<%BANNER%>

Nitrogen and Phosphorus Movement in Sandy Soils of South Florida Used for Sugarcane Production With Elevated Water Table

Permanent Link: http://ufdc.ufl.edu/UFE0044413/00001

Material Information

Title: Nitrogen and Phosphorus Movement in Sandy Soils of South Florida Used for Sugarcane Production With Elevated Water Table
Physical Description: 1 online resource (251 p.)
Language: english
Creator: Muwamba, Augustine
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: dripirrigation -- fertilizermixture -- tracer -- watertable
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: Farmers in southwest Florida need information related to distribution of soil characteristics in sugarcane fields with different soil orders since they affect accumulation of plant available nutrients within the root zone. Farmers need to know how fluctuating water table affects movement of plant available nutrients above and below the water table. Information on the distribution of nutrients leaching out of the root zone during water table management is needed by environmental scientists for the assessment of the adverse effects on the water quality. Water table depth is managed through pumping after rainfall events. Understanding of nutrient movement associated with decreases in water table depth is needed to minimize impacts on water quality by pumping activities. Reliable databases of constants like saturated conductivity values (Ksat) and phosphorus sorption coefficients (KD) are needed by scientists and farmers to accurately model water, nitrogen and phosphorus movement. Studies were conducted to document spatial distribution of soil characteristics in two mineral soils most widely used for sugar cane production, determine phosphorus sorption and kinetics associated with water table movement, and calibration of Hydrus 1D. The hypotheses of the studies were; (i) soil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary spatialy and with depth and this will lead to different patterns of phosphorus accumulations; (ii) phosphorus sorption coefficients (KD) when determined using different supporting electrolytes (0.01M KCl, 0.005M CaCl2, simulated Florida rain, deionized water, and fertilizer mixture) will significantly differ; (iii) reducing distance between water table and Bh horizon through lowering water table from 30 cm to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee soil; (iv) management of water table depth after rainfall events will lead to loss of plant available phosphorus and nitrogen out of the sugarcane plants’ root zone; (v) drip can be used to maintain high plant available nutrients within the root zone and minimize nutrients loss out of the root zone; and (vi) phosphorus leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolytes are used to model phosphorus movement. The objectives of the study were; (i) identifying the distribution of soil characteristics that affect accumulation of phosphorus and nitrogen in sugarcane fields with Immokalee fine sand and Margate fine sand; (ii) characterizing sorption of phosphorus using different electrolytes; (iii) studying movement of water determined by bromide tracer, phosphorus, and nitrogen in relation to fluctuating water table depth and drip irrigation, and (iv) modeling water (bromide), nitrogen, and phosphorus using linearized sorption coefficients determined using different electrolytes. Characterization of soil characteristics in sugarcane fields was conducted using 80 uniformly distributed (using a 38 m by 38 m grid) and 20 random sample positions. The two sugarcane fields (one with Margate soil and another with Immokalee soil) were 30 acres each. Soil samples sampled from 0-30cm, 30-60 cm, and 60-90 cm depths were analyzed for total carbon, total phosphorus, pH, oxalate iron, oxalate aluminum, oxalate phosphorus, and exchangeable calcium. The A horizon depths were fully explored and measured for the two sugarcane fields. The values of total phosphorus, total carbon, and pH were arranged in ascending order and clustered in to five clusters. One sample from each cluster was randomly selected for conducting phosphorus and ammonium sorption experiments. The A horizon from Immokalee and Margate soil, Bw horizon from Margate soil and Bh horizon from Immokalee soil were used for sorption experiments. The supporting electrolytes used to conduct phosphorus sorption experiments were; potassium chloride (0.01M KCl), calcium chloride (0.005M CaCl2), deionized water, simulated Florida rain, and fertilizer mixture (phosphorus, nitrogen, and potassium). The fertilizer mixture was prepared in simulated Florida rain using application rate, 50kg P2O5ha-1, 200kg Nha-1, and 200kg K2O ha-1. A saturated flow experiment where A horizon material was packed in a column (15 cm long and 7.5 cm in diameter) and fertilizer mixture pumped through at a rate of 10mL per minute was conducted. Column leaching experiments were conducted to identify changes in movement of phosphorus and nitrogen (ammonium and nitrate) with fluctuating water table (30 cm to 50 cm). In a lysimeter study, phosphorus, nitrogen, and potassium were applied to a lysimeter where sugarcane was planted and subjected to drip irrigation at water application rate of 2.3 Lh-1. The drip emitters’ spacing was 30.5 cm. The 15 cm depth increment (0-15 cm, 15-30cm, and 30-45 cm) was used for soil sampling. Bulk densities of soil horizons, saturated hydraulic conductivity (Ksat values), moisture release constants, and phosphorus sorption coefficients determined using different electrolytes and sorption kinetics parameters were used to calibrate Hydrus 1D. After calibration, column leaching results and lysimeter study results (bromide, phosphorus, ammonium, and nitrate) were used as validation data sets. Results after studying distribution of soil characteristics in two sugarcane fields have shown that unlike in a field with Immokalee soil where total carbon and oxalate aluminum influenced most total phosphorus distribution, total carbon influenced most total phosphorus distribution in Margate field. Soil characteristics were observed to vary spatially and with depth. Sorption of phosphorus by soil from least was deionized water, simulated Florida rain, potassium chloride (0.01M KCl), and calcium chloride (0.005M CaCl2). The calculated linearized sorption coefficient (0.01M KCl) compared well with linearized sorption coefficient (fertilizer mixture).  Negligible sorption of phosphorus was identified in E horizons sampled from two soil series (Margate and Immokalee).  The similarity in movement behavior of chloride and nitrate for saturated flow experiment showed that both can act as tracers. Since ammonium, nitrate, and chloride fit a convective-dispersive model, there was absence of physical non-equilibrium in saturated flow experiment. For both saturated and unsaturated flow experiment, phosphorus was more retarded than ammonium. Phosphorus and ammonium concentrations below the water table were higher when the water table was set at 50cm than 30cm. For the lysimeter study, differences in highest concentrations of bromide for 0-15 cm and 15-30 cm from irrigated (20 cm from center of plant row) and non-irrigated zone (50 cm from the center of plant row) were attributed to bromide uptake. Since high phosphorus and nitrogen concentrations were observed within the root zone (0-30 cm) and increasing concentrations in tissues with time, nutrients were managed within the root zone and plants responded to applied nutrients and moisture. The regression coefficient (R2) values for bromide (0.97) and ammonium (0.95) show that Hydrus 1D can be validated using data from column leaching experiment with water table depth set at 30 cm depth. For the irrigated zone in lysimeters (0-15 cm depth), 0.97 and 0.94 were the regression coefficient (R2) values for bromide and ammonium respectively. The low root mean square error (RMSE) values for linearized sorption coefficient value (0.01 M KCl) after validating Hydrus 1D with column data and lysimeter data show that modelers can model phosphorus movement with linearized sorption coefficient (0.01 M KCl). The significant results of this work were; (i) supporting electrolytes affect the sorption behavior of phosphorus in sandy soils and this has been shown in trend of linearized sorption coefficients; (ii) management of plant available nutrients within the root zone of sugarcane plants using drip irrigation; and (iii) a calibrated Hydrus 1D model for sugarcane production on sandy soils that can be used by modelers and farmers.
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 Augustine Muwamba.
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: UFE0044413:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044413/00001

Material Information

Title: Nitrogen and Phosphorus Movement in Sandy Soils of South Florida Used for Sugarcane Production With Elevated Water Table
Physical Description: 1 online resource (251 p.)
Language: english
Creator: Muwamba, Augustine
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: dripirrigation -- fertilizermixture -- tracer -- watertable
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: Farmers in southwest Florida need information related to distribution of soil characteristics in sugarcane fields with different soil orders since they affect accumulation of plant available nutrients within the root zone. Farmers need to know how fluctuating water table affects movement of plant available nutrients above and below the water table. Information on the distribution of nutrients leaching out of the root zone during water table management is needed by environmental scientists for the assessment of the adverse effects on the water quality. Water table depth is managed through pumping after rainfall events. Understanding of nutrient movement associated with decreases in water table depth is needed to minimize impacts on water quality by pumping activities. Reliable databases of constants like saturated conductivity values (Ksat) and phosphorus sorption coefficients (KD) are needed by scientists and farmers to accurately model water, nitrogen and phosphorus movement. Studies were conducted to document spatial distribution of soil characteristics in two mineral soils most widely used for sugar cane production, determine phosphorus sorption and kinetics associated with water table movement, and calibration of Hydrus 1D. The hypotheses of the studies were; (i) soil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary spatialy and with depth and this will lead to different patterns of phosphorus accumulations; (ii) phosphorus sorption coefficients (KD) when determined using different supporting electrolytes (0.01M KCl, 0.005M CaCl2, simulated Florida rain, deionized water, and fertilizer mixture) will significantly differ; (iii) reducing distance between water table and Bh horizon through lowering water table from 30 cm to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee soil; (iv) management of water table depth after rainfall events will lead to loss of plant available phosphorus and nitrogen out of the sugarcane plants’ root zone; (v) drip can be used to maintain high plant available nutrients within the root zone and minimize nutrients loss out of the root zone; and (vi) phosphorus leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolytes are used to model phosphorus movement. The objectives of the study were; (i) identifying the distribution of soil characteristics that affect accumulation of phosphorus and nitrogen in sugarcane fields with Immokalee fine sand and Margate fine sand; (ii) characterizing sorption of phosphorus using different electrolytes; (iii) studying movement of water determined by bromide tracer, phosphorus, and nitrogen in relation to fluctuating water table depth and drip irrigation, and (iv) modeling water (bromide), nitrogen, and phosphorus using linearized sorption coefficients determined using different electrolytes. Characterization of soil characteristics in sugarcane fields was conducted using 80 uniformly distributed (using a 38 m by 38 m grid) and 20 random sample positions. The two sugarcane fields (one with Margate soil and another with Immokalee soil) were 30 acres each. Soil samples sampled from 0-30cm, 30-60 cm, and 60-90 cm depths were analyzed for total carbon, total phosphorus, pH, oxalate iron, oxalate aluminum, oxalate phosphorus, and exchangeable calcium. The A horizon depths were fully explored and measured for the two sugarcane fields. The values of total phosphorus, total carbon, and pH were arranged in ascending order and clustered in to five clusters. One sample from each cluster was randomly selected for conducting phosphorus and ammonium sorption experiments. The A horizon from Immokalee and Margate soil, Bw horizon from Margate soil and Bh horizon from Immokalee soil were used for sorption experiments. The supporting electrolytes used to conduct phosphorus sorption experiments were; potassium chloride (0.01M KCl), calcium chloride (0.005M CaCl2), deionized water, simulated Florida rain, and fertilizer mixture (phosphorus, nitrogen, and potassium). The fertilizer mixture was prepared in simulated Florida rain using application rate, 50kg P2O5ha-1, 200kg Nha-1, and 200kg K2O ha-1. A saturated flow experiment where A horizon material was packed in a column (15 cm long and 7.5 cm in diameter) and fertilizer mixture pumped through at a rate of 10mL per minute was conducted. Column leaching experiments were conducted to identify changes in movement of phosphorus and nitrogen (ammonium and nitrate) with fluctuating water table (30 cm to 50 cm). In a lysimeter study, phosphorus, nitrogen, and potassium were applied to a lysimeter where sugarcane was planted and subjected to drip irrigation at water application rate of 2.3 Lh-1. The drip emitters’ spacing was 30.5 cm. The 15 cm depth increment (0-15 cm, 15-30cm, and 30-45 cm) was used for soil sampling. Bulk densities of soil horizons, saturated hydraulic conductivity (Ksat values), moisture release constants, and phosphorus sorption coefficients determined using different electrolytes and sorption kinetics parameters were used to calibrate Hydrus 1D. After calibration, column leaching results and lysimeter study results (bromide, phosphorus, ammonium, and nitrate) were used as validation data sets. Results after studying distribution of soil characteristics in two sugarcane fields have shown that unlike in a field with Immokalee soil where total carbon and oxalate aluminum influenced most total phosphorus distribution, total carbon influenced most total phosphorus distribution in Margate field. Soil characteristics were observed to vary spatially and with depth. Sorption of phosphorus by soil from least was deionized water, simulated Florida rain, potassium chloride (0.01M KCl), and calcium chloride (0.005M CaCl2). The calculated linearized sorption coefficient (0.01M KCl) compared well with linearized sorption coefficient (fertilizer mixture).  Negligible sorption of phosphorus was identified in E horizons sampled from two soil series (Margate and Immokalee).  The similarity in movement behavior of chloride and nitrate for saturated flow experiment showed that both can act as tracers. Since ammonium, nitrate, and chloride fit a convective-dispersive model, there was absence of physical non-equilibrium in saturated flow experiment. For both saturated and unsaturated flow experiment, phosphorus was more retarded than ammonium. Phosphorus and ammonium concentrations below the water table were higher when the water table was set at 50cm than 30cm. For the lysimeter study, differences in highest concentrations of bromide for 0-15 cm and 15-30 cm from irrigated (20 cm from center of plant row) and non-irrigated zone (50 cm from the center of plant row) were attributed to bromide uptake. Since high phosphorus and nitrogen concentrations were observed within the root zone (0-30 cm) and increasing concentrations in tissues with time, nutrients were managed within the root zone and plants responded to applied nutrients and moisture. The regression coefficient (R2) values for bromide (0.97) and ammonium (0.95) show that Hydrus 1D can be validated using data from column leaching experiment with water table depth set at 30 cm depth. For the irrigated zone in lysimeters (0-15 cm depth), 0.97 and 0.94 were the regression coefficient (R2) values for bromide and ammonium respectively. The low root mean square error (RMSE) values for linearized sorption coefficient value (0.01 M KCl) after validating Hydrus 1D with column data and lysimeter data show that modelers can model phosphorus movement with linearized sorption coefficient (0.01 M KCl). The significant results of this work were; (i) supporting electrolytes affect the sorption behavior of phosphorus in sandy soils and this has been shown in trend of linearized sorption coefficients; (ii) management of plant available nutrients within the root zone of sugarcane plants using drip irrigation; and (iii) a calibrated Hydrus 1D model for sugarcane production on sandy soils that can be used by modelers and farmers.
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 Augustine Muwamba.
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: UFE0044413:00001


This item has the following downloads:


Full Text

PAGE 1

1 NITROGEN AND PHOSPHORUS MOVEMENT IN SANDY SOILS OF SOUTH FLORIDA USED FOR SUGARCANE PRODUCTION WITH ELEVATED WATER TABLE By AUGUSTINE MUWAMBA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN P ARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Augustine M uwamba

PAGE 3

3 This dissertation is dedicated to my sister, Pauline Nankamb we, thank you for the inspiring words of wisdom.

PAGE 4

4 ACKNOWLEDGMENTS I am very happy that the chair of my s upervisory committee, Dr. Kelly T Morgan, provided incredible amounts of advice and funds for the program. I really landed on an i ncredible opportunity. I am glad that Dr. Kelly T. Morgan has been very patient with me. I appreciate that my co chair Prof. Peter Nkedi Kizza, taught me time management and being independent in conducting research activities. I would not further my grad uate school without Prof. Nkedi program. M y committee members, Prof. Howard Beck and Prof. Willie Harris, were very patient and provided advice in various components of this research. Thank you for the kindness and being pr ofessional. My parents did a good job in encouraging me to be brave and patient with school. I appreciated that my parents were very understanding for the times we have not been seeing each other.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ 4 LIST OF TA BLES ................................ ................................ ................................ ........... 8 LIST OF FIGURES ................................ ................................ ................................ ........ 9 ABSTRACT ................................ ................................ ................................ .................. 17 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ... 22 Study Overview ................................ ................................ ................................ ..... 23 Soil Nutrient Spatial Distribution ................................ ................................ ............ 26 Soil Nutrient Retention Characteristics ................................ ................................ .. 27 Modeling Water and Nutrient Movement ................................ ................................ 30 2 LITERATURE REVIEW ................................ ................................ ......................... 31 Sugarcane Production in United States ................................ ................................ 31 Spatial Variability of Soil Proper ties in Agricultural Fields ................................ ...... 31 Sorption of Phosphorus and Ammonium in Soils ................................ ................... 32 Sorption of Phosphorus ................................ ................................ ................... 32 Sorption of Ammonium ................................ ................................ .................... 34 Leaching of Phosphorus and Nitrogen in Soils ................................ ...................... 35 Leaching of Phosphor us ................................ ................................ .................. 35 Leaching of Nitrogen ................................ ................................ ....................... 36 Modeling Water, Bromide, Phosphorus, and Nitrogen Movement .......................... 37 3 SPATIAL DISTRIBUTION OF SOIL CHARACTERISTICS IN SUGARCANE FIELDS ................................ ................................ ................................ .................. 40 Background ................................ ................................ ................................ ........... 40 Materials and Methods ................................ ................................ .......................... 42 Soil Series ................................ ................................ ................................ ....... 42 Sampling and Measurement of Soil Properties ................................ ................ 43 R esults and Discussion ................................ ................................ ......................... 46 Concluding Remarks ................................ ................................ ............................. 54 4 SATURATED HYDRAULIC CONDUCTIVITY A N D SOIL WATER RETENTION CURVE ................................ ................................ ................................ .................. 60 Background ................................ ................................ ................................ ........... 60 Materials and Methods ................................ ................................ .......................... 61

PAGE 6

6 Results and Discussion ................................ ................................ ......................... 63 Concluding Remarks ................................ ................................ ............................. 65 5 EFFECTS OF SUPPORTING ELECTROLYTES AND FERTILIZER MIXTURE ON SORPTION BEHAVIOR OF PHOSPHORUS ................................ .................. 69 Background ................................ ................................ ................................ ........... 69 Materials and Methods ................................ ................................ .......................... 72 Soil Sampling and Soil Properties Measured ................................ ................... 72 Sorption Experiments ................................ ................................ ...................... 76 Results and Discussion ................................ ................................ ......................... 80 Identifying a Supporting Electrolyte that Mimics Fertilizer Mixture ................... 80 Parameters for Modeling Phosphorus and Ammonium Movement .................. 82 Concluding Remarks ................................ ................................ ............................. 83 6 EFFECT OF FLUCTUATING WATER TABLE ON PHOSPHORUS AND NITROGEN MOVEMENT ................................ ................................ ...................... 94 Background ................................ ................................ ................................ ........... 94 Materials and Methods ................................ ................................ .......................... 96 Saturated Flow Experiment ................................ ................................ ............. 97 Fluctuating Water Table ................................ ................................ ................ 100 Results and Discussion ................................ ................................ ....................... 102 Concluding Remarks ................................ ................................ ........................... 108 7 MANAGEMENT OF WATER AND NUTRIENTS WITHIN THE R OOT ZONE USING DRIP IRRIGATION ................................ ................................ .................. 127 Background ................................ ................................ ................................ ......... 127 Materials and Methods ................................ ................................ ........................ 130 Results and Discussion ................................ ................................ ....................... 133 Concluding Remarks ................................ ................................ ........................... 139 8 CALIBRATION AND VALIDATION OF HYDRUS 1D ................................ ........... 160 Background ................................ ................................ ................................ ......... 160 Materials and Methods ................................ ................................ ........................ 162 Results and Discussion ................................ ................................ ....................... 166 Concluding Remarks ................................ ................................ ........................... 169 9 CONCLUSIONS ................................ ................................ ................................ .. 201 APPENDIX A HISTOGRAMS FOR SPATIAL VARIABILITY DATA ................................ ............ 206 B GRAPHS FROM SORPTION EXPERIMENTS ................................ .................... 228

PAGE 7

7 C PROCEDURE FOR CALIBRATION AND VALIDATION OF HYDRUS 1D ........... 233 REFERENCES ................................ ................................ ................................ .......... 241 BIOGRAPHICAL SKETCH ................................ ................................ ......................... 251

PAGE 8

8 LIST OF TABLES Table page 2 1. Sugarcane production for sugar in million tons in States of United States ......... 39 2 2. Electrolytes in literature used by researchers ................................ .................... 39 3 1. Exploratory statistics of soil properties for Immokalee soil ................................ 56 3 2. Exploratory statistics of soil properties for Margate soil ................................ ..... 57 4 1. Parameters for modeling water movement ................................ ........................ 67 5 1. Physical and chemical properties of soils ................................ .......................... 85 5 2. Electrolytes used for phosphorus sorption experiments ................................ ..... 85 5 3. Maximum i nitial concentrations for phosphorus sorption isotherms ................... 85 5 4. Ionic streng th (I) and pH of the soil solution ................................ ....................... 85 5 5. Linear ized sorption coefficients (K D ) ................................ ................................ .. 86 5 6. Statistical differences between linear so rption coefficients (K D ) ......................... 86 5 7. Phosphorus sorption kinetics parameters: standard error in parentheses .......... 86 5 8. Sorption of phosph orus on E horizon of Immokalee soil ................................ .... 86 6 1. Properties of the A horizon used for column leaching experiment ................... 111 6 2. Peak times for bromide, ammonium nitrogen, and Phosphorus ...................... 111 7 1. Biomass accumulation in tissues as a function of time ................................ .... 141 7 2. Phosphorus accum ulation in tissues as a function of time ............................... 142 7 3. Nitrogen accumulation in tissues as a function of time ................................ .... 143 8 1. Model inputs for m odel calibration: modeling water movement ........................ 171 8 2. Model inputs for model validation: Lysimeter study ................................ .......... 171 8 3. Root mean square erro r and mean absolute e rror ................................ ........... 172 8 4. Root mean square error and mean absolute error: irrigated zone .................... 172 8 5. Root mean square error an d mean absolute error: non irrigated zone ............ 173

PAGE 9

9 LIST OF FIGURES Figure page 3 1. Sampling design from a sugarcane field with Margate soil series ..................... 58 3 2. Sampling design from a sugarcane field with Immokalee soil series. ................. 59 4 1. Moisture release curve for A horizon ................................ ................................ 67 4 2. Moisture release curve for E horizon ................................ ................................ 67 4 3. Moisture release curve for Bh horizon ................................ ............................... 68 5 1. Two site model ................................ ................................ ................................ .. 87 5 2. Phosphorus sorption isotherm in A and Bh horizons of Immokalee soil ............. 87 5 3. Phosphorus sorption isotherm in A and Bh horizons o f Immokalee soil ............. 88 5 4. Phosphorus sorption isotherm in A and Bw horizons of Margate soil ................. 88 5 5. Phosphorus sorption isotherm in A and Bw horizons of Margate soil ................. 89 5 6. Sorbed phosphorus versus equilibrium solution phosphorus ............................. 89 5 7. Sorbed phosphorus vers us equilibrium solution phosphorus ............................. 90 5 8. Sorbed phosphorus versus equilibrium solution phosphorus ............................. 90 5 9. Sorbed phosphorus vers us equilibrium solution phosphorus ............................. 91 5 10. Sorbed phosphorus versus equilibrium solution phosphorus ............................. 91 5 11. Relative concentrati on (C/C0) as a function of time with potassium chloride ..... 92 5 12. Relative concentration (C/C0) as a function of time with fertilizer mixture .......... 92 5 13. Sorbed ammonium versus equilibrium solution ammonium ............................... 93 6 1. Design of column leaching experiment for saturated flow ................................ 111 6 2. Design of column leaching experiment for unsaturated flow ............................ 112 6 3. Relative concentration (C/C0) as a function of pore volume for saturated flow 112 6 4. Relative ammonium concentration (C/CO) as a function of pore volume ......... 113 6 5. Relativ e phosphorus concentration (C/C0 ) as a function of pore volume ......... 113

PAGE 10

10 6 6. Bromide concentration as a function of time for cup A ................................ ..... 114 6 7. Nitrate nitrogen concentration as a function of time for cup A .......................... 114 6 8. Ammonium nitrogen concentration as a function of time for cup A .................. 115 6 9. Phosphorus concentration as a function of time for cup A ............................... 115 6 10. Bromide concentration as a function of time for 30 cm water table depth ........ 116 6 11. Nitrate nitrogen concentration as a func tion of time ................................ ......... 116 6 12. Ammonium nitrogen concentration as a function of time ................................ 117 6 13. Phosphorus concentration as a function of ti me for 30 cm water table depth .. 117 6 14. Bromide concentration as a function of time for 50 cm water table depth ........ 118 6 15. Nitrate n itrogen concentration as a function of time ................................ ........ 118 6 16. Ammonium nitrogen concentration as a function of time ................................ 119 6 17. Phosphorus c oncentration as a function of time for 50 cm water table depth .. 119 6 18. Bromide concentration as a function of time for cup E ................................ ..... 120 6 19. Bromide concentration as a function of time for cup E ................................ ..... 121 6 20. Nitrate nitrogen concentration as a functio n of time for cup E ......................... 121 6 21. Nitrate nitrogen concentration as a function of time for cu p E .......................... 121 6 22. Ammonium nitrogen concentration as a function of time for cu p E ................. 122 6 23. Ammonium nitrogen concentration as a function of time fo r cup E ................. 122 6 24. Phosphorus concentration as a function of time for cu p E ............................... 123 6 25. Phosphorus concentration as a function of time for cu p E ............................... 123 6 26. Bromide concentration as a function of time for cup B h ................................ ... 124 6 27. Bromide concentration as a function of time for cup Bh ................................ .. 124 6 28. Nitrate nitrogen concentration as a function of time for cup Bh ........................ 125 6 29. Nitrate nitrogen concentration as a function of time for cup Bh ........................ 125 6 30. Ammonium nitrogen concentration as a function of time for cup Bh ............... 126

PAGE 11

11 6 31. Ammonium nitrogen concentration as a function of time for cup Bh ................ 126 7 1. Layout of the lysimeter ................................ ................................ .................... 144 7 2. Moisture content as a function of time and depth for irrigated zone ................. 145 7 3. Moisture content as a fun ction of time and depth for non irrigated zone .......... 146 7 4. Bromide as a function of time and depth for irrigated zone .............................. 147 7 5. Bromide as a function of time and depth for non irrigated zo ne ...................... 148 7 6. Phosphorus as a function of time and depth for irrigated zone ........................ 149 7 7. Phosphoru s as a function of time for non irrigate d zone ................................ .. 150 7 8. Ammonium nitrogen as a function of time for irrigated zone ............................ 151 7 9. Ammonium nitrogen as a fun ction of time and dept h for non irrigated zone .... 152 7 10. Nitrate nitrogen as a function of time and depth for irrigated zone ................... 153 7 11. Nitrate nitrogen as a fun ction of time and depth for non irrigated zone ............ 154 7 12. Phosphorus in tissues of short plants as a function of time ............................. 155 7 13 Phosphorus in tissues of medium plants as a function of time ......................... 155 7 14. Phosphorus in tissues of tall plants as a function of time ................................ 156 7 15. Total nitrogen in tissues of short plants as a function of time ........................... 156 7 16. T otal nitrogen in tissues of medium plants as a function of time ...................... 157 7 17. Total nitrogen in tissues of tall plants as a function of time .............................. 157 7 18. Biomass accumulation for short plants as a function of time ............................ 158 7 19. Biomass accumulation for medium plants as a function of time ....................... 158 7 20. Biomass accumulation for tall plants as a function of time ............................... 159 8 1. Rain added to columns as a function of time for 30 cm water depth ................ 174 8 2. Experimental and simulated moisture as function of time ................................ 175 8 3. Experimental and simulated bromide as function of time for cup A .................. 175 8 4. Experimental and simulated ammonium as function of time for cup A ............. 175

PAGE 12

12 8 5. Experimental and simulated phosphorus as fu nction of time for cup A ............ 176 8 6. Experimental and simulated phosphorus as fu ncti on of time for cup A ............ 176 8 7. Experimental and simulated phosphorus as fu nction of time for cup A ........... 177 8 8. Experimental and simulat ed phosphorus as fu nction of time for cup A ........... 177 8 9. Experimental and simulated phosphorus as fu nction of time for cup A ............ 178 8 10. Experimental and simulated phosphorus as fu nction of time for cup A ............ 178 8 11. Natural rain received in Immokalee (6/6/11 7/9/11). ................................ ........ 179 8 12. Experimental and simulated bromide as function of time for irrigated zone ..... 179 8 13. Experimental and simulated bromide as function of time ................................ 180 8 14. Experimental and simulated bromide as function of time for irrigated zone ..... 180 8 15. Experimental and simulated bromide as functio n of time ................................ 181 8 16. Experimental and simulated nitrate as function of time for irrigated zone ........ 1 81 8 17. Experimental and simulated nitrate as function of time for non irri gated zone 182 8 18. Experimental and simulated nitrate as function of time for irrigated zone ........ 182 8 19. Experimental and simulated nitrate as function of time for non irrigated zone 183 8 20. Experimental and simulated ammonium as function of time for irrigated zone 183 8 21 Experimental and simulated ammonium as function of time ......................... 184 8 22. Experimental and simulated ammonium as function of time for irrigated zone 184 8 23. Experimental and simulated ammonium as function of time .......................... .. 185 8 24. Experimental and simulated ammonium as function of time for irrigated zone 185 8 25. Experimental and simulated phosphorus as function of time for irrigated zone 186 8 26. Experimental and simulated phosphorus as funct ion of time for irrigated zone .... 186 8 27. Experimental and simulated phosphorus as function of time for irrigated zone 187 8 28 Experimental and simulated phosphorus as function of time for irrigated zone 187 8 29. Experimental and simulated phosphorus as functio n of time ........................... 188

PAGE 13

13 8 3 0. Experimental and simulated phosphorus as function of time for irrigated zone .. 188 8 31. Experimental and simulated phosphorus as functio n of time ........................... 189 8 32. Experimental and simulated phosphorus as function of time for irrigated zone 189 8 33. Experimental and simulated phosphorus as function of ti me for irrigated zone 190 8 34. Experimental and simulated phosphorus as functio n of time .......................... 190 8 35. Experimental and simulated ph osphorus as function of time for irrigated zone 191 8 36. Experimental and simulated phosphorus as functio n of time .......................... 191 8 37. Expe rimental and simulated phosphorus as function of time for irrigated zone 192 8 38. Experimental and simulated phosphorus as function of time for irrigated zone 192 8 39. Experimental and simulated phosphorus as function of time ........................... 193 8 40. Experimental and simulated phosphorus as function of time for irrigated zone 193 8 41. Experimental and simulated phosphorus as functio n of time ........................... 194 8 42. Experimental and simulated phosphorus as function of time f or irrigated zone 194 8 43. Experimental and simulated phosphorus as function of time for iriigated zone 195 8 44. Experimental and simulated phosphorus as functio n of time .......................... 195 8 45. Experimental and simulated phosphorus as function of time for irrigated zone 196 8 46. Experimental and simulated phosphorus as functio n of time .......................... 196 8 47. Experimental and simulated phosphorus as function of time for irrigated zone 197 8 48. Experimental and simulated phosphorus as function of time for irrigated zone 197 8 49. Experimental and simulated phosphorus as function of time .......................... 198 8 50. Experimental and simulated phosphorus as function of time for irrigated zone 198 8 51. Experimental and simulated phosphorus as function of time .......................... 199 8 52. Experimental and simulated phosphorus as function of time for irrigated zone 199 8 53. Simulated p hosphorus in leachate as a function of horizon depths .................. 200 A 1. Histogram for A horizon depth from Immokalee soil ................................ ........ 206

PAGE 14

14 A 2. Histogram for total carbon (0 30cm) from Immokalee soil ................................ 206 A 3. Histogram for total carbon (30 60cm) from Immokalee soil .............................. 207 A 4. Histogram f or total carbon (60 90cm) from Immokalee soil .............................. 207 A 5. Histogram for pH (0 30cm) from Immokalee soil ................................ ............. 208 A 6 Histogram for pH (30 6 0cm) from Immokalee soil ................................ ........... 208 A 7. Histogram for pH (60 90cm) from Immokalee soil ................................ ........... 209 A 8. Histogram for oxalate aluminum (0 30 cm) f rom Immokalee soil ..................... 209 A 9. Histogram for oxalate aluminum (30 60 cm) from Immokalee soil ................... 210 A 1 0. Histogram for oxalate aluminum ( 60 90 cm) from Immokalee soil ................... 210 A 11. Histogram for oxalate iron (0 30 cm) from Immokalee soil ............................... 211 A 12. Histogram for oxalate iro n (30 60 cm) from Immokalee soil ............................. 211 A 13. Histogram for oxalate iron (60 90 cm) from Immokalee soil ............................. 212 A 14. Histogram for oxalate phosphorus (0 30 cm) from Immokalee soil .................. 212 A 15. Histogram for phosphorus saturation (0 30 cm) from Immokalee soil .............. 213 A 16. Hist ogram for total phosphorus (0 30 cm) from Immokalee soil ....................... 213 A 17. Histogram for total phosphorus (30 60 cm) from Immokalee soil ..................... 214 A 18. Histogram for total phosphorus (60 90 cm) from Immokalee soil ..................... 214 A 19. Histogram for exchangeable calcium (0 30 cm) from Immokalee soil .............. 215 A 20. Histogram for exchangeable calcium (30 60 cm) from Immokalee soil ............ 215 A 21. Histogram for exchangeable calcium (60 90 cm) from Immokalee soil ............ 216 A 22. Histogram for A horizon depth from Margate soil ................................ ............. 216 A 23. Histogram for total carbon (0 30 cm) from Margate soil ................................ ... 217 A 24. Histogram for total carbon (30 60 cm) from Margate soil ................................ 217 A 25. Histogram for total carbon (60 90 cm) from Margate soil ................................ 218 A 26. Histogram for pH (0 30 cm) from Margate soil ................................ ................. 218

