The Effects of Varying Levels of Deficit Irrigation and

The Effects of Varying Levels of Deficit Irrigation and Episodic Drought Stress on

West Texas Cotton Cultivars

by

Fulvio R Simao, B.Sc., M.Sc.

A Dissertation

In

PLANT AND SOIL SCIENCE

Submitted to the Graduate Faculty

of Texas Tech University in

Partial Fulfillment of

the Requirements for

the Degree of

DOCTOR OF PHILOSOPHY

Approved

Dr. Glen L. Ritchie

Chair of Committee

Dr. Craig W. Bednarz

Dr. Stephan J. Maas

Dr. Eric F. Hequet

Dr. Jeff Johnson

Dr. Dale T. Griffee

Dominick Joseph Casadonte, Jr.

Dean of the Graduate School

May, 2013

Copyright 2013, Fulvio Rodriguez Simao

Texas Tech University, Fulvio Rodriguez Simao, May 2013

ii

ACKNOWLEDGMENTS

Thanks to God for all his gifts and blessings, and for providing the strength

necessary for these scholastic achievements. Thanks to my advisor Dr. Glen Ritchie,

for his challenges and contributions to my professional growth. To Dr. Craig Bednarz,

who was co-responsible to the creation of the main ideas of the research used in this

dissertation.

All the other committee members, Dr. Stephen Maas, Dr. Eric Hequet, Dr. Jeff

Johnson, and Dr. Dale Griffee, also made valuable suggestions for my dissertation

research. Also, thanks to Dr. Tomas Thompson who in the occasion when Dr. Bednarz

started working at Bayer Cropscience assumed my Advisory Committee Chairperson

position, making it possible to continue my research in the 2010 season.

A special acknowledgement for the valuable financial support Cotton

Incorporated, the Ogallala Aquifer Research Consortium, and Bayer Cropscience,

provided to our research group. I also have to say thanks to Epamig and Embrapa for

sponsoring my doctoral degree.

I also point out my gratitude to Texas Tech University, along with the Plant

and Soil Science Department for accepting me in its graduate program, supporting me,

and also providing aid in various research and scholastic activities. The support of

TTU staff and other graduate and undergraduate students was crucial for the

development of all research activities and for my personal improvement. Special

thanks for Vinicius Buffon, Matthew Stroud, Chase Snowden, Tyler Painter, Heath

Reeves, Even Motley, Luke Obenhaus, Bablu Sharma, Curtis Schaefer, Jon Sherman,

Ryan Gregory, and all other members “Bednarz-Ritchie Crew” that provided support

for my field activities. TTU also supported me through the management of the New

Deal and Quaker Avenue research farms by Paul Green, Phil Brown, Reagan Anders,

and Steve Oswald.


iii

Benjamin Mullinix was also valuable for this research solving some of our

statistical analysis and experimental design questions. Thanks to Mary for the

technical writing review of my manuscripts as well.

Thanks for all my professors, classmates, and co-workers that made my

experience at TTU even more exciting. I also want to express my appreciation for all

that directly or indirectly supported my dissertation research and encouraged me to

pursue my doctoral degree.

Finally, I will like to thank all my family and friends that supported me during

all the difficult moments related to being in a foreign country and pursuing an

advanced graduate degree; especially to my wife, Luciana Simao, who agreed to leave

Brazil, and follow me to United States and Lubbock, and for our daughter Sofia. It is

to Sofia that this work is dedicated.


iv

TABLE OF CONTENTS

ACKNOWLEDGMENTS ............................................................................................. ii

ABSTRACT .................................................................................................................. ix

LIST OF TABLES ......................................................................................................... x

LIST OF FIGURES .................................................................................................... xiii

LIST OF ABBREVIATIONS .................................................................................... xxii

1. DISSERTATION INTRODUCTION ........................................................................ 1

1.1. Definition of the problem ..................................................................................... 2

1.2. Rationale for Gathering Original Data .............................................................. 3

1.3. Hypothesis ............................................................................................................. 3

1.4. Dissertation Objectives......................................................................................... 4

References ..................................................................................................................... 5

2. LITERATURE REVIEW........................................................................................... 6

2.1. Cotton Growth Habit ............................................................................................ 7

2.2. Cotton Water Use .................................................................................................. 8

2.3 Cotton Irrigation .................................................................................................. 10

2.4. Irrigation Management for Cotton ................................................................... 11

2.5. Irrigation Effects on Boll Distribution .............................................................. 15

2.6. Economics of Irrigation Water Use ................................................................... 16

2.7. Literature Review Conclusions .......................................................................... 20

References ................................................................................................................... 22


v

3. EPISODIC DROUGHT EFFECTS ON COTTON YIELD AND FIBER QUALITY

...................................................................................................................................... 26

3.1. Abstract ................................................................................................................ 26

3.2. Introduction ......................................................................................................... 27

3.3. Materials and Methods ....................................................................................... 29

3.4. Results .................................................................................................................. 32

3.4.1. Yields ................................................................................................................. 33

3.4.2. HVI .................................................................................................................... 38

3.4.3. AFIS ................................................................................................................... 46

3.5. Discussion ............................................................................................................. 51

3.6. Conclusions .......................................................................................................... 53

References ................................................................................................................... 55

4. AGRONOMIC WATER USE EFFICIENCY DIVERSE COTTON CULTIVARS

SUBJECTED TO VARYING IRRIGATION LEVELS.............................................. 58

4.1. Abstract ................................................................................................................ 58

4.2. Introduction ......................................................................................................... 59


4.4. Results .................................................................................................................. 63

4.4.1. Yields ................................................................................................................. 64

4.4.2. Agronomic Water Use Efficiency ................................................................... 66

4.4.3. Fiber Quality .................................................................................................... 68

4.5. Discussion ............................................................................................................. 72


vi

4.6. Conclusion ............................................................................................................ 74

References ................................................................................................................... 75

5. COTTON CULTIVARS FRUIT DISTRIBUTION UNDER EPISODIC

IRRIGATION INTERRUPTION ................................................................................ 77

5.1. Abstract ................................................................................................................ 77

5.2. Introduction ......................................................................................................... 78


5.4. Results .................................................................................................................. 83

5.5. Discussion ............................................................................................................. 87

5.6. Conclusion ............................................................................................................ 88

References ................................................................................................................... 89

6. COTTON PHYSIOLOGICAL PARAMETERS AFFECTED BY EPISODIC

IRRIGATION INTERRUPTION ................................................................................ 91

6.1. Abstract ................................................................................................................ 91

6.2. Introduction ......................................................................................................... 91


6.4. Results and Discussion ........................................................................................ 97

6.5. Conclusion .......................................................................................................... 112

References ................................................................................................................. 113

7. IRRIGATION AND EPISODIC DROUGHT MANAGEMENT EFFECTS ON

THE PROFITABILITY OF COTTON CULTIVARS IN WEST TEXAS ............... 115

7.1. Abstract .............................................................................................................. 115


vii

7.2. Observation ........................................................................................................ 116

7.3. Introduction ....................................................................................................... 116

7.3.1. Literature review ............................................................................................ 117

7.3.2. Economic Analysis Conceptual Framework ................................................ 123

7.3.3. Objectives ........................................................................................................ 128

7.4. Materials and Methods ..................................................................................... 128

7.4.1. Episodic trials ................................................................................................. 129

7.4.2. Irrigation trials ............................................................................................... 130

7.4.3. Measurements and Analysis .......................................................................... 131

7.5. Results ................................................................................................................ 133

7.5.1. 2011 .................................................................................................................. 134

7.5.3. 2012 .................................................................................................................. 149

7.6. Discussion ........................................................................................................... 165

7.7. Summary and Conclusion ................................................................................ 165

References ................................................................................................................. 169

8. DISSERTATION CONCLUSION ........................................................................ 171

9. BIBLIOGRAPHY .................................................................................................. 175

APPENDICES ........................................................................................................... 181

A. WEATHER PARAMETERS ................................................................................ 182

B. SOIL MOISTURE ................................................................................................. 186

C. TABLE OF AGROCHEMICALS APPLIED ....................................................... 187

D. IRRIGATED COTTON BUDGET ....................................................................... 188


viii

E. ADDITIONAL MEASUREMENTS DATA TABLES AND GRAPHS .............. 190

E.1. Plant grotht ....................................................................................................... 190

E.2. Total Water ....................................................................................................... 195

E.3. Yields ................................................................................................................. 198

E.4. Fiber Quality ..................................................................................................... 202

E.5. AWUE ................................................................................................................ 209

E.6. Bolls Distribution .............................................................................................. 210

E.7. Gas Exchange .................................................................................................... 211

E.8. Economics .......................................................................................................... 214


ix

ABSTRACT

Water availability and plant water use efficiency are important issues facing cotton

production in West Texas. Trials were conducted to compare the growth, gas

exchange, yield, and quality of West Texas cultivars subjected to different levels and

timings of water deficit. In the first experiment, cotton cultivars were subjected to

consistent levels of irrigation, ranging from full irrigation to dryland after first square.

In the second experiment, cultivars were exposed to episodic water deficit events

during the season. The effects of irrigation treatments upon physiological parameters

such as leaf photosynthesis, transpiration, and temperature, were measured, and

agronomic water use efficiency was calculated for each treatment. An economic

analysis was also performed to verify treatment profitability. Yield and most of the

quality parameters presented were significantly affected by drought episode and

cultivar selection. The data shows that in all years and locations the irrigation strategy

with no irrigation interruption provided the best yields. Regarding episodic drought

events, the highest reduction in yields was caused by irrigation interruption at the early

flowering stage. Agronomic water use efficiency was affected by varying irrigation

levels in two out of three years. In 2012, the agronomic water use efficiency observed

on fully irrigated cultivars was statistically higher than for other irrigation levels.

These results can be important in supporting the development of water management

strategies for irrigated cotton.


x

LIST OF TABLES

3.1. Total precipitation and irrigation applied per season and location. ...................... 33

3.2. Statistical significance (p-values) for the F test due to effects on

seed-cotton yield .......................................................................................................... 34

3.3. Seed-cotton yield (kg ha-1

) for the 2010, 2011, and 2012 seasons.

LSD values at the 0.05 level ........................................................................................ 34

3.4. Significance of the analysis of variance for HVI variables as

affected by the irrigation episodic drought, cultivar or their interaction ..................... 39

3.5. Micronaire (no units) observed during the 2010, 2011, and 2012

seasons. LSD values at the 0.05 level .......................................................................... 40

3.6. HVI UHML lengths (inches) observed during the 2010, 2011, and

2012 seasons. LSD values at the 0.05 level ................................................................. 43

3.7. Significance of the analysis of variance for AFIS variables as


3.8. Maturity ratios (%) observed in 2010 and 2011. LSD values at the

0.05 level ...................................................................................................................... 48

3.9. Fineness (mTex) observed in 2010 and 2011. LSD values at the

0.05 level ...................................................................................................................... 49

3.10. AFIS individual fiber length averaged by weight (inches)

observed in 2010 and 2011. LSD values at the 0.05 level ........................................... 50

4.1. Total precipitation and irrigation applied per season and location ....................... 64

4.2. Statistical significance (p-values) of the study factors and

interaction in the variable seed-cotton yield during the 2010, 2011, and

2012 seasons ................................................................................................................ 64

4.3. Seed-cotton yield (kg ha-1

) during the 2010, 2011, and 2012

seasons. LSD values at the 0.05 level. ......................................................................... 65

4.4. Statistical significance (p-values) of the study factors and

interaction in the variable AWUE during the 2010, 2011, and 2012

seasons ......................................................................................................................... 66

4.5. Agronomic water use efficiency (kg ha-1

mm-1

) during the 2010,

2011, and 2012 seasons. LSD values at the 0.05 level. 1 kg ha-1

mm-1

=

1 g daL-1

....................................................................................................................... 67


xi

4.6. Significance of the analysis of variance for HVI variables as


4.7. Significance of the analysis of variance for AFIS variables as

affected by the irrigation level, cultivar or their interaction ........................................ 71

7.1. Adjusted cotton enterprise budget (1 ha) ............................................................ 126

7.2. Episodic Drought trials. Statistical significance (p-values) for the F

test due to effects on seed-cotton yield ...................................................................... 133

7.3. Irrigation levels trials. Statistical significance (p-values) of the

study factors and interaction in the variable seed-cotton yield during the

2010, 2011 and 2012 seasons ..................................................................................... 134

7.4. Significance (p-values) of the analysis of variance for production

variables as affected by episodic irrigation interruption, cultivar or their

interaction in 2011...................................................................................................... 134


economics related variables as affected by episodic irrigation

interruption, cultivar or their interaction in 2011. ...................................................... 142


variables as affected by the irrigation episodic drought, cultivar or their

interaction in 2012...................................................................................................... 150


economics related variables as affected by irrigation management,

cultivar or their interaction in 2012. ........................................................................... 157

D.1. Estimated costs and returns per Acre Cotton, Drip Irrigated -

Herbicide-tolerant, Insect-resistant 2012 Projected Costs and Returns

per Acre (Source: Texas A&M University Department of Agricultural

Economics, 2012) ....................................................................................................... 188

E.1. Total precipitation and irrigation applied per season and location

(four irrigation levels analysis) .................................................................................. 198

E.2. Statistical significance (p-values) of the study factors and

interaction in the variable cotton seed yield during the 2010, 2011 and

2012 seasons. 20 cultivars in 2010, 5 in 2011 and 6 in 2012, and also

three irrigation levels in 2010, and four irrigation levels in 2011 and

2012. ........................................................................................................................... 200

E.3. Significance of the analysis of variance for HVI variables as

affected by the irrigation episodic drought, cultivar or their interaction

in 2011 ........................................................................................................................ 202


xii

E.4. significance of the analysis of variance for AFIS variables as

affected by the irrigation episodic drougth, cultivar or their interaction

in the 2010 season. Eight cultivars analysis ............................................................... 206

E.5. significance of the analysis of variance for AFIS variables as

affected by the irrigation episodic drougth, cultivar or their interaction

in the 2011 season ...................................................................................................... 206


xiii

LIST OF FIGURES

3.1. Accumulated Precipitation (mm) observed during the 2010 and

2012 seasons at the Texas Tech New Deal research farm and in 2011

and 2012 at the Texas Tech Quaker Avenue research farm. ........................................ 30

4.1. Accumulated Precipitation (mm) observed during the 2010 and

2011 seasons at the Texas Tech New Deal research farm (ND), and in

2012 at the Texas Tech Quaker Avenue research farm (Qk). ...................................... 63

5.1. Smoothed boll distribution averaged over four cultivars subjected

to five different irrigation interruptions periods during the 2011 (A) and

2012 (B) seasons. ......................................................................................................... 84

5.2. Smoothed Bolls distribution of four cultivars subjected averaged

over five different episodic drought periods during the 2011 (A) and

2012 (B) seasons. ......................................................................................................... 86

6.1. Accumulated Precipitation (mm) observed during the 2010 season

in the Texas Tech New Deal research farm and in 2011 and 2012 in the

Texas Tech Quaker Avenue research farm. ................................................................. 95

6.2. Photosynthesis measured on the FM9180 cultivar subjected to

three different irrigation regimes during the 2010 season at the Texas

Tech New Deal research farm. ..................................................................................... 98

6.3. Photosynthesis averaged on the FM9180 and DP0935 cultivars

submitted to different irrigation interruptions periods during the 2011

(a) and 2012 (b) seasons at the Texas Tech Quaker Avenue research

farm. ............................................................................................................................. 99

6.4. Photosynthesis measured in cultivars DP0935 and FM9180

submitted to five different irrigation interruptions periods during the

2011 season 105 days after planting at the Texas Tech Quaker Avenue

research farm. Black bars represent one Least Significant Difference

(LSD). LSD (p=0.05) = 2.49 µmol CO2 m-2

s-1

. ........................................................ 101

6.5. The effects of transpiration (a) and stomatal conductance (b) on

photosynthesis of cultivars DP0935 and FM9180 submitted to five

different irrigation interruptions periods during the 2012 season. ............................. 103

6.6. Relation of the air-leaf temperature difference on photosynthesis

of cultivars DP0935 and FM9180 submitted to five different irrigation

interruptions periods during the 2012 season. ........................................................... 105


xiv

6.7. Effect of the intercellular-ambient carbon dioxide concentration

ratio on photosynthesis of cultivars DP0935 and FM9180 submitted

to five different irrigation interruptions periods during the 2012 season................... 106

6.8. Physiological leaf water use efficiency averaged on the FM9180

and DP0935 cultivars submitted to different irrigation interruptions

periods during the 2012 seasons at the Texas Tech Quaker Avenue

research farm. ............................................................................................................. 107

6.9. Seed-cotton yield (a) and Agronomic water use efficiency of

cotton cultivars DP0935 and FM9180 subjected to five different

irrigation interruptions periods during the 2012 season. Black bars

represent one Least Significant Difference. LSD Yield (p=0.05) = 183

kg ha-1

. LSD AWUE (p=0.05) = 0.37 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1

g daL-1

. ....................................................................................................................... 109

7.1. Seed-cotton yield of four cotton cultivars subjected to five different

irrigation interruptions periods during the 2011 season at the Texas

Tech Quaker Avenue research farm. Black bars represent one least

significant difference. LSD(0.05) = 203 kg ha-1

. ....................................................... 135

7.2. Lint yield of four cotton cultivars subjected to five different



significant difference. LSD (0.05) = 91 kg ha-1

. ........................................................ 136

7.3. Lint percentage of four cotton cultivars subjected to five different



significant difference. LSD(0.05) = 1.9 %. ................................................................ 137

7.4. Agronomic seed-cotton water use efficiency of four cotton

cultivars subjected to five different irrigation interruptions periods

during the 2011 season at the Texas Tech Quaker avenue research

Farm. Black bars represent one least significant difference. LSD(0.05)

= 0.61 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

. ....................................................... 138

7.5. Agronomic lint water use efficiency of four cotton cultivars

subjected to five different irrigation interruptions periods during the

2011 season at the Texas Tech Quaker Avenue research farm. Black

bars represent one least significant difference. LSD(0.05) = 0.27 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

. .............................................................................. 138

7.6. Area to irrigate using four different irrigation interruptions periods

with a water volume equal to the used to fully irrigate 1 ha during the

2011 season at the Texas Tech Quaker Avenue research farm. ................................. 139


xv

7.7. Seed-cotton production of four cotton cultivars subjected to five

different irrigation interruptions periods in an area adjusted to use the

same volume of water of a fully irrigated ha during the 2011 season at

the Texas Tech Quaker Avenue research farm. Black bars represent one

least significant difference. LSD(0.05) = 262 kg. ...................................................... 140

7.8. Lint production of four cotton cultivars subjected to five different

irrigation interruptions periods in an area adjusted to use the same

volume of water of a fully irrigated ha during the 2011 season at the

Texas Tech Quaker Avenue research farm. Black bars represent one

least significant difference. LSD(0.05) = 117 kg. ...................................................... 140

7.9. Marginal lint yield to full irrigation of four cotton cultivars




mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

. .............................................................................. 141

7.10. Premium or discount received by four cotton cultivars subjected

to five different irrigation interruptions periods during the 2011 season

at the Texas Tech Quaker Avenue research farm. Black bars represent

one least significant difference. LSD(0.05) = US$ 0.05 kg-1

. .................................... 143

7.11. Average number of first positon bolls per node from four cultivars


2011 season. season at the Texas Tech Quaker Avenue Research Farm. .................. 144

7.12. Total revenue from four cotton cultivars subjected to five

different irrigation interruptions periods during the 2011 season at the

Texas Tech Quaker Avenue research farm. Black bars represent one

least significant difference. LSD(0.05) = US$ 203.88 ha-1

. ....................................... 145

7.13. Profit or loss of four cotton cultivars subjected to five different



significant difference. LSD(0.05) = US$ 203.88 ha-1

. ............................................... 146

7.14. Breakeven price of four cotton cultivars subjected to five different



significant difference. LSD(0.05) = US$ 0.60 kg-1

. ................................................... 147

7.15. Profit of an increased area irrigating using four different irrigation

interruptions periods with a water volume equal to the used to fully

irrigate 1 ha during the 2011 season at the Texas Tech Quaker Avenue

research farm. Black bars represent one least significant difference.

LSD (0.05) = US$ 264.86. ......................................................................................... 148


xvi

7.16. Marginal value product to full irrigation of four cotton cultivars



bars represent one least significant difference. LSD(0.05) = US$ 1.33

ha-1

mm-1

. US$ 1.00 ha-1

mm-1

= US$ 0.01 10-2

L-1

. ................................................ 149

7.17. Average seed-cotton yield over cultivars DP0395 and FM9180

subjected to eight different irrigation regimes during the 2012 season at

the Texas Tech Quaker Avenue research farm. Black bars represent one

least significant difference. LSD(0.05) = 338 kg ha-1

. ............................................... 150

7.18. Lint yield of cultivars DP0395 and FM9180 subjected to eight

different irrigation regimes during the 2012 season at the Texas Tech

Quaker Avenue research farm. Black bars represent one least

significant difference. LSD(0.05) = 152 kg ha-1

. ....................................................... 151

7.19. Average seed-cotton water use efficiency from cultivars DP0395

and FM9180 subjected to eight different irrigation regimes during the



mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

. .............................................................................. 152

7.20. Area equivalent to the use of the water volume in one full

irrigated ha for seven different irrigation regimes during the 2012

season at the Texas Tech Quaker Avenue Research Farm. ....................................... 153

7.21. Prodution of an expanded area for the same total water volume

averaged from cultivars DP0395 and FM9180 subjected to eight

different irrigation regimes during the 2012 season at the Texas Tech

Quaker Avenue research farm. Black bars represent one least

significant difference. LSD(0.05) = 379 kg. .............................................................. 154

7.22. Marginal from dryland lint yield of cultivars DP0395 and

FM9180 subjected to eight different irrigation regimes during the 2012

season at the Texas Tech Quaker Avenue research farm. Black bars

represent one least significant difference. LSD(0.05) = 0.39 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

. .............................................................................. 155

7.23. Marginal to full irrigation lint yield of cultivars DP0395 and



represent one least significant difference. LSD(0.05) = 0.37 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 0.1 g L-1

. ............................................................................... 156

7.24. Premium or discount of cultivars DP0395 and FM9180 subjected

to eight different irrigation regimes during the 2012 season at the Texas


xvii


significant difference. LSD (p=0.05) = US$ 0.0099 kg-1

........................................... 158

7.25. Average boll distribution per node from cultivars DP0395 and


season at the Texas Tech Quaker Avenue research farm. .......................................... 159

7.26. Revenue obtained from cultivars DP0395 and FM9180 subjected

to eight different irrigation regimes during the 2012 season at the Texas


significant difference. LSD (p=0.05) = US$ 326.00 ha-1

. .......................................... 160

7.27. Profit obtained from cultivars DP0395 and FM9180 subjected to

eight different irrigation regimes during the 2012 season at the Texas



. .......................................... 161

7.28. Breakeven price for cultivars DP0395 and FM9180 subjected to



significant difference. LSD (p=0.05) = US$ 0.26 kg-1

............................................... 162

7.29. Profit obtained from cultivars DP0395 and FM9180 subjected to



significant difference. LSD (p=0.05) = US$ 376.98. ................................................. 163

7.30. Marginal value product to full irrigation of cultivars DP0395 and

FM9180 subjected to seven different irrigation regimes during the 2012


represent one least significant difference. LSD (p=0.05) = US$ 0.88 ha-

1 mm

-1. US$ 1.00 ha

-1 mm

-1 = US$ 0.01 10

-2 L

-1. ..................................................... 164

A.1. Daily ET0 measured during the 2010 season at the Texas Tech

University New Deal research farm. .......................................................................... 182




University Quaker Avenue research farm. ................................................................. 183




University Quaker Avenue research farm. ................................................................. 184

A.6. Cumulative Degree Days for a 60 ºF base temperature at the Texas

Tech University research farms. ................................................................................ 184


xviii

A.7. Cumulative Precipitation at the Texas Tech University research

farms. .......................................................................................................................... 185

A.8. Daily average temperature measured during the 2010 season at the

Texas Tech University New Deal research farm. ...................................................... 185

B.1. Soil moisture content in treatments subjected to six (bottom) three

(middle) or zero (top) weeks of irrigation interruption, as a function of

days after planting, measured on the cultivar FM9180 during the 2010

season at the Texas Tech University New Deal research farm. ................................. 186

E.1. Average plant height of eight cotton cultivars subjected to different

irrigation interruptions during the 2010 season at the Texas Tech New

Deal research farm. Black bars represent standard errors. ......................................... 190

E.2. Average number of nodes from eight cotton cultivars subjected to

different irrigation interruptions during the 2010 season at the Texas

Tech New Deal research farm. Black bars represent standard errors. ....................... 191

E.3. Average plant height of 20 cultivars subjected to three different

irrigation levels during the 2010 season. Black bars represent standard

errors. ......................................................................................................................... 191

E.4. Average number of nodes from 20 cultivars subjected to three

different irrigation levels during the 2010 season. Black bars represent

standard errors. ........................................................................................................... 192

E.5. Height averaged from five cotton cultivars subjected to varying

irrigation regimes during the 2011 season at the Texas Tech New Deal

research farm. ............................................................................................................. 193

E.6. Number of nodes averaged from five cotton cultivars subjected to

varying irrigation regimes during the 2011 season at the Texas Tech

New Deal research farm. ............................................................................................ 193

E.7. Heights from five cotton cultivars averaged from four varying

irrigation regimes during the 2011 season at the Texas Tech New Deal

research farm. ............................................................................................................. 194

E.8. Number of nodes from five cotton cultivars averaged from four

varying irrigation regimes during the 2011 season at the Texas Tech

New Deal research farm. ............................................................................................ 194

E.9. Total water per treatment during the 2010 season in the Texas

Tech New Deal research farm. ................................................................................... 195


Tech Quaker Avenue research farm. .......................................................................... 195


xix


Tech Quaker Avenue research farm. .......................................................................... 196







E.15. Lint yield of eight cotton varieties subjected to three different


represent the standard error. ....................................................................................... 198

E.16. Lint yield of four cotton cultivars subjected to five different


represent the standard errors. ..................................................................................... 199

E.17. Yield of four cotton cultivars subjected to five different irrigation

interruptions periods during the 2012 season. Black bars represent the

standard errors. ........................................................................................................... 199

E.18. Yields of 20 cultivars subjected to three different irrigation levels

during the 2010 season. Black bars represent standard errors. .................................. 200

E.19. Seed-cotton yields of five cultivars subjected to four different


errors. ......................................................................................................................... 201

E.20. Average seed cotton yield of two cotton cultivars subjected to

five different irrigation interruptions periods during the 2011 season.

Different letters indicate LSD differences at p=0.05. ................................................ 201

E.21. Micronaire of eight cotton cultivars subjected to different



E.22. Micronaire of four cotton varieties subjected to five different



E.23. HVI UHML length of eight cotton cultivars subjected to different



E.24. HVI UHML length of four cotton cultivars subjected to different

irrigation interruptions during the 2011 season at the Texas Tech

Quaker Avenue research farm. Black bars represent standard errors. ....................... 204


xx

E.25. Micronaire of five cultivars subjected to four different irrigation

levels during the 2011 season. Black bars represent standard errors. ........................ 205

E.26. Maturity ratio of five cultivars subjected to four different


errors. ......................................................................................................................... 205

E.27. Fineness of eight cotton cultivars subjected to different irrigation

interruptions during the 2010 season at the Texas Tech New Deal

research farm. Black bars represent the standard errors. ........................................... 207

E.28. Maturity ratios of eight cotton varieties subjected to different

irrigation interruption periods during the 2010 season. Black bars


E.29. Fineness of four cotton cultivars subjected to different irrigation

interruptions during the 2011 season at the Texas Tech New Deal

research farm. ............................................................................................................. 208

E.30. Maturity ratios of four cotton varieties subjected to five different

irrigation interruption periods during the 2011 season. ............................................. 208

E.31. Agronomic Water Use Efficiency from 20 cultivars subjected to

three different irrigation levels during the 2010 season. Black bars

represent standard errors. ........................................................................................... 209

E.32. Smoothed Bolls distribution averaged on five cultivars subjected

to four different irrigation levels during the 2011 season. ......................................... 210

E.33. Smoothed Bolls distribution of five cultivars subjected averaged

over four different irrigation levels during the 2011 season. ..................................... 210

E.34. Average Photosynthesis of cultivars FM9180 and FM832

subjected to four different irrigation levels during the 2011 season.

Black bars represent standard errors. ......................................................................... 211

E.35. Photosynthesis averaged on the FM9180 and DP0935 cultivars

submitted to different irrigation interruptions periods during the 2012

season. Black bars represent standard errors.............................................................. 211

E.36. Average Photosynthesis of cultivars FM832, FM9180, and

DP0935 submitted to four different irrigation levels during the 2012

season. Black bars represent standard errors.............................................................. 212

E.37. Average Photosynthesis of cultivar FM9180 subjected to four

different irrigation levels during the 2012 season. Black bars represent

standard errors. ........................................................................................................... 212


xxi

E.38. Photosynthesis measured on two cultivars subjected to five

different irrigation interruptions periods during the 2011 season 110

days after planting. Black bars represent the standard deviation. .............................. 213

E.39. Profitability of eight cotton varieties subjected to three different


represent standard errors. ........................................................................................... 214

E.40. Breakeven price of eight cotton varieties subjected to three

different irrigation interruptions periods during the 2010 season. Black

bars represent standard errors..................................................................................... 214


xxii

LIST OF ABBREVIATIONS

A – Net leaf photosynthesis

AFIS – Advanced fiber information system

ANOVA – Analysis of variation

AWUE – Agronomic water use efficiency

Ci – Internal carbon dioxide

Ci/Ca – Intercellular-ambient carbon concentration ratio

DAP – Days after planting

DP – Delta and Pine

E – Leaf transpiration

ED – Episodic drought

Es – Seasonal irrigation efficiencies

ET – Evapotranspiration

ET0 – Reference evapotranspiration

ETC – Crop evapotranspiration

FM - Fibermax

gs – Stomatal conductance

HVI – High volume instrument

KC – Crop coefficient

LEPA – Low energy precision application

LSD – Least significant difference


xxiii

LSDirr – Least significant difference among irrigation treatments

LSDvar – Least significant difference among cultivars

LSDint – Least significant difference for irrigation-cultivar interactions

LWUE – Leaf physiological water use efficiency

MPP – Marginal physical product

MVP – Marginal value product

ND – Texas Tech New Deal research farm

PGR – Plant growth regulator

PWUE – Physiological water use efficiency

Qk – Texas Tech Quaker Avenue research farm

SSI – Subsurface drip irrigation

ST – Stoneville

TTU – Texas Tech University

USDA – United States Department of Agriculture

VPD – Vapor pressure deficit

WUE – Water use efficiency


1

CHAPTER I

DISSERTATION INTRODUCTION

Water management is a very important factor in any crop production system.

According to Denning et al. (2001), the Texas High Plains region is one of the most

important cotton producing areas in the United States. Farm-level yields in the Texas

High Plains are significantly influenced by a few critical factors, including irrigation

water application rates. These factors need to be collectively considered and managed

by producers in order to maximize yield and optimize water usage, including irrigation

application rates. Because water supplies and availability continue to decline in Texas,

it is imperative that we develop, determine, and adopt the most efficient irrigation

systems and management techniques (Bordovsky et al., 2000).

This dissertation outlines current irrigation management techniques and related

issues (Chapter 2); the effects of episodic drought on production and physiology of

cotton including yield and quality (Chapter 3), boll distribution (Chapter 5), and gas

exchange (Chapter 6); the effects of deficit irrigation on yield, quality, and water use

efficiency (Chapter 4); and, the economics of both episodic drought stress and deficit

irrigation (Chapter 7). Chapters 8 and 9 contain the comprehensive conclusions and

bibliography for the dissertation. Our results build upon previous research by

identifying the most critical developmental stages for stress, expanding upon the

current knowledge of the effects of water deficit on cotton physiology and phenology,


2

and outlining the economic impact of varying irrigation management regimes on

multiple cultivars of cotton.

1.1. Definition of the problem

Cotton is the primary field crop in West Texas, with 3 to 4 million acres

planted every year. According to USDA (2011), in 2009, Texas produced 4.5 million

bales out of the 11.8 million bales produced nationally (38%), and in 2010, Texas

produced 7.7 million bales out of the 17.6 million bales produced nationally (43%). In

West Texas, almost half of the cotton area is irrigated. The principal source of water

for irrigated cotton in West Texas is the Ogallala Aquifer, but withdrawals from the

aquifer in the Texas High Plains have exceeded the recharge rate for decades, resulting

in water depletion and making further withdrawals more difficult (Colaizzi et al.,

2009). Therefore, improving irrigation management practices is important for

extending the Ogallala sustainability (Howell et al., 2004; Torell et al., 1990), as well

as increasing farmers profit.

Additional information about crop response to water can help growers better

decide how to manage irrigation and improve overall water management and returns.

In order to improve cotton irrigation management, we conducted studies to determine

the effects of water restriction on irrigated commercial cotton cultivars; our studies

consisted of comparing the effect of varying irrigation levels or episodic water

restriction on a set of diverse cultivars. Our hope is that our results will lead to more

efficient water management strategies by cotton producers.


