Citation
Assessing Genotypic and Soil Management Resources for Improving Resilience in Southeastern United States Row Crop Production System

Material Information

Title:
Assessing Genotypic and Soil Management Resources for Improving Resilience in Southeastern United States Row Crop Production System
Creator:
Zurweller, Brendan A
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (158 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Agronomy
Committee Chair:
ROWLAND,DIANE L
Committee Co-Chair:
TILLMAN,BARRY
Committee Members:
ERICKSON,JOHN E
WRIGHT,DAVID L
MIGLIACCIO,KATI WHITE
PAYTON,PAXTON

Subjects

Subjects / Keywords:
adaptability -- irrigation -- nitrogen -- peanut -- roots -- water-deficit
Agronomy -- Dissertations, Academic -- UF
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Agronomy thesis, Ph.D.

Notes

Abstract:
Irrigation water and nitrogen (N) fertilizer are two of the most significant management inputs in the southeast agronomic production systems for mitigating abiotic stress. These inputs not only influence production amounts, but are a significant cost, both of which determine the profitability of a farmer. Additionally, both inputs impact water consumption and quality. As the human population increases so does the competition for quality freshwater between the residential and agricultural sectors, both which are necessary for maintaining quality of life. Research which assesses irrigation and N management optimization has the potential to positively influence these complex socio economic and environmental issues. One of the most challenging aspects of agronomic management is the complex interaction it has across production systems with varying crops, soil landscapes, and climatic conditions. Furthermore, stochastic yearly weather conditions often results in scenarios where management of water and N cannot meet crop demand resulting in yield loss due to abiotic stress. Fortunately, a large amount of genetic diversity exist among crop species possessing traits which can be utilized to mitigate yield loss under abiotic stress which can ultimately increase long term food and fiber production. Therefore, the proceeding chapters in this dissertation address both the optimization of water and N management, and evaluating whole plant genotypic trait responses for determining crop adaptability to environment and soil hydrological conditions. Chapter 1 of this dissertation will provide a general overview of the southern U.S. row crop system with an emphasis on nitrogen and water management practices. This chapter will also provide some information to the reader of the basic science which served as a foundation to the research chapters included in this dissertation. These research chapters included will address the following: (i) peanut genotypic pod yield adaptability to irrigation management across U.S. production regions; (ii) intrinsic above and below ground traits of disparate peanut genotypes for determining adaptability to soil hydrological conditions; (iii) the coordination of traits which improve peanut soil water acquisition and utilization in an southeast irrigation production system; (iv) utilizing irrigation scheduling tools and in-season N management strategies to optimize cotton production in the southeast. The final conclusions chapter will provide a brief synthesis of the preceding research chapters. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2017.
Local:
Adviser: ROWLAND,DIANE L.
Local:
Co-adviser: TILLMAN,BARRY.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2018-06-30
Statement of Responsibility:
by Brendan A Zurweller.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
6/30/2018
Classification:
LD1780 2017 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

ASSESSING GENOTYPIC AND SOIL MANAGEMENT RESOURCES FOR IMPROVING RESILIENCE IN SOUTHEASTERN UNITED STATES ROW CROP PRODUCTION SYSTEMS By BRENDAN ANTHONY ZURWELLER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVER SITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2017

PAGE 2

2017 Brendan Anthony Zurweller

PAGE 3

To my Mother and Father

PAGE 4

4 ACKNOWLEDGMENT S I would like to acknowledge my dissertation committee chair Dr. Diane Rowland for her gracious mentorship during my doctoral degree. Her guidance and wisdom have been fundamental to my success, and the development of my professional skills. Her impactf ul influence will undoubtedly follow me as I continue into my future endeavors. Furthermore, I am very thankful for the support she provided through a Graduate Research Assistant position Without this assistance, obtaining a doctoral degree would have no t been possible. I would also like to express my gratitude to my co chair Dr. Barry Tillman for providing me with his guidance during my doctoral program Not only did he provide assistance with my direct research focus, but provided me the exposure to hi s peanut breeding program which is knowledge I am grateful to have received. Additionally, I would like to thank all of my graduate research committee members for their insight, knowledge, and use of their resources over the duration of my degree. Finally I would like to thank peanut breeding, and PSREU employees who all contributed to the success of the research in this dissertation.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 1 1 ABS TRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 17 Introduction ................................ ................................ ................................ ............. 17 Dissertatio n Overview ................................ ................................ ............................. 24 2 EVALUATION OF PEANUT GENOTYPIC POD YIELD ADAPTABILITY TO IRRIGATION MANAGEMENT ACROSS U.S. PRODUCTION REGIONS .............. 26 I ntroduction ................................ ................................ ................................ ............. 26 Materials and Methods ................................ ................................ ............................ 31 Results ................................ ................................ ................................ .................... 35 Brownfield, TX ................................ ................................ ................................ .. 35 Lubbock, TX ................................ ................................ ................................ ..... 36 Citra, FL ................................ ................................ ................................ ............ 36 Genotypic Yield Stability ................................ ................................ ................... 37 Discussion ................................ ................................ ................................ .............. 38 Conclusion ................................ ................................ ................................ .............. 41 3 ASSESSMENT OF ABOVE AND BELOW GROUND TRAITS INTRINSIC TO DISPARATE PEANUT GENOTYPES FOR DETERMINING ADAPTABILITY TO SOIL HYDROLOGIC CONDITIONS ................................ ................................ ....... 51 Introduction ................................ ................................ ................................ ............. 51 Materials an d Methods ................................ ................................ ............................ 56 Results ................................ ................................ ................................ .................... 61 Discussion ................................ ................................ ................................ .............. 64 Conclusion ................................ ................................ ................................ .............. 68 4 COORDINATED ADAPTIVE PHENOTYPING FOR IMPROVING SOIL WATER ACQUISITION AND UTILIZATION ................................ ................................ ......... 84 Introduction ................................ ................................ ................................ ............. 84

PAGE 6

6 Materials and Methods ................................ ................................ ............................ 87 Results ................................ ................................ ................................ .................... 92 Discussion ................................ ................................ ................................ .............. 94 Conclusion ................................ ................................ ................................ .............. 99 5 OPTIMIZING COTTON IRRIGATION AND NITROGEN MANAGEMENT USING A SMARTPHONE IRRIGATION SCHEDULING TOOL AND IN SEASON NITROGEN APPLICATIONS ................................ ................................ ................ 112 Introduction ................................ ................................ ................................ ........... 112 Materials and Methods ................................ ................................ .......................... 117 Results ................................ ................................ ................................ .................. 121 Discussion ................................ ................................ ................................ ............ 125 Conclusion ................................ ................................ ................................ ............ 129 6 FINAL SUMMARY ................................ ................................ ................................ 142 7 REFERENCE LIST ................................ ................................ ............................... 146 8 BIOGRAPHICAL SKETCH ................................ ................................ ................... 158

PAGE 7

7 LIST OF TABLES Table page 2 1 Analysis of variance (ANOVA) for pod yield response variables at each site location in 2013 and 2014. ................................ ................................ ................... 44 2 2 Average genotypic pod yield for the Brownfield, Lubbock, and Citra location in 2013 and 2014 ................................ ................................ ................................ .... 45 2 3 Average (n=3) genotypic pod yield for the Irrigated (100%) and rainfed treatments at the Citra, FL location. ................................ ................................ ..... 48 2 4 Regre ssion analysis of pod yield response for increasing mm of total water received at the Citra, FL location. The regression coefficient (b) represents genotypic pod yield response for total water received across both years and irrigation treatments. ................................ ................................ ............................ 49 2 5 Average genotypic pod yield and stability across all three site locations. Regression coefficient (b) represents the pod yield stability (GxExM) which is the average pod yield for each genotype v ersus the average pod yield for each site location, year, and irrigation treatment. ................................ ................ 50 3 1 Amounts of water received through irrigation, irrigation plus precipitation, and the irrigation to pre cipitation ratio in 2015 and 2016 for each irrigation treatment at the PSREU near Citra, FL. ................................ ............................... 71 3 2 Analysis of variance (ANOVA) for response variables of TRL, TSA, and LAI. Root parameter data was summed across soil depths prior to running ANOVA. ................................ ................................ ................................ ............... 73 3 3 Analysis of variance (ANOVA) for response variables of TRL and TSA. Response variables were summed across 20 cm soil depth incremen ts prior to running ANOVA. ................................ ................................ .............................. 75 3 4 Analysis of variance (ANOVA) for pod yield response variables at the PSREU near Citra, FL. ................................ ................................ ................................ ...... 79 3 5 Average pod yields for each irrigation treatment at the PSREU near Citra, FL. Yields are averaged across peanut genotypes and years. ................................ .. 80 3 6 Average pod yields for each genotype and growing se ason at the PSREU near Citra, FL. The percent difference is the percentage of yield decline from 2015 to 2016. ................................ ................................ ................................ ....... 81 4 1 Analysis of variance (ANOVA) for response variables of TRL, TRV, soil water depletion, and soil water content measured at the Plant Science Research and Education Unit (PSREU) near Citra, FL in 2016. ................................ ........ 101

PAGE 8

8 4 2 Spearman correlation analysis between root parameters and so il water depletion (SWD) for each soil sampling depth on 75 DAP at the PSREU near Citra, FL in 2016. The volumetric soil water content (VSWC) on this day was high across all sampling depths and SWD was predominant in the top 30 cm of soil. ................................ ................................ ................................ ................ 106 4 3 Spearman correlation analysis between root parameters and soil water depletion (SWD) for each soil sampling depth on 82 DAP at the PSREU near Citra, FL in 2016. The volumetric soil water content (VSWC) on this day was low across all sampling depths and SWD was predominant below 30 cm of soil depth. ................................ ................................ ................................ .......... 107 4 4 Analysis of variance (ANOVA) for leaf level mid morning transpiration flux measured at the PSREU near Citra, FL in 2016. ................................ ............... 108 4 5 Spearman correlation analysis between average transpiration (TR) and volumetric soil water content (VSWC) across all sampling dates for each genotype at the PSREU near Citra, FL in 2016. ................................ ................ 110 5 1 Transient chlorophyll fluorescence (OJIP) parameter definitions as defined by Stirbet and Govindjee, (2011). ................................ ................................ ........... 130 5 2 Irrigation and total water received during the 2015 and 2016 growing season for each irrigation treatment at the PSREU near Citra, FL. Total water received is the sum of irrigation and precipitation. ................................ ............. 133 5 3 Analysis of variance (ANOVA) for SPAD chlorophyll content and leaf area index (LAI) variables at the PSREU near Citra, FL. ................................ ........... 134 5 4 Average specifi c and phenomenological energy fluxes for each N rate in 2015 and 2016. Data is showing the in season N rate, time within season, and year interaction. Parameters are averaged across both irrigation treatment and measurement date within a season. Measuremen ts averaged across dates included sampling which occurred from first developed boll to first open boll. ... 137 5 5 Analysis of variance (ANOVA) for N uptake, harvest index, and lint yield at the P SREU near Citra, FL. ................................ ................................ ....................... 139

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Precipitation and total water received for each irrigation treatment at the Brownfield, TX location for 2013 and 2014.. ................................ ........................ 43 2 2 Precipitation and total water received for each irrigation treatment at the Lubbock, TX l ocation during in 2013 and 2014 ................................ .................... 46 2 3 Precipitation and total water received for each irrigation treatment at the Citra, FL location during in 2013 and 2014.. ................................ ................................ .. 47 3 1 Cumulative precipitation from peanut planting to harvest in 2015 and 2016 at the Plant Science Research and Education Unit (PSREU) near Citra, FL.. ......... 70 3 2 Volu metric soil water content (left axis) and daily precipitation (r ight axis) during time periods when irrigation was applied in 2015 and 2 016 at the PSREU near Citra, FL ................................ ................................ .......................... 72 3 3 Average total genotypic root length growth at 21, 28, 35, 49, and 75 DAP at the PS REU.. ................................ ................................ ................................ ........ 74 3 4 Average genotypic root length summed across 20 cm depths at 21, 28, 35, Sig nificant Difference at P < 0.05 ................................ ................................ ......... 76 3 5 Average genotypic root surface area summed across 20 cm depths at 21, 28, 35, 49, and 75 DAP at the PSREU ................................ ................................ ...... 77 3 6 A) Average lea f area index (LAI) at 28, 35, 49, 64, and 77 DAP for each irrigation treatme nt and B) genotype at the PSREU ................................ ............ 78 3 7 Relationship between peanut pod yield and maximum leaf area index achieved for the 2015 and 2016 growing season at the PSREU near Citra, FL.. ................................ ................................ ................................ ....................... 82 3 8 Relationship among pod yield and maximum total root length achieved for the 2015 and 2016 growing season at the PSREU near Citra, FL.. ........................... 83 4 1 Total genotypic root length distribution for each root diameter class measures at the PSREU near Citra, FL in 2016. ................................ ................................ 102 4 2 A) Average total root length of each peanut genotype by soil depth measured at the PSREU near Citra, FL in 2016. B) Average total root volume of each peanut genotype by soil depth measured at the PSREU near Citra, FL in 2016.. ................................ ................................ ................................ ................. 103

PAGE 10

10 4 3 Total root volume architectural response of peanut genotype COC 041 to early season irrigation treatment measured at the PSREU near Citra, FL in 2016.. ................................ ................................ ................................ ................. 104 4 4 A) Soil water depletion by soil depth range for each sampling date. B) Corresponding volumetric soil water content by dep th range for each sampling date ................................ ................................ ................................ .... 105 4 5 (left y a xis) Average soil water content to a soil depth of 75 cm at each sampling date for each peanut genotype. (right y axis) Average leaf level mid morning transpiration flux of peanut genotypes at each sampling date for each peanut genotype. Data was measured at the (PSREU) near Citra, FL in 2016 109 4 6 A) Relationship between leaf level transpiration and soil water depletion for each peanut genotype. Data points include all weekly sampling points from 61 96 DAP measured at the PSREU near Citra, FL in 2016. B) Relationship between total root volume and soil water depletion for each peanut genotype.. 111 5 1 Depiction corresponding trans ient chlorophyll fluorescence (OJIP) parameters to the location within the light reactions of photosynthesis. ................................ 131 5 2 Cumulative precipitation from 1 May through 1 November for 2015 and 2016 at the Plant Science Research and Education Unit (PSREU) near Citra, FL. .... 132 5 3 Top) Average leaf area index (LAI) starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2015 at PSREU. Bottom) Average SPAD chlorophyll content starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2015 at PSREU . ................................ ................................ .......... 135 5 4 Top) Average leaf area index (LAI) starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2016 at PSREU.Bottom) Average SPAD chlorophyll content starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2016 at PSREU . ................................ ................................ .......... 136 5 5 Electron transport efficiencies from first developed boll to first open boll in 2015 (l eft) and 2016 (right). ................................ ................................ ............... 138 5 6 Average total N uptake for each N treatment in 2015 (top) and 2016 (bottom) at the PSREU near Citra, FL.. ................................ ................................ ............ 140 5 7 Average lint yield (top) and harvest index (bottom) for each N (left) and irrigation (right) treatment in 2015 and 2016 at the PSREU near Citra, FL. ...... 141

PAGE 11

11 LIST OF ABBREVIATIONS Akaike information criterion (A IC) Estimator of the relative quality/fit of a statistical model. Lower number defines model as a relatively better estimator. Analysis of variance (ANOVA) Statistical models used to analyze the variance among group means for determining differences. Ce llulose acetate butyrate (CAB) Material composition of mini rhizotron tubes. Coordinated adaptive phenotyping (CAP) Term referring to the assessment of pertinent crop traits functioning in unison to provide a biological advantage in a particular cropping production system. Used in context to soil water acquisition and utilization in the south east production system in this dissertation. Days after planting (DAP) Number of days after crop planting/sowing. Dehydration vulnerability scenario (DVS) This ter m refers to general water deficit conditions which are relevant to a particular production system as a result of climate, environmental conditions, and production management. Economically optimum nitrogen rate (EONR) Defined as the nitrogen rate that maxi mizes the value of crop produced at a given crop and fertilizer price. significant difference (LSD) A statistical method for determining confidence intervals between pairwise comparisons while controlling for a specified experiment wise er ror rate. General linear mixed model (GLIMMIX) Statistical model used for computing an analysis of variance. Genotype by environment interaction (GxE) Infers that crop genotypic traits are responding differently due to variation in environmental conditi ons. Genotype by environment by management interaction (GxExM) Infers that crop genotypic traits are responding differently due to variation in environmental conditions and management decisions. Genotype by management Infers that crop genotypic traits are responding differently due to variation in management decisions.

PAGE 12

12 Interaction (GxM) Genotype by year interaction (GxY) Infers that crop genotypic traits are responding differently due to differences among growing seasons. Harvest index (HI) Term reefi ng to the proportion of harvestable crop biomass to total aboveground biomass. Autoregressive (AR1) Covariance matrix structure used in repeated measures analysis of variance to avoid violating independence assumption. Structure has homogenous variance and correlation among variances in time decline exponentially with distance. Heterogeneous autoregressive (ARH1) Covariance matrix structure used in repeated measures analysis of variance to avoid violating independence assumption. Structure has hetero geneous variance and correlation among variances in time decline exponentially with distance. Nitrogen use efficiency (NUE) Proportion of nitrogen in above ground dry biomass to the total amount of nitrogen applied to the system. Internal nitrogen use ef ficiency (iNUE) Above ground biomass produced per unit of tissue nitrogen. Leaf area index (LAI) Unit area of crop canopy per soil area. New Mexico Valencia C (NMVC) Valencia peanut genotype of fastigiata descent. Plant available water (PAW) The length of water available to a plant at a particular time period during the growing season. Estimated between the product of the effective root zone (length units) and the available water holding capacity of the soil. Plant available water replacement (PAWR) T he percent of plant available water replacement from applying irrigation. Plant science research and education unit (PSREU) Research facility near Citra, FL where crop research was conducted. Primed acclimation (PA) Irrigation management practice of usi ng regulated deficit irrigation during the predominantly vegetative growth stages of a crop to alter homeostasis for improving later season drought resiliency. Regulated deficit An irrigation practice designed to reduce the amount of wat er

PAGE 13

13 irrigation (RDI) application below the full water requirement for optimum plant growth. Rain fed (RF) Irrigation treatment in where precipitation is the only form of water application. Root length density (RLD) Total longitudinal length of root segment per volume of soil. Root system architecture (RSA) Root parameters defined over the soil profile to make inference on root distribution through the soil. Soil water depletion (SWD) Measurement technique quantifying the amount of soil water depletion by examining th e difference in volumetric soil water content from pre dawn to pre dusk. Time domain reflectometry (TDR) Measurement technique to quantify volumetric soil water content by assessing the reflection of an electromagnetic pulse. Total root length (TRL) Total longitudinal length of root segment per distance of soil. Total root volume (TRV) Total root volume per distance of soil calculated where 2 *TRL Total surface area (TSA) Total cross sectional root area per distance of soil where CSA= Total water received as a percentage of reference evapotranspiration (TWRPET) Amount of total water (precip itation + irrigation) received in proportion to the amount of reference evapotranspiration defined by Penman Monteith equation. Transient chlorophyll fluorescence (OJIP) Method for quantifying both physical and chemical energy transfer within the light reaction of photosynthesis. Volumetric soil water content (VSWC) Volume of water in proportion to the volume of soil. Vapor pressure deficit (VPD) The difference in water potential between the sub stomatal cavity of a plant and the atmosphere. Varia ble rate Irrigation technology which allows for custom water application at

PAGE 14

14 irrigation specific location within a field. 100% Refers to positive control irrigation treatment. 60% Refer to season long reduction in irrigation application. 60%PA Primed acclimation treatment consisting of a 40% reduction of the 100% treatment.

PAGE 15

15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ASSESSING GENOTYPIC AND SOIL MANAGEMENT RESOURCES FOR IMPROVING RESILIENCE IN SOUTHEASTERN UNITED STATES ROW CROP PRODUCTION SYSTEMS By Brendan Anthony Zurweller December 2017 Chair: Diane Rowland Co chair : Barry Tillman Major: Agronomy I rrigation water and nitrogen (N) fertilizer are two of the most significant management inputs in southeast agronomic production systems for mitigating abiotic stress. These inputs not only influence production, but are a significant cost, both determining farmer profitability. Additionally, both inputs impact water consumption and quality. As the human population increases, so does the competition for quality freshwater between the residential and agricultural sectors. Research that assesses irrigation and N management optimization has the potential to positively influence these complex socio economic and environmental issues. One of the most challenging aspects of agronomic management is the complex interaction it has across production systems with varying crops, soil landscapes, and climatic conditions. Furthermore, stochastic yearly weather conditions often result in scenarios where management of water and N cannot meet crop demand resulting in yield loss due to abiotic stress. Fortunately, a large amount of genetic diversity exists within crop species possessing traits which can be utilized to mitigate yield loss under abiotic stress, and ultimately increasing long term food and fiber production.

PAGE 16

16 Therefore, the proceeding chapters in this dissertation add ress both the optimization of water and N management, and evaluation of whole plant genotypic trait responses for determining crop adaptability to environment and soil hydrologic conditions. Chapter 1 of this dissertation will provide a general overview of the southern U.S. row crop production system with an emphasis on nitrogen and water management practices, providing the reader an introduction to the research chapters included in this dissertation. The research chapters will specifically address: (i) pea nut genotypic pod yield adaptability to irrigation management across U.S. production regions; (ii) intrinsic above and below ground traits of disparate peanut genotypes for determining adaptability to soil hydrologic conditions; (iii) the coordination of traits which improve peanut soil water acquisition and utilization in the southeast irrigation production system; (iv) utilizing irrigation scheduling tools and in season N management strategies to optimize cotton production in the southeast. The final con clusions chapter provides a brief synthesis of the major findings from this research.