PAGE 15

15 A 27. Histogram for pH (30 60 cm) from Margate soil ................................ ............... 219 A 28. Histogram for pH (60 90 cm) from Margate soil ................................ ............... 219 A 29. Histogram for oxalate aluminum (0 30 cm) for Margate soil ............................. 220 A 30. Histogram for oxalate aluminum (30 60 cm) from Margate soil ........................ 220 A 31. Histogram for oxalate aluminum (60 90 cm) from Margate soil ........................ 221 A 32. Histogram for oxalate iron (0 30 cm) from Margate soil ................................ ... 221 A 33. Histogram for oxalate iron (30 60cm) from Margate soil ................................ .. 222 A 34. Histogram for oxalate iron (60 90cm) from Margate soil ................................ .. 222 A 35. Histogram for oxalate phosphorus (0 30cm) from Margate soil ....................... 223 A 36. Histogram for oxalate phosphorus (60 90 cm) from Margate soil ..................... 223 A 37. Histogram for phosphorus saturation ratio (0 30 cm) from Margate soil ........... 224 A 38. Histogram for phosphorus saturation ratio (60 90 cm) from Margate soil ......... 224 A 39. Histogram for total phosphorus (0 30 cm) from Margate soil ........................... 225 A 40. Histogram for total phosphorus (30 60 cm) from Margate soil ......................... 225 A 41. Histogram for total phosphorus (60 90 cm) from Margate soil ......................... 226 A 42. Histogram for exchangeable calcium (0 30 cm) from Margate soil .................. 226 A 4 3. Histogram for exchangeable calcium (30 60 cm) from Margate soi l ................ 227 A 44. Histogram for exchangeable calcium (60 90 cm) from Margate soil ................ 227 B 1. Phosphorus sorption isotherm in A and Bh horizo ns of Immokalee soil ........... 228 B 2. Phosphorus sorption isotherm in A and Bh horizons of Immokalee soil ........... 228 B 3. Phosphorus sorption isoth erm in A and Bh horizons of Immokalee soil ........... 229 B 4. Phosphorus sorption isotherm in A and Bw horizons of Margate soil ............... 22 9 B 5. Phosp horus sorption isotherm in A and Bw horizons of Margate soil ............... 230 B 6. Phosphorus sorption isotherm in A and Bw horizons of Margate soil ............... 230 B 7. Relative concentration (C/C0) as a function of time ................................ ......... 231

PAGE 16

16 B 8. Relative concentratio n (C/C0) as a function of time ................................ ......... 231 B 9. Relative concentration (C/C0) as a function of time with fertilizer mixture ........ 232 C 1. Main processes ................................ ................................ ............................... 233 C 2. Geometry information ................................ ................................ ...................... 233 C 3. Time information ................................ ................................ ............................. 234 C 4. Soil hydraulic model ................................ ................................ ........................ 234 C 5. Wate r flow parameters ................................ ................................ .................... 235 C 6. Water flow boundary conditions ................................ ................................ ....... 235 C 7. Non equilibrium solute transport models ................................ ......................... 236 C 8. Solute transport parameters ................................ ................................ ............ 236 C 9. Solute transport and reaction parameters ................................ ........................ 237 C 10 Solute transport boundary conditions ................................ .............................. 237 C 11. Root water and uptake model ................................ ................................ .......... 238 C 12. Root water uptake model parameters ................................ .............................. 238 C 13. Root growth parameter logistic growth function ................................ ............... 239 C 14. Time variable boundary conditions ................................ ................................ .. 239 C 15. Soil profile graphical editor ................................ ................................ ............ 240

PAGE 17

17 Abstract of Dissertation Presented to the Graduate School of the Un iversity o f Florida in Partial Fulfillment of the Requirements for the Degre e of Doctor of Philosophy NITROGEN AND PHOSPHORUS MOVEMENT IN SANDY SOILS OF SOUTH FLORIDA USED FOR SUGARCANE PRODUCTION WITH ELEVATED WATER TABLE By Augustine Muwamba August 2012 Chair: Kelly T. Morgan Cochair: Peter Nkedi Kizza Major: Soil and Water Science Farmers in southwest Florida need information related to distribution of soil characteristics in sugarcane fields with different soil orders since they affect accumul a tion of plant available nutrients with in the root zone Farmers need to know how fluctuating water table affects movement of plant available nutrients above and below the water table. Information on the distribution of nutrients leaching out of the root zone during water table management is needed by e n vironmental scientists for t he assessment of the adverse effec ts on the water quality Water table depth is managed through pumping after rainfall events Understanding of nutrient movement associated with decreases in water table depth is needed to minimize impacts on water quality by pumping activities. Reliable databases of constants like saturated conductivity values (Ksat) and phosphorus sorption coefficients (K D ) are needed by scientists and farmers to accurately model water, nitrogen and phosphorus movement. Studies were conduc ted to document spatial distribution of soil characteristics in two mineral soils

PAGE 18

18 most widely used for sugar cane production, determine phosphorus sorption and kinetics associated with water table movement, and calibration of Hydrus 1D. The hypotheses of t he stud ies were; (i) s oil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary spatially and with depth and this will lead to different patterns of phosphorus accumulations ; (ii) p hosphorus sorption coeffici ents (K D ) when determined using different supporting electrolytes (0.01M KCl, 0.005M CaCl 2 simulated Florida rain, deionized water, and fertilizer mixture) will significantly differ ; (iii ) r educing distance between water table and Bh horizon through lower ing water table from 30 cm to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee soil ; (iv) m anagement of water table depth after rainfall events will lead to loss of plant available phosphorus and nitrogen o ; (v) d rip can be used to maintain high plant available nutrients within the root zone and minimize nutrients loss out of the root zone ; and (vi) phosphorus leaching will be over predicted and or under predicted when so rption coefficients determined using different electrolytes are us ed to model phosphorus movement The objectives of the study were; (i) identifying the distribution of soil characteristics th at affect accumulation of phosphorus and nitrogen in sugarcane f ields with Immokalee fine sand and Margate fine sand ; (ii) characterizing sorption of phosphorus using different electrolytes ; (iii) studying movement of water determined by bromide tracer, phosphorus, and nitrogen in relation to fluctuating water table de pth and drip irrigation and (iv) modeling water ( bromide ) nitrogen, and phosphorus using linearized sorption coefficients determined using different electrolytes. Cha racterization of soil characteristics in sugarcane fields was conducted using 80 unifor m l y distributed

PAGE 19

19 (using a 38 m by 38 m grid) and 20 random sample positions. The two sugarcane fields (one with Margate soil and another with Imm okalee soil) were 30 acres each Soil samples sampled from 0 30c m, 30 60 cm, and 60 90 cm depths were analyzed fo r total carbo n, total phosphorus, pH, oxalate iron oxalate aluminum oxalate phosphorus and exchangeable calcium The A horizon depths were fully explored and measured for the two sugarcane fields. The values of total phosphorus, total carbon, and pH wer e arranged in ascending order and clustered in to five clusters. One sample from each cluster was randomly selected for conducting phosphorus and ammonium sorption experiments. The A horizon from Immokalee and Margate soil, Bw horizon from Margate soil and Bh horizon from Immokalee soil were used for sorption experiments. The supporting electrolytes used to conduct phosphorus sorption experiments were; potassium chloride (0.01M KCl), calcium chloride (0.005M CaCl 2 ) deionized water, simulated Florida rain and fertilizer mixture ( phosphorus, nitrogen, and potassium) The fertilizer mixture was prepared in simulated Florida rain using application rate, 50kg P 2 O 5 ha 1 200kg Nha 1 and 200kg K 2 O ha 1 A saturated flow experiment where A horizon material was pac ked in a column (15 cm long and 7.5 cm in diameter) and fertilizer mixture pumped through at a rate of 10 mL per minute was conducted Column leaching experiments were conducted to identify changes in movement of phosphorus and nitrogen (ammonium and nitrat e) with fluctuating water table (30 cm to 50 cm). In a lysimeter study, phosphorus, nitrogen, and potassium were applied to a lysimeter where sugarcane was planted and subjected to drip irrigati on at water application rate of 2.3 Lh 1 pacing was 30.5 cm. The 15 cm depth increment (0 15 cm, 15 30cm, and 30 45 cm) was used for soil sampling. B ulk densities of soil horizon s

PAGE 20

20 saturated hydraulic conductivity ( Ksat values ) moisture release constants, and phosphorus sorption coefficients det ermined using different electrolytes and sorption kinetics parameters were used to calibrate Hydrus 1D. After calibration, column leaching results and lysimeter study results ( bromide phosphorus, ammonium, and nitrate) were used as validation data sets. R esults after studying distribution of soil characteristics in two sugarcane fields have shown that u nlike in a field with Immokalee soil where total carbon and oxalate aluminum influenced most total phosphorus distribution, total carbon influenced most tot al phosphorus distribution in Margate field. Soil characteristics were observed to vary spatially and with depth. Sorption of phosphorus by soil from least was deionized water, simulated Florida rain, potassium chloride (0.01M KCl), and calcium chloride (0 .005M CaCl 2 ) The calculated linearized sorption coefficient (0.01M KCl) compared well with linearized sorption coefficient (fertilizer mixture). Negligible sorption of phosphorus was identified in E horizons sampled from two soil series (Margate and Immo kalee). The similarity in movement behavior of chloride and nitrate for saturated flow experiment showed that both can act as tracers. Since ammonium, nitrate, and chloride fit a convective dispersive model, there was absence of physical non equilibrium i n saturated flow experiment. For both saturated and unsaturated flow experiment, p hosphorus was more retarded than ammonium. Phosphorus and ammonium concentrations below the water table were higher when the water table was set at 50cm than 30cm. For the ly simeter study, d ifferences in highest concentrations of bromide for 0 15 cm and 15 30 cm from ir rigated (20 cm from center of plant row) and non irrigated zone (50 cm from the center of plant row) we re attributed to bromide uptake. Since high phosphorus an d nitrogen

PAGE 21

21 concentrations were observed within the root zone (0 30 cm) and increasing concentrations in tissues with time, nutrients were managed within the root zone and plants responded to applied nutrients and moisture. The regression coefficient (R 2 ) v alues for bromide (0.97) and ammonium (0.95) show that Hydrus 1D can be validated using data from column leaching experiment with water table depth set at 30 cm depth. For the irrigated zone in lysimeters (0 15 cm depth), 0.97 and 0.94 were the regression coefficient ( R 2 ) values for bromide and ammonium respectively. The low root mean square error ( RMSE ) values for linearized sorption coefficient value (0.01 M KCl) after validating Hydrus 1D with column data and lysimeter data show that modelers can model p hosphorus movement with linearized sorption coefficient (0.01 M KCl). The significant results of this work were; (i) s upporting electrolytes affect the sorption behavior of phosphorus in sandy soils and this has been shown in trend of li nearized sorption c oefficients; (ii) management of plant available nutrients within the root zone of sugarcane plants using drip irrigation; and (iii) a calibrated Hydrus 1D model for sugarcane production on sandy soils that can be used by modelers and farmers.

PAGE 22

22 CHAPT ER 1 INTRODUCTION The mineral soils used for sugarcane production in Florida are classified as Entisols, Mollisols, and Spodosols ( Obreza et al., 1998 ; Pitts et al. 1993 ). The soils have varying amounts of organic matter. The texture s range f rom coarse to fine textured sands ( Soil Survey Staff, 1996 ) Seepage irrigation by r aising the water table is possible due to the close proximity o f bed rock, and elevated water tables that are used to manage moisture content for plant uptake. Water table depth is lowe red from 30 cm depth t o 50 cm depth for good root aeration. The mineral soils demand proper water and fertility management before being used for sugarcane production. This is attributed to their low nutrient content and water holding capacity ( Soil survey staff, 1996 ) The average A horizon depths within the sugarcane fields differ due to land leveling. This leads to differences in availability of nutrients in the root zone of sugarcane plants. The A horizon depth is important in nutrient retention since it contains higher organic matter than the E horizon ( Li et al 1997 ) The E horizon s are eluted of nutrients and typically consist of uncoated sands that do not retain nutrients, thus act ing essentially as media through which nutrients pass to the water ta ble (Pant et al., 2002). The Bh horizon is the most reactive component of Spodosols. The Bh horizon accumulates amorphous mixtures of organic matter and aluminum (Soil Survey Staff 1996 ). Histosols in Palm Beach County, and Spodosols and Entisols in Hendr y and Glades C ounties are the dominant soil orders, used for sugarcane production in South Florida (Anderson, 1990, Obreza et al., 1998). The two soil series that are dominantly used for sugarcane production in Southwest Florida are Margate and Immokalee ( US Department of

PAGE 23

23 Agriculture. Natural Resources Conservation Service, 1995; Soil survey laboratory manual). In Florida 75 % of sugarcane acreage is in Palm Beach. Seventy five percent of the total harvested tonnage comes from Palm Beach Sugarca ne product ion (14 million tons) was reported in Florida (NASS 2011 ). Louisiana (11.3 million tons), Texas (1.7 million tons) and Hawaii (1.2 million tons) are o ther states in United States where sugarcane is grown commercially. Other countries that contribute most to the world sugar production are Brazil, China, India, and Thailand. In 2010, 161,900 metric tons (raw value) were the world sugar production. Brazil sugar production was 40475 metric tons (25% of the world sugar production). Brazil is the current world l argest producer of sugarcane. Sucrose, bagasse, ethanol, and molasses are the value added products from sugarcane. Sucrose is used as a sweetening agent. Molasses are used to feed animals. Bagasse is used as fuel for mills when burnt and for ethanol produc tion. Study Overview The commercial sugarcane growing area in Florida is located at sout hern tip of Lake Okeechobee. A vailability of phosphorus and nitrogen to plants can be obtained by monitoring phosphorus and n i trogen concentrations within and below the root zone The root zone is 30 cm from soil surface since the majori ty of the roots grow within the 30 cm (Smith et al., 2005). Monitoring phosphorus and nitrogen concentrations below the root zone helps in assessing ground water quality deterioration Th e phosphorus and nitrogen concentrations in perimeter ditches of sugarcane fields should be carefully monitored. The phosphorus and nitrogen transported by water canals from sugarcane fields in to surrounding waters can also lead to water quality deteriora tion ( Izuno et al., 1991). In this study, two sugarcane fields (each of 30 acres) were used to study the

PAGE 24

24 spatial distribution of soil physical and chemical characteristics. Soil samples from five sample positions determined to have different characteristi cs in each sugarcane field were used to study the phosphorus sorption behavior. Sample positions were determined by clusteri ng the values of the characteristics (total carbon pH, and total phosphorus ) in to five groups. Characteristic values were arranged in ascending order and divided into five groups (each group constituted of 20 sample positions). One sample from each of the five groups was obtained to represent each group rando ml y Deionized water, potassium chloride ( 0.01M KCl ) calcium chloride ( 0.00 5M CaCl 2 ) simulated Florida rain and fertilizer mixture were used as supporting electrolytes for sorption experiments. The data were subjected to Langmuir, Freundlich, and linear equations and linear ized sorption coefficients (K D ) calculated. A supporti ng electrolyte that yielded linearized sorption coefficients (K D ) value close to fertilizer mixture was identified. The linearized sorption coefficients (K D ) were used for modeling phosphorus movement under saturated and unsaturated conditions Column leac hing experiments (saturated and unsaturated) were conducted. For the saturated flow experiment, A horizon of Immokalee soil was packed in the column according to the field bulk density. For the unsaturated flow experiment, A, E, and Bh horizon materials fr om Immokalee soil were packed in the column. The water table depth for unsaturated flow experiment was set at 30 cm for six weeks and later lowered to 50 cm for six weeks. The masses of phosphorus and nitrogen for solutions collected from cup E and cup Bh with water table set at 30 cm were compared to masses with water table set a t 50 cm The bromide, phosphorus, and nitrogen (nitrate and ammonium) concentration data from column leaching experiments were used as validation datasets for Hydrus 1D. Break thro ugh

PAGE 25

25 curves were plotted from saturated flow experiment data and simulated. For phosphorus break through curves were modeled using linearized sorption coefficient ( potassium chloride ( 0.01M KCl ) and fertilizer mixture). A lys imeter study was used to invest igate the effect of drip irrigation on use efficiency of applied phosphorus and nitrogen The phosphorus, bromide, nitrate, and ammonium data sets were used to validate Hydrus 1D. The statistical tools, root mean square error ( RMSE ) and mean absolute erro r ( MAE ) were used to assess how well Hydrus 1D can be validated. The h ypo theses of this study were; (i) soil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary spatialy and with depth and this will lead t o different patterns of phosphorus accumulations; (ii) phosphorus sorption coefficients (K D ) when determined using different supporting electrolytes (0.01M KCl, 0.005M CaCl 2 simulated Florida rain, deionized water, and fertilizer mixture) will significant ly differ; (iii) reducing distance between water table and Bh horizon through lowering water table from 30 cm to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee soil; (iv) management of water table depth a fter rainfall events will lead to loss of plant available phosphorus high plant available nutrients within the root zone and minimize nutrients loss out of the root zone; and (vi) phosphorus leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolytes are used to model phosphorus movement.

PAGE 26

26 Soil Nutrient Spatial Distribution The two soil series used in this study were Margate and Immokalee. Unlike Immokalee with Bh horizon after E horizon, Margate is characterized by Bw after E horizon. The Bh horizon is characterized by accumulation of organic matter and aluminum with or without iron. The Bw is the transition hor izon that is undergoing development. That is transition from parent material to soil. The spatial variability study was conducted to understand and identify the differences between Immokalee and Margate soil series. This was achieved by quantifying the mag nitude of soil property values. The chemical characteristics measured were soil pH, total carbon, o xalate iron oxalate aluminum oxalate phosphorus exchangeable calcium, and total phosphorus The A horizon depth was the physical characteristic measured. The soil samples c haracterized for soil characteristics were sampled from 0 30cm, 30 60cm, and 60 90cm depths. The pH was measured using soil to water solution of 1:2. The total carbon was measured using combustion method with carbo n and nitrogen analyzer. Total phosphorus was determined using a cid digestion method. Oxalate iron, aluminum and phosphorus were extracted using oxalate solution (ammonium oxalate and oxalic acid) and analyzed using inductively coupled plasma (ICP). Exchangeable calcium was extr acted with 0.2M NH 4 Cl and measured using inductively coupled plasma The physical characteristic measured was A horizon depth. The same application rates of phosphorus, potassium, and nitrogen are used on sugarcane fields with different soil series ( Obreza et al., 1998 ) The amount of phosphorus, potassium, and nitrogen in discharge from a farm unit will depend on the soil characteristi cs (organic matter,

PAGE 27

27 oxalate aluminum, iron, and calcium ) of the soil series in that unit. The distribution of soil characte ristics in sugarcane fields helps researchers to identify positions in agricultural fields that are used for monitoring nutrient movement Soil Nutrient Retention Characteristics The natural rain and irrigation water dissolves the fertilizer mixtu re (nitr ogen, phosphorus, and potassium ) applied to meet plant nutrient requirements. The rainy season in s outhwest Florida is between June and August The solution that leaches out of th e root zone contains phosphorus, potassium, and nitrogen (ammonium, nitrate, and nitrite ) However the phosphorus sorption behavior in a mixture of fertilizers has not been explored by researchers. There is need to identify appropriate electrolyte with ionic strength that approximates liquid phase in sugarcane fields. The soil hor izons from five field positions were rando ml y selected from the two sugarcane fields used for spatial variability studies. The sorption experiments were conducted using A and Bh of Immokalee soil. The A and Bw horizons of Margate soil were used for sorptio n study. The E horizon acts as a medium through which phosphorus and nitrogen pass to the reactive Bw and Bh horizons. Therefore E horizon was not used for sorption study. The five soil samples were used as replicates from sugarcane fields. The supporting electrolytes that were used for sorption study were potassium chloride ( 0.01M KCl ) calcium chloride ( 0.005M CaCl 2 ) deionized water, and simulated Florida rain. The fertilizer mixture (nitrogen, phosphorus, and potassium ) was prepared in simulated Florida rain The concentrations of ions in Florida rain are 13.9, 30.7, 9.5, 21.7, 11.7, 18 L 1 of NO 3 SO 4 2 NH 4 + Ca 2+ Mg 2+ Na, K + and Cl respectively (Villapando,1997) The linear ized sorption coefficient ( K D ) is used by modelers to model

PAGE 28

28 phosphorus and amm onium movement in soils ( Simunek et al. 1999 ). The linearized sorption coefficient values from sorption isotherms using potassium chloride ( 0.01M KCl ) calcium chloride ( 0.0 0 5M CaCl 2 ) simulated Florida rain, and deionized wa ter we re compared with the fertilizer The phosphorus sorption kinetics experiments were conducted. The instantaneous sorption fraction (F) and k inetic rate coefficient for desorption constant (k 2 ) were used for modeling phosphorus movement The instantaneous so rption fraction (F) and k inetic rate coefficient for desorption constant (k 2 ) are inputs for Hydrus 1D since the two site (instantaneous and sorption kinetics) model is used to characterize phosphorus sorption In sugarcane fields, the water table is firs t set at approximately 30 cm from the soil surface for more than three weeks after planting management The water table is then lowered to approximately 50 cm for the remaining crop season Water table is lowered to 50 cm de pth through pumping and water channeled to perimeter ditches whenever rain is rec eived in sugarcane field. In this study, it is hypothesized that whenever the water table is lowered from 30 cm to 50 cm (closer to Bh horizon), diffusion of phosphorus and am monium below the water table is increased Additionally a fter rain fall events, s ignificant amounts of phosphorus, nitrogen, and potassium are hypothesized to leach out of the root zone through management of water table after rainfall events The water tab le depth s of the columns were maintained at 3 0 cm from soil surface for six weeks then lowered to 50 cm fr om the surface for another six weeks. A lysimeter study was used to monitor movement of water, bromide, phosphorus and nitrogen within the root zone and plant nutrient uptake. This study was used to

PAGE 29

29 investigat e whether water and nutrients (nitrogen, phosphorus, and potassium ) can be managed within the root zone using drip irrigation. The nutrient concentrations in sugarcane plants and biomass accumulat ion were also monitored The A and E horizons were packed in a lysimeter (3 00 c m by 4 00 c m) up to 80 cm from the soil surface T wo rows of s ugarcane plants were grown i n the lysimeter. At the bottom of the lysimeters were drainage channels for collecting l eachates. The sensors installed in the lysimeters were used to monitor moisture content. Fertilizer application was carried out after sampling soil and tissues for background nutrient concentrations. The fertilizers were surface applied close to the rows o f the sugarcane plants. The drip em i t ters were built within the sugarcane rows at a depth of 5 to 7.5 cm with a spacing of 30.5 cm and irrigation rate of 2271 mL s per hour Soil sampling was conducted using depth increments, 0 15, 15 30, and 30 45 cm. The soil samples were sampled from the irrigated zone (20 c m from the center of cane rows) and from non irrigated zone (50 cm from the center of cane rows). All the soil samples were analyzed for phosphorus nitrate, ammonium, bromide, and water content. Mois ture content was determined by weighing soil before and after oven drying. The moisture content = [(mass of wet soil mass of oven dry soil)/mass of oven dry soil]. Soil samples were extracted for phosphorus using M ehlich 1 solution (dilute HCl and H 2 SO 4 ) a nd analyzed using inductively coupled plasma (ICP). The soil samples were extracted for ammonium and nitrate using potassium ( 2 M KCl ) Ammonium and ni trate were analyzed using rapid flow method

PAGE 30

30 Modeling Water and Nutrient Movement Hydrus simulation envi ronment was used to model water and nutrient movement. This can be calibrated to specific soil characteristics to predict nutrient behavior in mineral soils and their impact o n water quality before conducting field experiments. Calibrated m odels are theref ore very helpful in providing data sets for long periods of time to determine potential impacts on water quality Soil cores were sampled from the site where soil samples for column leaching experiment were obtained. The cores were used to determine bulk d ensity, saturated hydraulic conductivity, and hydraulic functions The linearized sorption coefficient values were also used to study the partitioning of applied phosphorus and ammonium in the lysimeter study. The linearized sorption coefficient values fro m sorption experiments ( potassium chloride ( 0.01M KCl ) calcium chloride ( 0.005M CaCl 2 ) deionized water, simulated Florida rain and fertilizer mixture ) were used for modeling phosphorus movement The nutrient data from column leaching experiment and lysi meter study were used to validate Hydrus 1D. The root mean square error (RMSE) and mean absolute error (MAE) were used to identify th e linearized sorption coefficient value used to model phosphorus movement

PAGE 31

31 CH APTER 2 LITERATURE REVIEW Sugarcane Producti on in United States Sugarcane production can be reported in terms of yield and unit weight of cane or sugar per acre (NASS 2011 ). Yield per acre can be for sugar and for seed (NASS, 2011). Sugarcane production can also be for sugar and for seed planted to produce additional cane vegetatively (NASS, 2011). Table 2 1 show s the total production values for sugar in United States. Florida has consistently registered the total highest cane yield compared to other states (Hawaii, Louisiana, and Texas) where sugar cane is also commercially grown (Table 2 1) Spatial Variability of Soil Properties in Agricultural Fields Tillage implements induce changes in agricultural fields that are not uniform spatially (Casel, 1983 ). S oil of top layer is removed from some locati ons to other locations of the agricultural fields. The A horizon depths are reduced in different parts of the fields and increased in other parts by tillage and leveling ( Brye, 2003 ; Anderson, 1999; Casel, 1983) Soil carbon has been reported to vary withi n the soil orders and depth. This results in differences in nutrient retention potentials within the soil orders and depths (Guo et al. 2006 ; Davidson 1995 ). Land leveling also played a key role in the distribution of total carbon in surface layer (0 15c m) by exposing the sublayers and reducing the surface layers in some areas (Mann et al. 2011 ). The vegetative growth of sugarcane and gross yield were reported to show a positive correlation with organic matter accumulation in sugarcane fields of Lousiana (Johnson and Richard, 2005). The variability in the organic matter leads to variability of sugarcane yields in commercial sugarcane fields (Anderson et al., 1999; Johnson and Richard, 2005). This is attributed

PAGE 32

32 to un even vegetative growth of cane plants a ssociated with organic matter distribution. The retention of nutrients is affected by changes in organic matter because of its important role in reducing leaching of nutrients. Because of higher organic matter concentrations in the surface horizons, higher concentrations of phosphorus have been reported in surface layer (0 15cm) and this was attributed to fertilization (Mann et al., 2011).The immobile nature of phosphorus also expla ins the high phosphorus values in the surface layer (0 15 cm). Sorption of P hosphorus and Ammonium in Soils Sorption of Phosphorus The soil properties including exchangeable calcium, pH, soil organi c matter (Zhou et al., 1997; Axt and Walbridge, 1999), particle size distribution (Axt and Walbridge, 1999 ), amorphous aluminum and iron content (Richardson, 1985; Fox et al., 1990; Lockaby and Walbridge, 1998) determine the amount of phosphorus sorbed by soil. Sorption of phosphorus is affected by oxic and anoxic conditions. For example iron ph osphates are solubilized when iron (III) is reduced to iron (II) under anaerobic conditions. The phosphorus saturation ratio (PSR) is defined as molar ratio of M ehlich phosphorus to sum of M ehlich iron and M ehlich aluminum (Nair and Harris, 2004) The phosphorus saturation ratio (PSR) is al so def ined as molar ratio of oxalate phosphorus to sum of oxalate iron and oxalate aluminum (Chakraborty et al., 2011). (Change point PSR PSR)*( Mehlich1ironn +M ehlich1aluminum ) equals s oil phosphorus storage capacity (SPSC ). Mehlich1iron plus mehlich1aluminum is p roportional to phosphorus saturation ratio (Nair and Harris 2004 ). (Change point phosphorus saturation ratio phosphorus saturation ratio)*(oxalate iron +oxalate aluminum) equals soil phosphorus storage capacity (SPSC) (Chakraborty et al., 2011).

PAGE 33

33 When p hosphorus is added to soil, the amount of phosphorus that remains in soil solution is determined by s orption desorption mechanisms (Frossard et al., 1995). Sorption of phosphorus by Bh of Immokalee soil is higher than sorption by A and E horizons (Li et al ., 1997) This is due to the presence of higher amounts of aluminum/iron oxides and organic matter in Bh horizons (Mansell et al., 1991 ; Li et al., 1997). The Bh horizon act s as a sink for phosphorus under flooded and drained conditions (Villapando and Gra etz, 2001). Differences in sorption desorption capacities of different soil horizons has an impact on the extent and direction of phosphorus transport (He et al., 1999). Sorption of phosphorus on E horizon of a Spodosol was reported to be negligible compar ed to A and Bh horizons. Sorpt i on of phosphorus on Bh was greater than on A horizon (Li et al., 1997). The phosphorus sorption isotherm s that are used to explain the phosphorus sorp tion data are Langmuir, Freundlich and linear. For the Langmuir isotherm a plot of ration of equilibrium concentration to sorbed concentration (C /S ) versus equilibrium concentration ( C ) is first plotted The reciprocal of the slope is the sorption maximum (S max ). The constant, related to bonding energy is calculated using the formula (1/ (K*intercept). The equation, S = (k*S max *C)/ (1+k*C), is used to calculate the model for Langmuir isotherm. Sorption maximum and constant k related to bonding energy are calculated from Langmuir adsorpt ion equation (Nair et al 1998). To obta i n Freundlich isotherm, L nS versus LnC i s first plotted. Where S is sorbed concentration and C is equilibrium solution concentration. The parameter N is the slope and expon ential of the intercept is the F f of the plot. The model plot for Freundlich isotherm is S = K f C N The linear isotherm is the plot of equilibrium concentration (C) versus sorbed concentration (S)

PAGE 34

34 The ionic strength pH, electroconductivity, and nature of cations in supporting electrolytes used in sorption experimen ts affect phosphorus sorption behavior (Antelo et al., 2005; Namasivayam, 2004 ; Pardo et al., 1991 ). Confidence in applicability of sorption coefficients from sorption experiments is assured when electrolytes that best mimics field conditions are used. Tab le 2 2 shows the electrolytes in literature that have been used to conduct sorption experiments by researchers. The rate at which phosphorus react s is rapid for the first phase and is prolonged during the slow phase (Kuo and Lotse, 1974). The fast phase of phosphorus sorption is due to reactions on the surfaces of iron and aluminum (Li et al ., 1997 ). The slow phase is due to the amorphous hydroxides and oxides aluminum and iron that are developed during the experiment (Hsu 1964 ). Phosphate adsorption also was r eported to increase with time ( Kuo and Lotse, 1974). The exchange sites in soils are classified by the two site sorption concept a s type 1 ( sorption assumed to be instantaneous) and type 2 (sorption is time dependent) (Nkedi Kizza et al., 2006). A stu dy on two acid sandy soils showed that two site model describes the kinetics data for times of more than one hour (Fiskell et al., 1979). However the two site model broke down for contact times of less than one hour (Fiskell et al., 1979). Sorption of Amm onium The properties cation exchange capacity, total carbon, and texture were reported to affect ammonium adsorption (Vogeler et al 2011 ; Wang and Alva, 2000 ). Soil horizons with high organic carbon were reported to have higher potential to sorb ammoni um (Wang and Alva, 2000). The type (vermiculite, illite, k aolinite) and amount of clay found in soil determine the amount of ammonium fixed ( Evangelou and Lumbanraja, 2002 ). Cations like calcium and potassium ions have been reported to

PAGE 35

35 compete with ammoniu m for the sorption site ( Jellali et al., 2010) Ammonium sorption has been described by both linear isotherm (Jellali et al., 2010) and Freundlich isotherm s (Vogeler et al 2011 ). Leaching of P hosphorus and N itrogen in Soils Leaching of Phosphorus The pro cesses water transport, phosphorus release a nd sorption, greatly influence phosphorus leachi ng (Djodjic et al., 2004). The phosphorus from agricultural land partly contributes to eutrophication of fresh water bodies (Daniel et al., 1998; Izuno et al., 199 1; Parry, 1998 ). The sandy nature of Spodosols allows movement of phosphorus from A through the eluted E into Bh horizon (Weaver et al., 1988; Guertal et al., 1991 ). The anaerobic conditions created above the Bh horizon through fluctuating the water table changes the transport and sorption of phosphorus (Nair et al., 1999). For s oils with shallow water table, phosphorus can be transported horizontally above the Bh to ditches by drainage water (Villapando and Graetz, 2001). The trend for leaching of phosphor us was reported to be E > A > Bh for a Spodosol sampled from a citrus grove near Fort Pierce, St.Lucie County. This was attributed to very low concentratio ns of organic matter, clay, aluminum and iron ( Li et al 1997) in E and A horizons C olumn leaching studies have revealed the principles that explain the behavior of leaching and sorption of phosphorus Unlike conservative tracers (bromide, chloride), preferential flow is very important in movement of adsorbed solutes (Everts et al., 1989 ; Li and Ghodrat i, 1994 ). Tracers are very important in characterizing solute transport (Devine and McDonnell, 2005; Sumner et al., 1996 ; Lawrence and Richard, 2008 ) because they are non reactive. Tracer displacement by water is influenced by soil texture and rate of wate r application (Ghuman and Prihar, 1980). The dispersion

PAGE 36

36 coefficient which is velocity dependent also increases with the drainage pore sizes (Ghuman and Prihar, 1980). Dispersion coefficient and pore water velocities followed the order loamy sand > sandy lo am > silt loam (Ghuman and Prihar, 1980). A portion of the salts can be rapidly lost in soils that are well structured during heavy rains while the rest of it remains in immobile water of the soil mass (Tyler and Thomas, 1981). Phosphorus leaching in soil columns increases with the rate at wh ich phosphorus is applied and leaching intensity (Logan and Mclean, 1973). Leaching of Nitrogen The concentrations of nitrogen in leachates can exceed the recommended nitrate values established by U nited S tates E nviro nmental P rotection A gency (USEPA) for drinking water ( Muchovej et al., 2004). Although water table depth did not significantly affect ammonium release in Histosols, nitrogen release increases with increase in water table depth (Martin et al., 1997). Nitrif ication (conversion of ammonium to nitrate ) leads to rapid decrease in concentrations of ammonium nitrogen in soil solution when fertilization is coupled with irrigation (Mansell et al 1980). Recommended value for nitrogen fertilizers split applications to sandy soil s of south Florida is 70 kgN ha 1 Adhering to this limit helps in attaining EPA standards for drinking water (Muchovej et al., 2004). Nitrate was m i crobially denitrified at a velocity of 0.11cm hr 1 (21 % denitirified) (Corey et al 1967). H o wever there was no immobilization and denitrif ication at a velocity of 1.32 cm hr 1 in a study with long acrylic plastic columns containing soil (Corey et al., 1967). In a system with unsaturated flow, nitrate gained through minerali zation and nitrificati on exceeded the amount lost through immobilization and

PAGE 37

37 denitrification (Corey et al., 1967). This was attributed to aerobic conditions that are maintained through openings on the sides of the columns. The charge characteristics of soil were reported to in fluence the retardation in leaching and partitioning between soluble phases and adsorbed nitrate and chloride ( Sumner et al., 1996). Nitrate moves at slower velocity at low pH and the flow velocity increases when pH is increased (Qafoku et al., 2000). Reta rdation factor, R, = 1+ (K D D is linear ized content. Nitrate movement in a column was reported to be directly proportional to large pore spaces and inversely proportional to small pore spaces of soil (Bates and Tisdale, 1957). Flow velocity was reported to influence movement of nitrate and its denitrification moving through soils (Corey et al., 1967). The rate of flow in a porous media packed in a column affects the amount of ammonium a dsorbed. Ammonium adsorption increases with the amount of contact time in the column (Jellali et al 2010). Ammonium decreased with depth in a column study and this was attributed to microbial mediated transformations of ammonium to nitrate (Ardakani et a l., 1974). Modeling Water, Bromide P hosphorus and N itrogen Movement Hydrus simulation environment is used to model one and two dimensional movement of water, tracer (bromide and chloride ), phosphorus and nitrogen in a variably saturated media (Hassan et al 2008 ). For solute transport, sorption coefficients (K D ), ki netic parameters (k2 and F) of phosphorus and ammonium, phosphorus and nitrate fertilizer application rates (Simunek et al 199 9 ) are used to calibrate Hydrus 1D and 2D. quation (Richards, 1931) is used for modeling water flow. The sink term which is defined as the volume of water removed per unit time from a unit volume of soil due to plant uptake is embedded in Richards