3

1.2. Rationale for Gathering Original Data

Comparing the response of cotton cultivars to varying levels of deficit

irrigation and differing episodic water stress events will provide information that can

help to improve cotton production through better management strategies and cultivar

selection. Episodic drought has not been analyzed in depth in cotton, and our work

also expands on deficit irrigation research by increasing the number of irrigation

levels; of cultivars, several with unique growth habits; and gas with exchange

measurements throughout the season, contributing to knowledge gained from previous

research.

1.3. Hypothesis

Previous research has suggested that:

i) Commercial cotton cultivars exhibit a variety of responses to varying levels of

water deficit, with some displaying growth habits that make their yield less

affected by drought conditions.

ii) There may also be a differential response to episodic water deficit stress,

resulting in differences in yield and quality that are cultivar-specific.

iii) Episodic drought events affect yield and fiber quality differently, depending on

the timing of the events.


4

1.4. Dissertation Objectives

i) Determine the effects of episodes of water stress on cotton cultivars yield and

quality.

ii) Evaluate the Agronomic Water Use Efficiency of a diverse set of cotton

cultivars subjected to deficit irrigation management, compared to dry land and

fully irrigated crops.

iii) Describe the effects of episodic drought on fruiting distribution among cotton

cultivars.

iv) Quantify physiological processes, including leaf photosynthesis, transpiration,

temperature, and Physiological Water Use Efficiency (PWUE) for cotton

cultivars under multiple irrigation regimes.

v) Evaluate the profitability of cotton cultivars under several irrigation and

episodic drought regimes.


5

References

Bordovsky, J.P., Lyle W.M., Segarra E. (2000) Economic Evaluation of Texas High

Plains Cotton Irrigated by LEPA and Subsurface Drip. Texas Journal of

Agriculture and Natural Resources 13:67-73.

Colaizzi, P., Gowda P., Marek T., and Porter D. 2009. Irrigation in the Texas High

Plains: A brief history and potential reductions in demand. Irrigation and

Drainage 58:257-274.

Denning, M.L., Ramirez O.A., Carpio C. (2001) Impact of Quality on the Profitability

of Irrigated Cotton Production on the Texas High Plains, in: N. C. Council

(Ed.), Beltwide Cotton Conference, Memphis, TN. pp. 208-216.

Howell, T.A., Evett S.R., Tolk J.A., Schneider A.D. (2004) Evapotranspiration of

Full-Irrigated, Deficit-Irrigated, and Dryland Cotton on the Northern Texas

High Plains. Journal of Irrigation and Drainage Engineering 130:277-285.

DOI: 10.1061/(ASCE)0733-9437(2004)130:4(277).

Torell, L.A., Libbin J.D., Miller M.D. (1990) The Market Value of Water in the

Ogallala Aquifer. Land Economics 66:163-175.

USDA. (2011) New Release

<http://www.nass.usda.gov/Statistics_by_State/Texas/Publications/cg20311.pd

f>, USDA.

http://www.nass.usda.gov/Statistics_by_State/Texas/Publications/cg20311.pdf%3e



6

CHAPTER II

LITERATURE REVIEW

Cotton is the main crop in West Texas, and the Ogallala Aquifer is the main

water source (Colaizzi et al., 2009). Almost half of all cotton area in this region is

irrigated and the irrigated area is responsible for the majority of the production.

Mathis (2009) states that an imperative issue facing agriculture production in the

South Plains is water availability. Over time, water has been withdrawn out of the

Ogallala Aquifer (Texas High Plains) at a rate that exceeds its ability to recharge

(Colaizzi et al., 2009).

Depletion of water sources in dry climates leads to unsustainable irrigated

systems. The Texas High Plains exemplifies this challenge. Agriculture in this region

depends heavily on irrigation, using about 95% of the total water withdrawn from the

Ogallala Aquifer (Allen et al., 2007). Withdrawals from the aquifer in the Texas High

Plains have exceeded the recharge rate for decades, resulting in water depletion and

also making further withdrawals more difficult (Colaizzi et al., 2009). Therefore,

improving irrigation for cotton production is important for increasing farmers’ profits

and contributes for the Ogallala Aquifer’s sustainability in West Texas.


7

2.1. Cotton Growth Habit

Cotton is an indeterminate growth plant. It is difficult to clearly distinguish

between the crop growth periods, because vegetative growth, flowering, and boll

formation overlap. Some stages, however, have been described by Doorembos and

Kassam (1979). The duration of each of the plant development stages is a function of

the weather and the planting date.

Lanza and Penna (2007) state that the optimal planting date is a function of

region specific-climatic factors and can change each year depending on the rain

pattern. In West Texas, planting normally occurs between April and July, depending

on operational factors and soil temperature. For insurance purposes, the West Texas

planting cutoff month is June.

Doorembos and Kassam (1979) stated that cotton flowering is influenced by

temperature. They also stated that cotton plants in the flowering stage are very

sensitive to the effects of intensive rain during the flowering and bolls opening stages;

moisture compromises pollen exchange and reduces fiber quality.

Nutrient availability can also affect cotton Water and nitrogen are two of the

largest constraints to cotton production in the West Texas Southern High Plains

(Bronson et al., 2006), and both can affect growth habits (Ritchie et al. 2009; Mullins

and Burmester, 2010). Therefore, the cotton irrigation must be considered as an

important and integrated part of the crop management practices.


8

2.2. Cotton Water Use

Water is essential for the existence of life; plant growth depends more on the

amount of water available than on any other single environmental factor (Kramer and

Boyer, 1995). As stated by Kirkham (2005), water is the major environmental factor

limiting plant growth. Therefore, we can consider fundamental the study of plant-

water relations.

Probably the most important aspect of cotton crop irrigation management is the

plant’s water use. According to Fryxell (1986) locations where wild Gossypium

species are found in nature have diverse rainfall distribution and total amount. Jordan

(1986) states that for cotton (Gossypium hirsutum L.) soil water availability is one of

the primary factors that influences, and perhaps even controls, production of potential

fruiting points, retention of squares and bolls, and also crop yield.

The cotton water use also influences the crop practices. For example, due to

strong year-to-year variability of water supply and demand in most cotton growing

regions, farmers must be able to estimate relative crop water use, water availability,

and the impact of water deficit at varying growth stages to adapt their crop

management (Hake and Grimes, 2010).

According to Barreto et al. (2007), based on lysimeters measurements

throughout three seasons, the water use of cotton in the semi-arid region of Northeast

Brazil varies from 450 to 700 mm throughout the entire cycle. Martin et al. (1999) in a

similar research conduct in Maricopa, Arizona, reported an average seasonal Crop


9

Evapotranspiration (ETC), or crop water use, of 945 mm. While presenting a similar

system to measure daily crop water use, Lascano (2000) reported that daily cotton

water use peaked close to 9 mm in 1994 and more than 10 mm in 1995 in West Texas

conditions.

Evapotranspiration (ET) is the sum of the water evaporated from the soil and

transpired by plants. One method of modeling crop evapotranspiration (ETC) is

through the use of the reference evapotranspiration (ET0) and crop coefficient (KC)

using the method described by Allen et al. (1998). The ET0 is calculated from climate

parameters using the FAO Penman-Monteith equation, and resemble the ET of a well-

irrigated reference surface. ETC can then be calculated by multiplying the ET0 by the

KC, which is based on the relative evapotranspiration of the crop at any growth stage

compared to the reference surface.

For cotton, Allen et al. (1998) suggests a KC of 0.35 during early vegetative

growth initial stage (KC_ini), a KC from 1.15 to 1.20, for cotton in the intermediate

stage (KC mid) that occurs during flowering, and a KC between 0.5 and 0.7 for cotton in

the final stage (KC end) when boll development occurs. They considered a maximum

plant height from 1.2 to 1.5 m. It is also possible to estimate cotton water use using

remotely sensed crop ground cover, as indicated by Rajan (2007).

In a three year study conducted in Bushland, TX, Tolk and Howell (2010)

found that cotton lint yield correlated closely with ET in most soil types and years.

Hake and Grimes (2010) summarized that only with a solid understanding of optimal

plant-water relations at different growth stages can an optimal water-management


10

program be developed. This understanding of cotton-water relations is a key for the

study of cotton physiology, and for a successful cotton crop management.

2.3 Cotton Irrigation

The adoption of irrigation technologies can impact cotton production in many

ways. Ritchie et al. (2009) compared plant boll distribution in upland cotton that had

been irrigated with subsurface drip (SSI) to cotton irrigated with an overhead sprinkler

system. They concluded that SSI irrigation resulted in decreased early-season fruit

loss; this caused heavier carbohydrate sinks, and decreased plant height and boll

production near the top of the plant.

Two of the most prevalent irrigation systems in West Texas are the low energy

precision application (LEPA) and SSI. Since both systems can be considered localized

irrigation, it is expected for they would have potential seasonal irrigation efficiencies

(Es) as high as those described by Keller and Bliesner (1990) and Bernardo et al.

(2006) for trickle irrigation (Es ≥ 90%).

Previous studies reported irrigation system type affects cotton yield

(Bordovsky et al., 2000; Whitaker et al., 2008). Bordovsky et al. (2000) pointed out

that the advantages of LEPA over SSI include a lower initial cost and less

management and maintenance requirements. However, SSI increases cotton lint yield

and improves water use efficiency (WUE) (Bordovsky et al., 2000).

According to Bufon (2010), irrigation of crops using SSI continues to increase

in the Texas High Plains. However, information on the effect of drip-tape positioning


11

and irrigation strategies on the wetted soil area is needed to optimize rainwater

harvesting. Since Texas High Plains’ wells have limited water capacities, better use of

available resources is necessary to supply crops water needs.

In summary, water is the first limiting factor in any production system. This

fact alone underscores the importance of studies related to cotton water use and the

adoption of irrigation technologies. Furthermore, when it comes to irrigated cotton

production systems, implementing the best irrigation management becomes an

essential strategy to increase farmers’ profit and preserve scarce water resources.

2.4. Irrigation Management for Cotton

Water management is a critical issue for agricultural production. In both

dryland and irrigated production systems, farmers can conserve water through

management practices, such as cultivation with reduced tillage, crop and cultivar

selection, planting date, and crop rotation. In order to improve water management in

irrigated fields the use of more efficient irrigation methods and control of the irrigation

schedule are recommended. It is also important to note that irrigation management is

of key importance in determining the return on investment that a particular irrigation

system will provide.

The effects of water management in irrigated cotton production systems have

been reported by several authors. According to Silvertooth et al. (2000), the adequate

delivery of water to the crop is essential for establishing and maintaining good cotton

yield, particularly in arid or semiarid regions that are dependent upon irrigation.


12

In a study regarding the effect of irrigation levels on the herbaceous cotton

growth, de Sousa et al. (2008) concluded that the models of best fit, based on

application amount, followed a linear trend for plants height response, while following

a quadratic trend for stalk diameter and number of leaves.

In Turkey, Dagdelen et al. (2009) reported cotton response, in yield, to varying

irrigation levels. Bednarz et al. (2003) described that in Georgia, with proper

management, irrigation can increase lint yield by more than 350 kg ha-1

. In West

Texas, based on higher temperatures, and lower precipitation, we can expect even

higher cotton yield differences due to irrigation management. Mills (2010) reported

cotton responses in yield to varying irrigation levels in four commercial West Texas

cultivars, differences of more than 2000 kg ha-1

of lint cotton were observed between

irrigation levels.

Doorembos and Kassam (1979) recommend that during the flowering stage,

water must be supplied at times and in quantities to balance cotton vegetative and

reproductive growth. The authors suggested that water deficits in the initial stage until

flowering might cause more poignant effects in yield compared to those after this

stage; however, moderated water deficits just enough to limit vegetative growth during

flowering, result in good boll formation and high yields despite reduced number of

flowers. It is important to notice that the level of water deficit considered “moderated”

was not quantified.

Irrigating cotton plants when they have consumed lower percentages water of

the available soil can result in significant yield increase compared with yields


13

obtained from crops irrigated after an increased amount of soil moisture was

consumed (Jackson and Tilt, 1986).

Doorembos and Kassam (1979) also highlight that the water supply that will

maximize yield may be adjusted to the water needs of the crop in each growth period.

They suggest the optimum use of the available water after planting; the optimum water

use may be obtained by moistening the root zone to 1.8 m depth and completely using

available water in this layer. This strategy is supposed to increase the root system

activity.

Another water saving strategy proposed by Doorembos and Kassam (1979) is

carried out using all available water in the root zone. This is done by opportune

irrigation interruption in the end of the final growth period (early irrigation

termination). Cotton irrigation termination has been discussed in more detail in recent

studies (Martin et al., 2007; Silvertooth et al. 2000; Silvertooth et al. 2005; Sneed

2010) .

Another water saving strategy suggested by Doorembos and Kassam (1979)

was to stop water supply during the flowering until the crop had absorbed about 70%

of soil available water. They explain that the water supply should be adjusted to the

crop development stage, and that depending on the weather and total growth period,

cotton may need from 700 to 1300 mm of total water. These authors estimated that by

combining the irrigation management practices they proposed it may be possible to

save up to 20% of the water used by the crop without a big yield loss; however further

studies are necessary to verify the outcome of these strategies with the new cotton


14

cultivars that are now available and the selection of the more efficient irrigation

systems.

Irrigation management not only directly affects cotton production, but also

interacts with the effects of other crop practices. Balkcom et al. (2006), in a study

regarding cotton yield and fiber quality from irrigated tillage systems in the Tennessee

Valley, found that irrigation regimes influenced all yield and fiber quality variables

studied except the HVI strength of the cotton fibers. Likewise, Bronson et al. (2001)

reported that for most years of his study, cotton lint yields were influenced by

irrigation under different tillage systems.

Irrigation levels and cultivar interaction was also reported by Nunes Filho et al.

(1998) in a study conducted in Pernambuco state (Northwestern Brazil). Regression

equations adjusted to yield followed a quadratic model with high significance for the

variable irrigation depth with maximum yields of 3051, 2763, and 2423 kg ha-1

for

varieties CNPA 7H, CNPA Precoce 1 and CNPA 6H, with 836, 882 and 821 mm of

water applied, respectively (Nunes Filho et al., 1998). They used furrow irrigation. In

this study the fiber uniformity varied with irrigation management and intensity; while

length, strength, and fineness were not affected in the studied conditions with these

cultivars. Martin et al. (2007) summarized that proper irrigation of cotton requires

good water management as well as good crop management.

According to Doorembos and Kassam (1979) when cotton’s water supply is

limited, different from other plants such as bananas, it is possible to obtain a larger

total cotton production by increasing the planted area and providing partial water


15

needs, rather than fully supplying a limited area. In a banana plantation, for example, a

higher total production would be achieved using just the area were water is available

to completely supply the plants irrigation needs rather than using a larger area but not

fully irrigated. This is not the case with cotton.

Several factors account for differences in cotton yield among different

irrigation levels. One of the easiest ways to quantify the results of additional irrigation

is to analyze the distribution of fruit on the plants. Boll distribution can give insight

into crop growth habits. Bolls on different parts of the plant also may have varying

fiber qualities. Irrigation effects on cotton boll distribution will be discussed in more

detail in the following section.

2.5. Irrigation Effects on Boll Distribution

Mills (2010) conducted a study which showed that irrigation had a positive

effect on the number of bolls per plant. He concluded that irrigation affects yield on a

field level, on a plant level, and even within individual bolls.

Reducing the percentage of total yield produced at inner fruiting positions

through reduced plant densities increased the source-to-sink ratio during boll filling in

this region of the canopy, resulting in improved fiber properties (Bednarz et al., 2006).

Consequently, we can conclude that modifications in crop management may increase

the source-to-sink ratio during boll filling of the remaining fruiting positions (i.e. outer

fruiting positions), possibly resulting in greater improvements in fiber quality.


16

Irrigation management may be a crop practice with the potential to improve

cotton fiber quality from improved boll distribution. Ritchie et al. (2009) already

described that irrigation system selection alone can affect cotton boll distribution.

Cotton is a crop that requires close monitoring of water, nutrient, and pesticide

application to assure an optimum boll occurrence, and consequently an economically

viable yield at the end of the season (Martin et al., 2007). In order to make cotton

production profitable it is necessary to adjust complex irrigation and crop management

to production economics. Economics of irrigation water use is discussed in the next

section.

2.6. Economics of Irrigation Water Use

Since irrigated cotton production is an economic activity, farmers need to

maximize their profits as a function of inputs used. Hake and Grimes (2010)

concluded that for management of crop water supply, water-yield functions provide a

useful means of analyzing water use productivity.

Moore and Negri (1991) provided a policy simulation of a 10% reduction in

water allocation for ten major crops on farms served by the Bureau of Reclamation.

They indicated that production response to reduced water supply would affect the

national price of three out of ten major crops. These authors suggested that studies

regarding crop production and response to water can also be useful for water

allocation models and in defining water allocation policies. We can also assume that


17

developing more efficient water management strategies would reduce the impact of a

water supply reduction in the commodities prices.

Robinson et al. (2010) estimated the economic value of irrigation water

shortfalls and mitigation responses of farmers in the Lower Rio Grande Valley of

Texas. They showed that South Texas farmers react to risk by diversifying their crop

mix, which has implications for the imputed value of water and soil resources. From

Robinson et al. (2010) results we can see the importance of economic studies related

to water episodic drought, especially those relating water shortage scenarios.

Bordovsky et al. (2000) conducted an economic evaluation of Texas High

Plains cotton irrigated by LEPA and SSI. They stated that irrigation management

economic evaluations are as important as irrigation systems’ profitability studies.

The main conclusion of the Denning et al. (2001) study was that knowledge

and consideration of the effect of management decisions on lint quality can

substantially increase expected profitability and reduce profit variation. These authors

also concluded that better decision-making ability would improve profitability of farm

operations.

In research conducted by Mills (2010), it was also found that returns were

nonlinearly related to irrigation amount. Denning et al. (2001) also concluded cotton

had non-linear responses to irrigation amounts, so Mills (2010) suggested that the

level of water use to maximize profit is before maximum yield. This conclusions also

agrees with the basic concepts of production economics described by Doll and Orazen


18

(1992). Production Economics is the application of the principles of microeconomics

in agriculture (Doll and Orazem, 1992).

Mills (2010) also stated that the need for better irrigation management is

apparent and that, in the Texas High Plains, applying 0.64 cm of water per day is too

much in most years. They suggest that more conscientious water management would

increase profit margin. Based on his results we can also suggest that adding more

precise water management strategies, e.g. relating crop stages with an adequate

irrigation depth can increase farmers’ profits even more.

Mathis (2009) concluded, in a study regarding irrigation response in cotton in

the West Texas High Plains, that environmental factors heavily influence yield,

quality, and profitability of cotton, and that producers must try to achieve early crop

maturity without sacrificing yield and quality. In this study, conducted from 2007 to

2008, the returns from lint ranged from $1,308.34 to $ 2,851.61 per hectare.

Further observation of the effects of irrigation management on economics is

needed. Site-specific irrigation and nitrogen management study for cotton done by

Bronson at al. (2006), on the Texas High Plains, showed that dollar returns to

irrigation were greatest at the highest irrigation level, with exception of an

exceptionally wet year, such as 2004. This indicates the need of repeating this type of

study and analysis in different seasons and locations, which can produce different

results.

In Sneed’s (2010) thesis, he found that depending on the cost of irrigation

water, the most profitable time to terminate irrigation is not always the highest


19

yielding. Depending on the cost of irrigation water and the availability of water, a

producer could use less water and have equal or higher profits. Cultivar selection also

had an impact on profits according to Sneed’s (2010) research.

Wilde (2008), in a study about optimal economic combination of irrigation

technology and cotton varieties in the Texas High Plains, estimated gross margins and

net returns above total variable and fixed irrigation costs for varieties and irrigation

systems with different irrigation levels. He found that producers could increase gross

margins by adopting new varieties. His estimations showed that SSI irrigation can

produce higher net returns than LEPA center pivot systems. He also demonstrated the

importance of managing production according to environmental factors. Wilde’s

(2008) analysis emphasized the importance for cotton producers to be informed and

properly manage their production.

One final note from Denning et al. (2001) was a warning about using the

production response models estimated in his study. Though statistically sound,

Denning et al. (2001) models were based on just three years of experimental data from

Lubbock County. We can conclude from these recommendations that cotton response

to water needs to be studied for greater period of time and in different environments.

Summarizing, irrigation strategy studies and their economic relevance are

important from different stand points, especially in a scenario such as West Texas

where there is increased irrigation use in conjunction with water shortage.


20

2.7. Literature Review Conclusions

Water supply significantly influences irrigated cotton yield. We presented

some cotton crop characteristics related to its water consumption and provided

evidence that in an irrigated environment, cotton irrigation management must not be

considered isolated, but actually must be integrated with other crop practices.

Different water management strategies were suggested by Doorembos and

Kassam (1979) which make possible to reduce irrigation depth without significant

reductions on yields. Other cultivation practices, such as cultivar choice, can affect

optimum irrigation management (Nunes Filho et al., 1998; Bronson et al. 2001;

Balkcom et al. 2006).

Important studies on cotton response to water and related issues are

summarized in this literature review. All previous research about water response can

help cotton growers to decide where to place water and improve overall water

management and water returns. As for irrigation management of the new cotton

cultivars in West Texas, we suggested the need for determining whether or not cotton

yield would be affected by water restriction in different growth stages (episodic

drought). It would also be valuable to know the effects of varying irrigation levels on a

set of different cotton cultivars. This knowledge can lead to the development of more

efficient irrigation strategies for the farmers in the Texas High Plains.

In conclusion, it was important that we implemented the additional studies

presented in the following chapters on irrigation application and cultivar selection in


21

order to support the development of water management strategies for irrigated cotton

in West Texas.


22

References

Allen, R.G., Pereira L.S., Raes D., Smith M. (1998) Crop evapotranspiration:

Guidelines for computing crop water requirements. (FAO Irrigation an

drainaige paper 56)

Allen, V.G., Baker M.T., Segarra E., Brown C.P. (2007) Integrated Crop-Livestock

Systems in Dry Climates. Agronomy Journal 99:346-360.

Balkcom, K.S., Reeves D.W., Shaw J.N., Burmester C.H., Curtis L.M. (2006) Cotton

Yield and Fiber Quality from Irrigated Tillage Systems in the Tennessee

Valley. Agronomy Journal 98:596-602.

Barreto, A.N., do Amaral J.A.B., da Silva e Luz M.J. (2007) Consumo hidrico do

algodoeiro herbaceo nas diferentes fases fenologogicas no municipio Irece -

BA, VI Congresso Brasileiro do Algodão, Uberlandia. pp. 4.

Bednarz, C.W., Nichols R.L., Brown S.M. (2006) Plant density modifies within-

canopy cotton fiber quality. Crop Science 46:950-956. DOI:

10.2135/cropsci2005.08-0276.

Bednarz, C.W., Hook J., Yager R., Cromer S., Cook D., Griner I. (2003) Crop Water

Use and Irrigation Scheduling, in: A. S. e. a. Culpepper (Ed.), Cotton

Research-Extension Report UGA/CPES Research-Extension Publication,

UGA, Georgia. pp. 61-64.

Bernardo, S., Soares A.A., Mantovani E.C. (2006) Manual de Irrigacao. 8 ed. Editora

UFV.




Bronson, K.F., Brooker J.D., Bordovsky J.P., J.W. K., Wheeler T.A., Boman R.K.,

Parajulee M.N., Segarra E., Nichols R.L. (2006) Site-Specific Irrigation and

Nitrogen Management for Cotton Producion in the Southern High Plains.

Agronomy Journal 98:212-219.

Bronson, K.F., Onken A.B., Keeling J.W., Torbert. H.A. (2001) Nitrogen Response in

Cotton as Affected by Tillage System and Irrigation Level. Soil Society of

America Journal 65:1153-1163.


23

Bufon, V.B. (2010) Optimizing Subsurface Drip Irrigation Design and Management

with Hydrus-2D/3D Model, Plant and Soil Science, Texas Tech University,

Lubbock, TX. pp. 143.(Doctoral Dissertation)




Dagdelen, N., Basal H., Yilmaz E., Gurbuz T., Akcay S. (2009) Different drip

irrigation regimes affect cotton yield, water use efficiency and fiber quality in

western Turkey. Agricultural Water Management Journal 96:111-120.




de Sousa, P.S., de Medeiros J.F., de Matos J.d.A., de Melo S.B., Ferreira R.d.C.

(2008) Efeito de laminas de irrigacao sobre o crescimento do algodoeiro

herbaceo. Revista Verde 3:06-11.

Doorembos, J., Kassam A.H. (1979) Yield Response to Water FAO, Rome. (FAO

Irrigation and drainage paper 33)

Doll, J.P., Orazen F. (1992) Production economics: theory with applications. Second

Edition ed. Krieger Publishing Company, Florida.

Fryxell, P.A. (1986) Ecological adaptations of gossypium species, in: J. Mauney and J.

M. Stewart (Eds.), Cotton Physiology, The Cotton Foundation.

Hake, K.D., Grimes D.W. (2010) Crop water management to optimize growth and

yield, in: J. M. Stewart, et al. (Eds.), Physiology of Cotton, Springer.

Jackson, L.E.B., Tilt P.A. (1986) Effects of irrigation intensity and nitrogen level on

the performance of eight varieties of upland cotton, (Gossypium hirsutum L.).

Agronomy Journal 60:p. 13-17.

Keller, J., Bliesner R.D. (1990) Sprinkle and Tricke Irrigation. Chapman & Hall.

Kirkham, M. B. (2005) Principles of Soil Plant and Water Relations. Elsevier

Academic Press.

Kramer, P.J., Boyer J.S. (1995) Water relations of plants and soils. Elsevier Science.

Lanza, M.A., Penna J.C.V. (2007) Algodão (Gossypium hirsutum L.), in: J. T. de

Paula Júnior and M. Venzon (Eds.), 101 culturas: manual de tecnologias

agrícolas, Epamig, Belo Horizonte, MG - Brazil. pp. 63-74.


24

Lascano, R.J. (2000) A general system to measure and calculate daily crop water use.


Martin, E.C., Stephens W., Wiedenfeld R., Bittenbender H.C., Beasley Jr. J.P., Moore

J.M., Nibling H., Gallian J.J. (2007) Sugar, Oil, and Fiber, in: L. R.J. and

Sojka (Eds.), Irrigation of Agricultural Crops, American Society of Agronomy,

Inc., Crop Science Society of America, Inc., Soil Science Society of America,

Inc. pp. 279-335.

Mathis, G. (2009) Irrigation Response in Cotton to Optimize Yield, Quality and

Profitability in the Texas High Plains, Plant and Soil Science, Texas Tech

University, Lubbock, TX. pp. 105. (Master of Science Thesis)

Mills, C.I. (2010) Analysis of drought tolerance and Water Use Efficiency in Cotton,

Castor, and Sorghum, Plant and Soil Science, Texas Tech University, Lubbock

- TX. pp. 203. (Doctoral Dissertation)

Moore, M. R. and Negri D. H. (1991) “A Multicrop Production Model of Irrigated

Agriculture, Applied to Water Allocation Policy of the Bureau of

Reclamation”. Journal of Agricultural and Resource Economics, 17(1): 29-43

Mullins, G.L., Burmester C.H. (2010) Relation of Growth and Development to

Mineral Nutrition, in: J. M. Stewart, et al. (Eds.), Physiology of Cotton,

Springer.

Nunes Filho, J., de Lima e Sa V.A., de Oliveira Junior I.S., Coutinho J.L.B., dos

Santos V.F. (1998) Efeito de lâminas de irrigação sobre o rendimento e

qualidade da fibra de cultivares de algodoeiro herbáceo (Gossypium hirsutum

L. r. latifolium Hutch). Revista Brasileira de Engenharia Agrícola e Ambiental

2:295-299.

Rajan, N. (2007) Estimation of Crop Water Use for Different Cropping Systems in the

Texas High Plains Using Remote Sensing, Plant and Soil Science, Texas Tech

University, Lubbock - TX. pp. 175. (Doctoral Dissertation)

Ritchie, G.L., Whitaker J.R., Bednarz C.W., Hook J.E. (2009) Subsurface Drip and

Overhead Irrigation: A Comparison of Plant Boll Distribution in Upland

Cotton. Agronomy Journal 101:1336-1344. DOI: 10.2134/agronj2009.0075.

Robinson, J.R.C., Michelsen A.M., Gollehon N.R. (2010) Mitigating water shortages

in a multiple risk environment. Water Policy 12:114–128.

Sneed, J. (2010) Irrigation Termination to Improve Fiber Maturity on the Texas High

Plains, Plant and Soil Science, Texas Tech University, Lubbock - TX. pp. 119.

(Master of Science Thesis)


25

Silvertooth, J.C., Gladima A., Tronstad R. (2005) Evaluation of Irrigation Termination

Effects on Yield and Fiber Quality of Upland cotton, 2004. Arizona Cotton

Report:31-46.

Silvertooth, J.C., Galadima A., Norton E.R., Moser H. (2000) Evaluation of irrigation

termination effects on fiber micronaire and yield of upland cotton., Arizona

Cotton Report, Arizona.

Tolk, J.A., and Howell T.A (2010) Cotton Water Use and Lint Yield in Four Great

Plains Soils. Agronomy Journal 102:904-91.

Whitaker, J. R.; G.L. Ritchie; C.W. Bednarz; and C.I. Mills. 2008. Cotton Subsurface

Drip and Overhead Irrigation Efficiency, Maturity, Yield, and Quality. Agron.

J. 100:1763-1768.

Wilde, C. (2008) Optimal economic combination of irrigation technology and cotton

varieties on the High Plains of Texas, Agricultural and Applied Economics,

Texas Tech University, Lubbock, TX. pp. 158. (Master of Science Thesis)


26

CHAPTER III

EPISODIC DROUGHT EFFECTS ON COTTON YIELD AND FIBER QUALITY

3.1. Abstract

An aspect of cotton irrigation management that has not been studied in detail is the

effect of drought stress over a specific period of time in the crop development stage.

Field experiments were conducted in 2010, 2011, and 2012 at the New Deal and

Quaker Avenue Texas Tech Research farms located in Lubbock, Texas. The following

episodic drought treatments were used: a full irrigation throughout the season: non-

irrigation from squaring to flowering; 3 weeks of non-irrigation beginning at early

flowering; 3 weeks of non-irrigation beginning at peak bloom; and non-irrigation from

peak bloom to the crop termination. Cultivars DP0912, DP0935, FM9170, and

FM9180 were chosen for a multi-year analysis. Yield and most fiber quality

parameters presented were significantly affected by drought episode or cultivar. In all

years and locations the irrigation strategy with no irrigation interruption provided the

best yields. Early flowering water stress resulted in lower yields than obtained through

the other episodic drought treatments. The results can be important for developing

more efficient water management strategies for cotton in West Texas.


27

3.2. Introduction

Cotton is a major crop in West Texas. According to the USDA (2011), in 2010

nearly 40% of all cotton production in the USA occurred in West Texas. Almost half

of the West Texas cotton acreage is irrigated, and the irrigated area is responsible for

the majority of the production. The main water source in West Texas is the Ogallala

Aquifer, which has been decreasing in capacity over time (Howell et al., 2004; Torell

et al., 1990), leading to decreases in well capacity, additional acres or deficit-irrigated

and dryland fields, and an increasing urgency to develop more efficient irrigation

strategies.

In addition to the question of irrigation quantity, another factor is the timing of

irrigation. It is accepted that cotton growth and development are marked by differing

water use, as well as differing sensitivities to drought. However, quantifying these

differences has been problematic. An understanding of the effects of episodic water

stress on cotton yield can help with cultivar selection and other management decisions.

It has been shown that irrigation depth has a direct effect on cotton yield in

different regions (Nunes Filho et al., 1998; Bednarz et al., 2003; Dagdelen et al., 2009,

AbdelGadir et al., 2012), also depending on the irrigation systems (Bordovsky et al.,

2000; Whitaker et al. 2008). In West Texas, most of the research on irrigation levels

relating to cotton has focused on differing irrigation types and quantities that are

maintained throughout a growing season (Mills, 2010).


28

Irrigation effects on production interacts with crop management practices, and

cultivar selection (Bronson et al., 2001; Balkcom et al., 2006; Campbell and Bauer,

2007; Pettigrew and Dowd, 2012). Cotton irrigation termination was also discussed in

previous studies (Doorembos and Kassam, 1979; Martin et al., 2007; Silvertooth et al.

2000; Silvertooth et al. 2005; Sneed 2010).

Although previous research has suggested that irrigation timing can affect

cotton yield (Kriegg, 2000; Bauer et al., 2012; Collins and Hake, 2012), an aspect of

irrigation management that has not been studied in detail is the effect of water stress

over a specific period of time in the crop development stage (episodic drought) on

yield and fiber quality in irrigated cotton. Understanding this could prove valuable in

several ways. For example, knowing the relative sensitivity of differing growth stages

could help determine the effects of an episode of drought or irrigation downtime,

potentially saving water that would otherwise be wasted on a crop that has already

been irreparably damaged. Also, ascertaining the relative sensitivity of different

cultivars to drought episodes could be beneficial for cultivar selection, leading to more

efficient irrigation strategies for the farmers, as well as for better decision making

under water limiting circumstances.