PAGE 17

17 CHAPTER 1 LITERATURE REVIEW Introduction Southern U.S. cropping systems : Two of the predominant rotational row crops in the southern United States (U.S.) are peanu t ( Arachis hypogaea L.) and upland cotton ( Gossypium hirsutum L.).The southern U.S. can generally be delineated into specific regions with states of Florida, Georgia, Alabama, and Mississippi in the southeast; Virginia, North Carolina, and South Carolina i n the east; and Texas, New Mexico, and Oklahoma in the southwest. These three regions (southeast, east, and southwest) accounted for an average of 73, 16, and 11% of the total area planted to peanut from 2012 2014, respectively (USDA NASS, 2015) Cotton production in these three regions (southeast, east, and southwest) accounted for an average of 23, 9 and 68% of the total area planted in 2016, respectively (USDA NASS, 2015) .These regions also encompass extreme variation in disease pressure, temperature, humidity, rainfall patterns, and growing season duration. Two wide regional environmental characterizations can be made: 1) th e southwestern production areas are characterized as having a semi arid climate with low annual rainfall and high vapor pressure deficits (VPD); and 2) the southeast and eastern production areas are characterized as having humid conditions with high total annual precipitation and, in some years, high frequency of precipitation events. Although the humid southeastern and eastern regions typically receive more rainfall than the southwestern region, growing season rainfall patterns can be unpredictable and geo graphically isolated, which when combined with highly permeable soils in some areas, creates a need for supplemental irrigation. The potential for yield reductions across both regions due to large ranges of GxExM creates a need to

PAGE 18

18 explore the responses of genetically diverse crop genotypes in an effort to increase yield potential across both environments, and the examination of management practices specific to each region for optimizing inputs to increase grower profitability and reduce detriments to the en vironment. Peanut and cotton irrigation practices : One critically important management component that is prevalent across the southeast and southwest crop production systems is irrigation. Irrigated peanut acreage has recently accounted for 62 and 94% of the harvested peanut land area in the major peanut producing regions of the southeast and southwest, respectively (USDA NASS, 2014). Irrigated cotton acreage has recently accounted for 46 and 59% of the harvested cotton land area in the major peanut produc ing regions of the southeast and southwest, respectively (USDA NASS, 2014). This heavy reliance on irrigation across regions regardless of annual precipitation averages, shows the importance of supplemental water to combat the uncertainty in precipitation patterns that may decrease yield and impact grower profitability. Research in the southeastern U.S. has shown that overall yield increases are observed with irrigation; however obtaining a positive net return on irrigation cost can largely be dependent on peanut prices and crop rotation practices (Lamb et al., 1997; Lamb et al., 2007). Therefore, optimization of irrigation is critical in the southeast environment to ensure a positive net return on irrigation. Crop Phenological Sensitivity to Water Stress : P eanut is an indeterminate annual crop, meaning that vegetative and reproductive stages overlap with one another. Developmental stages of peanut are classified as: (V E ) emergence; (V o ) open cotyledons; (V1) first tetrafoliate; (R1) beginning bloom; (R2) beg inning peg; (R3)

PAGE 19

19 beginning pod; (R4) full pod; (R5) beginning seed; (R6) full seed; (R7) beginning maturity; (R8) harvest maturity (Boote, 1982). Peanut water use is low after emergence, and increases linearly with increasing LAI until maximal water use, w hich generally occurs from R4 R7 (Stansell et al., 1976). The peak reproductive development stages from R4 R7 is when peanut has been documented to be most susceptible to yield loss from water deficit, and reports have documented that water deficit stress during peg and pod development (R2 R4) most greatly reduce pod yield (Stirling et al., 1989; Patel and Gangavani, 1990; Meisner 1991). However, the pre flowering phase of peanut has been reported as a less sensitive time in phenological development to wate r deficit stress (Kulkarni et al., 1988; Patel and Golakiya, 1988; Rao et al., 1988). Cotton is also an indeterminate annual crop where ranges in reproductive sensitivity to water deficit stress have been documented. General developmental stages of cotton are vegetative development from seedling emergence to first square (start of reproductive growth), first bloom, first developed boll, and first open boll (Oosterhuis, 1990). Sensitive developmental time periods when cotton is most susceptible to lint yiel d loss due to water deficit stress are from square initiation through boll development, with the most sensitive time period being the first half of boll development (Hake and Grimes, 2010). Additionally, large fluctuations in soil water conditions can caus e abscission of reproductive fruits leading to lint yield losses (squares, flowers and bolls) (Hake and Grimes, 2010). Use of Regulated Deficit Irrigation (RDI) : Varying yield loss sensitivities to water deficit stress over the course of crop development h ave led to the examination of using developmental stage based RDI. Specifically, vegetative RDI is the practice of

PAGE 20

20 reducing water application below the full water requirement for optimum plant growth during vegetative growth when the crop is less sensitive to water deficit stress. This results in water savings with no yield reductions (Chai et al., 2016). Rowland et al. (2012) reported no decrease in peanut yield in one site year when plants were treated with early and mid season irrigation deficit followed by no late season irrigation deficit when compared to season long no deficit irrigation. Sustained peanut yields resulted in similar economic returns when peanuts were treated with a 50% irrigation reduction for the first 45 DAP, followed by no deficit ir rigation for the remainder of the growing season when compared to no deficit irrigation for the entire growing season. Furthermore, the use of vegetative RDI in peanut has been demonstrated to increase root proliferation deeper in the soil which could poss ibly increase water deficit stress resilience (Rowland et al., 2012; Thangthon et al., 2016). However, root distribution and pod yield responses have been reported to vary among peanut genotypes demonstrating variation in phenotypic plasticity (Jongrungkl ang et al., 2011). Furthermore, the phenotypic plasticity to increase root distribution deeper into the soil profile during vegetative water stress is likely heavily influenced by environmental conditions. This is of particular importance since success usi ng vegetative RDI has been documented in arid environments (Rowland et al., 2012), or under controlled conditions with rainout shelters (Jongrungklang et al., 2011; Thangthon et al., 2016). In both scenarios, total water received is primarily through irrig ation in contrast to peanut production in humid environments where stochastic precipitation is the primary water source, and water stress is more probable to be mild. Therefore, this previous work could be further advanced by examining both above and

PAGE 21

21 belo w ground response of disparate peanut genotypes to various irrigation management strategies in the southeast humid environment for determining optimal irrigation strategies, and possibly more water deficit stress tolerant germplasm. The use of vegetative RDI has been examined for cotton in the southeast production environment. Meeks et al (2017) reported no lint yield reductions when comparing pre bloom irrigation scheduling thresholds of 70 and 100 kPa to 20 kPa. Yield reductions were reported in one of the drier years of the study when water was withheld after the pre bloom period. These results indicate that reduction of water until mid bloom is a viable water saving strategy that can increase grower profitability and agronomic water use efficiency. However, the use of soil water tension sensors used to schedule irrigation, as employed in this study, can be costly, and may not be feasible in some cotton production systems. Free ET based soil water balance models have been developed for cotton (Vellidi s et al., 2016) which provides site specific recommendations for irrigation scheduling. Providing information in regards to RDI management using this tool could be useful for growers to optimize irrigation management. N requirement in cotto n: Another majo r variable cost in cotton production as compared to peanut is nitrogen (N) fertilization, which is the predominate contributor of the overall fertilization cost of cotton and usually exceeds the irrigation costs (Mullen et al., 2009). Estimates of the N re quirement per unit of lint produced range from 10 to 29 kg N 100 kg 1 lint across global cotton production regions (Miley and Oosterhius, 1989; Gerik et al., 1998). Typical application rates in cotton production regions of the United States (U.S.) have rep orted ranges from 71 220 kg N ha 1 (Taylor, 1995). Interestingly, cotton cultivar improvement since 1970 has resulted in cultivars with average lint yield

PAGE 22

22 increases of about 40%, and N uptake and internal nutrient use efficiency increases of approximately 20% (Rochester and Constable, 2015). This suggests that selective breeding for increased yields has indirectly resulted in increased N uptake and iNUE, and could possibly mean that N requirements could be less than previously reported. However, a major fac tor adding to the variability in estimates of N requirements is the overall susceptibility of N loss through NO 3 leaching and denitrification, both of which are heavily influenced by unpredictable seasonal environmental conditions and soil texture. Using in season N applications that split the total N applied into percentages applied at varying developmental stages is a management strategy that attempts to increase N efficiency by synchronizing soil N concentrations to plant N demand. Cotton nitrogen accu mulation is non linear with the threshold of increase reported to begin at first square, reaching peak accumulation rates at approximately early mid flower the developmental stage that accounts for up to 23 45% of the total N accumulation (Armstrong and Albert, 1931; Olson and Bledsoe, 1942; Bassett et al.,1970; Halevy, 1976; Oosterhuis et al., 1983; Halevy et al., 1987; Mullins and Burmester, 1991). Furthermore, N accumulation in the seeds and lint account for up to 43 60% of the total plant N, creating a large N demand sink relatively late in the growing season (Mullins and Burmester, 2010). Geng et al. (2015) reported a yield increase of 106 kg lint ha 1 in one of two growing seasons when comparing a split N application (180 kg N ha 1 total application rate) of 40% at pre plant and 60% at first flower compared to 100% of the N applied at pre plant. Additionally, Yang et al. (2011) observed that when splitting N application rates of 225 kg N ha 1 between pre plant and peak bloom (with a 40%

PAGE 23

23 application ra te utilized at first bloom across all treatments), lint yields were greatest when applied at 0 40 60 % split application. Other reports have observed declines in maximal obtainable lint yields when comparing split N applications of 50% at pre plant and fir st square, to yields obtained by applying 100% pre plant N, when utilizing N rates up to 180 kg N ha 1 (Reiter et al., 2008). However, splitting N generally resulted in a greater amount of lint produced per kilogram of N applied translating into a greater economic return. Soil texture has also been demonstrated to influence both the total in season N rates and the timing of application. Some studies utilize the concept of Economically Optimal Nitrogen Rate (EONR), defined as the N rate that maximizes the v alue of cotton produced at a given cotton and fertilizer price. The EONR concept is useful for evaluating the impact of N rates across different soil textures which have varying levels of productivity. For example, Scharf et al. (2012) reported a lower EON R for silt and sandy loams when compared to a clay soil. The EONR value was also useful for elucidating the impact of soil type on in season N applications applied at early square growth stage. This study also evaluated different timings on in season N app lications. No yield differences were observed when N was evenly split (56 kg N ha 1 ) between pre plant + early square or pre plant + early flower in sand and silty loams. However, optimal N split for a clay soil was 1/3 pre plant, 1/3 early square, and 1/3 early flower (Scharf et al., 2012). Individually, both N and irrigation management have been examined extensively in cotton; however, less research has examined these management inputs simultaneously as they occur in production systems. The Florida cott on production

PAGE 24

24 environment is characterized by Dissertation Overview Irrigation and N management are the two most costly m anagement inputs in U.S. peanut and cotton production systems for avoiding crop production reductions due to abiotic stress. The use of both vegetative RDI and in season N applications has been documented in the literature to increase the efficiency of irr igation water and N inputs per unit of yield gained. However, the specific amounts of water or N inputs while utilizing these management reduction strategies are dependent on the production environment of interest. To date, no examination of the simultaneo us management of both vegetative RDI and N management has occurred in the southeast cotton production system. Although the use of vegetative RDI has been explored in arid peanut production regions of the U.S. (Rowland et al., 2012), it has not been examine d in the sandy soils of the humid Southeast. Furthermore, the examination of whole plant peanut trait responses of disparate peanut genotypes to these irrigation practices has yet to be explored in this environment. The assessment of these genotypic traits responses is critical for identifying germplasm which may be better adapted for a

PAGE 25

25 particular soil hydrologic scenario and utilized in a breeding program. Therefore, the proceeding chapters of this dissertation will address these research needs of determin ing both genotypic adaptability and efficient water/N management strategies for optimizing resiliency in southeastern U.S. row crop production systems.

PAGE 26

26 CHAPTER 2 EVALUATION OF PEANUT GENOTYPIC POD YIELD ADAPTABIL IT Y TO IRRIGATION MANAGEMENT ACRO SS U.S. PRODUCTION REGIONS Introduction Over the past 3,500 years, cultivated peanut has expanded into the largest gene pool of the Arachis genus including subspecies fastigiata and hypogaea which are further separated into the botanical varieties of vulg aris, fastigiata, peruviana, aequatoriana, hypogaea and hirsuta (Pasupuleti and Nigam, 2013) Prior to the inception of germplasm collections in 1959, only a narrow portion of this genetic diversity had been incorporated into breeding programs within the U.S. This considerably narrowed the genetic diversity within cultivars such that Knauft an d Gorbet were germplasm sources for 100 and 90%, respectively, of all pedigrees in the U.S. contributed 50% of the germplasm to runner market type cultivars prior to 1969 (Knauft parentage, defined as the probability a random allele at any locus in one cultivar is identical to a descended cultivar (Kempthorne, 1969), was 0.20 and 0.21 for Runner and Virginia market types, respectively. This indicates progress was made in U.S. peanut breeding programs in introducing genetic variability in cultivars developed afte r 1969. Development of core and mini core collections have further advanced the ability of peanut breeding programs to utilize a reduced number of phenotypically diverse genotypes for practical trait screening while still incorporating improved levels of g enotypic diversity (Brown, 1989; Holbrook et al., 1993; Upadhyaya et al., 2003;

PAGE 27

27 Upadhyaya et al., 2005; Upadhyaya et al., 2006). The peanut mini core collection includes 184 accessions representing 10.8% of accessions in the core collection (Upadhyaya et a l., 2002), a collection representing the theoretical genetic variation of economically important traits in the U.S. germplasm collection. This has been estimated by using the Shannon Weaver index, a measure of allelic richness and evenness in a population, which was similar when comparing 13 morphological and agronomic descriptors between the core and mini core collections (Upadhyaya et al., 2002). Furthermore, the mini core has been reported to have a wide variation of genotypic pod yield ranging from 93 t o 2,287 kg ha 1 (Upadhyaya et al., 2002). Although this demonstrates that genetic factors significantly contribute to yield variability among peanut genotypes, it is still imperative to examine other factors that influence this complex, quantitative trait. Although genotypic variation may have a strong influence on both crop yield and yield components, these traits are of a polygenic nature and are highly influenced by the environment (Cattivelli et al., 2008). In peanut specifically, an analysis of 504 ac cessions across six environments in Asia resulted in broad sense heritability values of 35.5 and 56.9% for pod yield plant 1 and pod yield plot 1 respectively (Upadhyaya et al., 2005). This low to moderate range of broad sense heritability highlights the importance of examining phenotypic traits of different peanut genotypes across multiple environments for separating genotype and genotype by environment (GxE) influences to allow for the possible development of cultivars either for specific or broad rangin g environments (Mothilal 2012). A further refinement of this concept has occurred through improved understanding that the combination of both management and genetic gains

PAGE 28

28 has contributed to historical increases in yield production: a concept known as genot ype by management interaction (GxM) (Tollenaar and Lee, 2002). Therefore, it is critically important to take into account the influence of GxE and GxM across production regions for a given crop as a way to optimize the potential for success in improving cr op yield performance. Considering the environmental variability among the geographical regions in U.S. peanut production accounts for the GxE impact on yield potential. Similarly, accounting for GxM must consider important crop management strategies that may differ among regions. It is likely that GxM contributes to GxE and could be another measurable and manageable component of GxE. One critically important management component that is likely to have strong influence on peanut genotypic yield potential is irrigation. Irrigated peanut acreage has recently accounted for 46 and 95% of the harvested peanut land area in the major peanut producing regions of the southeast and southwest, respectively (USDA NASS, 2014). Heavy reliance on irrigation across regions regardless of annual precipitation, displays the importance of supplemental water to combat the uncertainty in precipitation patterns that may decrease yield and impact grower profitability. Research in the southeastern U.S. has shown that overall yield in creases are observed with irrigation; however obtaining a positive net return on irrigation cost can largely be dependent on peanut prices and crop rotation practices (Lamb et al., 1997; Lamb et al., 2007). It is important to examine the variability of pea nut germplasm in response to irrigation to determine the potential for water application to influence yield potential in different genotypes.

PAGE 29

29 The importance of both GxE and GxM to yield potential indicates that the optimal approach in breeding for a crop t hat is produced across large geographical ranges is a consideration of the genotype by environment by management interactions (GxExM). This approach is particularly important for peanut breeding programs in the U.S. because of the large geographical range of peanut production. Specifically, U.S. peanut production occurs primarily in the states of Florida, Georgia, Alabama, and Mississippi in the southeast; Virginia, North Carolina, and South Carolina in the east; and Texas, New Mexico, and Oklahoma in the s outhwest. These three regions (southeast, east, and southwest) accounted for an average of 73, 17, and 10% of the total area planted to peanut from 2014 2015, respectively (USDA NASS, 2015) These regions also encompass extreme variation in disease pressure, temperature range, humidity conditions, rainfall patterns, and growing season duration. For peanut, two wide regional environmental characterizations can be made: 1) the southwestern prod uction areas are characterized by a semi arid climate with low annual rainfall and high vapor pressure (VPD) deficits; and 2) the southeast and eastern production areas are characterized by humid conditions with high total annual precipitation and, in some years, frequent precipitation events. Although the humid southeastern and eastern regions typically receive more rainfall than the southwestern region, growing season rainfall patterns can be unpredictable and geographically isolated, which when combined with highly permeable soils in some areas, creates a potential need for supplemental irrigation. The susceptibility of yield reductions across both regions due to large ranges of GxExM provides a need to explore the responses of genetically diverse peanut genotypes in an effort to increase yield potential across both environments.

PAGE 30

30 Assessment of GxExM of diverse peanut genotypes produced across regions can be accomplished through the quantification of pod yield stability. Finlay and Wilkinson, (1963) develo ped an analysis which assesses yield stability across differing environments by plotting individual genotypic yield means against the population mean for specific site years, with the population mean across site locations and years being reflective of the relative yield potential. Thus, the regression coefficient for a specific genotype reflects its stability across varying environmental conditions, with a value of one representing average stability, and regression coefficients below or above one representi ng stable or unstable yield, respectively. Several studies have used this approach for assessing peanut pod yield stability across environments, as well as genotypic maturity and disease tolerance (Anderson et al., 1989; Adomou et al., 1997; Isleib et al., 2013; Narh et al., 2014). However, to our knowledge this analysis has not been used for examining genotypic peanut pod yield stability across varying environmental regions of the U.S. Therefore, to evaluate diverse peanut genotypes for their relative GxEx M effects on yield (with an emphasis on irrigation management as the most critical M component), we implemented research at multiple locations in the southeast and southwest U.S. utilizing genotypes from the peanut mini core collection. The overall objecti ve of this research was to evaluate yield stability of these peanut genotypes across sites varying in regional environmental characteristics and irrigation management. We hypothesized that a range of yield stability values would be present in different sit es, thus providing important information for breeding efforts aimed at developing regionally adapted cultivars.

PAGE 31

31 Materials and Methods Site Information : Field studies were conducted during the years of 2013 and Science Research and Education Unit near Citra, Florida (Citra) (29 24' 38" N, 82 10' 12" W); 2) the USDA ARS Cropping xas (Brownfield) (33 01' 55.76" N, 102 17' 48.52" W). Variation in soil texture occurred across sites: the Florida location was classified as Gainesville loamy sand (Hyperthermic, coated Typic Quartzipsamments); Brownfield was classified as Amarillo Clov is loamy fine sand complex ( Amarillo: Fine loamy, mixed, superactive, thermic Aridic Paleustalf; Clovis: Fine loamy, mixed, superactive, mesic Ustic Calciargids); and the Lubbock location was classified as Amarillo Acuff loamy fine sand (Amarillo: Fine loa my, mixed, superactive, thermic Aridic Paleustolls; Acuff: Fine loamy, mixed, superactive, thermic Aridic Paleustolls). Field Experiments: The experimental design at all locations consisted of a split plot arrangement in a randomized complete block design The whole plots consisted of irrigation treatments which were applied using a center pivot. The Brownfield and Lubbock locations had four replications, and the Citra location had three replications. For the Brownfield site, adequate irrigation (100%) was based on farmer well capacity, and the low irrigation (50%) treatment was implemented by decreasing the irrigation application frequency. At the Lubbock site, irrigation was set up to replicate the irrigation amounts applied at the Brownfield location, al though the low irrigation was applied at the same frequency of the 100% treatment, but involved reducing the irrigation application amount by 50%. Due to the greater probability of precipitation at the Citra

PAGE 32

32 location, and soils with lower water holding cap acity, irrigation treatments consisted of plots which received irrigation ranging from 12.7 25.4 mm per application (100%), while the reduced water application treatment was dependent on rainfall alone (rainfed control, RF). Irrigation was scheduled when s oil moisture tension at 15 cm soil depth reached 20 to 25 kPa. Nine peanut genotypes (C76 16, Chico, ICGS 76, COC 041, TMV 2, Serenut 5R, Serenut 6T, ICGV 86015, and ICGV 86388) were planted to the sub plots across all locations. Five additional genotyp es, New Mexico Valencia C (NMVC), FlavorRunner 458, FloRun TM TM Citra, four row plots were planted to a length of approximately 4.5 m with 0.9 m row spacing at approximately 1.8 seeds m 1 u sing a two row planter (Monosem, Inc., Edwardsville, KS). In furrow applications of Rhizobium inoculum at 1.31 l ha 1 (Advanced Biological Marketing, Van Wert, OH) were applied to enhance nodulation. To prevent thrips foliage feeding and the risk of develo ping Tomato spotted wilt virus phorate at 1.52 kg a.i, ha 1 (Amvac Chemical Corporation, Los Angelas, CA) was also applied; azoxystrobin at 0.21 kg a.i. ha 1 (Sygenta International AG, Basel, Switzerland) was applied at planting to prevent early fungal in fection. For the Brownfield and Lubbock trials, peanut genotypes were planted to the sub plots in four row plots to a length of approximately 6.0 m and 1.0 m row spacing at approximately 1.8 seeds m 1 In furrow applications of Rhizobium inoculum at 0.94 l ha 1 were made (Optimized Lift, EMD Crop Bioscience, Milwaukee, WI). At harvest, the lengths of the two center rows of each plot at the Citra location (quantified in length just after peanut emergence) were dug/inverted using a two row

PAGE 33

33 peanut digger (Kel ly Manufacturing Co., Tifton, GA). Inverted peanuts were allowed to air dry in the field until the canopy tissue dehydrated; plants were then threshed to remove pods from pegs and foliage using a peanut thresher (Kincaid, Haven Kansas). Pods were collected in burlap bags and dried to approximately 10.5% using a forced air drying wagon. At the Brownfield and Lubbock locations, rows were dug with a two row onion blade and inverted by hand. For the Brownfield trial, inverted plants were allowed to dehydrate in the field for one week. At the Lubbock location, inverted rows were threshed after drying in the field for approximately 6 hours using a peanut thresher (K.E.W., Kingaroy, QLD. Australia). Pods were collected in burlap bags and air dried to approximately 10.5%. Evaluation of the impact of water applied: Daily meteorological data were recorded using an automated weather station (Campbell Scientific, Logan, UT; Spectrum Technologies, Inc., Aurora, IL) located within 1500 m of the experimental plots at each site location. Daily weather data collected were used in determining the total water received (rainfall + irrigation) as a percentage of the ET o (TWRPET). Reference ET o was calculated using the Penman Monteith equation (FAO 56 Method) (Monteith, 1965; Alle n et al., 1998). To determine how TWRPET might affect critical periods of phenological development due to inherent variability in different sensitivities to water deficit stress in peanut (Reddy et al., 2003), TWRPET was calculated at 0 45, 45 90 DAP, and > 90 DAP. The 45 90 DAP encompasses a development period where pod yield declines are most sensitive to water stress, while >90 DAP is relatively insensitive (Patel and Golakiya, 1988). This approach aids in determining the relative

PAGE 34

34 potential impact of TWR PET on peanut yields; if TWRPET is low during developmental periods when stress sensitivity is high, the impact on yield is predicted to be great. Statistical Analysis: Statistical analysis was performed using SAS v. 9.4 statistical software (SAS Institute 2013). PROC GLIMMIX was used to compute analysis of variance (ANOVA) for pod yield traits. The data from each site location was analyzed separately. Random effects of rep nested within year, and rep crossed with irrigation nested within year were include d in the model. Normality and homogeneity were visually assessed by graphing the residual distribution, scatter plot of residuals, and Q Q plot of residuals. Data were pooled over factors if F test results had a P > 0.10. Multiple comparisons were performe Difference (LSD) at a P < 0.10 probability level. Regression analysis was performed using PROC REG (SAS v. 9.4; SAS Institute, 2013). The regression analysis at the Citra location was performed by predicting the response of genotypic pod yield to increasing total water received in the two site years and irrigation treatments as indicated by the regression coefficient (b). The amount of total water received for the two irrigation treatments across the two study ye ars was 574, 742, 812, and 951 mm so yield responses per unit of water are within the total water received range of 574 951 mm. Genotypic yield stability was performed for each genotype by plotting the average genotypic yield for each site year and irrigat ion treatment against the population mean for each site year and irrigation treatment, as according to Finlay and Wilkinson, (1963).