PAGE 38

38 equation. The phosphorus and nitrogen transport are modeled by the conv ection dispersion equations (Renduo, 2000). The root mean square error (RMSE) (Kwon et al., 2010) and mean absolute error (MAE) (Kandelous, 2010) are used to compare simulated values and experimental data

PAGE 39

39 Table 2 1. Sugarcane p roducti on for sugar in million tons in States of United States States/Year 2010 1 2009 1 2007 2 2006 2 Florida 14 13.3 13.5 13.7 Louisiana 11.3 12.6 11.9 11.1 Hawaii 1.2 1.3 1.5 1.6 Texas 1.7 1.3 1.4 1.6 Total 28.2 28.5 28.3 28.0 Source: 1. NASS, 2011. 2. Sou rce: NASS, 2009 Table 2 2 Electrolytes in l iterature u sed by r esearchers Electrolytes Source 0.01M KCl Rhue et al., 2006 0.02M KCl He et al ,1999 0.05M KCl Zhou and Li, 2001 0.01M CaCl 2 Dou et al., 2009 0.1 M CaCl 2 Rubio et al., 2008 Deionized water Drizo et al., 2002

PAGE 40

40 CHAPTER 3 SPATIAL DISTRIBUTION OF SOIL CHARACTERISTICS IN SUGARCANE FIELDS Background Management practices like tillage and leveling sugarcane fields reduces A horizon depth in some locations and increases the depths in other locatio ns. This also results in Ap horizon which is not a typical A horizon called a plow layer The A horizon is the surface horizon with higher organic matter than E horizon since most of the vegetation gro w with in A horizon (Johnson and Richard, 2005). Managem ent of phosphorus within the A horizon is very crucial since the majority of sugarcane roots grow within 30 cm from the soil surface (Pitts et al.,1993). The A horizon has a higher organic matter than the eluted E horizon that acts as a medium through which phosphorus moves to the water table. The refore the depth of A horizon determines how much phosphorus a soil will retain prior to lea ch ing through the E horizon and out of the root zone. The un even distribution of A horizon depths in th e sugarcane fields, leads to un even d istribution of plant available phosphorus in the root zone Rice yield reduction was observed when surface layer of soil was cut during land leveling and the authors attributed the r eduction to decrease in native phosphorus (Walker et al ., 2003). Amorphous aluminum iron content (Richardson, 198 5; Lockaby and Walbridge, 1998; Ige et al 2005), and soil organic matter (Zhou et al 1997 ; Axt and Walbridge, 1999; Ige et al., 2 005) are important in studying phosphorus sorption. O rganic matt er associates with iron, aluminum, and calcium that enhance the attraction of negatively charged phosphate ion on the exchange site. A luminum and iron phosphorus solubulizes at pH less than 3. Precipitation of phosphorus with iron and aluminum occurs at pH between 3 5.2. Phosphorus is most available for plant growth at pH

PAGE 41

41 bet ween 5.5 to 7.0. Calcium bound phosphorus is insoluble at pH greater than 7. Soil pH is therefore very important in determining the reactions of ir on, aluminum, and calcium with phospho rus According to Chakraborty et al 2011, sum of oxalate extractable aluminum and oxalate extractable iron (acid ammonium oxalate extractable, McKeague and Day, 1966) is used to calculate phosphorus saturation ratio (PSR). PSR = oxa late phosphorus/ (oxal ate aluminum + oxalate iron ). Soil phosphorus storage capacity (SPSC) was calculated using the equation, (0.15 PSR)*(M1Al+M1Fe) (Nair and Harris, 2004). The value 0.15, was reported to correspond to the critical phosphorus soluti on concentration value of 0.1mg L 1 (Nair and Harris, 2004). The PSR can also be calculated using the equation, M1P/ (M1Al+M1Fe) (Nair and Harris, 2004). The soil phosphorus storage capacity estimates the amount of phosphorus a soil can retain before exceeding a threshold soil equili brium concentration (Chakraborty et al., 2011). Depending on the change points, soil phosphorus storage capacity c an be calculated using oxalate phosphorus (Ox.P) Mehlich 1phosphorus (M1P) Mehlich 3 phosphorus (M3P) and their corresponding aluminum and i ron contents. Examples of equations are, SPSC Ox = (Change point PSROx soil PSROx)*(Ox Fe+Ox.Al)*31, SPSC M1 = (Change point PSR M1 soil PSR M1 )*(M1 Fe+M1 Al)*31*1.8, and SPSC M3 = (Change point PSR M3 soil PSR M3 )*(M3 Fe+M3 Al)*31*1.3 (Chakraborty e t al., 2011) A thorough understanding of the distribution of soil characteristics in sugarcane fields is important for simulation models that assume homogeneity in terms of soil characteristics Delineating agricultural fields according to soil characteri stics helps researchers to monitor nutrient movement to lower horizons In sugarcane fields,

PAGE 42

42 un even distribution of organic matter accumulation affects vegetative growth of sugarcane plants (Johnson and Richard, 2005). This affects yield of sugarcane and therefore sugarcane fie l ds ca n be divided into regions based on the level of production. Un even distribution of native phosphorus due to reduction of surface layer through land leveling was also reported to decrease rice yield (Walker et al 2003 ). In t his study, the chemical characteristics determined were soil pH, total carbon, t ota l phosphorus, oxalate iron oxalate aluminum oxalate phosphorus and exchangeable calcium. The physical characteristic determined was A horizon depth. The h ypothesi s of thi s study was ; (i) s oil characteristics for t wo dominant sandy soils used for sugarcane production in south Florida will vary spatialy and with depth and this will lead to different patterns of phosphorus accumulations The objectives of the study were; (i ) t o identify the pattern of distribution of soil characteristics (total phosphorus total carbon, pH oxalate iron, oxalate aluminum, oxalate phosphorus exchangeable calcium, and A horizon depth ) ; and (ii ) g roup the soil property values according to magnit ude of values (pH, total phosphorus and total carbon) and select soil samples for characterizing sorption of phosphorus and ammonium. Materials and Methods Soil S eries The soil series, Margate ( Sandy, s iliceous, hyperthermic Mollic Psammaquents) and Immok alee ( Sandy, s iliceous, hyperthermic Arenic Alaquods), are used for comm ercial sugarcane production in s outhwest Florida. Two sugarcane fields were used to identify the differences in the Margate and Immokalee soil series in terms of soil properties. The a rea of each sugarcane field was 30 acres. Soil pH (1:2 H 2 O), total

PAGE 43

43 phosphorus total carbon, oxalate extractable iron oxalate extractable aluminum oxalate extractable phosphorus exchangeable calcium, and A horizon depth were measured. The soil horizons of Margate are A E, and Bw. Margate fine sand is derived from marine sediments characterized by underlying limestone (Obreza et al., 1998). The soil horizons of Immokalee are A, E, and Bh. Margate (17.8 %) and Immokalee (11.3 %) contribute 30 % of the m ineral soils us ed for sugarcane production in s outhwest Florida ( U.S Department of Agriculture, Natural Resources Conservation Service, 1995 ). Sampling and Measurement of Soil Properties The commercial sugarcane production and processing company, US Suga r Inc., provided the two sugar cane fields with the two soil series. Out of the 100 sample positions 80 were sampled on a grid and 20 were rando ml y sampled. According to geostati stics analysis, 100 samples are enough for spatial evaluation. The sampling de signs are shown in Figures 3 1 and 3 2. A n a uger 3 cm in diameter was used to extract soil In the field, soil was displayed on the gutters to identify horizon designations. The surface A horizon had a darker color than the underlying E horizon. After ide ntifying the extent of A horizon, the depth was measured using a ruler. Soil was then divided into 0 30, 30 60, and 60 90 cm depths. Soil samples were analyzed for total phosphorus pH, oxalate extractable a luminum, oxalate extractable iron, oxalate extrac table phosphorus exchangeable calcium, and total carbon after air drying. The soil dryer room was set at 45 0 C and soil was kept in the room until when it dried Air dried soil can be sieved easily through a 2 mm sieve. Soil was then sieved through a 2 mm sieve.

PAGE 44

44 Soil pH was measured in a slurry with 1:2 soil to water solution ratio using a glass electrode ( Sparks, 1996 ). Air dried s oil samples (20 mL ) were scooped and poured into 90 mL plastic cups. Deionized water (40 mL s ) were added to the plastic cups a n d stirred with a glass rod. Soil plus deionized water were left to stand for 30 minutes but not more than 2 hours. The soil pH was measured while stirring the soil using a standardized pH meter ( model : AR15; manufacturer: Fisher Scientific ). The pH standar ds, 4, 7, and 10 were used to standardize the pH meter. Total carbon was determined by combustion method using elemental analyzer, Carbo Erba NA 2500 instrument (Model: NA 2500; manufacturer: CE instruments, Italy) ( Sparks, 1996) Air dried s oil sieved thr ough 2 mm was ground using a grinder in to powder form (soil particles that passed through a 0.5 mm sieve). Finely ground soil (0.05 g) was weighed and wrapped into aluminum foil and fed into the carbon nitrogen analyzer. The sample is passed through heli um stream set at 1000 o C followed by oxidation at 1800 o C through instantaneous combustion. The values of total carbon were obtained using a calibration curve derived from total carbon values of standard soil samples Different masses, 10 mg, 20 mg, and 40 m g, of standard whose total carbon content is known were used to make a calibration curve. Total phosphorus was determined by digestion method ( Bowman, 1988 ). Soil was first ground into soil particles that can pass through 0.5 mm sieve. Air dried ground so il (2 g) was added into dig estion tubes. Concentrated sulf uric acid (2 mL ) was added and the tubes inserted into a block heated at 340 o C for 1 hr. The d igestion tubes were then removed and allowed to cool for 5 minutes. Thirty p ercent hydrogen peroxide (0. 5 mL ) was then added and digested in the same block for an additional 15 minutes. This was

PAGE 45

45 followed by cooling. Additions of hydrogen peroxide, heating and cooling steps were repeated until solutions were clear. Hydrogen peroxide was then evaporated by he ating in blocks for 4 5 minutes. After cooling total phosphorus was measured using a spectrophotometer ( model : DR/4000U; manufacturer: HACH company ) Standard solutions of known phosphorus values (4, 8, 12, and 16 ppm) were added to known volumes of phosph ate reagent to form a characteristic blue color. Solutions put into vials were fed in to fittings in the spectrophotometer ( model : DR/4000U; manufacturer : HACH company ) to obtain a calibration curve (absorbance verses known concentrations) The calibration curve was used to obtain total phosphorus values corresponding to the absorbance of the samples. Samples were also added to phosphate reagent before feeding them in the spectrophotometer ( model : DR/4000U; manufacturer: HACH company ) Oxalate extractable i ron aluminum and phosphorus w ere determined by extraction with 0.1M oxalic acid + 0.175M ammonium oxalate (pH=3) solution (Mc Keague and Day, 1966). The extraction solution was prepared by combining 700 mL s of 0.2 M ammonium oxalate and 53 5 mL s of 0.2 M o xalic acid followed by adjusting pH to 3 by adding aliquots of either reagent ( McKeague and Day, 1966). One gram air dried soil was weighed into a 50 mL centrifuge tube A repipet dispensing bottle was used to add 25 mL of the acid oxalate solution (ammon ium oxalate and oxalic acid) and the tube capped tightly. The centrifuge tubes were wrapped with aluminu m foil to keep dark and prevent evolution of oxalate solution. The tubes were then placed on a reciprocal shaker ( manufacturer : Eberbach ) and shaken at low speed for four hours. The soil solution were centrifuged at 2000 rpm using a centrifuge ( model : Sorvall ST

PAGE 46

46 16R and manufacturer : Thermo Scientific ) for 20 minu tes and decanted into a container Inductively coupled plasma ( model : Optima 700 DV; manufact urer : Perkin Elmer ) was used to analyze solu tions for oxalate extractable iron aluminum, and phosphorus Exchangeable calcium was determined using a leaching tube method (soil survey staff, 1996) Five grams were weighed into 50 mL centrifuge tubes. A vo lume, 30 mL s, of 0.2 M NH 4 Cl were added, capped and shaken at high speed 60 rpm for 5 minutes using a reciprocating shaker ( manufacturer: Eberbach ) Soil solutions were centrifuged at about 4000 rpm and decanted into 250 mL volumetric flasks. Addition s of 30 mL s of 0.2 M NH 4 Cl, capping, shaking, centrifuging, and decanting were repeated three more times. The supernatants were combined and made up to the full volume (250 mL s) with 0.2 M NH 4 Cl. The solutions were analyzed for exchangeable c a lcium using induct ively coupled plasma ( model: Optima 700DV and manufacturer : Perkin Elmer ) The units of exchangeable calcium (Ca) was centimoles of c ation charge per kilogram (cmol kg 1 ), the formular used was, exchangeable Ca = (M n+ xVxn)/(WxA) where M n+ is concentration of cation in extract(ppm), V olume V is 250 mL s n is valence 2 W is weight of soil ( 5 g), and A is atomic weight of calcium (40) Results and Discussion The frequencies of the soil property values in sugarcane fields and the shapes of the histograms are s hown in figures A 1 to A 44 The A horizon maximum values for Immokalee and Margate fields were 70 cm and 80 cm respectively (Ta bles 3 1 and 3 2 ). The minimum values of A horizon for Immokalee and Margate fields were 13 cm and 10 cm respectively (Ta ble 3 1 and 3 2 ) and the range s (maximum minus minimum) values for Immokalee and Margate fields were 57 cm and 60 cm respectively. The variability in

PAGE 47

47 A horizon depth for both sugarcane fields is due to land leveling that is carried out to attain a relatively flat surface and tillage implements that are used for management practices ( Casel, 1983; Brye, 2006). Brye, ( 2006 ) found a decrease in organic matter, total carbon, soil extractable phosphorus and increase in soil pH in the top 10 cm after land leveling Th e de crease in soil extractable phosphorus organic matter, and total carbon is probably due to reduction in A horizon depth. The A horizon depths were normally distributed with mean of 36.4 cm for Im mokalee and 25.5 cm for Margate approximating median of 34.5 cm for Immokalee and 22 cm for Margate for both fields (Tables 3 1 and 3 2 ). The A horizon depth of 18 cm for Immokalee soil was reported by Hyde and Ford, 1989 near water well sites. Stutter et al., 2008 identified a positive correlation between O horizon depth and carbon. The authors identified differences in O horizon depth for different soil series. D ifferences in total carbon mean values (0 30 cm) for both soil series are due to differences in decompos ition of organic matter and land leveling (Tables 3 1 and 3 2) According to study by Brye ( 2006 ) land leveling significantly reduced organic matter a n d total carbon in the upper 10 cm. The author compared total carbon values from the pre and post leveled fields and attributed the reduction in values to the exposure of subsoil (E horizon) which is non retentive Total carbon values (0 30 cm ) for the two fields were normally distributed (Tables 3 1 and 3 2) Mean total carbon for 0 30cm (15.3 g k g 1 for Immo kalee and 13.4 g k g 1 for Margate) approximate d median values (15 g k g 1 for Immokalee and 12.7 g k g 1 for Margate). Mean to tal carbon value (0 15 cm), 12 g k g 1 for a sugarcane field with mostly Basinger sand, and small areas of Wabasso fine sand (sandy,siliceous, hyperthermic Albic Alaquods) and Mar gate fine

PAGE 48

48 sand was reported by Obreza et al.,( 1998 ) The matched t test was used to identify the significant difference at 95 % confidence interval in magnitude of values using with or without equal variance procedure ( Brye, 2006 ). For both fields, total c arbon values (0 30 c m) were significantly greater than total carbon (30 60 cm) (P < 0.01 for Immokalee and Margate ). The total carbon values at 0 30 cm for both fie lds were significantly greater than total carbon at 60 90 cm ( P < 0.01 for Immokalee and Mar gate ) This is due to the presence of A horizon at the 0 30 cm depth and E horizon in 30 60 cm and 60 90 cm depth s The similarity in total phosphorus at 0 30 cm for the two sugarcane fields (119.3 g g 1 for Immokalee and 125.1 g g 1 for Margate) is att ributed to the same phosphorus application rate used by the farmers. Farmers in s outhwest Florida use the typical application rates of 50 k g P 2 O 5 ha 1 200 k g K 2 O ha 1 and 200 k g N ha 1 ( Obreza et al., 1998 ) The high mean value of total phosphorus in the 0 30 cm compared to 30 60 cm and 60 90 cm (Table 3 1 and Table 3 2 ) are attributed to phosphorus application on the soil surface. Phosphorus is applied on the soil surface to meet plant nutrient requirements and is considered to be immobile. A study conduct ed by Mann et al 2011 reported higher values of phosphorus in the depth of 0 to 15 cm. The authors explained that the high values in the surface layer (0 15 c m) were due to phosphorus fertilization. Total phosphorus values at the 0 30 cm depth for the two fields were no rmally distributed. Mean total phosphorus values (119.3 g g 1 for Immokalee and 125.1 g g 1 for Margate) approximate median values (112.8 g g 1 for Immokalee and 117.2 g g 1 for Margate). Total phosphorus values at 0 30 cm for both fie l ds were significantly greater than values at 30 60 cm (P < 0.01 for Immokalee and Margate field). T otal phosphorus values at 0 30 cm were significantly greater than values at

PAGE 49

49 60 90 cm (P < 0.01 for Immok alee and P < 0.01 for Margate ). Soil phosphorus spatia lly varied in pastures and authors identified normal and lognormal distribution patterns in the data for soil samples collected from 0 15 cm depth (Daniels et al. 2001 ). The pH values at 0 30 cm for the two fields were normally distributed ( Table 3 1 and 3 2 ) Unlike the I mmokalee field where pH values at 0 30 cm were significantly greater than values at 30 60 cm (t statistic (8.93)>t critical (1.65) ), the values were not significantly greater (t statistic ( 3.0 5)
PAGE 50

50 ( 28.3 cmol c kg 1 for 0 30 cm and 19.1 cmol c kg 1 for 30 60 cm ). Mean exchangeable calcium values ( 53.7 cmol c kg 1 for 0 30 cm and 29.9 cmol c kg 1 for 30 60 cm) for Margate field do not approximate median values ( 46.3 cmol c kg 1 for 0 30 cm and 15.4 cmol c kg 1 for 30 60 cm). Exchangeable calcium values at 0 30 cm for both fields were significantly greater than values at 30 60 cm ( P < 0.01 for Immokalee and Margate ). Exchangeable c alcium values at 0 30 cm were not significantly different for Immokalee field ( P = 0.06 ) and significantly greater than values at 60 90 cm depth for Margate field (P < 0.01 ). Because E horizon was explored for the depth of 30 60 cm v alues of total carbon at 30 60 cm were lower than the values at 0 30cm for both fields (Tables 3 1 and 3 2 ) Mean total carbon values at 0 30 cm were 1 5 .3 g k g 1 soil and 1 3 .4 g k g 1 soil for Immokalee and Margate respectively (Tables 3 1 and 3 2 ). The mean total carbon values at 30 60 cm were 8 .5 g k g 1 soil and 3 .7 g k g 1 soil for Immokalee and Margate soil respectively. The E horizon typically has low values of total carbon or no carbon values ( Li et al., 1997 ) Values of total phosphorus, oxalate aluminum exchangeable cal cium, and iron at 30 60 cm were lower than values at 0 30 cm Since values of total carbon at 30 60 cm were lower than the 0 30 cm depth complexation of organic matter with phosphorus, aluminum, and iron was lower (Brye, 2006 ) Th ese soil concentrations explain the low values of total phosphorus, aluminum, and iron for the depth of 30 60 cm The total carbon at 30 60 cm for both fields were not signi ficantly different from values at 60 90 cm ( P = 0.08 for Immokalee and P = 0.25 for Margate). For both fie lds, E horizon was most explored for the depth of 60 90 cm. This explains the

PAGE 51

51 similarity in distribution of total carbon in both fields. Unlike in Immokalee soil where pH values at 30 60 cm depth were significantly greater than values at 60 90cm (t statist ic (2.68) >t critical (1.65)), the values were not significantly different for Margate (t statistic ( 1.89) < t critical (1.65)). Sampling portions of Bh in some positions o f sugarcane fields must have le d to the reductions of measured pH values at 60 90 c m depth. Total phosphorus values at 30 60 cm were significantly greater than values at 60 90 cm (t statistic (2.65) >t critical (1.65)) for Immokalee due to portions of A horiz on sampled and probably native phosphorus but were not significantly different f or Margate (t statistic (1.41) < t critical (1.65)). Oxalate extractable aluminum (30 60 cm) values were significantly lower than values at the 60 90 cm depth for Immokalee (P = 0.005 ) and Margate (P =<0.001 ). This is probably due to sampled portions of Bh horizon in Immokalee soil and thin layers of Bt horizon in Margate soil. Oxalate iron values at 30 60 cm were also significantly lower than values at 60 90 cm depth (P=<0.001 for Immokalee and greater for Margate ( P=0.042 ). The decrease in oxalate iron a n d aluminum with depth is due to the distribution of organic matter that follows the same trend. Organic matter surfaces (negatively charged) will react wi th positively charged metals (aluminum and iron ). Exchangeable calcium values at 30 60 cm were signifi cantly lower than values at 60 90 cm depth (P = 0.02 ) for Immokalee and were not significantly different for Margate field (P = 0.053). The maximum value of total carbon for dept h of 60 90 cm is higher for Immokalee soil (77.4 g k g 1 ) than for Margate soi l (19 g k g 1 ) (Table 3 1 and Table 3 2 ) This is attributed to the Bh horizon that is sampled at the depth increment of 60 90 cm Unlike Bw horizon which is a transition horizon Bh is the zone of organic matter accumulation.

PAGE 52

52 The pH values at 60 90 cm are higher for the field with Margate soil (7.9) than with Immokalee soil (7.1) (Tables 3 1 and 3 2 ) This is due to the underlying lime stone in the field with Margate soil. The mean aluminum values for Immokalee soil followed the order 60 90 cm (31 2.2 mg k g 1 ) > 0 30 cm (285 mg k g 1 ) > 30 60 cm (209.1 mg k g 1 ) (Table 3 1 ) The Bh horizon was sampled at a depth of 60 90 cm for most of the locations. The Bh horizon is the horizon in the Immokalee soil profile that accumulates organic matter, aluminum, and with or without iron. This explains the trend in the data. The mean values for oxalate iron for Immokalee soil followed the trend 0 30 cm > 30 60 cm > 60 90 cm (Table 3 1 ). The pH values for the t wo depths (30 60 cm and 60 90 cm) from the two sugarcane field s were normally distributed. The aluminum and iron maximum (1270 mg k g 1 for aluminum and 546.5 mg k g 1 for iron ) values for Immokalee soil at 60 90 cm were due to Bh horizon influence (Table 3 1 ). Percent organic matter for surface, subsurface, and deep h orizons decreased with depth (Iqbal et al., 2005). In a study that involved measurement of organic matter content at 0 15 cm 15 30 cm 30 45 cm and 45 60 cm depths in a heterogeneous citrus grove, values decreased with depth ( Mann et al., 2011). Total p hosphorus decreased with depth and the authors attributed the decr ease to the immobile nature of phosphorus (Walker et al., 2003). Mallarino and Borges 2006 ide ntified a decreasing soil test phosphorus values with depth in fields with no till and where ch isel disk tillage was adopted. The athours explained that the pattern in vert ical distribution of phosphorus was due to differences i n crop uptake, nutrient native phosphorus nutrient recycling and mobility in soil. For the Margate soil, correla tion coeff icients (R) at 95 % confidence interval

PAGE 53

53 ( = 0.05) of total phosphorus to the measured soil properties for 0 30 cm follows the trend, total carbon (0.48) > oxalate aluminum (0.32) > oxalate iron (0.21) > exchangeable calcium (0.1) > pH (0.022). For the 30 60 cm depth R values follows the trend total carbon (0.51) > oxalate aluminum (0.18) > pH (0.17) > exchangeable calcium (0.06) > oxalate iron (0.01). The correlation values for 60 90 cm depth followed the same trend as 0 30 cm depth total carbon (0.65) > oxalate extractable aluminum (0.18) > e xchangeable calcium (0.17) > oxalate extractable iron (0.04) > pH (0.03). The correlation values show that total carbon plays an important role in determining the distribution of total phosphorus in sugarcane fields with Margate soil. This is attributed to the role of organic c arbon in nutrient cycling. For the Immokalee soil, correlation values for 0 3 0 cm depth followed the trend pH (0.47) > exchangeable calcium (0.3) > total carbon (0.28) > oxalate extractable iron (0.23) > oxalate aluminum (0.17). The c orrelation values for 30 60 cm were oxalate extractable aluminum (0.30) > total carbon (0.28) > oxalate extractable iron (0.17) > exchangeable calcium (0.1) > pH (0.08) The order of correlation values for 60 90 cm depth was total carbon (0.38) > exchangea ble calcium (0.2) > oxalate extractable aluminum (0.22) > pH (0.14) > oxalate extractable iron (0.09). Unlike in Margate soil where total carbon de termined distribution of total phosphorus at all the depth increments, pH and total carbon determined the dis tribution at 0 30 cm depth oxalate extractable aluminum and total carbon at 30 60 cm and 60 90 cm depths Although other measured soil characteristics determined phosphorus distribution at different depths, total carbon remains a key factor in Immokalee soil. Roel and Plant, 2004 identified a positive cor relation between soil test phosphorus and

PAGE 54

54 organic matter and negative correlation between pH and organic matter in rice fields. In this study, the authors sampled soil from 0 30 cm depth using 30 m grid. When 8 cm increments were used to di vide the sugarcane fields according to A horizon depths, the percentage distribution in Margate field followed the trend, 19 27 cm (44 % ) > 10 18 cm (22 %) and 28 35 cm (22 %) > 37 45 cm (5.5%) > 50 58 cm (3.3 % ) and 60 80 cm (3.3 %) The trend, 22 30 cm (33 %) > 31 39 cm (28 % ) > 40 48 cm (17 % ) > 49 57 cm (8 % ) and 60 70 cm (8 %) was calculated for Immokalee field. For the Margate field, 88 % of A horizon depth range from 10 cm to 35 cm. Seventy eight percent of th e A horizon depth range from 22 cm to 48 cm for Immokalee field. Conc luding R emarks The measured values for the soil properties from two sugarcane fields revealed that soil characteristics vary spatia lly and with depths. Land leveling results in un e ven di stribution of A horizon depths in sugarcane fields. This will affect the availability of nutrients within the root zone if the A horizon depth is reduced to less than 30cm. The E horizon is categorized as a horizon that exh ibits minimal interaction with ph osphorus, potassium and ammonium. Characteristics like oxalate extractable iron and oxalate extractable aluminum tremendously reduced for 30 60 and 60 90 cm depth s both fields w here E horizon was most explored. Unlike in Margate soil, accumulation of orga nic matter in the Bh horizon has been reflected in maximum total carbon values for 60 90 cm depth for Immokalee soil. The underlying limestone in Margate soil is reflected in high pH values for 60 90 cm depth Conducting a study on distribution of soil pro perties in sugarcane fields is essent ial before studying nutrient movement The vertical and lateral distribution of total

PAGE 55

55 phosphorus was attributed to differences in nutrient cycling, depth of A horizon, and native phosphorus Whe n correlation coefficient s between total phosphorus and measured soil characteristics were compared, total carbon was the most important characteristic that determ ined the distribution of total phosphorus in both fields. This emphasizes the importance of A horizon depth in sugarc a ne fields. Exposure of the elut ed E horizon will result in leaching of applied nutrients and less avail able nutrients for plant uptake. This study has re vealed that modelers of nutrient movement in sugarcane fields should not assume that soil characteristi cs are uniformly distributed. Farmers should consider demarcating sugarcane fields acco rding soil characteristics to help farmers monitor sugarcane yield as function of s oil characteristics The hypothesis, s oil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary spatialy and with depth and this will lead to different patterns of phosphorus accumulations, has been proven after observing variability i n soil characteristics. After using two sugarcane fields wi th Immokalee and Margate soil and correlating soil characteristics with total phosphorus, total carbon determined most the accumulation of phosphorus.

PAGE 56

56 Table 3 1. Exploratory statistics of soil properties for Immokalee s oil Soil Property Mean Median StDev Min. Max. Skew. Kurtosi s Total carbon 1(g kg 1 ) 15.3 15.0 4.2 8.3 25.1 0.5 0.44 Total carbon 2 (g kg 1 ) 8.5 6.6 5 7 0.0 27.6 1.3 1.3 Total carbon 3 (g kg 1 ) 11.0 6.1 12 1.2 77.4 2.6 9.5 pH 1 7.0 6.9 0.44 6.3 8.2 0.66 0.16 pH 2 7.2 7.3 0.47 5.9 8.3 0.18 0.08 pH 3 7.1 7.0 0.52 5.9 8.4 0.5 0.10 Total P1(g g 1 ) 119.3 112.8 44.1 33.4 348.2 1.9 7.1 Total P2 (g g 1 ) 63.5 53 34.2 7.1 170.6 0.96 0.47 Total P3(g g 1 ) 51.0 40.7 32.5 3.6 172.3 1.4 2.3 Oxalate Al 1(mg kg 1 ) 285 285 59.7 195.2 484.8 0.86 0.34 Oxalate Al 2(mg kg 1 ) 209.1 182.3 103 74.3 870.8 3.4 18.3 Oxalate Al 3(mg kg 1 ) 312.2 203.8 262.8 69.2 1270 2.1 3.7 Oxalate Fe 1(mg kg 1 ) 237.6 224.6 48.6 155 377.8 0.78 0.14 Oxalate Fe 2(mg kg 1 ) 178.4 146.2 113.9 79.7 104.5 5.00 35.3 Oxalate F e 3(mg kg 1 ) 129.5 95.8 111.7 18.1 546.5 1.9 3.5 Oxalate P 1(mg kg 1 ) 32.70 31.0 20.3 2 84.7 0.46 0.32 PSR 0.059 0.059 0.034 0.004 0.123 0.118 1.03 Exchangeable Ca1(mg kg 1 ) 29.8 28.3 11.1 11.8 89.8 2.0 8.1 Exchangeable Ca2(mg kg 1 ) 20.8 19.1 9 6.0 4 7.2 0.8 0.2 Exchangeable Ca3(mg kg 1 ) 25.6 18.6 19.6 5.8 104.7 1.6 2.7 A horizon (cm) 36.4 34.5 11.9 13 70 0.85 0.35 1= 0 30 cm, 2= 30 60 cm, 3= 60 90 cm, StDev=Standard deviation, Min. =Minimum, Max.=Maximum, Skew. =Skewness, Ca = Calcium, PSR= Phospho rus saturation ration Al = Aluminum, Fe = Iron, P = Phosphorus

PAGE 57

57 Table 3 2. Exploratory statistics of soil p roperties for Margate soil Soil Property Mean Median StDev. Min. Max. Skew. Kurtosis Total carbon 1 (g kg 1 ) 13.4 12.7 6 4 1.9 35.5 0.40 0.4 Tota l carbon 2 (g kg 1 ) 3.7 2.0 5 4 0.6 38.8 4.60 25.2 Total carbon 3 (g kg 1 ) 3.0 1.7 3 5 0.6 19.0 2.90 9.0 pH 1 7.4 7.3 0.56 6.1 9.0 0.5 0.4 pH 2 7.7 7.5 0.68 6.0 9.3 0.5 0.2 pH 3 7.9 7.8 0.75 6.3 9.5 0.3 0.8 Total P1 (g g 1 ) 125.1 117.2 59.1 15.5 36 3.9 1.0 2.0 Total P2 (g g 1 ) 41.9 33.3 34.5 4.4 225.5 2.9 10.7 Total P3 (g g 1 ) 32.8 26.8 29.5 3.4 235.9 4.0 24.4 Oxalate Al1 (mg kg 1 ) 266.6 276.8 100.1 57.6 490.8 0.02 0.62 Oxalate Al 2 (mg kg 1 ) 110.5 88 78.3 36.6 462.3 2.57 8.1 Oxalate Al 3 (m g kg 1 ) 136.7 108.2 78.1 57.5 532 2.36 7.6 Oxalate Fe1(mg kg 1 ) 385.5 376 185.5 28.1 931.8 0.40 0.22 Oxalate Fe 2 (mg kg 1 ) 127.6 103 94.2 38.6 498.3 1.9 3.6 Oxalate Fe 3 (mg kg 1 ) 148.4 110 96 44 470.3 1.43 1.48 Oxalate P 1 (mg kg 1 ) 59.1 36 51.8 2.1 211 1.32 1.24 Oxalate P 3 (mg kg 1 ) 43.8 21.8 47.1 3.3 231.6 1.7 2.4 PSR 1 0.064 0.049 0.05 0.003 0.204 0.939 0.272 PSR 3 0.148 0.101 0.15 0.023 0.808 2.72 8.2 Exchangeable Ca 1(mg kg 1 ) 53.7 46.3 31.5 6.8 161.8 1.5 2.3 Exchangeable Ca 2 (mg kg 1 ) 29 .9 15.4 37.8 29.9 204.4 3.0 9.4 Exchangeable Ca 3 (mg kg 1 ) 33.1 12.9 44.3 1.30 199.6 2.1 4.1 A horizon (cm) 25.5 22 12.3 10 80 2.2 6.2 1= 0 30 cm, 2= 30 60 cm, 3= 60 90 cm, StDev=Standard deviation, Min. =Minimum, Max. =Maximum, Skew. =Skewness, Ca = C alcium, PSR = Phosphorus saturation ration Al = Aluminum, Fe = Iron, P = Phosphorus

PAGE 58

58 Figure 3 1 Sampling design from a sugarcane field with Margate soil series. Data source: www.fgdl.org Image source: www.labins.org

PAGE 59

59 Figure 3 2 Sampling design from a sugarcane field with Immokalee soil series. Data source: www.fgdl.org Image source: www.la bins.org

PAGE 60

60 CHAPTER 4 SATURATED HYDRAULIC CONDUCTIVITY AND SOIL WATER RETENTION CURVE Background Saturated water and residual water contents are used as model parameters for modeling water movement in soil ( Mmolawa and Or, 2003 ; Kandelous and Simunek 2010 ; vanGenuchten, 1980 ). Soil is saturated when all the por e spaces are filled with water and the water content is known as saturated water content. Residual water content is essentially the water content at the permanent wilting point ( pressure, h, = 1500cm). The residual water content is obtained using water retention curve at a large negative pressure value (van Genuchten, 1980). Conductivity is defined as so ability to transmit water, thus, the rate at which a saturated soil transmits water is called saturate d hydraulic conductivity (Ksat) and is used to model water movement ( Kandelous and Simunek, 2010). The saturated hydraulic conductivity ( Ksat ) remains constant regardless of the changes in water pressures and fluxes. Saturated water content, residual water content, and saturated hydraulic conductivity ( Ksat ) value can be determined experimentally under laboratory conditions. The hydraulic conductivity is affected by pore size distribution, structure, bulk density, and total porosity. When dete rmining saturated hydraulic constant (Ksat), soil cores are saturated and connected to a constant head device. The head is adjusted until water starts flowing through the soil core from the bottom. Water flowing through the core is measured as a function o f time. Flux (q) = Ksat (H1 H2)/(x1 x2). Where H (h+z) is the pressure head. The numbers, 1 and 2, represent bottom and top positions of the soil core. Flux (q) is c alculated by di viding the collected volume (Q) by product of cross section area (A) and

PAGE 61

61 ti me (t). The water content (g g 1 ) of a saturated soil under laboratory conditions is given by the equation, (mass of saturated soil mass of oven dry soil) /mass of oven dry soil. The water retention curve is the plot of moisture content versus pressure or pressure head (van Genuchten, 1980). Texture (particle size distribution) structure (arrangement of soil particles) and organic matter determine the water retention curve. Organic matter is hydrophilic in nature and affects the soil structure. When de termining moisture retention curve, soil cores are subjected to varying pressures and corresponding moisture contents determined. In this study, water retention curve and Ksat were determined. The hypothesis was; (i) the retention curve consta m) saturated water content, residual water content, and saturated hydraulic conductivity ( Ksat ) values for A E, and Bh horizons of sandy Immokalee soil will be different when determined experimentally T he objective s w ere;(i) determine moisture retention curve, saturated hydraulic conductivity ( Ksat ) values, saturated moisture contents, and residual moisture contents for A, E, and Bh horizons that are used for modeling water transport ; and (ii) Use vanGenuchten model to calculate parame ters m Materials and Methods The cores used to sample soil horizons we re first weighed. Soil cores of A, E, and Bh horizons were sampled from three locations. Triplicates of three horizons (A, E, and Bh) were therefore sampled. The soil cores were wrapped in wh ite plastic bags and put in a cooler. The cores were taken to the laboratory and prepared to attain flat surfaces on both side s A fla t knife was gently placed on top and bottom of the soil core and the soil tha t was not inside the core was removed. They were then wrapped in white

PAGE 62

62 cheese cloth After wrapping the cores, they were then satu r ated by placing the cores on water filled trays. The soil cores were saturated ove rnight. The cheese cloth was taken off the samples. The samples were put in tempe cells to fully saturate. The weights of the saturated cores were recorded. All the cores were subjected to pressures, 10, 18, 32, 56, 100, 316, 1000, 1500 Kpa. Different lengths of water columns were imposed on the core rings. The pressures were imposed on the cores until no water was witnessed dropping out of the cores. After each pressure treatment, cores were reweighed to obtain moisture content values. The soil water retention curve is the plot of moi sture conten ts versus pressures. The moisture data were modeled using vanGenuchten model (vanGenuchten, 1980). For saturated hydraulic conductivity determination triplicates of soil cores from soil horizo d height were 7.3 cm and 3 cm respectively. The calculated area an d volume of the cores were 22.91 cm 2 and 68.73 cm 3 respectively. The soil cores were saturated and water was introduced from the bottom until when a steady flow state was attained. After a s teady state flow state w as reached, the volume of water (Q) passing through the core at a specified time (t) was measured. The flux (q) is calculated using the equation Q/At. The flux also is calculated using the equation, Ksat (H1 H2)/(x1 x2). Where H1 a nd H2 are the pressure heads at bottom (x1) and top (x2) core positions. The texture of the soil was determined by weighing 50 g of air dried soil into 500 mL Erlemeyer flask. The organic matter was removed from A horizon by subjecting the soil to hydroge n peroxide The hydrometer method was used to determine percentages of clay, silt, and sand (soil survey laboratory methods manual, 1992).

