Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
4.3. Materials and Methods ....................................................................................... 60
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
5.3. Materials and Methods ....................................................................................... 80
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
6.3. Materials and Methods ....................................................................................... 94
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
affected by the irrigation episodic drought, cultivar or their interaction ..................... 47
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
xi
4.6. Significance of the analysis of variance for HVI variables as
affected by the irrigation episodic drought, cultivar or their interaction ..................... 69
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
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. ...................................................... 142
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...................................................................................................... 150
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. ........................................................................... 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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
. ........................................................ 136
7.3. Lint percentage 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) = 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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) = 1.09 kg ha-1
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
subjected to five different irrigation interruptions periods during the
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
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
. ............................................... 146
7.14. Breakeven price 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$ 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
xvi
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
. ................................................ 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
2012 season at the Texas Tech Quaker Avenue research farm. Black
bars represent one least significant difference. LSD(0.05) = 0.58 kg ha-1
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
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) = 0.37 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
xvii
Tech Quaker Avenue research farm. Black bars represent one least
significant difference. LSD (p=0.05) = US$ 0.0099 kg-1
........................................... 158
7.25. Average boll distribution per node from cultivars DP0395 and
FM9180 subjected to eight different irrigation regimes during the 2012
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
Tech Quaker Avenue research farm. Black bars represent one least
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
Tech Quaker Avenue research farm. Black bars represent one least
significant difference. LSD (p=0.05) = US$ 326.00 ha-1
. .......................................... 161
7.28. Breakeven price for 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$ 0.26 kg-1
............................................... 162
7.29. Profit obtained 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 (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
season at the Texas Tech Quaker Avenue research farm. Black bars
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
A.2. Daily ET0 measured during the 2011 season at the Texas Tech
University New Deal research farm. .......................................................................... 182
A.3. Daily ET0 measured during the 2011 season at the Texas Tech
University Quaker Avenue research farm. ................................................................. 183
A.4. Daily ET0 measured during the 2012 season at the Texas Tech
University New Deal research farm. .......................................................................... 183
A.5. Daily ET0 measured during the 2012 season at the Texas Tech
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
E.10. Total water per treatment during the 2011 season in the Texas
Tech Quaker Avenue research farm. .......................................................................... 195
Texas Tech University, Fulvio Rodriguez Simao, May 2013
xix
E.11. Total water per treatment during the 2012 season in the Texas
Tech Quaker Avenue research farm. .......................................................................... 196
E.12. Total water per treatment during the 2012 season in the Texas
Tech New Deal research farm. ................................................................................... 196
E.13. Total water per treatment during the 2010 season in the Texas
Tech New Deal research farm. ................................................................................... 197
E.14. Total water per treatment during the 2011 season in the Texas
Tech New Deal research farm. ................................................................................... 197
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. ....................................................................................... 198
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. ..................................................................................... 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
irrigation levels during the 2011 season. Black bars represent standard
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
irrigation interruptions during the 2010 season at the Texas Tech New
Deal research farm. Black bars represent standard errors. ......................................... 203
E.22. Micronaire of four cotton varieties subjected to five different
irrigation interruptions periods during the 2011 season. Black bars
represent the standard errors. ..................................................................................... 203
E.23. HVI UHML length 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. ......................................... 204
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
irrigation levels during the 2011 season. Black bars represent standard
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
represent the standard errors. ..................................................................................... 207
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
irrigation interruptions periods during the 2010 season. Black bars
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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,
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
21
order to support the development of water management strategies for irrigated cotton
in West Texas.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
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.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
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.
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.
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.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
24
Lascano, R.J. (2000) A general system to measure and calculate daily crop water use.
Agronomy Journal 92:821-832.
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
3 weeks at peak bloom 77 268 345
peak bloom to termination 77 231 308
2012 – Qk no drought 46 598 644
squaring 46 481 527
early flowering 46 376 422
3 weeks at peak bloom 46 445 491
peak bloom to termination 46 350 396
2012 – ND no drought 137 554 691
squaring 137 509 646
early flowering 137 422 559
3 weeks at peak bloom 137 457 594
peak bloom to termination 137 423 560
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Irrigation Interruption Period Environment
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
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
43
Table 3.6. HVI UHML lengths (inches) observed during 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
--------------------------- 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
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
47
Table 3.7. Significance of the analysis of variance for AFIS variables as affected by
the irrigation episodic drought, cultivar or their interaction
Environment Factor Irrigation Cultivar 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 **
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
48
Table 3.8. Maturity ratios (%) observed in 2010 and 2011. 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
----------------------------- 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
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Irrigation Interruption Period Environment
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
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Irrigation Interruption Period Environment
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
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.