PAGE 35

35 Results Brownfield, TX Total precipitation received was 33 and 345 mm in 2013 and 2014, respectively (Figure 2.1), requiri ng application of more irrigation water in 2013 than 2014 due to the substantially drier conditions. In 2013 were 260 and 340 mm was applied in the 50 and 100% irrigation treatments, respectively; while in 2014, 122 and 231 mm was applied in the 50 and 100 % treatments, respectively. Total water received as a percentage of ET o (TWRPET) was 31 and 39% in the 50% and 100% irrigation treatments in 2013, respectively. In 2014, the TWRPET was 54 and 66% in the 50 and 100% irrigation treatments, respectively. The effect of low yearly TWRPET may have been even more detrimental in 2014 than 2013 due to the pattern of low levels of TWRPET during critical developmental stages in 2014. When examining TWRPET at the critical period of phenological development between 45 90 DAP when pod development is occurring, the TWRPET in 2013 was 41 and 62% in the 50 and 100% irrigation treatments, respectively; while in 2014 for this same time period, was 21 and 33% in the 50 and 100% irrigation treatments, respectively. These result s indicate that the peanut genotypes experienced a greater degree of water stress during the critical 45 90 DAP development stage in 2014 versus 2013. The interaction of site year with irrigation on peanut yields also supports the argument that the level o f water stress during the 45 90 DAP period in 2014 contributed to a large yield decline (Table 2.1). Overall, average pod yields were reduced by 561 kg ha 1 comparing the 2014 growing season to 2013 (Table 2.2). Although the 100% irrigation treatment yield ed similarly in both years, the 50% irrigation treatment was 1351 kg ha 1 lower in 2014 than 2013. The TWRPET results for the 50%

PAGE 36

36 treatment suggest that the likely cause of this yield loss was the water availability during the 45 90 DAP period in 2014. L ubbock, TX Precipitation received in 2013 and 2014 was 229 and 399 mm, respectively (Figure 2.2). Irrigation amounts in 2013 were 108 and 191 mm in the 50 and 100% irrigation treatments, respectively; while in 2014, irrigation amounts were 159 and 305 mm i n the 50 and 100% treatments, respectively. The yearly TWRPET when averaged across irrigation treatments in 2013 and 2014 was 36 and 67%, respectively. The distribution of rainfall was also different among the years, with a lower total received in 2013, bu t a more even distribution than in 2014. In 2014, approximately half of the rainfall was received after 90 DAP, resulting in a TWRPET from 90 DAP until harvest of 127%, as compared to 20% for that same time period in 2013. Late season water deficit in 201 3 likely contributed to lower yield levels as compared to 2014. However, there was year by genotype interaction for yield, indicating that genotypes differed in their relative sensitivities to differing water conditions between the dry year of 2013 and the wet year of 2014 (Table 2.1). Only one genotype, ICGV 76, had greater pod yield in 2013 than 2014; while yield of others was less in 2013 compared to 2014, including Chico, ICGV 86388, and Serenut 5R. All other genotypes yielded similarly in both years (T able 2.2). Citra, FL Total precipitation amounts in 2013 and 2014 were 729 and 544 mm, respectively, at the Citra location (Figure 2.3). Irrigation amounts in 2013 were 12.7 and 222 mm in the 50 and 100% irrigation treatments, respectively; while in 2014, irrigation amounts were 31 and 268 mm in the 50 and 100% treatments, respectively. An

PAGE 37

37 irrigation amount applied in the rainfed treatment for both years was to ensure optimal germination and crop establishment. The high amounts of precipitation in 2013 res ulted in a seasonal average TWRPET of 134 and 172% in the rainfed and the 100% irrigation treatment, respectively. The TWRPET was relatively lower in 2014 with an amount of 90% in the RF treatment and 127% in the 100% treatment. In 2013, pod yield did not increase with irrigation, likely due to adequate precipitation amounts (729 mm) (Table 2.3). Pod yields increased by 588 kg ha 1 in 2014 when comparing the 100% irrigated treatment to the rainfed. When averaged across both years, genotypes differed in the ir yield response to irrigation, with increased pod yields in the 100% treatment for Georgia 06G and NMVC; and decreased pod yield for genotype ICGS 76. Genotypes Flavorunner 458, FloRun TM trend of decreased pod yields for each addi tional increase mm of total water received. The regression coefficient predicting genotypic pod yield response per unit of total water received over the two site years was negative for Flavorunner 458, FloRun TM C76 16; while those with a positiv e relationship were NMVC, Chico, and COC 041 (Table 2.4). Genotypic Yield Stability Yield stability was assessed across all environments in an effort to determine which genotypes may have characteristics allowing them to maintain yield across environment s. Stability was measured by examining how the regression coefficient (b) genotypes have spec ific adaptability (Finlay and Wilkinson, 1963). Regression coefficients when modeled for each genotype across all irrigation treatments and site

PAGE 38

38 years ranged from 0.87 to 1.3 (Table 2.5). Genotypes C76 16, Chico, and ICGS 766 had b < 0.9; while genotypes I CGV 86015, ICGV 86388, and TMV 2 had b ranging from 0.9 to 1.0. The genotypes Serenut 5R, Serenut 6T, and COC 041 had b > 1.1. Discussion A strong main effect of genotype (G) and its interaction with year (GxY) occurred within all locations indicating that genetics and yearly environmental variation are predominant factors influencing yield potential. This has been demonstrated for peanut by Upadyaya et al. (2005), who reported a broad sense heritability of 57% in a mini core collection conducted across env ironments in Asia. In the current study, few differential genotypic pod yield responses to irrigation occurred at each site location indicating a relatively low GxM interaction on pod yield. However, this result may have also been partly due to the lack of dry conditions and substantial differentials in the amount of total water received across most site years. There were instances in this study where water differentials occurred in total water received among irrigation treatments. For example, at the Brow nfield location, the average (across both years) differential in TWRPET between the 50% and 100% irrigation treatment during the 45 90 DAP period, when peanut pod yields are most susceptible to water deficit, was 31 and 48%, respectively. This likely contr ibuted to the 52% decrease in pod yield compared to the 100% treatment. At this site, there was also no irrigation by genotype (GxM) interaction. This lack of a GxM or GxMxY interaction at the Brownfield location demonstrates that, at more severe levels of water stress, all genotypes were susceptible to yield reductions, suggesting that the variability of peanut genotypic tolerance to water stress is more likely to occur at moderate stress levels. Upadhyaya et al. (2002) reported a Shannon Weaver index for pod yield of 0.60 and

PAGE 39

39 0.17 during rainy and post rainy season, respectively. Although a pod yield reduction occurred in the 50% irrigation treatment in both seasons at Brownfield in the current study, the yield decline was greater in the 2014 growing seaso n, despite a greater average TWRPET over the growing season. This was likely due to the severity of water stress received during the 45 90 DAP period, a developmental period in which water stress most greatly impacts pod yields (Rao et al., 1985; Patel and Golakiya, 1990). The Lubbock site in the dry year of 2013, also experienced a more severe water differential, but in this case, the genotype ICGS 76 increased pod yield production. This same genotype at the Citra site also had increased pod yield in the r ainfed compared to the 100% treatment. This response indicates that ICGS 76 may have drought tolerance under more severe water stress. Other studies have demonstrated ICGS 76 to exhibit both heat tolerance and increased osmotic adjustment capacity under dr ought which may contribute to its ability to withstand severe water scarcity (Selvaraj et al., 2011; Padmavathi and Rao, 2013). Differential genotypic yield responses to irrigation (GxM) was present in the current study under mild levels of water deficits at the Citra location. The relationship between genotypic pod yield and total water received showed both positive and negative responses among cultivars. Genotypes which had a positive relationship between pod yield and total water received were NMVC, Chi co, and COC 041, indicating a particular sensitivity in pod yield production to fluctuations in total water received. There is evidence of this relationship at least for COC 041 when examining its stomatal sensitivity to soil drying. Kotapalli et al. (2009 ) observed greater and faster reduction of stomatal conductance to the onset of soil drying in COC 041 compared to

PAGE 40

40 COC 166. However, COC 041 also exhibited rapid stomatal opening (leading to increased photosynthesis) following soil re wetting, indicating a n overall greater sensitivity for this genotype to soil water conditions. Genotypes with more sensitive stomatal control are likely to have positive pod yield responses in environments with intermittent drying and wetting cycles. When total water received exceeded 544 mm at the Citra location, the genotypes FloRun TM mm of water received, indicating a possible maladaptation to sites subject to periods of excessive precipitation. Yiel d declines could be due to genotypic physiological responses to excessive soil water content in combination with increased disease susceptibility. Reports have indicated the susceptibility of Flavorunner 458 to Sclerotinia minor (Woodward et al., 2015). In terestingly, genotypes in the current study that had positive pod yield responses to total water received were of fastigiata descent, and negative pod yield responses were of hypogaea descent. These results suggest differences in inherent adaptive mechanis ms between the two subspecies which may be related to their origin. Krapovickas and Gregory (2007), reported that evidence supports that subspecies hypo gaea had its origin in southeast Bolivia, whereas subspecies fastigiata differentiated itself further no rth, possibly Peru. Finally, the sensitivity of peanut genotypes to more than adequate levels of soil water content is novel, as most studies have been focused on water scarcity conditions only. These results demonstrate the importance of examining genotyp ic tolerance to both drought and excessive soil water content for improving cultivar development across peanut producing regions.

PAGE 41

41 Yield stability assessed by examining the interaction of GxExM varied considerably among peanut genotypes, with Serenut 6T, Se renut 5R, and COC 041 demonstrating the lowest stability across all environments. In production environments where the population pod yield mean was low, these genotypes tended to have the lowest yields, while large relative yield increases were observed i n the production environments with greater population pod yield means. These genotypes are all of subspecies fastigiata descent, further validating their specific adaptability to greater potential yield production scenarios. However, their yields in high y ielding environments were still lower than the population mean. The genotypes C76 16, Chico, and ICGS 76 performed the best across all production environments by possessing both high yield stability and overall yield levels. These genotypes have also been documented to have physiological responses which deem them as being water deficit stress tolerant (Ketring et al., 1990; Ratnakumear et al., 2009; Dang et al., 2013). Therefore, this germplasm may be best suited for production regions that are characterize d as either arid or humid environments. Conclusion The stability analysis of this study showed variation in the genotypic pod yield responses to irrigation and environment (GxExM). Genotypes C76 16, Chico, and ICGS 76 in this study had the greatest pod yi elds and stability indicating their broad adaptability. In contrast, COC 041, Serenut 5R, and Serenut 6T had relatively low yield stability across site locations and irrigation management, demonstrating their specific adaptability to the site locations wi th the greatest population pod yield means. The interaction of genotype by irrigation management (GxM) varied among the site locations and years which was likely due to the range in severity of water stress conditions

PAGE 42

42 among sites. At the Brownfield locatio n, severe water stress resulted in a pod yield reduction across all genotypes in the 50% irrigation treatment. This result, with a lack of genotype by irrigation interaction (GxM) implies that genotypic pod yield variation is most likely to occur at more m ild levels of water stress. This is further supported by the genotype by irrigation interaction (GxM) which occurred at the Citra location, where greater amounts of TWRPET occurred. At this location, both C76 16 and ICGS 76 responded negatively to increasi ng total water received, indicating some susceptibility of decreasing pod yields in environments which receive high amounts of precipitation. In contrary, genotypes of NMVC, Chico, and COC 041 had a positive pod yield response to increasing total water rec eived, demonstrating greater specific adaptability to the humid Citra environment. Further research which examines whole plant trait responses relating to genotypic pod yield may provide greater understanding and strengthen the assessment of both broad and specific adaptability of divergent peanut genotypes to environmental conditions and irrigation management.

PAGE 43

43 Figure 2 1. Precipitation and total water received for each irrigation treatment at the Brownfield, TX location for 2013 and 2014.Sh aded colors represent developmental time of 0 45, 45 90, and >90 days after planting. Percentages during developmental time periods for 100% and 50% irrigation treatments are the total water received as a percentage of ET o

PAGE 44

44 Table 2 1. Analysis of varianc e (ANOVA) for pod yield response variables at each site location in 2013 and 2014. GENO, genotype treatment. Pod Yield ANOVA by Location Effect Brownfield Lubbock Citra YEAR 0.0159 0.5719 0.0336 IR <0.0001 0.1265 0.3556 YEAR*IR 0.0034 0.1761 0.0657 GENO <0.0001 <0.0001 <0.000 1 YEAR*GENO <0.0001 0.0043 <0.0001 IR*GENO 0.3543 0.4576 0.0613 YEAR*IR*GENO 0.3393 0.8480 0.6462

PAGE 45

45 Table 2 2. Average genotypic pod yield for the Brownfield, Lubbock, and Citra location in 2013 and 2014. Least Significant at P < 0.10. Location Genotype Brownfield Lubbock Citra 2013 --------------------------------------kg ha 1 -------------------------------C76 16 4811a 3508a 2921bc C hico 4113b 2319bc 3261ac ICGS 76 3663bc 3387a 3170ab COC 041 3331cd 1534de 3296bc TMV 2 3118ce 1741cd 3938d Serenut 5R 3147ce 1055e 2703ab Serenut 6T 2928de 1572ce 3732bc ICGV 86015 2735e 2070b 3103a ICGV 86388 1672f 1203ce 2119cd Mean 3280 2043 31 38 2014 C76 16 2850ac 3006a 6297a Chico 3158a 2242b 2641de ICGS 76 3232a 1990bc 4085b COC 041 2771bc 1080e 2238e TMV 2 2336bc 1329de 2402de Serenut 5R 1827c 1255de 3852d Serenut 6T 2569ab 1406de 3803bc ICGV 86015 2529ab 1809bd 3178b ICGV 86388 2393ac 1462cd 3138cd Mean 2629 1731 3515

PAGE 46

46 Figure 2 2. Precipitation and total water received for each irrigation treatment at the Lubbock, TX location during in 2013 and 2014.Shaded colors represent developmental time of 0 45, 45 90, and >90 days after planting. Percentages during developmental time periods for 100% and 50% irrigation treatments are the total water received as a percentage of ET o

PAGE 47

47 Figure 2 3. Precipitation and total water received for each irrigation treatment at the Citra, FL location during in 2013 and 2014.Shaded colors represent developmental time of 0 45, 45 90, and >90 days af ter planting. Percentages during developmental time periods for 100% and RF irrigation treatments are the total water received as a percentage of ET o

PAGE 48

48 Table 2 3. Average (n=3) genotypic pod yield for the Irrigated (100%) and rainfed treatments at the Citra FL location. Least Significant at P < 0.10. Citra Location Genotype Irrigated Rainfed ------kg ha 1 ------Georgia 06G 5905a 5152b C76 16 4369a 4968a Chico 3187a 2597a COC 041 2920a 2380a Flavorunner 458 3226a 3968a FloRun TM 4830a 5472a ICGS 76 3332b 4228a ICGV 86 015 4141a 3697a ICGV 86388 3030a 2756a NMVC 2772a 1857b Serenut 5R 3565a 3133a Serenut 6T 3560a 3293a TMV 2 2343a 2136a TUFRunner TM 5510a 5049a Mean 3764 3621

PAGE 49

49 Table 2 4. Regression analysis of pod yield response for increasing mm of total water received at the Citra, FL location. The regression coefficient (b) represents genotypic pod yield response for total water recei ved across both years and irrigation treatments. Genotype Citra Yield Versus Total Water Received Regression Regression Coefficients (b) P value Model R 2 Georgia 06G 2.1 0.5773 0.04 FloRun TM 6.9 0.0293 0.47 TUFRunner TM 0.96 0.6619 0.0 3 C76 16 8.9 0.0696 0.35 NMVC 5.1 0.0153 0.54 Chico 3.2 0.0605 0.37 ICGS 76 4.2 0.1156 0.28 COC 041 3.7 0.0049 0.65 TMV 2 0.08 0.9359 0.00 Serenut 5R 2.8 0.1130 0.28 Serenut 6T 0.62 0.6866 0.02 ICGV 86015 1.2 0.2251 0.18 ICGV 86388 0.01 0.994 2 0.00 Flavorunner 458 6.8 0.0091 0.43

PAGE 50

50 Table 2 5. Average genotypic pod yield and stability across all three site locations. Regression coefficient (b) represents the pod yield s tability (GxExM) which is the average pod yield for each genotype versus the average pod yield for each site location, year, and irrigation treatment. Genotype Pod Yield and Stability Mean Regression Coefficients (b) R 2 C76 16 3814 0.88 0.25 ICGS 76 3289 0.90 0.42 Chico 2880 0.90 0.52 ICGV 86015 2752 0.93 0.49 Serenut 6T 2429 1.12 0.88 Sere nut 5R 2268 1.30 0.80 COC 041 2233 1.14 0.71 TMV 2 2143 0.90 0.76 ICGV 86388 2009 0.87 0.59 Mean 2646 0.99 0.60

PAGE 51

51 CHAPTER 3 ASSESSMENT OF ABOVE AND BELOW GROU ND TRAITS INTRINSIC TO DISPARATE PEANUT GENOTYPES FOR DETERMINING ADAPTABILITY TO SOIL HYDROLOGIC CONDITIONS Introduction Since 1950, peanut yield has increased approximately 46 kg ha 1 yr 1 (NASS, 2017). Major contributions to yield advances during this t ime have been attributed to the advancement of crop protection pesticides, disease tolerance, and the development of cultivars with greater yield potential (Isleib et al., 2001). Increasing yield potential by breeding new cultivars has been attributed to i ncreasing numbers of reproductive structures, reproductive efficiency (percentage of flowers resulting in pods and seeds), and seed weight (Coffelt et al., 1989; Seaton et al., 1992; Haro et al., 2013). An interesting historical change in global production practices occurred when the preference shifted from growing cultivars with erect canopy growth habits of fastigiata descent to cultivars with procumbent canopy growth habits of hypogaea descent (Haro et al., 2013). This shift to cultivars with procumbent canopy architectures has been linked to increases in LAI and light attenuation within the canopy, contributing to an overall increase in peanut above ground biomass and crop harvest index (Haro et al., 2017). Although there is much documentation on the inf luence peanut breeding has had on above ground traits, to our knowledge, no evidence exists on how peanut breeding trends have influenced root architecture. Changes in root architecture may also have been driven in commercial production by the shift from c ultivars of fastigiata descent to ones of hypogaea descent. Knowledge of subspecies root system architecture (RSA), and the relationship of this trait to above ground partitioning could be critical in genotypic selection for improving water deficit stress resiliency.

PAGE 52

52 Root trait responses that improve soil water acquisition and possibly reduce yield loss from water stress have been documented in the peanut literature. Many studies have reached similar conclusions showing that peanut genotypes with inherentl y deep root architectures, or that respond to water stress by proliferating roots deeper into the soil profile, have relatively greater pod yields, and under water stress conditions, are deemed as more water stress tolerant (Rucker et al., 1995; Songsri et al., 2008; Jongrungklang et al., 2011; Jongrungklang et al., 2012). These results suggest that water stress tolerance traits are attributed to not only a larger root system, but to proliferation of roots deeper into the soil profile. Further, the study co nducted by Jongrungklang et al. (2012) observed a negative relationship between pod yield and percent root length density (%RLD) in the top 30 cm of the soil when water stress was imposed during mid season development. This indicates that there may be a tr ade off between greater root production and yield under stress conditions. The interaction of genotype with environment and/or management has made it difficult to directly utilize the results of physiological studies on water stress in breeding programs ( Cattivelli et al., 2008) because the utility of a physiological adaptation depends on the specific environmental or management conditions the crop experiences. Water stress impacts vary in relation to the timing of stress with phenological development, the rate at which soil drying occurs to achieve water stress conditions, and the severity of water stress conditions imposed. Therefore, it is critical that experimental approaches to evaluate germplasm traits relative to water stress be consistent with the i ntended typical stress severity and timing for the region for which a specific cultivar is developed (Sinclair, 2011). Another critical factor when studying

PAGE 53

53 water stress resilience is to distinguish the difference between biological and agronomic water str ess tolerance. For example, although relative genotypic crop survival under severe water stress conditions may demonstrate water stress tolerance in a biological sense, survival of crop plants is irrelevant in a commercial setting because it does not neces sarily preserve harvestable yield and therefore economic viability for the farmer (Sinclair, 2011). Therefore, experiments that focus on trait assessment for improving water stress tolerance must be conducted within the region of interest, and under water stress conditions that are relevant to the particular deficit conditions commonly experienced in the area where the cultivar is to be produced. This ensures the development of cultivars with some level of local and relevant stress adaptation. One particul ar environment relevant to peanut production is the southeastern U.S. which encompasses an hectarage is planted in this area (NASS, 2015). The climate in this region is char acterized as humid, where the annual precipitation exceeds evaporation, and average seasonal precipitation is often greatest during the summer months coinciding with peanut production (Henry, 2005; SERCC, 2017). A large percentage of this region has soils with high sand content and low organic matter, characterizing them as well drained with little water holding capacity. Because of these edaphic conditions, water scarcity stress can occur quickly in periods of less frequent precipitation due to rapid soil water depletion; but dry periods are often intermittent due to the high probability of summer precipitation. However, in some drought prone years, the duration of water stress can be prolonged and cause more severe water scarcity stress. Mitigating the

PAGE 54

54 ris k of water scarcity in this region has led to an irrigation infrastructure on 60% of the harvested peanut hectarage, thus reducing the duration and the severity of crop water scarcity compared to rain fed production systems (NASS, 2014). However, both irri gated and rainfed production scenarios of this region have a heavy reliance on in season water supply from precipitation. An irrigated production environment also allows for more control of the level of water scarcity stress experienced by the crop over th e course of phenological development. For example, irrigated growers may choose to utilize regulated deficit irrigation (RDI), an irrigation practice designed to reduce the amount of water application below the full water requirement for optimum plant grow th (Chai et al., 2016). Developmental stage based RDI has been successful for peanut, particularly the use of reduced amounts of irrigation during early season vegetative growth, thus eliciting increased root proliferation deeper into the soil profile with out reducing pod yields (Rowland et al., 2012; Thangthon et al., 2016). However, root distribution and pod yield responses have been reported to vary across peanut genotypes demonstrating variation in phenotypic plasticity (Jongrungklang et al., 2011). Fur thermore, the phenotypic plasticity to increase root distribution deeper into the soil profile during vegetative water stress is likely heavily influenced by environmental conditions. This is of particular importance since success using vegetative RDI has been documented in arid environments (Rowland et al., 2012), or under controlled conditions with rainout shelters (Jongrungklang et al., 2011; Thangthon et al., 2016). In both scenarios, total water received is primarily through irrigation in contrast to p eanut production in humid

PAGE 55

55 environments where stochastic precipitation is the primary water source, and water stress is more likely to be mild. Major whole plant traits governing crop yield production under non stressed conditions are the amount of inciden t solar radiation which is transformed into chemical energy and partitioned into harvestable tissue (Hay and Porter, 2006). Leaf area index (LAI) largely influences the amount of intercepted solar energy, and therefore is positively related to dry matter p roduction when optimal growing conditions are present (Hay and Porter, 2006). However, under conditions of water deficit stress, stomatal closure reduces the amount of CO 2 substrate for photosynthesis which often reduces biomass production. A trait which c ould possibly mitigate water deficit stress is the RSA of a crop that includes deep rooting ability, allowing water acquisition from deep soil strata (Lynch and Wojciechowski, 2015). By characterizing LAI, RSA, and pod yields of peanut genotypes varying i n subspecies descent, knowledge of innate carbon partitioning can be obtained which could be useful for germplasm selection. Further evaluating these traits under various irrigation management strategies provides additional information on phenotypic plasti city which could possibly be utilized in determining genotypic adaptability to water deficit conditions, and possible irrigation management strategies which can be utilized for a particular production region. Therefore, the objectives of this field researc h trial were to: (i) assess phenotypic variation in RSA, LAI, and their relationship to pod yield among genotypes classified into peanut subspecies hypogaea and fastigiata ; and (ii) characterize the possible phenotypic interactions of these genotypic trait s with environmental variability under various irrigation management strategies relevant for a

PAGE 56

56 humid climate. We hypothesized that genotypic differences in root system architecture exist among subspecies of hypogaea and fastigiata descent which would provi de more or less resilience to water stress conditions, and that pod yield production would be negatively related to belowground traits due to trade offs in carbon allocation between root and shoot systems Furthermore, we suspected that above and below gr ound genotypic traits would interact with the range of irrigation treatments implemented in this study because of the disparate descents of these genotypes which may be adapted to different soil hydrological conditions. The information gained from this stu dy will provide knowledge of both above and belowground biomass partitioning between subspecies hypogaea and fastigiata that could be used for: i) selecting germplasm best adapted to a range in hydrologic conditions; and ii) evaluating the feasibility of utilizing reduced irrigation management strategies in humid climate peanut production regions. Materials and Methods Site Characterization : Field studies were conducted during 2015 and 2016 at n Unit (PSREU) in north central Florida (29 24' 38" N, 82 10' 12" W). The soil is classified as Candler sand (Hyperthermic, uncoated Lamellic Quartzipsamments). In both years, peanut was planted into a peanut cereal rye cotton rotation. Soil preparation consisted of conventional tillage approximately five weeks before planting. Immediately prior to planting, the field was surface tilled using a field cultivar with S tine sweeps. Seeding density of all peanut genotypes was 20 seeds m 1 with a row spacing o f 0.91 m. Plots consisted of four rows 7.5 m long. Planting was accomplished with a two row Monsem vacuum planter (Monosem Inc., Edwardsville, KS). Rhizobium inoculum (1.31 l ha 1 ; Advanced Biological Marketing, Van Wert, OH), and azoxystrobin fungicide (0.21 kg a.i.

