<%BANNER%>

Citrus Advanced Production System

Permanent Link: http://ufdc.ufl.edu/UFE0044365/00001

Material Information

Title: Citrus Advanced Production System Understanding Water and NPK Uptake and Leaching in Florida Flatwoods and Ridge Soils
Physical Description: 1 online resource (345 p.)
Language: english
Creator: Kadyampakeni, Davie Mayeso
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: citrus -- drip -- fertigation -- irrigation
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Florida citrus production is ranked number one in the nation, accounting for 63% of the 371,700 ha production area in the U.S. California, Texas, and Arizona account for 32.5%, 3.3% and 1.6%, respectively.  Citrus production in Florida has declined over the past 14 years from 342,077 ha in 1998 to 232,470 ha in 2011 largely due to increased urbanization, hurricanes, citrus canker (Xanthomonas axonopodis) and citrus greening (Liberibacter asiaticus).  The uneven rainfall distribution and sandy soils make water and nutrient management extremely difficult.  Thus, novel practices termed advanced citrus production systems (ACPS) using higher tree density, dwarfing rootstocks and a modified open hydroponics system (OHS) were developed to accelerate tree growth and bring young trees into production so growers can break-even within a few years of establishing a grove.  Several field and laboratory experiments coupled with computer simulations were conducted to compare the performance of the intensively managed drip and microsprinkler irrigation and fertigation systems with conventional grower practices on the Florida Flatwoods and Ridge soils.  The Ridge and Flatwoods field studies revealed higher but not significantly different water uptake with ACPS/OHS compared with grower practices. However, tissue nutrient concentration was greater for ACPS/OHS than the grower practices. In addition, ACPS/OHS practices, particularly on the Ridge, increased soil nutrient retention in the root zone by 60-90% compared with conventional fertigation or granular fertilization.  The soil cores indicated greater root length density for ACPS/OHS than grower practice, in the irrigated zones, and in the 0-15 cm soil layer.  HYDRUS-2D model, calibrated with field and laboratory data, showed reasonably good agreements between simulated and measured values suggesting that HYDRUS-2D could successfully be used as a tool for irrigation and nutrient management decision support. Overall, the results underline the importance of using innovative and carefully managed intensive fertigation practices in promoting tree growth and root length density, increasing nutrient and water uptake, and conserving environmental quality while sustaining high citrus yields on Florida’s sandy soils.  The results from the field experiments and computer simulations should allay any fears of potential groundwater contamination associated with proper use of the ACPS/OHS practices.
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.
Statement of Responsibility: by Davie Mayeso Kadyampakeni.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Morgan, Kelly Tindel.
Local: Co-adviser: Nkedi-Kizza, Peter.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044365:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044365/00001

Material Information

Title: Citrus Advanced Production System Understanding Water and NPK Uptake and Leaching in Florida Flatwoods and Ridge Soils
Physical Description: 1 online resource (345 p.)
Language: english
Creator: Kadyampakeni, Davie Mayeso
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: citrus -- drip -- fertigation -- irrigation
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Florida citrus production is ranked number one in the nation, accounting for 63% of the 371,700 ha production area in the U.S. California, Texas, and Arizona account for 32.5%, 3.3% and 1.6%, respectively.  Citrus production in Florida has declined over the past 14 years from 342,077 ha in 1998 to 232,470 ha in 2011 largely due to increased urbanization, hurricanes, citrus canker (Xanthomonas axonopodis) and citrus greening (Liberibacter asiaticus).  The uneven rainfall distribution and sandy soils make water and nutrient management extremely difficult.  Thus, novel practices termed advanced citrus production systems (ACPS) using higher tree density, dwarfing rootstocks and a modified open hydroponics system (OHS) were developed to accelerate tree growth and bring young trees into production so growers can break-even within a few years of establishing a grove.  Several field and laboratory experiments coupled with computer simulations were conducted to compare the performance of the intensively managed drip and microsprinkler irrigation and fertigation systems with conventional grower practices on the Florida Flatwoods and Ridge soils.  The Ridge and Flatwoods field studies revealed higher but not significantly different water uptake with ACPS/OHS compared with grower practices. However, tissue nutrient concentration was greater for ACPS/OHS than the grower practices. In addition, ACPS/OHS practices, particularly on the Ridge, increased soil nutrient retention in the root zone by 60-90% compared with conventional fertigation or granular fertilization.  The soil cores indicated greater root length density for ACPS/OHS than grower practice, in the irrigated zones, and in the 0-15 cm soil layer.  HYDRUS-2D model, calibrated with field and laboratory data, showed reasonably good agreements between simulated and measured values suggesting that HYDRUS-2D could successfully be used as a tool for irrigation and nutrient management decision support. Overall, the results underline the importance of using innovative and carefully managed intensive fertigation practices in promoting tree growth and root length density, increasing nutrient and water uptake, and conserving environmental quality while sustaining high citrus yields on Florida’s sandy soils.  The results from the field experiments and computer simulations should allay any fears of potential groundwater contamination associated with proper use of the ACPS/OHS practices.
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.
Statement of Responsibility: by Davie Mayeso Kadyampakeni.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Morgan, Kelly Tindel.
Local: Co-adviser: Nkedi-Kizza, Peter.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044365:00001


This item has the following downloads:


Full Text

PAGE 1

1 CITRUS ADVANCED PRODUCTION SYSTEM : UNDERSTANDING WATER AND NPK UPTAKE AND LEACHING IN FLORIDA FLATWOODS AND RIDGE SOILS By DAVIE MAYESO KADYAMPAKENI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 2

PAGE 2

2 201 2 D avie M ayeso K adyampakeni

PAGE 3

3 To my wife Iness, son Atikonda, dad Simfoliano and mum Ernestina Kadyampakeni and my siblings Do minic, Honoratus, Felicity, Perpetual Anthony, Auleria and Carnisius

PAGE 4

4 ACKNOWLEDG E MENTS First and foremost, I would like to thank the Almighty God for helping me get thus far on the academic ladder. In a special and grateful way, I wish to thank my c o a dvisors Drs. Kelly Morgan and Peter Nkedi Kizza for their generous financial, moral and material support. I would like to sincerely thank them for their patience and understanding (of my personal and professional lives) and their rare ability to comb ine critical evaluation of my write ups and/or manuscripts with warm personality and credible friendship. I feel privileged and program in Immokalee and gain hands on experien ce in using advanced laboratory equipment and software. I would like to thank Drs. Arnold Schumann, James Jawitz, Thomas Obreza and James Jones for accepting to be on my committee and providing material support, literature and useful suggestions for my wor k. Dr. Schumann is hereby acknowledged for giving me laboratory space and all the necessary help for my work at Lake Alfred. I would also like to thank Drs. Nkedi Kizza, Jawitz and Jones for their classroom instruction on Environmental Soil Physics, Cont aminant Subsurface Hydrology, and Biological and Agricultural Systems Simulation, respectively. Drs. Dean Rhue, William Harris, Jerry Sartain, and Samira Daroub are hereby tha nked for their classroom instruction. I would like to thank the Southwest Florida Water Management District for supporting this research and the Soil and Water Science Department for the matching assistantship. I am also grateful to the sponsors of Grin ter, William Robertson and Sam Polston Graduate Fellowships and the A.S. Herlong and Doris, Earl and Verna Lowe

PAGE 5

5 Scholarships. The help and friendship of Denise Bates, Kristy Sytsma, Janice Hill, Kevin Hill and Julie Carson in administrative, computing tec hnology and library and information services at Immokalee are gratefully acknowledged for making my work a lot easier. Rhiannon Pollard and Michael Sisk, the respective former and current Student Services Coordinator for the Soil and Water Science Departm ent are gratefully recognized for their support and timely advice on paper work regarding admission, course registration and graduation. Dr. Monica Ozores Hampton is also gratefully acknowledged for helping my family settle down in Immokalee. The author r ecognizes the friendship and support of Drs. Andrew Ogram (Graduate Coordinator of the Soil and Water Science Department), Mark Rieger (Associate Dean of the College of Agricultural and Life Sciences), David Sammons (Dean of the University of Florida Inter national Center) and Walter Bowen (Director of UF/IFAS International Programs). I would like to thank the following workmates and colleagues in Gainesville for their support in many ways: Drs. Gabriel Kasozi, Sampson Agyin Birikorang, Nicholas Kiggundu M ichael Miyittah, Hiral Gohil and Rajendra Paudel; Kafui Awuma, Jorge Leiva, Augustine Muwamba, Moshik Doron, Mike Jerauld, Lucy Ngatia Rish Prasad, and Jongsung Kim I am grateful to the following colleagues for their support with data collection, labora tory procedures, use of software and other equipments at Immokalee and Lake Alfred: Drs. Shinjiro Sato, Kamal Mahmoud and Kiran Mann; Laura Waldo, Smita Barkataky, Wafaa Mohamoud, Assma Zekri, Ann Summerals and Shengsen Wang. My friends at Immokalee Melis sa Benitez, Sunehali Sharma and Utpal Handque are also thanked for their friendship and help.

PAGE 6

6 I cherish the friendship and support of my fellow Malawian students who pursued various graduate programs at UF: Bonet Kamwana, Innocent Thindwa, Pearson Soko, Fi skani Nkana, Lucy Nyirenda, Jonathan Chiputula, Wycliffe Kumwenda, Aubrey Chinseu, Matrina Soko, Felix Makondi, Hamie Chakana, Chunala Njombwa, Suzgo Chapa and Donald Kazanga. The families of Dr. Nkedi Kizza, Dr. Chikagwa Malunga and Dr. Lergo are thanked for their friendship and company. My friend Thomson Paris, his dad Trevor, mum Cindy, brother Taylor and sister Sarah are thanked for being materially, spiritually, financially and morally so supportive to me and my family. I have vivid memories of my time in Jacksonville, St. Augustine, Chatanooga, Baltimore and Washington D.C. I am also grateful to Scott Croxton and his mum for their friendship and support. Over and above all, I would like to recognize and thank my wife Iness and son Atikonda for th eir company and incessant support during the final and most demanding times of my graduate program. My dad Simfoliano and mum Ernestina Kadyampakeni, my mother in law Martha Mhango and my siblings Dominic, Honoratus, Felicity, Perpetual, Anthony, Auleria and Carnisius; my in laws and many nephews and nieces, my uncles and aunts and cousins have always inspired and sustained my urge for higher academic accomplishment and professional advancement through their prayers, advice and encouragement.

PAGE 7

7 TABLE OF CO NTENTS page ACKNOWLEDGEMENTS ................................ ................................ ............................... 4 LIST OF TABLES ................................ ................................ ................................ .......... 11 LIST OF ABBREVIATIONS ................................ ................................ ........................... 17 ABSTRACT ................................ ................................ ................................ ................... 20 C HAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 22 Justification for Research on Citrus Irrigation and Nutrient Management in Florida ................................ ................................ ................................ .................. 24 ................................ ................ 24 Climate ................................ ................................ ................................ ............. 25 Citrus Canker and Greening Diseases ................................ ............................. 26 Planting Densities ................................ ................................ ............................. 28 Citrus Root Length Density ................................ ................................ ............... 28 Overview of the Dissertation ................................ ................................ ................... 29 General Research Goals and Hypotheses ................................ .............................. 31 Summ ary ................................ ................................ ................................ ................ 32 2 LITERATURE REVIEW ................................ ................................ .......................... 34 The Open Hydroponic System and Advanced Production Systems ....................... 35 Concepts ................................ ................................ ................................ .......... 35 Maximize water and nutrient efficiency ................................ ...................... 35 Concentrate roots in irrigated zone ................................ ............................ 36 Reduce nutrient leaching ................................ ................................ ........... 37 ....... 37 Tree density ................................ ................................ ............................... 39 Tree size control with rootstocks ................................ ................................ 39 Fertilizer Demand and Nutrient Uptake in Citrus ................................ ..................... 40 Biomass Development with Time ................................ ................................ ..... 40 Nutrient Requirements for Biomass and Fruit Production ................................ 42 Nutrient Uptake and Nutrient Use Efficiency ................................ ........................... 48 Citrus Nutrient Management ................................ ................................ ............. 48 Extraction Methods for N Forms, P, and K from Soils and Plant Tissue .......... 49 Irrigation Design and Scheduling Drip and Microsprinkler Irrigation ....................... 52 Evapotranspiration Ca lculations ................................ ................................ ....... 52 Citrus Crop Coefficients ................................ ................................ .................... 54 Water Use Efficiency ................................ ................................ ........................ 54

PAGE 8

8 B romide as a Tracer for Water Movement in the Soil ................................ ....... 55 Irrigation Methods ................................ ................................ ............................. 56 Irrigation Control Methods ................................ ................................ ................ 57 Citrus Root Density Distribution ................................ ................................ ........ 57 Process Oriented Models for Solute Transport, Water and Nutrient Uptake ........... 59 Types and Use of Models in Agriculture ................................ ........................... 59 Comparing Soil Water and Hydrologic Models ................................ ................. 61 Models Used for Citrus Production ................................ ................................ ... 62 Summary ................................ ................................ ................................ ................ 63 3 NUTRIENT UPTAKE EFFICIENCY AND DISTRIBUTION IN SITU FROM THE CITRUS ROOT ZONE ................................ ................................ ............................ 68 Materials and Methods ................................ ................................ ............................ 70 Site Conditions ................................ ................................ ................................ 70 Study Treatments and Experimental Design ................................ .................... 70 Plant Tissue and Soil Sampling Design and Analytical Methods ...................... 71 Soil sampling ................................ ................................ .............................. 71 Water sample collection and processing ................................ .................... 72 Extraction of NH 4 N, NO 3 N, P, Br and K ................................ ................... 73 Analysis of soil extracts and wa ter samples ................................ ............... 74 Plant tissue sampling and analysis ................................ ................................ ... 74 Leaf sampling ................................ ................................ ............................. 74 Destructive tree sampling and tissue processing ................................ ............. 75 Tissue analysis ................................ ................................ .......................... 76 Quality Control of Plant Tissue and Soil Sample Analysis ................................ 76 Data Analysis ................................ ................................ ................................ ... 77 Results and Discussion ................................ ................................ ........................... 77 Leaf NPK Concentratio n as a Function of Irrigation System ............................. 77 NPK Distribution in the Citrus Root Zone as a Function of Time, Depth and Lateral Distance ................................ ................................ ............................ 78 N, P, Br and K Leaching in the Irrigated and Nonirrigated Zones ..................... 83 Water Quality Analysis ................................ ................................ ..................... 89 Biomass and Nutrient Distribution as a Function of Irrigation Practice ............. 90 Summary ................................ ................................ ................................ ................ 92 4 EFFECTS OF FERTIGATION AND IRRIGATION RATES ON ROOT LENGTH DISTRIBUTION AND TREE SIZE ................................ ................................ ........ 122 Materials and Methods ................................ ................................ .......................... 125 Description of Study Sites and Treatments ................................ .................... 125 Root Sampling Methods ................................ ................................ ................. 126 Estimation of Tree Growth Characteristics ................................ ..................... 127 Statistical Analysis ................................ ................................ .......................... 127 Results and Discussion ................................ ................................ ......................... 128 Correlation of RLD Measured by Intersection Method versus Scanning Method ................................ ................................ ................................ ........ 128

PAGE 9

9 RLD Distribution as a Function of Irrigation Method, Time and Soil Depth ..... 129 Effect of Fertigation Method on Trunk Cross Sectional Area and Canopy Volume ................................ ................................ ................................ ........ 134 Summary ................................ ................................ ................................ .............. 135 5 EFFECTS OF IRRIGATION METHOD AND FREQUENCY ON CITRUS WATER UPTAKE AND SOIL MOISTURE DISTRIBUTION ................................ .. 147 Materials and Methods ................................ ................................ .......................... 150 Experimental design and irrigation methods ................................ ................... 150 Estimation of Soil M oisture ................................ ................................ ............. 150 Estimation of Crop Water Uptake and K c ................................ ........................ 151 Results and Discussion ................................ ................................ ......................... 153 Tree characteristics at Immokalee and Lake Alfred ................................ ........ 153 Water Uptake at Immokalee and Lake Alfred ................................ ................. 154 Soil moisture d istribution at Lake Alfred and Immokalee ................................ 161 Factors affecting water uptake on the two soils ................................ .............. 164 Summary ................................ ................................ ................................ .............. 165 6 CALIBRATION AND VALIDATION OF WATER, N, P, BR AND K MOVEMENT ON A FLORIDA SPODOSOL AND ENTISOL USING HYDRUS 2D .................... 194 Materials and Methods ................................ ................................ .......................... 195 Governing Equations and Parameters for Water Flow, Nutrient Transport and Uptake ................................ ................................ ................................ .. 195 Model Calibration Processes ................................ ................................ .......... 198 Sorption isotherms determination ................................ ............................ 198 Determination of soil water retention and hydraulic functions .................. 200 Sensitivity Analysis of Selected Parameters for HYDRUS 2D ........................ 203 Simulation Domain Microsprinkler irrigation ................................ ................... 206 Simul ation Domain Drip irrigation ................................ ................................ ... 206 Results and Discussion ................................ ................................ ......................... 207 Sensitivity analysis and calibration of selected model parameters ................. 207 Water, Br, K, P, NO 3 and NH 4 movement with drip and microsprinkler irrigation ................................ ................................ ................................ ...... 208 Phosphorus movement with microsprinkler irrigati on as function of K D value 209 Investigating bromide, nitrate and water movement using weather data from Immokalee and Lake Alfred. ................................ ................................ ........ 210 Summary ................................ ................................ ................................ .............. 211 7 CONCLUSIONS ................................ ................................ ................................ ... 228 APPENDIX A SUPPLEMENTARY FIGURES TO CHAPTERS 3, 4 AND 5 ................................ 234 B CHARACTERIZATION OF SORPTION ISOTHERMS FOR AMMONIUM N, K AND P ON THE FLATWOODS AND RIDGE SOILS ................................ ............ 278

PAGE 10

10 C SOIL WATER CHARACTERISTIC CURVE S AND HYDRAULIC FUNCTIONS ... 289 D SCHEMATIC FIELD DIAGRAM S SHOWING THE SET UP OF DRIP AND MIC R O SPR INKLER IRRIGATION S YSTE MS ................................ ...................... 296 E AVERAGE MONTHLY TEMPERATURE, RELATIVE HUMIDITY, RAINFALL, SOLAR RADIATION AND EVAPOTRANSPIRATION ................................ ......... 304 F CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA ................................ ................................ ................................ .................... 306 G EXPERIMENTAL SET UP FOR THE SORPTION STUDY ................................ ... 314 H SORPTION COEFFICIENTS FOR NH 4 + AN D K + ON IMMOKALEE AND CANDLER FINE SAND USING FERTILIZER MIXTURE IN TAP WATER ........... 315 I SORPTION COEFFICIENTS FOR P ON IMMOKALEE AND CANDLER FINE SAND ................................ ................................ ................................ .................... 316 LIST OF REFERENCES ................................ ................................ ............................. 317 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 344

PAGE 11

11 LIST OF TABLES Table page 2 1 Typical percent biomass distribution (dry weight basis) in oranges from different parts of the world ................................ ................................ .................. 65 2 2 Typical nutrient uptake rates in oranges ................................ ............................. 66 2 3 Soil test interpretation for soil P extraction methods compared with Mehlich 1 extractant § ................................ ................................ ................................ .......... 67 2 4 Guidelines for interpretations of orange tree leaf ana lysis based on 4 to 6 month old spring flush leaves from non fruiting twigs ................................ ........ 67 3 1 2M KCl extrac table NH4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in June 2009 at SW FREC ................................ ..................... 98 3 2 2M KCl extrac table NH4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in August 2009 at SWFREC ................................ .................. 99 3 3 2M KCl extrac table NH 4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in June 2010 at SWFREC ................................ ................... 100 3 4 2M KCl extrac table NH 4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in December 2009 at the Lake Alfred site ........................... 101 3 5 2M KCl extrac table NH 4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in July 2010 at the Lake A lfred site ................................ ..... 102 3 6 Fresh and dry tissue weight for samples collected in July 2011 at Immokalee 117 3 7 Fresh and dry tissue weight for samples collected in August 2011 at the Lake Alfred site ................................ ................................ ................................ ......... 118 3 8 N, P and K concentration in tissues collected in July 2011 at the Immokalee site ................................ ................................ ................................ .................... 119 3 9 N, P and K concentration in tissues collected in August 2011 at the Lake Alfred site ................................ ................................ ................................ ......... 1 19 3 10 Nitrogen, phosphorus and potassium accumulation on Immokalee sand ......... 120 3 11 N, P and K accumulation in 2011 at the Lake Alfred and Immokalee sites ....... 121 4 1 Models for RLD estimation at CREC ................................ ................................ 137 4 2 Models for RLD estimation at SWFREC ................................ ........................... 138

PAGE 12

12 4 3 RLD as a function of irrigation method, soil depth and distance from the tree at SWFREC in June 2009 ................................ ................................ ................ 139 4 4 RLD as a function of irrigation method, soil depth and distance from the tree at SWFREC in June 2010 ................................ ................................ ................ 140 4 5 RL D as a function of irrigation method, soil depth and distance from the tree at the Lake Alfred site in December 2009 ................................ ......................... 141 4 6 RLD as a function of irrigation method, soil depth and distance from the tree at the Lake Alfred site in July 2010 ................................ ................................ ... 142 4 7 Trunk cross sectional area as function of fertigation method at the Lake Alfred site ................................ ................................ ................................ ......... 146 5 1 Average leaf area ................................ ................................ ............................. 167 5 2 Tree canopy volume (CV), stem cross sectional area (SCA), and trunk cross sectional area (TCA) ................................ ................................ ......................... 167 5 3 Linear regression models relating cumulative water uptake to tree and soil characteristics at the Lake Alfred site in July 2010 and September 2011 ....... 192 5 4 Multiple linear regr ession model coefficients for cumulative water uptake ....... 193 6 1 Selected parameters for sensitivity analysis for simulating water flow and nutrient movement in citrus using HYDRUS 2D ................................ ............... 215 6 3 Simulation experiment scenarios for the Ridge and Flatwoods soils ................ 217 6 4 Soil physical characteristics and initial conditions of the Immokalee and Candler fine sands ................................ ................................ ............................ 217 6 5 Sensitivity indices for selected parameters for soil available water, P, ammonium and K movement using HYDRUS 2D ................................ ............ 218 6 6 Statistical comparison between the observed and simulated water contents in spring and summer on Candler and Immokalee sand ................................ ...... 225 6 7 Statistical comparison betwee n the observed and simulated Br, NO 3 NH 4 M1P and M1K on Candler and Immokalee sand ................................ .............. 226

PAGE 13

13 LIST OF FIGURES Figure page 1 1 Typic al soil orders of Florida ................................ ................................ ............... 33 3 1 Destructive tree sampling in July 2011 at Immokalee with the root zone of the tree marked to 30 cm depth ................................ ................................ ................ 95 3 2 Leaf NPK concentration deter mined in June 2009 at Immokalee ....................... 96 3 3 Leaf NPK concentration determined in August 2011 at the Lake Alfred site ...... 97 3 4 Lateral ammonium N distribution at 0 30 cm soil depth in June 2009 and 2010 at the Immokalee site ................................ ................................ ............... 103 3 5 Lateral nitrate N distribution in June 2009 and 2010 at Immokalee site ........... 104 3 6 Lateral ammonium N distribution in December 2009 on Candler fine sand ...... 105 3 7 Lateral ammonium N distrib ution in July 2010 at the Lake Alfred site .............. 106 3 8 Lateral nitrate N distribution in December 2009 at the Lake Alfred site ............ 107 3 9 Lateral nitrate N distribution in July 2010 at the Lake Alfred site ...................... 108 3 10 Lateral Mehlich 1 P distribution at Immokalee site in June 2009 and 2010 ...... 109 3 11 Lateral Mehlich 1 P distribution in the 0 30 cm depth layer at the Lake Alfred site in December 2009 ................................ ................................ ...................... 110 3 12 Lateral Mehlich 1 P distribution in the 0 30 cm depth layer at the Lake Alfred site in July 2010 ................................ ................................ ................................ 111 3 13 Lateral Mehlich 1 K distribution at Immokalee in June 2009 and 2010 ............. 112 3 14 Lateral Mehlich 1 K distribution at the Lake Alfred site in December 2009 ....... 113 3 15 Lateral Mehlich 1 K distribution at the Lake Alfred site in July 2010 ................. 114 3 16 Vertical nitrate N and ammonium N distribution in June 2010 at Immokalee site and in July 2010 at the Lake Alfred site ................................ ..................... 115 4 1 Canopy volume as a function of fer tilization practice at th e Lake Alfred site. ... 143 4 2 Trunk cross sectional area as a function of fertigation p ractice at the Immokalee site. ................................ ................................ ................................ 144 4 3 Canopy volume as a function of fertigation method at the Immokalee site ....... 145

PAGE 14

14 5 1 Linear correlations of leaf area index and canopy volume as a function of leaf area in March 2 011 at the Lake Alfred site ................................ ....................... 168 5 2 Correlations of leaf area index and canopy volume as a function of total leaf area at Immokalee site in March 2011 ................................ .............................. 169 5 3 Average hourly sap flow in July, 2010 and March, 2011 at Lake Alfred site ..... 170 5 4 Average daily sap flow in July, 2010 at the Lake Alfred site ............................. 171 5 5 Average daily sap flow in Marc h, 2011 at the Lake Alfred site ......................... 171 5 6 Average hourly sap flow in February March 2011 at SWFREC. ....................... 172 5 7 Average daily sap flow in February March 2011 at SWFREC. ......................... 173 5 8 Average hourly flow in June 2011 at the Immokalee site ................................ .. 174 5 9 Average hourly sap flow in August September, 2011 at the Lake Alfred site ... 175 5 10 Average daily sap flow in June 2011 at the Immokalee sit e ............................. 176 5 11 Average daily sap flow in August September 2011 at the Lake Alfred site ...... 177 5 12 Average hourly soil moisture distribu tion in July 2010 at the Lake Alfred site measured at 10 and 45 cm soil depth layers ................................ ................... 178 5 13 Average daily soil moisture distribution in July 2010 at the Lake Alfred site measured at 10 cm so il depth layer ................................ ................................ .. 179 5 14 Soil moisture distribution in July 2010 at the Lake Alfred site measured at 45 cm soil depth layer ................................ ................................ ............................ 179 5 1 5 Average hourly soil moisture distribution at the Lake Alfred site measured at 10 cm (top) and 45 cm (bottom) soil depth layers in March 2011. .................... 180 5 16 Daily soil moisture distribution at the Lake Alfred site measured at 10 cm soil depth layer in March 2011 ................................ ................................ ................ 181 5 17 Daily soil moisture distribution at the Lake Alfred site measured at 45 cm soil depth layer in March 2011 ................................ ................................ ................ 181 5 18 Average hourly soil moisture distribution at the Lake Alfred site measured at 10 and 45 cm soil depth layers in August September 2011 ............................ 182 5 19 Average daily soil moisture distribution at the Lake Alfred site measured at 10 cm soil depth layer in August September 2011 ................................ ........... 183

PAGE 15

15 5 20 Average daily soil moisture distribution at the Lake Alfred site measured at 45 cm soil depth layer in August September 2011 ................................ ........... 184 5 21 Soil moisture distribution for DOHS in February March 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers ....................... 185 5 22 Soil moisture distribution for DOHS in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers. ............................ 186 5 23 Soil moisture distribution for MOHS in February March 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers ....................... 187 5 24 Soi l moisture distribution for MOHS in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers ............................. 188 5 25 Soil moisture distribution for CMP in February March 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers ............................. 189 5 26 Soil moisture distribution for CMP in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soi l depth layers ................................ .............. 190 5 27 Correlation of water uptake and canopy volume at the Immokalee and Lake Alfred sites ................................ ................................ ................................ ........ 191 6 1 A forrester d iagram describing the conceptual model for water and nutrient uptake and movement processes ................................ ................................ ..... 213 6 2 Calibration of HYDRUS 2D for simulating soil water content at 10 cm soil depth at Lake Alfred site using drip irrigation ................................ .................... 214 6 3 Calibration of HYDRUS 2D for simulation soil water content at 40 cm soil depth at Lake Alfred site using microsprinkler irrigation ................................ ... 214 6 4 Calibration of HYDRUS model for simulating ammonium N movement on Candler fine sand ................................ ................................ ............................. 215 6 5 Soil Br monitored at 15 and 60 cm depth using drip irri gation at the Lake Alfred site ................................ ................................ ................................ ......... 219 6 6 Measured and simulated Br concentration at 15 and 60 cm at Immokalee site using microsprinkler irrigation ................................ ................................ ........... 220 6 7 Soil P monitored at 15 cm depth using drip irrigation at the Lake Alfred site .... 221 6 8 Simulated and measured cumulative nitrate concentration using microsprinkler irrigat ion at the Immokalee site ................................ .................. 221 6 9 Simulated and measured cumulative ammonium concentration using drip irrigation at the Immokalee site ................................ ................................ ......... 222

PAGE 16

16 6 10 Cumulative K distribution at 15 cm soil depth at Immokalee site using microsprinkler irrigation ................................ ................................ .................... 223 6 11 Cumulative K distribution at 15 cm soil depth at Immokalee site using drip irrigation ................................ ................................ ................................ ............ 223 6 12 Phosphorus movement on Candler and Immokalee fine sand depending on K D value estimated using HYDRUS 1D ................................ ............................ 224 6 13 Simulated nitrate, bromide and water movement over a 90 day period at 60 cm using grower practice ................................ ................................ .................. 227

PAGE 17

17 LIST OF ABBREVIATION S AB DTPA Ammonium bicarbonate diethylenetriamine pentaacetic acid ACPS Advanced Citrus Production Systems ANOVA Analysis of Variance ARS Agricultural Research Servi ce ASWD Available soil water depletion BMP Best Management Practice CDE Convection Dispersion Equation CEC Cation Exchange Capacity CMP Conventional Microsprinkler Practice CREC Citrus Research and Education Center CWMS Citrus Water Managemen t System DM Dry matter DMRT Duncan Multiple Range Test DOHS Drip Open Hydroponics System DSSAT Decision Support System for Agrotechnology Transfer DW Dry weight ER Effective rainfall E sap Daily sapflow per unit land area ET Evapotranspirati on ET o Reference evapotranspiration FRLD Fibrous root length density FW Fresh weight

PAGE 18

18 GLM General Linear Model GSA Global Sensitivity Analysis HLB Huanglongbing ICP Inductively Coupled Plasma ICP AES Inductively Coupled Plasma Atomic Emissio n Spectrometry IFP Intensive Fertigation Practice K c Crop coefficient LAI Leaf Area Index LEACHM Leaching and Chemistry Model M1K Mehlich 1 extractable potassium M1P Mehlich 1 extractable phosphorus MOHS Microsprinkler Open Hydroponics System NCSWAP Nitrogen, Carbon, Soil, Water and Plant OHS Open Hydroponics System NPV Net Present Value RAW Readily available water RLD Root length density RZWQM Root Zone Water Quality Model SHB Stem heat balance SWASIM Soil Water Simulation Mod el SWATRE Soil Water and Actual Transpiration, Extended SWFREC Southwest Florida Research and Education Center TAW Total available water

PAGE 19

19 TCA Trunk cross sectional area TSS Total Soluble Solids UF Upflux UF/IFAS University of Florida/Institute of Food and Agricultural Sciences USA United States of America USDA United States Department of Agriculture

PAGE 20

20 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requ irements for the Degree of Doctor of Philosophy CITRUS ADVANCED PRODUCTION SYSTEM: UNDERSTANDING WATER AND NPK UPTAKE AND LEACHING IN FLORIDA FLATWOODS AND RIDGE SOILS By Davie Mayeso Kadyampakeni August 2012 Chair: Kelly T. Morgan Cochair: Peter Nkedi Kizza Major: Soil and Water Science Florida citrus production is ranked number one in the nation, accounting for 63% of the 371,700 ha production area in the U.S California, Texas, and Arizona account for 32.5%, 3.3% and 1.6%, respectively. Citrus prod uction in Florida has declined over the past 14 years from 342,077 ha in 1998 to 232,470 ha in 2011 largely due to increased urbanization, hurricanes, citrus canker ( Xanthomonas axonopodis) and citrus greening ( Liberibacter asiaticus ). The uneven rainfall distribution and sandy soils make water and nutrient management extremely difficult. Thus, novel practices termed advanced citrus production systems (ACPS) using higher tree density, dwarfing rootstocks and a modified open hydroponics system (OHS) were d eveloped to accelerate tree growth and bring young trees into production so growers can break even within a few years of establishing a grove. Several field and laboratory experiments coupled with computer simulations were conducted to compare the perform ance of the intensively managed drip and microsprinkler irrigation and fertigation systems with conventional grower practices on the Florida Flatwoods and Ridge soils.

PAGE 21

21 The Ridge and Flatwoods field studies revealed higher but not significantly different water uptake with ACPS/OHS compared with grower practices. However, tissue nutrient concentration was greater for ACPS/OHS than the grower practices. In addition, ACPS/OHS practices, particularly on the Ridge, increased soil nutrient retention in the root zone by 60 90% compared with conventional fertigation or granular fertilization. The soil cores indicated greater root length density for ACPS/OHS than grower practice, in the irrigated zones, and in the 0 15 cm soil layer. HYDRUS 2D model, calibrated wi th field and laboratory data, showed reasonably good agreements between simulated and measured values suggesting that HYDRUS 2D could successfully be used as a tool for irrigation and nutrient management decision support. Overall, the results underline the importance of using innovative and carefully managed intensive fertigation practices in promoting tree growth and root length density, increasing nutrient and water uptake, and conserving environmental quality while sustaining high citrus yields on Florid experiments and computer simulations should allay any fears of potential groundwater contamination associated with proper use of the ACPS/OHS practices.

PAGE 22

22 CHAPTER 1 INTRODUCTION Citrus is one of the most importa nt crops in Florida with an annual value of $1.1 billion dollars (USDA, 2011). In 2010, Florida ranked number one in the nation for citrus production, accounting for 63% of the 371,700 ha production area in the U.S. while states of Arizona Texas and Cal ifornia accounted for 6,075 ha, 12,285 ha and 120, 870 ha, respectively. At a global scale, the U.S produced 12% of the 83 million ton world citrus production in 2010 (USDA, 2011). Research data from several studies show that increasing water costs and e nvironmental concerns create a need for more efficient management practices for citrus production (Lamb et al., 1999; Alva et al., 2003; Paramasivam et al., 2000a; 2001; Alva et al., 2006a, b). Irrigating to meet crop evapotranspiration (ET) demand and fe rtigation at optimal nutrient levels have the potential to increase production efficiency. The modifications to current irrigation water and nutrient management recommendations are termed open hydroponics system (OHS) and advanced citrus production system s (ACPS). OHS is an integrated system of irrigation, nutrition and horticultural practices that was developed in Spain to improve production on gravel based soils with low fertility (Martinez Valero and Fernandez, 2004; Falivene et al., 2005). According to Stover et al. (2008), OHS provides tight control over water and nutrient mediated plant growth and development using irrigation to train the root system into a limited area and fertigates with daily nutrient requirements. The ACPS is a short to medium term approach to citrus water and nutrient management being evaluated in Florida citrus groves for sustainable, profitable citrus production in the presence of greening and canker diseases with the goal of compressing and enhancing the citrus

PAGE 23

23 production c ycle so economic payback can be reached in fewer years to offset some of the disease losses (Schumann et al., 2009). Elements of OHS that have been incorporated into an Advanced Citrus Production System (ACPS) include a) intensive daily fertigation with c omplete balanced nutrient formula for early high yields and control of shoot and root growth; 2) high density planting to enable early high yields and early return to investment, and 3) a suitable rootstock adaptable to close spacing and intensive fertigat ion, capable of promoting vigorous tree growth and high root density in the fertigated zone (Morgan et al., 2009 b ). Muraro (2008) described the costs associated with shifting from current production systems to ACPS/OHS. The added costs to establish an AC PS/OHS with 890 trees ha 1 are about $13,541 ha 1 more than if a block is replanted to a more typical density of 371 trees ha 1 owing to buying 519 additional trees ha 1 planting costs, irrigation/bed preparation and young tree management (Muraro, 2008; R oka et al., 2009). However, net present value (NPV) analysis performed by Roka et al. (2009) over a 15 year horizon and a constant delivered in price of $1.20 per pound solids, showed a cumulative NPV of $7,949 ha 1 for 890 trees ha 1 planting and a negati ve ($833) ha 1 for a 371 trees ha 1 planting. The higher returns from ACPS affords a grower a greater cushion against low market prices than a 371 trees ha 1 planting. Thus enhancing production from young trees carries two benefits: 1) sustained higher fru it yield over time, and 2) increasing net returns earlier in the cashflow stream when discount rates are relatively higher (Roka et al., 2009). Despite these postulated notions, research studies on the effect of irrigating at various ET levels and specific NPK levels using OHS on the productivity of young citrus

PAGE 24

24 trees have not been adequately conducted on Florida Flatwoods and Ridge soils. This is key to understanding, in detail, interacting factors and processes that govern citrus water and nutrient uptak e and movement of nutrients such as N, P and K in the citrus root zone. Also, use of the OHS for fertilizer management in citrus production on Florida soils with high percentage of sand (>85%) and low organic matter content (<2%) may further reduce nutrie nt leaching and subsequent pollution of groundwater. This study hypothesizes that proper scheduling of irrigation water by drip or microsprinkler using OHS will improve water and nutrient use efficiency thus helping farmers more efficiently manage i nputs in an ecologically sound manner, and attain high citrus growth and/or yields. For optimum citrus tree growth and yield, fertilization rate and timing must be accompanied by efficient water management to avoid leaching of nutrients below the root zon e thus increasing nutrient use efficiency. The research objectives of the studies in the following chapters focus on 1) improved crop growth and fertilizer use efficiency, 2) irrigation management optimization, and 3) modeling of soil fertigation interact ions with the goals of (1) realizing lower water and fertilizer use, (2) ensuring sustainable citrus yields and (3) avoiding environmental pollution associated with nutrient leaching from the citrus root zone (Morgan and Hanlon, 2006). Justification for Research on Citrus Irrigation and Nutrient Management in Florida Soil Types i Citrus Growing Regions Most Florida citrus is grown on sandy soils that are unable to retain more than a minimal amount of soluble plant nutrients against leaching by rainfall or excessive irrigation (Obreza and Collins, 2008). Typical soil orders in the Florida citrus producing regions are Entisols on the Florida Ridge and Spodosols and Alfisols in the Flatwoods

PAGE 25

25 (Figure 1 1) Entisols, found mostly in central Florida are characterized by excessive drainage, good aeration and a deep root zone (Alva et al. 1998; Fares and Alva, 1999; 2000 a, b ; Morgan et al. 2006 b ; Fares et al. 2008; Obreza and Collins, 2008). Th ese soils are ascribed to high hydraulic conductivity for Entisols ranging from 15 215cm h 1 with percentage sand > 95% (Paramasivam et al. 2001; 2002; Obreza and Collins, 2008). In the Flatwoods, the Alfisols (except Winder soil series that has 85% sand) and Spodosols contain about 94 98% sand in the top 45 cm making irrigation water and nutrient management extremely difficult (Obreza and Collins, 2008). Generally, these soils have low water holding and nutrient retention capacity due to the sandy soil characteristic and low organic matter and thus require u se of intensive and well managed irrigation and fertigation systems that promote high water and nutrient use efficiency for high citrus yields. Climate Citrus trees in Florida must be irrigated to reach maximum production owing to uneven rainfall distrib ution and low soil water holding capacity (Morgan et al., 2006 b ). However, citrus irrigation and crop water requirements vary with climatic conditions and variety (Rogers and Barholic, 1976; Boman, 1994; Fares and Alva, 1999). Florida citrus water require ment is reported to range from 820 to 1280 mm yr 1 (Rogers et al., 1983) while 60% of the average annual rainfall (approximately 1386 mm) is distributed in the summer months of May through August ( Paramasivam et al., 2001; Obreza and Pitts, 2002; Paramasiv am et al. 2002) Thus, the rain is not distributed uniformly throughout the year stressing the need for supplementary irrigation.

PAGE 26

26 Citrus Canker a nd Greening Disease s According to USDA (2011), citrus production in Florida decreased from 386,137 ha in 1966 to 249,317 ha in 2010, as a result of increased urbanization, hurricanes, citrus canker ( Xanthomonas axonopodis) and citrus greening ( Liberibacter asiaticus ). The latter two diseases have eliminated 10 to 30% of trees and reduced yields of other trees in some citrus groves in Florida (Gottwald et al., 2002a b; Irey et al., 2008 ). In a study on the spread of citrus greening (also called Huanglongbing (HLB)) in Florida, Manjunath et al. (200 8 ) found that 9% of plant samples from 43 different counties test ed positive for Liberibacter asiaticus. Citrus bacterial canker disease is a quarantine pest for many citrus growing countries (Gottwald et al., 2002a). Citrus canker occurs primarily in tropical and subtropical climates where considerable rainfall acco mpanies warm temperatures as is the case with Florida (Polex et al. 2007). The disease is exacerbated when wet conditions occur during periods of shoot emergence and development of young citrus fruit (Halbert and Manjunath, 2004 ; Polex et al. 2007). Ci trus canker is mainly leaf spotting and rind blemishing disease characterized by defoliation, shoot dieback and fruit drop (Polex et al., 2007) and currently managed through eradication and exclusion of infected and exposed trees (Gottwald et al., 2002b). Citrus greening (also called Huanglongbing (HLB)) is a disease caused by several species of Candidatus Liberibacter consisting of phloem limited, uncultured bacteria (Zhao, 1981; da Graca and Korsten, 2004. HLB in Florida, vectored by the Asian psyllid ( Diaphorina citri ) (Zhao, 1981), mostly likely originated in China, where it was given its name because of its characteristic symptom, a yellowing of the new shoots in the green canopy (Polex et al., 2007). There is no cure for the infected trees which de cline and

PAGE 27

27 die within a few months or years (Chung and Brlansky, 2009). The fruit produced by the infected trees is not suitable for the fresh market or juice processing due to significant increase in acidity and bitter taste (Polex et al., 2007; Chung and Brlansky, 2009). HLB bacteria can infect most citrus cultivars, species, and hybrids as well as some citrus relatives (Halbert and Manjunath, 2004). Chronically infected trees display extensive twig and limb dieback, tend to drop fruit prematurely, and are sparsely foliated with small leaves that point upward (Bove, 2006; Polex et al. 2007). HLB infected fruits are frequently small, underdeveloped and misshapen (Polex et al. 2007). Management of HLB disease has proven to be very difficult, as a resul t, there are no cases of a completely successful eradication program to date (da Graca and Korsten, 2004; Halbert and Manjunath, 2004; Chung and Brlansky, 2009). Bove (2006) recommended the elimination of Liberibacteria inoculum by removing infected trees keeping psyllid populations as low as possible through use of contact and systemic insecticides, the use of healthy material for replanting and introduction of biological control predators. Bove (2006) estimated that an orchard with 30% symptomatic tree s, half of the trees are infected and will have to be pulled out sooner or later. Also, surveys conducted over an 8 year period in Reunion Island indicated that 65% of the trees were badly damaged and rendered unproductive within 7 years after planting (A ubert et al., 1996). In Thailand, citrus trees generally decline within 5 8 years after planting due to citrus greening, and yet, groves must be maintained for a minimum of 10 years in order to make a profit (Roistacher, 1996). The use of ACPS is an att empt to help growers optimize production in the face of the impact of canker and greening diseases on tree health and yields. One strategy

PAGE 28

28 being proposed is the use of intensive nutrient management to accelerate tree growth and bring young trees into prod uction so growers can break even within a few years of establishing a grove (Stover et al. 2008; Morgan et al., 2009 b ; Schumann et al. 2009). Planting Densities In Florida, studies on citrus tree densities have been done over the years and show that high planting density produced higher yields (Castle, 198 0 ; Whitney and Wheaton, 1984; Parsons and Wheaton, 2009) and utilized nutrients and irrigation water more efficiently (Parsons and Wheaton, 2009). However, most of the studies done in Florida used much lower densities (80 200 trees per acre) (Obreza, 1993; Obreza and Rouse, 1991; 1993; Alva and Paramasivam, 1998; Paramasivam et al. 2000b) and standard granular fertilization practice or infrequent fertigation at 112 280 kg N ha 1 yr 1 (Paramasivam et a l. 2000b; 2001; 2002) than the 250 trees or more per acre and very frequent fertilization practices proposed for OHS (Stover et al. 2008; Morgan et al. 2009b ; Schumann et al. 2009). Citrus Root Length Density Citrus root length density is a critica l indicator of the potential for water and nutrient uptake. Studies on root water and nutrient uptake are better described with root length density (Morgan et al. 2006 b ; 2007). Several researchers observed that roots of trees grown in the Flatwoods disp lay much stronger lateral than vertical development (Reitz and Long, 1955; Calvert et al., 1977; Bauer et al., 2004). Research on root length density (RLD) distribution has never been conducted on OHS/ACPS. The RLD data discussed in subsequent chapters w ill help define the potential of OHS/ACPS in promoting tree water and nutrient uptake while helping retain nutrients and water in the root zone.

PAGE 29

29 Overview o f t he Dissertation In view of the need for research on citrus irrigation and nutrient management to r educe the impact of g reening infection in Florida, the author presents literature review on the research done on citrus irrigation water and nutrient management, placing emphasis on the novel practices termed the open hydroponic systems (OHS) and advanced citrus production systems (ACPS) in Chapter 2 The review also details methods for nutrient analysis in soil, water and plant tissue samples. The work done in several countries on irrigation design and scheduling using drip and microsprinkler systems inc luding RLD distribution and nutrient uptake efficiencies is discussed. In the final part of the review, the author discusses the use of different models used in agriculture, specifically in citrus production and compares soil water and hydrologic models. The specific model of interest used in the study was HYDRUS 2D and is described and compared with other models used for studying water and solute transport and water uptake. In Chapter 3 aspects of nutrient use efficiency and nutrient distribution in sit u are addressed using data collected over two seasons on an Entisol and a Spodosol. The soil nutrient forms of interest included 2M KCl extractable NH 4 + N and NO 3 N and Mehlich 1 extractable K and P. The plant tissue samples presented relate to N, P and K concentration in above and below ground tissues collected in July 2011 and September 2011. The author compares the effects of irrigation and fertigation practices on citrus tree growth and root length density distribution in Chapter 4. In C hapter 4 t he author presented calibration equations for root length density (RLD) estimated with the intercept and scanning methods for both Ridge and Flatwoods sites for two of the four

PAGE 30

30 replications at each site and validated the equations using data collected from the remaining two replicates. Detailed results on spatial, temporal and vertical root length density distribution for the trees studied are discussed comparing irrigated and non irrigated zones for varying root diameters ranging from <0.5 mm to >3mm. Tr ee growth over time is described using data on trunk cross sectional areas and canopy volumes collected over the 2 years of the study. Chapter 5 shows the results on citrus water uptake estimated using the stem heat balance (SHB) technique for 10 to 21 da y periods over two to three seasons and the soil moisture distribution measured using capacitance probes. Critical measurements included in the SHB technique included leaf area, average hourly and daily transpiration and sapflows. The capacitance probes were calibrated gravimetrically to help estimate volumetric water content and soil moisture stress factor. Results and discussion on the investigation of water uptake and movement and Br movement on a Florida Spodosol and Entisol using HYDRUS 2D are pr esented in Chapter 6 In Chapter 6 the author also describes the sorption parameters for NH 4 + N, K and P on the Flatwoods and Ridge soils using three electrolytes: deionized water, 0.005M CaCl 2 and 0.01M KCl for calibrating HYDRUS 2D for solute transpor t The sorption isotherms were determined for P and fertilizer mixture for NH 4 + N, K and P for 24 h equilibration times using the selected electrolytes. Further a discussion and results on soil water retention characteristics and hydraulic functions for representative soils for soils on the Ridge and Flatwoods are presented. The soil physical characteristics presented are critical in determining sorption behavior of the soils and parameter estimation for computer model simulations The physical characte ristics determined in

PAGE 31

31 the laboratory experiment included 1) bulk density, 2) saturated and unsaturated hydraulic conductivities, 3) residual and saturated moisture contents and 4) soil moisture release curves. The HYDRUS 2D model was calibrated for water and nutrient movement with spring 2011data after sensitivity analysis using soil parameters e.g. residual and saturated moisture water content, bulk density (Obreza (unpublished data ); Carlisle et al. 1989; and Fares et al. 2008), maximum rooting depth (Ma ttos, 2000; Bauer et al. 2004) and water stress index (Simunek and Hopmans, 2009) and validated using the results collected in situ in June 2011 at the Flatwoods site and September 2011 at the Ridge site. The author presents a detailed procedure for sensi tivity analysis and parameter estimation and discusses implications of using HYDRUS 2D as a tool for decision support. The model simulations compared the performance of the conventional practices, microsprinkler OHS and drip OHS irrigation and fertigation scenarios to determine the most effective strategy for water and nutrient management. Outputs of interest from the model included soil water content NH 4 + NO 3 P, K and Br distribution. General Research Goals a nd Hypotheses To address the general resea rch objectives and goals listed above the following specific research goals were conceptualized: Develop optimum irrigation rate, method, and timing for young citrus trees. Determine growth and yield effects of fertigation on young citrus trees at selecte d frequencies. Measure effect of irrigation method and frequency on rooting patterns, nutrient retention, and water and nutrient uptake. Calibrate HYDRUS for water and nutrient movement using site specific soil hydraulic characteristics and nutrient sorpti on behavior.

PAGE 32

32 Characterize HYDRUS as a possible decision support system for predicting soil moisture distribution and solute transport in the vadose zone. The appropriate general hypotheses formulated to answer the above research goals are as follows: Micro sprinkler and drip OHS will increase citrus growth rate, above ground biomass, fruit yield and nutrient uptake resulting in higher plant N, P and K content than the conventional practice (Chapters 3 and 4 ). Spatial nutrient and root length density distrib ution will be significantly greater in irrigated zones of microsprinkler and drip OHS than conventional practice (Chapter 4 ). Citrus water use and K c increase with canopy volume and root length density in situ irrespective of the irrigation frequency and f ertigation method (Chapters 3 and 5 ). Phosphorus adsorption and NH 4 + N and K + exchange on the Flatwoods and Ridge soils do not adversely affect availability and uptake (Chapters 6 ). Measured soil water content, ET NH 4 + NO 3 P, K and Br correlate positiv ely with simulated outputs thus helping in decision support in citrus production systems (Chapters 6 ). Summary The first chapter (Chapter 1) highlights the need for further research in citrus production systems to adapt the ACPS/OHS practices in Florida t hrough use of intensive irrigation water and nutrient management practices to improved tree growth and productivity to increase short term citrus production. More research effort needs to be done to help growers contend with several natural and managerial scenarios outside their control namely 1) citrus canker and greening diseases, 2) uneven monthly rainfall distribution, 3) sandy soil characteristic, and 4) the need for sound environmental nutrient management practices according to USEPA specifications.