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
irrigation regimes affect cotton yield, water use efficiency and fiber quality in
western Turkey. Agricultural Water Management Journal. 96: 111-120.
Doorembos, J., Kassam A.H. (1979) Yield Response to Water FAO, Rome. (FAO
Irrigation and drainage paper 33)
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
56
High Plains. Journal of Irrigation and Drainage Engineering 130:277-285.
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.
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.
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)
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.
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.
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.
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
57
Torell, L.A., Libbin J.D., Miller M.D. (1990) The Market Value of Water in the
Ogallala Aquifer. Land Economics 66:163-175.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
4.3. Materials and Methods
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
The experimental design was a split-plot with irrigation treatments as the main
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
All plots were harvested and weighed with a cotton stripper equipped with load
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
2011 - ND Full irrigation 39 638 677
Deficit irrigation 39 473 512
No irrigation from squaring 39 355 394
2012 - Qk Full irrigation 46 583 629
Deficit irrigation 46 263 309
No irrigation from squaring 46 173 219 ND = Texas Tech New Deal Research Farm.
Qk = Texas Tech Quaker Avenue 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 *
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
LSDvar = Least significant difference for cultivar effects
LSDint = Least significant difference for irrigation-cultivar interaction effects
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
.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Factor p-value in 2010 p-value in 2011 p-value in 2012
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 **
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
LSDvar = Least significant difference for cultivar effects
LSDint = Least significant difference for irrigation-cultivar interaction effects
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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 *
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
71
Table 4.7. Significance of the analysis of variance for AFIS variables as affected by
the irrigation level, cultivar or their interaction
Environment Factor Irrigation Cultivar 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.
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
75
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.
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
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.
Dagdelen, N., Basal, H.; E. Yilmaz; T. Gurbuz; and S. Akcay. 2009. Different drip
irrigation regimes affect cotton yield, water use efficiency and fiber quality in
western Turkey. Agricultural Water Management Journal. 96: 111-120.
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.
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).
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
76
Pettigrew, W.T. (2004) Physiological Consequences of Moisture Deficit Stress in
Cotton. Crop Sci.. 44:1265-1272.
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.
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.
Torell, L.A., Libbin J.D., Miller M.D. (1990) The Market Value of Water in the
Ogallala Aquifer. Land Economics 66:163-175.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
at early flowering; 3 weeks of non-irrigation beginning at peak bloom; and non-
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
5.3. Materials and Methods
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
81
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 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
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. 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
89
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
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., Nichols R.L. (2005) Phenological and Morphological Components of
Cotton Crop Maturity. Crop Science 45:1497–1503
DOI:10.2135/cropsci2004.0321
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.
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.
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.
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)
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
90
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.
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
SAS Institute. 2010. The SAS system for Windows. Release 9.3. SAS Inst., Cary, NC.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
6.3. Materials and Methods
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
97
with weekly measurements 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. 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.
The statistical analysis was done 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).
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought squaring early
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
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom to
termination
Ag
ron
om
ic W
UE
(k
g h
a-1
mm
-1)
Irrigation interruption period
DP0935
FM9180
(b)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
113
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
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.; 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.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
114
Dagdelen, N., Basal, H.; E. Yilmaz; T. Gurbuz; and S. Akcay. 2009. Different drip
irrigation regimes affect cotton yield, water use efficiency and fiber quality in
western Turkey. Agricultural Water Management Journal. 96: 111-120.
Howell, T.A., Evett S.R., Tolk J.A., Schneider A.D. (2004) Evapotranspiration of
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).
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.
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.
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)
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.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
7.4. Materials and Methods
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-
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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).
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.
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
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
The statistical analysis was done 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).
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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 *
Qk = Texas Tech Quaker Avenue Research Farm.