PAGE 57

57 ha 1 ; Sygenta International AG, Basel, Switzerland) were applied in the seed furrow at planting. A soil test was conducted approximately six weeks prior to planting, and recommended nutrients were broadcast surface applied immediately after planting. Daily meteorological data was recorded using an automated weather station located within 1500 m of the experiment. Experimental Design : The experimental design consisted of a split plot arrangement in a randomized complete block design with four replications. The whole plots were irrigation treatments consisting of: (i) 1.9 cm per application (100%); (ii) a primed acclimation (PA) treatment consisting of 1.1 cm of water per application until mid bloom and then 1.9 cm of water for the remainder of the season (60%PA); (iii) 1.1 cm of water per application for the entire season (60%); and (iv) a rainfed control (RF). Irrigation was scheduled when tensiometers (Irrometer, Riverside, CA) placed in the 100% irrigation treatment of each replication reach ed an approximate average soil water tension of 25 kPa at a soil depth of 30 cm. Soil water tension level for irrigation scheduling was based on an estimated level of 50% maximum allowed depletion of 20 kPa for a sandy soil (NRCS, 2005). Tensiometer plac ement at 30 cm was selected as half of the estimated effective root zone. The effective root zone was a conservative estimate based on the FAO56 recommendation range of 0.5 1.0 m (Allen et al., 1998). The full irrigation (100%) amount (1.9 cm) was over est imated to ensure replenishment of soil water to a depth of 30 cm using an available water capacity of 0.04 cm H 2 O cm soil 1 (NRCS, 2017) assuming an irrigation efficiency factor of 0.80. Irrigation was applied using a lateral system equipped with variable rate irrigation (VRI) technology (Lindsay Corporation, Omaha, NE). Sub plots consisted of four

PAGE 58

58 genotypes, two Valencia genotypes ( Arachis hypogaea L. subsp. fastigiata Waldron var. fastigiata Waldron) COC 041 (PI 493631) and New Mexico Valencia C (NMVC) (R eg. No. 24, PI 565461), and two runner ( Arachis hypogaea L. subsp. hypogaea L var. hypogaea L ) commercial cultivars FloRun TM 127, PI 663993) and TUFRunner TM 131, PI 674432). Crop Measurements: Mini rhizotron tubes wer e used to examine RSA development over the growing season. Mini rhizotron tubes were installed directly under and parallel to the row at a 45 angle to the soil surface immediately after crop emergence using a hydraulic powered coring machine (Giddings Mac hine Company, Windsor, CO). At each measurement date, images were captured at 13.5 mm increments (resulting typically in 85 109 image frames) along the mini rhizotron tubes using a BTC 100X video camera and BTC I CAP image capture software (Bartz Technolog y Corporation, Carpinteria, CA). Total root length (TRL) and total surface area (TSA) analysis were conducted using WinRHIZO Tron software (Regent Instruments Inc., Quebec, Canada) by hand tracing of root segments within each image frame. Total length on e ach image frame was computed by estimating the linear length of each root segment. Total surface area was the product of total length and the cross sectional to the measur ed diameter of the projected root segment. In both years, root image sampling dates occurred at 21, 28, 35, 49, and 75 days after planting (DAP). Measurements of leaf area index (LAI) were recorded at 28, 35, 49, 64, and 77 days after planting (DAP) in bot h site years (LI COR, Lincoln, Nebraska) A single measurement of LAI for a plot was conducted by taking one above and four equidistant

PAGE 59

59 readings below the canopy as a perpendicular transect between the rows with the sensor view parallel to the row and repe ating this sequence with the sensor view perpendicular to the row to gain greater spatial averaging. A lens cap cover with a 45 angle was used on all measurement dates. Measurements were taken in early morning with a cast shadow over the plot area to prev ent diurnal bias and underestimation of LAI values. Plot lengths were measured after peanut emergence to determine the harvest length of the two center rows. Harvest was timed to optimal maturity as was determined by removing the exocarp and revealing the mesocarp layer of the pod using a pressure washer (Williams and Drexler, 1981). Pods were classified as either immature or mature by their mesocarp color; pods which were brown/black were classified as mature pods. When approximately 80% of the pods were mature, the peanut plants were dug/inverted using a two row peanut digger (Kelly Manufacturing Co., Tifton, GA). Inverted peanuts were allowed to dry in field until the canopy tissue dehydrated. Following this drying period, peanut pods were separated from the pegs using a hand thresher (Kincaid, Haven, Kansas), collected in burlap bags, and dried to approximately 10.5% using a forced air drying wagon. Each peanut market type had different digging/inversion dates in both years due to differences in maturity class. In 2015, the Runner and Valencia market types were inverted at 131 and 102 DAP, respectively. The 2016 inversion dates for the Runner and Valencia market types were 142 and 111, respectively. Statistical Analysis: Statistical analysis was performe d using SAS v. 9.4 statistical software (SAS Institute, 2013), with the GLIMMIX procedure computing an analysis of variance (ANOVA). Random effects of rep nested within year, and rep

PAGE 60

60 crossed with irrigation nested within year were included in the model; al l other factors were considered fixed. Repeated measures ANOVA was performed on measurements that were repeated over the growing season (TRL, TSA, and LAI). Covariance matrix th the lowest Akaike Information Criterion (AIC). A heterogeneous autoregressive (ARH1) covariance matrix structure was used for TSA and TRL measurements. An autoregressive (AR1) covariance matrix structure was used for LAI. Normality and homogeneity were visually assessed by graphing the residual distribution, scatter plot of residuals, and Q Q plot of residuals. Total root length and TSA data were heteroscedastic so a square root transformation was performed prior to running the repeated measures ANOVA. D ata were pooled over years and experimental treatments when appropriate as indicated by a non significant F test for all traits measured. Assessment of RSA was performed by summing parameters across 20 cm increments to a total soil depth of 80 cm. Root par ameters were summed across the entire 80 cm sampling depth for determining main and interactive effects not including the soil depth Least Significant Difference (LSD) at the P < 0.05 probability level. To assess the relationship between pod yield and LAI/TRL, r egression and correlation analyses were performed using PROC REG and PROC CORR in SAS v. 9.4 statistical software (SAS Institute, 2013). The last measurement dates at a pproximately 75 DAP for both LAI and TRL were used to examine the relationship of these variables with pod yield. These relationships were separated by year and examined across all genotype and irrigation treatments.

PAGE 61

61 Results Yearly precipitation and irrig ation : The annual precipitation between the two years was quite variable, with greater rates occurring during the 2015 growing season (834 mm) compared to the 2016 season (634 mm) (Figure 3.1). This resulted in substantial differences in the relative propo rtion of water received via irrigation and precipitation. The irrigation: precipitation ratio in 2015 and 2016 was 7 and 30%, respectively, in the 100% irrigation treatment (Table 3.1). The distribution pattern of precipitation also differed between the tw o years; for example, during the first 50 days after planting, approximately 125 mm more rainfall was received in 2016 compared to 2015. This difference is mostly attributed to one large precipitation event of 91 mm occurring on 5 June 2016. However, in bo th years soil water conditions required three irrigation treatments during the first 50 DAP. Precipitation following the first 50 DAP was substantial in 2015 and no additional irrigation was required (Figure 3.2). In contrast, very little precipitation was received from 50 95 DAP during the 2016 growing season, resulting in the need to apply seven irrigation treatments during this mid season period. Total Root Length: Differences in TRL growth were observed among the peanut genotypes by the first measureme nt day at 21 DAP (Table 3.2; Figure 3.3). Greater root growth occurred in COC 041 when compared to genotypes of hypogaea descent. At 28 DAP, greater separation in TRL occurred and the differences among genotypes continued to increase. Root growth rate appe ared to slow for most genotypes by 35 DAP, with maximum TRL achieved by the last measurement date at 75 DAP. Total root length measurements after 35 DAP showed delineations between the two peanut market types: TRL was similar between the two Valencia marke t types; however both were greater than the two runner market types. Differences in TRL growth also existed

PAGE 62

62 between the two runner market types with TUFRunner TM growth than FloRun TM Season long patterns of TRL also varied among peanut genotypes by soil depth (Table 3.3; Figure 3.4). From 0 20 cm soil depth, COC 041 had a greater TRL than genotypes of hypogaea descent across all measurement dates; while at the 20 40 cm depth, TRL was similar among genotypes, with the exception of FloRun TM exhibited a shorter TRL than the other genotypes. At 40 60 cm depth, COC 041 and NMVC had similar TRL which was longer than the TRL of FloRun TM TUFRunner TM zone. At 60 80 cm of soil depth, all genotypes had similar TRL until 35 DAP. The final TRL measurement at 75 DAP at this depth range was similar to the trends in the 40 60 cm depth: Valencia market type cultivars had longer TRL than the runner market type s, and genotypes within each market type were similar. Although these differences were observed for TRL, no significant differences in TSA where observed when making similar genotypic comparisons across the 20 cm soil depth increments over the growing seas on (Table 3.3; Figure 3.5). This result indicates that the TRL of genotypes of hypogaea descent generally have fewer large diameter roots, thus contributing to an overall similar amount of TSA for the genotypes. Leaf Area Index : The results in both years show that early LAI development was not impacted by irrigation, but by 64 DAP, the 100% irrigation treatment had larger LAI than the RF (Table 3.2; Figure 3.6a). By 77 DAP, the 60%PA had achieved greater LAI than the RF treatment as well. Genotypic differe nces in LAI emerged by 77 DAP, a pattern that was consistent between years (Figure 3.6b). At 77 DAP, both runner market

PAGE 63

63 types, FloRun TM TM other, and both had greater LAI than the two Valenci a market types, NMVC and COC 041, both of which had similar LAI values. Pod Yields : Irrigation treatments ultimately influenced peanut pod yields, a response that was similar across growing seasons and genotypes (Table 3.4). Treatments which received irri gation were statistically similar to one another, although all had greater yields than the RF treatment (Table 3.5). Irrigation treatments of 100%, 60%PA, and 60% had higher pod yields relative to the RF control of 23, 15, and 17%, respectively. Year also had a significant influence on genotypic pod yield (Table 3.6). In both years, yield followed an expectation of greater levels for the longer maturing differed between the two representative genotypes, with pod yield of FloRun TM being 672 kg ha 1 greater than TUFRunner TM among all genotypes in 2016, a likely impact of less mid season precipitation, resulting in yield declin es, when comparing 2016 to 2015, of 21, 30, 68, and 59% for FloRun TM TM Morphological Partitioning and Pod Yields : Leaf area index was a significant predictor of pod yield across all genotypes and irr igation treatments in both years (Figure 3.7). However, variability in the predictive strength of this relationship was observed between the two growing seasons, with R 2 values of 0.52 and 0.15 in 2015 and 2016, respectively. This positive relationship had regression equations of yield = 241 + 946*LAI (P < 0.0001) and yield = 281 + 704*LAI (P = 0.0015) in 2015 and 2016, respectively. There was also a relationship between yield and TRL, showing R 2 values

PAGE 64

64 of 0.15 and 0.21. This relationship was negative wit h regression equations of yield = 4398 1.10*TRL (P = 0.0015) and yield = 3422 1.37*TRL (P = 0.0002) in both 2015 and 2016, respectively (Figure 3.8). Discussion Reports have documented that the introduction of procumbent canopy growth habits of hypog aea over erect canopy growth habits of fastigiata descent have increased HI (Haro et al., 2013). Furthermore, Haro et al. (2017) documented that increases in HI were supported by increases in LAI, radiation intercepted by the crop, and a longer crop growin g cycle. Greater LAI of genotypes with procumbent growth habits was also observed in this study. This study further advances the growth habit knowledge base of these two peanut subspecies by quantifying both above growth and below ground traits. Genotypic RSA differences were observed between the disparate subspecies indicating that their belowground traits are inherently different. Specifically, genotypes of fastigiata descent had greater root growth following 35 DAP which lasted until the final root measu rement at 75 DAP. Additionally, the greater amount of TRL with genotypes of fastigiata descent occurred primarily at deeper soil depths ranging from 40 60 and 60 80 cm. Genotypic differences within subspecies hypogaea were observed in TRL (TUFRunner TM TM rhizotron area. However, TSA was similar among all genotypes indicating two different RSAs among the two subspecies. These architectures include subspecies fastigiata having possibly greater TRL as a r esult of an increased number of lateral roots, whereas genotypes of hypogaea have fewer, more coarse roots. Genotypic differences in TRL within hypogaea could also be a result of increased genetic variation within this subspecies due to more extensive bree ding and selection.

PAGE 65

65 Both TRL and LAI observed on the final measurement date were correlated with pod yields. In both years, there was a significant weak negative relationship between pod yield and TRL. This indicates that part of the increased HI of the hypogaea subspecies, and selection for high yielding cultivars, may have been achieved by partitioning less carbon to root systems. Studies comparing wheat landraces to modern cultivars, which have been extensively selected for yield, have observed decreas es in root biomass of the latter (Waines and Ehdaie, 2007). A positive relationship between pod yield and LAI was also observed in both years of this study. However, the relationship was much stronger (r=0.72) in 2015, whereas this relationship was weak (r =0.39) in 2016. A possible explanation for this is the variability in yearly environmental conditions. The 2015 season was characterized as relatively wet, with no irrigation applied after 47 DAP. Under non water deficit conditions, biomass production is p redominantly influenced by the fraction of intercepted photosynthetically active radiation (iPAR) and the efficiency of conversion of that solar energy into chemical energy (Charles Edwards, 1982; Monteith, 1977). The interception of iPAR is largely a func tion of LAI and canopy architecture (Haro et al., 2017). However, under water deficit conditions the major limitation on biomass production is not iPAR but the lack of CO 2 substrate stemming from stomatal closure (de Wit, 1958; Donald and Hamblin, 1976; Bl um, 2009). The yearly precipitation patterns influenced water deficit stress severity and ultimately impacted genotypic pod yields. When averaged across all genotypes, yield showed a 39% reduction in 2016 compared to 2015. This is likely due to the more severe water deficit conditions in 2016. Additionally, a year x genotype interaction

PAGE 66

66 provides evidence of differential pod yield susceptibility to water stress conditions. When comparing pod yields of 2016 to 2015, genotypes of subspecies fastigiata had yi eld declines of 64% when compared to a 25% pod yield reduction with genotypes of hypogaea descent. Genotypes of fastigiata descent may have greater susceptibility to pod yield reductions when subjected to more severe water deficit conditions. Interestingly these genotypes were those which had the largest amount of TRL growth at deeper soil depths (40 60 and 60 80 cm). This was surprising and contrary to other studies that examine root traits in relation to drought stress. For example, Koolachart et al. (20 13) examined the impacts of terminal water stress on peanut genotypes and demonstrated that peanut genotypes which maintain greater pod yields under water deficit stress are those which have more roots located deeper in the soil profile, and the ability to maintain transpiration rates. Other studies have demonstrated large variation among peanut genotypes in the fraction of transpirable water content when transpiration rates begin to decline due to soil drying (Devi et al., 2009). Therefore, it is plausible that, despite having greater amounts of TRL deeper in the soil profile, genotypes of fastigiata descent may have a greater sensitivity to soil drying leading to fast stomatal closure, thus necessarily limiting their pod yield potential under water deficit conditions. Furthermore, the possible fewer number of coarse lateral roots with subspecies hypogaea could be a more efficient strategy for maintaining water uptake under soil drying conditions in this environment. Large diameter roots may allow for improv ed root to soil water contact and hydraulic conductance under soil drying conditions in the porous sandy soils of this study.

PAGE 67

67 The sensitivity of above and belowground traits to change in soil hydrological conditions was further evaluated in this study by examining these responses across several of irrigation management strategies. Beginning at 64 DAP, a general trend of reductions in LAI with decreasing amounts of applied irrigation water became evident. However, irrigation had no influence on RSA develop ment, suggesting that leaf area development is more sensitive to water stress conditions than root growth. The sensitivity of leaf growth rate and an ability of roots to grow despite water stress has been documented (Boyer, 1968; Westgate and Boyer, 1985). Furthermore, Sharp et al. (1988) showed that, although reductions in leaf and root elongation rates can occur at the onset of soil drying in a controlled environment, continued soil drying reduced elongation of leaf tissues more than root tissues. Our res ults confirm the greater sensitivity of peanut canopy growth versus root growth to water deficit stress. This research also supports that, despite the inherent diversity among root and canopy morphology of these peanut subspecies, differential genotypic ro ot and canopy responses may not be evident at the levels and duration of water stress conditions which typically occur in a humid climate. A critical factor influencing root growth responses to irrigation management is the severity of water stress deficit at particular times in phenological development. The lack of root response to irrigation management factors in the present study could be due to similar amounts of irrigation applied during the first 50 DAP in both years; the period when peanut root devel opment was most rapid and the overall seasonal architectural pattern is established. In both years, three irrigation treatments were applied during this time period which is relatively little water in proportion to the amount of precipitation

PAGE 68

68 received. Thi s could have contributed to the contrasting results of the current study with those observed by Rowland et al. (2012), where reduced irrigation early in the growing season resulted in increased root growth deeper into the soil profile. The arid climate whe re this study was conducted required a greater proportion of irrigation to precipitation. In both years of the current study, the three irrigation treatments prior to 50 DAP were applied during a short period of time resulting in the stress severity and du ration being fairly mild or moderate. Plants which undergo moderate levels of water stress have been reported to have little difference in growth response compared to plants which have had no water stress (Padilla et al., 2009). A meta analysis conducted b y Poorter et al. (2012) demonstrated that a reduction in plant biomass and an increase in root mass fraction are often attributed to more severe water stress of longer durations. The authors hypothesized this to be an important adaptation of plants, becaus e rapid alteration of growth partitioning may result in growth habits that are sub optimal after the alleviation of water stress. This is applicable to the irrigated humid peanut production environment of this study where moderate intermittent water stress is often relieved by in season rainfall events. Therefore, rapid partitioning of carbon for deeper roots alone could be costly and may be less able to acquire surficial soil water. Conclusion This research demonstrates that different inherent above and b elowground traits exist among peanut genotypes of fastigiata and hypogaea descent, giving further insight into their whole plant carbon partitioning and adaptability to varying soil hydrological conditions. It has been documented that the greater yield pot ential of hypogaea subspecies is largely due to an increased crop cycle duration, and improvement of canopy dynamics to support greater biomass production and HI. This research also

PAGE 69

69 concludes that part of the greater above ground biomass production with hy pogaea subspecies may be due reduced amounts of TRL when compared to genotypes of fastigiata descent. Consequently, it could be hypothesized that decreases in TRL, particularly deeper in the soil profile, would result in the hypogaea subspecies to be less water deficit stress resilient. However, the results of this research demonstrate the contrary, at least to the level of water deficits experienced, where the pod yield reduction under increasing water deficit stress was greater in the fastigiata subspecie s. This could be due to varying water acquisition efficiencies among the contrasting root architectures of the two subspecies. That is, fewer coarse roots may be more effective at water acquisition in a humid environment, particularly in soils possessing a high sand content, low water holding capacity, and high frequency surficial water inputs. However, it is important to note that greater water acquisition of a particular RSA may not result in greater water utilization of a crop genotype which has high sto matal sensitivity to soil drying. Future research which links peanut phenotypic data with crop water use function will provide greater understanding of genotypic variation in water acquisition and utilization. This would likely benefit selection methods fo r improving agronomic water deficit stress resiliency.

PAGE 70

70 Figure 3 1. Cumulative precipitation from peanut planting to harvest in 2015 and 2016 at the Plant Science Research and Education Unit (PSREU) near Citra, FL. Vertical arrows pointing dow nward indicate when irrigation was applied. Dashed line represents 50 DAP when 60%PA irrigation treatment was similar to the 100%. Vertical arrows pointing up represent our two harvest dates for each peanut market type in both years (Abbreviations: 100%, 1 .9 cm per application; 60%PA, 1.1 cm per application until 50 DAP, followed by 1.9 cm per application for the remainder of the growing season).

PAGE 71

71 Table 3 1. Amounts of water received through irrigation, irrigation plus precipitation, and the irrigation to p recipitation ratio in 2015 and 2016 for each irrigation treatment at the PSREU near Citra, FL. DAP, followed by 1 .9 cm per application for the remainder of the growing season; 60%, 1.1 cm per application for the entire growing season; RF, rainfed control. Irrigation Water Applied Total Water Received Irrigation: Precipitation 2015 2016 2015 2016 2015 2016 Irrigation Treatment ------------------------------m m ------------------------------% --------100% 57 191 894 825 7 30 60%PA 34 168 871 802 4 27 60% 34 114 871 748 4 18 Rainfed 0 0 837 634 0 0

PAGE 72

72 Figure 3 2. Volumetric soil water content (left axis) and daily precipitation (right axis) during time periods when irrigation was applied in 2015 and 2016 at the PSREU near Citra, FL. (Abbreviations: 100%, 1.9 cm per application; 60%PA, 1.1 cm per application until 50 DAP, followed by 1.9 cm per application for the remainder of the growi ng season).

PAGE 73

73 Table 3 2. Analysis of variance (ANOVA) for response variables of TRL, TSA, and LAI. Root parameter data was summed across soil depths prior to running ANOVA. IR, irrigation treatment; GENO, genotype treatment; TM, time repeated measures within each growing season. Effect Total Root Length Total Surface Area Leaf Area Index year 0.3323 0.5180 0.0 145 IR 0.6748 0.7169 0.1053 year*IR 0.8525 0.8403 0.9444 GENO <0.0001 0.1122 0.0007 year*GENO 0.9164 0.7960 0.0007 IR*GENO 0.2123 0.1267 0.1236 year*IR*GENO 0.8085 0.9005 0.6393 TM <0.0001 <0.0001 <0.0001 year*TM <0.0001 <0.0001 0.3319 IR*TM 0.405 0 0.1295 0.0006 year*IR*TM 0.1266 0.1424 0.3961 GENO*TM 0.0014 0.4635 <0.0001 year*GENO*TM 0.7099 0.6589 0.2444 IR*GENO*TM 0.9508 0.9731 0.9908 year*IR*GENO*TM 0.8519 0.8154 0.8029

PAGE 74

74 Figure 3 3. Average total genotypic root length growth at 21, 28, 35, 49, and 75 DAP at the PSREU. Data is averaged across irrigation treatm ents and year. Error length is the sum of root lengths across the entire mini rhizotron tube to a soil depth of 80 cm (Abbreviations: NMVC, New Mexico Valencia C).

PAGE 75

75 Table 3 3. Analysis of variance (ANOVA) for response variables of TRL and TSA. Response variables were summed across 20 cm soil depth increments prior to runnin g ANOVA. treatment; GENO, genotype treatment; TM, time repeated measures within each growing season; Depth, sum of root parameters across 20 cm depth incre ments. Effect To tal Root Length Total Surface Area Year 0.1160 0.9928 IR 0.5521 0.3708 Year*IR 0.7393 0.9243 GENO <0.0001 0.0247 Year*GENO 0.8816 0.1852 IR*GENO 0.0123 0.3944 Year*IR*GENO 0.4221 0.0173 Depth <0.0001 <0.0001 Year*Depth 0.0447 0.2308 IR*Depth 0.49 82 0.9184 Year*IR*Depth 0.9885 0.2108 GENO*Depth 0.8995 0.9552 Year*GENO*Depth 0.3072 0.9196 IR*GENO*Depth 0.8337 0.8977 Year*IR*GENO*Depth 0.9851 0.2329 TM <0.0001 <0.0001 Year*TM <0.0001 0.0006 IR*TM 0.0689 0.0149 Year*IR*TM 0.0522 0.0583 GENO* TM <0.0001 0.0006 Year*GENO*TM 0.7531 0.4423 IR*GENO*TM 0.4807 <0.0001 Year*IR*GENO*TM 0.2078 0.3791 Depth*TM <0.0001 <0.0001 Year*Depth*TM <0.0001 0.0202 IR*Depth*TM 0.3291 0.1173 Year*IR*Depth*TM 0.5564 0.5885 GENO*Depth*TM 0.0345 0.9829 Year*GE NO*Depth*TM 0.6818 0.8541 IR*GENO*Depth*TM 0.9703 0.1983 Year*IR*GENO*Depth*TM 0.4530 0.9237

PAGE 76

76 Figure 3 4. Average genotypic root length summed across 20 cm depths at 21, 28, 35, Significant Difference at P < 0.05. Data is the averaged across irrigatio n treatments and years (Abbreviations: NMVC, New Mexico Valencia C).

PAGE 77

77 Figure 3 5. Average genotypic root surface area summed across 20 cm depths at 21, 28, 35, 49, and 75 DAP at the PSREU. No multiple comparisons error bars are dis played because F test results had a P > 0.05. Data is the averaged across all irrigation treatments and years. (Abbreviations: NMVC, New Mexico Valencia C).

PAGE 78

78 Figure 3 6. A) Average leaf area index (LAI) at 28, 35, 49, 64, and 77 DAP fo r each irrigation treatment and B) genotype at the PSREU. Data is averaged across years for graphs A & B (Abbreviations: 100%, 1.9 cm per application; 60%PA, 1.1 cm per application until 50 DAP, followed by 1.9 cm per application for the remainder of the g rowing season; 60%, 1.1 cm per application for the entire Significant Difference at P < 0.05).

PAGE 79

79 Table 3 4. Analysis of variance (ANOVA) for pod yield response variables at the PSREU near Ci tra, FL. GENO, genotype treatment; TM, time repeated measures within each growing season. Effect Pod Yield year 0.0022 IR 0.0054 year*IR 0.7077 Geno <0.0001 year*Geno 0.0301 IR*Geno 0.1581 year*IR*Geno 0.2575

PAGE 80

80 Table 3 5. Average pod yields for each irrigation treatment a t the PSREU near Citra, FL. Yields are averaged across peanut genotypes and years. DAP, followed by 1.9 cm per application for the remainder of the growing season; 60%, 1.1 cm per application for the entire growing season; RF, rainfed control. Pod Yields Irrigation Treatment Pod Yield ------kg ha 1 ------100% 3103 60%PA 2801 60% 2886 RF 2382 LSD (0.05) 369

PAGE 81

81 Table 3 6. Average pod yields for each genotype and growing season at the PSREU near Citra, FL. The percent difference is the percentage of yield decline from 2015 to 2016. Significant Difference at P < 0.05. Peanut Pod Yields Peanut Genotypes 2015 2016 Percent Difference --------------kg ha 1 ------------------------% ------FloRun TM 4718 3731 21 TUFRunner TM 4373 3059 30 NMVC 2532 815 68 COC 041 2227 907 59 LSD (0.05) ----------------609 -----------------

PAGE 82

82 Figure 3 7. Relationship between peanut pod yield and maximum leaf area index a chieved for the 2015 and 2016 growing season at the PSREU near Citra, FL. Individual data points represent all genotypes and irrigation treatments for each year.