PAGE 33

33 Figure 1 1. Typical soil orders of Florida (Source: K.T. Morgan)

PAGE 34

34 CHAPTER 2 LITERATURE REVIEW As defined earlier, OHS is an integrated system of irrigation, nutrition and horticultural practices that was developed in Spain to improve crop production on gravelly soils with low fertility (Martinez Valero and Fernandez, 2004; Yandilla, 2004 ; Falivene et al., 2005) while ACPS is a short to medium term approach to citrus water and nutrient management being evaluated in Florida citrus groves for sustainable, profitable citrus production in the presence of greening and canker diseases with the goal of compressing and enhancing the citrus production cycle so that economic payback can be reached in fewer years to offset some of the disease losses (Schumann et al ., 2009). OHS aims to increase productivity by continuously applying a balanced nutrient mixture through the irrigation system, limiting the root zone by restricting the number of drippers per tree and maintaining the soil moisture near field capacity (Fa livene, 2005). The combination of these practices is claimed to provide a greater control and manipulation of nutrient uptake at specific crop physiological stages and improved water uptake (Yandilla, 2004). OHS has been successfully used in the productio n of peaches, almonds, grapes, citrus, avocados and several vegetable crops in Spain, Australia, South Africa, Chile, Argentina, Morocco and California (USA) (Boland et al. 2000; Kruger et al., 2000a, b; Pijl, 2001; Kuperus et al., 2002; Schoeman, 2002; C arrasco et al. 2003; Martinez Valero and Fernandez, 2004; Falivene et al., 2005; Sluggett et al., unpublished). In South Africa, commercial growers have adapted the OHS through use of drip fertigation on daily basis during daylight hours (Pijl, 2001; Sch oeman, 2002) resulting in increased citrus yield and fruit size (Kruger et al., 2000a, b; Kuperus et al., 2002). OHS was introduced in Australia as an intensive fertigation

PAGE 35

35 practice (IFP) in citrus orchards (Falivene et al., 2005) that uses similar princi ples to OHS but is less intensive. Carrasco et al. (2003) in Chile found that cauliflower and cabbage grown in soil less media with hydroponics resulted in higher growth rate and dry matter yield than those grown in soil with traditional horticultural man agement. They attributed the lower yields of the cabbage and cauliflower grown in the soil media to reduction in water and nutrient uptake compared with transplants that were grown using traditional hydroponics. Jones (1997) described in detail the use o f hydroponics in the USA and early principles applied to this relatively new technology This chapter 1) reviews current open hydroponics system (OHS) management practices utilized in selected citrus producing countries around the world, 2) estimates citr us biomass accumulation and fertilizer demand in citrus 3) describes practices for improved water and fertilizer use efficiency, 4) discusses microsprinkler and drip irrigation system design and scheduling, 5) explains root distribution in response to soi l water and 6) describes various types of process oriented models for solute transport, water and nutrient uptake. The Open Hydroponic System a nd Advanced Production Systems Concepts Maximize water and nutrient efficiency Several studies that have been d one over the years have revealed that it is possible to increase yield, water use and nutrient use efficiency through use of water saving irrigation methods. In a study on water use efficiency and nutrient uptake on micro irrigated citrus, Grieve (1989) f ound that water uptake was limited by water availability rather than root density. Also, fertilizer injection with the micro sprinkler system significantly increased the efficiency of N and P uptake compared with surface

PAGE 36

36 application, whereas leaf K levels were lower under micro irrigation (Grieve, 1989). Multiple applications of N in relatively small amounts with drip irrigation results in lower residual mineral N concentrations and enhances N uptake efficiency by the citrus roots ( Klein and Spieler, 1987; Alva et al. 1998; Paramasivam et al. 2001). Xu et al. (2004) found that P and water uptake were also enhanced in lettuce by high fertigation frequency at low P level. In a three year study, Bryla et al. (2003) found that trees irrigated by surface and subsurface drip produced higher yields and had higher water use efficiency than those irrigated by microjets and furrow irrigation. Drip irrigation systems, in particular, are known to improve irrigation and fertilizer use efficiency because water and nu trients are applied directly to the root zone (Camp, 1998). The benefits of frequent fertigation and/or irrigations in achieving high water and nutrient use efficiency offered by drip irrigation can be negated by improper water placement as shown by the findings of Zekri and Parsons (1988) in grapefruits. Therefore, careful placement of water in the root zone is important in fruit production to ensure that water and nutrient uptake are optimized. Concentrate roots in irrigated zone The use of OHS with dr ip irrigation has the ability to limit root growth within the irrigated zone. Research studies into restricted root zones using physical constraints have shown a reduction in yield in fruit and vegetables (BarYosef et al., 1988; Ismail and Noor, 1996 ; Bol and et al., 2000 ). These studies attributed the yield reduction to reduced canopy growth. Reduced canopy growth or a reduction in yield per tree has not been observed to date in OHS. The wetted soil volume in OHS is considerably greater than the restric ted root zone studies mentioned above where significant reductions in vegetative growth and yield have been reported ( Falivene, 2005) The

PAGE 37

37 study by Boland et al. (2000) on peach in Australia showed a significant reduction in growth and yield when the root zone was restricted to 3% of its potential. In contrast, the wetted soil volume in OHS is approximately 8% to 15% of the potential root volume ( Falivene, 2005) These studies envisage that in an OHS situation the roots are redirected to grow more densel y in a smaller volume of soil, but the soil volume is sufficiently large enough to support active root growth and a productive tree. Reduce nutrient leaching Many researchers have attempted to study nutrient leaching to sustain environmental quality. Pa ramasivam et al. (2001) found that nitrate nitrogen leaching losses below the rooting depth increased with increasing rate of N application (112 to 280 N ha 1 yr 1 ) and the amount of water drained, and accounted for 1 to 16% of applied fertilizer N. Param asivam et al. (2001) noted that the leached nitrate nitrogen at 240 cm remained well below the maximum contaminant limit of 10 mg L 1 They ascribed their observations to careful irrigation management, split fertilizer applications and proper timing of th e application. Thus, it should be possible to reduce nutrient leaching with an OHS and/or IFP because in both scenarios water and nutrients are applied in correct quantities and close to the plant with less waste (Mason, 1990; Jones, 1997) and, at specifi c physiological stages of the plants (Harris, 1971). Paramasivam et al. (2002) used the Leaching and Chemistry Model (LEACHM) to show that 50% of water applied through rainfall and irriga tion drained beyond the root zone. Thus, Entisols require carefully planned and frequent irrigation scheduling during dry periods (Fares and Alva, 1999; Morgan et al. 2006 b ; Obreza and Collins, 2008) due to the inherent low water holding capacity of abou t 0.025 0.070cm 3 cm 3 (Obreza et al.

PAGE 38

38 1997 ; Morgan et al. 2006 b ; Obreza and Collins, 2008), and low organic matter content typically in the range of 0.5 to 1% (Obreza and Collins, 2008). Alfisols and Spodosols are poorly drained due to the presence of restrictive layers below the top and subsoil, respectively called, argillic and spodic horizons that lie at about 30 to 200 cm from the soil surface (Reitz and Long, 1955; Obreza and Admire, 1985; Obreza and Collins, 2008). The Alfisols and Spodosols have higher water holding capacity (particularly the Alfisols, with a water holding capacity ranging from 0.025 to 0.100 cm 3 cm 3 ) and natural fertility (with cation exchange capacity (CEC) ranging from 2 18 cmol(+) kg 1 ) compared with the Entisols (with CEC r anging from 2 4 cmol(+) kg 1 ) due to the presence of a high water table (Obreza and Pitts, 2002) and higher organic matter (typically ranging from 0.5 3%) (Obreza and Collins, 2008). In the Flatwoods, the soils require a combination of bedding and collect or ditches for drainage (Boman, 1994; Obreza and Admire, 1985; Obreza and Pitts, 2002). However, the Alfisols (except Winder soil series that has 85% sand) and Spodosols contain about 94 98% sand in the top 45cm making irrigation water and nutrient manage ment extremely difficult (Obreza and Collins, 2008). Thus, the use of OHS/ACPS needs to consider the unique soil and ecological characteristics for efficient water and nutrient management for high citrus production. The current best management practices ( BMPs) were developed based on low volume micro sprinkler irrigation systems (Lamb et al., 1999; Alva et al., 2003) and conventional fertilizer application practices (Obreza and Rouse, 1993; Alva and Paramasivam, 1998; Thompson and White, 2004). Yet, in cou ntries such as Australia and South Africa, the practices have been adapted through use of intensive and advanced fertigation methods using drip irrigation (Slugget, unpublished; Prinsloo,

PAGE 39

39 2007) to conform to the requirements of OHS. Thus, there is need to modify the current BMPs in the light of intensive fertigation practices that go with OHS in order to effectively sustain high yields in citrus groves and prevent nutrient leaching to groundwater. In Florida, citrus groves are established in the Flatwoods on poorly to very poorly drained Spodosols and Alfisols with a shallow water table (Obreza and Collins, 2008) and on the central Ridge on moderately to excessively well drained Entisols (Reitz and Long, 1955; Obreza and Collins, 2008). Thus, BMPs and nut rient management decisions devised for OHS must take into account these ecologically different zones. Tree density Martinez Valero and Fernandez (2004) provide yield results of some orchards using OHS in Spain in which Nova, Marisol and Delite mandarins were planted at high density (405 trees per acre). Yields in the sixth year were about 65 to 75 tons per hectare which is higher than a conventional orchard using low to medium density plantings (150 to 230 trees per acre) (Falivene et al., 2005). Robinso n et al. (2007) published results of planting densities ranging from 340 to 2178 trees per acre in New York. They found that the optimum economic density was between 1000 1200 trees per acre which is more than double the planting density proposed by Marti nez Valero and Fernandez (2004). The optimum density achieved improved yield and quality coupled with lower costs of production Thus there is a possibility of increasing yield per unit area using ACPS/OHS with densely planted orchards. Tree size control with rootstocks Rootstock selection along with tree planting is a key management element in the ACPS/OHS approach to the future. Citrus trees also require a certain amount of space

PAGE 40

40 to develop and flourish. When the allocated space is fixed, e.g. 1 acre of land, tree size becomes critical because the productive unit is the canopy and only a certain volume of canopy can be grown on 1 acre. Vigorous, large trees are neither compatible with close spacing nor productive in their younger years. Thus, in a wo rld of economic necessity dictated by early and robust returns, small, closely spaced trees become a required component of the new production concepts. Groves of closely spaced trees on vigorous to size controlling rootstocks have been extensively researc hed, but have had no commercial implementation in Florida (Morgan et al. 2009 b ; Schumann et al. 2009). From the research, it is apparent that proper matching of tree size with spacing and site conditions is critical for success. When that combination i s achieved, the higher density grove will outperform the more conventional one especially in the early years. In Florida, the conventional grove is spaced about 15 x 25 ft (116 trees/acre), the modern grove is at 10 x 20 ft (218 trees per acre), and the h igher density grove would be about 8 x 15 ft (363 trees per acre) (Morgan et al. 2009b ). Fertilizer Demand and Nutrient Uptake in Citrus Biomass Development with Time Tree growth and development change with time due to variable distribution of dry matter in both above and below ground tree components due to the growth of larger branches and trunks of older trees to support increased tree biomass (Richards, 1992; Morgan et al. 2006b). Mattos et al. (2003a b), studying the six year old Hamlin orange tre e [ Citrus sinensis (L.) Osb.] on Swingle citrumelo rootstock [ Poncirus trifoliata (L.) Raf. x Citrus paradise Macfad.], showed the following proportions of biomass distribution: fruit=30%, leaf=10%, twig=26%, trunk=6%, and root=28%. The biomass distributi on in other citrus cultivars is described in Table 2 1 ( Cameron and Appleman,

PAGE 41

41 1935; Cameron and Compton, 1945; Feigenbaum et al., 1987; Quiones et al. 2003 a ; 2005; Morgan et al. 2006 a ). Morgan et al. (2006 a ), for example, showed that the percent biomas s distribution in 14 year old Hamlin oranges in Florida on Carrizo and Swingle rootstocks, grown on Candler fine sand ranged from 12 13% in leaves, 52 61% in branches, twigs and the trunk, and 27 33% in the roots. In another study in Israel on 20 year old Shamouti oranges, percent biomass distribution ranged from 6 7% in leaves, 55 56% in branches, twigs and trunk, 8 13% in fruits and 24 31% in roots (Feigenbaum et al. 1987). Quiones et al. (2003 a ; 2005), studying eight year old Navelina orange trees in Spain under flood and drip irrigation systems on sandy loamy soil, found similar biomass distribution pattern in roots but noted higher biomass in leaves (13 16%) and fruits (21 27%), and lower biomass in branches (29 34%) compared with values reported by Feigenbaum et al. (1987). Earlier work on biomass distribution in 3.5 and 10 year old Valencia oranges was done in California (Cameron and Appleman, 1935; Cameron and Compton, 1945). In these early studies on biomass distribution on 3.5 year old Valenci a oranges, 31% of the biomass was found in both roots and leaves while the remaining biomass was accounted for in bark and woody tissues such as branches and the trunk (Cameron and Appleman, 1935). Contrasting results were noted on bearing 10 and 15 year old Valencia oranges where percent biomass distribution was approximately 18% in leaves, 61% in trunk and branches while 21% of the biomass was allocated to the below ground portion (Cameron and Appleman, 1935; Cameron and Compton, 1945). The accumulation of dry matter (DM) by various components of developing tamarillo ( Cyphomandra betacea ) was investigated by Clark and Richardson (2002).

PAGE 42

42 They found that percent DM accumulation in years 2 and 3 were 21 and 22% in roots, 37 and 33% in the stem, 23 and 15% in branches, 8% in leaves (both years), and, 12 and 13% in fruits. Richards (1992) studied the Cashew ( Anacardium occidentale ) tree nutrition as related to biomass accumulation, nutrient composition and nutrient cycling in sandy soils of Australia at 0 12 12 40 and 40 70 months after planting. He observed that the tops accounted for 75% of dry weight, with roots <20%, except at 12 months. Cashew apple and nuts account for <10% of tree total DM. Barnette et al. (1931) studied the biomass and mineral dist ribution of a 19 year old Marsh seedless grapefruit tree in Florida. They found that out of 273 kg dry weight per tree, the biomass distribution was as follows: fruits= 3%, leaves = 6% roots=34% and, trunk and branches=57%. Nutrient Requirements for Biomas s and Fruit Production Nutrient application rates for the majority of OHS and intensive fertigation practice (IFP) in citrus can be about 20% to 50% higher than conventional practices (Falivene et al., 2005). OHS and IFP use a more intensive nutrition pro gram with the goal of pushing trees into a higher level of vigor and productivity requiring higher nutrient application rates to maintain production. However, studies on fertilization practices on citrus in Florida have shown mixed results. Previous studi es on citrus nutrient management have shown that proper nutrient placement and timing (Koo, 1980; Koo et al. 1984 a ; Obreza and Rouse, 1993; Obreza et al. 1999; Kusakabe et al. 2006; Obreza and Tucker, 2006), applicatio n rate and frequency (Koo, 1980 ; Tu cker et al. 1995; Lamb et al. 1999; Paramasivam et al. 2000b; Mattos et al. 2003 a, c ; Tucker et al., 2006; ) and fertilizer application method (Alva et al. 2003; 2006a b) can substantially affect nutrient uptake, yield, yield quality and environmental quality in citrus. Obreza

PAGE 43

43 and Rouse (1993) showed that an increase in fertilizer rate resulted in a decrease in total soluble solids concentration and total soluble solids to acid ratio. Also, Koo and Smajstra (1984) made similar observations using trick le irrigation and fertigation on 26 (1980), in trials on sandy soil, found no significant differences due to fertigation frequencies (3 or 10 times a year) on 13 and Jifon (2001) studied fertigation in 6 and 80 times per year and found that fertigation frequency did not affect leaf nutrient concentration, canopy size, fruit yield or juice quality. Schumann et al. (2003) compared fertilizer application rates and methods for Hamlin oranges on Candler fine sand in central Florida. In the study, Schuman n and co workers showed that fertigation (applied 15 times) was superior to dry gran ular fertilization (applied in four splits) and control release fertilizer (applied once every fall) where optimal soluble solids concentration was obtained at 145, 180 and 190 kg N ha 1 and optimal fruit yields was realized at 138, 160 and 180 kg N ha 1 f or fertigation, dry granular fertilization and control release fertilizer, respectively. Fertigation resulted in 22 45 kg N ha 1 savings per year with leaf concentrations significantly higher per unit of N applied for fertigation>dry granular fertilizer>c ontrol release fertilizer, confirming the efficiency of fertigation practice with respect to optimal nutrient placement in the root zone and temporal distribution over the season. Morgan et al. (200 9 a ) studied the effect of fertigation (4 or 30 times ann ually), dry granular fertilization (applied in four splits) and control release fertilizer (applied once in February) on 1 5 yr

PAGE 44

44 method data showed that critical values of minimum N rates required to reach canopy volume plateau were 182, 198 and 199 kg ha 1 for fertigation (30 times annually), control release fertilizer and fertigation (4 times annually), respectively, representing canopy volumes of 8.4, 7.6, and 7.9 m 3 The more frequent fertigat ion practice produced larger trees with lower annual rates compared with both control release fertilizer and fertigation (4 times annually). Morgan and colleagues also noted reduced yield and tree size at higher dry granular fertilization rates suggesting improved nutrient use efficiency of trees fertilized by control release fertilizer and fertigation (30 times annually). Root injury observed under dry granular fertilization was ascribed to salt burn from excessive fertilizer distributed over a small are a. For maturing trees (6 10 years), Morgan et al. (200 9 a ) observed that citrus root systems were equally effective in capturing available N from frequent small fertilizer application (fertigation 30 times annually) or from 4 much larger applications. The y concluded that more frequent applications should result in increased fertilizer use efficiency and likely promote tree growth, albeit, little increase in fruit yield may be obtained in mature citrus. Tucker et al. (1995) and Alva et al. (2006c) recommend ed K rate for optimal production of bearing citrus trees ( > 4 years) in the range of 112 186 kg ha 1 for orange trees and 112 150 kg ha 1 for grapefruit trees. Alva et al. (2006c) observed that there are no consistent research results to make definitive co nclusions on potential differences between the dry granular, controlled release, or fertigation methods of K. They also described K concentration in 4 to 6 month old non fruiting citrus trees in the range of 12 17 g kg 1 as optimal for Florida citrus. A corollary method in some citrus producing parts of the world like South Africa and Brazil, nutritional status of the tree is

PAGE 45

45 determined using leaf analysis of fruiting terminals with optimal K status ranging 10 15 g kg 1 Tucker et al. (1995) recommended a minimum of 3 split applications for dry granular fertilizer and 10 times for fertigation practices. Tucker et al. (1995) recommended 120 240 g K tree 1 yr 1 (year 1), 240 480 (year 2) and 370 740 (year 3) for non bearing citrus trees. Criteria for sel ecting a rate within the recommended range include history of fertilization in the tree nursery, soil type, land history, and fertilizer placement. In other regions such as Arizona, Kusakabe et al. (2006) evaluated the response of 3 to 6 yr old microspr inkler and fertigation frequencies on coarse sand. In the study, Kusakabe and colleagues concluded that the maximum fruit yield of the trees occurred at N rates of 113 g N tree 1 for the fourth, 105 g N tree 1 yr 1 for the fifth, and 153 g N tree 1 yr 1 for the sixth growing season under the maximum fertigation rates (27 fertigations). The effect of timing of fertilizer application and irrigation system on nutrient use efficiency was investigated i n Spain (Quiones et al., 2005). Quiones et al. (2005) concluded that drip irrigation together with extensive splitting up of the N dosage may be the appropriate system for the N fertilization management in citrus as it offers greater fertilizer use effi ciency, smaller accumulations of residual nitrates in the soil, and 15% reduction in the amount of water applied, without impairing fruit yield and its commercial quality. Tucker et al. (1995) suggest P reduction or omission in fertilizer if soil test res ults indicate sufficient residual P. They observed that fertilizer applications in a number of doses generally increase nutrient uptake efficacy by providing available nutrients within the root zone over prolonged growing period and by reducing leaching t hat occurs due

PAGE 46

46 to excess rainfall and/or irrigation. Dry granular fertilizer may be applied in 4 6 doses during annual growing period, while liquid fertilizer could be split in 10 30 applications. Control release fertilizer can be applied at a reduced fr equency as preplant treatment, incorporated after planting, or broadcast to insure uniform distribution of nutrients throughout the enlarging root zone of young trees (Tucker et al., 1995). However, different scion and rootstock combinations respond di fferently to fertilization. For example, Mattos (2000) and Mattos et al. (2003a b) showed that e, response of Cox et al. (2001) demonstrated that same age of grapefruit trees on V olkamer lemon were larger than trees on sour orange rootstock and dry weight distribution of tree parts was affected by N fertilization and soil condition. From various studies, N, P and K nutrient distribution is mainly concentrated in the leaves or fruit s and roots (Cameron and Appleman, 1935; Cameron and Compton, 1945; Legaz et al. 1982 ; Dasberg, 1987; Feigenbaum et al. 1987; Legaz et al., 1995; Mattos et al. 2003a b; Quiones et al. 2003 a, b ; 2005; Morgan et al. 2006 a ) (Table 2 2). Earlier work o f Alva and Paramasivam (1998) and Paramasivam et al. (2000 c ) also showed predominance of N, P and K in fruits and leaves. In fruits, Alva and Paramasivam (1998) reported nutrient ranges for N (0.08 1.22%), P (0.14 0.15%) and K (1.17 1.23%) for four citrus varieties namely: Hamlin, Parson Brown, Valencia, and Sunburst. Also Paramasivam and colleagues (2000 c ) showed leaf concentrations of N

PAGE 47

47 (27.4 29.3 g kg 1 ), P (1.3 1.4g kg 1 ) and K (8.5 15.1 g kg 1 ) for the same varieties presented by Alva and Paramasivam (1998). Recent research showed nutrient concentrations in leaves and fruits of >4 year old Tahiti acid lime in Brazil (Mattos et al. 2010). Leaf N ranged from 14.7 23.1 g kg 1 while K varied from 11.2 17.1 g kg 1 Mattos and colleagues found N and K v alues in the range of 7.5 14.5 g kg 1 and 12.0 17.2 g kg 1 in fruits. Legaz et al. (1995) studied the mobilization of N from reserve organs (leaves, roots, branches and trunk) to developing organs at different moments of the growing cycle in three year ol d Valencia Late orange trees on siliceous sand in Spain. Legaz and colleagues found highest amounts of N in leaves and roots (33 42% and 30 38% ) respectively Alva et al. (2003) proposed a combined use of foliar fertilizer application and fertigation as the best management practice (BMP) for N because these were effective in reducing nitrate leaching to surficial groundwater. Nevertheless, the practices in the studies above are less intensive than a typical OHS in which 3 or more irrigations per day can be achieved (Falivene et al., 2005). More recently, novel, intensive fertigation methods termed Advanced Citrus sandy soil soils (Stover et al. 2008; Morgan et al. 2009 b ; Roka et al. 2009; Schumann et al. 2009). Preliminary results by Schumann et al. (2009) showed the benefits of ACPS on <1 yr old Hamlin oranges on swingle and C 35 rootstock grown on a Candler fine sand. Leaf nutrient concentrations for leaves sample d in 2009 had high, non limiting N concentrations (>3%). Additionally, N fertilizer applications were lower per tree relative to the benchmark N fertilizer applied, lower for drip (13%) and microsprinkler

PAGE 48

48 (20%) fertigation treatments than conventional grow er practice (100%). They concluded that the high nutrient and water use efficiencies possible with an ACPS in young planted citrus could improve overall profitability by reducing production costs and sustaining environmental quality. However, Schumann et al. (2009) note d that the possible limitation to successful implementation of ACPS/OHS in Florida include a unique combination of sandy soils and the distribution of more than half the high annual rainfall in the summer months, consequently, resulting in root growth in the nonirrigated zone. Nutrient Uptake and Nutrient Use Efficiency Citrus Nutrient Management In a study on Best Management Practices (BMPs) for N and P, Thompson and White (2004) noted that adequate supplies of N are necessary to optimize yields of young citrus trees. They reported higher nutrient use efficiency with micro irrigated citrus resulting in leaf N above the critical concentration of 2.5% when using surface irrigation. Thompson and White (2004) called for optimal levels of N and irrigation for trees to N P K fertilizer rates under field conditions in southwestern Florida, Obreza and Rouse (1993) found that an increase in fertilizer rate resulted i n a decrease in total soluble solids (TSS) concentration in juice and the TSS : acid ratio, but weight per fruit and TSS per tree increased. Several citrus fertilization experiments from other parts of the world indicate that an annual application of abou t 200 kg N ha 1 is sufficient to sustain optimal tree growth, and maintain high production (Dasberg, 1987). One of the options for improved citrus growth and yield is improved management of water and nutrient systems. Maximization of nutrient uptake effic iency and minimization of nutrient losses is a function of the rate, placement and timing of nutrient application ( Saka,

PAGE 49

49 1984; Alva and Paramasivam, 1998; Quiones et al., 2007). Zekri and Obreza (2003) observed that fertilization represents a relatively small percentage of the total costs of citrus production, but it has a large effect on potential profitability. Analyses of leaves and soil can be used to evaluate nutritional status of trees and nutrient availability in the soil to supply the trees nutrie nt requirement (Embleton et al., 1956; Alva and Paramasivam, 1998; Obreza et al., 1999). N is the key component in mineral fertilizers applied to citrus groves and has more influence on tree growth and appearance than any other element. N affects the absor ption and distribution of all essential nutrients (Zekri and Obreza, 2003). Quiones et al. (2003 a ) found that N uptake efficiency of the whole citrus tree was higher with drip irrigation (75%) than with flooding system (64%) showing that drip irrigation system was more efficient for improving water use and N uptake from fertilizer. This suggests that optimum nutrient management must take into account baseline information on the initial or residual soil nutrient composition of key elements such as N, P an d K. For citrus, K is important to yield, fruit size, and juice quality (Obreza and Morgan, 2008) such that its deficiency reduces fruit number, increases fruit creasing, plugging and drop and decreases juice soluble solids, acids and vitamin C content. Extraction Methods for N Forms, P and K from Soils and Plant Tissue Soil analysis is useful in formulating and improving a fertilization program over several consecutive years so that trends can be observed. Soil testing is particularly useful for P (a s shown in Table 2 3 and has no practical value for readily leached like N and K (Obreza et al., 2008a) because in many humid regions where annual precipitation exceeds evapotranspiration, leaching and denitrification reduce profile NO 3 N and K to levels often unreliable in fertilizer recommendation (Havlin et al., 2005; Obreza et al.

PAGE 50

50 2008 b ). Most recommendations call for soil testing about 3 years, with more frequent testing on sandy soils to determine whether the nutrient management program is adequate for optimum productivity. For instance, if soil test P is decreasing P application rate can be increased. If soil test P has risen to satisfactory level, application may be reduced to maintenance rates (Havlin et al., 2005). Havlin et al. (2005) reco mmends the use of Bray 1 and 2 P and Mehlich 3 P extraction on acid and neutral pH soils. A Mehlich 1 soil test is useful in regions with more highly weathered, low cation exchange capacity (CEC) soils. The Olsen P soil test is used in neutral and calcar eous soils (Havlin et al., 2005). Bray 1 and Mehlich P tests extract similar quantities of P while the Olsen P test extract about half as much P. The quantity of P dissolved by the extractants is calibrated with crop response. Sato et al. (2009 c ) collec ted soils from southwest Florida and compared available P levels by five different soil testing methods (Mehlich 1, Mehlich 3, Olsen, Bray 1, and ammonium bicarbonate DTPA). They observed that within a surface soil pH range of 6.4 and 8.6, correlation coef ficients between available P by Mehlich 1 and those by other 4 methods ranged from 0.61 and 0.73 (p < 0.001). Compared with Mehlich 1 method, all other 4 methods extracted less amounts of available P (59%, 22%, 51%, and 25% with Mehlich 3, Olsen, Bray 1, and AB DTPA, respectively). Alva (1993) compared methods for extraction of nutrient elements including P and K from the soil. He found that K extractable by Mehlich 3 was significantly correlated to extractions by either Mehlich 1 (r 2 =0.95), ammonium ace tate (AA) (r 2 =0.95), ammonium chloride (r 2 =0.97) or ammonium bicarbonate DTPA (AB DTPA) (r 2 =0.96) extractants. In the study, Mehlich 3 P significantly correlated with Mehlich 1 only (r 2 =0.65). Extractable

PAGE 51

51 P correlation between Mehlich 3 versus AB DTPA wa s weak (r 2 =0.18), non significant for Mehlich 3 vs AA and Mehlich 3 vs ammonium chloride. This was corroborated by earlier findings by Sartain (1978) who suggested that Mehlich 1 extractant solubilizes some of the calcium phosphate compounds which are not solubilized by ammonium acetate. The work by Elrashidi et al. (2001) also showed that Mehlich 3, Bray 1, or Mehlich 1 (double acid) were a good test for P concentration in water and soil. Bar Yosef and Akirir (1978 ) found that NaHCO 3 extraction is capable of providing simultaneously availability indices for NO 3 N, P, and K. The caveat for K with this extraction method is that it applies only when exchangeable K in the soil is greater than a given fraction of CEC of the soil. In a comparison of mechanical vacuum extraction with batch extraction method for estimation of CEC in soils, Huntington et al. (1990) found that the precision of the two methods was equivalent. The two extraction methods can be used for CEC estimation with consistently similar results Obreza et al. (2008 b ) explained the value of leaf tissue and soil analysis in determining fertilizer programs that increase fertilizer efficiency while maintaining maximum yields and desirable fruit quality in citrus. Leaf tissue analysis is used for qu antitative determination of the total mineral nutrient concentrations in the leaf. It is very useful in testing for N, P and K sufficiency. Guidelines for interpretation of tree leaf analysis are described by Koo et al. (1984 b ), Obreza et al. (1999) and O breza and Morgan (2008) in Table 2 4. Anderson and Henderson (1986; 1988) compared four methods for elemental analysis of plant tissues. The methods included sealed chamber digestion method, dry ash combustion, nitric/perchloric acid wet ash digestion, an d sulfuric acid/hydrogen

PAGE 52

52 peroxide wet ash digestion. They recommended the use of the former three methods whose use is dependent upon the preference of the user and availability of equipment. Sulfuric acid/hydrogen peroxide wet ash digestion appeared to give the poorest overall chemical analyses. Plank (1992) also indicated that nutrient content in the digests could be determined by Inductively Coupled Plasma (ICP). The plant nutrient uptake values (expressed as kg ha 1 ) could be obtained as the product of concentration (mg kg 1 plant) and dry matter yields (kg plant ha 1 ). Irrigation Design and Scheduling Drip and Microsprinkler Irrigation Evapotranspiration Calculations Citrus evapotranspiration (ET), like for any particular crop, is limited by atmosph eric demand, crop development stage, and available soil water content (Morgan et al. 2006 b ; Fares et al. 2008). It is estimated from daily reference evapotranspiration (ET o ) using the following equation: ET c =ET o K c K s ( 2 1 ) Where ET c is crop e vapotranspiration (mm d 1 ); ET o is potential evapotranspiration (mm d 1 ); K c is the crop coefficient and K s is the soil water depletion coefficient, which is also called the water stress function (Allen et al. 1998; Obreza and Pitts, 2002; Morgan et al. 2006 b ; Fares et al. 2008). The crop coefficient is defined as the ratio of ET c to ET o when soil water availability is nonlimiting, and thus, is proportional to atmospheric demand and plant development stage (Morgan et al. 2006 b ; Fares et al. 2008). Ac curate estimation of citrus ET is important in determining irrigation requirement (IRR, mm) calculations. Irrigation requirements for a particular crop are calculated as follows:

PAGE 53

53 IRR =ET c + S (UF+ER) ( 2 2 ) Where IRR (mm) is the irrigation requir ement, ER (mm) is effective rainfall, S (mm) is change in root zone soil water storage and UF (mm) is upward flux from the water table (if present) due to capillary rise. In the deep, well drained sandy soils of central Florida, UF is negligible (Fares et al. 2008) but is a critical factor in the poorly drained Flatwoods of southwest Florida (Obreza and Pitts, 2002). FC ) exists where water up take is not limited by soil water t ) is referred to as readily available water (RAW), and used it to estimate K s as the ratio of remaining available soil water to soil water that is not readi ly available (Allen et al., 1998; Morgan et al. 2006 b ): ( 2 3 ) where K s is soil water depletion coefficient (K s < FC WP is total available water (TAW) (cm 3 cm 3 WP is permanent wilting point soil wat er content (cm 3 cm 3 is soil water content (cm 3 cm 3 FC is field capacity soil water content (cm 3 cm 3 FC t is readily available water (RAW) (cm 3 cm 3 ) (Allen et al., 1998; Morgan et al. 2006a). Daily ET c of young citrus trees measured during the 1996 and 1997 cropping seasons were from 1.9 to 2.0 mm (Fares and Alva, 1999) and from 1.87 to 3.13 mm (Fares and Alva, 2000), respectively. For mature citrus, daily ET c ranged from 2.25 to

PAGE 54

54 3.52 mm (Rogers et al., 1983). However, reference ET c for m ature citrus was found to vary from 1.4 mm day 1 in December to 4.9 mm day 1 in May (Morgan et al., 2006 b ; Fares et al., 2008). Based on the studies conducted over the years in Florida, ET appears to be low from November to March and peaks from April to Oc tober. Citrus Crop Coefficients K c is defined as the ratio of crop evapotranspiration (ET c ) to potential evapotranspiration (ET o ) when soil water availability is non limiting and is a function of crop type, climate, soil evaporation and crop growth stage (Allen et al., 1998; Morgan et al., 2006 b ; Fares et al., 2008). Several studies estimated that K c values of citrus trees range from 0.6 in the fall and winter to 1.2 in the summer (Boman, 1994; Martin et al., 1997 ; Fares and Alva, 1999 ; Morgan et al., 20 06 b ). Jia et al. (2007) found that K c values may vary from location to location. For example, they found that annual average K c values were higher for the citrus grown in the Ridge regions (K c =0.88) than for the Flatwoods (K c = 0.72) in Florida, with mon thly recommended values ranging from 0.70 to 1.05 for the ridge and from 0.65 to 0.85 for the Flatwoods citrus, respectively. They attributed the differences to water logging in the root zone of the Flatwoods citrus owing to water table due to the presence of the spodic and/or argillic horizon. Water Use Efficiency Michelakis et al. (1993), studying avocado water use in a Mediterranean climate in Greece under drip irrigation, found that root percentage was generally higher in the upper 50cm soil layers a nd within 2 m from the drip line, where about 70 72% of the roots were located. They attributed the higher root percentage in the upper soil layers to biological factors and to the higher oxygen diffusion rate. In the study Michelakis et al. (1993) applie d irrigation water to each treatment using one drip lateral per row of trees

PAGE 55

55 with drippers of 4 l h 1 discharge rate placed 70cm apart. Coleman (2007) also observed that root length density in cottonwood, American sycamore, sweetgum and loblolly pine was dependent upon depth and position relative to drip emitter when fertilizers were applied and is greatest at the surface and in proximity to the drip line. The factors controlling root length density in the woody species studied included age, depth and pro ximity to the drip emitter. Partial soil wetting under drip irrigation generally leads to many agronomic benefits such as water and labor saving (Keller and Karmeli, 1974). However, the extent of the wetted soil volume is a function of the emitter discha rge and spacing but depends mainly on the soil type and the total water added (Warrick, 1986). High water use efficiency and water savings using high frequency drip and microsprinkler irrigation systems have also been reported in recent studies in Spain ( Quiones et al. 2003 ; 2005), California, USA (Bryla et al. 2003 ; 2005), Florida, USA (Zotarelli et al. 2008 a, b ; 2009 a, b ; Kiggundu et al. 201 1 ), Malawi (Fandika et al. 201 2 ) and Australia (Phogat et al. 2011). The principles underlying the restricti on of the roots to the wetted zone using drip irrigation are also applicable to OHS. Bromide as a Tracer for Water Movement in the Soil Bromide is one of the conservative anions generally applied to soils to trace water and solute movement in the soil. K hne and Gerke (2005) studied preferential Br movement in the soil. They found that Br was transported during physical equilibrium conditions, except for conditions of heavy rainfall that triggered preferential flow involving physical non equilibrium. A fyuni and Wagger ( 2006 ) also conducted an experiment on Br movement as a function of soil physical properties. They found that preferential flow via macropores appears to play a significant role in Br movement. Afyuni and Wagger ( 2006 ) postulated that under similar soil and environmental

PAGE 56

56 conditions, movement of mobile nonreactive anions such as NO 3 will occur if applied in concentrations exceeding those taken up by plants. Irrigation Methods Proper irrigation system design is important in advanced citr us production systems (ACPS) such as OHS and IFP to ensure that the system does not leak and/or fail at some point. There are two main types of irrigation scheduling programs in OHS: pulsing irrigation and continuous (Falivene et al., 2005). Pulsing irri gation management program involves short pulses of irrigation provided to the trees throughout the day while as continuous irrigation management program uses low output rates to match water use conditions in summer. The number and timings of pulses are ba sed on a calculation of readily available water (RAW) and average tree water use along with monitoring of irrigation scheduling devices like tensiometers, capacitance probes and trunk diameter measuring devices. In a restricted root zone situation up to ni ne or more pulses of irrigation could be scheduled throughout the day in summer (Falivene et al., 2005). Partial soil wetting under drip irrigation generally leads to many agronomic benefits such as water and labor saving (Keller and Karmeli, 1974). How ever, the extent of the wetted soil volume is a function of the emitter discharge and spacing but depends mainly on the soil type and the total water added (Warrick, 1986). I ncreasing the irrigation rate enhanced NO 3 N movement to deep layers under wheat (Charanjeet and Das, 1985; Recous et al. 1992). Quiones et al. (2007) reported similar observations in citrus. Several researchers have recommended the use of frequent fertigation combined with improved irrigation scheduling to improve fertilizer upta ke efficiency, to increase residence time of nutrients in the root zone and to reduce the potential for

PAGE 57

57 groundwater pollution (Graser and Allen, 1987; Ferguson et al., 1988; Obreza et al., 1999; Alva et al., 2003; Schumann et al., 2003). Also Bryla et al. (2003 ; 2005) showed that surface and subsurface drip scheduled daily increased fruit size and improved marketable yields of peach and reduced the number of nonmarketable fruit by 9% to 22% over more traditional furrow or microspray irrigation methods. Irr igation Control Methods Smajstrla et al. (2009) described the main components required in irrigation scheduling such as estimating evapotranspiration (ET), soil water storage capacity, and allowable water depletions. They recommended two irrigation schedu ling methods for Florida soils and climate 1) a water budget method requiring estimation of daily ET and soil water content, and 2) the use of soil moisture measurement instrumentation. Following the water budget principles, Morgan et al. (2009b) develope d an ET based scheduling tool for Florida that factors in soil characteristics and rooting depth for determining when to irrigate and how much water to apply. Researchers in Florida have also proposed methods of determining when to irrigate and how much w ater to apply using soil moisture measuring devices in the sandy soils (Alva and Fares, 1998 ; Migliaccio and Li, 2009 ; Munoz Carpena, 2009). Advances in the irrigation scheduling methods using microsprinklers can be adjusted to match the intensive irrigat ion practices used in OHS using drip irrigation. Citrus Root Density Distribution In Florida, citrus groves are established in both Flatwoods and Ridge regions. The Flatwoods soils are in the southern and coastal areas of the state, whereas the Ridge soils are in the northern and central citrus production areas of the state (Jia et al., 2007). Flatwoods are found in a flat landscape with low elevation where surface water

PAGE 58

58 drainage is slow. In these areas, citrus is normally grown on raised 2 row beds, and d rainage runs in ditches between beds (Boman, 1994). The Flatwoods Alfisols and Spodosols that support citrus are poorly to very poorly drained (Obreza and Collins, 2008). In contrast, Ridge citrus grows in a landscape of low hills, in which individual pl ots may be level. The Ridge Entisols are fine to coarse sands (Parsons and Morgan, 2004) that are moderately to excessively well drained (Reitz and Long, 1955; Obreza and Collins, 2008). Also, on the Ridge, mature citrus have at least half their roots in the top 90 cm (Reitz and Long, 1955; Fares and Alva, 2000 a, b ; Parsons and Morgan, 2004), while in the Flatwoods, over 95% of the roots are in the top 30 to 45 cm (Parsons and Morgan, 2004). Flatwoods citrus roots may be limited to the top 30 to 45 cm bec ause of the high water table and the presence of argillic or spodic horizons ( Obreza and Admire, 1985 ; Boman, 1994 ). For young citrus trees, most roots are in the top 30 to 60 cm (Parsons and Morgan, 2004). Kalmar and Laha v ( 1977 ) irrigated avocados with sprinklers at 7, 14, 21 and 28 day intervals and found that most water was absorbed from upper 60 cm soil layer suggesting that this was where most roots were concentrated. In a study on citrus water uptake dynamics on a sandy Florida Entisol, Morgan et al. (2006 b ) reported that roots were concentrated in the top 15 cm of soil under the tree canopy (0.71 to 1.16 cm roots cm 3 soil), where maximum soil water uptake was about 1.3 mm 3 mm 1 root 1 day 1 at field capacity, decreasing quadratically as moisture content decreased. Michelakis et al. (1993), studying avocado water use in a Mediterranean climate in Greece under drip irrigation, found that root percentage was generally higher in the upper 50 cm soil layers and within 2 m from the drip line, where abo ut 70 72% of the roots were located. They attributed the higher root percentage in

PAGE 59

59 the upper soil layers to biological factors and to the higher oxygen diffusion rate. Coleman (2007) also observed that root length density was dependent upon depth and pos ition relative to drip emitter when fertilizers were applied and is greatest at the surface and in proximity to the drip line. The factors controlling root length density included age, and depth and proximity to the drip emitter. Process Oriented Models f or Solute Transport, Water and Nutrient Uptake Types and Use of Models in Agriculture Several simulation models for predicting water and nutrient uptake and movement have been developed in recent years in recognition of the need to develop solutions for va rious agricultural and environmental management pro blems such as irrigation scheduling, design of drainage systems, crop management and pollution of 1999; Jones et al. 2003). The models may have some deficiencies in representing th e soil water plant atmosphere interaction and processes (Clemente et al., 1994) owing to the biases of their developers (Hutson, 2005) and simplifications associated with input data and variability in field data ( Hornsby et al., 1990; De Jong et al., 1992; Clemente et al., 1994). Nevertheless, the models help us examine and gain an understanding of the processes that cannot be subjected to experimentation. Most models have been developed in the past 20 years to help offer decision support in different cro pping systems (Jones et al. 2003), hydrologic systems (Hutson 1999 ; 2007) and soil water management (Ahuja et al. 1993). The decision support system for agrotechnology transfer (DSSAT) model simulates growth, developmen t and yield of a crop growing on a uniform area of land under prescribed or simulated management as well as the changes in soil water,

PAGE 60

60 carbon, and nitrogen that take place under the cropping system over time (Jones et al. 2003). The ARS Root Zone Water Qua lity Model (RZWQM) is used for predicting pesticides reactions and degradation, nutrient transformations, plant growth, and management practice effects (Decoursey and Rojas, 1990; Ahuja et al. 1993) 2D and 3D models to simulate the two and three dimensional movement of water, heat, and multiple solutes in variably variably s aturated water flow and convection dispersion equations for heat and solute transport. The flow equation incorporates a sink term to account for water uptake by ). Soil hydraulic parame ters of this model can be represented analytically using different hydraulic models such as the van Genuchten (1980) and Brooks and Corey (1964) equations. Several researchers have used HYDRUS model in irrigated systems (Fares et al. 2001; Grnens et al. 2005; Boivin et al., 2006; Fernndez Glvez and Simmonds, 2006; Hanson et al., 2006; Zhou et al. 201 1 ; Bufon et al. 2011; Phogat et al. 2011 ). Fares et al. (2001) simulated solute movement within the soil profile of Ridge and Flatwood soil types using the HYDRUS 2D model. In the simulations, they f ound that 25 % more water drained under the Flatwoods soil than the Ridge soil. Also, solute leaching was 2.5 greater under the Ridge soil type than Flatwood soil type. The results obtained by Fares et al. (2001), however, require further investigation an d validation by statistically correlating the measured outputs in situ versus the simulated outputs. Despite problems associated with identification of the actual physical processes when conducting simulation, Pang et

PAGE 61

61 al. (2000) found that HYDRUS model ac curately described soil water contents with minor discrepancies. Studies by Grnens et al. (2005) and Hanson et al. (2006) assessed fertigation strategies using HYDRUS 2D for nitrogen fertilizers. They found that HYDRUS 2D model described the movement o f urea, ammonium, and nitrate during irrigation and accounted for the reactions of hydrolysis, nitrification and ammonium adsorption. The HYDRUS 2D model was used in the study because it is appropriate for use in microsprinkler and drip fertigated syste ms Comparing Soil Water and Hydrologic Models Skaggs (1980) developed the DRAINMOD water management /drainage model for use in areas with high water tables. Using this model, Obreza and Boman (undated) simulated water table fluctuation, upward flux and c itrus ET on 12 citrus groves in the Flatwoods soils. They observed that a water table depth of 50 70 cm was sufficient to maintain a root zone soil moisture that did not limit citrus ET in the Flatwoods. Clemente et al. (1994) compared three models: SWATR E (Soil Water and Actual Transpiration, Extended), LEACHW and SWASIM (Soil Water Simulation Model). They concluded that model predictions and measured water content profiles were within the limits of acceptance and none of the models consistently outperfo rmed the others. They recommended the use of any of these models for prediction of water content in unsaturated soils. Several researchers have attempted to use LEACHM to simulate nutrient and water uptake and movement in various conditions. Jabro et a l. (1993) found that LEACHM (version 3.0) overestimates leached NO 3 due to its inability to estimate macropore flow effects. They also deemed the use of the water retention function fitted

PAGE 62

62 because it tends to overestimate soil water content. However, Jabro et al. (1993) concluded that NO 3 leaching was better simulated by LEACHM than by NCSWAP (Nitrogen, Carbon, Soil, Water and Plant). A study by Paramasivam et al. (2000 b ; 2002) also found a good agreement between the measured concentrations of NH 4 N and NO 3 N and the respective concentrations simulated by LEACHM. Soulsby and Reynolds (1992) also used LEACHM to model soil water flux in Al leaching study and found good agreement between mod el predictions and simulated data. Models Used for Citrus Production The Citrus Water Management System (CWMS) is a new soil water and nitrogen balance model that was developed to help citrus growers in irrigation scheduling and nutrient management (Morg an et al., 2006c ) basing on earlier work on the citrus growing regions in Florida (Fares and Alva, 2000; Obreza and Pitts, 2002; Scholberg et al 2002; Morgan et al., 2006 b ; Wheaton et al., 2006; Fares et al., 2008). The model estimates soil water and ni trogen balances in multiple soil compartments under a mature citrus tree utilizing empirical relationships for water and nitrogen uptake and movement in sandy soils. According to Morgan et al.( 2006c ), CWMS requires initial setup information such as daily reference evapotranspiration (ET o ), rainfall amounts, irrigation duration (hours:min), nitrogen inputs, and fertilizer application rates. Also, the user of the model is required to provide information on irrigation system output characteristics (spray dia meter, inch; wetting pattern and flow rate, gal h 1 ), soil series, tree spacing parameters (the in row and between row tree distances), tree age (for estimation of canopy volume and calculation of root distribution). The Candler and

PAGE 63

63 Immokalee soil series that are found at the Ridge and Flatwoods sites, respectively, are included in the model. The CWMS model postulates two assumptions: (1) that all trees in a given planting area are of the same average size and have the same growth activities such that wa ter and nutrients taken from one area is the average of all other areas of similar size in the planting; (2) that runoff and lateral water movement are negligible in the very sandy and well drained Florida Ridge soil (Morgan et al., 2006c). The model was designed for mature citrus under microsprinkler irrigation and needs to be calibrated for young citrus under both drip and microsprinkler irrigation. Equations governing water movement, water uptake, nitrogen movement and uptake in the model were based on earlier research work (Williams and Kissel, 1991; Allen et al., 1998; Scholberg et al., 2002; Morgan, 2004; Morgan et al., 2006a). Other models developed for citrus production have been used in insect pest management to predict population and crop damage caused by citrus pathogens (Timmer and Zitko, 1996), scale insects (Arias Reveron and Browning, 1995), and in water management for irrigation scheduling (Xin et al., 1997). Summary The chapter reviewed the work regarding 1) management options for use under ACPS/OHS, 2) biomass and nutrient distribution in citrus, 3) water and fertilizer use efficiency, 4) microsprinkler and drip irrigation system design and scheduling, 5) root distribution in response to soil water, and, 6) models used in agricultural manag ement systems. As discussed above, options for optimizing nutrient and water uptake and use of weather based irrigation scheduling methods. Use of computer mode ls has been

PAGE 64

64 reviewed as one option for aiding the decision making process in citrus nutrient and water management practices.