PAGE 63

63 Results and Discussion The saturated hydraulic conductivity ( Ksat ) values followed the order A (21.4 cm hr 1 ) > E ( 6 cm hr 1 ) > Bh (1.57 cm hr 1 ) (Table 4 1) The water needed to saturate soil horizons follows the order Bh (0.35 cm 3 cm 3 ) > A (0.3 cm 3 cm 3 ) > E (0.26 cm 3 cm 3 ) (Table 4 1 ; Figures 4 1, 4 2, and 4 3 ). Figures 4 1 4 2 and 4 3 show that increase in pressu re from 10 Kpa to 18 Kpa lead to drastic de crease in moisture content. The drastic decrease in moisture content was more pronounced in E horizon followed by A and least in Bh horizon. After 18 Kpa moisture content gradually decreased with increases in pre ssure. The texture for the different horizons are; A horizon (96.8 % sand, 2 % silt, and 1.2 % clay), E (98 % sand, 0.9 % silt, and 1.1% clay ) and Bh (89 % sand, 3 % silt, and 8 % clay) This means that Bh horizon will hold more water than A and E at satu ration and is attributed to higher clay content in Bh horizon. At much higher negative pressure, Bh horizon will hold more water than A and E horizons. A slight decrease in pressure causes a sharpest decr ease in moisture content for E horizon, followed b y A, and least in Bh and this is due to texture and structure of the E horizon (Figures 4 1 4 2 and 4 3) The E horizon essentially consists of large sized sand particles which on arrangement lead to large pore sizes that empty quickly with slight decrea se in pressure. Since E horizon has the highest percentage of san d compared to A and Bh horizons it holds the least amount of water at large negative pressure head ( 1500cm) (Figure 4 2) Unlike A horizon with more organic matter and Bh with higher amounts of clay, E has very low organic matter and very low clay percentage. Organic matter and clay reduc e the size of pores in soil horizons and for this study pore size follows the order, Bh > A > E. The shapes of moisture release

PAGE 64

64 curves (Figures 4 1 4 2 and 4 3) are determined by the pore sizes, pore size distribution, texture, and structure of the soil horizons of interest. The saturated hydraulic conductivity ( Ksat ) value for A (21.4 cm hr 1 ) horizon was significant ly greater than saturated hydraulic condu ctivity ( Ksat ) value for E (6 cm day 1 ) horizon (P = 0.002). The saturated hydraulic conductivity ( Ksat ) value for A (21.4 cm hr 1 ) horizon was also significantly greater than saturated hydraulic conductivity value for Bh (1.57 cm day 1 ) horizon (P=<0.001) The saturated hydraulic conductivity ( Ksat ) value for E hori zon (6 cm day 1 ) was significantly greater than saturated hydraulic conductivity ( Ksat ) value for Bh (1 .57 cm day 1 ) horizon (P=<0.001). The pattern, A (21.4 cm hr 1 ) > E (6 cm hr 1 ) >Bh (1.57 cm hr 1 ) in saturated hydraulic conductivity ( Ksat ) values explains the importance of determining saturated hydraulic conductivity ( Ksat ) values in the laboratory since it is one of the parameters used for modeling water movement in Hydrus 1D The range of s aturated hydraulic conductivity ( Ksat ) values, 0.152 60.5 cm h r 1 for sand texture was reported by B otros et al., 2009. The corresponding ranges for saturated and residual moisture contents were 0.222 0.399 and 0 0.166 cm 3 cm 3 for sand texture respectiv ely (Botros et al., 2009). A saturated hydraulic conductivity ( K sat ) value, 20.4 cm h r 1 was reported for A horizon of Immokalee fine sand (Nachabe et al., 2004) and this value comp are well with the saturated hydraulic conductivity ( Ksat ) (21.4 cm hr 1 ) d etermined in this study The saturated hydraulic conductivity ( Ksa t ) values, 38.8 cm h r 1 for 0 18 cm depth and 28.0 cm h r 1 for 18 64 cm depth, for Myakka fine sand which is also a sandy soil were

PAGE 65

65 also reported by Nachabe et al., 2004. The field capacity values from point on a moisture release curve (100Kpa,1020 cm pressure he ad) where rapid water drainage changes to slower water drainage were 0. 1 cm 3 cm 3 0.06 cm 3 cm 3 ,and 0. 21 cm 3 cm 3 for A, E, and Bh respectively Concluding R emarks Since the magnitu d e of the saturated hydraulic conductivity ( Ksat ) values, A (21.4 cm hr 1 ) E ( 6 cm hr 1 ) and Bh horizon (1.57 cm hr 1 ) are not the same for each soil horizon saturated hydraulic conductivity values should be determined in the laboratory to ensure good water movement predictions. The pressure at which water will be released from soil will depend on soil properties (Ksat and texture) of the soil horizons. Water release curves provide information on saturated and residual water contents of different soil h orizons. Measuring these soil properties help modelers to comfortably predict water movement in a study area of interest without necessarily relying on literature values. The shapes of the moisture reflect the properties of soil horizons. Properties like pore sizes, pore size distribution, texture, and structure. Information on how soil horizons retain water that is adequate for plant root uptake can also be obtained from moisture release curve. For example field capacity values for A, E, and Bh were 0. 1 0 cm 3 cm 3 0.06 cm 3 cm 3 and 0.2 0 cm 3 cm 3 respectively. Although the texture of soil horizons of Immokalee soil is dominated by sand, this study has revealed significant differences in saturated hydraulic conductivity (Ksat) values. Farmers should there fore determine values for other soil orders to obtain reliable values for modeling water transport. Hypothesis, the retention curve constants

PAGE 66

66 ( conductivity (Ksat), values for A E, and Bh horizons of sandy Immokalee soil will be different when determined experimentally, was proven after saturated hydraulic c onductivity values differing significantly between A, E, and Bh horizons.

PAGE 67

67 Table 4 1. Parameters for modeling water m ovement Horizon Ksat (cm hr 1 ) s (cm 3 cm 3 ) r (cm 3 cm 3 ) A 21.4 0.3 0.06 E 6 0.26 0.04 Bh 1.57 0.35 0.17 Figure 4 1. Moisture relea se curve for A horizon Figure 4 2. Moisture release curve for E horizon

PAGE 68

68 Figure 4 3. Moist ure release curve for Bh horizon

PAGE 69

69 CHAPTER 5 E FFECTS OF SUPPORTING ELECTROLYTES AND FER TILIZER MIXTURE ON SORPTION BEHAVIOR OF PHOSPHORUS Background Phosphorus fer tilizer is applied to sugarcane fields on the surface at an application rate of approximately 50 k g P 2 O 5 ha 1 Other fertilizers that are applied to s ugarcane fields with phosphorus fertilizers are potassium and nitrogen fertilizers at a rate of approximat ely 200 k g K 2 O ha 1 and 200 k g N ha 1 respectively ( Obreza et al., 1998 ) R ain fall and irrigation water dissolves the applied fertilizers and the solution that leaches from A horizon to the water table co ntains potassium, ammonium, nitrate nitrite and phos phorus. The phosphorus sorption behavior has been studied using solutions with phosphorus sources like KH 2 PO 4 prepared in different concentrations of for example potassium chloride and calcium chloride ( Rhue et al., 2006 ; Dou et al., 2009 ) The phosphorus sorption data from experiments conducted using solutions of fertilizer mixture are not available in literature. The ionic strength and composition of ions in solutions containing a fertilizer solution (phosphorus, nitrogen, and potassium ) are hypothesize d to change the interaction of phosphorus with soil and how much phosphorus moves to the water table. Linear ized sorption coefficient (K D ) is used to model phosphorus movement in soils The linearized sorption coefficient value is defined as a ratio n of sor bed concentration to equilibrium solution concentration after 24 hours of equilibration. The supporting e lectrolytes, KCl and CaCl 2 are frequently used by researchers to conduct phosphorus sorption experiments (Rhue et al 2006; Reddy et al., 1998; He et al., 1999; Dou et al., 2009; Ahmed et al., 2008; Penn et al., 2005; Rubio et al., 2008). To

PAGE 70

70 attain good predictions of phosphorus movement, the linearized sorption coefficient values should be determined using solutions with ionic streng th, composition, an d pH close to that of the liquid phase of the sugarcane fields. Decrease in pH of the sup porting electrolytes increased phosphorus sorption (Antelo et al., 2005). Sorption of phosphorus in soil has been reported to be higher when calcium chloride was used instead of potassium chloride (Pardo et al., 1991) This is due to the divalent cation (calcium ion) which has a high er affinity to the exchange site and polarizability than monovalent cation (potassium ion) (Curtin et al., 1992) Increasing the concentra tion of supporting electrolytes and higher charg ed electrolyte cations increases phosphorus sorption (Barrow et al., 1980 ; Curtin et al., 1992 ). The amount of hydrogen phosphate ions (H 2 PO 4 ) attracted to the exchange sites increases since the surface pote ntial becomes less negative (Curtin et al., 1992). Modelers need reliable databases of sorption coefficients values to assess movement of phosphorus in sugarcane fields and identify the adverse effects of soil applied phosphorus on ground water quality Th e sorption parameters kinetic rate coeff icient for desorption from the t ype 2 sites (k 2 ) (Cameron and Klute, 1977) instantaneous sorption fraction of type I sites (F) and are obtained from the sor ption kinetics experiments. The two site model (Nkedi Kizza et al. 2006) is then employed when optimizing the parameters from phosphorus sorption data. The two site model describes sorption domains as instantaneous for the one domain and ra te limited (kin etic reaction) for the other domain ( Brusseau and Rao, 1990 ; Travis and Etnier 1981 ; Lee et al., 1988 ) The parameters that are needed when modeling with two site model are type 1 fraction sites (F) and type 2 rate parameter which is also called mass

PAGE 71

71 tran sfer coefficient ( Kookana et al., 1993). The velocity dependence of the instantaneous sorption fraction of type I sites ( F ) and type 2 rate parameter has been reported and this is associated with diffusion of the solut e in to the soil matrix and limitation of sorption (Kookana et al. 1993 ). The tailing of the break through curves for the sorbing solutes has been attributed to the sorption kinetics (Kookana et al., 1993). For ammonium, sorption is time dependent for all sites therefore data are fit to one si te model (Jellali et al., 2010). The one site model is where all sorption sites are classified as time dependent ( Brusseau and Rao, 1990). The electrolytes 0.01M KCl, 0.005M CaCl 2 deionized water, and simulated Florida rain were used to conduct phosphoru s sorption experiments. The electrolyte that was close to fertilizer mixture ( phosphorus, nitrogen and potassium fertilizers) prepared in simulated Florida rain was identified The linear ized sorption coefficients (K D values) from potassium chloride ( 0.01 M KCl ) calcium chloride ( 0.005M CaCl 2 ) deionized water, and simulated Florida rain were compared with linearized sorption coefficient value from fertilizer mixture ( phosphorus nitrogen and potassium fertilizers). The sorption kinetics experiments were then conducted using the electrolyte close to fertilizer mixture and fertilizer mixture prepared in simulated Florida rain The hypothesis was ; (i) Phosphorus sorption coefficients (K D ) when determined using different s upporting electrolytes (0.01M KCl, 0 .005M CaCl 2 simulated Florida rain, deionized water and fertilizer mixture ) will significantly differ The objective of this study was ; (i) t o identify the support ing electrolyte that mim ics sorption behavior of phosphorus when applied as fertilizer mixt ure (nitrogen, phosphorus and potassium ).

PAGE 72

72 Materials and Methods Sorption of phosphors was characterized using potassium chloride ( 0.01M KCl ) calcium chloride ( 0.005M CaCl 2 ) deionized water, simulated Florida rain and fertilizer mixture. A fertilizer mix ture is a solution prepared by mixing nitrogen, phosphorus and potassium fertilizers and proportions are based on application rate, 50 k g P 2 O 5 ha 1 200 kg N ha 1 and 200 k g K 2 O ha 1 The linearized sorption coefficient value from fertilizer mixture was ca lculated and compared with linearized sorption coefficient values from the rest of the supporting electrolytes used A supporting electrolyte with a linearized sorption coefficient value close to the linearized sorption coefficient value from the fertilize r mixture was identified. Soil Sampling and Soil Properties Measured The commercial sugarcane growin g company in s outhwest Florida US sugar Inc ., provid ed two sugarcane fields where soil samples were obtained for spatial analysis (C hapter 3 ) Soil sample s were obtained from eighty sample positions on a grid and twenty sample positions that were rando ml y selected. Soil characteristic values (total carbon, total phosphorus and pH) were clustered in to five groups. The total carbon, total phosphorus and pH values were grouped in increasing order in magnitude of values The first group had the lowest values and the fifth group has the highest values. A a n d Bh horizons for Immokalee soil and the A a n d Bw for Margate soil at one sample location was rando ml y s elected from each group A 2 mm sieve was used to sieve the air dried soil horizons. The sorption experiments conducted using soil samples from the two fields were used to identify supporting electrolyte that mimics fertilizer mixture. After identifying t he supporting electrolyte that was close to fertilizer mixture, sorption

PAGE 73

73 experiments were conducted using the A horizon that was used for column leaching experiments. Fertilizer mixture and closest supporting electrolyte were used for sorption experiments to obtain linearized sorption coefficient valu es and kinetics parameters. Since Hydrus 1D was validated using leaching experiment data and using lysimeter data, soil horizons (A, E, and Bh) that were packed in the columns were sampled near lysimeters. Th ree pits (triplicates) were dug near the lysimeter and soil sa mples were taken for air drying in a dyer room set at 45 0 C. The soil characteristics determined prior to sorption experiment setups were pH (1:2 H 2 O), oxalate extractable aluminum, oxalate extra ctable iron, exchangeable calcium tot al carbon, texture, and total phosphorus (Table 5 1 ). A 1:2 soil to solution ratio by weight was used to measure pH using a pH meter ( Sparks, 1996 ). Four grams of soil and 8 mL s of water were used and soil solution sti rred. After 30 minutes, the pH was measured using a pH meter that was standardized with pH buffers, 4, 7and 10 ( model: AR15; manufacturer: Fisher Scientific ). The pH affects phosphorus reactions in soil for example high pH values promote precipitation of p hosphorus with calcium. Low pH values (acidic) promote precipitation of phosphorus with aluminum and iron. Ammonium oxalate (0.175 M) and oxalic acid (0.1 M) we re combined to extract oxalate extractable aluminum and iron Volumes, 700 mL s of 0.2 M ammoniu m oxalate and 535 mL s of 0.2 M oxalic acid, were combined to obtain extracting solution. The pH of the solution was first a djusted to 3 by either adding 0.2 M ammonium oxalate or 0.2 M oxalic acid ( McKeague and Day, 1966). A soil to solutio n ratio of 1: 25 (2 g of soil and 50 mL s of oxalate solution) was used. After shaking at low speed the soil solution for four hour s, the solution was subjected to centrifugation for 20 minutes at 2000 rpm.

PAGE 74

74 After filtration through a 42 Whatma n filter paper, solutions were analyzed using atomic absorption spectrophotometer ( model: Optima 700DV and manufacturer: Perkin Elmer ). Aluminum and iron oxides have high sorption for phosphorus. Total soil carbon was determined using an elemental analyzer, Carbo Erba NA 2500 instrumen t (Sparks, 1996) The soil used for carbon analysis was subjected to grinding and soil that passed through 0.5 mm sieve was used for total carbon analysis. Two g rams were wrapped into alum inum foil and fed into Carbo Erba NA 2500 instrument (model: NA 2500 ; manufacturer: CE instruments) subjected to high temperatures (1000 0 C followed by 1800 0 C) Values of total carbon were obtained using a calibration curve from soil standards of know n total carbon content. The negatively charged surface of organic matter associates with cations that bridge with negatively charged phosphate ions during phosphorus sorption The method used to determine particle size distribution was m icro pip e tte method ( Soil survey staff, 1996). Fifty grams of soil were weighed into 500 mL Erlemeyer flask. Hydrogen peroxide was subjected to samples with high organic matter content to remove organic carbon. Particle size distribution was determined using hydrometer method (soil survey laboratory methods manual, 1992). Soil samples with high a mounts of clay have a larger surface are for phosphorus exchange than samples with high sand percentages. For total phosphorus d etermination, concentrated sulf uric acid and hydrogen peroxide were used for digestion and oxidation of organic matter respectiv ely ( Bowman, 1988 ). Ground soil samples (2 g) that passed through 0.5mm mesh was subjected to 2 mL s of concentrated sulf uric acid and digestion tubes put into block set 340 0 C for one hour. Aft er cooling for 5 minutes, 0.5 mL s of hydrogen peroxide were adde d and digestion

PAGE 75

75 continued for another 15 minutes. Additions of hydrogen peroxide block heating, and cooling procedure continued until when a clear solution was attained. All the added hydrogen peroxide was evaporated by block heating for about an hour. S olutions added to phosphate reagent to obtain a phosphorus characteristic blue color. Sol utions were analyzed for total phosphorus using a spectrophotometer (model: DR/4000U; manufacturer: HACH company). A calibration curve obtained from plotting wavelengt hs and known phosphorus values were used to calculate the unknown phosphorus values of the samples. When determining e xchangeable calcium (Ca) and cation exchange capacity, 30 mL s of 0.2 M NH4Cl were added into 5g weighed into centrifuge tubes (Soil S u rvey S taff, 1996). After capping, centrifuge tubes were subjected to high speed (60 rpm) for 5 minutes using a reciprocating shaker (manufacturer : Eberbach ) centrifuged and decanted into a beaker. A volume, 30 mL s of 0.2 M NH4Cl, was added and procedure repea ted four more times. After c ombining the supernatants, the volume was made to 250 mL s with 0.2 M NH4Cl. The solutions were analyzed for exchangeable calcium magnesium (Mg), potassium (K), aluminum (Al), and sodium (Na), using inductively coupled plasma (m odel: Optima 700DV and manufacturer: Perkin Elmer). In centimoles of cation charge per kilogram (cmol/kg), the exchangeable cation is given by (M n+ x Vxn)/(WxA) where M n+ is concentration of cation in extract ( ppm), V is volume of extract ( 250 mL s ), n is val ence of cation W is weight of soil ( 5 g) ,and A is atomic weight of cation. The effective cation exchange capacity ( CEC ) is the sum of all the cations (Ca, Mg, K, Al, and Na). The positively charged cations associates with negatively charged soil surface t hat attracts negatively charged phosphate ions on the exchange sites.

PAGE 76

76 Sorption Experiments F our supporting electrolytes, potassium chloride ( 0.01M KCl ) calcium chloride ( 0.005 M CaCl 2 ) deionized water, and simulated Florida rain, were used to conduct s orption experiments. Simulated Florida rain chemistry used was 13.9, 30.7, 9.5, 21. 7, L 1 of NO 3 SO 4 2 NH 4 + Ca 2+ Mg 2+ Na, K + and Cl respectively (Villapando,1997) Electroconductivities of different electrolytes were measured using electro conductivity meter ( Pinnacle Series: M541P ) Four grams of soi l were added to 8 mL s in 50 mL polycarbonate centrifuge tubes and stirred using a glass rod The soil solution was left to stand for four hours. After four hours, the elctroconductivities of the solutions were determined using electrocon ductivity meter (mo del: AB30; manufacturer: Accumet ). A soil to solution ratio of 1:2 was used for sorption experiments. S oil was subjected to ranges of initial phosphorus c oncentrations (Table 5 3) prepared in four different electro lytes. The fertilizer mixture (nitrogen, p hosphorus and potassium ) was prepared in simulated Florida rain. Although some initial concentrations are slightly greater than others, we aimed at using the same initial concentrations for a given electrolyte and soil horizon. The maximum initial concent rations o f phosphorus, nitrogen, and potassium did not exceed the typical application rates used by local farmers ( 50 k g P 2 O 5 ha 1 200 k g K 2 O ha 1 and 200 kgN ha 1 ) Soil solution was subjected to shaking at low speed using a reciprocating shaker for 24 hours. Soil solution was then centrifuge d and filtered using 42 Whatman filter paper. Solutions were analyzed for phosphorus using a spectrophotometer (model: DR/400 0U; manufacturer: HACH company) Sorption isotherms were derived from phosphorus sorption data and linearized sorption coefficient values calculated using

PAGE 77

77 appropriate equations (linear, Langmuir, and Freundlich) Equations 5 2 5 4, and 5 5 were used to calculate linearized sorption coefficient values from Langmuir, Freundlich, and linear isoth erms. ( 5 1 ) ` ( 5 2 ) The Freundlich equation ( McGill et al 1999) is described by equations 5 3 and 5 4 ( 5 3 )

PAGE 78

78 S = K D C ( 5 5 ) The linearized sorption coefficient values calculated using different electrolytes were compared with the linearized sorption coefficient value from the fertilizer mixture. Where S = amo unt of phosphorus retained (mg kg 1 ), C = c oncentration of phosphorus after a certain p eriod of equilibration (mg L 1 ), and S max = phosphorus sorption maximum (mg kg 1 ), and k = a constant related to the bonding en ergy (L mg 1 P), and K f is the constant and N is a constant. This means that linearized sorption coefficient is also a function of equilibrium solution concentration (C) The ammonium data were fit to linear sorption isotherm (E quation 5 5 ). For the phosp horus sorption kinetics experiments potassium chloride ( 0.01M KCl ) and fertilizer mixture were used. For potassium chloride ( 0.01M KCl ) 19 ppm, 29.5 ppm, and 38.9 ppm of phosphorus were the initial concentrations used for the sorption kinetics. For fer tilizer mixture, 17.7 ppm and 35.5 ppm of phosphorus were the initial concentrations used for the sorption kinetics. The three concentrations were used to show that different concentrations do not affect the fraction of phosphorus due to instantaneous sorp tion. So il solutions were analyzed for phosphorus after 4hrs, 8hrs, 12hrs, 24hrs, and 48hrs The data was subjected to two site kinetics model to determine parameters (k 2 and F) that are used for modeling phosphorus movement using Hydrus simulation environ ment. Equation 5 6 was used to calculate the relative solution concentration, phosphorus availability index, (C/C in ).

PAGE 79

79 ( 5 6 ) Where, Where C is so lute concentration in water (mg L 1 ),C in is initial a dded solution concentration (mg L 1 ), R is ret instantaneous region,k 2 is kinetic rate coefficient for deso rption from the Type 2 sites ( h 1 ), t is time (h), Kp is linear ized sorption coefficient at equilibrium (L kg 1 ) V is volume of water (L), a nd M is mass of soil (kg). The model optimizes k 2 F, and Kp which are used to recalculate C Cin 1 The layout of the two site model is shown in Figure 5 1. Where K is linear ized sorption coefficient at equilibrium, S 1 is sorbed solute concentration on the s oil of the equilibrium region (S 1 = FKC at all times), k 1 is kinetic rate coefficient for sorption to the Type 1 sites, and S2 is sorbed solute concentration on the soil of the kinetic region. Other constants were defined earlier. The solution volu me (1 0 m L s), mass (5 g), total mass of phosphorus (190 g for example for C0 of 19 ppm), and equilibrium solution concentration as a function of time are the model inputs from the sorption experiments. The k 2 and F parameters are then used to model phosphorus move ment. When fertilizer mixture was used, solutions were analyzed for phosphorus, ammonium, nitrate and nitrite. Ammonium, nitrate, and nitrite was analyzed using Lachat ( model : Quikchem 8500 ; manufacturer: LACHAT company ) Lachat uses two calibration curve s, ammonium and nitrate calibration curves, to calculate values of the samples. The solutions for color development are prepared in such a way that solutions

PAGE 80

80 for ammonium and nitrate are different. The intensity of color is correlated with the concentratio ns on the calibration curve. Results and Discussion Identifying a Supporting Electrolyte that Mimics Fertilizer Mixture Total carbon tot al phosphorus texture pH (1:2H 2 O), oxalate extractable aluminum oxal ate extractable iron exchangeable calcium, and cation exchange capacity (CEC) of the soil us ed to characterize sorption of phosphorus were measured (Table 5 1) The ionic strengths and e lectroconductivities of the electrolytes used for phosphorus sorption experiments were also measured (Table 5 2) The i nitial concentrations used for phosphorus sorption experiments were within the same range for the same electrolyte ( Table 5 3 ) The ionic strength and pH of the highest initial concentrations were also measured ( Table 5 4 ) The equilibrium solution conce ntrations after 24 hours and calculated sorbed concentrations were recorded (Table 5 5). Isotherms (linear, Freundlich, and Langmuir) were derived from the phosphorus sorption data. The equations, K D = K f C max N 1 (Freundlich), K D = S max k/ (1+kC) (Langmuir), and K D = S/C (linear) were used to calculate sorption coefficients. Negligible sorption of phosphorus on E horizon was observed (Table 5 8). Phosphorus sorption data for A and Bh horizons were used to identify the behavior of phosphorus with different ele ctrolytes. For the Freundlich isotherms, the maximum solution concentration (C max ) used to calculate a linearized sorption coefficient (K D ) value was 10 ppm. The average linearized sorption coefficient value for each supporting electrolyte was calculated ( Table 5 6) The solution pH slightly increased after 24 hours of equilibration. The highest concentration (85 ppm) prepared in calcium chloride ( 0.005M CaCl 2 ) was fed in

PAGE 81

81 to the visual minteq to investigate the effect of calcium chloride ( 0.005 M CaCl 2 ) on i on speciation. Results indicated that there was no precipitation of phosphorus when calcium chloride ( 0.005 M CaCl 2 ) was used as an electrolyte. The pattern in the sorption results in this study was attributed to the nature of cations in supporting electro lytes Sorption of phosphorus by soils was quantified by measuring phosphorus in solution and the value used to calculate the sorbed concentration after 24 ho urs of equilibration. Freundlich equation described phosphorus sorption by all the soil horizons, A, Bh, and Bw (Figures 5 2 5 3 5 4 5 5 5 6 and 5 7) Sorption of phosphorus on soils followed the trend, calcium chloride ( 0.005M CaCl 2 ) > potassium chloride ( 0.01M KCl ) > simulated Flori da rain > deionized water (Table 5 6 ). The linearized sorption coefficient values (0.01M KCl) were closest to linearized sorption coefficient values determined using fertilizer mixture p repared in Florida rain (Table 5 6). Nair et al., 1984 reported that phosphorus sorption (0.005M CaCl 2 ) is greater than (0.01M KCl ) a nd authors attributed the pattern in sorption behavior to electrolyte cation charge Greater sorption of phosphorus with calcium ion ( Ca 2+ ) than potassium ion (K + ) was observed The higher affinity of calcium ion on the exchange site than potassium ion exp lains the trend in phosphorus sorption data ( Pardo et al., 1991 ). Higher negatively charged phosphate ions (H 2 PO 4 ) were attracted to positively charged calcium ion on the soil exchange site. The linearized sorption coefficient values for Bh horizon (Tabl e 5 6) were higher than the values for A horizon and this is due to higher sorption of phosphorus by Bh horizon (Li et al 1997). The Bh horizon accumulates organic matter with higher amounts of aluminum and iron than A horizon.

PAGE 82

82 The repulsion effect betwe en phosphate ion and soil surface is higher with potassium ion than with calcium ion (Pardo et al 1991). Surface potential of exchange site becomes less negative for divalent cation in supporting electrolyte than a monovalent cation (Curtin et al., 1992) This increases the attraction of negatively charged hydrogen phosphate ions (H 2 PO 4 ). In addition to potassium ion from potassium hydrogen phosphate ( KH 2 PO 4 ) used as a phosphorus source for sorption experiments, calcium chloride ( 0.005M CaCl 2 ) and potas sium chloride ( 0.01M KCl ) have additional cations in solution. This explains the higher sorption of phosphorus compared to deionized water and Florida rain The polarizability and hydration energy also follows the trend calcium chloride ( 0.005 M CaCl 2 ) > p otassium chloride ( 0.01M KCl ) (Pardo et al., 1991 ). Sorption of phosphorus with simulated Florida rain is higher than deionized water due to the cations added when preparing Florida rain. The K D values (0.01M KCl) were closest to linearized sorption coeffi cient values (fertilizer mixture) due to potassium chloride added in fertilizer mixture. Parameters for Modeling Ph osphorus and Ammonium Movement Sorption experiments were conducted using potassium chloride ( 0.01M KCl ) and fertilizer mixture to obtain the linearized sorption coefficient values for modeling phosphorus movement The linearized sorption coefficient value for ammonium was obtained using fertilizer mixture The linearized sorption coefficient value for ammonium was 0.2 mL g 1 (Figure 5 10 ). The low linearized sorption coefficient value is attributed to low carbon content of A horizon. Ammonium sorption was reported to increase with organic carbon content (Wang and Alva, 2000). The A and E horizon s were sampled near the lysimeters. The linearized sorption coefficient values from potassium chloride

PAGE 83

83 ( 0.01 M KCl ) fertilizer mixture calcium chloride ( 0.005 M CaCl 2 ) deionized water, and Florida rain were 1.6 mL g 1 1.4 mL g 1 2.4 mL g 1 0.6 mL g 1 and 0.6 mL g 1 respectively (Figures 5 6 5 7 ,5 8,5 9, and 5 10 ) The phosphorus sorption kinetics data fit the two site model (Figure s 5 8 to 5 9 ). The data in Table 5 8 show that different initial concentration s do not have an effect on the magnitude of k 2 and F values A similar effect was repor ted by Nkedi Kizza et al., 2006 The authors characterized sorption and sorption kinetics of organic pesticides on carbonatic soils of South Florida using different initial concentrations. The sorption kinetics parameters ( k 2 and F ) values did not vary wit h initial concentrations. Concl uding R emarks The supporting electrolyte that yielded linearized sorption coefficient values close to fertilizer mixture was potassium chloride ( 0.01M KCl ) The chemistry, ionic strength, pH, conductivity, and charge of catio ns, affect the amount of phosphorus adsorbed by soil. This will eventually affect the magnitude of the linearized sorption coefficient values When modeling phosphorus movement in sugarcane fields, the linearized sorption coefficient (0.005M CaCl 2 ) will un der predict phosphorus movement since the higher the linearized sorption coefficient val ue, the higher the sorption of phosphorus The linearized sorption coefficient (deionized water) and linearized sorption coefficient (simulated Florida rain) will over predict phosphorus movement since the lower the linearized sorption coefficient value, the lower the sor ption of phosphorus Modelers dealing with soil system that involves fertilizer mixture applied to sugarcane fields can comfortably use linearized sor ption coefficient (0.01 M KCl) and linearized sorption coefficient (fertilizer mixture) to model phosphorus movement.