ND = Texas Tech New Deal Research Farm.
* = Statistically significant at 0.05 level
** = Statistically significant at 0.01 level
n.s. = Not statistically significant at 0.05 level
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Factor p-value in 2010 p-value in 2011 p-value in 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 **
* = Statistically significant at 0.05 level
** = Statistically significant at 0.01 level
n.s. = Not statistically significant at 0.05 level
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 *
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
See
d -
cott
on
yie
ld (
kg
ha
-1)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
Lin
t y
ield
(k
g h
a-1
)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
137
Figure 7.3. Lint percentage 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) = 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%
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom to
termination
Lin
t p
ercen
tag
e
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Irrigation interruption period
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)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Irrigation interruption period
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
0
200
400
600
800
1000
1200
1400
1600
1800
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom to
terminationEq
ual
wate
r volu
me
lin
t p
rod
uct
ion
(k
g)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Quaker Avenue research farm. Black bars represent one least significant difference.
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
squaring early flowering 3 weeks at peak
bloom
peak bloom to
termination
Ma
rgin
al
lin
t y
ield
(k
g h
a-1
mm
-1)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Factor Irrigation Cultivar Interaction
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Quaker Avenue research farm. Black bars represent one least significant difference.
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
no drought squaring early
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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
.
0
500
1000
1500
2000
2500
3000
3500
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
To
tal
rev
enu
e (U
S$
ha
-1)
Irrigation interruption period
DP0935
FM9180
DP0912
FM9170
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought squaring early flowering 3 weeks at
peak bloom
peak bloom to
terminationPro
fit
or
loss
(U
S$
ha
-1)
Irrigation interruption period
DP0935
FM9180
DP0912
FM9170
Texas Tech University, Fulvio Rodriguez Simao, May 2013
147
Figure 7.14. Breakeven price 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$ 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
no drought squaring early flowering 3 weeks at
peak bloom
peak bloom to
termination
Bre
ak
even
pri
ce (
US
$ k
g-1
)
Irrigation interruption period
DP0935
FM9180
DP0912
FM9170
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought squaring early flowering 3 weeks at
peak bloom
peak bloom to
termination
Eq
ua
l ir
rga
tio
n v
olu
me
pro
fit
(US
$)
Irrigation interruption period
DP0935
FM9180
DP0912
FM9170
Texas Tech University, Fulvio Rodriguez Simao, May 2013
149
Figure 7.16. Marginal value product to full irrigation of four cotton cultivars subjected
to five different irrigation interruptions periods during the 2011 season at the Texas
Tech Quaker Avenue research farm. Black bars represent one least significant
difference. LSD(0.05) = US$ 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
squaring early flowering 3 weeks at peak
bloom
peak bloom to
termination
Ma
rgin
al
va
lue
pro
du
ct (
US
$ h
a-1
mm
-1)
Irrigation interruption period
DP0935
FM9180
DP0912
FM9170
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Factor Irrigation Cultivar Interaction
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 **
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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
Avenue research farm. Black bars represent one least significant difference.
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
difference. LSD(0.05) = 0.58 kg ha-1
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Irrigation regime or interruption period
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Irrigation regime or interruption period
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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
.
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
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
156
Figure 7.23. Marginal to full irrigation 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) = 0.37 kg ha-1
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
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Factor Irrigation Cultivar Interaction
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
159
Figure 7.25. Average boll distribution per node from cultivars DP0395 and FM9180
subjected to eight different irrigation regimes during the 2012 season at the Texas
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
160
Figure 7.26. Revenue obtained 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 (p=0.05) =
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)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
161
Figure 7.27. Profit obtained 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 (p=0.05) =
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
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
162
Figure 7.28. Breakeven price for 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$ 0.26 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
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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$ 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$1,000.00 superior than the
obtained at an increased area at a mild deficit irrigation regime.
-4000
-3000
-2000
-1000
0
1000
2000
3000
full irrigation mild deficit
irrigation
squaring 3 weeks at
peak bloom
peak bloom
totermination
early
flowering
deficit
irrigation
no irrigation
Eq
ua
l w
ate
r p
rofi
t (U
S$
)
Irrigation regime or interruption period
DP0935
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
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
.