PAGE 83

83 Figure 3 8. Relationship among pod yield and maximum total root length achieved for th e 2015 and 2016 growing season at the PSREU near Citra, FL. Individual data points represent all genotypes and irrigation treatments for each year.

PAGE 84

84 CHAPTER 4 COORDINATED ADAPTIVE PHENOTYPING FOR IMPROVING SOIL WATER ACQUISITION AND UTILIZATION Introducti on Drought stress is responsible for more crop loss than any other abiotic or biotic stress, thus making tolerance to water scarcity a critical goal in many breeding programs (Borm et al., 2012). But progress has been slow, primarily due to both the compl (Mwadzingeni et al., 2016), and because of the breadth of individual plant processes Sinc lair, 2011; Turner et al., 2014). In addition, the concept of successful drought tolerance in crop species diverges from many established paradigms regarding plant drought tolerance in general. This is because operationally, drought tolerance has been repe atedly defined as the ability to survive under research scenarios that compare well watered conditions to severe water reductions. However, this definition and method of evaluation is not relevant to crop species because they must do more than survive th ey must continue to yield under water scarcity, necessarily making genetic and physiological mechanisms of yield maintenance inherently different than those of survival (Skirycz et al., 2011). All these factors have led to the lack of true breakthroughs in solving the issue of breeding for crop drought tolerance. To successfully phenotype crop genotypes for water deficit stress tolerance, it is important to examine responses in both above and below ground traits, but perhaps more importantly, to assess the ir coordinated functioning. One of the most impactful aboveground traits which influences crop yield under water deficit stress is the degree of transpiration reduction which occurs under water deficit stress (de Wit, 1958;

PAGE 85

85 Passioura, 1996; Blum, 2009). Th e ability to maintain transpiration under dehydrating conditions allows for continued assimilation of CO 2 during photosynthesis, which is required for growth processes. Variability exists among peanut genotypes for the sensitivity of transpiration reductio ns under soil drying (Devi et al., 2009). Utilizing the relationship between stomatal closure and carbon isotope discrimination, Hubick et al. (1988) found a broad sense heritability of 81% in carbon isotope discrimination, suggesting that transpiration us e efficiency (a trait influenced by stomatal closure) is a heritable trait which can be utilized in breeding programs. These results indicate that selection and breeding for stomatal sensitivity to water deficit stress can be used to mitigate yield loss, o r possibly conserve water to avoid lethal levels of crop desiccation. For critical belowground traits, root system architecture (RSA) encompasses a suite of traits which have the potential to provide genotypic water deficit stress tolerance. For peanut, g enotypes with inherently deep root architectures and the ability to proliferate roots deep into the soil profile under dehydrating conditions have relatively greater pod yields under water stress conditions (Rucker et al., 1995; Songsri et al., 2008; Jongr ungklang et al., 2011; Jongrungklang et al., 2012; Koolachart et al., 2013). But through the increased emphasis on examining roots traits in general, research is showing that an examination that focuses solely on root length and depth is not correctly pred icting the complex response to soil drying and thus the overall tolerance to water deficits. Similarly for peanut, the relationship between extensive root architectures and dehydration tolerance is not always clear, including Ratnakumer and Vadez (2011) wh o reported a weak positive relationship between peanut root length density (RLD) and total water uptake under various water stress conditions. Therefore, concluding that

PAGE 86

86 deep and extensive RSAs are the primary or only phenotypes appropriate for drought tol erance is misguided. It is clear that individually, above and belowground traits play varying roles in providing adaptability to various dehydration vulnerability scenarios (DVS). However, what may be even more critical is the integration of function ac ross above and belowground processes. This idea is central to succeeding at identifying water deficit stress tolerant phenotypes because the entire water conductivity system of a plant is driven by transpiration, through a coordinated process of root wate r uptake from the soil, transport in the plant vascular system, and eventual water loss at the leaf surface. For example, a major resistance to transpiration rates under soil drying conditions is stomatal limitation, which would diminish root water uptake, regardless of where roots are distributed in the soil profile. Therefore, a genotype with an extensive root system that is able to access deep water resources but closes stomates quickly in a water scarce condition will not be able to fully capitalize on its extensive root system. In fact, while sensitive stomatal closure under developing soil drying conditions combined with a deep RSA may be excellently adapted for survival under drought, we could define this genotype as being maladapted to a commercial p roduction setting because rapid stomatal closure will not allow this genotype to efficiently utilize the energy invested into an extensive RSA, likely causing limits to yield potential. On the contrary, a shallow rooted genotype with delayed stomatal sensi tivity to soil drying would result in relatively low energy investment to RSA and thus increased yield potential. However, this same genotype would likely require adequate and reliable water application. Overall, agronomic water stress tolerance cannot be attributed to a single above or

PAGE 87

87 belowground trait, but that a coordinated combination of above and belowground traits is required to maintain yield under water stress conditions. While coordination of above and belowground traits is essential for deter mining water scarcity tolerance, the degree of variability in the coordination of stomatal aperture control coupled with RSA could be useful in determining specific adaptability to a particular (DVS). The first objective of this research was to: (i) develo p an in situ approach to quantify the coupled influence of root system architecture (RSA), soil water depletion (SWD), and stomatal limitation on soil water acquisition distribution within the soil profile over a series of soil wetting and drying; and (ii) determine if genotypic RSA responses to differential early season water management has a lasting effect on mid season soil water depletion zones and transpiration fluxes over a series of soil wetting and drying cycles. We hypothesized that under well wate red conditions with no stomatal limitation, soil water depletion would primarily occur at the soil depth with the greatest root density. Additionally, the degree of RSA plasticity to differential early season water management would vary among peanut genoty pes, and would ultimately determine the level of genotypic water deficit stress tolerance. Materials and Methods Site Characterization: A field study was conducted during 2016 at the University h Central Florida (29 24' 38" N, 82 10' 12" W). The soil is classified as Candler sand (Hyperthermic, uncoated Lamellic Quartzipsamments). Soil preparation consisted of conventional tillage approximately five weeks before planting. Immediately prior to p lanting, the field was surface tilled using a field cultivator with S tine sweeps. Planting was accomplished with a two row Monsem planter (Monosem Inc., Edwardsville, KS) at a seeding density for

PAGE 88

88 all peanut genotypes of 20 seeds m 1 with a row spacing of 0.91 m; plots consisted of four rows 6.0 m long. Rhizobium inoculum (1.31 l ha 1 ; Advanced Biological Marketing, Van Wert, OH)), and azoxystrobin fungicide (0.21 kg a.i. ha 1 ; Sygenta International AG, Basel, Switzerland) were applied in the seed furrow at planting. A soil test was conducted approximately six weeks prior to planting, and recommended nutrients were applied on the surface immediately after planting. Experimental Design: The experimental design consisted of a split plot arrangement in a ran domized complete block design with three replicates. The whole plots were irrigation treatments consisting of: 1) 1.9 cm per application for the entire season scheduled using tensiometers (100%); and 2) a primed acclimation (PA) treatment consisting of 1.1 cm of water per application for the first 50 days after planting (DAP) followed by 1.9 cm applications for the remainder of the season (60%PA), with all scheduling of irrigation events following the 100% treatment. Irrigation was scheduled when tensiomete rs placed in the 100% irrigation treatment of each replication reached an approximate average soil water tension of 25 kPa at a soil depth of 30 cm. Irrigation was applied using surface drip irrigation placed directly on the row with a 30 cm emitter spaci ng producing 3721 L hr 1 100 m 1 at 69 kPa (JAIN Irrigation Inc, Haines City, FL). Sub plots consisted of two genotypes, a Valencia genotype ( Arachis hypogaea L. var. fastigiata ) COC 041 (PI 493631), and a Runner ( Arachis hypogaea L. var. hypogaea ) commerc ial cultivar TUFRunner TM Rainout shelters were used in this experiment to exclude rainfall from the plots to ensure differential irrigation treatments were precise and that a series of soil drying events at controlled levels could be applied durin g the peak growth period when in

PAGE 89

89 depth physiological characterization took place. These physiological measurements occurred from approximately 60 90 DAP. This time period was established for two primary reasons: 1) RSA in peanut is well established and at its peak extension by 60 DAP; and 2) this time period coincides with the phenological development period when peanut is most susceptible to pod yield loss due to water deficit stress. Testing during these dry down periods allowed for a comparison of how ea rly season irrigation management influenced RSA, and if the established RSA has lasting impacts on mid season soil water acquisition. Crop Measurements : Two mini rhizotron tubes were installed in each plot directly into and parallel to the row at a 45 ang le with the soil surface immediately after crop emergence using a hydraulic powered coring machine (Giddings Machine Company, Windsor, CO). At each measurement date, images were captured at 13.5 mm increments along the mini rhizotron tubes using a BTC 100X video camera and BTC I CAP image capture software (Bartz Technology Corporation, Carpinteria, CA). Total root length (TRL) and total root volume (TRV) analysis was conducted using WinRHIZO Tron software (Regent Instruments Inc., Quebec, Canada) by hand tr acing of root segments within each image frame. Total root length on each image frame was computed by estimating the linear length of each root segment, while TRV was the 2 2 *TRL) determi ned from the traced images. Root image sampling dates occurred at 60, 67, 75, 82, and 88 days after planting (DAP) to coincide with the dry down periods. Measurements of leaf level gas exchange were recorded using a LICOR 6400XT portable photosynthesis sys tem at these same time periods coinciding with root

PAGE 90

90 imaging, specifically at 61, 62, 68, 69, 75, 76, 82, 83, 89, 90, 95, and 96 DAP (LICOR, Lincoln, Nebraska) Two separate leaf measurements were recorded on the upper most mature leaf (second nodal positio n) located on the terminal branch, using one leaflet from each leaf in order to fill the measurement chamber. Leaves were selected from plant canopies whose RSA was being quantified using the mini rhizotrons. Chamber conditions were consistent across sampl ing date, including a block temperature of 30 C, CO 2 concentration of 400 ppm, and a relative humidity maintained from 65 75%. The LED light source was set to 1800 mol m 2 s 1 The chamber flow rate was set to 500 mol s 1 Concomitant with root imaging and gas exchange measurements, VSWC was quantified using time domain reflectometry (TDR) with the TRIME PICO IPH/T3 probe (IMKO, Ettlingen, Germany). The probe sensor head was factory calibrated for the mini rhizotron tube cellulose acetate butyrate (CAB) material. This allowed for quantification of the VSWC around each mini rhizotron tube. The probe head measures the average VSWC at approximately 15.5 cm increments. These measurements were recorded on the same days as the gas exchange measurements: 61, 62 68, 69, 75, 76, 82, 83, 89, 90, 95, and 96 DAP. Soil water depletion was estimated by taking the difference between pre dawn and pre dusk for each quantified soil depth at 61, 68, 75, 82, 89, and 96 DAP. Statistical Analysis: Statistical analysis was pe rformed using SAS v. 9.4 statistical software (SAS Institute, 2013). The GLIMMIX procedure was used to compute analysis of variance (ANOVA). Random effects of rep, and rep crossed with irrigation were included in the model. A repeated measure ANOVA was per formed on

PAGE 91

91 measurements that were repeated over the growing season (TRL, TRV, SVWC, gas exchange, SWD), with a covariance matrix structure with the lowest Akaike Information Criterion (AIC). An autoregressive (ARH) covariance matrix structure was used for T RL and TRV measurements. A heterogeneous autoregressive (ARH1) covariance matrix structure was used for leaf level transpiration rates, SWD, and VSWC measurements. Normality and homogeneity were visually assessed by graphing the residual distribution, scat ter plot of residuals, and Q Q plot of residuals. The TRL and TRV traits were heteroscedastic, so a square root transformation was performed prior to running the repeated measures ANOVA. Data were pooled over experimental treatments when F tests were not s ignificant (specifically on irrigation treatments of SWD and Significant Difference (LSD) at the P < 0.05 probability level. To assess the bi variate relationships betwee n transpiration fluxes, root parameters, and soil water depletion, regression analysis was performed using SAS JMP Pro v. 12.2.0 statistical software (SAS Institute, 2015). A quadratic regression model was fit to the relationship between the daily transpir ation flux and total soil water depletion summed to a soil depth of 70 cm. A linear regression model was fit to the relationship between TRV with soil water depletion for each genotype. A spearman correlation analysis was performed assessing the relationsh ip between root parameters and SWD for all quantified soil depths and measurement dates. A spearman correlation was also used to estimate the relationship between VSWC and transpiration for each genotype across all measurement dates. A linear regression mo del and Pearson

PAGE 92

92 correlation analysis were performed examining the relationship between transpiration rates and pod yield Results Root System Architecture: The RSA varied among the two peanut genotypes in this study (Table 4.1). In general, genotype COC 041 had a greater TRL than TUFRunner TM TUFRunner TM 10 1 Genotypic RSA also interacted with soil depth. Both genoty pes COC 041 and TUFRunner TM while COC 041 had greater TRL at soil depths below 20 cm. Because it had more coarse roots, the TUFRunner TM 20 cm of soil depth was gr eater than COC 041 (Table 4.1), but below 20 cm depth, the TRV was similar among the two genotypes (Figure 4.2 b). Early season irrigation patterns influenced the genotypic RSA (Figure 4.3 a & b). The genotype COC 041 responded to the 60%PA treatment by h aving greater TRV at the deepest quantified soil depth range of 60 80 cm, while the TRV in the top 60 cm of soil was not affected for this genotype (Figure 4.3a). Similarly, early season irrigation treatments affected the RSA of TUFRunner TM with the 100% irrigation treatment having greater TRV in the soil depth ranges of 0 20 and 20 40 cm when compared to the 60%PA irrigation treatment, with the TRV for this genotype reduced in the 100% irrigation treatment at 60 80 cm of soil. Soil Water Depletion and Transpiration: The distribution of soil water depletion across the soil profile varied over the growing season and was dependent on the overall VSWC (Figure 4.4 a&b). Under high VSWC, the estimated percent of soil water

PAGE 93

93 depletion occurring in the top 30 cm of soil was 86, 66, and 71% at 61, 75, and 89 DAP, respectively. As the soil dried, most soil water depletion occurred at the depth of 30 60 cm. On these dates of 68, 82, and 96 DAP, the estimated percent of soil water depletion at 30 60 cm was 59, 67, and 76%, respectively. Significant correlations were observed between all rooting parameters and SWD. The measurement dates of 75 and 82 DAP which had contrasting VSWC levels demonstrate the general trend between SWD and the rooting parameters that was observed (Tables 4.2 & 4.3). The slope direction of the relationship was dependent on VSWC, such that on days where VSWC was high, and root uptake was predominately in the top 30 cm of soil, the amount of SWD increased with decreasing amount of ro ots root (negative slope) (all parameters). A positive slope was observed between SWD and root parameters when VSWC was low and the predominant amount of SWD was at soil depths greater that 30 cm. When examining transpiration rates on these same dates, TUF Runner TM had greater mid morning transpiration rates than COC 041 on the majority of the measurement dates (Table 4.4; Figure 4.5). At 61 and 83 DAP; TUFRunner TM greater mid morning transpiration rates despite similar values of VSWC from 0 70 cm of soil. However, greater VSWC with TUFRunner TM 90, and 96 DAP indicating less soil water depletion with this genotype. On days following irrigation treatments (75, 76, and 89 DAP) when VSWC had increased to high levels the mid morning transpiration rates among the two genotypes were similar. During soil dry down periods (62, 68, 69, 82, 90, 96, and 97 DAP) greater mid morning transpiration rates were observed in TUFRunner TM whe n a severe level of water deficit occurred (VSWC= 2.7%). The correlation between

PAGE 94

94 VSWV and transpiration showed a positive correlation for both genotypes (COC 041, r= 0.48; TUFRunner TM When examining the relationship between the daily transpiration flux and total soil water depletion summed to a soil depth of 70 cm (Figure 4.6), moderately high coefficients of determination (R 2 =0.61 and 0.67) were estimated for the genotypes COC 041 and TUFRunner TM tic regression equation was TR= 1.102 + 0.911*SWD 0.040*SWD 2 (P< 0.0001) for COC 041, and TR= 2.292 + 0.958*SWD 0.044*SWD 2 (P< 0.0001) for TUFRunner TM linear relationship between TRV and SWD, an R 2 =0.17 (TRV= 74.16 + 3.527*SWD ; P= 0.021) and 0.00 for genotypes COC 041 and TUFRunner observed. These results demonstrate that soil water depletion is more strongly related to daily transpiration flux than total root system size. Discussion To date, research interests focusing on examining variation in genotypic root architecture and plasticity has gained momentum due to its potential for identifying traits which could improve crop resilience to both impaired water and nutrient availability and ultimately cont ribute to resource use efficiency. However, very few of these studies attempt to quantify the relationship between root architecture and water uptake which leads to many assumed causal links regarding physical root presence and water uptake capacity. The l ack of directly relating root presence with quantified measures of water uptake is due to many logistical and expense restraints which make this measurement challenging. Estimations of crop water use using weighing lysimeters and sap flow rates are not onl y expensive (reducing the number of genotypes in which crop water use can be estimated), but do not provide information on the distribution of soil water depletion

PAGE 95

95 in relation to root water uptake. Utilizing soil water content sensors to examine soil water depletion can also be challenging if not properly paired with estimates of root architecture. Therefore, an objective of this study was to examine the feasibility of determining the direct relationship between RSA and SWD. The approach in this study was t o utilize a TDR soil water content senor which was calibrated for mini rhizotron tubes, allowing for the direct examination of the diurnal changes in soil water content in relation to quantified root presence. The relationship between estimated total SWD w as correlated with mid morning transpiration fluxes to validate that soil water depletion estimates were reflective of crop water uptake. The results showed a positive relationship between these variables for both genotypes (R 2 =0.61 for COC 041; R 2 = 0.67 f or TUFRunner TM in situ approach for quantifying SWD and RSA, allowing for mechanistic links to be made about relative differences in genotypic crop water use associated with root distribution in the soil The contrasting genotypic RSAs observed in this study provided suitable comparisons for examining the relationships between the structure and function of divergent root systems. The RSA of COC 041 can be generally characterized as a more fibrous root sy stem with greater TRL of small diameter (< 3mm) when compared to TUFRunner TM from 20 80 cm of soil depth than the runner market type. Despite these varying RSAs, no significant relations hip was observed between TRV and SWD, when TRV was summed across all sampling depths. Similar results observing the relationship between crop water uptake and total RLD have been reported (Ratnakumar and Vadez, 2011; Zaman Allah et al., 2011). Therefore, t hese results confirm that simply increasing the

PAGE 96

96 total amount of root mass, even at deep soil depths, may not necessarily lead to an increased capacity for water extraction. A plausible explanation for the lack of relationship between root system size (sum of TRV) and SWD could be due to water uptake responses to changes in soil hydrologic conditions. A general pattern in SWD distribution emerged among both genotypes in response to fluctuating soil water contents over the growing season. When soil water cont ent was high following a soil wetting event, the greatest proportion of SWD occurred in top 30 cm of soil. As soil drying occurred, water uptake of both genotypes moved deeper in the soil profile (30 60 cm). These results suggest that a primary influence o n SWD within the soil profile is strongly controlled by the soil water conditions, and the distribution of roots in relation to these dynamic SWD zones is a critical factor. The relationships between SWD and root parameters across all soil depth classifica tions (0 10 cm) in this study further emphasize this point. The correlation between SWD and rooting parameters was negative when VSWC was high following an irrigation event which is most likely due to the relationship between SWD and root architecture. A g eneral RSA observation in this study was an overall increase in roots at deeper soil depths (except the 100% TUFRunner TM VSWC, a negative correlation between rooting parameters and SWD occurred as a result of the predomi nant amount of SWD occurring where fewer of roots were observed in the soil from 0 30 cm. When water uptake patterns shifted to soil depths beyond 30 cm as a response to drying, these SWD zones contained large levels of TRL, likely resulting in the positiv e relationship between SWD and the increased presence of roots. Overall, these results support that the most effective RSA for maximizing soil water

PAGE 97

97 uptake depends on matching the root architecture to the hydrologic scenario of a particular production envi ronment. Studies which make the case that a more fibrous RSA with increased TRL at deep soil depths as being a critical phenotype for improving water deficit stress resilience, generally do so in the context of rainfed cropping systems where crop water de mand is predominately met through water stored deep in the soil profile (Wasson et al., 2012; Lynch, 2013; Lynch and Wojciechowski, 2015). The hydrologic conditions of this study are quite the opposite, reflecting a humid peanut production climate where wa ter deficit stress often arises quickly in surficial layers due to soils with low water retention. However, water deficit stress is generally intermittent due to crop water demand being met by in season receipt of precipitation and irrigation application. Several studies have demonstrated that effective RSA for maximizing water uptake in supply/pulse driven environments are those with more shallow root systems which have a high density of roots in the topsoil (Schwinning and Ehlerinder, 2001; Nakhforoosh et al., 2014; Tron et al., 2015). The current study also supports these findings when comparing the surficial (0 30 cm) soil water depletion of the 100% and 60%PA irrigation treatments of TUFRunner TM f roots systems in the surficial soils as compared to the other irrigation treatments also had a greater amount of soil water depletion in the surficial soil over the entire mid season measurement period. These findings highlight the importance that soil w ater uptake distribution is highly dependent on the soil hydrological conditions, and matching water acquisition.

PAGE 98

98 Although the coordination of RSA to soil hydrologi c conditions has the potential to maximize soil water acquisition, it is critical to note that the utilization of this water is dependent on transpiration providing the energy potential gradient controlling the ascent of xylem sap flow and root water absor ption (Boyer and Kramer, 1995). Furthermore, the rate of water transport through a plant is dependent on many physiological and anatomical factors of a plant influencing hydraulic conductivity (Vadez, 2014). In general, the genotype COC 041 had lower mid m orning transpiration fluxes than TUFRunner TM greater mid morning transpiration flux of TUFRunner TM water content between the two genotypes was similar, is li kely due to the regulation of hydraulic conductivity by COC 041 through reduction in stomatal aperture. However, there was an interaction between genotypic mid morning transpiration rates and measurement date, a result of COC 041 having similar transpirati on rates as TUFRunner TM genotypes typically occurred when soil water content was high following an irrigation event, and indicates that the hydraulic conductivity capability of COC 041 doe s have the potential to achieve similar rates of mid morning transpiration as TUFRunner TM Furthermore, this also demonstrates possible greater regulation of water flow of COC 041 caused by changes in soil water content than that exercised by TUFRunn er TM This is also supported by the observation of a stronger relationship between VSWC and transpiration rates with COC 041 than for TUFRunner TM point at which transpiration declines due to a decline in the fract ion of transpirable soil water has been reported among peanut genotypes (Devi et al., 2009). Interestingly,

PAGE 99

99 despite lower mid morning transpiration fluxes of COC 041, the rate of soil water decline was often greater than TUFRunner TM morning tran spiration fluxes were measured on the upper most mature leaf, and did not capture possible genotypic variation in transpiration rates among the lower and upper level canopy positons. For example, the erect canopy architecture of COC 041 could allow for gre ater light penetration and transpiration rates in the lower canopy resulting in a higher rate of soil water decline. Another possibility could be the combination of an extensive fibrous RSA of COC 041 exploring a great amount of soil volume, and greater tr anspiration fluxes than TUFRunner TM morning gas exchange measurements. Reports have documented limited transpiration phenotypes to vary among crop genotypes which allow for increased water use efficiency and soil water conservation by reducing transpiration rates during mid day when vapor pressure deficit is high (Sinclair et al., 2005; Devi et al., 2010; Sinclair et al., 2017). Conclusion There is increasing focus to assess genotypic root system architecture for identifying belowgr ound traits which could be utilized for improving water deficit stress tolerance. However, due to the coupled nature of both above and belowground traits fruitful info rmation for increasing possible water acquisition and utilization under water deficit stress conditions. Additionally, environmental and management factors which influence soil hydraulic factors strongly control the suitability of genotype for a particular production system. In this study we proposed a relatively cost effective in situ screening method for examining the relationship between RSA and soil water depletion distribution, which in combination with measurements of gas exchange can be used to

PAGE 100

100 deter mine genotypic suitability for a DVS of a particular production system, or coordinated adaptive phenotyping (CAPs).