PAGE 84

84 Before modeling nutrient movement, modelers should consider the ionic strength of the liquid phase of the system of interest. Researcher s dealing with the soils where sugarcane is grown in Southwest Florida should use potassium chloride ( 0.01 M KCl ) or fertilizer mixture (50 k g P 2 O 5 ha 1 200 kgN ha 1 a nd 200 k gK 2 O ha 1 ) when determining sorption coefficients. The phosphorus sorption kinet ics experiments should also be conducted using potassium chloride ( 0.01M KCl ) or fertilizer mixture. The sorption coefficient of ammonium, 0.2 mL g 1 shows that ammonium will move faster in soil than phosphorus with linearized sorption coefficient value o f 1.4 mL g 1 This study has shown that potassium chloride (0.01 M KCl) should be used as supporting electrolyte to characterize sorption of phosphorus in mineral soils (for example Immokalee and Margate soils) Phosphorus sorption experiments should be cond ucted using soil samples from other orders like Alfisols without relying on the data collected using Spodosol and Entisol In this study it is was hypothesized that, phosphorus sorption coefficients (K D ) when determined using different supporting electroly tes (0.01M KCl, 0.005M CaCl 2 simulated Florida rain, deionized water, and fertilizer mixture) will significantly differ. Phosphorus sorption experiments have proven that using different electrolytes affect sorption eq uilibria since linearized sorption co efficients (K D ) values were different in magnitude.

PAGE 85

85 Table 5 1. Physical and chemical p roperties of soils Property A I A M Bh Bw Total carbon (g kg 1 ) 15.6 12 34.1 4 Total Phosphorus (g g 1 ) 107.3 174.3 89 34.4 pH (1:2H 2 O) 6.9 8.1 6.9 8.5 %Sand 98 98 88. 5 98 .2 %Silt 1.2 1.5 3.4 0.8 %Clay 0.8 0.49 8.1 1.01 Ox.Al (mg kg 1 ) 282.4 310.3 307.0 90 Ox.Fe (mg kg 1 ) Exchangeable Ca ( cmol c kg 1 ) CEC ( cmol c kg 1 ) 236.0 21.6 27.9 663.1 68.3 71.8 116.4 54.4 88.7 151 30.5 36.4 I = Immokalee M = Margate Ox.Al = Oxalate aluminum Ox.Fe = Oxalate iron Table 5 2. Electrolytes used for phosphorus sorption e xperiments Electrolyte Ionic Strength (I) pH Electroconductivities (S cm 1 ) 0.01M KCl 0.01 4.7 3240 0.005M CaCl 2 0.03 4.7 2670 Deionized water 0 6.8 2.7 Simulated F lorida rain 0.000032 5.1 8.9 Table 5 3. Maximum Initial concentrations for phosphorus sorption i sotherms Horizon 0.01M KCl (ppm) 0.005M CaCl 2 (ppm) Deionized water (ppm) Fertilizer mixture (ppm) Florida Rain (ppm) A M 44 48 44.3 38 37.4 A I 43 59 41 29 37.4 Bw 38 46 28 36 28 Bh 74 85 84 85 46.8 M = Margate, I = Immokalee Table 5 4. Ionic strength (I) and pH of the soil s olution Electrolyte I pH Deionized water + 0.000021 M KH 2 PO 4 0.000021 6.1 0.01M KCl + 0.00002 M KH 2 PO 4 0.01002 5.7 0.005 M CaCl 2 + 0.00002 M KH 2 PO 4 0.015 02 5.5 Rain+0.000021 M KH 2 PO 4 0.0000525 5.9 Rain+0.000021 M KH 2 PO 4 +0.0000265M NH 4 NO 3 +0.00013 M KCl 0.000209 5.6

PAGE 86

86 Table 5 5 Linear ized sorption c oefficients (K D ) Horizons 0.01 M KCl 0.005M CaCl 2 Deionized water Florida rai n Fertilizer mixture A I 2.13 0.66 4.42 0.9 1.67 0.38 2.10 0.1 2.14 0.39 A M 2.83 0.17 5.15 0.40 2.77 0.32 2.33 0.49 2.82 0.36 Bh I 10.20 1.6 12.78 2.9 4.05 0.4 6.16 0.9 10.68 2.5 Bw M 2.22 0.54 5.21 0.23 1.64 0.82 2.14 0.45 2.15 0.50 M = Margate, I = Immokalee Table 5 6 Statistical differences b etween linear sorption c oefficients (K D ) P (A I ) P (A M) P(Bh) P(Bw) Fertilizer mixture verses FR 0.96 + 0.278 + 0.0 05 + + 0.98 + Fertilizer mixture verses H 2 O 0. 14 + 0.96 + 0.001 + + 0.41 + Fertilizer mixture vers es 0.005M CaCl 2 0.03 3 ++ 0.001 + + 0. 252 + 0.003 + + Fertil izer mixture verses 0.01M KCl 0.24 + 0.714 + 0. 790 + 0. 955 + M = Margate, I = Immokalee, + = No significant differe nce, ++ = significantly greater or lower FR = Florida rain, H 2 O = Deionized water, P= Pro bability value at 95 % confidence. Table 5 7 Phosphorus sorption kinetics parameters: standard error in p arentheses Electrolyte Initial Concentration (C0) k2 F Kp 0.01M KCl 19 0.1 (0.01) 0.5(0.02 ) 1.6 (0.02) 0.01M KCl 29.5 0.1 (0.01) 0.6 (0.01) 1.2 (0.01) 0.01M KCl 38.9 0.1(0.01 ) 0.6 (0.01) 1.0 (0.01) 0.01M KCl All C0 0.1(0.01) 0.6 (0.03) 1.25 (0.02) Fertilizer mixture 17.7 0.1 (0.01) 0.4(0.03 ) 1. 6 (0.03) Fertilizer mixture 35.5 0.1(0.01 ) 0.6 (0.02) 1.2 (0.0 3 ) Fertilizer mixture All C0 0.1 (0.01) 0.5 (0.0 1) 1.4 (0.02) Table 5 8. Sorption of phosphorus on E horizon of Immokalee soil Initial c oncentration (ppm) Mass (g) Volume ( mL s) Average e quilibrium c oncentration (ppm) 23 5 20 22.5 45 5 20 44.5 75 5 20 74.8

PAGE 87

87 Figure 5 1. Two site model Figure 5 2. Phosphorus sorption isotherm in A and Bh horizons of Immokalee soil using potassium chloride ( 0.01M KCl )

PAGE 88

88 Figure 5 3. Phosphorus sorption isotherm in A and Bh horizons of Immokalee soil using fertilizer mixture Figure 5 4 P hosphorus sorption iso therm in A and Bw horizons of M argate soil using potassium chloride ( 0.01M KCl )

PAGE 89

89 Figure 5 5 P hosphorus sorption isotherm in A and Bw horizons of M argate soil using fertilizer mixture Figure 5 6. Sorbed phosphorus versus equilibrium solution phosphor us for A horizon of Immokalee soil with potassium chloride ( 0.01M KCl )

PAGE 90

90 Figure 5 7. Sorbed phosphorus versus equilibrium solution phosphorus for A horizon of Immokalee soil with fertilizer mixture Figure 5 8. Sorbed phosphorus versus equilibrium s olution phosphorus for A horizon of Immokalee soil with calcium chloride ( 0.005 M CaCl 2 )

PAGE 91

91 Figure 5 9. Sorbed phosphorus versus equilibrium solution phosphorus for A horizon of Immokalee soil with deionized water Figure 5 10 Sorbed phosphorus versus eq uilibrium solution phosphorus for A horizon of Immokalee soil with Florida rain

PAGE 92

92 Figure 5 11 Relative concentra tion (C/C0) as a function of tim e for A horizon of Immokalee soil with potassium chloride ( 0.01M KCl ) Figure 5 12 Relative concentratio n (C/ C0) as a function of time for A horizon of Immokalee soil with fertilizer mixture

PAGE 93

93 Figure 5 13 Sorbed ammonium versus equilibrium solution ammonium for A horizon of Immokalee soil

PAGE 94

94 CHAPTER 6 EFFECT OF FLUCTUATING WATER TA BLE ON PHOSPHORUS AND NI TROGEN MOV EMENT Background Farmers in Southwest Florida maintain adequate moisture with in the root zone of sugarcane plants through fluctuating elevated water table (Obreza et al., 1998 ) Water table depth in sugarcane fields is first set near the soil surface (app roximately 30 cm deep ) to insure proper irrigation of newly planted seed pieces and then lower water table depth to approximately 50 cm as the crop root zone is established. Lowering the water table depth to 50 cm provides good aeration (prevents anaerobic conditions) and more accommodation of rain (Obreza et al., 1998). Farmers decide when to lower the water table Therefore the timing can vary from farmer to farmer. Roots function well in shallow water table for a short period (Glaz and Morris, 2010). The Spodosol, Immokalee series is the dominant mineral soil series used for sugarcane production in Southwest Florida. The most reactive soil horizon of Immokalee soil is Bh horizon ( Li et al 1997 ). Leaching of phosphorus is highest in E horizon followed by A and least in Bh horizon (Li et al 1997). In this study, it is hypothesized that phosphorus and nitrogen from fertilizer leach ed below the water table to Bh horizon is increased when water table depth is lowered. Phosphorus and nitrogen movement below t he water table is driven by concentration gradient The experimental design was such that water table was restored after each rainfall events by collecting solutions from the water table outlets The distance between the water table and the Bh horizon will be redu ced when water table is lowered. Diffusing phosphorus was reported to decrease wit h distance from the source of phosphorus (Bouldin and Black, 1954). Ammonium in the aerobic layer (unsaturated

PAGE 95

95 conditions) is oxidized to nitrate by microbes which ar e aerobic autotrophs (Reddy et al., 1976). In a soil environment with aerobic and anaerobic layers, ammonium diffuses to aerobic layer from anaerobic layer. Nitrate diffuses from aerobic to anaerobic layer (Reddy et al., 1976). Moisture content, bulk densi ty, cation exchange capacity of soil presence of reduced iron, and m anganese influence ionic diffusion of ammonium from anaerobic to aerobic layer (Reddy et al., 1976). Increasing flow velocity in a saturated flow system reduces adsorption of ammonium by soil (Jellali et al., 2010). Masses of ammonium and phosphorus that moved below the water table set at 30 cm and 50 cm depths from the soil surface were calculated. Tracer, bromide has been employed by researchers to monitor water movement ( Fortin et al. 2002; Lawrence and Richard 2008 ; Shinde et al., 2001 ). Tracers are used to study hydro dy namic characteristic of a given porous medium (Shinde et al., 2001) and water flow characterization (Kookana et al., 1993) The peclet number (P) is calculated using th e equation, vL/D, where v is pore water velocity, L is length of the column, and D is dispersion hydro dy namic coefficient (Shinde et al., 2001). Preferential flow plays a key role in water movement under unsaturated conditions (Lawrence and Richard, 20 0 8 ; Kelly and Wilson 2000 ). Bromide flow velocity is affected by bulk density and clay content of the soil (Poulsen et al. 2006 ). Although presence of physical non equilibrium due to mobile and immobile water flow zones can cause tailing in tracer break thro ugh curves, normally tracer break through curves are symmetrical (Shinde et al., 2001). In this study, water table was lowered from 30cm to 50cm after 6 weeks. The water table was maintained at 50cm depth from the soil surface for 6 weeks. The masses of ph osphorus and nitrogen from cups installed below the water table were

PAGE 96

96 compared for water table set at 30cm with 50cm depths from the soil surface The bromide, phosphorus and nitrogen (nitrate and ammonium) concentration s from solutions collected from cup A were used as validation data sets for Hydrus 1D The hypotheses of the study were ; (i) Reducing distance between water table and Bh horizon through lowering water table from 30 cm to 50 cm depth will increase d iffusion o f phosphorus and nitrogen below the water table for Immokalee soil ; and (ii) Management of water table depth after rainfall events will lead to loss of plant available phosphorus and nitrogen The objectives of the study were ;(i) t o investigate the ef fect of fluctuating water table on movement of phosphorus and nitrogen above and below the water table Materials and Methods Column leaching e xperiments with both saturated flow and unsaturated flow (with fluctuating water table), were conducted to inves tigate behavior of phosphorus and nitrogen when fertilizer mixture (nitrogen, phosphorus, and potassium ) is applied to the soil The saturated flow experiment was conducted to mimic the saturated zones of the soil after rainfall additions. Unlike in satura ted flow experiment where solutions were pumped through the soil packed in the column, f ertilizer mixture (rates of 50 k g P 2 O 5 ha 1 200kg N ha 1 and 200k g K 2 O ha 1 ) were applied on soil surface as soil in unsaturated flow experiment. Water table was low ered from 30 cm to 50 cm depth after 6 weeks. Water table depth was maintained at 50 cm depth from the soil surface for another 6 weeks. The phosphorus data were modeled with Hydrus 1D using the linearized sorption coefficient values from the fertilizer mixtur e and potassium chloride ( 0.01 M KCl ) The phosphorus data from unsaturated flow experiment were modeled

PAGE 97

97 with linearized sorption coefficient (fertilizer mixture) and linearized sorption coefficient (0.01 M KCl) In a saturated experiment, phosphorus movement was modeled considering a continuous flow rate at the bottom of the column and free drainage at t he out let (top of the column). The phosphorus movement was modeled from soil surface to water tabl e ( 30 cm and 50 cm depth). In unsatu rated flow expe riment, phosphorus was modeled considering variable flux on the soil surface and constant pressure at the water table. M agnitude of phosphorus and nitrogen masses collected below the water table for both water table depths (50 cm and 30 cm) were compared t o identify the influence of water table depth on phosphorus and nitrogen movement below the water table. Saturated Flow Experiment A column 15 cm long and 7.5 cm in diameter, was used for saturated flow experiment. The A horizon was packed in a column usi ng the same bulk density found in field soils (Figure 6 1 ). The mass and volume of soil packed in a column were 982 g and 663 cm 3 respectively. The bulk density and particle de nsity for A horizon were 1.48 g cm 3 and 2.56 g cm 3 respectively. The soil in the column was first saturated with two pore volumes (544 cm 3 ) of simulated Florida rain from the bottom followed by pumping 1047.2 cm 3 ( 4.68 pore volumes), of fertilizer mixture ( 200 k g N ha 1 50 k g P 2 O 5 ha 1 and 200 kg K 2 O ha 1 ) at a steady flow rate of 10 mL per minute. Pore volumes 12.2 of simulated Florida rain was introduced in to the column to allow desorption process Desor ption is the release of nutrients that had sorbed on soil into solution. Solutions (20 mL s) were collected every two minutes from the top of the column (Figure 6 1 ). Solutions were filtered and added to known volumes of phosphate reagent to develop a phosphorus characteristic blue color. Solutions of known concentrations

PAGE 98

98 were added to phosphate reagent to obtain a calibration cu rve. The absorbance of the blue color from unknown samples was correlated to the calibration curve to obtain the phosphorus values. The solutions were analyzed for phosphorus using a spectrophotometer ( model: DR/4000U; manufacturer: HACH company ) nitrogen (ammonium and nitrate ) using Lachat ( model : Quik Chem 8500; manufacturer : LACHAT ) Two calibration curves, one for ammonium and another for nit rate were obtained from the lachat software. The calibration is used to calculate values of the samples dependin g on the color intensity. Solutions used for color development are different for ammonium and nitrate. Solutions were analyzed for chloride using ion chroma tography. Break through curve s (relative concentration (C/C0) verses pore volumes) were plotted from the data collected from saturated flow experiment The ammonium and chloride data were fit to convective dispersion model which combines equilibrium and first order kinetic equations (Cameron and Klute, 1977). Equation 6 1 is a o ne dimensional equation t hat describes convective dispersion of a solute in uniform porous media at a steady hydraulic flow; E quation 6 2 is the equilibrium model; E quation 6 3 is the first order reversible kinetic model ; and E quation 6 5 is the combined equilibrium and first ord er kinetics.

PAGE 99

99 Where C s is the solution concentration of the mobile solute; S is the sorbed concentration which is defined as total adsorption in both kinetic and equilibrium fractions ; D is hydrodynamic dispersio n coefficient; v is seepage velocity; is the bulk density; is water content; K d (dimensionless) is the equilibrium constant; k 1 is adsorption constant; k 2 is desorption constant ; F is the fraction of type 1 sites; and R T is total retardation factor. The initial and boundary conditions for a soil column of length L are; S 1 S 1 1 1

PAGE 100

100 The two site model was used to model phosphorus break through curve. When modeling phosphorus data, pecle t number from ni trate and chloride data was used, pulse volume and pulse length were fixed and optimized beta ( ), omega ( ), and retardation factor (R) Other data inputs were pore volumes as a function of relative concentrations (C/C0) and isotherm exponent (N for Freun dlich isotherm). The equations for R were and Fluctuating Water Table P lexiglass columns, 100 cm long and 15 cm in diameter were designed with outlets placed at 30 cm and 50 c m from the soil surface to collect solutions off the water table ( Figure 6 2 ) P orous ceramic cups were installed just above the lower extent of each horizon to monitor phosphorus, nitrogen and bromide concentrations at each horizon ( Figure 6 2 ) The cera mic cups were installed at 13 cm, 60 cm, and 75 cm in A, E, and Bh horizon respectively. The moisture sensors (EC 5) connected to data loggers (EM50) from Decagon Devices, Inc, 2365 NE Hopkins Court, Pullman, Washington 99163 were installed at 8 cm, 41cm, and 71 cm from the soil surface in A, E, and Bh horizons respectively ( Figure 6 2 ) A glass tubing attached on the sides of columns was used to monitor the water table depth ( Figure 6 2 ) Soil samples (A, E, and Bh horizons) were taken from location near t he lysimeters to simulate water and nutrient concentrations under field conditions (documented in C hapter 7) The soil horizons, A E, and Bh were sampled from three pits (triplicates) from which a composite was obtained The soil samples were pla ced in t he air dryer room set at 45 0 C and monitored until complete drying was attained Soil horizons, A E

PAGE 101

101 and Bh for Immokalee were packed in the columns at 18 cm, 48 cm, and 14 cm thick, respectively, taking into consideration the bulk density of each horizon f ound in the field where they were removed The depth of A, E, and Bh horizons packed were 18 cm, 48 cm, and 14 cm. The soil was saturated from the bottom ( Villapando and Graetz, 2001 ) with simulated Florida rain using rain reservoir connected to the bo tto m tap. Introducing rain from the bottom helps to attain complete saturation and avoid entrapment of air in the soil. Rainwater chemistry data for Florida rain is 13.9, 30.7, 9.5, 21.7, L 1 of NO 3 SO 4 2 NH 4 + Ca 2+ Mg 2+ and K + re spectively ( Villapando 1997 ). The water table depth at 30 cm from the soil surface was restored by the opening the taps at the water table depth The column was left to stand for three days before fertilizer application to allow stabilization of water tab le at 30 cm depth Phosphorus (KH 2 PO 4 ) n it r ogen (NH 4 NO 3 ) and potassium (K Cl and KBr ) were applied to soil surface in the columns The fertilizer application rates, 50 k g P 2 O 5 ha 1 200 kg N ha 1 and 200 k g K 2 O ha 1 were used Masses, 165 mg of KH 2 PO 4 350.5 mg of KCl, 353 mg of NH 4 NO 3 and 238 mg KBr, were applied on the soil surface of each column. The tracer for water movement was bromide ion. Rainfall amounts in southern Florida are highest between June and August Average daily values of rain receiv ed in Immokalee were calculated for 2001 to 2010 In this experiment, rainy season (June, July, and August) was explored The prob abi lities of r eceiving rain were used to calculate rain distribution. Average daily amounts of rainfall were used to simulate rain when the chance of receiving rain was more than 50 %. S olution samples were collected from the solution samplers (Figure 6 2 ) installed in soil horizons and at the water table depth after every rain addition T he solutions were

PAGE 102

102 analyzed for phosphoru s, nitrogen (ammonium and nitrate ), and bromide The water table depth was restored to test depths after each simulated rainfall using outlets placed at the water table depth The water table was lowered from 30 cm to 50 cm depth after six weeks. Columns w ere emptied at the end of the experiment and soil samples collected from each horizon for further analysis. A wet sub sample was extracted with potassium chloride ( 2 M KCl ) for nitrogen (ammonium and nitrate ) analysis by rapid flow analys is. When analyzing for ammonium nitrogen and nitrate nitrogen, a sample is injected by the instrument needle, tubes are inserted in reagents used for color development continuously supply the reagents that mix with sample. The higher the intensity of color developed, the hi gher the concentrations of ammonium and nitrate. A subsample was air dried and analyzed for M ehlich 1phosphorus using spectrophotometer ( model : DR/4000U; manufacturer: HACH company ) B romide was analyzed using rapid flow method Results and Discussion The soil properties, bulk density particle density and porosity for the A horizon used for saturated and unsaturated flow are shown in T able 6 1. C hloride ni trate, ammonium, and phosphorus from saturated flow experiment was plotted to obtai n break through curve s (Figures 6 3, 6 4, and 6 5 ). The graphs also show that chloride, nitrate, and ammonium data were fit to convective dispersive model. Phosphorus data was also fitted with the two site kinetics model. Break through curves were plotted as relative conc entrations versus pore volumes. The maximum concentrations for tracers (chloride and nitrate), ammonium, and phosphorus were observed after 1.5, 2, and 5 pore v olumes respectively (Figures 6 3, 6 4, and 6 5 ). The sim ilarity in behavior for chloride

PAGE 103

103 and nit rate for saturated flow experiment shows that the y can both be used as tracers. Phosphorus concentrations peak after ammonium because it so rbs on soil more than ammonium. The linearized sorption coefficient ( fertilizer mixture ), phosphorus and ammonium, at A horizon were 1.4 mL g 1 and 0.2 mL g 1 respectively (Figures 5 7 and 5 10 ). The phosphorus, nitrogen (ammonium and nitrate ) and bromide concentrations of the solutions collected at the bottom of each horizon and off the water table for the unsaturated flow experiment were measured. The solutions out of the A horizon were collected only when the water table was set at 30 cm depth. The solutions out of the E and Bh were collected when water table was set at 30 cm and 50 cm depths. The greatest portion of the solutions was collected off the water table after e very rain fall event. The solution volume collected f rom each cup installed at the bottom of each horizon did not exceed 20 mL s after every sampling event. Bromide was used as a tracer in the unsatur ated flow experiment with fluctuating wat er table. Bromide phosphorus, and nitrogen concentrations increased with time as water dissolves fertilizer added to the soil surface reached peak and started decreasing when solutions were diluted with further ad ditions of water ( Figures 6 6 to 6 17 ) After lowering the water table to 50 cm, bromide, phosphorus, and nitrogen were transferred from upper 30 cm to water table (50 cm) and increased with time, peaked and started decreasing as solutions were diluted with water added after (Figures 6 18 to 6 31 ). Increases and decreases in bromide and ammonium concentrations for cup E and cup Bh (Figures 6 18 to 6 23, Figures 6 26 to 6 31) for water tables set at 30 cm and 50 cm depths are attributed to varying amounts of bromide brought to water table. The

PAGE 104

104 amounts of bromide brought to the water table from the soil surface are determined by added amounts of rain. Symmetrical nature that is equal halves on either side of the curve was observed for bromide (tracer ) data from solutions collected at 30 cm water table depth (Figure 6 5 ). However tailing was observed in bromide break through curves ( Figures 6 6 and 6 14 ) for solutions collected from cup A and water table depth, 50 cm, due to physical non equilibrium created by mo bile and immobile water flow zones (Shinde et al., 2001 ; Fisher and Healy,2008 ). Some water might have moved at a slower velocity. The bromide break through curves for cups E and Bh did not follow a systematic typical break through curve for a tracer be cau se the cups were located in a saturated zone Bromide is hypot hesized to have moved from unsaturated to saturated zone through diffus ion and displacement when solutions were sucked from cups E and Bh (Figures 6 18, 6 19, 6 26, and 6 27 ). A tracer, chloride followed a similar pattern (symmetrical break through curve) in a study with homogeneous columns conducted by Beigel and Di Pietro, 1999. Break through c urves for phosphorus ammonium and nitrate are spread ( Figures 6 7,6 8,6 9,6 11,6 12,6 13,6 15,6 16, and 6 17 ) and the asymmetrical nature of the curves is attributed to sorption on soil (Shinde et al. 2001 ) Phosphorus and ammonium are therefore said to be retarded. According to Shinde et al., 2001, the asymmetrical nature of the break through curv es f or interacting solutes was also attributed to presence of chemical non equilibrium. Beigel and Di Pietro, 1999 attributed the asymmetrical nature of break through curve of triticonazole in a solute movement experiment with homogeneous columns to non e quilibrium sorption. For cup A (inserted 13 cm from soil surface) the maximum concentration s (29.2 ppm) of phosphorus and

PAGE 105

105 ammonium (79.4 ppm) were ob served after 20 days (Figure 6 9 ;Table 6 2 ) and 11 days (Figure 6 8 ;Table 6 2 ) respectively. Bromide maxim um concentration was observed after 9 days ( Figure 6 6 ; Table 6 2 ) The times where maximum nutrient concentrations were observed indicate that phosphorus is more retarded than ammonium. When fertilizer mixture (nitrogen, phosphorus, and potassium ) and A h orizon of Immokalee soil were used to determine linearized sorption coefficient values, the calculated values for phosphorus and ammonium were 1.4 mL g 1 and 0.2 mL g 1 respectively ( Figures 5 7 and 5 10 ). There is higher sorption of phosphorus than ammoniu m due to the reactions with aluminum and iron. Ammonium which is positively charged will have to compete with potassium (cation) for the exchange sites. The maximum concentration s of ammonium (42.6 ppm) and phosphorus (14.6 ppm) were observed after 22 days and 37 days respectively for water 30 cm water table depth (Figures 6 12 and 6 1 3 ) The same ratio of 2.2 for time (maximum concentration of phosphorus ) to time (maximum concentration of ammonium) was calculated for cup A and water table (30 cm) conc entrations (Table 6 2) The maximum concentration for phosphorus at the depths of 30 cm (14.6 ppm) and 50 cm ( 7.5 ppm) were different due to the fact that phosphorus was applied once throughout the whole experiment (Figures 6 13 and 6 17 ) Maximum concent rations of bromide, phosphorus and ammonium for solutions collected water table depth of 50 cm were observed after 59 days, 6 8 days, and 74 days respectively (Figures 6 14 6 16 and 6 17 ; Table 6 2 ) Ratio of maximum concentration times for phosphorus to ammonium reduces to 1.3 from 2.2 for 30 cm water table depth. This is because the greatest portion of the 50 cm is E horizon (32 cm) which registers

PAGE 106

106 negligible sorption of phosphorus and ammonium. The sorption process was probably more pronounced when wat er table was set at 30 cm from the surface. The total amount of rain simulated was 23 cm when water table was set at 30 cm. The total amount of rain simulated was 17.4 cm when water table was set at 50 cm. The total mass es of phosphorus collected from cup E were 21.4 g and 74 g when water table was set at 30 cm and 50 cm depths respectively Although it takes longer to observe an increase in phosphorus concentration when water table is set at 50 cm depth total mass of phosphorus collected is higher than at 30 cm depth The distance between the soil surface and the water table i s increased to 50 cm therefore phosphorus has to move a longer distance. Once the concentrations of phosphorus have increased at water table depth (50 cm), the concentrations of cu p E also increased soon after. The distance between the water table and Bh horizon is reduced when water table is lowered to 50 cm. The Bh horizon influence on phosphorus movement through diffusion is increased since Bh horizon is the most reactive horizon No phosphorus was detected in solutions collected from the cup installed at the end of Bh horizon. The low concentrations of phosphorus that reaches Bh horizon were strongly adsorbed Phosphorus moved below the water table through diffusion. Phosphorus m oved in solution to the water table and therefore the water table has higher concentrations than depths below. The concentration gradient created acts as a driving force for diffusion For ammonium, 1.6 mg and 1.1 mg were collected when water table was set at 50 cm and 30 cm depths from the surface respectively. Unlike phosphorus ammonium was detected in solutions collected from cup inserted in Bh horizon ( Figures 6 30 and 6 31 ).

PAGE 107

107 Nitrate values were very low compared to ammonium values. Since most of time during the experiment, soil was close to saturation (mean moisture of 0.204) anaerobic conditions lead to denitrification of nitrate (Elmi et al. 2002 ) The low nitrate values were observed at different depths of the columns for unsaturated flow with wa ter table depths (Figures 6 7, 6 11, 6 15, 6 20, 6 21, 6 28 and 6 29 ) In a study conducted by Elmi et al., 2002, sub irrigation (water table set at 60 cm from the soil surface) reduced nitrate concentrations in the soil profile and this was attributed to saturation that created anaerobic conditions. The authors also found out that the denitrification rates were not affected by nitrogen application rates. Denitrifi cation rates, 0.17 to 4.23 kg N ha 1 day 1 0.26 to 2.45 k gN ha 1 day 1 and 0.44 k g N ha 1 day 1 (relatively uniform) were reported by Kliewer and Gilliam, 1995 for 15 cm, 30 cm, and 45 cm water table depths respectively. In this study, the authors subjected the three water table depths to undisturbed cores of Cape Fear loam (clayey, mixed, thermic Typic Umbraquult) that were set up in the field. Water table depth was reported to affect the upward capillary liquid flow which enhances continuous liquid pathways that connects water table to the soil surface ( Shokri and Salvucci, 2011). An inverse rela tionship between the water table and loss of water from soil surface was reported by the authors The shallower the water table, the higher the chances of creating anaerobic conditions above the water table that promotes denitrification. The ammonium, nit rate, and phosphorus concentrations for A E, and Bh horizons were quantified at the end of the experiment. Measured ammonium concentrations were 24.7 g g 1 15.46 g g 1 15.86 g g 1 for A, E, and Bh horizon respectively. Soil

PAGE 108

108 concentrations, 39.36 g g 1 (A horizon) 27.75 g g 1 (E horizon) 4.39 g g 1 (Bh horizon) were observed for nitrate. Phosphorus was not detected in A horizon, 7.78 g g 1 and 3.11 g g 1 were phosphorus values for E and Bh horizons respectively. Unlike nitrogen (ammonium and ni trate), A horizon was deprived of applied phosphorus through desorption at the end of the experiment. Since nitrogen and phosphorus were detected in E and Bh horizons, recorded values must have been due to diffusion. Conclu ding R emarks The phosphorus conce ntrations measured fro m solutions collected from E horizon would have b e e n lower if the plants were included in the experiment. Results rev ealed that lowering the water table depth from 30 cm to 50 cm increases movement of phosphorus below the water table. The Bh horizon is probably the most influencing factor since distance from the source (water table) to the Bh horizon is reduced by lowering the water table depth. A similar trend was observed for ammonium data. The amount of phosphorus that is adsorbed b y Bh horizon is higher than the A and E horizons. The phosphorus and ammonium concentration gradient is higher when the water table is lowered to 50 cm. Higher retardation factors ( for maximum phosphorus and ammonium to bromide) for phosphorus and ammoniu m were observed when water table is set at 30 cm depth than when set at 50 cm depth This is due to the sorption process that is more pronounced when water table is set 30 cm depth Although ammonium was detected in solutions collected from Bh h orizon, no phosphorus was detected in Bh solutions. Thi s is due to higher sorption of phosphorus by Bh horizon.

PAGE 109

109 Given similar linearized sorption coefficient value for A horizons, a farm unit with largest portion occupied by Immokalee soil will experience higher move ment of phosphorus and ammonium below the water table than Margate soil The same amount of phosphorus and ammonium will be transferred from upper horizons (A and E) up to the water table. The phosphorus and ammonium that will diffuse from water table to t he Bh horizon will experience higher sorption compared to sorption capacity of Bw horizon if A, E, and Bw were packed in the column. This is pronounced when water table is lowered to 50 cm from soil surface. Margate soil is expected not to follow the sam e trend in phosphorus movement since it lacks zone of total carbon and aluminum accumulation (Bh h orizon ). Absence of physical non equilibrium in saturated flow experiment was observed after ammonium, nitrate, and chloride data fitting convective dispersi ve model. However presence of physical non equilibrium in un saturated flow experiment has been reflected in tailing for the break through curves of bromide, nitrate, and ammonium. Sorption kinetics led to tailing in the break through curve of phosphorus. F or saturated flow experiments, phosphorus movement was described using sorption coefficient and sorption kinetics parameters determined using fertilizer mixture and potassium chloride (0.01 M KCl). Although it takes long time for movement of phosphorus and ammonium below the water table, movement of nutrients is more pronounced when water table is set at 50 cm depth for Immokalee soil than 30 cm depth. For this experiment, there was no plant and split fertilizer application was not used. Fewer nutrients wil l diffuse below the water table when split application is used in sugarcane fields. Farmers can minimize loss of nutrients below the water table and to perimeter ditches if a much deeper water

PAGE 110

110 table is maintained and adopt supplemental water application T he hypotheses for the column leaching experiment were; (i) reducing distance between water table and Bh horizon through lowering water table from 30 cm to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee so il; and (ii) restoring water table depth after rainfall events will lead to loss of plant available phosphorus and olumn leaching experiments that involved fluctuating water table, have proven th at ammonium and phosphorus diffuse below the water table with diffusion more pronounced when water table is closer to Bh horizon (50 cm depth). Although th ere was no crop in this experiment, ammonium and phosphorus were detected in solutions collected at w ater table.