Therefore, the objective of this study was to evaluate the effects of episodic

drought periods on the yield and fiber quality of a group of commercial cotton

cultivars in West Texas.


29

3.3. Materials and Methods

Field experiments were conducted in 2010 at the Texas Tech New Deal

Research farm (New Deal 2010), in 2011 at the Texas Tech Quaker Avenue Research

farm (Quaker 2011) located in Lubbock, Texas, and in 2012 in both locations (New

Deal 2012 and Quaker 2012). The soil at the Quaker Avenue farm is an Amarillo-

Acuff sandy clay loam (Fine-loamy, mixed, superactive, thermic Aridic Paleustolls),

and the soil at the New Deal farm is a Pullman-Olton clay loam (Fine, mixed,

superactive, thermic Aridic and Torrertic Paleustolls).

Planting dates were 25 May 2010, 13 May 2011, and 16 May 2012 at the

Texas Tech Quaker Avenue Research farm, and 24 May 2012 at the Texas Tech New

Deal Research farm. Harvest dates were 20 November 2010, 11 October 2011, and 14

October 2012 at the Texas Tech Quaker Avenue Research Farm, and 29 October 2012

at the Texas Tech New Deal Research Farm.

Agronomical practices followed Texas A&M AgriLife Extension

recommendations for the Texas High Plains. Fertilizer in the form of 28-0-0-5 was

applied at a rate of 90 kg N ha-1

. Weed control included herbicide applications

(glyphosate), and mechanical hoeing. A conventional tillage system was used. Plant

growth regulators (PGR) were applied in 2010.

The experimental design was a split-plot with irrigation treatments as the main

plot and cultivars as the split-plot. There were three blocks in 2010 and four in 2011

and 2012. The experimental unit was two rows of cotton, measuring 10.7 m in 2010


30

and 12.2 m in 2011 and 2012. There were also two rows serving as borders in the side

of each experimental unit. Row spacing was one meter. Subsurface drip irrigation

(SSI) tape placed 0.25 m below each planted row.

The following irrigation treatments were used: a full irrigation throughout the

season: non-irrigation from squaring to flowering; 3 weeks of non-irrigation beginning

at early flowering; 3 weeks of non-irrigation beginning at peak bloom; and non-

irrigation from peak bloom to the crop termination. We considered the squaring stage

begin to occur when the first square (floral bud) was observed, and the early flowering

stage to begin when the first opened flower was observed. During the 2010 season,

heavy precipitation early in the growing season (Figure 3.1 DAP 11 to 51) prevented

the establishment of the treatments with irrigation interruption at squaring and at early

flowering.

Figure 3.1. Accumulated Precipitation (mm) observed during the 2010 and 2012

seasons at the Texas Tech New Deal research farm and in 2011 and 2012 at the Texas

Tech Quaker Avenue research farm.

0

50

100

150

200

250

300

1 11 21 31 41 51 61 71 81 91 101 111 121

Acc

um

ula

ted

pre

cip

ita

tio

n (

mm

)

Days After Planting

2010

2011

2012 Qk

2012 ND

Lubbock 1981-2010

average


31

In 2010, 10 commercial upland cotton cultivars were screened for the study,

from which DP0912, DP0935, FM9170, and FM9180 were chosen for a multi-year

analysis.

Weather was monitored with an automated weather station located close to the

study. The station monitored precipitation (Fig. 3.1) and other parameters used to

calculate irrigation needs. The total water applied to each irrigation treatment was

calculated by individual hydrometers attached to each irrigation inlet. Soil moisture

content was monitored weekly for all irrigation treatments using a neutron probe

device, 503 DR Hydroprobe (CPN International, Inc., Concord, CA).

All plots were harvested and weighed with a cotton stripper equipped with load

cells, weights were used to calculate the yields. During harvest grab samples were

collected randomly from the cotton harvest using an adaptation on the stripper for this

purpose, the samples were ginned for the fiber quality analysis. Cotton samples were

analyzed using the high volume instrument (HVI) and advanced fiber information

system (AFIS) at the Fiber and Biopolymer Research Institute, in Lubbock, Texas.

The statistical analysis was performed using the GLIMMIX Procedure in

SAS® software (SAS Inst., 2010) with an ANOVA followed by a mean separation at

5% level of probability using the LSMEANS statement. The GLIMMIX model

performs estimation and statistical inference for generalized linear mixed models by

incorporating normally distributed random effects, allowing GLIMMIX procedure to

properly separate random and fixed effects. SAS programming statements followed

the recommendations provided by Littell et al. (1996).


32

3.4. Results

As shown in Fig. 3.1, precipitation totaled, from planting to crop termination,

325 mm in 2010, 77 mm in 2011, 46 mm in 2012 at the Quaker farm, and 78 mm in

2012 at the New Deal location.

Although we had a significant amount of total precipitation in 2010, most of it

was concentrated in the beginning of the season (30 to 50 DAP). For this reason, the

early episodic drought treatments did not result in any difference in irrigation and were

removed. Table 3.1 shows the total precipitation and irrigation for all irrigation

treatments in all years.


33

Table 3.1. Total precipitation and irrigation applied per season and location.

Season - Location Irrigation Interruption Precipitation

(mm)

Irrigation

(mm)

Total Water

(mm)

2010 – ND no drought 325 270 595

3 weeks at peak bloom 325 162 487

peak bloom to termination 325 56 381

2011 – Qk no drought 77 330 407

squaring 77 259 336

early flowering 77 250 327



2012 – Qk no drought 46 598 644

squaring 46 481 527




2012 – ND no drought 137 554 691

squaring 137 509 646




ND = Texas Tech New Deal Research Farm.

Qk = Texas Tech Quaker Avenue Research Farm.

3.4.1. Yields

As shown in Table 3.2, we observed a highly significant effect on seed-cotton

yield for the episodic drought periods and cultivar, and also a statistically significant

irrigation by cultivar interaction in 2010 and 2011. Irrigation (episodic drought)

effects were statistically significant in all environments. Similary, interactions of

irrigation-cultivar effects were significative in all years and locations of this study.


34

Table 3.2. Statistical significance (p-values) for the F test due to effects on seed-cotton

yield

Factor 2010 2011 2012(Qk) 2012(ND)

--------------------------- p-value ----------------------------------------------

Irrigation 0.0007 ** < 0.0001 ** < 0.0001 ** < 0.0001 **

Cultivars 0.0013 ** 0.0025 ** 0.93 n.s. 0.43 n.s.

Irrigation*Cultivar 0.0015 ** 0.041 * < 0.0001 ** 0.0215 *

* = Statistically significant at 0.05 level

** = Statistically significant at 0.01 level

n.s. = Not statistically significant at 0.05 level

Table 3.3. Seed-cotton yield (kg ha-1

) for the 2010, 2011, and 2012 seasons. LSD

values at the 0.05 level

Irrigation Interruption Period Environment

and LSDs

Cultivar No Drought Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom

to termination

Average

---------------------------------- Yield (kg ha-1

) ------------------------------

2010 (ND) DP0912 4224 - - 2070 1866 2720

LSDirr = 363 DP0935 4224 - - 3003 3166 3465

LSDvar = 302 FM9170 4260 - - 2232 1984 2826

LSDint = 467 FM9180 3576 - - 3378 2613 3190

Average 4071 - - 2671 2407

2011 (Qk) DP0912 3546 2635 1292 2565 2361 2480

LSDirr = 139 DP0935 3462 2863 1246 2356 2459 2477

LSDvar = 110 FM9170 3291 2705 1088 2333 2231 2329

LSDint = 200 FM9180 3072 2468 1092 2547 2366 2309

Average 3343 2668 1179 2450 2354

2012 (Qk) DP0912 4430 2695 1697 2384 1980 2637

LSDirr = 120 DP0935 4043 2681 1929 2424 2054 2626

LSDvar = 87 FM9170 4257 2812 1753 2305 1993 2624

LSDint = 171 FM9180 3862 2474 1729 2751 2212 2605

Average 4148 2665 1777 2466 2060

2012 (ND) DP0912 6486 5195 3278 4514 4444 4783

LSDirr = 263 DP0935 6430 5444 3361 4375 4264 4775

LSDvar = 134 FM9170 6556 5542 3153 4375 4139 4753

LSDint = 301 FM9180 6445 5056 2903 4611 4389 4681

Average 6479 5309 3174 4469 4309

LSDirr = Least significant difference for irrigation (episodic drought) treatments

LSDvar = Least significant difference for cultivar effects

LSDint = Least significant difference for irrigation-cultivar interaction effects


35

Regarding irrigation timing effects, cotton crops that were fully irrigated

throughout all the season had significantly higher yields than crops that received an

episodic drought, this was consistant for all years and locations. Irrigation interruption

at the early flowering stage caused the highest yield reduction at the environments

where we had this episodic drought treatment. However, episodic drought at squaring

caused the lowest yield reduction compared to the no drought treatment, yields from

crops that received an irrigation interruption in the squaring stage were significantly

higher than yields from crops that were water stressed on the other stages of this study.

In 2010, when we didn’t have the squaring and early flowering treatments,

yields from the 3 weeks at peak bloom were significantly higher than the yields

observed at the peak bloom to termination treatment. Although in the 2012 Quaker

enviroment yields from the 3 weeks at peak bloom were significant higher than the

peak bloom to termination treatment yields, in the 2011 Quaker and 2012 New Deal

environments yields from these two treatments were not statistically different. We can

assume that the main cause of this absence of significance in 2011 was due to the

smaller crop season occurred that year as consequence of the temperatures. In 2011,

due to higher temperatures, the amount of accumulated degree days was higher than in

2010, so there were just four weeks between the peak bloom and crop termination. In

2010, six weeks were observed from peak bloom to crop termination.

Cultivar ranking varyied troughout the environments. In average, DP0935 and

FM9180 provided significatively higher yields than DP0912 and FM9170, in 2010. In

2011, the average yields of the Delta and Pine cultivars (DP0912 and DP035) were


36

significantly higher than the yields of the Fibermax cultivars (FM9170 and FM9180).

In both 2012 locations, average yields from all cultivars were not statistically

different. 2010 had a higher than average amount of precipitation (Fig. 3.1), and 2011

had higher temperatures than the average, the weather variances may explain the

differences on the cultivar yields. Furthermore, interactions on irrigation and cultivar

effects on yields were observed in all environments.

Cultivar yield response to episodic water stress varied from the four

enviroments in wich the research was repeated, as shown in Table 3.3. From Table

3.3 we can observe that in 2010 the highest yield of all treatments was observed when

cultivars FM9170, DP0935, and DP0912, were subjected to no irrigation suspension.

When there was an irrigation suspension, cultivars FM9180 and DP0935 had

sgnificantly higher yields than the other cultivars. Although there was a significant

amount of precipitation in 2010, a negative response of seed cotton yield to periods of

irrigation suspension was still observed. In 2011, at the fully irrigated treatment, the

yield obtained with cultivar DP0912 was statisticaly higher than the obtained with the

Fibermax cultivars FM9170, and FM9180, but at the other irrigation timing

treatments, yields from the four cultivars were not significantly different.

In regard to irrigation-cultivar effects on yield, in the 2012 Quaker location,

when there was no drought, highest yields among cultivars were achieved by DP0912

and FM9170. When there was an irrigation interruption at the early flowering stage,

the cultivar DP0912 provided yields statistically lower than at the other cultivars.

When water stress occurred at the squaring stage, FM9180 provided the only


37

statistically lower yield when compared with the other cultivars. For 3 weeks at peak

bloom, episodic stress FM9180 yields were statistically superior. From peak bloom to

termination treatment, FM9180 yields were higher than the other cultivars, with

exception of DP0935.

The New Deal 2012 enviroment provided the highest seed-cotton yields from

all locations and years with an average of more than 6,400 kg ha-1

for the non-stressed

treatments. In that enviroment there was no cultivar effects for treatments fully

irrigated and treatments that had the irrigation interrupted at the peak bloom stage. At

New Deal 2012 for treatments that received and episodic drought at the squaring

stage, the yields from the cultivars FM9170 and DP0935 were significantly higher

than FM9180 yields; for treatments that received early flowering irrigation

interruption, the Delta and Pine cultivars (DP0912 and DP0935) had yields

statistically higher than FM9180 yield.

As described in the previous paragraphs, the irrigation-cultivar interaction on

yields, varied among the four environments of this study. In the rainy and longer 2010

season, FM9180, an early maturity cultivar, probably already had a higher part of its

yield formed before the peak bloom period, so its yield was not as reduced due to the

late season drought episodes. Similar effects might also have occurred in 2010 with

DP0935, although this cultivar is usually classified as a mid-maturity.

2011 had higher temperatures than usual for the region, so the highest yield

from DP0912 under full irrigation might indicate a genetic resistence to heat-stress

from this cultivar. We observed in all environments that DP0912 was very sensitive to


38

episodic drought in all environments, therefore, we can assume that the yield reduction

due to water stress compensated the better heat tolerance, explaining why at the

treatments that received irrigation interruption no significant differences among

cultivar yields were observed in 2011.

Similarly, for the other enviroments, we can assume that differences on how

cultivars responded to irrigation timing, to be result of a triple environment-irrigation-

cultivar interaction. Environment effects included the differences in climate among

years and the differences of the soils at the Quaker and New Deal locations.

3.4.2. HVI

A summary of some of the HVI fiber quality results statistical significance is

presented in Table 3.4. All of the parameters were statistically different among

cultivars in all four environments (Table 3.4), with the exception of uniformity ratio in

2010.

As shown in Table 3.4, irrigation affected micronaire, length, strength,

uniformity, and elongation at the Quaker 2012 location, and most of these parameters

in 2011 and at the New Deal 2012 location. In 2010, only micronaire was affected by

irrigation. An irrigation-cultivar interaction was observed for micronaire in 2011 and

in the Quaker 2012 location.

In the Quaker 2011, New Deal 2012, and Quaker 2012 environments, highest

uniformity ratios were observed in the fully irrigated treatments, and at the FM9180

cultivar, with uniformity ratio values higher than 80, level where potentially no


39

discounts would be applyied. No significant interactions were observed. In 2010,

when no statistical differences were observed, uniformity ratio values were higher

than 81 in all treatments.

Table 3.4. Significance of the analysis of variance for HVI variables as affected by the

irrigation episodic drought, cultivar or their interaction

Environment Factor Irrigation Cultivar Interaction

2010 (ND) Micronaire 0.0235 * 0.0272 * 0.1106 n.s.

Length 0.0628 n.s. < 0.0001 ** 0.0562 n.s.

Uniformity ratio 0.6485 n.s. 0.4454 n.s. 0.1858 n.s.

Strength 0.1204 n.s. 0.0001 ** 0.0925 n.s.

Elongation 0.3302 n.s. < 0.0001 ** 0.4223 n.s.

2011 (Qk) Micronaire 0.0002 ** < 0.0001 ** 0.0002 **

Length < 0.0001 ** < 0.0001 ** 0.1289 n.s.

Uniformity ratio 0.0022 ** 0.0005 ** 0.1584 n.s.

Strength 0.0013 ** < 0.0001 ** 0.1714 n.s.

Elongation 0.0803 n.s. < 0.0001 ** 0.3256 n.s.

2012 (ND) Micronaire < 0.0001 ** < 0.0001 ** 0.0606 n.s.

Length 0.0001 ** < 0.0001 ** 0.6308 n.s.

Uniformity ratio 0.0027 ** < 0.0001 ** 0.2823 n.s.

Strength 0.4165 n.s. < 0.0001 ** 0.5637 n.s.

Elongation 0.0170 * < 0.0001 ** 0.5389 n.s.

2012 (Qk) Micronaire < 0.0001 ** < 0.0001 ** 0.0061**

Length 0.0078 ** < 0.0001 ** 0.1134 n.s.

Uniformity ratio < 0.0001 ** < 0.0001 ** 0.8798 n.s.

Strength 0.0080 ** < 0.0001 ** 0.5732 n.s.

Elongation 0.0031 ** < 0.0001 ** 0.6233 n.s.

*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05


40

From Table 3.4 we can also observe that micronaire was affected by episodic

irrigation interruption and cultivar selection in all years and locations. The interaction

of the effects of irrigation and cultivar was observed just for micronaire in 2011 and in

the 2012 Quaker location. Based on these interactions, we decided to detail micronaire

values per year and treatment in Table 3.5.

Table 3.5. Micronaire (no units) observed during the 2010, 2011, and 2012 seasons.

LSD values at the 0.05 level


and LSDs

Cultivar No

Drought

Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom to

termination

Average

-------------------------------- Micronaire ---------------------------------

2010 (ND) DP0912 5.25 - - 4.60 4.44 4.76

LSDirr = 0.24 DP0935 4.98 - - 5.16 4.68 4.94

LSDvar = 0.21 FM9170 4.80 - - 4.42 4.61 4.61

LSDint = 0.34 FM9180 4.96 - - 4.46 4.18 4.53

Average 5.00 - - 4.66 4.48 2011 (Qk) DP0912 4.73 5.21 4.85 4.22 3.72 4.54

LSDirr = 0.27 DP0935 4.66 4.76 4.83 3.43 3.47 4.20

LSDvar = 0.16 FM9170 4.45 4.28 3.76 3.32 3.38 3.82

LSDint = 0.38 FM9180 4.70 4.66 3.60 3.91 3.52 4.04

Average 4.62 4.72 4.29 3.72 3.51 2012 (Qk) DP0912 4.87 5.01 5.48 3.90 3.77 4.66

LSDirr = 0.2 DP0935 4.20 4.11 4.92 3.60 3.22 4.53

LSDvar = 0.17 FM9170 4.39 4.20 5.08 3.03 3.10 3.98

LSDint = 0.36 FM9180 4.20 4.55 5.08 3.05 3.01 3.97

Average 4.76 5.38 4.77 3.66 2.86 2012 (ND) DP0912 5.21 5.92 5.01 4.22 2.94 4.61

LSDirr = 0.26 DP0935 4.74 5.52 5.22 3.89 3.30 4.01

LSDvar = 0.14 FM9170 4.40 4.73 4.75 3.30 2.74 3.96

LSDint = 0.33 FM9180 4.70 5.34 4.09 3.26 2.47 3.98

Average 4.43 4.49 5.15 3.38 3.28




Regarding the effects of irrigation interruption, on micronaire, the highest

micronaire values were observed at no drought in 2010; squaring irrigation

interruption, and no irrigation interruption in 2011; early flowering episodic drought in


41

2012 at New Deal; and when the irrigation was interrupted at the squaring stage in

2012 at the Quaker farm.

A premium could potentially be obtained in micronaire ranging from 3.7 to

4.2. In 2010 this included the average of cultivar FM9180, when subjected to episodic

water stress from the maximum flowering (peak bloom) stage until crop termination

this micronaire was not statistically different from DP0912 using the same irrigation

treatment. The same was true for cultivars FM9180 and FM9170 when using the 3

weeks at peak bloom episodic stress regime.

Average micronaire at the premium range (3.7 to 4.2) was also observed in

2011 for DP0935 with irrigation interruption from peak bloom to termination,

FM9180 with irrigation interruption for three weeks at peak bloom, and FM9170 with

irrigation suspension at the early flowering stage.

An interesting cultivar interaction was observed in 2011 for micronaire at the

early flowering irrigation interruption. Cultivars DP0912 and DP0935, which were

water stressed at early flowering, provided micronaire as high as was observed in the

fully irrigated and squaring drought treatments. However, cultivars FM9170 and

FM9180, which were water stressed for three weeks after the first flowers were

observed, had micronaire values as low as was observed when cultivars received

irrigation interruption treatments at peak bloom, in the potential premium range.

Micronaire in the premium range was also observed in 2012 at the Quaker

location for DP0912 subjected to episodic drought events at the peak bloom stage, and


42

also at the New Deal Location for DP0935 with the irrigation interrupted from three

weeks at the peak bloom stage.

The length measurements provided by HVI differs from the length

measurements provided by AFIS. While HVI provides a measurement of a cotton fiber

bundle, AFIS length is based on individual fibers measurements. We found interesting

to detail the values of HVI UHML length to compare them with the AFIS length

related parameters, described on the next section. The effects of episodic drought in

the HVI length were also consistent in 2011, and in both 2012 environments.

Therefore, the effects of the treatments on HVI length measurements can be observed

in Table 3.6.


43

Table 3.6. HVI UHML lengths (inches) observed during the 2010, 2011, and 2012

seasons. LSD values at the 0.05 level


and LSDs

Cultivar No

Drought

Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom

to termination

Average

--------------------------- HVI UHML Length (in) --------------------------

2010 (ND) DP0912 1.117 - - 1.053 1.060 1.077

LSDirr = 0.027 DP0935 1.113 - - 1.053 1.090 1.086

LSDvar = 0.019 FM9170 1.180 - - 1.123 1.103 1.136

LSDint = 0.032 FM9180 1.113 - - 1.117 1.117 1.116

Average 1.131 - - 1.087 1.093

2011 (Qk) DP0912 1.083 1.053 0.975 1.055 1.033 1.037

LSDirr = 0.016 DP0935 1.060 1.107 0.995 1.060 1.063 1.054

LSDvar = 0.013 FM9170 1.118 1.147 1.017 1.123 1.115 1.106

LSDint = 0.030 FM9180 1.127 1.093 1.028 1.120 1.093 1.090

Average 1.096 1.099 1.003 1.089 1.079

2012 (ND) DP0912 1.083 1.088 1.015 1.070 1.080 1.067

LSDirr = 0.017 DP0935 1.103 1.107 1.073 1.090 1.087 1.092

LSDvar = 0.015 FM9170 1.198 1.190 1.113 1.163 1.163 1.165

LSDint = 0.028 FM9180 1.185 1.180 1.108 1.158 1.150 1.156

Average 1.145 1.143 1.077 1.122 1.122

2012 (Qk) DP0912 1.045 1.063 1.018 1.050 1.040 1.043

LSDirr = 0.019 DP0935 1.073 1.090 1.055 1.038 1.038 1.059

LSDvar = 0.015 FM9170 1.153 1.118 1.095 1.113 1.115 1.119

LSDint = 0.034 FM9180 1.148 1.085 1.030 1.103 1.098 1.093

Average 1.104 1.089 1.049 1.076 1.073




HVI UHML lengths higher than staple 33 (0.97 inches) considering base

values of color and leaf content, could not receive a discount, the higher premium

could be achieved on higher than staple 37 (1.09 inches). In all years and locations,

average HVI lengths of 1.02 inches or higher were observed in all treatments.

Table 3.6 also shows that, in all years and locations, treatments that were fully

irrigated throughout the season had consistently higher HVI length than the treatments

that received irrigation interruptions, with the exception of the treatments that had

irrigation interrupted in the squaring stage. In the three environments that had a


44

squaring episodic drought treatment, Quaker2011, New Deal2012, and Quaker2012,

HVI lengths from that treatment were not significantly different from those observed

from the no drought treatment.

Regarding cultivar effects on HVI length, in all environments of this study, the

Fibermax cultivars FM9170 and FM9180 had consistently higher length than the Delta

and Pine cultivars DP0912 and DP0935. The same cultivar differences were observed

at the strength parameter, average strength values of 29 g tex-1

or higher were

observed for FM9170 and FM9180. Strength values of 29 g tex-1

or higher could

possibly receive a premium.

In 2011, the average strength of treatments water stressed at the early

flowering stage, 26 g tex-1

, at a possible discount level, were significantly lower than

observed at the other irrigation timing treatments that presented strengths of 28 g tex-1

or higher, above the discount level. We believe that the higher than normal

temperatures observed in 2011 increased the effects of the early flowering episodic

drought in fiber strength.

At Quaker 2012 environment, average strength of the fully irrigated treatment,

31 g tex-1

, was significantly higher than observed at the other timing treatments, with

exception of the squaring episodic drought. The squaring irrigation interruption

strength, 30 g tex-1

, was not statistically different from presented at the no drought

treatment (LSDirr = 0.7 g tex-1

). In 2010, and in the New Deal 2012 environment, the

effects of the episodic treatments were not significant, in these environments, all

irrigation treatments presented strengths of 29 g tex-1

or higher.


45

Strength effects should be observed together with elongation, because, at the

same strength level, a higher energy would be necessary to break a fiber with higher

elongation. In opposite of what was observed for HVI length and strength, in all years

and locations, Delta and Pine cultivars DP0912 and DP0935 had significantly higher

elongation than Fibermax cultivars FM9170 and FM9180.

Irrigation affected elongation in both 2012 environments, New Deal 2012 and

Quaker 2012. In New Deal 2012 average elongation of the treatments that were full

irrigated (10.24 %) was significantly higher than the observed at the two treatments

water stressed at the peak bloom stage (LSDirr=0.19 %). In the Quaker 2012

environment, average elongation observed at full irrigation (9.86%) was significantly

higher than the observed at all other episodic drought treatments, with exception of the

squaring stage (9.67 %).

Our results suggests that the peak bloom stage is the most sensitive to episodic

drought effects on elongation, while squaring might be the less sensitive. In 2010 we

had higher than normal precipitation, and in 2011 higher than normal temperatures.

The elongation values at 2010 and 2011 were also lower than the observed in 2012,

ranging from 7.51 to 9.47 %. Therefore, we can hypothesize that environmental

effects reduced the overall elongation values in 2010 and 2011, and reduced as well

the response in elongation due to episodic water stress.


46

3.4.3. AFIS

In addition to the information provided by HVI, AFIS parameters support the

understanding of fiber quality with information related to individual fiber length

distribution, and also measurements of fineness and maturity ratio, which relate to

micronaire. Based on the AFIS data standard fineness can be calculated dividing

fineness by the maturity ratio. Standard fineness not only separate the maturity effects,

but also correlates well with the fiber diameter. A summary of the statistical analysis

of some of these parameters, in 2010 and 2011, is presented in table 3.7.

Curiously, in 2010 the only irrigation effect shown on Table 3.7 was on the

averange length on the top 5% fibers, while in 2011 episodic drought affected all

parameters presented. The cultivar effect was significant in all parameters and

environments, with exception of short fiber content in 2010. An irrigation-cultivar

interaction was observed in 2010 for effects in average length, upper quartile length,

and longest 5% fibers lenth; three length distribution related parameters. In 2011

interactions occurred for the neps per gram, short fiber content, fineness, immature

fiber content, and maturity ratio.


47

Table 3.7. Significance of the analysis of variance for AFIS variables as affected by

the irrigation episodic drought, cultivar or their interaction


2010 (ND) Neps per gram 0.2568 n.s. 0.0340 * 0.8674 n.s.

Length average 0.0873 n.s. 0.0031 ** 0.0473 *

Upper quartile length 0.0557 n.s. < 0.0001 ** 0.0121 *

Short fiber content (w) 0.2254 n.s. 0.3154 n.s. 0.2906 n.s.

Longest 5% fibers (n) 0.0338 * < 0.0001 ** 0.0025 **

Fineness 0.0560 n.s. < 0.0001 ** 0.5112 n.s.

Standard fineness 0.1071 n.s. < 0.0001 ** 0.3848 n.s.

Immature fiber content 0.0791 n.s. 0.0026 ** 0.3366 n.s.

Maturity ratio 0.0514 n.s. 0.0053 ** 0.5342 n.s.

2011 (Qk) Neps per gram 0.0013 ** 0.0003 ** 0.0071 **

Length average 0.0007 ** 0.0003 ** 0.0714 n.s.

Upper quartile length 0.0003 ** < 0.0001 ** 0.0731 n.s.

Short fiber content (w) 0.0033 ** 0.0015** 0.0277 *

Longest 5% fibers (n) 0.0006 ** < 0.0001 ** 0.1322 n.s.

Fineness 0.0004 ** < 0.0001 ** 0.0312 *

Standard fineness 0.0015 ** < 0.0001 ** 0.2016 n.s.

Immature fiber content < 0.0001 ** < 0.0001 ** 0.0011 **

Maturity ratio < 0.0001 ** 0.0001 ** 0.0009 **


Since maturity ratio is one of the parameters that can influence micronaire, a

parameter discussed in the previous section, we decided to detail the averages per year

and treatments on Table 3.8. Cotton fibers with maturity ratio in the range of 0.8 to 1.0

are considered mature; this range includes all treatments from the 2010 and 2011

season.


48

Table 3.8. Maturity ratios (%) observed in 2010 and 2011. LSD values at the 0.05

level


and LSDs

Cultivar No

Drought

Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom to

termination

Average

----------------------------- Maturity ratio(%) ----------------------------

2010 (ND) DP0912 0.94 - - 0.88 0.89 0.90

LSDirr = 0.02 DP0935 0.94 - - 0.93 0.92 0.93

LSDvar = 0.01 FM9170 0.94 - - 0.92 0.91 0.92

LSDint = 0.02 FM9180 0.92 - - 0.90 0.88 0.90

Average 0.94 - - 0.91 0.90

2011 (Qk) DP0912 0.94 0.94 0.90 0.88 0.85 0.90

LSDirr = 0.02 DP0935 0.92 0.88 0.89 0.85 0.85 0.88

LSDvar = 0.01 FM9170 0.92 0.92 0.86 0.85 0.86 0.88

LSDint = 0.02 FM9180 0.93 0.91 0.85 0.87 0.86 0.88

Average 0.93 0.91 0.88 0.86 0.85




In both 2010 and 2011, maturity ratios from fully irrigated treatments were

significantly higher than the maturity ratio from other timing treatments (Table 3.8);

this probably was the cause of the highest micronaire values observed for these

treatments (Table 3.5).

DP0935 had significantly higher maturity ratio than the other cultivars in 2010.

In 2011, DP0912 had significantly higher maturity ratio than the other cultivars,

however, in 2011 a cultivar-irrigation interaction for maturity ratio was observed. In

2011, cultivars maturity ratios were not significantly different at the full irrigation and

peak bloom to termination irrigation regimes. In the 2011 squaring irrigation

interruption regime, maturity ratio from cultivar DP0912 was statistically higher than

the observed on DP0935. In the 2011 early flowering episodic drought DP0912 and

DP0935 maturity ratio was significantly higher that the observed with FM9170 and

FM9180. Finally, in the 2011 3 weeks at peak bloom water stress regime, maturity


49

ratio from DP0912 and FM9180 was statistically higher than the observed at cultivars

DP0935 and FM9170.

Fineness is another parameter obtained with AFIS that can influence

Micronaire, so we decided to also to detail fineness averages per year and treatments

on Table 3.9.

Table 3.9. Fineness (mTex) observed in 2010 and 2011. LSD values at the 0.05 level


and LSDs

Cultivar No

Drought

Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom to

termination

Average

----------------------------- Fineness (mTex) -----------------------------

2010 (ND) DP0912 194 - - 178 178 183

LSDirr = 7 DP0935 193 - - 192 183 190

LSDvar = 5 FM9170 175 - - 170 170 172

LSDint = 9 FM9180 180 - - 166 164 170

Average 186 - - 177 174

2011 (Qk) DP0912 180 185 173 163 155 171

LSDirr = 7 DP0935 176 166 164 153 149 161

LSDvar = 4 FM9170 163 160 151 141 138 150

LSDint = 8 FM9180 171 165 146 151 148 155

Average 172 169 159 152 147




Although in 2010 there was not a significant effect of irrigation timing in

fineness, in 2011 fineness of fully irrigated treatments and treatments that had

irrigation interrupted at the squaring stage was significantly higher than the observed

in the other episodic drought treatments. This would suggest a correlation of high

fineness with the high micronaire observed in the full irrigation and squaring during

the 2011 season (Table 3.5).

Ranging from 135 to 175 mTex cotton fibers are classified as fine. In 2010,

cultivars FM9180, and FM 9170 were classified in this range while DP0912 and


50

DP0935 had fineness significantly higher. DP0912 and DP0935 fibers could be

classified as average regarding fineness in 2010. In 2011, the fineness values of each

cultivar were statistically different, with the lowest fineness observed at FM9170;

however all cultivars could be classified as fine.

In 2011 there was also a cultivar-irrigation interaction for fineness effects,

although fineness of DP0912 was statistically higher than fineness from all other

cultivars in most irrigation regimes, at the full irrigation and early flowering irrigation

interruption regime, DP0912 fineness was not considered statistically different from

DP0935 fineness.

Finally, for the purpose of comparing the AFIS average individual fiber length

with the UHML length obtained with HVI, we decided to also detail AFIS fiber length

averages per year and treatments on Table 3.10.

Table 3.10. AFIS individual fiber length averaged by weight (inches) observed in

2010 and 2011. LSD values at the 0.05 level


and LSDs

Cultivar No

Drought

Squaring Early

Flowering

Peak Bloom

(3 weeks)

Peak Bloom to

termination

Average

-------------------------------- Length (in) ---------------------------------

2010 (ND) DP0912 1.003 - - 0.940 0.933 0.959

LSDirr = 0.034 DP0935 1.007 - - 0.943 0.977 0.976

LSDvar = 0.024 FM9170 1.057 - - 0.997 0.973 1.009

LSDint = 0.039 FM9180 0.993 - - 1.007 1.010 1.003

Average 1.015 - - 0.972 0.973

2011 (Qk) DP0912 0.950 0.957 0.865 0.910 0.913 0.915

LSDirr = 0.029 DP0935 0.925 0.927 0.898 0.898 0.915 0.912

LSDvar = 0.023 FM9170 0.988 1.000 0.840 0.963 0.980 0.958

LSDint = 0.044 FM9180 0.997 0.963 0.885 0.973 0.950 0.951

Average 0.964 0.962 0.874 0.941 0.934





51

Although the 2010 irrigation effect on AFIS length was not significant

(ANOVA F-test, Table 3.7), in 2011 lengths from fiber from cotton exposed to water

stress at the early flowering stage were significantly lower than those observed from

the other irrigation timing treatments.