PAGE 101

101 Table 4 1. Analysis of variance (ANOVA) for response variables of TRL, TRV, soil water depletion, and soil water content measured at the Plant Science Research and Education Unit (PSREU) near Citra, FL in 2016. treatment; GENO, genotype treatment; Depth, Soil depth increments; TM, repeated measures time within each growing season. Effect Total Root Length Total Root Volume Soil Water Depletion Soil Water Content IR 0.6712 0.4843 0.4805 0.8482 Geno <0.0001 0.4075 0. 0971 0.0237 IR*Geno 0.9768 0.1310 0.0640 <0.0001 Depth 0.0002 0.0080 <.0001 <0.0001 IR*Depth 0.0099 0.0001 0.7505 0.3190 Geno*Depth 0.0038 0.0303 0.0181 0.5024 IR*Geno*Depth 0.4672 0.0780 0.0038 0.8123 TM 0.0585 <0.0001 <0.0001 <0.0001 IR*TM 0.6632 0.1531 0.0044 <.0001 Geno*TM 0.4292 0.6512 0.0647 0.0349 IR*Geno*TM 0.0983 0.0182 <0.0001 0.1584 Depth*TM 0.5277 0.3693 <0.0001 <0.0001 IR*Depth*TM 0.8057 0.4350 0.6082 0.0149 Geno*Depth*TM 0.6048 0.4053 0.7068 0.9999 IR*Geno*Depth*TM 0.8795 0.6200 0 .3593 0.9850

PAGE 102

102 Figure 4 1. Total genotypic root length distribution for each root diameter class measures at the P SREU near Citra, FL in 2016. Total root length for each diameter class is summed to a soil depth of 80 cm.

PAGE 103

103 Figure 4 2. A) Average total root length of each peanut genotype by soil depth measured at the PSREU near Citra, FL in 2016. B) Aver age total root volume of each peanut genotype by soil depth measured at the PSREU near Citra,

PAGE 104

104 Figure 4 3. A)Total root volume architectural response of peanut genotype COC 041 to early season irrigation treatment measured at the PSREU near Citra, FL in 2016. A) Total root volume architectural response of peanut genotype TUFRunner TM Ci tra, FL in 2016. (Abbreviations: 100%, 1.9 cm per application; 60%PA, 1.1 cm per application until 50 DAP, followed by 1.9 cm per application for the Difference P < 0.05).

PAGE 105

105 Figure 4 4. A) Soil water depletion by soil depth range for each sampling date. B) Corresponding volumetric soil water content by depth range for each sampling date. Data measured at the PSREU near Citra, FL in 2016 and represents the sampling date by depth interaction which is averaged across both peanut genotypes and irrigation treatments.

PAGE 106

106 Table 4 2. Spearman correlation analysis between root parameters and soil water depletion (SWD) for each soil sampling depth on 75 DAP at the PSREU near Citra, FL in 2016. The volumetric soil water content (VSWC) on this day was high across all sampling depths and SWD was predominant in the top 30 cm of soil. root area; TRV, total root volume; ARD, average root diameter; SWD, soil water depletion. TRL PRA CSRA TRV ARD SWD P Value 0.20968 0.0556 0.21256 0.0522 0.21256 0.0522 0.18218 0. 0972 0.22814 0.0369

PAGE 107

10 7 Table 4 3. Spearman correlation analys is between root parameters and soil water depletion (SWD) for each soil sampling depth on 82 DAP at the PSREU near Citra, FL in 2016. The volumetric soil water content (VSWC) on this day was low across all sampling depths and SWD was predominant below 30 c m of soil depth. root area; TRV, to tal root volume; ARD, average root diameter; SWD, soil water depletion. TRL PRA CSRA TRV ARD SWD P Value 0.35132 0.0011 0.32214 0.0028 0.32214 0.00281 0.27520 0.0113 0.39756 0.0002

PAGE 108

108 Table 4 4. Analysis of variance (ANOVA) for leaf level mid morning transpiration flux measured at the PSREU near Citra, FL in 2016. measures time within each growing season. Effect Transpiration IR 0.4499 GENO < .0001 IR*GENO 0.1717 TM <.0001 IR*TM <.0001 GENO*TM <.0001 IR*GENO*TM 0.0780

PAGE 109

109 Figure 4 5. ( Left y axis) Average soil water content to a soil depth of 75 cm at each sampling date for each peanut genotype. ( Right y axis) Average leaf level mid morning transpiration flux of peanut genotypes at each sampling date for each peanut genotype. Data was measur ed at the (PSREU) near Citra, FL in 2016, and is averaged across early season irrigation treatments (Abbreviations: NS, denotes non Protected Least Significant Difference P < 0.05; All other time points are significantly differ Protected Least Significant Difference P < 0.05).

PAGE 110

110 Table 4 5. Spearman correlation analysis between average transpiration (TR) and volumetric soil water content (VSWC) across all sampling dates for each genotype at the PSREU near Citra, FL in 2016. COC 041 TUFRunner TM TR TR VSWC 0.4762 0.2792 P Value 0.0033 0.0994

PAGE 111

111 Figure 4 6. A) Relationship between leaf level transpiration and soil water depletion for each peanut genotyp e. Data points include all weekly sampling points from 61 96 DAP measured at the PSREU near Citra, FL in 2016. B) Relationship between total root volume and soil water depletion for each peanut genotype. Root volume is the sum of roots to a soil depth of 8 0 cm for each weekly sampling date.

PAGE 112

112 CHAPTER 5 OPTIMIZING COTTON IRRIGATION AND NITROGEN MANAGEMENT USING A SMARTPHONE IRRIGATION SCHEDULING TOOL AND IN SEASON NITROGEN APPLICATIONS Introduction Nitrogen is the most important applied nutrient in cotton production systems due to the potential of increasing lint yields by increasing N rates (Muharam et al., 2014). The possible economic gain by increasing N applications has led to many studies aimed at quantifying the N requirement per unit of lint produced resulting in estimates ranging from 10 to 29 kg N100 kg 1 lint across global cotton production regions (Miley and Oosterhius, 1989; Gerik et al., 1998). Typical application rates in cotton production regions of the United States (U.S.) have reported rang es from 71 220 kg N ha 1 (Taylor, 1995). A major factor adding to the variability in estimates of N requirements is the overall susceptibility of N loss through NO 3 leaching and denitrification, both which are heavily influenced by unpredictable seasonal environmental conditions and soil texture. Using in season N applications that split the total N applied into percentages applied at varying developmental stages is a management strategy that attempts to increase N efficiency by synchronizing soil N concen trations to plant N demand. Cotton nitrogen accumulation is non linear with the threshold of increase reported to begin at first square, reaching peak accumulation rates at approximately early mid flower the developmental stage that accounts for up to 23 45% of the total N accumulation (Armstrong and Albert, 1931; Olson and Bledsoe, 1942; Bassett et al., 1970; Halevy, 1976; Oosterhuis et al., 1983; Halevy et al., 1987; Mullins and Burmester, 1991). Furthermore, N accumulation in the seeds and lint account for up to 43 60% of the total plant N, creating a large N demand sink relatively late in the growing season (Mullins

PAGE 113

113 and Burmester, 2010). Geng et al. (2015) reported a yield increase of 106 kg lint ha 1 in one of two growing seasons when comparing a spli t N application (180 kg N ha 1 total application rate) of 40% at pre plant and 60% at first flower compared to 100% of the N applied at pre plant. Additionally, Yang et al. (2011) observed that when splitting N application rates of 225 kg N ha 1 between pr e plant and peak bloom (with a 40% application rate utilized at first bloom across all treatments), lint yields were greatest when applied at 0 40 60 % split application. Other reports have observed declines in maximal obtainable lint yields when comparing split N applications of 50% at pre plant and first square, to yields obtained by applying 100% pre plant N, when utilizing N rates up to 180 kg N ha 1 (Reiter et al., 2008). However, splitting N generally resulted in a greater amount of lint produced per kilogram of N applied translating into a greater economic return. Soil texture has also been demonstrated to influence both the total in season N rates and the timing of application. Some studies utilize the concept of Economically Optimal Nitrogen Rate ( EONR), defined as the N rate that maximizes the value of cotton produced at a given cotton and fertilizer price. The concept of EONR is useful for evaluating the impact of N rates across soil different soil textures which have varying levels of productivit y. For example, Scharf et al. (2012) reported a lower EONR for silt and sandy loams when compared to a clay soil. The EONR value was also useful for elucidating the impact of soil type on in season N applications applied at early square growth stage. This study also evaluated different timings on in season N applications. No yield differences were observed when N was evenly split (56 kg N ha 1 ) between pre plant + early square or pre plant + early flower in sand and silty loams. However,

PAGE 114

114 optimal N split for a clay soil was 1/3 pre plant, 1/3 early square, and 1/3 early flower (Scharf et al., 2012). Irrigation water can also be a costly management input in the approximately 40% of U.S cotton production systems which receive irrigation across the southeast an d southwest regions (Vellidis et al., 2016). Total water application for producing maximum lint yields in the southwest regions of the U.S. have been reported to be 74 cm; however, the relationship between lint yield and total water applied began to predi ct lint yield reductions when total water applied was greater than 80 cm (Wanjura et al., 2002). However, soil texture undoubtedly influences the lint yield responses to varying irrigation regimes (Tolk and Howell, 2010; Zhou et al., 2016). Additionally, t here has been research demonstrating the sensitivity of cotton lint production to timing of water stress to specific phenological stages (Hake and Grimes, 2010). Overall, these results suggest that the effectiveness of irrigation application rates on cotto n growth and yield are heavily influenced by soil type, regional and local climate conditions, and phenological development, demonstrating the need to develop irrigation scheduling solutions which encompass these variables. Some current irrigation decision tools have begun to account for these variables, and are being made readily available to growers through digital irrigation scheduling tools (Migliaccio et al., 2016). Individually, both N and irrigation management have been examined extensively in cotton ; however, less research has examined these management inputs simultaneously as they occur in production systems. Pettigrew and Zeng, (2014) reported that when all N was applied prior to planting, there was no additional gain in lint yield from furrow irri gation when no N was received, although lint yield increases with

PAGE 115

115 irrigation were obtained at N rates of 56 and 112 kg N ha 1 Singh et al. (2010) also reported interactive effects of irrigation and in season N management where seed cotton yield increased linearly with total N rate applications up to 200 kg N ha 1 with drip irrigation applied at 0.8 1.0 ET c In contrast, this study also reported that cotton plants treated with 0.5 0.6 ET c achieved maximal lint yields under 160 kg N ha 1 with slight yield d eclines at 200 kg N ha 1 Contrasting observations of a lack of interaction between irrigation and nitrogen have also been reported. For example, Bronson et al. (2006) conducted a study in west Texas and found no irrigation by N interaction when irrigation was applied using a center pivot at ET replacement rates of 63 93% and N rates of a non treated control, 164 kg N ha 1 applied prior to planting, and an even split application of 134 kg N ha 1 at pre plant and early square. The genetic yield potential is often never met in production systems due to a limitation of resources resulting in a failure to meet the genetic capacity for crop growth, development, and yield production (Lawlor, 2001). Management factors such as irrigation and N can have significant impacts on crop growth and production because they both have potential to influence the availability of substrates needed for photosynthesis and growth processes. For example, N provides substrate for chlorophyll production and protein synthesis which impa cts both CO 2 reduction and radiation capture. Additionally, irrigation has the potential to reduce water deficit stress which can reduce CO 2 assimilation, and cell turgor reducing growth rates. Furthermore, combination with excessive rates of N and irrigated have been reported to decrease lint yields as a result of reduced harvest index, and delayed maturity (Karam et al., 2006; Sabbe and Hodges, 2010). Therefore,

PAGE 116

116 quantification of both photosynthesis and abo ve ground partitioning in response to irrigation and N management practices has potential to provide insight possible optimal management levels to maintain the delicate balance between having sufficient levels to avoid deficiencies, without applying excess ive rates which can negatively impact yields and reduce production efficiency. The Florida cotton production environment is characterized by The overall objectiv es of this research were to: (i) evaluate varying levels of plant available water replacement (PAWR) estimated by the ET based soil water balance model Cotton SmartIrrigation smartphone application (Vellidis et al., 2016) which gives site specific recommen dation for irrigation scheduling; (ii) determine optimal in season N rates when split applications are made at first square and bloom; (iii) assess the possible interactive effects of various plant available water replenish (PAWR) and in season N rates whi ch may influence water and N management decisions. The effectiveness of these treatments were evaluated by quantifying their influence on radiation capture by quantifying photosynthetic responses using OJIP transient fluorescence protocol, SPAD chlorophyll content, and leaf area index (LAI). Additionally, N uptake, harvest index, and

PAGE 117

117 lint yield responses were measured for determining treatment impacts on nitrogen use efficiency, above ground biomass partitioning, and yield production. Materials and Methods Site Characterization : Field studies were conducted during the years of 2015 North Central Florida (29 24' 38" N, 82 10' 12" W). The soil is classified as Candler sand (Hyperthermic, uncoated Lamellic Quartzipsamments). In both years, cotton was planted into a peanut cereal rye cotton rotation. Soil preparation consisted of conventional tillage approximately five weeks before planting. Immediately prior to planting, the field was surface tilled using a field cultivar with S tine sweeps. The cotton cultivar PhytoGen 333 WRF was planted at a seeding density of 10 13 seeds m 1 with a row spacing of 0.91 m (Dow AgroSciences, Indianapolis, IN). Plots consisted of four rows 10. 6 m long. Planting was accomplished with a four row vacuum planter (John Deere., Moline, IL). A soil test was conducted approximately six weeks prior to planting, and recommended nutrients were broadcast surface applied immediately after planting. An even split broadcast surface application was made for potassium and sulfur recommendations which consisted of at planting + at first square. At first flower, a tank mix of 0.6 kg B ha 1 and 0.22 kg a.i. ha 1 of pyraclostrobin (BASF, Ludwigshafen, Germany) w as foliar applied. Pre emergence weed management spray consisted of 1.1 kg a.i. ha 1 pendimethalin (BASF, Ludwigshafen, Germany). Post emergence weed management spray consisted of a tank mix of 0.18 kg a.i. ha 1 glyphosate and 7.9 g a.i. ha 1 trifloxysulfu ron. Experimental Design: The experimental design consisted of a split plot arrangement in a randomized complete block design. The whole plots were irrigation

PAGE 118

118 treatments of plant available water replacement (PAWR), or the percent water replenishment of th e estimated root zone soil water deficit. These estimates provided by the ET based soil water balance model Cotton SmartIrrigation smartphone application (Vellidis et al., 2016; http://smartirrigationapps.org/cotton app/). Plant available water treatments consisted of (i) 100% of PAWR (100%); (ii) a primed acclimation (PA) treatment consisting of 50% of PAWR until first bloom and then 100% of PAWR (50%PA); (iii) 50% PAWR for the entire season (50%); (iv) a rain fed control (RF). Irrigation was applied using a lateral move system equipped with variable rate irrigation (VRI) technology (Lindsay Corporation, Omaha, NE). Sub plots consisted of four in season N treatments surface broadcast applied as NH 4 NO 3 at first square and flower at 0, 22, 34, and 45 kg N ha 1 per application. All plots received an initial N application broadcast applied as NH 4 NO 3 at 22 kg N ha 1 immediately after planting. Field Measurements : Leaf area index (LAI) was measured using an LAI 2200 (LI COR, Lincoln, Nebraska) and were initiated a t the first square phenological development stage (2015 40 DAP; 2016 37 DAP), with subsequent measurements occurring bi weekly until first fully developed boll (2015 77 DAP; 2016 75). A single measurement of LAI for a plot was conducted by taking one ab ove and four equidistant readings below the canopy as a perpendicular transect between the rows with the sensor view parallel to the row and repeating this sequence with the sensor view perpendicular to the row to gain greater spatial averaging. A lens cap cover with a 45 angle was used on all measurement dates. Measurements were taken in early morning with a cast shadow over the plot area to prevent diurnal bias and underestimation of LAI values.

PAGE 119

119 Leaf level measurements included SPAD chlorophyll content a nd OJIP chlorophyll fluorescence parameters. The SPAD measurements were initiated at first square phenological development stage with subsequent measurements occurring bi weekly until approximately first fully developed boll (Konica Minolta, Tokyo, Japan). These measurements occurred on the most newly formed mature leaf. Ten plants were randomly selected per plot and averaged. The OJIP measurements were initiated at first developed boll in both years, dark acclimated transient chlorophyll fluorescence (OJIP ) was measured and recorded weekly until boll opening (2015 112 DAP; 2016 110 DAP) using a pulse modulated fluorometer (Optisciences, Hudson, NH). Two plants per plot were measured using the most newly formed mature leaf and averaged. Measurements were rec orded in the evening (18:00 20:30), and dark acclimation clips were placed on leaves for approximately 60 minutes. Saturated light intensity was set for 3500 mols with a time duration of 3 seconds. Transient chlorophyll fluorescence parameters were calcul ated in as described by Stirbet and Govindjee, (2011). Parameters calculated ET2o RE1o, RE1o ; (ii) Specific energy fluxes per reaction center of J abs /RC, J o DI /RC, J o TR /RC, J o RE1 /RC, RC/J ABS ; and (iii) Phen omenological energy fluxes per cross section of J abs /CS o J o DI /CS, J o TR /CS, J o RE1 /CS, RC/CS (Table 5.1) (Figure 5.1). When all plots had approximately 30% of their bolls open by visual assessment, 0.75 m of row in each plot was harvested for determining N uptake by cutting the stems at the soil surface and fresh above ground biomass per unit area was determined. Whole plants were ground using a wood chipper and approximately 500 grams of fresh weight was collected as a sub sample. This sub sample was dried at 70 C for seven

PAGE 120

120 days and weighed for determining tissue water content which was used to calculate the dry above ground biomass per unit area. The dry sub sample was further ground through a Wiley Mill and plant material that passed through a 1mm sieve w as collected. This plant tissue was analyzed for total N (Waters Agricultural Laboratories, Inc., Camilla, GA) using the combustion method. Total percent N content was multiplied by the dry above ground biomass per area to determine total N uptake per area Decision to terminate cotton was done using the four nodes above the cracked boll method, and a harvest aid of 0.56 kg a.i. ha 1 paraquat dichloride was sprayed for defoliation. The two center rows of each plot were machine harvested using a spindle cot ton picker (John Deere, Moline, IL) for lint yield on 139 and 162 DAP in 2015 and 2016, respectively. The spindle picker was modified so cotton lint for each plot could be collected in mesh bags. Cotton lint collected was hand weighed for each plot and a 2 50 gram sub sample was collected and the seed removed from the lint using a table top cotton gin (manufacturer, location) to determine lint yield alone (gin turnout). The determined lint weight per area and above ground biomass per area harvested at 30% op en bolls was used for determining harvest index. Harvest index was calculated as the ratio of lint weight per area to total above ground biomass per area. Statistical Analysis : Statistical analysis was performed using SAS v. 9.4 statistical software (SAS Institute, 2013). PROC GLIMMIX was used to compute analysis of variance (ANOVA). Random effects of rep nested within year, and rep crossed with irrigation nested within year were included in the model. Repeated measures ANOVA was performed on measurements that were repeated over the growing season (SPAD, LAI, and chlorophyll fluorescence). Autoregressive covariance

PAGE 121

121 homogeneity were visually assessed by graphing the residual distribut ion, scatter plot of residuals, and Q Q plot of residuals. No data was pooled over the two site years due to the strong main effect of year. Data was pooled (when specifically) over other factors when appropriate as indicated by the F test results. Multipl e comparisons significance P < 0.05 probability level. Results Yearly precipitation and irrigation: Contrasting amounts of precipitation were received during the two growing seasons of 2015 and 2016; total cumulative rainfall received in 2015 and 2016 was 837 and 647 mm, respectively (Figure 5.2). The varying amounts of precipitation impacted the total number of irrigation events scheduled within the two growing seasons. In 20 15, a total amount of three irrigation applications were applied within the range of 41 63 days after planting (Table 5.2) for a total of 59 mm. Two of these irrigation applications occurred prior to first bloom. The third and final irrigation treatment wa s applied seven days after first bloom. In 2016, nine irrigation applications were applied ranging from 31 84 DAP for a total of 205 mm. Three of these applications occurred prior to first bloom, four in between first bloom and developed boll, and two in e arly boll development. Leaf area index and SPAD chlorophyll content : Differences in total water received in each year likely led to the differences in LAI between years (Table 5.3). What was similar between years was an observed trend where LAI continued to increase up until six weeks after first square (82 and 79 DAP in 2015 and 2016, respectively) to values of 4.7 and 2.9 (averaged across treatments) in 2015 and 2016,

PAGE 122

122 respectively (Figure 5.3 & 5.4). Following six weeks after first square, a decrease in LAI occurred in both years likely due to leaf senescence, to reductions of 32 and 19% in 2015 and 2016, respectively. Year also interacted with the irrigation and N treatments: in 2015, the 100% irrigation treatment had a greater LAI than all other irrigat ion treatments at two weeks after first square (Figure 5.3). Additionally, the 50%PA treatment in 2016 resulted in a greater LAI at the six and eight weeks after first square measurements when compared to the other irrigation treatments (Figure 5.4). For N treatments in both years, all treatments which received in season N (22, 34, and 45 kg N ha 1 ) maintained a greater LAI than the non treated control following first bloom; further, these in season N treatments maintained similar LAI across the entire seas on in 2015. Results were similar in 2016, with the exception of increased LAI when comparing the 34 kg N ha 1 to the 22 kg N ha 1 treatment at six weeks following first square. Differences in water treatments were not evident in 2015 for SPAD chlorophyll; whereas in 2016 the RF and 50% irrigation treatments maintain greater SPAD chlorophyll content later in the growing season when compared to the 100% and 50%PA treatments (Figure 5.3). Side dress N rates did impact SPAD chlorophyll content in 2015 which re sulted in the 22 kg N ha 1 treatment having lower chlorophyll content than the 45 kg N ha 1 at six and eight weeks after first square. The 45 kg N ha 1 treatment had similar SPAD chlorophyll content to the 34 kg N ha 1 treatment. In 2016, the SPAD chloroph yll content was similar among the side dress N treatments, except for the last measurement date at 8 weeks after first square (Figure 5.4). Overall when averaged across all treatments, there was a lower SPAD chlorophyll content during the

PAGE 123

123 2015 growi ng seas on with a marked decline after first square in this year, whereas SPAD chlorophyll content was relatively stable over the measurement period in 2016. Transient Chlorophyll Fluorescence (OJIP): Side dress N affected the energy fluxes of PSII photochemistry when measured during boll development in the 2015 growing season (Table 5.5). Phenomenological energy fluxes of J ABS/CS J o DI/CS and J o TR/CS were increased when comparing the greatest N rate (45 kg ha 1 ) to the treatment which received no in season N. The other two reduced N rates (22 and 34 kg N ha 1 ) were similar to the 45 kg N ha 1 treatment. These same energy fluxes expressed on a reaction center (RC) basis had a reverse trend where the non treated control had increased values when comparing to the gre atest side dress N rate at 45 kg ha 1 No differences were observed in the specific energy fluxes among the N treatments which received in season N. The amount of reaction centers per cross sectional area (RC/CSm) increased with increasing N rates. Increas ed N rates also had a similar trend of increased electron transport flux until PSI acceptors on both a reaction center J o RE1/RC and cross sectional basis J o RE1/CS In 2016, there was little influence of N treatment on the phenomenological or specific energ y fluxes. No differential influences of N treatments were also observed in the amount of RC/CSm or J o RE1/CS Irrigation treatments had little influence on the efficiency of electron transport during the 2015 growing season, while in 2016, there was in int eraction between irrigation and date (Figure 5.5). These interactions between irrigation and date may have been stronger for 2016 due to the relatively drier conditions experienced in this year requiring more frequent irrigations (Figure 5.2), particularly during boll development RE1o was greater when

PAGE 124

124 comparing all treatments which received irrigation to the RF treatment, but by the third week into boll development these relationships started to converge with the 100% irrigation treat RE1o when compared to the RF treatment (Figure 5.5). Within the electron transport chain, the transport efficiency from Q a to Q b ET2o ) was elevated in the 60%PA treatment at the two and three week sampling compared to the RF. Examini RE1o relationships among irrigation treatments show a similar RE1o, indicating that the mechanisms which drive electron transport efficiency from PSII to PSI are behaving differently depending their plant water status. Nitrogen Up take: There was a direct effect of N application rate on the total N uptake in both years, with no interaction between N and irrigation (Table 5.5) A reduced amount of N uptake was observed in the treatment that only received the base application at planti ng (22 kg N ha 1 ) compared to all the other treatments in both years (Figure 5.6). When comparing treatments that had received in season N, the patterns were different among years. For 2015, no differences in N uptake existed; while in 2016, the two mid le vel treatments (22 and 34 kg N ha 1 ) differed. When averaged across all treatments there was an increase in N uptake of 56 kg N ha 1 when comparing 2016 to 2015. Lint Characteristics and Harvest Index: Both N and irrigation had an interactive effect with y ear on lint yield (Table 5.5). In the relatively wet year of 2015, no lint yield differences were observed among irrigation treatments (Figure 5.7). However, during the relatively dry year of 2016, the 100% and 60%PA irrigation treatments had the greatest yields. For N treatments, 2015 showed differences in yield only between the 22 and 112 kg N ha 1 treatments, with the highest yield measured in the latter

PAGE 125

125 treatment. Surprisingly, the 112 kg N ha 1 treatment in 2016 had significantly lower yield than the o ther three application rates, even though this reduced yield was not reflected in the tissue N content. Influences of these yield patterns were evident in the HI as expected, since HI is calculated using yield. The impacts of irrigation were only evident i n 2016, with an overall lower HI for the RF treatment (Figure 5.7). Surprisingly, the pattern of HI in 2015 was opposite to that of yield: while the lowest N treatment had the lowest yield in that year, HI was greatest for this treatment. This indicates th at the relative allocation to aboveground biomass in the 22 kg N ha 1 treatment would have also significantly reduced total biomass, resulting in an elevated HI even though yield was low. For 2016, the yield and HI patterns were similar, with the highest a nd lowest HI values recorded for the 22 and 112 kg N ha 1 treatments, respectively (Figure 5.6). Discussion The contrasting environmental conditions among the two years of this study allowed for an opportunity to examine contrasting climatic conditions wh ich can make cotton N management challenging on the marginal sandy soils of north central Florida. The large amounts and intensity of precipitation events in 2015 occurring subsequent to both in season N applications were conducive for considerable NO 3 le aching. This was evident from the continual decrease in SPAD chlorophyll content in all N treatments following the last N application at first bloom (54 DAP). However, the greater amount of N uptake in treatments which received in season N prior to the per iod of intense rainfall at 63 DAP likely contributed to greater LAI. Late season plant N status had implications on photosynthetic performance (as measured by OJIP) in 2015. This is a critical period of boll development (77 112 DAP) when carbohydrate dema nd to support the formation and maturation of bolls is high.