PAGE 111

111 Table 6 1. Properties of the A horizon used for column leaching experiment Bulk density (g cm 3 ) Particle density (g cm 3 ) Porosity (cm 3 cm 3 ) Pore water v elocity (cm hr 1 ) Pulse (pore volume) 1.49 2.56 0.43 31.93 4.68 Table 6 2. Peak times for bromide, ammonium nitrogen, and Phosphorus Column Location Bromide (days) Ammonium nitrogen (days) Phosphorus (days) Cup A 9 11 20 Water table (30 cm) 20 2 2 37 Water table (50 cm) 59 68 74 Figure 6 1 Design of column leaching experiment for sa turated f low

PAGE 112

112 Figure 6 2 Design of column leaching experiment for unsaturated f low Figure 6 3 Relative concentration (C/C0) as a function of pore volume for saturated flow experiment with A horizon

PAGE 113

113 Figure 6 4 Relative ammonium concentration (C/CO) as a function of pore volume for saturated flow experiment with A horizon Figure 6 5 Relative phosphorus concentration (C/Co ) as a function of pore volume for saturated flow experiment with A horizon

PAGE 114

114 Figure 6 6 Br omide concentration as a function of time for cup A Figure 6 7 Nitrate nitrogen concentration as a function of time for cup A

PAGE 115

115 Figure 6 8. Ammonium nitrogen concentration as a function of time for cup A Figure 6 9 Phosphorus concentration as a function of time for cup A

PAGE 116

116 Figure 6 10 Bromide concentration as a function of time for 30 cm water table depth Figure 6 11 Nitrate nitrogen conce ntration as a function of time for 3 0 cm water table depth

PAGE 117

117 Figure 6 12. Ammonium nitrogen concentration as a function of time for 30 cm water table depth Figure 6 13. Phosphorus concentration as a function of time for 30 cm water table depth

PAGE 118

118 Figure 6 14 Bromide concentration as a function of time for 50 cm water table depth Figure 6 15 Nitrate nitrogen concentration as a function of time for 50 cm water table depth

PAGE 119

119 Figure 6 16. Ammonium nitrogen concentration as a function of time for 50 cm water table depth Figure 6 17. Phosphorus concentration as a function of time for 50 cm water table depth

PAGE 120

120 Figure 6 18. Bromide concentration as a function of time for cup E with water table set at 30cm Figure 6 19. Bromide concentration as a function of time for cup E with water table set at 50 cm

PAGE 121

121 Figure 6 20 Nitrate nitrogen concentration a s a function of time for cup E with water table set at 30cm Figure 6 21 Nitrate nitrogen concentration a s a function of time for cup E with water table set a t 50cm

PAGE 122

122 Figure 6 22. Ammonium nitrogen concentration as a function of time for cup E with water table set at 30cm Figure 6 23. Ammonium nitrogen concentration as a function of time for cup E with water table at 50cm

PAGE 123

123 Figure 6 24 Phosphorus concentration as a function of time for cup E with water table set at 30cm Figure 6 25 Phosphorus concentration as a function of time for cup E with water table set at 50c m

PAGE 124

124 Figure 6 26. Bromide concentration as a function of time for cup Bh with water table set at 30cm Figure 6 27 Bromide concentration a s a function of time for cup Bh with water table set at 50cm

PAGE 125

125 Figure 6 28 Nitrate nitrogen concentration as a function of time for cup Bh with water table set at 30cm Figure 6 29 Nitrate nitrogen concentration as a function of time for cup Bh w ith water table set at 5 0cm

PAGE 126

126 Figure 6 30 Ammonium nitrogen concentration as a function of time for cup Bh with water table set at 30cm Figure 6 3 1 Ammonium nitrogen concentration as a function of time for cup Bh wit h water table set at 50cm

PAGE 127

127 CHAPTER 7 MANAGEMENT OF WATER AND NUTRIENTS WITHIN THE ROOT ZONE USING DRIP IRRIGATION Background Drip irrigation is used to manage moist ure and nutrients within the root zone of plants (Ben Gal and Dudley, 2003 ; Camp, 1998 ; Skag gs et al., 2010 ; Assouline, 2002 ; Thompson et al., 2003 ; Thompson et al., 2000 ). In drip irrigation, the pipes that transmit water are buried in soil and water emitters are left on the surface (Or 1996) The water emitters are placed close to the plants t o achieve higher nutrient solution concentrations within the root zones (Ben Gal and Dudley, 2003). Water release from emitters at set time intervals increases nutrient and water uptake by plants (Ben Gal and Dudley, 2003). Unlike water release rate, amoun t of water applied through drip irrigation and texture determine the spread and direction of water (Skaggs et al., 2010). According to Skaggs et al. ( 2010 ) increases in water content increases spread of wat er content in vertical than the horizontal direc tion. The water application rate was reported to affect distribution of water in the soil profile (Assouline 2002) For example when the application rate of 0.25 L h 1 was used, the upper 0 20 cm layer was wettest (Assouline 2002 ) Zones below the emitt er were saturated when 8 L h 1 application rate was used. The surface drip line spacing has been reported to affect plant yield and drainage losses from the root zone ( Darusman et al., 1997). Decrease in corn yield coupled with increase in internal drainag e from the root zone beyond a drip line spacing of 1.5 m (Darusman et al., 1997). Fertilizers are applied close to sugarcane plants for the roots to access the nutrients easily (Ben Gal and Dudley, 2003). A ccording to Ben Gal and Dudley (2003) improved yie lds and phosphorus use efficiency reflected in plant tissue

PAGE 128

128 phosphorus content were registered when continuous point source irrigation was used. A study by Assouline et al., ( 2002 ) demonstrated improved yields and reduced water loss below the root zone wh en using drip irrigation. Fertilizer mixture (nitrogen, phosphorus, and potassium ) is applied on the soil surface to sugarcane fields in s outhwest Florida Split application of fertilizers is adopted to increase n utrient plant uptake efficiency. In the lys imeter study, fertilizer mixture and three split fertilizer applications were used. Water applied through irrigation and natural rain dissolves the fertilizer mixture allowing nutrients to move into the soil In this study, water was applied to plants thro ugh drip irrigation. Part of dissolved phosphorus will sorb on soil and the rest will leach below root zone. The mass balance for phosphorus can be calculated using the equation, PA = P+PD+PT. Where PA is phosphorus applied is change in total soil phosphorus PD is phosphorus in drainage water, and PT is phosphorus in plant tissue (Ben Gal and Dudley 2003 ). Nutrient concentrations were monitored within the root zone (0 30 cm) and below the root zone (30 45 cm). Nutrient concentrations were also monitored in aboveground tissues (leaves and stalks) to investigate the response of plants to the applied fertilizers. The goal is that plants should adequately utilize the applied nutrients and witness negligible loss of pla n t available nutrients from the root zone The plants will take up nitrogen as ammonium and nitrate ions Ammonium ions are converted into nitrate ion s through nitrification. Under high temperatures and alkaline pH, p art of applied ammonium will volatilize and be lost to the environment (Griggs et al., 2007 ; Horneck et al., 2011 ) Ammonium will compete with potassium for the exchange sites on soil during sorption process. Nitrate can undergo denitrification (conversion of nitrate to gaseous nitrogen) under saturated

PAGE 129

129 conditions Quantifying phosphorus and nitrogen forms (ammonium and nitrate) both in soil and tissues helps to understand the fate of applied phosphorus and nitrogen fertilizers. Unlike the traditional way of fluctuating elevated water table water was managed within the root zone through drip irrigation in this study The aim is to investigate whether water and nutrients needed for sugarcane plants can be managed without necessarily manipulating the water tabl e. The pumping activities of maintaining water table depth after rain fall events can be prevented thus reducing the management costs and potential loss of nutrients in drainage water Use of drip irrigation also reduces un even movement of water during p umping activities thus improving water use Water use efficiency can also be reduced if water moves at flow velocities higher than the normal flow velocity during fluctuating water table. Relatively uniform pore water velocities were reported when drip irr igation technique was used (Jaynes and Rice, 1993). In this study, m ovement of bromide, phosphorus and nitrogen (ammonium and nitrate) within (0 30 cm) and below the root zone (30 45 cm) was monitored as a function of time. The soil moisture content was al so measured as a functio n of depth and time. Nutrient (phosphorus and nitrogen ) concentrations and biomass accumulation in sugarcane leaves and stalks of sugarcane plants were monitored with time. The hypotheses of the study were ; (i) d rip irrigation can b e used to maintain high plant available nutrients within the root zone and minimize nutrients loss out of the root zone; (ii) s ugarcane plants will efficiently respond to applied fertilizers and moisture applied through drip irrigation. The objectives of t he study were; (i) t o assess movement of water (bromide) phosphorus, and nitrogen through the root zone of sugarcane plants ;

PAGE 130

130 (ii) t o monitor biomass accumulation and nutrient concentrations in sugarcane plants as a function of time. Materials and Methods A lysimeter study at Southwest Florida Research and Education Center ( SWFREC ) was designed to investigate the effect of drip irrigation on the inter action of applied fertilizers (nitrogen, phosphorus, and potassium ) with soil and sugarcane plant uptak e. In addition to natural rain, drip irrigation was used to manage water within the root zone of sugarcane plants. The soil phosphorus and nitrogen concentrations were monito red as a function of time. The phosphorus and nitrogen concentrations in the sugarcane plants were monitored as a function of time. Four lysimeters (replicates) were packed with A and E horizons to 80 cm from the soil surface. The dimensions of the lysimeters w ere 3 00 c m by 4 00 c m equipped with moisture sensors and a drainage system for co llecting leachates (Figure 7 1) Two rows of sugarcane were planted in each lysimeter and f ertilizer applications rates at an annual rate of 50 k g P 2 O 5 ha 1 200 k g K 2 O ha 1 and 200 kg N ha 1 were sp lit into three a pplications per year. The masses of fertili zer nutrients applied to each lysimeter at each application were 136.4 g of potassium nitrate 92 g of urea, and 32.6 g of triple super phosphate (46 % P). Sampling of soil and sug arcane plants were carried out to obtain the background nutrient concentrati ons before fertilizer application The fertilizers were surfac e applied on either side of each sugarcane row D rip emitters placed 5 to 7.5 cm below the surface irrigated at rate of 2271 mL per hour per emitter at 30.5 cm spacing The drip em i t ters were pl aced in sugarcane row The volume, 20 29 0 mLs (20.29 L), was applied to each lysimeters daily

PAGE 131

131 Soil was sampled 20 cm from the center of sugarcane rows (irrigated zone ) and 50 cm fr om the center of sugarcane rows (non irrigated zone ) using a 3 cm auger Soil samples were taken at 24 positions (12 were irrigated and 12 non irrigated) at t hree depth increments (0 15 cm, 15 30 cm, and 30 45 cm) Soil sampling period was between 6/6 /11 and 7/19/11. Soil was sampled every two days for two weeks and every week for three weeks. The dates of sampling were 6/6/11, 6/8/11, 6/10/11, 6/12/11, 6/14/11, 6/16/11, 6/18/11, 6/25/11, 7/2/11, and 7/9/11. At each sampling event, two positions for both irrigated and non irrigated zones per lysimeter were sampled. A composite of the two positions was obtained. The total number of samples from four lysimeters at each depth for each category (irrigated and non irrigated) was 40 Two hundred fo rty samples were sampled from irrig ated and non irrigated zones by the end of the experim ent. The soil was put in plastic bags. After sampling soil, the holes were filled with soil to preve nt preferential water movement to the bottom of the lysimeter A subsample was frozen and used for ammonium and nitrate extractions. The rest of the soil wa s air dried in the air drying room set at 45 0 C The air dried soil was u sed for M e hlich 1phosphorus extractions. Wet sub soil samples were extracted for nitrogen (ammonium and nitrate) using potassium chloride ( 2 M KCl ) Four grams of soil were weighed in to 50 mL centrifuge tubes and 40 mL s of potassium chloride ( 2M KCl ) were added. The soil solution was shaken using a reciprocating shaker ( manufacturer: Eberbach ) at low speed for 30 minutes. After filtering, the solutions were analyzed for a mmonium and n itrate using rapid flow method (model: Quikchem 8500 ; manufacturer: LACHAT ). In rapid flow method solutions that are used for color development are prepared separately and two

PAGE 132

132 calibration curves (ammonium and nitrate calibration curves) are obtained from the software. The gravimetric moisture content of all the soil samples was also determined. A known mass of wet soil was weighed into aluminum foil container and put into the oven set at 105 0 C for 24 hours. Gravimetric water content = [Weight of wet soil W eight of oven dry soil]/Weight of oven dry soil. Before analyzing for M ehlich 1phosphorus soil samples were air dried using a dyer room set at 45 0 C and sieved using a 2 mm sieve. The M ehlich 1 solution which is a double acid solution ( 0.05M HCl and 0.025M H 2 SO 4 ) was used to extract for phosphorus Five grams were weighed in to 50 mL centrifuge tubes and 20 mL s of M ehlich 1 solution was added. The soil solution was shaken for 5 minutes at high speed using a reciprocating shaker ( manufacturer : Eberbach ) Th e filtered solution was then analyzed using inductively coupled plasma ( model: Optima 7000DV; manufacturer : Perkin Elmer ) Tissue samples (leaves and stalks) were sampled biweekly for one and half months. Sixty one short plants, thirty five medium plants, and sixty seven tall plants were counted in each lysimeter before the beginning of the experiment. The number of plants was enough for sugarcane plant density. Plant sub samples of each category (short, medium, and tall) were cut. Three short, two medium, and four tall plants were sampled from each lysimeter at every sampling event The leaves were cut from the stalks and mid ribs removed Samples were air dried ( drying room set at 65 o C ) followed by grinding before analyzing for nutrients. The tissue sampl es were analyzed for phosphorus and nitrogen. Total nitrogen was analyzed using carbon nitrogen analyzer ( model: NA 2500 ; manufacturer: CE instruments, Italy ) after grinding air dried leaves and stalks. The ground tissue samples used passed through a 0.5 m m sieve.

PAGE 133

133 Phosphorus w as extracted by ashing method In the ashing method, 0.5 g of ground tissue was weighed into 20 mL vials. The vials with the tissue samples were put into the furnace set at 500 o C for 12 hours. After removing the samples from the furnac e, the sample s were digested by pouring 15 mL s of hydraulic acid (0.5 M of HCl) and left to stand for 24 hours. Solutions were analyzed for phosphorus using inductively coupled plasma ( model : Optima 7000DV; manufacturer: Perkin Elmer ). The weight of above ground biomass was also weighed using a weighing scale after air drying in a dryer room set 55 o C Sugarcane plants were categorized into short, medium, and tall. Subsamples of each category (3 short, 2 medium, and 4 tall) were sampled at every tissue sampl ing event. Results and Discussion Moisture content ( Figure s 7 2 and 7 3 ) and bromide concentration ( Figures 7 4 and 7 5 ) were monitored at 0 15 cm,15 30 cm, and 30 45 cm depths as a function of time for irrigated and non irrigated zones. The majority of sugarcane plant roots h ave been reported to grow between the soil surface and 30 cm (Smith et al., 2005). Thus, the 0 30 cm depth is considered sugarcane ro ot zone The phosphorus and nitrogen concentrations were monitored in soil from i rrigated and non irrigated zones (Figures 7 6 to 7 11). Phosphorus and nitrogen in p lant tissues (stalks and leaves) were monitored as a function of time for short, medium, and tall plants (Figure 7 12 to 7 17) Above ground biomass (weight of dry leaves and stalks) for sh ort, medium, and tall plants was also monitored as a function of time (Figures 7 18 to 7 20). The moisture content (0 15, 15 30 30 45 cm) for irrigated zone increased to 21.5% and decreased after 8 days (0 15 cm), 12% and decreased after 8 days (15 30

PAGE 134

134 cm), and 14.1 % and decreased after 10 days (30 45 cm) (Figure 7 2) Increases in water contents observed at the depth increments for the irrigated zone is due to precipitation and water applied through drip irrigation. Decreases in water contents a t diffe rent depth increments were observed when plant water uptake coupled with evapotranspiration exceeded precipitation and water from drip irrigation (Fisher and Healy, 2008). Although the initial moisture content s followed the order, 3.5% (0 15 cm) < 7 % (15 30 cm) < 11 % (30 45 cm), th e highest moisture content (21.5 %) was observed for 0 15 cm depth due to the placement depth of water em i t ters (Figure 7 2). The increase in initial moisture contents with depth is due to the exposure of first layer (0 15 cm) t o sunlight protecting deeper layers from losing water through evaporation. A similar trend in initial moisture contents (4.3 % for 0 15 cm, 5.9 % for 15 30 cm, and 10.1 % for 30 45 cm) was observed in non irrigated zone (Figure 7 3) A general increase in moisture content was observed at all depths (0 15 cm, 15 30 cm, 30 45 cm) for the non irrigated zone because of lower plant uptake due to reduced root density (Figure 7 4) The distribution of water within the 30 cm was observed by Assouline 2002 when 2 L h 1 water application rate was used. In this study, an application rate of 2.3 L h 1 was used. Phosphorus and nitrogen (ammonium and nitrate ) increase as precipitation and water from drip irrigation dissolves the fertilizer applied along the sugarcane row s and reaches a peak when all the fertilizer has dissolved ( Figures 7 6 and 7 8 ) Decreases in phosphorus and nitrogen concentrations are caused by plant uptake (Figures 7 6 and 7 8). Unlike phosphorus and nitrogen fertilizers, bromide was applied in liqui d phase on soil surface Bromide concentrations decreased as it was taken up by plants. Although

PAGE 135

135 the typical nature of bromide break through curve is symmetrical, tailing in bromide graphs (Figures 7 4 to 7 5) is due to physical non equilibrium created b y mobile and immobile water flow zones (Shinde et al., 2001 ;Fisher and Healy,2008 ). Some water probably moved at a slower rate than the average rate (Fisher and Healy, 2008). In figures, 7 4 and 7 5, a drastic increase was observed for bromide data and ta ils appeared behind the break through curves. The maximum bromide concentration (8.6 g g 1 soil for 0 15 cm, 4.22 g g 1 soil for 15 30 cm,3.82 g g 1 soil for 30 45 cm) were observed after 4, 6, and 4 days respectively (Figures 7 4 ) for irrigated zone. The maximum bromide concentration (9.7 g g 1 soil for 0 15 cm, 10.9 g g 1 soil for 15 30 cm, 4.4 g g 1 soil for 30 45 cm) were observed after 2, 4, and 4 days respectively (Figures 7 5 ) for non irrigated zone. The magnitudes in value of the highest bro mide concentration for all depths were higher for non irrigated zone than irrigated zone (Figures 7 4 to 7 5). This is probably due to water and bromide uptake by the plants in the root zon e and the presence of greater root density that slows water flow. I n a study conducted by Kung ( 1990 ) 53 % of applied bromide was reported to have been taken up by potato plants. In a corn field that was subjected to furrow and sprinkler irrigation, applied bromide concentrations were monitored and percent bromide recove ries in soil ranged from 20 % to 74 % (Butters et al., 2000). High concentrations were observed within the root zone and very low concentrations below the root zone. Percent bromide recoveries in corn plants averaged to 45 % and this was due to corn plant uptake (Butters et al., 2000). When greenhouse and field experiments were conducted with corn used as a crop, percent bromide recover ies in corn under field conditions ranged from 11 38 % and 85 % of bromide was recovered in corn plants under

PAGE 136

136 greenhouse co nditions ( Jemison Jr and Fox 1991 ). Since drip irrigation has been reported to reduce water loss through drainage below the root zone (Assouline, et al., 200 2 ), this can explain the low bromide concentrations below 30 cm depth Although phosphorus values at 0 15 cm and 15 30 cm increased with time ( Figure 7 6 ) values for phosphorus at 30 45 cm were lower than the background phosphorus for the irrigated zone Applied phosphorus was therefore effectively taken up by plants and/or were not leached below the root zone by drip irrigation Increases in phosphorus values for non irrigated zone (Figure 7 7) shows that phosphorus moved from irrigated zone to non irrigated zone through water from natural rain For non irrigated zone, increase in phosphorus values was more pronounced at 0 15 cm depth ( Figure 7 7 ). In the irrigated zone, maximum phosphorus concentration (32.6 mg kg 1 soil) was observed after 10 days for 0 15 cm 10 days for 15 30 cm (27.7 mg k g 1 ) and 8 days for 30 45 cm (20.7 mg k g 1 ) ( Figure 7 6 ) Phosphorus peaked at 30 45 cm depth before 0 15 cm and 15 30 cm depths probably due to absence of roots. In addition to time it takes to dissolve the phosphorus applied in solid forms, phosphorus movement is retarded. Phospho rus concentration increases a s phosphorus is dissolved and later decreases as phosphorus is taken up by plants (Figure 7 6). Phosphorus movement is more retarded than ammonium due to higher sorption capacity of phosphorus by soil. Hydrogen phosphate (H 2 PO 4 ) ions are negatively charge d therefore more attracted by positively charged cations on exchange site than positively charged ammonium. Unlike ammonium, phosphorus has a po tential of sorbing on metals, aluminum, iron and calcium

PAGE 137

137 For the irrigated zone, highest concentrations of amm on ium were observed aft er 4 days for 0 15 cm (25.6 mg k g 1 ) aft er 4 days for 15 30 cm (6.5 mg k g 1 ) and aft er 6 days for 30 45 cm (4.4 mg k g 1 ) (Figure 7 8) The maximum concentrations of n itrate for irrigated zone were observed after 2 days for 0 15 cm (12.5 mg k g 1 ) aft er 4 days for 15 30 cm (5.9 mg k g 1 ) and after 6 days for 30 45 cm (3.1 mg k g 1 ) ( Figure 7 10 ). Higher concentrations were measured for ammonium than nitrate for irrigated and non irrigated zones ( Figure 7 8 and 7 9 ). Unlike nitrate, a mmonium has a potential to react with negatively charged surfaces of organic matter. Nitrate therefore has a higher potential to leach than ammonium. This also exp lains why for irrigated and non irrigated zones, highest ammonium concentrations were observe d at later times than the highest nitrate concentrations (Figures 7 8 to 7 11). Plants might have preferred nitrate to ammonium and mineralization might have increased ammonium concentration within the root zone The high concentrations of ammonium and ni trate within the 30 cm depth compared to 30 45 cm depth show that drip irrigation maintains nutrients with in the root zone and prevents loss of nutrients below t he root zone (Assouline et al., 2002 ). Nitrogen (ammonium and nitrate) concentrations were rep orted to decrease with depth and time after fertigation with drip irrigation ( Khalil 2008 ). The author used an onion crop, emitter discharge rates (2 L h 1 and 4 L h 1 ) and monitored nitrogen as a function of depth (0 15 cm 15 30 cm 30 45 cm and 45 60 cm). Ajdary et al., ( 2007 ) found that nitr ogen (ammonium and nitrate) decreased with depth and time after fertilizer application wit hin the root zone of onion crop with emitter discharge rates of 1L h 1 2.5 L h 1 ,and 4 L h 1

PAGE 138

138 Phosphorus and nitrogen inc reased with time for short, medium and tall plants (Figures 7 12 to 7 17 Tables 7 2 and 7 3 ). F or short plants, tissue phosphorus and nitrogen increased up to 28 days and then decreased due to decreased biomass storage which affects nutrient storage. The percent increase in total nitrogen i n above ground tissues after 28 days was 159 % for short pla nts, 74.7 % for medium plants after 42 days, and 68.7 % for tall plants after 42 days. The percent increase in phosphorus in above ground tissues was 46.9 % aft er 28 days for short plants, 62.2 % after 42 days for medium plants and 118.3 % after 42 days for tall plants. According to Assouline et al., ( 2006 ) drip irrigation improved phosphorus and nitrogen uptake and this was reflected in increased concentrations with time. Drip irrigation also enhanced phosphorus uptake by corn plants (Ben Gal and Dudley, 2003). Increased uptake of nitrogen and above ground biomass was also reported by Thompson et al. ( 2002 ) when drip i rrigation was used to raise broccoli. Bell paper leaf nitrogen and phosphorus increased when the cr o p was subjected to drip irrigation (Assouline et al 2006). Above ground biomass gradually increased with time and drastic increase was observed after 28 days ( Figure s 7 18 to 7 20 ; Table 7 1 ) The p ercent increase in above ground biomass after 42 days was 140 % for tall plants, 86 % for medium plants, and 38 % for short plants. Table 7 1 shows the calculations for total biomass in a lysimeter. Drip irrigation was reported to increase yield in sweet c orn (Assouline, et al., 2002). Assouline, ( 2002 ) reported increase in yield for corn crop with drip irrigation at water appli cation rates of 0.25, 2, and 8 L h 1 Although there was no significant difference in yields for the three water application rates, highest yield in crease was observed with

PAGE 139

139 0.25 L h 1 application rate. Corn yield (Ben Gal and Dudley, 2003) and Bell paper yield (Assouline et al. 2006 ) were also reported to increase with drip irrigation Conclu ding R emarks The placement depth of the water e mi t ters determined the distribution of water over time observed in the irrigated zone. Highest moisture content was observed with 0 15 cm because the water em it t ers were installed approximately 7 cm from soil surface The water distribution for non irrigated zone was determined by rain. Since there was reduced plant uptake water from rain would drain in to lower soil layers. The trend for initial moisture content was 30 45 cm > 15 30 cm > 0 15 cm and moisture distribution was consistently the same throughout the experiment. Although bromide values for the 30 45 cm depth (3.82 g g 1 for irrigated and 4.4 g g 1 for non irrigated zone) did not significantly differ, values for irrigated zone at 0 15 cm (8.6 g g 1 ) were significantly greater than the 15 30 cm depth (4.22 g g 1 ) The pattern in bromide distribution is attribut ed to plant root uptake. Bromide concentrations did not significantly differ for 0 15 cm (9.7 g g 1 soil) and 15 30 cm (10.9 g g 1 soil) depths within the non irrigated zone. When using bromide as a tracer, r esearchers should know that a portion of applied bromide is taken up by plants. In this study, maximum bromide recovery within the root zone was 66 %, thus 34 % was probably absorbed by the sugarcane plants. Although the wa ter emitters were placed close to the sugarcane rows, nutrients can move to non irrigated zones through runoff by natural rain This has been reflected in the increases in nutrient concentrations for non irrigated zones. Increases in phosphorus concentrati ons with time were more pronounced in 0 15 cm than increases

PAGE 140

140 in nitrogen (ammonium and nitrate ) for 0 15 cm, 15 30 cm, and 30 45 cm for non irrigated zone This shows that phosphorus movement was more retarded than nitrogen (ammonium and nitrate ). The tren d in phosphorus and nitrogen concentrations at different depths shows that drip irrigation maintained applied nutrients within the root zone. Increases in biomass and nutrient concentrations of the above ground tissues as a function of time indicated that initial size of the plant influences uptake efficiency, thus larger plants are more efficient in obtaining sufficient soil nutrients. This study has shown that farmers can comfortably use drip irrigation knowing that plants will respond to irrigation techn ique, minimize loss of nutrients below the root zone, and minimize pumping activities during water table depth restoration. The hypotheses of this study were; ( i) d rip irrigation can be used to maintain high plant available nutrients within the root zone a nd minimize nutrients loss out of the root zone; and (ii) s ugarcane plants will efficiently respond to applied fertilizers and moisture applied through drip irrigation. The high concentration s of phosphorus and nitrogen for irrigated zone have proven that drip irrigation can be used as irrigation technique to manage moisture and nutrients within the root zone. Since phosphorus and nitrogen increased in tissues of plants, plants responded to drip irrigation and fertilizers.

PAGE 141

141 Table 7 1. Biomass accumula tion in tissues as a function of time Time(days) Tissue Weight per plant ( k g) Number of plants Total weight in a lysimeter ( k g) 0 S Leaves 0.011 61 0. 70 0 S Stalks 0.008 61 0. 46 0 Plant 0.019 61 1.16 14 S Leaves 0.013 61 0. 81 14 S Stalks 0.021 61 1 27 14 Plant 0.034 61 2.08 28 S Leaves 0.010 61 0. 58 28 S Stalks 0.038 61 2 29 28 Plant 0.047 61 2.87 42 S Leaves 0.010 61 0. 56 42 S Stalks 0.050 61 3 .05 42 Plant 0.060 61 3.61 0 M Leaves 0.017 35 0. 60 0 M Stalks 0.056 35 1.97 0 Plant 0.073 35 2.57 14 M Lea ves 0.023 35 0. 79 14 M Stalks 0.056 35 1.97 14 Plant 0.079 35 2.76 28 M Leaves 0.0 23 35 0.79 28 M Stalks 0.056 35 1.97 28 Plant 0.079 35 2.76 42 M Leaves 0.028 35 0.99 42 M Stalks 0.119 35 4.16 42 Plant 0.147 35 5.15 0 T Leaves 0. 078 67 5.23 0 T Stalks 0.147 67 9.84 0 Plant 0.225 67 15.07 14 T Leaves 0.094 67 6.28 14 T Stalks 0.191 67 12.77 14 Plant 0.284 67 19.05 28 T Leaves 0.100 67 6.7 0 28 T Stalks 0. 231 67 15.50 28 Plant 0.331 67 22.2 0 42 T Leaves 0. 109 67 7.30 42 T Stalks 0.509 67 34.13 42 Plant 0.619 67 41.46 S = Short M = Medium T = Tall

PAGE 142

142 Table 7 2 Phosphorus accumulation in tissues as a function of time Time (days) Tissue Concentration per plant (g k g 1 ) Weight of biomass per plant (kg) Number of plants Total weight of biomass in lysi meter (kg) Total phosphorus in lysimeter ( g ) 0 S Leaves 1.98 0.011 61 0.695 1.38 0 S Stalks 2.26 0.008 61 0.46 4 1.05 0 Plant 0.019 61 1.159 2.43 14 S Leaves 1.96 0.013 61 0.805 1.58 14 S Stalks 2.38 0.021 61 1.26 9 3.02 14 Plant 0.034 61 2.074 4.6 0 28 S Leaves 2.22 0.010 61 0.5 80 1.29 28 S Stalks 2.81 0.038 61 2.28 8 6.43 28 Plant 0.047 61 2.867 7.72 42 S Leaves 1.73 0.010 61 0.5 80 1.00 42 S Stalks 2.07 0.050 61 3.05 0 6.31 42 Plant 0.060 61 3.6 30 7.31 0 M Leaves 2.02 0.01 7 35 0.595 1.20 0 M Stalks 2.3 3 0.056 35 1.97 1 4.59 0 Plant 0.58 0 35 2.56 6 5.79 14 M Leaves 2.11 0.0 23 35 0.79 5 1.68 14 M Stalks 2.45 0.056 35 1.97 1 4.83 14 Plant 0.079 35 2.765 6.51 28 M Leaves 2.00 0.0 23 35 0.79 5 1.59 28 M Stalks 2.42 0.056 35 1.97 1 4.77 28 Plant 0.079 35 2.765 6.36 42 M Leaves 1.78 0.0 28 35 0.994 1.77 42 M Stalks 2.03 0.119 35 4.158 8.44 42 Plant 0.147 35 5.152 10.21 0 T Leaves 1.45 0.0 78 67 5.23 3 7.59 0 T Stalks 1.79 0.147 67 9.842 17.62 0 Plant 0.225 67 15.075 25.21 14 T Leaves 1.53 0.0 94 67 6.28 5 9.62 1 4 T Stalks 1.63 0.191 67 12.77 0 20.82 14 Plant 0.284 67 19.05 5 30.44 28 T Leaves 1.70 0. 100 67 6.700 11.39 28 T Stalks 1.79 0.231 67 15.497 27.74 28 Plant 0.331 67 22.197 39.13 42 T Leaves 1.44 0. 109 67 7.316 10.54 42 T Stalks 1.56 0.509 67 34.129 53.24 42 Plant 0.619 67 41.445 63.78 S = Short M = Medium T = Tall

PAGE 143

143 Table 7 3 Nitrogen accumulation in tissues as a function of time Time (days) Tissue Concentration per plant (g kg 1 ) Weight of biomass per plant (kg) Number of plants Total weight of biomass in lysimeter (kg) Total nitrogen in lysimeter (g) 0 S Leaves 7.5 0.011 61 0.695 5.22 0 S Stalks 2.7 0.008 61 0.464 1.25 0 Plant 0.019 61 1.159 6.47 14 S Leaves 9.2 0.013 61 0.805 7.41 14 S Stalks 6.7 0.021 61 1.269 8.50 14 Plant 0.034 61 2.074 15.91 28 S Leaves 13.9 0.010 61 0.580 8 .06 28 S Stalks 8.4 0.038 61 2.288 19.22 28 Plant 0.047 61 2.867 27.27 42 S Leaves 11.4 0.010 61 0.580 6.61 42 S Stalks 3.8 0.050 61 3.05 0 11.58 42 Plant 0.060 61 3.630 18.19 0 M Leaves 9.1 0.017 35 0.595 5.41 0 M Stalks 5.8 0.056 35 1.971 11.43 0 Plant 0.073 35 2.566 16.84 14 M Leaves 11.2 0.023 35 0.795 8.90 14 M Stalks 5.2 0.056 35 1.971 10.25 14 Plant 0.079 35 2.765 19 .15 28 M Leaves 13.4 0.023 35 0.795 10.65 28 M Stalks 5.6 0.056 35 1.971 11.03 28 Plant 0.079 35 2.765 21.68 42 M Leaves 13.0 0.028 35 0.994 12.92 42 M Stalks 4.0 0.119 35 4.158 16.63 42 Plant 0.147 35 5.152 29.55 0 T Leaves 8.1 0 0.078 67 5.233 42.38 0 T Stalks 3.5 0 0.147 67 9.842 34.45 0 Plant 0.225 67 15.075 76.83 14 T Leaves 10.6 0.094 67 6.285 66.62 14 T Stalks 3.10 0.191 67 12.770 39.59 14 Plant 0.284 67 19.055 106.20 28 T Leaves 11.9 0.10 0 67 6.700 79.73 28 T Stalks 3.40 0.231 67 15.497 52.69 28 Plant 0.331 67 22.197 132 .42 42 T Leaves 9.0 0 0.109 67 7.316 65.85 42 T Stalks 2.3 0 0 .509 67 34.129 7 8.50 42 Plant 0.619 67 41.445 144.34 S = Short M = Medium T = Tall

PAGE 144

144 Figure 7 1. Layout of the lysimeter

PAGE 145

145 Figure 7 2 Moisture content as a function of time and depth for irrigated zone

PAGE 146

146 Figure 7 3 Moisture content as a fun ction of time and depth for non irrigated zone

PAGE 147

147 Figure 7 4 Bromide as a function of time and depth for irrigated zone

PAGE 148

148 Figure 7 5 Bromide as a function of time and depth for n on irriga ted zone

PAGE 149

149 Figure 7 6 P hosphorus as a function of time and depth for irrigated zone

PAGE 150

150 Figure 7 7. Phosphoru s as a function of time for non irrigated zone

PAGE 151

151 Figure 7 8. Ammonium nitrogen as a function of time for irrigated zone

PAGE 152

152 Fig ure 7 9 Ammonium nit rogen as a function of time and depth for non irrigated zone

PAGE 153

153 F igure 7 10 Nitrate nitrogen as a function of time and depth for irrigated zone

PAGE 154

154 Figure 7 11 Nitrate nitrogen a s a function of time and depth for non irrigated z on e

PAGE 155

155 Fi gure 7 1 2 Phosphoru s in in tissues of short plants as a function of time Figure 7 1 3 Phosphorus in tissues of medium plants as a function of time

PAGE 156

156 Figure 7 1 4 Phosphorus in tissues of tall plants as a function of time Figure 7 15. Total nitrogen in tissues of short p lants as a function of time