Similar to what was observed for HVI UHML length (Table 3.6), AFIS

average individual fiber length of the Fibermax cultivars FM9170 and FM9180 had

consistently higher length than the Delta and Pine cultivars DP0912 and DP0935

lengths in 2010 and in 2011.

3.5. Discussion

The 2010 season had higher than normal precipitation and cooler temperatures,

whereas the other environments had lower than normal precipitation. Therefore, the

results observed in 2010 should be expected to occur just when a high precipitation

weather scenario occurs.

When water availability is reduced, the cotton plant closes its stomata,

reducing photosynthesis. This results in decreased carbon fixation, plant growth, lower

yields, and changes in fiber quality. Fiber quality parameters and yield showed

different responses to episodic drought periods and cultivar selection.

The effects of episodic drought on yield were similar to the previously

described for some of the irrigation reduced throughout the whole season by Nunes

Filho et al. (1998), Dagdelen et al. (2009), and Mills (2010).


52

Although we there was no early flowering water stress in 2010, based on 2011

and 2012 seasons’ data, we can assume that early flowering water stress can result in

lower yields than the ones obtained in the other episodic drought treatments.

Although a cultivar effect on seed-cotton yield was not observed in 2012, an

irrigation-cultivar interaction was observed in all years and locations from this study,

suggesting that the effects of episodic irrigation interruption should be observed

together with the cultivar selection seed-cotton yields.

With exception of micronaire, no other HVI parameter presented was

significantly affected by irrigation, or had a significant irrigation-cultivar interaction in

all environments. This suggests that the irrigation interruption cultivar selection

response on fiber quality is in general also a function of soil type and climatic

conditions.

Also, in each environment a different cultivar/irrigation combination provided

better micronaire values. Premium range micronaire was observed in 2010 FM9180,

when subjected to episodic water stress from the peak bloom to termination, in 2011

for DP0935 with irrigation interruption from peak bloom to termination, FM9180

with irrigation interruption for three weeks at peak bloom, and FM9170 with irrigation

suspension at the early flowering stage, in 2012 Quaker DP0912 subjected to episodic

drought events at the peak bloom stage, and in 2012 New Deal DP0935 with the

irrigation interrupted from three weeks at the peak bloom stage. Therefore, it is very

complex to manage irrigation and cultivar selection focusing on better micronaire.


53

Another difficulty for managing irrigation and cultivar selection for micronaire

is that, as indicated by the the AFIS analysis, micronaire is affected by both maturity

and fineness, and these can be affected differently for episodic water stress and

cultivar selection in diverse environments. Furthermore, in some environments, such

as the Quaker 2011 and Quaker 2012, irrigation-cultivar combinations that could

result in micronaire at the premium level would also cause a significant reduction on

yields. Mills (2010) also reported for varying irrigation levels micronaire at the

premium level to occur at a lower continuous irrigation level (around 5 mm per day),

which was related to reduced yield in the same study.

Since the micronaire HVI parameter used to classify cotton for commercial

purpose is affected by both Maturity and Fineness, AFIS analysis can provide a better

understanding of how environment, cultivar, and water stress affect each parameter.

In most of our environments, episodic drought affected fiber uniformity; this

was also similar to what was described for the effects of varying continuous irrigation

levels from Nunes Filho et al. (1998).

3.6. Conclusions

Yield and most of the HVI and AFIS parameters presented were significantly

affected by drought episode or cultivar. The data shows that in all years and locations

the irrigation strategy with no irrigation interruption provided the best yields.

In 2011 and in both 2012 locations, the irrigation interruption at the squaring

stage resulted in the smallest reduction in yield from the fully irrigated treatment. Also


54

in 2011 and in both 2012 locations, when the irrigation interruption occurred during

the early flowering treatment, the highest yield reduction was observed.

In 2010, under full irrigation, the varieties FM9170, DP0935 and DP0912

achieved the highest production compared to the other treatments. Under both 2010

stress periods, FM9180 and DP0935 showed the highest yield, indicating that these

cultivars might have some mechanism of resistance to late-occurring episodes of water

stress. In 2011, the cultivars DP0935 and DP0912 provided the highest seed cotton

yield among all cultivars used, across all irrigation suspension periods.

With exception of uniformity ratio, all other HVI variables presented were

affected by cultivar selection in all four environments. Uniformity ratio was not

affected by cultivar only in 2010. In all years and locations, average HVI UHML

lengths of 1.02 inches or higher were observed in all treatments.

Also in 2010, micronaire values statistically similar to the premium range were

observed on FM9180 and DP0912 when all were subjected to water stress until crop

termination. The same was true for cultivars FM9170 and FM9180, subjected to 3

weeks of drought. FM9170 and FM9180 were classified as fine regardless of the

irrigation treatment used. In 2010 and 2011 cotton fibers in all cultivars and water

regimes were classified as mature.

The results presented can be important for developing more efficient water

management strategies for cotton in West Texas.


55

References

AbdelGadir, A.H., Dougherty M., Fulton J.P., Curtis L.M., Tyson T.W., (2012) Effect

of Different Deficit-Irrigation Capabilities on Cotton Yield in the Tennessee

Valley. Irrigat Drainage Sys Eng 1:102. doi:10.4172/2168-9768.1000102

Balkcom, K. S.; D.W. Reeves; J.N. Shaw; C.H. Burmester and, L M. Curtis. 2006.

Cotton Yield and Fiber Quality from Irrigated Tillage Systems in the

Tennessee Valley. Agron. J. 98:596-602.

Bauer, P., Faircloth D.W., Rowland, D., Ritchie G.L. (2012) Water-sensitivity of

Cotton Groth Stages, in: C. Perry and Barnes E. (Eds.), Cotton Irrigation

Management of Humid Regions, Cotton Incorporated. p 17-20.

Bednarz, C.W.; Hook, J.; Yager, R.; Cromer, S.; Cook, D.; Griner, I. 2003. Crop

Water Use and Irrigation Scheduling. P. 61-64. In Culpepper, A.S. et al. (ed.)

2002 Cotton Research-Extension Report, UGA/CPES Research-Extension

Publication No. 4, April 2003.




Bronson, K. F.; A.B. Onken; J.W. Keeling; and, H. A. Torbert. 2001. Nitrogen

Response in Cotton as Affected by Tillage System and Irrigation Level. Soil

Sci. Soc. J. 65:1153-1163.

Campbell, T. B. and P.J. Bauer. 2007. Genetic Variation for Yield and Fiber Quality

Response to Supplemental Irrigation within the Pee Dee Upland Cotton

Germplasm Collection. Crop Sci. Soc. J. 47:591-599.

Collins, G. and K. Hake (2012) Management Considerations for Irrigated Cotton , in:

C. Perry and Barnes E. (Eds.), Cotton Irrigation Management of Humid

Regions, Cotton Incorporated. p 38-59.

Dagdelen, N., Basal, H.; E. Yilmaz; T. Gurbuz; and S. Akcay. 2009. Different drip


western Turkey. Agricultural Water Management Journal. 96: 111-120.






56


DOI: 10.1061/(ASCE)0733-9437(2004)130:4(277).

Krieg, D.R. 2000. Cotton water relations. P. 7-15. In Derrick Oosterhuis (ed) Special

Report 198 Proceedings of the 2000 Cotton Research Meeting and Summaries

of Cotton Research in Progress. University of Arkansas Division of

Agriculture, Fayetteville, Arkansas. (Available online

http://arkansasagnews.uark.edu/198ms3.pdf)

Littell, R.C., G.A. Milliken, W.W. Stroup, and R.D. Wolfinger. 1996. SAS system for

mixed models. SAS Inst., Cary, NC.





Inc. pp. 279-335.








2:295-299.

Pettigrew, W.T., and Dowd M.K. (2012) Interactions Between Irrigation Regimes and

Varieties Result in Altered Cottonseed Composition. The Journal of Cotton

Science 16:42-52.

SAS Institute. 2010. The SAS system for Windows. Release 9.3. SAS Inst., Cary, NC.



Report:31-46.








57





J. 100:1763-1768.


58

CHAPTER IV

AGRONOMIC WATER USE EFFICIENCY DIVERSE COTTON CULTIVARS

SUBJECTED TO VARYING IRRIGATION LEVELS

4.1. Abstract

The objective of this research was to investigate the response in yield and fiber quality

to varying irrigation levels of cotton cultivars not studied in previous research in West

Texas. The field experiment was conducted at the Texas Tech University research

farms from 2010 to 2012. The irrigation treatments were deficit irrigation, full

irrigation, and dryland. The cultivars used in all seasons were ACALA 1517-E2,

FM832, SIOKRA24, and ST506. A full continuous water supply was necessary to

achieve the highest seed cotton yield in most environments. An interaction of

irrigation-cultivar effects on seed-cotton yield was also observed in 2012. AWUE was

affected by the irrigation treatments in two years, in 2010 the highest AWUE occurred

under deficit irrigation, not statistically different from full irrigation, and in 2012 the

AWUE observed on fully irrigated cultivars was statistically higher than AWUE

observed under deficit irrigation and dryland. Fiber quality parameters were affected

by irrigation management and cultivar selection. This results can support the

development of more efficient water management strategies for cotton in West Texas.


59

4.2. Introduction

Previous research has emphasized the decrease in available water for West

Texas irrigated cotton, due to the decreasing of the Ogallala Aquifer (Howell et al.

2004; Torell et al., 1990). Water availability already impacts cotton production in

West Texas, and this region is responsible to nearly 40% of US cotton production.

Irrigation has a direct effect on yield (Bednarz et al., 2003; Dagdelen et al.,

2009, AbdelGadir et al., 2012), and also interacts with crop management practices,

and cultivar selection (Bronson et al., 2001; Balkcom et al., 2006; Campbell and

Bauer, 2007; Pettigrew and Dowd, 2012).

Cultivar selection is highly dependent on climate environment. For example,

Fritschi at al. (2004) stated that Pima cotton is more sensitive to delayed planting, has

a more pronounced indeterminate growth habit than Acala cotton, which limits Pima

production to regions with long growing seasons (Fritschi et al., 2004). Similarly,

Stiller et al. (2004), comparing irrigated and dryland environments in Australia,

reported a significant irrigation-cultivar interaction for yield and fiber quality with two

okra leaf cultivars yielding relatively more under dryland conditions. In a Stiller et al.

(2004) study, dryland cotton yielded 48% of irrigated cotton, and fiber lengths were

4% shorter. Pettigrew (2004) also tested okra-leaf genotypes in dryland and irrigated

conditions, reporting that genotypes physiology responded similarly in both soil

moisture regimes.


60

Based on this previous research evidence (Stiller et al., 2004; Pettigrew, 2004),

we hypothesized that this different kind of leaf might differ in photosynthesis,

transpiration and in consequence provide a different water use efficiency than

common leaves.

It was reported that, in Turkey, irrigation can improve cotton water use

efficiency (Basal et al., 2009). This was also been observed in West Texas (Mills,

2010), but the researches concentrated on genetically similar cultivars. Furthermore,

repeating irrigation research for a longer period of time is desirable, considering the

strong environmental effects on irrigation needs.

Therefore, the objective of this research is to investigate the response in yield

and fiber quality to varying irrigation levels of cotton cultivars, specifically focusing

on cultivars not studied in previous research in West Texas.


The field experiment was conducted at the Texas Tech University New Deal

research farm during 2010 (New Deal 2010), and 2011 (New Deal 2011) and at the

Texas Tech Quaker Avenue research farm in 2012 (Quaker 2012). These experimental

sites are located in the Texas High Plains which is considered a semi-arid area.

In both locations, the soil has high clay contents and high water holding

capacity. The soil at the Quaker Avenue farm is an Amarillo-Acuff sandy clay loam

(Fine-loamy, mixed, superactive, thermic Aridic Paleustolls), and the soil at the New


61

Deal farm is a Pullman-Olton clay loam (Fine, mixed, superactive, thermic Aridic and

Torrertic Paleustolls).

The planting dates were 27 May 2010, 18 May 2011, and 16May 2012.

Agronomical practices followed Texas A&M AgriLife Extension recommendations

for the Texas High Plains. Fertilizer in the form of 28-0-0-5 was applied at a rate of 90

kg N ha-1

. Weed control used mechanical hoeing. A conventional tillage system was

used. Plant growth regulators (PGR) were applied in 2010. Harvest dates were 17

November 2010, 3 October 2011, and 12 October 2012.


plot and cultivars as the split-plot. The experimental units were conducted in two

rows, measuring 10.7 m in 2010 and 12.2 m in 2011and 2012. Row spacing was one

meter. Subsurface drip irrigation (SSI) system was used. The irrigation tape was

placed on 0.25 m below the surface under every row.

The irrigation treatments were composed of deficit irrigation (nearly 40% of

ET replacement), a full irrigated management and dry land after establishment (after

the first squares were observed). From a set of 21 diverse cultivars and experimental

lines observed in 2010, we selected cultivars used in all seasons; they were ACALA

1517-E2, FM832, SIOKRA24, and ST506. The cultivars FM832 and SIOKRA24

have a different leaf form known as “Okra”. Based on previous research evidence

(Stiller et al., 2004; Pettigrew, 2004), we hypothesized that this different kind of leaf

might differ in photosynthesis and transpiration, and in consequence provide a

different WUE than other cultivars with common leaves.


62

Weather was monitored with an on-site automated weather station located in

the same field and included precipitation (Fig. 4.1) and other parameters used to

calculate evapotranspiration (ET) and irrigation needs. The total water applied to each

irrigation treatment was calculated by individual hydrometers attached to each

irrigation inlet. Soil volumetric water content was monitored with weekly

measurements using a neutron probe device, 503 DR Hydroprobe (CPN International,

Inc., Concord, CA).


cells, weights were used to calculate the yields. Agronomic Water Use Efficiency

(AWUE) was calculated dividing seed-cotton yield per the total water received per

treatment (precipitation plus irrigation). During harvest grab samples were collected,

the samples were ginned for the fiber quality analysis. Cotton samples were analyzed

using the high volume instrument (HVI) and advanced fiber information system

(AFIS) at the Fiber and Biopolymer Research Institute, in Lubbock, Texas.

The statistical analysis was done using the GLIMMIX Procedure in SAS®

software (SAS Inst., 2010) with an analysis of variance (ANOVA), followed by a

mean separation at 5% level of probability using the LSMEANS statement. The

GLIMMIX model performs estimation and statistical inference for generalized linear

mixed models by incorporating normally distributed random effects, allowing

GLIMMIX procedure to properly separate random and fixed effects. SAS

programming statements followed the recommendations provided by Littell et al.

(1996).


63

4.4. Results

The cumulative precipitations from all seasons of this study, compared with a

30 years average, are presented on Fig. 4.1.

Figure 4.1. Accumulated Precipitation (mm) observed during the 2010 and 2011

seasons at the Texas Tech New Deal research farm (ND), and in 2012 at the Texas

Tech Quaker Avenue research farm (Qk).

The total amount of precipitation and irrigation of each treatment in each year

are shown in Table 4.1. There was a higher than normal precipitation during the 2010

season totaling 325 mm. However, because the rainfall was concentrated in the

beginning of the season, differences in seed-cotton yield due to irrigation levels were

still observed (Table 4.2).

0

50

100

150

200

250

300

0 10 20 30 40 50 60 70 80 90 100 110 120

Acc

um

ula

ted

pre

cip

ita

tio

n (

mm

)

Days After Planting

2010 ND

2011 ND

2012 Qk

Lubbock 1981-

2010 average


64

Table 4.1. Total precipitation and irrigation applied per season and location

Season -

Location

Irrigation Treatment Precipitation

(mm)

Irrigation

(mm)

Total

Water

(mm)

2010 - ND Full irrigation 325 212 510

Deficit irrigation 325 118 446

No irrigation from squaring 325 27 353




2012 - Qk Full irrigation 46 583 629


No irrigation from squaring 46 173 219 ND = Texas Tech New Deal Research Farm.


4.4.1. Yields

As shown in the Table 4.2, we observed a highly significant effect on seed

cotton yield for the irrigation in all years. However, the single main effect of cultivar

was not observed during this study. In 2012 a irrigation-cultivar interaction was

observed.

Table 4.2. Statistical significance (p-values) of the study factors and interaction in the

variable seed-cotton yield during the 2010, 2011, and 2012 seasons

Factor p-value in 2010 p-value in 2011 p-value in 2012

Irrigation 0.0040 * 0.0009 ** < 0.0001 **

Cultivars 0.28 n.s. 0.11 n.s. 0.13 n.s.

Interaction 0.51 n.s. 0.68 n.s. 0.04 *



65

Table 4.3. Seed-cotton yield (kg ha-1

) during the 2010, 2011, and 2012 seasons. LSD

values at the 0.05 level.

Irrigation Level Season Cultivar Fully

Irrigated

Deficit

Irrigated

Dryland from

squaring

Average

2010 ACALA 1517-E2 3716 4232 1505 3151

LSDirr = 757 FM832 4932 4501 1745 3726

LSDvar = 599 SIOKRA24 4522 3264 1335 3040

LSDint = 1021 ST506 3744 3829 1773 3116

Average 4229 3957 1590

2011 ACALA 1517-E2 3890 2425 1651 2656

LSDirr = 406 FM832 4464 3317 1999 3260

LSDvar = 363 SIOKRA24 4425 2836 1891 2929

LSDint = 582 ST506 3774 3030 1984 3051

Average 4139 2902 1881

2012 ACALA 1517-E2 4540 968 465 1991

LSDirr = 147 FM832 4607 1284 663 2185

LSDvar = 128 SIOKRA24 4719 1010 440 2149

LSDint = 221 ST506 4365 1438 645 2056

Average 4558 1175 553

LSDirr = Least significant difference for irrigation treatments



In the three years, the lowest seed-cotton yields were observed when the

irrigation was interrupted after the crop establishment (dryland from squaring). In the

rainy 2010 season seed-cotton yield from fully irrigated crops were not statistically

higher than the observed under deficit irrigation. However, in 2011 and 2012 full

irrigation provided statistically higher seed-cotton yields than deficit irrigation, on

average more than 3300 kg ha-1

higher in 2012.

In 2012 the highest seed-cotton yield differences from full irrigation to dryland

after treatment establishment was observed, an average difference of more than 4000

kg ha-1

.


66

In 2012, the highest average yield was observed by cultivar SIOKRA24 fully

irrigated, and it was statistically higher than ST506 in the same irrigation regime.

Curiously, under deficit irrigation the highest yields were provided by ST506 and it

was statistically higher than yields from SIOKRA24 and ACALA 1517-E2 in the

same irrigation regime. Under dryland conditions the yields of all cultivars were not

found to be different based on a t-test at the 0.05 level.

Although the yields provide an interesting analysis, observing how efficient

cultivars are at using the precipitation and irrigation to achieve the described yields is

desirable, especially in regions with limited water supply, such as West Texas.

4.4.2. Agronomic Water Use Efficiency

The effects of the treatments on agronomic water use efficiency (AWUE), are

described in tables 4.4 and 4.5. As shown in Table 4.4 the response of AWUE to the

treatments varied; in 2011, no significant effects were observed, but in 2012 AWUE

was affect by irrigation, cultivar, and irrigation-cultivar interaction. In 2010, only

irrigation resulted in significant differences in AWUE.

Table 4.4. Statistical significance (p-values) of the study factors and interaction in the

variable AWUE during the 2010, 2011, and 2012 seasons


Irrigation 0.0075 ** 0.0733 n.s. < 0.0001 **

Cultivars 0.2195 n.s. 0.1490 n.s. 0.0001 **

Interaction 0.4710 n.s. 0.8101 n.s. 0.0038 **



67

Table 4.5. Agronomic water use efficiency (kg ha-1

mm-1

) during the 2010, 2011, and

2012 seasons. LSD values at the 0.05 level. 1 kg ha-1

mm-1

= 1 g daL-1

Irrigation Level Season Cultivar Fully

Irrigated

Deficit

Irrigated

Dryland from

squaring

Average

2010 ACALA 1517-

E2 7.29 9.49 4.26 7.01

LSDirr = 1.50 FM832 9.67 10.09 4.94 8.23

LSDvar = 1.22 SIOKRA24 8.87 7.32 3.78 6.66

LSDint = 2.06 ST506 7.34 8.59 5.02 6.98

Average 8.29 8.87 4.50 2011 ACALA 1517-

E2 5.75 4.74 4.19 4.98

LSDirr = 0.85 FM832 6.59 6.48 5.07 6.05

LSDvar = 0.74 SIOKRA24 6.54 5.54 4.80 5.63

LSDint = 1.44 ST506 5.58 5.92 5.04 5.51

Average 6.11 5.67 4.83 2012 ACALA 1517-

E2 7.22 3.13 2.12 4.16

LSDirr = 0.35 FM832 7.32 4.16 3.03 4.84

LSDvar = 0.28 SIOKRA24 7.50 3.27 2.01 4.26

LSDint = 0.48 ST506 6.94 4.65 2.95 4.85

Average 7.25 3.80 2.53

LSDirr = Least significant difference for irrigation treatments



In 2010, the hightest AWUE was observed for cultivars under deficit

irrigation, they were in average 4.37 kg ha-1

mm-1

higher than the dryland AWUE,

however, deficit irrigation AWUE was not statistically different from full irrigation

AWUE. However, in 2012, AWUE from the diverse irrigation treatments followed

the partern observed for seed-cotton yield, with higher values observed at full

irrigation statistically different from deficit irrigation and dryland

Also in 2012, AWUE from FM832 and ST506 was statistically higher than

AWUE from ACALA 1517-E2 and SIOKRA24. However, the cultivar difference in

AWUE was observed just when they were subject to deficit irrigation or dryland


68

conditions, under full irrigation AWUE from all cultivars were not statistically

different.

4.4.3. Fiber Quality

A summary of some of the HVI fiber quality differences that were statisticaly

significant is presented in Table 4.6. These results are important because some HVI

parameters are used in cotton classification for the marketing of the crop.

As shown in table 4.6, the response of the reported parameters to irrigation

management and cultivar selection was different over the years. In 2010 just cultivar

effects on uniformity, strength and elongation were observed. However in 2011 and

2012 all parameters, except elongation, were affected by irrigation. Irrigation-cultivar

interaction was observed just for micronaire and strength in 2011 and elongation in

2012.

The only effect that was consistent among the years was the difference in

strength among cultivars. However, in 2010 and 2011 the highest strength of all

cultivars was observed for the ACALA 1517-E2 cultivar (33.8 g tex-1

in 2010 and

32.1 g tex-1

in 2011). It was statistically higher than FM832 and ST506 in 2010 and

statistically different from all cultivars in 2011, in 2012 the highest strenght was

observed in FM832 (30.8 g tex-1

), but it was only significantly different from ST506

(28.3 g tex-1

), with a LSD of 1.3 mTex.


69

Table 4.6. Significance of the analysis of variance for HVI variables as affected by the

irrigation episodic drought, cultivar or their interaction

Season Factor Irrigation Cultivar Interaction

2010 Micronaire 0.6417 n.s. 0.1161 n.s. 0.9687 n.s.

Length 0.9981 n.s. 0.1018 n.s. 0.3484 n.s.

Uniformity ratio 0.8957 n.s. 0.0402 * 0.7353 n.s.

Strength 0.8954 n.s. < 0.0001 ** 0.4840 n.s.

Elongation 0.9848 n.s. 0.0030 ** 0.7603 n.s.

2011 Micronaire 0.0076 ** 0.0021 ** 0.0228 *

Length 0.0035 ** 0.0025 ** 0.2831 n.s.

Uniformity ratio 0.0218 * 0.1029 n.s. 0.5094 n.s.

Strength 0.0135 * < 0.0001 ** 0.0285 *

Elongation 0.1258 n.s. 0.4702 n.s. 0.7215 n.s.

2012 Micronaire 0.0049 ** 0.3191 n.s. 0.1252 n.s.

Length 0.0004 ** 0.1366 n.s. 0.6130 n.s.

Uniformity ratio 0.0006 ** 0.1008 n.s. 0.4890 n.s.

Strength 0.0024 ** 0.0412 * 0.4074 n.s.

Elongation 0.2741 n.s. 0.0238 * 0.0324 *


In 2011, average micronaire from the dryland treatments (4.06) was

significantly lower than in the other irrigation regimes. 2011 full and deficit irrigation

regimes provided average micronaire above the potential premium range (3.7 to 4.2).

However, in 2012 the only irrigation treatment in the premium range for micronaire

was the full irrigation (3.81) significantly lower than the other treatments. In 2012, the

only cultivar that provided average micronaire values at the premium range was


70

ACALA 1517-E2, significantly lower than the ones observed on ST506 and FM832

cultivars.

In 2011 and 2012, HVI length observed at full irrigation (1.11 and 1.18 inches)

was significantly higher than observed under deficit irrigation and dryland. In both

years length observed under deficit irrigation was also statistically higher than the

observed in the dryland treatment. In 2011, HVI length from ACALA 1517-E2 was

significantly higher than the ones provided by other cultivars.

In 2012 fully irrigated fiber average uniformity ratio (83.9), at the potential

premium level (82 or higher), was significantly higher than the observed under deficit

irrigation and dryland. However, these results may be influenced by our ginning

procedures that did not include an industrial gin, so they have to be analyzed with

caution.

The Advanced Fiber Information System analysis (AFIS) expands upon the

information provided by HVI parameters. For example, micronaire HVI parameter

used to classify cotton for commercial purpose is affected by both maturity and

fineness, both of which are measured by AFIS. AFIS also can provide individual fiber

length information, among other useful fiber quality information.


71

Table 4.7. Significance of the analysis of variance for AFIS variables as affected by

the irrigation level, cultivar or their interaction


2011 (ND) Neps per gram 0.3135 n.s 0.3362 n.s. 0.2641 n.s.

Length average 0.0086 ** 0.5304 n.s. 0.5955 n.s.

Upper Quartile Length 0.0054 ** 0.2855 n.s. 0.6597 n.s.

Short Fiber Content (w) 0.0656 n.s. 0.1293 n.s. 0.3745 n.s.

Longest 5% Fibers (n) 0.0072 ** 0.1338 n.s. 0.5934 n.s.

Fineness 0.0376 * < 0.0001 ** 0.3577 n.s.

Standard Fineness 0.0570 n.s. < 0.0001 ** 0.5009 n.s.

Immature Fiber Content 0.0614 n.s. 0.0248 * 0.7579 n.s.

Maturity Ratio 0.0612 n.s. 0.0342 * 0.3850 n.s.


Similar to HVI bundle length, in 2011, irrigation level affected average fiber

length, uperquartile length, short fiber content, longest 5% fibers length, and fineness,

parameters provided by AFIS (Table 4.7). AFIS data also inform that in 2011, cultivar

selection had a significant effect on fineness, standard fineness, immature fiber

content, and maturity ratio. There was not observed an irrigation-cultivar interaction

for any of the AFIS parameters presented in 2011.

In 2011, similar to 2011 HVI bundle length, AFIS average individual length

observed under full irrigation (0.99 inches) was significantly superior than the

observed under deficit irrigation and dryland. However, AFIS length observed under

deficit irrigation (0.95 inches) was not statistically different from the observed in the

dryland treatment (0.93 inches). Similarly, upper quartile length full irrigation (1.19

inches) was significantly superior than the observed under deficit irrigation and

dryland.


72

Ranging from 135 to 175 mTex cotton fibers are classified as fine, average

fineness on this range was observed in all irrigation regimes, although dryland

fineness (158 mTex) was considered significantly lower than the ones observed in the

irrigated treatments. All cultivars could be classifyied as fine, ACALA 1517-E2

provided significantly lower fineness (156 mTex) when compared to the other

cultivars.

Based on the AFIS data standard fineness can be calculated dividing fineness

by the maturity ratio. Standard fineness not only separate the maturity effects, but also

correlates well with the fiber diameter. ACALA 1517-E2 standard fineness was not

statistically different from the one observed for FM832, but it was significantly lower

than the one provided by ST506 and SIOKRA24.

Cotton fibers with maturity ratio in the range of 0.8 to 1.0 are considered

mature; this range includes all irrigation treatments and cultivars in the 2011 season.

4.5. Discussion

When water availability is reduced, the cotton plant closes its stomata,

reducing photosynthesis, and as a result, less carbon is fixed. This directly decreases

plant growth, and in consequence, the mass of the yield components at the end of the

season. Photosynthesis can also affect fiber quality parameters such as length and

maturity. Therefore, reduction in yield due to the continued application of smaller

amounts of water supports the information presented by Mills (2010).


73

The differences on the AWUE effects throughout the years of this study could

be due the climatic differences. Higher than average temperatures were observed in

2011, these extremely high temperatures might have reduced AWUE in all treatments,

covering the effects of irrigation level and cultivar choice in that year.

Stiller et al. (2004) defined Agronomic WUE as the lint yield divided by

cumulative ET summed from planting to the mean maturity date, and reported that

agronomic water use efficiency (WUE) varied among cultivars, with a full-season

okra leaf cultivar, SIOKRA L23, having the highest WUE (2.87 kg lint ha-1

mm-1

evaporation) and the highest yield. Although more related to the physiological

response to water use, the WUE calculation procedure used by Stiller et al. 2004 not

consider the irrigation system effects on efficiency and also ignores possible increases

on the seed yield that provides a secondary revenue to producers.

Also differently from Stiller et al. (2004), in 2010 and 2011 we didn’t observed

differences on average cultivar AWUE. However, in 2012 the AWUE from FM832

and ST506 was statistically higher than AWUE from ACALA 1517-E2 and

SIOKRA24 when they were subjected to reduced-moisture enviroments.

The effects of irrigation levels and cultivar selection on the reported HVI fiber

quality parameters were not consistent through the years. Even for the single main

effect of cultivar on strength, reported in the three years, cultivar ranking was diverse.

A better description of irrigation and cultivar effects on fiber quality was provided by

AFIS analysis in 2011.


74

4.6. Conclusion

A reduction in yield proportional to the irrigation depth applied was observed

on most years. We can conclude that a full continuous water supply is necessary to

achieve the highest seed cotton yield in most environments. An interaction of

irrigation-cultivar effects on seed-cotton yield was also observed in 2012.

AWUE was affected by the irrigation treatments in two years, in 2010 the

highest AWUE occurred under deficit irrigation, not statistically different from full

irrigation, and in 2012 the AWUE observed on fully irrigated cultivars was

statistically higher than AWUE observed under deficit irrigation and dryland. The

okra leaf cultivars FM832 and SIOKRA24 did not consitently provided increased

AWUE compared to the other cultivars of this study.

At least in one year, all HVI fiber quality parameters were affected by

irrigation management or cultivar selection. Similar to HVI bundle lenght, in 2011,

irrigation level affected the AFIS parameters, average individual fiber length,

uperquartile length, short fiber content, and the longest 5% fibers length. In a few

environments ACALA 1517-E2 provided some improved fiber quality properties.

This results can support the development of more efficient water management

strategies for cotton in West Texas.


75

References







Basal, H., Dagdelen N., Unay A., Yilmaz E. (2009) Effects if Deficit Drip Irrigation

Raios on Cotton (Gossypium hirsutum L.) Yield and Fibre Quality. J.

Agronomy & Crop Science 195:19-29. DOI:10.1111/j.1439-

037X.2008.00340.x







Sci. Soc. J. 65:1153-1163.







Fritschi, F. B.; B.A. Roberts; D.W. Rains; R.L. Travis and, R.B. Hutmacher. 2004.

Fate of Cof Nitrogen-15 Applied to Irrigated Acala and Pima Cotton. Agron. J.

96:646-655.




DOI: 10.1061/(ASCE)0733-9437(2004)130:4(277).





76

Pettigrew, W.T. (2004) Physiological Consequences of Moisture Deficit Stress in

Cotton. Crop Sci.. 44:1265-1272.



Science 16:42-52.

Stiller, W. N.; P.E. Reid and; G.A. Constable. 2004. Maturity and Leaf Shape as Traits

Influencing Cotton Cultivar Adaptation to Dryland Conditions. Agron. J.

96:656-664.




77

CHAPTER V

COTTON CULTIVARS FRUIT DISTRIBUTION UNDER EPISODIC

IRRIGATION INTERRUPTION

5.1. Abstract

The objective of this study was to describe cotton cultivars fruit distribution as

affected by episodic drought in different cotton development stages. Field experiments

were conducted in 2011 and 2012 at the Texas Tech Quaker Avenue Research farm.