PAGE 65

65 Table 2 1. Typical percent biomass distribution (dry weight basis) in oranges from different parts of the world Study area FL § FL a FL b SP a SP b IS a IS CA a CA b CA § § SP a SP b SP § §§ a SP § §§ b Age (years) 6 14 14 8 8 22 22 3.5 10 15 8 8 8 8 Plant tissue % Leaves 10 12 13 16 13 6 7 31 18 17 13 12 12 14 Branches, Twigs, and Trunk 32 61 52 32 30 55 56 38 61 61 33 34 29 32 F ruits 30 21 26 8 13 23 27 27 22 Roots 28 27 33 31 31 31 24 31 21 22 31 27 32 32 § Hamlin Swingle (25.0 kg tree 1 ) using microsprinkler irrigation (Mattos et al. 2003a b) Hamlin on Carrizo (104.3 kg tree 1 ) (a) and Swingle (82.6 kg tree 1 ) rootstocks (b) using microsprinkler irrigation (Morgan et al. 2006 a ) Navelina Carrizo (0.034 kg tree 1 ) using low frequency of N application combined with flood irrigation (a) and Navelina Carrizo (0.041 kg tree 1 ) using high frequency of N application c ombined with drip irrigation (b) (Quiones et al. 2003 a, b ) Shamouti (319.5 kg tree 1 ) with 223 g N tree 1 (a) and Shamouti (319.7 kg tree 1 ) with 763 g N tree 1 (b) using microsprinkler irrigation (Feigenbaum et al. 1987) Valencia (3.1 kg tree 1 ) (a ) and (80.1 kg tree 1 ) (b) Cameron and Appleman (1935) § § Valencia (94.6 kg tree 1 ) Cameron and Appleman (1945) Navelina Carrizo (39.15 kg tree 1 ) using two (a) and Navelina Carrizo (36.08 kg tree 1 ) using five (b) equal split applications of N with flo od irrigation (Quiones et al. 2005) § §§ Navelina Carrizo (41.03 kg tree 1 ) using drip irrigation by N demand (a) and Navelina Carrizo (37.49 kg tree 1 ) using drip irrigation by evapotranspiration (ET) demand (b) (Quiones et al. 2005) FL Florida, SP Spain, IS Israel, CA California = no data available

PAGE 66

66 Table 2 2. Typical nutrient uptake rates in oranges Plant tissue N K P N a N b N § a N §b N a N N N §§ a N §§ b N N b N a N b N N § §§ % Leaves 46 35 32 36 38 38 30 62 47 45 20 23 42 44 36 35 31 39 Branches twigs, trunk 15 15 24 39 32 20 25 21 39 35 44 42 18 15 25 23 22 24 Fruits 8 16 13 18 21 11 20 26 29 24 22 6 Roots 31 34 31 25 30 24 24 17 14 20 25 15 14 12 13 20 41 37 Tree age 6 6 6 14 14 8 8 3.5 10 15 22 22 8 8 8 8 5 3 Hamlin Swingle in Florida, % of total N Mattos et al. (2003a, b) using microsprinkler irrigation Hamlin Carrizo (a) and Hamlin Swingle (b) in Florida, % of total N Morgan et al. (2006 a ) using microsprinkler irrigation § Navelina Carrizo in Spain % of total N Quiones et al. (2003) using low frequency N application with flood irrigation (a) and high frequency N application with drip irrigation (b) Valencia in California, % of total N Cameron and Appleman (1935) for 3.5 year old (a) and 10 ye ar old (b) Valencia in California, % of total N Cameron and Compton (1945) §§ Shamouti in Israel, % of total N Feigenbaum et al. (1987) using microsprinkler irrigation with 223 g N tree 1 (a) and 763 g N tree 1 (b) Navelina Carrizo in Spain, % of tot al N Quiones et al. (2005) using flood irrigation schedules with two (a) and five (b) equal N splits Navelina Carrizo in Spain, % of total N Quiones et al. (2005) using drip irrigation by N demand (a) and ET demand (b) Calamondin in Spain, % of total N Legaz et al. (1982) § §§ Valencia in Spain, % of total N Legaz et al. (1995) = No data available

PAGE 67

67 Table 2 3. Soil test interpretation for soil P extraction methods compared with Mehlich 1 extractant § Extractant Soil test interpretation Very low Low Medium High Very high mg kg 1 Less than sufficient Sufficient Mehlich 1 <10 10 15 16 30 31 60 >60 Mehlich 3 <11 11 16 17 29 30 56 >56 Ammonium acetate pH 4.8 < 11 >11 Bray 1 P < 40 >40 Bray 2 P < 65 >65 § Koo et al., 1984 b ; Obreza et al ., 1999; 2008 b Table 2 4. Guidelines for interpretations of orange tree leaf analysis based on 4 to 6 month old spring flush leaves from non fruiting twigs Element Deficient Low Optimum High Excess % N <2.20 2.20 2.40 2.50 2.70 2.80 3.00 >3.00 P <0. 09 0.09 0.11 0.12 0.16 0.17 0.30 >0.30 K <0.70 0.70 1.10 1.20 1.70 1.80 2.40 >2.40 Koo et al. 1984 b ; Obreza et al. 1999; Obreza and Morgan, 2008.

PAGE 68

68 CHAPTER 3 NUTRIENT UPTAKE EFFI CIENCY AND DISTRIBUT ION IN SITU FROM THE CITRUS ROOT ZONE Intimately tied t o water management in citrus production systems is the need for efficient nutrient management strategies that enhance nutrient use efficiency while minimizing leaching losses in the root zone and sustain environmental quality. Several guidelines and crite ria are being and/or have been developed for managing water in concert with major nutrients in citrus production systems (Alva et al., 2003; Schumann et al., 2003; Alva et al., 2005; 2006a, b, c; Obreza et al., 2008 a, b ; 2010). Faced with the devastating citrus greening disease in Florida, researchers are attempting to explore ways of maximizing water and nutrient use efficiency by concentrating roots in the irrigated zones of microsprinklers or drip emitters, which should lead to high citrus yields and le ss nutrient leaching (Morgan et al., 2009 b). The concepts being promoted are termed Advanced Production Systems (APS) and Open Hydroponic Systems (OHS) (Stover et al. 2008, Morgan et al., 2009 b ). These two concepts are known to combine high density plant ings with intensive water and nutrient management thereby optimizing tree performance. An understanding of soil NH 4 + N, NO 3 N, P and K distribution patterns in the citrus root zone will help in devising ways of managing these critical nutrients for bette r horticultural, irrigation and environmental management. Leaching of NO 3 N, P and K are the greatest concern in all agricultural practices. Several researchers have shown the importance of applying recommended N rates to manage NO 3 N levels in groundw ater and soil (Lamb et al. 1999; Paramasivam et al., 2001; 2002; Alva et al., 2003; 2006a, b ; Sato et al., 2009a ) through use of carefully split N fertilizer applications (Quiones et al., 2003 a, b ; 2005; 2007) and well scheduled irrigation management (Al va

PAGE 69

69 et al., 2003; 2005; 2006b; Morgan et al. 200 9b ). Phosphorus (P) leaching has been identified recently as a threat to environmental quality (Sims et al., 1998; Boesch et al., 200 1 ; Agyin Birikorang et al., 2008). One strategy proposed by Obreza et al. (2008 a ) is for citrus producers to refrain from applying P fertilizer to young trees on Florida sandy soils if soil test P ranges from medium to very high levels according to University of Florida/Institute of Food and Agricultural Sciences recommendation s. Obreza et al. (2008 a ) observed that applying P fertilizer when it is not needed is wasteful and may cause undesirable enrichment of adjacent water bodies K is also considered a major nutrient in citrus production subject to leaching losses in the roo t zone. The extent of K leaching and distribution is mainly determined by drainage (Munson and Nelson, 1963), soil texture (Ylaranta et al., 1996) and irrigation practice (Sato et al., 200 9b ). Increasing nutrient availability in the irrigated zone will p robably lead to better water, N, P and K uptake and less nutrient leaching. This experiment was conducted to : 1) determine nutrient (NH 4 N, NO 3 N, Mehlich 1 P (M1P) and Mehlich 1 K (M1K)) and Br distribution patterns in the irrigated and non irrigated zon es as a function o f depth and fertigation method; 2) determine N, P and K concentration in below and above ground tissues. Using the OHS concept, we hypothesized that : 1) ammonium nitrogen (NH 4 + N) NO 3 N, Br, M1 P and M1K distribution will vary with dep th, distance from the tree and fertigation method; 2) ammonium nitrogen (NH 4 + N), NO 3 N, M1 P and M1 K will be higher in irrigated zones than nonirrigated zones and 3) plant N, P and K accumulation will be greater for OHS applied using microsprinklers or drip than grower practice.

PAGE 70

70 Materials and Methods Site Conditions The study was conducted at two locations: 1) University of Florida, Southwest Florida Research and Education Center, near Immokalee, Florida (Latitude 26.42N and Longitude 81.43W, at 10.4 1 m above sea level) and 2) a commercial grove near the University of Florida, Citrus Research and Education Center (SWFREC), near Lake Alfred, Florida (Latitude 28.09 o N, Longitude 81.75 o W, at 45.50 m above sea level) The soil at the Immokalee site was Im mokalee fine sand and consists of nearly level, poorly drained soils on the F latwoods formed in sandy marine sediments with slopes less than 2 percent (Obreza and Collins, 2008). These soils are classified as sandy, siliceous, hyperthemic Arenic Haplaquod s with the spodic horizon lying within 1m from the ground surface (USDA, 1990a). The soils at the research site near Lake Alfred was Candler fine sand and consists of excessively drained soils that formed in sandy marine or eolian deposits found on broad undulating upland ridges and knolls on flatwoods with slopes ranging from 0 8 percent. They are classified as hyperthermic, coated Typic Quartzipsamments (USDA, 1990b; Schumann et al., 2009). Study Treatments and Experimental Design At Immokalee, 3 yea r old citrus trees on S wingle rootstock were planted at 3.05 m between trees in a row and 6.71 m between tree rows Irrigation treatments at the Immokalee site were as follows: (1) Conventional practice (CMP) irrigated weekly and fertigated monthly; (2) D rip OHS (DOHS) irrigated and fertigated daily in small pulses; (3) Microsprinkler OHS (MOHS) irrigated daily and fertigated weekly. All the treatments were laid in a randomized complete block design replicated four times.

PAGE 71

71 A t the Lake Alfred site Haml in oranges were planted on Swingle rootstocks at 3.05 x 6.10m (~218 trees/acre) and on C35 rootstock at 2.44 x 5.49m (~302 trees/acre). The treatments imposed at the Lake Alfred site were similar to the set up at Immokalee except for the modification to t he conventional practice where the use of dry granular fertilizer applied under the canopy four times a year acted as a control for the experiment and also drip open hydroponic system was imposed on both Swingle and C 35 rootstock. The lay out of the trea tments are described in schematic diagrams (Appendi x D ). Plant Tissue and Soil Sampling Design and Analytical Methods Soil sampling In 2009 at Immokalee, twelve soil samples per replicate per treatment were collected in June and August for determination of NH 4 + N, NO 3 N, P and K concentration in each plot within a 30 cm x 45 cm grid in one quadrant of a given tree in a plot (Total number of soil cores = 3 treatments x 4 replicates x 12 cores samples per replicate x 2 profiles x 1 core per p rofile= 288 cores at Immokalee). Soil samples were collected at 0 to 15 cm and 15 to 30 cm depth s because this is where most roots of young citrus trees ( < 3 years old) are concentrated (Fares and Alva, 2000; Paramasivam et al., 2000c; Parsons and Morgan, 2004). In 2010, at Immokalee, soil samples were taken up to the 45 cm depth. S oil samples were also taken at 0 15, 15 30, 30 60 and 60 90 cm (June, 2010) from locations in the irrigated zones to analyze for NH 4 + N, NO 3 N, P and K. In June 2011, at Immo kalee, soil samples were taken in duplicates every two to three days at 0 15, 15 30, 30 45 and 45 60 cm at 15 cm and 45 cm from the tree to quantify nutrient movement in the irrigated and non irrigated zone using Br tracer.

PAGE 72

72 At the Lake Alfred site samples were collected in December 2009 at 4 locations per tree in a 15 cm x 15 cm grid (Total number of cores = 4 treatments x 4 replicates x 4 cores samples per replicate x 2 profiles x 1 core per profile= 128). In Ju ly 2010, at the Lake Alfred site, nine (9) samples were collected in a 30 cm x 30 cm grid per sampled tree in one replicate within the 0 30cm depth resulting in a total of 432 samples. S oil samples were also taken at 0 15, 15 30, 30 60 and 60 90 cm (Ju ly 2010) in the irrigated zones to analyze for NH 4 + N, NO 3 N, P and K. In August September 2011, at Lake Alfred, soil samples were taken in duplicates every two to three days at 0 15, 15 30, 30 45 and 45 60 cm at 15 and 45 cm from the tree to quantify nutrient movement in the irrigated and non irri gated zone using Br tracer. Nitrate N was compared with the maximum contaminant limit for drinking water standards (10 mg L 1 ) set by the U.S. Department of Health, Education and Welfare (1962) while P will be compared with numeric nutrient water quality c riteria explained by Obreza et al. (2010) and IFAS recommendations (Obreza and Morgan, 2008) Water sample collection and processing Water samples were collected every two days using suction lysimeters (Irrometer Co., Riverside, CA 92516) in July and Aug ust, 2009 at Immokalee for determination of NO 3 N leaching beyond the root zone ~50 cm at about 15 cm from the tree (irrigated zone) and 1 m away from the tree (non irrigated zone). The lysimeters were installed with a vacuum pressure pump for 5 minutes t o set a zone of lower pressure in the suction access tube to let soil solution flow into the lysimeter. The samples collected were filtered and later stored in a freezer at <4 o C until analysis.

PAGE 73

73 Extraction of NH 4 N, NO 3 N, P, Br and K To determine a mmoni um N and n itrate N concentration, 2 M KCl extraction procedure was used (Hanlon et al. 1997). Two wet subsamples per sample, one weighing approximately 4.5 g used for 2 M KCl extraction (in a ratio of 1 to 10 soil:solution ratio) and the other weighing 2 5 g for determination of oven dry soil weight (after drying for 24 h at 105 o C) to determine soil a mmonium N and n itrate N content on dry soil basis A 40 ml solution of 2 M KCl was added to the soil in each test tube, capped and shaken for 30 minutes. A fter shaking, all the sample solutions were allowed to settle for 30 minutes and filtered using Whatman filter paper # 42 into labeled vials, capped and stored in a freezer at <4 o C until analysis. Mehlich 1 extraction, a procedure recommended for soils w ith low organic matter and pH < 6.5, was used for determination of P and K (Mehlich, 1953). Air dried soil samples weighing 5.0 g (2 mm screened) were placed into extraction bottles and 20 ml of Mehlich 1 extracting solution was added to each sample and sha ken at high speed for 5 minutes at room temperature (252 C) and allowed to settle for 15 minutes. The extracts were filtered (Whatman filter paper # 42 ) and the supernatant was collected in labeled plastic vials and refrigerated Bromide was extracted u sing deionized water (soil:solution ratio of 1:2) by weighing about 5 g of dry soil and adding 10 ml of deionized water, shaking for 30 minutes and centrifuging at 5500 rpm. The suspension was filtered with Whatman filter paper # 42, capped and stored in p lastic vials until analysis according to the method described by Bogren and Smith (2003).

PAGE 74

74 Analysis of soil extracts and water samples Ammonium nitrogen (NH 4 N) and NO 3 N for soil samples were determined using a Flow Analyzer (Quich Chem 8500, Lachat Co.) at 660 nm and 520 nm, respectively (Harbridge, 2007a b) for samples collected in 2009, 2010 and 2011 at both sites Bromide was also analyzed using the Flow Analyzer (Quich Chem 8500, Lachat Co.) method (Bogren and Smith, 2003) Nitate nitrogen in water samples was also analyzed using the flow injection analysis method described by Harbridge ( 2007a ). Analysis for Mehlich 1 extractable P on samples collected in June 2009 at Immokalee was done by a DR/4000U Spectrophotometer (HACH INC.) at 880nm using a bla nk and four standards (4ppm, 8ppm, 12ppm and 16ppm) prepared in the Mehlich 1 extracting solution. Mehlich 1 extractable K for samples taken June 20 09 at Immokalee was determined by a 5100PC Atomic Absorption Spectrophotometer (Perkin Elmer Co.) at 766.5 nm prepared in the Mehlich 1 extraction solution using a blank and three standards (2 ppm, 6 ppm and 12 ppm). Samples collected at the Immokalee site in June 2010 and 2011 and those collected from the Lake Alfred site in December 2009, July 2010 and Augus t September 2011 were analyzed for M1P and M1K using Inductively Coupled Plasma (ICP) method (Hanlon et al. 1997) on a PerkinElmer Optical Emission Spectrometer Optim a 7000DV at 213.6 nm and 766.5 nm, for P and K, respectively. All results were expressed on oven dry soil mass basis. Plant tissue sampling and analysis Leaf sampling Leaf samples (a total of 20 leaves in four randomly sampled middle trees) were collected quarterly to determine nutrient uptake using the procedures outlined in Obreza and Mor gan (2008). Moist leaf samples were dried at 60C for 72 h and then passed

PAGE 75

75 through a stainless steel grinder with 20 and 60 mesh sieve and mixed thoroughly. Ground samples were stored at room temperature, but were redried at 60C for 2 h before weighing for analysis (Jones and Case, 1990; Plank, 1992; Hanlon et al. 1997). Destructive tree sampling and tissue processing Trees were destructively sampled in July 2011 at Immokalee (Figure 3 1) and August 2011 at the Lake Alfred site using methods adapted f rom Mattos (2000) and Morgan (2004). Before destroying the trees at Immokalee one representative tree per fertigation method was sampled and an area of 3.05 m x 3.05 m around each sampled tree was marked with a shovel to 30 cm depth. All trees were defo liated and the leaves were categorized into two: young or fully expanded, placed into separate plastic bags containing ice and taken to the laboratory. Twigs, f ruits small, medium and large branches were cut from each tree using clippers or manually, and placed in separate plastic bags. When all other tissues were collected from a particular tree, the trunk and roots were removed using an excavator. Thereafter, the soil was sifted and any remaining roots were collected. The roots were washed to remove any soil and debris before determining the fresh weight. All the tissues and tree parts were weighed for fresh weight determination. Thereafter, leaf tissues were dried for 72 h at 60 o C while larger tissue samples like the trunk, branches, fruits (fruits were cut into quarters after determining the fresh weight) and roots were dried at 60 o C for more than 14 days to constant weight. All the large tissues were cut into much smaller 1 cm wide pieces using a machete and an electric saw before passing them in to a larger grinder and then, the small ground tissues were passed through a stainless steel grinder with 20 and 60 mesh sieve and mixed thoroughly. At the Lake Alfred site because this was in a commercial grove and the trees could not be removed selec ted tissue s were sample d

PAGE 76

76 (twigs, leaves, fruits and roots) from one tree per irrigation method C ollected tissues samples we re handled as explained above. Tissue analysis Tissue N concentration (%) was determined using the NA2500 C/N Analyzer (Thermoques t CE Instruments). To accompli sh this, 5.0 mg of dry, ground tissue sample was weighed and compared to standards and blanks basing on calibration curve developed upon weighing and running approximately 2.5 mg, 5.0 mg and 10 mg of standard samples a nd two blanks on the analyzer. Tissue P and K concentration were determined using the dry ash combustion digestion method recommended by Anderson and Henderson (1988) for plant tissue analyses. Tissue K and P concentration were determined simultaneously by Indu ctively Coupled Plasma Atomic Emission Spectrometry ( ICP AES ). A 1.5 g sample of dried plant material was weighed and dry ashed at 500C for 16 h. The ash was equilibrated with 15 ml of 0.5 M HCl at room t emperature for h. Then the contents were gently swirled and allowed to settle for 1 h. The solution was decanted into 15 ml plastic disposable tubes for direct determination by ICP AES ( Munter and Grande, 1981; Munter et al., 1984 ; Fassel, and Kniseley, 1974). All samples were placed in a refrigerator at < 4 o C until extractions and analyses could be done (Plank, 1992; Morgan, 2004). Leaf N P, K concentration w ere compared with critical N PK levels for Florida Citrus (Obreza et al. 1999 ; Obreza and Morga n, 2008) and the concentration in all tissues was used to quantify the nutrient accumulation per tree. Quality Control of Plant Tissue and Soil Sample Analysis All sample collection/handling/chemical analysis was done according to standard procedures. A standard curve for certified standards (R 2 > 0.999) was developed for

PAGE 77

77 each set of samples. Method reagent blanks and duplicate samples were included for each 10 th and 20 th sample for soil samples while blanks, standards and duplicates were also included for each 10 th 20 th and 40 th sample of the tissue samples, respectively. Samples where blanks did not read as blank and/or where duplicates did not match with a relative standard deviation < 10% were re extracted and/or rerun until reasonably accurate results were obtained. Data Analysis All data were analyzed using PROC GLM Mixed model procedures using SAS 9.2. Means were separated using Duncan Multiple Range Test (DMRT). Results and Discussion Leaf NPK Concentration as a Function of Irrigation System L eaf N concentration using MOHS was similar to the other two treatments at Immokalee site (Figure 3 2) However, leaf N concentration using DOHS was significantly higher (p=0.03) than that observed under CMP. Thus, DOHS was effective in enhancing N conten t compared with the other two irrigation methods studied. Leaf P and K concentrations were similar across the treatments. Typical critical values for nutrient concentration in orange trees were suggested by Obreza and Morgan ( 2008 ). Most of the leaf N v alues observed in DOHS were either within the optimum (2.5 2.7%) or high (2.8 3.0%) ranges of nutrient concentration suggesting that nutritional requirements of the citrus tree were met by the irrigation and fertigation system. However, most leaf N values for MOHS and CMP were less than 2.5% showing that the two systems did not meet the critical N requirement of the tree. Leaf P concentration was high (0.17 0.30%) in all treatments while as leaf K concentration was within optimum and high (1.2 2.4%) range s suggesting that all the treatments met the P and K

PAGE 78

78 requirements (Figure 3 2) However, adjustments on fertilization programs proposed by Obreza et al. (2008 b ) need to be done on total N applied in the microsprinkler based systems so as to meet the N req uirement of the trees either by applying more fertilizer per unit time or increasing the fertigation frequency. It is apparent that the DOHS promoted more N concentration owing to the more frequent and localized fertigation. Another reason for the increa sed N concentration is ascribed to high root length density below the dripper (reported in the following chapters). It can be inferred that high root length density enhanced nutrient concentration The leaf N, P and K concentration at the Lake Alfred site were all in the optimum or high ranges greater than 2.5%, 0.12% and 1.2%, respectively. All the treatments grower practice and OHS fertigation practices met the tree nutrient requirements at the time of sampling (Figure 3 3). The fertilization practices at the Lake Alfred site did not show any significant differences in leaf concentrations between the four treatments showing that the fertilization rate was adequate in all the four methods used. NPK Distribution in the Citrus Root Zone as a Function of Ti me, Depth and Lateral Distance In June 2009 at Immokal ee, 2M KCl extractable ammonium N (p<0.001), M1P (p=0.0024) and M1K (p=0.001) were significantly different while 2 M KC l extractable nitrate N (p=0.27 ) was not significantly different among the fertiga tion methods (Table 3 1). Ammonium N, nitrate N and M1K were 9 to 34% higher using MOHS than CMP in the irrigated zone. Yet, using the same MOHS, M1P was 25% lower than conventional practice. DOHS significantly increased M1K, ammonium N, nitrate N and M1 P in the root zone by 44 to 133% over CMP. Ammonium N, nitrate N, M1P and M1K were significantly lower under MOHS by 4 to 73% compared with CMP in the non irrigated

PAGE 79

79 zone. In the non irrigated zone under DOHS, ammonium N and M1P were 3 to 6% lower than CMP while nitrate N and M1K were 4 to 48% higher than CMP (Table 3 1) Concentrations of 2M KCl extractable NH 4 + N and NO 3 N, M1 extractable K and P in soil samples decreased with depth in all treatments at SWFREC This is beneficial because root length density of citrus trees tends to be highest ne a r the soil surface and increase with depth as trees grow (Morgan et al., 2007) T hus, greater nutrient uptake has been found in the 0 30 cm horizon where, according to Paramasivam et al. (2000c), where 80% of the roots are concentrated. Except for ammonium N, it was observed that all the other forms of available nutrients decreased (p<0.05) with distance away from the tree (Figures 3 4 and 3 5) In August 2009 as shown in Table 3 2, no significant difference s in NH 4 + N and M1K were observed between fertigation methods, by depth and distance from the tree. Nitrate N (NO 3 N) decreased with depth in all treatments but was similar between fertigation methods. Mehlich 1 P (M1P) decreased with depth and distance f rom the tree, resulting in higher M1P distribution in the irrigat ed zone for MOHS and DOHS Thus M1P concentration in the top soil (0 15 cm) of the irrigated zones was sufficient and was low or medium at soil depth (15 30 cm) in all fertigation methods at Immokalee Results obtained from the soil samples collected in June 2010 are shown in Table 3 3. Ammonium N (NH 4 + N), NO 3 N and M1K differed by fertigation method (p<0.001), decreased with depth (p<0.001), and varied between irrigated vs. non irrigated zones. Mehlich 1P (M1P) decreased with depth and distance from the tree (p<0.001) and was very high in the 0 15 cm depth of irrigated zones of DOHS and MOHS ~80 mg kg 1 which was about twofold the M1P for CMP=48 mg kg 1 Significant interaction between

PAGE 80

80 f ertigation method x distance (p<0.05), depth x distance (p<0.001) for M1P, M1K and NH 4 + N suggests the importance of careful placement of the nutrients in the irrigated zones, proximal to the tree for better root uptake. The irrigated zones, as shown in Ch apter 4 showed high root length density, a tree characteristic that explains the potential for tree water and nutrient uptake. The 2M KCl extractable NH 4 + N and NO 3 N, M1P and M1K were significantly different among irrigation methods at the Lake Alfre d site in December 2009 (Table 3 4) and July 2010 (Table 3 5) Ammonium N ( NH 4 + N ) and M1K were higher at 0 15 cm than 15 30 cm in December 2009. In December 2009, NO 3 N was similar between DOHS Swingle, MOHS and DOHS C35 but all these were approximatel y one half of CMP~4 mg kg 1 The NO 3 N and NH 4 + N concentration was similar between irrigated and non irrigated zones of DOHS Swingle and DOHS C35 while the M1P was very high (>60 mg kg 1 ) in all treatments. In July 2010 at the Lake Alfred site NO 3 N a nd M1K decreased (p<0.001) with depth across all irrigation methods. Worth noting was the total NH 4 + N and NO 3 N that resulted in high total inorganic N concentration >25 ppm under conventional microsprinkler practice (CMP) in the 0 30 cm top soil that c ould pose an N leaching threat The M1P concentration at the Lake Alfred site at both sampling times were very high (>60 mg kg 1 ) across all the irrigation methods (Table 3 5) This suggests the need for lowering and/or adjusting the P application rate i n all irrigation methods, and also lowering the N application rate or changing the application method using CMP to minimize the risk of nutrient leaching beyond the root zone. There was significant

PAGE 81

81 interaction between irrigation method x distance from the tree for NO 3 N and M1K in December 2009 and July 2010, and M1P in December 2009 only. At the Immokalee site NH 4 + N and NO 3 N remained below 10 mg kg 1 in all fertigation methods in irrigated and nonirrigated zones in J une 2009 and June 2010 (Fig. 3 4 and 3 5 ). The M1P concentration was sufficient except in the irrigated zones of DOHS and MOHS where concentrations of 78 and 65 mg kg 1 were observed in 2009 and 2010, respectively (Fig. 3 10 ). The threat of P leaching was negated by high root length den sity associated with the irrigated zones of MOHS and DOHS as explained later in Chapter 4 Mehlich 1 K (M1K) concentrations were <20 mg kg 1 in all fertigation methods in both years suggesting increased K use and availability in the root zone. The lateral NH 4 + N and NO 3 N distribution remained below 2 mg kg 1 in all fertigation methods in 2009. The NH 4 + N and NO 3 N was uniformly distributed in the irrigated zone using CMP, and between the narrow north south irrigated stretch from 10 to 30 cm from the tre e under MOHS and just within 20 cm from dripper using DOHS. In 2010, the NH 4 + N and NO 3 N distribution pattern was similar to that of 2009 with concentrations <6 mg kg 1 but changed for DOHS due to the addition of another drip line and movement of the dr ippers to positions approximately 30 cm from the tree on the east and west of the tree. Nevertheless, nutrient concentrations using DOHS remained within 20 cm from the tree and below 3.5 mg kg 1 At the Lake Alfred site concentrations of NH 4 + N and NO 3 N were below 7 mg kg 1 and uniformly distributed across the irrigated and nonirrigated zones of all the irrigation methods in December 2009 (Table 3 4). In July 2010 at the same site, all NH 4 + N and NO 3 N concentrations were below 10 mg kg 1 in all irrig ation methods but CMP where

PAGE 82

82 NH 4 + N and NO 3 N concentrations as high as 17 and 29 mg kg 1 were observed due to granular fertilization compared with the fertigated methods. We noted fairly high concentrations of nutrients within 10 to 20 cm from the drip e mitter in the DOHS Swingle and DOHS C35 irrigation methods. Also M1P concentrations were very high (>60 mg kg 1 ) in December 2009 (Table 3 4) and July 2010 (Fig. 3 4) in both irrigated and nonirrigated zones suggesting the need for adjusting the P applicat ion rate to maintain the P concentration in the sufficiency range (<60 mg kg 1 ). The M1K concentration for CMP was >50 mg kg 1 while the M1K remained below 40 mg kg 1 for the fertigated methods decreasing in the order CMP>MOHS>DOHS C35>DOHS S wingle in Dec ember 2009 (Table 3 4). In July 2010, M1K was <35 mg kg 1 in all irrigation methods but CMP (Fig. 3 4) with concentration in the order CMP>DOHS S wingle >MOHS>DOHS C35. CMP showed M1K as high as 70 mg kg 1 Our results on drip irrigation also agree with th ose of Li et al. (2003) who found high nitrate concentrations within 15 cm from the dripper. However, Li et al. (2003) found that ammonium was concentrated just within 2.5 and 7.5 cm from the drip emitter suggesting that ammonium distribution was restricte d in a small volume, about 10 cm around the point source, probably due to adsorption and transformation to nitrate (Clothier et al., 1988; Clothier and Sauer, 1988) for an unsaturated soil. Earlier on, BarYosef and Sheikholslami (1976) found that NO 3 N wa s distributed to 16 cm radial distance within 21 h after first irrigation while phosphate movement was not directly proportional to water movement due to phosphate adsorption. Wang and Alva (1996) found that N leaching in microsprinkler irrigated citrus wa s a function of solubility of the fertilizer, soil type and duration of the leaching mechanisms. In their experiment, N

PAGE 83

83 leaching by N sources followed the order NH 4 NO 3 >Isobutylidene diurea>polyolefin resin coated urea, Meister. They also showed cumulative N leached from Wabasso sand (a Spodosol) was 58% of that from Candler sand. The proportion of N not found in the lower soil depth 45 60 cm can be attributed to tree root uptake, microbial immobilization, denitrification, soil N retention or ammonia volatil ization (Wang and Alva, 1996; Mattos et al. 2003c, Sato and Morgan, 2008; Morgan et al., 2009a). Our field experimental results somewhat contradict those of Wang and Alva (1996) who found NH 4 + N to be the dominant form of total N leached compared with NO 3 N in a greenhouse at 25 3 o C under intermittent leaching incubation conditions. Probably under the greenhouse conditions, the transformation of NH 4 + to NO 3 was somewhat limited by lack of air during saturated flow. Zvomuya et al. (2003) found that use of polylefin coated urea reduces N leaching but can also increase residual soil N on a loamy sand because NO 3 N leaching is associated with high rainfall and irrigation episodes. N, P Br and K L eaching in the Irrigated and Nonirrigated Zone s Leaching of NH 4 + N, NO 3 N, M1P, M1K at the Lake Alfred and Immokalee sites in 2010 in irrigated zones is described in Figures 3 16, 3 17 and 3 18 The results show that at SWFREC, soil ammonium N and NO 3 N (Figures 3 16 A and B) concentrations decreased with depth whi le remaining below 10 mg kg 1 that should lead to NO 3 N lower than the USEPA (2005) maximum contaminant limit of 10 mg L 1 At the Lake Alfred site NH 4 + N and NO 3 N was also below 10 mg kg 1 except for CMP where NO 3 N was >10 mg kg 1 throughout the 0 90 soil depth (Figures 3 16C and D) This underlined the gains of using fertigation compared with granular fertilization because residual NO 3 N was substantially increas ed nutrient retention with either drip

PAGE 84

84 or microsprinkler fertigation by 60 to 90%. These results support earlier findings in young and mature citrus (Alva et al. 1998; Scholberg et al. 2002; Morgan et al. 2009a) and other horticultural crops (Zotarelli et al., 2006; 2008a, b; 2009a, b; Scholberg et al., 2009). Muoz Carpena et al. (2 005) also noted that more frequent fertigation may be beneficial to sustain adequately high N concentrations and soil moisture content in a relatively small spherical soil volume surrounding the plants. According to Havlin et al. (2005) the problem of NO 3 leaching with CMP is explained by the fact that it is very soluble in water and is not strongly absorbed to the anion exchange capacity (AEC). Consequently, it is highly mobile and subject to leach ing losses when both soil N O 3 content and water movemen t are high. Thus, it is essential to minimize No 3 leaching by applying N synchronous with high crop N demand and peak N mineralization. Increasing N leaching potential occurs when inorganic profile N is present during periods of low evapontranspiration that coincides with periods of high precipitation, soil water content and drainage water. Timing N application to avoid periods of high water transport through the profile reduces leaching potential. Zotarelli et al. showed in their studies on pepper (2 006 ; 2007 ), tomato (2006 ; 2007 and 2009 a, b ) and sweet corn (200 8a ) that the fertigation systems help the root system to explore the entire soil volume efficiently and thus boost crop N uptake and growth culminating in potentially much more efficient utili zation of N fertilizer. Zotarelli et al. (2008b) demonstrated that N leaching increased when irrigation and N rates increased, with values ranging from 2 to 45 kg ha 1 of N. Thus, Zotarelli et al. (2008b; 2009a, b) showed that soil moisture based systems greatly improved irrigation water use efficiency thereby reducing irrigation water use and N leaching potential (NO 3 N leaching was reduced by 5 and 35 kg ha 1 )

PAGE 85

85 in tomato surface and subsurface irrigation systems. Also, Kiggundu et al. (2011) observed that NO 3 N leaching in a young avocado orchard appeared to be more influenced by the amount of water applied than the fertilizer rate suggesting that efficient irrigation methods have a greater potential to reduce nitrate leaching than reduced fertilizer appli cations. This speculation holds because under saturated flow, Havlin et al. (2005) contend that NO 3 ions move at similar speed as water molecules. Thus carefully managed irrigation schedules are critical in retaining nutrients in the root zone. Mehlich 1 P was high in the top 0 15 cm and decreased to <15 mg kg 1 at 60 90 cm depth (Figure 3 17A) It can be speculated that most of the P was available for uptake because most roots were concentrated in the top 0 15 cm and in the irrigated zones. Mehlich 1 P at SWFREC remained largely within the recommended levels (16 60 mg kg 1 ) in all treatments at 0 30 cm depth and was low (10 15 mg kg 1 ) or very low (<10 mg kg 1 ) in the 30 90 cm soil depth layer (Fig 3 17A). At the Lake Alfred site, CMP showed high M1P i n the range 31 60 mg kg 1 while OHS fertigation showed very high M1P (>60 mg kg 1 ) falling between 65 160 mg kg 1 Mehlich 1 P was very high in all the irrigation methods at the Lake Alfred site decreasing with depth (p<0.0001) suggesting the need for ad justing the annual P application rate with tree age. The very high M1 P concentrations will eventually affect P retention precipitation and adsorption mechanisms on the Ridge site with coated Candler fine sand With low solution P concentration, adsorptio n dominates while precipitation reactions proceed when solution P exceeds the solubility product (K sp ) of the specific P containing mineral. Thus, where water soluble fertilizers are applied, as was the case with grower practice, soil solution P concentrat ion increases greatly depending on P

PAGE 86

86 rate and method of application (Havlin et al., 2005) Nelson et al. (2005) attributed excessive P leaching on a loamy soil to over application of P, low P sorption capacity of the soil and rainfall exceeding evaporatio n. Soil solution P levels are buffered by adsorbed P on mineral surfaces (labile P), organic P mineralization and P mineral dissolution (Havlin et al., 2005) The P leaching in this study might be due to residual P, such that even with standard P applicati on rates and irrigation, excess P dissolves and becomes available for uptake and leaching as observed by several researchers (He et al., 2000; Nelson et al., 2005; Kiggundu et al., 2011). Thus understanding P fixation processes is important for optimum P n utrition and efficient fertilizer management. At Immokalee M1K decreased with depth and was <15 mg kg 1 throughout the 0 90 cm in all the treatments (Figure 3 18A) varying between 10 20 mg kg 1 in the 0 30 cm depth in all treatments. In the 30 90 cm soi l depth, K was ~15 mg kg 1 for CMP while that for MOHS and DOHS was lower than 10 mg kg 1 (Figure 3 18A). In July 2010, soil K at Lake Alfred site M1K remained between 15 and 45 mg kg 1 (Figure 3 18B). How ever Havlin et al. (2005) observed that K leachin g losses may be significant in coarse textured or organic soils in humid regions or under irrigation, as is the case with Florida. Thus carefully split applications rather than a buildup of soil K should be emphasized. Potassium ( K ) source ca n influence the amount K leached e.g. c ompared with KCl, the SO 4 2 and PO 4 2 sources exhibit greater anion adsorption to (+) exchange sites. Thus with fewer anions in solution available for leaching fewer K + would be leached Therefore, the efficient irrigation prac tices through ACPS/OHS practices would significantly reduce fertilizer lost through leaching by promoting uptake and reducing the occurrence of saturated flow and drainage as discussed later in Chapter 5.

PAGE 87

87 Measured soil Br distribution increased over time w ith depth on Immokalee sand accumulating in the bottom 30 60 cm layers in the irrigated zone. The soil Br spiked to 6 mg kg 1 in the irrigated and nonirrigated zone s on June 10, 2011 ( Appendix A, Figures A 1 and A 2) However, the initial soil Br was neg ligibly low ~0 1 mg kg 1 and started increasing immediately after Br application on June 8. A similar trend was observed within the non irrigated zone probably due to the rains in June 2011. At the Lake Alfred site soil Br was low (~0 1.5 mg kg 1 ), initi ally increasing to about 4.5 and 16 mg kg 1 after first application on August 23, 2011, and thereafter decreasing substantially due to heavy rains around the same time ( Appendix A, Figures A 3 and A 4 ) Thus most of the Br was washed away 6 to 10 days af ter application from the 0 60 cm soil profile The inorganic N (NH 4 ) remained between 0.5 and 14 .0 mg kg 1 in the irrigated and nonirrigated zone at 0 30 cm depth of Immokalee s ite ( Appendix A, Figure s A 5 and A 6 ). Nitrate N was consistently higher fo r DOHS ~ 6 1 2 mg kg 1 while CMP a nd MOHS remained between 1 and 6 mg kg 1 in the top 0 30 cm in both irrigated and nonirrigated zones on Immokalee sand ( Appendix A, Figures A 7 and A 8 ) The NH 4 N and NO 3 N concentration were similar in both irrigated and nonirrigated zones at 45 and 60 cm soil depths on Immokalee sand for CMP and MOHS while nitrate N was significantly higher in the irrigated than non irr i gated zones of DOHS At the Lake Alfred site NH 4 N ranged between 0.5 and 25 mg kg 1 in the top 0 30 cm depth in both irrigated and nonirrigated zones, with significantly higher values observed using CMP compared with ACPS/OHS treatments ( Appendix A, Figure s A 9 and A 10 ). The NH 4 N concentration at the Lake Alfred site varied between 0.5 and 9.0 mg kg 1 in the irrigated zone and between 0.2 and 3.0 mg kg 1 in the nonirrigated zone of the 30 60 cm soil depth ( Appendix A, Figure

PAGE 88

88 A 10 ). As shown in Figure s A 11 and A 12 the nitrate N values varied between 0.5 and 20 mg kg 1 in both irrigated and no nirrigated zones of the 0 30 cm soil depth layer for all treatments, except CMP where concentration as high as 30 mg kg 1 was noted in the irrigated zone. At 30 60 cm depth ( Appendix A, Figure s A 11 and A 12 ), all treatments but CMP, showed nitrate N var ying between 0 and 8 mg kg 1 while CMP nitrate N peaked to about 40 mg kg 1 at 30 45 cm depth( Appendix A, Figure A 11 ) in the irrigated zone vs. 12 mg kg 1 in the nonirrigated zone at 45 60 cm soil depth using CMP At the Immokalee site, Mehlich 1 P (M1P ) varied between 25 and 160 mg kg 1 in the top 0 30 cm of the irrigated zone ( Appendix A, Figure A 13 ), with very high M1P values decreasing in the order DOHS>CMP>MOHS in the 0 15 cm depth layer in the irrigated zone throughout the sampling period The M1 P in the nonirrigated zone was either high or very high for CMP and MOHS in the 0 30 cm soil depth and medium or low for DOHS ( Appendix A, Figure A 14 ). The M1P values remained between 0.1 and 50 mg kg 1 in the 30 60 cm depth of the nonirrigated zones sug gesting that most P might have been subjected to root uptake in the top 0 30 cm layer. Comparatively lower M1P values were noted in the 0 15 cm, 15 30 cm, and 30 45 cm and 45 60 cm soil depth layers of the nonirrigated zones M1P peak ed to ~145 mg kg 1 us ing DOHS, MOHS and CMP in the 0 45 cm segment in the irrigated zone and remained below 25 mg kg 1 in the lower 45 60 cm depth layer. This is a clear indication that P application rate ~50 70 kg P ha 1 for Immokalee sand is adequate for trees <5 yr old grow n on the Flatwoods using either conventional or ACPS/OHS practices without any serious P leaching threat. The M1P at the Lake Alfred site was very high with averages ranging from 100 to 250 mg kg 1 in the irrigated ( Appendix A, Figure A 15 ) and nonirrigat ed z ones of all

PAGE 89

89 treatments ( Appendix A, Figure A 16 ). The high values point to the need to lower the P application rate for the Lake Alfred site due to high residual P or the young tree age that might culminate in luxurious P consumption with no economic yield advantage. The irrigated zone of Immokalee sand ( Appendix A, Figure A 17 ) showed Mehlich 1 K (M1K) values in the range of 18 80 mg kg 1 in the 0 30 cm soil depth and between 12 40 mg kg 1 in the 30 45 cm soil depth suggesting a significant decrease in M1K with depth. Except for CMP at 0 15 cm soil depth where M1K~100 mg kg 1 was noted, most of the M1K values in the nonirrigated zones were <60 mg kg 1 ( Appendix A, Figure A 18 ). As explained earlier in Chapter 3, all the fertigated zones of CMP were irrigated implying a fairly uniform nutrient distribution around the tree using this system. A t the Lake Alfred site, significantly higher values of M1K (~75 175 mg kg 1 ) were noted using CMP in all the sampled depths in the irrigated zones while DOHS Swi ngle, MOHS and DOHS C35 showed values varying between 30 and 140 mg kg 1 in the 0 45 cm soil depth ( Appendix A, Figure A 19 ). Only in the 45 60 cm soil depth did M1K under MOHS vary between 25 and 110 mg kg 1 while DOHS Swingle and DOHS C35 remained betw een 25 and 55 mg kg 1 varying between 50 250 mg kg 1 at 0 15 cm, 60 100 mg kg 1 at 15 30 cm, 40 85 mg kg 1 at 30 45 cm, and 40 70 mg kg 1 at 45 60 cm soil depth layers. The MOHS, DOHS Swingle and DOHS C35 had values ranging between 20 75 mg kg 1 in all sampling depths except on August 24, 2011 when MOHS showed M1K~90 mg kg 1 (Figure Appendix A, A 20 ). Water Quality Analysis Nitrate N leaching was very low at Immokalee with nitrate N <1.7 mg L 1 in the irrigated zone of all treatments ( Appendix A, Figure A 21 ), much lower than the

PAGE 90

90 maximum contaminant limit suggested by USEPA of 10 mg NO 3 N L 1 At the Lake Alfred site, Schumann et al (2010) showed that CMP had nitrate concentration >10 ppm in July, August and September 2009 and March 2010 while MOHS had nitrate >10ppm in July 2009 and January 2010, and for DOHS <10 ppm was observed between July 2009 and January 2010. The results at the Lake Alfred site show that DOHS and carefully managed fe rtigated treatments should minimize N leaching, as also demonstrated at the Immokalee site. Biomass and Nutrient Distribution as a Function of Irrigation Practice Total above ground biomass (dry weight basis) contributed 69.4%, 75.4% and 76.0% while total below ground biomass (dry weight basis) accounted for 30.6%, 24.6% and 24.0% of total tree biomass using DOHS, MOHS and CMP at Immokalee respectively as shown in Table 3 6 The biomass under DOHS was distributed as follows: young leaves = 6.7%, fully e xpanded leaves = 12.8%, fruits = 11.6%, twigs = 11.2%, small branches = 4.3%, medium branches = 5.6%, large branches = 4.3%, trunk = 12.8%, small roots (<0.5 mm) = 1.0%, medium roots (0.5 1.0 mm) = 1.8 %, large roots (1.0 3.0 mm) = 3.6%, largest roots (>3 mm) = 24.2%. The biomass under MOHS was distributed as follows: young leaves = 8.0%, fully expanded leaves = 8.0%, fruits = 20.6%, twigs = 11.9%, small branches = 4.4%, medium branches = 3.6%, large branches = 6.6%, trunk = 12.3%, small roots (<0.5 mm) = 2.2%, medium roots (0.5 1.0 mm) = 0.6 %, large roots (1.0 3.0 mm) = 2.1%, largest roots (>3 mm) = 19.7%. The biomass under CMP was apportioned as follows: young leaves = 10.3%, fully expanded leaves = 2.2%, fruits = 23.5%, twigs = 9.5%, small branches = 4. 3%, medium branches = 4.1%, large branches = 8.3%, trunk = 13.8%, small roots (<0.5 mm) = 1.6%, medium roots (0.5 1.0 mm) = 0.4 %, large roots (1.0 3.0 mm) = 1.6%, largest roots (>3 mm) =

PAGE 91

91 20.3%. The subsamples of tissue samples for the Lake Alfred site are described in Table 3 7 Above ground tissues accounted for slightly above 90% of the total dry and fresh weight of subsamples while roots were <10% of total weight. The nutrient concentrations for the two research s ites are presented in Tables 3 8 and 3 9 With reference to guidelines of orange tree analysis in Table 2 4, N (%) was adequate for all treatments at sampling time. P and K were sufficient in all treatments. From the results, P was uniformly distributed among the various tissues in the treatme nts studied at Immokalee (Table 3 8) However, at the Lake Alfred site a fairly large amount of P was allocated to the roots for the S wingle rootstock (regardless of the fertigation method) and in the leaves for the C35 rootstock (Table 3 9) Generally, the N concentration was highest in the leaves followed by roots, fruits, twigs and branches. Potassium was distributed unif ormly across all tissues using Swingle rootstock but significantly low K (%) was noted in the roots of C 35 rootstock (<0.75%). Ov era ll, the study notes that most of the OHS treatments, including the conventional grower practices, meet orange tree nutrition requirements. As shown in the tree nutrient accumulation (Table 3 10) at Immokalee site, DOHS and MOHS accumulated about 44 % more N than CMP Thus, the nutrient accumulation showed lower N accumulation (~79 kg N ha 1 ) at Immokalee than DOHS (115 kg N ha 1 ) or MOHS (114 kg N ha 1 ) (Table 3 10). However CMP accumulated more P and K than DOHS and MOHS suggesting that even the grower pr actice was just as good in prompting nutrient accumulation. Nutrient accumulation at both sites analyzed in the leaves, fruits, twigs and roots showed that CMP at Immokalee had the lowest N accumulation in roots (5.6 13.8 g kg 1 ) while the ACPS/OHS practic es on Swingle

PAGE 92

92 rootstock at both sites and CMP at the Lake Alfred site had N contents ranging from 7.8 22.6 g kg 1 while the DOHS C35 had N content of ~22 g kg 1 in fibrous roots (<0.5 mm in diameter) and 5.8 9.9 g kg 1 in roots >0.5 mm in diameter. The N a ccumulation in twigs at Immokalee was 56 to 132% greater using ACPS than CMP. At the Lake Alfred site, N content in twigs was similar 10.6 13.6 g kg 1 in all the four fertilization methods, which was about 1.24 to 3.4 times greater than the Immokalee site The limited N accumulation in twigs might be ascribed to citrus greening in the Immokalee citrus trees in the third year of the study which might have limited N uptake. Nitrogen for C35 rootstock was largely allocated in the fruits (24.2 g kg 1 ) compared with trees in the other fertilization methods. The leaf N accumulation at both sites was between 25.2 and 37.7 g kg 1 At the Lake Alfred site, nutrient accumulation for N followed the order MOHS>DOHS C35>CMP>DOHS Swingle while P was DOHS C35>MOHS> DOHS S wingle >CMP and K was DOHS C35>MOHS> DOHS Swingle >CMP (Table 3 11).The P accumulation was similar among the fertilization methods at both sites, falling between 1.1 and 2.3 g kg 1 The K distribution in tissue shows fairly equal allocations to various pla nt parts using Swingle rootstocks while for C35, the K was largely allocated to the above ground parts (13.2 15.2 g kg 1 ) and lower portions (3.3 7.5 g kg 1 ) were allocated to the roots The only plausible explanation for high N, P and K accumulation of C MP would be the use of the granular fertilization (4 to 6 times annually) and controlled release fertilizer at the Lake Alfred site which might have promoted more N P and K absorption over time compared with monthly fertigated CMP at Immokalee. Summary Re sults over the 2 to 3 year studies showed that NH 4 + N, NO 3 N, M1P and M1K was uniformly distributed in the root zone of grower practices but was higher in the

PAGE 93

93 irrigated than nonirrigated zones of OHS fertigation practices. Overall, NH 4 + N, NO 3 N, M1P and M1K decreased with distance from the irrigated zone and with depth. This confirmed the NH 4 + N, NO 3 N, M1P and M1K would vary with depth, NH 4 + N, NO 3 N, M1P and M1K would be higher in irrigated increased nutrient retention and root uptake because the irrigated zone was associated with increased root de nsity as later discussed in Chapter 4 Nitrate N leaching was more pronounced for CMP at the Lake Alfred site with residual soil nitrate as high as 30 mg kg 1 but was largely minimal for all fertigation methods at Immokalee and the OHS fertigation methods at Lake Alfred. The use of Br suggested consistent trend s in the movement of N H 4 + N, NO 3 N, M1P and M1K in the irrigated and nonirrigated zones and could be used as an important guideline for making nutrient management decisions with regard to nutrient residence time M1P was very high at Lake Alfred site, despite applying the re commended rate probably because of the young tree age coated sands and residual P from previous tree plantings that could have become available from the sorbed or labile phases M1P application rate at the Lake Alfred site might need to be lowered over t ime to reduce P loading threat in to groundwater. T he citrus biomass distribution patterns were similar between the fertilization methods. All fertilization practices showed that leaf N, P and K concentrations were adequate. However propo r tion al nutrient accumulation patterns revealed that OHS fertigation increased N accumulation by 45% over grower practice at Immokalee, but P and K accumulation were fairly similar between the three practices, though CMP showed slightly higher P and K accumulation than OH S Thus, N accumulation

PAGE 94

94 confirmed the hypothesis that accumulation wo u ld be greater for OHS than grower practices The N, P and K concentration using granular fertilization at the Lake Alfred site suggests that grower practices are just as effective in promoting tissue nutrient concentration However, the grower practices (fertigated or under granular fertilization) might require more fertilizer and water applied per ha to achieve rapid tree de velopment within 1 to 5 years of establishing a grove compared with ACPS practices

PAGE 95

95 Figure 3 1. Destructive tree sampling in July 2011 at Immokalee with the root zone of the tree marked to 30 cm depth(A), tree after defoliation and fruit rem oval (B), tree after twig removal (C), fresh twigs (D), plucking the tree trun k and roots (E) and fresh roots (F) A B C D E F

PAGE 96

96 Figure 3 2. Leaf NPK concentration determined in June 2009 at Immokalee. Error bars denote one standa rd deviation of four replicates