PAGE 126

126 Under N limiting conditions, fewer PSII RC/CS were formed which was likely due to the high N demand required for synthesizing individual PSII reaction centers proteins. To compensate for fewer PS II reactions centers, the data indicated that larger antenna complexes were likely associated with each reaction center, resulting in a greater amount of trapping per reaction center (J o TR /RC). But the data also indicate that this compensation mechanism of larger antennae complexes was unsuccessful at increasing electron transport to the acceptors of PSI. This is shown by the greater amount of J O RE1/ RC when comparing the 45 to the 22 and 0 kg N ha 1 treatments, indicating that electron transport may be limi ted by intermediate steps between PSII and PSI in the 22 and 0 kg N ha 1 treatments. This may be due to N deficient plants having less Rubisco production which would result in decreased carboxylation, ATP synthase and NAD(P)H, causing an overall down regul ation of linear electron transport flux to PSI acceptors (Theobald et al., 1998). In 2015, two early season irrigation applications made prior to first bloom, followed by one irrigation application after first bloom, resulted in increased LAI in the 100% late season precipitation resulted in no late season irrigation applications. This resulted in no differences in photosynthetic performance among irrigation treatments during bol l development. Despite no differences in photosynthetic performance, the 100% treatment also had increased lint yield when comparing the 50% and RF irrigation treatments, however the 50%PA yielded similarly to the 100%, despite a lower LAI. These results i ndicate that the yield decline when comparing the 50% treatment to the 100% treatment can be attributed to the third and final irrigation treatment which

PAGE 127

127 occurred at approximately peak bloom. Peak flower has been documented to be a sensitive time for water stress due to the greatest number of young bolls being present on the plant, which are most sensitive to water deficit induced abscission (Grimes et al., 1970; Hake and Grimes, 2010). It also suggests that increased canopy growth in the 100% irrigation tr eatment did result in additional lint yield when compared to the 50%PA, indicating that the yield decline of the 50% and RF treatments was due to other factors than canopy growth, possibly a decreased number of bolls formed due to water stress disrupting c otton flower pollination. This would be consistent with studies showing that increasing water stress severity increases square/boll shedding and total number of bolls, thus reducing total lint yield (Cetin and Bilgel, 2002; Singh et al., 2010; Lockhande an d Reddy, 2014) Nonetheless, these results indicate that a reduction of up to 50% PAW replenishment can occur until first bloom in this sandy soil type without a decline in yield. The dry season of 2016 had little response in LAI with increasing in season N The absolute values of LAI were much lower when comparing 2015 to 2016 (2015 3.0; 2016 2.0). This was likely due to a reduction of carbon assimilation from water deficit stress as a result low precipitation during mid season cotton development. Late seas on N deficiency was not evident from SPAD chlorophyll readings, and approximately 56 kg N ha 1 more uptake when comparing average values in 2016 to 2015. In 2016, increasing N rates resulted in decreased HI values, most starkly evident in the 45kg N ha 1 t reatment that had a decreased lint yield of 56 kg lint ha 1 A combination of increased plant N levels and late season precipitation may have led to excess late season canopy growth and a delay in lint maturation in 2016. Regardless of the relative

PAGE 128

128 wet or dry conditions within a growing season these results show no additional yield gains were made when applying more than 22 kg N ha 1 in uniform split applications made at first square and bloom. Therefore, in years where precipitation and leaching potential is high on these sandy soils, it is likely that additional gains in lint yield from N management may only be made by increasing the number of N applications, not increasing the application rate. The very dry year of 2016 further supports the conclusion th at a 50% reduction in PAW until first bloom can be implemented without a reduction in yield. Three applications were made prior to first bloom, resulting in canopy growth differences among the treatments which received irrigation. Interestingly, the 50%PA irrigation treatments reached the greatest maximal LAI, despite receiving similar amounts of irrigation as the 100% irrigation treatment following first bloom. However, increased canopy growth did not correspond to increase or decreases in lint yield betwe en the 50%PA and 100% irrigation treatments. Whereas, the 50% and 100% had similar seasonal LAI values, but a yield decrease occurred with the 50% irrigation treatment. RE1o were similar when comparing the 100%, 50%PA, and 50% irrigation tre atments during boll development, indicating that the photosynthetic performance at this time was similar. These results show that the overall yield reduction in the 50% treatment may have been due to water stress impacts on reproductive development prior t o the first formed boll, and even greater lint yield reductions occurred in the RE1o was observed during the first two weeks of boll development. In both years, the research supports that a 50% reduction in PAWR can occur up until first bloom without any detriment to yield. However, when a 50%

PAGE 129

129 reduction in irrigation was extended past first bloom lint yield reductions did occur when compared to 100%PAWR, highlighting that water savings management should be targeted pr ior to first bloom. Conclusion The contrasting climatic condition between years allowed for a strong assessment of both N and irrigation management. Under both years, lint yield maintenance or gains occurred with making two split N applications at first s quare and bloom of 22 kg N ha 1 a 45 kg N ha 1 reduction in comparison to the traditional total N recommendation of 112 kg N ha 1 Yield maintenance also occurred when reducing the PAWR up until first bloom. These management strategies offer ways to cut i nput cost which are of economic importance when growing cotton in marginal production environments which often have an overall lower yield potential. An interesting research finding of this research was the lack of interaction between N and irrigation man agement on canopy development and yield. Particularly, in the dry year of 2016 this lack of interaction supports that the soil water content conditions did not limit N availability and uptake, and that increased canopy growth due to N application did not e xacerbate water stress. A possible explanation for these results could be differential root: shoot growth responses to both irrigation and N management. Further investigation into the interactive phenomenon of root: shoot growth responses, and synergisms o f both irrigation and N management under field conditions are needed.

PAGE 130

130 Table 5 1. Transient chlorophyll fluorescence (OJIP) parameter definitions as defined by Stirbet and Govindjee, (2011). OJIP Parameter RE1o ET2o RE1o J abs /RC J o DI /RC J o TR /RC J o RE1 /RC Specific energy fluxes per cross section J abs /CSo J o DI /CS J o TR /CS J o RE1 /CS RC/CS

PAGE 131

131 Figure 5 1. Depiction corresponding transient chlorophyll fluorescen ce (OJIP) parameters to the location within the light reactions of photosynthesis. Physical energy of light is absorbed by the antenna chlorophyll complex which is associated with a reaction center. Part of the absorbed energy is dissipated as heat and flu orescence (dashed transport chain.

PAGE 132

132 Figure 5 2. Cumulative precipitation from 1 May through 1 November for 2015 and 2016 at t he Plant Science Research and Education Unit (PSREU) near Citra, FL. Cumulative rainfall axis starts at planting for both years. Vertical arrows indicate when irrigation was applied.

PAGE 133

133 Table 5 2. Irrigation and total water received during the 2015 and 2016 growing season for each irrigation treatment at the PSREU near Citra, FL. Total water received is the sum of irrigation and precipitation. pplication until first flower followed by 100% PAWR per irrigation application; 50%, 50% PAWR per irrigation application for entire growing season; RF, rainfed control. Irrigation Water Applied Total Water Received 2015 2016 2015 2016 Irrigation Treatment -----------------------------------------mm ------------------------------------------100% 59 204 896 851 50%PA 46 179 883 826 50% 30 102 867 749 Rainfed 0 0 837 647

PAGE 134

134 Table 5 3. Analysis of variance (ANOVA) for SPAD chlorophyll content and lea f area index (LAI) variables at the PSREU near Citra, FL. on treatment; N in season N treatment; TM, time repeated measures within each growing season. Effect SPAD LAI year <0.0001 <0.0001 IR <0.0001 <0.0001 year*IR 0.0288 00.0004 N <0.0001 <0.0001 year*N 0.0127 0.1736 IR*N 0.8453 0.3829 year*IR*N 0.0917 0.8467 TM <0.0001 <0.0001 year*TM < .0001 <0.0001 IR*TM 0.0028 0.0005 year*IR*TM <0.0001 <0.0001 N*TM 0.0006 <0.0001 year*N*TM 0.8589 0.3582 IR*N*TM 0.9908 0.9746 year*IR*N*TM 0.9621 0.9912

PAGE 135

135 Figure 5 3. Top) Average leaf area index (LAI) starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2015 at PSREU. Bottom) Average SPAD chlorophyll content starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2015 at PSREU (Abbreviations: 100%, 100% PAWR per irrigation application; 50%PA, 50% PA WR per irrigation application until first flower followed by 100% PAWR per irrigation application; 50%, 50% PAWR per irrigation application for entire growing season; RF, rainfed control; Nitrogen rates are uniform in season N rates applied at first square and

PAGE 136

136 Figure 5 4. Top) Average leaf area index (LAI) starting at first square and measured until 8 weeks after first square for the irrigatio n and N treatments in 2016 at PSREU. Bottom) Average SPAD chlorophyll content starting at first square and measured until 8 weeks after first square for the irrigation and N treatments in 2016 at PSREU (Abbreviations: 100%, 100% PAWR per irrigation applica tion; 50%PA, 50% PAWR per irrigation application until first flower followed by 100% PAWR per irrigation application; 50%, 50% PAWR per irrigation application for entire growing season; RF, rainfed control; Nitrogen rates are uniform in season N rates appl ied at first square and Significant Difference at P < 0.05).

PAGE 137

137 Table 5 4. Average specific and phenomenological energy fluxes for each N rate in 2015 and 2016. Data is showing the in s eason N rate, time within season, and year interaction. Parameters are averaged across both irrigation treatment and measurement date within a season. Measurements averaged across dates included sampling which occurred from first developed boll to first op en boll. Year OJIP Parameters 2015 Nitrogen rate J abs /RC J abs /CS o J o DI /RC J o DI /CS J o TR /RC J o TR /CS J o RE1 /RC J o RE1 /CS RC/CS 0 kg N ha 1 1.53 744 0.33 158 1.20 586 0.38 182 497 22 kg N ha 1 1.52 761 0.33 162 1.19 599 0.38 191 513 34 kg N ha 1 1.48 757 0.32 161 1.16 596 0.39 197 525 45 kg N ha 1 1.46 764 0.31 160 1.15 603 0.41 214 534 LSD (0.05) 0.06 13.8 0.02 3.84 0.04 12.4 0.02 9.52 19.7 2016 Nitrogen rate 0 kg N ha 1 1.71 722 0.33 140 1.38 582 0.52 220 432 22 kg N ha 1 1.68 728 0.33 143 1.34 584 0.50 217 445 34 kg N ha 1 1.68 721 0.33 143 1.35 579 0.50 213 435 45 kg N ha 1 1.72 719 0.34 142 1.38 577 0.52 217 425 LSD (0.05) NS NS NS NS NS NS 0.02 NS NS season rates split at first square and fir st bloom. abs /RC average absorbed photon flux per PSII reaction center; J abs /CS o average absorbed photon flux per cross section; J o DI /RC, dissipated energy flux in processes other than trapping per PSII reaction center; J o DI /CS, dissip ated energy flux in processes other than trapping per cross section; J o TR /RC, maximum trapped exciton flux per PSII reaction center; J o TR /CS, maximum trapped exciton flux per cross section; J o RE1 /RC, electron transport flux until PSI acceptors per PSII rea ction center; J o RE1 /CS, electron transport flux until PSI acceptors per cross section; RC/CS, number

PAGE 138

138 Figure 5 5. Electron transport ef ficiencies from first developed boll to first open boll in 2015 (left) and 2016 (right) (Abbreviations: Abbreviations: 100%, 100% PAWR per irrigation application; 50%PA, 50% PAWR per irrigation application until first flower followed by 100% PAWR per irrig ation application; 50%, 50% PAWR per irrigation application for entire growing season; RF, rainfed control; RE1o efficiency that a PSII trapped electron is transferred ET2o efficiency that a PSII trapped electron is transferred fro m Q a to Q b; RE1o, efficiency that a PSII trapped electron is transferred from Q b to PSI acceptors; P < 0.05).

PAGE 139

139 Table 5 5. Analysis of variance (ANOVA) for N uptake, harvest index, and lint yield at the PSREU near Citra, FL. treatment; N in season N treatment. Effect N Uptake Harvest Index Lint Yield year 0.0012 0.0748 0.4268 IR 0.4703 0.0846 <0.0001 year*IR 0.3373 0.3222 0.0135 N <0.0001 <0.0001 0.3395 year*N 0.5407 0.2590 0.0002 IR*N 0.1102 0.5501 0.6780 year*IR*N 0.2099 0.563 9 0.1100

PAGE 140

140 Figure 5 6. Average total N uptake for each N treatment in 2015 (top) and 2016 (bottom) at the PSREU nea r Citra, FL. Nitrogen rates are uniform in season rates split at first square and first bloom. Data is average over each irrigation Protected Least Significant Difference at P < 0. 05).

PAGE 141

141 Figure 5 7. Average lint yield (top) and harvest index (bottom) for each N (left) and irrigation (right) treatment in 2015 and 2016 at the PSREU near Citra, FL. Significant Diffe rence at P < 0.05 (Abbreviations: Nitrogen rates are uniform in season rates split at first square and first bloom; 100%, 100% PAWR per irrigation application; 50%PA, 50% PAWR per irrigation application until first flower followed by 100% PAWR per irrigati on application; 50%, 50% PAWR per irrigation application for entire growing season; RF, rainfed control).

PAGE 142

142 CHAPTER 6 FINAL SUMMARY A primary objective of the research presented in this dissertation was to determine the degree of pod yield stability acro ss U.S. production regions and their relative response to irrigation within each environment. This objective was addressed in Chapter 2 by a pod yield stability analysis performed by examining the linear relationship between the average genotypic pod yield across all site years and irrigation treatments versus the population mean for each individual site year and irrigation treatment. Regression coefficients (b) of this linear relationship defining genotypic pod yield stability ranged from 0.88 to 1.30. Sta 1.10) genotypes having specific adaptability were COC 041, Serenut 6T, and Serenut 5R. A genotype by irrigation interaction occ urred at the Florida location, consisting of genotypes New Mexico Valencia C (NMVC), COC 041, and Chico responding positively to additional water applied, whereas, FloRun TM responded negatively. These results demonstrate both broad and specific adaptability of disparate peanut genotypes to U.S. production regions which could possibly be utilized by breeding programs for cultivar development. Variation of peanut germplasm pod yield stability across environments, and in re sponse to irrigation at the Florida location indicated diversity in genotypic phenotypes which may be contributing to these responses. This hypothesis was further evaluated in Chapter 3 of this dissertation by evaluating the above and belowground traits o f both genotypes which responded positively to increased water received and those which had no response. Genotypic total root length (TRL) and leaf area index (LAI) did not interact

PAGE 143

143 with irrigation treatments, but decreasing the total amount of irrigation over the growing season reduced LAI and pod yield, with no impact on TRL growth or distribution to 80 cm of soil depth. Genotypic effects influenced the TRL development over the growing season, and genotypes of subspecies fastigiata had greater TRL deeper in the soil profile. However, all genotypes had similar amounts of total surface area (TSA) distribution to 80 cm of soil depth. A positive relationship was observed between pod yield and maximal LAI in both study years, although this significant relations hip had a low correlation coefficient (r=0.39) in 2016 when greater water stress severity occurred during reproductive growth. A significant low negative correlation coefficient of 0.41 and 0.46 was observed between pod yield and maximal TRL in 2015 and 2016, respectively. The lack of interaction between irrigation and genotype for pod yield demonstrates that peanut genotypes which have more prolific root growth at depth may not necessarily have an advantage for increased amounts of water acquisition and utilization that are translated into yield. The results in the Chapter 3 study were interesting in that they went against some of the reports in the literature which emphasize that a deep root trait is superior for improving water deficit stress resilienc y. This lead to the hypothesis that simply more root presence, particularly deep in the soil profile may not necessary lead to greater crop water extraction. The research conducted in Chapter 4 set out to answer this question. A method was developed due ex amine the relationship between root architecture and soil water depletion distribution. The result of this experiment demonstrated that despite variations in genotypic RSA and responses to early season irrigation management, soil water depletion zones amon g the different RSA where

PAGE 144

144 similar for each sample date. However, the soil water depletion zones interacted with sample date indicating that volumetric soil water content mostly impacted where the soil water depletion was occurring in the soil profile. When soil water content was high following irrigation, approximately 74% of the quantified soil water depletion occurred in the top 30 cm of soil. Under soil drying conditions, the predominate proportion of soil water depletion occurred at 30 60 cm of soil dep th accounting for 67% of the total soil water uptake. The control of hydrologic conditions on soil water depletion zones is likely why no significant relationship were observed between total root volume (TRV) and total soil water depletion. Furthermore, co mparing daily volumetric soil water content among the two genotypes to mid morning transpiration fluxes provided evidence of differences in genotypic diurnal hydraulic conductivities. Overall, these results suggest that trait selection for improving utiliz ation in the humid southeast irrigated peanut production environment may consist of genotypes which have a greater proportion of its root density in surficial soils to capture in season water inputs, can obtain high maximal daily transpiration rates, but a re sensitive to decreasing transpiration rates one daily increases in vapor pressure deficit (VPD) occur. The final objective of the research in this dissertation was to determine optimal irrigation and in season N management strategies for the southeast cotton production system. Optimal irrigation management was determined by evaluate varying levels of plant available water replacement (PAWR) estimated by the ET based soil water balance model Cotton SmartIrrigation smartphone application. A strength of us ing this method is that this irrigation scheduling tool makes recommendations based on site specific environmental. Two years of analysis over a wet and dry year indicated that

PAGE 145

145 optimal irrigation management for this soil types was to apply 60% PAWR up unti l first bloom with subsequent irrigation amount being 100% PAWR Both years of this study also showed that lint yield maintenance or gains occurred with making two split N applications at first square and bloom of 22 kg N ha 1 a 45 kg N ha 1 reduction in comparison to the traditional total N recommendation of 112 kg N ha 1 The knowledge gained from this research has potential to improve peanut and crop production as a whole. This research has identified specific peanut genotypes which have relatively hig h yield and stability across production regions, and identified inherent biomass partitioning differences among peanut genotypes of varying taxonomic descent. Both of these contributions have potential to assist breeders in peanut cultivar development. Ano ther impact his research had is the development of a phenotyping protocol to examine the coordination of pertinent traits which can be used to determine specific adaptability of crop genotypes to climates and production regions of varying soil hydrological conditions. Overall these contributions have potential to improve the adaptability of crops, which when combined with site specific management support tools have the potential to improve yield production, grower profitability, and environmental sustainabi lity.

PAGE 146

146 REFERENCE LIST Adomou, M., Ntare, B.R., Williams, J.H., 1997. Stability of pod yields and parameters of a simple physiological model for yield among peanut lines in northern benin. Peanut Science 24, 107 112. Allen, R.G., Pereira, L.S. Raes, D., Smith, M., 1998. Crop Evapotranspiration. FAO irrigation and drainage paper 56, 1 300. Anderson, W.F., Mozingo, R.W., Wynne, J.C., 1989. Comparisons of stability statistics as criteria for cultivar development in peanut. Peanut Science 16, 21 2 5. Armstrong, G.M., Albert, W.B., 1931. A study of the cotton plant with special reference to its nitrogen content. J. Agric. Res. 42, 689 703. Bassett, D.M.; Anderson, W.D., Werkhoven, C.H.E, 1970. Dry matter production and nutrient uptake in irrigated co tton ( Gossypium hirsutum ). Agron. J. 62, 299 303. Bronson, K.F., Booker, J.D., Bordovsky, J.P., Keeling, J.W., Wheeler, T.A., Boman, R.K., Parajulee, M.N., Segarra, E., Nichols, R.L., 2006. Site Specific irrigation and nitrogen management for cotton produc tion in the southern high plains. Agronomy Journal 98, 212 219. Bauerle, T.L., J.H. Richards, D.R. Smart, and D.M. Eissenstat. 2008. Importance of internal hydraulic redistribution for prolonging the lifespan of roots in dry soil. Plant, Cell, and Environm ent 31: 177 186. Blum, A., 2009. Effective use of water (EUW) and not water use efficiency (WUE) is the target of crop yield improvement under drought stress. Field Crops Research 112, 119 123. Bonn, M.A. and F.W. Bell. 2003. Economic impact of selected Fl orida springs on surrounding local areas. Florida Department of Environmental Protection. Boote, K.J. 1982. Growth stages of peanut (Arachis Hypogaea L.). Peanut Science 9: 35 40. Borm, A., Ramalho, M.A.P., Fritsche Neto, R., 2012. Abiotic Stresses: Cha llenges for plant breeding in the coming decades, in: Fritsche Neto, R.,Borem, A.. (Eds), Plant breeding for abiotic stress tolerance, Springer, New York, p.p. 1 12. Boyer, J.S., 1968. Relationships of water potential to growth of leaves. Plant Physiology 43, 1056 1062. Boyer, J.S., and Kramer, P. J. 1995. Water relations of plants and soils. Academic Press, Inc.

PAGE 147

147 Brooks, J.R., F.C. Meinzer, J.M. Warren, J.C. Domec, and R. Coulombe. 2006. Hydraulic redistribution in a Gouglas fir forest: lessons from system manipulaions. Plant, Cell, and Environment 29: 138 150. Brown, A.H.D., 1989. Core collections: a practical approach to genectic resources management. Genome 31, 818 824. Cattivelli, L., Rizza, F., Badeck, F. W., Mazzucotelli, E., Mastrangelo, A.M., Francia E., Mar, C., Tondelli, A., Stanca, A.M., 2008. Drought tolerance improvement in crop plants: An integrated view from breeding to genomics. Field Crops Research 105, 1 14. Cetin, O., Bilgel, L. 2002. Effects of different irrigation methods on shedding an d yield of cotton. Agricultural Water Management 54, 1 15. Chai, Q., Gan, Y., Zhao, C., Xu, H., Waskon, R.M., Niu, Y., Siddique, K.H.M., 2016. Regulated deficit irrigation for crop production under drought stress. A review. Agron. Sustain. Dev. 36, 1 21. Charles Edwards, E.A., 1982. Physiological Determinants of Crop Growth. Academic Press, Sydney. Coffelt, T.A., Seaton, M.L., VanScoyoc, W., 1989. Reproductive efficiency of 14 Virginia type peanut cultivars. Crop Sci. 29, 1217 1220. Dang, P.M., Chen, C.Y., Holbrook, C.C., 2013. Evaluation of five peanut ( Arachis hypogaea ) genotypes to indentify drought repsonsive mechanisms utilising candidate gene approach. Functional Plant Biology 40, 1323 1333. de Wit, C.T., 1958. Transpiration and Crop Yields, vol. 64.6 Versl. Landbouwk. Onderz., Institute of Biological and Chemical Research on Field Crops and Herbage, Wageningen, The Netherlands, pp. 24. Devi, J.M., Sinclair, T., Vadez, V., 2010. Genotypic variation in peanut for transpiration response to vapor pressur e deficit. Crop Science 50, 191 196. Devi, J.M., Sinclair, T.R., Vadez, V., Krishnamurthy, L., 2009. Peanut genotypic variation in transpiration efficiency and decreased transpiration during progressive soil drying. Field Crops Research 114, 280 285. Donal d, C.M., Hamblin, J., 1976. The biological yield and harvest index of cereals as Finlay, K.W., Wilkinson, G.N., 1963. The analysis of adaptiation in a plant breeding programme. Australian Journal of Agriculutural Research 14, 742 754.