The episodic drought irrigation treatments included a full irrigation throughout the

season; non-irrigation from squaring to flowering; 3 weeks of non-irrigation beginning


irrigation beginning from peak bloom to the crop termination. The cultivars planted

for this study consisted of DP0912, DP0935, FM9170, and FM9180. Irrigation

interruption periods resulted in differences on the average number of bolls on nodes 8

to 13, these differences had consequences on seed-cotton yield. The effect of cultivar

selection on boll distribution was not as remarkable as the one observed in the

different irrigation interruption treatments. An episodic early flowering irrigation

interruption can reduce the number of bolls on nodes 8 to 13, reducing seed-cotton

yields. Studying fruit distribution provides insights on how episodic drought and

cultivar selection affects yield and fiber quality.


78

5.2. Introduction

Fruit distribution may potentially impact cotton yield and fiber quality.

Bednarz et al. (2006) hypothesized that reducing the percentage of the total yield

produced at inner fruiting positions through reduced plant densities increased the

source-to-sink ratio during boll filling in this region of the canopy, resulting in

improved fiber properties. Bednarz et al. (2006) continued suggesting that

modifications in crop management may increase the source-to-sink ratio during boll

filling of the remaining fruiting positions (i.e. outer fruiting positions), possibly

resulting in greater improvements in fiber quality.

Cotton fruit distribution is directly related to crop maturity. For example, in

Bednarz and Nichols (2005) research, early maturing cultivars produced a greater

percentage of their total lint yield at lower main stem nodes. It was also suggested that

full season cotton cultivars are better adapted to areas with decreased precipitation,

like in the southeastern U.S. In these areas, early maturing cultivars may not recover

from multiple episodic drought events (Bednarz and Nichols, 2005).

Mills et al. (2008) studied fruit distribution in Bollgard/Roundup Ready

(BG/RR) and Bollgard II/Roundup Ready Flex (BGII/RRF) cottons, and found that

BGII/RRF maturity was unaffected by the late glyphosate applications and produced a

higher percentage of plants having a harvestable boll in the lower canopy than BG/RR.

Mills et al. (2008) also concluded that, for the cultivars in his study, the BGII/RRF


79

cotton had increased boll number and weight in the first sympodial position at lower

main stem nodes while BG/RR produced more and heavier bolls on upper main stem

nodes.

Ritchie et al. (2009) found that the irrigation system type can affect cotton

bolls distribution. The irrigation method used had a significant impact on boll

distribution in the plants, with the overhead irrigation treatment consistently having

less cotton near the bottom of the plant and more cotton near the top than plants

irrigated with either of the subsurface drip irrigation (SSI) methods evaluated. Ritchie

at al. (2009) concluded that SSI irrigation decreases early-season fruit loss, resulting in

heavier carbohydrate sinks and decreased overall growth and upper boll filling on the

crop.

Another factor that can affect cotton fruit distribution is water stress. A crop

can be water stressed with a continuous reduced supply of water throughout the

season, for example with deficit irrigation, or can have the water supply completely

interrupted in a given stage (episodic drought).

Basal et al. (2009) reported that in Turkey, full irrigation and mild deficit

irrigation allowed plants to set more sympodial and additional monopodial branches

and bolls on the upper plant nodes than stronger deficit irrigation treatments, resulting

in a significant difference in boll numbers. Thus, in Basal et al. (2009) study, the

distribution of bolls on plants was consistently and significantly affected by irrigation

deficit. However, Basal et al. (2009) study concentrated on only cultivars Sahin-2000

and STN-8A. Additionally, Basal et al. (2009) didn’t examined episodic water deficit


80

effects; instead, the study focused on continuous irrigation deficits throughout the

season.

In West Texas, Mills (2010) stated that irrigation had a positive effect on the

number of bolls per plant, concluding that irrigation levels affect yield on a field level,

plant level, and within-boll yield components. However, Mills (2010) did not

measured episodic drought as a factor.

Although previous research suggested that irrigation timing can affect cotton

yield (Kriegg, 2000; Bauer et al., 2012; Collins and Hake, 2012), an aspect of

irrigation management that has not been studied in detail is the effect of episodic

drought on the fruit distribution of the new commercial cotton cultivars.

Therefore, the objective of this chapter is to study differences in commercial

cotton cultivars fruit distribution as affected by episodic drought occurred in several

plant development stages.


Field experiments were conducted in 2011 and 2012 at the Texas Tech Quaker

Avenue research farm (Quaker) located in Lubbock-TX. The soil at the Quaker

Avenue farm is an Amarillo-Acuff sandy clay loam (Fine-loamy, mixed, superactive,

thermic Aridic Paleustolls).

Planting dates were 13 May 2011, and 16 May 2012. Harvest dates were 11

October 2011, and 14 October 2012.


81








plot and cultivars as the split-plot. There were four blocks. The experimental units

were conducted in two rows measuring 12.2 m. Subsurface drip irrigation (SSI)

system was used. The tape is placed 0.25 m below the surface under every row.

The following irrigation treatments were used: a full irrigation throughout the



irrigation from peak bloom to the crop termination. The episodic drought irrigation

treatments included a full irrigation throughout the season; non-irrigation from

squaring to flowering; 3 weeks of non-irrigation beginning at early flowering; 3 weeks

of non-irrigation beginning at peak bloom; and non-irrigation beginning from peak

bloom to the crop termination. We considered the squaring stage begin to occur when

the first square (floral bud) was observed, and the early flowering stage to begin when

the first opened flower was observed. The cultivars of this study consisted of DP0912,

DP0935, FM9170, and FM9180.

Weather was monitored with an automated weather station located close to the

study. The station monitored precipitation (Fig. 5.1) and other parameters used to


82

calculate irrigation needs. The total water applied to each irrigation treatment was

calculated by individual hydrometers attached to each irrigation inlet. Soil moisture

content was monitored weekly for all irrigation treatments using a neutron probe

device, 503 DR Hydroprobe (CPN International, Inc., Concord, CA).

At the end of the season, the number of bolls (fruits) per node and their

position were counted in five plants per plot, number of plants considered adequate

once the procedures discussed by Ritchie et al. (2011) were considered. For sampling

plants were chosen randomly in the plot, however gaps were avoided since it can

impact the plant growth habit and impact boll distribution.

All plots were harvested and weighed with a cotton stripper equipped with

load cells, weights were used to calculate the yields. During harvest grab samples were

collected, the samples were ginned for the fiber quality analysis. Cotton samples were

analyzed using the high volume instrument (HVI) and advanced fiber information

system (AFIS) at the Fiber and Biopolymer Research Institute, in Lubbock, Texas.

A smoothing procedure consisting of a weighted average (40% weight on node

of interest, 20% on adjacent nodes, and 10% on nodes adjacent to those), as discussed

by Ritchie et al. (2011), was also used before analyzing boll distribution data. The

statistical analysis used the GLIMMIX Procedure in SAS® software (SAS Inst., 2010)

with an ANOVA followed by a mean separation at a 5% level of probability using the

LSMEANS statement. The GLIMMIX model performs estimation and statistical

inference for generalized linear mixed models by incorporating normally distributed

random effects, allowing GLIMMIX procedure to properly separate random and fixed


83

effects. SAS programming statements followed the recommendations provided by

Littell et al. (1996). Similar to Mills et al. (2008), we considered fruiting node a

random effect, and concentrated our analysis by node and year.

5.4. Results

Fig. 5.1 shows differences in boll distribution of cotton subjected to different

irrigation interruption periods in 2011. The differences in the average number of node

per boll due to irrigation interruption period were statistically significant at the 5%

level from nodes 7 to 14. At these nodes, the lowest boll retention occurred in the

cotton that had been stressed at early flowering. Despite the small stature of the cotton

that had irrigation interruption at the squaring stage, a reduction in boll retention was

observed on nodes 7 and 8, coupled with significantly higher boll retention on nodes

10 and 11 when compared with the other water stressed treatments. The retention at

the other nodes was not significantly different from the other treatments.


84

Figure 5.1. Smoothed boll distribution averaged over four cultivars subjected to five

different irrigation interruptions periods during the 2011 (A) and 2012 (B) seasons.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Aver

age

nu

mb

er o

f b

oll

s no drought

squaring

early flowering

3 weeks at

peak bloompeak bloom to

termination

A

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

5 6 7 8 9 10 11 12 13 14 15 16 17 18

Aver

age

nu

mb

er o

f b

oll

s

Node

B


85

In 2012 the effect of irrigation on boll number was statistically significant for

all nodes up to 16. Boll retention for individual irrigation treatments changed from

2011 to 2012 as well.

The effect of cultivar selection on boll distribution, averaged among the

different water regimes, is shown in Fig. 5.2. Cultivar was not as much of a factor as

the timing of irrigation interruption in determining boll distribution.

In 2011 the boll number of FM9180 was statistically lower than the other

cultivars at nodes 8 to 13. In 2012 cultivar effect on boll number was statistically

significant for nodes 5, 6, and 9 to 14.


86

Figure 5.2. Smoothed Bolls distribution of four cultivars subjected averaged over five

different episodic drought periods during the 2011 (A) and 2012 (B) seasons.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

5 6 7 8 9 10 11 12 13 14 15 16 17 18

Aver

age

nu

mb

er o

f b

oll

s DP0912

DP0935

FM9170

FM9180

A

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

5 6 7 8 9 10 11 12 13 14 15 16 17 18

Aver

age

nu

mb

er o

f b

oll

s

Node

B


87

5.5. Discussion

In addition to determining the total number of bolls and yield per plant, the

number of bolls per node gives insight into the overall health and maturity

characteristics of the crop. The reduced number of bolls observed on all nodes when

irrigation was interrupted during the early flowering stage (Fig. 5.1) corresponded

with the low cotton yields in Table 3.3.

However, the weight of individual bolls is also a necessary aspect of seed-

cotton yield. As shown on Table 3.3, although there were fewer bolls per node

throughout the plant for FM9180 (Fig. 5.2), the cultivar was observed to have larger

bolls than many of the other cultivars in the study, and did not have significantly lower

seed-cotton yields.

Furthermore, bolls from lower nodes (i.e. 5 to 8) have more time for their

fibers to be filled with cellulose than bolls located on higher nodes; therefore, they

tend to have a higher maturity ratio. Maturity ratio affects micronaire, which is used to

classify cotton for commercial purposes.

It might be expected that a reduced number of top bolls presented by the early

flowering irrigation interruption treatment observed on Fig. 5.1 to provide better fiber

quality, but it is not true. As shown in Table 3.4, early flowering episodic drought did

not improve micronaire values. Also on table 3.6 we can observe that the early

flowering water stress reduced fiber length. This indicates that the stress on the

developing bolls compromised fiber quality throughout the whole plant bolls.


88

5.6. Conclusion

Irrigation interruption periods resulted in differences in the average number of

bolls in several nodes and, consequently, impacted seed-cotton yield and fiber quality

as described on Chapter 3.

In both years FM9180 had slightly fewer bolls than the other cultivars from

nodes 8 to 15 and FM9170 retained more bolls at nodes 9 and 10 than the other

cultivars. However, the effect of cultivar selection on boll distribution was not as

remarkable as the observed in the different irrigation interruption treatments.

An episodic early flowering irrigation interruption can reduce the number of

bolls on nodes 8 to 13, reducing seed-cotton yields.

Studying fruit distribution provides insights on how episodic drought and

cultivar selection affects yield and fiber quality.


89

References




037X.2008.00340.x




Bednarz, C.W., Nichols R.L. (2005) Phenological and Morphological Components of

Cotton Crop Maturity. Crop Science 45:1497–1503

DOI:10.2135/cropsci2004.0321



10.2135/cropsci2005.08-0276.














Mills, C.I., Bednarz C.W., Ritchie, G.L., Whitaker, J.R. (2008) Yield, Quality, and

Fruit Distribution in Bollgard/Roundup Ready and Bollgard II/Roundup Ready

Flex Cottons. Agron.J. 100:35-41. DOI:10.2134/agronj2006.0299


90




Ritchie, G.L., Whitaker J.R., and Collins G.D. 2011. Effect of Sample Size on Cotton

Plant Mapping Analysis and Results. J. Cotton Sci. 15:224-232



91

CHAPTER VI

COTTON PHYSIOLOGICAL PARAMETERS AFFECTED BY EPISODIC

IRRIGATION INTERRUPTION

6.1. Abstract

Improving cotton irrigation management practices in West Texas is important for

increasing farmers’ profits and also for increasing the Ogallala life span. The objective

of this work was to evaluate the effects of in field controlled drought conditions on

cotton gas exchange. From 2010 to 2012 cotton cultivar FM9180 gas exchange was

measured throughout the season using a LiCor-6400 portable photosynthesis system.

In 2011 and 2012 measurements were also made on DP0935 cultivar. From the several

parameters evaluated some relationships were presented. Episodic drought periods can

affect leaf-level gas exchange and impact yield. Photosynthesis and yield were

particularly sensitive to water deficit at early flowering. Despite an increase in leaf

water use efficiency under water deficit, overall growth and yield were inhibited in all

treatments with a stress component. Understanding the relative sensitivity at different

growth stages can help with irrigation decisions when water resources are limited.

6.2. Introduction

Cotton is the main crop in West Texas, with 3 to 4 million acres planted every

year. In 2010, for example, Texas was responsible for nearly 40% of the United States


92

production. Almost half of the cotton area is irrigated, and irrigated cotton accounts

for most of the cotton production (USDA, 2011).

Over time, water has been drawn out of the Ogallala Aquifer at a rate that

exceeds its ability to recharge (Howell et al., 2004; Torell et al., 1990). Improving

irrigation management practices is important for increasing farmers’ profits and also

for increasing the Ogallala sustainability.

It has been proven that irrigation has a direct effect on yield (Bednarz et al.,

2003; Dagdelen et al., 2009, AbdelGadir et al., 2012). Irrigation effects on production

also interacts with crop management practices, and cultivar selection (Bronson et al.,

2001; Balkcom et al., 2006; Campbell and Bauer, 2007; Pettigrew and Dowd, 2012).

Based on previous plant-water relation basic knowledge (Kramer and Boyer, 1995;

Kirkham, 2005) it is fair to hypothesize that the impacts of irrigation on the production

is preceded by changes on plant level gas exchange parameters.

Baker et al. (2009) described a chamber system that can also be used to

calculate canopy transpiration (E) and photosynthetic net assimilation (A) in a whole

plant level for many field applications. However, gas exchange measurements at leaf

level are more commonly reported (Baker et al., 2007; Mills, 2010; da Costa and

Cothren, 2011).

A physiologically water-efficient crop will have increased carbon assimilation

per unit of transpiration, and these processes can be monitored with gas exchange

measurements. Cotton gas exchange response to drought has been reported by da


93

Costa and Cothren (2011), although their research was done under greenhouse

conditions, and only one drought level was reported.

Likewise, Mills (2010) observed a positive relationship between transpiration

and photosynthesis under varying irrigation levels. However, his treatments were

applied throughout the season, and he did not test episodic drought events caused by

complete interruption of irrigation within individual growth stages.

Reduction in transpiration can directly impact the leaf temperature (TL).

However, it was found that the TL minus Ambient Temperature (TA) differential (TL-

TA) together with the Vapor Pressure Deficit (VPD) provide better predictors of the

degree of water stress than TL alone (Baker et al., 2007).

Although previous research suggested that irrigation timing can affect cotton

yield (Kriegg, 2000; Bauer et al., 2012; Collins and Hake, 2012), an aspect of

irrigation management that has not been studied in detail is the effect of water stress

over a specific period of time in the crop development stage (episodic drought) on

cotton gas exchange.

We hypothesize that cotton photosynthesis, transpiration, and related

physiological parameters are affected differently by water deficit at different growth

stages. This will explain differences in yield based on the timing of stress. No

comprehensive research has been conducted to determine how cotton plants gas

exchange parameters, such photosynthesis and transpiration, respond to episodic

drought periods at different stages.


94

Therefore, the objective of this work was to evaluate the effects of controlled

drought in field conditions on gas exchange. Knowledge of episodic drought effects

can help cotton producers improve water management strategies.


Field experiments were conducted in 2010 at the Texas Tech New Deal

Research farm (New Deal), in 2011 and 2012 at the Texas Tech Quaker Avenue

Research farm (Quaker) located in Lubbock-TX. The soil at the Quaker Avenue farm

is an Amarillo-Acuff sandy clay loam (Fine-loamy, mixed, superactive, thermic Aridic

Paleustolls), and the soil at the New Deal farm is a Pullman-Olton clay loam (Fine,

mixed, superactive, thermic Aridic and Torrertic Paleustolls).

The planting dates were 25 May 2010, 13 May 2011, and 16 May 2012.

Agronomical practices followed Texas A&M AgriLife Extension recommendations

for the Texas High Plains. Fertilizer in the form of 28-0-0-5 was applied at a rate of 90

kg N ha-1

. Weed control included herbicide applications (glyphosate), and mechanical

hoeing. A conventional tillage system was used. Plant growth regulators (PGR) were

applied in 2010 only. Harvest dates were 20 November 2010, 11 October 2011, and 14

October 2012.

The experimental design was considered a Randomized Complete Block

Design (RCBD) in 2010 and split-plot in 2011 and 2012. Irrigation treatments were

the main plots and cultivars were the split-plots. Three blocks were used in 2010 and

four in 2011 and 2012. The experimental units were two rows, measuring 10.7 m in


95

2010 and 12.2 m in 2011 and 2012. Irrigation was applied using a subsurface drip

irrigation (SSI) system placed 0.25 m below the surface under every row.

During the 2010 season, heavy rainfall occurred from 11 to 51 days after

planting (DAP) (Fig. 6.1), resulting in no early-season stress and limiting the study to

three irrigation treatments: full irrigation throughout the season; 3 weeks of episodic

drought beginning at peak bloom (86 to 105 DAP); and irrigation interruption from

peak bloom stage until the crop termination.

Figure 6.1. Accumulated Precipitation (mm) observed during the 2010 season in the

Texas Tech New Deal research farm and in 2011 and 2012 in the Texas Tech Quaker

Avenue research farm.

In 2011 and 2012, the 2010 irrigation treatments were repeated (beginning at

89 DAP in 2011, and 83 DAP in 2012), and 3-week episodic drought events were

added at squaring (46 to 67 DAP in 2011, 41 to 54 DAP in 2012) and early flowering

(67 to 89 DAP in 2011, and 54 to 83 DAP in 2012). The squaring treatment was begun

0

50

100

150

200

250

300

1 11 21 31 41 51 61 71 81 91 101 111 121

Acc

um

ula

ted

pre

cip

ita

tio

n (

mm

)

Days After Planting

2010

2011

2012

Lubbock

1981-2010

average


96

when the first squares (floral buds) were observed on 50% of the plants in each

treatment, and the early flowering treatment was begun when opened flowers were

observed on 50% of the plants.

Gas exchange was measured throughout the season, using a LI-6400 XT gas

exchange system (LI-COR Inc., Lincoln, NE). The uppermost fully expanded leaf was

measured on three plants randomly selected per plot. In 2010, measurements were

conducted on FiberMax 9180B2F (FM9180) in each irrigation treatment. In 2011 and

2012, measurements were conducted on both FM9180 and DeltaPine 0935B2RF

(DP0935) cultivars.

Although plant sampling occurred at random, at the plant selection for

measures process we avoided plants located near big gaps, plants located near the

borders, and unhealthy plants and leaves.

Net photosynthesis (A), transpiration (E), leaf temperatures, and related

parameters rates were measured using the LI-6400. Leaf physiological water use

efficiency was calculated by dividing each measurement of A by to the corresponding

E. LI-6400 individual leaf data was also used to compute the differences in air and leaf

temperature at the measurement moment.

Weather was measured with an automatic station located in the same field and

included precipitation (Fig. 6.1) and other parameters used to calculate irrigation

needs. The total water applied to each irrigation treatment was calculated by individual

hydrometers attached to each irrigation inlet. Soil moisture content was monitored


97

with weekly measurements using a neutron probe device, 503 DR Hydroprobe (CPN

International, Inc., Concord, CA).


cells. Weights were used to calculate the yields. Adding precipitation (Fig. 6.1) to the

irrigation depths obtained using hydrometers measurements provided the total water

used by plot. The agronomic water use efficiency (AWUE) was obtained by the

division of the yield by the total water received in each plot.


software (SAS Inst., 2010) with an ANOVA followed by a mean separation at 5%

level of probability using the LSMEANS statement. The GLIMMIX model performs

estimation and statistical inference for generalized linear mixed models by




6.4. Results and Discussion

In all three years, we observed statistical significance for the episodic drought

and their interaction on photosynthesis or net assimilation (A) at p=0.05. Fig. 6.2

shows photosynthesis changes during the 2010 season. From 90 to 118 DAP a clear

difference for the irrigation treatments was observed, the single effect of irrigation

interruption was statistically significant (p<0.001). At 118 DAP each treatment was

different from the others at p=0.05. The same pattern was observed by transpiration.


98

Figure 6.2. Photosynthesis measured on the FM9180 cultivar subjected to three

different irrigation regimes during the 2010 season at the Texas Tech New Deal

research farm.

As shown in Fig. 6.3, in 2011 and 2012 water deficit resulted in decreased

photosynthesis, followed by a slow recovery after water application was continued.

0

5

10

15

20

25

30

75 85 95 105 115 125

Ph

oto

syn

thes

is (

µm

ol

CO

2 m

-2 s

-1)

Days after planting

peak bloom

to

termination

no drought

3 weeks at

peak bloom

drought

periods

begin

3 weeks

drought ends

Drought period


99

Figure 6.3. Photosynthesis averaged on the FM9180 and DP0935 cultivars submitted

to different irrigation interruptions periods during the 2011 (a) and 2012 (b) seasons at

the Texas Tech Quaker Avenue research farm.

0

5

10

15

20

25

30

35

40

45 65 85 105

Ph

oto

syn

thes

is (

µm

ol

CO

2 m

-2 s

-1) squaring early flowering peak bloom

0

5

10

15

20

25

30

35

40

41 51 61 71 81 91 101 111

Ph

oto

syn

thes

is (

µm

ol

CO

2 m

-2 s

-1)

Days after planting

no drought

squaring

early

flowering

3 weeks at

peak bloom

peak bloom to

termination

squaring early flowering peak bloom (b)

(a)


100

In 2011, when the episodic stress occurred at the squaring treatment (49 DAP)

we failed to reject the null hypothesis and concluded that there was not a treatment

effect on photosynthesis (p=0.604 for irrigation, p=0.72 for cultivar, and p=0.54 for

the interaction). The same occurred in 2012 squaring irrigation interruption (42 and 51

DAP).

Also in 2011, when the irrigation was interrupted during the early flowering

stage (68 to 82 DAP) there was a significant difference from the stressed plants

photosynthesis to the other treatments (p=0.0006 to p<0.0001) without a cultivar or

irrigation-cultivar interaction effect. Similar results were observed in 2012 early

flowering episodic drought, but just from DAP 65 to 82. Curiously at 2012 75 DAP

just photosynthesis from treatments that had irrigation interrupted at the squaring stage

were not statistically different from the treatments that were receiving an early

flowering stressed at that time at p=0.05.

From the 2011 and 2012 irrigation interruption at the peak bloom irrigation

effect was always statistically significant (p<0.0001). A cultivar effect was observed

at 2011 105 DAP when average photosynthesis from FM9180 was nearly 2 µmol CO2

m-2

s-1

higher. Neither cultivar nor an irrigation-cultivar interaction effect was

observed at 2012.

In 2011, at 105 DAP we observed a combined effect of the episodic drought

periods on the cultivars FM9180 and DP0935 on photosynthesis (Fig. 6.4).


101

Figure 6.4. Photosynthesis measured in cultivars DP0935 and FM9180 submitted to

five different irrigation interruptions periods during the 2011 season 105 days after

planting at the Texas Tech Quaker Avenue research farm. Black bars represent one

Least Significant Difference (LSD). LSD (p=0.05) = 2.49 µmol CO2 m-2

s-1

.

At 2011 DAP 105 (Fig. 6.4) there was both a single main effect for irrigation

interruption (p<0.0001) and cultivar (p=0.0088) on photosynthesis. Photosynthesis

from the treatments that received irrigation interruption at the squaring and early

flowering stages were as high as the observed on the fully irrigated treatment and

statistically higher than the treatments were receiving peak bloom water stress at that

time. The average difference in photosynthesis from the squaring irrigation

interruption treatment to the treatment that were receiving episodic drought at the peak

bloom stage was nearly 27 µmol CO2 m-2

s-1

.

Although the cultivar effect on photosynthesis at 2011 105 DAP (Fig. 6.4) was

not observed on any other irrigation regime, for the plants that received an irrigation

interruption until 105 DAP, cultivar FM9180 photosynthesis was more than 5 µmol

CO2 m-2

s-1

higher than DP0935 photosynthesis, statistically different at p=0.05.

0

5

10

15

20

25

30

35

40

no drought squaring early

flowering

3 weeks at

peak bloom

peak bloom to

termination

Ph

oto

syn

thes

is (

µm

ol

CO

2 m

-2 s

-1)

Irrigation interruption period

DP0935

FM9180


102

A linear relation between transpiration (E) and net carbon assimilation or

photosynthesis (A) was observed in 2011 (r2=0.806). In 2012 we also observed

photosynthesis increasing with transpiration (Fig. 6.5a) however, in that year the best

fit was obtained using a Michaelis-Menten equation. The linear relation observed in

2011 closely resembled the first part of the Michaelis-Menten equation, indicating that

this model, that is more suitable to this type of biological process, could possibly be

also adjusted to the 2011 data if we had a broader set of data in that year.


103

Transpiration (mmol H2O m-2

s-1

)

0 5 10 15 20 25

Ph

oto

syn

the

sis

(m

ol C

O2 m

-2 s

-1)

0

5

10

15

20

25

30

35

40

3 week peak bloom

Early flowering

Fully irrigated

Peak bloom to

terminationSquaring

PS = (51.3*Trans)/(11.7+Trans) A

Stomatal Conductance (gs) (mol H2O m-2

s-1

)

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Ph

oto

syn

the

sis

(m

ol C

O2 m

-2 s

-1)

0

5

10

15

20

25

30

35

40

3 week peak bloom

Early flowering

Fully irrigated

Peak bloom to

terminationSquaring

BPS = (40.3*gs)/(0.224+gs)

Figure 6.5. The effects of transpiration (a) and stomatal conductance (b) on

photosynthesis of cultivars DP0935 and FM9180 submitted to five different

irrigation interruptions periods during the 2012 season.


104

Since in both years we observed an increase in photosynthesis with

transpiration, it is fair to assume that stomata aperture is the common cause for the

correlation, a better effect of stomata effects (stomatal conductance – gs) on net

photosynthesis is shown in Fig. 6.5b. We can also observe photosynthesis increasing

with the reduction in stomatal resistance (increase in gs) that may be attributed to

stomatal opening. The model that better explained the photosinthesys gs relation was

the Michaelis-Menten.

We didn’t find models with good fit relating photosynthesis with leaf water use

efficiency (LWUE), nor with intercellular carbon dioxide (Ci), vapor pressure deficit

(VPD), leaf temperature, and nor with air temperature. Since when photosynthesis

occur, stomata are opened causing plant intercellular water to evaporate

(transpiration), this was partially explained by Fig. 6.5. The heat used to evaporate

liquid intercellular water reduces the leaf temperature; this explains the relation of air-

leaf temperature and photosynthesis (Fig. 6.6).


105

Figure 6.6. Relation of the air-leaf temperature difference on photosynthesis of

cultivars DP0935 and FM9180 submitted to five different irrigation interruptions

periods during the 2012 season.

At Fig. 6.6 a second degree polynomial model was used to relate

photosynthesis and air-leaf temperature difference (r2 = 0.693).

y = -1.1354x2 + 8.0586x + 18.949

-10

0

10

20

30

40

50

-4 -2 0 2 4 6

Photo

synth

esis

(µ

mol C

O2 m

-2 s

-1)

Air-leaf temperature difference (ºC)

r2= 0.693


106

Figure 6.7. Effect of the intercellular-ambient carbon dioxide concentration ratio on

photosynthesis of cultivars DP0935 and FM9180 submitted to five different

irrigation interruptions periods during the 2012 season.

When stomata are open, increased CO2 flux increases carbon assimilation

internal carbon dioxide (Ci). Therefore the Ci/Ca ratio approaches 1 under conditions

that promote stomatal opening. This explains the relationship between photosynthesis

and the intercellular-ambient carbon concentration ratio (Ci/Ca) (Fig. 6.7). In Fig. 6.7

we can observe a fairly good linear model fit (r2=0.61).

Taking all of these physiological measurements, described above, into

consideration, another parameter that can be interesting not only as an indicator of

stress but also for further physiological inferences, is the rate carbon dioxide

assimilated at photosynthesis for a unit of water lost by transpiration, or the leaf

physiological water use efficiency (LWUE). In Fig. 6.8 we can observe how LWUE

was affected by the several episodes of irrigation interruption.

y = 52.012x - 6.9422

0

5

10

15

20

25

30

35

40

45

0 0.2 0.4 0.6 0.8 1

Photo

synth

esis

(µ

mol C

O2 m

-2 s

-1)

Intercellular-ambient CO2 concentration ratio (Ci/Ca)

r2 = 0.61


107

Figure 6.8. Physiological leaf water use efficiency averaged on the FM9180 and

DP0935 cultivars submitted to different irrigation interruptions periods during the

2012 seasons at the Texas Tech Quaker Avenue research farm.

Similar to what occurred with photosynthesis (Fig. 6.3b), in 2012 the effects of

irrigation interruption on LWUE became statistically significant just at 51 DAP and

were significant until 91 DAP. At 58 and 65 DAP, higher LWUE of nearly 2.3 µmol

CO2 mmol H2O-1

was observed for the plants that received irrigation interruption at

squaring. At 72, 75, and 82 DAP, the treatments that received water stress at early

flowering together with the plants that had irrigation interrupted at squaring presented

a LWUE statistically higher than the other treatments (at p=0.05). However, at 91

DAP early flowering episodic stress presented LWUE statistically higher than the one

observed for the squaring treatment (nearly 0.58 µmol CO2 mmol H2O-1

more). At 91

0

0.5

1

1.5

2

2.5

3

3.5

4

41 51 61 71 81 91 101 111

Lea

f w

ate

r u

se e

ffic

ien

cy (

µm

ol

CO

2 m

mo

l H

2O

-1)

Days after planting

no drought

squaring

early flowering

3 weeks at

peak bloom

peak bloom to

terminationsquaring early flowering peak bloom


108

DAP early flowering LWUE was not statistically different from the ones observed for

the treatments who were receiving a peak bloom irrigation interruption (for p=0.05).

In 2012, a cultivar effect on LWUE was observed just on 51 DAP (p=0.0374)

when the average for FM9180 cultivar was higher than the presented by DP0935. At

51 DAP an irrigation-cultivar interaction was observed (p=0.0262), LWUE for fully

irrigated cultivars were not statistically different at p=0.05 but for the cultivars that

were receiving a squaring irrigation interruption FM9180 presented higher LWUE

than DP0935.

The effects of episodic drought on seed-cotton yield and agronomic water use

efficiency (AWUE) can be observed in Fig. 6.9. There was a significant effect of

irrigation on yield and AWUE (p<0.0001). Cultivar alone did not have a significant

effect on yield or AWUE; however, there was an irrigation-cultivar interaction for

both yield (p=0.0046) and AWUE (p=0.0041). Although the yields for FM9180 were

327 kg ha-1

higher than DP0935 in the 3-week peak bloom stress treatment, there were

no significant differences among cultivars at the other irrigation interruption

treatments.

The differences on yield (Fig. 6.9a) can be explained by the variations on

photosynthesis throughout the season caused by the irrigation interruption episodes

(Fig. 6.3b). When photosynthesis is reduced, the total carbon assimilated on the plant

biomass is reduced and fewer compounds can be partitioned into the yield

components. Therefore, yields from plants that did not receive any drought had yields


109

more than 2,100 kg ha-1

higher than the ones that had a severe photosynthesis

reduction from 68 to 82 DAP (Fig 6.3b).

Figure 6.9. Seed-cotton yield (a) and Agronomic water use efficiency of cotton

cultivars DP0935 and FM9180 subjected to five different irrigation interruptions

periods during the 2012 season. Black bars represent one Least Significant Difference.

LSD Yield (p=0.05) = 183 kg ha-1

. LSD AWUE (p=0.05) = 0.37 kg ha-1

mm-1

. 1 kg

ha-1

mm-1

= 1 g daL-1

.

0

500

1000

1500

2000

2500

3000

3500

4000

4500


flowering

3 weeks at

peak bloom

peak bloom

to termination

See

d-c

ott

on

yie

ld (

kg

ha

-1)

DP0935

FM9180

(a)

0

1

2

3

4

5

6

7


flowering

3 weeks at

peak bloom

peak bloom to

termination

Ag

ron

om

ic W

UE

(k

g h

a-1

mm

-1)


DP0935

FM9180

(b)


110

The increase of photosynthesis with transpiration described by Mills (2010)

was also observed with our data. Although in 2011 a linear relation was observed, in

2012 a Michaelis-Menten model, similar to the second degree polynomial described

by Mills (2010) was shown.

It is a fair assumption to assume that this good direct relation is based on the

fact that both processes are regulated by stomata aperture, which will vary in between

drought episodes, but keep a proportional effect on photosynthesis and transpiration.