PAGE 97

97 Figure 3 3 Leaf NPK concentration determined in August 2011 at the Lake Alfred site

PAGE 98

98 Table 3 1. 2M KCl extractable NH4 + N and NO 3 N, M1K and M1P concentrations of soil samples collecte d in June 2009 at SWFREC Fertigation method NH 4 + N NO 3 N M1P M1K IRR NI IRR NI IRR NI IRR NI mg kg 1 CMP 1.10 § 1.63 33.52 5.61 DOHS 1.65 1.07 2.45 1.69 78.16 31.47 8.06 8.30 MOHS 1.47 1.58 1.78 1.40 25.05 20.67 6.9 5.54 Soil depth (cm) 0 15 1.26 1.10 2.15 2.01 36.81 35.72 6.55 7.66 15 30 1.18 1.03 1.25 1.31 20.24 17.50 5.01 6.45 Statistics Fertigation method *** NS ** *** Depth NS ** Distance from the tree NS *** *** ** Fertigation method*Depth NS NS NS Fertigatio n method*Distance NS NS NS NS Depth*Distance NS NS NS NS Fertigation method*Depth*Distance NS NS NS NS IRR Irrigated, NI Non irrigated, § For conventional practice, all the sampled locations were irrigated. NS not significant p>0.05; p<0.05, p<0.01, *** p<0.001

PAGE 99

99 Table 3 2. 2M KCl extractable NH4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in August 2009 at SWFREC Irrigation method NH 4 + N NO 3 N M1P M1K Soil depth IRR NI IRR NI IRR NI IRR NI cm mg kg 1 CMP 0 15 1.47 § 1.35 32.10 13.10 15 30 1.12 1.20 12.80 11.27 DOHS 0 15 1.54 1.30 1.50 1.22 35.78 24.97 13.45 12.72 15 30 0.74 1.04 1.01 1.38 11.20 22.78 11.11 12.61 MOHS 0 15 2.03 1.81 1.63 1.51 50.62 22.10 13.55 13.19 15 30 1.54 1.61 1.08 1.2 3 11.98 9.61 11.67 9.06 Statistics Irrigation method NS NS NS Depth NS ** *** NS Distance from the tree NS NS NS Irrigation method*Depth NS NS NS NS Irrigation method*Distance NS NS NS NS Depth*Distance NS NS ** NS Irrigation method *Depth*Dis tance NS NS NS NS IRR Irrigated, NI Non irrigated, § For conventional practices, all the sampled locations were irrigated. NS not significant p>0.05; p<0.05, ** p<0.01, *** p<0.001

PAGE 100

100 Table 3 3. 2M KCl extractable NH 4 + N and NO 3 N, M1K and M1P conce ntrations of soil samples collected in June 2010 at SWFREC Fertigation method NH 4 + N NO 3 N M1P M1K Soil depth IRR NI IRR NI IRR NI IRR NI cm mg kg 1 CMP 0 15 2.07 § 6.44 48.04 14.88 15 30 1.75 4.70 27.93 15.53 30 45 1.32 3.50 17.15 12.28 DOHS 0 15 4.19 3.95 4.34 3.91 81.80 36.73 21.32 15.53 15 30 2.98 2.02 3.23 2.50 32.78 20.41 14.45 9.67 30 45 2.03 1.44 2.60 2.15 15.59 11.44 10.61 7.21 MOHS 0 15 3.33 3.69 5.24 4.70 81.24 30.89 14.21 10.23 15 30 1.89 2.97 3.98 7. 40 34.31 28.23 10.61 8.90 30 45 1.57 2.78 2.77 3.37 23.46 14.05 9.45 7.59 Statistics Fertigation method *** *** NS *** Depth *** *** *** *** Distance from the tree NS NS *** NS Fertigation method*Depth NS NS NS Fertigation method*Distance ** NS Depth*Distance NS NS *** NS Fertigation method*Depth*Distance NS NS NS NS IRR Irrigated, NI Non irrigated, § For conventional practices, all the sampled locations were irrigated. NS not significant p>0.05; p<0.05, ** p<0.01, *** p<0.001

PAGE 101

101 Table 3 4. 2M KCl extractable NH 4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in December 2009 at the Lake Alfred site Irrigation method NH 4 + N NO 3 N M1P M1K Soil depth IRR NI IRR NI IRR NI IRR NI cm mg kg 1 CMP 0 15 1.40 § 4.02 101.29 63.10 15 30 1.03 4.39 91.49 56.28 DOHS Swingle 0 15 1.18 1.23 1.65 1.74 154.74 104.10 10.70 13.48 15 30 0.92 0.83 1.48 1.43 136.33 122.42 7.06 8.98 DOHS C 35 0 15 1.59 1.75 1.71 1.49 209.24 104.72 31.75 22.90 15 30 1.35 1.10 1.38 1 .47 167.92 99.45 29.16 12.57 MOHS 0 15 1.12 1.53 147.01 37.97 15 30 0.94 1.62 139.35 24.58 Statistics Irrigation method *** *** *** Depth ** NS NS ** Distance from the tree NS NS *** NS Irrigation method*Depth NS NS NS NS Irrigatio n method*Distance NS ** *** ** Depth*Distance NS NS NS Irrigation method *Depth*Distance NS NS NS NS IRR Irrigated, NI Non irrigated, § For conventional practice and MOHS in December 20 09, all the sampled locations were irrigated, NS not significant p>0.05; p<0.05, ** p<0.01, *** p<0.001

PAGE 102

102 Table 3 5. 2M KCl extractable NH 4 + N and NO 3 N, M1K and M1P concentrations of soil samples collected in July 2010 at the Lake Alfred site Irrigation method NH 4 + N NO 3 N M1P M1K Soil depth IRR NI IRR NI IRR NI IRR NI cm mg kg 1 CMP 0 15 10.49 § 20.55 105.96 46.04 15 30 9.51 17.40 98.00 35.00 DOHS S wingle 0 15 2.64 2.12 7.59 6.31 105.88 88.52 35.56 24.55 15 30 2.98 2.77 3.97 2.71 99.08 93.51 27.29 18.13 DOHS C 35 0 15 0.82 0.59 6.83 5.56 158. 31 121.55 26.72 24.29 15 30 0.62 0.60 3.07 2.95 134.61 106.70 17.98 19.97 MOHS 0 15 2.09 1.63 6.64 7.35 114.69 115.87 24.79 25.98 15 30 2.21 1.48 3.55 3.55 126.04 131.63 21.42 21.39 Statistics Irrigation method *** *** *** *** Depth NS *** NS *** Distance from the tree NS NS NS NS Irrigation method*Depth NS NS NS NS Irrigation method*Distance NS ** NS Depth*Distance NS NS NS NS Irrigation method *Depth*Distance NS NS NS NS IRR Irrigated, NI Non irrigated, § For conventional practices, all the sampled locations were irrigated. NS not significant p>0.05; p<0.05, ** p<0.01, *** p<0.001

PAGE 103

103 Figure 3 4 Lateral ammonium N distribution at 0 30 cm soil depth in June 2009 and 2010 at the Immokalee site

PAGE 104

104 Figure 3 5 Lateral nitrate N distribution in June 2009 and 2010 at Immokalee s ite

PAGE 105

105 Figure 3 6 Lateral ammonium N distribution in December 2009 on Candler fine sand

PAGE 106

106 Figure 3 7 Lateral ammonium N distribution in July 2010 at the Lake Alfred site

PAGE 107

107 Figure 3 8 Lateral nitrate N distribution in December 2009 at the Lake Alfred site

PAGE 108

108 Figure 3 9 Lateral nitrate N distribution in July 2010 at the Lake Alfred site

PAGE 109

109 Figure 3 10 Lateral Mehlich 1 P distribution at Immokalee site in June 2009 and 2010

PAGE 110

110 Figure 3 11 Lateral Mehlich 1 P distribution in the 0 30 cm depth layer at the Lake Alfred site in December 2009

PAGE 111

111 Figure 3 12 Lateral Mehlich 1 P distribution in the 0 30 cm depth layer at the Lake Alfred site in July 2010

PAGE 112

112 Figure 3 13 Lateral Mehlich 1 K distribution at Immokalee in June 2009 and 2010

PAGE 113

113 Figure 3 14 Lateral Mehlich 1 K distribution at the Lake Alfred site in December 2009

PAGE 114

114 Figure 3 15 Lateral Mehlich 1 K distribution at the Lake Alfred site in July 2010

PAGE 115

115 Figure 3 16 Vertical nitrate N and ammonium N distribut ion in June 2010 at Immokalee site (A and B) and in July 2010 at the Lake Alfred site (C and D)

PAGE 116

116 Figure 3 17 Vertical M1P distribution in June 2010 at Immokalee site (A) and in July 2010 at the Lake Alfred site ( B ) Figure 3 18 V ertical M1 K distribution in June 2010 at Immokalee site (A) and in July 2010 at the Lake Alfred site ( B )

PAGE 117

117 Table 3 6 Fresh and dry tissue weight for samples collected in July 2011 at Immokalee Fertigation method CMP DOHS MOHS Tissue FW % DW % FW % DW % FW % DW % Young leaves 4,391.4 12.8 1,598.2 10.3 2,989.9 8.3 1,017.4 6.7 3,349.0 9.2 1,208.0 8.0 Fully expanded leaves 923.5 2.7 340.9 2.2 5,712.3 15.8 1,947.1 12.8 3,350.9 9.2 1,212.3 8.0 Fruits 12,922.7 37.5 3,657.3 23.5 9,685.8 26.8 1,763.8 11.6 13, 557.9 37.2 3,111.9 20.6 Twigs 2,890.0 8.4 1,480.0 9.5 3,394.6 9.4 1,700.0 11.2 3,589.9 9.9 1,800.0 11.9 Small branches 1,187.6 3.5 676.0 4.4 1,186.5 3.3 650.3 4.3 1,124.3 3.1 667.5 4.4 Medium branches 1,065.9 3.1 633.4 4.1 1,474.2 4.1 847.0 5.6 952.5 2. 6 542.2 3.6 Large branches 2,156.1 6.3 1,286.2 8.3 1,134.0 3.1 655.7 4.3 1,678.3 4.6 1,006.5 6.7 Trunk 3,447.3 10.0 2,143.9 13.8 3,243.2 9.0 1,934.2 12.8 3,164.4 8.7 1,869.5 12.4 Total above ground 28,984.5 84.2 11,815.7 76.0 28,820.4 79.7 10,515.5 69. 4 30,767.3 84.5 11,418.0 75.4 Roots (<0.5mm) 501.0 1.5 253.6 1.6 298.4 0.8 155.7 1.0 653.0 1.8 329.0 2.2 Roots (0.5 1mm) 84.9 0.3 67.3 0.4 449.2 1.2 273.5 1.8 100.9 0.3 86.0 0.6 Roots (1 3mm) 335.0 1.0 252.9 1.6 1,155.5 3.2 547.1 3.6 383.6 1.1 322.0 2.1 Roots (>3mm) 4,532.8 13.2 3,153.7 20.3 5,432.0 15.0 3,667.5 24.2 4,525.3 12.4 2,979.9 19.7 Total below ground 5,453.7 15.8 3,727.5 24.0 7,335.1 20.3 4,643.8 30.6 5,662.9 15.5 3,716.9 24.6 Total 34,438.2 100.0 15,543.2 100.0 36,155.5 100.0 15,159.3 10 0.0 36,430.1 100.0 15,134.9 100.0 FW Fresh weight, DW Dry weigh t in g

PAGE 118

118 Table 3 7 Fresh and dry tissue weight for samples collected in August 2011 at the Lake Alfred site Irrigation method CMP DOHS Swingle MOHS DOHS C35 Tissue FW % DW % FW % DW % FW % DW % FW % DW % Leaves 160.1 12.6 60.2 26.5 199.7 15.8 72.1 31.7 163.4 12.9 56.7 25.0 166.0 13.1 59.3 26.1 Fruits 811.5 64.1 104.7 46.1 1,003.8 79.3 123.8 54.5 850.6 67.2 104.4 45.9 844.4 66.7 103.7 45. 6 Twigs 24.6 1.9 10.5 4.6 43.1 3.4 19.8 8.7 32.9 2.6 14.2 6.2 32.8 2.6 14.3 6.3 Total above ground 996.1 97.1 175.5 90.8 1,246.6 98.4 215.7 94.9 1,046.9 96.0 175.3 94.7 1,043.2 97.8 177.3 91.7 Small roots (<0.5 mm) 4.0 0.3 2.8 1.2 6.1 0.5 4.7 2.1 5.4 0.4 4.5 2.0 4.5 0.4 3.1 1.4 Medium roots (0.5 1 mm) 1.2 0.1 0.6 0.3 1.3 0.1 0.3 0.1 1.3 0.1 0.9 0.4 0.9 0.1 2.0 0.9 Large roots (1 3 mm) 5.0 0.4 2.6 1.1 6.3 0.5 2.9 1.3 4.4 0.3 3.0 1.3 3.1 0.2 1.5 0.7 Largest root (>3 mm) 20.0 1.6 11.8 5.2 6.2 0.5 3.7 1.6 32.5 2.6 1.5 0.7 14.8 1.2 9.4 4.1 Total below ground 30.1 2.9 17.8 9.2 20.0 1.6 11.6 5.1 43.6 4.0 9.9 5.3 23.2 2.2 16.0 8.3 Total 1,026.3 100.0 193.3 100.0 1,266.6 100.0 227.3 100.0 1,090.5 100.0 185.2 100.0 1,066.5 100.0 193.3 100.0 FW Fresh weight, DW Dry weight in g

PAGE 119

119 Table 3 8 N, P and K concentration in tissues collected in July 2011 at the Immokal ee site Fertigation method CMP DOHS MOHS Tissue N P K N P K N P K % Young leaves 2. 58 0.14 1.83 3.25 0.11 1.11 3.3 3 0.12 1.36 Old, fully expanded leaves 2.3 1 0.15 1.35 3.53 0.12 1.20 2.9 4 0.13 1.17 Twigs 1.22 0.13 1.45 1.54 0.14 1.15 1.30 0.17 1.3 6 Small branches 0.37 0.13 0.88 0.58 0 11 1.13 0.86 0.13 0.89 Medium branches 0.34 0.14 1.29 0.71 0.11 0.92 0.32 0.13 1.12 Large branches 0.53 0.17 1.30 0.45 0.16 1.41 0.54 0.12 1.22 Trunk 0.60 0.18 1.35 0.99 0.14 1.36 0.95 0.15 1.15 Fruits 1.25 0.15 1.41 1.83 0.12 1.55 1.97 0.15 1.26 Small roots (<0.5 mm) 1.38 0.14 1.19 2.26 0.11 1.12 1.91 0.13 1.31 Medium roots (0.5 1 mm) 1.21 0.13 1.21 1.60 0.13 1.15 1.46 0.13 1.37 Large roots (1 3 mm) 0.65 0.14 1.19 1.11 0.15 1.55 2.62 0.14 1.43 Largest root (> 3 mm) 0.56 0.16 1.27 0.78 0.13 1.41 0.85 0.17 1.42 Table 3 9 N, P and K concentration in tissues collected in August 2011 at the Lake Alfred site Irrigation method CMP DOHS Swingle MOHS DOHS C 35 Tissue N P K N P K N P K N P K % Leaves 3.25 0. 13 1 53 2.52 0. 13 1 41 3.73 0. 14 1 33 2.62 0. 20 1 52 Fruits 1.27 0. 11 0.9 3 1.05 0. 12 1 01 1.69 0.1 3 1 33 2.42 0. 19 1.4 7 Twigs 1.26 0.1 1 1 13 1.06 0. 16 1 41 1.36 0. 14 1 09 1.07 0. 15 1 32 Small roots (<0.5 mm) 2.14 0. 18 1 25 2.05 0. 15 1 28 1.70 0. 13 1 43 2.2 1 0. 18 0. 45 Medium roots (0.5 1 mm) 1.78 0. 17 1 20 2.09 0. 23 2 28 0.99 0. 26 2 21 0.58 0.1 7 0. 75 Large roots (1 3 mm) 1.26 0. 12 1 09 1.26 0. 13 1.2 4 0.85 0. 14 1 47 0.86 0. 11 0. 33 Largest root (>3 mm) 1.11 0. 12 1 01 0.89 0. 14 1 53 1.70 0. 16 1 49 0.99 0. 1 1 0. 33

PAGE 120

120 Table 3 10 Nitrogen, phosphorus and potassium accumulation on Immokalee sand Fertigation method CMP DOHS MOHS CMP DOHS MOHS CMP DOHS MOHS Tissue N P K kg ha 1 Young leaves 20.16 16.17 19.67 1.09 0.55 0.71 14.30 5.52 8.03 Old, fully e xpanded leaves 3.85 33.61 17.43 0.25 1.14 0.77 2.25 11.43 6.94 Fruits 22.36 15.78 29.98 2.68 1.03 2.28 25.22 13.37 19.17 Twigs 8.83 12.80 11.44 0.94 1.16 1.50 10.49 9.56 11.97 Small branches 1.22 1.84 2.81 0.43 0.35 0.42 2.91 3.59 2.91 Medium branches 1.05 2.94 0.85 0.43 0.46 0.34 4.00 3.81 2.97 Large branches 3.33 1.44 2.66 1.07 0.51 0.59 8.18 4.52 6.00 Trunk 6.29 9.36 8.68 1.89 1.32 1.37 14.15 12.86 10.51 Small roots (<0.5 mm) 1.71 1.72 3.07 0.17 0.08 0.21 1.48 0.85 2.11 Medium roots (0.5 1 mm) 0. 40 2.14 0.61 0.04 0.17 0.05 0.40 1.54 0.58 Large roots (1 3 mm) 0.80 2.97 4.13 0.17 0.40 0.22 1.47 4.15 2.25 Largest root (>3 mm) 8.64 13.99 12.39 2.47 2.33 2.48 19.59 25.29 20.69 Total 78.65 114.78 113.72 11.64 9.52 10.95 104.43 96.49 94.13 Total N PK accumulation is fairly low compared with accumulation observed in a typical orange grove owing to citrus greening infection

PAGE 121

121 Table 3 11. N, P and K accumulation in 2011 at the Lake Alfred and Immokalee site s Irrigation method CMP DOHS Swingle MOHS DOHS C35 Lake Alfred N P K N P K N P K N P K g kg 1 Leaves 32.5 1.3 15.3 25.2 1.3 14.1 37.3 1.4 13.3 26.2 2.0 15.2 Fruits 12.7 1.1 9.3 10.5 1.2 10.1 16.9 1.3 13.3 24.2 1.9 14.7 Twigs 12.6 1.1 11.3 10.6 1.6 14.1 13.6 1.4 10.9 10.7 1.5 13.2 Small roots (<0.5 mm) 21.4 1.8 12.5 20.5 1.5 12.8 17.0 1.3 14.3 22.1 1.8 4.5 Medium roots (0.5 1 mm) 17.8 1.7 12.0 20.9 2.3 22.8 9.9 2.6 22.1 5.8 1.7 7.5 Large roots (1 3 mm) 12.6 1.2 10.9 12.6 1.3 12.4 8.5 1.4 14.7 8.6 1.1 3.3 Largest root (>3 mm) 11.1 1.2 10.1 8.9 1.4 15.3 17.0 1.6 14.9 9.9 1.1 3.3 Fertigation method CMP DOHS MOHS Immokalee N P K N P K N P K Young leaves 25.8 1.4 18.3 32.5 1.1 11.1 33.3 1.2 13.6 Fully expanded leaves 23.1 1.5 13.5 35.3 1.2 12.0 29.4 1.3 11.7 Fruits 12.2 1.3 14.5 15. 4 1.4 11.5 13.0 1.7 13.6 Twigs 3.7 1.3 8.8 5.8 1.1 11.3 8.6 1.3 8.9 Small branches 3.4 1.4 12.9 7.1 1.1 9.2 3.2 1.3 11.2 Medium branches 5.3 1.7 13.0 4.5 1.6 14.1 5.4 1.2 12.2 Large branches 6.0 1.8 13.5 9.9 1.4 13.6 9.5 1.5 11.5 Trunk 12.5 1.5 14.1 1 8.3 1.2 15.5 19.7 1.5 12.6 Roots (<0.5mm) 13.8 1.4 11.9 22.6 1.1 11.2 19.1 1.3 13.1 Roots (0.5 1mm) 12.1 1.3 12.1 16.0 1.3 11.5 14.6 1.3 13.7 Roots (1 3mm) 6.5 1.4 11.9 11.1 1.5 15.5 26.2 1.4 14.3 Roots (>3mm) 5.6 1.6 12.7 7.8 1.3 14.1 8.5 1.7 14.2

PAGE 122

122 CHAPTER 4 EFFECTS OF FERTIGATI ON AND IRRIGATION RA TES ON ROOT LENGTH DISTRIBUTION AND TRE E SIZE The use of automated irrigation systems and intensive nutrient management is critical to citrus production systems for achieving increased tree growth and yield Maintenance of soil moisture and nutrient concentrations in the tree root zone near optimum levels is known as the open hydroponic system (OHS) (Morgan et al., 2009 b ). Sound water and nutrient management is required in Florida soils with high sand cont ent (>94%) and low organic matter content because leaching and subsequent pollution of groundwater is a likely threat. Key to improving citrus nutrient and water uptake is the understanding of the root system dimensions, topological properties and distrib ution in the soil. Of these root properties, the property of greatest importance is root length density ( RLD ) distribution because it defines limits to the efficiency of a root system in absorbing water and nutrients (Tinker and Nye, 2000; Himmelbauer et al., 2004). Studies on tree RLD distribution done in Florida by Morgan et al. (2007) found that fibrous root length density (FRLD) distribution increased with soil depth and lateral distance as trees grew, resulting in mature trees with bimodal root syste ms. In their study, they classified fibrous roots as those roots whose diameter fell between 0 4 mm because such roots determine tree water and nutrient uptake efficiency. Morgan et al. (2007) reported that FRLD varied as a function of rootstock in which trees on Swingle citrumelo developed higher FRLD near the soil surface and lower FRLD below 0.3 m than trees on Carrizo citrange. Abrisqueta et al. (2008) studied root dynamics of young peach subjected to partial root zone drying and continuous deficit ir rigation in Spain. In the study, higher

PAGE 123

123 root length densities were recorded in non limiting irrigation conditions than under deficit irrigation where root growth was reduced. Two methods (plant based or soil based) have been used to estimate and describe root systems. The plant based method describes the way in which different parts of the root system are interconnected (Rose, 1983; Klepper, 1992). The second method describes root systems in the soil in terms of the distribution of RLD or mass throughou t the rooting zone and has been used as a standard way of measuring density in distributions of roots in field soils (Barraclough and Leigh, 1984; Vincent and Gregory, 1989a, b; Masse et al. 1991). Basing on the latter, researchers devised methods of soi l coring and root washing to provide the most practicable way of obtaining quantitative data on root system length and distribution in the field (Tinker and Nye, 2000). The main methods that have been used for measuring root length over the years are line intersect method (Newman, 1966); direct measurement and opisometer methods ( Reicosky et al 1970); photocopying and scanning (Collins et al. 1987; Kirchoff, 1992; Himmelbauer et al. 2004 ) and the stereological procedure (Wulfsohn et al. 2004). Desp ite its merits, the line intersect method uses a tedious operational procedure which includes insuring uniform root dispersal throughout a finite area and the repetitive use of short line intercepts ( Reicosky et al 1970 ). The study by Reicosky et al (19 70) showed significant gains in time by using the line intersection method over the direct and opisometer methods. Reicosky and colleagues found that there was little difference in precision between the line intersect, direct and op i someter methods for es timating root length but found more gains on time in using the first method (1.0 h) compared with the latter two where it took 5.0 h and 1.5 h for the direct and opisometer methods,

PAGE 124

124 respectively. Thus, through root scanning, the line intersect method can b e calibrated and used to predict root length with speed and greater precision (Collins et al. 1987; Bland and Mesarch, 1990). Studies done in central and south Florida showed that tree size w as a function of root density (Castle and Krezdon, 1975; Ford, 1954; 1964; 1972), root stock (Morgan et al., 2006a) and fertilization practice (Obreza and Rouse, 1991; 1993; 2006; Morgan et al., 2009a). Marler and Davies (1990) showed that canopy volume, trunk cross sectional area and root dry weight can be influenc ed by irrigation rate. In their study, canopy volume and trunk cross sectional area were similar at high (20 % of available soil water depletion) and moderate (45 % of available soil water depletion) levels in 2 of 3 years, but were reduced at low (65 % of available soil water depletion). More than 90 % of the roots were within 80 cm of the tree trunk at the end of the growing season. Parsons et al. (2001), in their study on the effect reclaimed water on citrus tree growth, found that tree growth was greate st at high irrigation rate (2500 mm) though fruit production per canopy volume was low compared with lower rates ~400 mm and 1250 mm. However, very little research, if any, has been conducted to determine the effect of irrigation rate and fertilization met hod on tree size in Florida using the modified ACPS/OHS practices. Documentation of the performance of ACPS/OHS practices with regard to tree size and root density is critical for their adaptation to Florida soil and climatic conditions. The objectives of the experiment were to: (1) calibrate line intersect method for determining RLD in 1 and 3 year old citrus using the digital scanning method on Florida Entisol and Spodosol,

PAGE 125

125 2) to determine the effect of fertigation frequency and irrigation method on RL D distribution, (3) v alidate the RLD estimate d based on root area using the intercept method (4) determine root distribution patterns in the irrigated and non irrigated zones as a function of fertigation method and depth, and, (5) determine the effect o f fertigation frequency and irrigation method on canopy volume and trunk cross sectional area. The following hypotheses were postulated: (1) root area using a flatbed scanner can be calibrated using the line intersect method and used to predict root len gth with speed and greater precision 2) spatial root length density distribution will be greater in irrigated zones of microsprinkler and drip OHS than conventional practice, and 3) microsprinkler and drip OHS will increase citrus growth rate resulting in canopy volumes and trunk cross sectional areas higher than conventional practice. Materials and Methods Description of Study Sites and Treatments Treatments and orchard locations for this study were the same as trees used in the nutrient distribution a nd accumulation study presented in Chapter 3. A t the SWFREC site treatments were : (1) Conventional practice irrigated weekly and fertigated monthly (CMP); (2) Drip OHS irrigated daily and fertigated weekly in small pulses (DOHS); (3) Microsprinkler OHS irrigated daily and fertigated weekly (MOHS). All the treatments were laid in a randomized complete block design replicated four times. The Hamlin oranges ( Citrus sinensis ) on S wingle rootstock were planted on a 3.05 x 6.71 m tree and row spacing. A s econd study was installed at a 15 acre Ridge site near the Citrus Research and Education Center (CREC), Lake Alfred, with Hamlin oranges on Swingle rootstocks at 3.05 x 6.10m (~218 trees/acre) and C 35 rootstock at 2.44 x 5.49m (~302 trees/acre). The treat ments imposed at the Lake Alfred site were

PAGE 126

126 similar to the set up at SWFREC except for the modification to the conventional practice where the use of dry granular fertilizer applied under the canopy four times a year acted as a control for the experiment an d also DOHS was imposed on both Swingle and C 35 rootstock. Root Sampling Methods Roots were sampled for RLD and average root diameter estimations of 3 year old citrus at SWFREC, o n June 9 through 17, 2009 in the 0 30 cm depth at 15 cm depth increments bec ause this is where most roots of young citrus trees (<3 years old) are concentrated (Fares and Alva, 2000; Paramasivam et al., 2000c; Parsons and Morgan, 2004). The samples were collected at 0, 15, 30, and 45 cm distance from the tree in the row in 15 cm increments up to 45 cm from the tree perpendicular to the planted row giving a total of 12 sampling locations in one quadrant for each tree (3 x 4 grid). On June 16 through 24, 2010 root samples at SWFREC were collected up to 45 cm depth using the above sampling scheme. At the Lake Alfred site fewer samples than those at SWFREC were collected using a 2 x 2 grid on December 22, 2009 and a 3 x 3 grid o n July 7 and 8, 2010 at 0 15 cm and 15 30 cm depth. All the cores were carefully bagged, labeled and store d in a refrigerator at < 4 C awaiting subsequent analysis. Roots were removed from the soil using a 2 mm diameter sieve. O ther debris passing through the sieve was removed manually. The roots were hydrated for 15 minutes and categorized into four groups according to diameter: <0.5 mm, 0.5 1.0 mm, 1 3 mm and >3 mm before using the line intersection method (adapted from Morgan, 2004). Root length for each root category was estimated using the grid system explained by Tennant (1975) by counting the number o f horizontal and vertical intersections of roots in a grid system of 1.0 x 1.0 cm which was

PAGE 127

127 multiplied by 11/14 and divided by volume of the corer to give root length density in cm cm 3 soil (Mattos, 2000). Estimation of Tree Growth Characteristics Tree canopy volumes were made by measuring the average canopy diameter using canopy width in the east west and north south directions and canopy height. Then, using the formula for a sphere= r 3 where r is the canopy radius (where the canopy width in the east west and north south directions and canopy height were averaged to give the canopy radius), canopy volume was calculated. Trunk diameter was estimated from averaging the diameter in the east west and north south directions and then calculating the a 2 assuming circular shape, where r is the trunk radius. Statistical Analysis Roots of selected root diameters from samples collected in two of the four replications at SWFREC were designated the calibration set and roots from the remaining two replications were designated the validation set. The RLD estimate using the intercept method (dependent variable) and mean root area using the scanning method (independent variable) for calibration set roots were correlated resulting in a c alibration curve for the root scanning method. To validate the calibration curve, estimated RLD calculated using the mean root areas of the validation set roots were correlated with the RLD estimate of the same roots using the intercept method. After val idating the procedure at Immokalee and getting reasonably good calibration curves, a calibration curve was also determined for estimating RLD at Lake Alfred using similar calibration and validation techniques All correlations were done using SAS 9.2

PAGE 128

128 PROC REG procedures (SAS Inst. 20 11 ) and SIGMA PLOT 10.0. The rest of the data were analyzed using PROC GLM Mixed Model Procedures (SAS Inst. 20 11 ) to determine the effect of the irrigation method and fertilization frequency on vertical and lateral root length density distribution, and tree growth. Results and Discussion Correlation of RLD Measured by Intersection Method versus Scanning Method Root length density measured using the modified Newman method (Tennant, 1975) correlated well with scanned area of the calibration set at both SWFREC and Lake Alfred sites ( Appendi x F ). RLDs predicted by the calibration scanning method also agreed with measured RLDs for the validation set at both SWFREC and the Lake Alfred site (Tables 4 1 and 4 2) The strong positive correlation for all diameters of roots collected at both locations shows the precision of using the scanning method upon calibration with the line intersection method. Only 1.61% of the scanned roots at CREC and 2.25% of the scanned roots at SWFREC were above the normal deviations (relative standard deviation 10%) showing that scanning roots in triplicate represent a fairly accurate and precise way of determining RLD The equations for the calibration set using scanned root area provided PL D with reasonable agreement to RLD measured using the intercept method ( greater than R 2 > 0.79) at both study sites. There was close agreement in the validation set between RLD estimated using the intercept method and that predicted using the calibration eq uation developed from root scanning (R 2 ranged from 0.88 to 1.00 at CREC and 0.87 to 0.97 at SWFREC). Lowest coefficient of determination for roots greater than 3 mm in diameter was noted at CREC owing to few roots in this root category ascribed to the y oung tree age. Examination of equation slopes revealed that of the validation models developed for both sites, only the models

PAGE 129

129 for root sizes <0.5. 0.5 1.0, and 1.0 3.0 mm yielded close approximations of predicted RLD with slopes > 0.78 thus nearly a 1:1 ra tio, adequately explaining the variability in the validation and versus calibration sets. Slopes for root sizes >3 mm explained about 50 to 60% of the variability probably because there were very few roots >3 mm at both sites. The scanning method reduces the time required to measure RLD in samples. In th is study for example, it took about 40 to 60 hours to scan root samples compared with the line intercept method that took 140 hours (about 10 hours per day, though not statistically done across different s ets of individuals). Considering the importance of accurate root length estimation, the scanning method offers a worthwhile alternative especially when the researcher has a large number of samples. Several researchers approve photocopying (Collins et al. 1987; Kirchoff, 1992) and scanning (Collins et al., 1987; Bland and Mesarch, 1990) in RLD determination to achieve as much accuracy in the shortest time possible. RLD Distribution as a Function of Irrigation Method, Time and Soil Depth Root samples colle cted at SWFREC in June 2009 showed that lateral RLD distribution for CMP decreased from 0.374 cm cm 3 near the tree to 0.084 at 45 cm away from the tree row (Table 4 3 and Appendix A, Figure A 22 ) The irrigated zones of DOHS and MOHS showed RLD as high as 0.386 cm cm 3 and 0.279 cm cm 3 that decreased to 0.139 and 0.053 in the non irrigated zone respectively (Table 4 1, Appendix A, Figures A 23 and A 2 4 ) Small roots (<0.5 mm and 0.5 1.0mm in diameter) 15 and 15 30 cm depths while largest root s (>3 mm) contributed <3% of the total RLD. For RLD samples collected at SWFRE C in June 2010, fibrous roots (roots <0.5 mm and 0.5 1.0mm in diameter) contributed >77% of RLD while largest roots (>3 mm)

PAGE 130

130 accounted for <2 % of t otal RLD (Table 4 4 ) For CMP, 49% of the roots were found in the 0 15 cm depth while 36% and 15% of the roots were observed in the 15 30 cm and 30 45 cm soil depth. About 54%, 35% and 10% were found in the 0 15 15 30 and 30 45 cm depth using DOHS (Ta ble 4 4 ) Most of the roots (60%) were found in the 0 15 cm depth using MOHS while 23% and 17% of the roots were distributed in the 15 30 and 30 45 cm depth (Table 4 4) Roots (<0.5 mm in diameter) for CMP were uniformly distributed in the irrigated zone s averaging about 0.089 cm cm 3 while the roots greater 0.5 mm in diameter were largely found in the irrigated zone at 15 to 30 cm from the tree The lateral root distribution (0.5 mm) for DOHS averaged 0.33 cm cm 3 in the irrigated and non irrigated zones of 0 15 cm soil depth layer (Table 4 4) probably because drippers were increased from 4 to 8 drippers per tree and due to the rainy season which supplied water even in the nonirrigated zone. However at 15 30 cm and 30 45 cm soil depths, RLD for (root <0 .5 mm in diameter) was approximately 2 times that observed in the non irrigated zone while for roots greater than 0.5 mm in diameter the RLD was similar between the irrigated and non irrigated zones (Table 4 4 ). Root density es timated for MOHS for roots (<0.5 mm in diameter) was two times higher in irrigated than nonirrigated zone at all depths (Table 4 4 ) decreasing from 0.33 cm cm 3 in the irrigated zone to about 0.022 cm cm 3 in the nonirrigated z one. I n June 2010 at SWFREC, roots were uniformly distributed laterally in the grid around the tree using CMP with RLD ranging from 0.154 cm cm 3 to 0.086 cm cm 3 ( Figure A 25 ) Root length density using DOHS decreased from 0.439 cm cm 3 below the dripper to 0.170 cm c m 3 at 45 cm from the tree ( F igure A 2 6 ) S i milar pat ter n for MO H S w as no ted (F igure A 2 7 ) The root densities under the drip at SWFREC were lower than Lake Alfred because the

PAGE 131

131 drippers were moved from near the tree trunk to 30 cm away from the tree, when the number of lines was increased from one to two in March 2010 p rior to the June sampling. In Dec ember 2009, at the Lake Alfred site RLD was high in irrigated zones of DOH S Swingle and DOHS C35. For example, positions below the dripper showed RLD of about 0.8 cm cm 3 (DOHS Swingle) ( Appendix A, Figure A 2 8 ) and 0.86 cm cm 3 (DOH S C35 ) ( Appendix A, Figure A 2 9 ) which, respectively, decreased to 0.41 cm cm 3 and 0.091 cm cm 3 with distance away from the tree. The microsprinkler irrigation methods, CMP ( Appendix A, Figure A 3 0 ) and MOHS ( Appendix A, Figure A 31 ) yield ed 3 closer to the tree that decreased laterally to 0.074 cm cm 3 and 0.27 cm cm 3 respectively (Table 4 5) All irrigation methods but CMP showed that >60% of the roots were concentrated in top 0 15 cm than the 15 30 cm soil depth. However, CMP showed that 65% of the roots dominated the 15 30 cm soil depth. All treatments showed that fibrous roots (<0.5 mm and 0.5 1 mm) contributed to > 80% of the total RLD at both 0 15 and 15 30 cm soil depth suggesting that it is likely that yo ung trees will develop small, fine roots to promote water and nutrient uptake and accelerate tree growth. Largest roots (>3 mm) contributed <1.1% of total RLD. These results are similar to those reported by Morgan et al. (2007) on 2 to 5 year old Hamlin and Valencia orange trees. In agreement with our findings, they also reported that citrus trees develop a dense root system within the upper 30 cm where fibrous RLD distribution increases with depth and lateral distance as trees grow. Results suggest th at the further away from the tree, the less likely we are to find roots as shown in the significant decrease in RLD with distance from the tree. Thus,

PAGE 132

132 irrigation methods such as drip which apply water and fertilizer frequently and in small pulses within a limited root zone offer a viable option for increasing root water and nutrient uptake compared with the microsprinkler based systems when the trees are small. Positions below drippers tended to have high root density as exemplified in the DOHS C35 and DOH S Swingle where RLD ~0.8 cm cm 3 was close to 2 times that obtained in the irrigated zones of CMP or MOHS at the Lake Alfred site. Thus, this should typify the potential for increasing root density and subsequent tree uptake using drip irrigation. In July 2010, lateral root distribution showed that RLD decreased gradually with distance from the tree at the Lake Alfred site. For CMP and MOHS, RLD near the tree was about 0.25 cm cm 3 and 0.45 cm cm 3 and decreased to 0.10 cm cm 3 and 0.19 cm cm 3 respective ly at 30 cm from the tree ( Appendix A, Figures A 32 and A 3 5 Table 4 6) The drip fertigated treatments showed high RLD of about 1.0 cm cm 3 in the irrigated zone that decreased to 0.20 cm cm 3 in the non irrigated zone ( Appendix A, Figures A 3 3 and A 3 4) Also, the high RLD in the non irrigated zones for the OHS based fertigation methods was not expected. We ascribe the presence of roots in the non irrigated zone to the high rainfall in Florida (approximately 1400 mm) which probably increased the am ount of available water including in the non irrigated zone thus promoting root growth and development. Obreza and Pitts (2002) also observed similar phenomena on root density distribution between the irrigated and non irrigated zones of southwest Florida The results show RLD < 1.3 cm cm 3 consistent with findings of other researchers in citrus (Mattos et al., 2003), apple (De Silva et al., 1999) and somewhat lower than the RLD reported by Coleman (2007) in other woody species. Positions below the dripper or

PAGE 133

133 in irrigated zones, except for CMP (probably due to infrequent irrigation), had very high root length density. Zhang et al. (1996 and1998) also reported that RLD of fibrous roots was significantly greater near the emitter and at 0 15 cm deep layer for grapefruit trees. Similar results were obtained in citrus (Alva and Syvertsen, 1991; Mattos et al., 2003a; Morgan et al., 2007) and in drip irrigated woody species Coleman (2007). At the Lake Alfred site, small roots (<0.5 mm) accounted for > 87% of the t otal RLD at 0 15 cm soil depth using all irrigation methods and > 64% at 15 30 cm depth while largest roots accounted for the smallest portion ( < 2%) of the total RLD. All other treatments but grower practice showed high RLD in the top 0 15 cm where 58%, 67 % and 57% of the roots were concentrated using DOHS Swingle, DOHS C35 and MOHS, respectively. Only 36% of the roots were found in the 0 15 cm using the grower practice. Nappi et al. (1985) and Bassoi et al. (2003) found that highest grapevine root presen ce was within the top 40 cm and within 40 cm radius from the trunk for drip and at 0.8 1 m distance from the trunk using microsprinkler irrigation. In their studies, roots with diameter < 2 mm corresponded to at least 80% of total root length. Overall, roo t length density found in this study was higher for drip than microsprinkler irrigated citrus, which is similar in relation to results from Australia (Stevens and Douglas, 1994) and Brazil (Bassoi et al., 2003). The maximum RLD values reported by Stevens and Douglas (1994) were 1.2 and 0.6 cm cm 3 for drip and microsprinkler irrigated 8 yr old vines, respectively. Thus, our values, particularly, on the Ridge site are somewhat greater for the tree age (<3 yr old) and point to the intensive irrigation and f ertigation rates. The study of Stevens and Douglas (1994) also revealed that 47% and 40% of the roots were found in the top 0 40

PAGE 134

134 cm in a 0 160 cm profile of microject and drip irrigation. These observations confirm our postulated hypothesis that RLD would be greater using OHS based fertigation methods. Effect of Fertigation Method on Trunk Cross Sectional Area and Canopy Volume DOHS treatments using S wingle and C 35 rootstocks increased canopy volumes from 0.45 0.05 m 3 to 1.87 0.20 m 3 while MOHS increase d canopy volume from 0.400.12 m 3 to 1.730.19 m 3 beginning 11/12/10 to 07/15/1 1 at CREC. Thus, compared with grower practice, DOHS S wingle and DOHS C 35 increased canopy volumes by 47 to 112% for the same sampling period, while MOHS increased canopy volum es by 36 to 87% ( Fig. 4 1 ) Consistent results with DOHS, MOHS and CMP treatments were also observed at SWFREC in July 2010 and August 2011 where canopy volumes were increased by 15, 20 and 9%, respectively (Fig. 4 2 ). All three fertigation treatments at SWFREC, CMP, DOHS, and MOHS increased TCA by 97, 123 and 122% in year 2 and by 44, 56 and 66% in year 3. All TCA measurements were similar in August 2009, July 2010 and August 2011. The results revealed that DOHS and MOHS treatments promoted vigorous t ree growth across the years of study at the Lake Alfred site probably as a result of increased water uptake and nutrient accumulation as described in Chapters Three and Five. At SWFREC, DOHS and CMP increased tree growth in a similar pattern. Noteworthy in the study is the fertilization practice at the Lake Alfred site where granular fertilizer was applied quarterly in the CMP while at SWFREC it was fertigated monthly suggesting that use of fertigation practice for the CMP will promote tree growth and can opy development as shown in Figures 4 1 through 4 3 The caveat for monthly fertigation is that more fertilizer has to be applied due to a larger irrigated and fertigated

PAGE 135

135 area for CMP (360 o ) than MOHS and DOHS where there is a limited irrigated and fertiga ted area and a restricted root zone. Trunk cross sectional areas (TCAs) at the Lake Alfred site were similar in 2009 (Table 4 7). In July 2010, MOHS, DOHS S wingle and DOHS C 35 increased TCA by 31%, 38% and 51% over CMP. In March 2011, MOHS and DOHS C 35 increased TCA by 28% while DOHS S wingle increased TCA by 44%. growth rate resulting in canopy volumes and trunk cross sectional areas higher than at the Lake Alfred site and SWFREC. Our results at SWFREC, while supporting this hypothesis, showed that TCA and canopy volume for CMP were similar to DOHS and better than MOHS probably because it was fertigated. However, the results at SWFREC show that the annual percent increments in TCA and canopy volume were higher using MOHS and DOHS than the grower practice. Summary The chapter has shown the importance of root scanning in determining RLD in young citrus trees. There was good agreement between root length density and scanned area and shorter time for measuring root length with a flatbed scanner than using a line root area using a flatbed scanner can be calibrated using the line intersect method a nd used to predict root length with speed and greater precision Thus, root densities measured using the line intersection method showed strong and positive correlation (R 2 > 0.79) with those predicted by the calibration equation relating RLD and scanned ro ot area. The results show ed that use of the scanning method could be used to increase the accuracy and reduce the time for determination of RLD. Generally, root length density was highest in

PAGE 136

136 the 0 15 cm depth and decreased with depth and distance away fr om the tree. Positions below the dripper of DOHS and in the irrigated zones of MOHS showed root length density twofold higher than non irrigated zones and even greater RLD than the irrigated zones of grower practices Despite having irrigated zones aroun d the tree using CMP, the infrequent irrigation probably resulted in lower RLD compared to the irrigated zones of DOHS and MOHS at both study sites. Thus, the hypothesis that spatial root length density distribution w ould be greater in irrigated zones of M OHS and DOHS than conventional practice holds However root densities at Immokalee we re three to four times lower than those observed at the Lake Alfred site largely because of citrus greening that infected all trees in the grove during the second year o f the study and probably because of the spodic horizon found at 60 70 cm from the soil surface Also the number of drippers for drip irrigated trees was increased from four to 8 drippers per trees around March 2010, just two months before the June 2010 sa mpling at Immokalee such that the roots might have not fully developed below the dripper. The results will increase citrus growth rate resulting in canopy volumes and trunk cross sectional true at the Lake Alfred and Immokalee sites Our results at Immokalee while supporting this hypothesis, showed that TCA and canopy volume for CMP were similar to DOHS and better than MOHS probably because CMP was fertigated However, annual increments in TCA respectively by CMP, DOHS, and MOHS were 97, 123 and 122% in year 2 and 44 % 56 % and 66% in year 3 at Immokalee suggesting vigorous tree growth with ACPS/OHS

PAGE 137

137 Table 4 1. Models for RLD estimation at CREC Regressed variables Model type Root size (mm) 0 1 n R 2 RMSE (cm cm 3 ) P value Measured RLD vs. scanned area Calibration set <0.5 0.0270 3.78exp( 6) 64 0.88 0.1156 <0.0001 Predicted RLD vs. measured RLD Validation set <0.5 0.0194 1.02 64 0.94 0.0857 <0.0001 Measured RLD vs. scanned area Calibration set 0.5 1.0 0.0036 3.49exp( 6) 64 0.92 0.0086 <0.0001 Predicted RLD vs. measured RLD Validation set 0.5 1.0 0.0095 0.78 64 0.88 0.0092 <0.0001 Measured RLD vs. scanned area Calibration set 1.0 3.0 0.000 8 1.76exp( 6) 64 0.90 0.0103 <0.0001 Predicted RLD vs. measured RLD Validation set 1.0 3.0 0.0003 0.95 64 0.92 0.0080 <0.0001 Measured RLD vs. scanned area Calibration set >3.0 0.0003 2.57exp( 7) 64 0.79 0.0021 <0.0001 Predicted RLD vs. measured RLD Va lidation set >3.0 0.0003 0.51 64 1.00 0.0001 <0.0001 0 is the y intercept, 1 is the slope, n is the number of samples, R 2 is the coefficient of determination, RMSE is the root mean square error

PAGE 138

138 Table 4 2. Models for RLD estimation at SWFREC Regresse d variables Model type Root size (mm) 0 1 n R 2 RMSE (cm cm 3 ) P value Measured RLD vs. scanned area Calibration set <0.5 0.0116 2.19 exp( 6) 144 0.92 0.0350 <0.0001 Predicted RLD vs. measured RLD Validation set <0.5 0.0128 0.91 144 0.87 0.0455 <0.000 1 Measured RLD vs. scanned area Calibration set 0.5 1.0 0.0025 1.88 exp( 6) 144 0.94 0.0089 <0.0001 Predicted RLD vs. measured RLD Validation set 0.5 1.0 0.0026 0.98 144 0.93 0.0102 <0.0001 Measured RLD vs. scanned area Calibration set 1.0 3.0 0.0021 1.03 exp( 6) 144 0.81 0.0113 <0.0001 Predicted RLD vs. measured RLD Validation set 1.0 3.0 0.0023 0.88 144 0.91 0.0095 <0.0001 Measured RLD vs. scanned area Calibration set >3.0 0.0005 2.19exp( 7) 144 0.84 0.0028 <0.001 Predicted RLD vs. measured RLD Validation set >3.0 0.0005 0.58 144 0.91 0.0013 <0.0001 0 is the y intercept, 1 is the slope, n is the number of samples, R 2 is the coefficient of determination, RMSE is the root mean square error

PAGE 139

139 Table 4 3. RLD as a function of irrigation method soil depth and distance from the tree at SWFREC in June 2009 Irrigation method Soil depth Root diameter <0.5 mm Root diameter 0.5 1.0 mm Root diameter 1.0 3.0 mm Root diameter >3.0 mm cm cm cm 3 IRR NI IRR NI IRR NI IRR NI CMP 0 15 0.154 § 0.045 0.018 0.002 15 30 0.105 0.047 0.032 0.005 DOHS 0 15 0.205 0.170 0.065 0.062 0.017 0.027 0.001 0.002 15 30 0.203 0.078 0.073 0.058 0.052 0.043 0.011 0.003 MOHS 0 15 0.168 0.136 0.033 0.035 0.008 0.013 0.001 0.001 15 30 0.155 0.0 55 0.061 0.033 0.031 0.015 0.006 0.001 Statistics Irrigation method NS *** *** NS Depth *** NS *** *** Distance from the tree *** NS NS NS Irrigation method *Depth NS NS NS NS Irrigation method*Distance NS NS NS NS Depth*Distance NS NS NS Irrigation method*Depth *Distance NS NS NS NS IRR Irrigated zone, NI Non irrigated zone. We did not observe many roots >3 mm in diameter at CREC in December 2009 § for conventional practices, all the s ampled positions were irrigated, Statistics: NS N on significant difference, p<0.05, ** p<0.01, *** p<0.001

PAGE 140

140 Table 4 4. RLD as a function of irrigation method, soil depth and distance from the tree at SWFREC in June 2010 Irrigation method Soil depth Root diameter <0.5 mm Root d iameter 0.5 1.0 mm Roo t d iameter 1.0 3.0 mm Root d iameter >3.0 mm cm cm cm 3 IRR NI IRR NI IRR NI IRR NI CMP 0 15 0.148 § 0.025 0.010 0.002 15 30 0.089 0.025 0.017 0.003 30 45 0.031 0.016 0.007 0.001 DOHS 0 15 0.318 0.346 0.036 0.054 0.028 0 .039 0.005 0.003 15 30 0.216 0.104 0.064 0.058 0.054 0.036 0.011 0.009 30 45 0.039 0.028 0.037 0.028 0.014 0.013 0.001 0.001 MOHS 0 15 0.330 0.206 0.035 0.032 0.024 0.018 0.005 0.001 15 30 0.103 0.061 0.033 0.022 0.013 0.010 0.005 0.002 30 45 0.09 0 0.022 0.032 0.009 0.015 0.005 0.004 0.001 Statistics Irrigation method *** ** *** Depth *** ** *** *** Distance from the tree NS *** Irrigation method *Depth *** NS *** Irrigation method*Distance NS NS NS NS Depth*Distance NS NS N S NS Irrigation method*Depth *Distance NS NS NS IRR Irrigated zone, NI Non irrigated zone. § For conventional practices, all the sampled positions were irrigated. Statistics: NS Not significantly different, p<0.05, ** p<0.01, *** p<0.001

PAGE 141

141 Table 4 5. RLD as a function of irrigation method, soil depth and distance from the tree at the Lake Alfred site in December 2009 Irrigation method Soil depth Root diameter <0.5 mm Root diameter 0.5 1.0 mm Root diameter 1.0 3.0 mm Root diameter >3.0 mm cm cm cm 3 IRR NI IRR NI IRR NI IRR NI CMP 0 15 0.113 § 0.026 0.013 0.002 15 30 0.220 0.038 0.026 0.000 DOHS SWINGLE 0 15 0.777 0.404 0.041 0.072 0.029 0.017 0.000 0.000 15 30 0.407 0.261 0.048 0.061 0.044 0.044 0.003 0.003 DOHS C 35 0 15 0.811 0.347 0.044 0.060 0.003 0.031 0.002 0.000 15 30 0.323 0.077 0.019 0.008 0.010 0.005 0.002 0.000 MOHS 0 15 0.428 0.351 0.079 0.054 0.015 0.012 0.000 0.004 15 30 0.124 0.109 0.037 0.029 0.040 0.013 0.000 0.000 Statistics Irrigation method NS *** *** NS Depth *** NS *** *** Distance from the tree *** NS NS NS Irrigation method *Depth NS NS NS NS Irrigation method*Distance NS NS NS NS Depth*Distance NS NS NS Irrigation method*Depth *Distance NS NS NS NS IRR Irrigated z one, NI Non irrigated zone. We did not observe many roots >3 mm in diameter at the Lake Alfred site in December 2009 § For conventional practices, all the sampled positions were irrigated Statistics: NS Not significantly different, p<0.05, ** p<0.01, ** p<0.001

PAGE 142

142 Table 4 6. RLD as a function of irrigation method, soil depth and distance from the tree at the Lake Alfred site in July 2010 Irrigation method Soil depth Root diameter <0.5 mm Root diameter 0.5 1.0 mm Root diameter 1.0 3.0 mm Root diame ter >3.0 mm cm cm cm 3 IRR NI IRR NI IRR NI IRR NI CMP 0 15 0.184 NA § 0.016 NA 0.011 NA 0.0007 NA 15 30 0.120 NA 0.043 NA 0.023 NA 0.0021 NA DOHS SWINGLE 0 15 1.172 0.885 0.033 0.036 0.026 0.019 0.0098 0.0003 15 30 0.543 0.185 0.030 0.027 0.040 0.016 0.0069 0.0006 DOHS C 35 0 15 1.195 0.582 0.008 0.017 0.010 0.017 0.0042 0.0047 15 30 0.293 0.165 0.012 0.017 0.008 0.016 0.0003 0.0009 MOHS 0 15 0.564 0.487 0.026 0.038 0.023 0.019 0.0037 0.0017 15 30 0.218 0.083 0.022 0.027 0.018 0.029 0.0062 0.0003 Statistics Irrigation method ** *** NS *** Depth *** NS NS NS Distance from the tree *** ** NS ** Irrigation method *Depth ** *** NS *** Irrigation method*Distance NS Depth*Distance ** NS ** Irrigation method*Depth* Dista nce NS NS IRR Irrigated zone, NI Non irrigated zone. §NA Not applicable, the whole sampled area was irrigated under CMP Statistics: NS Not significantly different, p<0.05, ** p<0.01, *** p<0.001

PAGE 143

143 Figure 4 1 Canopy volume as a function of fertilization practice at the Lake Alfred site Error bars denote one standard deviation of 4 replications

PAGE 144

144 August 2009 CMP vs. DOHS NS CMP vs. MOHS DOHS vs. MOHS N 12 July 2010 CMP vs. DOHS NS CMP vs. MOHS DOHS vs. MOHS N 60 August 2011 CMP vs. DOHS NS CMP vs. MOHS NS DOHS vs. MOHS NS N 60 indicates significance at p<0.05; NS indicates non significant differences; CMP Conventional microsprinkler practice, DOHS Drip open hydroponics system, MOHS Microsprinkler open hydroponics system Figure 4 2 Trunk cross sectional area as a function of fertigation practice at the Immokalee site Error bars denote one standard deviation

PAGE 145

145 July 2010, August 2011 CMP vs. DOHS NS CMP vs. MOHS NS DOHS vs. MOHS NS N 60 Mean one standard deviation, NS Not significant, p<0.05, ** p<0.01 CMP Conventional microsprinkler practice, DOHS Drip open hydroponics system, MOHS Microsprinkl er open hydroponics system Figure 4 3 Canopy volume as a function of fertigation method at the Immokalee site

PAGE 146

146 Table 4 7. Trunk cross sectional area as function of fertigation method at the Lake Alfred site Fertigation method TCA December 2009 TCA Ju ly 2010 TCA March 2011 cm 2 DOHS Swingle 4.340.73 8.990.66 14.440.94 CMP 4.371.66 6.510.61 10.021.55 MOHS 4.230.57 8.550.81 12.842.04 DOHS C 35 4.710.56 9.820.39 12.831.50 Significance § NS ** ** Mean one standard deviation, CMP Conve ntional microsprinkler practice, DOHS Swingle Drip open hydroponic system with Hamlins on Swingle rootstock, DOHS C 35 Drip open hydroponic system with Hamlins on C35 rootstock, MOHS Microsprinkler open hydroponic system, § NS Not significant, ** p<0.01 T CA Trunk cross sectional area

PAGE 147

147 CHAPTER 5 EFFECTS OF IRRIGATIO N METHOD AND FREQUEN CY ON CITRUS WATER UPTAKE AND SOIL MOIS TURE DISTRIBUTION Accurate estimation of plant water use could improve irrigation management (Gutierrez et al. 1994; Morgan et al. 2006 b ) leading to a better understanding of plant water interactions (Ham et al. 1990; Gutierrez et al. 1994). Plant water use typically called crop evapotranspiration (ET c ) can be determined with the stem heat balance (SHB) method. The SHB technique has been found to be reasonably accurate and dependable in estimating plant water use in pecan (Steinberg et al. 1990a, b), citrus (Steppe et al. 2006), Anacardium excelsum (Meinzer et al. 1993), cotton (Ham et al. 1990), coffee and koa (Gutirrez and M einzer, 1994; Gutirrez et al. 1994) and grapevines (Lascano et al. 1992; Heilman et al. 1994). The SHB approach provides a reliable method for measuring sap flow in the stems of herbaceous plants that is sufficiently accurate for application in many a gronomic and biological applications (Baker and van Bavel, 1987; Baker and Nieber, 1989). Using the SHB method, sap flow rates in trees have been found to be within 4 to 10% of transpiration loss (Baker and Nieber, 1989; Steinberg et al. 1989; Lascano et al. 1992; Devitt et al. 1993). Dugas et al. (1994) also showed that cumulative sap flow for 14 day periods was similar to cumulative evapotranspiration or transpiration calculated from a water balance in cotton. SHB technique has several advantages ov er other methods for measuring water use such as lysimetry and water balance. The technique is non intrusive, does not require calibration, responds quickly to plant water flow, can be used over long periods of time without damage to the plant (Baker and van Bavel, 1987; Steinberg et al. 1989; Gutierrez et al. 1994) and is simple to use with an appropriate digital datalogger (Baker and van Bavel, 1987).