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
169
References
Allen, V.G., Baker M.T., Segarra E., Brown C.P. (2007) Integrated Crop-Livestock
Systems in Dry Climates. Agronomy Journal 99:346-360.
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.
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)
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.
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.
Littell, R.C., G.A. Milliken, W.W. Stroup, and R.D. Wolfinger. 1996. SAS system for
mixed models. SAS Inst., Cary, NC.
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
Plains Cotton Growers Inc.
<http://www.plainscotton.org/2012Loan/2012LoanPandD.pdf> Access on
November 16th 2012.
Robinson, J.R.C., Michelsen A.M., Gollehon N.R. (2010) Mitigating water shortages
in a multiple risk environment. Water Policy 12:114–128.
Roseta-Palma, C. 2003. Joint Quantity/Quality Management of Ground Water.
Environmental and Resource Economics 26: 89-106.
SAS Institute. 2010. The SAS system for Windows. Release 9.3. SAS Inst., Cary, NC.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
170
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)
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
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
175
BIBLIOGRAPHY
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
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.
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.
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
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., Nichols R.L. (2005) Phenological and Morphological Components of
Cotton Crop Maturity. Crop Science 45:1497–1503
DOI:10.2135/cropsci2004.0321
Texas Tech University, Fulvio Rodriguez Simao, May 2013
176
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.
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.
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.
Agronomy Journal 98:212-219.
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)
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.
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.
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., 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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
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.
Doll, J.P., Orazen F. (1992) Production economics: theory with applications. Second
Edition ed. Krieger Publishing Company, Florida.
Doorembos, J., Kassam A.H. (1979) Yield Response to Water FAO, Rome. (FAO
Irrigation and drainage paper 33)
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.
Howell, T.A., Evett S.R., Tolk J.A., Schneider A.D. (2004) Evapotranspiration of
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.
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
178
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)
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.
Agronomy Journal 92:821-832.
Littell, R.C., G.A. Milliken, W.W. Stroup, and R.D. Wolfinger. 1996. SAS system for
mixed models. SAS Inst., Cary, NC.
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.
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)
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
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
179
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.
Pettigrew, W.T. (2004) Physiological Consequences of Moisture Deficit Stress in
Cotton. Crop Sci.. 44:1265-1272.
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.
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., 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.
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
Robinson, J.R.C., Michelsen A.M., Gollehon N.R. (2010) Mitigating water shortages
in a multiple risk environment. Water Policy 12:114–128.
Roseta-Palma, C. 2003. Joint Quantity/Quality Management of Ground Water.
Environmental and Resource Economics 26: 89-106.
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.
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)
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
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.
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
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)
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
181
APPENDICES
Texas Tech University, Fulvio Rodriguez Simao, May 2013
182
APPENDIX A
WEATHER PARAMETERS
Figure A.1. Daily ET0 measured during the 2010 season at the Texas Tech University
New Deal research farm.
Figure A.2. Daily ET0 measured during the 2011 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
183
Figure A.3. Daily ET0 measured during the 2011 season at the Texas Tech University
Quaker Avenue research farm.
Figure A.4. Daily ET0 measured during the 2012 season at the Texas Tech University
New Deal 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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
184
Figure A.5. Daily ET0 measured during the 2012 season at the Texas Tech University
Quaker Avenue research farm.
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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]
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
ITEM UNIT PRICE QUANTITY AMOUNT
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
189
Continuation of Table D1
ITEM UNIT PRICE QUANTITY AMOUNT
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
191
Figure 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.
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
3 weeks at peak bloom no drought peak bloom to
termination
Av
era
ge
nu
mb
er o
f n
od
es p
er p
lan
t
Period of irrigation interruption
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Quaker Avenue research farm.
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 Interruption Period
irrigation
precipitation
330 259 250 268
231
0
50
100
150
200
250
300
350
400
450
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
Irrig
ati
on
+ P
reci
pit
ati
on
(m
m)
Irrigation Interruption Period
irrigation
precipitation
Texas Tech University, Fulvio Rodriguez Simao, May 2013
196
Figure E.11. Total water per treatment during the 2012 season in the Texas Tech
Quaker Avenue research farm.