PAGE 157

157 Figure 7 16 Total nitrogen in tissues of medium plants as a function of time Figure 7 17 Total nitrogen in tissues of tall plants as a function of time

PAGE 158

158 Figure 7 18 Biomass accumulation for short plants as a function of time Figure 7 1 9 Biomass accumulation for medium plants as a function of time

PAGE 159

159 Figure 7 20 Biomass accumulation for tall plants as a function of time

PAGE 160

160 CHAPTER 8 CALIBRATION AND VALI DATION OF HYDRUS 1D Background Results from modeling with simu lation environments m ay be inaccurate if model parameters are not determined correctly. Parameters that are fed in the model should be experimentally determined using soil samples from the area of interest. Phosphorus sorption coefficients (K D ) that descri be partitioning of phosphorus between liquid and solid phases should also be determined using appropriate electrolyte that is similar to ionic strength of the system. Phosphorus movement was modeled using Hydrus 1D to identify effect of drip irrigation on management of phosphorus within the root zone of sugarcane plants and leaching of phosphorus below the root zone. Since variability of A horizon depths have been identified (Chapter 3), A horizon depth was varied in Hydrus 1D to investigate depths effect o n leaching of phosphorus. Phosphorus movement in a system that involved elevated water table was modeled using Hydrus 1D. Bulk density saturated hydraulic conductivity ( Ksat ) values, constants from moisture release curves ( m and n), sorption kinetics parameters, linear ized sorption coefficients (K D ) saturated water content and residual moisture content determined experimentally (C hapter 4 and 5 ) were used to calibrate Hydrus 1D Linear ized sorption coefficients (K D ), was used to describe the partitio ning of phosphorus between liquid and solid phases (Simunek, 1999). T he linearized sorption coefficient values calculated from sorption data determined exp erimentally using potassium chloride ( 0.01M KCl ) fertilizer mixture calcium chloride ( 0.005 M CaCl 2 ) deionized water, and Florida rain (C hapter 5 ) were used to model ph osphorus movement in a lysimeter In th e lysimeter study, d rip

PAGE 161

161 irrigation was used as irrigation technique to meet sugarcane plant water requirements. The different linearized sorption c oefficient values were also used to model phosphorus movement in a column where elevated water table was fixed. The assumptions adopted in Hydrus simulation environment to describe nutrient movement include; (i) solutes are assumed to exist in liqui d, soli d, and gaseous phase ; (ii) the nonlinear non equilibrium equations explain interactions between solid and li quid phase; (iii) s olutes are transported in liquid phase through convection, diffusion, and dispersion ; and (iv) there can be differences in decay and production processes in each phase (Simunek et al., 1999) In Hydrus water movement is modeled using a h ydraulic model, VanGenuchten Mualem (van Genuchten, 1980 ). After determining moisture release curves, the p arameters Genuchten, 1980) are estimated The equation, m + ( + ( n ] m is used to calculate water content. Where h is the pressure head and m=1 1/n. Modeli ng a system that involves a crop requires a modeler to account for plant uptake and evapotranspiration (Simunek and Hopmans, 2009). Tracers like bromide are modeled using equilibrium model and phosphorus sorption is characterized using two site model that describes instantaneous sorption and sorption kinetics In this study, Hydrus 1D was calibrated using parameters determined experimentally and o ne dimensional movement of water, bromide, phosphorus, and nitrogen (ammonium and nitrate) in a lysimeter and co lumn w as modeled. Out put results for phosphorus when using linearized sorption coefficient (0.01M KCl), linearized sorption coefficient (fertilizer mixture) linearized sorption coefficient

PAGE 162

162 (0.005 M CaCl2), linearized sorption coefficient (deionized wa ter), and linearized sorption coefficient ( Florida rain) were compared using root mean square errors and mean absolute errors The bromide, phosphorus and nitrogen (ammonium and nitrate) data sets from the lysimeter study and column leaching experiment re sults (Chapter 6 and Chapter 7 ) were used as the validation data set s for Hyd rus 1 D. In this study, it is hypothesized that phosphoru s leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolyte s are used to model phosphorus movement. The objective s w ere t o (i ) calibrate Hydrus 1D using the saturated hydraulic conductivity ( Ksat values ) moisture release curve constants ( n, and m ), saturated moisture con tent, residual moisture content, and bulk density, linearized sorption coefficient values, and sorption kinetics parameters (k 2 and F) and (ii) to validate Hydrus 1D using lysimeter study data and column data. Materials and Methods The parameters used to calibrate Hydrus 1D were; saturated hydraulic conductivity ( Ksat ) values, saturated water content, residual water content, bulk density, constants ( m and n) calculated from moisture release curve, and sorption paramet ers (K D values and sorption kinetic parameters) (Table s 8 1 and 8 2 ). Validation followed calibration of Hydrus 1D using nutrient concentrations from column leaching experiment and lysimeter study. Hydrus 1D was used assuming that in a column experiment an d at a discharge rate of 2.3 L h 1 in a lysimeter, water is moving in vertical direction (Skaggs et al., 2010 ; Vrugt et al. 2001 ) B romide, ammonium, phosphorus and water movement were modeled from the soil surface to the end of the A horizon for the col umn leaching experiment In a column leaching experiment, moisture was

PAGE 163

163 modeled at a point, 8 cm from soil surface since moisture sensor was installed at that depth. The water table depth was kept at 30 cm (18 cm for A horizon and 12 cm for E horizon) for t he first six weeks. The rain that was added to the columns was obtained by calculating the averages of the daily rainfall received in Immokalee for ten years (Figure 8 1) For the lysimeter study, water, bromide, nitrogen (ammonium and nitrate) and phospho rus were modeled for two months (period of sampling). Figure 8 11 shows the amount of rain received in Immokalee during the lysimeter study period. Phosphorus, nitrogen, bromide, and water were modeled upto 45 cm. The upper boundary condition for water flow was variable head/flux and lower boundary condition f or water flow was free drainage. The upper boundary condition for simulating phosphorus and bromide movement was concentration flux BC and lower boundary condition was concentration flux BC. The two si te non equilibrium mo del (Nkedi K izza et al., 2006 ; Simunek and van Genuchten, 2008 ) is used to describe phosphorus sorption behavior by Hydrus model I nstantaneous sorption (type 1site) and time dependent (type 2 site) are sites depicted by the two site model For type 1 site, S 1 =FK D C and for type 2 site, dS 2 F)K D C S 2 ]. Where F is fraction of type I sites, K D is linear ized adsorption coefficient ( mL /g), and order coefficient (s 1 ), S 1 is solute concentration in to solid phase for type I, and S 2 is solute concentration into solid phase for type II. T he two site non equilibrium s orption is used to describe non equilibrium adsorption desorption reactions. Convection dispersion equations model phosphorus transport (Renduo, 2000). One site chemical non equilibrium (kinetic adsorption reaction) w as used to mode l ammonium transport in soil (Jellali et al., 2010 ; Simunek and van Genuchten 2008 ). Equilibrium

PAGE 164

164 model was used to model bromide movement since bromide is used as a tracer 1931 ) is used to model water flow an d it contains a sink term that is a representation of the volume of water removed per unit time from a unit volume of soil due to plant uptake. Graphical editor in Hydrus 1D was set to have two soil layers1 (A horizon) and 2 (E horizon) for column leaching experiment and t hree soil layers (0 15 cm, 15 30 cm and 30 45 cm ) for lysimeter study For a column leaching experiment, 160 mg 177 mg and 38.5 mg were used as initial masses for modeling bromide ammonium and phosphorus movement respectively. Water b romide, and phosphorus movement w ere simulated up to six weeks when water table was set at 30 cm from soil surface. The linearized sorption coefficient ( K D ) values, 1.12 mL g 1 1.22 mL g 1 1. 42 mL g 1 0.6 mL g 1 and 0.6 mL g 1 determined using fertilizer mixture, potassium chloride ( 0.01 M KCl ) calcium chloride ( 0.005 M CaCl 2 ) Florida rain, and deionized water respectively were used for modeling phosphorus movement in a column leaching experiment For the un calibrated model, 1.05 mL g 1 was used to mod el phosphorus movement. The equation, K D = K f C max N 1 was used to calculate phosphorus sorption from the Freundlich isotherms The maximum concentration (C max ) from the validation data set is used in the equation. The fertilizer mixture K D (0.2 mL g 1 ) wa s used to model ammonium movement (Table 8 2). Unlike in validating Hydrus 1D model with column leaching experiment, water and nutrient uptake was a component when validating with lysimeter study The root zone is 30 cm from the surface. The linearized so rption coefficient values for modeling

PAGE 165

165 phosphorus movement in are shown in Tables 8 4 and 8 5. The linearized sorption coefficient value that was used to model ammonium movement was 0.2 mL g 1 The root water uptake model used was Feddes water uptake reduc tion m odel ( Simunek and Hopmans 2009 ). The critical stress index for water and solute uptake used was 1. The active solute uptake was used as ro o t solute uptake model and a value 0.5 was us ed as Michaelis Menten constant (Simunek and Hopmans 2009 ). The Feddes parameters, PO, POpt, P2H, P2L, P3, r2H, and r2L used were 0.1, 4, 5, 790, 1000, 0.5 cm day 1 and 0.1 cm d ay 1 respectively were used (Simunek and Hopmans 2009 ) Where PO is the value of the pressure head below which roots to extract water from the soil; POpt is the value of the pressure head below which roots extract water at the maximum possible rate; P2H is the value of the limiting pressure head below which roots can nolonger extract water at the maximum possible rate (assuming a potential tra nspiration rate of r2H); P2L is the value of the limiting pressure head below which roots can nolonger extract water at the maximum rate (assuming a potential transpiration rate of r2L); P3 is the value of the pressure h ead below which root water uptake ceas es (u sually at a wilting point); r2H is the potential transpiration rat e set at 0. 5 cm day 1 ;and r2L is the potential t ranspiration rate set at 0.1 cm day 1 The A horizon depths were adjusted to 30 cm, 40 cm, and 45 cm to identify effect of varying horizo n depths in sugarcane fields. The A horizon depths varied in the two sugarcane fields with Margate and Immokalee soil series. The step s for modeling water, bromide, phosphorus and nitrogen (ammonium and nitrate) are shown in figures C 1 and C 15.

PAGE 166

166 Root m ean square error (RMSE) was used to compare experiment data with simulation results from Hydrus 1D. w here n = number of pairs, O = observed values, S = simulated values, n = total number of observations and RMSE = root mean square error. Mean absolute error (MAE) was also used to assess how good Hydrus 1D predicts phosphorus movement. Hydrus predicts nutrient movement. Mean absolute error (MAE) = Where S = simulated values, O = observed values, and n = numbe r of pairs. Nutrient data that was obtained using Mehlich 1 extractions were considered to be adsorbed plus solution nutrient concentrations. Solution concentration of nutrients were calculated and compared with Hydrus results since simulation output is so lution concentrations. Results and Discussion After the calibration process, nutrient (phosphorus and nitrogen), water, and bromide data from column leaching experiment and lysimeter study were used to validate Hydrus 1D. The patterns of the measured and t he simulated moisture values for the column leaching experiment are shown in F igure 8 2 .The root mean square error and mean absolute error were 0.081 and 0.012 respectively of experimental and simulated values. The increase and decrease in water content r eflects the pattern for added rain (Fig ure 8 1). The low values of root mean square error and mean square error show that the measured moisture content compared well with the simulated results. The root mean square error, mean absolute errors, and regressi on coefficients (R 2 ) values for bromide, ammonium, and phosphorus for column data are shown in

PAGE 167

167 Table 8 3. The regression coefficient f or bromide (0.97) and ammonium (0.95) show strong agreement between experimental data and model results (Table 8 3; Figure s 8 3 and 8 4) The high root mean square error with linearized sorption coefficient using de ionized water and Florida rain (7.4) show over prediction of phosphorus movement (Table 8 3; Figures 8 9 and 8 10) The root mean square error values for lineariz ed sorption coefficient values obtained using potassium chloride ( 0.01 M KCl ) fertilizer mixture, and calcium chloride ( 0.005 M CaCl 2 ) were 3.40, 3.10, and 2.70 respectively (Table 8 3). The root mean square error value compare well with uncalibrated line arized sorption coefficient value, 2.65 (Table 8 3). Figures 8 3, 8 4, 8 5, 8 6, 8 7 8 8 8 9, and 8 10 for bromide, ammonium, and phosphorus show the patterns of experimental data and simu lated values. Values increases with increase in added rain, peak, and decrease as more rain is added. Although linearized sorption coefficient using calcium chloride ( 0.005 M CaCl 2 ) was the highest value, the root mean square error values from the predictions were closer to predictions when linearized sorption coefficie nt values using potassium chloride ( 0.0 1 M KCl ) and fertilizer mixtrure were used Modelers dealing with a system that involves fe rtilizer mixture can use linearized sorption coefficient values from potassium chloride ( 0.01M KCl ) to obtain similar predicti ons. Unlike the literature higher molarities of calcium chloride, 0.1M (Ru bio et al., 2008) and 0.01M (Dou et al., 2009), molarity value, 0.005 M, was used to obtain linearized sorption coefficient value used for modeling phosphorus movement. Probably redu ction in concentration of calcium chloride and low total carbon of A horizon does not result in greater sorption of phosphorus.

PAGE 168

168 Phosphorus and ammonium concentrations over time ( Figures 8 9 and 8 10 ) were steeper before 12 days and this is attributed to a ctive nutrient uptake after 12 days that leads to reduction in solution nutrient concentrations (Simunek and Hopmans 2009 ). Unlike in a controlled system (column leaching experiment with fluctuating water table), root mean square errors and mean absolute errors were greater (Table 8 4 ) for the lysimeter due to additional processes like plant uptake. Hydrus 1D was validated with lysimeter data (bromide, nitrogen, and phosphorus in Chapter 7 ) for 0 15 cm, 15 30 cm and 30 45 cm depths (Figures 8 12 to 8 52 ). The regression coefficient f or bromide ( 0.97), nitrate ( 0.97), and ammonium (R 2 = 0.94) show that Hydrus 1D was validated with the lysimeter data (0 15 cm) (Table 8 4 and 8 5). Regression coefficient values greater than 0.5 were also obtained for other de pths. For the irrigated zone, root mean square error for phosphorus results of for example 0 15 cm depth followed the order potassium chloride ( 0.01 M ) ( 3.0 ) < fertilizer mixture ( 3.6 ) < calcium chloride ( 0.005 M ) ( 5 ) < deionized water and Florida rain ( 1 9.9 ) (Table 8 4 ). Generally, a similar trend was observed for 15 30 cm and 30 45 cm depths. The root mean square error values for the non irrigated zone (0 15 cm) also followed the same trend. The uncalibrated linearized sorption coefficient value, 1.08 mL g 1 was closer to the linearized sorption coefficient using fertilizer mixture and potassium chloride ( 0.0 1 M KCl ) Unlike the column leaching experiment, where linearized sorption coefficient for ( 0.005M CaCl 2 ) reg istered similar predictions as linearized sorption coefficient using fertilizer mixture and potassium chloride ( 0.01 M KCl ) root mean square error values for linearized sorption coefficient using calcium chloride ( 0.005M CaCl 2 ) are relatively greater. From validating Hydrus 1D with column leachi ng experiment data and lysimeter data modelers can model

PAGE 169

169 phosphorus movement in sugarcane fields using either linearized sorption coefficient using fertilizer mixture or linearized sorption coefficient using potassium chloride ( 0.01 M KCl ) When the A hor izon depths were varied in the graphical editor concentrations in leachates decreased with increase in A horizon depth (Figure 53) Modelers can therefore model phosphorus movement using Hydrus 1D for the system s that involves dr ip irrigation and elevated water table coupled with surface fertilizer ( phosphorus potassium, and nitrogen ) application. Conclu ding R emarks Hydrus 1D can be calibrated with c olumn leaching experiments that involves unsaturated flow and fixed water table depth. When modeling water and solute transport, the model parameters should be determined using the soil samples packed in the columns. The linearized sorption coefficient values determined using potassium chloride (0.01M KCl ) and fertilizer mixture can be used for modeling phospho rus movement when phosphorus is applied to the field as fertilizer mixture The sorption kinetics para meters ( F and k 2 ) varied for potassium chloride ( 0.01M KCl ) and fertilizer mixture prepared in rain. However results from simulation with Hydrus 1D were not affecte d by the differences in values Since water, bromide and phosphorus compared well with the Hydrus 1D simulated values, the system that involves a fixed water table depth can be modeled using Hydrus 1D. Modeling phosphorus movement using varying phosphorus sorption coefficient (K D ) determined using different electrolytes over predicts and under predicts phosphorus leaching. Sorption coefficients determined using deionized water and Florida rain over predicts phosphorus leaching. Under

PAGE 170

170 predicting p hosphorus leaching was observed with sorption coefficients determined using calcium chloride (0.005 M CaCl 2 ). Hydrus 1D was validated using lysimeter parameters and soil moisture, bromide, phosphorus and ammonium data set. The sorption coefficient determ ined using fertilizer mixture was used to model phosphorus movement. Modelers can model changes in nutrient concentrations and water movement for a system that involves sugarcane raised using drip irrigation. However in this study, a lysimeter of dimension s 3 00 c m by 4 00 c m was used where two rows of sugarcane plants were raised. For this experiment, it was hypothesized that phosphorus leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolytes are used to model phosphorus movement. Results from simulating phosphorus movement with different sorption coefficients (K D ) determined using different electrolytes have proven that linearized sorption coefficients for deionized water and Florida rain over predicts leaching. Phosphorus leaching was under predicted with sorption coefficients for calcium chloride (0.005 M CaCl 2 ). Results from modeling phosphorus movement with sorption coefficients determined using potassium chloride (0.01 M KCl) and fertilize r mixture compared well with experimental data.

PAGE 171

171 Table 8 1 Model i nputs for m odel c alibration : m odeling w ater m ovement Model Input Value Saturated water content for A horizon 0.294 Saturated water content for A horizon 0.262 Residual water co 0.057 Residual water content for A horizon 0.036 Saturated hydraulic conductivity for A horizon (cm hr 1 ) 21.4 Saturated hydraulic conductivity for A horizon ( cm hr 1 ) 6 .0 retention function) 0.036 E horizon(parameter in the soil water retention function) 0.036 Bulk density for A horizon (g cm 3 ) 1.52 Bulk density for E horizon ( g cm 3 1.48 n for A horizon(obtained from soil water retention function 1.89 n for E horizon(obtained from soil water retention function 1. 56 l for A horizon (tortuosity parameter in conductivity function) 0.5 l for E horizon (tortuosity parameter in conductivity function 0.5 Table 8 2 Model i nputs for m odel validation : Lys imeter study column leaching experiment Value K D for A ho rizon (fertilizer mixture) for phosphorus 1.0 mL g 1 K D for A horizon ( 0.01 M KCl ) for phosphorus 1.1 mL g 1 K D for A horizon ( 0.005 M CaCl 2 ) for phosphorus 1.3 mL g 1 K D for A horizon ( deioniz ed water ) for phosphorus 0.6 mL g 1 K D for A horizon ( Florida rain ) for phosphorus 0.6 mL g 1 K D for A horizon (fertilizer mixture) for ammonium 0.2 mL g 1 Dis p ersivity coefficient for A horizon 0. 22 Dispersivity coefficient for E horizon 0. 10 Fractio n of type 1 sites for A horizon (f ertilizer mixture 0.01M KCl ) for phosphorus 0. 6 Fraction of t ype 1 sites for A horizon for bromide 1 1 between mobile and immobile liquid regions) for phosphorus 0.07 E horizon(mass transfer coefficient for solute exchange b etween mobile and immobile liquid regions) for phosphorus 0.07

PAGE 172

172 Table 8 3. Root mean square error and mean absolute error: column leaching experiment Depth Electrolyte for K D RMSE MAE R 2 cupA Br 9 .00 7 .00 0.97 cupA N Fertilizer mixture 6 .00 3.4 0 0 .95 CupA P Uncalibrated 2.65 2.40 0.66 CupA P Fertilizer mixture 3.1 0 2.5 0 0.66 cupA P 0.01 M KCl 3.4 0 2.8 0 0.65 cupA P 0.005 M CaCl 2 2.7 0 2.5 0 0.66 cupA P Deionized water 7.4 0 6.8 0 0.63 cupA P Simulated Florida rain 7.4 0 6.8 0 0.63 P = Phosphorus, N = A mmonium Br = Bromide RMSE = Root mean square error, and MAE = Mean absolute error. Table 8 4. Root mean square error and mean absolute error: irrigated zone Depth Electrolyte for K D K D ( mL g 1 ) Nutrient RMSE MAE R 2 0 15 cm Bromide 0.24 0.34 0.97 15 30 cm Bromide 0.58 0.52 0.81 0 15 cm Nitrate 1.96 1.23 0.97 15 30 cm Nitrate 2.11 1.46 0.96 0 15 cm Fertilizer mixture 0.2 Ammonium 3.9 3.6 0.94 15 30 cm Fertilizer mixture 0.2 Ammonium 1.1 0.84 0.62 30 45 cm Fertilizer mixture 0.2 Ammonium 0.30 0.28 0.88 0 15 cm Uncalibrated 1.08 Phosphorus 2.96 2.49 0.66 15 30 cm Uncalibrated 1.08 Phosphorus 3.2 2.79 0.56 30 45 cm Uncalibrated 1.08 Phosphorus 0.86 0.78 0.95 0 15 cm Fertilizer mixture 1.0 Phosphorus 3.6 2.9 0.66 15 30 cm Fertilizer m ixture 1.0 Phosphorus 2.3 2.1 0.49 30 45 cm Fertilizer mixture 1.0 Phosphorus 1.07 0. 97 0.95 0 15 cm 0.01 M KCl 1.1 Phosphorus 3.0 2.5 0.67 15 30 cm 0.01 M KCl 1.1 Phosphorus 2.6 2.2 0.48 30 45 cm 0.01 M KCl 1.1 Phosphorus 0.76 0.68 0.95 0 15 cm 0 .005 M CaCl 2 1.3 Phosphorus 5 4.5 0.67 15 30 cm 0.005 M CaCl 2 1.3 Phosphorus 3.6 3.1 0.47 30 45 cm 0.005 M CaCl 2 1.3 Phosphorus 0.68 0.48 0.95 0 15 cm Deionized water 0.6 Phosphorus 19.9 17.0 0.58 15 30 cm Deionized water 0.6 Phosphorus 7.6 6.3 0.56 30 45 cm Deionized water 0.6 Phosphorus 3.54 3.36 0.88 0 15 cm Florida rain 0. 6 Phosphorus 19.9 17.0 0.58 15 30 cm Florida rain 0. 6 Phosphorus 7.6 6.3 0.56 30 45 cm Florida rain 0. 6 Phosphorus 3.54 3.36 0.88

PAGE 173

173 Table 8 5. Root mean square error and mean absolute error: non irrigated zone Depth Electrolyte for K D K D ( mL g 1 ) Nutrient RMSE MAE R 2 0 15 cm Bromide 0.52 0.43 0.95 15 30 cm Bromide 0.41 0.36 0.54 0 15 cm Fertilizer mixture Nitrate 3.11 2.55 0.77 15 30 cm Fertilizer mixture Nitrate 2.89 1.75 0.77 0 15 cm Fertilizer mixture 0.2 Ammonium 0.17 0.14 0. 98 15 30 cm Fertilizer mixture 0.2 Ammonium 0.20 0.18 0.75 0 15 cm Fertilizer mixture 1.2 Phosphorus 1.62 1.41 0.73 15 30 cm Fertilizer mixture 1.2 Phosphorus 1.20 1.02 0.52 0 15 c m 0.01 M KCl 1.2 Phosphorus 1.62 1.41 0.73 15 30 cm 0.01 M KCl 1.2 Phosphorus 1.20 1.02 0.52 0 15 cm 0.005 M CaCl 2 1.6 Phosphorus 2.59 2.31 0.57 15 30 cm 0.005 M CaCl 2 1.6 Phosphorus 1.50 1.26 0.38 0 15 cm Deionized water 0.6 Phosphorus 4.49 4.39 0.7 0 15 30 cm Deionized water 0.6 Phosphorus 1.60 1.13 0.20 0 15 cm Florida rain 0.6 Phosphorus 4.49 4.39 0.70 15 30 cm Florida rain 0.6 Phosphorus 1.60 1.13 0.20

PAGE 174

174 Figure 8 1. Rain added to c olumns as a function of t ime for 30 cm water depth Source: fawn.ifas.ufl.edu Figure 8 2. Experimental and simulated moisture as function of time for a sensor at 8 cm depth

PAGE 175

175 Figure 8 3 Experimental and simulated bromide as function of time for cup A Figure 8 4 Experimental and simulated ammonium as function of time for cup A

PAGE 176

176 Figure 8 5 Experimental and simulated phosphorus as fu nction of time for cup A using linearized sorption coefficient (uncalibrated) Figure 8 6 Experimental and simulated phosphorus as function of time for cup A using li nearized sorption coefficient (fertilizer mixture)

PAGE 177

177 Figure 8 7 Experimental and simulated phosphorus as function of time for cup A using linearized sorption coefficient (0.01 M KCl) Figure 8 8 Experimental and simulated phosphorus as function of time for cup A using linearized sorption coefficient (0.005M CaCl 2 )

PAGE 178

178 Figure 8 9 Experimental and simulated phosphorus as function of time for cup A using linearized sorption coefficient (Deionized water) Figure 8 1 0 Experimental and simulated phosphorus as function of time for cup A using linearized sorption coefficient (Florida rain)

PAGE 179

179 Figure 8 11 Natural rain received in Immokalee (6/6/11 7/9/11 ) Source: fawn.ifas.ufl.edu Figure 8 12 Experimental and simulated bromide as function of time for irrig ated zone

PAGE 180

180 Figure 8 13 Experimental and simulated bromide as function of time for non irrigated zone Figure 8 14 Experimental and simulated bromide as function of time for irrigated zon e

PAGE 181

181 Figure 8 15 Experimental and simulated bromide as function of time for non irrigated zone Figure 8 16 Experimental and simulated nitrate as function of time for irrigated zone

PAGE 182

182 Figure 8 1 7. Experimental and simulated nitrate as function of time for non irrigated zone Figure 8 18 Experimental and simulated ni trate as function of time for irrigated zone

PAGE 183

183 Figure 8 19 Experimental and simulated nitrate as function of time for non irrigated zone Figure 8 20 Experimental and simulated ammonium as function of time for irrigated zone

PAGE 184

184 Figure 8 21 Experimenta l and simulated ammonium as function of time for non irrigated zone Figure 8 22 Experimental and simulated ammonium as function of time for irrigated zone

PAGE 185

185 Figure 8 23 Experimental and simulated ammonium as function of time for non irrigated zone Figure 8 24. Experimental and simulated ammonium as function of time for irrigated zone

PAGE 186

186 Figure 8 25 Experimental and simulated phosphorus as function of time for irrigated zone Figure 8 26 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 187

187 Figure 8 27 Experimental and simulated phosphorus as function of time for irrigated zone Figure 8 2 8. Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 188

188 Figure 8 29 Experimental and simulated phosphorus as function of time for non irrigated zone Figure 8 30 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 189

189 Figure 8 31 Experimental and simulated phosphorus as function of time for non irrigated zone Figure 8 32 Experimen tal and simulated phosphorus as function of time for irrigated zone

PAGE 190

190 Figure 8 33 Experimental and simulated phosphorus as function of time for irrigated zone Figure 8 34 Experimental and simulated phosphorus as function of time for non irrigated zone

PAGE 191

191 Figure 8 35 E xperimental and simulated phosphorus as function of time for irrigated zone Figure 8 36 Experimental and simulated phosphorus as function of time for non irrigated zone

PAGE 192

192 Figure 8 37 Experimental and simulated phosphorus as function o f time for irrigated zone Figure 8 38 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 193

193 Figure 8 39 Experimental and simulated phosphorus as function of time for non irrigated zone Figure 8 40 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 194

194 Figure 8 41 Experimental and simulated phosphorus as function of time for non irrigated zone Figure 8 42 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 195

195 Figure 8 43 Ex perimental and simulated phosphorus as function of time for iriigated zone Figure 8 44 Experimental and simulated phosphorus as function of time for non irrigated zone

PAGE 196

196 Figure 8 45 Experimental and simulated phosphorus as function of time for irrigate d zone Figure 8 46 Experimental and simulated phosphorus as function of time for non irrigated zone

PAGE 197

197 Figure 8 47 Experimental and simulated phosphorus as function of time for irrigated zone Figure 8 4 8. Experimental and simulated phosphorus as funct ion of time for irrigated zone

PAGE 198

198 Figure 8 49 Experimental and simulated phosphorus as function of time for non irrigated zone Figure 8 50 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 199

199 Figure 8 51 Experimental and simu lated phosphorus as function of time for non irrigated zone Figure 8 52 Experimental and simulated phosphorus as function of time for irrigated zone

PAGE 200

200 Figure 8 53 Simulated phosphorus in leachate as a function of horizon depths

PAGE 201

201 CHAPTER 9 CONCLUSIO NS Studying variability of soil characteristics in sugarcane fields is important. This helps in demarcating the fields according to the distribution of soil characteristics Nutrient movement and yield distribution can therefore be studied according to the distribution of soil characteristics This study has revealed that there are variations in A horizon depths in sugarcane fields with Immokalee and Margate soil series. The variability in A horizon depth helps farmers to know that different but predictable patterns of phosphorus and nitrogen losses occur below the root zone Unlike A and E horizons that exists in Margate and Immokalee soils the underlying horizons (Bh and Bw ) affect the distribution of total carbon, oxalate extractable aluminum oxalate ex tractable iron and pH. The limestone le d to high measured pH values (maximum pH value of 7.9) at 60 90 cm depth for Margate field. Sampling Bh horizon at 60 90 cm depth le d to high observed total carbon values (maximu m total carbon value of 77.4 g k g 1 s oil) for Immokalee field. Although the same fertilizer application rates are adopted by the farmers, the magnitude of tota l carbon values, iron, aluminum and pH will affect the sorption of phosphorus and nitrogen The leaching of nutrients below the root zone (0 30cm) will also depend on the capacity of surface horizon to sorb nutrients. After observing the variability of soil characteristics in sugarcane fields with different soil orders differences in patterns of phosphorus and nitrogen accumulations oc cur within the root zone. In future, farmers should study distribution of soil characteristics using other soil series that are used for sugarcane production. Before studying distribution of soil characteristics in two sugarcane fie lds, it was hypothesized that, soil characteristics for two dominant sandy soils used for sugarcane production in south Florida will vary

PAGE 202

202 spatialy and with depth and this will lead to different patterns of phosphorus accumulations. Variability of soil characteristics was observed in two sugarcane fields and correlation coefficients showed that soil characteristic that determined most distribution of total phosphorus was total carbon. Although constants calculated and residual mo isture content were not significantly different, saturated hydraulic conductivity (Ksat) values for A E, and Bh horizons of Immokalee soil were sig nificantly different. The h content, residual water content, and saturated hydraulic conductivity (Ksat), values for A E, and Bh horizons of sandy Immokalee soil will be different when determined e xperimentally, was proven when observed saturated hydraulic conductivity values for A, E, a nd Bh horizons were significantly different. Different electrolytes used by researchers result in different values of linearized sorption coefficient values for the same soil. The supporting electrolyte that should be used to mimic the ionic strength of fertilizer mixt ure (nitrogen phosphorus and potassium ) used by sugarcane growers is potassium chloride ( 0.01M KCl ) The initial phosphorus solutions should be prepared with the ionic strength close to liquid phases of the system of interest Changes i n s orption kinetics parameters ( k 2 and F ) when potassium chloride ( 0.01M KCl ) and fertilizer mixture are used do not affect modeling results. Farmers who model nutrient movement in sugarcane fields should use potassium chloride (0.01 M KCl) and soil samples from sugarcane fields of interest to characterize sorption equilibria and sorption kinetics of phosphorus. After observing significant differences in phosphorus sorption coefficient determined using different

PAGE 203

203 electrolytes, the hypothesis, phosphorus sorpti on coefficients (K D ) when determined using different supporting electrolytes (0.01M KCl, 0.005M CaCl 2 simulated Florida rain, deionized water, and fertilizer mixture) will significantly differ was proven. Ammonium, nitrate, and chloride data from satura ted flow experiment fit convective dispersive model and this shows absence of physical non equilibrium. The tailing in the break through curves of bromide and nitrate data shows presence of physical non equilibrium in unsaturated flow experiment. Tailing o f phosphorus data from saturated and unsaturated flow experiment shows presence of sorption kinetics. After lowering the water table to 50 cm depth (water table closer to Bh horizon), the concentration gradient was increased which led to increased diffusio n of phosphorus below the water table. Phosphorus diffusion below the water table increases with decrease in distance between the water table and Bh horizon Un like Immokalee soil with Bh horizon after A and E horizons Margate will not exhibit the same p attern in phosphorus movement. The Bw horizon does not sorb as much phosphorus as Bh horizon. Unlike Bw, Bh horizon w ill sorb significant amount of phosphorus due to organic ally complexed aluminum and iron Since it takes long time to observe high concent rations of phosphorus and ammonium below water table for Immokalee soil, a lternatives like drip irrigation should be used while maintaining much deeper water table depth Split fertilize r applications should be maintained by the farmers. Data from column l eaching experiments have shown that diffusion of ammonium and phosphorus below the water table is more pronounced when water table is lowered closer to Bh horizon (50 cm depth). This observation has proven that the hypothesis reducing distance between wat er table and Bh horizon through lowering water table from 30 cm

PAGE 204

204 to 50 cm depth will increase diffusion of phosphorus and nitrogen below the water table for Immokalee soil Detecting ammonium and phosphorus in solutions collected from 30 cm and 50 cm depth s shows that atleast a portion of applied nutrients will be lost during water table management. The observation proves the hypothesis, restoring water table depth after rainfall events will lead to loss of plant available phosphorus and nitrogen out of the Results from lysimeter study have demonstrated that, moisture and nutrients ( phosphorus and nitrogen ) were retained within the root zone by drip irrigation. Moisture content and nitrogen values within the root zone (0 30 cm) were observed to be c onsistently higher than values at 30 45 cm depth Phosphorus and nitrogen masses in tissues increased with time. This showed that sugarcane plants responded to applied phosphorus and nitrogen Increase in biomass with time also showed that, plant growth was enhanced by drip irrigation coupled by applied phosphorus and nitrogen To minimize loss of nutrients below the root zone and pumping activities, farmers should consider drip irrigation as an alternative for water management. Researc hers should know that a portion of bromide is taken up by sugarcane plants before it is used as a tracer. Throughout the whole lysimeter study, higher phosphorus and nitrogen concentrations were observed within 30 cm depth (root zone) than below the root z one (30 45 cm). This observation proves the hypotheses d rip irrigation can be used to maintain high plant available nutrients within the root zone and minimize nutrients loss out of the root zone Increases in nitrogen, phosphorus, and biomass with time f or leaves and stalks has proven the hypothesis, s ugarcane plants will efficiently respond to applied fertilizers and moisture applied through drip irrigation.