PAGE 148

148 Using reference evapotranspiration (ET o ), ET c can be accurately determined once a crop coefficient (K c ) and soil moisture depletion coefficient (K s ) are known (Allen et al. 1998). K s can be determined through periodic soil moisture measurement at selected depths of the plant root zone (Morgan et al. 2006 b ; Fares et al. 2008). K c is defined as the rati o of crop evapotranspiration (ET c ) to potential evapotranspiration (ET o ) when soil water availability is non limiting and is a function of crop type, climate, soil evaporation and crop growth stage (Allen et al., 1998; Morgan et al., 2006; Fares et al., 20 08). Several studies using water balance and drainage lysimeter methods estimated that K c values of citrus trees range from 0.6 in the fall and winter to 1.2 in the summer (Rogers et al. 1983; Boman, 1994; Martin et al., 1997, Fares and Alva, 1999, Morga n et al., 2006 b ). Jia et al. (2007) found that K c values may vary from location to location. For example, they found that annual average K c values were higher for the citrus grown in the Ridge regions (K c =0.88) than for the Flatwoods (K c = 0.72) in Flori da, with monthly recommended values ranging from 0.70 to 1.05 for the ridge and from 0.65 to 0.85 for the Flatwoods citrus, respectively. They attributed the differences due to water logging in the root zone of the Flatwoods citrus owing to water table due to the presence of the spodic and/or argillic horizon. In studies on citrus K c from other regions, different values have been reported depending on climate and method used. Values ranging from 0.80 to 0.90 have been reported using the water balance techn ique (Allen et al. 1998). For navel orange tree groves in California, Consoli et al. (2006) found that K c values ranged from 0.45 to 0.93 using an energy balance method. Rana et al. (2005), using the eddy correlation method, found that K c values ranged f rom 0.8 to 1.2, corresponding to citrus

PAGE 149

149 phenological growth stage and the effects of high wind speed and high vapor pressure deficit. Many studies in Florida on citrus tree water use have used other methods such as lysimetry water balance and the Florid a Automated Weather Network (FAWN) to estimate tree water use in citrus trees in the field without partitioning evaporation and transpiration from the ET component (Rogers et al., 1983; Boman, 1994; Obreza and Pitts, 2002; Jia et al., 2007; Morgan et al., 2006b; Fares et al., 2008) We attempted to estimate tree water use using the SHB technique to calculate K c values basing on plant transpiration and Leaf Area Index (LAI) and correlate the two with root length density (RLD) and canopy volume. Water use t hrough hourly and daily sap flow measurements would help in accurately predicting transpiration and devising ways of minimizing evaporation and percolation losses by synchronizing irrigation applications with peak tree water use. According to Morgan et al. (2006 b ), estimation of soil water uptake and resulting soil water depletion would allow for a more accurate assessment of soil water depletion, crop water uptake and soil moisture storage capacity. The hypotheses postulated were : 1) citrus water use inc rease s with canopy volume and root length density in situ irrespective of the irrigation frequency and fertigation method and, that, 2) soil water content will be greater using the drip and microsprinkler OHS than grower practice. The objectives of the st udy were to : 1) determine ET c and K c using SHB method on 1.5 and 4 year old citrus using three different irrigation methods and fertigation frequencies on Florida Spodosol and Entisol; 2) determine soil water distribution in the citrus irrigated root zon e.

PAGE 150

150 Materials and Methods Experimental design and irrigation methods A randomized complete block design consisting of three treatments at Immokalee s ite and two to four treatments at the Lake Alfred site was used, with three to four trees serving as replic ations. The irrigation treatments were applied to the replicate trees independently within a row. The irrigation treatments were as follows: (1) Conventional practice (CMP) irrigated weekly, with the microsprinkler placed at about 10 15 cm perpendicular to the tree; (2) Drip OHS (DOHS) irrigated daily in small pulses, with two drip lines spaced at 30 cm from the tree, each delivering four emitters on each side of the tree; (3) Microsprinkler OHS (MOHS) irrigated daily, with the microsprinkler placed at about 15 cm perpendicular to the tree. All the treatments were replicated four times The treatments imposed at Lake Alfred were similar to the set up at Immokalee s ite except for the modification to DOHS that had one drip line placed within the tree row, with one dripper placed at 15 cm on each side of the tree. The DOHS was imposed on both Swingle and C 35 rootstocks. Drip irrigation was provided with integral Uniram (Netafim) pressure compensating drip emitters (Netafim, Fresno, CA) (2.00 L h 1 ). At both sites, microsprinkler irrigation was provided with either a single 40 L h 1 Max 14 (Maxijet, Dundee, FL) fill in blue emitter for CMP or a 29 L h 1 Max 14 fill in orange emitter for MOHS at each tree ( Schumann et al., 2009; 2010). Estimation of Soil Moist ure Soil water sensors on Candler sand (VG400, Vegetronix, Sandy, UT) and Immokale e sand (RS 485, Portland, OR), using the capacitance method (Katul et al., 1997; Morgan et al.; 1999; 2002 ) of estimating volumetric water content were used to measure moistu re to determine treatment effects on soil water status. Soil moisture was

PAGE 151

151 measured every 30 minutes at 10 cm and 45 cm depths on Candler sand and 10 20 30 40 and 50 cm depths on Immokalee sand using capacitance probes and an automated logging system Volumetric water content was measured (%) (Hillel, 1998). Rainfall data and other climatic variables were collected from F AWN stations at Southwest Florida Research and Education Center ( SWFREC ) and Citrus Research and Education Center ( CREC ) ( http://fawn.ifas.ufl.edu/ ) ( Apprendi x E ). Estimation of Crop Water Uptake and K c Actual transpiration was measured on tree trunks or branches with a heat balance method using Dynagage Flow32 1K Sap Flow System t o evaluate tree water use. The direct transpiration readings were taken from July 2010 March 2011 and August to September 2011 at the Lake Alfred site and February, March and June 2011 at SWFREC K c was estimated for each site using the measured citrus t ranspiration and calculated reference ET o from FAWN data. Water uptake was meas ured using sap flow sensors ( Dynamax Inc. Houston, TX) on branches of four random trees per treatment (each tree serving as a replicate) at SWFREC. At SWFREC, four healthy tr ees per treatment were randomly selected to serve as replicates in the measurements. At Lake Alfred, due to limitation in the size of sensors, sap flow measurements on trunks of six trees were taken on Drip OHS (DOHS S wingle ) and Conventional microsprink ler practice (CMP) Prior to installation of the sensors, measurements were taken of branch and trunk diameter. Also, critical measurements of variables that characterize water use in citrus such as Leaf Area Index (LAI) and canopy volume were determined using a Leaf Area Meter and measuring tape. In the study, we used the Dynamax Flow32 1K sap flow system (Dynamax Inc., Houston, TX) with CR1000 data logger, including PC400 data logger support software ( Campbel l Scientific Inc Logan,

PAGE 152

152 UT). Trees at Lake Alfred had trunk diameters ranging from 24.92 mm to 31.40 and leaf area index ( LAI ) was a bout 1.770. 71 in July 2010. We used the gauges of SGB25 model at the Lake Alfred site on trunks in July 2010 because the tree trunks were greater than 24 mm in diam eter In the subsequent seasons, the following sizes of sensors were used: SGA13 ws, SGB16 ws, SGB19 ws and SGB25 ws for respective stem diameter ranges of 12 16, 15 19 18 23 and 24 32 mm. The thermocouple gaps specified were: 4.0 for SGA13, 5.0 for both SGB16 and SGB19, and 7.0 for SGB25. Tree canopy volumes were estimated by measuring the average canopy diameter using canopy width in the east west and north south directions and canopy height using the formula for a sphere= r 3 where r is the canopy radius. Trunk diameter was estimated from averaging the diameter in the east west and north south directions and 2 where r is the trunk radius. We adapted the approach for determin ing sap flow measurements from individual plants recommended by Lascano et al. (1992). Water use for trees was determined from measurement of sap flow in limbs by increasing measured sap flow by the proportion of leaf area of the measured limb over the lea f area of the entire tree. The m ean transpiration, was estimated by normalizing the stem flow data on a population per land area basis as: ( 5 1 ) Where is E sap =daily value of sap flow per unit land area (mm d 1 ), M=sap flow per plant (kg d 1 ), P=plant population m 2 water is water density, 1000 kg m 3

PAGE 153

153 The index sapflow crop coefficient, K c, was estimated follow ing the equation below: ( 5 2 ) where ET c is daily crop evapotranspiration (mm d 1 ), ET o is reference evapotranspiraton (mm d 1 ), E sap is the daily value of sap flow per unit land area (mm d 1 ), K s is soil water stress coefficient. Thus, assuming no water stress due to the automated irrigation, K s becomes unity. s at the Lake Alfred site (z 1 =0 cm and z 2 =4 5 cm) and Immokalee site (z 1 =0 cm, z 2 =10 cm, z 3 =20 cm, z 4 =30 cm, z 5 =40 cm, z 6 =50 cm) 1 t 2 ; i.e., 1 day was used) was calculated based on measured water content readings by the capacitance probes using the following equation already formulated by Fares and Alva (2000 b ) : ( 5 3 ) Results and Discussion Tree characteristics at Immokalee and Lake Alfred To determine leaf area in spring 2011, we categorized leaves at each site by size and measured the leaf area, length and width Leaf areas for small, medium and large leaves averaged 10.7 4.2, 28.7 8.4 67.4 16.8 cm 2 at Immokalee. Leaf areas for small, medium and large leaves averaged 15.1 4.7 36.2 8.0 68.9 16.0 cm 2 at Lake Alfred ( Table 5 1 ) T ree c anopy volumes ranged from 4.40 0.98 to 7.04 0.80 m 3 in February

PAGE 154

154 2011 and from 6.67 1.30 to 9.32 1.10 in June 2011 at Immokalee ( Table 5 2 ). At the Lake Alfred site, t ree canopy measurements showed that canopy volumes ranged from 0.90 0.20 to 1.42 0.32 m 3 in Jul y 2010 2.81 0.73 to 4.89 0.58 in March 2011 and 4.45 0.45 to 6.53 0.88 in August 2011 ( Table 5 2 ). Trunk cross sectional areas varied from 19.32 4.35 to 27.002.14 cm 2 in February 2011 and 25.723.92 to 31.512.32 cm 2 in June 2011 at Immokalee Trunk cros s sectional areas varied from 5.59 1.17 to 7.06 0.26, 10.02 1.55 to 14.44 0.9 4 cm 2 in March 2011 and 18.19 2.01 to 25.59 1.94 cm 2 in August 2011 at Lake Alfred ( Table 5 2 ) To estimate, leaf areas in later sap flow studies, we developed calibration equatio ns as shown in Figures 5 1 and 5 2. Water U ptake at Immokalee and Lake Alfred On average most days, we observed no sap flows in the two treatments at 0, 1, 6, 7, 8, 22 and 23 h as exemplified in Figure 5 3 in July 2010. Peak sap flow readings were noted b etween 10 and 20 h, ranging from 134 to 220 g h 1 under DOHS S wingle Sap flow readings under CMP peaked between 11 and 19 h, ranging from 110 to 133 g h 1 In March 2011, average hourly sap flows peaked at around 1100 h and 1200 h. On March 17, 2011, fo r example, peak sap flows recorded were 298, 329, 519 and 336 g h 1 for DOHS Swi ngle (at 1400 h), CMP and DOHS C 35 (at 1300 h), and MOHS (at 1200 h), respectively. On average hourly sap flow in March ranged from 19435 to 385152 g h 1 11736 to 29733 g h 1 17632 to 27646 g h 1 and from 15426 to 24846 g h 1 for DOHS C35, MOHS, DOHS S wingle and CMP (Figure 5 3) Similar to observations in July 2010, we noted that sap flows in spring 2011 also showed consistently high readings (>100 g h 1 ) between 10 00 h and 1800 h, probably due to increased solar radiation (averaging 239 and 254 W m 2 in July 2010 and March 2011) and temperature (25 30 o C in July 2010 and 11 22 o C March 2011) compared with the rest of day.

PAGE 155

155 Similar hourly sap fl ows among the treatmen ts using S wingle rootstock (DOHS Swingle, MOHS, CMP) were noted regardless of the fertigation method and irrigation frequency (F igure 5 3). However, DOHS on C 35 rootstock, had pea k hourly sap flows that were 58% to 61 % higher than CMP. Lowest daily sap fl ow was approximately 2.05 kg d 1 on July 12, 2010 while the highest sap flow was about 4.74 kg d 1 on July 18, 2010 under DOHS Swingle with mean daily sap flow readings averaging 3.960.74 g d 1 (Figure 5 4). Using CMP, the maximum and minimum values of t he average daily sap flows were about 3.83 kg d 1 and 1.71 kg d 1 on July 10 and July 25, 2010, respectively, with a mean of 2.750.59 kg d 1 (Figure 5 4). There was very high variability in the daily readings of grower practice as shown in Figure 5 4 whil e consistently high readings with less variability were observed using DOHS Swingle. On average, the sap flow was 44% higher under DOHS Swingle than CMP. It appears the trees under grower practice also had some reasonable variability in trunk cross secti onal area (Table 5 2). Lascano et al. (1992), in their study on grapevines, explained that such variability among trees exists and can be reduced by normalizing the total sap flow by leaf area. Furthermore, i n March 2011, all DOHS treatments on Swingle a nd C35 rootstocks showed sap flows greater than CMP by 7 to 150%. Sap flow for MOHS was higher than CMP on all days except on Julian d ays 77 and 78 when daily sap flow was 6% less than CMP suggesting that significant gains in water uptake on the ridge soi l lie with drip OHS (Figure 5 5) In March 2011, daily transpiration readings were lowest using CMP on March 11, 2011 and peaked on March 23, 2011 using DOHS C 35 ( Figure 5 5) Lowest average daily sap flows were observed on 03/11/2011 at the beginning o f the study. For example, sapflows for CMP,

PAGE 156

156 DOHS S wingle, DOHS C 35 and MOHS averaged 1.190.18, 2.7 71.42, 2.961.95 and 2.770.90 kg d 1 respectively. Average daily sap flows for the all the treatments but CMP at Lake Alfred peaked on 03/23/2011. CMP s howed a peak average sap flow of 4.840.94 kg d 1 on 03/18/2011. Peak sap flows recorded for DOHS S wingle DOHS C35 and MOHS were 5.731.42, 8.987.28 and 4.712.37 g d 1 Average hourly and daily sap flows readings for studies conducted at SWFREC in Febr uary and March 2011 are given in Figure s 5 6 and 5 7 For MOHS and CMP, we noted very low sap flow readings. The statuses of sensors reportedly ranged from 5 7 showing faulty readings that were identified late in the study. DOHS averaged hourly sap flow peaked to 1 361 g h 1 at 1100 h on February 27, 2011. MOHS and CMP peaked to about 210 g h 1 on February 27 at 1000 h. The data logger used for MOHS and CMP showed no readings most of the time resulting in the extremely low readings W e had to get this fixed at the end of the experiment. Thus, the readings for DOHS might actually represent the SWFREC site. Daily sap flow peaked to 21.6 kg d 1 on March 3, 2011 using DOHS. As indicated above, we also observed very low readings for MOHS with maximum dail y sap flow of 1.38 kg d 1 and for CMP where maximum daily sap flow was 1.09 kg d 1 Minimum daily sap flow readings were 213 and 220 g d 1 for MOHS and CMP, respectively. Average hourly sap flows (Figure 5 8 ) in June 2011 at SWFREC was high between 1000h and 1600h in all the three fertigation methods peaking to respective values of 3.55, 2.27 and 1.77 kg h 1 for DOHS, MOHS and CMP at 1600 h, 1400 h and 1300 h, respectively. Hourly sap flows peaked between 1000 h and 1900 h for DOHS and CMP and between 1000 h and 2000 h for MOHS.

PAGE 157

157 The average hourly sap flows at the Lake Alfred site in August September 2011 (Figure 5 9 ), peaked between 1000 h and 1700 h. Sap flows followed the pattern DOHS S wingle >MOHS> CMP>DOHS C35. DOHS S wingle MOHS and CMP peak ed to 2.53 kg h 1 1.62 kg h 1 and 1.19 kg h 1 at 1400 h and DOHS C35 peaked to 0.85 kg h 1 at 1300 h. The trees for the S wingle rootstock, including the grower practice (see canopy volumes in Table 5 2), had grown so much in f all 2011 compared with Marc h 2011 at the Lake Alfred site with trunk cross sectional area increments of 66, 77 and 94% and canopy volume increments of 49, 34 and 90% for the MOHS, DOHS and CMP respectively The grower practice, CMP, had the largest increase in canopy volume and si gnificant increase in leaf area, pro bably due to the use of controlled release fertilizer in summer 2010 and 2011. The tree size for C 35 rootstock did increase by only 16 and 42% in canopy volume and trunk cross sectional area suggesting a small increase in leaf area DOHS hourly sap flow was well above the other two fertigation methods in June 2011 at SWFREC. Daily sap flows (Figure 5 10 ) peaked in the following order: DOHS > 33 .31.7 kg d 1 33.816.6 and 23.510.6 kg d 1 and 26.614.2 kg d 1 and 14.511.4 kg d 1 All sap flows for DOHS ranged from 87 to 160% while for MOHS daily sap flow were 10 to 103% greater than CMP. In August September 2011, d aily sap flow averaged 35, 27, 14 and 13 kg d 1 for DOHS S wingle MOHS, DOHS C35, and CMP suggesting increments by 176%, 130% and 16% over CMP at the Lake Alfred site As explai ned above we expected much higher sap flows for DOHS C35 but a small increase in tree size and, probably leaf area

PAGE 158

158 compared with the other irrigation methods resulted in lower sap flow values compared with observations in March 2011 (Figure 5 11) Daily s ap flows per unit land area i n July 2010 ranged from 0.11 to 0.25 mm d 1 using DOHS S wingle and 0.09 to 0.21 mm d 1 using CMP with respective averages of 0.21 and 0.15 mm d 1 ( Appendix A, Figure A 3 6 ). Sap flows for the Swingles ranged from 0.20 to 0.31 mm d 1 and from 0.28 to 0.39 mm d 1 for DOHS C35 in March 2011 at the Lake Alfred site ( Appendix A, Figure A 37 ) The sap flow in February March 2011at Immokalee averaged 0.81 mm d 1 peaking to 1.06 mm d 1 on Julian day 61 ( Appendix A, Figure A 38 ) In Jun e 2011 at Immokalee daily sap flow averaged 2.3, 1.4 and 1.1 mm d 1 for DOHS, MOHS and CMP ( Appendix A, Figure A 39 ) In August September 2011 at the Lake Alfred site daily sap flow ranged from 1.030.67 to 2.802.21 mm d 1 0.230.08 to 1.110.42 mm d 1 0.740.03 to 1.970.28 mm d 1 and 0.620.21 to 1.380.64 mm d 1 for DOHS Swingle CMP, MOHS and DOHS C35 ( Appendix A, Figure A 40 ) Large canopies, leaf areas and increased temperatures (averaging 26 o C at both Immokalee and the Lake Alfred site ) accou nted for better uptake in the OHS fertigation methods than grower practices at both sites. Cumulative sap flows at the Lake Alfred site on the studies undertaken between Julian d ays 190 209 in 2010, 70 82 and 236 251 in 2011 showed that DOHS S wingle had cumulative sap flow of 4.3 mm on day 209 while cumulative sap flow of CMP was 3.0 mm representing percent increase in sap flow in DOHS S wingle of 20 to 56% over CMP between Julian d ays 190 and 209 ( Appendix A, Figure A 41 ) Cumulative sap flows were 43%, 35% and 80% higher than CMP for DOHS S wngle MOHS and DOHS C35 representing very high uptake using ACPS fertigation compared with conventional

PAGE 159

159 irrigation practice between Julian d ays 70 and 82 ( Appendix A, Figure A 42 ) The cumulative sap flows were 3.34, 2.58, 3.01 and 4.75 mm for DOHS S wingle CMP, MOHS and DOHS C35 between Days 70 and 82. In August September 2011, the trees had increased in trunk cross sectional area, canopy volume and leaf area resulting in cumulative sap flows that were 166%, 141% an d 65% higher than CMP using DOHS S wingle MOHS and DOHS C35, respectively. The cumulative sap flows on Julian Day 251 peaked to 30 11, 23 and 17 mm using DOHS Swingle CMP, MOHS and DOHS C35 ( Appendix A, Figure A 43 ) In March 2011 at Immokalee DOHS S wi ngle peaked from 0.77 mm on Julia n Day 48 to 11.31 mm on day 61 ( Appendix A, Figure A 44 ) The cumulative sap flows of 44 mm and 27 mm using DOHS and MOHS in June 2011 representing, on average, 115% and 37% higher sap flows than CMP, underlining the import ance of frequent fertigation as also shown on the ridge site ( Appendix A, Figure A 4 5 ) Index sap flow K c averaged 0.0290.014 and 0.0420.003 using CMP and DOHS S wingle respectively at the Lake Alfred site in J uly 2010 ( Appendix A, Figure A 46 ) incre asing to 0.060.01 and 0.080.02 in March 2011 ( Appendix A, Figure A 47 ) In March 2011, K c values for MOHS and DOHS C35 were 0.070.04 and 0.110.09 ( Appendix A, Figure A 47 ) The average K c peaked in August September ranging from 0.21 0.06 to 0.57 0.4 3 with high K c observed in the OHS irrigation methods compared with grower practice probably because of frequent irrigation, vigorous tree growth and large canopies ( Appendix A, Figure A 4 8 ) At SWFREC sap flow K c ranged from 0.250.10 to 0.340.15 in February March 2011 ( Appendix A, Figure A 49 ) The K c in June 2011, ranged from 0.300.11 to 0.540.26, 0.210.09 to 0.340.14 and 0.130.10

PAGE 160

16 0 to 0.250.15 using DOHS, MOHS, and CMP, respectively, suggesting that water uptake followed the order DOHS>MOHS>CM P (Figure A 5 0 ) The sap flow K c for the <2 .5 yr old trees at the Lake Alfred site suggests that transpiration accounted for about 3 to 1 0% of the actual evapotranspiration because the trees were young with small canopy volume s ( ranging from 0.71 to 1.72 m 3 ) and leaf area (LAI ranged from 1.230.42 to 2.300.49) and thus had little ground cover. S oil evaporation tends to account for the greatest part of actual transpiration for a uniformly wetted surface not covered by the canopy (Testi et al., 2004) Wit h tree s getting ol der ~3 years or older, the transpiration component, as expected, increased and accounted for about 25 to 70% of the actual evapotranspiration. This is because citrus K c for Florida conditions ranges from 0.6 in the fall and winter to 1.2 in the summer ( Rogers et al. 1983; Boman, 1994; Fares and Alva, 1999; Morgan et al., 2006 b ; Jia et al., 2007) and water use tends to increase with age and increase in canopy volume (Morgan et al. 2006 b ). It is important to assess actual tree water use fo r proper irrigation scheduling and planning because, depending on tree age, water may need to be applied in the actual root zone for tree uptake as was the case with the OHS treatments. The sap flow K c values in June/July and August/September (for trees>3 yr old) are close to or slightly lower than many crop c oefficients from other regions that included the evaporation component ( Hoffman et al. 1982; Castel et al. 1987; Sepaskhah and Kashefipour, 1995; Martin et al., 1997; Consoli et al., 2006 ; Petillo an l 2007 ) or split the evaporation and transpiration components (Villalobos et al., 2009) Our study focused on trees <5 yr old young trees while the studies from the other regions above focused on trees > 7 yr old mature trees. Rogers et al. (1983) explained that frequent rain s in

PAGE 161

161 Florida produce wet soil and leaf conditions that result in the actual ET being a great percentage of the potential ET than is true for semi arid or arid conditions of California, USA (Consoli et 1982), Arizona, USA (Martin et al., 1997), Iran (Sepaskhah and Kashefipour, 1995) Japan (Yang et al. 2003; 2010) a nd Spain (Castel et al., 1987; Testi et al., 2004; Villalobos et al., 2009). This suggest s that, ceteris paribus, Florida does have high water evaporative demand due to the hot humid climate and the deep drainage ascribed to the sandy soil characteristic. Soil moisture distribution at Lake Alfred and Immokalee The soil moi sture distribution pattern showed that there was ample soil moisture in the root zone in all the treatments in July 2010 at the Lake Alfred site For example, average soil moisture measurements as shown by time of day using grower practice (CMP) show that maximum soil moisture content was 8.42 % (at 17.0 h) and minimum of 6.31% (at 15.8 h) at 10 cm soil depth layer and a maximum of 9.02% (at 18.0 h) and a minimum of 6.88% (at 15.8 h) at the 45 cm soil depth layer. MOHS yielded maximum soil moisture of 13.65 % (at 9.8 h) and a minimum of 8.98% (at 8.0 h) at10 cm soil depth and 12.02% (at 17.5 h) and 11.26% (at 9.3 h) at 45 cm soil layer. The maxima and minima soil moisture using DOHS C35 were 19.50% (at 8.5 h) and 9.99% (at 13.0 h) at 10 cm soil depth and 12.4 3% (at 17.5 h) and 11.65% (around 8.3 8.8 h) at 45 cm soil depth (Figure 5 12 ) Daily soil moisture at 10 and 45 cm soil depths averaged 7.61.6 and 8.20.9 % 10.23.2 and 11.50.8 % and 11.84.5 and 12.00.4% ( Figures 5 13 and 5 14 ) using CMP, MOHS and DO HS C35. Lower average soil moisture content at 10 cm than 45 cm suggests water removals either through tree uptake, soil evaporation or downward drainage.

PAGE 162

162 In March 2011 at the Lake Alfred site t he soil moisture peaked to around 8.22, 12.08 and 14.79 % be tween 7.30 am and 8.30am at 10 cm depth, de creased to 6.44, 10.86 and 7.18 % in the afternoon and at nigh t in the respective treatments CMP, MOHS, DOHS C35 (Figure 5 15 ) At 45 cm soil depth soil moisture was higher than the upper top 10 cm soil layer (Fig ure 5 15 ) probably due to downward drainage D aily soil moisture averaged 7.3, 11.3 and 10.5% at 10 cm depth (Figure 5 16 ) and 10.5, 12.3 and 7.6% at 45 cm depth (Figure 5 17 ) using CMP, MOHS and DOHS C35 irrigation treatments in March 2011 Our own result s in Chapter 4 and those of Zhang et al. (1996) confirm that tree uptake should be greater in the 0 15 cm soil layer than lower horizons owing to high root density in the range of 55 67% on length basis (this study) and 70 75% on weight basis (Zhang et al. 1996) in the top 15 cm. Our observations are also supported by earlier studies ( Go l dberg et al., 197 1 ; Alva and Syvertsen, 1991; Khan et al. 1996; Alva et al., 1999; Fares and Alva, 2000 a, b ; Badr, 2007; Davenport et al., 2008; Badr and Abuarab, 2011). Khan et al. (1996) showed that soil water content increased up to 25 cm depth and 30 cm radial distance at application rates ranging from 1.5 2.5 L h 1 and input concentration falling between 100 and 500 mg L 1 on coarse loamy soil. They also showed that s olute concentration increased with high input concentration, applied volume and application rate up to about the same depth (~25 cm) and radial distance (~30 cm) as for soil water content. Davenport et al. (2008) further observed that soil moisture distrib ution for drip irrigated vineyards was adequate in the 0 45 cm depth and within 20 40 cm radius, either diagonal or perpendicular to the drip line. Our observations are also supported by Goldberg et al. (197 1 ) who concluded in their study that soil moistur e resulting from drip irrigation was two dimensional, with

PAGE 163

163 moisture contents high along and beneath the row and decreasing laterally. Thus, according to Goldberg et al. (197 1 ) the effect of shorter irrigation intervals, as was the case with drip and micros prinkler ACPS/OHS, with proportionally smaller amounts of water applied in a single irrigation, is to decrease the variations in moisture content in the root zone and establish a continuously higher moisture regime. Eventually, drip that was developed to c onserve water in arid environments (Goldberg and Shmueli, 1970) has been adapted to semi arid and humid regions to manage water in sandy soils with high conductivity and supplement water where rainfall is inadequate or is not uniformly distributed througho ut the year. In August and September 2011, a contrary soil moisture distribution trend was noted. The moisture content averaged 12.63 and 10.94% (CMP), 10.88 and 8.06% (MOHS) and 1 1.55 and 9.32% (DOHS C35) at 10 cm and 45 cm depth layers, respectively, suggesting that soil moisture decreased with depth probably because of the frequent rainfall that kept the top 10 cm layer wet throughout the study period (Figures 5 18 5 19 and 5 20 ) On Immokalee sand, the DOHS soil moist ure varied between 7.5 and 10.0 % in the top 10 30 cm and remained between 5 and 6.5 % at 40 and 50 cm depths in February and March 2011 (Figure 5 21 ) In June 2011, the moisture c ontents ranged from 7.5 to 12.0 % in the top 30% and between 6.5 and 7.7 at 40 and 50 cm soil depths (Figure 5 22 ). The soil water at Immokalee using MOHS ranged from 8.5 to 14% and around 6 to 8% in the 40 to 50 cm soil depths in February March 2011 (Figure 5 23 ) and June 2011 (Figure 5 24 ). The grower practice had soil moisture contents varying between 8 and 1 3 % in the top 20 cm, and between 6 and 7% in the 30 50 cm soil depth layers in

PAGE 164

164 February March 2011 (Figure 5 25 ). In June 2011, the soil moisture varied from 10 20% in the top 20 cm and ranged from 6 to 13% in the lower 30 50 cm soil depth (Figure 5 26 ). T he lower soil moisture contents in the lower 30 50 cm depth suggests that probably root water extraction in the top 30 cm resulted in less water percolating to lower soil depth layers. This might hold because the Immokalee sand has a shallow water table (O breza and Pitts, 2002) that limits root development in the top 30 cm (Bauer et al., 2004). Factors affecting water uptake on the two soils Linear and nonlinear analysis revealed the major factors controlling cumulative water uptake for young citrus trees a t Lake Alfred and Immokalee sites. In July 2010, when the trees at the Lake Alfred site were fairly small (<2 yr old) with small canopies (<1.74 m 3 ), cumulative water uptake was largely a function of trunk cross sectional area (R 2 =0.98, p<0.001) and canopy volume (R 2 =0.67, p=0.046) and less influence from soil water, leaf area and root length density (R 2 <0.56, p>0.05) (Table 5 3) At about 2.5 years, the trees at Lake Alfred showed that soil water at Lake Alfred (p<0.001) influenced water uptake to a larger extent while canopy volume, soil water at 45 cm, trunk cross sectional area and leaf area were less influential (p>0.05) (Table 5 4). This observation was also supported by results for 6 yr old trees at Immokalee in June 2011 and 3 yr old trees at Lake Al fred later in September 2011. For example, cumulative water uptake at Immokalee was largely influenced by soil water at 10, 20, 30, 40 and 50 cm soil depth (p<0.001) and not necessarily canopy volume (p=0.400), leaf area (p=0.96) and trunk cross sectional area (p=0.576). Also, the soil water at 10 cm (p=0.001) and 45 cm (p=0.002) depths at Lake Alfred in September 2011 exerted significant influence on water uptake compared with canopy volume (p=0.826), trunk

PAGE 165

165 cross sectional area (p=0.053) and leaf area (R 2 = 0.27) (Tables 5 3 and 5 4). However, longterm analysis of water uptake versus tree characteristics suggests that canopy volume (Figure 5 27 ) will be the major determinant of overall tree water uses matched with good irrigation practice. An exponential mod el adequately described the relationship between cumulative water uptake and canopy volume. Thus, it appears for young trees (<6 yr old) irrigation scheduling is a critical management practice especially for the sandy soil as shown by the good correlation with water uptake. Despite weak correlation with root length density, the results on root density showed increased root intensity in the top 0 30 cm soil depth layer indicating that water extraction would be enhanced with an increase in available water. S ummary The chapter described the citrus water uptake and soil moisture distribution patterns in the irrigated zone on the citrus producing regions of central and southwest Florida. The results showed that hourly, daily and cumulative sap flow w ere higher using the ACPS/OHS irrigation methods compared with the conventional grower practices (fertigated or receiving granular fertilization) albeit, no t significantly different The citrus water use in agreement with the postulated hypothesis, did increase wi th canopy volume and root length density in situ irrespective of the irrigation frequency and fertigation method and correlated strongly with soil moisture content, trunk crossectional area and canopy volume. The high uptake in the ACPS/OHS irrigation met hods is ascribed to the frequent irrigation and vigorous growth resulting in trees with large canopy volumes, leaf areas and trunk cross sectional areas compared with weekly irrigation associated with the grower practice The results support the thinking b ehind the novel ACPS/OHS practices that nutrient leaching would be minimized while

PAGE 166

166 accelerating tree growth as a result of enhanced water and corresponding nutrient uptake. Thus, the irrigated root zones of DOHS or MOHS which have about 4% and 20% of the area irrigated by CMP, respectively, show ed that the trees would not be stressed by the ACPS practices. The K c followed a similar pattern to that of sap flow and was generally higher using drip OHS compared with microsprinkler irrigation. For young trees 1 .5 to 2.3 y r old at the Lake Alfred site, i ndex sap flow K c averaged 0.0290.014 and 0.0420.003 using CMP and DOHS S wingle respectively at the Lake Alfred site in July 2010, increasing to 0.060.01 and 0.080.02 in March 2011 For older trees greater tha n 3 yr old, K c varied from 0.25 0.10 in March to 0.540.26 in June and 0.570.43 in September Thus, these studies revealed that tree water uptake accounted for a bout 3 to 10 % of the actual ET when the trees are small and over 60 % of the ET after three ye ars when the trees increased in size with regard to leaf area and canopy volume. The soil moisture distribution patterns in all the irrigation methods were similar and maintained soil moisture close to or slightly above field capacity l argely in the range of 7 and 15 % suggesting that soil moisture was non limiting at both sites Thus, the The increased availabi lity of water in the top 30 cm suggests that the leaching threat is minimal under such frequent irrigation practices due to increased root water and probably nutrient extraction from this layer. These results support intensive irrigation management practic es in young trees to insure ample water is available in the root zone.

PAGE 167

167 Table 5 1. Average leaf area Site Small Medium Large cm 2 Immokalee § 10.71 4.19 28.68 8.39 67.27 16.77 Lake Alfred 15.13 4.70 36.22 8.04 68.93 16.01 § All val ues are mean areas of 20 leaves one standard deviation Table 5 2. Tree canopy volume (CV), stem cross sectional area (SCA), and trunk cross sectional area (TCA) Immokalee site Irrigation method February 2011 June 2011 CV (m 3 ) SCA (cm 2 ) TCA (cm 2 ) CV (m 3 ) SCA (cm 2 ) TCA (cm 2 ) DOHS 7.040.80 2.480.96 27.002.14 9.321.10 3.301.24 31.512.32 MOHS 4.400.98 2.340.80 19.324.35 6.671.30 3.471.43 26.194.44 CMP 6.471.20 2.050.81 20.212.60 7.611.26 3.091.23 25.723.92 Lake Alfred site Irrigation method July 2010 Marc h 2011 August 2011 CV (m 3 ) TCA (cm 2 ) CV (m 3 ) TCA (cm 2 ) CV (m 3 ) TCA (cm 2 ) DOHS Swingle 1.420.32 7.060.26 4.890.58 14.440.94 6.530.88 25.591.94 CMP 0.900.20 5.591.17 2.810.73 10.021.55 5.330.48 19.461.60 MOHS NA NA 3.910.67 12.842.04 5.84 0.85 21.333.41 DOHS C 35 NA NA 3.830.79 12.831.50 4.450.45 18.192.01 Mean one standard deviation, n=3 per treatment for trees sampled in July 2010, n=4 for trees sampled in February, March and August 2011, mean 1 standard deviation CMP Convent ional microsprinkler practice, DOHS Swingle Drip open hydroponic system with Hamlins on Swingle rootstock, DOHS C 35 Drip open hydroponic system with Hamlins on C 35 rootstock, MOHS Microsprinkler open hydroponic system

PAGE 168

168 Figure 5 1. Linear correlations of leaf area index and canopy volume as a function of leaf area in March 2011 at the Lake Alfred site

PAGE 169

169 Figure 5 2. Correlations of leaf area index (LAI) and canopy volume as a funct ion of total leaf area at Immokalee site in March 2011

PAGE 170

170 Figure 5 3. Average hourly sap flow in July, 2010 (top) and March, 2011 (bottom) at Lake Alfred site

PAGE 171

171 Figure 5 4. Average daily sap flow in July, 2010 at the Lake Alfred site Error bars represent one standard deviation Figure 5 5 Average daily sap flow in March, 2011 at the Lake Alfred site Error bars represent one standard deviation

PAGE 172

172 Figure 5 6 Average hourl y sap flow in February March 2011 at SWFREC. Data logger used for CMP and MOHS had a fault and showed very low sap flow readings

PAGE 173

173 Figure 5 7 Average daily sap flow in February March 2011 at SWFREC. Error bars represent one standard deviation. Data logger used for CMP and MOHS had a fault and showed very low sap flow readings

PAGE 174

174 Figure 5 8 Average hourly flow in June 2011 at the Immokalee site

PAGE 175

175 Figure 5 9 Average hourly sap flow in August September, 2011 at the Lake Alfred site

PAGE 176

176 Figure 5 10 Average daily sap flow in June 2011 at the Immokalee site Error bars represent one s tandard deviation

PAGE 177

177 Figure 5 11 Average daily sap flow in August September, 2011 at the Lake Alfred site Error bars represent one standard deviation

PAGE 178

178 Figure 5 12 Average hourly s oil moisture distribution in July 2010 at the Lake Alfred site measured at 10 and 4 5 cm soil depth layers

PAGE 179

179 Figure 5 13 Average daily s oil moisture distribution in July 2010 at the Lake Alfred site measured at 10 cm soil depth layer Error bars denote one standard deviation Figure 5 14 Soil moisture distribution in July 2010 at the Lake Alfred site measured at 4 5 cm soil depth layer Error bars denote one standard deviation

PAGE 180

180 Figure 5 15 Average hourly soil moisture distribution at the Lake Alfred site measured at 10 cm (top) and 4 5 cm (bottom) soil depth layers in March 2011

PAGE 181

181 Figure 5 16 Daily soil moisture distribution at the Lake Alfred site measured at 10 cm soil depth layer in March 2011 Error bars denote one standard deviation Figure 5 17 Daily soil moisture distribution at the Lake Alfred site measured at 45 cm soil depth layer in March 2011 Error bars denote one standard deviation

PAGE 182

182 Figure 5 18 Average hourly soil moisture distribution at the Lake Alfred site measured at 10 and 4 5 cm soil depth layers in August Sept ember 2011

PAGE 183

183 Figure 5 19 Average daily soil moisture distribution at the Lake Alfred site measured at 10 cm soil depth layer in August September 2011 Error bars denote one standard deviation

PAGE 184

184 Figure 5 20 Average daily soil moisture distribution at the Lake Alfred site measured at 45 cm soil depth layer in August September 2011 Error bars denote one standard deviation

PAGE 185

185 Figure 5 21 Soil moisture distribution for DOHS in February March 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers Error bars denote one standard deviation

PAGE 186

186 Figure 5 22 Soil moisture distribution for DOHS in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers Error bars denote one standard deviation

PAGE 187

187 Figure 5 23 Soil moisture distribution for MOHS in February March 2011 at Immokalee si te measured at 10 20 30 40 and 50 cm soil depth layers Error bars denote one standard deviation

PAGE 188

188 Figure 5 24 Soil moisture distribution for MOHS in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers Error bars denote one standard deviation

PAGE 189

189 Figure 5 25 Soil moisture distribution for CMP in February March 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers E rror bars denote one standard deviation

PAGE 190

190 Figure 5 26 Soil moisture distribution for CMP in June 2011 at Immokalee site measured at 10 20 30 40 and 50 cm soil depth layers Error bars denote one standard deviat ion

PAGE 191

191 Figure 5 27 Correlation of water uptake and canopy volume at the Immokalee and Lake Alfred sites

PAGE 192

192 Table 5 3. Linear regress ion models relating cumulative water uptake to tree and soil characteristics at the La ke Alfred site in July 2010 and September 2011 Tree/soil characteristic 0 1 R 2 P value Canopy volume 0.55 2.63 0.67 0.046 Soil water at 10 cm 0.20 0.11 0.55 0.091 Soil water at 45 cm 0.79 0.13 0.56 0.086 Root length density 2.46 0.28 0.35 0.217 Trunk cross sectional area 1.33 0.80 0.98 <0.001 Leaf area 2010 1.78 0.3 6 0.49 0.12 0 Leaf area 2011 6.39 0.95 0.27 0.038 Only leaf area measured in September 2011 at Lake Alfred was included in this table, the rest are variables measured in July 2010, 0 Constant, 1 = C oefficient, SW Soil water, R 2 =coefficient of determinat ion

PAGE 193

193 Table 5 4 Multiple linear regression model coefficients for cumulative water uptake Site Date Y o Canopy volume SW at 10 cm SW at 20 cm SW at 30 cm SW at 40 cm SW at 45 cm SW at 50 cm TCA Leaf area RMSE R 2 Lake Alfred site March, 2011 4.4 0 .56 0.16 NA NA NA 0.132 NA 0.204 0.220 1.37 0.60 Immokalee June, 2011 184.82 0.0012 2.937 16.38 27.11 12.84 NA 9.79 0.00013 0.000004 0.001 1.00 Lake Alfred site September, 2011 0.60 0.43 1.11 NA NA NA 0.45 NA 1.14 NA 4.42 0.80 Y o Constant, SW Soil water, TCA Trunk cross sectional area RMSE=Root mean square error, R 2 =coefficient of determination

PAGE 194

194 CHAPTER 6 CALIBRATION AND VALI DATION OF WATER, N, P, BR AND K MOVEMENT ON A FLORIDA SPODOSOL AND ENTISOL USING HYDRUS 2D developed HYDRUS 2D model to simulate the two dimensional movement of water, heat, and multiple solutes in variably saturated media. saturated water flow and convection dispersion e quations for heat and solute transport. The flow 1999; 2007). Soil hydraulic parameters of this model can be represented analytically using different hydraulic models such as the Brooks and Corey (1964) and van Genuchten (1980) equations. Several researchers have used HYDRUS in irrigated systems in the last decade (Fares et al., 2001; Grdens et al., 2005; Bovin et al., 2006; Fernndez Glvez and Simmonds, 2006; Hans 2009). Despite problems associated with identification of the actual physical processes when conducting simulation, Pang et al. (2000) found that HYDRUS model was able to accurately describe soil water contents with mi nor discrepancies. Studies by Grdens et al. (2005) and Hanson et al. (2006) assessed fertigation strategies using HYDRUS 2D for nitrogen fertilizers. They found that HYDRUS 2D model described the movement of urea, ammonium, and nitrate during irrigatio n and accounted for the reactions of hydrolysis, nitrification and ammonium adsorption. Model simulations help to describe and predict complex processes and scenarios that are difficult to understand in nature. Simulation modeling can offer a viable alter native to predicting expected outcomes in various situations (such as changes in climate, crop type, age of crop, soil type, season etc) within a given set of parameters. The models are generally incomplete and not conclusive but with some degree of

PAGE 195

195 accura cy can help decision makers come up with rational and informed decisions such as sustaining environmental quality and ensuring high yields among commercial growers. The model simulations were performed to : c alibrate HYDRUS 2D for water and solute moveme nt as a possible decision support system for the Candler and Immokalee fine sand using data from conventional mi crosprinkler and drip irrigation methods, validate the performance of HYDRUS 2D using field results of microsprinkler and drip OHS irrigation me thods determine the effect of supporting electrolyte on K D for predicting phosphorus movement at 30 cm soil depth using HYDRUS 1D, investigate bromide, nitrate and water movement using weather data from Immokalee and Lake Alfred. The hypothes e s tested w e re that : M easured soil water content, Br ammonium N, nitrate N, phosphorus and potassium correlate well with simulated outputs thus helping in decision support in citrus production systems K D values for P sorption have an effect on P transport in the to p 0 30 cm soil depth and would vary depending on the supporting electrolyte Bromide, nitrate and water movement for Candler and Immokalee sand could provide the basis for determining fertilizer residence time in the 0 60 cm soil depth. Materials and Method s Governing Equations and Parameters for Water Flow, Nutrient Transport and Uptake The governing flow equations for water flow and nutrient transport are given by the Richards (1931) and convection s(h) ( 6 1 )