More details of these relations were described on Fig. 6.5 including the effects of

stomatal conductance (gs). This also agrees to the basic concepts of plant-water

relations (Kramer and Boyer, 1995; Kirkham, 2005).

Although not consistent throughout all the measurements, some differences on

cultivar response in gas exchange for episodic drought such the ones presented on Fig.

6.4 were observed in the last two years of this study.

Similar to the findings of Baker at al. (2007), ambient-leaf temperature

difference is a good indicator of water stress in our study. Modern infrared

thermometers and automatic weather stations have made measuring air and leaf

temperature simple, and we suspect that they can be used together with the equation

from Fig. 6.6 to estimate leaf photosynthesis in a broad number of situations.

Although following Fick’s law of diffusion differences in gradient can affect a

flux, such as the water vapor flux from the cell to the ambient during transpiration, we


111

were unable to find a good relation of vapor pressure deficit (VPD) that is and

transpiration. Since VPD is a measurement of the water vapor gradient it should

directly affect the evaporation of the water in the leaf level (transpiration). Based on

Baker et al. (2007) findings we were also expecting to find relations of VPD and

photosynthesis and transpiration. Therefore we can conclude that gradient is

important, but so is conductance: if you have leaves with different conductances, it

will influence your transpiration even if the vapor pressure deficit is the same.

Similar to the greenhouse research by da Costa and Cothren (2011), we

observed that episodes of drought decreased gas exchange impacting yield. When

water availability is reduced, the cotton plant closes its stomata, reducing

photosynthesis, and as a result, less carbon is fixed and less water is lost by

transpiration. This can decrease in-season plant growth and decrease yield at the end

of the season.

However, a comparison from the differences from photosynthesis and

transpiration measured in leaf scale such our instrument, and canopy level

measurements using a chamber system such as the one described by Baker et al.

(2009) is suggested. Although leaf level measurements indicated that drought affects

photosynthesis, and relations among physiological parameters, they cannot be directly

extrapolated to quantify gas exchange at a canopy level.

While leaf physiological water use efficiency (LWUE) varies throughout the

season and is a function of irrigation interruption regimes (Fig. 6.8), agronomic seed-

cotton water use efficiency (AWUE) is more a function of the crop seed-cotton yield


112

(Fig. 6.9). AWUE from fully irrigated crops were more than 1.8 kg ha-1

mm-1

higher

than the observed when irrigation was interrupted at the early flowering stage.

Further research relating the reported leaf level measurements throughout

episodes of irrigation interruption, crop level gas exchange, and the partitioning of

photo-assimilates are necessary to elucidate why the LWUE values were higher for

plants stressed at the flowering and peak bloom stages, but resulted in lower AWUE.

6.5. Conclusion

Understanding the relative sensitivity at different growth stages can help with

irrigation decisions with limited water resources.

Photosynthesis and other physiological related parameters were affected by

irrigation interruption episodes in the three years of this study. Yields were

particularly sensitive to the photosynthesis reduction caused by water deficit at early

flowering. Episodic drought periods can affect leaf-level gas exchange and impact

yield.

We also found good correlations of photosynthesis with transpiration, stomatal

conductance, air-leaf temperature difference, and internal-ambient carbon dioxide

concentration ratio. This can help to elucidate part of the complex physiological

phenomena that occurs in cotton fields subjected to different water levels. Furthering

the knowledge on cotton-water relations can be a tool for the development of

improved water management strategies.


113

References




Baker, J.T.; D.C. Gitz, P.Payton, D.F. Wanjura, and D.R. Upchurch. 2007. Using

Leaf Gas Exchange to Quantify Drought in Cotton Irrigated Based on Canopy

Temperature Measurements. Agron. J. 99:637–644.

doi:10.2134/agronj2006.0062

Baker, J.T.; S.V. Scott, D.C. Gitz, P. Payton, R.J. Lascano and B. McMichael. 2009.

Canopy Gas Exchange Measurements of Cotton in an Open System. Agron. J.

101:52–59. doi:10.2134/agronj2008.0007x




Bauer P., Faircloth D.W., Rowland, D., Ritchie G.L. (2012) Water-sensitivity of









Sci. Soc. J. 65:1153-1163.







da Costa, V.A.; and J.T. Cothren. 2011. Drought Effects on Gas Exchange,

Chlorophyll, and Plant Growth of 1-Methylcyclopropene Treated Cotton.

Agron. J. 103:1230-1241.


114





Full-, Deficit-Irrigated, and Dryland Cotton on the Northern Texas High

Plains. Journal of Irrigation and Drainage Engineering 130:277-285. DOI:

10.1061/(ASCE)0733-9437(2004)130:4(277).


Academic Press.

Kramer, P.J., Boyer J.S. (1995) Water relations of plants and soils. Elsevier Science.













Science 16:42-52.






f>, USDA.




115

CHAPTER VII

IRRIGATION AND EPISODIC DROUGHT MANAGEMENT EFFECTS ON THE

PROFITABILITY OF COTTON CULTIVARS IN WEST TEXAS

7.1. Abstract

Water management is a very important factor in irrigated cotton production systems.

The field experiments were conducted at the Texas Tech New Deal and Quaker

Avenue research farms during the 2010, 2011 and 2012 seasons. Eight cultivars were

studied in all the three years. In 2012 cultivars DP0935 and FM9180 were tested on all

irrigation regimes in the same location. Episodic drought treatments included a full

irrigation throughout the season; non-irrigation from squaring to flowering; 3 weeks of

non-irrigation beginning at early flowering; 3 weeks of non-irrigation beginning at

peak bloom; and non-irrigation beginning from peak bloom to the crop termination.

The studies also included deficit irrigation, mild deficit irrigation and a dry land

management started after crop establishment. The profitability and break-even price

were calculated based on the 2012 season costs and returns. In 2011, the reduction in

yield due to episodic drought treatments resulted in a reduction in profits. In 2012,

even if the area of mild deficit irrigation regime was increased to use the same amount

of water of a fully irrigated hectare, the profit of a fully irrigated hectare would be

higher. These results can be important to the development of more economically

efficient water management strategies.


116

7.2. Observation

This research data was based on small plots composed of two 12.2 m rows, which

were ginned with a table-top gin. The smaller scale can affect, compared with industry

data, the following: Turnouts (lint percentages) used to calculate lint yields; and the

impact of fiber quality parameters on premiums and discounts. Therefore, Chapter 7 is

a demonstration of the methodology that could be used with industrial scale data.

7.3. Introduction

Water management is a very important factor in any crop production system.

According to Denning et al. (2001), the Texas High Plains is one of the most

important cotton producing areas in the U.S., accounting for nearly 20% of the total

U.S. production during the past decade. In this region, almost half of the cotton area is

irrigated and the irrigated acreage is responsible for most of the production. The main

source of water to West Texas irrigated cotton is the Ogallala Aquifer that has been

depleted at a rate superior to its recharge. As stated by Bordovsky et al. (2000), as

water supplies and availability continue to decline in Texas, it is imperative to adopt

the most efficient irrigation systems and management techniques. Farm-level yields in

the Texas High Plains are significantly influenced irrigation water application rates.

Improving irrigation management practices can be profitable and may also increase

the usable life of the Ogallala Aquifer.

Information about irrigation costs and water response will help growers to

improve overall water management and water returns. No comprehensive studies


117

compared the effects of continuous and episodic water restriction on irrigated cotton

profitability. This knowledge can lead to more efficient irrigation strategies for

farmers.

7.3.1. Literature review

Water management is one of the most important issues related to crop

production systems. Irrigation is a practice that may be integrated with other crop

practices to maximize farmers’ profits. Irrigation of crops with subsurface drip

continues to increase in the Texas High Plains, so, irrigation strategies are needed to

optimize water use under limited water-well capacities (Buffon, 2010).

Mills (2010), in a study on water use efficiency (WUE) in castor, sorghum and

cotton, found that castor and sorghum yields provided significant linear responses to

increased irrigation depths. Cotton resulted in significant quadratic responses reaching

peak yields with 75 and 65 cm of total water in 2008 and 2009, respectively. Due to

heavy rainfall in 2008, irrigation produced varying results in cotton fiber quality.

Economic analysis suggested that profit maximizing yields were lower than the

maximum yields (Mills, 2010).

Irrigation termination is also an important aspect related to irrigation

management. According to Sneed (2010), in a study about irrigation termination to

improve fiber maturity on the Texas High Plains, irrigation termination had a

significant impact on yield, fiber maturity, and the overall quality. The earliest

irrigation termination matured the earliest and produced the lowest yields. Lint yields


118

were higher in the latest irrigation termination, but used more water and also reduced

fiber quality in most locations. Excessive rainfall and drought conditions that were

observed during this study impacted yield and quality.

Mills (2010) returns were nonlinearly related to irrigation amount. Denning et

al. (2001) also concluded cotton had non-linear responses to irrigation amounts; for

this reason, Mills (2010) suggested that profit maximizing levels of yield are lower

than yield maximizing levels. This also agrees with the basic concepts of the theory of

the firm (Doll and Orazem, 1992).

Mills (2010) results were different for each year due to the amount of rainfall

within each growing season. The curvature in all the results suggested that producers

can apply too much water, thereby decreasing net returns. With this information, Mills

(2010) concluded that the need for more irrigation management is apparent. He

suggested that, in the Texas High Plains, 0.64 cm of water per day is too much in most

years and that with high evaporation rates and low annual rainfall, more conscientious

water management would increase profit margin (Mills, 2010).

Sneed (2010) found that, depending on the cost of irrigation water, the most

profitable time to terminate irrigation is not always the highest yielding. Depending on

the cost of irrigation water and the availability of water, a producer could use less

water and still have equal or higher profits. Cultivar selection also had an impact on

profits.

Wilde (2008), in a study about optimal economic combination of irrigation

technology and cotton varieties on the High Plains of Texas, estimated gross margins


119

and net returns above total variable and fixed irrigation costs for varieties and

irrigation systems with differing irrigation levels. He found that producers could

increase gross margins by adopting new varieties. Estimations showed that subsurface

drip irrigation (SSI) can produce higher net returns than low energy precision

application (LEPA) center pivot systems. He also demonstrated the importance of

managing production according to environmental factors. His analysis emphasized the

importance for cotton producers to be informed and properly manage their production.

The main conclusion of the Denning et al. (2001) study was that knowledge

and consideration of the effect of management decisions on lint quality can

substantially increase expected profitability and reduce profit variation. The authors

also concluded that better decision-making ability will improve profitability of farm

operations. Availability of precise and up-to-date estimates of quality premiums and

discounts, implicit in the observed market lint prices, would be critical for these

purposes.

The final note from Denning et al. (2001) was caution about the production

response models estimated in his study. Though statistically sound, their models were

based on three years of experimental data from Lubbock County. The yield and

quality predictions from these models are imperfect due to the usual “random” error,

e.g., the effect of factors not included in the models. When applied in farm

management decisions, the predictions would also be subject to “extrapolation” error

caused by any major difference between the experimental site management and the

farm site management. Re-estimating the models on the basis of an expanded data set


120

that includes future-year observations from other Texas High Plains cotton-farming

areas could reduce this extrapolation error. We can conclude from these

recommendations that more research in related areas is needed.

Moore and Negri (1991) developed a multi-crop production model of irrigated

agriculture, applied to water allocation policy of the Bureau of Reclamation. By

formally modeling surface water as a fixed, allocable input to a multi-output firm, this

research captured the institutional constraints governing water allocation and,

established a cohesive approach to analyze the production effects of Bureau of

Reclamation allocation policy.

Econometric results from Moore and Negri (1991) showed that crop supply

and land allocation decisions are generally inelastic with respect to the water

constraint. Using the elasticities, a policy simulation of a 10% reduction in water

allocation indicates that production response to reduced water supply would affect the

national price of three out of ten major crops produced by Bureau of Reclamation-

served farms. This study suggests that studies about crop production and response to

water can also be useful for water allocation models and definition of water allocation

policies.

Robinson et al. (2010) estimated the economic value of irrigation water

shortfalls and mitigation responses of farmers in the Lower Rio Grande Valley of

Texas. In this work, water shortage levels closely correspond to supply shortfalls

experienced by the U.S. during the 1990s when Mexico fell behind on treaty delivery

obligations. They identified and evaluated a range of crop choices, appropriate


121

irrigation technology use, water source substitution, and other mitigation strategies

used by farmers to deal with water shortages. The effects of exogenous crop price and

yield risk, as well as other structural considerations were incorporated in the

estimation of the marginal value of irrigation water. Their results showed that South

Texas farmers react to risk by diversifying their crop mix, with implications for the

imputed value of water and soil resources.

The resulting marginal values found by Robison et al. (2010) reflected grower

adjustments for risk using crop mix, irrigation level, and irrigation technology. The

aggregate damage estimates using their approach are realistically smaller than

previous damage estimates that were based on fixed cropping patterns and average

water values. From their results, we can conclude about the importance of economic

studies related to water stress, for example related to water shortage scenarios and also

to the inclusion of different crops comparison when considering growers water

management studies.

According to Allen et al. (2007) water sources for irrigation, including surface

and groundwater, are declining in quality and quantity. Roseta-Palma (2003) in a

study of joint quantity/quality management of groundwater, presenting two alternative

models, optimal taxes were derived, and shown to be different from those in existing

quantity-only or quality-only models, in its conclusions it was emphasized that when

quality and quantity of aquifers are important, optimal policies must reflect the

relationship between them and the main features used in the characterization of

production and pollution functions draw heavily on agricultural literature, which relies


122

on the importance of agricultural water management studies to the development and

application of this kind of models.

Bordovsky et al. (2000) conducted an economic evaluation of cotton on the

Texas High Plains irrigated by LEPA and SSI. These authors explained that the

advantages of SSI over LEPA in cotton production are increased cotton lint yield and

improved water use efficiencies, particularly at very low irrigation capacities. They

also stated that irrigation management economic evaluations are as important as

irrigation systems’ themselves.

According to Allen et al. (2007), improvements in irrigation lead to

efficiencies that now exceed 95%; however, the improved efficiencies have often led

to increased water use, instead of water savings, since more systems have been

installed. Also, as groundwater becomes scarcer, more energy is required to extract

water from greater depths. Increasing demands for alternative water uses and depletion

of historic water sources make many irrigated systems in dry climates non-sustainable.

The Texas High Plains exemplifies these challenges, where agriculture depends

heavily on irrigation at non-sustainable rates of water extraction from the Ogallala

aquifer. Today, agriculture uses about 95% of total water withdrawn from this aquifer.

Doorembos and Kassam (1979) suggested that when the cotton water supply is

limited, different from other plants such as banana, it is possible to obtain a higher

total cotton production increasing the area and providing partially the water needs than

fully supplying a limited area. In a banana plantation, for example, a higher total

production would be achieved using just the area were there would be available water


123

to completely supply the plant irrigation needs than having a larger area not fully

irrigated.

In summary, irrigation strategies studies and their economic relevance are

extremely relevant from many different stand points; especially when considered in a

scenario of increased water use along with water shortages, such as the one facing

cotton production in West Texas.

7.3.2. Economic Analysis Conceptual Framework

One of the most important aspects related to enterprise management is

economic returns, making it very important to determine the effects of improved crop

practices, such as new water management strategies.

According to Doll and Orazem (1992), economics is the study of how

resources are used to satisfy the needs and desires of people. Its four tenets are:

scarcity, allocation (alternative uses), goals (ends or objectives) and choice indicators

(objective function or criterion). Therefore, economics can be defined as the allocation

of scarce resources among competing alternatives.

Agricultural Economics applies economic models to agricultural problems,

using economic principles in the biological nature of agriculture. It is assumed that the

goals of Agricultural Economics are to increase efficiency in agriculture, produce

needed products without wasting resources, increase the efficiency of farms and

increase the efficiency of resource use.


124

Production Economics is the application of the principles of microeconomics

in agriculture (Doll and Orazem, 1992). One of the goals of Production Economics is

to assist managers in determining the best use of their resources, given the changing

needs, values and goals of society. It is often assumed that the objective of the farm

manager is to maximize profit and increase efficiency in the use of resources.

To determine the profit of an enterprise under certain management, it is

necessary to develop an enterprise budget. Enterprise budgets provide an estimate of

costs and returns related to the production process. They also help to allocate land,

labor, and capital. Usually they are divided into “Total Revenue”, “Variable Costs” (or

direct expenses), “Fixed Costs” (or fixed expenses) and Income.

Total revenue (TR) is obtained multiplying product yield (Y) by its price:

TR = Y * Price .............................................................................................................. (1)

The enterprise budget also allows us to calculate the profit subtracting total

cost (TC) from total revenue (Eq. 2):

Profit = TR – TC ........................................................................................................... (2)

Other interesting factors that can be obtained from the data provided by an

enterprise budge are: the breakeven price and breakeven yield.

Breakeven price (BEP) is the price that, given an expected yield, equals total

revenue and total costs (Eq. 3):

Y

TCBEP ................................................................... ………………………………. (3)


125

Breakeven yield (BEY) is the yield that, given an expected price, also equals

total revenue and total costs (Eq. 4):

price

TCBEY ................................................................................................................. (4)

When there is secondary revenue (SR) such as cotton seeds the BEP and BEY

take the forms of equations 5 and 6:

Y

SRTCBEP

............................................................................................................ (5)

price

SRTCBEY

............................................................................................................ (6)

On a study, fiber quality premiums or discounts (PD) were included in the

prices used on the BEP and BEY analysis calculated on equations 5 and 6.

From extension services it is possible to find estimations for enterprise budgets

based on several conditions (Texas... 2012). For the conditions of this study, the most

appropriated budget was selected.

Using the budget for West Texas drip irrigated cotton from the West Texas

A&M AgriLife Extension Services web site (2012), we can develop an estimate of

how lint yields and lint prices will affect income, break-even price and returns above

irrigation costs, allowing us to make economic comparisons for different irrigation

amounts and the resultant yields.

Table 7.1 applies the extension service enterprise budget to create an adjusted

budget, with irrigation cost separated from the other costs. The total irrigation cost was

calculated by adding labor, fuel cost, and materials associated with drip irrigation


126

application and maintenance, with the assumption that all of these factors are directly

related to the amount of irrigation applied. The extension budget reports quantities in

conventional units, which have been converted to the International System (SI) in

Table 7.1.

Table 7.1. Adjusted cotton enterprise budget (1 ha)

ITEM UNIT PRICE QUANTITY AMOUNT

Dollars dollars

REVENUE

cotton lint kg 1.76 1681 $ 2,958.56

Secondary revenue

(cottonseed)

kg 0.20 1177 $ 235.40

---------------

TOTAL REVENUE $ 3,193.96

COSTS

Irrigation related variable

costs

mm 1.16 432 $ 501.12

Non-irrigation related costs ha 1,559.06 1 $ 1,559.06

---------------

TOTAL COSTS $ 2,060.18

PROFIT (eq. 2) ha 1 $ 1,133.78

Breakeven price (eq. 6) kg 1.09

Breakeven yield (eq. 8) kg 1037

Table 7.1 helps us to establish a series of relations that can be used to compare

irrigation management strategies.

Applying Eq. 6 to the Table 1 data we can find the Break-even Yield, as our

first criterion to evaluate the proposed irrigation management strategies as follow:

103776.1

40.23518.060,2

BEY .............................................................................................. (7)


127

We can conclude that in the proposed scenario, the breakeven yield is 1037 kg

ha-1

if the price is held at $1.76 per kg lint (Eq. 7). The same method can be used with

Eq. 6 to calculate the breakeven price as $1.09 per kg if the lint yield is 1681 kg ha-1

.

In Table 7.1, we assume non-irrigation related costs to be the same ($

1,559.06) for all irrigation treatments. To calculate TC for each irrigation management

strategy based on the irrigation depth (mm) applied per treatment (IAt) we can use Eq.

8:

TCt = 1559.06 + 1.16* IAt ............................................................................................ (8)

From combining equations 1, 2, and 8 we can have an equation to calculate

Profit per treatment (Pt) in dollars as a function of IAt, the seed yield per treatment

(Yst), and Yt, the field trials data:

Pt = TRt – TCt = 0.20 * Yst + 1.76 * Yt – 1559.06 – 1.16 * IAt ....................................... (9)

We can also combine Eq. 5 with Table 7.1 data to calculate a breakeven price

for each treatment (BEPt):

Yt

SRBEPt

IAt 1.16 + 1559.06 ........................................................................... ….. (10)

Treatments with the highest profit and the lowest breakeven prices should also

be the most economically beneficial irrigation management strategies. A complete set

of parameters must be observed to understand the effects of irrigation management

and Episodic Drought. Episodic Drought for this research is defined as a period when

the irrigation of an irrigated crop is completely interrupted.


128

7.3.3. Objectives

The general objective of this chapter is to determine optimum economic water

management strategies for cotton.

The specific objectives are:

i) Determine the effects of episodic water stress and irrigation management

strategies on the yield and production economics related variables for selected cotton

varieties.

ii) Determine if higher total production and profits can be achieved fully irrigating

a limited area or by using any of the proposed water saving strategies in a bigger area,

with the same total water applied.

iii) Relate irrigation and cultivar selection effects on boll distribution to fiber

quality, and profitability.


Yield, fiber quality, and boll distribution were obtained from a set of episodic

drought under different stages trials (Episodic trials) and from a set of varying

irrigation in diverse cultivars field trials (Irrigation trials) that we conducted in the

West Texas region from 2010 to 2012, as described on chapters 3 and 4.

The field experiments were conducted at the Texas Tech New Deal and

Quaker Avenue research farms during the 2010, 2011 and 2012 seasons. These

experimental sites are located in the Texas High Plains, considered a semi-arid area.

The soil at the Quaker Avenue farm is an Amarillo-Acuff sandy clay loam (Fine-


129

loamy, mixed, superactive, thermic Aridic Paleustolls), and the soil at the New Deal

farm is a Pullman-Olton clay loam (Fine, mixed, superactive, thermic Aridic and

Torrertic Paleustolls).







For both sets of trials the experimental design was a split-plot with water

treatments as the main plot and cultivars as the split-plot. For the experimental units

were conducted in two rows, measuring 10.7 m in 2010 and 12.2 m in 2011 and 2012.

Row spacing was one meter. SSI system was used. The tape was placed 0.25 m below

the surface under every row.

7.4.1. Episodic trials

In 2010 the Episodic trial was located at the Texas Tech New Deal research

farm (New Deal); in 2011, this trial was located at Quaker and; and in 2012 the trial

was repeated at both locations.

The Episodic Drought treatments included a full irrigation throughout the



irrigation beginning from peak bloom to the crop termination. We considered the


130

squaring stage to begin when the first square (floral bud) was observed, and the early

flowering stage to begin when the first opened flower was observed. At the 2010

season due to the occurrence of precipitation events in the early season the treatments

3 weeks of non-irrigation beginning at early flowering; and 3 weeks of non-irrigation

beginning at peak bloom were not present.

From a set of eight cultivars evaluated for the Episodic Drought trial in 2010,

we selected cultivars DP0912, DP0935, FM9170, and FM9180 for the 2011 and 2012

seasons.

7.4.2. Irrigation trials

In 2010 and 2011 the Irrigation trial was located at New Deal, and in 2012 the

Irrigation trial was located at Quaker.

Besides a fully irrigated treatment, in 2010 the Irrigation trial had a deficit

irrigated management, and a treatment with irrigation interrupted after the crop

establishment (squaring). In 2011 and 2012 Irrigation trial treatments also included a

second level of deficit irrigation.

From a set of 21 cultivars and experimental lines planted for the Irrigation trial

in 2010, cultivars ACALA 1517-E2, FM832, SIOKRA24, and ST506 were selected

for the 2011 and 2012 seasons. FM9180 was also tested on the 2011 and 2012

Irrigation trials. DP0935 was also tested on the 2012 Irrigation trial.


131

7.4.3. Measurements and Analysis

Weather climate parameters were also obtained from automatic weather

stations located in the experimental fields and include precipitation and other

measurements necessary to calculate the irrigation needs. The total water applied was

also measured with hydrometers. All plots were harvested with a cotton stripper and

weighed. Grab samples were taken during harvest, after ginning the samples were sent

for analysis at the using the High Volume Instrument (HVI) at the Fiber and

Biopolymer Research Institute (FBRI) in Lubbock, TX. During ginning seed-cotton,

lint, and seeds weights were taken for the turnout (lint percentage), lint yield, and seed

yield calculation.

Based on information obtained from the West Texas A&M AgriLife Extension

Service (2012) budgets and the equations presented on the “Conceptual Framework”

section it was possible to calculate Returns, Profits, and Break-even prices. Based on

the Plains Cotton Growers Inc. website (2012) it was possible to determinate

premiums and discounts based on HVI fiber quality parameters provided by the FBRI.

Seed-cotton water use efficiency (AWUE) was computed by the division of Seed-

cotton yield by the total water received per treatment (irrigation plus rainfall). Similar

procedure was used to calculate Lint Water Use Efficiency.

For testing the hypothesis that a higher cotton production can be achieved by

partially irrigating a big area, we first calculated for each irrigation level the area that

would be irrigated with the volume of water from the full irrigation treatment, and

then we multiplied this area to the yield, obtaining equivalent a water volume


132

production to be compared. Profits for equivalent water volumes were also computed

by the multiplication of the profit per hectare to the increased areas that would use

equal volumes of water.

Marginal Physical Product (MPP) is the ratio of the increase in physical yield

(e.g. lint yield) divided by the increase on the water applied. In this study the treatment

effects was analyzed by the MPP from the lower yield and irrigation treatment

(dryland) to a given water strategy yield and irrigation, and also from a given

treatment to the higher yield and irrigation treatment (full irrigation with no drought).

Similar to the MPP a Marginal Value Product (MVP) was obtained by the ratio of the

increase in revenue to the increase on water applied.

In 2012, we had DP0935 and FM9180 present in both trials at the Quaker

farm, so it was possible to compare the effects of all irrigation and episodic drought in

these two cultivars in that environment.


software (SAS Inst., 2010) with an ANOVA followed by a mean separation at 5%

level of probability using the LSMEANS statement. The GLIMMIX model performs

estimation and statistical inference for generalized linear mixed models by





133

7.5. Results

Tables 7.2 and 7.3 show the significance of the Seed Cotton Yield variable as a

function of the water treatment, cultivar selection, and irrigation-cultivar interaction

for all year-location environments of the Episodic Drought and Irrigation sets of trials.

For all environments and trials statistically significant (p<0.01) irrigation effect was

observed. A cultivar single main effect was observed for two environments of each

set of trials. While the irrigation-cultivar interaction was statistically significant for all

environments of the Episodic drought sets of trials, at the Irrigation levels set of trials

it was significant just in 2012.

Table 7.2. Episodic Drought trials. Statistical significance (p-values) for the F test due

to effects on seed-cotton yield

Factor p-value in

2010

p-value in

2011

p-value in

2012(Qk)

p-value in

2012(ND)

Irrigation 0.0007 ** < 0.0001 ** < 0.0001 ** < 0.0001 **

Cultivars 0.0013 ** < 0.0025 ** 0.926 n.s. 0.43 n.s.

Irrigation*Cultivar 0.0015 ** 0.04117 * < 0.0001 ** 0.0215 *


ND = Texas Tech New Deal Research Farm.





134

Table 7.3. Irrigation levels trials. Statistical significance (p-values) of the study factors

and interaction in the variable seed-cotton yield during the 2010, 2011 and 2012

seasons


Irrigation <0.0001 ** < 0.0001 ** < 0.0001 **

Cultivars <0.0001 ** 0.1071 n.s. < 0.0001 **

Irrigation*Cultivar 0.1297 n.s. 0.5667 n.s. 0.0006 **




7.5.1. 2011

Table 7.4 shows the significance of the treatments on yield related varibles in

the 2011 Episodic Drough Trial. These analyses included parameters used to evaluate

the physical efficiency of the water strategies such as agronomic water use efficiencies

(AWUE), productions obtained with an equal irrigation water volume over an

increased area, and marginal lint yield (MPP).

Table 7.4. Significance (p-values) of the analysis of variance for production variables

as affected by episodic irrigation interruption, cultivar or their interaction in 2011.

Factor Irrigation Cultivar Interaction

Seed-cotton yield < 0.0001 ** 0.0119 * 0.1166 n.s.

Lint yield < 0.0001 ** < 0.0001 ** 0.1672 n.s.

Lint percentage 0.0112 * < 0.0001 ** 0.0322 *

Seed-cotton water use efficiency < 0.0001 ** 0.0207 * 0.2119 n.s.

Lint water water use efficiency < 0.0001 ** < 0.0001 ** 0.3033 n.s.

Equal water seed cotton production < 0.0001 ** 0.0240 * 0.2395 n.s.

Equal water lint production < 0.0001 ** < 0.0001 ** 0.3392 n.s.

Marginal lint yield < 0.0001 ** 0.0008 ** 0.0390 *



135

Based on Table 7.4 we can conclude that, in the 2011 Episodic Drough Trial,

all yield related variables displayed were statistically affected by the irrigation period

and cultivar selection. However, an irrigation-interaction was only observed by the

effects on lint percentage and MPP. A more detailed view of the effects of each

treatment on each variable can be observerd on figures 7.1 to 7.9.

Figure 7.1. Seed-cotton yield of four cotton cultivars subjected to five different

irrigation interruptions periods during the 2011 season at the Texas Tech Quaker

Avenue research farm. Black bars represent one least significant difference.

LSD(0.05) = 203 kg ha-1

.

0

500

1000

1500

2000

2500

3000

3500

4000


flowering

3 weeks at

peak bloom

peak bloom

to

termination

See

d -

cott

on

yie

ld (

kg

ha

-1)


DP0912

DP0935

FM9170

FM9180


136

Figure 7.2. Lint yield of four cotton cultivars subjected to five different irrigation

interruptions periods during the 2011 season at the Texas Tech Quaker Avenue

research farm. Black bars represent one least significant difference. LSD (0.05) = 91

kg ha-1

.

In 2011, when there was no irrigation interruption, the yield (seed-cotton yield,

and lint yield) was statistically higher than all other treatments. the irrigation

interruption at the squaring stage resulted in the smallest reduction in yield (figures 7.1

and 7.2) from the fully irrigated treatment. When the irrigation interruption occurred

during the early flowering treatment, the highest yield reduction was observed, the

early flowering episodic drought caused a reduction of more than 800 kg of lint per

hectare. Regarding cultivar effect DP0912 provided the highest yields and was not

statistically different of DP0935, both cultivars provided average lint yields of more

than 1,000 kg ha-1

.

0

200

400

600

800

1000

1200

1400

1600

1800


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Lin

t y

ield

(k

g h

a-1

)


DP0912

DP0935

FM9170

FM9180


137

Figure 7.3. Lint percentage of four cotton cultivars subjected to five different



LSD(0.05) = 1.9 %.

Although the effects of the treatments on seed-cotton and lint yields were very

similar, a significant differences of lint percentages (turnount) were observed (Fig.

7.4). The highest lint of all treatments percentages were achived by cultivars DP0912

and DP0935 under an early flowering irrigation interruption.

36%

38%

40%

42%

44%

46%

48%


flowering

3 weeks at

peak bloom

peak bloom to

termination

Lin

t p

ercen

tag

e


DP0912

DP0935

FM9170

FM9180


138

Figure 7.4. Agronomic seed-cotton water use efficiency of four cotton cultivars

subjected to five different irrigation interruptions periods during the 2011 season at the

Texas Tech Quaker avenue research Farm. Black bars represent one least significant

difference. LSD(0.05) = 0.61 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

.

Figure 7.5. Agronomic lint water use efficiency of four cotton cultivars subjected to

five different irrigation interruptions periods during the 2011 season at the Texas Tech

Quaker Avenue research farm. Black bars represent one least significant difference.

LSD(0.05) = 0.27 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

.

0

1

2

3

4

5

6

7

8

9

10

no drought squaring early flowering 3 weeks at peak

bloom

peak bloom to

termination

See

d-c

ott

on

WU

E

(kg

ha

-1 m

m-1

)


DP0912

DP0935

FM9170

FM9180

0

0.5

1

1.5

2

2.5

3

3.5

4

no drought squaring early flowering 3 weeks at peak

bloom

peak bloom to

termination

Lin

t W

UE

(k

g h

a-1

mm

-1)


DP0912

DP0935

FM9170

FM9180


139

Similar to the effect of yields, the highest Lint and seed-cotton water use

efficiency was higher from cultivars DP0912 and DP0935. Althoug the WUE of the

no drought treatment was the highest among all irrigation treatments, it was not

statistically higher than the one observed when the irrigation was interrupted at the

squaring stage. However, an irrigation interruption at the early flowering stage caused

a reduction in seed-cotton WUE of more than 4 kg ha-1

mm-1

when compared to the

fully irrigated treatment.

Figure 7.6. Area to irrigate using four different irrigation interruptions periods with a

water volume equal to the used to fully irrigate 1 ha during the 2011 season at the

Texas Tech Quaker Avenue research farm.

1.27

1.32

1.23

1.43

1

1.05

1.1

1.15

1.2

1.25

1.3

1.35

1.4

1.45

squaring early flowering 3 weeks at peak

bloom

peak bloom to

termination

Are

a u

sin

g 1

ha

of

full

irr

iga

tio

n

wa

ter (

ha

)



140

Figure 7.7. Seed-cotton production of four cotton cultivars subjected to five different

irrigation interruptions periods in an area adjusted to use the same volume of water of

a fully irrigated ha during the 2011 season at the Texas Tech Quaker Avenue research

farm. Black bars represent one least significant difference. LSD(0.05) = 262 kg.