Figure E.12. Total water per treatment during the 2012 season in the Texas Tech New
Deal research farm.
598
481 376
445 350
0
100
200
300
400
500
600
700
no drought squaring early
flowering
3 weeks at
peak
bloom
peak
bloom to
termination
Irrig
ati
on
+ P
reci
pit
ati
on
(m
m)
Irrigation Interruption Period
Irrigation
Precipitation
554 509
422 457 423
0
100
200
300
400
500
600
700
800
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
Irrig
ati
on
+ P
reci
pit
ati
on
(m
m)
Irrigation Interruption Period
Irrigation
Precipitation
Texas Tech University, Fulvio Rodriguez Simao, May 2013
197
Figure E.13. Total water per treatment during the 2010 season in the Texas Tech New
Deal research farm.
Figure E.14. Total water per treatment during the 2011 season in the Texas Tech New
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 treatment
Irrigation
Precipitation
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
2010 - ND Full irrigation 325 212 510
Deficit irrigation 325 118 446
No irrigation from squaring 325 27 353
2011 - ND Full irrigation 39 638 677
Mild deficit irrigation 39 527 566
Deficit irrigation 39 473 512
No irrigation from squaring 39 355 394
2012 - Qk Full irrigation 46 583 629
Mild deficit irrigation 46 352 398
Deficit irrigation 46 263 309
No irrigation from squaring 46 173 219 ND = Texas Tech New Deal Research Farm.
Qk = Texas Tech Quaker Avenue 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
)
Period of irrigation interruption
DP0912
DP0924
DP0935
DP09R555
DP1028
FM1880
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
interruptions periods during the 2012 season. Black bars represent the standard errors.
0
500
1000
1500
2000
2500
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
Lin
t y
ield
(k
g h
a-1
)
Period of irrigation interruption
DP0912
DP0935
FM9170
FM9180
1000
1500
2000
2500
3000
3500
4000
4500
5000
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
See
d-c
ott
on
yie
ld (
kg
ha
-1)
Period of irrigation interruption
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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.
Factor p-value in 2010 p-value in 2011 p-value in 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)
Irrigation treatment
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Irrigation treatment
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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.
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
interruptions periods during the 2011 season. Black bars represent the standard errors.
3.7
3.9
4.1
4.3
4.5
4.7
4.9
5.1
5.3
5.5
no drought 3 weeks at peak bloom peak bloom to
termination
Mic
ron
air
e
Period of irrigation interruption
DP0912
DP0924
DP0935
DP09R555
DP1028
FM1880
FM9170
FM9180
2.5
3
3.5
4
4.5
5
5.5
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
Mic
ron
air
e
Period of irrigation interruption
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
204
Figure E.23. HVI UHML length 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.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
no drought 3 weeks at peak bloom peak bloom to
termination
HV
I U
HM
L l
eng
gth
(in
)
Irrigation interruption period
DP0912
DP0924
DP0935
DP09R555
DP1028
FM1880
FM9170
FM9180
0.9
0.95
1
1.05
1.1
1.15
1.2
no drought squaring early
flowering
3 weeks at
peak bloom
peak bloom
to
termination
HV
I U
HM
L l
eng
tth
(in
)
Irrigation interruption period
DP0912
DP0935
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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
*: significant at P = 0.05; **: significant at P=0.01; n.s.: not significant at P = 0.05
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 ** ** **
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
no drought 3 weeks at peak bloom peak bloom to
termination
Fin
enes
s (m
Tex
)
Period of irrigation interruption
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
3 weeks at peak bloom no drought peak bloom to
termination
Ma
turi
ty r
ati
o (
%)
Period of irrigation interruption
DP0912
DP0924
DP0935
DP09R555
DP1028
FM1880
FM9170
FM9180
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Period of irrigation interruption
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Irrigation treatment
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
mild deficit 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
mild deficit irrigation
deficit irrigation
no irrigation
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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)
Irrigation interruption period
DP0935B2RF
FM9180B2F
Cultivar
Texas Tech University, Fulvio Rodriguez Simao, May 2013
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
)
Period of irrigation interruption
DP0912
DP0924
DP0935
DP09R555
DP1028
FM1880
FM9170
FM9180