PAGE 196

196 3 L 3 ], h is the pressure head [L], x i (i=1, 2) are the spatial coordinates [L] for two dimensional flow, t is time [T], are components of a dimensionless anisotropy tensor (which reduc es to the unit matrix when the medium is isotropic), K is the unsaturated hydraulic conductivity function (LT 1 ), and s is a sink/source term [L 3 L 3 T 1 ], accounting for root water uptake (transpiration). The sink/source represents the volume of water remov ed per unit time from a unit volume of soil due to compensated citrus water uptake. The equation (CDE) governing transport of independent solutes i.e. single ion transport is given as: ( 6 2 ) Where c 1 and c 2 are solute concentrations in the solid (MM 1 ) and liquid (ML 3 ) phases, respectively; q i is the i th component of volumetric flux dens ity (LT 1 rate of change of mass per unit volume by chemical or biological reactions or other sources (negative) or sinks (positive) (ML 3 T 1 ), respectively, providing connections b is the soil bulk density (M L 3 ), D ij is the dispersion coefficient tensor for the liquid phase [L 2 T 1 ]. The term r a represents the root nutrient uptake (ML 3 T 1 ) which is the sum of actual active and passive nutrient uptake. The solid phase concentration, c 1 accounts for nutrient either sorbed in the solid phase or precipitated in various minerals. This is usually quantified by the adsorption isotherm relating c 1 and c 2 described by the linear equation of the form: ( 6 3 ) Where K D (L 3 M 1 ) is the distribut ion coefficient of species 1. Nitrate or a tracer (e.g. Bromide) are assumed to have a K D =0 cm 3 g 1 while ammonium has a K D in the

PAGE 197

197 range of 1.5 to 4.0 (Hanson et al., 2006; Paramasivam et al., 2002; Lotse et al., 1992). The first order decay constant range s from 0.36 0.56 d 1 (Ling and El Kadi, 1998). Rate coefficient for the nitrification of ammonium nitrate ranges from 0.02 0.72 d 1 (Jansson and Karlberg, 2001; Lotse et al., 1992; Selim and Iskandar, 1981; Ling and El Kadi, 1998; Misra et al., 1974). For phosphorus, K D is reportedly in the range of 19 to 185 cm 3 g 1 (Kadlec and Knight, 1996; Grosse et al., 1999). The K D for potassium is reported to be 28.7 cm 3 g 1 (Silberbush and Barber, 1983). Bulk density for the soil is in the range 1.59 1.72 g cm 3 ( Immokalee) and 1.55 1.93 g cm 3 (Lake Alfred) (T.A. Obreza, unpublished). The sink term, s, for the Richards equation represents the volume of water removed per unit time from a unit volume of soil due to plant water uptake. Thus, s is defined as: ( 6 4 ) Where the water stress response function is a prescribed dimensionless function of the soil water pressure head, b is the normalized water uptake distribution, L t is the width of the soil surface associated with the trans piration process and T p is the potential transpiration rate (LT 1 ) and w is the water stress index. The predictive equation for the unsaturated hydraulic function in terms of soil water retention parameters is given by van Genuchten (1980) as : ( 6 5 ) ( 6 6 ) Where

PAGE 198

198 ( 6 7 ) ( 6 8 ) r s K s and l are residual water content (L 3 L 3 ), saturated water content (L 3 L 3 ), saturated hydraulic conductivity (LT 1 ), and pore connectivity parameter (estimated to be an average of 0.5 for many soils). (L 1 ) and n are empirical coefficients affecting the shape of the hydraulic functions. We e stimated the hydraulic Genuchten model in Community Analyses System (CAS) 2007 (Bloom, 2009) developed for determination of soil hydraulic functions. Model Calibratio n Processes Sorption isotherm s determination HYDRUS 2D was calibrated using experiment a lly measured site specific values reported in Appendices B and C. The methods for calculating and estimating the parameters are also documented in Appendices B and C. S orption isotherms on the disturbed soil samples (0 15 cm, 15 30 cm) were determined using the batch equilibration procedure The initial solution concentratio ns for P in 0.005M CaCl 2 and 0.01M KCl were 10, 25, 50 ppm P. In the fertilizer mixture, the initi al concentrations were 6, 32 and 64 ppm N H 4 N 5, 25 and 50 ppm P and 6, 32 and 63 ppm K. The initial concentrations for N, P and K were chosen based on University of Florida IFAS recommendations for young, non bearing orange trees (Obreza and Morgan, 200 8) In this set of observations soil samples were obtained from 5 positions per site at two depths giving a total of 10 samples. Each sample was weighed in triplicates plus a blank check. A 10 g air dried, <2mm subsample of soil was placed in a centrif uge tube

PAGE 199

199 and equilibrated with 20 ml (soil solution ratio 1:2) of 3 initial concentrations of NH 4 + N, P and K solutions. The centrifuge tubes were shaken for 24 h, and centrifuged for 20 min and filtered. The supernatant was passed through a Whatman filte r paper (Q2). All these procedures were done at room temperature ~ 251 o C as recommended by Graetz and Nair (2009) but the filtrate was later stored at <4 o C until analysis for NH 4 + N, P and K. The samples from 0.005 M CaCl 2 and 0.01 M KCl were analyzed f or P while the fertilizer mixture was analyzed for NH 4 + N, P and K The amount of chemical sorbed t o the soil was calculated from the difference between the initial and equilibrium solution concentration: ( 6 9 ) Where S is the adsorbed concentration ( m g k g 1 ); V o is the volume of initial solution (L) ; m is the soil mass ( k g); C o is the initial concentration of the standard solution (m g L 1 ), and, C is the soil solution concentration at equilibrium ( m g L 1 ). KH 2 PO 4 w as u sed as a source for both P and K, while NH 4 NO 3 w as used as a source of NH 4 + N. The linear sorption isotherm wa s determined from the following model: S e =K D C e ( 6 10 ) Where K D =sorption distribution coefficient (L kg 1 ) Sorption isotherms for P w ere calculated using the Freundlich equation : ( 6 11 ) Where K f = the Freundlich sorption coefficient (mg 1 N kg 1 L N ) and N are empirical constants related to adsorption phenomena (Bowman, 1982) The linearized form of the Freundlich equation w as used to calculate K f and N:

PAGE 200

200 ( 6 12 ) Where S is the adsorbed equilibrium concentration (mg kg 1 ); C is the equilibrium concentration (mg L 1 ) and K f and N are calculated from the intercept a nd slope of Eq. B 4 To find average K D for the Freundlich isotherm, the integrated form of the equation was used : ( 6 13 ) The range of sorption coefficients used for potassium and am monium are presented in Appendix B, Table B 2. Ammonium adsorption for Immokalee and Candler fine sand followed a linear isotherm with distribution coefficients (K D ) o f 1. 1 2 0.42 and 1. 64 0.25 kg L 1 and 1.66 0.39 and 1.76 0.39 kg L 1 for the 0 15 and 15 30 cm depths, respectively. The range of linearized K D values for P for calibration are documented in Table B 3, with the three supporting electrolyte s P adsorption was well described by a Freundlich model with linearized K D ranging from 0.50 0.19 to 0.7 5 0.13 kg L 1 for Immokalee fine sand and from 1.73 0.15 to 4.43 0.50 kg L 1 for Candler fine sand. P sorption isotherm for Immokalee fine sand determined using fertilizer mixture was linear with K D averaging about 0.44 10 kg L 1 Determination of soil wat er retention and hydraulic functions Twenty undisturbed soil core samples were taken at 0 to 15 cm, 15 to 30 cm, 30 to 45 cm, and 45 to 60 cm at random locations at both Flatwoods and Ridge sites to determine soil water release curves (Klute, 1986; van Gen uchten, 1980; Paramasivam et al., 2002) and saturated hydraulic conductivity at each depth for each site (Klute and Dirksen, 1986). Soil physical parameters determined include bulk density, field capacity (at 5 kPa at the Ridge and at 8 kPa Flatwoods), av ailable water capacity, saturated

PAGE 201

201 hydraulic conductivity, and saturated water content. Textural classes were determined Reference Level at the 60 cm depth using average vol umetric water content at different soil depths for different treatments: ( 6 14 ) where K(h) = conductivity of the soil layer at suction (h, cm); (H 1 H 2 ) = differences in total water potential between two points in the soil profile; (X 1 X 2 ) = the thickness of the soil profile (cm); Soil water retention curves were determined in the laboratory according to the process described by Klute (1986) using Tempe Cells and were adapted from (Sanchez, 2004). Each sample wa s covered with a plastic bag and wrapped with a rubber band to avoid any soil loss. The samples were stored in the refrigerator to maintain the original soil water content until processing in the laboratory. To determine the water retention curves between 0 and 100 kPa, the soil cores were placed in the base cap of a Tempe cell containing a 0.5 bar porous ceramic plate. The soil sample was covered with the top cap of the Tempe cell. The Tempe cell was placed in a container with appropriate water level to s aturate the soil sample. After the samples reached saturation, the Tempe cells were removed from the water container and excess water was allowed to drain from the saturated samples under gravity. The Tempe cells were weighed and the initial weights were r ecorded. After the first point of equilibrium, the pressure line was connected to the top inlet of the Tempe cell. The weights were recorded, each time the Tempe cell reached equilibrium with the corresponding pressure applied. The Tempe cells were subjec ted to 13 levels of pressure: 0.3, 2.0,

PAGE 202

202 2.9, 4.4, 5.9, 7.8, 9.8, 14.7, 19.6, 33.8, 50.0, 70.0 and 100 kPa. The moisture content at 1500 kPa was determined from literature on earlier studies done on same soil series ( Carlisle et al., 1989; Obreza et al., 19 97, Obreza, unpublished). After applying the last level of pressure and reaching equilibrium, the Tempe cell was opened and the soil core was carefully removed. Then, the weight of the core was recorded. Saturated hydraulic conductivity was determined by constant head method. To determine saturated hydraulic conductivity, another brass ring was attached and sealed with a duct tape on top of the soil core. The surface of the soil sample in the cylinder was covered with a filter paper to avoid any disturbanc e during water application. The soil sample in the core assembly was rewetted in a water container. The core assembly was then transferred to the hydraulic conductivity apparatus where water was applied to the top cylinder and the water level was kept cons tant. Once a steady flow was established, the drainage water under the soil sample was collected for a known period of time for each sample. The volume of drained water and time was recorded and the saturated hydraulic conductivity determined. The soil wa ter desorption curves for both Immokalee and Candler fine sand were simulated using the VanGenuchten model described in Equations 6 5 and 6 6. Data c ollected related to residual and saturated moisture contents, moisture contents at field capacity, availabl e water content K sat and bulk density. The soil physical parameters were calculated to show the variation in soil physical characteristics as a function of depth and the soil water release curves developed using the nonlinear regression analysis using the CAS software developed by Bloom (2009)

PAGE 203

203 The range of soil water retention parameters and n, used for calibration are presented in Appendic C, Table C 1 The respective and n value ranges were 0.03 0.04 cm 1 and 1.29 2.06 for Immokalee fine sand,and 0.02 0.04 cm 1 and 1.70 2.22 for Candler fine sand. The l value used was 0.5, as recommended by Simunek et al. (1999). The soil physical parameters like residual and saturated moisture content, saturated hydraulic conductivity, and bulk density are documen ted in Table C 2. The residual moisture contents from literature are 0.013 and 0.009 cm 3 cm 3 for Immokalee and Candler fine sand (Carlisle et al., 19 89 ). The saturated moisture contents ranged from 0.318 to 0.390 cm 3 cm 3 on Immokalee fine sand and from 0.313 to 0.421 cm 3 cm 3 on Candler fine sand. The saturated hydraulic conductivity ranged from 13.22 to 15.82 cm h 1 on Immokalee and 14.76 to 15.94 cm h 1 on Candler fine sand. The bulk densities, similar for the two soils, ranged from 1.59 to 1.62 g cm 3 and from 1.57 to 1.68 g cm 3 for Immokalee and Candler fine sand, respectively. For model calibration, we based on spring 2011 soil water movement to avoid the effects of rainfall in summer 2011. All the parameters for use in the model for validation a ssuming a homogenous soil profile, are presented in Table 6 4. The bulk density, K sat sat r n, and l values were 1.61 and 1.64 g cm 3 14.40 and 15.49 cm h 1 0.35 and 0.36 cm 3 cm 3 0.01 cm 3 cm 3 0.033 and 0.028 cm 1 1.34 and 1.8, and 0.5 for Im mokalee and Candler fine sand, rerespectively. S orption coefficients for P, NH 4 + and K + for Immokalee and Candler fine sand were 0.44 and 0.98 L kg 1 1.37 and 1.89 L kg 1 and, 1.17 L kg 1 Sensitivity A nalysis of Selected Parameters for HYDRUS 2D The ai m of sensitivity analysis (SA) is to determine how sensitive the output of a model is, with respect to the elements of the model which are subject to uncertainty or

PAGE 204

204 variability (Monod et al., 2003). SA helps explore efficiently the model responses when th e input or parameter varies within given ranges (Sacks et al., 1989; Welch et al., 1992; Monod et al., 2003). The uncertainty in model structure, model parameters and input variables calls for SA to 1) check that the model output behaves as expected when the input varies; 2) identify which parameters need to be estimated more accurately and which input variables need to be measured with maximum accuracy; 3) identify which parameters have a small or large influence on the output; 4) detect and quantify inte raction effects between parameters, between input variates or between parameters and input variates (Saltelli et al., 2000; Monod et al., 2003 ) Two methods of conducting SA are well known: local and global sensitivity analysis. Local sensitivity analysis (LCA), on the one hand, is based on the local derivatives of output with respect to input variable or parameter which indicate how fast the output increases or decreases locally around given values of the input variable or parameter. In global sensitivity analysis (GSA), on the other hand, the output variability is evaluated when the input factors vary their whole uncertainty domains (Saltelli et al., 2000; Garnier, 2003; Monod et al., 2003 ; Saltelli et al., 2004). Of the two methods GSA if preferred beca use it helps the modelers identify inputs or parameters that deserve an accurate measure or estimation. One method to conduct a GSA is to vary one factor at a time, while other factors are fixed at their nominal values. The relationship between z i of fac tor Z i and the responses f(z 0,1 0, i 1 z i ,z 0,i+1 0,s ) determines a one at a time response profile. Each input factor or parameter z i takes k equispaced values from zmin, i to z max i with increments: ( 6 15 )

PAGE 205

205 The model responses f(z 0,1 0, i 1 z i,z0,i+1 0,s ) are then calculated for the k discretized values zi. Graphical representations and the Bauer and Hamby Index are used to determine the influence of the model parameters on the model output Th e Bauer and Hamby Index, I i BH (Bauer and Hamby, 1991) is approximated by the difference between maximum and minimum simulated values given as: ( 6 1 6 ) In this study, an att empt was made to conduct a GSA of HYDRUS 2D focusing on the following state variables: NO 3 N and Candler sand The simulations were done for 14 days to mimic the dynamics of a time of the field experiment at 1 d time step for drip and microsprinkler fertigation systems in a 50 cm wide and 60 cm deep transect subdivided into four layers each site and drippers located at 15 cm from the tree and microsprinklers irrigating the top 45 cm. The hypothesis governing the GSA i (NH 4 N), nitrate nitrogen (NO 3 N), phosphorus (P) and potassium (K) are reasonable within the given set of parameters. Once the parameters having a major influence on the outputs are known, a choice of which parameters to use for the various fertigation scenarios will be made based on the values that result in the least influence on the two study sites. GSA for the Immokalee sand was done separately from the Candler series near Lake Alfred due to the heterogeneity in drainage characteristics. Outputs of interest include d : soil water content, soil NO 3 N, NH 4 N, Br, P and K with depth. Data were analyzed using General Linear Model (GLM) and ProcReg procedures in SAS statistical package (SAS Institute, 20 11 ). Coefficients of determination (R 2 ) and root mean square errors (RMSE) between the simulated and measured values were

PAGE 206

206 determined to allow for statistical comparison of the correspondence between the measured and simulated data or between the results o f different models. Simulation Domain Microsprinkler irrigation The microsprinkler irrigation system for the two sites was simulated as a line source, planar two dimensional geometry perpendicular to the simulated domain assuming that the lateral flow on boundaries was zero (zero flux boundary condition) and the free drainage condition was imposed at the bottom boundary at each site with time variable flux surface boundary condition The simulation domain was 50 cm wide and 60 cm deep. The presence of a w ater table ~70 cm below the ground at Immokalee was assumed not to affect the drainage within the 60 cm simulation domain. The transport domain was discretized into 3834 triangular element s and 1918 nodes. The smallest finite element was 0.1 cm at the top of the simulation domain and the largest at the bottom of the domain was 2 cm. The non symmetry coefficients were assumed to be 1 and flow was assumed to be isotropic in both lateral and vertical directions. The maximum rooting depth was assumed to be 45 cm with maximum root intensity observed at 15 cm. Maximum citrus root lateral extension (<5 yr old) was assumed to be 45 cm while maximum lateral root intensity was found at 30 cm from the tree Detailed information related to the flow related parameters and experimental scenarios are presented in Tables 6 1 through 6 5. Simulation Domain Drip irrigation Drip irrigation was simulated as a point source, with an axi symmetrical two dimensional plane assuming that the lateral flow on boundaries was zero (zer o flux boundary condition). Like above, a free drainage condition was imposed along the bottom boundary at each site with a time variable flux boundary condition on the top

PAGE 207

207 surface. The simulation domain was also 50 cm radius and 60 cm deep. The presence o f a water table ~70 cm below the ground at Immokalee was assumed not to affect the drainage within the 60 cm simulation domain. The transport domain was discretized into 2462 triangular element and 1232 nodes. The smallest finite element was 0.1 cm and th e largest at the bottom of the domain was 2 cm. The non symmetry coefficients were assumed to be 1 and flow was assumed to be isotropic in both radial and vertical directions. The maximum rooting depth was assumed to be 45 cm with maximum root intensity ob served at 1 5 cm. Maximum citrus root lateral extension (<5 yr old) was assumed to be 45 cm while maximum lateral root intensity was found at 30 cm from the tree. Details related to the initial conditions and parameters are also presented in Tables 6 1 thro ugh 6 4 Results and Discussion Sensitivity analysis and calibration of selected model parameters The conceptual model for the uptake and movement of water, tracer Br and 6 1. M easured soil characteristic values (presented in Appendix C ) and soil nutrient sorption constants (presented in Appendix B ) were used to calibrate HYDRUS 2D for the Entisol and Spodosol at the Lake Alfred and Immokalee sites. The model was calibrated for b oth Candler and Immokalee sand for simulating water and solute transport as shown in Figures 6 2 6 3 and 6 4 The statistics reveal that the model outputs are close to the measured values with R 2 >0. 8 0 Sensitivity indices calculated suggest that saturated hydraulic conductivity and empirical parameter n were the most sensitive (sensitivity index=0.29) in predicting water movement (Table 6 5). Also, the simulation experiments on Candler fine sand

PAGE 208

208 suggest that any n>3.085 would yield no output with respect t o water content and uptake. Similarly, on Immokalee fine sand, no water content and water uptake values were obtained when n>4.63 (the nominal value) was used as a parameter. We also noted that no outputs on water content and uptake were obtained on Candl er fine sand sat <0.34 m 3 cm 3 was used. It is presumed that the parameter values for HYDRUS recommended for sandy soils and optimized using ROSETTA software (Carsel and Parrish, 1988; Schaap et al., 2001) are for soils than typical san four 0.15 m depth increments in the field to determine hydraulic functions. The values reported by several Florida researchers were close to our measured values because they were determined on similar soil series used in the study and were the basis for the global and local sensitivity analysis on both soils (Obreza, unpublished; Carl isle et al., 1989; Fares et al., 2008; Obreza and Collins, 2008). Most of literature values used for the sensitivity analysis of sorption coeffi ci ents with regard to P, K and NH 4 transport, were several times higher than what we estimated with soil samples collected from the research sites ( Appendix B ). Thus, the sorption confidents for P, K and NH 4 presented in Appendix B were used for the simulation experiments. Water Br, K, P, NO 3 and NH 4 movement with drip and microsprinkler irrigation To validate the calibrated model, measured water and solute movement were compared with model predicted values. Model predictions showed that with similar initial water contents and similar schedules, microsprinkler (in a line source, planar domain) and drip irrigation (with water from a point source, in a n axi symmetric domain), water movement were similar for both irrigation systems albeit, higher amounts of water

PAGE 209

209 were retained in the upper 0.15 m than when using the microsprinkler system. Very close agreement was obta ined (Table 6 6) between simulated and me asure values for the two systems where the predictions accounted for 90% of the measured water contents. Several researchers have reported good predictions on water in one and two dimensional domains using numerica l models (Angelakis et al., 1993; Andreu et a., 1997; Fares et al., 2001; Skaggs et al., 2004; Gardenas et al., 2005; Testi et al., 2006; Kandelous and Simunek, 2010a, b; Kandelous et al., 2011). Bromide distribution showed good agreements (R 2 ~0.63 0.90) with measured outputs with root mean square errors (RM SE) in the range of 0.04 7.57 (Figures 6 5 and 6 6 and Table 6 7). However, despite the good agreements, Br was under predicted by about 5 to 20% and there was very poor agreement at Immokalee, especial ly after 6 days of simulation Phosphorus was well predicted at Lake Alfred but poor correlations were noted at Immokalee. The phosphorus initial conditions were based on Mehlich 1 extractable P which might be several times greater than water soluble P (N air and Harris, 2004; Nair et al., 2004) and thus our prediction might have overestimat ed the actual leaching P potential. Nitrate and ammonium were well predicted by the model (Table 6 7, Figure s 6 8 and 6 9 ). Potassium, despite the under predictions, sho wed very good correlation at Immokalee, but poor correlation at Lake Alfred using microsprinkler (Table 6 7, Figure 6 10 and 6 1 1 ). Phosphorus movement with microsprinkler irrigation as function of K D value Phosphorus movement was predicted using three dif ferent K D s estimated with fertilizer mixture, 0.01M KCl and 0.005M CaCl 2 for a duration of 21 days, assuming no rainfall events (Figure 6 12) The assumption is that a K D value obtained using fertilizer mixture typifies that of field conditions with regard to chemical processes. The results on

PAGE 210

210 Candler fine sand at Lake Alfred showed that that P contents for the K D estimated with 0.01M KCl and 0.005M CaCl 2 were 10 15% higher than those predicted with a K D value measured with fertilizer mixture. The predicti ons on Immokalee fine sand showed that that P contents for the K D estimated with fertilizer mixture and 0.005M CaCl 2 were 12 20% higher than those predicted with a K D value measured with 0.01M KCl. The outputs with K D measured with 0.00 5 M CaCl 2 appear to be close to those predicted with a K D measured with fertilizer mixture. However, the analysis of the K D values across all electrolytes on the two soils studied revealed that 0.01M KCl is the electrolyte that yields K D values fairly close to fertilizer mix ture while 0.00 5 M CaCl 2 tends to give K D values two to threefold in magnitude to those determined with fertilizer mixture suggesting that the latter would overestimate P sorption and retardation during unsaturated or saturated flow than the former (0.01M K Cl). Thus, it would be Immokalee fine sand. Investigating bromide, nitrate and water movement using weather data from Immokalee and Lake Alfred. The nitrate, bromide and water movement as influenced by weather at Lake Alfred (August 22 to November 22, 2011 ) and Immokalee (June 4 to September 4, 2011) were predicted using climatic data obtained from the Florida Automated Weather Network for a 90 day period. The nitrate and bromi de at Lake Alfred (Figure 6 13A and B) was largely leached out beyond 60 cm depth within <20 days, a period corresponding with 158 mm of rain. The nitrate and bromide at Immokalee showed that most of the nitrate was leached i n 20 days and bromide leached after 25 days, dates corresponding with 57 and 108 mm of rain Mostly during the 90 days simulation, water contents remained

PAGE 211

211 between 15 and 25% and only went above 30% when it rained. The leaching of NO 3 in this case would be minimized if we accounted for uptake and transformation of nitrate into other forms. However, the incorporation of weather data into the simulation would serve as a guide in making decisions to apply mobile nutrients such as nitrate containing fertilizers when the weather forecast is g ood i.e. no chances of rainy events. The plausible approach with the irrigation practices used in this study is that they try to maintain soil moisture in the top 10 cm depth at near field capacity and applying nutrients in the morning hours when transpira tion and photosynthesis are high to avoid leaching losses (Schumann et al., 2010). Such irrigation and nutrient management decisions should be incorporated in the simulations ahead of a rainy season using say historical data to insure environmental quality is sustained. Summary The model showed reasonably good agreement between measured and simulated values for soil water content, Br ammonium N, nitrate N, phosphorus and potassium soil water content, B r ammonium N, nitrate N, phosphorus and potassium correlate well with simulated outputs thus helping in decision support in citrus production systems thus helping in decision support in citrus production systems The sorption K D value has a bearing on P transport in the root zone, the greater the value, the more retarded and adsorbed P is in the soil Thus, the use of 0.01M KCl which yielded K D values close to those of fertilizer mixture, appears to be the appropriate supporting electrolyte for Candler and Immokalee fine sand while 0.005M CaCl 2 tends to overestimate the P sorption process. The model could further be used as an important guideline for predicting Br or nutrient residence time. For example, th e Br at Immok alee leached between 15 to 25 d an d in

PAGE 212

212 less than 10 d at 60 cm depth near the Lake Alfred site. The NO 3 N leached between 15 to 20 d at Immokalee and between 10 12 d at the Lake Alfred site. Importantly, HYDRUS 2D could also be used for irrigation decision s upport if one could account for water use drainage and evaporation losses. The parameters used for HYDRUS should be carefully determined for meaningful predictions. When in doubt, own parameter estimation through laboratory or field measurements where time and resources permit should b e done. or over prediction were noted particularly for P, K, NO 3 and NH 4 probably due to transformations and adsorption The model could be successfully used for scheduling irrigation and predicting nutrient leaching for both microsprinkle Spodo so ls and Entisols. A correction factor may need to be used for the NH 4 NO 3 P and K outputs to account for soil processes such as chemical transformations ( largely considered negligible in HYDRUS) and sorpti on to successfully predict nutrient leaching, on case by case basis, according to soil type and management practice A dditional ly initial conditions for adsorbed solutes should probably be determined using water extraction to mimick natural conditions.

PAGE 213

213 Figure 6 1. A forrester diagram describing the conceptual model for water and nutrient uptake and movem e n t processes

PAGE 214

214 Figure 6 2 Calibration of HYDRUS 2D for simulati ng soil water content at 10 cm soil depth at Lak e Alfred site using drip irrigation Figure 6 3 Calibration of HYDRUS 2D for simulation soil water content at 40 cm soil depth at Lake Alfred site using microsprinkler irrigation

PAGE 215

215 Figure 6 4 Calibration of HYDRUS model for simulating ammonium N movement on Candler fine sand Table 6 1. Selected parameters for sensitivity analysis for simulating water flow and nutrient movement in citrus using HYDRUS 2D Parameter Units Nominal value Range Source rIM cm 3 cm 3 0.010 T.A. Obreza (unpublished data) sIM cm 3 cm 3 0.340 0.33 0.35 T.A. Obreza (unpublished data) IM cm 1 0.023 0.022 0.023 T.A. Obreza (unpublished data) n IM 4.631 3.19 6.07 T.A. Obreza (unpublished data) K sIM cm d 1 586.800 242.40 931.20 Carlisle et al., 1989 rLA cm 3 cm 3 0.009 T.A. Obreza (unpublished data) sLA cm 3 cm 3 0.385 0.36 0.41 T.A. Obreza (unpublished data) LA cm 1 0.061 0.04 0.079 T.A. Obreza (unpublished data) n LA 2.571 2.28 2.87 T.A. Obreza (unpublished data) K sLA cm d 1 600.500 455.00 746.00 Fares et al., 2008 K D NH4 cm 3 g 1 2.75 1.50 4.00 Selim and Iskandar, 1981; Lotse et al., 1992; Ling and El Kadi, 1998 K D P cm 3 g 1 102.0 19.00 185.00 Silberbush and Barber, 1983; Kadlec and Knigh t, 1996; Gross et al., 1999 K D K+ cm 3 g 1 28.7 11.48 45.92 Silberbush and Barber, 1983 w c 0.5 0.20 0.80 c 0.5 0.20 0.80 IM denotes soil hydraulic functions for Southwest Florida Research and Educational Center (SWFREC), Immokalee LA denotes soil hydraulic functions for Citrus Research and education Center (CREC), Lake Alfred

PAGE 216

216 Table 6 2. Irrigation system parameters for HYDRUS 2D for Immokalee and Candler fine sand Irrigation system parameter Drip Microsprinkler Irrigation Discharge rate (L/h) 2 40 Irrigation time (day) 0.13 0.08 Irrigation in terval (day) 1 1 Within x cross row tree spacing (cm) Immokalee 305 x 671 306 x 671 Within x cross row tree spacing (cm) Lake Alfred 305 x 610 306 x 610 Water use Transpiration February/March (mm/day) 1 1 Transpiration June/July (mm/day) 3 3 Ev aporation February/March (mm/day) 2 2 Evaporation June/July (mm/day) 2.5 2.5 Crop coefficient February/March (mm/day) 0.71 0.71 Crop coefficient June/July (mm/day) 0.83 0.83 Simulated domain Source Point Line Two dimensional geometry Axisymmetric al Planar Width (cm) NA 50 Radius (cm) 50 NA Depth (cm) 60 60 Number of triangular finite elements 2462 3834 Number of nodes 1232 1918 § Root water uptake Feddes pressure heads P0 (cm) 10 10 Popt (cm) 25 25 P2H (cm) 200 200 P2L (cm) 10 00 1000 P3 (cm) 8000 8000 r2H (cm/day) 0.5 0.5 r2L (cm/day) 0.1 0.1 Root zone parameters Root distribution model Vrugt Vrugt Maximum rooting depth (cm) 45 45 Depth with maximum root density (cm) 15 15 Maximum root lateral extension (cm) 45 45 Distance with maximum root density (cm) 30 30 Non symmetry coefficients, p z and p r 1 1 Obtained from Morgan et al. (2006 b ) ; § Obtained from Feddes et al. (1978 )

PAGE 217

217 Table 6 3. Simulation experiment scenarios for the Ridge and Flatwoods soils Irrigation sys tem Fertigation frequency Irrigation frequency Outputs of interest Drip Daily Daily NO 3 N, Br, soil water content, NH 4 N, P, K Microsprinkler Weekly Daily NO 3 N, Br, soil water content, NH 4 N, P, K Microsprinkler Weekly Daily NO 3 N, Br, soil water co ntent Outputs in the soil will be predicted on observation nodes at 15 cm and 60 cm for NO 3 N, Br, soil water content depth s while P and K will be predicted at 15 cm depth at 15 cm from the tree while Table 6 4. S oil physical characteristics and initial conditions of the Immokalee and Candler fine sands Soil # b K sat § sat §§ r n l ## K dP K dNH4 K dK §§§ Br NO 3 NH 4 M1P M1K Immokalee 1.61 14.40 0.35 0.01 0.033 1.34 0.5 0. 44 1.37 1.17 0.2 3.0 3.0 50.0 40.0 Candler 1.64 15.49 0.36 0.01 0.028 1.80 0.5 0 9 8 1.89 1.17 0.1 12.0 6.0 100.0 60.0 # b Bulk density, g cm 3 K sat saturated hydraulic cponductivity, cm h 1 § sat Saturated moisture content, cm 3 cm 3 §§ r Residual moisture content obtained from Obreza, unpublished data, cm 3 cm 3 ## K d Sorpt ion coefficient ( L kg 1 ) for P, K and NH 4 §§§ Br, NO 3 NH 4 M1P, M1K initial concentrations of Br, NO 3 NH 4 M1P, M1K, mg kg 1

PAGE 218

218 Table 6 5. Sensitivity indices for selected parameters for soil available water P, ammonium and K movement using HYDRUS 2D Par ameter Units Nominal value Bauer and Hamby s ensitivity index rIM cm 3 cm 3 0.01 0.04 sIM cm 3 cm 3 0.34 IM cm 1 0.02 n IM 4.63 K sIM cm d 1 586.80 rLA cm 3 cm 3 0.0 1 0.04 sLA cm 3 cm 3 0.385 0.08 LA cm 1 0.061 0.17 n LA 2.571 0.21 K sLA cm d 1 600.50 0.29 K dNH4 IM cm 3 g 1 2.75 0.31 K dP IM cm 3 g 1 102.0 0.21 K dK IM cm 3 g 1 28.7 0.33 K dNH4LA cm 3 g 1 2.75 0.19 K dPLA cm 3 g 1 102.0 0.06 K dK LA cm 3 g 1 28.7 0.0 3 IM denotes soil hydraulic functions for Southwest Florida Research and Educational Center (SWFREC), Immokalee LA denot es soil hydraulic functions for the site near Citrus Research and education Center (CREC), Lake Alfred

PAGE 219

219 Figure 6 5 Soil Br monitored at 15 and 60 cm depth using drip irrigation at the Lake Alfred site

PAGE 220

220 Figure 6 6 Measured and simulated Br concentration at 15 and 60 cm at Immokalee site us ing microsprinkler irrigation

PAGE 221

221 Figure 6 7 Soil P monitored at 15 cm depth using drip irrigation at the Lak e Alfred site Figure 6 8 Simulated and measured cumulative nitrate concentration using microsprinkler irrigation at the Immokalee site

PAGE 222

222 Figure 6 9 Simulated and measured cumulati ve ammonium concentration using drip irrigation at the Immokalee site

PAGE 223

223 Figure 6 10 Cumulative K distribution at 15 cm soil depth at Immokalee site using microsprinkler irrigation Figure 6 1 1 Cumulative K distribution at 15 cm soil depth at Immokalee site using drip irrigation

PAGE 224

224 Figure 6 12. Phosphorus movement on Candler and Immokalee fine sand depending on K D value estimated using HYDRUS 1D

PAGE 225

225 Table 6 6 Statistical comparison between the observed and simulated water contents and uptake in spring and summer on Candler and Immokalee sand Soil Comparison Soil available water (mm) R 2§ Candler OBS vs. MS spring at 10 cm 0.99 Candler OBS vs. MS spring at 40 cm 0.87 Candler OBS vs. DRIP spring at 10 cm 0.99 Candler OBS vs. DRIP spring at 40 cm 0.93 Candler DRIP vs. MS at 10 cm 1.00 Candler DRIP vs. MS at 40 cm 1.00 Immokalee OBS vs. MS spring at 10 cm 0.99 Immokalee OBS vs. DRIP spring at 10 cm 1.00 Immokalee OBS vs. MS spring at 40cm 1.00 Immokalee OBS vs. DRIP spring at 40 cm 0.95 Immokalee DRIP vs. MS spring at 10 cm 1.00 Immokalee Drip vs. MS spring at 40 cm 0.99 Immokalee OBS vs. MS summer at 10 cm 0.99 Immokalee OBS vs. DRIP sum mer at 10 cm 0.96 Immokalee OBS vs. MS summer at 40cm 0.99 Immokalee OBS vs. DRIP summer at 40 cm 1.00 OBS Observed or measured in the field, MS Microsprinkler irrigation, DRIP Drip irrigation § R 2 Coefficient of determination,

PAGE 226

226 Table 6 7 Statistical comparison between the observed and simulated Br, NO 3 NH 4 M1P and M1K on Candler and Immokalee sand Soil Comparison Br NO 3 NH 4 M1P M1K RMSE (mg kg 1 ) R 2§ RMSE §§ (mg kg 1 ) R 2 RMSE (mg kg 1 ) R 2 RMSE (mg kg 1 ) R 2 RMSE (mg kg 1 ) R 2 Candler OBS vs MS Fall at 15 cm 2.08 0.89 1.65 0.88 1.06 0.98 57.9 0.78 21.74 0.44 Candler OBS vs MS Fall at 60 cm 1.25 0.76 1.52 0.84 §§§ NA NA NA NA NA NA Candler OBS vs DRIP Fall at 15 cm 0.35 0. 96 5.48 0.98 1.98 0.91 6.74 0.74 14.64 0.99 Candler OBS vs DRIP Fall at 60 cm 0.86 0.75 1.90 0.66 NA NA NA NA NA NA Immokalee OBS vs MS summer at 15 cm 7.57 0.79 5.25 0.75 1.33 0.95 55.62 0.25 11.22 0.93 Immokalee OBS vs DRIP summer at 15 cm 0.44 0.90 1.66 0.91 1.15 0.93 15.74 0.69 10.64 0.94 Immokalee OBS vs MS summer at 60cm 0.06 0.74 4.88 0.82 NA NA NA NA NA NA Immokalee OBS vs DRIP summer at 60 cm 0.04 0.63 1.95 0.85 NA NA NA NA NA NA OBS Observed or measured in the field, MS Microsprinkler irrigation, DRIP Drip irrigation § R 2 Coefficient of determination, §§ RMSE R oot mean square error, mm, §§§ NA Not applicable

PAGE 227

227 Figure 6 13. Simulated nitrate (A), bromide (B) and water (C) movement over a 90 day period at 60 cm using grower practice

PAGE 228

228 CHAPTER 7 CONCLUSIONS The study had sought to address the general research objectives and goals as conceptualized: 1) d evelop optimum irrigation rate, method, an d timing for young citrus trees 2) d etermine growth and yield effects of fertigation on young citru s trees at selected frequencies 3) m eas ure effect of irrigation method and frequency on rooting patterns, nutrient retention and water and nutrient uptake 4) c haracterize the soil physical parameters and sorption of K, P and NH 4 of Immokalee and Candler sand 5) c alibrate HYDRUS for water and nutrient movement using site specific soil hydraulic characteristics and nutrient sorption behavior and 6) c haracterize HYDRUS as a possible decision support system for predicting soil moisture distribution and solute transport in the vadose zone. The a ppropriate general hypotheses formulated to answer the above research goals were as follows: 1) Microsprinkler and drip OHS will increase citrus growth rate, above ground biomass, and nutrient uptake resulting in higher plant N, P and K content than the co nventiona l practice. 2) Spatial nutrient and root length density distribution will be greater in irrigated zones of microsprinkler and drip OHS than conventional grower practices 3) Citrus water use and K c increase with canopy volume and root length densi ty in situ irrespective of the irrigation frequency and fertigation method. 4 ) Measured soil water content, ET and Br correlate positively with simulated outputs thus helping in decision support in citrus production systems. Overall, NH 4 + N, NO 3 N, M1P a nd M1K concentration and root length density decreased with distance from the irrigated zone and with depth, and were greater in patial nutrient and root length density distribution w o uld be greater in irrigated zones of microsprinkler

PAGE 229

229 This suggests the potential for increased nutrient retention and root uptake because the irrigated zone was associated with increased root density. Overa ll, the study found 60 90% increased nutrient retention with ACPS than grower practice. The use of Br suggested consistent trends in the movement of NH 4 + N, NO 3 N, M1P and M1K in the irrigated and nonirrigated zones, and could be used as an important gui deline for making nutrient management decisions with regard to nutrient residence time. The results at both sites showed increased tree size with ACPS than grower practices. For example the results at Immokalee showed that annual increments in trunk cro ss sectional area respectively for CMP, DOHS, and MOHS were 97, 123 and 122% in year 2 and 44 % 56 % and 66% in year 3 at Immokalee suggesting vigorous tree growth with ACPS/OHS This also underscored the Microsprinkler and drip OHS wi ll increase citrus growth rate and above and below ground biomass than the conventiona The gains on canopy volumes and trunk cross sectional area with ACPS and OHS compared with grower practices appear to be more pronounced during the first 3 ye ars of establishing a grove as shown by the results at Lake Alfred. P roportion al nutrient accumulation patterns revealed that OHS fertigation increased N accumulation by 45% over grower practice at Immokalee, but P and K accumulation were fairly similar be tween the three practices, though CMP showed slightly higher P and K accumulation than OHS. Thus, N accumulation confirmed the this hypothesis did not hold for P and K ac cumulation. The N, P and K concentration

PAGE 230

230 using granular fertilization at the Lake Alfred site suggests that grower practices are just as effective in promoting tissue nutrient concentration The biomass and nutrient ( N, P and K ) accumulation using granular fertilization or fertigation revealed that grower practices are just as effective in promoting nutrient and biomass accumulation. However, the grower practices do require more fertilizer and water applied per ha to achieve rapid tree development within 1 to 5 years of establishing a grove compared with ACPS practices Root length density measured using the line intersection method showed a positive correlation with those predicted by the calibration equation relating RLD and scanned root area. The results show that use of the scanning method could be used to increase the accuracy and reduce the time for determination of RLD. Generally, RLD was highest in the 0 15 cm depth and decreased with depth and distance away from the tree. Positions below the dripp er of DOHS and in the irrigated zones of MOHS showed higher root length density than non irrigated zones. Despite having irrigated zones around the tree using CMP, the infrequent irrigation probably resulted in lower RLD compared with the irrigated zones of DOHS and MOHS treatments at both study sites. The experiments on water uptake estimation showed that water uptake was higher using the ACPS/OHS fertigation methods compared with the conventional grower practices (fertigated or receiving granular fertili zation). The high uptake in the ACPS/OHS fertigation methods are ascribed to vigorous growth resulting in trees with large canopy volumes, leaf areas and trunk cross sectional areas. The results further support the thinking behind ACPS/OHS that nutrient leaching would be minimized while accelerating tree growth and fruit yield. Regression analysis further revealed that for

PAGE 231

231 young trees (< 5 yr old) irrigation scheduling is a critical management practice especially for the sandy soil as shown by the good cor relation of water uptake with soil moisture at 10, 20, 30, 40, 45 and 50cm soil depths. Tree size characteristics such as trunk cross sectional area and canopy volume also correlated well with water uptake. Despite weak correlation of cumulative sap flow w ith root length density, the results on root density showed increased root intensity in the top 0 30 cm soil depth layer indicating that water extraction would be enhanced with an increase in available water. The results from laboratory sorption work show that P adsorption in the top 0 30 cm was greater for Candler than Immokalee sand using tap wa ter in fertilizer mixture, 0.00 5M CaCl 2 and 0.01M KCl. The adsorption for P followed the Freundlich model and was best explained with 0.005M CaCl 2 as the supportin g electrolyte. The simulations with HYDRUS1D suggest that 0.005M CaCl 2 would be an appropriate electrolyte for Immokalee fine sand with low organic matter content (<0.65%) because the outputs were failry close to those of fertilizer mixture. For Candler fi ne sand, both 0.005M CaCl 2 and 0.01M KCl tend to over estimate P leaching making use of fertilizer mixture a viable option for estimating the sorption coefficients It appears the addition of a supporting electrolyte with a divalent or monovalent cation, unlike using fertilizer mixture increases the surface charge for adsorption of orthophosphate anions. The adsorption mechanism of both ammonium and potassium was linear and similar for both soils though ammonium adsorption coefficients were greater than those of potassium. The determination of the hydraulic conductivity and water retention characteristics yielded important site specific parameters like saturated and residual moisture contents,

PAGE 232

232 and hydraulic conductivity for use in the HYDRUS 2D model to describe water and solute transport to aid decision making in predicting environmental fate of fertilizers. The model simulations revealed that HYDRU 2D is a good model for predicting water and solute movement on Candler and Immokalee sand as long as it i s carefully calibrated with site specific parameters. However, the model appears to under predict most of the solutes of interest such as P K and NH 4 suggesting that a correction factor might need to be estimated with measured values. This under predictio n or over estimation is ascribed to the use of Mehlich 1 extractable P and K in the initial conditions for the simulations. Proba b ly, the use of water extractable values of P and cations of interest that give a better indication of leaching potential woul d be appropriate. Also, caution with the model relates to its inability to account for uptake in perennial crops like citrus and other transformation process of soil nutrients such as ammonium and nitrate. However, the HYDRUS 2D model could successfully b e used to determine fertilizer residence timeand for irrigation decisions if the modeler or grower has all the necessary parameters and climatic data for the site of interest. Based on the results from the field and laboratory experiments, the key points for citrus growers eager to try the novel practices of ACPS/OHS are documented here. First, water uptake with drip or microsprinkler OHS is similar to conventional microsprinkler practice but nutrient uptake particularly N, is increased with the former t wo than the fertigated grower practice. Also, the amount of water applied with drip or microspr i nkler OHS would be substantially less due to a limited root and irrigated zone, without stressing the tree with water deficit However, it appears one could use one drip line with two to four drippers per tree with in the first two to three years of installing the

PAGE 233

233 ACPS /OHS As the tree root and canopy volume expands with increase in tree age, there would be a need to increase irrigation frequency and the number of drip lines from one to two per tree row and the number of drippers per tree from two or four to eight or greater to effectively manage the greater tree sizes. This requires training of personnel in managing automated irrigation and fertigation, repairs and other maintenance procedures. Second, ACPS/OHS has the potential to accelerate tree growth and bring trees into production within the first five years after grove establishment. Third, ACPS/OHS installed on a coated sand like Candler fine sand present s greater p otentia l for vigorous tree growth and production due to better nutrient retention and higher soil organic matter (1.50 1.96%) than Immokalee fine sand with low nutrient retention and organic matter (0.40 0.61%) Last but not least, HYDRUS 2D co uld successfully be soils once the soil parameters are known.