Figure 7.8. Lint production of four cotton cultivars subjected to five different

irrigation interruptions periods in an area adjusted to use the same volume of water of

a fully irrigated ha during the 2011 season at the Texas Tech Quaker Avenue research

farm. Black bars represent one least significant difference. LSD(0.05) = 117 kg.

To analyse if using any a higher area with the same total water applied in a

fully irrigated treatment would provide a highter total production, we first calculated

0

500

1000

1500

2000

2500

3000

3500

4000

4500

no drought squaring early flowering 3 weeks at

peak bloom

peak bloom to

termination

See

d-c

ott

on

pro

du

ctio

n f

or

an

equ

al

wa

ter v

olu

me

(kg

)


DP0912

DP0935

FM9170

FM9180

0

200

400

600

800

1000

1200

1400

1600

1800


flowering

3 weeks at

peak bloom

peak bloom to

terminationEq

ual

wate

r volu

me

lin

t p

rod

uct

ion

(k

g)


DP0912

DP0935

FM9170

FM9180


141

the increased area (Fig. 7.6). Water savings from Episodic Drought treatments could

provide irrigation to areas more than 20% higher than a fully irrigation water

management. Regarding the total equal water lint and seed cotton production, although

the highest productions were observed under the squaring irrigation interruption, they

were not statistically different from the observed under no drought and peak bloom to

termination episodic drought. Even if the water saved in a early flowering irrigation

interruption was used to irrigate a bigger area, it would still provide a total production

reduced to more than 720 kg when compared with a equal water squaring episodic

drought, it would also mean a reduction of nearly 700 kg when compared with an fully

irrigated hectare yield.

Figure 7.9. Marginal lint yield to full irrigation of four cotton cultivars subjected to

five different irrigation interruptions periods during the 2011 season at the Texas Tech


LSD(0.05) = 1.09 kg ha-1

mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

.

In 2011, marginal lint yield was significativally higher from the early

flowering irrigation interruption treatment (Fig. 7.9), indicating that a crop subjected

0

2

4

6

8

10

12

14


bloom

peak bloom to

termination

Ma

rgin

al

lin

t y

ield

(k

g h

a-1

mm

-1)


DP0912

DP0935

FM9170

FM9180


142

to that episodic drought would have yields increased by nearly 7 kg ha-1

mm-1

more

than a crop with the irrigation interrupted at the squaring stage.

Table 7.1 established estimates of costs based on yields and total irrigation

depths proposed by the extension services, it is important to remember that field data

provided different lint yield and correspondent irrigation depth applied. Field data was

used in the models described in equations 9 to 10 to calculate production economics

related variables used in theses analyses.

Table 7.5 shows the significance of the treatments on Production Economics

related varibles in the 2011 Episodic Drought trial. These analyses included

parameters used to evaluate the economical efficiency of the water strategies such as

profits, breakeven prices, profits obtained with an equal irrigation water volume over

an increased area, and marginal value product (MVP).

Table 7.5. Significance (p-values) of the analysis of variance for production

economics related variables as affected by episodic irrigation interruption, cultivar or

their interaction in 2011.


Total premiums or discounts 0.0002 ** 0.0003 ** 0.1977 n.s.

Total revenue < 0.0001 ** 0.0001 ** 0.0710 n.s.

Total profit < 0.0001 ** 0.0001 ** 0.0710 n.s.

Breakeven price < 0.0001 ** 0.0056 ** 0.0300 *

Equal water volume profit < 0.0001 ** 0.0005 ** 0.1561 n.s.

Marginal value product < 0.0001 ** < 0.0001 ** 0.0128 *

*: significant at p = 0.05; **: significant at p = 0.01; n.s.: not significant at p = 0.05

Similar to the parameters from Table 7.4, based on Table 7.5 we can conclude

that, in the 2011 Episodic Drought trial, all production economics related variables


143

from this study were statistically affected by the irrigation irrigation period and

cultivar selection at p=0.01. However, an irrigation-interaction for p=0.05 was only

observed by the effects on breakeven Price and MVP. A more detailed view of the

effects of each treatment on each variable can be observerd on figures 7.10 to 7.16.

Figure 7.10. Premium or discount received by four cotton cultivars subjected to five

different irrigation interruptions periods during the 2011 season at the Texas Tech


LSD(0.05) = US$ 0.05 kg-1

.

Based on the actual system used for premiums and discounts (Plains… 2012)

in 2011 lowest discounts would be applied on cultivar FM9180 that would not

statistically differ from FM9170. Regarding episodic drought effects, the highest

premiums occurred on the no drought treatment (nearly US$ 0.01 kg-1

) and it was not

statistically different from discounts applied when the irrigation interrruption occurred

at the squaring and 3 weeks at peak bloom stages.

-0.2

-0.15

-0.1

-0.05

0

0.05

0.1


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Pre

miu

m o

r d

isco

un

t (U

S$

kg

-1)

Irrigation Interruption Period

DP0935

FM9180

DP0912

FM9170


144

The average of bolls per node in each episodic drought treatment is shown in

Fig. 7.11 in order to relate it to the fiber quality effects on premiums and discounts

presented on Fig. 7.10.

Figure 7.11. Average number of first positon bolls per node from four cultivars

subjected to five different irrigation interruptions periods during the 2011 season.

season at the Texas Tech Quaker Avenue Research Farm.

A fair assumption that could be made is that on the higher nodes of a plant,

cotton lint will be shorter and less mature, therefore more likely to receiving

discounts. However, based on 2011 episodic drought data we can not prove this

hypotheses. Fig. 7.11 shows that, when the irrigation was interrupted at the squaring

stage, plants compensated the early loss of bolls in the bottom (nodes 5 to 9) with

more bolls at the top (nodes 11 and 12), however, the effect of more bolls at the top

didn’t cause a statistically lower discount based on fiber quality parameters used in the

actual system. In Fig. 7.11 we also can notice that, although the early flowering

irrigation interruption caused a reduction on bolls in all nodes, the highest reduction

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Aver

age

nu

mb

er o

f b

oll

s

Node

no drought

squaring

early flowering

3 weeks at peak

bloom


145

occurred on the top bolls, but as shown on Fig. 7.10 this episodic drought did not

improve the premiums and actually had the highest discounts of all irrigation

interruption treatments.

Figure 7.12. Total revenue from four cotton cultivars subjected to five different



LSD(0.05) = US$ 203.88 ha-1

.

0

500

1000

1500

2000

2500

3000

3500


flowering

3 weeks at

peak bloom

peak bloom

to

termination

To

tal

rev

enu

e (U

S$

ha

-1)


DP0935

FM9180

DP0912

FM9170


146

Figure 7.13. Profit or loss of four cotton cultivars subjected to five different irrigation

interruptions periods during the 2011 season at the Texas Tech Quaker Avenue

research farm. Black bars represent one least significant difference. LSD(0.05) = US$

203.88 ha-1

.

In 2011, the highest revenue (Fig. 7.12) and profit (Fig. 7.13) were achieved at

the no drought treatment. While an average profit of US$ 740.50 per hectare was

observed when there was no drought, losses were observed when there were irrigation

interruptions on the peak bloom to termination and on early flowering stages (-US$

1034.1 per hectare in average). When there was irrigation interruption when the first

squares were observed, profit was reduced in nearly US$ 500 per hectare compared

with the no drought treatment.

-1500

-1000

-500

0

500

1000

1500


peak bloom

peak bloom to

terminationPro

fit

or

loss

(U

S$

ha

-1)


DP0935

FM9180

DP0912

FM9170


147

Figure 7.14. Breakeven price of four cotton cultivars subjected to five different



LSD(0.05) = US$ 0.60 kg-1

.

2011 break-even price per treatment is shown on Fig. 7.14. The price that

makes the revenue equal to the costs was statistically affected by the irrigation

interruption period (p<0.0001), and cultivar selection (p=0.0056). There was also an

irrigation-cultivar interaction (p=0.03). While when there was an episodic drought at

early flowering stage the highest break-even price was achieved from cultivars

FM9170 and was just not statistically different from FM9180 in the same water

regime. However, at the no drought, squaring and peak bloom irrigation treatments,

the break-even price was not statistically different among cultivars averaging US$

1.54 per kg of cotton lint.

0

1

2

3

4

5

6


peak bloom

peak bloom to

termination

Bre

ak

even

pri

ce (

US

$ k

g-1

)


DP0935

FM9180

DP0912

FM9170


148

Figure 7.15. Profit of an increased area irrigating using four different irrigation

interruptions periods with a water volume equal to the used to fully irrigate 1 ha

during the 2011 season at the Texas Tech Quaker Avenue research farm. Black bars

represent one least significant difference. LSD (0.05) = US$ 264.86.

Similar to the equal water volume lint yield presented on Fig. 7.8, Fig. 7.15

shows the profits that might be achieved on an increased area using the same irrigation

water volume. As we can observe the total profits where still higher when no episodic

drought occurred. When the episodic drought occurred at the early flowering the

losses were even bigger.

-2000

-1500

-1000

-500

0

500

1000

1500


peak bloom

peak bloom to

termination

Eq

ua

l ir

rga

tio

n v

olu

me

pro

fit

(US

$)


DP0935

FM9180

DP0912

FM9170


149

Figure 7.16. Marginal value product to full irrigation of four cotton cultivars subjected

to five different irrigation interruptions periods during the 2011 season at the Texas

Tech Quaker Avenue research farm. Black bars represent one least significant

difference. LSD(0.05) = US$ 1.33 ha-1

mm-1

. US$ 1.00 ha-1

mm-1

= US$ 0.01 10-2

L-1

.

Similar to the MPP (Fig. 7.9), Fig. 7.16 shows in 2011, MVP was significantly

higher from the early flowering irrigation interruption treatment, indicating that a crop

subjected to that episodic drought would have revenues increased in nearly US$ 7.77

ha-1

mm-1

more than a crop with the irrigation interrupted at the squaring stage.

7.5.3. 2012

In 2012, both episodic drought and different levels of irrigation were evaluated

for cultivars FM9180 and DP0935 at the Texas Tech Quaker Avenue Research Farm.

Table 7.6 Summarizes the significance of some factors that we can use to compare the

effects of different water allocation strategies on cotton yield. For all factors we found

extremely significant irrigation effects (p<0.0001). Cultivar selection had statistically

0

2

4

6

8

10

12

14

16


bloom

peak bloom to

termination

Ma

rgin

al

va

lue

pro

du

ct (

US

$ h

a-1

mm

-1)


DP0935

FM9180

DP0912

FM9170


150

significant effect on lint yield, lint percentage, and marginal lint yields (MPP).

Irrigation-cultivar interaction was observed for lint percentage and MPPs. More details

of the effect of the treatments on these variables can be observed on figures 7.17 to

7.23.

Table 7.6. Significance (p-values) of the analysis of variance for production variables

as affected by the irrigation episodic drought, cultivar or their interaction in 2012.


Seed-cotton yield < 0.0001 ** 0.6115 n.s. 0.7325 n.s.

Lint yield < 0.0001 ** 0.0008 ** 0.1021 n.s.

Seed-cotton water use efficiency < 0.0001 ** 0.3756 n.s. 0.3942 n.s.

Equivalent water area production < 0.0001 ** 0.3483 n.s. 0.3663 n.s.

Marginal lint yield from dryland < 0.0001 ** < 0.0001 ** 0.0001 **

Marginal lint yield to full irrigation < 0.0001 ** < 0.0001 ** < 0.0001 **


Figure 7.17. Average seed-cotton yield over cultivars DP0395 and FM9180 subjected

to eight different irrigation regimes during the 2012 season at the Texas Tech Quaker


LSD(0.05) = 338 kg ha-1

.

4,507

2,587 2,577 2,476

2,133 1,829 1,641

898

0

1000

2000

3000

4000

5000

6000

full

irrigation

3 weeks at

peak

bloom

squaring mild deficit

irrigation

peak

bloom to

termination

early

flowering

deficit

irrigation

dryland

after

squaring

See

d-c

ott

on

yie

ld (

kg

ha

-1)

Irrigation regime or interruption period

E


151

In 2012 seed-cotton yield was more than 3600 kg ha-1

higher in the full

irrigation treatment (highest yield treatment) when compared to the dryland after

irrigation management (Fig. 7.17). The seed-cotton yield when a 3 weeks at peak

bloom irrigation interruption occurred was not statistically different of the observed

from a squaring stage episodic drough and also didn’t differ from the yield from a

mild deficit irrigation management. The seed-cotton yield obtained after a early

flowering stage irrigation interruption was not statistically different from the obtained

from a deficit irrigation management.

Figure 7.18. Lint yield of cultivars DP0395 and FM9180 subjected to eight different

irrigation regimes during the 2012 season at the Texas Tech Quaker Avenue research

farm. Black bars represent one least significant difference. LSD(0.05) = 152 kg ha-1

.

In 2012 lint yield was more than 1400 kg ha-1

higher in the full irrigation

treatment (highest yield treatment) when compared to the dryland irrigation

0

500

1000

1500

2000

2500

full

irrigation

squaring mild deficit

irrigation

3 weeks at

peak bloom

peak bloom

to

termination

early

flowering

deficit

irrigation

dryland

after

squaring

Lin

t y

ield

(k

g h

a-1

)


DP0935

FM9180


152

management (Fig. 7.18). Similar to the seed-cotton yield, lint yield when a 3 weeks at

peak bloom irrigation interruption occurred was not statistically different of the

observed from a squaring stage episodic drough and also didn’t differ from the yield

from a mild deficit irrigation management. The lint yield obtained after peak bloom to

termination and early flowering stage irrigation interruption were not statistically

different from the obtained form a deficit irrigation management. The average yield of

DP0935 was 125 kg ha-1

higher than FM9180.

Figure 7.19. Average seed-cotton water use efficiency from cultivars DP0395 and

FM9180 subjected to eight different irrigation regimes during the 2012 season at the

Texas Tech Quaker Avenue research farm. Black bars represent one least significant


mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

.

In 2012, seed-cotton water use efficiency (AWUE) was higher at full

irrigation, followed by the mild deficit irrigation. The two lowest AWUE were

7.08

6.22

5.38 5.31 5.27 4.89

4.33 4.10

0

1

2

3

4

5

6

7

8

9

full

irrigation

mild deficit

irrigation

peak bloom

to

termination

deficit

irrigation

3 weeks at

peak bloom

squaring early

flowering

dryland

after

squaring

See

d -

cott

on

wa

ter u

se e

ffic

ien

cy (

kg

ha

-1 m

m-1

)

Irrigation Regime or Interruption Period


153

observed at the early flowering irrigation interruption and at the dryland treatments

and were not statistically different. All other irrigation treatments AWUE were not

statistically different averaging nearly 5 kg ha-1

mm-1

.

Figure 7.20. Area equivalent to the use of the water volume in one full irrigated ha for

seven different irrigation regimes during the 2012 season at the Texas Tech Quaker

Avenue Research Farm.

1.23 1.33 1.57 1.68 1.69

2.25

3.42

0

0.5

1

1.5

2

2.5

3

3.5

4

squaring 3 weeks at

peak bloom

early

flowering

mild deficit

irrigation

peak bloom

to

termination

deficit

irrigation

dryland after

squaring

Iso

wa

ter a

rea

(h

a)



154

Figure 7.21. Prodution of an expanded area for the same total water volume averaged

from cultivars DP0395 and FM9180 subjected to eight different irrigation regimes

during the 2012 season at the Texas Tech Quaker Avenue research farm. Black bars

represent one least significant difference. LSD(0.05) = 379 kg.

If the water saved on the several irrigation management strategies tested in

2012 were used to irrigate a larger area with the same total volume of water used in

one fully irrigated hectare, the increased area will have the dimensions described on

Fig. 7.20. These increased areas would provide the total production described on Fig.

7.21. The highest total production would still be achieved under one fully irrigated

hectare, but it was not statistically different from the total production obtained under a

mild deficit irrigation management.

0

1000

2000

3000

4000

5000

6000

full

irrigation

mild deficit

irrigation

deficit

irrigation

peak bloom

to

termination

3 weeks at

peak bloom

squaring dryland after

squaring

early

flowering

See

d -

cott

on

eq

ual

wate

r p

rod

uct

ion

(k

g)



155

Figure 7.22. Marginal from dryland lint yield of cultivars DP0395 and FM9180

subjected to eight different irrigation regimes during the 2012 season at the Texas



mm-1

. 1 kg ha-1

mm-1

= 1 g daL-1

.

0

1

2

3

4

5

6

mild deficit

irrigation

deficit

irrigation

full

irrigation

squaring 3 weeks at

peak bloom

peak bloom

to

termination

early

flowering

Marg

inal

from

dry

lan

d l

int

yie

ld (

kg

ha

-1 m

m-1

)


DP0935

FM9180


156

Figure 7.23. Marginal to full irrigation lint yield of cultivars DP0395 and FM9180




mm-1

. 1 kg ha-1

mm-1

= 0.1 g L-1

.

Figures 7.22 and 7.23 describe MPP’s. The highest increase in yields from the

dryland treatment (Fig. 7.22) would be achieved following the mild deficit irrigation

management (nearly 4 ha-1

mm-1

) and it was not statistically different from the deficit

irrigation management. In that scenario a higher MPP would also be observed on

cultivar DP0935, but if the increase in water would follow the fully irrigated

management, there would not be a cultivar difference. When the MPP was calculated

based on the deltas from the proposed treatment to the highest water and yield

treatment (full irrigation) the highest MPP was observed from the squaring episodic

0

1

2

3

4

5

6

7

8

9

mild deficit

irrigation

deficit

irrigation

dryland after

squaring

peak bloom to

termination

early

flowering

3 weeks at

peak bloom

squaring

Ma

rgin

al

to f

ull

irr

iga

tio

n l

int

yie

ld (

kg

ha

-1 m

m-1

)


DP0935

FM9180


157

drought treatment (Fig. 7.23) indicating that this treatment would be the bigger

beneficary of water increments.

On Table 7.7 the significance for production economics related variables is

presented. For all variables shown on Table 7.6 we had a signficant effect for

irrigation (p< 0.0001), and for cultivar at p=0.01. There was also a signficant

irrigation-cultivar interaction at p=0.05 for premiums, breakeven price, equal water

volume profit, and marginal value product.

Table 7.7. Significance (p-values) of the analysis of variance for production

economics related variables as affected by irrigation management, cultivar or their

interaction in 2012.


Total premiums or discounts < 0.0001 ** 0.0093 ** 0.0019 **

Total revenue < 0.0001 ** 0.0076 ** 0.2117 n.s.

Total profit < 0.0001 ** 0.0076 ** 0.2117 n.s.

Breakeven price < 0.0001 ** < 0.0001 ** 0.0008 **

Equal water volume profit < 0.0001 ** < 0.0001 ** 0.0124 *

Marginal value product < 0.0001 ** < 0.0001 ** < 0.0001 **

*: significant at p = 0.05; **: significant at p = 0.01; n.s.: not significant at p = 0.05


158

Figure 7.24. Premium or discount of cultivars DP0395 and FM9180 subjected to eight

different irrigation regimes during the 2012 season at the Texas Tech Quaker Avenue

research farm. Black bars represent one least significant difference. LSD (p=0.05) =

US$ 0.0099 kg-1

.

The effects of fiber quality in the price received in each treatment in 2012 is

described on Fig. 7.24. While only the full irritation would provide an increased price,

or premium, the higher discounts of nearly US$ 0.03 kg-1

would be received on the no

irrigation after the crop stablishment, and no irrigation from peak bloom to crop

termination, these discounts were not statistically different from the received at the

deficit irrigation. The single main effect of cultivar provided lower discounts for

FM9180, however, an irrigation-cultivar interaction was observed, and when there was

an episodic drought from peak bloom to termination the lower discount was obtained

by DP0935.

-0.05

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

full irrigation 3 weeks at

peak bloom

mild deficit

irrigation

squaring early

flowering

deficit

irrigation

no irrigation peak bloom

totermination

To

tal

pre

miu

m o

r d

isco

un

t (U

S$

kg

-1)


DP0935

FM9180


159

Figure 7.25. Average boll distribution per node from cultivars DP0395 and FM9180


Tech Quaker Avenue research farm.

The average number of bolls per node can be observed on Fig. 7.25. The

dryland, deficit irrigation and early flowering irrigation interruption present less bolls

on top wich can be normaly considered immature, so it was expected from they to

have higher premiums however, that was not confirmed by Fig. 7.24.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

5 6 7 8 9 10 11 12 13 14 15 16 17 18

Av

era

ge

nu

mb

er o

f b

oll

s

Node

full irrigation

3 weeks at peak bloom

peak bloom to termination

mild deficit irrigation

squaring

early flowering

deficit irrigation

no irrigation


160

Figure 7.26. Revenue obtained from cultivars DP0395 and FM9180 subjected to eight



US$ 326.00 ha-1

.

0

500

1000

1500

2000

2500

3000

3500

4000

4500

full irrigation squaring mild deficit

irrigation

3 weeks at

peak bloom

peak bloom

to termination

early

flowering

deficit

irrigation

no irrigation

To

tal

rev

enu

e (U

S$

ha

-1)


DP0935

FM9180


161

Figure 7.27. Profit obtained from cultivars DP0395 and FM9180 subjected to eight



US$ 326.00 ha-1

.

The effects of the treatments on the revenues (Fig. 7.26) were very similar to

the effect on profit (Fig.7.27). An average revenue difference of almost US$ 3,000 ha-1

(more than US$ 2,400.00 ha-1

in profit) was observed between the fully irrigated and

dryland water regimes. These were the highest and the lowest revenues and profits and

were statistically different from all other treatments. The effects of mild deficit

irrigation in revenue and profit were not statistically different from the obtained by

irrigation interruptions at the squaring stage and at the peak bloom stage from three

weeks. Similarly, the total revenue and profit obtained with a deficit irrigation regime

would be similar to the observed when there was an episodic water stress at the early

flowering stage and from the peak bloom until the crop termination.

-1500

-1000

-500

0

500

1000

1500

2000

2500

full irrigation mild deficit

irrigation

squaring 3 weeks at

peak bloom

peak bloom

totermination

deficit

irrigation

early

flowering

no irrigation

Pro

fit

(US

$ h

a-1

)


DP0935

FM9180


162

Figure 7.28. Breakeven price for cultivars DP0395 and FM9180 subjected to eight



US$ 0.26 kg-1

.

In 2012, the effect of the water treatments on breakeven price (Fig. 7.28) was

inversely related to the effect observed for revenue (Fig. 7.26) and profit (Fig. 7.27).

The lowest of all treatments breakeven price was observed for the full irrigation and

the highest for the dryland, and they were statistically different from all other

irrigation regimes or interruption periods. Similar to the effects of water in revenue

and profit, the breakeven price effects of mild deficit irrigation were not statistically

different from the obtained by irrigation interruptions at the squaring stage and at the

peak bloom stage from three weeks. Also the breakeven price for a deficit irrigation

regime would be similar to the observed when there was an episodic water stress at the

early flowering stage and from the peak bloom until the crop termination. Cultivar-

irrigation interaction was also observed for breakeven price (p=0.0008), although the

0

1

2

3

4

5

6

full

irrigation

mild deficit

irrigation

squaring 3 weeks at

peak bloom

peak bloom

to

termination

deficit

irrigation

early

flowering

no irrigation

Bre

ak

even

pri

ce

(US

$ k

g-1

)


DP0935

FM9180


163

single main effect of cultivar would indicate a breakeven price for FM9180 higher

than for DP0935, at full irrigation regime the breakeven price for both cultivars would

not be statistically different, averaging nearly US$ 0.93 kg-1

.

Figure 7.29. Profit obtained from cultivars DP0395 and FM9180 subjected to eight



US$ 376.98.

In 2012, considering unlimited area available, even if the area was increased to

use the same amount of water of a fully irrigated hectare (Fig. 7.20) The profits

obtained in a fully irrigated hectare would still be the highest of all water regimes

(Fig.7.29), indicating that is probably more profitable for a farmer, under 2012 similar

environmental conditions, to reduce the crop area but keep it fully irrigated, since the

profit of a fully irrigated hectare was more than a US$1,000.00 superior than the

obtained at an increased area at a mild deficit irrigation regime.

-4000

-3000

-2000

-1000

0

1000

2000

3000

full irrigation mild deficit

irrigation

squaring 3 weeks at

peak bloom

peak bloom

totermination

early

flowering

deficit

irrigation

no irrigation

Eq

ua

l w

ate

r p

rofi

t (U

S$

)


DP0935

FM9180


164

Figure 7.30. Marginal value product to full irrigation of cultivars DP0395 and

FM9180 subjected to seven different irrigation regimes during the 2012 season at

the Texas Tech Quaker Avenue research farm. Black bars represent one least


mm-1

. US$ 1.00 ha-1

mm-1

=

US$ 0.01 10-2

L-1

.

Fig. 7.30 shows that in 2012 the highest increase in revenue per additional

millimeter of irrigation depth applied would be obtained from a crop with irrigation

interrupted at the squaring stage, more than US$13 ha-1

mm-1

. The lowest MVP was

observed for a crop at a mild deficit, and it would not be statistically different from the

MVP for a deficit irrigation nor dryland.

0

2

4

6

8

10

12

14

16

18

mild deficit

irrigation

deficit

irrigation

no irrigation peak bloom to

termination

early

flowering

3 weeks at

peak bloom

squaring

Marg

inal

valu

e p

rod

uct

(U

S$

ha

-1 m

m-1

)

irrigation regime or interruption period

DP0935

FM9180


165

7.6. Discussion

Although the highest seed cotton yield from chapters 3 and 4 was only

achieved with the largest irrigation application, they also provided the highest profits

due to the consequent increase on revenues.

When the 2011 irrigation was interrupted at the squaring stage, plants

compensated the early loss of bolls in the bottom (nodes 5 to 9) with more bolls at the

top (nodes 11 and 12), however, the effect of more bolls at the top didn’t cause a

significantly lower discount based on fiber quality parameters used in the actual

system. When comparing 2012 premiums and bolls distribution, we also did not

observe a positive effect on the reduction on top bolls in the premium received.

The lowest breakeven prices are also desirable since they represent that

farmers who are using that irrigation cultivar selection treatment can potentially make

profits on a low-price scenario where other treatments would not. This study suggested

that the full irrigation, that provided the highest yields and profit, will also be

associated to the desirable lowest breakeven price.

7.7. Summary and Conclusion

For all environments and trials statistically significant (p<0.01) irrigation effect

on seed-cotton yield was observed.


166

A cultivar single main effect in seed-cotton yield was observed for two

environments of each set of trials.

While the seed-cotton irrigation-cultivar interaction was statistically significant

for all environments of the Episodic drought sets of trials, at the Irrigation levels set of

trials it was significant just in 2012.

In 2011 although the highest equal water increased area productions were

observed under the squaring irrigation interruption, they were not statistically different

from the observed under no drought and peak bloom to termination episodic drought.

However, the total profits where still higher when no episodic drought occurred.

In 2011 even if the water saved in a early flowering irrigation interruption was

used to irrigate a bigger area it would still provide a total production reduced in more

than 720 kg when compared with a equal water squaring episodic drought, it would

also mean a reduction of nearly 700 kg when compared with an fully irrigated hectare

yield.

In 2011 the highest revenue and profit were achieved at the no drought

treatment. While an average profit of US$ 740.50 per hectare was observed when

there was no drought, losses were observed when there were irrigation interruptions on

the peak bloom to termination and on early flowering stages (-US$ 1034.1 per hectare

in average).

When there was an episodic drought at early flowering stage in 2011, the

highest break-even price was achieved from cultivars FM9170 and was just not


167

statistically different from FM9180 in the same water regime. However, at the no

drought, squaring and peak bloom irrigation treatments, the break-even price was not

statistically different among cultivars, averaging US$ 1.54 per kg of cotton lint.

Similar to the MPP, in 2011, the MVP was significantly higher from the early

flowering irrigation interruption treatment, indicating that a crop subjected to that

episodic drought would have revenues increased in nearly US$ 7.77 ha-1

mm-1

more

than a crop with the irrigation interrupted at the squaring stage.

In 2012 the seed-cotton yield when a 3 weeks at peak bloom irrigation

interruption occurred was not statistically different of the observed from a squaring

stage episodic drought and also didn’t differ from the yield from a mild deficit

irrigation management.

If the water saved on the several irrigation management strategies tested in

2012 the highest total production would still be achieved under one fully irrigated

hectare, but it was not statistically different from the total production obtained under a

mild deficit irrigation management.

In 2012, the highest increase in yields from the dryland treatment (MPP) would

be achieved following the mild defict irrigation management (nearly 4 ha-1

mm-1

) and

it was not statistically different from the deficit irrigation management. In that

scenario a higher MPP would also be observed on cultivar DP0935, but if the increase

in water would follow the fully irrigated management, there would not be a cultivar

difference. When the MPP was calculated based on the deltas from the proposed


168

treatment to the highest water and yield treatment (full irrigation) the highest MPP was

observed from the squaring episodic drought treatment indicating that this treatment

would be the bigger beneficary of water increments.

2012 effects of the treatments on the revenues were very similar to the effect

on profit. An average revenue difference of almost US$ 3,000 ha-1

(more than US$

2,400.00 ha-1

in profit) was observed between the fully irrigated and dryland water

regimes.

The 2012 effects of mild deficit irrigation in revenue and profit were not

statistically different from those obtained by irrigation interruptions at the squaring

stage and at the peak bloom stage from three weeks. Similarly, the total revenue and

profit obtained with a deficit irrigation regime would be similar to the observed when

there was an episodic water stress at the early flowering stage and from the peak

bloom until the crop termination.

In 2012, even if the area of mild deficit irrigation regime was increased to use

the same amount of water of a fully irrigated hectare the profit of a fully irrigated

hectare would be superior in more than a thousand dollar. Also in 2012, the higher

MVP would be obtained from a crop with irrigation interrupted at the squaring stage,

more than US$13 ha-1

mm-1

.


169

References
























Plains Cotton Growers Inc.

<http://www.plainscotton.org/2012Loan/2012LoanPandD.pdf> Access on

November 16th 2012.



Roseta-Palma, C. 2003. Joint Quantity/Quality Management of Ground Water.

Environmental and Resource Economics 26: 89-106.


http://www.plainscotton.org/2012Loan/2012LoanPandD.pdf


170




West Texas A&M AgriLife Extension Service: Estimated costs and returns per Acre <

http://agecoext.tamu.edu/fileadmin/user_upload/Documents/Budgets/District/2

/2012/cottondihtir.pdf/>. Access on November 14th 2012




http://agecon.tamu.edu/%3e




171

CHAPTER VIII

DISSERTATION CONCLUSION

In most years, yield and most fiber quality parameters presented were

significantly affected by episodic drought, deficit irrigation, or cultivar selection. A

cultivar single main effect in seed-cotton yield was observed for two environments of

each set of trials.

The environments used in this study represented the climatic differences

observed in West Texas. While early in the 2010 we had a higher than normal

precipitation, in 2011 and 2012, in both locations (Quaker and New Deal), we

observed lower than average precipitation. The dry climatic condition observed in

2011 and 2012 is ideal for the development of irrigation management research,

especially for the studies involving episodes of irrigation interruption in different crop

development stages.

Episodic drought events provided bigger reductions on seed-cotton yields than

continuously reducing water depths. Yield and most of the HVI and AFIS parameters

presented were significantly affected by drought episode and cultivar. The data shows

that in all years and locations the irrigation strategy with no irrigation interruption

provided the best yields.

In 2011 and in both 2012 locations, the irrigation interruption at the squaring

stage resulted in the smallest reduction in yield from the fully irrigated treatment. Also


172

in 2011 and in both 2012 locations, when the irrigation interruption occurred during

the early flowering treatment, the highest yield reduction was observed.

Full continuous irrigation provided highest yields and AWUE than the

observed on deficit irrigation or episodic drought regimes. AWUE was affected by

varying irrigation levels in two out of three years. In 2010 the highest AWUE occurred

under deficit irrigation, not statistically different from full irrigation, and in 2012 the

AWUE observed on fully irrigated cultivars was statistically higher than the one

observed under deficit irrigation and dryland. At least in one year, all HVI fiber

quality parameters were affected by varying irrigation level or correspondent cultivar

selection on a diverse set of cultivars.

Yields were particularly sensitive to the photosynthesis reduction caused by

water deficit at early flowering. Differences in bolls distribution and gas exchange

were used to explaing the treatment effects on cotton yield.

Irrigation interruption periods resulted in differences on the average number of

bolls on nodes 8 to 13, these differences had consequences on seed-cotton yield. In

both years cultivar FM9180 presented slightly less bolls than the other cultivars from

nodes 8 to 15 and FM9170 presented more bolls on nodes 9 and 10 than the other

cultivars.