PAGE 148

148 Florida Climate Cente r. 2014. Florida statewide averaged precipitation data. Florida State University. Tallahassee, FL. Frigui, H., Nasraoui O. 2000. Simultaneous clustering and attribute discrimination. Ninth IEEE International Conference on FUZZ Systems 1, 158 163. doi: 10 .1109/FUZZY.2000.838651 Geng, J., Ma, Q., Zhang, M., Li, C., Liu, Z., Lyu, X, 2015. Synchronized relationships between nitrogen release of controlled released fertilizers and nitrogen requirements of cotton. Field Crops Research 184, 9 16. Gerik, T.J., Lem on, R.G., Favor, K.L., Hoelewyn, T.A., Jungman, M., 1998. Performance of ultra narrow row cotton in Central Texas. pp. 66 67. in : P. Dugger and D.A. Richter (Eds.). Proc. Beltwide Cotton Conf., National Cotton Council of America, Memphis, Tenn. Grimes, D.W. Miller, R.J., Dickens, L. 1970. Water stress during flowering of cotton. Calif. Agric. 24(3), 4 6. Hake, K.D., Grimes, D.W. 2010. Crop water management to optimize growth and yield, in: Stewart, James McD., Oosterhuis, D.M., Heitholt, J.J., Mauney, J.R.. (Eds.), Physiology of Cotton, Springer, New York, pp. 97 110. Halevey, J., A. Marani, and T. Markovitz. 1987. Growth and NPK uptake of high yielding cotton grown at different nitrogen levels in a permanent plot experiment. Plant and Soil 103: 39 44. Halev y, J., 1976. Growth rate and nutrient uptake of two cotton cultivars grown under irrigation. Agron. J. 68, 701 705. Haro, R.J., Baldessari, J., Otegui, M.E., 2013. Genetic improvement of peanut in Argentina between 1948 and 2004: seed yield and its compone nts. Field Crops Res. 149, 76 83. Haro, R.J., Baldessari, J., Otegui, M.E., 2017. Genetic improvement of peanut in Argentina between 1948 and 2004:light interception, biomass production and radiation use efficiency. Field Crops Res. 204, 222 228. Hay, R., Porter, J., 2006. The physiology of crop yield. Blackwell Publishing, Ames, IA. Henry, J.A., 2005. Humid Climates, in: Oliver, J.E. (Eds.), Encyclopedia of World Climatology, Springer, New York, p.p. 412 413. Holbrook, C.C., Anderson, W.F., Pittman R.N., 1 993. Selection of a core collection from the U.S. germplasm collection of peanut. Crop Science 33, 859 861. Howell, T (2003) Irrigation efficiency. Encyclopedia of Water Science (DOI:10.1081/E EWS 120010252).

PAGE 149

149 Hubick, K.T, Shorter, R., Farquhar, G.D. 1988. Heritability and genotype x environment interactions of carbon isotope discrimination and transpiration efficiency in peanut ( Arachis hypogaea L.). Australian Journal of Plant Physiology 15, 799 813. Isleib, T.G., Holbrook, C.C., Gorbet, D.W., 2001. Use o f plant introductions in peanut cultivar development. Peanut Science 28, 96 113. Isleib, T.G., Milla Lewis, S.R., Dong, W.B., Copeland, S.C., Shew, B.B., Arellano, C., 2013. Stability analysis of incidence of tomato spotted wilt in virginia type peanut cul tivars and breeding lines. Peanut Science 40, 24 30. Jongrungklang, N., Toomsan, B., Vorasoot, N., Jogloy, S., Boote, K.J., Hoogenboom, G., Patanothai, A., 2011. Rooting traits of peanut genotypes with different yield responses to pre flowering drought str ess. Field Crops Research 120, 262 270. Jongrungklang, N., Toomsan, B., Vorasoot, N., Jogloy, S., Boote, K.J., Hoogenboom, G., Patanothai, A., 2012. Classification of root distribution patterns and their contributions to yield in peanut genotypes under mid season drought stress. Field Crops Research 127, 181 190. Karam, F., Lahoud, R., Masaad, R., Daccache, A., Mounzer, O., Rouphael, Y., 2006. Water use and lint yield response of drip irrigated cotton to the length of irrigation season. Agricultural Water M anagement 85, 287 295. Katz, B.G., A. Alejandro Sepulveda, and R.J. Verdi. 2009. Estimating nitrogen loading to groundwater and assessing vulnerability to nitrate contamination in a large karstic springs basin, Florida. Journal of American Water Resources Association 45: 607 627. Kempthorne, O. 1969. An introduction to genectic statisitcs. Iowa State University Press. Ames, Iowa. Ketring, D.L., Erickson, P.I., Stone, J.F., 1990. Applied sap velocity in peanut genotypes under control and stress conditions. Peanut Science 17, 38 44. Knauft, D.A., Gorbet, D.W., 1989. Genetic diversity among peanut cultivars. Crop Science 29, 1417 1422. Koolachart, R., Jogloy, S., Vorasoot, N., Wongkaew, S., Holbrook, C.C., Jongrungklang, N., Kesmala, T., Patanothai, A., 2013. Rooting traits of peanut genotypes with different yield responses to terminal drought. Field Crops Research 149, 366 378. Kottapalli, K.R., Rakwal, R., Shibato, J., Burow, G., Tissue, D., Burke, J., Puppala, N., Burow, M., Payton, P., 2009. Physiology and proteomics of the water deficit stress response in three contrasting peanut genotypes. Plant, Cell & Environment 32, 380 407.

PAGE 150

150 Krapovickas, A., and Gregory, W.C., 2007. Taxonomy of the genus Arachis (Leguminosae). Translated by Williams, D.E. and Simpson, C.E. Bonplandia 16, 1 205. Kulkarni J.H., Ravindra V., Sojitra V.K. and Bhatt D.M. 1988. Growth, nodulation and N uptake of groundnut (Arachis hypogaea L.) as influenced by water deficits stress at different phenophase. Oleagineus 43 : 415 419. Lamb, M.C., Jr. Davidson, J.I., Childre, J.W., Jr. Martin, N.R., 1997. Comparison of peanut yield, quality, and net returns between nonirrigated and irrigated production. Peanut Science 24, 97 101. Lamb, M.C., Rowland, D.L., Sorenson, R.B., Butts, C.L., Faircloth, W. H., Nuti, R.C., 2007. Economic returns of irrigated and non irrigated peanut based cropping systems. Peanut Science 34, 10 16. Lawlor, D.W., 2002. Carbon and nitrogen assimilation in relation to yield: mechanisms are the key to understanding production sys tems. Journal of Experimental Botany 53, 773 787. Lenka D. and Mishra P.K. 1973. Response of groundnut (Arachis hypogaea L.) to irrigation. Indian J. Agron. 18: 492 497. Lynch, J., 2013. Steep, cheap and deep: an ideotype to optimize water and acquisition by maize root systems. Ann. Bot. 112, 347 357. Lynch, J., Wojciechowski, T., 2015. Opportunities and challegnes in the subsoil: pathway to deeper rooted crop. Journal of experimental Botany 66, 2199 2210. Marella, R. (2014) Water withdrawals, use, and tre nds in Florida, 2010. Unites States Geological Survey Scientific Investigations Report 2014 5088. (http://dx.doi.org/10.3133/sir20145088). Meeks, C.D., Snider, J.L., Porter, W.M., Vellidis, G., Hawkins, G., Rowland, D.L., 2017. Assessing the utility of pr imed acclimation for improving water savings in cotton using a sensor based irrigation scheduling system. Crop Science 57, 2117 2129. Meisner C.A. and Karnok K.J. 1991. Root hair occurrence and variation with environment. Agron J. 83: 814 818. Meisner C.A. and Karnok K.J. 1992. Peanut root response to drought stress. Agron. J. 84: 159 165. Migliaccio, K.W., Morgan, K.T., Vellidis, G., Zotarelli, L., Fraisse, C., Zurweller, B.A., Andreis, J.H., Crane, J.H., Rowland, D.L. 2016. Smartphone apps for irrigation scheduling. ASABE 59: 291 301.

PAGE 151

151 Miley, W.N., Oosterhuis, D.M., 1989. Nitrogen and carbohydrate nutrition of cotton. University of Arkansas, Cooperative Extension Service, Little Rock. Fact Sheet. 2045. p 4. Monteith, J.L., 1965. Evaporation and environment. Symp. Soc. Exp. Biol. 19, 205 234. Monteith, J.L., 1977. Climate and efficiency of crop production in Britain. Philos. Trans. R. Soc. Lond. B 281, 277 294. Mothilal, A., 2012. Groundnut. 323 395. Muharam, F.M., Bronson, K.F., Maas, S.J., Ritchie, G.L., 20 14, Inter relationship of cotton plant height, canopy width, ground cover and plant nitrogen status indicators 169, 58 69. Mullen, D.M., Y. Yu, and G. Hoogenboom. 2009. Estimating the demand for irrigation water in a humid climate: A case study from the s outheastern Unites States. Agriculute Water Management 96: 1421 1428. Mullens, G.L., and C.H. Burmester. 1990., Dry matter, nitrogen phosphorus, and potassium accumulation by four cotton varieties. Agron. J. 82: 729 736. Mullins, G.L., Burmester, C.H., 201 0. Relation of growth and development to mineral nutrition, in: Stewart, James McD., Oosterhuis, D.M., Heitholt, J.J., Mauney, J.R.. (Eds.), Physiology of Cotton, Springer, New York, pp. 97 110. Mwadzingeni, L. H. Shimelis, E. Dube, M.D. Laing, T.J. Tsilo, 2016. Breeding wheat for drought tolerance: Progress and technologies. J. Integ. Agri. 15, 935 943. Nakhforoosh, A., Grausgruber, H., Kaul, H., Bodner, G., 2014. Wheat root diversity and root functional characterization. Plant Soil 380, 211 229. Narh, S., M.D., Tillman, B.L., Brandenburg, R.L., Jordan, D.L., 2014. Yield Improvement and Genotype Environment Analyses of Peanut Cultivars in Multilocation Trials in West Africa. Crop Science 54, 2413. National Agricultural Statistics Service, 2014. Farm ranch and irrigation survey. 2012 Census of Agriculture 3, 1 206. National Agricultural Statistics Service, 2015.Acreage. United States Department of Agriculutre, 1 41. National Agricul tural Statistics Service, 2017. Quick Stats. Available at : https://quickstats.nass.usda.gov.

PAGE 152

152 Natural Resources Conservation Service, 2005. National Engineering Handbook: Irrigation Guide. Unites States Department of Agriculture. Available at : https://pol icy.nrcs.usda.gov/openNonWebContent.aspx?content=17837.wba. Oliveria, L.F., Scharf, P.C., Vories, E.D., Drummond, S.T., Dunn, D., Stevens, W.G., Bronson, K.F., Benson, N.R., Hubbard, V.C., Jones, A.S., 2012. Calibrating canopy reflectance sensors to predic t optimal mid season nitrogen rate for cotton. Soil Science Society of America 77, 173 183. Olsen, L.C. and R.P. Bledsoe. 1942. The chemical composition of the cotton plant and the uptake of nutrients of different stages of growth. Georgia. Agric. Exp. Stn Bull. 222. Oosterhuis, D.M, 1990. Growth and development of a cotton plant. in: Oosterhuis, D.M. and Miley, W.N. (Eds), Nitrogen Nutrition in Cotton: Practical Issues, ASA CSSA SSSA, Madison, WI, pp. 1 24. Oosterhuis, D.M., Chipamaunga, J., Bate, G.C., 1 983. Nitrogen uptake of field grown cotton. I. Distribution in plant components in relation to fertilization and yield. Expl. Agric. 19, 91 101. Padilla, F.M., Miranda, J.D., Jorquera, M.J., Pugnaire, F.I., 2009. Variability in amount and frequency of wate r supply affects roots but not growth of arid shrubs. Plant Ecol 204, 261 270. Padmavathi, T.A.V., and Rao, D.M., 2013. Differential accumulaiton of osmolytes in 4 cultivars of peanut ( Arachis hypogaea L.) under drought stress. Passioura, J.B., 1996. Droug ht and drought tolerance. Plant Growth Regul. 20, 79 83 Pasupuleti, J., Nigam, S.N., 2013. Phenotyping for Groundnut (Arachis hypogaea L.) Improvement. In: Panguluri, J., Kumar, K.K. (Eds.), Phenotyping for Plant Breeding: Applications of Phenotyping Metho ds for Crop Improvement. Springer New York, NY, pp. 129 167. Patel B.P., and Gangavani S.B., 1990. Effects of water stress imposed at various stages on yield of groundnut and sunflower. J. Maharashtra Agric. Univ. 15, 322 324. Patel M.S. and Golakiya B.A. 1988. Effect of water stress on yield attributes and yield of groundnut (Arachis hypogaea L.). Indian J. Agric. Sci. 58: 701 703. Pettigrew, W.T., Zeng, L., 2014. Integrations among irrigation and nitrogen fertility regimes on mid south cotton production. Agronomy Journal 106, 1614 1622. Pfannschmidt, T. and C. Yang. 2012. The hidden function of photosynthesis: a sensing system for environmental conditions that regulates plant acclimation responses. Protoplasm 249: S125 S136.

PAGE 153

153 Poorter H., Niklas, K.J., Reich P.B., Oleksyn, J., Poot, P., Mommer, L., 2012. Biomass allocation to leaves, stems and roots: metaanalyses of interspecific variation and environmental control. New Phytol 193, 30 50. Prieto, I., C. Armas, and F.I. Pugnaire. 2012. Water release through p lants: new insights into its consequences at the plant and ecosystem level. New Phytologist 193: 830 841. Prostko, E.P., editor. 2015. 2015 peanut update. University of Georgia Extension Peanut Team. Athens, GA. Rao Nageswara, R.C., Singh, S., Sivakumar, M .V.K., Srivastava, K.L., Williams, J.H., 1985. Effect of water deficit at different growth phases of peanut. I. Yield responses. Agronomy Journal 77, 782 786. Rao R.C.N., Williams S., Sivakumar M.V.K., Srivastava K.L. and Wadia K.D.R. 1988. Effect of water deficit at different growth phases of peanut. II. Response to drought during pre flowering phase. Agron. J. 80: 431 438. Ratnakumar, P., Vadez, V. 2011. Groundnut ( Arachis hypogaea ) genotypes tolerant to intermittent drought maintain a high harvest index and have small leaf canopy under stress. Functional Plant Biology 38, 2016 1023. Ratnakumar, P., Vadez, V., Nigram, S.N., Krishnamurthy, L., 2009. Assessment of transpiration efficiency in peanut ( Arachis hypogaea L.) under drought using a lysemtric syst em. Plant Biol (Stuttg) 1, 124 130. Reddy, T.Y., Reddy, V.R., Anbumozhi, V., 2003. Physiological responses of groundnut (Arachis hypogaea L.) to drought stress and its amelioration: a critical review. Plant Growth Regulation 41: 75 88. Reiter, M.S., Reeves D.W., Burmester, C.H., 2008. Cotton nitrogen management in a high residue conservation system: source, rate, method, and timing. Soil Science Society of America Journal 72, 1330 1336. Rochester, I.J. 2012. Using seed nitrogen concentration to estimate c rop N use efficiency in high yielding irrigated cotton. Field Crops Research 127: 140 145. Rochester, I.J., and G.A. Constable. 2015. Improvements in nutrient uptake and nutrient use efficiency in cotton cultivars released between 1973 and 2006. Field Crop s Research 173: 14 21. Rowland, D.L., Faircloth, W.H., Payton, P.P., Tissue, D.T., Ferrell, J.A., Sorenson, R.B., Butts, C.L., 2012. Primed acclimation of cultivated peanut (Arachis hypogaea L.) through the use of deficit irrigation timed to crop developme ntal periods. Agricultural Water Management 113, 85 95.

PAGE 154

154 Rucker, K. S., Kvien, C. K., Holbrook, C.C., Hook, J.E., 1995: Identification of peanut genotypes with improved drought avoidance traits. Peanut Sci. 21, 14 18. Schnepf, T., 2016. U.S. Peanut Program and Issues. Congressional Research Service, 1 20. Schwinning, S., Ehleringer, J.R., 2001. Water use trade offs and optimal adaptations to pulse driven arid ecosystems. Journal of Ecology 89, 464 480. Seaton, M.L., Coffelt, T.A., Van Scoyoc, S.W., 1992. Co mparison of vegetative and reproductive traits of 14 peanut cultivars. Oleagineux 47, 471 474. Selvaraj, M.G., Burow, G., Burke, J.J., Belamkar, V., Puppala, N., Burrow, M.D., 2011. Heat stress screening of peanut ( Arachis hypogaea L.) seedling for acquire d thermotolerance. Plant Growth Regulation 65, 83 91. Sharp, R.E., Silk, W.K., Hsiao, T.C., 1988. Growth of the maize primary root at low water potentials. I. Spatial distribution of expansive growth. Plant Physiology 87, 50 57. Shurley, D. and A. Smith. 2015. 2015 Georgia cotton cost and return budget estimates. University of Georgia Cooperative Extension. Athens, GA. Sinclair, T., Devi, J., Shekoofa, A., Choudhary, S., Sadok, W., Vadez, V., Riar, M., Rufty, T., 2017. Limited transpiration response to hi gh vapor pressure deficit in crop species. Plant Science 260, 109 118. Sinclair, T.R., 2011. Challenges in breeding for yield increase for drought. Trends in Plant Science 16, 289 293. Sinclair, T.R., Hammer, G.L., van Oosterom, E.J., 2005. Potential yiel d and water use efficiency benefits in sorghum from limited maximum transpiration rate, Funct. Plant Biol. 32, 945 952. Singh, Y., Rao, S.S., Regar, P.L., 2010. Deficit irrigation and nitrogen effects on seed cotton yield, water productivity and yield resp onse factor in shallow soils of semi arid environment. Agricultural Water Management 97, 965 970. Skinner, R.H., D.L. Gustine, and M.A. Sanderson. 2004. Growth, water relations, and nutritive value of pasture species mixtures under moisture stress. Crop S ci. 44:1361 1369. Skirycz, A., K. Vandenbroucke, P. Claux, K. Maleux, B. De Meyer, S. Dhondt, A. Pucci, N. Gonzalez, F. Hoeberichts, V.B. Tognetti, M. Galbiati, C. Tonelli, F. Van Breusegem, M. Vuylsteke, and D. Inz. 2011. Survival and growth of Arabidops is plants given limited water are not equal. Nature Biotech. 29, 212 214.

PAGE 155

155 Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Web soil survey. Available at: https://websoilsurvey.sc.egov.usda.gov/. Songsri, P ., Jogloy, S., Vorasoot, N., Akkasaeng, C., Patanothai, A., Holbrook, C.C., 2008. Root Distribution of Drought Resistant Peanut Genotypes in Response to Drought. Journal of Agronomy and Crop Science 194, 92 103. Stansell, J.R., Shephard, J.L., Pallas, J.E. Bruce, R.R., 1976. Peanut responses to soil water variables in the southeast. Peanut Science 3, 44 48. Tanou, G., V. Fotopoulos, and A. Molassiotis. 2012. Priming against environmental challenges and proteomics in pants: update and agricultural perspect ives. Frontiers in Plant Science 3: 1 5.United States Department of Agriculture National Agriculture Statistics Service (2015) Crop production 2014 summary. Thangthong, N., Jogloy, S., Pensuk, V., Kesmala, T., Vorasoot, N., 2016. Distribution patterns of p eanut roots under different duration of early season drought stress. Field Crops Research 198, 40 49. The Southeast Regional Climate Center, 2017. Monthly and Seasonal Climate Information. Available at: http://www.sercc.com/climateinfo/monthly_seasonal. T olk, J.A., Howell, T.A., 2010. Cotton water use and lint yield in four great plains soils. Agronomy Journal 102, 904 910. Tollenaar, M., Lee, E.A., 2002. Yield potential, yield stability and stress tolerance in maize. Field Crop Research 75, 161 169. Tron, S., Bodner, G., Laio, F., Ridolfi, L., Leitner, D. 2015. Can diversity in root architecture explain plant water use efficiency?. Ecological Modelling 312, 200 210. Turner, N.C., Blum, A., Cakir, M., Steduto, P., Tuberosa, R., Young, N. 2014. Strategies to increase the yield and yield stability of crop under drought are we making progress?. Functional Plant Biology 41, 1199 1206. Upadhyaya, H.D., Bramel, P.J., Ortiz, R., Singh, S., 2002. Developing a mini core of peanut for utilization of genectic resources. Crop Science 42, 2150 2156. Upadhyaya, H.D., Ortiz, R., Bramel, P.J., Singh, S., 2003. Develpoment of a groundnut core collec tion using taxonomical, geographical and morphological descriptors. Genectic Resources and Crop Evolution 50, 139 148. Upadhyaya, H.D., Reddy, L.J., Gowda, C.L.L., Singh, S., 2006. Identification of diverse groundnut germplasm: Sources of early maturity in a core collection. Field Crops Research 97, 261 271.

PAGE 156

156 Upadhyaya, H.D., Swamy, B.P.M., Goudar, P.V.K., Kullaiswamy, B.Y., Singh, S., 2005. Identification of diverse groundnut germplasm through multienvironment evaluation of a core collection for Asia. Field Crops Research 93, 293 299. Vadez, V., 2014. Root hydraulics: The forgotten side of roots in drought adaptation. Field Crops Research 165, 15 24. Vellidis, G., Liakos, V., Andreis, J.H., Perry, C.D., Porter, W.M., Barnes, E.M., Morgan, K.T., Fraisse, C., Migliaccio, K.W., 2016. Development and assessment of a smartphone application for irrigation scheduling in cotton. Computers and Electronics in Agriculture 127, 249 259. Waines, J.G., Ehdaie, B., 2007. Domestication and crop physiology: Roots of green re volution wheat. Annals of Botany 100, 991 998. Wanjura, D.F., Upchurch, D.R., Mahan, J.R., Burke, J.J., 2002. Cotton yield and applied water relationships under drip irrigation. Agricultural Water Management 55, 217 237. Wasson, A.P., Richards, R.A., Chatr ath, R., Misra, S.C., Sai Prasad, S.V., Rebetzke, G.J., Kirkegaard, J.A., Christopher, J., Watt, M., 2012. Traits and selection strategies to improve root systems and water uptake in water limited wheat crops. Journal of Experimental Botany 63, 3485 3498. Westgate, M.E., Boyer, J.S., 1985. Osmotic adjustment and the inhibition of leaf, root, stem and silk growth at low water potentials in maize. Plant Physiology 74, 882 889. Williams, E.J., Drexler, J.S., 1981. A non destructive method for determining pean ut pod maturity. Peanut Science 8, 134 141. Woodward, J.E., Russell, S.A., Baring, M.R., Cason, J.M., Baughman, T.A., 2015. Effects of Fungicides, Time of Application, and Application Method on Control of Sclerotinia Blight in Peanut. International Journal of Agronomy 2015, 1 8. Wright, D., J. Marois, and N. Default. 2003. Cotton cultural practices and fertility management. University of Florida Institute of Food and Agricultural Sciences Extension. Gainesville, FL. Yang, G., Tang, H., Nie, Y., Zhang, X., 2 011. Response of cotton growth, yield, and biomass to nitrogen split application ratio. European Journal of Agronomy 35, 164 170. Zaman Allah, M., Jenkinson, D., Vadez, V., 2011. A conservative pattern of water use, rather than deep or profuse rooting, is critical for the terminal drought tolerance of chickpea. Journal of Experimental Botany 62, 4239 4252.

PAGE 157

157 Zhou, Z.V., James, C.B., Larson, J.A., Leib, B., Gwathmey, O., 2016. Profit maximizing irrigation for cotton on high available water holding capacity so ils. Agronomy Journal 108, 1571 1580.

PAGE 158

158 BIOGRAPHICAL SKETCH Brendan Zurweller is a native of Missouri where he received his Bachelor of Science in Environmental Science at the University of Missouri Columbia. During his underg raduate education he had a strong interest in soil science which he applied during his graduate research at the University of Missouri Columbia. His research during this appointment was focused on examining the impacts of N management and soil waterlogging on both greenhouse gas emissions and corn grain production. During this research he developed an interest for crop responses to abiotic stress. This led him to pursue a doctoral degree with research encompassing crop trait responses to water and nitrogen management. Upon completing his doctoral degree, Brendan hopes to continue in a career where he can continue conducting agronomic related research.