PAGE 234

234 APPENDIX A SUPPLEMENTARY FIGURE S TO CHAPTERS 3, 4 A ND 5 Figure A 1 S oil Br distribution on Immokalee sand in the irrigated zone

PAGE 235

235 Figure A 2 Soil Br distribution on Immokalee sand in the non irrigated zone

PAGE 236

236 Figure A 3 Soil Br distribution on Ca ndler sand in the irrigated zone

PAGE 237

237 Figure A 4 Soil Br distribution on Candler sand in the nonirrigated zone

PAGE 238

238 Figure A 5 Soil ammonium N leaching on Immokalee sand in the irrig ated zone

PAGE 239

239 Figure A 6 Soil ammonium N leaching on Immokalee sand in the non irrigated zone

PAGE 240

240 Figure A 7 Soil nitrate N leaching on Immokalee sand in the irrigated zone

PAGE 241

241 Figure A 8 Soil nitrate N leaching on Immokalee sand in the non irrigated zone

PAGE 242

242 Figure A 9 Soil ammonium N leaching on Candler sand in the irrigated zone

PAGE 243

243 Figure A 10 Soil ammonium N leaching on Candler sand in the nonirrigated zone

PAGE 244

244 Figure A 11 Soil nitrate N leaching on Candler sand in the irrigated zone

PAGE 245

245 Figure A 1 2 Soil nitrate N leaching on Candler sand in the nonirrigated zone

PAGE 246

246 Figure A 1 3 Soil P leaching on Immokalee sand in the irrigated zone

PAGE 247

247 Figure A 14 Soil P leaching on Immokale e sand in the non irrigated zone

PAGE 248

248 Figure A 15 Soil P leaching on Candler sand in the irrigated zone

PAGE 249

249 Figure A 16 Soil P leaching on Candler sand in the nonirrigated zone

PAGE 250

250 Figure A 17 Soil K leaching on Immokalee sand in the irrigated zone

PAGE 251

251 Figure A 18 Soil K leaching on Immokalee sand in the non irrigated zone

PAGE 252

252 F igure A 19 Soil K leaching on Candler sand in the irrigated zone

PAGE 253

253 Figure A 20 Soil K leaching on Candler sand in the nonirrigated zone

PAGE 254

254 Figure A 21 Nitrate N leaching using water samples on Immokalee site in the irrigated zone

PAGE 255

255 Figure A 2 2 Lateral RLD distribution at the Immokalee site in June 2009 using CMP in the 0 30 cm soil depth layer. All color scales are in cm cm 3

PAGE 256

256 Figure A 23 Lateral RLD distribution at the Immokalee site in June 2009 using DOHS in the 0 30 cm soil depth layer. All color scales are in cm cm 3

PAGE 257

257 Figure A 2 4 Lateral RLD distr ibution at the Immokalee site in June 2009 using MOHS in the 0 30 cm soil depth layer. All color scales are in cm cm 3

PAGE 258

258 Figure A 2 5 Lateral RLD distribution at the Immokalee site in June 2010 using CMP in the 0 45 cm s oil depth layer. All color scales are in cm cm 3

PAGE 259

259 Figure A 26 Lateral RLD distribution at the Immokalee site in June 2010 using DOHS in the 0 45 cm soil depth layer. All color scales are in cm cm 3

PAGE 260

260 Figure A 27 Lateral RLD distribution at the Immokalee site in June 2010 using MOHS in the 0 45 cm soil depth layer. All color scales are in cm cm 3

PAGE 261

261 Figure A 28 Lateral RLD distribution as a funct ion of irrigation method at the Lake Alfred site in December 2009 using DOHS S wingle Error bars denote one standard deviation

PAGE 262

262 Figure A 2 9 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in December 2009 using DOHS C35. Error bars denote one standard deviation

PAGE 263

263 Figure A 30 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in December 2009 using CMP. Error bars denote one standard deviation

PAGE 264

264 Figure A 31 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in December 2009 using MOHS. Error bars denote one standard deviation

PAGE 265

265 Figure A 32 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in July 2010 using CMP. The color scale is in cm cm 3

PAGE 266

266 Figure A 33 Lateral RLD distribution as a f unction of irrigation method at the Lake Alfred site in July 2010 using DOHS S wingle The color scale is in cm cm 3

PAGE 267

267 Figure A 34 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in J uly 2010 using DOHS C35. The color scale is in cm cm 3

PAGE 268

268 Figure A 35 Lateral RLD distribution as a function of irrigation method at the Lake Alfred site in July 2010 using MOHS. The color scale is in cm cm 3

PAGE 269

269 Figure A 36 Average sap flow per unit land area at Lake Alfred site in July 2010. Error bars denote one standard deviation Figure A 37 Average sap flow per unit land area at Lake Alfre d site in March 2011. Error bars denote one standard deviation

PAGE 270

270 Figure A 38 Average sap flow per unit land area at Immokalee in March 2011. Error bars denote one standard deviation

PAGE 271

271 Figure A 39 Average sap flow per unit land area at Immokalee in June 2011 Error bars denote one standard deviation Figure A 40 Average sap flow per unit land area at Lake Alfred site in August September 2011. E rror bars denote one standard deviation

PAGE 272

272 Figure A 4 1 Cumulative sap flow at Lake Alfred site in July 201 0 Error bars denote one standard deviation Figure A 42 Cumulative sap fl ow at Lake Alfred site in March 2011. Error bars denote one standard deviation

PAGE 273

273 Figure A 4 3 Cumulative sap flow at Lake Alfred site in August September 2011. Error bars denote one standard deviation Figure A 44 Cumulative sap flow at Immokalee site in February March 2011. Error bars denote one standard deviation

PAGE 274

274 Figure A 4 5 Cumulative sap flow at Immokalee site in June 2011. Error bars den ote one standard deviation Figure A 46 Average index sap flow K c at the Lake Alfred site in July 2010 Error bars denote one standard deviation

PAGE 275

275 Figure A 4 7 Average index sap f low K c at the Lake Alfred site in March 2011 Error bars denote one standard deviation

PAGE 276

276 Figure A 4 8 Average index sap flow K c at the Lake Alfred site in August September 2011 Error bars denote one standard deviati on Figure A 4 9 Average index sap flow K c at the Immokalee site in February March 2011 Error bars denote one standard deviation

PAGE 277

277 Figure A5 0 Average index sap flow K c at the Imm okalee site in June 2011 Error bars denote one standard deviation

PAGE 278

278 APPENDIX B CHARACTERIZATION OF SORPTION ISOTHERMS F OR AMMONIUM N, K AND P ON THE FLATWOODS AND RIDGE SOILS The chemical characteristics of soils dominating the Flatwoods and Ridge regi ons of Florida are well described in Obreza and Collins (2008) and some were also determined in this study. The Immokalee and Candler fine sand are moderately acidic (pH ranging from 4.9 to 5.6), have low organic matter content (ranging from 0.41 to 0.61% on Immokalee fine sand and from 1.56 to 1.96% on Candler fine sand) and low cation exchange capacity (CEC) (ranging from 2 to 6 cmol (+) kg 1 ), have inorganic N in the range of 8.20 and 11.24 mg kg 1 moderate to very high P (in the range of 28.73 46.45 mg kg 1 for Immokalee sand and 112.79 115.82 mg kg 1 for Candler fine sand) and K in the range of 11.83 15.23 mg kg 1 for Immokalee fine sand and 23.03 29.70 mg kg 1 for Candler fine sand (Table B 1). The study speculates that the properties such as organic matter content and CEC are behind the adsorption processes of the nutrients in this study. Adsorption is the mechanism most commonly responsible for the retention of solutes by soils, particularly cations and phosphorus The sorption process tends to rest rict compound mobility and bioavailability (Essington, 2004). Thus, the procedure for determining the NH 4 N, P and K sorption isotherms could then provide information on their mobility in the soil The supporting electrolyte concentration is chosen to mimic that of soil solution. Most commonly 0.01 M CaCl 2 (Singh and Jones, 1975; Belmont et al., 2009), 0.01 N CaCl 2 (Bowman et al., 1981), 0.005 M CaCl 2 (Essington, 2004), 5 100 mg K L 1 KCl (Sparks et al., 1980), 0.05 M KCl (Harris et al., 1996; Zhou and Li, 2001), and 0.01 M KCl (Nair et al., 1998; Villapando and Graetz, 2001) have been used as electrolytes in studies on P and K sorption. Nair et al. (1984) reported that P sorption

PAGE 279

279 varies with ionic strength and cation species of the supporting electrol yte. For example, Nair et al. (1984) showed that P adsorption was generally lower with K + as the supporting electrolyte cation compared with Ca 2+ These studies and others have not explained the rationale behind use of a particular electrolyte other than equilibrating the solutions in deionized or tap water. This study attempted to 1) determine sorption isotherms for NH 4 + K and P on the Flatwoods and Ridge soils with the aim of predicting the mobility, availability and uptake of NH 4 + P and K in citrus pr oduction and 2) determine the effect of supporting electrolyte on P sorption We hypothesized that P adsorption and NH 4 + and K + exchange on the Flatwoods and Ridge soils do not adversely affect availability and uptake as a result of adsorption to soil co lloids. Results and Discussion The results of adsorption of K + NH 4 + N, and P are pr esented and described in Fig. B 1 through B 3, Tables B 2 and B 3, and Appendices G through I for Candler and Immokalee fine sand. Ammonium adsorption for Immokale e and Candler fine sand followed a linear isotherm with distribution coefficients (K D ) o f 1. 1 2 0.42 and 1. 64 0.25 kg L 1 and 1.66 0.39 and 1.76 0.39 kg L 1 for the 0 15 and 15 30 cm depths, respectively. P adsorption wa s described by a Freundlich model wi th linearized K D ranging from 0. 50 0. 19 to 0. 75 0. 1 3 kg L 1 for Immokalee fine sand and from 1.73 0.1 5 to 4.43 0.50 kg L 1 for Candler fine sand using a C max of 15 mg L 1 P sorption isotherm for Immokalee fine sand determined using fertilizer mixture was linear with K D averaging about 0.44 10 kg L 1 The adsorption of K + and NH 4 + was similar for both 0 15 and 15 30 cm soil depth layers while P adsorption was linear for the P concentration range studied on the Immokalee sand using fertilizer mixture Ammo nium K D was higher than that of

PAGE 280

280 potassium probably due to a larger hydrated radius in the former (ammonium ionic radius =0.56 nm and potassium ionic radius =0.53 nm). In other words more NH 4 + would be retained in the limited exchange sites of the colloidal fractions (due low organic matter content ~1 2%, and a small clay fraction ~0.5% (Obreza and Collins, 2008) while letting K + desorb into the soil solution for plant uptake. Lumbanraja and Evangelou (1990; 1994) also reported similar phenomena regarding K + and NH 4 + N on clay loam and silt loams soils of Kentucky, USA. They showed that the addition of K + stimulated the adsorption of NH 4 + N on high affinity sites while K + adsorption was suppressed by labile NH 4 + and Entisol, Wang and Alva (2000) showed that NH 4 + adsorption was great er for surface soils than that of the subsurface soils. They found that the potential NH 4 + buffering capacity was greater for Wabasso (at 0 30 and 60 90 cm) than the Candler soil (0 60 cm) owing to the presence of smectite in the former. Studies regarding ammonia sorption done over the years have yielded mixed observations. For example, Wagenet et al. (1977) assumed reversible, linear equilibrium sorption with distribution coefficients between 1 and 10 L kg 1 on a Tyndall silty loam. Yet, Rodrguez et al. (2005) found that representing ammonium adsorption desorption as a kinetic process better described their results. They noted that ammonium adsorption on the sandy clay loam soil was higher than adsorption on the loamy sand. The ammonium K D values found in this study agree with those proposed by several researchers (Wagenet et al., 1977; Selim and Iskandar, 1981; Lotse et al., 1992; Ling and El Kadi, 1998) Khakural and Alva (1996) studied transformation of urea and ammonium nitrate in an Entisol and a Spodosol under citrus production. The percentage of transformation of

PAGE 281

281 NH 4 + N into NO 3 N was 33 to 41 and 37 to 41% in the Candler fine sand and Wabasso sand, respectively, at applic ation rates of 1 g N kg 1 The rate of transformation of NH 4 in these sandy soils dictates the availability of NH 4 + N and NO 3 N forms of N for plant uptake and losses due to volatilization, leaching, and denitrification. We had speculated that some NH 4 + would volatilize and transform into NO 3 but 24 h equilibration time, under laboratory conditions at 251 o C renders this volatilization negligible while retaining the possible transformation due to nitrification. Freundlich sorption coefficients (K f ) ( A ppendix I ) were lower for Immokalee fine sand than for Candler High coefficients observed on Candler fine sand with K f eightfold greater than that of Immokalee fine sand. The K f value obtained with 0.005 M CaCl 2 was approximately twice that obtained w ith 0.01 M KCl and threefold that obtained in the fertilizer mixture suggesting the influence of the cation effect on P adsorption than with water. According to Zhou and Li (2001), the lower Freundlich sorption coefficients (K f ), indicate low P retention capacity at low P concentrations suggesting that the potential risk of subsurface P movement and leaching would be high when the concentration of P in surface soils is high. The K f and K D values reported in Appendix I are generally lower than those repo rted for carbonatic soils in south Florida (Zhou and Li, 2001) where K D ranged from 14.8 76.3 L kg 1 and K f from 12 58 mg 1 N kg 1 L N However, the results in this study agree with those of other researchers (Barrow et al., 1980; Nair et al., 1984 ; Havlin et al., 2005 ). According to Havlin et al. (2005), d ivalent cations on the CEC enhance P adsorption relative to monovalent cations because they increase the accessibility of (+) charged edges of clay minerals to P. This occurs at pH<6.5, because at great er soil pH Ca P minerals would precipitate. Barrow et al.

PAGE 282

282 (1980) also showed that at equal ionic strength below pH=6 there was more phosphate adsorption from CaCl 2 than from NaCl on goethite. This phenomenon, according to Barrow and colleagues, is caused because high concentration of positive charges near the negatively charged soil surface may be induced by replacing a monovalent cation with a divalent one and also if the added divalent cation has a specific affinity for the adsorption surface. Addition of cations from the supporting electrolyte, unlike using the fertilizer induced a greater negative charge for phosphate adsorption. The higher sorption coefficients for Candler might be due to high organic matter and some Fe/Al coatings that might bind P. This might explain, in part, wh y Mehlich 1 P was several times higher for Candler than for Immokalee fine sand as summarized in Table B 1 and discussed thoroughly in Chapter 3. The high K f value in the top 0 15 cm than the 15 30 cm layer is ascribed to higher organic carbon and organic matter in the former layer resulting in increased P adsorption. Summary The results show that P adsorption in the top 0 15 cm was greater for Candler than Immokalee sand using the fertilizer mixture, 0.00 5 M CaCl 2 a nd 0.01 M KCl. The distribution coefficients (K D ) for P estimated using 0.01 M KCl were similar to K D values determined using fertilizer mixture for Immokalee and Candler fine sand, respectively. The K D values determined using 0.00 5 M CaCl 2 as the supporti ng electrolyte were two to threefold greater than the K D of the fertilizer mixture on Immokalee and Candler fine sand suggesting that divalent Ca +2 might result in overestimation of P sorption on Candler and Immokalee sandy soils It appears the addition of a supporting electrolyte with a divalent or monovalent cation, unlike fertilizer mixture increases the surface charge for adsorption of orthophosphate anions. The adsorption isotherms of both

PAGE 283

283 ammonium and potassium w ere linear and greater for Candler than Immokalee sand probably due to Al and Fe coatings and higher organic matter in the former. For the two soils soils, a mmonium adsorption coefficients were greater than those of potassium.

PAGE 284

284 Table B 1 Selected soil chemical characteristics for Immokal ee and Candler sand Soil Soil depth (cm) pH OM § CEC NH 4 + NO 3 M1P M1K IN Immokalee 0 15 5.6 0.61 2 6 3.45 4.93 46.45 15.23 8.37 Immokalee 15 30 5.2 0.41 2 6 2.32 4.07 28.73 11.83 6.40 Candler 0 15 5.3 1.96 2 4 2.55 8.69 115.82 29.70 11.24 C andler 15 30 4.9 1.56 2 4 2.88 5.31 112.79 23.03 8.20 Soil to water ratio=1:2 (mass/volume), § OM organic matter expressed as a percentage, CEC cation exchange capacity expressed in cmol(+) kg 1 (CEC reported by Obreza and Collins, 2008) Mehlich 1 P ( mg kg 1 ), Mehlich 1 K (mg kg 1 ), IN=Inorganic N (mg kg 1 )

PAGE 285

285 Table B 2 Sorption coefficients for NH 4 + and K + on Immokalee and Candler fine sand using fertilizer mixture in tap water Soil Depth (cm) NH 4 + K + K D (L kg 1 ) K D (L kg 1 ) Immokalee 0 15 1 .12 0.42 0.91 0.38 Immokalee 15 30 1.64 0.25 0.87 0.74 Candler 0 15 1.66 0.39 1.65 0.56 Candler 15 30 1.76 0.39 0.93 0.28 K D =Mean one standard deviation of 3 replications Table B 3. Sorption coefficients for P on Immokalee and Candler fine sand Soil Depth (cm) Supporting e lectrolyte K D (L kg 1 ) Immokalee 0 15 0.01 M KCl Immokalee 15 30 0.01 M KCl Candler 0 15 0.01 M KCl Candler 15 30 0.01 M KCl Immokalee 0 15 0.005 M CaCl 2 Immokalee 1 5 30 0.005 M CaCl 2 Candler 0 15 0.005 M CaCl 2 Candler 15 30 0.005 M CaCl 2 Immokalee 0 15 Fertilizer mixture Immokalee 15 30 Fertilizer mixture Candler 0 15 Fertilizer mixture Can dler 15 30 Fertilizer mixture K D =Linearized K D using Equation 6 6 presented as m ean one standard deviation of 3 replications and a C max of 15 mg L 1

PAGE 286

286 Figure B 1. Selected linear isotherms for NH 4 + for I mmokalee and Candler sand

PAGE 287

287 Figure B 2 Selected Freundlich isotherms for P for Immokalee and Candler sand using 0.005M CaCl 2

PAGE 288

288 Figure B 3 Selected linear and Freundlich isotherms for P for Immokalee and Candler sand using 0.01M KCl

PAGE 289

289 APPENDIX C SOIL WATER CHARACTER ISTIC CURVE AND HYDR AULIC FUNCTIONS FOR THE IMMOKALEE AND CANDLE R SAND The soil properties that determine the behavior of soil water flow systems are the hydraulic conductivity and water retention characteristics. The relation between soil water content and the soil water suction is a fundamental part of the characterization of the hydraulic properties of soil (Klute, 1986). The conductivity of a soil depends on po re geometry and the properties of the fluid flowing through or retained in the pores. Viscosity and density are the two properties that directly affect hydraulic conductivity while soil porosity and water retention function are determined by soil texture a nd Law (Klute and Dirksen, 1986; Hillel, 1998) which for one dimensional vertical flow may be written as: (C 1 ) where q is the volume flux density, is the gradient of the hydraulic head H, and gradient of the hydraulic head composed of the gravi tational head, z, and the pressure head, h, mathematically given as: H=h+z (C 2 ) Mualem (1986) also explained that there are some independent variables of interest that describe soil water retention characteristics such as the degree of saturatio n (S), effective water content ( effective saturation could also be used to describe water retention characteristics.

PAGE 290

290 The amount of water retained in the soil at any given moment is dependent upon factors such as the type of plant cover, plant density, stage of plant growth, rooti ng depth, evaporation and transpiration rates, amount of water infiltrated, rate of wetting, nature of horizonation and the length of time since the last irrigation or rainfall event (Cassel and Nielsen, 1986). Amount of water available for plant use is de termined through estimation of available water capacity, field capacity and permanent wilting point. The traditional field capacity for well drained sandy soil under laboratory conditions is estimated at 10 kPa of soil water tension for a sandy soil and 33 kPa for medium or fine textured soil (Obreza et al., 1997). However, in their study on soil water holding characteristic on Florida Flatwoods and Ridge soils, Obreza and co workers showed that soil water tension of 5 kPa would be appropriate for the Ridg e and 8 kPa for the Flatwoods soil due to their inherent differences in porosity, conductivity and horizonation. Thus, the objective s of the laboratory experiments were to 1) determine water retention characteristics for the Immokalee and Candler sand and 2) calculate hydraulic parameters for use in HYDRUS model We hypothesized basing on literature and field observations that the soil water retention characteristics for the two sites would vary as a function of soil depth. Thus, it would be important to sample by depths of interest at each study site for use of selected site specific parameters in the simulation model. Results and Discussion The volumetric moisture contents at soil tensions ranging from 0 100 kPa (0 1020 cm) are presented in Fig. C 1 a nd C 2 The Van Genuchten model water retention parameters ( n and l ) are documented in Table C 1. The saturated and residual moisture contents, moisture contents at field capacity (10 kPa), available soil water

PAGE 291

291 content saturated hydraulic conductivity and bulk density are presented in Table C 2 The residual moisture contents from literature are 0.013 and 0.009 cm 3 cm 3 for Immokalee and Candler fine sand (Carlisle et al., 19 89 ). The saturated hydraulic conductivity ranged from 13.22 to 15.82 cm h 1 o n Immokalee and 14.76 to 15.94 cm h 1 on Candler fine sand. Field capacities averaged 0. 096 and 0. 09 3 cm 3 cm 3 for the two soils. Available water capacities ranged from 0.0 77 to 0. 087 and 0.0 65 to 0. 095 cm 3 cm 3 for Immokalee and Candler fine sand. The available water capacities were estimated using soil tensions of 10 kPa as field capacity and 1500 kPa as wilting point. The results suggested very high hydraulic conductivities, good drainage and permeability for both soils due to the strong sandy soil c haracteristic in the top 0.60 m soil depth. The soil desorption curves also indicate large soil pore sizes and a narrow pore size distribution in both soils (Klute, 1986; Klute and Dirksen, 1986; Obreza et al., 1997; Obreza and Pitts, 2002). The high hydra ulic conductivity values suggest the importance of careful water and nutrient management due to the potential threat of nutrient leaching and downward drainage of water beyond the plant root zone. Summary The soil the hydraulic conductivity and water retention characteristics are important for better nutrient and water management particularly in fertigated and irrigated systems. The experiment yielded important site specific parameters like alpha, n, m field capacity, available water capacity and hydr aulic conductivity for use in the HYDRUS 2D model to describe water and solute transport to aid decision making in predicting env ironmental fate of fertilizers.

PAGE 292

292 Figure C 1. Measured and simulated s oil water release cur ves for Candler fine sand

PAGE 293

293 Figure C 2 Measured and simulated s oil water release curves for Immokalee fine sand

PAGE 294

294 Table C 1 Soil water retention parameters of Immokalee and Candler fine sand estimated using CAS software developed by Bloom (2009) Soil Depth (cm) 1 ) n m l Immokalee 0 15 0.03 1.87 0.47 0.5 Immokalee 15 30 0. 0 4 1.29 0.23 0.5 Immokalee 30 45 0.03 2.06 0.52 0.5 Immokalee 45 60 0.03 1.71 0.42 0.5 Candler 0 15 0.03 2.22 0.55 0.5 Candler 15 30 0.04 1 .70 0.41 0.5 Candler 30 45 0.02 2.50 0.60 0.5 Candler 45 60 0.02 1.82 0.45 0.5 Pore connectivity parameter (estimated to be an average of 0.5 for many soils) (Simunek et al., 2007)

PAGE 295

295 Table C 2 Soil physical characteristics of the Immokalee and Candler fine sand Soil Depth (cm) Bulk density (g cm 3 ) K sat (cm h 1 ) § sat (cm 3 cm 3 ) §§ r (cm 3 cm 3 ) FC (cm 3 cm 3 ) AWC (cm 3 cm 3 ) Immokalee 0 15 1.62 15.82 0.343 0.013 0.090 0.077 Immokalee 15 30 1.62 13.97 0.362 0.013 0.100 0.087 Immokalee 30 45 1.59 13.22 0.390 0.013 0.100 0.087 Immokalee 45 60 1.61 14.57 0.318 0.01 3 0.095 0.082 Candler 0 15 1.65 15.53 0.362 0.009 0.074 0.065 Candler 15 30 1.64 15.94 0.330 0.009 0.104 0.095 Candler 30 45 1.57 14.76 0.313 0.009 0.100 0.091 Candler 45 60 1.68 15.73 0.421 0.009 0.094 0.085 K sat Saturated hydraulic conductivity § sat Saturated moisture content §§ r Residual moisture content obtained from Obreza, unpublished data FC Field capacity at 10 kPa AWC Available water content

PAGE 296

296 APPENDIX D A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF DRIP OPEN HYDR OPONIC SYSTEM AT IMMOKALEE IN 2009 t ree, sampling position, dripper, drip line, spacing between trees=3.05m, row spacing=6.71m, positions below the dripper within the sampling grid were also sampled 6.71 m 0.15 m 0.30m 3.05m 0.45m 0.15m 3.05m m

PAGE 297

297 A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF DRIP OPEN HYDR OPONIC SYSTEM AT IMM OKALEE IN 2010 AND THEREAFT ER t ree, sampling position, dripper, drip line, spacing between trees=3.05m, row spacing=6.71m, positions below the dripper within the sampling grid were also sampled 6.71 m 0. 15 m 0. 15m 0.45m 0.30m 3.05m

PAGE 298

298 A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF MICROSPRINKLER OPEN HYDROPONIC SYST EM ON IMMOKALEE SAND t ree, sampling position irrigation main line, microsprinkler emitter, spacing between trees= 3.05m, row spacing=6.71m, area between the dashed lines was the irrigated zone 6.71 m 0.15 m 0.15m 0.30m 0.45m 6.71 m 0.40m 3.05m 0.10m

PAGE 299

299 A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF CONVENTIONAL M IC ROSPRINKLER SYSTEM O N IMMOKALEE SAND t ree, sampling position irrigation main line, m icrosprinkler emitter, spacing between trees=3.05m, row spacing=6.71m, area wi thin the dashed circle was the irrigated zone 6.71 m 0.15m 0.15m 0.30m 0.45m 6.71m 0.10m 3.05m

PAGE 300

300 A SCHEMATIC FIELD DI AGRAM SHOWING THE SE T UP OF DRIP OPEN HYDR OPONIC SYSTEM (DOHS SWINGLE) ON CANDLER SAND t ree, sampling position, dripper, drip line, spacing between trees=3.05m, row spacing=6.10m, positions below the dripper within the sampling grid were also sampled 6.10 m 0.15 m 0.45m 0.15m 0.30m m 3.05m

PAGE 301

301 A SCHEMATIC FIE LD DIAGRAM SHOWING T HE SET UP OF DRIP OPEN HYDROPONIC SYST EM (DOHS C 35) ON CANDLER SAND t ree, sampling position, dripper, drip line, spacing between trees=2.44m, row spacing=5.49m, positions below the dripper within the sampling grid were also sampled 5.49m 0.45m 0.15m 0.30m 0.15m 2.44m

PAGE 302

302 A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF MICROSPRINKLER OPEN HYDROPONIC SYST EM ON CANDLER SAND tree, sampling position irrigation main line, Microsprinkler emitter, spacing between trees=3.05m, row spacing=6.10m, ar ea between the dashed lines was the irrigated zone 6.71 m 0.15m 0.15m 0.30m 0.45m 6.10m 0.40m 3.05m 0.10m

PAGE 303

303 A SCHEMATIC FIELD DIAGRAM SHOWIN G THE SET UP OF CONVENTIONAL M ICROSPRINKLER SYSTEM ON CANDLER SAND tree, sampling positio n irrigation main line, Microsprinkler emitter, spacing between trees=3.05m, row spacing=6.10m, area within the dashed circle was the irrigated zone 6.71 m 0.15 m 0.15m 0.30m 0.45m 6.10m 0.10m 3.05m

PAGE 304

304 A P P E N D I X E AVERAGE MONTHLY TEMP ERATURE, RELATIVE HU MIDITY, RAINFALL, SO LAR RADIATI ON AND EVAPOTRANSPIRATION A T IMMOKALEE SOURCED FROM THE FLORIDA AUT OMATED WEATHER NETWO RK (HTTP://FAWN.IFAS.UF L.EDU/) FROM 2009 TO 2011 Month Average temperature ( o C) Minimum temperature ( o C) Maximum temperature ( o C) Relative Humidity (%) Rainfall (mm) Sol ar radiation (W m 2 ) Evapotranspiration (mm d 1 ) January 15.51.4 1.81.7 28.90.3 77.72.1 35.829.5 154.35.7 1.50.0 February 16.92.5 1.13.3 28.91.2 74.74.5 27.935.3 193.816.8 2.40.3 March 19.02.0 1.92.4 31.62.0 74.02.0 94.0111.7 230. 412.3 3.00.3 April 22.91.3 9.63.1 32.72.1 73.03.0 94.090.3 267.314.8 4.10.4 May 25.51.1 15.84.0 35.10.6 76.01.7 116.464.9 277.421.0 4.80.5 June 27.01.3 19.72.8 36.00.5 80.02.0 202.2122.5 259.812.7 4.80.3 July 27.50.7 21.61.0 35 .70.4 83.01.0 137.851.3 239.83.8 4.60.0 August 27.50.5 22.41.1 36.10.6 85.01.0 133.98.9 223.813.6 4.30.3 September 26.90.4 20.50.8 34.70.9 84.71.5 138.953.4 213.92.8 3.90.1 October 24.21.1 9.91.3 33.51.3 79.33.5 74.6120.4 195.22 0.5 3.00.4 November 20.80.1 6.82.5 32.00.7 79.72.1 21.320.6 166.98.9 2.00.0 December 17.03.6 0.22.4 28.92.5 79.74.9 45.040.8 143.416.3 1.40.1 Mean 1 standard deviation

PAGE 305

305 AVERAGE MONTHLY TEMP ERATURE, RELATIVE HU MIDITY, RAINFALL, SOLAR RADI ATION AND EVAPOTRANSPIRATION A T LAKE ALFRED SOURCE D FROM THE FLORIDA A UTOMATED WEATHER NET WORK (HTTP://FAWN.IFAS.UF L.EDU/) FROM 2009 TO 2011 Month Average temperature ( o C) Minimum temperature ( o C) Maximum temperature ( o C) Relative Humi dity (%) Rainfall (mm) Solar radiation (W m 2 ) Evapotranspiration (mm d 1 ) January § 13.81.3 2.31.0 27.60.9 74.02.6 61.130.1 136.35.9 1.40.1 February 15.32.8 0.02.8 28.22.2 71.74.5 30.135.6 168.510.8 2.10.3 March 18.52.1 3.92.0 30.82. 8 70.32.5 149.7139.3 215.14.5 2.80.3 April 22.61.1 9.03.3 33.41.4 69.72.5 32.942.8 261.820.9 4.10.4 May 25.50.6 16.91.9 35.10.5 73.04.4 111.6102.6 268.035.0 4.70.5 June 27.50.5 20.20.5 36.90.6 76.71.5 137.359.0 258.514.1 4.80.3 July 27.70.3 21.20.0 36.00.4 80.01.0 110.835.3 235.412.1 4.60.3 August 27.60.1 22.60.5 35.90.1 82.71.5 249.263.4 220.67.7 4.20.1 September 26.50.2 18.91.5 34.40.3 81.30.6 109.842.5 210.211.0 3.70.1 October 23.11.6 10.52.4 33.02. 3 76.33.8 76.4128.1 193.626.1 2.90.3 November 19.50.4 6.51.6 30.20.4 78.31.5 23.017.0 149.313.8 1.80.0 December 15.24.2 0.94.2 28.12.2 78.06.1 41.041.5 128.324.6 1.30.3 § Mean 1 standard deviation

PAGE 306

306 APPENDIX F CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT CREC FOR ROOT DIAMETER LE SS THAN 0.5 MM

PAGE 307

307 CORRELATIONS BETWEEN RLD MEA SURED BY LINE INTERS ECTION METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT CREC FOR ROOT DIAMETER 0. 5 1.0 MM

PAGE 308

308 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT CREC FOR ROOT DIAMETER 1. 0 3.0 MM

PAGE 309

309 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTIO N METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT CREC FOR ROOT DIAMETER GREATER THAN 3 0 MM

PAGE 310

310 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT SWFR EC FOR ROOT DIAMETER LE SS THAN 0.5 MM

PAGE 311

311 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METH OD AND PREDICTED RLD BY SCANNING METHOD AND SCANNED AREA AT SWFR EC FOR ROOT DIAMETER 0. 5 1.0 MM

PAGE 312

312 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PRED ICTED RLD BY SCANNIN G METHOD AND SCANNED AREA AT SWFREC FOR ROOT DIAMETER 1. 0 3.0 MM

PAGE 313

313 CORRELATIONS BETWEEN RLD MEASURED BY LINE INTERSECTION METHOD AND PREDICTED RLD B Y SCANNING METHOD AN D SCANNED AREA AT SW FREC FOR ROOT DIAMETER GREATER THAN 3.0 MM

PAGE 314

314 APPENDIX G EXPERIMENTAL SET UP FOR THE SORPTION STU DY Ion/Element Co 1 g/mL Co 2 g/mL Co 3 g/mL P 10 25 50 P § 10 25 50 P §§ 5 25 50 K + 6 32 63 NH 4 + 6 32 64 Case A: In 20mL of 0.01M KCl § Case B: In 20mL of 0.005M CaCl 2 §§ Case C: In 20mL of fertilizer mixture

PAGE 315

315 APPENDIX H SORPTION COEFFICIENT S FOR NH 4 + AND K + ON IMMOKALEE AND CAN DLER FINE SAND USING FERTILIZER MIXTURE I N TAP WATER Soil Depth (cm) NH 4 + K + K D R 2 K D R 2 Immokalee 0 15 1.00 0.9 8 1.35 0.98 Immokalee 0 15 1.16 0.9 8 0.69 0.99 Immokalee 0 15 1.1 9 0.9 8 0.70 0.99 Immokalee 15 30 1. 9 5 0.9 4 1.70 0.99 Immokalee 15 30 1.6 2 0. 96 0.65 0.95 Immokalee 15 30 1. 3 5 0.9 9 0.26 0.93 Candler 0 15 1. 5 2 0.98 2.08 0.99 Candler 0 15 2.04 0. 98 1.01 0.95 Candler 0 15 1. 41 0.97 1.85 0.90 Candler 15 30 2. 20 1.00 1.05 0.99 Candler 15 30 1. 64 0. 98 1.13 0.89 Candler 15 30 1.44 0. 97 0.61 0.99

PAGE 316

316 APPENDIX I SORPTION COEFFICIENT S FOR P ON IMMOKALEE AND CANDLER FINE SAN D Soil Electrolyte Depth (cm) K f (mg 1 N kg 1 L N ) N K D (L kg 1 ) R 2 Immokalee 0.01 M KCl 0 15 2.60 0.47 0.62 0.99 Immokalee 0.01 M KCl 0 15 1.36 0.56 0.41 0.99 Immokalee 0.01 M KCl 0 15 3.98 0.28 0.57 0.99 Immokalee 0.01 M KCl 15 30 1.56 0.55 0.46 0.98 Immokalee 0.01 M KCl 15 30 2.24 0.30 0.34 0.93 Immokalee 0.01 M KCl 15 30 2.79 0.49 0.70 0.99 Candler 0.01 M KCl 0 15 9.01 0.61 3.13 0.99 Candler 0.01 M KCl 0 15 7.41 0.58 2.3 8 0.96 Candler 0.01 M KCl 0 15 11.7 0 0.51 3.10 0.97 Candler 0.01 M KCl 15 30 19.33 0.48 4.73 0.98 Candler 0.01 M KCl 15 30 15.66 0.46 3.63 0.99 Candler 0.01 M KCl 15 30 13.04 0.46 3.02 0.97 Immokalee 0.005 M CaCl 2 0 15 5.00 0.36 0.88 0.91 Immokalee 0 .005 M CaCl 2 0 15 3.25 0.45 0.73 0.98 Immokalee 0.005 M CaCl 2 0 15 2.18 0.54 0.63 0.96 Immokalee 0.005 M CaCl 2 15 30 3.56 0.42 0.74 0.99 Immokalee 0.005 M CaCl 2 15 30 1.85 0.45 0.42 0.94 Immokalee 0.005 M CaCl 2 15 30 3.97 0.51 1.05 0.95 Candler 0.005 M CaCl 2 0 15 29.85 0.21 3.51 0.97 Candler 0.005 M CaCl 2 0 15 18.04 0.31 2.78 0.97 Candler 0.005 M CaCl 2 0 15 30.31 0.26 4.09 0.99 Candler 0.005 M CaCl 2 15 30 31.02 0.23 3.86 0.99 Candler 0.005 M CaCl 2 15 30 35.69 0.25 4.68 1.00 Candler 0.005 M CaCl 2 1 5 30 33.39 0.28 4.75 0.96 Immokalee Fertilizer mixture 0 15 NA NA 0.56 1.00 Immokalee Fertilizer mixture 0 15 NA NA 0.41 0.95 Immokalee Fertilizer mixture 0 15 NA NA 0.38 0.99 Immokalee Fertilizer mixture 15 30 NA NA 0.26 0.91 Immokalee Fertilizer mix ture 15 30 NA NA 0.37 0.86 Immokalee Fertilizer mixture 15 30 NA NA 0.65 0.98 Candler Fertilizer mixture 0 15 5.17 0.57 1.61 0.95 Candler Fertilizer mixture 0 15 7.84 0.43 1.67 0.94 Candler Fertilizer mixture 0 15 15.22 0.23 1.89 0.91 Candler Fertiliz er mixture 15 30 2.73 0.69 1.18 0.98 Candler Fertilizer mixture 15 30 14.22 0.28 2.02 0.94 Candler Fertilizer mixture 15 30 14.6 0.41 2.95 0.99 K D Linearized K D estimated using a C max of 1 5 mg L 1 for Immokalee and Candler fine sand

PAGE 317

317 LIST OF REFERENCE S Abrisqueta, J., O. Mounzer, S. Alvarez, W. Conejero, Y. Garcia Orellana, L. Tapia, J. Vera, I. Abrisqueta, and M. Ruiz Sanchez. 2008. Root dynamics of peach trees submitted to partial rootzone drying and continuous deficit irrigation. Agric. Water Manage 95:959 967. Afyuni, M., and M. Wagger. 2006. Soil physical properties and bromide movement in relation to tillage system. Commun. Soil. Sci. Plant Anal. 37:541 556. Agyin Birikorang, S., G. O'Connor, and S. Brinton. 2008. Evaluation phosphorus loss from a Florida Spodosol as affected by phosphorus source application methods. J. Environ. Qual. 37:1180 1189. Ahuja, L., D. Decoursey, B. Barnes, and K. Rojas. 1993. Characteristics of macropore transport studied with the ARS root zone water quality model. Tran s. ASAE 36:369 380. Allen, R.G., L.S. Pereira, D. Raes, and M. Smith. 1998. Crop evapotranspiration Guidelines for computing crop water requirements. FAO, Rome. Alva, A. 1993. Comparison of Mehlich 3, Mehlich 1, Ammonium bicarbonate DTPA, 1.0M Ammonium ace tate, and 0.2M Ammonium chloride for extraction of calcium, magnesium, phosphorus, and potassium for a wide range of soils. Commun. Soil. Sci. Plant Anal. 24:603 612. Alva, A., and J. Syvertsen. 1991. Irrigation water salinity affects soil nutrient distrib ution, root density, and leaf nutrient levels of citrus under drip fertigation. J. Plant Nutr. 14:715 727. Alva, A., S. Paramasivam, and W. Graham. 1998. Impact of nitrogen management practices on nutritional status and yield of Valencia orange trees and g roundwater nitrate. J. Environ. Qual. 27:904 910. Alva, A.K., and A. Fares. 1998. A new technique for continuous monitoring of soil moisture content to improve citrus irrigation. Proc. Fla. State Hort. Soc. 111:113 117. Alva, A., O. Prakash, A. Fares, and A. Hornsby. 1999. Distribution of rainfall and soil moisture content in the soil profile under citrus tree canopy and at the dripline. Irrig. Sci. 18:109 115. Alva, A., S. Paramasivam, W. Graham, and T. Wheaton. 2003. Best nitrogen and irrigation managemen t practices for citrus production in sandy soils. Water Air Soil Pollut. 143:139 154.

PAGE 318

318 Alva, A., K, S. Paramasivam, A. Fares, J.A. Delgado, D.J. Mattos, and K.S. Sajwan. 2005. Nitrogen and irrigation management practices to improve nitrogen uptake efficienc y and minimize leaching losses. J. Crop Improvement 15:369 420. Alva, A.K., and D.P.H. Tucker. 1997. Impact of soil pH and cations availability on citrus production in low buffered soils. Trends Soil Sci. 2:21 37. Alva, A.K., and S. Paramasivam. 1998. An evaluation of nutrient removal by citrus fruits. Proc. Fla. State Hort. Sci. 111:126 128. Alva, A.K., S. Paramasivam, T.A. Obreza, and A.W. Schumann. 2006a. Nitrogen best management practice for citrus trees I. Fruit yield, quality, and leaf nutritional status. Sci. Hortic. 107:233 244. Alva, A.K., S. Paramasivam, A. Fares, T.A. Obreza, and A.W. Schumann. 2006b. Nitrogen best management practice for citrus trees II. Nitrogen fate, transport, and components of N budget. Sci. Hortic. 109:223 233. Alva, A. K., D.J. Mattos, S. Paramasivam, B. Patil, H. Dou, and K.S. Sajwan. 2006c. Potassium management for optimizing citrus production and quality. Int. J. Fruit Sci. 6:3 43. Anderson, D., and L. Henderson. 1986. Sealed chamber digestion for plant nutrient anal yses. Agron. J. 78:937 939. Anderson, D., and L. Henderson. 1988. Comparing sealed chamber digestion with other digestion methods used for plant tissue analysis. Agron. J. 80:549 552. Andreu, L., J. Hopmans, and L. Schwankl. 1997. Spatial and temporal dist ribution of soil water balance for a drip irrigated almond tree. Agricult. Water Manag. 35:123 146. Angelakis, A., D. Rolston, T. Kadir, and V. Scott. 1993. Soil water distribution under trickle source. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 119:484 500. Arias Reveron, J., and H. Browning. 1995. Development and mortality of the citrus snow scale (homoptera, diaspididae) under constant temperature and relative humidity. Environ. Entomol. 24:1189 1195. Aubert, B., M. Grisoni, M. Villemin, and G. Rossolin. 1 996. A case study of Huanglongbing (greening) control in Reunion. p. 276 278 In J. V. da Graca, et al. (eds.) Proc. 13th Conference of the International Organization of Citrus Virologists (IOCV), University of California, Riverside1996. Bacon, P., and B. D avey. 1989. Nutrient availability under trickle irrigation phosphate fertilization. Fert. Res. 19:159 167.

PAGE 319

319 Badr, A.E., and M.E. Abuarab. 2011. Soil moisture distribution patterns under surface and subsurface drip irrigation systems in sandy soil using neut ron scattering technique. Irrigation Science DOI 10.1007/s00271 011 0306 0. Badr, M.A. 2007. Spatial distribution of water and nutrients in the root zone under surface and subsurface drip irrigation and cantaloupe yield. World J. Agric. Sci. 3:747 756. B aker, J., and C. Vanbavel. 1987. Measurement of mass flow of water in the stems of herbaceous plants. Plant Cell Environ. 10:777 782. Baker, J., and J. Nieber. 1989. An analysis of the steady state heat balance method for measuring sap flow in plants. Agri c. For. Meteorol. 48:93 109. Barnette, R.M., E.F. DeBusk, J.B. Hester, and W.W. Jones. 1931. The mineral analysis of nineteen year old Marsh seedless grapefruit tree. Citrus Ind. 12:5 6. Barraclough, P., and R. Leigh. 1984. The growth and activity of winte r wheat roots in the field the effect of sowing date and soil type on root growth of high yielding crops. J. Agric. Sci. 103:59 74. Barrow, N., J. Bowden, A. Posner, and J. Quirk. 1980. Describing the effects of electrolyte on adsorption of phosphate by a variable charge surface. Aust. J. Soil Res. 18:395 404. Baryosef, B., and M. Sheikholslami. 1976. Distribution of water and ions in soils irrigated and fertilized from a trickle source Soil Sci. Soc. Am. J. 40:575 582. Baryosef, B., and B. Akiri. 1978. S odium bicarbonate extraction to estimate nitrogen, phosphorus, and potassium availability in soils. Soil Sci. Soc. Am. J. 42:319 323. Baryosef, B., S. Schwartz, T. Markovich, B. Lucas, and R. Assaf. 1988. Effect of root volume and nitrate solution concentr ation on growth, fruit yield, and temporal N and water uptake rates by apple trees. Plant Soil 107:49 56. Bassoi, L.H., J.W. Hopmans, L.A.C. Jorge, C.M. Alencar, and J.A.M. Silva. 2003. Grapevine root distribution in drip and microsprinkler irrigation. Sc i. Agric. 60:377 387. Bauer, L.R., and K.J. Hamby. 2003. Relative sensitivities of existing and novel model parameters in atmospheric tritium dose estimates. Radiat. Prot. Dosimetry 37:253 260. Bauer, M.G., W.S. Castle, B.J. Boman, T.A. Obreza, and E.W. S tover. 2004. Root systems of healthy and declining citrus trees on swingle citrumelo rootstock growing in the southern Florida flatwoods. Proc. Fla. State Hort. Soc. 117:103 109.

PAGE 320

320 Belmont, M., J. White, and K. Reddy. 2009. Phosphorus Sorption and Potential Phosphorus Storage in Sediments of Lake Istokpoga and the Upper Chain of Lakes, Florida, USA. J. Environ. Qual. 38:987 996. Bland, W., and M. Mesarch. 1990. Counting error in the line intercept method of measuring root length. Plant Soil 125:155 157. Bloo m, S.A. 2009. Water release curve fitting using CAS. Gainesville, Florida, USA Boesch, D., R. Brinsfield, and R. Magnien. 2001. Chesapeake Bay eutrophication: Scientific understanding, ecosystem restoration, and challenges for agriculture. J. Environ. Qual 30:303 320. Bogren, K. and P. Smith. 2003. Determination of bromide by flow injection analysis. QuickChem Method 10 135 214 2 B. Lachat Instruments 5600 Lindburgh Drive, Loveland, Colorado, USA. Boivin, A., J. Simunek, M. Schiavon, and M. van Genuchten. 2006. Comparison of pesticide transport processes in three tile drained field soils using HYDRUS 2D. Vadose Zone J. 5:838 849. Boland, A., P. Jerie, P. Mitchell, and I. Goodwin. 2000. Long term effects of restricted root volume and regulated deficit irriga tion on peach: I. Growth and mineral nutrition. J. Am. Soc. Hortic. Sci.125:135 142. Boman, B.J. 1994. Evapotranspiration by young Florida flatwoods citrus trees. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 120:80 88. Bove, J. 2006. Huanglongbing: A destructi ve, newly emerging, century old disease of citrus. J. Plant Pathol. 88:7 37. Bowman, B.T. 1982. Conversion of Freundlich adsorption K values to the mole fraction format and the use of SY values to express relative adsorption of pesticides. Soil Sci. Soc. A m. J. 46:740 743. Bowman, R., M. Essington, and G. Oconnor. 1981. Soil sorption of nickel influence of solution composition. Soil Sci. Soc. Am. J. 45:860 865. Brooks, R.H., and A.T. Corey. 1964. Hydraulic properties of porous media Hydrology paper No 3, Colorado State University, Fort Collins. Bryla, D., T. Trout, J. Ayars, and R. Johnson. 2003. Growth and production of young peach trees irrigated by furrow, microjet, surface drip, or subsurface drip systems. HortScience 38:1112 1116. Bryla, D., E. Dickso n, R. Shenk, R. Johnson, C. Crisosto, and T. Trout. 2005. Influence of irrigation method and scheduling on patterns of soil and tree water status and its relation to yield and fruit quality in peach. HortScience 40:2118 2124.

PAGE 321

321 Bufon, V.B., R.J. Lascano, C. Bednarz, J.D. Booker, and D.C. Gitz. 2011. Soil water content on drip irrigated cotton: comparison of measured and simulated values obtained with the Hydrus 2 D model. Irrig. Sci. DOI: 10.1007/s00271 011 0279. Calvert, D.V., H.W. Ford, E.H. Stewart, and F .G. Martin. 1977. Growth response of twelve citrus rootstock scion combinations on a Spodosol modified by deep tillage and profile drainage. Int. Soc. Citric. 1:79 84. Cameron, S., and O. Compton. 1945. Nitrogen in bearing orange trees. Proc. Am. Soc. Hort ic. Sci. 46:60 68. Cameron, S.H., and D. Appleman. 1935. The distribution of total nitrogen in the orange tree. Proc. Am. Soc. Hortic. Sci. 30:341 348. Camp, C. 1998. Subsurface drip irrigation: A review. Trans. ASAE 41:1353 1367. Campbell, G. 1974. Simple method for determining unsaturated conductivity from moisture retention data. Soil Sci. 117:311 314. Carlisle, V.W., F. Sodek, M. Collins, L.C. Hammond, and W.G. Harris. 1989a. Characterization data for selected Florida Soils. Gainesville, University of F lorida, Florida. Carlisle, V.W., F. Sodek, M. Collin, L.C. Hammond, and W.G. Harris. 1989b. Characterization data for selected Florida soils. University of Florida, Gainesville, FL. Carrasco, G., O. Marquez, M. Urrestarazu, and M.C. Salas. 2003. Transplant s grown hydroponically are an alternative for soil. Acta Hortic. :407 410. Carsel, R., and R. Parrish. 1988. Developing joint probability distributions of soil water retention characteristics. Water Resour. Res. 24:755 769. Cassel, D.K., and D.R. Nielsen. 1986. Field capacity and available water capacity, p. 901 926, In A. Klute, (ed.) Methods of Soil Analysis Part 1 2nd edition. ed. Agronomy Ser. 9. Madison, Wisconsin, USA. Castel, J.R., I. Bautista, C. Ramos, and G. Cruz. 1987. Evapotranspiration and irri gation efficiency of mature orange orchards in Valencia (Spain). Irrig. Drain. Syst. 3:205 217. Castle, W.S. 1980. Citrus rootstocks for tree size control and higher density plantings in Florida. Proc. Fla. State Hort. Soc. 93:24 27. Castle, W.S. and A.H. Krezdorn. 1975. Effect of citrus rootstocks on root distribution 4.

PAGE 322

322 Charanjeet, S., and D.K. Das. 1985. Nitrate nitrogen distribution in soil and nitrogen uptake by wheat u nder varying levels of water and nitrogen supply. Ann. Agric. Res. 6:91 97. Chung, K.R., and R.H. Brlansky. 2009. Citrus diseases exotic to Florida: Huanglongbing (Citrus Greening), PP210 ed. UF/IFAS Extension Cooperative Service, Gainesville, Florida. Cla rk, C., and A. Richardson. 2002. Biomass and mineral nutrient partitioning in a developing tamarillo (Cyphomandra betacea) crop. Sci. Hortic. 94:41 51. Clemente, R., R. Dejong, H. Hayhoe, W. Reynolds, and M. Hares. 1994. Testing and comparison of 3 unsatur ated soil water flow models. Agric. Water Manage. 25:135 152. Clothier, B., and T. Sauer. 1988. Nitrogen transport during drip fertigation with urea. Soil Sci. Soc. Am. J. 52:345 349. Clothier, B., T. Sauer, and S. Green. 1988. The movement of ammonium nit rate into unsaturated soil during unsteady absorption. Soil Sci. Soc. Am. J. 52:340 345. Coleman, M. 2007. Spatial and temporal patterns of root distribution in developing stands of four woody crop species grown with drip irrigation and fertilization. Plan t Soil 299:195 213. Collins, R., P. Gregory, H. Rowse, A. Morgan, and B. Lancashire. 1987. Improved methods of estimating root length using a photocopier, a light box and a bar code reader. Plant Soil 103:277 280. Consoli, S., N. O'Connell, and R. Snyder. 2006. Estimation of evapotranspiration of different sized navel orange tree orchards using energy balance. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 132:2 8. da Graca, J., and L. Korsten. 2004. Citrus huanglongbing: Review, present status and future strateg ies. Diseases Fruit Vegetables 1:229 245. Dasberg, S. 1987. Nitrogen fertilization in citrus orchards. Plant Soil 100:1 9. Davenport, J., R. Stevens, and K. Whitley. 2008. Spatial and temporal distribution of soil moisture in drip irrigated vineyards. Hor tScience 43:229 235. De Jong, R., G.C. Topp, and W.D. Reynolds. (eds.) 1992. The use of measured and estimated hydraulic properties in the simulation of soil and water movement a case study. US Department of Agriculture and Department of Soil and Environ mental Sciences, University of California, Riverside.

PAGE 323

323 De Silva, H., A. Hall, D. Tustin, and P. Gandar. 1999. Analysis of distribution of root length density of apple trees on different dwarfing rootstocks. Ann. Bot. 83:335 345. Decoursey, D.G., and K.W. R ojas. 1990. RZWQM A model for simulating the movement of water and in solutes in the root zone. p. 813 821 In D. G. Decoursey (ed.) Proc. International Symposium on Water Quality Modeling of Agricultural Nonpoint Sources, Washington, DC.1990. USDA Agricult ural Research Service ARS 81, Washington, DC. Devitt, D., M. Berkowitz, P. Schulte, and R. Morris. 1993. Estimating transpiration for 3 woody ornamental tree species using stem flow gauges and lysimetry. HortScience 28:320 322. Elrashidi, M., A. Alva, Y. H uang, D. Calvert, T. Obreza, and Z. He. 2001. Accumulation and downward transport of phosphorus in Florida soils and relationship to water quality. Commun. Soil. Sci. Plant Anal. 32:3099 3119. Embleton, T.W., J.D. Kirkpatrick, W.W. Jones, and C.B. Cree. 19 56. Influence of applications of dolomite, potash and phosphate on yield and size of fruit and on composition of leaves of Valencia orange trees. Proc. Amer. Soc. Hort. Sci. 67:183 190. Essington, M.E. 2004. Soil and water chemistry: an integrative approac h CRC Press, Boca Raton, FL. Falivene, S. 2005. Open hydroponics: risks and opportunities. Land and Water Australia through the National Program of Sustainable Irrigation, Australia. http://www.arapahocitrus.com/files/OHS_Stage_Publications.pdf Falivene, S ., I. Goodwin, D. Williams, and A.M. Boland. 2005. Open hydroponics: risks and opportunities. Stage 1: General principles and literature review. http://www.arapahocitrus.com/files/OHS_Stage_Publications.pdf Fandika I.R., D. Kadyampakeni, and S. Zingore. 20 12. Performance of bucket drip irrigation by treadle pump on tomato and maize/bean production in Malawi. Irrig. Sci. 30:57 68. Fares, A., and A. Alva. 1999. Estimation of citrus evapotranspiration by soil water mass balance. Soil Sci. 164:302 310. Fares, A ., and A. Alva. 2000a. Evaluation of capacitance probes for optimal irrigation of citrus through soil moisture monitoring in an Entisol profile. Irrig. Sci. 19:57 64. Fares, A., and A. Alva. 2000b. Soil water components based on capacitance probes in a san dy soil. Soil Sci. Soc. Am. J. 64:311 318.