PAGE 205

205 Hydrus 1D can be used to model one dimensional movement of phosphorus nitrogen and bromide in an environment that involves a fixed water table depth and drip irrigation as a water management technique for sugarcane The saturated hydraulic conductivity ( Ksat ) saturated water co ntent, residual water content, linearized sorption coefficient values, and kinetics parameters (K 2 and F) should be experimentally determined. The soil materials used for column leaching experiments should be used to determine the model parameters. Hydrus 1D was validated using the nutrient data collected from lysimeter study and column leaching experiment that involved fluctuating water table Hydrus 1D predicted well nutrient movement within the root zone. Modelers can therefore model phosphorus and nitrogen m ovement however the appropriate linearized sorption coefficient va lue for phosphorus should be determined using potassium chloride (0.01 M KCl). Farmers should generate databases of constants like saturated hydraulic conductivity (Ksat) and sorption coefficients using soil samples f rom different soil orders. Reliable con stants can be used to model water and nutrient Modeling phosphorus movement with linearized sorption coefficient (Florida rain and deionized water) and linearized sorption coefficient (0.005 M CaCl 2 ) over predicts and under predicts leaching respectively. The observation has proven the hypothesis, phosphorus leaching will be over predicted and or under predicted when sorption coefficients determined using different electrolytes are used to model phosphorus movement.

PAGE 206

206 A PPENDIX A H ISTOGRAMS FOR SPATIA L VARIABILITY DATA Figure A 1. Histogram for A horizon depth from Immokalee s oil Figure A 2 Histogram for total carbon (0 30cm) from Immokalee s oil

PAGE 207

207 Figure A 3. Histog ram for total car bon (30 60cm) from Immokalee s oil Figure A 4. Histogram for total carbon (60 90cm) from Immokalee s oil

PAGE 208

208 Figure A 5. Histogram for pH (0 30cm) from Immokalee s oil Figure A 6. Histo gram for pH (30 60cm) from Immokalee s oil

PAGE 209

209 Figure A 7. Histogram for pH (60 90cm) from Immokalee s oil Figure A 8 Histogram for o x alate a luminum ( 0 30 cm) from Immokalee s oil

PAGE 210

210 Figure A 9 Histogram for oxalate alu minum (30 60 cm) from Immokalee s oil Figure A 10 Histogram for oxalate aluminum (60 90 cm) from Immokalee s oil

PAGE 211

211 Figure A 11. Histogram for o xalate i ron (0 30 cm) from Immokalee s oil Figure A 1 2 Histogram for oxalate i ron ( 30 60 cm) from Immokalee s oil

PAGE 212

212 Figure A 13. Histogram for oxalate iron (60 90 cm) from Immokalee s oil Figure A 14. Histogram for o xalate p hosphorus (0 30 cm) from Immokalee s oil

PAGE 213

213 Figure A 15. Histogram for phosphorus saturation ratio (0 30 cm) from Immok alee s oil Figure A 16 Histogram for total phosphorus (0 30 cm) from Immokalee s oil

PAGE 214

214 Figure A 17 Histogram for t otal p hosph orus (30 60 cm) from Immokalee s oil Figure A 18 Histogram for t otal p hosphorus (60 90 cm) from Immokalee s oil

PAGE 215

215 Figure A 19 Histogram for exchangeable calcium ( 0 30 cm) from Immokalee soil Figure A 20 Histogram for exchangeable calcium ( 30 60 cm) from Immokalee soil

PAGE 216

216 Figure A 21 His togram for exchangeable calcium ( 60 90 cm) from Immokalee soil Figure A 22 Histogram for A horizon depth from Margate s oil

PAGE 217

217 Figure A 2 3 Histogram for total carbon (0 30 cm) from Mar gate s oil Figure A 2 4 Histogram for t otal carbon ( 30 60 cm) from Margate s oil

PAGE 218

218 Figure A 2 5 Histogram for t otal carbon (60 90 cm) from Margate s oi l Figure A 2 6 Histogram for pH (0 30 cm) from Margate s oil

PAGE 219

219 Figure A 2 7 Histogram for pH (30 60 cm) from Margate s oil Figure A 2 8 Histogram for pH (60 90 cm) from Margate s oil

PAGE 220

220 Figure A 2 9 Histogram for o xalate a luminum (0 30 cm) for Margate s oil Figure A 30 Histogram for o xalat e a luminum (30 60 cm) from Margate s oil

PAGE 221

221 Figure A 31 Histogram for o xalate a luminum (60 90 cm) from Margate s oil Figure A 32 Histogram for o xalate i ron (0 30 cm) from Margate s oil

PAGE 222

222 Figure A 3 3. Histogram for o xalate i ron (30 60cm) from Margate s oil Figure A 3 4 Histogram for o xalate i ron (60 90cm) from Margate s oil

PAGE 223

223 Figure A 3 5 Histogram for o xalate p hosphorus (0 30cm ) from Margate s oil Figure A 3 6 Histogram for o xalate p hosphorus (60 90 cm) from Margate s oil

PAGE 224

224 Figure A 3 7 Histogram for phosph orus saturation ratio ( 0 30 cm) from Margate s oil Figure A 3 8 Histogram for phosphorus saturation ratio (60 90 cm) from Margate s oil

PAGE 225

225 Figure A 3 9 Histogram for total phosphorus (0 30 cm) from Margate s oil Figure A 40 Histogram for total phosphorus (30 60 cm ) from Margate s oil

PAGE 226

226 Figure A 41 Histogram for total phosphorus (60 90 cm) from Margate s oil Figure A 42 Histogram for exchangeable calcium (0 30 cm) from Margate soil

PAGE 227

2 27 Figure A 43 Histogram for exchangeable calcium ( 30 60 cm) from Margate soil Figure A 44. Histogram for exchangeable calcium (60 90 cm) from Margate soil

PAGE 228

228 APPENDIX B GRAPHS FROM SORPTION EXPERIMENTS Figure B 1 P ho sphorus s orption i sotherm in A and B h h orizons of Immokalee s oil u sing calcium chloride ( 0.005M CaCl 2 ) Figure B 2 P hosphorus s orption i sotherm in A and B h h orizons of Immokalee s oil u sing d eionized w ater

PAGE 229

229 Figure B 3 P hosphorus s orption i sotherm in A and Bh h orizons of Immokalee s oil u sing Florida r ain Figure B 4 P hosphorus s orption i sotherm in A and Bw h orizons of Margate s oil u sing calcium chloride ( 0.005 M CaCl 2 )

PAGE 230

230 Figure B 5 P hosphorus s orption i sotherm in A and Bw h orizons of Margate s oil u si ng d eionized water Figure B 6 Phosphorus s orption i sotherm in A and Bw h orizons of Margate s oil u sing Florida rain

PAGE 231

231 Figure B 7 Relative c oncentration (C/C0) as a f unction of t ime with potassium chloride ( 0.01M KCl ) Figure B 8 Relative c o ncentratio n (C/C0) as a f unction of t ime with potassium chloride ( 0.01M KCl )

PAGE 232

232 Figure B 9 Relative c oncentration (C/C0) as a f unc tion of time with f ertilizer mixture

PAGE 233

233 APPENDIX C PROCEDURE FOR CALIBRATION AND VALI DATION OF HYDRUS 1D Figure C 1 Main p rocesses F igure C 2 Geometry i nformation

PAGE 234

234 Figure C 3 Time i nformation Figure C 4 Soil h ydraulic m odel

PAGE 235

235 Figure C 5 Water f low p arameters Figure C 6 Water f low b oundary c onditions

PAGE 236

236 Figure C 7 Non equilibrium s olute t ransport m odels Figure C 8 Solute t ransport p arameters

PAGE 237

237 Figure C 9 Solute t ransport and r eaction p aramete r s Figure C 10 Solute t ransport b oundary c onditions

PAGE 238

238 Figure C 11 Root w ater and u ptake m odel Figure C 12 Root w ater u ptake m odel p arameters

PAGE 239

239 Figure C 13 Root g rowth p arameter l ogistic g rowth f unction Figure C 14 Time v ariable b oundary conditions

PAGE 240

240 Figure C 15 Soil p rofile g raphical e ditor

PAGE 241

241 R EFERENCES Adarkani, M.S Rehbock, J.T and Mclaren A.D.1974. Oxidation of ammonium and nitrate in a soil column. Soil Sci.Soc.Amer.P roc.38:96 99. Ahmed M.F Kennedy, I.R, Choudhury, A .T.M.A, Kecskes, M.L, and Deaker, R. 2008. Phosphorus adsorption in some Australian soils and influence of Bacteria on the desorption of phosphorus. Communications in Soil Science and Plant Analysis.39:1269 1 294 Ajdary, K Singh, D.K, Singh, A.K, Khanna M.2007 Modelling of nitrogen leaching from experimental onion field under drip fertigation Agricultural Water Management.89:15 28. Anderson, D.L Portier, K.M, Obreza, T.A, Collins, M.E, and Pitts, D.J. 1999. Tree regression analysis to determine effects of soil variability on sugarcane yields. Soil Sci.Soc.Am.J. 63:592 600. Anderson, D.L. 1990. A Review: Soils, nutrition and fertility practices of the Florida sugarcane industry. Soil Crop Sci. Soc. Fla. 49:78 87. Antelo, J Avena M Fiol S, Lopez R, and Arce F 2005. Effects of pH and ionic strength on the adsorption of phosphate and arsenate at the goethite water interface. Journal of Colloid and Interface Science 285: 476 486. Assouline S.2002 The effects of microdrip and conventional drip irrigation on water distribution and uptake.Soil Sci.Soc.Am.J.66:1630 1636. Assouline S Cohen S Meerbach D Harodi T and Rosner M 2002. Microdrip irrigation of field crops: Effect on yield, water uptake, and drainage i n sweet corn. Soil Sci.Soc.Am.J.66:228 235. Assouline, S Moller, M, Cohen, S, Ben Hur, M, Grava, A, Narkis, K, Silber, A. 2006. Soil plant system response to pulsed drip irrigation and salinity: Bell pepper case study. Soil Sci.Soc.Am.J.70:1556 1568. Axt, J.R an d Walbridge, M.R. 1999. Phosphate removal capacity of palustrine forested wetlands and adjacent uplands in Virginia. Soil Sci.Soc.Am. J. 63: 1019 1031. Barrow, N.J Bowden, J.W, Posner, A.M, Quarker, J.P. 1980. Describing the effects of electrolyte on adsorp tion of phosphate by a variable charge surface. Aust.J.Soil Res.18:395 404. Bates T.E and Tisdale S.L.1957 The movement of nitrate nitrogen through columns of coarse textured soil materials. Soil Science Society Proceedings. Page: 525 528. Ben Gal, A and Dudley, L.M. 2003. Phosphorus availability under contin u ous point source irrigation. Soil Sci.Soc.Am.J.67:1449 1456.

PAGE 242

242 Beigel, C and Di Pietro, L. 1999.Transport of triticonazole in homogeneous soil columns: Influence of non equilibrium sorption.63:1077 1086. Bo tros, F.E Harter, T, Onsoy Y.S Tuli A Hopmans, J.W. 2009. Spatial variability of hydraulic properties and sediment characteristics in a deep alluvial unsaturated zone. Vadoze zone.8:276 289. Bouldin, D.R and Black C.A.1954 Phosphorus diffusion in soils. Soil Science Society Proceedings Page: 255 259. Bowman, R.A.1988 A rapid method to determine total P in soils. Soil Sci. Soc. AM. J. 52: 1301 1304. Brusseau, M.L and Rao P.S.C.1990 Modeling solute transport in structured soils: A review.Geoderma.46:169 19 2. Brye, K.R Slaton, N.A, Savin, M.C, Norman, R.J, and Miller, D.M. 2003. Short term effects of land leveling on soil physical properties and microbial biomas s Soil Sci.Soc.Am.J. 67:1405 1417. Brye K.R. 2006. Soil biochemical properties as affected by land lev eling in a clayey Aquert. Soil Sci.Soc.Am.J.70:1129 1139. Butters, G.L Benjamin, J.G, Ahuja, L.R, Ran, H. 2000. Bromide and atrazine leaching in furrow and sprinkler irrigated corn. Soil Sci.Soc.Am.J.64:1723 1732. Cameron D.R and Klute A.1977 Convective dis persive solute transport with a combined equilibrium and kinetic adsorption model. Water resources research.13:183 188. Camp, C.R.1998. Subsurface drip irrigation: A review. American Society of Agricultural Engineers. 45:1353 1367. Casel, D.K. 1983. Spatial a nd temporal variability of soil physical properties following tillage of Norfolk loamy sand. Soil Sci. Soc.Am.J. 47:196 201. Chakraborty, D Nair V.D, Chrysotome, M and Harris W.G. 2011.Soil phosphorus storage capacity in manure impacted Alaquods : Implica tions for water table management. Agriculture Ecosystems and Environment.142:167 175. Corey, J .C Nielsen, D.R and Kirkham, D. 1967.Miscible displacement of nitrate through soil columns.Soil Sci.Soc.Am.Proc.31:497 501. Crop Production (2010 Summary), Nationa l Agricultural Statistics Service (NASS), USDA, 2011. Crop Production (2009 Summary), National Agricultural Statistics Service (NASS), USDA, 20 09

PAGE 243

243 Curtin D Syers J.K and Bolan N.S. 1992 Phosphate sorption by soil in relation to cation exchangeable cati on composition and pH. Aust.J.Soil Res.31:137 149 Daniel T.C Sharpley, A.N., and Lemunyon, J.L. 1998. Agricultural phosphorus and eutrophication: A symposium overview. J. Environ. Qual. 27: 251 257. Daniels, M.B Delaune, P, Moore, P .A, Mauromoustakos, A, Ch apman, S.L, and Langston, J.M. 2001. Soil phosphorus variability in pastures: Implications for sampling and environmental management strategies.J.Environ.Qual.30:2157 2165. Darusmnan Khan, A.H, Stone, L.R, Lamm, F.R. 1997.Water flux below the root zone vs. dr ip line spacing in drip irrigated zone. Soil Sci.Soc.Am.J.61:1755 1760. Davidson, E.A. 1995. Spatial covariation of soil organic carbon, clay content and drainage class at a regional scale. Landscape ecology.10 (6):349 362. Devine, C.E and McDonnell, 2005. T he future of applied tracers in hydrogeology.Hydrogeol.J.13:255 258. Djodjic, F Borling, K, and Bergstrom, L. 2004. Phosphorus leaching in relation to soil type and soil phosphorus content. J. Environ.Qual.33:678 684. Dou, Z Ramberg, C.F, Toth, J.D, Wang Y, Sharpley, A.N, Bod, S.E, Chen, C.R, Williams, D, Xu, Z.H.2009. Phosphorus speciation and sorption desorption characteristics in heavily manured soils. Soil Sci.Soc.Am.J. 73:93 101. Drizo, A Comeau Y, Forget, C, and Chapuis, R.P. 2002. Phosphorus saturation poten tial: A parameter for estimating the longevity of constructed wetland systems. Environ. Sci. Technol. 36: 4642 4648. Elmi, A.A Madramootoo, C, Egeh, M, Liu, A, and Hamel C. 2002 Environmental and agronomic implications of water table and nitrogen fertiliza tion management.J.Environ.Qual.31:1858 1867. Evangelou V.P and Lumbanraja, J. 2002. Ammonium Potassium Calcium Exchange on Vermiculite and Hydroxy aluminum Vermiculite.Soil Sci.Soc.Am.J.66:445 455 Everts C.J Kanwar, R.S, Al exander E.C, and Alexander, S.C. 1989. Comparison of tracer mobilities under laboratory and field conditions. J.Environ.Qual.18:491 498. Fisher, L.H and Healy, R.W. 2008. Water movement within the unsaturated zone in four agricultural areas of the United States.J.Environ.Qual.37 :1051 1063. Fiskell, J.G.A Mansell, R.S, Selim, H.M, and Martin, F.G. 1979. Kinetic behavior of phosphate sorption by acid, sandy soil.8:579 584.

PAGE 244

244 Fox, T.R Comerford, N.B, and Mcfee, W.W. 1990. Kinetics of phosphorus release from Spodosols: Effects of oxalate and fo rmate. Soil Sci.Soc.Am.J.54:1441 1446. Fortin, J Gagnon Batrand, E, Vezina, L and Rompre, M. 2002. Preferential bromide and pesticide movement to tile drains under different cropping practices.J.Environ.Qual.31:1940 1952. F rossard, E Brossard, M, Hedley, M.J and Metherell, A. 1995. Reactions controlling the cycling of P in soils.p. 107 138. In H. Tiessen (ed.) Phosphorus in the global environment: Transfers, cycles and management. John Wiley and Sons, New York. Ghuman, B.S and Prihar, S.S. 1980. Chloride displa cement by water in homogeneous columns of three soils. Soil Sci.Soc.Am.J.44:17 21. Glaz, B and Morris, D.R. 2010. Sugarcane responses to water table depth and periodic flood.Agron.J.102:372 380. Griggs, B.R Norman, R.J, William, C.E, and Slaton, N.A. 2007. Am monium volatilization and nitrogen uptake for conventional and conservation tilled dry seeded, delayed flood rice. Soil Sci.Soc.Am.J.71:745 751. Guertal, E.A Eckert, D.J, Traina, S.J, and Logan, T.J. 1991. Differential phosphorus retention in soil profiles u nder no till crop production. Soil Sci. Soc. Am. J. 55: 410 413. Guo, Y Amundson, R, Gong, P, and Yu, Q. 2006. Quantity and spatial variability of soil carbon in the conterminous united states.Soil Sci.Soc.Am.J.70:590 600. Hassan, G Reneau, R.B, Hagedorn, C, and Jantrania, A.R. 2008. Modeling effluent distribution and nitrate transport through an on site wastewater system. J.Environ.Qual.37:1937 1948 He, Z.L Alva, A.K, Li, Y.C, Calvert, D.V, and Banks, D.J. 1999. Sorption desorption and solution concentration o f phosphorus in a fertilized sandy soil. J.Environ.Qual.28:1804 1810. Horneck D.A Holcomb, J, Sullivan D, and Clough G. 2011 Ammonia Volatilization in grass forages. Proceedings, 2011Western Alfalfa and Forage Symposium, Las Vegas,NV ,11 13. Hsu, P.H.1964. Adsorption of phosphate by aluminum and iron in soils. Soil Science Society Proceedings.474 478. Hyde, A.G and Ford, R.D.1989. Water table fluctuation in representative Immokalee and Zolfo soils of Florida. Soil Sci.Soc.Am.J.53:1475 1478.

PAGE 245

245 Ige, D.V Akinremi O.O, and Flaten, D.N. 2005. Environmental index for estimating the risk of phosphorus loss in calcareous soils of Manitoba. 34:1944 1951. Iqbal, J Thomasson, J.A, Jenkins, J.N, Owens, P.R, Whisler, F.D. 2005. Spatial variability analysis of soil physical prop erties of alluvial soils. Soil Sci.Soc.Am.J.69:1338 1350. Izuno, F.T Sanchez, C.A, Coale, F.J, Bottcher, A.B, and Jones, D.B. 1991. Phosphorus concentrations in drainage water in the Everglades agricultural area.J.Environ.Qual.20:608 619. Jaynes, D.B and Rice R.C. 1993. Transport of solutes as affected by irrigation method. Soil Sci.Soc.Am.J.57:1348 1352. Jellali, S Diamantopoulos, E, Kallali, S, Bennaceur, S, Anane, M, and Jedidi, N. 2010. Dynamic sorption of ammonium by sandy soil in fixed bed columns: Evaluati on of equilibrium and non equilibrium transport processes. Journal of Environmental Management.91:897 905. Jemison Jr, J.M and Fox, R.H.1991. Corn uptake of bromide under green house and field conditions. Communications in Soil Science and Plant Analysis.2 2 (3and 4):283 297. Johnson, R.M and Richard Jr E.P. 2005.Sugarcane yield sugarcane quality and soil variability in Louisiana.Agron.J.97:760 771. Kandelous, M.M and Simunek, J. 2010. Comparison of numerical, analytical and empirical models to estimate we tting patterns for surface and subsurface drip irrigation. Irrig.Sci.28:435 444. Kelly, W.R and Wilson, S.D.2000.Movement of bromide, nitrogen 15 and atrazine through flooded soils.J.Environ.Qual.29:1085 1094. Khalil, A. 2008. Simulation of nitrogen distri bution in soil with drip irrigation system. Journal of Applied Sciences, 8:3157 3165. Kliewer, B.A and Gilliam, J.W. 1995. Water table management effects on denitrification and nitrous oxide evolution. Soil Sci.Soc.Am.J.59:1694 1701. Kookana, R.S Schuller, R .D, and Aylmore, L.A.G. 1993. Simulation of simazine transport through soil columns using time dependent sorption data measured under flow conditions. Journal of Contaminant Hydrology.14:93 115. Kung, K J,S.1990. Influence of plant uptake on the performance of bromide tracer. Soil Sci.Soc.Am.J.54:975 979. Kuo, S and Lotse, E.G. 1974. Kinetics of phosphate adsorption and desorption by lake sediments.Soil.Sci.Soc.Amer.Proc.38:50 54.

PAGE 246

246 Kwon, H Grunwald, S, B eck, H.W, Jung, Y D aroub, S.H, Lang, T.A, Morgan, K.T. 2010. M odeling of phosphorus loads in sugarcane in a low relief landscape using ontology based simulation.J.Environ.Qual.39:1 11. Lawrence, H.F and Richard, W.H. 2008.Water movement within the unsaturated zone in four agricultural areas of the United States.J.Envir on.Qual.37:1051 1063. Lee, L.S Rao, P.S.C, Brusseau, M.L, and Ogwada, R.A. 1988. Non equilibrium sorption of organic contaminants during flow through columns of aquifer materials. Environmental Toxicology and Chemistry.7:779 793. Li, Y and Ghodrati, M. 1994. Pr eferential transport of nitrate through soil columns containing root channels. Soil Sci.Soc.Am.J.58:653 659. Li, Y. C, Alva, A.K, and Calvert, D.V 1997. Transport of phosphorus and fractionation of residual phosphorus in various horizons of a Spodosol. Water Air Soil Pollu. 109:303 312. Lock aby, B.G., and Walbridge, M.R. 1998. Biogeochemistry.p.149 172.In M.G. Messina and W.H.Conner (ed.) Southern forested wetlands: Ecology and management. Lewis Publ., Boca Raton, FL. Logan, T.J and Mclean, E.O. 1973 Effects of phosphorus application rate, soil properties, and leaching mode on 32P movement in soil columns. Soil Sci.Soc.Am.Proc.37:371 374. Mallarion, A.P, and Borges, R. 2006. Phosphorus and potassium distribution in soil following long term deep band fertilization i n different tillage systems. Soil Sci.Soc.Am.J.70:702 707. Mann, K.K Schumann, A.W, Obreza, T.A, Tepliski, M, Harris, W.G and Sartain, J.B. 2011. Spatial variability of soil chemical and biological properties in Florida citrus production.Soil Sci.Soc.Am.J.75:1 863 1873. Mansell, R.S Bloom, S.A, and Burgoa, B. 1991. Phosphorus transport with water flow in acid, sandy soils; 271 314.In Jacob and M.Y. Corapcioglu (ed.) Transport processes in porous media. Kluwer Academic Publ., Dorecht, the Netherlands. Mansell, R .S Sel im, H.M, Calvert, D.V, Stewart, E.H, Allen, L.H, Graetz, D .A, Fiskell, J.G.A, and Rogers, J.S. 1980. Soil Sci. Soc. Am. J. 44:95 102. Martin, H.W Ivanoff,D.B, Graetz,D.A, and Reddy, K.R. 1997. Water table effects on Histosol drainage water carbon, nitr ogen, and phosphorus. J.Environ.Qual. 26:1062 1071. McKeague, J.A and Day, J.H. 1966. Dithionite and oxalate extractable Fe and Al as aids in differentiating various classes of soils. Can.J.Soil Sci. 46: 13 22.

PAGE 247

247 Mmolawa K and Or, D. 2003. Experimental and n umerical evaluation of analytical volume balance model for soil water dynamics under drip irrigation. Soil Sci.Soc.Am.J. 67:1657 1671. Muchovej, R.M, and Newman, P.R. 2004. Nitrogen fertilization of sugarcane on a sandy soil: II. Soil and ground water analy ses. Journal American Society Sugarcane Technologists. 24:225 240. Nachabe, M Masek, C, and Obeysekera, J. 2004. Observations and modeling of profile soil water storage above a shallow water table. Soil Sci.Soc.Am.J.68:719 724. Nair, V.D and Harris, W.G. 200 4.A capacity factor as an alternative to soil test phosphorus in phosphorus risk assessment.NewZealand Journal of Ag ricultural Research.47:491 497. Nair, V.D Villapando, R.R and Graetz, D.A. 1999. P retention capacity of the spodic horizon under varyin g environmental conditions. J.Environ.Qual. 28:1308 1313. Nair, V.D Graetz, D.A, and Reddy, K.R. 1998. Dairy manure influences on phosphorus retention capacity of spodosols. J.Environ.Qual. 27:522 527. Nair,P.S Logan,T.J, Sharpley,A.N, Sommers,L.E, Tabatab ai, M.A, and Yuan, T.L. 1984.Interlaboratory comparison of a standardized phosphorus adsorption proced ure. J.Environ.Qual.13:591 595. Namasivayam, C and Sangeetha, D. 2004. Equilibrium and kinetic studies of adsorption of phosphate on ZnCl 2 activated coir pi th carbon. Journal of colloid and interface science.280: 359 365. Nkedi Kizza, P Shinde, D, Savabi, M.R, Ouyang, Y and Nieves ,L. 2006. Sorption kinetics and equilibria of organic pesticides in carbonatic soils from South Florida. J. Environ.Qual. 35:268 276 Obreza, T.A Anderson, D.L, and Pitts, D.J. 1998. Water and nitrogen management of sugarcane growth on sandy, high water table soil. Soil Sci.Soc.Am.J. 62:992 999. Or, D.1996.Drip irrigation in heterogeneous soils: Steady state field experiments for stochas tic model evaluation. Soil Sci.Soc.Am.J.60:1339 1349. Pant H.K Nair, V.D, Reddy, K.R, Graetz, D.A, and Villapando, R.R. 2002. Influence of flooding on phosphorus mobility in manure impacted soil. J.Environ.Qual.31:1399 1405. Pardo, M.T Guadalix, M.E, and G arcia Gonzalez, M.T. 1991. Effect of pH and background electrolyte on P sorption by variable charge soils. Geordama 54: 275 284.

PAGE 248

248 Parry, R. 1998. Agricultural phosphorus and water quality: A.U.S environmental protection agency perspective. J.Environ.Qual.27: 258 261. Penn, C.J Mullins, G.L, and Zelazny, L.W. 2005. Mineralogy in relation to phosphorus sorption and dissolved phosphorus losses in runoff. Soil Sc i .Soc.Am.J. 69:1532 1540. Pitts, D.J Tsai Y.J, Myhre, D.L, Anderson, D.L, Shih, S.F. 1993. Influence of w ater table depth on sugarcane grown in sandy soils in Florida.ASAE. 36(3):777 782. Poulsen, T.G Moldrup, P, de Jonge, L.W, and Komatsu, T. 2006.Colloid and bromide transport in undisturbed soil columns: Application of two regional model.Vadose Zone Journal.5 :649 656. Qafoku, N.P Sumner, M.E, and Radcliffe, D.E.2000 Anion transport in columns of variable charge soils: nitrate and chloride.J.Environ.Qual.29: 484 493. Reddy, K.R OConnor, G.A, and Gale P.M .1998. Phosphorus sorption capacities of wetlands soils and stream sediments impacted by diary effluent. J.Environ.Qual. 27:438 447. Reddy, K.R Patrick, W.H, and Philips, R.E. 1976. Ammonium diffusion as a factor in nitrogen loss from flooded soils. Soil Sci. Soc.AM.J.40:528 533. Renduo, Z. 2000. Generalized tra nsfer function model for solute transport in heterogeneous soils. Soil Sci.Soc.Am.J.64:1595 1602. Rhue, R.D Harris, W.G, and Nair, V.D. 2006. A retardation based model for phosphorus transport in sandy soil. Soil Science.171:293 304. Richardson, C.J.1985. M echanisms controlling phosphorus retention capacity in fresh water wetlands. Science (Washington, D.C) 228: 142 1427. Richards, L.A.1931. Capillary conduction of liquids through porous mediums. Physics1:318 333. Roel, A and Plant, R.E. 2004. Factors underlyi ng yield variability in two California rice fields. Agron. J.96:1481 1494. Roberts, T Lazarovitch, N, Warrick, A.W, and Thompson, T.L. 2008. Modeling salt accumulation with subsurface drip irrigation using Hydrus 2D. Rubio, G Cabello, M.J, Boem, F.H.G, and Munaro, E. 2008. Estimating available soil phosphorus increases after phosphorus additions in Mollisols. Soil Sci.Soc.Am.J. 72:1721 1727.

PAGE 249

249 Shinde, D Savabi, M.R, Nkedi Kizza, P, and Vazquez, A.2001.Modeling transport of atrazine through calcareous soils from South Florida. American Society of Agricultural Engineers.44 ( 2):251 258. Shokri, N and Salvucci, G.D. 2011. Evaporation of porous media in the presence of a water table. Vadose Zone J. 10:1309 1318. Simunek, J and Hopmans, J.W. 2009.Modeling compensated root water and nutrient uptake. Ecological Modeling.220:505 521. Simunek, J and van Genucten, M,Th. 2008.Modeling nonequilibrium flow and transport processes using Hydrus. Vadose Zone.J.7:782 797. Simunek, J Sejna, J, and vanGenuchten, Th.1999. The HYDRUS 2D sof tware package for simulating the two and one dimensional movement of water, heat, and multiple solutes in variably saturated media. U.S. Salinity Laboratory Agricultural Research Service.USDA, Riveerside California. Skaggs, T.H Trout, T.J, and Rothfuss, Y. 2010. Drip irrigation water distribution patterns: Effects of emmiter rate, pulsing, and antecedent water. Soil Sci.Soc.Am.J.74:1886 1896. Smith, D.M Inma n Bamber, N.G, and Thorburn, P.J. 2005. Growth and function of the sugarcane root system. Field Crops Research.92:169 183. Sparks, D.L. 1996. Soil Science Society of America Series (SSSA) 5: Methods of soil analysis. Soil Survey Staff.1996. Keys to Soil Taxonomy. U.S.Gov.Print.Office,Washington,DC. Stutter, M.I Lumsdon, D.G, Billet, M.F, Low, D, Deeks, L.K. 2 008. Spatial variability in properties affecting organic horizon carbon storage in upland soils. Soil Sci.Soc.Am.J. 73:1724 1732. Sumner, M.E Bellini, G, Radcliffe, D.E, and Qafoku, N.P. 1996. Anion transport through columns of highly weathered acid soil: ad sorption and retardation.Soil Sci.Soc.Am.J.60:132 137. Thompson, T.L White, S.A, Walworth, J, and Sower, G.J. 2003. Fertigation frequency for subsurface drip irrigated Broccoli Thompson, T.L Doerge, T.A, and Godin, R.E 2000. Nitrogen and water interactions i n subsurface drip irrigated cauliflower: I. Plant response. Soil Sci.Soc.Am.J. 64:406 411. Thompson, T.L Doerge, T.A, and Godin, R.E. 2002. Subsurface drip irrigation and fertigation of broccoli: I. Yield, quality, and nitrogen uptake. Soil Sci.Soc.Am.J.66:18 6 192.

PAGE 250

250 Travis, C.C and Etnier, E.L. 1981. A survey of sorption relationships for reactive solutes in soil.J.Environ.Qual.10:8 17. Tyler, D.D and Thomas, G.W.1981. Chloride movement in undisturbed soil columns. Soil Sci.Soc.Am.J.45:459 461. US Department of A griculture. Natural Resources Conservation Service, 1995. Soil survey laboratory manual. Van Genuchten M.Th.1980 A closed form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci.Soc.Am.J.44: 892 898. Villapando R.R, and Grae tz, D.A. 2001. Phosphorus and desorption properties of the spodic horizon from selected Florida Spodosols.Soil Sci.Soc. Am.J. 65:331 339. Villapando R.R. 1997. Reactivity of inorganic phosphorus in the spodic horizon. PhD dissertation. University of Florid a, Gainesville. Vogeler, I Cichota, R, Snow, V.O, Dutton, T, and Daly, B. 2011.Pedotransfer functions for estimating ammonium adsorption in soils. Soil Sci.Soc.Am.J.75:324 331. Vruget, J.A Hopmans, J.W, and Simunek J. 2001.Calibration of a two dimensional roo t water uptake model. Soil Sci.Soc.Am.J.65:1027 1037. Walker,T.W Kingery,W.L,Street,J.E, Cox,M.S,Oldham,J.L,Gerard,P.D, and Han,F.X.2003. Rice yield and soil chemical properties as affected by precision land leveling in alluvial soils. Agron.J.95:1483 14 88. Wang, F.L and Alva, A.K. 2000. Ammonium adsorption and desorption in sandy soils. Soil Sci.Soc.Am.J.64:1669 1674. Weaver D.M., Ritchie, G.S.P., Anderson, G.C, and Deeley, C.M. 1988. Phosphorus leaching in sandy soils: Short term effects of fertilizer applications and environmental conditions. Aust. J. Soil Res. 26: 177 190. Zhou, M and Li, Y. 2001 Phosphorus sorption characteristics of calcareous soils and limestone from the southern Everglades and adjacent far ml ands.Soil Sci.Soc.Am.J.65:1404 1412. Z hou M Rhue, R.D, and Harris, W.G. 1997. Phosphorus sorption characteristics of Bh and Bt horizons from sandy coastal plain soils. Soil Sci.Soc.Am.J. 61:1364 1369

PAGE 251

251 BIOGRAPHICAL SKETCH Augustine Muwamba was born in Uganda. He graduated with a Bachelor of Sc ience in agriculture from Makerere University, Kampala Uganda in 2005. He joined University of Florida, Gaine sville (Soil and Water Science D epartment) in August 2005. He graduated with 2007. He continued with a PhD program in 2008 in Soil and Water Science department at University of Florida, Gainesville. He graduate d with Doctor of Philosophy in the summer of 2012.