Photosynthesis and other physiological related parameters were affected by

irrigation interruption episodes in the three years of this study. Yields were sensitive to

the photosynthesis reduction caused by water deficit at early flowering and peak


173

bloom stages. Episodic drought periods can affect leaf-level gas exchange and impact

yield. We also found fairly good relations of photosynthesis with transpiration,

stomatal conductance, air-leaf temperature difference, and internal-ambient carbon

dioxide concentration ratio.

In different years, the cultivar and irrigation the response to variables from this

study varied indicating that their effect is also affected by the year to year

environmental differences.

Data suggested that full irrigation might possibly provide higher total

production and profits than an increased area under deficit irrigation treatments using

the same amount of water.

The economic analysis provided bases to indicate the more profitable cultivar

selection and irrigation management combination. 2012 effects of the treatments on

the revenues were very similar to the effect on profit. An average revenue difference

of almost US$ 3,000 ha-1

(more than US$ 2,400.00 ha-1

in profit) was suggested

between the fully irrigated and dryland water regimes.

Effects of mild deficit irrigation in revenue and profit were not statistically

different from the obtained by irrigation interruptions at the squaring stage and at the

peak bloom stage from three weeks. Similarly, the total revenue and profit obtained

with a deficit irrigation regime would be similar to the observed when there was an

episodic water stress at the early flowering stage and from the peak bloom until the

crop termination.


174

In 2012, even if the area of mild deficit irrigation regime was increased to use

the same amount of water of a fully irrigated hectare, the profit of a fully irrigated

hectare would possibly be superior in more than a thousand dollar. 2012 data also

suggested that the higher MVP would be obtained from a crop with irrigation

interrupted at the squaring stage, nearly US$13 ha-1

mm-1

.

This dissertation presented a broad set of evaluations and our studies were

repeated in several environments, however, we can indicate the need of some future

cotton irrigation management studies. Since this research data was based on small

plots, composed of two 12.2 m rows, ginned with a table-top gin, we recommend

using the methodology described on Chapter 7 with industrial scale data for more

accurate analysis of the production economics parameters.

Further research can also include the study of episodic drought effects on the

establishment stage, diverse levels of water stress in each stage, and smaller periods of

irrigation interruption. We can also suggest repeating the trials described in this

dissertation including other cultivars such as Phytogen cultivars, different Stoneville

cultivars, and different newer cultivars, especially DP1044B2RF and FM2011GT that

are recently in use for a significant amount of farmers in West Texas.


175

BIBLIOGRAPHY




Allen, R.G., Pereira L.S., Raes D., Smith M. (1998) Crop evapotranspiration:

Guidelines for computing crop water requirements. (FAO Irrigation an

drainaige paper 56)



Baker, J.T.; D.C. Gitz, P.Payton, D.F. Wanjura, and D.R. Upchurch. 2007. Using

Leaf Gas Exchange to Quantify Drought in Cotton Irrigated Based on Canopy

Temperature Measurements. Agron. J. 99:637–644.

doi:10.2134/agronj2006.0062

Baker, J.T.; S.V. Scott, D.C. Gitz, P. Payton, R.J. Lascano and B. McMichael. 2009.

Canopy Gas Exchange Measurements of Cotton in an Open System. Agron. J.

101:52–59. doi:10.2134/agronj2008.0007x

Balkcom, K.S., Reeves D.W., Shaw J.N., Burmester C.H., Curtis L.M. (2006) Cotton

Yield and Fiber Quality from Irrigated Tillage Systems in the Tennessee

Valley. Agronomy Journal 98:596-602.

Barreto, A.N., do Amaral J.A.B., da Silva e Luz M.J. (2007) Consumo Hidrico do

Algodoeiro Herbaceo nas Diferentes Fases Fenologogicas no municipio Irece -

BA, VI Congresso Brasileiro do Algodão, Uberlandia. pp. 4.




037X.2008.00340.x




Bednarz, C.W., Nichols R.L. (2005) Phenological and Morphological Components of

Cotton Crop Maturity. Crop Science 45:1497–1503

DOI:10.2135/cropsci2004.0321


176



10.2135/cropsci2005.08-0276.

Bednarz, C.W., Hook J., Yager R., Cromer S., Cook D., Griner I. (2003) Crop Water

Use and Irrigation Scheduling, in: A. S. e. a. Culpepper (Ed.), Cotton

Research-Extension Report UGA/CPES Research-Extension Publication,

UGA, Georgia. pp. 61-64.

Bernardo, S., Soares A.A., Mantovani E.C. (2006) Manual de Irrigacao. 8 ed. Editora

UFV.




Bronson, K.F., Onken A.B., Keeling J.W., Torbert. H.A. (2001) Nitrogen Response in

Cotton as Affected by Tillage System and Irrigation Level. Soil Society of

America Journal 65:1153-1163.

Bronson, K.F., Brooker J.D., Bordovsky J.P., J.W. K., Wheeler T.A., Boman R.K.,

Parajulee M.N., Segarra E., Nichols R.L. (2006) Site-Specific Irrigation and

Nitrogen Management for Cotton Producion in the Southern High Plains.











Collins, G. and K. Hake (2012) Management Considerations for Irrigated Cotton, in:



Dagdelen, N., Basal H., Yilmaz E., Gurbuz T., Akcay S. (2009) Different drip


western Turkey. Agricultural Water Management Journal 96:111-120.


177

da Costa, V.A.; and J.T. Cothren. 2011. Drought Effects on Gas Exchange,

Chlorophyll, and Plant Growth of 1-Methylcyclopropene Treated Cotton.

Agron. J. 103:1230-1241.

de Sousa, P.S., de Medeiros J.F., de Matos J.d.A., de Melo S.B., Ferreira R.d.C.

(2008) Efeito de laminas de irrigacao sobre o crescimento do algodoeiro

herbaceo. Revista Verde 3:06-11.








Fritschi, F. B.; B.A. Roberts; D.W. Rains; R.L. Travis and, R.B. Hutmacher. 2004.

Fate of Cof Nitrogen-15 Applied to Irrigated Acala and Pima Cotton. Agron. J.

96:646-655.

Fryxell, P.A. (1986) Ecological adaptations of gossypium species, in: J. Mauney and J.

M. Stewart (Eds.), Cotton Physiology, The Cotton Foundation.

Hake, K.D., Grimes D.W. (2010) Crop water management to optimize growth and

yield, in: J. M. Stewart, et al. (Eds.), Physiology of Cotton, Springer.


Full-, Deficit-Irrigated, and Dryland Cotton on the Northern Texas High

Plains. Journal of Irrigation and Drainage Engineering 130:277-285. DOI:

10.1061/(ASCE)0733-9437(2004)130:4(277).

Jackson, L.E.B., Tilt P.A. (1986) Effects of irrigation intensity and nitrogen level on

the performance of eight varieties of upland cotton, (Gossypium hirsutum L.).

Agronomy Journal 60:p. 13-17.

Jordan, R.W. (1986) Water deficits and reproduction, in: J. Mauney and J. M. D.

Stewart (Eds.), Cotton Physiology, The Cotton Foundation.

Keller, J., Bliesner R.D. (1990) Sprinkle and Tricke Irrigation Chapman & Hall.


Academic Press.

Kramer, P.J., Boyer J.S. (1995) Water relations of plants and soils Elsevier Science.


178






Lanza, M.A., Penna J.C.V. (2007) Algodão (Gossypium hirsutum L.), in: J. T. de

Paula Júnior and M. Venzon (Eds.), 101 culturas: manual de tecnologias

agrícolas, Epamig, Belo Horizonte, MG - Brazil. pp. 63-74.

Lascano, R.J. (2000) A general system to measure and calculate daily crop water use.




Martin, E.C., Obermyer K.C., Pegelow E.J., Watson J. (1999) Using drainage

lisimeters to evaluate irrigation and N interactions in cotton production. Cotton

report:204-210.





Inc. pp. 279-335.

Mathis, G. (2009) Irrigation Response in Cotton to Optimize Yield, Quality and

Profitability in the Texas High Plains, Plant and Soil Science, Texas Tech

University, Lubbock, TX. pp. 105. (Master of Science Thesis)




Mills, C.I., Bednarz C.W., Ritchie, G.L., Whitaker, J.R. (2008) Yield, Quality, and

Fruit Distribution in Bollgard/Roundup Ready and Bollgard II/Roundup Ready

Flex Cottons. Agron.J. 100:35-41. DOI:10.2134/agronj2006.0299




Mullins, G.L., Burmester C.H. (2010) Relation of Growth and Development to

Mineral Nutrition, in: J. M. Stewart, et al. (Eds.), Physiology of Cotton,

Springer.


179





2:295-299.

Pettigrew, W.T. (2004) Physiological Consequences of Moisture Deficit Stress in

Cotton. Crop Sci.. 44:1265-1272.



Science 16:42-52.

Plains Cotton Growers Inc.

<http://www.plainscotton.org/2012Loan/2012LoanPandD.pdf> Access on

November 16th 2012.

Rajan, N. (2007) Estimation of Crop Water Use for Different Cropping Systems in the

Texas High Plains Using Remote Sensing, Plant and Soil Science, Texas Tech

University, Lubbock - TX. pp. 175. (Doctoral Dissertation)




Ritchie, G.L., Whitaker J.R., and Collins G.D. 2011. Effect of Sample Size on Cotton

Plant Mapping Analysis and Results. J. Cotton Sci. 15:224-232



Roseta-Palma, C. 2003. Joint Quantity/Quality Management of Ground Water.

Environmental and Resource Economics 26: 89-106.



Report:31-46.







http://www.plainscotton.org/2012Loan/2012LoanPandD.pdf


180

Stiller, W. N.; P.E. Reid and; G.A. Constable. 2004. Maturity and Leaf Shape as Traits

Influencing Cotton Cultivar Adaptation to Dryland Conditions. Agron. J.

96:656-664.

Tolk, J.A., and Howell T.A (2010) Cotton Water Use and Lint Yield in Four Great

Plains Soils. Agronomy Journal 102:904-91.





f>, USDA.

West Texas A&M AgriLife Extension Service: Estimated costs and returns per Acre <

http://agecoext.tamu.edu/fileadmin/user_upload/Documents/Budgets/District/2

/2012/cottondihtir.pdf/>. Access on November 14th 2012






J. 100:1763-1768.







181

APPENDICES


182

APPENDIX A

WEATHER PARAMETERS

Figure A.1. Daily ET0 measured during the 2010 season at the Texas Tech University

New Deal research farm.



0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

25-May-10 25-Jun-10 25-Jul-10 25-Aug-10 25-Sep-10 25-Oct-10

Da

ily

ET

o (

mm

)

Day

0

2

4

6

8

10

12

14

5/18/2011 6/18/2011 7/18/2011 8/18/2011 9/18/2011 10/18/2011

Da

ily

ET

o (

mm

)

Day


183


Quaker Avenue research farm.



0

2

4

6

8

10

12

5/13/2011 6/13/2011 7/13/2011 8/13/2011 9/13/2011 10/13/2011

Da

ily

ET

o (

mm

)

Day

0

2

4

6

8

10

12

5/24/2012 6/24/2012 7/24/2012 8/24/2012 9/24/2012 10/24/2012

Da

ily

ET

o (

mm

)

Day


184



Figure A.6. Cumulative Degree Days for a 60 ºF base temperature at the Texas Tech

University research farms.

0

1

2

3

4

5

6

7

8

9

10

5/16/2012 6/16/2012 7/16/2012 8/16/2012 9/16/2012

Da

ily

ET

o (

mm

)

Day

0

500

1000

1500

2000

2500

3000

3500

1 8

15

22

29

36

43

50

57

64

71

78

85

92

99

106

113

120

127

134

141

148

155

162

169

176

Cu

mu

lati

ve

DD

60

(ºF

)

Days after planting

Normals

ND 2010

ND 2011

Qk 2011

ND 2012

Qk 2012


185

Figure A.7. Cumulative Precipitation at the Texas Tech University research farms.

Figure A.8. Daily average temperature measured during the 2010 season at the Texas

Tech University New Deal research farm.

0

50

100

150

200

250

300

0 5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

105

110

115

120

Cu

mu

lati

ve

pre

cip

ita

tio

n (

mm

)

Days after planting

Normals

2010 ND

2011 ND

2011 Qk

2012 ND

2012 Qk

0

5

10

15

20

25

30

35

25-May-10 25-Jun-10 25-Jul-10 25-Aug-10 25-Sep-10 25-Oct-10

Tem

per

atu

re (

ºC)

Day


186

APPENDIX B

SOIL MOISTURE

Figure B.1. Soil moisture content in treatments subjected to six (bottom) three

(middle) or zero (top) weeks of irrigation interruption, as a function of days after

planting, measured on the cultivar FM9180 during the 2010 season at the Texas Tech

University New Deal research farm.

Depth

(cm

)

-120

-100

-80

-60

-40

-20

16

18

20

22

24

Depth

(cm

)

-120

-100

-80

-60

-40

-20

16

18

20

22

24

Days after Planting

70 80 90 100 110 120

Depth

(cm

)

-120

-100

-80

-60

-40

-20

16

18

20

22

24


187

APPENDIX C

TABLE OF AGROCHEMICALS APPLIED

Trade name Active Ingredient

Herbicide

Roundup glyphosate N-(phosphonomethyl)glycine

Plant growth regulator

Stance cyclanilide and mepiquat

Insecticide

Temik aldicarb [2-methyl-2-(methylthio)

propionaldehyde O-

(methylcarbamoyl)oxime]


188

APPENDIX D

IRRIGATED COTTON BUDGET

Table D.1. Estimated costs and returns per Acre Cotton, Drip Irrigated - Herbicide-

tolerant, Insect-resistant 2012 Projected Costs and Returns per Acre (Source: Texas

A&M University Department of Agricultural Economics, 2012)


dollars

dollars

INCOME

cotton lint lb. $ 0.80 1500.0000 $ 1,200.00

cottonseed ton $ 200.00 1.0500 $ 210.00

--------

TOTAL INCOME

$ 1,410.00

DIRECT EXPENSES

SEED

seed - cotton thou $ 1.30 52.0000 $ 67.60

FERTILIZER

fert. (P) lb. $ 0.70 25.0000 $ 17.50

fert. (N) lb. $ 0.70 150.0000 $ 105.00

CUSTOM

fert appl. acre $ 5.00 1.0000 $ 5.00

preplant herb. appl acre $ 4.50 1.0000 $ 4.50

post emerg herb+appl acre $ 16.00 1.0000 $ 16.00

insec+appl - cotton appl $ 12.00 1.0000 $ 12.00

harvaid appl-cot dri acre $ 30.00 1.0000 $ 30.00

strip & module-cotto lb. $ 0.08 1500.0000 $ 120.00

ginning - cotton cwt. $ 3.00 53.5700 $ 160.71

CROP INSURANCE

cotton - CP acre $ 30.00 1.0000 $ 30.00

BOLL WEEVIL ASSESS

Irrigated acre $ 1.00 1.0000 $ 1.00

OPERATOR LABOR

Implements hour $ 10.00 0.6358 $ 6.35

Tractors hour $ 10.00 0.7322 $ 7.32


189

Continuation of Table D1


HAND LABOR

Implements hour $ 10.00 0.1908 $ 1.90

IRRIGATION LABOR

Drip Irrigated hour $ 10.00 1.0880 $ 10.88

DIESEL FUEL

Tractors gal $ 3.20 3.2862 $ 10.51

GASOLINE

Pickup gal $ 3.20 2.0100 $ 6.43

IRRIGATION FUEL

Drip Irrigated ac-in $ 9.00 17.0000 $ 153.00

REPAIR & MAINTENANCE

Implements Acre $ 4.08 1.0000 $ 4.08

Tractors Acre $ 5.46 1.0000 $ 5.46

Pickup Acre $ 1.00 1.0000 $ 1.00

Drip Irrigated ac-in $ 2.25 17.0000 $ 38.25

INTEREST ON OP. CAP. Acre $ 18.52 1.0000 $ 18.52

--------

TOTAL DIRECT EXPENSES

$ 833.06

RETURNS ABOVE DIRECT EXPENSES

$ 576.93

FIXED EXPENSES

Implements Acre $ 9.84 1.0000 $ 9.84

Tractors Acre $ 11.10 1.0000 $ 11.10

Pickup Acre $ 1.25 1.0000 $ 1.25

Drip Irrigated Acre $ 60.00 1.0000 $ 60.00

--------

TOTAL FIXED EXPENSES

$ 82.20

--------

TOTAL SPECIFIED EXPENSES

$ 915.27

RETURNS ABOVE TOTAL SPECIFIED EXPENSES $ 494.72

ALLOCATED COST ITEMS

cash rent - cottondr acre $ 100.00 1.0000 $ 100.00

RESIDUAL RETURNS

$ 394.72


190

APPENDIX E

ADDITIONAL MEASUREMENTS DATA TABLES AND GRAPHS

E.1. Plant grotht

Figure E.1. Average plant height of eight cotton cultivars subjected to different

irrigation interruptions during the 2010 season at the Texas Tech New Deal research

farm. Black bars represent standard errors.

0

20

40

60

80

100

120

3 weeks at peak bloom no drought peak bloom to

termination

Av

era

ge

pla

nt

hei

gth

(cm

)

Period of irrigation interruption

DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180


191

Figure E.2. Average number of nodes from eight cotton cultivars subjected to different



Figure E.3. Average plant height of 20 cultivars subjected to three different irrigation

levels during the 2010 season. Black bars represent standard errors.

15

16

17

18

19

20

21

22

23

24

25


termination

Av

era

ge

nu

mb

er o

f n

od

es p

er p

lan

t


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180

70

75

80

85

90

95

100

352 443 537

Pla

nt

hei

gh

t (c

m)

Total water (mm)

1031518

1060799

1060894

04M082

06TC065

ACALA1517_95

ACALA1517_E2

CHUNG_MEIN_SUE#7

FM832

FM989

IV4F_91057

M7646

MAC13

MAC95


192

Figure E.4. Average number of nodes from 20 cultivars subjected to three different

irrigation levels during the 2010 season. Black bars represent standard errors.

12

13

14

15

16

17

18

19

20

352 443 537

Nu

mb

er o

f n

od

es

Total water (mm)

1031518

1060799

1060894

04M082

06TC065

ACALA1517_95

ACALA1517_E2

CHUNG_MEIN_SUE#7

FM832

FM989

IV4F_91057

M7646

MAC13

MAC95

PM303

SIOKRA24

ST468

ST474

ST506

TAMCOTCD3H


193

Figure E.5. Height averaged from five cotton cultivars subjected to varying irrigation

regimes during the 2011 season at the Texas Tech New Deal research farm.

Figure E.6. Number of nodes averaged from five cotton cultivars subjected to varying

irrigation regimes during the 2011 season at the Texas Tech New Deal research farm.

0

10

20

30

40

50

60

70

80

37 56 69

Av

era

ge

pla

nt

hei

gh

t (c

m)

Days after planting

0

33

66

99

0

2

4

6

8

10

12

14

16

18

20

37 56 69

Av

era

ge

nu

mb

er o

f n

od

es

Days after planting

0

33

66

99


194

Figure E.7. Heights from five cotton cultivars averaged from four varying irrigation

regimes during the 2011 season at the Texas Tech New Deal research farm.

Figure E.8. Number of nodes from five cotton cultivars averaged from four varying

irrigation regimes during the 2011 season at the Texas Tech New Deal research farm.

0

10

20

30

40

50

60

70

80

37 56 69

Av

era

ge

pla

nt

hei

gh

t

Days after planting

ACALA 1517-E2

FM832

FM9180

Siokra24

ST506

0

2

4

6

8

10

12

14

16

18

20

37 56 69

Av

era

ge

nu

mb

er o

f n

od

es p

er p

lan

t

Days after planting

ACALA 1517-E2

FM832

FM9180

Siokra24

ST506


195

E.2. Total Water

Figure E.9. Total water per treatment during the 2010 season in the Texas Tech New

Deal research farm.

Figure E.10. Total water per treatment during the 2011 season in the Texas Tech


270

162 56

0

100

200

300

400

500

600

700

no drought 3 weeks at peak

bloom

peak bloom to

termination

Irrig

ati

on

+ P

reci

pit

ati

on

(m

m)


irrigation

precipitation

330 259 250 268

231

0

50

100

150

200

250

300

350

400

450


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Irrig

ati

on

+ P

reci

pit

ati

on

(m

m)


irrigation

precipitation


196

Figure E.11. Total water per treatment during the 2012 season in the Texas Tech



Deal research farm.

598

481 376

445 350

0

100

200

300

400

500

600

700


flowering

3 weeks at

peak

bloom

peak

bloom to

termination

Irrig

ati

on

+ P

reci

pit

ati

on

(m

m)


Irrigation

Precipitation

554 509

422 457 423

0

100

200

300

400

500

600

700

800


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Irrig

ati

on

+ P

reci

pit

ati

on

(m

m)


Irrigation

Precipitation


197


Deal research farm.


Deal research farm.

27

118

212

0

100

200

300

400

500

600

No irrigation from

squaring

Deficit irrigation Full irrigation

Irrig

ati

on

+ p

recip

ita

tio

n (

mm

)

Irrigation treatment

Irrigation

Precipitation

355

473 527

638

0

100

200

300

400

500

600

700

800

No irrigation

from

squaring

Deficit

irrigation

Mild deficit

irrigation

Full irrigation

Irrig

ati

on

+ p

recip

ita

tio

n (

mm

)


Irrigation

Precipitation


198

Table E.1. Total precipitation and irrigation applied per season and location (four

irrigation levels analysis)

Environment Irrigation Treatment Precipitation

(mm)

Irrigation

(mm)

Total

Water

(mm)





Mild deficit irrigation 39 527 566



2012 - Qk Full irrigation 46 583 629

Mild deficit irrigation 46 352 398


No irrigation from squaring 46 173 219 ND = Texas Tech New Deal Research Farm.


E.3. Yields

Figure E.15. Lint yield of eight cotton varieties subjected to three different irrigation

interruptions periods during the 2010 season. Black bars represent the standard error.

0

200

400

600

800

1000

1200

1400

1600

no drought 3 weeks at peak bloom peak bloom to

termination

Lin

t y

ield

(k

g h

a-1

)


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180


199

Figure E.16. Lint yield of four cotton cultivars subjected to five different irrigation

interruptions periods during the 2011 season. Black bars represent the standard errors.

Figure E.17. Yield of four cotton cultivars subjected to five different irrigation


0

500

1000

1500

2000

2500


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Lin

t y

ield

(k

g h

a-1

)


DP0912

DP0935

FM9170

FM9180

1000

1500

2000

2500

3000

3500

4000

4500

5000


flowering

3 weeks at

peak bloom

peak bloom

to

termination

See

d-c

ott

on

yie

ld (

kg

ha

-1)


DP0912

DP0935

FM9170

FM9180


200

Table E.2. Statistical significance (p-values) of the study factors and interaction in the

variable cotton seed yield during the 2010, 2011 and 2012 seasons. 20 cultivars in

2010, 5 in 2011 and 6 in 2012, and also three irrigation levels in 2010, and four

irrigation levels in 2011 and 2012.


Irrigation <0.0001 ** < 0.0001 ** < 0.0001 **

Cultivars <0.0001 ** 0.1071 n.s. < 0.0001 **

Irrigation*Cultivar 0.1297 n.s. 0.5667 n.s. 0.0006 **

Figure E.18. Yields of 20 cultivars subjected to three different irrigation levels during

the 2010 season. Black bars represent standard errors.

0

1000

2000

3000

4000

5000

6000

7000

8000

Irrigation interrupted

from squaring

Deficit irrigated Fully irrigated

See

d-c

ott

on

yie

ld (

kg

ha

-1)


1031518

1060799

1060894

04M082

06TC065

ACALA1517_95

ACALA1517_E2

CHUNG_MEIN_SUE#7

FM832

FM989

IV4F_91057

M7646

MAC13

MAC95

PM303

SIOKRA24

ST468

ST474

ST506

TAMCOTCD3H


201

Figure E.19. Seed-cotton yields of five cultivars subjected to four different irrigation

levels during the 2011 season. Black bars represent standard errors.

Figure E.20. Average seed cotton yield of two cotton cultivars subjected to five

different irrigation interruptions periods during the 2011 season. Different letters

indicate LSD differences at p=0.05.

0

1000

2000

3000

4000

5000

6000

Irrigation

interrupted from

squaring

Deficit irrigated Mild deficit Fully irrigated

See

d-c

ott

on

yie

ld (

kg

ha

-1)


Acala1517-E2

FM832

FM9180

Siokra24

ST506

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

3 weeks at peak

bloom

early flowering no drought peak bloom to

termination

squaring

See

d-c

ott

on y

ield

(kg h

a-1)

Irrigation interruption stage

a

b c c

d


202

E.4. Fiber Quality

Table E.3. Significance of the analysis of variance for HVI variables as affected by the

irrigation episodic drought, cultivar or their interaction in 2011


Factor Irrigation Cultivar Irr*Cult Interaction

Micronaire ** ** **

Length ** ** n.s.

Uniformity ratio ** ** n.s.

Bundle strength ** ** n.s.

Elongation * ** n.s.

Color Rd n.s. ** n.s.

Color +b * ** n.s.

Leaf n.s. n.s. n.s.


203

Figure E.21. Micronaire of eight cotton cultivars subjected to different irrigation

interruptions during the 2010 season at the Texas Tech New Deal research farm. Black

bars represent standard errors.

Figure E.22. Micronaire of four cotton varieties subjected to five different irrigation


3.7

3.9

4.1

4.3

4.5

4.7

4.9

5.1

5.3

5.5


termination

Mic

ron

air

e


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180

2.5

3

3.5

4

4.5

5

5.5


flowering

3 weeks at

peak bloom

peak bloom

to

termination

Mic

ron

air

e


DP0912

DP0935

FM9170

FM9180


204

Figure E.23. HVI UHML length of eight cotton cultivars subjected to different



Figure E.24. HVI UHML length of four cotton cultivars subjected to different

irrigation interruptions during the 2011 season at the Texas Tech Quaker Avenue

research farm. Black bars represent standard errors.

0.9

0.95

1

1.05

1.1

1.15

1.2

1.25


termination

HV

I U

HM

L l

eng

gth

(in

)


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180

0.9

0.95

1

1.05

1.1

1.15

1.2


flowering

3 weeks at

peak bloom

peak bloom

to

termination

HV

I U

HM

L l

eng

tth

(in

)


DP0912

DP0935

FM9170

FM9180


205

Figure E.25. Micronaire of five cultivars subjected to four different irrigation levels

during the 2011 season. Black bars represent standard errors.

Figure E.26. Maturity ratio of five cultivars subjected to four different irrigation levels

during the 2011 season. Black bars represent standard errors.

3

3.2

3.4

3.6

3.8

4

4.2

4.4

4.6

4.8

5

0 33 66 99

Mic

ronai

re

Irrigation (%)

ACALA 1517-E2

FM832

FM9180

Siokra24

ST506

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0 33 66 99

Mat

uri

ty r

atio

Irrigation (%)

ACALA 1517-E2

FM832

FM9180

Siokra24

ST506


206

Table E.4. significance of the analysis of variance for AFIS variables as affected by

the irrigation episodic drougth, cultivar or their interaction in the 2010 season. Eight

cultivars analysis


Factor Irrigation Cultivar Irr*Cult Int.

Neps per gram n.s. ** n.s.

Length per weight * ** n.s.

Upper Quartile Length * ** *

Short Fiber Content by weight n.s. ** n.s.

Longest 5% Fibers by number ** ** **

Fineness * ** n.s.

Standard Fineness ** ** n.s.

Immature Fiber Content (%) * ** n.s.

Maturity Ratio * ** n.s.

Table E.5. significance of the analysis of variance for AFIS variables as affected by

the irrigation episodic drougth, cultivar or their interaction in the 2011 season


Factor Irrigation Cultivar Irr*Cult Int.

Neps per gram ** ** *

Nep size ** n.s. n.s.

Length per weight ** ** n.s.

Upper Quartile Length ** ** n.s.

Short Fiber Content by weight ** ** *

Longest 5% Fibers by number ** ** n.s.

Fineness ** ** n.s.

Standard Fineness ** ** n.s.

Immature Fiber Content (%) ** ** **

Maturity ratio ** ** **


207

Figure E.27. Fineness of eight cotton cultivars subjected to different irrigation

interruptions during the 2010 season at the Texas Tech New Deal research farm. Black

bars represent the standard errors.

Figure E.28. Maturity ratios of eight cotton varieties subjected to different irrigation

interruption periods during the 2010 season. Black bars represent the standard errors.

135

145

155

165

175

185

195


termination

Fin

enes

s (m

Tex

)


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180

0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1


termination

Ma

turi

ty r

ati

o (

%)


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180


208

Figure E.29. Fineness of four cotton cultivars subjected to different irrigation

interruptions during the 2011 season at the Texas Tech New Deal research farm.

Figure E.30. Maturity ratios of four cotton varieties subjected to five different

irrigation interruption periods during the 2011 season.

135

145

155

165

175

185

195

3 weeks at

peak bloom

early

flowering

no drought peak bloom to

termination

squaring

Fin

enes

s (m

Tex

)


DP0912

DP0935

FM9170

FM9180

0.8

0.82

0.84

0.86

0.88

0.9

0.92

0.94

0.96

3 weeks at

peak bloom

early

flowering

no drought peak bloom to

termination

squaring

Ma

turi

ty r

ati

o

Episodic drought period

DP0912

DP0935

FM9170

FM9180


209

E.5. AWUE

Figure E.31. Agronomic Water Use Efficiency from 20 cultivars subjected to three

different irrigation levels during the 2010 season. Black bars represent standard errors.

0

2

4

6

8

10

12

14

Irrigation interrupted

from squaring

Deficit irrigated Fully irrigated

Ag

ron

om

ica

l W

ate

r U

se E

ffic

ien

cy (

kg

ha

-1 m

m-1

)


1031518

1060799

1060894

04M082

06TC065

ACALA1517_95

ACALA1517_E2

CHUNG_MEIN_SUE#7

FM832

FM989

IV4F_91057

M7646

MAC13

MAC95

PM303

SIOKRA24

ST468

ST474

ST506

TAMCOTCD3H


210

E.6. Bolls Distribution

Figure E.32. Smoothed Bolls distribution averaged on five cultivars subjected to four

different irrigation levels during the 2011 season.

Figure E.33. Smoothed Bolls distribution of five cultivars subjected averaged over

four different irrigation levels during the 2011 season.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Aver

age

num

ber

of

boll

s

Node

0

33

66

99

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Aver

age

num

ber

of

boll

s

Node

ACALA 1517-E2

FM832

FM9180

Siokra24

ST506


211

E.7. Gas Exchange

Figure E.34. Average Photosynthesis of cultivars FM9180 and FM832 subjected to

four different irrigation levels during the 2011 season. Black bars represent standard

errors.

Figure E.35. Photosynthesis averaged on the FM9180 and DP0935 cultivars submitted

to different irrigation interruptions periods during the 2012 season. Black bars

represent standard errors.

0

5

10

15

20

25

30

35

40

45

50 60 70 80 90 100 110 120

Pho

sosy

nth

esis

(µ

mol

CO

2 m

-2 s

-1)

Days after planting

0

33

66

99

0

5

10

15

20

25

30

35

40

42 51 58 65 72 75 82 91 98

Pho

tosy

nth

esis

(µ

mo

l C

O2 m

-2 s

-1)

Days after planting

no drought

squaring

early flowering

3 weeks at peak bloom

peak bloom to termination


212

Figure E.36. Average Photosynthesis of cultivars FM832, FM9180, and DP0935

submitted to four different irrigation levels during the 2012 season. Black bars

represent standard errors.

Figure E.37. Average Photosynthesis of cultivar FM9180 subjected to four different

irrigation levels during the 2012 season. Black bars represent standard errors.

0

5

10

15

20

25

30

35

40

47 64 76 90 96

Pho

tosy

nth

esis

(µ

mo

l C

O2 m

-2 s

-1)

Days after planting

full irrigation


deficit irrigation

no irrigation

0

5

10

15

20

25

30

35

40

47 64 76 90 96

Pho

tosy

nth

esis

(µ

mo

l C

O2 m

-2 s

-1)

Days after planting

full irrigation


deficit irrigation

no irrigation


213

Figure E.38. Photosynthesis measured on two cultivars subjected to five different

irrigation interruptions periods during the 2011 season 110 days after planting. Black

bars represent the standard deviation.

0

5

10

15

20

25

30

35

40

early flowering peak bloom to

termination

3 weeks at

peak bloom

no drought squaring

Photo

synth

esis

(

mol

CO

2 m

-2 s

-1)


DP0935B2RF

FM9180B2F

Cultivar


214

E.8. Economics

Figure E.39. Profitability of eight cotton varieties subjected to three different irrigation

interruptions periods during the 2010 season. Black bars represent standard errors.

Figure E.40. Breakeven price of eight cotton varieties subjected to three different

irrigation interruptions periods during the 2010 season. Black bars represent standard

errors.

-600

-400

-200

0

200

400

3 weeks at peak bloom No drought Peak bloom to

termination

Pro

fit

(US

$/a

c)

Drought period

DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

3 weeks at peak bloom No drought Peak bloom to

termination

Bre

ak

-ev

en p

rice

(U

S$

/lb

)


DP0912

DP0924

DP0935

DP09R555

DP1028

FM1880

FM9170

FM9180