PAGE 324

324 Fares, A., J. Simunek, M.T. van Genuchten, L.R. Parsons, T.A. Wheaton, and K.T. Morgan. 2001. Effect of emitter patterns on solute transport. Soil Crop Sci. Soc. Proc. 61:46 56. Fares, A., A. Dogan, F. Abbas, L. Parsons, T. Obreza, and K. Morgan. 2008. Water balance components in a mature citrus orchard. Soil Sci. Soc. Am. J. 72:578 585. Fassel, V., and R. Kniseley. 1974. Inductively coupled plasma optical emission spectroscopy. Anal. Chem. 46:1110 1120. Feddes R.A., P.J. Kowalik, and H. Zaradny. 1978. Simulation of field water use and crop yield John Wiley and Sons, New York, USA. Feigenbaum, S., H. Bielorai, Y. Erner, and S. Dasberg. 1987. The fate of N 15 labeled nitrogen applied to mature citrus trees. Plan t Soil 97:179 187. Ferguson, J.J. 2002. Young Florida dooryard citrus guide young tree care. Gainesville, FL. Ferguson, J.J., F.S. Davies, C.H. Matthews, and R.M. Davis. 1988. Controlled release fertilizers and growth of young 'Hamlin' orange trees. Proc. Fla. State Hort. Soc. 101:17 20. Fernandez Galvez, J., and L. Simmonds. 2006. Monitoring and modelling the three dimensional flow of water under drip irrigation. Agric. Water Manage. 83:197 208. Ford, H.W. 1954. Root distribution in relation to the water table. Proc. Fla. State Hort. Soc. 67:30 33. Ford, H.W. 1964. The effect of root stock, soil and soil pH on citrus root growth in soils subject to flooding. Proc. Fla. State Hort. Soc. 77:41 45. Ford, H.W. 1972. Eight years of root injury from water table fluctuations. Proc. Fla. State Hort. Soc. 85:65 68. Gardenas, A., J. Hopmans, B. Hanson, and J. Simunek. 2005. Two dimensional modeling of nitrate leaching for various fertigation scenarios under micro irrigation. Agric. Water Manage. 74:219 242. Garnier, P. 2003. Sensitivity analysis of PASTIS, a model of nitrogen transport in the soil, p. 367 375, In D. Wallach, et al., (eds.) Working with Dynamic Models: application, analysis, parameterization, and applications. ed., Amsterdam, The Netherlands. Goldberg, D., and M. Shmueli. 1970. Drip irrigation a method used under arid and desert conditions of high water and soil salinity. Trans. ASAE 13:38 41.

PAGE 325

325 Goldberg, S., M. Rinot, and N. Karu. 1971. Effect of trickle irrigation intervals on distribution and utilizat ion of soil moisture in a vineyard. Soil Sci. Soc. Am. Proc. 35:127 &. Gottwald, T., J. Graham, and T. Schubert. 2002a. Citrus canker: The pathogen and its impact. American Phytopathological Society, St. Paul, Minnesota. Gottwald, T., X. Sun, T. Riley, J. Graham, F. Ferrandino, and E. Taylor. 2002b. Geo referenced spatiotemporal analysis of the urban citrus canker epidemic in Florida. Phytopathology 92:361 377. Graetz, D.A., and V.D. Nair. 2009. Phosphorus sorption isotherm determination, p. 33 37, In J. L. Kovar and G. M. Pierzynski, (eds.) Methods of P Analysis for Soils, Sediments, Residuals, and Waters 2nd ed. Virginia Tech University, Blacksburg, Virginia. Graser, E.A., and L.H. Allen. 1987. Water management for citrus production in the Florida Flatwood s. Proc. Fla. State Hort. Soc. 100:126 136. Grieve, A. 1989. Water use efficiency, nutrient uptake and productivity of micro irrigated citrus. Austr. J. Exp. Agric. 29:111 118. Grosse, W., F. Wissing, R. Perfler, Z. Wu, J. Chang, and Z. Lei. 1999. Biotech nological approach to water quality improvement in tropical and subtropical areas for reuse and rehabilitation of aquatic ecosystems. Cologne, Germany. Gutierrez, M., and F. Meinzer. 1994. Estimating water use and irrigation requirements of coffee in Hawai i. Journal of the American Society For Horticultural Science 119:652 657. Gutierrez, M., R. Harrington, F. Meinzer, and J. Fownes. 1994. The effect of environmentally induced stem temperature gradients on transpiration estimates from the heat balance meth od in 2 tropical woody species. Tree Physiol. 14:179 190. Halbert, S., and K. Manjunath. 2004. Asian citrus psyllids (Sternorrhyncha : Psyllidae) and greening disease of citrus: A literature review and assessment of risk in Florida. Fla. Entomol. 87:330 35 3. Ham, J., J. Heilman, and R. Lascano. 1990. Determination of soil water evaporation and transpiration from energy balance and stem flow measurements. Agric. For. Meteorol. 52:287 301. Hanlon, E.A., J.S. Gonzalez, and J.M. Bartos. 1997. Mehlich 1 extracta ble P, Ca, Mg, Mn, Cu and Zn. University of Florida, Gainesville, FL.

PAGE 326

326 Hanson, B., J. Simunek, and J. Hopmans. 2006. Evaluation of urea ammonium nitrate fertigation with drip irrigation using numerical modeling. Agric. Water Manage. 86:102 113. Harbridge, J 2007a. Determination of nitrate in 2M KCl soil extracts by flow injection analysis QuickChem Method 12 107 04 1 J. Lachat Instruments 5600 Lindburgh Drive, Loveland, Colorado, USA. Harbridge, J. 2007b. Determination of ammonia in 2M KCl soil extracts by flow injection analysis QuickChem Method 12 107 04 1 J. Lachat Instruments 5600 Lindburgh Drive, Loveland, Colorado, USA. Harris, W., R. Rhue, G. Kidder, R. Brown, and R. Littell. 1996. Phosphorus retention as related to morphology of sandy coastal plain s oil materials. Soil Sci. Soc. Am. J. 60:1513 1521. Havlin, J.L., J.D. Beaton, S.L. Tisdale, and W.L. Nelson. 2005. Soil fertility and fertilizers: An introduction to nutrient management. 7th ed., Prentice Hall, New Delhi. He, Z., D. Calvert, A. Alva, D. Ba nks, and Y. Li. 2000. Nutrient leaching potential of mature grapefruit trees in a sandy soil. Soil Sci. 165:748 758. Heilman, J., K. Mcinnes, M. Savage, R. Gesch, and R. Lascano. 1994. Soil and canopy energy balances in a west texas vineyard. Agric. For. M eteorol. 71:99 114. Hillel, D. 1998. Environmental Soil Physics Academic Press, San Diego, CA. Himmelbauer, M., W. Loiskandl, and F. Kastanek. 2004. Estimating length, average diameter and surface area of roots using two different Image analyses systems. P lant Soil 260:111 120. Hoffman, G.J., J.D. Oster, and W.J. Alves. 1982. Evapotranspiration of Mature Orange Trees Under Drip Irrigation in an Arid Climate. Trans. ASAE 25:992 996. Hornsby, A.G., P.S.C. Rao, J.G. Booth, P.V. Rao, K.D. Pennell, R.E. Jessup, and G.D. Means. 1990. Evaluation of models for predicting fate of pesticides. Soil Science Department and Statistics Department, University of Florida, Tallahasee, FL. Huntington, T.G., A.H. Johnson, and T.N. Schwartzman. 1990. Mechanical vacuum extractio n vs. batch equilibration for estimation of exchangeable cations. Soil Sci. Soc. Am. J. 54:381 385. Hutson, J., and R. Wagenet. 1991. Simulating nitrogen dynamics in soils using a deterministic model. Soil Use Manage. 7:74 78.

PAGE 327

327 Hutson, J.L. 2005. Leaching a nd Chemistry Model (LEACHM): A process based model of water and solute movement, transformations, plant uptake and chemical reactions in the unsaturated zone. Cornell University, Ithaca, New York. Irey, M., P. Mai, J. Graham, and J. Johnson. 2008. Data tre nds and results from an HLB testing laboratory that has processed over 64,000 commercial and research samples over a two year period in Florida. Abstract IRCHLB Proc. 103. Ismail, M., and K. Noor. 1996. Growth, water relations and physiological processes of starfruit (Averrhoa carambola L) plants under root growth restriction. Sci. Hortic. 66:51 58. Jabro, J., J. Jemison, L. Lengnick, R. Fox, and D. Fritton. 1993. Field validation and comparison of leachm and ncswap models for predicting nitrate leaching. Trans. ASAE 36:1651 1657. Jansson, P.E., and L. Karlberg. 2001. Coupled heat and mass transfer model for soil plant atmosphere systems. Royal Institute, Stockholm. Jia, X., A. Swancar, J.A. Jacobs, M.D. Dukes, and K. Morgan. 2007. Comparison of evapotransp iration rates for flatwoods and ridge citrus. Trans. ASABE 50:83 94. Jones, J., G. Hoogenboom, C. Porter, K. Boote, W. Batchelor, L. Hunt, P. Wilkens, U. Singh, A. Gijsman, and J. Ritchie. 2003. The DSSAT cropping system model. Eur. J. Agron. 18:235 265. Jones, J.B. 1997. Hydroponics: A practical guide for the soilless grower St. Lucie Press, FL. Jones, J.B.J., and V.W. Case. 1990. Sample, handling, and analyzing plant tissue samples, p. 389 427, In R. L. Westerman, (ed.) Soil Testing and Plant Analysis. e d. Soil Science Society of America, Madison, Wisconsin. Kadlec, R.H., and R.L. Knight. 1996. Treatment wetlands CRC Press, Boca Raton, FL, USA. Kalmar, D., and E. Lahav. 1977. Water requirements of avocado in Israel .1. Tree and soil parameters. Austr. J. Agric. Res. 28:859 868. Kandelous, M., and J. Simunek. 2010a. Comparison of numerical, analytical, and empirical models to estimate wetting patterns for surface and subsurface drip irrigation. Irrig. Sci. 28:435 444. Kandelous, M., and J. Simunek. 2010b. N umerical simulations of water movement in a subsurface drip irrigation system under field and laboratory conditions using HYDRUS 2D. Agric. Water Manage. 97:1070 1076.

PAGE 328

328 Kandelous, M., J. Simunek, M. van Genuchten, and K. Malek. 2011. Soil water content dist ributions between two emitters of a subsurface drip irrigation system. Soil Sci. Soc. Am. J. 75:488 497. Katul, G., P. Todd, D. Pataki, Z. Kabala, and R. Oren. 1997. Soil water depletion by oak trees and the influence of root water uptake on the moisture c ontent spatial statistics. Water Resour. Res. 33:611 623. Keller, J., and D. Karmeli. 1974. Trickle irrigation design for optimal soil wetting. Proc. International Drip Irrigation Congress, San Diego, California, USA. Khakural, B., and A. Alva. 1996. Trans formation of urea and ammonium nitrate in an Entisol and a Spodosol under citrus production. Commun. Soil. Sci. Plant Anal. 27:3045 3057. Khan, A., M. Yitayew, and A. Warrick. 1996. Field evaluation of water and solute distribution from a point source. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 122:221 227. Kiggundu, N., K.W. Migliaccio, B. Schaffer, Y. Li, and J.H. Crane. 2011. Water savings, nutrient leaching, and fruit yield in ayoung avocado orchard as affected by irrigation and nutrient management. Irr ig. Sci. DOI: 10.1007/s00271 011 0280 6. Kirchoff, G. 1992. Measurement of root length and thickness using a handheld counter scanner. Field Crops Res. 29:79 88. Klein, I., and G. Spieler. 1987. Fertigation of apples with nitrate or ammonium nitrogen under drip irrigation .2. Nutrient distribution in the soil. Commun. Soil. Sci. Plant Anal. 18:323 339. Klein, I., A. Meimon, and D. Skedi. 1999. Drip nitrogen, phosphorus, and potassium fertigation of 'Spadona' pear. J. Plant Nutr. 22:489 499. Klepper, B. 1992 Development and growth of crop root systems. Adv. Soil Sci. 19:1 25. Klute, A. 1986. Water retention: Laboratory methods, p. 635 662, In A. Klute, (ed) Methods of soil analysis. Part 1. Physical and Mineralogical Methods. Agronomy No. 9. 2nd ed. America n Society of Agronomy, Madison, WI. Klute, A. and C. Dirksen. 1986. Hydraulic conductivity and diffusivity: laboratory methods, p. 687 734, In A. Klute (ed.) Methods of soil analysis. Part 1. Physical and Mineralogical Methods. Agronomy No. 9. 2nd ed. Amer ican Society of Agronomy, Madison, WI. Kohne, J., and H. Gerke. 2005. Spatial and temporal dynamics of preferential bromide movement towards a tile drain. Vadose Zone J. 4:79 88.

PAGE 329

329 Koo, R.C.J. 1980. Results of citrus fertigation studies. Proc. Fla. State Hor t. Sci. 93:33 36. Koo, R.C.J., and A.G. Smajstrla. 1984. Effects of trickle irrigation and fertigation on fruit production and juice quality of 'Valencia' orange. Proc. Fla. State Hort. Soc. 97:8 10. Koo, R.C.J., C.A. Anderson, I. Stewart, D.P.H. Tucker, D.V. Calvert, and H.K. Wutscher. 1984. Recommended fertilizer and nutrional sprays for citrus. Gainesville, Florida. Kruger, J.A., C.D. Tolmay. and K. Britz. 2000a. Effects of fertigation frequencies and irrigation systems on performance of' Valencia' ora nges in two subtropical areas of South Africa. Proc. Int. Soc. Citric. Congr. p. 232 235. Kruger, J.A., K. Britz, C.D. Tolmay, and S F. du Plessis. 2000b. Evaluation of an open hydroponics system (OHS) for citrus in South Africa: Preliminary results. Proc. Int. Soc. Citric. Congr. p. 239 242. Kuperus, K.H., N. Combrink, K. Britz, and J. Ngalo. 2002. Evaluation of an open hydroponics system (OHS) for citrus in South Africa: Preliminary results, pp. 239 242 Annual Research Report. Citrus Research Internationa l (Pty) Ltd. Kusakabe, A., S. White, J. Walworth, G. Wright, and T. Thompson. 2006. Response of microsprinkler irrigated navel oranges to fertigated nitrogen rate and frequency. Soil Sci. Soc. Am. J. 70:1623 1628. Lamb, S., W. Graham, C. Harrison, and A. A lva. 1999. Impact of alternative citrus management practices on groundwater nitrate in the Central Florida Ridge I. Field investigation. Trans. ASAE 42:1653 1668. Lascano, R., R. Baumhardt, and W. Lipe. 1992. Measurement of water flow in young grapevines using the stem heat balance method. Am. J. Enol. Vitic. 43:159 165. Lea Cox, J.D., J.P. Syvertsen, and D.A. Graetz. 2001. Springtime 15Nitrogen uptake, partitioning, and leaching losses from young bearing Citrus trees of differing nitrogen status. J. Amer Soc. Hort. Sci. 126:242 251. Legaz, F., M. Serna, and E. Primomillo. 1995. Mobilization of the reserve in citrus. Plant Soil 173:205 210. Legaz, F., E. Primomillo, E. Primoyufera, C. Gil, and J. Rubio. 1982. Nitrogen fertilization in citrus .1. Absorptio n and distribution of nitrogen in calamondin trees (Citrus mitis BL), during flowering, fruit set and initial fruit development periods. Plant Soil 66:339 351.

PAGE 330

330 Li, J., and Y. Liu. 2011. Water and nitrate distributions as affected by layered textural soil a nd buried dripline depth under subsurface drip fertigation. Irrig. Sci. 29:469 478. Li, J., J. Zhang, and L. Ren. 2003. Water and nitrogen distribution as affected by fertigation of ammonium nitrate from a point source. Irrig. Sci. 22:19 30. Ling, G., and A. El Kadi. 1998. A lumped parameter model for nitrogen transformation in the unsaturated zone. Water Resour. Res. 34:203 212. Lotse, E.,J. Jabro, K. Simmons, and D. Baker. 1992. Simulation of nitrogen dynamics and leaching from arable soils. J. Contam. Hydrol. 10:183 196. Lu, J., and L. Wu. 2003. Visualizing bromide and iodide water tracer in soil profiles by spray methods. J.Environ. Qual. 32:363 367. Lumbanraja, J., and V. Evangelou. 1990. Binary and ternary exchange behavior of potassium and ammonium on Kentucky subsoils. Soil Sci. Soc. Am. J. 54:698 705. Lumbanraja, J., and V. Evangelou. 1994. Adsorption desorption of potassium and ammonium at low cation concentrations in 3 Kentucky subsoils. Soil Sci. 157:269 278. Manjunath, K.L., S.E. Halbert, C. Ra madugu, S. Webb, and R.F. Lee. 2008. Detection management of citrus huanglongbing in Florida. Phytopathology 98:387 396. Marler, T.E. and F.S. Davies. 1990. Microsprinkler Amer. Soc. Hort. Sci. 115(1):45 51. Martin, E.C., A.K. Hla, P.M. Waller, and D.C. Slack. 1997. Heat unit based crop coefficient for grape fruit trees. Appl. Eng. Agric. 13:485 489. Martinez Valero, R., and C. Fernandez 2004. Preliminary results in citrus groves grown under the MOHT system. Int. Citrus Cong.103. Masse J., F. Tardieu, and C. Colnenne. 1991. Rooting depth and spatial arrangement of roots in winter wheat. In McMichael B.L. and Persson H. (eds.) Plant Roo ts and their Environment, pp. 480 486. Amsterdam: Elsevier. Mattos, D., D. Graetz, and A. Alva. 2003a. Biomass distribution and nitrogen 15 partitioning in citrus trees on a sandy entisol. Soil Sci. Soc. Am. J. 67:555 563. Mattos, D., J.A. Quaggio, H. Cant arella, and A.K. Alva. 2003b. Nutrient content of biomass components of Hamlin sweet orange tree. Sci. Agric. 60:155 160.

PAGE 331

3 31 Mattos, D., A. Alva, S. Paramasivam, and D. Graetz. 2003c. Nitrogen volatilization and mineralization in a sandy entisol of Florida u nder citrus. Commun. Soil. Sci. Plant Anal. 34:1803 1824. Mattos Junior, D. 2000. Citrus response functions to N, P, and K fertilization and N uptake dynamics. Thesis (Ph D ), University of Florida, 2000. Mattos, D., T.F. Milanze, F.A. Azevedo, J.A. Quaggi o. 2010. Soil nutrient availability and its impact on fruit quality of Tahiti acid lime. Rev. Bras. Frutic. 32:335 342. Mehlich, A. 1953. Determination of P, K, Ca, Mg, and NH4. Soil Test Div. Mimeo, N.C. Dept. of Agriculture, Raleigh. www.ncagr.com/agr onomi/pdffiles/mehlich53.pdf Meinzer, F., G. Goldstein, N. Holbrook, P. Jackson, and J. Cavelier. 1993. Stomatal and environmental control of transpiration in a lowland tropical forest tree. Plant Cell Environ. 16:429 436. Michelakis, N., E. Vougioucalou, and G. Clapaki. 1993. Water use, wetted soil volume, root distribution and yield of avocado under drip irrigation. Agric. Water Manage. 24:119 131. Migliaccio. K.C. and Y.C. Li. 2009. Irrigation scheduling for tropical fruit groves in south Florida. Univ. of Florida Coop. Ext. Serv., Fact Sheet TR001. Gainesville, FL. Misra, C., D. Nielsen, and J. Biggar. 1974. Nitrogen transformations in soil during leaching .1. Theoretical considerations. Soil Sci. Soc. Am. J. 38:289 293. Monod, H., C. Naud, and D. Makow ski. 2003. Uncertainty and sensitivity analysis for crop models, p. 55 99, In D. Wallach, et al., (eds.) Working with Dynamic Crop Models: Application, Analysis, Parameterization, and Applications. ed., Amsterdam, The Netherlands. Morgan, K. H. Beck, J. Sc holberg, and S. Grunwald. 2006c. In Season Irrigation and Nutrient Decision Support System for Citrus Production. In F. Zazueta, J. Kin, S. Ninomiya and G. Schiefer (eds) Computers in Agriculture and Natural Resources, Proceedings of the 4th World Congress 24 26 July 2006, Orlando, Florida USA. Morgan, K. T. and E. A. Hanlon. 2006. Improving citrus nitrogen uptake efficiency: understanding citrus nitrogen requirements. SL 240. UF/IFAS Cooperative and Extension Service, Gainesville, FL. Morgan, K., T. Obrez a, T. Wheaton, and L. Parsons. 2002. Comparison of soil matric potential measurements using tensiometric and resistance methods. Soil Crop Sci. Soc. Proc. 61:63 66.

PAGE 332

332 Morgan, K., J. Scholberg, T. Obreza, and T. Wheaton. 2006a. Size, biomass, and nitrogen rel ationships with sweet orange tree growth. J. Amer. Soc. Hort. Sci. 131:149 156. Morgan, K., L. Parsons, T. Wheaton, D. Pitts, and T. Obreza. 1999. Field calibration of a capacitance water content probe in fine sand soils. Soil Sci. Soc. Am. J. 63:987 989. Morgan, K., T. Obreza, J. Scholberg, L. Parsons, and T. Wheaton. 2006b. Citrus water uptake dynamics on a sandy Florida Entisol. Soil Sci. Soc. Am. J. 70:90 97. Morgan, K.T. 2004. Nitrogen and biomass distribution, and nitrogen and water uptake parameters for citrus. Thesis (Ph D ), University of Florida, 2004. Morgan, K.T., T.A. Obreza, and J.M.S. Scholberg. 2007. Orange tree fibrous root length distribution in space and time. J. Amer. Soc. Hort. Sci. 132:262 269. Morgan, K.T., T.A. Wheaton, W.S. Castle, and L.R. Parsons. 2009a. Response of young and maturing citrus trees grown on a sandy soil to irrigation scheduling, nitrogen fertilizer rate, and nitrogen application method. HortScience 44:145 150. Morgan, K.T., A.W. Schumann, W.S. Castle, E.W. Stover, D. Kadyampakeni, P. Spyke, F.M. Roka, R. Muraro, and A. Morris. 2009b. Citrus production systems to survive greening Horticultural practices. Proc. Fla. State Hort. Soc. 122:114 121. Morgan, K.T., E.A. Hanlon, and T.A. Obreza. 2009c. A Web Based Irrigatio n Scheduling Model to Improve Water Use Efficiency and Reduce Nutrient Leaching for Florida Citrus. UF/IFAS Extension, Publication # SL286, Gainesville, FL. Mualem, Y. 1986. Hydraulic conductivity of unsaturated soils: prediction and formulas, p. 799 824, In A. Klute, (ed.) Methods of Soil Analysis: Part 1 Physical and Mineralogical Methods. Second Edition ed. Agronomy. American Society of Agronomy, Soil Science Society of America, Madison, Wisconsin. Munoz Carpena, R. 2009. Field devices for monitoring soi l water content. Univ. of Florida Coop. Ext. Serv. Fact Sheet BUL343. Gainesville, FL. Munoz Carpena, R., M. Dukes, Y. Li, and W. Klassen. 2005. Field comparison of tensiometer and granular matrix sensor automatic drip irrigation on tomato. HortTechnology 15:584 590. Munson, R., and W. Nelson. 1963. Movement of applied potassium in soils. J. Agric. Food Chem. 11:193 201. Munter, R., T. Halverson, and R. Anderson. 1984. Quality assurance for plant tissue analysis by ICP AES. Commun. Soil. Sci. Plant Anal. 15 :1285 1322.

PAGE 333

333 Munter, R.C., and R.A. Grande. 1981. Plant tissue and soil extract analysis by ICP AES, p. 653 673, In R. M. Barnes, (ed.) Developments in atomic plasma spectrochemical analysis. ed. Heydon and Son, Philadelphia, PA. Muraro, R.P. 2008. Summary of 2007 08 citrus budget for the Southwest Florida production region. http://www.crec.ifas.ufl.edu/extension/economics/pdf/SW_Budget_Summ_2007_ 2008.pdf Nair P., T. Logan, A. Sharpley, L. Sommers, M. Tabatabai, and T. Yuan. 1984. Interlaboratory comparison of a standardized phosphorus adsorption procedure. J. Environ. Qual. 13:591 595. Nair, V., D. Graetz, and K. Reddy. 1998. Dairy manure influences on phosph orus retention capacity of Spodosols. J. Environ. Qual. 27:522 527. Nair, V.D., and W.G. Harris. 2004. A capacity factor as an alternative to soil test phosphorus in phosphorus risk assessment. N. Z. J. Agric. Res. 47:491 497. Nair, V.D.; K.M. Portier, D. A. Graetz, and M.L. Walker. 2004. An environmental threshold for degree of phosphorus saturation in sandy soils. J. Environ. Qual. 33:107 113. Nappi, P., R. Jodice, A. Luzzati, and L. Corino. 1985. Grapevine root system and VA mycorrhizae in some soils of Piedmont (Italy). Plant Soil 85:205 210. Nelson, N., J. Parsons, and R. Mikkelsen. 2005. Field scale evaluation of phosphorus leaching in acid sandy soils receiving swine waste. J. Environ. Qual. 34:2024 2035. Newman, E. 1966. A method of estimating total length of root in a sample. J. Appl. Ecol. 3:139 145. Obreza, T. 1993. Program fertilization for establishment of orange trees. J. Prod. Agric. 6:546 552. Obreza, T., and R. Rouse. 1993. Fertilizer effects on early growth and yield of Hamlin orange trees. HortScience 28:111 114. Obreza, T., and D. Pitts. 2002. Effective rainfall in poorly drained microirrigated citrus orchards. Soil Sci. Soc. Am. J. 66:212 221. Obreza, T.A., and D.P.H. Tucker. 2006. Nutrient management, p. 167 182, In Tucker, D.P.H., J.S. R ogers, E.W. Stover, and M.R. Ziegler (eds), Florida citrus: A comprehensive guide. SP278, University of Florida, IFAS Extension, Gainesville, FL.

PAGE 334

334 Obreza, T., R. Rouse, and K. Morgan. 2008a. Managing Phosphorus for Citrus Yield and Fruit Quality in Developi ng Orchards. HortScience 43:2162 2166. Obreza, T., M. Clark, B. Boman, T. Borisova, M. Cohen, M. Dukes, T. Frazer, E. Hanlon, K. Havens, C. Martinez, K. Migliaccio, S. Shukla, and A. Wright. 2010. A guide to EPA's proposed numeric nutrient water quality cr iteria for Florida. UF/IFAS Extension, Publication #316, Gainesville, FL. Obreza, T.A., and K.E. Admire. 1985. Shallow water table fluctuations in response to rainfall, irrigations and evapotranspiration in flatwoods citrus. Proc. Fla. State Hort. Soc. 98: 32 37. Obreza, T.A., and R.E. Rouse. 1991. Controlled release fertilizer use on young 'hamlin' orange trees. Soil Crop Sci. Soc. Proc. 51:64 68. Obreza, T.A., and R.E. Rouse. 2006. Long term response of 'Hamlin' orange trees to controlled release nitrogen fertilizers. HortScience 41:423 426. Obreza, T.A., and M.E. Collins. 2008. Common soils used for citrus production in Florida. University of Florida Cooperative Extension Service, Gainesville, Florida. Obreza, T.A., and K.T. Morgan. 2008. Nutrition of Flo rida citrus trees, Cooperative Extension Service, University of Florida, Institute of Food and Agricultural Sciences, Gainesville, FL. Obreza, T.A., M. Zekri, and E.A. Hanlon. 2008b. Soil and leaf testing, p. 24 32, In T. A. Obreza and K. T. Morgan, (eds.) Nutrition of Florida Trees. Second Edition ed. University of Florida/IFAS, Gainesville. Obreza, T.A., A.K. Alva, E.A. Hanlon, and R.E. Rouse. 1999. Citrus grove leaf tissue and soil testing: Sampling, analysis and interpretation., Gainesville, FL. Obreza, T.A., D.J. Pitts, L.R. Parsons, T.A. Wheaton, and K.T. Morgan. 1997. Soil water holding characteristics affect citrus irrigation strategy. Proc. Fla. State Hort. Soc. 110:36 39. Pang, L., M. Close, J. Watt, and K. Vincent. 2000. Simulation of picloram, at razine, and simazine leaching through two New Zealand soils and into groundwater using HYDRUS 2D. J. Contam. Hydrol. 44:19 46. Paramasivam, S., A. Alva, and A. Fares. 2000a. An evaluation of soil water status using tensiometers in a sandy soil profile unde r citrus production. Soil Sci. 165:343 353. Paramasivam, S., A. Alva, and A. Fares. 2000b. Transformation and transport of nitrogen forms in a sandy entisol following a heavy loading of ammonium nitrate solution: Field measurements and model simulations. J Soil Contam. 9:65 86.

PAGE 335

335 Paramasivam, S., A. Alva, A. Fares, and K. Sajwan. 2001. Estimation of nitrate leaching in an entisol under optimum citrus production. Soil Sci. Soc. Am. J. 65:914 921. Paramasivam, S., A. Alva, A. Fares, and K. Sajwan. 2002. Fate o f nitrate and bromide in an unsaturated zone of a sandy soil under citrus production. J. Environ. Qual. 31:671 681. Paramasivam, S., A. Alva, K. Hostler, G. Easterwood, and J. Southwell. 2000c. Fruit nutrient accumulation of four orange varieties during fr uit development. J. Plant Nutr. 23:313 327. Parsons, L.R., and K.T. Morgan. 2004. Management of microsprinkler systems for Florida citrus. University of Florida, Gainesville, Florida. Parsons, L.R., and T.A. Wheaton. 2009. Tree density, hedging and topping Gainesville, Florida. Parsons, L.R., T.A. Wheaton, and W.S. Castle. 2001. High Application Rates of Reclaimed Water Benefit Citrus Tree Growth and Fruit Production. HortScience 36(7):1273 1277. Petillo, M., and J. Castel. 2007. Water balance and crop coe fficient estimation of a citrus orchard in Uruguay. Spanish J. Agric. Res. 5:232 243. Phogat, V., M. Mahadevan, M. Skewes, and J.W. Cox. 2011. Modelling soil water and salt dynamics under pulsed and continuous surface drip irrigation of almond and implicat ions of system design. Irrig. Sci. DOI: 10.1007/s00271 011 0284 2 Pijl, I. 2001. Drip fertigation: Effects on water movement, soil characteristics and root distribution MS, University of Stellenbosch, South Africa. Plank, C.O. 1992. Plant analysis referen ce procedures for the southern region of the United States. Southern Coop. Ser. Bull. 368. The University of Georgia, Crop & Soil Science Dept.Athens, GA, URL: http://www.cropsoil.uga.edu/~oplank/sera368.pdf Polex, M., G. Vidalakis, and K. Godfrey. 2007. Citrus bacteria canker disease and Huanglongbing (citrus greening). University of California. Prinsloo, J.A. 2007. Ecophysiological responses of citrus tree and sugar accumulation of fruit in response to altered plant water relations. M. S. Thesis, Unive rsity of Stellenbosch, South Africa. p. 92. Quinones, A., B. Martinez Alcantara, and F. Legaz. 2007. Influence of irrigation system and fertilization management on seasonal distribution of N in the soil profile and on N uptake by citrus trees. Agric. Ecosy st. Environ. 122:399 409.

PAGE 336

336 Quinones, A., J. Banuls, E.P. Millo, and F. Legaz. 2003a. Effects of N 15 application frequency on nitrogen uptake efficiency in Citrus trees. J. Plant Physiol.:1429 1434. Quinones, A., J. Banuls, E. Primo Millo, and F. Legaz. 2003b. Seasonal dynamics of 15N applied as nitrate with different irrigation systems and fertilizer management in citrus plants. J. Food Agric. Environ. 1:155 161. Quinones, A., J. Banuls, E. Primo Mil lo, and F. Legaz. 2005. Recovery of the N 15 labelled fertiliser in citrus trees in relation with timing of application and irrigation system. Plant Soil 268:367 376. Rana, G., N. Katerji, and F. de Lorenzi. 2005. Measurement and modelling of evapotranspir ation of irrigated citrus orchard under Mediterranean conditions. Agric. For. Meteorol. 128:199 209. Recous, S., J. Machet, and B. Mary. 1992. The partitioning of fertilizer N between soil and crop comparison of ammonium and nitrate applications. Plant S oil 144:101 111. Reicosky, D., R.J. Millington, and D. Peters. 1970. A comparison of 3 methods for estimating root length. Agron. J. 62:451 453. Reitz, H.J., and W.T. Long. 1955. Water table fluctuations and depth of rooting of citrus trees in the Indian River area. Proc. Fla. State Hort. Soc. 68:24 29. Richards, L.A. 1931. Capillary conduction of fluid through porous medium. Physics 1:318 333. Richards, N. 1992. Cashew tree nutrition related to biomass accumulation, nutrient composition and nutrient cy cling in sandy red earths of northern territory, Australia. Sci. Hortic. 52:125 142. Robinson, T.L., S.A. Hoying, A. DeMarree, K. Iungerman, and M. Fargione. 2007. The evolution towards more competitive apple orchard systems in New York. Fruit Q. 15:3 9. Rodriguez, S., A. Alonso Gaite, and J. Alvarez Benedi. 2005. Characterization of nitrogen transformations, sorption and volatilization processes in urea fertilized soils. Vadose Zone J. 4:329 336. Rogers, J., and J. Bartholic. 1976. Estimated evapotranspir ation and irrigation requirements for citrus. Soil Crop Sci. Soc. Proc. 35:111 117. Rogers, J., L. Allen, and D. Calvert. 1983. Evapotranspiration from a humid region developing citrus grove with grass cover. Trans. ASAE 26:1778 1783.

PAGE 337

337 Roistacher, C.N. 1996 The economics of living with citrus diseases: Huanglongbing (greening) in Thailand, p. 279 285, In J. V. da Graca, et al., (eds.) 13th Conference of the International Organization of Citrus Virologists (IOCV). ed., University of California, Riverside. Ro ka, F., R. Muraro, R.A. Morris, P. Spyke, K. Morgan, A. Schumann, W. Castle, and E. Stover. 2009. Citrus production systems to survive greening: Economic thresholds. Proc. Fla. State Hort. Soc. 122:122 126. Rose, D. 1983. The description of the growth of root systems. Plant Soil 75:405 415. Sacks, J., W. Welch, T. Mitchell, and H.P. Wynn. 1989. Design and analysis of computer experiments. Stat. Sci. 4:409 435. Saka, A.R. 1984. Nitrogen movement, retention and uptake in the corn (Zea mays L.) root zone as influenced by cultivation and water management. Thesis (Ph D ), University of Florida, Gainesville, FL. Saltelli, A., S. Tarantola, and F. Campolongo. 2000. Sensitivity analysis as an ingredient of modeling. Stat. Sci. 15:377 395. Saltelli, A., S. Tarantol a, and F. Campolomgo. 2004. Sensitivity analysis in practice Wiley and Sons, New York. Sanchez, J.F. 2004. Water and nitrate movement in poultry litter amended soils. University of Florida, [Gainesville, Fla.]. Sartain, J. 1978. Adaptability of the double acid extractant to florida soils. Soil Crop Sci. Soc. Proc. 37:204 208. SAS Institute. 2011. The SAS system for Microsoft Windows Release 9.3.188. Sato, S. and K.T. Morgan. 2008. Nitrogen recovery and transformation from a surface or sub surface applicatio n of controlled release fertilizer on a sandy soil. J. Plant Nutr. 31:2214 2231. Sato, S., K. Morgan, M. Ozores Hampton, and E. Simonne. 2009a. Spatial and Temporal Distributions in Sandy Soils with Seepage Irrigation: I. Ammonium and Nitrate. Soil Sci. So c. Am. J. 73:1440 1440. Sato, S., K. Morgan, M. Ozores Hampton, and E. Simonne. 2009b. Spatial and Temporal Distributions in Sandy Soils with Seepage Irrigation: II. Phosphorus and Potassium. Soil Sci. Soc. Am. J. 73:1440 1440. Sato, S., K. Morgan, and K. Mahmoud. 2009c. Available phosphorus by five different soil testing methods and fractionation in southwest Florida vegetable production. ASA CSSA SSSA Oral Presentation, November 1 5, Pittsburgh, PA.

PAGE 338

338 Schaap, M., F. Leij, and M. van Genuchten. 2001. ROSETTA : a computer program for estimating soil hydraulic parameters with hierarchical pedotransfer functions. J. Hydrol. 251:163 176. Schoeman, S.P. 2002. Physiological measurements of daily daylight fertigated citrus trees. MS, University of Stellenbosch, South Africa. Scholberg, J., L. Parsons, T. Wheaton, B. McNeal, and K. Morgan. 2002. Soil temperature, nitrogen concentration, and residence time affect nitrogen uptake efficiency in citrus. J. Environ. Qual. 31:759 768. Scholberg, J., L. Zotarelli, R. Tubbs, M Dukes, and R. Munoz Carpena. 2009. Nitrogen Uptake Efficiency and Growth of Bell Pepper in Relation to Time of Exposure to Fertilizer Solution. Commun. Soil. Sci. Plant Anal. 40:2111 2131. Schumann, A.W., K. Hostler, L. Waldo and K. Mann. 2010. Update on Advanced Citrus Production System Research in Florida. Citrus Ind. 91:6 11. Schumann, A.W., J.P. Syvertsen, and K.T. Morgan. 2009. Implementing advanced citrus pproduction systems in Florida Early results. Proc. Fla. State Hort. Soc. 122:108 113. Schum ann, A.W., A. Fares, A.K. Alva, and S. Paramasivam. 2003. Response of 'Hamlin' orange to fertilizer source, annual rate, and irrigation area. Proc. Fla. State Hort. Soc. 116:256 260. Selim, H., and I. Iskandar. 1981. Modeling nitrogen transport and transf ormations in soils .1. Theoretical considerations. Soil Sci. 131:233 241. Sepaskhah, A., and S. Kashefipour. 1995. Evapotranspiration and crop coefficient of sweet lime under drip irrigation. Agric. Water Manage. 27:331 340. Silberbush, M. and S.A. Barber. 1983. Prediction of phosphorus and potassium uptake by soybeans with a mechanistic mathematical model. Soil Sci. Soc. Am. J. 47:262 265. Sims, J.T., R.R. Simard, and B.C. Joern. 1998. Phosphorus loss in agricultural drainage: Historical perspective and cu rrent research. J. Environ. Qual. 27:277 293. Simunek, J., and J. Hopmans. 2009. Modeling compensated root water and nutrient uptake. Ecol. Modell. 220:505 521. Simunek, J., M. Sejna, and M.T. van Genuchten. 1999. The HYDRUS 2D software package for simulat ing two dimensional movement of water, heat, and multiple solutes in variably saturated media, International Groundwater Modeling Center, Colorado School of Mines, Golden.

PAGE 339

339 Simunek, J., M. Sejna, and M.T. van Genuchten. 2007. The HYDRUS software package for simulating two and three dimensional movement of water, heat, and multiple solutes in variably saturated media PC Progress, Prague, Czech Republic. Singh, B., and J. Jones. 1975. Use of sorption isotherms for evaluating potassium requirements of plants. Soil Sci. Soc. Am. J. 39:881 886. Skaggs, R.W. 1980. DRAINMOD reference report: Methods for design and evaluation of drainage water management systems for soils with high water tables. South National Technical Center, Ft. Worth, Texas. Skaggs, T., T. Trout J. Simunek, and P. Shouse. 2004. Comparison of HYDRUS 2D simulations of drip irrigation with experimental observations. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 130:304 310. Smajstrla, A.G., B.J. Boman, D.Z. Haman, F.T. Izuno, D.J. Pitts, and F.S. Zazuet a. 2009. Basic irrigation scheduling in Florida. University of Florida Cooperative Extension Service, Gainesville, Florida. Snyder, R.L., and N.V. O'Connell. 2007. Crop Coefficients for Microsprinkler irrigated Clean Cultivated, Mature Citrus in an Arid Cl imate. J. Irrig. Drain. Div. Am. Soc. Civ. Eng. 133:43 52. Soulsby, C., and B. Reynolds. 1992. Modeling hydrological processes and aluminum leaching in an acid soil at Llyn Brianne, mid Wales. J. Hydrol. 138:409 429. Sparks, D., L. Zelazny, and D. Martens. 1980. Kinetics of potassium exchange in a Paleudult from the coastal plain of Virginia. Soil Sci. Soc. Am. J. 44:37 40. Steinberg, S., C. Vanbavel, and M. Mcfarland. 1989. A gauge to measure mass flow rate of sap in stems and trunks of woody plants. J. Am er. Soc. Hort. Sci. 114:466 472. Steinberg, S., M. Mcfarland, and J. Worthington. 1990a. Comparison of trunk and branch sap flow with canopy transpiration in pecan. J. Exp. Bot. 41:653 659. Steinberg, S., C. Vanbavel, and M. Mcfarland. 1990b. Improved sap flow gauge for woody and herbaceous plants. Agron. J. 82:851 854. Steppe, K., S. Dzikiti, R. Lemeur, and J. Milford. 2006. Stomatal oscillations in orange trees under natural climatic conditions. Ann. Bot. 97:831 835. Stevens, R., and T. Douglas. 1994. Dis tribution of grapevine roots and salt under drip and full ground cover microjet irrigation systems. Irrig. Sci. 15:147 152. Stover, E., W.S. Castle, and P. Spyke. 2008. The citrus grove of the future and its implications for Huanglongbing management. Proc Fla. State Hort. Soc. 121:155 159.

PAGE 340

340 Syversten. J.P. and J.L. Jifon. 2001. Frequent fertigation does not affect citrus tree growth, fruit yield, nitrogen uptake, and leaching losses. Proc. Fla. State Hort. Soc. 114:88 93. Tennant, D. 1975. Test of a modif ied line intersect method of estimating root length. J. Ecol. 63:995 1001. Testi, L., F. Villalobos, and F. Orgaz. 2004. Evapotranspiration of a young irrigated olive orchard in southern Spain. Agric. For. Meteorol. 121:1 18. Testi, L., F. Villalobos, F. O rgaz, and E. Fereres. 2006. Water requirements of olive orchards: I simulation of daily evapotranspiration for scenario analysis. Irrig. Sci. 24:69 76. Thompson, T.L., and S.A. White. 2004. Nitrogen and phosphorus fertilizer requirements for young, bearing microsprinkler irrigated citrus, University of Arizona, Tucson, AZ. Timmer, L., and S. Zitko. 1996. Evaluation of a model for prediction of postbloom fruit drop of citrus. Plant Disease 80:380 383. Tinker, P.B., and P.H. Nye. 2000. Solute movement in the rhizosphere. New York, USA. Tucker, D.P., A.K. Alva, L.K. Jackson, and T.A. Wheaton. 1995. Nutrition of Florida citrus trees. University of Florida/IFAS Cooperative Extension Service, Gainesville, FL. U.S. Department of Health, E.a.W. 1962. Public health s ervice drinking water standards. Washington, DC, USA. USDA. 1990a. Soil Survey of Polk County. Florida, USA. USDA. 1990b. Soil Survey of Collier County Area. Florida, USA. USDA. 2011. Florida citrus statistics 2009 2010. Florida Department of Agriculture a nd Consumer Services, Tallahasee, Florida. USEPA. 2005. Groundwater and drinking water, United States Environmental Protection Agency, USA. VanGenuchten, M. 1980. A closed form equation for predicting the hydraulic conductivity of unsaturated soils. Soil S ci. Soc. Am. J. 44:892 898. Villalobos, F., L. Testi, and M. Moreno Perez. 2009. Evaporation and canopy conductance of citrus orchards. Agric. Water Manage. 96:565 573.

PAGE 341

341 Villapando, R., and D. Graetz. 2001. Phosphorus sorption and desorption properties of t he spodic horizon from selected Florida Spodosols. Soil Sci. Soc. Am. J. 65:331 339. Vincent, C., and P. Gregory. 1989a. Effects of temperature on the development and growth of winter wheat roots .1. Controlled glasshouse studies of temperature, nitrogen a nd irradiance. Plant Soil 119:87 97. Vincent, C., and P. Gregory. 1989b. Effects of temperature on the development and growth of winter wheat roots .2. Field studies of temperature, nitrogen and irradiance. Plant Soil 119:99 110. Wagenet, R., J. Biggar, an d D. Nielsen. 1977. Tracing transformations of urea fertilizer during leaching. Soil Sci. Soc. Am. J. 41:896 902. Wang, F., and A. Alva. 1996. Leaching of nitrogen from slow release urea sources in sandy soils. Soil Sci. Soc. Am. J. 60:1454 1458. Wang, F., and A. Alva. 2000. Ammonium adsorption and desorption in sandy soils. Soil Sci. Soc. Am. J. 64:1669 1674. Warrick, A.W. 1986. Soil water distribution, In Nakayama and Bucks, (eds.) Trickle Irrigation for Crop Production. ed. Elsevier, Amsterdam. Wheaton, T.A., L.R. Parsons, and K.T. Morgan. 2006. Simulating annual irrigation requirement for citrus on excessively drained soils. HortScience 41:1487 1492. Whitney, J.D., and T.A. Wheaton. 1984. Tree spacing affects citrus fruit distribution and yield. Proc. F la. State Hort. Soc. 97:44 47. Williams, J.R., and D.E. Kissel. 1991. Water percolation: An indicator of nitrogen leaching potential, In R. F. Follet, et al., (eds.) Managing Nitrogen for Groundwater Quality and Farm Profitability. ed. Soil Science Society of America, Madison, Wisconsin. Xin, J., F.S. Zazueta, A.G. Smajstrla, T.A. Wheaton, J.W. Jones, P.H. Jones, and D.D. Dankel. 1997. CIMS An integrated real time computer system for citrus micro irrigation management. Appl. Eng. Agric. 13:785 790. Xu, G., I. Levkovitch, S. Soriano, R. Wallach, and A. Silber. 2004. Integrated effect of irrigation frequency and phosphorus level on lettuce: P uptake, root growth and yield. Plant Soil 263:297 309. Yandilla. 2004. Martinez Open Hydroponic Technology. http://www.yandillapark.com.au/Growers/ohs_main.htm Yang, S., M. Aydin, T. Yano, and X. Li. 2003. Evapotranspiration of orange trees in greenhouse lysimeters. Irrig. Sci. 21:145 149.

PAGE 342

342 Yang, S.L., M. Aydin, Y. Kitamura, and T. Yano. 2010. The impact of irrigation water quality on water uptake by orange trees. African J. Agric. Res. 5:2661 2667. Ylaranta, T., J. UusiKamppa, and A. Jaakkola. 1996. Leaching of phosphorus, calcium, magnesium and potassiu m in barley, grass and fallow lysimeters. Acta Agric. Scand. 46:9 17. Zekri, M. and L. R. Parsons. 1988. Water relations of grapefruit trees in response to drip, microsprinkler, and overhead sprinkler irrigation. J. Amer. Soc. Hort. Sci. 113:819 823. Zekri M., and T.A. Obreza. 2003. Plant nutrients for citrus trees. UF/IFAS Extension, Gainesville, Florida. Zhang, M., A. Alva, and Y. Li. 1998. Fertilizer rates change root distribution of grapefruit trees on a poorly drained soil. J. Plant Nutr. 21:1 11. Zha ng, M., A. Alva, Y. Li, and D. Calvert. 1996. Root distribution of grapefruit trees under dry granular broadcast vs fertigation method. Plant Soil 183:79 84. Zhao, X. 1981. Citrus yellow shoot disease (Huanglongbing) A review. International Society of C itriculture 1:466 469. Zhou, M., and Y. Li. 2001. Phosphorus sorption characteristics of calcareous soils and limestone from the southern Everglades and adjacent farmlands. Soil Sci. Soc. Am. J. 65:1404 1412. Zhou, Q., S. Kang, L. Zhang, and F. Li. 2007. C omparison of APRI and Hydrus 2D models to simulate soil water dynamics in a vineyard under alternate partial root zone drip irrigation. Plant Soil 291:211 223. Zotarelli, L., J. Scholberg, M. Dukes, and R. Munoz Carpena. 2007. Monitoring of nitrate leachin g in sandy soils: Comparison of three methods. J. Environ. Qual. 36:953 962. Zotarelli, L., J. Scholberg, M. Dukes, and R. Munoz Carpena. 2008a. Fertilizer residence time affects nitrogen uptake efficiency and growth of sweet corn. J. Environ. Qual. 37:127 1 1278. Zotarelli, L., M. Dukes, J. Scholberg, R. Munoz Carpena, and J. Icerman. 2009a. Tomato nitrogen accumulation and fertilizer use efficiency on a sandy soil, as affected by nitrogen rate and irrigation scheduling. Agric. Water Manage. 96:1247 1258. Z otarelli, L., J. Scholberg, M. Dukes, R. Munoz Carpena, and J. Icerman. 2009b. Tomato yield, biomass accumulation, root distribution and irrigation water use efficiency on a sandy soil, as affected by nitrogen rate and irrigation scheduling. Agric. Water M anage. 96:23 34.

PAGE 343

343 Zotarelli, L., J. Scholberg, M. Dukes, H. Snyder, R. Munoz Carpena, and E. Simonne. 2006. Interaction between water and nitrogen application on yields and water use efficiency of tomato and pepper in sandy soil. HortScience 41:981 981. Zot arelli, L., M. Dukes, J. Scholberg, T. Hanselman, K. Le Femminella, and R. Munoz Carpena. 2008b. Nitrogen and water use efficiency of zucchini squash for a plastic mulch bed system on a sandy soil. Sci. Hortic. 116:8 16. Zvomuya, F., C. Rosen, M. Russelle, and S. Gupta. 2003. Nitrate leaching and nitrogen recovery following application of polyolefin coated urea to potato. J. Environ. Qual. 32:480 489.

PAGE 344

344 BIOGRAPHICAL SKETCH Davie Kadyampakeni was born in 1979 in Dedza district in central Malawi. He is a 5 th born in a family of 5 sons and 3 daughters He pursued his primary education (elementary education) at Dzenza Primary School and Kasina Preparatory Seminary in the same district between 1984 and 1993. Davie completed his junior secondary education (middl e school) at St. Kizito Seminary in Dedza district (1993 1995) and completed his senior secondary education at Ntcheu Secondary School (High School) from 1996 to 1997. Upon completion of high school education as the best student of that year Davie was se lected to the University of Malawi in April 1998 to pursue a Bachelor of Science in Agriculture at Bunda College of Agriculture, specializing in Agricultural Engineering. He completed his first degree in June 2002 receiving several an for Capacity Building in Agriculture (RUFORUM) Fellowship to pursue a Master of Science (MS) in a gronomy at the same college starting in October 2002. Davie completed his MS degree in September 200 4 and joined the Malawi Ministry of Agriculture and Food Security as an Irrigation Agronomist in the Department of Agricultural Research Services at Kasinthula Experiment Research Station, Chikwawa, Malawi in October of the same year. In January 2007 he j oined the International Crops Research Institute for the Semi Arid Tropics (ICRISAT) at Chitedze Agricultural Research Station Lilongwe, Malawi working as a Regional Scientific Officer for Malawi and Tanzania up to December 2007. From January 2008, he wo rk ed for the World Bank IFAD funded Irrigation, Rural Livelihoods and Agricultural Development Project in Zomba, Malawi until August 2008 when he enrolled in the Ph.D. in Soil and Water Science program at the University of Florida. Davie i s married to Ine ss Mhango and

PAGE 345

345 t hey are blessed with a son Atikonda who was born in the final stages of his doctoral program on October 16, 2011 in Naples, Florida. Upon completion and graduation, Davie plans to publish his work in refereed journals before joining the Con sultative Group of International Agricultural Research Centers as a Scientist to solve food security problems and poverty in developing countries through soil fertility amelioration and improved irrigation management.