. 1985. Description and performance of CERES-Wheat: A user oriented
wheat yield model. USDA-ARS. ARS-38. pp. 159-175.

215

Ritchie, J.T., U. Singh, D. Godwin and L Hunt. 1989. A user’s guide to CERES-Maize -
V2.10. Int. Fert. Dev. Center, Muscle Shoals, AL 86 p.

Russelle, M.P., EJ. Deibert, R.D. Hauck, M. Stevanovic, and RA. Olson. 1981. Effects of water and
nitrogen management on yield and NIS-depleted fertilizer use efficiency of irrigated corn.
Agron. J. 45:553-558.

Russelle, M.P., R.D. Hauck, and RA. Olson. 1983. Nitrogen accumulation rates in irrigated maize.
Agron. J. 75:593-598.

Saul, MR. 1986. Effect of tillage, nitrogen fertilization, and cover crop management in continuous
corn production. M.S. thesis, Univ. of Kentucky, Lexington, KY.

Schepers, J .S. 1988. Cropping systems and environmental quality. pp. 167-178. 1;: W.L Hargrove (ed.)
Cropping strategies for efficient use of water and nitrogen. ASA Special Publication Number
51. ASA, CSSA, and SSSA, Madison, WI.

Schepers, J.S. and D1. Martin. 1983. Management practises to reduce nitrate leaching with irrigated
corn. pp. 85-91. l_n: Proceedings of the NWA Western Regional Conference. Oct 23-26,
1983. San Diego, CA.

Schepers, J.S., K.D. Frank and DO. Watts. 1983. Influence of irrigation and nitrogen fertilization on
ground water quality. In; Proc. XVII General Assembly Inter. Union Geodesy and
Geophysics, Aug. 15-22, 1983. Hamburg, Federal Republic of Germany.

Shayya, W.H., Bralts, V.F., and TR. Olmsted. A General Irrigation Scheduling Package for
Microcomputers. Computer and Electronics in Agriculture (in press).

Shoultz, D. 1985. An evaluation of parent delay techniques, pp.57-64. l_n: Proc. 40th Annual
Ind. Res. Cont. ASTA. Washington, DC.

Smika DE, DE Heerrnann, H.R. Duke, and AR. Bathchelder. 1977. Nitrate-N percolation through
irrigated sandy soil as affected by water management. Agron. J. 69:623-626.

Stapper, M. and Arkin, OF. 1980. CORNF: A dynamic growth and development model for
maize (Zea mays L). 91 p. Texas Agricultural Experiment Station, College Station,
TX. Program and model documentation No. 80-2.

Timmons, DR. and Dylla, AS. 1979. Nitrate Leaching on an Irrigated Soil in a Subhumid
Climate, pp. 95-102. _Ig: Proc. Tech. Conf., Irrigation Assn. Feb. 18-21, 1979. San
Francisco, CA.

Uehara, G. 1985. The International Benchmark Site Network for Agrotechnology Transfer,
pp. 271-274. l_n: W. Day and RK. Atkin (eds.) Wheat Growth and Modelling.
Plenum Publishing, New York, NY.

Utomo, M. 1986. Role of legume cover crops in no-tillage and conventional tillage corn production.
Ph.D. diss. Univ. of Kentucky, lexington.

Wagger, MG. and Mengel, DB. 1988. The role of nonleguminous cover crops in the efficient use of
water and nitrogen. pp. 115-127. 1g: W. L Hargrove (ed.) Cropping strategies for efficient use
of water and nitrogen. ASA Special Publimtion number 51. ASA, CSSA, and SSSA, Madison,
WI.

216

Wallach, D. and Goffinet, B. 1989. Mean squared error of prediction as a criterion for evaluating and
comparing system models. Ecol. Modelling, 42:299-306.

Watts D.G. and Hanks RJ. 1978. A soil-water-nitrogen model for irrigated corn on sandy soils.
Agron. J. 42:492-499.

Watts, D.G. and Martin, D.L 1981. Effects of water and nitrogen management on nitrate leaching
loss from sands. Trans, ASAE 24:911-916.

Wilkerson, G.G., J.W. Jones, K.J. Boote, and J.W. Mishoe. 1983. SOYGRO V5.0: Soybean
crop growth and yield model. 253 p. Technical Documentation. Ag. Egr. Dept. Univ.
of Florida, Gainsville, FL

Williams, J .R. and Renard, K.G. 1985. Assessment of Soil erosion and crop productivity with
process models (EPIC). pp. 67-103. Lg: R.F. Follett and BA. Stewart (eds) Soil
erosion and crop productivity. ASA, Madison, WI.

Worman, F.O., A.W. Biere, M.L Hooker, R.L Vanderlip and ET. Kanemasu. 1988.
Simulation analysis using physiological crop-response models: Alternative cropping
strategies for southwest Kansas. Kansas Agric. Exp. Stn., Bulletin 653.

Wych, RD. 1988. Production of hybrid seed corn. 1;: Corn and corn development, ASA
Monograph no. 18, 3rd edition. ASA, Madison, WI.

‘iiilllllilillllll“